The Lipids of the Common House Crickmt, Acheta Domesticus L. II

The Lipids of the Common House Crickmt, Acheta domesticus L. II. Hydrocarbons

RODERICK F. N. HUTCHINS' and MICHAEL M. MARTIN,: Depart~hent of Chemistry, University of Michigan, Ann Arbor, Michigan 48104

ABSTRACT methyl~lkanes, highly branched alkanes, and The hydrocarbons of the common house cycloalKanes have now been recognized as sig- cricket Acheta domesticus L. consist of n- nificant components of a number of natural alkanes (4%), 2-methylalkanes (20%), waxes (1-10). By far the most thorough study x-methylalkanes (59%), an unidentified of a n~turally occurring hydrocarbon mixture series of alkanes (trace), and olefins (16- has beeaa by Mold et al. (11,12), who utilized 18%). The major n-alkane is n-nonaco- the hydk'ocarbon fraction of wool wax. They sane (3.5%). The major 2-methylalkanes have characterized 93 components of a hydro- are 2-methyloctacosane (8%) and 2- carbon fraction, which constitutes only 32% methyltriacontane (11%). The members of the total hydrocarbons present. of the homologous x-methylalkane series The characterization of the hydrocarbons of consist of mixtures of methylalkanes in insect duticle is of interest for a number of which the methyl side-chain is located on reasons A knowledge of the constituents of the 13th, 15th, and 17th carbon atom in cuticular wax is clearly basic to an understand- the chain. The major x-methylalkanes are ing of the role of the cuticle in maintaining the homologues containing 31 (3%), 33 water balance. Likewise basic to any bio- (6.5%), 35 (12.5%), 37 (27%), and 39 chemical study designed to establish the me- (2.5%) carbon atoms. The olefins are a tabolic pathways available for hydrocarbon mixture of straight-chain and 2-methyl- synthest is the characterization of the meta- alkenes and alkadienes. The major olefins bolic end-products. Hydrocarbon biosynthesis contain 31 (3%) and 37 (7.5%) carbon is an area of lipid metabolism in which little atoms. has been done. Robbins et al. (13) and Piek (14) have demonstrated the incorporation of INTRODUCTION 1~C from 1-14C acetate into the hydrocarbons of the wmx of Musca domesticus and Apis mel- HE ADVENT OF analytical and preparative li/era ra pectively, but any characterization of T gas-liquid chromatography (GLC), cou- the anabolic pathways connecting acetate and pled with mass spectral techniques, has per- the marw constituents of the hydrocarbon mix- mitted detailed analyses of naturally occurring tures prmsent is lacking. Finally chemical tax- hydrocarbon mixtures heretofore considered onomy, utilizing hydrocarbon patterns, prom- beyond the realm of possibility. Whereas early ises to l~e a valuable adjunct to classical mor- studies of the hydrocarbon constituents of phologioll taxonomy when applied to large various plant and insect waxes had succeeded groups of similar and closely related species (7, in recognizing the presence of an homologous 15). series of alkanes, usually characterized as pre- This paper discusses the characterization of dominantly odd carbon-numbered normal al- the hydJpcarbons of the common house crick- kanes in the 27 to 33 carbon range, more re- et, Achefa domesticus I_. As a consequence of cent studies have demonstrated a much greater the relalive ease of maintaining laboratory degree of complexity. colonies (16,17) and because of its commercial Even-numbered as well as odd-carbon-num- availability, A. domesticus is an attractive bered alkanes, alkenes, alkadienes, mono- species t~r chemical, biochemical, and biologi- methylalkanes with the methyl side-chain lo- cal studll~s. cated at a number of different positions, di-

MATERIALS AND METHODS

1Postdoctoral Research Associate. Department of Chem- Isolation of the Hydrocarbon Fraction istry, University of Michigan, 1965-1967. 2Associate Professor. Deoartment of Chemistry, Univer- In a I~pical run, 5.0 g of lipid extract, ob- sity of Michigan; Alfred P. Sloan Foundation Fellow. to tained a,~ described in the previous paper (18), whom inquiries regarding this publication should be ad- dressed. was adsorbed onto a column of 400 g of Flori- 250 LIPIDS OF (,t',OMMON HOUSE CRICKET. II 251

sil (Fisher) to which 28 g of water had pre- analyses of prepared mixtures of known com- viously been added (19). A hydrocarbon frac- position it was found that the most -accurate tion weighing 290 mg was eluted with petro- quantitative data were obtained by cutting out leum ether (40-60(2, redistilled). This hydro- the individual peaks and weighing them. Such carbon fraction showed no absorption in its in- a procedure was far superior to triangulation or frared spectrum characteristic of a carbonyl or disc integration. The quantitative data present- hydroxyl function. ed in Table I are averages of three or four different determinations run at different tem- Separation of Saturated from Unsaturated Hydrocarbons (20) peratures. In this and all subsequently described opera- Removal of n-Alkanes from an Alkane Mixture tions anhyrous reagent grade ether and redis- In order to confirm the identification of the tilled reagent grade petroleum ether were util- n-alkane series in the saturated alkane mixture, ized. The glassware was carefully cleaned. n-alkanes were removed by treatment with a In a typical run 1.15 g of cricket hydrooarbons molecular sieve (Linde, 5 A). Those peaks ab- was placed on a column of 50 g of silica gel, sent from the GLC of a sample so treated cor- impregnated with 25% silver nitrate (Ad- responded exactly to those which had been iden- sorbosil-CABN, Applied Science Laboratories). tified as n-alkanes by comparison with stand- Saturated hydrocarbons (0.97 g) were eluted ards and by graphing of the logarithms of re- with petroleum ether, unsaturated hydrocar- tention times against carbon number. In a bons (0.19 g) with ether: petroleum ether typical experiment 101 mg of an alkane mix- (10:90). Infrared spectra of both fractions in- ture, from which unsaturates had been re- dicated an efficient separation of saturated from moved, was dissolved in 7 ml of isooctane, and unsaturated hydrocarbons. the solution was shaken with 6 g of Linde 5 A Hydrogenation of Unsaturated Hydrocarbons molecular sieves (1/16-in. pellets) overnight. The unsaturated hydrocarbon fractior, (100 The mixture was then filtered, and the sieves rag), dissolved in 25 ml of acetic acid to which were washed twice with an equal volume of iso- just enough petroleum ether was added to ef- octane. The filtrate was concentrated to 7 ml, fect solution, was hydrogenated at room tem- 6 g of Linde 5 A molecular sieve were added, perature and atmospheric pressure for 2 hr us- and the mixture was shaken for 24 hr, after ing 100 mg of 5% platinum on carbon (Engel- which it was filtered and washed as described hard Industries Newark, N. J.) as catalyst. previously. Upon concentrating the isooctane The mixture was filtered, water was added, and filtrates, a hydrocarbon mixture was obtained the hydrogenated product, was extracted with which contained only the Series II and the petroleum ether. Removal of the petroleum Series III alkanes. The largest peaks in the ether solvent gave a hydrocarbon product de- gas chromatogram which were removed were void of any infrared absorption attributable to the two with retention times identical to n-octa- a doub!e bond. cosane and n-nonacosane. GLC Conditions Oxidative Cleavage of the Unsaturated Hydrocarbons (21) The GLC analyses were conducted on a dual column F & M Model 810 gas chromatograph, To a solution of 135 mg of unsaturated hydrocarbons in 50 ml of t-butyl alcohol was add- equipped with dual flame ionization detectors. Analyses were run isothermally at temperatures ed a solution of 600 mg of sodium periodate between 220(2 and 330C on 6 ft. • 1~ in. o.d. and 40 mg of potassium oermanganate dissolved in 50 ml of water. The oH of the mix- 10% Apiezon L on acid-washed silanized ture was adjusted to 8-9 by the addition of Chromosorb W columns. High operating temperatures were advantageous if the analyses aqueous potassium carbonate, then the solution was stirred for four days. The solution was ex- were concerned with higher homologues, but lower temperatures were used for the lower tracted with methylene chloride to remove any homologues. neutral materials, of which there appeared to be none. The aqueous solution was then de- Determination of Amounts of Each Component colorized and neutralized with an aqueous The relative amounts of the various com- solution of sodium bisulfite and extracted with ponents in the saturated and unsaturated hy- chloroform. The chloroform solution was dried drocarbon mixture were determined by quanti- over sodium sulfate, filtered, and evaporated tative GLC. The higher homologues in these to dryness, leaving 117 mg of acids. This resi- mixtures gave broad, rounded peaks, and by due was esterified using boron trifluoride/

LIPIDS, VOL. 3, NO. 3 252 RODERICK F. N. HUTCHINS AND MICHAEL M. MARTIN

TABLE I The Hydrocarbons of the Common House Cricket, Acheta domesticus L., Percentage Composition a

Series II Series III Carbon Series I (2-methyl- (x-methyl- Series IV Number ( n-alkanes ) alkanes) alkanes) (unidentified) ( Olefins )

15 Trace Trace -- Trace -- 16 Trace Trace -- Trace -- 17 Trace Trace -- Trace -- 18 Trace Trace -- Trace Trace 19 Trace Trace -- Trace Trace 20 Trace Trace -- Trace Trace 21 Trace Trace -- -- Trace 22 Trace Trace -- -- Trace 23 Trace Trace -- -- Trace 24 Trace Trace -- -- Trace 25 Trace Trace -- -- Trace 26 Trace Trace -- -- Trace 27 Trace Trace Trace -- Trace 28 <0.5 Trace -- -- Trace 29 3.5 8.0 Trace -- 1.5

30 Trace 0.5 Trace -- <0.5

31 -- 11.5 3.0 -- 3.0 32 -- -- 0.5 -- Trace

33 -- 0.5 6.5 -- 1.0b

34 -- -- 1.5 -- Trace

35 -- -- 12.5 -- 0.5 e

36 -- -- 4.0 -- <0.5

37 -- -- 27.0 -- 7.5a

38 -- -- 1.5 -- 1.5

39 -- -- 2,5 -- 1.5

Total % 4.0 20,5 59,0 Trace 17.5

Accurate to 2-3 absolute percentage for components present to the extent of 3% or more. Percentages for minor components have a larger error associated with them. bThree overlapping peaks on GLC, one minor and two of equal magnitude, eTwo overlapping peaks of equal magnitude on GLC. dTwo peaks on GLC, one very minor, the other major (> 97%).

methanol (22) and analyzed by GLC with the homologous series, designated Series I, ]I, and use of 6-ft. • a/8-in O.D. 6% LAC 728 col- IIIY umns. Series I was identified as homologous normal alkanes; every carbon number from 15 to 30 RESULTS AND DISCUSSION was represented although most of them were In Table I is presented the distribution of the only in trace amounts. The line resulting from cricket hydrocarbons. The weight percentage the logarithmic plot coincided precisely with composition values have been rounded off to one constructed from authentic reference sam- the nearest half per cent. Those components ples. Furthermore the two major components designated as "trace" constituents were clearly of the Series I hydrocarbons had retention identifiable peaks on the gas-liquid chromato- times on GLC identical to authentic n-octaco- grams and fit nicely onto the logarithmic plots sane and n-nonacosane. Treatment of the sat- but were so minor as to make any attempt at urated hydrocarbon mixture with the Linde 5 quantitation pointless. A molecular sieve completely removed all of The hydrocarbon mixture initially obtained the components present in Series I. These hy- from the crickets was subdivided into a satu- drocarbons are therefore normal alkanes. The rated and an unsaturated fraction by column chromatography on silica gel impregnated with .~Authentic samples utilized for direct comparison and silver nitrate. GLC of the saturated hydrocar- for the construction of logarithmic plots included n-hexane, bon fraction indicated a complex mixture. n-octane, n-decane, n-tetradecane, and n-hexadecane (F & M Scientific Corporation); n-octadecane and n-tetracosane However, by plotting the logarithm of the re- (Applied Science Laboratories); n-octacosane, n-dotria- tention times of the various components against contane, and n-hexatriacontane (Aldrich Chemical Com- pany); n-nonacosane and n-hentriacontane (donated by E. a linear ordinate representing the number of Ritchie, Department of Organic Chemistry, University of carbon atoms present in the hydrocarbon, three Sydney, Sydney, Australia) ; 2-methyltricosane, 2-methyl- parallel lines resulted, indicating that the sat- octacosane, 3-methyltetracosane, and 12-methylnonacosane (donated by James Mold, Research Department, Liggett urated hydrocarbon mixture consists of three and Myers Tobacco Company, Durham, N. C.). LIPIDS, VOL. 3, No. 3 LIPIDS OF COMMON HOUSE CRICKET. II 253

n-alkanes, which so often constitute the bulk The 36-carbon homologue submitted for of naturally occurring hydrocarbon mixtures, mass spectral analysis was unfortunately con- make up only 4% of the hydrocarbons of taminated with significant quantities of 35-car- Acheta domesticus L. bon Series IfI alkanes. Thus the mass spectrum Series II was identified as homologous 2- is less definitive than the one discussed above. methylalkanes (isoalkanes). An authentic sam- Nonetheless it is informative. Prominent frag- ple of 2-methyloctacosane had a retention time mentation peaks at 337 (C24H4~) and 197 identical to the 29-carbon Series II hydrocar- (C~4H2o) indicate the presence of 13-methyl- bon. An authentic sample of 3-methyltetraco- pentatriacontane; peaks at 309 (CzzH4~) and sane had a different retention time from the 225 (C1~H~3) indicate 15-methylpentatriacon- trace Series If component with 25 carbons. Al- tane; and peaks at 281 (C20H4~) and 253 (C~ so, a mass spectrum of the saturated hydrocar- H~7) indicate 17-methylpentatriacontane. The bon mixture had parent peaks at 408 and 436 35 carbon contaminants appear to be 13- mass units, corresponding to C29Hq q and C~ methyltetratriacontane, 15-methyltetratriacon- H64 hydrocarbons (the two major Series II al- tane, and 17-methyltetratriacontane. kanes); large fragment peaks at 393 and 421, Clearly the mass spectrum of a mixture as corresponding to the loss of 15 mass units; a complex as this one is not subject to rigorous methyl group from each of parent ions but not interpretation. However the absence of prom- fragment peaks at 379 and 407, which would inent fragmentation peaks at 155 (C11Ho~), have resulted from the loss of 29 mass units; 183 (Cl~H27), 211 (CI~H~,), and 239 (C1~ an ethyl group from the parent ions of 3- H35) requires that no methyl branches occur methyloctacosane and 3-methyltriacontane. on carbon atoms number 10, 12, 14, or 16 in Thus it can be concluded that the Series II the alkane chain. Likewise the absence of any hydrocarbons are 2-methylalkanes (isoalkanes) major fragment peaks above 337 (C~H~,) re- a-d that 3-methylalkanes (anteisoalkanes) are quires that no methyl group occur closer to the absent from the hydrocarbon mixture. The end of a 35-carbon chain than carbon atom isoalkanes, a class of hydrocarbons which is number 13. being encountered more and more frequently in The mass spectrum of the 35-carbon Series naturally occurring hydrocarbon mixtures, con- III alkane, which is contaminated with 34- and stitutes 20.5% of the cricket hydrocarbons. 36-carbon homologues, is even more complex The Series II1 hydrocarbons constitute 59% and less rigorously interpreted. However the of cricket hydrocarbons. The 37-carbon mem- pattern is the same. The mass spectrum is con- ber of this series alone constitutes 27% of sistent with a mixture of 13-methyltetratriacon- the total. This series was established to contane (peaks at 323 and 197 owing to C~H,~ sist of methylalkanes having the methyl side- and C~4H~,), 15-methyltetratriacontane (peaks chain toward the middle of the carbon chain. at 295 and 225 owing to C2aH,3 and C~Ho~), In fact, each homologous member of this series 17-methyltetratriacontane (peaks at 267 and consists of a mixture of alkanes having the 253 owing to C~,H~9 and C,,Ho~), 13-methyl- methyl side-chain at different positions in the tritriacontane (peaks at 309 and 197 owing to carbon chain. This was established by collect- Co.H4~ and Ca,H~9), 15-methyltritriacontane ing and submitting to mass spectral analysis (peaks at 281 and 225 owing to Co,H,~ and the 35, 36, and 37 carbon members of the C~H~), 17-methyltritriacontane (one peak at series. The mass spectra were interpreted in the 253 owing to CxsH3r), and 13-methylpenta- same way that Mold et al. (12) interpreted triacontane (peaks at 337 and 197 owing to mass spectra of similar hydrocarbons. C,,_~H~, and C14H29 ). The 37-carbon homologue, with a parent Although the suggested composition of this peak at 520 (C:~7H7~), had prominent fragmen- mixture may be subject to debate, what the tation peaks at 323 (C~oH4~), 295 (C2~H~.~), mass spectrum indicates incontestably is that 253 (C, sH~7), and 225 (C~H~3). This pattern there are no major fragment ions with the suggests that the sample is a mixture of 15- formulas C~H2.~ (155), C~.~H2~ (183), C~sH~ methylhexatriacontane and 17-methylhexatri- (211), and C~H3.5 (239), thus reinforcing the acontane. The major fragment ions result from conclusion arrived at earlier that, in the Series cleavage of the parent ion on either side of the III alkanes, methyl groups occur at the odd- methyl group. There may be traces of 13- numbered carbon atoms 13, 15, and 17 and methylhexatriacontane since there are slightly not at the even-numbered carbon atoms 10, enhanced fragment peaks at 351 (C2.~H~) and 12, 14, and 16. 197 (C14H29) . Hydrocarbons of this type, in which the LIPIDS, VOL. 3, NO. 3 254 RODERICK F. N. HUTCHINS AND MICHAEL M. MARTIN methyl group is located at a carbon atom sinte they would have been lost in the work- toward the center of the chain, have been up The predominant straight-chain acids were identified in wool wax (12), where they com- th*t 7, 12, 14, 24, and 26 carbon acids. prise 4.5% of the total saturated hydrocarbons, [Methyl-substituted acids with from 20 to 26 and may have been detected in sugar cane ca bon atoms were also produced; the 22 and wax (9) and in the waxes derived from three 24 carbon homologues predominated. Thus plants, Humulus lupulus, Populus nigra, and in the 2-methylalkadiene series it appears that Rosa spec. (5, 6), but in none of these cases ur'~aturation gets no closer to the branched do they constitute the major hydrocarbon type. en{l of the chain than 19 carbon atoms. The The Series IV hydrocarbons, which are trace str)tight-chain acids suggest a similar location components in the 15 to 20 carbon range were of the unsaturation in the normal alkenes and not identified. They fall on the same logarith- alJ/adienes. A more complete structural analy- mic plot as the Series III hydrocarbons, but sis of this olefin mixture would require isola- since no Series III or IV hydrocarbons in the ti0h of individual homologues. Alkanes and 21 to 26 carbon range could be detected even alJtadienes have been isolated from a number as trace components, it does not seem likely of natural sources, but 2-methylalkenes and 2- that the two series are related anabolically. m,thylalkadienes do not appear to have been Besides, since the Series III alkanes have their rel~orted previously. methyl group at least 13 carbon atoms from The hydrocarbons of the crickets' food were the nearest end, no 15 or 20 carbon compound alJb surveyed. In addition to large amounts of could class as a member of such a series. No hishly branched material, probably largely attempt was made to characterize these trace isobrenoid in nature, which would not form a constituents. solhble urea clathrate compound, there was also present some material which did form a The olefinic hydrocarbon fraction constitutes soluble clathrate. This material was almost 16-18% of the total hydrocarbons. The GLC ent)rely composed of n-alkanes, from n-non- of the olefin fraction clearly indicated one adecane through n-tritriacontane. At most only major and possibly two minor homologous traces of Series II isoalkanes were present, and series. Hydrogenation of the olefin mixture none of the Series Ill hydrocarbons were de- gave an alkane mixture composed of n-alkanes testable. Thus the n-alkanes in the cricket may from 18 through 39 carbon atoms and 2- be present as a consequence of the selective methylalkanes from 22 through 39 carbon up take of dietary hydrocarbons. However the atoms. The n-alkanes made up about 67% of Series II hydrocarbons appear to be, and the the hydrogenation mixture, the isoalkanes Series III hydrocarbons most certainly are, about 33%. No 3-methylalkanes or Series III products of anabolic hydrocarbon metabolism. alkanes were detectable. The major n-alkanes So little is known about hydrocarbon bio- were n-nonacosane, n-hentriacontane, n-hepta- synthesis that it is hardly even fruitful to specu- triacontane, n-octatriacontane, and n-nonatria- lat, about the origin of the cricket hydrocar- contane. All of the isoalkanes were present in bons. Long-chain fatty acids are plausible small quantities except for 2-methyldotriacon- precursors. Because of the ease of raising A. lane, 2-methyltetratriacontane, and 2-methyl- domesticus, the ease with which radioactive hexatriacontane, which were the three major fatty acids could be fed to them, and the ease isoalkanes. with which the various hydrocarbon classes can A mass spectrum of the hydrocarbon mix- be separated, crickets would appear to be an ture obtained initially from the crickets had ideal organism for the study of hydrocarbon parent peaks at 544 (C:~,,HT.). 530 (C~8H7~), metabolism. 516 (C:~7H~2), 460 (C:~:~H,.~), 432 (C:~H,o), and 404 (C.2,0H.~,0, indicating a prevalence of ACKNOWLEDGMENT dienes. An ultraviolet spectrum of the olefin mixture had a weak maximum at ~...... of 230 Supported in part by the University of Michigan Cancer /~ (~ ~ 500, assuming an average molecular ReSearch Institute and the American Cancer Society. weight of 486). Thus there is less than 5% conjugated diene present. The oxidative cleav- REFERENCES age of the olefin mixture is informative al- 1# Baker, G., J. H. Pepper, L. H. Johnson and E. Hast- though the complexity of the mixture precluded ings. J. Ins. Physiol., S, 47-60 (1960). complete structural assignment. Straight-chain 2. Baker, G. L., H. E. Vroman and J. Padmore, Bio- chem. Biophys. Res. Commun. 13, 360-365 (1963). acids from 6 to 26 carbon atoms were obtained. 3, Carruthers, W.. and R. A. W. Johnstone, Nature 184, The absence of lower acids is not significant 1131-I132 (1959).

Livins, VOL. 3, NO. 3 LIPIDS OF COMMON HOUSE CRICKET. II 255

4. Hallgren, B., and S. Larsson, Acta Chem. Scan& 17, 13. Robbins, W. E., J. N. Kaplanis, S. J. Louloudes and 1822-1824 (1963). R. W. Monroe, Ann. Entomol. S0c. Am. 53, 128-129 (1960). 5. Jarolimek, P., V. Wollrab, M. Streibl and F. Sorm, 14. Piek, T., J. Ins. Physiol. 10, 563-572 (1964). Chem. Ind. 237-238 (1964). 15. Faurot-Bouchet, E., and G. Michel, Bull. Soc. Chim. 6. Jarolimek, P., V. Wollrab and M. Streibl, Coll. Czech. Biol. 47, 93-97 (1965). Chem. Commun. 29, 2528-2536 (1964). 16. Ghouri, A. S. K., and J. E. McFarlane, Can. Ento- 7. Louloudes, S. J., D. L. Chambers, D. B. Moyer and mol. 90, 158-165 (1958). J. H. Starkey, Ann. Entomol. Soc. Am. 55, 442-448 (1962). 17. McFarlane, J. E., Can. J. Zool. 42, 645-647 (1964). 8. Romanuk, M., V. Herout and F. Sorm, Coll. Czech. 18. Hutchins, R. /7. N., and M. M. Martin, Lipids, in Chem. Commun. 24, 2018 (1959). press. 9. Sorm, F., V. Wollrab, P. Jarolimek and M. Streibl, 19. Carroll, K. K., J. Lipid Res. 2, 135-141 (1961). Chem. Ind. 1833-1834 (1964). 20. deVries, B., Chem. and Ind. 1049-1050 (1962). 10. Waldron, J. D., D. S. Gowers, A. C. Chibnall and S. H. Piper, Biochern, J. 78, 435-442 (1961). 21. Chang, T. C. L., and C. C. Sweely, J. Lipid Res. 3, 170-176 (1962). 11. Mold, J. D., R. E. Means, R. K. Stevens and J. M. 22. Metcalf, J. P., and A. A. Schmitz, Anal. Chem. 33, Ruth, Biochemistry 3, 1293-1298 (1964). 363-364 (1961). 12. Mold, J. D., R. E. Means, R. K. Stevens and J. M. Ruth, /bid. 5, 455-461 (1966). [Received July 17, 1967]

LIPIDS, VOL. 3, No. 3