Biochem. J. (1968) 109, 449Printed in Great Britain

Biosynthesis of Cyclopropyl Long-Chain Fatty Acids fromCyclopropanecarboxylic Acid by Mammalian Tissues in vitro

By W. G. DUNCOMBE AND T. J. RISINGThe Wellcome Re8search Laboratorie8, Beckenham, Kent BR3 3BS

(Received 1 April 1968)

1. Radioactivity from cyclopropane[14C]carboxylic acid is incorporated intofatty acids in vitro by rat and guinea-pig adipose tissue, by rat liver slices and bythe supematant fraction ofrat liver homogenate. 2. The labelled acids are differentfrom endogenous straight-chain fatty acids, and evidence is produced that theyconsist of a cyclopropyl ring in the c-position, the remainder of the chain beingbuilt up from C2 units (not derived from cyclopropanecarboxylic acid) in thenormal way via the malonate pathway. 3. It is suggested that these unnaturalacids have some metabolic effect related to the hypoglycaemic action of cyclo-propanecarboxylic acid.

Cyclopropanecarboxylate is an active oral hypo-glycaemic agent in the guinea pig and the monkey.In some other species, including the rat and man, itis much less effective, but in all species studied itproduces a marked synergistic effect with injectedinsulin (Stewart, 1962). The mechanism of its hypo-glycaemic action is obscure, and some of its otherbiochemical effects, both in vivo and in vitro, are

qualitatively different from those produced by a

number of other hypoglyeaemic agents (hypoglycin,tolbutamide, phenformin) (Stewart, 1962).

In view of the interactions between glucose andfatty acid metabolism in tissues, and their relation-ship to the control of glucose and fatty acid concen-trations in blood and of insulin sensitivity (Randle,Garland, Hales & Newsholme, 1964), it seemed thatinformation about the relation of cyclopropane-carboxylate to lipid metabolism might be of interestin connexion with its hypoglycaemic effect andsynergism with insulin. We started by investigatingthe metabolism of cyclopropanecarboxylate viafatty acids in some mammalian tissues in vitro, theresults of which are reported in this paper.

MATERIALS

Synthe8i8 of 14C-labelled cyclopropanecarboxylic acid.Cyclopropyl bromide was prepared by the reaction of silvercyclopropanecarboxylate with bromine, with dichlorodi-fluoromethane as solvent (Roberts & Chambers, 1951). Pre-paration of the Grignard reagent followed by carbonationwith 14CO2 gave cyclopropanecarboxylic acid labelled in thecarboxyl group. The methods used were similar to thosedescribed byRenk, Shafer, Graham, Mazur & Roberts (1961)for the preparation of the same acid labelled with 13C. In a

typical preparation the radiochemical yield was 76.5%,

starting with 3mc of 14CO2, and the acid had a specific radio-activity of 10.4mc/m-mole. Purity was established byinfrared spectroscopy, by thin-layer chromatography on

silicic acid with the ethanol-water-ammonia system ofBraun & Geenen (1962) followed by radioautography, andby gas-liquid chromatography of the ethyl ester, fractionsbeing collected for radioactivity determinations as describedbelow; the polyester column was for this purpose run

at 1400.Other labelled materials were obtained from The Radio-

chemical Centre, Amersham, Bucks. Avidin was fromNutritional Biochemicals Corp., Cleveland, Ohio, U.S.A.

METHODS

Fatty acid 8ynthesi8 from labelled precursors in rat andguinea-pig adipo8e ti88ue and liver in vitro. General pro-

cedures for incubating epididymal fat pads and for isolatingtotal fatty acids were as described by Duncombe (1968);similar methods were used for liver slices. About 20ml. ofKrebs-Ringer bicarbonate buffer (Krebs & Henseleit, 1932)was used/g. of tissue, with substrates and tracers as detailedin the Tables and Figures. When labelled cyclopropane-carboxylate, acetate or isobutyrate was used, fatty acidsfrom incubated tissue were isolated by the method ofDuncombe (1968). When labelled long-chain fatty acidswere used, total lipids of adipose tissue were extracted bythe procedure of Folch, Lees & Sloane-Stanley (1957), andthe neutral lipids in the extract (predominantly triglycer-ides) were purified by chromatography on silicic acid treatedwith .propan-2-ol-KOH, to remove contaminating labelledexogenous fatty acids (McCarthy & Duthie, 1962). Puritywas checked by thin-layer chromatography on silica gel G inhexane-ether-acetic acid (30:20:1, by vol.) and radio-autography. Methyl esters for gas-liquid-chromatographicanalysis were prepared from neutral lipid fractions or fromtotal tissue fatty acids by acid-catalysed methanolysis(Bowyer, Leat, Howard & Gresham, 1963), by treatment

449

W. G. DUNCOMBE AND T. J. RISINGwith boron trifluoride-methanol (deMan, 1967), or by reac-tion with diazomethane (Gellerman & Schlenk, 1965).Completeness of esterification was checked by thin-layerchromatography on silica gel G by the two-step develop-ment of Skipski, Smolowe, Sullivan & Barclay (1965).Radioactivity was determined in lipids by dissolvingweighed amounts in a toluene-based scintillator and count-ing in a Packard Tri-Carb liquid-scintillation spectrometer,followed by internal standardization and calculation ofabsolute radioactivities.

Oxidation of labelled sub8tratea by ti8aues in vitro. Incuba-tions were carried out by the same general methods asdescribed above, but with the apparatus and techniques ofMoss (1961), except that CO2 was trapped in a solution con-taining ethanolamine, and its radioactivity was determinedby liquid-scintillation counting (Jeffay & Alvarez, 1961).

Lipid 8ynthesi8 from labelled precursors in rat liverhomogenate fraction8. All operations before incubation werecarried out under ice-cold conditions. Liver was passedthrough a small stainless-steel tissue mincer into homogeniz-ing medium (Fletcher & Myant, 1961), about 3-5ml. ofmedium/g. of tissue being used. The suspension was homo-genized in a Potter-Elvehjem homogenizer and centrifugedat 800g for 10min. to remove tissue debris and nuclei, andthe supernatant was centrifuged at lOOOOg for 30min. tosediment mitochondria. The resulting supernatant (includ-ing the soluble fraction and microsomes) was designatedfraction SLo. The incubation medium used was the Fletcher& Myant (1961) optimum medium for fatty acid synthesis.Variations from this and additions of inhibitors and labelledsubstrates were as detailed in the Results section. Incuba-tions were carried out at 370 for 2 hr. under air, each tubecontaining lml. ofincubation medium and 2-5 ml. of fractionSlo. Blanks consisted of incubation medium+radioactivesubstrates only, and 2-5ml. of fraction Slowas added to eachof these after incubation. Lipid biosynthesis was measuredin the incubated samples by the method of Goldfine (1966),in which 0-Iml. samples were pipetted on to 25 mm. disks ofWhatman 3MM paper, dried and immersed successively inice-cold 10% (w/v) trichloroacetic acid (30min.), ice-cold5% (w/v) trichloroacetic acid (15min.) and ice-cold water(two washes each of 10min.). After being dried, the diskswere immersed in a toluene-based scintillator solution incounting vials and the radioactivity was determined. Forcomparing different conditions within one experimentobserved counts/min. were used, since the nature of thesamples precluded internal standardization. With the com-plete incubation system, the rate of lipid synthesis fromradioactive acetate was constant between 30 and 120min.,and at 180min. was more than 80% of the value expected atconstant rate.

Gas-liquid-chromatographic separation offatty acid methylesters. This was carried out with a Griffin Mk. III Katharo-meter Chromatograph, with Iin. x 6ft. columns with flashheaters. Two systems were used: (a) Celite (60-72 mesh)-diethylene glycol succinate (4:1) + 0-5% polyethylene glycol1500; carrier gas, H2 at 51b./in.2 or He at lOlb./in.2;column temperature 184°. (b) Celite (80-120 mesh)-ApiezonL grease (4:1); H2 at 15lb./in.2; column temperature 1970.Radioactive fractions were collected at 0-5 or 1min. intervalsby means of a Packard gas-chromatography fraction col-lector with glass cartridges (approx. 45mm. x 9mm.).Instead of the customary filling of silicone-coatedanthracene crystals, which permits the cartridge to be

counted directly in a liquid-scintillation counter, eachcartridge contained two cotton-wool cigarette filters, one ateach end. The one receiving the effluent from the gas chro-matograph was damped with a few drops of toluene, andafter collection of the fraction this filter was removed fromthe cartridge, the surrounding paper slit open and the com-plete filter dropped into a vial of toluene-based scintillatorfor counting. Lipids were readily extracted from the cottonwool into the scintillator solution and internal standardiza-tion for absolute measurements could be carried out ifrequired. This technique gives a greater counting efficiencythan the anthracene method and the efficiency of trappingcan be readily measured. As only about 1% of the radio-activity passed through the first filter and was trapped onthe second, the second was not counted. It was always putinto each cartridge, however, so that the trapping efficiencycould be checked when necessary, and also because it wasthought that its omission might decrease the trappingefficiency of the first filter by increasing the gas flow rate.

Hydrogenation of fatty acid methyl esters. Three methodswere used: (1) with Pd-CaCO3 in ethanol (Nunn, 1952),which reduces the double bond in cyclopropene rings withoutaffecting olefinic double bonds; (2) with PtO2 in methanol,which reduces both cyclopropene and olefinic double bonds(Gellerman & Schlenk, 1966); (3) with PtO2 in acetic acid,which in addition opens cyclopropane rings, giving pre-dominantly a mixture ofbranched-chain isomers (Gellerman& Schlenk, 1966; MeCloskey & Law, 1967).

Decarboxylation of 14C-labelled fatty acids. This wascarried out by the method of Brady, Bradley & Tramis(1960).

RESULTS

Fatty acid synthesis from cyclopropanecarboxylatein rat and guinea-pig ti88ues in vitro. The resultsshown in Table 1 demonstrate the incorporation ofcarbon from cyclopropane[14C]carboxylate intolong-chain fatty acids of rat adipose tissue in vitro.The incorporation increased considerably when glu-cose was also present in the medium, and in thisrespect was similar to the incorporation of acetateinto rat adipose-tissue fatty acids (Masoro, 1962).Moreover, the two compounds were incorporated toa similar degree (Table 1).To investigate the nature of the labelled fatty

acids formed from cyclopropanecarboxylate, theacids were methylated and separated by gas-liquidchromatography, and successive fractions of the ef-fluent were examined for radioactivity. Fig. 1 showsa typical gas chromatogram (b) and the correspond-ing radioactive scan (a) for material from rat adiposetissue incubated with cyclopropane[14C]carboxylateplus non-radioactive glucose. A notable feature wasthe appearance in the radioactivity scan of five un-usual labelled fatty acids (peaks A-E). The typicalpattern shown was found quite consistently in aboutten different experiments, and virtually identicalpatterns appeared also with samples from rat liverand guinea-pig adipose tissue. Close examination ofall the results showed that none of the radioactive

1968450

LIPOGENESIS FROM CYCLOPROPANECARBOXYLIC ACID

Table 1. Incorporation of cyclopropane[14C]carboxylate and [carboxy-14C]acetate intofatty acid8 of rat adipo8eti8ue in vitro

Experimental details were as given in the Methods section. Some flasks contained 16.7mM-glucose in additionto the radioactive substrate.

Expt. Radioactive substrate and conen.no. (mM)1 Cyclopropanecarboxylate (0 06)2 Cyclopropanecarboxylate (0 1)3 Cyclopropanecarboxylate (24)4 Cyclopropanecarboxylate (24)

5-11 Acetate (041)* Mean+ S.D.

Carboxyl C incorporated (pqtg.atoms/mg.of fatty acids)

Without glucose44, 49, 87

101, 79, 74

With glucose223, 274, 337354, 300, 370240, 300, 360380, 320, 400338+ 148* (7 expts.)

0

;a

Ca~~~~~~~~~

E

(b)

0

30 25 20 15 10 5 0

Elution time (min.)

Fig. 1. Gas-chromatographic separation of fatty acids (asmethyl esters) from rat adipose tissue incubated with cyclo-propane[14C]carboxylic acid. General conditions of incuba-tion and sample preparation were as described in theMethods section. Additions were: sodium cyclopropane-[14C]carboxylate (24mm; 7,uc in 20ml.); glucose (16 7mM).(a) Radioactivity; (b) mass-detector record.

p3aks coincided with the major endogenous tissuefatty acids (14:0, 16:0, 16:1, 18:0, 18:1, 18:2) or

with the positions of added straight-chain odd-numbered fatty acids. The pattern was unchangedwhen the fatty acid methyl esters were made by anyof the three methods described, or when helium

was used instead of hydrogen as the flow gas in thegas-liquid-chromatographic separations. It there-fore seems unlikely that the typical pattern was ananalytical artifact.The gas-liquid-chromatographic trace (Fig. lb)

showed no peak corresponding to even the mostradioactive fraction; the weight of the latter syn-thesized during the incubation was thus very smallcompared with the weight ofendogenous fatty acids.The question whether these new fatty acids

were synthesized exclusively from cyclopropane-carboxylate or whether other small molecules couldbe involved was investigated by incubating rat adi-pose tissue with [1-14C]acetate together with non-radioactive cyclopropanecarboxylate. The results(Fig. 2) showed that radioactivity appeared not onlyin the fatty acids expected to be synthesized fromacetate (e.g. 14: 0, 16: 0, 16: 1, 18: 0, 18: 1) but alsoin two positions (peaks D and E) correspondingexactly to peaks D and E of the radioactive cyclo-propanecarboxylate experiment (Fig. 1). Peakscorresponding to peaks A, B and C of that ex-periment would have been obscured by the radio-active fatty acids normallysynthesizedfromacetate.When radioactive acetate was incubated in the ab-sence of cyclopropanecarboxylate no radioactivitywas seen at positionsD and E. It is therefore evidentthat, for the biosynthesis of the fatty acids D andE(and possibly also acids A, B and C), both cyclo-propanecarboxylate and endogenous acetate arerequired.

Similar experiments with other labelled and un-labelled compounds were carried out in an attemptto show whether the cyclopropanecarboxylate wasmetabolically degraded before utilization for fattyacid synthesis. The radioactive acids D and E werenot produced when adipose tissue was incubatedwith labelled acetate plus unlabelled propionate orisobutyrate, or with labelled isobutyrate (in all casesin the presence ofglucose). Incubation with labelled

451Vol. 109

W. G. DUNCOMBE AND T. J. RISING

.~(a)

0~~~~~~~~~~~1:

o D

S~~~~~~~~~~~~~~~~~~

E

16:0.

(b)

30 25 2018:118:2 1:

'5

14:0

18:0

30 25 20 15 10 5 0Elution time (min.)

Fig. 2. Details were similar to those given for Fig. 1, but thetissue was incubated with [1-14C]acetate (10(M; 20,uc in20ml.) together with non-radioactive cyclopropane-carboxylate (0 5mm) and glucose (16-7mM).

Table 2. Incorporation of radioactivity from cyclo-propane[14C]carboxylate into lipid8 of 10OOOg 8uper-

natantfrom rat liver homogenateExperimental details were as given in theMethods section.

Each tube also contained cyclopropane[14C]carboxylate(11/lmoles; 11-5,uc). Other additions (quantity per tube)and omissions were as detailed below.

Radioactivity in lipids(counts/min.)

SystemExpt. 1 Expt. 2

Complete system

Avidin (150,kg.) added

Blank

Malonate omitted

HC03- omitted

Malonate and HCO3- omitted

353432123617897843919384504253

1093095909700

310280180

499050004630307030303950182015901510

laurate or palmitate plus unlabelled cyclopropane-carboxylate also failed to produce them. These re-sults suggest that cyclopropanecarboxylate is notconverted into propionate or isobutyrate before in-corporation into fatty acids, and also that the fattyacids that are formed do not arise by elongation ofexisting long-chain acids. The possibility that thecyclopropane ring is opened after incorporation intolong-chain fatty acids is not supported by thechanges in the radioactive acids caused by hydro-genation (see below).On chemical decarboxylation of the fatty acids

derived from adipose tissue incubated with cyclo-propane[14C]carboxylate, less than 5% of the radio-activity appeared in the carbon dioxide evolved.

Further evidence relating to the structure of thenew acids was sought by comparing the gas-liquid-chromatographic radioactivity patterns of the fattyacid methyl esters from adipose tissue incubationsbefore and after selective hydrogenation pro-cedures. Owing to the very small amounts ofmaterial available, it was not possible to measurehydrogen uptake quantitatively, so that only quali-tative conclusions could be drawn. The followingobservations were made. (1) Hydrogenation forcyclopropene double bonds produced no differencein the pattern obtained on the polyester column;

(2) reduction of olefinic double bonds (as confirmedby disappearance ofthe mass peaks for the endogen-ous unsaturated fatty acids) caused a diminutionofthe radioactive peakD (Fig. 1) and the appearanceof a faster-running radioactive peak ('peak X')having an elution time slightly longer than that ofpalnitic acid; (3) hydrogenation capable also ofbreaking cyclopropane rings again produced thepeak X, together with an increase in peak C (Fig. 1).With an increase in the time of hydrogenation peakX disappeared and there was substantial conversionof acid D into acid C (or into a compound having asimilar elution time). Changes in the other radio-active peaks of Fig. 1 were difficult to follow, owingto their smaller size. With non-polar columns similarqualitative changes were seen.The simplest explanation for the observed be-

haviour can be found by assuming that acid D is along-chain fatty acid containing one or more olefinicdouble bonds and one or more cyclopropane rings,but no cyclopropene ring. Reduction of the un-saturation would give acidX, containing a saturatedchain and a cyclopropane ring or rings, and furtherreduction, also breaking the rings, would give risepredominantly to a mixture of isomeric long-chainacids, bearing methyl-group side chains, appearing

452 1968

LIPOGENESIS FROM CYCLOPROPANECARBOXYLIC ACID

at position C. The present evidence is insufficient toshow whether this radioactive hydrogenation prod-uct appearing at peak C is identical with the com-pound found at peak C before hydrogenation.

Lipogene8i8 from cyclopropanecarboxylate in cell-free preparation8: mechani8m of bio8ynthe8i8. Sincewe usually observe much greater lipogenic activityin liver homogenates than in adipose-tissue hom-ogenates, on a basis of fresh tissue weight, we usedrat liver preparations to investigate the mechanismof biosynthesis from cyclopropanecarboxylate.

Table 2 shows the results of experiments underconditions known to influence fatty acid synthesisfrom acetate in this preparation (Vagelos, 1964).The requirement for HCO3- and malonate and theinhibitory effect ofavidin suggest that a mechanismsimilar to the malonate pathway is involved in theincorporation of cyclopropanecarboxylate intofatty acids.

Oxidation ofcyclopropanecarboxylate by rat adipo8eti88ue in vitro. The conversion of the carboxylcarbon of cyclopropanecarboxylate into carbondioxide was found to be about 1% of its incorpora-tion into fatty acids.

DISCUSSION

The incorporation of the radioactivity of cyclo-propane[14C]carboxylate into tissue fatty acidsraises three main questions. (1) In what chemicalform does the incorporation of the labelled carbonatom occur? (2) What is the mechanism involved?(3) Are these findings relevant to the hypoglycaemicaction of cyclopropanecarboxylate?

Chemical identification of the labelled biosyn-thetic acids was difficult because of the very smallamounts produced compared with the endogenousacids. The radioactive material isolated by gas-liquid chromatography was undoubtedly contami-nated by comparatively large quantities of non-radioactive material arising from tailing of themajor fatty acid peaks and from 'bleeding' of thestationary phase, making spectroscopic examina-tion of little value. Evidence about chemical struc-ture was therefore obtained indirectly by observingchanges in the chromatographic behaviour ofradio-active samples subjected to various hydrogenationprocedures.The results of such experiments show that the

cyclopropane-containing acid D also contains unitsderived from acetate, and at least part of the incor-poration of cyclopropanecarboxylate is due to amechanism stimulated by HCO3- and malonate andsensitive to avidin. This is presumably analogousto, or identical with, the malonate pathway for fattyacid synthesis from acetate. The enzymic forma-tion from cyclopropanecarboxylate ofa dicarboxylicacid, similar to the formation of malonate from

15

acetate, seems rather unlikely. Further, if the fattyacid chain were made up by the addition of acetateand cyclopropanecarboxylate units randomly, aconsiderable variety of species would be expected,their character depending on the number andposition of cyclopropyl groups included and thetotal chain length. In fact, only five acids wereobserved, and these appeared consistently in anumber of different experiments with differenttissues and animal species. The most plausiblemechanism for the formation of these acids is thatcyclopropanecarboxylate forms the start of thechain, analogous to the methyl end of a long-chainfatty acid synthesized from acetate, and that therest of the chain is made up in the usual way fromacetate units by the malonate pathway. Supportfor this as the major pathway is given by the findingof less than 5% ofthe total radioactivity ofthe acidsin the carboxyl group. By analogy with thenaturally occurring acids synthesized from acetateunits, it seems likely that all five acids contain thec-cyclopropyl group, differing only in unsaturationor the number of C2 units in the chain (or both).

If we assume that these acids contain one cyclo-propane group, some deductions can be made abouttheir chain length. In agas-liquid-chromatographicsystem similar to the one used here (polyestercolumn), the elution time ofa normal straight-chainfatty acid is shorter than that ofa monocyclopropylacid ofthe same carbon number but longer than thatofthe branched-chain acid resulting from hydrogen-ation of the cyclopropyl ring (Kaneshiro & Marr,1961). Consideration of the present results suggeststhat the major peak D might correspond to a C16atom fatty acid containing an w-cyclopropyl groupand probably also some unsaturation, for whichevidence from hydrogenation experiments has beenpresented. The other radioactive peaks would corre-spond to saturated or unsaturated (or both) even-numbered homologues.

Other possible modes of synthesis of these acidsseem to be much less likely. No elongation ofpalmitate or laurate by cyclopropanecarboxylatewas demonstrated. No mechanism involving degra-dation of the cyclopropyl ring before fatty acidsynthesis seems to occur, since labelled propionateor isobutyrate did not give rise to the novel acids;further, ring fission to give n-butyrate or two acetateunits does not occur since this would result only inndrmal even-chain-length acids, which were notdetected. Decarboxylation before incorporation isconceivable, since the radioactive label would belost and the cyclopropyl group would give rise tounlabelled acids; the very small conversion of thelabelled carboxyl group into carbon dioxide (1%)compared with its incorporation into fatty acidsmakes this unlikely as a major pathway.We are unaware ofany reports oflong-chain fatty

Biochem. 1968, 109

453Vol. 109

454 W. G. DUNCOMBE AND T. J. RISING 1968

acids containing an w-cyclopropyl group, eithernatural or synthetic. Naturally occurring cyclo-propyl and cyclopropenyl long-chain fatty acidsare well known in plants and micro-organisms, butthe ring is never terminal. In micro-organismstypical acids are cis-11,12-methyleneoctadecanoate(lactobacillate) and cia-9,1O-methylenehexadecano-ate, either or both of which have been found inlactobacilli, Eacherichia coli, Salmonella typhi-murium and other organisms (Chalk & Kodicek,1961; Kaneshiro & Marr, 1961; Macfarlane, 1962;O'Leary, 1962). In plant families (Sterculaceae andMalvaceae) unsaturated fatty acids such as 9,10-methyleneoctadec-9-enoate (sterculate) and 8,9-methyleneheptadec-8-enoate (malvalate) acid arecommonly found (Shenstone & Vickery, 1961). Inacids of both micro-organisms and plants the ringis believed to be formed by addition ofa C1 fragment(e.g. formate or methionine methyl) across an un-saturated bond (Liu & Hofmann, 1962; Chalk &Kodicek, 1961; Smith & Bu'Lock, 1964).Our finding that the cyclopropane ring is not

degraded during the incorporation of cyclopropane-carboxylate intQ animal-tissue fatty acids is con-sistent with the findings of other workers. Chung(1966) found no conversion into carbon dioxide ofthe cyclopropane methylene carbon in labelled cia-9,10-methylenehexadecanoate or c8-9,10-methyl-eneoctadecanoate (biosynthesized by Cloatridiumbutyricum) when these acids were incubated withrat liver mitochondria or whole-liver homogenates,though the carbon chain was shortened. Wood &Reiser (1965) reported the accumulation of cw- ortrana-3,4-methylenedodecanoate in the adiposetissue of rats fed with synthetic ci- or tran8-9,10-methyleneoctadecanoate respectively. Anotherhypoglycaemic agent, 2-amino-3-methylenecyclo-propylpropionate (hypoglycin), is rapidly metab-olized by rats in vivo, ultimately giving rise tomethylenecyclopropylacetate (Holt, 1966). Cyclo-propane itselfand its methyl, ethyl and vinyl ethersare biochemically inert and are eliminated un-changed (Williams, 1959). All these results suggestthat the cyclopropane ring is not readily broken inmammalian tissues.The present demonstration that cyclopropane-

carboxylate can lead to the synthesis of unnaturalfatty acids in the body does not throw any directlight on the mechanism of its hypoglyeaemic action.Such action of the cyclopropyl amino acid, hypo-glycin, has been attributed to the inhibition of fattyacid oxidation by its metabolite, methylenecyclo-propylacetate (Holt, Holt & Bohm, 1966). Also, itis known that cyclopropanecarboxylate can causeinhibition of fatty acid oxidation (Williamson &Wilson, 1965; Senior & Sherratt, 1967). One pos-sibility would therefore seem to be that cyclo-propanecarboxylate exerts its hypoglycaemic effect

through interference with the metabolism ofendogenous fatty acids by the unnatural acidsformed from it. Certainly the onset of hypo-glycaemia in sensitive species is slow (Stewart,1962), which is consistent with the active agent'sbeing a metabolite not rapidly synthesized. Ourfinding that synthesis of the unnatural acids issimilar in tissues both of rat and of guinea pig,whereas the hypoglycaemic response is very dif-ferent in the two species, indicates that if thispossibility is correct then considerable differencesmust exist between these two species as far as themagnitude of the metabolic interference is con-cerned. As yet there is no definite evidence for thesuggested mechanism or that the synthesis of thecyclopropyl long-chain acids has any physiologicalsignificance.

We are indebted to Mr A. R. Elphick for the preparationof cyclopropyl bromide.

REFERENCES

Bowyer, D. E., Leat, W. M. F., Howard, A. N. & Gresham,G. A. (1963). Biochim. biophys. Acta, 70,423.

Brady, R. O., Bradley, R. M. & Trams, E. G. (1960). J. biol.Chem. 235, 3093.

Braun, D. & Geenen, H. (1962). J. Chromat. 7,56.Chalk, K. J. I., & Kodicek, E. (1961). Biochim. biophys.

Acta, 50, 579.Chung, A. E. (1966). Biochim. biophys. Acta, 116, 205.deMan, J. M. (1967). Lab. Practice, 16, 150.Duncombe, W. G. (1968). Biochem. J. 106, 179.Fletcher, K. & Myant, N. B. (1961). J. Physiol. 155,498.Folch, J., Lees, M. & Sloane-Stanley, G. H. (1957). J. biol.

Chem. 226, 497.Gellerman, J. L. & Schlenk, H. <1965). J. Protozool.-12, 178.Gellerman, J. L. & Schlenk, H. (1966). Analyt. Chem. 88,72.Goldfine, H. (1966). J. Lipid es. 7, 146.Holt, C. von (1966). Biochim. biophys. Acta, 125, 1.Holt, C. von, Holt, M. von & Bohm, H. (1966). Biochim.

biophys. Acta, 125, 11.Jeffay, H. & Alvarez, J. (1961). Analyt. Chem. 33,612.Kaneshiro, T. & Marr, A. G. (1961). J. biol. Chem. 236,2615.Krebs, H. A. & Henseleit, K. (1932). Hoppe-Seyl. Z. 210,33.Liu, T. Y. & Hofmann, K. (1962). Biochemistry, 1, 189.McCarthy, R. D. & Duthie, A. H. (1962). J. Lipid Res. 3,

117.McCloskey, J. A. & Law, J. H. (1967). Lipids, 2, 225.Macfarlane, M. G. (1962). Biochem. J. 82, 40P.Masoro, E. J. (1962). J. Lipid Res. 3, 149.Moss, G. (1961). Int. J. appl. Radiat. Isotopes, 11, 47.Nunn, J. R. (19-52). J. chem. Soc. p. 313.O'Leary, W. M. (1962). Biochem. biophys. Res. Commuen. 18,

87.Randle, P. J., Garland, P. B., Hales, C. N. & Newsholme,

E. A. (1964). CibaFound. Colloq.: Endocrinology, 15, 192.Renk, E., Shafer, P. R., Graham, W. H., Mazur, R. H. &

Roberts, J. D. (1961). J. Amer. chem. Soc. 83, 1987.Roberts, J. D. & Chambers, V. C. (1951). J. Amer. chem.

Soc. 78, 3176.

Vol. 109 LIPOGENESIS FROM CYCLOPROPANECARBOXYLIC ACID 455Senior, A. E. & Sherratt, H. S. A. (1967). Biochem. J. 104,

56P.Shenstone, F. S. & Vickery, J. R. (1961). Nature, Lond.,

190, 168.Skipski, V. P., Smolowe, A. F., Sullivan, R. C. & Barclay, M.

(1965). Biochim. biophy8. Acta, 106, 386.Smith, G. N. & Bu'Lock, J. D. (1964). Biochem. biophy8.

Re8. Commun. 17, 433.

Stewart, G. A. (1962). Anglo-Germ. med. Rev. 1, 334.Vagelos, P. R. (1964). Annu. Rev. Biochem. 33, 139.Williams, R. T. (1959). Detoxication Mechani8ms, 2nd ed.,

p. 114. London: Chapman and Hall Ltd.Williamson, D. A. & Wilson, M. B. (1965). Biochem. J. 94,

19c.Wood, R. & Reiser, R. (1965). J. Amer. Oil Chem. Soc. 42,

315.