Mechanisms of HIV Latency in CD4 T cells Good morning My project is focused on establishing a novel model for HIV latency in primary human T cells and using such a system to characterize the molecular mechanism of HIV latency establishment, maintenance and reactivation.

HIV latency has emerged in recent years as one of the main barriers in our effort to eradicate HIV infection. Patients treated successfully with antiHIV agents show a near complete suppression of HIV replication but the longterm persistence of latently infected cells. These latently infected cells are thought to allow the reseding of the infection upon cessation of therapy. During the past few years, we have learned that: All patients contain latently infected cells Latently infected cells are rare HIV persists in TCM and TTM CD4 T cells Persistence is maintained by T cell survival and homeostatic proliferation Clearance of the latent reservoir should take more than 60 years ….

TTM From the work of Dr. Sekaly, we have recently learned that latent HIV in infected patients predominantly resides in central memory T cells and in effector memory T cells. If we want to establish a model for latent HIV, we need to focus on these cells. A key unanswered question is how do these cells become infected. The most commonly accepted model is that naïve T cells are infected as they become activated and that latency happens as the cell returns to the quiescent stage as a central memory T cells. However,, we know that latent HIV is largely of the CCR5 type but naïve T cells do not express CCR5. In fact, all models for HIV latency in primeary cells are using CXCR4 viruses.

The HIV Life Cycle Preintegration vs. Postintegration latency t1/2=1-2 weeks t1/2=44 months The HIV Life Cycle We have also learned that this long term HIV latency is established at the transcriptional level. On this slide, I am showing you a simplified version of the HIV lifecycle illustrating the two different outcomes of an HIV infection at the transcriptional level: 1- active transcription leads to efficient synthesis of viral proteins and viral production 2. Inactive transcription leads to a shutdown of virus production.

The HIV Life Cycle What is Latency? Importantly this shutdown is reversible and under the proper stimulating conditions, virus transcription resumes and leads to a fully productive infection

To understand HIV latency, we therefore need to understand HIV transcription which has been the focus of my laboratory for the past 20 years. There are two phases in HIV transcription, an early phase dependent on cellular factors binding to the HIV promoter. This phase is inefficient due to an inability of the polymerase to successfully elongate and aborted transcription. In some rare cases, however, the polymerase elongates through the whole genome leading to the synthesis of the mRNA of the viral transactivator Tat. Tat binds to an RNA element at the 5 prime extremity of the viral genome and dramatically enhances the elongation efficiency. This results in more Tat synthesis and in a high rate of transcription.

We can also see from this graph that if basal HIV expression proceed at a lower basal rate, the threshold for Tat-mediated activation of HIV expression might not occur. These are conditions that would lead to reversible HIV silencing of HIV transcription, or latency.

The advantages of this experimental approach are several; The experiments can be conducted on a small number of cells (one thousand). This means that very rare cells can be studied The real time monitoring will allow us to examine HIV expression over time. I presented earlier a possible model of latency establishment based on the de novo silencing of the HIV promoter following integration. However, we have considered an alternative hypothesis shown on the next slide.

It is indeed possible that HIV latency is established after a period of productive HIV infection. If we imagine that a productively infected and activated lymphocyte returns to a quiescent state, we can hypothesize that the return to quiescent could be inducing a return to latency of the HIV genome integrated in this cell. This technical approach will allow us to determine whether this truly occurs. Determining whether this actually occurs has potentially important implications. If this model shown here indeed occurs, it indicates that some or all of latently-infected cells go through a productive infection first. The expression of several viral proteins, such as Tat, Nef, Vpr are know to induce changes in cellular gene expression during infection. A latent cell derived from a productively-infected cells would be modified in its pattern of gene expression. Such change in gene expression could used in the future to isolate and characterize such cells and possibly to selectively target them.

To understand HIV latency, we therefore have to understand the cellular factors that contribute to basal HIV transcription. There are two broad classes of factors that I have categorized into cis- and trans-acting factors. Cis-acting factors represent the variable integration sites and distinct chromatin environments with variable transcription potential. While HIV has evolved a unique mechanism that allows it to specifically target transcriptionally favorable environments, this targeting mechanism is not fool proof and, under some conditions, the virus may integrate in a transcriptionally unfavorable environment where it comes under the activity of chromatin associated factors that repress its transcription. Transacting factors refers to transcription factors such as NFAT and NFKB that recognize the HIV promoter and that are regulated by the state of T cell activation. Importantly, the relative contribution of both effects, cis and trans effects might differ between transformed cell lines, in which latency has been studied most intensely thus far, and primary lymphoid tissues, which is the true and relevant site of HIV latency.

To understand HIV latency, we therefore have to understand the cellular factors that contribute to basal HIV transcription. There are two broad classes of factors that I have categorized into cis- and trans-acting factors. Cis-acting factors represent the variable integration sites and distinct chromatin environments with variable transcription potential. While HIV has evolved a unique mechanism that allows it to specifically target transcriptionally favorable environments, this targeting mechanism is not fool proof and, under some conditions, the virus may integrate in a transcriptionally unfavorable environment where it comes under the activity of chromatin associated factors that repress its transcription. Transacting factors refers to transcription factors such as NFAT and NFKB that recognize the HIV promoter and that are regulated by the state of T cell activation. Importantly, the relative contribution of both effects, cis and trans effects might differ between transformed cell lines, in which latency has been studied most intensely thus far, and primary lymphoid tissues, which is the true and relevant site of HIV latency.

To understand HIV latency, we therefore have to understand the cellular factors that contribute to basal HIV transcription. There are two broad classes of factors that I have categorized into cis- and trans-acting factors. Cis-acting factors represent the variable integration sites and distinct chromatin environments with variable transcription potential. While HIV has evolved a unique mechanism that allows it to specifically target transcriptionally favorable environments, this targeting mechanism is not fool proof and, under some conditions, the virus may integrate in a transcriptionally unfavorable environment where it comes under the activity of chromatin associated factors that repress its transcription. Transacting factors refers to transcription factors such as NFAT and NFKB that recognize the HIV promoter and that are regulated by the state of T cell activation. Importantly, the relative contribution of both effects, cis and trans effects might differ between transformed cell lines, in which latency has been studied most intensely thus far, and primary lymphoid tissues, which is the true and relevant site of HIV latency. DNA methylation miRNA

Heterogeneity in HIV Expression: Cis- vs. Trans- Effects

0.5-2%

Latent HIV integration sites in J-Lat cells Heterochromatin mediated silencing ~10% Transcriptional interference ~10% Virus integration-mediated mutagenesis <5% Gene deserts, long intergenic regions ~15% 5. Unknown (integration within genes) ~60%

Promoter

HIV CpG Islands Are Methylated in Latently Infected Primary CD4+ Cells Quiescent nonpolarized Reactivated nonpolarized Mock Mock 1.6 56 21 days post infection

Conclusions -HIV Latency is established at the transcriptional level in primary memory T cells -Cis-acting and Trans-acting mechanism are likely to contribute to the establishment of HIV latency in primary T cells -While trans-acting factors are likely to be homogeneous within a homogeneous lymphocyte population (i.e. resting TCM), different integration sites are likely to suppress HIV transcription via distinct mechanisms. This is due to the variety of repressive chromatin environments that exists in cells. -HIV latency is therefore likely to be heterogeneous (therapeutic implications) -The HIV genome is methylated in a large fraction of latently-infected cells (both transformed and primary lymphoid cell models) -Inhibitors of methylation potently synergize with trans-acting factors (TNF, prostratin) to reactivate latent HIV -Future efforts will need to better define the mechanism of HIV latency in primary lymphoid cells either isolated from patients or from appropriate in vitro model systems.

Gladstone Institute of Virology and immunology UCSF T. Mahmoudi D. Bisgrove A. Jordan S. Kauder S. Hakre L. Chavez U. Penn M. Lewinski F. Bushman Utah School of Medicine A. Bosque V. Planelles