Tuberculosis (TB) is a chronic infectious disease that mainly affects the lungs and causes extensive human morbidity and mortality. It results from infection with Mycobacterium tuberculosis, a slow-growing intracellular pathogen that can replicate and survive inside macrophages. M. tuberculosis relies on the specialised ESX-1 secretion system to export virulence factors needed for intracellular spread and pathogenesis. The ESX-1 apparatus is a multi-subunit nanomachine composed of ~20 polypeptides including membrane proteins, ATPases, proteases, chaperones and substrates. Although many of the individual components of this secretion system have been characterised, the overall mechanism underlying ESX-1 secretion is still far from clear. To obtain a more comprehensive picture of the functioning of the ESX-1 apparatus, we have studied various components which were largely unexplored in M. tuberculosis. The structural component EccE1, the ESX-1 specific protein EspL and the transcriptional regulator WhiB6 have been investigated in this thesis using an integrative approach involving genetics, biochemistry, proteomics and microscopy. We have demonstrated that EccE1 is a membrane- and cell-wall associated protein critical for secretion of ESX-1 substrates and M. tuberculosis-mediated cell lysis. Deletion of eccE1 from the chromosome severely compromised secretion of EsxA, EsxB, EspA and EspC but not EspB. Localization studies using a florescent-fusion protein showed that EccE1 localises to the poles of M. tuberculosis in the presence of an active ESX-1 system. Our study also shows that EspL is a cytosolic protein needed for stabilising EspE, EspF and EspH, suggesting that it acts as a specific chaperone of the ESX-1 secretion system. Moreover, EspL was shown to interact with EspD and to be important for the secretion of ESX-1 substrates. Lack of EspL resulted in a growth defect ex vivo, loss of cytotoxicity and reduction of innate cytokine production demonstrating its critical role in M. tuberculosis virulence. Analysis of the transcriptional response revealed that the only gene deregulated in the absence of espL was whiB6, encoding a transcriptional factor that positively controls ESX-1 genes. To explore the role of this regulator in M. tuberculosis virulence, we generated a deletion mutant of whiB6 and discovered that its loss resulted in severe reduction of cytotoxicity ex vivo. Overall this investigation improves our current understanding of the ESX-1 secretion system and of the molecular basis of M. tuberculosis virulence. The increased knowledge of the complex interactions between the pathogen and the human host will hopefully translate into new strategies to control the spread of TB.