Research Briefs

CMV Vaccine Shows Impressive Control in Nonhuman Primates

While many researchers believe an AIDS vaccine should prevent acquisition of HIV, a recent study in nonhuman primates (NHPs) suggests it may also be possible to use vaccination to suppress the virus indefinitely following HIV transmission.

The study, led by Louis Picker, a professor of pathology at the Oregon Health & Science University, showed that 12 of 24 Indian rhesus macaques vaccinated with a replication-competent rhesus cytomegalovirus (rhCMV) viral vector vaccine candidate encoding the simian immunodeficiency virus (SIV)mac239 proteins Env, Pol, Gag, and Vpr/Vpx demonstrated early and complete control of viral replication for more than a year after repeat, homologous, low dose SIVmac239 challenge (1).

The study compared the immunogenicity of the rhCMV vaccine candidate in a four-arm trial involving 61 rhesus macaques previously exposed to CMV. Twelve macaques were given the rhCMV/SIV viral vector-based vaccine; 12 received an rhCMV/SIV vector-based candidate followed by a replication-defective adenovirus serotype 5 (Ad5) vector-based candidate encoding the full SIVmac239 genome; nine received a DNA prime/Ad5 boost (encoding the full SIVmac239 genome) regimen; and 28 control animals remained unvaccinated.

Nearly 14 months (59 weeks) after the initial vaccination, all 61 animals were challenged rectally, and while the study arms demonstrated no measurable differences in the number of challenges needed to infect the animals, the course of infection was markedly different in the different arms. Picker and colleagues found that after having plasma viral loads ranging from 60 copies/ml to 10 million copies/ml, 13 of the 24 macaques that received the CMV candidate, either alone or in combination with Ad5, showed complete control of SIV. And, despite one or two episodes of transient viremia, all but one of the 13 macaques sustained viral control for more than a year following challenge.

In contrast, 27 of the 28 unvaccinated control animals exhibited typical progressive SIV infection, as did all nine of the macaques that received the DNA/Ad5 prime-boost regimen.

Necropsy results from the CMV-vaccinated animals showed that SIV could rarely be found in the tissues of these animals. In 72% of specimens collected from four of the rhCMV/SIV vaccinated macaques, there was no evidence of SIV DNA or RNA in cells taken from the gut, lymph nodes, or other tissues. Picker compares the control achieved in the animals to that of human elite controllers or individuals whose viral loads are suppressed by antiretroviral therapy.

“I hesitate to say that the [rhesus macaques] cleared the virus, but there has never been an infected human or animal that has had that low a level of HIV or SIV before,” says Picker. “This is really unique.”

Ronald Veazey, a professor of pathology at the Tulane National Primate Research Center who was not involved in the study, was equally impressed with the results. He described the findings as “one of the most remarkable demonstrations of protection” that has been observed thus far.

Yet Veazey cautioned against over-interpreting the findings. “We know that persistent antigen at low levels seems to keep the immune system stimulated,” he says. “But if the virus levels diminish, the immune system dampens and quits fighting it. So I wouldn’t be surprised if some of those macaques currently controlling eventually progress to AIDS.”

Also, only half of the 12 rhCMV vaccinated macaques and seven of the 12 rhCMV/Ad5 vaccinated animals exhibited impressive control of viral replication, a finding Veazey found quite interesting. Previous work by Picker’s lab has shown that rhCMV/SIV induces effector memory T cells, which are better at protecting from challenge virus in mucosal tissues than central memory T cells that are most commonly induced by non-replicating vectors (seeResearch Briefs, IAVI Report, Mar.-Apr. 2009). He says the failure of some CMV-vaccinated animals to control infection in this latest study could be because they did not generate enough effector memory T cells early enough. The macaques that controlled SIV infection did not have protective major histocompatability complex alleles orTRIM5 polymorphisms associated with SIV control.

Picker and colleagues noted that the total SIV-specific CD8+ T-cell response to Gag and Pol antigens remained consistently high throughout the one year follow-up period. However, SIV-specific responses to Vif, an antigen not included in the vaccine candidate, which were initially as high as for Gag and Pol, waned over time, raising the intriguing possibility that the number of SIV-infected cells might be declining.

Picker and colleagues are now developing attenuated versions of the RhCMV candidate. One such candidate is now being evaluated in the fetuses of pregnant rhesus macaques. The attenuated Δpp71(rh110)RhCMV candidate lacks the rhesus CMV protein pp71 that is crucial for efficient viral replication. Picker says his laboratory is also looking at the immunogenicity of this attenuated vaccine in adult rhesus macaques to see whether weakening the vaccine also makes it less responsive to SIV. —Regina McEnery

1. Nature 473, 523, 2011

Correlates of Protection from SIV Challenge Identified in Monkeys

The immune responses that correlate with protection from HIV infection in humans are still elusive. But a recent study in Indian rhesus macaques allowed researchers to identify immune and genetic correlates of protection from challenge with simian immunodeficiency virus (SIV) after vaccination with a prime-boost vaccine regimen that is similar to the DNA/adenovirus serotype 5 (Ad5) prime-boost regimen currently being tested in HVTN 505, a Phase II trial conducted by the HIV Vaccine Trials Network (HVTN).

Norman Letvin, a professor of medicine at Harvard Medical School, and colleagues vaccinated 64 rhesus macaques with SIVmac239 Gag, Pol, and Env immunogens first delivered as three DNA injections, followed by an injection of Ad5 carrying the same immunogens. An additional 65 animals received a sham vaccination containing vaccine constructs without the SIV gene inserts. About four months after the boost, the animals were challenged rectally with up to 12 weekly, low-doses of SIVmac251 or SIVsmE660 (1). While SIVmac251 is very similar in sequence to the immunogens used in the vaccine, it is quite difficult to neutralize, whereas SIVsmE660 is more genetically different from the vaccine immunogens, but easier to neutralize.

The vaccine didn’t protect any of the animals challenged with SIVmac251. But about half of the animals challenged with SIVsmE660 were protected, and a low level of neutralizing antibodies to Env, and an Env-specific CD4+ T-cell response correlated with this protective effect. The vaccine was also more likely to protect monkeys with two alleles of the TRIM5 gene that restrict SIV replication than monkeys that had at least one permissive allele. “It is, I think, the first study large enough to allow us to dissect the correlates of immune protection and the first to demonstrate that a genetic trait can contribute to whether one sees or doesn’t see vaccine protection,” Letvin says. “I was not at all surprised that neutralizing antibody levels can contribute to protection against viral acquisition. What I was surprised by was the profound genetic effect on acquisition.”

The animals challenged with SIVmac251 did not have a major histocompatibility class I allele called Mamu A*01 that is known to be associated with control of viremia. All 20 vaccinated animals became infected after up to 12 repeat, low-dose rectal challenges, but showed a one to two log reduction of peak viremia compared with the 20 sham-vaccinated control animals, all of which were infected as well.

There were two groups of animals challenged with SIVsmE660, one with and one without Mamu A*01. In the Mamu A*01 negative group, 12 of the 25 vaccinated animals became infected after up to 12 repeat, low-dose rectal challenges, compared with 22 of the 25 sham-vaccinated control animals. In the Mamu A*01 positive group, seven of the 19 vaccinated animals became infected after up to 12 challenges, compared with 15 of the 20 sham-vaccinated control animals.

In both groups, slightly more than half, or 25 of the 44 vaccinated animals challenged with E660 were protected. But only in the Mamu A*01 positive group did the vaccinated animals that became infected have a lower peak viral load than the sham-vaccinated control animals that became infected, a finding that, Letvin says, underscores the importance of CD8+ T lymphocytes in the control of SIV and HIV replication once an infection has been established.

Because both the Mamu A*01 positive and negative vaccinated monkeys showed about 50% protection from E660 challenge, the researchers combined both groups for their analysis of correlates of protection. They observed that a very low neutralizing antibody titer could differentiate those that were protected from those that were not protected, Letvin says, which shows that “a neutralizing antibody response can mediate protection against acquisition of SIV. This is frank sterilizing protection.”

Letvin says that the restrictive TRIM5 alleles that were found to be a genetic correlate of protection made the animals more likely to be protected whether they received vaccine or placebo. “It’s sort of hard to infect those animals to begin with,” Letvin says. “If you then vaccinate, it becomes even more difficult to infect those animals.”

It’s unlikely that TRIM5 has any similar effects in humans, according to Letvin, because humans don’t show the same variability of the TRIM5 gene as rhesus monkeys. “The important take home [message] for humans is that a gene can contribute to protection or susceptibility to infection and that that can have a profound effect on vaccine efficacy,” Letvin says.

This study is the “first appropriately powered study that shows protection from acquisition by a prime-boost vaccine [in nonhuman primates],” says Louis Picker, a professor at Oregon Health & Science University, who was not involved in Letvin’s study, adding that it modeled the observations in RV144, the prime boost trial in Thailand that for the first time showed—albeit modest—protection from HIV in humans. “It shows that the monkey model can show what was observed in humans in RV144,” Picker says. “I’d be willing to bet we are looking at the same phenomena [in both] where you have an antibody response that’s relatively weak in terms of neutralization but still able to prevent acquisition.”

The study also suggests that challenge experiments with SIVsmE660 in monkeys need to account for possible protective effects of certain TRIM5 alleles, Picker says. “With 660 you have to take that into consideration,” Picker says. “It also helps us go back and interpret other 660 experiments.”

While Letvin and colleagues didn’t observe any CD8+ T-cell responses as a correlate of protection in their study, Picker has been working on a replicating rhesus cytomegalovirus (rhCMV) vector vaccine candidate that induces CD8+ effector memory T cells, which can control viral replication to undetectable levels (see CMV Vaccine Shows Impressive Control in Nonhuman Primates, above). Picker says that the antibody vaccine approach by Letvin and colleagues and the CMV vaccine approach complement each other and could be combined. “They conceivably could work together,” he says. “I imagine there would be synergy.” —Andreas von Bubnoff

1. Sci. Transl. Med. 3, 81ra36, 2011

Researchers Identify Host Restriction Factor that is Target of Vpx

It has been known for some time that HIV-1 cannot replicate in certain cells such as dendritic cells (DCs), and that HIV-1 replication in macrophages is not very efficient. In contrast, HIV-2 and certain strains of simian immunodeficiency virus (SIV) can productively infect these cells because they have a protein called Vpx. Researchers have suspected that the role of Vpx was to counteract an unknown cellular host restriction factor that keeps HIV-1 from replicating in such cells, which in turn keeps DCs from generating innate immune responses to HIV-1.

Now, two research groups, one led by Monsef Benkirane at the Institut de Génétique Humaine in Montpellier, France, the other led by Jacek Skowronski, a professor of molecular biology and microbiology at Case Western Reserve University, have identified a protein called SAMHD1 as the cellular restriction factor that is targeted by Vpx.

Benkirane and colleagues identified SAMHD1 in a human cell line called THP-1, which, once treated with certain chemicals, becomes more permissive for HIV-1 infection only if Vpx is added. To identify the host restriction factor, Benkirane and colleagues expressed sooty mangabey Vpx in these cells, purified Vpx and the proteins that bound to it, and identified the proteins by mass spectrometry (1).

They found about 50 proteins that bound to Vpx, but when the researchers saw that SAMHD1 was among the proteins, they knew it must be the right factor, Benkirane says. “When we saw this protein, we knew that it’s going to be that protein because of the little we knew about this protein,” he says.

Precisely how SAMHD1 restricts HIV-1 replication isn’t known yet, but researchers believe it might do so by degrading viral DNA because mutations in SAMHD1 can lead to Aicardi-Goutières syndrome (AGS), in which excess nucleic acid accumulation in cells is thought to lead to inflammatory immune responses. In addition, both Benkirane’s and Skowronski’s group found that inhibiting SAMHD1 in HIV-1 infected cells leads to a increase in HIV-1 DNA.

“[SAMHD1] is exactly the kind of molecule that you might suspect would be involved because it’s a molecule that restricts the synthesis of viral DNA,” says Dan Littman, an investigator at the Howard Hughes Medical Institute at the Skirball Institute at New York University School of Medicine, who was not involved in the most recent Vpx studies. SAMHD1 seems to have a similar biological role to the protein TREX1, Littman adds, in that TREX1 mutations can also cause AGS. TREX1 degrades HIV-1 DNA in infected cells, thereby helping HIV-1 to avoid inducing an innate immune response in infected CD4+ T cells and macrophages, according to a study conducted last year by Judy Lieberman of Harvard Medical School and colleagues (see Research Briefs, IAVI Report, Sep.-Oct. 2010).

To confirm that SAMHD1 was indeed the elusive factor that restricts HIV-1 infection, Benkirane and colleagues showed that inhibiting SAMHD1 expression in human DCs by siRNA made the DCs fully infectable by HIV-1, and expression of SAMHD1 in cells that don’t normally express it inhibited HIV-1 infection of these cells.

While Vpx proteins from some SIV strains can overcome restriction of HIV-1 replication in DCs, others cannot, and the researchers showed that only the Vpx proteins from SIV strains that can overcome restriction can induce degradation of SAMHD1 in THP-1 cells. “We showed that there is really a strict correlation,” Benkirane says.

In a separate study, Skowronski and colleagues identified SAMHD1 as the target of Vpx and found that Vpx not only binds to SAMHD1, but that the Vpx-SAMHD1 protein complex also binds to a protein called DCAF1 that targets proteins for degradation (2). This suggests that Vpx induces degradation of SAMHD1 by bringing SAMHD1 in contact with DCAF1. “We know how Vpx disposes of SAMHD1,” Skowronski says.

Skowronski’s group also found that Vpx depletes SAMHD1 in human DCs and macrophages, and that SAMHD1 is also required for HIV-1 restriction in macrophages. Together with the findings by Benkirane and colleagues, this suggests that SAMHD1 is responsible for restricting HIV-1 infection of both human DCs and macrophages.

But while SAMHD1 is clearly required for HIV-1 restriction, it’s not always sufficient, Skowronski says. Many cell types, including a fraction of macrophages and perhaps even activated CD4+ T cells, the main target cells of HIV-1, don’t restrict infection even though they express SAMHD1. “Clearly SAMHD1 is not sufficient to block infection, so what else is there?” asks Skowronski.

The inability of HIV-1 to infect cells like DCs enables HIV-1 to go undetected by the innate immune system, says Littman, whose group showed last year that infecting DCs with HIV-1 in the presence of Vpx can induce innate immune responses (see Research Briefs, IAVI Report, Sep.-Oct. 2010). The identification of SAMHD1 therefore suggests that perhaps inhibiting SAMHD1 could now lead to better vaccines or treatments for HIV by improving the innate immune response to the virus. “If we can figure out ways of manipulating this, for example by blocking the activity of SAMHD1 either in people who are being vaccinated with a replication-defective or an attenuated type of virus, or in people who are already infected, that could lead to a more effective immune response against the virus,” Littman says.

Benkirane is now testing whether the effects of vaccination can be improved by inhibiting SAMHD1 in dendritic cells in humanized mice. —Andreas von Bubnoff

1. Nature 474, 654, 2011
2. Nature 474, 658, 2011