A requirement for septins and the autophagy receptor p62 in the proliferation of intracellular Shigella

Abstract Shigella flexneri, a Gram‐negative enteroinvasive pathogen, causes inflammatory destruction of the human intestinal epithelium. During infection of epithelial cells, Shigella escape from the phagosome to the cytosol, where they reroute host cell glycolysis to obtain nutrients for proliferation. Septins, a poorly understood component of the cytoskeleton, can entrap cytosolic Shigella targeted to autophagy in cage‐like structures to restrict bacterial proliferation. Although bacterial entrapment by septin caging has been the subject of intense investigation, the role of septins and the autophagy machinery in the proliferation of noncaged Shigella is mostly unknown. Here, we found that intracellular Shigella fail to efficiently proliferate in SEPT2‐, SEPT7‐, or p62/SQSTM1‐depleted cells. Consistent with a failure to proliferate, single cell analysis of bacteria not entrapped in septin cages showed that the number of metabolically active Shigella in septin‐ or p62‐depleted cells is reduced. Targeted metabolomic analysis revealed that host cell glycolysis is dysregulated in septin‐depleted cells, suggesting a key role for septins in modulation of glycolysis. Together, these results suggest that septins and the autophagy machinery may regulate metabolic pathways that promote the proliferation of intracellular Shigella not entrapped in septin cages.

Although host cell metabolism is not significantly affected by the proliferation of intracellular Shigella, rapid catabolism of pyruvate into acetate is a hallmark of Shigella infection (Kentner et al., 2014).
Interestingly, Shigella mutants unable to catabolize pyruvate into acetate can still proliferate intracellularly, albeit at a reduced rate (Kentner et al., 2014), and it has been suggested that alternative metabolic routes and/or carbon energy sources can be used by Shigella to sustain proliferation (Waligora et al., 2014).
During S. flexneri infection, septins entrap actin-polymerizing bacteria targeted to autophagy in cage-like structures (Mostowy et al., 2010;Sirianni et al., 2016). Although bacterial entrapment by septin caging has been the subject of intense investigation, the role of septins and the autophagy machinery in the proliferation of noncaged Shigella is mostly unknown. In this study, we tested the role of septins (SEPT2, SEPT7) and the autophagy receptor p62/SQSTM1 in the proliferation of intracellular Shigella not entrapped in septin cages. Our data reveal a new role for septins and p62 in Shigella proliferation, and an unexpected role for septins in host cell glycolysis.

| R E SULTS AN D DISC USSION
2.1 | A requirement for septins and p62 in the proliferation and metabolic activity of intracellular Shigella not entrapped in septin cages To investigate a role for septins in the proliferation of intracellular Shigella, we performed gentamicin survival assays (Krokowski & Mostowy, 2016).
HeLa cells were treated with small interfering RNA (siRNA) sequence specific for SEPT2 (Figure 1a; Supporting Information Figure S1), and infected with S. flexneri M90T for up to 4 h 40 min. Considering that septins regulate bacterial entry into host cells bacterial burden at 4 h 40 min postinfection was normalized to values at 1 h 40 min postinfection as previously described (Mostowy et al., 2009a,b). Unexpectedly, the depletion of SEPT2 ( Figure 1a) resulted in significantly reduced (2.6 6 0.5-fold) bacterial proliferation, as compared to control cells (Figure 1b). To confirm this unexpected role for septins in Shigella proliferation, we infected HeLa cells treated with siRNA sequence specific for SEPT7 ( Figure 1a). As previously shown (Estey et al., 2010), the depletion of SEPT7 also reduced levels of SEPT2 ( Figure 1a; Supporting Information Figure S1). In agreement with results obtained for SEPT2-depleted cells, Shigella proliferated significantly less (1.5 6 0.1-fold) in SEPT7-depleted cells, as compared to control cells ( Figure 1b). Septins mediate host cell division (Hartwell, 1971;Surka, Tsang, & Trimble, 2002), and may play a role in host cell viability (and therefore Shigella intracellular proliferation). To test this, we measured host cell viability in control-, SEPT2-, or SEPT7-siRNA treated cells. We did not observe significant differences in viability between control and septin depleted cells (Supporting Information Figure S2a).
Factors controlling the proliferation of intracellular Shigella are not fully known. Previous work has shown that septins can modulate autophagy (Barve et al., 2018;Mostowy et al., 2010). Additionally, p62 has been demonstrated to facilitate Salmonella replication inside HeLa cells (Yu et al., 2014). Considering this, we tested the role of p62 in the proliferation of intracellular Shigella. We infected HeLa cells treated with siRNA sequence specific for p62 ( Figure 1c). Consistent with septin-autophagy interactions, the depletion of p62 resulted in significantly reduced (1.7 6 0.2-fold) bacterial proliferation, as compared to control cells ( Figure 1d).
In agreement with results obtained using septin-depleted cells, p62 depletion did not significantly affect host cell viability (Supporting Information Figure S2b). Together, these data suggest a requirement for septins and p62 in the proliferation of intracellular Shigella.
We reasoned that a reduction in the number of metabolically active intracellular bacteria may contribute to the decreased proliferation of Shigella in septin-or p62-depleted cells. To test this, we used a S. flexneri strain (called 'x-light') carrying a GFP-encoding plasmid, where gfp expression is induced upon IPTG exposure (Schlosser-Silverman et al., 2000;Sirianni et al., 2016). In this system, only metabolically active bacteria synthesize GFP upon IPTG exposure. Control, SEPT2-, SEPT7-, or p62-depleted cells were infected with x-light S. flexneri for 4 h 10 min, and IPTG was added for 30 min prior to fixation. After this, the percentage of bacteria responding to IPTG (and therefore metabol-

| Septins control host cell glycolysis
Why is Shigella dependent on septins for intracellular proliferation?
Considering that both the proliferation and metabolic activity of Shigella are reduced by septin depletion, we reasoned that host cell glycolysis (the primary energy source used by intracellular Shigella for proliferation) is also reduced. To test this, we used liquid chromatography-mass spectrometry accurate mass retention time These results suggest that glycolysis may be impaired in septindepleted cells, and therefore glucose cannot be converted into H6P.
Alternatively, glycolytic activity may be increased in septin-depleted cells, and therefore H6P is consumed faster. To distinguish between these two possibilities, we tested the cellular levels of lactate (the final product of glucose fermentation) by LC-MS AMRT analysis.
Consistent with a role for septins in suppression of glycolytic activity, we observed significantly more lactate (2.2 6 0.3 and 2.1 6 0.3fold) in SEPT2-and SEPT7-depleted cells, respectively, as compared to control cells ( Figure 3b).
To support intracellular proliferation in epithelial cells, S. flexneri consumes glycolysis-derived pyruvate (Kentner et al., 2014). Surprisingly, we observed increased pyruvate levels (1.7 6 0.2-fold and 1.5 6 0.2-fold) in SEPT2-and SEPT7-depleted cells, respectively, as we failed to detect significant differences between control and SEPT2-or SEPT7-depleted cells, for any of five TCA cycle metabolites tested (i.e., citrate, aconitate, succinate, fumarate, and malate) ( Figure 3c). Decreased levels of H6P and increased levels of pyruvate and lactate in septin-depleted cells may reflect altered protein levels of glycolytic enzymes. To test this, we measured the amount of four key enzymes of the glycolytic pathway, that is, hexokinase (HK), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), pyruvate kinase (PK), and lactate dehydrogenase (LDH), in control-, SEPT2-, or SEPT7-depleted cells by Western blot. We did not observe any change in the protein levels of HK, GAPDH, PK, or LDH upon septin depletion (Supporting information Figure S3).
Collectively, mass spectrometry analysis revealed that host cell glycolysis, crucial for the proliferation of intracellular Shigella, is dysregulated in septin-depleted cells. Previous work has shown that septin recruitment to intracellular Shigella is dependent on p62 recruitment, and vice versa (Mostowy et al., 2010). Considering the established link between septins and p62, we hypothesized that p62 may modulate glycolysis similarly to septins. We employed LC-MS AMRT to analyze metabolites of the glycolytic pathway in control or p62-depleted cells. In this case, we did not observe significant differences in levels of the glycolytic metabolites tested, indicating a specific role for septins in modulating glycolysis (Supporting Information Figure S4). These results demonstrate that, in the absence of septins or p62, bacterial proliferation is not compromised because of Interestingly, T. gondii induces the autophagy of lipid droplets (i.e., lipophagy) to obtain fatty acids from the host cell required for its proliferation. Similarly, autophagy-derived fatty acids may support the proliferation of intracellular Shigella, as also suggested in the case of Salmonella (Yu et al., 2014). Future work will be required to investigate the underlying requirement for septins and p62 in the metabolism and proliferation of intracellular Shigella.
What is the role of the cytoskeleton in host cell glycolysis? The glycolytic pathway is viewed to constitute a 'metabolon', a multien- work has shown that actin filaments bind glyceraldehyde-3phosphate dehydrogenase (GAPDH), increasing its enzymatic activity (Poglazov & Livanova, 1986). What is the role of septins in glycolysis? Septins form molecular scaffolds and diffusion barriers, interacting with actin for cellular compartmentalization (Mostowy & Cossart, 2012;Saarikangas & Barral, 2011). Considering the roles for actin in glycolysis (Araiza-Olivera et al., 2013;Mejean et al., 1989;Waingeh et al., 2006), it is tempting to speculate that septins can also modulate glycolysis. In agreement with this, our results show that septin depletion significantly increases glycolytic activity ( Figure 3b) without affecting the protein levels of glycolytic enzymes (Supporting Information Figure 3). Here, septins may inhibit glycolytic enzymes, such as hexokinase (that mediates the phosphorylation of glucose into H6P), and therefore inhibit glycolysis. However, in the absence of septins, such enzymes would be free to interact with other components of the pathway, enabling rapid glucose catabolism. In support of this, recent work has shown that actin filaments can regulate glycolysis by inhibiting the glycolytic enzyme aldolase (Hu et al., 2016). In a separate study using human adipocytes, work has shown that SEPT11 interacts with  Kentner et al., 2014;Warburg, 1956).
Other cell types, such as primary colonocytes, can present a different metabolism in which the TCA cycle is primarily used for energy production (Donohoe et al., 2012;Zhang, Wu, Chapkin, & Lupton, 1998). Considering this, it would be interesting to see how septins influence glycolysis and Shigella proliferation in human colonocytes,

| Gentamicin survival assays
HeLa cells were seeded in 6-well plates (Thermo Scientific) and treated with siRNAs as described above. Cell cultures were infected with S.
flexneri at a multiplicity of infection (MOI, bacteria: cell) of 100:1. Bacteria and cells were centrifuged at 1103g for 10 min at room temperature, and then placed at 378C and 5% CO 2 for 30 min. To quantify IPTG-responsive bacteria, images were processed by ImageJ software. Before quantifying GFP positive bacteria brightness and contrast were adjusted for all images to remove noise signal from GFP channel, so only bacteria but not host cells could be seen. The total number of bacteria were counted using Hoechst stain. Then, GFP-negative bacteria were quantified.

| Quantification of dead cells
Cells were seeded in 6-well plates and treated with siRNAs as 1:2 in trypan blue and the number of dead cells was measured by trypan blue dye exclusion using a hemocytometer; dead cells become blue in the presence of the dye.

| Targeted metabolite analysis
HeLa cells were grown in 6-well plates as described above. psig and the nitrogen drying gas flow rate was set at 5 L/min. The drying gas temperature was maintained at 3008C. Data were collected in the centroid mode in the 4 GHz (extended dynamic range) mode. Ion counts were normalized to the amount of biomass of individual samples determined by the residual protein content of metabolite extracts using the BCA assays kit (Thermo ® ).

| Statistics
Statistical analyses were performed using GraphPad Prism (v7, La Jolla, USA). Data are presented as mean 6 standard error of the mean (SEM) from at least three independent experiments per treatment. One-way ANOVA and Student's t test were used to compare values, with p < .05 considered as significant.