PTEN and the PTEN-like phosphatase CnrN have both distinct and overlapping roles in a Dictyostelium chemorepulsion pathway Open Access

The directed movement of eukaryotic cells towards or away from an external stimulus is crucial for neuronal migration, embryogenesis and the trafficking of immune cells during inflammation (Borrell, 2019; García-Cuesta et al., 2019; Glass et al., 2020; SenGupta et al., 2021). Studies on the movement of the model eukaryote Dictyostelium discoideum have revealed mechanisms of chemoattraction, where cells move towards a stimulus. Migrating D. discoideum cells extend pseudopods to adhere to the substrate and contract the trailing edge of the cells to force them forward (Uchida et al., 2003; Van Haastert and Devreotes, 2004).

During development, a gradient of the chemoattractant cyclic adenosine monophosphate (cAMP) activates phosphoinositide 3-kinases (PI3Ks), which increase levels of phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P₃] and active Ras at the D. discoideum cell membrane to form polymerized actin-rich pseudopods projecting towards the source of cAMP (Cheng et al., 2020; Chung et al., 2000; Huang et al., 2003; Kortholt et al., 2013; Park et al., 2004; Sasaki et al., 2004; Zigmond et al., 1997). D. discoideum cells utilize phosphatase and tensin homolog (PTEN) to dephosphorylate phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P₃] into phosphatidylinositol 4,5-bisphosphate [PI(4,5)P₂], which prevents PI(3,4,5)P₃-dependent filamentous actin (F-actin) accumulation and pseudopod formation at the side of the cell facing away from the source of cAMP, causing biased movement of cells towards cAMP (Chung et al., 2000; Funamoto et al., 2002; Huang et al., 2003; Iijima and Devreotes, 2002; Park et al., 2004; Sasaki et al., 2004; Wessels et al., 2007; Zigmond et al., 1997). Also at the side of the cell facing away from the source of cAMP, polymerized myosin II provides contractile force for the cell movement (Egelhoff et al., 1993; Heissler and Sellers, 2016; Levi et al., 2002; Liang et al., 1999; Pasternak et al., 1989). The PTEN-like phosphatase cell number regulator N (CnrN) is a D. discoideum PTEN-like protein with phosphatidylinositol phosphatase activity, sharing 23–25% sequence identity with PTENs, including D. discoideum PTEN (Tang and Gomer, 2008). Loss of CnrN in D. discoideum cells leads to an increase in cAMP production during development (Tang and Gomer, 2008). The loss of PTEN or CnrN causes increased PI(3,4,5)P₃ accumulation, and the loss of PTEN increases F-actin levels after cAMP stimulation (Huang et al., 2003; Iijima and Devreotes, 2002; Tang and Gomer, 2008).

Proliferating D. discoideum cells secrete a protein called autocrine proliferation repressor protein A (AprA) (Brock and Gomer, 2005; Choe et al., 2009; Phillips and Gomer, 2012). AprA inhibits D. discoideum proliferation (Brock and Gomer, 2005). In addition, AprA is a chemorepellent for D. discoideum cells, causing them to move in a biased direction away from a source of AprA (Phillips and Gomer, 2012). In a colony of cells, there will be a high extracellular concentration of AprA at the center, and low concentrations or AprA outside the colony, causing a gradient of AprA at the edge of the colony (Gomer, 2019; Kirolos and Gomer, 2022; Rijal et al., 2019). This gradient of a chemorepellent causes cells at the colony edge to move away from the colony, potentially in search of new food sources (Phillips and Gomer, 2012). In a gradient of AprA, the side of the cell sensing the highest concentration of AprA shows an inhibition of Ras activation, F-actin formation and pseudopod formation, inhibiting the movement of cells towards the higher concentration of AprA, and thus biased movement of cells away from the higher concentration of AprA (Kirolos and Gomer, 2022; Rijal et al., 2019).

In a uniformly high concentration of extracellular AprA, the AprA inhibition of Ras activation and pseudopod formation occurs on all sides of a cell, and this causes cells to round up and stop moving (Kirolos and Gomer, 2022). At the center of a large colony of densely populated cells, the cells have overgrown the local food source and no longer need to move around and hunt for food. In this region of a colony, the AprA concentration is high and relatively uniform (Kirolos and Gomer, 2022). Presumably to conserve energy, the high AprA concentrations causes cells to stop moving and become round (Kirolos and Gomer, 2022; Phillips and Gomer, 2012).

Although PTEN and CnrN both dephosphorylate PI(3,4,5)P₃ into PI(4,5)P₂, and would thus appear to be redundant, AprA-induced chemorepulsion requires both of these enzymes (Herlihy et al., 2013; Phillips and Gomer, 2014; Rijal et al., 2019; Tang et al., 2018). In this report, we elucidate the overlapping and different roles of PTEN and CnrN in the AprA signal transduction pathway.

Dictyostelium cells exposed to a gradient of the chemorepellent AprA exhibit a biased movement away from AprA (Phillips and Gomer, 2012). Both PTEN and the PTEN-like phosphatase CnrN are required for AprA-induced chemorepulsion (Herlihy et al., 2013; Rijal et al., 2019). Expressing PTEN–GFP in pten⁻ cells (pten⁻/pten-gfp) (Iijima and Devreotes, 2002) (confirmed by PCR using gene-specific primers, Fig. 1A,B) restored the ability of cells to move away from AprA (Fig. 2A). This suggests that the insensitivity to AprA observed in pten⁻ cells is specifically due to the loss of PTEN, and not influenced by secondary mutations. Consistent with previous observations in wild-type Ax2, cnrN⁻ and pten⁻ cells (Rijal et al., 2019), AprA did not affect the speed and the persistence of pten⁻/pten-gfp cells (Fig. 2B,C). CnrN localizes to the side of the cell facing AprA in an AprA gradient (Rijal et al., 2019). We observed a uniform distribution of PTEN at the cell periphery in unstimulated cells, and in an AprA gradient, the percentage of cells with PTEN at the side facing AprA increased, whereas the percentage of cells with PTEN at the side facing away from AprA decreased (Fig. 2D,E). Together, these data indicate that in a gradient of AprA, like CnrN, PTEN tends to localize to the side of the cell towards the source of AprA.

AprA inhibits the proliferation of wild-type Dictyostelium cells, and the loss of CnrN abolishes this effect (Brock and Gomer, 2005; Herlihy et al., 2013). To determine whether AprA requires PTEN to inhibit proliferation, proliferating Ax2, cnrN⁻ and pten⁻ cells were incubated with AprA, and the decrease in cell density compared to that seen in a buffer control after a 24-h incubation was determined. As previously observed, AprA inhibited the proliferation of Ax2 cells (Brock and Gomer, 2005) (Fig. 2F). Both cnrN⁻ and pten⁻ cells had a slower proliferation than Ax2, and AprA did not significantly affect their proliferation (Fig. 2F). It is possible that the effect of AprA on the proliferation of cnrN⁻ or pten⁻ cells is obscured due to the ∼two-fold reduced rate of proliferation of already seen for cnrN⁻ or pten⁻ cells compared to that in Ax2 cells. Together, these data suggest that AprA might require both CnrN and PTEN to inhibit proliferation.

PTEN dephosphorylates PI(3,4,5)P₃ to PI(4,5)P₂ at the side of Dictyostelium cells facing away from cAMP, inhibiting the formation of filamentous actin (F-actin) and preventing pseudopod formation (Funamoto et al., 2002; Iijima and Devreotes, 2002; Matsuoka and Ueda, 2018; Sasaki et al., 2007; Wessels et al., 2007). Loss of either PTEN or CnrN in Dictyostelium cells causes overaccumulation of PI(3,4,5)P₃ in response to cAMP stimulation (Iijima and Devreotes, 2002; Tang and Gomer, 2008). To determine how two enzymes with redundant functions, PTEN and CnrN, regulate PI(4,5)P₂ and PI(3,4,5)P₃ levels during chemorepulsion, Ax2, pten⁻, cnrN⁻, pten⁻/pten-gfp and cnrN⁻/cnrN-gfp cells (verified by PCR using gene-specific primers, Fig. 1A,B) were exposed to a uniform concentration of AprA for the indicated amounts of time or an equivalent volume of 20 mM NaPO₄, pH 6.3 buffer for 0 s (control), and phosphoinositide levels were measured. The basal levels of PI(4,5)P₂ were higher in pten⁻ and pten⁻/pten-gfp cells than in Ax2, cnrN⁻ or cnrN⁻/cnrN-gfp cells (Fig. 3A). Both pten⁻ and cnrN⁻ cells had increased basal PI(3,4,5)P₃ levels compared to that seen for Ax2 cells (Fig. 3B). Expressing pten-gfp in pten⁻ cells or expressing cnrN-gfp in cnrN⁻ cells reduced PI(3,4,5)P₃ levels (Fig. 3B). AprA increased PI(4,5)P₂ levels at 5 and 10 min in Ax2 cells (Fig. 3C), and that level was at a maximum at 10 min. Therefore, we opted to expose pten⁻ or cnrN⁻ cells overexpressing PTEN–GFP or CnrN–GFP to AprA for 10 min. In pten⁻ cells, AprA reduced PI(4,5)P₂ levels at 5, 10 and 20 min (Fig. 3D). Expression of PTEN–GFP in pten⁻ cells, as with Ax2 cells, caused AprA to increase PI(4,5)P₂ levels at 10 min (Fig. 3E). AprA did not significantly affect PI(4,5)P₂ levels at any time in cnrN⁻ cells, but as with Ax2 cells, increased PI(4,5)P₂ levels in cnrN⁻/cnrN-gfp cells at 10 min (Fig. 3F,G).

cAMP stimulation of Dictyostelium cells causes an increase in the levels of PIP₃ (Clark et al., 2014) as early as 5 s, and that level of PIP₃ drops to normal after 20 min, but cAMP does not alter the levels of PIP₂ (Clark et al., 2014). In contrast, AprA caused a decrease in PI(3,4,5)P₃ levels at 10 and 20 min in Ax2 cells (Fig. 3H). In pten⁻ cells, AprA did not affect PI(3,4,5)P₃ levels, but as with Ax2 cells, AprA reduced PI(3,4,5)P₃ levels in pten⁻/pten-gfp cells at 10 min (Fig. 3I,J). Similarly, AprA did not affect PI(3,4,5)P₃ levels in cnrN⁻ cells but reduced PI(3,4,5)P₃ levels in pten⁻/pten-gfp cells at 10 min (Fig. 3K,L). Together, these data indicate that PTEN but not CnrN decreases basal levels of PI(4,5)P₂. Assuming that the PTEN levels in pten⁻/pten-gfp are not exactly the same as in Ax2 cells, the proper levels of PTEN are needed to maintain basal levels of PI(4,5)P₂. Both PTEN and CnrN decrease basal levels of PI(3,4,5)P₃. The effect of AprA on the levels of PI(4,5)P₂ depends on CnrN and not on PTEN. In the absence of PTEN but not CnrN, AprA causes an unknown enzyme to decrease levels of PI(4,5)P₂. The effect of AprA on the levels of PI(3,4,5)P₃ depends on both CnrN and PTEN.

In a cAMP gradient, a localized increase in PI(3,4,5)P₃ levels at the side of the cell closest to the source of cAMP increases actin polymerization and pseudopod formation, causing biased movement of cells towards cAMP (Cheng et al., 2020; Chung et al., 2000; Huang et al., 2003; Park et al., 2004; Sasaki et al., 2004; Zigmond et al., 1997). Myosin II stabilizes the cytoskeleton by associating with the actin meshwork and provides force on actin filaments (Egelhoff et al., 1993; Heissler and Sellers, 2016; Levi et al., 2002; Liang et al., 1999; Pasternak et al., 1989). Myosin II is active in its filamentous form, which is negatively regulated by phosphorylation (Liang et al., 1999). Both cnrN⁻ and pten⁻ cells had normal basal total actin and myosin II levels (Fig. 4A,B). In contrast, cnrN⁻ and pten⁻ cells had increased basal F-actin and polymerized myosin II levels (Fig. 4C,D). Loss of CnrN but not PTEN decreased basal phosphorylated myosin levels compared to that seen in Ax2 (Fig. 4E). Similar to its effects on Ax2 cells, AprA did not alter levels of total actin, total myosin II, F-actin and polymerized myosin II in pten⁻ and cnrN⁻ cells (Fig. S1). AprA decreased levels of phosphorylated myosin II after 5 min in Ax2 cells (Rijal et al., 2019) (Fig. 4F). Loss of CnrN or PTEN prevented that decrease in phosphorylated myosin II levels (Fig. 4G,H). Together, these data suggest that both PTEN and CnrN are required for AprA-induced myosin II dephosphorylation.

In a cAMP gradient, the activation of Ras at the front of the cell (the side closest to the source of the cAMP attractant) leads to localized PI3K activation and PI(3,4,5)P₃ production at the front, resulting in actin polymerization and pseudopod formation (Cheng et al., 2020). During chemorepulsion, AprA prevents pseudopod formation by inhibiting Ras activation at the side of the cells facing towards AprA, causing biased movement of cells (Kirolos and Gomer, 2022). To determine whether AprA requires CnrN and PTEN to inhibit Ras activation, Ax2, cnrN⁻ and pten⁻ cells were treated with AprA for 0, 10 and 30 min, and Ras activation was assessed using a pulldown assay of active Ras with Raf-RBD affinity beads. The loss of CnrN or PTEN did not significantly affect the basal levels of total and active Ras (Fig. 5A–C). AprA did not significantly affect total Ras levels in cnrN⁻ and pten⁻ cells at 10 and 30 min (Fig. 5D). As previously observed (Kirolos and Gomer, 2022), AprA reduced active Ras levels in Ax2 cells within 30 min (Fig. 5E). The loss of CnrN or PTEN abolished this effect (Fig. 5E). Together, these data suggest that AprA requires both CnrN and PTEN to inhibit Ras activation.

In a gradient of AprA, Dictyostelium cells inhibit Ras activation and pseudopod formation at the side of the cell facing the source of AprA (Kirolos and Gomer, 2022; Rijal et al., 2019). Prolonged exposure (60 min) of cells to uniform concentration of AprA causes cells to become rounder (Kirolos and Gomer, 2022). To determine whether AprA-induced cell roundness is dependent on cell density, the roundness of Ax2 cells at 1×10⁵, 1.5×10⁵, 5×10⁵ and 10×10⁵ cells/ml densities was measured by determining the ratio of the short and long axes of the cell (short/long) before adding AprA, and 30 min after adding AprA or buffer control. Cell densities did not significantly affect the roundness of cells before adding AprA, but after further incubation for 30 min in the absence of AprA, cells became rounder as cell densities increased (Fig. 6A,B). AprA further increased the roundness of cells as cell densities increased (Fig. 6A,B). To determine whether AprA requires CnrN and/or PTEN to induce cell rounding, Ax2, cnrN⁻ and pten⁻ cells were treated with uniform concentration of AprA or buffer control for 30 min, and cell roundness was determined. AprA increased the roundness of Ax2 and cnrN⁻ but not pten⁻ cells (Fig. 6C,D). Together, these data suggest that AprA induces cell roundness in a cell density-dependent and PTEN-dependent manner.

Dictyostelium cells increase formation of pseudopods and filopods at the front of the cell in a steep gradient of cAMP but do not alter pseudopod formation in a shallow gradient of cAMP (Bosgraaf and Van Haastert, 2009; Heid et al., 2005). Dictyostelium cells use cup-shaped ruffles to uptake liquid nutrient in a process called macropinocytosis (Williams and Kay, 2018). Fast-moving cells have a slow rate of macropinocytosis, and vice versa (Veltman, 2015). An AprA gradient does not alter the rate of pseudopod formation but increases the rate of filopod projections (Rijal et al., 2019). To determine whether AprA requires CnrN and PTEN to regulate membrane protrusions, Ax2, cnrN⁻ and pten⁻ cells were incubated in HL5 medium on a coverslip-bottommed Petri plate in the presence or absence of a uniform concentration of AprA, and images of cells were taken for 5 min. The numbers, sizes and lifespans of filopods, pseudopods and macropinosomes (Fig. S2A) were then determined. Compared to Ax2 cells, cnrN⁻ cells had a reduced number of filopods, and pten⁻ cells had increased filopod size and pseudopod numbers (Fig. S2). In this present study, the size of the pseudopods in Ax2 cells were ∼10 times larger than we found for Ax2 cells previously (Rijal et al., 2019). In the previous study, we used cells taken from shaking suspension culture, and in this study, we used cells that had been growing on surfaces in stationary submerged culture. The difference might thus be due to the different culture conditions. Together, these data suggest that CnrN but not PTEN increases the number of filopods, and that PTEN but not CnrN decreases filopod sizes and the number of pseudopods.

Ax2 cells had an increased number of filopods after 60 min, and AprA did not significantly affect that number (Fig. S3A). This increase might have been due to hypoxia in the confined environment under the condenser. Compared to what was seen in Ax2 cells, AprA increased the number of filopods in cnrN⁻ but not pten⁻ cells over the first 5 min (time 0), and that effect of AprA was lost at 30 and 60 min (Fig. S3B,C). As previously observed (Rijal et al., 2019), AprA did not significantly affect the number of pseudopods. Compared to what was seen in Ax2 cells, AprA decreased the number of pseudopods in cnrN⁻ at 30 min and pten⁻ cells at 0 and 30 min (Fig. S3D–F). In untreated pten⁻ cells at 60 min, the basal number of pseudopods was reduced compared to that for cells at 0 and 30 min (Fig. S3F). As above, this may have been due to hypoxia. Ax2 cells had an increased number of macropinosomes at 30 and 60 min, and AprA prevented that increase (Fig. S3G). AprA did not significantly affect the number of macropinosomes in cnrN⁻ and pten⁻ cells (Fig. S3G–I). Together, these data suggest that for unknown reasons, AprA transiently increases the number of filopods in cnrN⁻ cells. AprA transiently increases the number of pseudopods, and this effect is reversed in cells lacking either CnrN or PTEN. AprA increases the number of macropinosomes, and this effect is reversed in cells lacking either PTEN or CnrN.

AprA increased filopod size at time 0 in Ax2 cells but not in cnrN⁻ and pten⁻ cells (Fig. S4A–C). AprA did not significantly affect the sizes of pseudopods or macropinosomes in Ax2, cnrN⁻ and pten⁻ cells (Fig. S4D–I). AprA increased the lifespan of filopods and pseudopods in pten⁻ but not Ax2 or cnrN⁻ cells at 60 min and did not affect lifespan of macropinosomes in Ax2, cnrN⁻ and pten⁻ cells (Fig. S5A–I). Together, these data suggest that AprA transiently increases the filopods size, and this effect requires either CnrN or PTEN. AprA increases the lifespan of filopods and pseudopods in cells lacking PTEN.

PTEN and the PTEN-like phosphatase CnrN, which dephosphorylate PI(3,4,5)P₃ to PI(4,5)P₂, are necessary for AprA-induced chemorepulsion (Herlihy et al., 2013; Phillips and Gomer, 2014; Rijal et al., 2019; Tang et al., 2018). In this report, we determined how D. discoideum cells utilize PTEN and CnrN in response to AprA, as prevailing evolutional ideology has been that superfluous proteins and functions are lost, as seen by the disappearance of CnrN in higher-level eukaryotes (Tang and Gomer, 2008). We found PTEN decreases basal levels of PI(4,5)P₂ and PI(3,4,5)P₃, and CnrN decreases basal levels of PI(3,4,5)P₃. Both PTEN and CnrN are required to increase the number of macropinosomes by an unknown mechanism, and AprA prevents that increase. AprA requires both PTEN and CnrN to increase PI(4,5)P₂ levels, decrease PI(3,4,5)P₃ levels, inhibit proliferation, decrease myosin II phosphorylation, inhibit Ras activation and increase in filopod sizes, but only requires PTEN to induce cell roundness.

PTEN and CnrN are cytosolic and/or uniformly distributed on the cytosolic side of the plasma membrane of unstimulated D. discoideum cells (Funamoto et al., 2002; Tang and Gomer, 2008). In a cAMP gradient, PTEN localizes to the side of the cell facing away from the source of cAMP, prevents PI(3,4,5)P₃ accumulation via conversion of PI(3,4,5)P₃ into PI(4,5)P₂, and suppresses localized F-actin accumulation and lateral pseudopod formation (Cheng et al., 2020; Funamoto et al., 2002; Iijima and Devreotes, 2002; Matsuoka and Ueda, 2018). In an AprA gradient, similar to CnrN (Rijal et al., 2019), PTEN localizes to the side of the cell facing towards the source of AprA, suggesting that these proteins functions concurrently to suppress PI(3,4,5)P₃-dependent F-actin polymerization and pseudopod formation in the region of the cell closest to the source of AprA.

Our previous work has established that PI3K is not needed during AprA-mediated chemorepulsion (Phillips and Gomer, 2012), and it does not affect total F-actin levels during chemorepulsion. Instead, it induces localization of F-actin and active Ras binding RBD–Raf1–GFP towards the side of the cell opposite to the AprA source (Kirolos and Gomer, 2022; Rijal et al., 2019). Building upon those previous findings, PTEN–GFP localization at the side of cells facing the source of AprA in an AprA gradient suggests a reciprocal regulation mechanism between PTEN and Ras.

Loss or overexpression of PTEN but not CnrN increased basal PI(4,5)P₂ levels, possibly due to dysregulation of CnrN activity as a result of either loss of PTEN or overexpression of PTEN in cells lacking PTEN, and optimum PTEN levels appears to be necessary to maintain basal PI(4,5)P₂ levels. Loss of either PTEN or CnrN increased basal PI(3,4,5)P₃ levels, indicating that both PTEN and CnrN are required to maintain the basal PI(3,4,5)P₃ levels.

AprA increased PI(4,5)P₂ levels and decreased PI(3,4,5)P₃ levels in Ax2 cells. AprA alters PI(4,5)P₂ and PI(3,4,5)P₃ levels at timepoints when the cells start to respond and migrate away from AprA (Phillips and Gomer, 2012; Rijal et al., 2019), and it is possible that cells move away from the source of AprA by inhibiting PI(3,4,5)P₃-dependent pseudopod extension at the side of the cells facing towards the source of AprA. Cells lacking PTEN or CnrN do not move away from the source of AprA (Herlihy et al., 2013; Rijal et al., 2019). AprA reduced PI(4,5)P₂ levels in cells lacking PTEN, but not in cells lacking CnrN, suggesting that loss of PTEN might activate an unknown pathway to cause AprA to decrease PI(4,5)P₂ levels.

The basal levels of total actin were unaltered in pten⁻ and cnrN⁻ cells; however, the proportion of F-actin to total actin was increased in both pten⁻ and cnrN⁻ cells. This increase in F-actin could be due to the increased levels of PI(3,4,5)P₃ levels in pten⁻ and cnrN⁻ cells, which might be the reason why pten⁻ cells have decreased cell migration speed and persistence, and increased basal number of pseudopods and filopods. PTEN might have compensated for the loss of CnrN in cnrN⁻ cells to regulate cell speed, persistence and pseudopods (Rijal et al., 2019).

The proportion of polymerized myosin II to total myosin II was increased in both pten⁻ and cnrN⁻ cells, possibly due to the increased F-actin accumulation. When the myosin II tail is phosphorylated, the phosphorylated amino acids prevent myosin from forming contractile myosin filaments between the F-actin filaments that allow for the contractile activity at the ‘rear’ of the cell during migration (Levi et al., 2002; Liang et al., 1999; Pasternak et al., 1989). We found that cnrN⁻ cells had decreased basal phosphorylated myosin levels compared to Ax2 and pten⁻ cells, suggesting that myosin II is likely to be highly bundled in cnrN⁻ cells and that CnrN negatively regulates myosin dephosphorylation in D. discoideum.

Reduced levels of phosphorylated myosin II and filopod numbers in cnrN⁻ cells suggest that PTEN but not CnrN is indispensable for maintaining cell speed and persistence, and both CnrN and PTEN are necessary to maintain basal F-actin and polymerized myosin II levels, and CnrN, but not PTEN, is necessary for maintaining basal phosphorylated myosin II levels. However, loss of CnrN or PTEN abolished the AprA-mediated decrease in phosphorylated myosin, indicating that both CnrN and PTEN are required for AprA to reduce levels of phosphorylated myosin (Rijal et al., 2019).

In a gradient of AprA, Dictyostelium cells inhibit Ras activation and pseudopod formation at the side of the cell facing the source of AprA (Kirolos and Gomer, 2022; Rijal et al., 2019). Prolonged exposure (60 min) of cells to a uniform concentration of AprA causes cells to become rounder (Kirolos and Gomer, 2022). Loss of PTEN, but not CnrN, caused cells to become rounder in the presence of AprA, suggesting that AprA induces rounding of cells not by inhibiting Ras activation, but instead by activating an unknown PTEN-dependent pathway. It is possible that AprA decreases the number of pseudopods in pten⁻ and cnrN⁻ cells after 0 and/or 30 min exposure of cells to AprA by Ras-independent mechanisms, and the subtle or no effect of AprA on cellular projections such as filopods, pseudopods and macropinosomes indicates that AprA inhibition of Ras activation causes biased movement of cells not by altering number, size and lifespan of these projections, but rather by modifying the location of these projections (Kirolos and Gomer, 2022; Rijal et al., 2019) during cell movement.

Our results indicate that CnrN and PTEN are both needed for AprA-induced chemorepulsion, AprA-mediated inhibition of proliferation and AprA-mediated inhibition of Ras activation. How two enzymes with similar properties could both be needed for these effects is unclear. One possibility is that neither CnrN nor PTEN is present at sufficiently high levels and thus has enough activity alone, and for the effects, cells need both activities. Alternatively, although they both localize to the side of the cell closest to the source of AprA in an AprA gradient, CnrN and PTEN might function in different membrane environments (such as small lipid rafts) that cannot be distinguished by optical microscopy. For instance, D. discoideum possesses five class I PI3Ks (Eichinger et al., 2005). PI3K 1 and 2 producs PI(3,4,5)P₃ in the membrane domains involved in formation of macropinosome ruffles, whereas PI3K 4 produces PI(3,4,5)P₃ in the vesicle membrane at the later stage of formation of macropinosomes (Hoeller et al., 2013), and two different Ras GTPases regulate local PI(3,4,5)P₃ (Hoeller et al., 2013). In support of the idea that CnrN cleaves PI(3,4,5)P₃ that is present in one type of membrane domain and that PTEN cleaves PI(3,4,5)P₃ that is present in a different type of membrane domain, we observed that CnrN and PTEN have different effects on basal levels of PI(4,5)P₂. In addition, CnrN and PTEN have different effects on the ability of AprA to increase the roundness of cells, the number of filopods and pseudopods and the sizes of filopods. In conclusion, the distinct effects of CnrN and PTEN suggest that these enzymes play different roles in D. discoideum signaling pathways, possibly by dephosphorylating PI(3,4,5)P₃ in different membrane domains to regulate cell responses to AprA.

D. discoideum strains were obtained from the Dictyostelium Stock Center (Fey et al., 2019), and were wild-type Ax2, cnrN⁻ (DBS0302655; Tang and Gomer, 2008), pten⁻ (DBS0236830; Iijima and Devreotes, 2002), pten⁻/pten-gfp (DBS0236831; Iijima and Devreotes, 2002), and cnrN⁻/cnrN-gfp (DBS0302656; Tang and Gomer, 2008). Cells were grown at 21°C in shaking culture in HL5 medium (growth medium) (Formedium, UK) and on SM/5 agar plates [2 g/l glucose (VWR, Solon, OH, USA), 2 g/l bacto peptone (BD, USA), 0.2 G/l yeast extract (Hardy Diagnostics, Santa Maria, CA, USA), 0.2 g/l MgSO₄·7H₂O (Thermo Fisher Scientific), 1.9 g/l KH₂PO₄ (VWR), 1 g/l K₂HPO₄ (VWR), 15 g/l agar (Hardy Diagnostics)] (Sussman, 1966) with a lawn of Escherichia coli B/R20 (Dictyostelium Stock Center). 100 µg/ml ampicillin (cat. no A-301-25; Gold Bio; St. Louis, MO) and 100 µg/ml dihydrostreptomycin (cat. no D5155; Sigma, St. Louis, MO) were used to kill E. coli in cultures of Dictyostelium transferred from SM/5 agar plates with shaking (Brock and Gomer, 1999). Cells expressing a selectable marker were grown under selection with the appropriate antibiotics [5 µg/ml blasticidin (cat. no B-800-25; Gold Bio) and 5–10 µg/ml G418 (cat. no N-6386; Sigma)]. GFP-expressing cells were grown under constant selection, and the expression of GFP was confirmed by fluorescence microscopy.

Recombinant AprA was expressed in E. coli, purified, stored in 20 mM NaPO₄ pH 6.2, and checked for purity as described previously (Brock and Gomer, 2005). Chemorepulsion assays and FMI calculations were performed using an Insall chamber as previously described (Rijal et al., 2019).

Imaging of GFP-expressing cells in an AprA gradient was performed as previously described (Rijal et al., 2019). Briefly, pten⁻/pten-gfp cells were maintained in log phase (2×10⁶ to 4×10⁶ cells/ml) in HL5 medium prior to the experiment. 2.4×10⁴ cells per well in a volume of 300 µl were allowed to settle in 8-well slides (# 354118, Corning, Big Flat, NY, USA) for 1 h in a humid chamber. A volume of 1.8 µl of a 50 µg/ml stock of recombinant AprA in 20 mM NaPO₄ pH 6.2 was carefully added to the corner of each well and then left to sit undisturbed in the humid chamber for 20 min to let the gradient establish. Medium was removed from the well and 300 µl of 4% paraformaldehyde (#19210, Electron Microscopy Sciences, Hatfield, PA, USA) in PBS was added to the well and cells were fixed for 10 min. The fixative was gently removed, and cells were washed twice with 300 µl of PBS for 5 min each and then permeabilized for 5 min with PBS containing 0.1% Triton X-100 (# J66624, Alfa Aesar, Ward Hill, MA, USA). Cells were then washed three times for 5 min each with PBS and then stained with a 1:3000 dilution of Phalloidin–Alexa Fluor 555 (#ab176756, Abcam, Cambridge, UK) in 300 µl PBS for 30 min. Cells were washed three times for 5 min each with PBS, and then coverslips were mounted with Vectashield hardset mounting medium with 4′,6-diamidino-2-phenylinodole (DAPI) (Vector Laboratories, Burlingame, CA, USA) following the manufacturer's directions. Images were captured with a 20× objective using a Ti2-Eclipse (Nikon, Kyoto, Japan) inverted fluorescence microscope. Deconvolution of images was undertaken with a Richardson–Lucy algorithm in NIS-Elements AR software (Laasmaa et al., 2011). Random cells from more than four fields of view in individual experiment were scored by observers that were not aware of the experimental conditions as having cytosolic GFP fluorescence, or as having localized GFP fluorescence in the 90° sector of the cell periphery either closest to the AprA source, furthest from the AprA source or the two the 90° sectors perpendicular to the AprA source. For all microscopy, images of a calibration slide (Swift, Carlsbad, CA, USA) were used to generate size bars and calibrate measurements.

Dictyostelium cells in the presence or absence of 300 ng/ml AprA were prepared as previously described (Choe et al., 2009), and the cell density was measured at day 0 and day 1 using a hemocytometer. The percentage proliferation was calculated by dividing the cell density at day 1 by the cell density at day 0.

For the phosphatidylinositol extractions, 1.0×10⁷ cells were stimulated with 300 ng/ml AprA for the indicated time. For controls where buffer not containing AprA was added to cells, the buffer was added, the culture was swirled to mix, and the culture was then immediately harvested, and this time was designated as 0 s. The reaction was stopped with an equal volume of ice cold 1 M trichloroacetic acid (TCA) and incubated on ice for 5 min. Phosphatidylinositol extractions and ELISAs were performed following the manufacturer's directions for the phosphatidylinositol 4,5-bisphosphate (PI(4,5)P₂) Mass ELISA kit (#K-4500) and the phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P₃] Mass ELISA kit (#K-2500s, Echelon Biosciences Inc, Salt Lake City, UT). The PI(4,5)P₂ and PI(3,4,5)P₃ of known concentrations provided in the kit were used as positive controls.

The validation of strains was performed as described (Rijal et al., 2019). Briefly, total RNA was extracted from Ax2, cnrN⁻, cnrN⁻/ cnrN-gfp, pten⁻, and pten⁻/ pten-gfp strains using a Quick-RNA miniprep kit (# R1054, Zymo Research, Irvine, CA), and cDNA was synthesized using a Maxima H minus first-strand cDNA synthesis kit (#K1652, Thermo Fisher Scientific). A PCR was performed to confirm the presence or absence of cDNA from the strains using gene specific primers. A gpdA primer pair served as a positive control. Oligonucleotides for validating strains by PCR were: gpdA forward, 5′-ACCGTTCACGCCATCACTGCC-3′ and reverse, 5′-GACGGACGGTTAAATCGACGACTG-3′; cnrN forward, 5′-ACAGGCTTAGAAGCAAGTTGGAGA-3′ and reverse: 5′-ACGTTGTTGTGAAGGTTGAGTTACA-3′; pten forward, 5′-AGTTGCAGTCTCTAAACAAAAGAG-3′ and reverse: 5′-GGTGCGTCTGATGCTACAAC-3′. Molecular mass standards for gels were 100 bp and 1 kb DNA ladders (GoldBio, St Louis, MO, USA).

Whole-cell actin and myosin II, filamentous actin (F-actin), polymerized myosin II, and phosphorylated myosin II levels were determined exactly as described in Rijal et al. (2019).

For cell roundness measurements, Ax2 cells at 10⁵, 1.5×10⁵, 5×10⁵, 10⁶, and 5×10⁶ cells/ml were cultured in a 96-well, black/clear, tissue-culture-treated, glass-bottom plate (#353219, Corning) in 300 μl HL5 medium. After allowing cells to settle for 30 min, images of cells were taken using a 40× objective on a Ti2-Eclipse (Nikon) inverted fluorescence microscope, imaging at least 40 cells per assay. AprA (or an equivalent volume of 20 mM NaPO₄ pH 6.2) was added to a final concentration of 300 ng/ml. After 30 min, cells were imaged as above. The short (As) and long (Al) axes of cells were measured using Fiji (ImageJ) (Schindelin et al., 2012). Roundness was quantified by calculating the ratio of As/Al. To determine whether AprA requires CnrN and PTEN to induce roundness, Ax2, cnrN⁻ and pten⁻ cells at 1.5×10⁵ cells/ml were assayed as above.

Ras activity in Ax2, cnrN⁻ and pten⁻ cells was assessed using a pull-down assay kit (#BK008-S, Cytoskeleton, Denver, CO, USA) following the manufacturer's instructions as previously described (Kirolos and Gomer, 2022). Total cell lysates or Raf-RBD affinity bead pull-down samples were run on SDS-PAGE gels, western blots of the gels were stained with anti-pan Ras antibodies (1:2000; cat. no AESA02, Cytoskeleton), and densitometry was used to estimate levels of Ras in the Raf-RBD affinity bead pulldown assays. The protein content in the total cell lysates was estimated using a BCA protein assay kit (#23227, Thermo Fisher Scientific) following the manufacturer's instructions. For an accurate comparison of Ras levels in all samples, Ras levels from densitometric analysis of western blots were normalized by dividing Ras levels by the protein content in the total cell lysates.

A hole was punched in the bottom of a 100 mm type 25384-302 petri plate (VWR) with a gas flame-heated 13 mm glass test tube. After sanding the resulting burr and washing the plate with distilled water, a 25×25 mm glass coverslip was attached to the bottom of the plate covering the hole with heated paraffin wax (Gulf Lite, Memphis, TN, USA). Dictyostelium cells from log-phase cultures were washed twice in HL5 medium by centrifugation at 500 g for 3 min and resuspension in 1 ml, and were diluted to 0.15×10⁶ cells/ml in 1 ml. A volume of 300 µl of cells was placed on the coverslip in the Petri dish and allowed to adhere for 30 min. AprA (or an equivalent volume of 20 mM NaPO₄ pH 6.2) was added to the cells to a final concentration of 300 ng/ml. Images of cells were then captured starting at 0, 30 and 60 min in the presence or absence of AprA for 5 min, with 2-s intervals, using a 100× oil immersion Hoffman modulation lens (Modulation Optics, Greenvale, NY, USA) on a Diaphot inverted microscope (Nikon), with the Hoffman condenser in the liquid over the cells to obtain the illumination required for the Hoffman imaging. Images were analyzed using ImageJ (Schneider et al., 2012) to assess the size, lifespan, and count of filopods, pseudopods and macropinosomes. To avoid discrepancies, we scored cells with projections as shown in Fig. S2A. Any cells which had projections that were difficult to distinguish as a macropinosome or a pseudopod, were not scored. The ImageJ manual tracking tool was used to measure both the count and lifespan of the structures, the ImageJ freehand selections tool was used to measure the area of pseudopods and macropinosomes, and the straight measure tool was used to measure the size of filopods.

Statistical analyses were performed using Prism 10 (GraphPad Software, Boston, MA) or Microsoft Excel. A P<0.05 was considered significant.

We thank Dr Sara Milligan for educational conversations and assisting in experiment design.

The peer review history is available online at https://journals.biologists.com/jcs/lookup/doi/10.1242/jcs.262054.reviewer-comments.pdf