The Stereochemistry of Dehydro-β-Carotene. Studies on the Interaction of Some Carotenoids with Boron Trifluoride and N-Bromosuccinimide. Provitamin A Activity of a C₄₂-Carotenoid Citation Wallcave, Lawrence (1953) The Stereochemistry of Dehydro-β-Carotene. Studies on the Interaction of Some Carotenoids with Boron Trifluoride and N-Bromosuccinimide. Provitamin A Activity of a C₄₂-Carotenoid. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/0TZT-6C69. https://resolver.caltech.edu/CaltechETD:etd-05092003-171754 Abstract NOTE: Text or symbols not renderable in plain ASCII are indicated by [...]. Abstract is included in .pdf document. An extensive study of the stereoisomerization of dehydro-[Beta]-carotene is presented. Two of the cis isomers have been isolated in crystalline form and described. The nature of "isocarotene" has been clarified. Treatment of [Beta]-carotene with boron trifluoride under defined conditions leads chiefly either to its stereoisomerization or chemical conversion. Two of the several conversion products have been crystallized; a third, exhibiting strong fluorescence in ultraviolet light is also described. Three new carotenoids, termed isocryptoxanthin, isocryptoxanthin methyl ether, and dimethoxy-[Beta]-carotene, formed by the addition of water or methanol to dehydro-[Beta]-carotene in the presence of boron trifluoride are described. The structures of the first two compounds have been determined; the location of the substituent groups in the third is, at present, uncertain. Dehydro-[Beta]-carotene, bisdehydro-[Beta]-carotene, and anhydroeschscholtzxanthin have been isolated from the end products of the reaction of [Beta]-carotene with N-bromosuccinimide. Bisdehydro-[Beta]-carotene is a new carotenoid whose probable structure is discussed. The considerable provitamin A potency of 16,16'-homo-[Beta]-carotene is discussed from the standpoint of its novel structural features. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry and Physics) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Minor Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Zechmeister, Laszlo Thesis Committee: Unknown, Unknown Defense Date: 1 January 1953 Additional Information: Title varies in the 1953 Caltech Commencement program: I. The Stereochemistry of Dehydro-Beta-Carotene. II. Studies on the Interaction of Some Carotenoids with Boron Trifluoride and N-Bromosuccinimide. Ill. Provitamin A Activity of a C₄₂-Carotenoid. Record Number: CaltechETD:etd-05092003-171754 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-05092003-171754 DOI: 10.7907/0TZT-6C69 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 1700 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 14 May 2003 Last Modified: 26 May 2023 19:06 Thesis Files Preview PDF (Wallcave_l_1953.pdf) - Final Version See Usage Policy. 9MB Repository Staff Only: item control page

THE STEREOCHEMISTRY OF DEHYDRO- p “CAROTENE. STUDIES ON THE INTERACTION OF SOME CAROTENOIDS WITH BORON TRIFLUORIDE AND N-BROMOSUCCINIMIDE. PROVITAMIN A ACTIVITY OF A C,5-CAROTENOID. Thesis by lawrence Nallcavae in Partial Fulfiliment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1953 The auther wishes to express his desp appreciation to Professor Zechmeister for his interest and guidance in the problema connected with this research, ACKNOWLEDGMENTS I wish to thank the California Institute of Technology for generous financial assistance throughout the period of the author's graduate studies, I am indebted to Professor H. J. Deuel, Jr. and Dr. J. Ganguly who carried out the bioassays for provitamin A activities and to Dr. A. Elek and Mr. G. Swinehart for the microanalytical determinations. ABSTRACT An extensive study of the stereolsomerization of dehydro fh ~carotene is presented, Two of the cig isomers have been isolated in crystalline form and described, The nature of “isocarotene"™ has been clarified, Treatment of fp-carctene with boron trifluoride under defined conditions leads chiefly sither to its stereoisomerization or chemical conversion. Two of the several conversion products have been crystallized; a third, exhibiting strong fluorescence in ultraviolet light is also described. Three new carotenoids, termed isocryptoxanthin, isocryptoxanthin methyl ether, and dime thary- fi “carotene, formed by the addition of water or methanol to dehydro- p “carotene in the presence of boron trifluoride, are described. The structures of the first two compounds have been determined; the location of the substituent groups in the third is, at present, uncertain. Dehydro~ B ~carotens, bisdehydro- fi “carotene, and anhydro-~ sschscholtzxanthin have been isolated from the end products of the reaction of P-carotens with N-bromosuccinimide,. Bisdehydro- “carotene is a new carotenoid whose probable structure is discussed. The considerable provitamin A potency of 16,16'-homo~ 3 “carotene is discussed from the standpoint of its novel structural features, Table of Contents I. A Study of Stereoisomeric Dehydro- ff ~carotenes, A. Theoretical Part ...csessececseececssescrecevesscssesseces ob A. Introduction cescscececscsseseesesececectusecessssenee od 2. Steric Possibilities in the Dehydro- > -carotens Set . 2 3. Sterecisomrization of Dehydro- B -carctens eseeeseeee 5 1. Materials and Methods ....sccccsscccrcscreseseasssecns fa 2. Preparation of Dehydro- f-carotene cece eee ceeenseeeae BA 3, Stereocisorerization of Dehydro- p WCATOTENG eeesssevee BL ai. A Study of the Interaction of Boron Trifluoride with f-Cerctene. Ae The oretical Part @aeeoet eet sees sete te eee tant r es haorsrsrseesee £0 l. Introduction Pseaer see es, ete ee ev saeser ese set erer ees etvaesaee aan 40 2. Stereoisomerization of [§-Carotene by the Action of Boron Trifluoride .c.dscscccsnnvccessvecrancearenten 40 3. Some New Compounds Obtained from |9-Garotene by Prolonged Treatment with Boron Trifluoride .......0.0e. 4l B, Experimental Part .s.secsesesscesavcsesnsseccesssrscccoses £8 1, Materials ............00. wneeeeeees se rserevenrasevenes AG 2. Stereolsomerization Experiments cicssesenceccsenscceee 48 3. Formation of Some Conversion Products of B-Garctens.. 50 III, Some New Carotenoids Produced by the Action _of Water or Methanol on the Boron Trifluoride-Dehydro- (> ~carotene Complex. A. Theoretical Part crceccecsescccanescencecssesesencessatsne 50 1. Introduction ...ccecccceee ee ensccrecreseccecscsesceres 56 BR. Iscoryptoxanthin ove eet ae see, ea ere eee tse ees aeeaeeanevrease ae 57 3. 4. 5. Isocryptoxanthin Methyl Ether and Dimethoaxy- P-carotens... eee eee & es e@e 9 Gee eens @evt@ Peete eevee wp aa aene 64 Two Proposed Mechanisms for tha Formation of Isocryptoxanthin and ite Methyl Ether from Dehydro~ (} -carotene meetzteove ae evesceeene ea taeetreasravaeereeaete T1 Frovitamin A Activities of Isocryptoaxanthin, Isoeryptoxanthin Methyl Ether, and Dimethoxy~ ~carotene @aereeeee eee a seesr eae taeseeeesareeneevaeteee ee ene 75 Experimental Part Per ecrer ese ces eee eevee eset saasreseeseaaevoeaeeas 78 B. i. 2. 3. 4. De IV. On the Ae Materials co. ccecseccaeseccesecsccersccsscssessencsses 18 Artethaxy- 8 -oarotene and Dimethaxy- p “carotene .... 78 Tsooryptoxanthin ...sccsnsccescecessscenseceseescasese 83 Action of Acid Chloroform .......cssenecveresecesecees GO Freparation of Isocryptoxanthin, Isceryptoxanthin Methyl Ether, and Dimethaxy~ f} -carotene fron a Steresoisomeric Mixture of Dehydro- (} “Garotenss ...00. 93 Interaction of fp-arctens with N-Bromosuccinimide. Theoretical Part Ones aeereeenaeet eter eeteazeer eevee tseeeveaeseenee 95 1. 2. 3. Introduction and General Remarks .cscseccereccceassese 9S Structure of Bisdehydro- (3 “CATOTONG sescececsocessese 9 Anhydro-eschacholtemanthin .....-e sever erste ecereverce 104 B. Experimental Part Pave eaeeegne sees est eaten ene vr eaveeseeereeveteseet 106 Appendix, i. 2. Bs - 4. Materials Soave sere e sen ae ee ser eee estar ateaeehneesneeeeeeotesne eae 106 Praparation of Bisdehydro- 8 “carotene and Anhydro~ aschgoholtzxanthin Peter ae ee Geeta eee ee eseaesenen ses eeeeeaa a 106 Some Stereolsomers of Bisdehydro- 6 ~carotene ........ 111 Some Stereolsomers of Anhydro~eschscholtzxanthin ..... 112 Provitamin A Potency of a P-Carotene Homologue 1. Re Introduction and General Remarks .ecseccceacsseerseces LLB Experimental @*e se eevre se caer teeraneeaeteoat tense ees see eae ueae te 118 I, ~"L~ A STUDY OF STEREOISOMERIG DEHYDRO- p -CAROTSES. A. THEORETICAL PART 1. jntroduction, In 1932, Kuhn and lederer (21) had prepared a carotenoid, which they called "isocarotena", by a spontaneous conversion of b-carotene tetraicdide when standing in acetone sclution, After a treatment with thiosulfate and, eventually, by repeated crystallizations from benzene-methanol thase authors isolated a product which melted at 181° and showed spectral maxima in light petroleum (v.p. 70-80°) at 504, 475, and 447 m, By further oryetallization of "“isocarotene" from light petroleum Karrer, Schopp, and Morf (16) obtained glittering violet needles (m.p. 193°). They found that this compound had 12 conjugated double bonds and corresponded to the formula CyoHs, » In 1940 Karrer and Schwab (17) proposed the now generally accepted structure for the compound which they correctly tarmed "dehydro~ p ~carotens " (Formola I.). a le (3 GH. ie (3 BON rs p67 ~“ =CH-CH2C-CH2CH~CH2C~CH=CH-CH2CH-C=2CH-CH=CH-C=cH-cH=0~ cH BR 2 3 4 HAG o-0H i1,0-6 Cll. ™~ XZ 3 FN 2 Formula I, Dehydro-P-carotene, (The sterically unhindered double benda are designated by italicized numbers.) In this Thesis it will be shown that Kuhn's "lsocarotene" preparation was a mixture of (mainly) three stereolsoreric dehydro- 6 “carotenes and unreacted p-carotene. Because all-trang-dehydro~B -carotens is very mach less soluble than any of its cis isomers or P-carotene, it is possitile, in principle, to concentrate these latter compounds in the mother liquors and thus to obtain a relatively pure crystalline all-trans-dehydro~ -carotene, ~Q0 The melting point given by Karrer (16) correspands to that of the all=trang form and indicates that he was probably able to eliminate al] other components of Kuhn's "“isecarotene" without applying chromatography. Since, as will be shown, the ali~irangs form constitutes but a small part of "isocarotene", the crystallization method of purification could not have been very efficient. It is interesting to note that neither of the authors mentioned reported the yields obtained after repeated crystallizations. The application of chromatography as employed in this Thesis in addition to offering a simple means for the isolation of all=trang-dehydro- B -cerotene also led to the detection and isclation in pure form of two crystalline cis isomers of this carotenoid, Neither of the previous authors reported any inhomogeneity in their final preparations, although Kuhn (21) had reported the presence of p-carotere in some crude preparations of "isocarotene" before he had developed the proper experimental conditions, Although many preparations of "isocarotene" were made following the directions of Kuhn and Lederer a p-carotene free sample was never obtained (of. Table 1.). 2. Sterie Possibilities in it ia new well established that carotenoids and other compounds with long conjugated polyene systems are capable of undergoing reversible sis- trans isowerization,. The most extensive work in this field has been done by Zechmeister and his eollaberateors, espeeially Eseue, Iemmen, LeRosen, Pinckard, Polgaér, and Schroeder (32). The stereochemical study of dehydro- b-carotene is of special interest because, so far as is known, it is the first time that a Cr07 earotancld with a single bond in the central position has been investigated from this ahs fo viewpoint, although Oroshnik et al. (25) have discussed certain stereo~ chemical aspects of retro-vitamin A, a Cg- compound with one end grouping identical to that in dehydro- f -carotene. Table i. Composition of Kuhn's "Isccarotens™ Component Maxima in hexane MP. Amount (in % of (Beckman) starting material) All~trans~dehydro- 502,471, 445 m. 193° 16 p-carotene Ne o-dehydro~ 497, 467, 442 161° 15 p -carctens Ax Neo-dahydro~ 494, 464, 440 178° 10 P-carotens D* p-Carotene 479, 451 - 46 Unaccounted for - “ 13 According to Pauling (26) only those double bonds which do not have a methyl group attached to the next adjacent carbon atom are capable of assuming cis configurations easily. The presence of such a methyl group would, because of a spatial conflict, diminish the probability of a cis configuration being formed if the usual methods are employed in the trang—e cis isomerization experiments. However, in the synthesis of carotenolds as developed by Inhoffen, in Braunschweig, and Karrer, in Zurich, * This nomenclature is based on their respective positions in the chrom atogram (on lime-celita) of the stereoisomeric mixture obtained by iodine catalysis of any steric form of dehydro~ p ~carotene. whem certain double bende are formed by the partial hydrogenation of triple bends by means of specific catalysts, Such a conversion leads to the formation of a double bond with a oig configuration (9 ). Recently, by this method Garbers, Eugstear, and Karrer (5) hava synthesized a carotenoid-like molecule in which one or two "sterically hindered" double bonds in the conjugated system have a gig configuration, These compounds show degraded spectra, i.e., no fine structure is present in their extinction curves; furthermore the extinction coefficients are vary much lower than that of the all-trans form, and the maxima are shifted to much lower wave lengths. When catalyzed with iodine, these "hindered" isomers disappear and the same stereolsomeric mixture is obtained as would also be produced by a similar catalytic treatment of the all-trans form. The formation of a compound with a "hindersd" cig bond by the partial hydrogenation of an acetylenic bond is not surprising from the thermo- dynamic standpoint, for this step represents a decrease in the energy content of the molecule. In the case of the iodine catalyzed stereo~ isomerization of an all-trans carotenoid, however, the energy requirezenta for the formation of a hindered cis bond would be greater than for an unhindered ois bond; therefore the probability of such a configuration being assumed would be small.* | Formula I shows the locatio of the four sterically unhindered double bonds in dehydro- B carotene. Calculated on this basis, only 10 unhindered spatial forms are possible, In the course of the experiments to be * <A brief discussion of a similar steric hindrance caused by certain hydrogen atoms in the benzene rings of tha diphenyloctatetraene molecule has been given by Zechmeister and LeRosen (33). For the trans-gigs isomerization of diphenylbutadiene see Pinckard, Wille and zechmeister (27). wo Fue described the ten isomers listed in Table 2 were observed, Table 2, Observed Stereoisomeric Dehydro~ -carotenes (in order of descending chromatographic sequence on lime-cealite) Sterecisonsr Maxima in hexane solution (Beckman) All~irang-dehydro- 6 - carotene 502, 471, 445 m. Neo-dehydro- f -carotene A 497, 467, 442 Neo-dehydro- f -carotene B 494, 463, 439 Neo-dehydro~ P “carotene 6 491, 461, 435-6 Neo-dehydro- p carotene D 494, 464, 439 Neo-dehydro- fh ~carotene E 487, 457, 432 Ne o-dehydro~ B ~carotene F 489, 459, 433-4 Neo-dehydro- 6 ~carotene G 482, 453, 430 Neo-dehydro- B -carotene H 479, 452, 429 Ne o-dehydre- p-carotene J (degraded, no fine structure) The molecular extinction curves of these isomers are given in Figs, 1-7, The stereoisomrs listed in Table 2, except for neo J, were obtained by iodine catalysis. The neo J isomer was isolated from melt isomerization products, and will be discussed further in Section 3d, 3. Stereoisomerization of Dehydron P-oarotens.. As mentioned earlier threa aasily crystallizable stereolsomeric dehydro ~carotenes were found in "Isocgrotene", Stereolsomrization min axperinents were performed with all three of these compounds. It should be mentioned at this point that the ois isomers of most previously studied carotenoids which were adsorbed not much below their respective all-trans forms were crystallizable with difficulty, if at all, a. Stereoisomerization by iodine catalysis. In these experiments a solution of the carotenoid in hexane was treated with iodine (about 1% of the pigment weight) and illuminated under standardized conditions. Each of the three crystalline dehydro-~ P-carctenes gave stereolsoneric mixtures of nearly identical composition. The percentage of all-trang-dehydro-p ~garotene present in the stereo~ isomeric mixture wags remarkably emall, In analogous equilibria resulting from other C,o~ carotenoids the all-trans form was the predominant component as shown by the following figures (32): lycopene, 57% all-trans; 6 -carctene, 48%; andQ-carotene, 52% In the dehydro- Pb ~carotens series, as will be noted from Table 3, a cig form termed neo A is predominant (28-30%) while the all-trang compoumd constitutes only 17-18% of the stareolsomeric mixture. >. Stereoisomerization by Refluxing Solutions. Preliminary experiments showed that none of the three crystalline dehydro~ P -carotenes was appreciably iscmerized when allowed to remain in hexane solution at room temperatures for 24 hours. However, when such solutions were refluxed (in the absence of light) for 1 hour extensive stereoisomrization took place as shown in Table 4. The stability of all-trans-dehydro- P -carotene toward thermal isomerization is not as great as that of X- or P-carotene, but is about the same as that shown by lycopene. In similar reflux experiments (32) ~7- Table 3 Composition of the Stereolsomeria Equilibrium Mixtures Obtained Iedine Catalysis in Light of Three Crystalline Dehydro- P-carotenes, (The figures designate percent of starting material as established photometrically. In numerous experiments these values did not vary more than 1-2%,) teric form treated with All-trans-dehydro Ne o~dehydro~ Neo-dehydro~ p-carctens -carotene A b-carotene D All-trans 17 18 17 Neo A 28 30 30 Neo B 4 5 3 Neo G 10 11 10 Neo D 12 15 15 Neo E 4 6 5 Neo F 6 6 3 Neo G 3 z 1 Neo H 1 1 1 Unaccounted for 15 7 16 the following percentages of the respective all-trans starting materials were recovered: A-carotene, 92%; f-carctens, 86%; lycopene, 45% (the last figure refers to a 30 mimite reflux period), As shown by Table 4, half of the all-trans and neo A forms of dehydro- f-carotene were present in their unchanged configurations after refluxing; wo Bue Table 4, Composition of the Stereoisomeric Mixture Obtained By Refluxing Hexane Solutions of the Three Crystalline Dehydro~ P=carotenes for 1 Hour, (in @ of starting material) Starting Material All-trang~-dehydro Ne o-dehydro~ Neo-dehydro- fi-csrotene P-carotene A B-carotene D All-trang 53 5 2 Neo A 24, 50 13 Neo B - 4 - Neo © - (*) - Neo D 4 16 * 70 Neo E 2 10 8 Unaccounted for 17 13 7 the still higher stability of the neo D isomer as compared with that of the all-trans form is an unusual feature, In most carotenoid sets those cig isomers which can be produced from the all-trans form by stereochemical conversions often show a degree of thermostability which may approach but never exceed that of the all-trans form. co. Photochemical Sterecisomerization. The hexane solutions were illuminated with two 40 watt fluorescent lamps at room temperature, Of all the isomerization methods employed, such * A small amount of neo C was present in this zone and estimated together with neo D, -~O- illumination caused the least amount of steric rearrangement (Table 45). Tabla §. Composition of the Stereoisomeric Mixtures Obtained by Illumination of Hexane Solutions of the Three Crystalline Dehydro~ Dp -carotenes for 8 Hours. (in % of starting 'material) Starting Material All-trans-dehydro- Neo-dehydro~ Neo-dehydro~ b-carotene Proazotene A P~carotene D All-trang 84 5 1 Neo & 8 86 6 Neo D 2 3 91 Unaceeunted for 6 6 2 Insolation (i.e., exposure to intense sunlight) of carotenoid solutions offers another means for causing their stereoiacnerizaticon, Because of incenetant atmospheric conditions (smog, rain, cloudiness) it was not possible to carry out an extensive study of the effect cf insolaticn on dehydro- f ~carotenes. During a three hour insolation pericd with the all- trans compound more than half of the pigment was converted into unidentified colorless preducts. As established chromatographically, the remaining Pigment mixture showed the following composition: all-trans, 48%3 neo A, 40%3 and other cig isomers (mainly nec D), 12%. d. Sterecisomerization by Melting Crystale, The fusion of a sample of a1) Sreng-dehydve- fi ~caretene (mp. 193°) in an evacuated glass tuke caused extensive stersoilscomerization, As expected, a large fraction of the plgment was destroyed. The composition of the stereolsonerized mixture obtained from numerous experiments was quite uniform, Except for the absence of any neo H, all the steric forms of dehyaro- P carotene producad by iedine catalysis were also present here (Table 6). In the fusion experiments, however, an additional gig isomer was formed consistently and, since it constituted the lowest zone on the colum, it was termed ne o-dehydro~ B -carotens J, Table 6, Composition of the Steraclsomeric Mixture Obtained by Fusing Crystals of All-trang-dehydro~ ~carotene at 200° for lz Minutes (in @ of starting Material) All-trans 8 Neo A 14 Neo B 2 Neo 6 12 Neo D 14 Neo E 3 Neo F | 1 Neo G 1 Neo J i Unaccounted for 46 The neo J isomer differed from all the othera in that its extinction curve showed a degraded spectrum, i.6., a single broad band with almost no ell= fine structure, Although in different experinents some minor variations at the top part cf the curve were observed, the maximum was consistently at 446-448 mi. in hexane (Fig. 5). These smell spectral variations in the extinction curves were protably due to its rather rapid stereclsomerization in solutions, In one exceptional instance the maximum was located at 436 mi. indicating that a different isomer was formed (Fig. 8). When a dilute solution of the neo J isomer in hexane was treated with a trace of iodine and illuminated for one mirmmte, a deepening of the color from a pale yellow to a bright yellow or orange was easily noticeable to the naked aye, Yhen this change was followed with the visual apectroscope, the appearance of sharp bands of 496 and 467 mi. could be observed, Chromatographic analysis of such solutions showed that the usual iodine Catalyzed sterecisomeric mixture was present, A nurber of ois carotenoids are known which also lack definite fine structure and which behave similarly when subjected to iodine catalysis, @.8., Prolycopens and pro-Y -carotene (24), both naturally occurring pigments, but, so far as is know, ‘tthe observations reported here represent the first instance of an in vitro conversion of an all-trang C,o- carotenoid into a gis isomer presenting the described spectral characteristics, Neo- dehydro- fh ~carotene J was also formed by melting crystals of the neo A and neo D isomers (Figs. 6 and 7). Although the absence of fine structure in the extinction curve of the neo J isomer and the low value of its molecular extinction coefficient (about 60% of that of the all-trans form) are indicative of a polycis structure, such a configuration can not be elaired with certainty for this compound, since, as mentioned earlier, Garbers, Eugster, and Karrer (5) alee have demonstrated that a carotenoid whose molecule contains a sterically hindered cis double bond shows very similar characteristics, Possibly, under the energetic conditions of the fusion experiments such a gis double bond could be formed, It is interesting to note that no neo J isorer appeared when neo-dehydro= P carotene A was welted at 170°, although the yield of neo J amounted to 13 at 200°. —/3— xlO-4 mol E Icom res te ee , Be oat y BE ey aa merce Rms on ao & # Fisure 1. Uolecilar extinction curva in herane of ell . SERS wseeewmimnis 2 2 ry : MCETOUSHS F smmery CEOS soluitiany onee, go 2 L hew £4, 1 a or 1 hourg sesees, after ielire catalrais in light, bee —4—. \ D: | jt yy ad OO 300 = 350 400 450 500 mu dolaquiar extinctien curve in bexame of nececehwdre= mcarotens At ~—, fresh solitiong e-r-, after refluxing ott hoe L hourg venene, after lodive eatalyels in lett. —/5— x 1o~4 mol Icm E SOO my Figure 3, Molecular extinction curve 4n hexane of nec-debydro- Beonrotens Di wee, fresh sclutiong “~~, after rofliccing foxy L hourg eewees, after fodine catalyaia in light. —{6— 300 500 mu 450 350 Holecular extinction ourwe in vemane of aa gig B-carotene ~ Ae Sea Soaks Fag ° ‘ad gp sr, nee By oom y nes Cs wowewe, ms Dy omens y nec Be tame aan RO G. ioure &, ~—~/T— "SUE TL WT SPRATERSO eappoT ce_re Swecee SUM TLL TOS qSGT; fommem E(SUEET@Tle go uopsny soar) p¢ euegecrto. d | aQapdyep-oeu JO GUO UT GAINS WTZOUTINe acmmoepay *¢ amit OSr OOv OG OOF OG¢d a a ee ee ee ee ™. ™, | I | Te) a3 NX | ,-Ol X . — ol —18— T T my TT TT L [\ J T / 4 ] . , 7 L. \ = \ lO F- ~ ‘ — \ 1 \ \ _ { x< L. ' _ SE / ul oF : | 7 / 5 / a / -_= ~ / — j~ St / ft iz - an 7] A . f Y 7 \ = \ 7 { | | L on | j 4 H L | j | i Lf | { ~: 300 350 400 450 500 My Pisure ©, ~tarotens J S (tree f 2Posn golntion; 17 x|O°4 mol F Icm 300 350 «| 400 450 500 Me Firures 7, Molecular extinction curv in hexon one of raoeCeltrdrae “carotens J (fren fusion of nec D LgOmRR) Rey, Sresh goluiLong eewrre, after to aine catalesls in Diget. —20— a T . ee _ ‘comand . —-,-" _ z Cy : a —_— 4 pone —_ 450 dehydro-Becarotens formed by fusion of the alle fresh solution «rrene, alter 400 - \ ike! * \ | ™m = C7 _ | 18 -— . _ nm \ eo) | ; 1S — WN a = _ — “a _ | \ L 1 | i ! mee ee ee ee © LO =_ w9| p-Ol x jou 4 7 Figure &, Molecular extinction curve in hexane of nee isonor on a . (ens COrr B. EXPERIMENTAL PART 1, Materials and Methods, In all of the experimental work reported in this Thesis, certain standard techniques, equipment, and reagents were used, In order to eliminate undue repetition a description of these is given in this section which applies to all the work herein, unless stated otherwise, ae Adaorbents and Solvents. The brand of. caleium hydroxide used was Sierra Hydrated Lime, Superfine, United States Lime Products Corp., Los Angeles. In order to obtain an increased flow rate in chromatograms this material was mixed with Celite, No. 545, Johns-Manvillle Co. The term lime-celite refers to a 2:1 mixture (by weight). The alumina used was Activated Alumina, Grade F, 80-200 mesh, Aluminum Ore Co., East St. Louia, T11., reground to 200 mesh. When the flow rate of some reground batches was very low, the alumina was mixed with celite (4:1 by weight). In general, the developed colwm was removed | by extrusion with s wooden pestle. For chromatographic work hexane (commercial grade), acetone (C.P.= USP), and benzene (Reagent Grade) were used, Acetone-hexane or benzene- hexane mixtures were often used as developers, The figures given in these cases (e.g., "5% acetone in hexane") refer to volume percent. Acetone was used for all elutions; in some instances acetone-methanol 5:1 was employed in order to elute strongly adsorbed pigments. For cry~ stallizations only reagent grade solvents were used (except for hexane, which was commercial grade). For spectroscopic work in the ultraviolet region a purified hexane was prepared, FPhillips' comrercial grade was treated with fuming sulfuric acid until its optical density (after washing and a final distillation) was close to that of distilled water in the region 200-300 m1. b, General Notes. Evaporations and concentrations were perforred in vacuo (water aspirator) while a stream of pure dry nitrogen was bubbled through the solution, Sintered glass funnels were used for elutions, and where possible, all-glass apparatus was used for evaporations, washings, etc, Visual spectroscopic readings were made with an Evaluating Grating Spectroscope, Zeisa (leewe-Schumm modification) with light filter BG-7, Extinction curves were taken with the Beckman photoelectric spectro~ photometer, Model DY. The quantitative extinction curves are basad on two or more independent weighings, All melting points were taken on an electrically heated Berl block in evacuated capillaries and are corrected, 2. Preparation of Dehvdro- P -carotene, a, P-carotene The b-carotene used was obtained from a crude orystalline carotene preparation available in 1 gm. ampules from Barnett Laboratories, Long Beach, Galif. Hach gram of this material was chromatographed on lize- celite from 3 hexane solution, in a large conic percolator, (48 x 21 x 8 on.) or, alternatively, on six, 30 x § om, lime-celite columis. Except for 29 minor zones near the top only p-carotene (upper main zone) and A-carotene were observed in these chromatograms (ratio, 3:1). After development with hexane was complete, the percolator was inverted and the adsorbent cake released by gentle tapping of the glass walls. The P-carotene Z2One was cut out and eluted. The pigment was then transferred to hexane by the addition of water and the solution thoroughly washed in the LeRosen apparatus (23). After drying over anhydrous sodium sulfate, the solution was evaporated to dryness and the residue crystallized from the mininun anount of hot benzene or chloroform by the addition of methanol {4 to 5 volumes). The crystals were centrifuged, washed once in the centrifuge tube with hot methancl and then dried in the Abderhalden apparatus at 55° and 0.1 mm. pressure for 1 hour. Yield, 350-400 mg. In some other experiments the commercial crude carotene mixture was twice recrystallized from benzene-methanol (or chloroform-methanol) without a preliminary chromatographio isolation of pure P-carotens, Such samples were found to contain 70% P-carotene and were used directly for the preparation of dehydro- B-carotens since the A-carotene content did not affect the next step, One gram of crude carotene crystals gave about 650 mg. of twice erystallized product, b. Formation of Kuhn's "Isocarotene" from P-carotene. The method descrited here is a modification of that given by Kuhn and lederer (21), Although a crystalline preparation containing 70% P-carotene {see above) was used in later experiments, only those experiments which were based on the use of a pure crystalline P-carctene for the preparaticn of dehydro- [> ~carotene will be described in detail. The use of the 70% p> -carotene mixture merely involved reducing the amounts of all necessary mh solvents and reagents by 1/3 for a given amount of starting material, To 500 ml. of a solution of 800 mg, of iodine in hexane, contained ina31. rowd bottom flask and cooled to -10°, was added, with good mechanical stirring, a solution of 800 mg. of B-carotene in 1 liter of hexane, also cooled to -10°, The P-carotene was added over a one minute period and stirring was continued for 2 more mlnutes, The black ‘precipitate of B-carotene iodide was then filtered off without suction on a large fluted filter paper (diam., 40 om.) and washed into 15 1, of acetone with $ 1. of the same solvent, Ocvusional awirllng was necyssary to bring all the iodide into solution, The dark brown solution was allowed to stand at room temperature for 20-30 min. The solution was then trans- ferred to a large separatory funnel and vigorously shaken with a solution of 5 gm. of sodium thiosulfate in 250 ml. of water, This last procedure caused the solution to turn red. Four hundred milliliters of hexane was then added followed by sufficient water to transfer nearly all of the pigment to the epiphase. The aquecus~acetone hypophase, which remained somewhat yellow was discarded. The hexane layer was washed in the continuous apparatus and dried. After evaporating to dryness, the bright red, powdery residue was transferred to a centrifuge tube with the minimum amount of hot benzene (50°) and crystallized out by the dropwise addition of warm methanol, Yield, 250-300 mg. of the pigments constituting the “isocarotene" described by Kuhn and Lederer. ce. Chromatographie analysis of Kuhn's "Isocarotene", Bighty~two milligrams of recrystallized “isocarotene" (m.p. 172°) was dissolved in 150 ml. of hexane and developed on a 30 x 8 em. lime-celite colum with 8% acetone in hexane, A description of the chromatogram appears below (figures to left indicate width of zones in mm.): 4 reddish-brown 45 interzone 70 orange-red - - - = = Zones I 2 interzone 30 reddish-crange - - - Zone IT 2 interzone 35 reddish-crange ~ - - Zone III 15 intergone 5 pale yellow - --~ - Zone IV oe aa ay ae a ce ee ME OE Br ss OS OS a ee NS Each of the numbered zones was cut out, eluted, and transferred with water to hexane. The colored filtrate, which amounted to several hundred milliliters, was washed free of acetone, then, after drying, concentrated to about 50 ml. and develored with hexane on a 30 x 8 em. lime~-celite colum. 1 brownish 60 several yellowish zones (traces) 20 yellow ----- ~~ Zone V 3. interzone 40 orange-brown - - - = Zone VI 6 interzone 10 orange~yellow - ~ = Zone VIT 5 interzone (slightly colored) 10 yellow---<---- Zone VIII The numbered zones from the above chromatogram were also cut out, eluted and transferred to hexane, The amount of pigment in each zone was estimated photometrically on the basis of previously determined molecular extinction coefficients (see the section below for the characterization of individual dehydro- F-carctene isomers). The composition of the “Asocurotene" produced in this typleal experiment was as follows: Dba Fraction I: All-trang-dehydro- B carotene : 16% of "isocarotend! mixture. Fraction II: Neo-dehydro- [p-carctene Ai 154 Fraction III: Neo-dehydro- B -carotane Ds 10% Fraction TV: A miner starsolsomer of dehydro~ f -earotens : 0.4% Fractions V-VII: Unreacted P-carotene and some of its stereoisomers: 46% Starting material unaccounted for: 13% A chromatographic resolution of "isocarotana" without preliminary ra- crystallization showed approximately the same composition: All~trane-dehyaro- /> “carotene: 11% Ne onde hydr o- ~carotene A: 12% Ne o-dehydro- -carotene D: 8% Other stereoisomeric dehydro- P ~carotenes: 1Z Unreacted- f) -carotene (including stereoisomers): 50% Unaccounted for: 184 d. Isolation and characterization of all-trang-dehydro- P-carotene and two crystalline stereoisomers (neo A and D), For the purpose of isolating pure all-trang-dehydro-p “carotene and some stereoisomers, the hexane solution, obtained from 800 mg, of prearotsne by its reaction with iodine and subsequent treatment with thiosulfate and acetone atc. as described in section »., was concentrated to 300 ml. and developed directly on lime-cslite with 6% acetone in hexane. Two 30 x 8 cm, colurms were required, the corresponding zones of which were conbined, =2)= 6 brownish=red 4 interzone 4 yellow 5 interzone QO orange-red -"= 7-7 ~ Zone I 2 interzone 65 reddish-orange - - - ~ Zone Ii 2 Interzone 50 orange-"-*--=-* 7-5 Zone III 12 orange-ysllow 4 yellow 3 interzone 5 purple 5 interzone 10 pale yellow Oo on an we me ae we Ee Te TER Filtrate Sh on ce =e Dee Oe ce a oe ne Each of the zones I-ITI was eut ont, eluted and transferred to hexane. The three fractions were crystallized as described below for the casa of the ell-trang-dehyaro-P- carotene fraction (from Zone I). The solution was evaporated to dryness and the dark red, crystalline residue transferred to a 15 ml. centrifuge tube with a few milliliters of warm chloroform. (Fractions II and III which are more easily soluble than the trans form may be transferred with warm benzene, The centrifuge tube was placed in a bath at 50° and the chloroform completely evaporated by directing a gentle stream of dry nitrogen at its surface through a capillary tubs, The residue was then dissolved in the minimum amount of benzene (about 2 ml.) at 50° and crystallized with constant stirring by Slow dropwise addition of methanol down the walls of the centrifuge tube. Thus, the methanol was warmed to about the bath temperature before reaching the benzene solution. Dark purple-red, glittering crystals of dehydro- P-carotene appeared after a few drops of methanol had been introduced, but the addition was continued until the total volume amounted to 4-5 milliliters. (Fractions II and III crystallized less easily and required ~2h= somewhat larger amounts of methanol.) The tube was removed from the bath and allowed to stand at room temperature for 1 hr. The crystals were then sharply centrifuged and dried in an Abderhalden apparatus at the temperature of refluxing acetone and 0.1 ma. pressure for 1 hr, Yields: 28 mg. of the all-trans form (plus 2 mg, in mother liquor) ; 48 mg. of neo A form (plus 15 mg. in mothar liquor); 18 mg. of neo-dehydro~ P-carctene D (plus 15 mg. in mother liquor), All the unreacted p-carotane (about 300 mg.) was collected in the filtrate of the chromatograms, Thus about 25% of the 800 mg, of P-carctene which had reacted had been converted to dehydro- P -carotenes; the 400 ng. of P-carotene which was unaccounted for had evidently been converted to colorless destruction products and, to a lesser degrees, to wnidentified pigments. Characterization of the crystalline products, 1.) All-trane-dehydro- Prearctene . The compound crystallized from benzene-methanol in long red plates with characteristically cut ends (Fig. 9). Macroscopically, the glittering orystals were purplish red in color. Melting point: 192-193° (corr.). Solubility: Sparingly soluble in hexane, more easily in benzene or (especially) chloroform, Insoluble in methanol. Partition test: On partitioning between hexane and 95% methanol the compound is found entirely in the epiphase, Analysis: Calculated for Cy ots, 2 C, 89.83; H, 10.17. Found: 6, 89.80, 89.913 Hy 9404, 9095. Spectral maxima, in hexane: 504, 474 mi. (visual); 502, A471, 445 ma. (Beckman). Molecular extinction coefficient (at A max): Ee = 16.6 x 104 (in hexane). Chromatographic behavior: On lime-celite the compound requires 6~8% acetone in hexane for development, Its sone was located well above that of P-carctene but below lycopens on this adsorbent, 2.) Neo-dehydro- } ~carotene A. This isomar crystallized from benzene-methanal in red diemond shaped plates (Fig. 10). Maecrosnopically, the crystels had a glittery, dark red appearance. Melting point: 161-163° (corr.). Solubility: More soluble in hexane, benzene, and chloroform than the all-trans form but less so than Prcarctene. Practically insoluble in rethanol, Partition test: Completely epiphasic. Analysig: Calculated for Qype)? C, 89.833 H, 10.17. Found: 6, 89.67, 89,703 H, 10.31, 10.22, Spectral mexina, in hexane: 500, 470 m. (visual); 497, 467, 442 mp. (Beckman). Molecular extinction coefficient (at N mes)? Bron = 16,0 x 104 (hexane), Chromgtographic behavior: Separates from all-trang form on lime-celite upon developing with 6-8% acetone in hexane and occupies a lower position than the trang isomer. 3.) Ne o-dehydro~ (b-carotene De This isomer crystallized from benzene~rethanol in almond shaped crystals (Fig. 11); the individual crystals were smaller than the corresponding units of the trang or neo A forms. Melti oint: 177-178° (corr.) HEATIDE PON "S O€7 UoTP BOTs useY pue $y eusqoiso- * TOUR} ON eueZtEq WOAs PeZTTLeyeAIO g sueqoaso-¢ ~OUDPATSp-6 SU. ¢| ~xpdyop~cen ‘guezoauo-¢]-oupAyep-HUBAL-TLY ¢(QUeTt oF 456T) TI-6 Sematy Solubility: Easily soluble in hexane, benzene, and chloroform. Almost insoluble in methanol, Partition test: Then partitioned between hexane and 95% methanol only a trace of coloration is imparted to the hypophase. Analysis: Calculated for Crofts, 2 C, 89.833; H, 10.17. Found: C, 89.02, 89,663 H, 10.21, 10.07. Spectral maxing, in hexane: 497, 467 m (visual); 494, 464, 440 ma. (Beckman). Noleoular extinction coefficient (at A max.): Eon = 15.6 x 104 (hexane ), Chromatographic behavior: This isomer separates from, and is located below, the neo A form when developed om line-celite with 6-8% acetone in hexane. 3. Stereoisomrization of dehydro- b-carotene. &. Iodine catalyzed sterecisomerizaticon, In all quantitative studies of the cis-trans isomerization of dehydro~ p-carctene the following procedure was useds A weighed sample of the pigment was dissolved in hexane to give a concentration of 10-15 mg./L00 ml. A few drops of a solution of iodine in hexane, caleulated to contain about 1% of the pigment weight, was then added and thoroughly mixed with the solution in a suitably sized Pyrex yolumtrio flask. The flask was then placed at a distance of 60 om, from a pair of 3500° Mazda white fluorescent lamps (40 watts; 120 om, long) and illuminated for 30 mimtes, After illumination the mixture of isomers formed was resolved chromatographically and the amount of the individual components determined photometrically, Since the molecular extinction coefficients of only the three crystalline steric forme of dehydro- = F200 Pe f-carotens were determined quantitatively, those of the other stereo- isomers had to be determined indirectly on the basis of the known extinction value for the iodine catalyzed stereoisomeric mixture (tor. = 14.0 x 104 at Nuax.» 1 hexane). Just before taking the absorption spectra of the various cis~trans isomers they were rechromato~ graphed and any minor zones found on the column were eliminated. A11 the neo isomers except A and D were found to be very susceptible to thermal isomrization, even at room temperature. Iodine catalysis of all-trans-dehydro-f-carotene, Typical experiment: A solution of 11.7 mg. of tha pigment in hexane was isomerized and then developed on a 28 x 5.8 om, lime-celite column with 84 acetone in hexane, In the description of the chromatogram which follows the figures to the right indicate the quantities of the individual isomers expressed in 4% of the starting material, The spectral data refer to the positions of the main maxima in hexane solution as determined with the Beckman spectro- photome ter. 1 yellow-brown 45 interzone 28 orange-red (all-trans) 502, 471 m. 17% 4 interzone 30 yreddish-orange (neo A) 497, 467 28% 2 yellow (neo B) 494, 463 4% 2 interzone 20 yellow-orange (neo C) 491, 461 10% 10 orange to pink (neo D) 494, 464 12% 4 interzone 8 yellow (neo E) 487, 457 Lf 1 interzone 6 orange (neo F) 489, 459 6% 3 interzone 7 pale yellow (neo G) 482, 453 3% 3 interzone 6 very pale orange (neo H) 479, 452 1% 15% of the starting material could not be accounted for. «33a Each of the pigments was shown to be a dehydro- -carctene and not a structurally altered product by treating the hexane solution with a trace of iodine and recovering (chromatographically) the all-trang form of dehydro- b-carotene. The latter was identified by its spectrum and in the mixed chromatogran test, The isomers termed neo G and neo D could not be separated well enough on line-celite to permit an accurate determination of their amounts, I+ was found, however, that complete separation could be achieved on alumina when developed with a benzene-hexane mixture, 3:1, It is interesting that in the latter chromatogram the relative pooitiong of the two isomers were inverted, The neo A and neo B forms do not separate well either, even on alumina colums. In this case, since the amount of the neo B isomer formed was small, any errors made in its quantitative estimation could not have significantly affected the accuracy of the overall analysis of the stereo~ isomeric rixture., For purposes of spectral readings und extinction estimations the top part of the neo B zone was neglected when it was cut out of the colum; thus the possibility of its contamination by the neo A isomer was minimized. lodine catalysis of neo-dehydro- P-carotens A, Typical experiment: 8.7 ng. of the carotenoid in hexane was isomerized and then developed on a 28 x 5,8 cm. liwe-cellite colum with 64 acetone in hexane, 1 yellow-brown 35 interzone 20 orange-red (all-trans) 18% 2 interzonse 28 reddish-orange (neo A) 30% 2 yellow (neo B) 5% 3 interzone 20 yellow (neo C) 11% 8 orange (with pinkish leading edge) (neo D) 15 2 dinterzone 6 pale yellow (neo E) 6% 3 interzone 6 pale orange (neo F) 6% 5 interzgone 10 very pale yellow (neo G) 2% 6&6 interzone 10 very pale buff (neo H) 1% Starting material unaccounted for, 7%, Iodine catalysis of meoctehdre- boesrotene D, Typical experiment: 13.9 mg. of the carotenoid in hexane solution was isomerized and then chromatographed aa stated for the neo A form, 1 yellow-brown 40 interzone 25 orange-red (all-trang) 17% 3 interzone 25 reddish-orange (neo A) 30% 3 yellow (neo B) 3% 2 interzone 15 yellow {neo C) 10% 10 orange (with pinkish leading edge) (nao D) 154% 3 interzone B yellow {neo EB) 5% 2 interzona 5 orange {neo F) 3%, § interzone 8 very pale yellow (neo G) 1% 5 interzone 10 very pale buff {neo H) 1% Unaccounted for: 16% b. Stereoisomerization by fusion, A sample of the carotenoid (15-17 mg.) was sealed, in yacuo, in a glass tube (diam., 7 mm.) and kept in an oil bath at 200° for 14 minutes. (& longer period cf immersion led to excassive destruction.) after cooling in lee water the tuke was opened; the contents were dissolved in 2-3 ml, of benzene and diluted to 50 ml, with hexane, This soluticn wag develored on a line-celite column with 67% ucetone in hexane, The composition of the mixture was determined in the same way as stated shove for the isomerization by iodine catalysis, Tsoner ‘vation of ali~ trans-dehydro- ~carotene, Typical experinent: Meade es oe. ne one ee i A BO EERE Be De MEDD SD CAL iL AS the The bexane sclution obtuined from the fusion of 15 mg, of compound was developed on a 24 x 4.8 cm, lime-celite columm with 6% acetone in hexane, Pigures to the right show vercentuges composition, 1 yellow 4 brewnish-red £0 interzone 18 orange-red {all-trans} 8% 2 ianterazcne 15 reddish-orange (neo 4) 14% 1 yellow (neo B) 2% 3 dintersone lz yellow-orange (neo G) 124 4 orange-pink (neo D) 124 10° yellow (neo 5) 3% 1 dinterzone 5 pale orange (neo F) 1% 4. intersore 8 very pale yellow (neo G) 1% 45 interzone & very pale buff (nec J, possibly a poly-gig isover) 18 FPiltrate pale yellow-green OP are Oe Ore: ma Mae we OO Ht ae ON oe BR All the fractions, excapt the all-trang end neo A forms, had to be re~ chromatographed tefore the spectra were taken, After eluticn and transfer into hexane, some contaminants absorbing in the ultraviolet regions were elininated by repeatedly washing with 95% methanol, Some larger scale experiments were undertaken to obtain the "poly-gis” ~36= isomer in ancunts sufficient for crystallization; however, this substance could not be crystallized since all attempts led to almost complete stereo- isorerization, The pigment collected in the filtrate did not contain any dehydro- P-carotene isomers and was, undoubtedly, a decomposition product of the starting material, In all, 46% of the all-trans-dehydro~ 6 ~carotene was unaccounted for, Stereoisonerlzation of neo-Gehydro~ B mcarotene As It was found that 9 temperature of 170° was not sufficiently high to cause the formation of any “poly-gig" isomer, whereas at a temperature of 200° approximately 12 of the starting material appeared in the "poly~cig" form. A detailed chromatographic analysis of the isomeric mixture, beyond the estiration of "poly-cig" compound, was not undertaken. Qualitatively, the chronmate- gram was very similar to that shown in the case of the all-trang melt. As was demonstrated in the theoretical section of this Part, the "poly~cig"-dehydro- P-carctene has a degraded spectrum tut develops the usvel maxima upon icdine catalysis. Stereolsonerization of neo-dehyare- B-csrotens. BD, A melt isomerization of this isomer was undertaken with 17 mg. of substance to ascertain whether a "poly-cig" isorer was formed, This was found to be the case; the yield amounts to approximately 1% of the starting material upon fusion at 200° for 14 mimites, Isomerization of a sterecisomeric mixture of dehydro-|y-carotenes. A "poly-cig" form was obtained also by fusion of a sample of Kuhn's "iso- carotene", from which all p-carotene had been removed chromatographically,. Again, only about 12 of this form was isolated. -37- eo, Sterecisomerization by refluxing solutions. The general procedure was as follows: A dilute solution of the carotenoid in hexane was refluxed in an all glass apparatus for 1 hour. Tha heating souree vas a Glas-Col heating mantle and the voltage was adjusted so that the hexane vapors rose about 1 inch in the condenser. The flask and condenser were covered with a black cloth to exclude light. The isomerized solution was resolved chromatographically and its composition determined spectrophotometrically. Stereoisormerization of all-trans-dehydro- B-carctene. The refluxed ~ r golution of 8,5 mg. of substance in 200 ml, of hexane was developed on a 28 x 5.8 om. lime-celite colum with 6% acetone in hexane. 2 browmish 45 interzone 45 orange-red (all-trans) 53% 2 interzone 25 pale orange {neo A) 245 2 interzone 10 pale yellow {nec D) 4% 2 interzone 4 very pale yollew {neo E) 2% Seventeen percent of the starting material used was not recovered, Sterecisomerization of neo-dehydro- Prosrotens A, The starting material amounted to 4.2 mg. and wes dissolved in 100 ml. of hexane. The isomerized solution was developed on a 24 x 4.8 em. lime-celite column as stated above: 1 orange-brown 25 interzone 6 orange-red (all-trans) 5% 1 interzone 40 reddish-crange (neo A) 50% 2 yellow (neo B) 4% 2 interzone 8 pale orange (neos G and D) 16% & interzone 15 yellow-brown (neo E) 10% 2 interzone 8 very pale orange (neo F) 2% Thirteen percent of the starting material could not be accounted for. ~38- Sterecisomerization of neo-dehyaze- b-carotene D. 8.0 mg, of substance was dissolved in 200 ml, of hexane, The refluxed solution was developed on a 28 x 5.8 om, lire-celite eolum as stated above, 1 yellewigsh 30 dinterzone 10 pale orange (all-trang) 2% 4 interzone 20 pale orange (neo A) 13% § inierzone 45 orange (nao D) 70% 1 interzone 10 yellow . (neo BE) 8% Only 7% of the starting material could not be accounted for, d, Photochemical stereolsomerizatian, In each of the experiments a solution of 2 mg. of pigment in 25 ml, of hexane, contained in a Pyrex volumetric flask, was illuminated for 8 hours at a distance of 30 em, from the light source described in the section on iedine catalyzed stereoisomerization, After illumination the solutions were developed om 20 x 3,8 om, lime-celite colums with 6% acetone in hexane, Sterecleomrization of all-trang-dehydro- 2) carotene , { 1 pale yellow 15 interzone 55 orange-red (all-trans) 84% 1 interzcne 8 pale reddish-orange (neo A) 84 2 intergone 5 pale yellow-orange (neo D) 2% Six percent of the starting material was unaccounted for, -39= Stere ocisoverization of neo-delydren f-serotene As 1. pale yellow-brown 20 interzone 8 pale orange (all-trans) 5% 1 interzone 50 reddish~orange (neo A) 86% 2 intergone 6 pale yellow-orange (neo D) 3% Six percent of the starting material was unaccounted fer, Stereoisomerization of neo-delvéro- B-carotene dD. 1 pale yellow 20 interzone 5 very pale pink (all-trans) 1% 2 interzone 6 pale orange-red {neo A) 6% 3 interzone 35 orange (neo D) O1¢ Two percent of the starting material was unaccounted for, ~£0- II, A STUDY OF THE INTERACTION OF BORON THIFLUCRIDE ‘TITH P-CaRcrmie, A, THEORSTICAL PaRT i. Intreducticn, Twelve years ago H, H. Strair, (30) reported that sore carotenoids in curbon tetrachloride sclutions were converted by boron trifluoride into Nunstable blue pigments" which, by treatment with alcohol, were reconverted to yeliow plerents that “did not exhibit definite absorption maxina"™, Ags far as is known, there is no further literature on this subject, Tt will be shown that a brief treatment of p-carotene in hexane solution with boron trifluoride leads principally to its gis~trans iscners’, while a more prolonged treatnant with this reagent completely converts the starting material to other preducts, sone colorless, Two of the pigrents present in the complex reaction mixture obtained from the latter experinrent have bean isolated in crystalline form, 2, Steracisorerization of P-Carotene by the Action of Boron Trifluoride. When @ hexane solution of P-carotene was shaken with boren tri- fluoride etherate for 1 to 4 minutes the reagent (hype-) phase turned dark blue, Upon mixing the liquid with 956 methanol, the original orange-red color of the P-carotene solution reappeared, after transferring the pigrents to hexane, chroratographic analysis showed that the Prcarotene was extensively sterecisomerized, The results of a 1 minute and a 5 minute experinent are compared in Table 7, The extent of stereoisorerization within a given tize was essentially constant, whether the experinents were performed in total darkness or in normal indoor lighting, The boren Ek a AE a a RE REL a Fa NE we a a a * Some preliminary observations were made in this laboratory by J. Leemann, -41- trifluoride catalyzed cis~trang isomerization of p-carotene, therefore, must proceed by means of a mechanism different from that of the light dependent iedine catelyzed isorerization, and is possibly related to acid catalyzed stereciscmerizations (28). Table 7, Pigment Mixture Obtained from a~Carotene by a Brief Treatment with Boron Trifluoride’ at Room Temperature. (The figures refer to % of starting material as estimated spectrophotome trically.) Duration of Treatment Component of tha Mixture 1 Min. & Win. leo b-carotene U 19 17 All-tran ap =carotene 36 35 Heo- fh searctane A 6 6 Neo~ Bj -carotene EB 14 12 Other reo isomers 5 3 of p-carotene Pigrents not belonging to the . 1 2 “carotene set (estimated as " P-carotens" ) Unaccounted for (including some 19 25 fluorescent destruction preducts) 3. Some New Compounds Obtained from PrGarotens by Prolonged Treatment with Boron Trifluoride les As mentioned above, in the boron trifluoride sterecilsomerization experivents some plements not belonging to the Prearotene set apreared in small quantities on the column; furthermore, a colorless preduct which shoved blue fluorescence in ultra-violet light was collected in the chromatographic filtrate, Since the 5 mimite treatment preduced more of these new substances than did shorter experiments, the shaking tine was extended for the purpose of further increasing the yields, Jt was found that upon shaking the hexane solution of Prcarctene with the reagent for a full hour, the hexane phase became almost colorless and, after the addition of methanol, the recovered pigment mixture did not include either Prcarotene or any of its steric forms, Subsequent chromatograms showed the presence of at least five new pigment zones, The main sone was much more weakly adsorbed than Prearctene and was collected in the filtrate together with the fluorescent substance previously mentioned, More strongly adsorbed were three pale, greenish-yellow zones, the upper two of which were located above Prearotene in the mixed chromatogram test (Table 8), Pigment IV did not form a homogeneous zone and, except for demon- strating ita non-identity with P-carotene in a mixed chromatogram test, waa not investigated further, | Pigment I orystallized easily from benzene-rethanol and melted sharply at 214° (corr.)3 its spectral curve is given in Fig. 12. Pigment IT (Fig. 13) was established to be a cigs form of I; because it was formed only in very small amcunts no attempt was made to erystallize it, Pigrent III crystallized with sore difficulty from benzene-rethanol 3 it melted at 157-148° (corr.) (spectral owrve, Fig. 14). , eo) anne 14 hid Fg Sh cy) Of fe Pe eee netien cure in homane of Pd a wele ene altar dedine catelysis i SOLTLONS wewewes, gator Tecine —4y4— my 450 400 350 a stinction curve in hexane of Pigment TI: -—-, ores: clutiong wewene, aftar Galina emtsiyaie in Light. Fimve 13, & 2 —Y5— “FACTT UP STSATHARO AUTPSZ Leg ZH fememee FO yil QBGLT Semen PTT VIOLET YE TO RUBKAL UE eae mopRouyng ° YE eandyg OSb O0b sage oo¢ os2 Table 8. Visually Observed Spectral Maxima in Hexane of Some New Pigments Obtained from B-Carotene by Treatment with Boron Trifluoride and Subsequently with 95% Methanol. - (In descending order of adsorption on lime-celite) Pigment I 448, ‘422 m. (very sharp maxima) Pigment II. 444, 419 m. (very sharp maxima) Pigment III | 468, 441 mp. (very sharp maxima) Pigment IV 423, 455, 428 mp. (somewhat blurred maxima) Analysis showed that Pigment I was not a hydrocarbon but had the approximate composition C 404560 Pigment III, on the other hand, was a hydrocarbon whose analysis was indicative of the formula Cy Hag or Cc 407545 Since, however, the positions of the spectral maxima of Pigments I and IIT indicate the presence of, respectively, 8 and 9 con~ ' fugated double bonds further experiments would be required for a structural clarification. The colorless fluorescent zone mentioned was separated from some pigmented compounds by adsorption on alumina. The absorption curves as obtained from two different experiments are shown in Fig. 15. It may be noted that maxiina at 228 mp. are characteristic for a number of naphthalene derivatives (ef, the pertinent curves in (4)). The conversion of carotenoids to naphthalene derivatives is not unknown, Kuhn and Winterstein (22) having isolated 2, 6~dimethylnaphthalene, among other compounds, in trace amounts by the thermal decomposition of these pigments. Figwe 15. Eetimtion cur: estainsa in se teitinuerice on eo CAT GLANG » im hosare of fluoresce: ee 7 gts oat gi pM £00 350 300 250 B, | EXPERIMENTAL PART 1, Materials. The boron trifluoride etherate used was either the Eastman Kodak product, redistilled, or prepared by passing the gas into ethyl ether (Reagent Grade) ae described in Part III. p-Carctene was prepared from a commercial carotene sample (Barnett Labs., Long Beach, Calif.) by erystallization of the chromatographically isolated product. 2. Stereocisomrization Experiments, One minute experiment: A solution of 20 ng. of p-carotens in 25 ml. of hexane contained in a 50 ml. volumetric flask was shaken (by hand) with 2ml. of the etherate for 1 minute. The reagent phase turned dark blue immediately, The contents were then added to 100-200 ml. of 95% methanol in an Erlenmeyer flask and swirled for a few seconds until a bright orange-red color appeared, The pigment mixture was transferred to the hexane phase by the addition of water and the colorless aqueous-methanol hypophase discarded. The solution was washed in the IeRosen apparatus for 10 minutes, dried over sodium sulfate, and developed with hexane on a 28 x 5,8 om, lime-celite colum (Figures to right indicate spectral mexima in hexane (Beckman) ). 20 colorless 3. orange unidentified; 478, 451 m. 35 interzone . 50 yellow neo P ~csrotene U3 475, 447 45 orange all-trans~ [-carotene; 480, 452 15 pale yellow neo~ -carotene A; 467, 440 25 orange-yellow neo--carotene B; 470, 444 5 interzone — 5 orange a cis--carotene; 468, 442 2 yellow a cis--carotens; - , 442 ee ae aD Se IT HE AT Filtrate | slightly fluorescent in ultraviolet light ~L9~ Each zone was cut out, eluted, and transferred to hexane solution, The pigments obtained from the two bottom zones could not be identified with certainty but were definitely members of the p-carotene set as shown by visual observation of the spectral shift upon addition of iodine to their hexane solutions. The pigment of the top sone could not be identified either but all others were well-known members of the P-carotene set, The composition of the stereolsomerized mixture was determined epectrophotometrically; for this purpose the unidentified pigment adsorbed at the top of the colum was assumed to have the p-carotene extinction coefficient at A The following results were obtained (in % of the max. * starting material): p-caratene, 36%: neo~ P-carotane U, 19%: neo A, 623 neo B, 14%, other cis-P-carotenes, 5%; unidentified pigment, 1%; unaccounted for, 19% | Five minute experiment: A solution of 10 mg. of p-carotene in 25 ml. of hexane contained in a 50 ml, volumetric flask was vigorously shaken with 1 ml. of boron trifluoride etherate for 5 minutes. The liquid was then treated as described for the 1 minute experiment; the resulting hexane solution was developed with hexane on a 20 x 3,8 om, lime-celite colum. (See description of chromatogram on following page.) Except for the appearance of another unidentified pigment zone, the chromatogram was similar to that obtained from the 1 minute experiment, It was noted, however, that the filtrate was more strongly fluorescent. The isomerized mixture showed the following composition (in % of the starting material): P-carotene, 3583 neo- f carotene U, 17%; neo A, 6%3 neo B, 12%3 other gis-P -carotenes, 3%; unidentified pigments, 2%3 -unaceounted for, 25%, ~50- 20 colorless 3 orange _ unidentified; 478, 451 m. 15 interzone 6 pale yellow unidentified; 452 15 interzone - 40 yellow neo-p “carotene U 2 interzone 40 orange all-trans- “carotene 2 interzone 10 pale yellow neo-P ~carotene A 15 orange~yellow neo- 6 ~carotene B 2 interzone 5 pale orange a cis P-carotene 2 pale yellow a cis $-carotene Filtrate fluorescent in ultraviolet light 3. Formation of Some Conversion Products of P-Ceretene. Each of six, 30 mg. samples of P-carotene, dissolved in 60 ml, of hexane in a 125 ml. glass stoppered Erlenmeyer flask was shaken mechanically with 3 ml. of boron trifluoride etherate for 1 hour, At the end of this period the hexane epiphase waa almost deeolorized while the etherate leyer was dark blue, The contents of each flask were added to 200 mL. of 95% methanol and the pigment mixture transferred to the hexane phase by the addition of water. The etx hexane solutions were then combined, washed, dried, and concentrated to 100 ml, This solution was then developed on two, 30 x 8 em, lime-celite columns with hexane. (See description of chromatogram on following page.) The corresponding zones from the two columns were combined, eluted and transferred to hexane. The two filtrates were also combined (cf, Saction b).. 2 reddish~brom 30 interzone 3 pale red LO yellowish-green Pigment I 5 interzone 10 very paleo yollowish-green Pigment TT 35 intersena 45 pale greenish-yellow Pigment IIT R orange 10 interzone 75 pale yellow to pale yellow-orange 60 several very pale blurred zonss Filtratea, yellowish; strongly fluorescent in ultra- violet light a, Troatment of the pigment zones. 1.) Pigment I, After evaporation of its hexane solution, the residue was dissolved in the minimum amount of hot benzene (1-2 ml.) and orystallized in a 5 ml, centrifuge tubs by slow, dropwise addition of methanol at room temperature, Yield, 5 mg. Melting point: 214° (corr,) Solubility: Sparingly soluble in hexane; soluble in benzene but distinctly less so than Prearotens. Insoluble in methanol. Partition test: When partitioned between hexane and 95% methanol the pigment waa found completely in the apiphase. analysis: G, 86.90; H, 10,33. Chromatographic behavior: Moved slowly on lime~celite with hexane; adsorbed wall above P-carotens but below dehydro- “carotene, Grystal form: Pigment I crystallized easily from benzene~methanol in yellow rectangular plates (Fig. 16). MNacroscopically, the crystals appeared dull orange in color. 2.) Pigment II bad maxima in hexane at 444, and 419 m. (visual), After the addition of a trace of iodine the maxima shifted to 446, and 421 mi, Chromatographic resolution of the isomerized solution showed the presence of a pigment that was chromatographically and spectro- scoplcally indistinguishable from Pigment I, Hence, Pigment II was a gis forn of I, 3.) Pigment III initially failed to crystallize fron benzene-methancol, After it had been rechromatographed on alumina, it was possible to crystallize it by rapid addition of methanol to its warm benzene solution. At first, only a cloudiness appeared but crystals wera forued after allowing the liquid to stand at 4° for a few hours, Yield, 5 mg. Melting point: 158-159° (corr. ) Solubllity; Very soluble In hexane and benzene; insoluble in methanol, Partition test: Completely opiphasic in the hexane-95% methanol system, Analysis: ©, 89.333; H, 19.17 Chromatographia behavior: Adsorbed below but not easily separable from A-carotene, Spectrum, in hexane: 468, 441 ma, (visual); 467, 437, 414 ma. (Beckman). Crystal form: The crystals from benzene-rethanol were light yellowish~ orange in color, very small, and of rather irregular shape, They tended to form clumped aggragates (Fig.17). Chromatographic analysis of the iodine-treated hexane solution of Pigment III showed that the pigment had remained mainly unaltered. The Figure 16 (top), Pigment I from benzene-methanol (430 x). Figure 17 (bottom). Piement ITI from benzene-nethanol (430 x). only other zone on the column had maxima in hexane at some 4 mi. lower wave length than those of the starting material (Fig, 18). The location of the spectral maxima of this new pigment (462, 433, 410 m.) was unaltered by treatment with a trace of iodine; only a slight decrease in its extinction values could be observed. 4.) Pigment IV did not form a homogeneous zone on lime- celite but discrete zones could not be obtained, The visually observed maxima (blurred) in hexane were located at 483, 455, and 428 m. In the mixed chromatogram test this pigment was adsorbed well below s ~carotene, b, Treatment of the flucrescent filtrates. The combined chromatographic filtrates (Sec. 3) were concentrated to 50 ml, After washing and drying, the yellowish~orange solution was developed with hexane on a 28 x 5.8 om. alumina colum. Under these conditions a weakly adsorbed colorless zone (detected by a dull blue fluorescence in ultraviolet light) was easily separated from the pigments, all of which were strongly adsorbed on the column. Only one fluorescent zone was observed. The latter was cut out, eluted, transferred to hexane and completely evaporated; all attempts to crystallize the very pale Yellow, oily residue failed. In hexane solution the oll, which had an intense, blue fluorescence in ultraviolet light, showed a single distinct maximum at 228 mi. —55— "QUOFT UE Opsépoyoo euypoy Ay TIT quewdzy may Ay rsceneray . POLLO SOURLRENA JO SUuKey UT GAO UPTLOUPESE ‘ey aleitg OSb 0b os =i (ss a a ee ee ~56= III, SOME NEW CAROTENOIDS FRODUCED BY THE ACTION OF WATER CR METHANOL ON THE BORON TRIFLUOR IDE-DEHYDRO-. p “CAROTENE COMPLEX. A, THEORETICAL PART 1, introduction, When boron trifluoride etherate is added to a hexane solution of dohydro- 6 ~earotene, Cy os 4? a dark blue or greenish=blue color appears in both phases owing to the formation of a complex between the two com- pounds, Complexes are also formed between boron trifluoride and other carotenoids, .e.g., lycopene or frcarotens. Carotenoids, in general, also form blue-colored complexes with some other strongly electrophilic compounds such as antimony or arsenic trichloride (in chloroform solution) and with concentrated sulfuric acid. Furthermore, carotenoid epoxides (5) and carotenoid aldehydes, as well as polyhydroxylated carotenoids, form blue-colored complexes with concentrated HCl, also, The nature of many of these complexes is unknowm; however, the p-carotens complex with antimony trichloride has been discussed by Meunier (24). ‘In general, these complexes are broken by water or methanol, As was shown in Part II, a few minutes treatment of p-carotens solutions with boron trifluoride etherate and subsequently with methanol led chiefly to stereolsoner ization, It was found, however, that a similar brief treatment of dehydro- p -carotene did not yield stereoisomers of the starting material. When methanol was added a complicated mixture was obtained, one third ef which consisted of colorless breakdown products not investigated in this Thesis. The colored components of the end product were found, by chromatographic analysis of the hexane solution, to consist of three main pigments (and their stereoisomers), These were P-carotene (18% of the starting material), a mono~nethaxy~ -earotene C, Hs 50CH, ~57= (40%, half of 1t in the all-trans form and half in ois forms), and a dime thexy- P-carotene, C,olis,(OCH,), (6%). The latter two compounds were previously unknow, When the ‘dehydro- h -earotene-beron trifluoride complex was decomposed with water instead of mthanol, a different mixture was obtained whose hexane solution was resolved chromatographically. The total quantity of pigment components in this instance amounted to about 45% of the starting material, The product contained P-carotene (12%, including stereoisomrs), but the main pigment zone in the chromatogram was a mono-hydroxylated Presretene, CyoHs,0H (33%, about a third in eig forms). The product obtained as mntioned was termed isocryptoxanthin, and those obtainsd from the methanol reaction, isceryptoxanthin methyl ether and dime thoxy- b-carotene, respectively. None of these compounds had been mentioned earlier in the literature. 2. Isoeryptoxanthin (4-hydroxy- f =carotene), Isocryptoxanthin (and its neo A isomer) were adsorbed well above all- Szans~ b-carotene on the lime-celite colum, It crystallized easily from benzene-methanol, Its absorption spectrum was very similar to that of P ~carotens (Fig. 19), which indicated identical chromophores. When partitioned between hexane and 95% methanol isocryptoxanthin was found to be about equally distributed between the two phases. Such distribution is a characteristic feature of mono-hydroxylated carotenoids. Analysis showed the composition, Cy of560- The compound was less soluble in hexane than P ~carotens. As further confirmation of the hydroxylic nature of isocryptoxanthin, a well-crystallized acetyl derivative was prepared, As would be expected, this derivative was completely epiphasic and, in contrast * (@ua00R0« ¢| ~Snipliey) Up umiogdccnosy yo euxey UT eam WTLCUT Qe septoepg *CY ear ug Mu OOS OSt OOr ose wd jow E) »-Ol X ol S| SaaS hae to isocryptoxanthin, could be developed easily on lime~celite with 1 or 2% acetone in hexane, Furthermore, when the acetate was saponified with 15% nethanolie KOH, isocryptoxanthin was recovered, Isceryptoxanthin was thus established to be a nono-hydroxy- fh carotene. One such compound was already well known, nane ly the naturally occurring cryptoxanthin. In order to compare isocryptoxanthin with eryptoxanthin, 30 mg. of a pure crystalline preparation of the latter was isolated from 4.5 Kgs. of ripe persimmons (Diospyrog kaki L.) according to Chatterjee and Zechmeister (1). The two compounds were extraordinarily similar, They melted only 1° apart (isccryptoxanthin, 164°; cryptoxanthin, 165°) and could not be distinguished spectroscopical ly or in the partition test. It was possible, however, to separate a mixture of the two compounds chromatographically on the lime~celite colum, but only after prolonged development, Although two distinct zones were observed, no sharp colorless interzone appeared, for once separated. the two zones moved down the column at vory oimilar rates, In order to determine the adsorption “gequence, unequal amounts of cryptoxanthin and isecryptoxanthin were used in this mixed chromatogram test. When isocryptoxanthin was the major component, the uppermost zone was the larger one. Isooryptoxanthin, therefore, is adsorbed above eryptoxanthin in the system used. | _ The next step was to determine the location of the hydroxyl group in the isocryptoxanthin molecule, Since this compound resembled crypto~ xenthin so closely, it was reasonable to assume that the group was located in a ring. (No carotenoids are mown which have substituent groups in the polyene chain linking the two P-tonone rings or in any of the various ne thyl groups. ) In the P-cerotene molecule there are only three positions avallable in either ring for substitusnt groups, namely, 2, 3 and 4. (see Formula II). H.C GH H.C GH 3\7 3 3 3 V CH, Cl cH, cH, \/ ATS { | Pa: H56 z 6 i ee ! pou, H ie 4 2C-CH H ole 4 abe 2 No 3 | 3 “Oh 2 Formula II. Pr carotene, showing the numbering of the P-donone rings. It had been proposed by Kuhn and Grundmann (20), and generally accepted, that the hydroxyl group in cryptoxanthin occupies position 3. Recently it has been suggested by Goodwin and Taha (6) that cryptexanthin is not 3-, but 4-hydroxy~ p-carctone 3 however, Karrer (11), reviewing the comple te evidence available concerning the location of the hydroxyl group showed conclusively that there was no valid basis for Goodwin's suggestion, As will be shown here isoeryptoxanthin, not cryptoxanthin, has its hydroxyl group located in the 4 position. It would be difficult, indeed, to justify assigning the hydroxyl group in isocryptoxanthin to position 2, for no satisfactory explanation of its formation from dehydro- f ~earotene could then be offered, On the other hand, if isooryptoxanthin were 4-hydroxy- p-carotene ,» its formation conld be accounted for by a Thiele addition of HOH to the ends of the conjugated -61~ system in dehyaro- 6 ~carotene under the conditions described for the boron trifluoride experiment (Scheme I), Although boron trifluoride is an effective catalyst for many reactions (e.g., polymerizations and esterifications) this is the first instance in which the application of this reagent has caused the addition of the elements of water to the ends of a conjugated polyene system. AN oi, CH. CHy Hy CH. BIN Zokg HC ™$=8-Of0-GHAGH- Gls CHSC CHSC O=OH-ORSOR- OSCR CRG ‘ Hy "25 Fie) HaC-Gy of 2 HOH Bgl. oH CHs CH CH, CH HCN ug AN | a 3} (~CH=CH-C=CH-CH=CH-G=CH-CH=CH-CH=C-CH=CH-CH=C-CH=CH-C~ 1? HoC C-CH HiC-C CH No 3 3 ‘of, 2 on Scheme I, Formation of isccryptoxanthin from dehydro- h-carotene, ~62= It was emphasized by Karrer (11) that compounds containing a hydroxyl group adjacent to a conjugated system (i.e., an allylic hydroxyl group) undergo dehydration with extraordinary ease under the influence of acid, with the formation of an additional double bond. He reported (14) that eschscholtzxanthin, Cy ogo (OH) os a pigment detected in the blossoms of Eschscholtzia californica by Strain (31), was dehydrated to a hydro- carbon with 14 double bonds merely by acidifying a chloroform solution with a trace of HOl (Scheme If, below), Karrer also found (12) that A-hydroxy- 6 ~Aonone was dehydrated under similar conditions (Scheme III, below.) In his discussion of the position of the hydroxyl group in eryptoxanthin (11) this author pointed out that since cryptoxanthin did not show this property lts hydroxyl group waa very probably not in the 4 position, isocryptoxanthin, however, did show this oroverty; the dehydration was accomplished by acidifying a chloroform solution of the plement and allowing it to stand for a few minutes at room temperature. The HCl was introducad by adding a drop or two of HCl saturated chloroform to tha solution, A stereoisomeric mixture of dehyaro- B -carotenes was formed in 70% yield together with about 5% of Proxrotene, as determined by chromatographic and spectrophotometric methods, Excapt for some minor breoymish destruction preduete which romained at the top of the colum during developrant, no pigments other than those mentioned were found, "PN "3 ie [? i CH Ns ai 0H=CH=U=GH=UH=CH=U=CH=CH-CH=GH=G=CH-CHSCH-C=CH=CH=G~ [2 RK 7H we-tH HaC-6Qy HON A 3 eschscholtzxanthin 3 So ~~2 HOH Ages, ollg f? fis i? ie Ni T Nogsou-one ~CH=GH-CH=C -CH=GH-CH=CH-C=CH~-CH=CH-C=cH- aa Nex Sehere II, Dehydration of eschscholtzxanthin, le ie ai {3 Hi a Nog-cHech-co — HOH Hg (-CHSCH-CO mmeccteccerann tenis sten se I> I, Scheme III. Dehydration of Achydrcesy= f~tonone This tacid chloroferm test” offers the only rapid means for aifferentiating isocryptoxanthin from eryptexanthin, One or two milli- liters of a dilute isoeryptoxanthin solution (only a few micrograns are : needed) ehange in color from yellow to oranse-red within 30 seconds upon the addition of one drop of chloroform (C.P.) saturated with HCl. The color changa is a result of the formation of a system with 12 conjugated : doubles bonds from ony composeyd of 11 conjugated double bonds. Grypto- xanthin wundergess no visible alteration when subjected to this test, | In addition to the all-trang form of isooryptoxanthin, neo-isocrypto~ xanthin, as well as isocryptoxanthin acetate, also give the same color reaction, £11 the evidence presented is consistent with the proposed 4~hydroxy- S p -carotene structure for isocryptoxanthin, The contention of Goodwin (6) that cryptoxanthin is the jehydroay compound is thus disproved on the basis of new observations. 3, Isoeryptaxanthin Methyl. Ether Dime th b-carotene ee “Be Isceryptexanthin methyl ether, The main pignent isolated from the interaction of the dehydro- P-earotene-boron trifluoride complex with methanol was crystallized from penzene-nethancl . It had an absorption spectrum (Fig. 20) which was nearly indistinguishable from that of Prearatene, Although this new compound was adsorbed above, and easily separated from A-carctene on lime oelite, it was possible to causa it to migrate with pure hexane, This behavior sarved to differantiate it from iscoryptoxanthin, Analysis of the compound showed that it had one oxygen atom. Since it was much more easily developed on lime-celite than isocryptoxanthin it did not seem probable that the new pigment had a free hydroxyl group, This was confirmed by partitioning the compound between hexane and 95% methanol; it showed much more pronameead epiphasice character than would be expected from a hydroxylated carotenoid, The possibility of ita being an epacide of dehydro~ P -carotene was eliminated when it was found that ita ether solution did not color the acid phase whan shaken with concentrated HO]. It had been found by Karrer (10)that all know earotenold epoxides impart a blue coloration to the acid in this test, Finally, a Zeiseal determination showed the presence of one methoxyl group. When a chloroform solution of thea ue thorry- b-carotene was acidified with a trace of acid chloroform, according to the method described above for the dehydration of isoeryptoxanthin, a stereoisomeric mixture of dehydro-6 =~carotenas was formed in 73% yield, together with 9% of P-carctens. Thus 1+ waa concluded that tha mettheryl group was located in 4 position, For confirmation, some isocrypteoxanthin was methylated directly (based on the method described by Karrer (19) for the nethylation of zeaxanthin, a naturally occurring dihydrcory-f -cer otene) * (G10, 0.000— q ncsouzamy) coj.e Téen upramemogdLaosey jo eumtey UP GAO UOT OUYgKe avpnoepae “oe ext Tu OOS OSY OOF OSE | l \ \ Ol wd] jow 3 4 O1X SI ~67- ‘This isocryptoxanthin methyl ether was found to be identical with 4- me thoxy= p “carotene, obtained via the BF complex, both spectroscopically, in melting point (135°), in the mixed chromatogram test, and the "acid chloroform test", In addition, som cryptoxanthin methyl ether (so far not described in the literature) was prepared from naturally occurring eryptoxanthin in order to compare it with seme thoxy- B-carotens Although the two compeunds dould not be differentiated eithar spactrosenpically or in the mixed chromatogram test, eryptoxanthin methyl ether, in contrast to the 4-methoxy compound did not undergo any color change in the "acid chloroform test", In this respect, the two methylated derivatives behaved in an analogous manner to the respective unmethylated compounds, b, Dine theny- f ~earotene Adsorbed above isocryptoxanthin methyl ether, on the chromatogram of the reaction mixture obtained by treating the boron trifluoride~dehydro- P-carotene complex with methanol, was a second new carotenoid showing a P-carotene spectrum (Fig.21). In the mixed chromatogram test the new compound was located below isocryptoxanthin and easily separated from it, Analysis and a Zeisel estimation indicated the presence of two oxygen atoms and the formula, C,oH5, (OCH) 5» Although somewhat more hypophasic than isocryptoxanthin methyl ether, the compound was considerably less hypophasic than isocryptoxanthin when partioned between hexane and 95% methanol, ‘Thus, all the evidence was consistent with the formulation of the new compound as a dim thoxy- ~carotens. The position of the two methoxy groups could not be determined with as much certainty as were the positions of the single substituent groups in {socryptoxanthin and its methyl ether, for in the present case a simple, if unusual, type of Thiele addition of methanol to the conjugated system of 400 450 500 my 350 prraritreprpryr turpis oO wo 0 _ wi] ,-Ol* jow 3 Figere 21, Molecular extinetien curve in hexane of Gira thanr> B “CRT OTANE » ~69- . debyaro- f -carotene would not explain the formation of a compound with two methoxyl groups and 11 conjugated double bonds. A dilute, yellow-colored chloroform solution of Aime thoxy- 6 =~carotene turned rod when a trace of HCl was introduced, Aftor transfer into hexane, chronatographic analysis showed that some of the dime thoxy- fh ~carotene had remained unchanged, whereas, under the same conditions isocryptoxanthin or isocryptoxanthin methyl ether were completely converted to other products, mainly debydro- f-carotenes. Among the conversion products isolated from the chromatogram just mentioned, two could be identified by spectroscopic and mixed chromatogram tests, nanely, dehydro- carotene and © bisdehydro-6 ~carotens. | In addition, a carotenoid which was spectro~ seopically identical to dehydro- 6 -carotene but adsorbed above it was obtained. Because of the small yield it was not possible to characterize this compound further, It is believed, however, that it represents a methoxylated dehydro-~ p ~carotene. The available evidence at the present time concerning the location of the methoxyl groups in dime thaxy- } carotene is inconclusive, and further experiments will be required before lts structure can be clarified. The formation of some bisdehydro~ fh -carotene (a carotenoid with 13 conjugated double bonds; see Schene IV, below) would indicate that these groups are in the 4 and 4' positions, by analogy with the dehydration of eschscholtzxanthin, mentioned above, The reaction product mentioned which appeared above the dehydro- By-carotone gone in the ehromatogram could possibly be interpreted as having resulted from the elimination of the elements of one molecule of methanol (Scheme V, below). I, Ais f ie | fs ik BN, Ay H¢ yo =CH-CH=CH-O=CH-CH=GH-CH=¢-CH=CH-CE= CHO i Ho. _-C=CH HoC-C 2b. f 3 oN OCH. . OCH, — 2 CH,OH p< Hye g-cH=CH-C= CHC cso ns-oorest-osotat ott CH, Hi C-CH H,C-C. 7 ~ or a: Scheme IV. Formation of bisdehydro- [9-carotene from 4,4'-dime thoxy- P-carote 4,41 -dime thoxy- p -carotene ; i< 743 GH GH, CH. He a "2 sl =CH=CH= CH~C=CH-CH=CH-C=CH-CH= o~ Ho batt Scheme V. Formation of 4~methoxy-dehydro- 6 -carotene from 4,4'-dimethaxy~ p-carotens -Ti= Some bicassays have shown that atzetheg-B ~carotene has about 1/6 the provitamin A activity of P-carotene in the rat (see Section 4). On the other hand, zeaxanthin (3,3!=aihyareogf -carotone) and, in general, all carotencids which do not possess at least one unsubstituted Pricncne ring. show no provitamin A effect. This would indicate that both methoxyl groups were in the same ring, were it not for two additional facts: 1. Almost all carotencids, in the all-trans configuration, which do have one unsubstituted p-doncne group display about half the provitamin A activity of prearotene. 2. Exceptional cases in which carotenolds with substituents in both p-Acmone rings show some provitamin A effect are knawm, @.¢., P vearotens diepoxide which according to Karrar, at al (13), has a potency equal to about 1/7 that of Prserotens. 4. Two Proposed Mechanisms for the Formation of Isocryptoxanthin and itg Methyl Ether from Dehydro- fe carotene. In the following discussion two mechanisms, g and b, are proposed, but the possibility of both occurring simultanecusly is: not excluded. | ae) In the ne chani.sm proposed here the water or methanol reacts with the complex of boron trifluoride and dehydro- p -carotene, Step le Boron trifluoride is attracted to one end of the chromophoric system in dehydro- carotene. Additional electrons are drawn from the adjacent carbon atom te this site which has now become momentarily electron deficient because of the proximity of the boron tri- fluoride, Thus a localized dipole is formed. However, bacanse of the extraordinary nobility of the slectrons in a long conjugated polyene system, polarization of the whole molecule takes place: re 73s CH, REN wae Hoo gi seveesereeses COECH-CHEC™” CH FN gortily Hee-G\ Ay Pee F Step 2. The positive end of the molecule is then subject to nucleophilic substitution by water or methanol (R =H or cH,). CH z BPN sc aa Hy HN A + ROH pm 2. =C-CH=CH- i {2 +H H c-Gl CH Sof? sed Nm i) | \ -73- Step 3. Ono of the hydrogen jena, abundantly. present in the acid mediwna because of the hydrolysis (or methanolysis) of boron trifluoride, displaces BF which is removed, primarily, as a complex with ether, water or methanol: C | l G oN “™N H.C C-CHSCH=Gi + « © © 0 6 ew ee oe Se GHRCH-6 CHa Hy S) 7 - OH H30-0 Glig “ep 3 3 “é 4% : cR ROR°BF3 H be) The second possible mechanism proposed is an extension of that generally accepted for the acid catalyzed hydration of the olefinic bond. | | Step 1, Hydrogen ion is made available as the result of complex ion formation between boron trifluoride and hydroxyl or methoxyl ion: | _ : BF; + ROX—> RO:BF, + H Step 2. A hydrogen ion attacks the conjugated system in dehydro~ p -carotens with the resulting formation of a carbonium ion, The ~The positive charga appears at the other end of the systam bacanse of induead electron shifts throughout the chromophore: an ie Fg BSN OH a... . HC “ofcH-cuec-.. 2... . . 20H c=¢~ oH, — \ + Ht ——— Hots | o-cH HeC-Ox fH iF 3 3 off HC. Oy Oly CH, Hg 08g CHa Step 3. The nucleophilic ion, RO:BF3°, then attacks the molecule at the point of winimum electron density, whereby Bs is regenerated; Ho “NG OHECH*C. oo ne Ko e “Goseeee HC. @ -C-CH 4 RO:EF, ee HC. _-C=CH + BF 2 Now 3 3 NG 3 3 cR 5. Provitamin A Activities of Igceryotoxanthin, Isocryptoxanth As is well known, Brcarctene has strong provitamin A activity by virtue of the fact that the animal organism is able to split this . molecule enzymatically at its central double bond and thus produces two molecules of vitamin 4, This reaction is represented achematically, below, By Big CH oi GH, CH, Hs Jig HAC ¢~CH=CH~C=CH-CH=CH-C7CH-CH=CH~CH=C~CH=CH-CH=C-CH=CH-C~ cu, H..0 G=-CH Hat-G OH 2a 3 # “Rof * + 2 8,0 HC. CH, CH ie Ho¢~ promos ceo-cimon-teci- gH a 7O-CHy ICH (Vitamin A) According to somes authors, an oxidative cleavage takes place in the center of the f-carotens molecule and the resulting aldehyde group is reduced in a second step, Evidently, no other carotenoid can show as strong a provitamin A effect as P-carctene. Those carotenoids tested, half of whose molecules are structurally identical with one half of the frcaratene noleecule, also show strong provitamin A activity, but only about 50% of that exhibited by the latter, while if no unsubstituted P-ienone ring is present the compound is inactive. Cryptoxanthin, as would be expected, exhibits OO it Ld streng provitemin A activity (57%); therefore it wae of interest to determine the provitamin A activity of the isomeric isocryptoxanthin, Isceryptoxanthin methyl ether and dime thexy- ~carotene were also tested; these were of special interest since the provitamin A activities. of ne thoxylated P-carotencs have not been reported heretofore, The rat. bleassays, were kindly carried out at the University of Southern California by Drs. H. J. Dauely: drs. and J, Ganguly. It was found that isocryptoxanthin had 48% of the vitamin A potency of Prcarotene, and the methyl ether: 52%. Both results are in complete agresrent with the structures assigned to these compounds in this Thesis. Dine thexy~ fi -carotene also showed provitamin A activity, namely, 16% that of p-carotene, This result can not readily be interpreted at the present tima, for 1f the two methoxyl groups were located in the same ring tho bioactivity should amount to about 50% (p -carotene = 100%), while if each ring had one methoxyl no provitamin A potency should be observed. Although the well crystellized sample: showed chromatographic homogeneity, the possibility that it is a mixture of two (or more) dimethoxy derivatives of the types mentioned can not be excluded, B, EXTERINMENTAL Part 1. Wateriais. The delydro-R-carotene used was of high purity, being chromato- graphically honegene ous and free from colorless impurities as show by microscopic examination of the crystals. Tts preparation from B-carotexe is described in Part I, Boren trifluoride siherate was prepared from the gas avallable in small cylinders from the Ohio Chemical & Manufacturing Co., Cleveland, and Reagent Grade diethyl ether, Except for the use of commercially available boron trifluoride gas, the directions given by Hennion (7) were followed, The etherate darkened on long standing, but this was found to have little or no sffect on its potency. Re & Ne theny= ~carotens and Dine thony-B sgarotene . a. Formation To a solution of 60 mg. of all-trans-dehydro- [} -carctene in 250 ml, of hexane, contained in a 750 ml, Erlenmeyer flask, was added 5 ml. of boron trifluoride etherate. The flask was vigorously swirled by hand for 60 seconds (the etherate is not appreciably soluble in hexane ) and then 300 ml. of 95% methanol was added; the dark greenish~blue color of the carotenoid-boron trifluoride complex immediately changed to dark red, The liquid was transferred to a separatory funnel and water was added to trensfer all pigments to the upper phase. The aqueous-rethanol hypophase was discarded and the epiphase was washed in a continuous apparatus, ‘After drying with anhydrous sodium sulfate, the hexane solution was adsorbed at the top of a 30 x 8 om. lime-celite column and developed with 2©3% acetone. in hexane. «Qa Description ef the chromatogram (the figuros designate thickness in mm, )% 1 yellow i purple 10 orange 3 purple 10 interzore § yale pink 20 interzone 20 orange Zone I (divethary- carters ) 15 interzene 25 yellow zone II 5 initerzone 60 orange Zone TIT (ome thcay-B -carctens) 2 interzons 25 yellow Zone IV 2 interzone 15 orange Zone V ( b-carotene ) 2. interzone 10 pale yellow Zone VI 10 interzons 10 pale yellow Zone VII Bach of tke zones I-VII was cut out and eluted with acetone. After transferring to hexane, each fraction was rechromatographed on a smaller column, Repeated rechromatography of fraction III was sometines found to be necessary before it crystallized easily. No attempts were made to orystallize fractions IT, IV, Vi, and VII as these were found to be stereoisomers of the others (cf. section ¢,, belcw). Fractions I, ITI, and V were crystallized according to the procedure deserited below for Jone thory-[}-carotene. The hexane solution (about 50 mi.) was evaporated to dryness in vacuo. (The temperature of the bath should not exceed 45° since excessive stereoigomerization of carotenoids would increasa erystallization difficulties.) The dry, powdery residue was dissolved in 2-3 ml, of cold benzene and transferred toa 5 ml, centrifuge tube .* The tube was placed in a bath of water kept at about 40° and the solvent evaporated by directing, through a capillary, a gentle stream of pure dry nitrogen at the surface, Tho residue was then dissolved in 10 drops of benzene;. and methanol was slowly added from a dropper by allowing it to run dom the side of the tube, Thus, the methanol was warmed approximately to the bath temperature: before reaching the benzene solution, During these operations the solution was stirred constantly with a thin pointed glass red, In all, 4 to 5 volumes ef methanol were added, and, in favorable cases, glittering red crystals appeared before this addition was complete. The tube was then allewed to stand at room temperature for 30 minutes, (Stronger cooling was not found to be advisable, as at times, a colorless or pale yellow oryetalline material precipitated which was difficult to remove by recrystallization.) The red crystals were centrifuged down and the mother liquor decanted, The pigment was washed twice, in the centrifuge tube, with methanol at room temperature, and then dried in an Abderhalden apparatus, in the presence of P20,, in good vacuum (0.1 mm. ), at 55° (refluxing acetone), for 1 hour. Yield: 8-9 mg, of 4-methaxy= B-cerctene, 2-3 me. of dimethoxy~B -earotene, and 5-6 mg. of B-carotene, in all, 27-28% of the starting material, The mother liquors were found to contain 1-3 mg. of additional pigments. * If, after the hexane has been evaporated, the residue is of oily or partially oily nature, difficulties in crystallization will be encountered later. In such a case the residue should be diasclved in hexane and rechromatographed on lime-celite, -S1- b. Characterization 1.) Dime thaxy~ B -carotene The crystals from benzene-methanol formed very small irregular plates, Undar the microscope (magnification 430 x), single crystals appeared light orange, while clumpsd groups were dark red, Macroscopically,. the dried crystals were dark red, Melting points 152° (Berl block: evac. capillary; corr. ) Solubility: Soluble in hexane, more easily in benzene or chloro~ form, Almost insoluble in methanol, Partition test: When partitioned between hexane and 95% methanol, the compound was mainly epiphasic, However it was more hypophasic than P-carctens. Analysig: Calculated for CoHs,(OCH,)2: C, 84.503 H, 10.13. Found: C, 84.333 H, 10.02. C, 84.433 H, 10.84. (Second result corrected for 1,5% ash). Methoxyl determination (Zeisel): Calculated: 10.37%, Found: 9.45%, Spectrum: Maxima in hexane: 484, 453 mu. (visual); 479, 450-451 mu. (Beckman). | Molecular extinction coefficient: gnole = 13.4 x 104 atd mar. Chroma tographic behavior: Easily developed on lime-celite with 23% acetone in hexane, Adsorbed well above B-carotene in the mixed chromato~ gram test, 2.) 4-Mothaxy- f -carotene This compound crystallized from benzene-msthanol in oval shaped clusters (Fig.22). Macroscopieally, the dry crystals appeared dark red. Melting point: 133-135° Solubility: Soluble in hexane; very soluble in benzene or chloro- form; almost insoluble in methanol, Fartiiijon fast: The compmmd showed an essentially, but not fully epliphasic behavior, In this raspect it stands between Prearotene and dimo thoxy~f ~carotene, | Analysis: Galeulated for Gjci,,0CHg: G, 86.863 H, 10,31. Found: G, 86.03, 86.113 H, 10.60, 10,09. ethoxy 1 determination: Caleulated: 5.46%, Found: 5,054 Spectrum: Maxima in hexane, 484, 453 m. (visual); 479-480, 451 m. (Beekman). | Molecular extinction goofficients B[C™* = 13.5 x 104 at A max, Chromatographie behavior: On lime-eelite the zone of this compound appears between those of dine thaxy- p ~earotene and P-earotens. 3.) P-carotens, The compound erystallized very easily from benzere-methanol in red — hexagonal plates characteristic of Prcarotens. The crystals were dark red in color, | Solubilitye Soluble in hexane, but somwhat less so than the two methoxy derivatives described above. Partition: Only a trace of hypophasic behavior was observed, Identical in thia respect with authentie B-carctons. sis: Calculated for ©, oflsg! C, 89.483 H, 10,37. Founds C, 89.35; H, 10,37. Spectrums Tho absorption spectrum in hexane was identical with that of an authentic sample of Brearotens. #83= Chromatographic behavior: Not separable from authentic prosretene in tha mixed chromatogram test, c, Identification of Fractions II, IV, VI, and Vii. | 1.). Fraction II. . This fraction was saparabla from dena theory B -earotens only with. some difficulty, It had maxima in hexane at 473, 449 mp. (Beckman), and showed a small cis-peak at 340 mi, These values corresponded closely with those observed for neo~ fh -carotens U (28). Upon isomerizing this fraction with iodine, all-tgane- heme thoary-p -carotere was formed and isolated chromatographically, It was identified by specroscopic readings and in the mixed chromatogram test, Consequently, Fraction II was termed neo-h~ne thary-P ~carotene-U. 2.) Fraction IV. This fraction showed maxima at 470 and 445 mi. (Beckman, in hexane) and a strong gig-peak at 338 mi. Upon iodine catalysis the main maximum shifted to 446-447 mi. and the height of the cig-peak decreased. From a chromatogram of the isomrized mixture 4-nethaxy- p ~carotene and P-carotene were isolated (ratio 2:1), Fraction IV was thus identified asa nixture of noo~h-notheay~ p ~carotene and neo- f ~arotene Ue -3.) Fractions VI and VII. Tests similar to those described above showed these fractions to be neo isomers of Prearctens. | 3. Isecryptexanthin a, Formation. To a solution of 35 mg. of all-trang-dehydro- P-carotene in 100 ul, of hexane, contained in a 300 ml, Erlenmeyer flask, was added 3 ml, of boren trifluoride etharate, The mixture was vigorously swirled by hand for 60 seconds and then transferred to a l-liter separatory funnel... containing 400 ml,. of aca tone . Water was added slowly until two phaseg separated and the upper (hexane) phase contained nearly the total pigment, The aque ous~ace tone hypophase Was. discardad, and the. hexane layer washed in the continuous washing apparatus until the.yellowish-brom color changed to a clear dark red, This usually took about 15 mimtes, After drying, ths hexane. solution was. filtered from a reddish-brom, gumny material which precipitated during the washing process, The filtrate (about 100 ml.) was developed on a 27 x 5.8 om, lime-celite colum with 6-8% acetone in hexane. brown bromish=red interzone purple interzons yellow interzona purple interzone yellow interzons reddish-orange Zone I (isecryptoxanthin) interzone yellow. Zone II (neo-isocryptoxanthin) interzons very pale yellow interzone very pale yellow interzone very pale yellow 35 interzone 15 yellow Zone III ( B-carotens ) avatathbrwr ds Sbeurwrurh be Zones I-III were cut, eluted, and, after transfer to hexane, each was rechromatographed on lime-celite, Fractlons I and II required 8% acatone for development, and fraction III, pure hexane, after ory~ gstallization of I and III from benzene-methanol (according to the method dascribted above for dmne'tharry- Pp -carotene ), the gamples wera dried in tha Atderhalden apparatus at 80° (refluwsing benzene) for an hour, Yield, 5 mg. of isceryptoxanthin and 2 mg. of Prcarotene. Fhotemetris reasurea= manta showed en additional 9 me, in tha mether Liquers and seqla lsomera {not crystallized), b, Gharacterization of Isocryptexanthin and p-Garotens 1.) Isoeryptoxanthin The crystals obtained from benzere-methanol are shown in Fig, 23. Under the microsnope (430 x), singles crystals appeared orange-brown, but where they crossed a light to dark red color was observed depending on the thickness of the conglomerate, The general shape was that of oval plates with blunted ends, Macroscopically, the somewhat glittery orystals appeared dark red, Welting point: 162-164° Solubility: Rather sparingly soluble in hexane; more easily in benzene or chloroform, Slightly soluble in hot methanol. Partition test: Strongly, but not completely, hypophasic. OH G, 86.973 Hy 10.22. Pound: G, Analysis: Calculated for Gy ots 5€ 86.97, 86.523 H, 10.54, 10.24. Spectrum: Maxima in hexane, 484, 453 m. (visual); 479, 451 m. (Beckman), Molecular extinction coefficient: B™l+ = 13.5 x 104 at ns icon, Xe Chromatographic behavior: Strongly adsorbed on lime-celite. = Figure 22 (above): Isoeryptoxanthin methyl ether from benzene-methanol (430 x). Figure 23 (below): Isooryptoxanthin from benzene-methanol (430 x). Separated with difficulty from eryptoxanthin in the mixed chromatogram test; adsorbed above the latter, Frolonged developmant with 8% acetone in hexane wags required, 2.) p-carotens. tha crystals from benzena-methanol were cleaniy formed hexagonal plates. Tha compound was in all respects identical with authentic P-carotene. Analysis: Calculated for Crolsg 3 C6, 89,483 H, 10,52. Found: 0, 89,323 H, 10,76, o, Identification of Fraction TI. it is claimed that this fraction was a nao isomar of laocryptoxanthin for the following reasons: The maxima Im hexane were located at 477 and 446 mp. (visual), i.e., at about 6 mi. shorter wave lengths than the corresponding maxima of ali-trans-isoceryptoxanthin, Uponisomerizing this fraction with a trace of iodine (in light), the main isomr which was formed, and separated chromatographically, was found to te all-trang-iso~ eryptoxanthin as showm by spectral observation and in the mixed chromato~ gram test. d. Acetate and Methyl Ether of Isocryptoxanthin. 1.) Isoeryptoxanthin acetats. The method used was based on that of Kuhn and Grundmann (20) for the preparation of aryptoxanthin acetate, with some minor modifications, Thirty milligrams of isocryptaxanthin were dissolved in 2,5 ml. of dry pyridine (c.p.), in a 15 ml, centrifuge tube. One ml, of acetic anhydride (c.p.) was added and the mixture heated at 70° for 45 minutes, The solution was then transferred to a small separatory funnel, diluted with 50 nl, of hexens, and washed free of pyridine and acetic anhydrides (about 10 nin, washing tise was required), The dried hexane gclution waa developed on a 27 x 5,8 on, line-celite column with 2% acetone in hexane, The acetyl derivative forwed en corange-brom: aone which moved down the colmm at a moderate rate together wlth some pale orange and yellew miner sterecisoners, In contrast, unreacted isceryptuxanthin remained rear the top during development, The acetele zone was cut cut, eluted, and trans- ferred with water te hexane, The dried hexane seluticn was completely evaporated, The reaidue was crystallized in a emall centrifuge tube fron penzene-rethanol (134) and dried for 1 heur at 55° in the Abderhalden apparatus, Yield: 4 mg, Fartition test: Entirely epiphasic. Chromatographic behavior: Easily developed on line-celite with 23 acetone in hexane, Adscorbed above p-carotene but well below isocrypto~ xanthin, Spectrum: Identical to that of isocryptexanthin, (Maxima at 479 and 451 mi. in Beckman. ) Seponification: About half of the sample underwent saponification when 10 ml, of a dilute hexane solution was shaken with 10 ml, of 15% methanolic KOH for 10 min, at room tempersture, This estimate was made photometrically after chromatographic resclution of the reaction product. 2.) Isccryptexanthin methyl ether. The method used was based cn the reported by Kerrer and Takahashi (19) for the methylation of zeaxanthin, with a number cf modifications in the procedure. To 8.5 ng, of iseeryptexanthin contained in a 5 ml. cantrifuge tute was added 0.5 ml, of tert.amyl alcohol (dried ty distilling from metallic potassium). the mixture was heated to 90" and toluere (dried over Pete) was added dropyise, with constant stirring, vntil ali the isocryptacanthin dissolved, Two pieces of potassium retel, each about the size of a grain of riea, wave intreduced and stirred about until disselved, The soluticn waa then allowed to cool to 35°, and 3 ml, of methyl iodide (free from iodine} were added, ‘The corked centrifuge tube was kept at 35° for 30 mizmtes, The liquid was transferred to 4 mall seraratory funnel and diluted with 25 ml. of hexans, After washing and drying, the hexane sclution was developed on a 24 x 4.8 om, lime-celite colwm with 2% acetcme in hexane, The isoeryptexanthin vethyl ether appeared as a yalilon- brown zone which moved dow the colurm quite readily, The major uone which was unreacted isccryptoxanthin, remained near the top, No stereo- isomers of either cf theae two oubstanoes were ebserved, presumably be- cause the methylation had been carried out in a basic medium, Yield, about 0.5 mg. Melting points 133-137° Partition test: Slightly hypophasic. Spectrum: In hexane, 479, 450-451 mi. (Beckman). Chromatographic behavicg: The compound did not separate from an authentic sample of dene thoxy~ B carotene or fron cryptoxanthin methyl ether in the mixed chromatogram test on lime-celite, To a solution of 2-4 mg, of the compound (isoeryptoxanthin, 4-rethary- p-carctene or dine thaxy- 6 -carotens ) in 10 ml. of chloroform (c.p.), 4 drops of chloroform saturated with HCl were added. The solution was then kept at room temperature for 25 minutes, and shaken with 10 ml. of a 5% NaHCO; solution, After drying over sodium sulfate, the chloroform was evaporated, the dry residue dissolved in 10-20 ml. of hexane and developed on a 20 x 3.8 om, lire-celite colurm with 6% acetone in hexane. Description of the chromatogram obtained from isocrypteoxanthin as starting material: 1 brewnish-yellow 25 intergcne oo 15 pale red Zone I 3 interzone *? lew Fore a cut out together 8 interzone 20 orange-red Zone III (contains two stereoisorers) 6 interazone § orange-yellow Zone IV 60 interzone (inel. a few very pale zones) § yellow gone V The chromatogran obtained fron 4-nethory- B “carotene in the acid chloroform reaction was qualitatively identical to that described above, there being an exact correspondence zone for zone, Zone I was all-trans- dehydro~ B ~carotene, and Zones II-IV were stereoisomers of this compound. Zone V was Proarotene (including stereoisomers), The identity of the respective fractions was established by spectra and mixed chromatogram tests, Estimates of the amount of substance in the various fractions wera made photometrically (Tables 9 and 10). Table 9. Conversion Products Obtained frem Isceryptexanthin by Treatwent with Acid Chloroform (% of Starting Material) 41l-trang-dehydro- 8 -carotene 16 Neo A +B isomers 30 Neo C + D isomers 18 Other isoners 5 A-Carotene (including isomerg) 5 Unaccounted for 26 Table 10, Conversion Preducts Obtained fron 4-Vethary-8 -carotene by Treatment with Acid Chloroform (@ of Starting Material) All-trang-dehydro~ 8 -carotene 19 Neo A and neo B isomers 30 Neo G and neo D iscrers 19 Other steric forms 6 A-Garotene (including isomers) 4 Unancounted for 22 The interaction of dine thexy-8-carotene with acid chlcroform was nore complex as shown by the following chromatogram: aad a RORON CWUWU OWWNWWH a) up yellow-brown inter zone pale red 5 6g ow te tt we tt tt Zone pale red 6 6 5 ee ee ee ee ee ee ew) =6BONE inter zone pale orange-red . . 6 51 6 ee ee ee ee ) 6 6BOMO pale red with yellow leading edgo ....» Zone inter zones very pale yellow interzone very pale pink inter zone red-<ourple ..... Fr Zone inter zone several very pale zones interzone yellow interzone orange-brewm with pink trailing edge - +--+ + Zone pale yellow Ii III VI All nmunhered zones were cut out, eluted and transferred to hexane, The mixture of Fractiona IT and II was rechromatographed to obtain a more satisfactory separation, following spectral maxina were observed (visual, in hexane): Fraction VY was also rechromatographed, The Fraction I 503, 473 mp. unidentified; possibly a methoany- dehydro~ 8 -carotene Fraction IJ 498, 469 m. unidentified; possibly a neo isomer of above Fraction ITI 504, 474 m. all-trang-dehydro- -carotene Fraction TV 497, 467 m. mixture of neo-dshydro-S~carotenes Fraction V 519, 487, 460 mp. bisdehydro-S-carotene Fraction VI 483, 453 m. unchanged dinethoxy-8 -carotene Fraction V did not separate in the mixed chromatogram test from an authentic sample of bisdehydro-@-carctene (a carotenoid with 13 con- jJugated double bends, cf. Part IV), prepared from S-carotene by dehydro- genation with N-bromosuccininids, No quantitative estimate of the conversicn products could be made because of the complexity of the product and small amounts of its compuoents, It was evident, however, that the conversion was less extensive, if more complex, than in the case of iscoryptoxanthin or its methyl ether. >» froparation of Isocryptosenthin, Isocryptoxanthin Nethyl Ether, and Dimethacy~@~-carotene from a Stersoisorerio Mixture of Dehydro-48 -carotenes, In the experiments previously descrited, a pure crystalline all-trans- dehydre-4~-carotene was employed as starting material for the preparation of isccryptoxanthin, isocryptoxanthin methyl eather, and dinethoxy- A-carctene. It was found that a stereoisoseric mixture of dehydro- Ancarotenes could also be used for the preparation of these compounds. Although its use resulted in chromatograms of increased complexity, the over all yields, based on S-carotene, were better owing to the greatly increased amount of starting material made available, {As was shown in Part I, the all-trans form constituted but a relatively small fractim of the sterecisomeric dehydro-S-carotenes produced from Axncarotene iodide hy the action of acetone.) Preparation and treatment of a stereoisomeric mixture of dehydro- #-carotenas, The three principal steric forms of dehydre-A-carotene (i.e., all- trang, neo A, and neo D) produced from S-carotene as descrited in Part I were separated from extranecus products by adsorption on lime-celite. About 400 mg, of these pigments were ultimately obtained in hexane solution ~9L~ (photometrdo estinate) and, after evaporation to dryness, 150-200 ng. eculd be crystallized from benzene-methanol, These crystals were treated ao follows; To 30-40 mg. of the crystalline mixture dissolved in 100 ml. of hexane were added 5 ml, of boron trifluoride etherate, The flack was vigorously swirled for 2 mimites and then treated with water or methanol to form isocryptexanthin, or its methyl ether and dinethaxy-8-carotene, exactly as prescribed for the pure all-trans compound, The mother liquor remaining from the crystallization of the dehydre- 8 - carotene mixture was also utilized for the preparation of the three new substituted S-carotenes in the following manners After dilution to 100 ml. with hexane, the mother liquor was shaken with four, 100 ml. portions of 95% methanol, discarding the methanol phase after each washing, The hexane phase was then washed with water, dried, and diluted to 500 ml. with additional sclvent. Each 100 m1. portion of this sclution was treated with 5 ml. of boron trifluoride etherate in the same manner as described above for the crystalline starting material. Iv, ON THE TITERAGTION oF A-CAROTUNE AND N-BROMOSUCCIN IMIDE A. THEORETICAL PART. 1. dJntredustion and Generel Renarks., The formation of dehydro-lycopena, Oz gtsas an entirely aliphatic carotenold with 15 aonjugated double bonds, by the action of N-bromo- succinimide on lycopene, CyoHs¢, was observed by Karrer and Rutschnann (15) eight years ago. Up to the present time this saems to be the only instance reported in which a C10" carotenoid has been dehydrogenated by this method. The corresponding reaction of S-carotene (only partially aliphatic) with N-bromosuccinimide described in this Thesis was carried out by refluxing in carbon tetrachloride, The resulting mixture was resolved by means of chromatographic analysis and the individual pigments estimated spectrophotonetrically, A typical experiment of this type led to a complicated mixture, the composition of which appears jin Table 11. About one half of each of the three dehydrogenated B-carotenes was isolated in pure, crystalline all-trans form, showing the following characteristics: Dehydro~@-carotsna, mp. 193°, maxima in hexane (Beckman): 502, 471, 445 mp. Bisdehydro-@~carotene, mp, 204°, maxima: 519, 487, 461, Anhydro-eschacholtzxanthin, n.p, 193°, maxima: 531, 498, 472. The ease with which Karrer and Rutschmann dehydroganated lycopene compared with the difficulty involved in dehydrogenating A-carotene may te attributed to the following factors: a) The bromination of lycopena Table 11. Gonposition of the Reaction Product of #-Carotene and N-Bromosuccinimide (the figures indicate % of starting material) Unchanged #-carotene 25 Unidentified products (pigmentea showing extraordinarily strong adsorption affinities and colorless, non-fluorescent compounds; both types probably including halogenated substances) 60 Dehydro- 6 -carotene 9 Bisdehydro- 8 -carotana Z Anhvdre~eschscholtgxanthin 2 Cian MO Os kk ae RE MSM mS probably occurs at the carbon atoms within the chain which are next adjacent to the isolated double bonds; it is well knovm that allylic broninations by N-bromosuccinimide are difficult when the double bonds are part of a conjugated system. »b) The necessary dehydrobromination is facilitated by the presence of double bonds on both sides of each of the brominated oarbon atoms which then fall into conjugation with tha newly forred double bonds (Scheme VI). c) Since dehydro-lycopene is insoluble in carbon tetrachloride it is continuously removed fron the reaction mixture as it formed, Nona of these factors is oparative in the casa of the A-carotene dehydrogenation, The use of N=-phenylmorpholine to facilite the elimination of HBr has been desoribed by Farrar at al. (3) for the dehydrogenation of Vitamin A acid methyl ester (Scheme VII.). A tentative applisation of this method to A-carotene actually decreased the yield somewhat; at the -97- HAC. a Che GB ~ CH, ot he Gy HC % lycopene | (N-bromosucceinimide ) "e Notccitar=cHly- a aan a HECH-P=CH-CH H 3 |- HBr SE eC eC GHEGH- PGE ORS CH PE CH-GH Hg H H rN Hg a 4 dehydro-lycopene Scheme VI. present tine no explanation of this phenomenon can be offered since, in principle, the two reactions should be very similar (Scheme VIJI). Since N-bromosuccininide is thought to react via a free radical mechanism, many workers have used light or peraxides to facilitate free radical formation. However, neither illumination of the refluxing reaction mixture nor the use of benzoyl percuide affected the yields in the Avoarotene dehydrogenation, It was found that the addition of NaHCog or Bal, did increase the yieids slightly, A refluxing period of 6 hours and 2:lmole ratio of N-bromosuccininide: A-carotene gave the best yields, Hg AH GH, GH, 7 | HG Ng~GH=CH~G=CH-CH=CH- =CH-COOGH Ho. 7 e-tk, Vitamin A acid methyl ester of, | (N-bromosuccinimide ) HQ” Q-GHECH-O=CH-CHECH-C=CH-COOCH, HN 7 O-Gy ; Br (N-phenylmorpholine ) ( - HBr) tig Als Hi, fie a » prosson-teai-oeaH-OraH-coodE H oN a Scheme VII Hg Aly fig f’ HQ” G-GHSCH-O=0H-CH=CH-C=CH-CH= Hoty _C-6H + B-carotene CH, 3 (N-bromosucoinimide ) | (= HBr) Hg Ally Hy Hy #307 > fronmni-GecH- one os-Se0H-OR= q 0-03 i ale Scheme VIII 2. Structure of Bisdehydro- -carctens. Bisdehydro-@-carctene is thought to have the structure represented by either Formula III or IV on the basis of the following considerations: a) Analysis showed that the compound is a hydrocarbon, b) Catalytic hydrogenation proved the presence of 13 double bonds, c) The location of the spectre] maxima in hexane, viz. 519, 487, 461 mp. (Fig. 24), indicates that the double bonds are conjugated, Dehydre=A-carctene heaving 12 conjugated double bends hes its maxima Located about 17 mi, lower than the corresponding maxima of the bisdehydro compound, d) Bisdehydre-8-carotene is leas strongly adsorbed than lycopene (which has only 11 conjugated double bonds tut a completely aliphatic structure) thus indicating that neither of the ring systems of the A-carotene molecule had been opened during the conversion, At the present tine it is not possible to decide between the tive proposed structures (Formulas Illand IV). The marked similarity in the shapes of the bisdehydro= and dehydro-8~carotens extinction curves and the absence of distinctive ois peaks in the corresponding curves of their neo isomers is most plausibly explained on the basis of the unsymmetrical structure (Formula IV ), since then the arrangement of the double bonds in the two compounds would be identical, If bisdehydro- Sncarotene had the symmetrical structure represented by Formula IIT, it would bear the same relationship to vitamin A, (3) as A-carotene doas to vitamin A, (of, Schane Ii), Bisdehydro-A-carotene could then be expacted to show considerable growth promoting properties in the rat, m - pmol y io lcm —100—— Figure /. uolecclar extinetion curve in hexene of all 350 ~ 400 450 900 *Caratenes sew, fragh solutions cee, after refluxing for Mn Hew, woweme, efter dedino catalreales in Light, x|0-4 mol lem E _ Uelecular axtinetion curves in hemane ef sams gig biedehydve~ P ~cerotence: me, THO AZ wee, NBO Bz wrmenr, MAO, —f02— | \ J _ | | | H | Ke) | © LJ Figero £6, Extinction curves in hose of anhyiro-ecchac oligganthins women, alivtrangy vom, neo ABs seren, hoe Bis 400 450 500 550 Mu 350 ~103~ aie a ie bi ie) M3 BSN 8 07 No-CHEOH-C=OH-CHECH-b=CH-CHLCH-CHb-CHEOH-cHSb-oRSoMO” Ng 2 “ CH H | C-CH HaC-C E f° Forma III RF Loe f° | B (" i Hye 0H H G~ C=CH-CH=C-CH=CH-CH=G~CH=CH-CH=CH-C=CH-CH=CE- =O8-CF ai H CO 0-CH | HaC-Cy CH ‘N< 3 Formila IV ; ~~ eS ls cH, ie “sf G-CH=CH-C=CH-CH=CH-C=CH-CH5H H C-CH Nh? Vitamin Ay i-. 73 3 i? i ‘“ i es ar H. I ce 10 3 % A-Carotene By Als é ie He Dei ~CH-CHECH-C=CH-CH0H H “CH, NC “Vitamin Ay By Jig CH, H #07 Danneel wae Hy /C-CH Noff 3 % Bisdehydro~ 8-caroctene (see text) Scheme IX -104- Should this be the case, then its formulation as a symmetrical molecule would be established, A negative result, however, would not te indicative of any definite structure, Bioassays of this nature are now being carried out at the University of Southern California and will be reported later, 3. Anhydro~esohscholtzxanthin. By the dehydration of a naturally occurring hydroxylated carotenoid, via, eschscholtzxanthin, Karrer and Leumann (14) obtained a compound, CroHsq. termed "anhydro-eschscholtaxanthin" Formula V). The method of dehydration used by these authors was mentioned in Part III of this Thesia in connection with the dehydration of isocryptoxanthin, HG. ae) 13 12 [2 ie HCY CH, — ‘“ “=CH-CH=C-GH-CH~CH=C-CH=CH-CH=CH- en | it HG C-0H HaG= CH RF 3 NOS 4 Formula V Anhydr o-a schscholtzxanthin It is believed that one of the products obtained from S-carotene by the action of N-bromosuccinimide is identical with Karrer and leumann's anhydro~sschscholtzxanthin on the basis of the comparative data presented in Table 12, 4s can be seen from Formula V, only a single arrangement of the 14 double bonds is possible unless the N-bremosuceinimide reaction had opensd one or both of the rings in the P-carotene molecule, However this -105- Table 12, Comparison of Som Characteristics of Anhydro- eschscholtzxanthin Preparationg Sample of Karrer Sample obtained in the and Leumann present study Spectral maxima in hexane (visual) 531, 500 m, 534, 499 m. Melting point 196° 193° Analysis (Caled. ao" H, 9.50) C, 89.69; H, 9.63 G, 89.89; H, 10,10 possibility may be excluded on the basis of the markedly weaker adsorption affinity of the compound in comparison with lycopene, -106~ B, EXFERTMENTAL PaRT i, Haterials, The N-bromosuccinimide was obtained from Arapahoe Chemicals, Ina,, Boulder, Colo, Reorystallization of this product from water offered no advantages. 2. Preparation of Bisdehydro-f -carotene and Anhydro= sschsgholtzxanthin, To a solution of 200 mg. of S-carotene in 60 ml, of carbon tetra- chloride were added 132 mg. of N-bromosuccinimide and 300 mg. of sodium bicarbonate, This liquid, contained in a 100 ml. round bottom flask, was refluxed for 6 hours using a 200 watt incandescent lamp as a source of heat, after refluxing, the solution, then somewhat darker than before the reaction, was cooled to room temperature under the tap and filtered. The filtrate was concentrated, in yacuo, to ahowt 20 nl., diluted with 80 nl, of hexane and developed on a 30 x 8 om, lire~celite aolum with 3% aseetone in hexane: 15 browmish-red 10 interzone 75 several red and orange zones Section I 40 Interzone 45° yellewish-orange Section IT Sectioens I and II were cut out separately, eluted and transferred to hexane solution, Section II consisted of unreacted S-carctene of which, according to a spectrophotoretrica estimation, 52 mg, were re- covered, The hexane solution of Section I was evaporated to dryneas, the crystalline resicue transferred with the minimum amount of hot benzene, to 9 15 wl. centrifuge tube and erystallized from the hot soluticn by drop- -107- wise addition of methanol with constant stirring, Yield, 28 lige These crystals were dissolved in 150 nl, of hexane and developed on a 30 x 8 on, lime~celite colum with benzene-hexane (1:3). The following rather blurred chromatogran was obtained (figures to the right indicate the visually observed spectral maxima in hexane): or 3 brownish-red 20 interzoure 1G pale red 15 interacne 25° orange~red zone I 5h, ATL Tp. 50 reddish<purple Zone IT 522, 490 25 purple Zone IIT 534, 499 15 orange gone IV 500, 469 10 orange-red zone V 518, 489 5 pale orange Zone VI 494, 461 3 interzcns 10 pale orange Zone VII 515, 284 5 yellowish-crange Zone VIUT 491, 460 5 intersone 10 pale orange Zone IX 510, 481 Bach numbered zone was cut cut, eluted and transferred to hexane The pigments contained in the various fractions are listed below, Fracticn I ali~trans-dehydro- A-cerotene Fraction II all-trans-bisdehydro- A-carotene Fraction ITI anhyarc-eschscholtzxanthin Fractions IV, VI and VIII is isomers of dehydro~ 4-caroitene Fractiona V, VII and Ix gig isomers of bisdehydro~A -carotene The dehydro~M-carotenes, which constituted the bulk of the dehydro- genated pigments formed, were well known on the basis of observations reported in Part T of this Thesis. The eryatellizatien and identification of bisdehydro~-carctene and anhydro-eschscholtzxanthin are desoribed below in Sections a, and b. ~108~ a. Bisdehydao~ A-carotene. The benzene-hexane solution of this compound (Fraction II) was completely evaporated in vacuo and the reddish-purple crystalline residue transferred to a 3 ml. centrifuge tube ty means of 1+2 nl, of warm benzene, The tube was then kept in a bath at about 40° and the henzene evaporated by directing a gentle stream of nitrogen at ite surface until about 4 ml, of solution remained, ‘The carotenoid wag then crystellized from the warm solution by dropwise addition of methanol down the sides of the centrifuge tube. After 32/4 volurea of methanol had been added the tuse was removed from the bath and kept at rocn temperature for an hour, The crystals were centrifuged and, after decanting the mother liquor, washed with methanol, centrifuged again and dried in yacuo (0.2 mm,) over P25 at the temperature of refluxing acetone for one hour, Yield, 2-2.5 mg, A totel of 35 mg. of bisdehydreo- Axcarotens vas isolated fron nurerous similar experinenits, Gharesoteri zation. Crvystel form: Crystallized from benzene-methanol in oblong, pale purple plates (Fig, 27). Macroscopically, the crystals appeared dark purple. Meiting point: after sintering at 190° the crystals melted sharply at 204°, Solubility: Sparingly soluble in hexane, more easily in benzene, Tnsolutile in prethancl, Partition test: When partitioned between hexane and 95% methanol the compound was found entirely in the epiphase, Analysis: Calowlated for Co oHss' C, 90.16; H, 9.84. Founds 6, 89.40; H, 10.37, Spectral maxima, in hexane: 522, 490 m (visual); 519, 487, 461 my (Backman). Molecular extinotion ecefficient at » may? TRCLs = 16,6 x 104 (in hexane), Hvdregonption: 5.269 mg. of substance (with 11.53 mg. of Pt0, catalyst) added 2.76 ml, of hydrogen (0°, 760 m.). Calculated for GC) Hso! 13 double bends, Found: 12.5 dcuble bonds, As a control, a parallel hydrogenation was rum with 5,056 mg. of S-carotene and 11,09 mg. of the catalyst. Calculated for Cr Hag? ij double bonds, Found: 10,6 deuble bonds. GShromatographic behavior: Adsorbed below dehydre=- S-carotene and above G-carotene on line-celite, when developed with 3% acatone in hexane cor a 133 benzene-hexane mixture, b, Anhydro~e schscholtzxanthin. This compoumd was crystallized from benzene=rethanol according to the procedure described above for bisdehydro-S-carotene, Yield, about img, A total of 14 mg. was isolated from numerous similar experiments, Gharacterization, Crystal forms Individual crystals from benzene-methancel wers browmish=purple; overlapping crystals appeared almost black. In general, the crystals formed bulky conglomerates but occasionslly parallel twinning and fan-like groupings were found, (Pig. 28). Macroscopically, the crystals appeared very dark purple, Melting point: 192-1939 Solubility: More soluble in hexane and benzene than bisdehydro- ns teat ts a togeg sahechol ia * Arcarotens; inscluble in methanol. fartition test: Untirely epiphasic in the hexane-95% methanol systen, Analysis: Calculated for SoHsq3 6, 90.503; H, 9.50. Found: 6, 89,893 H, 10.10 Srectral maxiza, in hexane: 534, 499 mp. (visual); 531, 499,472 m. (Eeckman). Molecular extinction ecefficient, at d mere? ar = 16.0 x 104, Chromatographic behavier: Adsorbed telow bisdehydro- 4-carotene but above S-carotene on live-celite, when developed with 3% acetone in hexane. 3. Sone Sterecisovers of Bisdehydro-4-carotene. A golution of 3 mg. of bisdehydro- 4-carotens in 100 ml. of hexane containing about 1% of tha pigment weight in icdire was illuminated under the same conditions as prescrited for the iodine catalysis cf dehydro~ Arcarctene (Part I, Section B., 3a.), The isomerized scluticn was develored on a 28 x 5 cm. line-celite coluwm with benzene-hexane (1:3): 5 very pale purple 10 interzone 5 very pale pink 15 red all-trang 521, 490, 457 mm. 10 orange-red neo A 518, 485, 456 (rig. 25) 20 interzcne 15 pale orange~red neo B 514, 483 (Fig. 25) 20 intersone 20 pale orange neo G 510, 481 (Fig. 25) Before extrusion, the colurm was freed from benzene by washing with hexane, Each zones was cut out, eluted, and transferred to hexane, Zstingtes of the amount of pigment in each fraction were made spectro= “l12- photometrically. The composition of the stereoligomeric mixture (in % of starting raterial) is given below: Ali~trens~bisdehydro~ A-carotene 24% Neo-~bisdehydro- 8 -carotene A 25 Neo B 13 Neo 0 6 Unaccounted for 32 The atnormally large percentage of destroyed pigment is indicative of the sensitivity of this compound. 4. Sore Stereocisomers of Anhydro~eschschol taxanthins Systenatia sterecisomerization experiments were not carried out with this compound, However, chromatographic resolution of a hexane solution which had teen kept at 4° for about 3 days showed the presence of small amounts of two new substances both adsorbed below anhydro-eschacholtz- xanthin. They were cig isomers of the latter since chromatographic resolution of their iodine catalyzed hexane solutions showed the presence of a pigment that was chromatographically and spectroscopically identical with anhydre-eschscholtzxanthin, The more strongly adsorbed of the cigs isomers had spectral maxima identical to that of the all-trans compound (cf, Fig. 26), ~113- APPEND Te A. PROVITAHIN a POTENCY oF A A-CAROTENE HOMOLOGIE , 1. ntreduction and General Remarks, Recently, a homologue of 4-carotene was obtained by total synthesis ty Inhoffen, Bohlmann, Aldag, Bork, and leibner (8). This new compound, termed 16,16'-hono- A=csrotena (Formula VI) differs from @~carotana in the presance of two axtra CH groups in the central part of the chromophore and is the first representative of a carotenold of this type. Although the last step in this synthesis (partial saturation of acstyleaic bonds by means of a poisoned catalyst) led to the formation of a nolecule with two centrally located gis bonds, these authors wera also able to obtain the all<trang fern by iodine catalysis (Fig. 29). HE Jilly (' By fis (" Hp Ay Hof ' PG-CH=CH~C2CH-CH=GH-G2cH-C)=CH~GH=CH-CH=0-CHZCH-CH=U-cHZGH-0% "DoH, 7 iS 16 6’ 15° 7 | Hot2 4 £0-CH HaO=O8 4 DOH. aN tf, 3 3 “age 2 tw Formula VI 16,16'~homo-~4-carotene, {The numbering system follows the nomenclature of Karrer,) Bioassays of the all-trans and di-gis spatial forms were carried out with rats, In accordance with the usual procedure, the supplements were atuiulstered In Wesson oll sulutions*, As shows in Table 13.4, both asc a cam ae . Se ef a ae Soe Ee en ee ee a ee OR * This work has keen published in collaboration with H. J. Deuel, Jr., H, H. Inhoffen, J. Ganguly and L. Zechmeister (2). Determinations of the spectral characteristics and the degree of stability, under atand- artigad conditions, of the compomds in ‘esson oll were carried out in this laboratory, while the bioassays were performed at the University of Southern California, JI wish to thank Prof, H, H. Inhoffen, Braunschvelg, Gernany for the two synthetic sanples. “lli- *(STepou woxsTeys) (MoTeq) teuost STo-tp [eiquso 8)7 pue ‘(aaoge) ate_018d— of ~ONIO4~.9T ‘g[-BUBTR-TLT¥ “62 eanaTy ~115- spatial forms were active, Table 13, Relative Provitamin A activities in tha Rat SRAM EB at MERE DO A a ECR Ga ie ME SS RES IR i SOM LC A Sa RR Compound Relative potenay (4) All~trang- § -carotena 100 All-trans-16,16!-homo~ 8 -carotana 22 15, 16,15',161-di-gig-homo- B -carotcone 20 Ne i me MME RS ESSER RR Ee me a RR MR EGET TORRE RD OIE BE On the basis of the considerable provitamin A potency of homno~ frcarotane the following conclusions may be draws a) The conversion of a carotenoid to vitamin 4 takes place even when the nunber of carbon atoms in the C-C skaleton of the provitamin molecule exceseda forty, (It had been found eariler by Karrer and Solmssan (18) that A-apo~2=carotenal, a carotenoid with thirty carbon atoms showed strong provitamin A potency. ) b) The absence of an isoprenic structure in the central portion of the chromophore desea not prevent enaymatic attack in that region, . eo) It is not essential, for a successful enzyratic attack, that the provitamin A have a double bond at the eenter of the molecule, the practically equal biopoteneies of the all-trans and di-gis forma ot homo~ B “carotene can be explained, tentatively, by the fact that. both have otvaight molecular shapes. lO. x1O* mol lom E 6 Figure 30, Udlecular extinction curve in Wesson of] of all«grang- LG, 16%*haioe p “carotene s NT x1O~* mol E cm Oo Figure 31, Meleevlar extinction curve in veason of2 ef Uegiselh, 16) <horme- p “OO POTHO « ~118- BE Ee RT ' In order to prepare the supplements, the carotenoids were dissolved in peroxide free diethyl ether and transferred to flasks having ground- glass stoppers; the ether was then removed under gentle suction, leaving the pigment as a thin film on the walls of the flask, About 90-100 ml, of lesson oi1 (containing 0.5 mg, of X-tocopherol ver 0.1 ml. of solution) was immediately introduced, the air displaced by nitrogen, and the flasks stopvered, They were allowed to stand for 3-4 days in the dark at room temperature, and were shaken occasionally. The oily solutions were then passed through sintered-glass funnels of medium porosity and the con- centration of the pigment in each filtrate determined spectrophoto- metrically. They were then diluted to give concentrations of 1, 2, or 4 pg./ 0.2 ml. of solution in the case of the synthetic Cro compounds, and 0.5 and 1.0 pg. in the case of all-trans-8-carotene, All solutions were stored under nitrogen in 10 ml, browm bottles at 5°, New samples were used twice weekly during the assay period. Spectral readings were made at d max, with a Beckman spectrophoto- rater, Model DU, before and after the bioassay. No significant changes could be noted, demonstrating that both the di-cis and all-trans con- figurations of homo~-carotene show remarkable resistance toward thermal stereoisomerizations in Wesson oi] solutions, The values for the molecular extinction coefficients in Wesson oil ares 4All-trans-16,16'-homo--carotene, pnol. = 11,5 x 104 (at 477 me ) 3 di-gig isoner, BPol. = 10.0 x 104 (at 466-467 m.), For the molecular extinction curves, cf, Figs. 30 and 31. le 2, Be Ae 3e 6. 7. 8. 9. 10. 12, 13. 15. 16, 17, 18, 19. 20. ~119- REFERENCES Chatterjee, A. and Zechmeister, L: J. Am. Chem. Soc. 72, 254 (1950). Devel, H. J. dr., Inhoffen, H. H., Ganguly, J., Wellcave, L. and Zechmeister, L,: Arch, Biochem, and Biophys. 40, 352 (1952). Farrar, K., Hamlet, J. C., Henbest, H., B. and Jones, E.R. He: J. Chem. Soc. (Lond.) 1952, 2657. Friedel, R. A, and Orchin, M.: Ultraviolet Spectra of Organic Compounds, John Wiley & Sons, New York, 1951. Garbers, G. F., Eugster, C. H. and Karrer, P.: Helv. chim. Acta 35, 1850 (1952). Goodwin, T. W. and Taha, M. M,: Biochem. J. 48, 513 (1951). Hennion, G. F. Hinton, H. D. and Nieuland, J. A.: J. Am. Chem, Soc. 55, 2858 (1933). Inhoffen, H. H., Bohlmann, F., Aldag, H. J., Bork, S. and Leibner, G.: Ann, 573, 1 (1951). Inhoffen, H, H. and Siemer, H.: Fortschr. Chem. org. Naturstoffe 9, 1 (1952). Karrer, P.: Fortschr. Chem. org. Naturstoffe 5, 1 (1948). Karrer, P.: Helv. chim, Acta 34, 2160 (1951), Karrer, P, and Rugster, C. H.: Helv, chim. Acta 34, 1400 (1951). Karrer, P,., Jucker, E., Rutschmann, J. and Steinlein, K.: Helv. chim, Acta 28, 1150 (1945). Karrer, P. and leumann, E.: Helv, chim, Acta 34, 445 (1951). Karrer, P, and Rutschmann, J.: Helv. chim, Acta 28, 793 (1945). Karrer, P., Sehépp, K. and Morf, R.: Helv. chim, Acta 15, 1158 (1932). Karrer, P, and Schwab, G.: Helv. chim, Acta 23, 578 (1940). | Karrer, FP. and Solmssen, U.: Helv. chim, Acta 20, 682 (1937) ‘Karrer, P, and Takahashi, T.: Helv. chim, Acta 16, 1163 (1933). Kuhn, R. and Grundmann, C.: Ber. 66, 1746 (1933). 25. 30. 31. 3260 33.6 =120~ Kulm, R. and Lederer, Be: Ber. 65, 637 (1932). Kuhn, R. and Winterstein, Aet Ber. 66, 1733 (1933). — IeRosen, A. L.t Ind. and Eng. Chem. (Anal. Ed.) 14, 165 (1942). Meunter, P, and Vinet, P.: Chromatographie et Mesomerie, Masson & Cie, Paris, 1947. Oroshnik, ¥., Karmas, G. and Mebane, A. D.: J. Am, Chem, Soc. 7A, 295 (1952). | Pauling, L.: Fortschr, Chem. org. Naturstoffe 3, 203 (1939). Pinckard, J. H., Wille, B, and Zechmeister, L.: J. Am. Chem. Soc. 70, 1938 (1948). Polgar, A. and Zechmeister, Le: J. Am, Ghem, Soo, 64, 1856 (1942). Polgar, A. and Zechmeister, L.: J. Am. Chem. Soc. 65, 1528 (1943). Strain, H. H.: J. Am. Chem, Soc. 63, 3448 (1941). Strain, H. H.: J. Biol. Chem. 123, 425 (1938). gechmeister, Le: Chem, Rev. 34, 267 (1943). Zechmeister, L. and LeRosen, A. Lez J. Am. Chem, Soc. 64, 2755 (1942). Zechmeister, Le, LeRosen, A. L., Sehroeder, W. A., Polgar, A. and Pauling, Le: J. Am. Chem, Soc. 65, 1940 (1943). -121= Propositions submitted by L. Walleave 1. Within recent years certain organic compounds in crystalline form or in solutions have been used, in conjunction with photo- miltipliers, as scintillation counters for ¥ -radiation or energetic particles. So far, only aromatic or partially aromatic hydrocarbons such as stilbene, terphenyl, naphthalene, and anthracene have been investigated, (Cf. R. Hofstadter: Nucleonics 6, (No. 5), 70 (1950)). I propose these investigations be extended to include 4 -dithienyl and &-terthienyl, compounds containing sulfur and exhibiting some aromatic character, Terthienyl is strongly fluorescent in ultraviolet light. | HO——CH || Q~ TERTHIENYL CH HO—-CH HC——CH igagm ‘s’ sf ‘s’ 2. When a solution of phosphotungstic acid, HP®, 20, o» was photochemically reduced in the presence of isopropyl alcohol the colorless solution turned black, (L. Chalkley: J. Chem. Phys. 56, 1084 (1952)). Upon oxidation with silver nitrate the original acid was again formed, The composition of the black acid could not be determined precisely, but Chalkley suggested the formulas, Hy Py 20) 9 or Hp PH 203g- I propose that the reduced acid, as indicated by its color, contains tungsten in a mixture of two or more oxidation states, 3, A white solid showing the empirical composition AlH3 can be isolated from the reaction mixture formed by the interaction of aluminum chloride and lithium hydride. (A. B, Finholt, A, C. Bond, Jr. ~l22= and H. I. Schlesinger: J. Am, Chem. Soc. 69, 1199 (1947)). I propose that X-ray diffraction studies be made with this substance to supplement available information on the nature of the covalent aluminum bonds in such compounds as AlCl, and Ala (CH3) ¢- Ae Myxoxanthin, G 4045 hd isa ke to~carotenoid of undetermined constitution but which may have the Y-carotene carbon skeleton. (Cf. P, Karrer and E, Jucker: Carotenoids, Elsevier, New York, 1950, p. 225). I propose that the carbonyl group in myxoxanthin be reduced by means of the Wolff-Kishner mthod and the resultant hydrocarbon compared with authentic T-carotens. 5, The Thiele addition of water or methanol to the conjugated system in dehydro- B ~caratene in the presence of boron trifluoride (reported in this Thesis} should also occur with other electrophilic reagents such as aluminum chloride or antimony trichloride. Furthermore, the use of boron trifluoride as a catalyst for the addition of water or alcohols to conjugated dienes or polyenes should be investigated, 6, The adsorption affinities on lime of dehydro- fh ~carotene (CyoHs, 3 12 conjugated double bonds), bisdehydro-B ~carotene (C, oHga3 13 conjugated double bonds), and anhydro-eschscholtzxanthin (C,oH593 14 conjugated double bonds) decrease in that order (reported in this Thesis). Since these compounds all have the same carbon skeleton, I propose that this unexpected sequence is associated with the steric effects of the methyl groups located near the ends of the conjugated systen, | 7. %T. We Goodwin has reported (Biochem. J. 51, 458 (1952)) the presence of a colorless, fluorescent polyene with a phytofluene like -123- absorption spectrum in the berries of Lonicers japonica. Since it was absorbed less strongly than phytofluens, which is dodecahydro-lycopene (C,oH6g), Goodwin suggested that the new compound is dodecahydro- i “carotene. I suggest that the new compound may well be the cis phytofluone reported by F. J. Potracok and L, Zechmeister (J. Am, Chom. Sec. 74, 184 (1952)). I also suggest that dode cakydro- ~carotene, should it exist, would be found in carrots, in which f-carotens is the main pigment and where lycopene is found only occasionally, in small amounts, Therefore, I propose that the phytofluene from carrots be compared with that obtained from tomatoes, in which lycopene is predominant. The situation in the case of Lonicera japonica is unclear because considerable amounts of both p-carotene and lycopene are found in these berries and thus might be expected to contain both phytofluene (dodecahydro-lycopene) and dodecahydro~ B “carotene, 8 According to M. J. S, Dewar (Electronic Theory of Organic Chemistry, Oxford Univ. Press, London, 1949, p. 136) the Meerwein~ Ponndorf reduction using aluminum isopropoxide may proceed by means of a hydride ion transfer within a cyclic intermediate, as follows: r\ RoC juve 0 0 Al(0-isoPr) 5 I propose the use of boron isopropoxide in the reaction to test this hypothesis considering the fact that boron alkoxides form stronger "acceptor" bonds with carbonyl groups than do the corresponding -124- aluminum compounds. (Cf. N. V. Sidgwick: The Chemical Elements and Their Compounds, Vol. 1, Oxford Univ. Press, London, 1950, p. 421). 9. In the metalation of triphenylamine with butyl lithium the metal atom substitutes a ring in the meta position. In corresponding reactions with aniline, diphenylamine, and N-phenylcarbazole the substitution occurs in the ortho position. (H. Gilman and S. M. Spatzs J. Org. Chem. 17, 860 (1952)). These authors suggest that the . exception to the "ortho rule" in the case of triphenylamine results from "the steric hindrance of the free rotating phenyl groups". I propose that this explanation is at best only partly correct since the metalation should then occur at the para position, I further propose that the metal atom is. directed into the meta position by the presence of a considerable positive charge on the nitrogen atom owing to resonance effects; thus this process would be independent of any steric factors, 10. A serious problem in the analysis of many organic compounds is the preparation of ash-free samples. I propose the use, for this purpose, of ethylenediamine N-tetra-acetic acid, (HOOCGH> ) oNCH>- CHaN(CHCOOH) 55 sold under the trade name "Sequestrene®,