A transcription co-factor integrates cell adhesion and motility with the p53 response

Abstract

Despite its obvious importance in tumorigenesis, little information is available on the mechanisms that integrate cell motility and adhesion with nuclear events. JMY is a transcription co-factor that regulates the p53 response. In addition, JMY contains a series of WH2 domains that facilitate in vitro actin nucleation. We show here that the ability of JMY to influence cell motility is dependent, in part, on its control of cadherin expression as well as the WH2 domains. In DNA damage conditions JMY undergoes nuclear accumulation, which drives the p53 transcription response but reduces its influence on cell motility. Consequently, the role of JMY in actin nucleation is less in damaged cells, although the WH2 domains remain functional in the nucleus where they impact on p53 activity. Together, these findings demonstrate a pathway that links the cytoskeleton with the p53 response, and further suggest that the ability of JMY to regulate actin and cadherin is instrumental in coordinating cell motility with the p53 response.

It is widely recognized that cell motility plays a fundamental role in tumorigenesis by allowing tumor cells to invade and colonize surrounding, as well as distant, healthy tissue (1). JMY (junction-mediating and regulatory protein) is a transcription co-factor, originally identified as a p300-binding protein, that augments the p53 response (2). Upon DNA damage JMY forms a complex with Strap and p300, which recruits PRMT5 into a co-activator complex that drives the p53 response (2–5). By regulating p53-dependent transcription and the functional outcome of the p53 response, JMY takes on a central role during the DNA damage response (2).

The Mdm2 oncoprotein is a key regulator of the p53 response, a function it achieves by targeting p53 at multiple levels, including transcription, protein stability, and subcellular localization (6). Recent studies identified JMY as a target through which Mdm2 regulates p53 activity, where Mdm2 prevents JMY from activating p53, in part by targeting JMY for degradation (7). Upon DNA damage JMY is released from Mdm2, allowing it to contribute to the p53 response (7).

The actin cytoskeleton provides an essential cellular mechanism linked to many physiological activities like cell motility and invasion (8). WH2 [Wiskott Aldrich Syndrome protein (WASp)-homology 2] domains are actin-binding motifs found in WASp family members, including neuronal (N)-WASp and WAVE/SCAR isoforms (9, 10), as well as Spir (11). WASp family members activate the Arp2/3 (actin-related protein 2/3) complex, a major regulator of actin filament assembly (12). Arp2/3 does not initiate actin nucleation unless it is activated, in the presence of ATP, by a nucleation promoting protein such as WASp that, via its WH2 domains, facilitates the assembly of actin monomers into newly formed actin filaments (12).

Cadherins, such as E- and N-cadherin, are adherens junction proteins responsible for cell-cell adhesion (13). Cell-cell adhesion is intimately connected to cell migration, invasion, and tumor progression; loss of E-cadherin expression correlates with tumor invasion and metastasis (14). At cell-cell junctions cadherin proteins form transmembrane contacts with cadherin molecules in neighboring cells, while the cytoplasmic domain associates with cytosolic catenins (α, β, and p120). This complex is functionally linked to the actin cytoskeleton via α-catenin (13).

The actin cytoskeleton not only maintains cellular architecture but also signaling pathways that regulate cell motility and growth, as well as survival and apoptosis. Actin cytoskeletal reorganization can influence apoptosis and during the cellular response to DNA damage, the actin cytoskeleton is reorganized (15–18). In this respect, it is known that p53 affects cell migration, as loss of p53 promotes cell migration and invasion (19). In addition, nuclear actin and actin-related proteins (Arps) play important roles in several nuclear processes including chromatin remodeling and transcription (20). Actin is required for efficient transcription by RNA polymerases I, II, and III (21–23), and actin and Arps are also associated with SWI/SNF-like chromatin remodeling complexes (24). In this respect, we have described a biochemical role for the p53 co-factor JMY in regulating actin nucleation (25). JMY contains three carboxyl-terminal WH2 domains which nucleate actin in vitro and JMY can direct the assembly of actin filaments in vitro in an Arp2/3-independent fashion.

Here we report evidence that links JMY with a role in cell adhesion. The ability of JMY to influence cell motility is dependent, in part, on its control of cadherin expression. Under DNA damage conditions JMY becomes nuclear, which drives the p53 transcription response while reducing its influence on cell motility. The role of JMY in actin nucleation is less in damaged cells, although the WH2 domains are functional in the nucleus where they impact on p53 activity. Together these findings demonstrate a pathway that links the cytoskeleton with the p53 response, and further suggest that the ability of JMY to regulate actin and cadherin is instrumental in coordinating cell motility with the p53 response.

Results

JMY Regulates Cadherin Levels.

JMY contains three WH2 domains in its C-terminal region, and a central and acidic domain with similarity to other activators of Arp2/3 such as WASp and Scar [(25) and Fig. 1A]. Depletion of JMY resulted in altered cell morphology (Fig. 1 B and C), suggesting that there might be alterations in proteins involved in cellular adhesion. Since cadherins play an important role in cell-cell adhesion and expression of E-cadherin can suppress tumor cell motility (26, 27), we investigated the level of E-cadherin in JMY depleted cells. A marked upregulation of E-cadherin in JMY siRNA treated MCF-7 cells was apparent (Fig. 1 D and E). As E-cadherin is known to regulate cell motility and invasion (28), we reasoned that the upregulation of E-cadherin might be involved in the ability of JMY to influence cell motility. Indeed, simultaneous depletion of both JMY and E-cadherin rescued the decrease in cell motility seen upon JMY depletion alone (Fig. 1F).

Fig. 1.

JMY depletion influences cell morphology. (A) Schematic representation of JMY. WH2 = WASp Homology 2, P = proline. C-terminal region of JMY is highly similar to other WH2 domain containing proteins. C = Central, A = Acidic. Green shaded box indicates previously described p300 and Strap interacting regions (2, 3). Red shading indicates WH2 domains, yellow shading central and magenta shading acidic regions. (B) HEK 293 (i) or MCF-7 (ii) cells were treated with JMY or control (con) siRNA for 72 h. JMY was detected with anti-JMY antibody L-16 and actin was used as a loading control. (C) MCF-7 and HEK 293 (HEK) cells were treated with control or JMY siRNA as above and photographs were taken 72 h later. (D) MCF-7 cells were treated with control (con) or JMY siRNA for 72 h and harvested at a similar density. JMY was detected using anti-JMY antibody L-16, E-cadherin was detected using anti-E-cadherin antibody and actin was used as a loading control. (E) MCF-7 cells were plated onto glass coverslips and treated as in (D) before fixing and processing for immunostaining. E-cadherin was detected using anti-E-cadherin antibody and DAPI was used to visualize nuclei. (Scale bar, 10 μm.) (F) MCF-7 cells were treated with control, JMY, E-cadherin, or E-cadherin and JMY siRNA for 72 h before performing scratch wound assays. Graph represents percentage wound closure after 24 h. JMY siRNA treatment resulted in a 33% decrease in wound closure rate. Values represent mean ± s.e.m, of a representative experiment, n = 7–12, *, P < 0.05, Student's t-test. Blots below show levels of JMY and E-cadherin knockdown. Con = control siRNA, E-cad = E-cadherin siRNA. Actin was used as a loading control. (G) U2OS cells were treated with control (con) or JMY siRNA for 72 h and harvested at a similar density. JMY was detected using anti-JMY antibody L-16 and N-cadherin with anti-N-cadherin antibody. Actin was used as a loading control. (H) U2OS cells were plated onto glass coverslips and treated as in (F) before fixing and processing for immunostaining. N-cadherin was detected using anti-N-cadherin antibody and DAPI was used to visualize nuclei. (Scale bar, 10 μm.) (I) U2OS cells were treated with control, JMY, N-cadherin or N-cadherin and JMY siRNA for 72 h before performing scratch wound assays. Graph represents percentage wound closure after 24 h. JMY siRNA treatment resulted in a 20% decrease in wound closure rate. Values represent mean ± s.e.m, of a representative experiment, n = 10, *, P < 0.05, Student's t-test. Blots below show levels of JMY and N-cadherin knockdown. Con = control siRNA, N-cad = N-cadherin siRNA. Actin was used as a loading control. (J i and ii) U2OS cells were treated with control (NTC) or JMY siRNA for 72 h before treating with vehicle (-) or 100 μM cyclohexamide (CHX) for the appropriate time point before harvesting. N-cadherin was detected using N-cadherin antibody and actin was used as a loading control. Graph (i) represents the percentage cadherin expression at the time points denoted (with the vehicle control set as 100%) after normalising to actin levels using Scion Image (Scion Corporation).

JMY also influences cell motility in the N-cadherin expressing U2OS and SAOS2 osteosarcoma cells (Fig. 1I and Fig. S1A) and cell morphology (Fig. S1B). N-cadherin levels were also found to be upregulated upon JMY depletion in U2OS and SAOS2 cells (Fig. 1 G and H and Fig. S1C). In osteosarcoma cells N-cadherin has been shown to be a negative regulator of invasion and metastasis (29) and depletion of both N-cadherin and JMY reversed the ability of JMY to inhibit cell motility (Fig. 1I). It was of further interest that depletion of E-cadherin resulted in enhanced JMY levels (Fig. 1F and Fig. S1 D and E), suggesting a reciprocal relationship between JMY and cadherin. Moreover, JMY depletion resulted in an increase in N-cadherin half-life in U2OS cells (Fig. 1J i and ii). After 6 h of cyclohexamide treatment N-cadherin levels were reduced by approximately 50% in control treated cells whereas, at this time point in JMY depleted cells, the N-cadherin level had undergone a 6% decrease (Fig. 1Ji). Taken together these results suggest that JMY regulates motility by influencing cadherin levels through altered cadherin stability/protein turnover.

JMY Actin Nucleation Influences Cell Motility.

When JMY was depleted from MCF-7 human breast cancer cells, there was a marked decrease in the ability of cells to migrate into newly formed wounds (Fig. 2A i and ii). Control siRNA treated cells had an average wound closure rate of 42% compared to an average closure rate of 13% in the JMY siRNA treated cells (Fig. 2Aiii). To substantiate these results, we studied cellular motility in inducible stable JMY shRNA cell lines, which allowed the conditional depletion of endogenous JMY (Fig. 2Bi). Similar decreased wound closure was apparent when JMY shRNA was induced in MCF-7 cells (Fig. 2B ii and iii). Conversely, in a stable cell line in which ectopic JMY was inducible, increased wound healing rates were observed (Fig. 2C), and ectopic JMY was also able to rescue the decreased motility seen upon JMY depletion (Fig. S1F). Importantly, the increased motility observed when ectopic JMY was induced was abolished when the WH2 domains were removed from JMY (Fig. 2C), confirming that the effect on cell motility depends on the WH2 domain containing region. Additional evidence that JMY plays a role in cell motility was obtained from studying the ability of cells to migrate through a Matrigel reconstituted basement membrane. In keeping with the fact that JMY influences cell motility, there was a marked decrease in the ability of cells to migrate into the Matrigel upon depletion of JMY (Fig. 2D). These results indicate that JMY WH2 domains are directly involved in regulating cell motility.

Fig. 2.

JMY influences cell motility through its WH2 domains. (Ai) MCF-7 cells were treated with control (con) or JMY siRNA. JMY was detected with anti-JMY antibody L-16 and GAPDH was used as a loading control. (ii) MCF-7 cells were treated as in (i) before scratch wounding. Photographs were taken immediately after wounding (time 0) and after 24 h. Graph in (iii) represents percentage wound closure after 24 h. Values represent mean ± s.e.m., *, P < 0.0002, n = 9, Student's t-test. (Bi) MCF-7 JMY shRNA stable cell line was treated with 1 μg/mL doxycycline (dox) for 72 h. JMY was detected using anti-JMY antibody L-16 and GAPDH was used as a loading control. (ii) MCF-7 JMY shRNA stable cell line was treated as in (i) before scratch wounding. Photographs were taken immediately after wounding (time 0) and after 24 h. Graph in (iii) represents percentage wound closure after 24 or 48 h as denoted. Values represent mean ± s.e.m., *, P < 0.02, n = 7, Student's t-test. (Ci) U2OS stable inducible JMY cells were treated with 1 μg/mL doxycycline for 24 h before wounding. JMY was detected with anti-JMY antibody M300. Actin was used as a loading control. Vec = vector control, WT = wild-type, ΔWH2 = JMYΔWH2. (ii) Cells treated as in (i) were scratch wounded and photographs were taken immediately after wounding (time 0) and after 24 h. Graph in (iii) represents percentage wound closure after 24 h. Values represent mean ± s.e.m., *, P < 0.005, **, P < 0.002, n = 6, Student's t-test. (Di) U2OS cells were treated with control (con) or JMY siRNA for 72 h. JMY was detected using anti-JMY antibody L-16 and GAPDH was used as a loading control. (ii) U2OS cells were treated as in (i) before replating 0.5 × 10⁵ cells into both control and Matrigel containing chambers. Cells were allowed to invade for 16 h before fixing and staining with propidium iodide to visualize nuclei. Results are representative of two independent experiments. Graph in (iii) represents percentage of cells invaded with either control (con) or JMY siRNA treatments. (E) U2OS cells plated onto coverslips were transfected with HA-tagged wild-type JMY or JMY derivatives as denoted. Before fixation cells were subjected to in situ actin incorporation assays with labeled G-actin for 5 min. Following actin incorporation cells were fixed and processed for immunostaining (without a further permeabilization step) and JMY was detected using anti-HA antibody HA11. DAPI was used to visualize nuclei. (Scale bar, 10 μm.)

To assess whether JMY influences actin incorporation in cells, in situ G-actin incorporation assays were used (30). Wild-type JMY and a JMY mutant derivative devoid of nuclear localization, JMYΔ468–558, could facilitate G-actin incorporation into the sites of localization (Fig. 2E), as well as enhance G-actin incorporation compared to surrounding cells that did not express the ectopic protein (Fig. 2E). JMY 1–504 did not stimulate G-actin incorporation and in fact appeared to cause a decrease in overall G-actin incorporation (Fig. 2E). Importantly, a JMY derivative lacking its WH2 domains (ΔWH2) failed to stimulate incorporation of G-actin, as did JMYΔ468–558ΔWH2 (Fig. 2E). Taken together, these results indicate that JMY can direct actin incorporation at sites of its intracellular localization.

Additionally, cytoplasmic JMY localized to actin-containing membrane ruffles and ectopic JMY could also induce the formation of actin structures (indicated by arrows, Fig. S2A). Moreover, JMYΔ468–558 exhibited a more prominent co-localization with actin (Fig. S2A). Importantly, removal of the WH2 domain region from JMY (ΔWH2) and JMYΔ468–558 (JMYΔ468–558ΔWH2) resulted in a loss of actin co-localization (Fig. S2A, summarized in Fig. S2D). In addition, actin was detected in JMY immunoprecipitates (Fig. S2 B and C), which is consistent with a physical interaction between JMY and actin.

DNA Damage Augments the Nuclear Role of JMY.

JMY is a damage responsive protein (2, 7). Biochemical fractionation of cells demonstrated that upon DNA damage endogenous JMY undergoes nuclear accumulation (Fig. 3A and Fig. S3 A and B). Similarly, by immunostaining JMY was shown to acquire a nuclear location in DNA damaged cells (Fig. 3B), and ectopic JMY behaved in a similar fashion (Fig. 3C and Fig. S3 C and D). Because of this, we reasoned that the contribution that JMY makes toward cell adhesion and motility might be diminished under DNA damage conditions. To explore this idea, scratch wound assays were performed using a low dose of UV irradiation (1 J UV/m²), sufficient to activate the DNA damage response (31). In DNA damaged cells, ectopic JMY was not able to augment cell motility (Fig. 3D), contrasting with the increased cell motility observed in non-DNA damage conditions (Fig. 3D). Moreover, there was much less effect on cell motility when JMY was depleted during the DNA damage response, contrasting with the effect of JMY depletion in untreated cells (Fig. 3E).

Fig. 3.

DNA damage regulates JMY activity. (A) NIH 3T3 Ras (i) or MCF-7 (ii) cells were subjected to fractionation to obtain nuclear and cytoplasmic extracts with (+) or without (-) UV treatment (I, 50J/m², 16 h ii, 30J/m², 8 h). Endogenous JMY was detected with anti-JMY antibody M300 (i) or anti-JMY antibody L-16 (ii). Actin was used to show protein loading. The nucleolar protein nucleophosmin (NPM) is localized exclusively to the nucleus. C = cytoplasmic, n = nuclear. (B) Endogenous JMY immunostaining in wild-type (WT) MEF and NIH 3T3 Ras cells using anti-JMY antibody M300. Cells were left untreated (-) or treated (+) with UV (50J/m², 16 h). (Scale bar, 10 μm.) (C) U2OS cells were transfected with HA-tagged JMY and treated with UV (50J/m²) for the times denoted. JMY was detected using anti-HA antibody HA11. (Scale bar, 10 μm.) (Di) U2OS cells stably expressing inducible JMY were treated with doxycycline (dox; 1 μg/mL) for 48 h to induce expression of JMY before scratch-wounding. Cells were either left untreated (-) or treated with UV (1 J/m², 16 h). JMY was detected with anti-JMY antibody L-16 and PCNA was used as a loading control. (ii) Graph represents percentage wound closure with and without JMY induction. Values represent mean ± s.e.m, *, P < 0.002, n = 6, Student's t-test. (Ei) U2OS cells treated with either control or JMY siRNA (48 h) were treated with or without 1 J/m² UV at the time of scratch-wounding. JMY was detected using anti-JMY antibody L-16 and actin was used as a loading control. (ii) Graph represents the percentage wound closure after 24 h. Values represent mean ± s.e.m., **, P < 0.005, n = 8–9, Student's t-test. (Fi) U2OS cells stably expressing wild-type (WT) JMY or JMY-NLS (NLS) were wounded with or without 1 J UV/m² treatment. JMY was detected with anti-HA antibody HA11 and actin was used as a loading control. (ii) Graph represents percentage wound closure. Values represent mean ± s.e.m., *, P < 0.05, n = 9, Student's t-test. iii U2OS cells stably expressing JMY or derivatives were plated onto glass coverslips. JMY was detected using anti-HA antibody HA11 and DAPI was used to visualize nuclei.

We hypothesized that the nuclear localization of JMY prevents or hinders its influence on cell motility, and therefore assessed the contribution of wild-type JMY and a JMY derivative targeted to the nucleus (with an artificial nuclear localization signal, JMY-NLS; Fig. S3E) to cell motility. Cells expressing JMY-NLS displayed reduced motility compared to wild-type JMY expressing cells (Fig. 3F). In contrast, during the DNA damage response JMY-NLS displayed a wound closure rate similar to that of wild-type JMY (Fig. 3F). These results indicate that the nuclear location of JMY in cells undergoing the DNA damage response coincides with a decreased influence on cell motility.

Nuclear JMY Is Required for p53 Transcriptional Activity.

To investigate the role of JMY in p53 transcriptional activity, we examined the ability of p53 to activate its well-described target gene, Bax, under conditions when JMY was depleted. There was a significant reduction in p53 activity (Fig. 4A), supporting the fact that JMY plays a functional role in p53 transcriptional activity. Ectopic JMY enhanced p53 transcriptional activity (2) (Fig. 4B), and targeting JMY to the nucleus (JMY-NLS) (Fig. S3E), increased further the ability of JMY to enhance p53 transcriptional activity (Fig. 4B). Nuclear JMY therefore augments p53 activity during the DNA damage response.

Fig. 4.

JMY WH2 domains can function in the nucleus. (A) U2OS cells were treated with control (con) or JMY siRNA for 48 h before transfecting 200 ng Bax-luciferase and 100 ng pCMV-p53. Cells were harvested after 24 h and reporter assays performed. β-galactosidase activity was used to normalize for transfection efficiency. p53 inputs are shown below. Results represent mean ± s.e.m., n = 3, *, P < 0.001, Student's t-test. (B) U2OS cells were transfected with Bax-luciferase with or without p53 (100 ng) or JMY (200 ng) expression constructs as denoted. β-galactosidase activity was used to normalize for transfection efficiency. Results represent mean ± s.e.m., *, P < 0.05, two independent experiments. JMY and p53 inputs are shown below. (C) U2OS cells plated onto coverslips were transfected with JMY-NLS or JMYW981A-NLS. Before fixation cells were subjected to in situ actin incorporation assays with labeled G-actin for 5 min. Following actin incorporation cells were fixed and processed for immunostaining (without a further permeabilization step) and JMY was detected using anti-HA antibody HA11. DAPI was used to visualize nuclei. (Scale bar, 10 μm.) (Di) U2OS cells were transfected with Bax-luciferase with or without p53 (50 ng) or JMY expression constructs (200 ng) as denoted. Cells were treated with or without latrunculin A (1.5 h, 1 μM) before harvesting for luciferase assay. β-galactosidase activity was used to normalize for transfection efficiency. Vec = vector, NLS = JMY-NLS. Results represent mean ± s.e.m., n = 3. (ii) Blots represent JMY and p53 input levels either untreated or with latrunculin A as denoted. (E) U2OS cells were transfected with Bax-luciferase with or without p53 (50 ng) or JMY expression constructs (200 or 500 ng) as denoted. Cells were harvested after 24 h and reporter assays performed. β-galactosidase activity was used to normalize for transfection efficiency. Results represent mean ± s.e.m., n = 3. JMY and p53 inputs are shown below. (F) U2OS cells stably expressing inducible JMY or JMYW981A (W981A) were treated with doxycycline (dox; 1 μg/mL) for 48 h to induce expression of JMY before scratch-wounding. Graph represents percentage wound closure with and without JMY induction. Values represent mean ± s.e.m, *, P < 0.05, n = 10, Student's t-test. (G) Model depicts potential mechanism involved in JMY regulation. In non-stressed cells, cytoplasmic JMY influences actin cytoskeletal events and cadherin expression. During the DNA damage response JMY becomes nuclear where it functions as a p53 transcription co-factor, reducing its contribution to cell motility.

Given the role of JMY as a p53 transcription co-factor, we reasoned that the WH2 domains might exhibit nuclear activity. To test this idea, we assessed if targeting JMY to the nucleus resulted in nuclear G-actin incorporation. Nuclear JMY resulted in the appearance of actin-containing structures (Fig. 4C), suggesting that the WH2 domain region of JMY is active in the nucleus. To explore the role of actin in JMY nuclear activity, p53 transcription was monitored in the presence of latrunculin A, an inhibitor of actin polymerization (32). While latruculin A had only marginal effects on p53 activity alone, it significantly reduced the enhanced p53 activity due to the co-activator role of JMY-NLS (Fig. 4D).

Moreover, because JMY can nucleate actin in the absence of Arp2/3 in vitro (25), we investigated the requirement for Arp2/3 in nuclear JMY function. A highly conserved tryptophan in the WCA (WH2 domain, central-acidic) region of WH2 containing proteins such as JMY is important for Arp2/3-mediated, but not intrinsic, actin assembly (25). To explore the role of Arp2/3 in JMY activity, we mutated the conserved tryptophan (W981A) in JMY-NLS and assessed nuclear activity. JMY was still able to incorporate nuclear actin (Fig. 4C) and W981A was still able to enhance p53 transcription (Fig. 4E). This contrasted with JMY cytoplasmic function where mutation of the tryptophan resulted in decreased ability of JMY to enhance cell motility (Fig. 4F). These results are consistent with a role for the WH2 domains in both the cytoplasmic and nuclear roles of JMY, although the mechanistic details relating to the role of Arp2/3 are distinct. Thus JMY is uniquely positioned as an actin nucleating protein that is able to function as both an effector of cell adhesion and a regulator of nuclear events during the DNA damage response, dependent on its subcellular localization.

Discussion

JMY: A WH2 Domain-Containing Protein.

We describe here a link between JMY and cell adhesion and motility. JMY, by influencing cadherin levels, is able to alter cell adhesion which in turn, influences cell motility. The presence of the three C-terminal WH2 domains is required for JMY to enhance cell motility. WH2 domain-containing proteins promote actin nucleation or elongation (9). JMY, in common with other WH2 domain containing proteins, such as N-WASp and WAVE/SCAR, harbors a C-terminal region comprised of WH2 domains and a central-acidic region. This region binds monomeric actin, through the WH2 domains, and activates the Arp2/3 complex via the central-acidic region, to induce actin filament formation (33, 34). JMY nucleates actin in vitro (25) and induces actin filament formation in vivo, which is dependent on the presence of its WH2 domains. Significantly, JMY can nucleate actin in the absence of Arp2/3 (25).

Cadherins mediate cell-cell adhesion and as such play a vital role in normal tissue, as well as malignant disease. E-cadherin is required for the formation and maintenance of epithelia and, during tumor progression, loss of function of E-cadherin is associated with increased invasion and metastasis (35). Cadherin based complexes are linked in a dynamic fashion to the actin cytoskeleton, although the mechanisms involved are unclear (36). That JMY plays a role in cell motility and invasion is likely to reflect its actin nucleation properties and its ability to regulate cadherin levels, mediated by its WH2 domains, as removal of the WH2 domains in JMY reduced its ability to enhance motility. Other actin nucleating proteins have also been shown to influence cell motility. For example, N-WASp can positively influence cell motility (37–39). Similarly, WAVE/SCAR isoforms have been shown to be required for lamellipodia formation and promote cell motility and invasion (40–42). A striking difference however, between these proteins and JMY, is the nuclear role that JMY has in regulating the p53 response. E-cadherin levels can impact on Wnt/β-catenin signaling, with high cadherin expression resulting in attenuated Wnt/β-catenin signaling (14). It is therefore tempting to speculate that JMY may also influence Wnt/β-catenin signaling, and future experiments will aim to address this intriguing possibility.

JMY Coordinates Cytoskeletal Events with Nuclear p53 Response.

p53 and components of its signaling pathways have been implicated in tumor metastasis and motility (19, 43, 44). Whilst JMY enhances motility, in general, p53 acts negatively to prevent motility and tumor metastasis, as anticipated for a protein that acts in tumor suppression (45). For example, loss of p53 has been shown to cooperate with activated Ras(V12) to promote cell motility (46). Moreover, targeting JMY to the nucleus enhances its role as a p53 co-factor, while diminishing its ability to enhance cell motility. Thus, although we cannot rule out that JMY is able to influence cell motility through its role as a transcription co-factor, our data suggest this is unlikely to be the case. JMY may act as a ‘damage sensor’ to relay signals to the nucleus and in this way JMY is uniquely positioned with a dual function as both a protein that influences cytoplasmic cytoskeletal events which can, during the stress response, localize to the nucleus to regulate p53 activity (Fig. 4G).

While the polymerization status of nuclear actin is the subject of much debate, our study adds to the growing body of evidence that actin polymerization activity functions in the nucleus (47, 48), since JMY WH2 domain activity directs nuclear actin incorporation. The fact that blocking actin polymerization with latrunculin A negated the positive effects of JMY on p53 activity, support a positive role for actin nucleation. Moreover, our results suggest that the nuclear function of JMY WH2 domains is independent of Arp2/3, although it is possible that JMY interacts with other nuclear Arps to influence p53 transcriptional activity. This is particularly intriguing, as thus far, only two other actin nucleation factors with a similar Arp2/3 independent nucleation activity have been described: Spir and formins (49).

In conclusion, our results have established that the WH2 domains in JMY, which function in actin nucleation, influence cellular motility and adhesion. The nuclear accumulation upon DNA damage facilitates transcriptional control during the p53 response, while limiting JMY's contribution to cell motility. Thus, JMY co-ordinates cytoskeletal events and motility with transcriptional control during the DNA damage response.

Materials and Methods

Plasmids, Antibodies, and Reagents.

The following plasmids have been previously described (2); pcDNA3 HA-JMY, HA-JMYΔ468–558, HA-JMY-NLS, and HA-JMY 1–504. ΔWH2, Δ468–558ΔWH2, and NLS-ΔWH2 were generated by engineering a stop codon at amino acid 848 using the Stratagene QuikChange Site-directed mutagenesis kit according to manufacturer's instructions. JMYW981A and JMYW981A-NLS were generated using the Stratagene QuikChange Site-directed mutagenesis kit. All constructs were verified by sequencing. Goat anti-JMY L-16, rabbit anti-JMY M300, goat anti-GAPDH V18, mouse anti-p53 DO-1, rabbit N-cadherin, mouse E-cadherin, and mouse anti-PCNA antibodies were from Santa Cruz. Rabbit anti-JMY antibody 1289 has been previously described (7). Mouse anti-HA antibody HA11 was from BAbCO. Mouse anti-actin, anti-flag M2 and anti-nucleophosmin antibodies were from Sigma. HRP-conjugated secondary antibody was from DAKO. Alexa Fluor conjugated secondary antibodies and Alexa Fluor 488 conjugated actin were from Molecular Probes. Phalloidin-TRITC was from Sigma.

Cell Lines and Generation of Stable Cell Lines.

HEK 293, MCF-7, U2OS, SAOS2, NIH 3T3 Ras transformed (Cancer Research U.K.), and wild-type MEFS (a gift from K. Vousden, Beatson Institute for Cancer Research) were grown in 5% FCS-DMEM plus antibiotics under 5% CO₂. A stable JMY shRNA cell line was created in MCF-7 and HEK 293 cells using the Clontech Knockout^TM single vector inducible RNAi system according to manufacturer's instructions. The JMY shRNA was generated from the JMY siRNA, sense 5′ GCA ACU AGA AAG CAU CAA AUU 3′, according to manufacturer's instructions. JMY shRNA was induced using 1 μg/mL doxycycline. JMY stable inducible overexpression and empty vector control cell lines were created in U2OS cells using TET-On gene expression system (Clontech) and expression was induced by the addition of 1 μg/ml doxycycline. U2OS cells stably expressing wild-type JMY and JMY-NLS were obtained after transfection of the appropriate constructs into U2OS cells and carrying out selection with 600 μg/mL G418.

See SI Methods for additional details.