The Evolution & Medicine Review

According to estimates by the World Health Organization, in 2013 on the order of 35 million people were infected with HIV worldwide (  Globally, about 1.5 million people are believed to have died from AIDS-related diseases in that year.  Substantial, although perhaps not insurmountable, obstacles to the development of a highly effective vaccine for HIV-1 have increased interest in curative strategies.  A key challenge to cure strategies is that infected people harbor a latent reservoir of infected CD4+ memory T cells that do not express significant amounts of viral proteins.  The paucity of viral proteins in these cells makes it more difficult to identify infected cells and eradicate them.  A new study (Deng et al., 2015) in Nature from Robert Siliciano’s lab at Johns Hopkins identifies an additional difficulty faced by one of the currently popular approaches to curative therapy but also, more optimistically, suggests a way to overcome this challenge.The authors note that methods for reactivating the HIV-1 provirus, i.e. the virus genome integrated into a host cell chromosome in latently infected cells, have been devised.  This development has generated interest in a cure strategy that would employ reactivation of latently activated cells followed by an immunological attack, a two-step approach commonly referred to as “shock and kill.”  So, it becomes crucial to know whether or not activation of immunological mechanisms, especially CD8+ cytotoxic T lymphocytes (CTLs) that are known to contribute to control of viral replication early in disease, can eliminate the cells with reactivated proviruses, only a minority of which are replication-competent.

Virus-specific CTLs recognize peptide fragments from virus proteins non-covalently associated with self-class I major histocompatibility complex (MHC) antigens.  CTLs can secrete cytokines, such as interferon-gamma (IFN-g), and kill cells infected with actively replicating virus displaying such peptide-MHC complexes on their cell surfaces.  Previous work has established that some of these CTL epitopes elicit stronger and some elicit weaker immune responses.  The former epitopes are referred to as dominant and the latter as sub-dominant.  With time, in any given individual infected HIV-1, viruses with escape mutations in dominant CTL epitopes begin to increase in frequency indicating that CTLs mediate strong selective pressure on the virus.  Thus, HIV-1 in a single infected host detectably evolves in response to both antibody- and CTL-mediated selection in a time frame as short as weeks to months.

The authors find that the mutational status of key portions of proviral genome sequences depends on how long after initial infection effective antiretroviral chemotherapy is initiated.  Individuals treated with antiretroviral therapy (ART) within three months of infection (acute phase, AP, treatment) do not have large numbers of proviruses with escape mutations in the portions of viral genes encoding dominant CTL epitopes.  In contrast, individuals not treated with ART within three months of infection (chronic phase, CP, treatment) do have large numbers of proviruses with escape mutations in the portions of viral genes encoding dominant CTL epitopes.

So, Deng et al. tested whether CTLs from infected patients could respond, as assessed by measuring secretion of IFN-g, to peptides corresponding to mutated dominant versus unmutated (dominant or sub-dominant) epitopes.  They found that the patient CTL responded much more strongly to the unmutated epitopes and only minimally to the mutated epitopes.

Next, the authors isolated replication-competent viruses from latent reservoirs from nine CP-treated patients.  They found that these viruses shared the same epitope variants as the defective proviruses in latently infected cells.  This result is important because it suggests that following any intervention that can elicit reactivation much of the replicating virus would contain escape mutations rendering CTL responses to wild-type (i.e. unmutated) dominant epitopes unlikely to be effective and controlling the virus population.

When CTL from 13 CP-treated patients were stimulated in vitro with a mixture of peptides from a key HIV-1 protein, Gag, these CTLs could kill infected autologous CD4+ T cells.  Further analysis revealed that only CTLs targeting non-dominant epitopes that had not mutated exhibited significant cytotoxic activity for the autologous CD4+ T cells.  CTLs targeting epitopes with known in vivo escape mutations exhibited negligible cytotoxicity.

Having demonstrated that CTLs to unmutated epitopes can eliminate cells infected with autologous virus in vitro, Deng et al. used partially humanized mice that can support human hematopoietic cell lineages to assess the ability of CTLs to unmutated sub-dominant versus mutated dominant epitopes to kill virus-infected cells and control virus replication in vivo.  Over a period of weeks, each mouse was given bone marrow from a given HIV-1-infected subject, later infected with virus from that same individual, and then administered CTLs, after in vitro antigen stimulation, from the bone marrow and virus donor.  The results of these rather complex experiments yielded results parallel to the in vitro results: only the CTLs specific for viral protein sequences that had not mutated in the patients could control virus replication in the mice.

The authors conclude that a broadly based CTL response that targets sub-dominant epitopes less likely to have suffered escape mutations could eliminate the latently infected cells in an HIV-1-infected subject following appropriate intervention to reactivate the latently infected cells.  They also suggest that more standard vaccination strategies for CTLs that activate primarily T-cell clones specific for unmutated dominant epitopes (that are likely to have suffered mutations in the infected individual and be represented in the latent virus reservoir) will likely fail to effectively eliminate latently infected CD4+ T cells.  In addition, Siliciano and colleagues suggest their results strongly support initiating ART as early as possible after infection by HIV-1 to prevent the latent reservoir from including proviruses with escape mutations in dominant CTL epitopes.

While the focus of Siliciano and his collaborators in this study is reasonably confined to immune responses by CTLs, there are additional options for designing immunologically-based cures for individuals infected by HIV-1.  Other cell types that might be able to direct cytotoxicity to (i.e., kill) CD4+ cells carrying reactivated latent proviruses include natural killer (NK) cells and T cells expressing gamma and delta T-cell antigen-specific receptors (TCRs) instead of alpha and beta TCRs (which are used by classical CD8+ CTLs).  These types of cells would probably require substantial ex vivo amplification to generate adequate numbers of cells for re-infusion, but pertinent techniques now exist for NK cells (Knorr et al., 2013) as well as T cells (Dallas et al., 2007).  It remains to be determined how NK and gamma-delta T cells might compare to CTLs in terms of overall efficacy in eliminating the cells harboring the latent HIV-1 reservoir.

Another possible strategy is to genetically engineer T cells or NK cells to enhance their abilities to recognize the infected cells of the latent reservoir and destroy them.  For example, methods exist for designing genes that encode novel antigen-recognizing receptors (generally referred to as chimeric antigen receptors, CAR), introducing such genes into the cells of interest, and adequately regulating the synthesis, cell surface display, and function of the receptors encoded by the designed genes (Gill and June, 2015).  In the next several years, it should become clearer if any of these curative strategies will be likely to succeed to a sufficient extent and frequently enough to justify their costs and complexities of implementation in the context of HIV-1 infection.


Global Health Observatory (GHO): HIV/AIDS. Global situation and trends. (last accessed on 1/31/15)

Deng K, Pertea M, Rongvaux A, Wang L, Durand CM, Ghiaur G, Lai J, McHugh HL, Hao H, Zhang H, Margolick JB, Gurer C, Murphy AJ, Valenzuela DM, Yancopoulos GD, Deeks SG, Strowig T, Kumar P, Siliciano JD, Salzberg SL, Flavell RA, Shan L, Siliciano RF. Broad CTL response is required to clear latent HIV-1 due to dominance of escape mutations. Nature. 2015 Jan 15;517(7534):381-5. doi: 10.1038/nature14053. Epub 2015 Jan 7. PubMed PMID: 25561180.

Knorr DA, Ni Z, Hermanson D, Hexum MK, Bendzick L, Cooper LJ, Lee DA, Kaufman  DS. Clinical-scale derivation of natural killer cells from human pluripotent stem cells for cancer therapy. Stem Cells Transl Med. 2013 Apr;2(4):274-83. doi: 10.5966/sctm.2012-0084. Epub 2013 Mar 20. PubMed PMID: 23515118; PubMed Central PMCID: PMC3659832.

Dallas MH, Varnum-Finney B, Martin PJ, Bernstein ID. Enhanced T-cell reconstitution by hematopoietic progenitors expanded ex vivo using the Notch ligand Delta1. Blood. 2007 Apr 15;109(8):3579-87. Epub 2007 Jan 9. PubMed PMID: 17213287; PubMed Central PMCID: PMC1852253.

Gill S, June CH. Going viral: chimeric antigen receptor T-cell therapy for hematological malignancies. Immunol Rev. 2015 Jan;263(1):68-89. doi: 10.1111/imr.12243. PubMed PMID: 25510272.