c-Src inactivation reduces renal epithelial cell-matrix adhesion, proliferation, and cyst formation

Abstract

c-Src is a non-receptor tyrosine kinase whose activity is induced by phosphorylation at Y418 and translocation from the cytoplasm to the cell membrane. Increased activity of c-Src has been associated with cell proliferation, matrix adhesion, motility, and apoptosis in tumors. Immunohistochemistry suggested that activated (pY⁴¹⁸)-Src activity is increased in cyst-lining autosomal dominant polycystic kidney disease (ADPKD) epithelial cells in human and mouse ADPKD. Western blot analysis showed that SKI-606 (Wyeth) is a specific inhibitor of pY⁴¹⁸-Src without demonstrable effects on epidermal growth factor receptor or ErbB2 activity in renal epithelia. In vitro studies on mouse inner medullary collecting duct (mIMCD) cells and human ADPKD cyst-lining epithelial cells showed that SKI-606 inhibited epithelial cell proliferation over a 24-h time frame. In addition, SKI-606 treatment caused a striking statistically significant decrease in adhesion of mIMCD and human ADPKD to extracellular collagen matrix. Retained viability of unattached cells was consistent with a primary effect on epithelial cell anchorage dependence mediated by the loss of extracellular matrix (ECM)-attachment due to α₂β₁-integrin function. SKI-606-mediated attenuation of the human ADPKD hyperproliferative and hyper-ECM-adhesive epithelial cell phenotype in vitro was paralleled by retardation of the renal cystic phenotype of Pkd1 orthologous ADPKD heterozygous mice in vivo. This suggests that SKI-606 has dual effects on cystic epithelial cell proliferation and ECM adhesion and may have therapeutic potential for ADPKD patients.

c-src is the signature member of the Src family of non-receptor tyrosine kinases that play key roles in cell attachment, motility, proliferation, differentiation, and survival (16, 28, 38). c-Src contains a protein-tyrosine kinase domain adjacent to a COOH-terminal regulatory tail, four Src homology (SH) domains, a unique segment, and an NH₂-terminal 14-carbon myristoyl (myr) sequence. When c-Src is autophosphorylated on tyrosine (Y)⁴¹⁸, the protein is allowed to unfold and become catalytically active. Inactive c-Src is mostly located in the perinuclear region of the cell, whereas the activated protein translocates to the cell membrane, to which it becomes tethered by its NH₂-terminal myr sequence (9). c-Src can be activated in response to stimulation of many receptor tyrosine kinases, including epidermal growth factor (EGF) receptor (EGFR), platelet-derived (PD)GFR, and insulin-like-1 (IGF-1R) as well as by integrins, cell adhesion complexes, G protein-coupled receptors, and cytokines (17).

Autosomal dominant polycystic kidney disease (ADPKD) is a very common monogenetic disease of the kidney caused by mutations in the PKD1 gene in 85% of cases or in the PKD2 gene in 15%. The resultant abnormalities in expression and function of the respective encoded proteins, polycystin (PC)-1 or PC-2, lead to aberrant development of the kidney, with a loss of control of normal tubular lumen diameters leading to cystic dilation due to epithelial hyperproliferation and abnormalities in ion and fluid secretion, epithelial cell polarity, and cell-matrix interactions (29, 36, 45, 47). PC-1 is developmentally regulated and highly expressed in the ureteric bud-derived epithelia of fetal kidneys, where it forms multiprotein complexes with many proteins at the cell membrane at sites of mechanosensation and transduction in the apical primary cilium, at cell-cell adherent junctions, and at the focal adhesions at the cell-matrix interface (45).

Focal adhesion complexes are sites of interaction of many cellular proteins that function to translate integrin engagement through intracellular signaling into cell-matrix attachment, spreading, and motility (4, 17). c-Src forms an active complex with focal adhesion kinase (FAK) following cellular integrin engagement at the extracellular matrix (ECM) or after ligand stimulation of tyrosine kinase receptors by EGF or PDGF. The resulting autophosphorylation of FAK at Y³⁹⁷ provides a nexus for recruitment of c-Src and other SH2-containing signaling molecules and leads to the activation of c-Src and phosphorylation of the adhesion-related cytoskeletal adaptor proteins paxillin and pl30cas (12, 35). In patients with ADPKD, FAK/Src interactions are abnormal since there is failure to recruit FAK to focal adhesion complexes, associated with a loss of FAK-Y³⁹⁷ autophosphorylation, increased cell-matrix adhesion via α₂β₁-integrin receptors, and decreased growth factor-mediated directional migration (29, 45, 50).

Src activity is significantly elevated in many human epithelial cancers (24, 39, 40, 42, 44), where it is often associated not only with cell proliferation but also with tumor cell migration (9, 14, 21). Correspondingly, both Src-negative (^−/−) and FAK^−/− fibroblasts display impaired cell migration (5, 15). Activation of c-Src leads to downstream activation of the Ras/MEK/ERK pathway, which has also been implicated in both proliferation and cell motility (33, 34). Src has also been shown to contribute to the regulation of cytoskeletal dynamics (4, 7). For instance, FAK/Src complex formation is important for the release of cytoskeletal tension, which, during cell spreading, stimulates plasma-membrane protrusion and inhibits cytoskeletal contractility (1, 2).

Several studies using various Src kinase inhibitors and dominant-negative mutant constructs in tumor cells have shown that inhibition of c-Src activity can block cell proliferation and decrease cell migration associated with metastasis, thereby suggesting c-Src as an attractive molecular target for anticancer therapy (20, 32). Since the progressive expansion of cysts in ADPKD is characterized by increased epithelial cell proliferation and increased cell-matrix adhesion coupled with abnormalities in cellular migration (8, 26, 29, 45, 49), this study was designed to examine the role of c-Src in these processes with a view to assessment of its value as a therapeutic target. We show that the Wyeth c-Src inhibitor SKI-606 inhibits mouse renal collecting duct and human ADPKD cyst-lining epithelial cell proliferation and reduces cell-matrix adhesion in vitro and can retard cystic expansion in a Pkd1 heterozygous mouse model of ADPKD in vivo.

METHODS

Cell lines and tissues.

The mouse inner medullary collecting duct (mIMCD) cell line was procured from American Type Culture Collection and cultured in Dulbecco's modified Eagle's medium-F-12 containing 4.5 g/l glucose and 10% fetal bovine serum (FBS), 50 U/ml penicillin, and 50 μg/ml streptomycin. Normal and ADPKD kidney tissues were procured at source by National Disease Research Interchange and prepared for microdissection; primary and conditionally immortalized [temperature sensitive (ts)] epithelial cell culture and characterization are as previously described (30, 46, 48). ADPKD cells were cultured in HEPES-buffered Click-RPMI media (Quality Biologicals, Gaithersburg, MD) containing transferrin (5 μg/ml, Sigma, St. Louis, MO), dexamethasone (5 × 10⁻⁸ M, Sigma), insulin (5 μg/ml, BD Biosciences, Bedford, MA), tri-iodothyronine (5 × 10⁻¹² M, Sigma), and 1% FBS (Crystalgen, Plainview, NJ) (48). Conditionally immortalized ts-ADPKD cells were grown at 33°C and transferred to the nonpermissive 37°C for 3 days for differentiation before experimentation.

Pkd1 null heterozygous (^+/−) mice (23) were backcrossed onto the C57/BL6 background to reduce variability and accelerate cystic development that had previously been noted after 12–16 mo of age (Dr. J. Zhou, personal communication). For characterization studies, at least six animals [Pkd1^+/− and wild-type (WT) littermate] were euthanized for pathology, cystic, and marker analysis at 2, 3, 4, 5, 6, 8, 11, and 12 mo of age. Renal cysts (defined as a >3-fold increase in tubule lumen diameter) were seen in some mice by 3 mo of age and in all mice by 5 mo of age, and this development was accompanied by an increase in renal weight (WT 0.45 ± 0.03 g vs. Pkd1^+/− 0.64 ± 0.06 g, n = 10). Some mice were monitored for longer, but since by 14 mo 20% had died, all animals were euthanized at 16 mo of age. All animal studies were approved by the Institutional Animal Care and Use Committee of Mount Sinai School of Medicine.

Cell and animal treatments.

mIMCD and ADPKD cells were plated at confluent, 75% or 50% subconfluent density in standard cell-type media, in media containing 2.5 μM SKI-606 (Wyeth, generous gift of Dr. E. Avner, Medical College of Wisconsin), dissolved in minimal volume of 0.05% DMSO; or in media containing the vehicle (0.05% DMSO) alone. Wild-type (WT, ^+/+) and heterozygous Pkd1 (^+/−) mice were identified by genomic PCR, and six groups of five animals were treated for 3 mo from 5 mo of age and euthanized at 8 mo, and from 8 mo of age and euthanized at 11 mo. Pkd1 (^+/−) and WT (^+/+) littermates were treated with SKI-606 (30 mg·kg⁻¹·day⁻¹) or vehicle alone (0.2% DMSO) in their daily drinking water and compared with untreated mice. After 3 mo, the mice were euthanized and the kidneys were fixed for 4 h at 4°C in 4% paraformaldehyde (electron microscopy grade, EM Sciences, Pasadena, CA) in phosphate-buffered saline (PBS, pH 7.4), dehydrated, and embedded in low-temperature paraffin before sectioning at 5 μm.

Cell-matrix adhesion assays.

mIMCD and ADPKD cells were subjected to matrix attachment assays for 2 to 4 h. A total of 5,000–8,000 cells per well in 96- or 24-well plates were seeded onto collagen type I matrix (300 μg/ml, BD Biosciences) in 100 μl of the media in the presence or absence of SKI-606 or DMSO vehicle. After 4 h of attachment, unattached cells were gently removed by aspiration, and numbers of attached cells were determined either by cell counting of fixed, hematoxylin and eosin (H&E)-stained culture plates or by an automated colorimetric plate reader assay after 4 h of incubation in MTS (Cell Titer 96 AQ, Promega, Madison, WI). Color development was read at 492 nm (model 550, Bio-Rad, Bucks, UK) and values were subtracted from parallel control plates lacking cells. Each experiment was repeated three to six times with four to eight replicates of each condition. In addition, unattached cells released into the media were collected by aspiration and low-speed centrifugation and were counted using a hemocytometer. Attached cells from the same plates were collected by trypsinization and counted, total cell numbers were summed, and percentages of unattached floating cells were calculated.

Cell viability assays.

Trypan blue exclusion assays were carried out to determine viability of attached and unattached cells. Trypan blue (0.1 ml; 0.4%) was added to 0.5 ml of cell suspension and left to stand at room temperature for 30 min before hemocytometer counting of viable (unstained) and nonviable (blue) cells.

Proliferation assays.

Multiple replicates of 2,000 cells were plated in 100 μl per well in 96-well plates and left to attach in standard media for 24 h and in serum-free media for a further 24 h. Cell proliferation was then measured in the presence or absence of SKI-606 (2.5 μM) or vehicle (DMSO, 0.05%) at 4, 8, and 24 h by the addition of 20 μl MTS (Cell Titer 96 AQ), incubation at 37°C for 4 h, microplate reader determination of color development at 492 nm, and subtraction of no cell blanks. These experiments were repeated three times with eight replicates in each group.

Western immunoblot analysis.

Total cell protein extracts were prepared by lysis in buffer containing 1% Triton X-100, 150 mM NaCl, 10 mM Tris, pH 7.2, 1 mM EGTA, 1 mM EDTA, and a comprehensive protease and phosphatase inhibitor cocktail (Sigma). Protein concentrations were determined using the Bradford assay (Pierce). Cell proteins were separated by 12% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE). Lysate (20–50 μg) was loaded per lane, and separated proteins were transferred to nitrocellulose or polyvinylidene difluoride membranes. After blocking in nonfat milk in TBST (10 mM Tris base, 150 mM NaCl, 0.1% Tween 20, pH 8) or 5% BSA, membranes were incubated in primary antibody for 1.5 h at room temperature or overnight at 4°C and washed in TBST. The primary antibodies used were as follows: anti-Src (1:1,500, mouse monoclonal, Upstate Biotechnologies, Lake Placid, NY), anti-pY⁴¹⁸-Src (1:1,000, rabbit polyclonal, Biosource, Camarillo, CA), anti-EGFR (1:1,000, rabbit polyclonal, Cell Signaling, Danvers, MA), anti-pY¹⁰⁶⁸ EGFR (1:500, mouse monoclonal, Cell Signaling); anti-ErbB2 (1:1,000, rabbit, AbCam, Cambridge, UK), anti-pY⁸⁷⁷ and ^1021/2 ErbB2 (1:500 rabbit, Cell Signaling), and anti-β-actin (1:5,000, rabbit, Abcam). Conjugated horseradish peroxidase-coupled goat anti-mouse (1:2,000, Pierce) or goat anti-rabbit IgG (1:40,000, Sigma) were used as secondary antibodies and immunoreactivity detected by enhanced chemiluminescence (ECL, Thermoscientific, Rockford, IL).

Immunohistochemistry.

Paraffin sections (5 μm) of age-matched human and mouse normal and ADPKD kidneys, fixed for 4 h at 4°C in 4% paraformaldehyde (EM Sciences) in PBS, pH 7.4, were subjected to optimized immunohistochemistry, as described previously (29) using an optimized indirect, avidin biotin-enhanced technique and aminoethylcarbazole detection yielding a red reaction product (Vectastain, Vector Laboratories). Anti-Src (1:150, GD11 monoclonal, Upstate) and anti-pY⁴¹⁸-Src (1:100); AQP-1 (1:500, Alpha Diagnostic, Caguas); calbindin (mouse monoclonal D28K, 1:300, Sigma), and AQP-2 (1:75, Alpha Diagnostic) were used and compared with matched controls lacking primary antibody or incubated with preabsorbed and irrelevant antibodies. Some sections were also stained with the peroxidase-coupled lectins DBA, PNA, and LTA (Sigma) or for alkaline phosphatase using the nitro blue tetrazolium (NBT-BCIP) technique (Vector).

Microscopy.

Sections were viewed under a Nikon Eclipse 800s microscope by bright field, or differential interference contrast (DIC) optics. At least 10 random fields were examined per section and 10 sections per block of normal, WT, or cystic kidneys for photography and counting of cysts (> 3-fold expansion of normal tubule diameter).

Statistics.

Data were collected and plotted as means ± SE to compare pairwise the difference in the untreated and treated groups of cells. Statistical analysis was performed using Student's t-test or analysis of variance, and significance was determined as a P value. P < 0.05 was considered significant.

RESULTS

Activated c-Src is increased in human and mouse ADPKD cyst-lining epithelia.

Since an increase in expression and activation of c-Src by phosphorylation on its autophosphorylation site at Y⁴¹⁸ is associated with increased cell proliferation, spreading, and migration in numerous tumors and cancer cell lines, we first analyzed its distribution and activation in human and mouse ADPKD. Immunohistochemistry was carried out using well-characterized, anti-pY⁴¹⁸-Src antibodies (Fig. 1) and demonstrated strongly increased levels of active c-Src in cyst-lining ADPKD epithelial cells when compared with age-matched normal human kidneys (Fig. 1, B vs. A) as well as in C57/Pkd1 heterozygous (^+/−) compared with WT (^+/+) littermate mouse kidneys from 5 mo of age (Fig. 1, D vs. C).

Fig. 1.

A–D: immunohistochemistry of pY⁴¹⁸-Src in normal human adult (46-yr-old) kidney (A); human end-stage-autosomal dominant polycystic kidney disease (ADPKD; 46-yr-old) kidney (B); normal adult (5-mo-old) mouse kidney (C); and C57/Pkd1^+/− 5-mo-old mouse kidney (D). Bar, 10 μm. Red reaction product.

SKI-606 is a specific inhibitor of activated (p)-Src.

The recent development of cell-permeable, orally available, small molecule inhibitors of tyrosine kinases has led to advances in the analysis, understanding, and inhibition of proliferative conditions such as cancer. To determine the profile of specificity of the c-Src inhibitor SKI-606, with regard to ADPKD-related proliferative pathways, we carried out comparative Western blot analysis mIMCD cells (Fig. 2A). SKI-606 (2.5 μM in 0.05% DMSO-containing media) (lane 3 vs. lane 1), but not of DMSO (0.05%) vehicle alone (lane 2), resulted in specific and complete inhibition of active pY⁴¹⁸-Src but had no effect on the levels of expression of the other important ADPKD-related receptor tyrosine kinases, EGFR or ErbB2. Levels of total c-Src protein, which would include active (phosphorylated) and inactive c-Src, were unchanged. SKI-606 also inhibited pY⁴¹⁸-Src in cultured human ADPKD cells (Fig. 2B, lane 2 vs. lane 1).

Fig. 2.

A: Western blot of effects of SKI-606 on mouse inner medullary collecting duct (mIMCD) cell proteins. Lane 1, untreated; lane 2, treated with 0.05% DMSO vehicle; lane 3, treated with SKI-606. EGFR, epidermal growth factor receptor. B: Western blot of effects of SKI-606 on cultured human ADPKD cell proteins. Lane 1, untreated; lane 2, treated with SKI-606.

Inhibition of active c-Src reduces normal renal tubule epithelial cell-matrix attachment and proliferation.

Renal epithelial cell-matrix adhesion is essential for and mediates subsequent proliferative responses during normal tubule development and differentiation in vitro and in vivo. Matrix adhesion assays were carried out to determine the role of activated c-Src in normal renal epithelial cells by quantitative comparison of adherent cell numbers after incubation for 4 h in the presence or absence of the c-Src inhibitor SKI-606 (Fig. 3A). Although this compound is cell permeable and orally available, it must be solubilized in low concentrations (0.05%) of DMSO vehicle. Control studies showed that 0.05% DMSO alone caused no statistically significant alteration in adhesion properties (Fig. 3A, bar 2 compared with bar 1) while the presence of SKI-606 resulted in a statistically significant decrease in cell-matrix attachment (Fig. 3A, bar 3 compared with bars 1 and 2; P < 0.05). This result was confirmed by the demonstration of statistically significant increases of nonadherent floating cells collected from the media of SKI-606-treated cells (26 ± 6%) compared with DMSO (3.8 ± 0.4%) vehicle-treated or untreated cells (4.2 ± 0.7%; P < 0.05) (Fig. 3B). In addition, Trypan blue exclusion viability studies demonstrated that these effects were not due to a toxic or apoptotic effect of SKI-606 since only low levels (15.2 ± 3.5%) of nonviable mIMCD cells were seen in the unattached populations after 2 h of treatment with SKI-606, a value that was not significantly different from those measured for unattached mIMCD cells before treatment (12.5 ± 2.8%). These results are consistent with a specific effect of pY⁴¹⁸-Src inhibition on cell-matrix attachment (anchorage dependence) of normal renal epithelial cells.

Fig. 3.

Effects of SKI-606 on mIMCD cell adhesion to matrix. A: attached cells. B: unattached cells from culture media. White bars, untreated; gray bars, treated with 0.05% DMSO vehicle; black bars, treated with 2.5 μM SKI-606. 4-h Attachment to type I collagen. SRCI, c-Src inhibitor. *P < 0.05.

Longer-term studies were carried out to determine the effects of SKI-606 on proliferation in serum-starved mIMCD cells over a 24-h period (Fig. 4). These experiments demonstrated a statistically significant decrease in mIMCD cell proliferation after 8 h of treatment that was exacerbated after 24 h. DMSO vehicle alone had no effect of mIMCD cell proliferation under identical conditions.

Fig. 4.

Effects of SKI-606 on mIMCD cell proliferation. All cells were serum starved for 24 h before plating of 2,000 cells per well. Colorimetric MTS assay; 8 replicates per condition; 3 experiments. *P < 0.05.

Inhibition of active c-Src reduces ADPKD cyst epithelial cell-matrix attachment and proliferation.

Since active pY⁴¹⁸-Src is apparently highly expressed in ADPKD epithelia (Fig. 1) and human ADPKD epithelia had previously been shown to exhibit increased cell-matrix adhesion (29, 49) and increased epithelial cell proliferation (8), we next sought to determine whether SKI-606 could normalize the human ADPKD cell phenotype in vitro. Five thousand cells were plated per well, allowed to settle, and serum starved overnight before addition of the compound. Interestingly, significant decreases in ADPKD epithelial cell-matrix adhesion were seen after 4 h of attachment in the presence of 2.5 μM SKI-606 (bar 3 compared with bars 1 or 2, Fig. 5A). Corresponding significant increases in unattached ADPKD cells were measured in SKI-606-treated ADPKD cells compared with vehicle or untreated control cultures (bar 3 compared with bars 1 or 2, Fig. 5B). Trypan exclusion assessment of viability detected low levels of cell death (12.5 ± 3.1%) in these floating SKI-606-treated cells that were not significantly different from those values calculated for untreated controls (17.1 ± 2.5%). These results are again consistent with an effect of SKI-606 on ADPKD cell anchorage-dependence (loss of ECM attachment). In addition, 24 h studies on serum-starved monolayers showed an effect of SKI-606 on human ADPKD epithelial cell proliferation which was significantly reduced after 24 h of treatment (Fig. 6).

Fig. 5.

Effects of SKI-606 on human ADPKD epithelial cell adhesion to matrix. A: attached cells. B: unattached cells from culture media. Bar 1, untreated; bar 2, treated with 0.05% DMSO vehicle; bar 3, treated with 2.5 μM SKI-606. 4-h Attachment to type I collagen. *P < 0.05.

Fig. 6.

Effects of SKI-606 on human ADPKD epithelial cell proliferation. All cells were serum starved for 24 h before plating of 2,000 cells per well. White bars, treated with 0.05% DMSO vehicle; black bars, treated with 2.5 μM SKI-606. Colorimetric MTS assay; 8 replicates per condition; 3 experiments. *P < 0.05.

Inhibition of active c-Src reduces cystic development in ADPKD mice.

Although complete knockout of the PKD1 gene and PC-1 protein causes embryonic lethality in mice and humans, its haplo-insufficiency and loss of function results in progressive renal cystic disease. Modification of an orthologous Pkd1-null mouse (23) by backcrossing onto a homogeneous C57/BL6 background and characterization of the heterozygous offspring has led to the development of a Pkd1 mouse model that closely recapitulates the progressive time scale and phenotypic properties of renal cystic (and interstitial fibrotic) development seen in human ADPKD (Fig. 7, A–C and L). By 6 mo of age, 100% of C57/Pkd1^+/− mice had cystic kidneys with an average of 2-fold increases in kidney volume and 1.5-fold increases in kidney weight compared with their unaffected littermates. Immunohistochemical marker analysis demonstrated the presence of aquaporin 1 (AQP-1), alkaline phosphatase, calbindin, DBA, and AQP-2-positive cysts (Fig. 7, D–K, and data not shown), reflecting the patterns seen in human ADPKD in which cysts are derived from every segment of the nephron (45, 47).

Fig. 7.

Characterization of heterozygous C57/Pkd1 mouse model of ADPKD. A, D, and G: unaffected littermates. B, E, and H: 8-mo-old Pkd1^+/− mice. C, F, I, J, K, and L: 12-mo-old Pkd1^+/− mice. A, B, and C: cultures stained with hematoxylin and eosin (H&E); D, E, and F: immunohistochemistry of aquaporin-1. Red reaction product. G, H, and I: immunohistochemistry of aquaporin-2. Red reaction product. J: nitro blue tetrazolium-5-bromo-4-chloro-3-indolyl phosphate (NBT-BCIP) staining of alkaline phosphatase. Blue/purple reaction product. K: DBA- peroxidase staining of collecting duct lectin. Red reaction product. L: Masson's trichrome staining of fibrosis. Blue reaction product. Bars, 25 μm (A, D, G, and J) and 10 μm (B, C, E, F, H, I, K, and L).

Treatment of C57/Pkd1^+/− mice with the pY⁴¹⁸-Src inhibitor SKI-606 for 3 mo from 5 to 8 mo of age resulted in significant reduction of cyst numbers (defined as a ≥3-fold increase in lumen diameter) by comparison to untreated or vehicle-treated controls (untreated 89.45 ± 6.5; DMSO-treated 92.64 ± 7.9; SKI-606-treated 22.35 ± 4.2; P < 0.05 treatment vs. DMSO or treatment vs. untreated). Assessment was carried out by two blinded individuals and 10 fields were examined per section, with 5 sections per kidney from 5 mice per condition. Cystic reduction in the SKI-606-treated animals was associated with the normalization of kidney morphology (Fig. 8), volume, and weight (WT at 8 mo: untreated, 0.50 ± 0.05 g; vehicle-treated, 0.47 ± 0.03 g; and SKI 606-treated, 0.51 ± 0.08 g, vs. Pkd1 at 8 mo: untreated, 0.74 ± 0.06 g; vehicle-treated, 0.72 ± 0.07 g; and SKI 606-treated, 0.51 ± 0.08 g).

Fig. 8.

Effects of SKI-606 on ADPKD cyst expansion in heterozygous C57/Pkd1 mouse model of ADPKD. A: unaffected littermate kidney. H&E. B: Pkd1^+/− untreated mouse kidney untreated. H&E. C: Pkd1^+/− SKI-606-treated mouse kidney. H&E. All mice were 8 mo of age and treatments were 3 mo in duration. Bars, 10 μm.

DISCUSSION

Src was first identified as an oncogene and subsequently the cellular Src family of kinases has been implicated in cell growth regulation, particularly through the activation of the extracellular regulated kinase (MEK/ERK)/mitogen-activated kinase (MAPK) signaling pathway (31) and shown to be activated via cell-matrix adhesion or growth factor receptor activation. Both integrin-dependent adhesion- and growth factor induced-signaling leads to Src-mediated phosphorylation of FAK at Y⁹²⁵ that then creates a binding site for the SH2 domain of the adaptor protein Grb2 thus coupling Src to the activation of the Ras/MEK/ERK pathway (34). ADPKD is characterized by progressive expansion of renal tubular cysts due in part to increased cell proliferation and fluid secretion associated with abnormalities in cytoskeletal organization, cell-matrix adhesion, and growth factor-dependent migration (45). Increased EGFR, ErbB2, and ERK expression and activation are also characteristic of expanding ADPKD cystic epithelia (6, 8, 36, 51). Previous studies have shown that inhibition of EGFR or ErbB2 can inhibit ADPKD cystic epithelial cell proliferation cell in vitro and retard cystic development in mouse models of ADPKD and autosomal recessive (AR)PKD in vivo (37, 51). Interestingly, c-Src has been shown to physically interact with the ErbB2 catalytic domain and to modulate cell polarity (19), noted phenotypic characteristics of ADPKD epithelia. The potential ability of c-Src inhibitors to inhibit both proliferation and cell-matrix adhesion, with an impact on growth-factor-mediated migration, provides an attractive proposition to target several components of the ADPKD cell phenotype.

Our studies show that c-Src activity is increased and translocated to the apical cell plasma-membranes of epithelial cells lining ADPKD cysts in human and mouse kidneys and that inhibition of this activity by the specific inhibitor SKI-606 reduces cystic enlargement in the orthologous C57/Pkd1 mouse model of ADPKD. The PKD1-encoded PC-1 is a large membrane protein with a long extracellular NH₂-terminal portion, 11 transmembrane domains, and a short intracellular COOH-terminal portion that contains many protein binding and phosphorylation sites that can be activated by G proteins, serine/threonine, and tyrosine kinases including c-Src at Y4237 and FAK at Y4127 (22, 27).

Our in vitro studies showed that, at the cellular level, in addition to the expected ability to inhibit renal epithelial cell proliferation, a primary mechanism of action of SKI-606 was to reduce epithelial cell-ECM adhesion (anchorage dependence) without significant early loss of viability or direct toxicity. Loss of ECM attachment occurs as a result of alterations in integrin function (41) and has previously been associated with alterations of α₂β₁-integrin content and distribution in ADPKD (3, 29, 50). The PKD1-encoded protein PC-1 forms multimolecular complexes with many membrane proteins in normal renal epithelia, including α₂β₁-integrin and components of the focal adhesion/cytoskeletal complex including c-Src, FAK, and paxillin (10, 50). Dynamic interactions of the polycystin complex with cell-ECM, cell-cell adhesion, and cilial complexes that interface with the actin cytoskeleton are thought to confer mechanosensory function and mediate transduction of extracellular information into the cell by eliciting intracellular signaling cascades resulting in gene transcription. Integration of extracellular signals from these basal, lateral, and apical cell membrane sites results in regulation of cell shape, differentiation, and lumen diameter during renal development (45).

ADPKD epithelial cells show several phenotypic characteristics that distinguish them from their normal renal tubular epithelial cell counterparts. ADPKD epithelia are more proliferative in response to growth factors, and second messengers including EGF and cAMP (8, 13) are more adherent to ECM proteins, particularly collagen type I and IV (49), and are less migratory in response to growth factor gradients (29). In vivo, PKD1 knockout results in disruption of ureteric bud branching morphogenesis and faulty renal tubule differentiation leading to cystic dilation (3, 29, 36). It is well established that overexpression or overactivation of c-Src is often associated with carcinogenesis and plays multiple roles in the regulation of proliferation, cell-ECM adhesion, invasion, and motility (52). Furthermore, the c-Src tyrosine kinase has been shown to interact with the catalytic domain of ErbB2, another major mediator of increased proliferation in ADPKD and tumors (19, 53). This led to the hypothesis that specific inhibition of activated pY⁴¹⁸-Src could be therapeutic for ADPKD and selection of the SKI-606 inhibitor. Since c-Src is the only member of the Src kinase family that is significantly increased in ADPKD (data not shown), it is unlikely that other potential targets for this drug such as Fyn, Yes, or Lyn play a significant role in its ability to restore normal cell function to ADPKD epithelia in vitro and retard cystic expansion in vivo (11).

This compound has previously been found to retard cystic expansion in rodent models of ARPKD including the phenotypic, nonorthologous bpk mouse model and the genotypic orthologous Pck rat model of ARPKD (37) where its mechanism of action was concluded to be antiproliferative via the ERK/MAPK pathway. ARPKD in human patients is caused by mutational loss of the PKHD1 gene and shares some phenotypic features with ADPKD, particularly with regard to hyperproliferation of collecting duct proliferation and morphogenesis (25). Recent studies have also shown that ARPKD epithelial are characterized by hyperactive pY⁴¹⁸-Src, increased ECM adhesion, and reduced growth-factor-mediated directional migration (18) and that the PKHD1-encoded protein fibrocystin-1 participates in the PC-2/PC-1/integrin multiprotein complex (43). Furthermore, the PC-1-protein has been shown to be phosphorylated on Y4237 in its intracellular tail in normal renal collecting tubule cells, where phospho-activation of c-Src has also been detected, particularly in fetal stages of development (10, 22). Taken together, these results suggest a role for overactivation of c-Src in rodent models of both ARPKD and ADPKD.

Our studies suggest that SKI-606 may be a dual function inhibitor of renal epithelial cell proliferation and matrix-mediated cell anchorage and that it may provide a promising therapeutic agent for the future retardation therapy of cystic expansion in ADPKD.

GRANTS

This work was supported by National Institutes of Health Grant P01DK62345 (to P. D. Wilson).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).