Dissecting the chitinolytic activity of mammalian chitinases

Chitin is an abundant polysaccharide used by a large range of organisms for structural rigidity and water repulsion. As such, the insoluble crystalline structure of chitin poses significant challenges for enzymatic degradation. Vertebrates do not produce chitin, but do express chitin degrading enzymes. Acidic mammalian chitinase, the primary enzyme involved in the degradation of environmental chitin in mammalian lungs, is a processive glycosyl hydrolase that may be able to make multiple hydrolysis events for each binding event. Mutations to acidic mammalian chitinase have been associated with asthma, and genetic deletion of the enzyme in mice results in significantly increased morbidity and mortality with age. We initially set out to reverse this phenotype by engineering hyperactive acidic mammalian chitinase variants. Using a directed evolution screening approach using commercial fluorogenic substrates, we identified mutations with consistent increases in activity. To determine whether the activity increases observed with oligomeric substrates were consistent with more biologically relevant chitin substrates, we developed new assays to quantify chitinase activity with colloidal crystalline chitin, and identified a high throughput fluorogenic assay that gives sufficient signal to noise advantages to quantify changes to activity due to the addition or removal of a chitin binding domain to the enzyme. We show that the activity increasing mutations derived from our directed evolution screen were lost when crystalline substrates were used. In contrast, naturally occurring gain-of-function mutations gave similar results with oligomeric and crystalline substrates. We also show that the activity differences between acidic mammalian chitinase and chitotriosidase are reduced in the context of crystalline substrate, suggesting that previously reported activity differences with oligomeric substrates may have been largely driven by differential substrate specificity for the oligomers. These results highlight the need for assays against more physiological substrates when engineering complex metabolic enzymes, and provide a new approach that may be broadly applicable to engineering glycosyl hydrolases.