Viral replication fitness is defined as the capacity of a virus to produce infectious progeny in a given environment1 and is an important contributing factor in determining the prevalence of a virus variant at the population level over time2. As in vivo fitness studies are not feasible with pathogenic human viruses, such as HIV-1, various in vitro and ex vivo replication fitness assays have been developed to study the effects on fitness arising from drug resistance and immune escape mutations, epistasis and the evolution of viral populations3-6. Among different fitness assays, growth competition assays are recognized to yield more sensitive and valid measures of fitness differences, as two or more viral variants compete for the same cell population under precisely the same environmental conditions, as occurs in vivo1,7,8. Before starting growth competition experiments, several variables need to be determined, including the use of different multiplicities of infection (MOI), viral input ratio, and timing of sampling for analysis. We have studied the effects of these parameters on viral growth kinetics and on the outcome of competition experiments, and have identified key factors necessary for robust measurements of HIV-1 fitness in cell culture9.
In addition to assay variables, there are a variety of methods for quantitating viral variants in growth competition experiments. Bulk10,11 or clonal sequencing12,13 has been used to determine the ratio of the competing viruses based on the nucleotide frequencies at the site(s) of interest. Relative fitness is derived from changes in this ratio over time. This method is convenient as DNA sequencing services are widely available. The parallel allele-specific sequencing (PASS) method enables sequencing at multiple sites and the detection of recombinants14, but it also requires specifically developed reagents and detection systems. Essentially, these methods were developed to study viral strains with a small number of nucleotide differences in a region of interest. Other methods use a small reporter gene15 or synonymous mutations4,16-18 as tags to distinguish the competing viruses by sequencing, heteroduplex tracking assays (HTA)4,7,17,19 or reverse transcription-quantitative real-time polymerase chain reaction (RT-qPCR)15,16,18,20, all of which can be made applicable to study competing strains regardless of sequence similarity. An additional step is required to introduce tags into viral genomes and the RT-qPCR assay also requires specific reagents and instrumentation. We have found that bulk Sanger sequencing yields comparable results9.
Following growth competition, viral replication fitness is presented as relative fitness, or a fitness ratio between two viral variants. The relative fitness of a virus can be defined as the final proportion of a viral variant normalized by its initial proportion in the inoculum or as the net growth rate difference between the two competing viruses. We found that the latter method, using longitudinal data points only within the exponential growth phase, produced the most robust results9,20.
In vitro fitness assays are used primarily to study biological clones6-8 and infectious molecular clones of HIV-1. The latter, being amenable to genetic manipulation, are often employed to study the effect on fitness from particular mutations or specific sequences of interest3-5,21,22. The following protocols describe a workflow from the point of constructing new full-length infectious HIV-1 molecular clones using HIV-1 vectors containing a sequence tag, introducing mutations of interest, making viral stocks and establishing viral growth kinetics, to performing the growth competition assay and calculating relative fitness (Figure 1).
Using our optimized procedures, we created three recombinant HIV-1 mutants and determined their replication fitness. The recombinant molecular clone was first constructed by replacing the HIV-1 gag-p24 gene region of pNL4-3, a plasmid containing a full length infectious genome of HIV-1 lab strain NL4-3, with a synthetic COTB (Center-Of-Tree, subtype B) gag-p24 sequence23 to create the prototype strain. Single amino changes (T186M, T242N, and I256V) were then introduced to create three mutant clones. Each mutant was competed against the prototype virus to observe the fitness impact of each mutation in the given genetic background. The three mutants demonstrated varying levels of replication fsitness from slight to significantly lower than the prototype virus. The T242N mutation was previously reported to have a moderate fitness cost24-26, similar to the result shown in this study. The fitness cost of the other two mutations had not been reported previously.