Cellular morphological changes and cytoskeletal actin disorganization in the absence of Pten.
Throughout the years of studying PTEN and culturing adherent mammalian cells with and without PTEN, we have noticed a morphological feature associated with PTEN loss, i.e., cell rounding and detaching from culture plates or surrounding cells. To investigate and document morphological alterations following PTEN depletion, we traced single cell cytoskeletal architecture by immunofluorescence of F-actin using an isogenic pair of mouse embryonic fibroblast cells derived from PtenWT and PtenKO mice.
The cell shape is visibly rounder in cells depleted of PTEN (Fig. 1a). We measured the cell length and width and found decreased cell length but increased cell width in PtenKO cells. The length-to-width (L:W) ratio in PtenKO cells is decreased significantly (Fig. 1b). There is also a significantly higher number of round cells in PtenKO compared to PtenWT cells (Fig. 1c). We excluded mitotic cells that would more likely be round using the DAPI stain.
Additionally, we report higher frequency of poorly aligning actin filaments pointing in different directions in cells depleted of PTEN. This is contrasted with PtenWT cells, which we found have more well-aligned F-actin bundles (Fig. 1d). Furthermore, there is more F-actin accumulated at the cell cortex in PtenKO cells compared to PtenWT cells (Fig. 1a and 1e). There appears to be a correlation between loss of Pten and enhanced enrichment of F-actin along the plasma membrane. These results suggest that PTEN is required for proper formation and distribution of actin cytoskeleton to support normal cell morphology.
Higher levels of pMoe and increased formation of pMoe aggregates in PtenKO cells. F-actin is crosslinked to the plasma membrane through the adapter protein Moe23. PTEN has been reported to regulate Moe during mitosis24. However, the interaction between PTEN and Moe during interphase is not well studied. The crosslinking activity of Moe is dependent on its phosphorylation of a conserved threonine residue T558 that induces a conformational change to an “active” state30. We thus examined the activation status of Moe in PtenWT and PtenKO MEFs by focusing on pMoe levels.
The level of pMoe is significantly different between PtenWT and PtenKO MEFs cells (Fig. 2a). We used immunofluorescence to assess quantity of pMoe and F-actin expression. Figure 2b shows both increased F-actin membrane enrichment and increased overall pMoe levels to the naked eye in cells depleted of Pten. An overlay of the two targets shows visibly different morphological phenotypes in PtenKO cells as compared to PtenWT cells. However, we were unable to document a visible difference in Moe after Pten loss (Fig. 2c).
We measured the quantity of pMoe levels and found significantly higher intensity of pMoe in PtenKO cells, which agrees with our visual assessment (Fig. 2d). In addition to higher pMoe intensity, there is a higher number of cells with prominent, compact, and punctate pattern of pMoe deposition in PtenKO cells (Fig. 2e). We called these “aggregates,” or highly concentrated collections of pMoe which was also visually deciphered (Fig. 2b). These results support that Pten deletion may alter F-actin-binding protein Moe, manifested both biochemically (i.e., increased pMoe levels) and morphologically (pMoe aggregates). Thus, it is likely that Moe is a signaling target of PTEN.
Higher pMoe expression found near the nucleus as opposed to membrane in PtenKO cells. In addition to altered activity of Moe, it is possible that Pten loss changes the localization of pMoe in the cell. We performed cellular fractionation experiments to measure pMoe levels in the cytoplasm and membrane (Fig. 3a). In PtenWT cells, a non-uniform pattern of Moe and pMoe was found. More pMoe is present at the membrane. However, in PtenKO cells, the opposite trend was found. A lower level of pMoe was found at the membrane compared to the cytoplasm. No significant difference between Moe levels near the membrane and cytoplasm was found between PtenWT and PtenKO cells (Fig. 3a).
We further categorized the distribution of Moe and pMoe into two specific regions: near the plasma membrane (juxtamembrane region) and near the nucleus (perinuclear region, Fig. 3b). Lower levels of pMoe in the juxtamembrane region and higher levels in the perinuclear region were found in PtenKO cells (Fig. 3c and 3d). We further partitioned the cytoplasm to investigate the cellular distribution of pMoe and Moe in PtenKO cells (Fig. 3e). The two areas are the peripheral area (P) and the central area (C). We found significantly higher levels of pMoe in the central area in PtenKO cells compared to PtenWT cells (Fig. 3f). This corresponds with an increased pMoe C:P ratio in PtenKO cells (Fig. 3g).
Additionally, we measured the levels F-actin in the central and peripheral areas as well to determine if there might be correlation with the significant changes in pMoe localization we found between PtenKO and PtenWT cells. Indeed, we found a significant decrease in the F-actin C:P ratio in PtenKO cells (Fig. 3h). In other words, more actin filaments accumulate in the peripheral area after Pten loss, which is in agreement with greater F-actin enrichment at the membrane (Fig. 1e). The inverse redistribution of pMoe and F-actin suggests that their cross-linking may be hampered in the absence of Pten, which may lead to the impairment of cell cortical architecture.
Reversible F-actin cortical enrichment and greater cell rounding in PtenKO cells. Given previous findings of increased actin filaments at the cortex in PtenKO cells, we hypothesize that loss of Pten may affect actin polymerization. To test this hypothesis, we treated PtenKO cells with cytochalasin D, an inhibitor of actin polymerization, to determine whether it can reverse the morphological and F-actin distribution aberrations in PtenKO cells.
Both PtenWT and PtenKO cells were treated with cytochalasin D for assessment of F-actin expression at the membrane and cytoplasm. We found visible differences in both cell shape and F-actin signals after treatment between PtenKO and PtenWT cells (Fig. 4a). Quantitative imaging analysis shows a nearly complete reverse of F-actin membrane enrichment in PtenKO cells treated with cytochalasin D (Fig. 4b). There is also a significant increase in the F-actin C:P ratio after cytochalasin D treatment (Fig. 4c), indicating that inhibition of actin polymerization in PtenKO cells displaces F-actin from the membrane and that more F-actin becomes present in the cytoplasm.
Despite the correction of aberrant cortical enrichment and subcellular distribution of F-actin by inhibition of actin polymerization (Fig. 4a-4c), the organization of F-actin fibers (i.e., how well fibers aligned with each other and homogeneity of direction) remains poorly aligned in PtenKO cells even after cytochalasin D treatment (Fig. 4d). Interestingly, there is a significantly lower number of round cells in PtenKO treated with cytochalasin D compared to PtenKO without treatment (Fig. 4e). These data suggest that impairment of actin filament alignment per se may not be responsible for the cell shape deregulation in PtenKO cells.
Reversible Moe phosphorylation and subcellular localization in PtenKO cells in response to inhibition of actin polymerization. Finally, we strived to determine if correcting aberrant actin polymerization can ameliorate the abnormal pMoe mislocalization in PtenKO cells. We used cytochalasin D to inhibit actin polymerization and compared pMoe levels and distribution in PtenKO and PtenWT cells. First, we found a time-dependent reduction of pMoe in PtenKO cells following cytochalasin D (Fig. 5a). We also found visibly decrease size and intensity of compact and punctate pMoe deposition after cytochalasin D in both PtenKO cells (Fig. 5b). There is a significant reduction of measured pMoe levels in PtenKO cells after cytochalasin D, which is in agreement with our visual impression (Fig. 5c).
The cellular localization of pMoe was also categorized into central (C), peripheral (P), perinuclear, and juxtamembrane. We found significantly decreased levels of pMoe in the central area (lower C:P ratio), decreased levels in the perinuclear region, and increased levels in the juxtamembrane region after cytochalasin D in PtenKO cells (Fig. 5d-5f). Following inhibition of actin polymerization, the reversed elevation of pMoe levels and membrane dissociation of Moe (Fig. 5) co-occur with the reverse of F-actin cortical enrichment and cell rounding (Fig. 4). These data collectively suggest that Moe deregulation and corresponding aberrant cell morphology may be associated with dysfunction of cortical actin cytoskeleton in cells lacking Pten.