Monkey Matters

Nonhuman primate researchers gathered in New Orleans to discuss the growing number of models for HIV infection and pathogenesis

By Andreas von Bubnoff

Almost 200 participants gathered for this year’s 28th Annual Symposium on Nonhuman Primate (NHP) Models for AIDS, held October 19-22 in New Orleans. The meeting was scheduled to take place here three years ago but was relocated to another primate center because of the devastating effects Hurricane Katrina had on the city.

Andrew Lackner, director of the Tulane National Primate Research Center (TNPRC), which hosted the conference, said New Orleans has rebounded quite strongly. “There are actually more restaurants in the city than there were before Katrina,” he said. In addition to the opportunity to savor the fine restaurants of the Big Easy, attendees were also served a diverse selection of research updates, including insights from using NHPs as models for many different aspects of HIV infection and pathogenesis, as well as genetic characterization of NHP hosts.

Monkey models of transmission

An increasing number of NHP models are now available that can mimic different aspects of HIV transmission, and these models are important for testing vaccine, microbicide, or drug candidates.

An important component of these NHP models is the choice of challenge virus, and one characteristic of the challenge virus that is useful to understand is how sensitive it is to neutralization by antibodies.

To accurately predict the immunogenicity of candidate vaccines in challenge experiments, a challenge stock should ideally be moderately sensitive to neutralization by antibodies, said Katharine Bar, an assistant professor for infectious diseases at the University of Alabama at Birmingham. A challenge stock that is too resistant to neutralization—as appears to be the case for the commonly used challenge stocks simian immunodeficiency virus (SIV)mac239 and 251—might make it difficult to show protection even with good vaccine candidates, she said. Conversely, if a challenge stock is too sensitive to neutralization, a candidate vaccine that shows protection in animal experiments might not be protective when tested in humans.

Bar reported on the neutralization sensitivity of SIVsmE660, a challenge stock for which the neutralization sensitivity is still poorly characterized. “Characterizing the neutralization sensitivity of this virus will be helpful in figuring out if it’s an appropriate challenge,” said Bar. “Just understanding what we are dealing with is important for us to figure out how to interpret the results of past studies, and then going forward how to create a model that most closely recapitulates HIV-1.” SIVsmE660 is similar to HIV in that it is a swarm, which means it contains many different virus variants. Bar and colleagues used single genome amplification (SGA) to isolate different variants from SIVsmE660 stocks, as well as the transmitted founder viruses of animals rectally infected with SIVsmE660. They then determined the neutralization sensitivity of several variants to monoclonal antibodies directed to different parts of the SIV Env protein, and to serum taken from macaques that were infected with SIVsmE660. Preliminary results suggest that most variants are rather sensitive to neutralization, while others show intermediate or high resistance. This suggests that SIVsmE660 is a mix of variants with varying degrees of resistance to neutralization, and could explain some of the variable pathogenicity of this challenge stock after low-dose mucosal challenge.

Researchers are also trying to develop animal models of different viral transmission routes. Brandon Keele, head of the viral evolution and genomics core at the National Cancer Institute, gave an update on the development of an NHP model of penile HIV transmission he is working on in collaboration with Chris Miller at the University of California, Davis. In the model, the penises of male rhesus macaques were exposed to different doses of SIVmac251 by immersing their flaccid penis in a virus solution for up to 20 minutes; in some cases, the researchers also put a small amount of the solution onto the foreskin at the tip of the penis. According to Keele, SGA showed that all seven animals that were infected this way with three different doses (the lowest dose required multiple challenges) had just one transmitted founder virus, indicating this model of penile infection allows for single variant infections like most of the heterosexual transmissions of HIV in humans, Keele said.

At similar doses, vaginal infection of rhesus macaques is more likely to result in several transmitted founder viruses, the number of which is more likely to vary, Keele said. One possible reason for the difference could be that female rhesus macaques may be in different stages of their menstrual cycle. It is also possible that when immersing a penis in a virus solution, not all of the viruses in the solution touch the penis, whereas solution put inside the vagina might have more time to interact with the inner surface of the vagina. “You need a lot of virus to infect in the penile route,” Keele said, “which means that the penile mucosa is a good anatomical barrier to infection and recapitulates human infection because you get a single infection, but it’s difficult because you have to use a lot of virus.”

Co-infection models

NHP models are also useful in modeling co-infection of HIV with other pathogens. A better understanding of HIV/tuberculosis (TB) coinfection in humans is especially important because TB is thought to be a leading cause of death among people with HIV/AIDS (see Deadly Synergy, IAVI Report, Sep.-Oct. 2009). In many cases, TB becomes latent when the TB bacteria hide in granulomas in places such as the lungs, but HIV infection, by weakening the immune system, can lead to reactivation of latent TB infection, said Smriti Mehra, a research scientist at the TNPRC, adding that she and her colleagues are developing the first HIV/TB coinfection model in rhesus macaques.

Mehra and her colleagues were able to induce SIV-induced reactivation of TB in rhesus macaques. They infected the macaques by letting them breathe in tuberculosis bacteria, which is how humans become infected. Once the infection had become latent for nine weeks, the researchers infected the animals intravenously with SIVmac239. They found that the animals developed clinical signs of TB within two weeks. The researchers want to use these animals to study the gene expression in the granulomas in the lungs and see which genes are expressed when TB is reactivated.

NHP models of HIV co-infection with sexually transmitted infections (STIs) are also important because STIs are associated with an increased risk and rate of HIV infection. Tara Henning, a postdoctoral fellow at the US Centers for Disease Control and Prevention (CDC), reported that she and her colleagues did the first successful triple infection of female pigtail macaques (PTMs) with the SIV/HIV hybrid SHIVSF162P3, the single-celled protozoan parasite Trichomonas vaginalis, and the bacterium Chlamydia trachomatis, with clinical presentation of genital STI symptoms that were similar to those observed in humans. An NHP model of STI-HIV coinfection would make it possible to investigate how STIs might contribute to enhanced susceptibility to HIV infection, and to test prevention strategies that target the genital mucosa in the context of STIs, Henning said.

Clues about pathogenesis

NHPs are also studied as a model of the pathogenic effects of HIV. One important hallmark of HIV infection is chronic immune activation. Structural damage to the gut is thought to contribute to this immune activation because of translocation of microbial products from the gut into the blood. Although the underlying mechanisms of this are still unclear, some clues are coming from studying the pathogenic effects of SIV infection in NHPs.

Nichole Klatt, a postdoctoral fellow in the lab of Jason Brenchley at the National Institute of Allergy and Infectious Diseases, reported on the identification of NK17 cells, a type of cell in the gut that can produce interleukin (IL)-17, a cytokine that is concentrated in mucosal tissues and produced in response to bacterial and fungal antigens. These cells had not been described before.

Klatt showed evidence that in SIV-infected rhesus macaques, loss of IL-17 producing CD4+ T cells and CD8+ T cells in the colon is associated with damage to the colon epithelium and with immune activation. However, loss of NK17 cells actually shows the strongest association with damage of the colon epithelium and with immune activation. This suggests a role for these IL-17 producing cells in maintaining the structural barrier of the gastro-intestinal tract and in preventing chronic immune activation, Klatt said.

Loss of NK17 cells in the gut might also play a role in causing the bacterial translocation and chronic immune activation that is often observed in PTMs, even if they are not SIV infected, according to Klatt (Mucosal Immunol. 3, 387, 2010). Chronic immune activation is thought to be one reason why PTMs that are infected with SIV, such as SIVmac239, progress to disease more quickly than rhesus macaques. Klatt reported that in SIV-uninfected PTMs, only loss of NK17 cells, but not of IL-17 producing CD4+ and CD8+ T cells, was significantly correlated with damage of the colon epithelium. This suggests an important role for NK17 cells in the maintenance of the tight epithelial barrier of the colon, Klatt said. “Taken together, this study really is one of the first potential mechanisms showing why or how this damage may possibly occur during SIV pathogenesis,” she added.

Controlling infection

Researchers are also studying NHPs infected with live-attenuated versions of SIV as a model for how the immune system can control viruses like HIV. One live-attenuated SIV that has been studied for many years is SIVmac239∆nef, which has a deletion in the nef gene. Andrea Jordan, a research specialist in the laboratory of James Hoxie, a professor of medicine at the University of Pennsylvania, reported on studies with another live-attenuated version of SIVmac239 called ∆GY, where just two amino acids are deleted from the cytoplasmic tail of the Envelope protein.

Given that PTMs progress to disease more quickly than rhesus macaques when infected with SIVmac239, it was surprising that Jordan reported that PTMs infected with ∆GY control this live-attenuated virus better than rhesus macaques. The researchers found that all nine PTMs they had intravenously infected with ∆GY had undetectable or extremely low viral loads for several months to years, whereas in ∆GY-infected rhesus macaques, viral loads were low, but still detectable in all cases. In addition, ∆GY-infected PTMs were not only protected from homologous intravenous challenge with SIVmac239, but could also bring viremia from a heterologous intravenous challenge with SIVsmE660 to very low or undetectable levels. In contrast, SIVmac239∆nef only provides partial or poor control from intravenous heterologous E660 challenge in rhesus macaques, Hoxie said.

This suggests that ∆GY-infected PTMs could serve as an interesting model system to identify immunological correlates of protection. “We have a very interesting puzzle here why an animal species that is prone to early deaths and AIDS faster than the rhesus appears to be our preferred model for complete control and protection with ∆GY,” Hoxie said. “We’ve made a very small change and have totally changed pathogenesis. We’ve rendered a highly pathogenic virus to be totally controllable.”

When the researchers depleted the CD8+ T cells in the ∆GY-infected PTMs, they found that this led to reappearance of ∆GY, after two years in which ∆GY RNA was undetectable in plasma, suggesting that cellular immunity has a role in the protection, although it’s also possible that antibody-dependent cellular cytotoxicity (ADCC) plays a role, Jordan and Hoxie said. Hoxie said the ∆GY-infected PTMs also resemble elite controllers, HIV-infected individuals who control virus replication below detectable levels without treatment, because ∆GY initially replicates similar to wild type levels of SIVmac239 but is later rendered controllable by host immune responses.

Cristian Apetrei, an associate professor of microbiology and molecular genetics at the University of Pittsburgh, described the development of another NHP model of elite controllers. Apetrei and colleagues intravenously infected five Indian rhesus macaques with SIVagm, the SIV that naturally infects African green monkeys without causing disease. The SIVagm-infected rhesus macaques resembled human elite controllers, in that they had undetectable viral loads for four years, completely recovered levels of mucosal CD4+ T cells by four years after infection, and no disease progression. Depletion of CD8+ T cells four years after infection resulted in a transient rebound of viral loads, suggesting that SIVagm was controlled by the immune system. Apetrei said that having such a model is important because it makes it possible to study how elite control develops early after infection.

SNPs and genomes

Another theme of this year’s meeting was efforts to genetically characterize NHP hosts. When it comes to the availability of genetic tools, NHPs have a long way to go compared to mice or fruit flies, according to Jessica Satkoski Trask, a postdoctoral research associate at the University of California, Davis. “If you want a knockout mouse, you call the knockout mouse store and they send you the mice,” she said, referring to mice that have a gene knocked out. In contrast, “rhesus macaques and primates are not set up as a genetic model yet. There is going to be a need for a more and more diverse set of genetic tools.”

One such tool is called single nucleotide polymorphisms (SNPs), variations in the genetic code that geneticists use to find candidate genes for certain genetic traits or phenotypes. This could be important to identify, for example, the genetic factors that are the basis for different responses of Chinese and Indian rhesus macaques to infection with SIV.

Traditionally, most studies that use NHPs as a model for HIV infection have used Indian rhesus macaques, and Chinese rhesus macaques are also gaining in popularity, Satkoski said. However, she said, until recently just a few hundred SNPs were known for Chinese and Indian rhesus macaques, while over 25 million are known for humans and about 14 million for mice. Now Satkoski and colleagues have identified more than 4,000 additional SNPs in Indian and Chinese rhesus macaques that are unique and consistently observed with different frequencies in the two groups. They also made a linkage map, which means that they determined where those SNPs are located on the chromosomes and which SNPs are likely inherited together.

The availability of more SNPs on a linkage map will accelerate the process of finding candidate genes for inherited traits, Satkoski said. She and her colleagues have already used their more detailed SNP linkage map to identify some candidate genes, including one that likely has an impact on elite controller status in Chinese rhesus macaques.

The newly identified SNPs will also allow researchers to screen animals that are going into research protocols, allowing them to determine the degree to which an animal is really a Chinese and not an Indian rhesus macaque. In the future, it may also be possible to use the new SNPs to determine the likelihood that a Chinese rhesus macaque is going to be an elite controller. “[This] will allow you to have a more uniform research population,” Satkoski said.

Chinese rhesus macaques might be the more relevant model for AIDS in humans than Indian rhesus macaques because they don’t progress to AIDS as quickly, said Bianca Mothé, an associate professor of biology at the California State University San Marcos. In addition, a larger percentage of Chinese rhesus macaques become elite controllers, Mothé said. One possible explanation for the different disease progression in Chinese rhesus macaques could be their MHC class I alleles. MHC class I alleles are important for the cellular immune response because antigen presenting cells use MHC class I to present antigens to CD8+ T cells, but so far these MHC alleles have not been well characterized in Chinese rhesus macaques.

Mothé and colleagues sequenced the MHC class I alleles of 50 Chinese rhesus macaques from different primate centers and found that Chinese rhesus macaques have a more varied MHC composition than Indian rhesus macaques. In addition, when Mothé and colleagues predicted the peptides that can be bound by these MHC alleles, they found that three of the four most frequent MHC alleles in Chinese rhesus macaques bind the same peptides as the three most common HLA motifs in humans. Together, this suggests that Chinese rhesus macaques are more similar to humans in their MHC class I alleles than Indian rhesus macaques, which may explain why Chinese rhesus macaques are more similar to humans than Indian rhesus macaques in their disease progression to AIDS (Immunogenetics 62, 451, 2010).