HIV reservoir: CD4 T stem cells may facilitate long-term persistence

A report in Nature Medicine is important for suggesting that a recently discovered and little understood subset of CD4 T cells are a long-term hiding place for the HIV reservoir.

Maria Buzon and colleagues at the Massachusetts General Hospital and Massachusetts Institute of Technology, in Boston, describe their analysis of the HIV reservoir in different T cell subsets before and after several years of ART [1].

T cells are divided into different maturational subsets according to their exposure to antigen.

These are:

  • Naive T cells (TNA), which have never encountered antigen;
  • Central memory T cells (TCM), which have encountered antigen and undergone clonal expansion, but which are in a resting non-activated state;
  • Effector memory T cells (TEM), which have encountered antigen and are activated, providing effector functions such as cytokine release; and
  • Terminally differentiated T cells (TTD), which are memory cells that have reached the end of their capacity to divide.

Over the last 5 years or so, evidence has accumulated that there is an additional maturational subset of T cells with stem cell-like properties, called stem central memory T cells (TSCM). These cells may form the first and most long-lived developmental stage of memory T cells as they can subsequently differentiate into either TCM or TEM cells. TSCM are highly resistant to cell death, they live for extremely long periods and they maintain their pool size through proliferation and self-renewal.

Considering these characteristics, Buzon and colleagues hypothesised that TSCM cells might represent a nook that harbours a significant proportion of the long-term persistent HIV reservoir. In order to investigate this, the researchers measured the abundance of HIV DNA in each of the different T cell subsets over time, and conducted detailed comparisons of the genetic sequences of this DNA, and how they related to plasma viruses before and after ART.

Before assessing the contribution of TSCM cells to the HIVreservoir, they first determined that TSCM cells from uninfected donors could be infected with lab-adapted strains of HIV. The observation that TSCM cells can be infected with HIV, has also been made recently by others [2]. Buzon et al found that HIV-1 RNA can be found in TSCM cells from HIV-infected donors who were not on ART, demonstrating that HIV infection of TSCM cells occurs in vivo.

The investigators went on to assess the contribution of TSCM to the HIV reservoir in HIV-infected subjects who had been receiving ART for a median of 7 years. The amounts of HIV DNA present per cell were significantly higher in these donors TSCM cells than in any of the other T cell subsets: TNA, TCM, TEM or TTD cells. Despite this, the overall contribution of TSCM cells to the HIV reservoir size was calculated at only 8%. Most of the contribution to the reservoir was found in TCM and TEM cells. This likely reflects the very low frequency of TSCM cells in the overall T cell pool. While the numerical contribution of TSCM to the overall reservoir may be small, this does not have bearing on the potential contribution of these cells to reservoir persistence.  Indeed, the researchers found that even when, in some individuals, the reservoir size in TCM and TEM cells was relatively limited, its size in TSCM cells was sustained at generally higher levels, increasing the contribution of TSCM cells to the overall reservoir in these subjects. The researchers propose that this suggests TSCM cells are “a not necessarily large but very stable and durable component of the CD4+ T cell reservoir”.

Replication-competent virus could be recovered from TSCM cells from all HIV-infected donors whose cells were used for viral outgrowth assays. These donors had been on suppressive ART for a median of 28 months. The decay rates of HIV DNA were also found to be far more stable in TSCM cells than other T cell subsets over several years of ART. HIV DNA levels were assessed in 8 subjects who had initiated ART during primary HIV infection. Samples were collected for HIV DNA assessment at a median of 1 year after initiation of ART and a second time point of 9 years after ART. Rates of HIV DNA decline in TCM and TNA cells were slightly faster than TSCM cells, while rates in TEM and TTD declined significantly faster. Furthermore the contribution of infected TSCM cells to the reservoir significantly increased from 14% at the one year post-ART time point to 24% at the nine year time point, despite the fact that these cells did not change as a proportion of the overall T cell pool. In contrast the contribution of TEM to the reservoir significantly decreased from 35% to 26% during these two time points, despite the fact that TEM cells significantly increased as a proportion of the T cell pool, from 31% to 39%. This further suggests that the HIV reservoir in TSCM is stable and outlasts the reservoir in other T cell subsets.

Lastly, in three patients who had initiated ART during chronic infection, a number of proviral Env sequences were observed in both TCM and TSCM cells at the beginning of ART that were still present after 4-8 years of ART. No proviral sequences identified at the beginning of ART in TNA or TTD cells were observed again at later time points. Analysis of diversity in genetic sequences of both provirus and plasma virus found that there was a close relationship between proviral DNA sequences in TSCM and TCM after 6 – 12 years of ART and plasma virus sequences from early in infection. This suggests that HIV that infects TSCM early in the course of infection is retained in these cells for very long periods of time.

While this data is based on small numbers of patients, it supports the hypothesis that the TSCM cell population is a stable and durable niche for the HIV reservoir. Furthermore, TSCM may represent an increasing relative proportion of the HIV reservoir over years of ART, as the reservoir in other T cell subsets gradually declines.


  1. Buzon MJ et al. HIV-1 persistence in CD4+ T cells with stem cell-like properties. Nat Med, Feb 2014, Volume 20, Issue 2. Page 139-42


  1. Tabler CO et al. CD4+ Memory Stem Cells (TSCM) are Infected by HIV-1 in a Manner Regulated in Part by SAMHD1 Expression. J Virol, 19 Feb 2014. Epub ahead of print.


First published in HIV Treatment Bulletin Mar/Apr 2014 (

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.