Two drugs from a class new to HIV medicine called BH3 mimetics were unveiled at July’s 12th International AIDS Society Conference on HIV Science (IAS 2023) in Brisbane. They may contribute to a cure for HIV by killing off long-lived cells that contain HIV genes in their DNA. Notably, venetoclax (Venclexta) and obatoclax only killed off cells containing intact DNA, capable of giving rise to new viruses, and did not delete cells containing defective, harmless DNA.
A number of drugs and treatments from the anti-cancer arsenal have been investigated as HIV cure research such as HDAC inhibitors, PD-1 inhibitors and therapeutic vaccines. (And, of course, the six successful cures so far have used the radical cancer therapy of a stem cell (bone marrow) transplant.)
This is not coincidental: cancer and AIDS are both the end result of mutations in the DNA of some of our cells. In the case of cancer they arise in the host DNA and in HIV infection they are introduced by a virus, but both are the result of ‘rogue genes’ (some other viruses, such as HPV, directly cause cancers).
Cures for cancer and HIV share one other property. They are difficult because the immune system finds it harder to ‘see’ cancer and HIV-infected reservoir cells as they do not show most of the obvious distress signals that virally infected cells usually do. Drugs that specifically target HIV-infected cells would take us a long way towards a cure.
In the case of the BH3 mimetics, HIV cure researchers noticed a similarity between cancer and HIV-reservoir cells in that both seem to be ‘immortalised’. Something seems to stop them undergoing the normal and necessary ‘pruning’ of cells called apoptosis, which kills off cells that have outrun their usefulness or whose senescence may cause malfunctions.
It was already known that the immortality of some cancer cells depends on proteins belonging to a family called BCL-2, which switch off the process of apoptosis. BCL-2 proteins are necessary for growth and healing when needed, but are hugely overabundant in cancer cells. They were first spotted in B-Cell Lymphoma, hence the name.
Normally the activity of BCL-2 proteins is regulated by members of the same family called BAX proteins that act as their ‘anti-selves’ – slotting into a cleft in the BCL-2 and ‘locking’ the protein’s anti-apoptotic action. Cancer researchers discovered a class of small molecules called BH3 mimetics that slot into BCL-2 proteins in the same place as BAX (it’s called the BCL-2 Hydrophobic-pocket number Three, hence BH3).
Hypothesising that HIV-infected cells become long-lived reservoir cells that evade immune surveillance due to the action of BCL-2, researchers have wanted to see if BH3 mimetics could disrupt that process.
Venetoclax is furthest along in the research pathway because it is already a licensed treatment for certain types of leukaemia under the brand names Venclexta and Venclyxto, and has thus cleared safety hurdles.
Dr Philip Arandjelovic of Australia’s Walter and Eliza Hall Institute of Medical Research presented the results of a study of venetoclax in 14 mice ‘humanised’ so their immune systems were receptive to HIV infection.
The mice were infected with HIV, then placed on three-class antiretroviral therapy (tenofovir/emtricitabine/rilpivirine/raltegravir). As it would with humans, it took about a month for them to become fully virally suppressed. While still on ART, eight of them were given venetoclax and six of them a placebo, over seven weeks. These were given on a five-days-on, two-days-off basis, and there was also a pause of a week in the middle. This was to avoid intolerance as far as possible; cytotoxic drugs like BH3 mimetics almost inevitably have some side effects, including a general reduction in all cellular blood components (pancytopenia).
“This suggests venetoclax is doing a lot more than indiscriminately destroying cellular viral DNA”
After the last venetoclax dose, the mice were then taken off ART. HIV viral loads in humanised mice normally rebound very fast (they have fewer cells to re-infect), and all the placebo animals had detectable viral loads within a week. Rebound was slower in the mice receiving venetoclax – one rebounded after a week, five after two weeks and two after three weeks.
The same study was repeated with venetoclax and another drug called S63145, an inhibitor of a different apoptosis suppressor called MCL-1 (found in another cancer, Myeloid Cell Leukaemia). The extra drug did nothing in itself, but when the two drugs were combined and given to four mice, two rebounded virologically after two weeks but the other two not till four weeks (so the median time to rebound was two weeks on one drug but three weeks on both).
In ‘ex vivo’ tests, DNA-positive CD4 cells were donated by 10 people with HIV and cultured in a lab dish along with three different does of venetoclax (5, 10 or 100 nanomols) or with the inactive substance dimethyl sulphoxide (DMSO). DNA assays measured the amount of total HIV DNA, intact DNA capable of forming new viruses, and defective DNA.
At the highest venetoclax dose, the total amount of DNA declined, relative to cells treated with DMSO, by 61% and the amount of intact DNA by 42%. The former result was statistically significant (p=0.031), the second result not quite so (p=0.064, where 0.05 is the conventional upper limit of statistical probability, or less than one chance in 20 this was a random finding). In contrast, there was no decline at all in the amount of defective DNA in the cells.
This last finding is the most interesting one as it suggests venetoclax is doing a lot more than indiscriminately destroying cellular viral DNA. It suggests that by selectively inhibiting BCL-2 in cells that contain too much of it, it also selectively marks out only those immune cells that contain viable HIV DNA and ignores cells containing only fragments of it. This ability to target active or potentially target active reservoir cells is one the most essential requirements of the most feasible and scalable HIV cures.
Obatoclax – and others
A second lab-dish study presented at IAS 2023 suggested venetoclax might not be the most powerful BH3 mimetic in terms of curing HIV. Obatoclax is another member of the BH3 mimetic drug family and is an inhibitor of both BCL-2 and MCL-1. It is owned by Teva Pharmaceuticals and is under investigation for leukaemia and lymphoma, but has yet to be licensed.
Dr Steven Yukl of the University of California, San Francisco took CD4 cells from nine donors living with HIV and cultured them for six days, either with standard ART plus DMSO as a buffer, or with DMSO plus a panel of 10 drugs with immunomodulatory effects, most of which have been used for cancer, but have diverse mechanisms of action.
- Sapanisertib, a drug currently under trial for a wide range of cancers, and related to the established anti-inflammatory drug sirolimus.
- Bortezomib (Velcade), a drug that stops cancer cells being able to eject waste products, thus poisoning the cell.
- Birinapant, a drug under trial for breast cancers which mimics SMAC, an antagonist of another anti-apoptosis protein family called IAPs.
- Nivolumab (Opdivo), one of the many monoclonal antibody drugs that have revolutionised cancer treatment. A PD-1 inhibitor, it prevents cells being able to ‘hide’ from the immune system by becoming latent.
- Vesatolimod, an antiviral rather than anticancer drug, which provokes a strong immune reaction to viruses including HIV, hepatitis B and possibly rotavirus. It is under investigation by Gilead Sciences.
- Acitretin (Soriatane), an immunomodulatory drug licensed to treat psoriasis.
- Auranofin (Ridaura), a gold-containing anti-inflammatory drug used to treat rheumatoid arthritis. It has been under investigation since 2011 as a drug capable of shrinking the hidden reservoir of HIV cells.
- Interferon-alfa and interferon gamma, two types of the family of proteins made by the body’s innate immune system as a first line of defence against infections. Many different formulations of artificial interferon-alfa have been used to treat cancers and viral infections. Interferon gamma is more of a cytokine, a substance that directs the activities of other parts of the immune system and is used, for instance, as an indicator of immunogenicity in vaccines.
The results were interestingly different from the venetoclax study. Obatoclax did not reduce the total amount of DNA in the cell samples – and neither did any of the other drugs (obatoclax did reduce total DNA by 25%, but this was not statistically significant). But it did very significantly reduce the amount of intact viral DNA in the cell samples, by an average of 83%.
Obatoclax also produced the most consistent results. The amount of intact viral DNA shrank to zero in cells from some of the patients treated with obatoclax – and also with auranofin, interferon-alfa, birinapant, bortezomib and sapanisertib. But with all drugs except obatoclax the amount of detectable viral DNA in the cells from some patients actually rose – and in some cases rose considerably, by 50% to 200%, reminding us that immunomodulatory drugs can produce paradoxical effects. In contrast there was no increase in intact DNA in any cells treated with obatoclax, although in three cases the decline in intact DNA was less than 50% and on one case only 20%.
One possible caution about obatoclax is that it has not yet been fully evaluated for safety. Although no drug produced serious cellular toxicity, there were slight signs of it with obatoclax (and auranofin) and the lab-dish doses were reduced – which did not seem to reduce the efficacy of obatoclax greatly.
As with venetoclax, the interesting thing about obatoclax is that it seems to be able to produce results that distinguish between intact viral DNA and harmless, defective DNA. There are also other BH3 mimetics on the horizon such as navitoclax.
Research into the HIV cure potential of these drugs is at an early stage. The first study of venetoclax in 18 HIV-positive human volunteers is due to start this month, with results due late next year but, as the history of translating cancer cures into HIV cures has already taught us, it may be a long time before we develop the perfect HIV-killing drug package.
Arandjelovic P et al. Venetoclax, alone and in combination with the BH3-mimetic S63845, depletes HIV-1 latently infected cells and delays rebound in humanized mice. 12th IAS Conference on HIV Science, Brisbane, abstract OALBA0503, 2023.
Following the conference presentation, the research has also been published in a peer-reviewed journal:
Arandjelovic P et al. Venetoclax, alone and in combination with the BH3 mimetic S63845, depletes HIV-1 latently infected cells and delays rebound in humanized mice. Cell Reports Medicine 101178, 2023 (open access).
Yukl S et al. Differential susceptibility of cells infected with defective and intact proviruses to HIV-selective cell killing by small molecule therapies. 12th IAS Conference on HIV Science, Brisbane, abstract OAA0302, 2023.