LLO intoxication induces tyrosine phosphorylation of NMHC2A and activation of Src kinase

Given that Listeria infection activates Src kinase [19] and leads to Src-dependent phosphorylation of NMHC2A Tyr¹⁵⁸ [17], and because its pore-forming toxin Listeriolysin O (LLO) remodels the NM2A cytoskeleton [4,10], we asked whether LLO is necessary and sufficient to trigger NMHC2A tyrosine phosphorylation and activate Src kinase. Using an anti-pTyr antibody, the total tyrosine phosphorylated proteins were immunoprecipitated (IP) from lysates of HeLa cells under control non-intoxicated conditions (NI) or after incubation with a sub-lytic dose of purified LLO (0.5 nM, 15 min), a condition that allows PM damage repair and prevents cell death [5]. While the levels of NMHC2A in whole cell lysates (WCL) were comparable in NI and LLO-intoxicated samples, NMHC2A was significantly enriched in pTyr IP fractions from LLO-intoxicated cells (Fig 1A and 1B). Similar results were obtained in HeLa cells infected with the wild-type (WT) L. monocytogenes strain (1h, MOI 200) but not in cells infected with the isogenic strain lacking LLO (∆LLO) (S1A and S1B Fig). These indicate that LLO is necessary and sufficient to trigger NMHC2A tyrosine phosphorylation. To further extend our observations, we assessed if other LLO-related bacterial PFTs [1], such as perfringolysin O (PFO) produced by Clostridium perfringens [16] and pneumolysin (PLY) released by Streptococcus pneumoniae [3], also trigger NMHC2A tyrosine phosphorylation. We found that both PFO and PLY trigger the tyrosine phosphorylation of NMHC2A upon intoxication of HeLa cells (S2A-S2D Fig). However, we showed that NMHC2A phosphorylation is impaired upon intoxication of HeLa cells with a pore-deficient PFO variant (PFO inactive) or with LLO in calcium-free medium, demonstrating that this event depends on pore formation and the consequent influx of calcium (S2A and S2B Fig). These findings indicate that NMHC2A tyrosine phosphorylation represents a conserved host cell response to toxin-induced pore formation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. LLO promotes NMHC2A tyrosine phosphorylation and Src activation in HeLa cells.

(A, B) Levels of NMHC2A measured by immunoblots on whole-cell lysates (WCL) and immunoprecipitated (IP) fractions of pTyr proteins (IP pTyr) from non-intoxicated (NI) or LLO-intoxicated (LLO; 0.5 nM, 15 min) HeLa cells. Actin was used as loading control. (B) Levels of NMHC2A in the IP pTyr fraction (NMHC2A pTyr) were quantified and normalized to those detected in the WCL (NMHC2A WCL). Each dot corresponds to an independent experiment. Data correspond to mean ± SEM (n = 5); p-value was calculated using two-tailed unpaired Student’s t-test, **p < 0.01. (C) Levels of Tyr⁴¹⁹-phosphorylated Src (Src pTyr⁴¹⁹) and total Src (Src Total) assessed by immunoblots on WCL of HeLa cells non-intoxicated (NI) or LLO-intoxicated (0.5 nM LLO, 5 min). Actin was used as loading control, long (LE) and short (SE) exposures are shown. (D) Levels of Src pTyr⁴¹⁹ signals were quantified and normalized to the levels of total Src. Each dot corresponds to an independent experiment. Data correspond to mean ± SEM (n = 6); p-values were calculated using two-tailed unpaired Student’s t-test, **p < 0.01. (E) Schematic representation of Src biosensor composed of ECFP, a SH2 domain, a flexible linker, a substrate peptide derived from p130cas, and YPet (22). In conditions where Src is not activated ECFP and YPet are close and FRET occurs. If Src is activated, the SH2 and p130cas domains interact thus increasing the distance between ECFP and Ypet. Consequently, no FRET occurs and the ECFP/YPet ratio increases. (F) Kinetics of donor/acceptor fluorescence emission (ECFP/YPet) of the Src biosensor in NI and LLO-intoxicated cells. LLO was added to the culture medium 10 min after the acquisition started. Values correspond to the mean ± SEM (n > 30); p-values were calculated using one-way ANOVA with Dunnett’s post hoc analysis, ***p < 0.001. (G) Sequential frames of time-lapse confocal microscopy FRET analysis of NI or LLO-intoxicated HeLa cells expressing ECFP/YPet-based Src biosensor. LLO was added to the culture medium 10 min after the acquisition started. Arrows indicate LLO-induced plasma membrane (PM) blebbing sites. Scale bar, 10 µm. (H) Confocal microscopy images of NI or LLO-intoxicated (0.5 nM, 15 min) HeLa cells, immunolabeled for active Src (red) and NMHC2A (green) and stained with DAPI (blue). Insets show LLO-induced cortical accumulations of NMHC2A enriched in active Src. Scale bar, 10 µm.

https://doi.org/10.1371/journal.ppat.1013945.g001

To determine whether LLO-induced NMHC2A phosphorylation correlates with Src activation, we evaluated the levels of activated Src in non-intoxicated (NI) and in LLO-intoxicated HeLa cells. Full activation of Src requires both dephosphorylation of its tyrosine 530 (Tyr⁵³⁰) and phosphorylation of its tyrosine 419 (Tyr⁴¹⁹) [20,21], which can be detected by immunoblot. We found that the levels of Src phosphorylated on Tyr⁴¹⁹ (Src pTyr⁴¹⁹) and dephosphorylated on Tyr⁵³⁰ (Active Src) were increased upon 5 min-LLO intoxication of HeLa cells (Figs 1C, 1D, S1C and S1D). Activation of Src kinase was also detected in HeLa cells intoxicated by PLY, suggesting that Src signaling is a common response to various PFTs (S2E and S2F Fig). In addition, activated Src kinase was detected in HeLa cells infected with the wild-type (WT) L. monocytogenes strain (10 min, MOI 200) but not in those infected with the ∆LLO strain (S1E and S1F Fig). Altogether, these results indicate that LLO is necessary and sufficient to induce Src activation, which presumably might occur upstream of the tyrosine phosphorylation of NMHC2A.

To further confirm this, we examined the activation of Src in single cells using the specific FRET-based ECFP/YPet KRas-Src biosensor [22]. Briefly, this biosensor contains the c-Src SH2 domain, a flexible linker, and a substrate peptide derived from p130cas, which are concatenated between FRET donor ECFP and acceptor YPet (Fig 1E). This Src biosensor has a high FRET signal at rest, due to a strong YPet emission signal when ECFP is excited. As the activated Src phosphorylates the substrate, the biosensor undergoes conformational changes, resulting in the decrease of FRET and YPet emission signals. Therefore, Src activation status can be visualized and quantified by calculating the ratio of CFP/YPet [22]. HeLa cells expressing the biosensor were imaged starting 10 min before LLO was added and the ratio of CFP/YPet signals was calculated over time. The activation levels of Src remained stable before the addition of LLO (Fig 1F and 1G and S1 Movie). However, quickly upon LLO intoxication, the CFP/YPet ratio increased and remained high over time (Fig 1F and 1G and S1 Movie), reflecting the global and lasting Src activation in LLO-intoxicated cells. Of note, increased CFP/YPet values were also detected associated with the PM, where the Src kinase activity was often reported [23] and where the cell is responding to LLO-induced damage by forming blebs (Fig 1G, arrows) [4]. Concurrently, confocal microscopy analysis of NI and LLO-intoxicated HeLa cells showed that, whereas active Src is predominantly cytoplasmic in NI cells, LLO exposure induces its strong enrichment at the cell cortex, colocalizing with the NMHC2A cortical accumulations that serve as hallmarks of PFT-induced PM damage [4,16] (Fig 1H). Similar observations were made in PLY-intoxicated cells (S2G Fig). These results suggest that Src kinase activity at the cell periphery may regulate cytoskeletal remodeling in response to LLO-induced damage. In addition, these data indicate that LLO is sufficient to activate Src, which in turn may control the tyrosine phosphorylation of NMHC2A, as it occurs during Listeria infection [17].

Src kinase activity drives LLO-induced NMHC2A tyrosine phosphorylation

To determine if Src kinase is necessary for the tyrosine phosphorylation of NMHC2A mediated by LLO, we manipulated Src activity by: 1) pharmacological inhibition using Dasatinib (Dasa), 2) downregulation of Src expression using specific shRNAs (shSrc) [17,24], and 3) overexpression of Src kinase mutants. The use of multiple approaches to modulate Src activity ensures data robustness and mitigates concerns related to inhibitor specificity or potential off-target effects. Levels of NMHC2A were assessed in IP pTyr fractions obtained from NI and LLO-intoxicated cells in control or Src perturbed conditions. In control conditions, LLO consistently induced an enrichment of NMHC2A in the IP pTyr fraction (Fig 2A-2F). In contrast, NMHC2A pTyr was reduced upon Dasatinib treatment (Dasa; Fig 2A and 2B), downregulation of Src kinase expression (shSrc; Fig 2C and 2D) or ectopic overexpression of a Src kinase dead mutant (KD; Fig 2E and 2F). Inversely, upon LLO intoxication, the ectopic overexpression of a constitutively active (CA) Src mutant increased NMHC2A pTyr to levels higher than those detected in cells expressing wild-type Src (WT) (Fig 2E and 2F). Notably, low expression levels of Src CA and Src KD when compared to that of WT Src were sufficient to modulate the levels of NMHC2A pTyr (Fig 2E and 2F). These results show that the tyrosine phosphorylation of NMHC2A triggered by LLO depends on the activation of Src kinase.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. LLO-induced tyrosine phosphorylation of NMHC2A requires Src activity and occurs in Tyr¹⁵⁸.

(A, C, E) Immunoblots for NMHC2A, total Src and Src pTyr⁴¹⁹ on WCL or IP pTyr fractions of non-intoxicated (-) or LLO-intoxicated (0.5 nM, 15 min) (+) HeLa cells (A) in the absence (Mock) or presence of Dasatinib (300 nM, 1 h) (Dasa); (C) expressing control (shCtr) or Src-specific oligos (shSrc), and (E) ectopically expressing wild-type Src (WT), constitutively active Src (CA) or kinase dead Src (KD) variants. Actin was used as loading control. (B, D, F) Shows the quantified levels of NMHC2A in IP fractions (NMHC2A pTyr) normalized to those detected in the WCL (NMHC2A WCL), under the different conditions described above. Each dot corresponds to an independent experiment. Values correspond to the mean ± SEM (n ≥ 3) and p-values were calculated using (B, F) one-way ANOVA with Dunnett’s post hoc analysis, *p < 0.05, **p < 0.01 or (D) two-tailed unpaired Student’s t-test, ***p < 0.001. (G, H) Confocal microscopy images of NI or LLO-intoxicated (0.5 nM, 15 min) HeLa cells, immunolabeled for NMHC2A phosphorylated in Tyr¹⁵⁸ (pY158, green) and total NMHC2A (red in G), and stained for F-actin (red in H). DAPI was used to stain the nuclei (blue). Insets in NI cells show defined puncta structures at the cell periphery, which are associated to the extremities of stress fibers detected by F-actin staining (H). The insets in LLO-intoxicated cells show cortical accumulations of NMHC2A phosphorylated in Tyr¹⁵⁸ (pY158, green in G, H), total NMHC2A (red in G) and F-actin (red in H). Scale bar, 10 µm. (I, J) Levels of NMHC2A phosphorylated in Tyr¹⁵⁸ detected at the cortex of NI and LLO-intoxicated cells. (I) Schematic representation of the protocol used to quantify the pY158 signal specifically at the cortical and sub-cortical regions, detailed in the Material and Methods section. (J) Quantification of Y158 signal (integrated density values) in confocal images similar to those shown in G and H. Values correspond to the mean ± SEM (n = 83) and p-values were calculated using two-way ANOVA with Tukey’s post hoc analysis, *p < 0.05, ***p < 0.001 (K) Dot blots for pY158 NMHC2A and total NMHC2A performed using 10 and 20 µg of total protein extracts from NI or LLO-intoxicated (0.5 nM, 10 min) HeLa cells.

https://doi.org/10.1371/journal.ppat.1013945.g002

We previously demonstrated that Src-dependent phosphorylation of NMHC2A triggered by Listeria infection, occurs at tyrosine residue in position 158 (Tyr¹⁵⁸) [17]. Given our findings described above, we hypothesized that LLO is the bacterial factor that drives the Src-dependent phosphorylation of NMHC2A Tyr¹⁵⁸. To verify this, HeLa cells, either NI or incubated with LLO were fixed and immunostained using an antibody raised against phosphorylated NMHC2A Tyr¹⁵⁸ (anti-pY158) [17]. In NI conditions, the anti-pY158 antibody revealed faint cytoplasmic labeling with defined puncta located at the cell periphery (Fig 2G). We detected pY158 labeling associated with the extremities of stress fibers detected by F-actin staining (Fig 2H). These observations are in agreement with our previous data showing that cells overexpressing a non-phosphorylatable form of NMHC2A (NMHC2A Y158F) display altered focal adhesions [24]. In response to LLO intoxication, the pY158 labeling significantly accumulates at the cell cortex at sites where NMHC2A is also accumulating (Fig 2G) [4] and where actomyosin undergoes remodeling (Fig 2H) [4]. Quantification of the pY158 signal at the cortex of the cell revealed a significant increase in LLO intoxicated cells as compared to NI resting cells (Fig 2I and 2J). The increased phosphorylation in NMHC2A residue Tyr¹⁵⁸ was also confirmed by dot blot using the anti-pY158 (Fig 2K). While the levels of total NMHC2A remain similar in NI and LLO-intoxicated cells (Fig 2K, right panel), the levels of NMHC2A pY158 are higher in intoxicated cells (Fig 2K, left panel). These data indicate that LLO induced the Src-dependent phosphorylation of NMHC2A in the residue Tyr¹⁵⁸.

Src kinase coordinates LLO-induced NMHC2A remodeling

Next, we investigated whether Src kinase interferes with the accumulation of cortical NMHC2A induced by LLO, which constitutes a hallmark for PM repair and for the establishment of host survival responses to PFTs [4]. HeLa cells pre-incubated with the Src pharmacological inhibitor Dasatinib (Dasa) [25] or expressing low levels of Src (through specific shRNAs) were challenged with LLO, immunostained for NMHC2A and analyzed by confocal microscopy (Figs 3A and S3A). The percentage of cells displaying LLO-induced cortical accumulations of NMHC2A was not altered by perturbing either Src activity or expression (Fig 3B). However, while the majority of control cells displayed 1 or 2 NMHC2A accumulations, the percentage of cells displaying at least 3 accumulations was significantly increased in Src-deficient cells (Fig 3C and 3D). These data suggest that, while dispensable for the cortical accumulation of NMHC2A induced by LLO, Src activity is necessary to limit the remodeling of the cytoskeleton in response to intoxication. We thus hypothesized that Src activity is critical to restrain PM damage. We measured, by flow cytometry, the PM permeabilization induced by sub-lytic doses of LLO following propidium iodide (PI) incorporation in NI or LLO-intoxicated HeLa cells, under control or Src-deficient conditions. In such conditions LLO intoxication does not cause cell death but repair machinery is activated [5]. As expected, the percentage of PI-positive cells significantly increased after LLO-intoxication (Figs 3E and S4). Despite the moderate differences, this increase was significantly more pronounced in cells with impaired Src activity (Dasa treated) or with low Src expression (shSrc) (Fig 3E), indicating that Src activity limits LLO-induced PM damage. Additionally, we measured, by high-content microscopy, the PM permeabilization induced by increasing concentrations of LLO over time following Draq7 incorporation in NI or LLO-intoxicated HeLa cells, under control or Src-impaired conditions. Draq7 is a non-toxic, far-red fluorescent, membrane-impermeable DNA dye that selectively labels cells with compromised membrane integrity, analogous to PI [26]. As before, the percentage of Draq7-positive cells significantly increased over time after intoxication with 0.5 nM and 1 nM of LLO (S3B Fig). This increase was significantly more pronounced in cells with impaired Src activity (Dasa treated). Importantly, the LLO concentrations used did not induce extensive cell death over the duration of the experiment, as the total number of cells in the end of the experiment remained similar across conditions (S3C Fig) and remained unaltered throughout time (S3D Fig), indicating that Draq7 positivity reflects transient PM permeabilization rather than loss of cell viability. These results are consistent with a role for Src activity in regulating the actomyosin cytoskeleton remodeling, and contributing to limiting PM damage upon LLO intoxication, either by restricting pore formation and/or by promoting cellular responses that facilitate pore removal or membrane repair.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Src kinase orchestrates the response to LLO-intoxication by controlling actomyosin remodeling in HeLa cells.

(A) Confocal microscopy images of LLO-intoxicated (0.5 nM, 15 min) HeLa cells in control (Mock and shCtr, respectively black circles and black triangles) and Src-impaired (Dasatinib-treated and shSrc, respectively blue circles and blue triangles) conditions. Cells were immunolabeled for NMHC2A (greyscale) and stained with DAPI (blue). White arrows indicate NMHC2A cortical accumulations. Scale bar, 10 µm. (B) Percentage of LLO-intoxicated cells displaying cortical accumulations of NMHC2A under the conditions described in (A). Values are the mean ± SEM (n = 3). (C, D) Percentage of LLO-intoxicated cells showing 1-2, 3-5 or >5 NMHC2A accumulations, under the conditions described in (A). Values are the mean ± SEM (n = 3); p-values were calculated using two-way ANOVA with Sidak’s post hoc analysis, *p < 0.05, **p < 0.01, ***p < 0.001. (C) Shows data from cells under Src chemical inhibition (Dasa). (D) Shows data from cells in which Src expression was downregulated through shRNAs. (E) Percentage of PI-positive HeLa cells in non-intoxicated (NI) conditions or LLO-intoxicated (LLO) as in (A), measured by flow cytometry analysis. Values are the mean ± SEM (n = 4); p-values were calculated using one-way ANOVA with Dunnett’s post hoc analysis, *p < 0.05.

https://doi.org/10.1371/journal.ppat.1013945.g003

NMHC2A pTyr does not affect the kinetic and structural properties of NM2A in vitro

We previously reported that the phosphorylation status of NMHC2A Tyr¹⁵⁸ regulates cytoskeletal organization, affects the assembly/disassembly of focal adhesions, and interferes with cell migration [24]. Tyr¹⁵⁸ is localized in the proximity of the NM2A ATPase domain [17] (S5A Fig) and is highly conserved across evolution (S5B Fig). These observations led us to hypothesize that the phosphorylation status of Tyr¹⁵⁸ may directly interfere with NMHC2A activity. We investigated if phosphorylation of Tyr¹⁵⁸ perturbs NM2A ATPase activity or actin-binding capacity, by determining the mechanical and kinetic parameters of human NMHC2A heavy meromyosin (HMM) fragments, which do not form filaments but retain ATPase and actin-binding/translocating activities. We used HMM fragments derived from wild-type NMHC2A (NM2A-GFP-Flag^HMM-WT), a non-phosphorylatable mutant in which Tyr¹⁵⁸ was substituted by a phenylalanine (Phe, NM2A-GFP-Flag^HMM-Y158F), and a mutant mimicking a permanent phosphorylation in which Tyr¹⁵⁸ was replaced by a glutamate (Glu, NM2A-GFP-Flag^HMM-Y158E). Such HMM fragments were co-expressed and co-purified with human RLC and mouse ELC from Sf9 insect cells (S6A Fig). Purified molecules were analyzed by electron microscopy and no strong conformational changes were detected among the different HMM variants (NM2A-GFP-Flag^{HMM-WT/Y158F/Y158E}, S6B Fig), suggesting that the substitution of Tyr¹⁵⁸ by a Phe or a Glu does not affect the overall HMM structure. The enzymatic activity of the NM2A-GFP-Flag^HMM variants was measured in a steady-state ATPase assay. Although the mutant molecules appeared to hydrolyze ATP at a slightly higher rate than the WT NM2A-GFP-Flag^HMM (Fig 4A), this difference was not significant and the K_ATPase (concentration of actin required for one-half V_max) and V_max values obtained were similar (Fig 4B), and comparable to those previously described for WT NM2A [27]. This indicates that the tyrosine phosphorylation status of the NMHC2A does not affect the ATPase activity of NM2A. The binding affinity of NM2A-GFP-Flag^HMM variants to actin was then assessed by co-sedimentation assays in the absence or presence of ATP. Actin alone was pelleted by ultracentrifugation, whereas NM2A-GFP-Flag^HMM variants remained in the supernatant in the absence of actin (Fig 4C). NM2A-GFP-Flag^HMM molecules co-sedimented with actin in the absence of ATP but remained in supernatants in the presence of 1 mM ATP (Fig 4C). The fraction of each NM2A-GFP-Flag^HMM variant bound to actin in the presence of ATP was comparable in all conditions (Fig 4D). Corroborating these data, electron microscopy analysis of mixtures of actin and the different NM2A-GFP-Flag^HMM variants in the presence or absence of ATP showed similar behaviors (S6C Fig). Finally, we assessed if NMHC2A Tyr¹⁵⁸ is involved in the velocity with which actin filaments slide on NM2A-GFP-Flag^HMM. We performed in vitro actin-gliding assays using coverslip-immobilized NM2A-GFP-Flag^HMM. All the NM2A-GFP-Flag^HMM variants translocated actin at similar rates (Fig 4E and S2 Movie), comparable to those previously reported for human NM2A [28].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Phosphorylation of NMHC2A in Tyr¹⁵⁸ does not affect NM2A mechanical or kinetic properties.

(A) Actin-activated Mg-ATPase activity of heavy meromyosin (HMM) NM2A-GFP-Flag^HMM -WT, -Y158F and -Y158E determined by the conversion of NADH into NAD⁺. Three independent preparations of each NM2A-GFP-Flag^HMM variant were used in at least 3 independent experiments. Data sets were fitted to a hyperbolic equation to determine the kinetic constants, K_ATPase and V_max. (B) Plot of the K_ATPase and V_max values. Data represent the mean ± SEM of at least 6 independent experiments (per actin concentration) and using the 3 different preparations of NM2A-GFP-Flag^HMM variants. Each dot represents the value obtained for a single preparation. Values are the mean ± SEM (n = 3). (C) Representative actin co-sedimentation assay of NM2A-GFP-Flag^HMM WT, -Y158F and -Y158E. Supernatants (S) and pellets (P) obtained from ultracentrifugation of mixtures of either NM2A-GFP-Flag^HMM -WT, -Y158F or -Y158E with or without 10 µM actin and 1mM ATP, are shown by Coomassie blue staining. Actin alone is also shown. Regulatory light chain (RLC); Essential light chain (ELC). (D) Quantification of co-sedimentation assays showing myosin fraction bound to actin (pellet) in the presence of ATP. Each dot corresponds to a single preparation of each NM2A-GFP-Flag^HMM variant. Values are the mean ± SEM (n = 3). (E) Gaussian distribution of the velocity of actin filaments moving on top of either NM2A-GFP-Flag^HMM-WT, -Y158F or -Y158E. Values were obtained from at least 5 independent motility assays (S2 Movie) performed on 3 independent NM2A-GFP-Flag^HMM preparations. 500 to 1300 filaments were quantified for each condition and per independent experiment.

https://doi.org/10.1371/journal.ppat.1013945.g004

These results demonstrate that the substitution of the Tyr¹⁵⁸ by a Phe or Glu does not significantly affect the in vitro biochemical properties of NM2A, including ATPase activity and actin binding, indicating that NMHC2A pTyr¹⁵⁸ does not directly regulate these intrinsic motor functions under purified conditions. While these findings establish that Tyr158 phosphorylation is dispensable for NM2A motor activity in vitro, they do not allow direct assessment of the relevance of these biochemical properties in cells. Thus, our data are consistent with the idea that LLO-induced remodeling of NMHC2A is not driven by changes in ATPase activity but may instead involve regulation of filament assembly and disassembly [24]; this interpretation remains to be fully established.

Tyrosine phosphorylation of the C. elegans NMY-2 is required for survival responses triggered by pore-forming toxins and heat shock

Our data led us to hypothesize that, although phosphorylation of NMHC2A at Tyr¹⁵⁸ does not affect the biochemical properties of NM2A, it may play an important role in cellular responses to infections by PFT-producing bacteria. We therefore used the C. elegans model to test whether NMHC2A pTyr¹⁵⁸ contributes to organismal resistance to PFTs.

C. elegans NMY-2 is a homolog to the human NMHC2A (S7 Fig), and NMY-2 function has been linked to Src-mediated signaling [29]. Using CRISPR/Cas9 genome editing, we introduced point mutations at the conserved tyrosine 163 (Tyr¹⁶³), corresponding to Tyr¹⁵⁸ in NMHC2A (S7 Fig), by replacing it with either a non-phosphorylatable (Y163F) or a phosphomimetic (Y163E) residue in the endogenous nmy-2 locus.

We successfully generated a homozygous strain expressing NMY-2(Y163F), which showed normal brood size and embryonic viability comparable to the wild-type (Fig 5A). In contrast, we were unable to propagate nmy-2(Y163E) animals, as homozygous animals were sterile and therefore unable to lay embryos, suggesting that constitutive phosphorylation of Tyr¹⁶³ may impact gonad function. As the animals expressing NMY-2(Y163E) could not be propagated, we focused our analysis on wild-type and nmy-2(Y163F) animals.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Regulation of NMY2 pTyr¹⁶³ is necessary for the survival of C. elegans to intoxication and heat stress.

(A) Mean brood size and embryonic viability ± SEM in C. elegans expressing wild-type NMY-2(WT), NMY-2(Y163F) or NMY-2(Y163E). On the left, each dot represents the brood size of a single animal. On the right, each dot represents the embryonic viability (in %) of embryos laid by each animal. (B-D) Viability scores based on movement of wild-type (WT) and nmy-2(Y163F) animals after 5-day exposure to growing concentrations of (B) LLO (C) PFO and (D) PFO inactive. Data are from three independent experiments performed in triplicate. p-values were calculated using two-way ANOVA with Sidak’s post hoc analysis, *p < 0.05. (E) Mean percentage of survival ± SD of wild-type (WT) and nmy-2(Y163F) animals fed for 5 days with control E. coli (NI WT; NI Y163F) or with E. coli expressing the Cry5Ba toxin (Cry5Ba WT; Cry5Ba Y163F). Each curve represents the mean of 3 independent experiments with 30 animals per condition. p-values were calculated using two-way ANOVA with Sidak’s post hoc analysis, ***p < 0.001. (F) Mean percentage of survival ± SEM of wild-type and nmy-2(Y163F) animals 24 h post a 1 h heat shock at 35 ºC. Each dot represents one independent experiment with 8-10 animals per condition. p-values were calculated using two-tailed paired Student’s t-test, *p < 0.05. (G) Mean brood size ± SEM of the animals that survived the 1 h heat shock at 35 ºC. Each dot represents the brood size of a single animal. (H) Mean percentage of viability ± SEM of the embryos laid by the animals that survived the 1 h heat shock at 35 ºC. Each dot represents the percentage of viable embryos laid by a single animal. p-values were calculated using two-tailed paired Student’s t-test, *p < 0.05.

https://doi.org/10.1371/journal.ppat.1013945.g005

We assessed whole-animal intoxication using dose–response survival assays across increasing concentrations of different PFTs, including LLO, PFO and a pore-deficient PFO (PFO inactive). Survival of wild-type animals was not significantly affected by any of the toxins at the concentrations tested (Fig 5B-5D). In contrast, nmy-2(Y163F) animals showed a dose-dependent reduction in survival in response to increasing concentrations of both LLO and PFO (Fig 5B and 5C). Importantly, the inactive PFO variant did not cause significant toxicity in either wild-type or nmy-2(Y163F) animals (Fig 5D), indicating that the observed effects depend on pore-forming activity. Together, these data support a conserved functional role for NMHC2A/NMY-2 tyrosine phosphorylation in organismal responses to PFTs intoxication.

C. elegans is naturally susceptible to infection by Bacillus thuringiensis, which produces the PFT Cry5Ba and reduces nematode viability [30]. This infection system is a well-established in vivo model for dissecting host defense pathways against PFTs [31–34]. Motivated by the strong track record of this model, we examined the susceptibility of nmy-2(Y163F) mutant animals to Cry5Ba. Wild-type and nmy-2(Y163F) animals were fed with control or Cry5Ba-expressing E. coli starting with L4 larvae and the number of survivors was quantified daily, over a period of 5 days. While wild-type animals fed with control E. coli (NI WT) survived and appeared healthy over time, those fed with Cry5Ba-expressing E. coli (Cry5Ba WT) showed decreased survival (Fig 5E). Similarly, nmy-2(Y163F) animals showed normal survival when fed with control E. coli, but were significantly more susceptible to Cry5Ba-expressing E. coli than those expressing wild-type NMY-2 (Fig 5E). Our data suggest that proper regulation of NMY-2 phosphorylation at Tyr¹⁶³ contributes to mounting an effective response to Cry5Ba intoxication in this system.

To address the possibility that NMY-2 phosphorylation at Tyr¹⁶³ might also be important for the response to other sources of stress, we assessed its role in heat shock conditions. Testing heat shock stress was a logical choice, as our previous work demonstrated that PFTs target the endoplasmic reticulum (ER) and affect heat shock proteins, including the ER-resident chaperone Gp96 [4]. L4 larvae were subjected to heat shock for 1 h at 35 ºC and worm survival was assessed 24 h later. nmy-2(Y163F) animals showed increased susceptibility to heat shock (Fig 5F). Moreover, the number of embryos laid by heat-shocked wild-type and nmy-2(Y163F) animals was similar (Fig 5G), but embryo viability was lower for heat-shocked nmy-2(Y163F) progenitors (Fig 5H).

Altogether, these data demonstrate that in C. elegans the regulation of NMY-2 Tyr¹⁶³ phosphorylation is important for survival responses to heat shock and PFT intoxication, suggesting a broad involvement in response to different stresses.

Cell lines

HeLa cells (ATCC CCL‐2) were grown in DMEM with glucose and L‐glutamine (Lonza), supplemented with 10% fetal bovine serum (FBS; Biowest) and maintained at 37 °C in a 5% CO₂ atmosphere. Sf9 cells (ATCC CRL-1711) were grown in Grace’s Insect Medium (Gibco) supplemented with 10% FBS (Biowest) and maintained at 27 °C in a humidified atmosphere without CO₂.

Reagents, toxins, antibodies and dyes

Dasatinib (Santa Cruz Biotechnology) was used at 300 nM in complete medium for 1 h prior to LLO intoxication. LLO, PLY, PFO and PFO inactive were expressed in E. coli BL21(DE3) and purified as described in [59]. Intoxications and washes were carried out in Hanks’ balanced salt solution (HBSS, Lonza). The following antibodies were used at 1/200 for immunofluorescence microscopy (IF) or 1/1,000 for immunoblotting (IB): rabbit anti-Src pTyr⁴¹⁹ (#44-660G, Thermo Fisher), mouse anti-active Src (#AHO0051, Thermo Fisher), rabbit anti-Src (#sc-18, Santa Cruz), mouse anti‐β-actin (#A1978, Sigma); mouse anti-phosphotyrosine (#05–321, Millipore); rabbit anti‐NMHC2A (#M8064, Sigma); rabbit anti-NMHC2A pY158 (#AP3775a, Abgent). For IF analysis, DNA was stained with 4′,6‐diamidino‐2‐phenylindole dihydrochloride (DAPI; Sigma) and actin with Rhodamine Phalloidin (Thermo Fisher Scientific) at 1/200; secondary antibodies were used at 1/500: goat anti‐rabbit Alexa Fluor 488 (Invitrogen), goat anti-rabbit Alexa Fluor 594 (Invitrogen), goat anti‐mouse Cy3 (Jackson ImmunoResearch). For IB, goat anti‐rabbit or anti‐mouse HRP (PARIS) and donkey anti-rabbit or anti-mouse IRDye 680 or 800 (LI-COR Biosciences) were used at 1/10,000. For dye uptake assays using high-content microscopy, PhenoVue Draq7 (Perkin Elmer ILC) was used at 60 nM and Hoechst 33342 (Perkin Elmer ILC) was used at 100 ng/ml.

Plasmids

Kras-Src FRET biosensor was a gift from Yingxiao Wang (Addgene plasmid # 78302) [22]. Plasmids allowing the expression of wild-type Src kinase (WT), constitutively active Src (CA) and kinase dead Src (KD) were kindly provided by S. J. Parsons (University of Virginia) [60]. pFastBac-NMHC2A^HMM-GFP-Flag-WT, pFastBac-RLC and -ELC were described in [61]. Insertion of point mutations, Y158F and -Y158E, was achieved by site-directed mutagenesis using QuickChange II Site-directed mutagenesis kit (Agilent) as described in [17]. pET28 His6-mEGFP-D4, allowing the expression and purification of PFO inactive, was a gift from Catherine Tomasetto (Addgene plasmid # 226372) [62].

Infection experiments

Wild‐type (EGDe, WT) and ΔLLO strains of L. monocytogenes were grown overnight at 37°C, with shaking, in brain–heart infusion (BHI; BD‐Difco), sub‐cultured at a 1:10 dilution until optical density of 0.6–0.8, washed in DMEM and inoculated into cells at a multiplicity of infection of 200.

Immunoprecipitation

HeLa cells (6 × 10⁶ cells) treated as indicated or infected for 10 minutes or 1h, were washed with PBS 1x and lysed in RIPA buffer (Santa Cruz Biotechnology). Lysates were recovered after centrifugation at 15,000 xg (10 min at 4 °C). Cell lysates (approximately 500 μg) were pre‐cleared with protein A‐immobilized PureProteome magnetic beads (Millipore) and incubated overnight (4 °C) with 1.5 μg of anti‐phosphotyrosine antibody (4G10 Millipore). Immune complexes were captured with 50 μl of protein G‐immobilized magnetic beads (Millipore) and processed for immunoblotting.

Immunoblot assays

Protein extracts of HeLa cells were recovered in sample buffer (0.25 mM Tris-Cl, pH 6.8; 10% SDS; 50% glycerol; and 5% β‐mercaptoethanol), resolved by SDS–PAGE (10% acrylamide) and transferred onto nitrocellulose membranes using a TransBlot Turbo system (BioRad), at 0.3 A for 1 h. Primary and secondary HRP‐conjugated antibodies were diluted in TBS-Tween 0.1% (150 mM NaCl; 50 mM Tris–HCl, pH 7.4; and 0.1% Tween) with 5% (m/v) milk. Washes were performed with TBS-Tween 0.2%. Signal was detected using Western Blotting Substrate (Thermo Fisher Scientific) and collected in a ChemiDoc XRS+ System with Image Lab Software (BioRad). Alternatively, the signal was detected using an Odyssey CLx Imager with Image Studio v5.2 Software (LI-COR Biosciences). For dot blots, samples were lysed in RIPA buffer (Santa Cruz Biotechnology) supplemented with protease and phosphatase inhibitors (Roche). Protein concentration for each sample was quantified using Bradford reagent (Biorad) and 10 and 20 μg of protein were spotted onto a nitrocellulose membrane. After complete drying for 30 min at room temperature, the membranes were blocked and primary and secondary HRP‐conjugated antibodies were diluted in TBS-Tween 0.1% with 5% BSA (m/v). The signal was developed using Western Blotting Substrate (Thermo Fisher Scientific) and detected in Chemidoc Touch Gel Imaging System (BioRad).

Immunofluorescence microscopy

Cells were fixed in 3% paraformaldehyde (15 min), quenched with 20 mM NH₄Cl (1 h) and permeabilized with 0.1% Triton X‐100 (5 min). Coverslips were incubated for 1 h with primary antibodies, washed three times in PBS 1x and incubated 45 min with secondary antibodies and when indicated, stained with Rhodamine Phalloidin and DAPI. Antibodies and dyes were diluted in PBS containing 1% BSA. Coverslips were mounted onto microscope slides with Aqua‐Poly/Mount (Polysciences). Images were collected with a confocal laser‐scanning microscope (Leica SP5 II or SP8) and processed using Fiji or Adobe Photoshop software. The percentage of cells with NMHC2A accumulations and the number of those accumulations per cell were blindly quantified in at least 200 cells per sample. Positive cells displayed at least a distinct cortical focus enriched in NMHC2A, each focus was quantified as a single NMHC2A accumulation. The intensity of the pY158 signal at cortical and sub-cortical regions was calculated using ImageJ software. Cell outlines were manually defined based on F-actin staining to generate total cell ROIs. A reduced ROI was produced by proportionally shrinking the original ROI using the Enlarge function. The PM-associated region was defined by subtracting the reduced ROI from the total cell ROI using the XOR operation, yielding a peripheral band of constant width in which pY158 integrated density was measured. Sub-cortical pY158 was quantified by further proportionally reducing and subtracting the cortical ROI, generating a second concentric band immediately beneath the PM.

Dye Incorporation assay coupled to high-content microscopy

HeLa cells (5 × 10⁴) treated with DMSO or 300 nM Dasatinib were left non-intoxicated or were intoxicated with increasing concentrations of LLO (0.25 nM, 0.5 nM, and 1 nM diluted in HBSS with calcium and magnesium) in the presence of Hoechst (100 ng/mL) and Draq7 (60 nM) nuclear dyes. Fluorescence intensity was measured using a high-content widefield microscope (Operetta CLS Revvity) using a 20x water and 1.0 NA objective, capturing 17 fields per well (approximately 3000 cells per well), with a time step of 5 minutes for a total of 45 minutes. Data were analyz ed using Harmony software (Revvity). Nuclei labeled by Hoechst (total cells) and by Draq7 (LLO-permeabilized cells) were automatically segmented. Draq7 uptake was expressed as the percentage of Draq7-positive nuclei relative to Hoechst-positive nuclei.

Fluorescence resonance energy transfer (FRET)

HeLa cells were seeded into Ibitreat μ‐dishes (Ibidi), transfected to express ECFP/YPet Kras-Src biosensor [22], maintained in HBSS or HBSS supplemented with LLO (2 nM) at 37 °C in 5% CO₂, and imaged using a confocal laser‐scanning microscope (Leica SP8) equipped with a 40x 1.4 NA objective. Cells were imaged for 10 min prior LLO addition. Image acquisition under LLO-intoxicated conditions started 20 s after LLO addition. ECFP/YPet fluorescence data sets with 0.42 μm z‐steps were acquired every 30 s. Fiji was used for image sequence analysis and video assembly. The ECFP/YPet ratio was quantified on the entire surface of at least 30 cells.

Flow cytometry

For flow cytometry, 5x10⁵ cells seeded in six‐well plates 24 h before treatment, were intoxicated as indicated and washed twice in cold PBS 1x, trypsinized and resuspended in 0.5 ml of cold PBS 1x supplemented with 2% FBS. They were further incubated with 2 μg/ml of Propidium Iodide (PI, Sigma) 5 min before flow cytometry analysis. At least 30000 events per sample were analyzed in a FACS Calibur cytometer (BD Biosciences), and data were analyz ed using FlowJo (TreeStar).

NM2A-GFP-Flag^HMM-WT, -Y158F and -Y158E expression in Sf9 cells

NM2A-GFP-Flag^HMM variants were expressed in Sf9 insect cells, following the Bac-to-Bac Baculovirus Expression System (Life Technologies) as recommended. Briefly, Sf9 cells were co-transfected with the pFastBac-NMHC2A^HMM-GFP-Flag-WT, or -Y158F, or -Y158E, pFastBac-RLC and -ELC bacmids, and incubated at 27 ºC for 96–120 h. Human RLC (NP_291024.1) and mouse ELC (NP_034990.1) were used [63]. Supernatants were collected to harvest the viral stock (P1). Sf9 cells in suspension were then infected with P1 (MOI = 1) to obtain a higher viral titer (P2) and incubated at 27 ºC with shaking. The process was repeated and the viral titer P3 was used to co-infect Sf9 cells for further expression and purification of the HMMs. Sf9 P3-infected cells were harvested after 72 h and used for purification or stored at -80 ºC.

NM2A-GFP-Flag^HMM-WT, -Y158F and -Y158E purification

Sf9 infected-cell pellets were resuspended in extraction buffer (15 mM MOPS pH 7.3, 200 mM NaCl, 10 mM MgCl₂, 1 mM EGTA, 3 mM NaN_3, 1mM DTT, 0.1 mM PMSF, 5 μg/ml leupeptin, 2 phosphatase inhibitors tablets) and lysed by sonication. The lysate was incubated with 1mM ATP for 15 min, centrifuged at 19000 rpm for 20 min and incubated with pre-washed anti-Flag-resin (#A2220, Sigma) for 3 h at 4 ºC. Flag-resin was washed with high ionic strength buffer (10 mM MOPS pH 7.3, 0.1 mM EGTA, 3 mM NaN₃, 0.5 M NaCl, 1mM ATP, 5 mM MgCl₂ and 0.1 mM PMSF), loaded into a column (4 ºC) and further washed with 3 column volumes of low ionic strength buffer (10 mM MOPS pH 7.3, 0.1 mM EGTA, 3 mM NaN₃ and 0.1 mM PMSF). Proteins were eluted with 2 column volumes of Flag-peptide buffer (0.5 mg/ml Flag peptide, 10 mM MOPS pH 7.3, 0.1 mM EGTA, 3 mM NaN₃, 100 mM NaCl and 0.1 mM PMSF). Eluted fractions were concentrated and dialyzed overnight in the dialysis buffer (10 mM MOPS pH 7.3, 0.1 mM EGTA, 3 mM NaN₃, 0.5 M NaCl and 1mM DTT) at 4 ºC. Aliquots of the purified proteins were quickly frozen in liquid nitrogen and stored at -80 ºC.

Single molecule electron microscopy

For NM2A-GFP-Flag^HMM molecules in the absence of actin, samples were diluted to 50 nM HMM in 10 mM MOPS pH 7.0, 50 mM NaCl, 2 mM MgCl₂, 0.1 mM EGTA. NM2A-GFP-Flag^HMM molecules were mixed with F-actin to give final concentrations of 100 nM of HMM and 500 nM of F-actin. Samples were handled as previously described [64]. Briefly, 5 μl of the sample was applied to a carbon-coated copper grid (pre-exposed to UV light to produce a hydrophilic surface) and stained with 1% uranyl acetate. Data were recorded at room temperature on a JEOL 1200EX II microscope equipped with an AMT XR-60 CCD camera. The Fiji FFT bandpass filter (40 – 2 pixels, auto-scaled) was applied to images prior to making figures, in order to enhance the clarity of individual molecules.

ATPase activity assay

Actin-activated Mg²⁺-ATPase activities were determined by previously described NADH-linked assay [65,66]. Purified NM2A-GFP-Flag^HMM variants were phosphorylated for at least 30 min at room temperature, in adequate buffer conditions (0.2 mM ATP, 0.2 mM CaCl₂, 10 µg/ml MLCK and 1 µM CaM). Mg²⁺-ATPase activities were determined in 10 mM MOPS, 0.1 mM EGTA, 50 mM NaCl, 2mM MgCl₂, 3.4 mM ATP, growing actin concentrations (0–120 µM), 40 units/ml lactate dehydrogenase, 200 units/ml pyruvate kinase, 1 mM phosphoenolpyruvate, 0.2 mM NADH and non-phosphorylated or phosphorylated HMMs (0.1 – 0.2 µM). The conversion of NADH into NAD⁺ was determined by measuring the A₃₄₀ (∊ = 6220 M^{-1 cm-1}) every second for 20 min. Data were corrected for the background Mg²⁺-ATPase activity of actin alone. The Mg-ATPase activity of the different NM2A-GFP-Flag^HMM variants at 0 µM actin concentration was further subtracted to each data point. To determine the kinetic constants V_max and K_ATPase, the experimental data sets were fitted to the Michaelis-Menten mathematical equation using Prism 8.

Co-sedimentation assay

2 µM of each NM2A-GFP-Flag^HMM variant was incubated for 30 min at room temperature with 10 µM f-actin in the absence or presence of 1 mM ATP, in a buffer containing 10 mM MOPS, 0.1 mM EGTA, 100 mM NaCl, 2 mM MgCl₂, 3 mM NaN₃, and 1 mM DTT. Samples containing only the NM2A-GFP-Flag^HMM variants or filamentous actin were used as sedimentation controls. Samples were centrifuged at 100,000 xg for 30 min at 4 ºC. Supernatant and pellet fractions were recovered and the loading buffer was added equivalently. Each sample (20 µl) was analyzed on a 4–12% Bis-Tris gel (Life Technologies), stained with Coomassie Blue and scanned on an Odyssey system (Li-Cor Biosciences).

In vitro motility assay

In vitro motility assays were performed as described [67]. Briefly, 60 μl of NM2A-GFP-Flag^HMM molecules (0.2 mg/ml) in motility buffer (MB: 20 mM MOPS, pH 7.3, 0.1 mM EGTA, 2 mM MgCl₂) with 0.5 M NaCl were trapped in a flow cell. NM2A-GFP-Flag^HMM molecules attached to the coverslip were further incubated with 50 μl MB with 0.5 M NaCl and 1 mg/ml BSA, washed in MB with 50 mM NaCl (LS buffer) and incubated for 4 min with 35 μl of 1 mM ATP, 0.2 mM CaCl₂, 1mM ATP, 1 μM CaM, 1 nM MLCK and 10 μM of unlabeled actin in LS buffer. Coverslips were washed in LS buffer and incubated with 30 μl of labeled actin filaments (20 nM Rhodamine phalloidin actin in 50 mM DTT) for 35 s. The excess solution was removed and the flow cell was loaded for 1 min with 40 μl of MB with 0.7% methylcellulose, 1mM ATP, 50 mM KCl, 50 mM DTT, 2.5 mg/ml glucose, 2.5 μg/ml glucose oxidase, and 45 μg/ml catalase. The slides were imaged at 25 ºC, at 5 s intervals for 2 min, on an inverted Nikon Eclipse Ti-E microscope with an H-TIRF module attachment, a CFI60 Apochromat TIRF 100x Oil Immersion Objective Lens (N.A. 1.49, W.D. 0.12 mm, F.O.V 22 mm) and an EMCCD camera (Andor iXon Ultra 888 EMCCD, 1024 × 1024 array, pixel size: 13 μm). Velocity of the actin filaments on top of the different NM2A-GFP-Flag^HMM molecules was quantified using the FAST algorithm described in [68].

C. elegans strains

C. elegans strains were maintained at 20 ºC on nematode growth medium (NGM) plates previously seeded with OP50 E. coli. Strains harboring NMY-2 point mutations (Y163F and Y163E) were generated by CRISPR/Cas9 (S1 Table). Briefly, the gonads of young adult wild-type animals (N2 strain) were injected with three different single guide RNA sequences (sgRNAs) and a single-stranded repair template (ssODN; IDT ultramer) to edit the nmy-2 locus (details available in S2 Table). The sgRNAs were individually expressed from the pDD162 vector, which also contains the Cas-9 gene re-encoded for C. elegans [69]. The injection mix also contained a sgRNA and a single-stranded repair template carrying the R92C mutation on the dpy-10 gene, to allow the identification of successfully injected worms through the described roller phenotype [70,71]. To screen for animals carrying the targeted modifications, amplification of a nmy-2 fragment was performed using specific primers (Diagnosis primers, available in the supplementary material), followed by digestion with the HindIII or Eco47III restriction enzymes that recognize the restriction site present in modified animals only. The successful editing of the nmy-2 locus was confirmed by genomic DNA sequencing. nmy-2 mutant strains were outcrossed six times with N2 animals to eliminate possible off-target mutations.

Brood size and embryo viability in C. elegans

Synchronized L1 animals of N2 and GCP693 strains were grown in NGM plates seeded with OP-50 E. coli at 20 ºC for 72 h (1-day adults). Adult animals were singled out onto fresh NGM plates and the total number of eggs they laid over 24 h was counted for brood size quantification. For quantification of embryonic viability, the percentage of embryos that hatched was determined 24 h after embryos were laid.

Toxin soaking assay in C. elegans

Toxin exposure experiments were performed by incubating worms in liquid medium supplemented with defined concentrations of toxins and food [72]. For each condition, ~ 25 L4 stage animals of N2 and GCP693 strains were transferred into individual wells of a 96-well plate (in triplicate). Each well contained a total volume of 200 µl of S-medium supplemented with OP-50 E. coli (OD₆₀₀ ~ 2–3) and 50 µM 5-fluoro-20-deoxyuridine (FUdR) to inhibit progeny production. Toxins (LLO, PFO or PFO inactive) were prepared in 20 mM HEPES (pH 8) at final concentrations of 50, 100 and 250 µg/ml. Plates were incubated at 25 °C in a humid chamber for 5 days. Animals were scored for viability by touch-provoked movement.

Cry5Ba intoxication assay

Synchronized L1 animals of N2 and GCP693 strains were grown in OP-50 E. coli-seeded NGM plates at 20 ºC for 50 h (L4 stage) and then transferred to NGM plates previously seeded with control (empty vector pQE) or Cry5Ba-expressing BL21(DE3) E. coli [72]. Adult animals were transferred into freshly-seeded plates every 2 days. Moving and dead animals were counted every 24 h, the ones not responsive to touch were considered dead.

Briefly, saturated cultures of E. coli transformed with the pQE empty vector or pQE-Cry5Ba vector were diluted (10x) in LB with 100 μg/ml ampicillin and grown for 1 h at 37 ºC with shaking. Expression of Cry5Ba was induced with 1 mM IPTG for 3 h at 30 ºC. 100 µl of each culture at OD = 2 ± 0.1 were plated onto NGM plates containing 2 mM IPTG and 100 μg/ml ampicillin, and incubated at 25 ºC overnight.

Heat shock assay

Synchronized L1 animals of N2 and GCP693 strains were grown as described above for 50 h. L4 animals of each strain were transferred to fresh NGM plates and maintained at 20 ºC or incubated in a water bath at 35 ºC for 1 h. Plates were then kept at 20 ºC for 24 h. Surviving animals were singled out on fresh NGM plates and allowed to lay eggs for 24 h at 20 ºC. Brood size and embryo viability were assessed 24 h later.

Statistical analysis

Statistical analyses were carried out with Prism 8 (version 8.1.1, GraphPad Software, La Jolla California USA), using one‐way ANOVA with Dunnett’s post hoc analysis to compare different means in relation to a control sample and Tukey’s post hoc analysis for pairwise comparisons of more than two different means. Two-way ANOVA with Šídák’s post hoc analysis was used to compare each sample mean with another sample mean, in the same condition. Two‐tailed unpaired and paired Student’s t‐test was used for comparison of means between two samples.