Recombinant bovine rhodanese: purification and comparison with bovine liver rhodanese

286

Birx'himh'a et Biophvsita Actu, 1121 (19921286-2tJ2 ~t~ 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4838/92/$05.(10

BBAPRO M218

Recombinant bovine rhodanese: purification and comparison with bovine liver rhodanese David M. Miller ~, Gary P. Kurzban ~, Jose A. Mendoza a, John M. Chirgwin a,b, Stephen C. Hardies a and Paul M. Horowitz " Department oJ'Bioc'hemisto'. Unirer.~'ityo f Tl:ras Heahh Science Center at San Antonio, San Antunio, TX qUSA) and h Department o f Medicine. Unit'ersit.v o] h:ta.s Health Sc'iem'e Center at Suu Atttollhh Sail Al~touio, TX qUSAJ

(Received 27 Newember Iqt~l)

Key words: Bovine liver rhodanese: Recombinant rhodanese: Protein purification; Protein expression; Chaperonin; Protein folding: ( E. c'oli )

Recombinant bovine rhodancsc (thiosuifate:cyanide sulfurtransfcrasc, EC 2.8.1.1) has been purified to homogeneity from Escherichia coil BL21(DE3) by cation-exchange chromatography. Recombinant and bovine liver rhodanese coelectrophorese

under denaturing conditions, with an apparent subunit molecular weight of 33000. The amino terminal seven residues of the recombinant protein are identical to those of the bovine enzyme, indicating that E. coil also removes the N-terminal methionine. The K m for thiosulfate is the same for the two proteins. The specific activity of the recombinant enzyme is 12% higher (816 IU/mg) than that of the bovine enzyme (730 IU/mg). The two proteins are indistinguishable as to their ultraviolet absorbance and their intrinsic fluorescence. The ability of the two proteins to refold from 8 M urea to cnzymatically active species was similar both for unassisted refolding, and when folding was assisted either by the detergent, lauryl maltoside or by the E. coil chaperonin system composed of cpn60 and cpnl0. Bovine rhodanese is known to have multiple eleetrophoretic forms under native conditions, in contrast, the recombinant protein has only one form, which comigrates with the least negatively charged of the bovine liver isoforms. This is consistent with the retention of the carboxy terminal residues in the recombinant protein that are frequently removed from the bovine liver protein.

Introduction Rhodanese (thiosulfate:cyanide sulfurtransferase, EC 2.8.1.1) is a mitochondrial matrix enzyme that is thought to be involved in the synthesis and control of iron-sulfur proteins, and in cellular sulfur metabolism [I-8]. The X-ray crystal structure of bovine rhodanese is known [2,9-13]. The protein consists of a single polypeptide chain of 296 amino acids [14]. The active site is located in a cleft between the protein's two domains. In vitro, rhodanese catalyzes the transfer of the outer sulfur atom of thiosulfate to nucleophilic acceptors, such as cyanide. During catalysis, this sulfur atom is bound in a stable persulfide linkage to cysteine 247.

Abbreviations: SDS, sodium dodecyl sulfate; rhodanese A. least negatively charged of rhodanese isoforms; rhodanese B. second least negatively charged of rhodanese i~forms. Correspondence: P.M. Horowitz, Depart;nent of Biochemistry, University of Texas Health Science Center at San Antonio, San Antonio. TX 78284-7760. USA.

Rhodanese has become a useful model for protein folding [15-21)] and protein oxidation [21-24]. To gain further insights, it is desirable to produce unique rhodanese species by molecular biological approaches. Previously, we reported the cloning and expression of a full length eDNA encoding for bovine adrenal rhodanese [14]. This eDNA encodes three residues at the C-terminus that had not been observed in protein sequencing or X-ray crystallography studies [9,10]. Removal of these residues by proteolysis may account for the interconversion of two major electrophoretic variants of rhodanese [14,25]. At the amino terminus, a putative initiator methionine is the only residue deduced from the cDNA sequence that is not present in the purified bovine liver enzyme. Here, we report the expression, purification and characterization of recombinant rhodanese from E. coll. By a one column procedure, a high yield of recombinant rhodanese has been purified to homogeneity, as judged by SDS-polyacrylamide gel electrophoresis. By several criteria, the recombinant rhodanese is indistinguishable from bovine liver rhodanese. These criteria include mobility during SDS-

287 electrophoresis, amino terminal sequence, K m for thiosulfate, ultraviolet absorbance spectrum, intrinsic fluorescence emission spectrum and the success of refolding from urea under several refolding conditions. Two differences have been observed between recombinant and bovine liver rhodanese. First, the recombinant enzyme has 12% higher specific activity than the bovine liver rhodanese, which may reflect differences in protein conformation, or some inactivation of the bovine liver enzyme. Second, the recombinant rhodanese is homogeneous during electrophoresis under native conditions, and co-electrophoreses with the least negatively charged of the isoforms of the bovine liver enzyme. This is consistent with the retention by the recombinant rhodanese of the three C-terminal residues deduced from the eDNA sequence, a n d / o r the absence of protein phosphorylation and deamidation of glutamine and asparagine residues. The methods and results reported here indicate that the recombinant protein is a valid and accessible model for continuing studies of rhodanese, and that the expression and purification of mutant rhodaneses should be achievable. Materials and Methods

Materials. The pET-1 ld plasmid and expression system were from Novagene (Madison, WI); molecular weight standards and chemicals for electrophoresis were from Bio-Rad Laboratories (Richmond, CA); restriction endonucleases and T4 ligase were from New England Biolabs (Beverly, MA); isopropyl /3-D-thiogalactopyranoside was from Sigma (St. Louis, MO); CM 52 cellulose was from Whatman LabSale (Hillsboro, OR) and lauryl maltoside was from Boehringer Mannheim (Indianapolis, IN). Purification of bovine liver rhodanese. Bovine liver rhodanese was prepared as described previously [26]

Barn HI

T7 terminator Nco I

pJSSJD (6834 bp)

TTlrc 1 promoter

ori

Fig. 1. Expression vector pJSSJD, pJSSJD was derived from plasmids pSJKE and pET-l Id, as described in Materials and Methods.

and stored at - 7 0 ° C as a crystalline suspension in 1.8 M ammonium sulfate containing 1 mM thiosuifate. Expression of rhodanese eDNA in E. coil, and rhodanese purification. Plasmid pSJKE [14] was digested with Ncol (which cuts at the initiator methionine) and BamH! (which cuts downstream of the 3'-untranslated region), producing a 1.1 kb fragment that contained the complete coding sequence of bovine adrenal rhodanese. This fragment was subcloned into the Ncol/BamHi sites of pET-l ld. The resultant plasmid, pJSSJD (Fig. l) was used to express rhodanese from a T7 promoter upon addition of isopropyl fl-o-thiogalactopyranoside. E. coil B L 2 1 ( D E 3 ) w e r e transformed with pJSSJD to ampicillin resistance. This expression system requires the use of hosts that arc bacteriophage ADE3 lysogens, which provide T7 RNA polymerase from the lac UV5 promoter [27]. Rhodanese production is efficiently initiated by the lac inducer isopropyi /~-D-thiogalactopyranoside (IPTG).

TABLE 1

Purification of recombinant rhodanese from E. coli Step

Volume (ml)

Total J protein (ml)

Total activity (units)

Yield (%)

Specific activity (units/rag)

Purification (fold)

(I) Lysis supernatant (2) pH 4.5, 0.4 M ammonium sulfate supernatant (3) 1.4 M ammonium sulfate supernatant (4) 2.25 M ammonium sulfate precipitate (5) Dialysis supernatant (6) CM 52 pool (7) Washed precipitate

30 30

1~ 3 ! 727

48144 45 370

(100) 94

~ 26

(1.0) 1.0

30

1558

43248

90

28

I.I

8

893

41045

85

46

1.8

8 9.5 0.5

677 49 43

39576 37536 35 088

82 78 73

58 766 816

2.3 .30.6 32.6

Protein was determined by the method of Bradford [311 for steps 1-5. For steps 6 and 7, the absorbance at 280 nm was employed, assuming the sample to be composed of pure rhodanese.

288 For expression of rhodanese, freshly transformed BL21(DE3) E. coil were employed. A single colony was used to inoculate 3 ml of Terrific Broth [28] containing 200 # g / m l ampicillin, and was grown to late-log phase at 37°C. Due to the secretion of fliactamase by the pJSSJD/BL21(DE3) host, this culture was centrifuged, washed with 2 mi of media, and resuspended in l ml of media prior to inoculation of larger cultures. At an absorbance of 1 at 600 rim, the expression of rhodanese was then induced with 0.4 mM isopropyl fl-o-thiogalactopyranoside. After an additional 3 h of growth, cells from a 1.5 I culture were harvested by centrifugation. Recombinant rhodanese was purified as follows. The E. coli were lysed by treatment with lysozyme and deoxycholate, as described [29], and then centrifuged at ~ 0 0 0 x g for 15 min at 4°C. The lysis supernatant (step l, Table 1) was brought to pH 4.5 at 0-4°C by the addition of 2 M glycine sulfate (pH 2.5) and made 0.4 M in ammonium sulfate. After a 30 min incubation, the suspension was centrifuged, and the resulting supernatant was made 1.4 M in ammonium sulfate, incubated for 30 min and centrifuged. Solid ammonium sulfate was then added to the supernatant until a test centrifugation indicated that about 95% of the rhodanese had precipitated. Typically, 2.25 M ammonium sulfate was employed. After centrifugation, the pellet was resuspended in a minimal volume of dialysis buffer (2 mM sodium acetate, 2 mM disodium thiosulfate, pH at room temperature adjusted to 5.4 with acetic acid), and dialyzed overnight at 4°C against 4 ! of the dialysis buffer, with one change. After dialysis, the turbid sample was centrifuged for 30 min at 22°C. At low levels of expression, rhodanese was found predominantly in the supernatant. However, at high levels of expression, substantial rhodanese was found in both the supernatant and pellet. When rhodanese was predominantly in the supernatant, the supernatant was made 60 mM in sodium acetate (pH 5.4) and loaded directly onto the CM 52 column. When substantial rhodanese was also present in the pellet, rhodanese from both the supernatant and pellet was combined prior to chromatography, as follows. The pellet was incubated briefly with buffer A (60 mM sodium acetate, 2 mM disodium thiosulfate, pH 5.4, with acetic acid) containing 0.25 M sodium chloride, and then centrifuged. Rhodanese was rapidly and selectively solubilized by this procedure. The resulting supernatant was then combined with the supernatant from the dialysis, and the concentration of sodium was adjusted prior to chromatography to that of buffer A by the addition of concentrated sodium acetate buffer (pH 5.4). Chromatography was effective whether or not the rhodanese had precipitated during the dialysis. Chromatography on CM 52 cellulose was performed

at room temperature using a 1.0 × 27 cm, 21 ml column that had been poured according to the manufacturer's instructions and then equilibrated in buffer A. After loading the rhodanese, the column was washed with one bed volume of buffer A and then eluted with a linear gradient of NaCI in buffer A, with a slope of 30 mM per bed volume. The column was monitored at 275 nm with an LKB 2138 Uvicord S flow-through detector. Fractions with high rhodanese activity were pooled, precipitated with 2.5 M ammonium sulfate (by the addition of 100% saturated ammonium sulfate, pH 5.4), and stored at -70°C. A final purification was achieved by pelleting the rhodanese (5 min, 10000 × g at 4°C in a microfuge), and then washing the pellet twice with 1.8 M ammonium sulfate (pH 5.4). The final pellet was dissolved in 0.05 M sodium phosphate, pH 7.5. Unfolding and refolding of rhodanese. The procedures for the unfolding, refolding and subsequent assay of rhodanese were essentially as described [18]. Each type of rhodanese at 300 ~ g / m l was unfolded by incubation for at least 30 min in 8 M urea. Refolding was performed at 25°C, and was initiated by diluting 3 ~tl of the unfolded protein into 250 /zl of 50 mM Tris-HCl (pH 7.8) containing 200 mM fl-mercaptoethanol and 50 mM disodium thiosulfate. Successful refolding was defined as the regain of enzymatic activity, and was assessed as a function of time to be sure that the refolding process was complete. Successful refolding was expressed as a percentage of the activity of an equal amount of each type of native, never-unfolded enzyme, that had undergone a similar sequence of dilutions, omitting urea. For detergent-assisted refolding, the refolding buffer additionally contained 5 mg/ml lauryl maltoside. For chaperonin assisted refolding, the refolding mixture did not contain detergent, but additionally contained 2.5/zM cpn60 (protomere), 2.4 ~ M cpnl0 (protomers), 2 mM ATP, l0 mM MgCI 2 and 10 mM KCI. Analytical techniques. Ultraviolet absorbance was measured using a Spectronic 3000 spectrophotometer (Milton Roy, Arvada, CO). Fluorescence emissions were measured and analyzed as described previously [30] with l nm excitation and 5 nm emission band passes, using an SPF-500C spectrofluorometer (SLM Instruments, Urbana, IL). Protein concentrations of early steps in the rhodanese purification were determined by the method of Bradford [31]. Rhodancse activity was assayed colorimetrically and the concentration of purified samples was determined using an absorbance at 280 nm of 1.75 cm-n at I mg/ml [32,33]. Electrophoresis under denaturing conditions was performed in the presence of SDS as described by Laemmli [34], using a 12% resolving gel. Elecuophore-

289 sis under native conditions was performed as described previously [14]. Gels were stained with 0.05% Coomassie blue R-250, 25% isopropanol, 10% acetic acid. Rhodanese was sequenced using an Applied Biosystems 477A protein sequencer. To determine the K= for thiosulfate, data were fit by a nonlinear least squares procedure (Minsq; MicroMath Scientific Software, Salt Lake City, UT) to the following equation: V = I/~,,:,,([thiosulfate]/([thiosulfate] + Kin)) where V is the observed activity and Vm~~ is the maximum activity. Results and Discussion

Expression and purification of recombinant rhodanese in E. coll. We previously reported the cloning of a full length eDNA encoding bovine adrenal rhodanese [14]. Here, we subcloned the rhodanese eDNA into the pET-l ld vector. The resulting plasmid-host expression system (pJSSD, Fig. l in E. coil BL21(DE3)) stringently regulates rhodanese expression, with high levels of expression upon induction [27]. 1.5 l of E. coil (10 g wet weight) yielded rhodanese activity in the supernatant of the lysate equivalent to 59 mg of pure recombinant enzyme. Most of the rhodanese was present in the soluble phase. Additional, inactive rhodanese was also found in insoluble inclusion bodies (as determined by Western analysis, data not shown). The specific activity in the lysis supernatant (step ! in Table l, lane l in Fig. 3)was 3% that of pure rhodanese. The pH of the supernatant was lowered to 4.5, and all subsequent purification procedures were performed at low pH so as to stabilize the rhodanese activity [35,36]. Rhodanese has net negative charge at physiological pH, and net positive charge throughout the purification procedure. The substrate, thiosulfate, was added to buffers to further stabilize the rhodanese. {:

2.0

0.5~

~

1.6

0.4 ~

o ~ .<

1.2

0.3~

¢-

ui 0.a o,1 z '< O O "r tr

0.2 C

0.1

0.4

d

0.0 0

p~ 04 0.0 <

~

20

40

60

80

100

VOLUME. ml Fig. 2. Cation-exchange chromatography on CM 52 at pH 5.4. E. coil were grown, harvested, lysed, fractionated at low pH with ammonium sulfate and dialyzed, as de~ribed in Materials and Methods. A typical chromatogram is shown. Conductivity ( - - - - - ) ; absorbance, in arbitrary units ( ~, rhodanese activity (e . . . . . . e). The flow rate was 10 ml per h. Rhodanese activity eluted at 0.{l M total sodium.

1

2

m

3

m

4

5

~ tram=m,

:1 Fig. 3. SDS-polyacrylamide gel electrophoresis of recombinant and bovine liver rhodanese. Lane I, supernatant of E. coil lysis (step 1 in Table I): lane 2, purified recombinant rhodanese a5 pooled from the CM 52 column. 1.8 big (step 6, Table !); lane 3, a mixture of recombinant and bovine liver rhodanese, employing 0.75/,tg of each: lane 4. bovine liver rhodanese. 1.8 ~g: lane 5. molecular weight markers, phosphot3,1ase (97400). bovine ,serum albumin (66200), ovalbumin (45000), carbonic anhydrase (31 [100), soy trypsin inhibitor (21500) and lysozyme (14400).

The supernatant of the lysate was fractionated with ammonium sulfate, and the precipitated rhodanese was then dialyzed against a low ionic strength buffer so as to facilitate cation exchange chromatography. The combined pH, ammonium sulfate fractionation and dialysis resulted in an approximately 2-fold purification (Table 1). The dialyzed material was subjected to cation exchange chromatography at pH 5.4 (Fig. 2). All of the rhodanese activity bound to the column. The vast majority of the E. coil proteins failed to bind tightly to the column. A~s a result, this step was highly effective in purifying the rhodanese. Rhodanese eluted in a single peak and was pooled based upon a combination of activity measurements and SDS-electrophoresis, and was then precipitated with ammonium sulfate. As needed, remaining impurities were removed by washing the ammonium sulfate pellet with an intermediate concentration of ammonium sulf,~.te [26]. The resulting rhodanese was homogeneous as judged by electrophoresis under denaturing conditions (Fig. 3). The overall yield of the purification procedure was 73% of the initially soluble, active rhodanese (Table l). There appears to be no reason that the procedure cannot be scaled up, especially since the one column had a 21 ml bed volume. We note that the cation exchange column can be used repeatedly, and can be stored in sodium azide. Comparison of recombinant and bovine liver rhodanese. Recombinant and bovine liver rhodanese were compared as to their K m for thiosulfate, specific activ-

290 ity, electrophoretic mobility under denaturing conditions, mobility and heterogeneity during native electrophoresis, ultraviolet absorbance, intrinsic fluorescence and ability to refold from 8 M urea spontaneously, or with the assistance by a detergent or the E. coli chaperonins, cpnl0 and cpn60. Similarities between recombinant and bocine liter rhodanese. Interactions between rhodanese and the substrate, thiosulfate, were tested by performing the standard rhodanese assay at various concentrations of thiosulfate. The K m for thiosulfate was 16.8 + 0.8 and 15.9-2-0.8 mM, for recombinant rhodanese and bovine liver rhodanese, respectively (data not shown). The molecular weights of the two proteins were compared by SDS-electrophoresis (Fig. 3). The two proteins co-electrophoresed, and only one band was seen when the samples were mixed prior to electrophoresis. The apparent molecular weight was 33 000, as expected for rhodanese. Electrophoresis might fail to detect small differences in the length of the two polypeptide chains. The N-terminal valine of purified bovine liver rhodanese [9,10] corresponds to residue two of the sequence deduced from the eDNA sequence [14]. That is, the bovine liver enzyme, as purified, has had the putative initiator methionine, and only that methionine, removed from the amino terminus. We asked whether E. coli also removes the N-terminal initiator methionine from the recombinant rhodanese. The recombinant protein was sequenced. Yields were essentially quantitative, indicating a miaimum level of N-terminal blocked species. The sequence of the first seven residues of the recombinant protein was: VaI-His-GInVaI-Leu-Tyr-Arg-. This sequence corresponds to residues 2-8 of the eDNA for rhodanese and residues 1-7 of the bovine liver enzyme. Thus, the N-terminal methionine, and only this residue, was removed from the recombinant rhodanese by E. coli. As a result, the recombinant and bovine liver enzymes have identical N-terminal sequences. The ultraviolet absorbance (250-300 nm) and intrinsic fluorescence emission spectra of the recombinant and bovine liver rhodanese were compared (data not shown). Total fluorescence emissions, normalized to the absorbance at 280 nm and integrated from 300-450 rim, differed by less than 1%. The fluorescence emission maxima were indistinguishable (0.2 nm difference), as were bandwidths at half height. The ultraviolet absorbance spectra were indistinguishable. The refolding of recombinant and bovine liver rhodanese were compared (Table If). Rhodanese was unfolded in urea, and then diluted so as to allow the protein to refold (see Materials and Methods). Refolding was performed in the presence of a detergent (lauryl maltoside), or an E. coli chaperonin system, or

TABLE II

Refoldhtg of recomb#rant and borhw lirer thodanese Rhodanese was unfolded and refolded as described in Materials and Methods. Activity was measured 80, 2711or 80 rain after the initiation of refolding for unassisted, detergent assisted and chaperonin assisted rcfolding, respectively. Protein

Recombinant Bovine liver

Percent successful refolding unassisted

detergent assisted

chaperonin assisted

25 26

68 7i

80 78

in the absence of these additives. The amount of successful refolding, to active rhodanese, was similar for the two sources of rhodanese for each of the three refolding protocols (Table II). Differences between recombinant and bocine liter rhodanese. The specific activity of recombinant rhodanese (816 I U / m g ) was 12% higher than that of bovine rhodanese (730 IU/mg). The reason for this difference is not clear. Possibilities include conformational differences, or the inactivation of some bovine liver enzyme. The latter possibility is consistent with the lack of difference in the amount of refolding (see above). Whatever the cause, differences in the method of purification could be responsible for the difference in specific activity. Bovine liver rhodanese is heterogeneous as judged by electrophoresis under native conditions, and as many as four isoforms have been reported [37]. Explanations for this heterogeneity have included differential deamidation of glutamine a n d / o r asparagine residues [38], phosphorylation [6] and proteolysis at the carboxy terminus [14,25]. We have previously studied the two major, least negatively charged isoforms. The least negatively charged form, 'rhodanese A', can be converted into a species that co-electrophoreses with the other major form, 'rhodanese B', by treatment with the C-terminus exopeptidase, carboxypeptidase B [25]. Further, the sequence of the eDNA encoding for bovine adrenal rhodanese indicates the presence of 296 residues [14], rather than the 293 residues determined by protein sequencing [9] and in the X-ray crystal structure [10]. The extra three residues are at the C-terminus and include one charged residue, lysine 295. The removal of lysine 295 by proteolysis should confer charge differences consistent with those seen when converting rhodanese A to rhodanese B. Indeed, carboxypeptidase B converts recombinant rhodanese A into rhodanese B [14]. Thus, we have developed a consistent picture that the difference between rhodanese A and rhodanese B is the absence of positively charged residue(s) from the C-terminal of rhedanese B.

291

1

2

Fig. 4. Native electrophoresis of recombinant and bovine liver rhodanese. Lane I. 15 lzg of recombinant rhodanese: lane 2. 15 u g of bov/ne liver rhoclanese. The anode is at the hoUom.

Native electrophoresis of the recombinant and bovine liver rhodanese was performed so as to extend our understanding of the interrelationship between the rhodanese isoforms (Fig. 4). Bovine rhodanese elcctrophoresed as two major bands, as is generally observed. The recombinant protein was homogeneous, and had the same mobility as bovine liver rhodanese A, the least negatively charged isoform. This result is consistent with our understanding of the charge isoforms of rhodanese (see above), and indicates that the purified recombinant enzyme retains the C-terminal tripeptide.

Suitability of recombinant rhodanese for physical studies. For recombinant rhodanese to be useful for further physical studies, two criteria must be satisfied. First, a procedure for facile expression and purification must be developed. Second, physical differences between the recombinant and bovine forms must be either minimal or favorable. The results presented here indicate that these conditions have been satisfied. The synthesis and folding of the recombinant rhodanese result in the production of active, soluble protein that was readily purified by a single column procedure (Table 1). The purification procedure for recombinant rhodanese is simpler than methods for bovine liver rhodanese. For bovine liver rhodanese, procedures generally employ repeated ammonium sulfate fractionations and the crystallization of rhodanese [26,32,33,38-42], augmented by chromatography on Affi-gel blue [26]. These procedures require substantial time and expertise. We have recently developed a column-based purification method for bovine liver rhodanese [35]. In this method, the ammonium sulfate precipitations are replaced with a three column procedure. This new procedure is still more arduous than the one column procedure for recombinant rhodanese. Thus, the purification procedure for the recombinant rhodanese is simpler and quicker than all methods

reported previously for bovine liver rhodanese. The simpler purification protocol is due to the higher initial specific activity of the recombinant rhodanese, and perhaps to the absence in E. coli of bovine liver proteins that interfere with purification procedures due to interactions with the rhodanese. The second criterion for the utility of recombinant rhodanese is that differences from the liver enzyme be minimal or favorable. Because proteins are subjected to covalent modifications both in vivo and during purification procedures, it is possible for recombinant rhodanese to have a different primary structure than rhodanese purified from bovine liver. We have found no evidence for covalent differences, apart from the heterogeneity of the bovine liver enzyme. The amino terminal sequences of the two proteins were identical, molecular weights were similar, and the recombinant enzyme most likely retains three carboxy terminal residues (residues 294-296 using the numbering of the bovine liver enzyme), whose removal likcly account for the heterogeneity of the bovine liver enzyme. The heterogeneity of bovine liver rhodanese in native electrophoresis is likely due to changes in the covalent structure of the protein. Possible causes include deamidation of glutamine and asparagine residues [38], phosphorylation [6] and proteolysis at the carboxy terminus (Refs. 14, 25, this report). These events may occur, in part, during the involved purification of the bovine liver enzyme. Although the bovine liver rhodanese is heterogeneous, the recombinant enzyme is homogeneous during native electrophoresis, and coelectrophoreses with the least negatively charged bovine isoform. The least negatively charged species will be that with the least amount of deamidation (to negatively charged aspartates and glutamates), the lowest phosphate content and the most intact earboxy terminus. Thus, recombinant rhodanese most likely contains the complete amino acid sequence dictated by the eDNA sequence (less the initiator methionine), without the deamidations, phosphorylations or C-terminal proteolysis that may alter bovine liver rhodanese. Conformational differences between the two sources of rhodanese also appear to be minimal. Both sources have the same K m towards thiosulfate. Ultraviolet absorbance and intrinsic fluorescence emissions were indistinguishable between the recombinant and bovine liver proteins. The 12% higher specific activity of the recombinant rhodanese could be due to a conformational difference, but could also be due to the presence of some inactive enzyme in the bovine material. Such species would not affect the measured K m towards thiosulfate, and as minor species could be undetected by the spectral techniques. Both forms of rhodanese could refold from 8 M urea, under selected conditions, to native, active rho-

292

danese and both could be assisted in their refolding by either a detergent or by the E. coli chaperonin system. The extent of successful refolding was similar for the two sources of rhodanese. These experiments are consistent with the proteins sharing a similar covalent structure. Since the unfolding/refolding cycle eliminates any conformational differences between the proteins, the identical recoveries are consistent with the slightly higher specific activity of the recombinant rhodanese being due to the presence of some bovine rhodanese that is inactive, and that does not refold to active species. There appears to be no reason why recombinant rhodancse cannot be used for further studies of rhodanese. Specifically, site-directed mutagenesis of rhodanese should be useful in testing hypotheses concerning both folding and oxidative inactivation. Such studies are currently in progress. Acknowledgements This work was supported by research grants GM25177 and ES05729 from the National Institutes of Health and AQ-723 from the Robert A. Welch Foundation to PMH, Grants HG00190 from the National Institutes of Health and AQ-1107 from the Robert A. Welch Foundation (to SCH) and Grant AQ-1123 from the Robert A. Welch Foundation and a Veterans Administration merit award (to JMC). GPK is a Robert A. Welch Fellow. We thank Peggy Rifleman of the Biopolymer Sequencing and Synthesis Facility, University of Texas Health Science Center at San Antonio, for sequencing the amino terminus of recombinant rhodanese. References I Westley. J. 119731 Adv. En~mol. Rel. Areas Mol. Biol. 39, 327-368. 2 Hol, W.G.J., Lijk. L.J. and Kalk, K.H. (19831 Fundam. Appl. Toxicol. 3. 370-376. 3 Westley, J., Adler. H., "~Vestley, L. and Nishida, C. (1983) Fundam. Appl. Toxicol. 3. 377-382. 4 Cerletti, P. (1986)Trends Biochem. Soc. 11,369-372. 5 0 g a t a , K. and Volini, M. (1986) J. Prot. Chem. 5, Z~;9-246. 6 0 g a t a , K., Dai, X. and Volini. M. (1989) J. Biol. Chem. 264. 2718-2725. 7 Toobey, J.I. (1989) Biocbem J. 264, 625-632. 8 0 g a t a . K and Volini. M. (1990) J. Biol. Chem. 265, 8087-8093. 9 Rus~ll. J.. Weng. L., Keim. P.S. and Heinrikson, R.L. (1978) J. Biol. Chem. 253, 8102-8108.

10 Ploegman, J.H., Drent, G.. Kalk. K.H., Hol. W.G.J., l-leinrik~n,

R.L., Keim. P.. Weng. L. and Russell. J. 119781 Nature 273, 124-129. I I Lijk. L.J.. Kalk. K.H., Brandenburg. N.P. and l tol. W.G.J. (1983) Biochemistry 22. 2952-2957. 12 Lijk. L.J.. Toffs. C,A.. Kalk. K.H., De Maeyer. M.C.H. and Hol. W.G.J. (1984) Eur. J. Biochem. 142. 399-408. 13 K~ystra, P.J.U.. Kalk, K.H. and Hol, W.G.J. (19881 Eur. J. Biocbem. 177, 345-349. 14 Miller. D.M.. Delgado, R., Chirgwin, J.M., tlardies, S.C. and Hort~vitz. P.M. (1991)J. Biol. Chem. 266, 4686-4691. 15 Tandon, S. and Horowitz. P.M. (1990)J. Biol. Chem. 265. 5967597l}, 16 Horowitz. P.M. and Criscimagna, N.L. (1~{}) J. Biol. Chem. 265, ~76-2583. 17 Martin, J.. Langer, T., I~teva, R.. SchrameL A.. Horwich, A.L. and itartl. F.-U. (1991) Nature 352. 36-42. lg Mendoza, J.A.. Rogers. E., Lot(met, G.H. and Horowilz, P.M. ( 1991 ) J. Biol. (.'hem. 266. 13044-13(M9. 19 Mendoza. J.A.. Rogers, E.. Lorimer, G.H. and Horowitz, P.M. (1991) J. Biol. Chem. 266, 13587-13591. 2{) Mcndoza. J.A., Lorimer, G.H. and Horowitz. P.M. (1991)J. Biol. Chem. 266. 16973-16976. 21 Horowiiz. P.M. and Bowman. S. (1987) J. Biol. Chem. 262, 8728-8733. 22 Horowitz~ P.M. and Bowman. S. (1987) J. Biol. (.'hem. 262. 14544-14548. 23 Merrill, G.A., Horowitz, P.M., Bowman. S., Bentley, K. and Klebe, R. (1988) J. Biol. Chem. 263, 19324-19330. 24 Horowitz, P.M. and B~'man, S. (1989) J. Biol. Chem. 264, 3311-3316. 25 Horowitz, P.M. and Falksen. K. (1983)J. Biol. Chem. Z58. 16141618. 26 Horowitz, PM, (1978) Anal. Biochem. 86. 751-753. 27 Studier. F.W., Rosenberg, A.H.. Dunn. J.J. and Dubendorff, J.W. (19911) Methods EnzFmol. 185, 611-89. 28 Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: a Laboratow Manual, 2nd Edn., p A.2, Cold Spring Harbor Laboratow Press. Cold Spring Harbor, NY. 29 Sambrook. J.. Fritsch. E.F. and Man(at(s, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd Edn., p 17.38, Cold Spring Harbor Laboratory Press, Cold Spring Harl~r. NY. 30 Kurzban, G.P., Gillin, G., Bayer, E.A., Wilchek. M. and Horowitz, P.M. (1990) J. Protein Chem. 9, 673-682. 31 Bradford, M.M. 119761 Anal. Biochem. 72, 248-254. 32 S6rbo, B.H. (1953)Acta Chem. Stand. 7, 1129-1136. 33 Sorbo, B.H. (1955) Methods Enzymol. 2, 334-337. M Laemmli, U.K. (1970) Nature 227, 680-685. 35 Kurzban. G.P. and Horowitz, P.M. (1992) J. Protein Express. Purif., in press. 36 Reference has been deleted. 37 Canella, C., Costa, M., Pen~, B., Rice(, G., Pccci, L. and Cavallini, D. 119811Eur. J. Biochem. 119, 491-495. 38 Blumenthal. K.M. and Heinrikson, R.L. (1971) J. Biol. Chem. 246, 2430-2437. 39 Westley, J. and Green, J.R. (19591J. Biol. Chem. 234, 2325-2326. 40 Leininger, K.R. and Westley, J. (19681J. Biol. Chem. 243, 18921899. 41 Horowitz, P.M. and DeToma, F. (19701 J. Biol. Chem. 245. 984-985. 42 Wang, S.-F. and Volini, M. (19731J. Biol. Chem. 248, 7376-7385.