The Generation of Peripheral Neuron Diversity from Mammalian Progenitor Cells In Vitro

Citation

Greenwood, Amy L. (2000) The Generation of Peripheral Neuron Diversity from Mammalian Progenitor Cells In Vitro. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/6pn4-h629. https://resolver.caltech.edu/CaltechTHESIS:08062025-230309605

Abstract

The vertebrate peripheral nervous system is comprised of many different kinds of neurons that develop from a common progenitor pool, the neural crest. Factors that control the specification of most neuronal subtypes are not well understood. The studies presented in this thesis describe how and when sensory precursors become different from those of the autonomic lineage, and how subtype diversity is generated within the sensory neuron class.

The development of peripheral neurons was assayed in vitro using rat neural tube explant cultures as a source of neural crest, and antibody staining for lineage-specific transcription factors to identify neuronal subtypes. A chemically defined medium supported the differentiation of sensory but not autonomic neurons in these explants. When cultures were challenged with the extrinsic factor BMP2 to induce autonomic neurons, the division and differentiation of sensory precursors was unperturbed. Therefore, the neural crest contains a population of di vi ding precursors that are operationally determined to the sensory neuron fate.

The majority of sensory neurons that differentiated in defined medium displayed characteristics of the muscle afferent subtype including dependence on the neurotrophins BDNF and NT-3, expression of the marker ER81, and lack of expression of trkA. In contrast, the addition of serum or BMP2 induced the development of many neurons that expressed the cutaneous afferent marker, trkA. Neither serum nor BMP2 is sufficient to induce trkA in post-migratory, differentiated neurons; instead, these factors induce the neural tube to produce additional sensory precursors, some of which differentiate into cutaneous afferents. We also observed that the neurotrophins BDNF and NT-3, which could directly regulate ER81 expression in differentiated neurons, did not prevent the development of trkA expressing neurons. This implies that serum/BMP2 have early affects on the specification of sensory subtypes while neurotrophins regulate acquisition of particular subtype characteristics.

In addition to their ability to promote the development of the cutaneous afferent sensory subtype, serum!BMP2 induced patterning in neural tube explants. In the presence of these factors, the location of sensory and autonomic neurons in the neural crest outgrowth of the explant resembled the arrangement of these neuronal subtypes along the dorsoventral axis in vivo.

Item Type:	Thesis (Dissertation (Ph.D.))
Subject Keywords:	(Biology)
Degree Grantor:	California Institute of Technology
Division:	Biology
Major Option:	Biology
Thesis Availability:	Public (worldwide access)
Research Advisor(s):	Anderson, David J.
Thesis Committee:	Anderson, David J. (chair) Patterson, Paul H. Wold, Barbara J. Zinn, Kai George Fraser, Scott E.
Defense Date:	1 April 2000
Record Number:	CaltechTHESIS:08062025-230309605
Persistent URL:	https://resolver.caltech.edu/CaltechTHESIS:08062025-230309605
DOI:	10.7907/6pn4-h629
Default Usage Policy:	No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:	17602
Collection:	CaltechTHESIS
Deposited By:	Benjamin Perez
Deposited On:	08 Aug 2025 19:28
Last Modified:	08 Aug 2025 19:28

Thesis Files

PDF - Final Version
See Usage Policy.
53MB

Repository Staff Only: item control page

THE GENERATION OF PERIPHERAL NEURON DIVERSITY FROM MAMMALIAN PROGENITOR CELLS IN VITRO Thesis by Amy L. Greenwood In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 2000 (submitted April, 2000) II © 2000 Amy L. Greenwood All rights reserved 111 To my parents, Gary and Janice Greenwood, and my sisters, Erica and Amanda IV Acknowledgements I have had many teachers. First, I am indebted to my advisor, David Anderson, for the countless hours of shared discussion, note scribbling, argument, data analysis, and whiteboard drawing that introduced me to experimental strategy. Few people are as intellectually engaged and as fearless, and few advisors give their students as much attention; I have benefited enormously by having had the General with me in the trenches (or at the microscope, as it were). I admire his enthusiasm, creativity, and scientific aesthetic, and I am grateful for the time he spent to share those things with me. In addition to his direct efforts, David has provided for my training by running a lab that draws talented post-docs, students, and staff from whom I have had the honor to receive both wisdom and friendship. I have enjoyed working with all Anderson lab members, past and present, but a few deserve special thanks. In particular, during his time as a post-doc in the lab, Andy Groves was a consistent source of practical, solid advice that helped me through the troublesome period as I was toddling toward experimental independence. I am extremely grateful for his confidence in me. Similarly, post-docs Mohendra Rao, my first teacher in the lab, and Joe Verdi were supportive of my early efforts. Later post-docs Quifu Ma, Sean Morrison, and Sherry Perez contributed excellent advice that shaped the course of my project. Due to the similarity of our natures, having Sherry in the lab was a blessing. Throughout the years, Li Ching Lo has happily shared her wisdom with me. I very much admire her integrity and skill. Friends in the lab have provided emotional support and motivation; moreover, they demonstrate how many ways there are to do successful science. Pat White, my benchmate for many years, is the most tolerant person I know and has a heart of gold. I am grateful for her optimism. I thank Alice Paquette for her high level of interest, V loyalty, and sense of fairness. Nirao Shah, Sebastian Gerety, Emma Dormand, Jae Kim, Mariela Zirlinger, and Gina Mancuso have contributed much in the way of humor and camaraderie. I thank Gaby Mosconi for creating order and for going out of her way for me and for the lab. I have also had many teachers outside the lab. My committee members Paul Patterson, Barbara Wold, Kai Zinn, and especially Scott Fraser have been individually and collectively helpful with my project. Scott has been particularly supportive and has provided a crucial perspective that gives balance to my work. Also, his post-docs Rusty Lansford, Mary Dickinson, and Dave Crotty have kindly contributed to my social life. Roian Egnor has been my true friend from the start. She is a model of character, joie de vivre, and coolness. I am grateful to Roian, Dan Fain, Pat Chuang, and a few others for their continuing support since our common beginnings at Caltech. During this past year I have been immensely grateful to Joel Pomerantz for knowing how to make things better and for his calming influence. He is the definition of thoughtfulness. Most importantly, however, are my first teachers. My father taught me about honesty, integrity, and humor. My mother taught me about pride and unconditional love. My sister Erica taught me about motivation and following one's heart. My sister Amanda taught me about giving excellent gifts and about having a sense of family and place. VI Abstract The vertebrate peripheral nervous system is comprised of many different kinds of neurons that develop from a common progenitor pool, the neural crest. Factors that control the specification of most neuronal subtypes are not well understood. The studies presented in this thesis describe how and when sensory precursors become different from those of the autonomic lineage, and how subtype diversity is generated within the sensory neuron class. The development of peripheral neurons was assayed in vitro using rat neural tube explant cultures as a source of neural crest, and antibody staining for lineage-specific transcription factors to identify neuronal subtypes. A chemically defined medium supported the differentiation of sensory but not autonomic neurons in these ex plants. When cultures were challenged with the extrinsic factor BMP2 to induce autonomic neurons, the division and differentiation of sensory precursors was unperturbed. Therefore, the neural crest contains a population of di vi ding precursors that are operationally determined to the sensory neuron fate . The majority of sensory neurons that differentiated in defined medium displayed characteristics of the muscle afferent subtype including dependence on the neurotrophins BDNF and NT-3, expression of the marker ER81, and lack of expression of trkA . In contrast, the addition of serum or BMP2 induced the development of many neurons that expressed the cutaneous afferent marker, trkA. Neither serum nor BMP2 is sufficient to induce trkA in post-migratory, differentiated neurons; instead, these factors induce the neural tube to produce additional sensory precursors, some of which differentiate into cutaneous afferents. We also observed that the neurotrophins BDNF and NT-3 , which could directly regulate ER81 expression in differentiated neurons, did not prevent the development of trkA expressing neurons . This implies that serum/BMP2 have early affects on the specification of sensory subtypes while neurotrophins regulate acquisition of particular subtype characteristics. In addition to their ability to promote the development of the cutaneous afferent sensory subtype, serum!BMP2 induced patterning in neural tube explants. In the presence of these factors, the location of sensory and autonomic neurons in the neural crest outgrowth of the explant resembled the arrangement of these neuronal subtypes along the dorsoventral axis in vivo . Vlll Table of Contents Dedication .. .................... . .................................................... . ................. .iii Acknowledgements ............................................ . ..................................... iv Abstract. ............................ . .................................................................. vi List of Tables and Illustrations ................... .................. . .............................. viii Chapter 1 .................................... .................................................. . ........ 1 Introduction Chapter 2 ................................. ........................................................... ..38 The identification of dividing, determined sensory neuron precursors in the mammalian neural crest. Chapter 3 ..... . ............... . ............... ............................................ .. ........... 54 The regulation of subtype characteristics in sensory neurons generated from mammalian neural crest in vitro. Chapter 4 . . . ............................. . ............................................................ 88 Summary and future directions Appendix 1 . .. .... .................................................................................. ..95 The extent of sensory neuron development in explants depends on the presence of the neural tube as a source of precursors. Appendix 2 .......................................................................................... 104 Patterning of explants in vitro resembles the patterning of sensory and autonomic derivatives in vivo and is affected by BMP2 acting on the neural tube. Appendix 3 ........ . ................................................................................ .126 Identification by differential RT-PCR of a novel paired homeodomain protein specifically expressed in sensory neurons and a subset of their CNS targets. IX List of Tables and Illustrations Chapter 1 Fig. 1 Fig. 2 Chapter 2 Table 1 Fig. 1 Fig. 2 Fig. 3 Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 Fig. 9 Fig. 10 Fig. 11 Chapter 3 Fig. 1 Fig. 2 Fig. 3 Neuronal diversity in the vertebrate peripheral nervous system ........... .. .4 Sensory neuron subtypes ..... . .. ......... . ..... ... ... ... ........................... 7 Marker combinations identify peripheral sensory and autonomic neurons .. .. .... .... . .... ... ... ....... .......... .... .... ....... . . ............... . ..... 42 Two populations of neurons develop in rat neural tube ex plant cultures .. ... ............... .... .......... ... .. ................. .. ................ . ... 41 Neural tube explant cultures contain peripheral sensory and autonomic neurons . ...... ............................. ..... ............... ...... ...42 Defined medium supports sensory but not autonomic neuronal differentiation ...... .................... . ... .... ... ... . .. .. .... .. ....... .... . .... ... 43 Sensory precursors differentiate into neurons by day 2-3 in defined medium .............................................. . . ... . ...................... . .. 43 Sensory precursors divide during the first two days of culture ............ .44 BMP2 induces Phox2a but does not prevent expression of Bm-3.0 or division of sensory precursors ... ........ . .. .. .... ........ ... . ...... .45 Early sensory precursor-specific transcription factors are expressed in neural tube ex plant cultures at day 1. . .. ... ... . .... . .. ............... ..... .... 46 Sensory precursor-specific markers persist in the presence of BMP2 . ... ..47 NT-3 and BDNF, not NGF or LIF, support significant populations of ex plant-derived sensory neurons . ........ .. . . ... ... ... .. .......... ......... ..48 Explant-derived sensory neurons express proprioceptive, but not nociceptive, sensory subtype markers .... .... .. ...... .. .... ... . ...... ......... .48 Origins of sensory neurons in vitro and in vivo .............................. .49 Defined medium supports the development of sensory neurons that have characteristics of the muscle afferent subtype ..... ....... ... ... .. ...... .62 A one hour exposure to NT-3 or BDNF in the absence of the neural tube is sufficient to induce ER81 expression in sensory neurons grown in defined medium ....... . .... . ....... ...... ...... . . ...... ................ ... ...... 64 Serum-containing medium supports the development of sensory neurons that have characteristics of the cutaneous afferent subtype .. ..... 65 X Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 Fig. 9 Appendix 1 Fig. 1 Fig. 2 Appendix 2 Fig. 1 Fig. 2 Fig. 3 Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 Fig. 9 Serum induces the development of trkA + sensory neurons in ex plant cultures, but is not sufficient to induce trkA in sensory neurons that have differentiated in defined medium .. . .... ..... ..... ......... 65 BMP2 induces the development of trkA + sensory neurons in ex plant cultures grown in defined medium .... ......... ..... ..... .. .. ......... .70 Serum acts on the neural tube rather than on already-migrated neural crest to induce the development of trkA + sensory neurons . ..... .... 73 NT-3 and BDNF induce ER81 expression but do not prevent the development of trkA + sensory neurons ............ ..... .... .. .... .......... .... 75 Models in which sensory subtypes are specified before precursors migrate from the neural tube ....... ..... .. ... .. ... .... ........................... .78 Models in which sensory subtypes are specified after precursors migrate from the neural tube ............ .. .. ... .... .. .......... ...... ..... .... ... 80 The neural tube is required for the extent of sensory neuron development in ex plant cultures ... . .. . ..... .. .. . ... .................. ........ .. 98 Sensory precursors migrate from the neural tube for numerous days in explant cultures ... .. . ................... ... . ............ . ... .. .... . ... . . 100 BMP2 induces P-D patterning of sensory and autonomic neurons in the outgrowth of neural tube ex plants . ....... .. .. .. .. .. .. . ........ .. ... . .. . 106 BMP2 induces P-D patterning of sensory and autonomic neurons in ex plants even after initial neural crest migration .. . . ... ... .. . .... ........ 108 The patterning of explants in vitro resembles the patterning of sensory and autonomic ganglia in vivo .... . ... . ......... ...... .. .... ..... ... . 109 The neural tube is required for the P-D patterning of explants by BMP2 ...... ..... . .... ...... .. .. . . . ............. . .... . .. .. ...... . ... .. .. .. .. .. ... 111 The P-D position of sensory neurons in the outgrowth depends on the length of exposure to serum ...... .. .......... .. . ....... ... .. .. ....... ... .. . 114 The percentage of sensory neurons along the R-C axis is higher in the caudal part of ex plants grown in defined medium .... . ... ........... 115 Both trkA + and trkA sensory neurons develop preferentially in the caudal part of ex plants grown in serum-washout conditions ... .... .. .. . ... 117 Autonomic and sensory neurons form in different parts of neural tube explants grown in serum-washout conditions .................. .. ..... . 118 Different doses of BMP2 induce the expression of Brn-3 .0 and Phox2a in cultures of dissociated neural tube cells .. . ....................... 121 1 Chapter 1 Introduction The specification of sensory neuron fates in the vertebrate peripheral nervous system 2 Specification is the process by which cells choose developmental pathways and become different from each other. A single cell generates the vast diversity of cell types found in a mature organism through many consecutive specification events. Significant progress has been made in identifying the molecules that act as intrinsic and extrinsic cues in the specification of particular cell fates. However, such cues do not have the same effects on all progenitor cells because the potential of a cell is limited by its developmental history. Therefore, when studying the specification of a cell type it is necessary to identify the precursors that can and cannot respond in a predicted manner to particular cues. In this way, one is able determine how the timing and nature of specification events are coordinated to achieve the precise patterning of cell types in the embryo. One system in which the study of cell fate specification is particularly appropriate is in the development of the peripheral nervous system (PNS) of vertebrate embryos. The neuronal types in the PNS are highly diverse but derive largely from a common progenitor pool: the neural crest. The neural crest consists of precursors that delaminate from the neural tube and then migrate through the embryo before reaching their sites of differentiation (Horstadius, 1950; Le Douarin, 1982). A longstanding question about the development of this system is centered around the origin of heterogeneity among different types of neurons (reviewed in Le Douarin, 1982; Le Douarin, 1986; Anderson, 1989; Sieber-Blum, 1990; Marusich and Weston, 1991; Anderson, 1993; Bronner-Fraser, 1993; Le Douarin et al., 1994; Anderson, 1997; Groves and Bronner-Fraser, 1999). Specifically, how much of the heterogeneity is due to inherent early biases among the neural crest, and how much is due to the differential response of multi potent cells to environmental signals present at sites of differentiation? The goal of my thesis is to address this question in the context of the specification of sensory neurons, one particular class of cells in the PNS. The work presented in Chapters 2 and 3 describes how the specification of sensory neurons and the regulation of some characteristics of sensory 3 subtype are affected by different environmental conditions in vitro. These data emphasize the existence of heterogeneity among neural crest precursors that contribute to sensory neurons. The intention of this chapter is to put this work in context. The first section provides a brief overview of neuronal diversity in the PNS. It is followed by a discussion that reviews the evidence pertaining to how and when sensory neuron fates become specified during development. Overview of neuronal subtypes in the PNS The PNS contains two main classes of neurons (Fig. 1): sensory (somatic) and autonomic (Kandel et al., 1991). The general function of sensory neurons is to transduce stimuli from the external environment into signals that are relayed to the central nervous system (CNS). Sensory neurons are found in cranial ganglia, which serve sensory functions in the head and neck, and dorsal root ganglia (DRGs) that occur in a metameric pattern within the vertebral column adjacent to the spinal cord, and which carry somatic information from the rest of the body. Many subtypes of sensory neurons can be found within each DRG. Conversely, the three classes of autonomic neurons, sympathetic, parasympathetic, and enteric, are generally found in separate ganglia at various locations in the body. Autonomic neurons regulate homeostatic processes such as heart rate, body temperature, and digestion among others. Despite the heterogeneity within each class, sensory and autonomic neurons can be distinguished from each other based on location, function, and, more recently, gene expression. Significant effort has been put into the molecular characterization of sensory and autonomic neurons and has led to the identification of combinations of markers that can identify subsets of neurons with reasonable certainty. The ability to distinguish between 4 Neuronal Diversity in the PNS Neural Crest Sensory nociceptive .i:,1... mechanoreceptive proprioceptive 0 .i:,1... - Autonomic ~ sympathetic ~ Fig. 1 Neuronal diversity in the vertebrate peripheral nervous system (PNS). The schematic represents a cross-sectional view through an embryo (dorsal at top; ventral at bottom). There are two main classes of neurons: sensory and autonomic (right side). Different subtypes of sensory neurons, nociceptive, mechanoreceptive, and proprioceptive, are located together in dorsal root ganglia (DRG) adjacent to the neural tube (NT). In contrast, autonomic neuron subtypes, sympathetic, parasympathetic, and enteric, are generally located in separate ganglia at various sites in the ventral part of the body. All these neuronal types derive from neural crest cells that migrate from the dorsal neural tube earlier in development (left side). 5 sensory and autonomic neurons based on gene expression has been an important advance in the study of the development of these subtypes in vitro. In particular, differentially expressed transcription factors have been useful as specific markers of the classes of neurons in the PNS (reviewed in Anderson, 1997). Some important examples include members of the basic helix-loop-helix (bHLH) transcription factor family. For example, neurogenin-1 (ngnl; Ma et al., 1996), neurogenin-2 (ngn2; Gradwohl et al., 1996; Sommer et al., 1996), NeuroD (Lee et al., 1995), NSCL-1 (Begley et al., 1992) and NSCL -2 (Gobel et al., 1992), are expressed in sensory precursors and developing sensory neurons, but not in autonomic precursors or neurons. Conversely, the bHLH transcription factors MASRI (Lo et al., 1991), eHAND (Cserjesi et al., 1995) and dHAND (Srivastava et al., 1995) are expressed in autonomic but not sensory precursors. Other transcription factors are maintained in differentiated neurons and thus provide good markers of mature neurons of each subtype. In the DRG, the POU-domain transcription factor Bm-3.0 (Fedtsova and Turner, 1995; Xiang et al., 1995) and the paired homeodomain transcription factor DRG-11 (Saito et al., 1995) are expressed in most sensory neurons. The mature neurons in autonomic ganglia express the pairedhomeodomain transcription factors Phox2a (Valarche et al., 1993) and Phox2b (Pattyn et al., 1997), which are not expressed in sensory neurons of the trunk. Many of these transcription factors are required for the development of the subsets of neurons in which they are expressed (Guillemot et al., 1993; Blaugrund et al., 1996; Fode et al., 1998; Ma et al., 1998; Pattyn et al., 1999). Both the sensory and autonomic classes can be further divided into many subtypes of neurons. The heterogeneity of sensory neurons in the DRG is apparent in terms of functional, structural, and molecular differences (reviewed in Scott, 1992), and it has been a major challenge to relate these characteristics (Fig. 2). Individual sensory neurons in the DRG mediate unique sensory modalities; they transduce information about pain 6 (nociception), thermal sensation (thermoreception), touch (mechanoreception), or spatial sense (proprioception). Moreover, they may innervate skin (cutaneous afferents), muscle (muscle afferents), or certain structures in the viscera, and display a wide range of electrophysiological properties. Although numerous methods have been employed to try to categorize sensory neurons, the heterogeneity among this population is large enough that it resists simple groupings. There are, however, general classes that can be described with the caveat that there is a fair amount of heterogeneity in each. The development of neuronal subtypes in the sensory ganglia is inherently related to the question of sensory neuron specification; it is not clear whether sensory neurons are specified generally as a class and then acquire different subtype fates, or whether sensory neuron and sensory subtype fate are specified simultaneously. The subject of autonomic subtype specification is beyond the scope of this thesis, however, but has recently been addressed in White and Anderson (1999). Significant evidence has accumulated that sensory neurons with different functions have different neurotrophic requirements for survival (for reviews see Scott, 1992; Snider, 1994; Lewin, 1996). Thus, subtypes of sensory neurons can be characterized in part by their expression of high-affinity neurotrophin receptors. One class of sensory neurons expresses trkA and depends on nerve growth factor (NGF). These neurons have small to medium sized cell bodies, tend to express neuropeptides such as substance P (SP) and calcitonin gene related peptide (CORP), are excited by capsaicin via the vanillinoid receptor VRl, and typically innervate the skin to serve nociceptive or thermal functions . This class of neurons is lost in mice homozygous for the targeted deletion of trkA (Smeyne et al., 1994). Also lost in trkA-/- mice, however, is another population of nociceptive sensory neurons that do not express any trk receptors in 7 Sensory Neuron Subtypes NT nociceptive / thermoreceptive cutaneous afferents small/intermediate diameter soma trkA+, NGF dependent c-ret~ GDNF dependent peptidergic: SP+, CGRP+ VR1+, capsaicin sensitive mechanoreceptlve cutaneous afferents intermediate diameter soma multiple trk expression multiple neurotrophin dependence trkB, BDNF important for function proprioceptive muscle afferents intermediate/large diameter soma trkC+, NT-3 dependent express ETS-domain transcription factors: Era1+, PEA3+ strongly express neurofilament Fig. 2 Sensory neuron subtypes. The schematic represents a cross-sectional view of the neural tube (NT) and dorsal root ganglia (DRG) containing three classes of sensory neurons (dorsal, top; ventral, bottom). Some properties of each class are listed on the right. Nociceptive sensory neurons mediate painful stimuli and make central projections in the dorsal horn of the spinal cord. Mechanoreceptive sensory neurons mediate different types of touch; these too innervate the dorsal horn. Proprioceptive sensory neurons mediate spatial sense and make central projections to the ventral horn of the spinal cord, some synapsing directly onto motor neurons. 8 the adult (Silos-Santiago et al., 1995). These have small cell bodies and can be identified by their binding to the lectin BSI-B4. It appears that these neurons express trkA and depend on NGF during development, but then undergo a trophic switch to become dependent instead on glial derived neurotrophic factor (GDNF) late in development (Molliver et al., 1997; Fundin et al., 1999). The trkA+ and trkA nociceptors both make central connections in the dorsal horn of the spinal cord, but terminate in different layers indicating they play different roles in transducing painful stimuli (Stucky and Lewin, 1999). Many of the sensory neurons that express trkC and depend on neurotrophin-3 (NT-3) for survival are large-diameter muscle afferents that serve a proprioceptive function (Mu et al., 1993; Klein et al., 1994; McMahon et al., 1994). Large-diameter sensory neurons project to muscle spindles and Golgi tendon organs in the periphery and have central projections that terminate in the ventral horn of the spinal cord. These neurons express members of the ETS-domain family of transcription factors, such as ER81 and PEA3 , that may play a role in the specificity of connections between sensory and motor neuron pools innervating the same muscles (Lin et al., 1998). Mice homozygous for the targeted deletion of trkC have obvious proprioceptive defects and are missing all innervation to the ventral horn of the spinal cord (Klein et al., 1994). In addition to proprioceptive defects, trkC -/- mice also have some defects in mechanosensation. Such a deficit is consistent with the observed loss of both large and intermediate sized sensory neurons in trkC-/- mice, and the fact that trkC is sometimes co-expressed with trkA and/or trkB (Wright and Snider, 1995). The final class of sensory neurons is the most poorly defined. Sensory neurons that strongly express trkB have medium sized cell bodies and comprise less than 10% of the adult DRG (Mu et al., 1993; Wright and Snider, 1995). There are many cells, however, that co-express weaker levels of trkB with trkA and/or trkC, and it is likely that 9 this overlap is higher during development. A small number of sensory neurons are lost in mice with targeted deletions of trkB (Klein et al., 1993), but the function of these has not been determined. Although it was initially thought that mechanoreceptive cutaneous afferents would express trkB and depend on BDNF for survival, the story is much more complex. Different subtypes within the mechanoreceptive class appear to have multiple and overlapping neurotrophic dependencies (Fundin et al., 1997; Carroll et al., 1998). In addition, recent analyses of the combinatorial effects of different neurotrophin knockouts suggests that BDNF-dependent and NT-3-dependent neurons may derive from a common precursor that produces trkB+ positive neurons in response to NT-4 signaling and trkC+ neurons in the absence (Liebl et al., 2000). The development of sensory neurons is an interesting, if complex, system in which to study how and when neural crest cells become specified to their different fates. Sensory neurons are just one general class of neural crest derivatives. The neural crest also produces the autonomic class of neurons as well as many non-neuronal cell types such as glia, smooth muscle, neuroendocrine cells, and melanocytes. It is equally important to identify cues that regulate particular aspects of sensory neuron identity as it is to characterize the developmental potentials of sensory precursors in terms of their abilities to contribute to sensory subtypes and other neural crest lineages. Levels of cell fate specification The concept that progenitor cells acquire intrinsic developmental biases has been the subject of considerable discussion among biologists. The problem is that there seem to be different degrees to which a cell may be biased toward a particular fate. Operationally, there are three levels of bias that may affect a cell's potential. First, a cell is unspecified (unbiased) if it is multipotent and responds equally to different instructive cues or yields different cell types via stochastic events that are equally probable. Second, a cell is specified (biased) if it produces the same cell type(s) when it is placed in a "neutral" environment as it would in vivo (Slack, 1991). Such a bias may not be irreversible. Finally, a cell is committed or determined (biased) to a particular fate if it does not respond to instructive cues or environments that promote alternate fates. Thus, to conclude that a progenitor is determined to a particular fate, experiments must be performed to challenge the cell with instructive signals or new embryonic environments. Note that it is possible for a cell to be both multi potent and biased if it produces more of one cell type than another by preferentially responding to particular cues or by stochastic events that are not equally probable. Although commonly used, I find that the term restricted, when unqualified, is not adequately descriptive and can lead to confusion. Cells can be restricted from or restricted to particular cell types both in terms of fate and potential. In this discussion I will use the other terminology defined above. Specification of the sensory class of peripheral neurons Evidence for specified or determined sensory precursors Single-cell fate mapping in vivo has demonstrated that both premigratory and newly emigrated neural crest contains some cells that contribute exclusively to the DRG rather than to autonomic ganglia, the ventral root, or the dorsal migration pathway (BronnerFraser and Fraser, 1988; Bronner-Fraser and Fraser, 1989; Fraser and Bronner-Fraser, 1991; Raible and Eisen, 1994). Such data are consistent with the possibility that there are specified sensory precursors among the neural crest, but do not provide direct evidence toward such a conclusion. In vivo experiments have been performed that challenge the sensory potential of populations of neural crest. For example, it has been observed that the neural crest cells that migrate first along the ventromedial pathway contribute to autonomic derivatives while later migrating neural crest contribute to sensory ganglia (Weston, 1963; Serbedzija et al., 1989; Serbedzija et al., 1990). To determine whether 11 early and late migrating neural crest contain only committed autonomic or sensory precursors, respectively, grafting experiments were performed in which labeled donor neural tube fragments were heterochronically transplanted to host embryos (Weston and Butler, 1966). It was observed that both "young" (early migrating) and "old" (later migrating) neural crest have the potential to contribute to both sensory and autonomic ganglia. These results indicate that, as a population, early and late migrating progenitors are not strictly committed to colonizing particular ganglia. This study does not rule out the possibility that the neural crest at each stage is a mixture of specified sensory and autonomic precursors. Neural crest cells that have recently delaminated from the neural tube are heterogeneous with respect to their expression of some sensory markers (Ciment and Weston, 1982; Gradwohl et al., 1996; Perez et al., 1999), and it has been hypothesized that the precursors that express sensory markers are specified to become sensory neurons. In the case of the neurogenins (ngns), it has been shown that ectopic expression of these genes in vivo biases the migration of neural crest cells to the DRG and induces expression of sensory markers in other cell types (Perez et al., 1999). Since the neurogenins are expressed in the neural tube and in delaminating neural crest, it is possible that endogenous ngns function to bias premigratory neural crest cells toward a sensory neuron fate. Ectopic expression of ngnl in the zebrafish CNS induces different types of neurons, the subtype of which is dependent on local cues (Blader et al., 1997). Preliminary studies of ectopic ngn expression in mammalian cells in vitro yield similar results (L. Lo and DJ.A., unpublished observation), underscoring the importance of identifying appropriately responsive precursors when studying the effects of a particular "determinant." Regardless, the ngns may play a role in the specification of the sensory fate from certain neural crest or neural tube precursors. We do not yet know the level of specification of ngn expressing cells in vivo. In particular, it will be important to 12 determine if all ngn expressing neural crest cells go on to make only sensory neurons and whether these ngn+ cells are committed to the sensory fate. Studies performed in vitro provide the strongest evidence for specified sensory precursors among recently emigrated neural crest. When neural crest was directly explanted from the midbrain region of quail embryos, neurons with substance P (SP) immunoreactivity were observed to differentiate from postmitotic precursors (Ziller et al., 1983; Ziller et al., 1987). The differentiation of such neurons was promoted by growth in a defined medium whereas these neurons did not appear in serum-containing medium. Although these data were interpreted as evidence for committed sensory precursors, the failure of sensory neurons to appear in medium that contained serum could be due to the differentiation of sensory precursors into another cell type. Other studies, however, have described similar neurons in cultures of both quail and mouse truncal neural crest (Sieber-Blum, 1989; Matsumoto, 1994a; Matsumoto, 1994b). These neurons were reported to differentiate from postmitotic precursors under conditions that were permissive for the division and differentiation of other neuronal types. The main reason for concluding that the precursors were specified to a sensory neuron fate was because they differentiated without dividing; by definition they did not produce any other cell types besides sensory neurons. Similarly, specification of neuronal identity is thought to occur concomitant with cell cycle arrest in other parts of the nervous system (McConnell and Kaznowski, 1991; Ezzeddine et al., 1997). Recently, we have observed evidence of dividing, determined sensory precursors in cultures of rat neural crest (see Chapter 2, Greenwood et al., 1999). Sensory precursors were characterized in these cultures with a number of transcription factor markers that could unambiguously distinguish them from autonomic neurons. We observed that sensory neurons arose from dividing precursors even in the presence of the autonomic-inducing cue, bone morphogenetic protein-2 (BMP2). This study is the first 13 to challenge sensory precursors with conditions known to be instructive for a non-sensory fate. Operationally, these dividing sensory precursors appear to be committed to the sensory fate. The caveat, which exists for all experiments that challenge the potential of a cell, is that there may exist other instructive signals that can prevent these precursors from adopting a sensory fate. Additional experiments that challenge these precursors with other cues and environments will be required to more completely describe their state of commitment. Evidence for multipotent progenitors that contribute to sensory neurons as well as other major neural crest lineages Clonal analyses performed both in vivo and in vitro provide strong evidence that the neural crest does contain some precursors that produce sensory neurons in the DRG as well as contribute cells to other neural crest derivatives. When either individual neural tube cells or individual migrating neural crest cells in chicken embryos were labeled with a vital dye, at least some of the resulting clones were observed to contain cells in multiple neural crest derivatives including: neurons and non-neuronal cells in the DRG and sympathetic ganglion, glia in the ventral root, cells in the adrenal gland, and pigment cells beneath the ectoderm (Bronner-Fraser and Fraser, 1988; Bronner-Fraser and Fraser, 1989; Fraser and Bronner-Fraser, 1991 ). Similarly, crest cells that generate sensory neurons as well as other types of progeny have been observed in zebrafish (Raible and Eisen, 1994). In vitro clonal analyses of quail neural crest cells have demonstrated that clones containing putative sensory cells, as identified by SSEA-1 immunoreactivity, can also contain pigmented cells and cells expressing autonomic markers such as tyrosine hydroxylase (TH) or dopamine-~-hydroxylase (DBH) (Sieber-Blum, 1989). These data suggest that individual neural crest cells can give rise to both sensory and autonomic neurons as well as other cell types . To date, it has not been possible to analyze individual 14 mammalian neural crest cells that have sensory potential. Sensory neurons have not been observed from mammalian cells grown at clonal density (A.LG., unpublished observation). The identification of conditions under which mammalian sensory neurons can develop in clonal cultures would represent an important advance in neural crest biology. One major question concerning such multipotent sensory precursors is when they become specified to the sensory lineage. Clonal analysis of the progenitors found in quail DRG and sympathetic ganglia suggests that multipotent precursors persist within the ganglia for quite a long time in development (Duff et al., 1991). Precursors from the DRG produced clones that contained SSEA-1 + sensory-like cells, pigments cells, and autonomic-like cells even at two days after gangliogenesis (E6, st. 29). By four days after DRG formation (E8, st. 34) progenitors produced unpigmented colonies (without melanocytes) with both sensory- and autonomic-like immunoreactivities. The percentage of multipotent colonies appeared to decrease with time; this could either reflect the progressive restriction of a homogeneous pool of multipotent precursors or the loss of multipotent precursors from a heterogeneous pool. Quite similarly, an in vivo study in which DRGs were backgrafted to the crest migration pathway demonstrated that the grafted ganglia could contribute to neurons and glia in both the host DRG and sympathetic ganglia, but no longer could contribute to sensory ganglia if they were taken from the donor after E8 (Schweizer et al., 1983). Again, this study does not distinguish between the progressive restriction of multipotent precursors and the loss of a specified group of sensory progenitors. Taken together, however, these data suggest that some multipotent progenitors with sensory potential may persist in the DRG until E8 in quail. It is interesting that cells with autonomic potential can be found in sensory ganglia well after the sensory neurons are born (Schweizer et al., 1983; Deville et al., 1992). The normal fate of these cells has not been established, but presumably they give 15 rise to glia or other support cells and could be similar or identical to precursors isolated from the sciatic nerve that have autonomic, glial, and myofibroblast potentials (Morrison et al., 1999). Whether autonomic ganglia contain cells with sensory potential is less certain. Autonomic ganglia from quail can produce colonies containing SSEA-1 + cells for a limited time in development (Duff et al., 1991). By contrast, backgrafting studies in which autonomic ganglia were placed at the top of the crest migration pathway showed that these ganglia do not contribute cells to the DRG (Le Lievre et al., 1980). It is possible, however, that this assay does not provide an efficient enough challenge of precursors to reveal their sensory potential. Alternatively, culture conditions may have induced sensory capacity in Duff et. al. (1991 ). Identification of a strong, sensory neuron inducing cue would be of great use in resolving the question of which cells have sensory potential. Extrinsic signals implicated in the specification of the sensory neuron fate Although many growth factors appear to affect the development of sensory neurons and precursors, two in particular have been implicated in the specification of the sensory fate from neural crest cells: BDNF and leukemia inhibitory factor (LIF). The neurotrophin BDNF has been reported to upregulate the expression of the sensory marker SSEA-1 in clonal cultures of quail neural crest (Sieber-Blum, 1991). Importantly, this induction occurred without a concurrent increase of colony size. These data suggest that the effect of BDNF was to instruct neural crest precursors to adopt a sensory fate at the expense of other fates, rather than to promote the survival or proliferation of sensory precursors. It should be noted, however, that most of the colonies were three weeks old when they were analyzed for cell number, time enough to amplify slight differences in growth rate and survival. Such an instructive effect of BDNF has not yet been shown for mammalian precursors. In contrast, mammalian neural crest precursors respond to LIF by producing 16 sensory neurons. Addition of murine LIF to mouse neural crest cultures was observed to strongly promote the differentiation of CORP+ neurons from non-dividing precursors (Murphy et al., 1991). Although neuropeptides such as CORP can be expressed in cultured autonomic neurons and thus are not specific sensory markers (Nawa and Patterson, 1990; Fann and Patterson, 1994b; Fann and Patterson, 1994a), a later study has shown that murine LIF promotes the differentiation of neurons that express the pansensory marker Bm-3.0 and have capsaicin sensitivity, a characteristic of a subtype of sensory neurons (Carey and Matsumoto, 1999). The effect of LIF to promote differentiation of non-dividing precursors into sensory neurons could be consistent with an instructive effect. Alternately, it may act permissively to allow the differentiation of precursors that were already specified for a sensory fate . A clonal analysis that determines the effect of LIF on other non-sensory fates, such as glia or autonomic neurons, could distinguish between these possibilities. It is also interesting to note that LIF appears to promote the differentiation of sensory neurons with a nociceptive phenotype. Recent genetic evidence from zebrafish suggests a role for BMP family members in the specification of some sensory neurons during the neural plate stage, rather than after neural crest migration. In zebrafish, there are two types of sensory neurons: RohonBeard (RB) cells that serve mechanosensory functions in larvae and DRO neurons that derive from the classically recognized neural crest and survive into adulthood. The progenitors of RB cells and neural crest both arise from the same area of the lateral neural plate, the numbers of these cells generated is affected by levels of BMP activity (Barth et al., 1999; Nguyen et al., 2000). Recently it has been found that the gene narrowminded (nrd') is autonomously required for the development of both RB neurons and early migrating neural crest cells, suggesting, though not proving, the existence of a common progenitor for these cells (Artinger et al., 1999). Furthermore, parallels have been drawn 17 between the ngn2 expressing sensory neurons in mammals and the RB cells because both develop from early emigrating precursors. Interestingly, ectopic expression of BMP4 or mutations which dorsalize embryos but have residual BMP activity greatly expand the areas of zebrafish ngn expression and the number of RB cells (Neave et al., 1997; Barth et al., 1999; Nguyen et al., 2000). This genetic analysis is consistent with a role of BMP family members in specifying a responsive zone in which sensory progenitors can form. It is likely that additional cues are necessary to actually specify the sensory neuron fate from these precursors because alterations of BMP levels do not change the density of neurons in the expanded neurogenic regions. It is tempting to speculate, however, that BMPs acting in the neural tube may affect sensory neuron production in mammals as well. Intriguingly, we have observed that treatment of neural tube explants with BMP2 also appears to expand the domain of ngn expression (see Fig. 8 in Chapter 2) and that exposure of neural tube cells to BMP2 can induce the expression of Bm-3.0 (see Fig. 9 in Appendix 2). Specification of sensory neuron subtypes The question of whether particular sensory subtypes are specified at the time of neural crest migration is important in terms of understanding the initial source of heterogeneity among sensory neurons. At one extreme, all sensory neurons may be specified generally as a class and only later acquire subtype characteristics. In this case, sensory subtype fate could be specified by interactions between postmigratory sensory precursors, environmental influences at the time of differentiation, or interactions with peripheral or central targets after differentiation. This scenario predicts a fair amount of plasticity in the types of neurons that form under different conditions. At the other extreme, all precursors may acquire a subtype fate at the same time that they become specified as sensory neurons, perhaps some even before they migrate from the neural tube. This scenario predicts that there will be very early differences between classes of sensory 18 neurons and little plasticity. Clearly, there are many possibilities that lie between the two extremes. The way to discriminate between these possibilities would be to challenge sensory precursors with cues that are known to induce particular subtype fates and assess their response. At the present time, this sort of analysis is beyond our abilities. We are only beginning to identify the cues that regulate particular sensory characteristics and it is not known what factors are responsible for instructing particular subtype fates. Additionally, there are no known intrinsic determinants that distinguish between precursors fated to produce sensory neurons with different functions. Without these tools we cannot directly address the question, but an analysis of when sensory subtype heterogeneity first becomes apparent and how subtype characteristics are affected under certain conditions may provide important clues and allow us to rule out certain possibilities. Evidence for precursors that are specified to a particular sensory subtype fate Heterogeneity among sensory neurons in the DRG is apparent very early in gangliogenesis. Different types of neurons are born (become postmitotic) at different times: the first neurons born are fated to contribute to the large-diameter subclass and tend to occupy the ventrolateral (VL) part of the DRG. The small diameter sensory neurons are born a day later and occur in the dorsomedial (DM) part of the DRG (Lawson et al., 1974; Lawson and Biscoe, 1979; Kitao et al., 1996). Although this is generally true, sensory neurogenesis may go on longer than originally thought due to contribution of new progenitors from the neural tube (Sharma et al., 1995; Kitao et al., 1996). Clonal analysis in vivo of neural tube precursors that contribute to the DRG suggests that there are two types of precursors with respect to their subtype fates (Frank and Sanes, 1991). One class of precursors is observed when labeling is performed early. These cells divide only a few times and are fated to give rise to only large-diameter sensory neurons in the 19 VL part of the DRG ("VL only" clones). The second type of precursor is apparent when the labeling is done later. These typically undergo many divisions and produce sensory neurons with both large and small diameters located in both the VL and DM parts of the DRG ("VL+DM" clones). In this study, no precursors were observed to give rise to only small diameter sensory neurons . It is possible that no "DM only" clones were observed because the precursors analyzed did not include the latest wave of migrating cells. Regardless, the existence of clones that contain only one sensory subtype (VL only) is consistent with the hypothesis that some sensory precursors may be specified to the largediameter (presumably proprioceptive, muscle afferent) fate when they emigrate from the neural tube. The ability to determine the potential of such precursors awaits a method to prospectively identify and then challenge them with cues that induce the small-diameter (nociceptive, cutaneous afferent) subtype fate. Similar sensory precursors have been revealed in the analysis of ngnl and ngn2 mutant mice. Sensory neurogenesis in mice depends on the ngns since ngnl-/-;ngn2-/embryos lack DRGs but contain autonomic ganglia (Ma et al., 1999). In ngn2-/- mutants, neurogenesis is delayed, indicating a defect in the early wave of precursors. However, the DRGs of these fetuses contain the full complement of sensory subtypes, suggesting that ngnr precursors have the potential to make both large and small diameter subtypes and may be similar to the multipotent precursors identified by clonal analysis in chick. In contrast, ngnl-1- mutants are missing the later wave of neurogenesis. The DR Gs of these mice are depleted of trkA-expressing neurons (small-diameter, nociceptive). Thus, the ngn2+ cells do not compensate for the loss of ngnl-dependent precursors. This phenotype is consistent with the notion that ngn2-dependent precursors do not have the potential to make trkA+ sensory neurons; however, it is also possible that the early wave of neurogenesis is complete and ngn2 expression is extinguished by the time the appropriate nociceptive-inducing signals appear. This question might be resolved by knocking the 20 ngn2 gene into the ngnl locus to see if trkA + neurons can develop from ngn2-expressing cells. Arguments for the early specification of sensory subtypes have also been put forth based on the projection patterns of sensory neurons identified by retrograde labeling of target tissues (Honig, 1982; Honig et al. , 1998). The case that developing sensory neurons are prespecified to innervate particular muscle or cutaneous targets rests mainly on the observations that 1.) sensory axons projecting to a particular target appear to be bundled from the outset and to make distinct pathway choices within a very short distance from the DRG, and 2.) the cell bodies of sensory neurons projecting to a particular target are not all located near each other in the DRG (based on retrograde labeling of neurons with soma in the L V part of the DRG). An additional observation cited as support for this hypothesis is that individual neurons that project to a particular target often have cell bodies that appear in pairs or small clusters, suggesting that a prespecified precursor divided to produce a clone of neurons that all actively chose the same target. However, there are other interpretations of the data that do not require the early specification of precursors or neurons to a particular subtype fate. Also possible, I believe, is a scenario in which unspecified sensory precursors send out axons that follow the nearest available guidance cue. The first born sensory neurons in the VL part of the DRG would encounter a variety of different guidance cues, many of which are likely to be a motor axons on their way to the muscle. Sensory neurons born late in development would encounter a more "crowded" environment and would be more likely to follow other sensory axons, most of which had innervated a cutaneous target because the muscle axons were occupied. This model predicts that the cell bodies of the later born neurons should be more grouped with respect to their peripheral targets because the pathways of innervation are already well established by the time they are extending axons . 21 Retrograde labeling of targets that resulted in labeled cell bodies in the DM part of the DRG was not reported; it is possible these neurons are more clustered. Although the available evidence does not directly demonstrate the existence of precursors that are specified to a particular sensory subtype, it does suggest that there are early differences in the types of precursors that contribute to the DRG. The ability to isolate different types of sensory precursors and the identification of cues involved in the specification of sensory subtypes would be an important step in resolving these questions. Chapter 3 of this thesis provides an initial characterization of some cues that affect the development of particular sensory subtype characteristics. Evidence for precursors among the neural crest that can generate multiple sensory subtypes Lineage and genetic analyses presented above suggest that there may be precursors among the neural crest that can make multiple types of sensory neurons . Individual progenitors labeled in the neural tube produce clones of cells in the DRG that include both large and small diameter sensory neurons (Frank and Sanes, 1991). Similarly, the population of ngnl expressing precursors can generate all subtypes of sensory neurons as assayed by expression of the trk receptors (Ma et al., 1999). It is likely that these studies reveal sensory precursors among the neural crest that are multipotent with respect to sensory subtype fate. Single cell analyses of migrating neural crest would bolster this conclusion. For example, it would be interesting to determine the range of sensory subtype fates of individual neural crest cells in the context of wild-type and ngn2 mutant embryos. The recent advent of inducible, stable lineage markers, such as the Crerecombinase system (Sauer and Henderson, 1988), may make such an analysis possible if these methods can be adapted to follow the progeny of individual cells. 22 Studies performed in vivo that alter the available targets for sensory innervation reveal a certain amount of subtype plasticity. In the normal course of development, each DRG innervates only targets within a particular, well-defined dermatome (segment) of the body. DRGs along the neuraxis differ in the composition of sensory neuron subtypes they contain based on the types of targets present in that particular segment. For example, in the bullfrog, DRG2 innervates the arm and contains a significant number of muscle afferents, whereas DRG3 innervates the ventral part of the thorax and contains mostly cutaneous afferents (Frank and Westerfield, 1982b). If sensory neurons were strictly specified to innervate particular targets, one might expect that they would not innervate structures in a different dermatome than their origin. It has been demonstrated, however, that upon removal of DRG2, the adjacent ganglia, DRG3 , will innervate many of the muscles in the arm and form appropriate connections in the spinal cord (Frank and Westerfield, 1982a). Furthermore, if entire ganglia are transplanted from the midthoracic region to the location of DRG2, they also innervate peripheral and central targets appropriate for the new location (Smith and Frank, 1987). These results are particularl y interesting since most of the neurons in the thoracic DRGs are cutaneous and not muscle afferents . The authors favored the interpretation that sensory subtype fate is plastic until late in development, and that the environment plays an important role in the specification of different subtypes. These analyses, however, did not follow the fate of individual neurons. It is therefore equally possible that the thoracic DRGs contain a mixed population of specified precursors or neurons and that the environment plays a selective, rather than instructive, role in matching the subtype of sensory neuron with the type of available target. Regardless, it is clear that sensory neurons are not restricted to innervating only targets in the dermatome from which they originate. In vitro evidence has also been interpreted to suggest that early differentiating sensory neurons may not be initially specified with respect to their targets (Adams and 23 Scott, 1998). Cranial sensory ganglia, unlike DRGs, contain areas that can be prospectively identified to produce neurons of a particular subtype. When sensory neurons are isolated early in development from two areas that are fated to innervate either muscle or cutaneous targets, the growth cones of these axons respond similarly in vitro with respect to their avoidance of and attraction to particular target choices. However, if the sensory neurons from these structures are isolated at later times they respond differently to the targets. The authors suggested that one reason for such observations might be if the neurons isolated at early times did not have rigidly specified identities. Although this is possible, it is not clear how well such a target-choice assay challenges the identity of a neuron. The inability of immature neurons to discriminate between targets could simply indicate that their pathfinding machinery is not fully developed. Given that the number of mature subtypes of neurons in the sensory ganglia is quite large, it is likely that some precursors are "multipotent" with regards to some aspects of identity. It is also possible that some subtype characteristics are specified early in neural crest development. The analysis of when particular sensory subtype characteristics become determined awaits the ability to efficiently follow the progeny of individual cells and the assess these clones for expression of subtype markers under conditions that challenge their fate. Extrinsic signals implicated in the specification of sensory neuron subtype fates Due to their ability to support the survival of different types of sensory neurons, it has been hypothesized that neurotrophins may play a role in the specification of subtype fate (Lewin, 1996). This hypothesis is intuitively appealing, but what is the evidence? Both the neurotrophins NT-3 and BDNF have the opportunity to exert early effects on subtype fate because they are present in developing DRGs (Emfors et al., 1992; Schecterson and Bothwell, 1992). They have also been shown to affect the proliferation and maturation of 24 sensory neuroblasts before the period of neurotrophin dependence (Wright et al., 1992; Memberg and Hall, 1995), but there is no evidence that they directly specify the fates of NT-3- and BDNF-dependent neurons . Indeed, the observation that some sensory neurons appear to switch their neurotrophin dependency from NT-3 or BDNF to NGF (Buchman and Davies, 1993; Buj-Bello et al., 1994; Pinon et al., 1996) suggests that NT-3 and BDNF are not sufficient to specify subtype survival properties. Furthermore, when NT-3 is ectopically expressed in vivo using a muscle or skin specific promoter (Albers et al., 1996; Wright et al., 1997), it selectively increases the number of neurons within appropriate subpopulations without depleting the others. Evidence presented in Chapter 3 indicates that the continued presence of NT-3 and BDNF does not prevent the differentiation of trkA + neurons in vitro. Although we cannot rule out the possibility that NT-3 or BDNF plays an instructive role in specifying the subtype fate of some precursors, it does not seem to be a general rule that sensory precursors or neurons become dependent on the first neurotrophins they are exposed to. In the case of NGF, there is direct evidence to suggest that this signaling pathway is not required for the initial specification of the small-diameter nociceptive subtype. If it were, one would expect that the nociceptive fate would not be specified in the absence of NGF or trkA. The phenotype of trkA-/- and NGF-/- mutants, however, precludes the analysis of sensory subtype because the neurons are lost during the period of programmed cell death before they have any obvious identity (Smeyne et al., 1994). A recent and elegant study has addressed this problem: when the trkA-/- phenotype is analyzed in the context of the BAX-/- mutation, which prevents apoptotic death, many small-diameter neurons are found in DRGs that make appropriate connections to the dorsal horn of the spinal cord (Patel et al., 2000). These neurons do have defects in nociceptive function, presumably due to lack of peripheral innervation and expression of neuropeptides, but they appear to have retained a fundamentally nociceptive identity 25 based on their morphology, location, and central projections. Neurotrophins have pleiotropic and important effects on the development of sensory neuron characteristics, but it appears that they are not generally required for the initial specification of sensory subtypes. It will be very interesting to see how sensory subtype characteristics are affected in trkB-/- and trkC-/- mutants crossed into the BAX-/- background. Concluding remarks The neurons in the PNS are an exceedingly heterogeneous group of cells and the sensory lineage especially so. Although many intrinsic and extrinsic cues have been discovered that regulate particular aspects of peripheral neuron identity, the challenge still remains to understand the main forces that initiate the diversity. Given that specification, as a general process, seems to work by the progressive subdivision of a progenitor population into groups that have different response properties, if we are to make sense of PNS subtype diversity it seems important to focus on the earliest specification events. Evidence presented here and elsewhere indicates that some, though certainly not all, of these decisions affecting the PNS may occur among premigratory or early migrating neural crest. This in itself emphasizes the need to characterize the developmental potentials of precursors within the neural tube and neural crest at different times. The isolation of precursors based on gene or protein expression should provide an important tool in this analysis. Additionally, the increasing ability to apply genetic analysis in mice and zebrafish to questions of PNS development will greatly contribute to the understanding of initial specification events. It should also be pointed out that lineage relationships in the PNS may not correspond to progressive restrictions within the categories we have arbitrarily defined in this system. For example, one type of precursor may be limited to producing a subset of 26 sensory neuron types, while another may have the potential to produce a different subset of sensory subtypes plus autonomic, glial, and or myofibroblast fates . The characterization of the precursors that contribute to the PNS is not only important in terms of understanding what cells can and cannot respond to particular cues, but it may also reveal an underlying logic to the development of the system. 27 References Adams, D. H. and Scott, S. A. (1998). Response of "naive" cutaneous and muscle afferents to potential targets in vitro. Dev Biol 203: 210-220. Anderson, D. J. (1989). The neural crest cell lineage problem: Neuropoiesis? Neuron 3: 1-12. Anderson, D. J. (1993). Cell and molecular biology of neural crest cell lineage diversification. Curr. Op. Neurobiol. 3: 8-13. Anderson, D. J. (1997). Cellular and molecular biology of neural crest cell lineage determination. Trends Genet. 13: 276-280. Artinger, K. B., Chitnis, A. B., Mercola, M. and Driever, W. (1999). Zebrafish narrowminded suggests a genetic link between formation of neural crest and primary sensory neurons . Development 126: 3969-3979. Barth, K. A., Kishimoto, Y., Rohr, K. B., Seydler, C., Schulte-Merker, S. and Wilson, S. W. (1999). Bmp activity establishes a gradient of positional information throughout the entire neural plate. Development 126: 4977-4987. Begley, C. G., Lipkowitz, S., Gobel, V., Mahon, K. A., Bertness, V., Green, A. R., Gough, N. M. and Kirsch, I. R. (1992). Molecular characterization of NSCL, a gene encoding a helix-loop-helix protein expressed in the developing nervous system. Proc. Natl. Acad. Sci. USA 89: 38-42. Blader, P., Fischer, N., Gradwohl, G., Guillemot, F. and Strahle, U. (1997). The activity of Neurogeninl is controlled by local cues in the zebrafish embryo. Development 124: 4557-4569. 28 Blaugrund, E., Pham, T. D., Tennyson, V. M., Lo, L., Sommer, L., Anderson, D. J. and Gershon, M. D. (1996). Distinct subpopulations of enteric neuronal progenitors defined by time of development, sympathoadrenal lineage markers and Mash-I-dependence. Development 122: 309-320. Bronner-Fraser, M. and Fraser, S. (1988). Cell lineage analysis shows multipotentiality of some avian neural crest cells. Nature 335: 161-164. Bronner-Fraser, M. and Fraser, S. (1989). Developmental potential of avian trunk neural crest cells in situ. Neuron 3: 755-766. Bronner-Fraser, M. E. (1993). Segregation of cell lineage in the neural crest. Curr. Op. Gen. Devel. 3: 641-647. Buchman, V. L. and Davies, A. M. (1993). Different neurotrophins are expressed and act in a developmental sequence to promote the survival of embryonic sensory neurons. Development 118: 989-1001. Buj-Bello, A., Pinon, L. G. P. and Davies, A. M. (1994). The survival of NGFdependent but not BDNF-dependent cranial sensory neurons is promoted by several different neurotrophins early in their development. Development 120: 1573-1580. Carey, M. B. and Matsumoto, S. G. (1999). Neurons differentiating from murine neural crest in culture exhibit sensory or sympathetic-like calcium currents. J Neurobiol 39: 501-514. Ciment, G. and Weston, J. A. (1982). Early appearance in neural crest and crestderived cells of an antigenic determinant present in avian neurons. Dev. Biol. 93: 355-367. 29 Cserjesi, P., Brown, D., Lyons, G. E. and Olson, E. N. (1995). Expression of the novel basic helix-loop-helix gene eHAND in neural crest derivatives and extraembryonic membranes during mouse development. Devel. Biol. 170: 664678 . Deville, F. S.-S. -. C., Ziller, C. and Le Douarin, N. (1992). Developmental potentialities of cells derived from the truncal neural crest in clonal cultures. Dev. Brain Res. 66: 1-10. Duff, R. S., Langtimm, C. J., Richardson, M. K. and Sieber-Blum, M. (1991). In Vitro Clonal Analysis of Progenitor Cell Patterns in Dorsal Root and Sympathetic Ganglia of the Quail Embryo. Devel. Biol. 147: 451-459. Ernfors, P., Merlio, J.-P. and Persson, H. (1992). Cells expressing mRNA for neurotrophins and their receptors during embryonic rat development. Eur. J. Neurosci. 4: 1140-1158. Ezzeddine, Z., Yang, X., DeChiara, T., Yancopoulos, G. and Cepko, C. (1997). Postmitotic cells fated to become rod photoreceptors can be respecified by CNTF treatment of the retina. Development 124: 1055-1067. Fann, M.-J. and Patterson, P.H. (1994a). Depolarization differentially regulates the effects of bone morphogenetic protein (BMP)-2, BMP-6, and Activin-A on sympathetic neuronal phenotype. J. Neurochem. 63: 2074-2079. Fann, M.-J. and Patterson, P.H. (1994b). Neuropoietic cytokines and activin A differentially regulate the phenotype of cultured sympathetic neurons. Proc. Natl. Acad. Sci. USA 91: 43-47. Fedtsova, N. G. and Turner, E. E. (1995). Brn-3.0 expression identified early postmitotic CNS neurons and sensory neural precursors. Mech. Dev. 53: 291-304. 30 Fode, C., Gradwohl, G., Morin, X., Dierich, A., LeMeur, M., Goridis, C. and Guillemot, F. (1998). The bHLH protein NEUROGENIN 2 is a determination factor for epibranchial placode-derived sensory neurons. Neuron 20: 483-494. Frank, E. and Sanes, J. R. (1991). Lineage of neurons and glia in chick dorsal root ganglia: analysis in vivo with a recombinant retrovirus. Development 111: 895908. Fraser, S. E. and Bronner-Fraser, M. E. (1991). Migrating neural crest cells in the trunk of the avian embryo are multipotent. Development 112: 913-920. Fundin, B. T., Mikaels, A., Westphal, H. and Ernfors, P. (1999). A rapid and dynamic regulation of GDNF-family ligands and receptors correlate with the developmental dependency of cutaneous sensory innervation. Development 126: 2597-2610. Fundin, B. T., Silos-Santiago, I., Ernfors, P., Fagan, A. M., Aldskogius, H., DeChiara, T. M., Phillips, H. S., Barbacid, M., Yancopoulos, G.D. and Rice, F. L. (1997). Differential dependency of cutaneous mechanoreceptors on neurotrophins, trk receptors and p75LNGFR. Dev. Biol. 190: 94-116. Gobel, V., Lipkowitz, S., Kozak, C. A. and Kirsch, I. R. (1992). NSCL-2: a basic domain helix-loop-helix gene expressed in early neurogenesis. Cell Growth Differ 3: 143-148. Gradwohl, G., Fode, C. and Guillemot, F. (1996). Restricted expression of a novel murine atonal-related bHLH protein in undifferentiated neural precursors. Dev. Biol. 180: 227-241. Greenwood, A. L., Turner, E. E. and Anderson, D. J. (1999). Identification of dividing, determined sensory neuron precursors in the mammalian neural crest. Development 126: 3545-3559. 31 Groves, A. K. and Bronner-Fraser, M. (1999). Neural crest diversification. Curr Top Dev Biol 43: 221-258. Guillemot, F., Lo, L.-C., Johnson, J.E., Auerbach, A., Anderson, D. J. and Joyner, A. L. (1993). Mammalian achaete-scute homolog-1 is required for the early development of olfactory and autonomic neurons . Cell 75: 463-476. Honig, M. G. (1982). The development of sensory projection patterns in embryonic chick hindlimb. J. Physiol. 330: 175-202. Honig, M. G., Frase, P.A. and Camilli, S. J. (1998). The spatial relationships among cutaneous, muscle sensory and motor neuron axons during the development of the chick hindlimb. Development 125: 995-1004. Horstadius, S. (1950). The Neural Crest. Oxford University Press, London. Kandel, E. R., Schwartz, J. H. and Jessell, T. M. (1991). Principles of Neural Science. Elsevier, New York. Kitao, Y., Robertson, B., Kudo, M. and Grant, G. (1996). Neurogenesis of subpopulations of rat lumbar dorsal root ganglion neurons including neurons projecting to the dorsal column nuclei. J Comp Neural 371: 249-257. Klein, R., Silos-Santiago, I., Smeyne, R. J., Lira, S. A., Brambilla, R., Bryant, S., Zhang, L., Snider, W. D. and M., B. (1994). Disruption of the neurotrophin-3 receptor gene trkC eliminates Ia muscle afferents and results in abnormal movements. Nature 368: 249-251. Klein, R., Smeyne, R. J., Wurst, W., Long, L. K., Auerbach, B. A., Joyner, A. L. and Barbacid, M. (1993). Targeted disruption of the trkB neurotrophin receptor gene results in nervous system lesions and neonatal death. Cell 75: 113-122. Lawson, S. N. and Biscoe, T. J. (1979). Development of mouse dorsal root ganglia: an autoradiographic and quantitative study. J. Neurocytol. 8: 265-274. 32 Lawson, S. N., Caddy, K. W. T. and Biscoe, T. J. (1974). Development of rat dorsal root ganglion neurones: studies of cell birthdays and changes in mean cell diameters. Cell Tiss. Res. 153: 399-413. Le Douarin, N. M. (1982). The Neural Crest. Cambridge University Press, Cambridge, UK. Le Douarin, N. M. (1986) . Cell line segregation during peripheral nervous system ontogeny. Science 231: 1515-1522. Le Douarin, N. M., Dupin, E. and Ziller, C. (1994) . Genetic and epigenetic control in neural crest development. Curr. Op. Gene. Dev. 4: 685-695. Le Lievre, C. S., Schweizer, G. G., Ziller, C. M. and Le Douarin, N. M. (1980). Restrictions in developmental capabilities in neural crest cell derivatives as tested by in vivo transplantation experiments. Dev.Biol. 77: 362-378. Lee, J.E., Hollenberg, S. M., Snider, L., Turner, D. L., Lipnick, N. and Weintraub, H. (1995). Conversion of Xenopu s ectoderm into neurons by NeuroD, a basic helix-loop-helix protein. Science 268: 836-844. Lewin, G. R. (1996). Neurotrophins and the specification of neuronal phenotype. Philos Trans R Soc Land B Biol Sci 351: 405-411. Lin, J. H., Saito, T., Anderson, D. J., Lance-Jones, C., Jessell, T. M. and Arber, S. (1998) . Functionally-related motor neuron pool and muscle sensory afferent subtypes defined by coordinate ETS gene expression. Cell 95: 393-407. Lo, L.-C., Johnson, J.E., Wuenschell, C. W., Saito, T. and Anderson, D. J. (1991). Mammalian achaete-scute homolog 1 is transiently expressed by spatiallyrestricted subsets of early neuroepithelial and neural crest cells. Genes & Dev. 5: 1524-1537. 33 Ma, Q., Chen, Z. F., Barrantes, I. B., de la Pompa, J. L. and Anderson, D. J. (1998). Neurogenin 1 is essential for the determination of neuronal precursors for proximal cranial sensory ganglia. Neuron 20: 469-482. Ma, Q., Fode, C., Guillemot, F. and Anderson, D. J. (1999). NEUROGENINl and NEUROGENIN2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes & Dev. 13: in press. Ma, Q., Kintner, C. and Anderson, D. J. (1996). Identification of neurogenin, a vertebrate neuronal determination gene. Cell 87: 43-52. Marusich, M. F. and Weston, J. A. (1991). Development of the neural crest. Curr Op Genet Devel 1: 221-229. Matsumoto, S. G. (1994a). Neuronal differentiation in cultures of murine neural crest. II. Development of capsaicin-sensitive neurons. Dev. Brain Res. 83: 17-27. Matsumoto, S. G. (1994b). Neuronal differentiation in cultures of murine neural crest. I. Neurotransmitter expression. Dev. Brain. Res. 83: 1-16. McConnell, S. K. and Kaznowski, C. E. ( 1991 ). Cell cycle dependence of laminar determination in developing neocortex. Science 254: 282-285. McMahon, S. B., Armanini, M. P., Ling, L. H. and Phillips, H. S. (1994). Expression and coexpression of Trk receptors in subpopulations of adult primary sensory neurons projecting to identified peripheral targets. Neuron 12: 1161-1171. Memberg, S. P. and Hall, A. K. (1995) . Proliferation, differentiation, and survival of rat sensory neuron precursors in vitro require specific trophic factors. Mal Cell Neurosci 6: 323-335. Molliver, D. C., Wright, D. E., Leitner, M. L., Parsadanian, A. S., Doster, K., Wen, D., Yan, Q. and Snider, W. D. (1997). IB4-binding DRG neurons switch from NGF to GDNF dependence in early postnatal life. Neuron 19: 849-861. 34 Morrison, S. J., White, P. M., Zock, C. and Anderson, D. J. (1999) . Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96: 737-749. Mu, X., Silos-Santiago, I., Carrol, S. L. and Snider, W. D. (1993). Neurotrophin receptor genes are expressed in distinct patterns in developing dorsal root ganglia. J. Neurosci. 13: 4029-4041. Murphy, M., Reid, K., Hilton, D. J. and Bartlett, P. F. (1991). Generation of sensory neurons is stimulated by leukemia inhibitory factor. Proc. Natl. Acad. Sci. USA 88: 3498-3501. Nawa, H. and Patterson, P. H. (1990). Separation and partial characterization of neuropeptide-inducing factors in heart cell conditioned medium. Neuron 4: 269277. Neave, B., Holder, N. and Patient, R. (1997). A graded response to BMP-4 spatially coordinates patterning of the mesoderm and ectoderm in the zebrafish. Mech Dev 62: 183-195. Nguyen, V. H., Trout, J., Connors, S. A., Andermann, P., Weinberg, E. and Mullins, M. C. (2000). Dorsal and intermediate neuronal cell types of the spinal cord are established by a BMP signaling pathway. Development 127: 1209-1220. Patel, T. D., Jackman, A., Rice, F. L., Kucera, J. and Snider, W. D. (2000). Development of sensory neurons in the absence of NGF/TrkA signaling in vivo. Neuron 25: 345-357. Pattyn, A., Morin, X., Cremer, H., Goridis, C. and Brunet, J.-F. (1997). Expression and interactions of the two closely related homeobox genes Phox2a and Phox2b during neurogenesis. Development 124: 4065-4075 . 35 Pattyn, A., Morin, X., Cremer, H., Goridis, C. and Brunet, J.-F. (1999). The homeobox gene Phox2b is essential for the development of autonomic neural crest derivatives. Nature 399: 366-370. Perez, S. E., Rebelo, S. and Anderson, D. J. (1999). Early specification of sensory neuron fate revealed by expression and function of neurogenins in the chick embryo. Development 126: 1715-1728. Pinon, L. G., Minichiello, L., Klein, R. and Davies, A. M. (1996). Timing of neuronal death in trkA, trkB and trkC mutant embryos reveals developmental changes in sensory neuron dependence on Trk signalling. Development 122: 3255-3261. Raible, D. W. and Eisen, J. S. (1994). Restriction of neural crest cell fate in the trunk of the embryonic zebrafish. Development 120: 495-503. Saito, T., Greenwood, A., Sun, Q. and Anderson, D. J. (1995). Identification by differential RT-PCR of a novel paired homeodomain protein specifically expressed in sensory neurons and a subset of their CNS targets. Mol. Cell. Neurosci. 6: 280-292. Schecterson, L. C. and Bothwell, M. (1992) . Novel roles for neurotrophins are suggested by BDNF and NT-3 mRNA expression in developing neurons. Neuron 9: 449-463 . Schweizer, Ayer-LeLievre and Douarin, L. (1983). Restrictions of developmental capacities in the dorsal root ganglia during the course of development. Cell Differentiation 13: 191-200. Scott, S. A. (1992) . Sensory neurons: Diversity, development and plasticity. Oxford University Press, Oxford. Serbedzija, G. N., Bronner-Fraser, M. and Fraser, S. E. (1989). A vital dye analysis of the timing and pathways of neural crest cell migration. Development 106: 809-819. 36 Serbedzija, G. N., Fraser, S. E. and Bronner-Fraser, M. (1990). Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labeling. Development 108: 605-612. Sharma, K., Korade, Z. and Frank, E. (1995). Late-migrating neuroepithelial cells from the spinal-cord differentiate into sensory ganglion-cells and melanocytes. Neuron 14: 143-152. Sieber-Blum, M. (1989). Commitment of neural crest cells to the sensory neuron lineage. Science 243: 1608-1610. Sieber-Blum, M. (1990). Mechanisms of Neural Crest Diversification. In: Comments Developmental Neurobiology , Vol. 1 . Gordon and Breach Science Publishers, SA, Great Britain. pp. 225-249. Sieber-Blum, M. (1991). Role of the neurotrophic factors BDNF and NGF in the commitment of pluripotent neural crest cells. Neuron 6: 949-955. Silos-Santiago, I., Molliver, D. C., Ozaki, S., Smeyne, R. J., Fagan, A. M., Barbacid, M. and Snider, W. D. (1995). Non-TrkA-expressing small DRG neurons are lost in TrkA deficient mice. J Neurosci 15: 5929-5942. Slack, J.M. W. (1991). From egg to embryo. Cambridge University Press, Cambridge. Smeyne, R. J., Klein, R., Schnapp, A., Long, L. K., Bryant, S., Lewin, A., Lira, S. A. and Barbacid, M. (1994). Severe sensory and sympathetic neuropathies in mice carrying a disrupted Trk/NGF receptor gene [see comments] . Nature 368: 246249. Snider, W. D. (1994). Functions of the neurotrophins during nervous system development - what the knockouts are teaching us. Cell 77: 627-638. Snider, W. D. and Silos-Santiago, I. (1996). Dorsal root ganglion neurons require functional neurotrophin receptors for survival during development. Philos Trans R Soc Lond B Biol Sci 351: 395-403. 37 Sommer, L., Ma, Q. and Anderson, D. J. (1996). neurogenins, a novel family of atonal-related bHLH transcription factors, are putative mammalian neuronal determination genes that reveal progenitor cell heterogeneity in the developing CNS and PNS. Mal. Cell. Neurosci. 8: 221-241. Srivastava, D., Cserjesi, P. and Olson, E. N. (1995). A subclass of bHLH proteins required for cardiac morphogenesis. Science 270: 1995-1999. Stucky, C. L. and Lewin, G. R. (1999). Isolectin B(4)-positive and -negative nociceptors are functionally distinct. J Neurosci 19: 6497-6505. Valarche, I., Tissier-Seta, J.-P., Hirsch, M.-R., Martinez, S., Goridis, C. and Brunet, J.-F. (1993). The mouse homeodomain protein Phox2 regulates NCAM promoter activity in concert with Cux/CDP and is a putative determinant of neurotransmitter phenotype. Development 119: 881-896. Weston, J. A. (1963). A radiographic analysis of the migration and localization of trunk neural crest cells in the chick. Devel. Biol. 6: 279-310. Weston, J. A. and Butler, S. L. (1966). Temporal factors affecting localization of neural crest cells in the chicken embryo. Dev. Biol. 14: 246-266. White, P. M. and Anderson, D. J. (1999) . In vivo transplantation of mammalian neural crest cells into chick hosts reveals a new autonomic sublineage restriction. Development 126: 4351-63. Wright, D. E. and Snider, W. D. (1995) . Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. J. Comp. Neural. 351 : 329-338. Wright, E. M., Vogel, K. S. and Davies, A. M. (1992). Neurotrophic factors promote the maturation of developing sensory neruons before they become dependent on these factors for survival. Neuron 9: 1-20. 38 Xiang, M. Q., Zhou, L. J., Macke, J.P., Yoshioka, T., Hendry, S. H. C., Eddy, R. L., Shows, T. B. and Nathans, J. (1995). The Bm-3 family of POU-domain factors - primary structure, binding-specificity, and expression in subsets of retinal ganglion-cells and somatosensory neurons. J. Neurosci. 15: 4762-4785. Ziller, C., Dupin, E., Brazeau, P., Paulin, D. and Le Douarin, N. M. (1983). Early segregation of a neural precursor cell line in the neural crest as revealed by culture in a chemically defined medium. Cell 32: 627-638. Ziller, C., Fauquet, M., Kalcheim, C., Smith, J. and Le Douarin, N. M. (1987). Cell lineages in peripheral nervous system ontogeny: medium-induced modulation of neuronal phenotypic expression in neural crest cultures. Developmental Biology 120: 101-111. 39 Chapter 2 Identification of dividing, determined sensory neuron precursors in the mammalian neural crest 40 3545 Devetopmen1 126 . 3545-355911999 ) Prin1ed 1n Great Britain© The Company ot B101og1sts L1m 1ted 1999 DEV9668 Identification of dividing, determined sensory neuron precursors in the mammalian neural crest Amy L. Greenwood1, Eric E. Turner 2 and David J. Anderson 1 ·• ' Division of Biology 216-76 . Howard Hughes Med ical Institute. California Institute of Technology. Pasadena . CA 91125 . USA 2 Department of Psychiatry. 0603. University of California. San Diego. San Diego VA Medical Center. La Jolla. CA 92093 . USA ·Author for correspondence (e-mail : mancusog @cal1ech.edu) Accepted 28 May: publ1shea on WWW 19 July 1999 SUMMARY Sensory and autonomic neurons of the vertebrate peripheral ncnous system arc dcrh·ed from the neural crest. Herc we use the expression of lineagc-specifir transcription !'actors as a means to idcntif) neuronal suhtypcs that de\'clop in rat neural crest cultures grown in a defined medium . Sensory neurons. identified h) expression of the POL-domain transcription factor Brn3.0. de\'elop from di\'iding precursors that differentiate within .:? days following emigration from the neural tuhc . i\lost of these precursors generate sensory neurons nen when challenged with ll\lP.:?. a factor that induces autonomic ncurogenesis in many other cell s in the explants. \loreonr. ll\lP2 rails to prevent expression or the sensor)• specific hasic helix-loop-helix ibHLH I transcription factors 11euro~e11i11 I . 11e11roge11i112 and neuroD. although it induces expression of the autonomic-specific hHLH factor \1..1.SH I and the paired homeodomain factor Phox2a in other cell s. These data suggest that there arc mitoticall) active precursors in the mammalian neural crest that can gl'ncrall' sensor~ neurons cYcn in the presence of a strong autonomir-inducing cue. Further charartcrization of the neurons generated front such precursors indicates that. under these culture conditions. they exhibit a proprioccpth·c and/or mechanosensory. hut not nociccpti\'e . phenotype. Such precursors may therefore correspond to a lineally I Frank. E. and Sanes . .I . 1199 I I 1Jerelop111e11t 111 . 893-90111 and genetically (:\la. Q .. Fode. C.. Guillemot. F. and Anderson . D. J. i 19991 Genes Dn. I.,. in press I distinct suhsct of earl) -differentiating precursors of large-diameter sensory neurons identified in \'irn. INTRODUCTION 1Bronnt'r-Fr~i.--1,;r ~inJ Fra:,.1..'f. l 9~S . i 9X9: Fra,r.:r anJ Bn111111..·r- i,.;, ,,:~ \ \llr1.1' · :-..-c:ur:tl .... r~,l. s~ll:--llr~ 111.: Ufll:!t' ll t.:,1,. \"t:Ufll:!t'Olll. Hrn- :;.11, B\11'2 . '.\T-_;_ R,1 Fra,cr. llJlJl I and in ,·i1r<> 1S1ebcr-B lum . llJXlJ: Duff cI al. .-\ funJamcntal ~uc:-.tion in tlc:,·elopment i:-- ho,, a di,·c:rse arra: of c..liffcrc:n11att:J ct.:111ypc:-- can be: prnJw.:t:J from ~1 multiplHL'nt populatiPn o!' pnll,!t.:llitl1 r 1..:el\.,. In ,·c-rtehratt: :--. thi :-- i--:--uc ha:,., ht.:t.:n intt.:11 :-- 1, d: im·t.:,tig~1ti.:J i n th e neura l L·rc,t. .-\ lthnugh ;.i Lir~ t.: hoJ, of e, i Jt.: 11 1.:L' :--uppnrt :-- the iJl!a th ~11 thL' n1.:ural crc.::--t cl11~t~1in:-- ·mu ltiputt.:nl progL·ni1or, who,(' d 1niL'(' of fatt.: i:-t.ktc"rmint.:J tn cn,·irtH1mcntal si!.!nab. tht.:rt.: i:-- ~d sll L·,· iJt.: 1h.:i.: 1hal 1hc nc.:ura·I crc:--t ....:n nt;.i 11h ll) f ·c,cntu~.d l: gt.:nt.:ri.llL':•.J :,,.u h ,t.:t:-of p1"t:-1.·ur:--urs ,\·ho:,.r.: fate':,. appc.:ar rr.::,.trictcd ! for rl.'\ il.'\\.., :,.l.'1..' Le.: Douarin. llJ X6: AnJcrs<rn. l<JXlJ : Sid,er-Blu111. l <J<JIJ: Le Douarin el al .. 1')'11 . ~lan" ich and \\'c , lon. l<JlJI: \\·c,Ion. 199 1: ..\ 111.kr:,.u n. 199]: Bronncr-Frasc.:r. llJlJ3: L e.: DlHl~lflll l.'l ~d.. 199--L .-\nJc.: r-.un. 1997 l. I t i, thc.:rc.:fur1.."' important h ) unJl.'r-. tanJ w hen parti1..·tilar nc.:u ra l crc.:,q cdl r~11r.:, h1..•cnm1..· 1.kk·rmi n1..·d in rc.:latiun to othi.:r,. anJ him lin i..·:.t~L' ,c~rc.:gatlllll j:,. n:gulatl.'d duri n~ 111.: ural 1..-rcs1 1.k\1,;lnpmL'nt . On1..· -.l.'tli n~ in whirh ,ud1 i:,.:,.ui::-- h~1,·c bc.:en aJJrc.:,:,.r.:J i:,. in th e ~pt!c itl1..·a;il)ll o f the: 1wo m~1jnr nc.:urun ~il d~i-. ... c.: ., 1)1° til l· peripheral ncrrnu , ",tclll I Pi'<S I. the ,cIhllr\ anJ the at1 tlHHl mi1..· 1Le DlHlarin. I t.JS6: Sic.:hc.:r-Blum c.: t al. . \ LJlJ]: Li..· Douarin cl al .. l<JlJ-.I. Clonal analyses of neural crcsi in ,·i,o llJl)l I h~l\ l.' iJc ntillcJ indi,·iJua l rrc.: 1..·uf;',l) [" ;<, lh~1t L'Jfl gi,·1.• ri-. 1. ' to hnlh ,1.;11:,.nr: anJ aut1)11om 11. · nc.:uron -. . Such nh:,.1.;na t i1111,. h1m·c,t:r. Ju nnt pr1.;cluJi: the L'\t:,.k•ni..·i: of "L' !Xlfatl.' ..,i_:n,or: ~111J autunumic pri:cursor-.. Prl.'\"inu, rt.:,ult, from had~-gra!"tin~ :,.luJic.:, in a, 1an i:mh rylh ha ,l.' ,uggt:,tc J lhl.' t'\hll.'ll1..·1..· Pl Ji,·it.lin~ rrr.:cur.,1lr, 1hat i..:an co ntrihut c.: lt) thl.' autunum 1r hu1 lll)t si: 11 .,1 )f \ J1.:.•n,·ati,c.:, of thl.' nt:ural l'rt::,. t. !LI..' L1c\rt.: l.'t ~ti .. l lJi\tl: S1..·h"t:.izl.'r l.' I al.. llJX.': L1..· Dou~inn. l lJS61. \\"hi:lht:r th1. · nc-ural L'f(':,.l 1..·untai11:,. Ji,·iJing pn)~t'llltnr, lh~ll i.."~1ll ~i:111.;r;.1(1..' ,l.'!hUr: ncurnn, t.:,l.'n in the pr1..·,cnl·l.' oi" autonomic-i11Ju1..·111~ 1..·ul.'", i, rw t knnwn. T11 ht.:ttl.'r unJc.:r ... tanJ !ht: ml.'d1a11i:,.llb tha t rc~ulatt.: the :,.t: !.!rC!.! atitlll of th1..· ~en:,.or, lini:a~c. it i, ni:ct: "Jf\ lil hl' ahk l l 1 iJ~nti·f~ ;md m;.rnipulak· • ,1..· 11--0;: pr1..·c ur,or,. L't1~ilkn~l.' them with ~1Ull)fh)llli1.·-1nJu1..·in~ , 1~n~d, ~1nJ dc1c.:rmin1..· " hi:thcr th1..·: ~i:neratc :,.c.: 11:,.on or aCt 1H1~1mi 1. · neuron ,. Rt:ccntl,. ,c, l.'Ld ; ra,i:,.c ription fa L~or, ha,1..· hct.:n 1..·h~1r;.1c tt:n1.l.'J that c~n hi: u,i:d tn i1.kntih :,.t:1bon neuro n:,. anJ/or their pr1.;cur:-.nr, anJ Ji,tingui:,.h 1hcm ·from autonomic neuron:,. and/or tht:ir p rr.:1..·u r,ur,. 8a:,.i1..· hdi , -ln1lp-hdi , t hHLH I 1ran,1..·ri ptinn factor, :,. uch J :,. thl' 11e11ro!:e11i11., 111:.,:ns: G raJ wohl l.'t al. . \i.JlJ6: 41 3546 A. L. Greenwood. E. E. Turner and D. J. Anderson Ma et al.. 1996: Sommer cl al.. 19961 anJ 11,·11mD t La et al .. 199:i 1. a., well a, the POLi-J<>main 1ranscrir1ion factor Bm-3.0 I FcJ1so\'a anJ Turner. 199:i: Xiang cl al .. I lJlJ:i I arc srcc·ilicall) expressed in thi.: sensory lineage. Other transcription f~H.:tor~ arc cxpresscU by autonomic hut nol sensory celb. including the bHLH factor i\1.-\SH I tJohn,on et al.. llJlJll: Shah cl al .. 1'!%1 anJ the paireJ homeoJomain factor Phox2a t Valarch0 ct ,ii .. 1993 : Lo et al .. 19981. Mcmhcr., of the h,mc morrhngcnetic rrotcin t BMP1 ,uhfamily. such a, Bi\lP2 anJ B\IP.J. ha,·c hccn iJe111i1ieJ a, auto1Hlmi1..: nc:unm-im.iuL·ing signab both in \·iro anJ in ,·itro. Such factors can induci.: autonumic diff~rcntiation in ma:--s cultures of a,·ian neural crest I Rci"man ct al .. I 996: Yark, anJ Maxwell. 19961 anJ in d<>nal nrlture, of mammalian neural crest stem cdb tShah et al .. 1996: Shah anJ .-\nJcr.,on. 1997: Morrison et al.. 19991. In the laller S\Sk'lll. B\IP2 \\·a, L·or1Ji1inri.... thL'\ \\cr,.: \\a,ht:J l'.\ ti.:n ... i\~]\ in JL·li11L·d 1111,.:d1u111 hL·i'11rL· shown to :.u..-t inslrui.:ti\l~I~. promnling autonomic nc:urogc:n..::--i:-- µ,:!iilll from muhipntc:nt neural crest L·c:lb at 1hc c.\pen:--i.: nl' 01hi.:r fati:;-,. O,·c:n:xpre:-.:-.ion of B\IP-+ in n,·n alsn i1H..lucc:-- ~1uto11omi . .: marker, in neural crest c·ell, tRei"man et al .. I99/n 8\11'2/-l arL· liki:l y ln be inJuct.:r" of tht.: ~tllllHlomic fat 1, .• in \ ·in1. a., tilt.': are e.\prt.::-..,1.;J h: ti:,. _ ..:.uc., ncighhuring autnnnrnic ~a11gli~1 1:-.ud1 '" the Jors;rl anna: Rei,,111;,11 ct ;ii.. ltJtJ<,: Sh;rh ct al .. l'J'-J(, I and these ti ... ,ui::-. can mimii...: tht.: inJuctin.· i:ffect:-. nf B\IP, Pll 11t:ur;.il crest cell:- 1Gron:., anJ .-\11Uersl1n. 1996: R1,;1,,m~1n t.:t ;11.. l'-J'-J61 in a m;inncr 1h;r1 i, inhihitahk h: the B\IP2/-l-.1111;i~o11i,1 noggin 1Zimmcrma11 ct al.. 1996: A. Gro\(', anJ D. J. .-\ .. unpuhlisheJ J;rt;r I. \\"1.; ha,\~ now used transcription factnr markt:r-. 10 ,tuJ: the Lk\·clopmt!nl nf ,t:nsory nt:urnns in 1..:ulturc, of rat neur~d cn:,t. We h,l\·c formula1ed a fulh JclineJ .:ulwre meJium that p1.;rmit:-. the t.lifferentiatinn pf ., e1hor: neuron, from prolikr;.iting precursors that migrati: from neural tuh1..· e\planh . ...\utonnmi1: neuron:- Jo not Lh:-\"t:·lop unLkr tht:,L' conJition, unkss the, ar,· inJuccJ I<> Jilfrre111i;i1c h,· ;rJJition ot' B\IP2 . Stri king I). B\IP2 Joe, 11<>1 "rrcar 10 pre,en1 the Jilfrrc111ia1i,H1 of mos! cJi,·itling :-.1.!n:-.ory precur~or ... or the c.\prc:-,,ion ol ~c1i...ury precursor marker., . Tht::sc u~1la suggc., t th;tt the rnigraling nc-ural cn: ... t inL'lutlc, a ,uhpupul~11inn uf di, icJing prt.:1.:ur,ur, that arc. opcralion;.ill;. committed tn ~1 ,cIhPr; Lite with rc.,pcL·t lll the autonomic-inducing ... ign:ll B\1P2. Interesting ly. th1,.;sc precur:--urs gin:- ri,c Ill scnsur; I1t..:uroI1-. th;ll c,hihit ~1 prnpriocc-pti,·L~ anJ/or mcd1;11lllrL·L'L'Pli\'\..:. ra1hcr than ;1 noLi \,.·epti,c phcno1ypc. Thc; may thcrc!'orc corrc,p1u1J tn ~1 Ji,1inl:t ,uh:-,ct ni' se11:-,nry prcL'lIr,or ... prc, i,n1,I; iJcntificJ ti; rctnn iral lint.:agc lraL·ing in chich. crnhr; o, 1 Fr;1nh. and Sanc,. ] LJLJ I, anJ nHlft.: rt.'L't.:lltl; h; gcnctiL· anal; :-,c, in mice I \Li ct al .. I 'J'l'J 1. pb1111,:! . E\pl:1;11cJ ncur:11 lllht.', \\i.:rL· i,lati.:d in .'5 mm Ji:1111~1L·r Corn in~ Ji . . h1.·, on a lihr1111i.:~tin I I3Tl 1 ,uh ... tratum ,rnd ~I'll\\ n ;It .' 7 C in 5r; CO:/lJ5 r ; ;1i r. \Vhi..:ri.: ,tati.:J. the ni.:ural tuh~ \\ ;1, ~rt:nHwcJ from thc L'UllurL· :.if1er 2~ huur, witll :i tUn,:!,tcn nct.'Jk. hut othcn\ i ,1..· thL' C:\plJlll \\;1, ldt in1.1,.:l Media and additives Defined medium ThL· ri.:L·irL· t'nr JctincJ rnl'Jium I D\t 1 \\;i, nlliJilit.:J from th;11 J1..·,L-rihi.:J pr1.:.·, iou,I~ 1St~mpk anJ :\nJt."r,011. I l)l)~ J and i, ,1, i'nlhm , · to L- l )CO~ , H;1,,·rut ;.inJ Patti:r..;un. [ lJ7lJ! ;1J<l I 00 u~/rnl tr:111'r'L·rrin !C;ilh1t1L·ht.'llll. 5 P,:!lml in ... ulin 1Si,:!lllJl. 16 pg/ml pu.tr~ ... L·inL'. 1Sig111;11. ~() 11.\I rrogL' ... tL'roni.: 1Sig111;1J. 30 n\1 ., ckninti... :iciJ 1Si,:!ma1 I mg / ml h1n·inL' ,1..·rum alhumin. (t..'r~ ..;1;illi zcJ. GihL·o/BRLi. 3lJ p,:!lrnl dt.:\a111i..:tha,1111L' 1Sigma1.) pg/1111 u-J- l-t11cophL'n1I I Si:::111,11. h.' ~tg / ml ~-h:Jrn,: hu1~ rt.Ill· r Si,:!111;11. hiillin (Sig11u1. gl:ci.:rnl. ~:i ng/ml l.·ohal t L'11loriJi.: ( S1g111;1 l. 10 rl,:! / ml ukiL· aciJ 1Si:::ma1. .' .(1 l 111~ / 111 ! I ()(J 11,:!lml u-rni.:latHlL·~ IL'• ... 11111ub1in,:!-ht1rn1nr1L· I S1,:!111;11. 1(/ ng/ml pn"',taglandin [ I 1.-\lJriL·h 1. l:h.,11.: tihn1hb,1 ,:!fll\\ 111 L1L·11 H thl"GF: Ill 11,:!l llll. L:811 \\;1, lllL:luJL·J in thL' 111i.:d1u111 t'11r 1l1L· 1ir,1 ~ J;1.\,. bul lllll th1.'l'L':1t'tLT \\'hcrL' indiL·;1ti.:d. D.\I \\;1, ,urrk.'lllL'lltL'd \\ ith rL·u1111hi11:1111 hu111;1n B.\lP~ 1)() 11:-:/ml. ,::if1 frnrn GL'lll..' lll, !1i...tituh:1 Undefined medium L· nLklineJ 111L'Jiu111 \\a, ;tl,11 moJilii.:J fnHll th:it dL·,uihL·J prL'\ inu,I~ 1S11.:tnpk :rnd :\nJa,011. I l)lJ~ 1. It L't1111;1ini.:d lhL· ,;1111L· h:1,:d L lllllpt11lL' llh ;1, lhi.: 1)\1 JL'., (rihi.:J i.lPlHL' plu, hrGF 1 ➔ 11g / 111I. Rl\:I) 0 S~,[L' llb l rh.:. l. i.:r1Ji.:nn;d ,:!l'!l\\(h LKt11r tEGr: 1110 11:::l ml. S1g111.11 . rL'lllltll,.' .1i..:1J ! 3.'- 11:..! / 1111. Si:_!111;1 I. IIL'r\L' ~l'tl\\ th (;1l·t11r (\'Gr:,',(} 11:..: / 111I. L'BIJ. hr;11n-Ji.:rl\L·:1 ~f(l\\li1 f;1(t11r tBl)\'r: : ~) 11g / ml1. 111..·un1tr11r111n· 1 \T.' : ~5 11,:!l ml. RD,r: and ,T.~ hoth pr1n idt.·d h~ Dr G .'. Y:1n,.:Pp11ul11-- at RL',:!L'lll"rtnl Ph:1rn1aL'L'lltic;d,1. kuk.i:mi:1 1111lih1111r~ L11..·t11r 1Llt:: ~_.:; ll:.,! / 111I. R&D s,,ll' l11' 1111.:.1 ;1nJ 1()•; d1id, L'lllhn,i i.:,1raL·1 1CEE1 !'Pr -tllL· lir,t ~ ➔ lll.1ur., . 0 . 1', CEE 1hai.:afti.:r. • lmmunocytochemistry CulturL'' \\L'rL' ti,L·d Ill ➔', rar;11°11rn1;1l l kh~1.k :ind thL'll rn11.·~, ... 1..·d t'11r ,in:::k 1ir Jouhk Jrll1hPd~ lah~lin,:! 11r 111 ,itu h~ hnJit.:11111111,L'L' hL•h1\\ 1 fulh1\\L'd b~ ;1111ihnd~ l;1hi.:lin~ . A11 tib11J~-,111llh11J) J11uhk l:lhL·l111::: i11\Phi.:J Ji.:li.:l·tion 111 a1111~~11' u,in~ , i 0.-\[3 and 0.-\B ,uh,tr;11L·, a, Jl·,L'rib~d rri..:,iPu,I~ 1Sh-:1h t.'t :ii.-. ll)l) ➔ i. Fin;dl). L'UlllH'L'' \\L'rl· i111..·uh:llL'J \\ i1h Ill p:::11111 !) ..\Pl f11r I~ 111i11u1~, ;11 ftltllll ll'lllrL'f:lllll'L' \!I iJL·1111t~ ntllL·r nu,: ki :111J 1ln1, thL· L'\lL'lll 111° tllL' nu1~r11\, 111 f r P111 th,: Pnmar> antihPJ~ 111L"uh;11H1n, \\L'rL· L·;1rri~d out ;11 ➔ (' ,i\L'fll !;:'.lll . .-\t'li1111~-puntiL·J r;1hhi1 ;1111i-Brn -3JJ ;1111 1hod~ t I ::"!I() : FL·d1,P,.1 ;irhl TurrJL·r. I lJ'J:, 1 :111d ll:1111'1L'r a1111-DR(i 1 I Ii> hnJ11111:1 ,upL·m.1L.1JH t S.1111 i t.'l :d .. 11Jl);" 1 \\l'fL• u,i.:J tti JL·IL·L·t 'L·1i--,1r~ 11L•ur1,11, . .\lt11111,:l1111. d ;111tih11J1~, 111 SCG 111 1ll\ hriJ,1111;1 ,u1ii.:rn.11a111: L. LP .11;d D. J :\ unpuhli,hL·d J:1l.1 l. 11~un.11il;i111l'lll lhfl I I :251l : ~Fl<,O. 1.· h1n1.· '\'\ I S. Si,:!m;11. 11r \~u, 1 I :.'-00. Chi.:11111..·11111 \\L rL' u,1.·d ;1, :!L'llL'r;d IIL'lln111.i l 111:irk.L·r, _ PL'.riphl·r:11 llL'lln11 1, \\l'rL· 1JL·1111liL·J u,111~ the ,.:1 1111h i 11;1 11,111 ,,1 Hrn-.'.O i.:.\prL'"lllll ;111J rl 1]~ L'lt111al an11-pt:rtphai11 I I -~II! t( J: Ch1.·11111..·11n 1 Pr 111111111d11rul ;11111-1.·· RET I h~ hriJ11111;1 ,up~rna1;1111: l.11 ;111J _.\11d1..T,1111. 1q 1J_.:; 1;1111ih1 ,Jii.:, . K:1hhi1 p1 ii~ i..:h111:d :11111-PhP,2:1 :1111ih11d~ , 1::-rn u ,. ,::11 1 t'n111 1 l)r, . C. G11nd1, ;1nJ J. F. HrunL't: Tl\LT,111 ~I a l .. 11J1Jh 1 \\;1, u,1.·~ l MATERIALS AND METHODS Rat neural tube explant cultures \i.:ural tub~, \\~rL· j,lil:IIL'd frn11 1 tlh: trunk. ll'\1.:I nt' ElO . .' Spr.1:::LIL'• tn 1d~111if\ ;1u111n11111 i1. · 1.·L'lk .-\ull111111111,,: rrL·1.·ur"1r, \\L' l'L' ,1;1111L·d u,111~ :1mi-.\ t.-\SH I ll\ hri.d11111 ;1 ,upcrn;H:1111 , Li 1 L·t :d .. I \llJ I J. ThL' 111111111L·li in.ii D;n\ k\ rah ;1, Ji.:,LTih~d rrL'\ iPu,1~ 1Sti.:mpk and .·\nJL·r,1111. I 1N~ i :int1ii11J> RTll-·11::-,1111 : ::: i f t t'nim Dr J. .'.: . \\,1,1d : L;i\\,1H1 L't :d .. llJS-l 1 \\ 'o~th1n~tin~, ;ind Ji,ra.-.i.: (0.77 u/ 111I. -Ciih1..·111 in H:1111--·, h;il;11i..:i.:J ., J.ll anJ r;1hhit pPl~..:h1n;tl :.i111i - ERS 1 ;1n11h,1d~ 1 I I ~.tlllll : ~il l lr11111 Dr T Lin 1,,;1 al. . I 1NS, \\L'fL' u,L·J ;1, 111;1rl--l·r, 11 f till· lar:::i.:. pn1priPL·~pl1\l' ,uht~ PL' pf 'L'i1,or:- rlL'UrPl1' . R;1hl">1t p1d;1..·lt 1n:d ,olution ·, HBSS l . ThL' ti.....,ut.' \\ a, i111..·ub:i1L·d f11r 5 mim11L·, in IL L'•1..·nkl ;intih11dii.:~ 111 Sul"t;ini..:L· P 1SP: I 11)()( ): l11L· ,1;1r1. L·;1ki111n111 :!l.'IIL' rL·LltL·J i.;nL\nll' so lution ,,ith tntur;1tion. thi.:11 i1lL·ub;11L·d ll1r 11) m111u1L·, .11 pcptid~ 1CGRP : Wil;ll 11.:mp~r•llur~. thi.:n ,:!~1111~ triturat~d furtllL·r until tllL' n~ur,d tuhL'' ( 1:)0(): ~1r1 frt1lll Dr L \\L'rL· frc:t' pf nthL·r ti ...... uc, \\ .hen 11L•ur;JI tll1'L'' \\L' rL· ,:!rtl\\ n 111 Ji:11111..·d mark.i.:r, ~tlt' tilt.' :,..11\;dl JlllL'l1.."qHi\1.:.' ,Llbllr~· ,uht~ ri..: hul \~ 1th lhL' t:nt\ matiL·;1lh rl1llt1\\ ill~ m.:atLd \\ i~th n111di tiL·;1t111n ., : 1.25 md111l trunk. 'L'L'IIPl1' \\ l"rl' i..:ulb:..!L·na,1..· 1 I /2 u/ 111~. 0 fr,,c:11: I 10110 : P~n111,ul;1 L1hl1r:11t1nL·, . ln 1.· . 1 ;1nJ Tri-...-\ Rcid1ard1: ll:1r, L'l JI.. llJ 1 J➔ 1 \\1.'l'L' u,1.·d .1, 42 Determination of the sensory lineage in vitro 3547 All secondary antibodies were used at a dilution of I :200 and incubated for I hour at room temperature. Rabbit polyclonal antibodies were t.letc:t.:ted using an HRP-conjugare<l. goat ant i-rabbit secondary antibody purchased from Vector Labs. Inc.. The monoclonal antibodies to SCG 10. NF! 60 and MASH I were detected with an HRP-conjugatcd. goat anti-moust! lgG secondary from Chem icon. Hamster monoc lonal antibodies 10 c-RET and DRG 11 were dctecteU by a biotin ylatl!d goat anti-mouse lgG secondary antibody (Jackson Laboratories. Inc.) fol lowed by the ABC amplilication system (Wctastain ABC kit . Vector Labs. Inc.). In situ hybridization In situ hybridization on cxplants was performl!<l using a modification of procedures dc\'eloped by Dr R. Li (Signal Pharmaceutica ls. personal communication) and Dr D. Henrique ( Henrique ct al.. I 995 ). Briefly. the explants were pretreated with acetic anhydride and 0.2 N HCI then prehybridized fur I hour at room 1empera1,irc. Hybrid izat ion in 1-2 µg/ ml of digoxigenin (DIG)-labekd probe was carried out at 65°C overnight. After the hybridization. cxplants were washed at high stringency (0.2x SSC at 65°C for I hourJ. incubated with alkaline phosph,1t..ise (AP)-conjugatt!d anti-<ligoxigenin antibody (BOt!hringcr) and the signal developed using NBT/BCIP reagents. A more detailed protocol is available upon request. The probes used in this study include rut 11g11 I (Ma et al.. 19961. mouse 11g112 (Sommer et al .. 1996 1 and mouse lleHmD (Lee ct al. . 1995). For in situ antibody double labeling. cxplants were processec.l first for in situ detection of mRNA. then for immunocytochemistry as c.lcscribed above . BrdU analysis To detenni111.! the mitotic statu s of sensory precursors in vitro. ex plants were grown in OM and supplemented with 175 nM BrdU at the time of plating (day 0) or after I. 2 or 3 days of culture. After initial crest migration (24 hours). cxplants were motlilkd by remo ving the neural tube from culture with a tun!.!sten needle. Cultures were !.!rown in the presence of BrdU for 2-1 • hours. then fixed and p;;,ccsscd for fluorescent Brn-3.0/BrdU double labeling . Rabbit anti-Brn-3.0 antibody (see above) was detected using a FITC-conjugatcd goat antirabbit seconc.lary from Jackson Laboratories. After Brn-3.0 staining. BrdU incorporation was detected as described previously (Novitch el al .. 19961 using a mouse anti-BrdU primary antibody 11: I 00. Caltagl anc.l Cy5-,onjuga1ed goat anti -m ouse secondary antibody ( I :200. Jackson). Sti.linin!.! was observed using u ➔ IO Ziess Axiovert LSM confocal micros~ope . Doubk-pnsiti;e nuclei were scored by identifying individual Brn -3.0+ cells in one channel (green) and checking for an identically shaped. BrdU• nucleus in the other channel (red). In cases where it was not possible to clearly determine if a Brn-3 .0• cell had incorporated BrdU. these cells were omitted from lhc ;111alysis: this situation occurred most often when thi: ,ells were tightly clustered in cxplants at day I . Trophic support analysis Neural crest explants were cultured in OM supplemented with 10 ng/ml bFGF for 2 days. On day 3. OM l without bFGF) was supplememed with growth factors sing ly or in combination at the following concentrations: NGF ( JOO ng/mll. BDNF (50 ng/ml). NT3 (50 ng/ml ) and/or LIF (50 ng/ml). The medium and factors were replac";,d e,·ery 2-3 days u111il -the es plants were fixed at day 8 and prm:csscd for Brn-3 .0 antibody swining. The numbt!r of Brn-3.0• <.:ells in the dorsal outgrowth \vas scored for each explant. exc lud ing ct!lls lightly associated with the neurul tube . RESULTS Both sensory and autonomic neurons are generated in rat neural tube explant cultures To characterize the types or neurons that differentiated in rat Fig. I. Two populations or neurons devc:lop in rat neural tube explant cu ltu res. The view shown includes the dorsal ll(!Ural tube INT. top) and neural crest outgrowth (bi:low dashed li~1e1 of an explanl grown for 8 days in CEEcontaining ml!dium plus NGF. BDNF. NT3 and LIF. The explant is stained for both Brn-3 .0 (dark blue. nuclear) and SCG 10 (brown. cytop lasmi,) via immunocytochemistry. Brn-3.o+/SCG JO+ cells are located in 1he outgrowth proximal to the neural tube and have large ce ll bodies (region X). Brn-J.0-/SCG JO+ cells an! sma ller and often found in clusters in the periphery of the out growth ( rt.!gion X'). Th; punctatc ;taining in this latter region represents perinuclear SCG I O and not nuclear Brn -3.0 staining: see Fig . 2B'-D' for higher magnification images showin~ thut the neurons in this 1;uion do not express Brn-3 .0. ~' ~~/; .'.•,•:·•·' 1 .,. II ~ \ i~:. '. neural tube explant cultu re s. we examined the expression of lineage-specific transcription factor markers in combination with general markers of peripheral neurons (Table I). Both Brn-3.0 . a POU-domain transcription factor. and DRG I I. a paired homeodomain factor. arc expressed by sensory but not autonomic neurons in the PNS in vivo (Fedtsova and Turner. 1995: Saito et al.. 1995). Moreover. Brn-3.0 is genetically important for some aspects of sensory neuron development (McEvilly e( al.. 1996: Xiang et al. . 1996). Although these transcription factors are also expressed by dorsal spinal cord neurons. Brn-3.0• or DRG I I+ periphe ral se nsory neuron s can be distinguished from these CNS neurons by their expression 43 3548 A. L. Greenwood, E. E. Turner and D. J. Anderson Brn-3.0 I Phox2a DRG11/ SCG10 Brn-3.0 I Peripherin .. • ~- • .... . •. ,..: .. ..' ·•.. -· t .. , .. I •B Brn-3.0 I c-RET ••• C .. D . . .... • .J,· . " It ... A' C' .•"'f ..,,r •t . 1 .•, Fig. 2. Neu ral tube explanl cu ltures contain periphl.!ral st!nsory an<l autonomic neurons. Double-immunrn.:ytrn.:ht:mistry shows that largc:diameter cel ls in the net1ral tube-proximal region of the outgrowth (Fig . I. region X) express the sensory markers DRG 11 (Al and Brn-3.0 (BD: dark blue. nuclear). They also express SCG IO (AJ. periphcrin CC) and c-RET (D: brown. cytoplasmic I. Smaller-diameter cells in the same cultures (Fig. I . region X') express neither DRG 11 (A'J or nor Brn-3 .0 (B'.C'.D'). but do express SCG 10 (A'). peripherin (C'). c-RET (D') and the autonomic marker Phox2a (B': brown. nuclear). of peripherin (Parysek and Goldman. 1988) or c-RET (Pachn is et al.. 1993). Peripheral autonomic neurons can be identilied by their expression of the paired homeodomain factor Phox2a (Valarche et al.. 1993: Tiveron et al.. I 996) and either peripherin or c-RET. as well as by their lack of sensory markers (Table I). We lirst app lied this array of markers to neural tube explants grown for 8 days in standard CEE-containing ('undelined·) medium supplemented wi th the neurotrophic factors NGF. LIF. BDNF and NT3 (see Materials and Methods). Two populations of morphologically and antigcnically distinct neurons were evident in the outgrowth of these cultures. Brn-3.0•tSCG Io• neurons had large somata and tended to be located proximal 10 the neural tube (Fig. IX). Neurons in chis locacion also expressed DRG 11 (Fig. 2A), peripherin (Fig. 2C) and c-RET (Fig. 2D). We conclude that che large. neural tube-proximal neurons are sensory in nacure. It was not possible to double- Table l. Marker combinations identify peripheral sensory and autonomic neurons Brn-3.0·' DRG I I h Phox2:.f c-RETJ Pcriphcrinr PNS sensory nt!urons autonomk nt!urons + + + + + + + + + CNS dnr.;al neural tuhi! prt!g:.mg.lionic autonomic neurons motor neurons label suc h neurons for Brn-3.0 and Phox2a (Fig. 28.B'J because the chromogen used to visualize Brn-3.0 labeling obscured that used for Phox2a. Howewr. in cultures labeled only for Phox2a. che neurons wi th large cel l bodies localed proxima l to che neural tube generally did noc express che auconomic marker (daca not shown). A separate population of small er neurons was obse rved scat1ered throughout che ou1grow1h. often found in 1igh1 cluscers in che periphery (Fig. IX'). These smaller neurons did not express Brn-3.0 (Figs IX'. 28'.C'.D') or DRG 11 (Fig. 2A'). but did express Phox2a (Fig . 28'). These small. Phox2a• neurons expressed peripherin (Fig. 2C.C') and c-RET (Fig. 2D.D'). indicacing a peripheral. autonomic phenocype . Thus . the use of specific transcripcion faccor markers indicaces thac neurons of both sensory and auconomic lineages are generated in neural tube exp lant cultures grown in CEE-concaining medium. Qualitac ive ly. many more auconomic than sensory neurons appeared to develo p under chese condicions. It is scrik ing chat the sensory and auconomic neurons in these explancs are located proximal and discal co che neural cube. respeccively. as are sensory and auconomic derivacives in vivo . + + n.<l. 11.<l. + The tahlc summ:.trit.t!S previously published expression d:11:.1 in the ro<lc.:-111 l.!mhryo ,11 tnmk level. (a) Fet.Jtsova ;mt.J Turner. 1995: Xiang ct :ii.. 1995. (bJ Saito ct ,1\.. llJY5. (,,:) V;darchC cl al.. 199J; Tiwron ct .ii.. 1996. (<l) Pachnis cl al.. 199J: Lo ;md An<lcr:,,on. 1995. kl P;1rysck an<l Goh.Iman. 1988. (+) t!xprcssion in tissue in<lkatc<l. (-l no expression in tissue in<licalc<l. I n.d.) 1101 <ll.'.'te!nninc<l. Defined medium permits sensory but not autonomic neurogenesis A previous scudy suggested 1ha1 the differen1ia1i on of sensory neurons may be promoced by growth in delined medium lacking se rum or CEE (Ziller ec al.. I 987). We therefore formulated a de lined medium ( OM) wich the incention of decennining che growch requirements of sensory and autonomic neurons in the absence of CEE (see Material s and Mechodsl. When neural cube ex plants were grown for ➔ days in such DM . greater than 959'r of the neurons in chc ou1~row1h were Brnas determ ined by double labe ling for B~·n-3.0 and NeuN. a pan-neuronal marker (data not shown) . The Brn-3.0+ cel l, had a neuronal morphology. expressed NFI60 (Fig. 3A. inset' :i.o• 44 Determination of the sensory lineage in vitro 3549 Brn-3.0/ NF160 Fig. 4. Sensory precursors different iate into neurons by day :!-3 in Uefine<l medium. (A-Ci Double-labeling for Brn-3.0 (dark blue. nuclear) antl NF! 60 • 4 neura l c.: re sl ou tgrowth outgrowth ( minu s neu-r:.il tube) is cu ltured Fig. 3. Defined medium supports sensory but not autonomic neuronal differentiation. (A. Bi Cells st.tined with an an tibody to Brn-3.0 (dark blue. nuclear) are fo unU in tht: crest ou tgrowth or ex plants grown for -I days in defined med ium. The cultures in (A) have been counter- indicates the average number of Brn-J .O• cells per ex plant at each day. c.Jividetl into Brn3.0•JNF I 60· (blue part of barJ and Brn3.o+/NF 16()+ (brown part of bar) fractions. Note that as the tntal stained \Vith an antibod y to NFl60 (brown filamentous staining}. The number of Brn-3 .0+ dashed line divides the neural tube (NT. top) and the outgrowth (below dashed line ). (A. inset) Brn-3 .0• cells co-express NF160 (brown. cytoplasmic: see also Fig. -1). (Bl Cells in the outgrowth of these explants are not labeled with an antibody to Phox2a (brown. cel ls per explan t irn.:rcascs rrum duy I to day 2. the proporti on of nuclenr). Phox2a sta in ing on BMP2-tre~Jtetl cultures (see also Fig. 6) saves as a positi ve control (inse t in 8 ). U nlabeled l.'t::lls an:: \'isuulizcd with DAPI liuorescem:t! (aqui.l, nuclear). thcsc cells thut express NF 160 im;reascs as wt:11. (n) is the number of cxplants scored from B day 1 • •· co-express NFl60 (A. arrow). When this until Jay 2 {Bl or day .1 (Ci. a majority of BrnJ.O+ cells co-express NFl60. (DJ The graph t • antiucns shows that ·- $ I A ( brown. cy toplusmk l afte; I day of cult ure. few Brn-3.0• ce lls in the -' - ·- • ... •• .... •• - day 2 •• • day i C ~ ~ u. 500 n..S 400 z- +£ t 300 +~ t 200 z Q. +LL +;:; ID two t!xperiments . Error and peripherin (data not shown). and made up 6.6±2.8'7e of cells in the outgrowth. In contrast. autonomic (Brn-J.o-; Phox2a•) cells we re almost never obse rved under these condit ions (0.06±0.12~ of cells in the out~rowth. Fi£. 3B 1. Thus . this DM wax permissive for sensory but not aut;nomic neu ronal differentiation. The sensory neurons in DM cu ltures were not restricted to any particular region along the proximodistal axis of the outgrowth. in contrast to CEEcontaini ng cultures where they were located proximal to the neural tube . Sensory neurons differentiate de nova from migratory precursors It was formal ly possible that the sensory neurons that we observed in v itro did not differentiate de novo. but derived from already-differentiated neurons that simpl y migrated onto the culture substratum. To address this possibility. we compared the expression of Brn-3.0. which ix fo und in both sensory precursors and sensory neurons (Fedtsova and Turner. 1995: Xiang et al. . 1995 ). to that of NF160. an earl y marker of burs represent s,~rn<lar<l deviations. D day 2 day 3 neuronal differentiation. In cult ures 2rown for 2-+ hours in DM . only a minority of the Brn-3.0• cells that had migrated onto the substratum co-expressed NF160 (Fig . 4A.D) . However. when suc h 2-+ hour outgrowth wax cultured for an additional day following remo val o f the neural tube (to prevent the emigration of addi ti ona l precu rsors). most or the Brn-3.0• cells co-expressed NF 160 and had acquired a neuronal morphology (Fig . .+B.D). By day 3. no further increase in NFI 60 expression was apparent. althoug h the neuronal somata appeared larger (Fig . .+C.D). The increase in the percentage of Brn-3 .0• cells tha t co-exp ressed NFI 60 between day I and day 2 did not appear to be due to the selec ti ve death of undiffere ntiated Brn 3.0+/NF l 60- cells. since the avera~e number of Brn-3.o+ cel ls per explant increased. ra ther than ·decreased. during that time period (Fig. 4D) . These data suggest that the Brn-3.Q+/N Fl6o+ sensory neurons in these explants differentiate de novo from Brn-3 .0•JN Fl60- precursors between day I and day 2 of 45 3550 A. L. Greenwood, E. E. Turner and D. J. Anderson culture. Such differentiation can occur in a fully defined medium lacking exogenously supplied neurotrophic factors . Sensory neurons develop from dividing precursors We were interested to determine whether the sensory precursors in our cultures divided prior to differentiating. Previous stud ies have reported that the sensory neurons that differentiate rapidly in neural crest cultures are derived from postmitotic precursors (Ziller et al.. 1983. I 987: Matsumoto. 1994b). To determine the mitotic activity of the sensory precursors found in our cultures. explants were exposed to 24hour pulses of 175 nM BrdU beginning at the time of plating (day 0). or at day I. day 2 or day 3. Immed iately after the pulse. the cells were fixed and processed for Brn-3.0/BrdU double immunofluorescence. Again. the neural tube was removed from the cultures after 24 hours to avoid the influx of new precursors into the outgrowth. Surprisingly, when BrdU was applied at the time of plating. greater than 95'7r- of the Brn-3.0+ cells in the outgrowth had incorporated the analog by the end of day I (Fig. 5A.EJ. During the next day (day 1-2). BrdU was incorporated into approximately 50% or Brn-3.0+ cells (Fig. 5B.E). The percentage of BrdU+/Brn-3.0• cells dropped dramatically on subsequent days of incubation: less than 5% of Brn-3 .0• cells were labeled with BrdU when it was applied during day 2-3 (Fig. 5C,E). and there was essentially no labeling when it was applied during day 3-4 (Fig. 5D.E). These results indicate that most sensory precursors in our cultures are mitotically active as they leave the neural tube. Moreover. many of these cel ls divide during the subsequent day in culture . After two days. most of the sensory precursors have become postmitotic . suggest that the sensory precursors present in these cultures cannot be diverted to an autonomic fate by BMP2, although this factor clearly induces autonomic neurogenesis in many other cells in the explants. As shown above, approximately 50% of the Brn-3.0• cells in cultures grown in DM come from precursors that divided between day I and day 2. (see Fig. 5B,E). Although BMP2 did not drastically reduce the percentage or number of Brn-3.0+ cells pe r explant. the scatter in the data might conceal the inhibition by BMP2 of sensory differentiation from a subset of dividing precursors. When BMP2-treated cultures were labeled with BrdU during day 2. however. we found that the factor did not prevent the differentiation of Brn-3.0+ neurons from BMP2 does not prevent either dividing or postmitotic precursors from executing a sensory fate If the sensory and autonomic neurons that <lifferentiate in cxplant cultures (Figs I. 2) arise from a common precursor. conditions that promote differentiation of autonomic neurons might do so at the expense of sensory neurons. To explore this possibility. we challenged sensory precursors with 50 ng/ml of BMP2. a signal that instructs some neural crest precursors to become autonomic neurons at the expense of other cell fates (Shah et al.. 1996: Shah and Anderson . 1997: Morrison et al.. 1999). After removal of the neural tube at 2.i hours. BMP2 was applied during the second day. a period during which similar cultures grown in DM contain many dividing sensory precursors (see Figs -iA. 5B.E. respectively). Following a 2.i hour exposure of such cultures to BMP2. a vast number of cells expressed Phox2a (Fig. 6B. brown nuclei). approximately 5- to I 0-fold more than the number of cells expressing Brn-3.0. Thus. althou~h cultures grown in DM do not contain autonomic ne~1rons. treatm~nt of such cultures with BMP2 reveals a large population of neural crest cells with autonomic capacity. Despite th is huge induction of autonomic neurogenesis by BMP2. the percentage of Brn-3.0• neurons per explant did not differ significantly when cultures were grown in the absence (Fig . 6A.C) or presence (Fig. 6B.C) of the factor. Furthermore. the exposure of explants to BMP2 did not seem to affect the total number of Brn-3.0• cells per explant (~=0.995, Mann-Whitney U-test. data not shown). These data + 100 n=S 0 ,.; n=14 C: 75 m ..... + 50 ::) "C 25 :,g 0 0 ai n=6 day 0-1 day 1-2 day 2-3 day 3-4 E Fig. 5. Sensory precursors div ide during the fi rst 1wo days or cul1ure. (A-DJ Confocal images show Brn-3.0 (green ) an<l Br<lU (red) double-immunoHuorescence in culturt!s of neural crest out~rnwth pulsed wi1h BrdU for 2-l hours at ei1her 1he lime or pla1ing-,n <lay 0-1 ( A J. from <lay 1-2 (B ). day 2-3 (C) or day 3--l I Dl. S1rongl y double labeled cells appear yellow. IE) The percentage or Brn-J.Q+ cells th:H incorporated BrdU during the previous 2-+ hours was qui..lntitmed per ex plant and is represented as an average! . n is the number of c:xplunt s from 2-3 experiments. t!rror is gi ven as stand.ird dt!\'iation. Note that the high pcrcen1age of Brn-3 .0• cells labeled by BrdU at day 0- 1 and 1-2 is consistent with the mort.'.! than doubling in the total number of Brn-3.o+ cells per explant <luring thi s period (Fig. -lD ). 46 Determination of the sensory lineage in vitro No Add Fig. 6. 8MP2 induces Phoxla but does not prevent expression or Brn-.lO or division of sensory precursors. (A.8) Double labe ling for 8rn-3.0 and Phoxla at d.iy 1 indicates that 8rn-3.o+ cells (dark blue. nuclear) are present in cultures of neural crest outgrov.'th wh ich have bt!en ~rown for 24 hours in t!ither the absence (A) or prcse~ce (8) or 8MP2. Phoxla• ce ll s (brown. nuclear) are only found in 8MP2-treated cu ltures (8). Unlabeled 3551 +BMP2 ~ ---C\IOl X 0 .i::. C. . -· ·= . cells arc visuulized with DAPI nuorescencc {aqua ma.:l ei --.,; in A.8). Note that the massive induction or Phoxla by 8MP2 in 2-l hours is unlikely to be expla ined by the a5 A 0 C: selecti ve surv ivu l or prolifcr:llion of a subst:t or Phox2a+ cells present in control cultures: therefore BMP2 is likel y to act instructively nllht.:r thun selective ly. (CJ Quantitation of the _ average percentage of Brn -3.Q+ expression per explunt at tluy 2. after 2-+ hours of growth '· 15 n=9 n=7 100 +o in ·No Add' or in 50 ne/111I 8MP2. The data show that .,; 10 a5 +:::, 5 "O a5 25 ~ 0 0 sig ni~ cantly (* P=0.00-1. 1-test). in the presence or BM Pl. In C anti D. n is the number or exp lants counted over 3 experiments: error is represented as standard deviation. n=10 50 --- Simi lar results wen~ obtained when the total number of Brn-3.o+ cells per explant. rather than the percentage of n= 14 C: there is no statisti cally ~,ignilkant reJuction in the percentage of Brn-3.u+ cells by 8MP2I1'=0.705. 1-test). Brn-J.O+ cel ls per cxp lant. was si:orcd tnol shown). ( D) The average percentage of BrJU incorporation into Brn-3.o+- l'Clls per cxpl:.mt dcc:rcases slightl y. but 75 0 No Add C dividing precursors. although it did reduce slight ly (by about 35'i'c) the proportion of Brn-3.0• cells labeled by BrdU (Fig. 6D). These data suggest that the majority of dividing sensory precursors in these cu ltures cannot be respecified by an autonomic ncuron-inducin2 si2 nal. The sl i2ht decrease in the percentage of Brn-3.0• c;ll s ·that incorp;rated BrJU could indi cate that BMP2 either promotes cell cyc le arrest in a minority of sensory precursors. kills some of the cycling precursors. or else diverts a small subpopu lation of th em to an autonomic fate. Iden ti fication of neu ral crest cells expressing early sensory precursor-specific markers The previous data indicated that neural tube explants grown in DM contain a population of mitotieally active sensory neuron precursors that cannot be diverted to an autonomic fate by exposure to BMP2 during the seco nd day of culture. However. these data did not exclude the possibility that more immature sensory progenitors. suc h as those that do not yet express Brn3.0. may be re specified by the factor. We therdorc sough t to directly visualize such precursors in order to examine their response to BMP2. To do this. we used the bHLH transcription factors 11g11 I. 11g112 and 11e11mD as early markers of sensory precursors. In vivo. 11g11:! is transiently expressed by so me migr.tting crest ce ll s. whi le 11g11 I is expressed slight ly later rollowing condensation ot' the sensory ganglia (Sommer et al. . 1996: Perez et al. . I 999: Ma et al.. 1999). neumD expression is downstream of and genetically ,Jcpcndcnt on the 11g11s (Ma et al.. I 996. 1998: Fode et al.. I 998). Expression of 11e11mD appears to overlap the period during which Brn-3.0 is expressed durin~ DRG formation in 1·irn I Fedtso1·a and Turner. 1995). In ;i tu hybridization of 2-t hour explant cultures revealed No Add +BMP2 +BMP2 D cells expressing 11g112 mRNA both in the neu ral tube and in the immediately adjacent outgrowth (Fig. 7B.E. arrows). Combined in situ hybridization and antibod y staining indicated that most ot· the 11g112• cells in the outgrowth did not co-express Brn-3.0 (Fig. 7E. arrowheads). although a rew examp les of 11g11J+/Brn-3.0• cells could be found (Fig. 7E. inset). This relative lack of overlap is consistent with the fact that express ion of 11g112 is initiated prior to that of Brn-3.0 in vivo ( Fedtsova and Turner. I 995: Sommer et al.. 1996: Ma ct al. . 1999). but the identification ot· double-labeled cells supports the idea that at least some Brn-3.0• sensory neurons derive from 11g112-expressing precursors (sec below). Expression of 11g11 I was not detected among em igrating crest cells in such 2-t hour explants (Fig . 7A). but was detected a day later (Fig. SC). consistent with the fact that its express ion follows that of 11g112 in vivo (Sommer et al.. 1996: Ma et al. . 1999). Numerous 11e11roO• cells were al so found throughout the outgrowth (Fig. 7C.F arrows) as we l l as in the neural tube (data not-shown). T11e 11eumD• ce lls in the outgrow th tended to be located more distal to the neural tube than 11g112• cells ( Fig . 7B.C). suggesting that the 11<' 11roD• cells may have emigrated earl y and have already down regu lated express ion of 11g112 ( Lee et al.. 1995: Ma et al.. 1996. 1998: Fode et al.. I 99S). In contrast to 11g112. most 11e1m,o • cel ls co-expressed Brn-3.o+ (Fig . 7F. arrows and arrowheads: inset). Given the genetic dependence of 11e11roD express ion on 11g112. the 11g11]+/Brn-J.Oand the neuroO•tBrn-3.o+ precursors likely represent successive stages in sensory development. Consistent with this. forced express ion or 11g11s can induce expression of Brn-3.0 both in vivo (S. Perez and D. J. A .. unpublished data ) and in vitro (L. Lo ~nd D. J. A .. unpublished data). Thus expression of 11g11s can be used to identify senso ry precursors at an even 47 3552 A. L. Greenwood, E. E. Turner and D. J. Anderson earlier stage than they can be identified by expression of Brn3.0. BMP2 does not prevent expression of early, sensory precursor-specific markers To determine if BMP2 might prevent the development of the early sensory precursors identified by expression of ngn2 or neuroD, we exposed explants to this factor at the time of plating (before neural crest emigration) and assayed expression of these and other markers after 24 hours. Although BMP2 induced expression of both Phox2a and MASH! in the outgrowth of neural tube explants (Fig. 8F,G, arrows, compare Fig. 8A,B), it did not abolish the expression of ngnl, ngn2 and neuroD in cells adjacent to the neural tube (Fig. 8H-J). If anything, the express ion of these sensory precursor-specific genes in the explants may have been slightly elevated by exposure to BMP2 (Fig. 8, compare C versus H, D versus I, E versus J). As expected from the results described earlier, addition of BMP2 at the time of plating also did not prevent the expression of Brn-3 .0 (Fig. 8A,F, arrowheads). These data indicate, qualitatively, that the survival and/or proliferation of early sensory precursors identified by expression of ngns and neuroD was not prevented by addition of BMP2, and that the factor did not suppress expression of these sensory bHLH genes. Moreover, as ngn2 is expressed earlier than Brn-3 .0, these results suggest that sensory precursors may be refractory to the autonomic-inducing activity of BMP2 even before they express Brn-3 .0. Strikingly, the MASH!+ autonomic precursors revealed by addition of BMP2 occupied a more distal region of the outgrowth relative to the neural tube than did the sensory precursors identified by expression of the ngns (cf. Fig. 80, arrows versus 8H, I and J). This complementary distribution is similar to that of the differentiated sensory and autonomic neurons that develop from these precursors (Fig. 1, compare X versus X', Fig. 8F, compare arrowheads versus arrows). The reasons for this patterned distribution are unclear but are currently under investigation. The phenotype of sensory neurons generated in defined medium includes the proprioceptive and/or mechanoreceptive, but not nociceptive, sensory subtypes The relative rapidity with which the Brn-3.0+ sensory neurons developed in our cultures suggested that they may correspond to the early-differentiating subset of sensory neurons in vivo. Early-differentiating neurons are large-diameter cells that comprise primarily the proprioceptive and mechanoreceptive sensory subtypes (Lawson et al., 1974; Lawson and Biscoe, 1979; Lawson, 1992). We were therefore interested in defining the subtype(s) of the sensory neurons that differentiate in our cultures. Since it is known that different subsets of sensory neurons require specific neurotrophins for their survival (reviewed in Snider, 1994), we sought to determine the trophic support profile of the sensory neurons generated in vitro as one criterion to identify their subtype. Proprioceptive or mechanoreceptive sensory neurons require NT-3 or BDNF, while nociceptive sensory neurons require NGF or LIF (for review see Snider, 1994). Explants were grown for 8 days in DM or DM supplemented on day 3 with NGF, LIF, NT3 or BDNF alone or in combination. Explants grown in DM contain hundreds of sensory neurons at day 4 (Fig. 3A,B and data not shown), but very few Brn-3 .0+ cells survived after 8 days of culture (Fig. 9, 'No Add'), suggesting that these neurons die in the absence of exogenous trophic support. Addition of 100 ng/ml of NGF or 50 ng/ml of LIF at day 3 did not effect a significant increase in the number of surviving Brn-3.0+ cells (Fig. 9, NGF and LIF) . Medium supplemented with 50 ng/ml of NT3 or BDNF, however, supported the survival of significantly more sensory neurons than were obtained in NGF or LIF (Fig. 9, NT3 and BDNF). When explants were grown in both BDNF and NT3, more sensory neurons survived than with either factor alone (Fig. 9, +NT3 +BDNF). Similar numbers of Brn-3.0+ cells were observed in cultures grown in NT3 plus BDNF compared to those cultured in the entire cocktail of factors (Fig. 9, +NGF +LIF +NT3 +BDNF). Therefore, our cultures contain sensory Fig. 7. Early sensory precursor-specific transcription factors are expressed in neural tube explant cultures at day 1. (A-F) Cells in the neural crest outgrowth (below dashed line) of explants grown in defined medium for 1 day express in situ signal (blue) A for ngn2 (B,E, arrows), and neuroD (C,F, arrows), and are immunopositive for Brn-3 .0 ..,__ (D, blue nuclei; E,F, brown nuclei, arrowheads). (A) At this stage ngnl expression is restricted to the neural tube (cells above dashed line). (E) Most 11g112+ cells (arrows) do not co-express E D Brn-3.0 (arrowheads), although a few double-labeled cells can be found (inset). (F) In contrast, most neuroD+ cells (arrows) do express Brn-3.0 (arrowheads; inset). The extent of the outgrowth is visualized with DAPI fluorescence (aqua). 48 Determination of the sensory lineage in vitro 3553 neurons that can be supported by NT3 and BDNF, but not by NGF or LIF. Different subpopulations of sensory neurons are also known to have distinct profiles of gene expression (e.g., see Lawson, 1992 for review). We therefore characterized the sensory neurons that appeared in explants grown for 8 days in DM supplemented with all four neurotrophic factors to determine if they expressed markers known to correlate with a particular class of sensory neurons. We found that sensory neurons grown under these conditions expressed a highly phosphorylated form of neurofi lament recognized by the monoclonal antibody RT97 (Fig. lOA), the expression of which in vivo correlates with the large proprioceptive subpopulation of sensory neurons (Lawson et al., 1984). Additionally, the large neurons in neural ne uroD / D,..,.? 1 MASH1 / 01'.l.PI .. . .. ~ "C "C cl: 0 z • ., . ~ .. ·i -'•., A -. .. I ~ -·.. 8 C ~ - - --_,,,,,. E D ........ -- - - - - - - - - - ....... •• I H J Fig. 8. Sensory precursor-specific markers persist in the presence of BMP2. Neural tube explants were grown in DM for 2 days in the absence (A-E) or presence (F-J) ofBMP2. (A,F) Double labeling for Brn-3.0 (dark blue, arrowheads) and Phox2a (brown, arrows) shows that Brn-3.0+ cells (arrowheads) are found in both conditions (A,F, arrowheads), but Phox2a+ cells (arrows) are found only in the presence of BMP2 (F). (B,G) Similarly, cells immunopositive for MASH! (arrows, purple nuclei) are found only in BMP2-treated explants (G) but not in 'No Add' conditions (B). (C-E versus H-J, purple in situ signal) At this stage, explants grown in the absence of BMP2 (C-E) contain a few cells in the outgrowth expressing ngnl (C, arrows), and many more expressing ngn2 (D) and neuroD (E). Explants grown in the presence of BMP2 (H-J), contain cells expressing similar or greater levels of 11g11J (H), 11g112 (I), and neuroD (J). 49 3554 A. L. Greenwood, E. E. Turner and D. J. Anderson 1500 -r-------------------~ d n.fl Brn-3.0/RT97 ER81 • • • .... .. - • • •• • d • n■9 ll "' • •• .. • " •• ••• • • .. B Brn-3.0/CGRP Brn-3.0/SP . • •NT3 +BONF +Nll +BDNF +NOF +LIF +BDNF • Fig. 9. NT3 and BDNF, not NGF or LIF, support significant crest explants expressed ER8 l (Fig. !OB), an ETS domain transcription factor also restricted to the proprioceptive sensory subtype (Lin et al., 1998). Conversely, these neurons did not show detectable expression of markers that co1Telate with the nociceptive phenotype (for reviews see Hunt et al. , 1992; Akopian et al., 1996): the neuropeptides , SP (Fig. IOC) and CORP (Fig. !OD), or trkA, the high-affinity receptor for NOF (data not shown). .. • ,NT3 populations of explant-derived sensory neurons. The graph represents the total number of Brn-3.0+ cells in the neural crest outgrowth per ex plant cultured for 8 days in defined medium in the absence or presence of different neurotrophins. There are more Brn3.Q+ cells in cultures grown in NT3 or BDNF than in cultures grown in DM ('No Add'), NGF or LlF. Cultures grown in NT3+BDNF and NGF+LIF+NT3+BDNF contain significantly more Brn-3.0+ cells than other conditions. n is the number of explants counted from 2-4 experiments. Error is represented as standard deviation. Statistics were compared by ANOVA followed by pair-wise I-tests. Statistically significant differences (P<0.05) are indicated by letters (e.g., a is different from b, but not from a or a, b). ~ , • . • 0 Fig. 10. Explant-derived sensory neurons express proprioceptive but not nociceptive, sensory subtype markers. (A-D) Neural tube ex plants were grown in defined medium in the presence of NGF, LIF, NT3 and BDNF for 8 days and stained for sensory subtype markers. (A) Double immunostaining for Brn-3.0 (dark blue nuclear) and RT97 (brown, cytoplasmic) reveals ce lls that co-express both markers (arrows) as well as many RT97+ processes. (B) The large diameter cells in these explants also express ER8 I (dark blue, nuclear), a marker of proprioceptive sensory neurons (Lin et al., 1998). (C,D) In contrast, Brn-3.0+ cells in these cu ltures (dark blue, nuclear) do not express detectab le levels of SP (C) or CORP (D) staining (brown, cytoplasmic), compared to dissociated postnatal DRGs stained in parallel as a positive control (C,D insets). nociceptive, phenotype. Such precursors may therefore correspond to a lineally (Frank and Sanes, 1991) and genetically (Ma et al., 1999) distinct subset of earlydifferentiating sensory precursors recently identified in vivo. Mammalian sensory neurogenesis in vitro DISCUSSION The timing and mechanism of sensory and autonomic lineage segregation have been of interest for many years (Le Lievre et al., 1980; Schweizer et al., 1983; Le Douarin, 1986; Le Douarin et al., 1994). Here we have established a culture system in which mammalian sensory neurons differentiate from proliferating precursors in a defined medium. Although BMP2 induces massive autonomic neurogenesis in such explants, it does not prevent sensory neurogenesis . Furthermore, the survival and/or proliferation of early sensory precursors, identified by expression of several sensory-specific transcription factors, is not abolished by BMP2. These data provide evidence that a subset of neural crest cells are, operationally, committed to a sensory fate with respect to an autonomic inducing cue that is likely to be relevant in vivo (Reissman et al. , 1996; Shah et al., 1996). Strikingly, the sensory neurons that develop from these precursors appear to assume a proprioceptive and/or mechanoreceptive, rather than Considerable ambiguity has surrounded the identification of sensory neurons in mammalian neural crest cultures due to a lack of definitive markers. For example, large numbers of 'sensory-like' neurons , identified by expression of calcitonin gene-related peptide (CORP) and SubstanceP (SP) were shown to differentiate in mouse neural tube explant cultures (Murphy et al., 1991, 1994). These authors acknowledged that CORP and SP are not definitive sensory markers; for example, they can be expressed by sympathetic neurons under certain conditions (Nawa and Patterson, 1990; Fann and Patterson, 1994a,b). Indeed, other studies have shown that CORP and SP can be co-expressed with autonomic markers in individual neurons that develop from mouse neural crest cells when they are grown in an undefined medium (Matsumoto, 1994a,b). In our hands, neural crest cultures grown for long pe1iods (8 days) in an undefined medium contain, qualitatively, many more autonomic than sensory neurons . Therefore, it is possible that many, if not all, of the 'sensory-like' neurons observed in cultures of mouse neural crest grown in undefined conditions 50 Determination of the sensory lineage in vitro 3555 (Murphy et al., 1991, 1994) may in fact have been autonomic The neural crest contains dividing precursors that in nature. Alternatively, these neurons may have been of the generate sensory neurons even in the presence of nociceptive sensory subtype, which express CORP and SP in an autonomic-inducing signal vivo and which do not develop under the culture conditions Our results point to the existence of a subpopulation of dividing described here. neural crest cells that execute a sensory neuron fate whether or In both mammalian and avian systems, early-differentiating not they are challenged with BMP2, an autonomic-inducing sensory-like neurons in neural crest cultures have been cue. In this sense, these precursors appear 'committed' to a reported to arise from postmitotic precursors (Ziller et al., sensory neuron fate. Under these defined culture conditions, 1987; Sieber-Blum, 1989; Matsumoto, 1994b). Using BrdU autonomic neurons do not differentiate unless BMP2 is added labeling, we have demonstrated that virtually all the Brn-3.Q+ to the medium. The simplest explanation for these observations sensory precursors present in our cultures at day I are is that there are separate precursors for sensory and autonomic proliferating and that they continue to divide for the next 1-2 neurons in our cultures (Fig. I IA, Ps and PAGM, respectively) . days. Consistent with this, cell division also persists after the However, it is also formally possible that the cultures contain initial expression of Brn-3.0 by sensory precursors in vivo multipotent progenitors that generate both sensory and (Fedtsova and Turner, 1995). A likely explanation for the autonomic neurons in BMP2 (Fig. 1 lB , PsAGM). In order to difference between our results and those obtained in mouse explain why BMP2 would not induce autonomic neurogenesis neural crest cultures (Matsumoto, 1994b) is that the from such progenitors at the expense of a sensory fate, it is concentration of BrdU used in the earlier study (10 µM) was necessary to make two independent assumptions: first, these toxic to sensory precursors. In this study, we used 175 nM multipotent cells must be themselves unresponsive to BMP2 BrdU to label sensory precursors; we found that 10 µM BrdU (Fig. 11B , PsAGM), but divide to generate an autonomic disrupted explant morphology and abolished Brn-3.0 expression, although BrdU was incorporated into A. SEPARATE SENSORY AND AUTONOMIC many cells (A . G. and D. J. A., unpublished data). PRECURSORS The establishment of a system in which mammalian ( a (BMP2 sensory neurons differentiate de novo from ~ non-responsive) proliferating precursors in a defined medium now '- ? opens the way to more detailed studies of the cellin trins ic and cell-extrinsic factors controlling their + development. ,. e "G Fig. 11. Origins of sensory neurons in vitro and in vivo. (A,B) Alternative explanations for our in vitro observations; (C) represents a speculative model for sensory neurogenesis in vivo. In this schematic, straight arrows indicate potential and curved arrows indicate division, not self-renewal. (A) Sensory precursors (Ps) and their progeny do not have the potential to execute an autonomic neuronal fate (NA) in response to BMP2. Instead, they generate only sensory neurons (Ns) or sensory neurons and glia (G). In this model, autonomic neurons are derived from a separate population of precursors (PAGM), which are assumed to have glial and myofibroblast (M) potential like neural crest stem cells (Shah et al. , 1996; Morrison et al., 1999). (B) A common precursor to sensory and autonomic neurons (PsAGM) whose sensory differentiation is not inhibited by BMP2. Such a model assumes that this cell undergoes obligate asymmetric divisions that generate a sensory precursor (Ps) and a BMP2-responsive autonomic progenitor (PAGM). The sensory daughter of the asymmetric division may divide symmetrically before it differentiates (not illustrated). (C) Two classes of sensory progenitors are postulated to exist in vivo: those committed to a sensory fate (Ps 1) and those that can generate both autonomic and sensory neurons (Ps21A) . The former are 11g11-2-dependent (Ma et al. , 1999) and are fated to generate primarily largediameter (proprioceptive/mechanoreceptive) sensory neurons (Frank and Sanes, I 991 ), but are not necessarily committed to this sensory subtype. The latter give rise to both autonomic progenitors (PAGM) and ngnl-dependent sensory precursors (Ps2). These sensory precursors generate both large-diameter and small-diameter (nociceptive) sensory neurons (Frank and Sanes, I 99 I; Ma et al. , I 999). B. COMMON BMP2-UNRESPONSIVE PRECURSOR ~ (BMP2 non-responsive) / ~ a t o r y asymmetric division (BMP2 responsive) C. TWO KINDS OF SENSORY PROGENITORS Ce Q9 I~ P• 872 """'+ ngn1-I dependentt e 0 8 ~ s V JZ!oo", ~ - W, Large-diameter sensory neurons Small-diameter sensory neurons 51 3556 A. L. Greenwood, E. E. Turner and D. J. Anderson progenitor that is BMP2-responsive (Fig. l lB, PAGM ): second, such cell divisions must be asymmetric, so as to always generate a daughter that differentiates into a sensory neuron even in the presence of BMP2 (Fig. 11 B, Ps). (Alternatively, equivalent cells could divide symmetrically to generate with equal probability either sensory or autonomic precursors.) In this way, such precursors would be ' committed' to generating sensory neurons in BMP2. but not be restricted from an autonomic fate. The inability to obtain sensory neuron differentiation in clonal cultures (A. G. and D. J. A., unpublished observations) and the low percentage of sensory neurons that develop in our explants makes it difficult to distinguish between these models by direct lineage analysis. However, we favor the former model (Fig. 11 A) because it makes fewer assumptions. Our use of BMP2 to challenge sensory precursors raises the question of whether this factor is a biologically relevant signal for testing the commitment state of these cells. BMPfamily members have been shown to induce autonomic differentiation in a highly robust manner in multiple contexts. BMP2, BMP4 and BMP7 are able to induce expression ot· autonomic markers in cultures of neural crest cells from either avian (Varley et al., 1995; Reissman et al., 1996; Varley and Maxwell, 1996) or mammalian (Shah et al.. 1996: Shah and Anderson, 1997; Morrison et al. , 1999) embryos. Overexpression of BMPs in vivo has also been s hown to induce autonomic markers (Reissman et al., 1996). In vivo, BMPs are expressed by tissues (e.g., dorsal aorta) near s ites of autonomic differentiation (Reissman et al., 1996; Shah et al., 1996) and these tissues can mimic the inductive effects of BMPs (Reissman et al., 1996). Taken together. these data argue that BMP family members are likely acting as autonomic inducers in vivo. Therefore , although the sensory precursors that we have studied may not be committed with respect to other inductive cues, their insensitivity to BMP2 is likely to be of particular significance for sensory development in vivo. BMP2 also constitutes a particularly stringent challenge of sensory precursors because it has been shown to act instructivel y on rat neural crest stem cells, causing them to adopt an autonomic neuronal fate at the expense of alternative fates (S hah et al., 1996). Because our assay for sensory neurogenesis could not be performed us ing clonal cultures. we were unable directly demonstrate that BMP2 acts instructivel y in thi s syste m. However, the large induction of autonomic markers that occurs during a 24 hour exposure to this factor (approximately, 0% Phox2a+ cells in 'No Add ' ve rsus 25-50% in BMP2 ; see also Fig. 6) is consistent with an instructive effect of BMP2. Gi ven that BMP2 has been shown to act instructively on rat neural crest cells isolated from two different sources and grown in two different culture media (Shah et al., I 996; Morrison et al., 1999), it is highly likely that the factor acts instructivel y in this system as well. Earlier studies in avian embryos have been interpreted to suggest that the neural crest contains a subset of precursors restricted to a sensory fate (Ziller et al. , 1987; Sieber-Blum, 1989 : Duff et al., I 991 ). Although this general conclusion is similar to that which we have drawn here, the evidence supporting it differs in several important respects from our own results. In one series of experiments, sensory-like neuron s were observed to differentiate in a defined medium; sensory differentiation was prevented by serum. which promoted autonomic neurogenesis (Ziller et al., 1983 , 1987 ). These observations were interpreted to suggest that sensory neurons developed from separate precursors that were restricted from an autonomic fate. However. since sensory neurons failed to appear in serum-containing medium , it could not be determined whether the sensory precursors died or were converted to other fates. Other investigators have also concluded that some sensory precursors are restricted from an autonomic fate because they do not produce autonomic neurons under culture conditions that allow autonomic neurogenesis (Sieber-Blum, 1989: Duff et al., 1991 ). In neither of these studies, however, were the sensory precursors challenged by addition of defined, instructive autonomicinducing signals. The fact that autonomic neurons differentiate under a given set of culture conditions may s imply reflect the presence of factors permissive for such differentiation. in which case they are not an adequate challenge of the state of commitment of sensory precursors. Multiple sensory sublineages in the neural crest The ability of some neural crest cells to give rise to both sensory and autonomic neurons in vivo is well-documented (Bronner-Fraser and Fraser, 1988, 1989 ; Fraser and BronnerFrase r. 1991 ). Nevertheless, the notion that migrating neural crest also contains a subpopulation of precursors determined for a sensory fate is still consistent with the results of such lineage analyses. In one study, almost 50% (8/17) of migrating avian neural crest cells injected with a lineage tracer produced neurons exclusively in the DRG, rather than in both the DRG and sympathetic ganglia (Fraser and Bronner-Fraser, 1991 ). From such a lineage analysis alone, it is not possible to distinguish whether these clonal differences represent stochastic variations in fate among a developmentall y equivalent precursor population. or rather distinct subpopulations that are intrinsically different. The present observations are consistent with the possibility that some lineage-marked crest cells that generated only sensory neurons in vivo were in fact determined for a se nsory fate. lnterestingly. a rece nt clonal analysis of migrating crest in vitro suggests that many precursors are fated to generate only neuronal progeny. although whether these neuron s were sensory or autonomic was not investigated (Henion and Weston. 1997 ). Our observations. taken together with previous studies. support the notion that there may be at least two classes of sensory precursors in the neural crest: those that are determined for a sensory fate early in neural crest ontogeny, and those that have a wider range of potentials including both se nsory and autonomic fates (Bronner-Fraser and Fraser, 1988, 1989 : Sieber-Blum, 1989 ; Fraser and Bronner-Fraser, 1991 ). Interestingly, sensory precursors in our cultures generate neurons that are predominantly of the proprioceptive or mechanoreceptive subtypes. Such cells may therefore correspond to a distinct sublineage of early-differentiating precursors that generate large-diameter sensory neurons (Fig. I IC , Psi) that have been identified by a previous in vivo retroviral clonal analysis in chick (Frank and Sanes, 1991 ). Similar precursors are genetically dependent on 11g112 but not 11g11 I in mice (Ma et al., 1999). This correspondence does not mean that 11g112-dependent precursors are necessarily committed to a proprioceptive/ mechanoreceptive subtype; for 52 Determination of the sensory lineage in vitro 3557 example, they may simply differentiate at a time when signals that promote these subtypes are present. If the progenitors that appear committed to a sensory fate in our cultures correspond to those fated to generate earlydifferentiating proprioceptive and mechanoreceptive neurons in vivo, then perhaps the common progenitors of sensory and autonomic neurons (Fig. I IC, Ps21A) identified by in vivo lineage tracing (Bronner-Fraser and Fraser, 1988, 1989; Fraser and Bronner-Fraser, I 99 I ) are cells that give rise to the laterdifferentiating subpopu lation of sensory precursors (Fig. I IC, Ps2; Lawson et al. , 1974; Carr and Simpson, 1978; Lawson and Biscoe, 1979; Frank and Sanes, I 99 I). Such latedifferentiating sensory precursors have been shown to be ngn !dependent (Ma et al., 1999), and to give rise to both smalldiameter (nociceptive) and large-diameter (proprioceptive/ mechanoreceptive) neurons (Frank and Sanes, 1991; E. Frank, personal communication). At present it is not clear why only the early-differentiating sensory precursors develop in our cultures. One possibility is that the later-differentiating precursors may not migrate from the explanted neural tube. Alternately, they may migrate but not differentiate into, or survive as, nociceptive sensory neurons. In the absence of conditions to elicit differentiation of such nociceptive neurons, we cannot say whether the BMP2-responsive autonomic precursors do or do not have sensory potential. These cells may correspond to a common sensory-autonomic precursor from which we are currently able to elicit autonomic but not sensory neurogenesis (Fig. I IC, Ps21A); alternately, they may already be restricted to an autonomic fate (Fig. I IC, PAGM). Early determination of some sensory precursors may prevent them from responding to autonomicinducing signals as they migrate When do sensory precursors become determined? Our data reveal a correlation between the resistance of sensory precursors to an autonomic inducing signal and the persistence of 11g112-expressing cells despite the presence of this signal. Forced expression of ngns in chick neural crest cells in ovo can bias their distribution to the DRG and can induce sensory-specific genes including Brn-3.0 (Perez et al., 1999; S. Perez and D. J. A., unpubli shed data). These data are consistent with the idea that some progenitors may become determined for a sensory fate as early as the time they exp ress ngn2. In vivo, ngn2 is expressed by a subset of neural crest cells early in migration (Gradwohl et al., I 996; Sommer et al., 1996; Perez et al., 1999; Ma et al., 1999). Therefore, early expression of ngn2 may identify a subset of neural crest cells that are specified for a sensory fate short ly after they emigrate from the neural tube, a conclusion consistent with some results from the in vivo lineage-tracing experiments mentioned earlier (Fraser and Bronner-Fraser, 1991 ). These ce ll s may be similar or equivalent to those that we have studied here in vitro. The early determination of some sensory precursors among the migrating neural crest may serve to make them resistant to the autonomic-inducing effects of locally expressed BMPfamily members. Although these factors are likely to be important for autonomic neurogenesis in vivo, these same BMPs are also expressed in many tissues in the dorsal part of the embryo where autonomic neurons do not develop, including dorsal ectoderm (Liem et al., 1995), the dorsal neural tube (Liem et al., 1997) and the developing somites (Reshef et al., 1998). Because migrating neural crest cells are likely to encounter BMP family members in the dorsal part of the embryo, mechanisms must exist that prevent them from differentiating into autonomic neurons in this location. The early determination of some sensory progenitors with respect to B MPs may represent one such strategy. The fact that such determination occurs while the precursors are still mitotically active, moreover, would allow for the continued expansion of the sensory neuron population by postmigratory cells during the growth of the DRG (Duff et al., 1991 ). In this way, the early determination of a subset of dividing neural crest cells may play an important role in the growth and development of sensory ganglia. We thank Li Ching Lo, Andy Groves, Sean Morrison, Nirao Shah and Lukas Sommer for helpful suggestions and advice pertaining to ·tissue cu lture and medium formulation. We thank H. Phan and L. Wang for technical help and G. Mosconi for laboratory management. Thanks also to S. Arber. J. F. Brunet. C. Goridis, T. Jessel!, L. Reichardt and J. N. Wood for supplying essential antibody reagents, to G. Yancopoulos/Regeneron for NT-3 and BDNF, and to Y. Rossen/Genetics Institute for BMP2. We thank S. Perez, A. Paquette, S. Gerety, P. Patterson, R. Lahav, S. Fraser and M. Bronner-Fraser for critical reading of the manuscript. This work was supported by a grant from the March of Dimes Foundation. D. J. A. is an Investigator of the Howard Hughes Medical Institute. REFERENCES Akopian, A. N., Abson, N. C. and Wood, J. N. ( 1996). Molecular genetic approaches to nociceptor development and funclion . Trends Newvsci. 19, 240-246. Anderson, 0. J. ( 1989). The neural crest cell lineage problem: Neuropoiesis' Neuron 3. 1- 12. Anderson, 0. J . ( 1993). Cell and molecular biology of neural crest cell lineage diversification. Cun: Opin. Neurobiol. 3. 8- 13. Anderson, 0. J. ( 1997). Cellular and molecular biology of neural crest cell lineage determination. Trends Gener. 13. 276-280. Bronner-Fraser, M. and Fraser, S. ( 1988). Cell lineage analysis shows multipotentiality of some avian neural crest cells. Narure 335, 161-164. Bronner-Fraser, M. and Fraser, S. ( 1989). Developmental potential of avian trunk neural crest ce ll s in situ. Net1ro11 3, 755-766. Bronner-Fraser, M. E. ( 1993). Segregation of cell lineage in the neural crest. Cun: Opin. Gen. Devel. 3, 641-647. Carr, V. M. and Simpson, S. 8. ( 1978). Proli ferative and degenerative events in the early development of chick dorsal root ganglia. J. Co111p. Ne,irol. 182. 727-740. Clary, 0. 0 ., Weskamp, G., Austin, L. R. and Reichardt, L. F. ( 1994). TrkA cross-linking mimics neuronal responses to nerve growth factor. Mo/. Biol. Cell 5, 549-563. Duff, R. S., Langtimm, C. J., Richardson, M. K. and Sieber-Blum, M . ( 1991 ). In vitro clonal analysis of progenitor cell patterns in dorsal root and sympathetic ga ngli a of the quail embryo. Dev. Biol. 147. 451-459. Fann, M.-J. and Patterson, P. H. ( I 994a). Depolarizati on differentially regulates the effects of bone morphogenetic protein (BMP)-2. BMP-6. and Activin-A on sympathetic neuronal phenotype. J. Neurochem. 63. 2074- 2079. Fann, M.-J. and Patterson, P.H. ( 1994b). Neuropoietic cytokines and activin A differentially regulate the phenotype of cultured sympathetic neurons. Proc. Nari. Acad. Sci. USA 91, 43-47. Fedtsova, N. G. and Turner, E. E. ( 1995). Brn-3.0 expression identified early post-mitotic CNS neurons and sensory neural precursors. Mech. Dev. 53. 291-304. Fode, C., Gradwohl, G., Morin, X., Oierich, A., LeMeur, M., Goridis, C. and Guillemot, F. ( 1998). The bHLH protein NEUROGENIN 2 is a determination factor for epibranchial placode-derived sensory neurons. Neuron 20. 483-494. 53 3558 A. L. Greenwood, E. E. Turner and D. J. Anderson Frank, E. and Sanes, J. R. ( 1991 ). Lineage of neurons and glia in chick dors;.il root ga nglia: analysis in vivo with a rec.:o mbinanl retrovirus. Develop111e111 111. 895-908. Fraser, S. E. and Bronner-Fraser, M. E. ( 199 I). Migrating neura l crest cells in th e trunk of th e avian embryo are mu hi potent. Deve!op111e11t 112. 9 l 3920. Gradwohl, G., Fode, C. and Guillemot, F. ( 1996 ). Restricted exp ression of a novel murine ato11a/-re lated bHLH protei n in undifferen1ia1ed neural precursors. Dei•. Biol. 180, 227-241. Groves, A. K. and Anderson, D. J. ( 1996). Ro le of environmental sig nals and tran scrip tional regulators in neura l crest development. Oe1'. Ge11e1. 18. 6472. Hawrot, E. and Patterson, P. H. ( 1979). Lone-term culture of dissoci at ed sympathetic neurons. Methods in E11cy11w/og_,~58, 574-58 3. Henion, P. D. and Weston, J. A. ( 1997). Timing and pattern of ce ll fate restrictions in lhe neural crest lineage. Developmem 124. 435 1-4359. Henrique, D., Adam, J., Myat, A., Chitnis, A., Lewis, J. and l s h•Horowicz, D. ( 1995). Expression of a Delra homo logue in prospective neurons in th e chick. Natnre 375 , 787-790. Hunt , S. P., Mantyh , P. W. and Priestly, J. V. ( 1992). The organization of biochemically characterized sensory neurons. In Sen.wry Ne11ro11s: Dil'ersity. Devefopme111, 011d Pf{lsticity. (ec.l. S. A. Scou). pp . 60•76. New York: Oxford Uni ve rsi ty Press. Johnson, J.E., Birren, S. J. a nd Anderson, D. J . ( 1990). Two rat homologue, or Drosophila aclwete-scu1e speci ficall y expre ssed in neu ronal precu rsors. Nmure 346, 858-86 1. Lawson, S. N. ( 1992). Morphological and Biochemica l Cell Types of Sensory Ne urons. In Sensory Neurons: Diversity, Del'elop111e11t. and Plasticity. (ec.l. S. A. Scott). pp. 27-59. New York: Oxford Uni ve rsity Press. Lawson, S. N. and Biscoe, T. J . ( 1979). Deve lopment of mou se dorsal root ganglia: an autorac.liographic anc.l quantitative study. J. Ne11roc:yro/. 8. 265· 274. Lawson, S. N., Caddy, K. W. T. and Biscoe, T. J. ( 1974). De ve lopment of rai dorsal root ganglion neurones: studies of cell birthdays anti c hanges in mean ce ll diam ete rs. Cell Tiss. Res. 153. 399-413. Lawson, S. N., Harper, A. A., Harper, E. I., Garson, J. A. and A nderton, 8. H. (1984 ). A mon oclonal ant ibody against neurofilament protein specifically labels a subpop ulation of rat se nso ry neurones. J. Comp. Ne11rol. 228 . 263-272. Le Douarin, N., Dulac, C., Du pin, E. and Cameron-Curry, P. ( 1991 ). Glial ce ll lineages in the neu ral crest Glia 4. 175-1 84. Le Douarin, N. M. ( 1986). Cell line segregation during peripheral nervou s system ontogeny. Science 231 , 1515-1522. Le Douarin, N. M. , Dupin, E. and Ziller, C. ( 1994 ). Genetic and epige neti c co nt ro l in ne ural cres t deve lo pment. Cun: Opin. Gene. Dev. 4, 685-695. Le Lievre, C. S. 1 Schweizer, G. G., Ziller, C. M. and Le Douarin, N. M. ( I 980). Restri ctions of develop me nta l capabilities in neural crest cell deri vatives tested by in vivo tra nspl antat ion experim ents. De1 •. Biol. 77. 362378. Lee, J. E., Holle nberg, S. M., Snider, L., Turner, D. L., Lipnick, N. and Weintraub, H. ( 1995). Conversion of Xenop11J ec1oderm inco neurons by NeuroD. a basic hel ix-loop-hel ix protein . Science 268. 836-844. Liem, K. F., Jr., Tremmel, G. and Jessell, T. M. ( 1997) . A ro le for the roof plate and its resident TGF~-related proteins in neuronal pauerning in the dorsa l spinal cord. Cell 91. 127-138. Liem, K. F., Tremml, G., Roelink, H. and Jessell, T. M. ( 1995). Dorsal di ffe rentiati on of neural plate ce lls induced by BMP•mediatcd sig nal s from epidermal ectoderm. Cell 82. 969-979. Lin, J . H. 1 Saito, T., Anderson, D. J., Lance-Jones, C., Jessell, T. M. a nd Arber, S. ( 1998). Functionall y- rel ated motor neuron pool ,rnc.l muscle senso ry affe ren t subt ypes defined by coord inate £TS gene ex.pression. Cell 95. 393-407 . Lo, L., Tiveron, M.-C. and Anderso n, D. J. ( 1998). MAS H I activates ex press io n o f th e paired hom eodo mai n transc ription factor Phox2a. anc.l couples pan-neuronal and subtype-specific compo nents o f auto nom ic neuronal identil y. Developmelll 125, 609-620. Lo, L.-C. and Anderson, D. J. ( 1995). Postmigratory neural crest ce lls expressing c-ret display rest ricted deve lopm en ta l anc.l proliferative capacities. Nemvn 15. 527-539. Lo, L.-C., Johnson , J. E., Wuenschell, C. W., Saito, T. and Anderson, D. J. ( 1991 ). Mammalian aclwete•scute homo log I is tra nsientl y expressed by spat iall y•restricted subse ts of ea rl y neuroepithelia l and ne ural crest ce lls. Genes Dev. 5. 1524-1537 . Ma, Q., Fode, C., Guillemot, F. and Anderson, D. J . ( 1999) NEUROGENIN I and NEU ROGENI N2 control two distinc t waves of neurogenesis in developing dorsal rout ga ng li a. Genes Del' (in press). Ma, Q., Chen, Z. F., Barrantes, I. B., de la Pompa, .J. L. and Anderson, D. J. ( 1998). Ne11rogeni11 I is essential for th e determination of neuronal precursors for proximal crania l sensory ganglia. Ne11ro11 20, 469-482 . Ma, Q., Kintner, C. and Anderson, D. J. ( 1996). Identification of 11e111vgeni11. a verte brate neurona l determination ge ne. Cefl 87 . 43 -52. Marusich, M. F. and Weston, J. A. ( 199 1). Development of the neural crest. Cu n : Opi11 . Genet. Dev. l. 22 1-229. Matsumoto, S. G. ( 1994a). Ne urona l diffe rentiat ion in cultures of murin e neural crest. II. De velopmenl of capsaicin-sensitive neuron s. De1•. Brain Res. 83. 17 -27. Matsumoto, S. G. ( 1994b ). Ne uronal differentiation in cu lt ures of murine neural cresl. I. Ne urotran smitte r ex press ion. De1·. Bmi11. Res. 83. 1-1 6. McEvilly, R. J. , Erkman, L., Luo, L., Sawchenko, P. E., Ryan, A. F. and Rosenfeld, M. G. ( 1996). Requirement for Brn-3.0 in differentiation a nd survival of sensory and motor-ne uron s. Nature 384. 574.577_ Morrison, S. J., White , P. M. , Zock, C. and Anderson, D. J. ( 1999). Prospective identitica1ion. isolation by flow cytome1ry. and in vivo se lfren ewal of mult ipotent m,1mmalian neural cres1 stem cells . Cell 96. 737. 749 . Murphy, M., Reid , K., Ford, M., Furness, J.B. and Bartlett, P. F. ( 1994). FGF2 regulat es proliferati on of neural crest cells. with subsequent neuronal differen1iation regu la1ed by LI F o r related factors. Deve/opmelll 120. 35 19 3528. Murphy, M. , Reid, K., Hilton , D. J. and Bartlett, P. F. ( 1991 ). Gener.uion of se nsory neuron s is s1imulatec.l by leukemia in hibi mry factor. Proc. Nari. Acad. Sci. USA 88 , 3498-3501 . Nawa, H. and Patterson, P. H. ( 1990). Sepa rati on and partial characterization of neuropeptide-inc.lucing fac1ors in hea rt cell conditioned medium. Ne11m11 4, 269-277 . Novitch , B. G., Mulligan, G. J., Jacks, T. and Lassar, A. ll. ( 1996). Skeletal musc le ce ll s lacking the retinob lastuma protein display defeL"ts in mu scle gene expression anti accumula1e in S anc.l G2 phases of the cell cycle. J. Cell Biol. 135 . 441-456. Pachnis, V., Mankoo, 8. and Costantini, F. ( 1993 ). Exp ress ion o f the c-r<'I prolO•o ncogene during mouse embryoge nesis. De1·c:lo1,111e111 119, 1005101 7. Parysek, L. M. and Goldman, R. D. ( 1988). Distribution of a novel 57 kDa ln1ermediatc Filament (IF) protein in the nervous system. J. Ne11rusci. 8. 55 5-563. Perez, S. E., Rebelo, S. and Anderson , D. J. ( 1999 ). Earl y specification of senso ry neu ron fate revealed by express ion and functi on of neuroge nin s in the chick embryo. De.,eJop111e111 126. 17 15-1 728 . Reissman. E., Ernsberger, U. , Francis-West, P. H., Rueger, D., Brickell , P. D. and Rohrer1 H. ( 1996). Invol veme nt of bone morphogenetic protei n•4 anti bone morphogenelic protein-? in the differe ntiation of th e adre nerg ic phenotype in developing sy mpathelic neurons. Develop111e111 122. 20792088. Reshef, R., Maroto, M. and Lassar, A. B. ( 1998). Regu lat ion of dorsal semitic cell fntes: BMPs anti Noggin co ntrol the timin g and panern of myogenic regulator expre ssion. Genes De1~ 12. 290-303. Saito, T., Greenwood, A., Sun, Q. and Anderson, D. J. ( 1995). Ide nt ificat ion by differential RT•PCR of a novel paired homeoc.lornain protein specificall y expressed in se nsory neuron s and a subset of th eir CNS targets. Mo /. Cell. Ne11rosci. 6. 280-292. Schweizer, Ayer-LeLievre a nd Douarin, L. ( 1983). Restrictions of developmenrnl capacities in the dorsal root ganglia dur ing the course of development. Cell Differe111iatio11 13. 191-200. Shah, N. M. and Anderson, D. J . ( 1997). In tegration of multi ple instruct ive cues by ne ural crest stem ce lls re veals ce ll- in tri nsic biases in re lmive grow1h facto r respo nsiveness. Proc. Natl. Acad Sc:i. USA 94. 11369-1 1374. Shah, N. M., Groves, A. and Anderson, D. J. ( 1996). Alternative neural crest ce ll fates are instructi ve ly promoted by TGFP su perfam il y mem be rs. Cell 85 . 331-343 . Shah, N. M., Marchionni, M.A., Isaacs, I., Stroobant, P. W. and Anderson1 D. J. ( 1994 ). Glial growth factor res tricts mammalian neu ral crest stem cells to a glial fate. Cell 77. 349-360. Sieber-Blum, M. ( 1989). Com mitm enr o f neural cres t cells to the sensory neuron lineage. Science 243. 1608-1610. Sieber-Blum, M. ( 1990 ). Mec hanisms of Neural Crest Di ve rsitication . In Comments Deve/opmenra/ Neurobiology. Vol. 1 pp. 225- 249. Gordon and Breac h Science Publishers. SA. Great Britai n. Sieber-Blum, M., Ito, K., Richardson, M. K., Langtimm, C. J. and Duff, 54 Determination of the sensory lineage in vitro 3559 R. S. ( 1993). Distribution of pluripotent neural crest cells in the embryo and the role of brain-derived neurotrophic factor in the commitment to the primary sensory neuron lineage. J. Neurobio/. 24. 173-184. Snider, W. D. ( 1994). Functions of th e neurotrophins during nervous system deve lopment - what the knockouts are teaching us. Cell 77. 627638. Sommer, L., Ma, Q. and Anderson, D. J. ( 1996). 11eurogen i11s. a novel family of ato11al-related bHLH transcription factors, are putative mammalian neuronal determ ination genes taht revea l progenitor cell heterogeneity in the deve loping CNS and PNS . Mo/. Cell. Neurosci. 8. 221-241 . Stemple, D. L. and Anderson, D. J. ( I 992). Isolation of a stem cell for neurons and glia from the mammalian neural crest. Ce// 71. 973-985. Tiveron, M.-C., Hirsch, M.-R. and Brunet, J.-F. ( 1996). The expression pattern of the transcription factor Phox2a delineates sy naptic pathways of the autonomic nervous sys tem . J. Neurosci. 16, 7649-7660. Valarcht!, I., Tissier-Seta, J.-P., Hirsch, M.-R., Martinez, S., Goridis, C. and Brunet, J.-F. ( I 993). The mouse homeodomain protein Phox2 regulates NCAM prommer activity in concert with Cux/CDP and is a putative determinanr of neurotransmitter phenotype. Developme111 119. 881-896. Varley, J.E. and Maxwell, G.D. (1996). BMP-2 and BMP-4. but not BMP6, increase the number of adrenergic cells which develop in quail trunk neural crest cu ltures. Exp. Neural. 140, 84-94. Varley, J. E., Wehby, R. G., Rueger, D. C. and Maxwell, G.D. (1995). Number of adrenergic and islet- I immunoreacLive cells is increased in avian trunk neural crest cultures in the presence of human recombinant osteogenic protein-I. Dev. Dy1wm. 203. 434-447. Weston, J. A. ( I 99 I). Sequential segregation and fate of developmentally restricted intermediate cell populations in the neural crest lineage. Cun: Topics Dev. Biol. 25 . 133-153. Xiang, M. Q., Gan, L., Zhou, L., Klein, W. H. and Nathans, J. ( 1996). Targeted deletion of the mouse POU domain gene Brn-3a causes a selective loss of neurons in the brain-stem and trigerninal ganglion, uncoordinated limb movement, and impaired suckling. Proc. Nari. Acad. Sci. USA 93. I 1950-1 1955. Xiang, M. Q., Zhou, L. J., Macke, J. P., Yoshioka , T., Hendry, S. H. C., Eddy, R. L., Shows, T. B. and Nathans, J. ( 1995). The Brn-3 family of POU-domain factors - primary structure, binding-specificity. and expression in subsets of retinal ganglion-cells and somatosensory neurons. J. Neurosci. IS, 4762-4785. Ziller, C., Dupin, E., Brazeau, P., Paulin, D. and Le Douarin, N. M. ( 1983 ). Earl y segregation of a neural precursor cell line in the neural crest as revealed by culture in a chemicall y defined medium. Cell 32, 627-638. Ziller, C., Fauquet, M., Kalcheim, C., Smith, J. and Le Douarin, N. M. ( I 987). Cell lineages in pe ripheral nervous system on togeny: mediuminduced modulation of neurona l phenotypic expression in neurnl cres t cultures. Dev. Biol. 120. 10 I - I 11. Zimmerman, L. B., Dejesus-Escobar, J. M. and Harland, R. M. ( 1996). The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein-4. Ce// 86. 599-606. 55 Chapter 3 The regulation of subtype characteristics in sensory neurons generated from mammalian neural crest in vitro 56 SUMMARY Sensory neurons that differentiate from rat neural crest cells in a chemically defined medium exhibit characteristics of the muscle afferent sensory subtype, including trophic dependence on NT-3 and BDNF, expression of the ETS-domain transcription factor ER81 , and lack of trkA expression (Greenwood, A. L., Turner, E. E., and Anderson , D. J. (1999) Development 126, 3545-3559). NT-3 and BDNF can dramatically upregulate the expression of ER81 in such neurons after they have differentiated. In contrast, in medium containing either serum or BMP2 many of the sensory neurons that differentiate do not express muscle afferent markers , but rather express trkA and VRl , markers of the cutaneous afferent sensory subtype. Unlike the neurotrophins, serum acts on the neural tube to induce the production or migration of progenitors of trkA + (as well as trkA) sensory neurons. Importantly, NT-3 and BDNF do not prevent the development of trkA+ sensory neurons in serum-containing cultures. Furthermore, these neurotrophins are not required for the development of trkA sensory neurons . These data suggest that NT-3 and BDNF are neither necessary nor sufficient to specify sensory subtypes in vitro, but rather up-regulate expression of certain markers in neurons already specified for a muscle afferent fate . In contrast, serum and BMP2 act on the neural tube to influence, directly or indirectly, the fate of neural crest cells that subsequently migrate. These data provide evidence that the generation of different subclasses of sensory neurons from neural crest cells can be controlled by manipulation of the culture environment, and are consistent with the idea that these subclasses are generated from multiple subpopulations of sensory neuron precursors. INTRODUCTION An important goal of developmental neurobiology is to identify extrinsic signals that regulate particular aspects of neuronal identity and to discover when and how they exert 57 their effects. Such signals may regulate the acquisition of either single or multiple neuronal characteristics and may act on undifferentiated precursors or on alreadydifferentiated neurons. It is also important to identify which cells can and cannot respond to these signals. Environmental cues that affect the development of the two main classes of neurons derived from the neural crest, sensory and autonomic neurons, have received significant attention (Patterson, 1992; Le Douarin et al., 1994; Anderson, 1997; Groves and Bronner-Fraser, 1999). However, the factors that regulate neuronal subtype diversity within each class are still largely unknown. Sensory neurons of the dorsal root ganglia (DRG) exhibit considerable phenotypic heterogeneity. In particular, two major sensory subtypes have been well described (Scott, 1992). One subtype consists of muscle afferents that receive information about spatial position (proprioception) and send their central projections to the ventral spinal cord. These neurons tend to have large-diameter cell bodies in vivo, depend on neurotrophin-3 (NT-3) for survival and express the corresponding receptor, trkC (Mu et al., 1993; Klein et al., 1994; McMahon et al., 1994). Many of these neurons also express ETS-domain transcription factors, such as ER81 and PEA3, the expression of which is regulated by unidentified target-derived factors in vivo (Lin et al., 1998). Another class of sensory neurons is comprised of the cutaneous afferents that transduce painful stimuli (nociception) and make central connections in the dorsal horn of the spinal cord. Many of these neurons have small-diameter cell bodies, respond to capsaicin, and express the vanillinoid receptor VRl that mediates this sensitivity (Tominaga et al., 1998). The survival of most nociceptive, cutaneous afferent neurons depends on nerve growth factor (NGF) during at least some point in their development, and many of these neurons express the high affinity NGF receptor, trkA (Smeyne et al., 1994). Other types of sensory neurons, such as those that mediate the sense of touch (mechanoreception), appear to be more heterogeneous in their properties. 58 It is not known how sensory subtypes are initially specified during development. It has been suggested that neurotrophins may specify sensory subtype identity as well as supporting the survival of different subtypes (Lewin, 1996). Although this model is intuitively appealing, it has been recently demonstrated that the initial specification of nociceptive sensory neurons does not require NGF (Patel et al., 2000). The situation is not so clear for other neurotrophins. Brain derived neurotrophic factor (BDNF) has been implicated in the specification of generic sensory neurons from quail neural crest cells both in vitro and in vivo (Sieber-Blum, 1991). It has not been determined, however, whether early exposure of sensory precursors to NT-3 or BDNF affects the subsequent development of sensory subtype characteristics. We have begun to investigate how extrinsic factors regulate sensory subtype characteristics in neurons that develop from rat neural crest cells in vitro. Previously, we characterized the sensory neurons that develop in a chemically defined medium. We found that these neurons resemble muscle afferent sensory neurons by numerous criteria including trophic dependence on NT-3 and/or BDNF, the expression of the ETS-domain transcription factor ER81, and the lack of expression of trkA (Greenwood et al., 1999). Cutaneous afferent-like sensory neurons were, however, conspicuously absent from these cultures. This observation prompted us to look for factors that would promote the development of the cutaneous afferent-like sensory subtype. We were interested to know whether signals that could positively affect the development of one sensory subtype would negatively affect the development of the other. Such an observation would imply that both muscle and cutaneous afferent sensory neurons derive from a common precursor pool ; alternately , the independent regulation of muscle and cutaneous afferent development might imply that these neurons come from different precursors among the neural crest. We have now identified conditions that promote the development of the cutaneous afferent sensory neuron subtype. Our data are consistent with the idea that 59 there are different subclasses of sensory neuron precursors in the neural crest, though these precursors are not necessarily restricted to particular sensory subtypes. MATERIALS AND METHODS Rat neural tube explant cultures Neural tubes were isolated from the trunk level of El0.5 Sprague-Dawley rats as described previously (Stemple and Anderson, 1992; Greenwood et al., 1999). Explanted neural tubes were plated in 35 mm diameter Coming dishes on a fibronectin (BTI) substrate, and grown at 37°C in 5% CO/ 95 % air. Medium and Additives Defined medium The recipe for defined medium (DM) was modified from that described previously (Stemple and Anderson, 1992; Greenwood et al. , 1999) and is as follows: to L-15CO 2 (Hawrot and Patterson, 1979) add 100 µg/ml transferrin (Calbiochem), 5 µg/ml insulin (Sigma), 16 µg/ml putrescine (Sigma), 20 nM progesterone (Sigma) , 30 nM selenious acid (Sigma) 1 mg/ml bovine serum albumin, crystallized (Gibco/BRL), 39 pg/ml dexamethasone (Sigma), 5 µg/ml oc-d-1-tocopherol (Sigma), 63 µg/ml Bhydroxybutyrate (Sigma), 25 ng/ml cobalt chloride (Sigma), I µg/ml biotin (Sigma), 10 ng/ml oleic acid (Sigma), 3.6 mg/ml glycerol, 100 ng/ml oc-melanocyte-stimulatinghormone (Sigma) , 10 ng/ml prostaglandin El (Sigma), 67.5 ng/ml triiodothyronine (Aldrich), and 10 ng/ml basic fibroblast growth factor (bFGF, UBI). Additives Where indicated, DM was supplemented with one or more of the following: recombinant human bone morphogenetic protein-2 (BMP2 at 10 ng/ml; gift from Genetics Institute), brain derived growth factor (BDNF at 25 ng/ml) , neurotrophin-3, (NT3 at 25 ng/ml; BDNF and NT3 both provided by Dr. G. Yancopoulos at Regeneron Pharmaceuticals), 60 nerve growth factor (NGF at 25 ng/ml; UBI), and/or 10% fetal bovine serum (FBS; Hyclone, lot AHC817). Immunocytochemistry Cultures were fixed in 4% paraformaldehyde and then processed for double antibody labeling. All primary antibodies were incubated overnight at 4 °C. All rabbit polyclonal antibodies were detected using an HRP-conjugated, goat anti-rabbit secondary antibody purchased from Vector Labs, Inc., at 1:200 dilution for 1 hour at room temperature. Staining was detected using NiDAB and DAB substrates as described previously (Shah et al., 1994). In all cases the first antibody applied was developed with NiDAB (dark blue) and the second with DAB (brown). Affinity-purified rabbit anti-Brn-3 .0 (1: 1000; (Fedtsova and Turner, 1995)) was used to detect general sensory neurons. Rabbit polyclonal antibodies to ER81 (1:12,000; gift from Dr. T . Jessell; (Lin et al., 1998)), TrkA (1:5000; gift from Dr. L. Reichardt, (Clary et al., 1994)), and VRl (1:5000; gift from Dr. D. Julius) were used to detect these sensory subtype-specific antigens. RESULTS Defined medium promotes the development of sensory neurons that display characteristics of the muscle afferent subtype Our previous study indicated that a chemically defined medium can support the differentiation of sensory neurons from rodent neural crest cells derived from neural tube explants (Greenwood et al., 1999). However, these sensory neurons do not represent the full array of sensory subtypes that appear in vivo. Explant-derived sensory neurons, as identified by expression of the POU-domain transcription factor Brn-3.0 (Fedtsova and Turner, 1995; Xiang et al., 1995), differentiate in the absence of neurotrophins, but depend on NT-3 and BDNF, but not NGF, for long-term survival in culture. This 61 observation indicates that these neurons may be of the muscle afferent or mechanosensory fate, but not of the nociceptive cutaneous afferent fate. In addition, we have observed that sensory neurons that develop in defined medium in the presence of NT-3 and BDNF express ER8 l, (Fig. 1C) a marker that largely correlates with the muscle afferent fate in vivo (Lin et al., 1998), and do not express trkA (Fig. IA, B), a marker of the cutaneous afferent fate. Although it is not possible to precisely determine the functional specificity of the sensory neurons that develop in these cultures due to lack of appropriate targets, the ability to be supported by NT-3, the expression of ER81, and the lack of expression of trkA are consistent with these sensory neurons having a muscle afferent identity. To determine whether ER81 expression is an inherent characteristic of sensory neurons that differentiate in defined medium or whether it is regulated by NT-3 and BDNF, we looked at the expression of this transcription factor in newly-differentiated sensory neurons before the period of trophic dependence. Neural tube explant cultures were grown in defined medium in the absence of neurotrophins. The neural tube was removed from culture on day 2 to isolate the neural crest, and the cultures analyzed on day 3 for the presence of ER8 l and Bm-3.0 expression by antibody staining. As shown previously, the majority of Bm-3.0+ neurons in similar cultures are post-mitotic and express neurofilament at this time (Greenwood et al., 1999). In the absence of neurotrophins, the cultures contained many neurons that expressed Bm-3 .0 (Fig. 2A, arrows), but only a few that expressed low levels of ER81 (Fig 2A, arrowheads). However, as little as a 1 hour exposure to BDNF and/or NT3 on day 3 resulted in a striking upregulation of ER8 l in the majority of sensory neurons (Fig. 2B-D). Such a brief exposure to neurotrophins is unlikely to significantly affect survival or proliferation. Therefore, these data indicate that exogenously added NT-3 and BDNF are sufficient to 62 Brn-3.0 /trkA Brn-3.0 /trkA ,, B Er81 / trkA / . ,t.. ., . . • ·· / . • . •.• -. . . . .. . . -. .. . ~ •' •.. . - C Fig. 1 Defined medium supports the development of sensory neurons that have characteristics of the muscle afferent subtype. Rat neural tube explants were grown for 6 days in defined medium supplemented with NT-3 and BDNF (25 ng/ml ea.) and then double-labeled for Brn-3.0 (A, B; blue, nuclear) or Er81 (C; blue, nuclear) and trkA (A, B; brown, cytoplasmic) by immunocytochemistry. (A) The view shown includes the dorsal neural tube (NT; top) and neural crest outgrowth containing many Brn-3.0+/trkAcells (arrows). (B) Higher magnification shows that the Brn-3.0+ cells clearly do not stain for trkA. (B, inset) TrkA staining on serum-treated cultures serves as a positive control. (C) The neurons grown under these conditions express Er8 l. 63 regulate ER81 expression in differentiated sensory neurons, but are not required for their initial development. Serum promotes the development of sensory neurons that display characteristics of the cutaneous afferent subtype Other studies of mammalian sensory neurons that differentiate in neural crest cultures have observed that these neurons express markers of the cutaneous afferent subtype (Murphy et al., 1991; Matsumoto, 1994a; Matsumoto, 1994b ). In these studies, serum was included as a component of their medium. Therefore, we added serum to our medium to determine whether it would promote expression of cutaneous afferent markers and/or inhibit expression of muscle afferent markers. When neural tube explants were grown for 6 days in medium supplemented with 10% fetal bovine serum (FBS), we found that many Brn-3.0+ neurons developed in the neural crest outgrowth. Most of these neurons were also immunopositive for trkA (Fig. 3A, B) and some expressed VRl (Fig. 3C arrowheads), which is another marker of the cutaneous afferent phenotype. The Brn3.0+/trkA + sensory neurons in our explant cultures were found in large clusters adjacent to the neural tube (Fig. 3A). In the absence of exogenously added neurotrophins, sensory neurons in these cultures were only rarely immunopositive for ER8 l (Fig. 3D). When these cultures were exposed for 1 hour to BDNF and NT-3 on day 6, ER81 was upregulated in minority of the sensory neurons (Fig. 3E). Most trkA + neurons did not express ER81 (Fig. 3E, -/+; 85.2 ± 3.7% of total trkA + cells), but of the cells that were induced to express ER81, 80.4 ± 5.2% were trkA+ (Fig. 3E, +/+) and 19.6 ± 5.2% were trkA (Fig. 3E, +/-). These data suggest that most of the sensory neurons that develop in serum-containing medium have a phenotype that is consistent with the cutaneous afferent fate; that is, they express trkA but not ER81 even in the presence of NT-3 and BDNF. The neurons that co-express trkA and ER81 in these cultures may represent a distinct 64 Er81 / Brn-3.0 No Add / +BDNF 1hr • > / • / / > A • • B +NT-3 1hr +BDNF & NT-3 1hr • • C D Fig. 2 A one hour exposure to NT-3 or BDNF in the absence of the neural tube is sufficient to induce ER81 expression in sensory neurons grown in defined medium. The neural tubes were removed at day 2 from explants grown in defined medium; cultures were exposed to NT-3 and/or BDNF (25 ng/ml ea.) for 1 hour on day 3, fixed immediately afterward and processed for ER81 (blue, nuclear) and Brn-3.0 (brown, nuclear) antibody staining. (A) Only a small percentage of the sensory neurons (Brn-3.0 cells; arrows) are weakly positive for ER81 (arrowheads) in the absence of these factors. Cultures exposed for 1 hour to BDNF (B) or NT-3 (C) or both BDNF and NT-3 together (D) contain a much higher percentage of sensory neurons that strongly express ER81 . 65 Fig. 3 Serum-containing medium supports the development of sensory neurons that have characteristics of the cutaneous afferent subtype. Neural tube explants were grown for 6 days in medium supplemented with 10% FBS. (A) The view from the dorsal neural tube (NT, top) across the outgrowth shows many cells that co-express Brn-3.0 (blue, nuclear) and trkA (brown, cytoplasmic) in the area adjacent to the neural tube. (B) Higher magnification of similarly stained cultures shows that most Brn-3.0+ cells express trkA. (C) Some of the Brn-3.0+ cells in these cultures also express VRl (brown, cytoplasmic; arrowheads). (D) Most of the tr~ cells in these cultures do not express Er81 (blue, nuclear), although an occasional weakly Er81+ cell is observed (arrowhead). (E) When such cultures are exposed for 1 hour to NT-3 and BDNF (25 ng/ml ea.) on day 6, Er81 is upregulated in some trkA (+/- arrowhead) and some tr~ (+/+ arrowhead) cells. Most trkX cells do not express Er81 under these conditions (-/+ arrowhead). 66 subpopulation of trkA + neurons that co-express trkB or trkC, such as has been observed in vivo (McMahon et al., 1994; Wright and Snider, 1995). Serum is required for the development of significant numbers of trkA + sensory neurons in explant cultures To understand why trkA + sensory neurons develop in medium that contains serum, we examined whether serum is required and/or sufficient for the expression of trkA. As mentioned above, explants grown for 6 days in serum-free medium supplemented with NT-3 and BDNF do not contain trkA+ cells, suggesting that defined medium does not support the development of trkA + sensory neurons. Consistent with this observation, we found that explants grown for 4 days in defined medium in the absence of neurotrophins contained mostly trkA/Bm-3 .0+ cells. These cultures, however, do contain a small fraction of Bm-3.0+ cells that expressed trkA (approximately 10% ), but most of these neurons had only weak expression and only occasionally did we find a strongly positive cell (Fig. 4A, arrowhead). Overall, the trkA + neurons found in explants grown in defined medium were far less numerous and had much weaker staining than those observed in explants that had been exposed to 10% FBS for the first two days of culture. In such serum-treated cultures, 28 .0 ± 3% of the Bm-3.0+ cells co-expressed trkA, and many of these had strongly stained cell bodies and fibers (Fig. 4B, arrowheads) . Qualitatively, these cultures contained many more sensory neurons (Bm-3.0+ cells) of both trkA + and trkA (Fig. 4B arrows) subtypes, especially in the caudal part of the explants. Thus, exposure to serum for the first 2 days of culture is required for the development of large numbers of strongly trkA + sensory neurons. There are two general mechanisms by which serum may promote the development of significant numbers of trkA + cells in ex plant cultures . First, serum may directly induce trkA expression in some sensory neurons or precursors that have already 67 migrated from the neural tube. Second, serum may induce the neural tube to produce sensory precursors, some of which have the potential to differentiate into the trkA + subtype. To address the first possibility, we allowed neural crest to migrate from the neural tube in defined medium for two days and then cultured these explants in medium containing 10% FBS until day 4. In contrast to explants that were exposed to serum for the first two days of culture, explants that were exposed to serum during the second two days of culture contained very few trkA+/Bm-3 .o+ cells (Fig. 4C). Occasionally, however, a few trkA + cells were found adjacent to the neural tube under these conditions (Fig. 4C, inset). We hypothesized that these cells might have migrated from the neural tube after addition of the serum. Consistent with this idea, when the neural tube was removed from explant cultures at day 2 (before the exposure to serum) none of the Bm3.0+ cells in the outgrowth expressed trkA at the end of the culture period at day 4 (Fig. 4D). Serum, therefore, is required for development of significant numbers of trkA + sensory neurons in explant cultures, but is not sufficient to induce the expression of trkA in sensory neurons that have differentiated in defined medium. TrkA + sensory neurons develop in defined medium from explants that have been treated with BMP2 One additional feature of serum-treated explants was that they contained autonomic neurons as assayed by Phox2a expression (data not shown). Such neurons are absent in explants grown in serum-free medium. As shown previously, autonomic neurons can be induced to develop in explant cultures grown in serum-free medium by exogenous addition of bone morphogenetic protein-2 (BMP-2; Greenwood et al., 1999). We were therefore interested to know if BMP2, like serum, could promote the development of trkA + sensory neurons in ex plants grown in defined medium in addition to promoting the development of autonomic neurons. As mentioned above, there is a low incidence of 68 TrkA / Brn-3a0 DM FBS--DM DM--FBS DM remove NT• FBS A C Fig. 4 Serum induces the development of trkA+ sensory neurons in explant cultures, but is not sufficient to induce trkA in sensory neurons that have previously differentiated in defined medium. Explants were cultured for 4 days then double-labeled for trkA (blue, cytoplasmic) and Bm-3.0 (brown, nuclear) by immunocytochemistry. (A) Most Brn-3.o+ cells found in cultures grown continuously in defined medium (DM) for 4 days do not express trkA, although a few double-positive cells are observed (arrowhead). (B) Cultures grown for the first 2 days in the presence of serum (FBS) then in DM until day 4 contain many trkK/Brn-3.0+cells (arrowheads) and many trkA-/Brn-3.0+cells (arrows). (C) However, when serum is added instead during the second two days of culture, most of the Brn-3.0+ cells do not express trkA. (C, inset) A few trkK cells in these cultures are observed near the neural tube. (D) Cultures from which the neural tube was removed at day 2 before the addition of serum contained only Brn-3.0+cells that did not express trkA. 69 trkA expression in the sensory neurons that differentiate in defined medium in the absence of exogenously added BMP2 (Fig. SA, C). When BMP2 is included in the medium for the first two days of culture, the percentage of Bm-3.0+ cells that express trkA at day 4 increases significantly from 10.6 ± 8.4% to 31.2 ± 11.8% (Fig. 5B, C). Due to the large and variable number of sensory neurons per explant (many hundreds to thousands), it has proven impractical to quantitate the effect of different conditions on total sensory neuron number. Qualitatively, however, the number of sensory neurons (total Bm-3.0+ cells) appeared to increase in cultures treated with BMP2, though not as dramatically as in cultures treated with serum. In addition, BMP2, like serum, does not induce the expression of trkA in sensory neurons that have already differentiated in defined medium (c.f. Fig. 4; data not shown). These results indicate that BMP2 can affect the development of sensory neurons and their acquisition of subtype characteristics in a manner similar to serum. Serum acts on the neural tube to promote its ability to generate trkA + sensory neurons The foregoing experiments have demonstrated that when neural tube explants are exposed to serum during the first two days of culture, many trkA + sensory neurons can be found in the outgrowth at day 4 (Fig. 4B). As mentioned previously, it is possible that serum may increase the number of trkA + sensory neurons that develop in the outgrowth by affecting the type or number of precursors that migrate from the neural tube. In this case, the main effect of serum would be to act on the neural tube, the source of neural crest, rather than on already-emigrated neural crest cells. Examination of explants during the culture period suggested that the neural tube likely responds to serum by producing precursors that have the potential to differentiate into trkA + neurons. When ex plants were cultured for the first 2 days in medium supplemented with 10% FBS, the neural crest 70 TrkA /Brn-3.0 A t C\I a.. :E co B Fig. 5 BMP2 induces the development of trkK sensory neurons in explant cultures grown in defined medium. Explants were cultured for 4 days in either defined medium (DM) or DM supplemented with BMP2 (10 ng/ml) for the first two days. (A, B) Explants were double-labeled with antibodies for trk.A (blue, cytoplasmic) and Bm-3.0 (brown, nuclear). (A) Explants grown in DM contain Brn-3.0+ cells that are mostly trkA . (B) Explants exposed to BMP2 for the first two days contain a larger percentage of Brn-3.0+ cells that express trkA. (C) The graph indicates the percentage of Brn-3.0+ cells in each condition that express trkA strongly (dark blue part of bar) and weakly (light blue part of bar). n=6 explants for each condition; *P<0.01. Error is represented as standard deviation. 71 outgrowth was most prevalent along the rostral portion of the explant but only beginning to migrate from the caudal portion. However, following withdrawal of serum at day 2 and further culture until day 4, many cells migrate from the caudal portion of the explants. Most importantly, it is the caudal region that contains the most trkA and Brn3.0 expression, although some sensory neurons were observed at all levels along the neural tube ex plant. These observations suggest that the precursors of trkA + sensory neurons are likely to migrate from the caudal neural tube after the serum withdrawal at day 2. To directly demonstrate that trkA + neurons are derived from cells that migrate from the neural tube after serum is withdrawn, we removed the neural crest outgrowth from explants at day 2 at the time of serum withdrawal and assayed whether the subsequent outgrowth contained trkA + neurons. As mentioned previously, by day 4 the caudal region of unoperated ex plants contained significant numbers of trkA +/Brn-3.0+ (Fig. 6A, arrowheads) and trkA/Brn-3.0+ (Fig. 6A, arrows) cells. Similarly, the caudal outgrowth of explants from which neural crest had been removed at day 2 contained many trkA+ (Fig. 6B, arrowheads) and trkA· (Fig. 6B, arrows) sensory neurons. These explants contained almost as many cells as seen in unoperated cultures. Therefore, sensory cells can migrate from the neural tube after serum has been withdrawn from culture and then develop into sensory neurons that express trkA. Note that the continued presence of serum is not required for the late migration of these precursors and their differentiation into trkA+ cells. By contrast, when the neural tube, rather than the neural crest, was removed from explants at day 2 at the time of serum withdrawal, the isolated outgrowth contained a small population of Brn3.0+ sensory neurons (Fig. 6C, arrows) that only rarely expressed trkA (Fig. 6C, arrowheads). This reciprocal experiment suggests that the neural tube is required for the development of trkA + sensory neurons because it produces most of the precursors to these neurons after the second day of culture. 72 The neurotrophins BDNF and NT-3 induce expression of ER81 but do not prevent expression of trkA in sensory neurons We have shown that NT-3 and BDNF regulate the expression of ER81, one characteristic of the muscle afferent fate. Although al hr exposure to BDNF and NT-3 did not abolish the expression of trkA + in sensory neurons that had been cultured for 6 days in serum (Fig. 3), it was not clear whether these neurotrophins might prevent the development of trkA + sensory neurons if added at earlier times or for longer durations. Therefore, to determine whether these factors would prevent the development of trkA + sensory neurons, we included these factors in the medium for the duration of the culture period. Cultures grown in serum for 2 days and then in defined medium until day 4 contain many trkA and trkA + sensory neurons in the absence of exogenously added neurotrophins (Fig. 7A). The expression of ER81 under such conditions is low (approximately 15%) and typically weak (Fig. 7B, note lack of blue nuclei). When explants were grown in similar conditions in the continued presence of 25 ng/ml of BDNF and NT-3 for four days, ER81 was strongly upregulated in many sensory neurons (Fig. 7D). However, such NT3/BDNF treated cultures still contained many trkA +/Brn-3.o+ cells (Fig. 7C). However, such NT-3/BDNF treated cultures still contained many trkA+/Brn-3.o+ cells (Fig. 7C). BDNF and NT-3 have a similar affect on cultures grown in the continued presence of serum for 6 days; they do not abolish trkA expression or delay its onset although they do promote the presence (likely via survival) of a very large number of trkA neurons that express ER81 (data not shown). These data indicate that NT-3 and BDNF can positively regulate the development of the muscle afferent subtype without preventing the development of the cutaneous afferent subtype. 73 ventral rostral t trkA / Brn-3.0 caudal 1 - - - - - - - - - - - . - - - - - - - - - - T - - - - - - - - - - - l NT removed dorsal unoperated --- - ....... day2 / / day4 / / crest removed --day2 NT removed day2 / day 4 A B C Fig. 6 Serum acts on the neural tube rather than on already-migrated neural crest to induce the development of trkK sensory neurons. Explants were grown in medium containing 10% FBS for two days, at which time the neural crest or the neural tube was removed from some cultures. Operated and unoperated explants were then cultured until day 4 in defined medium and stained for trkA (blue, cytoplasmic) and Brn-3.0 (brown, nuclear) expression by immunocytochemistry. Schematics to the left indicate the manipulations performed and the resulting morphology of the explants. (A) Unoperated 74 ex plants contain many Brn-3.0+/trkA + (arrowheads) and Brn-3.0+/trkA (arrows) cells in the outgrowth from the caudal neural tube (NT, top). (B) Explants from which the neural crest was removed at day 2 also contain many trkA+ (arrowheads) and trkA· (arrows) sensory neurons (Brn-3.o+ cells) in the outgrowth from the caudal part of the neural tube. (C) Cultures from which the neural tube was removed at day 2 do not contain as many Brn-3.0+ cells. Most of these do not stain for trkA (arrows) although a few Brn-3.0+/trkA+ cells can be found (arrowheads). DISCUSSION We have developed an in vitro assay system in which to study the factors that regulate the subtype characteristics of sensory neurons that differentiate de novo from the neural crest. Here we show that both serum and BMP2 promote the development of the cutaneous afferent sensory neuron subtype, as assayed by expression of trkA. Serum appears to exert its effects by inducing the neural tube to produce sensory precursors that then differentiate into both trkA + and trkA· sensory neurons. Neither serum nor BMP2 induces trkA expression directly in neurons that have previously differentiated in defined medium. In contrast, we have found that BDNF and NT-3 can directly induce expression of ER8 l, a marker of the muscle afferent sensory neuron phenotype, in alreadydifferentiated neurons. NT-3 and BDNF do not prevent the development of sensory neurons that express trkA. Furthermore, serum-treated cultures contain many neurons that can respond to BDNF and NT-3 by upregulating ER8 l. These data indicate that serum/BMP2 modulate, either directly or indirectly, early aspects of sensory subtype identity while neurotrophins modulate later characteristics of sensory subtype identity. 75 '"C '"C <( 0 z C") i-!. z ~ LL z 0 en + Fig. 7 NT-3 and BDNF induce Er81 expression but do not prevent the development of trkK sensory neurons. Explants were grown in the absence (No add) or continued presence of NT-3 and BDNF (25 ng/ml ea.) in serum washout conditions (2 days in medium containing 10% PBS followed by 2 days in defined medium), and then doublelabeled with antibodies to trkA (blue, cytoplasmic) or Er81 (blue, nuclear) and Brn-3.0 (brown, nuclear). (A) Explants grown in the absence of NT-3 and BDNF contain many trkK/Brn-3.0+and trkA /Brn-3.0+cells. (B) Similar explants grown without factors contain sensory neurons (Brn-3.0+ cells) that express very little Er81. (C) In the continued presence of NT-3 and BDNF, many trk.K /Brn-3.0+ cells still are observed. (D) A large percentage of the sensory neurons in cultures exposed to NT-3 and BDNF express Er8 l. 76 Models for the mechanism by which serum and BMP2 affect the development of sensory neuron subtypes The ability of transforming growth factor-B (TGF-B) family members to affect neural tube patterning (Liem et al., 1997; Neave et al., 1997; Lee et al., 1998; McMahon et al., 1998; Nguyen et al., 2000), migration of the neural crest (Sela-Donenfeld and Kalcheim, 1999), and specification of peripheral neuron fates from the neural crest (Reiss man et al. , 1996; Shah et al., 1996; Schneider et al., 1999) has been well established. It is not surprising then that we have observed that BMP2, a member of the TGF-B family, and serum, which has been shown to have BMP-like activities (Ai et al., 1999), affect the development of sensory neurons in neural tube explant cultures. Serum and BMP2 have similar effects on the development of peripheral neuron subtypes in these cultures. Both induce the neural tube to produce more sensory precursors than would otherwise migrate in defined medium; these develop into both trkA + and trkA neurons. Both induce the development of autonomic neurons from emigrated neural crest cells. Our evidence suggests that the ability of serum/BMP2 to affect the development of sensory subtypes in explant cultures is a result of an early action on the neural tube, and not due to an action on cells that have already migrated from the neural tube. There are, however, a number of cellular mechanisms by which this effect could be achieved. There are two classes of possible mechanisms by which serum and BMP2 may act on the neural tube to cause it to produce sensory precursors, some of which differentiate into trkA+ sensory neurons. These models must also account for the fact that many trkA sensory neurons develop in the absence of serum/BMP2. The first class of models supposes that sensory neuron subtype fates are specified before neural crest precursors migrate from the neural tube. Serum/BMP2 could play a direct role in the specification of the cutaneous afferent fate from some neural tube cells only so long as there were already inherent differences that assured the development of the muscle afferent subtype 77 (Fig. SA). Alternately, serum/BMP2 may not instruct the initial specification of sensory precursors, but may promote the migration of precursors that are already specified to produce trkA+ neurons (Fig. SB). Specification could also be achieved indirectly via another cell type (Fig. SC). In addition, serum/BMP2 could somehow increase the number of premigratory neural crest, and lateral interactions within this population would specify some precursors to the cutaneous afferent fate (Fig. SD). Given the previously demonstrated actions of BMP family members on the development of the neural tube and crest, this class of models seems plausible, though difficult to test. Possibly a clonal analysis of the sensory subtype fates of premigratory neural crest could distinguish between these models. At this point, we favor the model in which serum/BMP2 affects the migration of precursors from the neural tube (Fig. B) because the outgrowth of serum/BMP2 treated explants appears generally larger and contains more sensory neurons than in ex plants grown in the absence of these factors. In the second class of models, serum may affect some aspect of neural tube or neural crest development that, in tum, specifies the subtype fate of sensory precursors after they migrate from the neural tube. For example, serum may induce the neural tube to produce some sort of signal that instructs sensory precursors or neurons in the outgrowth to assume a cutaneous afferent fate. A soluble signal (Fig. 9A) seems unlikely, however, since it should constrain the location of trkA + cells to the area immediately adjacent to the neural tube. Although we do observe such a phenomenon in cultures grown in the continued presence of serum, the trkA+ cells in cultures exposed to serum or BMP2 for only 2 days seem well-distributed in the outgrowth, often far from the neural tube. It is also possible that the neural tube could produce a trkA-inducing cue in the form of a particular type of dorsal neural tube cell that, when contacted by sensory axons, specifies the cutaneous subtype fate (Fig. 9B). This too seems extremely unlikely, especially in light of the fact that sensory neurons in vivo express trkA immediately after 78 A. Serum specifies a subytype of sensory precursors within the NT. B. Serum causes a specified subtype of sensory precursor to migrate from the NT. •• ••• ~y · ~ : • • y •• • • •• C. Serum causes a non-sensory cell within the NT to specify the subtype of sensory precursors. . ./ y' °-.Af~ i'!' 1• •'. / • .., 1'. i' • • •• •• • • •• • ••• • • • •• • • • •• D. Serum produces more unspecified sensory precursors within the NT and lateral interactions specify subtype before they migrate. . ·~ ~ , QK:. • • . . . ve • • • •• • • •• • ••• • ... .. k_ y ./ y' ' y ' : '' Fig. 8 Models in which sensory subtypes are specified before precursors migrate from the neural tube. Grey regions indicate the neural tube. Circles indicate cells: purple represents unspecified neural crest, blue represents sensory precursors specified to the trkA" fate, red represents sensory precursors specified to the trkA+ sensory fate. An "X" represents an act of specification. These models assume that the default sensory subtype fate is to become a trkA" sensory neuron. (A) Serurn/BMP2 may directly specify the trkK fate only if there is already subtype heterogeneity among the potentials of premigratory sensory precursors. (B) Subtype fates may be prespecified among sensory precursors within the neural tube by an unknown mechanism. Serurn/BMP2 may cause sensory precursors to migrate from the neural tube including those with the potential to 79 make trkA + sensory neurons. (C) Serum/BMP2 may directly or indirectly cause a nonsensory cell within the neural tube to specify the trkA+ subtype fate from some premigratory neural crest precursors. (D) Serum/BMP2 may promote the development of more premigratory neural crest cells that are not specified to any particular sensory subtype fate. Lateral interactions within this population may specify some to produce trkA + sensory neurons upon migration. differentiation before their axons reach central targets (Farinas et al., 1998). Other ways in which transient exposure of the neural tube to serum could affect the subtype fate of sensory precursors after they have migrated would be by increasing the density of these precursors to promote lateral interactions (Fig. 9C), or by promoting the specification or migration of a non-sensory cell that directly induces the cutaneous afferent fate (Fig. 9D). The ability to isolate sensory precursors from the migratory neural crest and grow them at different densities in the presence or absence of the neural tube would allow one to distinguish between these models. The role of NT-3 and BDNF in the development of sensory neuron subtypes Our results point to a role for the neurotrophins BDNF and NT-3 in the regulation of the muscle afferent phenotype. We have shown previously that the explant-derived sensory neurons that develop in defined medium depend on these factors for their long term survival in culture (Greenwood et al., 1999). In this study we find that a short (1 hour) exposure of these sensory neurons to NT-3 and BDNF is sufficient to upregulate expression of ER8 l . Given the rapidity of the response, it is likely that these factors are acting directly on the neurons themselves to upregulate expression of ER8 l rather than promoting the survival or proliferation of ER81 + cells. Interestingly, in chick, ER81 80 A. Serum causes the NT to produce a subtype-inducing diffusable signal. B. Serum causes the NT to produce a subtype-inducing contact-mediated signal. _,,, ......_______ •••• • • • • •• • • •• • •• C. Serum causes the production or migration of more sensory precursors. Subtype is specified after migration via lateral interactions among these cells. D. Serum causes the production or migration of a non-sensory cell that specifies subtype fate of sensory precursors. • • •• • y --· •I • • o ••• , _,,, o• oo • ·• : • • Jv • • a• • • • • •0 •i • • • k_ y' • • y y . : Fig. 9 Models in which sensory subtype fates are specified after precursors migrate from the neural tube. (A) In the presence of serum/BMP2, the neural tube may produce a soluble factor that specifies the trkA+ subtype fate from sensory precursors in the outgrowth. (B) In the presence of serum/BMP2, a cell type develops within the neural tube that specifies the trkA+ subtype fate from sensory neurons upon contact with their central projections. (C) Serum/BMP2 may cause the neural tube to produce more sensory precursors. Lateral interactions within this population after they migrate specifies sensory subtype fate. (D) Serurn/BMP2 may promote the development of a non-sensory, migratory progenitor. Signals from this cell type specify the trkA+ subtype fate from sensory precursors after the cells have migrated from the neural tube. 81 expression is found in only a subset of the neurons that express trkC (Lin et al., 1998), and therefore there appear to be some neurons that are dependent upon NT-3 for survival but which do not upregulate ER8 l. Perhaps in this system there is a lower threshold of signaling required for survival than for ER81 expression. In contrast, recent data from mouse indicates that all trkC+ sensory neurons express ER81 (S. Arber and T. Jessell, in preparation). We have observed that the ability to induce ER8 l is not strictly specific to NT-3 and BDNF. A l hour exposure to NGF can elicit a modest but significant response (data not shown), and ER81 is also expressed in cultures treated with glial cell line derived neurotrophic factor (GDNF; data not shown). It is possible that multiple types of receptor tyrosine kinases can transduce signals to upregulate this gene. The mechanisms that restrict the expression of ER8 l are not yet understood but could involve lineage restrictions that prevent subtypes of sensory neurons from expressing the gene in response to neurotrophins. Our data suggest that BDNF and NT-3 are not required for the initial specification of the muscle afferent phenotype. Exogenous addition of these factors is not required for the differentiation of sensory neurons that ultimately will depend on them for survival and will respond to them by upregulating ER81. We cannot rule out the formal possibility that BDNF and NT-3 play an instructive role in specifying the fate of a subset of sensory precursors, but our results suggest they are not categorically required for development of the muscle afferent fate. Furthermore, these neurotrophins do not prevent the development of the cutaneous afferent (trkA +) sensory neuron fate . This observation may imply that BDNF and NT-3 are not instructive factors for sensory subtype fates because they do not seem to affect the decision of precursors to become muscle or cutaneous afferents. Alternatively, BDNF and NT-3 might be instructive for the muscle afferent fate but they do not prevent the development of trkA+ sensory 82 neurons because some sensory precursors are determined for the cutaneous afferent fate when they migrate from the neural tube. Clearly, neurotrophins play extremely important roles in the overall development of sensory neuron heterogeneity even if they are not responsible for the initial differences between sensory subtypes. It will be important to determine when sensory neuron subtypes first become different from each other and what signals impose the initial heterogeneity. 83 REFERENCES Ai, X., Cappuzzello, J. and Hall, A. K. (1999). Activin and bone morphogenetic proteins induce calcitonin gene-related peptide in embryonic sensory neurons in vitro [In Process Citation]. Mol Cell Neurosci 14: 506-518. Anderson, D. J. ( 1997). Cellular and molecular biology of neural crest cell lineage determination. Trends Genet. 13: 276-280. Carroll, P., Lewin, G. R., Koltzenburg, M., Toyka, K. V. and Thoenen, H. (1998). A role for BDNF in mechanosensation [see comments]. Nat Neurosci 1: 42-46. Clary, D. 0., Weskamp, G., Austin, L. R. and Reichardt, L. F. (1994). TrkA crosslinking mimics neuronal responses to nerve growth factor. Mol. Biol. Cell 5: 549-563. Ernfors, P., Merlio, J.-P. and Persson, H. (1992). Cells expressing mRNA for neurotrophins and their receptors during embryonic rat development. Eur. J. Neurosci. 4: 1140-1158. Farinas, I., Wilkinson, G. A., Backus, C., Reichardt, L. F. and Patapoutian, A. (1998). Characterization of neurotrophin and trk receptor functions in developing sensory ganglia: direct NT-3 activation of trkB neurons in vivo. Neuron 21: 325-334. Fedtsova, N. G. and Turner, E. E. (1995) . Bm-3.0 expression identified early postmitotic CNS neurons and sensory neural precursors. Mech. Dev. 53: 291-304. Greenwood, A. L., Turner, E. E. and Anderson, D. J. (1999). Identification of dividing, determined sensory neuron precursors in the mammalian neural crest. Development 126: 3545-3559. Groves, A. K. and Bronner-Fraser, M. (1999). Neural crest diversification. Curr Top Dev Biol 43: 221-258. 84 Hawrot, E. and Patterson, P.H. (1979). Long-term culture of dissociated sympathetic neurons. Methods in Enzymology 58: 574-583. Klein, R., Silos-Santiago, I., Smeyne, R. J., Lira, S. A., Brambilla, R., Bryant, S., Zhang, L., Snider, W. D. and M., B. (1994). Disruption of the neurotrophin-3 receptor gene trkC eliminates Ia muscle afferents and results in abnormal movements. Nature 368: 249-251. Klein, R., Smeyne, R. J., Wurst, W., Long, L. K., Auerbach, B. A., Joyner, A. L. and Barbacid, M. (1993). Targeted disruption of the trkB neurotrophin receptor gene results in nervous system lesions and neonatal death. Cell 75: 113-122. Le Douarin, N. M., Dupin, E. and Ziller, C. (1994). Genetic and epigenetic control in neural crest development. Curr. Op. Gene. Dev. 4: 685-695. Lee, K. J., Mendelsohn, M. and Jessell, T. M. (1998). Neuronal patterning by BMPs: a requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev 12: 3394-3407. Lewin, G. R. (1996). Neurotrophins and the specification of neuronal phenotype. Philos Trans R Soc Land B Biol Sci 351: 405-411. Liem, K. F., Jr., Tremmel, G. and Jessen, T. M. (1997). A role for the roof plate and its resident TGFB-related proteins in neuronal patterning in the dorsal spinal cord. Cell 91: 127-138. Lin, J. H., Saito, T., Anderson, D. J., Lance-Jones, C., Jessen, T. M. and Arber, S. (1998). Functionally-related motor neuron pool and muscle sensory afferent subtypes defined by coordinate ETS gene expression. Cell 95: 393-407. Matsumoto, S. G. (1994a). Neuronal differentiation in cultures of murine neural crest. II. Development of capsaicin-sensiti ve neurons. Dev. Brain Res. 83: 17-27. Matsumoto, S. G. ( 1994b ). Neuronal differentiation in cultures of murine neural crest. I. Neurotransmitter expression. Dev. Brain. Res. 83: 1-16. 85 McMahon, J. A., Takada, S., Zimmerman, L.B., Fan, C. M., Harland, R. M. and McMahon, A. P. (1998). Noggin-mediated antagonism of BMP signaling is required for growth and patterning of the neural tube and somite. Genes Dev 12: 1438-1452. McMahon, S. B., Armanini, M. P., Ling, L. H. and Phillips, H. S. (1994). Expression and coexpression of Trk receptors in subpopulations of adult primary sensory neurons projecting to identified peripheral targets. Neuron 12: 1161-1171. Mu, X., Silos-Santiago, I., Carrol, S. L. and Snider, W. D. (1993). Neurotrophin receptor genes are expressed in distinct patterns in developing dorsal root ganglia. J. Neurosci. 13: 4029-4041. Murphy, M., Reid, K., Hilton, D. J. and Bartlett, P. F. (1991). Generation of sensory neurons is stimulated by leukemia inhibitory factor. Proc. Natl. Acad. Sci. USA 88: 3498-3501. Neave, B., Holder, N. and Patient, R. (1997) . A graded response to BMP-4 spatially coordinates patterning of the mesoderm and ectoderm in the zebrafish. Mech Dev 62: 183-195. Nguyen, V. H., Trout, J., Connors, S. A., Andermann, P., Weinberg, E. and Mullins, M. C. (2000). Dorsal and intermediate neuronal cell types of the spinal cord are established by a BMP signaling pathway. Development 127: 1209-1220. Patel, T. D., Jackman, A., Rice, F. L., Kucera, J. and Snider, W. D. (2000). Development of sensory neurons in the absence of NGFffrkA signaling in vivo. Neuron 25: 345-357. Patterson, P.H. (1992). New blossoms in the neural crest field: trophic and transcription factors. Curr. Op. Cell Biol. 4: in press. 86 Schecterson, L. C. and Bothwell, M. (1992). Novel roles for neurotrophins are suggested by BDNF and NT-3 mRNA expression in developing neurons. Neuron 9: 449-463. Scott, S. A. (1992). Sensory neurons: Diversity, development and plasticity. Oxford University Press, Oxford. Sela-Donenfeld, D. and Kalcheim, C. (1999). Regulation of the onset of neural crest migration by coordinated activity of BMP4 and Noggin in the dorsal neural tube. Development 126: 4749-4762. Shah, N. M., Groves, A. and Anderson, D. J. (1996). Alternative neural crest cell fates are instructively promoted by TGFB superfamily members. Cell 85: 331-343. Shah, N. M., Marchionni, M.A., Isaacs, I., Stroobant, P. W. and Anderson, D. J. (1994). Olia! growth factor restricts mammalian neural crest stem cells to a glial fate. Cell 77: 349-360. Smeyne, R. J., Klein, R., Schnapp, A., Long, L. K., Bryant, S., Lewin, A., Lira, S. A. and Barbacid, M. (1994). Severe sensory and sympathetic neuropathies in mice carrying a disrupted Trk/NGF receptor gene [see comments]. Nature 368: 246249. Stemple, D. L. and Anderson, D. J. (1992). Isolation of a stem cell for neurons and glia from the mammalian neural crest. Cell 71 : 973-985 . Tominaga, M., Caterina, M. J., Malmberg, A. B., Rosen, T. A., Gilbert, H., Skinner, K., Raumann, B. E., Basbaum, A. I. and Julius, D. (1998). The cloned capsaicin receptor integrates multiple pain-producing stimuli [see comments]. Neuron 21: 53 1-543. Wright, D. E. and Snider, W. D. (1995). Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. 1. Comp. Neural. 351 : 329-338. 87 Wright, E. M., Vogel, K. S. and Davies, A. M. (1992). Neurotrophic factors promote the maturation of developing sensory neruons before they become dependent on these factors for survival. Neuron 9: 1-20. Xiang, M. Q., Zhou, L. J., Macke, J.P., Yoshioka, T., Hendry, S. H. C., Eddy, R. L., Shows, T. B. and Nathans, J. (1995). The Bm-3 family of POU-domain factors - primary structure, binding-specificity, and expression in subsets of retinal ganglion-cells and somatosensory neurons. J. Neurosci. 15: 4762-4785. 88 Chapter 4 Summary and future directions 89 For any patterning or specification event in developmental biology, the objective is to find the source of information. Because the neural crest is a population of migratory cells, the goal is to learn what information about prospective fate is contained within individual premigratory or migratory progenitors, and what information is imparted by signals present in the embryonic environment. It appears that there is not an absolute rule governing the timing and nature of cell fate specification of neural crest precursors. Some migratory neural crest cells remain multi potent for significant periods of time, such as the stem cells that can be isolated from sciatic nerve that retain the potential to make glia, myofibroblasts, and autonomic neurons (Morrison et al., 1999). Others are determined to produce only one cell type before or shortly after they migrate from the neural tube, as is likely the case for the sensory progenitors described in Chapter 2 (Greenwood et al., 1999). Therefore, neural crest cells are heterogeneous in terms of their potentials even early in their development. The central theme that can be drawn from the data presented in this thesis is that the diversity of peripheral neuron types is likely established early in neural crest development in a way that can be modulated by BMP signaling. This is not to say that individual progenitors are all determined for their peripheral neuron fates, but that interactions among neural tube and neural crest cells are sufficient to produce numerous peripheral neuron types in an orderly way that often resembles the in vivo patterning of these neurons. We have observed four ways in which BMP2 affects the development of peripheral neurons in neural tube explant cultures. 1.) BMP2 induces the autonomic neuron fate from migrating neural crest progenitors without preventing the differentiation of some sensory precursors (Chapter 2). 90 2.) We have noted, qualitatively, that BMP2 increases the number of sensory neurons present in explant cultures (Chapter 3). 3.) BMP2-treated cultures have an increased percentage of sensory neurons that exhibit characteristics of the cutaneous afferent phenotype (Chapter 3). 4.) BMP2 acts to pattern the proximodistal axis of the outgrowth such that sensory neurons form in the area adjacent to the neural tube while autonomic neurons form only in the periphery of the explant (Appendix 2). Previous studies have shown that BMP2 can act instructively to specify the autonomic fate from some neural crest cells at the expense of glial and myofibroblast lineages (Shah et al. , 1996). The induction of autonomic markers in the neural crest outgrowth is the only action of BMP2 we have observed that occurs independently of the neural tube. In contrast, the last three effects of BMP2 on the development of ex plant cultures are all dependent on the presence of the neural tube. These data indicate that the neural tube and the emigrated neural crest contain precursors that have inherently different responses to BMP signaling and/or that precursors rapidly change their responsiveness to BMPs. Such data emphasizes an emerging focus in developmental biology: how can the same signaling molecules (or families of molecules) be used reiteratively during development to create diversity? Given any specification event, it is overwhelmingly likely that TGF-~, FGF, Wnt, hedgehog, and/or Notch/Delta signals or modifiers will be involved. How is it possible to use the same signals again and again in the context of a very small space to induce disparate developmental events without creating mass confusion? Within the development of the neural crest, there are numerous questions that can be posed. For instance, BMP signaling induces the neural tube to produce sensory 91 precursors and it also instructs neural crest to produce autonomic neurons. Since some sensory precursors are not shunted to an autonomic fate by BMP2, does this mean that they are all determined to the sensory fate? How can we reconcile this with the ability of some migrating neural crest cells to produce both sensory and autonomic neurons? If BMP signaling induces the migration of neural crest cells, some of which have autonomic potential, and it also induces them to become autonomic neurons, what prevents precursors from differentiating into autonomic neurons as soon as they migrate from the neural tube (i.e., in the area of the DRG)? Clearly, BMP responses must be well regulated. There are two aspects of BMP signaling that may provide explanations for such questions. First, it is well known that BMPs can act as morphogens; depending on the dose, they can induce different responses from equivalent cells (for reviews see Smith, 1995; Dale and Jones, 1999). It is possible then that neural crest cells near the neural tube are exposed to a lower effective dose of BMPs than those in the periphery. Indeed, our preliminary data suggest that low doses of BMP2 induce the formation of sensory neurons from dissociated neural tube cells while higher doses induce the development of autonomic neurons (Appendix 2). This analysis should be extended, however, using better marker combinations to more carefully identify the different types of peripheral neurons that form under each condition. In contrast, when neural tube explants are cultured in the same high dose of BMP2, sensory neurons form near the neural tube and autonomic neurons form in the periphery. It seems important, though somewhat difficult, to determine the level of BMP signaling that is occurring in areas proximal and distal to the neural tube. Perhaps it may be possible to assay for the presence of activated SMAD proteins, which are known to transduce BMP signals upon phosphorylation (for review see (Massague and Chen, 2000), by staining with antibodies generated against phosphorylated SMADs or by monitoring the amount that translocate to the nucleus. A 92 lower level of activated SMAD proteins in cells near the neural tube would indicate that they were transducing less signal from BMP2 receptors. If such a difference in BMP activity were observed in the proximal and distal portions of the outgrowth, it may have arisen by an increased expression of BMP antagonists, such as noggin, in the neural tube and adjacent areas. In situ analysis of the expression patterns of BMP antagonists in neural tube explants grown in the presence and absence of noggin would be very interesting. Another possibility for the difference in response to BMP signaling by different parts of the explant is that, instead of these areas being exposed to different effective levels of signaling, they may express different co-factors for activated SMADs that change their response to the incoming signal. It has been recently shown that transcription factors of the homeodomain and winged-helix families can act as co-factors for activated SMAD proteins to recruit them to distinct promoter elements (Germain et al., 2000). This implies an interesting strategy for characterizing the molecular differences between similar cell types that respond differently to BMPs (provided these cells can be isolated to obtain RNA with which to construct libraries). By using activated SMAD protein as a bait in an interaction screen, it may be possible to identify co-factors that are specific to each tissue type. Moreover, since assembly of transcriptional complexes may be dependent on the particular promoter, one could study the potential SMAD co-factors using a promoter element that contained SMAD binding sequences from a gene already known to be differentially expressed between the two cell types. For example, the neurogenin and Mash genes may be regulated directly by SMADs. If so, an analysis of the co-factors required for SMAD binding to their promoters may identify additional transcription factors involved in the specification of these particular cell fates. The ideal experimental scenario would involve the ability to prospectively isolate and culture individual multi potent precursors from the rodent neural tube or neural crest 93 that can produce both sensory and autonomic neurons. At present, this requires that many conditions be tested empirically to define those that support the growth and direct the differentiation of these cells. Furthermore, the specific markers to identify and isolate such multipotent precursors have not yet been identified. It may therefore be valuable to explore whether any of the known genes which are differentially expressed in the neural tube can serve as markers for such multipotent precursors. This might be assayed by the use of transgenic reporter mice in which expression of lacZ, GFP, or some other sortable marker, is driven by regulatory elements of the candidate gene. The ability to isolate multi potent progenitors and direct their differentiation is the true test of how well we understand the processes that control development. In addition, such technology holds significant promise for exciting clinical applications. 94 REFERENCES Dale, L. and Jones, C. M. (1999) . BMP signalling in early Xenopus development. Bioessays 21: 751-760. Germain, S., Howell, M., Esslemont, G. M. and Hill, C. S. (2000). Homeodomain and winged-helix transcription factors recruit activated Smads to distinct promoter elements via a common Smad interaction motif. Genes Dev 14: 435-45 I. Greenwood, A. L., Turner, E. E. and Anderson, D. J. (1999). Identification of dividing, determined sensory neuron precursors in the mammalian neural crest. Development 126: 3545-3559. Massague, J. and Chen, Y. G. (2000). Controlling TGF-beta signaling. Genes Dev 14: 627-644. Morrison, S. J., White, P. M., Zock, C. and Anderson, D. J. (1999). Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96: 737-749. Shah, N. M., Groves, A. and Anderson, D. J. (1996). Alternative neural crest cell fates are instructively promoted by TGFB superfamily members. Cell 85: 331-343. Smith, J.C. (1995) . Mesoderm-inducing factors and mesodermal patterning. Curr Opin Cell Biol 7: 856-861. 95 Appendix 1 The extent of sensory neuron development in explants depends on the presence of the neural tube as a source of precursors 96 INTRODUCTION The neural tube is the ultimate source of sensory progenitors in the trunk region of the embryo. In addition, previous studies have indicated that the neural tube may provide signals important for the development of sensory ganglia in vivo (Kalcheim and Le Douarin, 1986; for review see Kalcheim, 1996). To further define the relationship between the neural tube and sensory neurogenesis, we took advantage of the ability to easily extirpate tissues from neural tube explants in vitro. We sought to address the following questions: first, is the neural tube required for the development of sensory neurons in ex plant cultures? Second, if the neural tube is required does it serve as a necessary source of signals, precursors, or both? Third, is the neural tube sufficient to support the generation or survival of sensory neurons in the absence of exogenously added neurotrophic factors? RESULTS AND DISCUSSION The neural tube is required for the extent of sensory neuron development in explant cultures To determine if the neural tube was required for the development of sensory neurons in our cultures, we assayed the number of sensory neurons in the outgrowth of explants in which the neural tube had been removed after various durations of time. Ex plants were grown in a complex medium (containing chicken embryo extract and growth factors) that we had previously determined was permissive for sensory neurogenesis (see Materials and Methods in Chapter 2). The neural tube was removed from explants using a tungsten needle after 1, 2, 3, or 4 days of culture. The number of sensory neurons, identified as Bm-3.0+/SCG 10+ cells by immunohistochemistry, was quantified per ex plant after 8 days of culture and compared to the number found in control explants in which the explant was left intact (see Materials and Methods in Chapter 2). Approximately a thousand 97 sensory neurons could be identified in the outgrowth of such undisturbed explants (Fig. I , Left in; Fig. 2, Left in). However, when the neural tube was removed after only one day of culture, there were strikingly fewer sensory neurons found in the outgrowth at day 8 (27± 13 Bm-3.0+/SCG 10+ cells; Fig. 1, Out day 1). The number of sensory neurons found in the outgrowth increased progressively with the length of time the neural tube remained in culture (Fig. 1, Out day 2, 3, and 4 ). This result indicates that the extent of sensory neurogenesis in explants depends heavily on the presence of the neural tube. However, the neural tube is not absolutely required for the presence of sensory neurons under these conditions, i.e., for survival. The positive correlation between the number of sensory neurons and the presence of the neural tube in explants can be explained by two possible mechanisms. First, the neural tube may provide an extrinsic factor (or factors) that is necessary for the differentiation or proliferation of sensory precursors, but is not required for survival of sensory neurons under these conditions. Second, and not mutually exclusive with the first, the neural tube may contribute sensory precursors to the outgrowth in a progressive manner. The latter mechanism predicts that the thousand or so sensory neurons in an 8 day explant derive from precursors that migrate from the neural tube over an extended period of time. In attempt to provide evidence for the model that the neural tube supplies critical factors for sensory neurogenesis, we looked for an ability of neural tube conditioned medium or of neural tubes grown on transfilters to induce sensory neurogenesis from neural crest. We did not observe a significant increase in the number of sensory neurons in either case. Furthermore, data presented in Chapter 2 (discussed below) indicate that the neural tube is not absolutely required for the division or differentiation of sensory precursors in our cultures. 98 1500 - n=8 CX) n=B > -- ct! "C ct! C ct! a. >< Cl) 1000 ,- n=B I., Cl) a. "' n=B Cl) 0 +o T"" C, u (J) 500 .... +- 0 MI C I., ID n=B 0 ~ 0 - Out day 1 I I Out day 2 I Out day 3 I Out day 4 I Left in Treatment of neural tube Fig. 1 The neural tube is required for the extent of sensory neuron development in ex plant cultures. Bars indicate the total number of sensory neurons (Bm-3.o+; SCG 1o+ cells) in the neural crest outgrowth per ex plant after 8 days of cul ture in complex medium supplemented with neurotrophic factors (NGF, BDNF, NT-3, and LIF). There are very few sensory neurons present in cultures in which the neural tube was removed at day 1 (Out day 1) compared to cultures left intact (Left in). The number of sensory neurons per explant increases progressively with the time the neural tube is left in culture. Error is represented as standard deviation. (n) is the number of explants scored from two experiments. 99 The neural tube is a long-term source of sensory precursors in explant culture To address the role of the neural tube as a continual source of sensory precursors in vitro, we determined the number of sensory neurons produced by late migrating cells. The neural crest outgrowth was removed from explants using a tungsten needle at day 1, 2, 3, or 4. More cells were observed to migrate from the neural tube after this procedure, and the number of sensory neurons found in this new outgrowth was assayed at day 8 (Bm3.0+/SCG 10+ cells). When the neural crest was removed at day 1, the number of sensory neurons found in the outgrowth at day 8 was only slightly less than the number found in explants that were left intact (Fig. 2, Out day I vs. Left in). This suggests that the majority of sensory precursors that contribute to the sensory neuron population in vitro migrate from the neural tube after the first day in culture. When the neural crest was removed on subsequent days, progressively fewer sensory neurons were observed in explants at day 8 (Fig. 2, Out day 2, 3, and 4). There were very few sensory neurons found in 8 day explants in which the neural crest had been removed at day 4 (15 ±12 Bm3.0+/SCGlO+ cells; Fig. 2, Out day 4), suggesting that the bulk of sensory precursors migrate during the first three days of culture. This experiment suggests that one important reason the extent of sensory neurogenesis in explants depends on the neural tube is because it serves as a source of sensory precursors for numerous days in culture. A possible caveat is that the procedure of neural crest removal may somehow stimulate the migration of new precursors that normally would not have emigrated from the neural tube. Given that late migrating cells with sensory neuron potential have been observed in vivo (Sharma et al., 1995), it is unlikely that late migrating sensory precursors observed here in vitro are merely artifacts of the procedure. One way to directly address this issue, however, would be to label neural tube cells in an intact explant and observe when they migrate. 100 C0 1500 ~ n=6 > (ti -- "C ( ti C ~ a. >< Q) n=S 1000 ,_ I,.. n=8 Q) a. - 1/) Q) +CJ 0 ~ c., u CJ) 500 ,_ n=7 +- 0 MI C I,.. a::i 0 '#: n=9 I 0 Left in I Out day 1 I Out day 2 - I Out day3 I Out day4 Treatment of neural crest outgrowth Fig. 2 Sensory precursors migrate from the neural tube for numerous days in explant culture. Bars represent the total number of sensory neurons (Bm-3.0+/ SCG 10tells) in the neural crest outgrowth per explant after 8 days of culture in complex medium supplemented with neurotrophic fac tors (NGF, BDNF, NT-3, and LIF). When the neural crest outgrowth is removed at day 1 (Out day 1), more cells migrate from the neural tube and produce nearly as many sensory neurons as are found in intact explants (Left in). T he number of sensory neurons per explant at day 8 decreases as the neural crest is removed on subsequent days of culture. E rror is represented as standard deviation. (n) is the number of ex plants scored fro m two experiments. 101 The neural tube is not required for sensory neurogenesis and is not sufficient for the long-term survival of sensory neurons The following data were presented in Chapter 2, but are discussed here in terms of the requirements of sensory neurons for the neural tube. In Chapter 2 we demonstrated that sensory precursors could differentiate in a fully defi ned medium in the absence of many exogenously added neurotrophic factors, such as nerve growth factor (NGF), brain derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), and leukemia inhibitory factor (LIF; see Chapter 2, especially Figs. 3 and 4). We found that the neural tube is not required for the process of sensory neurogenesis under these conditions since sensory precursors in the outgrowth can divide and differentiate even after the neural tube is removed from the culture (see Chapter 2, Figs. 4 and 5). Furthermore, the neural tube is not sufficient to support the long-term survival of sensory neurons under these conditions: although hundreds of sensory neurons could be found in explants after four days in culture, very few remained by day 8 unless the medium was supplemented with either neurotrophic factor-3 (NT-3) or brain derived neurotrophic factor (BDNF; see Chapter 2, Fig. 9). The experiments presented above emphasize that the neural tube is important for the process of sensory neuron development in vitro because it is a source of precursors for an extended period of time. These data are consistent with the report that late migrating sensory precursors can exit the neural tube through the dorsal root entry zone (DREZ) to reach the sensory ganglia (Sharma et al., 1995). In addition, we have observed that sensory precursors can divide and differentiate in the absence of the neural tube in our defined medium. Superficially, this second observation appears to contradict a previous study that suggests neural tube derived signals are important for the development of sensory neurons in vivo (Kalcheim and Le Douarin, 1986). When silastic membranes were inserted between the neural tube and migrated neural crest, 102 markers of neural crest disappeared from the location of the presumptive DRG across from such membranes unless the membranes were coated with neural tube extract, BDNF, or basic fibroblast growth factor (bFGF) (Kalcheim and Le Douarin, 1986; Kalcheim et al., 1987; Kalcheim, 1989). It is important to note that the neural tubeindependent sensory neurogenesis we observe in vitro occurs in a defined medium that does contain bFGF (though not BDNF or other neurotrophins). Explants grown in the absence of any bFGF did not survive well or produce much outgrowth. It would be interesting to know if withdrawal of bFGF from our cultures would have any effect on proliferation or survival or sensory precursors and whether neural tube derived factors could rescue such an effect. 103 REFERENCES Kalcheim, C. (1989). Basic fibroblast growth factor stimulates survival of nonneuronal cells developing from trunk neural crest. Dev Biol 134: 1-10. Kalcheim, C. (1996). The role of neurotrophins in development of neural-crest cells that become sensory ganglia. Philos Trans R Soc Lond B Biol Sci 351: 375-381. Kalcheim, C., Barde, Y. A., Thoenen, H. and Le Douarin, N. M. (1987). In vivo effect of brain-derived neurotrophic factor on the survival of developing dorsal root ganglion cells. Embo J 6: 2871-2873. Kalcheim, C. and Le Douarin, N. M. (1986). Requirement of a neural tube signal for the differentiation of neural crest cells into dorsal root ganglia. Dev Biol 116: 451-466. Sharma, K., Korade, Z. and Frank, E. (1995). Late-migrating neuroepithelial cells from the spinal cord differentiate into sensory ganglion cells and melanocytes. Neuron 14: 143-152. 104 Appendix 2 Patterning of explants in vitro resembles the patterning of sensory and autonomic derivatives in vivo and is affected by BMP2 acting on the neural tube 105 INTRODUCTION The pattern of neural derivatives in the vertebrate PNS is manifested largely in the placement of different types of neuronal ganglia throughout the body. In contrast to patterning of the CNS, which is governed by interactions of cells within a neuroepithelial sheet, the arrangement of different neuronal types in the PNS is thought to depend on interactions between migrating neural crest cells and the embryonic environment. Our in vitro system afforded the opportunity to address whether interactions among the neural tube and neural crest were sufficient to impose a pattern on peripheral neuron fates, and whether such patterning could be influenced by manipulation of the culture medium. Any patterning observed in explants must be a result of cellular information contained only within the tissues present in culture: the neural tube and the neural crest. Surprisingly, we observed a large degree of patterning of sensory and autonomic neurons along both proximodistal (P-D) and rostrocaudal (R-C) axes of neural tube explants that resembled patterning of sensory and autonomic derivatives in vivo. P-D patterning was induced by BMP2 or serum in a manner that requires the presence of the neural tube. In addition, we found that neural tube cells respond to different doses of BMP2 by expressing markers consistent with sensory or autonomic neurons. These data suggest that regulation of BMP-responsiveness among neural tube and crest cells is important to create the pattern of peripheral neuron derivatives observed in vivo. RESULTS AND DISCUSSION Patterning along the proximodistal axis of explants resembles the dorsoventral patterning of sensory and autonomic derivatives in vivo Explants grown in defined medium lack overt P-D patterning. Sensory neurons, as identified by immunoreactivity for Brn-3.0, are found throughout the P-D axis of the outgrowth at both day 2 (Fig. IA) and day 4 of culture (Fig. 2A). Under these conditions 106 Brn-3.0 /Phox2a/ API .... ... s •, .: ... .. -. .. -· -· .. ... .,. .. _,>,,;, . ,.. ,,P.~ , - .,,,_.,...,. .... .• A ~ ' #, - ~ ~ ~ ·t ~-~· -~ , \; • •t f ... ... B Fig. 1 BMP2 induces P-D patterning of sensory and autonomic neurons in the outgrowth of neural tube explants. The views shown extend from the dorsal neural tube (NT; top) across the ougrowth (below dashed line) of explants grown in defined medium (DM) for 2 days. Double-labeling for Brn-3.0 (dark blue, nuclear) and Phox2a (brown,nuclear) shows the position of sensory and autonomic neurons, respectively. Other cells in the outgrowth are visualized by DAPI staining (aqua). (A) In explants grown without BMP2, Bm-3.0+cells are scattered throughout the outgrowth and Phox2a is not expressed. (B) Explants grown in the presence of 50 ng/ml of BMP2 contain many Phoxza+ cells in only the distal part of the outgrowth; Brn-3.0+cells are present in the area adjacent to the neural tube where Phoxza+ cells do not form. 107 no autonomic neurons (Phox2a+ cells) are observed. Surprisingly, however, when the autonomic-inducing cue BMP2 is included in the medium, the outgrowth appears patterned: sensory neurons form only in the area adjacent to the neural tube and autonomjc neurons form in the distal part of the outgrowth (Fig. 1B, 2B). Such patterning is also apparent when explants are grown in complex medium in the presence of chicken embryo extract (CEE; Chapter 2, Fig. 1) or serum (Chapter 3, Fig. 3). Note that the patterning has two potentially independent features: first, sensory neurons are located near the neural tube and, second, autonomic neurons are not located near the neural tube. The induction of autonomic neurons in the distal part of the outgrowth occurs both when BMP2 is added at the time of plating and explants are fixed at day 2 (Fig. 1B) or when BMP2 is added at day 2 after significant crest migration and then explants are analyzed at day 4 (Fig. 2B). The later the addition of the autonomicinducing cue, however, the less likely are sensory neurons to be strictly localized near the neural tube (data not shown). The arrangement of neural derivatives along the P-D axis in BMP-treated cultures bears striking resemblance to the arrangement of sensory and autonomic ganglia along the dorsoventral (D-V) axis in vivo. In explants, sensory neurons form near the neural tube; autonomic neurons form farther away from it (Fig. 3A). Similarly, in vivo, sensory neurons are found in dorsal root ganglia (DRGs) located immediately adjacent to the neural tube while autonomic neurons are located far from the neural tube in ganglia in the ventral part of the embryo (Fig. 3B). Although it has been demonstrated that DRGs contain cells with autonomic potential (Le Lievre et al., 1980; Schweizer et al., 1983; Duff et al. , 1991 ), autonomic neurons do not differentiate in DR Gs. Thus, autonomic neurons are excluded from the dorsal area near the neural tube in vivo even though there are cells that reside in or pass through the area that have autonomic potential. The similarity between the patterning of neuronal derivatives in vivo and in vitro raises the 108 Brn-3.0 / Phox2a / DAPI B Fig. 2 BMP2 induces P-D patterning of sensory and autonomic neurons in explants even after initial neural crest migration. Explants were grown in defined medium (DM) for 4 days and then double-labeled for Brn-3.0 (dark blue) and Phox2a (brown) to identify sensory and autonomic neurons, respectively. The dorsal neural tube appears at the top (NT); neural crest outgrowth is below the dashed line. DAPI staining (aqua) reveals the extent of the outgrowth. (A) Explants that were grown in the absence of BMP2 contain many Brn-3.o+ cells at all levels along the P-D axis. (B) Explants that were grown in the absence of BMP2 for 2 days and then exposed to 50 ng/ml ofBMP2 contain Bm-3.0+cells in the area near the neural tube and Phox2a+ cells in the distal region of the outgrowth. 109 A B NT sensory 0 0 0 •• •• •• \e • .,' "'" \e.19 ., • ti• Oe .... autonomic 0 aorta autonomic I- .... gut Fig. 3 The patterning of explants in vitro resembles the patterning of sensory and autonomic ganglia in vivo. (A) In neural tube explants that have been treated with an autonomic-inducing cue, sensory neurons (blue circles) form in the area adjacent to the neural tube (NT) while autonomic neurons (red circles) form only the distal part of the outgrowth. (B) Schematic of the location of peripheral ganglia in vertebrate embryos. Sensory neurons (blue circles) are located in dorsal root ganglia (DRG) adjacent to the neural tube (NT) while autonomic ganglia (red circles) are located distal to the neural tube in the ventral part of the embryo. Purple arrows indicate the direction of neural crest migration. 110 possibility that similar forces are working in both systems. If so, interactions among cells of the neural tube and migrating neural crest may be more important in generating the pattern of peripheral ganglia than factors encountered locally by neural crest cells at the site of gangliogenesis. The neural tube is required for the proximodistal patterning of explants by BMP2 Given that BMP family members are known to act on the neural tube to specify dorsal CNS cell fates (Liem et al., 1997; Lee et al., 1998) and thought to promote neural crest outgrowth (Sela-Donenfeld and Kalcheim, 1999), we were interested to determine whether BMP2 was patterning explants through an action on the neural tube. Alternatively, BMP2 could directly pattern the neural crest or simply reveal an inherent heterogeneity already present among neural crest cells. To address this question, explants were grown for 24 hours in defined medium in either the presence or absence of BMP2. During this time the neural tube is necessarily a part of the explant, and BMP2, if present, has an opportunity to act on this tissue. At the end of day 1, the neural tube was removed from the explant and the isolated outgrowth was cultured for another 24 hours in either the presence or absence of BMP2. (Experimental design is represented in Fig. 4E). Cultures that were never exposed to BMP2 had many sensory (Brn-3.0+) cells scattered along the P-D axis of the outgrowth, but did not contain any autonomic (Phox2a+) cells (Fig. 4A). When neural crest cells that had migrated in the absence of BMP2 were subsequently exposed to the factor after removal of the neural tube, many Phox2a+ cells were observed, but these were freely interspersed with the Brn-3.0+ cells along the P-D axis (Fig. 4B). Therefore, although autonomic cells were generated, the cultures were not patterned. Conversely, explants grown in conditions in which they were exposed to BMP2 during the first 24 hours showed a striking degree of pattering (Fig. 4C, D). Bm3.0+ cells were only located in the part of the explant that had been near the neural tube. 111 Brn-3.0 / Phox2a / DAPI DM-BMP2 A BMP2-DM ...t: f J.o ' D G B .. BMP2-BMP2 ~: ~,..,.c ~- .. ..... --~ '•; • .,. I , , . .. -... ~.. \ .... '!! ' ,. '• , . . \ I .., . E • .,. .. ; t~.. 24hrs ) +/- BMP2 .• ( ..... Fig. 4 The neural tube is required for the P-0 patterning of explants by BMP2. (A-0) Views show day 2 outgrowth stained for Brn-3.0 (dark blue) and Phox2a (brown) to identify sensory and autonomic neurons, respectively (outgrowth that was near the neural tube is at the top). OAPI staining (aqua) shows other nuclei. (A) Neural crest grown in defined medium (OM) without BMP2 was not patterned and contained only Brn-3.o+ cells. (B) Neural crest that migrated in the absence of BMP2 and was then exposed to the factor after removal of the neural tube was not patterned and contained both Bm-3.o+and Phox2a+ cells. (C) Neural crest that migrated in the presence of BMP2 and was then switched into OM is patterned; Brn-3.o+cells form in the area that was near the neural tube while Phox2a+ cells are found in the distal outgrowth. (0) When the explants were continually exposed to BMP2, the outgrowth is patterned as in (C). (E) Schematic of the experimental design: neural tube explants were plated in the presence or absence of 50 ng/ml BMP2. The neural tube was removed from culture on day 1 when the explants were switched into new medium that did or did not contain BMP2 and grown until day 2. 112 This region also contained many cells that expressed neither neuronal marker. Phox2a+ cells were only located in the distal portion of the explant. From these data we conclude that interactions between BMP2 and the neural tube are required to pattern the outgrowth into proximodistal regions that express either sensory or autonomic markers. There are two general mechanisms by which the neural tube may be reacting to BMP2 to create an area that does not contain autonomic neurons. First, upon exposure to BMP2 the neural tube may inhibit the cells in the neighboring neural crest from becoming autonomic neurons. For example, in response to BMP2 the neural tube itself may produce an inhibitor, such as the BMP2/4 antagonist noggin, that prevents the induction of autonomic neurons in nearby regions. It could also, perhaps, begin a cascade of inhibitory cell-cell interactions by inducing expression of a BMP antagonist in the cells adjacent to the neural tube. In either case, it would be interesting to look at the expression of BMP antagonists in explant cultures grown in the presence or absence of BMPs. The second mechanism by which the neural tube could pattern the explant would be if BMP2 induced the migration of a different type of neural crest cell than normally migrates in the absence of BMP2. Given that the outgrowth appears larger in explants that have been treated with BMP2 (compare Fig. 4C, D to 4A, B), it seems likely that the neural tube is also producing more neural crest when it is exposed to BMP2. If so, the most interesting question is why those particular cells do not respond to BMP2 by becoming autonomic neurons. Although the model presented above is directed at why autonomic neurons do not form near the neural tube, the question of why sensory neurons form only near the neural tube under autonomic-inducing conditions also needs to be addressed. We have observed that the placement of sensory neurons along the P-D axis of the outgrowth depends on how long they have been exposed to an autonomic-inducing cue such as BMP2 or serum. When explants are exposed to serum for only one day and then grown in defined medium 113 until day 4, both trkA + and trkA- sensory neurons at the caudal end of explants are observed to be scattered throughout the outgrowth (Fig. 5A). When explants are exposed to serum for two days before they are switched into defined medium, many more sensory neurons are found in the caudal outgrowth; these too are distributed evenly along the P-D axis (Fig. 5B). However, when explants are exposed to serum for three days before they switched into defined medium, many sensory neurons are observed only in the area adjacent to the neural tube (Fig. 5C). Because the presence of serum promotes the differentiation of autonomic neurons, one possibility is that autonomic neurons on the distal edge of the outgrowth inhibit the migration or differentiation of sensory precursors into that area. This hypothesis predicts, however, that clusters of sensory neurons near the neural tube should always be flanked by autonomic neurons, and this situation is not always observed. Another possibility is that autonomic inducing cues may directly affect the migration ability of sensory precursors, perhaps by promoting differentiation. This hypothesis is consistent with the observed small decrease in BrdU incorporation by sensory precursors in the presence of BMP2 (Chapter 2, Fig. 6D), but remains to be fully explored. Autonomic and sensory neurons form preferentially in the rostral and caudal portions of neural tube explants, respectively In addition to patterning along the P-D axis of neural tube explants, there is a significant degree of patterning along the R-C axis as well. In a defined medium that does not support autonomic differentiation, a greater percentage of sensory neurons (Brn-3.0+ cells) are found in the caudal portion of the outgrowth than in the rostral portion in day 4 explants (Fig. 6). This discrepancy is increased in autonomic-inducing conditions, such as when explants are grown for two days in the presence of serum before they are switched into defined medium. Under these conditions, a significant amount of neural 114 TrkA/ Brn-3.0 :• .,!;~.~.,, :;Jti\,-,~; - . ,· . ·- ... ·• A D Fig. 5 The P-D position of sensory neurons in the outgrowth depends on the length of exposure to serum. (A-C) View from the neural tube (top) across the outgrowth of 4 day explants double-labeled with Brn-3.0 (brown), which stains all sensory neurons, and trk.A (blue, cytoplasmic), which stains a subset of sensory neurons. (A) Explants from which serum containing medium was withdrawn at day 1 contain Brn-3.0+ cells that are found throughout the outgrowth. (B) Explants from which serum was withdrawn at day 2 contain many Brn-3.0+cells which are also dispersed along the P-D axis of the explant. (C) The Brn-3.0+cells in cultures from which serum was withdrawn at day 3 are located in the area adjacent to the neural tube but not in the distal outgrowth. (D-F) High magnification views of Brn-3.o+; trk.A+ cells in (A-C), respectively. 115 B 20 .!! 'ii () 16 'ii 12 - 0 0 0 c-; 8 . I C m '#. 4 R C Fig. 6 The percentage of sensory neurons along the R-C axis is higher in the caudal part of explants grown in defined medium (DM). (A) Phase microscopy of a day 4 explant grown in DM indicating the rostral (R), intermediate (I), and caudal (C) areas of the outgrowth from which the percentage of sensory marker expression was determined. (B) The graph represents the percentage of total cells (as identified by DAPI staining) that express Brn-3.0 in the R, I, and C parts of the outgrowth from 4 explants. Error is shown as standard deviation. 116 crest migrates first from the rostral part of the explant (Fig. 7 A) and only later (after serum washout) from the caudal part (Fig. 7B). Most of the Phox2a+ cells are found rostrally (data not shown) and most of the Brn-3 .o+ cells (both trkA + and trkA-) are found in the caudal outgrowth that migrated after day 2 (Fig. 7C, D). The distribution of neuronal types can most easily be seen when the neural tube is divided evenly into rostral, intermediate, and caudal segments before culturing them as explants. When the segments are grown in serum-washout conditions, the rostral and intermediate explants produce overwhelmingly more Phox2a+ cells than the caudal segments (Fig. 8A-C). Conversely, the caudal portion contains more Brn-3.0+ cells than the rostral or intermediate portions, although some sensory neurons are found at all levels (Fig. 8D-E). Finally, trkA +cells, which have been shown to express the sensory marker Brn-3.0 in other experiments, are localized almost exclusively to the caudal explants (Fig. 8O-I). Simply put, the rostral, intermediate and caudal portions of the neural tube produce different kinds of neurons when cultured under the same conditions. These data suggest that different parts of the neural tube are predisposed to produce different amounts of autonomic and sensory neurons. The observation that different proportions of sensory and autonomic neurons come from different parts of the neural tube in vitro could reflect two aspects of peripheral neurogenesis that have been observed in vivo . First, neural crest migration and subsequent gangliogenesis proceed in a rostral to caudal gradient such that the more rostral structures develop before caudal structures. At any particular rostrocaudal level of the trunk, the earliest migrating neural crest are fated to travel to the ventral part of the embryo and become autonomic neurons, while later migrating neural crest give rise to sensory neurons (Weston, 1963; Serbedzija et al., 1990). It is an interesting correlation that in vitro the early-migrating cells give rise preferentially to autonomic neurons and the later-migrating cells to sensory neurons. This suggests, perhaps, that early and late 117 + ventral rostml caudal dorsal A day 2 \if:~-.., ~-,t_ •• ·. :::}~:.~ .~ r'·, .... .;a ~ B ._. :C rostral D caudal Fig. 7 Both trkA+ and trkA sensory neurons develop preferentially in the caudal part of explants grown in serum-washout conditions. (A) Phase microscopy of an explant grown for 2 days in medium containing serum shows that more neural crest has migrated from the rostral part (red line) than the caudal part (blue line) of the neural tube (NT). (B) The same explant at day 4 after subsequent growth in defined medium has significant outgrowth in both rostral (red stripe) and caudal (blue stripe) regions. (C-D) Rostral and caudal views of the outgrowth stained for Brn-3.0 (brown) and trkA (blue) shows that there are more sensory neurons of both subtypes in the caudal (D) versus the rostral (C) parts. 118 intermediate rostral caudal m C\I >< 0 .c c.. t • \ .... , , -" . . 0 -· • ,: . ·\,; . . ' C ~· . (Y) I : E .. al D E- •, . .F.. :- . . ,, ~a.. I • •. ., ,, , ... ~ .... • ,. • , • "'""i . G ~ H Fig. 8 Autonomic and sensory neurons form in different parts of neural tube explants grown in serum-washout conditions. Explants were separated into rostral, intermediate, and caudal segments, then grown for 2 days in medium containing serum followed by 2 days in defined medium. (A-C) Labeling for the autonomic marker Phox2a reveals that there are many positive cells in the rostral (A) and intermediate (B) explants, but few in the caudal (C) segment. (D-F) The sensory marker Brn-3.0 is expressed in all segments, but more in the outgrowth of caudal (F) segment than in intermediate (E) or rostral (D) parts. (G-1) The sensory subtype marker trkA is expressed almost exclusively in caudal explants (I vs. G and H). 119 migrating crest have intrinsic biases to produce autonomic and sensory neurons, respectively. However, such a model predicts that the earliest migrating cells at all rostrocaudal levels will be biased to the autonomic fate. Instead, we observe that the first cells that migrate from the caudal part of the ex plant produce sensory instead of autonomic neurons. If the R-C patterning of explants is a reflection of the R-C gradient of development, then perhaps the reason it does not faithfully reproduce the in vivo pattern is because important environmental cues that control the timing of development are lacking. Another possible explanation for the differences in the sensory and autonomic potential of rostral and caudal neural crest is that these differences may actually reflect the normal proportion in which these neurons produced in vivo. Although all segments in the trunk contain sensory ganglia, DRGs are much larger at limb levels. This size difference is apparent before the onset of cell death, which indicates that more sensory neurons are born (rather than simply survive) at limb levels (Goldstein, 1993). In our experiments, we isolate the neural tubes from the section of trunk which is caudal to the forelimb and thus represents only the mid-thoracic and lumbar segments. If different areas of the neural tube give rise to different proportions of sensory and autonomic neurons, it is possible that some of this information is contained within the neural tube itself and is not imposed only by somatic structures. It is intriguing that the cultured neural tube explants clearly produce more autonomic neurons from the rostral part and more sensory neurons from the caudal part. The relationship between this phenomenon and the in vivo processes is not yet understood. 120 Different doses of BMP2 induce the expression of Brn-3.0 and Phox2a in cultures of dissociated neural tube cells The generation of both sensory and autonomic neurons from cells that migrate from the neural tube suggests that some precursors may be biased toward a particular neuronal fate at the time that they migrate. Furthermore, as shown above the patterning of these neurons appears to depend on interactions between the neural tube and BMP2. We were therefore interested to determine if BMP2 could have an affect on the types of peripheral neurons produced from neural tube cells. Members of the transforming growth factor-~ (TGF-~) superfamily, which includes the BMPs, are known to pattern cell fates within the dorsal neural tube in a dose-dependent manner (Liem et al., 1997; Lee et al., 1998). Given this dose responsiveness , we looked at the ability of different doses of BMP2 to induce Brn-3 .0 and Phox2a in dissociated cultures of neural tube cells grown in defined medium for 4 days. In the absence of BMP2, there was little expression of the markers by neural tube cells (Fig. 9A). In the presence of low doses of BMP2 (5-10 ng/ml) , the cultures contained many Brn-3.0+ cells and very few Phox2a+ cells (Fig. 9B, C). When grown in 20 -30 ng/ml of BMP2, the cultures contained both Brn-3.o+ and Phox2a+ cells (Fig. 9D, E). In 50 ng/ml of BMP2, the cultures contained many Phox2a+ cells and very few Brn-3 .0+ cells (Fig. 9F). Therefore, different doses of BMP2 can differentially affect the expression of these two markers in cultured neural tube cells: low doses promote Brn3.0 expression, high doses promote Phox2a expression. Although it is possible that different doses of BMP2 could specify different peripheral neuron fates from multipotent precursors within the neural tube (the most exciting interpretation), there are many caveats and alternate explanations that must be considered. First, the markers used in this experiment, Brn-3 .0 and Phox2a, are expressed not only in sensory and autonomic neurons, respectively, but also in neurons of the dorsal neural tube . It is therefore possible that different doses of BMP2 are 121 Brn-3.0 / Phox2a Ong/ml . . 5 ng/ml 10 ng/ml ·~. ..,, .... ' .I.• • t • A 20 ng/ml 50 ng/ml 30 ng/ml ,... ..... D . F • Fig. 9 Different doses of BMP2 induce the expression of Brn-3.0 and Phox2a in cultures of dissociated neural tube cells. Double-labeling shows Brn-3.0 (dark blue) and Phox2a (brown) expression in dissociated neural tube cultures grown in defined medium for 4 days. (A) Little expression of either marker is seen in cultures that were not exposed to BMP2. (B, C) Many Brn-3.o+ cells are found in cultures grown in low doses of BMP2 (5 and 10 ng/ml), but few Phox2a+ cells are observed. (D, E) At intermediate doses of BMP2 (20 and 30 ng/ml), cultures contain both Brn-3.o+ and Phox2a+cells. (F) At high doses of BMP2 (50 ng/ml), there are many Phox2a+cells and few Brn-3.0+cells. 122 modulating only the fate of dorsal neural tube cells, and not the production of peripheral neurons from premigratory neural crest precursors. To partially address this concern, dissociated neural tube cultures were grown in the presence of low doses of BMP2 (10 ng/ml) to promote differentiation of Brn-3.0+ cells, then cultured in the presence of the neurotrophins NT-3 and BDNF (25 ng/ml) until day 8. When these cultures were stained for Brn-3.0 and peripherin, there were many neurons that expressed both markers indicating that at least some peripheral sensory neurons are produced in these cultures. A more detailed analysis in which neural crest derivatives can be distinguished from neural tube cell types will be important for the interpretation of these results. In addition to the importance of identifying the different types of cells that arise in such cultures, it is also necessary to determine the mechanism by which BMP2 induces different markers. We are interested to know if it is acting instructively, selectively, or permissively on different sorts of neural tube progenitors. The standard method for analyzing such a question is via a clonal analysis ; for technical reasons, however, this has not been possible with these cells. In the absence of such data, it is not possible to say how different doses of BMP2 are affecting neural tube cells. The preliminary observations, however, raise the interesting idea that heterogeneity among migrating neural crest cells, in terms of their abilities and biases to give rise to different kinds of peripheral neurons, may originate in the neural tube as a consequence of exposure to different factors and cell interactions. Such factors may include a differential response to BMP family members. 123 MATERIALS AND METHODS All immunocytochemical and tissue culture procedures described here are identical to those presented in Chapters 2 and 3. Cultures of dissociated neural tube cells Neural tubes were isolated from the trunk region of El0.5 Sprague-Dawley rats as described previously (Greenwood et al., 1999). They were collected in defined medium in an Eppendorf tube and gently microcentrifuged at 5000 rpm for 1 minute. The medium was removed with a pipette and replaced with 1 ml of 0.5% Trypsin-EDTA (GibcoBRL), gently pipetted to resuspend the neural tubes . The Eppendorf tube was placed on its side in a tissue culture incubator and incubated at 37°C for approximately 10 minutes. (The timing of digestion is critical and should be determined for the amount of tissue and batch of enzyme being used). After incubation, medium containing 10% FBS was added to the tube, and then the tissue was again microcentrifuged to pellet the cells. The neural tube cells were washed again with medium containing serum and then twice with defined medium. Finally, the tissue was resuspended in 0.5 ml of defined medium and triturated with a fire-polished, pre-wet Pasteur pipette. Cells were plated on 35 mm plates coated with a fibronectin substrate (BTI). 1.5 ml of defined medium was added to plates, then greaseless, sterile cloning cylinders (4 mm internal diameter; Bellco) were placed on the dishes. 15,000 cells were delivered to each cloning cylinder (2-3 per dish) . Cells were allowed to attach for 2-6 hours before removal of the cylinder. BMP2 was added after plating. 124 REFERENCES Duff, R. S., Langtimm, C. J., Richardson, M. K. and Sieber-Blum, M. (1991). In Vitro Clonal Analysis of Progenitor Cell Patterns in Dorsal Root and Sympathetic Ganglia of the Quail Embryo. Devel. Biol. 147: 451-459. Goldstein, R. S. (1993). Axial level-dependent differences in size of avian dorsal root ganglia are present from gangliogenesis. J Neurobiol 24: 1121-1129. Greenwood, A. L., Turner, E. E. and Anderson, D. J. (1999). Identification of dividing, determined sensory neuron precursors in the mammalian neural crest. Development 126: 3545-3559. Le Lievre, C. S., Schweizer, G. G., Ziller, C. M. and Le Douarin, N. M. (1980). Restrictions in developmental capabilities in neural crest cell derivatives as tested by in vivo transplantation experiments. Dev.Biol. 77: 362-378. Lee, K. J., Mendelsohn, M. and Jessen, T. M. (1998). Neuronal patterning by BMPs: a requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev 12: 3394-3407. Liem, K. F., Jr., Tremmel, G. and Jessen, T. M. (1997). A role for the roof plate and its resident TGFB-related proteins in neuronal patterning in the dorsal spinal cord. Cell 91: 127-138. Schweizer, Ayer-LeLievre and Douarin, L. (1983). Restrictions of developmental capacities in the dorsal root ganglia during the course of development. Cell Differentiation 13: 191-200. Sela-Donenfeld, D. and Kalcheim, C. (1999). Regulation of the onset of neural crest migration by coordinated activity of BMP4 and Noggin in the dorsal neural tube. Development 126: 4749-4762. 125 Serbedzija, G. N., Fraser, S. E. and Bronner-Fraser, M. (1990). Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labeling. Development 108: 605-612. Weston, J. A. (1963). A radiographic analysis of the migration and localization of trunk neural crest cells in the chick. Devel. Biol. 6: 279-310. 126 Appendix 3 Identification by differential RT-PCR of a novel paired homeodomain protein specifically expressed in sensory neurons and a subset of their CNS targets 127 MCN Molecular and Cellular Neuroscience 6, 280-292 (1995) Identification by Differential RT -PCR of a Novel Paired Homeodomain Protein Specifically Expressed in Sensory Neurons and a Subset of Their CNS Targets Tetsuichiro Saito,*·1 Amy Greenwood , Qi Sun, and David J. Anderson*· 2 Division of BiologtJ 216-76 and *Howard Hughes Medical Institute, California Institute of Technology, Pasadena, California 91125 Sensory neurons are a major derivative of the neural crest for which there have been no definitive molecular markers in mammals. We have developed a method that combines differential hybridization with degenerate RT -PCR to rapidly screen gene families for members exhibiting differential expression among tissues or cell types. We used this approach to search for transcription factor-encoding genes specifically expressed in mammalian sensory neurons. A novel paired homeodomain protein, called DRG11, was identified. DRG11 is expressed in most sensory neurons, including trkA-expressing neurons, but not in glia or sympathetic neurons. Unexpectedly, it is also expressed in the dorsal horn of the spinal cord, a region to which NGF-dependent sensory neurons project. These data suggest that DRG11 is not only a useful marker for sensory neurons, but may also function in the establishment or maintenance of connectivity between some of these neurons and their central nervous system targets. INTRODUCTION The study of cellular diversification during neurogenesis requires markers to identify different neural cell types. Recently it has become clear that transcription factors can serve as useful markers of neuronal identity. For example, the bHLH protein MASHl identifies autonomic progenitors in the peripheral nervous system (Lo et al., 1994). Similarly, Islet-1 (Ericson et al., 1992) and addi1 Present address: Tsukuba Life Science Center, RJKEN, 3-1-1 Koyadai, Tsukuba, lbaraki 305, Japan. 2 To whom correspondence should be addressed . Fax: (818) 564-8243. E-mail: andersond@starbasel.caltech.edu. 280 tional recently characterized proteins in the lim homeodomain family mark subsets of functionally distinct motor neurons (Tsuchida et al., 1994). These data suggest that the diversity of neuronal. cell types can be defined in terms of an underlying diversity of expression of members of a transcription factor gene family. Polymerase chain reaction (PCR)-based cloning using degenerate oligonucleotide primers has proven useful as a method to rapidly isolate members of a gene family (see, for example, Libert et al., 1989; Wilkie and Simon, 1991). When cDNA rather than genomic DNA is used as the template, the method can selectively identify those members of a gene family expressed in a given tissue or cell type Uohnson et al., 1990; Lai and Lemke, 1991). A limitation of this approach is that one must often sequence tens if not hundreds of PCR products to identify novel genes and then examine their expression patterns to identify those appropriate for further study. This makes it labor-intensive to apply this approach to several gene families simultaneously. We have developed and applied a method, called differential reverse-transcriptase-based PCR (dRT-PCR), which combines RT-PCR with differential hybridization. This method identifies members of a gene family which are specifically expressed in a given tissue or cell type. We have applied this method to search for markers of mammalian sensory neurons, a major derivative of the neural crest for which there have been no definitive molecular markers described. Neuropeptides such as CGRP (Murphy et al., 1991) and Substance P (Ito et al., 1993) have been used as sensory markers, but these neuropeptides are not in fact sensory neuron-specific: for example, they 1044•7431/95 $6.00 Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved. 128 A Sensory Neuron-Specific Homeodomain Protein A 281 Cell lines/tissues B, C, D... etc. Cell line/tissue A t t RNA RNA I RT•PCA with degenerate primers I T for gene tamlly of Interest T ♦ Gel-Isolate expected bends RT-PCR products, ♦ RT-PCR products j ::::·:.i.,•'"'"" j labeled probe • labeled probe B 1 2 3 4 DAG-probe / Dlfferentlal hybridization NCM 1-probe i MAH-probe A-specific clones Rat1-probe FIG. 1. (A) Schematic representation of the differential RT-PCR procedure and (B) an example of differential hybridization. See text for details. can be induced in cultured sympathetic neurons by some cytokines (Fann and Patterson, 1994). In the absence of definitive sensory neuron markers, unambiguous identification of sensory neurons in mammalian neural crest cultures has been difficult (Matsumoto, 1994). Using dRT-PCR, we screened four families of transcription factors for members specifically expressed in different neural crest derivatives and have thereby cloned a cDNA encoding a novel paired homeodomain protein called DRGll. DRGll is specifically expressed in sensory neurons but not in autonomic neurons or glia. Unexpectedly, DRGll is also expressed in a subset of the known functional targets of sensory neurons in the spinal cord. This transcription factor therefore provides a molecular marker to identify neurons in the peripheral sensory lineage. Moreover, its pattern of expression in the central nervous system suggests that it may function in regulating some aspect of the connectivity between these neurons and their central targets. RESULTS Design and Application of a Differential RT-PCR Cloning Strategy A schematic outline of the dRT-PCR procedure is illustrated in Fig. lA. cDNA from two or more cell types or tissues is amplified using degenerate oligonucleotide primers flanking a conserved domain from a given gene family. The PCR products from the "positive" tissue of interest (A cell in Fig. lA) are subcloned, and the colonies are grown in a 12 X 8 array in a microtiter plate. Multiple replicas of this array are prepared by spotting small aliquots of each liquid bacterial culture onto nylon filters (Wilkie and Simon, 1991). These filters are then annealed with 32P probes made from the same RT-PCR products, as well as from the RT-PCR products derived from one or more "negative" tissues (B, C, D ... etc. cells in Fig. lA). Clones displaying differential hybridization (A +B-c-o-) are picked for sequencing. For simplicity, examples of only four clones (Fig. lB, columns 1-4) derived from paired homeodomain (PHO) primer amplification of Embryonic Day 13.5 (E13.5) rat DRG cDNA, annealed in quadruplicate with four different probes, are illustrated. The probes used were PHO-amplified PCR products from: DRG, NCM-1 cells [a glial progenitor cell line (Lo et al., 1990)], MAH cells [a sympathoadrenal progenitor cell line (Birren and Anderson, 1990)] and Rat-1 fibroblasts (a nonneural cell line). Two pieces of information are derived from this screen. First, the relative intensities of the hybridization signals exhibited by a single clone annealed with multiple probes gives an indication of its relative abundance within its gene family in different cells or tissues (Fig. lB, read vertically). Second, the relative intensities of the hybridization signals exhibited by multiple clones annealed with a single probe provide an indication of their relative abundance within that cell or tissue type (Fig. lB, read horizontally). Taken together, such information provides a unique "fingerprint" for each clone. Such fingerprint data allow a preliminary assessment of the number of distinct differentially expressed sequences within an array of clones. In the example shown in Fig. lB, clone 1 is strongly expressed in DRG, weakly in NCM-1 cells, and not above background in MAH and Rat-1 cells (background is de- 129 282 fined by the weakest hybridization signal observed on a clone array with a given probe); this clone was later identified as Pax3 (Goulding et al., 1991). Clones 2 and 3 are expressed strongly in DRG, and not above background in NCM-1, MAH, and Rat-1 cells; this clone was subsequently identified as a novel PHD protein which we have called DRGll. Clone 4 is expressed in DRG and NCM-1 cells and weakly in Rat-1 cells. This clone encodes a novel PHD protein we have called NCM3 (data not shown) . Similar analyses were carried out for three other transcription factor gene families [Ets-domain (Nye et al., 1992), forkhead (Lai et al., 1991), and scalloped/ TEF (Campbell et al. , 1992) families] (data not shown). In all cases, differentially expressed sequences were obtained. Only the results of the PHD family screen will be presented here. To further confirm the specificity of expression of the clones identified in the initial differential screen, the hybridization procedure was reversed: inserts from three of the clones (DRGll, Pax3, and NCM3) were used as hybridization probes on gel blots of total PHD RT - PCR products from various cell lines or tissues at different stages (Fig. 2). Consistent with the clone blot data (Fig. lB), the DRGll probe annealed to a single band in PHD RT-PCR products from DRG (Fig. 2C, lanes 7 and 8), but not in MAH, Ratl, or NCM-1 cells (Fig. 2C, lanes 1-6). By this analysis, DRGll mRNA appears to be expressed in sensory ganglia but not in cell lines representing sympathetic neuron or glial precursors or in fibroblasts. As expected from the clone blot analysis, hybridization patterns distinct from that of DRGll were obtained with the Pax3 and NCM3 probes (Figs. 2D and E). To isolate full-length DRGll cDNA clones a cDNA library was constructed using PCR-amplified cDNA from El3.5 DRG (see Experimental Methods) and screened using the DRGll PCR product as a probe. The deduced amino acid sequence of the longest cDNA (2.4 kb) obtained is presented in Fig. 3A. The clone encodes a novel 28.6-kDa protein in the PHD family. As the 280bp sequence upstream of the most N-terminal methionine contains an in-frame translational termination codon, we have tentatively assigned this methionine as the initiation codon. Attempts to determine the size of native DRGll mRNA by Northern blotting of either total or poly(A +) RNA from rat embryos were unsuccessful, most likely due to the low abundance of the transcript. Moreover, the hamster anti-rat DRGll monoclonal antibodies we generated (see below) did not work well in Western blotting experiments; therefore we are unable to establish the size of the native DRGll protein. Thus, we cannot rule out the possibility that this protein is larger than that predicted by the deduced amino acid sequence from our cDNA clone. Saito et al. A B 1 2 3 4 5 6 7 8 9 101112 C D E FIG. 2. (A) RT-PCR of paired homeodomain genes and (C-E) differential expression of the paired homeodomain genes. (A) Agarose electrophoresis of PCR products obtained from Rat 1 fibroblasts (lanes 1, 2), MAH (3, 4) and NCMl cells (5, 6), E13.5 rat DRG (7, 8), Ell.5 rat embryo (9, 10), and E13.5 embryo (11, 12) cDNA using the same degenerate primers as were used to clone the paired homeodomain genes. One microliter (odd-numbered lanes) or 0.1 µl (even-numbered lanes) of cDNA mixture equivalent to 40 or 4 ng, respectively, of total RNA (see Experimental Methods) were used as template for PCR. (B) Control RT-PCR from the same cDNA sources as in (A) performed using actin primers. (C-E) the PHO dRT-PCR products (A) were transferred to GeneScreen filters and hybridized using probes containing the homeodomains of (C) DRGll, (D) Pax3, and (E) NCM-3. 130 A Sensory Neuron-Specific Homeodomain Protein 283 A MFYFHCPPQLEGTAPFGNHSTGDFDDGFL~ RKQRRNRTTFALQQLEALEAVFAQTHYPDV 60 FTREELAMKINLTEARVQVWFQNRRAKWRK 90 TERGASDQEPGAKEPMAEVTPPPVRNINSP 120 PPGDQARGKKEALEAQQSLGRTVGPAGPFF 150 PSCLPGTLLNTATYAQALSHVASLKGGPLC 180 SCCVPDPMGLSFLPTYGCQSNRTASVAALR 210 MKAREHSEAVLQSANLLPSTSSSPGPASKQ 240 VPPEGSQDKPSPTKEQSEGEKSV* B PRETTYPLOT of: @Phd.Fil Plurality: 6.00 30 Threshold: 1.00 AveWeight: December 7, 1.00 264 1994 AveMatch: 16:49 0.54 AvMisMatch: -0.40 Drgll Phox2 Paxl Cartl Phoxl 50 50 50 50 SB 50 50 50 so Pax6 Dal Dprd Drepo Cunc4 Drgll Phox2 Pax3 Cartl Phoxl SB Pax6 Dal Dprd Drepo Cunc4 so 50 50 V Q V W F Q N R R A K W R K T[El R G A S V Q V W F _Q_ N R R A K F R K Ql..fil R A A S V Q V W Fl.2J N R R Ai]) W R K Q A G A N Q V Q V W F Q N R R A K W R K R~ R Y G Q VQVWFQNRRAKFRRNERAML VQVWF NRRAKFRRNERAML I Q V W FCT) N R R A K W R R E E K L R N IQ V W F _Q_N RR AK WR K Q EK VG P I Q V WF N R R Ai]) L R K Q H T S V S V Q V W F Q N R R A K W R K H[El P P R K V V WF N R R A K W R K Rl..fil Q N R N L2J 70 10 10 10 10 10 10 10 10 10 70 FIG. 3. (A) Deduced amino acid sequence of DRGl l protein; the homeodomain is underlined. (B) Comparison of homeodomain sequences among paired homeodomain proteins. The sources of the sequences illustrated are: Phox2 (Valarche et nl. , 1993); Cart] (Zhao et n/., 1993); Phoxl (Grueneberg et nl. , 1992); S8 (Opstelten et al., 1991); Pax6 (Walther and Gruss, 1991); Dal (Schneitz et nl., 1993); Dprd (Frigerio et n/., 1986); Drepo (Xiong et nl., 1994); Cunc4 (Miller et nl., 1992). DRGll is closely related in sequence to Pax3 as well as to several other paired homeodomain proteins (Fig. 3B). Unlike the Pax genes, however, DRG11 lacks a paired domain (Fig. 3A). Interestingly, DRGll has a Gin residue instead of a Ser at position 9 of the recognition helix (residue 56 in Fig. 3B); a substitution of Gin for Ser at this position also occurs in several other family members that lack a paired domain (e.g., Phox2 and Cartl in Fig. 3B). DRG11 Is Specifically Expressed in Sensory Neurons and in a Subset of Their CNS Targets We next performed in situ hybridization experiments using digoxygenin-labeled full-length DRG11 cRNA probes, on sections from rat embryos of different ages. The earliest stage at which DRGll mRNA could be detected was at E12.5, 1-2 days following the initial condensation of neural crest cells to form dorsal root ganglia. Expression at this age was restricted to the nervous system and within the trunk region was restricted to the dorsal root ganglia; no expression was detected in the neural tube (Fig. 4A). By contrast, other transcription factors expressed in sensory neurons at this stage, such as Isl-1 and Pax3 (Fig. 4C, D) are also expressed in the neural tube. By E15.5, DRG11 expression was strong in trunk DRG (Fig. SA, arrowheads) and was also detected in the dorsal spinal cord. Positive cells were located both laterally near 131 284 the dorsal root entry zone and more medially near the ventricular zone (Fig. SA, arrows). Between these two zones, there is a region containing only scattered DRGllpositive cells (Fig. SA, open arrow), although this region does contain neurons as shown by SCGlO hybridization (Fig. SB, solid arrow). This pattern suggests that DRGll is expressed by cells which display the known migration pattern of newly born dorsal horn neurons (Langman and Haden, 1970): these cells are initially generated in the ventricular zone, then migrate laterally through an intermediate region to take up their final position at the dorsolateral margins of the spinal cord, where they create a second region of high DRGll expression. The DRGll expression pattern is distinct from that of a related PHD protein, Pax3 (Goulding et al., 1991), which is also expressed in the dorsal spinal cord at this stage but only in the ventricular zone (Fig. SC, arrowhead). Although DRGl 1 was detected in sensory but not sympathetic ganglia (not shown) at ElS.5, this apparent specificity could simply reflect the timing of differentiation in the two groups of neurons; sympathetic development is known to lag behind sensory development by several days. For example, trkA mRNA is initially expressed in sensory but not sympathetic neurons at E12.5 (MartinZanca et al., 1990), but is later detected in sympathetic ganglia beginning on E16.5-E17.5 (Birren et al., 1993). To determine whether this is also true for DRGll, we examined its expression in sections through the anterior trunk region of El7.S embryos. These sections contain large sympathetic ganglia expressing SCGlO (Fig. 6B), trkA (Fig. 6D), and Isl-1 (Fig. 6C), but not DRGll (Fig. 6A, arrows). This specificity of DRGll expression in the PNS is also maintained at Postnatal Day 3 (not shown). At E17.5, DRGll expression in the spinal cord appeared increased in the dorsal horns relative to ElS.5 (Fig. 6A, open arrow). Interestingly, this region of the spinal cord receives synaptic input from the DRG (Kandel et al., 1991). Examination at high magnification of sections hybridized with the DRGll cRNA probe revealed that the DRGll + cells have a process-bearing morphology, suggesting that they are neurons (not shown). These data suggest that DRGll is expressed both by sensory neurons in the DRG and by a subset of their synaptic target neurons in the dorsal spinal cord. At no time, however, did we detect DRGll expression in the ventral spinal cord which also receives sensory innervation. No DRGll expression was detected outside of the nervous system at any of the stages examined. The expression of DRGll in sensory ganglia as well as in a subset of their CNS targets raised the question of whether DRGll is expressed only by those sensory neurons that project to the dorsal spinal cord. These neurons include the nociceptive, NGF-dependent subset Saito et al. (Ruit et al., 1992) which expresses trkA. These neurons can be distinguished from other sensory neurons which express trkC, are NT-3-dependent, and project to the ventral spinal cord, many of which are proprioceptive (for review, see Snider, 1994). The expression of DRGll in E13.5 trigeminal sensory ganglia (Fig. 7A, arrow) as well as in trunk DRG (Fig. 6A, arrowheads) appeared broader than that of trks A, B, or C in nearby sections, however (Figs. 6D, 7D-7F, and data not shown). Rather, the extent of DRGl 1 expression was similar to that of SCGlO (Stein et al., 1988) or Isl-1, two markers which label all neurons in these sensory ganglia (Figs. 7B and 7C). In addition, DRGll expression was detected in both small and large DRG neurons at E17.5; in contrast the larger neurons expressed trks Band C but not A (Figs. 7D-7F and data not shown). This suggests that, at least at this stage, DRGll expression includes, but is not restricted to, the NGF-responsive subset of sensory neurons. That such neurons do express DRGl 1, however, is supported by double-label immunocytochemical labeling experiments (see below). To determine whether DRGll expression within ganglia is restricted to neurons or is common to neurons and nonneuronal (glial) cells, we examined its expression in dissociated postnatal DRG cultures using a monoclonal antibody raised against bacterially expressed DRGll protein (see Experimental Methods). Staining of perinatal rat dissociated DRG cultures revealed nuclear immunoreactivity in many sensory neurons, but not in glia (Figs. SA and SB). Approximately 60% of the neurons were labeled by the antibody under these conditions (Fig. SA, arrowheads). Whether the apparent lack of DRGl 1 expression in other sensory neurons reflects distinct sensory lineages or rather environmental modulation in culture is currently being investigated. As an additional control for the specificity of the antibody, similar dissociated cultures of superior cervical sympathetic ganglia (SCG) were stained. No labeling of any cells was detected (Figs. SC and 8D). Thus, as predicted by its initial selection in the dRT-PCR screen based on expression in DRG but not in MAH or NCM-1 cell cDNA, DRGll is expressed in many sensory neurons but not in sympathetic neurons or glial cells. The availability of a monoclonal antibody to DRGll allowed a preliminary assessment of whether this transcription factor is indeed expressed by the trkA-expressing subset of DRG sensory neurons, as suggested by the in situ hybridization data (see above). To address this question, dissociated cultures of E16.5 DRG were doublelabeled with anti-DRGl 1 and a specific polyclonal antiserum to trkA (Clary et al., 1994). Many trkA + neurons (Fig. 9A, arrowheads) coexpressed DRGll in their nuclei (Fig. 9A, open arrows) . Conversely, it appeared that the 132 A Sensory Neuron-Specific Homeodomain Protein majority of DRGll + cells also expressed trkA, although a few DRGll + trkA - cells could be observed. Finally, some neurons in the cultures expressed neither marker (Figs. 9A and 9B, solid arrow). These data confirm that DRGll is expressed by NGF-responsive DRG sensory neurons, many of which are nociceptive neurons that project to the dorsal horn of the spinal cord (Snider, 1994) where DRGll is also expressed. However, DRGll expression is also detected in some sensory neurons that do not express trkA, consistent with the in situ hybridization data. DISCUSSION Transcription factors have proven to be useful markers of cell type in the nervous system. Here we have developed and applied a method that combines degenerate RT-PCR with differential hybridization and used it to screen several families of transcription factors for genes that are differentially expressed among neural crest derivatives. We have identified a novel paired homeodomain gene, called DRGll, that is expressed in sensory but not sympathetic neurons and in dorsal horn neurons in the spinal cord. This transcription factor provides a molecular marker for the sensory lineage in the mammalian peripheral nervous system. Moreover, the expression pattern of DRGll suggests that it may function in some aspect of the establishment or maintenance of appropriate connectivity between some sensory neurons and a subset of their central targets. DRG11 Is a Sensory Neuron-Specific Paired Homeodomain Protein DRGll is a transcription factor expressed in sensory but not in autonomic neurons. While other transcription factors have been reported to be specifically expressed in sensory neurons [such as Pax3 (Goulding et al., 1991), the POU-domain protein Brn 3.0 (Gerrero et al., 1993) and the bHLH proteins NSCLl, NSCL2 (Begley et al., 1992; Gobel et al., 1992), and NeuroD (Lee et al., 1995)], in most of these cases a detailed temporal analysis of expression in autonomic ganglia was not reported. As mentioned earlier, the expression of several genes (such as trkA) in sympathetic ganglia (Birren et al. , 1993) lags 285 4-5 days behind their initial expression in sensory ganglia (Martin-Zanca et al. , 1990), perhaps in part because mitotic arrest of sympathetic neurons similarly lags behind that in sensory neurons (Rohrer and Thoenen, 1987; Verdi and Anderson, 1994). Most of the studies cited above examined expression at early gestational stages (El l.5-E13.5), and therefore expression of these genes in sympathetic ganglia at later stages might have been missed. Here we have shown that DRGll is expressed by sensory neurons, but not by sympathetic and other autonomic neurons (such as parasympathetic and enteric neurons) even at late gestational and postnatal stages (Fig. 6 and data not shown). The timing of DRGll expression, 1-2 days after neurons are first detected in the DRG by expression of panneuronal markers such as SCGlO, suggests that this putative transcriptional regulator is unlikely to be required for initial neuronal differentiation. Rather, this protein is likely to regulate later-developing aspects of sensory neuron phenotype or function. In this respect, the homology of DRGll to another PHD protein, Phox2 (Fig. 3B), is of interest. Phox2 is expressed specifically in developing autonomic ganglia but not in trunk sensory ganglia (Valarche et al., 1993), a pattern which is strikingly complementary to that of DRGl 1. Correlative data from in vivo and in vitro experiments, as well as DNAbinding data, suggest that Phox2 may be involved in the expression of neurotransmitter synthesizing enzymes such as DBH (Tissier-Seta et al., 1993). By analogy, DRGll could play a role in specifying some aspect of the complex neurotransmitter phenotype of sensory neurons. Alternatively, if as suggested below, DRGll function is important in appropriate synapse formation, its downstream targets could include cell surface proteins important in establishing or maintaining proper connectivity. In this respect it is of interest that Phox2 has also been shown to regulate the promoter of Ncam, a cell surface adhesion molecule (Tissier-Seta et al., 1993; Valarche et al., 1993). An intriguing feature of DRGll expression is that it is detected both in sensory neurons and in a subset of their target neurons in the spinal cord, specifically those in the dorsal horn (Fig. 6A). A subset of these central neurons receive input from the NGF-dependent population of sensory neurons, which (as shown by double-labeling with antibodies to trkA) also expresses DRGll. This sug- FIG. 4. Expression of DRGl l mRNA in the trunk spinal cord region of a rat El2.5 embryo. Serial sections were hybridized with anti-sense cRNA probes for (A) DRGll, (B) SCGlO, (C) PAX3, and (D) Islet-!. Control hybridization using a DRGll sense-strand probe gave no signal (not shown). FIG. 5. Expression of DRGll mRNA in the trunk spinal cord of an EJS.5 rat embryo. Sections were hybridized using anti-sense cRNA probes for (A) DRGll , (B) SCGJO, and (C) PAX3. Arrowheads indicate the DRG, arrows DRGll expression domains in the dorsal spinal cord. No DRGll expression was detected outside of the nervous system or in autonomic ganglia (not shown). ~ Q) C A _ -~ ~ .... .. .,.-: ' i -·..c ;;J ,~ • .-~ . ~,;. PAX3 DRG11 ' D -- B -~ \ ~ ,, • .c ~ ~ - Isl-1 ··-- " _SCG10 . ·------------~ ··- ' ,.., ~-:'\ , . "' ~'i. '6·. r 5 f ! f , , l ~r,;{ l' ;, i 't _I • • • -~ • .• {'.;.,. (. . . .~ ·t · i. ~ . J {!: I·' , • PAX3 SCG10 ' I • ~ W'"' , , . ,\ ',f\l C~•.•~:.•i O"· -~ }. ••1..,~ I •.'' •.d(y · · ., ;lii" DRG1 1 r.(; , •\,:.,.'\:i, ,14\ ' .~... -- _ 1.- [1 1f:} -~~--;.· '?, .j ; i ;:·:>,'· . 1· . ,f >.::;- '\1-,;, / 'Cl- ;t,.\~• '. ,, •;:,•,., \ W.,,,,··· , ,!t:•_~~i.·A'.:,. ,~. •. f I, .; :. r$ ,''--tt • • )-' •.._i , 'L-1- \ ,, : ~d-tj{;..:ii' ;,1·.· . ,,., _.<,i,i/. /J · , . 1: , ,?,)! i ·J ,:i,.. ,t\~ ~ •·· ~- . ~-- C B A w w ~- r,.I:.. DRG11 'i~ ,;;·.:, TrkA SCG10 ,· ,. 6 T rkB E Isl-1 ~ J . .. . , •• • '!. ,. , ...,, \ tt ·Y ; . , .,\ .,,;,- '; , :·, ~i'{it ff(-' •• .. ~;-;\.~'J:;·. ')?:· •.'At• ·~·~~· •.,.,! v:.'J...' ,,., , t•·t ~:.:·, .,, -~ ...,· ~""' /, ~ .... TrkC TrkA 1.V ,. . .. _ , D C \; SCG10 DRG11 • .. ·• • '•\?er~{; • 4-~•t,·7 '•1.- -a~ ,, I • :'r , ~ •\ ·1·• '( ......... 1' · ~ • rt-.\\ • !'"•· ~ ~\ ~ 1°J. :ilf .:tir21 •• ,?; •~t ,.;:,,.~,,•.\ .: • , . · •.,,.ii ."r . • . - ~• #_ y.. ,, ·'·,: • .-.._. . :: $.)~ ~;._... >fkT.1•,•., •• • t •~ •1 T • ' \; • ·;f \\..,.,.~[. •if:. •t•~:.~ ,;}, FIG. 6. Expression of DRGll mRNA in the trunk region of a ra}: E17.5 embryo. Sections were hybridized with anti-sense cRNA probes for (A) DRGll, (BJ SCGlO, (C) Islet-1, and (D) TrkA. Arrowheads (A, D) indicate DRG, filled arrows (A-DJ the sympathetic ganglia, and open arrow (A) the dorsal horn of the spinal cord. FIG. 7. Expression of DRGll mRNA in E13.5 trigeminal sensory ganglia. Serial sections through the same ganglion were hybridized with the probes indicated at the bottom. The number of cells expressing DRGll appears greater than that obtained using Irks A, B, or C probes (D- F) and comparable to that obtained using the panneuronal probes (B) SCGlO and (C) Isl-1. I sl- 1 >: :A; :i ;./,. ,L'. ..... \ ~- - - \' C •)- ~•: l,.) ..... ~ 135 288 gests that DRGll might function in regulating some aspect of synapse formation between sensory neurons and their central targets. However, DRGll expression within sensory ganglia was more extensive than that of trkA (Figs. 6 and 7); for example it was detected in large neurons which typically project to the ventral spinal cord (Snider, 1994). This suggests that DRGll expression is not restricted to NGF-dependent sensory neurons. This indicates that DRGll expression cannot be sufficient to specify the central connectivity of such neurons. However it could be necessary for some aspect of this process. This notion is strengthened by the fact that DRGll first appears in the dorsal spinal cord at E15.5, about the time at which the first sensory afferents are growing into this region of the CNS. Studies in the chick indicate that some aspects of dorsal horn development depend upon sensory axon ingrowth (Sharma et al., 1994). Experiments in progress are aimed at determining whether DRGll expression in dorsal spinal cord neurons is dependent upon such ingrowth, or is, rather, autonomously controlled. In turn, functional perturbation experiments currently underway should address the role of DRGll in the development of sensory neurons and their connections with the spinal cord. Strengths and Limitations of dRT-PCR as a Method for Identifying Cell- Type -Specific Genes The dRT-PCR method we have developed has several advantages over conventional brute force RT-PCR using degenerate primers. First and foremost, only those RT PCR products that display differential expression between cell types or tissues are selected for sequencing. This greatly reduces the number of PCR products that must be sequenced for any set of degenerate primers to identify interesting genes. Second, the intensity of the hybridization signals exhibited by a given clone annealed with multiple probes, as well as compared to other clones annealed with the same probes, provides a unique "fingerprint" for that clone. In our experience, clones displaying similar fingerprints usually contain the same insert sequence. Thus, by selecting for sequencing only those clones which exhibit different hybridization fingerprints, the characterization of redundant inserts is reduced. Third, although sequencing of inserts from a PCR experiment may no longer be rate-limiting due to automation, the analysis of the expression pattern of each of the PCR products still represents a major bottleneck. dRT-PCR reduces the number of genes whose expression patterns ha ve to be characterized by identifying differentially expressed sequences at an early stage in the procedure. These three factors combined allow a relatively rapid assessment of whether a set of degenerate Saito et al. primers will identify gene family members differentially expressed among a set of tissues or cell types. This in turn permits rapid screening of multiple pairs of degenerate primers, either within a single gene family or among several different families. Similar approaches related to dRT - PCR have been described previously by others (Boehm, 1993; Lai and Lemke, 1991; Wilkie and Simon, 1991). An unexpected problem we encountered is that the in vivo expression pattern of several differentially expressed dRT-PCR products was inconsistent with expectations based on their differential distribution in the original screen (T. Saito, L. Sommer, and D. Anderson, unpublished observations). In some cases, a gene was not detectably expressed in the expected tissue but rather was highly expressed elsewhere. For example, the NCM3 sequence, which behaved as if it were glial-specific in the dRT - PCR screen, was not detectably expressed by peripheral glial cells in vivo, but rather was abundantl y expressed in cartilage, a tissue not examined in the original dRT-PCR screen (data not shown). Similar situations were encountered in a screen of receptor tyrosine phosphatases (L. Sommer and D. Anderson, unpublished data). The reason(s) for this discrepancy is not clear. It may reflect aberrant expression of genes in some of the cell lines used as sources of cDNA. Alternatively, in cases where dissected tissues were used it could reflect the efficient amplification of sequences present in minor or contaminating cell types. These are biological problems rather than problems with the dRT-PCR method per se, but they illustrate the importance of judiciously selecting multiple tissues and /or cell lines for the initial dRTPCR screen. Another potential explanation is that some genes may be preferentially amplified by a specific primer set, which would be a problem intrinsic to the PCR method rather than to dRT - PCR. A final possibility, specific to dRT-PCR, derives from the fact that hybridization signal intensity in the clone blots actually reflects the relative abundance of that sequence within the gene family of interest, rather than in the cDNA population as a whole. In the limiting case where a given cell type expresses only a single family member, that sequence would represent 100% of the hybridization probe (and therefore would yield an intense hybridization signal from that cell type) even if it is actually expressed at very low levels in that cell or tissue. Such artifacts can be minimized by performing RNase protection experiments for individual clones on different tissues. In any case, the pattern of specificity exhibited by the dRT-PCR procedure should be considered tentative until confirmed by in situ hybridization data. This in turn raises another problem, in that the short (typically <250- 136 A Sensory Neuron-Specific Homeodomain Protein bp) PCR products provide insufficiently sensitive probes for the nonradioactive in situ hybridization procedure, necessitating additional time and effort to isolate longer cDNAs. The value of this method would be improved if sensitive and reliable nonradioactive in situ hybridization procedures could be performed directly with the PCR products. Nevertheless, the prescreening of clones in dRT-PCR still reduces the number of genes that have to be characterized by these more labor-intensive procedures. In summary, our data support the idea that transcription factors provide useful and specific markers for subtypes of neurons in the vertebrate nervous system. We have developed and applied a method that combines degenerate PCR amplification with differential hybridization to allow rapid screening of gene families for members which are differentially expressed among populations of neural cell types. This screen has yielded a novel paired homeodomain protein, DRGl 1, that represents the first sensory neuron-specific transcription factor identified in mammals. The expression pattern of DRGl 1 suggests that its mechanism of regulation and function should be interesting subjects for future study. In addition, this gene may provide a useful marker for cell biological and genetic studies of the development of sensory and dorsal horn neurons. 289 EXPERIMENTAL METHODS for PCR: CGGGA TCCTT(TC)ACIGCIT A(TC)CA(GA)(TC)TIGA and CGGAA TTC(GT)(GA)TT(TC)TG(GA)AACCAIAC(TC)TG. DNA was amplified by Taq DNA polymerase under the following conditions: 5 reduced stringency cycles using 94°C for 1 min, 42°C for 1 min, 55°C for 1 min, and 72°C for 1 min; followed by 33 cycles using 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min. PCR products were fractionated by agarose gels and DNA fragments of the expected size ( ~ 150 bp) were purified using the Mermaid kit (BiolOl). Purified DNA fragments were reamplified using 28 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min, purified by phenol extraction, and digested by EcoRI and BamHI. Digested DNA fragments with a size of ~ 150 bp were purified from agarose gels (see above). An aliquot of the digested DNA fragments was ligated into pBluescript for subcloning and transformed into Escherichia coli. Another aliquot of these DNA fragments was saved for use as a probe in differential hyb ridization. Each transformant was transferred into 70 µ,l of LB-amp medium in a well of a 96-well plate and cultured at 37°C for 14 h with shaking. Five microliters of the liquid culture was spotted onto replica filters (Genescreen, Dupont) using a multichannel pipettor. The filter was treated according to the manufacturer's protocol and hybridized with a 32 P-labeled probe prepared by random-primed labeling of the other aliquot of the restriction-digested PCR products. Nucleotide sequences of the clones which showed differential distribution were determined. Materials Southern Blotting of PCR Products Rat embryos and pups were obtained from timed pregnant Sprague-Dawley rats (Simonson Laboratories). MAH, NCMl, and Ratl cells were maintained as described previously (Birren et al. , 1993; Lo et al. , 1990). PCR products of RT-PCR were fractionated by agarose gel electrophoresis, transferred onto replicate Genescreen filters, and hybridized with 32 P-labeled probes derived from the DNA fragments encoding the homeodomains of DRGll, Pax3, or NCM3. As a control for the amount of amplified cDNA in the gel blot experiments, parallel aliquots were amplified for ,6-actin using genespecific primers. The PCR product corresponded to a 310-bp fragment spanning nucleotides 767 -1077 of the rat ,6-actin mRNA and was amplified with the following primers: TCATGAAGTGTGACGTTGACATCC and GTAAAACGCAGCTCAGT AACAGTC. Conditions were 20 cycles of 94°C for 1 min, 60°C for 1 min, 72°C for 1 min. Differential RT-PCR Molecular cloning was performed according to standard procedures (Sambrook et al. , 1989), with minor modifications. Total RNA was prepared by acid guanidinum thiocyanate method as described (Chomczynski and Sacchi, 1987) with slight modification. cDNA was synthesized from 2 µ,g of total RNA in 50 µ,l by using random hexamer primers and reverse transcriptase, and 1 µ,l of this reaction mixture was used for PCR. Degenerate oligonucleotide primers corresponding to the sequence coding for amino acids FT AYQLE and the complementary sequence coding for amino acids QVWFQNR [N-terminal and C-terminal portions of the paired type homeodomain from the Drosophila protein RK2 / repo (Campbell et al., 1994; Xiong et al., 1994)] were used cDNA Library Construction and Screening Poly(A +) RNA was purified from 1 µ,g of total RNA from El3.5 rat dorsal root ganglion by using oligo(dT) magnetic beads (Dyna!) and converted to cDNA using the Superscript choice system (Gibco BRL). The cDNA was ligated to a pre-annealed mixture of oligonucleo- •· SCG C DAG A •• .t: y •._.. t " • 1. 11 \ "" ' t ' " • -· ' • ' k• 1.\' ; ~ ,.(' ~· . ,.._ ' ' -.: t, . ,' .,..,. ,A_ \ f l ' '"" . , .. '• . t • ' , ~ • • -. • • ' # . • -~ /: ' , ' ir; ,./ . I "j . t I, L / • ~ , .. "" \ • / , 1.· : ,~ ..... . , I., ,. _. ; \ f I -,- ' (I t I ·, • ,• I , • • ~ - , ,.- ' 1. ' < . Cl •• 1 • • , , -~ , ") ".: - .s ,;'f • . ,!ff . • •, , ,. • ·• ._ • ...; _ # •a/ \ .- . ;...~.. ' f.I )'-,,., ' ' ." l •' 8 {~ . \· '- f'. f-1 ,! 7 I ' • · ·· ' ,,. ._5:; '1~•/\~ • • , ".., , ' . . ··: .... _;_ ). I~~ J')-'t' .,~ ,-.._.._, ~. II•/; ' ,, • . ' f:,,\I'.'' • ·I -;' ··•\ - • , " '\\ \ ei ·,.~;-✓ ,' /· .~,i,. 1 ' • , , --,f ► I •; 1):1~ .. • ,., " • • {- I ·i· ..,_, \ < { , ,; I , , • " ' -~'• It_ , .. •'· • ~,.-· ... . ..,. J 'I I • ·1. " \ · . " " ·· ,:· . · , 1- .~ • .r ~ I ,-~·. ' . I ...... ' t • •• i\ •' • I o •, • •,· . ...I , ~ •. . ~/ •' ~. • I :.,,,,~-. _.-. ' \ :_;., • I , ,/ I ' -V .' ' •• . • I r· i,.) .,. , ~. • ' (• ·, ... , 'I -~ : • "' t Q ' t r ,{I' ,' .;) ·, r. .. . • .,..,r- .- ., - - r.~ - ..; :.:.;~ '{>,f ,~- • •• • \ . : - ( , . . ' • ••f' ' ~ '· ~,t .. • ; . - "'~ . • .- :_,-(.• . J , ., ; ., ' ·•· -":•, ·~ ,- I ' , • i /,111 •;:~• ;'°1/\ • , I\ _1., . , ·. :',n1 • ( \ • •• • - ~_ .., _. \ • • ·, I , ' ..... • D .·\ , . · j ~ _ q . •·- .: ,•/ ; ..\ :~~I-. ,• .,. •.\,. - • .. : ;;':,,----;:_..:.~ .1:i,~·-J .~..~·-:.~.7/ ... ~ '° • -- ·~>t•·.r, ! \••:,:"'~:;:.~',/,, ..-. _., . . .:-,. ,~ 'f / ..-·: to_ ~: ~ ~ , · ' <" ~ -.--..:· \ • ~ ..... ' ' _ ......, - ..~ ,. 1' ,1 4,. r -- .,,, , . _:·..·-.. .\\ . • .. . .... ' -. •~ _: ..~J· : 1_ , ~,.. ~ r,: , ~J. } .. , ./ ..._ ~ , .._ •, \ -~ .. . ,,..,. " • / •• ✓ ,.r ._ ··~· , .. ,', ' ..... - :-~-:.···, -~- 11_ ~ ,: ' \ ' ,. -· · -.:-, , ,- '~ , ... " . r · .• ~ -.~,· - . ,. ,. ?~·., ·.:.::--.;-..-. . . , .... .• ·· • "" ..... t ,.,' t , { · .;. • ,j, j. , •• .. ' t . ' • \~ ' ' . , ' .-. ~ ~ ; · ·-x: ·· ~ ''J>v:J'J, '~ ., • -\ f , . .-1:.v~ . ·' . Ji~ ~ ,~i ."'.! l-.: ... ,,, , . . .. .. ~ I ' Jr,t.r.. . • --_,.., I ,7 . A,"' ,.' J.{ t •, ~ "'1> . . ..i: l ·<4 ;:,, , • '. ,,.. ·~1 "j1\~ - . I 7t!' ~ .,;r, • 1 ~ <~--·t ~If,<-{!:'~ ..,..-.,. ~"' • \~-~t:-,~ · .. ~ · -- .,~ ::#-r<:~ t: . ,J ' ' ~-.• •.,..,.,.. ~~~~_:,-~)~~ ~ &i'~r;.:eA,r-4 . r. r i' ~-.~- --. •- tt'1 ' - FIG. 8. Expression of DRGll protein in dissociated sensory neurons. Postnatal Day 2 (A, B) dorsal root ganglia and (C, D) superior cervical ganglia were cultured overnight, fixed, and stained with monoclonal antibody to DRGll protein. Note that the antibody stains the nuclei of several neurons in the DRG cultures (A, arrowheads), but not those in the SCG cultures (C, arrowheads), which exhibit only a low level of nonspecific background staining as determined by comparison to controls stained with an irrelevant primary antibody (not shown). (B, D) Phase-contrast micrographs of the bright-field images shown in (A, C) are provided to reveal glial cells (small arrowheads) which are not labeled in either culture. FIG. 9. Coexpression of DRGll and trkA in some sensory neurons. Embryonic Day 16.5 DRG were dissociated, cultured overnight, fixed, and doublelabeled with monoclonal antibody to DRGll (purple staining) and polyclonal antibody to trkA (brown staining). Open arrows (A) indicate neurons containing DRGll in their nuclei which also express trkA (solid arrowheads). (A, B) Solid arrow indicates a neuron that expressed neither protein. The majority of trkA • neurons coexpressed DRGll at this stage, but some DRGll • cells did not express trkA. ~ .~ . • ;Er." -l w I--' 138 A Sensory Neuron-Specific Homeodomain Protein tides ACTGAAGCCAAGGTAGGATCCG and (phosphorylated) CGGA TCCT ACCTTGGCTTCAGTAG. The ligated cDNA was purified using the Spinbind PCR purification system (FMC) and amplified by 3 rounds of PCR (16 cycles, 12 cycles, and 9 cycles) using a phosphorylated oligonucleotide (CTACTGAAGCCAAGGTAGGATCCG) under the condition of 95°C for 1.5 min, 64°C for 2 min, and 72°C for 7 min. After each round of PCR, a small aliquot of the amplified DNA (one-tenth of the total reaction mixture) was used for the next round of PCR. (32 P]dCTP was added in the third round of PCR to calculate the amount of amplified DNA. Approximately 1.3 µg of DNA was obtained after the third round of PCR. cDNA fragments longer than 500 bp were separated on a size-fractionation column (Gibco BRL) and cloned into the 'A. Zap!I vector (Stratagene). The library was screened using a DNA probe containing the homeodomain of the DRGll RT-PCR product. Two cDNA clones containing 2.4- and 1.0-kb inserts were obtained. Nucleotide sequences were determined on both strands using Sequenase (USB). The sequence of DRGl 1 has been submitted to GenBank under Accession No. U29174. In Situ Hybridization Nonradioactive in situ hybridization was performed as described previously (Birren et al., 1993). The following probes were used: SCGlO (Stein et al. , 1988); trkA, from pDM97 (a gift from L. Parada); trkB, from pFRK16; trkC, from pRtrkc8 (gifts from G. Yancopoulos). Islet-1 and PAX3 clones, which carry 0.9 and 1.2 kb of cDNA, respectively, were obtained by RT-PCR. Antibody Production DRGll protein expressed in bacteria was gel-purified and used to immunize Armenian hamsters. Hamster spleen cells were fused with P3X63Ag8u.l mouse myeloma cells. Supernatants were first screened on dot blots of recombinant DRGll protein, and positives from this screen were rescreened by immunofluorescence staining of CHO cells transfected with a mammalian DRGll expression construct. Details of staining procedures, secondary antibodies, etc., are available on request. ACKNOWLEDGMENTS We thank Steven Padilla for technical assistance, Susan Ou, Ling Wang, and Liching Lo for advice and help in the production of monoclonal antibodies, Andy Groves for advice on immunostaining procedures, and Lukas Sommer and Kai Zinn for their helpful comments on 291 the manuscript. O.).A . is an Associate Investigator of the Howard Hughes Medical Institute. Note added in proof We have recently succeeded in identifying DRGl 1 transcripts by Northern blot. A single species of -2.4 kb was detected in mRNA from DRG and neural tube. The size of this transcript is similar to the size of the DRGll cDNA clone that was sequenced. This suggests that the clone is close to full-length and therefore that the deduced amino acid sequence of DRGll represents the intact polypeptide. REFERENCES Begley, C. G., Lipkowitz, S., Gobel, V., Mahon, K. A., Bertness, V., Green, A. R., Gough, N. M., and Kirsch, I. R. (1992). Molecular characterization of NSCL, a gene encoding a helix-loop-helix protein expressed in the developing nervous system. Proc. Natl. Acad. Sci. USA 89: 38-42. Birren, S. )., and Anderson, D. J. (1990). A v-myc-immortalized sympathoadrenal progenitor cell line in which neuronal differentiation is initia ted by FGF but not NGF. Ne11ro11 4: 189-201. Birren, S. J., Lo, L. C., and Anderson, D. J. (1993). Sympathetic neurons undergo a developmental switch in trophic dependence. Development 119: 597-610. Boehm, T. (1993). Analysis of multigene families by DNA fingerprinting of conserved domains: Directed cloning of tissue-specific protein tyrosine phosphatases. 011coge11e 8: 1385-1390. Campbell, G., Goring, H., Lin, T., Spana, E., Andersson, S., Doe, C. Q., and Tomlinson, A. (1994). RK2, a glial-specific homeodomain protein required for embryonic nerve cord condensation and viability in Drosophila. Development 120: 2957-2966. Campbell, S., Inamdar, M., Rodrigues, V., Raghavan, V., Palazzolo, M., and Chovnick, A. (1992). The scalloped gene encodes a novel, evolutionarily conserved transcription factor required for sensory organ differentiation in Drosophifa. Genes Dev. 6: 367-379. Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thioc ya nate-phenol-chloroform extraction. Anal. Bioclie111. 162: 156-159. Clary, D. 0 ., Weskamp, G., Austin, L. R., and Reichardt, L. F. (1994). Trk.A cross-linking mimics neuronal responses to nerve growth fac• tor. Mo/. Biol. Cell 5: 549-563. Ericson, J., Thor, 5., Edlund, T., Jessel!, T. M., and Yamada, T. (1992). Early stages of motor neuron differentiation revealed by expression of homeobox gene /s/et-1. Science 256: 1555-1560. Fann, M.-J., and Patterson, P. H. (1994). Neuropoietic cytokines and activin A differentially regulate the phenotype of cultured sympathetic neurons. Proc. Natl. Acad. Sci. USA 91: 43-47. Frigerio, G., Burri, M., Bopp, D., Baumgartner, S., and Noll, M. (1986). Structure of the segmentation gene paired and the Drosophila PRO gene set as part of a gene network. Cell 47: 735-746. Gerrero, M. R., McEvilly, R. J., Turner, E., Lin, C. R., O'Connell. S., Jenne, K. J., Hobbs, M. V., and Rosenfeld, M. G. (1993). Brn-3.0: A POU-domain protein expressed in the sensory, immune, and endocrine systems that functions on elements distinct from known oc· tamer motifs. Proc. Natl. Acad. Sci. USA 90: 10841-10845. Gobel, V., Lipkowitz, S., Kozak, C. A., and Kirsch, I. R. (1992). NSCL2: A basic domain helix-loop-helix gene expressed in early neurogenesis. Cell Growth Diff 3: 143-148. Goulding, M., Chalepakis, G., Deutsch, U., Erselius, J., and Gruss, P. (1991). Pax-3, a novel murine DNA binding protein expressed during early neurogenesis. EMBO /. 10: 1135-1147. 139 292 Grueneberg, D. A., Natesan, S., Alexandre, C., and Gilman, M. Z. (1992). Human and Drosophila homeodomain proteins that enhance the DNA-binding activity of serum response factor. Science 257: 10891095. Ito, K., Morita, T., and Sieber-Blum, M. (1993). In vitro clonal analysis of mouse neural crest development. Dev. Biol. 157: 517- 525. Johnson, J. E., Birren, 5. J., and Anderson, D. J. (1990). Two rat homologues of Drosophila acl,aete-sci,/e specifically expressed in neuronal precursors. Nnt11re 346: 858-861. Kandel, E. R., Schwartz, J. H., and Jessell, T. M. (1991 ). Principles of Ne11rnl Science. Elsevier, New York. Lai, C., and Lemke, G. (1991). An extended family of protein-tyrosine kinase genes differentially expressed in the vertebrate nervous system. Ne11ron 6: 691-704. Lai, E., Prezioso, V. R., Tao, W., Chen, W. S., and Darnell, J.E. Jr. (1991) . Hepatocyte nuclear factor 3a belongs to a gene family in mammals that is homologous to the Drosophila homeotic gene fork !,end. Genes Dev. 5: 416-427. Langman, J., and Haden, C. C. (1970). Formation and migration of neuroblasts in the spinal cord of the chick embryo. /. Co111p. Ne,, rol. 138: 419-431. Lee, J. E., Hollenberg, S. M., Snider, L., Turner, D. L., Lipnick, N., and Weintraub, H. (1995). Conversion of Xenop11s extoderm into neurons by NeuroD, a basic helix- loop-helix protein. Science 268: 836-844. Libert, F., Parmentier, M., Lefort, A., Dinsart, C., Vansande, J., Maenhaut, C., Simons, M. J., Dumont, J. E., and Vassart, G. (1989). Selective amplification and cloning of 4 new members of the G-protein coupled receptor family . Science 244: 569-572. Lo, L., Guillemot, F., Joyner, A. L., and Anderson, D. J. (1994). MASH-!: A marker and a mutation for mammalian neural crest development. Perspect. Dev. Neural. 2: 191-201. Lo, L.-C., Birren, S. J., and Anderson, D. J. (1990). V-111yc immortalization of early rat neural crest cells yields a clonal cell line which generates both glial and adrenergic progenitor cells. Dev. Biol. 145: 139- 153. Martin-Zanca, D., Barbacid, M., and Parada, L. F. (1990). Expression of the Irk proto-oncogene is restricted to the sensory cranial and spinal ganglia of neural crest origin in mouse development. Genes Dev. 4: 683-694. Matsumoto, S. G. (1994). Neuronal differentiation in cultures of murine neural crest. II. Development of capsaicin-sensitive neurons. Dev. Brain Res. 83: 17-27. Miller, D. M., Shen, M. M., Shamu, C. E., Burglin, T. R., Ruvkun, G., Dubois, M. L., Ghee, M., and Wilson, L. (1992). C. elegance unc-4 gene encodes a homeodomain protein that determines the pattern of synaptic input to specific motor neurons. Nnt11re 355: 841-845. Murphy, M., Reid, K., Hilton, D. J., and Bartlett, P. F. (1991). Generation of sensory neurons is stimulated by leukemia inhibitory factor. Proc. Natl. Acnd. Sci. USA 88: 3498-3501. Nye, J. A., Petersen, J. M., Gunther, C. V., Jansen, M. D., and Graves, B. J. (1992). Interaction of murine Ets-1 with GGA-binding sites establishes an ETS domain as a ne w DNA-binding motif. Genes Dev. 6: 975-990. Opstelten, 0 . J., Vogels, R., Robert, B., Kalkhoven, E., Zwartkruis, F., De Laaf, L., Destree, 0. H., Deschamps, J., Lawson, K. A., and Meijlink, F. Saito et al. (1991). The mouse homeobox gene, S8, is expressed during embryogenesis predominantly in mesenchyme. Mech. Dev. 34: 29-41. Rohrer, H., and Thoenen, H. (1987). Relationship between differentiation and terminal mitosis; Chick sensory and ciliary neurons differen- tiate after terminal mitosis of precursor cells, whe reas sympathetic neurons continue to divide after differentiation./. Ne11rosci. 7: 37393748. Ruit, K. G., Elliott, J. L., Osborne, P.A., Yan, Q., and Snider, W. D. {l 992) . Selective dependence of mammalian dorsa l root ganglion neurons on nerve growth factor during embryonic development. Neuron 8: 573 587. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Lnborntory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schneitz, K., Spielmann, P., and Noll, M. (1993). Molecular genetics of aristaless, a prd-type homeobox gene involved in the morphogenesis of proximal and distal pattern elements in a subset of appendages in Drosophila. Genes Dev. 7: 114-129. Sharma, K., Korade, z., and Frank, E. {] 994). Development of specific muscle and cutaneous sensory projections in cultured segments of spinal cord. Develop111ent 120: 1315-1323. Snider, W. D. (1994). Functions of the neurotrophins during nervous system development-What the knockouts a re teaching us. Ce/I 77: 627-638. Stein, R., Mori, N., Matthews, K., Lo, L.-C., and Anderson, D. J. (1988). The NGF-inducible SCGl0 mRN A encodes a novel membrane-bound protein present in growth cones and abundant in developing neurons. Neuron 1: 463-476. Tissier-Seta, J.-P., Hirsch, M.-R., Valarche, I., Brunet, J.-F., and Goridis, C. (1993). A possible link between cell adhesion receptors, homeodomain proteins and neuronal identity. C. R. Acnd. Sci. Paris 316: 13061315. Tsuchida, T., Ensini, M., Morton, S. B., Baldassare, M., Edlund, T., Jessell, T. M., and Pfaff, S. L. (1994). Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes. Cell 79: 957- 970. Valarche, I., Tissier-Seta, J.-P., Hirsch, M.-R., Martinez, S., Goridis, C., and Brunet, J.-F. (1993). The mouse homeodomain protein Phox2 regu lates NCAM promoter activity in concert with Cux / CDP and is a putative determinant of neurotransmitter phenotype. Develop111ent 119: 881 - 896. Verdi, J.M., and Anderson, D. J. (1994). Neurotrophins regulate sequential changes in neurotrophin receptor expression by sympathetic neuroblasts. Ne11ron, in press. Walther, C., and Gruss, P. (1991). Pax-6, a murine paired box gene, is expressed in the developing CNS. Develop111ent 113: 1435-1449. Wilkie, T. M., and Simon, M. I. (1991). Cloning multigene families with degenerate PCR primers. Methods: Companion Methods Enzy1110/. 2: 3241. Xiong, W. C., Okano, H., Patel, N. H., Blendy, J. A., and Montell, C. (1994). Repo encodes a glial-specific homeodomain protein required in the Drosophila nervous system. Genes Dev. 8: 981-994. Zhao, G.-Q., Zhou, X., Eberspaecher, H., Solursh, M., and de Crombrugghe, B. (1993). Cartilage homeoprotein 1, a homeoprotein selectively expressed in chondrocytes. Proc. Na tl . Acnd. Sci. USA 90: 8633-8637. Received for publication May 26 , 1995