Available via license: CC BY 4.0
Content may be subject to copyright.
THE JOURNAL OF BIOLOGICAL CHEVI~TRY
Vol. 246, No. 12, Issue of June 25, BP. 3961-3966, 1971
Printed in U.S.A.
Serine Transhydroxymethylase
AFFINITY OF THE ACTIVE SITE FOR SUBSTRATES, SUBSTRATE ANALOGUES, AND ANIONS*
(Received for publication, December 21, 1971)
LAVERNE SCHIRCH
AND
ANN DULLER
Department of Chemistry, Bluj’ton College, B&&n, Ohio 45817
SUMMARY
A number of substrate analogues of serine transhydroxy-
methylase were tested as competitive inhibitors and sub-
strates. Our previous results show that L-se&e, L-threo-
nine, allothreonine, and D-alanine serve as substrates. In
this study, we found that
threo-
and erythro-fi-phenylserine
were cleaved to benzaldehyde and glycine. Tetrahydro-
folate was not required for this reaction. Substrate ana-
logues which contained an amino group adjacent to a car-
boxy& phosphonate, or sulfonate group were bound by the
enzyme. Analogues which lacked either the amino group or
a carboxyl group of the amino acid had little if any affinity
for the active site. The only exceptions to this were several
sulfhydryl compounds. Several D-amino acids were com-
petitive inhibitors with n-cysteine being the most effective.
This shows that the configuration around the o-carbon of the
amino acid is not critical for substrate binding.
The variation of
V,,,
and K,,, with pH for the substrate
allothreonine was determined. Both
V,,,
and K, were
found not to change from pH 5.8 to 7.6. However, an anion
binding site on the enzyme with a pK of 6.2 was found.
This site showed little specificity for anions with all anions
tested serving as competitive inhibitors.
The enzyme-glycine complex exhibits absorption maxima
at 343, 425, and 495 rnp. The equilibrium between the
complexes absorbing at 343 and 425 rnp was found to be
sensitive to both temperature and pH. A group on the en-
zyme-glycine complex with a pK of 6.9 was found to be ln-
volved. The temperature sensitivity was observed to be
due in part to a large increase in entropy, which suggests that
conformational changes in the protein are also involved in the
interconversion of the enzyme-glycine complexes.
Our studies on the purification and properties of serine transhy-
droxymethylase (n-serine: tetrahydrofolate 5: IO-hydroxymethyl-
transferase, EC 2.1.2.1) during the past several years have
shown that the enzyme catalyzes several reactions (l-5). Reac-
tions 1 through 4 are aldol type reactions while Reaction 5 is
* This investigation was supported by Grant GB-7195 from the
National Science Foundation and a grant from the Central Ohio
Heart Association.
classified as a transamination reaction. Tetrahydrofolate is
required for Reactions 1 and 2 but not 3 through 5.
n-Serine + FHI * glycine f 5,10-methylene-FHd (1)
cu-Methylserine + FHI * n-alanine + 5, lo-methylene-
(2)
FJL
n-Threonine * glycine -I- acetaldehyde
Allothreonine e glycine + acetaldehyde
n-Alanine + holoenzyme a pyruvate + pyridoxamine
phosphate + apoenzyme
(3)
(4)
(5)
where FH4 is tetrahydrofolate. Because of this apparent lack of
specificity by the enzyme we have undertaken a more systematic
study of the interaction of substrates with the active site. By
studying the inhibitory and substrate properties of a number of
structural analogues of serine, threonine, and glycine, we have
been able to show that both the cr-amino and carboxyl groups of
the amino acid are essential for substrate binding. The con-
figuration around the a-carbon is not critical for binding since
several n-amino acids were found to be competitive inhibitors.
The configuration of the hydroxyl group of the P-carbon of the
substrate is not critical for either binding or substrate activity.
In extension of our previous finding that both threonine and
allothreonine are substrates, we report in this paper that erythro-
and three-/3-phenylserine are also cleaved by serine transhydroxy-
methylase.
We have also studied the variation of
K,
and V,,, with pH
to determine whether any ionizable groups are involved in the
interaction of the enzyme with substrates and inhibitors. The
data show that both
K,
and V,,,,, do not change in the pH range
5.8 to 7.6. However, there is an anion binding site on the
enzyme with a pK of 6.2 and an ionizable group on the enzyme-
glycine complex with a pK of 6.9. Possible mechanistic inter-
pretations of these results are discussed.
EXPERIMENTAL PROCEDURE
Materials and Methods-AlIothreonine, n-threonine, and
D-
alanine were purchased from Nutritional Biochemicals. Amino-
methyl phosphonic acid was obtained from Calbiochem and
/3-chloro-n-alanine was purchased from Cycle Chemicals. All
other amino acids were acquired from Sigma and the amines
from Aldrich. Serine transhydroxymethylase was purified from
rabbit liver as previously described, except the heat step was
3961
This is an Open Access article under the CC BY license.
Serine Transhydroxymthylase
Vol. 246, No. 12
TABLE I
Aflnity constants and K, values for substrate analogues and
substrates of serine transhydroxymethylase at pH 7.8
Group andcompound
1. n-Alanine. . . . .
Aminomethylphosphonate
Aminomethylsulfonate.. . .
a-Amino isobutyrate.. . . .
n-Alanine. . . .
n-Serine.....................
o-Threonine.................
2. n-Cysteine. . . .
n-Cysteine. . . .
Cysteamine , .
Mercaptoethanol. . .
3-Mercaptopropionate . . .
3. L-Serine.....................
nn-Allothreonine .
n-Threonine.................
Glycine. .
ol-Methylserine .
nn-erythro-&Phenylserine. . .
Dr.-threo-P-Phenylserine. .
-
--
Ki
?nM
14
0.95
170
13
50
31
300
0.3
0.4
25b
36
7
6
0.8 9.6
1.50 2.4
40 0.8
7
>125
>125
0.9
0 Micromoles of product formed per mg of enzyme per min.
b Noncompetitive inhibitor.
c This number is one-half of the experimental value on the as-
sumption that only the L isomer is active.
performed at 63” in the presence of 0.02 M nn-serine and the
purified enzyme was crystallized, as previously described (2).
The enzyme was pure by the criteria of starch gel electrophoresis
and ultracentrifugal analysis. The sedimentation constant was
9.0 S. The enzyme preparations exhibited an absorption maxi-
mum at 428 rnp and an A280:A428 ratio of 5.5 when fully satu-
rated with pyridoxal5-phosphate. We have previously reported
an A280:A428 ratio of 9 to 10 (5). However, the enzyme under
these conditions was only about 65% active in the absence of
added pyridoxal phosphate (2).
Fujioka isolated supernatant and mitochondrial isozymes of
serine transhydroxymethylase in rabbit liver (6). The two
isozymes were found in their studies to be essentially identical
in their physical and chemical properties. The only major
difference found for the two isozymes was their immunochemical
behavior. Studies on amino acid analysis, peptide maps, and
subunit structure indicate that our preparations of serine trans-
hydroxymethylase consist of a homogeneous population of pro-
tein molecu1es.l
Kinetic studies were performed with the assay previously
described for Reaction 4 in which the acetaldehyde is reduced to
ethanol with DPNH and alcohol dehydrogenase (5). The
cleavage of the /3-phenylserines was followed by measuring the
increase in absorbance at 252 rnp due to the formation of the
product benzaldehyde. Kinetic studies were followed with a
Ca.ry model 15 or a Beckman DB spectrophotometer at 30”. A
Cary model 15 was used for the spectral studies.
The pH of reaction solutions was determined both before and
1 L. Schirch and P. Fasella, unpublished results.
after the reaction rate was recorded. Because all anions tested
were inhibitors, those reactions measuring anion inhibition were
performed with albumin neutralized to the appropriate pH with
imidazole as the buffer. The albumin-imidazole buffer did not
inhibit the reaction in the pH range 6.0 to 8.0. The organic
acids used in the anion inhibition studies were neutralized to the
appropriate pH with imidazole. In the studies with amines the
pH was adjusted with HCl. An equivalent amount of potassium
chloride was added to the control so that any observed inhibition
could be attributed to the amine rather than chloride ion.
RESULTS
Substrate
Speciificity-A series of structural analogues of glycine
and serine were tested as competitive inhibitors of allothreonine
in reaction 4 at pH 7.2. We found that analogues which lacked
either the carboxyl group or the a-amino group did not show
measurable inhibition when tested at a concentration of 50 mM
with 1.25 mM nn-allothreonine as the substrate. These com-
pounds included 2-amino ethanol, 1-amino-2-propanol; L-2-
amino-1-propanol, methyl amine, 3-hydroxybutyrate, and ace-
tate. The ability of the threonine analogue, 1-amino-2-propanol,
to serve as a substrate in Reaction 4 was tested. At high enzyme
and substrate concentration no evidence of acetaldehyde forma-
tion could be detected. Similar experiments to determine
whether ethanolamine could serve as a substrate in Reaction 1
also proved negative. The failure of these compounds to show
measurable affinity for the active site shows that the presence of
both the carboxyl and amino groups of the substrate is necessary
for substrate binding.
Analogues characterized by having an amino group adjacent to
a negatively charged carboxyl, phosphonate, or sulfonate group
were found to be competitive inhibitors (Group 1, Table I). The
large differences in affinities for methyl amine, aminomethyl
sulfonate, aminomethyl phosphonate, and glycine show that the
negatively charged group plays an important role in substrate-
enzyme interaction. The glycine analogue aminomethyl phos-
phonate was found to shift the 428 ml* peak of the enzyme to
422 mp. This spectral shift and the small
Ki
for aminomethyl
phosphonate suggest that a Schiff base has been formed with the
enzyme-bound pyridoxal phosphate. However, this compound
does not form complexes absorbing at 343 and 495 rnp which have
been observed with glycine (2). Since it binds almost an order
of magnitude more tightly than glycine we tested it for substrate
activity. Solutions of aminomethyl phosphonate, 0.1 M, and
enzyme, 1 mg per ml, when incubated with either acetaldehyde
or tetrahydrofolate and formaldehyde, formed no new com-
pounds which could be detected by paper chromatography in
three solvents. The failure of this analogue to form multiple
intermediates with the enzyme as observed with glycine and the
lack of evidence of the formation of a hydroxymethylated product
upon the addition of an aldehyde suggest that the carboxyl group
of glycine plays a role in the mechanism of the reaction which
cannot be duplicated by the negatively charged phosphonate.
It is also interesting to note in the Group 1 compounds that
several n-amino acids are competitive inhibitors. We have
previously shown that n-alanine behaves as an analogue of
glycine, forming complexes with the enzyme which exhibit prop-
erties similar to those of the enzyme-glycine complexes (3). If
the n-amino acids bind with their amino and carboxyl groups in
the same manner as the n-amino acids, then the enzyme site has
Issue of June 25, 1971
L. Schirch and A. Diller
3963
sufficient free space to accommodate small P-carbon groups
oriented in either of the two possible positions.
The second group of compounds listed in Table I is thiol ana-
logues of the substrates. High concentrations of L-cysteine have
been shown previously to form a thiazolidine complex with the
enzyme-bound pyridoxal phosphate (1). The thiazolidine ab-
sorbs at 332 rnp and appears not to be bound to the enzyme.
This would suggest that n-cysteine may be a noncompetitive
inhibitor. However, we have observed in this study that both
L- and n-cysteine are extremely effective competitive inhibitors.
Spectra of the enzyme-cysteine complexes at pH 7.3 show that
both L- and n-cysteine at saturation levels react to form com-
plexes which absorb near 340 rnp and 428 mp. These spectra
show that little if any thiazolidine complexes are free in solution.
However, the spectra of the enzyme and n-cysteine solution
slowly changed over a 30-min period to give a compound ab-
sorbing at 332 rnp. This is presumably the thiazolidine complex
which we had previously found (1). o-Cysteine does not form
the 332 rnp complex. Both L- and n-cysteine were tested for
substrate activity in Reaction 1. We found no evidence for the
production of glycine at enzyme concentrations as high as 2 mg
per ml.
Mercaptoethanol and 3-mercaptopropionate were found to be
less effective competitive inhibitors than L- and .n-cysteine (Group
2, Table I). They also form complexes with the enzyme which
absorb at 345 mp. This probably reflects the addition of the
sulfhydryl group to the imine double bond of the bound pyridoxal
phosphate. Cysteamine is a noncompetitive inhibitor and forms
a complex absorbing at 332 rnp which is probably the free thiazoli-
dine compound. The observation that L- and n-cysteine have
lower Ki values than the other sulfhydryl compounds supports
the above observations that the carboxyl and amino groups of
the amino acid are required for binding and that both D- and
n-amino acids can bind to the active site.
The third group of compounds listed in Table I is substrates.
In addition to the compounds previously shown to be substrates
we found that erythro- and threo-P-phenylserine were cleaved to
benzaldehyde and glycine. The benzaldehyde was identified by
its ultraviolet absorption spectrum and glycine was identified by
paper chromatography in a moving phase of pyridine-HzO,
80:20. Addition of either erythro- or threo-P-phenylserine to a
2 mg per ml solution of enzyme results in a spectrum with ab-
sorption maxima at 495, 425, and 343 rnp. This is identical
with the enzyme-glycine spectrum and is additional evidence for
the presence of glycine as a product of the reaction.
The rate of cleavage of both ,&phenylserines was proportional
to substrate concentration from 5 X 10e4 M to 2 X 1O-2 M. Also,
the fl-phenylserines do not inhibit the cleavage of allothreonine
in Reaction 4. These data show that these compounds have
very little affinity for the active site. However, the P-phenyl-
serines are cleaved rapidly by the enzyme in comparison to the
rates observed for the other substrates. At the highest con-
centrations tried, 2 x lo+ M, threo- and erythro#-phenylserine
had activities of 3.1 and 1.7 pmoles of product formed per mg
of enzyme per min, respectively. This is compared to maximal
activities of 0.8 for L-threonine, 2.4 for allothreonine, 0.9 for cy-
methylserine, and 9.6 for L-serine.
The possibility exists that a contaminating enzyme in our
preparations is cleaving the fl-phenylserines. To test this pos-
sibility we took advantage of the ability of n-alanine to inactivate
serine transhydroxymethylase by transaminating the enzyme-
100
80
60
z
0
L 40
2
W
a
s 20
I I I I I 9 I I I I I
80
160 240 320 400 .
TIME (MIN)
FIG. 1. Correlation of the loss of enzymatic activity in cleaving
P-phenylserine and the decrease in absorption at 428 m due to
transamination by n-alanine (Reaction 5). To a solution of en-
zyme, 2.1 mg per ml, in 0.05 M potassium phosphate, pH 7.5, was
added n-alanine to 0.2 M. Spectra were recorded at several inter-
vals to follow Reaction 5 (3). After each spectral run a 0.02~ml
aliquot was removed and added to 2.0 ml of potassium phosphate
buffer containing 5 X 10-3 M threo+phenylserine. The rate of
this reaction was followed by measuring theincrease in absorbance
at 252 m due to the formation of benzaldehvde. The rrrauhshows
the correlation of the extent of the transamination-reaction as
measured by the disappearance of the 428 rnp peak (O-O) and
the rate of benzaldehyde formation (O-O) in the cleavage of
threo-fl-phenylserine.
bound pyridoxal phosphate to pyridoxamine phosphate (Reac-
tion 5). The rate of this reaction is determined by following the
disappearance of the 428 rnp absorption peak (3). Fig. 1 shows
that the rate of loss of activity for cleavage of threo-/3-phenylserine
is the same as the transamination reaction. We conclude from
this that both reactions are catalyzed at the same active site.
In addition to the compounds listed in Table I, erythro- and
threo-&hydroxyaspartate, ,&chloroalanine, homoserine, and 3-
phosphoserine were tested for inhibitory and substrate activity
at a concentration of 25 mM. In each case the data were nega-
tive. The failure of ,&chloroalanine, serine phosphate, and L-
cysteine to show substrate activity indicates that the enzyme is
specific for a ,&hydroxyl group.
Variation in Substrate Binding and Vm,, with pH-A bell-
shaped pH activity curve with a maximum at pH 7.3 has been
shown for Reactions 1 and 4 (6, 7). Experiments were per-
formed in the present study to determine whether the pH curve
could be accounted for by ionizable groups in the enzyme which
affect Km and V,,,. Because we had observed inhibition by
phosphate, we extrapolated all kinetic data to zero buffer con-
centration according to the method of Dixon (8). The rate of
Reaction 4 was determined as a function of allothreonine and
phosphate concentration from pH 5.8 to 8.4. The reciprocal of
the rate of Reaction 4 was plotted against the phosphate con-
centration for each level of allothreonine. The data for experi-
ments performed at pH 6.0 and 8.0 are recorded in Fig. 2, A and
B. The KJ for phosphate can be obtained from the point of
3964 Serine Transhydroxymethylase Vol. 246, No. 12
c 48
-k4
0 0.08 0.16 0 200 400
PHOSPHATE IhLLOTHREONINE
FIG.
2. Phosphate inhibition of allothreonine in reaction 4 at
pH 6.0 and 8.0. The reaction was followed by measuring the rate
of disappearance of DPNH at 340 rnp. A and B show the effect
of increasing concentrations of potassium phosphate at pH 6.0
and 8.0, respectively, at the following four mM concentrations of
nn-allothreonine: O-O, 2.5; A--A; 5.0, cl-0, 10;
O-O, 25. C and D are secondary plots of the ordinate inter-
cepts of A and B against the reciprocal of the allothreonine con-
centrations.
intersection of the lines in the left-hand quadrant according to
the method of Dixon (8). A large decrease in phosphate inhibi-
tion between pH 6.0 and 8.0 is readily observed. To obtain the
K, for allothreonine at zero phosphate concentration the ordinate
intercepts are plotted in a secondary graph against the reciprocal
of the allothreonine concentration (Fig. 2, C and 0). The data
show that only a small change occurred in the K, for allothreo-
nine between pH 6.0 and 8.0. The reciprocal of V,,,,, at zero
phosphate concentration is obtained from the ordinate intercept
in the secondary graphs (Fig. 2, C and D). Studies identical
with those presented in Fig. 2 were performed at several pH
values between 5.8 and 8.4. Fig. 3 summarizes these data
according to the method of Dixon by correlating pK,, pK;, and
log vnmx with pH (8). The data in Fig. 3a show that the V,,,
for the reaction is constant between pH 5.8 and 7.6. The pKi
for the phosphate, Fig. 3b, decreases sharply between pH 5.8
and 7.0. According to the rules outlined by Dixon for inter-
preting these curves the decrease in phosphate inhibition can be
accounted for by the ionization of a group on the enzyme with a
pK, of 6.2 (8). Data presented later in the paper show that the
change in pKi cannot be explained by the ionization of monobasic
phosphate to dibasic phosphate. The variation of pK,, Fig. 3a,
with pH shows a small change between pH 5.8 and 6.5 which can
probably be ascribed to inhibition by small amounts of anions
other than phosphate in the reaction mixture. The changes in
pK, and log I’,,, (Fig. 3a) above pH 7.6 could be due to the
ionization of allothreonine or ionizable groups on the enzyme.
Attempts to obtain data at higher pH values gave nonreproduc-
ible results.
Glycine is a competitive inhibitor of allothreonine in Reaction
4. Using the method of Lineweaver and Burk the KC for glycine
was determined at several pH values between pH 6.0 and 8.2
2.0 I I I I I I I
b
pKi 1 .o
3.0 1 I I I I I 1
I I I I I I 1
7.0 PH 81)
FIQ.
3. Variation of V,,, and pK, for allothreonine, and pKi
for phosphate and glycine with pH. a, the values for pK,
(O-O) and P,,,,= (A---A) were obtained from graphs calcu-
lated by the method used in Fig. 2, C and D. b, the values for
pKi for phosphate were determined from the intercept of tbe lines
as shown in the graph of Fig. 2A. The points represented as n
in a and 6 were obtained from the experiments reported in Table
III. c, the values for the pK; for glycine were calculated from
competitive inhibition studies of glycine in Reaction 4. The
buffers used in these experiments were imidazole and albumin.
TABLE
II
Inhibition of swine transhydroxmethylase by anions at pH 6.3
Anion Ki
?nM
Chloride. .
Iodide. .
Cacodylate . .
Phosphate.
Sulfate. .
Formate. . .
Acetate. .
Citrate. . . .
.
36
35
140
56
47
90
83
140
(9). The results in Fig. 3c show that the pKi of glycine gradually
increases between pH 6.0 and 7.6. Above pH 7.6 the curve is
essentially the same as the one for allothreonine.
The experimental results from Fig. 3 show that the bell-shaped
pH activity curve previously observed can be explained by a
decrease in phosphate inhibition between pH 5.8 and 7.3 and a
decrease in V,,, above pH 7.6.
Anion Inhibition-A variety of anions were tested as competi-
tive inhibitors of allothreonine at pH 6.3. Table II records the
Issue of June 25, 1971 L. Schirch and A. Biller 3965
TABLE
III
Competitive inhibition of allothreonine cleavage by phosphate,
chloride, and sulfate between pH 6.0 and 7.2
PH
6.0
6.3
6.7
7.2
Chloride
29
36
97
130
Ki
Sulfate
9nM
22
47
82
220
Phosphate
38
56
115
180
Ki for these anions. There is little specificity shown for the
structure and charge of the anion. The inhibitory effects of
chloride and sulfate ions were tested over the pH range 6.0 to 7.2
to determine whether the variation in phosphate inhibition
(Fig. 3a) was due to the ionization of monobasic phosphate
(Table III). Since the results show that chloride and sulfate
behave like phosphate, the variation in inhibition is due to ionic
interaction with a group on the enzyme with a pK of 6.2, rather
than to the ionization of phosphate or the type of anion inhibitor.
Spectra of Enzyme-Glycine Compkxes--We have previously re-
ported that the spectrum of the enzyme-glycine complex varies
with pH (4). We have repeated this study in more detail to
determine whether the spectral change can be attributed to the
ionization of a single group of the enzyme-substrate complex.
Fig. 4 records the spectral changes observed when the pH of a
solution of enzyme and 0.2
M
glycine is varied. At pH 5.8 the
spectrum differs from that of the free enzyme by a shift of the
428 rnp peak to 425 rnp and a small increase in absorbance at
343 rnp. As the pH is increased the 425 rnp peak decreases and
the peaks at 343 and 495 rnp increase. Above pH 8.4 the 425 rnE.1
and 343 rnp peaks are insensitive to pH increases, but the 495
rnp peak continues to increase. This observation is similar to
our previous work with the glycine analogue n-alanine which
forms a pH-sensitive complex absorbing at 505 rnp. The pK
for the formation of this complex was found to be 9.1 (5). We
were unable to do similar studies with the 495 rnp absorbing
complex of glycine because the changes in absorbance were too
small and at pH values above 9.0 were not stable. However, a
qualitative examination of the spectra suggests that if the forma-
tion of the complex at 495 rnp is attributable to the ionization of
a single proton the pK for this group would be above 9.0.
Evidence that the shift from 425 to 343 rnp with increasing
pH can be attributed to the loss of a single proton by the enzyme-
glycine complex absorbing at 425 rnb is shown in the inset of
Fig. 4. The graphical method is based on the method of Sizer
and Jenkins for the ionization of a single proton EX. H a EX +
H+ (10). The method predicts a linear relationship between
the reciprocal of the absorption change and hydronium ion with
the negative abscissa intercept being equal to -K,. The data
show that the changes in absorbance at 343 and 425 rnp both
give straight lines with almost identical K, values. We conclude
that the spectral shifts are due to the dissociation of a single
proton from the enzyme-glycine complex with a pK of 6.9.
We have also observed that the spectrum of the enzyme-
glycine complex varies with temperature. At pH 7.6 and 37” a
1.5 mg per ml solution of enzyme and glycine, 0.2
M,
exhibits an
absorption of 0.150 at 425 rnp and 0.115 at 343 rnp. Upon
lowering the temperature to 8” the 425 rnp absorption decreases
320 380 440 500
WAVELENGTH (MIJ)
FIG. 4. Variation of the spectrum of the enzyme-glycine com-
plex with pH. Solutions of enzyme, 0.5 mg per ml, and glycine,
0.2
M,
in 0.01
M
potassium phosphate were adjusted to the appro-
priate pH by the addition of KOH. Spectra were recorded at 30”
in l-ml cuvettes with a l-cm light path. Inset, graphical analysis
of the changes in absorbance at 343 and 425 rnp with pH as shown
in Fig. 4. The data in Figure 4 were graphed according to the
equation
l/AA - (Et)EEH = l/@t)(EE - EEH) + H+/ZCJ@~)(EE - eEH)I
(E) and (EH) are the concentrations of the unprotonated and
protonated enzyme-glycine complexes, respectively. AA -
(E J ezH is the absorbance change between the sbectru& of EH and
the eauilibrium mixture of E and EH. K, is the dissociation
constant for the reaction EH G E + H+ and can be obtained from
the negative abscissa intercept. The spectrum of EH was as-
sumed to be the same as that of the free enzyme since at pH 5.8
there is only a negligible difference.
to 0.047 and the absorption at 343 rnp increases to 0.195. Eight
spectra were taken of the solution between the above-mentioned
temperatures. We calculated equilibrium constants for the
complexes absorbing at 343 and 425 rnp at each temperature.
A plot of log K against the reciprocal of the absolute temperature
is linear. From this graph we calculated that the enthalpy
change for the temperature-sensitive reaction EX343 + EX4z5 is
13.3 kcal per mole. At 37” the free energy change is -670 cal
per mole and the change in entropy is 45 e.u. The large positive
entropy change drives the reaction to the right and suggests that
a protein conformation is involved.
DISCUSSION
Amino acid substrates bind to B6 enzymes by forming a Schiff
base with the enzyme-bound pyridoxal phosphate. Scheme I
shows a proposed mechanism for this reaction. As previously
discussed by Karpeisky and Ivanov (11)) in Structure I the amino
acid substrate is above the plane of the pyridinium ring, while in
Structure III both are in the same plane. It is clear that in going
from St.ructure I to III something must move. Karpeisky and
Ivanov conclude that in glutamic-aspartic transaminase the
transition from I to III is accomplished by a rotation of the
pyridoxal phosphate, with the amino acid remaining stationary.
3966 Serine Transhydroxymethylase Vol. 246, No. 12
Is
SCHEME 1
This mechanism implies that in Structure I the amino and car-
boxy1 groups of the substrate are bound to the protein. An
alternative mechanism is that the pyridoxal phosphate remains
stationary and the amino acid moves into the plane of the pyri-
dinium ring. This mechanism suggests that the amino group
binds first in Structure I and that the carboxyl group of the sub-
strate binds as the molecule moves into the plane of the ring.
Yet another mechanism would involve a change in conformation
of the protein as the major factor in bringing the amino acid and
pyridoxal phosphat,e into the same plane.
Although the experiments in this paper do not rule out any of
the mechanisms they favor the last two alternatives mentioned.
The data in Table I show that both the amino and carboxyl
groups of the amino acid are required for binding. The failure
of such compounds as ethanolamine and 3-hydroxybutyrate to
show measurable affinity for the active site can be explained by
sequential binding where the carboxyl group binds only after an
initial attack by the amino group. This mechanism is also
supported by the anion inhibition data in Figs. 2 and 3. These
experiments show that the anion binding site is not the site which
binds the carboxyl group since anion binding is pH-sensitive but
substrate binding is not. In fact, above pH 7.4 anions appar-
ently do not bind at the active site. This must mean that, if
there is a cationic site on the enzyme which binds the carboxyl
group of the substrate, it is either masked or generated subse-
quent to the first formed complex.
We do not know the structure of the enzyme-glycine complex
absorbing at 343 rnp but several interesting properties are re-
vealed about it in this paper. In Fig. 4 the data show that at
pH 8.2 where the pH titration of the ionizable group affecting
the spectral shift is essentially complete about one-half of the
complex absorbing at 425 rnp remains. However, if one conducts
the titration at a very low temperature, the spectral shift from
425 mp to 343 rnp is complete. This is accounted for by the
temperature sensitivity of the equilibrium.
We
interpret this
data to mean that the proton being dissociated in the enayme-
glycine complex is not associated with a group on the chromo-
phore. Apparently the unprotonated enzyme-glycine complex
can exist in two conformational states, one favoring the 343 rnp
absorbing species and the other the 425 rnp absorbing species.
The temperature sensitivity of the enzyme-glycine complexes
may give us a spectral probe of these enzyme conformational
changes and how they relate to the mechanism of the reaction.
We are also investigating the possibility that the complexes of
D- and L-cysteine which absorb near 340 rnp may be analogous
structures to the 343 mp absorbing enzyme-glycine complex.
The two cysteine isomers may be valuable in elucidating the
interaction of inhibitors and substrates at the active site since
they have three centers of interaction. Our observation that
only n-cysteine forms a thiazolidine complex could give us a
method of determining the spatial relationship of the carboxyl,
amino, and sulfhydryl groups.
REFERENCES
1. SCHIRCH, L., AND MASON, M., J. Biol. Chem., 237, 2578 (1962).
2. SCHIRCH, L., AND MASON, M., J. Biol. Chem., 238, 1032 (1963).
3. SCHIRCH, L., AND JENKINS, W. T., J. Biol. Chem., 239, 3797
(1964).
4. SCHIRCH: L., AND JENKINS, W. T., J. Biol. Chem., 239, 3801
(1964).
5. SCHIRCH, L., AND GROSS, T., J. Biol. Chem., 243, 5651 (1968).
6. FUJIOKA, M., Biochim. Biophys. Acta, 186, 338 (1969).
7. SCHIRCH, L., Ph.D. thesis, The University of Michigan, 1963.
8. DIXON, M., Biochem. J., 55, 170 (1953).
9. LINEWEAVER, H., AND BURIC, D., J. Amer. Chem. Sot., 66, 658
(1934).
10. SIZER, I. W., AND JENKINS, W. T., in E. E. SNELL, P. M.
FASELLA. A. E. BRAUNSTEIN, AND A. ROSSI-FANELLI (Edi-
tars), Chemical and biological aspects of pyridoxal catalysis,
Pergamon Press, New York, 1963, p. 123.
11. KARPEISKY, M. Y., AND IVANOV, V. I., Nature, 210, 493 (1966).