Rice team makes a stable Cas13a 100-fold more sensitive by editing one of two regions

The starting enzyme was stable and insensitive. It withstood harsh assay conditions but could not detect very small amounts of target RNA ^[2]. First author Seye Oladeji, a Rice graduate student, said: "Our stable Cas13a variant was just not sensitive enough: a perfect problem for an engineering approach." ^[4] Yang Gao, the corresponding author, set out the design constraint: "Any enzyme used in an assay needs to be both sensitive and stable." ^[3]

Cas13a moves a lot as it binds. That makes it hard to capture in the detail needed to resolve a structure, so Oladeji designed a way to hold the enzyme and its target RNA in place while collecting data ^[5]. The structure let the team pick two regions it believed would raise sensitivity, and it tested the results in the lab ^[6]. Altering both made the enzyme a little too sensitive: it signaled even when no target RNA was present ^[7]. Altering just one raised sensitivity 100-fold with no loss of stability ^[8]. "It was exciting and a good reminder to be careful not to overengineer a protein," Oladeji said ^[9].

The result does not show how the enzyme behaves outside a controlled sample. According to the phys.org account of the study, the 100-fold improvement was measured against a synthetic, noninfectious HIV sample in the team's own detection assay ^[10]. The write-up does not include replicate numbers or an absolute detection limit. Oladeji described the ideal assay as one that can tell you if there is as little as one strand of RNA per microliter of liquid ^[11]. A 100-fold gain from an unstated baseline cannot be checked against that target.

Only HIV was tested. The enzyme is paired with a guide programmed to a specific RNA sequence. The team says it could reprogram that guide to detect SARS-CoV-2, influenza, Ebola, Zika, respiratory syncytial virus or hepatitis C instead ^[13]. In our view that is sound design logic, and each of those targets still needs its own test. Gao said: "Our version of Cas13a is optimized for sensitive and stable RNA detection assays and could potentially be used in commercial assays like RT-qPCR and lateral flow." ^[14]