Truncations at the NH2 Terminus of Rhodanese Destabilize the Enzyme and Decrease Its Heterologous Expression*

Rhodanese mutants containing sequential NH2-terminal deletions were constructed to test the distinct contributions of this region of the protein to expression, folding, and stability. The results indicate that the first 11 residues are nonessential for folding to the active conformation, but they are necessary for attaining an active, stable structure when expressed inEscherichia coli. Rhodanese species with up to 9 residues deleted were expressed and purified. Kinetic parameters for the mutants were similar to those of the full-length enzyme. Compared with shorter truncations, mutants missing 7 or 9 residues were (a) increasingly inactivated by urea denaturation, (b) more susceptible to inactivation by dithiothreitol, (c) less able to be reactivated, and (d) less rapidly inactivated by incubation at 37 °C. Immunoprecipitation showed that mutants lacking 10–23 NH2-terminal amino acids were expressed as inactive species of the expected size but were rapidly eliminated. Cell-free transcription/translation at 37 °C showed mutants deleted through residue 9 were enzymatically active, but they were inactive when deleted further, just as in vivo. However, at 30 °Cin vitro, both Δ1–10 and Δ1–11 showed considerable activity. Truncations in the NH2 terminus affect the chemical stability of the distantly located active site. Residues Ser-11 through Gly-22, which form the NH2-proximal α-helix, contribute to folding to an active conformation, to resisting degradation during heterologous expression, and to chemical stability in vitro.

since they are poorly resolved in the x-ray structure. Residues Val-4 to Arg-7 are fixed to the surface of the NH 2 domain. Residues Ala-8 to Val-10 contribute one of five strands of a ␤-sheet that runs into the interior of the NH 2 domain, and residues Ser-11 to Gly-22 form one of five ␣-helices that surround the ␤-sheet. The residue Lys-23 starts a turn into the interior of the NH 2 domain. All of the residues required for catalysis, including the essential Cys-247, are on the C-terminal domain. In one view then, the C-terminal domain may be active alone. This has never been observed, and it has not been possible to produce stable COOH-terminal domains either proteolytically or recombinantly. Rhodanese offers a unique opportunity to investigate the influence of the NH 2 -terminal portion on the activity of the enzyme.
The majority of proteins imported into mitochondria have their NH 2 -terminal sequences removed proteolytically (7,8). Rhodanese purified from bovine liver has the same NH 2 -terminal sequence as that predicted from the mRNA except it is missing the initiating methionine. This lack of proteolytic processing is unusual but not unique, however, as other mitochondrial proteins share this property, e.g., 3-oxoacyl-CoA thiolase and cytochrome oxidase VIc. It is unclear why certain proteins retain their NH 2 -terminal leader sequences. When the first 23 amino acid residues of rhodanese were fused to the NH 2 terminus of dihydrofolate reductase, which is not normally found in mitochondria, the enzyme was imported (9). The NH 2 -terminal sequence could be cleaved only when a mitochondrial processing peptidase cleavage site was introduced (10). It was proposed that the ␣-helical structure contained in the NH 2terminal sequence confers the structure necessary for import (11). Retention of the NH 2 -terminal sequence of rhodanese in the mature protein suggests it serves functions in addition to import.
A role for the NH 2 -terminal sequence in stabilizing the protein has been suggested by studies with rhodanese that contained a site-directed mutation that replaced glutamic acid at position 17 with proline. The purified protein showed increased susceptibility to denaturation by urea and a decreased efficiency of refolding in vitro (12). However, the mutant protein was stable, active, and purifiable from Escherichia coli. The E17P mutant does not directly address the role of the NH 2terminal segment in folding to a stable conformation during heterologous expression.
We expressed NH 2 -terminal truncations of rhodanese in E. coli and in an E. coli-based, cell-free transcription/translation system to investigate contributions of the NH 2 -terminal segment to folding, stability, and resistance to degradation during heterologous expression. E. coli is considered particularly appropriate for rhodanese, since the bacterial cytosol provides a folding environment similar to the mitochondrial matrix (13,14), and many useful recombinant proteins are produced in E. coli host cells. Full-length rhodanese is stable and active when expressed in E. coli (15). We now report that the first 9 residues of the NH 2 terminus of rhodanese are nonessential for the activity or resistance to elimination of the protein expressed in E. coli. Deletions of 7 or 9 NH 2 -terminal residues alter the chemical stability of the isolated protein. Deletions of residues 10 and beyond resulted in loss of activity when the protein was expressed in E. coli at 37°C. These same mutant proteins were not degraded in the cell-free system, and they were synthesized in similar quantities without degradation at either 30 or 37°C. At 37°C, the synthesized rhodanese was inactive when 10 or more NH 2 -terminal residues were removed. Expression at 30°C, however, resulted in active protein even when 11 residues were deleted, showing that under conducive conditions rhodanese can fold to an active conformation even when the truncation extends into the first ␣-helix. In sum, the interactions involving the NH 2 -terminal sequence of rhodanese can have global effects on the protein altering the reactivity of the active site.

EXPERIMENTAL PROCEDURES
Materials-All chemicals (analytical grade) and antibiotics were obtained from Sigma. All media components were purchased from Life Technologies, Inc. H 2 35 SO 4 and 14 C leucine were supplied by NEN Life Science Products. Polyclonal antibodies were previously described (16). Oligonucleotide primers were made by Genosys or Life Technologies, Inc. Restriction endonucleases were obtained from New England Biolabs. Chromatographic steps for purification of mutant proteins were performed using a Pharmacia fast protein liquid chromatography system.
Expression in a Cell-free Transcription/Translation System-The coupled transcription/translation system used to express the rhodanese truncation mutants was prepared and employed as described previously (30 -32). All coding regions for the mutants were ligated as 5Ј PstI to 3Ј BamHI fragments into pSP64 (Promega). Equal aliquots of the cell-free samples for each mutant were assayed for rhodanese activity and for total protein concentration to quantify the relative amount of active protein made. SDS-PAGE 1 gels were run to verify the presence of the mutant proteins in similar quantities as wild type.
Rhodanese Assay in Cell Lysates-Activity was calculated from the extinction coefficient at 460 nm of the complex formed between the reaction product, thiocyanate, and ferric ion (33). The assay was modified for bacterial cellular lysates. Host cell culture (1 ml) was collected by centrifugation and resuspended in 250 l of 0.1% Triton X-100 (v/v), 10 mM Tris-HCl, pH 8, 1 mM EDTA, 50 mM Na 2 S 2 O 3 , 100 mM NaCl for 10 min to lyse the cells. An aliquot of the cellular lysate was then assayed following a brief clarification centrifugation step just before the absorbance reading. An equal aliquot was assayed for total protein concentration.
Immunoprecipitation of Expressed, Radiolabeled Proteins-After a 2-h induction period, samples were centrifuged and the medium decanted. The cell pellets were resuspended and grown in radioactive M9 minimal medium (  The amino acid sequence in single letter code is given for the NH 2 -terminal sequence of all rhodanese truncation mutants and wild type. (M) indicates the purified protein has the initiating methionine excised. M indicates the purified protein retains the initiating methionine. "M' indicates excision or retention of the initiating methionine on the protein was not investigated. Letter codes in italics and underlined indicate amino acid residue substitutions from the wild-type sequence.
50 g/ml ampicillin, 35 g/ml chloramphenicol, 150 g/ml rifampicin, pH 7.4) for 15 min. The samples were then switched to nonradioactive M9 medium containing 2 mM MgSO 4 . Parallel samples without radiolabeling were used for activity determinations. At time points after the addition of nonradioactive medium, 1-ml aliquots were centrifuged and the supernatants decanted. The pellets were resuspended in 100 l of Laemmli sample buffer (34) and placed in a boiling water bath for 5 min. Samples were then centrifuged to pellet insoluble debris, and the supernatants were immunoprecipitated at 4°C by adding 900 l of immunoprecipitation buffer (100 mM KCl, 100 mM MgCl 2 , 1% Triton X-100 (v/v), 0.5% SDS (w/v)). Polyclonal antiserum raised against purified rhodanese was added (1:1000 dilution), and after 18 h the samples were briefly centrifuged. The pellets were discarded, and the supernatants were mixed with 100 l of a 50% (w/v) suspension of protein A-Sepharose (Sigma) equilibrated with immunoprecipitation buffer. After 2 h of incubation with the protein A-Sepharose, the samples were centrifuged, and the pellets were washed with 1 ml of immunoprecipitation buffer three separate times. The washed pellets were resuspended in 30 l of Laemmli sample buffer, incubated 5 min in a boiling water bath, and centrifuged. The pellets were discarded, and the supernatants were subjected to SDS-PAGE. The gels were fixed, treated with En 3 hance (Dupont), dried, and exposed to autoradiographic film (Kodak) or analyzed with a PhosphorImager. Images were digitally acquired either with a GBC video camera (CCTV Corp.) for autoradiographs or with the PhosphorImager (Molecular Dynamics). Signal intensities were quantified using NIH Image software or ImageQuant software (Molecular Dynamics).
Mutant Protein Purification-Purification protocols for ⌬1-3, ⌬1-7, and ⌬1-9 mutant proteins were based on the method described previously to purify recombinant rhodanese (35). The procedure was unchanged up to the point of dialysis in low ionic strength buffer of the resuspended pellet from the final (NH 4 ) 2 SO 4 precipitations of the cellular lysates. During dialysis, the mutant proteins formed precipitates that contained the enzyme activity. These precipitates were extracted with 1.5 M (NH 4 ) 2 SO 4 , 5 mM Na 2 S 2 O 3 , 5 mM NaC 2 H 3 O 2 , 10 mM MES, 1 mM EDTA, pH 5.0, and applied to a hydrophobic interaction column (Pharmacia CL-4B phenyl Sepharose) equilibrated with the extraction buffer. Elution of the active mutant proteins was performed by adding 5 mM Na 2 S 2 O 3 , 5 mM NaC 2 H 3 O 2 , 10 mM MES, 1 mM EDTA, pH 5.0, in a 0 -100% gradient with an increase of 1.5% of the elution buffer for every milliliter added to the column. SDS-PAGE revealed significant contamination of the eluted active fractions, which were combined and dialyzed in the above elution buffer prior to loading onto a dye-binding column (Bio-Rad Affi-Gel Blue) equilibrated in the same buffer (36). Elution of the active mutant proteins was achieved by adding 1.5 M NaCl, 5 mM Na 2 S 2 O 3 , 5 mM NaC 2 H 3 O 2 , 10 mM MES, 1 mM EDTA, pH 5.0, to the previous gradient conditions. The ⌬1-3 and ⌬1-7 mutant proteins were free of contaminating proteins as judged by an SDS-PAGE gel stained with AgNO 3 . ⌬1-9 still contained three to four lower molecular weight proteins, which were not detected in a Western blot using polyclonal antibodies raised against wild-type rhodanese. The active ⌬1-9 fractions were combined and dialyzed in 5 mM Na 2 S 2 O 3 , 5 mM NaC 2 H 3 O 2 , 10 mM MES, 1 mM EDTA, pH 5.0, then loaded onto a cation exchange column (Whatman CM-52) equilibrated in the same buffer. Elution was performed by adding 400 mM NaCl, 5 mM Na 2 S 2 O 3 , 5 mM NaC 2 H 3 O 2 , 10 mM MES, 1 mM EDTA, pH 5.0, in a 30 -100% gradient with an increase of 1.5% of the elution buffer for every milliliter added. The eluted, active fractions were Ͼ95% homogeneous, as judged by AgNO 3 staining of a 12% SDS-PAGE gel. Purified proteins were precipitated with 2.5 M (NH 4 ) 2 SO 4 and were stored at Ϫ70°C. NH 2 Terminus Analysis-The amino-terminal sequence of each purified, mutant protein was determined using an Applied Biosystems 477A Sequencer coupled to an ABI 120 HPLC Analyzer using a reversed phase C18 PTH column.
Enzyme Kinetic Determinations-Absorbance at 280 nm of 1.75 cm Ϫ1 at 1 mg/ml was used to determine concentrations of the rhodanese species that were purified (up to ⌬1-9) (37). These purified mutants (up to ⌬1-9) that were analyzed retained all tryptophan residues found in the wild type. Activity was determined using a standard rhodanese assay (33). The formation of SCN Ϫ was determined as ferric thiocyanate with a molar extinction coefficient at 460 nm of 4.2 mM Ϫ1 cm Ϫ1 . For determination of K m , the concentration of thiosulfate was varied between 1 and 10 mM, and the data were fit to the equation Curves were fit to data based on a two-state transition mechanism (40). Table II. Deletions through 9 amino acid residues from the NH 2 terminus of rhodanese resulted in levels of activity similar to the wild-type control. Deletions past 9 residues resulted in no detectable activity compared with the negative control at 37°C. Fig. 1 shows a plot of activity as a function of time at 37°C for  active, stable truncation mutants. Active mutants showed little loss of activity for up to 4 h after induction. ⌬1-9 was less stable as seen by a decrease in activity over time. When the experiment was repeated at 30°C (Table II), ⌬1-10 had detectable activity, but it was only slightly above background.

Expression in E. coli-Activities in cell lysates are shown in
Immunoprecipitation of Radiolabeled Proteins-Labeled cell proteins were immunoprecipitated at increasing times after induction with polyclonal antibodies raised against purified rhodanese. Fig. 2 shows representative autoradiographs of SDS gels of the immunoprecipitations of wild-type, mock-transformed, ⌬1-23, and ⌬1-9 proteins. The signal intensities of wild-type and ⌬1-9 proteins did not decrease significantly during a 1-h chase. The amount of the radiolabeled ⌬1-23 protein decreased rapidly after the 15-min labeling period, suggesting that it was quickly degraded in the host cells. The signal intensity of ⌬1-23 at t ϭ 0 was less than that of wild type. The data suggest that, although this mutant is synthesized as a full-length protein of the expected size (31 kDa), it has a halflife shorter than the 15-min labeling period. The mock-transformed host cell control showed only a lower molecular mass (28 kDa), nonspecific protein present in all samples. Non-induced samples ("N" lanes) expressed little rhodanese. The same radiolabeling protocol was employed with the inactive mutants. Fig. 3 shows plots of signal intensity as a function of time at 37°C for the inactive mutants compared with wild-type and the active truncation mutant, ⌬1-9. Inactive mutant proteins were rapidly eliminated in the host cells, although they could all be detected at t ϭ 0 by immunoprecipitation.
Expression in the Cell-free System-Wild-type and selected truncation mutants were expressed at 37 or 30°C in a cell-free, coupled transcription/translation system derived from E. coli. At 37°C, activity (Table III) was undetectable when 10 or more residues were deleted, which was similar to results obtained when these mutants were expressed in E. coli. At 30°C, activity was detected with up to 11 residues deleted. All samples contained similar amounts of protein, and little or no degradation of mutant proteins was observed (data not shown).
Purification of Rhodanese Truncation Mutants-Mutant proteins not eliminated in E. coli were purified to homogeneity to compare their catalytic characteristics. The purification method for the wild-type enzyme required some modification, since the mutant proteins tended to precipitate at lower (NH 4 ) 2 SO 4 concentrations than wild-type rhodanese. Both hydrophobic interaction chromatography and dye-binding affinity chromatography were needed to purify ⌬1-3 and ⌬1-7, whereas wild type could be purified to apparent homogeneity using only cation exchange chromatography. Purification of ⌬1-9 required all three chromatographic steps. As seen in Table IV, the final yields were greatest for wild type and ⌬1-3. The percent recoveries for ⌬1-7 and ⌬1-9 were significantly less, with ⌬1-9 being the most difficult to separate from contaminating proteins. Table IV shows that the K m values and specific activities for the purified mutants did not differ significantly from wild type. The concentrations of urea representing the denaturation transition midpoints (U 1/2 ) shown in Fig. 4 were the same for wild type and ⌬1-3 (each 3.6 M). However, for ⌬1-7 (U 1/2 ϭ 2.9 M) and ⌬1-9 (U 1/2 ϭ 2.45 M) there was a progressive shift to lower urea concentrations. Values for the free energy of unfolding estimated for each protein at 0 M urea buffer conditions ⌬G D H2O ) and the measure of dependence of ⌬G on urea concentration (M) were significantly lower for ⌬1-9 and ⌬1-7 than for wild type and ⌬1-3 (See Table IV).

Kinetics and Stability Measurements of Purified Truncation Mutants-
Inactivation and Reactivation of Purified Truncation Mutants-Although DTT is a reductant, at low DTT and low protein concentrations, it has the paradoxical effect of inactivating rhodanese by a complex series of reactions at the active site that include oxidation (38,39). This has been ascribed to the facts that 1) reduced DTT is analogous to the dithiol acceptor substrate, dihydrolipoic acid, and it can remove the persulfide sulfur from the rhodanese active site to produce a labile form of the enzyme; 2) the autooxidation of DTT leads to formation of partially reduced reactive oxygen species such as hydrogen peroxide, which can oxidize the active site of rhodanese, thereby inactivating the enzyme through a series of states that become progressively refractory to reduction; and 3) DTT, through its resemblance to dihydrolipoic acid, can specifically interact with the active site and form a stabilized disulfide-bonded adduct. Fig. 5 displays the percent activity of rhodanese as a function of DTT concentration. At the lowest DTT concentration of 20 M, 50% of wild type was inactivated, whereas no more than 10% of any of the mutants was inactivated. The remaining 50% of wild-type activity was more resistant to inactivation, but complete inactivation was achieved at 15 mM DTT (data not shown). Relative to this resistant phase of the wild-type inactivation, the mutants were less sensitive with ⌬1-3 being more sensitive than ⌬1-7 and ⌬1-9, which were almost equal. The relative sensitivity of the proteins can be defined in terms of the DTT concentration giving 50% inactivation. These were as follows: wild type (low sensitivity response), 7.5 mM; ⌬1-3, 1.3 mM; ⌬1-7 and ⌬1-9, 0.25 mM.
The ability of the sulfur-donor substrate, Na 2 S 2 O 3 , to reactivate the DTT-inactivated enzymes was studied. Fig. 6 displays the time dependence of the reactivation of rhodanese species that had been completely inactivated by DTT. Wild type and ⌬1-3 could be reactivated by Na 2 S 2 O 3 to 55-60% activity, whereas ⌬1-7 was reactivated to 40%. On the other hand, ⌬1-9 had a maximum reactivation of only 6%, suggesting that the active site of ⌬1-9 was able to be more highly oxidized, e.g., to a sulfinic acid, so that thiosulfate reactivation was less effec-tive. It is interesting that the maximum observed reactivation was only to a level that was observed for the less sensitive inactivation component of the wild-type protein.
Spontaneous inactivation at elevated temperature reflects reactivity at the active site of rhodanese (41,42). The spontaneous inactivation of the mutant and wild-type enzymes is shown in Fig. 7. Wild-type samples were inactivated in 2 h, whereas for ⌬1-3, inactivation took 6 h. ⌬1-7 and ⌬1-9 were incompletely inactivated even after 6 h at 37°C. Thus, the mutants were much less sensitive to oxidation than the wildtype protein. DISCUSSION It has been speculated that the amino-terminal 23 residues of mitochondrial rhodanese contribute to the global stability of the folded protein. In one study, for example, an E17P mutant was active but less stable than wild-type rhodanese, a result that was suggested to be due to destabilization of the ␣-helix proximal to the amino terminus (12). In another study, limited tryptic digestion could produce an active rhodanese in which the peptide composed of the NH 2 -terminal 45 residues was non-covalently bound. After denaturation, this form of the enzyme was not able to refold under conditions appropriate for reactivation of the native enzyme. These results were taken to indicate that portions of the NH 2 terminus are important for the folding of the enzyme to an active form (43). The present studies used site-directed deletions to extend these earlier suggestions and to investigate effects including (a) the role of the NH 2 -terminal region of rhodanese in the expression of active enzyme, i.e., its resistance to elimination in vivo; (b) the effect on the enzyme kinetic parameters of active mutants; and (c) the stabilities of mutants that could be purified.
Active rhodanese was expressed with deletions of up to the first nine residues. The crystal structure of wild-type rhodanese shows that these residues are situated parallel to, and form part of, the bi-domain interface, and they are the only residues out of the first 23 that interact hydrophobically with the carboxyl domain. These residues participate in a hydrophobic cluster that includes the following interactions: His-2 with Pro-266 and Asp-267; Leu-5 with Tyr-261, Leu-262, Gly-264, and Pro-266; Tyr-6 with Tyr-261; Ala-8 with Leu-258, Tyr-261, and Leu-262; and Leu-9 with Glu-222, Leu-258, and Leu-262. Hydrogen bonding of the first 9 residues is primarily with other residues of the NH 2 domain. Tyr-6 is the only one of these that H-bonds to the carboxyl domain, and it makes two H bonds to Tyr-261. Thus, it could be expected that these residues may be important for acquisition or maintenance of the active enzyme. Interestingly, the present results show that these residues are non-essential for the enzymatically active conformation during heterologous expression in E. coli or for the development of the kinetic parameters that characterize the wild-type enzyme. However, the results do indicate that deletions through residue TABLE III Rhodanese activity in cell-free, transcription translation expression system Activities were measured 30 min after induction of expression in the cell-free, coupled transcription/translation system as described under "Experimental Procedures" and previously (30 -32). Aliquots of the reaction mixtures were analyzed by SDS-PAGE followed by autoradiography. All quantities are mol min Ϫ1 mg Ϫ1 synthesized rhodanese in the reaction mixtures and are averages of at least three trials with standard deviations. Measurements that were not done are indicated by ND, and measurements that gave no activity after at least three trials are indicated with a 0.  IV Characteristics of purified rhodanese Percent yield of activity is percent recovery from initial cell lysis to final (NH 4 ) 2 SO 4 precipitation in the purification protocols. K m is for thiosulfate after at least three separate determinations with standard deviation. Specific activity is equivalent to V max for the enzymes purified to apparent homogeneity and is a result of at least three determinations with standard deviation. U 1/2 is the urea denaturation transition midpoint where half of the protein population is denatured and half is folded to the active conformation. ⌬G O H 2 O is an extrapolated estimate of the free energy of unfolding for the native state (0 M urea buffer conditions) of the protein (41). The m value is the measure of dependence of ⌬G on urea concentration (41 The crystal structure shows that residues 11-22 form an amphipathic ␣-helix that interacts hydrophobically solely with the NH 2 -terminal domain. In our experiments, truncations through Val-10 and beyond to Lys-23 resulted in mutant proteins that were rapidly degraded when expressed in E. coli. Deletion of the residues immediately preceding the start of the helix at S11 may destabilize the N-capping sequence region of the ␣-helix, which is oriented perpendicular to the bi-domain interface with the N-cap region closest to, and forming part of, the interface. Other interactions made by the ␣-helix are with residues in the region spanning from Ser-124 to His-138, which are also located predominately at the bi-domain interface region. Thus, it is likely that changes in the NH 2 -terminal sequence can perturb the global structure not only through effects on the NH 2 -terminal domain but also through effects on the bi-domain interface. That these effects cannot occur by direct interaction is suggested by the fact that Val-10 is Ͼ18 Å from the active site sulfhydryl group. All mutant proteins with deletions through Lys-23 were detected by immunoprecipitation as proteins of the predicted sizes (e.g., Fig. 2), indicating that removal of these residues did not result in cotranslational degradation in E. coli at 37°C. Further, the protocol included solubilizing conditions so that it was clear that the mutants were degraded and not transferred into inclusion bodies. When degradation occurred, no discrete intermediates were observed. At 37°C, deletions beyond residue 9 produced inactive enzyme both in vitro and in vivo, which was rapidly degraded in host cells. On the other hand, in the cell-free system, similar amounts of translated rhodanese, including ⌬1-15, were observed in all experiments, indicating that lack of activity is a consequence of failure to fold completely and not due to degradation. Expression in the cell-free system at 30°C showed that at least two more NH 2 -terminal residues, Val-10 and to a lesser extent Ser-11, could be removed and still permit synthesis of active enzyme. These two mutants were inactive when expressed at 37°C in the cell-free system; yet, when they were expressed at 30°C and then brought to 37°C for 45 min, the proteins retained activity. This implies that truncation influences folding to the active conformation rather than stability of the native state. A late step in the folding pathway may require prior formation of the Ser-11-Gly-22 helix. Residues Val-10 and Ser-11 could be essential for stabilization of this helix at 37°C but not at 30°C, thus conferring the temperature-sensitive folding phenotype that was observed.
The NH 2 -terminal sequence can modulate the unassisted and the chaperonin-assisted refolding of rhodanese in vitro (44). It has been suggested that the NH 2 -terminal sequence binds to the body of the NH 2 -terminal domain late in folding, when the protein has adopted a near-native conformation (12). The present study suggests that if there is insufficient NH 2terminal sequence left after truncation to aid in stabilization, and if the first ␣-helix is perturbed, then the final step from near-native to native conformation may be blocked. The NH 2terminal sequence through its ability to bind to the chaperonin GroEL may be involved in the signal that E. coli uses in degrading rhodanese (45,46).
The active site sulfhydryl group on Cys-247 is very reactive in part due to the conformation and side chain interactions in this region of the protein. Thus, Cys-247 can participate in complex reactions with other sulfhydryl groups (especially Cys-254 and Cys-263) in the protein and with reactive oxygen species (47)(48)(49)(50). All sulfhydryl groups are reduced in the active enzyme. The sensitivity of rhodanese to inactivation and its subsequent reactivation reflect the conformation and flexibility at the active site. This behavior is clear in the loss of activity observed at 37°C in which the greater the truncation, the less inactivation is observed, e.g., the rates of inactivation are in the order WTϾ⌬1-3Ͼ⌬1-7Ͼ⌬1-9. The inactivation with the sulfur acceptor substrate, DTT, although incompletely understood, has been shown also to reflect specific interactions at the active site. The results here demonstrate that DTT inactivation is clearly influenced by the degree of truncation near the NH 2 terminus. These inactivations are particularly noteworthy since activity differences are difficult to discern in the folded proteins with these truncations. Thus, events at the active site Cys-247, which is on the C-terminal domain of this 296-residue protein, are influenced by the state of the first 9 residues of the protein, thus suggesting that the effect of a local structural perturbation in the NH 2 -terminal domain can be transmitted to the active site.
The present work supports several conclusions. Although residues Val-1-Leu-9 may be essential for mitochondrial import of rhodanese, they are not required for correct folding to the native state when the protein is expressed in E. coli, although they do affect the stability of the purified protein. The amino-proximal ␣-helix in bovine rhodanese contributes to folding, global stability, and resistance to degradation in addition to any function in import. The ability of the NH 2 -terminal sequence of rhodanese to perform these multiple functions may help to explain why it is not removed after mitochondrial import.