Introduction

Phosphatase and tensin homolog (PTEN) is a dual-specificity lipid and protein phosphatase. Mutations in PTEN are associated with various cancers (Steck et al. 1997). PTEN regulates the dephosphorylation of phosphatidylinositol 3,4,5-triphosphate (PIP3) to form phosphatidylinositol 4,5-bisphosphate (PIP2) (Campbell et al. 2003). Thus, PTEN can suppress the phosphoinositide 3-kinase (PI3K) pathway and inhibit downstream signaling events that regulate cell migration (Stambolic et al. 1998; Tamura et al. 1998). In addition, PTEN plays an important role in modulating actin dynamics (Li et al. 2014), which is important for cell movement and mediating the direction of cell migration (Fife et al. 2014).

Chemotaxis is the process by which cells migrate directionally along gradients of chemoattractants, such as adenosine 3′,5′- cyclic monophosphate (cAMP) (Xu et al. 2005). The cytoplasmic distribution of PTEN is essential for the directed migration of single cells (Funamoto et al. 2002). In experiments with Dictyostelium, individual cells moved rapidly toward a micropipette approximately 40 s after the addition of 1 µmol L−1 cAMP (Parent et al. 1998). Before this directed migration began, PTEN decreased at the front edge of the cell and accumulated at the rear of the cell 15 s after the cells were exposed to a chemoattractant gradient and PTEN was continuously distributed at the rear of the cell during directed migration (Funamoto et al. 2002). In contrast, PIP3, a substrate of PTEN, was localized at the front edge of neutrophil-differentiated HL-60 cells 13 s after the addition of the chemoattractant before the directional migration began and continued to accumulate at the front edge of the cells (Servant et al. 2000). This may be related to localized actin polymerization and pseudopodia formation. In addition, loss of PTEN promoted cytoskeletal rearrangement (Lin et al. 2015). PTEN-deficient Dictyostelium cells did not migrate toward the micropipette following cAMP exposure and filamentous actin levels were sixfold higher compared with that of wild-type cells 120 s after the addition of 1 µmol L−1 cAMP; this suggests that defective chemotaxis is specifically the result of the loss of PTEN (Iijima and Devreotes 2002). The local distribution of PTEN and PIP3 yields cells with functionally and morphologically distinct front and rear edges (Arai et al. 2010). PTEN is involved in directional sensing and PTEN is localized at an early stage before directed migration in single cells.

In confluent cell monolayers, directed migration can be induced by mechanical scratching or wounding. Surrounding cells initiate migration to the wounded area when the monolayer is scratched with a micropipette tip. In particular, when confluent MCF7 cells are mechanically wounded, surrounding cells migrate to the site of the wound within 24 h and a loss of PTEN can enhance this directed migration (Heering et al. 2009). Besides, a loss of PTEN enhanced the velocity and distance of migration in a monolayer of astrocyte cells 24 h after scratching by a microinjection needle, which was correlated with PTEN controlling cell–cell junctions to influence intracellular actin in moving cells (Peglion et al. 2022). In addition, after the monolayer was scratched, human glioma cells migrated to the wound area within 16 h. PTEN inhibits this directed migration by disturbing the process through its C2 domain (Raftopoulou et al. 2004). In lung epithelial monolayers, PTEN inhibition resulted in a decrease in cell stiffness near the wound edge, which in turn significantly enhanced the directed migration 2 h after scratching (Mihai et al. 2012). Although the inhibitory effect of PTEN has been well-studied, there are no reports describing the intracellular localization of PTEN in monolayers, particularly before directed migration.

The localization of PTEN in the cell nucleus has been found to be related to cell proliferation, apoptosis, and cellular stability (Planchon et al. 2008). The PTP and the C2 domains of PTEN contain several nuclear localization signal (NLS)–like sequences (Chung et al. 2005; Chung and Eng 2005), which are important for controlling PTEN’s subcellular localization and function (Gil et al. 2007). Mutations in the PTEN N-terminal region affect its nuclear localization and the ability of PTEN to regulate mammalian cell growth (Denning et al. 2007; Gil et al. 2015). The presence of PTEN in the nucleus is regulated by complex mechanisms, including diffusion (Liu et al. 2005), monoubiquitylation-dependent import (Baker 2007), and interactions with major vault protein (MVP) (Chung et al. 2005). Furthermore, the binding and dissociation of the small GTPase Ran to importin-β proteins are necessary for importin-mediated nuclear transport (Görlich and Kutay 1999). Coexpression of a dominant negative Ran GTPase protein has been shown to inhibit PTEN accumulation in the nucleus, suggesting that importin-β may play an active role in PTEN nuclear accumulation (Gil et al. 2006). Therefore, clarifying the relationship between importin proteins and PTEN nuclear localization is crucial for a more comprehensive insight into PTEN’s subcellular localization mechanisms. Specifically, we investigated the influence of importin-β inhibition on PTEN nuclear localization before directed cell migration.

Single-cell migration is based on the formation and rearrangement of front-rear polarity orientation, which is related to the intracellular distribution of PIP3 and PTEN (Gerisch et al. 2012). In particular, the spatiotemporal distribution of PTEN and PIP3 has been studied in single cells. The response of PTEN occurs at an early stage before directed migration begins, highlighting their important roles in determining cell directionality. However, in confluent cell monolayers, unlike single cells, each cell is surrounded by neighboring cells, in which stable interconnections and mutual constraints are present (Scarpa and Mayor 2016). Thus, cell migration in monolayers exhibits different mechanisms compared with that of single cells. However, there is little information regarding the status of PTEN specifically before directed migration, although there have been studies describing the role of PTEN during directed migration. Therefore, understanding its state before migration is essential for initiating cell migration in monolayers. In this study, we wounded one cell in confluent fetal rat skin keratinocytes (FRSKs) and examined the spatiotemporal distribution of PTEN in the cells adjacent to the wound before the start of directed migration. In addition, we determined the effects of the actin cytoskeleton and importin-β on the dynamics of PTEN.

Materials and methods

Cell culture

Fetal rat skin keratinocytes (FRSKs) were purchased from the Japanese Cancer Research Bank (Japan) in September 2023 and were serially passaged, proliferating for less than 6 mo. FRSKs were cultured using Dulbecco’s modified Eagle’s medium (DMEM; Gibco™, Thermo Fisher Scientific, Waltham, MA) containing 10% fetal bovine serum (Biological Industries, Kibbutz Bei t-Haemek, Israel) and a 1% antibiotic–antimycotic solution (10,000 U ml−1 penicillin, 10,000 μg ml−1 streptomycin, and 25 μg ml−1 amphotericin B, Gibco™, Thermo Fisher Scientific, Waltham) in a humidified incubator containing 5% CO2 at 37 °C.

cDNA and transfection

GFP-PTEN was a gift from Alonzo Ross (Addgene plasmid #13,039; http://n2t.net/addgene:13039; RRID:Addgene_13039). PH-Akt-GFP was a gift from Tamas Balla (Addgene plasmid #51,465; http://n2t.net/addgene:51465; RRID:Addgene_51465). pcDNA3-mRFP was a gift from Doug Golenbock (Addgene plasmid #13,032; http://n2t.net/addgene:13032; RRID:Addgene_13032). pcDNA3-EGFP-Rac1(wt) was a gift from Klaus Hahn (Addgene plasmid #13,719; http://n2t.net/addgene:13719; RRID:Addgene_13719).

FRSKs were seeded into a 35 mm glass-bottomed dish and cultured to 60–70% confluence prior to transfection. The cells were washed twice with Opti-MEM (Gibco™; Thermo Fisher Scientific) and cultured for 1 h with 1 ml Opti-MEM as the medium. Then, 2 μl of 1 μg μl−1 DNA (GFP-PTEN, PH-Akt-GFP, pcDNA3-EGFP-Rac1, or pcDNA3-mRFP) and 4 μl of HilyMax (Dojindo Laboratories, Kumamoto, Japan) were mixed in 100 μl of Opti-MEM and the mixture was added to the glass dish. After a 4-h incubation, the medium was replaced with 2 ml of DMEM (10% fetal bovine serum, no antibiotics) at 37 °C. The FRSKs grown to confluence were used for experiments 24 h following transfection.

Calcium imaging

Fura-2-acetoxymethyl ester (Fura-2-AM; Invitrogen, Carlsbad, CA) was used as a Ca2+ indicator. Transfected FRSKs in each dish were washed twice with 1 ml 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)–buffered saline (HBS) and loaded with 5 µmol L−1 Fura-2-AM in HBS for 30 min at 25 °C. Transfected FRSKs in each dish were washed twice with 1 ml HBS and 2 ml of HBS was added as the buffer for observation. Images of Fura2-AM were observed following excitation at 380 nm and fluorescence at 510 nm.

Treatment with inhibitors

N-(1-Phenylethyl)-2-(pyrrolidin-1-yl)quinazolin-4-amine (Importazole; Focus Biomolecules, Plymouth Meeting, PA) is a specific inhibitor for importin-β-mediated nuclear import in cells. A 20 mmol L−1 Importazole solution was diluted to 40 µmol L−1 before treatment. Transfected FRSKs in some dishes were washed twice with 1 ml HBS and a 40 µmol L−1 Importazole solution was added. Then, the FRSKs were prepared for experiments after a 30-min incubation.

(3S,3aR,4S,6S,6aR,7E,10S,12R,13E,15R,15aR)-15-(acetyloxy)-3,3a,4,5,6,6a,9,10,12,15-decahydro-6,12-dihydroxy-4,10,12-trimethyl-5-methylene-3-(phenylmethyl)-1H-cycloundec[d]isoindole-1,11(2H)-dione (cytochalasin D; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) was used as an inhibitor of the actin cytoskeleton. A 1.0 mmol L−1 cytochalasin D solution was diluted to 2.5 µmol L−1 before treatment. Transfected FRSKs in some dishes were washed twice with 1 ml HBS and 2.5 µmol L−1 cytochalasin D solution was added. Then, the FRSKs were prepared for experiments after a 30-min incubation.

Mechanical wounding and imaging

In confluent cell monolayers, cells tend to migrate toward the wounded area after a wound is generated with a fine glass micropipette (Arai et al. 2017). A micropipette puller (P-97; Sutter Instrument Company, Novato, CA) was used to melt and stretch a borosilicate tube for making the fine glass micropipette, which was less than 3 µm in diameter. A single cell was mechanically wounded using a fine glass micropipette in a monolayer FRSKs. A manipulator (InjectMan 4; Eppendorf, Germany) was used to control the micropipette approach to the cell. When the micropipette tip approached the target cell, the cell was stimulated by immediately dropping the micropipette. We determined whether the cell was wounded successfully by observing a decrease in the fluorescent intensity of the Ca2+ indicator Fura2-AM (Supplement Movie 1). The results of fluorescence recovery were considered an unsuccessful wound and were excluded from the statistical analysis. On each experimental day, we prepared 3 culture dishes and measured 2–3 cells for each dish during the experiment. N represents the total number of culture dishes and n represents the total number of cells. An inverted microscope (Eclipse TE2000-S; Nikon Corp., Tokyo, Japan) fitted with a 40 × oil-immersion lens (S-Fluor, numerical aperture = 1.3; Nikon Corp.) was used to observe the expression of GFP-PTEN, PH-Akt-GFP, and the change in protein distribution within the cell. The excitation wavelength was cycled between 380, 488, and 583 nm to observe the fluorescence of intracellular Ca2+, Green Fluorescent Protein (GFP), and Monomer Red Fluorescent Protein (mRFP). ImageJ software was used to process and analyze the fluorescent images (Schneider et al. 2012). The images of GFP and mRFP were visually displayed in green and red, respectively. Because fluorescence gradually decreases with time, all of the figures in this study were subject to fading correction by dividing the fade ratio (Supplement Fig. 1) at the corresponding time using “Process-Math” in ImageJ. The fluorescence intensity at any time divided by the initial fluorescence intensity at 0 min was used to define the relative intensity, which represents the level of the observed subjects in the regions of interest (ROI). Images of translocation are obtained by adding a value of 128 to the original image and then subtracting the original image captured at t = 0 min using the “Process-Image Calculator” in ImageJ. The initial cell outline is utilized as a reference for images depicting translocation, preventing the impact of local cellular deformations on the resultant image. Furthermore, the initial background value of the resulting image is set to 0.

Figure 1.
figure 1

Cell elongation at the edge after neighbor cell wounding. (A) Images of the cell including EGFP-Rac1 after neighboring cell wounding. Cell boundaries at 0 min are indicated by the dotted line. White scale bar: 5 μm. (B) Quantification of the cell elongation area in a time series after wounding. N = 9, n = 16. Mean ± standard error of the mean. *P < 0.05 vs. 1 min (two-way ANOVA with post hoc Dunnett’s test).

Full size image

Quantitative analysis of cell elongation for determining the initiation of cell migration

Rac1 localizes anterior to the front edge of migrating cells (Yamao et al. 2015). Therefore, we analyzed the area including EGFP-Rac1 as the area of cell elongation after wounding. The elongation of FRSKs was calculated from the resulting images. Because cell migration begins with anterior signaling and protrusion formation (Ridley et al. 2003), we analyzed the area protruding after 1 min as the cell elongation after wounding. When cells significantly elongate, it is considered that cell migration has begun.

Statistical analysis

Statistical analyses included an analysis of variance. Multiple comparisons were performed using a two-way ANOVA followed by Dunnett’s test or Sidak’s test. A P value < 0.05 was considered a statistically significant difference, whereas ns indicates not significant. All graphs are presented as the mean ± standard error of the mean. The number of glass dishes and cells are denoted by N and n, respectively.

Results

Cells start to migrate 10 min after neighboring cells are wounded

To determine the time before initiation of the directed migration in confluent cell monolayers, we examined the area of cell elongation. Figure 1A shows images of FRSKs, including EGFP-Rac1, after wounding its neighboring cells in monolayers. The FRSKs were gradually elongated over time toward the wounded area. Figure 1B shows that the average elongation area per cell increased over time after wounding. There was no statistically significant difference within the first 10 min following wounding; however, the elongation area after 20 min was statistically larger compared with that after 1 min. Because elongation at the edges of the cells is considered an early event in directed cell migration, we defined the time between 0 and 10 min after wounding as the period before cell migration. During this period, we examined the spatiotemporal distribution of PTEN and its substrate PIP3.

PTEN translocated to the nucleus before directed migration

Figure 2 shows the dynamics of GFP-PTEN before directed migration in FRSKs. Figure 2A and B show the fluorescent images of the GFP-PTEN translocation within the cells. The fluorescence of GFP-PTEN decreased in the area near the wounded cell and increased in the nucleus (Supplement Movie 2 and 3). To eliminate the potential influence of fluorescent protein on PTEN nuclear localization, we also observed the Monomer Red Fluorescent Protein (mRFP) before directed migration by co-transfecting pcDNA3-mRFP and GFP-PTEN. In contrast to GFP-PTEN, the mRFP did not exhibit any sustained changes in the cell nucleus (Fig. 2C). Considering the cytoplasmic and nuclear localization of PTEN, we designated three regions of interest (ROIs) as near, far, and nucleus in the cells based on the distance from the wounded cell (Fig. 2D); the near and far ROI were within a maximum distance of 4 µm away from the edge of the cell (Arai et al. 2017). Figure 2E shows the changes in the relative fluorescent intensity of GFP-PTEN. The relative fluorescence intensity was significantly lower after 40 s in the near ROI than that at 0 min, and reached a minimum value of 0.85 ± 0.03 at 2 min. In contrast, the relative fluorescence intensity in the nucleus ROI increased and reached a maximum value of 1.11 ± 0.01 at 10 min. The relative fluorescence intensity after 2 min was significantly higher than that at 0 min in the nucleus. Furthermore, we examined the relative fluorescent intensity of mRFP in the nucleus. The relative fluorescent intensity of mRFP remained close to the initial value of 1.0 after 5 min and was not significantly different from that at 0 min (Fig. 2F). These results suggest that PTEN translocates from the cytoplasm to the nucleus after neighboring cell wounding and before directed migration in monolayers.

Figure 2.
figure 2

The dynamics of GFP-PTEN before directed migration in monolayers. (A) Representative fluorescence images of GFP-PTEN before directed migration. (B) Images of GFP-PTEN translocation obtained by subtracting the original image in (A) captured at t = 0 min and setting the initial value as 0. The intensity of GFP-PTEN is shown using a pseudo color scale. Greater than 0 is considered an increase compared with the initial value. (C) The fluorescence images of pcDNA3-mRFP before directed migration. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (D) The region of near, far, and nucleus in the cells. (E) Relative fluorescent intensity of GFP-PTEN in a time series. (F) Relative fluorescence intensity of pcDNA3-mRFP in a time series. N = 8, n = 16. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test).

Full size image

Effects of Importazole on PTEN nuclear translocation and directed cell migration after neighboring cell wounding

The nuclear translocation of PTEN may require the involvement of transport proteins. To gain further insight into the PTEN response, we used Importazole as an inhibitor for importin-β-mediated nuclear import and examined the dynamics of PTEN after neighboring cell wounding during Importazole treatment (Fig. 3A, B). We examined the relative fluorescence intensity, and GFP-PTEN showed no difference in relative fluorescence intensity compared to 0 min in the nucleus (Fig. 3C). Moreover, there was a significant difference in the relative fluorescence intensity of GFP-PTEN after 4 min between the Importazole-treated and untreated groups (Fig. 3D). Importazole treatment inhibited the PTEN response, indicating that importin-β may be involved in the nuclear translocation of PTEN after wounding in confluent cell monolayers. The effect of Importazole treatment on directed cell migration was also examined. Figure 4A is an image of FRSK after neighboring cell wounding with Importazole treatment. Figure 4B shows that the average elongation area per cell increased with time after neighboring cell wounding; the elongation area at 30 min was significantly smaller than that of the untreated control. Importazole treatment inhibited the increase in cell elongation, suggesting an active role in the nuclear localization of PTEN in directed cell migration in cell monolayers.

Figure 3.
figure 3

The effect of the importin-β inhibitor Importazole. (A) Representative fluorescence images of GFP-PTEN during treatment with Importazole (40 µmol L−1) before directed migration. (B) Images of GFP-PTEN translocation obtained by subtracting the original image in (A) captured at t = 0 min and setting the initial value as 0. The intensity of GFP-PTEN is shown using a pseudo color scale. Greater than 0 is considered an increase compared with the initial value. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (C) Relative fluorescent intensity of GFP-PTEN under Importazole treatment. N = 8, n = 16. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test). (D) Comparison of GFP-PTEN relative fluorescence intensity in the nucleus with Fig. 2E. N = 8, n = 16. Mean ± standard error of the mean. *P < 0.05 vs. untreated (two-way ANOVA with post hoc Sidak’s test).

Full size image
Figure 4.
figure 4

Cell elongation at the edge after neighbor cell wounding under Importazole treatment. (A) Images of the cell including EGFP-Rac1 after neighbor cell wounding with Importazole treatment (40 µmol L−1). Cell boundaries at 0 min are indicated by the dotted line. White scale bar: 5 μm. (B) Quantification of the cell elongation area in a time series after wounding. N = 8, n = 11. Mean ± standard error of the mean. *P < 0.05 vs. untreated (two-way ANOVA with post hoc Sidak’s test).

Full size image

Accumulation of PIP3 before directed migration

The PIP3 signal may play a potential role in PTEN localization; we examined PIP3 dynamics using the indicator PH-Akt-GFP before directed migration (Fig. 5). Figure 5A and B show the fluorescent images of the PH-Akt-GFP translocation in the cells. The fluorescence of PH-Akt-GFP increased in the region near the wounded cell (Supplement Movie 4 and 5). We examined the relative fluorescence intensity in the near, far, and nucleus ROI in the cells based on the distance from the wounded cell (Fig. 5C). The relative fluorescence intensity of PH-Akt-GFP was increased in the near ROI and reached a maximum value of 1.36 ± 0.03 at 10 min (Fig. 5D). The relative fluorescence intensity after 2 min in the near ROI was significantly higher than that at 0 min. In contrast, the relative fluorescent intensity in the nucleus ROI and far ROI showed no significant change compared to that at 0 min. Both the increase in PH-Akt-GFP and the decrease in GFP-PTEN occurred in the near ROI, suggesting that the translocation of PTEN may contribute to the directional accumulation of PIP3 before directed migration in monolayers. To investigate the effect of PTEN nuclear localization on PIP3 accumulation, we examined the intracellular translocation of PH-Akt-GFP after neighbor cell wounding with addition of Importazole (Fig. 6). Figure 6A and B show PH-Akt-GFP fluorescence images of Importazole-treated cells, with a slight accumulation at the cell edge near the wounded cells; the relative fluorescence intensity of PH-Akt-GFP increased in the near ROI and reached a maximum value of 1.13 ± 0.02 at 10 min (Fig. 6C). Although there was an increase in relative fluorescence intensity in the near ROI, the Importazole treatment group showed significantly lower intensity than the untreated group (Fig. 6D). These results indicate that inhibition of PTEN nuclear translocation inhibits PIP3 accumulation, suggesting that the nuclear localization of PTEN may play a role in regulating PIP3 accumulation in cell monolayers.

Figure 5.
figure 5

The dynamics of PH-Akt-GFP before directed migration in monolayers. (A) Representative fluorescence images of PH-Akt-GFP before directed migration. (B) Images of PH-Akt-GFP translocation obtained by subtracting the original image in (A) captured at t = 0 min and setting the initial value as 0. The intensity of PH-Akt-GFP is shown using a pseudo color scale. Greater than 0 are considered an increase compared with the initial value. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (C) The region of near, far, and nucleus in the cells. (D) Relative fluorescent intensity of PH-Akt-GFP in a time series. N = 14, n = 32. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test).

Full size image
Figure 6.
figure 6

The dynamics of PH-Akt-GFP after neighbor cell wounding under Importazole treatment. (A) Representative fluorescence images of PH-Akt-GFP during treatment with Importazole (40 µmol L−1) before directed migration. (B) Images of PH-Akt-GFP translocation obtained by subtracting the original image in (A) captured at t = 0 min and setting the initial value as 0. The intensity of PH-Akt-GFP is shown using a pseudo color scale. Greater than 0 are considered an increase compared with the initial value. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (C) Relative fluorescent intensity of PH-Akt-GFP under Importazole treatment. N = 8, n = 14. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test). (D) Comparison of PH-Akt-GFP relative fluorescence intensity in the near ROI with Fig. 5D. N = 8, n = 14. Mean ± standard error of the mean. *P < 0.05 vs. untreated (two-way ANOVA with post hoc Sidak’s test).

Full size image

PTEN and PIP3 do not translocate after wounding when cells are separated

Neighbor cell wounding is accompanied by leakage of cellular content and the disruption of cell–cell adhesion. To determine what induces the response of PTEN and PIP3, we wounded cells near, but without direct cell–cell adhesion, to the target cell. Figure 7 shows the dynamics of PTEN and PIP3 after wounding when the cells were separated. Figure 7B and D represent the fluorescent images using the GFP-PTEN and PH-Akt-GFP constructs, respectively. PTEN did not translocate to the nucleus and PIP3 did not accumulate at the cell edge near the wounded cells. Using the same definition of ROIs as in Fig. 2D, the relative fluorescence intensity of GFP-PTEN (Fig. 7E) or PH-Akt-GFP (Fig. 7F) was not significantly different in any of the ROIs within 10 min. Even if nearby cells were wounded, there was no translocation of PTEN or PIP3 in the absence of intercellular adhesion. This suggests that it is not the leakage of cell contents, but the disruption of cell–cell adhesion that induces the localized distribution of PTEN and PIP3.

Figure 7.
figure 7

The dynamics of GFP-PTEN and PH-Akt-GFP after wounding when cells are separated. (A) Images of cells before wounding. (B) Representative fluorescence images of GFP-PTEN. (C) Images of cells before wounding. (D) Representative fluorescent images of PH-Akt-GFP. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (E) Relative fluorescence intensity of GFP-PTEN in a time series. N = 6, n = 9. (F) Relative fluorescence intensity of PH-Akt-GFP in a time series. N = 6, n = 10. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test).

Full size image

Actin cytoskeletal affects PTEN translocation

The actin cytoskeleton plays an important role in cell–cell adhesion and is responsible for generating the force that holds cells together (Vasioukhin et al. 2000). Cytochalasin D destabilizes and displaces actin filaments, causing them to detach from their normal attachment points, which results in the disruption of the actin cytoskeleton (Wakatsuki 2001). To determine the effect of the actin cytoskeleton on PTEN dynamics, we used cytochalasin D as an actin inhibitor and examined PTEN localization after neighbor cell wounding in monolayers. Figure 8A and B show the fluorescence images of cytochalasin D–treated cells. Using the same definition for the ROIs as Fig. 2D, we examined the relative fluorescence intensity in the near, far, and nucleus ROI (Fig. 8C). Except for a very slight decrease in the near ROI at 1–2 min after wounding, the relative fluorescence intensity of GFP-PTEN showed no significant difference in all ROIs compared with that at 0 min. The relative fluorescent intensity of GFP-PTEN was compared with the untreated group for the nucleus ROI (Fig. 8D). There was a significant difference after 5 min between the cytochalasin D–treated group and the untreated group. These results suggest that the inhibition of actin suppresses the nuclear translocation of PTEN in monolayers.

Figure 8.
figure 8

The effect of the actin inhibitor cytochalasin D. (A) Representative fluorescence images of GFP-PTEN under treatment with cytochalasin D (2.5 µmol L−1) before directed migration. (B) Images of GFP-PTEN translocation obtained by subtracting the original image in (A) captured at t = 0 min and setting the initial value as 0. The intensity of GFP-PTEN is shown using a pseudo color scale. Greater than 0 are considered an increase compared with the initial value. Cell boundaries and nucleus boundaries at 0 min are indicated by white line. “ × ” indicates the wounded cell. White scale bar: 10 μm. (C) Relative fluorescence intensity of GFP-PTEN in time series. N = 6, n = 12. Mean ± standard error of the mean. *P < 0.05 vs. 0 min (two-way ANOVA with post hoc Dunnett’s test). (D) Comparison of GFP-PTEN relative fluorescence intensity in the nucleus with Fig. 2E. N = 6, n = 12. Mean ± standard error of the mean. *P < 0.05 vs. untreated (two-way ANOVA with post hoc Sidak’s test).

Full size image

Discussion

As opposed to directed migration in single cells, this study represents the first report demonstrating that PTEN translocates to the nucleus before directed migration begins in cell monolayers, without accumulating at the rear of the cell. In contrast, PIP3, a substrate for PTEN, accumulates in the periphery closest to the wounded cells. Furthermore, we found that Importazole treatment inhibits PTEN nuclear translocation and directs cell migration following neighbor cell wounding. Moreover, cell–cell adhesion is essential for the translocation of PTEN and PIP3; specifically, PTEN and PIP3 exhibited no response after wounding when the cells were separated.

Directed migration in single cells correlates with actin polymerization and disassembly on the sides of the cell (Stock and Pauli 2021). There are no cell–cell adhesions in single cells and the response of PTEN occurs at an early stage before directed migration begins. However, cells in a monolayer are closely packed and cell–cell adhesion links adjacent cells together, which provides mechanical support and facilitates communication (Mayor and Etienne-Manneville 2016). Cell–cell adhesion is a significant factor that influences directed migration (Saraswathibhatla et al. 2020). Therefore, we believe that the different phenomenon observed for PTEN in the present study compared with single cells is the result of cell–cell adhesion.

In monolayers, PTEN translocates to the nucleus after neighbor cell wounding (Fig. 2) and this nuclear translocation is suppressed by an actin inhibitor (Fig. 8). We previously reported that wounding of neighboring cells leads directly to the unloading of intercellular stress and the disruption of intercellular adhesion results in a larger strain field of the substrate in areas adjacent to wounded cells (Sera et al. 2019). In addition, mechanical forces generated by the dynamic rearrangements of cell–cell adhesion and the cytoskeleton cause changes in cell shape and motility (Jamora and Fuchs 2002). In the present study, neighbor cell wounding was accompanied by the disruption of cell–cell adhesion. When the actin cytoskeleton was inhibited with cytochalasin D, cell–cell adhesion and the mechanical properties of the cells were negatively affected and the nuclear translocation of PTEN was significantly inhibited. In contrast, there was no translocation of PTEN when the cells were subconfluent (Fig. 7), suggesting a dependence upon mechanical force sensing for cell–cell adhesion. These results indicate the nuclear translocation of PTEN requires mechanical stimulation and the intact actin cytoskeleton, which regulates the spatial distribution of PTEN.

Unlike its distribution in single cells, PTEN does not accumulate at the rear of the cell before directed migration begins in monolayers (Fig. 2A, B). PTEN can negatively regulate cell interactions with the extracellular matrix (ECM) by decreasing the formation of actin filaments and focal adhesions (Tamura et al. 1998). In addition, cell-ECM adhesion or cytoskeletal contractility can affect cell surface scaffolds, which regulates single-cell and collective-cell migration (Friedl and Alexander 2011). Impaired intercellular adhesion and loss of adherens junctions result in the detachment of cells (Young 2003). Therefore, cells may need to stably adhere to the ECM and surrounding cells in monolayers, unlike the situation in single-cell migration. PTEN does not accumulate at the rear of the cell to mediate rear contractility, guaranteeing high mechanical stability in monolayers.

An inhibitor of importin-β significantly suppressed the PTEN nuclear translocation and directed migration following wounding (Figs. 3 and 4). The precise mechanism governing the nuclear translocation of PTEN remains unclear; however, it may be associated with the following mechanisms. PTEN nuclear accumulation is dependent on Ran-GTPase activity, suggesting the essential function of importin-β in this transport process (Gil et al. 2006). The importin-β family member Importin-11 (IPO11) transports PTEN into the nucleus and protects it from cytoplasmic proteins that cause PTEN degradation (Chen et al. 2017). In addition, cells can sense mechanical forces through a mechanosensitive complex and generate reactive oxygen species (ROS) (Chatterjee and Fisher 2014). Mechanical forces could induce the direct opening of hemichannels. Adenosine triphosphate (ATP) is released outside of the cell from hemichannels (Takada et al. 2014). The production of ROS and cellular ATP depletion results in increased nuclear PTEN (Du et al. 2017). Moreover, ubiquitination is known to control PTEN’s nuclear localization (Baker 2007). Nedd4-1, an E3 ubiquitin-protein ligase, mediates PTEN monoubiquitination, regulating the nuclear import of PTEN (Trotman et al. 2007). In this study, after neighbor cell wounding, Importin-β and several other mechanisms could be associated with the entry of PTEN into the nucleus. Further studies are needed to determine whether importin-β directly interacts with NLS-like sequences and transports PTEN into the nucleus. Inhibition of PTEN nuclear localization inhibits directed migration after neighbor cell wounding, suggesting that, besides cell proliferation and cell stability, nuclear-localized PTEN may also be involved in directed cell migration in confluent cell monolayers.

PIP3 accumulated at the cell edge near the wounded cells (Fig. 5); Importazole suppressed the nuclear localization of PTEN and the directional accumulation of PIP3 (Fig. 6). The mutually inhibitory relationship between PTEN and PIP3 regulates their localized levels, resulting in their spatial separation (Matsuoka and Ueda 2018). Additionally, PI3K signaling and PIP3 accumulation at the front edge of cells are important for promoting pseudopodia formation and directed cell migration (Chung et al. 2001; Blitsman et al. 2024). When importin-β is inhibited, cytoplasmic PTEN is not transported into the nucleus. Consequently, in the region near the wounded cells, undiminished cytoplasmic PTEN converts PIP3 to PIP2, reducing PIP3 accumulation and thereby inhibiting cell elongation after wounding. These results suggest that the nuclear translocation of PTEN may regulate PIP3 accumulation by modulating cytoplasmic PTEN levels, thereby affecting cell migration in cell monolayers.

On the other hand, although the mechanisms of how signal transduction regulates PIP3 and PTEN are unclear, Protein kinase Cα (PKCα) and myristoylated alanine-rich C kinase substrate (MARCKS) may play a role. PKCα accumulates at the edges of the cells adjacent to wounded cells at an early stage (Sasaki et al. 2021). MARCKS, a major PKC substrate, binds to and modulates free PIP2 levels, which in turn, affects the recruitment of PI3K to the membrane (Ziemba et al. 2016). The displacement of MARCKS from the membrane results in increased PIP3 levels (Ziemba and Falke 2018). Therefore, after neighbor cell wounding, active PKCα accumulates at the edge near wounded cells, causing the dissociation of MARCKS from the membrane, and then, active PI3K phosphorylates PIP2 to form PIP3, which accumulates while PTEN decreases from the cell edge near wounded cells. The PKCα-MARCKS-PIP3 signaling transduction may play an important role in mediating the spatial distribution of PTEN.

Conclusions

Our study presents the first evidence of PTEN nuclear translocation before directed migration begins in cell monolayers, a process different from that in single cells. This translocation is critically dependent on cell–cell adhesion and mediated by importin-β and actin cytoskeleton signaling pathways, offering new insights into the regulatory mechanisms underlying directed migration in cell monolayers. Our findings highlight the pivotal role of PTEN translocation in mechanotransduction mechanisms within confluent monolayers. The intricate interplay between cell–cell adhesion, signal transduction, and PTEN’s spatiotemporal distribution provides potential avenues for investigating tissue dynamics and cellular responses to wounding.