Constructing Enzymes With Novel Functions Through Domain Swapping

Constructing enzymes with desired properties is a major goal for proteinengineers. Recombination is an important mechanism for natural evolution of proteins,and presents an effective strategy for exploring protein sequence space. Diversefunctional domains in proteins provide a resource for designing novel biocatalysts.The recombination of functional protein modules, such as domains or subdomains,from diverse homologous enzymes by conventional domain recombination has led tochimeric enzymes with novel activities, altered substrate specificities, and improvedstability. Recombining proteins with less sequence similarity would offer betteropportunities for creating novel proteins. The challenge has been that the domainsfrom parental proteins with less sequence identity often have structurallynon-compatible interfaces when they are recombined, resulting in inactive chimeras.Thus, further advances in rational protein engineering require a profoundunderstanding of how proteins evolve and new strategies for the efficient generationof chimeras. In this work, we constructed chimeras through two different strategies:traditional domain recombination combining with domain interface refinement; andKey Motif Directed Recombination.AFEST is a carboxylesterase from Archaeoglobus fulgidus, hydrolyzesp-nitrophenyl esters (pNP) with short acyl chain lengths, and shows the highestactivity towards pNPC4(para-nitrophenyl-butyrate). apAPH is an acylpeptidehydrolase from Aeropyrum pernix K1(apAPH) and shows a promiscuous esteraseactivity with a preference for middle chain length substrates, whose favorite ester substrate is pNPC8(para-nitrophenyl-caprylate). In order to study thestructure-function relationship of the hyperthermophilic esterases apAPH and AFEST,we exchanged substrate binding domains of the two enzymes, and optimized thenewly formed domain interfaces through site directed mutations. We obtained twofunctional chimeric enzymes PAR and AAM7. Characterization of these two chimericenzymes showed that they inherited the thermophilic properties of the parent proteins.The kinetic parameters indicated that, the esterase substrate specificities of PAR andAAM7were similar to the parent which provided them with the substrate bindingdomain, respectively. Our experimental results showed that: a, the substrate-bindingdomains play a dominant role for the enzyme substrate specificity; b, the optimizationof newly formed domain interface is an important guarantee for obtaining correctlyfolded and stable chimeric enzymes constructed by domain swapping of parents withlow sequence identity.In one superfamily, proteins can differ greatly in their primary amino acidsequences and in their biological functions. Traces of ancient functional amino acidsequence motifs and structural motifs are, however, often detectable, suggesting thatat least some members of a protein superfamily may be derived from one or morecommon ancestors. By comparing the sequence and structure homologues of theα/β-hydrolase fold superfamily, we discovered the existence of key motifs, which areconserved amino acid sequences within a conserved structural motif. Key motifsrepresent a localized similar structural environment in distant proteins within a proteinsuperfamily. Here, we explored the idea of recombining distant sequence proteins atkey motif regions to engineer novel enzymes. We developed a new domainrecombination strategy-Key Motif Directed Recombination (KMDR). ThroughKMDR, We designed and constructed chimeric proteins from three catalyticallydistinct members of the α/β hydrolase fold superfamily: two hyperthermophilicesterases (AFEST and apAPH) and a mesophilic lipase (Lip1from Candida rugosa)with less than20%sequence identity. The chimeras LAf and LAp retained thesubstrate preference of Lip1for long acyl chain ester substrates and the thermophilic property of the parent AFEST and apAPH.Recruiting evolved structural domains or subdomains with desired functions frommembers of a superfamily provides a powerful approach for generating largefunctional leaps. As an innovation for domain recombination at inner, rigid areas witha key motif, rather than flexible regions between domains, KMDR thus represents apractical method for the efficient manipulation and recombination of structuraldomains with distantly related sequences. The ability to efficiently design proteinscombining distantly related sequences provides a means to create new biocatalyststhat are useful for fundamental research, and accelerating the commercialization ofvarious products, from pharmaceuticals to fine chemicals. Our work not only shedlight on the efficient recombination of very dissimilar protein sequences within asuperfamily, but also provided the first evidence that key motifs may have playedimportant roles in natural protein evolution.