Mechanisms For The Impact Of T Cell Escape Mutations And Their Preexisting Compensatory Amino Acids On HIV-1Fitness

Human immunodeficiency virus type1(HIV-1) is a retrovirus that infectsprimarily CD4+T cells and causes the acquired immunodeficiency syndrome(AIDS). HIV-1infection is primary caused by a single transmitted/founder (T/F)virus. However, due to the high mutation rate (3×10-5substitutions/site/generation),the rapid virus turnover (108to109virions per day), large numbers of infected cells(107to108infected cells), and a high level of recombination, HIV-1can rapidlygenerate a population of genetically distinct but related variants called aquasispecies soon after infection. The continous increase in genetic diversity enablesHIV-1to quickly adapt to a variety of the host immune pressures. According topopulation genetics, HIV-1fitness is defined as a variant’s ability to contribute tosuccessive generations. Within a particular infected host, HIV-1fitness is defined byhow viruses affect fecundity (or replication) and survival within a particular hostimmune environment. As one of the most important viral intrinsic characteristics, amodest change in HIV-1fitness can have a significant impact on viral evolution,pathogenesis, transmission and disease progression.Soon after HIV-1infection, the host will mount potent cytolytic T lymphocyte(CTL) responses. However, under strong selection pressures from these CTLresponses, viruses can quickly develop resistance through mutations, such as theT242N mutation in the TW10epitope (TSTLQEQIGW, Gag240~249) inHIV-1-infected individuals carrying HLA-B*57/5801alleles, to render the CTLresponses ineffective. But the CTL escape mutation T242N can result in significantfitness loss which may lead to long-term HIV-1control and a decreased probabilityof transmission to new hosts. Moreover, if a T242N mutant is transmitted into newhosts, a better clinical outcome is expected, although the effect may not be sustainedinto chronic infection. Therefore, epitopes with significant fitness loss due to CTL escape mutations will be desirable targets for developing effective HIV-1vaccines.However, recent results have suggested that the fitness costs due to the T242Nmutation can be partially or completely restored by its related compensatorymutations in or outside of the TW10epitope. Furthermore, reversion of the escapemutation T242N back to the subtype ancestral state is frequently detected in thenewly infected hosts without HLA-B*57/5801. It has been considered that reversionof T242N mutationcan render the viruses more fit, although it generally takes morethan one year to occur in vivo.However, the fitness costs of the CTL escape mutation T242N have only beentested in heterologous viral genomes that may have contained compensatorymutations. More critically, previous methods have been used to study the impact ofthe T242N mutation on unrelated viruses instead of the cognate viruses from whichit was selected in vivo. Thus, the impact of different context in diverse HIV-1strainson the fitness costs due to the T242N mutation has not been accurately andsystematically investigated. In this study, to better understand the extent of thefitness costs of the T242N mutation and the repair of fitness loss through itsassociated compensatory amino acids, we investigated its fitness impact in differentcognateT/F viruses. First, we analyzed whole genome sequences of longitudinalsamples from4acutely infected individuals (AHIs). We mapped the CTL epitopesand generated infectious molecular clones (IMCs) based on the inferred T/F viralgenome sequences. The T242N mutation was selected in CH58during infectionwhile the T242N mutation was present in the T/F virus in CH131but reverted backto the wild type N242T post infection; the T242N mutation was not selected inCH470and CH40because both subjects have no restriction HLA alleles B*57/5801.Using a newly established PASS fitness assay (PFA) that can preciselydetermine the fitness impact of CTL escape mutations in their cognate viral genomesand the repair of fitness loss by compensatory mutations. We found that the T242Nmutation resulted in various levels of fitness loss in different T/F viruses. However, the fitness costs were significantly compromised by preexisting compensatoryamino acids in (isoleucineat position247in Gag) or outside (glutamine at position219in Gag) the TW10epitope.Moreover, the transmitted T242N escape mutant in subject CH131was as fit asthe reversion mutant N242T and the elimination of the compensatory amino acidI247in the T/F viral genome resulted in significant fitness cost, suggesting thefitness loss caused by the T242N mutation had been fully repaired in the donor attransmission. These findings may explain why the CTL escape mutation T242N isprevalent in sizable proportions in the global HIV-1population.During HIV-1infection, mutations are often strongly selected and fixed in theviral population under strong host selection pressures, primarily from CTL andneutralizing antibody (nAb) responses. We identified two strongly selectedmutations K43R and N323TC in Gag gene of CH131T/Fduring infection. Using theex vivo IFN-γ enzyme-linked immunospot (ELISpot) assay, we found no T cellresponses targeting those sites, suggesting that they were not selected by CTLresponses. It could not be selected by the nAb responses since no nAbs target theGag protein. However, the K43R mutation caused a significant fitness loss in itscognate T/F virus while another strongly selected synonymous mutation N323TChad no detectable impact on viral fitness. The quick fixation of the K43R mutationwith the significant fitness loss in the viral population demonstrated that such amutation was strongly selected in vivo. This finding indicates that host factors otherthan immune responses can have a strong selection pressure on viral evolution anddrive the genetic diversification of HIV-1in vivo.In summary, we demonstrated that the fitness loss caused by theT242Nmutation could be significantly affected by the differing context of the viralbackbones because of preexisting compensatory amino acids. Thus, high prevalenceof the compensatory amino acids could significantly mitigate the benefits of thefitness costs caused by the T242N escape mutation and the T242N mutation might become more prevalent at the population level. Importantly, the transmitted T242Nescape mutant could be as fit as the reversion mutant, suggesting its fitness losscould have been fully repaired by the compensatory amino acids in the donor at thetransmission. With little or no fitness costs, some CTL escape mutants could be astransmissible and pathogenic as the wild type viruses. Other studies showed that Tcell escape mutations with little or no fitness costs could be prevalent in the HLAmatched and mismatched populationsor persisted for a long time without revertingback to the wild type. Thus, the increased prevalence of the CTL escape mutationsconsequent on preexisting compensatory amino acids at the population level mayfurther reduce the benefits of the protective HLA alleles to slow HIV diseaseprogression and pose challenges for designing T cell based vaccines.