Inhibition of Src Kinase Blocks High Glucose–Induced EGFR Transactivation and Collagen Synthesis in Mesangial Cells and Prevents Diabetic Nephropathy in Mice

Primary rat glomerular mesangial cells (passages 8–12) were isolated, characterized, and grown as described (9). At 70–80% confluence, cells were growth-arrested in Dulbecco’s modified Eagle’s medium (DMEM) containing 0.1% FBS, and 5.6 mmol/L (normal glucose) or 25 mmol/L (high glucose) d-glucose or normal glucose plus 19.4 mmol/L mannitol as an osmotic control. For inhibitor studies, cells were treated as follows: PP2 (2 μmol/L) and SU6656 (2.5 μmol/L), TAPI-2 (100 μmol/L) (Calbiochem, San Diego, CA). For experiments with 48-h exposure to high glucose, PP2 and SU6656 were added for the final 24 h. For time course studies of 24 h or less, these inhibitors were added 1 h before high glucose. TAPI-2 and AG1478 were added 1 h before high glucose in all experiments. AG1478 (200 nmol/L) (Biomol, Plymouth Meeting, PA) was added 30 min before EGF. All inhibitors were dissolved in DMSO. Control cells received an equal amount of DMSO.

A stealth negative universal control scrambled (Src), two different Src-specific Stealth RNAi duplex oligoribonucleotides (Src-RSS331230-1), and a Fyn-specific Stealth RNAi duplex oligoribonucleotide (Fyn-RSS303099) were predesigned (Invitrogen). Reverse transfections were performed according to the manufacturer’s protocol. Briefly, small interfering RNA (siRNA) (5 nmol/L) was mixed with lipofectamine RNAiMax in serum-free Opti-MEM and incubated for 20 min at room temperature. Then, 200 μL was added into each well containing mesangial cells (2 × 10⁵) in 1.8 mL DMEM (10% FBS without antibiotics). After 24 h, cells were growth-arrested in 5.6 mmol/L or 25 mmol/L glucose for 48 h.

Total cell lysates (500 μg) were incubated with 3 μg anti-Src monoclonal antibodies (Cell Signaling), anti-Fyn (Novus Biologicals), or anti-Yes (BD Biosciences) and protein-A Sepharose (GE Healthcare) for 1 h at 4°C. Immune complexes were washed and eluted with cell lysis buffer (120 μL). Gel loading buffer (40 μL of 4× Laemmli buffer) was added, and samples were boiled at 100°C for 5 min. Equal amounts were separated by SDS-PAGE and immunoblotted.

Kidney cortex, mesangial cell total lysates, and membrane fractions were prepared and underwent immunoblotting, as previously described (9). Primary antibodies against SrcpTyr-416, EGFRpTyr-845, EGFRpTyr-1173, phospho-ERK, phospho-p38, and total EGFR, Src, ERK, p38, TACE, Fyn, Yes (Cell Signaling), and collagen IV (Cederlane, Burlington, Ontario, Canada) were used at 1:1,000 dilution. p-ERK (1:3,000) goat anti–HB-EGF (Sc-21591) (1:1,000), β-actin, and horseradish peroxidase–conjugated secondary antibodies were from Santa Cruz Biotechnology, Inc.

Collagen IV and TACE were detected as described (9) by growing mesangial cells on coverslips in six-well plates and staining with primary collagen IV (1:300, Rockland Immunochemicals) or TACE (1:200, Abcam) antibodies, followed by fluorescein isothiocyanate–conjugated goat anti-rabbit IgG (Jackson ImmunoResearch). FM 4-64FX dye, a lipophilic styryl compound (Invitrogen), was used for labeling the plasma membrane of the cells according to the manufacturer’s instructions. Immunofluorescence was detected by confocal microscopy, and the fluorescence intensities of plasma membranes and total cells were analyzed by Scion Image Software (Scion Corporation).

Mesangial cell lysates and kidney cortex homogenates were analyzed for TACE activity using SensoLyte520 TACE Activity Assay Kit (AnaSpec Inc.), following the manufacturer’s instructions. Enzymatic activity was measured using a fluorescence resonance energy transfer peptide containing the fluorescent probe 5-FAM quenched by QXL520. TACE cleaves the peptide and releases 5-FAM, leading to increased fluorescence.

Eight-week-old male DBA2/J mice (20–25 g) (The Jackson Laboratory) received five consecutive daily intraperitoneal injections of STZ (Sigma) (40 mg/kg, freshly prepared in 100 mmol/L citrate buffer, pH 4.5) after a 6-h fast (28). Control mice were injected with buffer alone. Six-hour fasting glucose was monitored after 3 weeks via tail vein using a glucometer (OneTouch Ultra, LifeScan) and was more than 15 mmol/L in all of the mice with diabetes. The mice were divided into four groups: control (n = 22), diabetes (n = 25), control + PP2 (n = 23), or diabetes + PP2 (n = 25). The PP2 groups received 2 mg/kg PP2 dissolved in DMSO three times per week by intraperitoneal injection. At 16 weeks of age, individual mice were placed in metabolic cages to obtain urine collections. Urinary albumin excretion was assessed by the urinary albumin-to-creatinine ratio using the Albuwell M kit and The Creatinine Companion (Exocell, Philadelphia, PA). Mice were killed at 18 to 24 weeks after STZ injection. Animal protocols were approved by the University Health Network Animal Care Committee and carried out in accordance with the guidelines of the Canadian Council of Animal Care.

Aortic root and left ventricular blood pressure measurements were performed in triplicate on a subset of mice (n = five) under isoflurane anesthesia using a Millar pressure-transducing catheter introduced through the right common carotid artery before they were killed, as described (29).

Kidneys were rapidly dissected, fixed overnight in 10% buffered formalin, and embedded in paraffin, and 4-μm sections underwent periodic acid-Schiff (PAS) staining. Glomerular volume and mesangial matrix area relative to total glomerular area were quantified in 20–30 glomeruli from each group by ImageJ software or Image Probe. The thickness of the glomerular basement membrane (GBM) was detected by electron microscopy (Department of Pathology, Mount Sinai Hospital, Toronto, Ontario, Canada). Images were captured by an Olympus ATM1600 camera system, and the thickness of the GBM was analyzed using the ruler function of the software (Olympus XL Docu). The average GBM thickness for each mouse was calculated from 95–225 measurements obtained from different GBM sites, from 9 images of 3 random glomeruli from each mouse.

Mouse glomeruli were isolated as previously described (30). Briefly, mice were anesthetized and perfused through the heart with 40 mL magnetic 4.5-μm diameter Dynabeads (Invitrogen) in PBS. Kidneys were minced into small pieces and digested with collagenase A (Roche) at 37°C for 30 min, filtered with a 100-μm cell strainer, and the glomeruli containing the Dynabeads were collected by a magnetic particle concentrator.

Collagen IV immunohistochemistry was performed on 4-μm sections that were deparaffinized, rehydrated, and blocked with 3% H₂O₂, followed by antigen retrieval in a microwave oven. Staining was done using 1:2,000 rabbit polyclonal collagen IV antibody and visualized using a Super Sensitive Polymer-HRP IHC Detection System/DAB kit (BioGenex) according to the manufacturer’s instructions. Slides were digitized using a bright field scanner (ScanScope XT) and analyzed with Image-Pro software (Media Cybernetics). WT-1 staining was performed using WT-1 antibody (1:100; Santa Cruz Biotechnology). Color development was done with freshly prepared Nova Red solution (Vector Laboratories). Sections were counterstained lightly with Mayer’s hematoxylin, dehydrated in alcohols, cleared in xylene, and mounted in Permount (Fisher Scientific) (31). The podocyte number was determined by counting the average number of positive-staining nuclei per glomerulus, from 40–50 glomeruli/mouse (n = five per group). All pathological analyses were performed blinded and verified by two observers.

The experimental data are reported as means ± SEM. Statistical differences among groups were assessed by ANOVA with a Tukey post hoc test for multiple comparisons. P < 0.05 was considered to be statistically significant.

Treatment of mesangial cells with high glucose resulted in a time-dependent increase in Src Tyr-416 phosphorylation, a marker of Src activation (11,12), determined by immunoprecipitation, followed by immunoblotting with SrcpTyr-416 antibodies (Fig. 1A). This was not due to osmotic effects because mannitol had no effect (Supplementary Fig. 1A). Because Src has been reported to participate in EGFR transactivation, we examined EGFR tyrosine phosphorylation using site-specific antibodies. Phosphorylation of EGFR Tyr-845 was stimulated by high glucose, peaking at 48 h (Supplementary Fig. 1B). Two structurally distinct small molecule Src kinase inhibitors, PP2 (32) and SU6656 (33) used at low concentrations, 2 and 2.5 μmol/L, respectively, blocked high glucose–induced EGFR Tyr-845 phosphorylation (a Src kinase site) (Fig. 1B). As depicted in Fig. 1C and Supplementary Fig. 1C, high glucose also enhanced EGFR phosphorylation at Tyr-1173 (an EGFR autophosphorylation site), and this was also prevented by PP2. Because PP2 does not affect EGF-stimulated EGFR phosphorylation (34 and data not shown), these results indicate that EGFR kinase activity activated by high glucose required Src kinase. EGFRTyr-845 phosphorylation by Src appears to require ligand binding and activation (35). Consistent with this notion, AG1478, an EGFR kinase inhibitor, blunted EGFR Tyr-845 phosphorylation stimulated by high glucose (Fig. 1D). Together, these results suggested that high glucose induced mesangial cell EGFR transactivation through Src kinase stimulation.

Ectodomain shedding of transmembrane EGFR ligands can be evoked by ADAMs (23–27). To investigate the transactivation of EGFR by high glucose, we focused on ADAM-17/TACE because it has been documented to mediate EGFR transactivation by serotonin (25). The hydroxamic acid–based TACE inhibitor, TAPI-2 (36,37), diminished the stimulatory effect of high glucose on EGFR Tyr-845 and Tyr-1173 phosphorylation (Fig. 1E). These results suggest that TACE plays a functional role in EGFR transactivation in this context.

To confirm the role of TACE and determine the requirement of Src, TACE activity and its membrane translocation were assessed. A TACE assay that used a fluorogenic substrate peptide with a TACE cleavage site revealed that high glucose stimulated a fourfold increase of activity, which was blocked by the Src inhibitors PP2 and SU6656 (Fig. 2A). To determine the specific role of Src kinase in high glucose using a genetic approach, mesangial cells were transfected with Src-specific siRNAs (siRNA1, siRNA2) or with a negative universal control siRNA to suppress the expression of endogenous Src. Transfection of siRNA2 decreased Src protein content by 77% (0.33 ± 0.08 compared with scrambled siRNA) and did not alter the SFKs, Fyn, or Yes (Supplementary Fig. 2A and B), and so was used in subsequent experiments. Src siRNA2 inhibited TACE activation in high glucose (Fig. 2B). In addition, immunoblotting (Fig. 2C and D) and confocal immunofluorescence microscopy (Fig. 2E and F) demonstrated increased TACE protein content in cell membrane fractions and TACE translocation from cytosol to the cell membrane in response to high glucose. The plasma membrane compartments were confirmed by immunoblotting of cell fractions for the plasma membrane marker, Na⁺/K⁺ ATPase α1 (Supplementary Fig.2C), and by loading cells with the plasma membrane dye, FM 4–64 Fx, for immunofluorescence (Supplementary Fig. 3). These effects on TACE were attenuated by chemical inhibition of Src kinase (PP2 and SU) and by knockdown with Src-siRNA (Fig. 2). These data, together with the inhibition of EGFR transactivation by blocking TACE, indicate that Src plays a key role in TACE-mediated ectodomain shedding of EGFR ligand in mesangial cells, leading to EGFR transactivation by high glucose.

MAPKs, such as ERK1/2 (ERK) and p38 MAPK, are high glucose effectors that mediate mesangial cell ECM protein synthesis (8). High glucose stimulated a dose-dependent increase in ERK1/2 and p38 phosphorylation (Supplementary Fig 1D). Pretreatment of mesangial cells with either of the Src inhibitors, PP2 or SU6656 (Fig. 3A), eliminated ERK phosphorylation in the presence of high glucose. In accordance with ERK and p38 being downstream of EGFR in the setting of high glucose, antagonists of EGFR kinase, AG1478 and TACE (TAP1-2), all attenuated ERK and p38 phosphorylation elicited by high glucose (Fig. 3B and C).

Of several members of the family of SFKs (11,12), three ubiquitous members, Src, Yes, and Fyn, were expressed in the cultured mesangial cells (Supplementary Fig. 4). Immunoprecipitation with specific antibodies and immunoblotting with anti-Src pY-416 (which cross-reacts with the pY activation sites of other SFKs) revealed a transient activation by high glucose of Fyn (which reached maximum at 3–12 h) and no effect on Yes (Supplementary Fig. 4A–D). However, Fyn-specific siRNA, which had no effect on Src protein expression (Supplementary Fig. 5A), did not block high glucose–induced phosphorylation of EGFR Tyr-1173 or Tyr-845 (Supplementary Fig. 5B) or high glucose–stimulated p38 or ERK1/2 phosphorylation (Supplementary Fig. 5C and D). Consistent with the proposed specificity of Src to activate TACE and release EGFR ligand, the transfection of Src-specific siRNA2 resulted in a significant reduction of EGFR (Fig. 4A) and MAPK (p38 and ERK) (Fig. 4B) phosphorylation in high glucose. Together, these data indicate that in mesangial cells, Src is the major SFK responsible for signaling to MAPK in high glucose.

Excessive glomerular collagen IV deposition consequent to hyperglycemia is one of the key events in the development of diabetic glomerulosclerosis (1–6). To evaluate collagen IV levels, mesangial cells underwent immunofluorescence staining for collagen IV detected by confocal microscopy and immunoblotting of cell lysates. Collagen IV protein expression was increased by high glucose (Fig. 5). This effect was blocked in cells treated with the Src inhibitor, PP2 (Fig. 5A and B), and in cells transfected with Src-specific siRNA (Fig. 5C and D). Inhibition of EGFR with AG1478 or TACE with TAPI-2 also prevented high glucose–stimulated collagen IV expression (Fig. 5A). Thus, analogous to the GPCR-triggered TMS pathway (36–38), high glucose engages MAPKs and stimulates collagen IV synthesis through Src, TACE, and EGFR.

To address the in vivo role of Src, DBA2/J mice, reported to be particularly susceptible to diabetic nephropathy (28,39), were treated with five low-dose injections of STZ as indicated by the Animal Models of Diabetes Complications Consortium protocol (28). PP2 was administered beginning 3 weeks after the STZ injections. Six-hour fasting blood glucose levels 3 weeks after STZ administration were elevated (range 15–33 mmol/L) compared with vehicle-injected control mice (range 8.4–9.8 mmol/L) and were not altered by PP2 administration (Supplementary Table 1A). Glucose levels remained similarly elevated in diabetic mice until they were killed (not shown). Untreated control (nondiabetic) and PP2-treated nondiabetic mice demonstrated similar weight gain, as expected. However, neither untreated nor PP2-treated diabetic groups gained weight after STZ administration (Supplementary Table 1B). Kidney weight-to-body weight ratio was significantly increased only in the untreated diabetic mice (Supplementary Table 1C). To exclude effects of PP2 on blood pressure, aortic and left ventricular pressures were measured as described (29) after 6 weeks of diabetes. No significant differences were observed between PP2 treated and untreated groups (Supplementary Table 1D). The tendency to a lower blood pressure in both diabetic groups was likely due to a hyperglycemia-induced modest degree of fluid loss.

To determine the role of Src in vivo, renal cortical extracts were immunoblotted after 18–24 weeks of diabetes and showed an increase in Src Tyr-416 phosphorylation in total lysates (Fig. 6A) as well as after immunoprecipitation with Src-specific antibodies (Supplementary Fig. 6A), but there was no change in total or Tyr phosphorylated Fyn (Supplementary Fig. 6B and C). The Src Y-416 phosphorylation was abrogated by the Src inhibitor, PP2, for up to 24 h after administration, confirming its vivo efficacy (Fig. 6A). We also observed enhanced EGFR (Tyr-845 and Tyr-1173) phosphorylation in renal cortical membrane fractions from the diabetic mice, which was abolished by PP2 treatment (Figs. 6B and 7C). Furthermore, PP2 diminished the increase in phosphorylation of MAPKs (ERK1/2 and p38) in the renal cortex (Figs. 6D and 7E). Because the renal cortex contains tubules as well as glomeruli, and Src, EGFR, and MAPKs have been reported to be activated in renal tubular cells by angiotensin (40), glomeruli were isolated from these groups of mice, and phosphorylation was assessed by immunoblotting. Diabetes augmented and treatment with PP2 inhibited the phosphorylation of Src Tyr-416 (Supplementary Fig. 7A), EGFR Tyr-1173 (Supplementary Fig. 7B), and p38 and ERK1/2 (Supplementary Fig. 7C and D).

To confirm the role of TACE in vivo, TACE activity was measured in the kidney cortex homogenates. TACE activity was enhanced approximately twofold (Fig. 7A), and immunoblotting demonstrated increased TACE protein accumulation in the membrane fractions (Fig. 7B). These changes were attenuated by treatment with PP2. TACE is known to cleave HB-EGF to release ligand from the cell membrane. Immunoblotting of kidney cortex membranes with anti–full-length/mature HB-EGF revealed a marked decrease in the presence of diabetes, consistent with cleavage and loss of the mature fragment, which was blocked by PP2 administration (Fig. 7C). These data support the activation of a Src-TACE-HB–EGF-EGFR-ERK–signaling pathway in the diabetic kidney.

Urinary albumin excretion, measured as albumin-to-creatinine ratio, was significantly elevated in vehicle-treated diabetic mice and markedly reduced by administration of PP2 (Fig. 7D). Glomerular volume was increased by diabetes, and this was significantly attenuated by PP2 treatment (Supplementary Table 1E). Glomerular ECM protein was evaluated by quantitative analysis of PAS and collagen IV staining. As expected, PAS staining (quantified as mesangial matrix fraction) and collagen IV expression were elevated in the glomeruli of mice with STZ-induced diabetes and significantly diminished by PP2 treatment (Figs. 7E and 8A and B and Supplementary Fig. 8). To assess podocyte injury, the kidney sections were stained with WT-1, and morphological changes were examined by electron microscopy as described in 1research design and methods. The number of podocytes determined by WT-1 staining was decreased by 28.5% in diabetic compared with nondiabetic mice. Treatment with PP2 preserved podocyte number in the diabetic mice (Fig. 8C). In addition, podocyte foot process effacement, along with increased GBM thickness, was found in diabetic mice, and PP2 administration inhibited these changes (Fig. 8D). These data demonstrate that in vivo Src inhibition can block or significantly reduce the manifestations of experimental diabetic nephropathy.

Current interventions for diabetic nephropathy, namely, control of blood glucose and hypertension and blockade of the renin-angiotensin system, are able to hinder but not stop the progressive loss of renal function that frequently accompanies diabetic nephropathy (41). Accordingly, new disease mediators and treatments continue to be sought. In this study, we explored the role of Src, an integrator of signaling pathways relevant to diabetic nephropathy. High glucose increased Src Tyr-416 phosphorylation in cultured mesangial cells as well as in the kidney cortex and glomeruli of mice with type 1 diabetes. Although we found evidence for transient activation of Fyn by high glucose in cultured mesangial cells, there was no evidence of long-term activation of this SFK member in the kidney of the diabetic mice. The functional consequences of Src-specific activation were demonstrated by chemical inhibitors as well as by siRNA-mediated Src knockdown. Thus, activation of TACE, EGFR, MAPK (ERK1/2 and p38), and collagen IV accumulation were all inhibited. Significantly, we showed that high glucose mimicked the TMS pathway, transactivating the EGFR via Src and TACE. High glucose–induced EGFR transactivation was manifested by EGFR Tyr-1173 and Tyr-845 phosphorylation. It should be noted that although EGFR activation may also stimulate Src, these data place Src upstream of EGFR signaling in the context of high glucose. The inhibition of phosphorylation of EGFRY845, a Src phosphorylation site, by the EGFR inhibitor, AG1478, is consistent with a ligand binding–dependent conformational change and activation of EGFR for this phosphorylation to take place (35). Transactivation by high glucose in cultured mesangial cells (10) and endothelin-dependent EGFR transactivation in the renal cortices of rats with type 2 diabetes (42) have been observed. Here, we demonstrated the involvement of Src in EGFR transactivation by high glucose using two distinct chemical inhibitors, PP2 and SU6656, as well as Src-siRNA. The further activation of EGFR, with maximum activation at 48 h (Supplementary Fig. 1) compared with maximum activation of Src at 12 h (Fig. 1), suggests that other downstream effectors of EGFR are also stimulated by high glucose.

TACE is the probable metalloproteinase releasing EGFR ligand in high glucose–treated mesangial cells based on the inhibition by TAPI-2 and the ability of high glucose in vitro and hyperglycemia in vivo to stimulate TACE activity and membrane translocation. In support of this concept, a number of studies have shown that TACE is involved in EGFR transactivation (37,38,43,44), and TACE/ADAM17 was the main sheddase for HB-EGF in a study of embryonic stem cells lacking various ADAMs (26). In mesangial cells, ADAM17, but not other ADAMs, immunoprecipitated with HB-EGF (25). Proteolytic release of a number of different cell surface–tethered EGFR ligands, such as TGF-α, HB-EGF, and amphiregulin, has been described to transactivate EGFRs (27,28,36). Our data in vivo, demonstrating the release of HB-EGF from kidney membranes of diabetic mice, are consistent with HB-EGF cleavage in high glucose. While this study was underway, Uttarwar et al. (10) reported in rat mesangial cells that ADAM17 (TACE) cleaves pro–HB-EGF, leading to high glucose–induced activation of the EGFR.

It has been proposed that TACE activity is upregulated by Ser/Thr phosphorylation of its cytoplasmic domain (43,45,46) by phosphoinositide-dependent kinase 1 (43) or ERK (47). However, in the current study, ERK was activated downstream rather than upstream of EGFR. As a reported target of TGF-β receptors, TACE activation was associated with translocation to the plasma membrane (48), similar to our findings in high glucose. Increased TACE expression, observed in angiotensin-infused mice (44), may also contribute to increased TACE activity. High glucose was also reported to stimulate TACE activity through PKC (27,45), but the role of Src in these studies was not examined. Although previous studies showed that HB-EGF transactivates EGFR in mesangial cells exposed to angiotensin, serotonin, or TGF-β (25,40,46,49), the current study is the first to demonstrate stimulation of TACE and HB-EGF ligand shedding in the diabetic kidney in vivo and its attenuation by inhibition of Src.

The upregulation of mesangial cell collagen IV mediated by Src and EGFR in the setting of high glucose is consistent with other observations. Thus, EGFR was linked to high glucose–induced synthesis of collagen I, which accumulates in late diabetic nephropathy (10). Src has also been implicated as a downstream mediator in TGF-β–stimulated collagen I and in angiotensin II–mediated type I and IV collagen and fibronectin synthesis in mesangial cells (13–15,18). In a recent study, the stimulation by angiotensin II was suggested to involve EGFR transactivation, and similar to our data, Src activation was implicated (40). However, in contrast to these findings in mesangial cells in high glucose, chronic EGFR activation in proximal tubular cells exposed to angiotensin was associated with Tyr-845 phosphorylation but not that of Tyr-1173, suggesting an alternate mechanism of activation (40).

We previously reported that the high glucose–induced increase in collagen IV in mesangial cells required NADPH oxidase (9). This ROS-generating pathway may interact with Src by stimulation of Src activity by NADPH oxidase (14) or, conversely, by an Src-mediated activation of NADPH oxidase (16). Further experiments will be required to define the role of NADPH oxidase and to identify whether Src is upstream and/or downstream of this enzyme in the setting of high glucose. In addition, Src could contribute to glomerular collagen IV accumulation by increasing PAI-1, an inhibitor of ECM protein degradation (7,50). Thus, although Src is a component of multiple signaling pathways that are activated in the kidney in the setting of diabetes (e.g., angiotensin II, TGF-β₁, platelet-derived growth factor, and vascular endothelial growth factor), it has not been implicated as a critical initiator or formally examined as a therapeutic target in diabetic nephropathy.

In STZ-induced diabetic mice, we detected elevated Src Tyr-416 phosphorylation in the kidney cortex and in isolated glomeruli, consistent with reports of increased glomerular Src phosphorylation in diabetic rats (17,18). However, in STZ-induced diabetic rats and mesangial cells exposed to angiotensin II, treatment with an angiotensin II receptor blocker, olmesartan, decreased glomerular Src Tyr-416 phosphorylation (18). In contrast, we observed that although angiotensin II stimulation of Src Tyr-416 phosphorylation in mesangial cells was blocked by losartan, stimulation by high glucose was not (Supplementary Fig. 9). This finding indicates that angiotensin II is not the sole mediator of high glucose signaling to Src. In the current study, we demonstrate for the first time that in vivo inhibition of Src prevents several manifestations of diabetic nephropathy, including albuminuria and glomerular pathologic changes (i.e., mesangial matrix protein and collagen IV accumulation and podocyte depletion). Podocyte loss in diabetic nephropathy has been associated with apoptosis (28). In contrast, Src activation has been found to promote cell survival and proliferation, particularly in the context of cancer (12,22,24,51). It should be noted that the data in this study do not implicate Src as a direct mediator of podocyte loss. Thus, whether Src is activated in podocytes by high glucose and its relationship to podocyte apoptosis in diabetic nephropathy requires further study. The above in vivo findings were accompanied by inhibition of the proposed signaling pathway, namely, a marked decrease in TACE activation and HB-EGF shedding along with EGFR and ERK phosphorylation in diabetic mice treated with PP2.

Although the specific role of Src in cell culture is supported by the siRNA knockdown experiments, the in vivo data with PP2 must be interpreted with caution because PP2 is not completely specific for Src and can inhibit several other kinases (52). Additional experiments with newer, more specific Src inhibitors as well as those compounds with different nonoverlapping kinase targets will be necessary to confirm that Src is a major target to prevent diabetic nephropathy in mice.

In summary, Src activation occurs in mesangial cells exposed to high glucose and in the kidneys of mice with STZ-induced type 1 diabetes. Activated Src mediates a high glucose–stimulated pathway leading to EGFR transactivation, MAPK activation, and collagen IV synthesis. In vivo treatment of type 1 diabetic mice with the Src inhibitor, PP2, prevented albuminuria, glomerular matrix protein accumulation, GBM thickening, and podocyte depletion, which correlate with progressive nephropathy. Thus, because Src may be activated by multiple pathways in the setting of high glucose, it may play a central role in pathogenesis and serve as an effective therapeutic target.

This work was funded by grants from the Canadian Institutes of Health Research (CIHR) MOP 4l9409 and CCD 83025 (I.G.F.), the Canadian Diabetes Association (I.G.F.), and the Japan Diabetes Foundation and Uehara Memorial Foundation (K.T.), and in part by the University of Toronto Fred Sorkin Fund for Diabetes Research. E.A.Y.M. was supported in part by a postdoctoral fellowship from the Canadian Diabetes Association. K.W.K.L., A.S., and L.S. were supported in part by University of Toronto Banting and Best Diabetes Centre Novo Nordisk studentships.

No potential conflicts of interest relevant to this article were reported.

K.T. designed and conducted experiments and analyzed data. L.X. conducted experiments, analyzed data, and wrote the manuscript. H.J.G. analyzed data and wrote the manuscript. K.W.K.L. (in vitro), A.S., L.S., E.A.Y.M., A.M., E.A.S., and R.J. (in vivo) conducted experiments and analyzed data. M.H. contributed to discussion, analyzed data, and edited the manuscript. I.G.F. planned and supervised the study, designed experiments, analyzed data, and wrote the manuscript. I.G.F. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Parts of this study were presented in abstract form at the 72nd Scientific Sessions of the American Diabetes Association, Philadelphia, Pennsylvania, 8–12 June 2012.

The authors thank Albert Liang for technical help.