CUTL1 promotes tumor cell migration by decreasing proteasome-mediated Src degradation

Abstract

Recently, we identified the homeodomain transcription factor CUTL1 as important mediator of cell migration and tumor invasion downstream of transforming growth factor β (TGFβ). The molecular mechanisms and effectors mediating the pro-migratory and pro-invasive phenotype induced by CUTL1 have not been elucidated so far. Therefore, the aim of this study was to identify signaling pathways downstream of CUTL1 which are responsible for its effects on tumor cell migration. We found that the reduced motility seen after knock down of CUTL1 by RNA interference is accompanied by a delay in tumor cell spreading. This spreading defect is paralleled by a marked reduction of Src protein levels. We show that CUTL1 leads to Src protein stabilization and activation of Src-regulated downstream signaling molecules such as RhoA, Rac1, Cdc42 and ROCK. In addition, we demonstrate that CUTL1 decreases proteasome-mediated Src protein degradation, possibly via transcriptionally upregulating C-terminal Src kinase (Csk). Based on experiments using Src knockout cells (SYF), we present evidence that Src plays a crucial role in CUTL1-induced tumor cell migration. In conclusion, our findings linking the pro-invasive transcription factor CUTL1 and the Src pathway provide important new insights in the molecular effector pathways mediating CUTL-induced migration and invasion.

Introduction

CUTL1, also known as CCAAT displacement protein (CDP) or Cux-1, is a homeobox transcription factor involved in the control of embryonic development and differentiation (Nepveu, 2001). Recently, our group has shown that CUTL1 is highly expressed in various epithelial cancers and its expression seems to be negatively correlated with tumor differentiation and patient survival (Michl et al., 2005). In addition, we characterized CUTL1 as a transcriptional target of transforming growth factor β (TGFβ) and as an important mediator of the TGFβ-induced cell migration and invasion (Michl et al., 2005). So far, the molecular mechanisms and effectors mediating the pro-migratory and pro-invasive phenotype induced by CUTL1 remain to be elucidated.

Cell migration is a critical feature of numerous physiological and pathological phenomena, including development, wound repair, angiogenesis and metastasis (Price et al., 1997). The ability of a cell to migrate on extracellular matrix (ECM) substrates depends on adhesion receptors known as ‘integrins’ (Hynes, 1992; Hynes and Lander, 1992). Upon activation of integrins by ECM proteins, Src family protein tyrosine kinases such as Src, Yes and Fyn (SFKs) get activated and localize to focal adhesions (Burridge and Chrzanowska-Wodnicka, 1996; Lowell and Soriano, 1996; Parsons, 1996; Parsons and Parsons, 1997; Thomas and Brugge, 1997). Cells lacking the Src family protein tyrosine kinases Src, Yes and Fyn (SYF triple mutant cells) show reduced motility and reduced spreading in vitro, indicating that Src, Yes and Fyn kinases play an important role as mediators of pro-migratory signals initiated by ECM proteins (Klinghoffer et al., 1999).

Under basal conditions, 90–95% of Src is phosphorylated by the C-terminal src kinase (Csk) at the C-terminal Tyr527 in mice or Tyr530 in humans (Zheng et al., 2000). Through phosphorylation at Tyr527, Csk stabilizes Src in a dormant form with low basal activity (Roskoski, 2004). Csk, however, is not able to prevent Src activation through conformational changes on external stimuli such as integrin activation by ECM proteins (Moarefi et al., 1997; Ma et al., 2000).

Based on our previous findings which identified the transcription factor CUTL1 as important mediator of cell migration and tumor invasion downstream of TGFβ1, we aimed to elucidate the molecular mechanisms and effectors which are crucial for CUTL1-induced tumor cell migration. In this study, we show that CUTL1 upregulates Csk mRNA and protein expression and decreases proteasome-mediated Src degradation thereby stabilizing Src protein levels. Elevated Src levels induced by CUTL1 led to markedly enhanced migration upon activation by ECM stimuli.

Results

Generation of CUTL1 knockdown cells

The mesenchymal HT1080 fibrosarcoma cells as well as the epithelial pancreatic carcinoma cell line Panc1 were stably transfected with CUTL1 short hairpin RNA (shRNA), thereby obtaining clones with a constitutive reduction in CUTL1 protein levels of approximately 40–60%. The two clones which showed the best CUTL1 reduction were used for our experiments. To ensure that our results were not biased by clonal selection, most experiments were confirmed by transient assays using two different silencing sequences of CUTL1 small interfering RNA.

CUTL1 modulates cell spreading

To investigate whether CUTL1 influences the adhesive properties of tumor cells, we first compared the ability of CUTL1 knockdown and control cells to adhere to different components of ECM such as fibronectin, collagen I, collagen IV, vitronectin or laminin. Both cell lines show the strongest affinities for fibronectin as determined by adhesion assays. As adhesion was not significantly altered by the presence or absence of CUTL1 (data not shown), we subsequently evaluated cell spreading differences of CUTL1 knockdown and control cells on fibronectin-coated dishes by time-lapse microscopy. As demonstrated in Figure 1a for PANC1 cells, CUTL1 knockdown cells show a significant spreading deficit on fibronectin, with CUTL1 knockdown cells spreading up to 67% less than control cells 8 min after seeding (Figure 1b). Similar results were obtained for HT1080 cells (data not shown). The differences in spreading between CUTL1 knockdown cells and control cells persisted up to 20 min after seeding. However, 120 min after seeding, the differences in spreading diminished, indicating that CUTL1 knock down induces a spreading delay rather than an absolute spreading defect.

Figure 1
figure1

CUTL1 modulates cell spreading. (a) Representative pictures derived from time-lapse microscopy of stable CUTL1 knockdown PANC1 cells (CUTL1, lower panel) versus control cells (Control, upper panel) on fibronectin-coated slides 0, 8 and 16 min after seeding. Note a significant spreading deficit of CUTL1 cells as soon as 8 min after seeding. (b) Quantification of the differences in cell spreading 8 min after seeding, as shown in Figure 1a. *Indicates P<0.05 as compared with control cells, as assessed by double-sided unpaired t-test. Data are representative for three independent experiments. The lower panel shows the level of CUTL1 knockdown in stable CUTL1 shRNA and control Panc1 cells, as assessed by Western blot with a specific antibody. To control for equal loading the blots were reprobed with β-actin antibody. (c) Representative pictures derived from time-lapse microscopy of Panc1 cells after transient knockdown of CUTL1 by Cy3-labeled siRNA (CUTL1, right panel) versus Cy3-labeled nonsilencing RNAi in control cells (Control, left panel) on fibronectin-coated slides. Cell spreading was depicted by DIC (upper panel) 16 min after seeding, transfection efficiency with Cy3-siRNA's was confirmed by fluorescence microscopy (lower panel). Note a significant spreading deficit of CUTL1 cells.

To ensure that the differences in spreading were not owing to clonal biases, we transiently transfected Panc1 cells with Cy3-labeled CUTL1 siRNA or Cy3-labeled nonsilencing control siRNA. Forty-eight hours after transfection, we performed spreading experiments on fibronectin-coated dishes, as described above. Fluorescence microscopy indicated that the majority of cells was successfully transfected with siRNA (Figure 1c, lower panel), and CUTL1 knockdown was verified by Western blot (data not shown). Using this transient approach, we could confirm a markedly decreased spreading in CUTL1 knockdown cells 16 min after plating (Figure 1c, upper panel).

CUTL1 modulates Src protein levels

To investigate whether the spreading defects observed in CUTL1 knockdown cells are due to alterations in the integrin profile, we first tested the effect of CUTL1 on the expression of several alpha and beta subunits of integrins (alphaV, alpha1-6, beta1, beta3 and beta5) by immunoblotting and real-time reverse transcriptase-polymerase chain reaction (RT-PCR). We found that CUTL1 does not regulate the expression of the integrins tested (data not shown). Next, we analysed the regulation of promigratory signaling molecules known to act downstream of integrins, such as Src, small GTPases and ROCK.

Immunoblotting analysis of Src protein levels revealed markedly decreased Src protein levels after CUTL1 knockdown (Figure 2a and b). Reduced Src protein levels were confirmed after both stable (Figure 2a, left panels) and transient (Figure 2a, right panels) suppression of CUTL1 in both Panc1 and HT1080 cells. The decrease in total Src levels was paralleled by a decrease in Src phosphorylated at Tyr416, a residue phosphorylated in activated Src (Figure 2a, left lower panels). In addition to Src downregulation after CUTL1 knockdown, we were able to show upregulation of Src protein levels upon CUTL1 overexpression (Figure 2c). To evaluate whether CUTL1 regulates src transcription, we analysed src mRNA levels in control versus CUTL1 knockdown cells by real-time PCR. However, our data did not show any transcriptional regulation of src mRNA by CUTL1 (data not shown). Thus, it appears that CUTL1 modulates Src protein levels, but independent of transcriptional regulation of Src mRNA.

Figure 2
figure2

CUTL1 regulates Src protein levels. (a) Detection of Src by immunoblotting after stable (left, stable clones) and transient (right) CUTL1 suppression. Cell lysates obtained from CUTL1 knockdown cells (CUTL1) and control cells (C) were analysed for Src protein expression by immunoblot analysis using an anti-Src antibody. Note that both stable and transient CUTL1 suppression in Panc1 and HT1080 cells leads to reduced Src protein levels. To control for equal loading the blots were reprobed with β-actin antibody (middle panels). The left lower panels show phospho-Src levels (Tyr416) in both Panc1 and HT1080 CUTL1 knockdown (CUTL1) and control cells (C). The upper band of the phospho-Src blot shows the levels of phosphorylated Src, the lower band most likely represents phospho-Fyn. To control for equal loading the blots were reprobed with β-actin antibody (lower left panels). Data are representative for three independent experiments. (b) Semiquantitative densitometry of the immunoblots of stable Panc1 and HT1080 clones shown in Figure 2a. Three immunoblots each were scanned, analysed by semiquantitative densitometry and normalized to β-actin expression. Data are shown as mean±s.e.m. (c) Detection of Src by immunoblotting after transient CUTL1 upregulation. Panc1 and HT1080 cells were transiently transfected with wild-type CUTL1 plasmid (+CUTL1) or with pcDNA3 (c). CUTL1 and Src protein levels were analysed by immunoblots using anti-CUTL1 or anti-Src antibodies. The lower panel shows β-actin as loading control. Data are representative for three independent experiments.

CUTL1 modulates the activities of small GTPases (RhoA, Rac1 and Cdc42) and of ROCK

To investigate the functional relevance of the CUTL1 effect on Src protein levels, we evaluated the activity of promigratory signaling cascades acting downstream of Src 20 min after plating cells on fibronectin-coated dishes. The activity of the Rho-family of small GTPases was investigated by pull-down analyses, which measure the amount of active, GTP-bound of RhoA, Cdc42 or Rac1 (Figure 3a). Our results show that RhoA, Cdc42 and Rac1 activity is dependent on the presence of CUTL1 in Panc1 (Figure 3a) and HT1080 cell lines (data not shown). Further downstream, RhoA activation is known to stimulate actin-myosin contractility through activation of the Rho-effector ROCK (Amano et al., 1996; Riento and Ridley, 2003). ROCK activity was also enhanced in the presence of CUTL1, as assessed by an in vitro kinase assay using MYPT1 as a substrate for immunoprecipitated ROCK (Figure 3b). These results suggest an important role for Src and its targets RhoA, cdc42, Rac1 and ROCK in CUTL1-regulated cell migration.

Figure 3
figure3

CUTL1-induced Src regulation is associated with activation of Rho GTPases and ROCK. (a) Activity assays of small Rho-GTPases. CUTL1 stable knockdown (CUTL1) and control Panc1 cells (C) were lysed in the appropriate buffer and lysates were subjected to affinity precipitation assays to determine RhoA (RhoA-GTP), Cdc42 (cdc42-GTP) and Rac1 (Rac1-GTP) activity in the presence of GST-Rhotekin (RhoA) or GST-PAK (Cdc42 and Rac1). Equal protein loading was confirmed by immunoblotting analysis of total cell extracts using antibodies against total RhoA, Rac1, or Cdc42. Similar results were obtained in HT1080 cells. Data are representative for three independent experiments. (b) Immuncomplex kinase assay of ROCK activity. Activity of ROCK was assayed in CUTL1 knockdown (CUTL1) versus control cells (C) by an immune complex kinase assay using MYPT1 as substrate. The upper panel indicates the activated MYPT1 substrate visualized by autoradiography (*pMYPT1). The lower panel shows the blot reprobed with anti-ROCK2 antibody to assess the efficiency of immunoprecipitation used as a loading control (Rock IP/IB). Data are representative for three independent experiments.

CUTL1 regulates Src protein levels by enhancing its stability

As Src is not transcriptionally regulated by CUTL1, we aimed to elucidate whether Src protein levels may be regulated by post-translational modifications. It has already been shown that Src tyrosine kinase can be degraded by the ATP-dependent 26S proteasome, which represents a common way of protein degradation following ubiquitinylation (Hakak and Martin, 1999). Therefore, we investigated if the reduced Src protein levels in CUTL1 knockdown cells are due to its faster degradation via the proteasome. To block the activity of the proteasome complex, we used the proteasome inhibitor MG-132 (Hakak and Martin, 1999) and analysed Src protein levels in control versus CUTL1 knockdown cells in both Panc1 (Figure 4a) and HT1080 (Figure 4b) cell lines. In both cells lines, the results demonstrate that Src proteins after proteasome inhibition accumulated in CUTL1 knockdown cells reaching similar levels as in control cells. Accumulation started as early as 1 h after proteasome inhibition, reaching control Src protein levels after 3 h of MG-132 treatment (see Figure 4). Immunoblotting for β-actin showed that this was not due to a general protein accumulation (Figure 4a and b, lower panels). These data showing enhanced re-accumulation of Src after proteasome-inhibition in CUTL1 knockdown cells suggest that the activity of the transcription factor CUTL1 is associated with reduced proteasome-mediated degradation of the Src protein.

Figure 4
figure4

CUTL1 regulates Src protein levels by enhancing its stability. Panc1 (a) or HT1080 (b) CUTL1 stable knockdown cells (CUTL1) and control cells (C) were treated with 40 μ M of the proteasome inhibitor MG-132 for 1, 2 and 3 h or with ethanol (EtOH) as a control. Aliquots of total lysates were analysed by immunoblotting with anti-Src antibody. Note that proteasome inhibition leads to faster accumulation of Src protein in CUTL1 knockdown cells. The blot was reprobed with β-actin antibody to control for equal loading (lower panel). Data are representative for three independent experiments. Semiquantitative densitometry of the immunoblots are shown at the bottom part. Three immunoblots each were scanned, analysed by semiquantitative densitometry and normalized to β-actin expression. Data are shown as mean±s.e.m.

Re-expression of Src in CUTL1-knockdown cells restores spreading and migration

To explore whether defects in spreading and migration observed in HT1080 and Panc1 CUTL1 knockdown cells could be rescued be re-expressing Src, we first stably transfected CUTL1 knockdown clones with wild-type Src and tested the spreading and migratory activities of three resulting double transgenic clones (CUTL1 src+) of each cell line. As shown in (Figure 5a and b), re-expression of Src into CUTL1 knockdown cells significantly restores the cell spreading in Panc1 cells. In addition, re-expression of Src significantly re-enhances the impaired migration in Panc1 cells with CUTL1 knockdown (Figure 5c). Similar results were obtained with HT1080 cells (not shown).

Figure 5
figure5

Re-expression of Src into CUTL1 knockdown cells restores spreading and migration. (a) Panc1 CUTL1 knockdown cells stably transfected with wild-type Src (CUTL1/Src) or pcDNA3.1 (CUTL1/mock) and Panc1 control cells transfected with pcDNA3.1 (Control/mock) were plated on fibronectin-coated slides. Representative pictures derived from time-lapse microscopy are shown 30 min after seeding. Results are representative for two independent clones each. (b) Percentage of spread cells 30 min after seeding on fibronectin-coated dishes, as shown in Figure 1a and as counted in five independent visual fields. *Indicates P<0.05 as compared with CUTL cells, as assessed by double-sided unpaired t-test. Results are representative for two independent clones each. (c) Panc1 CUTL1 knockdown cells stably transfected with wild-type Src (CUTL1/Src) or pcDNA3.1 vector (CUTL1/mock) and Panc1 control cells transfected with pcDNA3.1 (Control/mock) were assaysed for migration differences. Migration of six double-transfected clones (CUTL1Src+) and six control clones (C) was assayed by in vitro migration assays. After 4 h, the number of migrating cells was measured by CellTiter-Glo Luminescent Cell Viability Assay as described. The results represent average migration of six different clones±s.d. measured in triplicates. *Indicates P<0.05 as compared with CUTL1 cells, as assessed by double-sided unpaired t-test. Data are representative for three independent experiments. (d) Panc1 CUTL1 knockdown or control cells were transiently transfected with wild-type Src or pcDNA3.1 and LacZ vectors. In vitro migration assays were performed by using Boyden-chamber inserts. Cells were fixed after 8 h of migration, stained with X-Gal and migrated transfected cells (blue cells) was counted in five independent visual fields. The results were normalized to the total number of transfected cells. Data are shown as mean±s.d. and are representative for three independent experiments. *Indicates P<0.05 as compared with CUTL cells, as assessed by double-sided unpaired t-test. The knockdown of CUTL1 and the over-expression of Src was verified by Western blot. Equal loading was confirmed with a β-actin antibody (lower panels).

To ensure that the differences in spreading were not due to clonal biases in the Src-overexpressing Panc1 clones, we transiently overexpressed Src in CUTL1 knockdown and control Panc1 cells. Transient Src overexpression in CUTL1 knockdown cells significantly increased cell migration in two-chamber assays (Figure 5d). Interestingly, transient Src-overexpression in control cells enhanced migration only to a lesser extent, which did not reach significance (Figure 5d). These data indicate that Src plays a central role in CUTL1-induced migration.

Csk is transcriptionally regulated by CUTL1 in several cell lines

Csk is able to phosphorylate human Src at Tyr530 which is located in the C-terminal tail of Src (Zheng et al., 2000). More than 95% of cellular Src is phosphorylated at the Csk-specific tyrosine residue 530 representing an important mechanism in the regulation of basal Src activity (Zheng et al., 2000). When phosphorylated at Tyr530, the activity of Src is reduced to low basal levels, thereby protecting cells from the potential oncogenic activity of the src protooncogene (Zheng et al., 2000). The activity of Src phosphorylated at the C-terminal Tyr530 can be increased either by dephosphorylation of Tyr530 or by phosphorylation at other tyrosine residues, for example, after the stimulation of integrin receptors with fibronectin (Moarefi et al., 1997; Ma et al., 2000). It has been shown that Src shows significantly greater protein stability when phosphorylated by Csk (Roskoski, 2004). To investigate if Csk is a transcriptional target of CUTL1 thereby mediating its effects on Src stability, we analysed the level of csk mRNA expression after transient CUTL1 suppression by siRNA (Figure 6a). Real-time PCR shows in both HT1080 and Panc1 cells that reduced CUTL1 expression leads to significantly reduced csk mRNA transcription (Figure 6a). The transcriptional regulation of Csk by CUTL1 is also seen on protein level, as demonstrated by immunoblotting analyses for Csk after transient down or upregulation of CUTL1 (Figure 6b and c, respectively). This clearly demonstrates transcriptional upregulation of Csk by CUTL1. Moreover, the phosphorylated form of Src at Tyr530, the known phosphorylation site of Csk, was decreased upon reduction of CUTL1 (Figure 6d), indicating that CUTL1 increases Src protein stability via upregulation of Csk, which in turn phosphorylates and stabilizes Src.

Figure 6
figure6

CUTL1 transcriptionally regulates Csk. (a) Quantitative real-time PCR of Csk mRNA expression after transient CUTL1 suppression. mRNAs isolated from control (C) versus CUTL1 knockdown cells (CUTL1) from Panc1 and HT1080 cells were subjected to real-time PCR analyses to assess the levels of Csk mRNA expression after CUTL1 suppression. Data represent relative values, normalized to the levels of Cyclophilin mRNA. Assays were performed in triplicate and are representative for three independent experiments and were performed using two different silencing sequences. *Indicates P<0.05 as compared with control cells, as assessed by double-sided unpaired t-test. (b) Detection of Csk protein expression after transient downregulation of CUTL1. Cell lysates obtained from control cells (C) or cells with transiently downregulated CUTL1 levels (CUTL1) in HT1080 and Panc1 cells were tested for Csk protein levels by immunoblots with the anti-Csk antibody. The blot was reprobed with β-actin antibody to control for equal loading. Data are representative for three independent experiments. (c) Csk immunoblotting after transient up-regulation of CUTL1. HT1080 and Panc1 cells were transiently transfected with CUTL1 plasmid (+CUTL1) or with pcDNA (c). The levels of Csk protein were assayed by anti-Csk immunoblotting. The blot was reprobed with β-actin antibody as loading control (lower panel). Data are representative for three independent experiments. (d) Detection of Src phosphorylation at Tyr530 in stable CUTL1 knockdown (CUTL1) or control HT1080 cells or cells with overexpression of CUTL1 after transient transfection (+CUTL). Cell lysates were tested for phospho-Tyr530 Src protein levels by immunoblots with the anti-phospho-Tyr530 Src antibody. Levels of CUTL1 and total Src were detected by the respective antibodies. The blot was reprobed with β-actin antibody to control for equal loading. Data are representative for three independent experiments. (e) Panc1 CUTL1 knockdown cells were transfected with Csk or pcDNA3.1 and LacZ vectors. In vitro migration assays were performed by using Boyden-chamber inserts. Cells were fixed after 8 h of migration, stained with X-Gal and migrated transfected cells (blue cells) was counted in five independent visual fields. The results were normalized to the total number of transfected cells. Data are shown as mean±s.d. and are representative for three independent experiments. *Indicates P<0.05 as compared with CUTL cells, as assessed by double-sided unpaired t-test. The knockdown of CUTL1 and the overexpression of Csk was verified by Western blot. Equal loading was confirmed with β-actin antibody (lower panels).

Overexpression of Csk enhances cell migration depending on the presence of CUTL1

To assess the physiological significance of Csk as modulator of migration in our cell systems, we transiently overexpressed Csk in Panc1 cells with stable knockdown of CUTL1 and in mock-transfected Panc1 cells. In CUTL1 knockdown cells with low endogenous Csk, expression of exogenous Csk significantly increased migration (Figure 6e). In contrast, the effect of exogenous Csk in mock-transfected cells with high endogenous CUTL1 levels was markedly less (Figure 6e) These data suggest that Csk enhances migration up to a certain expression level which is – at least in part – regulated by CUTL1. Supraphysiological expression levels, however, are not able to further enhance the effect on migration.

Src is a crucial mediator of CUTL1-induced migration

Both Src and Csk are known as important mediators of cell migration and actin cytoskeletal reorganization. Interestingly, Csk is able to regulate actin cytoskeleton organization both via src-dependent and src-independent pathways (Lowry et al., 2002). Therefore, we were interested whether CUTL1-induced cell migration is mainly the effect of Csk on Src protein stability, or whether Csk influences CUTL1-induced migration independently of Src. To address this issue, we used two different mouse embryonic cell lines derived from double or triple knockout mice: the src++ (yes fyn) cell line shows normal expression of endogenous Src but lacks Yes and Fyn. In contrast, the SYF (src yes fyn) cell line lacks endogenous Src, Yes and Fyn through triple knockout mutations (Klinghoffer et al., 1999). It has been shown that SYF cells show significantly reduced cell migration compared with src++ cells (Klinghoffer et al., 1999) indicating the important role of Src in migration. To assess the relevance of Src, we transiently upregulated CUTL1 in both SYF and src++ cell lines. Upregulation of CUTL1 in src++ cells leads to increased Src protein levels (Figure 7a). Moreover, CUTL1 upregulation also leads to increased Csk protein levels in both src++ and SYF cells (Figure 7b). Using this cell system, we analysed cell migration using a modified Boyden-chamber migration assay. As shown in Figure 7c, upregulation of CUTL1 in the absence of Src in SYF cells failed to enhance migration despite of increasing Csk levels. In contrast, CUTL1 significantly increased migration in src++ cells paralleled by an increase in Src protein levels (Figure 7c). These data identify Src as a crucial effector of CUTL1-induced migration. The fact that CUTL1 has no significant effects on migration in the absence of Src despite of a marked upregulation of Csk suggests that Csk induced by CUTL1 does not exert any significant Src-independent effects on migration (Figure 7c).

Figure 7
figure7

CUTL1 modulates migration in Src++ but not in SYF cells. (a) Upregulation of Src in src++ double knockout cells (Fyn, Yes) by CUTL1. Src++ cells which lack endogenous Yes and Fyn, were transiently transfected with CUTL1 plasmid (+CUTL1) or pcDNA (C), lysed 24 h after the transfection and assayed by immunoblotting using anti-Src antibody. The blot was reprobed with β-actin antibody as loading control (lower panel). Data are representative for three independent experiments. (b) Upregulation of Csk in src++ double knockout cells (Fyn, Yes) and SYF triple knockout (Fyn, Yes, Fyn) cells by CUTL1. Src++ and SYF cells were transiently transfected with CUTL1 plasmid (+CUTL1) or pcDNA (C), lysed 24 h after transfection and assayed by immunoblot using anti-Csk antibody. The blot was reprobed with β-actin antibody as loading control (lower panel). Data are representative for three independent experiments. (c) Boyden chamber migration assays of src++ (Yes, Fyn) and SYF (Src, Yes and Fyn) cells upon CUTL1 upregulation. Src++ and SYF cells were co-transfected with CUTL1 and LacZ plasmids (+CUTL1) or with pcDNA and LacZ plasmid (c) and in vitro migration assays were performed by using Boyden-chamber inserts. Cells were fixed after 2 h of migration, stained with X-Gal and migrated transfected cells (blue cells) was counted in four independent visual fields. The results were normalized to the total number of transfected cells. Data are shown as mean±s.d. and are representative for three independent experiments. *Indicates P<0.05 as compared with Src++ control cells, as assessed by double-sided unpaired t-test.

Discussion

The ability to migrate and invade into surrounding tissues, blood and lymphatic vessels is a prerequisite for local tumor progression and metastatic spread. We have recently shown that reduced expression of the transcription factor CUTL1 severely impairs the ability of tumor cells to migrate and invade (Michl et al., 2005). The molecular mechanisms and effectors mediating the promigratory and proinvasive phenotype induced by CUTL1 have not been elucidated so far. Therefore, the aim of this study was to identify signaling pathways downstream of CUTL1, which are responsible for its effects on the tumor cell migration.

First, we found that knockdown of CUTL1 results in significantly delayed spreading of tumor cells, which was particularly pronounced when cells were spread on the ECM compound fibronectin. Difficulties in cell spreading on ECM substrates might be caused by disturbed signaling pathways downstream of integrins. Among others, we investigated a possible involvement of the Src pathway, known as a central component in the regulation of cell motility. We found that CUTL1 transcriptionally activates Csk and increases the stability of Src protein. Our data identify Src as a crucial effector of CUTL1-promoted tumor cell migration. Furthermore, our experiments using SYF cells with knockout of Src indicate that Src-independent pathways do not appear to play a major role in CUTL1-induced tumor cell migration.

Most of our experiments are based on knockdown strategies, which avoid potential biases introduced by overexpression techniques. To control off-target effects of the RNA interference oligonucleotides, we used several strategies: first, we confirmed all our experiments that used stable clones with experiments using transient suppression of CUTL1. The transient experiments were performed with siRNA oligos using two distinct silencing sequences, which both were different from the silencing sequence used for generating the stable clones with vector-based shRNA. Therefore, we used a total of three different silencing sequences which all showed similar results. In addition, we had previously performed microarray experiments using stable clones transfected with the same vector-based shRNA as used in this manuscript (Michl et al., 2005). In this gene expression profile, we did not observe significant upregulation of interferon-response genes which are known to be typical ‘off target’ genes. Moreover, the sequence used for generating the stable RNAi clones did not have any perfect matches other than CUTL1 in the relevant mammalian genomes, and the genes showing partial matches (13–15/19) to the RNAi sequence by BLAST search did not show any suppression of expression in the microarray experiments. Likewise, the sequences used for transient knockdown strategies did not have any match other than CUTL1 in relevant mammalian genomes.

Before migration, cells adhere and spread on numerous components of ECM such as fibronectin (Hynes and Lander, 1992; Parsons and Parsons, 1997; Thomas and Brugge, 1997). As our aim was to elucidate effector pathways downstream CUTL1, our first step was to investigate whether the expression of CUTL1 leads to alterations in cell adhesion or spreading on ECM substrates. Our data indicate that CUTL1 activity is strongly associated with enhanced cell spreading. As alterations in the expression profile of integrins have been implicated in reduced cell spreading (Kaido et al., 2004), we analysed possible effects of CUTL1 on the expression of various integrins. Our findings indicate, however, that CUTL1 does not modulate the expression of integrins.

Previous studies have demonstrated that adhesion of cells to fibronectin leads to translocation of the Src-family of tyrosine kinases such as Src, Yes and Fyn (SFKs) to the site of integrins (Parsons and Parsons, 1997). Cells lacking Src, Yes and Fyn (SYF triple-mutant cells) show reduced motility and reduced spreading in vitro (Lowell and Soriano, 1996; Parsons and Parsons, 1997; Thomas and Brugge, 1997; Klinghoffer et al., 1999). Our results suggest that Src has a crucial role in CUTL1-induced tumor cell migration: migration of SYF triple-mutant cells, which lack Src, Yes and Fyn was not noticeably affected by CUTL1 upregulation. In contrast, the migration of src++ cells, which express Src, but still lack Yes and Fyn, was significantly enhanced upon CUTL1 upregulation. These results, along with the fact that stable or transient Src re-expression into CUTL1 knockdown cells restored cell migration to a significant extent support the important role of Src in CUTL1-induced migration.

It has already been reported that Src can be degraded via the proteasome-ubiquitin-dependent pathway (Hakak and Martin, 1999). As we did not observe a transcriptional regulation of Src mRNA by CUTL1, we investigated possible mechanisms of post-translational Src stabilization by CUTL1. Therefore, we analysed the effects of proteasome inhibition on Src protein levels in CUTL1 knockdown versus control cells. Our data indicate that depletion of CUTL1 leads to destabilization of Src protein and facilitates its degradation via the proteasome complex.

Csk is known to stabilize Src in a dormant form with low basal activity through phosphorylation at Tyr527 in mice or Tyr530 in humans (Roskoski, 2004). This stabilized form of Src shows low basal activity, but ECM components such as fibronectin may activate it by inducing conformational changes (Moarefi et al., 1997; Ma et al., 2000). Moreover, it has been described that Src achieves the conformation accessible for the activating phosphorylation by different signaling molecules, such as Cdc2, only after prior phosphorylation at the Csk-specific Tyr527 residue (Stover et al., 1994). According to our results, CUTL1 is able to transcriptionally upregulate the expression of Csk, which is paralleled by stabilization of Src protein levels. As it has been described that phosphorylation by Csk protects Src from being recognized by the ubiquitination machinery (Hakak and Martin, 1999), we postulate that transcriptional upregulation of Csk by CUTL1 may stabilize Src by decreasing its ubiquitinylation.

Our data proposing Src stabilization as new effector pathway of CUTL1, which mediates tumor invasion, are in accordance with a recent report demonstrating that Src protein upregulation and activation by ErbB2 is due to increased Src protein stability (Tan et al., 2005). Stabilizing Src protein levels might therefore represent a novel mechanism by which tumor-promoting signaling cascades exert their effect on tumor cell migration and invasion. In contrast to ErbB2, which increased Src stability mainly through the inhibition of calpain protease activity (Tan et al., 2005), we show that in our cell systems the proteasome activity is essential for the CUTL1-dependent modulation of Src protein levels. In addition to decreased Src protein levels after CUTL1 knockdown, the activity of known Src-regulated downstream effectors such as small Rho GTPases (Rho, Rac and Cdc42) and ROCK is markedly impaired, culminating in reduced cell spreading and motility. The fact that the Src-targets Rho GTPases and ROCK activities are reduced upon knockdown of CUTL1, is suggestive that CUTL1 mediates its effects on the Rho GTPases and ROCK via Src, but further experiments are necessary to confirm this interaction and Src-independent effects of CUTL1 on Rho GTPases and ROCK cannot be ruled out so far.

Previously, we found that the transcription factor CUTL1 is highly expressed in a variety of epithelial cancers and is associated with high tumor grade and poor patient survival (Michl et al., 2005). However, direct therapeutic targeting of tumor-promoting transcription factors, for example, by small molecule inhibitors is difficult. Our findings linking the proinvasive transcription factor CUTL1 and the Src pathway provide not only important new insights in the molecular effector pathways mediating CUTL-induced migration and invasion but also suggest that CUTL1-positive tumors such as highly invasive pancreatic or breast cancers might be particularly susceptible to small molecule inhibitors targeting Src tyrosine kinases, which are currently in preclinical and clinical evaluation (Johnson et al., 2005). Further studies are warranted to elucidate the clinical significance of Src activation in CUTL1-positive tumors and its potential applicability for therapeutic targeting.

Materials and methods

Cell culture

The human fibrosarcoma cell line HT1080 and the human pancreatic carcinoma cell line Panc1 were maintained in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Paisley, Scotland) supplied with 10% fetal calf serum (FCS) (Invitrogen, Paisley, Scotland) and 1% streptomycin/penicillin (Sigma-Aldrich, St Louis, MO, USA) at 37°C in 7% CO2. Clones stably expressing Cutl1 shRNA were selected in growth medium supplemented with 2.5 μg/ml of Puromycin (Calbiochem, Merck Biosciences, Darmstadt, Gemany). SYF cells (mouse embryonic fibroblasts derived from triple knockout (KO) mice for Src, Yes and Fyn) and Src++ cells (mouse embryonic fibroblasts derived from double KO mice for Yes and Fyn) were obtained from ATCC (ATCC, Cat.No. 30-2002) and maintained in modified DMEM supplemented with 10% FCS and 1% streptomycin/penicillin at 37°C in 7% CO2.

In vitro migration assays

In vitro migration assays were performed in modified Boyden-chambers (8 μm pores, Falcon BD Biosciences, Bedford, MA, USA) essentially as described (Michl et al., 2005). In these assays, cells which migrated through the pores were counted under the microscope. To selectively assess the migration of transiently transfected cells, co-transfections with LacZ plasmid were performed. After 2 h (HT1080) or 8 h (Panc1), cells which had migrated through the pores were fixed and stained by X-Gal stain (Sigma, Poole, Dorset, England). Migration was calculated by counting blue-stained cells. The number of migrated cells was normalized to the total number of transfected cells in a adjacent well.

Migration of stably transfected clones in Boyden-chamber assays was measured by the CellTiter-Glo Luminescent Cell Viability Assay (Promega, Madison, WI, USA) that determines the relative number of cells, which migrated through the pores by measuring the amount of ATP in those cells. The assay was performed according to the manufacturer's instructions.

Adhesion and spreading assays

The adhesive and spreading properties of control cells versus CUTL1 knockdown cells on different components of ECM were tested using 24-well plates coated with fibronectin, collagen I, collagen IV, laminin or vitronectin (Roche, Mannheim, Germany). After trypsination, cells were incubated in serum-free medium supplemented with 0.5% bovine serum albumin (BSA) at 37°C for 30 min. Subsequently, cells were plated onto coated plates and fixed after indicated time points (3′, 5′, 10′, 15′, 20′, 30′ and 1 h) with 3.7% formaldehyde/phosphate-buffered saline (PBS) for 10 min. After fixation, cells were washed with PBS and covered with glycerol/PBS (1:1). Fixed cells were analysed by light microscopy (Zeiss, Axiovert 135, Oberkochen, Germany) for the efficiency of adhesion and spreading.

For time-lapse microscopy, coverslips were coated with 150 μl of 50 μg/ml of fibronectin (full-length human fibronectin, Roche) solution and incubated for 1 h at 37°C. Cells were detached with trypsin/ethylenediaminetetraacetic acid (EDTA) (0.05% for 1 min). After adding growth medium to inhibit trypsin, cells were centrifugated, cell pellets were suspended in DMEM supplemented with 0.5% BSA and incubated for 30 min before plating on the coated coverslips. Imaging started immediately after plating the cells on coated coverslips.

Cells were imaged simultaneously with × 10 objectives in DIC modes on Olympus IX-71 microscope. Time-lapse sequences were obtained digitally, using SimplePCI software (Compix Inc., Lake Oswego, OR, USA). In experiments using cells transfected with Cy5-labeled siRNA, successfully transfected cells were visualized simultaneously in DIC and fluorescence modes to analyse cell spreading selectively in transfected cells.

GTPase activity assays

For detection of GTP-bound Rac1, Cdc42 and RhoA, control cells and cells stably expressing CUTL1-shRNA were plated on fibronectin-coated plates for 20 min at 37°C. Subsequently, cells were lysed for Rac1/Ccd42-GTP pull-down experiments in 500 μl Rac1/Cdc42-RIPA buffer (50 mmol/l Tris-HCl, pH 7.2, 150 mmol/l NaCl, 10 mmol/l MgCl2, 1% (v/v) Triton X-100, 0.5% (w/v) sodium deoxycholate, 10 μg/μl aprotinin, 10 μg/μl leupeptin, 0.1 mmol/l phenylmethylsulfonyl fluoride (PMSF)) or for RhoA-GTP pull-down assays in RhoA-RIPA buffer (50 mmol/l Tris-HCl, pH 7.2, 500 mmol/l NaCl, 10 mmol/l MgCl2, 1% (v/v) Triton X-100, 0.5% (w/v) sodium deoxycholate, 10 μg/μl aprotinin, 10 μg/μl leupeptin, 0.1 mmol/l PMSF). Cell lysates were cleared by centrifugation at 15 800 g for 15 min at +4°C. Lysates were then incubated with 20–40 μg GST-PAK (Rac1 and Cdc42 assays) or GST-Rhotekin (RhoA assay) for 45 min at +4°C. GST-PAK and GST-Rhotekin were prepared as previously described (Stahle et al., 2003). Beads were then washed four times with washing buffer (50 mmol/l Tris-HCl, pH 7.2, 150 mmol/l NaCl, 10 mmol/l MgCl2, 1% (v/v) Triton X-100, 10 μg/μl aprotinin, 10 μg/μl leupeptin, 0.1 mmol/l PMSF). Precipitated proteins were separated on 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels, transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA, USA) and detected by using the appropriate antibodies (see below).

In vitro ROCK kinase assay

Cells stably expressing CUTL1 shRNA and control cells were plated on fibronectin-coated plates for 20 min at 37°C. After washing with PBS, cells were lysed in ROCK lysis buffer (50 mmol/l M Tris-HCl pH 7.4, 100 mmol/l NaCl, 10% (v/v) glycerol, 0.5% (v/v) Triton X-100, 10 mmol/l MgCl2, 1 mmol/l Na3VO4, 10 mmol/l NaF, 10 μg/μl aprotinin, 10 μg/μl leupeptin, 0.1 mmol/l PMSF). Lysates were incubated with anti-ROCK antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 2 h at +4°C, followed by incubation with protein-G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) for one additional hour. Beads were washed twice with ROCK lysis buffer and twice with ROCK kinase buffer (50 mmol/l Tris pH 7.4, 100 mmol/l NaCl, 10% (v/v) glycerol, 0.05% (v/v) Triton X-100, 2 mmol/l MgCl2, 2 mmol/l MnCl2, 1 mmol/l Na3VO4, 10 mmol/l NaF, 10 μg/μl aprotinin, 10 μg/μl leupeptin, 100 nmol/l PMSF). In vitro kinase assays of precipitated ROCK were performed in 30 μl ROCK kinase buffer with 1 μg recombinant MYPT-1 substrate (Upstate, Charlottesville, VA, USA) and 5 μCi radioactively labelled [32P]γ-ATP (Amersham, Little Chalfont, UK) for 30 min at 30°C. Phosphorylated substrate was separated on 8% SDS-polyacrylamide gel, transferred to a PVDF (Millipore, Bedford, MA, USA) membrane and detected by autoradiography.

Immunoblotting

Immunoblotting was performed as described previously (Weber et al., 2000). Unless stated otherwise, cells were incubated in lysis buffer (25 mM Tris-HCl (pH 8), 150 mM NaCl, 2 mM EDTA, 10% glycerin, 1% Triton-X-100) supplemented with a cocktail of protease inhibitors (1:200; Sigma, Cat.No. P 8340) and the tyrosine phosphatase inhibitor Na3VO4 (final concentration 1 mM), homogenized and cleared by centrifugation. Cell lysates were used for immunoblotting analyses as described.

The following antibodies were used: anti-ROCK-2 and anti-RhoA from Santa Cruz Biotechnology (Santa Cruz, Santa Cruz, CA, USA); anti-Rac1, anti-Cdc42 and anti-Src from Upstate, the β-actin antibody and anti-Csk (C-20) from Santa Cruz Biotechnology; anti-phospho-Src (Tyr416), anti-phospho-Src (Tyr530) and anti-Csk from Cell Signaling (Beverly, MA, USA). The anti-CUTL1 antibody was prepared by our group as described recently (Michl et al., 2005).

Isolation of RNA and reverse transcription

Total RNA was extracted from cell lysates using the RNeasy Midi Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The concentration of RNA samples was determined using a spectrophotometer, and samples were stored at −80°C.

Total RNA of 2 μg was subjected to reverse transcription according to the manufacturer's instructions (Omniscript RT Kit (50), Qiagen). The reaction was performed at 37°C for 1 h and inactivated by boiling at 95°C for 5 min. cDNA was stored at −20°C.

Real-time RT-PCR

cDNA obtained from 2 μg total RNA was diluted 1:8 and subjected to real-time PCR analysis to quantify expression of mRNA of different integrins, src, pak, mypt, PP1 and csk in control cells compared with cells with stably or transiently reduced CUTL1 expression. The reactions were performed in a final volume of 50 μl as follows: 25 μl × 2 Sybr Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 5 μl diluted cDNA, 5 μl 10-fold primer mixture (15 pmol of each primer) and distilled water. Amplification was performed under the following conditions: 2 min at 50°C, 10 min at 95°C, 15 s at 95°C, 1 min at 60°C for 40 cycles, using an ABI PRISM 7700 Sequence Detection System (Applied Biosystems). The following primers were used: csk (F-5′-IndexTermgcg att acc gag gga aca aag-3′, R-5′-IndexTermtgg cgt cgt tct taa tgc act-3′); cyclophilin (F-5′-IndexTermccc tcc acc cat ttg ct-3′, R-5′IndexTerm-caa tcc agc tag gca tgg ga-3′).

Transient transfection of plasmid DNA

HT1080 and Panc1 cells were transfected in 10 cm cell culture dishes with 4 μg of DNA from different plasmids: Myc-tagged full-length human CUTL1 in pMX (kind gift from A Nepveu, McGill University, Canada), wild-type src in pcDNA3.1 (kind gift from Dr K Bonham, University of Saskatchewan, Canada), wild-type human Csk (kind gift from Dr XY Huang, Cornell University, New York) and control pcDNA3.1 (kind gift from Dr C Weber, University of Ulm). The Transfast Transfection Reagent Kit (Promega, Mannheim, Germany) was used according to the manufacturer's instructions.

Stable transfection of CUTL1 shRNA and transient transfection of CUTL1 siRNA

For stable suppression of CUTL1 expression by shRNA, the silencing sequence (sense IndexTermGAAGAACACTCCAGAGGAT) was cloned into the puromycin-resistent pRetrosuper vector according to the manufacturer's instructions (Oligoengine, Seattle, WA, USA).

To obtain stable clones, growth medium was replaced by selection medium containing 2.5 μg/ml puromycin (Calbiochem) 24 h after transient transfection with pRetrosuper expressing CUTL1 shRNA or empty vector. Selection with both puromycin (2.5 μg/ml) and hygromycin (300 μg/ml, Invitrogen, Karlsruhe, Germany) was used to obtain double clones CUTL1Src+ after transfection of the CUTL1-shRNA clones with the src-wt plasmid containing hygromycin resistance.

To obtain transient reductions in CUTL1 mRNA levels, the following CUTL1 silencing sequences were used, which are different from the sequence used for generating stable CUTL1 shRNA clones: sequence 1, sense IndexTermGGAACAGAAGUUACAGAAUtt and sequence 2, sense IndexTermGGAAGCUGAAGCAGCUUUCtt. siRNA oligos were obtained from Ambion. As control, nonsilencing siRNA was obtained from Ambion. For visualization of successfully transfected cells in time-lapse microscopy experiments, Cy5-labeled siRNA oligos for CUTL1 as well as nonsilencing control siRNA were obtained from Ambion. For transient siRNA transfection, cells were transfected using TransMessenger Transfection Reagent (Qiagen, Hilden, Germany) and 20 μ M siRNA according to the manufacturer's instructions.

References

  1. Amano M, Ito M, Kimura K, Fukata Y, Chihara K, Nakano T et al. (1996). Phosphorylation and activation of myosin by Rho-associated kinase (Rho-kinase). J Biol Chem 271: 20246–20249.

  2. Burridge K, Chrzanowska-Wodnicka M . (1996). Focal adhesions, contractility, and signaling. Annu Rev Cell Dev Biol 12: 463–518.

  3. Hakak Y, Martin GS . (1999). Ubiquitin-dependent degradation of active Src. Curr Biol 9: 1039–1042.

  4. Hynes RO . (1992). Integrins: versatility and signaling in cell adhesion. Cell 69: 11–25.

  5. Hynes RO, Lander AD . (1992). Contact and adhesive specificities in the associations, migrations, and targeting of cells and axons. Cell 68: 303–322.

  6. Johnson FM, Saigal B, Talpaz M, Donato NJ . (2005). Dasatinib (BMS-354825) tyrosine kinase inhibitor suppresses invasion and induces cell cycle arrest and apoptosis of head and neck squamous cell carcinoma and non-small cell lung cancer cells. Clin Cancer Res 11: 6924–6932.

  7. Kaido T, Yebra M, Cirulli V, Montgomery AM . (2004). Regulation of human beta-cell adhesion, motility, and insulin secretion by collagen IV and its receptor alpha1beta1. J Biol Chem 279: 53762–53769.

  8. Klinghoffer RA, Sachsenmaier C, Cooper JA, Soriano P . (1999). Src family kinases are required for integrin but not PDGFR signal transduction. EMBO J 18: 2459–2471.

  9. Lowell CA, Soriano P . (1996). Knockouts of Src-family kinases: stiff bones, wimpy T cells, and bad memories. Genes Dev 10: 1845–1857.

  10. Lowry WE, Huang J, Ma YC, Ali S, Wang D, Williams DM et al. (2002). Csk, a critical link of g protein signals to actin cytoskeletal reorganization. Dev Cell 2: 733–744.

  11. Ma YC, Huang J, Ali S, Lowry W, Huang XY . (2000). Src tyrosine kinase is a novel direct effector of G proteins. Cell 102: 635–646.

  12. Michl P, Ramjaun AR, Pardo OE, Warne PH, Wagner M, Poulsom R et al. (2005). CUTL1 is a target of TGFβ signaling that enhances cancer cell motility and invasiveness. Cancer Cell 7: 521–532.

  13. Moarefi I, LaFevre-Bernt M, Sicheri F, Huse M, Lee CH, Kuriyan J et al. (1997). Activation of the Src-family tyrosine kinase Hck by SH3 domain displacement. Nature 385: 650–653.

  14. Nepveu A . (2001). Role of the multifunctional CDP/Cut/Cux homeodomain transcription factor in regulating differentiation, cell growth and development. Gene 270: 1–15.

  15. Parsons JT . (1996). Integrin-mediated signalling: regulation by protein tyrosine kinases and small GTP-binding proteins. Curr Opin Cell Biol 8: 146–152.

  16. Parsons JT, Parsons SJ . (1997). Src family protein tyrosine kinases: cooperating with growth factor and adhesion signaling pathways. Curr Opin Cell Biol 9: 187–192.

  17. Price JT, Bonovich MT, Kohn EC . (1997). The biochemistry of cancer dissemination. Crit Rev Biochem Mol Biol 32: 175–253.

  18. Riento K, Ridley AJ . (2003). Rocks: multifunctional kinases in cell behaviour. Nat Rev Mol Cell Biol 4: 446–456.

  19. Roskoski Jr R . (2004). Src protein-tyrosine kinase structure and regulation. Biochem Biophys Res Commun 324: 1155–1164.

  20. Stahle M, Veit C, Bachfischer U, Schierling K, Skripczynski B, Hall A et al. (2003). Mechanisms in LPA-induced tumor cell migration: critical role of phosphorylated ERK. J Cell Sci 116: 3835–3846.

  21. Stover DR, Liebetanz J, Lydon NB . (1994). Cdc2-mediated modulation of pp60c-src activity. J Biol Chem 269: 26885–26889.

  22. Tan M, Li P, Klos KS, Lu J, Lan KH, Nagata Y et al. (2005). ErbB2 promotes Src synthesis and stability: novel mechanisms of Src activation that confer breast cancer metastasis. Cancer Res 65: 1858–1867.

  23. Thomas SM, Brugge JS . (1997). Cellular functions regulated by Src family kinases. Annu Rev Cell Dev Biol 13: 513–609.

  24. Weber CK, Liptay S, Wirth T, Adler G, Schmid RM . (2000). Suppression of NF-kB activity by sulfasalazine is mediated by direct inhibition of IkB kinases alpha and beta. Gastroenterology 119: 1209–1218.

  25. Zheng XM, Resnick RJ, Shalloway D . (2000). A phosphotyrosine displacement mechanism for activation of Src by PTPalpha. EMBO J 19: 964–978.

Download references

Acknowledgements

This work was supported by a grant of the Deutsche Forschungsgemeinschaft (DFG) to PM. We thank Susanne Braun for expert technical assistance.

Author information

Correspondence to P Michl.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Aleksic, T., Bechtel, M., Krndija, D. et al. CUTL1 promotes tumor cell migration by decreasing proteasome-mediated Src degradation. Oncogene 26, 5939–5949 (2007) doi:10.1038/sj.onc.1210398

Download citation

Keywords

  • CUTL1
  • migration
  • Src
  • Csk
  • spreading

Further reading