What Integration Sites Tell Us about HIV Persistence

Abstract

Advances in technology have made it possible to analyze integration sites in cells from HIV-infected patients. A significant fraction of infected cells in patients on long-term therapy are clonally expanded; in some cases the integrated viral DNA contributes to the clonal expansion of the infected cells. Although the large majority (>95%) of the HIV proviruses in treated patients are defective, expanded clones can carry replication-competent proviruses, and cells from these clones can release infectious virus. As discussed in this Perspective, it is likely that cells that produce virus are strongly selected against in vivo, and cells with replication competent proviruses expand and survive because only a small fraction of the cells produce virus. These findings have implications for strategies that are intended to eliminate the reservoir of infected cells that has made it almost impossible to cure HIV-infected patients.

Figure 1

Retroviral DNA Integration. A. The steps in retroviral integration. At the top is a drawing of the full length linear viral DNA. In the second panel, IN has removed 2 nucleotides from the 3’ ends of the viral DNA (the 3’ Processing reaction). In the third panel, the newly exposed 3’ ends of the viral DNA are aligned with the host genome for the Strand Transfer or ST reaction (shown in panel 4). The five dots represent the nucleotides of the host DNA that are duplicated in the HIV integration reaction. In the bottom two panels, host repair enzymes repair the mismatches and nicks in the DNA left by the ST reaction, creating the target site duplication. B . The crystal structure (Maertens et al., 2010) of a complex of PFV IN, two pieces of DNA that represent the ends of the virus DNA (vDNA), and the target DNA (tDNA). There are four IN subunits in the complex. All of the domains of the two inner subunits, which are directly involved in the ST reaction, can be seen; these two are colored blue and green. Only a portion of the two outer subunits (colored gold) can be seen in the crystal structure; these are the central catalytic domains (CCD); the CCDs of the two inner subunits are also marked. Compared to the IN of HIV, PFV IN has an N-terminal extension domain (NED)(seen only on the two inner INs). C. Structure of the host and virus DNA with the IN protein removed. The tDNA is bent, as can be seen more clearly in this panel, because the DNAs are rotated 90° relative to panel B. Panel C also shows the position of the host strands that participated in the ST reaction after the ST reactions have taken place. Reprinted with permission from (Brown, 1997) (A) and (Maertens et al., 2010) (B and C).

Figure 2

Viral RNA from blood and viral DNA from PBMCs of patients on ART. A. Kinetics of decay and rebound. Note that the Y-axis is a log scale. Immediately following the start of therapy, there is a rapid loss of the viral RNA and the level of the viral RNA decays in several stages (see text). The level of the viral DNA also decays, but the loss of viral DNA is much smaller. Interruption of therapy, even after many years, invariably leads to a return of the virus to levels very near the pretherapy values. B. Comparison of HIV DNA sequences before and after ART. The drawings show neighbor-joining trees representing the relatedness of gag-pol sequences from viral DNA isolated from blood cells. The open dots show viral sequences from cells taken either before or just after the initiation of ART. All of the patients had been infected for a long time, and no two pretherapy viral DNA sequences are alike. After prolonged ART (>7 years) another sample was taken from the blood of all 4 patients. After long term ART, clusters of identical viral DNA sequences are seen in samples of blood from each of the patients (filled dots, circled in red) Reprinted from (Hilldorfer et al., 2012).

Figure 3

Only a small fraction of the cells in an expanded clone that carries an infectious provirus make virus at any one time. A group of infected cells is shown on the left. In patients, only about one cell in a thousand is infected; for simplicity, uninfected cells are not shown. Most of the infected cells carry defective proviruses; a cell that carries an infectious provirus is shown in blue. The blue cell divides, giving rise to a clone. Most of the cells in the clone do not make viral RNA or produce virus (blue cells in the upper part of the drawing); however, by an unknown mechanism, some of the cells in the clone are occasionally activated to make virus (bottom). These cells most likely die rapidly (dying cells are shown in faded red). In treated patients, effective therapy (cART) prevents the virus that is released from the cells from infecting new cells. The clone is maintained by division of the (blue) cells that do not make virus; from this reservoir, cells that can make virus are continually produced. Although the virus producing cells die rapidly, new virus producing cells arise, and these cells maintain a low level of virus in the blood (red, lower right). The authors thank Francesco Simonetti for this figure.

Figure 4

Integration sites in the MKL2 gene. A diagram of the MKL2 gene on chromosome 16 is shown; the exons are indicated as small vertical bars. Integration sites in PBMCs taken from healthy human volunteers and infected in culture are shown above the diagram of the gene. The orientation of the proviruses relative to the gene is indicated by the direction of the arrowheads and by their color. The integration sites obtained from patient 1 of Maldarelli et al. (2014) are shown below the gene. The 7 kb portion of the gene, encompassing 3 small introns where there were integration sites in the patient, is expanded and is shown at the bottom of the figure. Circled arrowheads indicate integrations that were shown to be clonally expanded. The integration site data for patient 1 is reprinted from (Maldarelli et al., 2014).