Methenyltetrahydrofolate cyclohydrolase catalyzes the synthesis of (6S)-5-formyltetrahydrofolate

Article abstract of DOI:10.1006/bioo.1996.0020

The methenyltetrahydrofolate cyclohydrolase activity of the bifunctional methylenetetrahydrofolate dehydrogenase/cyclohydrolase domain catalyzes the conversion of 5,10-methenyltetrahydrofolate to 10-formyltetrahydrofolate. We have observed that in the presence of this domain, 10-formyltetrahydrofolate is converted to 5-formyltetrahydrofolate. The effects of nucleotide analogs on the initial rates of this reaction were found to be similar to their effects on the cyclohydrolase activity, establishing that 5-formyltetrahydrofolate production is dependent on the cyclohydrolase. The specific activity of 5-formyltetrahydrofolate production is approximately 7 x 10⁴-fold lower than that of the cyclohydrolase activity but can be used to obtain quantitative conversion of 10-formylH₄folate to 5-formyltetrahydrofolate, in mg amounts. This 'side-reaction' may contribute to the in vivo production of (6S)-5-formyltetrahydrofolate.

Full text of DOI:10.1006/bioo.1996.0020

BIOORGANIC CHEMISTRY 24, 220–228 (1996)
ARTICLE NO. 0020
Methenyltetrahydrofolate Cyclohydrolase Catalyzes the
Synthesis of (6S)-5-Formyltetrahydrofolate
J
OELLE N. PELLETIER AND
R
OBERT E. MAC
K
ENZIE
*
*Department of Biochemistry, McGill University and Montreal Joint Center for Structural Biology,
Montre´al, Que´bec, Canada H3G 1Y6
Received January 25, 1996
The methenyltetrahydrofolate cyclohydrolase activity of the bifunctional methylenetetra-
hydrofolate dehydrogenase/cyclohydrolase domain catalyzes the conversion of 5,10-methenyl-
tetrahydrofolate to 10-formyltetrahydrofolate. We have observed that in the presence of this
domain, 10-formyltetrahydrofolate is converted to 5-formyltetrahydrofolate. The effects of
nucleotide analogs on the initial rates of this reaction were found to be similar to their
effects on the cyclohydrolase activity, establishing that 5-formyltetrahydrofolate production
is dependent on the cyclohydrolase. The specific activity of 5-formyltetrahydrofolate
production is approximately 7 ϫ 104-fold lower than that of the cyclohydrolase activity but can
be used to obtain quantitative conversion of 10-formylH4folate to 5-formyltetrahydrofolate,
in mg amounts. This ‘‘side-reaction’’ may contribute to the in vivo production of (6S)-5-
formyltetrahydrofolate.
1996 Academic Press, Inc.
INTRODUCTION
The function of tetrahydrofolate (H4folate)1 is to transfer one-carbon units at
the oxidation states of methanol, formaldehyde, and formate to acceptor molecules
for use in biosynthetic reactions. The one-carbon unit is covalently bound at N5 of
the pteridine ring, N10 of the p-aminobenzoyl moiety, or to both forming a N5–N10
bridge. Figure 1 illustrates the enzymatic addition of formate to H4folate [1] in the
ATP-dependent 10-formylH4folate synthetase reaction, forming 10-formylH4folate
[2] (reviewed in 1). This product can be used in the biosynthesis of purines or can be
reversibly converted to 5,10-methenylH4folate [3] by the action of methenylH4folate
cyclohydrolase without the need of a cofactor or second substrate. 5,10-MethenylH4-
folate can be converted to 5,10-methyleneH4folate by NAD(P)-dependent methyl-
eneH4folate dehydrogenase, for use in biosynthesis of thymidylate, serine, or methi-
onine. 5,10-MethenylH4folate [3] is also converted chemically to 5-formylH4folate
[4], with Keq 4/3 ϭ 1.5 ϫ 104 (2). This chemical conversion has been proposed to
contribute to the in vivo occurrence of (6S)-5-formylH4folate (3). However, the
1 Abbreviations used: H4folate, tetrahydrofolate; D, methyleneH4folate dehydrogenase; C, methenyl-
H4folate cyclohydrolase; S, 10-formylH4folate synthetase; SHMT, serine hydroxymethyl transferase;
TCA, trichloroacetic acid; Tris, tris(hydroxymethyl)aminomethane; FPLC, fast protein liquid chromatog-
raphy; EDTA, ethylenediaminetetraacetic acid; AADP, 3-aminopyridine adenine dinucleotide phos-
phate; NADP, nicotinamide adenine dinucleotide 2Ј-phosphate; 2Ј,5Ј-ADP, 2Ј-phosphoadenosine 5Ј-
phosphate.
220
0045-2068/96 $18.00
Copyright 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

ENZYMATIC SYNTHESIS OF FORMYLTETRAHYDROFOLATE
221
F
IG. 1. 1, H4folate; 2, 10-formylH4folate; 3, 5,10-methenylH4folate; 4, 5-formylH4folate; S, 10-formyl-
H4folate synthetase activity; C, 5,10-methenylH4folate cyclohydrolase activity. It should be noted that
one-carbon substitution at N10 changes the (R,S) designation at position 6.
rate appears to be too slow to account for the concentrations detected (4), which
vary from about 5% of the folate pool in rat liver (5) up to 85% of that in Neurospora
crassa conidiospores (6). Serine hydroxymethyltransferase (SHMT) produces 5-
formylH4folate as a side-reaction (4) and this reaction was shown to be the main
source in N. crassa and Escherichia coli (6, 7)
5-FormylH4folate [4] is of considerable interest as its metabolic role has long
been a source of debate. Unlike other forms of H4folate, the 1-carbon unit carried
by 5-formylH4folate does not participate in any known biosynthetic reaction. This
prompted the proposal that it serves as a stable storage form of H4folate (8).
The enzyme methenylH4folate synthetase is the only enzyme known to utilize 5-
formylH4folate as a substrate. It converts 5-formylH4folate [4] to 5,10-methenylH4-
folate [3] using ATP to overcome the thermodynamically unfavorable reaction;
this results in cycling the stored H4folate back into the pool of metabolically useful
H4folate forms. This reaction forms the basis for the use of 5-formylH4folate as an
agent to rescue normal cells from methotrexate toxicity during cancer treatment
(for a review, see 9). Currently, the racemic mixture of 5-formylH4folate, leucovorin,
is clinically administered. Although the unnatural isomer seems to have no impact
on rescue from methotrexate (9), it is a weak inhibitor of cellular uptake of the
natural isomer, (6S)-5-formylH4folate (9, 10), and it accumulates in plasma relative
to the natural isomer (11). Therefore, it may be an advantage to administer only
the natural isomer (12, 13).
We have found that methenylH4folate cyclohydrolase (C) catalyzes the stereospe-
cific synthesis of (6S)-5-formylH4folate. In many organisms, the C and methyl-
eneH4folate dehydrogenase (D) activities belong to a single polypeptide (D/C)
and appear to share a common active site (14). Furthermore, the D/C domain in
eukaryotes is covalently associated with a 10-formylH4folate synthetase (S) domain,

222
PELLETIER AND MACKENZIE
forming a trifunctional D/C/S enzyme. The human D/C/S has been cloned and
overexpressed in E. coli in this laboratory and the recombinant D/C domain has
been separately expressed and retains the kinetic characteristics of the D/C/S
(15–17). In this study, we examine the synthesis of (6S)-5-formylH4folate by the
C activity of the human D/C domain in order to assess its putative role in contribut-
ing to in vivo 5-formylH4folate levels, and report an alternative method to the
stereospecific synthesis of (6S)-5-formylH4folate [4].
MATERIALS AND METHODS
All the reagents used were of analytical grade. DEAE-cellulose DE-23 was
purchased from Whatman. Centriprep-10 concentrators are Amicon products. So-
dium formate was from J.T. Baker Chemical Co., ͱ-mercaptoethanol from Kodak,
and ATP from Sigma. All other reagents were BDH products. Recombinant human
D/C/S enzyme and its bifunctional D/C domain (D/C301) were purified from E.
coli crude lysate in a 1-day purification as described previously (16, 17). Enzyme
was made NADP-free by desalting and was concentrated by centrifugation using
Centriprep-10. Typical purifications of D/C/S enzyme and of D/C domain yield
approximately 10 and 60 mg of enzyme, respectively (10 and 35 U/mg dehydrogenase
activity, respectively) from 10 g of cells.
Synthesis of (6R)-10-formylH4 folate [2]. The enzymatic synthesis is similar to
that reported by Ho and Jones (18). (6R,S)-H4folate [1] is the starting material
and in this protocol, the purified recombinant human D/C/S enzyme was used as
the source of 10-formylH4folate synthetase (S) activity. Formate is the source of
the one-carbon unit and ATP is the cofactor for the reaction, which is illustrated
in Fig. 1. A solution of (6R,S)-H4folate [1] (0.125 mmol in 12.5 ml) containing 5
m
M
Tris buffer, pH 7.5, 0.36
67 m sodium formate, 5 m
pH 8.0, and 1.5 m ATP, in a final volume of 48 ml. The solution was incubated
at 37ЊC and 15 U synthetase activity (in 2 ml of 30% glycerol, 50 m potassium
phosphate, pH 7.3, and 35 m ͱ-mercaptoethanol) was added. The reaction was
M
ͱ-mercaptoethanol, and 250 m
M
NaCl was made to
M
M
MgCl2 , 33 m KCl, 33 m triethanolamine и HCl,
M
M
M
M
M
placed under vacuum at 37ЊC. Conversion to 2 was monitored by periodic acidifica-
tion of aliquots and spectroscopic quantitation of 3 (described below). The reaction
was complete after 40 min. The reaction mixture was diluted with 167 ml of 0.1
M
ͱ-mercaptoethanol at 4ЊC and applied to a 75-ml DEAE-cellulose column. The
product was purified and eluted in NH4OAc as previously outlined (18).
Conversion of (6R)-10-formylH4 folate [2] to (6R)-methenylH4folate [3]. For sta-
ble storage, 2 was converted to 3 by acidification essentially as described in (18).
The purified 2, in approximately 0.175
M
NH4OAc and 0.1
M
ͱ-mercaptoethanol,
was lyophilized. The resulting yellow powder was redissolved in 5.0 ml 0.1
N
HCl
with heating to 55ЊC. The solution, now at pH 1.7, was placed under vacuum and
precipitated at 4ЊC over 48 h. The yellow crystals were washed with ice-cold EtOH
and dried. A 64% yield of 3 was obtained from the biologically active isomer of 1.
Conversion of (6R)-methenylH4 folate [3] to (6R)-10-formylH4 folate [2]. Immedi-
ately before use 3 was dissolved in a solution of 0.1
M
potassium phosphate, pH

ENZYMATIC SYNTHESIS OF FORMYLTETRAHYDROFOLATE
223
7.3, and 0.1
M
ͱ-mercaptoethanol and was stirred gently at 23ЊC under vacuum, for
1 h, for conversion to 2.
Synthesis of (6S)-5-formylH4 folate [4] using cyclohydrolase activity. The source
of cyclohydrolase activity used in the synthesis was purified D/C domain as it is
obtained in higher yields than the D/C/S. 2 (7.5 Ȑmol) was combined with 1500 U
of cyclohydrolase activity (500 U dehydrogenase activity) in 100 m
anol, 25 m potassium phosphate, pH 7.3, and 0.1% ascorbate in a final volume of
20 ml. The concentrations of substrate and enzyme were 375 and 20 Ȑ , respectively,
M
ͱ-mercaptoeth-
M
M
or a 19 : 1 ratio. The extent of the reaction was followed by differential acidification
(described below). When the reaction was complete, the mixture was made enzyme-
free by filtration to dryness in a Centriprep-10 microfiltration device. The filtrate
was diluted 1 : 1 with water and purified according to the parameters defined by
Moran et al. (19) adapted to purification by FPLC at 4ЊC.
Differential acidification. An aliquot of reaction mix was made to 3.5% TCA and
placed in boiling water for 55 s. Under these conditions, 2 and 4 are both rapidly
1
1
converted to 3, which is quantified by spectrophotometry (E350 ϭ 24.9 mMϪ cmϪ ).
A second aliquot of reaction mix was made to 0.5
M
sodium acetate, pH 3.3, at
23ЊC. 2 is rapidly converted to 99.7% 3, according to the following equilibrium:
1
[3]/[2] ϫ [Hϩ] ϭ 9 ϫ 105 MϪ (2). However, at pH 3, the half-time of cyclization
of 4 to 3 is 270 min (20); thus its conversion is negligeable. The difference between
the concentration of 3 determined in TCA and in acetate provides the concentration
of 4.
Characterization of [6S]-5-formylH4 folate synthetic activity. Initial rates were mea-
sured using 200 Ȑ
of nucleotide or in the presence of saturating nucleotide (40 Ȑ
NADP, or 200 Ȑ 2Ј,5Ј-ADP). When maximum conversion of 2 to 5-formylH4folate
M
(R,S)-10-formylH4folate [2] and 5 Ȑ
M
D/C301 in the absence
M
AADP, 200 Ȑ
M
M
[4] was sought, equivalent concentrations of substrate and enzyme were used. The
amount of product formed was determined by differential acidification. It was not
possible to study classical initial rate kinetics of this activity as its turnover is very
low compared to that of the interconversion of 2 and 3 by the cyclohydrolase activity.
RESULTS
Enzymatic Synthesis of (6S)-5-formylH4 folate
We have observed that 10-formylH4folate [2] is converted to 5-formylH4folate
[4] in the presence of the human D/C domain. In order to demonstrate the enzymatic
nature of the synthesis, we measured the reaction rate in the presence of varying
D/C domain concentrations. Figure 2 illustrates that the initial rate of the reaction
is directly proportional to the enzyme concentration, at saturating substrate concen-
trations. Further evidence of the enzymatic nature of the reaction is provided by
demonstrating its stereospecificity, illustrated in Fig. 3. The natural isomer, (6R)-
10-formylH4folate, is entirely converted to product under conditions where only
50% of the racemate is converted. The chemical rate of interconversion is negligible
compared to that of the enzymatic reaction (Fig. 3).
The very slow rate observed for the previously unrecognized activity prompted

224
PELLETIER AND MACKENZIE
F
IG. 2. Linearity of synthetic activity with enzyme concentration. Initial rates were determined with
increasing concentrations of enzyme (ᮀ ϭ 2 Ȑ , ᭿ ϭ 4 Ȑ , ᭛ ϭ 8 Ȑ ).
M
M
M
F
IG. 3. Stereospecificity of 5-formylH4folate synthesis. Comparison of synthesis in the absence (open
symbols) or presence (closed symbols) of enzyme, using racemic substrate (ᮀ, ᭿) or the naturally
occurring isomer (᭛, ᭜).

ENZYMATIC SYNTHESIS OF FORMYLTETRAHYDROFOLATE
225
TABLE 1
Effect of Nucleotides on Initial Rate of Enzymatic Activities
methenylH4folate Ǟ
10-formylH4folate Ǟ
10-formylH4folate
5-formylH4folate
No nucleotide
ϩNADP
ϩAADP
1
1
0.6
0.2
1.3
0.3
0.1
1.5
ϩ2Ј,5Ј-ADP
Note. Rates are given relative to the initial rate with no nucleotide present.
Initial rates were measured with saturating nucleotide.
the hypothesis that 4 is a side-product of the D/C reactions. As its synthesis is not
nucleotide-dependent, 4 is unlikely to be a side-product of the NADP-dependent
dehydrogenase activity. To test whether 4 is a side-product of the cyclohydrolase
activity, we determined the kinetic effects of various nucleotides on the initial rate
of production and compared them to the effects against the cyclohydrolase activity
(Table 1). It is apparent that the synthetic activity is dependent upon the cyclohydro-
lase activity, as the effects of the nucleotides are virtually identical toward the rates
of both reactions. The conversion of 2 to 4 by the D/C domain occurs with a specific
activity that is approximately 7 ϫ 104-fold lower than the cyclohydrolase activity
of the domain (Table 2). However, as 4 is thermodynamically favored, the reaction
is essentially irreversible and the product accumulates.
Our observation of (6S)-5-formylH4folate production by the D/C domain was
timely, as we required mg amounts of high quality product for kinetic and binding
experiments (14). We devised the simple synthetic scheme that is outlined under
Materials and Methods. Because racemic mixtures of 10-formylH4folate can contain
small amounts of (6R)-5-formylH4folate (2), we enzymatically synthesized (6R)-10-
formylH4folate as starting material to ensure that the final product would be a pure
stereoisomer. After 4.5 h at 37ЊC, the reaction mix contained at least 97% (6S)-5-
formylH4folate. Figure 4 illustrates the elution profile of the FPLC purification.
The elution profile is consistent with the identity of the small peak being starting
material [2] and the main peak being 4, according to Moran et al. (19). The small
TABLE 2
Specific Activity of the Reactions Catalyzed by the D/C301 Domain
Specific activity
1
1
Reaction catalyzed
Activity
(Ȑmol minϪ mgϪ
)
[6R]-methyleneH4folate Ǟ [6R]-methenylH4folate
[6R]-methenylH4folate Ǟ [6R]-10-formylH4folate
[6R]-10-formylH4folate Ǟ [6S]-5-formylH4folate
Dehydrogenase
Cyclohydrolase
Cyclohydrolase
37
120 (16)
3
1.8 ϫ 10Ϫ
Note. Specific activities were determined under saturating substrate conditions with limiting D/C301.

226
PELLETIER AND MACKENZIE
F
IG. 4. FPLC profile for purification of [6S]-5-formylH4folate by chromatography. Absorbance was
recorded in arbitrary units. The sample was loaded onto a 1.5 ϫ 4.5-cm column of DEAE-cellulose
packed with H2O and washed with 50 ml 0.01 ammonium acetate at a rate of 0.8 ml/min, until
background absorbance was eliminated. An ammonium acetate gradient was applied. The small peak
of starting material eluted off at 0.3 ammonium acetate, followed by the main peak which eluted at
approximately 0.45 ammonium acetate, as expected (19).
M
M
M
and the main peaks were characterized by absorbance with differential acidification.
The small peak gave equivalent absorbance at 350 nm upon acidification of the
product in sodium acetate, pH 3.3, and in TCA with boiling, consistent with the
properties of 2. For the main peak, there was no significant absorbance at 350 nm
upon acidification of the product in sodium acetate, pH 3.3, relative to acidification
in TCA with boiling, consistent with the characteristics of 5-formylH4folate. The
product was further characterized by spectral analysis. The ␭max(neutral pH) ϭ 287
nm (21) and the shape of the spectrum confirm the identity of the product. The
pooled fractions containing 4 were lyophilized. The reaction produced 6.1 Ȑmol of
product (2.8 mg), representing an 83% yield from 2. As the D/C domain shows
little variation in affinity toward polyglutamylated forms of folate (22, 23), these
could also be used as substrates for (6S)-5-formyltetrahydropteroylpolyglutamate
synthesis.
DISCUSSION
The cyclohydrolase activity has allowed us to develop an alternative synthesis of
the natural isomer of 5-formylH4folate. Previously published syntheses require
chromatographic separation of the chemically prepared racemate (24) or synthesis
of (6R)-methenylH4folate followed by conversion to product by boiling (21) or
autoclaving (25), often resulting in low yield with a high amount of degradation
products. Stover and Schirch have overcome this problem and report near quantita-
tive conversion to product with heat (26). A gentle, enzymatic conversion provides
a favorable alternative to chemical conversion and also circumvents the possibility

ENZYMATIC SYNTHESIS OF FORMYLTETRAHYDROFOLATE
227
of obtaining degradation products. A previously published enzymatic synthesis of
(6S)-5-formylH4folate is catalyzed by a low specific activity side-reaction of the
enzyme formiminoglutamate : H4folate formiminotransferase : formiminoH4folate
cyclodeaminase (27). It requires synthesis of N-formylglutamate as a starting sub-
strate, which may outweigh the simplicity of the reaction scheme. The enzymatic
scheme that we have developed is efficient, requires enzyme purified in one step,
and can conveniently be scaled up to provide mg amounts of product.
The observation that the cyclohydrolase activity of the human D/C/S enzyme
produces 5-formylH4folate in appreciable amount leads us to speculate on its physio-
logical role. The production of 5-formylH4folate in N. crassa has been shown to
decrease by 70% in a mutant lacking an active cytosolic SHMT, clearly demonstra-
ting that this enzyme is the main source of 5-formylH4folate (6). The remaining
30% may be at least partially accounted for by the mitochondrial isoform of SHMT
(6). Stover and Schirch (28) demonstrated that SHMT hydrolyzes tetraglutamylated
1
methenylH4folate [2] as well as a hydrated form of 2 to 4 with kcat of 0.06 sϪ .
However, the low in vivo concentration of 2 combined with its comparatively high
Km with SHMT, pointed out by the authors, raises interesting questions about the
physiological use of these substrates in production of 4. The rate of production of
4 from 2 by SHMT determined in a coupled assay with D/C/S and pentaglutamylated
H4folate under conditions that approximate physiological concentrations of sub-
strate and enzyme was 0.0004 sϪ1 (calculated from Fig. 3 in 4). The rate of synthesis
of 4 by the D/C domain, using monoglutamylated substrate under similar conditions,
1
is in the same range, approximately 0.001 sϪ (calculated from Fig. 1). As the D/C
domain has little preference for polyglutamylated substrate (22, 23) this rate should
not be enhanced significantly with the polyglutamylated substrate. The similarity
in rates of synthesis under these conditions and the high cellular concentration of
D/C/S, which is in the Ȑ
M
range (29, 30), suggests that this enzyme could also
contribute to in vivo synthesis of 4.
In the conversion of 2 to 4 by the D/C domain, methenylH4folate [3] must be
an intermediate, but the actual mechanism of synthesis of 4 from 3 is still unclear.
It is possible that a small fraction of methenylH4folate is enzymatically hydrolyzed
to yield 4 while the vast majority is converted to 2. Alternatively, maintenance of
a rapid equilibration between 2 and 3 by the cyclohydrolase may enable a higher
rate of chemical hydrolysis of 3 to 4 than in the absence of the enzyme. Regardless
of the mechanism, the properties of the cyclohydrolase provide a practical tool in
vitro and may be a contributor to (6S)-5-formylH4folate synthesis in vivo.
ACKNOWLEDGMENT
This work was supported by MRC Canada Grant MT4479.
REFERENCES
1. MAC
R. L., and Benkovic, S. J., Eds.), Vol. 1, pp. 225–306, Wiley, Philadelphia.
2. KAY, L. D., OSBORN, M. J., HATEFI, Y., AND UENNEKENS, F. M. (1960) J. Biol. Chem. 235, 195–201.
KENZIE, R. E. (1984) in Folates and Pterins: Chemistry and Biochemistry of Folates (Blakley,
H

228
PELLETIER AND MACKENZIE
3. HOPKINS, S., AND
4. STOVER, P., AND
5. HORNE, D. W., PATTERSON, D., AND
S
CHIRCH, V. (1984) J. Biol. Chem. 259, 5618–5622.
S
CHIRCH, V. (1990) J. Biol. Chem. 265, 14227–14233.
OOK, R. J. (1989) Arch. Biochem. Biophys. 270, 729–733.
ONALD, D., COSSINS, E. A., AND CHIRCH, V. (1994) J. Biol. Chem. 269, 28757–
C
6. KRUSCHWITZ, H., M
28763.
C
D
S
7. STOVER, P., KRUSCHWITZ, H., AND
SCHIRCH, V. (1993) in Chemistry and Biology of Pteridines and
Folates (Ayling, J. E., Ed.), pp. 679–685, Plenum, New York.
8. STOVER, P., AND
9. JOLIVET, J., AND
S
B
CHIRCH, V. (1993) Trends Biochem. Sci. 18, 102–106.
ERTRAND, R. (1993) Adv. Exp. Med. Biol. 339, 23–30.
10. SIROTNAK, F. M., CHELLO, P. L., MOCCIO, D. M., KISLIUK, R. L., COMBEPINE, G., GAUMONT, Y.,
ONTGOMERY, J. A. (1979) Biochem. Pharm. 28, 2993–2997.
11. ZHANG, Z. G., AND USTUM, Y. M. (1991) Cancer Res. 51, 3476–3481.
AND
M
R
12. ETIENNE, M. C., THYSS, A., BERTRAND, Y., TOURAINE, R., RUBIE, H., ROBERT, A., AND
MILANO
,
G. (1992) J. Nat. Cancer Inst. 84, 1190–1195.
13. MACHOVER, D., GRISON, X., GOLDSCHMIDT, E., ZITTOUN, J., LOTZ, J. P., METZGER, G., RICHAUD
J., HANNOUN, L., MARQUET, J., AND UILLOT T. (1992) J. Natl. Cancer Inst. 84, 321–327.
14. PELLETIER, J. N., AND AC ENZIE, R. E. (1995) Biochemistry 34, 12673–12680.
15. HUM, D. W., BELL, A. W., ROZEN, R., AND ENZIE, R. E. (1988) J. Biol. Chem. 263, 15946–15950.
16. HUM, D. W., AND ENZIE, R. E. (1991) Protein Eng. 4, 493–500.
17. PELLETIER, J. N., AND ENZIE, R. E. (1994) Biochemistry 33, 1900–1906.
, P. P. K. AND ONES, L. (1967) Biochim. Biophys. Acta 148, 622–628.
19. MORAN, R. G., WERKHEISER, W. C., AND AKRZEWSKI, S. F. (1976) J. Biol. Chem. 251, 3569–3575.
20. BEAVON, J. R. G., AND LAIR, J. A. (1972) Br. J. Nutr. 28, 385–390.
21. TEMPLE, C., ELLIOTT, R. D., ROSE, J. D., AND ONTGOMERY, J. A. (1979) J. Med. Chem. 22, 731–734.
ENZIE, R. E., AND ATTHEWS, R. G. (1984) Biochemistry
,
G
M
K
M
ACK
M
ACK
M
ACK
18. H
O
J
Z
B
M
22. ROSS, J., GREEN, J., BAUGH, C. M., MAC
K
M
23, 1796–1801.
23. MAC
24. FEENY, J., BIRDSALL, B., ALBRAND, J. P., ROBERT, G. C. K., BURGEN, A. S. V., CHARLTON, P. A.,
OUNG, D. W. (1981) Biochemistry 20, 1837–1842.
25. NIXON, P. F., AND ERTINO, J. R. (1980) Met. Enzymol. 66, 547–553.
26. STOVER, P., AND CHIRCH, V. (1992) Anal. Biochem. 202, 82–88.
27. BORTOLUZZI, L. C., AND ENZIE, R. E. (1983) Can. J. Biochem. Cell Biol. 61, 248–253.
28. STOVER, P., AND CHIRCH, V. (1992) Biochemistry 31, 2155–2164.
29. GREEN, J. M., MAC ENZIE, R. E., AND ATTHEWS, R. G. (1988) Biochemistry 27, 8014–8022.
30. STRONG, W. B., AND CHIRCH, V. (1989) Biochemistry 28, 9430–9439.
KENZIE, R. E., AND BAUGH, C. M. (1980) Biochim. Biophys. Acta 611, 187–195.
AND
Y
B
S
M
ACK
S
K
M
S