A high throughput system for phenotypic analysis of NS3 sequences
J. Hong1; X. Qin1; S. Lim1; L. M. Blatt1; S. Seiwert1
1. Intermune, Inc, Brisbane, CA, USA.
Background: Systems to evaluate the phenotypic characteristics of NS3 sequences have many uses including the evaluation of drug sensitivity of patient-isolated NS3 following drug treatment. ITMN-191 (R7227) is a potent, slowly dissociating macrocyclic inhibitor of NS3/4A protease that has displayed robust antiviral activity in initial clinical studies. To support these studies, as well as larger future studies, we have developed a novel, scalable system to characterize drug sensitivity of NS3 sequences isolated from HCV patients.
Methods: NS3 protease domain is amplified using nested PCR. Amplicons allow determination of a consensus or population sequence and are cloned into a “phenotyping” plasmid which inserts NS3 sequence upstream of an NS4A sequence, an endoplasmic reticulum (ER) tethering sequence and a secreted luciferase (sLuc) sequence, each of which are separated by a NS3 cleavage site. Transfection into 293F suspension cells enables homogenous, solution based detection of NS3 activity: NS3 activity liberates sLuc from its ER tether allowing its secretion from cells. Dose response curves can be determined since cells treated with greater amounts of NS3 inhibitor release less sLuc into media. Automation of the above process allows determination of ~100 EC50 values per day.
Results: Repeated analysis of drug sensitivity of a single NS3 genotype 1b sequence over three days provided consistent EC50 values (EC50 = 1.5 ± 1.0 nM, n = 28, CV ~66%, average R2 of 0.93 for dose response curves), indicating the assay has a high degree of precision. Analysis of 72, 48, and 24 quasi-species derived from three patients harboring genotype 1b HCV provided average EC50 = 1.5 nM, 0.6 nM, and 0.9 nM, respectively, for a close analog of ITMN-191. Importantly, mixing of quasi species prior to mammalian cell transfection in a population-based assay yielded drug sensitivities that were within 2-fold of the average drug sensitivities when individual sequences were characterized in a clonal approach, indicating population-based phenotyping is viable. Characterization of NS3 sequence variants identified in the course of in vitro selection experiments demonstrated the phenotyping assay recapitulates the rank order potency observed when these substitutions are introduced into HCV replicons.
Conclusions: The methods reported here to analyze NS3 sequence and drug sensitivity provide a robust and scalable solution for the characterization of NS3 sequences derived from clinical samples. This system accurately reflects drug sensitivities observed in the HCV replicon system and will be deployed to analyze NS3 sequences obtained from ongoing clinical studies of ITMN-191.
