Regular Article
The structural basis for enhanced stability and reduced DNA binding seen in engineered second-generation cro monomers and dimers1

https://doi.org/10.1006/jmbi.1999.3498Get rights and content

Abstract

It was previously shown that the Cro repressor from phage λ, which is a dimer, can be converted into a stable monomer by a five-amino acid insertion. Phe58 is the key residue involved in this transition, switching from interactions which stabilize the dimer to those which stabilize the monomer. Structural studies, however, suggested that Phe58 did not penetrate into the core of the monomer as well as it did into the native dimer. This was strongly supported by the finding that certain core-repacking mutations, including in particular, Phe58→Trp, increased the stability of the monomer. Unexpectedly, the same substitution also increased the stability of the native dimer. At the same time it decreased the affinity of the dimer for operator DNA. Here we describe the crystal structures of the Cro F58W mutant, both as the monomer and as the dimer. The F58W monomer crystallized in a form different from that of the original monomer. In contrast to that structure, which resembled the DNA-bound form of Cro, the F58W monomer is closer in structure to wild-type (i.e. non-bound) Cro. The F58W dimer also crystallizes in a form different from the native dimer but has a remarkably similar overall structure which tends to confirm the large changes in conformation of Cro on binding DNA. Introduction of Trp58 perturbs the position occupied by the side-chain of Arg38, a DNA-contact residue, providing a structural explanation for the reduction in DNA-binding affinity.

The improved thermal stability is seen to be due to the enhanced solvent transfer free energy of Trp58 relative to Phe58, supplemented in the dimer structure, although not the monomer, by a reduction in volume of internal cavities.

Introduction

Native proteins have presumably evolved to achieve an optimal balance between stability and activity. Mutation can increase stability or activity, but it is difficult to simultaneously improve both. For a number of reasons the Cro protein from bacteriophage λ serves as a useful system to explore this issue. The DNA binding of the protein has been extensively characterized Ptashne 1986, Takeda et al 1989. Also, the structure of the wild-type Cro dimer and a number of variants in free and DNA-bound forms have been determined by X-ray crystallography and NMR Anderson et al 1981, Brennan et al 1990, Matsuo et al 1995, Albright and Matthews 1998, Mossing 1998, Ohlendorf et al 1998. A stable Cro monomer has also been engineered, demonstrating that the intermolecular contacts that stabilize the native dimer can be substituted by intramolecular interactions within a single polypeptide chain Mossing and Sauer 1990, Albright et al 1996, Mossing 1998.

Cro represses transcription by binding as a homodimer to palindromic DNA operator sites in phage λ (Ptashne, 1986). The X-ray structure (Anderson et al., 1981) showed that the Cro dimer is held together by two antiparallel β-strands, with Phe58 of one subunit buried within the core of the other subunit (cf. Figure 1(a)). The connection between the subunits seems inherently flexible and, upon binding, DNA was observed to twist and flatten, resulting in a 40° rotation of one subunit relative to the other (Figures 1(a) and (b))Brennan et al 1990, Albright and Matthews 1998.

By inserting five amino acids into the third β-strand, Mossing & Sauer (1990) successfully designed a stable monomeric version of Cro (cf. Figure 1(c)) which was more stable than the native dimer. A high-resolution structure determination (Albright et al., 1996) confirmed the design strategy, but revealed several unexpected features. One of these was an apparent “loosening” of packing within the core of the monomer due to the “retraction” of Phe58 by about 1 Å.

This prompted Mollah et al. (1996) to try to achieve better packing of the core by randomization of the three spatially adjacent residues Ala29, Ala33 and Phe58 itself. This proved successful, the largest gain in stability coming from a single substitution of Phe58 to Trp. The substitution increased the melting temperature by 5 deg. C to give 0.74 kcal/mol of additional stability.

Given these findings, the Phe58→Trp mutation was then introduced into the native Cro dimer and found to increase substantially its stability, but to reduce operator binding. Interestingly, the monomer-dimer equilibrium is only slightly shifted by the F58W substitution, implying that both the folded monomer and the folded dimer are similarly stabilized. Taking the dimerization effect into consideration the intrinsic affinity for DNA is reduced approximately ten to 25-fold (Jana et al., 1997).

In order to provide a structural basis for these observations we have determined the monomeric and dimeric structures of Cro containing the Phe58→Trp mutation. In both cases the introduction of the mutation causes the protein to crystallize in a new form.

Section snippets

Nomenclature

The original monomer of Cro designed by Mossing & Sauer (1990) is designated “monomer” or, more formally, “Cro K56-[DGEVK]“, indicating that the five amino acid residues within the square parentheses have been inserted following Lys56. The full designation of the monomer including the substitution Phe58→Trp is “Cro K56-[DGEVK]-F58W” or, more simply, “mutant monomer”. Wild-type Cro dimer is simply designated “Cro” or “Cro dimer”. The mutant version of the dimer is identified as “Cro-F58W” or,

Discussion

Protein stability can, in principle, be achieved by substituting larger more hydrophobic residues into the core of a protein so long as strain is not introduced (Karpusas et al., 1989). Based on inspection of the respective structures, the substitution of Phe58→Trp appears to introduce little strain. The χ1 and χ2 angles change little at the mutation site, and are comparable to average values in protein structures Ponder and Richards 1987, Laskowski et al 1993. At the same time, there are

Crystallization and data collection

Cro K56-[DGEVK]-F58W protein was expressed and purified in 20 mM K2HPO4 (pH 7.0), 200 mM KCl and 0.1 mM EDTA as described by Mollah et al. (1996). The protein was dialyzed into 200 mM KCl, 20 mM Hepes (pH 7.0) and 0.1 mM EDTA and concentrated to 25 mg/ml with a 3000 MW cut-off Centricon filter and centrifugation. Diffraction-quality crystals were obtained using the hanging drop method by equilibrating over a precipitant buffer of 4.8 M sodium formate and 0.5 % (w/v) β-octylglucoside. Crystals

Acknowledgements

We thank Dr Enoch Baldwin for assistance in data collection for the monomer structure and Drs Martin Sagermann and Mike Quillin for assistance in computing issues. This work was supported in part by NIH grants GM46513 to M.C.M. and GM20066 to B.W.M.

References (28)

  • W.F. Anderson et al.

    Structure of the Cro repressor from bacteriophage λ and its interaction with DNA

    Nature

    (1981)
  • R.G. Brennan et al.

    Protein-DNA conformational changes in the crystal structure of a lambda Cro-operator complex

    Proc. Natl Acad. Sci. USA

    (1990)
  • M.L. Connolly

    Analytical molecular surface calculation

    J. Appl. Crystallog.

    (1983)
  • D. Eisenberg et al.

    Solvation energy in protein folding and binding

    Nature

    (1986)
  • Cited by (7)

    • Analysis of self-association of West Nile virus capsid protein and the crucial role played by Trp 69 in homodimerization

      2005, Biochemical and Biophysical Research Communications
      Citation Excerpt :

      When Trp 69 was replaced by another hydrophobic residue Phe, the self-association of WN virus C protein was significantly decreased (p = 0.001) as shown in Fig. 5A. This might be due to the significant alteration of the backbone structure, changes in solvent accessible surface area, solvent transfer free energy or the conformational changes induced by the substitution [30]. When Trp 69 was replaced with Phe, dimerization was retained, demonstrating the importance of a hydrophobic environment at this position.

    View all citing articles on Scopus
    1

    Edited by T. Richmond

    2

    Present addresses: P. B. Rupert, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N, A3015, PO Box 19024, Seattle, WA 98109-1024, USA; A. K. M. M. Mollah, Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA; M. C. Mossing, Department of Chemistry, 452 Coulter Hall, University of Mississippi, University, MS 38677, USA.

    View full text