34 Biochimica et Biophysica Acta, 957 (1988) 34-46 Elsevier

BBA 33215

Inactivation of yeast hexokinase by o-phthalaldehyde: evidence for the presence of a cysteine and a iysine at or near the active site

Rajinder N. Purl *, Deepak Bhatnagar and Robert Roskoski, Jr. Department of Biochemistry and Molecular Biology, Louisiana State University Medical Center, New Orleans, LA (U.S.A.)

(Received 28 March 1988)

Key words: Yeast hexokinase; o-Phthalaldehyde: Fluorescent labeling: Inactivation

Yeast bexokinase (ATP:D-hexose 6-phosphotransferase, EC 2.7.1.1), a homodimer, was rapidly and irreversibly inactivated by o-phthalaldehyde at 25°C (pH 7.3). The reaction foiloweci pseudo-first-order kinetics over a wide range of the inhibitor concentration. The second-order-rate constant for the inactivation of hexokinase was estimated to be 45 M - t . s - t . Hexokinase was protected more by sugar substrates than by nudcoside triphosphates during inactivation by o-phthalaldehyde. Absorption spectrum (~'ma,, 338 nm), and fluorescence excitation (~m,, 363 nm) and emission (Xm.~ 403 nm) spectra of the bexokinase-o-phthal- aldehyde adduct were consistent with the formation of an isoindole derivative. These results also suggest that sulfhydryl and e-amino functions of the cysteine and lysine residues, respectively, participating in the isoindole formation are about 3 A apart in the native enzyme. About 2 real of the isoindole per real of hexokinase dimmer were formed following complete loss of the phospbotransferase activity. Chemical modification of hexokinase by iodoacetamide in the presence of mannose resulted in the modification of six sulfhydryl groups per maul of hexokinase with retention of the phospbotransferase activity. Subsequent reaction of the iodeaeetamide modified hexokinase with o-phthalaldehyde resulted in complete loss of the phosphotransferase activity with concomitant modification of the remaining two sulfhydryl groups of hexokinase. Chemical modification of hexokinase by iodnacetamide in the absence of mannose resulted in complete inactivation of the enzyme. The iodoacetamide inactivated hexokinase failed to react with o-phthaluldehyde as evidenced by the absence of a fluorescence emission maximum characteristic of the isoindele derivative. The boloenzyme failed to react with [5 '-( p-fluorosulfonyl)henzoylladenosine. The dissociated hexokinase could be inactivated by 15 '-( p-fluorosulfonyl)benzoylladeuosine; the degree of inactivation paralleled the extent of reaction between o-phthalaldehyde and the nucleotide-analog modified enzyme. Thus, it is concluded that two cysteines and lysines at or near the active site of the hexokinase were involved in reaction with o-phthalaldehyde following complete loss of the phosphotransferase activity. An impmlant finding of this investigation is that the lysines, involved in isoindole formation, located at or near the active site are probably buried. The molar transition energy of hexokinase-o-phthalaldehyde adduct was estimated to he 116 k J / m o l which compares favorably with a value of 127 k J / r e a l for the synthetic

Abbreviations: cAMP, adenosine cyclic-Y,5'-monophosphate; cGMP, guanosine cyclic-3',5'-monophosphate: AATP, (9-/~-D- arahinofuranosyl)adenine 5'-triphosphate; 6.NMe2-ATP , N6.dimethyladenosine triphosphate; AMPPNP, /~-~,-imidoadenosine tri- phosphate; AMPPCP, P-¥-methyleneadenosine triphosphate; FSBA, [5'-(p-fluorosnlfonyl)]benzoyladenosine; Ser-peptide, Leu-Arg- Arg-Ala-Ser-Leu.Gly; EA, 1-(/~-hydroxylethylthiol)-2-O-hydroxyethylisoindole; DTNB, 5,5'-dithiobis(2-nitrobenzoic acid); PLP, pyridoxal phosphate; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesuifonic acid.

Correspondence: R.N. Purl, Thrombosis Research Center, Temple University Health Sciences Center, 3400 North Broad Street, Philadelphia, PA 19140, U.S.A.

0168-4838/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

35

isoindole, l-[(p-hydroxyethyl)thiol-2-(j0-hydroxyethy|)isoindole, in hexane, indicating the hyarophobic en- vironments of the cysteine and lysine residues participating in the isolndo|e formation in the reaction between hexokinase and o-phthala|dehyde.

introduction

Hexokinase (ATP : o-hexose-6-phosphotrans- ferase, EC 2.7.1.1) catalyzes the transfer of the ~,-phosphoryl group of ATP to glucose in the presence of Mg 2+, and plays an important role in the overall energy-producing metabolism of the cell. The yeast hexokinase is a dimer consisting of two identical subunits [1]. The extensive literature available concerning the interaction of substrates and products with specific amino acid residues of the enzyme is not unequivocal, and our under- standing of the structural organization of the ac- tive site of hexokinase remains incomplete. Otieno et al. [2] demonstrated that alkylation of a single thiol per monomer of the hexokinase by the affin- ity reagent, N-bromoacc,yl-2-D-galactosamine, is sufficient to cause complete inactivation, and that the other three thiols per monomer are truly non- essential. The essential thiol, since then, has been identified as Cys-244 in the incomplete amino acid sequence of hexokinase determined by Gray et al. [3]. Chemical modification of yeast hexokinase by 2,4,6-trinitrobenzenesuifonate leads to a loss of its phosphotransferase activity. This has been attri- buted to modification of e-amino groups of the protein [4]. However, the presence of a lysine residue(s) at or near the active site of yeast hexokinase has never been unequivocally demon- strated. There are also considerable disagreements between portions of the sequence of yeast hexokinase determined by X-ray methods [5] and chemical methods [3].

Previous work in our laboratory demonstrated the usefulness of o-phthalaldehyde [6-8] in study- ing the structural organization of the active site of the catalytic subunit of cAMP-dependent protein kinase [9], active and regulatory sites of cGMP-de- pendent protein kinase [10] and the active site of fructose-l,6-bisphosphatase [11]. We also demon- strated the usefulness of this reagent in obtaining information about isoindoles derived from low- molecular-weight aminothiols, such as glutathione [12]. o-Phthalaldehyde, a homobifunctional cross-

linking reagent is a useful probe for examining those cysteine and lysin¢ residues in a protein whose sulfhydryl and e-amino functions, respec- tively, are about 3 ,A apart [9,10]. We investigated the inactivation of yeast hexokinase by o-phthal- aldehyde and found the presence of a cysteine and a lysine at or near its active site. Preliminary reports of this work have previously appeared [13,14].

Materials and Methods

Yeast hexokinase, glucose-6-phosphate dehy- drogenase (yeast), sugars, nucleosides and nucleo- tides were obtained from Sigma. All other chem- icals used were of reagent grade. Concentrations of solutions of ATP, o-phthalaldehyde and isoin- dole ring were determined spectrophotometrically as described previously [9,10]. o-Phthalaldehyde was prepared freshly in methanol prior to each experiment. Fluorescence measurements and ab- sorption spectroscopy measurements were made with a SLM 4800 spectrofluorometer and Gilford Model 2600 spectrophotometer, respectively, as described by Purl et al. [9,10]. Proteins were esti- mated by the method of Lowry et al. [15] or by using a specific absorption of 0.947 cm2/mg ~ [16]. Commercial samples of yeast hexokinase used in this work were homogeneous as judged by poly- acrylamide gel electrophoresis in the presence and absence of sodium dodecyl sulfate. Size exclusion chromatography on a Sephacryl S-200 column also showed the presence of only one protein of Mr 100000 as evidenced from the elution profile which showed the presence of a single symmetrical peak. Commercial samples of yeast hexokinase have been suitably t/sed to study the mechanism of catalysis [17-20]. Molecular weights of 100000 for the di- meric [21] and 50000 for the monomeric [22] forms were used in the calculations.

Enzyme assay Hexokinase activity was measured spectropho-

tometrically by coupling with glucose-6-phosphate

36

dehydrogenase and following reduction of NADP + at 340 nm [23]. A total volume of I ml of the assay mixture contained the following: glucose, 1 raM; ATP, 2 mM; MgCI 2, 10 mM; Tris-HCl (pH 8.5), 50 mM; NADP +, 0.5 mM and glucose-6-phos- phate dehydrogenase, 1-3 units. All assays were performed at 25 ° C. Typical concentrations of the enzyme in the assay mixture were 2-5 nM. The commercial samples of yeast hexokinase had 290-400 units of activity per mg protein. One unit of activity is defined as the amount of enzyme that catalyzes the phosphorylation of 1 ~mol of glu- cose per min at 25 ° C.

Results

Inactivation of hexokinase by o-phthalaldehyde The time-course of inactivation of hexokinase

by o-phthalaldehyde at 25 °C is shown in Fig. 1. o-Phthalaldehyde, in the concentration range 0.15-2.4 raM, inhibited hexokinase rapidly. When the logarithm of the percent residual activity was plotted against the time of incubation with o- phthalaldehyde, linear relationships were obtained over a 16-fold range of the concentration of the inhibitor. When the pseudo-first-order rate con- stants (kobsd) calculated from the slopes of such plots, were plotted against the concentration of o-phthalaldehyde, a linear relationship was ob- tained (Fig. 1, inset). The slope of this plot yielded a second-order rate constant, K, equal to 45 M -

• s - t . Furthermore, reagents containing sulfhydryl, and sulfhydryl and amino functions, e.g., fl- mercnptoethanol and cysteine, did not reverse the reaction. Instead, these reagents were used to terminate the reaction between o-phthalaldehyde and hexokinase at any given time during the pro- gress. Thus, it is concluded that the reaction of hexokinase and o-phthalaldehyde is irreversible. Linear dependence of pseudo-first-order reaction rates over a wide range of the concentration of o-phthalaldehyde (Fig. 1, inset) suggests that the inactivation reaction did not follow saturation kinetics. NoAcetyl-D-glucosamine ( K i = 1 mM) is a competitive inhibitor of hexokinase [24,25]. It was found that 10-20 mM N-acetyl-D-glucos- amine did not protect hexokinase from inactiva- tion by o-phthalaldehyde. The second-order-rate constant for inactivation of hexokinase by o-

O~

c o E

>

0.5 1.0 15 2.0 2.5 3,0 3.5 Time (rain)

Fig. l. Time-course of inactivation of yeast hexokinase by o-phthalaldehyde. The hexokinase (15 p g / | 0 0 pl) was incubated in 100 mM Hepes-NaOH (pH 7.3), 0.1 mM EDTA and 1~ methanol at 25°C. At the times specified, the reaction was terminated by transferring a portion of the incubation mixture into a solution of cysteine and fl-mercaptoethanol in the same buffer yielding a final concentration of 20 and 5 raM, respectively. Residual phosphotransferase activity was then measured as described under Materials and Methods. The data are p[otted as the natural logarithm of percent activity remain- ing vs. time. o-Phthalaldehyde concentrations were as follows: (o), 0 raM; (v), 0.15mM; (A), 0.3 mM; (A), 0.6 raM; (D), 0.9 raM, (i), 1.2 mM; (t), 2.4 mM. The inset shows a plot of pseudo-first-order rate constants vs. concentration of o-phtha-

laldehyde.

phthalaldehyde in the presence of 40 mM N- acetyl-D-glucosamine was found to be 42 M - t . s - 1 (data not shown). It is thus concluded that reac- tion between o-phthalaldehyde and hexokinase followed rapid second-order kinetics and did not involve the formation of an enzyme-inhibitor com- plex before the isoindole formation. The hexokinase-o-phthalaldehyde adduct was found to be fluorescent (as described later). The second- order rate constant, determined spectrofluoromet- rically, for inactivation of hexokinase by o-phtha- laldehyde was also 45 M -1. s - l [12].

TABLE I

EFFECT OF VARIOUS SUGARS ON THE PROTECTION OF HEXOKINASE FROM INACTIVATION BY o-PHTHALALDEHYDE

Solutions containing 0.1 tiM hexokinase, 100 mM Hepes-NaOH (pH 7.3), 0.1 mM EDTA, 1~ methanol and 0.8 mM o-phthal- aldehyde were incubated for 30 s at 25 o C in the presence of specified concentration of various sugars. The reactions were terminated as described in Fig. 1. Appropriate controls devoid of o-phthalaldehyde were included in each case. Residual phosphotransferase activity in the reacaon mixtures was subse- quently measured as described under Materials and Methods.

Addition Concentration • Activity (mM) remaining

None 26 Glucose 1 38 Glucose 10 73 Mannose 1 54 Mannose 10 100 Fructose 1 30 Fructose 10 63 Galactose 1 26 Galactose I0 26 Arabinose 1 27 Arabinose 10 26 Lyxose 10 30 Giucosamine 10 46 Glucose 6-phosphate 10 66 1,5-Anhydroglucitol 15 27 1,5-Anhydromannitol 15 27 2-Deoxyglucose 10 58 6-Deoxyglucose 10 27

Effect of substrates and substrate analogs on the reaction between hexokinase and o-phthalaldehyde

Results of the effect of various sugars on the inactivation of hexokinase by o-phthalaldehyde are summarized in Table I. In general, those sugars that are good substrates for hexokinase provided considerable protection to the enzyme from in- activation by o-phthalaldehyde. At 1 mM or 10 mM concentration, protection provided by sugar substrates followed the order of their relative sub- strate affinities, i.e, mannose > glucose > fructose [26]. The fact that glucosamine and 2-deoxyglu- cose provided protection is consistent with the findings that they are also substrates for the phos- photransferase activity of hexokinase [27]. Galac- tose [28], 1,5-anhydroglucitol and 1,5-anhydro- mannitol [19,29], and arabinose and 6-deoxyglu- cose [19] are poor substrates in phosphorylation

37

reaction catalyzed by hexokinase. They provided little or no protection to hexokinase from inactiva- tion by o-phthalaldehyde. Lyxose provided little protection to hexokinase in the current studies. We found that 10 mM glucose 6-phosphate, one of the products of hexokinase reaction, provided significant protection to the enzyme from inactiva- tion by o-phtha!aldehyde.

Results of the investigation of the effect of nucleotides, nucleosides, chelators and metal ions are presented in Table If. The protection provided

TABLE I!

EFFECT OF VARIOUS NUCLEOTIDES, NUCLEOSIDES. CHELATING AGENTS AND METAL IONS ON THE PROTECTION OF THE HEXOKINASE FROM IN- ACTIVATION BY o-FHTHALALDEHVDE

Solutions containing 0.1/t M hexokinase, 100 mM Hepes-NaOH (pH 7.3), 0.1 mM EDTA, 1~ methanol and 0.8 mM o-phthal- aldehyde were incubated for 30 s at 25 o C in the presence of above reagents. The other experimental details are the same as those described in Table I.

Addition Concentration % Activity (mM) remaining

None 24

Nucleotides and nucleosides: AMP 1 23 AMP 5 23 ADP 1 26 ADP 5 30 ATP 1 38 ATP 5 53 AATP 5 38 ITP I0 40 GTP I0 22 UTP 10 23 CTP 10 22 AMPPNP 5 23 AMPPCP 5 20 6-NMe2-ATP 1 43 cAMP 1 24 cGMP 1 25 Adenosine 1 23 Guanosine 1 22

Chelators: EGTA 1 26 EDTA 1 22

Metal ions: Ca 2+ 5 21 Mn 2+ 5 18 Mn 2 + + EDTA 7 26

2.0 by 5 mM ATP was greater than that provided by ADP under similar conditions. No protection was observed with 5 mM ATP. AATP was less effec- tive than ATP in protecting hexokinase from in- activation. Among the other nucleoside triphos- phates examined, only 10 mM ITP afforded noticeable protection to the hexokinase from in- activation. N-Dimethyl adenosine triphosphate at 1 mM concentration provided significant protec- tion. Our investigations showed that AMPPNP and AMPPCP, the ATP analogs, did not protect hexokinase from inactivation.

Characterization of the nature of product formed in the reaction between hexokinase and o-phthalalde- hyde

The absorption spectrum of the hexokinase-o- phthalaldehyde adduct exhibited a shoulder at 280 and a maximum 337 nm (Fig. 2A). Excitation of the adduct at either 295 or 338 nm led to the appearance of a fluorescence emission maximum at 403 nm (Fig. 2B). The excitation spectrum of the hexokinase-o-phthalaldehyde adduct (emis- sion wavelength at 403 nm) showed the presence of a major excitation band at 363 nm and a shoulder at 290 nm (Fig. 2C). Spectral characteris- tics of hexokinase-o-phthalaldehyde adduct are consistent with formation of an isoindole ring by the covalent chemical interaction of SH and e-NH 2 functions of cysteine and lysine residues, respec- tively, with o-phthalaldehyde [8-12]. The hexokin- ase-o-phthalaldehyde adduct was stable over a period of 24 h.

Fig. 2. Absorption spectrum and fluorescence excitation and emission spectra of the hexokinase-o=phthalaldehyde adduct. Hexokinase (I mg/ml) was incubated with 1.6 mM o-phthal- aldehyde for 1.5 rain at 25 °C and the reaction terminated as described in Fig. I. The hexokinase-o-phthalaldehyde adduct was purified by gel filtration as described previously by Puri et al. [9]. (A) Absorption spectra: hexokinase (. ), and hexokinase-o-phthalaldehyde adduct ( . . . . . . ). (B) Fluores- cence emission spectra: ( . . . . . ) hexokinase; ( ), hexokinase-o-phthalaldehyde adduct, excitation at 338 nm; ( . . . . . . ) hexokinase-o-phthalaldehyde adduct, excitation at 295 nm; (C) Fluorescence excitation spectra: ( ) hexokinase-o-phthalaldehyde adduct, emission wavelength at

403 nm.

1.6.

1 .2 ,

O . a L o oa- .,Q <

04-

1.0

0 4

210

A

\

~\\" ", . . . . . . . . . . . . . . . . . .

2~3 2~6 ~9 3~2 Wovelengt h (nm)

375

08-

250

i ~ 06-

L

.,-, Q 4 .

O

u. 02.

1.0

36o

i

'i

/ 1 ~ ,, ~./ . , ..... ,.~,.....~_"-;

7"<1 350 400 450 500

Wavelength (rim)

B

5 5 0

t - ~ o6- - - E - ~3 u E ~

~ 04-

OL

02.

250 360 3~0 4(JO Wavelength (nm)

38

450

39

Stoichiometry of the reaction between the hexokinase and o-phthalaldehyde

The stoichiometry of the hexokinase-o-phthal- aldehyde adduct was determined by using an ab- sorption coefficient of 7.66 mM -1- cm-1 at 337 nm as described previously by Puri et al. [9]. At various times during inactivation of yeast hexokinase by o-phthalaldehyde, the incubation mixture was removed to assay the residual phos- photransferase activity and quantitate the isoin- dole derivative formed. In Fig. 3, percentage of residual phosphotransferase activity was plotted as a function of amount of o-phthalaldehyde in- corporated per mol of the enzyme. When the data were extrapolated to zero enzyme activity, ap- proximately, 1.95 mol of reagent were incorpo- rated per tool of the hexokinase dimer. When different amounts of hexokinase were completely inactivated by o-phthalaldehyde and the ab- sorbance at 337 nm was plotted as a function of

Binding s~tes (/aM) 1 0 0 ~ 2 4 6 ,8 1~)1.2

I ~ E o 3 o ' ' ' , .

\ =~ ~o~o

.-,, o,o . ,

~ <005 .,

~ 4 0 - \ ~ 2 a 4 5 6 ~ . Hexoklnclse ( iJM )

20 . , , , , • ,

0 4 0 8 12 1.6 2.0 2 4 Mof o -Ph tho lo !dehyde / Mol hexok inose

Fig. 3. Stoicbiometry of the reaction between the hexokinase and o-phthalaldehyde. Hexokinase (40/Lg/3 ml) was incubated with 0.3 mM o-phthalaldehyde as described in Fig. 1. Aliquots corresponding to 190 ~l were removed at various times and the reaction quenched as described in Fig. 1. ~;mall portions of the incubation mixture were appropriately diluted and assayed for residual phosphotransferase activity and the remainder dialyzed for 20 h against the same buffer (x 3) as used during incuba- tion with o-phthalaldehyde. The number of isoindole groups were quantitated as described previously by Purl et al. [9,10]. The inset shows a plot of the various amounts of the hexokinase vs. absorbance (o) at 337 nm following complete inactivation of the hexokinase and a plot of the number of binding sites (or

isoindole groups formed) vs. absorbance (e).

the amount of isoindole derivative, a linear rela- tionship was obtained (Fig. 3, inset). A plot of the binding sites vs. the absorbance was also linear. A value of 2.10+0.21 moi of isoindole groups formed per mol of the hexokinase was obtained. It is thus concluded that complete inactivation of hexokinase by o-phthalaldehyde resulted in the chemical modification of two cysteine and two lysine residues per mol of the enzyme.

Characterization of the site of reaction between hexokinase and o-phthaialdehyde

Results of covalent chemical modification of thiols of hexokinase are presented in Table III. When hexokinase was treated with 1 mM DTNB for 1.5 h, it lost 90~ of its phosphotransferase activity and 1.8 mol of suifhydryl groups per tool of the enzyme were modified (experiment B, Table III). Over a period of 20 h, under the conditions described, DTNB was found to react with 8.30 mol of sulfhydryl groups per mol of the enzyme (experiment C, Table III). Hexokinase pretreated with mannose was not inactivated by treatment with DTNB, although it modified about six sulf- hydryl groups per mol of the enzyme (experiment D, Table III). Mannose was earlier shown to protect the hexokinase from inactivation by o- phthalaldehyde (Table I). Treatment of the hexokinase with iodoacetamide for a brief period of time (experiment E, Table III) led to almost complete loss of the phosphotransferase activity with concomitant modification of four thiols per mol of the enzyme. Prolonged treatment with iodoacetamide (experiment F, Table III) alkylated the remaining four thiols of the enzyme. However, pretreatment of hexokinase with mannose fol- lowed by prolonged incubation with iodoacetam- ide modified six thiols of hexokinase dimer with almost complete retention of the phosphotrans- ferase activity (experiment G, Table IID. When the iodoacetamide modified hexokinase was treated with o-phthalaldehyde (experiment H, Ta- ble IIl), all of the phosphotransferase activity was abolished with concomitant modification of the remaining two thiols of the enzyme. The data presented in Table III point to the fact that in- activation of hexokinase by o-phthalaldehyde is more likely a consequence of modification of thiol I at or near the active site of the enzyme subunit.

40

TABLE II1

SULFHYDRYL GROUPS MODIFIED IN THE REACTION BETWEEN THE HEXOKINASE AND VARIOUS REAGENTS

In experiments (A) --* (G), hexokinase (2.7/~M) was incubated with DTNB/o-phthalaidehyde/iodoacetamide in a solution contain- ing 50 mM glycine-NaOH (pH 8.5) and 100 mM NaCl under the conditions described in the table. Incubation mixtures containing o-phthalaldehyde also contained lC£ methanol. In experiment H, hexokinase (17.8 ~M) was first incubated, in the buffer system described above, with mannose and iodoacetamide. Subsequently, the reaction mixture was subjected to gel filtration on a Sephadex G-25 column (0.45 × 18.5 mm) with a flow rate of 2 ml/min. Protein containing fractions were used for enzyme activity measurement and subsequent reaction with o-phthalaldehyde.

In the case of reaction of hexokinase with o-phthalaldehyde, sulfhydryl groups participating in the isoindole formation were quantitated by the procedure described previously by Puri et al. [9,10]. Sulfhydryl groups participating in reaction with DTNB were estimated by using an absorption coefficient of 1.36.104 M - I . c m - I at 412 nm [50].

Experiment First Second Reaction • activity mol SH modified/ treatment treatment conditions remaining moi hexokinase

A hexokinase.t- 25°C, 1.5 min 0 2.03 o-phthalaldehyde

B hexokinase + 30 o C, 1.5 h 10 1.80 DTNB (1 raM)

C hexoldnase + 35 o C, 20 h 0 8.30 DTNB (1 mM)

D hexokinase + 35 o C, 2h 100 mannose (7 raM)

DTNB 35°C, 20 h 100 6.16 (lmM)

E hexokinase + 35 o C, lh 3 3.90 iodoacetamide (6 mM)

F hexokinase + 35 o C, 20 h 0 7.84 iodoacetamide (6 mM) hexokinase + mannose (7 mM)

G

H hexokinase + mannose (7 raM) + iodoacetamide (6 raM)

iodoace- tamide (6 raM)

o-phthal- aldehyde

35 o C, 2h 100

35 o C, 20 h 93 5.85

35 o C, 20 h 97 5.93

25°C, 1.5 min 0 2.13

When hexoldnase was alkylated with iodo- acetamide, as described in experiment E (Table Ill), there was almost complete loss of the phos- photransferase activity. Subsequent treatment of the iodoacetamide modified enzyme with o- phthalaldehyde failed to produce the fluorescence emission band characteristic of hexoldnase-o- phthalaldehyde adduct observed in control experi- ment (Fig. 4A) under similar experimental condi- tions. This result combined with the evidence pre- sented in the preceding sections strengthen our

belief that it was the thiol present at or near the active site of hexokinase that more likely reacted with o-phthalaldehyde following complete loss of the phosphotransferase activity.

When hexokinase was incubated with FSBA for 20 h, there was no loss of the phosphotransferase activity. Subsequent reaction of the above incuba- tion mixture with o-phthalaldehyde produced the fluorescence emission maximum whose intensity was comparable with the one obtained by incubat- ing untreated hexokinase with o-phthalaldehyde

1.0.

0.8,

1,0,

0.6,

L 0.4 P ~. 0.2

250

QS'

250

o8,

u_ 0.2,

380

1,0

> 0.8]

~ ~ o6] 8 ~o.4 °=~02 E

250

360 350 400 450 500 V~wlength (nrn)

, " , i

360 350 aoo 4~50 500 V~velen,3t h (rim)

3~50 4()0 450 560 V~velength (nrn)

41

A

550

B

550

C

5.50

under identical conditions (Fig. 4B). When hexokinase was dissociated in glycine-NaOH-NaC1 ( I = 0.6) in the presence of 10 mM glucose, as described by Derechin et ~,!. [22], and the dissoci- ated enzyme treated with 4 mM FSBA at 25 °C for 20 h, there was 67% loss of the phos- photransferase activity. The FSBA modified en- zyme so obtained on treatment with o-phtha- laldehyde exhibited a fluorescence maximum whose intensity was diminished by almost 67% compared with the one obtained by treating a control under identical conditions. These results show that although ATP partially protected hexokinase from inactivation by o-phthalaldehyde, the ATP-binding site was not amenable to FSBA modification in the holoenzyme. The ATP-binding site only became susceptible to FSBA modificl- tion in the dissociated hexokinase.

Treatment of yeast hexokinase with 6 M urea for 30 rain led to its irreversible inactivation. The urea denatured enzyme on treatment with o- phthalaldehyde did not form an isoindole deriva- tive as evidenced by the absence of the fluores- cence emission maximum at 403 nm (Fig. 4C).

Fig. 4. Fluorescence emission spectra of the chemically rood° ified hexokinase. (A) lodoacetamide treatment. A solution containing the hexokinase (269/Lg/ml). 50 mM glycine-NaOH (pH 8.5) and 100 mM NaC! was incubated with 6 mM iodo- acetamide at 35 ° C for 1 h. The incubation mixture containing iodoacetamide modified-hexokinase was subjected to gel filtra- tion on a gephadex G-25 column as described previously [9]. Protein-containing fractions were collected and the fraction corresponding to the peak protein was incubated with !.2 mM o-phthalaldehyde at 25°C for 1.5 rain and its fluorescence emission spectra recorded. A control consisting of the unmod- ified hexokinase was carried through the same reaction se- quence. The iodoacetamide-modified hexokinase ( . . . . . ); the iodoacetamide-modified hexokinase incubated with o-phthal- aldehyde ( . . . . . . ); and hexokinase incubated with o-phthal- aldehyde ( ~ ) . (B) FSBA treatment. The hexoHnase was treated with 4 mM FSBA as described previously [9]. It was subsequently incubated with o-phthalaldehyde as described above. The figure shows the emission spectra of FSBA-mod- ified hexokinase ( . . . . . ); The FSBA-treated hexokinase incubated with o-phthalaldehyde ( . . . . . . ); and hexokinase incubated with o-phthalaldehyde ( ~ ) . (C) Urea treat- ment. The denaturation of the hexokinase by 6 M urea was carried out as described previously [9]. The emission spectra of the urea-denatured hexokinase ( . . . . . . ); urea incubated with o-phthalaldehyde ( . . . . . ); the urea-denatured hexokinase with o-phthalaldehyde ( - - - - - ) ; and hexokinase incubated with

o-phthalaldehyde ( ~ ) are shown in the figure.

42

Our experience has shown that urea itself sometimes produces a low intensity fluorescence emission maximum in the presence of o-phthal- aldehyde. The reasons for this are not understood at the present time. These results are consistent with the fact that the SH function of the cysteine and ~-NH2 function of the lysine participating in an isoindole formation during inactivation of hexokinase by o-phthalaldehyde maintain their spatial proximity necessary to react with the in- hibitor only in the tertiary structure of the en- zyme.

Effect of pyridoxal phosphate on yeast hexokinase The data presented so far indicate the presence

of a lysine residue in close proximity to a cysteine residue at or near the active site of a hexokinase subunit. However, efforts to modify the holoen- zyme with FSBA were unsuccessful. This prompted us to examine further, by other methods, the effect of chemical modification of iysine residue(s) on the phosphotransferase activity of hexokinase. It was found that the loss of phosphotransferase activity following chemical modification of the hexokinase by PLP was both time and concentra- tion dependent. The results of the time-dependent inactivation of hexokinase by PLP are presented in Table IV. The hexokinase lost only 25~ of its phosphotransferase activity in 20 h when in-

TABLE IV

TIME-DEPENDENT INACTIVATION OF HEXOKINASE BY 23 mM PYRIDOXAL PHOSPHATE

A solution containing 5 ~M hexokinase, 50 mM Tris-HCI (pH 7,5) and pyridoxal phosphate was incubated at 25°C. A small aliquot of the incubation mixture was transferred, at time intervals indicated into 1 ml of the assay mixture and the residual phosphotransferase activity determined as described under Materials and Methods. The concentration of the PLP solution was determined spectrophotometrically by using a molar absorption coefficient of 4990 at 388 nm [51].

Time ~ Activity remaining

0 100 2 rain 50

30 rain 46 3h 44 7 h 37

20h 13

TABLE V

CONCENTRATION DEPENDENT INACTIVATION OF HEXOKINASE BY PYRIDOXAL PHOSPHATE

In each case, a solution containing 0.5/~M hexokinase, 50 mM Tris-HCl (pH 7.5) and pyridoxal phosphate was incubated for 2 rain at 25 o C. Subsequently, a small aliquot of the incubation mixture was directly transferred into I ml of the assay mixture and residual phosphotransferase activity determined as de- scribed under Materials and Methods. in experiments involv- ing the determination of lysine residues modified, 1.2 /iM hexokinase was similarly incubated followed by reduction with sodium borohydride in the presence of 2-octanol as described by Poulose and Kolattukudy [30]. Quantitation of lysine was performed using a molar absorption coefficient of 9700 at 325 nm for the reduced Schiff base [52].

Pyridoxal ~ Activity mol phosphate remaining lysine/mol (raM) hexokinase

0 100 0 2.5 88 1.71 5.0 70 3.05

10.0 57 5.95 23.0 48 12.80

cubated with 2.5 mM PLP (data not shown). We found that incubation with 23 mM PLP for 20 h resulted in 87% loss of the phosphotransferase activity of the enzyme. Examination of the data in Table IV also show that loss of the phos- photransferase activity as a result of incubation of hexokinase with 23 mM PLP, probably, followed biphasic kinetics. About 50~ loss of phos- photransferase activity occurred in 2 min and the rest occurred over a prolonged time period. Hexokinase was better protected by glucose than ATP during inactivation by PLP. The results of inactivation of the hexokinase at various con- centrations of PLP are presented in Table V. The data show that incubation of hexokinase with 2.5-23.0 mM PLP for 2 rain resulted in 12-52~ inhibition of the enzyme activity. The fact that the inactivation of hexokinase by PLP was indeed due to formation of the Schiff base and not due to nonspecifie effects, e.g., noncovalent interactions, was ascertained by spectroscopic methods. The PLP modified hexokinase on subsequent reduc- tion with sodium borohydride exhibited an ab- sorption band with ~ m a x at 325 nm and a fluores- cence emission band with a ~ma~ at 395 nm, characteristic of a reduced Schiff base [30]. Con-

43

sistent with these observations were tile findings that 25 mM pyridoxamine phosphate (with no aldehyde function) or 25 mM pyridoxal (with no phosphate function) failed to inhibit hexokinase. A possible explanation is that pyridoxal exists as a hemi-acetal and hence the aldehyde function is not available to form a lysine-Schiff base. The data presented in Table IV show that modification of hexokinase by 10 mM PLP resulted in about 457o inhibition of the enzyme with concomitant modification of six lysine residues per mol of the enzyme. These results bear out the fact that chemical modification of three lysine residues per subunit of the enzyme resulted in substantial loss of phosphotransferase activity and further modifi- cation of the lysine residues produced less pro- nounced effect on the inhibition of the enzyme. The data, thus, suggest that chemical modification of lysine residues in the yeast hexoldnase in some way is associated with loss of its phos- photransferase activity.

Environment of the cysteine and lysine residues par- ticipating in reaction between the hexokinase and o-phthalaldehyde

Linear free energy relationships between the molar transition energies (E T) measured by Dim- roth et al. [31] and fluorescence emission maxi- mum of the o-phthalaldehyde adduct of a syn- thetic isoindole, 1-[(fl-hydroxyethyl)thio]-2-(fl-hy- droxyethyl)isoindole (EA) have been used by us to probe the degree of polarity of the microenviron- ment of the cysteine and lysine residues participat- ing in isoindole formation in reaction between various enzymes and o-phthalaldehyde [9-11]. The molar transition energy of the hexokinase-o- phthalaldehyde adduct was found to be 1.16 k J/tool. Molar transition energy of the o-phthal- aldehyde adducts of the hexokinase, and cyclic nucleotide-dependent protein kinases compare favorably with that of EA in hexane (127 k J/tool) (Refs. 9 and 10; other references cited therein). These results show that the cysteine and lysine residues participating in isoindole formation in reaction with o-phthalaldehyde are located in hy- drophobic environment of hexokinase. To our surprise, molar transition energy of o-phthalalde- hyde adduct derived from glutathione (L-y- glutamyl-L-cysteinyl-glycine-a tripeptide) was

estimated to be 136 k J/reel [12], a value also close to the value obtained for the synthetic isoindoie, EA, in hexane.

Discussion

Hexokinase was found to be rapidly and irre- versibly inactivated by o-phthalaldehyde. The sec- ond-order rate constant for the inactivation of hexokinase (K = 45 M- 1. s - i) was comparable with those determined for fructose-l,6-bis- phosphatase [11] and cGMP-dependent protein kinase [10]. Absence of saturation kinetics over a wide range of concentration of o-phthalaldehyde, irreversible nature of the reaction, and inability of N-acetylglucosamine to significantly affect the second-order-rate constant for the inactivation of hexokinase by o-phthalaldehyde, suggest that the primary inhibitory process in this system appears to be the chemical modification of cysteine and lysine residues at or near the active site of the enzyme via a simple bimoleeular reaction. The second-order rate constant for the inactivation of hexokinase resembles that for the reaction be- tween glutathione and o-phthalaldehyde (K = 25 M - i . s - l ) [12]. Similar results have been obtained by other investigators [32]. These similarities have been discussed in detail elsewhere [12].

Absorption and fluorescence spectral properties of the hexokinase-o-phthalaldehyde adduct were similar to those of the o-phthalaldehyde adduet of the catalytic subunit of cAMP-dependent protein kinase [9] and cGMP-dependent protein kinase-o-phthalaldehyde adduct [10], and fruc- tose-l,6-phosphatase [11 ]. Polyacrylamide gel elec- trophoresis in the absence of sodium dodecyl sulfate and size-exclusion chromatography on a Sephacryl S-200 column of the hexokinase-o- phthalaldehyde adduct showed the absence of products with molecular weight higher than that of the hexokinase. It is, therefore, reasonable to conclude that the reaction between hexokinase and o-phthalaldehyde is intramolecular in nature (cf. Puri et al. [9,10]). Investigation of the stoichiom¢.iy of the reaction between the hexokinase and o-phthalaldehyde following com- plete inactivation of the enzyme, revealed forma- tion of about 2 mol of isoindole groups per reel of the hexokinase dimer (Fig. 3 and Table lII).

44

The results of the protection studies sum- marized in Table I show that the presence and proper orientation of the hydroxyls at 1, 3, 4 and 6 position in the hexoses is important for the protec- tion provided to hexokinase by them. Fructose provided good protection to the hexokinase and is also a good substrate in phosphorylation reaction catalyzed by the hexokinase. Lyxose, a competi- tive inhibitor of glucose in the hexokinase reaction [33], provided minor protection. The fact that hydroxyl at C-2 in glucose is relatively unim- portant is consistent with the observation that mannose (C-2 epimer of glucose) provided the best protection to hexokinase from inactivation. Similar observations have been made by Danen- berg and Cleland [19] during their investigation of the protective elfect of 2,5-anhydro-D-mannitol and 2,$-anhydro-v-glucitol on the inhibition of hexokinase by CrATP. The importance of the role of binding of hydroxyl at carbons 1, 3, 4 and 6 in the sugar moiety to hexokinase has been discussed in detail elsewhere [34,35]. The protection pro- vided by glucose-6 phosphate may be associated with conformational changes induced by interac- tion of the sugar phosphate with hexokinase [36,37]. At 5 mM concentration, ATP was a better protector than ADP, while AMP was found to be completely ineffective (Table II). We have previ- ously shown that during inactivation of the cata- lytic subunit of cAMP-dependent protein kinase [9] and cGMP-dependent protein kinase [I0], the protection provided by the nucleotides to the en- zymes from inactivation by o-phthalaldehyde fol- lowed the order of their respective K d values, i.e., ATP > ADP > AMP [38,39]. Partial protection provided by ITP may be due to weak interaction of the nudeotide with hexokinase, for ITP was previously shown to stimulate the ATPase activity of hexokinase [33], N-Dimethyladenosine triphos- phate provided significant protection to hexok/nase from inactivation. Dorgan and Schus- ter [40] demonstrated that N-dimet]~yladenosine triphosphate was capable of binding tightly to hexokinase. Whereas interactions of the adenine moiety of ATP play a more important role in the binding of the nudcotides to the catalytic subunit of the cAMP-dependent protein kinase [38,41] and cGMP-dependent protein kinase [39], the interac- tions of the ribose moiety play a more important

role in the binding of the nucleotide to hexokinase [42-44]. During the course of this work, it was found that the presence or absence of Mg -'+ did not affect the extent of protection provided by the nucleotides. These observations are consistent with the results of other investigators who showed that, in the case of hexokinase, Mg 2+ was not required for the binding of nuclcotides [27,43] but was required for catalysis [45-47].

Although ATP partially protected hexokinase from inactivation by o-phthalaldehyde, attempts to inactivate the holoenzyme by FSBA were un- successful (Table II and Fig. 4B). However, it was possible to inactivate the dissociated hexokinase by FSBA (see Results). This implies that ATP binding site in the holoenzyme is probably buried in hydrophobic environment. The dissociation of hexokinase into monomers could expose this site, thus making the buried lysine amenable to reac- tion with FSBA. These arguments are attractive in the light of the fact that the cysteine residue, i.e., the thiol I, is known to reside in the hydrophobic cleft of the hexokinase dimer [2,43] and the lysine residue, participating in the isoindole formation during inactivation by o-phthalaldehyde, must also be resident in the vicinity of this cysteine. If the thiol I site of the hexokinase is only partially protected by ATP from inactivation by o-phthal- aldehyde, then, how does ATP bind effectively to the enzyme during catalysis? The answer to this lies in the fact that binding of glucose to hexokinase causes a conformational change [34,48] and that a further substantial conformational change occurs as a result of ATP binding to the glucose-hexokinase complex [34]. During catalysis, the y-phosphate of ATP is placed in the correct position for phosphoryl transfer only after the sugar substrate is bound. That the cysteine and iysine residues in hexokinase participating in the isoindole formation in reaction with o-phthal- aldehyde are in close proximity only in the native enzyme, was confirmed by the experime~ in_ which the urea-denatured hexokinase failed to show the formation of an isoindole derivative when in- cubated with o-phthalaldehyde (Fig. 4C).

There are four thiols (designated I, If, Ill and IV) in the monomeric hexokinase [32]. Mannose was earlier shown to protect hexokinase from in- activation by o-phthalaldehyde (Table I). Man-

nose was also found to protect the essential thiols of hexokinase from inactivation by iodoacetamide or DTNB (experiments D and G, Table III). Of interest was the finding that chemical modificao tion of hexokinase by iodoacetamide in the pres- ence of mannose resulted in alkylation of about six thiols per dimer without significant loss of the phosphotransferase activity; subsequent treatment of the iodoacetamide modified hexokinase with o-phthalaldehyde resulted in the modification of two more thiols with complet~ loss of the enzyme activity (experiment H, Table III). Incubation of hexokinase with 2-(N-bromoacetyl)-D-galactos- amine, an affinity reagent, under the described conditions, resulted in the modification of only thiol I and complete loss of the phosphotrans- ferase activity [2]. These experiments clearly established the fact that only thiol I is critical for the catalytic function of hexokinase. Our results show that the chemical modification of hexokinase by DTNB at or below 30°C caused complete inactivation of the enzyme in about 2 h with concomitant modification of the two thiols per enzyme dimer. It is, therefore, likely that it is the thiol I that was modified during short time in- cubation of hexokinase with DTNB. Results pre- sented in Table III (experiment H) lead us to the same conclusion for the modification reaction in- volving o-phthalaldehyde. This was further con- firmed by the experiment in which the iodacetam- ide modified hexokinase (experiment E, Table Ill) failed to react with o-phthalaldehyde as evidenced from the absence of a fluorescence emission maxi- mum at 403 nm (Fig. 4A).

The time-dependent inactivation of yeast hexokinase by PLP showed biphasic kinetics (Ta- ble IV). It required fairly high concentrations of PLP, i.e., 10-20 mM, to inactivate hexokinase. Even after 20 h of incubation with PLP, 15~ residual phosphotransferase activity was detecta- ble. Frerman et al. [49] pointed out that the per- sistence of residual activity, even at apparently saturating concentrations of PLP, is a universal feature observed in the modification of enzymes by PLP. These results further suggest that these lysines are probably buried.

The molar transition energy of the |texokin- ase-o-phthalaldehyde adduct was found to be 116 k J/ tool and compares favorably with that of EA

45

in hexane (127 kJ/mol). These results indicate that cysteine and lysine residues participating in the isoindole formation are located in a hydro- phobic environment [cf. Ref. 9-11]. This is in agreement with the earlier results of the chemical modification studies that show that the cysteine and iysine residues participating in the isoindole formation are buried in a hydrophobic cleft of the hexokinase. It is interesting to note that Bennett and Steitz [48] showed that a hydrophobic effect alone favors the active conformation of the hexokinase in the absence or presence of glucose.

In conclusion, the inactivation of yeast hexo- kinase by o-phthalaldehyde (a) proceeds rapidly and irreversibly through the formation of an isoin- dole derivative, (b) the second-order-rate (K = 45 M-1 . s - l ) of inactivation does not follow Micha- elis-Menten kinetics, and (c) the cysteine and lysine residues involved in the isoindole derivative for- mation are located in the hydrophobic environ- ment of the active site. An important finding of this investigation is that there is a lysine residue at or near the active site, and it is not readily accessi- ble in the holoenzyme.

Acknowledgements

This work was supported by US. Public Health Service grant NS-15994 (R.R.) and Fe!lowships (R.N.P. and D.B.) from the American H~z~rt As- sociation-Louisiana. We thank Ms. Gail Daniels and Ms. Cathy Spiotta for typing the manuscript.

References

1 Rustum, Y.M., Marsaro, E.T. and Barnard, E.A. (1971) Biochemistry 10, 3509-3516.

20 t i eno , S., Bhargava, A.K., Serelis, D. and Barnard, E,A. (1977) Biochemistry 16, 4249-4255.

3 Gray, A.J.W.. Barton, N.M. and Barnard. E.A, (1983) Biosci. Rep. 3, 963-971.

4 Menezes, L.C. and Pudles. C. (1976) Biochimie 58, 51-58. 5 Anderson, C.M., McDonald, R.C. and Steitz. T.A. (1978) J.

Mol. Biol. 123, 1-13. 6 Simons, S.S.0 Jr. and Johnson. D.F. (1978) J. Org. Chem.

43, 2886-2891. 7 Sirnons, S.S.. Jl ~. and Johnson, D.F. (]978) Anal Bi~hem.

90. 705-725. 8 Simons, S.S,, Jr,, Thompson, E.B. and Johnson, D.F. (1979)

Biochemistry 18, 4915-4922.

46

9 Puff, R., Bhatnagar, D. and Roskoski, R., Jr. (1985) Bio- chemistry 24. 6499-6508.

10 Puri, R., Bhatnagar, D., Glass, D.B. and Roskoski, R., Jr. (1985) Biochemistry 24, 6508-6514.

11 Puri, R, and Roskoski, R., Jr. (1988) Biochem. Biophys. Res. Commun, 150, 1088-1095.

12 Puri, R. and Roskoski, R., Jr. (1988) Anal. Biochem. (in press).

13 Purl. R., Bhatnagar, D. and Roskoski, R., Jr. (1984) 40th Southwest ACS Regional Meeting, Lubbock, Texas, p. 36, American Chemical Society.

14 Puri, R., Bhamagar, D. and Roskoski, R., Jr. (1985) Bio- chemistry 24, 3360,

15 Lowry, O.H,, Rosebrough, N.J., Farr, A,L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-235.

16 Hoggett, J.G. and Kellett, G.L. (1976) Eur. J. Biochem, 66, 65-77.

17 Dela Fuente, G. and Sols, A. (1970) Eur. J. Biochem. 16, 234-239.

18 Hohnadel, D.C. and Cooper, C. (1972) Eur. J. Biochem, 31, 180-185.

19 Danenberg, K.D, and Cleland, W.W. (1975) Biochemistry 14, 28-39.

20 Cornelius, R.D, and Cleland, W.W. (1978) Biochemistry 17, 3279-3286.

21 Schultze, I.T. and Colowick, S.P. (1969) J. Biol. Chem. 244, 2306-2316.

22 Derechin, M., Rustum, Y.M. and Barnard, E.A. (1972) Biochemistry 11, 1793-1797.

23 Schmidt, JJ. and Colowick, S.P. (1973) Arch. Biochem. Biophys. 158, 458-470.

24 Maley, F, and Lardy, H.A. (1955) J. Biol. Chem. 2154, 765-772,

25 Sols, A., Dela Fuente, G., Villar-Palasi, C. and Asensio, C. (1958) Biochim. Biophys, Acta 30, 92-101.

26 Slein, M.W., Cori, G.T, and Cori, C.F. (1950) J. Biol. Chem, 186, 763-780.

27 Purieh, D.L. and Fromm. H.J, (1973) Adv. Enzymol. 39, 249-326,

28 Sols, A, and Crane, R.K. (1954) J. Biol, Chem. 210, 581-596. 29 Crane, R,K, (1960) in The Enzymes, Vol. 6, 2nd Edn, p, 47,

Academic Press, New York,

30 Poulose, A.J. and Kolattukudy, P.E. (1980) Arch. Biochem. Biophys. 201,313-321.

31 Dimroth, K., Reichardt, C., Srepmann, T. and Bohlmann, F. (i963) Ann. Chim. 66, 1-37.

32 Jones, J.G., Otieno, S., Barnard, E.A. and Bhargava, A.K. (1975) Biochemistry 14, 2396-2403.

33 Dela Fuente, G, and Sols, A. (1963) Biochem. J. 89, 36P. 34 Steitz, T.A., Anderson, W.F., Fletterick, R.J. and Ander-

son, C.M. (1977) J. Biol. Chem. 252, 4496-4500. 35 Anderson, C.M., Stenkamp, R.E., McDonald, R.C. and

Steitz, T.A. (1985) J. Mol. Biol. 123, 207-219. 36 McDonald, R.C., Steitz, T.A. and Engelman, D.M. (1979)

Biochemistry 18, 338-342. 37 Jarori, G.K., Kasturi, S.R. and Keukare, V.W. (1981) Arch.

Biochem. Biophys. 211, 258-268. 38 Bhatnagar, D., Roskoski, R., Jr., Rosendail, M.S. and

Leonard, NJ. (1983) Biochemistry 22, 6310-6317. 39 Bhatnagar, D,, Glass. D.B., Roskoski, R., Jr., Lessor, R.

and Leonard, N.J. (1985) Biochemistry 24, 1122-1129. 40 Dorgan, L.J. and Schuster, S.M. (1981) Arch. Biochem.

Biophys. 207, 165-174. 41 Bhatnagar, D., Hartl, F.T., Roskoski, R., Jr. Lessor, R.A.

and Leonard, N.J. (1984) Biochemistry 23, 4350-4357. 42 Shoham, M. and Steitz, T.A. (1980) J. Mol. Biol. 140, 1-14. 43 Steitz, T.A., Shoham, M. and Bennett, W.S., Jr. (1981) Phil.

Trans. R. Soc. Lond. B293, 43-52. 44 Bhatnagar, D., Cromwell, W.C. and Roskoski, R., Jr. (1985)

Fed. Proc., Fed. Am. Soc. Exp. Biol. 44, 1812. 45 Melchior, N.C. and Melchior, J.B. (1958) J. Biol. Chem.

231,609-623. 46 Hammes, G.G. and Kocharu, D. (1962) J. Am. Chem. Soc.

84, 2073-2076. 47 Noat, G. and Richard, J. (1968) Eur. J. Biochem. 5, 71-72. 48 Bennett, W.S., Jr. and Steitz, T.A. (1978) Proc. Natl. Acad.

Sci. USA 75, 4848-4852. 49 Frerman, F.E., Andreone, P. and Nielke, D. (1977) Arch.

Biochem. Biophys. 181, 508-515. 50 Elhnan, G.L. (1959) Arch. Biochem. Biophys. 82, 70-77. 51 Sober, H.A. (ed.) (1973) Handbook of Biochemistry, p.

K-60, CRC Press, Cleveland, OH. 52 Goldin, B.R. and Frieden, C. (1972) J. Biol. Chem. 247,

2139-2144.