A PKC that controls polyphosphate levels, pinocytosis and exocytosis, regulates stationary phase onset in Dictyostelium

Although many organisms can pause their cell cycle, entering either a stationary or a quiescent phase, very little is known about the mechanisms controlling stationary phase (SP) entry in protozoans. A stronger understanding of SP entry in model eukaryotes other than yeasts will reveal whether mechanisms of cell cycle pausing are evolutionarily conserved and provide insights into cell survival under stressful conditions (Brauer et al., 2008). It could also add to insights into cancer biology that have emerged from the pioneering cell cycle research done in yeast (Cazzanelli et al., 2018; Guaragnella et al., 2014; Matuo et al., 2012) and perhaps help in the management of protozoan infections such as malaria.

Different kinases are implicated in SP entry in both bacteria (Jiang et al., 2000; Gaidenko et al., 2002) and yeast (Reinders et al., 1998), but little is known about the role of protein kinase C (PKC). Both entry and survival in SP in Saccharomyces cerevisiae are dependent on the serine/threonine kinase Rim15p (Reinders et al., 1998). A cAMP-dependent PKA phosphorylates Rim15p and inhibits its kinase activity, and this inhibited state is necessary for SP entry. This is achieved by increased expression of Bcyl (a regulatory subunit of PKA), which suppresses PKA activity (Werner-Washburne et al., 1991). PKC belongs to a family of serine/threonine kinases and PKC isoforms are known to either promote or suppress cell proliferation depending on the context (Steinberg, 2008). For example, PKCδ inhibits proliferation of vascular smooth muscle cells, but promotes cancer stem cell proliferation in breast cancer cells (Fukumoto et al., 1997; Chen et al., 2014). In many cancer cells, the activation, or inactivation, of a PKC isoform is an important step in signaling pathways influencing cell division, migration, invasion, cell survival and quiescence (Fukumoto et al., 1997; Chen et al., 2014; Kang, 2014; Musashi et al., 2000). In yeast, Pkc1 is required for cell viability and integrity once SP has been entered (Heinisch et al., 1999; Krause and Gray, 2002), but whether it has a role in SP entry in yeast, or in any other microorganism, is unknown.

The model unicellular eukaryote Dictyostelium discoideum enters SP and stops dividing after reaching a density of ∼2×10⁷–2.5×10⁷ cells/ml in suspension cultures (Soll et al., 1976; Yarger et al., 1974). Prior to the onset of SP, the concentration of factors controlling proliferation, such as extracellular polyphosphates (polyP), autocrine proliferation repressor A (AprA) and counting factor associated proteins (CfaD) increase, thereby restricting further cell division (Suess and Gomer, 2016; Brock and Gomer, 2005; Bakthavatsalam et al., 2008). As a function of increasing cell density, polyP levels rise, inhibiting further proliferation (Suess and Gomer, 2016). The action of polyP is mediated by the G-protein-coupled receptor GrlD and the small GTPase RasC (Suess et al., 2017, 2019).

Here we show that a putative Dictyostelium PKC (pkcA; DDB_G0288147), which is also known to have a significant role in multicellular development (Mohamed et al., 2015; Singh et al., 2017) can control SP entry by regulating the rates of pinocytosis and exocytosis, down and up, respectively. This should result in a reduced access to nutrients, thus mimicking or amplifying, the starvation conditions that are the normal trigger for increased extracellular polyP, and SP onset.

To determine whether PkcA plays a role in entry into SP, the growth kinetics of wild-type Ax2 and pkcA⁻ cells were examined. Throughout most of the growth phase, the doubling time of pkcA⁻ was significantly higher than Ax2 cells (Fig. 1A). Prior to entering log phase (<1×10⁶ cells/ml), the doubling time of Ax2 was 12.5±0.4 h (mean±s.e.m.) compared to 19.0±0.7 h for pkcA⁻. During log phase growth (1×10⁶–10×10⁶ cells/ml), the doubling times were 9.8±0.5 h and 14.6±1.1 h, respectively. However, beyond a density of 10×10⁶ cells/ml, the doubling times of Ax2 and pkcA⁻ equalized at 18.9±2.0 h and 18.2±0.5 h, respectively. In addition to having a longer doubling time, pkcA⁻ cells reached SP at half the cell density of Ax2 cells; 12.6±0.4×10⁶ cells/ml compared to 25.1±1.2×10⁶ cells/ml (Fig. 1A,B; Table S1). Both of the growth defects of pkcA⁻ cells were rescued by overexpression of pkcA, suggesting that the defects are specifically due to loss of PkcA (Fig. 1A,B). Both defects were also mimicked in Ax2 cells by the addition of a PKC-specific inhibitor [GF109203X; bisindolylmaleimide I (Bis I) (Toullec et al., 1991)] (Fig. 1D,E; Fig. S1B, Table S1) strongly implying that the kinase activity of PkcA is indeed responsible for regulating SP onset.

Individually, either the regulatory or catalytic domain of PkcA failed to rescue the pkcA⁻ growth defects, implying that both domains are critical for cell density-dependent onset of SP (Fig. S1C; Table S1). The decreased density at which pkcA⁻ cells enter SP was observed in cultures with different inoculum densities and aeration rates, suggesting that this is an intrinsic defect (Fig. S1D,E; Table S1). Interestingly, the growth defects associated with loss of PkcA were not observed when the cells were grown on plates or with bacteria, suggesting that the defect is specific to culturing conditions (Fig. S1F,G).

As the precocious SP entry of pkcA⁻ cells was only apparent in suspension cultures, we examined whether the shear stress associated with growth in suspension culture played a role in the phenotype. By varying the rotation speed at which the cultures were grown, we found that at a lower speed (50 rpm) the difference in the SP cell density (at SP onset) between Ax2 and pkcA⁻ cells was not significant. However, as the rotation speed increased, the SP-onset densities of Ax2 and pkcA⁻ increasingly diverged (Fig. 1G,H; Table S1). This experiment suggests that PkcA is important for cell proliferation in high shear stress conditions.

Given that the loss of PkcA caused an increased doubling time and precocious SP entry, we examined whether PkcA expression and/or activity varied during cell proliferation in the wild type. Although there was no significant difference in pkcA expression in mid-log (ML) cells compared to SP cells (Fig. 1F), PKC activity in SP cells was only 46.1±3.9% (mean±s.e.m.) of that seen in ML cells (Fig. 1C), suggesting that PKC activity decreases when cells are in SP. Our observations are similar to the results reported in neuronal and skeletal muscle cells, which show a significant difference in PKC activity without any change in the expression of PKC isoforms (Rose et al., 2004; Chopra et al., 2018). Interestingly, the PKC activity of pkcA⁻ cells in ML is less than that for Ax2 cells (60.1±1.7% of that in ML Ax2 cells), which suggests the presence of other kinases with similar functions (Fig. 1C). However, the PKC activity drops to 41.6±5.8% in SP pkcA⁻ cells (Fig. 1C). Taken together, these results indicate that lowered PKC activity is associated with low proliferation rates and early SP onset, which is consistent with PkcA activity being involved in their regulation.

Given that (1) the growth defects of pkcA⁻ cells only occur under conditions of high shear stress, and (2) that SP onset and lower proliferation are associated with lowered PKC activity, we examined how the levels of PKC activity in ML cells respond to increased shear stress. Surprisingly, as the shaking speed was increased from 0 rpm to 150 rpm, the PKC activity in both Ax2 and pkcA⁻ cells increased dramatically; the increase was, however, much higher for the wild-type cells (Fig. 1I). Thus, the severity of the defects seen in pkcA⁻ cells do not appear to be related to low absolute levels of PKC activity, but rather correlate with the decreased ability to upregulate PKC activity in response to shear stress.

During growth, Dictyostelium cells secrete various factors into the medium that control their own proliferation (Bakthavatsalam et al., 2008; Brock and Gomer, 2005; Soll et al., 1976; Yarger et al., 1974; Suess and Gomer, 2016). Thus, the growth defects seen in the pkcA⁻ cells may be due in part to changes in the factors that have been secreted into the growth medium. In order to address this possibility, we grew Ax2 and pkcA⁻ cells until approximately day 5, when both cell types are in SP and examined the effects of replacing the initial growth medium with either: (1) fresh medium (HL5), which will not have any of the extracellular growth factors; or (2) the original growth medium [medium in which the cells were grown until they reached SP; also called conditioned medium (CM)] (Fig. 2A–C). As summarized in Fig. 2A,C (first six bars), both Ax2 and pkcA⁻ cells continued to grow after resuspension in the HL5 medium, and reached SP at a considerably higher density than did either (1) the original pre-replacement cultures, or (2) cultures where the cells were resuspended in their original conditioned media. In the latter case, the Ax2 and pkcA⁻ cells ceased growth. These results clearly support the proposal that secreted factors are responsible for the precocious onset of SP in pkcA⁻ cells.

Next, we repeated the above experiments, except that the cells were resuspended at a quarter of the density used previously and again, the dilution was done on day 5 (Fig. 2B,C, final four bars). In fresh HL5 medium, the cells of both Ax2 and the pkcA⁻ cells continued to grow until they reached the same density as they had before collection (Fig. 2C; compare bars 1 vs 7, and 2 vs 8). However, when the cells were resuspended in their original conditioned media, the cultures did not grow further (Fig. 2C, bars 7 and 8). These results indicate that it is not cell density per se that is responsible for the precocious SP onset, but factors in the growth medium (Table S1).

To further explore the role of the CM as the source of inhibitory factors, we also performed several CM replacement experiments where: (1) the CM from the opposite strain was used (e.g. the Ax2-CM was replaced with pkcA⁻-CM); (2) CM from both ML and SP cultures were tested on ML cells; and (3) finally, after replacement, the cultures were propagated under either high (shaking; Fig. S2A,B) or low (static; Fig. 2D–F) shear stress conditions. The low shear stress experiment was included because, in this case, any effect on growth should only be due to factors secreted into the CM when the culture was exposed, earlier, to high shear stress. When the post-replacement culturing occurred in high shear stress conditions, replacement of the Ax2-CM with the pkcA⁻-CM led to growth inhibition; conversely, replacement of the pkcA⁻-CM with the Ax2-CM led to growth (Fig. 2B). When post-replacement culturing occurred under low shear stress, the same patterns were seen (Fig. 2E,F). The degree of the respective inhibition or stimulation was also greater under high shear stress conditions. As expected, the cell densities observed after replacement with various ML-conditioned media led to higher cell densities than when SP-CM was used (Fig. 2E,F).

A gradual increase in polyP levels during Dictyostelium growth eventually leads to the cells entering SP (Suess and Gomer, 2016). Given that pkcA⁻ cells entered SP at a low cell density, we examined whether pkcA⁻ cells also have excess extracellular polyP. For this, the polyP levels in the CM of both Ax2 and pkcA⁻ cells grown in minimal medium (FM) were quantified, as HL5 medium would give background fluorescence (Aschar-Sobbi et al., 2008; Franke and Kessin, 1977). As seen in HL5 cultures, pkcA⁻ cells in FM medium also reached SP at a lower cell density than did Ax2 cells (Fig. S1H). The polyP levels of both the ML and SP-CM pkcA⁻ cultures were also 1.4 to 1.75 times higher than those of Ax2, while pkcA⁻/pkcA-OE (OE, is overexpression) strains had levels similar to the wild type (Fig. 3A,B). The high polyP levels seen in pkcA⁻ appear to be due in part to higher polyP synthesis, as intracellular levels of polyP are also higher than in Ax2 cells (Fig. 3C). Thus, pkcA⁻ cells synthesize and secrete more polyP than wild-type cells.

To determine whether excess polyP is responsible for the inhibitory effects of CM from pkcA⁻ cells, we used the recombinant S. cerevisiae exopolyphosphatase ScPPX1, which cleaves polyP (Gray et al., 2014). ScPPX1 treatment (0.15 μg/ml) for 4 h at 37°C (Suess and Gomer, 2016) indeed reduced the extracellular polyP levels in SP-pkcA⁻-CM [56.9±0.1% (mean±s.e.m.) relative to untreated SP-pkcA⁻-CM] (Fig. 3D). Ax2 cells were then incubated for 24 h with ScPPX1-treated and buffer-treated SP-pkcA⁻-CM, and cell density was measured. Consistent with earlier results, buffer-treated SP-pkcA⁻-CM restricted Ax2 growth to 33.7±3.2% of that in the control. Inhibition by ScPPX1-treated SP-pkcA⁻-CM was much less (63.4±0.8% of control level; Fig. 3E). Furthermore, daily addition of 0.15 μg/ml ScPPX1 increased the SP density of pkcA⁻ from 14.0±0.4×10⁶ cells/ml to 24.8±2.9×10⁶ cells/ml, and Ax2 cells from 19.6±2.3×10⁶ cells/ml to 37.5±2.0×10⁶ cells/ml (Fig. 3F; Fig. S1I, Table S1). Taken together, these results suggest that pkcA⁻ cells produce excess polyP, significantly contributing to their entry into SP at low cell density.

The accumulation of extracellular polyP is facilitated by two enzymes, polyphosphate kinase 1 (Ppk1), which catalyzes polyphosphate synthesis, and inositol hexakisphosphate kinase (I6kA), which is involved in synthesis of inositol pyrophosphates (IP7 and IP8) from IP6 (Suess and Gomer, 2016; Zhang et al., 2007; Pisani et al., 2014). The SP-CM from both ppk1⁻ and i6kA⁻ cells had a marginal inhibitory effect on pkcA⁻ proliferation compared to SP-pkcA⁻-CM. On the contrary, SP-pkcA⁻-CM suppressed proliferation in all cell types more than their respective conditioned media (Fig. S2C–F). This suggests that the reduced amount of extracellular polyP in ppk1⁻ and i6kA⁻ cells complements the excess of polyP in pkcA⁻ cells. To identify the particular enzyme through which PkcA controls polyP production, the effect of PKC inhibitor (5 μM Bis I) on the growth of ppk1⁻ and i6kA⁻ cells was tested. ppk1⁻ cells reached SP at a density of 12.6×10⁶±0.6×10⁶ cells/ml and 27.7×10⁶±4.2×10⁶ cells/ml (mean±s.e.m.) with and without Bis I, respectively. In contrast, Bis I treatment of i6kA⁻ cells had no significant impact, with cells entering SP at densities of 35.4×10⁶±2.0×10⁶ cells/ml and 31.9×10⁶±1.0×10⁶ cells/ml in the absence and presence of Bis I, respectively (Fig. 4A,B). Concomitantly, Bis I treatment increased the extracellular polyP levels in SP-Ax2-CM and SP-ppk1⁻-CM but had no effect on polyP levels in SP-i6kA⁻-CM (Fig. 4C). The lack of effect on i6kA⁻ cells in both cases argues that i6kA is acting downstream of Bis I inhibition. In other words, PkcA requires I6kA to regulate polyP production and cell density-dependent SP entry.

Previous studies report that PkcA works through the phospholipase D ortholog PldB during Dictyostelium development (Singh et al., 2017). To determine whether this relationship also exists during growth, we examined the effect of Bis I on the growth of pldB⁻ cells as well as their ability to produce extracellular polyP. Much like with i6kA⁻ cells, Bis I treatment had no significant impact on the growth of pldB⁻ cells, with SP densities of 13.5×10⁶±1.2×10⁶ cells/ml and 13.5×10⁶±0.5×10⁶ cells/ml in the absence and presence, of Bis I, respectively (Fig. 4A,B). Consistent with this, Bis I treatment had no impact on the extracellular polyP levels in pldB⁻ cells (Fig. 4C). Also the inhibitory effect of SP-pldB⁻-CM on each of the cell types (Ax2, pkcA⁻ and pldB⁻) was stronger than SP-pkcA⁻-CM (Fig. S2G,H). Taken together, these results suggest that PkcA acts via PldB to regulate polyP production and cell density-dependent SP entry.

During Dictyostelium growth, secreted factors such as AprA and CfaD slow down cell proliferation once they have reached a threshold concentration (Bakthavatsalam et al., 2008; Brock and Gomer, 2005). Thus, the levels of these factors were determined in pkcA⁻-CM. Although AprA levels in the ML-pkcA⁻-CM were significantly higher (a 2.1±0.5-fold increase) (Fig. S3A,B), there were no differences in the intracellular levels of either protein in Ax2 and pkcA⁻ cells (Fig. S3C,D). This suggests that levels of AprA, but not CfaD, correlate with the high levels of polyP in pkcA⁻ cells during growth.

Next, to ascertain if the inhibitory activity of SP-pkcA⁻-CM is due to excess AprA, the CM was treated with anti-AprA antibody (1:150) and thereafter, cell proliferation was examined in Ax2 and pkcA⁻ cells. Surprisingly, there was no difference in the proliferation rates in both cell types (Fig. S3E). Taken together, our results indicate that high AprA levels in ML-pkcA⁻-CM are not responsible for the defective growth of the mutant.

Dictyostelium cells deprived of nutrients are known to have excess extracellular polyP (Suess and Gomer, 2016). If pkcA⁻ cells are defective in taking up the available nutrients, then providing them with more nutrients may reverse the phenotype. To address this possibility, we grew Ax2 and pkcA⁻ cells in concentrated medium (2× HL5) and measured their cell density and polyP levels at SP. Growth in 2× HL5 had a small but significant effect on Ax2 cells. In contrast, pkcA⁻ cells, reached SP at drastically higher cell density in 2× HL5 medium (Fig. 5A; Fig. S1J, Table S1). There was also a significant reduction (31.2±7.9%; mean±s.e.m.) in extracellular polyP levels (Fig. 5B). These results suggest that the pkcA⁻ phenotype may be due to a defect in nutrient uptake or sensing.

Nutrient assimilation and cell growth significantly depend on effective pinocytosis and exocytosis. To determine whether the pkcA⁻ phenotype may be due to impaired pinocytosis or exocytosis, cells grown in shaking HL5 cultures were incubated with the FITC–dextran (2 mg/ml, 70,000 M_r), which was collected at the indicated time points, and the uptake of FITC–dextran was measured using a spectrofluorimeter (492 nm ex/525 nm em). ML pkcA⁻ cells had a reduced pinocytosis rate, resulting in only 56.0±3.0% (mean±s.e.m.) of the FITC-dextran uptake seen in Ax2, at 120 min (Fig. 5C). Interestingly, pkcA⁻ cells had higher exocytosis rates, releasing 50% of the FITC–dextran in 30 min, whereas Ax2 cells took nearly 60 min to exocytose the same percentage of FITC–dextran (Fig. 5D).

The increased exocytosis in pkcA⁻ cells could be due to faster postlysosomal neutralization and release of fluids from endo-lysosomal compartments (Neuhaus et al., 2002; Seastone et al., 2001). To test this, the vesicular pH of fluid phase-harboring compartments was measured (Seastone et al., 2001). In the first 10 min, fluid entry was into the acidic compartments (pH<5) in both cell types. In the subsequent 30 min, FITC–dextran entered the neutral compartments (pH>5.5) of Ax2 cells, but in pkcA⁻ cells, dextran showed a delayed entry into the neutral postlysosomes (PLs) and was detected from 50 min. From this time, the vesicular pH in pkcA⁻ cells (5.5±0.4) was not significantly lower than that of vesicles in Ax2 cells (5.7±0.1) (Fig. 5E). This indicates that rapid postlysosomal neutralization in these cells does not occur, and thus cannot be the reason for the increased exocytosis in pkcA⁻ cells.

To determine whether low pinocytosis is associated with early SP onset, we measured the cell density at SP of several mutants impaired in pinocytosis or exocytosis. At the same time, the extracellular polyP levels were also quantified. Like pkcA⁻, mutants with reduced pinocytosis rates such as the double-mutant abpA⁻/C⁻ (Rivero et al., 1999), lvsB⁻ (Charette and Cosson, 2007), the double-mutant proA⁻/B⁻ (Temesvari et al., 2000) and scarA⁻ (Seastone et al., 2001), also entered the SP at a low cell density (Fig. S4A, Table S2). Similar to pkcA⁻, abpA⁻/C⁻, lvsB⁻ and scarA⁻ also have high extracellular polyP levels (Fig. S4B, Table S2). Only proA⁻/B⁻ cells had low polyP levels. Taken together, the results point to a significant correlation between defective pinocytosis, high levels of extracellular polyP and entry into SP at low cell density.

The actin cytoskeleton performs important roles during pinocytosis and exocytosis (Seastone et al., 2001; Veltman et al., 2016; Wight et al., 2020). As pkcA⁻ cells have defective pinocytosis and exocytosis, we examined whether F-actin levels or actin assembly were also affected in the mutant. Confocal imaging and SDS-PAGE analysis of the cytoskeletal protein fraction showed higher F-actin levels in pkcA⁻ than in Ax2 cells (1.8±0.2 fold higher) (Fig. 6A,B). Interestingly, in non-shaking conditions, the F-actin levels in Ax2 and pkcA⁻ cells were not significantly different (Fig. 6C) suggesting that PkcA regulates F-actin polymerization differentially in different conditions. The impaired F-actin polymerization in pkcA⁻ cells was observed only when cells were grown in shaking conditions, and such cells were impaired in cytokinesis, were larger and more of them were multinuclear (Fig. S5A–D) (Singh et al., 2017). However, neither myosin II assembly nor its functions were impaired in pkcA⁻ cells (Fig. S6A–D). As polyphosphates are known to affect actin cytoskeleton assembly and polymerization in D. discoideum (Suess et al., 2017), and as pkcA⁻ cells have excess polyP, we then examined whether reducing the F-actin levels in pkcA⁻ cells restores extracellular polyP levels and normal growth. This was done by overexpressing coronin A (corA), an actin-binding protein involved in F-actin depolymerization (Fig. 6D) (Shina et al., 2011). Overexpression of corA in pkcA⁻ cells indeed reduced extracellular polyP levels and restored SP entry, such that it was similar to that in the wild type (Fig. 6E,F; Fig. S1K, Table S1). These results suggest that PkcA, by maintaining actin polymerization, regulates pinocytosis, exocytosis and extracellular polyP levels, thus controlling SP entry.

Seeking further insights into the cause of the increased exocytosis in the pkcA⁻, we examined subcellular localization of PkcA by imaging vegetative cells expressing a PkcA–GFP fusion protein (Ax2/act15::pkcA-GFP). In ML cells, PkcA–GFP was confined to vesicles and the plasma membrane (Fig. 7A; Movie 1), but in SP cells, PkcA–GFP, while largely showing the same distribution, was also partially diffused in the cytoplasm (Fig. 7A). The C1 domain (pkcA-C1–RFP) was only localized to the vesicles; membrane localization was not observed. The catalytic kinase domain (pkcA-cat–GFP) was enriched in the vesicles and also had a diffuse staining in the cytosol (Fig. S7A). The staining patterns suggest that the regulatory C1 domain is important for restricting the vesicle specific localization of PkcA and the full-length protein is important for plasma membrane localization. Confocal imaging also suggests some PkcA–GFP is secreted out of the cells (Movie 2). Extracellular PKC, or any other kinase, would require also the secretion of ATP for its activity; Dictyostelium is known to release micromolar levels of ATP into the extracellular environment (Parish and Weibel, 1980; Sivaramakrishnan and Fountain, 2015; Consalvo et al., 2019). Furthermore, we observed PKC localization in lysosomes, the contents of which are known to be exocytosed in other systems (Buratta et al., 2020; Tancini et al., 2020). The CM, which will have all the secreted proteome, shows PKC activity both in ML and SP cells of wild-type and pkcA⁻ cells (Fig. S7B). PKC activity did not affect the localization of PkcA–GFP (Movies 3 and 4). However, F-actin polymerization is important for the movement of vesicles harboring PkcA–GFP (Movies 3 and 5).

To identify the organelles harboring PkcA–GFP, we used MitoTracker RED, which stains mitochondria, and either Acridine Orange (AO) or LysoTracker RED for staining acidic lysosomes. PkcA–GFP did not colocalize with MitoTracker Red (Fig. S7C) but did colocalize with AO- and LysoTracker RED-containing vesicles (Fig. 7B,C). This suggests that PkcA–GFP localization is confined onto lysosomes. Furthermore, PkcA–GFP vesicles colocalized with TRITC–dextran-containing late endosomes (Fig. S7D).

Given the localization of PkcA–GFP on lysosomes and the fact that pkcA⁻ cells show decreased pinocytosis and increased exocytosis, we examined whether the fluid transfer along the endo-lysosomal pathway is defective in pkcA⁻ cells. Ax2 and pkcA⁻ cells were independently incubated with FITC–dextran and AO for 10 min and observed at regular intervals for 1 h using a confocal microscope. Large intense FITC-fluorescent vesicles are known to be less acidic (generally neutral PLs of ∼2 µm diameter) than the smaller vesicles with reduced FITC fluorescence (Seastone et al., 2001). By 60 min, 67.9±1.2% (mean±s.e.m.) of Ax2 cells showed the presence of large FITC–dextran-labeled vesicles with reduced AO fluorescence (less acidic or neutral vesicles) and 35.5±7.7% of pkcA⁻ cells showed vesicles with reduced AO fluorescence (Fig. 7E,F). In the remainder of pkcA⁻ cells, intense AO fluorescence was confined to acidic vesicles. Although the average number of PLs per cell is not significantly different in pkcA⁻ (1.8±0.1) compared to in Ax2 cells (1.9±0.3, P=0.322) (Fig. 7G), pkcA⁻ cells had smaller PLs with an average diameter of 1.2±0.02 µm, whereas, in Ax2 cells, the average diameter of PLs was 1.5±0.03 µm (Fig. 7H). These observations, together with the faster exocytosis rates, suggest the PLs are prematurely exocytosed from pkcA⁻ cells and thus PkcA seems to be involved in regulating the maturation of PLs, and exocytosis. Taken together, the results suggest that PkcA regulates nutrient accumulation and thus cell growth by ensuring that vesicles are retained for an optimal time, such that nutrients can be properly assimilated.

PkcA–GFP is localized in the lysosomes and, given that one of the lysosome-dependent processes, exocytosis, is affected in pkcA⁻ cells, we wanted to ascertain whether any other lysosome-involving processes are affected. One such process is the lysosomal degradation of large aggregates of misfolded proteins transported to lysosomes by the autophagy machinery (Jackson and Hewitt, 2016). PKC isoforms localized in the lysosomes have previously been documented to play an important role in clearing protein aggregates (Li et al., 2016). In Dictyostelium, which has a Q/N-rich proteome, protein aggregation, while limited, does occur (Malinovska et al., 2015; Santarriaga et al., 2015). To examine whether the clearance of protein aggregates is PkcA dependent, we transformed Ax2 and pkcA⁻ cells with a Q103–GFP construct (polyQ-containing exon 1 of human huntingtin with 103 consecutive glutamine residues), to measure protein aggregate formation. Q103–GFP cells form protein aggregates and GFP spots representing polyQ foci that can be identified microscopically. Confocal imaging and subsequent counting of GFP spots suggest that pkcA⁻ cells, at both ML and SP, have higher protein aggregation than the corresponding Ax2 cells (Fig. S7E,F). The fraction of ML-pkcA⁻ and SP-pkcA⁻ cells with polyQ foci were respectively, 4.3±0.8% and 45.8±1.7% (mean±s.e.m.) compared to the Ax2 cells being 1.9±0.2% in ML and 21.5±2.4% in SP. These two observations, the lysosomal localization of PkcA and the fact that pkcA⁻ cells had a larger fraction of cells with protein aggregates, are both consistent with the idea that impaired protein aggregate clearance in the pkcA⁻ cells is due to defective lysosomal function. However, protein aggregate clearance is also mediated by autophagy (Jackson and Hewitt, 2016), so the relative contribution of these two processes needs to be examined.

Upon nutrient deprivation, many microorganisms cease cell division and enter SP (Herman, 2002). No previous work has examined the possible role of PKC in the onset of SP in Dictyostelium, or any of the other model microorganisms. In this study, we demonstrate that pkcA (a Dictyostelium PKC analog) is important for SP entry in conditions of high shear stress. Our results indicate that, in the shift between ML and SP, PKC activity is reduced, and SP onset is accompanied by higher levels of polyP within the cell, and also, perhaps as a direct and/or indirect consequence, in the medium. As previously shown, a high extracellular level of polyP is one of the crucial signals inducing the coordinated entry of Dictyostelium cells into SP (Suess and Gomer, 2016; Suess et al., 2017).

Multiple lines of our evidence support, or are consistent with, the idea that a lowered level of PKC activity is an important intermediate step in the transition from ML to SP. The most important observations were the early onset of SP in the pkcA⁻ cells and the pattern of PKC activity in the wild type during the ML-SP transition. The involvement of PKC and polyP was also demonstrated by the high levels of polyP in the CM of the early-SP-onset (pkcA⁻) line and by the ability of its CM to induce early onset when used to replace the medium of wild-type cells. Although a role for PKC in regulating a microbial cell cycle is novel, PKC is well known to be involved in cell proliferation and cell cycle arrest in many mammalian tissues (Nanos-Webb et al., 2016; Graham et al., 2000; Fukumoto et al., 1997; Chen et al., 2014). In yeast, several Ras proteins, small GTP-binding proteins that activate protein kinase A (PKA) and TOR signaling, are involved in regulating SP entry (Herman, 2002; Werner-Washburne et al., 1991). The onset of SP also involves many kinases, such as Rim15p protein kinase, S6 kinase and PKA (Herman, 2002; Werner-Washburne et al., 1991; Reinders et al., 1998). Both pkcA⁻ and Ax2/pkcA-OE cells show severe growth defects and enter SP at a lower cell density. However, Ax2/pkcA-OE cells survive for a longer time in SP (5–6 days compared to 2–3 days for Ax2) suggesting that pkcA is important for both SP onset and survival in SP. This is similar to the case in yeast cells, where Pkc1 is required for viability and cellular integrity during nutrient starvation (Heinisch et al., 1999; Krause and Gray, 2002). Thus, either an increase or decrease in pkcA levels affects growth. PkcA may be regulating different targets at different stages of cell proliferation. For example, both up- and down-regulation of SR-protein specific kinase 1 (SPRK1) promotes cell proliferation in mammalian cells, such as mouse embryonic fibroblasts (Wang et al., 2014).

Reduced levels of pkcA are associated with increased levels of polyP, both intracellularly, and even more importantly, extracellularly. In our model, PkcA drives this relationship and is responsible for regulating polyP levels (Fig. 8A,B). Although it is possible that the reverse is true, and that increased extracellular polyP results in lowered pkcA, previous proteomics work argues against it (Suess et al., 2017). It should also be noted that polyP at different concentrations (125 μM and 150 μM) suppresses cell proliferation of pkcA⁻ and wild-type cells equally; thus, the sensitivity of pkcA⁻ cells to polyP is similar to that of the wild type, suggesting that pkcA is not required for polyP-mediated proliferation inhibition (Tang et al., 2021). Also, of the two genes i6kA and ppk1, which are involved in regulating extracellular polyP levels (Suess and Gomer, 2016; Zhang et al., 2007), we show that only i6kA and pkcA are epistatically related, suggesting that they share an enzymatic pathway. We argue that prior to SP, at a low cell density, PkcA inhibits I6kA-dependent polyP production, whereas at high densities, reduced PkcA activity leads to increased polyP production via I6kA. This is the first report suggesting that a PKC regulates I6kA. PkcA⁻ cells had reduced expression of i6kA (Fig. S8A). Possibly, I6kA protein activity is either not compromised or the reduced expression is in fact a negative feedback response to increased I6kA activity. Alternatively, even reduced levels of i6kA may be sufficient to support the function of the protein. Although the mechanism by which PkcA regulates I6kA is yet to be ascertained, there are two possible PkcA phosphorylation sites (Ser9 and Ser572) in I6kA (Netphos 3.1) (Fig. S8B). Additionally, our work indicates that PldB also acts downstream of PkcA to regulate extracellular polyP levels and proliferation; PldB has previously been shown to interact with PkcA during the development stage of Dictyostelium (Singh et al., 2017). Neuronal PLD3 localizes in lysosomes and regulates the structure and function of lysosomes (Fazzari et al., 2017).

Our work suggests that an important step in the process by which reduced PkcA leads to higher polyP is a rise in actin polymerization. We documented higher levels of actin in the pkcA⁻ line, and showed that when the actin-depolymerizing protein CorA was overexpressed in this line, the extracellular polyP levels of the mutant were lower, and SP onset reverted to that of the wild type. For the following step in our model, that is, disrupted pinocytosis and exocytosis leading to nutrient depletion, we do not have direct evidence that the observed rate changes in the pkcA⁻ line were due to the observed increase in actin levels. However, previous studies on Dictyostelium using actin-regulating-protein mutants (such as scar⁻) and drug treatments, such as cytochalasin treatment, have demonstrated the importance of F-actin levels in pinocytosis and exocytosis (Seastone et al., 2001; Wight et al., 2020; Veltman et al., 2016). Similarly, PKC inactivation in neuronal cells is known to induce F-actin polymerization (Yang et al., 2013). In Dictyostelium, an increased level of actin does not, however, appear to be an obligatory step in SP onset. Although 2× HL5 rescued the cell density-dependent SP defect of the pkcA⁻ line, the F-actin levels remained high. This implies that nutrient levels can control polyP levels independently of PkcA-regulated actin polymerization. It is also important to note that adding polyP to the growth medium has been reported to lead to lowered actin levels in Dictyostelium (Suess et al., 2017).

Regardless of whether actin is directly involved, the rates of pinocytosis and exocytosis in the pkcA⁻ were lower and higher, respectively, than in the wild type, and each of these changes is a potential driver of early SP onset. Reduced pinocytosis and increased exocytosis are likely to result in, respectively, less access to, or retention of, nutrients, thus accentuating the starvation conditions associated with rising cell density. The involvement of reduced pinocytosis in polyP-driven SP onset was also supported by the early onset seen in four pinocytosis-defective mutants, three of which also had high extracellular polyP. Furthermore, the rescue of the pkcA⁻ growth defect in concentrated medium (2× HL5) supports the hypothesis that reduced nutrient retention is one of the main features of the pkcA⁻ growth defect.

Exocytosis-defective mutants showed drastic variation in extracellular polyP levels, suggesting exocytosis might not be directly involved in regulating the polyP levels. Rather, in the case of pkcA⁻ cells, increased exocytosis could have resulted in decreased retention time of nutrients, leading to starvation conditions.

PkcA–GFP localization in LysoTracker RED-positive vesicles and at the plasma membrane in Dictyostelium does suggest a direct involvement of PkcA in vesicle trafficking, as has also been reported in HeLa cell lines (Li et al., 2016). The C1 domain of PkcA is important for confined vesicular localization. Also, we show that both regulatory and catalytic domains are required to rescue the pkcA⁻ growth defect, consistent with previous observations (Steinberg, 2008). Our study also shows that only 35% of pkcA⁻ cells had neutral PLs of reduced size, indicating that there is a role for PkcA in regulating their maturation.

PL maturation might well be the stage affected by increased levels of actin. In Dictyostelium, the actin coat on late endosomes is crucial for PL neutralization and lysosomal fusion leading to exocytosis (Rivero, 2008). Thus, increased F-actin levels in pkcA⁻ cells might be preventing the fusion events necessary for PL maturation, with PLs thus remaining small. Postlysosomal fusion events are necessary for lysosome maturation and, in pkcA⁻ cells, PLs are exocytosed earlier, impairing the fusion events, as observed in cpnA⁻ cells (Wight et al., 2020; De Araujo et al., 2020). In pancreatic β-cells, PKC enhances the exocytosis of insulin by causing the rearrangement of cortical actin, and phosphorylation of exocytotic proteins (Trexler and Taraska, 2017). In platelets, partial inhibition of PKC restores dense granules secretion (Unsworth et al., 2011). While our model assumes that the main consequence of increased exocytosis in the pkcA⁻ line is to accentuate the starvation conditions, it is also possible that the vesicles carry out excess polyP into the medium (Fig. 8A,B).

One of our most intriguing results was that pkcA was only important when cultures were grown under shaking conditions, in which Dictyostelium cells experience considerable shear stress (Taira and Yumura, 2017). Although it is unclear what relevance this might have in the natural history of this organism, our results should be of interest to cancer researchers. If a tumor is to become truly metastatic, any of the cells that will enter the blood or lymph must complete the epithelial-to-mesenchymal transition (EMT), part of which involves the cell switching to a non-proliferative mode. This is more appropriate during the migration phase, and will itself be reversed (in the mesenchymal-to-epithelial transition) if the migratory cell ever finds a suitable site to restart proliferation. The shear stress experienced during migration is critical for completion of the EMT (Liu et al., 2016; Xin et al., 2020). Also, previous studies in endothelial cells have shown that shear stress caused by blood flow induces PKC activation, and the association of PKC with F-actin (Hu and Chien, 1997). The Dictyostelium genome also has many genes related to human cancer genes, which will be helpful in further studies of any interaction between them and pkcA (Kuspa et al., 2001; Eichinger et al., 2005). PKC-targeting drugs that have been considered for treating several types of cancer (Kim et al., 2008; Kreisl et al., 2010) may be more potent in high shear stress conditions, a point that deserves further study.

The D. discoideum strains used in this study include: wild-type Ax2 (DBS0235534), NC4A2 (DBS034992), HPS400 (DBS023631), pkcA⁻ (Mohamed et al., 2015), i6kA⁻ (DBS0236426) (Luo et al., 2003), ppk1⁻ (DBS0350686) (Zhang et al., 2007), pldB⁻ (DBS0236796) (Chen et al., 2005), abpA⁻/C⁻ (DBS0235456) (Rivero et al., 1999), lvsB⁻ (DBS0236521) (Harris et al., 2002), proA⁻/proB⁻ (DBS0236827) (Temesvari et al., 2000) and scar⁻ (DBS0236924) (Seastone et al., 2001). The proliferation kinetics of all the parental Ax2 strains, DBS0235534 (pkcA⁻, abpA⁻/C⁻ and proA⁻/B⁻), DBS0238015 (ppk1⁻), DBS0235522 (pldB⁻) and DBS0350762 (i6kA⁻) used to generate the mutants were analyzed and found to be highly similar (Fig. S1A). The cells were grown axenically in shaking cultures or Petri dishes in maltose-HL5 medium (28.4 g bacteriological peptone, 15 g yeast extract, 18 g maltose monohydrate, 0.641 g Na₂HPO₄ and 0.49 g KH₂PO₄ per liter, pH 6.4) or minimal medium FM (Formedium, Norfolk, UK). Alternatively, Dictyostelium cells were also grown in SM/5 agar plates (2 g glucose, 2 g protease peptone, 0.4 g yeast extract, 1 g MgSO₄·7H₂O, 0.66 g K₂HPO₄, 2.225 g KH₂PO₄ and 1.5 g bacto agar per litre, pH 6.4) in association with K. aerogenes at 22°C. Scar⁻ and HPS400 cells were grown in HL5 supplemented with 100 μg/ml thymidine. The mutant strains carrying the antibiotic resistant markers were grown with either blasticidin (10 μg/ml) or G418 (20 μg/ml) (MP Biomedicals, USA). The fine chemicals, antibiotics and salts were obtained from Sigma-Aldrich, USA, MP Biomedicals, USA or Thermo Fisher Scientific, USA.

To generate the PkcA–GFP construct, the full-length pkcA gene (4044 bp) was PCR amplified from the genomic DNA and cloned into the pDM317 vector (Veltman et al., 2009) with GFP at the N-terminus. This gene cassette was cloned between BglII sites such that it was driven by a constitutive actin15 promoter. A 918 bp fragment encoding the catalytic domain of pkcA was PCR amplified and cloned in the pTX-GFP (Dicty Stock Center, USA; N-terminal GFP) vector by exploiting BamHI and XhoI restriction sites. For cloning the C1 domain (pkcA-C1-RFP), a 162 bp coding sequence was PCR amplified and engineered in the pTX-RFPmars (N-terminal RFP) vector (Fischer et al., 2004), between BamHI and XhoI restriction sites. For all PCR reactions, Expand PCR High Fidelity System (Roche, Indianapolis, USA) was used as per the manufacturer's protocols. Ax2 and pkcA⁻ cells were independently transformed with pkcA–GFP, pkcA-cat–GFP and pkcA-C1–RFP constructs and the positive clones were selected with G418 (20 μg/ml).

Plasmid vectors were transformed into the cells by electroporation using a BTX ECM830 electroporator (Harvard Apparatus, USA). The cell suspension in HL5 medium was incubated at 22°C for 24 h, after which the cells were supplemented with 20 μg/ml G418 (MP Biomedicals, India) and were screened for G418-resistant clones. Subsequent to transformation, the clones were grown in HL5 medium supplemented with 10 μg/ml blasticidin (for selecting pkcA⁻ cells) and 20 μg/ml G418 (all other strains).

Dictyostelium cells overexpressing PkcA (pkcA-OE) (Mohamed et al., 2015), coroninA (corA-OE) (Shina et al., 2011) and polyQ-containing exon 1 of human huntingtin with 103 consecutive glutamine residues (Q103–GFP; for the protein aggregation assay) (Malinovska et al., 2015) were grown in HL5 supplemented with 20 μg/ml G418. Ax2/corA-OE, pkcA⁻/corA-OE, Ax2/Q103–GFP and pkcA⁻/Q103–GFP cells were generated by transforming Ax2 and pkcA⁻ cells by electroporating the corA-OE construct (Shina et al., 2011) (a kind gift from Dr Annette Müller-Taubenberger, LMU, Germany) and the Q103–GFP construct (Malinovska et al., 2015) (a kind gift from Dr Simon Alberti, MPI-CBG, Germany).

ML cells were resuspended in HL5 medium at a density of 1×10⁵ cells/ml and kept in a shaker again at 150 rpm. For the proliferation assay in FM medium, cells were resuspended at 1×10⁶ cells/ml. Every day, the cell density was determined using a hemocytometer. For the proliferation assay in Petri dishes, the ML cells were diluted to 1×10⁴ cells/ml in Petri dishes containing 10 ml HL5. For the proliferation assay on a bacterial lawn, approximately 1000 Dictyostelium cells were mixed with 200 μl Klebsiella aerogenes and spread on three 90 mm Petri dishes in SM/5 medium. At the indicated time points, one of the plates was washed and the cells were counted using a hemocytometer. For all the conditions, the cells were incubated at 22°C.

Growth assays after drug treatments were performed as described in the figure legends. ML cells grown in HL5 suspension cultures were tested with 1, 3, 5, 10, 20 and 40 μM bisindolylmaleimide GF109203X (Bis I) (Mohamed et al., 2015; Toullec et al., 1991) (Sigma-Aldrich, USA) and the minimum effective inhibitory dose was found to be 5 μM.

The PKC activity was measured from ML and SP cell lysates as per the manufacturer's protocol (V5330, Promega, USA). Dictyostelium cells harvested from HL5 medium were adjusted to a density of 1×10⁸ cells/ml in KK2 buffer, lysed in 500 μl–1 ml of PKC extraction buffer [25 mM Tris-HCl pH 7.4, 0.5 mM EDTA, 0.5 mM EGTA, 0.05% Triton X-100, 10 mM β-mercaptoethanol, 1× protease inhibitor cocktail (Sigma-Aldrich, USA) and 1× phosphatase inhibitor (Sigma-Aldrich, USA)]. A total of 9 μl of cell lysate was mixed with the PKC reaction mix containing PepTag C1 peptide and incubated at 30°C. After 30 min, the reaction was stopped by heating the samples at 95°C for 10 min. Then, 1 μl of 80% glycerol was added, the samples were run on a 0.8% agarose gel and were imaged. The kit uses a colored, fluorescent peptide substrate that is specific for PKC. Phosphorylation by PKC (cell lysates) alters the net charge of the peptide substrate, which allows the phosphorylated and un-phosphorylated versions of peptide substrate to rapidly separate on a agarose gel. The fluorescence intensity of phosphorylated peptide (a measure of PKC activity), was measured from the cell lysates using ImageJ (NIH, USA). The phosphorylated bands were analyzed and the intensity of each band was normalized to the total protein content. Total protein levels were assayed using Bradford reagent.

To determine the PKC activity in CM, at the indicated time points, the cell suspension was centrifuged (500 g for 7 min), the supernatant was concentrated 10-fold using 10 kDa cutoff filters (Amicon Ultracel-10K; Sigma-Aldrich, USA). 9 μl of concentrated CM was mixed with the PKC reaction mix for the PKC assay, which was performed as mentioned above.

Ax2 and pkcA⁻ cells at ML and SP were harvested and RNA was extracted using FavorPrep™ Tri-RNA Reagent (Favorgen, Taiwan) (Pilcher et al., 2007). 1 μg of total RNA was used to synthesize cDNA using a cDNA synthesis kit as per manufacturer's protocol (Verso, Thermo Scientific, USA). Quantitative real-time PCR (qRT-PCR) was carried out with 1 μl of cDNA using DyNAmo Flash SYBR Green qPCR kit (Thermo Scientific, USA) to analyze the expression levels of pkcA (FW, 5′-TAATATGTCTCGGTCCACCA-3′ and REV, 5′-ATCAACTCTCATCACATCGAC) and i6kA (FW, 5′-AAGCAATAGTGGTAACTTTAGCGG-3′ and REV, 5′-CAAATAGCCATCATCTTCTTGAGC-3′) using the Applied Biosystems^® QuantStudio Flex 7 (Thermo Fisher Scientific, USA). Ig7 (FW, 5′-TCCAAGAGGAAGAGGAGAACTGC-3′ and REV, 5′-TGGGGAGGTCGTTACACCATTC-3′) was used as an mRNA amplification control. Real-time data was analyzed as described previoulsy (Schmittgen and Livak, 2008).

The proliferation inhibition by CM was assayed by incubating the cells in 50% CM and 50% HL5 for 24 h at 22°C in non-shaking/shaking conditions (Suess and Gomer, 2016). The CM was harvested as mentioned in the figure legends. Cells grown in HL5 were considered as the control.

CM preparation and polyP quantification were carried out as described previously (Suess and Gomer, 2016) with slight modifications. For fluorimetric measurements of extracellular polyP, cells were grown in FM medium to reduce the background fluorescence of HL5 medium (Franke and Kessin, 1977; Suess and Gomer, 2016). The CM was clarified by passing it through a 0.22 μm PVDF syringe filter, and the polyP levels were measured by staining the samples with 25 μg/ml DAPI for 5 min. Fluorescence intensity was measured at 415 nm excitation and 550 nm emission (Aschar-Sobbi et al., 2008) using a spectrofluorometer (Molecular Devices, USA). The polyP concentrations were obtained by using standards (kindly provided by Dr Toshikazu Shiba, RegeneTiss Inc., Japan). Intracellular polyP was extracted following the previously described protocol (Suess and Gomer, 2016) and was analyzed on a 25–30% polyacrylamide gel.

PolyP levels were analyzed on polyacrylamide gels as previously described (Pisani et al., 2014) with the following modifications. ML- and SP-CM were purified through Midiprep columns (Qiagen Midiprep kit, Germany) (Desfougères et al., 2016). The polyP was eluted with the buffer provided by the manufacturer and polyP standards were run for reference. The gel images were quantified using ImageJ (NIH, USA).

The plasmid vector encoding the yeast recombinant exopolyphosphatase protein ScPPX1 was provided by Dr Michael Gray (University of Michigan, USA) (Gray et al., 2014). The CM was incubated with recombinant ScPPX1 at a concentration of 0.15 μg/ml along with 5 mM MgCl₂. Using 0.22 μm filters, the CM was filter sterilized and used directly as mentioned above. For the proliferation assay with ScPPX1, ML cells were inoculated at 1×10⁵ cells/ml and 0.15 μg/ml ScPPX1 or buffer was added daily to cell culture during growth.

To examine the extracellular levels of AprA and CfaD, equal volumes of ML- and SP-CM were boiled with 6× SDS gel loading buffer at 100°C for 5 min. The protein samples were separated in a 12% SDS-polyacrylamide gel (Laemmli, 1970) and a parallel gel was run for silver staining. For measuring intracellular AprA and CfaD levels, cells were resuspended in lysis buffer [62.5 mM Tris-HCl pH 6.8, 2% SDS, 0.1 M DTT and 10% (v/v) glycerol] and boiled for 5 min. An equal concentration of protein was loaded on a 12% SDS-polyacrylamide gel and another gel was run for Coomassie staining. After electrophoresis, the proteins were transferred to a nitrocellulose membrane (BioRad, Laboratories, USA) and immunoblotted with anti-AprA (1:3000; Brock and Gomer, 2005) or anti-CfaD (1:1500; Bakthavatsalam et al., 2008) primary antibodies (a kind gift from Dr Richard Gomer, Texas A&M University, USA). After incubation with HRP-conjugated secondary antibody (1:5000; Bangalore Genei, India), the membrane was then treated with Clarity Western ECL substrate (BioRad Laboratories, USA) and the chemiluminescent signals were captured using ChemiDoc imaging systems (BioRad Laboratories, USA). The images were quantified using ImageJ (NIH, USA). Full images of gels used to quantify F-actin and myosin levels are given in Fig. S9.

To deplete extracellular AprA, the CM was treated with the anti-AprA antibody in a ratio of 1:150 for 2 h in 22°C. Subsequently, the proliferation inhibition assay was performed using the treated CM.

Fluid-phase endocytosis and exocytosis assays were carried out as described previously (Rivero and Maniak, 2006) with slight modifications. Log phase cells were harvested and resuspended at a concentration of 5×10⁶ cells/ml in fresh HL5 medium and incubated for 15 min at 22°C at 150 rpm and then FITC–dextran (70,000 M_r; Sigma-Aldrich, USA) was added to a final concentration of 2 mg/ml. In the first 2 h, endocytosis of FITC–dextran was monitored by collecting the samples at indicated time points. After 2 h, cells were harvested, washed twice and resuspended in fresh HL5 medium. Exocytosis of FITC–dextran was monitored for the next 2 h. At the indicated time points, 500 μl of the cell sample was drawn and added to a tube containing 1.5 ml Sörensen's buffer (16.7 mM Na₂HPO₄/KH₂PO₄, pH 6.0), centrifuged (500 g for 3 min), washed twice with Sörensen's buffer, once with wash buffer (5 mM glycine, 100 mM sucrose, pH 8.5) and then stored on ice. The cells were lysed using 1 ml of lysis buffer containing 0.2% (v/v) Triton X-100. The fluorescence was measured at 492 nm excitation/525 nm emission; the intensity of fluorescence is a measure of the FITC–dextran level within the cells.

Endosomal pH was determined by the dual excitation ratio method as described previously (Rivero and Maniak, 2006) with slight modification. Briefly, ML cells were harvested and resuspended at a concentration of 3×10⁶ cells/ml in HL5 medium containing FITC–dextran (2 mg/ml). After a 10 min pulse, cells were harvested and resuspended in HL5 medium without FITC–dextran. At 10 min intervals, 500 µl of cells was collected and added to a tube containing ice-cold MES buffer (50 mM), washed twice and then resuspended in 1 ml MES. The fluorescence intensity was measured after dual excitation at 450 nm and 495 nm, and emission at 525 nm. The fluorescence ratio (value after excitation at 495 nm/value after excitation at 450 nm) was calculated and the endosomal pH was obtained by comparing to a standard curve. To obtain the standard curve for pH, the fluorescence of FITC–dextran in various pH conditions (pH 4–7) was measured.

ML cells were harvested, seeded onto coverslips with HL5 and incubated at 22°C for 2 h. The cells were then washed with KK2 buffer, fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS) for 10 min, washed with PBS, permeabilized with 0.5% Triton-X 100 in PBS and washed with PBS again. To visualize F-actin, the cells were stained with 300 nM TRITC–phalloidin (Sigma-Aldrich, USA) for 1 h in dark conditions. The samples were washed with PBS and mounted on a slide with AntiFade Gold (ThermoFisher Scientific, USA) mounting solution. The images were taken with an Olympus FV 3000 confocal microscope using a 63× objective lens and the images were processed with ImageJ software (NIH, USA).

The cytoskeletal proteins were isolated as detergent-insoluble fractions using a NP-40 lysis buffer (Chung and Firtel, 1999). ML cells were harvested from HL5 medium, resuspended at 10⁸ cells/ml and lysed with 2× lysis buffer (50 mM Tris-HCl pH 7.6, at room temperature, 200 mM NaCl, 20 mM NaF, 2 mM sodium vanadate, 6 mM sodium pyrophosphate, 2 mM EDTA, 2 mM EGTA, 1× protease inhibitor cocktail, 2% NP-40, 20% glycerol, 2 mM DTT). The lysed samples were vortexed, placed on ice for 10 min, then kept at room temperature for 10 min. The samples were then spun for 4 min at 11,000 g and the collected pellet was washed twice with 1× lysis buffer, and dissolved in 2× SDS gel loading buffer by boiling at 100°C for 10 min. The samples were run on a 12% gel (for F-actin) or a 8% gel (for myosin). The protein bands were stained with Coomassie Blue R250, scanned and changes in F-actin and myosin levels were measured densitometrically, using ImageJ (NIH, USA).

To check myosin function (Tuxworth et al., 1997), axenically grown cells were harvested and resuspended in KK2 buffer, seeded onto six-well plates and allowed to settle. The buffer was replaced with 0.01% sodium azide in KK2 and incubated for 30 min. The fraction of detached and adhered cells was counted after 30 min.

PkcA⁻ cells were grown for 5 days in shaking conditions and an aliquot was transferred to the coverslips. The cells were allowed to settle and, once the cells showed cytofission, they were fixed, stained and imaged as described above (Tuxworth et al., 1997).

The DNA content was determined by fixing ML cells with ice cold methanol for 20 min in −20°C. The fixed cells were placed on a glass slide using DAPI-containing mounting solution (Sigma-Aldrich, USA), imaged on a Nikon Eclipse 80i upright microscope (Nikon, Japan) and analyzed using ImageJ (NIH, USA).

Flow cytometry was carried out to examine cell size and cell cycle as previously described (Chen and Kuspa, 2005). Dictyostelium cells were grown in HL5 medium at 22°C in Petri dishes or with constant shaking (150 rpm). Cells were harvested, washed twice and 1×10⁷ cells were resuspended in KK2 buffer. Then, the cells were fixed with 3 ml 70% ice-cold ethanol and incubated for 30 min at −20°C. Fixed cells were pelleted and resuspended in 1 ml of propidium iodide (PI; Sigma-Aldrich, USA; 10 μg/ml in KK2 buffer). Then, 1 μl of 100 mg/ml DNase-free RNase (Sigma-Aldrich, USA) was added and cells were incubated at 37°C. Flow cytometry was carried out with a CytoFLEX apparatus (Beckman-Coulter, USA). Cell size was determined by forward scatter and cell cycle was analyzed by measuring the fraction of PI stained population. At least 10,000 cells were counted for each analysis and were sampled in biological triplicates.

To identify the subcellular organelles where PkcA–GFP was localized, we used 3 μM Acridine Orange (AO) (Sigma-Aldrich) marking acidic vesicles (Padh et al., 1989), or 200 nM LysoTracker RED (Thermo Fisher Scientific) marking lysosomes (Li et al., 2016) or 5 μM MitoTracker RED (Thermo Fisher Scientific) marking mitochondria (Schimmel et al., 2012). Endosomal vesicles were marked by either FITC–dextran (2 mg/ml) or TRITC–dextran (2 mg/ml). To test the effect of F-actin polymerization on PkcA–GFP localization, 10 μM latrunculin B (Kriebel et al., 2008) (Sigma-Aldrich) was used. To test the effect of PKC activity on PkcA-GFP localization, 10 μM Bis I was used on the cells. The cells were imaged using a Zeiss laser scanning LSM 710 confocal microscope. Imaging was carried out using a pinhole of 1 Airy unit, or, for MitoTracker, a pinhole of 1.36 Airy units (Schimmel et al., 2012).

Dictyostelium cells containing Q103–GFP (polyQ-containing exon 1 of human huntingtin with 103 consecutive glutamines) vector were harvested from ML and at SP, seeded onto coverslips with HL5 and incubated at 22°C for 30 min. Thereafter, the cells were washed with KK2 buffer and fixed in 3.7% formaldehyde in phosphate-buffered saline (PBS) for 10 min, and washed with PBS. The samples were observed using the 63× objective lens of a confocal microscope Nikon A1 HD. Z-stacks of 20 slices were taken and the maximum intensity was projected using the slices. The RGB images were converted to 8-bit images and then analyzed for the presence of Q103–GFP protein aggregates. Q103–GFP protein aggregates form fluorescent foci in the cells. Cells with the foci were counted in each condition and percentage of cells containing the aggregate foci was calculated.

Data analysis was done using GraphPad Prism software (GraphPad Prism software, San Diego, CA). Statistical significance was assessed using a Student's t-test, or by ANOVA as indicated in the figure legends. Significance was defined as P<0.05.

We gratefully acknowledge the help of Dictyostelium Stock Center (Northwestern University, USA) for providing the strains and plasmids used in the study, Dr Richard Gomer for the anti-AprA and anti-CfaD antibodies, Dr Annette Müller-Taubenberger for the coronin A overexpression (corA-OE) plasmid, Dr Simon Alberti for the Q103-GFP plasmid, Dr Michael Gray for the ScPPX1 plasmid, Dr Toshikazu Shiba of RegeneTiss Inc for the polyP standards. We acknowledge the facilities, the scientific technical assistance of the Advanced Microscopy Facility at NCTB, Department of Biotechnology, IIT Madras. S.U. acknowledges the help of Rakesh Mani, Dr Nasna Nassir, Vignesh R., Prajna A. Rai and Pavani Hathi. We gratefully acknowledge Industrial Consultancy and Sponsored Research (ICSR) and IIT Madras for all the support.