Journal of Protein Chemistry, Vol. 19, No. 5, 2000

Differences in the Denaturation Behavior of Ribonuclease A Induced by Temperature and Guanidine Hydrochloride Ulrich Arnold1,2 and Renate Ulbrich-Hofmann1,3

Received June 7, 2000

Moderate temperatures or low concentrations of denaturants diminish the catalytic activity of some enzymes before spectroscopic methods indicate protein unfolding. To discriminate between possible reasons for the inactivation of ribonuclease A, we investigated the influence of temperature and guanidine hydrochloride on its proteolytic susceptibility to proteinase K by determining the proteolytic rate constants and fragment patterns. The results were related to changes of activity and spectroscopic properties of ribonuclease A. With thermal denaturation, the changes in activity and in the rate constants of proteolytic degradation coincide and occur slightly before the spectroscopically observable transition. In the case of guanidine hydrochloride-induced denaturation, however, proteolytic resistance of ribonuclease A initially increases accompanied by a drastic activity decrease far before unfolding of the protein is detected by spectroscopy or proteolysis. In addition to ionic effects, a tightening of the protein structure at low guanidine hydrochloride concentrations is suggested to be responsible for ribonuclease A inactivation. KEY WORDS: Ribonuclease A; limited proteolysis; temperature; guanidine hydrochloride; unfolding.

1. INTRODUCTION

via ubiquitination and proteasomes (Ciechanover, 1998), and the conformational rearrangement of prion proteins and Alzheimer’s β-amyloid peptides (Cohen and Prusiner, 1998) inspired investigations of the unfolding processes. In addition, the biotechnological use of enzymes and the optimization of their properties require knowledge about their stability and unfolding under the conditions of practical application (Janecek, 1993; Imoto, 1997; Ulbrich-Hofman et al., 1999). Because of the cooperative nature of protein denaturation, the protein structure tolerates changes of temperature or addition of chemical denaturants within certain limits, beyond which global unfolding occurs. Even in the range of tolerance, however, the dynamic structure of the native protein is influenced and minor local structural

In contrast to the thoroughly studied protein folding processes (Anfinsen, 1972; Creighton, 1990), the investigation of unfolding processes and the forces responsible for maintaining the native protein structure have drawn increasing attention only recently. Findings like the recognition of (partly) unfolded structures by molecular chaperones (Fenton and Horwich, 1997), the requirement of protein unfolding for membrane crossing (Schatz, 1996), the proteolytic degradation of proteins

1

Department of Biochemistry/Biotechnology, Martin-Luther University Halle-Wittenberg, D-06120 Halle, Germany. 2 Present address: Department of Biochemistry, University of WisconsinMadison, Madison, Wisconsin 53706. 3 To whom correspondence should be addressed at Department of Biochemistry/Biotechnology, Martin-Luther University HalleWittenberg, Kurt-Mothes-Str. 3, D-06120 Halle, Germany. e-mail: [email protected]

4

Abbreviations: CAM-RNase A, carbamidomethylated ribonuclease A; DTE, 1,4-dithioerythreitol; GdnHCl, guanidine hydrochloride; NaDOC, sodium deoxycholate; RNase A, ribonuclease A; TCA, trichloroacetic acid.

345 0277-8033/00/0700-0345$18.00/0 © 2000 Plenum Publishing Corporation

346 changes may occur. These effects are not reflected in the spectroscopic properties of the enzyme, but are often manifested in changes of the activity (Creighton, 1990; Tsou, 1995). In case of chemical denaturants, however, the enzymatic activity may be also impaired by binding of denaturant molecules, which thus act as inhibitors. To differentiate these effects, methods are valuable that allow the observation of conformational changes of the protein molecule on the level of structural regions. Limited proteolysis has proven to be a very sensitive tool for detecting local conformational changes of the protein structure (Price and Johnson, 1990). In addition to the specificity of the protease used, the proteolytic susceptibility of proteins strongly depends on the flexibility and accessibility of the structural region to be investigated (Hubbard, 1998). While specific proteases are mostly not able to cleave proteins in their native compact conformation, but require unfolding of the protein molecule by denaturants (Hubbard, 1998; Arnold et al., 1996), unspecific proteases often attack even native proteins. This attack is possible due to the large number of theoretically possible cleavage sites and the less stringent requirement for the number and conformation of subsites. Since flexible regions of the protein molecule are cleaved more probably, localization of the primary cleavage sites yields information on the conformation of the protein molecule (Hubbard, 1998). Under denaturing conditions resulting in a global unfolding of the protein molecule, the proteolytic fragment pattern changes and becomes more complex because of an increase of accessible peptide bonds (Wang et al., 1998). Additionally, the rate of proteolytic degradation increases (Yang and Tsou, 1995; Yang et al., 1998). As known so far, ribonuclease A (RNase A)4 is cleaved by unspecific proteases such as subtilisin (Richards and Vithayathil, 1959), elastase (Klee, 1965), and proteinase K (Arnold et al., 1996) under native conditions (25°C, pH 8.0) at the peptide bonds before or after Ala20. Although RNase A is rather thermostable (transition temperature Tm = 60.4°C; Arnold and Ulbrich-Hofmann, 1997) and considerably resistant to denaturation by GdnHCl (free energy of unfolding ∆G becomes zero at about 3.0 M; Pace et al., 1990), this cleavage is enabled due to a high flexibility of the loop region around Ala20 (Wlodawer, 1982; Rico et al., 1989). RNase A activity, in contrast, already drastically decreases at relatively low GdnHCl concentrations (Liu and Tsou, 1987; Yang and Tsou, 1995; Kasumov et al., 1990), the reasons for which are controversial. In the present paper, both qualitative and quantitative analyses of the proteolytic degradation by the unspecific proteinase K have been used to ascertain the

Arnold and Ulbrich-Hofmann structural regions (RNase A) affected by temperature and guanidine hydrochloride (GdnHCl). The changes of the proteolytic rate constants and of the primary cleavage sites in the degradation of RNase A by proteinase K are discussed in relation to spectroscopic investigations as well as measurements of ribonucleolytic activity, thus providing new insight into specific denaturant effects.

2. MATERIALS AND METHODS 2.1. Materials RNase A was purchased from Sigma and purified to homogeneity using an FPLC MONO S column (Pharmacia). Proteinase K, soybean trypsin inhibitor, cytochrome c, and aprotinin were purchased from Sigma. GdnHCl (ultra pure) was from ICN. 1,4-Dithioerythreitol (DTE) was from Sigma. All other chemicals were the purest grade commercially available.

2.2. Denaturation and Proteolysis To 80 µl of 50 mM Tris–HCl buffer, pH 8.0 (25– 65°C, containing 0–6.25 M GdnHCl), 10 µl of proteinase K (2 µg ml−1 to 2 mg ml−1) in 50mM Tris–HCl buffer, 10 mM CaCl2, pH 8.0, and 10 µl of RNase A (2.0 mg ml−1) were added. After defined time intervals, samples (10 µl) were removed, mixed with 5 µl of 50 mM phenylmethanesulfonyl fluoride (dissolved in isopropanol), and dried under nitrogen.

2.3. Precipitation, SDS–PAGE, Densitometric Evaluation, RP-HPLC, and MALDI-MS Desalting and removal of GdnHCl were performed by precipitation with sodium deoxycholate (NaDOC)/ trichloroacetic acid (TCA) according to Arnold and Ulbrich-Hofmann (1999). SDS–PAGE and densitometric evaluation were carried out as described in Arnold and Ulbrich-Hofmann (1997). RP-HPLC and MALDI mass spectrometry were performed as described previously (Arnold et al., 1998).

2.4. Determination of Proteolysis Rate Constants The rate constants of proteolysis (kp) were calculated from the decrease of the peak areas of intact RNase A in the scanned SDS–PAGE gels as a function of time of

Denaturation Behavior of RNase A proteolysis, which followed pseudo-first-order reactions. Determination of kp values was made from seven data points by nonlinear regression and repeated at least twice. The obtained kp values were corrected by the proteinase K activity and linearly transformed to an RNase A/proteinase K ratio of 100:1 (w/w) to yield kp′, which was proven to be permissible from studies on the dependence of kp on the proteinase K concentration under all conditions applied. 2.5. Preparation of Carbamidomethylated Ribonuclease A (CAM-RNase A) In a typical preparation, 75 µl of RNase A (2 mg ml−1 50 mM Tris–HCl buffer, pH 8.0) was mixed with 20 µl of 1.88 M Tris–HCl buffer, pH 8.8, 100 µl of 8 M aqueous GdnHCl, and 10 µl of 200 mM aqueous DTE. After incubation for 2 hr, iodoacetamide (3.6 mg) was added followed by a further incubation of 30 min. Both operations were performed in the dark under nitrogen at room temperature. Desalting was performed by precipitation with NaDOC/TCA (Arnold and Ulbrich-Hofmann, 1999). CAM-RNase A was resolubilized in 75 µl of 50 mM Tris–HCl buffer, pH 8.0, containing 0.5 M GdnHCl. 2.6. Activity Assay for Proteinase K The kp of CAM-RNase A was used as a measure of proteinase K activity. kp values were determined for different temperatures or GdnHCl concentrations in the same way as described above for RNase A.

347 Calculation of ∆G values was done by use of the Gibbs– Helmholtz equation for thermal denaturation and by use of the linear extrapolation method (Santoro and Bolen, 1988) for GdnHCl-induced denaturation as described in Arnold and Ulbrich-Hofmann (1997).

3. RESULTS 3.1. Proteinase K Activity as a Function of Temperature or GdnHCl Concentration If rate constants of the proteolytic degradation of a protein are to be used as an indicator of conformational changes, the protease activity must be identical under all conditions or the rate constants of proteolysis must be corrected regarding activity changes of the protease. For this reason, we measured the effect of temperature (25–65°C) and concentration of GdnHCl (0–5.0 M at 25°C) on kp of proteinase K acting on CAM-RNase A. CAM-RNase A has the same amino acid sequence as RNase A, but lacks tertiary and secondary structure and disulfide bonds (Qi et al., 1998) which could impede proteolytic degradation. Therefore, it was chosen as the most authentic substrate with respect to the degradation of RNase A studied in this paper. As can be seen in Fig. 1, protease activity is hardly changed with increasing temperatures or in the presence of ≤1.0 M GdnHCl. In the presence of more than 1.0 M GdnHCl, proteinase K activity toward CAM-RNase A gradually decreases (Fig. 1). In the following, the relative activities obtained from this analysis have been used for the correction of

2.7. Activity Assay for RNase A RNase A (5–100 µl of a 38.15 µM solution in 50 mM Tris–HCl buffer, pH 8.0) was added to solutions containing 2′,3′cCMP and GdnHCl, NaCl, or urea in 50 mM Tris–HCl buffer, pH 8.0, to give a final volume of 400 µl with 0.3 mM 2′,3′cCMP and 0–2 M GdnHCl, 0–3 M NaCl, or 0–5 M urea. The changes in absorbance were followed in a thermostated quartz cuvette (1 cm length) at 286 nm using a Hitachi U2000 spectrophotometer. kcat and KM were determined from initial reactions rates for 0– 0.6 mM 2′,3′cCMP according to the Michaelis– Menten equation by nonlinear regression. For calculation of kcat, ∆ε = 1450 M−1 cm−1 was used (delCardayré and Raines, 1995). 2.8. UV Spectroscopy and Determination of ∆G Data for transition curves determined at 287 nm were taken from Arnold and Ulbrich-Hofmann (1997).

Fig. 1. Residual activity of proteinase K toward CAM-RNase A as a function of temperature (●) and GdnHCl concentration (■).

348

Arnold and Ulbrich-Hofmann

the kp values measured for RNase A to yield the corresponding kp′ values. 3.2. Conformational and Activity Changes of RNase A as a Function of Temperature 3.2.1. Proteolytic Susceptibility Even under native conditions (25°C, pH 8.0), RNase A is degraded by proteinase K as determined by densitometric evaluation of Coomassie-stained SDS–PAGE gels (Fig. 2A). Up to 45°C, there is only a slight increase in kp′. With further increasing temperature, however, kp′ strongly increases, indicating a changed manner of RNase A proteolysis. At 65°C, the kp′ value for RNase A approaches that of CAM-RNase A (4.26 × 10−2 s−1), reflecting the upper limit of proteolysis rate under the applied conditions. 3.2.2. Activity The activity of RNase A with 2′,3′cCMP as a substrate shows an optimum near 40°C. Below this temperature, RNase A activity only slightly decreases, but a strong decrease was detected at temperatures above 40°C (Fig. 2A). 3.2.3. UV Spectroscopy The thermal transition curve of RNase A reveals a pretransition region up to 50°C and a sharp transition

zone between 56°C and 64°C (Fig. 2A) with a transition temperature at 60.4°C. The corresponding ∆G values are indicated as perpendicular lines. 3.2.4. Proteolytic Fragmentation As shown in Figs. 3a and 3b, the proteolytic fragment pattern of RNase A by proteinase K changes with increasing temperature. As verified by MALDI-MS analysis after HPLC separation of the fragments, under native conditions (below 35°C; ∆G > 35 kJ mol−1), RNase A is cleaved by proteinase K to yield Lys1–Ala20 and Ser21–Val124 as the only pair of complementary N- and C-terminal fragments. These fragments identify Ala20 – Ser21 as the only primary cleavage site under native conditions. In the SDS–PAGE gel, only the 11.5-kDa fragment Ser21–Val124 can be detected. Under denaturing conditions, on the other hand, several fragments were detected by SDS–PAGE as demonstrated for 65°C in Fig. 3b. By MALDI-MS analysis, numerous fragments could be aligned to the RNase A sequence. In addition to fragments resulting from secondary cleavage events, seven pairs of complementary N- and C-terminal fragments were identified suggesting the following peptide bonds as primary cleavage sites: Ala20–Ser21, Arg33–Asn34, Phe46–Val47, Gln55–Ala56, Val57– Cys58, Tyr76–Ser77, and Met79–Ser80 (Fig. 4). By SDS–PAGE, these fragments could not be completely separated since some of the larger fragments seem to comigrate due to the small differences in their molecular weights (Fig. 3b). The smaller, complementary fragments

Fig. 2. Spectroscopic transition (fN) of RNase A (●), residual activity toward 2′,3′cCMP (u), and rate of proteolytic degradation by proteinase K (▲) as a function of temperature (A) and GdnHCl concentration (B). The numbers in parentheses denote ∆G values (kJ mol−1) derived from spectroscopic data (Arnold and UlbrichHofmann, 1997).

Denaturation Behavior of RNase A

349 3.3. Conformational and Activity Changes of RNase A as a Function of GdnHCl Concentration 3.3.1. Proteolytic Susceptibility

Fig. 3. SDS–PAGE gels of RNase A incubated in the presence of proteinase K (a) at a ratio of 100:1 (w/w) at 25°C for 60 min, (b) at a ratio of 1000:1 (w/w) at 65°C for 5:30 min, and (c) at a ratio of 100:1 (w/w) at 25°C, in the presence of 5 M GdnHCl for 35 min. (d) Reference proteins soybean trypsin inhibitor (21 kDa), cytochrome c (12.4 kDa), and aprotinin (6.5 kDa).

At 25°C, between 0 and 1.5 M GdnHCl, a slightly decreased susceptibility of RNase A to proteinase K can be detected by densitometric evaluation of Coomassiestained SDS–PAGE gels (Fig. 2B). The minimum is at 1.0 M GdnHCl, where kp′ decreases to 64 ± 9% of that at 0 M GdnHCl. Beyond the minimum, kp′ strongly increases approaching the kp′ value of CAM-RNase A (4.26 × 10−2 s−1) above 4.0 M GdnHCl. This reflects the upper limit of the RNase A degradation rate under the applied conditions. To interpret the unexpected initial increase of the proteolytic resistance, we compared the degradation of RNase A by proteinase K in the presence of GdnHCl to that in the presence of NaCl as ionic component and urea as chaotropic agent (Fig. 5A) as suggested recently (Fan et al., 1999). While addition of NaCl did not lead to an altered degradation rate compared to that in the absence of additives, addition of urea resulted in nearly the same decrease of the degradation rate as observed with the addition of GdnHCl (Fig. 5A). Similarly, addition of 17.5 mM 2′,3′cCMP resulted in a decrease of the degradation rate by more than 50% (not shown). 3.3.2. Activity

Fig. 4. Tertiary structure of RNase A. The model was taken from the Brookhaven protein data bank and drawn with MOLSCRIPT (Kraulis, 1991). Arrows indicate positions of suggested primary cleavage sites for proteinase K at 65°C or 5 M GdnHCl. Asterisk indicates primary cleavage site Ser89–Ser90, which was found in 5 M GdnHCl only.

cannot be seen in the SDS–PAGE. At intermediate temperatures (40–55°C), no large fragments could be found in SDS–PAGE or by MALDI-MS (not shown). Therefore, we examined the proteolytic degradation of RNase S, which is the result of the first cleavage in the native RNase A (Richards and Vithayathil, 1959) and has a lower thermal stability (Tm = 47.7°C; Tsong et al., 1970; Neira et al., 1999) than intact RNase A. RNase S was found to be degraded 5–50 times faster than RNase A under these conditions. From this result, it can be concluded that the absence of RNase A fragments at 40–55°C is due to the fast degradation of the arising cleavage products.

The activity of RNase A toward 2′,3′cCMP rapidly decreases with the addition of GdnHCl (Fig. 2B) in a clear biphasic manner which can be described by the sum of two exponential terms. To interpret the effects induced by GdnHCl, we also determined the dependence of RNase A activity on NaCl and urea concentration (Fig. 5B). Addition of urea results in a monophasic inactivation profile, whereas addition of NaCl results in a biphasic inactivation profile. Comparison of the curves reveals that the initial phases of activity loss with the addition of GdnHCl and NaCl roughly coincide. RNase A activity loss with the addition of urea, however, proceeds much more gradually and is similar to the second phase of RNase A activity loss under addition of GdnHCl. Further information was obtained from the determination of the Michaelis–Menten parameters of RNase A in the presence of 1 M GdnHCl, NaCl, or urea in comparison to that in buffer alone (Table I). The transition curve (Fig. 2B) reveals that the concentration of native enzyme under these conditions corresponds to E0. Here, kcat remains unchanged in the presence of urea, whereas kcat in the presence of both GdnHCl and NaCl decreases

350

Arnold and Ulbrich-Hofmann

Fig. 5. (A) Rate of proteolytic degradation of RNase A by proteinase K and (B) residual activity of RNase A toward 2′,3′cCMP as a function of the concentration of GdnHCl (●), NaCl (■), and urea (▲).

Table I. Michaelis –Menten Parameters of RNase A for 2′,3′cCMP in the Presence of GdnHCl, NaCl, and Urea Without additive −1

kcat (s ) 1.18 ± 0.24 KM (mM) 1.8 ± 0.5

1 M GdnHCl

1 M NaCl

1 M urea

0.50 ± 0.18 6.5 ± 2.6

0.48 ± 0.19 2.3 ± 1.0

1.18 ± 0.35 3.6 ± 1.3

to about 40%. On the other hand, KM is only slightly increased in the presence of NaCl, whereas the KM values of RNase A in the presence of urea and GdnHCl increase by factors of 2.0 and 3.6.

3.3.3. UV Spectroscopy The transition curve reveals a pretransition region up to 2.0 M GdnHCl and a midpoint (GdnHCl1/2 for [N] = [D]) at 3.0 M GdnHCl. The urea-induced transition (not shown), on the other hand, possesses a midpoint at 7.5 M urea, i.e., the denaturing strength of urea is 2.5 times less than that of GdnHCl in the case of RNase A.

3.3.4. Proteolytic Fragmentation As for 25°C under thermal denaturation, the analysis of the degradation of RNase A by proteinase K below 1.0 M GdnHCl (Fig. 3a; ∆G > 28.5 kJ mol−1) identified Ala20–Ser21 as the only primary cleavage site. At 5.0 M GdnHCI (Fig. 3c), apart from numerous fragments resulting from secondary cleavage events, the following seven new primary cleavage sites were localized by identification of the complementary N- and C-terminal frag-

ments by MALDI-MS: Arg33–Asn34, Phe46–Val47, Gln55–Ala56, Val57–Cys58, Tyr76–Ser77, Met79– Ser80, and Ser89–Ser90. This cleavage pattern denotes one additional pair of primary fragments in comparison to thermal denaturation, namely Lys1–Ser89 + Ser90– Val124. The fragment pattern in SDS–PAGE after limited proteolysis in 5 M GdnHCl (Fig. 3c) is similar to that after thermal denaturation at 65°C (Fig. 3b). The additional fragment Ser90–Val124 is too small to be detected by SDS–PAGE, while the fragment Lys1–Ser89 likely comigrates with the Asn34–Val124 fragment. At intermediate GdnHCl concentrations (1.5–3.0 M), no large fragments could be found. Proteolytic degradation of RNase S (not shown) revealed that it is degraded 5–30 times faster than RNase A under these conditions. From this result, it can be concluded that the absence of RNase A fragments at 1.5–3.0 M GdnHCl is due to the fast degradation of the arising cleavage products, as found for thermal denaturation at 40–55°C.

4. DISCUSSION The comparison of changes of several characteristics which reflect the integrity of the protein molecule as a function of the concentration of denaturants allows the differentiation between local and global effects of the denaturants on the protein structure. Furthermore, the dependence of these parameters on the kind of denaturant can disclose specific denaturant effects. With thermal denaturation of RNase A, remarkable enzyme inactivation commences at 45°C, whereas global unfolding can be detected only above 55°C (Fig. 2A). In the same way as RNase A activity decreases, prote-

Denaturation Behavior of RNase A olytic susceptibility toward proteinase K increases. Consequently, limited proteolysis detects subtle local conformational changes of the RNase A molecule that result in a decreased catalytic activity, but do not significantly influence its spectroscopic properties. These changes are attributed to increased local fluctuations and/or limited unfolding of confined regions. The distribution of the primary cleavage sites under denaturing conditions (Fig. 4) reveals that here structural changes affect the main part of the protein molecule. Fortunately, the identification of the Phe46–Val47 peptide bond as a primary cleavage site under denaturing conditions gives a sensitive probe for detecting changes of the RNase A active site which includes the neighbored residue Thr45 in the B1 binding site (Nogués et al., 1995). Unfolding of the active site therefore results in an exposure of the Phe46–Val47 peptide bond. The influence of GdnHCl, on the other hand, is more complex. The consideration of the protease activity allowed again the quantification of the structural changes which the RNase A molecule undergoes due to the influence of the denaturant. While the increase of the proteolytic susceptibility occurs again slightly before global changes are detected by spectroscopy, activity loss is already observed at very low concentrations of GdnHCl (Fig. 2B). It is noteworthy that at ∆G = 30 kJ mol−1 (0.9 M GdnHCl) RNase A activity is drastically decreased (to 18%), whereas proteolytic resistance persists and even increases. In contrast, with thermal denaturation at ∆G = 30 kJ mol−1 (39.5°C), both the activity and the rate of proteolytic degradation of RNase A are marginally increased in comparison with their values at 25°C. By comparing the changes of RNase A activity in GdnHCl with those in NaCl and urea to differentiate ionic and chaotropic effects of GdnHCl, we found that at least half of the activity loss is caused by ionic effects. Indications for similar interpretations were obtained upon comparing GdnHCl-induced inactivation of RNase A with that induced by LiCl (Kasumov et al., 1990). Additionally, the activity of RNase A is also affected by the nonionic chaotrope urea due to an increase of KM, which is similar to that in the presence of GdnHCl. Since, at least up to 1.0 M GdnHCl, Ala20–Ser21 was identified as the only primary cleavage site, unfolding of the RNase A molecule can be ruled out as a reason for its inactivation. A distortion of the “network of water” in the active site (Brunger et al., 1985), interferences with the Coulombic interactions responsible for substrate binding and catalysis (Fisher et al., 1998), and/or an increased rigidity (Sackett et al., 1994) of the RNase A molecule might be reasons for this inactivation. The latter argument is supported by the findings of a decreased kp′ in the presence of low concentrations of

351 GdnHCl (with a minimum of kp′ at 1 M GdnHCl), whereas kp′ is not influenced by NaCl. Since changes of the proteinase K activity have been taken into account, these effects have to be unambigiously attributed to changes of the RNase A molecule. In contrast to the thermal denaturation, where the unfolding of the active site is accompanied by an increased cleavage rate of the remote Ala20–Ser21 peptide bond, the decelerated cleavage rate of this peptide bond in the presence of low concentrations of GdnHCl points to a loss of flexibility of the active site which contributes to the inactivation of the enzyme. An interplay between the active site and the remote Ala20–Ser21 peptide bond is also observed if proteolysis is followed in the presence of 2′,3′cCMP. The interaction of the substrate with the P1 and B1 binding sites in the RNase A molecule (Nogués et al., 1995) results in a decelerated cleavage rate by proteinase K. These findings, however, exclude local unfolding of the active site from causing enzyme inactivation in the pretransition region. Sackett et al. (1994) also interpreted similar results on tubulin in the presence of low concentrations of urea as due to a “tightening” of the protein structure. Moreover, the exposure of the Phe46–Val47 peptide bond to proteinase K digestion as a probe of the integrity of the active site occurs only at considerably higher GdnHCl concentrations than RNase A inactivation, again indicating that the inactivation of RNase A at low GdnHCl concentrations is not coupled with an unfolding of the active site. While our results on RNase A inactivation in GdnHCl confirm those by the group of Tsou (Liu and Tsou, 1987; Yang and Tsou, 1995), our results on proteolysis of RNase A are in distinct contradiction. These authors found an increased proteolytic susceptibility of RNase A even at 1 M GdnHCl and held local unfolding of the active site responsible for the inactivation of the enzyme at low GdnHCl concentrations. While we applied 25°C throughout all experiments concerning the GdnHCl dependence of RNase A, the different temperatures (0–35°C) applied by Tsou et al. for the activity assay and proteolytic degradation have not been taken into account. Even though the enzyme appears to be native by spectroscopy, the resulting ∆G values considerably differ due to their temperature dependence. Consequently, the increased rate of proteolytic degradation in 1 M Gdn HCl at 35°C has likely to be attributed to a general decreased enzyme stability (∆G) and not to local unfolding of the active-site region. In summary, under thermal denaturation, the decrease of RNase A activity coincides with the increase of its proteolytic susceptibility to proteinase K. Both processes

352 slightly precede global unfolding due to local conformational changes. The localization of the primary cleavage sites under denaturing conditions reveals that here the main part of the RNase A molecule becomes accessible to proteinase K. In the presence of low concentrations of GdnHCl, inactivation of RNase A is accompanied by a decrease of the proteolytic susceptibility, whereas its spectroscopic properties remain unchanged. Here, inactivation of RNase A cannot to be attributed to local unfolding, but is probably caused by an increased rigidity of the molecule in addition to disturbances influencing kcat by ionic effects. ACKNOWLEDGMENTS We thank M. Kipping, Max-Planck-Forschungsstelle “Enzymologie der Proteinfaltung,” Halle, Germany, for performing MALDI-MS measurements. The work was supported by the Deutsche Forschungsgemeinschaft, Bonn, Germany (Ul 130/2–3). REFERENCES Anfinsen, C. B. (1972). Biochem. J. 128, 737–749. Arnold, U. and Ulbrich-Hofmann, R. (1997). Biochemistry 36, 2166–2172. Arnold, U. and Ulbrich-Hofmann, R. (1999). Anal. Biochem. 271, 197–199. Arnold, U., Rücknagel, K.-P., Schierhorn, A., and Ulbrich-Hofmann, R. (1996). Eur. J. Biochem. 237, 862–869. Arnold, U., Schierhorn, A., and Ulbrich-Hofmann, R. (1998). J. Protein Chem. 17, 397– 405. Brunger, A. T., Brooks, C. L., and Karplus, M. (1985). Proc. Natl. Acad. Sci. USA 82, 8458–8462. Ciechanover, A. (1998). EMBO J. 17, 7151–7160. Cohen, F. E. and Prusiner, S. B. (1998). Annu. Rev. Biochem. 67, 793–819. Creighton, T. E. (1990). Biochem. J. 270, 1–16.

Arnold and Ulbrich-Hofmann delCardayré, S. B. and Raines, R. T. (1995). J. Mol. Biol. 252, 328–336. Fan, X.-Y., McPhie, P., and Miles, E. W. (1999). Biochemistry 38, 7881–7890. Fenton, W. A. and Horwich, A. L. (1997). Protein Sci. 6, 743–760. Fisher, B. M., Schultz, L. W., and Raines, R. T. (1998). Biochemistry 37, 17386–17401. Hubbard, S. J. (1998). Biochim. Biophys. Acta 1382, 191–206. Imoto, T. (1997). Cell. Mol. Life Sci. 53, 215–223. Janecek, S. (1993). Proc. Biochem. 28, 435– 445. Kasumov, E. A., Volynskaya, A. V., Shishkov, A. V., and Goldanskii, V. I. (1990). Studia Biophys. 136, 167–170. Klee, W. A. (1965). J. Biol. Chem. 240, 2900 –2906. Kraulis, P. J. (1991). J. Appl. Crystallogr. 24, 946–950. Liu, W. and Tsou, C.-L. (1987). Biochim. Biophys. Acta 916, 455– 464. Neira, J. L., Sevilla, P., Menéndez, M., Bruix, M., and Rico, M. (1999). J. Mol. Biol. 285, 627–643. Nogués, M. V., Vilanova, M., and Cuchillo, C. M. (1995). Biochim. Biophys. Acta 1253, 16–24. Pace, C. N., Laurents, D. V., and Thomson, J. A. (1990). Biochemistry 29, 2564 –2572. Price, N. C. and Johnson, C. M. (1990). In Proteolytic Enzymes—A Practical Approach (Beynon, R. J., and Bond, J. S., eds.), IRL Press, Oxford, 1990, pp. 163–180. Qi, P. X., Sosnick, T. R., and Englander, S. W. (1998). Nature Struct. Biol. 5, 882–884. Richards, F. M. and Vithayathil, P. J. (1959). J. Biol. Chem. 270, 17180 –17184. Rico, M., Bruix, M., Santoro, J., Gonzales, C., Neira, J. L., Nieto, J. L., and Herranz, J. (1989). Eur. J. Biochem. 183, 623–638. Sackett, D. L., Bhattacharyya, B., and Wolff, J. (1994). Biochemistry 33, 12868–12878. Santoro, M. M. and Bolen, D. W. (1988). Biochemistry 27, 8063–8068. Schatz, G. (1996). J. Biol. Chem. 271, 31763–31766. Tsong, T. Y., Hearn, R. P., Wrathal, D. P., Sturtevant, J. M. (1970). Biochemistry 9, 2666–2677. Tsou, C.-L. (1995). Biochim. Biophys. Acta 1253, 151–162. Ulbrich-Hofmann, R., Arnold, U., and Mansfeld, J. (1999). J. Mol. Catal. B Enzymatic 286, 125–131. Wang, L., Chen, R. X., and Kallenbach, N. R. (1998). Proteins 30, 435– 441. Wlodawer, A., Bott, R., and Sjolin, L. (1982). J. Biol. Chem. 257, 1325–1332. Yang, H.-J. and Tsou, C.-L. (1995). Biochem. J. 305, 379–384. Yang, H.-J., Li, X. C., Amft, M., and Grotemeyer, J. (1998). Anal. Biochem. 258, 118–126.