Research Briefs

CMV-based vaccine elicits a new kind of T-cell response in macaques

By Andreas von Bubnoff

A recent study could change much of our understanding of CD8+ T-cell responses and give vaccine developers new tools to manipulate them.

The current understanding of these responses goes something like this: The body makes millions of CD8+ T cells, each of which expresses a different T-cell receptor. Each of those receptors can recognize a distinct set of peptides that is presented by MHC class I molecules on an infected target cell. When that happens, the CD8+ T cells multiply and develop into effector cells that kill the infected cells or perform other antiviral functions. CD4+ T helper cells, by contrast, are activated by peptides that are presented by MHC class II molecules. Each person has a different set of variants or “alleles” of MHC class I and II molecules. Each can only bind and present certain kinds of peptides. So different people can only make CD8+ T-cell responses to a small subset  of the peptide sequences a pathogen contains.

Now a study by Louis Picker of Oregon Health & Science University and colleagues challenges these dogmas. Previously, Picker and his colleagues found that half of rhesus macaques vaccinated with a cytomegalovirus (CMV) derived vector called rhCMV 68-1 that expresses SIV genes could suppress viral load to undetectable levels after repeat rectal challenge (Nature 473, 523, 2011).

In the current study, they report that two thirds of the CMV vaccine-induced CD8+ T-cell responses are not directed at peptides presented by MHC class I, but by MHC class II. These responses are also directed at peptides that are completely different from the “canonical” peptide targets of the CD8+ T-cell responses to SIV infection itself or to other types of vaccines that have been used to deliver SIV genes, including adenovirus vector vaccines. The responses are about three to four times broader than what’s usually observed after vaccination with these other vaccines (Science 2013, doi: 10.1126/science.1237874).

“It’s paradigm breaking,” Picker says. “It means that there is a lot more flexibility to [CD8+] T-cell responses than we think. Nothing has been seen before as promiscuous as these responses.” Indeed, the responses covered about two thirds of the entire SIV Gag protein, and the responses to some Gag peptides were so common that they were observed in all of about 100 vaccinated monkeys tested. 

Apparently, the CMV 68-1 vector itself manipulates the host immune response to behave in such an unusual way: One rhCMV gene inhibits the MHC class I type responses to the “canonical” peptides, while another set of genes enables the broader responses to the  unconventional peptides to occur, including the responses to the class II presented peptides. “This is the first time any agent has been found that controls its own recognition by CD8+ T cells,” Picker says.

The value of the study, Picker says, lies in the fact that it adds to the vaccine developer’s tool box. It might now be possible to make vaccine vectors that modify CD8+ T-cell responses at will.

Both the better breadth and the fact that the peptide targets of the responses differ from the targets to natural SIV infection would make it much harder for the virus to escape from CMV-vaccine-elicited responses, which makes this type of vaccine one of the best current candidates for a therapeutic vaccine for HIV, Picker says.

In a related commentary that appeared in the same issue of Science, Nilu Goonetilleke and Andrew McMichael of Oxford University—both not involved in Picker’s study—wrote that the findings “could not be more timely,” given the recent termination of HVTN 505. That trial used a DNA/Ad5 vaccine regimen, one of several failed HIV candidate regimens that induce the classical narrower MHC class I based CD8+ T-cell responses.

The findings are a “big advance to the field” in terms of the understanding of T-cell immunology in the context of host-pathogen interactions, says Tom Hope of Northwestern University, who was not involved in the study. “It is rather remarkable,” he says, “how the CMV vector can change the recognized T-cell epitopes.” Still, he adds, the study doesn’t clarify which part of these unusual responses helped protect 50% of the vaccinated animals, and why the other half of the animals could not control viral load. “What is not clear is how [the findings relate] to the observed protection,” Hope says.

The reason that the study couldn’t identify the correlates of the observed viral control in half the animals, Picker says, is that it’s difficult to check for immune responses in the tissues relevant for protection such as the mucosa, something Picker and colleagues are doing now. However, Picker adds, all vaccinated animals showed these responses, regardless of whether they were later challenged (and protected) or not, and these were the only immune responses that could be detected in the vaccinated animals. To Picker, this suggests that these responses must be relevant to the observed protection.

So why couldn’t some of the animals control the virus, even though all vaccinated animals showed these unusual immune responses? One explanation, says Picker, could be that by chance, the virus might have infected some animals at sites with too few vaccine-induced virus-fighting effector memory CD8+ T cells to contain it.

One concern is that because the induced responses are so broad, the vaccine might induce autoimmune responses. Picker says that that’s something to watch for, but adds that so far, there are no signs of autoimmune responses in any of his vaccinated rhesus macaques.

Next, Picker wants to find out how CMV manipulates the CD8+ T-cell response. He has also been working on human CMV vectors that are attenuated to reduce any risks associated with their use in immune-compromised individuals, because the vaccine is based on a replicating vector that permanently expresses its immunogens. “I hope,” he says, “to have a phase I clinical trial maybe in the third year from now."