Tuberculosis (TB) is currently and historically an enormous public health problem. Approximately one-third of the world's population are currently infected with Mycobacterium tuberculosis (M. tuberculosis) and TB accounts for over 25% of preventable adult deaths world-wide. Despite the high infection rate, only about 10% of people infected with M.tb ever become sick with active TB. Evidence suggests that progression to active TB is influenced by host genetic factors. For example, the epidemiology of TB suggests that genetic selection takes place after introduction of M. tuberculosis to the population; genetically susceptible individuals succumb to the infection and relatively resistant individuals survive to reproduce. As well, twin studies demonstrate higher concordance rates for TB among identical twins, compared to fraternal twins. Mouse models of mycobacterial infection have identified several potential susceptibility loci, such as the gene named Nramp1, as well as several cytokine and cytokine receptor genes. Family-based linkage studies and case-control studies of candidate genes in humans suggest roles for these and other genes associated with development of TB in humans. In light of these observations, we propose a family-based association study of candidate genes for TB susceptibility. To accomplish the goal of identifying genes influencing susceptibility to TB we specifically propose to: 1) Ascertain 1,000 parent- child triads (500 Caucasian, 500 African-American) from North and South Carolina for genetic studies of TB susceptibility genes. 2) Test candidate genes in the first 500 parent-child triads. Multiple single nucleotide polymorphisms (SNPs) will be genotyped in each gene and analyzed using family- based tests of association; significant results will be followed-up in the remaining 500 triads. 3) Examine the relationship between candidate genes and other clinical variables such as PPD skin test results, disease severity, treatment relapse and failure, and presence of extrapulmonary disease. 4) Evaluate gene-gene and gene-environment interactions using multivariable models and data reduction techniques such as the multifactor dimensionality reduction (MDR) method.
