Secretion Mechanisms Potentially Responsible for Transporting HtpB to Extracytoplasmic Locations
We determined very early in our research that HtpB could be found in extracytoplasmic locations (Garduno et al. 1998a), as could other chaperonins then reported (Garduno et al. 2011). To date, many extra-cytoplasmic chaperonins have been identified, and it is no longer surprising to find more bacterial species in which chaperonins are associated with the cell envelope, exposed on the bacterial cell surface, or present in culture supernatants, as most recently reported in Rickettsia spp. (Qi et al. 2013; Gong et al. 2014), Streptococcus gordonii (Maddi et al. 2014), and Edwardsiella tarda (Kumar et al. 2014). Moreover, at least two reports have shown that the process by which chaperonins reach extra-cytoplasmic locations involves active secretion, as opposed to passive release by cell lysis (Vanet and Labigne 1998; Yang et al. 2011). It is therefore accepted now that Cpn60s are secreted by some bacterial species, but the mechanism by which this is accomplished largely remains a mystery. Before discussing the proposed mechanisms of Cpn60 secretion, it seems reasonable to tackle a fundamental question: why are Cpn60s secreted in the first place? Since ATP would not be readily available in extra-cytoplasmic bacterial locations, and protein folding is critically dependent on a supply of ATP (Clare and Saibil 2013), it could be argued that secretion of Cpn60s is not primarily intended to provide proteinfolding assistance beyond the cytoplasmic membrane. Although this argument would not be valid for intracellular pathogens and endosymbionts (where their extracellular compartment actually is the ATP-rich intracellular milieu of their host cells), an additional barrier seems to exist. That is, protein folding by Cpn60s happens in large 14-mer complexes that require the participation of a 7-mer co-chaperonin (Cpn10) complex, so that the co-secretion of cognate Cpn10s would need to be ensured. It therefore seems reasonable to surmise that Cpn60 secretion co-evolved with Cpn60 moonlighting rather than with protein-folding. This view is also supported by the many reported moonlighting Cpn60 functions that are compatible with an extra-cytoplasmic location, particularly the bacterial cell surface (Henderson et al. 2013). In conclusion, the evolution of Cpn60 secretion mechanisms seems to be driven by the bacteria-in-question's niche, their particular needs when occupying such niche, and the contribution of extra-cytoplasmic moonlighting Cpn60s to meeting those needs.
One limiting condition in the elucidation of Cpn60-secreting mechanisms is that, by design, these would have to be inefficient. That is, due to the essential role that Cpn60s play in the cytoplasm of bacterial species that possess a single cpn60 gene (Lund 2009), these bacteria could not afford to have dangerously low cytoplasmic levels of their chaperonin as a consequence of an efficient Cpn60 secretion process. Although this argument would not be valid for the numerous bacterial species with multiple cpn60 genes and functional Cpn60 diversity (Lund 2009), it could still be argued that an efficient secretion system would be deleterious due to a potential inability to distinguish between the various highly conserved Cpn60s in the cell. We would therefore expect to find only small amounts of secreted Cpn60s at any given time, making it difficult to distinguish their measured levels from the background levels of Cpn60s passively released by lysis. In spite of this experimental limitation, mechanistic hints for chaperonin secretion do exist.
One possible (and favored) mechanism by which chaperonins reach culture supernatants (i.e., the extracellular milieu) is via outer membrane vesicles (OMVs) (Ferrari et al. 2006; Joshi et al. 2008; McCaig et al. 2013; Berleman et al. 2014; Li et al. 2015). However, it remains to be determined how Cpn60s end up either in the lumen of OMVs or associated with their membrane, as OMVs primarily consist of periplasmic and outer membrane material. HtpB is also found in OMVs (Galka et al. 2008), and we have determined that L. pneumophila mutants with a defective Dot/Icm system (a type IV secretion system essential for virulence) accumulate periplasmic HtpB (Chong et al. 2006), and have increased amounts of OMV-associated HtpB (unpublished results). Based on these observations and other experimental evidence for the periplasmic residence of HtpB (Allan 2002), we have proposed that a portion of the total HtpB in L. pneumophila naturally translocates across the cytoplasmic membrane into the L. pneumophila periplasm, from where it is then packed as cargo in OMVs. Besides accumulating HtpB in the periplasm, Dot/Icm mutants do not display surface-exposed HtpB, and we have therefore hypothesized that HtpB is also mobilized from the periplasm across the outer membrane by the Dot/Icm system. However, it is not known how HtpB gets into the L. pneumophila periplasm.