Nearly one-third of the world's population is infected with Mycobacterium tuberculosis, the bacterium that causes tuberculosis. To establish and maintain infection, M. tuberculosis uses surface and secreted proteins to modulate the host immune response. There are several dedicated export machines that transport surface and secreted proteins from their site of synthesis in the cytoplasm across the bacterial cytoplasmic membrane. The bulk of protein export across the cytoplasmic membrane is carried out by the Sec pathway. Energy for Sec-dependent protein export is provided by the essential ATPase SecA. Recently, a small subset of Gram positive bacteria and mycobacteria were found to possess two SecA homologs, SecA1 and SecA2. In M. tuberculosis and the non-pathogenic model mycobacterium M. smegmatis, SecA1 is essential for general protein export and is the presumed "housekeeping" SecA, while SecA2 is an accessory SecA specific for a subset of exported proteins. In this work, we describe our initial attempts to characterize the mechanism behind SecA2-mediated protein export in mycobacteria. We began by comparing similarities and differences between SecA1 and SecA2. One striking difference we discovered was that SecA1 and SecA2 localize to different cellular compartments. By comparing a structural prediction of SecA2 to the published crystal structure of SecA1, we identified putative structural differences between the two SecAs. Both SecA1 and SecA2 have predicted ATP binding sites. We showed that M. tuberculosis SecA1 and SecA2 bind and hydrolyze ATP in vitro. By constructing a secA2 mutant that encodes a protein defective in ATP binding, we also demonstrated that ATP binding is required for normal SecA2 function in vivo. Upon subsequent analysis, we found that SecA2 mutants unable to bind ATP were dominant negative. We used this dominant negative phenotype as a tool to study SecA2 by performing a suppressor screen. We isolated intragenic suppressors and used them to identify structural subdomains of SecA2 that are important for function. While we have not yet identified the extragenic suppressors, we believe they represent a powerful tool for identifying SecA2-interacting proteins. Identifying SecA2-interacting proteins will be crucial to fully understand the accessory SecA2 protein export pathway of mycobacteria. Finally, we show for the first time in any bacterium with two SecA proteins that the canonical SecA1 protein is also required to export SecA2-dependent substrates. By developing a better understanding of the SecA2 pathway, we hope to increase our knowledge of mycobacterial physiology to enable the design of new anti-tuberculosis therapies.