The immunological basis of susceptibility to nontyphoidal Salmonella bacteraemia among HIV-infected adults in Uganda

Nontyphoidal strains of Salmonella are a major cause of fatal bloodstream infection in Africa, particularly among young children and people living with HIV/AIDS. These Salmonella infections are difficult to diagnose and antibiotic resistance is an increasing problem, so an effective vaccine against nontyphoidal Salmonella (NTS) has the potential to save many lives. Vaccines currently in development against NTS aim to work by stimulating the production of antibodies to a sugar known as O-antigen which is part of the lipopolysaccharide (LPS) molecule and found on the Salmonella surface. My previous research has shown that some HIV-infected adults in Africa have very high levels of these antibodies which then block killing of the Salmonella bacteria. Hence there is a risk that these vaccines may cause harm to people with HIV.

This project aims to understand the basis for susceptibility to NTS bloodstream infections in HIV-infected Africans, and in particular to understand the phenomenon of blocking antibodies, with a view to designing improved vaccines that will protect HIV-infected groups against NTS. I hypothesise that NTS is present at increased levels in the gut in HIV infection and that because the gut wall does not function properly in these patients, the Salmonella bacteria are able to invade the bloodstream. Here they can go on to cause severe disease and death due to the inability of the immune system to kill these bacteria, because of the presence of blocking levels of antibodies. I hypothesise that the continual movement of Salmonella and Salmonella LPS into the circulation drives the production of these high levels of blocking antibodies, because antibody production is no longer appropriately regulated in HIV infection.

I will test this hypothesis in a group of adults newly-diagnosed with HIV infection in Uganda, studying them for two years after they commence antiretroviral therapy (ART) and comparing them with HIV-uninfected subjects. Incidence of NTS bloodstream infections initially rises when starting ART and then falls, but remains much higher than in HIV-uninfected subjects. Therefore, looking at specific immune parameters with time on ART will help understand which are most relevant to susceptibility to NTS infection. Since compromise of the barrier state of the gut is key to my hypothesis, I will study samples of gut tissue taken at endoscopy, as well as peripheral blood.

First, I will test to see whether NTS is present in the gut of HIV-infected participants and whether it occurs more frequently and at higher levels than in HIV-uninfected controls. Then I will investigate the transfer of Salmonella from the gut to the bloodstream and compare this with the immune status of the gut wall, seeing whether this transfer is associated with loss of particular groups of immune cells known as Th17 and mucosal-associated invariant T (MAIT) cells. Finally, I will study the levels and function of antibodies targeting LPS O-antigen, to determine whether they block or kill Salmonella, comparing them with antibodies to a protein known as flagellin which is also present on the Salmonella surface. I will test whether levels and function of these antibodies associate with changes in the numbers and function of two further groups of immune cells, Tfh and Treg cells, which are infected by HIV and which regulate antibody production.

I will follow changes in these immune parameters in the group of HIV-infected participants over time on ART, and look for associations at time points with known differences in susceptibility to NTS. This will enable me to better understand which factors are most relevant for susceptibility to NTS, the mechanism of susceptibility and production of blocking antibodies, and what vaccine or other therapeutic strategies might best protect HIV-infected populations against NTS.

Nontyphoidal Salmonellae (NTS) cause fatal bacteraemia among HIV-infected Africans. Surprisingly, susceptibility initially increases on antiretroviral therapy (ART) and, although subsequently diminishes, it remains over 50-fold higher than in HIV-uninfected Africans. Candidate vaccines aim to induce antibodies against Salmonella lipopolysaccharide (LPS) O-antigen. Previously, I found that these antibodies occur in excess in some HIV-infected Africans and block killing of NTS, raising safety concerns.

Blocking antibodies may represent one of a number of factors contributing to susceptibility. I propose an integrated hypothesis that NTS bacteraemia in HIV-infected Africans occurs due to:
1. high levels of NTS in the gastrointestinal (GI) tract,
2. translocation of NTS to the circulation secondary to loss of CD4+ Th17 cells and mucosal-associated invariant T (MAIT) cells,
3. dysregulated production of anti-O-antigen antibodies, secondary to translocated LPS and Tfh cell dysregulation; these antibodies are unable to kill translocated bacteria, resulting in bacteraemia.

The experimental approach exploits the changes in susceptibility to NTS with ART and investigates blood, GI mucosa collected by endoscopy, and stool in a cohort of treatment-naïve HIV-infected adults in Uganda, immediately before and for 24 months after initiating ART, in comparison with HIV-uninfected adults. I will examine Salmonella in the GI tract by metagenomics, and its translocation to the circulation by 16S qPCR, and immunological and biochemical markers. Concentration and function of blocking antibodies will be determined by ELISA and bactericidal assays, while frequency and function of T cell subsets, focusing on CD4+ Th17 and Tfh, and MAIT cells, will be explored by flow cytometry. I will test for association between these parameters and findings will be correlated with time on ART. The detailed knowledge obtained will help guide the development of appropriate vaccines and therapies.

The Global Health Vaccine Industry
By seeking to understand the immunological basis of susceptibility to NTS bacteraemia in HIV-infected Africans, the study addresses a key question for the global health vaccine community engaged in developing vaccines against NTS, including the GSK Vaccines Institute for Global Health, Siena, and the Center for Vaccine Development, University of Maryland. NTS bacteraemia in African adults is a major cause of morbidity and mortality. Following our previous finding of excess blocking levels of antibodies to Salmonella LPS O-antigen in some HIV-infected Africans, there is major uncertainty as to whether NTS vaccines currently in development and designed to induce such antibodies with be beneficial, harmful or have no effect in HIV-infected groups. Study findings will help indicate whether an alternative vaccine approach is required and/or possible to protect HIV-infected populations against NTS. With current vaccines in development due to be tested in clinical trials during the lifetime of the fellowship, this project is particularly timely and would have immediate impact. An improved vaccine approach that protects HIV-infected people, as well as immune-naïve children, will have a significantly broader impact in relation to global burden of disease than vaccine approaches that only protect the latter group.

Health policy makers
As the research will inform global health policy makers concerning the potential of NTS vaccines in development for protecting HIV-infected individuals against invasive NTS (iNTS) disease, it will help indicate what high-level direction is indicated regarding the development of alternative vaccine strategies. Our findings will help understand why HIV-infected patients in Africa become more susceptible to iNTS during the first three months on ART, and why, following immune reconstitution, they remain far more susceptible than HIV-uninfected adults. This is important for guiding strategies other than vaccination, such as nutritional and anti-microbial interventions, to prevent iNTS disease in this population and will have implications for understanding other causes of morbidity and mortality among HIV-infected subjects receiving ART.

Clinical researchers
The work will be of interest and benefit to clinical researchers spanning disciplines of immunology, microbiology, infectious disease, gastroenterology and HIV medicine, as well as vaccinology, as described in the 'Academic beneficiaries' section.

HIV-infected population
The HIV/AIDS community takes great interest in clinical research related to HIV infection and the work will be of interest to this group, and will help inform, in addition to the potential benefit of vaccines in development, what self-action may be beneficial in relation to minimising risk of Salmonella disease and which HIV-infected sub-groups of patients are most susceptible to NTS. As mentioned above, translation of the findings of this project into improved NTS vaccine design has the potential, over around a ten year time frame, to have major impact in terms of decreased morbidity and mortality from NTS disease among Africans living with HIV/AIDS.

Project staff
Finally, a project with this broad cross-discipline approach, and study of clinical material in the developing world, provides excellent opportunities for members of the study team in Oxford and Entebbe to increase their range of skills and expertise. This learning will be enhanced by the opportunities to learn from world-leading collaborators and specific workshops, and to attend training at Oxford and MRC/UVRI Uganda.