**7. Genetic association studies and the post-genomic era**

Most of the protective variants are thought to have emerged in populations living in regions endemic for malaria as a result of the high selective pressure due to the parasite (Kwiatkowski, 2005). The past decade has seen growing evidence of ethnic differences in susceptibility to malaria and of the diverse genetic adaptations to malaria that have arisen in different populations. The fact that different malaria-resistance alleles have arisen in different places suggests that a great deal of selection by malaria has happened relatively recently in human history and certainly since human migration out of Africa (Eid et al., 2010). Such population differences in susceptibility to malaria are becoming more amenable to study since the development of high through-put genetic technology, thereby allowing us to genetically dissect the outcome of infection.

#### **7.1 Candidate and genome wide association studies**

Association studies are used for identifying genes and their common allelic variants involved in predisposition to a disease. Such studies are performed after localization of susceptible loci by linkage analysis. This method compares the allele frequency of a genetic marker of affected and non-affected individuals, chosen at random in a population (casecontrol study). The marker might be the causal polymorphism or any polymorphism in linkage disequilibrium (LD) with the causal one. A positive association with one marker suggests that this marker is in LD with the causal polymorphism. The LD between two markers is defined by the existence of a combination of alleles of these markers more often than expected by chance. The choice of the control population is one of the most important problems of case-control studies: if the control group is not from the same population as the affected individuals, uncontrolled environmental factors or population stratification might

Human Genetic Contribution to the Outcome of Infection with Malaria Parasites 285

The ultimate goal of studying the human genetics underlying infectious disease is to identify key biological pathways that determine the outcome of infection. Human genetics studies of malaria have, to date, almost exclusively focussed on severe disease caused by *P. falciparum* and yet this is a relatively rare event that regroups many pathologies. The majority of infections cause mild clinical symptoms or are asymptomatic. Asymptomatic infections are evidence for the development of clinical immunity (premunition), acquired at a rate proportional to exposure. Quantifying the frequency of a symptomatic *vs*. an asymptomatic outcome of infection and the parasite densities during such outcomes provides a measure of the host-parasite interaction. More detailed parasite phenotypes, such as those looking at particular parasite stages, provide additional insight into how the human response to infection determines the phenotype. Longitudinal family-based cohort studies offer the opportunity to generate non-severe malaria-related phenotypes for individuals over time, allowing the progression towards premunition to be examined. Family-based studies offer the possibility of assessing the robustness of phenotypes through simple heritability analysis. Fine-tuning phenotypes as such will save time and money during the human genotyping stage and subsequent analysis. Our cohort studies revealed surprisingly similar human genetic contributions to the clinical outcome of infection despite very large differences in transmission intensity. Identifying and accounting for confounding factors and covariates is essential and the similarity of the heritability values is reassuring for the robustness of the phenotype and implicitly suggest that covariates have been correctly accounted for. However, the potential that such covariates interact with human genes of interest should not be forgotten and future focus on gene-environment interactions is paramount. In addition, we identified for the first time a human genetic contribution to the transmission of the parasite. This opens up exciting possibilities for targeting transmission as well as disease. From our subsequent linkage studies, we found evidence that similar biological pathways govern the clinical outcome of a *P. falciparum* infection and allergy. In addition to the potential therapeutic possibilities, that common biological pathways govern multiple diseases offers the huge potential for combining efforts across multiple domains.

We are grateful to the villagers of Dielmo, Ndiop and Suan Phung for their participation and continued collaboration in this project. We thank the administrative authority of Institut Pasteur of Dakar, Senegal and Mahidol University, Thailand for their continuous support. Funding by provided by Institut Pasteur and the Ecole des Hautes Etudes en Santé

Abecasis, G.R., Cherny, S.S., Cookson, W.O., & Cardon, L.R. (2001) Merlin-- rapid analysis of dense genetic maps using sparse gene flow trees. *Nat Genet*, 30, 97-101. Agarwal, A., Guindo, A., Cissoko, Y., Taylor, J.G., Coulibaly, D., et al. (2000) Hemoglobin C

associated with protection from severe malaria in the Dogon of Mali, a West African population with a low prevalence of hemoglobin S. *Blood*, 96, 2358-2363.

**8. Conclusion** 

**9. Acknowledgements** 

Publique.

**10. References** 

induce false positive association. Association studies are the most widely used contemporary approach to relate genetic variation to phenotypic diversity. This is due to their higher power and lower cost to detect a susceptibility locus than linkage analysis.

The genome era has heralded unparalleled possibilities to identify genetic variants that underpin disease (http://www.genome.gov/gwastudies). The majority of these initial studies have concerned non-infectious diseases and, for the most part, examined dichotomous disease phenotypes, nominally affected and unaffected. Application to infectious diseases has only been relatively recent (Thye et al., 2010; Zhang et al., 2009) and for malaria has focussed on severe disease (Jallow et al., 2009). Genome wide association studies of clinical malaria and biological phenotypes are currently underway in our laboratory and whilst certain to reveal many novel candidate genes of importance, we know this is not enough. It is widely recognized that common multifactorial diseases are caused by multiple genetic and environmental factors and interactions among all these factors.

#### **7.2 Multiple loci and gene x environment interactions**

Following the success of identifying genes underlying diseases resulting from single locus mutations and inherited in a Mendelian fashion, it has become clear that there are many complex diseases that are inherited in a non-Mendelian fashion. That there are many loci affecting a trait is no surprise, but the simultaneous analysis of many loci is problematic for several reasons, especially when searching for novel genes as in the case of GWAS. The first is the curse of dimensionality - there are more candidate loci than there are observations. The second is the extent to which genes exert their effects independently, or whether there exist interactions among genes with respect to the phenotype. Thirdly, there is the question of whether there exists gene x environment interactions.

With the development of genotyping technologies, GWAS have become the method of choice to identify complex disease associated genes using SNPs as biomarkers (Hardy & Singleton, 2009). In the past three years, about 400 GWAS have successfully identified more than 531 genetic variants associated with various traits or diseases (Manolio et al., 2009). Standard analytical approaches in GWAS have proceeded by individually testing each SNP of the hundreds of thousands of genotyped SNPs. Thus, only SNPs that have a relatively strong marginal effect have been detected, explaining only a small part of the heritability. Hence, other SNPs that have weaker association with disease and/or act primarily through a complex mechanism involving interactions with other genetic variants and environmental factors have yet to be discovered. There is an increasing number of statistical methods and software that are being proposed to allow analysing multiple genetic markers and their interactions simultaneously (Cordell, 2009), but the identification of these interactions remain very challenging. Powerful methods for conducting genome-wide interactions studies are therefore needed. One of the interesting features of GWAS is that the same loci were found associated with several diseases (e.g. cancers, cardio-vascular diseases, autoimmune diseases), suggesting that genes with a pleiotropic effect may be more frequent than anticipated and may play a key role in basic physiopathological mechanisms underlying a number of diseases. The identification of pleiotropic genes which are likely to influence master regulators of biological processes is therefore of major importance. Studying together diseases that are supposed to share common genetic determinants can facilitate the characterization of such genes.
