Biohydrogenation of Unsaturated Fatty Acids

8/17/2019 Biohydrogenation of Unsaturated Fatty Acids

1/6

THE JOUIWAL OF BIOLOGICAL CHEMIBTRY

Vol. 246, No. 16, Issue of Augu st 25, pp. 502.55030, 1971

Printedin U.S.A.

Biohydrogenation of Unsaturated Fatty Acids

VI. SOURCE OF HYDROGEN AND STEREOSPECIFICITY OF REDUCTION*

(Received for publicat,ion, March 9, 1971)

1. S. ROSENFELI) ANU S. B. TOVE~.

From the Department oj Biochemistry, North Camha State Uwiuersity, Raleigh, North Carolina 2760?’

SUMMARY

The biohydrogenation of either linoleic acid or cis-9, trans-

11 ,cis-13-octadecatrienoic acid (punicic acid) by Butyriuibrio

jbrisolvens results in the formation of trans-11-octadecenoic

acid. Incubation of whole cells with tritiated formate, triti-

ated succinate, and glucose labeled with tritium in various

positions failed to result in the labeling of the monoenoic acid

product. In contrast, experiments performed in DzO indi-

cated that deuterium was incorporated at the cis double

bond(s) reduced by the microorganism. This reduction,

which takes place stereospec ifically, was found to occur by

cis addition to the D side of cis-9, tram-1 1-octadecadienoic

acid, an intermediate in the biohydrogenation of linoleic acid.

The distribution of deuterium at the reduced carbon atoms

shows an isotope ef fec t and leads to the speculation that re-

duction occurs by addition of a proton and hydride ion medi-

ated by an unknown carrier.

The pathway of biohydrogen:rtion of linoleic acid by the

anaerobic rumen bacterium, Butyrivibrio Jibrisolvens, consists of

at least two reactions: (a) an initial isomerization to L-9,

trans-ll-octadecadienoic acid and (b) the subsequent hydrogena-

tion of this compound to trans-ll-octadecenoic acid (1, 2).

Partial purification of linoleic acid isomerase, the enzyme that

cntalyzes the isomerization, has been achieved and some of its

properties have been investigated. It shows marked speci ficit y

for a free carboxyl group and a cis-Q,cis-12 pentadiene system

(3). These studies were greatly facilitated by the finding that

this reaction takes place under aerobic conditions.

In contrast

to this, t,he hydrogenation reaction appears to be obligately an-

aerobic, and active cell-free preparations have been diff icul t to

* This work is a contribution from the Department of Bio-

chemis try, School of Agriculture and Life Sciences and School of

Physical and Mathematical Sciences. It is Paper 3421 of the Jour-

nal Series of the, North Carolina State University Agricultural

Experiment Station, Raleigh, North Carolina. This work was

sunnorted in oart bv Public Health Service Research Grant AM-

02483 from &e Naiional Institute of Arthritis and Metabolic

Disecxs. High resolution mass spectrometry was done at the

Research Triangle Institute Cenier for Mass Spectrometry under

Grant PR 330 from the Biotechnology Resources Branch of the

National Institutes of Health.

t To whom correspondence should be addressed.

prepare. This report deals with the hydrogenation reaction with

intact cells in which the source of hydrogen and stereospecificity

of the reduction of the double bond were investigated.

EXPERIMENTAL PROCEDURE

Bacterial Culture

B. fibr isolvens strain A-38 was grown and maintained as

pre-

viously described (2) except that the oxidation-reduction poten-

tial dye, resazurin, was not included and the media was gassed

with an atmosphere of oxygen-free 95% COZ and 50/O H, for 2

hours prior to inoculation.

The cells were harvested by centri-

fugation in 250~ml capped polypropylene bottles in a Sorvall

GSA rotor at 14,600O X g for 15 min.

Chlorella vulgar is was grown and maintained as described by

Harris and James (4).

Substrates

Linoleic and a-eleostearic acids were obtained from the Hormel

Institute . The tritiated substrate cis-9, trans-ll-[Q, lo-3H]

octadecadienoic acid was prepared by reduction of octadec-Q-

yn, trans-ll-enoic acid with tritium gas and was the generous

gif t of Dr. L. J. Morris, Unilever, Shambrook, Bedford, Eng-

land.

Punicic acid (c&Q, trans-11 ,cis-13-octadccatrienoic acid) was

isolated from the seed oil of Punica granatum (pomegranate)

purchased in a local market. The outer covers of the pome-

granates were removed and the fru it was allowed to soak in water

for 1 to 2 days. The frui t was then squeezed by hand to remove

the fleshy coating; and the small, hard, white seeds were dried

in a vacuum desiccator over PZOS. The seeds were ground

in a Wiley Mill and extracted under nitrogen with petro-

leum ether (b.p. 40-60”) in a Soxhlet apparatus for 24 hours.

The acid was isolated by low temperature crystallization as de-

scribed by Crombie and Jacklin (5). The white crystalline

product melted at 43” (lit. m.p. 40-42’) (5) and gave the ex-

pected ultraviolet spectrum with maxima at 264, 274 and 285

nm.

The alcohol derivative of punicic acid was prepared from the

methyl ester by treatment with

LiAIHt

(6).

The alcohol gave

the same absorption spectrum as punicic acid and migrated as

a single spot on thin layer chromatoplates of silica gel with

heptane-isopropyl ether-acetic acid (6 :4 :0.3).

The infrared ;spec-

trum exhibited characteristic peaks at 3600.0 cm-1 (OH) and

at 981.4 aud 932.0 cm-’ (cis, truns-conjugated oublebond sys-

tern).

No peaks were observed in t,he carbonyl region.

5025

y g ,

,

j g

http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/


2/6

5026

Source of Hydrogen and Stereospecificity of Reduction Vol. 246 No. 16

c1a320

Nuclidic mass

Calculated: 264.2453

Found : 264.2448

The ci.s-9, trans-11 , cis-13-octadecatriene was prepared by

LiAIHd reduction of the mesylate ester of the alcohol (7, 8).

The

hydrocarbon had an absorption spectrum identical with that of

punicic acid and gave a single spot when chromatographed on

silica gel plates with hexane as the solvent. When subjected to

gas-liquid chromatography, a single peak was observed. The

infrared spectrum showed no peaks in the carbonyl region, but

the same doublet, characteristic of the cis-trans double bond sys -

tem, was observed.

Nuclidic mass Calculated: 248.2504

Found : 248.2509

Deuterium oxide was supplied by Stohler Isotope Chemicals

and the acid hydrolysate of algae cells grown on D20 was obtained

from Merck.

Tritiated sodium formate, 2, 3-3H-succinic acid, and 5-3H-

glucose were obtained f rom Amersham-Searle. Glucose labeled

with tritium in positions 1, 2, 3, and 6 was obtained from New

England Nuclear.

The standard paraff ins, 9-nonadecene and %heptadecene, were

obtained from the Chemical Samples Company.

Methods

Incubations-A solution of 5 mg of the fa tt y acid or derivat ive

in benzene was added to a 125-ml Erlenmeyer flask and the

solvent was removed with a stream o f nitrogen. After the

benzene had evaporated, 12 ml of 0.05 M potassium phosphate

buffer, pH 6.6, containing 0.48 g of bovine serum albumin (Frac-

tion V) was added. Twelve millili ters of a bacterial suspension

in 0.1 M phosphate buffer, pH 6.6, were added and the flask was

stoppered with a rubber stopper equipped with two short glass

tubes, on which were placed 2-inch pieces of thin walled rubber

tubing. The flasks were placed in an ice bath and flushed with

hydrogen for 20 min, after which the rubber tubes were closed

with a pinch clamp. Incubation was carried out with gentle

agitation for 4 hours at 37”.

Undue exposure to air was avoided during the preparation of

the bacterial suspension.

Following centrifugation, the bacterial

pellet was suspended in 13 ml o f 0.1 1\~phosphate buffer, pH 6.6,

that had been thoroughly flushed with hydrogen. The tube

containing the cells was flushed with hydrogen for 5 min, stop-

pered, and shaken to disperse the bacteria. The suspension was

diluted with thoroughly gassed buffer such that a 1: 100 dilution

gave an absorbance of 1 at 420 nm.

When the tritium-labeled substrates were used, 100 PCi were

added as an aqueous solution to the buffered albumin. When

cis-9, truns-11[9, 10-3H]octadecadienoic acid was incubated, vol-

umes one-third the usual size were used.

In experiments in which DzO was used, the buffer solution

was evaporated to dryness and the buffer salts were dissolved in

the appropriate volume o f DZO.

In experiments conducted with the alcohol or paraffin deriva-

tive of punicic acid, the substrate was dispersed by sonic oscil-

lation (Branson) in a small amount of buffer prior to incuba-

tion.

Isolation of Reaction Products-Following incubation, the reac-

tion mixture was extracted according to the method of Dole (9).

The products of the fa tty acid substrates were methylated by

diazomethane and the monoenoic acids were isolated by chroma-

tography of their methyl esters on silicic acid-silver nitrate

columns (10, 11). In each case, a single component was ob-

served when examined by gas-liquid chromatography. When

the alcohol or paraffin derivat ives of punicic acid were used as

substrates, their hydrogenation products were separated by

chromatography on Florisil (12). The monoene paraff in prod-

uct was indicated by its retention time during gas-liquid chro-

matography with 9-nonadecene and S-heptadecene as standards.

The monoene alcohol was indicated by its cochromatography

with trans-11-octadecenol on silicic acid-silver nitrate thin layer

plates (13).

Stereospecijicity Studies-In these studies cis-9, trans-ll-

[9,10-3H]octsdecadienoic acid was used as the substrate. The

labeled trans-11-octadecenoic acid was isolated, methylated, and

reduced to methyl stearate by hydrazine (14). After saponifica-

tion, l-14C-stearic acid was added and the doubly labeled stearic

acid incubated with a suspension of Chlorella as described by

Morris et al. (15). The algal suspension was then extracted

with chloroform-methanol (2: 1)) and the methyl esters of the

fa tt y acids were prepared by transmethylation (16). Methyl

oleate and methyl linoleate were isolated by argentation column

chromatography (11). Each gave a single peak upon gas-liquid

chromatography.

To ensure that the tritium label had not moved during the hy-

drogenation of the Ag-bond, l-14C-labeled stearic acid was omitted

from the Chlorella incubation and the tritiated oleic acid was

isolated from the Chlorella suspension as previously described.

Carrier methyl oleate was added and reductive ozonolysis was

accomplished by the method of Edwards (17), except that the

2,4-dinitrophenylhydrazine reagent of Johnson (18) was used.

The dinitrophenylhydrazone derivat ives of the aldehyde and

aldehydo-ester fragments were separated by chromatography on

alumina (19). The purity was established by the single spot

obtained for each fragment when chromatographed on thin layer

plates of Microcel-T38 (20). To determine the tritium in each

fragment, the nonanal-dinitrophenylhydrazone and the methyl-

9-oxononanoate dinitrophenylhydrazone were completely oxi-

dized (21) and the tritiated water was absorbed in 20 ml of a

solution of 30% methanol in toluene that contained 6 g of Omni-

fluor (New England Nuclear) per liter.

Oxidative cleavage of the 3H-labeled methyl oleate to nonanoic

acid and monomethyl azelaic acid was accomplished according

to the procedure of Castle and Ackman (22). The nonanoic

acid was isolated by steam distillation and the monomethyl

azelaic acid was isolated by thin layer chromatography on

silica gel plates withheptane-isopropyl ether-acetic acid (6:4 :0.3).

The monocarboxylic acid was extracted with ether and trans-

ferred to a counting vial.

The spot corresponding to the mono-

methyl azelaic acid was scraped of f and the product was eluted

with methanol and counted.

Mass Spectrometry-Following extraction, methylation, and

isolation of the product of either a deuterated substrate or fa tty

acid substrate incubated in DzO, mass spectra were obtained by

means of a AEI-12 mass spectrometer.

The 11,12-dimethoxy methyl octadecanoate derivat ive o f the

methyl truns-11-octadecenoate obtained from the incubation of

linoleic acid with B. jibrisolvens in DzO was prepared and isolated

y g ,

,

j g



3/6

Issue of August 25, 1971

I. X. Rosenfeld and S. B. Tove

as described by Neihaus and Ryhage (23).

The monoene f rac-

tion from the incubation of punicic acid in D20 was reduced to

the paraffin via the alcohol and mesylate ester (7,8), as previously

described and oxidatively cleaved by a modified method of

Scheuerbrandt and Block (24). Since the paraff in was insoluble

in their reaction mixture, the solvent was removed and the

the following solutions were added per 5 mg of unsaturated hy-

drocarbon: 0.8 ml of t-butyl alcohol, 0.3 ml of a mixture of 0.02

M

Khln04 and 0.19

M

NaI04, 0.12 ml of 0.04

M

K&Ox, and

fina lly 0.6 ml of water. The flask was sealed and stirred for 2

hours at room temperature and the acid fragments were isolated

(24). Mass spectra of their methyl esters were obtained by

using the gas chromatographic inlet system of a model 9000

LKB mass spectrometer. A four-foot column of ethylene glycol

succinate-HaPOd was used with temperature programming.

Several scans were obtained for all samples and the peaks of

int’erest were corrected for natural abundance.

Gus-Liquid Chromatography--The methyl esters of the acids

obtained from incubations with linoleic acid, cr-eleostearic acid,

and the c&runs-conjugated acid mixture were analyzed by

gas-liquid chromatography. The paraf fins isolated from incu-

bation of B. jibrisolvens with cis-9, truns-11 , cis-13-octadecatriene

were also subjected to gas-liquid chromatography. An F and M

model 700 flame ionization instrument equipped with four-foot

columns of 10% diethylene glycol succinate on Chromosorb W

was used.

Other Analytical Procedures-Ester groups were determined by

the procedure of Snyder and Stephens (25).

Radioactiv ity was measured in a Packard Tri-Carb liquid

scintillation spectrometer by usin, 0‘ a scintillation solution of

Omnifluor (New England Nuclear) in toluene (4 g per liter).

Infrared spectra were measured in a Beckman IR-8 in carbon

disulfide solution.

Jloleculnr formulas were determined by accurate mass meas-

urement on a MS-902 mass spectrometer.

RESULTS

Hydrogenation of Punicic Acid---Linoleic acid isomerase, the

enzyme that catalyzes the first reaction in the biohydrogenation

pathway, has marked substrate specifi city requirements (3).

Since B. fibrisolvens was able to hydrogenate a mixture of cis-

frans conjugated dienes (A9~11,A1’J,1z,A8~10) l), it appeared that the

spec ific ity properties for the hydrogenation reaction were likely

to be less stringent.

Accordingly, the naturally occurring con-

jugated octadecatrienoic acid, punicic acid, with a cis-9, truns-

11 ,&s-13 double bond system seemed likely to serve as a sub-

strate. When punicic acid was incubated with the bacteria,

analysis of the methyl esters of the free fatty acids isolated from

the incubation mixture showed a complete disappearance of the

conjugated triene and the appearance of a peak coincident with

methyl oleate. After isolation of this product by argentation

chromatography, it was subjected to analysis by infrared spec-

troscopy and mass spectrometry. In each case the spectra ob-

tained were identical with those of the trans-ll-octadecenoate

product o f the linoleic acid incubation. Moreover, reductive

cleavage of the methyl ester yielded heptaldehyde and methyl-

1 l-osoundecanoate, which indicated the position of unsatu-

ration to be at C-11.

In contrast to punicic acid, cis-9, trans-11 , trans-13-octadec-

atrienoic acid (a-eleostearic acid) was not changed during incu-

bation with the bacteria. Thus, it would appear that the

TABL E I

Recovery of aH frqm products of &saturation of doubly labeled

stearic acid by Chlorella vulgaris

Experiments with

cis-9, trams-11[9, 10-3Hloctadecadienoic acid

were as described in the text. The biohydrogenation product

was reduced to stearate and incubated with Chlorella. Oleic and

linoleic acids were isolated and counted.

Experim ent and acid =H

“C SH: “C

apm x 10-z dJ%Pz x NJ-’

1. Substrate 18:O~. .

501.8 149.3 3.35

Product 18: 1. . 35.3 11.2

3.15

Product 18:2.. _. . 6.6 2.0

3.30

2. Substrate 18:O.. .

467.5 40.6 11.50

Product 18: 1. 243.1 23.3

10.40

Product 18:2. . . 28.0 2.5

11.20

a The number to the left of the colon represents the number of

carbon atoms in the chain; the number to the right of the colon

designates the number of double bonds.

presence o f the trans bond at C-13 prevented the hydrogenation

of the cis-9 bond.

Hydrogenation of Parafin and Alcohol Derivatives of Punicie

Acid--Gas-liquid chromatography of the hydrocarbons isolated

after incubation of B. Jibrisolvens with cis-9, truns-11

,cis-13-

octadecatriene showed the appearance of a peak not observed in

the hydrocarbon fraction from a zero time control. This peak,

amounting to 21.5% of the hydrocarbon fraction, exhibited a

retention time corresponding to that calculated for an octa-

decene.

The alcohol derivative of punicic acid also appears to be re-

duced, since analysis of the reaction products by argentation

thin layer chromatography showed a spot that corresponded to

truns-11-octadecenol.

Stereospecijcity of Biohydrogenation Reaction-Stereospecific

desaturation of stearic acid by C. vulgaris (15) provided the

rationale by which the stereospecific ity of the reduction of the

cis-9 double bond of cis-9, truns-ll-octadecadienoic acid was

studied. In these experiments cis - 9, truns - 11[9,10 - 3H]octa-

decadienoic acid was incubated with B. brisolvens. The la-

beled truns-11-octadecenoate product was isolated as the methyl

ester and converted to stearic acid. Following the addition of

lJ4C-stearic acid, the doubly labeled stearic acid was incubated

with Chlorella. In each of two experiments, the ratio of 3H:14C

in the oleic and linoleic acids isolated from the algae was the

same as the 3H:14C of the stearic acid substrate (Table I). To

ensure that migration of the labeled hydrogens had not occurred

during incubation with B. Jibrisolvens, stearic acid containing

only the tritium label was incubated with Chlorella. The dini-

trophenylhydrazone derivatives of the aldehyde fragments ob-

tained from reductive ozonolysis of the tritiated methyl oleate

were found to contain almost equal amounts of tritium (Table

II). Another portion of the labeled oleate was oxidat ively

cleaved. No radioactivity was observed in either the nonanoic

or monomethyl azelaic acid fragments . These results show that

the tritium in the cis-9, truns-1119, 10-3H]octadecadienoate had

remained at positions 9,lO during incubation with B. fibrisol-

vens.

Source of Hydrogen in Hydrogenation Reaction---Initial at-

tempts to ascertain the source of the reducing hydrogen were

made by incubating B. jibrisolvens with a series of tritiated

y g ,

,

j g



4/6

5028 Source of Hydrogen and Stereospecificity of Reduction Vol. 246, No. 16

TABLE II

Trilium. in reductive ozonolysis fragments of methyl oleate isolated

S:orn Zhlorella after incubation with cis-9, trans-11[9,1 O-aH]-

octaclecad ienoic acid with Butyrivibrio jibrisolvens

The oeonide of methyl oleate was reduced with 2,4-dinitro-

phenylhydrazine and the dinitrophenylhydrazon es of nonan al

and methyl 9-oxononanoate were oxidized and the water from

each was collected and counted. The specif ic activity of the 9, lo-

di-3H-cis-9, trans-11-octadecadienoic acid was 60 mCi per mmole.

Fragment Tritium

&5m/Jmw1e x 10-a

Nonanal................................. 100.0

Methyl 9-oxononanoate. . . 80.0

TABLE II I

Incorporation of 3H from 1 -3H-glucose and VH-glucose into trans-

11 -octadecenoic acid and the

saturated

fatty acids by

Butyrivibrio jibrisolvens

Incubations were carried out with linoleic acid and the labeled

glucose as described in the text. The methyl esters of the satu-

rated fa tt y acids and trans-11-octadecenoic acid were isolated by

sili cic acid-silver nitrate column chromatography. The frac-

tions emerging from the column first were taken as saturated fatty

esters. Ester concentrations were determined on a portion of

the

sample, and another portion was counted in a liquid scintil la-

tion spectrometer. Radioactivity measurements were corrected

for background and quench. The specif ic acti vity of each trit-

iated glucose was 100 $Zi per mmole.

Substrate

PH-Glucose

3-3H-Glucose.

Saturated acids

cpm//mde

7279

366

trans-11.18: 1

C@&/j.Hde

40

58

TABLE IV

Deuterium in methyl trans-11 -octadecenoate isolated after incubation

of linoleic.acid and punicic acid

Ex-

Per cent of parent ions containing

Substrate peri-

ment

XoD 1D 2D 3D 4D

atoms atom atoms atoms atoms

~-__-

cis, &s-18:2 (Agn12) 1

15 33 43 9 0

2

15 16 60 7 0

3 9 40 39 3 0

cis, trans,cis-18:3 (Ag.11v13) 1 7 19 34 27 13

2 8

24 36 25 7

substrates. No tritium was incorporated from glucose labeled

in positions 1 2 3 5 and 6 or from tritiated succinate or formate.

> 7 9 9

,Evidence that l-3H-glucose and 3-3H-glucose were metabolized

as expected, i.e. provided reducing equivalents for fa tt y acid

synthesis, comes from the observation that tritium was found in

the fa tt y acids synthesized by the cell (Table III) .

To determine whether or not water provides the hydrogens

for reduction, incubations of

I?. jlbriso lvens in DzO were per-

formed. When incubated in D20, a single deuterium atom was

found to be incorporated at C-13 during the isomerization of

linoleic acid to cis-9, truns-11-octadecadienoic acid (2), but the

hydrogenated product was not examined. More recent experi-

TABLE V

Deuterium cgntent of fragments of methyl end and carboxyl end of

11 ,I$-dimethoxu octade canoa te prepared from methyl trans-il-

octadecenoate isolated after linoleic acid incubation with

Butyrivibrio $brisolvens

Mass spectra (70 e.v.) were obtained with an AEI-12 spectrome-

ter and the peaks of interest were corrected for natural abund ance.

Per cent of parent ions containing

No D atoms / 1 D atom 1 2Datoms

TABLE VI

Distribution of deuterium in trans-li-octadecene prepared from

methyl trans-11 -octadecenoate product of linoleic and punicic

acid incubations

After isolation, methyl trans-11-octadecenoate was converted

to the paraffin derivative and oxidatively cleaved to hepta noic

and undecanoic acids. The methyl esters of these acids were

subjected to gas-liquid chromatography-mass spectrometer

analyses on a LKB model 9000 spectrometer (70 e.v.). The peaks

of interest were corrected for natural abundance. The positions

refer to the original trans-11-octadecenoic acid, positions 9 and 10

coming from undecanoic acid and positions 13 and 14 coming from

heptanoic acid.

Substrate

Deuterium at positions

1 9 1 10 1 13 / 14

cis-cis-18:2 (A9.12)

cis-trans-&s-18:3 (As.ll.lr)

ments in which approximately 3 ml of bacterial pellet were

suspended in 12 ml of 99% DzO indicate that, during reduction

of the cis-9, trans-11-octadecadienoic acid in DzO, 2 additional

atoms of deuterium were present in the resulting truns-ll-

octadecenoic acid (Table IV). The actual level of deuterium

incorporated reflects not only the specif ic activ ity of the water

but the rate of equilibration of deuterium with the active hydro-

gens of the bacterial cell.

Similar experiments were performed with punicic acid as a

substrate. Since punicic acid does not undergo isomerization

prior to its hydrogenation to the truns-11-monoene and since

both c is bonds are reduced, it was expected that 4 deuterium

atoms would be incorporated. The results in Table IV show

this to be the case.

These experiments analyzed by mass spectrometry indicated

that deuterium was incorporated during the reduction of the

cis double bond(s) but did not reveal the positions of substitu-

tion. To localize the incorporated deuterium, the methyl ester

of truns-11-monoenoic acid resulting from linoleic incubation

was converted to the 11,12-dimethoxy derivat ive and subjected

to

mass spectrometry (23). When treated in this manner, the

dimethoxy compound undergoes cleavage between the meth-

oxy l groups, yielding 2 ions (m/e 129 and m/e 229) corresponding

to the methyl end and the carboxyl end of the methyl truns-ll-

octadecenoate (23). The results (Table V) indicate that the deu-

terium atoms incorporated during hydrogenation of the A-9,

y g ,

,

j g



5/6

trans-11-octadecadienoic acid were located in the carboxyl por-

acid were hydrogenated and, thus, support this conjecture.

tion of the molecule.

Some naturally occurring compounds that contain a truns-con-

The distinct positions of substitution were obtained by reduc- jugated double bond system, such as the carotenes, escape

ing the deutcrated truns-11-octadecenoic acid from the punicic

hydrogenation in the rumen (27). The findings with cu-eleo-

acid and linoltic acid incubations to the trans-11-octadecene. stearic acid and the punicic acid derivat ives suggest that it is

Oxidative cleavage and mass spectrometry of the methyl esters

the presence of the truns configuration rather than the absence

of the heptanoic acid and undecanoic acid fragments allowed of a carboxyl group which accounts for their lack of hydrogena-

the use of the McLaffer ty rearrangement to determine the tion.

location of deuterium in the original truns-1 l-octadecenoic acid. Experiments with tritiated glucose (labeled in positions 1, 2, 3,

The major peak of methyl esters longer than Cs is due to a 5, and 6) showed that the hydrogen that reduces the double bond

rearranged ion of m/e 74.

This ion contains 3 hydrogen atoms: did not come directly from glucose. Similarly, absence of tritium

2 from the a-carbon and 1 from the y-carbon of the fa tty acid incorporation in truns-11-octadecenoic acid from labeled formate

methyl ester (26). In this case, the two hydrogens bonded to and succinate, as well as the absence of deuterium incorporation

the a-carbon of the methyl heptanoate fragment represent the from a totally deuterated algal hydrolysate, indicated that the

hydrogen atoms at C-13 of the truns-11-octadecenoic acid.

direct addition of hydrogen from an organic substrate was un-

Those bonded at C-10 of the trans-11-octadecenoic acid would likely . In contrast, the fac t that 2 deuterium atoms were incor-

correspond to a-hydrogens of the methyl undecanoate fragment.

porated from D,O during the hydrogenation of cis-9, truns-ll-

The appearance of a large peak at m/e 75 in the spectrum of octadecadienoic acid and 4 deuterium atoms from DzO were

each of the monocarboxylic methyl esters from both substrates incorporated in the biohydrogenation of punicic acid indicates

indicates that hydrogen from HZ0 is incorporated at C-10 of

that water is the immediate source of hydrogens used to reduce

linoleic acid and at C-10 and C-13 of punicic acid. the cis bond(s). These results, however, do not preclude the

From the ratio of the m/e 74 ion to the m/e 75 ion and the direct reduction of a carrier by an organic substrate if the hydro-

assumption that all of the deuterium incorporated was bonded gen carrier can undergo rapid exchange with water.

to the carbons o f the cis double bond(s), the distribution of Examination of the isotope distribution in the reduced products

deuterium at each of the positions of the double bond could be showed that the position(s) adjacent to the truns double bond

calculated. The results (Table VI) show that the carbons adja-

contains less deuterium than the distal position(s). This dis-

cent to the truns double bond contain less deuterium than those tribution indicates that discrimination against deuterium had oc-

distal to the truns bond.

From the mass peaks associated with curred at C-10 of cis-9,truns-ll-octadecadienoic acid, and at (‘-10

the parent ions, it may be calculated that 3056 of the hy-drogen and C-13 of punicic acid. Therefore, the hydrogens added at

atoms at C-13 and C-14 and 28T1 of the hydrogen at C-9 and C-10 or C-13 must have experienced at least one more bond-

C-10 were replaced by deuterium.

These results, together with

breaking event than those added at C-9 or C-14. These results

the similarity in distribution, suggest that both of the cis bonds

lead to the suggestion that the mechanism of biohydrogenation

of punicic acid were hydrogenated by the same system. involves the addition of a proton to the cis bond at the position

distal to the truns bond and that reduction of the double bond is

DISCUSSION

fina lly completed by a hydride ion provided by an unknown car-

The hydrogenation of linolcic acid initially involves the isom- rier.

erization of linoleic acid to a cis-9, truns-1 l-octadecadienoic

Since ferredoxin occurs commonly in anaerobic organisms, one

acid. Several reports on the natire and characteristics of lino-

might expect this electron carrier to be involved in biohydrogena-

leic acid isomerase, the enzyme that catalyzes this reaction,

tion. However, we were unable to observe a ferredoxin band on

have been published (2, 3), but, until now, none of the findings

a DEAE-cellulose column following chromatography (28) of cell

concerning the reduction of the conjugated intermediate to truns-

extracts of B. brisolvens.

11.octadecenoic acid have been reported.

Upon biohydrogenation and reduction of the monoenoic acid to

As there is no readily available source of this cis-9,truns-ll-

stearic acid, i t is possible, with the stearic acid as a substrate for

octadecadienoic acid intermediate, punicic acid, cis-9, truns-ll ,

Chlorellu, to determine the stereochemistry of hydrogen addition

cis-13.octadecatrienoic acid represents a unique substrate which

by B. jibrisolvens. Morris has used this approach to study the

facilita tes the investigation of t,he reductive reaction. It has

stereospecific ity of the biohydrogenation of oleic and elaidic acids

been shown (Table IV) that both c is double bonds are hydro-

by mixed rumen flora (29), and Schroepfer, using Corynebucterium

genated, resulting in the same product as that obtained from

diphtheriue instead of ChZoreZZu, determined the stereospecific ity

linoleic hydrogenation. It is interesting to note that, when

of the hydroxylation of oleic acid (30). As reported in the pre-

ar-eleostearic acid (cis-9, truns-11 , trans.13.octadecatrienoic acid) vious paper of this series, the same approach was used to show

is used as a substrate, no reaction occurs.

The inactive a- the stereospecific ity of hydrogen addition at C-13 of linoleic acid

eleostearic acid is a conjugated triene similar to punicic acid

during its isomerization (31).

differing only in that the configuration of the Al3 bond is truns

If the biohydrogenation of cis-9, truns-11-octadecadienoic acid

instead of cis. It is apparent, therefore, that the configuration

occurs by cis addition, then either DD or LL-9, 10-di-aH-truns-ll-

of the conjugated truns double bond system imparts a degree of

octadecenoic acid would result. Desaturation by Chore&x of the

alteration to the molecule such that the organism is incapable of

stearic acid derived from the truns-ll-octadecenoic acid would

reducing the cis bond of the conjugated triene.

yield oleic and linoleic acids showing either complete recovery o f

The similarity of deuterium distribution at both cis bonds

the tritium for the D-labeled enantiomer or complete loss of trit-

indicates that, unlike linoleic acid isomerase, the carboxyl group

ium for the L-labeled enantiomer. The truns addition of hydro-

is a dispensable feature of the substrate. Preliminary experi-

gen by B. fibrisolvens would yield threo-di-aH-truns-ll-octudccc-

ments showed that the alcohol and paraffin derivat ives of punicic

noic acid, and the oleic acid isolated from Chlorellu would be

Issue of August 25, 1971

I. X. Rosenfeld and S. B. Tove

5029

y g ,

,

j g



6/6

5030 Source of Hydrogen and SkreospeciJicity of Reduction

Vol. 246, No. 16

expected to show one-half of the tritium label. The results (Ta-

ble I) showed complete recovery, and reductive and oxidative

ozonolysis of the oleic acid showed that the tritium label had not

moved during incubation. We conclude, therefore, that the bio-

hydrogenation of L-9, trns-1 I-octadecadienoic acid by B.

fibrisolvens occurs by cis addition to the D side of carbons 9 and 10.

Morris (29) has shown that the biohydrogenation of oleic acid

involves cis addition to the

L

side.

However, B. fibrisolvens is

unable to hydrogenate oleic acid (32). Consequently, although

the biohydrogenation of oleic and &s-9, trans-11-octadecadienoic

acids is similar in that both involve cis addition to a cis double

bond, it is clear that, the two systems are different.

Studies with a cell-free system capable of carrying out biohy-

drogenation are in progress.

Particular eff ort is being directed

toward the elucidation of the nature of the electron donor and

carrier.

Acknowledgments-We wish to thank Dr. Marion Miles of the

Department of Chemist ry for some of the mass spectrometric

analyses. We also thank Drs. D. P. Schwartz and 0. W. Parks

of the USDA, Washington, D. C., for helping us separate the

2,4-dinitrophenylhydrazone derivatives and further apprec iation

is extended to Dr. Parks for his help in gas-liquid mass spec-

trometry. We are also indebted to Dr. L. J. Morris for his many

helpful comments and discussions.

7. BAUMANN , W. J., AND MANGOLD, H. K., J. Org. Chem., 29,

3055 (1964).

8. BAUMANN, W. J., JONES. L. L.. BARNUM. B. E.. AND MANGOLD.

9.

10.

11.

12.

13.

DEVRIES, B., J. Amer. Oil Chem. Sot., 40, 184 (1963).

CARROLL, K. K., J. LimZ Res.. 2. 135 (1961).

MORRIS, L. J., in A. ‘I . JAM& AAD L.‘J.

I~~ORRIS

(Editors),

New biochem ical separations,

D.

Van Nostrand, New York,

1964, p. 300.

14.

ROEHM, J. N., AND PRIVE TT, 0. S., J. Lipid Res., 10,245 (1969).

15.

MORRIS, L. J., HARRIS, R. V., KELLY, W., AND JAMES, A. T.,

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

H. K., &em. khys. Lipids, i, 63 (1966). ’

DOLE, V. P., J. Clin. Invest., 36, 150 (1956).

SCHLENIC, H.. AND GELLERMAN. J. L.. Anal. Chem.. 32. 1412

(1960):

Biochem. J., 109, 673 (1968).

TOVE, S. B., J. Nutr., 76, 361 (1961).

EIIWARDS, H. M., JR., Lipids, 1, 1 (1966).

JOHNSON, G. D., J. Amer. Chem. Sot., 73, 5888 (1951).

SCHWARTZ,

D. P.,

WEIHRAUCH, J. L., AND BURGWALD, L. H.,

Anal. Chum., 4i, 984 (1969).’

SCHWARTZ, D. P.. SHAMEY. J.. BREWINGTON. C. R.. AND PARI~S.

0. W., ikicrochem. J., i3, 407 (1968). ’ ’

PETERSON, J. I., Anal. Biochem., 31, 189 (1969).

CASTLE, J. D., AND ACKMAN, R. G., Can. J. Chem., 45, 1405

(1967).

NEIHAUS, W. J., JR., AND RYHAGE, R., Anal. Chem., 40, 1840

(1968).

SCHEUERBRAND T, G., AND BLOCK. K.. J. Biol. C hem., 237,

2064 (1962).

, ,

SNYDER, F., AND STEPHE NS, N., Biochim. Biophys. Acta, 34,

244 (1959).

1.

2.

3.

4.

5.

6.

26.

REFERENCES

RYHAGE, R., AND STENHAGEN, E., in F. W. MCLAFFERTY

(Editor), Mass spectroscopy of organic ions, Academ ic Press,

New York, 1963, p. 400.

KEPLER, C. R., HIRONS, K. P., MCNEILL, J. J., AND Tove, 27.

S. B., J. Bid. Chem., 241, 1350 (1966).

SHORLAND, F. B., WEENINK, It. O., JOHNS, A . T., AND Mc-

KEPLER, C. R., AND Tovs, S. B., J. Biol. Chum., 242, 5686

DONALD, I. R. C., Biochem. J., 67, 328 (1957).

28.

(1967)

MOIZTENSON, L. E., VALENTINE , R. C., AND CARNHAN, J. E.,

Biochem. Biophys. Res. Comm un., 7, 448 (1962).

KEPLER, C. R., TUCI~ER, W. P., AND TOVE, S. B., J. Biol.

29.

Chem., 246, 3612 (1970).

MORRIS, L. J., Biochem. J., 118, 681 (1970).

HARRIS, R. V., AND JAMES, A. T., Biochim . Biophys. Acta, 106,

30. SCHROEPFER, G. J., JR., J. Biol. Chem., 241, 5441 (1966).

456 (1965).

31. KEPLER, C. R., TU CKER, W. P., AND TOVE, S. B., J. Biol.

CROMBIE, L., AND JACICLIN, 9. G., J. Chem. Sot., 1632 (1957).

Chem., 246, 2765 (1971).

FARQUHAR, J. W., J. Lipid Res., 3, 21 (1962).

32. POLAN, C. E., MCNEILL, J. J., AND TOVE, S. B., J. Bacterial.,

88, 1056 (1964).

y g ,

,

j g


Biohydrogenation of Unsaturated Fatty Acids

Documents