Reversible inactivation of pancreatic deoxyribonuclease A by sodium dodecyl sulfate. Removal of COOH-terminal residues from the denatured protein by carboxypeptidase A.

In the course of experiments on the role of the COOH-terminal residues in pancreatic deoxyribonuclease, we undertook to ascertain whether the presence of sodium dodecyl sulfate would render the normally unavailable terminus susceptible to hydrolysis by carboxypeptidase A. When DNase A is dissolved in 0.005% sodium dodecyl sulfate the protein becomes enzymically inactive when assayed against DNA in the same sodium dodecyl sulfate concentration. The loss of activity caused by treatment with sodium dodecyl sulfate for 1 hour at 45 degrees can be fully restored if the detergent-containing solution is diluted 10-fold into 6 M guanidinium chloride and then 10-fold into a pH 7.0 buffer, 10 mM in CaCl2, prior to a 100-fold dilution for assay. The presence of Ca2+ is essential for the refolding process. If the same degree of dilution is made into sodium dodecyl sulfate-free buffer without the guanidinium chloride step, there is very little reversal of the inactivation. An almost complete loss of regenerable activity is caused by 1 hour of digestion by carboxypeptidase at 45 degrees in the presence of 0.03% sodium dodecyl sulfate. Although up to 6 amino acid residues can be removed from the COOH terminus, the loss of activity can be correlated with the removal of either 1 or 2 amino acid residues (-Leu-Thr) from the COOH-terminal sequence. Thus, DNase A is one of the several enzymes in which residues at the COOH terminus are essential to the active conformation. If the enzyme minus 2 to 6 terminal residues was mixed with a 15-residue COOH-terminal peptide (obtained by cyanogen bromide cleavage), only about 2% activity could be regenerated.

inactive when assayed against DNA in the same sodium dodecyl sulfate concentration.
The loss of activity caused by treatment with sodium dodecyl sulfate for 1 hour at 45" can be fully restored if the detergentcontaining solution is diluted lo-fold into 6 M guanidinium chloride and then lo-fold into a pH 7.0 buffer, 10 mM in CaC12, prior to a loo-fold dilution for assay. The presence of Ca2+ is essential for the refolding process. If the same degree of dilution is made into sodium dodecyl sulfate-free buffer without the guanidinium chloride step, there is very little reversal of the inactivation.
An almost complete loss of regenerable activity is caused by 1 hour of digestion by carboxypeptidase at 45" in the presence of 0.03 % sodium dodecyl sulfate. Although up to 6 amino acid residues can be removed from the COOH terminus, the loss of activity can be correlated with the removal of either 1 or 2 amino acid residues (-Leu-Thr) from the COOH-terminal sequence. Thus, DNase A is one of the several enzymes in which residues at the COOH terminus are essential to the active conformation. If the enzyme minus 2 to 6 terminal residues was mixed with a 15-residue COOH-terminal peptide (obtained by cyanogen bromide cleavage), only about 2% activity could be regenerated.
The present experiments began with the aim of studying the essentiality of residues near the COOH-terminal end of bovine pancreatic deoxyribonuclease A  Earlier experiments have shown that one of the disulfide bonds is essential for activity (3,4) and that modification of His-118 (3,5) or Tyr-62 (3,6) results in complete inactivation of the enzyme. Controlled proteolysis by chymotrypsin (7) has led to a single cleavage of the chain between residues 178 and 179, without inactivation, and 5 amino acid residues could be removed by carboxypeptidase Y at this point of cleavage without loss of activity. The COOH terminus of the native enzyme is unavailable to carboxypeptidase action (7). We have undertaken to unfold the protein sufficiently by sodium dodecyl sulfate to render the COOH terminus susceptible to hydrolysis by carboxypeptidase A; this approach has led to a study of the reversibility of the denaturation by Na dodecyl-S04,1 the effects of other denaturing agents on DNase, and the consequence of removal of residues from the COOH terminus of the enzyme. Successful renaturation of several other enzymes from solution in 6 M urea after denaturation by Na dodecyl-SO4 has been described by Weber and Kuter (8).

Materials-DNase
A was prepared by chromatography on cellulose phosphate according to the procedure of Salnikow el al. (l), starting with DP grade bovine pancreatic DNase from Worthington. Traces of proteolytic act,ivity were removed by affinity elution with Ca*+ from DEAE-cellulose (9). Calf thymus DNA and carboxypeptidase A (COADFP grade) were from Worthington.
DNase Assay-The activity was measured by the hyperchromicity method of Kunitz (12), modified for use at pH 7 (9). The rate of hydrolysis was measured at 260 nm on a recording spectro- and CNBr-5 (the COOH-terminal peptide) were partially separated; the material isolated from the right half of the peak contained CNBr-5 and 23oj, (on a molar basis) CNBr-1, as judged from amino acid analysis.

RESULTS
E$ect of Na Dodecyl-SO4 on DNase-When DNase is exposed to 0.005% Na dodecyl-SO4 at 25" for 1 min (Fig. 1 Under these conditions, in which the same Na dodecyl-SOa concentration is maintained in the substrate solution, 50% inactivation is obtained with 0.0025% Na dodecyl-S04. However, as indicated in the inset in Fig. 1, there is a further change with time; 0.0025% Na dodecyl-SO1 leads to 82% inactivation at 4 hours. Gottesfeld et al. (13)  of DNase A with or without Na dodecyl-SO4 were assayed in media with or without the detergent. The recorder tracings are shown in Fig. 2. In one of the controls (Curve 16), both the DNase and the DNA solutions were 0.7 mM in EDTA (the same molarity as 0.01% Na dodecyl-S0.J in order to be sure that any slight effect of Na dodecyl-SO4 on the bivalent metal concentration in the medium would not be critical. When Na dodecyl-SO4 was present only in the enzyme solution (Curue 6) the initial rate was decreased, but the activity increased with time, which indicated that the enzyme was recovering when the Na dodecyl-SO1 concentration was diluted 100-fold after only 1 to 2 min of exposure to the detergent. When the Na dodecyl-SOc was present initially only in the substrate solution (Curve S), the initial rate of catalysis was rapid, but dropped almost to zero within 2 to 3 min in the 0.01 y0 Na dodecyl-SOr. Curve 4 was a second control at one-tenth the DNase concentration used for Curue 1. When the Na dodecyl-SO4 was present in both solutions (Curue 6), there was no activity at all.
These results show that the effect of Na dodecyl-SO1 is on 0.14 0. DNase. The spectra were taken on an Aminco spectrophotometer, model DW-2, at 0.05 absorbance full scale, l-cm path length. Curve 1 represents a control with 0.25 mg/ml DNase in 0.1 M Tris-HCl, pH 7.0, in each cell. &roe 2 was taken after the solution in the sample position had been incubated with 0.01% Na dodecyl-SO4 for 1 hour at 25". A blank with 0.01% Na dodecyl-SOa did not show appreciable absorbance. I?lset, the absorption spectrum of DNase without Na dodecyl-Sol. The molar extinction coefficients were calculated b.y assuming a molecular weight of 30,000 (3). l ---0, the l-hour samples showed little or no increased regain of activity over 7 hours after the Na dodecyl-SO4 was diluted out in the initial CaClt-containing diluent.
the enzyme. A difference spectrum (Fig. 3) shows a Na dodecyl-Sod-induced perturbation similar to that observed by Zimmerman and Coleman (14) when DNase was heated at 79" in 2.5 mM HCl for 10 min. The time dependence of the inhibition by Na dodecyl-SO4 was then studied in more detail (Fig. 4). Samples were diluted prior to assay into Na dodecyl-Sod-free pH 7.0 buffer. 10 rnM in CaCl.. in order to ascertain the notential

Effects of Guanidinium
Chloride on DNase A-Full reversibility of the denaturation of the enzyme by Na dodecyl-SO1 has been attained through experiments with guanidinium chloride. When DNase is exposed to various concentrations of guanidinium chloride (up to 8 M) for as long as 1 hour and then is diluted into CaCLcontaining buffer, the activity remains unchanged (Fig. 5). If DNase that had been inhibited by Na dodecyl-SO4 was first diluted IO-fold into 8 M guanidinium chloride and within 1 or 2 min was diluted lo-fold into CaClp-containing pH 7.0 buffer and then loo-fold for assay, full activity was regained. The presence of Ca*+ is essential for this regeneration; if CaClz was omitted there was no regain of activity. When 6 M sodium chloride was substituted for 6 M guanidinium chloride, no activity was regenerated.
The dashed line in Fig. 5 shows the recovery of activity gained through dilution of Na dodecyl-Sod-treated DNase with intermediate concentrations of guanidinium chloride. About 6 M guanidinium chloride is needed to permit nearly complete reversal of the inhibition by Na dodecyl-SOa. for full recovery of activity when the detergent was diluted lO,OOO-fold. Ca2+ was included in the diluent because of its major effect in facilitating the refolding of reduced (4) or proteolyzed (7) DNase; CaClz could not be tested in the presence of 0.01% Na dodecyl-SO4 since the Ca2f ions precipitate the anionic detergent at that concentration.
Even when 7 hours were given for recovery, the regain was less than 50% after 1 hour of Na dodecyl-SOa exposure at 25" and was only about 12'% when the temperature of the initial incubation was 45".
When DNase treated with 0.025% Na dodecyl-SOa for 1 hour at 45" was gel-filtered on a Na dodecyl-S04-free Sephadex G-100 column (1.5 x 90 cm; flow rate, 25 ml/hour) at pH 8.0, nearly all of the protein emerged at the void volume, with only a very slight amount of active enzyme at the usual position of elution of DNase.  (17) have also shown that with tetradecyltrimethylammonium chloride a IO-fold higher concentration is required for the denaturation of serum albumin than is needed with Na dodecyl-S04. The results in Table I show that neutral detergents, such as Brij-35 and Triton X-100, do not have any effect at concentrations at which Na dodecyl-SO4 inactivates the enzyme. With the alkyl sulfates, the length of the hydrophobic side chain is important; there is little effect with the Cs and Cl0 analogs of Na dodecyl-S04. Quaternary ammonium bases are much less effective than the acidic detergents.

Action of Carboxypeptidase A on Na Dodecyl-SO&eated DNase
A-With the knowledge that fully active DNase can be recovered from Na dodecyl-Sob-treated DNase, it was possible to study the effect upon activity of the removal of COOH-terminal residues from the enzyme by carboxypeptidase A. Guidotti (18) has shown that carboxypeptidase A is active in the presence of 1.6% Na dodecyl-S04, and he used the enzyme in the presence of the detergent to remove COOH-terminal residues from the (Y and p chains of hemoglobin. With DNase in 0.03% Na dodecyl-SO,, the action of carboxypeptidase A was slow at 25" but appreciable at 45" (Fig. 6). The action of carboxypeptidase A can be expected to stop at Pro-251 in the COOH-terminal sequence -Pro-Val-Glu-Val-Thr-Leu-Thr (3). Also plotted in Fig. 6 is the percentage loss of activity accompanying this hydrolysis after dilution into guanidinium chloride to give full opportunity for the native Na dodecyl-Sod-treated protein to regain activity. The activity is very sensitive to the removal of the COOH-terminal residues; the loss of activity parallels very closely the removal of leucine, indicating that the removal of only 2 residues leads to almost complete inactivation of DNase.
This conclusion is valid, of course, only if there has been no proteolysis in other parts of the DNase chain. The product of carboxypeptidase A treatment was examined by gel filtration in 6 M guanidinium chloride (Sephadex G-100 column, 1.5 x 90 cm; flow rate, 3 ml/hour). The protein material had the same elution volume as DNase in the control and no separation of fragments was observed. In the control and the carboxypeptidase A experiments the recoveries of protein were about equal, based upon absorbance at 280 nm. In order to be sure that -S-S-bridges were not holding proteolyzed fragments together, gel electrophoresis (19) in the presence of Na dodecyl-SO4 and mercaptoethanol was performed; the control and the carboxypeptidasetreated DNase gave bands in identical positions. The activities (after solution in 6 M guanidinium chloride and dilution into CaCls-containing buffer) and the amino acid compositions of the gel-filtered proteins were compared. There was a significant decrease in activity of the control after the gel filtration (Table II) which may indicate that prolonged (24 hours) exposure to 6 M guanidinium chloride is not innocuous. For amino acid analysis of the proteins, the 6 M guanidinium chloride of the eluent used for the gel filtration was removed by dialysis against water. If a fine precipitate formed, it was removed by centrifugation and resuspended in water for the pipetting of aliquots of the suspension. In the control with DNase, the major proportion of the protein remained water-soluble; the carboxypeptidase Atreated protein, however, became largely water-insoluble.
The specific activities of the fractions indicated 987, inactivation of the insoluble carboxypeptidase A-treated protein (after solubilization for assay), which was isolated in 66 To yield.
The amino acid analyses summarized in Column 4 of Table III  indicate that the insoluble carboxypeptidase A-treated protein almost completely lacks the last 2 COOH-terminal residues (-Leu-Thr; e.g. Leu-0.8) and that about one-half of the molecules have also had the next 4 residues (-Val-Glu-Val-Thr; e.g. Glu-0.7, Val-1.2) removed. The amounts of free amino acids liberated within 60 min in this experiment are somewhat greater than those found for Fig. 6, since the substrate concentration was increased 3-fold.
Possible Regeneration of Activity by Add&on of a COOKTermi-na2 Peptide-Since RNase A from which 4 to 9 COOH-terminal residues have been removed can be as much as 90% reactivated by admixture with COOH-terminal peptides of 9 to 14 residues (20-23), we have undertaken a similar experiment with DNase. A 15.residue COOH-terminal peptide from DNase, CNBr-5 (3), has been isolated. When the precipitated, carboxypeptidase A-treated DNase (1.5 nmol) was dissolved in 200 ml of 7 M guanidinium chloride containing 40 eq of CNBr-5 (with approximately 10 eq of CNBr-1 as an impurity) and the mixture was diluted lo-fold into CaCl,-containing buffer prior to assay, there was only a 2% regeneration of activity. Increase in the ratio of peptide to protein or of the concentrations of the constituents did not increase the yield. The 2% result is reproducible, but the percentage of reactivation under our conditions has been extremely low. DISCUSSION Enzymes vary widely in the effect of the binding of Na dodecyl-SOa upon activity (24). DNase A is one of the most sensitive, since inactivation is complete of 0.005% Na dodecyl-SOa. The aim of reversing this inactivation was achieved only by introducing the transient exposure to 6 M guanidinium chloride; this step makes it possible to obtain, upon further dilution in the presence of Ca2+, the monomeric, active enzyme instead of inactive aggregates.
The results on the importance of residues at the COOH terminus of DNase add this enzyme to the list of proteins that possess this property; the group includes ribonuclease (25,26), aldolases (27,28), isocitrate lyase (29), and thymidylate synthetase (30). The COOH-terminal residues are the last to be added in the biosynthetic process and in many instances they have a determining effect upon the folding of the chain into the active conformation.
The chances for regeneration of major activity through replacement by adsorption of a COOH-terminal peptide would probably be increased if a longer terminal segment had been removed from DNase; with ribonuclease (20) the optimum regain is obtained when 6 amino acid residues have been removed from the COOH terminus.