BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 239, 789–793 (1997)ARTICLE NO. RC977558

Site and Significance of Cysteine Residues in XyloseReductase from Neurospora crassa as Deduced byFluorescence Studies

Urmila Rawat and Mala Rao1

Division of Biochemical Sciences, National Chemical Laboratory, Pune 411008, India

Received September 2, 1997

N. crassa catalyses the conversion of D-xylose to xylitol,Inactivation of xylose reductase (XR) by p-hydroxy- initial crucial reaction in the major pathway for the

mercury benzoate (PHMB) was found to be biphasic conversion of xylose to ethanol (3). Recently, the en-with second-order rate constants of 80 and 6 M01s01 for zyme has gained importance due to its application inthe fast (kf) and slow (ks) phase respectively. Spectro- the synthesis of xylitol an acariogenic noncaloricscopic studies indicated that the inactivation was due sweetner used in many food products (4). Despite theto modification of one Cys residue per molecule of XR

biotechnological application of XR virtually nothing isand not due to subsequent disruption of the quater-known about the amino acid residues constituting thenary structure. The binding of NADPH to XR (Kd 0.9active site or participating in the catalytic process. WemM) was depressed on modification of the enzyme byhave purified and characterized XR from N. crassa andPHMB (Kd 2.3 mM). The dependence of PHMB inducedstudied the involvement of a tryptophan residue at theinactivation of XR in the presence of alcohols and vary-low affinity NADPH binding site (5). The role of aning temperature revealed that the Cys residue is situ-essential lysine at the active site of XR and the confor-ated in a hydrophobic microenvironment and is notmation and microenvironment of this site has also beeninvolved in hydrogen bonding. Our present investiga-assessed (6). Present article is the first report on thetion using o-phthalaldehyde (OPTA) as the chemical

initiator for fluorescent chemo-affinity labeling and differential location of cysteine residues and their con-double inhibition studies indicates that Cys residues tribution to the structure and function of XR an enzymeinvolved in the reaction with PHMB (SHI) and OPTA belonging to the important class of oxido-reductases.(SHII) are distinctly different. Experimental evidencepresented here serves to implicate that SHI located in MATERIALS AND METHODSa hydrophobic microenvironment at the high affinityNADPH binding site of XR plays a role in the binding

Isolation and purification of XR from N. crassa. XR was purifiedof the coenzyme to XR, whereas SHII serves to maintain and assayed as described earlier (5). Enzyme units are defined asthe conformation of the active site essential for cataly- mmoles of NADPH oxidized per min. Protein was determined by thesis by interacting with the NH2 group of an essential method of Bradford (7) with bovine serum albumin as the standard.

The molar concentration of XR was calculated assuming a Mr oflysine residue. q 1997 Academic Press60 000 (5).

Inactivation kinetics of XR by PHMB. XR (20 mg) was incubatedwith varying concentrations of PHMB in 50 mM phosphate buffer,pH 7.0 at 257C and assayed for XR activity at different time intervals.In view of the rapid depletion of fossil resources, the Control tubes having only enzyme or only inhibitor or inhibitor and

efficient utilization of hemicellulose, the major compo- substrate were incubated under identical conditions. Protection stud-ies were performed by incubating the enzyme with the substratesnent of plant biomass, is necessary for the generationNADPH or xylose for 10 min prior to the addition of the modifier.of chemicals and fuels. However, D-xylose, the predom-The inactivation data was analyzed according to (8). Pseudo first-inant sugar of hemicellulose, is not readily utilized fororder rate constants for the fast (kf) and slow phase (ks) of inactiva-

the production of ethanol by micro-organisms (1). We tion were obtained by plotting ln a or ln [a-C2 exp (0kst)] versus timehave identified Neurospora crassa NCIM 870 a novel (t). The second-order rate constants for the fast and slow phase were

obtained from slope of the plots of kf and ks against varying concentra-organism with the ability to directly convert biomasstions of PHMB.to ethanol (2). Xylose reductase (XR; EC 1.1.1.21) from

PHMB titration of XR. XR (1.5 1 1005 M) in 50 mM phosphatebuffer, pH 7.0 was titrated with 10 ml volumes of PHMB (1 mM) andthe progress of the reaction was monitored spectrophotometrically1 To whom correspondence should be addressed.

0006-291X/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

789

AID BBRC 7558 / 693c$$$561 10-21-97 10:12:42 bbrcg AP: BBRC

Vol. 239, No. 3, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ment with excess b-mercaptoethanol (b-ME) (data notshown).

Monitoring conformational change. The inactiva-tion of XR by PHMB did not alter the fluorescence pro-file (lex 295 nm; lem 326 nm) indicating no shift of tryp-tophan residues from polar to nonpolar environmentand vice versa (Fig. 3A). The CD measurements re-vealed no effect on the a-helix and b-sheet content ofXR (Fig. 3B). Hence the PHMB induced inactivation ofXR is a result of direct chemical modification of anessential Cys and cannot be attributed to subsequentdisruption of the quaternary structure.

Substrate protection studies. Xylose (300mM) af-forded no protection against inactivation by PHMB.This may be due to the fact that the enzyme followsan iso-ordered bi bi mechanism (5). NADPH (20 mM)protected maximum 90% of XR activity. Thus the inac-

FIG. 1. (A) Plot of ln a versus t (j) for inactivation of XR (20 mg) tivation may be due to modification of Cys residue (SHI)by PHMB (60 mM). The values for the amplitude (C2) and pseudosituated at or near the NADPH binding site of XR.first-order rate constant for the slow phase (ks) of inactivation were

obtained from the vertical intercept and slope of this plot. (B) Plot NADPH binding studies. Native and PHMB modi-of (kf) (l) and ks (s) versus varying concentrations of PHMB. Second- fied XR were titrated with varying concentrations oforder rate constant for the fast and slow phase of inactivation were

NADPH and the formation of E-coenzyme complex wascalculated from the slope of these plots. The values for ks were ob-studied by monitoring the quenching of the XR fluores-tained as described in materials and methods.cence (lex 295 nm; lem 326 nm) (Fig. 4). The dissociationconstant for XR-NADPH complex was calculated andvalues of 0.9 mM and 2.3 mM were obtained for theat 250 nm. Simultaneously aliquots of the reaction mixture werenative and modified XR respectively (Fig. 4 inset). Thuswithdrawn to assay the residual activity. The number of Cys residues

modified were determined by the method of (9). the inactivation was accompanied by a concomitant de-crease in the ability of XR to bind the phosphorylatedFluorescence and circular dichroism measurements. Fluores-

cence and CD spectras were recorded at 257C using a AMINCO SPF- coenzyme.500 spectrofluorometer and JASCO J600 model spectropolarimeter.

Microenvironment of (SHI) in XR. The inactivationThe dissociation constant of E-NADPH complex was calculated byof XR by PHMB in dilute aqueous solutions of variousmonitoring the tryptophanyl fluorescence of XR at varying concentra-

tions of the coenzyme (10).

Effect of reagents containing SH and NH2 groups to terminate thereaction between XR and OPTA. XR (1.8 mM) was incubated with15 ml of 10 mM OPTA in 50 mM phosphate buffer, pH 7.0 at 257Cfor 5 min at which time 3 mM each of the reagent was added to thereaction mixture and the residual activity measured.

RESULTS AND DISCUSSION

Inactivation of XR by PHMB. Inactivation of XRby varying concentrations of PHMB was found to bebiphasic. Fig. 1A shows a plot of lna versus time forinactivation of XR at a fixed concentration of PHMB.Applying the analysis described earlier (8) the pseudofirst-order rate constant values of 5.8 1 1003 and 4.151 1004 s01 and second order-rate constant values of 80and 6 M01 s01 were obtained for the fast (kf) and slow(ks) phases of inactivation (Fig. 1B). Values of 0.27 and0.72 were obtained for the amplitude of fast (C1) andslow phase (C2) of inactivation. Plot of residual activityagainst number of Cys residues modified (Fig. 2) re-vealed that the inactivation was a result of modificationof 0.90 Cys residues per molecule of XR. PHMB induced FIG. 2. Plot of residual activity against number of Cys residues

modified by PHMB per mole of XR.inactivation of XR was found to be reversible by treat-

790

AID BBRC 7558 / 693c$$$562 10-21-97 10:12:42 bbrcg AP: BBRC

Vol. 239, No. 3, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 4. Native (l) PHMB (s) and NEM modified (n) XR each(0.15 mM) in 50 mM phosphate buffer, pH 6.3, were titrated withvarying concentrations of NADPH and the changes in the fluores-cence spectra recorded (lex 295 nm; lem 326 nm). (inset) Plot of 1/1-a against NADPH/a for the native (N) and PHMB modified XR (M),reciprocal slope of these plots yielded the value of Kd for XR-NADPHcomplex. a is the ratio of fluorescence reduction at a specified NADPHconcentration and limiting amount of fluorescence at fully saturatingcoenzyme concentration.

each alcohol decreases (Table 1). Addition of alcohol toaqueous solution results in the decrease in the dielec-tric constant. This would cause an increase in thestrength of electrostatic interactions and hydrogenbonds, the strength of the later is increased by thereduced competition of water molecules. An increase inXR inactivation in presence of alcohols reveals thatneither of these types of bonding are responsible forthe reduced reactivity of XR. With the increase in thelength of the carbon chain the alcohols become morehydrophobic. It was evident from our results that thevarious alcohols act by reducing the hydrophobic inter-

TABLE 1

Dependence of PHMB-Induced Inactivation of XR in thePresence of Dilute Aqueous Solutions of Alcohols

Pseudo first-order rate constants

Alcohol kf 1 1003 s01 ks 1 1004 s01

FIG. 3. (A) Fluorescence (lex 295 nm; lem 326 nm) (B) far-UV CDcontrol 2.1 2.5(200–300 nm) spectra at XR concentrations of 0.5 and 1.5 mM.methanol (1.23 M) 5.7 2.6( ) and (---) represent the spectra of native and PHMB modifiedt-butanol (0.27 M) 5.74 2.65XR.ethanol (0.85 M) 6.5 3.16n-propanol (0.67 M) 6.68 6.2n-butanol (0.27 M) 6.82 7.1

alcohols was studied. The pseudo first-order rate con-XR (20 mg) in 50 mM phosphate buffer, pH 7.0 was incubated withstant for the fast (kf) and slow phase (ks) of inactivation

PHMB (20 mM) in the presence of various alcohols listed above. Theincreased in the order methanol, t-butanol, ethanol, n- extent of inactivation at different time intervals was measured andpropanol and n-butanol, despite the fact that with an the values of pseudo first-order and second-order rate constants were

calculated as described earlier.increase in chain length the molar concentration of

791

AID BBRC 7558 / 693c$$$562 10-21-97 10:12:42 bbrcg AP: BBRC

Vol. 239, No. 3, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABLE 2 (Fig. 5). Chemical modification of XR with 5 mM N-ethylmaleimide (NEM) (2 h) a Cys modifier, at pH 7.5Effect of b-ME and Cysteine on the XR Inactivated

by PHMB and OPTA resulted in inactivation of XR. Addition of NADPH toNEM-modified XR resulted in the quenching of XR flu-

Inhibitor Cys / b-ME Residual activity (%) orescence indicating that the inactivation did not alterthe formation of XR-NADPH complex (Fig. 4). How-Control 0 100ever, the modified enzyme failed to form the XR-isoin-PHMB 0 21dole derivative on reaction with OPTA (Fig. 5) indicat-PHMB / 74

OPTA 0 22 ing that the Cys residue involved in the reaction withOPTA / 4 NEM and OPTA is the same (SHII). These results impli-PHMB r OPTA 0 2 cate that SHII plays no role in the binding of NADPHPHMB r OPTA / 2

to XR but its integrity is essential for the catalyticXR (0.16 mM) was incubated with PHMB (50 mM) and OPTA (50 mechanism of the enzyme.

mM) each for 5 min. The reaction mixture was further incubated withProximity of cysteine and lysine residues in XR-isoin-1 mM Cys and 5 mM b-ME and the enzyme activity measured.

dole derivative. The XR-isoindole derivative formedwas stable for 24 h under the experimental conditions.This is probably due to the ease of formation of anactions to varying extents which otherwise hinder theisoindole ring system, from the benzenoid ring of OPTAreaction of SHI. The rate of inactivation of XR byand not vice-versa. Several reagents that contain SHPHMB was greater in n-butanol than t-butanol. Bothand NH2 functions in varying distance were examinedthese isomers have same dielectric constant and molec-for their ability to terminate the reaction between XRular weight, however, t-butanol is completely miscibleand OPTA. On addition of excess cysteine a residualin water and hence disrupts hydrophobic bonds to aactivity of 68% was obtained. This indicates that cys-lesser extent than n-butanol resulting in the differen-teine effectively terminated the reaction and the prox-tial rate of inactivation. The temperature dependenceimity of SH and NH2 functions of the participating Cysof hydrophobic interactions in proteins has been stud-and Lys residues in the XR molecule and that of cys-ied earlier by Baldwin (11). Maximum stabilization ofteine are comparable to that between the two -CHOthese interactions is observed at higher temperaturesfunctions in OPTA. Based on the bond distances andand as the temperature is decreased the interactionsbond angles of the two -CHO functions in OPTA (13)weaken. In the present investigation it was observedour studies indicate that the SH and NH2 functions inthat the pseudo first-order rate constants values forXR involved in the formation of XR-isoindole derivativethe fast (ks) and slow phase (kf) of PHMB induced inac-and cysteine are probably 3-rA apart. Homocysteinetivation of XR at 187C were 1.1 and 1.4 fold higher than

that observed at 257C. Thus the decrease in tempera-ture was accompanied by a concomitant increase in theavailability of SHI for reaction with PHMB. Altogetherthese results support the notion that the microenviron-ment of SHI in XR is hydrophobic.

Double inhibition studies. o-phthalaldehyde (OPTA)is a bifunctional agent with absolute specificity for SHand NH2 groups (12). XR reacts with OPTA resultingin the formation of a fluorescent XR-isoindole deriva-tive (6). Studies were performed to ascertain whetherreactive Cys binding to PHMB is also involved in OPTAreaction. Two sets of experiments were performedwherein the enzyme was treated with PHMB prior toOPTA modification or vice versa followed by cysteine,a terminator of OPTA reaction and b-ME that reversesthe inhibition due to PHMB. As shown in Table 2 theenzyme was inactivated in both cases indicating thatthe Cys residues reacting with PHMB (SHI) and OPTA(SHII) are distinctly different. If they were same thenreactivation would have been observed in the casewhere the enzyme was modified with PHMB prior toOPTA. This was further confirmed by the observation FIG. 5. Isoindole fluorescence (lex 338 nm; lem 410 nm) of nativethat XR modified by PHMB retained its ability to form ( ), PHMB (---) and NEM inactivated (-r-) XR each 0.5 mM on

reaction with 20 mM OPTA.the XR-isoindole derivative on reaction with OPTA

792

AID BBRC 7558 / 693c$$$562 10-21-97 10:12:42 bbrcg AP: BBRC

Vol. 239, No. 3, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

and glutathione were not efficient terminators with a with the NH2 group of this residue situated at a dis-tance of 3-rA and thus serves to maintain the conforma-residual activity of 35%. This can be explained by the

fact that the SH and NH2 functions in these reagents tion of the active site essential for catalysis. The micro-environment of the active site of XR which binds thedo not assume the 3-rA distance with respect to each

other as in cysteine. Glycine and Cystine completely XR-isoindole derivative is nonpolar in nature (6) sug-gesting that SHII is located in a hydrophobic environ-lack the ability to terminate the reaction because of the

absence of a NH2 and a free SH function respectively. ment. A detailed knowledge of the active site of XRthat imparts the specificity for NADPH/NADH will beCrystallographic studies will further confirm the dis-

tance of 3-rA between the essential SH and NH2 func- useful in redesigning a protein with high NADH-linkedactivity crucial for alcohol production.tional groups of Cys and Lys residues in XR.

Cysteine residues are known to play a crucial role inenzyme catalysis. Investigations that shed light on REFERENCEStheir location are of importance in the crystallographic

1. Jeffries, T. W. (1983) Adv. Biochem. Eng./Biotechnol. 27, 1–32.context of locating binding sites for Hg atoms in heavy2. Mishra, C., Keskar, R., and Rao, M. (1984) Appl. Environ. Micro-metal derivatives. Our present investigation using flu-

biol. 48, 224–228.orescence studies reveals the presence of two essential3. Rawat, U. B., Bodhe, A., Deshpande, V., and Rao, M. (1993) Bio-Cys residues in XR from N. crassa. Earlier we have

technol. Lett. 15, 1173–1178.reported the presence of a low (Kd 0.85 mM) and high

4. Bernd, N., Wilfried, N., Kulbe, D., and Klaus, D. (1996) Biotech-affinity (Kd 72 mM) NADPH binding site in XR from N. nol. Bioeng. 52, 387–396.crassa (5). In the present study it was observed that 5. Rawat, U. B., and Rao, M. (1996) Biochem. Biophys. Acta 1293,NADPH at a concentration of 20 mM afforded complete 222–230.protection against inactivation by PHMB. Considering 6. Rawat, U. B., and Rao, M. (1997) Eur. J. Biochem. 246, 344–

349.a Kd value of 0.85 mM for NADPH, 99% of the enzyme7. Bradford, M. N. (1976) Anal. Biochem. 72, 248–254.may be assumed to be in XR-NADPH form at this coen-8. Topham, C. M., and Dalziel, K. (1986) Eur. J. Biochem. 155, 87–zyme concentration. This indicates that SHI is located

94.at the high affinity coenzyme binding site of XR. Our9. Riordan, J. F., and Vallee, B. L. (1972) Methods Enzymol 25,results also reveal that this residue is involved in main-

449–456.taining the hydrophobicity of the NADPH binding site10. Stinson, R. A., and Holbrook, J. J. (1973) Biochem. J. 131, 719–and permitting its efficient binding to the enzyme. Us- 728.

ing the bifunctional reagent OPTA as the chemical ini- 11. Baldwin, R. L. (1986) Proc. Natl. Acad. Sci. USA 83, 8069–8072.tiator for fluorescent chemo-affinity labeling we have 12. Simons, S. S. Jr., Thompson, E. B., and Johnson, D. F. (1979)also identified a Cys residue (SHII) at the active site of Biochemistry 18, 4915–4922.XR. Earlier we have reported the role of an essential 13. March, J. (1977) in Advanced Organic Chemistry: Reactions

Mechanisms and Structure, pp. 22, McGraw–Hill, New York.Lys that reacts with OPTA (6). SHII probably interacts

793

AID BBRC 7558 / 693c$$$562 10-21-97 10:12:42 bbrcg AP: BBRC