Accumulation of mesangial matrix is a pivotal event in the pathophysiology of diabetic nephropathy. The molecular triggers for matrix production are still being defined. Here, suppression subtractive hybridization identified 15 genes differentially induced when primary human mesangial cells are exposed to high glucose (30 mm versus 5 mm) in vitro. These genes included (a) known regulators of mesangial cell activation in diabetic nephropathy (fibronectin, caldesmon, thrombospondin, and plasminogen activator inhibitor-1), (b) novel genes, and (c) known genes whose induction by high glucose has not been reported. Prominent among the latter were genes encoding cytoskeleton-associated proteins and connective tissue growth factor (CTGF), a modulator of fibroblast matrix production. In parallel experiments, elevated CTGF mRNA levels were demonstrated in glomeruli of rats with streptozotocin-induced diabetic nephropathy. Mannitol provoked less mesangial cell CTGF expression in vitro than high glucose, excluding hyperosmolality as the key stimulus. The addition of recombinant CTGF to cultured mesangial cells enhanced expression of extracellular matrix proteins. High glucose stimulated expression of transforming growth factor β1 (TGF-β1), and addition of TGF-β1 to mesangial cells triggered CTGF expression. CTGF expression induced by high glucose was partially suppressed by anti-TGF-β1 antibody and by the protein kinase C inhibitor GF 109203X. Together, these data suggest that 1) high glucose stimulates mesangial CTGF expression by TGFβ1-dependent and protein kinase C dependent pathways, and 2) CTGF may be a mediator of TGFβ1-driven matrix production within a diabetic milieu. Accumulation of mesangial matrix is a pivotal event in the pathophysiology of diabetic nephropathy. The molecular triggers for matrix production are still being defined. Here, suppression subtractive hybridization identified 15 genes differentially induced when primary human mesangial cells are exposed to high glucose (30 mm versus 5 mm) in vitro. These genes included (a) known regulators of mesangial cell activation in diabetic nephropathy (fibronectin, caldesmon, thrombospondin, and plasminogen activator inhibitor-1), (b) novel genes, and (c) known genes whose induction by high glucose has not been reported. Prominent among the latter were genes encoding cytoskeleton-associated proteins and connective tissue growth factor (CTGF), a modulator of fibroblast matrix production. In parallel experiments, elevated CTGF mRNA levels were demonstrated in glomeruli of rats with streptozotocin-induced diabetic nephropathy. Mannitol provoked less mesangial cell CTGF expression in vitro than high glucose, excluding hyperosmolality as the key stimulus. The addition of recombinant CTGF to cultured mesangial cells enhanced expression of extracellular matrix proteins. High glucose stimulated expression of transforming growth factor β1 (TGF-β1), and addition of TGF-β1 to mesangial cells triggered CTGF expression. CTGF expression induced by high glucose was partially suppressed by anti-TGF-β1 antibody and by the protein kinase C inhibitor GF 109203X. Together, these data suggest that 1) high glucose stimulates mesangial CTGF expression by TGFβ1-dependent and protein kinase C dependent pathways, and 2) CTGF may be a mediator of TGFβ1-driven matrix production within a diabetic milieu. extracellular matrix actin-related protein-3 adenylyl cyclase-associated protein connective tissue growth factor fetal bovine serum glyceraldehyde-3-phosphate dehydrogenase myosin regulatory light chain protein kinase C plasminogen activator inhibitor suppression subtractive hybridization streptozotocin transforming growth factor-β1 translationally controlled tumor protein reverse transcription-polymerase chain reaction kilobase(s) Diabetic nephropathy accounts for over 30% of end stage renal failure. The pathological hallmark of diabetic nephropathy is glomerulosclerosis due to accumulation of extracellular matrix in the glomerular mesangium (1Kreisberg J.I. Ayo S.H. Kidney Int. 1993; 43: 109-113Abstract Full Text PDF PubMed Scopus (84) Google Scholar). Mesangial matrix accumulation reflects both increased synthesis and decreased degradation of extracellular matrix (ECM),1 is associated with the development of proteinuria, and is followed by subsequent development of hypertension and progressive kidney failure (2Parring H.-H. Osterby R. Anderson P.W. Hsueh W.A. Bienner B.M. Brenner and Rector's The Kidney. 5 th Ed. W. B. Saunders, Philadelphia1996: 1864-1883Google Scholar). Hyperglycemia is a major stimulus for mesangial cell matrix production in diabetic nephropathy. The mechanisms by which hyperglycemia perturb mesangial cell function are still being appreciated and include direct effects of high extracellular glucose levels and indirect effects transduced through alterations in glomerular hemodynamics and through the actions of advanced glycosylation end products. Propagation of mesangial cells under conditions of high ambient glucose has proved a useful in vitro model with which to probe the molecular basis for mesangial matrix accumulation in diabetes, attributable to hyperglycemia. Specifically, exposure of cultured mesangial cells to high glucose stimulates de novo synthesis of ECM components, such as type IV collagen, fibronectin, and laminin, and other products that are accumulated in vivo (3Ayo S.H. Radnik R.A. Garoni J.A. Glass W.F. Kreisberg J.I. Am. J. Pathol. 1990; 136: 1339-1348PubMed Google Scholar, 4Wahab N.A. Harper K. Mason R.M. Biochem. J. 1996; 316: 985-992Crossref PubMed Scopus (80) Google Scholar, 5Ayo S.H. Radnik R.A. Glass W.F. Garoni J.A. Rampt E.R. Appling D.R. Kreisberg J.I. Am. J. Physiol. 1991; 260: F185-F191PubMed Google Scholar). Here, we used suppression subtractive hybridization (SSH) (6Diatchenko L. Lau Y.F. Campbell A.P. Chenchik A. Moqadam F. Huang B. Lukyanov S. Lukyanov K. Gurskaya N. Sverdlov E.D. Siebert P. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 6025-6030Crossref PubMed Scopus (2748) Google Scholar) to identify mesangial cell genes that are differentially induced by high glucose and have focused on genes encoding potential modulators of ECM production in diabetic nephropathy. SSH is a method based on suppressive PCR that allows creation of subtracted cDNA libraries for the identification of genes differentially expressed in response to an experimental stimulus (7Gurskaya N.G. Diatchenko L. Chenchik A. Siebert P.D. Khaspekov G.L. Lukyanov K.A. Vagner L.L. Ermolaeva O.D. Lukyanov S.A. Sverdlov E.D. Anal. Biochem. 1996; 240: 90-97Crossref PubMed Scopus (233) Google Scholar). SSH differs from earlier subtractive methods by including a normalization step that equalizes for relative abundance of cDNAs within a target population. This modification should enhance the probability of identifying increased expression of low abundance transcripts and represents a potential advantage over other methods for identifying differentially regulated genes such as differential display-PCR (8Liang P. Pardee A.B. Science. 1992; 257: 967-971Crossref PubMed Scopus (4727) Google Scholar) and cDNA representation difference analysis (9Hubank M. Schatz D.G. Nucleic Acids Res. 1994; 22: 5640-5648Crossref PubMed Scopus (781) Google Scholar). Primary human mesangial cells (Clonetics) were cultured as reported previously (10Brady H.R. Denton M.D. Jimenez W. Takata S. Palliser D. Brenner B.M. Kidney Int. 1992; 42: 480-487Abstract Full Text PDF PubMed Scopus (64) Google Scholar, 11Denton M.D. Marsden P.A. Luscinskas F.W. Brenner B.M. Brady H.R. Am. J. Physiol. 1991; 261: F1071-F1079PubMed Google Scholar). Cells (passage 7–11) were maintained in medium containing either 5 or 30 mmd-glucose for 7 days. Culture medium was replenished three times during this period to maintain glucose levels in the desired range. To control for the effects of hyperosmolality, mesangial cells were cultured in medium containing 5 mm glucose supplemented with 25 mmmannitol. Male Munich-Wistar rats (260–290 g, Simonsen Laboratories) were rendered diabetic by treatment with streptozotocin (STZ; Sigma), 50 g/kg intravenously as described previously (12Zatz R. Meyer T.W. Rennke H.G. Brenner B.M. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 5963-5967Crossref PubMed Scopus (523) Google Scholar). At months 2 and 4 after induction of diabetes mellitus, rats were anaesthetized by intraperitoneal injection of pentobarbital (50 mg/kg), and the right kidney was excised and weighed immediately. Glomeruli were isolated from renal cortex by standard sieving techniques (10Brady H.R. Denton M.D. Jimenez W. Takata S. Palliser D. Brenner B.M. Kidney Int. 1992; 42: 480-487Abstract Full Text PDF PubMed Scopus (64) Google Scholar, 11Denton M.D. Marsden P.A. Luscinskas F.W. Brenner B.M. Brady H.R. Am. J. Physiol. 1991; 261: F1071-F1079PubMed Google Scholar). Glomerular isolation was completed within 20 min of removing the kidneys. RNA extraction proceeded immediately thereafter. Polyadenlyated RNA was isolated from mesangial cells using the Microfast Track kit (Invitrogen). Total RNA was isolated from glomeruli using RNAzol solution (TEL-test Inc.). SSH was performed with the PCR-SELECT cDNA subtraction kit (CLONTECH) as directed by the manufacturer with the modification that a 4-fold greater than recommended amount of driver cDNA was added to the second hybridization. Starting material consisted of 2 μg of mesangial cell mRNA cultured in 30 mmd-glucose for 7 days as “tester” and 2 μg of mesangial cell mRNA cultured in 5 mmd-glucose for 7 days as “driver.” Thirty primary PCR cycles and 12 secondary PCR cycles were performed. PCR products generated by SSH were subcloned into the PCR 2.1 vector using the original TA cloning kit (Invitrogen). Subcloned cDNAs were isolated by colony PCR amplification. Sequencing was performed using an automated ABI 370A DNA sequencing system. Sequence reactions were carried out with the ABI prism dye terminator cycle sequencing ready reaction kit (Perkin-Elmer). The sequences obtained were compared against GenBankTM/EBI and expressed sequence tag data bases using BLAST searches. Northern blots were performed using formaldehyde denaturation according to standard protocols and quantitated using a phosphorimager (Bio-Rad). For RT-PCR, chromosomal DNA was removed from total RNA using DNase I (Life Technologies, Inc.). To generate a rat-specific CTGF probe, 2 μg of total RNA was transcribed as cDNA using standard procedures. Equal amounts of cDNA were subsequently amplified by PCR using specific primers for GAPDH (GenBankTM/EBI accession numberM17851, sense: ACCACAGTCCATGCCATCAC; antisense: TCCACCACCCTGTTGCTGTA), collagen I (GenBankTM/EBI accession number X55525, sense: GGTCTTCCTGGCTTAAAGGG; antisense: GCTGGTCAGCCCTGTAGAAG), collagen IV (GenBankTM/EBI accession number M11315, sense: CCAGGAGTTCCAGGATTTCA; antisense: TTTTGGTCCCAGAAGGACAC), fibronectin (GenBankTM/EBI accession number X02761, sense: CGAAATCACAGCCAGTAG, antisense: ATCACATCCACACGGTAG), and the mouse cDNA sequence for the CTGF homologue, Fisp–12 (GenBankTM/EBI accession number M70642, sense: CTAAGACCTGTGGAATGGGC; antisense: CTCAAAGATGTCATTGTCCCC). To assess the role of TGF-β1 as a stimulus for CTGF expression in response to high glucose, cells were incubated in either 5 mm glucose, 30 mm glucose plus 1 μg/ml anti-TGF-β1 antibody for 7 days with three changes of medium. In additional experiments cells were serum starved for 24 h in RPMI 1640 and 0.5% FBS and subsequently treated with 10 ng/ml TGF-β1 (Calbiochem) or 10 ng/ml TGF-β1 preadsorbed with 1 μg/ml neutralizing anti-TGF-β1 polyclonal antibody (R & D Systems) and CTGF expression assessed. The role of protein kinase C (PKC) on CTGF expression in response to high glucose was investigated by culturing the mesangial cells in either 5 mm, 30 mm glucose, or 30 mm glucose and the PKC inhibitor GF 109203X (13Toullec D. Pianetti P. Coste H. Bellevergue P. Grand-Perret T. Ajakane M. Baudet V. Boissin P. Boursier E. Loriolle F. Duhamel L. Charon D. Kirilovsky J. J. Biol. Chem. 1991; 266: 15771-15781Abstract Full Text PDF PubMed Google Scholar). The effect of baculovirus expressed recombinant human CTGF (rhCTGF) (14Frazier K. Williams S. Kothapalli D. Klapper H. Grotendorst G.R. J. Invest. Dermatol. 1996; 107: 404-411Abstract Full Text PDF PubMed Scopus (680) Google Scholar) (8 ng/ml for 24 h; generously supplied by Dr. Gary Grotendorst) on the expression of collagens and fibronectin was determined in serum-starved mesangial cells as described above. SSH suggested differential induction of 16 mRNAs in primary cultures of human mesangial cells propagated for 7 days in 30 mm glucose. Differential expression of 15 of the 16 subcloned fragments was confirmed by Northern blot (TableI). Sequence analysis revealed induction of four genes implicated previously in the pathogenesis of diabetic nephropathy: fibronectin, caldesmon, plasminogen activator inhibitor-1 (PAI-1), and thrombospondin (15Hong C.Y. Chia K.S. J. Diabet. Complicat. 1998; 12: 43-60Crossref Scopus (138) Google Scholar, 16Makino H. Kashihara N. Sugiyama H. Kanao K. Sekikawa T. Okamoto K. Maeshima Y. Ota Z. Nagai R. Diabetes. 1996; 45: 488-495Crossref PubMed Google Scholar, 17Holmes D.I. Abdel Wahab N. Mason R.M. Biochem. Biophys. Res. Commun. 1997; 238: 179-184Crossref PubMed Scopus (36) Google Scholar, 18Tada H. Isogai S. Nephron. 1998; 79: 38-43Crossref PubMed Scopus (39) Google Scholar). Of 11 other cDNA fragments, two encoded novel genes, designated IHG-1 andIHG-2 (increased in highglucose), and nine encoded known genes whose induction by high glucose had not been reported previously. Prominent among the latter genes were CTGF and several cytoskeleton-associated proteins, namely profilin, caldesmon, adenylyl cyclase-associated protein (CAP), actin-related protein-3 (ARP3), T-plastin, and myosin regulatory light chain (MRLC). Subsequent studies focused on induction of CTGF, a regulator of matrix production in several model systems (14Frazier K. Williams S. Kothapalli D. Klapper H. Grotendorst G.R. J. Invest. Dermatol. 1996; 107: 404-411Abstract Full Text PDF PubMed Scopus (680) Google Scholar, 19Grotendorst G.R. Cytokine Growth Factor Rev. 1997; 8: 171-179Crossref PubMed Scopus (681) Google Scholar, 20Bradham D.M. Igarashi A. Potter R.L. Grotendorst G.R. J. Cell Biol. 1991; 114: 1285-1294Crossref PubMed Scopus (820) Google Scholar) as discussed below. The prominence of genes encoding multiple actin-binding proteins is also noteworthy, given recent reports implicating F-actin disassembly in the pathogenesis of mesangial cell dysfunction and glomerular hypertension in diabetic nephropathy (21Zhou X. Hurst R.D. Templeton D. Whiteside C.I. Lab. Invest. 1995; 73: 372-383PubMed Google Scholar,22Zhou X. Lai C. Dlugosz J. Kapor-Drezic J. Munk S. Whiteside C. Kidney Int. 1997; 51: 1797-1808Abstract Full Text PDF PubMed Scopus (37) Google Scholar). The induction of profilin expression is particularly interesting given its role as a regulator of actin polymerization under conditions of cell stress (23Sohn R.H. Goldschmidt-Clermont P.J. Bioessays. 1994; 16: 465-472Crossref PubMed Scopus (163) Google Scholar).Table ISummary of cDNAs identified by SSH as being induced in mesangial cells cultured in high glucoseGeneaSequence identity based on comparison with Genbank™ EBI data base.mRNAbEstimate of size (kb) of mRNA by Northern blot analysis.Differential expressioncExpression based on Northern blot analysis of primary human mesangial cells cultured under indicated conditions relative to expression in cells cultured in 5 mm glucose. Values were obtained by phosphorimaging and were normalized by comparison with GAPDH (dash indicates detected in mesangial cells cultured in 5 mm glucose + 25 mm mannitol and 30 mm glucose, but not in 5 mm glucose, -fold expression is degree of expression relative to that found in 5 mM glucose).30 mm glucose25 mmmannitol + 5 mm glucosekb-foldExtracellular matrix proteinsFibronectin7.02.11.5Thrombospondin6.07.08.0Actin-binding proteinsMRLC0.93.91.6T-plastin1.24.21.8Caldesmon3.63.32.8Profilin1.02.22.3CAP2.61.51.7ARP32.52.01.0Growth factorsCTGF2.43.01.5OthersPAI-12.01.21.03.03.92.0RBM31.51.81.4Ubiquitin3.02.31.7TCTP0.84.32.5IHG-1 (GenBank™ accession no. AF110136)dIHG-1 shows identity to EST GenBank™ accession no.AA312756.3.4——IHG-2 (GenBank™ accession no. AF110137)eIHG-2 shows identity to EST GenBank™ accession no.AA599801.2.52.02.0a Sequence identity based on comparison with Genbank™ EBI data base.b Estimate of size (kb) of mRNA by Northern blot analysis.c Expression based on Northern blot analysis of primary human mesangial cells cultured under indicated conditions relative to expression in cells cultured in 5 mm glucose. Values were obtained by phosphorimaging and were normalized by comparison with GAPDH (dash indicates detected in mesangial cells cultured in 5 mm glucose + 25 mm mannitol and 30 mm glucose, but not in 5 mm glucose, -fold expression is degree of expression relative to that found in 5 mM glucose).d IHG-1 shows identity to EST GenBank™ accession no.AA312756.e IHG-2 shows identity to EST GenBank™ accession no.AA599801. Open table in a new tab Within a diabetic milieu, high glucose levels may perturb cellular function through glucose-specific actions or by increasing the osmolality of extracellular fluids. The role of hyperosmolality as a mediator of gene induction by high glucose was assessed by comparing mRNA levels, as determined by Northern blot, in cells cultured in either 30 mm glucose or in 5 mm glucose supplemented with 25 mm mannitol. High glucose was more effective than high osmolality at inducing expression of CTGF, MRLC, ARP3, T-plastin, and translationally controlled tumor protein (TCTP). High glucose and mannitol-induced hyperosmolality afforded equivalent induction of the other products. CTGF is a 38-kDa cysteine-rich secreted peptide known to modulate ECM production in some extrarenal cell types (14Frazier K. Williams S. Kothapalli D. Klapper H. Grotendorst G.R. J. Invest. Dermatol. 1996; 107: 404-411Abstract Full Text PDF PubMed Scopus (680) Google Scholar, 19Grotendorst G.R. Cytokine Growth Factor Rev. 1997; 8: 171-179Crossref PubMed Scopus (681) Google Scholar, 20Bradham D.M. Igarashi A. Potter R.L. Grotendorst G.R. J. Cell Biol. 1991; 114: 1285-1294Crossref PubMed Scopus (820) Google Scholar). Here, SSH identified a cDNA fragment of 250 base pairs, which was identical to bases 814–1061 of the human CTGF cDNA. In separate experiments, induction of CTGF mRNA expression in primary human mesangial cells cultured in high glucose was confirmed by Northern blotting (Fig. 1). CTGF mRNA expression was between 2.5- and 3.3-fold higher in mesangial cells cultured in 30 mm glucose as compared with 5 mmglucose. CTGF is a member of a small family of highly homologous proteins termed the CCN family (for CTGF/fisp-12,cef10/cyr61 and Nov) (24Bork P. FEBS Lett. 1993; 327: 125-130Crossref PubMed Scopus (558) Google Scholar). These peptides are characterized by conservation of 38 cysteine residues that constitute more than 10% of the amino acid content. All members have signal peptides and appear to be secreted via orthodox secretory pathways (20Bradham D.M. Igarashi A. Potter R.L. Grotendorst G.R. J. Cell Biol. 1991; 114: 1285-1294Crossref PubMed Scopus (820) Google Scholar). In the context of diabetic nephropathy, it is intriguing that CTGF stimulates the proliferation of NRK fibroblasts in vitro and the expression of type 1 collagen, fibronectin, and α5 integrin by these cells (19Grotendorst G.R. Cytokine Growth Factor Rev. 1997; 8: 171-179Crossref PubMed Scopus (681) Google Scholar). To investigate the direct effects of CTGF on matrix production mesangial cells were incubated with recombinant protein. rhCTGF up-regulates mesangial cell collagens I and IV and fibronectin (Fig. 1 B). These proteins typify matrix accumulation seen in diabetic nephropathy (1Kreisberg J.I. Ayo S.H. Kidney Int. 1993; 43: 109-113Abstract Full Text PDF PubMed Scopus (84) Google Scholar, 2Parring H.-H. Osterby R. Anderson P.W. Hsueh W.A. Bienner B.M. Brenner and Rector's The Kidney. 5 th Ed. W. B. Saunders, Philadelphia1996: 1864-1883Google Scholar, 25Poncelet A.C. Schnaper H.W. Am. J. Physiol. 1998; 275: F458-F466Crossref PubMed Google Scholar). These data demonstrate the potential of CTGF as a stimulus for increased ECM synthesis (19Grotendorst G.R. Cytokine Growth Factor Rev. 1997; 8: 171-179Crossref PubMed Scopus (681) Google Scholar) and mesangial expansion in diabetic nephropathy. To assess CTGF expression in diabetic nephropathy in vivo, CTGF mRNA levels were measured in RNA isolated from the cortex of rats with STZ-induced diabetes mellitus (Fig. 2). To this end, PCR primers for rat CTGF were designed from the sequence of the mouse CTGF homologue, fisp12 (26Ryseck R.P. Macdonald-Bravo H. Mattei M.G. Bravo R. Cell Growth Differ. 1991; 2: 225-233PubMed Google Scholar). The product was confirmed as rat CTGF by sequence analysis; the 340-base pair product being 94% identical at the nucleotide level (bases 783–1123) and 99% identical at the amino acid level to mouse fisp12 (Fig. 2, A andB). Induction of CTGF mRNA was observed in the renal cortex of rats with STZ-induced diabetic nephropathy at four months after administration of STZ, coincident with mesangial expansion and proteinuria (Fig. 2 C and data not shown). CTGF expression was further localized to glomeruli by Northern analysis of RNA extracted from glomeruli isolated by differential sieving from renal cortex of rats with STZ-induced diabetic nephropathy. Glomerular levels of CTGF mRNA were increased by 2.5- and 1.6-fold after 2 and 4 months of diabetes, respectively, by comparison with age- and sex-matched controls. The significance of these observations are further supported by a recent report demonstrating CTGF expression in a screen of human renal diseases, including diabetic nephropathy (27Ito Y. Aten J. Bende R.J. Oemar B.S. Rabelink T.J. Weening J.J. Goldschmeding R. Kidney Int. 1998; 53: 853-861Abstract Full Text Full Text PDF PubMed Scopus (519) Google Scholar). TGF-β1 is a stimulus for mesangial matrix accumulation in diabetic nephropathy and triggers CTGF release in several other systems (19Grotendorst G.R. Cytokine Growth Factor Rev. 1997; 8: 171-179Crossref PubMed Scopus (681) Google Scholar, 28Kothapalli D. Frazier K. Grotendorst G.R. Cell Growth Differ. 1997; 8: 61-68PubMed Google Scholar, 29Yamamoto T. Nakamura T. Noble N.A. Ruoslahti E. Border W.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 1814-1818Crossref PubMed Scopus (828) Google Scholar). In our experimental model, high glucose concentrations provoked induction of TGF-β1 mRNA expression in cultured human mesangial cells over the same temporal framework as CTGF expression (data not shown). To explore the potential role of TGF-β1 as a regulator of CTGF expression induced by high glucose, mesangial cells were first incubated with 10 ng/ml TGF-β1 for 24 h in the presence of 5 mm glucose. Northern analysis revealed that TGF-β1 was a potent inducer of increased CTGF mRNA levels under these conditions (Fig. 3 A). This effect was inhibited by the addition of a neutralizing anti-TGF-β1 antibody (Fig. 3 A). Interestingly, the PKC inhibitor GF 109203X was without effect on TGF-β1-induced CTGF expression in our system (data not shown). This is consistent with a recent report indicating that TGF-β1-induced CTGF expression is PKC-independent in fibroblasts (30Kothapalli D. Hayashi N. Grotendorst G.R. FASEB J. 1998; 12: 1151-1161Crossref PubMed Scopus (100) Google Scholar). Intriguingly, our data indicate that the TGF-β1 antibody only partially attenuated the glucose-induced increase in CTGF transcript level in mesangial cells grown in 30 mm glucose for 7 days (Fig. 3 B), suggesting that high glucose triggers mesangial cell CTGF expression through TGF-β1-dependent and -independent pathways. Several studies have reported increased expression of TGF-β1 in renal glomeruli in human and experimental models of diabetes (31Sharma K. Ziyadeh F.N. Diabetes. 1995; 44: 1139-1146Crossref PubMed Scopus (558) Google Scholar, 32Park I. Kiyomoto H. Abboud S.L. Abboud H.E. Diabetes. 1997; 46: 473-480Crossref PubMed Google Scholar). Short term administration of TGF-β1 neutralizing antibodies attenuates overexpression of mRNAs encoding matrix components and glomerulosclerosis in the STZ mouse model of diabetes (33Sharma K. Jin Y. Guo J. Ziyadeh F.N. Diabetes. 1996; 45: 522-530Crossref PubMed Scopus (0) Google Scholar). CTGF shares some of the biological actions of TGF-β1 such as stimulation of cell proliferation and extracellular matrix protein synthesis in fibroblasts. When considered in this context, our results suggest that TGF-β1 may promote mesangial matrix production, in part, by inducing CTGF synthesis. TGF-β1 has a complex profile of biological activities that includes pro-inflammatory, pro-fibrotic, and anti-inflammatory effects. By targeting CTGF it may be possible to attenuate the sclerosis-inducing effects of TGFβ, while preserving its more desirable anti-inflammatory activities. Whereas the PKC inhibitor GF 109203X did not influence TGF-β1 mediated up-regulation of CTGF (see above), this compound afforded partial inhibition of high glucose-induced CTGF expression (Fig.3 C). These data are particularly interesting given the evidence that high glucose induces de novo diacylglycerol synthesis and subsequent PKC activation in a variety of cell types, including mesangial cells (34DeRubertis F.R. Craven P. Diabetes. 1994; 43: 1-8Crossref PubMed Google Scholar, 35Fumo P. Kuncio G.S. Ziyadeh F.N. Am. J. Physiol. 1994; 267: F632-F638PubMed Google Scholar). In fact, PKC has been identified as a potential therapeutic target in the treatment of diabetic complications (36Ishii H. Jirousek M.R. Koya D. Takagai C. Xia P. Clermont A. Bursell S.-E. Kern T.S. Ballas L.M. Heath W.F. Stramm L.E. Feener E.P. King G.L. Science. 1995; 272: 728-731Crossref Scopus (1100) Google Scholar, 37Murphy M. McGinty A. Godson C. Curr. Opin. Nephrol. Hypertens. 1998; 7: 563-570Crossref PubMed Scopus (34) Google Scholar). In summary, SSH proved a useful technique for the identification of mesangial cell genes induced by high glucose. The relative speed, efficiency, and convenience of SSH suggests that it offers an attractive method for comparative gene expression that complements more comprehensive but technically demanding techniques such as serial analysis of gene expression (38Velculescu V.E. Zhang L. Volgestein B. Kinzler K.W. Science. 1995; 270: 484-487Crossref PubMed Scopus (3636) Google Scholar). SSH identified 15 mesangial cell genes whose expression is induced by high glucose. Four of these genes encoded for proteins already implicated in the pathophysiology of diabetic nephropathy, nine encoded known proteins whose induction had not been reported previously by high glucose, and two were novel genes. Among the nine known genes was CTGF. This pro-fibrotic factor was expressed in renal cortex and glomeruli of kidneys with diabetic nephropathy and appeared to be induced in vitro through TGF-β1-dependent and PKC-dependent pathways. rhCTGF increased ECM production by mesangial cells. CTGF may therefore be an attractive target for design of novel anti-sclerotic therapies for diabetic glomerulosclerosis. We thank Drs. Gary Grotendorst and Shawn Williams, Department of Cell Biology and Anatomy, University of Miami, School of Medicine for the generous gift of recombinant human CTGF. This study was funded by The Health Research Board, Ireland, Diabetic Nephropathy Research Unit.