Search

Fitness landscapes reflect the adaptive potential of viruses. There is no information on how fitness peaks evolve when a virus replicates extensively in a controlled cell culture environment. Here we report the construction of Self-Organized Maps (SOMs), based on deep sequencing reads of three amplicons of the NS5A-NS5B-coding region of hepatitis C virus (HCV). A two-dimensional neural network was constructed and organized according to sequence relatedness. The third dimension of the fitness profile was given by the haplotype frequencies at each neuron. Fitness maps were derived for 44 HCV populations that share a common ancestor that was passaged up to 210 times in human hepatoma Huh-7.5 cells. As the virus increased its adaptation to the cells, the number of fitness peaks expanded, and their distribution shifted in sequence space. The landscape consisted of an extended basal platform, and a lower number of protruding higher fitness peaks. The function that relates fitness level and peak abundance corresponds a power law, a relationship observed with other complex natural phenomena. The dense basal platform may serve as spring-board to attain high fitness peaks. The study documents a highly dynamic, double-layer fitness landscape of HCV when evolving in a monotonous cell culture environment. This information may help interpreting HCV fitness landscapes in complex in vivo environments.IMPORTANCEThe study provides for the first time the fitness landscape of a virus in the course of its adaptation to a cell culture environment, in absence of external selective constraints. The deep sequencing-based self-organized maps document a two-layer fitness distribution with an ample basal platform, and a lower number of protruding, high fitness peaks. This landscape structure offers potential benefits for virus resilience to mutational inputs.

Despite the high virological response rates achieved with current directly-acting antiviral agents (DAAs) against hepatitis C virus (HCV), around 2% to 5% of treated patients do not achieve a sustained viral response. Identification of amino acid substitutions associated with treatment failure requires analytical designs, such as subtype-specific ultra-deep sequencing (UDS) methods for HCV characterization and patient management. Using this procedure, we have identified six highly represented amino acid substitutions (HRSs) in NS5A and NS5B of HCV from 220 patients who failed therapy, which are not bona fide resistance-associated substitutions (RAS). They were present frequently in basal and post-treatment virus of patients who failed therapy to different DAA-based therapies. Contrary to several RAS, HRSs belong to the acceptable subset of substitutions according to the PAM250 replacement matrix. Coherently, their mutant frequency, measured by the number of deep sequencing reads within the HCV quasispecies that encode the relevant substitutions, ranged between 90% and 100% in most cases. Also, they have limited predicted disruptive effects on the three-dimensional structures of the proteins harboring them. Possible mechanisms of HRS origin and dominance, as well as their potential predictive value of treatment response are discussed.