Significance and taxonomic value of iso and anteiso monoenoic fatty acids and branded beta-hydroxy acids in Desulfovibrio desulfuricans

(1)

Copyright01977 American Society for Microbiology PrintedinU.S.A.

Significance

and Taxonomic Value of Iso and Anteiso

Monoenoic Fatty Acids

and

Branched

f8-Hydroxy

Acids in

Desulfovibrio

desulfuricans

JAAP J. BOON,* J. W. DR LEEUW, G. J. V. D. HOEK, AND J. H. VOSJAN

Department of Chemistry and Chemical Engineering,DelftUniversityof Technology,de Vries van Heystplantsoen 2, Delft,* and Netherlands Institute for Sea Research, 't Horntje, Texel,TheNetherlands

Received for publication 30 November 1976

The fatty acids obtained from extractable lipids of the anaerobic sulfate

bacteriumDesulfovibrio desulfuricans wereidentified. Saturated and

monoe-noic iso (C15-C01) and anteiso (C15, C17) fattyacids andsaturated normal

(C14-C1,)

and monoenoicnormal

(C16,

C08) fatty acidswere showntobe present by

capillary gas chromatography-mass spectrometry. Iso and anteiso 8-hydroxy

fattyacidswereanalyzedastrimethylsilylethersinthesameway.Theposition

ofmethyl branchesinthe monoenoic fattyacidswasdetermined from

character-isticfragmentions inthemass spectraof their methylesters.Disilyloxymethyl

esters, prepared by derivatization of themono unsaturated methyl esters and

analyzed by capillary gas chromatography-mass spectrometry, provided the

position ofdouble bonds. The monoenoic fatty acids identifiedinthisway were

normal

(A7-C16:1,

A9-C16:1,

A9-C18:1,

AlO1-Cl8:),

iso

(A7-C05:1,

A9-C06:1, A9-C17:1,

A11-C18:1,

Al1-C19:

l), andanteiso

(A7-C15:,

A91-C17

). Iso

A9-C17:1

fattyacid ispresent asthe

major

component. The occurrenceof these monoenoic fattyacidsinthis

bacterium isof taxonomicalimportance.

Desulfovibrio species are omnipresent bacte-ria in sediments of the marine environment.

Our interest in their fatty acids was aroused

after analysis ofa recent marine sediment in

which sulfate-reducing bacteria were

impor-tant links in the final decomposition of the

organicmatter. Thepresence of monoenoic iso

C05and iso C17 fatty acids (4)and ofa setofiso

and anteiso

(3-hydroxy

fattyacids (5) in these

sediments prompted research into the lipids of

Desulfovibrio desulfuricans, seen as a possible source of theseacidsinthissedimentary envi-ronment. Moreover, the occurrence ofiso and

anteiso monoenoicfatty acidsamongthemajor

fatty acids ofD.desulfuricans has considerable

taxonomic and biochemical significance.

The iso andanteiso saturated fatty acidsare

widely foundin gram-positivebacteria.

Gram-negative bacteria very often contain

straight-chain fatty acids with one double bond. Such

characteristic features in the alkyl chain of

fatty acidsareusedasaguidetoclassification

in bacterial taxonomy (23). Unsaturated iso

andanteisofatty acidsare stillrare in nature. Sofar, theyhave been foundinphospholipidsof the aerobic gram-positive Bacillus cereus (15), as anexcretionproduct of Myxococcus xanthus

(18),andas anacylconstituant in two

antibiot-ics (25).

Their distribution may prove to be more

widespreadinthe future, whenmoreadvanced

techniques, such as high-resolution gas

chro-matography and combined gas

chromatogra-phy-mass spectrometry, find theirway among

taxonomically orientedbacteriologists. An

ex-clusive presence as

major

fatty acids in

Desul-fovibrio species would point,onthe other hand,

to averyspecial phylogenetic position with

re-spect toother microorganisms.

Desulfovibrio species are thought to be

prim-itive incharacter. Anearlyoffspringfrom the

phylogenetic tree is indicated by its

cyto-chromes (10) and primitiveenergy metabolism

(14, 29). Desulfovibrio species are able to

bio-synthesize iso and anteiso precursors for

pro-cessing along the anaerobic chain elongation

pathway, as is indicated by the double-bond

position in their fatty acids. This points to a

dehydrase enzyme system (3) able to handle

branched-chainsubstrates. Thepresenceoftwo

enzyme systems in one bacterium which are

foundtooperateseparately in otherstrains of

bacteria could be anotheroneof those "vestiges

ofancestral biochemistry" that havebeen

enu-merated forthe sulfate-reducingbacteria (17).

MATERIALS AND METHODS

Organismand culture conditions. A strain of the

sulfate-reducing bacteria D. desulfuricans was iso-lated from the anaerobic black sediment of the

1183

at BIBLIOTHEEK TU DELFT on January 20, 2009

jb.asm.org

(2)

1184 BOON ET AL.

Dutch Wadden Sea (J. H. Vosjan, Ph.D. thesis,

University of Groningen, Groningen, The Nether-lands, 1975). ThecriteriaofPostgateandCampbell

(21)wereused for identification, and the puritywas

checked by usingthemethodof Postgate(20). Theorganismwasgrowninbatch cultureonthe

following medium: NaCl, 20 g; K2HPO4, 0.1 g;

MgCl2-6H20, 0.1 g; CaC02 2H20, 0.1 g; yeast

ex-tract, 1 g;tris(hydroxymethyl)aminomethane, 6.06 g;FeSO3* 6H20,0.003g;(NH4)2SO4,3.3g;Li-lactate,

3.3g;concentratedHCl, 3.8 ml; dissolved in1liter of

water and distilled over KMnO4. The pH was

ad-justedto7.5. Cultureswere grownanaerobicallyat

300C. The cells ofan18-liter batchculturewere

har-vested bycontinuouscentrifugationattheend of the exponential growth phase. The bacterial pellets (16.1 g) were freeze-dried, and the dried bacteria werestoredinvacuum-sealed ampoules.

Extraction and isolation of lipid material. Freeze-dried bacteria (1.882 g)wereextracted bya

modification of the Bligh and Dyer (2) procedure.

The bacterial residue were homogenized

subse-quently inaWaring blender in20mlofwater,40ml ofmethanol, and 80 ml of chloroform. After filtra-tiontheresiduewasextracted with50ml of

metha-nol. Centrifugation yieldedapellet, whichwas

ul-trasonically treated in 30mlofchloroform. The

com-binedextractsyieldedamonophasic solution,which wasseparated intotwolayers by addition of40ml of water.Thechloroformlayer, containing total lipids,

waswashedonce with water. A 0.269-gamountof

totallipidextractwasobtainedby solvent evapora-tion,usingarotatoryevaporator.

Fatty acids and hydroxy acidsin the total lipid

extract.Aportionof the totallipidextract (34 mg)

wassaponifiedin1 ml of1 NKOHin 95%ethanol

under reflux for 30 min. Thenon-acid fraction was

extracted withdiethyl ether. About13 mgof fatty

acidswasobtainedafter acidification and extraction

withdiethyl ether. Theseacidsweremethylated by

diazomethaneindiethyl ether.

Preparative thin-layer chromatography (TLC) yielded a monocarboxylic and ahydroxy

monocar-boxylic methylesterfraction(26).The monocarbox-ylic methylesterswereanalyzed bygas-liquid

chro-matography (GLC) and gas chromatography-mass

spectrometry (GC-MS).

Furthercharacterization was achieved by

appli-cation of the Os04-H2S method (see below)for the

derivatization of the unsaturated methyl esters. Diol methyl esters obtained after TLC separation ofthe reaction mixture weretreated with

bis(tri-methylsilyl)acetamide (Pierce Chemical Co.) and

analyzed by GC-MS as disilyloxy methyl esters.

Thehydroxy fatty acid methylestersweretreated

withbis(trimethylsilyl)acetamide, and the silyloxy

derivativeswereanalyzed by GLC and GC-MS.

0s04-H2S methodforpreparation of diolesters. About 5mgoffatty acid methylesterswastakenup

in 2ml ofdioxane-pyridine(8:1,vol/vol)ina

centri-fuge tube. Asmall crystal ofOSO4wasadded, and

thesolutionwasoccasionallyswirled byhand. The

yellow solution turned orange-red during 2 h of reaction in the dark at room temperature (about

2000). Diol fatty acid methylesterswereproduced

from osmate esters by bubbling gaseous H2S through the solution until saturation. Excess H2S wasremoved by bubbling nitrogen for 1h. Ablack osmium salt was precipitated from suspension by centrifugation. The supernatant was concentrated under a stream of nitrogen to a volume suitable for TLC and/or GC-MS after silylation by

bis-(trimethylsilyl)acetamide.

TLCseparation of reactionproducts. Separation ofreaction products resulting from the OsO4-H2S

and OsO4-Na2SO3 methods (8, 9) was achieved by

silicagel TLC (0.25 mm) usinghexane-ethylacetate

(7:3, vol/vol) asdeveloper. TLC bands were

visual-ized byiodine vaporand/or 0.05% rhodamine 6G in 95% ethanol. In this TLC system diol fatty acid methyl estershad an R, of 0.20, whereassaturated methyl esters had anRfof 0.80.

GLCandGC-MSconditions. GCwascarried out on a Perkin-Elmer 990 equippedwith 30-m, stain-less-steel open tubular columns (ID, 0.25 mm) coated with OV-101. Nitrogen wasused as carrier gas.

Monocarboxylicfattyacidmethylesters were

an-alyzedusing temperatureprogrammingfrom 140to

2500C at a rate of 2°C/min, 3-silyloxy fatty acid

methyl esters wereanalyzed using temperature pro-grammingfrom 170 to2600C at a rate of4°C/min, anddisilyloxy fatty acidmethyl esters were

ana-lyzedusing temperatureprogramming from 200to

300°C at a rate of2°C/min. Quantitative datawere

obtained by applying an Infotronics CRS 101 elec-tronicintegrator tothe gaschromatograph.

GC-MSwascarriedoutusingaVarian-MAT111

instrumentequippedwitheithera21-m, stainless-steel open tubular column(ID,0.25mm) coated with SP 2250 or a 30-m, stainless-steelopentubular

col-umn(ID, 0.25 mm)coatedwithOV-101.The carrier gasusedwasHe. Themassspectrometerwasused under standard conditions at 80 eV. Temperature

programming wascarried out underconditions

com-parabletothosedescribed above for the GLC

analy-sis.

Accurate massanalysis andmetastableanalysis.

For accurate massanalysis, the mixture of

disilyl-oxy methyl esters, obtained by the Os04-NaSO3

method, wasused. This mixture, consisting almost

exclusively of iso and anteiso 9,10-disilyloxy-C07

methyl ester and the normal 9,10-disilyloxy-C06

methylester, wasanalyzed bydirectprobeanalysis

on aMAT 311 Amass spectrometer.

Accurate masses determined were: m/e 111

C8H15 (experimental, 111.11712; calculated,

111.117370); mle 201 = C,,H25OSi (experimental, 201.17187; calculated, 201.174458);mle 109 = C8H13

(experimental, 109.10393; calculated, 109.101720); m/e 187 = C,OH23OSi (experimental, 187.15146; cal-culated, 187.158809).

Metastable analysis wasappliedtothis mixture

todetermine the origin of fragment ionsmle111,m/

e109,andm/e97both with thedefocusingtechnique and with direct analysis ofdaughterions (seeFig.

1).

Standard reference compounds. Some saturated normal, iso, and anteiso fatty acid methyl esters

werepurchasedfromApplied Science Laboratories.

jb.asm.org

(3)

VOL. 129, 1977

A mixture of branched fatty acid methyl esters (5), prepared from Mackerel (Scromber scombrus) lipids in our laboratory, was also used for comparison.

Mass spectra and equivalent drain length values were obtained from a mixture of normal, iso, and anteiso /8-silyloxyfatty acid methyl esters, synthe-sized from the corresponding fatty acid methyl esters (5). Normal 9-C18:1,

7-C16:1,

and A5-Ci4:1 fatty acid methyl esters were purchased from Ap-plied Science Laboratories. These were used for preparation of corresponding disilyloxy compounds, using the Os04-H2S method. Normal

A9-C171,

isO

A9-C17:1,

15-deutero-iso

A9-C17:1,

and anteiso A9-C17:1 methyl esters were synthesized, and their mass spectra were used for comparison (J. J. Boon etal., submitted for publication).

RESULTS

Fatty acids of D. desulfuricans. Figure 2 presentsthegas chromatogram of the fattyacid methyl esters obtained from the total lipid

ex-tract of D. desulfuricans. The numbers

indi-Defocusing 208 201 - o 111 157 DADI 185 < - 201 187 - K97 - 187 155 - .109

FIG. 1. Origin of some fragment ions from

di-silyloxy methylesters,determined by defocusingand the direct analysis of daughter ions (DADI) tech-niqueort aVarian MAT311Amassspectrometer.

22

u

FATTY ACIDS IN D. DESULFURICANS 1185

cated correspond with those in Table 1, listing the compounds and their relative abundances. The identification isbased on mass spectral data of the methyl esters and the disilyloxy methyl esters, which were prepared to deter-mine the double-bondpositions in the

monoe-noic fatty acid methyl esters. Iso and anteiso methyl branches in the saturated methyl esters canbe judged from the mass spectral databy

theenhancedintensity of[M+-43]and [M+-57],

the intensity reversal of the fragment ions [M+-29] and[M+-31],and the fragment ions [M+-61] and [M+-79] present in the mass spectra of

an-teiso methyl esters (19). Equivalent chain

length values aided in the identification.

Straight-chain monoenoic methyl esters are characterized by their parent ion and the high-intensity fragment ions [M+-32], [M+-74], and

[M+-116] (22). Branched-chain monoenoic

methyl esters present a somewhat different

mass spectrum (see Fig.3A andB). The isoC,7,:

methyl ester is also characterized by m/e 227

[M+-55],m/e 195[M+-87],andm/e177 [M+-105].

The anteiso C17:1 methyl ester mass spectrum shows a similar set offragpent ions but is 14 mass units lower. This ispointed out bym/e213

[M+-69], mle 181 [M+-101], and m/e 163

[M+-119]. Comparable fragment ions are found in the mass spectra of iso

C151,

anteiso

C15:1,

isO

C16:1,

and iso

C18:1.

The mass spectrum of iso A11-Ci7: methyl esterisolated from M. xanthus(18) presents the

samecharacteristics.Theexplanationfor

frag-ment ion [M+-87], considered by theseauthors

II I I I

2300 220° 21 200- 1i 1.00 1700 1°

FIG. 2. Gas chromatogram of the fatty acid methyl esters obtained from the total lipid extract of D.

desulfuricans.Numbered peaks are listed in Table 1. A30-miOV-101-coatedcapillary column was used for

analysis, and column temperature was programmed from 160 to 250°C at a rate of 2°C/min.

jb.asm.org

(4)

TABLE 1. Fatty acids of D. desulfuricans

Compound Systematicname Relative abundance Peak no. inFig.2

NormalC,4:0 Tetradecanoic 0.008 1

IsoA7-C1,: 13-Methyltetradec-7-enoic 0.051 2

AnteisoA7-C,5,1 12-Methyltetradec-7-enoic 0.013 3

Iso C150 13-Methyltetradecanoic 0.583 4

AnteisoC15:0 12-Methyltetradecanoic 0.281 5

NormalC15:0 Pentadecanoic 0.006 6

IsoA7-C,6:1 14-Methylpentadec-7-enoic 0.002 7

IsoA9-C16:1 14-Methylpentadec-9-enoic 0.028 8

IsoC15:0 14-Methylpentadecanoic 0.058 9

NormalA7-C,161 Hexadec-7-enoic Unresolved

trace amount 1

Normal

A9-CM16

Hexadec-9-enoic 0.150

NormalC,0.0 Hexadecanoic 0.503 11

IsoA9-C17:1 15-Methylhexadec-9-enoic 1.000 12

AnteisoA9-C17:1 14-Methylhexadec-9-enoic 0.260 13

IsoC17.0 15-Methylhexadecanoic 0.447 14

AnteisoC17:0 14-Methylhexadecanoic 0.238 15

Normal

C17:0

Heptadecanoic 0.027 16

Unknown 17

Iso All-C18.1 16-Methylheptadec-11-enoic 0.004 18

Iso C,80 16-Methylheptadecanoic 0.002 19

NormalA9-C8:, Octadec-9-enoic

0.071

20

NormalAll-C,8., Octadec-11-enoic 21

NormalCl,8 Octadecanoic 0.142 22

Iso ll-C19.1 17-Methyloctadec-11-enoic 0.004 23

AnteisoA11-C19.1 16-Methyloctadec-11-enoic 0.002 24

IsoC,,k0 17-Methyloctadecanoic 0.001 25 69 100 90 70-7' 60-40 152 30 Ii166 20 ~ ~206 250 A 282 0L 100 -90 -80 -70 -60 -50 -40 -30 20 -10 -0 *0 60 80 100 120 1L 160 180 2 220 240 260 290 3M 55 163 181 203 21 AA.A250 40 60 80 100 120 140 160 180 200 220 2.0 260 290

FIG. 3. Massspectrum of iso

A9-C,7,:

fatty acid methyl ester (A) andanteiso

A9-C,,7:

fatty acidmethyl ester (B). Both mass spectra were obtained by capillary GC-MS of methyl esters obtainedfromthe total lipid extract

ofD.desulfuricans.

jb.asm.org

(5)

to be loss of

CH,OCOCH2CH2,

is to be

criti-cized. Itdoes not explainthe loss of 101 mass

units in the case ofthe anteiso methyl ester.

Evidence from direct analysis of daughter ions,

appliedto synthetic standards and the

15-deu-tero-iso

A9-C,7,1

methylester points to another

mechanism of fragmentation: the tertiary H or

D (in the 15 position) is transferred tothe

car-boxyl end ofthe molecule (J. J. Boon et al.,

submitted for publication). This initial

rear-rangement is followed by the loss of 55 or 69

atomic mass unitsdependent on, respectively,

the iso or anteiso structure of the unsaturated

fatty acid methyl ester. This isshown in Fig.

3AandBby fragments

mle

227and

mle

213.

Methanol and water are lost subsequently

fromthese fragment ions, thus generatingmie

195andmle177 inthe caseof the isocompound

and mle 181 and mle 163 in the case ofthe

anteisocompound. Inthis way the isoand

an-teiso structure of branched monoenoic fatty

acidmethyl esters canbedeterminedfrom the

presenceofdefinitefragment ions,whereas in

case of the corresponding saturated esters

in-tensity differences from some fragment ions

must beused for theidentification. Itshouldbe

notedthat[M+-61]and[M+-79]arealso present

in the mass spectra of anteiso

mono-unsatu-rated methyl esters.

Hydroxyacidsof D.desulfuricans. Hydroxy

fattyacids were found in the acid fraction

ob-tained after alkaline hydrolysis of the total

lipid extract. Figure 4 presents the gas

chro-matogram of thehydroxy acids analyzed as

sil-yloxy methyl esters. Table 2 lists the

com-pounds identified.

The mass spectra agree with those of

stan-dards (11). Iso, anteiso, and normal

f3-silyloxy

methyl esters cannot be distinguishedbytheir

mass spectra(5). Comparison with ECLvalues,

obtained from a mixture of iso and anteiso

13-silyloxy methyl esters, aided in the

identifica-tion. The

a,,f-unsaturated

methyl ester

identi-fied is considered to be the result of elimination

of water from the(8-hydroxymethyl esters after

TLC separation. Theother methyl esters found

intrace quantity in this fraction are considered

to be impurities resulting from tailing on the

TLC plate. The relative amount of the

83-hy-droxy fatty acids as compared to the fatty acids

issmall.

Usefulness of the disilyloxy derivatives. Un-equivocal determination of the double-bond

po-sition in mono-unsaturated methyl esters is

possibleby modification of thedoublebond with OSO4 and ultimate analysis of the reaction

products asdisilyloxymethyl estersby GC-MS.

The orginal double-bond position is marked by high-intensity mass spectral fragment ion

peaksgeneratedas a result of the cleavage of

the carbon bond between the two silyloxy

TABLE 2. Hydroxy fatty acids of D. desulfuricans

Compound P_ir f-OH-isoC150 ... ,f-OH-anteiso C15:0 ... 13-OH-isoC16:0 ...

,1-OH-normal C16:0

... 13-OH-isoC,,O ... 13-OH-anteiso C17. ... ,8-OH-normal C180 ...

a,,8-Unsaturated methyl ester ...

Normal C16:0methyl ester ... Normal C18:omethyl ester ...

3 10 'eakno. nFig.4 3 4 5 6 8 9 10 1 2 7 I I I 270 250 230 210 10S me

FIG. 4. Gas chromatogram of the1-hydroxyacids ofD.desulfuricans analyzed astrimethylsilyloxymethyl esterson a 30-mOV-101-coated capillary column. Column temperature was programmed from 170 to 270°C at

arateof4°CImin.Numbered peaks are listed in Table 2.

VOL. 129, 1977

jb.asm.org

(6)

groups (8, 11). So far, this method has been

applied to single pure compounds of various kinds (11, 16, 31). Weappliedthis modification reaction to a mixture ofsaturated and

mono-unsaturatedmethylesters.Animportant modi-fication of the earlier procedure used is the reduction of osmates. The

Na2SO,-methanol-watermethodwasreplacedbyareductionwith

H2S(1)toavoid the loss of volatilecompounds,

toincreaseeaseofhandling,andtomake

possi-bleascalingdowntosmall samples. The

reac-tion mixtureobtainedcanbeseparatedbyTLC into saturated methyl esters and diol methyl

esters or, alternatively, can be analyzed as

such. The latterapproachsavestimeand vola-tilecompoundsbut isdependentonthe compo-sitionof the mixtureanalyzed(unlessthemass

spectraldataacquisitioniscomputerized).Both

approacheshave beenusedsuccessfullyinthis work onDesulfovibrio acids.

In the courseof the mass spectral examina-tion ofthe disilyloxy compounds, nodirect ob-vious differences were observed between the

mass spectra of normal and branched-chain

compounds. However, in Fig.5, comparingthe normaland iso9,10-disilyloxy methylester,

anenhancement offragmentmle 97isobserved in the mass spectrum of the iso compound.

This phenomenon is also found in the other brancheddisilyloxy compounds (Table 3).

100 90 60 70 60 50 1.0 30 20 10-I 0 100 90 so0-70 60 50 1.0 30 20 1 0-0 1.0 s0 100 120 160 260 220 240 260 3 330 3 420

FIG. 5. Massspectrum ofnormal9,10-disilyloxyC,16 methylester (A)and iso9,10-disilyloxy C16 methyl

ester(B).

Itcould be shownbyaccuratemass measure-mentsand theapplicationofdefocusingandthe directanalysisofdaughterionstechniquethat

mle 187 and mle 201 produce, mle 97 and mie

111 (seeFig. 1),respectively. Fragmentionsml e 187 and mle 201 are ions consisting of the

alkyl partofthedisilyloxycompound contain-ing one trimethylsilyloxy group. Preterminal

branching in this alkyl part is apparently

enough toenhance the loss oftrimethylsilanol

(90 massunits) fromthesefragmentions. DISCUSSION

D. desulfuricans can becharacterizedbyits

straight-chain and branched-chain fatty acids.

Preliminary results by TLC and field

desorp-tionMS indicate that thesefattyacidsare pres-ent in phospholipids, plasmalogens, and as

acylglucosaminederivatives. Thephospholipid

composition is in qualitative agreement with data on otherDesulfovibrio species (19). The branched-chain fattyacids of D. desulfuricans

are the most characteristic part of its fatty

acids. About 34% of allfattyacidspresentshow the unusual combination ofmethyl branching

and unsaturation in thealkyl chain.

The determination of the double-bond

posi-tion in thesefatty acids wasconsidered

neces-sary tounderstandaspectsof theirbiosynthesis

andtoincrease theirtaxonomical value.

There-73 A 155 97 109 I67 259 ES 1.13 1.0 60 9 0 21010/0 26 2202 2 60 260 3M 320 3W1.o 0 1.30 73 97 B 155 I ~~~~~~~~~~~~~~~259 109 167 11

jb.asm.org

(7)

TABLE 3. Major fragment ions and relative intensity ofsome normal and branched-chain disilyloxy fatty acid

methylesters

I

RV RRX

(CH3)3SiO

OSi(CH3)3

OCH3

Ry, Ry-90 R.

Originalcomp,ound Fragment Intensity Fragment Intensity Fragment Intensity

ion (%) ion (%) ion (%)

Branched" A70-C1: 201 36 111 69 1.91 231 57 Branched A9-C,6: 187 39 97 91 2.34 259 48 Normal A9-C16:1 187 79 97 33 0.42 259 63 BranchedA9-C17:1 201 39 111 60 1.54 259 69 BranchedA11-C18:1 187 11 97 53 4.82 287 6 Normal A9-C18.1 215 13 125 4 0.31 259 10 Normal All-C18:1 187 47 97 35 1.34 287 14 Branched

A11-C,9:

201 9.7 111 40 4.13 287 11

aBranched =isooranteiso.

fore,wemodifiedanexistingtechniqueusedfor

determination ofdouble bonds in single pure

compoundsinsuchawaythatitcouldbe used

for mixtures of fatty acids. The anaerobic

me-tabolism ofDesulfovibrio species indicates that the so-called "anaerobicpathway of monoenoic

fattyacid synthesis"isthebiosynthetic

mecha-nism to consider. Bloch and co-workers

thor-oughlyinvestigatedthispathwayforastrainof

Escherichiacoli (K-12) (3) andforClostridium kluyveri (25). Other microorganisms

investi-gated byusing thispathwayhavebeen

summa-rized in areview (12).

The usual fatty acid synthesis sequence (30)

is split at the 8-hydroxy

C1o

level by a

D-,3-hydroxydecanoyl thioester dehydrase(7) in

cis-f3,y-decenoyl-ACP

and

trans-a,/3-decenoyl-ACP.

The

cis-,8,-y-decanoyl

derivative iselongated

andleadsto

A5C,2:l, A7-C14:1,

A9-Cl6:l,

and

All-Cl08:,

theso-calledcis-vaccenicacidfamily. The

doublebondsinthefatty acids ofDesulfovibrio

makecertainthat thisbiosynthetic principleis

operative not only on straight-chain, but also

onbranched-chainfattyacidsinthese bacteria.

Thepresenceof thebranched-chainvaccenic acidfamilyinDesulfovibriomusthaveadirect

relation to the properties of its

D-,3-hydroxy-decanoyl thioester dehydrase enzyme.

Assum-ing that a dehydrase similarto the one in E.

coli andC.kluyveri ispresent in D.

desulfuri-cans, then an

8-methyl-j8-hydroxydecanoyl-ACP is the precursor for anteiso

A7-C15:l

and

anteiso

A9-Cl7:

.The

9-methyl-f3hydroxydecano-yl-ACP is theprecursor foriso

A7-Ci5:,

iso

A9-C17:1, and

All-C19:l.

The precursors for iso

A7-C16:1

and that foriso

A&9-C16:1

and iso

1b11-C18:1

are the

10-methyl-,3-hydroxyundecanoyl-ACP

andthe

8-methyl-,8-hydroxynonanoyl-ACP,

re-spectively. The E. coli dehydrase (24) was

shown to be highly specific for the

straight-chain ,B-hydroxydecanoyl-ACP derivative; Cg

and

C,1

weremuch lessactive, whereas 08 and

C12were almost inertsubstrates.

InD. desulfuricans the dehydrase

preferen-tially attacks the normal8-OH-C10 compound, judged from the preponderance ofnormal

A9-C16:1

and normal

A11-C18:1.

The enzyme is

pre-sumably not hindered by the presence of the

methyl group in the

9-methyl-Clo

derivative,

whichleadsto alargerelative abundance of the

oddisocompounds.

The

8-methyl-Clo

compound (alternatively

"interpreted" by the enzyme as an

8-ethyl-Cg

compound) and the 8-methyl-C9 and the

10-methyl-C,,

derivatives are less suitable sub-strates.Onlyverysmallamountsof thenormal

9-C18:1

andnormalA7-C16: arepresent, which means that the 8-OH-C12 compound is a very

poorsubstrate. Theseconclusions, basedon

in-directevidence,need further confirmationwith

thepurified dehydrase.Thepresenceofseveral

dehydrasesfor the generation ofstraight- and

branched-chainfatty acidscannotberuledout.

Little comparative work has been done

con-cerningthefattyacids ofDesulfovibrio-species,

although thereis atrendinbacteriologyto use

fattyacidspectrafortaxonomicalpurposes(23). Unfortunately, extracts from bacteria

some-timessuffer fromtheabsence of detailed

struc-tural identification. The excellent GLC traces

of thefatty acids ofDesulfovibriovulgaris

Hil-VOL. 129, 1977

jb.asm.org

(8)

1190

denborough, D. desulfuricans Essex 6, and D.

desulfuricansBerreSol inthework of Han (J.

C.-Y. Han, Ph.D. thesis, University of

Califor-nia, Berkeley, 1970), however, allow

compari-sonwithourresults.

The presence of the anaerobic pathway of monoenoic fatty acid synthesis was shown in

these bacteria by GC identification ofdouble

bonds in the straight-chain C16 and C08 acids,

using the corresponding standards. Although iso and anteiso

C15:0

and

C,7,:

fatty acidswere identified, the identification of several major

peaks, whichcorrespond in our view with the monoenoic isoand anteisofattyacids, was not

reported. The GLC traceof theHildenborough

strainshowsclose resemblanceto ourspeciesof

D. desulfuricans. A comparative study of the

fatty acids in several species of Desulfovibrio

and,preferably, alsoinspeciesof

Desulfotoma-culum (21) would be of taxonomical and

bio-chemical interest. This approach would enable

the evaluation of the abilityoftheseorganisms

to use branched-chain f8-hydroxy

C0,

C,o

and

Cl,

compoundsassubstrates for their

3-hydrox-ydecanoyl thioester dehydrase. The properties

ofthedehydraseenzyme systemitselfcanalso

beusedas ameans tounderstand the

relation-shipbetween the bacterialstrainsthatusethe

anaerobic pathway ofmonoenoicfattyacid

syn-thesis.

The distribution of this pathway is not

re-stricted to anaerobic microorganisms but

fol-lows evolutionary lines (24). This pathway has

beenshownto occurinaerobic, facultative

an-aerobic,anaerobic, andphotosynthetic

microor-ganisms, e.g., Pseudomonas fluorescens (24),

E.coliK-12(3),lactobacilli (27),pediococci(27),

C.kluyveri (24), Rhodospirillum rubrum (32),

and Rhodopseudomonas sphaeroides (32).

Branched-chain fatty acids have not been

re-portedamongthefatty acids of these

microor-ganisms. If their dehydrase enzyme systems

wereableto usebranched-chain substrates,one

could think ofarelationship withDesulfovibrio

species, whicharethoughttobe closeto "the"

ancestral mother microbe(17).

The composition of fatty acids present in

phospholipids, whichplayanimportantrolein

the cell membrane ofmicroorganisms,

deter-mines to alargeextent thefluidityofthe

mem-brane. Thisfluidityisnecessaryfor theproper

function of membrane-related metabolic

proc-esses (28). Several bacteriaareableto change

the degreeofunsaturation of theirlipidsupon

lowering the growthtemperatureby enhanced

desaturase activity (13). Anotherway to lower

themeltingtemperature oflipidsis toincrease

the amount of iso and anteiso fatty acids in

theselipids. Theamountofbranched saturated

andunsaturatedfattyacids inphospholipidsof

D.desulfuricans is surprisingly high inview of

the growthtemperature in the culture (30°0).

Several explanations for this phenomenon, such asinabilitytoregulate fattyacid composi-tion in the cell membrane lipids, a special structureof thecell wall, the influence of

cul-turefactors, etc., areunder investigation.

The branched-chain 18-hydroxy fatty acids

present aproblem inrespect totheiroriginand roleintheorganism. Theanalytical procedure

points to a free or esterified (via the carboxyl

group) occurrencein the totallipidextract. No

free hydroxy acids have been observed in the

neutral lipids after column chromatography.

,8-Hydroxy

fattyacidsarepresent in the cell

wall of gram-negative bacteria, usuallybound

tothe lipidAcomponent (6).Thislipidis

usu-allynot extractable by mildorganic solvents.

Further work is necessary to pinpoint these

hydroxyacidsin the bacterium.

ACKNOWLEDGMENTS

We gratefully acknowledge Chris N. Keynon and H. Duine fortheir valuablesuggestions aftercritically reading themanuscript. P. A. Schenck is thanked for the opportu-nity to carry outthis work in his laboratory.

LITERATURE CITED

1. Barton, D. H. R., and D. Elad. 1956. Colombo Root Bitterprinciples. Part I. The functional groups of columbin. J.Chem. Soc.,p.2085-2090.

2. Bligh,E. G., and W. J.Dyer.1959. Arapid method for totallipidextractionand purification. Can. J. Bio-chem.Physiol. 37:911-917.

3. Bloch, K. 1969. Enzymatic synthesis of monounsatu-rated fatty acids.Acc. Chem. Res. 2:193-202. 4. Boon, J. J., J. W.de Leeuw, and P. A. Schenck. 1975.

Organic geochemistry ofWalvis Baydiatomaceous ooze.I.Occurrence andsignificanceof thefatty acids. Geochim. Cosmochim. Acta 39:1559-1565.

5. Boon, J. J., F. de Lange, P. J. W.Schuyl, J. W. de Leeuw, and P. A. Schenck. 1977. Organic geochemis-try ofWalvisBaydiatomaceousooze.II. Occurrence and significance of the hydroxy fatty acids. In E. Campos (ed.), Advances in organic geochemistry 1975,Proceedingsofthe 7thInternational Meeting on Organic GeochemistryinMadrid.

6. Braun,V., and K. Hantke. 1974.Biochemistryof bacte-rial cell envelopes.Annu. Rev. Biochem. 43:89-121. 7. Brock, D. J. H., L. R. Kass, and K. Bloch. 1967.

(3-Hydroxydecanoylthioesterdehydrase.II.Modeof ac-tion. J. Biol. Chem. 242:4432-4440.

8. Capella,P., andC. M. Zorzut. 1968. Determination of doublebond positioninmonounsaturated fatty acid esters by mass spectrometry of their trimethylsil-yloxy derivatives. Anal.Chem.40:1458-1464. 9. Criegee,R. 1936.Osmiumsiure-esterals

Zwischenpro-dukte beiOxydationen.Justus Liebigs Ann. Chem. 522:75-96.

10.'Dickerson, R. E., and R. Timkovich. 1975. Cytochrome C,p. 397-547. InP.D.Boyer(ed.), The enzymes, vol. 11A.Academic Press Inc., New York.

11. Eglinton, G., D. H. Hunneman, and A. McCormick. 1968. Gaschromatographic-mass spectrometric stud-ies oflongchainhydroxy acids.I.Themass spectra of the methylesters trimethyl-silylethers of aliphatic

jb.asm.org

(9)

VOL. 129, 1977

hydroxy acids. A facile method of double bond loca-tion. Org. Mass Spectrom. 1:593-611.

12. Erwin, J. A. 1973. Comparative biochemistry of fatty acidsineukaryoticmicroorganisms, p. 41-145. In J. A. Erwin (ed.), Lipids and biomembranes of eukar-yotic microorganisms. Academic Press Inc., New York.

13. Fulco, A. J. 1969. The biosynthesis of unsaturated fatty acids by Bacilli. I. Temperature induction of the de-saturation. J. Biol. Chem. 244:889-895.

14. Gottschalk, G. 1968. The stereospecificity of the citrate synthase in sulfate-reducing and photosynthetic bac-teria. Eur. J. Biochem. 5:346-351.

15. Kaneda, T. 1972. Positional preference of fatty acids in phospholipids of Bacillus cereus and its relation to growth temperature. Biochim. Biophys. Acta 280:297-305.

16. Karlander, S. G., K. A.Karlsson,H. Leffler, A.Lilja, B. E.Samuelson,and G. 0.Steen. 1972.The struc-ture ofsphingomyelin of the honey bee (Apis melli-fera). Biophys. Biochim. Acta 270:117-131.

17. LeGall, J., and J. R. Postgate. 1973. The physiology of sulphate-reducingbacteria, p. 81-134. In A. H. Rose and D. W. Tempest (ed.), Advances in microbial physiology, vol. 10. Academic Press Inc., New York. 18. Noren, B., and G. Odham. 1973. Antagonistic effects of

Myxococcus zanthus on fungi. II. Isolation and char-acterisation of inhibitory lipid factors. Lipids 8:573-583.

19. Makula,R. A., and W. R. Finnerty. 1975.Phospholipid composition of Desulfovibrio species. J. Bacteriol. 120:1279-1283.

20. Postgate, J. R. 1965. Enrichment and isolation of sul-phatereducingbacteria, p. 190-205. In H. G. Schlegel (ed.), Anreicherungkultur und Mutantenauslese. GustavFischer Verlag, Stuttgart.

21. Postgate, J. R., andL. L. Campbell. 1966. Classifica-tionofDesulfovibriospecies,the nonsporulating

sul-DESULFURICANS 1191

fate-reducing bacteria. Bacteriol. Rev. 30:732-738. 22. Ryhage, R., and E. Stenhage. 1960. Mass spectrometry

in lipid research. J. Lipid Res. 1:361-390.

23. Shaw, N. 1974.Lipid composition as a guide to classifi-cation ofbacteria, p. 63-109. In D. Perlman (ed.), Advances in applied microbiology, vol. 17. Academic Press Inc., New York.

24. Scheuerbrandt, G., and K. Bloch. 1962. Unsaturated fatty acids in microorganisms. J. Biol. Chem. 273:2064-2068.

25. Shoji, J., and H. Otsuka. 1969. Studies on Tsushimycin. II. The structures of constituent fatty acids. J. Anti-biot. 12:473-479.

26. Skipski,V. P., A. F. Smolowne, R. C. Sullivan, and M. Barclay. 1965. Separation of lipid classes by thin-layer chromatography. Biochim. Biophys. Acta 106:386-396.

27. Uchida,K., and K. Mogi. 1972.Cellularfattyacid spec-traof Pediococcusspecies inrelationtotheir taxon-omy.J. Gen.Appl.Microbiol. 18:109-129.

28. Vessey,D.A.,and D. Zakim.1974.Membranefluidity andtheregulationofmembraneboundenzymes, p. 138-175. In E. Quagliarello, F. Palmieri, and T. P. Singer (ed.), Horizons inbiochemistry and biophys-ics, vol. 1. Addison-Wesley Publishing Co., Reading, Mass.

29. Vosian, J. H. 1970. ATP generationbyelectron trans-port inDesulfovibrio desulfuricans. Anthonie van Leeuwenhoek 36:585-587.

30. Wakil, S. J.(ed.). 1970. Lipidmetabolism. Academic Press Inc., NewYork.

31. Wolff, R. E., G. Wolff, and J. A. McCloskey. 1966. Characterisation of unsaturated hydrocarbons by mass spectrometry.Tetrahedron22:3093-3101. 32. Wood, B. J. B., B. W. Nichols, and A. T. James. 1965.

The lipids and fatty acid metabolismof photosyn-theticbacteria.Biochim. Biophys.Acta 106:261-273.

jb.asm.org