Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Finding latent needles in a haystack


Antiretroviral therapy can keep HIV at bay, but a few cells remain infected, so the disease cannot be cured. The discovery of a protein that marks out these infected cells will facilitate crucial studies of this latent viral reservoir. See Letter p.564

The development of combination antiretroviral therapy, which suppresses the replication of HIV in immune-system cells called CD4 T lymphocytes, is a major achievement of modern medicine1. However, a small proportion of these CD4 T cells retains the virus despite ongoing therapy, forming a latent HIV reservoir that can reactivate if treatment is stopped. Ways of eradicating this reservoir are an intense focus of research, but efforts have been hampered by the difficulty of analysing latently infected cells — of the order of one in a million CD4 T cells in an individual receiving antiretroviral therapy is latently infected, and markers for these cells have been lacking2,3. On page 564, Descours et al.4 have identified just such a marker for about half of the latently infected CD4 T cells in the blood.

During its replication cycle, HIV generates, through reverse transcription, a DNA version of its RNA genome, which integrates into a host-cell chromosome. Replication of the HIV genome in CD4 T cells produces viruses that destroy the host cells. The rare cells that survive infection constitute the latent reservoir. HIV DNA remains integrated in their genomes, but might not produce viral RNA or protein.

To analyse this reservoir, the investigators developed an in vitro model of HIV latency using CD4 T cells infected with an HIV-derived genetic construct that produced green fluorescent protein (GFP) as a marker of infection. An important point to note when considering this study is that the cells used were in a resting state — they were not activated or proliferating as are most CD4 T cells propagating high levels of HIV. An unresolved controversy has been whether latency results from infection of resting cells or (in line with the predominant view) is a rare consequence of a cell surviving active HIV replication, and then converting to a resting state5. The current study should prompt more investigation to address this issue.

Descours et al. screened the infected, GFP-expressing cells using ultradeep RNA sequencing to search for host gene transcripts that were increased in this population compared with uninfected cells. They identified 103 differentially expressed genes, including 16 that encode transmembrane proteins, which make cell markers that are readily amenable to identification and cell-sorting techniques. Of these, the most highly expressed was FCGR2A, which encodes the protein CD32a.

CD32a is a low-affinity receptor for a fragment, Fc, of immunoglobulin G (IgG) antibodies, and, like other Fc receptors, is expressed on the surface of dendritic cells and macrophages6. These immune cells internalize Fc receptors, along with their bound IgGs and the specific antigen molecules that associate with the IgGs, and then present the processed antigens to lymphocytes. This triggers an immune response specifically against cells expressing that antigen6. Descours and colleagues showed that CD32a is also expressed on about 50% of latently infected CD4 T cells — but not on uninfected T cells or those with an active HIV infection. CD32a can be bound by a commercially available antibody, enabling the authors to separate CD32a-expressing CD4 T cells from other cells.

“This highly selective marker for latently infected T cells could at last allow researchers to investigate the mysterious mechanisms of latency.”

This highly selective marker for latently infected CD4 T cells could at last allow researchers to investigate the mysterious mechanisms of latency, without needing to find the one cell of interest in a million (Fig. 1). The marker might also be helpful in analysing the success of candidate drugs that aim to target this reservoir. Isolation of CD32a-expressing CD4 T cells would concentrate the population of latently infected cells by about 1,000 times. But although this process would remove the vast majority of irrelevant cells, a large amount of blood would still be needed to obtain enough latently infected cells for analysis — 100 millilitres or more, depending on the type of investigation.

Figure 1: Hunting down an elusive cell population.

A person chronically infected with HIV and receiving antiretroviral therapy has about 200 billion immune cells called CD4 T cells, of which of the order of one in a million remains infected with a latent, replication-competent form of HIV despite treatment2,3. About 2% of the body's CD4 T cells (about 4 billion) are present in the approximately 5 litres of blood in an adult — 100 millilitres, as might be extracted in a blood sample in human studies, therefore contains roughly 80 million CD4 T cells, of which only about 80 are latent cells. Descours et al.4 identified the protein CD32a as a marker of about 50% of all latently infected CD4 T cells. By using a sorting technique to separate cells expressing the protein from other cells, they can selectively concentrate the latent population by more than 1,000-fold to between 2,000 and 16,000 in a million CD4 T cells, making many new experimental studies of this population feasible.

As with any breakthrough, new questions arise and new experiments become feasible. First, Descours and colleagues raise an interesting question: can CD32a be used to selectively target the rare, latently infected cells? This might be a basis for strategies aimed at eradicating the latent reservoir. One problem, however, is that CD32a is a marker for only 50% of the reservoir, whereas the eradication of latent HIV would require a much greater reduction in the number of latently infected cells in the body. Moreover, targeting CD32a would also make the antigen-presenting cells that normally express CD32a vulnerable to destruction, which might well cause unwanted or harmful side effects.

Second, the authors studied CD4 lymphocytes from the blood, but these circulating cells account for 2%, at most, of the CD4 T cells in the body2. It remains to be seen whether CD32a is as good a marker for latently infected cells in the lymph nodes, bone marrow, gut and other tissues. Perhaps more markers could be identified from the 103 differentially expressed genes found in the researchers' screen — analysis of these proteins in combination with CD32a might increase the total proportion of identifiable latent cells.

Third, does the presence of CD32a vary between latently infected CD4 T-cell subsets? CD4 T cells differentiate into a complex variety of specialized subsets, some of which may not express CD32a when latently infected. It would also be interesting to characterize the mechanism by which latent HIV — thought not to be transcribed — can increase transcription of FCGR2A and other differentially expressed genes. Such an insight might help to reveal how HIV latency is maintained or reversed. Another question that merits careful investigation is whether CD32a is ever expressed on lymphocytes other than those latently infected with HIV.

Finally, the antibody used by the authors does not react with the CD32a protein found in rhesus macaques, which are used as a model of latency with simian immunodeficiency virus. An alternative antibody must therefore be produced if this finding is to be exploited for studies in monkeys.

A good understanding of HIV latency and interventions to manage it would provide remarkable benefits, but the field has so far progressed too slowly to capitalize on this. This potentially seminal study promises to facilitate the study of HIV latency and accelerate the generation of insights and therapeutic approaches. Footnote 1


  1. 1.

    See all news & views


  1. 1

  2. 2

    Haase, A. T. Annu. Rev. Immunol. 17, 625–656 (1999).

    CAS  Article  Google Scholar 

  3. 3

    Richman, D. D. et al. Science 323, 1304–1307 (2009).

    CAS  Article  ADS  Google Scholar 

  4. 4

    Descours, B. et al. Nature 543, 564–567 (2017).

    CAS  Article  ADS  Google Scholar 

  5. 5

    Han, Y. et al. Nature Rev. Microbiol. 5, 95–106 (2007).

    CAS  Article  Google Scholar 

  6. 6

    Guilliams, M., Bruhns, P., Saeys, Y., Hammad, H. & Lambrecht, B. N. Nature Rev. Immunol. 14, 94–108 (2014).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Douglas D. Richman.

Related audio

Related links

Related links

Related links in Nature Research

HIV: Cure by killing

HIV: Seeking ultimate victory

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Richman, D. Finding latent needles in a haystack. Nature 543, 499–500 (2017).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing