HIV Evolution: CTL Escape Mutation and Reversion After Transmission

A J Leslie; K J Pfafferott; P Chetty; R Draenert; M M Addo; M Feeney; Y Tang; E C Holmes; T Allen; J G Prado; M Altfeld; C Brander; C Dixon; D Ramduth; P Jeena; S A Thomas; A St John; T A Roach; B Kupfer; G Luzzi; A Edwards; G Taylor; H Lyall; G Tudor-Williams; V Novelli; J Martinez-Picado; P Kiepiela; B D Walker; P J R Goulder


Nat Med. 2004;10(3) 

In This Article


These data provide the first clear evidence of reversion after transmission of an HIV-1 CTL epitope in which escape has occurred in the donor. We show that T242N is an escape mutation in the TW10 epitope that is selected in the majority of individuals possessing either HLA-B57 or HLA-5801. B57/5801-positive individuals can transmit the T242N mutant, and such transmission has occurred in numerous instances. The T242N mutation was found in 0 of 187 of HLA-B57/5801-negative individuals with chronic HIV infection, compared with 4 of 19 individuals in the acute phase of infection (P = 0.00005), providing strong evidence that T242N is transmitted and, in the absence of HLA-B57/5801, reverts to wild type after acute infection. We present longitudinal data from three HLA-B57/5801-negative subjects to demonstrate that the transmitted T242N variant is replaced by wild-type virus over time. Further evidence of T242N transmission, followed by reversion, derives from the analysis of genetic footprints that are linked with T242N in HLA-B57/5801-positive subjects, and that can be identified in B57/5801-negative subjects in whom T242N is no longer present.

A previous study examining escape mutations in the HLA-B27-restricted epitope KK10, a second epitope associated with long-term control of HIV infection, indicated that the occurrence of reversion after transmission in this instance is rare.[10] Intrapatient reversion after progression to AIDS is also rare, having been suggested by only a single clone out of >150 clones analyzed from that study and others.[10,24] Thus, the B27-KK10 escape variants generally remain stable in the absence of the evolutionary pressure that originally selected them.[10,24] This contrasts with T242N, which consistently reverts after transmission to a B57/5801-negative individual. This distinction highlights the importance of determining the fate of CTL escape mutations. If they are transmissible and stable after transmission, they will spread through the population and the epitope will eventually become extinct. Conversely, if the escape mutant reverts, the epitope will remain useful for CTL-based vaccine design.

The likely explanation for T242N reversion is that there is an associated fitness cost. Drug-resistance mutations present an obvious analogy, as they typically do not persist in untreated patients, despite conferring greater fitness in the presence of ART, because they impair the replication of wild-type virus.[25] Thus, ART-resistant mutants may pay an 'evolutionary penalty' for their survival.[26] Similarly, the T242N mutation may represent the evolutionary penalty paid by HIV to escape TW10-specific CTLs. TW10 is the immunodominant CTL response in acute infection,[19] so even an attenuated virus would have a selection advantage over one that is recognized and eliminated by TW10-specific CTLs. Recent data indicate that in the rare HLA-B57-positive individuals in whom the TW10 epitope has not mutated, viremia is successfully suppressed for decades to <50 copies/ml (ref. 27). Once selection pressure from TW10-specific CTLs is removed, as in transmission to a non-B57/5801-expressing individual, one would expect natural selection to return the virus to its fitter, wild-type state. Even minor enhancements of viral fitness can quickly lead to replacement of the less fit variants in a population.[28]

One might further speculate that the early dominance of the TW10-specific response in acute infection, followed by the emergence of escape variants conferring an attenuated replicative state upon the virus, could be causally linked to the association of HLA-B57/B5801 with long-term suppression of HIV infection. In our ART-naive C-clade-infected study group, HLA-B57/5801-expressing subjects whose virus encoded T242N had a median viral load of 10,342 HIV RNA copies per ml plasma, still significantly lower than in the HLA-B57/5801-negative group (median viral load = 28,125 HIV RNA copies per ml plasma; P = 0.02 by Mann-Whitney test). This T242N mutation may therefore contribute to the control of viremia observed in infected persons expressing HLA-B57/B5801, despite the loss of CTL recognition through this TW10 specificity.

The precise mechanism by which the T242N mutation might affect viral fitness is unknown. Gag residue 242 falls within the HIV capsid protein, p24, which has crucial roles in viral function[29] and is a highly conserved region of the HIV genome.[30] Single-point mutations in p24 can have a profound effect on viral fitness,[31,32,33,34] and linker-insertion mutations between residues 241-242 and 242-243 destroy HIV's ability to replicate.[32,35,36] T242N mutant viruses can persist for at least 8 years, however, replicating and evolving (Fig. 2b). Under such circumstances, one might expect HIV to accumulate other mutations that restore viral fitness to some degree, as suggested for the late-developing B27-KK10 escape mutation.[24] The H219X mutation, universally observed in association with T242N in B57/5801-positive subjects, may represent a similar compensatory mutation. If so, however, the fact that 22 B57/5801-negative subjects possessed H219X in the absence of T242N suggests that, unlike the compensatory changes associated with KK10 escape, H219X does not fully restore viral fitness. H219 lies within the cyclophilin A binding loop, which essential to HIV's life cycle.[35,36,37] H219 is directly involved in cyclophilin A binding, and the H219Q mutation results in a 4.8-fold decrease in binding affinity.[37] It would thus be unlikely for the H219Q mutation to be selected unless it was linked to some separate fitness advantage -- in this case, hypothetically, reducing the negative effect of T242N.

In these studies, we have illustrated two contrasting outcomes of CTL escape mutation. In the case of the T242N mutation, which we hypothesize is associated with a fitness cost to the virus, we observe reversion after transmission to B57/5801-negative recipients. Conversely, in the case of the G248A mutation, occurring within the same TW10 epitope, we observe maintenance after transmission to recipients lacking B57/5801. The inference that this CTL escape mutation arises at an insignificant fitness cost to the virus is supported by a study that examined this particular G248A mutation and found no effect on viral fitness.[34] Between these two extremes, there is likely to be a broad spectrum of CTL escape mutations whose presence may be associated with greater or lesser degrees of constraint on viral replication. Our findings illustrate that intrapatient evolution of HIV driven by CTL escape does not necessarily translate into evolution of HIV at the population level. The degree to which escape mutations revert or are maintained will determine the extent to which they contribute to the evolution of HIV as it spreads through a population.

In conclusion, we have identified two escape mutations in the immunodominant HLA-B57/5801-restricted Gag epitope TW10, one of which undergoes reversion and one of which is stable in the absence of B57/5801. This highlights the diverse outcomes of CTL escape mutation, and confirms the need to clarify the fate of escape mutations in order to determine their future relevance to vaccine design. These findings also suggest that the role of CTLs in shaping HIV evolution may not be a simple one, and that intrapatient viral evolution does not necessarily translate to evolution of HIV at the population level. Rather, it is the complex interaction between different selection forces -- positive selection pressure from CTLs on one hand, and purifying selection in the virus on the other -- that determines the role of CTLs in shaping HIV evolution.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.