The ever-emerging complexity of α-toxin’s interaction with host cells

For more than a century, Staphylococcus aureus α-toxin, also known as α-hemolysin or Hla, has been recognized as an important virulence factor (1). α-Toxin exhibits many roles in the virulence of this formidable pathogen. In the past few years, our understanding of these roles has increased immensely. This wave of discovery began with the identification of α-toxin as a critical determinant of community-associated methicillin-resistant S. aureus pneumonia (2, 3). These studies were then followed by the identification of the metalloprotease a disintegrin and metalloprotease 10 (ADAM10) as the receptor for α-toxin (4), and the discovery that α-toxin targeting of ADAM10 has a critical role in the pathogenesis of S. aureus pneumonia and skin infection (5, 6), through the disruption of epithelial barriers. These studies demonstrated that α-toxin disrupts barriers not only by lysing cells but also by the more subtle mechanism of activating ADAM10 to cleave the ectodomains of E-cadherin molecules, breaking adherens junctions and causing dramatic changes of the actin cytoskeleton (4–6). The pathogenic functions of α-toxin have been further extended to include several roles in sepsis, including disruption of endothelial barriers and interactions with platelets and leukocytes (7, 8). Further, α-toxin interacts with the development of adaptive immunity and has a role in recurrent skin and soft tissue infections (9). In PNAS, Popov et al. (10) deepen our understanding of the interaction between α-toxin and cellular junctions, finding that several components of the adherens junctions, including pleckstrin-homology domain containing protein 7 (PLEKHA7), are necessary for cell death upon intoxication with this toxin.

Popov et al. (10) performed a whole-genome screen in search of additional host factors that facilitate the cytotoxic function of α-toxin by taking advantage of the mutagenized haploid human cells HAP1 (11). Importantly, the top hit of this unbiased screen was ADAM10, which both validates the screen and provides further support for ADAM10 as the key receptor for α-toxin (4). Interestingly, the next most significant hit was PLEKHA7, a component of adherens junctions. Several other proteins associated with adherens junctions were also identified in the screen, including N-cadherin and α-catenin. The authors (10) further investigated the requirement for PLEKHA7 for the lethal effect of α-toxin, showing that the phenotype can be complemented by PLEKHA7 expression in trans, and that expression of the first 284 amino acids of the protein is sufficient for this complementation.

It seems that α-toxin has evolved specific mechanisms to target cells that are engaged in cell–cell junctions.

Unexpectedly, Popov et al. (10) show that α-toxin formed pores in the plasma membrane of ΔPLEKHA7 cells but that these cells are able to recover from injury. The cells experience K⁺ flux and cytopathic changes, as well as a loss of intracellular ATP similarly to WT intoxicated cells; however, by 8 h after intoxication, they begin showing signs of recovery, and about 80% of them survive the insult. The ΔPLEKHA7 cells do phosphorylate p38 MAPK upon α-toxin exposure, indicating that they are activating recovery pathways (12).

The authors (10) nicely extend their findings to the in vivo setting by taking advantage of PLEKHA7^−/− mice. In an ear skin S. aureus infection model, PLEKHA7^−/− mice develop a similar lesion size compared with WT mice initially but then resolve the infection with less tissue loss. Similarly, the mice exhibit decreased lethality in an S. aureus pneumonia model without exhibiting a significant decrease in lung bacterial burdens. Thus, PLEKHA7 exacerbates α-toxin cytotoxicity both in vitro and in vivo.

Cellular Responses to α-Toxin

This study adds to the growing body of work demonstrating that α-toxin, as well as bacterial pore-forming toxins in general, elicits cellular responses that are far more complex than simple pore-mediated osmotic lysis (13, 14). Exposure to α-toxin can cause cellular death by necrosis, apoptosis, or pyroptosis, through activation of different cellular pathways (15, 16). Cells are capable of recovery from α-toxin exposure, which has been shown to involve p38 MAPK activity and endocytosis (12, 17, 18). Autophagy, which can facilitate a cell’s capacity to survive stress, can also be triggered by α-toxin exposure (19). Of note, inactivation of autophagy renders host endothelium more susceptible to α-toxin, a phenotype linked to increased surface exposure of ADAM10 (20). Lastly, α-toxin exposure can even cause cell proliferation (21). Thus, the responses to α-toxin may be cell type-specific; whereas erythrocytes and leukocytes may lyse relatively easily upon exposure to α-toxin, “hardier” cells, such as epithelial cells, may have more robust mechanisms for surviving membrane assault. Interestingly, a key difference between epithelial cells compared with erythrocytes and leukocytes is that epithelial cells are attached to other cells by cell–cell junctions in physiological settings, whereas erythrocytes float by themselves and many leukocytes spend much of their life unattached to other cells. It seems that α-toxin has evolved specific mechanisms to target cells that are engaged in cell–cell junctions, perhaps because epithelial cells are relatively resistant. On the other hand, leukocytes are involved in cell–cell junctions at very crucial points in their lifetime, including when they roll along the endothelium and when they form immune synapses. As such, the targeting of cell–cell junctions by α-toxin brings up the interesting possibility that α-toxin may be specifically interfering with these processes. Indeed, targeting of cell–cell junctions could be a mechanism by which α-toxin inhibits the formation of protective immunity, facilitating recurrent skin infections (9).

The study by Popov et al. (10) brings up many fascinating new questions. Chief among these questions is whether, in the absence of PLEKHA7, α-toxin still activates ADAM10 to cleave E-cadherin. If, as seems likely, the answer to this question is “yes,” then ΔPLEKHA7 cells and PLEKHA7^−/− mice are tools that could be used to tease apart which actions of α-toxin are mediated through ADAM10 cleavage, as opposed to overt cell lysis. If the answer is “no,” then it is possible that PLEKHA7 has a role in the regulation of epithelial shedding in general. Similarly, it will be interesting to see what effects of α-toxin beyond cell death are prevented in ΔPLEKHA7 cells, including loss of epithelial and endothelial electrical resistance, inflammasome activation, release of the inflammatory cytokines IL-1β and IL-18, and the impact these changes might have on the formation of an immune response during infection. It remains possible that PLEKHA7’s facilitation of cell death upon α-toxin exposure is a protective mechanism in some situations, perhaps by changing the nature of the immune response.

How Does PLEKHA7 Render Cells Sensitive to α-Toxin?

Perhaps most obviously, the work by Popov et al. (10) raises the question of how PLEKHA7 renders cells sensitive to α-toxin–mediated cell death. Does PLEKHA7 negatively regulate one of the survival pathways mentioned above? Membrane reshuffling by endocytosis is an intriguing possibility, because this process has similar slow recovery kinetics (8 h) to what is observed in the ΔPLKHA7 cells (18). It seems unlikely that a p38 MAPK-dependent mechanism is responsible, because the phosphorylation of p38 in the ΔPLEKHA7 cells was weaker than the phosphorylation of p38 in the WT cells. K⁺ efflux, on the other hand, showed a trend toward being stronger in the ΔPLKHA7 cells, although this trend did not reach significance. Aerolysin, a small pore-forming toxin produced by Aeromonas hydrophila, has been shown to induce K⁺ efflux (22). K⁺ efflux, in turn, triggers, through the assembly of the NALP inflammasome, the activation of sterol regulatory element-binding proteins (SREBPs), key regulators of membrane synthesis, which facilitate recovery from intoxication (22). Interestingly, SREBF2 was also identified in the screen by Popov et al. (10). Thus, it is possible that PLEKHA7 promotes cell death by dampening the K⁺-mediated membrane synthesis in response to α-toxin. Further, K⁺ fluctuations seem to be necessary for the phosphorylation of AMP-activated protein kinase, a key player in the induction of autophagy in response to pore-forming toxins (19), raising the possibility that PLKHA7 negatively regulates the activation of this response. Autophagy protects cells from α-toxin by decreasing the levels of ADAM10 (20), so if PLKHA7 mutes the proautophagy response to α-toxin, this effect would increase the susceptibility of the cell to further attacks by α-toxin.

The other hits of the screen on mutagenized HAP1 cells also bring up interesting new avenues for investigation. The fact that several adherens junction-associated proteins were identified in the screen raises the question of whether PLEKHA7 itself specifically mediates α-toxin susceptibility or if the adherens junction as a whole is involved. Another area of interest would be to investigate which genes are regulated by the transcription factors and chromatin-associated proteins that were enriched in the screen, which may point to new pathways that are important in regulating the susceptibility to α-toxin. Two glucosidases were identified in the screen, both of which are involved in N-glycosylation, which raises the possibility that the glycosylation pattern of ADAM10 could be important for α-toxin binding, or perhaps that glycosylation of adherens junction proteins plays a role in mediating susceptibility.

Finally, important clinical questions are raised by this study as well. Does genetic variation in the genes encoding PLEKHA7 or other adherens junction-associated proteins confer differential susceptibility to S. aureus infection? Additionally, of course, could PLEKHA7 be a drug target? In contrast to ADAM10, PLEKHA7^−/− mice are viable and appear to be normal, so interfering with the function of this protein is a potential therapeutic strategy.

Altogether, the work by Popov et al. (10) provides additional insight into the increasingly complex picture of the interaction between S. aureus toxins and the host.