PTEN is required in the electrotactic response of Dictyostelium.
In the absence of EF, AX2 wild type (WT) cells moved randomly with the migration directedness close to zero (Fig. 1G). EFs triggered an obvious electrotactic response of Dictyostelium cells toward the cathode in a voltage-dependent manner (Figs. 1A, 1D and 1G; Video S1). 5V/cm EF treatment triggered a clear electrotactic response of WT cells toward cathode when compared with non-treated cells (Fig. 1G, P < 0.05, one-way ANOVA). Electrotactic response of WT cells maximized at 10 V/cm EF, with migration directedness approaching to 1, which indicates that the majority of the cells migrate directionally towards EF vector (Fig. 1G).
Compared with WT cells, pten− cells showed significantly reduced electrotactic response when treated with EFs between 5–15 V/cm (Figs. 1B, E and G, Video S2; P < 0.05 for 5 V/cm, P < 0.01 for 10 and 15 V/cm, one-way ANOVA). The electrotactic response of pten− cells gradually elevated at 20 V/cm, and was fully restored to WT cells level when treated with 30V/cm of EF (Fig. 1G). PTEN re-expression in pten− cells (PTEN/pten−, or pten rescue), on the other hand, completely reinstated the reduced electrotactic response of pten− cells back to WT cells level under all experimental conditions (Figs. 1C, 1F and 1G; Video S3). EF treatment at 10 V/cm or higher significantly increased the trajectory speed of both WT and PTEN/pten− cells, which dictates increased motility of the migrating cells in EFs (Fig. 1H). In contrast, disrupting PTEN abolished the EF-promoted motility enhancement at 20 and 30 V/cm (Fig. 1H; ##, P < 0.01, one-way ANOVA). Putting together, the data above suggest that PTEN is required for electrotaxis of Dictyostelium.
Electrotaxis of Dictyostelium cells requires asymmetrical plasma membrane translocation and activation of PTEN
Using human PTEN-GFP expressing Dictyostelium cells, we monitored the dynamic plasma membrane translocation and redistribution of PTEN during electrotaxis. Similarly as demonstrated in previous studies, asymmetric PTEN plasma membrane recruitment was not observed on Dictyostelium cells in the absence of EF [51, 16]. Compared with no EF control which showed minimal plasma membrane translocation (Figs. 2A and 2B), EF stimulation triggered significantly more posterior plasma membrane translocation of PTEN-GFP than the anterior region (Figs. 2C, 2D; 2J, P < 0.01; Video S4). GFP intensity line scan analysis confirmed the asymmetrical redistribution of PTEN-GFP at the posterior plasma membrane of the cells (Fig. 2E). Fluorescence intensity ratio analysis (posterior vs anterior plasma membrane) further revealed that the PTEN plasma membrane translocation and posterior redistribution started at 42±9.4 sec and peaked at 119±10.7 sec post EF stimulation (Fig. 4A, red line). Latrunculin A (LatA) treatment completely abolished the EF-triggered PTEN posterior plasma membrane redistribution and electrotaxis of the cells (Figs. 2F, 2G; Video S5), which is a fully reversible event when LatA was washed out (Figs. 2H, 2I, 4C, 4D and 4G; Video S6). Interestingly, although LatA fully abolished EF-triggered posterior redistribution of PTEN-GFP (Fig. 2H, P > 0.05 between posterior and anterior within “LatA + EF” group), it did not affect the plasma membrane translocation of PTEN in EF (Fig. 2J, ## P < 0.01 compared between “LatA + EF” and anterior of “EF only” group). The fact that blocking actin function triggered significantly elevated but even translocation of PTEN-GFP to the plasma membrane, suggesting that PTEN mediated electrotaxis requires actin polymerization, via maintaining the biased PTEN signals to the rear region of directionally migrating cells. PTEN mediated electrotaxis was further supported by the following observations: 1). membrane fraction of PTEN expression was detectable as early as 120 sec post EF stimulation (Fig. 2K). 2). EF-triggered PIP3-C8 phosphatase activity peaked at 120 sec post EF treatment (Fig. 2L). 3). Activation of PTEN phosphorylation from both plasma membrane fraction and whole-cell lysate (Fig. 2K) also coincided with the PTEN posterior plasma membrane translocation at 120 sec after EF exposure (Fig. 4A, red line). And 4). EF-triggered PTEN posterior relocalization preceded the electrotactic response of WT cells (Fig. 4B, red line, peaked at 200 sec post EF stimulation), which was abolished by PTEN knockout (Fig. 4B, blue line).
PTEN-dependent posterior plasma membrane translocation of myosin II in EF
Myosin II is another posterior signaling regulator in close association with PTEN, which redistributed to the posterior plasma membrane of chemotaxing cells [42, 52]. To explore the role of myosin II in electrotaxis, we tested the dynamic redistribution of myosin and its correlation with PTEN during electrotaxis of Dictyostelium on myosin II-GFP/WT and myosin II-GFP/pten− cells. Similar with the findings from PTEN-GFP, localized plasma membrane recruitment of myosin II was also detected at the rear of electrotaxing Dictyostelium when treated with EFs at 10 V/cm (Figs. 3A-3E; 3R, P < 0.01 compared between posterior and anterior myosin II-GFP within the “EF only” group; Video S7). Fluorescence intensity ratio analysis (posterior vs anterior plasma membrane) further revealed that the myosin II plasma membrane translocation and posterior redistribution started at 102±8.9 sec and peaked at 183±11.3-sec post EF stimulation (Fig. 4A, blue line). LatA-treatment abolished EF-triggered asymmetric redistribution of myosin II-GFP (Figs. 3F and 3G; Video S8), suggesting myosin II relocalization during electrotaxis is actin-dependent. This is a reversible event since washout of latA fully restored the asymmetrical redistribution of myosin II-GFP to the posterior plasma membrane of the electrotaxing cells (Figs. 3H-3K;3R, P < 0.01 compared between “washout” and “LatA” groups; Video S9). In contrast, Pten null cells lost posterior membrane translocation of myosin II-GFP in EF completely compared with WT cells evenly in the highest EF tested (Figs. 3L-3Q; 4A, orange line; 3R, P < 0.01 compared between “pten null” and “WT EF only” groups; Video S10), suggesting EF-triggered myosin posterior redistribution is PTEN dependent.
Pten asymmetric redistribution is an early event preceded myosin II in electrotaxis
To further explore the spatial and temporal correlation between PTEN and myosin II during electrotaxis, we conducted a time-lapse analysis to exam the plasma membrane translocation of PTEN-GFP vs myosin II-GFP in WT cells under 10 V/cm EF. Posterior vs anterior GFP intensity ratio analysis was performed at all time points to elucidate the asymmetric translocation of the GFP signals toward posterior of the electrotaxing cells. The higher the intensity ratio is, the more asymmetric posterior redistribution of GFP signals the cells generate. Both PTEN-GFP and myosin II-GFP showed a gradually increased intensity ratio for 80 seconds until they reached the maximum level in electrotaxing WT cells (Fig. 4A, red & blue lines).
Interestingly, the asymmetric membrane relocalization of PTEN-GFP in EF was an early event than the asymmetric recruitment of myosin II-GFP. In comparison with the PTEN-GFP intensity ratio elevation which started from 42±9.4-sec post EF treatment (Fig. 4A, red line), there was ~ 60 seconds delay of the myosin II-GFP rear redistribution when treated with the same EF (Fig. 4A, blue line). The time-lapse immunoblotting analysis further confirmed that compared with phospho-PTEN (pS380) expression within the plasma membrane fraction which became detectable from 120 sec (Fig. 2K), the myosin II was only expressed on plasma membrane from 180-sec post EF stimulation (Figs. 4H, 4I). EF-triggered membrane translocation and posterior redistribution of myosin II-GFP were fully abolished in Pten null cells (Figs. 3L-Q; 4A; 4H; 4I, orange lines). Putting together, these data suggest that PTEN asymmetric activation might be an upstream regulator of myosin II in the electrotaxis of Dictyostelium cells.
Further time-lapse analysis comparing between electrotactic response and PTEN / myosin II membrane translocation of the cells revealed that, PTEN/myosin asymmetric redistribution were earlier events than the electrotactic response of the cells. The electrotactic response of cells initiated at 124±8.3 sec and peaked at 211±10.4 sec (Fig. 4B, red line), which is ~ 80 sec or ~ 20 sec behind the PTEN-GFP or myosin II-GFP membrane redistribution, respectively (Fig. 4A, red and blue lines).
To further support the causal link of the delayed myosin II redistribution downstream of PTEN rather than due to a slower accumulation of myosin II activation in EF, we temporarily blocked the asymmetric redistribution of PTEN-GFP & myosin II-GFP with LatA while maintaining their saturated activation with continuous EF treatment, then washed out LatA to restart the asymmetric redistribution analysis. LatA washout restored the EF-controlled posterior membrane translocation of PTEN-GFP from 122±8.7 sec and peaked at 181±9.4 sec (Fig. 4G, red line). In comparison, LatA washout also reinstated the posterior redistribution of myosin II-GFP in EF, but with ~ 60 sec delay after PTEN-GFP (Fig. 4G, blue line).
Taken together, the spatial-temporal dynamic observation above suggests the causal link that electric signals triggered PTEN asymmetric plasma membrane translocation and activation, and subsequent myosin II posterior redistribution to regulate the electrotactic response of Dictyostelium.
PTEN dependent anterior plasma membrane translocation of PH-Crac in EF
Since PH-Crac binds to PIP3 and asymmetrically redistributed to the leading edge of migrating cells [39], we examined PHCrac-GFP expression as an indicator of PIP3 signaling in WT and pten null cells during electrotaxis. Dictyostelium cells treated with EFs at 15 V/cm or lower did not show an asymmetric redistribution of PHCrac-GFP, similarly as previously reported [7]. In contrast, Higher EF at 20 V/cm and above triggered electrotaxis of the cells with asymmetric recruitment of PHCrac-GFP to the anterior plasma membrane of electrotaxing cells (Figs. 5A and 5B; 5F, P < 0.01 compared between the anterior and posterior membrane of EF-treated WT cells; Video S11). Interestingly, in contrast with the actin-dependent PTEN and myosin II posterior redistribution in EFs (Figs. 2D and 3F), PH-Crac anterior redistribution did not require actin polymerization when treated with LatA in EF (Figs. 5C; 5F, P < 0.01 compared between the anterior and posterior membrane of LatA-treated WT cells in EF; Video S11). EF-triggered PH-Crac anterior redistribution is PTEN dependent since PH-Crac translocated evenly to the entire plasma membrane of pten− cells in EF (Figs. 5D; 5F, P > 0.05 compared between the anterior and posterior membrane of EF-treated pten− cells; Video S12). Time-lapse analysis on GFP intensity ratio revealed that PH-Crac anterior redistribution peaked at ~ 180 sec post EF treatment which was ~ 60 sec later than PTEN-GFP posterior relocalization (Fig. 5E). This is in agreement with the membrane fraction immunoblotting analysis that PHCrac-GFP was only detectable from 180 sec post EF treatment (Fig. 5G), compared with PTEN-GFP which was shown from 120 sec onwards in EF (Fig. 2K). These data further suggest that PTEN asymmetric redistribution might be an upstream event regulating PH-Crac activity in electrotaxis.
PTEN mediated electrotaxis via maintenance of biased PHCrac-GFP expressing pseudopod protrusion toward the cathode
EF triggered a biased distribution of PHCrac-GFP positive pseudopod protrusion at the anterior region of the electrotaxing cells (Fig. 6). Compared with the no EF control group which showed randomly distributed pseudopod protrusions (Fig. 6C), there was a clear shift of the newly formed PHCrac-GFP pseudopods at the anterior region of the migrating WT cells toward cathode (Figs. 6A-6C). We recorded 51% of the total pseudopods orientating toward EF vector (0–30 degree) compared with 18% in the control group (0–30 degree; Fig. 6K). Interestingly, compared with the pseudopod production rate of no-EF WT control cells at 23±5.1 sec per pseudopod (s/ps), EF-triggered a much longer lifetime of the PHCrac-GFP positive pseudopods at the forwarding direction in EF (0–30 degrees) than the rest of the orientations (Fig. 6D), with a much higher average production rate at 55±6.2 s/ps within 0–30 degrees toward EF than the protrusions against EF (90–180 degrees) at 24±5.5 s/ps (Fig. 6K). This is in sharp contrast with the chemotaxing cells, which showed uniform pseudopod production rate at ~ 23 s/ps irresponsive to chemoattractant gradient [53]. More strikingly, on the contrary to the speculation that EF might have promoted more pseudopods to expedite the sharp membrane translocation of myosin / F-actin and increased motility in electrotaxis, our data demonstrated that EF significantly suppressed the total number of PHCrac-GFP pseudopod protrusions (Fig. 6I, P < 0.05), while significantly increased the total protrusion lifetime compared to no EF control (Fig. 6J, P < 0.01). These data above suggested that EF persistently maintains more persistent PHCrac-GFP expressing pseudopods in the forwarding direction rather than generates an increased number of protrusions to facilitate the electrotactic response.
In contrast with WT cells (Figs. 6A-6D), pten− cells completely lost biased redistribution of PHCrac-GFP positive pseudopods in EF, as shown by the evenly distributed protrusions across all orientations (Figs. 6E-6G). Compared with WT cells which had 51% PHCrac-GFP pseudopods concentrated in the forwarding direction in EF (0–30 degrees), there was a significantly reduced proportion of protrusions with forwarding orientation (18%) in pten− cells when treated with the same EF (Fig. 6K). On the other hand, WT cells showed significantly reduced PHCrac-GFP positive pseudopods at the posterior regions of the electrotaxing cells (19%; Fig. 6K), while pten knockout produced evenly distributed PHCrac-GFP pseudopods and significantly more pseudopods at the posterior regions of the cells (48%; Fig. 6K), with the total protrusion number more than doubled compared with WT cells (Fig. 6I, P < 0.01; Fig. 6K).
Interestingly, EF-promoted protrusions in pten− cells are short-lived. Compared with the average lifetime of WT pseudopods of all orientation at 45±5.8 sec, pten− cells showed a significantly reduced average lifetime at 20±5.9 sec (Fig. 6K). This was further confirmed when comparing the average lifetime of pseudopods in the forwarding direction (0–30 degrees) of EF: pten− cells showed a reduced average lifetime of pseudopods at 24±4.8 s/ps compared with WT cells at 55±6.5 s/ps in the forwarding direction (Fig. 6K). As a combined effect of the observations above, EF triggered significantly increased overall lifetime of PHCrac-GFP pseudopods facing 0–30 degrees of electrotaxing direction (62% or 13380 sec), which was notably diminished by pten knockout (22% or 4580 sec). On the other hand, EF significantly suppressed the overall PHCrac-GFP pseudopod lifetime in the posterior region of the WT cells (10% or 2200 sec) compared with that of pten− cells (46% or 9560 sec). These data above is in harmony with the dynamic changes of pseudopod protrusion length: EF-triggered significantly reduced pseudopod length at the posterior region, and increased protrusion length at the anterior area of the electrotaxing cells (Fig. 4A: black = anterior; green = posterior).
PTEN promoted anterior redistribution of F-actin colocalized with PH-Crac signaling
EF triggered posterior colocalization of myosin and phospho-PTEN (Figs. 7A-7D). Phospho-PTEN and PH-Crac localized in the opposite directions toward anode and cathode, respectively (Figs. 7E-7H). At the same time, EF also triggered anterior redistribution of F-actin, which colocalized with PH-Crac at the leading edge of electrotaxing cells (Figs. 7I-7L). Pten knockout fully abolished the F-actin redistribution response to EF treatment (Fig. 7M), with evenly distributed number of pseudopods across all directions (Fig. 7N, grey columns) compared with the anterior relocalization of WT pseudopods (Fig. 7N, red columns).