| |
Blood and Guts and HIV: Preferential HIV Persistence in GI Mucosa
EDITORIAL COMMENTARY
|
| |
| |
The Journal of Infectious Diseases Feb 2008;197
Steven Yukl and Joseph K. Wong
Veterans Affairs Medical Center San Francisco and University of California San Francisco
"The disproportionate concentration of HIV infected cells in GALT supports the premise that cryptic HIV replication in GALT continues despite the receipt of combination ART......The effect of treatment intensification or novel treatment strategies on the reservoir of HIV+ cells in GALT should be explored. Newly available compounds, such as integrase inhibitors and inhibitors of viral entry, may add to antiviral potency and provide as yet undefined pharmacodynamic advantages due to their action on distinct parts of the viral life cycle (such as integration) or, in the case of entry inhibitors, an extracellular site of action....infection in the GI mucosa establishes a cycle of transmigration of proinflammatory substances that sustain local and systemic T cell activation and increase permissiveness for HIV infection that, in turn, promotes continued mucosal disruption. If so, reducing residual replication in GALT could produce results that extend beyond local effects, such as systemic CD4 cell recovery and reduced reactivation of latently-infected cells"
Twelve years after the introduction of potent combination antiretroviral therapy (ART) for HIV infection, it remains unclear whether suppressive ART that drives plasma viral RNA below 50 copies/mL also results in complete inhibition of HIV replication. The issue is not merely of academic interest; resolving this controversy could help guide future strategies aimed at eradicating HIV and inform novel management approaches such as induction-maintenance. In this issue of the Journal, Chun and colleagues [1] uncover new evidence for persistent, cryptic HIV replication by showing that cells in gut-associated lymphoid tissues (GALT) are disproportionately infected with HIV, despite the fact that the study patients had been receiving continuous suppressive therapy for up to 10 years. These findings raise new questions, such as: (1) how low-level replication in the gut, if it occurs, might qualitatively differ from replication at other sites; (2) why replication persists despite potent antiretroviral therapy; (3) how therapies could better inhibit HIV replication in the gut; and (4) whether further reductions of HIV replication in the gut will translate into new clinical benefits.
Within weeks of starting ART, the plasma viral RNA level often drops to <50 copies/mL, the level of detection of most commercial assays. With the use of more sensitive assays, it is possible to show that HIV RNA often persists at low levels in the peripheral blood [2-5]. Although some studies show a slow decline of plasma HIV RNA levels over several years, most studies suggest that the plasma HIV RNA level reaches an equilibrium level after 1-2 years of treatment [5, 6]. It is unclear whether this residual viral expression in patients who receive ART reflects the incomplete suppression of ongoing viral replication, intermittent production from stable reservoirs of chronically infected cells, or viral expression from self-limited reactivation of latently infected cells. HIV DNA is readily detectable in peripheral blood mononuclear cells (PBMCs), a subset of which constitutes a reservoir of latently infected, resting memory CD4 lymphocytes whose stability has been attributed to varying degrees of replenishment by ongoing replication [7-12]. These latently infected cells likely represent a barrier to eradication that is distinct from incomplete inhibition of viral replication.
There is evidence both for and against the presence of persistent viral replication in patients who receive long-term suppressive ART. Unless the residual HIV comes from sanctuary sites with extremely limited drug penetration, any level of residual replication should be accompanied by viral evolution and selection for increasingly resistant viral variants, as has been reported for patients with intermittent viremia at higher levels of 50 to several hundred copies/mL [13, 14]. Such evidence of drug resistance and evolution is typically not seen in the residual RNA and DNA of patients with prolonged viral loads <50 copies/mL who receive ART [15, 16]. However, in support of ongoing viral replication, if patients who receive long-term ART and who have prolonged viral loads <50 copies/mL (but detectable at stable levels by more sensitive assays) are given 1 or more new drugs (i.e., intensification), there is a new, further reduction in viral load [6]. Also, indirect evidence comes from the many patients who continue to have heightened levels of immune activation despite having received suppressive ART for years [17].
It has long been appreciated that lymphoid tissues are the principal sites of HIV replication and production [18, 19]. The GALT is quantitatively the largest collection of lymphoid tissues in the body and collectively represents the single largest source of HIV in the absence of ART [20]. During the acute phase of infection, the gastrointestinal (GI) tract has been found to be particularly permissive of replication for both HIV and simian immunodeficiency virus (SIV) [21-24]. The vulnerability of T cells at this site may be the result of high levels of CCR5 expression on terminally-differentiated effector memory T cells, as well as the result of conditions that facilitate T cell activation [25]. Earlier studies that sampled both central lymphoid tissue and GALT had shown that potent suppressive antiretrovirals not only reduced plasma viremia to undetectable levels but also greatly reduced HIV levels in lymphoid tissues [26-30]. These studies relied on in situ hybridization and polymerase chain reaction-based detection from bulk nucleic acid extracts. Whereas some studies were unable to identify much residual virus in tissues from treated patients, others suggested that at least a subset of patients continue to have moderately high levels of detectable virus and persistent abnormalities in mucosal function, integrity, and cellularity, despite receipt of apparently effective ART [21, 31-32].
To investigate the GI mucosa as a potential sanctuary site for HIV replication in patients who had been receiving ART, Chun and colleagues [1] performed a cross-sectional study of 8 patients who had been treated for up to 9.9 years with combination ART. In contrast with earlier studies that used DNA extracted in bulk from mucosal biopsy samples or total cellular extracts, they isolated mononuclear cells from GI mucosal biopsy samples and then enriched for CD4 cells using CD8-depletion. They found a gradient of HIV DNA content in which the lowest levels were present in resting CD4 cells from blood, intermediate levels were present in activated CD4 cells from the blood, and the highest levels of HIV DNA were present in CD8-depleted cells from GALT. It should be noted that CD8-depleted cells from GALT had the highest HIV DNA content, even though these cells were not purified CD4 cells and were not segregated for markers of cellular activation. The observed differences are, therefore, likely to underestimate the differences that would be encountered if they had compared only activated CD4 cells from both blood and GALT. From these data, Chun et al. [1] inferred that there was persistent viral replication in CD4+ lymphocytes in the GALT. An alternative explanation might be that there was division of cells with integrated proviral DNA but no active HIV replication. However, cells that contain HIV DNA should not divide more than those without HIV DNA, so such a mechanism should not result in preferential enrichment of either cell type. The increased concentration of HIV DNA-bearing cells in GALT is best explained by amplification from local viral replication. The fact that these aviremic individuals showed incomplete recovery of GALT CD4 cells, even when the peripheral blood CD4 counts were normal or near normal (in half of the cases), also argues for ongoing viral replication in the gut.
Having found evidence for local viral replication, Chun and colleagues [1] next looked for differences between viral sequences in GALT and blood by analysis of the viral env gene. They did not find evidence of distinct viral variants in GALT, but instead found evidence for "cross-infection" between blood and GALT, both by phylogenetic analysis and by measures of viral diversity. Because no longitudinal samples were included in this study, it was not possible to address viral sequence evolution. However, because earlier studies showed a lack of sequence evolution in HIV DNA and RNA in the blood, the similarity of blood and GALT sequences suggests that there is likely to be a similar lack of evolution of HIV in the GALT.
The conclusion that viral replication is ongoing, yet most often not accompanied by demonstrable sequence evolution or selection for drug resistance, requires explanation. If antiretroviral concentrations at GI mucosal sites were exceedingly low, one might propose that such "sanctuary replication" would fail to select for drug resistance. However, notwithstanding moderate differences in levels of some drugs, it seems unlikely that drug levels in GI mucosa would be low enough not to select for resistance, yet high enough to sizably and durably reduce viral load [29, 30]. Alternatively, viral replication at low levels may differ qualitatively from replication at higher levels. Grossman and colleagues [33] have proposed that HIV replication occurs in lymphoid tissues in waves or bursts that are temporally and spatially discrete. At the very low levels attained during ART, this discontinuity of replication may significantly constrain the viral evolutionary machinery that makes HIV adaptation so efficient at higher levels of replication [34].
The disproportionate concentration of HIV infected cells in GALT supports the premise that cryptic HIV replication in GALT continues despite the receipt of combination ART. Future larger-scale studies are needed to confirm and extend these findings, perhaps with the addition of tests to assess drug levels in the GALT, HIV RNA levels and sequences, the presence of replication-competent virus, immunophenotypes of infected cells, the numbers of latently-infected cells, and correlation with clinical parameters. Longitudinal studies would help to address whether the concentration of infected cells continues to decline with extended treatment and would permit assessment of viral evolution at this site. The effect of treatment intensification or novel treatment strategies on the reservoir of HIV+ cells in GALT should be explored. Newly available compounds, such as integrase inhibitors and inhibitors of viral entry, may add to antiviral potency and provide as yet undefined pharmacodynamic advantages due to their action on distinct parts of the viral life cycle (such as integration) or, in the case of entry inhibitors, an extracellular site of action. Brenchley et al. [35] postulated that infection in the GI mucosa establishes a cycle of transmigration of proinflammatory substances that sustain local and systemic T cell activation and increase permissiveness for HIV infection that, in turn, promotes continued mucosal disruption. If so, reducing residual replication in GALT could produce results that extend beyond local effects, such as systemic CD4 cell recovery and reduced reactivation of latently-infected cells. Thus, we should not treat the findings of Chun and colleagues [1] as an incidental observation to be filed away or simply as a cause for concern about the limitation of existing treatments, but rather as an opportunity to better understand the persistence of HIV and as a guide for improving existing HIV treatment paradigms.
|
|
| |
| |
|
|
|
