Indian lou mal of Biochemistry & Biophysics Vol. 36, December 1999, pp. 433-441

UDP-galactose 4-epimerase from Kluyveromyces fragilis: Equilibrium unfolding studies

Nilesh Ranj an Maity, Bhaswati Bara! and Debasish Bhattacharyya*

Division of Protein Engineering, Indian Institute of Chemical Biology, 4, Raja S.c. Mullick Road, Ja lavpur. Ca lcutta 7000:12

Received 22 February 1999; revised 29 illlle 1999

UDP-galactose 4-epimerase from yeast CKluyverolllyces fragilis) is a homodimer of total molecular mass 150 kDa having possibl y one mole of NAD/dimer acting as a cofactor. The molecule could be dissociated and denatured hy ~ M urca at pH 7.0 and cou ld be fun ctionall y reconstituted after dilution with butTer having extraneous NAD. Thc unfolded and refolded equilibrium intermed iates of the enzyme between 0-8 M urea have been characteri zed in terms of catalyti c act ivi ty. NADH li ke characteristi c coenzyme fluorescence, interacti on with extrinsic flu orescence probe I-anilino ~- naphthe l cne

sui phonic acid CANS). far UV circular dichroism spectra, fluorescence emission spectra of aromatic residucs and suhunit dissociation. While denaturation monitored by parameters associated wi th active si te region e.g. inactivation and coenzyme fluorescence, were found to be cooperative having I'lG between -8.8 to -4.4 kcals/mole, the overal l dcnaturatioJl process in terms of second ary and terti ary structure was however continuous without having a transiti un pui nt. At 3 M L1rea a stabl e dimeri c apoenzyme was formed having 65% of native secondary struc ture which was dissociated to monomer at (1 M urea with 12% of the said structure. The unfolding and refolding pathways involved identica l structures except near the linal stage of refoldi ng where catalytic ac tivity reappeared .

UDP-galactose 4-ep imerase (hereafter ca lled epimerase) revers ibl y convert s UDP-galactose (UDP­gal) to UDP-glucose (UDP-glu) . It is an essenti a l enzyme for galactose metabolism and is ubiquitously present in a ll organisms . Thi s enzyme from yeast Kluyveromyces fragilis is a homod imer of molecular mass 75 kDaisubunit and is assoc iated with NAD by strong noncovalent binding. In contrast to c lassica l dehydrogenases (class I ox idoreductase) which uses NAD/NADH as cosubstrate, in epimerase (c lass II oxidoreductase) the nuc leotide acts as a true cofactor. It is a well studied enzyme particularly in terms of its 'oxidoreducti ve' type of mechani sm t

-, though many of its fundamental properties like the stoichiometry of the bound cofac tor has not yet been stud ied.

In the perspective of global interest on protein fo lding, we selected epi merase not onl y for its unique quarte rnary structure but a lso as it serves as a prototype model of c lass II ox idoreductases2

. Earlier conditions for revers ible foldin g of this enzyme from 8 M urea was estab li s hed~ and the distribution of population of its different unfolded states ha ve been analysed on the basis of refolding kine tics). Dissociation of the subunits and cofactor by parachloromercuribenzoate under nondenaturing condition and reconstituti on of the holoenzyme structure offered an inte resting system to be followed

• Author for correspondence: Fax: 9 1-D-47:l-5 197

for its cofactor binding6. Thi s mi ght have significance

on the maturation of the mo lecu le during refo lding as recruitment of NAD on apoenzyme was kn own to be rate limiting step in reactivat i o n ~ . Here we report characterization of equ ilibrium inte rmediates of epimerase formed in presence o f 0-8 M urea to e lucidate the sequence of events happe ning ell /'OllIe

denaturation and some energy ca lculati ons of unfolding.

Materials and Methods Urea (E. Merck, India, AR) was rec rys talli zed fro m

hot ethanol to remove possible contam in at ion of cyanate ions. It was dried to constant weight and stock solutions were made grav i metricall y. A II fine chemicals e.g. UDP-ga l, NAD. ANS ( I-anilino 8-naphthalene sulfonic acid) , EDT A, 2-mercaptoethano l (2-MCE), Sephadex G-SO and mo lecu lar we ight markers were from S igma (USA). Other laboratory reagents (analytical grade) were purc hased locally. UDP-g lucose dehyd rogenase was partia ll y purifi ed from beef li ver after Za liti s el ({f 7 upto the heat denaturat ion step. The preparation was left at storage buffer for 15 days when the background epimerase activity was lost.

Ellzyme purificalioll alld assay Epimerase was puri fied fro m yeast stra in

Kluyveromyces frag ilis (renamed as KfllvverolllYCeS

434 INDI AN 1. BIOCHEM , BIOPHYS, VOL. 36, DECEMBER 1999

marx ianus var. l17arx ianus, ATCC strain no, 10022 and Microbial Type Culture Collection, Institute of Microbial Technology, Chandigarh) after Darrow and Rodstrom8 and was extended upto a size-exclusion (SE)-HPLC step4, It was essentiall y a homogeneous preparation showing a single band in SDS-PAGE corresponding to molecular mass of 75 kDa, Enzyme activity was assayed by measuring conversion of UDP-gal to UDP-glu after coupling with UDP­glucose dehydrogenase and NAD and foll owing absorbance change at 340 nmx. Details of enzyme purification steps, specific acti vity, storage and assay conditions and protein determination were mentioned earlier4

.

Unfo lding and refolding

Epimerase was equilibrated with 0-8 M urea in 20 mM K-phosphate buffer, pH 8.0 between 16-20 hr at 25°C when the protein reached a stable conformation as judged by non variati on of signals with further incubation time in different ph ysical techniques like inacti vation, circular dichroism (CD) and fluorescence emission. To initiate refolding, 5 III ( 10-1 5 mgfml ) of epimerase was first treated with 8 M urea in phosphate buffer for 10 min at 25"C to ensure complete unfolding. Under such conditions 85% of the original secondary structure was lost with dissoc iation of the subunits and the cofactor. It was subsequently diluted with the same buffe r containing 0-7 M urea and 10 mM NAD for 16 hr to attend equilibrium refolded states. Addition of extraneous NAD was essential fo r reacti vati on. The reversible refolding process was free from aggregation as judged from over 90% recovery of activity and SE-HPLC of the refolded state. Native, denatured and fully refolded states of the enzy me ha ve been defined and characteri sed earlier4

. Inclusion of ANS in the refolding buffer did not change the refolding profile. Reactivation was complete and the kinetics of renatutation in presence and absence of 100 11M of ANS as judged by aromatic amino acid flu orescence quenching (ex: 280 nm; em: 335 nm) were indi stin­gui shable.

Spect roscopic measu remellt s

Optical measurements including enzy me assay were done either with a Beckman DU6 or Hi tachi 3200U recording spectrophotometer. Secondary structure of proteins were measured by CD spectroscopy between 2 10-250 nm using a J asco 720 spectropolarimeter. All flu orescence measurements

were done with a Hitac hi F4020 spectroflu orimeter. Interaction of ANS ( 100 11M) with di fferent protein conformers were monitored f1u orimetri ca ll y (ex: 375 nm; em: 400-550 nm). The emi ss ion spectrulll of buffers had low intensities and in no case corrected spectrum was recorded. All experiments were done at 25°C.

Size-exclllsioll -HPLC Elution profiles of epimerase pre-equil ibrated under

different denaturing condi tions were foll owed by passing through a LKB Ultropack TSK G3000SW SE-HPLC column (0,75x30 cm) in presence of 20 mM Na-phosphate buffe r, pH 7,{) at a fl ow ra te of 0.5 ml/min and at 280 nm.

Calculation of thermodynamic P ((JWl1 l' /crs

At equilibrium in presence of di ffe rent concentra­tions of urea foll owing denaturati on, I'll (frac ti on nati ve) was ca lculated using the relati on') :

f = (R - R ) f (R - R ) II e 1I II 1I

. , . (cq, I )

where Ru, Rn and Re were un fo lded (denatured enzyme) , nati ve enzy me and enzy me equilibrated in urea in terms of a phys ica l parameter like act ivit y. coenzyme flu orescence or dimeri c structure. In the enzyme acti vity ex periments the residual ac tivity between 6-8 hr in the pre-trans iti on region or the linear extrapolati on of the first order kinetics of inacti vation in the post transiti on region were employed to generate Re. For all equilibriu lll ex periments, the equilibrium constant KIJ and free energy of denaturati on (6CJ)) for the transiti on was calculated as fo ll ows:

KD= ( I-fn)ffn (eq, 2)

and 6 CD = -RT In K D ,. ' (eq, 3)

KD values within the range of 0. 1 to 10 were in the analysis of the transition, The limiti ng va lues of the transition were determined th rough regress ion analysis of the line given by:

, . , (eq, 4)

where , 6C ~ , free energy of denaturati on under

standard condition, 111 , the siope and [D], concentration of denaturant. Parameters from the

regressed line (i.e., 6C ~ and 111 ) were llsed to

genetate the model curve fo r the data (i .e. f" al gi ven values of D) given by:

6 CIO) - /11 [D] = - RT In [( I - f ) f f ]

II II '" (eq. 5)

MAlTY el al.: UDP-GALACTOSE 4-EPIMERASE FROM K. FRAGILIS 435

The model curve was then plotted with the raw data to examine the fit of the base line. Transition midpoint

was calculated where ~G ~ = 0 i,e, f" = I 12 or K D = I.

In the inactivation experiments the corresponding activation energy was derived from rate constant k, using the relation :

~Go =-RTln (khlkBT ) ... (eq.6)

where h, kB, R and T are the Planck, Boltzman and gas constants and absolute temperature respectively. The activation energy values were then extrapolated to zero urea concentration to obtain activation energy in absence of denaturant for the step. Here ~Go was assumed to be a linear function with respect to denaturant concentrati on.

Other methods

Proteins ( 100 Ill ) were separated from ligands or denaturant by loading onto a pre-spun (0.7x2.9 cm) Sephadex G-50 column (usuall y known as 'spin column') equilibrated with standard renaturing buffer and was eluted with low speed centrifugation 10. Recovery was between 90-95%. NAD bound to nati ve or parti ally denatured epimerase was es timated by pass ing the protein through a spin column to remove excess urea and presumably di ssociated NAD. The recovered protein was subsequentl y di ssociated from cofactor by 70% ethanol , dried and residual NAD, if any, was extracted with 0.2 M glycyl glycine buffer, pH 8.8. Finally NAD was estimated spectroflu oro­metrically after reacti on with NaOH (ex: 360 nm;em: 460 nm)" .

1.0 o

3 .5M / 3 .25M

20 Sec

60 5 Min

3M '\,

Results Denaturation ofepimerase at the (lcrive site

The active site of epimerase has been arbitraril y but convenientl y di vided into three subs ites to be followed separately-inacti vati on for catalyti c site; specific interac ti on of the f1u orophore ANS at the substrate binding site and its quenching by 5 '-UM P, a strong competiti ve inhibitor for the substrate UDP­gal'2. '4; and the very characteristi c NA DH li ke coenzyme fluorescence ari sin g from AD .. . HS­Cysteine charge transfer complex on the surface of the enzymel. 12

Epimerase remains full y acti ve upto I M urea at pH 7.0 at 25° C for 6 hr. Slow inacti vati on starts after interacti on with 1.25 M urea while the enzy me was completely inacti ve within 30 sec by 3.25 M urea. The .i nitial first order kineti cs of in acti vati ons ha ve been illustrated in Fig. I. Corresponding rate constants and the res idual acti vity at equilibrium in presence of 1-3.5 M urea have been presented in Tab le I . It shows that 5'-UMP can parti all y protect the enzy me from inacti vation, for example, 2 mM of the li gand reduces the rate by a fac tor of 3.4 in presence of 2 M urea indicating an induced stabili ty at the catalytic site. The ~Go of the respecti ve steps showed a I i near dependence with urea between 1.25-3 .0 M and the extrapolated ~Go(H20) was found to be 25.3 kcal s/mole (Fig. 2) . It may be menti oned that under such denaturing conditions loss of secondary and tertiary structures were very slow and the monophas ic dependence of the relati on suggests absence of interference of these factors with inacti va ti on.

1.5 M

2M

1.7SM

100 180 260 Min

Fig. \- Time course of inacti vation or epi merase by varying concentrations of urea. [Molar concentralion or Ihe dcn alUr~lI1l 1i ~ls heen represented by the number corresponding to each set of ki netics. Protei n conc. was 1.3."\ pM I

436 INDI AN 1. BIOCHEM . BIOPHYS , VOL. 36, DECEMBER 1999

Table I - Kineti c parameters associated with denaturati on with epimerase

Region where Detected by Solvent Observed Ii rst Re lllark s denaturation (M urea) order kineti cs was monitored (min -I)

Residual ac ti vi lY at cqu il ihri um ('Yr ,)

Catalyti c site Inac tivati on I 0 lJ ()· I ()() 1.25 0.003 4.')

1.5 0.006 I() 1.75 0.011 .')

2 0.01 7 ()

2+5 mM 5'· UMP 0.005 N.D. 2.5 0.078 ()

2.75 0.138 ()

3 0.23 N. D. 3.25/3.5 2.3 N. D.

Cofactor binding site Loss of coenzyme flu orescence

Dissociati on of subuni ts Size exclusion HPLC

N. D. : Not determined

Interaction of part ia ll y or full y denatured epimerase with ANS was foll owed first by treating the enzy me with 0-8 M urea for 6 hr and then with ANS . In cases where there was enhancement of fluorescence

intensity, effect of quenching by 0.5 mM 5'-UMP was checked . It was observed that partiall y unfo lded structures upto 4 M urea react w ith ANS. However the prote in-ligand complex modifi ed between 1.5--4 M urea cou ld not be repl aced by 5'-UMP. Since the enzy me becomes inac ti ve at thi s stage and A NS is known to reac t strongly with partiall y fo lded prote ins ' 5, it is like ly that these inte rac ti ons were nonspec ific. The results have been presented in Fi g. 3 .

The e mi ss ion spec trum of epimerase at equilibrium between 0-4 M urea afte r exc itati on at 353 nm have been illustrated in Fig. 4 _ It shows that between 2-3M urea the fluorescence was lost by 70% . To corre late the loss of coenzy me flu orescence with other events like inactivat ion , the kinetics of quenching at 3 M urea was measured which y ie lded a fi rs t order rate constant , k = 0 .28 min - I at 25°C (Fig. 4 . inset). Thi s is c lose to the kinetics of inacti vati on at 3 M urea, k=O.23 min· l

.

Dissociation of cofactor Loss of coenzy me fluorescence of epimerase

before atta ining comple te unfolding ra ises a question of dissoc iati on of the cofacLOr fro m parti a ll y unfo lded

3

3 4 5 6

0.28

Residu al dimel'ic Siru cture at equil ihriu lll ('Ir ,)

a I ()O 0.00 1 () O

0.006 () :I

0.17 ()

state. Alternati ve ly conformati ona l c hange a round the nucleotide may a lso lead to quenchin g as well. Therefore estimati on of bound cofactor to the enzy me molecule was done aft er incubati o n with ] M urea for I hr. NAD content o f the native and pa rtiall y fo lded e nzyme were found to be 0 .89 and O. I mo le/mo le o f dimeric prote in , indi cating an apoenzy me format ion.

Secondary struc f/lre Secondary struc tures o f nati ve. parti a ll y and full y

unfo lded epimerase by 0-8 M urea were monit ored hy CD spec troscopy aft e r equilibrati on (Fig . 5 ). Na ti ve epimerase shows maximum negative 1110 la r e ll ipti c it y at 222 nm typ ica l o f an ·-he li x containin g prote in. Therefore re lat ive loss o f secondary ~ tru c ture of the inte rmed iates were monitored at th at: wave length that shows a continuous patte rn o f un fo ldi ng (Fig. 6A ).

Tertimy structure

The tertiary struc ture o f the in te rmed iates were monitored from flu orescence e miss io n spectrum (ex :

280 nm). A continu o Ll s red shift o f e rni ss ion

max imum from 335-7350 nm and a d iscontinuous change of emi ss ion intens ity at 335 nm we re o bserved with progress of unfold ing (Fig .6 B and C). The unfolding profile of Fig. 6C d id no! c hange s ignificantl y if the intensities at the respective emiss ion max ima were cons ide red .

MAlTY el at.: UDP-GALACTOSE 4-EPIMERASE FROM K. FRAGILIS 437

2S

-.24 (lJ ....... o E

-....... III --'

d 23 u ..::s:. ........

22

21

, ,

o

\ , \ , ,

\ \ , , ,

\ \ , ,

2

Urea (M) 3

Fi g. 2- Dependence of energy of activation of the inactivation process of epimerase by urea.

Quarternary structure The elution profiles of epimerase equilibrated with

different concentrations of urea from a TSK G3000SW SE-HPLC co lumn have been shown in Fig. 7. The enzyme in native and unfolded states upto 3 M urea appeared as a single peak of retention time (Vt) = \3. \ ±0.05 min while unfolded between 6-8 M urea appeared at Vt= \5 . \ ±0.08 min . They were designated as 'N' and ' 0 ' respecti vely. These posttlOns corresponded to molecul ar mass of 158 kDa (dimeric state) and 69 kDa (monomeric state) respecti vely and were characterised in detail earli er4

. As illustrated, onl y two spec ies could be identified in the chromato­gram. Urea was excluded from the equilibrating buffer because of development of hi gh back pressure and also to enhance column life. It has been assumed that neither associat ion nor dissoc iation of the protein occurs during chromatography for the followin g reasons-urea was removed from the di ssociating mixture in the chromatography by 0.5 min as observed from the calibrati on of the column with

120 ,...------------,

:::> <l

~80 C/I c QJ

of­

C

QJ U C QJ U

~ 40 .... o ::J

LL

"Aell:375nm "Aem:450nm

OL...----L----'-----'----'-' 2 4

Ureo(M) 6 8

Fig. 3-Fluorescence enh ancement or ANS wi th nati ve and partially denatured epimerase during unfo ldi ng. (e ): and quenching by 5 '-U MP. (0). Thc same phenomcnoll duri ng refoldi ng have been rep re~en t ecl hy ( ... ) and (t,).

epimerase and low molecular weight UV absorbing material e.g. tryptophan ; at such low concentrati on of protein (approx imate ly I ~M) the rate of dimerization after removal of urea was ineffi cient (unpubli shed observation); in absence of the cofactor the rate of dimerization was even slower-l and finall y no association-dissociation type of profi les were observed J6

. The fi rst order kinet ics of dissociation and the percentage of res idual dimeric structure at equilibrium have been presented in Table I.

Cooperativity and Iherl11odynarnico/ IWUll/ l e la.\"

associated with denaturatio/l

The transiti on curves of unfolding of epimerase using the four parameters viz. inacti va ti on (Fig. I), coenzyme flu orescence (Fig.4), secondary structure (Fig. 6A) and dissociation of subunits (Fig. 7) aga inst f" (fracti on nat ive, eq. l) have been shown in Fig. 8A assuming a two state deri va ti on of unfolding. It shows that inactivat ion and quench ing of coenzyme flu orescence are cooperati ve in nature having transition midpoints at 1.3 and 2.5 M urea. Transi ti on of di ssociation also appea rs to be cooperati ve though not sharp and asymmetri ca l having a transition midpoint at 5.1 M. A wide difference was observed in

438 INDIAN J. BIOCHEM. BIOPHYS, VOL. 36, DECEMBER 1999

150 "Aex:353nm

..-..

~IOO <t --->----II)

c Q)

---C

Q) U C Q)

50 u II)

Q) L-0 :J

LL

OL--_~~ __ --'-__ ---' 400 450 500 55C

Wavelength A (nm)

Fig. 4--Coenzyme flu orescence of nati ve and part iall y denat red epimerase by varying concentration of urea. [Molar concentr:n ion of the denaturant has been rep resented by the number corresponding to each spectrum. B! and B2 denote spectrum of buffer in presence and absence of urea whil e N stands for the nati ve state. (In set): Kinetics of quenching of coenzyme flu orescence by 3 M urea that follows a tlrst ordcr rate constant, k = 0.28 min ' ! at 25"C. Protein concentration was 4 ~MJ

case of loss of secondary structure which fo ll owed a contineous pattern . Since none of the four transiti ons coincides with each other, it is obvious that denaturation was a non-cooperative process. Typical plots of the influence of denaturant concentrat ion on the change of free energy assoc iated wit h denaturation have been shown in Fig. 8B . Parameters derived from the linear plots in addition to the standard free energy of denaturation are the tran siti on midpoints (0 112) and the steepness or cooperativity indexed by the slope,

-111. These have been summari sed in Tabl e 2 .

Reversibility of unfolding Any reversible process may proceed us ing same or

1-0

E "D

C\J E u

Ol Q)

"D .........

t<)

'0

x ('< ..,

<D l-1

1

N /0

~---______ -1. __________ ~

200 250 "Anm

300

Fig. 5--CD measurement s of epilllerase eq uili hrat ed with I-X M urea. The respecti ve numbers denote denaturalll e()ll eell tr ~ll i()ll s

and N stands for naii ve state.

d iffe rent inte rmediates (c./". Carnots cyc le of the second law of thermodynamics). To c heck th is the inte rmedi ates at equilibrium during denaturati o n and renaturation back from 8 M urea ha ve been characterized in terms of molar e llip tic it y at 222 nm and fluorescence spectrum (ex: 280 nm : e m: :ns nm ) (Fig. 6 A-C) . Within limits of e rror it was observed that the renaturation process from 8 to 3 M urea involves intermediates that are identi cal with those observed in the forward process. A min or but signifi cant difference in structure occur o nl y near the final state of folding i.e .. be low 3 M ure,!.

Association of subunits and coF{lctor

In order to check whet he r partiall y re folded epimerase could form a d imeric holoe nzy me structure, the denatured prote in was re fo lded by 3 M

MAlTY el at.: UDP-GALACTOSE 4-EPIMERASE FROM K. FRACtLiS

urea and was passed th rough the SE-HPLC system as described earl ier. The reason for se lecting urea at 3 M was that the nati ve enzyme could wi thstand its dimeric structure under such condition . However dimerization was sluggish (20% dimer formed after 3 hr of refolding) and was independent of 0- \0 mM extraneous NAD. Under such condition incorporat ion

(lJ

0-C o .c v o -o

o ~

E c

(\J (\J (\J

ro <D L.J

E c c 0 Vl Vl

E (lJ ---x 0

E r<

(lJ-; v:) c . (lJ<t v-Vl >-(lJ -L VI o C ::J (lJ

LLc

100~---.----.----r---.

80

A 60

40

20

0

346

342 B

334

60

C 4 0

20 0 2 4 6 8

Urea (M) Fig. 6--Characterization of native, partially and fully denat ured and renatured epimerase. (A): Secondary structure monitored by CD spectroscbpy using mol ar ell ipticity at 222 nm. (8): Change of fluorescence emission maxi ma and (C): change of Iluorescence emission intensi ty at 335 nm of the conformers while exciting at 280 nm. [In all cases (e ) and (0 ) represent the process of unfold ing nd refolding respectivelyl

0 .03 A 280nm

O' 10 Elurion rime (min)

o

N

'---~6

3

20

Fig. 7-Elution profi les of nati ve, partia ll y and full y dissoc iated epimerase equilibrated by urea through SE- HPLC. IN and D represent el ution posit ions of nati ve climeri c and folded monomeric epimerase]

of NAD was a lso insigni ficant-O.OS mole/di mer. Cofactor incorporati on was marginall y improved to 0 . 1 mole/dimer when urea concentrati on was reduced to J M.

Reversibility offolding at Ihe ((clive sile r eg ion

To identify the sequence in whi ch the ac ti ve site region was formed during folding , specific properti es as described earlier were tested. Effect ive reac ti vat ion permits maximum denaturant concentrati on upto 0.4 M, even then it takes 2 hr to express maximum activity (k=J.48x I0 :; M·t S·I at 25°C)~·~ . Reacti va tion under I M urea becomes comp licated because of very s low reactivation that enters into competitio n with stability factors. Also results were not reproducible . Secondly in none of the react ivat ion experiments the coenzyme fluorescence could be recovered. in contrast partia lly refolded epimerase reacts distinctly with ANS .

Epimerase refolded at 0-3 M urea reacts with ANS havi ng characteristic flu orescence intensity and bl ue

shi ft of emi ss ion maxima. However 5'-UMP can replace ANS unless the denaturant concentrati on was kept be low 0.5 M (50% quenc hing by 5 mM MP) (Fig.3). This spec ific rep lacement can take place within 5 min of refolding indicating th at the substrate binding subsite is formed in the refo lded inte rmediate much ahead of reactivati on.

Discussion

The act ive site of epimerase from KII/v vemlllvces

fragi lis has so far been proposed to const itute by sharing of two subunits and the cofactor AD I.6. I ~ . The resu lts from equilibrium denaturation studies be tween 0-8 M urea as prese nted here show that the

440 INDIAN 1. BIOCHEM. BIOPHYS. VOL. 36, DECEM BER 1999

A

c '+--<lJ > +-0 z c 0.5 0

:;:. u 0 L.

LL

-/°r-B 8\

\ \ \ ,..... \

0

E 6 .........

0 u :-:s:: 4

0 <:) <J 2

+

o

2 4 6 8 2

o 2 4

Urea (M) Fig. H A): Urea induced unfolding parameters of epimerase. Denaturati on was foll owed by inac ti va tio n. (I); coenzY llle llu<Jrescence. (2) ; retent ion of dimeri c structure, (3) and molar elli pticit y, (4). (B): Influence of urea concentration on the change 111 i'1-ee energy or denaturation fo llowing enzyme activ ity, ( I) and coenzyme Iluorescence, (2).

Table 2-Thermodynamical parameters associated with denaturatio n with epimerase

[(0112) "1'1' is the transition midpoint; - Il l""" is the cooperat ivity index related with the steepness of the transit io n prorile; L'.G"""" is the apparent change in free energy of denat urat ion undcr standard condit ions)

Parameter

(Old "pp (M, urea) -//lapp (kcal. mol. M) L'.G"app (kcal/mole)

Acti vit y data

1.3 5.67 -8 .8

Coenzyme Iluorescence data

2.5 1.9

-4.6

transition of inact ivation, quenching of coenzyme fluorescence poss ibl y assoc iated with di ssoc iati on of NAD and subunit dissociation are di fferent (Fig. 8A) . A 0-8 M transverse urea gradi ent gel electrophoresis of epimerase shows a sharp transition at 3.8 M urea apparentl y due to subunit dissoc iation (B.Barat and D. Bhatlacharyya, man uscript communicated). These suggest that the catalytic site, coenzyme binding site and subunit contact regions are distinct though overlappings are poss ible. Small single domain proteins e.g. chicken egg whi te lysozy me, ribonuclease and • -l actalbumin show high degree of cooperati vity upon unfolding induced by 0-8 M urea or 0-6 M guanidi ne, HCI as observed by CD or

fluorescence spectroscopy or urea grad ierl t gel I I . ? 17- 11) CI ,. AC l' I e ectrop lOreSIS . . lange 0 - il I II or t lese

transiti ons vary betweeen -5 to - 10 kca ls/mole. Ln case of epimerase tJ.GD for inact iva ti on and quenching of coenzyme flu orescence were found to be -8.8 and - 4.4 kcals/mole respecti ve ly (Fig. 8B). Thus these two sites may ari se from unfoldi ng of small do mains likely to be overlapped. The inacti vati on process has a relati vely hi gh energy of act iva ti on of 25.3 kca ls/mole indicating its stability. The .6CD of the react ivation process of the same enzy me from 8 M urea is 15.!. kcal s/mole4

.

In contrast to the acti ve site reg ion. ep imerm;e does not show sharp transition of denaturat ion in terms of secondary and terti ary structures (Fig. 6 A-C). Deviati on from the two state behaviour (nati ve H

denatured) of unfolding have been reported for other prote ins of comparable size e.g. puramyos in (46 kDa) by guanidine, HCI20, 3,4-dih ydroxyp henyl al anine (DOPA) decarboxy lase (d imeI' , 54 kDa/subu nit) by

J I » urea- , phosphorylase B (92 kDa ) and by urea-- . Complex denaturation pattern of epi merase IS

expected to be ari sing from its NAD bound di1ller ic structure of moderately bu lky ~; I ze suitable of'

MAlTY et al.: UDP-GALACTOSE 4-EPIMERASE FROM K. FRAGILIS 44 1

multidomain structure . It has al so been demonstrated that deviation from perfect reversibility of the process of unfolding as judged by physical characters of the intermediates starts near the final state of folding i.e ., below 3 M urea (Fig. 6). The overall findings are consistent with the idea that for a large protein to be refolded properly, major process of folding occurs at an early stage followed by minor structural changes at the end23

-24

.

Based on the equilibrium studies reported here, the whole denaturation process could be devided into three phases: (a) , in the first phase between 0-3 M urea, the molecule becomes inacti ve, the substrate binding site is modified and the cofactor NAD is detatched off. At the end of this phase about 65 % of the secondary structure remains ; (b), in the second phase between 3-6 M urea further denaturation occurs as a result of which 30% of the secondary structure remains and the subunits are rapidly dissociated and (c), in the third phase between 6-8 M urea the enzyme attains an unstable structure which is completely abolished by 8 M urea. This can be summarised and compared with the refolding process4 as follows. N--e--N ~ I - I + e ~ 2 U I ~ 2 U (unfolding)

N--e--N f- h-o-h f- II - II f-2 II f- 2 U (refolding) where N--e--N, native dimeric holoenzyme; e ,

cofactor NAD; I - I, partially denatured intermediate (apoenzyme) at 3 M urea; U I , di ssociated intermediate (monomer) at 6 M urea; U, complete ly denatured monomer; II , refolded monomer; II-II , refolded dimer; h-o-h, refolded dimer (apoenzyme) having NAD binding site and 0 , cofactor binding site.

Acknowledgement NRM and BB were supported by CSIR-NET

fellowships. Research was partially assisted by a DST grant (SP/SOID-45/93 ) awarded to DB.

References I Gabriel 0 , Kalckar H M & Darrow R A ( 1075) in SlIlJ/{//il

En zymes: Biochemistr.\' and FlInction (Ebner K B cd) pp X5-134, Marcel Dekker, Inc. Ncw York

2 Frey P A (1987) in Pyridine NlIcleotide Coen ~rlll es: Ch(,lIIiml biochemical and medical aspects (Dolph in D. Poul son R & Avannovie 0 eds) vol 2B , pp 462-447, Wi Icy. New York

3 Frey P A ( 1996) FASEB J 10.461-470 4 Bhattacharyya D ( 1993) Biochelllistrv 32. 0726-07:14 5 Dutta S, Maity N R & Bhallacharyya D ( 10<)7 ) Biochilll

Biophys Acta 1343,251-262 6 Majumdar S. Bhallachaljee H, Bhallacharyya D & Bhad uri A

(199 8) EliI' J Biochem 257, 427-4:13 7 Zaliti s J, Uram M, Bowser A M & Feingo ld D S (1072 )

Methods E,IZYIl10/2 8. 430-4:15 8 Darrow R A & Rodstrom R ( I 06X) Biochl'lllisilT 7, 1645-

1654 9 Tanford C ( 1968) Ad" Prot Chelll 23. 12 1-2X2

10 Mani atis S, Fritsch E F & Sambrook .J ( 10X2) MoleclIlar Cloning:: A Labomtorv M{(IlI ((I1 Co ld Spring Harbour Laboratory Press, N. Y

II Udenfield S ( 1962) in FllIoresCl'nce Asstn' ill Biologr and Medicine vol I, pp 249, Academic Prcss . Ncw York

12 Bhallacharyya H & Bhaduri A ( 1092) .I lliol C/I('III 267 . I I 7 14- I I 720

13 Samanta A & Bhaduri A (1982) In diall .I Bioc/}(!1I1 Biophr.I' 19, 320-323

14 Samanta A & Bhaduri A ( I 0W\) .I Riol Cill'IIl 2SX, I I I I X-11122

15 Bychkova V E & Ptit sin (1 995) FEBS Let! 3 )<), 6-X 16 Ackers G K (1975) in Th e Proteill s vo l!. ( curath H & Hill R

L eds) pp. 2-92, Academi c Prcss 17 Goldenberg D P & Creighton T E ( 19X4) An{/I BlI!chelll D X,

1-18 18 Smith F X (1992) in Protein Folding (Crei ght on T Eed ) pp

197-24 1 W H Freeman and Co 19 Creighton T E ( 1990) Bioche/ll .1 270. 1- 16 20 Riddiford L M ( 1966 ) .1 Bioi Chelll 24 1, I 702-2X02 21 Dominici P. Moore P S & Bori Volt all nrni C (1003) Biochell l

J 295, 493-500 22 Chignell D A, Azhir A & Gratzer W ( 1072) ElIr .l Bioch('lll

26,37-42 23 Muller k & Garel J-R ( 1984) BiochelllisilT 23.655-660 24 Martel A & Garel J-R (1984) .I Bioi Chl'lll 25').49 17-4')21 25 Fi scher G & Schmid F X (1990) Bi(}chl'lIIi.lt lr 20. 2205-2212