Monitor behavior of latent proviruses in the frame of antiretroviral therapy
In many cases of viruses, the status of viral transcription shows a strong correlation to their pathogenesis. Viruses, such as hepatitis B virus, herpes virus and HIV[83–85]require a transcriptionally silent state, so-called latency throughout their life cycles, allowing viruses to persist in host cells for a long time and establish chronic infections. In the case of HIV, viral rebound frequently occurs when the antiretroviral therapy is interrupted, subsequently leading to the AIDS symptoms resurface. It thus becomes an obvious obstacle for curing HIV-1 infection. In the previous section we talked about molecular barcodes used to quantitatively measure single virus transcription. Here we are going to give several examples that illustrate how molecular barcodes help us to interfere with translation research involving antiretroviral therapy in addition to the measurement of viral transcription.
The “shock and kill” antiretroviral therapy has been the most attractive strategy for purging HIV latent infection as it was first proposed in 2012[86]. The concept of this strategy is to give HIV-1-infected individuals latency reversing agents (LRAs) such as histone deacetylase inhibitors to deliberately reactivate proviral transcription in latently infected cells, rendering reactivated cells destroyed by the host immune system. The success of this strategy relies on the drugs used to reactivate latent proviruses, yet several exploratory clinical trials have already shown limited evidence of efficacy[87]. Since molecular barcodes allow us to quantitative measure HIV transcription at a single-virus level, Chen and colleagues thus attempted to use them for estimating the spectrum of vorinostat, an histone deacetylase inhibitor used in clinical trials[88]as a proof of concept. Comparing with phytohemagglutinin, a strong activator for T cells, it turned out that the same provirus can either respond more strongly to phytohemagglutinin than vorinostat, orvice versa , indicating that vorinostat fully reactivate only a subset of the proviruses in a cell population. Proviruses that were prone to be reactivated by vorinostat were more frequently in the vicinity of active regulatory elements (H3K27ac and histone H3 trimethylated at K4 and monomethylated at K4)[25]. This observation was later confirmed by Battivelli and colleagues[89]using a second generation of full-length dual-fluorescence reporter HIV, HIVGKO, to investigate the reactivation potential of various LRAs, including panobinostat[90], JQ1[91–95]and bryostatin[96,97]in pure latent populations. They showed that less than 5% of the latently infected primary CD4+ T cells were able to be reactivated by LRAs and confirmed that the proviruses refractory to be reactivated appeared to be present closer to heterochromatin (histone H3 trimethylated at K27 and at K9) and non-accessible regions (DNase hyposensitivity)[89].
In vivo barcoded HIV has also been used to track viral rebounds in animal models. Marsden and colleagues[61]infected bone marrow-liver-thymus mice, an animal model that can represent the human immune system[98], with barcoded nearly full-length and replication-competent HIV and measured viral rebound after giving a synthetic bryostatin analog, SUW133[99], which is a PKC inhibitor[100]. It was observed that mice treated with SUW133 showed a delay of viral rebound after antiretroviral treatment was stopped compared with vehicle control. In addition, barcode diversity present in reactivated viruses showed a decrease compared with the mice treated with vehicle control, meaning that the administration of SUW133 can eliminate a subset of latently infected cells in this in vivo model.
Nonhuman primates are another useful animal model to study retroviruses. SIV-infected nonhuman primates are similar in their pathophysiology and viral dynamics to humans infected with HIV-1[101–105]and therefore make ideal models for HIV studies in many aspects. Fennessey and colleagues[63]successfully employed the model of rhesus macaques infected with barcoded SIV to identify and quantify the dynamics of the establishment of viral reservoir and viral rebound after combination antiretroviral therapy administration and interruption, demonstrating that the time of initiation and duration of therapy administration in nonhuman primates can alter the size of the reservoir. Soon afterwards Khanal and colleagues[106]generated a derivative based on this barcoded SIV. At present barcoded SIV continues to be applied on animal models of rhesus macaques to tackle several important questions in the field of HIV/AIDS in many aspects[107–111].