THE FATTY ACIDS OF BACTERIA

WILLIAM M. O'LEARY

Department of Microbiology and Immunology, Cornell University Medical College,New York, New York

I. Introduction .......................................... 421II. Isolation and Analysis of Fatty Acids .................................................. 422

A. Extraction of Fatty Acids from Bacterial Cells ....................................... 4221. General considerations............................................................. 4222. Extraction methods............................................................... 423

B. Separation and Identification of Bacterial Fatty Acids................................ 4241. Distillation........................................................................ 4242. Crystallization.................................................................... 4253. Countercurrent distribution ....................................................... 4254. Chromatography.................................................................. 4255. Infrared spectroscopy............................................................. 4276. Other procedures.................................................................. 4277. Evaluation........................................................................ 427

III. Fatty Acids of Bacteria ..................................... 428A. Saturated Straight Chain Acids...................................................... 428B. Hydroxy Acids ....................................... 431C. Branched Chain Acids .................................... 431D. Cyclopropane Acids ................... 431E. Unsaturated Acids ................... 432

IV. Fatty Acid Spectra of Some Representative Bacteria ...................... 433V. Chemical Nature of Bacterial Lipids .............................. 433

A. Free Fatty Acids ...................................... 434B. Fatty Acid Polymers ..................... ......................... 434C. Glycerides .......................................... 434D. Waxes ............................................ 434E. Phospholipids ....................................................................... 434F. Glycolipids............ 435G. Lipopeptides and Lipoproteins ................................ 435H. Bound Lipids ........................................ 435

VI. Intracellular Distribution of Fatty Acids ................................................ 435A. Cytoplasmic Inclusions ................................................ 435B. Cell Walls........................................................................... 436C. Protoplasmic Membranes ................ ................................ 436D. Protoplasts and L-Forms ................................... 436

VII. Aspects of Fatty Acid Biochemistry and Metabolism ...................... 436A. Degradation and Synthesis................................................ 437B. Stimulation and Inhibition of Growth................................................ 438C. Metabolism of Cyclopropane Acids................................................ 439D. Other Findings of Interest .................................. 440

VIII. Conclusion.............................................................................. 441IX. Addendum.. ........................................................................... 441X. Literature Cited ...................... .......................... 442

I. INT'RODUCTION monographs dealing with bacterial physiology toperceive the paucity of readily available informa-

For many years the accumulation of knowledge tion on bacterial lpids and on their componentregarding the chemistryand metabolism of lipids fatty acids. There appear to have been threehas lagged far behind the progress made in the main reasons for this lack of activity and progressstudies of carbohydrates and proteins. One has in the study of lipids. First of all, low water solu-only to examine any of the current textbooks or bility and other inconvenient physical character-

421

WILLIAM M. O'LEARY

istics of lipoid materials often make work withsuch compounds discouragingly difficult. Sec-ondly, until relatively recently few really satis-factory techniques were available for the isola-tion, purification, and identification of lipids andtheir components. Lastly, and possibly to someextent as a consequence of the first two reasons,there has been a tendency to regard the lipids asless interesting than other cellular constituents,both from the chemical and physiological pointsof view.Today a variety of techniques are available

that make work with the lipids less arduous andmore accurate than ever before. Also, it is nowrecognized that the lipids and fatty acids, andparticularly those of microorganisms, are muchmore chemically complex and metabolicallyactive than was suspected in the past. Conse-quently, in the years since World War II therehas been a progressively increasing tempo of re-search in lipid chemistry and metabolism, andseveral reviews attest to the fact that thisactivity has extended to the microorganisms(9, 11, 12, 46, 153). As these reviews indicate,recent years have seen research done on manyaspects of microbial lipids: synthesis and degrada-tion, intracellular distribution and function, im-munological properties, the natures of thevarious lipid complexes, and so on.However, essential to real understanding of

each of these aspects is an exact and detailedknowledge of those fatty acids which are thefundamental components of the bacterial lipids.Such knowledge is the sine qua non for furtheraccurate work on microbial fats, and as such isdeserving of compilation and study. Much re-search has been done on the fatty acids of bac-teria, and a considerable body of information hasbeen accumulated. Unfortunately, this informa-tion is scattered among many journals and mono-graphs in the world's chemical and biologicalliterature. It is, therefore, the purpose of thisreview to collect and summarize the presentlyavailable knowledge regarding bacterial fattyacids, including their chemical nature, certainbiochemical considerations, and the methods nowin use for the study of these compounds.Those familiar with the subject of microbial

fats will note that little is here reported on thelipids of the mycobacteria. As is well known, theseorganisms contain an astonishing variety of lipidconstituents, a far more complex assortment than

one finds in other bacteria. Indeed, the study ofmycobacterial lipids could well constitute a sub-specialty in itself. Fortunately the vast body ofliterature on this complicated subject has beenexcellently and exhaustively reviewed elsewhere(5, 9, 11, 12, 135). Therefore I have intentionallyomitted all but a few references to the myco-bacteria and have limited this review to con-sideration of the fatty acids of less well surveyedbacteria, particularly species of the ordersEubacteriales and Pseudomonadales.

II. ISOLATION AND ANALYSIS OF BACTERIALFATTY ACIDS

A. Extraction of Fatty Acids from Bacterial Cells

1. General considerations. To evaluate betterthe data available in this field and to avoid someof the pitfalls involved in this type of research,it is important to be aware of several character-istics peculiar to bacterial fats.

Of primary importance is the fact that thefatty acid composition of bacteria is markedlyaffected both quantitatively and qualitatively bythe nature of the medium and by the conditionsunder which the culture is grown. Many authorshave commented in detail on this phenomenon(12, 57, 117, 134, 153). Among the various nutri-tional and environmental factors which theyhave reported to influence the fatty acid contentof bacteria are the amounts and relative propor-tions of acetate, glycerol, carbohydrate, lipid,and nitrogenous substances present in the me-dium; oxygen supply; pH; and the age of theculture when harvested. Unawareness of the pro-found effects that these factors can have on fattyacid composition has led to considerable confu-sion over the many inconsistencies and apparentcontradictions that have appeared in the litera-ture on bacterial fats. For many years thesevariations in fat content seemed largely randomand unpredictable, and certainly there is still agreat deal to be learned. However, as the body ofknowledge regarding bacterial metabolism in-creases, many of the vagaries in the fat contentof bacteria grown under different conditions arebecoming more understandable. Conversely, incertain cases careful study of the variations inthe fatty acid contents of cells grown under dif-ferent but carefully defined conditions has led toincreased understanding of metabolic processes.An example of this is the work of Hofmann and

422 [VOL. 26

FATTY ACIDS OF BACTERIA

his colleagues (61, 68) in which the fatty acidcontents of different media were correlated withthe fatty acids found in the bacteria grown ineach medium to obtain an insight into the bio-synthesis of lactobacillic acid. The accuracy ofthe predictions based on this work was laterverified by other procedures (95, 112, 113).From the foregoing it is obvious that a de-

tailed knowledge of the medium and culture con-ditions employed is essential to any valid inter-pretation of data regarding bacterial fats, andthat, to be comparable, successive experimentsmust duplicate conditions exactly. This meansthat particularly for initial studies of an organisma chemically defined and exactly reproduciblemedium is necessary for meaningful results. Itshould be borne in mind, however, that syntheticmedia may differ markedly from the natural en-vironments of the organisms being studied, andthat this difference may affect the compositionof the bacterial cells. Asselineau and Lederer (12)and Lovern (97) have pointed out that the lipidcontents of such "artificial bacteria" can be quitedissimilar to those of the same species grown innatural surroundings. For example, it has beenobserved that cells of Mycobacterium tuberculosisgrown in lung tissue differ from cells of the samestrain grown in laboratory medium with respectboth to relative lipid content and to the utiliza-tion of various fatty acids (127). Therefore, aftercarefully studying the fatty acids of any givenmicroorganism grown on synthetic media, it maybe desirable to proceed to further studies of theorganism grown under more nearly natural con-ditions or, in the case of the pathogens, to orga-nisms grown in vivo.Another important problem involved in study-

ing bacterial fatty acids is the difficulty one en-counters in extracting all of the fatty acids presentin the cell. In every bacterial species some per-centage of the fatty acids occurs free and can bereadily removed from the cells simply by extract-ing with an appropriate organic solvent. Thepercentage of free fatty acids varies with thespecies and may be quite high (83). However, asignificant proportion of the fatty acids is boundvery firmly in various protein and carbohydratecomplexes which are not soluble in organic sol-vents. Consequently, merely extracting bacterialcells with some lipid solvent will fail to recoverconsiderable portions of fatty acids. To obtainall the fatty acids present, it is necessary first to

subject the cells to some hydrolytic procedurewhich will free the so-called "bound" fatty acidsso that they can be removed by the extractingsolvent. Studies of bacterial fats in which thisfact was not taken into consideration are oflimited value, since they of necessity consideronly those acids which occur free in the cell.There is no evidence to show that the free acidsare quantitatively or even qualitatively repre-sentative of the total cellular content of fattyacids.One further circumstance worthy of mention

in any general discussion of bacterial fats is thefact that, with the notable exception of Azoto-bacter chroococcum (128), none of the true bac-teria have been shown conclusively to containsteroids. In fact, most bacteria contain littleunsaponifiable material of any kind. The absenceof steroids in bacteria is particularly interestingin that various species have been shown to syn-thesize or to require for growth mevalonic acid(130, 137, 138), which is known to be a sterolprecursor in other forms of life.

2. Extraction methods. In attempting a carefulstudy of bacterial fats, the investigator is facedwith somewhat of a dilemma. If extraction con-ditions are too mild, he will not recover all theconstituents present; if, on the other hand, con-ditions are too rigorous, he may damage ordestroy some of the more labile constituents. Thisis especially a problem when the object is tostudy specific lipid complexes. Fortunately, thefatty acids are relatively resistant to the hy-drolytic procedures necessary to insure theircomplete removal from the cell. Nevertheless, ex-posure to any extreme conditions should be mini-mized to prevent alterations of the more labileacids such as the unsaturated and ring-containingcompounds. Exposure to oxygen is particularlyto be avoided and is prevented whenever possibleby conducting procedures in an atmosphere ofnitrogen. This precaution should include eventhe purging of solvents with nitrogen to removedissolved oxygen.

Several references are available (26, 38, 58, 81,119, 132) which discuss in much greater detailthan is possible in the present paper the generalprinciples and methods of lipid and fatty acidchemistry including extraction procedures, andthese should be consulted for backgroundinformation.The extraction techniques employed for the

4231962]

WILLIAM M. O'LEARY

fatty acids of bacteria are typified by the methodemployed by Hofmann et al. (62) and by manyothers with generally satisfactory results. In thisprocedure, dried bacterial cells are suspended in2 N sulfuric acid and hydrolyzed by autoclavingfor 90 min. After the hydrolyzate has cooled andthe insoluble cell debris has been removed byfiltration, the crude lipoid material is extractedfrom the hydrolyzate with ether. The lipoid ma-terial is then saponified with potassium hydroxidein ethanol to form water-soluble potassium saltsof the fatty acids. Water is added, and the un-saponifiable material present, if any, is removedby a second ether extraction. Finally, the aqueousphase, now containing only salts of fatty acids,is reacidified to convert the fatty acid salts backto free fatty acids. The acids are then recoveredby ether extraction. Evaporation of the etheryields the mixed fatty acids which had beenpresent in the bacterial cells.

In another method used successfully by severalworkers, the hydrolysis is accomplished usingethanolic potassium hydroxide instead of sulfuricacid (61). This both hydrolyzes and saponifies inone step, producing potassium salts of the cellularfatty acids which are then handled as describedabove. One advantage of this method is thesaving in time as a consequence of fewer manipu-lations. However, an even more important ad-vantage has recently been pointed out by Law(87), who, by applying various extraction pro-cedures to Escherichia coli, found that acidhydrolysis sometimes produced artifacts in therecovered fatty acids. He found that the pro-cedure that caused the least alteration of fattyacids was the use of methanolic potassium hy-droxide for hydrolysis followed by extraction ofunsaponifiable material, acidification, and re-covery of the mixed acids.

It must be borne in mind that the methodsjust described are for the recovery of total cellu-lar fatty acids only, without regard to the com-plexes in which they occur. Naturally, for therecovery of specific lipid complexes such as theglycerides, phospholipids, glycolipids, waxes, andso on much more gentle and selective proceduresmust be employed. The classic example of lipidclass fractionation in bacteria is the elegantscheme originally devised for the study of myco-bacterial lipids (5, 12).

B. Separation and Identification of BacterialFatty Acids

As mentioned above, the inability readily andreliably to separate and identify fatty acids wasfor many years a major reason for the limited ac-tivity and progress in lipid research. Fortunately,today a number of techniques are availablewhich, in various combinations with each otherand with the standard methods of analyticalorganic chemistry, make possible quite accurateanalyses of fatty acid mixtures. Methods forseparation and identification will be discussedtogether in this review since in many cases thecharacteristics which make it possible to separateone fatty acid from others are also some of thecharacteristics by which it can be identified.

1. Distillation. Highly efficient fractional dis-tillation can be used to separate fatty acids onthe basis of the lengths of their carbon chains. Itis usual to convert fatty acids to lower boilingmethyl esters and to perform the distillations invacuo to use as low temperatures as possible. Thevarious types of apparatus suitable for this useand their operation have been described byBowman and Tipson (18) and Murray (105).The temperature at which each component is

obtained indicates the number of carbon atomspresent in the molecules being distilled. However,one cannot distinguish between saturated andunsaturated compounds having the same chainlength. Thus each fraction must be checked byother techniques to determine whether more thanone compound is present. Because the generalshape of the distillation curve can be predicted,any unusual departure from the expected curvecan aid in the detection of new or unusual com-pounds. It was in this way that cyclopropanering-containing fatty acids were first discoveredin bacterial lipids (65, 66).

In addition to being useful for the separationand analysis of fatty acids, distillation is also avaluable means of obtaining amounts of indi-vidual fractions large enough to be used forfurther chemical and biological purposes. Mostof the other methods to be described below yieldamounts that are impractically small for anyfurther manipulation.

Paradoxically, however, its relatively largescale is the major disadvantage of distillation inthe study of bacterial fatty acids. That is, thesize of the fatty acid sample required is quiteappreciable in terms of the amount of cells that

424 [VOL. 26

FATTY ACIDS OF BACTERIA

must be amassed. Because its sample require-ments are relatively modest in comparison withthose of other distillation assemblies, the semi-micro spinning band still, particularly the Piros-Glover model, has been preferred for this type ofwork. The use of this device for the analysis ofbacterial fats has been described in detail byLucas (98). However, even for the spinning bandapparatus, one still must have from 1 to 10 g ofmixed fatty acid esters. Since the total lipidcontent of most bacteria is less, usually muchless, than 10% of the dry cell weight, such sam-

ple requirements entail the collection of largeamounts of bacteria. Though this may be reason-

able for preparative purposes, it is obviously notconvenient for routine analyses.

2. Crystallization. For many years a standardprocedure in fatty acid chemistry has been theseparation of saturated acids from unsaturatedacids by differential crystallization of their leadsalts (81, 119). Today, increasing use is beingmade of low temperature fractional crystalliza-tion of free fatty acids (20). With this procedureit is possible to achieve quite satisfactory separa-

tions of saturated from unsaturated acids, and toresolve further each of these two major groups.

Low temperature crystallization procedures havethe considerable advantage of simplicity, re-

quiring only facilities for working at -20 to-70 C and appropriate solvents. Also, there islittle danger of any chemical alteration of fattyacids occurring during the crystallization opera-

tions. There is, however, the problem of possiblecocrystallization of similar compounds which can

make precise separations difficult in some cases.

Earlier methods required considerable amountsof starting material, but techniques have now

been devised which require as little as 20 mg ofsample (41, 139).Another useful crystallization technique in-

volves the formation of urea inclusion complexesof mixed fatty acids (125). These inclusion com-

plexes, or adducts, can be separated by crystalli-zation at higher temperatures than the free fattyacids (i.e., around 0 C). After fractionation, theadducts of the individual fatty acids can be easilydecomposed by warming, thus liberating the un-

altered fatty acids. In addition to resolving fattyacids with respect to degree of unsaturation,urea adducts are useful in separating straightchain compounds from those with branchedchains.

3. Countercurrent distribution. Extensive de-scriptions of countercurrent distribution and ofits application to fatty acids can be found inpapers by Craig and Craig (29), Dutton (35),and Ahrens (1). In essence this method consistsof subjecting a sample to a large number of suc-cessive liquid-liquid extractions in a multiple-tube apparatus which permits one of the solventsto advance to the next tube of the series inde-pendently of the other solvent. The componentsof the sample are distributed according to theirrelative solubilities in each solvent.The separation of fatty acids by countercurrent

distribution, as well as other liquid-liquid parti-tion procedures, is affected by both the chainlength and the degree of unsaturation of thecompounds in the sample. The presence of onedouble bond has the same solubility effect as doesshortening the chain by two carbon atoms. Thusa monounsaturated 18-carbon acid will migratein the same way as the saturated 16-carbon acid.Aside from this problem for which there are ade-quate remedies including crystallization pro-cedures, countercurrent distribution can separatecomplex mixtures into their various constituentsin high purity. Further, there are the advantagesof accomplishing this separation under themildest of conditions and of being able to obtainrelatively large amounts of sample componentsshould the need arise. Countercurrent distribu-tion does have the disadvantages of requiringcomplicated and expensive equipment and ofyielding a large number of individual liquid frac-tions which must be further manipulated.

4. Chromatography. A surprisingly large num-ber of chromatographic methods have been de-vised for fatty acids, and these have been re-viewed periodically by various authors (8, 15, 17,74, 90). These techniques, which vary widely inapplication and effectiveness, for present pur-poses may be divided into three groups: adsorp-tion chromatography, partition chromatography,and gas chromatography.

Adsorption techniques using charcoal, silicicacid, or alumina are widely used for the separa-tion of lipid complexes (38). However, suchmethods have not proved as successful in theanalysis of the fatty acids themselves, and conse-quently they are not as commonly employed forthis purpose as are partition and gas chromato-graphic techniques. One variation of adsorptionchromatography that has been successfully used

42519621

WILLIAM M. O'LEARY

for fatty acid analysis is the displacement tech-nique employed by Holman (74) and others. Twonew methods that appear quite promising are thechromatography of fatty acid esters on glass fiberpaper impregnated with silica gel (52) and theso-called "thin-layer chromatography" whichhas been used in Europe for some time and is nowfinding enthusiastic supporters in this country(152, 154).

Partition chromatography both in columnsand on paper has been very useful for fatty acidresearch, and this is particularly true of reversed-phase procedures. Boldingh's basic technique(16), using powdered rubber as a support for thenonpolar stationary phase and acetone-watermixtures as eluting solvents, has found manyapplications. A modification of this method de-vised by Hofmann and his associates (62) hasgiven excellent results in the study of the fats ofmany bacteria (62, 68, 112, 113). Unsaturatedacids in the sample are hydroxylated so that theycan be separated from the shorter saturated acidswith which they would otherwise be eluted. Theeluate is monitored by titration of each fraction.Another useful variation of Boldingh's techniqueis that of Hirsch (59) in which methyl estersrather than the free acids are passed throughrather lengthy columns. In this procedure theeluate is continuously monitored by a recordingdifferential refractometer.

Reversed-phase paper chromatography isuseful when dealing with very small samples.Unfortunately, it is necessary to impregnate thepaper with various materials such as latex (16),hydrocarbons (7, 81), or silicones (126), whichintroduce other variables into an already com-plex operation. Spot development is also morecomplicated than in other types of paper chroma-tography (7, 124).

All the chromatographic methods mentionedso far, although exhibiting the same order ofefficiency customarily associated with otherchromatographic procedures, generally fail toseparate an unsaturated acid from a saturatedacid having two less carbon atoms (e.g., oleicacid cannot be separated from palmitic acid).For this reason, it is necessary to remove onegroup of acids before chromatographing theother, or to alter chemically the unsaturatedacids so that they can be resolved. One suchchemical alteration is the hydroxylation em-ployed in Hofmann's reversed-phase column

chromatography described above. Several othermethods have also been devised (38). If the aimof the investigator is solely analysis of a fattyacid mixture, such chemical alterations pose noproblem. But, if further use is to be made of theindividual fatty acids, chemical changes may beundesirable.

Gas-liquid chromatography is actually anotherform of partition chromatography. This tech-nique, first introduced by James and Martin(78), has been widely and enthusiasticallyadopted by lipid chemists as the most nearlyideal analytical method now available for fattyacids. A number of reviews have appeared dis-cussing the principles of this technique and inparticular its application to fatty acid analysis(10, 54, 77, 94, 115).In this method, methyl esters of fatty acids

are introduced in minute amounts into a heatedcolumn containing the nonpolar stationary phasecoated on an inert support such as Celite or evensimply the column wall. The esters are vaporizedand moved through the column by a stream ofgas such as nitrogen or helium. As the individualesters emerge from the column at times reflectingtheir affinities for the stationary phase, they passthrough a detection device which is connected toa recorder. The time after injection of sample atwhich each peak appears on the recorder identi-fies the ester, and quantitative data can be ob-tained from the heights or areas of each peak. Theindividual esters can be recovered after leavingthe detector. Several types of detectors and manycolumns are available, each with varying applica-tions and efficiencies (54, 116).Very high efficiencies can be obtained with gas

chromatographic apparatus which not onlyfacilitate the detection of trace amounts of fattyacids, but also make it possible to separate un-saturated acids from saturated acids withoutpreliminary chemical alterations.The advantages of this method are: very small

sample requirements, speed and ease in perform-ing an analysis that is simultaneously quantita-tive and qualitative, and recovery of unalteredsample constituents. The major disadvantage,which is consequent to a small sample size, isthat only minute amounts of each constituentcan be recovered. Thus, gas chromatography isprimarily an analytical method, although pre-parative scale gas chromatography can be per-

426 [VOL. 26

FATTY ACIDS OF BACTERIA

formed in which capacity is increased at theexpense of efficiency.

5. Infrared spectroscopy. In recent years thistechnique has become one of the major tools offatty acid analysis (104, 148). The examinationof infrared absorption spectra has three mainapplications: in determining the details of mo-lecular structure of a compound, in detecting thepresence of unidentified substances even in com-plex mixtures, and in identifying individual com-pounds by comparing their spectra with those ofknown compounds. This method has a numberof advantages including the fact that a con-siderable amount of information can be obtainedwithout affecting the sample (which consequentlycan be recovered and used for other purposes), theability to use a very small sample (as little as 0.5mg) if necessary, and the ability to differentiatereadily between cis and trans isomers, which isof prime importance in the study of fatty acidscontaining double bonds or cyclopropane rings.

It has become common practice to apply infra-red spectroscopy to gas chromatographic frac-tions as a means of obtaining a maximum ofinformation from amounts of material too minutefor any appreciable treatment by conventional"wet" analytical procedures.

6. Other procedures. In addition to thosemethods already described, there are severalother instrumental procedures that are in use orthat are now coming into use for special purposesin the analysis of fatty acids. These proceduresinclude X-ray diffraction, ultraviolet spectros-copy, microwave analysis, nuclear magneticresonance, and mass spectroscopy. Space limita-tions preclude any detailed description of theserather complex methods here, but they have beenrecently and thoroughly discussed by O'Connor(109) and by Ryhage and Stenhagen (120).In any study in which one is attempting to

identify fatty acids, the techniques described sofar should, of course, be used in conjunction withthe classical methods of lipid chemistry and or-ganic analysis to arrive at an unequivocal identi-fication (50, 53, 58, 144). These methods include,among many, determinations of the variousphysical properties for both the fatty acid asisolated and for appropriate derivatives; ele-mental analysis; determinations of neutralequivalent; iodine number; hydrogenation up-take; and identification of the hydrogenationproduct. In studying compounds containing

double bonds, cyclic configurations, and otherlabile structural features, the molecule should bedegraded and the resultant fragments identifiedto perceive the structure of the original molecule.In many cases microbiological assay procedurescan be used to advantage for both quantitativeand qualitative studies of fatty acids (61, 70).Whenever possible, it is best to compare the

various properties of the fatty acids under studywith those determined simultaneously for highpurity preparations of known compounds (i.e.,"authentic samples") rather than to rely exclu-sively on values appearing in the literature.The ultimate proof of structure is accomplished

by synthesizing with certainty the suspectedmolecular configuration and demonstrating thatthe synthetic compound is identical with thecompound originally isolated (42, 49). This pro-cedure is particularly important in the study ofnew or unusual fatty acids when authentic sam-ples are not available.

7. Evaluation. The particular methods ofseparation and identification that should be em-ployed in any given study depend on many dif-ferent considerations. Prominent among theseare: the amount of sample available, whether ornot significant amounts of each sample compo-nent must be recovered, whether some chemicalalteration of some components is acceptable, theinformation that each method used is capable ofsupplying, and, most important, the amount ofinformation that is necessary to establish un-equivocally the identity to be reported for anygiven compound.The importance of this last consideration

cannot be stressed too much since the lack ofsufficiently rigorous criteria in the identificationof fatty acids has impaired the value of much ofthe literature in the field. For example, there aremany reports in the older literature of the occur-rence of oleic acid in various species of bacteria.In many if not most cases, the "identification"of this compound was based on insufficient datararely exceeding a neutral equivalent, determina-tion of hydrogen uptake, and perhaps the identi-fication of the hydrogenation product as stearicacid. Actually, even with the a priori assumptionthat the substance under study was a straightchain, monocarboxylic fatty acid, such informa-tion only indicated that the compound could bean 18-carbon acid with one double bond. Thisdescription is equally applicable to several posi-

42719621

WILLIAM M. O'LEARY

tional isomers of octadecenoic acid, only one ofwhich is oleic acid (i.e., cis-9-octadecenoic acid).More elaborate investigations in which the loca-tion of the double bond was established withcertainty have shown that in many bacteria thesole or major unsaturated fatty acid is cis-vaccenic acid in which the double bond is in the11-12 position rather than in the 9-10 position asis the case in oleic acid. This is not to say thatoleic acid does not occur in bacteria, but it doespoint out the necessity of excluding other possiblecompounds before making a specific identifica-tion. The literature records a number of instancesin which fatty acids have been incorrectly identi-fied or in which the opportunity to discover anunusual and important compound has beenmissed because of insufficiently critical evaluationof analytical data.

It is, therefore, essential that the investigatorgive constant consideration both to the criteriathat must be satisfied for the unequivocal identi-fication of any given compound, and to the limi-tations as well as the capabilities of each pro-cedure he is employing.

III. FATTY ACIDS OF BACTERIA

Much of the earlier work on bacterial lipidswas limited to determinations of "per cent totallipid" for various organisms. Because figures forany species can vary considerably depending onthe strains studied and the culture conditionsemployed, such data are of limited value andsignificance. However, from these studies onemay make the general observation that, with afew exceptions, the total lipid content in mostspecies of bacteria ranges between 1 and 10% ofthe dry cell weight (83, 117).Of considerably more interest and utility are

reports of the chemical composition of bacteriallipids, particularly with respect to the com-ponent fatty acids. Porter (117) has reviewed thework in this field up to 1946, and further infor-mation on bacterial fatty acids has been in-cluded in several more recent lipid reviews (9, 11,12, 60, 153). Of parallel interest are the reviewson the lipids of the yeasts (36) and of the molds(40).One purpose of the present review is to cata-

logue and to describe briefly those fatty acidswhich, on the basis of reasonably satisfactoryevidence, are believed to occur in bacteria. Thereader should bear in mind that the same amountof analytical information has not been accumu-

lated for each of the fatty acids to be listed. Theidentities of most of these compounds has beenconclusively established, but some still requirefurther study and may be subject to revision inthe light of later findings. Also, with the adventof the newer analytical methods, the number offatty acids known to occur in bacteria has beenincreasing rapidly. For this reason, it is probablethat by the time this review appears in print thecatalogue of bacterial fatty acids which it con-tains will already be incomplete.The fatty acids to be listed are grouped ac-

cording to their chemical natures. Their struc-tural formulas are given in Table 1.

A. Saturated Straight Chain Acids

These compounds have long been known tooccur in bacteria and are identical to the satu-rated fatty acids found in other forms of life.

Acids having chains of less than 12 carbonatoms have been detected in small amounts invirtually all species in which they have beensought. Included in this group are the followingacids: formic (C1), acetic (C2), propionic (C3),butyric (C4), caproic (C6), caprylic (C8), andcapric (C1o).The higher fatty acids constitute a much larger

proportion of the total fatty acid content of thecell. The acids included in this group are lauric(C12), myristic (C14), palmitic (C16), stearic (C18),arachidic (C20), behenic (C22), lignoceric (C24),and octacosanoic (C28). No straight chain satu-rated acids longer than C2s have been reportedother than in the mycobacteria. Palmitic acidoccurs more frequently and usually in largeramounts than any other saturated acid found inbacteria. Stearic, myristic, and lauric acids arealso common though lesser constituents of mi-crobial lipids, whereas the C20.28 acids areencountered only in a few species. Arachidic acidhas been reported in Pseudomonas aeruginosa(117), behenic in Corynebacterium diphtheriae (6),lignoceric in C. diphtheriae (3, 6) and Lactobacillusacidophilus (30), and octacosanoic in C. diph-theriae (6).

It will be noted that the higher acids listedabove all contain even numbers of carbon atoms,as is usually expected in fatty acids extractedfrom biological materials. However, there havebeen a few reports of fatty acids with odd num-bers of carbon atoms occurring in bacteria. Intheir study of fatty acids recovered from culturesof P. aeruginosa, James and Martin (79) found

428 [VOL. 26

FATTY ACIDS OF BACTERIA

TABLE 1. Fatty acids of bacteria

Common name Systematic name Catoms Formu

Higher straightchain satu-rated acids

LauricMyristicPalmiticStearicArachidicBehenicLignocericMontanic

Hydroxy acids- Hydroxy-butyric

Mevalonic(o -a - Dihy-hydroxy -, -methyl-valeric)

- Hydroxy-myristiC

Di - hydroxy-stearic

Corynomycolic

Corynomyco-lenic

Corynolic

DodecanoicTetradecanoicHexadecanoicOctadecanoicEicosanoicDocosanoicTetracosanoicOctacosanoic

3-Hydroxybutanoic

3,5 - Dihydroxy - 3 -methyl pentanoic

- Hydroxyoctanoic(3 - hydroxyocta-noic)- Hydroxydecanoic(3 - hydroxydeca-noic)

13 - Hydroxydodeca-noic (3 - hydroxy-dodecanoic)

3 - Hydroxytetra-decanoic

Di - hydroxyocta-decanoic

1214161820222428

4

6

8

10

12

14

18

32

32

52

CH3-(CH2)10-COOHCH3-(CH2)12-COOHCH3-(CH2)514-COOHCH3-(CH2)1-COOHCH3-(CH2)18-COOHCH3-(CH2)20-COOHCH3-(CH2)22-COOHCH3-(CH2)26-COOH

CH3-CH(OH)- CH2-COOH

OH

HO-CH2-CH2-C CH2-COOH

CH3

CH3-(CH2)4-CH(OH)-CH2-COOH

CH3-(CH2)6-CH(OH)-CH2-COOH

CH3-(CH2)8--CH(OH)-CH2-COOH

CH3-(CH2)0o-CH(OH)-CH2-COOH

C17H53 (OH)2COOH

CH3-(CH2)14-CH(OH)-CH-COOH

U14H29

CH3-(CH2) 6-CH=-CH-(CH2)7-CH(OH)-CH-COOH

U14H29

CH3-CH(OH)-(CH2)7-CH CH(OH)-CH CH-CH3

CH3 CR3

CH3-(CH2)14-CH-(CH2)17-C COOHHCH3 Hi

1962] 429

WILLIAM M. O'LEARY

TABLE 1.-(Continued)

Common name Systematic name atoms Formula

Branched chainacidsIsooctanoic

CorinnicDiphtheric

Cyclopropaneacids

Lactobacillic

Unsaturatedacids

PalmitoleicPalmitvac-

cenic

Oleiccis-Vaccenic

Methylheptanoic

6-Methyloctanoic

13 - Methyltetra-decanoic

15 - Methylhexa-decanoic

Methylene - hexa-decanoic

cis - 11,12 - Methy-lene - octadecanoic

TetradecenoicHexadecenoiccis-9-Hexadecenoiccis-11-Hexadecenoic

Octadecenoiccis-9-Octadecenoiccis-11-OctadecenoicEicosenoicHeneicosenoicDocosenoicTetracosenoicOctacosenoic

8

9

15

17

3535

17

19

14161616

1818182021222428

CH3-(CH2)X-CH (CH2),-COOH (x + y = 4)

OH3

CH3-CH2-CH--(CH2)4-COOH

CH3

CH3-CH-(CH2)11-COOH

CH3

CH3-CH-(CH2)13-COOH

CH3

C35H6802C3sH6802

H H

CH3-(CH2). C (CH2)y-COOH (x+ y = 12)

C-- C

H H

H H

CH.3-(CH2)5 C (CH2)9-COOH

C ~C

H H

CH3-(CH2)t-CH=CH-(CH2),-COOH (x + y = 10)CH3-(CH2).--CH=CH-(CH2)y-COOH (x + y = 12)CHz3-(CH2)65-CH=CH-(CH2)7-COOHCH3-(CH2)3-CH-CH-(CH2)9-COOH

CH3-(CH2).--CH=-CH-(CH2)y--COOH (x + y = 14)CH3-(CH2)7-CH-=CH-(CH2)7-COOHCH3-(CH2)r5-CH=CH-(CH2)s--COOHCH3-(CH2)5--CH=CH-(CH2)y-COOH (x + y = 16)CH3-(CH2),-CH-CH-(CH2)y-COOH (x + y = 17)CH3-(CH2),-CH=CH-(CH2)y-COOH (x + y = 18)CH3-(CH2)1-CHCH-(CH2)y-COOH (x + y = 20)CH3-(CH2)1-CH-CH-(CH2)y-COOH (x + y = 24)

430 [VOL. 26

FATTY ACIDS OF BACTERIA

evidence of C15 and C17 straight chain saturatedacids. Goldfine and Bloch (45) noted what maybe a C15 acid of this type in Clostridium butyricum.None of these compounds has yet been identifiedwith certainty.

B. Hydroxy Acids

,3-Hydroxybutyric acid has been found inseveral bacteria including species of Azotobacter(39, 92), Chromobacterium (39), Bacillus (91, 146,151), Pseudomonas (39), and Micrococcus (131).Mevalonic acid (f3-5-dihydroxy-,3 methyl-

valeric acid) has been isolated from cultures oflactobacilli and other microorganisms (137, 138).Since it is known that some bacteria require onlyvery small amounts of this compound for maxi-mal growth, it may well be that it has escapeddetection in many organisms.

,B-Hydroxydecanoic acid occurs in E. coli (87),Serratia sp. (23), and in various species ofPseudomonas (14, 80). Small amounts of ,B-hy-droxyoctanoic and ,3-hydroxydodecanoic acidshave been detected in Pseudomonas pyocyanea

(P. aeruginosa) (14). ,B-Hydroxymyristic acidhas been found in E. coli by Ikawa et al. (76)and by Law (87).

Dihydroxystearic acid occurs, apparently as

the free acid, in L. acidophilus. Although 30 yearshave elapsed since this compound was discoveredby Crowder and Anderson (31), the location ofthe hydroxyl groups is still unknown.

Several complex hydroxy acids are peculiar toC. diphtheriae. These compounds, which havebeen discussed in detail by Asselineau (9) andAsselineau and Lederer (12), are corynomycolicacid (C32H6403), its monounsaturated form calledcorynomycolenic acid (C32H6203), and corynolicacid, a dihydroxy compound also known as

corynine (C52H10404). These substances, as theirnames would suggest, have many similarities tothe mycolic acid found in mycobacteria. In fact,the complicated lipid composition of the coryne-bacteria is in many ways more like that of themycobacteria than that of the species in theorders Eubacteriales and Pseudomnadwaes. Aninteresting summary of the many lipoid sub-stances isolated from the corynebacteria is thatof Pustovalov (118).

C. Branched Chain Acids

Mevalonic acid, discussed in the previous sec-

tion with the hydroxy fatty acids, should be men-

tioned here as it is also a branched compound.

Hausmann and Craig (56) have shown thatisooctanoic, or methylheptanoic, acid (C8H1202)is produced by Bacillus polymyxa as a componentof polymyxin B. The same paper reported thatthe major fatty acid constituent of the variouspolymyxins is a methyloctanoic acid. The ac-cumulated results of numerous workers haveshown that this compound is (+)-6-(L)methyl-octanoic acid.

Akashi and Saito (2) detected a branchedchain C15 acid in the lipids of a strain of Sarcina.This compound was tentatively named sarcinicacid. A similar C15 acid and a C17 compound werefound to "abound" in Bacillus subtilis andBacillus natto. In subsequent studies on the acidsfound in the Bacillus species, Saito (121, 122)showed that the C15 acid was 13-methyltetra-decanoic acid and that the C17 acid was15-methylhexadecanoic acid. Whether the C15acid found in Bacillus species is the same as thatdetected in Sarcina species is not yet known.

In a paper dealing with gas chromatographicanalyses of the fatty acids from several microbialspecies, Asselineau (10) has reported finding twobranched chain acids in B. subtilis, one a C15compound and the other a C16 substance. Thesewere not further characterized. In the samepaper there is mention of a C19 branched acidwhich occurs in Pasteurella pestis. This also hasnot been studied in greater detail.A number of complex branched chain acids

have been isolated from the corynebacteria(9, 12, 48). These include corinnic and diphthericacids, both of which have the same empiricalformula, C35H6802, although they appear to differin structural detail. Several derivatives of theseacids have been found in the same organisms.Further, the reader will recall that corynomy-colic, corynomycolenic, and corynolic acids,which were discussed with the hydroxy acids, arealso branched chain compounds.

D. Cyclopropane AcidsThe first acid of this type to be discovered in

bacteria, and the only one for which any detailedstructural information is available, has beennamed lactobacillic acid after the organisms fromwhich it was originally isolated. This compoundwas first reported in 1950 by Hofmann and Lucas(65) and is now known to be cis-11 , 12-methylene-octadecanoic acid (63, 67, 69, 70). It has not yetbeen possible to determine whether this acidoccurs naturally as the D or the L isomer. Organic

4311962]

WILLIAM M. O'LEARY

synthesis of this molecule (69) produces a DLmixture, and resolution studies are impeded bythe unfortunate fact that the optical rotation ofthe natural material is so small as to be un-detectable.

Lactobacillic acid has been unequivocallyidentified in Lactobacillus arabinosus (62, 66),Lactobacillus casei (62, 71), Lactobacillus del-brueckii (62), and Agrobacterium tumefaciens(62, 73). The report on A. tumefaciens is of par-ticular interest. In 1944 Velick (141, 142) re-ported the discovery in Phytomonas (nowAgrobacterium) tumefaciens of a 10- or 11-methyl-nonadecanoic acid which he named "phytomonicacid." Eleven years later, Hofmann and Tausig(73) demonstrated that this substance was a 19-carbon cyclopropane compound rather than abranched methyl acid, and that, in fact, phyto-monic and lactobacillic acids were identical.

Other organisms have been shown to containC19 acids similar to lactobacillic acid althoughthey have not been completely characterized. Thebacteria in which such compounds have beenreported are E. coli (87, 113), C. butyricum (44,45), and possibly in P. pestis (10).For some time it was believed that only C19

cyclopropane acids occurred in bacteria. Thisbelief was based on the fact that no other suchacids had been encountered in all the organismsthat had been studied. However, more sensitivetechniques have shown in recent years that othercyclopropane acids do occur in microorganisms.In 1959 the author detected what appeared to bea C17 cyclopropane acid in the lipids of E. coli(113). Shortly thereafter, Dauchy and Asselineau(32) showed that such an acid was indeed presentin E. coli. The location of the ring in this com-pound, as well as other structural details areunknown. C17 cyclopropane acids have since beenfound in B. subtilis and P. pestis (10), and in C.butyricum (44, 45). Recently evidence has ap-peared suggesting the possible occurrence of C,,and C15 cyclopropane acids in the lipids of C.butyricum (44, 45).

These shorter cyclopropane acids occur inonly small amounts and are most difficult toseparate from other acids of similar chemical andphysical characteristics. These considerationsprobably account for the fact that they are beingencountered only now with the advent of highlyefficient separation techniques. One cannot helpbut speculate on the acids that may not yet have

come to light. Of interest in this respect is thepaper of James and Martin (79) in which theyreport the detection in extracts of a culture ofP. aeruginosa of a number of branched chainacids containing an odd number of carbon atomsranging in length from C8 to C19. None of theseacids was specifically identified, nor was it shownwhether any of these compounds did, in fact,contain a cyclopropane ring. This would be amost interesting matter to investigate further.

Cyclopropene acids, such as sterculic acid(cis-9,10-methylene-9-octadecenoic acid) whichhas been isolated from the oil of Sterculia foetida(108), have not been demonstrated in the truebacteria although such compounds have beensuggested as intermediates in the formation ofsaturated cyclopropane acids. However, recentstudies indicate that cyclopropene acids mayoccur in the lipids of pleuropneumonia-likeorganisms (114a).

E. Unsaturated Acids

These compounds are among the most im-portant fatty acids from a biochemical stand-point and constitute a major portion of the fattyacids of all bacteria. Yet it is with respect to theseacids that the literature is least reliable. There isa curious readiness on the part of many investi-gators to "identify" unsaturated fatty acids onthe basis of the most superficial analytical data.This is by no means always the case, but ithappens often enough to warrant the mostcritical examination of papers purporting to listthe unsaturated fatty acids of bacteria.Let us first consider the octadecenoic acids.

The reader has already been cautioned at somelength about the tendency in the past, and tono small extent in the present, to call any mono-unsaturated C18 fatty acid "oleic acid." In 1952it was found (66) that the octadecenoic acid ofL. arabinosus was in actuality cis-vaccenic acid(i.e., cis-11-octadecenoic) rather than oleic acid(cis-9-octadecenoic acid). In rapid succession itwas shown that cis-vaccenic acid was the soleoctadecenoic acid present in L. casei (71) andin A. tumefaciens (73), two organisms thatdiffer greatly biologically. In a study of thelipids of a group C streptococcus, Hofmann andTausig (72) found that the major octadecenoicacid was, again, cis-vaccenic accompanied by asmall amount of oleic. Law reported that the

432 [VOL. 26

FATTY ACIDS OF BACTERIA

octadecenoic acid in E. coli was cis-vaccenicacid (87).

Oleic acid has been definitely identified inbacteria in certain instances. For example, thiscompound was unequivocally detected in strepto-cocci (72). Also there is the most interesting andconclusive identification of this compound inthe tubercle bacillus by Cason and Tavs (24).Other cases in which oleic acid was specificallyidentified have been listed by Asselineau (9).

Hexadecenoic acids are almost as frequentlyencountered in bacteria as are the octadecenoicacids. On occasion palmitoleic acid (cis-9-hexadecenoic acid) has been identified (6, 72,87, 118), but in most studies the evidence justifiesonly the generic term "hexadecenoic acid."There has been one report of cis-11-hexadecenoicacid in streptococci (72). This compound was

named palmitvaccenic acid because of thesimilarity between the carboxyl end of its mole-cule and that of cis-vaccenic acid.

There are few reports of unsaturated acids offewer than 16 carbon atoms in bacteria. Theoccurrence of an unidentified tetradecenoic acidin C. diphtheriae has been reported by Asanoand Takahashi (6). Gubarev et al. (47) haveevidence of a C10 or C12 unsaturated acid inBrucella suis.

Unsaturated fatty acids with chain lengthsgreater than C18 have been noted only in thecorynebacteria; C20 acids have been reportedby Alimova (3) and by Pustovalov (118), C21by Asano and Takahashi (6), C22-24 by Alimova(3), and C28 again by Asano and Takahashi (6).

IV. FATTY ACID SPECTRA OF SOMEREPRESENTATIVE BACTERIA

Although the identification of specific fattyacids in various bacterial species is interestingand important if not indispensable, the mostuseful information in this field is the exhaustiveanalysis of all the fatty acids present in a givenorganism. Such catalogues of the fatty acidcomposition of individual species are often re-

ferred to as fatty acid spectra. At the presenttime, only a few such spectra are available. Theorganisms which have been most thoroughlycharacterized in this respect are A. tumefaciens(73), C. butyricum (45), E. coli (32, 87, 113), L.arabinosus (66), L. casei (71), L. delbrueckii(62), and a group C streptococcus (72). Thefatty acid composition of these organisms is

TABLE 2. Fatty acid spectra of somerepresentative bacteria

Fattyacids a

SaturatedC10 1.16 2.1 0.5 0.5 0.5 0.3 0.9C12 2.3 2.8 1.1 5.2 2.4 0.3 4.0C14 1.2 2.1 2.5 4.4 0.4 0.7 1.1C16 18.7 24.3 27.5 26.6 49.0 8.2C16 + C17C 85.4C18 + C19C 7.0 10.8 18.0 6.2018 + C19c 0.5Unsaturated016 17.001s 7.9C16 + C18 35.6 37.6 45.5 38.0 11.6 63.6

Cyclopro-pane

013 0.4018 1.5C17 9.0 +dC19 30.1 12.6 6.4 5.2 +d 9.4

a References.b All figures refer to per cent of total fatty acids.e Cyclopropane acid.d + = Present but amount not determined.

shown in Table 2. In addition, there are severalother bacteria whose lipids have been somewhatless intensively studied but regarding which thereis nevertheless considerable information avail-able. Among these are B. subtilis (2, 10, 121),B. suis (47), C. diphtheriae (3, 4, 6, 10, 48, 118),Corynebacterium ovis (10), P. pestis (10), and aspecies of Sarcina (2).In discussing determinations of the total

complement of fatty acids in various bacterialspecies, it is germane to mention again brieflythe fact that marked quantitative and sometimeseven qualitative changes can be brought aboutin fatty acid spectra by alterations in media orgrowth procedures (see, for example, (61) and(68)).V. CHEMICAL NATURE OF BACTERIAL LIPIDS

It is not strictly within the purview of thispaper to discuss in any detail the chemical com-binations in which bacterial fatty acids occur.

19621 433

WILLIAM M. O'LEARY

Several reviews have appeared in recent yearswhich have dealt with the chemical nature ofbacterial lipid complexes (9, 11, 12), and theinterested reader is referred to these articles forextensive information on this subject. However,for the purposes of the present discussion, itseems apropos at least to enumerate and describebriefly the various forms in which bacterialfatty acids are encountered.

Various authors including Asselineau andLederer (12) and Lovern (97) have emphasizedthe fact that bacterial lipids differ radically fromthose of higher forms of life in several respects.These include: the presence of uncommonlylarge proportions of free fatty acids, the frequentpresence of unusual fatty acids not seen in otherorganisms, the absence of sterols, and the absenceof classical lecithins and cephalins. In manystrains the phospholipids are markedly low innitrogen and rich in carbohydrate. These generalcharacteristics are reflected in the discussions ofthe different classes of bacterial lipids thatfollow.A few reports have appeared from time to

time regarding the specific fatty acid(s) detectedin individual lipid complexes, but in generalthere is little information on the distribution ofthe fatty acids among the various types oflipids found in bacteria. This is understandablewhen one considers that investigators are stillstriving to establish merely the identity andrelative abundance of each fatty acid in thebacterial cell as a whole. Now that a clearerpicture is emerging of the fatty acids that onecan expect to encounter, in all likelihood one ofthe next steps in this field will be to determinewhich acids are present in which lipids.

A. Free Fatty AcidsAs just noted, the proportion of free, or un-

esterified, fatty acids found in bacterial cells isunusually high in comparison with the amountsfound in the cells of other forms of life. A numberof bacteria are known in which more than 20%of the total fatty acids occur free. Among theseare C. diphtheriae, L. acidophilus, Bacillus mega-therium, A. tumefaciens, Hemophilus (Bordetella)pertussis, and Salmonella typhimurium (12, 83).The lipid fraction of the last named organismhas been reported to consist almost entirely offree fatty acids.The major constituents of free fatty acid

fractions are usually found to be palmitic,stearic, and octadecenoic acids, although smallamounts of many more fatty acids have beendetected in one organism or another in theunesterified form (83). The purposes whichunesterified fatty acids serve in bacteria arenot known. Equally uncertain are the cellularsites at which they may be located.

B. Fatty Acid PolymersLipoid material consisting of polymerized

/3-hydroxybutyric acid has been isolated frommany bacteria including species of Azotobacter(39, 92), Chromobacterium (39), Pseudomonas(39), Bacillus (91, 146, 151), and Micrococcus(131). Law and Slepecky (88) have summarizeda number of recent papers dealing with poly-p-hydroxybutyric lipid. With respect to the pos-sible function of this polymer, Luria (99) hascommented, "This is an interesting storagedevice, which makes available a metabolite lessrich in energy, but more readily utilizable thanthe fully reduced chains of saturated fatty acids."There is at present, however, no detailed in-formation on the metabolism of this type oflipid.

C. GlyceridesMono-, di-, and triglycerides have been re-

ported in many bacteria (12, 117). However, itis noteworthy that the concentrations of theseneutral lipids are much lower in bacteria thanin plant and animal cells, and that, in fact, manybacteria apparently contain no glycerides at all(11, 83).

D. WaxesAs has been pointed out recently by Asselineau

and Lederer (12), the occurrence of true waxes(i.e., esters of long chain fatty acids and longchain fatty alcohols) is quite uncommon inbacteria. Such substances appear to be limitedprimarily to the mycobacteria and to the coryne-bacteria.

E. PhospholipidsFor a discussion of the chemistry of the

phospholipids, the reader is referred to a mono-graph by Lovern (96). Representatives of thislipid class are widespread in bacteria and fre-quently comprise major portions of the totallipid contents of the cells. Unfortunately, few

434 [VOL. 26

FATTY ACIDS OF BACTERIA

bacterial phospholipids, or phosphatides asthey are also called, have been studied thoroughlyby modern analytical methods. Carbohydrates,inositol, and glycerol are common constituents ofthose bacterial phospholipids which have beenstudied. Of the nitrogen-containing components,ethanolamine is most frequently detected, cholinebeing found in some instances (12). Bases suchas sphingosine have not been found in bacteria.Lovern (97) has commented on the relativesimplicity of many bacterial phospholipids whichhe describes as being essentially fatty acid estersof phosphorylated carbohydrates. Phosphatidicacids (diacyl glycerophosphoric acids) are alsocommonly found in microorganisms.

F. Glycolipids

These important substances have been re-viewed recently by Lederer (89) and Law (86).Such compounds are characterized by thecombination of one or more molecules of fattyacids or fatty alcohols and one or more moleculesof sugar. Many carbohydrates have been de-tected in bacterial glycolipids including glucose,galactose, arabinose, mannose, rhamnose, andtrehalose. Amino sugars are also frequently ob-served.

Strictly speaking, the glycolipids are com-pounds which are soluble in such solvents asether or benzene, but not soluble in water.However, the water-soluble lipopolysaccharidesare usually considered in discussions of glyco-lipids because the lipoid moieties of the lipo-polysaccharides are themselves glycolipids. Manyof the glycolipids have pronounced biologicalproperties as, for example, the antibiotic sub-stances that have been isolated from variousspecies of Pseudomonas and the lipopolysac-charide endotoxins of gram-negative bacteria(89).

G. Lipopeptides and LipoproteinsAs their names suggest, these are complexes

consisting of fatty acids and amino acids. Puristsrequire that these terms be used only for com-plexes having the solubility characteristics ofpeptides or proteins, whereas complexes withlipid solubilities should be called peptidolipidsand proteolipids. It has been shown that in-dividual complexes of this general type havefixed, characteristic proportions of lipid andprotein, so that they seem to have some specific

structure (51). At present, however, very littleis known of the nature of these substances inbacteria.

H. Bound Lipids

This is a frequently encountered but ill definedterm that refers more to solubility characteristicsthan to chemical nature. "Bound lipids" aregenerally understood to be lipoid constituents ofthe cell that are firmly combined with carbo-hydrates and proteins and that, consequently,cannot be extracted with organic solvents untilfirst liberated by some hydrolytic procedure.As has been noted in a previous section, thesebound lipids often constitute large proportions ofthe total cellular lipids.

VI. INTRACELLULAR DISTRIBUTION OFFATTY ACIDS

Information regarding the chemical distribu-tion of bacterial fatty acids is scanty, but evenless is known about the intracellular distributionof these compounds. Certain cellular structuresare known to be relatively rich in lipid, but evenin these instances little is known of the specificfatty acids present.

A. Cytoplasmic InclusionsLipid inclusions are commonly detected by

differential staining using a dye such as Sudanblack B. However, as Lamanna and Mallette(85) have pointed out, a globule staining withsuch a dye is not necessarily composed exclusivelyof lipid. As lipids have a tendency to collect atexposed surfaces, it is possible that many "lipidinclusions" may be nonlipid granules coated bylipid.

In any event, lipid inclusions have been re-ported in many bacteria. According to Burdon(21), the accumulation of lipid globules withinthe cell is a consistent characteristic of variousspecies. Such globules are especially conspicuousin species of Azotobacter, Rhizobium, Spirillum,and Bacillus. Stainable intracellular lipid hasalso been found in C. diphtheriae, certain speciesof both gram-positive and gram-negative cocci,and in some anaerobic bacilli. Few gram-negativerods have been found to contain visible deposi-tions of lipoid material. Murray (106) has pointedout that the number of lipid globules per cellappears to increase with increasing physiological

19621 435

WILLIAM M. O'LEARY

age and is influenced to some extent by thenature of the medium.With respect to the chemical nature of lipid

inclusions, the literature appears to be limitedto the demonstrations by Lemoigne and hisassociates (91) and by Williamson and Wilkin-son (151) that the lipoid granules characteristicof Bacillus species are composed almost exclu-sively of polymerized f3-hydroxybutyric acid.

B. Cell Walls

In his recent and authoritative reviews on thecell wall (123, 124), Salton has pointed out, ashave many others, that there is a marked differ-ence in lipid content between the cell walls ofgram-positive and gram-negative bacteria. Onlysmall amounts of lipid have been found asso-ciated with the isolated cell walls of gram-positive organisms, usually on the order of 1 to3%; it has been suggested that such a smallquantity of lipid might be merely contaminationfrom the rest of the cell and that in actuality thegram-positive cell wall may not contain anylipid at all. In contrast, gram-negative cell wallscontain between 10 to 20% lipid. The physiologi-cal significance of this difference is not known.

In 1956, Salton (123) stated bluntly, "Thereis no information about the chemical nature ofcell wall lipids." Unfortunately, little has hap-pened in the intervening years to modify thisbleak evaluation. Asselineau and Lederer (12)have commented that gram-negative cell wallsseem to be composed largely of lipoprotein andlipopolysaccharide complexes. According toWeidel and Primosigh (147), the cell wall ofE. coli contains 80% lipoprotein. In what maybe the first attempts at actual analyses of cellwall lipids, Alimova (3, 4) has found that thewalls of C. diphtheriae contain a variety of fattyacids ranging in chain length from C14 to C24.The acids occur free, as esters of trehalose, orin both forms (3, 4).

C. Protoplasmic MembranesThese structures, which are believed to be the

selective, semipermeable osmotic barriers ofbacterial cells, are notably rich in lipid materials.Studies of a variety of bacteria show that thelipid content of their protoplasmic membranesvaries between 10 and 30% (12, 106). Gilbyand his associates (43) found, for example, thatthe protoplasmic membranes of Micrococcus

lysodeikticus contained 28% lipid which con-sisted largely of phosphatidic acids. Weibull(146) has shown that the membranes from B.megaterium strain M contained 55 to 75% ofthe total cellular lipid despite the fact that themembranes represented only about 15% of thetotal cell weight. These membrane lipids con-sisted of free fatty acids, neutral fats, andphosphatidic acids.

D. Protoplasts and L-Forms

It is of interest that these protoplasmic ele-ments, whose chief characteristic is the lack ofa cell wall, frequently contain large amounts oflipids (102). Vendreley and Tulasne (143) haveshown, for example, that L-forms derived fromProteus species contain 24% lipid although thebacterial forms contain only 6 to 8%. As yetno data are available regarding the chemicalnature of the lipids found in bacterial protoplasts,L-forms, or mycoplasmas (pleuropneumonia-like organisms; PPLO). This is doubtless dueto the great difficulty of growing sizable amountsof these entities.

VII. ASPECTS OF FATrY ACID BIOCHEMISTRYAND METABOLISM

The burgeoning activities in fatty acid chem-istry have 'been paralleled by equally intenseresearch on the metabolism of these compounds.As has been the case in the chemical studies,metabolic research has been greatly facilitatedand accelerated by the availability of new in-struments and techniques. Of particular im-portance have been the various isotopic tracertechniques, using both stable and radioactivenuclides, which have permitted the following ofatoms and molecules through many of the in-tricate pathways of fatty acid metabolism. Meadand Howton (103) have discussed in a recentand excellent monograph the many uses ofradioisotopes in this field.The studies of fatty acid biochemistry and

metabolism in bacteria not only have contributedto our knowledge of bacteria per se, but alsohave been most useful in providing clues to thedetails of fat metabolism in higher animals. Inspite of the progress that has been made, how-ever, there still exists a great deal of uncertaintyregarding this aspect of bacterial physiology.Prior to the 1950's, very little was known aboutthe synthesis or utilization of fatty acids in

436 [VOL. 26

FATTY ACIDS OF BACTERIA

bacteria, or in other organisms for that matter.Then, with the discovery of the coenzyme A(CoA) cycle and the elucidation of many detailsregarding it, there was a widespread feeling thatthe mystery of fatty acid metabolism had beensuddenly and largely solved. In recent years,however, as data accumulated and promisinghypotheses were tested and found inadequate, ithas become increasingly apparent that the pictureis much more complex than was originally be-lieved, and that many pieces of the puzzle haveyet to be found and fitted into place.We are fortunate that a number of excellent

reviews have appeared lately discussing thegeneral field of fatty acid metabolism in somedetail (28, 82, 100, 101, 136). It is my intent,therefore, to discuss below only certain aspectsof the subject that are today receiving particularattention from workers in the field.

A. Degradation and SynthesisThe (3-oxidative reactions associated with

CoA have been demonstrated in various bacteria(133, 136), and it is generally assumed that thismechanism is operative in the bacterial degrada-tion of fatty acids. Indeed, much of our presentknowledge of (3-oxidation and of the functioningof the CoA cycle has been obtained from studiesusing microorganisms (133).Although a large amount of information has

been gathered regarding the oxidation of fattyacids, considerably less is known about thesynthesis of these compounds. Since the variousreactions involved in (3-oxidation are known tobe individually reversible, it was a popular con-cept for some years that synthesis was merelythe reverse of degradation. It is now evidentthat the synthesis of fatty acids is a more com-plicated process than was previously suspected,and that it involves considerably more thanmerely "backing up" the p3-oxidation pathway.Wakil (145) has discussed this problem ex-haustively and, on the basis of studies so farlimited to preparations of animal tissues or ofyeast, has proposed the existence of at least twosynthetic mechanisms. One of these appears toinvolve enzymes of the (3-oxidation system andis believed to be primarily responsible for theelongation of existing fatty acids. The othersystem converts acetyl-CoA to malonyl-CoAwhich is then condensed with itself, acetyl-CoA,or propionyl-CoA to form long chain fatty acids.

The malonyl-CoA is formed by condensingHC03- with acetyl-CoA in a reaction catalyzedby acetyl-CoA carboxylase (a biotin-containingenzyme). Condensations involving propionyl-CoA are of special interest in this context inthat they could account for branched chain acidswhich, as has been pointed out, are common inbacteria.

Another dichotomy in bacterial lipid me-tabolism seems to exist between the synthesis ofsaturated and unsaturated fatty acids. It hasbeen found that in animal metabolism long chainunsaturated fatty acids are formed largely byenzymatic dehydrogenation of the correspondinglong chain saturated acid, e.g., oleic acid isproduced by dehydrogenation of stearic acid.For some time it was assumed that bacteria

formed their long chain unsaturated fatty acidsby a similar procedure, and there have been afew reports of such dehydrogenations in a smallnumber of species. However, the infrequentobservation of this reaction in bacteria seemedcurious in view of the fact that unsaturated fattyacids comprise large portions of the total fattyacid content of most microorganisms. Conse-quently, many investigators began to suspectthat in bacteria dehydrogenation was not themajor pathway to unsaturated acids, that insteadsome other more complicated mechanism wasinvolved. The literature pertaining to thisproblem has been reviewed elsewhere (111).

Investigations of lactobacilli, in which themajor unsaturated fatty acid is cis-vaccenic acid,have suggested the interesting possibility thatthis compound may be formed by the progressivelengthening, two carbons at a time, of someunknown, already unsaturated short chainprecursor (68, 114). As will be discussed below,it has been shown that cis-vaccenic acid is theprecursor of lactobacillic acid in these organisms.These observations suggest the possibility of apathway such as is outlined in Fig. 1.

It will be noted that in each of the compoundsindicated, the double bond is located in the sameposition with respect to the methyl end of thechain. Thus lengthening the chains on thecarboxyl ends would eventually produce palmi-toleic and cis-vaccenic acids, both of which areroutinely found in lactobacilli.Although this tentative pathway is an in-

teresting and plausible explanation of unsaturatedfatty acid biosynthesis in bacteria, it is em-

4371962]

WILLIAM M. O'LEARY

LUnknownunsaturatedprecursor

I- + " ,,

CH3-(CH2)5 (CH2)7-COOH

C0=CH I

cis-9-hexadecenoic acid(palnitoleic acid)

+

ICH -(CH)5 (CH)9-COOH

\ /

C=CH H

cis-11-octadecenoic acid(cis-vaccenic acid)

CH3-(CH2)5 /(CH2)r-COOHI

-C

H H

cis-5-dodecenoic acid

4-. "C2" +

+ "C22

1CH3-(CH2)5 (CH2)5-COOH

C-C

H H

cis-7-tetradecenoic acid

H H

04-(CH2)5\C/ (CH2)9-COOH

H" ,Hcis-1l, 12-methylene octadecanoic

acid (lactobacillic) acid

FIG. 1. Possible pathway for the biosynthesis of unsaturated and cyclopropane fatty acids in bacteria.

phasized that these studies are as yet preliminary,not conclusive, and that further investigation ofthis matter is essential. Some recent studies ofC. butyricum and several other organisms haveprovided further evidence for separate anddistinct biosynthetic pathways for saturatedand unsaturated fatty acids (44, 45).A third divergence in bacterial lipid metab-

olism is the fact that the metabolic pathways ofthe lower fatty acids (C2 to Cl0) differ in manyways from those of the higher acids (C12 to C18).It is, for example, possible to inhibit the oxida-tion of higher acids without affecting oxidationof the shorter compounds. Further, unsaturatedfatty acids ranging in chain length from C18down to C12 have been shown to stimulategrowth of certain bacteria, whereas shorterunsaturated acids do not possess this capability(68, 111). Oginsky and Umbreit (110) havesuggested a spatial arrangement of the fattyacid molecule which, if correct, would make itimprobable that the longer acids would besuitable substrates for enzymes that could

affect the shorter compounds. Also, of course,there are marked differences in the solubilitycharacteristics of the two groups that mightinfluence their metabolism.

B. Stimulation and Inhibition of GrowthIt has been known for some time that fatty

acids in low concentrations can have markedeffects on the growth of bacteria. Such phe-nomena have been reviewed by Kodicek (84),Nieman (107), and Deuel (33). The nature andmagnitude of the effect in any given instancedepend upon the specific fatty acid, on its con-centration, and upon the bacterial species in-volved. In general, gram-positive bacteria aremore sensitive to all effects of fatty acids thanare the gram-negative organisms. Both saturatedand unsaturated fatty acids can cause inhibitionof growth. In the unsaturated acids, inhibitoryeffects increase with the number of doublebonds in the molecule. Also, cis isomers are moreinhibitory than the corresponding trans forms.Stimulation of growth is a property largely

438 [VOL. 26

FATTY ACIDS OF BACTERIA

limited to unsaturated long chain acids (i.e.,C12 to C18) and to certain cyclopropane acids.Stimulatory effects decrease with increasingunsaturation, and, once again, cis isomers aremore effective than trans. Acids which arestimulatory in low concentrations are never-theless inhibitory if present in only slightly higherconcentrations.

There has been much disagreement regardingthe nature of the mechanisms by which fattyacids affect bacterial growth. These argumentsare discussed in detail in the references justcited. Here, for purposes of brevity, they canbe reduced to two schools of thought, one favor-ing nonspecific physicochemical mechanisms andthe other favoring more specific effects on met-abolic reactions. For example, it has been gen-erally accepted that inhibitory effects of fattyacids and their compounds are due to theirsurface-active properties, which cause nonspecificalterations in cell permeability and consequentloss of essential cellular constituents. However,two reports have appeared (22, 68) which in-dicate that, at least in lactobacilli, inhibition ofgrowth by saturated acids and the reversal ofthis inhibition by unsaturated acids may bespecific antimetabolite-metabolite interrelation-ships.

Similarly, some have chosen to regard thegrowth-stimulating effects of unsaturated fattyacids as merely the consequences of nonspecificenhancement of cell permeability, which fa-cilitates the assimilation of essential nutrients.Others, including the author, feel that the pre-ponderance of experimental evidence shows thatfatty acids stimulate growth because they areuseful and even essential metabolites for theorganisms so affected.The ability of unsaturated fatty acids to re-

place biotin in the nutrition of certain bacteriais pertinent in this respect. The literature re-garding this interesting aspect of bacterial lipidmetabolism has been reviewed extensively else-where (107, 111). Many fatty acids have beenfound to replace biotin including oleic, elaidic,cis- and trans-vaccenic, linoleic, linolenic, andother unsaturated C18 acids (13, 25, 70, 149),hexadecenoic acid (55, 68), dodecenoic andtetradecenoic acids (68), and even certaincyclopropane acids (70). The ability of eachacid to serve in lieu of biotin depends upon the

bacterial species, but is quite constant for anygiven organism.

Considerable evidence has been accumulatedin support of the concept that biotin is somehowinvolved in the biosynthesis of long chain un-saturated fatty acids in many bacteria, possiblyin the formation of the short chain unsaturatedprecursors (see (68)), and that if the unsaturatedacids are supplied preformed, the biotin require-ment is correspondingly reduced (19, 68, 70).Though this concept has gained considerable

credence and currency, it does not satisfy equallyall students of the subject (84, 93, 140, 150). Itmay well be that the observed effect of exogenousfatty acid on bacterial growth in any giveninstance is actually the summation of changesbrought about by many diverse mechanisms.Only further studies can say.

C. Metabolism of Cyclopropane Acids

When lactobacillic acid was discovered in1950 (65), it was the first fatty acid containing acyclopropane ring to be encountered in biologicalmaterials. Even today, as has already beenpointed out, only a few compounds of thistype are known. Also, lactobacillic acid was thefirst compound other than long chain unsaturatedfatty acids shown to have a biotin-like activityfor certain bacteria (70). Understandably, in therelatively short time that they have been knownto exist, the cyclopropane acids have posed anumber of interesting metabolic questions.One major problem associated with these com-

pounds has been the manner in which they arebiosynthesized. Some years ago Hofmann andhis co-workers (61) proposed, on the basis ofindirect evidence, that lactobacillic acid mightbe produced in bacteria by the addition of a"C1" fragment across the double bond of cis-vaccenic acid. This addition accompanied byreduction of the double bond would then resultin a cyclopropane ring in the correct chainlocation. It was later shown by radioisotopetechniques that when cis-vaccenic acid is suppliedunder appropriate conditions to actively growinglactobacilli, it is taken into the cells unalteredand there used as the chain moiety of lactobacillicacid (112). This, then, accounted for the chainand 18 of the 19 carbon atoms of lactobacillicacid. Shortly thereafter, also using tracer pro-cedures, it was shown that the 19th carbon atom,i.e., the one added to form the cyclopropane

4391962]

WILLIAM M. O'LEARY

ring, was derived from the methyl group ofmethionine (113). Subsequent studies haveconfirmed and expanded these observations(64, 95).The concept of cis-vaccenic acid serving as

the immediate precursor of lactobacillic acid isconsonant with the fact that the double bond incis-vaccenic acid and the cyclopropane ring inlactobacillic acid are located at the same placein the chain, i.e., between the 11th and 12thcarbon atoms. Seemingly at variance with thisis the observation that labeled oleic acid suppliedin the medium also gives rise to labeled octa-decenoic and cyclopropane acids in lactobacilli(112). It will be recalled that in oleic acid thedouble bond is located between the 9th and 10thcarbon atoms. Preliminary data presented inthis paper suggest the interesting possibilitythat even when oleic acid is supplied, the acidsin the bacterial cells are cis-vaccenic and lacto-bacillic. This possible interconversion meritsfurther investigation. In their studies of fattyacid metabolism in clostridia, Goldfine andBloch (45) have also observed the conversion ofoleic acid to a C19 cyclopropane fatty acid.An even more intriguing problem is the use

to which lactobacillic acid is put by bacteriawhich contain it. Various pieces of evidenceargue, somewhat teleologically it is true, infavor of some specific and important function.The nature of the molecule requires the expendi-ture of considerable energy in its synthesis. Themolecule is a highly specific structure in thatalterations in chain length, cis-trans isomerism, orpositional isomerism have quite marked effectson its biological activity (70). Studies of syn-thetic DL mixtures indicate that the active ma-terial is exclusively either the D or the L con-figuration, the contrary form being withoutbiological activity (69). Lastly, it has been shownthat, in bacteria containing this compound, it ispossible by appropriate cultural procedures toincrease, decrease, or even eliminate altogetherthe amounts of the other major acids bothsaturated and unsaturated. However, it is notpossible by any means so far tried to eliminateor even to reduce below certain levels the amountof lactobacillic acid in these cells (61, 68). Allof these observations alike suggest some im-portant, indispensable function for lactobacillicacid. The nature of this function is totally un-known at present. It has been suggested, in

view of the rather labile ring carbon which isthe conspicuous and unusual feature of thismolecule, that lactobacillic acid may function asa carrier of 1-carbon fragments or even as astorage form of such units, but there is at thistime no experimental evidence to support suchhypotheses.The information summarized in the preceding

paragraphs pertains largely to lactobacillicacid, and most of the data have been gatheredin studies on the lactobacilli. It is still too earlyto tell how much of this will be applicable tothe shorter cyclopropane compounds found inother genera. We now predicate that the syn-thesis of lactobacillic acid is accomplished bythe formation of an 18-carbon unsaturated acidwith a double bond between the 11th and 12thcarbon atoms and then the bridging of thisdouble bond with a methylene group by someprocess resulting in a saturated ring. Are each ofthe shorter acids terminal products of similarsyntheses in which the immediate precursors arecorresponding unsaturated acids, or is some othermechanism involved? Much will depend uponthe as yet undetermined locations of the cyclo-propane rings in the shorter compounds. Arethe shorter acids capable of replacing biotin orof producing any other growth effects? Do thevarious cyclopropane compounds have similaror different functions? What are the details ofthe biosynthetic reactions involved in con-structing a cyclopropane ring on a carbon-to-carbon double bond? Obviously many interestingfindings lie ahead in the study of these sub-stances.

D. Other Findings of Interest

In this section are included several phenomenaassociated with bacterial fats which may wellprove to be of major importance but regardingwhich little is definitely known at the presenttime. Recently two reports have appeared re-garding the possible involvement of bacterialfatty acids and their complexes in the transportof amino acids. Hunter and Goodsall (75) havepresented evidence indicating that in B. meg-aterium lipid complexes transport amino acidsfrom the site of activation to the site of proteinsynthesis. Silberman and Gaby (129) havepublished some more detailed observations on asimilar action in P. aeruginosa. These reports

440 [VOL. 26

FATTY ACIDS OF BACTERIA

are quite recent, and as yet no further elucidationof these observations has appeared.The continuing search for molecular explana-

tions of the mode of action of antibiotics hasprovided evidence that reactions with lipoidmaterials may play a role in the actions of someof the antibiotics. Cooper (27) has shown thatthe so-called penicillin combining component inmany gram-positive bacteria is a lipid substancewhich is still unidentified. A similar reactionappears to be involved in the initial combinationof the polymyxins with sensitive bacteria (37).It might be mentioned here that at least oneantibiotic substance, pyolipic acid isolated fromP. pyocyanea (aeruginosa), is a glycolipid con-sisting of 1 molecule of L-rhamnose and 1 moleculeof D-3-hydroxydecanoic acid (14).The ability of bacterial fatty acids and their

complexes to produce a variety of pharmacologi-cal and immunological effects in higher animalsis another fertile field for further investigation.Asselineau and Lederer (12) have recently re-viewed these effects which include enzyme in-hibition, inhibitory and stimulatory effects onleukocytes, local tissue reactions, general toxicity(as exhibited by the lipopolysaccharide endo-toxins of gram-negative bacteria), antigenicity,and adjuvant action. Observations on theseproperties, although numerous, are almost with-out exception empirical. Little is known of thechemical composition of the fractions producingthese effects, of the mechanisms whereby theeffects are elicited, or of the relationships betweenspecific composition and specific effect.A recent paper by Dubrovskaia et al. (34)

suggests that one aspect of the effect of phageon microorganisms may be a disturbance of lipidmetabolism. In studies employing Brucelkaabortus, these workers noted that cells infectedwith phage contained 26% more lipid than didnormal cells. At the same time they observed ashift in the proportions of different type oflipids in that phage infection caused an increasein phosphatide and a concomitant decrease inneutral fat.

VIII. CONCLUSION

The very considerable increase in interest,research, and knowledge that has taken placein recent years regarding the fatty acids ofbacteria is indeed encouraging, especially incomparison with the years of relative quiescence

in this field. Nevertheless, as has been aptly saidin another context, all we know is still infinitelyless than all that yet remains to be known.Many intriguing problems still await attention

and solution. Among these, one may cite severalas being of particular and immediate interest:the manner in which the various acids are dis-tributed among the different lipid classes, thecytological distribution of fatty acids, the detailsof biosynthesis and metabolism particularly withrespect to the unsaturated and cyclopropanecompounds, and the specific functions of thefatty acids and their compounds in the bacterialcell.The complexity and importance of these

questions are ample indication that, in spite ofthe progress made thus far, the major accom-plishments in the study of bacterial fatty acidsand lipids still lie ahead. The writer, for one,awaits future developments with enthused an-ticipation.

IX. ADDENDUM

Since the completion of this review, numerouspapers have appeared dealing with variousaspects of microbial lipids. Among these, thefollowing are of especial interest.Two papers have added significantly to

knowledge of the cyclopropane acids. In one,Liu and Hofmann (95a) demonstrated un-equivocally that lactobacillic acid biosynthesisinvolves the addition of a 1-carbon fragmentacross the double bond of cis-vaccenic acid. Inthe other, Kaneshiro and Marr (80a) have shownthat the C17 cyclopropane acid found in thephospholipids of E. coli is cis-9 ,10-methylenehexadecanoic acid, a compound analogous topalmitoleic acid.

Problems of the biosynthesis of unsaturatedfatty acids have been extensively discussed byBloch and his associates (14a). Their findingsindicate that in microorganisms there are twopathways for the synthesis of monoethenoicacids. One involves oxidative desaturation andoccurs in yeast, blue-green algae, and someaerobic bacteria, whereas the other is an anaerobicmechanism involving elongation of already un-saturated acids. The latter pathway is found inanaerobes, facultative aerobes, and some obli-gate aerobes. In a later paper, Scheuerbrandtand Bloch (124a) postulate the involvement ofoctenoic and decenoic acids in the biosynthesis

4411962]

WILLIAM M. O'LEARY

of higher unsaturated and cyclopropane acidsin a series of pathways similar to the schemeshown in Fig. 1 although much expanded indetails.

Lastly, there have been attempts recently toanalyze the fatty acids of microorganisms smallerthan the true bacteria. Examination of the lipidsof PPLO strain 07 (114a) has revealed a fattyacid composition differing both qualitatively andquantitatively from that of most bacteria. InPPLO, the concentrations of unsaturated andcyclopropane acids are much lower than inbacteria, and the concentration of palmitic acidis correspondingly increased. Kates and co-workers (80b) have catalogued the fatty acidsfound in influenza virus and have concludedthat this virus incorporates lipid componentspresent in the host cells before infection as wellas those synthesized after infection. They furtherconcluded that influenza virus is not able todirect cellular lipid synthesis as it does nucleicacid and protein synthesis.

X. LITERATURE CITED

1. AHRENS, E. H. 1956. Application of counter-current distribution to the study of lipids,p. 30-41. In G. Popjak and E. Le Breton[ed.], Biochemical problems of lipids. Inter-science Publishers, Inc., New York.

2. AKASHI, S., AND K. SAITO. 1960. A branchedsaturated C15 acid (sarcinic acid) fromSarcina phospholipids and a similar acidfrom several microbial lipids. J. Biochem.(Tokyo) 47:222-229.

3. ALIMOVA, E. K. 1958. The distribution oflipids between the cell membrane and othercomponent parts of the cell in diphtheriamicrobes. Biochemistry (U. S. S. R.)23:193-198.

4. ALIMOVA, E. K. 1959. The surface layer ofdiphtheria bacteria cells and its toxiclipids. Biochemistry (U. S. S. R.) 24:722-725.

5. ANDERSON, R. J. 1943. The chemistry of thelipids of the tubercle bacillus. Yale J. Biol.and Med. 15:311-345.

6. ASANo, M., AND H. TAKAHASHI. 1945. Bac-terial components of Corynebacteriumdiphtheriae. I. Studies of fats. J. Pharm.Soc. Japan 65:17-19. (Chem. Abstr. 45:3906 (1951).)

7. ASHLEY, B. D., AND U. WESTPHAL. 1955.Separation of small quantities of saturatedhigher fatty acids by reversed-phase paper

chromatography. Arch. Biochem. Biophys.56:1-10.

8. ASSELINEAU, J. 1957. Applications of chroma-tography to fatty acids. Chim. anal. 39:375-383.

9. ASSELINEAU, J. 1957. Les lipides bacteriens,p. 90-108. In W. Ruhland [ed.], Handbuchder Pflanzenphysiologie, v. VII. Springer-Verlag, Berlin.

10. ASSELINEAU, J. 1961. Sur quelques applica-tions de la chromatographie en phasegazeus6 a l'6tude d'acides gras bacteriens.Ann. inst. Pasteur 100:109-119.

11. ASSELINEAU, J., AND E. LEDERER. 1953.Chimie des lipides bacteriens. Fortschr.Chem. org. Naturstoffe 10:170-273.

12. ASSELINEAU, J., AND E. LEDERER. 1960.Chemistry and metabolism of bacteriallipides, p. 337-406. In K. Bloch [ed.], Lipidemetabolism. John Wiley & Sons, Inc., NewYork.

13. AXELROD, A. E., M. A. MITZ, ANDK. HOFMANN. 1948. The chemical nature offat-soluble materials with biotin activityin human plasma. Additional studies onlipide stimulation of microbial growth. J.Biol. Chem. 175:265-274.

14. BERGSTROM, S., H. THEORELL, AND H.DAVIDE. 1946. Pyolipic acid, a metabolicproduct of Pseudomonas pyocyanea, activeagainst Mycobacterium tuberculosis. Arch.Biochem. 10:165-166.

14a. BLOCH, K., P. BARONOWSKY, H. GOLDFINE,W. J. LENNARZ, R. LIGHT, A. T. NORRIS,AND G. SCHEUERBRANDT. 1961. Biosynthesisand metabolism of unsaturated fatty acids.Federation Proc. 20:921-927.

15. BLOCK, R. J., E. L. DURRUM, AND G. ZWEIG.1958. Paper chromatography and paperelectrophoresis, 2nd ed. Academic Press,Inc., New York.

16. BOLDINGH, J. 1950. Fatty acid analysis bypartition chromatography. Rec. trav.chim. 69:247-261.

17. BOLDINGH, J., 1953. The separation of fattyacids by chromatography. Proc. 1st Intern.Conf. Biochem. Prob. Lipides (Brussels),64-81.

18. BOWMAN, J. R., AND R. S. TIPSON. 1951.Distillation, p. 463-494. In A. Weissberger[ed.], Technique of organic chemistry,v. IV. Interscience Publishers, Inc., NewYork.

19. BROQUIST, H. P., AND E. E. SNELL. 1951.Biotin and bacterial growth. I. Relationto aspartate, oleate and carbon dioxide.J. Biol. Chem. 188:431-444.

442 [VOL. 26

FATTY ACIDS OF BACTERIA

20. BROWN, J. B., AND D. K. KOLB. 1955. Applica-tions of low temperature crystallization inthe separation of the fatty acids and theircompounds. Progr. in Chem. Fats Lipids3:57-94.

21. BURDON, K. L. 1958. Textbook of microbi-ology, p. 100-101. 4th ed. The MacmillanCo., New York.

22. CAMIEN, M. N., AND M. S. DUNN. 1957. Satu-rated fatty acids as bacterial antimetabo-lites. Arch. Biochem. Biophys. 70:327-345.

23. CARTWRIGHT, N. J. 1957. The structure ofserratamic acid. Biochem. J. 67:663-669.

24. CASON, J., AND P. TAVS. 1959. Separation offatty acids from tubercle bacillus by gaschromatography: identification of oleicacid. J. Biol. Chem. 234:1401-1405.

25. CHENG, A. L. S., S. M. GREENBERG, H. J.DEUEL, AND D. MELNICK. 1951. Biotin-likeactivity of positional and stereoisomers ofoctadecanoic acids. J. Biol. Chem. 192:611-622.

26. COLOWICK, S. P., AND N. 0. KAPLAN. 1957.Methods in enzymology, p. 299-382. v. III.Academic Press, Inc., New York.

27. COOPER, P. D. 1956. Site of action of radio-penicillin. Bacteriol. Rev. 20:28-48.

28. CORNFORTH, J. W. 1959. Biosynthesis of fattyacids and cholesterol considered as chemi-cal processes. J. Lipid Research 1:3-28.

29. CRAIG, L. C., AND D. CRAIG. 1956. Laboratoryextraction and countercurrent distribu-tion, p. 149-332. In A. Weissberger [ed.],Technique of organic chemistry. v. III.Interscience Publishers, Inc., New York.

30. CROWDER, J. A., AND R. J. ANDERSON. 1932.A contribution to the chemistry of Lacto-bacillus acidophilus. I. The occurrence offree, optically active, dihydroxystearicacid in the fat extracted from Lactobacillusacidophilus. J. Biol. Chem. 97:393-401.

31. CROWDER, J. A., AND R. J. ANDERSON. 1934.A contribution to the chemistry of Lacto-bacillus acidophilus. III. The compositionof the phosphatide fraction. J. Biol. Chem.104:487-495.

32. DAUCHY, S., AND J. ASSELINEAU. 1960. Surles acides gras des lipides de Escherichiacoli. Existence d'un acide C17H3202 conte-nant un cycle propanique. Compt. rend.250:2635-2637.

33. DEUEL, H. J., JR. 1957. The lipids, p. 887-896.v. III. Interscience Publishers, Inc., NewYork.

34. DUrBROVSKAIA, I. I., N. N. OSTROVSKAIA,AND A. I. GLUIBOKINA. 1958. Effect of phageon the chemical composition of Brucella

organisms. Biochemistry (U. S. S. R.)23:489-493.

35. DUTTON, H. J. 1954. Countercurrent frac-tionation of lipides. Progr. in Chem. FatsLipids 2:292-325.

36. EDDY, A. A. 1958. Aspects of the chemicalcomposition of yeast, p. 157-249. A. H.Cook [ed.], In The chemistry and biologyof yeasts. Academic Press, Inc., New York.

37. FEW, A. V. 1955. The interaction of poly-myxin E with bacterial and other lipids.Biochim. Biophys. Acta 16:137-145.

38. FONTELL, K., R. T. HOLMAN, AND G.LAMBERTSEN. 1961. Some new methods forseparation and analysis of fatty acids andother lipids. J. Lipid Research 1:391-404.

39. FORSYTH, W. G. C., A. C. HAYWARD, ANDJ. B. ROBERTS. 1958. Occurrence of poly-j6-hydroxybutyric acid in aerobic gram-negative bacteria. Nature 182:800-801.

40. FOSTER, J. W. 1949. The chemical nature ofmold mycelium, p. 76-147. In Chemicalactivities of fungi. Academic Press, Inc.,New York.

41. FRIEDRICH, J. P. 1961. Low temperaturecrystallization apparatus for semimicroquantitative work. Anal. Chem. 33:974-975.

42. GENSLER, W. J. 1957. Recent developmentsin the synthesis of fatty acids. Chem. Rev.57:191-280.

43. GILBY, A. R., A. V. FEW, AND K. MCQUILLEN.1958. The chemical composition of theprotoplast membrane of Micrococcus lyso-deikticus. Biochim. Biophys. Acta 29:21-29.

44. GOLDFINE, H. 1961. Fatty acid metabolismin Clostridium butyricum. Federation Proc.20:273.

45. GOLDFINE, H., AND K. BLOCH. 1961. On theorigin of unsaturated fatty acids in CMs-tridia. J. Biol. Chem. 236:2596-2601.

46. GUBAREV, E. M. 1958. Bakterielle Lipoide,p. 1-36. In Bakteriochemie. Fischer Verlag,Jena.

47. GUBAREV, E. M., G. D. BOLGOVA, AND E. K.ALIMOVA. 1959. A study of free and boundlipid fractions of the Brucella type suis 44by the method of chromatography. Bio-chemistry (U. S. S. R.) 24:185-189.

48. GABAREV, E. M., E. K. LUBENETS, A. A.KANCHUKH, AND Y. V. GALAEV. 1951.Fractionation and composition of somelipide fractions of diphtheria bacteria.Biokhimiya 16:139-145.

49. GUNSTONE, F. D. 1957. The constitution andsynthesis of fatty acids. Progr. in Chem.Fats Lipids 4:1-44.

50. GUNSTONE, F. D. 1958. An introduction to

44319621

WILLIAM M. O'LEARY

the chemistry of fats and fatty acids. JohnWiley & Sons, Inc., New York.

51. GURD, F. R. N. 1960. Association of lipideswith proteins, p. 208-259. In D. J. Hanahan[ed.], Lipide chemistry. John Wiley &Sons, Inc., New York.

52. HAMILTON, J. G., J. R. SWARTWOUT, 0. N.MILLER, AND J. E. MULDREY. 1961. A silicagel impregnated glass fiber paper and itsuse for the separation of cholesterol, tri-glycerides, and the cholesteryl and methylesters of fatty acids. Biochem. Biophys.Research Commun. 5:226-227.

53. HANAHAN, D. J. 1960. Lipide chemistry. JohnWiley & Sons, Inc., New York.

54. HARDY, C. J., AND F. H. POLLARD. 1959. Re-view of gas-liquid chromatography. J.Chromatog. 2:1-43.

55. HASSINEN, J. B., G. T. DURBIN, AND F. W.BERNHART. 1950. Hexadecenoic acid as agrowth factor for lactic acid bacteria.Arch. Biochem. 25:91-96.

56. HAUSMANN, W., AND L. C. CRAIG. 1954.Polymyxin B,: fractionation, molecular-weight determination, amino acid and fattyacid composition. J. Am. Chem. Soc. 76:4892-4896.

57. HERBERT, D. 1961. The chemical compositionof micro-organisms as a function of theirenvironment, p. 391-416. In G. G. Meynelland H. Gooder [ed.], Microbial reaction toenvironment. Cambridge University Press,New York.

58. HILDITCH, T. P. 1956. The chemical constitu-tion of natural fats. 3rd ed. John Wiley &Sons, Inc., New York.

59. HIRSCH, J. 1959. Preparative and analyticalseparation of fatty acids by rubber chroma-tography. Federation Proc. 18:246.

60. HOFMANN, K. 1953. The chemical nature offatty acids of bacterial origin. RecordChem. Progr. 14:7-17.

61. HOFMANN, K., D. B. HENIS, AND C. PANOS.1957. Fatty acid interconversions in lacto-bacilli. J. Biol. Chem. 228:349-355.

62. HOFMANN, K., C. Y. HSIAO, D. B. HENIS,AND C. PANOS. 1955. The estimation of thefatty acid composition of bacterial lipides.J. Biol. Chem. 217:49-60.

63. HOFMANN, K., 0. JUCKER, W. R. MILLER,A. C. YOUNG, JR., AND F. TAUSIG. 1954.On the structure of lactobacillic acid. J.Am. Chem. Soc. 76:1799-1804.

64. HOFMANN, K., AND T. Y. LIu. 1960. Lacto-bacillic acid biosynthesis. Biochim. Bio-phys. Acta 37:364-365.

65. HOFMANN, K., AND R. A. LUCAS. 1950. The

chemical nature of a unique fatty acid.J. Am. Chem. Soc. 72:4328.

66. HOFMANN, K., R. A. LUCAS, AND S. M. SAX.1952. The chemical nature of the fatty acidsof Lactobacillus arabinosus. J. Biol. Chem.195:473-485.

67. HOFMANN, K., G. J. MARCO, AND G. A.JEFFREY. 1958. Studies on the structure oflactobacillic acid. III. Position of thecyclopropane ring. J. Am. Chem. Soc.80:5717-5721.

68. HOFMANN, K., W. M. O'LEARY, C. W. YOHO,AND T-Y. LIU. 1959. Further observationson lipide stimulation of bacterial growth.J. Biol. Chem. 234:1672-1677.

69. HOFMANN, K., S. F. OROCHENA, AND C. W.YoHo. 1957. Unequivocal synthesis ofDL - cis - 9,10 - methyleneoctadecanoic acid(dihydrosterculic acid) and DL-ciS-11,12-methyleneoctadecanoic acid. J. Am. Chem.Soc. 79:3608.

70. HOFMANN, K., AND C. PANOS. 1954. Thebiotin-like activity of lactobacillic acidand related compounds. J. Biol. Chem.210:687-693.

71. HOFMANN, K., AND S. M. SAX. 1953. Thechemical nature of the fatty acids of Lacto-bacillus casei. J. Biol. Chem. 205:55-63.

72. HOFMANN, K., AND F. TAUSIG. 1955. Thechemical nature of the fatty acids of agroup C Streptococcus species. J. Biol.Chem. 213:415-423.

73. HOFMANN, K., AND F. TAUSIG. 1955. On theidentity of phytomonic and lactobacillicacids. A reinvestigation of the fatty acidspectrum of Agrobacterium (Phytomonas)tumefaciens. J. Biol. Chem. 213:425-432.

74. HOLMAN, R. T. 1952. Chromatography offatty acids and related substances. Progr.in Chem. Fats Lipids 1:104-126.

75. HUNTER, G. D., AND R. A. GOODSALL. 1960.Protein synthesis in protoplasts of Bacillusmegatherium. The passage of '4C-labelledamino acids through phospholipid frac-tions. Biochem. J. 74:34P.

76. IKAWA, M., J. B. KOEPFLI, S. G. MUDD,AND C. NIEMANN. 1953. An agent fromEscherichia coli causing hemorrhage andregression of an experimental mouse tumor.III. The component fatty acids of thephospholipide moiety. J. Am. Chem. Soc.75:1035-1038.

77. JAMES, A. T. 1960. Qualitative and quantita-tive determination of the fatty acids bygas-liquid chromatography, p. 1-59. InD. Glick [ed.], Methods of biochemical

444 [VOL. 26

FATTY ACIDS OF BACTERIA

analysis. Interscience Publishers, Inc.,New York.

78. JAMES, A. T., AND A. J. P. MARTIN. 1952.Gas-liquid chromatography. The separa-tion and microestimation of volatile fattyacids from formic acid to dodecanoic acid.Biochem. J. 50:679-690.

79. JAMES, A. T., AND A. J. P. MARTIN. 1956.Gas-liquid chromatography. The separa-tion and identification of the methyl estersof saturated and unsaturated acids fromformic acid to n-octadecanoic acid. Bio-chem. J. 63:144-152.

80. JARVIS, F. G., AND M. J. JOHNSON. 1949. Aglyco-lipide produced by Pseudomonasaeruginosa. J. Am. Chem. Soc. 71:4124-4126.

80a. KANESHIRO, T., AND A. G. MARR. 1961.Ci8-9, 10-methylene hexadecanoic acid fromthe phospholipids of Escherichia coli. J.Biol. Chem. 236:2615-2619.

80b. KATES, M., A. C. ALLISON, D. A. J.TYRRELL, AND A. T. JAMES. 1961. Lipids ofinfluenza virus and their relation to thoseof the host cell. Biochim. Biophys. Acta52:455-466.

81. KAUFMANN, H. P. 1958. Analyse der Fetteund fett Produkte. Springer-Verlag, Berlin.

82. KENNEDY, E. P. 1957. Metabolism of lipides.Ann. Rev. Biochem. 26:119-148.

83. KNAYSI, G. 1951. Chemistry of the bacterialcell, p. 1-27. In C. H. Werkman and P. W.Wilson [ed.], Bacterial physiology. Aca-demic Press, Inc., New York.

84. KODICEK, E. 1956. The effect of unsaturatedfatty acids, of vitamin D and other sterolson gram-positive bacteria, p. 401-406. InG. Popjak and E. Le Breton [ed.1, Bio-chemical problems of lipids. IntersciencePublishers, Inc., New York.

85. LAMANNA, C., AND M. F. MALLETTE. 1959.Basic bacteriology, 2nd ed. The Williams& Wilkins Co., Baltimore.

86. LAW, J. H. 1960. Glycolipids. Ann. Rev.Biochem. 29:131-150.

87. LAW, J. H. 1961. Lipids of Escherichia coli.Bacteriol. Proc. 129.

88. LAW, J. H., AND R. A. SLEPECKY. 1961. Assayof poly-g3-hydroxybutyric acid. J. Bacteriol.82:33-36.

89. LEDERER, E. 1958. Glycolipids of bacteria,plants, and lower animals, p. 119-146. InColloqium der Gesellschaft fur physiol.Chemie 8. Springer-Verlag, Berlin.

90. LEDERER, E., AND M. LEDERER. 1954. Chro-matography. Elsevier Publishing Co., NewYork.

91. LEMOIGNE, M., B. DELAPORTE, AND M.

CROSON. 1944. Contribution a l'6tudebotanique et biochimique des bacteriens dugenre Bacillus. Ann. inst. Pasteur 70:224-233.

92. LEMOIGNE, M., AND H. GIRARD. 1943. Re-serves lipidiques 0-hydroxybutyriques.Compt. rend. 217:557-559.

93. LICHSTEIN, H. C. 1960. Microbial nutrition.Ann. Rev. Microbiol. 14:17-42.

94. LIPSKY, S. R., AND R. A. LANDOWNE. 1960.Gas chromatography. Biochemical applica-tions. Ann. Rev. Biochem. 29:649-668.

95. LIu, T. Y., AND K. HOFMANN. 1960. Bio-synthesis of lactobacillic acid. FederationProc. 19:227.

95a. Liu, T-Y., AND K. HOFMANN. 1962. Cyclo-propane ring biosynthesis. Biochemistry1:189-191.

96. LOVERN, J. A. 1955. The chemistry of lipidsof biochemical significance. John Wiley &Sons, Inc., New York.

97. LOVERN, J. A. 1957. The phosphatides andglycolipids, p. 376-392. In W. Ruhland[ed.], Handbuch der Pflanzenphysiologie,v. VII. Springer-Verlag, Berlin.

98. LUCAS, R. A. 1951. The chemical nature offatty acids of Lactobacillus arabinosus.Ph.D. Thesis. Univ. of Pittsburgh, Pitts-burgh, Pa.

99. LURIA, S. E. 1960. The bacterial protoplasm:composition and organization, p. 1-34. InI. C. Gunsalus and R. Y. Stanier [ed.], Thebacteria. v. I. Academic Press, Inc., NewYork.

100. LYNEN, F. 1955. Lipide metabolism. Ann.Rev. Biochem. 24:653-688.

101. LYNEN, F. 1959. Participation of acyl-CoA inin carbon chain biosynthesis. J. CellularComp. Physiol. 54 (Suppl. 1):33-49.

102. MCQUILLEN, K. 1960. Bacterial protoplasts,p. 249-359. In I. C. Gunsalus and R. Y.Stanier [ed.], The bacteria. v. I. AcademicPress, Inc., New York.

103. MEAD, J. F., AND D. R. HOWTON. 1960. Radio-isotope studies of fatty acid metabolism.Pergamon Press, New York.

104. MEIKELJOHN, R. A., R. J. MEYER, S. M.ARONOVIC, H. A. SCHUETTE, AND V. W.MELOCHE. 1957. Characterization of longchain fatty acids by infrared spectroscopy.Anal. Chem. 29:329-334.

105. MURRAY, K. E. 1955. Low pressure fractionaldistillation and its use in the investigationof lipids. Progr. in Chem. Fats Lipids,3:243-273.

106. MURRAY, R. G. E. 1960. The internal struc-ture of the cell, p. 35-96. In I. C. Gunsalus

4451962]

WILLIAM M. O'LEARY

and R. Y. Stanier [ed.], The bacteria. v. I.Academic Press, Inc., New York.

107. NIEMAN, C. 1954. Influence of trace amountsof fatty acids on the growth of microorgan-isms. Bacteriol. Rev. 18:147-163.

108. NUNN, J. R. 1952. The structure of sterculicacid. J. Chem. Soc. p. 313-318.

109. O'CONNOR, R. T. 1960. Spectral properties,p. 380-498. In K. S. Markley [ed.], Fattyacids. 2nd ed., part 1. Interscience Pub-lishers, Inc., New York.

110. OGINSKY, E. L., AND W. W. UMBREIT. 1959.An introduction to bacterial physiology.2nd ed. W. H. Freeman and Co., San Fran-cisco.

111. O'LEARY, W. M. 1957. Bacterial metabolismof unsaturated fatty acids. Ph.D. Thesis.Univ. of Pittsburgh, Pittsburgh, Pa.

112. O'LEARY, W. M. 1959. Studies of the utiliza-tion of C14-labeled octadecenoic acids byLactobacillus arabinosus. J. Bacteriol.77:367-373.

113. O'LEARY, W. M. 1959. Involvement of methi-onine in bacterial lipid synthesis. J. Bac-teriol. 78:709-713.

114. O'LEARY, W. M., AND K. HOFMANN. 1957.Short chain monoethenoid fatty acids aspossible precursors in bacterial biosynthe-sis of cis-vaccenic acid. Federation Proc.16:228.

114a. O'LEARY, W. M. 1962. On the fatty acids ofpleuropneumonialike organisms. Biochem.Biophys. Research Commun. 8:87-91.

115. ORR, C. H., AND J. E. CALLEN. 1959. Recentadvances in the gas chromatographicseparation of methyl esters of fatty acids.Ann. N. Y. Acad. Sci. 72:649-665.

116. PECSOK, R. L. 1959. Principles and practiceof gas chromatography. John Wiley &Sons, Inc., New York.

117. PORTER, J. R. 1946. The chemical composi-tion of microorganisms, p. 352-450. InBacterial chemistry and physiology. JohnWiley & Sons, Inc., New York.

118. PUSTOVALOV, V. L. 1956. Higher fatty acidsof the diphtheria bacillus. Biochemistry(U. S. S. R.) 21:33-42.

119. RALSTON, A. W. 1948. Fatty acids and theirderivatives. John Wiley & Sons, Inc.,New York.

120. RYHAGE, R., AND E. STENHAGEN. 1960. Massspectrometry in lipid research. J. LipidResearch 1:361-390.

121. SAITO, K. 1960. Chromatographic studies onbacterial fatty acids. J. Biochem. (Tokyo)47:699-709.

122. SAITO, K. 1960. Studies on bacterial fatty

acids; the structure of subtilopentadec-anoic and subtiloheptadecanoic acids. J.Biochem. (Tokyo) 47:710-719.

123. SALTON, M. R. J. 1956. Bacterial cell walls,p. 81-110. In E. T. C. Spooner and B. A. D.Stocker [ed.], Bacterial anatomy. Cam-bridge Univ. Press, New York.

124. SALTON, M. R. J. 1960. Surface layers of thebacterial cell, p. 97-151. In I. C. Gunsalusand R. Y. Stanier [ed.], The bacteria. v. 1.Academic Press, Inc., New York.

124a. SCHEUERBRANDT, G., AND K. BLOCH. 1962.Unsaturated fatty acids in microorgan-isms. J. Biol. Chem. 237:2064-2068.

125. SCHLENK, H. 1954. Urea inclusion compoundsof fatty acids. Progr. in Chem. Fats Lipids2:243-267.

126. SCHLENK, H., J. L. GELLERMAN, J. A. TILLOT-SON, AND H. K. MANGOLD. 1957. Paperchromatography of lipides. J. Am. OilChemists' Soc. 34:377-386.

127. SEGAL, W., AND H. BLOCH. 1956. Biochemicaldifferentiation of Mycobacterium tuberculo-sis grown in vivo and in vitro. J. Bacteriol.72:132-141.

128. SIFFERD, R. H., AND R. J. ANDERSON. 1936.Uber das Vorkommen von Sterinen inBakterien. Z. physiol. Chem. 239:270-272.

129. SILBERMAN, R., AND W. L. GABY. 1961. Theuptake of amino acids by lipids of Pseu-domonas aeruginosa. J. Lipid Research2:172-176.

130. SKEGGS, H. R., L. D. WRIGHT, E. L. CRES-SON, G. D. MACRAE, C. H. HOFFMAN,D. E. WOLF, AND K. FOLKERS. 1956. Dis-covery of a new acetate-replacing factor.J. Bacteriol. 72:519-524.

131. SMITHIES, W. R., N. E. GIBBONS, AND S. T.BAYLEY. 1955. The chemical compositionof the cell and cell wall of some halophilicbacteria. Can. J. Microbiol. 1:605-613.

132. SPERRY, W. M. 1955. Lipide analysis, p.83-111. In D. Glick [ed.], Methods of bio-chemical analysis. v. 2. Interscience Pub-lishers, Inc., New York.

133. STADTMAN, E. R. 1954. Studies on the bio-chemical mechanism of fatty acid oxida-tion and synthesis. Record Chem. Progr.15:1-17.

134. STEPHENSON, M. 1949. Bacterial metabolism.3rd ed. Longmans, Green and Co., Ltd.,London.

135. STODOLA, F. H. 1958. Chemical transforma-tions by microorganisms. John Wiley &Sons, Inc., New York.

136. STUMPF, P. K. 1960. Lipid metabolism. Ann.Rev. Biochem. 29:261-294.

446 [VOL. 26

FATTY ACIDS OF BACTERIA

137. TAMURA, G. 1956. Hiochic acid, a new growthfactor for Lactobacillus homohiochi andLactobacillus heterohiochi. J. Gen. Appl.Microbiol. 2:431-434.

138. TAMURA, G., AND K. FOLKERS. 1958. Identityof mevalonic and hiochic acids. J. Org.Chem., 23:772.

139. TAUSIG, F. 1955. The chemical nature offatty acids of microbiological origin. Ph.D.Thesis. Univ. of Pittsburgh, Pittsburgh,Pa.

140. TRAUB, A., AND H. C. LICHSTEIN. 1956.Cell permeability. A factor in the biotin-oleate relationship in Lactobacillus arab-inosus. Arch. Biochem. Biophys. 62:222-233.

141. VELICK, S. F. 1944. The chemistry of Phy-tomonas tumefaciens. III. Phytomonic acid,a new branched chain fatty acid. J. Biol.Chem. 152:535-538.

142. VELICK, S. F. 1944. The chemistry of Phy-tomonas tumefaciens. IV. Concerning thestructure of phytomonic acid. J. Biol.Chem. 156:101-107.

143.' VENDRELY, R., AND R. TULASNE. 1953.Chemical constitution of the L-forms ofbacteria. Nature 171:262-263.

144. VOGEL, A. I. 1956. A textbook of practicalorganic chemistry including qualitativeorganic analysis. 3rd ed. Longmans, Greenand Co., London.

145. WAKIL, S. J. 1961. Mechanism of fatty acidsynthesis. J. Lipid Research 2:1-24.

146. WEIBULL, C. 1957. The lipids of a lysozymesensitive Bacillus species (Bacillus "M").Acta Chem. Scand. 11:881-892.

147. WEIDEL, W., AND J. PRIMOSIGH. 1958. Bio-chemical parallels between lysis by virulentphage and lysis by penicillin. J. Gen.Microbiol. 18:513-517.

148. WHEELER, D. H. 1954. Infrared absorptionspectroscopy in fats and oils. Progr. inChem. Fats Lipids 2:268-291.

149. WILLIAMS, V. R., AND E. A. FIEGER. 1946.Oleic acid as a growth stimulant for Lacto-bacillus casei. J. Biol. Chem. 166:335-343.

150. WILLIAMS, V. R., AND E. A. FIEGER. 1949.Further studies on lipide stimulation ofLactobacillus casei. II. J. Biol. Chem.177:739-744.

151. WILLIAMSON, D. H., AND J. F. WILKINSON.1958. The isolation and estimation of thepoly-j3-hydroxybutyrate inclusions of Ba-cillus species. J. Gen. Microbiol. 19:198-209.

152. WOLLISH, E. G., M. SCHMALL, AND M. HA-WRYLYSHYN. 1961. Thin-layer chromatog-raphy. Anal. Chem. 33:1138-1142.

153. WOODBINE, M. 1959. Microbial fat. Micro-organisms as potential fat producers.Progr. in Ind. Microbiol. 1:181-245.

154. WREN, J. J. 1960. Chromatography of lipidson silicic acid. J. Chromatog. 4:173-195.

1962] 447