Towards a cure for all: How we might do it

In the second part of this two-part feature, Gus Cairns investigates current research into finding a cure for HIV.

Last month, we looked at the case of Timothy Ray Brown, a leukaemia patient who became the first person ever to be cured of HIV infection.¹

We explained why this is so difficult: even under the most intensive current therapy, a silent ‘reservoir’ of a type of CD4 cell called ‘memory cells’ remains infected with HIV. These are like sleeper cells in a resistance organisation – their job is to spring into action when a specific infection they are primed to recognise turns up. In other medical conditions, vaccines work by tricking cells to ‘recognise’ an infection without actually having had it. The trouble is, when HIV-infected memory cells spring into action, they start spewing out HIV.

We can flush HIV-infected memory cells out of hiding by activating them and then kill them: but the burst of HIV they produce in this process causes more CD4 cells to be infected.

Last month, we explained how Brown’s doctor, Gero Hütter, got round this by destroying Brown’s CD4 cells and then re-introducing others, via a bone marrow transplant, from a donor naturally resistant to HIV (missing the CCR5 co-receptor, which HIV grabs on to). However, a bone marrow transplant, while the standard second-line treatment for leukaemia, is far too toxic – and expensive – for general use and indeed nearly killed Brown.    

It is, however, proof that a cure is possible. The most promising approach towards a cure for all is to do at least one of the two things Dr Hütter did, but in a much more subtle way.

1. Re-engineer CD4 cells

One approach could be to take bone marrow cells from the patient’s own body, and by means of enzymes and genetic tools, engineer them to become CCR5-negative, thus protecting them against further HIV infection. You then re-introduce them into the patient’s body, in a so-called ‘autologous’ – meaning self-donated – transplant.

The hope is that the CCR5-negative cells would slowly start to take over from the CCR5-positive cells and HIV would slowly be starved of the cells it needs in order to reproduce.

Even people on effective antiretrovirals maintain a viral load averaging three copies/ml and this appears to contribute to keeping the immune system in a permanently higher state of activation than in HIV-negative people. This activation kills off some HIV-infected cells but infects others, keeping the reservoir topped up, or so the theory goes. If, however, a population of infection-proof cells were introduced, they would come to predominate as there would be fewer cells to infect as time went by.

This approach has actually been trialled successfully, in mice genetically modified to be susceptible to HIV. Researcher Paula Cannon and her team from the University of Southern California used a drug called SB728, a so-called zinc finger nuclease enzyme, to snip out CCR5 from mature CD4 cells and then re-introduce them into the blood.² They then infected these mice and a control group with HIV. The control mice lost their CD4 cells and developed AIDS within 8 to 12 weeks but the mice given CCR5-negative cells maintained normal CD4 counts and undetectable HIV viral loads.

Many scientists are sceptical that the reservoir of HIV-infected cells could be replaced by HIV-proof cells unaided. The CCR5 cells in Cannon’s mice were by no means eliminated, especially the all-important progenitor cells in the bone marrow. Steven Deeks, a prominent cure researcher from the University of California, San Francisco, says: “They did the transplant first and then the infection.” It might not work in people already infected, where there is an established reservoir of HIV-infected cells.

Even if it does work, it could take a long time for one cell population to replace another: “In mice it happens in months, in people it could take years,” Deeks told HTU.

Nonetheless Cannon and her colleague John Zaia are now leading a Phase I trial in patients with lymphoma, using bone marrow transplants of patients’ own genetically engineered progenitor cells to try to ensure the growth of a CCR5-negative cell population.³

2. Delete infected cells

Alternatively, one approach could be to concentrate more on the immune-destruction part of Timothy Ray Brown’s therapy instead of the CCR5-deletion bit. The idea would not be to crudely annihilate all the cells HIV might infect. Instead we could:

‘Purge’. This strategy involves enticing reservoir cells out of hiding using drugs that ‘switch on’ reservoir cells so they become activated and therefore detectable, while keeping patients on antiretrovirals so that the activated cells do not go on to seed new infection. The HIV-infected activated cells would then destroy themselves, and the idea is that repeated cycles of activation would deplete the reservoir beyond the point at which it can replenish HIV – a strategy that’s been called ‘purge’.

Experiments were done more than five years ago using the drug valproic acid (Depakote). This is a member of a class of drugs called HDAC inhibitors, which take the genetic brakes off resting cells. In one study, three out of four subjects given valproic acid achieved a 70% reduction in the number of HIV-infected reservoir cells.⁴ It appears, however, that this reduction may only be temporary: two larger studies in 2008 showed no long-term reduction in the number of HIV-infected reservoir cells in other patients.^5,6

This may be because valproic acid is not strong enough. Trials are planned of a stronger HDAC inhibitor called vorinostat (Zolinza), a cancer drug already used for some types of lymphoma and which is being trialled for anal cancer.⁷ “Vorinostat is a tremendously powerful drug,” says Deeks.

If HDAC inhibitors turn out not to work, there is a second family of drugs called HMT inhibitors, some of them already in use as cancer drugs, that reawaken latently infected cells in a different way. They are only just starting to be studied.⁸

‘Kill’. We don’t yet know if activating HIV-infected cells would cause so many to commit cellular suicide that HIV would be purged from the body. Instead of enticing cells out of hiding by activating them and seeing if they blow themselves up, how about a more aggressive strategy of directly seeking them out and killing them in their sanctuary sites? Amazingly, attempts to do this date from as long ago as 1988, when a group devised a drug ‘missile’ that combined an antibody that locked on to the CD4 molecule with a cell-killing toxin derived from the pneumonia bacterium Pseudomonas. It wasn’t taken further because it wasn’t selective enough, targeting all CD4 cells.⁹

By 2002, we were able to make more specific antibodies that only locked on to the memory cells that form the reservoir, and a team devised a similar cell-missile that eliminated a proportion of latently HIV-infected cells in the test tube, from blood taken from patients with HIV. The trouble is that while it cut the number of HIV-infected reservoir cells by at least 80%, it probably didn’t eliminate enough, while at the same time picking off rather a lot of non-infected memory cells.¹⁰

‘Shock and kill’. We still don’t have a way of infallibly identifying only those one-in-a-million memory cells latently infected with HIV, so we can’t kill them and only them. So researchers are devising combination drug missiles that would both entice HIV-infected cells out of hiding and then seek them out actively for destruction. The idea is to devise a three-component therapy that would combine an immune stimulant, an antibody that seeks out activated cells, and a toxin to destroy the targeted cell, a strategy that’s been called ‘shock and kill’.     

One of the possible problems with both ‘purge’ and ‘shock and kill’ is that anything strong enough to activate enough immune cells might be too toxic to use – as has already proved to be the case with drugs like IL-2. In particular, some researchers are concerned that it may cause inflammation in places like the brain which may have been what happened to Timothy Ray Brown: an opinion piece warning about this appeared recently in the journal AIDS, recommending that attempts to deplete the reservoir this way should be started gradually.¹¹

What we really need is a drug that stops cells from being ‘latent’ and gets them to rejoin the actively circulating, and therefore visible and vulnerable, force of T-cells without widespread immune activation. Researcher Robert Siliciano and his team at Johns Hopkins University in Baltimore are involved in identifying small molecules that could manage this feat, gently teasing the immune cells out of hiding instead of shocking them, and in 2009 identified the first one, a compound called 5HN.¹²

3. Delete resting cells

Another strategy is to try and find markers that uniquely identify infected reservoir cells while they are still resting, and kill them without ever having to activate them. Just because we have found no such markers yet does not mean they don’t exist. Researcher Rafick-Pierre Sékaly, scientific director of the recently established Vaccine and Gene Therapy Institute of Florida, is investigating possible chemical markers, including an enzyme called PDI (protein disulfide isomerase), which might betray the location of resting HIV-infected cells. Sékaly has identified a multiplicity of active genes that characterise resting cells and appear to keep them quiescent, and has also discovered that the presence of another kind of cell called myeloid dendritic cells may be necessary to keep them that way.¹³

Equally, HIV may gravitate towards cells that display particular kinds of biomarkers already, other than the ones we already know, and we could become able to characterise the subset of cells that is most likely to become infected with HIV and target just those for destruction. The cellular receptors CCR4 and CXCR3 have already been found to characterise immune cells in the gut that are more likely to become infected.¹⁴

4. Dry up the reservoir

Cells don’t just passively stop producing HIV and go into quiescent mode by themselves. The process through which a small minority of CD4 cells join the reservoir of resting memory cells is controlled by a complex chemical pathway whereby specific genes are turned off – just like the lights at bedtime. Instead of trying to prod the resting cells to come out of hiding, we could keep these genes active and stop them ever going into hiding in the first place. Protein disulfide isomerase (PDI) is in a family of enzymes that seem to be involved in this process, but there are many more.

One old favourite is a molecule called nuclear factor kappa B (NFκB), a ubiquitous gene activator that was first investigated as a possible target for HIV drugs 20 years ago. Aspirin is a NFκB inhibitor, though its effect is far too weak and non-specific for HIV therapy. Low levels of NFκB are generated during the low-level viral replication seen in antiretroviral therapy, and these levels appear to help keep the HIV reservoir replenished. If you could find a drug that had a much more specific effect on NFκB or one of the other molecules in the cell suppression/activation pathway, you might be able to stop cells joining the reservoir. Conversely, if you stimulate NFκB or related molecules with a stimulant drug like the plant derivative prostratin,¹⁵ you turn reservoir cells into activated ones – another example of the ‘purge’ approach.

However, that also illustrates a problem: some of these cellular proteins, like NFκB, do tremendously complex cellular jobs containing many feedback loops. In one situation they are activators, in another, suppressors, and you may find that inhibiting them has the opposite effect to the one you want. Scientists are therefore investigating drugs that inhibit the mechanism whereby the reservoir gets replenished in other ways. Amongst these is a drug called hexamethylene bisacetamide (HMBA) which might be able to stimulate HIV-infected reservoir cells without activating non-infected ones.¹⁶

Prostratin is quite an exciting drug.This is because, while it stimulates cells to come out of hiding and therefore makes them vulnerable to self-destruction or attack, it also ‘downregulates’ the CCR5 receptor, and indeed another receptor called CXCR4 which some types of HIV use to get into cells. This means that it could be our best shot yet at a drug that purges infected cells but makes other cells less likely to be infected. Prostratin itself looks rather toxic and until recently, has only been available as an expensive extract from the bark of a tree from Samoa, where it has been used to treat liver disease for centuries. Scientists have recently discovered how to make a cheap synthetic version, which means they can start doing bulk searches of similar molecules to find less toxic drugs of the same type.¹⁷

The 'combo' cure

We’re used to combination therapy against HIV and have more recently started talking about combination prevention. A cure for HIV is also unlikely to involve one ‘magic bullet’. Any cure is likely to involve several different approaches, used together or sequentially.

For instance, we don’t yet know if there is a threshold number of infected cells below which active HIV replication is very unlikely to restart. It’s like cancer: can we tolerate a few infected cells in the body, or will the presence of even one eventually lead to the return of HIV?

We could therefore use HDAC and NFκB inhibitors to flush out the majority of infected cells, use engineered CCR5-negative cells to try and replace them, and use a therapeutic vaccine to mount continued surveillance against whatever small minority of HIV-infected cells might still remain. Or – since one of the problems with therapeutic vaccination is that it depends on enhancing an immune response, which may lead to more infection – use an immune-suppressant drug to ‘lock down’ the infected remainder.

There have been a number of attempts already to deliver several HIV eliminators in one package. For instance, the Australian biotech company Benitec has devised a combination consisting of an enzyme that snips out CCR5 from CD4 cells, combined with sections of ‘interfering’ RNA that delete HIV’s reverse transcriptase enzyme and its Tat protein, the viral toxin that over-excites CD4 cells into an HIV-receptive state in the first place. This is all wrapped up in a vector, the shell of an HIV-like virus that infects cells with the genetic products and gets them to start making them. Zaia’s team at the City of Hope Hospital in Duarte, California, has already done a Phase I proof-of-concept trial in lymphoma patients in which the genetically modified cells produced the HIV-disabling products for over two years, though only at low levels.¹⁸

We are only as yet on the first steps of a journey towards making a cure practicable for all, though in researching this article I sensed a new confidence amongst researchers that it might be possible. Many refused to guess at timelines, but Steven Deeks told me that a usable cure strategy would take “at least ten years”.

Sharon Lewin of Monash University in Melbourne, Australia, made a keynote address at the opening of the International AIDS Conference in Vienna last year,¹⁹ and, with Nobel Laureate and co-discoverer of HIV, Françoise Barré-Sinoussi, was instrumental in pulling together a pre-conference two-day workshop on strategies towards a cure.²⁰

In her keynote address she said she was encouraged by two major cure-research initiatives now underway: amfAR’s ARCHE initiative, which had a budget of $1m, and the Martin Delaney Collaboratory, a public/private partnership of research labs funded to the tune of $8.5 million by the US National Institutes of Health and named after the late AIDS activist who founded Project Inform. However, she pointed out that less than 10% of the current funding for an HIV preventive vaccine is currently devoted to curing HIV.

“Cure research doesn’t have to be hugely expensive,” she told HTU. “You don’t need the big trials with tens of thousands of people you need for vaccine and biomedical prevention studies. The initial discoveries can be made with studies of 100 people. But we do need large, multidisciplinary consortia like the Martin Delaney project to ensure that research is co-ordinated and not wasteful.”

The final question, though, is one only Deeks addressed, among the researchers I talked to. We can control HIV and the illness caused by it, but it’s becoming apparent we may never be able to treat everyone because of the massive levels of funding, human resources and healthcare provision needed. Will the same be true of a cure?

“A cure is going to be expensive,” he said. “If we were going to do it with aspirin we’d have done it by now. It may also carry with it a degree of risk, and researchers and patients may have to ask themselves how much risk they are prepared to tolerate if the result is going to be elimination of HIV.

“But it’s going to be a lot more affordable than lifelong antiretrovirals in resource-rich countries. As to whether it would be scalable for poor countries, though – ah, that’s a very different question.”