Introduction

Metastasis represents a critical step in carcinogenesis, and is responsible for greater than 90% of cancer-related deaths (Sporn, 1996). Prostate cancer is the most common form of cancer in men in the United States, the second largest cause of cancer-related death in men, and is thus a major public heath problem (Jemal et al., 2005). Death owing to prostate cancer is caused by metastatic spread of disease (Carroll et al., 2001). Alterations in cell attachment and invasion represent early steps in the metastatic cascade (see for reviews, Ruoslahti, 1996; Woodhouse et al., 1997). The regulation of metastatic behavior, as well as the dependent processes of adhesion and invasion, are not well understood in general, and not for prostate cancer in particular.

Transforming growth factor (TGF), a 25 kDa disulde-linked homodimeric extracellular cytokine, is an important regulator of cell adhesion and motility in a variety of cell types, including prostate (Shi and Massague, 2003). As TGF not only exerts pleiotropic effects within a given cell, but appears to act through different pathways in different cell types, it is important to investigate its role in a cell type-specific manner (Attisano and Wrana, 2002). We have previously shown that both Smad3 and p38 MAP kinase were necessary for TGF-mediated increases in cell adhesion in human prostate (Hayes et al., 2003). Most recently, we have shown that p38 MAP kinase is necessary for TGF-mediated up-regulation of matrix metalloproteinase type 2 (MMP-2), as well as TGF-dependent increases in prostate cell invasion (Huang et al., 2005).

As prior studies established the importance of p38 MAP kinase in regulating cell invasion, the current study was undertaken to identify downstream effectors of p38 MAP kinase as they relate to TGF-mediated regulation of invasion. p38 MAP kinase is a serine/threonine kinase, which is activated in a time-dependent manner after TGF treatment (for review see, Platanias, 2003). It therefore follows that downstream effector proteins must be phosphorylated on serine and/or threonine in a time-dependent manner after TGF treatment. Furthermore, if proteins were in fact specific downstream effectors of p38 MAP kinase, then inhibition of p38 MAP kinase activity should block phosphorylation.

The current study describes the results of such a screening, and how it led to the initial identification of mitogen-activated protein kinase-activated protein kinase 2 (MAPKAPK2), and heat-shock protein 27 (HSP27). Investigating further, using chemical inhibitors of p38 MAP kinase, wild-type and dominant-negative constructs of MAPKAPK2, wild-type and mutant, non-phosphorylateable, HSP27, as well as siRNA-mediated knock down of both MAPKAPK2 and HSP27, we show for the first time in any cell type that both MAPKAPK2 and HSP27 are necessary for TGF-mediated increases in MMP-2 activity, as well as cell invasion.

Results

Screening for p38 MAP kinase downstream effectors

We have previously shown that p38 MAP kinase is phosphorylated (i.e., activated) in a time-dependent manner after treatment of PC3-M cells with TGF (Hayes et al., 2003). To further refine the time course of activation, and to expand this analysis to other cell lines, we first sought to characterize the time course of TGF-dependent activation of p38 MAP kinase in PC3 and PC3-M prostate cancer cells (Figure 1a). In both cell lines tested, phospho-p38 MAP kinase levels began to increase within 10 min of TGF treatment, peaking at 60 min in PC3 cells, and at 90 min in PC3-M cells, and declining thereafter.

Figure 1.

Kinetics of protein phosphorylation after TGF treatment. One day after plating, PC3 and PC3-M cells were changed to serum-free media, and 24 h later, treated as indicated. (a) Kinetics of p38 MAP kinase activation. PC3 and PC3-M cells were treated with TGF (or buffer for controls) for the indicated times, and equal amounts of protein from resultant cell lysates probed for phosho-p38 MAP kinase (pp38) by Western blot, as described in Materials and methods. After stripping, the blot was reprobed for total p38 MAP kinase (p38). Similar results were seen in a replicate experiment performed at a separate time. (b) Screening for signaling proteins downstream of p38 MAP kinase. PC3 cells were pre-treated with either SB203580 for 1 h, with genistein for 21 h, or not, then treated with TGF for the indicated times, and equal amounts of protein from resultant lysates were separated on 4–15% linear gradient gel under reducing conditions and probed for phosphothreonine.

Full figure and legend (164K)

p38 MAP kinase is a serine/threonine kinase, is activated by TGF in a time-dependent manner, and has been shown by us to be necessary for TGF-dependent increases in prostate cell invasion (Huang et al., 2005). It therefore stands to reason that effectors downstream of p38 MAP kinase, which are responsible for TGF-dependent regulation of invasion, should meet the following criteria: (1) be phosphorylated on serine and/or threonine, (2) be phosphorylated in a time-dependent manner after TGF treatment and (3) not be phosphorylated in the face of p38 MAP kinase inhibition. Two different inhibitors of p38 MAP kinase with different mechanisms of action were tested. SB203580 is a chemical inhibitor that inhibits the kinase activity of and isoforms of p38 MAP kinase, which have high homology and are expressed in a relatively ubiquitous manner (Tong et al., 1997; Wilson et al., 1997). We have shown that SB203580 is a specific inhibitor of p38 MAP kinase in PC3 and PC3-M cells, and its regulation of prostate cell adhesion and invasion (Hayes et al., 2003; Huang et al., 2005). We have also shown that genistein (4,5,7-trihydroxyisoflavone) blocks TGF-mediated activation of both isoforms of p38 MAP kinase, thus inhibiting cell invasion (Huang et al., 2005). SB203580 and genistein studies were performed in both PC3 and PC3-M cells.

Based upon the kinetics of p38 MAP kinase phosphorylation observed after TGF stimulation, Figure 1a, several time points were selected for screening purposes. PC3-M and PC3 cells were pre-treated (or not) with either SB203580 or genistein, and then treated with TGF (or not) for up to 180 min, and resultant protein lysates probed for either phosphoserine or phosphothreonine by Western blot. No protein bands, which met the above criteria, were evident in PC3-M cells probed for either phosphoserine or phosphothreonine, or in PC3 cells probed for phospho-serine (data not shown). However, in PC3 cells probed for phospho-threonine, a 27 kDa band exhibited increased phosphorylation at the 45 and 90 min time points after TGF treatment (Figure 1b). Increased phosphorylation of the 27 kDa band was not present in cells pre-treated with either SB203580 or with genistein, nor were any changes evident in cells treated with TGF for only 30 min (either in the presence or absence of SB203580 or genistein). No changes in phosphothreonine levels were seen in cells treated with either SB203580 or genistein in the absence of TGF (data not shown).

Heat-shock protein 27 (HSP27) has previously been reported to be phosphorylated by MAPKAPK2 (Stokoe et al., 1992b), and MAPKAPK2 is a known substrate of p38 MAP kinase (Krump et al., 1997). HSP27 was therefore evaluated as a potential regulator of cell invasion, which was downstream of p38 MAP kinase, by first assessing whether it was phosphorylated in a time-dependent manner after TGF treatment. As can be seen (Figure 2), phospho-HSP27 levels began to increase 10–20 min after treatment with TGF, in both cell lines tested. In PC3 cells, phospho-HSP27 levels peaked at 90 min, and declined thereafter, consistent with changes observed on the screening phosphothreonine Western blot (Figure 1b). In PC3-M cells, however, levels of phospho-HSP27 continued to increase throughout the 3 h time course of the experiment.

Figure 2.

TGF leads to phosphorylation of HSP27. PC3 and PC3-M cells were treated with TGF, as in Figure 1a, and Western blot performed for phospho-HSP27 and total HSP27. Similar results were seen in a replicate experiment performed at a separate time.

Full figure and legend (70K)

To assess whether MAPKAPK2 and/or HSP27 activation by TGF was dependent upon p38 MAP kinase, we pre-treated cells with SB203580 or with the inactive chemical analog, SB202474, then with TGF (or not), and measured the effects upon phosphorylation of p38 MAP kinase, MAPKAPK2 and HSP27 by Western blot (Figure 3). As can be seen, SB203580 treatment was associated with increases in phosphorylation of p38 MAP kinase. This is because SB203580 inhibits the kinase activity of p38 MAP kinase, but does not prevent it from being phosphorylated (Ravanti et al., 1999; Hatakeyama et al., 2002). As previously described by us (Huang et al., 2005) and others (Munshi et al., 2004), inhibition of the kinase activity of p38 MAP kinase, by SB203580, is associated with increased phosphorylation of p38 MAP kinase, thus indicating that biologically relevant concentrations of SB203580 were being utilized. Importantly, TGF-mediated increases in phosphorylation of both MAPKAPK2 and HSP27 were abrogated by SB203580, whereas SB202474 chemical control had little-to-no effect, in both cell lines tested. Thus, MAPKAPK2 and HSP27 are activated downstream of p38 MAP kinase, after TGF treatment, in prostate cancer cells.

Figure 3.

MAPKAPK2 and HSP27 are phosphorylated downstream of p38 MAP kinase. PC3 and PC3-M cells were cultured, pretreated for 1 h with SB203580 (p38 MAP kinase inhibitor), SB202474 (inactive chemical analog), or not (control), and then treated with TGF for 10 min (for MAPKAPK2) or 1 h (p38 MAP kinase, HSP27). Cell lysates were then probed for total and phospho-p38 MAP kinase (p38 and pp38), total and phospho-MAPKAPK2 (pMK2 and ppMK2), total and phospho-HSP27 (pHSP27 and ppHSP27), or GAPDH, by Western blot. Individual bands from two separate experiments were quantitated, expressed as the ratio of phosphoprotein to total protein and resultant data graphed as the mean+s.d. N=2 in the bottom panel.

Full figure and legend (191K)

MAPKAPK2 and HSP27 regulate TGF-mediated increases of MMP-2 activity and cell invasion

Above findings raised the notion that MAPKAPK2 and/or HSP27 may be downstream signaling effectors of p38 MAP kinase, with respect to TGF-mediated regulation of matrix metalloproteinase type 2 (MMP-2) activity and cell invasion. If this were the case, then disruption of either MAPKAPK2 or HSP27 signaling should also inhibit TGF-mediated increases of MMP-2 activity and cell invasion. A series of experiments was therefore performed to test this hypothesis. First, cells were transiently transfected with either wild-type MAPKAPK2 (MK2WT), a kinase-inactive dominant-negative mutant (MK2K76R), a constitutively active mutant (MK2EE) or with empty vector control (VC), and effects upon phosphorylation of HSP27 measured by Western blot (Figure 4). Treatment with TGF increased phosphorylation of MAPKAPK2 in wild-type, dominant-negative (which is kinase inactive, but can still be phosphorylated) and VC-transfected cells. In contrast, high baseline levels of phosphorylation and no further increase with TGF were seen with constitutively active mutant-transfected cells. Importantly, TGF treatment led to increases in phospho-HSP27 (ppHSP27) levels in cells transfected with either wild-type MAPKAPK2 or with VC. In contrast, in cells transfected with dominant-negative MAPKAPK2, the effects of TGF were abrogated. Finally, in cells transfected with constitutively active MAPKAPK2, the baseline level of ppHSP27 was high, and no further significant increase was seen after TGF treatment. Thus, HSP27 is activated downstream of MAPKAPK2 after TGF treatment in prostate cancer cells.

Figure 4.

MAPKAPK2 phosphorylates HSP27. PC3 and PC3-M cells were transfected with MAPKAPK2 wild-type, pcDNA3mycMK2WT (MK2WT), dominant-negative kinase-inactive mutant, pcDNA3mycMK2K76R (MK2K76R), or constitutively active mutant, pcDNA3mycMK2T205E317E (MK2EE), or with empty VC. Twenty-four hours after transfection, cells were changed to serum-free media, and 24 h later, cells were treated (or not) with TGF for either 10 min (for HSP27) or for 1 h (MAPKAPK2), and total and phospho-MAPKAPK2 (pMK2 and ppMK2), total and phospho-HSP27 (pHSP27 and ppHSP27), and GAPDH were measured by Western blot. Similar results were seen in a separate experiment, performed at a different time. Cell transfection and Western blotting were performed as described in Materials and methods.

Full figure and legend (80K)

We next investigated the role of MAPKAPK2, and then of HSP27, in regulating MMP-2 and cell invasion. For MAPKAPK2, TGF increased the activity of MMP-2 in both wild-type and VC-transfected cells, Figure 5a. Importantly, in MAPKAPK2 dominant-negative cells, no increase in MMP-2 activity was observed following TGF treatment. Conversely, in cells transfected with constitutively active MAPKAPK2, baseline levels of MMP-2 were elevated, but did not respond further to TGF. Similar effects were observed in both cell lines tested. The characteristic multiple zymogens of MMP-2 are evident under the current assay conditions in PC3-M cells, as evidenced by the presence of multiple bands on the gel. Increased gel loading and/or over-exposure would also reveal them in PC3 cells (data not shown), as described previously (Huang et al., 2005).

Figure 5.

MAPKAPK2 regulates MMP-2 activation and cell invasion. Cells were transfected with MAPKAPK2 wild-type (MK2WT), dominant-negative kinase-inactive mutant (MK2K76R), or constitutively active mutant (MK2EE), or with empty VC, as described in Figure 4. (a) MAPKAPK2 regulates MMP-2 activity. Cells were changed to serum-free media, treated with TGF (or not) for 24 h, conditioned media collected, and MMP-2 activity of equal amounts of media measured by zymography. Similar results were obtained in a separate experiment. (b) MAPKAPK2 regulates cell invasion. Cells were treated (or not) with TGF for 24 h, and cell invasion measured. (c and d) MAPKAPK2 is dependent upon p38 MAP kinase activity. One hour before addition to TGF (or not), cells were pre-treated with SB203580 or SB202474, as indicated. The effect upon cell invasion was then measured in cells transfected with either wild-type MAPKAPK2, (c) or with constitutively active MAPKAPK2. (d) Cell invasion is depicted as the percent of invading cells, relative to untreated VC cells, and is the mean+s.e.m. N=4 of a single experiment. Invasion assays were repeated at a different time, also with N=4, with similar results. ^*P-value for difference between values is <0.05 (two-sided t-test); NS, denotes not significant. Cell transfection, zymography, and invasion assays were performed as described in Materials and methods.

Full figure and legend (144K)

As the activity of MMPs is closely linked to cancer invasion and metastatic behavior, the functional significance of changes in MMP-2 were then evaluated by assessing effect upon cell invasion (Figures 5b–d). TGF increased cell invasion in both VC and wild-type MAPKAPK2-transfected cells (Figure 5b). In contrast, in dominant-negative transfected cells, baseline levels of invasion were below those seen in both VC- and MAPKAPK2-transfected cells. More importantly, in the presence of TGF, the invasion of dominant-negative cells was significantly below that for VCs, as well as for wild-type cells. In addition, there was a complete abrogation of TGF-mediated increases in invasion in dominant-negative cells. Finally, with constitutively active MAPKAPK2 transfectants, baseline levels of invasion were high, and TGF did not further increase invasion. Similar effects were observed in both cell lines tested.

Above findings demonstrate that MAPKAPK2 is important in regulating cell invasion, particularly that responsive to TGF. If MAPKAPK2-related effects were dependent upon p38 MAP kinase, then wild-type MAPKAPK2 should not rescue invasion in the face of p38 MAP kinase inhibition, whereas constitutively active MAPKAPK2 should be able to do so. By measuring invasion in wild-type (Figure 5c) and constitutively active (Figure 5d) transfectants, treated with SB203580, this notion is supported. Specifically, inhibition of p38 MAP kinase activity by SB203580 abrogated TGF-mediated increases in invasion in wild-type transfectants, compared to cells treated with inactive chemical control, SB202474. However, in constitutively active cells, the inhibitory effects of SB203580 were markedly reduced, having no significant effect in PC3 cells, and only a small, albeit significant, effect in PC3-M cells.

HSP27 was investigated next (Figure 6). First, cells were transiently transfected with wild-type HSP27 (HSP27 WT), with a non-phosphorylateable mutant (HSP27 3G), or with empty VC. High levels of HSP27 were observed in both wild-type and mutant transfectants (Figure 6a). TGF treatment increased phospho-HSP27 levels in VC and wild-type cells, but had minimal to no effect in mutant-transfected cells. Next, effects upon MMP-2 activity were measured (Figure 6b). TGF increased MMP-2 activity in both wild-type and control cells, but had no effect in mutant-transfected cells (Figure 6b). Similarly, TGF increased invasion in both wild-type and control cells (Figure 6c), in both cell lines tested. In mutant-transfected cells, baseline levels of invasion were significantly decreased, and response to TGF almost completely abrogated. Specifically, there was a small, but significant, increase in invasion in response to TGF treatment in PC3 cells transfected with mutant HSP27, whereas in PC3-M cells, there was no significant increase. Finally, inhibition of p38 MAP kinase activity by SB203580 abrogated TGF-mediated increases in invasion in wild-type transfectants, compared to cells treated with inactive chemical control, SB202474 (Figure 6d). Similar findings were observed in both cell lines evaluated.

Figure 6.

HSP27 regulates MMP-2 activation and cell invasion. PC3 and PC3-M cells were transfected with either HSP27 wild-type, pcDNA3.1-HSP27 WT (HSP27 WT), non-phosphorylateable mutant, pcDNA3.1-HSP27 3G (HSP27 3G), or with VC, as indicated, and total and phospho-HSP27 (pHSP27 and ppHSP27) and GAPDH protein measured by Western Blot (a). (b) HSP27 regulates MMP-2 activity. Cells were changed to serum-free media, treated with TGF (or not) for 24 h, conditioned media collected and MMP-2 activity measured by zymography as in Figure 5. Similar results were obtained in a separate experiment. (c) HSP27 regulates cell invasion. Cells were treated (or not) with TGF for 24 h, and cell invasion measured. (d) HSP27 is dependent upon p38 MAP kinase activity. One hour before addition to TGF (or not) cells were pre-treated with SB203580 or SB202474, as indicated, and cell invasion measured as in Figure 5. Cell invasion is depicted as the percent of invading cells, relative to untreated VC cells and is the mean+s.e.m. (N=4) of a single experiment. Invasion assays were repeated at a different time, also with N=4, with similar results. ^*P-value for difference between values is <0.05 (two-sided t-test); NS, denotes not significant.

Full figure and legend (145K)

To further evaluate the functional importance of MAPKAPK2 and HSP27 in the regulation of cell invasion, a separate series of experiments was performed wherein siRNA methodology was used to knock down the expression of MAPKAPK2, and of HSP27, and effects upon invasion measured. In cells transfected with MAPKAPK2-directed siRNA, near complete suppression of MAPKAPK2 protein was observed (Figure 7a). Suppression was specific, as there was little effect upon either HSP27 or MAPKAPK3 (MK3, which has 70% homology to MAPKAPK2; (Ludwig et al., 1996), and no effect upon either p38 MAP kinase or GAPDH protein levels. In addition, no suppression of targeted proteins was observed in cells treated with siRNA to green fluorescent protein (GFP) or in sham-transfected controls. Conversely, specific, and essentially total, suppression of HSP27 protein was observed in cells exposed to HSP27 siRNA. Finally, when siRNA to both MAPKAPK2 and HSP27 were combined, efficacy and specificity were both preserved.

Figure 7.

Knockdown of either MAPKAPK2 or HSP27 inhibits cell invasion. PC3 and PC3-M cells were transfected with siRNA to either MAPKAPK2 (siMK2), HSP27 (siHSP27) or to green fluorescent protein (GFP; siGFP), with siRNA to both MAPKAPK2 and HSP27 (siMK2+siHSP27), or sham transfected (control), as indicated. (a) MAPKAPK2 and HSP27 siRNA is effective and specific. Two days after transfection with the indicated siRNA species, expression of HSP27, p38 MAP kinase (p38), GAPDH and MAPKAPK3 (MK3) protein were measured by Western blot. Similar results were seen in a separate experiment. (b) Knockdown of MAPKAPK2 or HSP27 inhibits cell invasion. Two days after transfection with the indicated siRNA species, cells were treated with TGF (or not), and cell invasion measured as described in Figure 5. Cell invasion is depicted as the percent of invading cells, relative to siGFP-treated control cells, and is the mean+s.e.m. (N=4) of a single experiment. Similar results were obtained in a replicate experiment, performed at a different time (also, N=4). ^*P-value for difference between values is <0.05 (two sided t-test); ns, denotes not significant.

Full figure and legend (108K)

Compared to GFP siRNA cells, both MAPKAPK2 and HSP27 siRNA cells exhibited decreased baseline levels of invasion, as well as significantly lower levels of invasion after TGF treatment, in both cell lines tested (Figure 7b). If HSP27 and MAPKAPK2 were independent regulators of TGF's effects upon invasion, then simultaneous knock down of HSP27 and MAPKAPK2 should be associated with a further decrease in invasion. As can be seen, however, no further decrease in invasion is evident when MAPKAPK2 siRNA is given along with HSP27 siRNA, as compared to cells treated with HSP27 siRNA alone.

HSP27 is upregulated in aggressive prostate cancer

Several investigators have shown that HSP27 is upregulated during prostate cancer cell progression, and that high levels of HSP27 in prostate cancer biopsy specimens are associated with a poor clinical outcome (Cornford et al., 2000; Rocchi et al., 2004). The current study demonstrates that HSP27 regulates cell invasion. Cell invasion is associated with metastatic behavior and is associated with prostate cancer progression. Taken together, these facts support the notion that dysregulated HSP27 contributes to prostate cancer progression. If this were the case, then higher levels of HSP27 would be expected in PC3-M metastatic variant cells as compared to the parental PC3 cell line. This was shown to be the case by comparing HSP27 protein expression between PC3-M and PC3 cells by Western blot, demonstrating almost two-fold higher levels in PC3-M cells (Figure 8).

Figure 8.

HSP27 protein is elevated in metastatic variant cells. Equal amounts of total cell lysate from PC3 parental and PC3-M metastatic variant cells were subjected to Western blotting for HSP27, p38 MAP kinase (p38), MAPKAPK2 (MK2) and -tubulin (loading control). HSP27 bands were quantitated, normalized to PC3 cells and mean+s.d. percent increase in PC3-M cells determined (N=two separate experiments); two-sided t-test value <0.05 for differences between PC3 and PC3-M cells.

Full figure and legend (25K)

Discussion

In the current study, we sought to identify downstream effectors of p38 MAP kinase, as they relate to its regulation of MMP-2 activity and cell invasion. Based upon the fact that p38 MAP kinase is a serine/threonine kinase, which is activated in a time-dependent manner after TGF treatment, we undertook a successful proteomics based screening approach. We therefore sought to identify time-dependent phosphorylation on serine and/or threonine, in both PC3 and PC3-M cells, after TGF treatment, which was dependent upon the activity of p38 MAP kinase.

As screening approaches are limited by the resultant number of false positives, we used two separate inhibitors of p38 MAP kinase, which we have also shown to inhibit cell invasion (Huang et al., 2005). SB203580 inhibits the kinase function of p38 MAP kinase (Ravanti et al., 1999; Hatakeyama et al., 2002), whereas genistein blocks its activation (i.e., phosphorylation) (Huang et al., 2005). That positive findings were not seen with PC3-M cells, or with PC3 cells probed for phosphoserine, but only with PC3 cells probed for phosphothreonine, is a measure of the stringency of the screening approach utilized. This notion is also supported by the fact that MAPKAPK2 was not identified on phospho-protein-based screening investigations. Thus, it is possible that there are additional proteins of importance, but the current level of stringency did not permit their detection.

Importantly, we went on to specifically demonstrate that HSP27 does in fact exhibit time-dependent phosphorylation after TGF treatment in both PC3 and PC3-M cells. Interestingly, the kinetics of HSP27 dephosphorylation differs between the two cell types investigated. While the meaning of this, if any, is not clear at this time, it highlights the fact that there is differential signaling between different cell types, and underscores the importance of investigating any single pathway in multiple cell types. In this regard, it is important to note that the current study employed two separate human prostate cell lines, and that findings in each line were similar, if not identical, throughout the series of investigations performed.

MAPKAPK2 is known to be a p38 MAP kinase downstream signaling protein (Stokoe et al., 1992a). We have previously shown that p38 MAP kinase is necessary for the regulation of TGF-mediated increases in cell adhesion and invasion (Hayes et al., 2003; Huang et al., 2005). While it therefore stands to reason that MAPKAPK2 would regulate cell invasion, to our knowledge, this is the first report demonstrating that TGF activates MAPKAPK2 through p38 MAP kinase. Thus, current findings represent an extension of prior findings by us, and are in agreement with that of Kotlyarov et al. (2002), who show that migration of MAPKAPK2-deficient mouse embryonic fibroblasts and smooth muscle cells on fibronectin is dramatically reduced.

p38 MAP kinase phosphorylation, which is necessary for the activation of the kinase function, was increased when cells were treated with SB203580. SB203580 is a chemical inhibitor that inhibits the kinase activity of p38 MAP kinase, but does not prevent p38 MAP kinase from being phosphorylated (Tong et al., 1997; Wilson et al., 1997). Increased p38 MAP kinase phosphorylation in the face of SB203580-mediated inhibition of its kinase function has previously been described by us (Huang et al., 2005) and others (Munshi et al., 2004), and is not cell type specific. A likely explanation would be the presence of a feedback inhibition regulatory loop. SB203580-mediated inhibition of p38 MAP kinase's enzyme activity therefore leads to loss of feedback inhibition and increased phosphorylation of p38 MAP kinase. As SB203580 is a reversible inhibitor, immunoprecipitation-kinase assays are not able to detect effects, and thus in vivo measures must be employed. It was therefore important that we went on to show that SB203580-mediated inhibition of p38 MAP kinase led to inhibition of phosphorylation of MAPKAPK2, an in vivo substrate of p38 MAP kinase (Krump et al., 1997).

HSP27 is known to be a substrate of MAPKAPK2 (Stokoe et al., 1992b). While the function of HSP27 is not clear, its phosphorylation appears to stimulate the polymerization of actin (Guay et al., 1997; Gerthoffer and Gunst, 2001). It also appears to have a role in regulating cell invasion, but its function in this regard may be cell type specific. Specifically, Aldrian et al. (2002) demonstrated that HSP27 overexpression decreased invasion of melanoma cells. In contrast, others have shown that HSP27 increases cell migration in both breast and smooth muscle cells (Hedges et al., 1999; Hansen et al., 2001). It was therefore important that we investigated the role of HSP27 in human prostate cancer, a common form of cancer that causes death by metastasizing, and did so using multiple cell lines. Our findings of increased HSP27 expression in aggressive prostate cancer are in agreement with similar prior findings by other investigators (Cornford et al., 2000; Rocchi et al., 2004). As increased cell invasion is a hallmark of the metastatic phenotype, and thus a measure of aggressiveness, the current study provides a likely mechanistic link to this previously identified association.

We demonstrated for the first time, in any cell type, that HSP27 mediates TGF-dependent increases in cell invasion and MMP-2 activity. It has previously been shown that inhibition of p38 MAP kinase blocks HSP27 phosphorylation, and decreases invasion in smooth muscle cells (Hedges et al., 1999). Our findings are in direct agreement with these, and demonstrate for the first time that signaling goes from p38 MAP kinase to MAPKAPK2 to HSP27, and then to the regulation of MMP-2 and cell invasion.

In summary, our findings demonstrate that prostate cancer cell invasion is mediated through a pathway in which p38 MAP kinase activates MAPKAPK2, which in turn leads to the phosphorylation of HSP27, which in turn regulates MMP-2 activation and cell invasion. TGF activates each of these pathway members, and these members, in turn, regulate its effects upon MMP-2 and cell invasion. This study provides new information about the biological regulation of cell invasion and MMP-2 activity in human prostate cancer, in particular that regulated by TGF.

Materials and methods

Materials

Phosphatase inhibitor cocktail I and II, genistein and gelatin (used in invasion assays) were purchased from Sigma (St Louis, MO, USA). The p38 MAP kinase inhibitor, SB203580, and the structurally related inactive analog, SB202474, were from Calbiochem (San Diego, CA, USA). Genistein, SB203580 and SB202474 were stored as stocks in DMSO, and were thawed just before use. Antibodies were obtained from the following sources: phosphoserine, Q5, and phosphothreonine, Q7, Qiagen (Valencia, CA, USA); p38 MAP kinase, clone C-20 (Santa Cruz Biotechnology, Santa Cruz, CA, USA); phospho-p38 MAP kinase, clone 9211S (recognizes Thr180 and Tyr 182), phospho-MAPKAPK2, clone 3041 (recognizes Thr334), phospho-HSP27, clone 2401S (recognizes Ser82), MAPKAPK2, clone 3042, and HSP27, clone 2402 (Cell Signaling Technology, Beverly, MA, USA); GAPDH, CSA-335E (Stressgen, Victoria, CA, USA); anti-mouse Ig-horseradish peroxidase (HRP) and anti-rabbit Ig-horseradish peroxidase, were part of the ECL Western Blotting System (Amersham Pharmacia Biotech), and were used to detect proteins in Western blots. Recombinant human TGF1 (R&D Systems; Minneapolis, MN, USA) was resuspended and stored according to the manufacturer's instructions, and was used at a final concentration of 2 ng/ml.

Constitutively active -galactosidase expression vector, pCMV--gal, was from Stratagene (La Jolla, CA, USA). MAPKAPK2 wild-type, pcDNA3mycMK2WT, dominant-negative kinase-inactive mutant, pcDNA3mycMK2K76R, and constitutively active mutant, pcDNA3mycMK2EE, plasmids were generous gifts from Matthias Gaestel (Institute of Biochemistry, Medical School, Hannover, Germany), and have been described (Winzen et al., 1999). HSP27 wild-type, pcDNA3.1-HSP27 WT, and non-phosphorylateable mutant, pcDNA3.1-HSP27 3G, plasmids were kindly provided by Rainer Benndorf (University of Michigan), and have been described earlier (Schafer et al., 1999).

Cell culture and transfection

The origin, characteristics and culture conditions for PC3 and PC3-M established cell lines have been described previously (Liu et al., 2000). Cells were maintained at 37°C in a humidified atmosphere of 5% carbon dioxide, with biweekly media changes. Cells were maintained at 37°C in a humidified atmosphere of 5% carbon dioxide, with biweekly media changes, were drawn from stored stocks, replenished on a standardized periodic basis and were routinely monitored for mycoplasma. Cell viability was routinely monitored under all experimental conditions, by counting the number of trypan blue excluding cells, under an inverted microscope, using a hemocytometer, and unless otherwise stated, was not affected.

In some experiments, cells were treated with various agents. Cells were exponentially growing and non-confluent, and, unless otherwise stated, were pre-treated for 1 h with 10 M SB203580 or SB202474, or 50 M genistein, and then with 2 ng/ml TGF (or not) for various time periods, as indicated. For Western blots, plasmids were transfected into cells, which had been plated into six well plates the previous day, with TransIT^®-LT1 Transfection Reagent (Mirus, Madison, WI, USA), as per the manufacturer's instructions, using 2 g of the indicated expression plasmid. For measurement of MMP-2 activity by zymography, cells were transfected using LipofectAmine-2000 (Invitrogen, Carlsbad, CA, USA), as per the manufacturer's instructions, using 4 g of the indicated expression plasmid. For cell invasion assays, cells were transfected using LipofectAmine-2000 (Invitrogen, Carlsbad, CA, USA), as per the manufacturer's instructions, using 3 g of the indicated expression plasmid, along with 1 g of pCMV--gal. The use of a constitutively active -galactosidase vector allowed identification of transfected cells in the context of a three-dimensional matrix, used in invasion assays, as described previously (Huang et al., 2005). For all assays, cells were treated, as indicated, 24 h after transfection.

For siRNA studies, cells were plated into six-well plates the previous day, and transfected with 15 l GeneSilencer reagent (Gene Therapy Systems, San Diego, CA, USA), according to the manufacturer's instructions, using 1 g of siRNA along with 1 g of pCMV--gal. In some experiments, cells were transfected with siRNA to both MAPKAPK2 and HSP27, as indicated. In those circumstances, cells were transfected with 1 g of MAPKAPK2 siRNA and 1 g of HSP27 siRNA along with 1 g of pCMV--gal. After 24 h, cells were changed to fresh culture media, and after 60 h, cells were harvested, and analysed as described.

Cell lysis and Western blot analysis

Cells were lysed, and Western blots performed, as described previously, with modifications (Liu et al., 2001). Briefly, cells were lysed at 4°C in cell lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM -glycerolphosphate) in the presence of protease inhibitors (1 g/ml leupeptin, 1 g/ml aprotinin, 1 mM PMSF; all from Sigma) and phosphatase inhibitors (10 mM NaF, 1 mM orthovanadate, phosphatase inhibitor cocktail I and cocktail II, both at 1:100 dilution; all from Sigma). Resultant lysates were centrifuged at 14 000g for 30 min at 4°C, and protein concentration in the supernatant determined by Bradford dye-binding assay (Bio-Rad).

When screening for protein-serine or -threonine phosphorylation by Western blot, equal amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), on a 4–15% linear gradient gel, under reducing conditions. For all other Western blots, equal amounts of protein were separated by SDS-PAGE on a 12% gel, under reducing conditions. Separated proteins were transferred onto 0.45 m nitrocellulose (Schleicher & Schuell, Keene, NH, USA), in a wet transfer cell. Blots were blocked with 5% milk in TBST (10 mM Tris-HCl, pH 7.6, 80 mM NaCl, 0.1% Tween-20), for 1 h, at room temperature, and probed overnight at 4°C with primary antibody, diluted in TBST with 5% BSA (fraction V; Sigma). Anti-phosphoserine, clone Q5, was diluted 1:250; anti-phosphothreonine, clone Q7, 1:250; anti-phospho-p38 MAP kinase, clone 9211s, 1:1000; anti-phospho-MAPKAPK2, clone 3041, 1:1000 and anti-phospho-HSP27, clone 2401s, 1:1000. After washing, membranes were incubated for 1 h at room temperature with anti-rabbit-HRP-conjugated secondary antibody or anti-mouse-HRP-conjugated secondary antibody, and visualized by chemiluminescence, using the ECL Western Blotting Kit (Enhanced Chemiluminescence; Amersham Pharmacia Biotech), as per the manufacturer's instructions. Membranes were then stripped by treating with stripping buffer (100 mM -mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7) at 50°C for 30 min. After washing with TBST at room temperature, membranes were re-exposed (after re-addition of HRP substrate to confirm removal of prior antibody) before being re-blocked and re-probed, as indicated. Specifically, for total p38 MAP kinase, using antibody clone C-20 diluted 1:1000; and for MAPKAPK2, clone 3042, 1:1000; for HSP27, clone 2402, 1:1000; for GAPDH, clone CSA-335E, 1:5000. Each experiment was repeated one or more times, with similar results. Individual bands on Western blots were quantified using AlphaEaseFC™ software (Alpha Innotech Corporation, San Leandro, CA, USA). All data were within the linear range of detection for each antibody used.

Generation of siRNA for MAPKAPK2 and HSP27

Total RNA from PC3 cells was first isolated with TRIzol^® Reagent (Invitrogen, Carlsbad, CA, USA; cat no. 15596-026), and reverse transcribed using SuperScript™ III Reverse Transcriptase (Invitrogen; cat. no. 18080-044), at 50°C for 60 min in a total volume of 20 l, in the presence of oligo d(T) primers, as described previously (Schwartz et al., 1998). Next, gene-specific dsDNA products that could support in vitro transcription were generated by polymerase chain reaction (PCR). For PCR reactions, the following primers were used: MAPKAPK2 forward primer – 5'-GCGTAATACGACTCACTATAGGGAGAAGGAGACCAGGCATTCACAG-3' and reverse primer – 5'-GCGTAATACGACTCACTATAGGGAGATGCAAAATGAGCACAGGTTC-3', yielding a 989 bp product; HSP27 forward primer – 5'-GCGTAATACGACTCACTATAGGGAGAAGAGCAGAGTCAGCCAGCAT-3'and reverse primer – 5'-CGTAATACGACTCACTATAGGGAGATTACAGGTGGTTGCTTTGAAC-3', yielding a 756 bp product. GFP was amplified from GFP plasmid, using primers provided in the Dicer siRNA generation kit (Gene Therapy Systems, San Diego, CA, UDS; cat. no. T510001), yielding a 740 bp product. A SuperMix High Fidelity kit (Invitrogen, Carlsbad, CA, USA; cat. no. 10790-020) was used to perform PCR reactions. Reaction conditions were, 94°C for 3 min, then 35 cycles of 94°C for 30 s 60°C for 30 s, and 68°C for 50 s, followed by 72°C for 10 min. All PCR products were analysed by agarose electrophoresis, and identity confirmed by sequencing, before being purified using QIAquick PCR Purification Kit (Qiagen, Valencia, CA, USA; cat. no. 28104).

In vitro transcription was performed as per the manufacturer's instructions in the Dicer siRNA generation kit (Gene Therapy Systems, San Diego, CA, USA; cat. no. T510001). Briefly, in vitro transcription reactions were performed in a volume of 20 l, using 1 g PCR product as a template, and were incubated at 37°C for 4 h, followed by purification with Qiagen RNeasy RNA kit (Qiagen; cat. no. 74104). Finally, the Dicer reaction was performed in a reaction volume of 40 l, using 10 g dsRNA, and was incubated 18 h at 37°C. Resultant siRNA were then purified, quantified and transfected according to the manufacture's instruction.

Zymography

Zymography assays were performed as described previously, with modifications (Huang et al., 2005). Briefly, 24 h after plating, cells were transfected with the indicated expression vector. One day later, cells were washed three times with serum-free media, changed into serum-free media, treated as indicated, and cultured for an additional 24 h. Conditioned media were collected, centrifuged at 3000 r.p.m. for 10 min to remove debris and concentrated by placing in a Microcon YM-30 Centrifugal Filter (Millipore, Billerica, MA, USA), and spinning at 14 000 g for 12 min. Concentrated media were separated by mixing with 2 sample dilution buffer (125 mM Tris, pH 6.8, 1% SDS, 0.002% bromophenol blue, 10% glycerol), incubating 15 min at room temperature, and then running on a 9% SDS-PAGE containing 1 mg/ml gelatin, under non-reducing conditions. Gels were washed with 2.5% Triton X-100 for 30 min, and incubated for up to 48 h at 37°C in 20 mM glycine, pH 8.3, 10 mM CaCl₂ and 1 M ZnCI₂. Gels were then stained with 0.5% Coomassie Brilliant Blue G solution, containing 10% acetic acid and 20% methanol, for 30 min, destained with 10% acetic acid and 20% methanol, and visualized.

Cell invasion assays

Cell invasion assays were performed as described previously, with modifications (Liu et al., 2002; Huang et al., 2005). Briefly, 24 h after plating, cells were transfected with the indicated expression vector, along with a constitutively active -galactosidase expression vector, thus allowing detection of transfected cells. One day later, cells were treated as indicated, detached by treatment with trypsin/EDTA, washed, resuspended in RPMI 1640 cell culture media (Gibco, Grand Island, NY, USA) with 0.1% BSA and 52 l of cell suspension were placed into the upper chamber of a 48-well Boyden chamber unit (i.e., 1 10⁴ cells/well). Cells were allowed to migrate for 15 h through a Nuclepore Track-Etch Membrane (NC 983-1643; Whatman, Clifton, NJ, USA), which contained 8 m pores and was coated with gelatin, towards serum-free NIH-3T3 conditioned medium, present in the lower chamber.

Membranes were fixed with 2% formaldehyde and 0.2% glutaraldehyde in phospate-buffered saline (PBS) for 10 min at room temperature, rinsed with PBS and -galactosidase was detected using a -galactosidase staining kit from Stratagene (cat no. 200384), according to the manufacturer's instructions. Cells were then stained with solution no. 1 (i.e., xanthane dye) using Diff-Quick cell staining kit (Dade Beharing AG; Dudingen, Switzerland). Membranes were mounted onto slides, using Permount™ (Fisher Scientific, Hampton, NH, USA). Using predetermined field coordinates, the number of invading and non-invading cells were then counted in each of five high-power fields (i.e. 100 ) for a given well, 4 wells for each treatment condition (N=4). All statistical tests of invasion were two-sided, and changes were only considered statistically significant for t-test P-values of 0.05 or less. All experiments were repeated at least once, at a separate time, in replicates (N=4).

References

1. Aldrian S, Trautinger F, Frohlich I, Berger W, Micksche M, Kindas-Mugge I. (2002). Cell Stress Chaperones 7: 177–185. | Article | PubMed | ISI | ChemPort |
2. Attisano L, Wrana JL. (2002). Science 296: 1646–1647. | Article | PubMed | ISI | ChemPort |
3. Carroll PR, Lee KL, Fuks ZY, Kantoff PW. (2001). In: DeVita VT, Hellman S, Rosenberg SA (eds). CANCER: Principals and Practices of Oncology. Lippincott-Raven: New York, pp 1418–1479.
4. Cornford PA, Dodson AR, Parsons KF, Desmond AD, Woolfenden A, Fordham M et al. (2000). Cancer Res 60: 7099–7105. | PubMed | ISI | ChemPort |
5. Gerthoffer WT, Gunst SJ. (2001). J Appl Physiol 91: 963–972. | PubMed | ISI | ChemPort |
6. Guay J, Lambert H, Gingras-Breton G, Lavoie JN, Huot J, Landry J. (1997). J Cell Sci 110(Part 3): 357–368. | PubMed | ISI | ChemPort |
7. Hansen RK, Parra I, Hilsenbeck SG, Himelstein B, Fuqua SA. (2001). Biochem Biophys Res Commun 282: 186–193. | Article | PubMed | ISI | ChemPort |
8. Hatakeyama D, Kozawa O, Niwa M, Matsuno H, Ito H, Kato K et al. (2002). Biochim Biophys Acta 1589: 15–30. | PubMed | ISI | ChemPort |
9. Hayes SA, Huang X, Kambhampati S, Platanias LC, Bergan RC. (2003). Oncogene 22: 4841–4850. | Article | PubMed | ISI | ChemPort |
10. Hedges JC, Dechert MA, Yamboliev IA, Martin JL, Hickey E, Weber LA et al. (1999). J Biol Chem 274: 24211–24219. | Article | PubMed | ISI | ChemPort |
11. Huang X, Chen S, Xu L, Liu YQ, Deb DK, Platanias LC et al. (2005). Cancer Res 65: 3470–3478. | PubMed | ISI | ChemPort |
12. Jemal A, Murray T, Ward E, Samuels A, Tiwari RC, Ghafoor A et al. (2005). CA Cancer J Clin 55: 10–30. | PubMed | ISI |
13. Kotlyarov A, Yannoni Y, Fritz S, Laass K, Telliez JB, Pitman D et al. (2002). Mol Cell Biol 22: 4827–4835. | Article | PubMed | ISI | ChemPort |
14. Krump E, Sanghera JS, Pelech SL, Furuya W, Grinstein S. (1997). J Biol Chem 272: 937–944. | Article | PubMed | ISI | ChemPort |
15. Liu Y, Jovanovic B, Pins M, Lee C, Bergan RC. (2002). Oncogene 21: 8272–8281. | Article | PubMed | ISI | ChemPort |
16. Liu YQ, Kyle E, Patel S, Housseau F, Hakim F, Lieberman R et al. (2001). Prostate Cancer Prostatic Dis 4: 81–91. | Article | PubMed | ISI | ChemPort |
17. Liu YU, Kyle E, Lieberman R, Crowell J, Kelloff G, Bergan RC. (2000). Clin Exp Metast 18: 203–212. | Article | ISI | ChemPort |
18. Ludwig S, Engel K, Hoffmeyer A, Sithanandam G, Neufeld B, Palm D et al. (1996). Mol Cell Biol 16: 6687–6697. | PubMed | ISI | ChemPort |
19. Munshi HG, Wu YI, Mukhopadhyay S, Ottaviano AJ, Sassano A, Koblinski JE et al. (2004). J Biol Chem 279: 39042–39050. | Article | PubMed | ISI | ChemPort |
20. Platanias LC. (2003). Blood 101: 4667–4679. | Article | PubMed | ISI | ChemPort |
21. Ravanti L, Hakkinen L, Larjava H, Saarialho-Kere U, Foschi M, Han J et al. (1999). J Biol Chem 274: 37292–37300. | Article | PubMed | ISI | ChemPort |
22. Rocchi P, So A, Kojima S, Signaevsky M, Beraldi E, Fazli L et al. (2004). Cancer Res 64: 6595–6602. | Article | PubMed | ISI | ChemPort |
23. Ruoslahti E. (1996). Sci Am 275: 72–77. | PubMed | ISI | ChemPort |
24. Schafer C, Clapp P, Welsh MJ, Benndorf R, Williams JA. (1999). Am J Physiol 277: C1032–C1043. | PubMed | ISI | ChemPort |
25. Schwartz GN, Liu YQ, Tisdale J, Walshe K, Fowler D, Gress R et al. (1998). Antisense Nucleic Acid Drug Dev 8: 329–339. | PubMed | ISI | ChemPort |
26. Shi Y, Massague J. (2003). Cell 113: 685–700. | Article | PubMed | ISI | ChemPort |
27. Sporn MB. (1996). Lancet 347: 1377–1381. | Article | PubMed | ISI | ChemPort |
28. Stokoe D, Campbell DG, Nakielny S, Hidaka H, Leevers SJ, Marshall C et al. (1992a). EMBO J 11: 3985–3994. | PubMed | ISI | ChemPort |
29. Stokoe D, Engel K, Campbell DG, Cohen P, Gaestel M. (1992b). FEBS Lett 313: 307–313. | Article | PubMed | ISI | ChemPort |
30. Tong L, Pav S, White DM, Rogers S, Crane KM, Cywin CL et al. (1997). Nat Struct Biol 4: 311–316. | Article | PubMed | ISI | ChemPort |
31. Wilson KP, McCaffrey PG, Hsiao K, Pazhanisamy S, Galullo V, Bemis GW et al. (1997). Chem Biol 4: 423–431. | Article | PubMed | ISI | ChemPort |
32. Winzen R, Kracht M, Ritter B, Wilhelm A, Chen CY, Shyu AB et al. (1999). EMBO J 18: 4969–4980. | Article | PubMed | ISI | ChemPort |
33. Woodhouse EC, Chuaqui RF, Liotta LA. (1997). Cancer 80: 1529–1537. | Article | PubMed | ISI | ChemPort |

Acknowledgements

This work was funded by the following grants to Raymond C Bergan: Specialized Program of Research Excellence (SPORE) Grant CA90386, from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, and by a merit review award from the Veterans Administration. We wish to thank Professor Matthias Gaestel (Institute of Biochemistry, Medical School, Hannover, Germany) for kindly providing us plasmids for wild-type of MAPKAPK2, pcDNA3mycMK2WT, dominant-negative kinase-inactive mutant, pcDNA3mycMK2K76R, and constitutively active mutant, pcDNA3mycMK2T205E317E. We also wish to thank Professor Rainer R Benndorf (Departments of Cell and Developmental Biology, University of Michigan Medical School, Ann Arbor, MI, USA) for kindly providing us plasmids for the wild type of HSP27, pcDNA3.1-HSP27 WT and non-phosphorylateable mutant of HSP27, pcDNA3.1-HSP27 3G.