The Role of Chemokines in Establishing HIV Latency
This report is part of a series of focused summaries from the “5th International Workshop on HIV Persistence, Reservoirs & Eradication Strategies” held in St Maarten, December 6-9, 2011. This presentation was given by Professor Sharon Lewin, Director, Infectious Disease Unit, Alfred Hospital, Professor, Department of Medicine, Monash University, Co-head, Centre for Virology, Burnet Institute, Melbourne, Australia.
Ongoing infection of resting T-cells in patients on cART may potentially replenish the pool of latently infected cells. Consequently, identification of host targets may lead to new ways to block establishment of latency.
Resting CD4+ T cells are relatively resistant to HIV infection (1, 2). Infection of these cells ex vivo leads to entry and reverse transcription but rarely integration.
However, in ex vivo tissue blocks, resting T-cells can be infected and integration occurs. Also, stable integrated virus is clearly detected at low frequency in blood and at high frequency in tissue (3, 4).
During her talk, Professor Lewin addressed 4 main questions:
Is it really established in vitro that chemokines induce latency?
The authors have previously demonstrated that chemokines play an important role in establishing post integration latency. Treatment of resting CD4+ T-cells with chemokines led to high level of HIV-1 integration with minimal productive infection (5, 6).
Multiple chemokines can establish latency in resting CD4+ T-cells and, in particular, CCL19 and CCL21. The absence of productive infection was further confirmed by staining for intracellular p24 expression where CCL19-treated infected CD4+ T-cells resulted in <1% p24-positive cells in contrast to IL-2/PHA activated infected CD4+ T-cells (~6-9%). Finally, following infection with WT NL4.3 of PHA/IL-2 treated, CCL19-treated cells, the fold change in US RNA at day 4 compared to day 0 was 21.2 and 1.1 respectively, demonstrating a significant block in synthesis of US RNA in CCL19 treated cells. The increase in MS-RNA was 56,000, 5,000 and <200 copies/million cells in IL-2/PHA activated, CCL19-treated and unactivated CD4+ T-cells respectively .
How do chemokines facilitate efficient integration in resting T-cells?
Chemokines activate multiple signalling pathways. The PI3K pathway is critical for integration.
Duverger and colleagues (7) showed that in cell lines, the generation of latent or productive infection was dependent on the relative amount of NF-kB. When present at high levels, NFKB promoted HIV transcription and productive infection but at basal levels NFKB was required for HIV integration. In addition, the sites of integration of other retroviruses has been shown to be dependent on the concentration of local transcription factors (8).
Recent data has also shown that JNK phosphorylates HIV integrase via PIN-1 leading to increased stability of integrase allowing for integration in activated CD4+ T-cells. It is possible this same pathway is also important for integration in resting cells (9).
These signalling pathways are summarized in Figure 1:
What role do dendritic cells (DCs) play in infection of resting T-cells?
In the absence of stimulation the authors observed low levels of latently infected cells when the resting CD4+ T cells were cultured alone.
However when the resting CD4+ T cells were co-cultured with DCs they observed a significant increase in the number of latently infected cells.
Surprisingly, a similar but enhanced effect was observed in the presence of SEB, which may be due to an increase in the secretion of soluble factors by the DCs or an increase in DC-T cell contact following SEB stimulation.
In order to confirm that the sorted SNARFhi EGFP- cells were indeed latently and not productively infected resting CD4+ T cells that had not yet produced EGFP, in some experiments they also cultured the sorted cells in the absence of any stimulus for 24 hours
Under these conditions they did not detect EGFP+ cells in any of the cultures, which is consistent with amplification of virus from latently infected and not from productively infected cells.
Blood DCs can be broadly grouped into 2 subsets: plasmacytoid (pDC) and myeloid (mDC).
In order to determine which DC subset was inducing latency the authors used either purified pDCs or mDCs in place of bulk DCs in their in vitro model.
Both in the absence and presence of SEB, latently infected cells were only identified in the non-proliferating CD4+ T cells following co-culture with mDCs but not pDCs where the number of latently infected cells was similar to the CD4+ T cells cultured alone.
These results indicated that at least within their model mDC not pDC induce HIV-1 latency in non-proliferating CD4+ T cells.
The authors next asked whether DC-T cell contact was required to establish DC-induced latency.
To answer this question they cultured the DCs within a transwell above the resting CD4+ T cells for 24h, at which time they then exposed both the DCs and CD4+ T cells to HIV.
When DC-T cell contact was prevented, there was a decrease in the number of latently infected CD4+ T cells compared to the co-cultures without membranes.
However, the number of latently infected cells was still significantly greater than that observed when the CD4+ T cells were cultured alone indicating that soluble factors play a role in DC induced latency.
The incomplete reduction in latently infected cells could then be due to either an additional need for DC-T cell contact or a decreased concentration of soluble factors brought about by the increased distance the transwell system creates between the CD4+ T cells and DCs.
How can these models be used to identify novel cellular factors that maintain latency?
The authors studied what candidate genes are important to DC-induced latency.
They identified 461 genes that were differentially expressed in latently infected, non-proliferating CD4 T cells when compared to mock infected T cells (Figure 2, 3).
Control of cell proliferation and cycle were the most significantly enriched biological functions, of which the majority (70%) were down-regulated, including numerous genes critical to active cell cycle.
They also observed an up-regulation of multiple genes known to arrest cell cycle including IL-15, which arrests cell cycle before the G0/G1 phase transition and can also protect against resting cell apoptosis, and the histone linker H1F0 that also delays cells in G0 and has been correlated with cellular quiescence.
Additionally they observed the suppression of multiple genes associated with the activation of NF-B.
These included genes such as:
EDAR and CD27 – that have a role in inducing apoptosis:
-Down-regulation of apoptotic genes in U61 latently infected cell line
-Survival of latently infected cells?
PRKCA – activation of NF-kB and NFAT:
-NFAT signalling – reactivation of latency
-Inhibition of NFAT – inhibition of HIV gene expression
When they explored differentially expressed genes that encode for cellular proteins with known interactions with HIV-1, they identified the up-regulation of 4 genes known to inhibit HIV-1 replication at various stages post-integration. All these genes encode for proteins with known interactions with Tat and included:
TXNRD1 - negative regulator of Tat-dependent transcription
IL1RN - blocks IL-1 mediated reactivation of HIV-1 latency in the chronically infected U1 cell line, inhibits HIV-1 replication in PBMC
IRF7 - suppresses transcriptional activation, associated with Epstein-Barr virus latency
TRIM22 - down-regulates transcription from the HIV-1 LTR, possibly through inhibition of NFAT
Together, these genes may contribute to latency by:
 Maintaining cellular quiescence;
 Repressing viral replication.
In conclusion, such in vitro models of direct infection of resting cells leading to “pre-activation” latency show that:
- Activation of PI3K (Jnk and NF-kB) is critical for integration in resting CD4+ T-cells;
- Myeloid DCs mediate HIV latency in resting memory CD4+ T-cells, involving soluble factors and cell-cell interaction;
- These in vitro models can be used to screen for novel genes to block the establishment and/or maintenance of latency.
1- Zack et al. Cell 1990; 61(2): 213-22
2- Swiggard et al. J Virol 2005; 79(22): 14179-88
3- Eckstein et al. Immunity 2001; 15: 671
4- Kreisberg et al. J Exp Med 2006; 203:865
5- Saleh et al. Blood 2007; 110:416
6- Cameron et al. Proc Natl Acad Sci 2010; 107(39): 16934-9
7- Duverger et al. J Virol 2009; 83(7): 3078-93
8- Felice et al. Plos One 2009; 4(2): e4571
9- Managanaro et al. Nat Med 2010; 16(3): 329-33
Key words: HIV chemokines, HIV dendritic cells, HIV latency, HIV persistence, HIV reservoirs, Sharon Lewin