Diverse pathogens encode proteases with crucial functions during infection, but knowledge on secreted proteases and their activities in
H. pylori is very limited. In many bacteria, HtrA is a well-recognized factor in the periplasm, which contains chaperone and proteolytic functions with important roles in protein quality control involved in stress tolerance and bacterial survival [
23,
25‐
29]. In addition, it was demonstrated that HtrA has a significant impact on the virulence of multiple bacterial pathogens including
Borrelia, Burkholderia,
Campylobacter, Chlamydia,
Klebsiella, Legionella, Listeria, Salmonella,
Shigella and
Yersinia species. Interestingly,
htrA does not appear as an essential gene in each of these bacteria because Δ
htrA knockout mutation has been described [
31,
32,
34,
43‐
50]. In contrast, inactivation of the
htrA gene in
H. pylori has been unsuccessful in more than one hundred worldwide isolates, but the reasons for this failure are still unclear [
37,
51,
52]. Remarkably, it was also demonstrated that pharmacological inhibition of HtrA protease activity effectively killed
H. pylori, while it did not affect the growth and viability of other Gram-negative pathogens including
Salmonella and
Shigella [
52].
Research progress on
H. pylori HtrA is mainly hampered by the lack of Δ
htrA knockout mutants. Thus, other genetic manipulation strategies are required to study HtrA function during the infection process. Here we developed a genetic approach to overexpress HtrA in two clinical isolates, P12 and 26695. For this purpose, a second
htrA gene copy was introduced into the
H. pylori chromosome and placed under an IPTG-inducible promotor [
53]. Once the HtrA proteins are translated by the bacteria they are delivered into the periplasm and subsequently secreted into the extracellular environment. This important new aspect seems to be conserved among a wide range of worldwide
H. pylori isolates [
52]. We could show here that overexpression of HtrA enhanced not only its proteolytic activity by up to ~2.5-fold, but also the secretion of the protease by ~1.8-fold. Interestingly, the secretion of other well-known bacterial virulence determinants, VacA and GGT, was not affected by HtrA overexpression, suggesting that the secretion of these factors follow different, non-linked pathways. In addition, we could demonstrate that various virulence-associated properties of
H. pylori were also not affected including bacterial attachment to the epithelial cells and induction of pro-inflammatory responses such as the secretion of IL-8. In contrast, the transepithelial migration of
H. pylori overexpressing HtrA increasing significantly up to ~2.2-fold compared to the control bacteria. This phenotype was accompanied by significantly enhanced damage to the adherens junction protein E-cadherin. Our Western blotting data demonstrated that HtrA-mediated cleavage of full-length E-cadherin was enhanced, leading to elevated levels of the 90 kDa E-cadherin NTF-fragment in the supernatants of infected cells. Immunofluorescence microscopy confirmed these observations and showed that the cell-to-cell junctions of infected Caco-2 cells were significantly more disrupted after 24 h compared to the wild-type control infection, explaining why higher numbers of bacteria can cross the epithelial barrier and reach basolateral compartments. Finally, we observed that the levels of CagA translocation and phosphorylation increased up to ~twofold in HtrA-overexpressing
H. pylori compared to the control bacteria. These observations can be explained by reports showing that CagA delivery into host cells requires a receptor, which was identified as the basolateral integrin member α
5β
1 [
56‐
62]. Integrins are well-known mammalian cell adhesion receptors, which facilitate anchoring of host cells to the extracellular matrix and which are absent at apical surfaces [
63,
64]. These findings let us to suggest a novel mechanism how the T4SS of
H. pylori works in polarized epithelial cells by cooperating with the secreted serine protease HtrA, which opens cell-to-cell junctions. Using an inducible genetic system to overexpress HtrA, we could enhance the proteolytic activity of HtrA, necessary for elevated paracellular transmigration of
H. pylori across the polarized epithelial cells to reach basolateral membranes and inject CagA in an integrin-dependent fashion. Extensive research has shown in recent years that the above discussed features basically resemble a phenotype, called epithelial-mesenchymal transition (EMT). Gastric cancerogenesis is known for its aggressiveness and tendency to metastasize. EMT is the initial step in metastasis, orchestrated by various cellular factors [
65]. We proposed that the activity of secreted HtrA is maybe the initial step in a signaling cascade, followed by CagA and probably others, that triggers EMT in gastric epithelial cells. Translocated CagA can then deregulate cell polarity and scattering, by various pathways including the interaction with partioning kinase Par1b changing cell polarity [
66] and by stabilizing Snail, a transcriptional repressor of E-cadherin expression [
67]. Taken together, these data provide for the first time genetic evidence that HtrA is a major novel virulence factor of
H. pylori, controlling multiple pathogenic activities of this important microbe.