This document summarizes a study on the interaction of pyridoxal-5'-phosphate (PLP) with the apo form of sheep liver serine hydroxymethyltransferase (SHMT). The key findings are: 1) Removal of PLP from the holoenzyme converts it to the inactive apoenzyme and addition of PLP back restores full enzyme activity, demonstrating PLP's role in catalysis. 2) PLP binding to the apoenzyme occurs in two phases, a very rapid initial phase and a slower secondary phase, forming an internal aldimine linkage critical for activity. 3) While the secondary structures of the apo and holo forms are identical, they have different

1. ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS
Vol. 330, No. 2, June 15, pp. 363–372, 1996
Article No. 0263
Interaction of Sheep Liver Apo-serine
Hydroxymethyltransferase with Pyridoxal-5؅-phosphate:
A Physicochemical, Kinetic, and Thermodynamic Study
Bhaskar Brahatheeswaran,*,1,2
V. Prakash,† Handanahal S. Savithri,* and N. Appaji Rao*
*Department of Biochemistry, Indian Institute of Science, Bangalore 560 012, India; and †Department of Protein
Technology, Central Food Technological Research Institute, Mysore 570 013, India
Received December 26, 1995, and in revised form April 1, 1996
zyme could be converted into an active holoenzyme
while the dimeric form could not be reconstituted intoSheep liver serine hydroxymethyltransferase (EC
an active enzyme. These results demonstrate that PLP2.1.2.1) is a homotetramer of Mr 213,000 requiring pyri-
plays an important role in maintaining the structuraldoxal-5؅-phosphate (PLP) as cofactor. Removal of PLP
integrity of the enzyme by preventing the dissociationfrom the holoenzyme converted the enzyme to the apo
of the enzyme into subunits, in addition to its functionform which, in addition to being inactive, was devoid
in catalysis. ᭧ 1996 Academic Press, Inc.of the characteristic absorption spectrum. Upon the
Key Words: apo-serine hydroxymethyltransferase;addition of PLP to the apoenzyme, complete activity
PLP interaction; internal aldimine; dissociation.was restored and the visible absorption spectrum with
a maximum at 425 nm was regained. The interaction
of PLP with the apoenzyme revealed two phases of
reaction with pseudo-first-order rate constants of 20 {
Cytosolic serine hydroxymethyltransferase (SHMT)3
5 s01
and 12.2 { 2.0 1 1003
s01
, respectively. However,
(EC 2.1.2.1) is a homotetrameric enzyme of identicaladdition of PLP to the apoenzyme did not cause gross
subunits of Mr 53,000. The pyridoxal-5؅-phosphateconformational changes as evidenced by circular di-
(PLP) tightly bound at the active site of each subunitchroic and fluorescence spectroscopy. Although con-
participates in a reasonably well-understood mannerformationally apoenzyme and holoenzyme were indis-
in catalysis (1). In an earlier study of the interactiontinguishable, they had distinct apparent melting tem-
of the enzyme with several compounds, such as D-peratures of 51 { 2 and 58 { 2ЊC, respectively, and the
cycloserine (2), o-amino-D-serine (OADS) (3), me-reconstituted holoenzyme was thermally as stable as
thoxyamine (4), and thiosemicarbazide (5), it wasthe native holoenzyme. These results suggested that
there was no apparent difference in the secondary shown that these compounds bind to the active site
structure of holoenzyme, apoenzyme, and reconstitu- PLP in two distinct steps to generate an external aldi-
ted holoenzyme. However, sedimentation analysis of mine or corresponding derivatives such as hydrazone
the apoenzyme revealed the presence of two peaks of and oxime. These derivatives because of their lower
S20,w values of 8.7 { 0.5 and 5.7 { 0.3 S, respectively. A affinity to the enzyme dissociate from the active site
similar pattern was observed when the apoenzyme leading to the formation of the apoenzyme. Previous
was chromatographed on a calibrated Sephadex G-150 studies have shown that the addition of serine con-
column. The first peak corresponded to the tetrameric ferred thermal as well as conformational stability to
form (Mr 200,000 { 15,000) while the second peak had the enzyme and converts the enzyme from an ‘‘open’’
a Mr of 130,000 { 10,000. Reconstitution experiments
to a ‘‘closed’’ form, thereby giving reaction specificity
revealed that only the tetrameric form of the apoen-
to the enzyme (2, 6, 7).
1
Present address: Department of Biochemistry, Uniformed Ser- 3
Abbreviations used: apo-SHMT, apo-serine hydroxymethyltrans-
ferase; holo-SHMT, holo-serine hydroxymethyltransferase; PLP, pyr-vices University of the Health Sciences, 4301 Jones Bridge Road,
Bethesda, MD 20814-4799. idoxal-5؅-phosphate; EDTA, ethylenediaminetetraacetic acid; DTT,
dithiothreitol; app.Tm , apparent melting temperature; H4-folate, tet-2
To whom correspondence should be addressed. Fax: (301) 295-
3512; E-mail: BHASKARB@usuhsb.usuhs.mil, BHASKARB@mx1.u rahydrofolate; OADS, o-amino-D-serine; DCS, D-cycloserine; TSC,
thiosemicarbazide.suhs.mil.
3630003-9861/96 $18.00
Copyright ᭧ 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

2. 364 BRAHATHEESWARAN ET AL.
solutions were mixed under nitrogen pressure (4 kg/cm2
). The deadAmong the PLP-dependent enzymes, aspartate ami-
time of the instrument was fixed at 10 ms and the slit width wasnotransferase (EC 2.6.1.1) and tryptophan synthase
set at 1.4 nm in all the experiments. All the reaction curves presented
(EC 4.2.1.20) are the most widely studied enzymes with were the average of at least five sets of experiments. The k values
respect to their oligomeric state and the role of PLP in were calculated manually by the procedure of Hiromi (16).
maintaining the structure (8–10). In the case of trypto- Thermal denaturation studies. The thermal denaturation of
sheep liver apo-SHMT in the presence or absence of PLP and holo-phan synthase, it was observed that removal of PLP
SHMT was carried out in a Gilford Response II spectrophotometerled to the dissociation of the native enzyme into a and
from Ciba Corning Diagnostics, U.S.A. A clear solution (250 ml of 0.3
b subunits which could be reconstituted upon the addi-
mg/ml) of the protein in 50 mM potassium phosphate buffer, pH 7.2,
tion of PLP (9). SHMT being a homotetramer, it was containing 1 mM EDTA and 1 mM DTT with or without PLP with
of interest to examine the effect of removal of PLP on appropriate blanks was taken in matched thermal quartz cuvettes
and allowed to equilibrate at 25ЊC in the instrument to obtain theits oligomeric state and the role of PLP in maintaining
baseline. The samples were heated from 25 to 95ЊC at the rate ofthe structure of the protein. For this purpose, a compar-
1ЊC/min using the software built into the instrument. The absorbance
ative analysis of the thermal stability and conforma- change in each case was monitored at 287 nm and data were aver-
tional states of the holoenzyme, apoenzyme, and recon- aged from three experiments. The first derivative of the thermal
stituted holoenzyme was undertaken. denaturation profiles was used to evaluate the apparent transition
temperatures (app.Tm) of the enzyme.
The thermal denaturation results were analyzed according to the
EXPERIMENTAL PROCEDURES method of White and Olsen (17) in which the fraction of protein in
the denatured state (FD) is given by
Materials. The following biochemicals were obtained from Sigma
Chemical Company (St. Louis, MO): 2-mercaptoethanol, DL-dithio-
threitol (DTT), ethylenediaminetetraacetic acid (EDTA), PLP, L-ser-
FD Å
AI 0 AN
AD 0 AN
, [1]
ine, L-cysteine, D-alanine, pyridoxamine phosphate, pyruvate, 5,5-
dimethyl-1,3-cyclohexanedione, 2,5-diphenyl oxazole, and carboxy-
methyl–Sephadex C-50. Sephacryl S-200 and Sephadex G-150 were where AN is the absorbance of protein solution at 20ЊC, AD is the
purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). Tet- absorbance of protein solution in the plateau region (70ЊC), and AI
rahydrofolate (H4-folate) was prepared by the method of Hatefi et al. is the absorbance of the protein solution at different temperatures
(11). L-[3-14
C]Serine (specific activity 53 mCi/mmol) was purchased between 20 and 70ЊC. The app.Tm was defined as the temperature
from Amersham International (Buckinghamshire, UK). All other at which the value of FD was 0.5.
chemicals were of analytical grade.
Sedimentation velocity. Sedimentation velocity experiments were
Enzyme purification and preparation of apoenzyme. Sheep liver performed in a Beckman Model E analytical ultracentrifuge equipped
SHMT was purified as described by Baskaran et al. (3). Apoenzyme with a rotor temperature indicator and control unit. Experiments
of SHMT was prepared according to the method of Schirch et al. were carried out at 27 { 2ЊC and at 50,740 rpm. Schlieren patterns
(12) with minor modifications. D-Alanine (200 mM) was added to the were recorded on Agfa Panachrome film 100 ASA. The S20,w values
holoenzyme in 50 mM potassium phosphate buffer, pH 7.2, containing were calculated according to the standard procedure (18).
1 mM EDTA, 1 mM DTT, and 200 mM ammonium sulfate and incu-
Circular dichroism. Circular dichroic (CD) measurements werebated at 37ЊC for 3 h. The reaction mixture was then dialyzed over-
made in a Jasco J-500 A automated recording spectropolarimeter.night in cold with two changes against the same buffer without am-
The spectropolarimeter was continuously purged with nitrogen be-monium sulfate.
fore and during the experiments. Slits were programmed to yield 10The apoenzyme was also prepared by the method of Jones and
A˚ bandwidth at each wavelength.Priest (13). The holoenzyme was mixed with 100 mM L-cysteine and
The CD spectra obtained were plotted as molar ellipticity values20% ammonium sulfate and dialyzed for 6 h against 500 ml of 50
assuming a relative Mr of 213,000 for the sheep liver enzyme withmM potassium phosphate buffer, pH 7.2, containing 1 mM EDTA, 1
a mean residue weight of 110.24. The u value was calculated ac-mM DTT, 100 mM L-cysteine, and 20% ammonium sulfate. Following
cording to the method of Greenfield and Fasman (19). All CD spectrathis, the enzyme was dialyzed extensively against 500 ml of 50 mM
were recorded at 22 { 2ЊC in 50 mM potassium phosphate buffer,potassium phosphate buffer, pH 7.2, containing 1 mM EDTA and
pH 7.2, containing 1 mM EDTA and 1 mM DTT using the same buffer1 mM DTT but without L-cysteine in cold with two changes. The
as blank. The protein concentrations used were 7.0 mM for far-UVapoenzyme prepared by both methods had minimal absorbance
CD spectra and 14.0 mM for visible CD spectra, respectively.(õ0.01) in the visible region (350–500 nm) and negligible enzyme
Titration microcalorimetry. The titration of sheep liver apo-activity.
SHMT (32.8 mM) with PLP (500 mM) was carried out in a Microcal
Enzyme assay. The enzyme activity was determined as described
Omega ultrasensitive calorimeter (Microcal, Inc., Northampton,
by Manohar and Appaji Rao (2) using L-[3-14
C]serine as substrate
MA). A window-based software package (ORIGIN Version 1.1) also
(14). Protein concentration was determined by the method of Lowry
supplied by Microcal, Inc., was used to analyze and plot the data.
et al. (15) using bovine serum albumin as the standard.
The reaction cell was equilibrated at 15ЊC using insulated constant
Absorption spectroscopy. All spectral measurements were carried temperature circulating waterbath. Internal calibration was per-
out in 50 mM potassium phosphate buffer, pH 7.2, containing 1 mM formed for all constants and other parameters of the microcalori-
EDTA and 1 mM DTT at 25 { 2ЊC. Absorption spectra were recorded meter. PLP (500 mM) was taken in 250-ml computer-controlled injec-
in a Shimadzu UV-240 Graphicord double-beam spectrophotometer. tion syringe and 12 injections of 20 ml each at 3-min intervals with
Enzyme solutions were extensively dialyzed against the buffer men- a delivery time of 15 s were set by the program in titrating the
tioned above before spectral measurements. protein against the ligand in the automatic mode. The data analysis
was done according to the method of Wiseman et al. (20) using theStopped-flow spectrophotometry. Rapid kinetic measurements
were performed in a Union-Giken RA 401 stopped-flow spectropho- ORIGIN software supplied by Microcal, Inc. Similarly, the calorimet-
ric analysis was also carried out for the holoenzyme (32.8 mM) withtometer equipped with a 10-mm cell. The data were collected using
a NEC-PC 9801E computer interfaced to the spectrophotometer. The PLP (500 mM) as control under similar experimental conditions.
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

3. 365SHEEP LIVER APO-SERINE HYDROXYMETHYLTRANSFERASE
RESULTS PLP (1 mM) under pseudo-first-order conditions and
the reaction followed in the millisecond range, the
Reconstitution of Apoenzyme with PLP as Monitored
appearance of 423-nm absorbance was very rapid
by Enzyme Activity Measurements
reaching a near saturation value within 60 ms (Fig.
One of the critical tests for the presence of a coen- 2A). From the derivative plot of the data, a pseudo-
zyme on an enzyme and its involvement in catalysis is first-order rate constant was calculated to be 20 { 5
its removal resulting in loss of activity and its readdi- s01
which was very rapid compared to many other
tion leading to complete restoration of activity. The interactions studied so far ([3]; Bhaskar, unpub-
holoenzyme had a specific activity of 6.61 which was lished results). This rapidity could be attributed to
normalized to 100%. The apoenzyme had a specific ac- the chemical reaction involved in the formation of a
tivity of 0.37. This was probably due to incomplete re- covalent linkage between Lys 256 of the enzyme and
moval of PLP from the holoenzyme. When the apoen- PLP to form the internal aldimine (21).
zyme, which was inactive, was preincubated with 100 When the subsequent slower reaction was monitored
mM PLP for 30 s and assayed by starting the reaction in the seconds range, the increase in absorbance at 423
with the addition of H4-folate, 70% of the original en- nm was gradual and continued beyond 170 s. Further
zyme activity was regained. However, when incubation progress of the reaction could not be monitored by
with PLP was carried out for 4 and 10 min and assayed stopped-flow spectrophotometry as the upper time limit
for activity, a greater percentage of activity, viz. 87 and of measurement of the instrument was reached (Fig.
98%, was restored. 2B). The pseudo-first-order rate constant for this phase
of the reaction was calculated from the derivative plot
and was 12.2 { 2.0 1 1003
s01
which was nearly 3Kinetics of Reconstitution of Apoenzyme with
orders of magnitude slower than the first, faster step.Pyridoxal-5؅-Phosphate
This phase probably represented a conformational
Visible absorption spectroscopy. Reconstitution of change in the interaction completing reconstitution of
apoenzyme with PLP was monitored by the appearance apoenzyme to holoenzyme.
of the visible absorption spectrum. Figure 1 (inset)
shows the reconstitution of apo-SHMT with PLP. Un- Titration Microcalorimetry of Binding of PLP
der stoichiometric concentrations of 2.5 mM protein and to Apo-SHMT
10 mM PLP, the reconstitution, reflected by the appear-
The microcalorimetric titrations were carried out byance of a 425-nm peak due to the formation of internal
titrating PLP with the apoenzyme and holoenzyme.aldimine (Schiff base linkage between aldehyde group
Figure 3 shows the profile of titration and the energyof PLP and e–NH2 group of lysine 256 at the active
changes in the system as a result of PLP binding. Thesite of the enzyme—Usha et al. [21]), was initially very
apoenzyme titrates PLP significantly and the energyrapid followed by a gradual increase which was com-
changes saturate by the 10th injection (Fig. 3, curveplete in about 5 min (Fig. 1 inset). When DA425 was
b), whereas in the holoenzyme no significant changesplotted against time (Fig. 1), a typical hyperbolic satu-
are observed (Fig. 3, curve a). The difference of nearlyration pattern reaching a maximum value in about 10
30 cal/mol of the enzyme suggests the interaction to bemin was observed. The rate constant from this part of
thermodynamically favorable toward binding of PLP tothe curve was calculated to be 0.63 min01
. However,
the apoenzyme, which could reflect the chemical naturewhen the reaction was carried out under pseudo-first-
of the reaction involved in the formation of internalorder conditions of 2.5 mM enzyme and 100 mM PLP,
aldimine.the reaction was complete within the time range of
the regular spectrophotometer. However, under these
Thermal Stability of Apo-SHMT in the Absence andconditions there was an immediate increase in ab-
Presence of PLPsorbance at 425 nm followed by a slow increase which
indicated that there could be a possibility of a presence In order to evaluate the role of PLP in maintaining
of a faster reaction component in this reconstitution the structural integrity of SHMT, a comparative analy-
process not detectable by ordinary spectrophotometer. sis of thermal stability of the holoenzyme, apoenzyme,
These results suggested that PLP was reacting in a and reconstituted holoenzyme was carried out by moni-
concentration-dependent but possibly phased manner. toring the absorbance changes at 287 nm upon pro-
In order to determine this faster reaction component, grammed heating of the samples. Figure 4A shows the
rapid reaction kinetic measurements were carried out. effect of temperature on absorbance changes of holoen-
zyme, apoenzyme, and reconstituted holoenzyme. ItStopped-flow spectrophotometry. The rapid phase
in the interaction of PLP with the apoenzyme was can be seen from first-derivative plots of this change
(Fig. 4B), that the app.Tm of holoenzyme was 58 { 2ЊC.monitored kinetically by stopped-flow spectropho-
tometry. When the apoenzyme (5 mM) was mixed with Upon the removal of PLP from the holoenzyme, the
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

4. 366 BRAHATHEESWARAN ET AL.
FIG. 1. Reconstitution of sheep liver apo-SHMT with PLP. Plot of D absorbance at 425 nm as a function of time. (Inset) The sheep liver
apo-SHMT (500 mg/ml) was taken in 50 mM potassium phosphate buffer, pH 7.2, containing 1 mM EDTA and 1 mM DTT, and its absorption
spectrum was recorded in the range 330–550 nm in a Shimadzu UV–visible recording spectrophotometer UV-240 Graphicord. PLP (10 mM)
was added to the enzyme and mixed well, and spectrum was recorded every minute after addition.
FIG. 2. Rapid reaction kinetics of reconstitution of apo-SHMT with PLP. (A and B) Apo-SHMT (1 mg/ml) in 50 mM potassium phosphate
buffer, pH 7.2, containing 1 mM EDTA and 1 mM DTT was taken in reservoir A and PLP (1 mM) prepared in the same buffer in reservoir
B of a Union-Giken RA 401 stopped-flow spectrophotometer. The solutions were mixed and the absorbance changes were recorded at 423
nm in (A) milliseconds and (B) seconds ranges. The tracings shown are an average of at least five different experiments.
AID ARCH 9472 / 6b1a$$9472 05-11-96 08:14:43 arca AP: Archives

5. 367SHEEP LIVER APO-SERINE HYDROXYMETHYLTRANSFERASE
Effect of Removal of PLP on the Velocity
Sedimentation of Apo- and Holo-SHMT
The oligomeric state of the holoenzyme, apoenzyme,
and reconstituted holoenzyme was monitored by sedi-
mentation velocity experiments. Any structural change
brought about by removal of PLP from the tetrameric
holoenzyme would in turn affect the frictional coeffi-
cient of the enzyme and hence the S20,w value. Results
of velocity sedimentation of apoenzyme and holoen-
zyme in the absence and presence of PLP showed that
the holoenzyme sedimented as a single symmetrical
peak (Fig. 6A, bottom) with an S20,w value of 8.7 { 0.5
S. However, the apoenzyme (Fig. 6A, top) sedimented
as a bimodal pattern with the fast moving peak having
an S20,w value of 8.7 { 0.5 S and the slow moving peak
FIG. 3. Titration microcalorimetry of binding of PLP to apo-SHMT. having an S20,w value of 5.7 { 0.3 S.
Plot of processed data in the derivative format obtained for 12 auto- When 200 mM PLP was added to both the apoenzyme
matic injections, 20 ml each of 500 mM PLP, into the sample cell
and the holoenzyme and subjected to analytical ultra-containing sheep liver apo-SHMT or holo-SHMT at a concentration
centrifugation, the holoenzyme continued to sedimentof 32.8 mM. Other conditions used were 23ЊC, 50 mM potassium phos-
as a single peak with an S20,w value of 8.7 S (Fig. 6B,phate buffer, pH 7.2, containing 1 mM EDTA and 1 mM DTT. Total
run time was 40 min. (Curve a) holo-SHMT, (curve b) apo-SHMT. bottom), whereas the reconstituted holoenzyme still
sedimented as two discrete peaks with similar S20,w
values of 8.7 and 5.7 S as was observed for the apoen-
zyme (Fig. 6B, top).
apoenzyme was less thermally stable with an app.Tm
of 51 { 2ЊC. However, upon readdition of PLP to the
Oligomeric Nature of SHMTapoenzyme followed by extensive dialysis against
buffer containing 100 mM PLP, the reconstituted holo- The holoenzyme (60 mM) upon gel filtration using a
enzyme had an app.Tm of 57 { 2ЊC. This value was calibrated Sephadex G-150 column (0.6 1 90 cm) gave
nearly equal to that of the native holoenzyme, clearly a single symmetrical peak (Mr Å 200,000 { 15,000)
indicating the restoration of the thermal stability of when absorbance at 280 and 423 nm was monitored
the enzyme. (Fig. 7A). When the apoenzyme (60 mM) was chromato-
graphed on the same calibrated Sephadex G-150 col-
umn, it separated into two peaks—Peak I (Mr Å
Effect of Removal of PLP on the Secondary Structure
200,000 { 15,000), which corresponded to the tetra-
of SHMT
meric form of the enzyme, and Peak II which had a Mr
130,000 { 10,000 (Fig. 7B). When the apoenzyme (67.5Figure 5A shows the far-UV CD spectra of apoen-
zyme as well as the holoenzyme. There was no marked mM) was loaded onto the same column equilibrated with
100 mM PLP so that reconstitution could take place ondifference in the secondary structure upon removal of
PLP from the holoenzyme, either visibly or by evalua- the column, the reconstituted holoenzyme still sepa-
rated as two peaks, albeit Peak I was in greatertion of the a, b, and aperiodic contents (by CD Estima
method of Fasman—program courtesy of Professor amounts compared to Peak II (Fig. 7C), suggesting that
PLP prevented the dissociation of the tetrameric en-G. D. Fasman, Brandeis University, Waltham, MA, to
Dr. V. Prakash, one of the authors) which were 15 { zyme.
When the two peaks (Peaks I and II), separated by2, 65 { 3, and 20 { 2%, for the holoenzyme, apoenzyme,
and reconstituted holoenzyme. This suggested that re- gel filtration on Sephadex G-150, were subjected to sed-
imentation analysis separately, Peak I again showedmoval of PLP from the holoenzyme did not have any
significant effect on the secondary structure of the apo- dissociation into two equal peaks of S20,w values of 8.7
and 5.7 S (Fig. 6C, top), respectively, indicating thatenzyme.
However, the positive CD band of the holoenzyme dissociation was an equilibrium phenomenon. On the
other hand Peak II sedimented as a single peak of S20,wwith a maximum at 430 nm, which was lost as a result
of removal of PLP during the preparation of the apoen- value of 5.7 S (Fig. 6C, bottom). When 200 mM was
added to both peaks and then subjected to sedimenta-zyme (Fig. 5B, curve a), was regained upon the external
addition of 100 mM PLP (Fig. 5B, curve b). This was tion analysis, Peaks I and II still sedimented in a simi-
lar fashion as before with similar S20,w values (Fig. 6D,another indication of complete reconstitution of apoen-
zyme to holoenzyme upon the addition of PLP. bottom).
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

6. 368 BRAHATHEESWARAN ET AL.
FIG. 4. Thermal denaturation profiles of apo-SHMT, holo-SHMT, and reconstituted holo-SHMT determined using Gilford Response II Tm
spectrophotometer at 287 nm. About 250 ml of clear protein solution of 300 mg/ml in 50 mM potassium phosphate buffer, pH 7.2, containing
1 mM EDTA and 1 mM DTT with or without PLP prepared in the same buffer along with appropriate blanks was taken in thermal quartz
cuvettes which were equilibrated to 25ЊC in the instrument to obtain baseline. The samples were heated from 25 to 95ЊC at the rate of 1ЊC/
min. The profiles were obtained after smoothing. (A and B) 1, Apo-SHMT; 2, holo-SHMT; 3, reconstituted holo-SHMT.
To establish conditions for dissociation of the holoen- then subjected to sedimentation analysis. As shown in
Figs. 6E–6G the holoenzyme, which was a single homo-zyme, the enzyme was dialyzed against 50 mM Tris–
HCl containing 1 mM EDTA, 1 mM DTT, and 100 mM geneous peak at pH 7.2 (Fig. 6A), continued to sediment
as a homogeneous peak at pH 7.8 (Fig. 6E, bottom) andPLP at pH values of 7.2, 7.8, 8.3, 8.55, 8.8, and 9.3 and
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

7. 369SHEEP LIVER APO-SERINE HYDROXYMETHYLTRANSFERASE
enzyme has lost most of its activity. The pH-dependent
dissociation of the holoenzyme into dimeric form sug-
gests that electrostatic interactions are predominantly
responsible for holding the subunits together as a tetra-
mer, even though other interactions such as hydrogen
bonds, hydrophobic interactions, van der Waal’s forces,
and other minor interactions could play a role in sub-
unit interactions.
DISCUSSION
The present investigation has focused its attention
on the effect of removal of PLP from the holoenzyme
and readdition of PLP to the apoenzyme on the struc-
ture and function of the enzyme. Although the apoen-
zyme could be completely reactivated by the addition
of PLP, the kinetics of reactivation revealed at least
two steps in this process. Reconstitution is not always
reversible in all cases, as shown in the case of 2-amino-
3-ketobutyrate-CoA ligase of beef liver mitochondria
(22), wherein after making the apoenzyme by treating
the holoenzyme with hydroxylamine, the apoenzyme
could not be reconstituted with PLP, suggesting that
PLP may have another role in addition to its function
in catalysis. The initial fast phase with a pseudo-first-
order rate constant k of 20 { 5 s01
followed by the
relatively slower phase with a pseudo-first-order rate
constant k of 12.2 { 2.0 1 1003
s01
reflected the chemi-
cal and conformational changes, respectively, in rear-
rangement of the enzyme into an active form. The fast
rate in the reconstitution with PLP was comparable to
the rate of formation of external aldimine with L-serine
(Bhaskar, unpublished results) and OADS—12.0 { 1.0
s01
(3). However, this rate was considerably faster than
the formation of the external aldimine with D-cycloser-
ine (DCS) (23) and thiosemicarbazide (TSC) (5). The
rate constant for the slower phase calculated byFIG. 5. Effect of PLP on the CD spectra of sheep liver apo-SHMT.
stopped-flow spectrophotometry compared well withThe apo-SHMT (1.5 mg/ml far-UV CD spectra and 3.0 mg/ml visible
the rate constant of 0.63 min01
calculated from Fig. 1.CD spectra) in 50 mM potassium phosphate buffer, pH 7.2, containing
1 mM EDTA and 1 mM DTT was used to record the CD spectrum in Similar values for the slower phase of the interaction
a JASCO J-500 A spectropolarimeter equipped with a DP-501N data of apoenzyme with PLP have been reported earlier (24).
processor and Kenwood Oscilloscope CO-1530A. PLP (100 mM) was
The rate for the slower reaction also corresponded welladded to the apoenzyme solution and mixed well, and CD spectrum
with the final step in the formation of PLP–oximes aswas recorded again. The time taken for each scan was 8 min. Each
spectrum was an average of at least four different scans. (A) Far- was observed with DCS—5.4 1 1003
s01
(23), OADS—
UV CD spectra (190–250 nm); (B) visible CD spectra (350–500 nm); 2.2 1 1003
s01
(3), and TSC—17.3 1 1003
s01
(5). The
(A and B) curve a, apo-SHMT; curve b, apo-SHMT / 100 mM PLP.
two-phase interaction of PLP with the apoenzyme was
analogous to the reaction of substrate, L-serine, with
the enzyme, which had a very fast rate well within the
dead time of the instrument and a relatively slower8.3 (Fig. 6F, bottom). However, it showed a tendency
to dissociate at pH 8.55 (Fig. 6G, bottom) and was disso- rate with pseudo-first-order rate constant k of 49 { 5
1 1003
s01
at 50 mM serine.ciated into two equal peaks at pH 8.8 (Fig. 6E, top) and
dissociated completely at a pH value of 9.3 (Fig. 6G, In parallel with the regain of enzyme activity and
spectral characteristics, the results presented in Fig. 4top). When the sedimentation coefficient values were
plotted as a function of pH (Fig. 6H) they showed a showed that there was complete regain of the thermal
stability of SHMT upon the addition of PLP to the apo-clear decrease with increase in pH. Dissociation of the
holoenzyme began at a pH value of 8.55, wherein the enzyme. This indicated that the reconstituted holoen-
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives

8. 370 BRAHATHEESWARAN ET AL.
FIG. 6. Sedimentation velocity patterns of apo-SHMT, holo-SHMT, and reconstituted holo-SHMT. Sedimentation velocity experiments
were performed in a Beckman Model E analytical ultracentrifuge at 50,740 rpm and at 25 { 2ЊC. Schlieren patterns were recorded and
photographs taken after attainment of two-thirds maximum speed. The times at which photographs were taken are shown below. S20,w
values were calculated according to the method of Schachman (18). (A) Top, apo-SHMT (68 min); bottom, holo-SHMT (68 min). (B) Top,
apo-SHMT / 200 mM PLP (68 min); bottom, holo-SHMT / 200 mM PLP (68 min). (C) Top, Peak I (apo-SHMT) (58 min); bottom, Peak II
(apo-SHMT) (58 min). (D) Top, Peak I (apo-SHMT / 100 mM PLP) (58 min); bottom, Peak II (apo-SHMT / 100 mM PLP) (58 min). (E) Top,
holo-SHMT, pH 8.8; bottom, holo-SHMT, pH 7.8. (F) Holo-SHMT, pH 8.3. (G) Top, holo-SHMT, pH 9.3; bottom, holo-SHMT, pH 8.55. (H)
Plot of sedimentation coefficient (S20,w) values as a function of pH.
AID ARCH 9472 / 6b1a$$9472
05-11-96 08:14:43 arca AP: Archives

9. 371SHEEP LIVER APO-SERINE HYDROXYMETHYLTRANSFERASE
FIG.7.Gelfiltrationofapo-SHMT,holo-SHMT,andreconstitutedholo-SHMT.(A–C)SamplesofSHMTwereappliedtoacolumn(0.6190
cm)ofSephadexG-150superfineandelutedwith50mMpotassiumphosphatebuffer,pH7.2,containing1mMEDTA,1mMDTT,and100mM
NaClat5ml/h,collecting2-mlfractions.(A)1.5mlof12mg/mlholo-SHMT.(B)1.5mlof12mg/mlapo-SHMT.(C)1.3mlof13.5mg/mlapo-
SHMTwasappliedtothecolumnandelutedwith50mMpotassiumphosphatebuffer,pH7.2,containing1mMEDTA,1mMDTT,100mMPLP,
and100mMNaCl.(᭺)Absorbanceat280nm;(l)absorbanceat423nm.
AID ARCH 9472 / 6b1a$$9472 05-11-96 08:14:43 arca AP: Archives

10. 372 BRAHATHEESWARAN ET AL.
hindra Rao, and Mr. Ramesh Kumar Parigi in carrying out some ofzyme was thermally as stable as the holoenzyme. This
the experiments described in this paper is gratefully acknowledged.was confirmed by the absence of any perceptible change
in the a, b, and aperiodic structures of the protein (Fig.
REFERENCES5A). In addition to this, there was no change in the
1. Schirch, V. (1984) in Folates and Pterins, Vol. I, Chemistry andtryptophan environment of the protein upon removal
Biochemistry of Folates (Blakley, R. L., and Benkovic, S. J.,
of PLP as revealed by fluorescence measurements (data
Eds.), p. 399, Wiley, New York.
not shown). All three observations support that upon
2. Manohar, R., and Appaji Rao, N. (1987) Ind. J. Biochem. Bio-
removal of PLP from the holoenzyme any changes oc- phys. 24, 6–11.
curring in structure were probably taking place with- 3. Baskaran, N., Prakash, V., Appu Rao, A. G., Radhakrishnan,
out affecting the secondary structure as well as the A. N., Savithri, H. S., and Appaji Rao, N. (1989) Biochemistry
28, 9607–9612.overall conformation of the protein.
4. Acharya, J. K., Prakash, V., Appu Rao, A. G., Savithri, H. S.,In some instances of multimeric PLP enzymes as in
and Appaji Rao, N. (1991) Ind. J. Biochem. Biophys. 28, 381–the case of aspartate aminotransferase (8) and trypto-
388.
phan synthase (9), removal of PLP results in changes
5. Acharya, J. K., and Appaji Rao, N. (1992) J. Biol. Chem. 267,
in the oligomeric structure of the protein. In the case 19066–19071.
of SHMT, there is clear evidence of such a phenomenon 6. Schirch, V., Shostak, K., Zamora, M., and Gautam-Basak, M.
taking place. The apoenzyme sedimented as a doublet (1991) J. Biol. Chem. 266, 759–764.
(Fig. 6A, top) and addition of PLP did not alter the 7. Bhaskar, B., Prakash, V., Savithri, H. S., and Appaji Rao, N.
(1994) Biochim. Biophys. Acta 1209, 40–50.sedimentation pattern significantly (Fig. 6A, bottom).
8. Braunstein, A. E. (1973) in The Enzymes (Boyer, P. D., Ed.),The absence of complete separation of the two peaks
Vol. 9, pp. 379–480, Longmans, Green, London.implied that this dissociation could be an equilibrium
9. Yanofsky, C., and Crawford, I. P. (1972) in The Enzymes (Boyer,phenomenon. Gel filtration chromatography of the apo-
P. D., Ed.), Vol. 7, pp. 1–31, Longmans, Green, London.
enzyme on a calibrated Sephadex G-150 column con-
10. Miles, E. W., and Moriguchi, M. (1977) J. Biol. Chem. 252, 6594–
firmed the presence of more than one form of the en- 6599.
zyme, probably as tetramers and dimers, and the in- 11. Hatefi, Y., Talbert, P. T., Osborn, M. J., and Huennekens, F. M.
ability to separate these two peaks quantitatively (1959) Biochem. Prep. 7, 89–92.
suggested that they could be present in an equilibrium 12. Schirch, D., Fratte, S. D., Iuresia, S., Angellaccio, S., Contesta-
mixture. The enriched dimeric form was surprisingly bile, R., Bossa, F., and Schirch, V. (1993) J. Biol. Chem. 268,
23132–23138.inactive, indicating that only the tetrameric form could
13. Jones, C. W., and Priest, D. G. (1976) Arch. Biochem. Biophys.be reconstituted to enzyme activity. This was also con-
174, 305–311.firmed by staining for enzyme activity on nondenatur-
14. Taylor, R. T., and Weissbach, H. (1965) Anal. Biochem. 13, 80–
ing gels using b-phenylserine as substrate (data not
84.
shown). Further, the SDS–PAGE analysis of both
15. Lowry, O. H., Rosenbrough, N. J., Farr, A. L., and Randall,
forms of the apoenzyme showed a single protein band R. J. (1951) J. Biol. Chem. 193, 265–275.
corresponding to 53 kDa (data not shown), indicating 16. Hiromi, K. (1979) Kinetics of Enzyme Reactions—Theory and
that the observed peaks are not a result of protein deg- Practice, p. 189, Wiley, New York.
radation. 17. White, F. L., and Olsen, K. W. (1987) Arch. Biochem. Biophys.
258, 51–57.The results presented in this paper demonstrate that
18. Schachman, H. K. (1959) Ultracentrifugation in Biochemistry,the removal of PLP from the holoenzyme of SHMT
Academic Press, New York.leads to the formation of apoenzyme which could be
19. Greenfield, N., and Fasman, G. D. (1969) Biochemistry 8, 4108–reconstituted to regain complete enzyme activity which
4116.
is in two distinct kinetic steps. The removal of PLP
20. Wiseman, T., Williston, S., Brandts, J. F., and Lin, L-N. (1989)
while not affecting the tertiary structure of the enzyme, Anal. Biochem. 179, 131–137.
however, results in alteration of the oligomeric struc- 21. Usha, R., Savithri, H. S., and Appaji Rao, N. (1994) Biochim.
ture of the enzyme. Biophys. Acta. 1204, 75–83.
22. Tong, H., and Davis, L. (1994) J. Biol. Chem. 269, 4057–4064.
23. Manohar, R. (1983) Ph.D. thesis, Indian Institute of Science,
ACKNOWLEDGMENTS
Bangalore.
24. Hansen, J., and Davis, L. (1979) Biochim. Biophys. Acta 568,We thank the Department of Science and Technology for financial
assistance. The technical assistance of Ms. Seetha Murthy, Dr. Sud- 321–330.
AID ARCH 9472 / 6b1a$$$261 05-11-96 08:14:43 arca AP: Archives