Gene Editing for HIV Therapies

by Robert Kruse

Human Immunodeficiency Virus (HIV) has been a truly intractable disease for researchers to eradicate since it's discovery over 35 years ago. HIV is a lentivirus that integrates into host T cells, along with monocytes and possibly other cell types causing lifelong infection. The advent of HAART (or highly active antiretroviral therapy) was able to suppress the production of newly formed virus preventing new cells from being infected. The drugs target multiple different processes important for the viral life cycle, including reverse transcription, protease activity, and more recently, integrase activity. This causes a plummet in viral loads to undetectable levels in the serum, representing a functional cure for the patient. Unfortunately, with the withdrawal of anti-retroviral drugs, the viral levels quickly rebound, being derived from stably integrated virus in T cells. Initially, there was hope that T cells with integrated virus could be waited out, such that they would all die out after a couple years, and then treatment could be stopped. Those hopes were dashed with the realization that HIV integrated into long lived memory T cells and possibly other cell types and represents a long term reservoir with a half-life of several years. Thus, for the foreseeable future, antiretroviral drugs must be maintained.

This best describes the hurdles and challenges of the current state of HIV medications. Into this frontier, many people are now hyping up using gene editing as a tool to treat HIV. The gene editing strategies can basically be broken down into two camps. The first targets host genes, in order to facilitate an advantage against HIV. The second targets the virus itself, in order to purge it from host cells. Both have received attention from academia and biotech alike.

1)  Gene Editing of Host Genome in order to Protect against HIV

The furthest along gene editing strategy against HIV targets removal of host CCR5 expression in T cells in order to protect those cells from HIV entry. HIV enters cells by first attaching to CD4 through gp120, and subsequently binding to CCR5 in order to induce a conformational change resulting in membrane fusion. In the human population, individuals with a deletion in CCR5 have been found to be resistant from HIV infection, through public health studies. This observation was then turned into a therapy with the treatment of the "Berlin patient" with a bone marrow transplant from a donor with CCR5 del32 genotype. The patient had concomitant leukemia, and necessitated undergoing a bone marrow transplant anyways. His physician hypothesized that infusing a new immune system with the CCR5 deletion would prevent residual HIV from infecting the new cells leading toward eradication. Remarkably, the strategy worked and the patient has been hailed as a paradigm toward the cure for HIV.

The excitement over the result is potentially dampened by a few questions. It's uncertain how much contribution the chemotherapy the patient was administered to treat his cancer and in turn wipe out all of his immune cells, actually helped eliminate his HIV via eliminating the reservoir cells, versus the cure happening because the new immune cells were resistant to infection; GVHD could have also helped eliminate HIV infected cells. Given that the patient took HAART throughout this clinical course, before withdrawing medication, it makes it difficult to determine. This fact matters a lot when contemplating how to translate this strategy into the clinic. Bone marrow transplants with fully ablative regimens are too dangerous to use on an otherwise healthy individuals. Instead the clinical application requires modifying ex vivo a small number of hematopoietic stem cells or T cells. These cells will initially represent a small proportion of the patient's immune system, and then could grow through a selective advantage of not being eliminated by HIV, in theory preventing loss of CD4 T cell levels. Another possibility is that the few protected, uninfected cells could restore enough of the patient's immune system to reverse course and clear HIV permanently. Both of these hypotheses remain to be born out in human patients, however. One concern is that the majority of T cell loss occurs without those T cells ever becoming infected, so how could a small minority exert any effect on this process? If any company will find out, it will be Sangamo Biosciences ($SGMO), with its zinc finger nuclease technology already has this strategy of gene editing knockout targeting CCR5 in Phase II clinical trials, and with several other trials planned or ongoing that consist of optimized versions of the approach. The Phase I trial results showed that the cells could be safely administered with a modest survival advantage.

2) Gene editing to Excise HIV from the Host genome

As discussed above, the main hurdle toward eradicating HIV is that it integrates into long-lived immune cells, such as memory helper T cells, which don't actively express all HIV genes, with the HIV genome remaining relatively latent inside the cell. HIV gene expression is tied to the activation state of the T cell, affording the possibility of latency periods in T cell lineages. There has also been some speculation of HIV integrations causing increased T cell divisions in an oncogenic effect to increase persistence.

An elegant solution to all these problems would be to simply cut the HIV genome out of the host cell, whereafter it forms episomal DNA that is lost with cell division. With the rise of CRISPR, one solution has been to use Cas9 to cut HIV out of the genome, as demonstrated here and here. A similar strategy has used TALENs to edit HIV from the genome. Just a couple weeks ago, a paper was published online showed in vivo administration of an AAV expressing Cas9 and gRNA could remove HIV sequences in a transgenic HIV mouse model where the target was in every chromosome, in a fraction of the lymphocytes and cells of other organs. However, the strategy has a number of limitations. A simple one is that the HIV sequence is heterogeneous, so a small fraction of circulating virus will not be cut. Alternatively, Cas9 cleavage itself is mutagenic, meaning that it can induce mutations that cause the resultant HIV genome to be resistant to further Cas9 cleavage. In essence, the therapy is aiding and abetting the escape of the criminal.

A strategy that accomplishes genome editing in a unique way that might be less resistant to mutational changes is using evolved recombinases that can recognize the HIV long terminal repeats (LTRs). The ends of the HIV genome are called LTRs, formed during the process of reverse transcription and harboring the promoter and polyadenylation elements respectively. They also harbor palindromic sequences, which are similar to the LoxP sites targeted by Cre recombinase. Researchers have been able to adapt Cre recombinase via evolutionary selection in bacteria to instead target HIV LTRs. The end result is efficient excision of the entire genome into an episome, very similar to what many science researchers use today with Cre excision of different exons or stop cassettes in transgenic mouse models.

The real question with both of these HIV targeted strategies are whether they are even practical and worth pursuing. HIV infected cells are a needle in a haystack. During active infection, levels might only be 1% of CD4 T cells, and the percentage is much lower for the targeted latently infected T cell types that are the real goal for this type of approach. These cells look functionally similar to non-infected cells, so no specific targeting approach is likely possible. One would need to introduce CRISPR into billions of your T cells, hoping some of them harbor HIV. Even withdrawing blood in an apheresis setting, the total would only be a fraction of the T cells in your body. A vector that could target memory T cell markers and cell types would also be useful, but again, if you only hit a portion, and the virus could grow back anyways that you didn't hit, what was the point exactly. Similar challenges present themselves for HSV and HBV applications, but those viruses at least operate under different replication kinetics than HIV. Either way, the HIV focused strategy has significant delivery challenges that might preclude it from ever seeing the clinic.

Conclusions

Sangamo's trials targeting CCR5 offer great insight into the immunology of HIV. I would hesitate that it's still too early to gauge success. HIV leads to disrupted lymph node architectures and dysregulation of the entire immune system, meaning the gene edited cells, even if protected, might not stand much of a chance of long term survival in a hostile environment. That said, it does offer some hope for the community. CRISPR strategies targeting CCR5 have already been published and might soon move into the clinic as well. On the other hand, the direct HIV targeting strategy has significant delivery challenges that might preclude it from ever seeing the clinic. The best hope would be a procedure to lower the latent provirus enough in patients where the remaining infected cells limited half-life could lead to a sterilizing cure under HAART without off-treatment rebound. Either way, by expanding our tool kit with genome editing, the ability for HIV to hide is continuing to diminish, and new treatments may be heading toward patients.