Original Paper | Published:

RNA-mediated gene silencing of the RON receptor tyrosine kinase alters oncogenic phenotypes of human colorectal carcinoma cells

Abstract

Altered expression of receptor tyrosine kinases contributes to tumorigenic behaviors of epithelial cancers. In this study, the pathogenic roles of receptor tyrosine kinase RON (recepteur d'origine nantais) in regulating oncogenic phenotypes in colorectal epithelial cells were studied. Increased expression of RON and its variants resulted in colony formation and motile activities of colonic epithelial AA/C1 cells as evident in soft-agar and migration assays, respectively. These results suggest that overexpression of wild-type RON mediates the transformed phenotypes in immortalized colon epithelial cells. In colorectal cancer cells (HT-29, HCT116, and SW620) that naturally express RON, the RON gene expression was silenced by RNA interference. The introduction of RON-specific small interfering (si) RNA significantly affected cancer cell proliferation, motility, and led to increased apoptotic cell death. Focus-forming activities and anchorage-independent growth of colon cancer cells were also dramatically reduced. Moreover, it was demonstrated in tumor growth assays that silencing RON gene expression significantly reduces tumorigenic activities of SW620 cells in vivo. By analysing signaling proteins involved in colon carcinogenesis, we found that the effect of RON-specific siRNA is associated with diminished expression of β-catenin, a critical component in the Wnt signaling pathway. Taken together, our results demonstrate that altered expression of RON in colon cancer cells is required to maintain tumorigenic phenotypes. Thus, silencing RON gene expression could have potential to reverse malignant activities of colon tumors in vivo.

Introduction

Colorectal cancers, almost all of them are adenocarcinomas, arise through multiple genetic alterations including mutational activation of oncogenes and inactivation of tumor suppressor genes (Kinzler and Vogelstein, 1998). Tumors begin as in situ lesions and grow into different morphological patterns (Crawford, 1994). Malignancy of colorectal cancers is evident by their abilities to invade adjacent structures and to metastasize into remote organs (Portera et al., 1998; Yokota, 2000). These processes, collectively known as invasive growth (Comoglio and Trusolino, 2002), ultimately cause death (Crawford, 1994). At present, the precise mechanisms that govern invasion and metastasis of colorectal cancers are still largely unknown. However, it is believed that accumulation of multiple genetic alterations, together with cellular disorganization, determines the malignant behaviors of tumors (Bogenrieder and Herlyn, 2003).

Invasive growth is a complicated process involving a variety of factors (Dorudi and Hart, 1993; Yokota, 2000). Altered expression of cellular proteins, such as receptor tyrosine kinases, contributes significantly to invasive growth of cancers (Birchmeier et al., 1997; Vande Woude et al., 1997). Abnormal activation of the epidermal growth factor receptor (EGFR) in various types of tumors is a typical example (Earp et al., 2003). Activated EFGR not only leads the transformation of normal cells, but also causes tumors with increased invasiveness (Humphreys and Hennighausen, 2000). Similarly, MET, the receptor for hepatocyte growth factor/scatter factor (HGF/SF) is involved in regulating malignant phenotypes in epithelial cancers (Vande Woude et al., 1997). Thus, understanding the roles of receptor tyrosine kinases in regulating cancer malignant phenotypes should provide insight into mechanisms of cancer progression.

The receptor tyrosine kinase RON (recepteur d'origine nantais) is the member of the MET proto-oncogene family (Ronsin et al., 1993; Rubin et al., 1993). The product of the RON gene is a 180 kDa heterodimeric protein composed of a 40 kDa extracellular α-chain and a 150 kDa transmembrane β-chain with intrinsic tyrosine kinase activity (Ronsin et al., 1993). The ligand for RON was identified as macrophage-stimulating protein (MSP) (Skeel et al., 1991; Wang et al., 1994; Gaudino et al., 1994), also known as HGF-like protein (Han et al., 1991). The RON gene is normally transcribed at relatively low levels in cells from epithelial origin (Gaudino et al., 1994). Recent studies have shown that RON expression is significantly altered in the majority of primary colorectal carcinoma samples (Zhou et al., 2003) and in established colorectal cancer cell lines (Chen et al., 2000). Abnormal RON accumulation is also accompanied with generation of alternatively spliced RON variants with oncogenic potentials (Wang et al., 2000; Zhou et al., 2003). In vitro studies have shown that activated RON transduces multiple signals that regulate various functions in colon epithelial cells. These activities include increased resistance to apoptotic signals and enhanced cell motile/invasive activities (Chen et al., 2000; Wang et al., 2000; 2003; Zhou et al., 2003). Thus, altered RON expression and its aberrant signaling contribute significantly to the pathogenesis and carcinogenesis of colorectal cancers (Wang et al., 2003).

The present work is to determine if RON and its variants play a role in cellular transformation and malignant behavior of colorectal epithelial cells. Using human colorectal epithelial and established cancer cells as a model, we demonstrated that overexpression of RON leads to transforming phenotypes in immortalized colorectal epithelial cells. Silencing RON gene expression by RNA interference (RNAi) (Fire et al., 1998) impairs proliferation, migration, focus formation, anchorage-independent growth, and in vivo tumor formation of established colon cancer cells. Thus, altered expression of RON is critical in regulating tumorigenic phenotypes in certain colorectal epithelial cancers.

Results

Expression of RON and its variants in immortalized human colon epithelial cells

To determine if RON expression regulates cellular activities, human colon epithelial AA/C1 cells were chosen to express RON and its variants. AA/C1 cells originated from the colonic adenoma cell line PC/AA (Williams et al., 1990) and express very little of RON as evident in RT–PCR and Western blot analysis (data not shown). AA/C1 cells display relatively normal colonic epithelial properties and replicate in an anchorage-dependent manner (Williams et al., 1990).

Expression of human RON, RON mutant m1254t (RONmt), or RON splicing variants, (RONΔ165, Δ160, or Δ155) (Santoro et al., 1998; Zhou et al., 2003) was achieved by plasmid transfection techniques. Results in Figure 1 showed the levels of RON or its variants expressed in individual cell lines as determined by Western blot analysis. The sizes of proteins were identical to those reported previously (Zhou et al., 2003). RON was phosphorylated upon MSP stimulation (data not shown). However, RONmt and all three RON variants showed spontaneous tyrosine phosphorylation, consistent with those reported previously (Zhou et al., 2003).

Figure 1
figure1

Expression of RON and its variants in colon epithelial AA/C1 cells: generation of stable AA/C1 cells expressing RON or its variants was performed as described previously (Wang et al., 2000). Cellular proteins (50 μg/sample) from individual cell lines were separated in an 8% SDS–PAGE under reduced conditions. RON and its variants were detected in Western blotting by rabbit IgG antibodies specific to the C-terminus of RON. M-RE7 or parental AA/C1 cells were used as the positive or negative control, respectively. The same membrane was reprobed with rabbit IgG antibodies to β-actin for loading controls

Using transfected AA/C1 cells, we studied cell motile activities in the migration assays. Results are shown in Figure 2. Parental AA/C1 cells did not migrate in the presence of MSP. However, when RON was overexpressed and stimulated with MSP, AA/C1 cells migrate in a concentration-dependent manner. These studies demonstrate that activation of RON in AA/C1 cells results in increased motile activities.

Figure 2
figure2

Increased migration of AA/C1 cells expressing RON. The cell migration assays were performed as detailed in Materials and methods. Similar results were also seen when other transfected cells were used. One of two experiments with similar results

Effect of RON and its variants on anchorage-independent growth of AA/C1 cells

To test if RON and its variants have the ability to cause transforming phenotypes, the colony formation by transfected AA/C1 cells was studied in the soft-agar assay. AA/C1-RONmt cells were used as positive control. Results are shown in Figure 3. The parental AA/C1 and control vector-transfected cells did not produce colonies in the soft agar (Figure 3a). In contrast, numerous colonies were observed in AA/C1-RON, AA/C1-RONΔ160, or AA/C1-RONΔ155 cells (Figure 3a). The number of colonies formed by AA/C1-RON cells was much less than those by AA/C1-RONΔ160 or RONΔ155 cells, but is significant in comparison with those in control cells (Figure 3b). Surprisingly, no colonies were formed by AA/C1-RONΔ165 cells (Figure 3). These results suggest that overexpression of RON, RONΔ160, or RONΔ155 results in anchorage-independent growth of AA/C1 cells in soft agar.

Figure 3
figure3

Anchorage-independent growth of AA/C1 cells expressing RON or its variants: the experiments were conducted using stable AA/C1-RON or RON variant cells. Cells at 1 × 104 cells/dish were suspended in 2 ml DMEM containing 2% FBS and 0.35% agar and then placed on the top of 2.5 ml solidified 0.6% agar. The plates were incubated at 37°C. Fresh media were added once a week. Cells were tested in triplicate. The colonies were counted and photographed (magnification × 40) after cells were cultured for 21 days. (a) Colony formation in soft agar. Parental AA/C1 cells that did not form colonies as pTCB-transfected cells are not shown. (b) Numbers of colonies from triplicate (mean±s.d.). One of three experiments with similar results

Inhibition of RON-mediated AA/C1 cell colony formation by small interfering (si) RNA

To determine if knockdown of RON expression reverses the transforming phenotypes of transfected AA/C1 cells, a 21-nt sequence within the RON gene exon 1 (Angeloni et al., 2000) was chosen as the siRNA target. This 21-nt target sequence, separated by a 7-nt spacer from its reverse complement sequence, was introduced into the expression vector psiRNA-hl (InvivoGen, San Diego, CA, USA), resulting in the vector psiR-RON. The control vector (psiRmRON) containing the same sequence with four mutations was also generated. Transfection of psiR-RON into AA/C1-RON cells resulted in the complete silencing of RON expression (Figure 4a). No silencing effect was seen when pisR-mRON was used. The specific activity of siRNA on the RON expression was confirmed by Western blot analysis, in which, the expression of MET, a protein with similar structures to RON, was not affected (data not shown).

Figure 4
figure4

Inhibitory effect of siRNA-mediated RON silencing on colony formation of AA/C1-RON cells: (a) knockdown of RON expression by siRNA. Stable AA/C1-RON cells were left alone or transiently transfected with psiR-RON or psiRmRON (5 μg of the plasmid DNA/60 mm diameter dish) for 48 h. RON expression was determined by rabbit IgG antibodies to RON in Western blotting. The same membrane was reprobed with antibodies to β-actin for loading controls. (b) Effect of siRNA on colony formation by AA/C1-RON cells. Cells were transfected with psiR-RNA as described in (a). The colony formation was determined as described in Figure 3. Similar results were also observed in other transfected cells. One of three experiments with similar results

We then tested if siRNA has abilities to inhibit RON-mediated colony formation by AA/C1 cells. AA/C1-RON cells were transiently transfected with psiR-RON or psiRmRON. Results are shown in Figure 4b. Control AA/C1-RON cells grew colonies in soft agar (data not shown). Transfection of cells with psiRmRON had no effect on colony formation. The inhibitory effect was only seen when AA/C1-RON cells were transfected with the psiR-RON vector. The numbers of colonies were significantly reduced. Similar results were also observed when AA/C1 cells expressing RON variants were used. These results, thus, demonstrate that RON expression is responsible for the transforming phenotypes of AA/C1 cells.

Effect of chemical inhibitors on RON and its variants-mediated colony formation

To determine the potential signaling components involved in RON-mediated transforming activities, chemical inhibitors PD98059, Wortmannin, or SB203580, specific to MAP kinase (MEK), PI-3 kinase, or p38 kinase, respectively, were included in the soft-agar assay. The concentrations of these chemicals completely inhibited the phosphorylation of corresponding proteins (data not shown). Results are shown in Figure 5. PD98059 at 50 μ M completely inhibited the RON-mediated colony formation by AA/C1 cells. Wortmannin also inhibited the colony formation. Significant inhibition (74%) was achieved when 100 nM of Wortmannin was used. No inhibitory effect was observed when SB203580 (up to 5 μ M) was used. Similar results were also observed in AA/C1 cells expressing RON variants (data not shown). These results indicate that MAP kinase and PI-3 kinase is involved in RON or its variants-mediated cell-transforming activities by AA/C1 cells.

Figure 5
figure5

Effect of chemical inhibitors on RON-mediated colony formation: AA/C1-RON cells were monitored for colony formation as described in Figure 3. Individual chemical inhibitors (PD98059, 50 μ M; Wortmannin, 100 nM; or SB203580, 5 μ M, respectively) were added into the culture dishes after initiation of cell cultures. Results (mean±s.d.) shown here are from one of three experiments. CTL, control. Magnification × 40

Silencing of RON expression by siRNA in human colon cancer cell lines

To determine siRNA-mediated activities in more detail, human colon cancer cell lines (HT-29, HCT116, and SW620) that express RON endogenously were used (Chen et al., 2000). The effect of siRNA on the RON gene expression was tested. As shown in Figure 6, siRNA affects RON expression in all three cancer cell lines. However, complete knockdown of the RON gene expression was only observed in HCT116 and SW620 cells. Again, no silencing effect was seen when the control vector psiRmRON was introduced into tumor cells. Western blot analysis also confirmed that siRNA had no inhibitory effect on MET expression (data not shown). These results suggest that siRNA is effective not only in AA/C1 cells that artificially express RON, but also in cancer cells that naturally transcribe RON.

Figure 6
figure6

Silencing of the RON gene expression by siRNA in human colon cancer cell lines: three colon cancer cell lines (HT-29, HCT116, and SW620) were cultured and transiently transfected with psiR-RON or the control vector psiRmRNA. Cells were incubated for 48 h and RON was detected by rabbit IgG antibodies to RON in Western blot analysis. The same membrane was reprobed with antibodies to β-actin for loading controls. One of three experiments with similar results

Effect of RON-specific siRNA on replication and apoptosis of colon cancer cells

To study if silencing RON expression affects the oncogenic phenotypes of colon cancer cells, we first determined the effect of siRNA on cell proliferation using BrdU incorporation as a marker. As shown in Figure 7a and b, siRNA introduction resulted in significant inhibition of cell proliferation in all three cancer cell lines. The percentages of BrdU-positive cells were reduced, in average, up to 60% in comparison with those of control cells. The BrdU incorporation was not affected in cancer cells transfected by the psiRmRON vector.

Figure 7
figure7

Effect of siRNA-mediated RON silencing on proliferation and apoptosis of colon cancer cells: (a, b) Effect of siRNA on cell BrdU incorporation. Cells were cultured in medium alone or transiently transfected with 5 μg/dish of psiR-RNA or psiRmRNA in triplicate, respectively. After incubation for 48 h, BrdU was added into the cell cultures for 4 h. The BrdU-positive cells with dark-brown nuclei were detected as detailed in Materials and methods and photographed (magnification × 200). (a) The percentages of BrdU-positive cells in (b) (mean±s.d.) were calculated from one of three experiments with similar results. (c) Effect of RON knockdown on the kinetic growth of SW620 cells. SW620 cells stably expressing psiR-RNA (1 × 104 cells/well) were cultured in DMEM containing 5% FBS in a 96-well culture dish for 5 days. At indicating time intervals, cells were fixed, stained with HEMA-3 staining solutions, and lysed with 100 μl/well of 1% SDS solution. Color intensity was measured at 570 nM in an ELISA reader. Absorbance was converted into cell number by reference to a stained cell concentration standard curve (Wang et al., 1996). (d) Increased apoptotic death of colon cancer cells upon siRNA expression. Cells were transiently transfected with psiR-RNA for 48 h as described in (a). Apoptotic cells were identified by using the TUNEL assay (pointed by open arrows) under microscopy (magnification × 200) as described in Materials and methods. Experiments were performed in triplicate and repeated twice with similar results

To further confirm these results, a pool of SW620 cells stably expressing siRNA was employed. Kinetic studies of the replication status showed that SW620-siRNA cells grew significantly slower than parental SW620 or psiRmRNA-transfected cells (Figure 7c). The numbers of SW620-siRNA cells at the end of incubation period (day 5) is significantly lower than those of control cells.

We then tested if siRNA expression increases apoptosis. As shown in Figure 7d, Control HT-29, HCT116, or SW620 cells replicate without visible apoptotic cell death. Similar results were also seen in cells transfected with psiRmRON. However, when cells were transfected with psiR-RON, the increased nuclear condensation in a number of cells was observed. These results, together with those in Figure 7a–c, suggest that knockdown of the RON gene expression in established colon cancer cells not only impairs cell replication, but also results in increased apoptotic death.

Reduction of migration activities of colon cancer cells stably expressing siRNA

Using stable SW620-siRNA cells, we tested if RON silencing affects cell motile activities. Results in Figure 8 showed that untransfected SW620 cells have high levels of spontaneous migration. These activities were further enhanced, in a dose-dependent manner, when MSP in added. The knockdown of RON expression completely abolished the MSP-induced migration of SW620 cells, but had no effect on spontaneously motile activities. Similar results were also obtained when HCT116 cells were used. Thus, the effect of siRNA on colon cancer cell migration is limited in a RON-dependent manner.

Figure 8
figure8

Effect of RON-specific siRNA on MSP-induced migration of colon cancer cells: the migration assay was detailed in Materials and methods. Stable SW620-siRNA cells (2 × 105 cells/well) were allowed to migrate for 3 h in the presence or absence of MSP. High levels of spontaneous migration were observed. Migrated cells were counted in three randomly selected areas. One of three experiments with similar results

Effect of siRNA on focus-forming activities, anchorage-independent growth, and in vivo tumorigenesis of colon cancer cells

Since siRNA inhibited the RON-mediated colony formation in AA/C1 cells, we wanted to know if it had a similar effect on colon cancer cells. We first used the focus-forming assay to test this possibility. As shown in Figure 9a, siRNA inhibited the focus-forming activities of SW620 cells. The numbers of foci from siRNA expressing SW620 cells were dramatically reduced in comparison with those from controls cells. We then used the soft-agar assay to determine the effect of siRNA on anchorage-independent growth of SW620 cells. Results in Figure 9b show that after psiR-RON transfection, the number of colonies grown in the soft agar was significantly decreased. Further, we used stable SW620-siRNA cells in colony formation assays. Results in Figure 9c showed that the colony formation was reduced in the absence of RON. Similar results were also obtained when HCT116 cells were used. The effect of RON-specific siRNA on in vivo tumorigenesis of stably transfected SW620 cells was also tested (Table 1). As documented, SW620 cells are highly tumorigenic in nude mice with tumor formation in less than 15 days after inoculation (Fogh et al., 1977). The data from SW620 cells harboring the control vector psiRmRON confirm that this is the case. Tumors grew rapidly and reached the size of 17 mm in diameter within a very short period. In contrast, the tumor growth by SW620 cells expressing the psiR-RON vector was significantly delayed. Although tumors were grown in all three mice, the days of tumor onset was dramatically prolonged. The sizes of tumors were significantly smaller than those seen in psiRmRON-transfected SW620 cells. Taken together, these results suggest that the RON expression is required for the focus-forming activities, the anchorage-independent growth, and the in vivo tumor formation of colon cancer cells.

Figure 9
figure9

Effect of RON-specific siRNA on focus and colony formation of colon cancer cells: (a) Expression of siRNA reduces focus-forming activities of SW620 cells. SW620 cells (4 × 105 cells/dish) were transfected with psiRNA or psiRmRNA as described in Figure 4. Cells were harvested 24 h later and replated as low density in the presence of Zeocin. The foci were counted after cells were incubated for 21 days. (b) Decreased colony formation of SW620 cells transfected with psiR-RON. Cells were transfected with plasmids in triplicate as described above and seeded at a density of 2 × 104 per dish containing Zeocin. After incubation for 21 days, the colonies larger than 250 μm in diameter were counted. Results shown here are from one of three experiments with similar results. (c) Colony-forming activities of SW620 cells stably expressing RON-specific siRNA. Experimental conditions were similar as those described in (b) except SW620 or stable SW620-siRNA cells were used. Results shown are one of three experiments with similar results

Table 1 Effect of RON-specific siRNA on tumorigenesis of SW620 cells in vivo

Effect of siRNA-mediated RON silencing on β-catenin expression in colon cancer cells

To determine if RON silencing affects the expression of signaling proteins associated with tumorigenic phenotypes, the levels of several cellular proteins such as MAP kinase, PI-3 kinase, p38, and others were determined. We found that the expression of β-catenin, a key component of the Wnt signaling pathway, is diminished in AA/C1 cells expressing RON or RON variants (Figure 10a). Complete loss of β-catenin expression was also observed in HT-29, HCT116, and SW620 cells transiently transfected with psiR-RON (Figure 10b). This effect was not seen when the control vector was used. Moreover, we observed by Western blot analysis that expression of cMYC, the downstream target of the β-catenin signaling pathway (He et al., 1998), was also diminished (data not shown). To further confirm these results, stable SW620-siRNA cells, in different passages, were used to monitor β-catenin expression in correlation with RON expression. We have found that effect of siRNA on RON expression was gradually diminished during the passages (our unpublished data). As shown in Figure 10c, parental SW620 cells express high levels of β-catenin. However, when siRNA is expressed, β-catenin expression was diminished. No β-catenin expression was observed in SW620-siRNA cells at early passages (up to 8 weeks). In this period, RON expression was also silenced. However, when SW620-siRNA cells at later passages (12 weeks) were evaluated, RON expression gradually reappeared. The appearance of RON is accompanied with re-expression of β-catenin. Thus, knockdown of RON expression is accompanied with reduction of β-catenin expression in colon cancer cells.

Figure 10
figure10

Effect of RON-specific siRNA on β-catenin expression: transient transfection of psiR-RON in AA/C1-RON cells (a) or colon cancer cells (b) were performed as described above. SW620 cells stably expressing siRNA were also used (c). Expression of β-catenin or RON was determined by antibodies specific to β-catenin or RON, respectively. The same membrane was reprobed with antibodies to β-actin for loading controls. One of three experiments with similar results

Discussion

Previous studies have demonstrated that altered RON expression, accompanied with generation of oncogenic forms of RON variants, exists in the majority of primary human colorectal carcinoma samples (Zhou et al., 2003). The current study is a continuation of the previous project aimed at determining the requirement of RON and its variants in maintaining tumorigenic phenotypes in human colon epithelial cells. Our results demonstrated that overexpression of wild-type RON causes transforming phenotypes in immortalized human colon epithelial AA/C1 cells. Silencing the RON gene expression by siRNA techniques in a panel of colon cancer cells significantly affects cell replication, survival, migration, focus formation, anchorage-independent growth, and tumor formation in vivo. Our data also provide evidence indicating that the oncogenic activities of RON or its variants are probably mediated by RON-transduced signals that regulate cellular signaling components such as β-catenin in the Wnt signaling pathway. Thus, altered RON expression is one of the factors contributing to the carcinogenesis of colorectal epithelial cells.

Early studies have shown that wild-type RON does not have cell-transforming activities in vitro when expressed in rodent fibroblast cells (Santoro et al., 1996). However, RON mutants with tumorigenic activities can be generated by replacing certain conserved amino acids in the kinase domains (Santoro et al., 1998; Williams et al., 1999) or by mRNA splicing that deletes a critical region in the extracellular domain (Zhou et al., 2003). The results from our current studies (Figure 3) demonstrated that wild-type RON has the ability to cause colony formation by immortalized colon epithelial AA/C1 cells. First, AA/C1-RON cells, like AA/C1-RONmt and AA/C1-RONΔ160 cells, forms numerous colonies in soft agar. This activity is not observed in parental or vector control AA/C1 cells. Second, knockdown of RON expression by RON-specific siRNA abolishes the colony formation. Third, the RON-mediated colony formation was inhibited by chemical inhibitors at concentrations specific to PI-3 kinase or MAP kinase, indicating that these signaling proteins are involved in RON-mediated cell transformation.

Consistent with results described above, recent in vitro and in vivo studies have indicated that RON expression and activation are tumorigenic under certain conditions (Peace et al., 2001; Chen et al., 2002b). Human or mouse wild-type RON has been shown to cause lung tumors in transgenic mice and to mediate cellular transformation in mouse cells (Peace et al., 2001; Chen et al., 2002b). Thus, abnormal accumulation and activation of RON leads to transforming phenotypes in certain types of epithelial cells. However, since increased RON expression is only observed in colon cancer cells but not in normal or premalignant lesions (Zhou et al., 2003) and the RON-mediated transforming activities are relatively low, the importance of the transforming activities of RON should be judged with caution. From available information about tumorigenesis of RON in different epithelial cells including those from colon, lung, and breast (Maggiora et al., 1998; Chen et al., 2002a; Zhou et al., 2003), it is believed that the transforming activities of RON might not be a major player during the initiation stage of colon epithelial transformation. They may be more relevant to the maintenance of malignant phenotypes in colon cancer cells at later stages.

One of the interesting observations is that RONΔ165 does not have colony-forming activities when expressed in AA/C1 cells. RONΔ165 was identified in a stomach cancer cell line (Collesi et al., 1996) and in normal colon epithelial mucosa (Okino et al., 1999). Our analysis confirmed that the inability of RONΔ165 in induction of colony formation is not related to its expression level or phosphorylation status. The possible explanation might come from its unique structure (Collesi et al., 1996; Zhou et al., 2003). RONΔ165 is a splicing variant lacking 49 amino acids in the extracellular domain coded by exon 11 (Collesi et al., 1996; Zhou et al., 2003). The deletion of exon 11 results in a prematured, single-chain RON precursor retains in cytoplasm that has weak kinase activities (Collesi et al., 1996). Thus, RONΔ165 is a unique variant with distinct biological properties. Further study of RONΔ165 with regard to its inability in cellular transformation might provide important clues about functional basis in RON oncogenesis.

To analyse cellular activities related to malignant behaviors in colon cancer cell lines, we employed siRNA techniques. RNAi is a potent method to suppress gene expression and to elucidate gene function in mammalian cells (McManus and Sharp, 2002). We have selected several siRNA sequences specifically to silence the RON gene expression in colon cancer cells. The effect of siRNA on RON expression either by transfected AA/C1 cells or by colon cancer cells is promising and relatively specific. RON expression is knocked down by RON-specific siRNA but not by mutant siRNA. Also, the RON-specific siRNA has no effect on MET expression, the receptor for HGF/SF with high sequence similarity to RON. Moreover, introduction of siRNA in different colon cancer cell lines results in profound phenotypic changes. Thus, application of RON-specific siRNA has potentials in the future to elucidate the pathogenic mechanisms of RON in colon cancers as well as in other types of epithelial tumors.

By silencing RON expression in HT-29, HCT116, and SW620 cells, we have demonstrated that altered RON expression plays a pivotal role in maintaining growth, survival, and other tumorigenic activities in these colon cancer cells. As shown in Figures 7, 8 and 9 and Table 1, silencing RON expression significantly impairs cell replication, which is accompanied with increased apoptotic cell death. Also, MSP-induced motilities were clearly abolished by siRNA even though the spontaneous migration of tumor cells was not affected. The abilities of colon cancers to form foci and colonies are blocked upon knockdown of the RON gene expression. The in vivo tumorigenic assay further demonstrated that silencing RON expression impairs tumorigenic activities of SW620 cells with delayed tumor formation and reduced tumor volumes. Since increased RON expression results in activation of multiple signaling pathways including Ras, MAP kinase, PI-3 kinase, and others (Wang et al., 2003) and these signaling proteins are essential in malignant progression of colon cancer cells (Oving and Clevers, 2002), it is believed that RON-transduced signals are integrated into the signaling network and are required for sustaining oncogenesis of colon cancers. Disruption of RON-mediated signals effectively impairs the aberrant signaling network leading to diminished tumorigenic phenotypes.

The findings that silencing RON expression causes diminished β-catenin expression are unexpected. β-Catenin is a critical component in the Wnt signaling pathway (Bienz and Clevers, 2000). Abnormal activation of β-catenin is essential in the initiation of colorectal tumors and its progression towards malignancy (Bright-Thomas and Hargest, 2003). The data from our present studies indicate that RON expression is important in modulating β-catenin expression both in immortalized AA/C1 cells and in colon cancer cells. Since expression of other signaling proteins like p38 and PI-3 kinase was not affected and the mutant siRNA does not have this activity, we speculate that the observed effect is relatively specific to β-catenin. At present, we do not know the mechanisms of how knockdown of the RON gene expression leads to diminished β-catenin expression. It is possible that altered β-catenin expression is due to its increased degradation during the process of silencing the RON gene expression. A previous report has shown that expression of oncogenic RON results in the association of RON with β-catenin in epithelial cells (Danilkovitch-Miagkova et al., 2001). The association causes β-catenin phosphorylation, which reduces the degradation of β-catenin leading to its abnormal accumulation (Danilkovitch-miagkova et al., 2001). Thus, phosphorylation is one of the mechanisms that stabilize β-catenin. We have showed previously that RON or its variants are constitutively active in colon cancer cells including HCT116, HT-29, and SW620 (Chen et al., 2000; Wang et al., 2000). Also, knockdown of the RON gene expression significantly alters protein phosphorylation patterns in three colon cancer cell lines (our unpublished data). These data indicate that RON-transduced signals are important in maintaining cellular phosphorylation status, which is critical in regulating β-catenin stability (Danilkovitch-Miagkova et al., 2001). However, since our knowledge about the effect of siRNA is still limited and the siRNA-mediated off-target effect on unrelated gene expression has been documented (Jackson et al., 2003; Persengiev et al., 2004), it is possible that the downregulated β-catenin expression is a side effect during the processing of the RON gene silencing. Nevertheless, our data provide an important clue indicating that the diminished β-catenin expression in association with RON silencing could be a novel mechanism by which RON regulates oncogenic phenotypes of colon cancer cells. If this is the case, targeting RON expression might have therapeutic potential in the reversal of malignant behavior in colon cancer cells.

Materials and methods

Cells and reagents

M-RE7 cells are derived from MDCK cells transfected with human RON cDNA (Wang et al., 1994). AA/C1 cells were from Dr C Gespach (Hôpital Saint-Antoine, Paris, France). Human colorectal carcinoma cell lines, HT-29, HCT116, and SW620, were from ATCC (Rockville, MD, USA). The RON mutant cDNA (RONm1254t) (Santoro et al., 1998) was from Dr G Gaudino (Universita di Torino, Novara, Italy). The cDNA encoding three RON variants (RONΔ165, RONΔ160, or RONΔ155), respectively, was used as previously described (Zhou et al., 2003). Human MSP was from Dr EJ Leonard (National Cancer Institute, Frederick, MD, USA). Mouse mAb ID2 and Rabbit IgG antibodies specific to human RON were used as described previously (Montero-Julian et al., 1998; Wang et al., 2000). Mouse mAb P-Tyr-100 to phospho-tyrosine or phosphor-p42/44 and rabbit IgG antibodies to p38 or phosphor-p38 were from Cell Signaling Inc. (Beverly, MA, USA). Mouse mAb to β-catenin and pan-ERK were from Transduction Laboratories (San Diego, CA, USA). Goat IgG antibodies to β-actin were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). PD98059, SB203580, and Wortmannin were from CalBiochem (San Diego, CA, USA). ECL detection reagents were from Amersham-Pharmacia (Arlington Heights, IL, USA).

Establishment of AA/C1 cell lines expressing RON, RONm1254t, or RON variants

The cDNA encoding RON, RONm1254t, RONΔp165, RONΔp160, or RONΔp155 were inserted into the vector pTracer™-CMV/Bsd (Invitrogen, San Diego, CA, USA). Transfection of PC/AA/C1 cells with individual plasmids or the empty vector was performed as described previously (Wang et al., 2000). Cells were selected with 4 μg/ml of Blasticidin and expanded into cell lines. Expression of RON, RON-T, or individual RON variants were determined by Western blotting using the rabbit IgG antibody to RON.

Generation of RON mRNA silencing vectors and cell lines

The psiRNA-hH1zeo vector was from Invivogen (San Diego, CA, USA). Three siRNA sequences were designed using standard selection rules (Elbashir et al., 2002). A 55-nucleotide that encodes two complementary sequences of 21 nt, corresponding to the RON coding sequence in exon 1, separated by a hairpin structure, was inserted into psiRNA-hH1zeo, yielding the vector psiR-RON. A control vector psiRmRON containing a mutant RON sequence was also generated. The effect of siRNA on silencing RON expression was determined by Western blot analysis. AA/C1 or other cells stably expressing siRNA were obtained by selection with Zeocin. A pool of cells with reduced RON expression was used for functional analysis.

Western blot analysis

The methods were performed as described previously (Wang et al., 2000). Cellular proteins (50 μg/sample) were separated in an 8 or 12% polyacrylamide gel under reducing conditions and transferred to Immobilon-P membrane (Millipore, Bedford, MA, USA). Rabbit or mouse IgG to RON or other proteins were used as primary antibodies followed by goat anti-rabbit or mouse IgG conjugated with horse radish peroxidase. The reaction was developed with enhanced ECL reagents and recorded on film. In some experiments, the membrane was treated with the erasure buffer (Wang et al., 1994) and reprobed with other antibodies.

Cell growth assays

Two assays were used to monitor cell proliferation. One is the BrdU incorporation assay. Briefly, HT-29 or other cells were transfected with psiR-RON or control vectors, labeled with BrdU, and then incubated with biotinylated mAb to BrdU (Zymed Laboratories Inc., South San Francisco, CA, USA). Nuclei with positive staining were counted in three random selected areas under a microscope and photographed. Results were expressed as the percentage of BrdU-positive cells. The other assay is the color-based cell growth assay (Wang et al., 1996). SW620 cells stably expressing psiR-RON (1 × 104 cells/well) were cultured in a 96-well culture plate, fixed at different intervals, and staining with HEMA-3 Stain Set (Biochemical Sciences Inc., Swedesboro, NJ, USA). Cells transfected with control vectors were used as controls. The optical densities of each well were measured by an ELISA reader at 570 nm.

Apoptotic assays

Apoptotic cells were determined by the TUNEL assay using a DeadEnd colorimetric kit (Promega, Madison, MI, USA). A minimum of 200 cells was counted in each sample. Apoptotic cells were expressed as a percentage of the total counted nuclei.

Cell transformation assays

The focus-forming assay was performed as described (Santoro et al., 1998). The numbers of foci in experimental or control cells were determined 21 days after initiation of cell cultures. The soft-agar assay was carried out as previously described (Zhou et al., 2003). The number of colonies larger than 250 μm in diameter was counted and photographed after cells were cultured for 21 days.

Cell migration assays

A multiwell chamber was used to monitor cell transmembrane migration as detailed previously (Wang et al., 1994). Migrated cells were counted in three randomly selected areas. Results were expressed as the percentage of input cells that migrated.

In vivo tumorigenesis assays

The assays were performed as previously described (Zhou et al., 2003). Briefly, SW620 cells stably expressing psiR-RON or psiRmRON were suspended in 0.2 ml serum-free DMEM and inoculated subcutaneously into the posterior flank of athymic nude mice (2 × 106 cells per mouse, three mice per group). The tumor formation was monitored daily. The latency was determined as the period of time required by tumors to reach a diameter of 8 mm. Mice that did not form tumors or the sizes of tumors less than <8 mm in diameter were observed for an additional 2 weeks.

References

  1. Angeloni D, Danilkovitch A, Ivanov SV, Breathnach R, Johnson BE, Leonard EJ and Lerman MI . (2000). Genes Chromosomes Cancer, 29, 147–156.

  2. Bienz M and Clevers H . (2000). Cell, 103, 311–320.

  3. Birchmeier W, Brinkmann V, Niemann C, Meiners S, DiCesare S, Naundorf H and Sachs M . (1997). Ciba Found. Symp., 212, 230–240.

  4. Bogenrieder T and Herlyn M . (2003). Oncogene, 22, 6524–6536.

  5. Bright-Thomas RM and Hargest R . (2003). Eur. J. Surg. Oncol., 29, 107–117.

  6. Chen Y-Q, Zhou Y-Q, Angeloni-Andreazzoli D, Kurtz AL, Qiang X-Z and Wang M-H . (2000). Exp. Cell Res., 261, 229–238.

  7. Chen Y-Q, Zhou Y-Q, Fisher JH and Wang M-H . (2002a). Oncogene, 21, 6382–6386.

  8. Chen Y-Q, Zhou Y-Q, Fu L-H, Wang D and Wang M-H . (2002b). Carcinogenesis, 23, 1811–1891.

  9. Collesi C, Santoro MM, Gaudino G and Comoglio PM . (1996). Mol. Cell. Biol., 16, 5518–5526.

  10. Comoglio PM and Trusolino L . (2002). J. Clin. Invest., 109, 857–862.

  11. Crawford JM . (1994). Pathological Basis of Disease: The Gastrointestinal Tract Cotran RZ, Kumar V, Robbins SL (eds). WB Saunders: Philadelphia, PA, pp. 755–829.

  12. Danilkovitch-Miagkova A, Miagkov A, Skeel A, Nakaigawa N, Zbar B and Leonard EJ . (2001). Mol. Cell. Biol., 21, 5857–5868.

  13. Dorudi S and Hart IR . (1993). Curr. Opin. Oncol., 5, 130–135.

  14. Earp HS, Calvo BF and Sartor CI . (2003). Trans. Am. Clin. Assoc., 114, 315–333.

  15. Elbashir SM, Harborth J, Weber K and Tuschl Y . (2002). Methods, 26, 199–213.

  16. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE and Mello CC . (1998). Nature, 391, 806–811.

  17. Fogh J, Fogh JM and Orfeo T . (1977). J. Natl. Cancer Inst., 59, 221–226.

  18. Gaudino G, Follenzi A, Naldini L, Collesi C, Santoro M, Gallo KA, Godowski PJ and Comoglio PM . (1994). EMBO J., 13, 3524–3532.

  19. Han S, Stuart LA and Degen SJF . (1991). Biochemistry, 30, 9768–9780.

  20. He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT, Morin PJ, Vogelstein B and Kinzler KW . (1998). Science, 281, 1509–1512.

  21. Humphreys RC and Hennighausen L . (2000). Oncogene, 19, 1085–1091.

  22. Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M, Li B, Cavet G and Linsley PS . (2003). Nat. Biotechnol., 21, 635–637.

  23. Kinzler KW and Vogelstein B . (1998). The Genetic Basis of Human Cancer: Colorectal Tumors Vogelstein B, Kinzler KW (eds). McGraw-Hill: New York, pp. 565–587.

  24. Maggiora P, Marchio S, Stella MC, Giai M, Belfiore A, De Bortoli M, Di Renzo MF, Costantino A, Sismondi P and Comoglio PM . (1998). Oncogene, 16, 2927–2933.

  25. McManus MT and Sharp PA . (2002). Nat. Rev.Genet., 3, 737–747.

  26. Montero-Julian FA, Dauny I, Flavetta S, Ronsin C, Andre F, Xerri L, Wang M-H, Marvaldi J, Breathnach R and Brailly H . (1998). Hybridoma, 17, 541–551.

  27. Okino T, Egami H, Ohmachi H, Takai E, Tamori Y, Nakagawa K, Nakano S, Akagi J, Sakamoto O, Suda T and Ogawa M . (1999). Intern. J. Oncol., 15, 709–714.

  28. Oving IM and Clevers HC . (2002). Eur. J. Clin. Invest., 32, 448–457.

  29. Peace BE, Hughes MJ, Degen SJF and Waltz SE . (2001). Oncogene, 20, 6142–6151.

  30. Persengiev SP, Zhu X and Green MR . (2004). RNA, 10, 12–18.

  31. Portera Jr CA, Berman RS and Ellis LM . (1998). Surg. Oncol., 7, 183–195.

  32. Ronsin C, Muscatelli F, Mattei MG and Breathnach R . (1993). Oncogene, 8, 1195–1202.

  33. Rubin JS, Bottaro DP and Aaronson SA . (1993). Biochem. Biophys. Acta, 1155, 357–371.

  34. Santoro MM, Collesi C, Grisendi S, Gaudino G and Comoglio PM . (1996). Mol. Cell. Biol., 16, 7072–7083.

  35. Santoro MM, Penengo L, Minetto M, Orecchia S, Cilli M and Gaudino G . (1998). Oncogene, 17, 741–749.

  36. Skeel A, Yoshimura T, Showalter SD, Tanaka S, Appella E and Leonard EJ . (1991). J. Exp. Med., 173, 1227–1234.

  37. Vande Woude GF, Jeffers M, Cortner J, Alvord G, Tsarfaty I and Resau J . (1997). Ciba Found. Symp., 212, 119–130.

  38. Wang M-H, Dlugosz AA, Sun Y, Suda T, Skeel A and Leonard EJ . (1996). Exp. Cell Res., 226, 39–46.

  39. Wang M-H, Kurtz AL and Chen Y-Q . (2000). Carcinogenesis, 21, 1507–1512.

  40. Wang M-H, Ronsin C, Gesnel MC, Coupeym L, Skeel A, Leonard EJ and Breathnach R . (1994). Science, 266, 117–119.

  41. Wang M-H, Wang D and Chen Y-Q . (2003). Carcinogenesis, 24, 1291–1300.

  42. Williams AC, Harper SJ and Paraskeva C . (1990). Cancer Res., 50, 4724–4730.

  43. Williams TA, Longati P, Pugliese L, Gual P, Bardelli A and Michieli P . (1999). J. Cell. Physiol., 181, 507–514.

  44. Yokota J . (2000). Carcinogenesis, 21, 497–503.

  45. Zhou Y-Q, He C, Chen Y-Q, Wang D and Wang M-H . (2003). Oncogene, 22, 186–197.

Download references

Acknowledgements

This work was performed in part at the University of Colorado School of Medicine. We thank Drs C Gespach (Hôpital Saint-Antoine, Paris, France) for AA/C1 cells; G Gaudino (Universita di Torino, Novara, Italy) for the RONm1254t cDNA; and EJ Leonard (National Cancer Institute, Frederick, MD, USA) for human MSP. We also thank Dr JS Lindsey (Texas Tech University Health Sciences Center) for critical reading of the manuscript. This work was supported by NIH Grants R01 CA91980, Amarillo Area Foundation, and Foundation of Cheung Kong Scholars Program from the Ministry of Education, PR China.

Author information

Correspondence to Ming-Hai Wang.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • colon cancer
  • oncogenic phenotype
  • receptor tyrosine kinase
  • RNA interference

Further reading