Vaccinomics, predictive vaccinology and the future of vaccine development

The concept of vaccinomics

Vaccinology has rapidly progressed from the first empirical smallpox vaccine application in 1796 by Edward Jenner to the ‘genomic era’ of reverse vaccinology, computational vaccinology and recombinant vaccines (hepatitis B vaccine), immunogenic refinements (pneumococcal and meningococcal conjugate vaccines, virus-like particle vaccines such as the human papillomavirus [HPV] vaccine) and recently the prospect of ‘universal vaccines’, based on highly conserved peptide or saccharide sequences, against a defined species or group of pathogens [1]. We believe that advanced vaccine development and implementation is not achievable without recognizing the mechanisms underlying the heterogeneity of immune responses to vaccines, as informed by population genetic studies, and have extensively published on the role of immune response gene families on humoral, cellular and even innate responses to vaccines [2–4]. Our previous research has highlighted a large and growing family of immune response genes that are critical to vaccine immune responses, such as classical human leukocyte antigen (HLA) genes, cytokine and cytokine receptors, Toll-like receptors and related signaling molecules, viral cell-surface receptor genes, vitamin A and D receptors, antiviral effectors and other genes associated with innate and adaptive immunity, merged into a complex interconnected ‘immune-response network’ [2,4,5]. The ability to understand and comprehensively elucidate variations in host response to immunization will likely dramatically affect the design of new and more effective vaccines, and the development, further refinement and use of both new and existing vaccines and vaccine adjuvants to address the effect(s) of common and/or rare genotypes on vaccine response phenotype (‘personalized vaccinology’). We have previously defined the emerging field of ‘vaccinomics’ as the application of immunogenetics and immunogenomics to understanding the biologic and immunologic basis of vaccine response [2,3]. The current level of growth of this area of inquiry demands the integration of advanced immunology approaches, systems biology, immune profiling, functional validation studies and sophisticated statistical and bioinformatics analysis methods and modeling algorithms (immunoinformatics) to decipher the complexity of immune responses to vaccines [6]. As a result the continuously evolving field of vaccinomics encompasses the discovery, replication, validation and interpretation of known and/or novel immunogenetic signatures and immune profiles that explain the variance in immune response and/or discriminate high from low responders to vaccines. Our ability to understand variations in immune responses to vaccines is critical to directed vaccine development and use of vaccines. For example, knowledge of replicated ‘true-positive’ associations between variations in vaccine-induced immune response and polymorphisms in the functional domains of immune response genes (e.g., cytokine or cytokine receptor genes), could potentially allow the design of a vaccine (such as a cytokine adjuvanted vaccine) that circumvents immunogenetic restrictions and minimizes vaccine failure [2,7].

Predictive vaccinology in the dawn of new technology

We have used the term ‘predictive vaccinology’ to denote predictions of novel vaccine candidates, such as promiscuous HLA-binding peptides, as targets for vaccine development, via mass spectrometry-based approaches and predictive computational models. We introduced and further expanded this concept to ‘predictive vaccine response profiling’, intended to characterize and/or discriminate high from low vaccine-induced immune responses and to answer (at an individual, group or population level) important clinical questions regarding vaccine immunogenicity, efficacy, number of doses needed, potency of dose, route of application and plausible adverse events, all at a reasonable cost [8]. The comprehensive characterization and prediction of complex traits such as immunity and vaccine reactivity is not possible without the implementation of innovative approaches and technologies to develop very large genotype/phenotype datasets, integrated and explored by the use of advanced bioinformatics algorithms and statistical analysis methods.

During recent years, applications of novel and advanced high-dimensional technologies in the medical field have been rapidly emerging. New high-throughput assays, such as advanced immunologic assays (simultaneously measuring the expression of hundreds of proteins), genome-wide genotyping and sequencing, gene expression profiling arrays and epigenetic assays, combined with sensitive proteomic detection and identification techniques are radically transforming vaccine research and vaccine development. The use of next-generation sequencing (NGS) for whole-genome sequencing, targeted deep sequencing (for focused high-resolution interrogation of narrow genomic intervals of interest), mRNA whole-transcriptome gene expression and chromatin immunoprecipitation (CHIP)-sequencing (mapping global binding sites for any protein of interest) offers unimaginable sensitivity, breadth and precision in expression profiling and genome analysis of both host and pathogens. In particular, NGS allows a more comprehensive host-transcriptome interrogation, including the detection of low abundance and novel transcripts (including small-noncoding RNAs), expressed mutations/polymorphisms, alternative splice forms and fusion genes. We and others have acknowledged the potential of this technology for providing a more detailed multidimensional view of host–pathogen interactions, and for adding new insights into infection pathogenesis and vaccine-induced immune responses [8]. Our group has spent years developing and optimizing the ability (both in terms of technology and analysis) to identify specific genetic, protein (naturally processed, HLA-bound pathogen-derived peptides) and transcriptome signatures that are associated with, or can ultimately predict, variance in vaccine-induced immunity. Using mRNA-Seq, we recently simultaneously assessed host cell transcriptomes and levels of viral replication in high and low antibody responders to rubella, smallpox and influenza vaccines, and were able to identify host modulation of genes and pathways integral to immune function that accounts for differences in immune response to these viral vaccines [Poland GA et al., Unpublished Data] [9]. The next step will be to apply similar and/or novel analytic approaches to a larger sample size in order to validate these, and/or other key gene signatures as distinguishing ‘profiles’ that identify and predict high or low immune response. These findings will further expand our understanding of the mechanisms and drivers of vaccine-induced immunity specific for particular vaccines and vaccine immunity in general. In addition, the resulting NGS-specific findings for different vaccines can be compared, contrasted or integrated with epigenetic, genome-specific and/or protein expression-specific data to potentially produce a ‘universal viral vaccine signature’ or profile(s), explaining or predicting vaccine-induced immunity. Such work may also lead to more general principles that advance our understanding of vaccine immunogenetics. The discovery and validation of such signatures/profiles is likely to have an immense impact on vaccine research and could be readily translated into clinical practice, and used for predicting robust and/or weak humoral or cellular protective immune responses following vaccination. Furthermore, these immune profiles/signatures could be used to predict adverse events or serve as markers of vaccine efficacy. After validation, identified vaccine-specific or universal signatures/profiles could readily serve as improved biomarkers of vaccine immunogenicity, safety and effectiveness, allowing for quick and inexpensive clinical testing for rapid acceptance/rejection of new vaccine candidates and facilitating directed vaccine development and testing.

An example of predictive vaccinology is our advanced mass-spectrometry-based approach for isolation and identification of pathogen-derived viral peptides (e.g., vaccinia virus and measles virus peptides) from specific high- or low-responder-associated HLA molecules as candidate antigens for directed vaccine development [10,11]. In addition, we have defined associations between genotypes/haplotypes in a variety of immune response genes and immune response phenotypes for measles, mumps, rubella, influenza and more recently, smallpox vaccines [2,4].

Other examples of predictive vaccinology addressing vaccine-induced immune response have also appeared in the literature. Using a systems biology approach Querec et al. reported the discovery and validation of a distinct signature (including B-cell growth factor TNFRSF17), which predicted the subsequent adaptive immune response (neutralizing antibodies) following yellow fever vaccine with up to 100% accuracy, thus uncovering new correlates of vaccine efficacy [12]. In addition to new discoveries, efforts should be made to replicate and functionally validate identified novel and existing targets. We believe that the emerging field of vaccinomics and predictive vaccinology will be further advanced by the developing efforts to create functional ex vivo immune system modules, such as vaccination site equivalents, lymphoid tissue equivalents and artificial engineering of secondary lymphoid organs, which will ultimately allow functional validation studies and representative ex vivo testing of immunogenicity and vaccine response [13].

Future endeavors in the field should consider:

• ▪ Genetic association and immune profiling studies, designed and powered to detect scientifically and clinically relevant associations and correlates across populations;

• ▪ Studies using high-dimensional, high-throughput technologies, designed to take full advantage of genetic, transcriptomic and proteomic information;

• ▪ Studies using sophisticated bioinformatics, and statistical analysis and modeling approaches, for integration and systems-level analysis of disparate data;

• ▪ Studies designed for follow-up replication and functional validation of results;

• ▪ Studies directed at developing, sharing and using public databases, biobanks of clinical specimens and analysis tools to decrease associated costs and allow for rapid, inexpensive and reliable testing.

Conclusion

The new paradigm of vaccinomics and predictive vaccinology has brought irrevocable changes to the field of vaccine research with the advent of novel technology, advanced systems-level analytic approaches and the wealth of information regarding vaccine-induced immune responses that they provide. These technical and methodological advances have prompted and are driving the development of novel computational and analytic tools to identify, characterize, discriminate and predict immune response profiles and phenotypes following vaccination. In turn, we need to accept and apply a new strategy for vaccine testing and vaccination, based on predictive vaccinology and increasingly personalized vaccine approaches informed by these new paradigms and scientific discoveries.