J Am Soc Nephrol 14: 28–36, 2003 Ca2ϩ/-Dependent Protein II Stimulates c-fos Transcription and DNA Synthesis by a Src-Based Mechanism in Glomerular Mesangial Cells

YUAN WANG, RANGNATH MISHRA, and MICHAEL S. SIMONSON Department of Medicine, Division of Nephrology, School of Medicine, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, Ohio.

Abstract. Mesangial cell growth factors elevate intracellular free CaMK inhibitor, KN-93, did not block activation of the c-fos 2ϩ 2ϩ [Ca ]i, but mechanisms linking [Ca ]i to gene expression and promoter by ectopically expressed v-Src. Stimulation of c-Src by DNA synthesis are unclear. This study investigated the hypothesis endothelin-1 required CaMK II activity, further supporting the that Ca2ϩ/calmodulin-dependent II (CaMK II), notion that CaMK II acts upstream of Src in a signaling cassette. 2ϩ which is activated by elevated [Ca ]i, increases c-fos transcrip- Activation of the c-fos promoter by CaMKII290 and Src required tion and DNA synthesis via a Src-based mechanism. In cultured the c-fos serum response element. Dominant negative SrcKϪ also rat mesangial cells, dominant negative Src (SrcKϪ) blocked ac- blocked induction of DNA synthesis in mesangial cells by CaMK tivation of the c-fos gene promoter by CaMK II 290, a constitu- II 290. Collectively, these results suggest that in mesangial cells tively active form of CaMK II␣. Activation of the c-fos promoter Src protein tyrosine act downstream of CaMKII in a 2ϩ by CaMK II 290 was also blocked by COOH-terminal Src kinase, signaling pathway in which [Ca ]i induces the c-fos promoter which phosphorylates and inactivates c-Src. A pharmacologic and increases DNA synthesis. [email protected]

2ϩ Mesangial cells and the associated mesangial matrix are essen- The molecular mechanisms linking [Ca ]i to gene expres- tial for normal structure and function of the glomerular capil- sion and DNA synthesis in mesangial cells are unclear. In renal laries (1–5). The mesangial phenotype is highly plastic, and and nonrenal cells, recent evidence suggests a possible role in control of mesangial cell growth is critical for normal devel- Ca2ϩ signaling for non-receptor protein-tyrosine kinases of the opment of the glomerular tuft in metanephrogenesis and in the Src family. Ca2ϩ influx through Ca2ϩ-permeable ion channels response of adult kidneys to glomerular injury (6). In the adult activates c-Src tyrosine kinase activity in epithelial cells kidney, mesangial cells are largely quiescent, with a renewal (12,13), in neurons (14), and in mesangial cells (15). The rate of less than 1% (2–4). Glomerular injury can alter the ability of dominant negative c-Src to block c-fos immediate phenotype of mesangial cells, resulting in hypertrophy, hyper- early gene induction and neurite outgrowth by Ca2ϩ influx in plasia, and/or expansion of the mesangial matrix. This injury- neurons suggests an important role for Src in cell signaling by dependent phenotypic switch of mesangial cells contributes to Ca2ϩ (14). In mesangial cells, Ca2ϩ influx and subsequent glomerulosclerosis and is mediated in part by growth factors activation of Src are necessary for c-fos immediate early gene and inflammatory mediators that evoke immediate early gene induction by the endothelin-1 -coupled receptor expression and DNA synthesis. The path- (15,16), which has been suggested to stimulate mesangial cell ways that control mesangial cell growth are complex but seem growth in glomerular injury (17,18). These observations sug- to share as an early event a rapid and transient increase in ϩ gest c-Src as a candidate in signal transduction pathways that 2 ϩ intracellular free [Ca ]i (7,8). A variety of experiments in 2 link elevation of cytosolic free [Ca ]i to c-fos induction in cultured mesangial cells and in experimental models of glo- mesangial cells, but a functional role for c-Src in stimulating merular injury with mesangial expansion demonstrate that in- DNA synthesis by specific effectors of Ca2ϩ signaling has not hibiting the rise in intracellular free [Ca2ϩ] attenuates mesan- i been demonstrated. gial cell proliferation (2,4,9–11). One of the primary mechanisms by which intracellular 2ϩ [Ca ]i regulates signal transduction is through activation of the Ca2ϩ/calmodulin-dependent protein kinases (CaMK) (19). Received May 30, 2002. Accepted October 2, 2002. In this study, we asked whether a CaMK II signaling pathway Correspondence to Dr. Michael S. Simonson, Department of Medicine, Divi- activates c-fos transcription and mesangial cell growth by a sion of Nephrology, Biomedical Research Building, Rm. 427, Case Western Reserve University, 2065 Adelbert Road, Cleveland, OH 44106. Phone: 216- Src-based mechanism. We report here that dominant negative 368-1251; Fax: 216-368-1249; E-Mail: [email protected] mutants of Src and the Src-inactivating kinase Csk block 1046-6673/1401-0028 activation of the c-fos promoter by CaMKII. Dominant nega- Journal of the American Society of Nephrology tive Src mutants also inhibit mesangial cell DNA synthesis Copyright © 2002 by the American Society of Nephrology stimulated by an active CaMKII290 mutant. We conclude, DOI: 10.1097/01.ASN.0000043180.18456.47 therefore, that CaMKII signals through Src and suggest that J Am Soc Nephrol 14: 28–36, 2003 CaMK-Src Signaling in Mesangial Cells 29 this pathway regulates genes that control cell growth in the After extensive washing, the appropriate peroxidase-labeled second- mesangium. ary antibodies in blocking buffer (1:10,000) were added, and the proteins were detected by chemiluminescense (ECL; Amersham, Ar- lington Heights, IL). Typical exposure times were 30 to 60 s. The Materials and Methods ␤ Cell Culture same blot was reprobed with an antibody that recognizes rat -actin (Sigma, mAb A5316) to confirm equal protein loading. Mesangial cell strains from male Sprague-Dawley rats were iso- lated and characterized as previously reported (20). Cells were main- tained in RPMI 1640 medium supplemented with 17% fetal bovine Inhibition of Endothelin-1-Activated c-Src by a serum, 100 U/ml penicillin, 100 ␮g/ml streptomycin, 5 ␮g/ml each of CaMKII Inhibitor insulin and transferrin, and 5 ng/ml of sodium selenite at 37°Cin5% Serum-starved cells (24 h; 0.5% FBS) in 60-mm plates were

CO2 incubator. HeLa cells, A7r5 vascular smooth muscle cells, and stimulated with 100 nM endothelin-1 (ET-1), and the cells were bovine pulmonary artery endothelial cells were obtained from the prepared for Western blotting as described above. To block CaMK II American Type Culture Collection and were cultured in DMEM activity, cells were preincubated for 1 h with 20 uM of a myristolyated supplemented with 10% fetal bovine serum, 24 mM HEPES (pH 7.4), autocamtide-2-inhibitory peptide (myr-AIP), a nonphosphorylatable 100 U/ml penicillin, and 100 ␮g/ml streptomycin. analog of autocamtide-2 that corresponds to the autophosphorylation site of CaMK II (Calbiochem). This protocol with myr-AIP has Plasmids, Transient Transfections, and Reporter Gene previously been shown to inhibit CaMK II activation in cultured cells Assays (27,28). The membranes were probed with 1:2000 of an affinity- purified polyclonal phosphospecific antibody that recognizes the ac- The following plasmids were as described: p-356wt/fosLUC (16); tive form (P-Tyr 416) of c-Src exactly as described by the manufac- point mutants (pm) of the c-fos promoter, pmxfosLUC (16,21); turer (Cell Signaling Technology, Beverly, MA). To confirm equal pSrcKϪ,Kϩ (22,23); pCMVCsk (24); pCMVCaMKII␣ 290 (21); amounts of total c-Src, the blot was reprobed with mAb clone 327 pMv-Src (25). Cells were transfected using the calcium phosphate (Calbiochem) as described by Lipsich et al. (29). method as previously reported (16,26). Cells in 6-well plates (35 mm; 2 ϫ 105 cells/well) were transfected with 0.5 ␮g of the p-356wt/ fosLUC, 0.1 ␮g pRSV␤Gal internal control, and 2 ␮g of pCMV or Measurements of DNA Synthesis in Transiently pCMVCaMKII290. When indicated, 2 ␮g pSrcKϩ/KϪ or pCMVCsk Transfected Cells was also added. Total DNA was adjusted to 6 ␮g of DNA/well with Mesangial cells in 6-well plates were transfected as described pUC19. Total promoter strength in control transfections was held above, except that the amount of RSV␤Gal was increased to 1 ␮g per equivalent by inclusion of expression plasmids lacking a cDNA insert. well, and the fos luciferase reporter was omitted. Incorporation of After incubating with precipitates for 16 to 18 h at 37°C, cells were 5-bromo-2-deoxyuridine (BrdU; Sigma, St. Louis, MO) into DNA in then washed three times with DMEM and incubated with DMEM/ transiently transfected cells (i.e., ␤ galactosidase–positive cells) was 0.5% FBS for an additional 24 h before cell lysis (RLB Buffer, measured exactly as described previously (30). Briefly, transfection Promega). Luciferase activity was assessed as described previously cells were labeled for 24 h with 20 ␮M BrdU; Sigma, St. Louis, MO). (26), and relative light units were measured in a Berthold Lumat Monolayers were then washed twice with ice-cold DPBS and fixed for Luminometer (Wallac, Gaithersburg, MD) for two 10-s intervals. 5 min on ice with 2.0% formaldehyde/0.2% glutaraldehyde in DPBS Transfection efficiency was determined using a pRSV␤Gal construct (Dulbecco phosphate-buffered saline). Cells expressing ␤ galactosi- that directed ␤ galactosidase expression from the viral LTR. ␤ Ga- dase were detected by histochemical analysis (30), and under these lactosidase activity was measured using the Galacto-Light protocol as conditions 1 to 3% of mesangial cells were transfected (i.e., ␤ galac- described by the manufacturer (Tropix, Bedford, MA). All reporter tosidase–positive). BrdU incorporation into DNA was then identified gene assays were in the linear range. In experiments where KN-93 by immunocytochemistry using 6 ␮g/ml IgG of a specific monoclonal (Calbiochem, San Diego, CA) was used to inhibit endogenous CaMK antibody (Boehringer Mannheim, Indianapolis, IN) as described (30). activity, KN-93 at 10 ␮M was added 8 h after DNA addition and was Cells were visualized under brightfield microscopy (Nikon Diophot) present throughout the remaining transfection period. with a neutral filter. Red cells (i.e., expressing ␤ galactosidase– positive) with dark nuclei (BrdU-positive) and light nuclei (BrdU- Western Blotting of Epitope-Tagged CaMKII290 negative) were counted, and percent inhibition of DNA synthesis was To determine if SrcKϪ repressed expression of CaMKII290, mes- calculated as follows: angial cells were transiently transfected with pCMV CaMK II 290 ͑N Ϫ N ͒/N (containing an HA epitope tag (21)) and with increasing amounts of v e v the plasmid expressing dominant negative SrcKϪ. Forty-eight hours ϭ where Nv % BrdU-positive nuclei in cells transfected with vector after transfection, cells in 100-mm dishes were washed once in DPBS ϭ only and Ne % BrdU-positive nuclei in cells transfected with and scraped into 1 ml of sample lysis buffer (0.5 M Tris-HCl [pH 6.8], expression plasmid. Statistical significance was analyzed by the ␹2 2.5 ml; 10% (wt/vol) SDS, 4.0 ml; glycerol, 2 ml; 0.1% bromophenol test using InStat for Macintosh (GraphPad, SanDiego, CA). blue, 0.5 ml; ␤-mercaptoethanol, 0.5 ml; distilled water, to 10 ml). The lysate was vortexed, boiled for 5 min, and aliquots were resolved Results on 8 to 16% SDS-PAGE gradient gels and proteins transferred to A Dominant Negative Src Mutant (SrcKϪ) Blocks 0.2-␮m nitrocellulose filters. Transferred proteins were stained with Ponseau S, and the filter was blocked in blocking buffer (1.0% BSA, Activation of the c-fos Promoter by CaMK II 10 mM Tris [pH 7.5], 100 mM NaCl, 0.1%Tween 20) with shaking Our first approach was to ask if dominant negative mutants overnight at 4°C. To detect the HA epitope-tagged CaMKII290 in of Src would interfere with activation of the c-fos promoter by transfected cells, the blots were incubated with a monoclonal anti-HA CaMK II. Kinase-inactive forms of Src (SrcKϪ) function as antibody (clone 12CA5; Boehringer Mannheim, Indianapolis, IN). dominant negative mutants and inhibit DNA synthesis stimulated 30 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 28–36, 2003 by colony stimulating factor or platelet-derived growth factor in suggest that CaMK II acts upstream of Src in a signaling cascade fibroblasts (23,31) and block stimulation of the c-fos promoter by that activates the c-fos promoter. the G protein-coupled ET-1 receptor in mesangial cells (16). We One potential interpretation of the inhibitory effects of therefore tested whether dominant negative SrcKϪ (Lys 295 to SrcKϪ is that SrcKϪ might block expression of CaMKII290. Met) would block stimulation of the c-fos promoter by a consti- Therefore, in mesangial cells transfected with CaMKII290 and tutively active mutant of CaMK II, CaMK II 290 (Figure 1A). SrcKϪ, we measured CaMKII290 levels by Western blotting

Mesangial cells were transiently transfected with a plasmid con- using the HA epitope tag fused to the NH2 terminus (Figure taining a genomic DNA fragment of the c-fos promoter (Ϫ356 to 1A). As shown in Figure 2, SrcKϪ did not inhibit CaMKII290 ϩ109) that drives transcription of a luciferase reporter gene (Fig- expression at any concentration used in these experiments (lane ure 1A) (16). Transfection with a plasmid vector expressing 2 versus lanes 3 and 4 with SrcKϪ, Figure 2). Coexpression of CaMK II 290 stimulated a 3.9-fold increase in c-fos promoter SrcKϩ (lanes 5 and 6, Figure 2) did not significantly alter activity (Figure 1B). Cotransfection with a vector expressing CaMKII 290 levels, which suggests that the Src expression SrcKϪ completely blocked activation of the c-fos promoter by vector was without any apparent effect on CaMKII 290 ex- CaMK II 290 (Figure 1B). Inhibition by SrcKϪ was apparently pression under control of the CMV promoter/enhancer. We specific as transfection with wild-type SrcKϩ did not block conclude that SrcKϪ does not inhibit expression of CaMK II 290-stimulated c-fos promoter activity (Figure 1B). CaMKII290 and that CaMKII does indeed act upstream of Src SrcKϩ did not potentiate the effect of CaMK II 290. These results in the pathway to c-fos.

Figure 1. SrcKϪ dominant negative mutant blocks Ca2ϩ/calmodulin-dependent protein kinase II (CaMK II) signaling to c-fos. (A) Schematic diagram of plasmids used in transient transfections. p-356wt/fos LUC is a reporter construct containing Ϫ356 to ϩ109 of the mouse c-fos promoter driving transcription of a luciferase reporter gene. SIE, sis-inducible element; SRE, serum response element; Ca/CRE, Ca2ϩ-cAMP response element. pCMV CaMK II 290 and pSrcKϪ are expression plasmids encoding a truncated, constitutively activated form of CaMK II and a dominant negative Src mutant, respectively. (B) Glomerular mesangial cells were co-transfected with p-356wt/fos LUC and either pCMV or pCMV CaMK II 290 as described below. As indicated in the legend, plasmids encoding SrcKϪ or wild-type Src Kϩ were also added. Forty-eight hours after DNA addition, the cells were lysed and processed for luciferase and ␤ galactosidase expression. Data are mean Ϯ SEM from four independent experiments in duplicate. ** P Ͻ 0.01 by t test. J Am Soc Nephrol 14: 28–36, 2003 CaMK-Src Signaling in Mesangial Cells 31

Activation of Src by ET-1 Requires CaMK II To determine whether CaMK II participates in Src activation in mesangial cells by physiologic ligands, we treated serum- starved cells with 100 nM ET-1and measured Tyr-416 phos- pho-Src in the presence and absence of myr-AIP to block CaMK II activity. ET-1 increased phospho-Src levels at 10 2ϩ Figure 2. CaMK II 290 expression is unaffected by SrcKϪ. Mesan- min, and Src remained activated at 60 min (Figure 4A). Ca gial cells were transfected with pCMV (lane 1) or pCMV CaMK II ionophore A23187, which increases CaMK II activity, also 290 (lanes 2 to 6) as in Figure 1 but with co-transfection of SrcKϪ (2 increased phospho-Src levels but with a more rapid time ␮g, lane 3; 4 ␮g, lane 4) or SrcKϩ (2 ␮g, lane 5; 4 ␮g, lane 6). Levels course. Preincubation of cells with myr-AIP before addition of of HA epitope-tagged CaMKII290 (p33kDa) were then assessed by ET-1 greatly attenuated Src activation (Figure 4B); reprobing Western blotting. Results are representative of three independent of the blot for total Src protein confirmed the presence of experiments. unactivated Src. These results suggest that CaMK II is up- stream of Src and participates in a pathway of Src activation by ET-1. COOH-Terminal Src Kinase Also Inhibits Activation of the c-fos Promoter by CaMK II An independent approach to inactivate c-Src is to express CaMK II Acts Upstream of Src in Diverse Cell Types Csk, which phosphorylates the COOH-terminal tyrosine We next asked if the ability of CaMK II to function upstream (Y527) in c-Src and maintains the kinase in an inactive con- of Src was cell type–specific or was instead a common feature formation (24,32). When a vector expressing Csk (Figure 3A) of diverse cell types. We conducted the same cotransfection was transiently co-transfected into cells with the c-fos lucif- experiments described above in Figure 1 with CaMK II 290 erase reporter, Csk inhibited activation of the c-fos promoter by and SrcKϪ in A7r5 vascular smooth muscle cells, porcine CaMK II 290 (Figure 3B). As shown previously (16), Csk does aortic endothelial cells, and HeLa cells. SrcKϪ, but not not nonspecifically inhibit the c-fos promoter because Csk does SrcKϩ, blocked activation of the c-fos promoter by CaMK II not block activation of the c-fos promoter by fetal bovine 290 in A7r5 vascular smooth muscle and endothelial cells serum or a constitutively active mutant of Raf-1 kinase (16). (Figure 5). SrcKϪ also blocked activation of the c-fos pro- Also, Csk had no effect on c-fos luciferase activity in the moter by CaMK II 290 in HeLa cells, but SrcKϩ potentiated absence of CaMK II 290, (i.e., pCMV; Figure 3B). Collec- activation by CaMK II 290 (Figure 5). It is not clear why tively, the results with SrcKϪ and Csk strongly suggest that SrcKϩ potentiated stimulation by CaMK II 290 in HeLa cells CaMK II functions upstream of Src in a signaling pathway to and not in the other cell types we studied, but this probably the c-fos promoter. reflects cell type-specific differences in cross-talk between

Figure 3. Src-inactivating kinase Csk blocks induction of c-fos promoter by CaMK II 290. Mesangial cells were transfected with p-356wt/fos LUC and pCMV or pCMV CaMK II 290 as in Figure 1 but with co-transfection of a plasmid encoding full-length Csk (pCMVCsk) or the control vector lacking a cDNA insert (A). (B) Results (mean Ϯ SEM) of three independent experiments in duplicate. ** P Ͻ 0.01. 32 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 28–36, 2003

nist of CaMK, KN-93 (33), did not block stimulation of the c-fos promoter by constitutively active v-Src (Figure 6). Eight hours after co-transfecting mesangial cells with v-Src and the fos-luciferase reporter, KN-93 at 10 ␮M was added and main- tained throughout the remainder of the transfection. KN-93 did not significantly reduce activation of the c-fos promoter by v-Src and did not affect basal promoter activity in cells trans- fected with the expression vector alone (i.e., pM-MuLV, Fig- ure 6). We have previously shown that 10 ␮M KN-93 blocks stimulation of the c-fos promoter by ET-1 (21), a finding that was confirmed here in concurrent experiments to demonstrate that KN-93 was indeed effective (Figure 6). The inability of a CaMK antagonist to block c-fos promoter activation by Src is further evidence that Src functions downstream of CaMKII on a signaling pathway to c-fos.

CaMKII/Src Signaling to c-fos Requires the c-fos Figure 4. Blockade of CaMK II inhibits Src activation by endothelin-1 Serum Response Element (SRE) (ET-1). (A) Serum-starved cells were stimulated with 100 nM ET-1 or We next sought to determine which cis-element(s) of the 2ϩ 10 ␮MCa ionophore A23187 and levels of phospho-Tyr-416 Src c-fos promoter is targeted by CaMKII/Src signaling. Mesangial were measured by Western blotting with an activation state-specific cells were transfected with c-fos reporter constructs in which antibody. The blot was reprobed with mAb against total Src to point mutations have been introduced to inactivate specific confirm equal protein loading (bottom panel). (B) Cells were prein- cis-elements as shown in Figure 7A. Inactivation of the SIE cubated with myr-AIP (20 ␮M for 1 h), a cell-permeable peptide that blocks CaMKII activation, before addition of ET-1 and measurement and FAP cis-elements had no significant effect on activation of of phospho-Tyr-416 Src levels as in panel A. Identical results were the c-fos promoter by either constitutively active CaMKII290 observed in two independent experiments. or by v-Src (pm6 and pm9, Figure 7A). Point mutations in the Ca/CRE, which responds to Ca2ϩ and cAMP-based signaling pathways (34), slightly inhibited the response to CaMKII290 (i.e., 6.4-fold versus 4.3-fold with pm3) but did not alter the response to v-Src. In contrast, point mutations in the SRE completely inhibited stimulation of the c-fos promoter by either

Figure 5. Dominant negative SrcKϪ blocks activation of the c-fos promoter by CaMK II 290 in different cell types. Embryonic rat A7r5 vascular smooth muscle cells, porcine aortic endothelial cells, and human HeLa cells were transfected with p-356wt/fos LUC, pCMV CaMK II 290, and SrcKϪ or SrcKϩ exactly as described in Figure 1. Data are mean Ϯ SEM from three experiments in duplicate. Inhibition by SrcKϪ was significant (P Ͻ 0.01) in all cell types.

CaMK II or Src. However, the key finding is that dominant negative SrcKϪ blocked activation of the c-fos promoter by Figure 6. A CaMK antagonist (KN-93) fails to block c-fos promoter CaMK II 290 in four distinct cell types and suggests that a activation by v-Src. Mesangial cells were co-transfected with p-356wt/fos LUC and p-v-Src (2 ␮g) or the parent expression vector CaMK II-Src signaling cassette is conserved. pM-MuLV. Eight hours later, KN-93 (10 ␮M) was added and main- tained throughout the transfection. In experiments to demonstrate that A Pharmacologic Antagonist of CaMK Fails to Block KN-93 was effective, mesangial cells transfected with p-356wt/fos Activation of the c-fos Promoter by v-Src LUC were treated with KN-93 followed by 100 nM endothelin-1 Additional evidence placing CaMKII upstream of Src came (ET-1). Results (mean Ϯ SEM) of 4 independent experiments in from experiments in which a selective pharmacologic antago- duplicate. ** P Ͻ 0.01. J Am Soc Nephrol 14: 28–36, 2003 CaMK-Src Signaling in Mesangial Cells 33

Figure 7. The c-fos SRE is necessary for activation of the c-fos promoter by CaMKII/Src signaling. (A) Mesangial cells were transfected with c-fos promoter LUC constructs with inactivating point mutations (pm) as indicated schematically at left. Cells were cotransfected with pCMV (open bars), pCMVCaMKII290 (solid bars), or pSrcKϪ (hatched bars). Data are means from three independent experiments; error bars are omitted for clarity but were never more than 20% of the mean. (B) Cells were transfected with a minimal promoter construct (p56LUC, 2 ␮g) or the same vector with a single copy of the c-fos SRE upstream of TATA (pSRELUC, 2 ␮g). Cotransfections were with pCMV alone (no addition), CaMKII290, or CaMKII290 plus dominant negative SrcKϪ. Data are mean Ϯ SEM from four independent experiments.

CaMKII290 or v-Src (pm12, Figure 7A). Thus the SRE ap- activation of the c-fos promoter by CaMKII290 requires both pears to be the critical cis-element required for the CaMKII/Src the SRE and Ca/CRE cis-elements. signaling pathway. It is important to note, however, that full To further investigate the role of the SRE in CaMKII/Src 34 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 28–36, 2003 signaling, we transfected mesangial cells with a luciferase the concept that Src acts downstream of CaMK II in a signaling reporter construct in which transcription is controlled by a cassette that controls gene expression and cell growth. single copy of the SRE upstream of TATA, pSRELUC (26). Cotransfection of CaMKII290 or CaMKII290 plus SrcKϪ Discussion failed to alter luciferase expression from the parent vector, Mesangial hypertrophy and hyperplasia are adaptive re- p56LUC (Figure 7B). However, in cells transfected with sponses to glomerular injury and are mediated by a variety of 2ϩ pSRELUC, CaMKII290 stimulated a 4.7-fold increase in SRE paracrine and autocrine growth factors that increase [Ca ]i. cis-element activity that was markedly inhibited by SrcK- (i.e., Mesangial cell growth requires coupling of [Ca2ϩ]i to tran- 1.6-fold, Figure 7B). Taken together, these results suggest that scription factors that activate and maintain hypertrophy and CaMKII/Src signaling targets the SRE of the c-fos promoter. hyperplasia, but elucidation of the intracellular signals in- volved represents an important challenge. The results pre- Dominant Negative Src KϪ Inhibits DNA Synthesis sented here demonstrate a novel mechanisms whereby the Stimulated by CamK II 290 Ca2ϩ-dependent effector CaMKII functions upstream of Src in To assess the biologic importance of Src in signaling by a signaling pathway that activates the c-fos promoter and CaMK II, we asked whether Src is a downstream effector in stimulates DNA synthesis in cultured mesangial cells. pathways whereby CaMK II controls mesangial cell growth. Several lines of evidence suggest that CaMK II lies upstream To this end, we measured bromodeoxyuridine uptake into of Src in mesangial cells. We blocked CaMK II-stimulated DNA in cells transiently transfected with CaMK II 290. Trans- c-fos induction and DNA synthesis by expressing dominant fected cells were identified by cotransfection with a lac Z gene negative SrcKϪ and Csk to inactivate Src. SrcKϪ and Csk and histochemical detection of expressed ␤ galactosidase (Fig- interfere with Src activity by distinct molecular mechanisms, ure 8, A and B). BrdU was incorporated into DNA in 71% of thereby providing an important internal control for possible cells transfected with CaMK II 290 (Figure 8, A and C) versus nonspecific effects of either protein. Transfection with the 22% of cells with the control vector (Figure 8C). Expression of empty expression vector likewise did not reduce luciferase Src KϪ attenuated the increase in DNA synthesis in cells activity. That inactivation by SrcKϪ was specific was further expressing CaMK II 290 (Figure 8, B and C; 35% versus 71%, supported by the finding that overexpression of wild-type P Ͻ 0.01 by ␹2 analysis). In contrast, SrcKϩ had no effect on SrcKϩ did not alter activation of the c-fos promoter by CaMK CaMK II 290-stimulated DNA synthesis. These results support II 290. SrcKϪ did not block expression of CaMKII290 as verified by Western blotting of HA epitope-tagged CaMKII290 in mesangial cells transfected with SrcKϪ. Moreover, we have previously shown that SrcKϪ and Csk do not inhibit activation of the c-fos promoter by Raf-1, which functions downstream of Src and Ras (16). A pharmacologic CaMK inhibitor (KN-93) did not significantly block activation of the c-fos promoter by v-Src, providing additional evidence that Src lies downstream of CaMKII in this signaling pathway. We also demonstrated that CaMKII acts upstream of Src in several different cell types. We had previously shown that Src activation by ET-1 requires Ca2ϩ influx and elevation of cytosolic free [Ca2ϩ] (15), and the present studies in mesangial cells demonstrated that Src activation by ET-1 was inhibited in the presence of myr-AIP to block CaMKII. Taken together, these results sug- gest that CaMK II functions upstream of Src in a Ca2ϩ- dependent signaling pathway. An important caveat is that SrcKϪ and Csk might also inactivate closely related Src PTK family members (23,24,31,32), so we cannot formally rule out participation of Src-related kinases in CaMK II signaling. In addition, our experiments have not addressed the question of how Ca2ϩ and Figure 8. SrcKϪ inhibits DNA synthesis stimulated by CaMK II 290. CaMK II might activate Src. We were unable to demonstrate a Mesangial cells were transfected with pRSV lacZ, pCMV CaMK II direct protein-protein interaction between CaMK II and cSrc Ϫ ϩ 290, and either pSrcK or pSrcK . After 24 h, DNA was labeled by (Wang and Simonson, unpublished results). Src contains a incorporation of BrdU for 24 h at 37°C. Transfected cells were then consensus site for CaMKs (R-X-X-S/T) in the identified by histochemical detection of expressed ␤ galactosidase, and DNA synthesis was assessed by immunocytochemical detection unique NH2 terminal sequence, raising the possibility of phos- of BrdU. Transfected cells with BrdU-positive nuclei (A) and BrdU- phorylation and control of Src activity by CaMK II. An indirect negative nuclei (B) were counted in three independent experiments, n mechanism by which CaMK II activates Src might depend on Ͼ 300 cells for each condition taken from three independent experi- some as yet unidentified or protein as part of an ments (C). Statistical significance was calculated by the ␹2 test. heterooligomeric protein complex. Indeed, Zhao et al. (12) J Am Soc Nephrol 14: 28–36, 2003 CaMK-Src Signaling in Mesangial Cells 35 propose that Ca2ϩ ionophore A23187 increases Src activity in 2. Mene P, Simonson MS, Dunn MJ: Physiology of the mesangial keratinocytes by an indirect mechanism possibly involving a cell. Physiol Rev 69: 1347–1424, 1989 Ca2ϩ-stimulated tyrosine phosphatase activity or the Ca2ϩ- 3. Johnson RJ, Floege J, Yoshimura A, Iida H, Couser WG, Alpers dependent association of three unidentified proteins with c-Src. CE: The activated mesangial cell: A glomerular “myofibro- Another possibility is that CaMKII activates Ca2ϩ channels blast”? J Am Soc Nephrol 2[Suppl 2]: S190–S197, 1992 2ϩ 4. Haas CS, Schocklmann HO, Lang S, Kralewski M, Sterzel RB: and/or ion pumps (19), thereby elevating [Ca ]i and stimu- lating Src PTK activity. Regulatory mechanism in glomerular mesangial cell prolifera- Our results with point mutants of the c-fos promoter suggest tion. J Nephrol 12: 405–415, 1999 5. Fogo AB: Mesangial matrix modulation and glomerulosclerosis. that the SRE is required for CaMKII/Src signaling to the c-fos Exp Nephrol 7: 147–159, 1999 immediate early gene. Point mutations in the SRE blocked 6. Mene P: Mesangial cell cultures. J Nephrol 14: 198–203, 2000 activation of the c-fos promoter by CaMKII290 and Src. Mu- 7. Bonventre J: Calcium and calcium-related signaling pathways tation of the Ca/CRE only modestly blocked the response to in glomerular mesangial cells. Clin Exp Pharmacol Physiol CaMKII 290 and did not alter activation of the c-fos promoter 23: 65–70, 1996 by Src. The c-fos SRE is activated by the Ras/MAPK pathway, 8. Ling B, Matsunaga H, Ma H, Eaton D: Role of growth factors in and it seems possible that CaMKII/Src signaling also activates mesangial cell ion channel regulation. Kidney Int 48: 1158–1166, 1995 the Ras/MAPK pathway to c-fos. Indeed, Rosen et al. (35) 9. Sabbatini M, Vitaioli L, Baldoni E, Amenta F: Nephroprotective ϩ have previously shown in PC12 cells that Ca2 influx activates effect of treatment with calcium channel blockers in spontaneously the Ras/MAPK pathway, and Rusanescu et al. (14) have fur- hypertensive rats. J Pharmacol Exp Ther 294: 948–954, 2000 ϩ ther demonstrated that Src is required for Ca2 activation of 10. Sugiura T, Imai E, Moriyama T, Horio M, Hori M: Calcium Ras/MAPK. It is not entirely clear how Ca2ϩ influx and Src channel blockers inhibit proliferation and matrix production in activate Ras, but the mechanisms probably involve Src-medi- rat mesangial cells: Possible mechanism of suppression of AP-1 ated phosphorylation of the Shc adapter protein and subsequent and CREB activities. Nephron 85: 71–80, 2000 recruitment of Grb2-Sos complexes (14). The new finding of 11. Whiteside C, Munk S, Zhou X, Miralem T, Templeton DM: the present experiments concerns the role of CaMKII contrib- Chelation of intracellular calcium prevents mesangial cell pro- uting to Src-based signaling to the c-fos SRE. liferative responses. J Am Soc Nephrol 9: 14–25, 1998 An important aspect of our study was to test a possible role 12. Zhao Y, Sudol M, Hanafusa H, Krueger J: Increased tyrosine for the CaMKII-Src signaling cassette in mesangial cell kinase activity of c-Src during calcium-induced keratinocyte differentiation. Proc Natl Acad Sci 89: 8298–8302, 1992 growth, which is a pathologic hallmark of many forms of 13. Zhao Y, Uyttendaele, H., Krueger J, Sudol M, Hanafusa H: glomerular disease. Previous experiments in cultured mesan- Inactivation of c-Yes tyrosine kinase by elevation of intracellular gial cells and in experimental models of glomerular injury with Ca2ϩ levels. Molec Cell Biol 13: 7507–7514, 1993 mesangial growth demonstrate that inhibiting a rise in intra- 14. Rusanescu G, Qi H, Thomas SM, Brugge JS, Halegoua S: 2ϩ cellular free [Ca ]i attenuates mesangial cell proliferation Calcium influx induces neurite growth through a Src-Ras signal- (2,4,9–11), but the molecular mechanisms and effectors of ing cassette. Neuron 15: 1415–1425, 1995 2ϩ Ca signaling that underlie these results remain unclear. We 15. Simonson MS, Wang Y, Herman WH: Ca2ϩ channels mediate showed that ectopic expression of CaMKII 290 stimulated protein tyrosine kinase activation by endothelin-1. Am J Physiol DNA synthesis in quiescent mesangial cells that was effec- 270: F790–F797, 1996 tively blocked by SrcKϪ. The control vector lacking SrcKϪ 16. Simonson MS, Wang Y, Herman WH: Nuclear signaling by and the vector expressing SrcKϩ had no effect on DNA endothelin-1 requires Src protein tyrosine kinases. J Biol Chem synthesis driven by CaMKII 290. These results are the first 271: 77–82, 1996 direct demonstration in mesangial cells that activated CaMKII 17. Benigi A: Endothelin antagonists in renal disease. Kidney Int 57: drives DNA synthesis, and the results also support a role for 1778–1794, 2000 Src in this signal transduction pathway. Many mesangial cell 18. Hunley TE, Kon V: Update on endothelins- biology and clinical growth factors stimulate a rapid and transient increase in in- implications. Pediatr Nephrol 16: 752–762, 2001 2ϩ 19. Hanson PI, Schulman H: Neuronal Ca2ϩ/calmodulin-dependent tracellular free [Ca ]I (7,8), and our experiments suggest that a CaMKII-Src signaling cassette is one mechanism coupling protein kinases. Ann Rev Biochem 61: 559–601, 1992 the rise in [Ca2ϩ]i to DNA synthesis. 20. Simonson MS, Dunn MJ: Eicosanoid biochemistry in cultured glomerular mesangial cells. Meth Enzymol 187: 544–553, 1990 Acknowledgments 21. Wang Y, Simonson MS. Voltage-insensitive Ca2ϩ channels and We thank Sara Courtneidge, Hidesaburo Hanafusa, G. Stanley Ca2ϩ/calmodulin-dependent protein kinases propagate signals McKnight, and Michael G. Gilman for providing plasmids. We also from endothelin-1 receptors to the c-fos promoter. Molec Cell thank William Herman for excellent technical assistance, and George Biol 16: 5915–5923, 1996 Dubyak, Hsing-Jien Kung, and Mark Kester for helpful comments. 22. Twamley-Stein GM, Pepperkok R, Ansorge W, Courtneidge SA: Supported by grant DK-46939 from the National Institutes of Health. The Src family tyrosine kinases are required for platelet-derived growth factor-mediated signal transduction in NIH 3T3 cells. References Proc Natl Acad Sci USA 90: 7696–7700, 1993 1. Latta H: An approach to the structure and function of the glo- 23. Roche S, Koegl M, Barone MV, Roussel MF, Courtneidge SA: merular mesangium. J Am Soc Nephrol 2: S65-S73, 1992 DNA synthesis induced by some but not all growth factors 36 Journal of the American Society of Nephrology J Am Soc Nephrol 14: 28–36, 2003

requires Src family protein tyrosine kinases. Molec Cell Biol 15: 30. Simonson MS, LePage DF, Walsh K: Rapid characterization of 1102–1109, 1995 growth-arrest genes in transient transfection assays. Biotech- 24. Sabe H, Knudsen B, Okada M, Nada S, Nakagawa H, Hanafusa niques 18: 434–442, 1995 H: Molecular cloning and expression of chicken C-terminal Src 31. Vittoria Barone M, Courtneidge SA: Myc but not Fos rescue of kinase: Lack of stable association with c-Src protein. Proc Natl PDGF signaling block caused by kinase-inactive Src. Nature Acad Sci USA 89: 2190–2194, 1992 378: 509–512, 1995 25. Johnson PJ, Coussens PM, Danko AV, Shalloway, D: Overexpressed 32. Nada S, Okada M, MacAuley A, Cooper J, Nakagawa, H: Clon- pp60c-src can induce focus formation without complete transformation ing of a complementary DNA for a protein-tyrosine kinase that of NIH 3T3 cells. Molec Cell Biol 5: 1073–1083, 1985 specifically phosphorylates a negative regulatory site of pp60c- 26. Herman WH, Simonson MS: Nuclear signaling by endothelin-1: src. Nature 351: 69–72, 1991 A Ras pathway for activation of the c-fos serum response ele- 33. Mamiya N, Goldenring J R, Tsunoda Y, Modlin IM, Yasui K, ment. J Biol Chem 270: 11654–11661, 1995 Usuda N, Ishikawa T, Natsume A, Hidaka H: Inhbition of acid 27. Gailly P: Ca2ϩ entry in CHO cells, after Ca2ϩ stores depletion, secretion in gastric parietal cells by the Ca2ϩ/calmodulin-depen- is mediated by arachidonic acid. Cell Calcium 24: 293–304, 1998 dent protein kinase II inhibitor KN-93. Biochim Biophys Res 28. Ishida A, Kameshita I, Okuno S, Kitani T, Fugisawa H: A novel, Comm 195: 608–615, 1993 highly specific a potent inhibitor of calmodulin-dependent pro- 34. Sheng M, McFadden G, Greenberg ME: Membrane depolariza- tein kinase II. Biochem Biophys Res Commun 212: 806–812, tion and clacium induce c-fos transcription via phosphorylation 1995 of transcription factor CREB. Neuron 4: 571–582, 1990 29. Lipsich LA, Lewis AJ, Brugge JS: Isolation of monoclonal 35. Rosen LB, Ginty DD, Weber MJ, Greenberg ME: Membrane antibodies that recognize the transforming proteins of avian depolarization and calcium influx stimulate MEK and MAP sarcoma virus. J Virol 48: 352–360, 1983 kinase via activation of Ras. Neuron 12: 1207–1221, 1994