Engineered T-cells may kill off reservoir cells, mimicking elusive CD8 vaccine effect

Killer T-cells (CD8 cells or cytotoxic T-lymphocytes) engineered to mount a rare and specific reaction to HIV-infected cells that was first seen in an experimental vaccine more than a decade ago were able to halve the number of cells containing intact viral DNA, the 13th International AIDS Society Conference on HIV Science (IAS 2025) in Kigali heard yesterday.

“We’re very excited about our findings,” Professor J. Victor Garcia of the University of Birmingham, Alabama told the conference.

Background – the vaccine that cured monkeys

Although the experiment was in mice, the reason why the result is a possible breakthrough becomes clear if you consider research done over a decade ago. A novel vaccine that generated a strong CD8 response to SIV, the monkey equivalent of HIV, did not prevent a group of animals becoming infected – but did appear to cure half of them, with viral loads vanishing in 48 out of 96 monkeys over the following three years. The vaccine acted like a cure.

This is not necessarily a surprise. The majority of candidate vaccines are designed to elicit antibodies. Antibodies are the gatekeepers of the body, protecting cells from invasion by viruses and other invaders. A vaccine that elicits antibodies acts like a ‘trial run’ of a real infection, causing these small Y-shaped proteins to be manufactured in their billions, where they either directly neutralise invaders or recruit other parts of the immune system that will.

Antibodies may not totally prevent viruses from infecting cells. This may not matter that much with a virus like SARS-COV-2, where blunting a short-lived infection is usually enough to stop it overwhelming the immune system. But in HIV even one HIV-infected cell can give rise to a new wave of HIV years later.

This is why we need a CD8 vaccine too. If antibodies are the uniformed branch of the immune system, CD8 cells are its detectives. They alone are able to find and destroy cells that secretly harbour viruses – alerted by epitopes, cellular ‘distress signals’ consisting of bits of viral proteins that only CD8 cells can read. In the case of a slow-acting virus like HIV, a CD8 vaccine might eliminate an established infection gradually, but maybe completely.

The original CD8 vaccine achieved its unusual result, it was thought, because it had an unusual vector – a vector being the shell of another virus that can get inside cells, then exposes them to the HIV components it needs to sensitise them to. This vector was CMV (cytomegalovirus) – a virus that most people have been infected by, but which only causes trouble in people with impaired immunity.

It was found that the CMV vector gave rise to a very unusual kind of CD8 response.

Viral epitopes are ‘presented’ to CD8 cells by HLA (human leucocyte antigen) molecules, cell-surface molecules that act like display cases. They are a tremendously variable group of molecules, probably because they have to be in order to compete in the arms race between the immune system and pathogens. The ones that display epitopes form three broad classes, A, B and C. The CD8 cells that respond to them are very variable too and are often only active against narrow strains of virus which, in the case of the hyper-variable HIV, is a problem.

CMV, however, is a virus almost as tricky as HIV, and it was forcing the proliferation of a different kind of CD8 cell that only responded to an HLA called HLA-E – so-called HLA-E restricted cells. HLA-E is not variable – its genetic sequence differs by only one ‘letter’ in all populations with European ancestry. CD8 cells attuned to HLA-E are rare, but appear able to respond to a wide variety of viral strains.

The tale of human trials of the CMV-vector CD8 vaccine has been a frustrating one, however, with many failures. Eventually – many years after those original experiments in monkeys – it was found that the CMV vector originally used, fortuitously, had a very specific genetic profile; it lacked a set of genes that CMV normally uses to suppress the immune response against it.

Engineering HLA-E restricted CD8 cells

CMV vaccines correcting for this are in clinical trials, but Garcia and his colleagues thought: why not do away with the CMV vector altogether? What if it was possible to artificially expand a population of HLA-E receptive CD8 cells and then re-infuse them, as a ready-made immune response, back into the body?

They took T-cells from mice and induced a small proportion of them already attuned to HLA-E to respond specifically to two particular HIV epitopes.

They expanded the proportion of cells that was HLA-E restricted and then infused them into humanised mice, who were then infected with HIV. The number and proportion of CD8 cells in the mice expanded over a two-week period before, by four weeks after this single infusion, declining to levels seen in mice not given the cell infusions. The CD8 cells particularly homed in on the bone marrow and lymph nodes, where you’d want them to be, with 35 times the level seen in control mice.

Starting two to three weeks after infusion, the HIV viral load in the blood of the CD8-infused mice fell by 80%, as did HIV RNA levels in tissues. The mice were euthanised at day 28, so we don’t know if their viral load would have continued to remain lower than in control mice, or even decline further.

In a second experiment, four doses of HLA-E restricted CD8 cells were given weekly to mice that already had HIV and were on antiretroviral therapy. Because they were already virally suppressed, there was only a slight further fall in viral load in the tissues of the infused mice, compared with control mice, which was not statistically significant.

The total integrated HIV proviral DNA in the mice’s cells was also measured and this was not different to levels in control mice. But what was significant was that the amount of intact proviral DNA, capable of forming new viruses, was nearly halved, and this was significant. This suggests that the HLA-E restricted CD8 cells were able to detect and kill off cells producing HIV – a first step towards reducing the ‘reservoir’ of HIV-infected cells or even towards a cure.

These results were strong enough to elicit some surprise in the audience. Veteran HIV cure Professor Steve Deeks commented that he had not previously seen evidence that HLA-E-restricted CD8 cells could in themselves lead to falls in HIV RNA in blood. But Victor Garcia commented that these only made him more convinced of the promise of this approach, especially as it was potentially applicable to many strains of HIV.

“HLA-E based cell therapy, with its ‘universal’ nature, holds significant potential as a novel approach to curing HIV,” he commented.