Different PE_PGRS mutant proteins were expressed in different model organisms to study their localization: i) M. smegmatis, whose chromosome encodes neither ESX-5 nor PE_PGRS proteins, but was previously shown to be able to export PE_PGRS33 and chimeric proteins based on its PE domain ; ii) M. marinum, whose chromosome encodes ESX-5 and many PE_PGRS proteins, but not an apparent PE_PGRS33 orthologue, and iii) M. bovis BCG and M. tuberculosis, whose chromosomes encodes both ESX-5 and PE_PGRS33. In different genetic backgrounds to investigate the conservation of the PE-exporting pathway. As expected, the protein lacking the entire PE domain was not able to translocate and was exclusively found in the bacterial cytoplasm. The same expression profile was found in the mutant protein missing the first 30 amino acids of the PE domain, suggesting that this portion of the protein is essential for the translocation. Surprisingly, mutation of the SF or of the PE (+)-JQ1 abmole bioscience conserved residues did not result in any clear phenotype with the exception of a strong instability of the resulting proteins in M. marinum. It is worth noting that PE_PGRS33 was found in Genapol extracts of M. marinum, but not in those of the other tested mycobacterial species suggesting a weaker association of this protein to the cell wall in this species. Of course it is still possible that using different experimental procedures PE_PGRS33 and/or some of the mutant protein might be extracted with this detergent even in M. smegmatis, M. tuberculosis or M. bovis BCG. These results confirm previous proteomics data on Genapol extracts from M. marinum as compared to M. tuberculosis or M. smegmatis. Moreover, the size of PE_PGRS33 in M. marinum Genapol extracts showed a molecular weight lower than that predicted for the entire protein, suggesting a maturation process, as we recently reported for LipY, another PE protein not belonging to the PE_PGRS family. In that case the protein was also processed in M. tuberculosis, but only when bacteria were grown inside macrophages, suggesting that still unknown factors only expressed during infection are required for full export and maturation of PE proteins in M. tuberculosis. For this reason, PE_PGRS33 might also be expected to be processed in M. tuberculosis during growth inside macrophages. In M. smegmatis some of the recombinant proteins showed multiple bands. However, since bands of the same size were present in both cytoplasmic and cell wall fractions, they were likely not due to processing during translocation but rather the result of degradation during cell lysis. The lack of processing observed in M. smegmatis might be due to the absence of ESX-5 in this species. Finally, the release of PE_PGRS33 was totally abrogated in the M. marinum ESX-5 mutant, confirming that PE protein translocation in M. marinum is dependent on this secretion system.