News & Views | Published:


Tired T cells turn around

HIV-1 prompts a massive cellular immune response, but eventually it tires the immune cells. Blocking the activation of a cell receptor called PD-1 might restore these exhausted cells.

There are only the pursued, the pursuing, the busy, and the tired.

F. Scott Fitzgerald, The Great Gatsby.

When confronted with a pathogen, the immune system must respond with appropriate force, sufficient to clear the pathogen but not so vigorous as to damage the host through excessive inflammation. So, the effector arms of the immune system — the cells and molecules that carry out the immune response — are carefully regulated to achieve a balance between pathogen clearance and immunopathology. Failure to achieve this balance can lead to persistent infection, often accompanied by chronic inflammation and disease, as seen in human infection with hepatitis viruses B and C and the human immunodeficiency virus HIV-1. Moreover, the mechanisms that control the balance between effective response and immune-mediated damage to the host can be subverted by pathogens — through millennia of coevolution, pathogens have acquired substantially more profound insights into the complexity of host immune responses than human immunologists have yet achieved.

Maintaining the balance of the immune response is especially critical in regulating T cells, which secrete inflammatory molecules (cytokines) and kill infected cells. In chronic infection, the responding T cells often show features of what immunologists term 'exhaustion', with varying degrees of impairment of their effector functions. A paper in this issue (page 350)1 by Day et al. and one by Trautman et al. just published on the Nature Medicine website2 show how an essential immune mechanism that regulates the response of T cells is undermined in HIV-1 infection.

One of the unexplained paradoxes in HIV infection has been why a huge cellular immune response — in which as many as one in five circulating killer T cells are responding to HIV antigens — fails to control or clear the virus. Nevertheless, most immunologists believe that virus-specific killer T cells are a key component of immunity to HIV-1, particularly in the early stages of infection, and many of the HIV vaccine candidates currently in the early stages of clinical trials have been designed to stimulate a killer-T-cell response. Can the paradox be explained by a progressive loss of function of HIV-specific killer T cells (as suggested in some studies)3,4,5 — and, more importantly, can this process be reversed?

The mouse model of chronic infection with lymphocytic choriomeningitis virus (LCMV) has many parallels with human HIV-1 infection. Earlier this year, Barber et al.6 used this model to show that exhausted killer T cells are characterized by expression of an inhibitory molecule called programmed death-1 (PD-1; Fig. 1a). This cell-surface molecule is part of the CD28 family and is thought to be a regulator of the T-cell response to invading pathogens, interacting with its ligands (PD-L1 and PD-L2) to inhibit activation of T cells once the T-cell surface receptors have recognized the virus.

Figure 1: Immune exhaustion and PD-1.

a, Although immune cells called T cells initially respond to an HIV infection by killing the infected cells and secreting inflammatory molecules, they seem to gradually give up the fight against the virus. In a mouse model of the disease, these 'exhausted' T cells are characterized by expression of the PD-1 receptor on their cell surface. b, Day et al.1 and Trautman et al.2 studied the role of PD-1 in cultured cells from HIV-1-infected patients. Day et al. show that blocking the interaction between PD-1 and its ligand (PD-L1) seems to restore certain of the T-cell responses, at least at the level of the bulk culture.

This mechanism probably has a physiological role in limiting the expansion in the numbers of T cells responding to an acute infection. But in the dysregulated setting of chronic infection, PD-1 expression on virus-specific T cells seems to be a signal that the T cells are giving up the fight and can no longer perform vital functions to control the infection. Remarkably, in the LCMV-infected mice, administration of an antibody that blocked the interaction of PD-1 with its ligands allowed these exhausted T cells to regain the ability to secrete cytokines, to kill other cells and to proliferate, leading to control of the virus. Moreover, this strategy was effective even in animals lacking 'T-cell help', the essential supportive functions provided by T cells expressing the CD4 molecule that makes them a prime target for HIV infection. This raises the possibility that such an antibody regime might be relevant in addressing HIV-1 infection, as a cardinal feature of the infection is a lack of helper T cells.

Following on from this important finding, Day et al.1 studied a cohort of untreated HIV-1-infected patients in South Africa. They now report that PD-1 is significantly upregulated on killer T cells in the blood that recognize HIV-1 antigens. Both the proportion of killer T cells expressing PD-1 and the levels of PD-1 on the cell surface showed a strong correlation with the viral load in the blood plasma — currently the best indicator of disease progression — and decreased when the viral load was controlled by anti-retroviral therapy. More strikingly, they found that by blocking the interaction between PD-1 and its ligands, several effector functions of killer T cells, notably proliferation and the ability to secrete antiviral cytokines, could be restored in vitro (at least in bulk cell culture, if not on an individual cell basis; Fig. 1b).

Similar studies by Trautmann and colleagues2 show restoration of a range of effector molecules in the presence of an antibody that blocks the PD-1 ligands. These experiments confirm that functional exhaustion is a feature of virus-specific T cells that are responding to HIV-1 infection, and show that impaired T-cell function is magnified by increasing exposure to replicating virus.

Both reports1,2 show that PD-1 expression levels vary in different T-cell populations, even within the same individual. This raises the possibility that some responding T cells can maintain function better than others, as described in long-term HIV-infected patients who do not develop disease or any signs of immune suppression7.

The exciting conclusion from both studies is that, despite the damage to T cells that results from their encounter with HIV-1, the cells are not terminally incapacitated and can potentially be revived. The work will undoubtedly prompt therapeutic strategies to restore responding T cells to full health and function, not only in HIV-1 infection, but in other settings where T-cell exhaustion is believed to play a part in disease, such as chronic infection with hepatitis B and C viruses. Similar approaches are being explored in cancer immunotherapy8, because upregulation of the PD-1 ligands is a feature of many tumours and is associated with a poor prognosis.

These strategies will need to be approached with caution, as studies in mice suggest that blocking the interactions between PD-1 and its ligands can exacerbate autoimmune disease. Intriguingly, several human autoimmune conditions are associated with a mutation (a single nucleotide polymorphism) in the regulatory region of the PD-1 gene9. T cells from donors with this genetic variant seem to be less susceptible to PD-1-mediated suppression of T-cell function9, so could this mutation provide an advantage to a host infected with a persistent virus or faced with a tumour that overexpresses a PD-1 ligand? Just as PD-1 ligand expression seems to be a manoeuvre on the part of tumours to avoid recognition by T cells, several viruses have developed mechanisms to upregulate PD-1 ligand in the tissues. HIV-1 is already known to possess an extraordinarily diverse repertoire of immune evasion strategies: will the stimulation of PD-1 ligand expression in T cells prove to be a further example of viral subterfuge?


  1. 1

    Day, C. L. et al. Nature 443, 350–354 (2006).

  2. 2

    Trautman, L. et al. Nature Med. 10.1038/nm1482(2006).

  3. 3

    Appay, V. et al. J. Exp. Med. 192, 63–75 (2000).

  4. 4

    Shankar, P. et al. Blood 96, 3094–3101 (2000).

  5. 5

    Kostense, S. et al. Eur. J. Immunol. 31, 677–686 (2001).

  6. 6

    Barber, D. L. et al. Nature 439, 682–687 (2006).

  7. 7

    Migueles, S. A. et al. Nature Immunol. 3, 1061–1068 (2002).

  8. 8

    Geng, H. et al. Int. J. Cancer 118, 2657–2664 (2006).

  9. 9

    Kroner, A. et al. Ann. Neurol. 58, 50–57 (2005).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

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.