Coping with genetic diversity: the contribution of pathogen and human genomics to modern vaccinology

Lemaire, D.; Barbosa, T.; Rihet, P.

This concept has stimulated the development of gene prioritization tools to propose candidate genes lying within a chromosomal region linked to the disease. The approaches used imply that most genes involved in a disease share functional properties and/or can be mapped in the same gene or protein network. The most frequent approach is to rank candidates on the basis of their similarity to a set of training genes that have been demonstrated to be involved in the disease (reviewed in Ref. 58) . The biological concept of the gene network implies the existence of gene-gene interactions that are generally not taken into account in genome-wide linkage and association analyses. Genegene interaction can be explicitly included in the models used in linkage and association analyses. Statistical approaches have been proposed to detect such interactions also in studies of vaccine responses (59) . Furthermore, the influence of TLR2 and TIRAP polymorphisms on clinical disease caused by M. tuberculosis has been found to depend on pathogen genotypes, suggesting the existence of bacterial gene-human gene interactions (60) . Obviwww.bjournal.com.br Braz J Med Biol Res 45 (5) 2012 ously, the gene-gene interaction studies at the genome scale pose additional multitest problems that must be tackled. In addition, such approaches are generally limited to low order interactions, and efforts are now needed to develop statistical approaches that make use of the whole biological network. The analysis of genetic variation and gene expression levels could also provide help in understanding how polymorphisms could perturb a transcriptional network in infectious diseases (61) . Microarray studies that are based on mice infected with either virulent or nonvirulent pathogens have identified gene expression patterns associated with some infectious diseases, indicating the existence of at least one transcriptional network perturbed by a virulent strain. Moreover, gene expression profiles have been shown to discriminate between genetically cerebral malaria-resistant versus genetically cerebral malaria-susceptible mice at early and late stages of infection with P. berghei ANKA (62, 63) . This further indicates that genetic variants direct at least one transcriptional network towards a state conducting to cerebral malaria and death in susceptible mice after infection. In humans, blood cell gene expression profiles have been associated with the clinical status of patients infected with different pathogens (64) , and have been used to monitor the response to BCG vaccination in children (65) .