Structure-function relationships at the active site of nuclease-T'.

Abstract Solid phase peptide synthesis has been used to study structural requirements at the active site of staphylococcal nuclease-T', the noncovalent complex of the polypeptides nuclease-T-(6-48) and nuclease-T-(49-149) (residues 6 through 48 and 49 through 149 of native nuclease, respectively). In light of the importance of glutamic acid 43 in nuclease-T' catalysis (Chaiken, I. M., and Sanchez, G. R. (1972) J. Biol. Chem. 247, 6743–747), analogues of nuclease-T-(6-48) with progressively longer deletions at the COOH terminus have been synthesized, and their binding to native nuclease-T-(49-149) to produce an active complex has been investigated. The results indicate that residues glutamic acid 43, threonine 44, lysine 45, histidine 46, proline 47, and lysine 48 are not necessary for formation of a nuclease-T'-like complex. On the other hand, threonine 44 is critical for normal enzymic activity. The effect of threonine 44 appears to be due to its contribution to the peptide bond with glutamic acid 43 and not to its specific side chain moiety. The synthesis of an analogue of nuclease-T-(6-48) with alanine at position 19 in place of aspartic acid was also undertaken. No enzymic activity is generated by this peptide in the presence of nuclease-T-(49-149), although complex formation appears to occur. This finding is consistent with the view, based on the crystal structure of the nuclease-Ca++-deoxythymidine 3',5'-diphosphate complex (Cotton, F. A., Bier, C. J., Day, V. W., Hazen, E. E., Jr., and Larsen, S. (1972) Cold Spring Harbor Symp. Quant. Biol. 36, 243–249), that the side chain carboxyl group of aspartic acid 19, along with those of aspartic acid residues 21 and 40 and glutamic acid 43, participates in the binding of the essential calcium ion.

native enzyme in its specificity toward substrates, in its requirement for calcium, and in over-all conformation (2)(3)(4)(5). Recent work (6) has indicated that the crystal structure of nuclease-T' with Ca++ and the strong inhibitor deoxythymidine 3') 5'diphosphate bound is isomorphous with that of the native nuclease-Ca++-pdTp' complex, the structure of which has been determined to approximately 2.5 A resolution (7). On this basis, the high resolution model of nuclease has been used in considerations of structure-function relationships in nuclease-T'. In native nuclease, according to the model, the NH%-terminal third of the polypeptide chain contributes a number of important residues to the interaction between the protein, Ca++, and pdTp. In particular, the side chain carboxyl groups of aspartyl residues 19, 21, and 40, and glutamyl residue 43 are located very close to the essential Ca++, while the guanidino group of arginine 35 appears to interact with the 5'.phosphate group of the inhibitor. Because of the close similarity between nuclease and nuclease-T', these interactions presumably play an important role in the latter complex as well. In fact, studies with synthetic analogues of nuclease-T-(6&48),2 the smaller polypeptide component of nuclease-T', have shown that aspartic acid 21 and 40, glutamic acid 43, and arginine 35 are critical for enzymic activity (9,10). Furthermore, t,he results have suggested participation of glutamic acid 43 in the formation of the active site (11).
In an attempt to investigate further the structural requirements at the active site of nuclease-T' and the role of glutamic acid 43 at this locus, we have undertaken the synthesis of analogues of nuclease-T-  with progressively larger deletions at the COOH terminus and studied the ability of these peptides to bind to nuclease-T-  and to generate nuclease-T' activity.
Amino acid blocking groups, coupling procedures, cleavage of the peptide from the resin with hydrogen fluoride, and removal of e-trifluoroacetyl groups of lysine residues with piperidine were as previously described (9,17). Removal of t-butyloxycarbonyl blocking groups was carried out with 59% trifluoroacetic acid in methylene chloride (18). The deblocked peptide mixture was passed through Sephadex G-25 (fine), and the peak fraction was rechromatographed on the same column.
Analogues were always synthesized in parallel with synthetic- , the peptide corresponding to the native sequence (omitting residue lysine 48, which has been shown not to be necessary for nuclease-T' activity (3,4)), in order to control the syntheses.
DNase and RNase activities were measured using the assay system of Cuatrecasas et al. (21). Synthetic peptide (0.01 Mmole)4 was incubated for 1 hour with nuclease-T-(49-149) (0.01 pmole) in 0.025 M Tris-Cl, pH 8.0, in a total volume of 100 ~1. Aliquots of the incubation mixture were assayed against DNA and RNA assay solutions prepared as previously described (21). When the effect of calcium ion (as CaClJ and pdTp was studied, these ligands were included in the incubation mixture at concentrations of 10 mM and 0.10 mM, respectively. For the determination of the fluorescence emission characteristics of semisynthetic peptide mixtures, synthetic peptide (0.05 pmole) was incubated with nuclease-T-(49P149) (0.0025 pmole) in 0.025 M Tris-Cl, pH 8.0, in the presence of 10 mM CaC12 and 0.10 mM pdTp in a total volume of 200 ~1. The 3 Histidine 46 has been shown to be nonessential for nuclease-T' activity (20). 4 The reported molar amounts of synthetic analogues in the experiments are based on the expected molecular weights for the peptides in question. samples were incubated for 1 hour and the fluorescence spectrum, upon excitation at 295 nm (22), was determined with an Aminco-Bowman spectrophotofluorometer (23). Appropriate blanks were run to correct for small contributions of the synthetic peptide, buffer, and ligands.
The resistance to trypsin of the activity of mixtures of synthetic peptides and nuclease-T-  was tested by addition of 5 ~1 of 0.2% trypsin (Worthington; diisopropylfluorophosphate-treated) to a solution containing synthetic peptide (0.025 pmole) and nuclease-T-(49-149) (0.0012 pmole) in 0.025 M Tris-Cl, pH 8.0, in the presence and in the absence of CaClz (10 mM) and pdTp (0.10 mM) in a total volume of 100 ~1. Aliquots were taken at time intervals and assayed for DNase activity as described above.
The resistance to trypsin of the fluorescence of mixtures of synthetic peptides and nuclease-T-  was tested by addition of 5 ~1 of 0.1% trypsin to a solution containing synthetic peptide (0.05 pmole) and nuclease-T-(49-149) (0.0025 pmole) in 0.025 M Tris-Cl, pH 8.0, in the presence and absence of CaClz (10 mM) and pdTp (0.10 mM) in a total volume of 200 ~1 (11). The fluorescence spectra of samples were determined as a function of time and corrected for the contribution of synthetic peptide, trypsin, ligands, and buffer.
Semisynthetic complexes of synthetic-(6-44)) -(6-43)) and -(6-42) with native nuclease-T-  were purified by a functional purification method developed previously (4,11). In this method, a mixture of synthetic peptide and native nuclease-T-(49-149) is treated with trypsin in the presence of Ca++ and pdTp and then fractionated by phos'phocellulose chromatography. Synthetic-  (150 mg) and native n&ease-T-(49-149) (15 mg) were dissolved in 15 ml of 0.4ye NH4HCOz containing 0.1 mM pdTp and 0.01 M CaCL at pH 8.0. After incubation at room temperature for 15 min, 2 mg of trypsin (diisopropylfluorophosphate-treated) were added. The solution was allowed to incubate for 150 min, during which time the DNase activity of the mixture fell to 84% of the initial value. At this time, 9 mg of soybean trypsin inhibitor were added. After 10 min, the solution was lyophilized.
Soybean trypsin inhibitor (3.3 mg and 5.2 mg, respectively) was then added.
After 15 min, the solutions were lyophilized.
The lyophilized mistures were dissolved in small volumes of 0.3 M ammonium acetate buffer, pH 5.8 (starting buffer), and applied to phosphocellulose columns equilibrated with the same buffer (column sizes: 22 x 1.0 cm for synthetic-(6-44) mixture, 7 x 1.0 cm for synthetic-  and (6-42) mixtures). Elution was carried out first with starting buffer (until essentially complete elution of the initial peak occurred) and then with a gradient consisting of 200 cc of starting buffer and 200 cc of 1.0 M ammonium acetate buffer, pH 8.0. In each case, an absorbance peak was observed eluting at a position very similar to that observed for native nuclease-T' (corresponding to buffer conductivity in the range of 26 to 31 mmhos).
The peak fractions were pooled and lyophilized.
A white powder was obtained in each case.
The complex of the native fragments nuclease-T-(6-48) and nuclease-T-  was prepared in essentially the same manner as the semisynthetic complexes, as described previously (4).
Stereo drawings were prepared with a Calcomp plotter,5 utilizing the high resolution coordinates for the nuclease-Ca++ pdTp model of Cotton et al. (7).

Characterization
of Crude Synthetic-(6-47) Analogue Peptides Activity of Synthetic Peptides-Each crude synthetic peptide was assayed for its ability to generate DNase and RNase activity after incubation with nuclease-T-(49-149) in a 1: 1 molar ratio.
In addition, synthetic peptides were assayed after incubation with nuclease-T-  in the presence of 10 mM CaClQ and 0.10 mM pdTp.
The results are shown in Table I. Synthetic- (6-47)) [Ala4$ynthetic- , and synthetic-  produce significant levels of DNase and RNase activity. Synthetic-  and -(6-42) produce essentially no activity at this level of purity.
[Asp43]synthetic-  also fails to produce enzymic activity, as might be expected from previous observations (10,11). The presence of the active site ligand, Ca++, and of the inhibitor, pdTp, in the incubation mixture enhances the DNase activity effected by nuclease-T-  and by synthetic-(6-47), [Ala4Q]synthetic- , and synthetic-(B-44). This effect has been interpreted as being due to the stabilization of the active complex formed (18). The slight inhibition of RKase activity observed is believed to reflect the ability of pdTp to compete effectively with RSh under the conditions used (18).
It appears that deletion of the terminal threonine 44 residue causes loss of the ability to effect normal activity in the incubation mixture.
In order to determine whether this loss is a result of the loss of a function performed by the side chain of threonine 44 at the active site, the analogues synthetic-  and [Ala44]synthetic-  were prepared and assayed for DNase and RNase activity.
The results are shown in Table I. The data indicate that the side chain of threonine 44 is not essential for enzymic activity. Therefore, the difference in the enzymic activities effected by synthetic-  and synthetic-(6-43) must be due, mainly, to factors other than the absence of the side chain of this amino acid.
The data in Table I also show that the analogue [AlaiQ]syn thetic-  fails to produce enzymic activity. For this peptide, as well as for synthetic-(6-43), [hsp43]synthetic-(6-43), and synthetic- , the absence of significant activity is evident even at 20 : 1 molar ratios of synthetic peptide to nuclease-T-  in the presence of Ca++ and pdTp. Fluorescence of Semisynthetic Mixtures-Specific binding of synthetic-  analogue peptides to nuclease-T-(49-149) has been measured by determining the fluorescence spectral characteristics of mixtures of synthetic peptide with nuclease-T-(49-149) in the presence of Ca++ and pdTp.
Native nuclease-T-(6-48) (which contains no tryptophan) produces a blue shift and an enhancement of the fluorescence of the single-tryptophanyl residue in nuclease-T-(49-149) (at position 140) when added to this peptide.
This effect has been interpreted as being due to the insertion of the tryptophanyl residue into a nonpolar pocket (24) in the active conformation of nuclease and nuclease-T'. The synthetic peptides, therefore, were assayed for their ability to induce a specific conformation in nuclease-T-(49-149), as indicated by the fluorescence effect produced.
The results are shown in Fig. 2. All of the synthetic peptides with truncations at the COOH terminus shift the emission maximum of the tryptophan residue and produce a large fluorescence enhancement. Even those truncated peptides which are not capable of producing activity show the characteristic fluorescence behavior.
The results are taken to indicate that all of the truncated analogues bind nuclease-T-  to form semisynthetic complexes. It is noteworthy that the truncated sequence, synthetic-(6-42)) still contains the information required for the stabilization of the specific three-dimensional structure of nuclease-T-(49149) in nuclease-T'.
The [Alaig]synthetic-  analogue shows a smaller fluorescence effect than does synthetic- , although the effect is still considerable.
[Asp43]synthetic- , not shown in the figure for the sake of clarity, gives a fluorescence enhancement of the same magnitude as that of synthetic- . Clearly, this enhancement is at least as great as that of synthetic-  and -   I  I  I  I  I  I   90   IO   I  I  I  I  I  I  300  320  340  360 (2). Semisynthetic nuclease-T' also exhibits this property (4). In order to characterize further the mixtures of synthetic peptides with nuclease-T-(49-149), the resistance to trypsin of these semisynthetic mixtures was determined.
Two types of experiments were carried out. For the synthetic peptides that generated enzymic activity with nuclease-T-(49-149), measurements were made of the enzymic activity remaining at different times after addition of trypsin in the presence and absence of ligands.
For the truncated synthetic peptides which generated no enzymic activity with nuclease-T-(49-149) but which still bound this fragment (as determined from the fluorescence effect), measurements were made of the fluorescence intensity of the mixture at different times after addition of trypsin, in the presence and absence of ligands (11). The results of the first type of experiment are shown in Table II, and those of the second type in Fig. 3.
In the absence of ligands, on the other hand, enzymic activity disappeared within 1 hour. The resistance of the enzymic activity of the semisynthetic mixtures to tryptic attack in the presence of ligands parallels that observed for the native mixture under similar conditions  -(49-149) in the presence of CaC12 and pdTp, in Tris-Cl buffer, pH 8.0. Trypsin was added, and the DNase activity of the mixture was followed as a function of time. See "Experimental Procedures" for other details.
(In the absence of ligands, essentially all activity was lost within 1 hour after addition of trypsin.) in the presence (0) and absence (0) of Ca++ and pdTp. Experimental conditions were as described under "Experimental Procedures." (2,20). The specific ligand effect suggests that, under the experimental conditions, these synthetic analogues combine uith nuclease-T-  to form semisynthetic complexes very similar to nuclease-T' in over-all conformation and ligand binding.
The data in Fig. 3 show the resistance to trypsin of the fluorescence of mixtures of nuclease-T-  with native nuclease-T- , synthetic-(6-44), synthetic- , and synthetic-  in the presence and absence of ligands. For the two active semisynthetic mixtures (Fig. 3, A and B), an initial decrease in fluorescence occurs which is not paralleled by a decrease in activity (Table II).
This initial decrease is probably due mainly to the proteolysis of excess nuclease-T-  or synthetic- .
This has been observed as the drastic decrease in the scattering component in the fluorescence emission spectrum, which has been attributed to these peptides and which contributes significantly to the measured fluorescence emission at 330 nm. Once this initial decrease is complete, the remaining fluorescence due directly to nuclease-T-like complexes is stable, as expected, for both active semisynthetic mixtures.
The initial rapid fluorescence decrease is apparent also for the inactive mixtures for synthetic-  and -(6-42) (Fig. 3, C and D, respectively).
However, here, the fluorescence emission surviving the initial period of digestion continues to decrease steadily, apparently due to a decreased resistance of the nuclease-T-like complexes produced in these cases. Thus, while the stabilization of the fluorescence effect by ligands in Fig. 3, C and D indicates that the inactive semisynthetic complexes formed do bind ligands, the lowered absolute resistance to trypsin in the presence of ligands suggests that the mode of binding for these complexes has perhaps been altered relative to binding for the active complexes. This behavior has been found previously for [ASHES]synthet,ic-  (II).

Properties of Xemisynthetic
Complexes-It has been shown previously that the resistance to trypsin of the enzymic activity and of the fluorescence emission properties of mixtures of crude synthetic- .related peptides and native nuclease-T-(49-149) in the presence of Ca++ and pdTp provides the basis for the isolation of semisynthetic analogues of the nuclease-T' complex (4,11). Inasmuch as the peptides involved in the COOHterminal activity transition, namely synthetic-(G-44), -(6-43), and -(6-42), effect trypsin-resistant nuclease-T/-like properties when mixed with native nuclease-T-(49-149), such semisynthetic complexes were prepared for these peptides using the functional method (see "Experimental Procedures"). The resultant compleses were all isolated on phosphocellulose and identified by enzymic activity when appropriate and by amino acid compositiou. The yields were 4.0, 0.7, and 1.0 mg, respectively, for the  All of the semisynthetic complexes were assayed for DNase and RNase activity, along with native nuclease-T', with the results shown in Table III.
[Des 45-471.semisynthetic nuclease-T', containing synthetic-  in place of native nuclease-T- , is approximately 82% as active catalytically as native nuclease-T'. The  complexes show approximately 3 to 5yo and 0.4 to OX%, respectively, of the enzymic activity of nuclease-T'. Normal semisynthetic nuclease-T' (containing synthetic-(6-47) of normal sequence), 3657 prepared by the same method as in previous work, showed about 90% the enzymic activity of nuclease-T' (4). The sharp transition in the enzymic activity upon loss of threonine 44 observed with crude peptides (Table I) is again  quite apparent with the purified semisynthetic complexes. This observation is in agreement with the preliminary conclusion (based on activity measurements before purification) that threonine 44 performs an important role in this portion cf the active site of nuclease-T'.
On the other hand, both  nuclease-T' do exhibit low levels of enzymic activity, in contrast to the essentially total lack of activity shown by the crude synthetic-  and -  peptides in the presence of nuclease-T-(49-149) ( Table I). Such low activities do not appear to be due to dissociation of otherwise highly active complexes, since fluorescence emission properties of these two analogues appear to be of the same character as those of nuclease-T' and of [Des 4547lsemisynthetic nuclease-T', namely, a blue shift in the emission maximum and a large intensification of the emission spectrum of the single tryptophanyl residue.
Based on our previous experience (ll), it is probable that the very low level of enzymic activity associated with the [Des 43-47lsemisynthetic nuclease-T complex is due to the presence of a trace amount of native nuclease-T', produced during the trypsin purification step as a result of the presence of either native nuclease or nuclease-T-  in the nuclease-T-(49-149) used. The activity associated with [Des 4447]semisynthetic nuclease-T' is subject to similar considerations, although, again based on our previous experience, the percentage relative to nuclease-T' appears to be too high to be fully accounted for in the same manner.
Rather, we feel that [Des 44-47lsemisynthetic nuclease-T' does in fact possess intrinsic enzymic activity, the low level relative to nuclease-T' being due to the loss of the peptide bond between positions 43 and 44. The lack of significant activity effected by mixtures of crude synthetic-  with nuclease-T-(49-149) ( Table I) probably reflects competition of nonfunctional peptide impurities for available nuclease-T-(49-149).
The effect of elevated temperature on the enzymic activity of the purified compleses was determined as previously described (4). As already found for native nuclease-T' and normal semisynthetic nuclease-T' (4), the DNase activity associated with [Des 4547lsemisynthetic nuclease-T' varies biphasically, with a drastic falloff occurring above 40", presumably due to dissociation of the complex.
A similar behavior was found for the [Des 44-471 and [Des 43-471 semisynthetic complexes.
For the latter two cases, the data indicate that the low levels of enzymic activity are not due to contamination of the isolated complexes by native nuclease.

DISCUSSION
The arrangement of the side chain carboxyl groups of residues aspartic acid 19, 21, and 40, and glutamic acid 43 around the essential Ca+f at the active site of the nuclease-Ca++-pdTp complex is shown in a stereo drawing in Fig. 4. The location of the bound inhibitor, pdTp, is also shown. Because studies in solution (2)(3)(4)(5) and in the crystal state (6,7) have shown that the two enzyme species are closely related in over-all structure, the relative position of residues at the active site of nuclease-T' is believed to be very similar to that shown in the figure.
Any major differences would be expected to exist near the points of trypsin cleavage. By analogy with the ribonuclease-ribonuclease-S system (25, 26), the residues in the COOH-terminal region of nuclease-T-  and in the NHz-terminal region of nuclease-T-  would be expected to be arranged in a more disordered manner in nuclease-T' than in nuclease, where the peptide bond between residues 48 and 49 restricts the orientations which the polypeptide chain can assume in this region. The finding that residues lysine 45, histidine 46, proline 47, and lysine 48 can be left out of the COOH terminus of nuclease-T-(6-48) without significant loss of activity is in agreement with the idea of a random, essentially unimportant conformation in this portion of the nuclease-T' molecule.
The fluorescence experiments (Figs. 2 and 3) indicate that, in addition to the above residues, threonine 44 and glutamic acid 43 can also be left out of nuclease-T-  without a significant change in the over-all conformation of the complex formed.6 The fact that enzymic activity is almost completely lost when residues threonine 44 through lysine 48 are missing, while a nuclease-T'-like complex is still formed, suggests that loss of threonine 44 causes a critical perturbation of glutamic acid 43. The results with the [Ala44]synthetic-  analogue indicate that lack of activity is not due to the absence of the side chain of threonine 44. Rather, threonine 44 appears to maintain the critical properties of the T-COOH group.
The somewhat decreased stability to trypsin of the fluorescence of mixtures of synthetic-  and native nuclease-T-  in the presence of ligands very likely reflects altered ligand binding upon loss of threonine 44. In fact, the trypsin stability of the fluorescence effect for synthetic-  is very similar to that for synthetic-(6-42), which lacks glutamic acid 43 and therefore would not interact with Ca+f through the side chain of this residue. These findings strongly support the view (9-11) that glutamic acid 43 is performing an important role during enzymic catalysis by nuclease-T', either directly as a catalytic group or indirectly through the essential calcium ion.
One structural element common to all of the truncated analogues which bind native nuclease-T-  efficiently is the sequence of residues 13 through 35. This segment in nuclease 6 That a complex very similar in structure to nuclease is formed between synthetic-  and nuclease-T-(49-149) was indicated also in antibody inhibition experiments performed in our laboratory by Dr. David H. Sachs. Antibody against conformational features in the region of residues 99 through 126 of native nuclease is inhibited by an equimolar mixture of synthetk(6-43) and nuclease-T-(49-149) but not significantly by either peptide alone (27).
(and presumably in nuclease-T', as well) forms a triple-stranded anti-paralleled P-pleated sheet array. It is likely that this P-structure contributes, in a significant manner, towards the stabilization of the nuclease-T' complex.
We have found7 that synthetic-  brings about a fluorescence effect with nuclease-T-  in the presence of ligands, whereas synthetic-(9-35) and synthetic-  do not. Since synthetic-  also gives rise to a fluorescence effect, it is possible that a particular residue in the sequence 38 through 41 is important for complex formation with nuclease-T-(49-149).
The functioning of aspartic acid 40 as an "anchoring" residue is an interesting possibility. The finding that replacement of aspartic acid 19 with alanine leads to loss of activity parallels the results of substitutions (with and without preservation of charge) of aspartic acid residues 21 and 40 and glutamic acid 43 in synthetic-(6-47) (9,10). Based on the high resolution crystal structure of the nuclease-Ca++-pdTp complex (7), the side chain carboxyls of these 4 residues (Fig. 4) are believed to be sufficiently close to the calcium ion to be able to interact with it, although the exact nature of this interaction is not known (14). The results with the [Alal analogue of synthetic- , together with the earlier observations (g-11), indicate the need for the presence of these four carboxyl groups around the essential Ca++ during the catalytic process.
The resu1t.s presented here, and those previously reported (g-11)) have provided a qualitative indication of the importance of aspartic acid 19, 21, and 40 and glutamic acid 43 in the active site of nuclease-T'.
Purification of some of these synthetic analogues, as reported previously (11) and in the present work, suggests that when large differences in properties exist between analogue and normal synthetic peptides at the crude level of purity such differences remain after purification.
As described earlier (II)., the availability of these purified semisynthetic analogues of nuclease-T', with modifications at key loci in the active site region in the molecule, should allow a quantitative description of the participation of certain residues in ligand binding and enzymic activity. for the crystal