SHMT_BBA_1994

et Biophysica Aftra ELSEVIER Biochimica et Biophysica Acta 1209 (1994) 40-50 , ,

Interactions of L-serine at the active site of serine hydroxymethyltransferases: induction of thermal stability

Brahatheeswaran Bhaskar a, V. Prakash b, Handanahal S. Savithri a, N. Appaji Rao a,. a Department of Biochemistry, Indian Institute of Science, Bangalore 560012, India

b Department of Protein Technology, Central Food Technological Research Institute, Mysore 570 013, India

Received 11 April 1994; revised 28 June 1994

Abstract

Serine hydroxymethyltransferase (SHMT), EC 2.1.2.1, exhibits broad substrate and reaction specificity. In addition to cleaving many 3-hydroxyamino acids to glycine and an aldehyde, the enzyme also catalyzed the decarboxylation, transamination and racemization of several substrate analogues of amino acids. To elucidate the mechanism of interaction of substrates, especially L-serine with the enzyme, a comparative study of interaction of L-serine with the enzyme from sheep liver and Escherichia coli, was carried out. The heat stability of both the enzymes was enhanced in the presence of serine, although to different extents. Thermal denaturation monitored by spectral changes indicated an alteration in the apparent T m of sheep liver and E. coli SHMTs from 55 + I°C to 72 + 3°C at 40 mM serine and from 67 + I°C to 72 + I°C at 20 mM serine, respectively. Using stopped flow spectrophotometry k values of (49 + 5). 10 -3 s- 1 and (69 + 7)" 10 -3 S- 1 for sheep liver and E. coli enzymes were determined at 50 mM serine. The binding of serine monitored by intrinsic fluorescence and sedimentation velocity measurements indicated that there was no generalized change in the structure of both proteins. However, visible CD measurements indicated a change in the asymmetric environment of pyridoxal 5'-phosphate at the active site upon binding of serine to both the enzymes. The formation of an external aldimine was accompanied by a change in the secondary structure of the enzymes monitored by far UV-CD spectra. Titration microcalorimetric studies in the presence of serine (8 mM) also demonstrated a single class of binding and the conformational changes accompanying the binding of serine to the enzyme resulted in a more compact structure leading to increased thermal stability of the enzyme.

Keywords: Serine hydroxymethyltransferase; Serine interaction; Thermal stability; Aldimine, internal and external

1. Introduction

Serine is a versatile amino acid with many functions. In addition to being a part of the protein structure, it is also a gluconeogenic amino acid, provides one-carbon fragments for the biosynthesis of purines and methyl group for thymidine and methionine [1]. Serine hydroxymethyltransferase, EC 2.1.2.1 (SHMT), which is a key enzyme in the pathway for interconversion of folates, has attracted

Abbreviations: Serine hydroxymethyltransferase, SHMT; pyridoxal 5'-phosphate, PLP; ethylenediaminetetraacetic acid, EDTA; dithiothreitol, DTI'; thiosemicarbazide, TSC; nicotinamide adenine dinucleotide, NAD +; 2-mercaptoethanol, 2-ME; 5,5-dimethyl-l,3-cyclohexane-dione, dimedone; 2,5-diphenyloxazole, PPO; carboxymethyl-Sephadex, CM-Sep- hadex C-50.

* Corresponding author. E-mail: bcnar@bi°chem'iisc'ernet'in' Fax: + 91 80 3341683.

0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0 1 6 7 - 4 8 3 8 ( 9 4 ) 0 0 1 3 5 - 9

widespread attention as a model pyridoxal 5'-phosphate (PLP) protein and as a possible target for cancer chemo- therapy [2]. For these reasons, several laboratories have examined different facets of the structure-function rela- tionship of the enzyme, including elucidation of the pri- mary structure [3-5], interactions at the active site [6-9] and cloning and expression of the enzyme from various sources [10-15]. Although details of the interactions of the inhibitors such as D-cycloserine [16], O-amino-D-serine [7] and methoxyamine [17] at the active site have received extensive attention, the role of substrates and substrate analogues in protecting the enzyme against thermal inactivation has not been investigated extensively, except for the tetrameric rabbit liver and the dimeric Escherichia coli enzymes. [18-20]. In this communication, we report the results of experiments aimed at understanding the mechanism by which serine stabilizes the enzyme against thermal denaturation.

B. Bhaskar et al. / Biochimica et Biophysica Acta 1209 (1994) 40-50 41

2. Experimental procedures

2.1. Materials

The following biochemicals were obtained from Sigma, St. Louis, MO, USA: 2-mercaptoethanol (2-ME), DL-dithiothreitol (DTT), ethylenediaminetetraacetic acid (EDTA), pyridoxal 5'-phosphate (PLP), L-serine, glycine, L-threonine, D-alanine, 5,5-dimethyl-l,3-cyclohexane-dione (dimedone), 2,5-diphenyloxazole (PPO), and carboxymethyl (CM)-Sephadex C-50. Sephacryl S-200 was purchased from Pharmacia, Uppasala, Sweden. Tetrahy- drofolate (H4-folate) was prepared by the method of Hatefi et al. [21]. L-[3-14C]Serine (specific radioactivity 53 mCi/mmol) was purchased from Amersham, Bucks, UK. All other chemicals were of analytical reagent grade.

2.2. Methods

Bacterial cultures. The bacterial strain used for the isolation of SHMT was GS245, a derivative of E. coli K12 and is pheA905 araD139 81acU169 8glyA strA thi. Host bacteria were transformed with plasmid pGS29 by the CaC12 procedure of Mandel and Higu [22]. pGS29 is a derivative of pBR322 and contains the E. coli glyA gene on a 3.3 kb SalI-EcoRI fragment [10]. Transformant bacteria were screened for ampicilin resistance and complemen- tation of glyA deletion. Bacterial cultures were maintained as Luria-Bertani (LB) agar stabs and glycerol cultures containing 100 /xg /ml ampicilin at 4°C. Host bacteria and plasmid DNA were kindly supplied by Dr. George Stauffer of Iowa University, Iowa City, IO, USA.

Enzyme purification. Sheep liver SHMT was purified as described by Baskaran et al. [7]. E. coli SHMT was purified as described by Schirch et al. [23] with minor modifications in the final step of purification.

Enzyme assay. The enzyme activity was determined as described by Manohar and Appaji Rao [19] using L-[3- 14C]serine as substrate [24]. Protein concentration was determined by the method of Lowry et al. [25] using bovine serum albumin (BSA) as the standard.

Absorption spectroscopy. All spectral measurements were carried out in 50 mM phosphate buffer (pH 7.2) with 1 mM EDTA and 1 mM DTT at 25 __+ I°C. Absorption spectra were recorded in a Shimadzu UV-240 Graphicord double beam spectrophotometer. Enzyme solutions were extensively dialyzed against the buffer mentioned above before spectral measurements.

Heat inactivation. Sheep liver (3.4 mg/ml) or E. coli SHMT (2 mg/ml) in 50 mM potassium phosphate buffer (pH 7.2) was kept in a thermostatically controlled water bath at different temperatures. At different time intervals (0-15 min) aliquots of 20 /zl were withdrawn and diluted to 1 ml and chilled in ice. Aliquots (20/.tl) of this solution were assayed at 37°C for residual enzyme activity after adding the remaining components of enzyme assay mix-

ture [19]. Results of heat inactivation experiments were expressed as percent activity remaining compared to the control value obtained at zero time of incubation.

Thermal denaturation studies. The thermal denaturation of sheep liver SHMT or E. coli SHMT in the absence and in the presence of the ligands L-serine, glycine, folate, D-alanine, L-threonine and thiosemicarbazide (TSC) was carried out by measuring absorbance changes at 287 nm in a Gilford Response II spectrophotometer from Ms. Ciba Coming, USA. A clear solution of 0.3 mg/ml of the protein in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTF was prepared. About 250 /zl of protein with the ligand was used with appropriate blanks in the thermal quartz cuvettes and equilibrated to 25°C in the instrument to obtain the base- line. The samples were heated from 25°C to 95°C at the rate of l °C/min using the software available with the instrument. The absorbance change in each case was monitored at 287 nm and data were averaged from three experiments. The first derivative of the denaturation profile was used to evaluate the apparent transition temperatures (T m) using the software supplied along with the instrument.

The results were analyzed according to the method suggested by White and Olsen [26] in which the fraction of protein in the denatured state (F D) is given by

A l - A N F D -- - - (1)

A D -A N

where A N is the absorbance of protein solution at 20°C, A D is the absorbance of protein solution of the plateau region (in this case 80°C and A 1 is the absorbance of protein solution at different temperatures between 20°C and 80°C. The apparent denaturation temperature (app. T m) was defined as the temperature at which the value of F o was 0.5.

Fluorescence spectroscopy. Fluorescence excitation and emission spectra were recorded in a Shimadzu RF-500 spectrofluorophotometer. All the fluorescence measurements were made using quartz cuvettes (3 ml) with 1 cm path length at 25 _ 1°C.

In the fluorescence titration experiments, the titrant was delivered in 3 /xl aliquots into the sample cuvettes. The concentrations of the protein used were 0.7/zM for sheep liver SHMT and 1.5 /xM for E. coli SHMT. The sample solution was mixed well inside the cuvette holder which had a magnetic stirrer attached. At least 5-10 min time was given for stabilization of the reading. Appropriate corrections were made for dilution of the protein sample upon addition of the ligand. The protein was excited at its excitation maximum of 285 nm and emission monitored at 338 nm.

Circular dichroism (CD) spectra. CD measurements were made in a Jasco J-500A automated recording spectropolarimeter. The spectropolarimeter was continuously

42 B. Bhaskar et al. / Biochimica et Biophysica Acta 1209 (1994) 40-50

purged with nitrogen before and dur!ng the experiments. Slits were programmed to yield 10 A bandwidth at each wavelength.

The enzyme CD-spectra were plotted as molar ellipticity values assuming a relative M r of 213 000 for sheep liver SHMT and 97 000 for E. coli SHMT, respectively with mean residue weights (mrw) of 110.24 for sheep enzyme and 116.30 for E. coli enzyme, respectively and O was calculated using the following equation [27]:

[O]mrw = [ O] × mrw/lO × l× c (2)

where O is the observed ellipticity in degrees, I is the optical path length in cm and c is the concentration of enzymes in mg/ml . All CD spectra were recorded at 22 + I°C in 50 mM phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTI" using the same buffer as blank. The protein concentrations used were 7.0 /xM and 15.0 /xM for obtaining far UV-CD spectra and 14.0 /xM and 30.0 / ,M for visible CD spectra in the case of sheep liver and E. coli enzymes, respectively.

Sedimentation velocity. Sedimentation velocity experiments were performed in a Beckman Model E analytical ultracentrifuge equipped with an RTIC unit. Experiments were carried out in a Kel F coated aluminium single sector centre piece with quartz windows, at 25 + I°C and at 59780 rpm. Schlieren patterns were recorded on Agfa Parachrome films. The S20,w values were calculated according to the standard procedure [28].

Stopped-flow spectrophotometry. The stopped-flow experiments were performed in a Union-Giken RA 401 stopped-flow spectrophotometer equipped with a 10 mm cell. The data were collected using a NEC 9801E computer interfaced to the spectrophotometer. The solutions were mixed under nitrogen pressure (4 kg/cm2). The dead time of the instrument was 40 msec and the slit width was set at 1.4 nm in all the experiments. All the reaction curves presented were the average of at least 5 sets of experiments. The k values calculated using the specific software agreed with the values calculated manually by the procedure of Hiromi [29].

Titration calorimetry. The titration of sheep liver SHMT (27.2 / ,M) with L-serine (8 mM) was carried out in a Microcal Omega Ultrasensitive calorimeter (Microcal, Northampton, MA). A window-based software package (ORIGIN version 1.1) also supplied by Microcal was used to analyze and plot the data. The reaction cell was equilibrated at 15°C using insulated constant temperature circu- lating water bath. Internal calibration was performed for cell constants and other parameters of the microcalorime- ter. L-Serine (8 mM) was taken in 250/zl injection syringe and 18 injections of 15 /,1 each at 3 min interval with a delivery time of 15 s was programmed in titrating the protein against the ligand using the instrument software. The data analysis was done according to the method of Wiseman et al. [30].

3. Results

The conformational and functional features of sheep liver and E. coli SHMTs were examined by monitoring: (a) thermal inactivation of the enzymes in the absence and presence of serine and other ligands; (b) interaction of serine with both the enzymes measured by changes in its fluorescence and circular dichroism; (c) alterations in the sedimentation coefficient of the enzymes; (d) fast reaction kinetics of the interaction of serine with the enzymes by stopped-flow spectrophotometry and (e) heat capacity changes as a result of interaction of the enzyme with serine by titration microcalorimetry.

3.1. Thermal stability of SHMT-L-serine complex

It is well known that substrates and effectors either increase or decrease stability of enzymes. Sheep liver SHMT (3.4 mg/ml) was incubated separately at 62 °, 65 ° and 67°C for different periods of time and the residual enzyme activity estimated. It can be seen from Fig. 1A that although the sheep liver enzyme was stable at 60°C, it lost 85% of its activity within 10 min when the temperature was increased to 65°C. When a similar experiment was carried out using E. coli SHMT, it was observed that increasing temperatures above 60°C inactivated the enzyme and more than 90% activity was lost at 75°C (Fig. 1C). When the sheep liver SHMT was incubated at 65°C in the presence of either, 1 or 10 mM serine, considerable protection (65% and 73%, respectively) of enzyme activity was observed during 15 min (Fig. 1B). In the case of E. coli SHMT, at 65°C only 35% activity was lost which could be prevented by the addition of serine. On the other hand at 75°C, 95% of the enzyme activity was lost in the absence of serine which could be prevented to the extent of 25% in the presence of 10 mM serine (Fig. 1D). These results suggested that serine protected both sheep liver and E. coli SHMTs against heat inactivation.

In addition to serine, SHMT interacted with a number of amino acids, nucleotides and folate derivatives [20]' It was, therefore, of interest to study the effect of these ligands on the temperature induced denaturation of the enzyme. The inactivation of enzymes in the presence of different concentrations of serine, 10 mM glycine and 10 mM NAD + is given in Table 1. It is evident from the table that increasing concentrations of serine brings about increased protection of both sheep liver and E. coli SHMTs and almost complete protection is observed at 40 mM serine (Table 1). Glycine (10 raM) protected the enzyme to the extent of 60%. Similarly, NAD + also protected the enzymes, but the protection was not as significant as that of serine. Several other ligands such as folic acid, Cibacron blue F3GA, L-threonine, D-alanine and thiosemicarbazide (TSC) even at high concentrations failed to protect either of the enzymes significantly (data not shown). As serine


maximally protected the enzymes against heat inactivation

detailed investigations were carried out with this ligand.

3.2. Thermal denaturation o f S H M T

Fig. 2 (A and B) show the effects of temperature on absorbance changes of the enzymes at 287 nm. In Fig. 2A and B (insets) representative first derivative plots of sheep liver and E. coli SHMT, respectively, are shown. In the case of sheep liver SHMT, in buffer alone the apparent T m was 55 + I°C. Upon the addition of 10 mM serine, a shift

in the transition curve was observed. An apparent T m of 69 + 2°C at 10 mM serine and an apparent T m of 72 + 3°C at 40 mM serine suggested that the enzyme was stabilized by serine against thermal denaturation.

I0C

6o

0 0 0 60

5 10 15 TIME (rnln)

~ I00

2C [ ] .

t I L l I i Ii 5 10 15

TIME (min)

IO0

g 20

0 0 0 6(

,,. 5 10 15

TI ME (rnln)

6s'c

- - 1 0

0 5 I0 15 TIME (rain)

Fig. 1. Heat inactivation of SHMT in the absence and presence of L-serine. (A) Sheep liver SHMT (3.4 mg/ml) in 200 /xl of 50 mM potassium phosphate buffer (pH 7.2) was incubated at 60, 62, 65 and 67°C, respectively. Aliquots (20 /zl) were withdrawn at time intervals indicated in the figure and rapidly chilled in ice. At the end of time periods indicated, aliquots were assayed for the residual enzyme activity as described in Section 2 [19,24]. (B) Sheep liver SHMT (3.4 mg/ml) in 200/xl of 50 mM potassium phosphate buffer (pH 7.2) was incubated at 65°C in the presence of 1 mM or 10 mM serine. Aliquots were withdrawn at regular intervals as indicated in the figure and assayed for residual enzyme activity [19]. (C) In a separate experiment E. coli enzyme (2 mg/ml) in 200 /xl of 50 mM potassium phosphate buffer (pH 7,2) was incubated at 60, 65, 70 and 75°C, respectively. Aliquots (20 /zl) were withdrawn at time intervals indicated in the figure and assayed for residual activity [19]. (D) E. coli SHMT (2 mg/ml) in 200 #1 of 50 mM potassium phosphate buffer (pH 7.2) was incubated at 65 and 75°C in the presence of 10 mM serine, Aliquots (20 /zl) were withdrawn at time intervals indicated in the figure and assayed for the residual activity [19]. Results in all the above experiments are expressed as percent activity remaining over the control.

Table 1 Heat inactivation of (SHMT) in presence of different ligands

Ligand Percent activity remaining a

sheep liver E. coli (65°C) (70°C)

Enzyme alone 16 36 Enzyme + 0.5 mM serine 50 46 Enzyme + 1.0 mM serine 46 60 Enzyme + 5.0 mM serine 78 70 Enzyme + 40.0 mM serine 82 78

Enzyme alone 15 33 Enzyme + 10.0 mM glycine 60 68

Enzyme alone 25 38 Enzyme + 10.0 mM NAD + 55 45

a These are independent measurements of activity at the end of 15 minat the specified temperature and ligand concentration. Sheep liver SHMT (4 mg/ml) in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTI r was incubated at 65°C in the absence and in the presence of 0.5 mM, 1 mM, 5 mM, and 40 mM L-serine, 10 mM glycine or 10 mM NAD +, respectively. After 15 min incubation the residual enzyme activity was determined. A similar experiment was carried out with the E.coli enzyme (2 mg/ml) but inactivation was carried out at 70°C both in the presence and absence of the ligands mentioned above.

In the case of E. coli SHMT, the native enzyme in buffer had an apparent T m of 67 _+ 1°C, which is 12°C

higher compared to the apparent T m for the sheep liver enzyme. However, the addition of serine increased the apparent T~ for the E. coli enzyme by 5 + 1°C (from 67 to 72°C, see Fig. 2B). A comparison of denaturation curves both in convoluted and deconvoluted states showed a well defined hump in the control enzymes, which became more pronounced with increasing serine concentration. How- ever, the reason for the hump is not very clear.

An analysis of data on the denaturation of the enzymes in the presence of different concentrations of serine and

glycine and at a single concentration of o-alanine, L- threonine and TSC (Fig. 3A and B) showed that compared to serine, glycine was a poor protector of both sheep liver and E. coli enzymes. L-Threonine, o-alanine and TSC

which interacted at the active site of the enzyme like serine or glycine, did not increase the thermal stability of the enzymes. On the other hand, the two enzymes in the presence of these ligands were more susceptible to heat denaturation. A common feature of interaction of all these ligands including serine was that they interacted with PLP at the active site and generated characteristic intermediates. In spite of this commonality, the external aldimine formed with serine and to a lesser extent with glycine appeared to induce changes in the enzyme structure leading to the formation of a more stable enzyme, whereas the formation of a similar complex with other ligands led to the genera- tion of a less stable structure indicating that the equilib- rium between different structures of enzymes was affected by the presence of ligands at the active site. From Fig. 3, it is evident that in the concentration range of 1 to 40 mM,


the apparent T m increased linearly in the presence of L-serine. On the other hand, in the presence of 1 to 10 mM folic acid the apparent T m decreased sharply (Fig. 3B) suggesting destabilization of the sheep liver enzyme. Higher concentrations of folic acid (above 10 mM) could not be used due to its limited solubility at pH 7.2.

3.3. Effect of serine on fluorescence spectra of SHMT

The results thus far described suggested that binding of serine probably brought about subtle changes in the structure of the enzymes. It was, therefore of interest to determine whether conformational changes were a prerequisite to binding of serine to the enzyme. Intrinsic fluorescence changes provided a convenient handle to examine this question. Sheep liver enzyme contained 1 tryptophan residue per subunit amounting to 4 residues per mole of

2.C )3 -.- , _ ._ . . . . j . ~

6~=C 1 ~ '~

7I ,:, tA\ I / / / , V ,: t J / t k I ,' 7 I . : • i

w : - " /

<1.o . ~ 1 i i , i ~1/ / /

O ! i I I I 25 36 47 58 69 80

TEMPERATURE(°C) 2.0

67°C l /'/X~. 3 , /

/ t~ : - - : I il 2

? J S ~3

25 36 47 58 69 80 ./ j / -

-o.2 . . . . i 25 36 47 58 69 80

"TEMPERATURE (°C)

Fig. 2. Representative thermal denaturation profiles of SHMT determined using Gilford Response II T m spectrophotometer at 287 nm. About 250 /zl of a clear solution of 0.3 A280 nm per ml of the protein in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTT with or without serine prepared in the same buffer along with appropriate blanks were taken in thermal quartz cuvettes which were equilibrated to 25°C in the instrument. The samples after equilibration were heated from 25 ° to 95°C at the rate of I°C per rain. The profiles were obtained after smoothening. (A) Sheep liver SHMT: 1, enzyme alone; 2, enzyme + 10 mM serine; and 3, enzyme + 40 mM serine. (B) E. coli SHMT: 1, enzyme alone; 2, enzyme + 5 mM serine, 3; and enzyme + 20 mM serine. Insets of (A) and (B) show first-derivative plots analyzed to evaluate transition temperatures (app. T m).

0

~ - 2

- 4

r . - , e , . . . . . . . • L Cb) - t ' - O

1 ~ I I I -1.o o 1 .o

LOG [ 5ERINE ] , m M

A

B 0 - -

¢...1

I i I r ~ I 0 0 .5 1.0 1.5 2 . 0

LOG [ LIGAND ] , mM

Fig. 3. A logar i thmic plot o f apparent midpoint o f transit ion o f thermal

denaturation vs. ligand concentration. Apparent T m values obtained from first-derivative analyses after heat denaturing the enzyme in the presence of different concentrations of ligands were plotted against the concentration of ligands. (A) in the presence of L-serine. (B) In the presence of other ligands. T~ = app. T m set to maximum at the highest ligand concentration. T o = app. T m at specified concentration of the ligand. (a) Sheep liver SHMT + L-serine (0.1-40 mM), (b) E. coli SHMT + L-serine (0.5- 20 mM), (c) Sheep liver SHMT+ folic acid (1-10 mM), (d) Sheep liver SHMT+glycine (1-40 mM), and (e) E. coli SHMT+glycine (1-32 mM).

the tetramer [5], whereas the E. coli enzyme has 3 tryptophan residues per subunit amounting to 6 residues per mole of the dimer [10].

Fig. 4 (Inset) shows fluorescence emission spectra of both sheep liver and E. coli SHMT at different concentrations of serine. Upon progressive addition of L-serine significant quenching was observed in both the cases. However, there was no marked difference in the extent of quenching with increasing concentrations of serine. The data were analyzed by Lehrer and Fasman's method [31] for the binding parameters (Fig. 4A and B). The K a values were 26.7 + 3.5 M -1 and 97.6 -t- 10 M -1 and AG values were - 1 . 8 8 + 0.15 kca l /mol and - 2 . 6 9 ± 0.11 kcal/mol, respectively for the sheep liver and E. coli enzymes. These values suggest that the interactions of L-serine with the enzyme may not have specifically altered the environment around the tryptophan residues during the formation of premediated or postmediated complex. How- ever, the energy transfer between tryptophan residues and bound pyridoxal phosphate cofactor cannot be the cause since the apoenzyme also showed similar quenching phenomenon upon titrating with increasing concentrations of serine (Bhaskar, unpublished data). This does not exclude conformational changes in regions devoid of tryptophan.

B. Bhaskar et al. / Biochimica et Biophysica Acta 1209 (1994) 40-50 4 5

3.4. Effect of serine on the velocity sedimentation of SHMT A B

Any conformational change upon binding of serine in different domains of the enzyme can alter the shape resulting in increase or decrease of the frictional coefficient of the enzyme. Results of velocity sedimentation of both sheep liver and E. coli enzymes in the absence and presence of serine is shown in Fig. 5. Sheep liver SHMT had an S20,w value of 8.0 + 0.2, whereas in the presence of 50 mM serine it decreased to 7 . 8 _ 0.2. These results suggested that upon binding of serine, the enzyme molecule had an increased frictional coefficient thereby decreasing

the S20,w value of the protein. On the other hand, the S20,w value of E. coli enzyme which was 5 . 0 _ 0.15 did not change upon binding of 50 m M serine. These changes in the sedimentation coefficients are not very significant due to error bars involved in the calculation and measurements.

Fig. 5. Sedimentation velocity patterns of SHMT in the absence and presence of L-serine. The photographs were taken after reaching two thirds maximum speed. The time at which photographs were taken are shown against each frame. (A) Sheep liver SHMT: upper trace, enzyme + 50 mM serine (71 min); lower trace, native enzyme (71 min). (B) E. coli SHMT: upper trace, native enzyme (49 min); lower trace, enzyme + 50 mM serine (49 min).

0 . 4

m. 0.2

oo z u

25

0 10

1

o - - - - . L _ 2 _ . _ L _ 300 350 4C0

2 0 3 0 5 0 FSERINE] f tee

I /oOO 1.5 B

• o/ue

0 . 5 7/" /.

I 0 10

~o

d a:

0 J I i 300 350 400

WAVELENGTH (rim)

I I j ~ t 20 30 50

[SERINE'] f r e e

Fig. 4. The effect of serine on the fluorescence emission spectrum of sheep liver and E. coli SHMT. The enzyme (0.15 A280/ml) in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTT was excited at 285 nm and fluorescence emission spectrum was recorded in a Shimadzu RF-500 spectrofluorometer attached with a recorder. To the enzyme 3 /xl aliquots of 500 mM L-serine were added, mixed well inside the chamber which had a magnetic stirrer attached to it and the emission spectrum was recorded in the range 300-400 nm and emission maximum at 338 nm was recorded. The data was then analyzed by Lehrer and Fasman method [31]. The graph shown in this figure gives a plot of / 3 / 1 - f l vs. [serine]fre¢, where /3 is the fractional change occuring in fluorescence upon addition of L-serine. (A) Sheep liver SHMT, and (B) E. coli SHMT. Insets of (A) and (B): fluorescence emission spectrum of sheep liver and E. coli SHMT recorded after the addition of each aliquot of L-serine (3 Ixl of 500 mM L-serine).

3.5. Effect of serine on the CD spectra of SHMT

Earlier results on the interaction of serine with the enzyme showed that it produced an external aldimine at the active site of the enzyme [1]. The spectral properties of PLP-Schiff ' s base provided a convenient probe to monitor the changes at the active site of the enzyme. It can be seen from Fig. 6A that the addition of serine caused a very significant decrease in molar ellipticity of PLP at the active site of the enzyme. Similar changes were also seen in the case of E. coli enzyme (Fig. 6B); however, the extent of change was less as compared to the sheep liver enzyme. These results indicated that the decreased positive band at 430 nm was due to the alteration in the orientation of PLP with respect to the neighbouring groups as a result of the interaction with L-serine. These results also indicated the extent of conformational change at the active site of the enzyme.

In Fig. 7 the far UV-CD spectra of the enzymes from sheep liver and E. coli in the absence and presence of 50 mM L-serine is shown. The data was analyzed by the CD Estima method of Fasman (program courtesy of Prof. G.D. Fasman, Brandeis University, Waltham, MA, USA). In the case of sheep liver enzyme upon the addition of serine, the /3-pleated sheet decreased by 50% of its value in the native enzyme with a concomitant increase in /3-turns. However, in the case of E. coli enzyme the decrease in the/3-pleated sheet was much more drastic from 25% to 6% which was accompanied by an increase in fl-turns. However, the other secondary structural parameters of c t -he l ix and aperiodic components were not significantly altered by the addition of L-serine in both cases. Hence it was apparent that the regions of fl-pleated structure were altered in both the enzymes upon the addition of L-serine. Conformational change was thermodynamical ly stabilized by the binding of serine to the enzyme irrespective of its source, (either sheep liver or E. coli) suggesting similar mechanisms might be responsible for the change.


3.6. Spectroscopic changes upon the addition of L-serine to SHMT

(A) Visible absorption spectroscopy. Fig. 8 shows the visible absorption spectra of SHMT from sheep liver as well as E. coli (Insets IA and IIA) with the absorption maximum at 425 nm due to lysine-PLP-Schiff's base. An increase in absorption at 425 nm was observed as a function of serine concentration with a concomitant decrease at 343 nm and no change was observed at 280 nm.

(B) Stopped-flow spectrophotometry. The visible spectroscopy studies dearly suggested an increase in absorption at 425 nm occurred duc to the interaction of serine with PLP bound to the enzyme. Serine-PLP-Schiff's base had an absorption maximum at 425 nm like the lysine- PLP-Schiff's base in both enzymes (Fig. 8). It was observed in the visible spectrum that within one min after the addition of L-serine at 50 mM concentration there was already a large increase in the absorbance at 425 urn. This raised the question of existence of a rapid reaction compo- nent in the protein-ligand complex formation which could

4C

-o E

2o

o

A ENZYME

rnM SER

I i I i I I I 3 5 0 4 0 0 4 5 0 5 0 0

WAVELENGTH (nm)

25.0 - B ENZYME

~15.0 e,,E $ER

u oo

-o 5.0 !

0

-5.0 - - I I [ I [ I [ 350 400 ,,50 500'

WAVELENGTH (rim)

Fig. 6. The effect of L-serine on the visible CD-spectrum of sheep liver and E. coli SHMT. The enzyme (3 m g / m l ) in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM D'IT was used to record the CI) spectrum in the range 350 to 500 nm in a Jasco J-500A spectropolarimeter equipped with a DP-501N data processor and Kenwood Oscilloscope CO-1530A. To the enzyme solution L-serine (50 /zL of 1000 mM) was added, mixed well and CD spectrum was recorded again. The time taken for each scan was 8 min. Each spectrum given in the figure is an average of 4 scans. (A) Sheep liver SHMT, and (B) E. coli SHMT.

0 - -

% u

~ - 4 . 0

~-~-8.0

I 200

\v(- ~z+ so mM S E R / - ~, /(

LENZYME I i ]

220 240 260 WAVELENGTH (rim)

T "6 E 0

o J" E , j

L ~'- 4 .0 1D

% x

-8 .0

\

\

\ ENZ+ 50mM SER /,/~

~EHZYME

I I ] 200 220 240

W A V E L E N G T H ( n m )

Fig. 7. The effect of L-serine on the far UV-CD spectrum of sheep liver and E. coli SHMT. The enzyme (1.5 m g / m l ) 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 the range 193 to 250 nm in a Jasco J-500 A spectropolarimeter equipped with a DP-501N data processor and Kenwood Oscilloscope CO-1530A. To the enzyme solution L-serine (50 p.! of 1000 mM) was added, mixed well and the CD spectrum was recorded again. The time taken for each scan was 8 min. Each spectrum given in the figure is an average of 4 independent scans. (A) Sheep liver SHMT, and (B) E. coli SHMT.

be best studied by stopped-flow spectrophotometry. In order to analyze the kinetics of the fast reaction, the change in absorption of the enzyme-serine complex, the reaction was monitored at 425 nm under pseudo-first order conditions by adding a large excess of L-serine. Represen- tative curves are shown in Fig. 8 insets IB and IIB. Fig. 8 insets IC and IIC show derivative plots from which rate constants were calculated. A comparison of k values of the two enzymes at 50 mM serine concentration gave a value of (49 + 5) . 10 -3 s-1 for sheep enzyme compared to the value o f ( 6 9 + 7)- 10 -3 s -1 for the E. coli enzyme. These results showed that kinetics of interaction of sheep enzyme with serine was slower compared to E. coli enzyme.

At lower concentrations of L-serine, namely at 1 mM the k values were ( 1 9 + 2 ) - 1 0 -3 s-1 and ( 1 6 . 5 + 1 ) . 10 -3 s - l ; and at 10 mM the values were (33.5 + 4)- 10 -3 s -1 and ( 2 7 + 3 ) . 10 -3 s - ] for sheep liver and E. coli


enzymes, respectively. The lower k value for the sheep enzyme was probably due to the difference in accessibility of serine to PLP between the two enzymes. The nature of reaction being fast implied that the reaction was probably electronic in nature as compared to the slow reactions in many other systems due to long range factors and Van der Waals' forces stabilizing such interactions [32,33].

3. 7. Titration calorimetry.

Fig. 9 shows a calorimetric titration of 2.2 ml (5.8 mg/ml) of sheep liver SHMT with a ligand solution of 8

0.100

0.099 E t,.-

~ o.o98 Z

II1 ¢Y

~ o.o97

0.096

0.102

7 <

3 5 0 4 5 0 5 5 0 5 10 15 20 WAVELENGTH (nm) TIME (see)

I ~ I ~ I ~ I 20 60 100 140

TIME (s¢c)

I1 {s)

o.loo

¢J Z <( m 0.098 "" "~0 - ,,° \ -.;,

. . . . . .I J,, . . . . I *-_,--: X . 0.096 0.0625 ___j _ , , , , , . . , . , . ,=, . . i t .

350 450 550 5 10 15 20 0 . 0 9 4 WAVEL GTH (rim) TIME(Jet)

20 60 100 140 180 T I M E ( s e e )

Fig. 8. The effect of L-serine on the absorption properties of sheep liver and E. coli SHMT. Insets IA and IIA show the visible absorption spectra: the sheep liver SHMT (0.5 m g / m l ) was taken in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTI" and its spectrum was recorded in the range 200 to 550 nm in a Shimadzu UV-visible recording spectrophotometer UV-240 Graphicord. To the enzyme serine to a final concentration of 50 mM was added, mixed well and again the spectrum was recorded. Insets IB and IIB depict the rapid reaction kinetics. The enzyme (0.3 m g / m l ) in 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM DTI" was taken in reservoir A and L-serine (100 raM) prepared in the same buffer in reservoir B of the Union-Giken RA 401 stopped-flow spectrophotometer. The solutions were mixed and the absorbance change was recorded at 425 nm. Insets IC and IIC show the first order plots were constructed for the absorbance changes measured at 425 nm. The tracings shown in the figure are an average of 5 different experiments.

- 2 0 o~

- 4 0

1 o

I I 4 8 12 INJECTION NUMBER

I I 16 20

Fig. 9. Titration calorimetry of binding of L-sefine to SHMT. Plot of processed data in the derivative format obtained for 18 automatic injections, each of 15 /xl of 8 mM I,- serine into the sample cell containing sheep liver SHMT solution at a concentration of of 0.0272 raM. Other conditions used were 28°C 50 mM potassium phosphate buffer (pH 7.2) containing 1 mM EDTA and 1 mM D'Iq'. The total duration of the experiment was 74 rain.

mM L-serine in the same buffer. Using the instrument software, 18 injections of 15 /xl each of 8 mM serine at 3 rain interval with a delivery time of 15 s was programmed in titrating the protein against the ligand. Immediately following the injection of serine an initial exothermic phase was observed which increased upon each successive addition of serine reaching a plateau after 15 injections. The energy value at the plateau region was 25 + 3 kcal /mol of serine. An analysis of data using the o~6[N software revealed a single class of binding. Our earlier data indicated that conformational changes induced in the enzyme by serine was occurring due to its interaction at the active site (Figs. 6 and 8).

4. D i s c u s s i o n

The stability of an enzyme and kinetics of the reaction, especially its specificity is profoundly affected by the presence of ligands. Serine, one of the substrates of SHMT protects the enzyme from various sources against heat inactivation [18-20,34]. The results (Figs. 1-3 and 6 and Table 1) described in this paper showed that the interaction of serine with the enzyme enhanced the stability of the enzyme due to the formation of external aldimine with PLP at the active site of the enzyme. Earlier work with aspartate aminotransferase and SHMT from rabbit liver and E. coli [20,35] suggested that the reaction specificity was profoundly regulated by the presence of substrates. In addition to the transfer of hydroxymethyl group of serine


to form 5,10-methylenetetrahydrofolate, a physiologically important reaction, the enzyme also catalyzed several other reactions such as decarboxylation and transamination with amino acid analogues [1,36,37]. Although E. coli and sheep liver enzymes have different subunit structures they react similarly in the presence of serine. In the present investigation changes in absorbance values as a function of temperature were used to determine the denaturation status of the SHMTs in the absence and presence of ligands. This method was earlier used in the case of lysozyme [38].

A second method of monitoring ligand binding is by measuring catalytic activity of the enzyme and also monitoring alterations in a reporter group present at the active site. In this study, unlike in the earlier work of Schirch et al. [20], changes in stability of the enzyme consequent to binding of serine was monitored by activity measurements. Activity is the most sensitive parameter to measure the integrity of catalytic centre of enzymes. It is evident from data presented (Fig. 1 and Table 1) that serine protected both sheep liver and E. coli enzymes very significantly but to different extents. In fact, upon heating the sheep liver enzyme in the absence of serine for 10 min at 65°C almost complete loss of activity was observed, however, in the presence of serine, nearly 80% of the activity was present (Fig. 1B). Similarly in the case of E. coli enzyme serine protected the enzyme (Fig. 1D), whereas glycine - - the other substrate - - was partly effective (Table 1). On the other hand, other ligands which were shown earlier to bind to the enzyme, were least effective in protecting the enzyme against heat inactivation (data not shown). In general, E. coli enzyme was found to be more stable than the sheep liver enzyme as evident from the apparent T m values of 67 _+ l°C and 55 +_ I°C, respectively (Fig. 2). Addition of serine to a final concentration of 40 mM led to a large increase (from 55 _+ 1°C to 72 _+ 3°C) in the apparent T m value of sheep liver enzyme. The increase was not as large in the case of E. coli enzyme (from 67 _+ 1°C to 72 _+ I°C) (Fig. 3).

Does this mean that thermal denaturation of the enzyme was initiated at a single point and the protein cooperatively denatured rapidly? It could be postulated that a conformational change upon ligand binding resulted in this site of interaction becoming inaccessible and inactivation started from a different site leading to increased or decreased stability. When substrates were used as ligands it was necessary to distinguish between binding of substrates at the active site and possibility of binding elsewhere and causing a conformational change resulting in protection against heat inactivation. These two possibilities could be examined by probing changes at the active site and/or monitoring gross conformational changes.

The conformational changes in the protein were monitored by using intrinsic tryptophanyl fluorescence as a probe (Fig. 4). It can be seen that the fluorescence emission spectra of the enzyme in the presence and absence of serine for both sheep liver and E. coli (Fig. 4) were

similar suggesting that conformational change induced by serine did not involve changes in the microenvironment around the tryptophanyl residue. This was also reflected in the magnitude of K a values determined for both the enzymes in the presence of serine. The conformational analyses of the enzymes in the far UV-CD spectra indicated a decrease in /3-sheet and an increase in 0-turns upon binding of serine for both the enzymes. All these studies clearly indicated that the binding of serine to the enzyme leading to increased stability was accompanied by a generalized change in structure of the protein.

In order to pinpoint the origin of structural changes which could be occurring at the active site, interactions with PLP-Schiff's base were monitored by a variety of methods. The absorbance in the region at 425 nm was increased when serine was added to the enzyme (Fig. 8, IA and IIA). This change was attributed to the formation of an external aldimine of serine with the enzyme [1,39]. This absorbance change was accompanied by a slight shift in absorption maximum to around 430 nm (Fig. 8). It was suggested [1,39]) that there was initial formation of a geminal diamine from an internal aldimine which was subsequently converted to an external aldimine. The kinetics of these steps were monitored by stopped flow spectrophotometry. The initial fast phase of the reaction leading to the formation of geminal diamine was accompanied by a decrease in absorbance at 422 nm with a concomitant increase at 343 nm [40]. However, this reaction was very rapid in the case of sheep liver enzyme and could be observed only at 8°C when stoichiometric amounts of the enzyme and ligand were used. The very rapid rate of the reaction which occurred very close to the dead time of the instrument made it difficult to evaluate the kinetic constants accurately for this phase of the reaction. It was evident from Fig. 8 that there was a reasonably rapid change in absorbance at 425 nm upon the addition of serine to either of the enzymes. The reaction reached a near stationary phase within 20 s. The rate constants of the reaction for both enzymes were calculated and it was found that the rate constants increased with increasing concentrations of serine. Qualitatively this increase was similar to the increased protection against heat inactivation afforded by high concentrations of serine (Figs. 1 and 2 and Table 1). The reaction was relatively rapid with E. coli enzyme compared to the sheep liver enzyme (Fig. 8, IIB) indicating that the constraints in the structure of the sheep liver enzyme slowed down the rate of reaction of this enzyme with serine. These results suggested that the rapid kinetic step of formation of external aldimine of serine with PLP at the active site of the enzyme was a prerequisite for the final conformational change of the enzyme molecule upon binding of serine.

Yet another specific method for monitoring changes at the active site of SHMTs was the visible CD spectrum of the enzymes. The structure of the active site causes an asymmetric orientation of PLP and the external aldimine


formed in the presence of serine is still bound to the active site residues and this property provides a convenient tool for measuring changes in the PLP environment. It can be seen from Fig. 6A that serine caused a large change in visible CD spectrum of the sheep liver enzyme. Similar change was also seen with the E. cold enzyme (Fig. 6B). The titration microcalorimetry experiments presented in this paper show that binding of serine to SHMT resulted in an exothermic reaction with a single class of binding leading to conformational changes that induced stability against thermal denaturation.

Brandts [41] showed that negative heat capacity changes were associated with burial of hydrophobic amino acids. The data obtained in Fig. 9 were obtained at a single protein concentration and at one temperature (22°C). In order to arrive at the various thermodynamic parameters experiments have to be conducted at different protein concentrations and different temperatures. The limited availability of the enzyme has precluded these experiments. In SHMT it was clearly shown that the enzyme molecule had an altered conformation upon binding of serine and also had a compact structure indicated by sedimentation velocity experiments. The noninvolvement of the lone tryptophan residue in the protein was con- firmed by fluorescence spectroscopy and the heat capacity changes that have arisen must be derived from other hydrophobic amino acids other than tryptophan. These results comprehensively showed that conformational changes accompanying binding of serine led to the formation of a more compact structure. This could explain the exothermic phase in the two kinetic pathways detected using rapid kinetic measurements.

The existence of SHMT in 'open' and 'closed' forms was postulated by Schirch et al. [20] based on their studies with rabbit liver enzyme using differential scanning calorimetry. Similar 'open' and 'closed' structures were proposed in case of aspartate aminotransferase [35], tryptophan synthase [42] and triosephosphate isomerase [43]. Our results from thermal stability parameters, rapid kinetic measurements, spectroscopic studies and activity measurements of both sheep liver and E. cold enzymes clearly demonstrate that this phenomenon is more widespread and is a common feature of SHMT from several sources and may be a common phenomenon in case of several PLP enzymes.

Acknowledgments

We thank the Department of Science and Technology for the financial assistance. The technical assistance of Ms. Seetha Murthy, Mr. Ramesh Kumar, Mr. Sudhindra Rao, Mr. Muralidhar and Mr. Srinivasulu in carrying out some of the experiments described in this paper is gratefully acknowledged.

This work was supported by the Department of Science and Technology, Government of India, New Delhi, India.

References

[1] Schirch, L. (1984) in Folates and Pterins Vol. 1, Chemistry and Biochemistry of Folates (Blakley, R.L. and Benkovic, S.J., eds.), 399 pp., Wiley-lnterscience, New York.

[2] Appaji Rao, N. (1991) New Trends in Biological Chemistry (Ozawa, T., ed.) 333, Japan Scientific Press, Tokyo.

[3] Martini, F., Angellacio, S., Pascarella, S., Barra, D., Bossa, F. and Schirch, L. (1987) J. Biol. Chem. 262, 5499-5509.

[4] Martini, F., Maras, B., Tanci, P., Angellacio, S., Pascarella, S., Barra, D., Bossa, F. and Schirch, V. (1989) J. Biol. Chem. 264, 8509-8519.

[5] Usha, R., Savithri, H.S. and Appaji Rao, N. (1994) Biochem. Biophys. Acta. 1204, 75-83.

[6] Manohar, R. and Appaji Rao, N. (1984) Biochem. J. 224, 703-707. [7] Baskaran, N., Prakash, V., Appu Rao, A.G., Radhakrishnan, A.N.,

Savithri, H.S. and Appaji Rao, N. (1989) Biochemistry, 28, 9607- 9612.

[8] Acharya, J.K. and Appaji Rao, N. (1992) J. Biol. Chem. 267, 19066-19071.

[9] Usha, R., Savithri, H.S. and Appaji Rao, N. (1992) Ind. J. Biochem. Biophys. 29, 183-188.

[10] Plamann, M., Stauffer, L.T., Urbanowski, M.L. and Stauffer, G.V., (1983) Nucleic Acids Res. 11, 2065-2075.

[11] Turner, S.R., Ireland, R., Morgan, C. and Rawsthorne, S. (1992) J. Biol. Chem. 267, 13528-13534.

[12] Byrne, P.C., Sanders, P.G. and Snell, K. (1992) Biochem. J. 286, 117-123.

[13] McClung, C.R., Davis, C.R., Page, K.M. and Denome, S.A. (1992) Mol. Cell. Biol. 12, 1412-1421.

[14] Garrow, T.A., Brenner, A.A., Whitehead, V.M., Xiao-Ning Chen, Duncan, R.G., Korenberg, J.R. and Shane, B. (1993) J. Biol. Chem. 268, 11910-11916.

[15] Miyata, A., Yoshida, T., Yamaguchi, C., Tanabe, T., Toh, H., Mitsunaga, T. and Izumi, Y. (1993) Eur. J. Biochem. 212, 745-750.

[16] Manohar, R., Appu Rao, A.G. and Appaji Rao, N. (1984) Biochem- istry, 23, 4116-4122.

[17] Acharya, J.K., Prakash, V., Appu Rao, A.G., Savithri, H.S. and Appaji Rao, N. (1991) Ind. J. Biochem. Biophys. 28, 381- 388.

[18] Schirch, V., Schirch, D., Martini, F. and Bossa, F. (1986) Eur. J. Biochem. 161, 45-49.

[19] Manohar, R. and Appaji Rao, N. (1987) Ind. J. Biochem. Biophys. 24, 6-11.

[20] Schirch, V., Shostak, K., Zamora, M. and Gantam-Basak, M. (1991) J. Biol. Chem. 266, 759-764.

[21] Hatefi, Y., Talbert, P.T., Osborn, M.J. and Huennekens, F.M. (1959) Biochem. Prep. 7, 89-92.

[22] Mandel, M. and Higa, A. (1970) J. Mol. Biol. 53, 154-157. [23] Schirch, V., Hopkins, S., Villar, E. and Angellacio, S. (1985) J.

Bacteriol. 163, 1-7. [24] Taylor, R.T. and Weissbach, H. (1965) Anal. Biochem. 13, 80-84. [25] Lowry, O.H., Rosenbrough, N.J., Farr, A.L. and Randall, R.J.

(1951) J. Biol. Chem. 193, 265-275. [26] White, F.L. and Olsen, K.W. (1987) Arch. Biochem. Biophys. 258,

51-57. [27] Greenfield, N. and Fasman, G.D. (1969) Biochemistry, 8, 4108-

4116. [28] Schachman, H.K. (1959) Ultracentrifugation in Biochemistry, Aca-

demic Press, New York. [29] Hiromi, K. (1979) Kinetics of Enzyme Reactions--Theory and

Practice, 189 pp., Halsted Press/Wiley, New York. [30] Wiseman, T., Williston, S., Brandts, J.F. and Lung-Nan Lin (1989)

Anal. Biochem. 179, 131-137.


[31] Lehrer, S.S. and Fasman, G.D. (1966) Biochem. Biophys. Res. Commun. 23, 133-138.

[32] Jencks, W.P. and Gilchrist, M. (1968) J. Am. Chem. Soc. 90, 2622-2637.

[33] Jencks, W.P. (1969) Catalysis in Chemistry and Enzymology, Mc- Graw-Hill, NY, USA.

[34] Hopkins, S. and Schirch, V. (1986) J. Biol. Chem. 261, 3363-3369. [35] Relimpo, A., Iriate, A., Clalebowski, J.F. and Martinez-Carrion, M.

(1981) J. Biol. Chem. 256, 4478-4488. [36] Palekar, A.G., Tate, S.S. and Meister, A. (1973) J. Biol. Chem. 248,

1158-1167.

[37] Shostak, K. and Schirch, L. (1988) Biochemistry 27, 8007-8014. [38] Arakawa, T. and Timasheff, S.N. (1984) J. Biol. Chem. 259, 4979-

4986. [39] Schirch, L. and Mason, M. (1963) J. Biol. Chem. 238, 1032-1037. [40] Angellacio, S., Pascarella, S., Fattori, E., Bossa, F., Strong, W. and

Schirch, V. (1992) Biochemistry 31, 155-162. [41] Brandts, J.F. (1964) J. Am. Chem. Soc. 86, 4302-4311. [42] Xiang-Jiao Yang and Miles, E.W. (1992) J. Biol. Chem. 267,

7520-7528. [43] Wierenga, R.K., Noble, M.E., Postma, J.P.M., Graendyjk, H., Kalk,

K.H., Hol, W.G.J. and Opperdoes, F.R. (1991) Proteins 10, 33-49.

SHMT_BBA_1994

Documents

Transcript of SHMT_BBA_1994