Skip to main content

Identification of novel correlates of protection to accelerate the introduction of vaccines against typhoid into populations with high disease burden

Final Report Summary - MCF_IIF BLOHMKE 2012 (Identification of novel correlates of protection to accelerate the introduction of vaccines against typhoid into populations with high disease burden.)

Introduction and Rational
Typhoid fever caused by Salmonella enterica serovar Typhi (S. Typhi) is a significant cause of morbidity and mortality throughout the world, particularly in resource-limited settings. Despite decades of research into disease pathogenesis and two available licensed vaccines against typhoid fever, the disease is still endemic in many resource-limited settings. Licensed vaccines are only moderately efficacious and cannot be used in young children, a population carrying a large part of the burden of disease. The lack of knowledge regarding disease pathogenesis of this biologically complex infection as well as the absence of suitable correlates of protection hinder the development of improved vaccines. The Oxford Vaccine Group (OVG) at the Univeristy of Oxford recently established a controlled human infection (challenge) model for typhoid fever. In this model, healthy, consenting adult volunteers are deliberately infected (challenged) with a dose of disease causing S. Typhi, monitored following infection and subsequently treated with a course of antibiotics. This model is now being used to screen vaccine candidates for efficacy to facilitate the deployment of novel vaccines into the field.
The overarching goal of the present project was to establish analyses pipelines for transcriptional data in participants with acute typhoid following challenge. The analysis strategies are then applied to data derived from volunteers vaccinated with two live attenuated oral vaccines against typhoid, Ty21a and M01ZH09. We specifically focused on molecular host response by analyzing gene expression data at several time points after vaccination and relating these transcriptional profiles to immunogenicity as well as clinical outcome (typhoid diagnosis) following subsequent infection of vaccinees.
Results
We analyzed several datasets comprising of samples derived from participants with acute typhoid, participants challenged but who stayed well following challenge, and participants vaccinated with the two live attenuated oral vaccines. Using data from solely challenged participants we established integrative computational analysis scripts in order to mine multi-faceted datasets consisting of transcriptional, serological, cytokine and clinical data to develop knowledge regarding disease pathogenesis. Several novel key insights were yielded from these analyses:
A) Model ‘validation’: A criticism is that the responses in a challenge model using naïve volunteers will not mirror the host-responses in the field where high background infection with other disease may obscure the response pattern. Comparison of transcriptional profiles from patients in the field (Vietnam) with our data indicated a marked degree of similarity between acute sample from the field and the challenge model.
B) The data shows a potential involvement impact of the host’s metabolism on the immune response to typhoid. This previously unrecognized link may be integral to the interaction between host and pathogen.
C) We are the first to identify host responses to oral challenge 12hrs after ingestion of the pathogen. This entirely unknown signal has previously not been reported and may provide insight into how the host protects itself from developing typhoid fever.

The computational analysis algorithms were subsequently adapted to interrogate the transcriptional profiles to vaccination. Host responses to vaccination are profoundly more subtle compared to those measured during acute disease. Using more sensitive computational methods and customized analyses approaches, we investigated the transcriptional patterns overall (classic approach) and on a single sample basis. This sophisticated approach allows integration of the gene expression with immunogenicity, namely plasma cell and antibody responses 7 and 28 days following completion of the vaccine schedule, respectively. The two vaccines, M01ZH09 and Ty21a, showed marked differences in the magnitude of immunogenicity (M01ZH09 > Ty21a). This was reflected in the transcriptional profiles with 144 and 110 genes differentially expressed (p<0.001) after M01ZH09 and Ty21a vaccination, respectively. Detailed Gene Set Enrichment Analysis (GSEA) of blood transcriptional modules (BTMs) previously described indicated distinct differences between the vaccines:

• BTMs reflecting NK cell transcripts were significantly up-regulated after Ty21a and down-regulated after M01ZH09 vaccination.
• BTMs reflecting the cell cycle were significantly up-regulated following M01ZH09 but not Ty21a vaccination.

These are entirely novel insights into the response to live attenuated oral vaccines. Interestingly, despite being derived from the same parent S. Typhi strain (Ty2), the vaccines elicit distinctly different profiles, which may or may not be due to vaccine dose or molecular differences caused by the attenuation of the vaccine strains. Specifically correlation analysis highlighted a strong association of the cell cycle modules with humoral responses thus predictive of immunogenicity.

Impact and Novelty
This project developed considerable systems vaccinology expertise within OVG. Through this fellowship, we established fruitful, cross-sectional collaborations with leading bioinformaticians, immunologists and clinicians around the world. The expertise is already a significant contribution to the ERA with several publications being prepared at the moment. The novel insights into disease pathogenesis and responses to live attenuated oral vaccines against typhoid are likely to be invaluable for further vaccine development, identification of novel diagnostic tools and treatment opportunities.