Additivity of Protein-Guanine Interactions in Ribonuclease T1*

(Received for publication, December 31, 1996, and in revised form, January 27, 1997)

Stefan Loverix , Jan Doumen and Jan Steyaert Dagger

From the Dienst Ultrastruktuur, Vlaams Interuniversitair Instituut Biotechnologie, Vrije Universiteit Brussel, Paardenstraat 65, B-1640 Sint-Genesius-Rode, Belgium

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS AND DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

It has been established that Tyr-42, Tyr-45, and Glu-46 take part in a structural motif that renders guanine specificity to ribonuclease T1. We report on the impact of Tyr-42, Tyr-45, and Glu-46 substitutions on the guanine specificity of RNase T1. The Y42A and E46A mutations profoundly affect substrate binding. No such effect is observed for Y45A RNase T1. From the kinetics of the Y42A/Y45A and Y42A/E46A double mutants, we conclude that these pairs of residues contribute to guanine specificity in a mutually independent way. From our results, it appears that the energetic contribution of aromatic face-to-face stacking interactions may be significant if polycyclic molecules, such as guanine, are involved.


INTRODUCTION

The energetics of virtually all binding properties in proteins is the culmination of a complex set of intermolecular interactions. The individual contributions and the mutual interdependence of these interactions are currently being probed through protein engineering by many research groups. The most powerful experimental approach is to analyze the effects of single mutations on binding, turnover, or conformational stability and to compare these with the properties of proteins where multiple mutations are combined in one molecule (1-4).

The functional role and cooperative interplay between catalytic residues have been investigated in detail for a number of enzymes including subtilisin (5, 6), staphylococcal nuclease (7, 8), and ribonuclease T1 (9). In each case, the free energy barriers to substrate turnover introduced by mutations of catalytic residues are not additive in the corresponding double (or multiple) mutants. Residues involved in the process of breaking and forming covalent bonds appear to contribute to turnover in a mutually dependent way. Far less information is available on the additivity of molecular interactions involved in substrate binding. In this study, we investigate the mutual dependence of interactions at the guanine-binding site of ribonuclease T1 by protein engineering.

Ribonuclease T1 (RNase T1; EC 3.1.27.3) of the slime mold Aspergillus oryzae (10) is the best known representative of a large family of homologous microbial ribonucleases with members in the prokaryotic and eukaryotic worlds (11, 12). RNase T1 has a pronounced specificity for the base guanine; kinetic studies on the trans-esterification of dinucleoside phosphates revealed that the specificity constant (kcat/Km) for GpN1 substrates is ~106-fold greater than for corresponding ApN substrates and at least 108-fold greater than for CpN and UpN substrates (13). The three-dimensional structure of RNase T1 complexed with the competitive inhibitor 2'-GMP (14, 15) provides a structural basis for understanding the enzyme's specificity. In the complex, the hydrogen-bonding potential of the guanine base is completely saturated by complementary donor/acceptor sites on the enzyme involving the backbone atoms of Asn-43, Asn-44, and Asn-98 and the side chain carboxyl group of Glu-46 (Fig. 1). The N(1)-H-Glu-46 O-epsilon 1, N(2)-H-Glu-46 O-epsilon 2, and the N(2)-H-Asn-98 O hydrogen bonds have apparent contributions of 2.7, 1.1, and 1.2 kcal/mol to the specificity of RNase T1 for guanosine, respectively (16). Moreover, the guanine base is stacked between the phenolic side chains of tyrosines 42 and 45 (14, 15, 17-20).


Fig. 1. Pattern of the intermolecular interactions found in the crystalline structure of the wild-type RNase T1·2'-GMP complex. The guanine base; the backbone atoms of Tyr-42, Asn-43, Asn-44, Tyr-45, Glu-46, and Asn-98; and the side chains of Tyr-42, Tyr-45, and Glu-46 are shown. The oxygen atoms are shown in black, the nitrogen atoms in light gray, and the carbon atoms in medium gray. The residue number is shown near the corresponding C-alpha atom. The base numbering follows IUPAC conventions. Aliphatic protons are omitted for clarity. Dashed lines represent hydrogen bonds. Coordinates were taken from Ref. 15.
[View Larger Version of this Image (22K GIF file)]


In this study, Tyr-42, Tyr-45, and Glu-46 (the only residues that interact with the guanine base through their side chains) have been replaced by alanine. The effects of the single mutations are discussed below. To measure the functional cooperativity between these residues, single and multiple mutations have been analyzed by double mutant cycles. The structural implications of the Y42A mutation have been investigated by x-ray diffraction analysis.


EXPERIMENTAL PROCEDURES

Chemicals

Oligonucleotides were bought from Eurogentec. The 3',5'-dinucleoside phosphate substrate GpU was from Sigma. Common reagents were purchased at the highest purity available.

Overproduction and Oligonucleotide-directed Mutagenesis

The overproduction of wild-type RNase T1 and the E46A mutant as secretory proteins in Escherichia coli has previously been described (16, 21). The Y42A and Y45A mutants, the corresponding double mutant, as well as the Y42A/E46A double mutant were constructed via a polymerase chain reaction-based site-directed mutagenesis technique (22). Mutations were identified by DNA sequence determination (23); the entire sequence of each mutant gene was determined to check that no additional unwanted mutations had arisen during the polymerase chain reaction steps. The wild-type and mutant enzymes used in this study are of the isoform containing a lysine at position 25; they were purified to homogeneity as described (24).

Kinetics of Dinucleotide Phosphate trans-Esterification

All experiments were performed at 35 °C in 50 mM imidazole, 50 mM NaCl, and 2.5 mM EDTA at pH 6.0 (ionic strength = 0.1 M). The protein concentrations were determined spectrophotometrically at 278 nm, where A0.1% = 1.54 (25). For the mutants lacking one or two tyrosine side chains, this value was recalculated according to Pace et al. (26). The obtained A0.1% values were 1.42 for mutants with one truncated tyrosine and 1.30 for the mutant with two tyrosine side chains removed. The kinetic parameters for the trans-esterification of GpU were determined from initial velocities by measuring the absorbance increase at 280 nm (27). GpU concentrations varied between 10 µM and 1 mM. Reactions were started by adding enzyme to final concentrations ranging from 3 × 10-10 to 5 × 10-6 M depending on the enzyme used, except for the Y42A/E46A mutant, for which a concentration of 397 µM had to be used. For this mutant, the second-order rate constant (kcat/Km) for the trans-esterification of GpU was derived from a progress curve run overnight at a substrate concentration of 408 µM (much lower than the Km). Indeed, the Km for the single E46A mutant exceeds 5 mM (16). In experiments requiring high substrate concentrations, 0.5-, 0.2-, or 0.1-cm path length cuvettes were used to diminish the background absorbance. Experimental data were analyzed with the program GraFit (28).

Crystallization, Data Collection, and Structure Elucidation of the Y42A Mutant

Single crystals were grown by vapor diffusion in sitting drops at 20 mg/ml protein in 50 mM sodium acetate buffer, pH 4.2, containing 0.125% (w/v) 2'-GMP and 1.25% (w/v) CaCl2 using 50% (v/v) 2-methyl-2,4-pentanediol as a precipitant. X-ray data up to a resolution of 2.3 Å were measured on a MAR image plate and processed using the CCP4 (Collaborative Computational Project 4) suite of programs (29). The crystal was found to belong to space group P212121, with cell dimensions a = 49.61, b = 48.49, and c = 40.61 Å (see Table I).

Table I.

Crystallization, data collection, data processing, and refinement parameters of the Y42A RNase T1 · 2'-GMP complex


Crystallization
  Crystallization conditions 20 mg/ml enzyme, 50 mM NaOAc, pH 4.2, 0.125% (w/v) 2'-GMP, 1.25% CaCl2, 50% (v/v) MPD,a
  Space group P212121
  Cell parameters a = 49.61, b = 48.49, c = 40.61 Å 
Data collection/processing
  Resolution 2.3 Å 
  Rsym (last resolution shell, 2.38 to 2.30 Å) 0.107 (0.239)
  Completeness (last resolution shell, 2.38 to 2.30 Å) 93.0% (49.7%)
  No. of unique reflections 4313
  Initial model pdb1bir.ent (30)
Final structure
  No. of water molecules 82
  R-factor 0.216
  Rfree 0.257

a MPD, 2-methyl-2,4-pentanediol.

Because the cell constants of the F100A RNase T1 crystals (30) are very similar, we used the coordinates of this structure (Protein Data Bank code pdb1bir.ent) as our initial model for molecular replacement. After removing all non-protein atoms and the side chain of Tyr-42, rigid body refinement was carried out using X-PLOR (31), resulting in an R-factor of 0.316. The structure has been refined to a final R-factor of 0.216 and an Rfree factor of 0.257 by successive sessions of stereochemically restrained least-square refinements using X-PLOR and manual model revisions using the program O (32) on a Silicon Graphics Inc. graphics station. The phenylalanine at position 100 and 2'-GMP were fitted manually into the corresponding difference densities after the first couple of refinement cycles. 82 water molecules were inserted in spherical difference densities where suitable hydrogen-bonding partners were available.


RESULTS AND DISCUSSION

Experimental Strategy

The interaction of an individual side chain with the substrate may be analyzed by comparing the kinetics for the wild-type enzyme with those for a mutant in which the side chain has been truncated (2, 3). The obtained apparent binding energy is always a measure of the specificity of binding or catalysis (33). If the mutation does not induce structural changes and allows the access of bulk water to the cavity in the enzyme-substrate complex, the apparent binding energy may be a crude measure of the incremental binding energy of the interaction, i.e. the net free energy contribution of the interaction to transfer the substrate from bulk water to the enzyme complex. Free energy changes calculated from the specificity constants (Delta G = -RT ln((kcat/Km)/(kcat/Km)groupright-arrow 0)) equal the apparent contribution to substrate specificity of the group under investigation. In this study, Tyr-42, Tyr-45, and Glu-46 have been replaced by alanines. The effects of these mutations on the steady-state kinetic parameters are discussed below.

This paper also addresses the mutual interplay of amino acid residues involved in substrate binding. For this purpose, we constructed the Y42A/Y45A and Y42A/E46A double mutants and measured their kinetics. The free energy barriers to substrate specificity introduced by the single and multiple mutations have been analyzed by double mutant cycles (1, 4, 34). The degree to which one mutation affects the contribution of a second mutation (quantified by the coupling term Delta Delta G) measures the mutual component of the interaction energy between both residues and the substrate, provided that none of the mutations gives rise to a disrupted spatial arrangement of other amino acid residues (1). Simple additivity (Delta Delta G = 0) is observed when the two residues contribute to binding/turnover in a functionally independent way (4).

Non-disruptive Nature of the Introduced Deletions

To discuss the apparent effects of mutations on the steady-state kinetics in terms of binding energy and specificity (33), it is a requisite to analyze the structural implications of these mutations (see above). In the ideal case, a mutation causes the removal of a simple interaction with no perturbation of the structure. Below, we argue that the mutations we introduced in RNase T1 are non-disruptive deletions (according to the classification of Fersht et al. (35)). Therefore, the apparent energetic contributions calculated from the effects of these mutations are genuine estimates of changes in incremental interaction energies.

The phenolic side chain of Tyr-42 forms the basis of the guanine-binding site (Fig. 1). The aromatic ring lies on top of a hydrophobic core involving the side chains of Phe-50, Val-79, Ile-90, and Phe-100. Because of its buried location in the enzyme-substrate complex and the dramatic effect of the Y42A mutation on the kinetics (see below), we were concerned that this mutation induces major changes in the overall structure of the protein. For this reason, we examined the structural implications of the Y42A mutation by x-ray crystallography. The complex of Y42A RNase T1 with the specific inhibitor 2'-GMP has been refined to an R-factor of 0.216 using x-ray diffraction data to 2.3 Å (Table I). The Y42A mutation does not disrupt the overall structure of the enzyme-inhibitor complex (overall root mean square deviation = 0.349 Å). The local structural effects caused by the Y42A mutation are summarized in Fig. 2. The apparent structural perturbations at lysine 25 and glutamate 28, two solvent-exposed residues, are due to alternative side chain conformations. Table II compares the intermolecular hydrogen bonds for wild-type and Y42A RNase T1 in complex with 2'-GMP. A water molecule (Wat-134) occupies part of the free space available after removal of the tyrosine side chain (Fig. 3). The water molecule takes up the same position as the Tyr-42 O-eta does in the wild-type enzyme and forms a new hydrogen bond of 2.92 Å (Table II) with O-6 of the base (Fig. 4). The sugar pucker in the Y42A RNase T1·2'-GMP complex is C-2' endo (Table III), as in the case of the wild-type complex (15).


Fig. 2. Root mean square deviations between Y42A RNase T1 and wild-type RNase T1 complexed with 2'-GMP. For each residue, the root mean square (RMS) deviation for the main chain atoms (above) and side chain atoms (below) are plotted. Coordinates of the wild-type enzyme were taken from Ref. 15.
[View Larger Version of this Image (41K GIF file)]


Table II.

Comparison of the intermolecular hydrogen bonds between wild-type and Y42A RNase T1 complexed with 2'-GMP


Y42A
Wild-typea
Bond Length Bond Length

Å Å
Sugar
  O-5 Wat-158 2.83
  O-2 Wat-148 3.50
  O-3 Wat-121 3.23 Wat-148 3.49
Phosphate
  O-1P His-40 N-epsilon 2 3.05 His-40 N-epsilon 2 2.84
  O-1P Tyr-38 OH 2.68 Tyr-38 OH 2.70
  O-2P Arg-77 N-epsilon 2 2.97
  O-2P His-92 N-epsilon 2 3.22
  O-2P Wat-158 2.63
  O-2P Wat-121 3.23 Wat-179 3.02
  O-2P Wat-132 2.74 Wat-186 2.73
Base
  O-6 Asn-44 N 2.93 Asn-44 N 2.78
  O-6 Tyr-45 N 3.35 Tyr-45 N 2.82
  O-6 Wat-134 2.92
  N-1 Glu-46 O-epsilon 1 2.91 Glu-46 O-epsilon 1 2.73
  N-2 Glu-46 O-epsilon 2 2.90 Glu-46 O-epsilon 2 2.96
  N-2 Asn-98 O 2.85 Asn-98 O 2.86
  N-7 Asn-43 O-delta 1 3.48

a Protein Data Bank code pdb1rnt.ent (15).


Fig. 3. Ball and stick representation of Ala-42 and Tyr-45 and of the nucleotide in the Y42A mutant. The thin lines represent the mutated tyrosine 42 of the superimposed wild-type enzyme (15). The spherical electron density is from a water molecule occupying the space available after removal of the tyrosine side chain.
[View Larger Version of this Image (20K GIF file)]



Fig. 4. Hydrogen bonding pattern around the guanine base as found in the Y42A mutant. The backbone atoms of Ala-42, Asn-44, Glu-46, Tyr-45, and Asn-98 and the side chain of Tyr-45 are shown. The oxygen atoms are shown in black, the nitrogen atoms in light gray and the carbon atoms in medium gray. The residue number is shown near the C-alpha atom. The base numbering follows IUPAC conventions. Aliphatic protons are omitted for clarity. Dashed lines represent hydrogen bonds.
[View Larger Version of this Image (18K GIF file)]


Table III.

Nucleotide conformation in the Y42A RNase T1 · 2'-GMP complex


Sugar puckera

 tau 0 6.64°
 tau 1  -16.64°
 tau 2 19.83°
 tau 3  -16.52°
 tau 4 6.64°
P 166.35°  (C-2' endo)
Torsion angles

C-5'-C-5'-C-4'-C-3' (gamma )  -144.1°
C-5'-C-4'-C-3'-O-3' (delta ) 100.82°
C-5'-C-4'-C-3'-O-2'  -129.98°
C-4'-C-3'-C-2'-O-2'  -92.09°
C-3'-C-2'-O-2'-P  -94.23°
O-4'-C-1'-N-9'-O-4 (chi ) 77.51°  (syn)

a Pseudorotation parameters calculated according to Ref. 50.

Several crystallographic (15, 18, 19) and NMR (36) studies on RNase T1 indicate that the Tyr-45 side chain is flexible and solvent-exposed. Removal of this side chain is therefore not likely to disturb the enzyme's three-dimensional structure. Glu-46, another solvent-exposed residue in the free enzyme, swings into hydrogen bond the guanine base (see the Introduction) upon substrate binding (19). The invagination resulting from the removal of the Glu-46 side chain is probably sufficiently large to allow access of bulk water (16). All structural data corroborate the view that the Y42A, Y45A, and E46A mutations correspond to non-disruptive deletions.

Tyr-42 and Glu-46 Contribute to Substrate Binding Rather than to Turnover

Table IV lists the steady-state kinetic parameters for the trans-esterification of GpU obtained for the wild-type enzyme and for the Y42A, Y45A, and E46A mutants. The Y45A mutation has minor effects on the steady-state kinetic parameters of the enzyme, indicating that this residue does not contribute significantly to the incremental binding energy of the substrate. Consistent with this notion, the Y45F and Y45W mutations have been found to induce marginal effects on the trans-esterification kinetics of pGpC (37-39).

Table IV.

Steady-state kinetic parameters of Y42A, Y45A, Y42A/Y45A E6A, Y42A/E46A, and wild-type RNase T1 for the substrate GpU

Measurements were performed in imidazole buffer (0.1 M ionic strength), pII 6.0, at 35 °C.


Km kcat kcat/Km

s-1 mM-1 s-1
Wild-type 33  ± 3 µM 29  ± 1 879  ± 85
Y45A 44  ± 11 µM 28.6  ± 1.8 647  ± 279
Y42A >1 mM 0.127  ± 0.007
Y42A/Y45A >1 mM 0.079  ± 0.002
E46Aa >1 mM 1.66  ± 0.08
Y42A/E46A >1 mM 1.260 E-4  ± 4.6 E-6

a Data taken from Ref. 16 and corrected for A0.1chi  = 1.54 instead of 1.9.

In contrast to the Y45A mutation, the Y42A and E46A deletions were found to affect the steady-state kinetic parameters of the enzyme significantly. The trans-esterification rates of Y42A and E46A are linearly proportional to the substrate concentration over the entire concentration range that is experimentally accessible. From these linear relationships, we calculated the second-order rate constant for the trans-esterification reaction (kcat/Km). We defined a minimum value of Km equal to 1 mM (16). From our data, it appears that substrate binding, rather than turnover, is impaired upon deletion of the Tyr-42 or Glu-46 side chain.

The six-membered aromatic rings of Tyr-42 and the guanine base are involved in a parallel face-to-face stacking interaction (dihedral angle = 12°; centroid separation = 5.0 Å). The Y42A mutation has a devastating effect (Delta G = -5.4 kcal/mol) on the enzyme's incremental binding energy toward guanosine. This effect is much larger than normally observed for intramolecular aromatic-aromatic interactions between side chains of phenylalanine, tyrosine, and tryptophan, mostly of the perpendicular edge-to-face type (40-42). Indeed, simple aromatic rings prefer to associate via enthalpically favorable edge-to-face interactions of ~1.5 kcal/mol. This is in contrast to parallel stacking, which contributes almost zero when the rings are monocyclic. The large contribution of Tyr-42 to substrate binding suggests that face-to-face stacking can be significant when polycyclic molecules such as guanine are involved. The contribution of Glu-46 to substrate binding and guanine specificity (see the Introduction) has been discussed in detail previously (16, 43).

Substrate Binding Is Not Cooperative

Studies on model compounds indicate that hydrogen bonding and stacking interactions may mutually reinforce each other (44). The large effects of the E46A and Y42A mutants could theoretically be explained by such a type of cooperativity. Hydrogen bonds with the periphery of the base (Glu-46) might enhance a stacking interaction with this base (Tyr-42) and vice versa. To investigate this hypothesis, we performed a double mutant cycle analysis involving Tyr-42 and Glu-46 (Fig. 5). The effect of the Y42A mutation was measured in the presence and absence of the Glu-46 side chain. The coupling energy Delta Delta G, measured by comparing the free energy differences corresponding to parallel transitions of this cycle, is 0.394 kcal/mol. This is a small value (<10%) compared with the effects of the individual mutations, indicating that the contributions of Tyr-42 and Glu-46 to guanine binding are mutually independent.


Fig. 5. Thermodynamic cycle analyzing the mutual dependence between Tyr-42 and Glu-46. Delta G values (kcal/mol) equal the group's incremental binding energies and were calculated from the effect on kcat/Km. Delta Delta G equals the mutual component of this cycle.
[View Larger Version of this Image (14K GIF file)]


Crystallographic data suggest that specific recognition of guanosine by RNase T1 involves sandwich-like parallel stacking of the guanosine base between the phenolic side chains of Tyr-42 and Tyr-45 (14, 15, 17-20). Our data indicate that this view is no longer valid. The Tyr-45 phenolic side chain does not contribute to substrate binding or to substrate turnover. By constructing the Y42A/Y45A double mutant, we also investigated whether the stacking interaction between the guanine base and Tyr-42 depends on the presence of the Tyr-45 phenolic ring. The kcat/Km ratio for the double mutant is very similar to that for Y42A RNase T1 (Table IV), indicating that the strength of the Tyr-42-guanine base interaction does not depend on Tyr-45. Taken together, all data show that the Tyr-45 side chain, often referred to as the lid of the guanine-binding site (15, 38, 39), does not contribute to the incremental binding energy for guanosine.

From our double mutant cycles, it appears that the Tyr-42/Tyr-45 and Tyr-42/Glu-46 pairs contribute to substrate binding in a mutually independent way. This is consistent with a large data base for free energy changes that result when single mutants are combined (4). Indeed, if two residues do not interact with each other by direct or indirect contact, the sum of the free energy changes derived from the single mutations is nearly always equal to the free energy change measured in the multiple mutant. One major exception where such simple additivity does not apply exists for catalytic residues which, instead, act in a highly cooperative way (5, 7-9).

Evolutionary Implications

Structural conservation of particular enzyme residues among homologous family members is indicative of their structural or functional importance (45). Among the guanine-specific members of the homologous family of microbial ribonucleases, Tyr-42, Tyr-45, and Glu-46 are conserved (11, 12). Both tyrosines are consistently observed at equivalent positions in all eukaryotic members. In the prokaryotic ribonucleases, a phenylalanine occupies position 42 (RNase T1 numbering), and an arginine is found at position 45. Glu-46 is found throughout the family. The conservation of an aromatic ring (Tyr, Phe) at position 42 (RNase T1 numbering) is compatible with the observation that the guanine base interacts with the phenyl part of the side chain. The tyrosine hydroxyl group makes no direct intermolecular contacts with the guanine base. The strict conservation of Glu-46 is easily explained in terms of two specific hydrogen bonds with the guanidinium part of the guanine base. It has been established that the Glu-46 carboxylate takes part in an invariant structural motif that renders guanine specificity (46).

The fact that Tyr-45 is conserved but does not contribute to substrate binding or to catalysis indicates that it fulfills another function. We can only speculate on its role. The Tyr-45 side chain may be involved in the folding and stability of RNase T1. It may prevent other nucleotides from binding the catalytic site. Comparison of the expression of the authentic RNase T1 gene in Saccharomyces cerevisiae and A. oryzae (47) suggests the existence of an intracellular inhibitor for the enzyme. If A. oryzae has a specific inhibitor for its ribonuclease, Tyr-45 may be essential in RNase inhibitor recognition. Arg-59 (barnase numbering), the equivalent of Tyr-45 in barnase, interacts with five barstar residues and contributes ~5 kcal/mol to the intermolecular interaction energy of the barnase-barstar complex (48, 49).


FOOTNOTES

*   This work was supported by the Vlaams Interuniversitair Instituut voor Biotechnologie and the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom correspondence and reprint requests should be addressed. Tel.: 32-2-359-02-48; Fax: 32-2-359-02-89; E-mail: jsteyaer{at}vub.ac.be.
1   The abbreviations used are: NpN, 3',5'-linked dinucleoside phosphate compounds (N represents any of the four common nucleosides); 2'-GMP, 2'-guanylic acid; Wat, water.

ACKNOWLEDGEMENTS

We thank Remy Loris, Rex Palmer, and Andrew Hemmings for assistance during x-ray data collection.


REFERENCES

  1. Ackers, G. K., and Smith, F. R. (1985) Annu. Rev. Biochem. 54, 597-629 [CrossRef][Medline] [Order article via Infotrieve]
  2. Gerlt, J. A. (1987) Chem. Rev. 87, 1079-1105
  3. Knowles, J. R. (1987) Science 236, 1252-1257 [Medline] [Order article via Infotrieve]
  4. Wells, J. A. (1990) Biochemistry 29, 8509-8517 [Medline] [Order article via Infotrieve]
  5. Carter, P., and Wells, J. A. (1988) Nature 332, 564-568 [CrossRef][Medline] [Order article via Infotrieve]
  6. Carter, P., and Wells, J. A. (1990) Proteins Struct. Funct. Genet. 7, 335-342 [Medline] [Order article via Infotrieve]
  7. Weber, D. J., Serpersu, E. H., Shortle, D., and Mildvan, A. S. (1990) Biochemistry 29, 8632-8642 [Medline] [Order article via Infotrieve]
  8. Weber, D. J., Meeker, A. K., and Mildvan, A. S. (1991) Biochemistry 30, 6103-6114 [Medline] [Order article via Infotrieve]
  9. Steyaert, J., and Wyns, L. (1993) J. Mol. Biol. 229, 770-781 [CrossRef][Medline] [Order article via Infotrieve]
  10. Sato, K., and Egami, F. (1957) J. Biochem. (Tokyo) 44, 753-767
  11. Hartley, R. W. (1980) J. Mol. Evol. 15, 355-358 [Medline] [Order article via Infotrieve]
  12. Hill, C., Dodson, G., Heinemann, U., Saenger, W., Mitsui, Y., Nakamura, K., Borisov, S., Tischenko, G., Polyakov, K., and Pavlovsky, S. (1983) Trends Biochem. Sci. 8, 364-369 [CrossRef]
  13. Walz, F. G., Osterman, H. L., and Libertin, C. (1979) Arch. Biochem. Biophys. 195, 95-102 [Medline] [Order article via Infotrieve]
  14. Heinemann, U., and Saenger, W. (1982) Nature 299, 27-31 [Medline] [Order article via Infotrieve]
  15. Arni, R., Heinemann, U., Tokuoka, R., and Saenger, W. (1988) J. Biol. Chem. 263, 15358-15368 [Abstract/Free Full Text]
  16. Steyaert, J., Opsomer, C., Wyns, L., and Stanssens, P. (1991) Biochemistry 30, 494-499 [Medline] [Order article via Infotrieve]
  17. Hakoshima, T., Toda, S., Sugio, S., Tomita, K.-I., Nishikawa, S., Morioka, H., Fushimura, K., Kimura, T., Uesugi, S.-I., Ohtsuka, E., and Ikehara, M. (1988) Protein Eng. 2, 55-61 [Abstract]
  18. Kostrewa, D., Choe, H.-W., Heinemann, U., and Saenger, W. (1989) Biochemistry 28, 7592-7600 [Medline] [Order article via Infotrieve]
  19. Martinez-Oyanedel, J., Choe, H.-W., Heinemann, U., and Saenger, W. (1991) J. Mol. Biol. 222, 335-352 [Medline] [Order article via Infotrieve]
  20. Zegers, I., Haikal, A. F., Palmer, R., and Wyns, L. (1994) J. Biol. Chem. 269, 127-133 [Abstract/Free Full Text]
  21. Steyaert, J., Hallenga, K., Wyns, L., and Stanssens, P. (1990) Biochemistry 29, 9064-9072 [Medline] [Order article via Infotrieve]
  22. Chen, B., and Przybyla, A. E. (1994) BioTechniques 17, 657-659 [Medline] [Order article via Infotrieve]
  23. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 5463-5467 [Abstract]
  24. Mayr, L. M., and Schmid, F. X. (1993) Protein Expression Purif. 4, 52-58 [CrossRef][Medline] [Order article via Infotrieve]
  25. Shirley, B. A., and Laurents, D. V. (1990) J. Biochem. Biophys. Methods 20, 181-188 [Medline] [Order article via Infotrieve]
  26. Pace, C. N., Vajdos, F., Fee, L., Grimsley, G., and Gray, T. (1995) Protein Sci. 4, 2411-2423 [Abstract/Free Full Text]
  27. Zabinski, M., and Walz, F. G. (1976) Arch. Biochem. Biophys. 175, 558-564 [Medline] [Order article via Infotrieve]
  28. Leatherbarrow, R. J. (1991) GraFit, Version 2.07, Biosoft, Cambridge, United Kingdom
  29. CCP4 (1994) Acta Crystallogr. Sect. D 50, 760-763 [CrossRef][Medline] [Order article via Infotrieve]
  30. Doumen, J., Gonciarz, M., Zegers, I., Loris, R., Wyns, L., and Steyaert, J. (1996) Protein Sci. 5, 1523-1530 [Abstract/Free Full Text]
  31. Brünger, A. T. (1992) X-PLOR, Version 3.1, Yale University, New Haven, CT
  32. Jones, T. A., Zou, J. Y., Cowan, S. W., and Kjeldgaard, M. (1991) Acta Crystallogr. Sect. A 47, 110-119 [CrossRef][Medline] [Order article via Infotrieve]
  33. Fersht, A. R. (1988) Biochemistry 27, 1577-1580 [Medline] [Order article via Infotrieve]
  34. Carter, P. J., Winter, G., Wilkinson, A. J., and Fersht, A. R. (1984) Cell 38, 835-840 [Medline] [Order article via Infotrieve]
  35. Fersht, A. R., Leatherbarrow, R. J., and Wells, T. N. C. (1987) Biochemistry 26, 6030-6038 [Medline] [Order article via Infotrieve]
  36. Shimada, I., and Inagaki, F. (1990) Biochemistry 29, 757-764 [Medline] [Order article via Infotrieve]
  37. Ikehara, M., Ohtsuka, E., Tokunaga, T., Nishikawa, S., Uesugi, S., Tanaka, T., Aoyama, Y., Kikyodani, S., Fujimoto, K., Yanase, K., Fuchimura, K., and Morioka, H. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 4695-4699 [Abstract]
  38. Nishikawa, S., Morioka, H., Kimura, T., Ueda, Y., Tanaka, T., Uesugi, S., Hakoshima, T., Tomita, K.-I., Ohtsuka, E., and Ikehara, M. (1988) Eur. J. Biochem. 173, 389-394 [Abstract]
  39. Hakoshima, T., Tanaka, M., Itoh, T., Tomita, K.-I., Amisaki, T., Nishikawa, S., Morioka, H., Uesugi, S.-I., Ohtsuka, E., and Ikehara, M. (1991) Protein Eng. 4, 793-799 [Abstract]
  40. Burley, S. K., and Petsko, G. A. (1985) Science 229, 23-28 [Medline] [Order article via Infotrieve]
  41. Burley, S. K., and Petsko, G. A. (1988) Adv. Protein. Chem. 39, 125-189 [Medline] [Order article via Infotrieve]
  42. Serrano, L., Bycroft, M., and Fersht, A. R. (1991) J. Mol. Biol. 218, 465-475 [Medline] [Order article via Infotrieve]
  43. Hirono, S., and Kollman, P. A. (1991) Protein Eng. 4, 233-243 [Abstract]
  44. Ishida, T., Tarui, M., In, Y., Ogiyama, M., Doi, M., and Inoue, M. (1993) FEBS Lett. 333, 214-216 [CrossRef][Medline] [Order article via Infotrieve]
  45. Poteete, A. R., Rennell, D., and Bouvier, S. E. (1992) Proteins Struct. Funct. Genet. 13, 38-40 [Medline] [Order article via Infotrieve]
  46. Sevcik, J., Sanishvilli, R. G., Pavlovsky, A. G., and Polyakov, K. M. (1990) Trends Biochem. Sci. 15, 158-162 [CrossRef][Medline] [Order article via Infotrieve]
  47. Fujii, T., Yamaoka, H., Gomi, K., Kitamoto, K., and Kumagai, C. (1995) Biosci. Biotechnol. Biochem. 59, 1869-1874 [Medline] [Order article via Infotrieve]
  48. Guillet, V., Lapthorn, A., Hartley, R. W., and Mauguen, Y. (1993) Structure 1, 165-176
  49. Schreiber, G., and Fersht, A. R. (1995) J. Mol. Biol. 248, 478-486 [CrossRef][Medline] [Order article via Infotrieve]
  50. Altona, C., and Sundaralingam, M. (1972) J. Am. Chem. Soc. 49, 8205-8212

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.