PROPIONIC ACID METABOLISM: MECHANISM OF THE

METHYLMALONYL ISOMERASE REACTION AND THE REDUCTION

OF ACRYLYL COENZYME A TO PROPIONYL COENZYME A IN
PROPIONIBA CTERIA
By ROBERT W. SWICK
DIVISION OF BIOLOGICAL AND MEDICAL RESEARCH, ARGONNE NATIONAL LABORATORY, ARGONNE,
ILLINOIS

Communicated by W. H. Taliaferro, October 2, 1961

The pathways of propionate formation in the propionibacteria and of propionate
oxidation in animal tissues have been well documented. In the bacteria the final
step is the transfer of a carboxyl group from methylmalonyl CoAl to pyruvate with
the formation of oxalacetate and propionyl CoA." I The reaction was found to
involve biotin;2 the purified enzyme, methylmalonyl-oxalacetic transcarboxylase,
has now been shown to contain this factor.4 Methylmalonyl CoA arises from
succinyl CoA by an isomerization the mechanism of which remains obscure. The
original proposal that it occurred via a transcarboxylation5 was questioned on
theoretical grounds as well as by indirect experimental evidence.2 The discovery
that. the reaction was dependent on the presence of the various cobamide coen-
zymes6 suggested that the rearrangement was analogous to the isomerization of
glutamic acid to f-methylaspartic acid observed in Clostridium tetanomorphum.)
In the latter case the rearrangement can be considered to be a cleavage of a pro-
pionate moiety with subsequent reattachment of the glycine residue to the a
position of propionate. Using an extract of propionibacteria, Eggerer et al.8 showed
that the isomerization of methylmalonyl CoA involved a similar migration of the
formyl thioester group from one methylene position of the propionate moiety to
the other. Similar results were obtained in our laboratory with methylmalonyl
isomerase isolated from sheep kidneys, as well as from propionibacteria.
Eggerer et al.8 suggested that the rearrangement of methylmalonyl CoA involved
a free radical intermediate. In accord with this hypothesis, Phares et al.9 recently
reported an inhibition of the reaction by chlorpromazine, an efficient electron donor.
It has also been suggested10 that the reaction might involve a dehydrogenation
which would weaken the bond of the formyl CoA group, permitting its translocation
to the other side of the double bond. Since there has been no successful demon-
stration of the existence of formyl CoA, it seemed likely that the thioester would be
bound to the enzyme and that free acrylic acid might then occur as a transient
intermediate. This hypothesis has been tested by carrying out the isomerization of
methylmalonyl CoA in the presence of isotopic acrylic acid and examining the
succinate formed for radioactivity. When crude extracts of Propionibacterium
shermanii were used as a source of the transcarboxylation system for the synthesis of
methylmalonyl CoA, acrylic acid-C'4 was incorporated into succinate. However,
subsequent study showed that the route of its incorporation was not via the methyl-
malonyl isomerase reaction. This was revealed by the failure to obtain incorpora-
tion of acrylate when synthetic methylmalonyl CoA was used or when the extracts
were treated with ammonium sulfate before incubation.
Further investigation revealed that acrylyl CoA was formed from acrylic acid,
288
VOL. 48, 1962 BIOCHEMISTRY: R. W. SWICK 289

reduced to propionyl CoA, and then converted to succinyl CoA. The reduction of
acrylyl CoA by DPNH has not previously been demonstrated in the propioni-
bacteria. It apparently was the only reaction involving the carbon chain of acrylyl
CoA catalyzed by our preparations, however.
Methods.-P. shermanii was cultured and harvested as described previously.2 Cell-free ex-
tracts were prepared by suspending 5 gm of lyophilized or 20 gm of fresh cells in 30 ml of 0.12 M
KCl, pH 7, containing 0.001 M cysteine, and subjecting the suspension to a Raytheon disintegrator
(10 kc/second) for 10-15 min while the temperature was maintained near 0C. The suspension
was centrifuged at 20,000 X g for 15 minutes and the unbroken cells resuspended and treated
again with the disintegrator. After centrifuging the cells, the cloudy supernatant solutions from
both sonic treatments were combined and centrifuged at 25,000 X g for 20 min. The clear
extracts were either used immediately, or more often dialyzed overnight against 100 volumes of
0.1 M phosphate buffer, pH 7.2. The protein fraction precipitated by 90 per cent saturation
with ammonium sulfate was obtained in the usual manner.
Methylmalonyl isomerase was prepared from sheep kidneys as described by Beck et al."
Methylmalonic acid labeled with C14 in the methyl group was synthesized by the condensation
of C14HJI with sodium-diethylmalonic ester. The ester was saponified, and the free acid purified
by partition chromatography on Celite."2 Acrylic acid-2-C'4 was prepared from f-alanine-2-Cl4 by
deamination with HN02 to form ,-hydroxypropionic acid, which was then dehydrated by dis-
tillation from 50% H2SO4.13 The distillate was neutralized, concentrated to a small volume, and
taken to dryness by lyophilization. Acrylic acid was purified by adsorption on Celite equilibrated
with 1.5 M KH2P04 and elution with CHC13 saturated with the phosphate solution.14 The eluate
was neutralized and the aqueous phase was lyophilized, dissolved in buffer, and stored at - 100.
This chromatographic procedure was also used to separate acrylic acid from propionic acid in
subsequent experiments.
The coenzyme A derivative of methylmalonic acid was prepared by the mixed anhydride pro-
cedure11 and acrylyl CoA was prepared from acrylic acid anhydride by a modification15 of the
method of Simon and Shemin.16
Methylmalonyl-C'4 CoA was incubated with the purified isomerase for 15 min at 300 at pH 7
and the reaction terminated by the addition of NH40H to a final concentration of 2 N." This
procedure converted succinyl CoA to succinamic acid, thus preserving the asymmetry of the
isotope distribution in the methylene carbons of succinyl CoA. Succinamic acid was isolated by
extraction into acetone at pH 4.5,11 evaporation of the acetone, and solution in 0.2 N H2S04.
(Excessive acid will hydrolyze succinamic acid to succinic acid.) Carrier succinamic acid was
added and the succinamic acid-Cl4 was purified by partition chromatography. After treatment
successively with CHC13, CHC13-5% n-butanol, CHC13-7.5% n-butanol, and CHCL,-10% n-
butanol, methylmalamic acid was eluted with CHC13-15% n-butanol and succinamic acid with
CHC13-25% n-butanol. The sodium salt was evaporated to dryness and degraded to s-alanine
with sodium hypobromite. B-Alanine was converted to acrylic acid as described above and the
acrylic acid reduced to propionic acid. Propionic acid was next degraded by successive Schmidt
decarboxylations,17 each carbon being obtained as CO2. The specific radioactivity of each fraction
was determined by proportional gas counting.
A sample of succinamic acid was also obtained which had been prepared from propionate-3-C14
using purified propionyl carboxylase and methylmalonyl isomerase, the latter enzyme having been
prepared from P. shermanii."8 This preparation, the generous gift of Drs. R. Stjernholm and
H. G. Wood, Western Reserve University, was also degraded and the products counted.
Results.-The isomerization of methylmalonyl CoA: The results of the degrada-
tion of succinyl CoA arising by the isomerization of methyl-C14-malonyl CoA are
shown in Table 1. The isomerization of methylmalonyl CoA appears to occur by a
migration of a formyl CoA group from the a position of the propionic acid moiety to
the f position, thus confirming the results of Eggerer et al.8 The mechanism is
similar to that involved in the isomerization of glutamate. The occurrence of some
activity in the other methylene carbon of succinamic acid may be due either to
290 BIOCHEMISTRY: R. W. SWICK PROC. N. A. S.

TABLE 1
ASYMMETRIC DISTRIBUTION OF RADIOACTIVITY IN SUCCINYL CoA
COOH 4
COOH CH2 3
H3C14-CH &2 2
co bo 1
SCoA SCoA
Distribution of radioactivity in
methylene carbons of succinyl CoA
Methylmalonyl isomerase formed from methyl-C14-malonyl CoA
prepared from C2, cpm C3, cpm
Sheep kidneys 55 6
P. shermanii 889 103

excessive oxidation during the Schmidt degradation or to the presence of small

amounts of CoA transferase in the enzyme preparations. Such contamination
would remove CoA from succinate, resulting in the randomization of isotope upon
reacylation. In our hands, crude extracts obtained from propionibacteria produce
nearly complete randomization.
Acrylic acid as an intermediate in the isomerization: The possible participation
of free acrylic acid in the isomerization was tested by the following experiments.
The simplest procedure was to incubate acrylate-C14 with the extracts and the
intermediates of the transcarboxylation-isomerization system: propionate, co-
enzyme A, ATP, and oxalacetate. Under these conditions, succinate-C14 was
readily formed (Table 2). However, treatment of the extracts with ammonium
TABLE 2
SUCCINATE FORMATION FROM PROPIONATE-C04 AND ACRYLATE-C14
_.SuccinateC14, pmoles,t formed from:--
Acrylate
Additions* Propionate
Crude extract + oxalacetate + CoA + ATP 0.580 0.330
Treated extract + oxalacetate + CoA + ATP 0.560 0.009
Treated extract + methylmalonyl CoA (0. 277) T 0. 001
* When indicated, the system contained (in Mmoles): propionate-3-C'4 or acrylate-2-C14, 1; oxal-
acetate, 10; CoA, 1; ATP, 3; methylmalonyl CoA (active isomer), 0.5; and in addition potassium
phosphate, pH 7, 100; enzyme with 8 mg of protein. Final volume 1.0 ml; incubated 15 min at 300.
t The ,umoles are calculated from the radioactivity recovered in succinate and the specific activity
of the propionate or acrylate added.
$ Performed previously (ref. 2).

sulfate to isolate the protein fraction resulted in the total loss of activity for the
incorporation of acrylate; in control incubations propionate-C14 was still converted
to succinate (Table 2). Furthermore, incubation of these otherwise active ex-
tracts with methylmalonyl CoA and acrylate-C14 resulted in no incorporation of
radioactivity into succinate. Hence, one may conclude that the incorporation of
acrylate observed with crude extracts was not occurring via the suggested mecha-
nism and that free acrylate is not an intermediate in the methylmalonyl isomerase
reaction. Previous experiments had shown indirectly that propionate is not an
intermediate.2 Barker19 recently reported the failure of his group to observe
incorporation of acrylate, glycine, or propionate into f3-methylaspartate during
the isomerization of glutamate.
The conversion of acrylate to succinate: It was of interest to elucidate the pathway
of acrylate incorporation into succinate. With crude extracts the incorporation
VOL. 48, 1962 BIOCHEMISTRY: R. W. SWICK 291

was completely inhibited by avidin and was partially dependent on the addition of
oxalacetate and coenzyme A. The loss in activity observed when the extracts
were treated with ammonium sulfate suggested that some other essential factor was
being removed. There is ample evidence that acrylyl CoA can be reduced to
propionyl CoA in a number of systems,20-22 but previous attempts to demonstrate
the reaction in the propionibacteria had been negative. When either TPNH or
DPNH was added to the treated extract in the presence of CoA, ATP, and oxal-
acetate, the reduction of acrylate to propionate and subsequent conversion to
succinate occurred (Table 3).
TABLE 3
REQUIREMENTS FOR THE CONVERSION OFACRYLATE-C14 TO SUCCINATE AND PROPIONATE
Succinate formed, Propionate formed,
Additions jsolesj jsmolest
Complete system* 0.44 ...
-DPNH 0.01 ...
-DPNH +TPNH 0.12
-oxalacetate 0.02 0.30
- oxalacetate, - DPNH ... 0.00
- CoA, - oxalacetate ... 0.00
* The system contained (in jsmoles): acrylate-2-C'4, 1; DPNH (or TPNH, where indicated), 5;
oxalacetate, 10, CoA, 1; ATP, 3; potassium phosphate, pH 7, 100; and enzyme with 8 mg of pro-
tein. Final volume, 1 ml; incubated 15 min at 30°.
t Calculated as described in footnote to Table 2.
The reduction of acrylate to propionate was shown directly by the isolation of
labeled propionate when acrylate-C14 was incubated in the same system minus
oxalacetate (Table 3).
The reaction could be followed more easily by observing the decrease in optical
density at 340 mp resulting from the oxidation of reduced pyridine nucleotide in
the presence of acrylyl CoA (Table 4). Here only DPNH was active in the reduc-
TABLE 4
SPECIFICITY OF PROPIONYL CoA DEHYDROGENASE
Additions* A 340 mjut
Acrylyl CoA + DPNH 0.109
Acrylyl CoA + TPNH 0.000
Acrylyl CoA + TPNH + DPN 0.000
Acrylyl CoA + DPNH 0.199
Crotonyl CoA + DPNH 0.051
Acrylyl CoA + DPNH 0.126
Acrylyl pantetheine + DPNH 0.029
* The system contained (in pmoles): acrylyl CoA, crotonyl CoA, acrylyl
pantetheine, 1; DPNH, TPNH, DPN, 0.4; Tris buffer, pH 7.2, 50; enzyme
containing 1 mg of protein. Final volume, 3 ml, incubated 10 min at 300.
* Change in optical density at 340 mjA per 10 min as measured in the Beckman
spectrophotometer. Corrected for the oxidation of DPNH in the absence of
thioester.

tion; TPNH was inactive, in contrast to the results of the isotope experiments.
The lack of activity when DPN was added with TPNH indicates that there is no
transhydrogenase activity; therefore the isotopic evidence for the participation of
TPNH in the reaction appears to be an artifact. The enzyme involved most
likely is propionyl dehydrogenase since it exhibits much less activity toward
crotonyl CoA. The enzyme also displays only slight but significant activity
toward acrylyl pantetheine (Table 4). We are currently attempting to purify and
to characterize this enzyme from the propionibacteria.
292 BIOCHEMISTRY: R. W. SWICK PROC. N. A. S.

The enzyme for the conversion of acrylic acid to acrylyl CoA is present in our
extracts, but the activity is not as high as that seen with propionate (Table 5).
Acrylic acid actually appears to inhibit the esterification of propionic acid, perhaps
TABLE 5
ESTERIFICATION OF PROPIONATE AND ACRYLATE BY CoA AND ATP AS MEASURED BY
HYDROXAMATE FORMATION
Hydroxamate formed,
Acid /Amoles
Propionate 4.61
Acrylate 0.93
Propionate + acrylate 3.12
The system contained (in Mmoles): Tris buffer, pH 7.4, 250; CoA, 0.5; ATP, 5;
MgCl2, 8; glutathione, 20; hydroxylamine, 1000; propionate, acrylate, 100; and
enzyme with 4 mg of protein. Final volume, 2.0 ml; incubated 30 min at 30°.
Hydroxamic acid was determined by the method of Lipmann and Tuttle (ref. 23).

by the removal of CoA by addition across the double bond.20 The activity for
acylation was too low to be determined by coupling the reaction with the reduction
of acrylyl CoA as measured spectrophotometrically.
The metabolism of acrylic acid by whole cells: To determine whether acrylate
could be metabolized by whole cells, 100 ,moles of acrylate-2-C'4 (5,200 cpm) was
incubated with 1 gm of whole cells for 16 hours at 30°. Most of the activity re-
covered was in propionate (1,590 cpm), succinate (950 cpm), and acetate (360 cpm).
Since lactic acid migrates with succinate on a Celite column, the succinate fraction
was treated with KMnO.4 to destroy any lactic acid present. Since there was no
loss in activity, apparently no labeled lactic acid accumulated; possibly none was
formed from acrylate. No attempt was made to recover unreacted acrylate.
Discussion.-The DPNH-dependent reduction of acrylyl CoA to propionyl CoA
by the propionibacteria has not been described previously. No other reactions in-
volving the carbon chain of acrylyl CoA were observed to be catalyzed by our ex-
tracts. So far we have been unable to detect the amination of acrylyl pantetheine
to beta-alanine or its hydration to lactyl pantetheine or to beta-hydroxy propionyl
pantetheine, reactions known to occur in other organisms.24-26 Further study will
be required to determine the function of the reduction of acrylyl CoA in the metab-
olism of the propionibacteria.
Summary.-Methylmalonyl CoA is converted to succinyl CoA by the migration
of the thioester moiety from one methylene carbon of the propionate group to the
other. The isomerization was shown not to involve free acrylic acid as an inter-
mediate. Acrylic acid was shown, however, to be subject to esterification to acrylyl
CoA and reduction to propionyl CoA by cell-free extracts of P. shermanii.
The author is deeply grateful to Mrs. Isabelle S. Waldron and Mrs. Geraldine G. Moss for
their invaluable technical assistance in the performance of these experiments.
*
Performed under the auspices of the U. S. Atomic Energy Commission. A preliminary report
was presented at the meeting of the American Society of Biological Chemists, Atlantic City,
April 10-14, 1961.
1 Abbreviations used are CoA (coenzyme A), ATP (adenosine triphosphate), DPNH (diphos-
phopyridine nucleotide, reduced), TPNH (triphosphopyridine nucleotide, reduced), Tris (tris
(hydroxymethyl) methylamine).
2 Swick, R. W., and H. G. Wood, these PROCEEDINGS, 46, 28 (1960).
VOL. 48, 1962 BIOCHEMISTRY: WIBERG ET AL. 293

3Wood, H. G., and R. Stjernholm, these PROCEEDINGS, 47, 289 (1961).

4Stjernholm, R., and H. G. Wood, Fed. Proc., 20, 235 (1961).
5 Beck, W. S., and S. Ochoa, J. Biol. Chem., 232, 931 (1958).
6 Smith, R. M., and K. J. Monty, Biochem. Biophys. Research Commun., 1, 105 (1959); Stadt-
man, E. R., P. Overath, H. Eggerer, and F. Lynen, Biochem. Biophys. Research Commun., 2,
1 (1960); Lengyel, P., R. Mazumder, and S. Ochoa, these PROCEEDINGS, 46, 1312 (1960); Gurnani,
S., S. P. Mistry, and B. C. Johnson, Biochim. et Biophys. Acta, 38, 187 (1960); Stern, J. R., and
D. L. Friedman, Biochem. Biophys. Research Commun., 2, 82 (1960).
7 Munch-Petersen, A., and H. A. Barker, J. Biol. Chem., 230, 649 (1958).
8 Eggerer, H., P. Overath, F. Lynen, and E. R. Stadtman, J. Am. Chem. Soc., 82, 2643 (1960).
9 Phares, E. F., M. V. Long, and S. F. Carson, Bact. Proc., 61, 184 (1961).
10 Kosower, E. M., personal communication.
11 Beck, W. S., M. Flavin, and S. Ochoa, J. Biol. Chem., 229, 997 (1957).
12 Swim, H. E., and L. 0. Krampitz, J. Bact., 67, 426 (1954).
13 Lagerkvist, U., Acta Chem. Scand., 7, 114 (1953).
14 Bueding, E., and H. W. Yale, J. Biol. Chem., 193, 411 (1951).
15 Stadtman, E. R., in Methods in Enzymology, eds. S. P. Colowick and N. 0. Kaplan (New
York: Academic Press, 1957), vol. 3, pp. 931-941.
16 Simon, E. J., and D. Shemin, J. Am. Chem. Soc., 75, 2520 (1953).

17 Phares, E. F., Arch. Biochem., 33, 173 (1951).

18 Stjernholm, R., and H. G. Wood, these PROCEEDINGS, 47, 303 (1961).
19 Barker, H. A., Federation Proc., 20, 956 (1961).
20 Stadtman, E. R., and P R. Vagelos, International Symposium on Enzyme Chemistry (Tokyo
and Kyoto: Pan-Pacific Press, 1957), p. 86.
21 Rendina, G., and M. J. Coon, J. Biol. Chem., 225, 523 (1957).
22 Giovanelli, J., and P. K. Stumpf, J. Biol. Chem., 231, 411 (1958).
23 Lipmann, F., and L. C. Tuttle, J. Biol. Chem., 159, 21 (1945).
24 Stadtman, E. R., J. Am. Chem. Soc., 77, 5765 (1955).
25 Vagelos, P. R., J. M. Earl, and E. R. Stadtman, J. Biol. Chem., 234, 765 (1959).
26 Vagelos, P. R., and J. M. Earl, J. Biol. Chem., 234, 2272 (1959).

EARLY ENZYME SYNTHESIS AND ITS CONTROL IN E. COLI

INFECTED WITH SOME AMBER MUTANTS OF BACTERIOPHAGE T4*
BY JOHN S. WIBERG, MARIE-LUISE DIRKSEN, RICHARD H. EPSTEINJt S. E. LURIA,
AND JOHN M. BUCHANAN
DIVISION OF BIOCHEMISTRY, DEPARTMENT OF BIOLOGY,
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Communicated January 2, 1962
The "amber" (am) mutants of bacteriophage T4 are a recently discovered class
of mutants that can replicate in certain derivatives of Escherichia coli strain K-12
but not in E. coli B. ' The bacterial strains that support growth of the am mutants,
thus making their isolation and propagation possible, are called the "permissive"
hosts. The biochemical basis of this permissiveness has not yet been clarified.
With some of the am mutants, infection of E. coli B leads to a subnormal production
of phage DNA; with others, DNA is formed but infection is abortive because of
failure of late stages of phage development.2
It is known that infection of E. coli with a T-even bacteriophage leads to the ap-
pearance of several new enzyme activities and to an increase in several other enzyme
activities, all related to synthesis of phage DNA ("early enzymes").312 These