The a-lytic protease of Lysobacter enzymogenes was successfully expressed in Escherichia cofi by huing the promoter and dgnal sequence of the E. coil phoA gene to the proenzyme portion of the a-lytic preten gene.Following induction, active enzyme was found both within cells and in the extracellular ue , where it slowly accumulated to high levels. Use of a similar gene fusion to express the protease domain alone produced inactive enzyme, indicatin that the large amino-terminal pro region is necessary for activity. The cinations for protein foding are discussed. Furthermore, inactivation of the protse by mutation of the catalytic serine residue resulted in the production of a higher-molecular-weight form of the a-lytic protease, suggesing that the enzyme is self-processing in E. coli. a-Lytic protease is one of a battery of extracellular enzymes secreted by the gram-negative bacterium Lysobacter enzymogenes to lyse and degrade soil microorganisms. By virtue of its well-studied Asp-His-Ser catalytic triad mechanism (1, 2, 5), its degree of structural homology to mammalian serine proteases (6), and its amenity to nuclear magnetic resonance studies (1, 2, 30), a-lytic protease is an ideal candidate for site-specific mutagenesis studies of substrate specificity and structure-function relationships. We recently reported the cloning and sequence analysis of the oa-lytic protease gene from L. enzymogenes (33). The nucleotide sequence contained a large open reading frame 5' of the coding sequence of the mature enzyme, and we proposed that the additional 199 amino acids made up a 33-amino-acid signal sequence (pre) and a 166-amino-acid pro region. Recent studies with secreted proteases from both grampositive and gram-negative bacteria, including several Bacillus species (17,32,34,(36)(37)(38)41), Neisseria gonorrhoeae (28), Streptomyces griseus (13), and Serratia marcescens (40), have shown that all of these bacterial proteases are synthesized as precursors, although the pro region varies in its amino-or carboxyl-terminal location. Recently, the 77-amino-acid amino-terminal pro region of Bacillus subtilis subtilisin E has been shown to be necessary for the production of active protease, suggesting a critical role for the pro region in folding (15). In this report we provide evidence that the pro region of a-lytic protease has a similar function.We subcloned regions of the a-lytic protease gene behind the promoter and signal sequence of the Escherichia coli phoA gene, allowing production of a-lytic protease in E. coli under conditions of phosphate depletion (16). In addition to confirming the requirement for the pro region, this approach has provided evidence that expression of active enzyme is temperature sensitive in E. coli and that a-lytic protease has the ability to proteolytically process itself. The implications of these results for protein folding are discussed. Strains, media, and expression. DG98, an F'-carrying strain used for M13 phage growth, was described previously (33). MH1 (araD139 AlacX74 galU galK hsr rps...

Binding pocket mutants of alpha-lytic protease (Met 192----Ala and Met 213----Ala) have been constructed recently in an effort to create a protease specific for Met just prior to the scissile bond. Instead, mutation resulted in proteases with extraordinarily broad specificity profiles and high activity [Bone, R., Silen, J. L., & Agard, D. A. (1989) Nature 339, 191-195]. To understand the structural basis for the unexpected specificity profiles of these mutants, high-resolution X-ray crystal structures have been determined for complexes of each mutant with a series of systematically varying peptidylboronic acids. These inhibitory analogues of high-energy reaction intermediates provide models for how substrates with different side chains interact with the enzyme during the transition state. Fifteen structures have been analyzed qualitatively and quantitatively with respect to enzyme-inhibitor hydrogen-bond lengths, buried hydrophobic surface area, unfilled cavity volume, and the magnitude of inhibitor accommodating conformational adjustments (particularly in the region of another binding pocket residue, Val 217A). Comparison of these four parameters with the Ki of each inhibitor and the kcat and Km of the analogous substrates indicates that while no single structural parameter consistently correlates with activity or inhibition, the observed data can be understood as a combination of effects. Furthermore, the relative contribution of each term differs for the three enzymes, reflecting the altered conformational energetics of each mutant. From the extensive structural analysis, it is clear that enzyme flexibility, especially in the region of Val 217A, is primarily responsible for the exceptionally broad specificity observed in either mutant. Taken together, the observed patterns of substrate specificity can be understood to arise directly from interactions between the substrate and the residues lining the specificity pocket and indirectly from interactions between peripheral regions of the protein and the active-site region that serve to modulate active-site flexibility.

Abstract. ot-Lytic protease is a bacterial serine protease of the trypsin family that is synthesized as a 39-kD preproenzyme (Silen, J. L., C. N. McGrath, K. R. Smith, and D. A. Agard. 1988. Gene (Amst.). 69: 23%244). The 198-amino acid mature protease is secreted into the culture medium by the native host, Lysobacter enzymogenes (Whitaker, D. R. 1970. Methods Enzymol. 19:599-613). Expression experiments in Escherichia coli revealed that the 166-amino acid pro region is transiently required either in cis ( Here we show that full-length precursor produced at nonpermissive temperatures is tightly associated with the E. coli outer membrane. The active site mutant Ser 195--'Ala (SA195), which is incapable of self-processing, also accumulates as a precursor in the outer membrane, even when expressed at permissive temperatures. When the protease domain is expressed in the absence of the pro region, the misfolded, inactive protease also cofractionates with the outer membrane. However, when the folding requirement for either wild-type or mutant protease domains is provided by expressing the pro region in trans, both are efficiently secreted into the extracellular medium. Attempts to separate folding and secretion functions by extensive deletion mutagenesis within the pro region were unsuccessful. Taken together, these results suggest that only properly folded and processed forms of ot-lytic protease are efficiently transported to the medium.

No abstract