Expression of Connective Tissue Growth Factor Is Increased in Injured Myocardium Associated With Protein Kinase C β2 Activation and Diabetes

Construction of the adult cardiac myocyte-specific α myosin heavy chain promoter PKCβ2 transgene as well as generation of transgenic mice overexpressing PKCβ2 in the myocardium and their characterization are as previously described (13). In this study, heterozygous transgenic line 4 mice and nontransgenic littermate controls were used at 4, 8, 12, and 16 weeks of age. A separate group of mice were treated with the PKCβ-specific inhibitor LY333531 (0.18% in diet) (Eli Lilly, Indianapolis, IN), which was administered at weaning (3 weeks) and maintained until death at 8 weeks of age.

Diabetes was induced in 8-week-old male FVB mice (Taconic, Germantown, NY) by injection with streptozotocin (STZ; 100 mg/kg i.p.) (Sigma, St Louis, MO) on 2 consecutive days after overnight fast. A separate group of STZ-treated mice were also treated with LY333531 (0.18% in diet) for 8 weeks. Diabetes was induced in 6-week-old male Sprague-Dawley rats (Taconic) by a single STZ injection (60 mg/kg i.p.) after overnight fast. All control animals received citrate vehicle (0.01 mol/l, pH 4.5). Two groups of STZ-treated rats were established to receive either insulin pellet implant (Linplant; Linshin, Scarborough, Ontario, Canada) or LY333531 (0.062% in diet) for 4 weeks. Successful induction of diabetes was confirmed by elevated blood glucose levels (>250 mg/dl) (Glucometer Elite XL; Bayer, Elkhart, IN). Mice were killed 8 weeks after injection, and rats were killed at 2, 4, and 8 weeks. All procedures described are in accordance with the Joslin Diabetes Center’s Animal Care and Use Committee guidelines.

Total RNA was isolated by the phenol/guanidine thiocynate/chloroform method (13). Northern blot analysis was performed using total RNA (15–25 μg) following standard techniques (26). Hybridization was performed with the following cDNA probes: human CTGF cDNA EcoRI fragment (0.8 kb) (26); mouse α1(VI) collagen cDNA EcoRI fragment (2 kb) (13); and mouse fibronectin cDNA EcoRI and XhoI fragment (mouse fibronectin clone ID 524915; American Type Culture Collection, Manassas, VA). Differences in RNA loading were normalized using a PstI fragment of human acidic ribosomal phosphoprotein cDNA (36B4) as the internal control gene, and data were expressed as fold change from control values.

Total RNA (500 ng) was reverse transcribed at 42°C in a 25-μl volume that included 50 ng random hexamer primers (Life Technologies, Rockville, MD) and 200 units Superscript II RT (Life Technologies). The following mouse primers were used: TGFβ 5′-CAC.CTG.CAA.GAC.CAT.CGA.CAT, 3′ -AAA.GAC.AGC.CAC.TCA.GGC.GTA.TC (Genebank accession no. NM-011577); TATA binding protein (TBP) 5′-ACC.CTT.CAC.CAA.TGA.CTC.CTA.TG, 3′ -ATG.ATG.ACT.GCA.GCA.AAT.CGC (Genebank accession no. D01034). Signals to TGFβ were normalized using TBP as the internal standard. Band identity was confirmed using Southern Blot analysis and DNA sequencing. Multiplex quantitative PCR was performed using cDNA (20 ng) in a 47-μl mix that included 1.5 mmol/l MgCl₂ (Perkin Elmer, Foster City, CA), 80 μmol/l dNTP, 10 pmol oligonucleotide primer, 5 units AmpliTaq Gold DNA polymerase (Perkin Elmer), and 2.5 μCi [α-³²P]dCTP. Reaction conditions commenced for 10 min at 94°C followed by 24 cycles for 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C and then final incubation at 72°C. Amplimers were separated on a polyacrylamide gel for analysis.

Ventricle samples were prepared as described (27), and 50 μg SDS extract (S2) was resolved by 4–12% SDS/PAGE. Rat heart collagen VI standard was used (provided by Dr. M.J. Spiro) (4), and membranes were incubated with collagen VI primary antibody (Santa Cruz Biotechnology, Santa Cruz, CA).

Formalin (10%) fixed hearts were bisected transversely at the mid-ventricular level and embedded in paraffin. Sections (∼5 μm thick) were stained with hematoxylin and eosin as well as Masson’s trichrome stain. To determine cardiac myocyte size, the diameter (microns) of 25 myocytes was assessed under ×400 magnification within the left ventricle and a mean was obtained. Fibrosis was quantitated using trichrome-stained sections. Dark blue staining collagen fibers were quantitated as a measure of fibrosis. Color images (six to eight fields per heart section) were obtained under ×40 magnification and digitized using Image-Pro Plus version 3.0 software (Media Cybernetics, Silver Spring, MD). Areas of fibrosis were divided by the area of tissue section measured.

Before immunostaining, antigen retrieval was performed on sections by immersion in Target Retrieval Solution (15 min at 85°C) (Dako, Carpinteria, CA). Endogenous peroxide was blocked using 3% H₂O₂ (5 min). After 45 min of blocking using protein blocking solution (Dako), sections were incubated with 10 μg/ml primary antibody to PKCβ2 (Santa Cruz Biotechnology) or CTGF (Cell Sciences, Norwood, MA). For immunostaining, biotinylatedlinked streptavidin-horse radish peroxidase (Dako LSAB2) antibody was used with DAB chromagen. Slides were counterstained with hematoxylin. Sections for immunofluorescence studies were double stained. CTGF was detected with anti-rabbit Alexa 488 and PKCβ2 with anti-rabbit Alexa 568 (Dako).

Statistical analyses were performed using unpaired Student’s t test and one-way ANOVA, and P values <0.05 indicated significance (SigmaStat Software; Jandel Scientific). Values shown are means ± SE, and n indicates the number of animals used.

Body weight (g) and whole heart weight (mg) were assessed from control mice and transgenic mice overexpressing PKCβ2 in the myocardium at 4, 8, 12, and 16 weeks of age. No significant differences in body weights were observed between the control and transgenic groups at any age (data not shown). Heart weights from transgenic mice increased significantly (P < 0.05) at 8 and 12 weeks of age (138.4 ± 7.1 and 156.1 ± 13.2 mg, respectively) compared with the control group (109.8 ± 3.8 and 114.4 ± 4.7 mg, respectively; n = 12–22). No significant differences (P > 0.05) in heart weights were observed between the control and transgenic groups at 4 weeks (110.2 ± 5.6 and 106.4 ± 4.7 mg, n = 17–19) or 16 weeks (109.3 ± 3.4 and 111.6 ± 3.9 mg, n = 18–23). Similarly, significantly increased (P < 0.05) heart-to-body weight ratios were observed in transgenic mice at 8 weeks (5.7 ± 0.5) and 12 weeks (5.6 ± 0.4) compared with the control group (4.4 ± 0.1 and 4.3 ± 0.1, respectively). At 16 weeks, atrophy of transgenic hearts resulted in similar ratios between the control (4.1 ± 0.1) and transgenic groups (4.3 ± 0.1). A separated group of mice were treated with the PKCβ-specific inhibitor LY333531 for 8 weeks. No differences in body weight, heart weight, or heart-to-body weight ratios were observed between the control (26.4 ± 1.2 g, 129.1 ± 7.0 mg, and 4.9 ± 0.1, respectively; n = 5) and transgenic groups (25.9 ± 1.0 g, 135.5 ± 7.8 mg, and 5.2 ± 0.1, respectively; n = 6).

Northern blot analysis of total RNA isolated from ventricle from transgenic mice at 4, 8, 12, and 16 weeks of age revealed a single 3.6-kb transcript for PKCβ2 mRNA (data not shown). Endogenous PKCβ2 mRNA levels from control ventricles at all durations were almost undetectable. Expression of collagen VI (Fig. 1A) and fibronectin (Fig. 1B) mRNA levels increased significantly (average of 1.6-fold and 2.3-fold, respectively; P < 0.05) in ventricles from transgenic mice at 4, 8, and 12 weeks of age compared with control mice. Collagen VI protein expression also increased significantly (P < 0.05) in ventricle from 8-week transgenic mice by 1.7 ± 0.2-fold compared with controls (1.0 ± 0.0-fold, n = 5) (Fig. 1C). Treatment of control and transgenic mice with LY333531 for 8 weeks reduced this increase in collagen VI protein expression in transgenic hearts to 1.3 ± 0.2-fold, which did not differ significantly (P > 0.05) from control values (1.0 ± 0.1-fold, n = 4) (Fig. 1C). However, expression of collagen VI mRNA remained elevated by 2.1 ± 0.2-fold compared with controls (1.0 ± 0.1-fold, n = 5–6). At 16 weeks, mRNA levels for collagen VI and fibronectin did not differ significantly (P > 0.05) between the groups (Fig. 1A and B). In contrast, collagen VI protein expression in 16-week transgenic hearts revealed a 1.4-fold (n = 2) increase in expression compared with the control group (Fig. 1C).

Significant increases (P < 0.05) in CTGF mRNA expression, from 2.5- to 3.3-fold, were observed in ventricle obtained from transgenic mice at 4, 8, and 12 weeks duration compared with the control group (Fig. 2A). At 16 weeks, a 1.5-fold increase in CTGF expression was observed and remained significantly greater (P < 0.05) than the control group. At 4 weeks duration, TGFβ mRNA levels did not differ between groups, and at 8 and 12 weeks, significant increases (P < 0.05) in TGFβ expression between 1.4- and 1.9-fold were observed compared with controls (Fig. 2B). However, after 16 weeks, no difference in TGFβ expression between the groups was apparent.

Treatment of control and transgenic mice with LY333531 for 8 weeks reduced the increase in expression of CTGF from 3.3-fold in untreated transgenic mice to 1.7 ± 0.3-fold in transgenic mice treated with LY333531, which did not differ significantly from controls treated with LY333531 (1.0 ± 0.2-fold) (Fig. 2C). Interpretation of the effect of LY333531 treatment on TGFβ expression is confusing (Fig. 2D). Expression of TGFβ did not differ significantly between control and transgenic mice treated with LY333531 (Fig. 2D). LY333531 induced a slight reduction in expression of TGFβ, from 1.4 ± 0.1-fold in the untreated transgenic group to 1.3 ± 0.1-fold in the treated group, which did not differ from the control group treated with LY333531 (1.0 ± 0.1-fold). We interpret these findings to indicate that LY333531 did not affect expression of TGFβ in transgenic hearts compared with controls.

Ventricular myocardium from 4- to 16-week control mice had normal histology (Fig. 3A). In contrast, ventricular sections from transgenic mice at all time points displayed myocyte hypertrophy with focal myofiber disarray (Fig. 3B). In 4-week transgenic hearts (Fig. 3C), sites of dystrophic calcification associated with inflammatory cell infiltrate and necrosis commonly occurred. In addition, patchy fibrosis was noted in transgenic hearts at all durations (Fig. 3D and E). Myocardial pathology evolved from 8 to 16 weeks with a reduction in cell infiltration and progressive scar maturity by 16 weeks (Fig. 3F). Increased cardiac myocyte size, indicating myocyte hypertrophy, was determined in ventricles from transgenic mice at all durations and reached significance at 4, 12, and 16 weeks of age (P < 0.05) (Fig. 4A). Quantitation of the area of fibrosis/area measured (percent) confirmed that there was a significant increase in sections from transgenic mice at all durations (P < 0.05) (Fig. 4B) compared with controls. In all sections from transgenic mice quantitated, 2–3% of the measured area was shown to be fibrotic, whereas <0.1% of ventricle sections from control mice exhibited fibrosis.

Ventricular sections from 8-week transgenic mice displayed patchy or scattered immunostaining for PKCβ2. Intense staining was observed in some cardiac myocytes, with adjacent myocytes staining weakly (Fig. 5A). In sections from control mice, staining for PKCβ2 could not be detected (Fig. 5B). Immunofluorescence was used to assess for colocalization of PKCβ2 and CTGF within cardiac myocytes. A similar pattern of patchy staining for PKCβ2 to cardiac myocytes was observed in transgenic sections after immunofluorescent staining (Fig. 5C). Immunofluorescent staining for CTGF was also observed in cardiac myocytes from transgenic hearts (Fig. 5D); staining was weaker compared with that for PKCβ2, and some bleed-over fluorescence could be seen. Using double-labeled immunofluorescence, colocalization of PKCβ2 and CTGF in cardiac myocytes was also observed in some cells (Fig. 5E). To further clarify the colocalization, the double-stained slides were examined using confocal microscopy to eliminate the fluorescent bleed over between the wavelengths used. The confocal images revealed colocalization similar to that seen with epifluorescence (Fig. 5F).

The effect of diabetes on CTGF mRNA expression in ventricle was assessed in 8-week diabetic mice and in 2-, 4-, and 8-week diabetic rats. At all durations, blood glucose levels of STZ-treated mice and rats were significantly elevated (P < 0.05) compared with the corresponding vehicle-treated group (Table 1). Insulin treatment of 4-week diabetic rats reduced the increase in blood glucose levels, whereas for diabetic mice or rats treated with LY333531, levels remained significantly greater (P < 0.05) than the control group. Increases in body weight of STZ-treated mice and rats at all durations tested were significantly reduced (P < 0.05) compared with vehicle-treated animals, although a steady increase in body weight was observed in both the control and diabetic groups (Table 1). Insulin treatment of diabetic rats prevented body weight loss, whereas LY333531 did not affect the body weight of diabetic animals.

Ventricle from 8-week diabetic mice showed a significant 2.4-fold increase (P < 0.05) in CTGF mRNA expression compared with the control group (Fig. 6A). Assessment of collagen VI or fibronectin mRNA levels revealed no difference between 8-week control and diabetic mice (data not shown). Treatment of diabetic mice with LY333531 for 8 weeks did not affect the increase in CTGF mRNA expression compared with controls. TGFβ mRNA expression was also significantly increased, by 3.5-fold, in diabetic mice compared with controls (Fig. 6B). Treatment of diabetic mice with LY333531 for 8 weeks prevented the increase in TGFβ mRNA expression compared with the control group (Fig. 6B).

Ventricle from 2-, 4-, and 8-week diabetic rats showed significant increases (P < 0.05) in CTGF expression of 1.3-, 1.3-, and 1.5-fold, respectively, compared with controls (Fig. 7A and B). Expression of collagen VI mRNA in heart from 4-week diabetic rats did not differ from the control group (data not shown). Treatment of 4-week diabetic rats with insulin prevented the increase in CTGF mRNA expression (Fig. 7C), whereas treatment with LY333531 had little effect on CTGF expression such that levels did not differ significantly from either the untreated diabetic or control groups (Fig. 7C).

In this study, we have characterized the effect of PKCβ2 isoform activation on the expression of CTGF, TGFβ, and ECM proteins and on pathological fibrosis using PKCβ2 transgenic mice and have identified the potential for similar alterations in rodent models of diabetes. Preferential activation of PKCβ2 isoform was identified in diabetic rat heart (10), which leads to the generation of transgenic mice with sustained cardiac activation of PKCβ2 as an accelerated cardiac model of later-stage diabetes. We have reported that ventricle from these mice developed hypertrophy, fibrosis, and contractile dysfunction at 8–12 weeks of age (13). Studies by Bowman et al. (14), in which PKCβ was constitutively activated in adult mouse heart after tetracycline induction, noted progressive cardiac hypertrophy and dysfunction, although no fibrosis by 9 months. Our study focused on the changes in cytokines, ECM, and fibrosis in transgenic mice and diabetic rodents in order to determine a relationship between PKCβ isoform activation and CTGF with fibrosis because such pathologies are observed in cardiomyopathy associated with diabetes and heart failure.

In PKCβ2 transgenic mice, gross pathological changes and increased heart weight were not observed until 8 weeks of age, although increased PKCβ2 expression was observed from 2 to 3 weeks of age postpartum (13). In addition, immunoblot analysis of membranous heart fraction confirmed no change in activity of other PKC isoforms, including PKCα, -δ, or -ε (13). Histological abnormalities, such as fibrosis and myocyte hypertrophy, occurred at 4 weeks and probably were present at 1–2 weeks of age. Cardiac myocyte size was increased at all durations, whereas heart weight increased significantly at 8 and 12 weeks of age. By 16 weeks, atrophy or normalization of heart weight occurs, although cardiac myocyte hypertrophy was still apparent at this duration. At 16 weeks, heart weight may also be influenced by contraction of replacement fibrosis of injured myocytes and by a decrease in new fibrotic sites resulting from loss of myocytes overexpressing PKCβ2. It is interesting that the early lesions manifested as patchy areas of leukocyte infiltrate, calcium deposits, and myofiber disarray with occasional myocyte lysis, indicating an inflammatory response to injury. The exhibition of focalized lesions suggested that the expression of PKCβ2 in the myocardium is not homogenous, and immunohistochemical results demonstrating patchy staining for PKCβ2 to cardiac myocytes support this. This is unexpected because the α myosin heavy chain promoter was used for targeting PKCβ2 isoform expression and suggests that the location of transgene incorporation in the chromosome may influence its expression.

Activation of the PKCβ isoform clearly induced cardiomyopathy with hypertrophy of myocytes and fibrosis. The presence of fibrosis was further supported by the quantitation of mRNA levels of collagen type VI and fibronectin. Increased expression of these proteins in 4- to 12-week transgenic mice coincided with the increase in cardiac myocyte size and the histological appearance of fibrosis; however, by 16 weeks, expression did not differ from control levels. Normalization in ECM expression occurs even though elevated PKCβ2 expression persists. However, both histological analysis and immunoblot assessment showed increased fibrosis and collagen VI protein in heart from 8-week transgenic mice, which continued until 16 weeks, indicating that pathology was still present at this latter duration. An association of PKCβ with these pathologies was confirmed using the PKCβ selective inhibitor LY333531, which reduced the increase in heart weight and collagen VI protein expression and, as previously reported, reduced the size of fibrotic lesions (13). We believe that LY333531 acts to reduce rather than prevent potential cardiac alterations, as it is administered to 3-week-old weaned mice in which myocardial changes may be initiated. It is unclear why the increase in collagen VI mRNA was not reduced by drug treatment; it is possible that transcript half-life or stability is altered or that continued drug treatment may be necessary to reduce levels. Therefore, initial PKCβ2 activation parallels the induction of a fibrotic response that may involve stimulation of cytokine expression, and this coincides with myocyte injury and hypertrophy. Indeed, an increase in collagens and fibronectin is linked to cardiac failure, hypertrophy, infarction, and diabetes, and collagens and fibronectin also exhibit activation or increased expression of the PKCβ isoform (2–5,22,28–30).

Our investigation of cytokines, which can induce the fibrotic response, revealed two potential growth factors: TGFβ, which is known to be regulated by PKC, and CTGF, which has not been linked to PKC; however, both factors possess AP-1 binding sites in their promoter regions. TGFβ is induced in cardiac pathologies, including fibrosis, hypertrophy, and infarction, and precedes elevation of ECM proteins (16,30–33). Recently, the time-course for increased TGFβ expression after myocardial infarction was associated with increased expression of CTGF and was attributed to a TGFβ response element in the CTGF promoter region (16,17). Both correlated with increased fibronectin and collagen expression.

Since its identification and classification, CTGF has been associated with fibrotic disorders of the pancreas, skin, lung, liver, and atherosclerotic plaques (34–40). Moreover, addition of TGFβ increases CTGF mRNA in cardiac fibroblasts, rat neonatal cardiac myocytes, and human cardiac fibroblasts (16). Reports have also shown a lack of dependence among these cytokines. For example, dexamethasone increases CTGF expression and downregulates TGFβ in Balb 3T3 cells (41); increased CTGF and collagen expression in the absence of a TGFβ transcript was reported in pseudoscleroderma associated with lung cancer (42); and CTGF can autoinduce CTGF expression in mesangial cells without affecting TGFβ levels (25).

Alterations in cardiac CTGF expression were of particular interest given recent reports indicating its role in myocardial infarction and fibrosis, elevated expression in heart from ischemic patients (15,16), and localization to cardiac myocytes, fibroblasts, and myoblasts and sites of fibrotic lesions in rodent heart (15). In our study, increased expression of TGFβ in transgenic hearts was observed at 8 and 12 weeks of age, corresponding to increased heart weight. At 16 weeks, no difference in TGFβ expression between the control and transgenic groups was observed. In contrast, significant increases in CTGF expression occurred at all durations, demonstrating that increased expression of CTGF coincided with increased expression of PKCβ2, collagen VI, and fibronectin, before the elevation of TGFβ, and persisted even when TGFβ levels normalized at 16 weeks. Therefore, in this model, cardiac CTGF expression can be modulated independently of TGFβ and may contribute to elevated ECM gene expression and fibrosis. The reduced expression of CTGF in transgenic mice treated with LY333531 indicates increased PKCβ expression associates with increased CTGF expression and myocardial injury. To further confirm an association between PKCβ2 and CTGF, immunohistochemical localization of CTGF to cardiac myocytes colocalized in some cells with staining for PKCβ2. LY333531 did not appear to affect TGFβ expression; expression remained elevated, but the difference was not significant. A longer treatment period may be effective in reducing levels further. Therefore, increased CTGF gene expression associated with PKCβ2 activation occurs early, indicating a potential role in inducing fibrotic change; furthermore, CTGF expression is sustained longer than TGFβ, indicating potential for a role in chronic fibrosis.

Whether elevated CTGF expression alone is sufficient to induce fibrotic changes in this model or an associated increase in TGFβ is required is unknown. Further studies are required to more definitely assess the association between these cytokines and the production of cardiac fibrosis, although previous studies indicate a strong link between increased TGFβ and CTGF expression and increased ECM components in cardiac and renal tissue (15,16,24, 25). Furthermore, although a parallel exists between increased PKCβ2, CTGF, and TGFβ expression with myocardial injury in transgenic mouse heart, it remains to be determined whether PKCβ2 activation can directly induce these cytokines or whether the increase is secondary to PKCβ2-induced cardiac cytoxicity, including necrosis and apoptosis.

The effect of diabetes/high glucose on CTGF expression has been assessed in renal tissue only. Mesangial cells exposed to high glucose show increased expression of CTGF, which could be reduced with an anti-TGFβ antibody and by PKC inhibition, indicating that both TGFβ-dependent and -independent mechanisms were involved (24, 25). In addition, increased CTGF expression was reported in diabetic rat and mouse renal cortex and glomeruli (24,25). We now report for the first time that CTGF expression is increased in ventricle from diabetic mouse and rat heart and that insulin can prevent this increase. In heart from diabetic mice, a corresponding increase in TGFβ expression was noted; similar diabetes-induced increases in TGFβ have been reported in the diabetic kidney and heart, and it has been determined that its promoter region possesses a glucose response element (18–23). Unlike heart from our transgenic model, which showed a fivefold increase in PKC activity (13), we did not identify increased expression of collagen VI or fibronectin mRNA in heart from diabetic rodents. We have previously reported that hearts from 2-week diabetic rats show a 21% increase in membranous PKC activity that is significantly less than that in the transgenic model. Longer durations of diabetes than those used in this study may be necessary to see pathological alterations comparable to those observed in the transgenic mice. For instance, 5-month diabetic rats were used to identify increased cardiac fibrosis, apoptosis, and dysfunction (43). To determine an association between increased cardiac PKCβ expression in the diabetic heart with cytokine expression, animals were treated with LY333531. Inhibition of PKCβ prevented the increase in cardiac expression of TGFβ; however, CTGF expression remained increased in the diabetic rat and mouse heart, indicating a lack of association between PKCβ and increased CTGF expression. These findings could indicate that longer durations of treatment with LY333531 may be necessary to see further reductions in expression of CTGF or that other isoforms of PKC in the diabetic heart are associated with cardiac pathologies. PKC isoforms, including PKCα, -δ, -ε, and -ζ, have been reported to show increased membranous expression in diabetic heart (44–47). In addition, roles for PKCα and -ε have been identified in cardiac hypertrophy and heart failure, and it was reported that different levels of PKCε activation could induce dichotomous cardiac phenotypes (12,48–50). Furthermore, the findings could indicate that other diabetes-associated mechanisms unrelated to PKC activation may be involved. Nevertheless, an association between increased PKCβ and TGFβ expression is indicated in diabetic heart, whereas hyperglycemia is associated with increased cardiac CTGF expression. Future long-term diabetes studies will help to determine the interdependence of these cytokines and their contribution to cardiac pathologies.

In this study, we have shown that PKCβ2 activation is associated with increased myocardial injury and increased expression of the cytokines CTGF and TGFβ. It remains to be determined whether PKCβ2 activity induces cytokine expression to then produce cardiac injury, such as fibrosis and hypertrophy, or whether PKCβ2 is directly toxic to the myocardium, leading to increased cytokine expression. We report for the first time that in diabetic heart, CTGF expression is increased and may play a role in associated pathologies. Although the mechanisms involved in contributing to the diabetes-induced increase in CTGF expression remain unclear, an increase in cardiac CTGF expression may be a marker of myocardial injury. Given that increased cardiac fibrosis ultimately leads to ventricular stiffness regardless of cardiac hypertrophy, therapies aimed to reduced ECM accumulation are required. We now provide evidence to suggest that inhibition of PKCβ activation or the expression of cytokines, such as CTGF, could prevent cardiac pathologies associated with ECM accumulation, particularly in diabetes or heart failure.

This study was supported by the grants NIH EY5110, JDF 22380811, and NIH DK59725-01. K.J.W. is the recipient of a JDFI Postdoctoral Fellowship.

The authors thank Dr. Myra Lipes and Edward Boschetti from the Transgenic Core at the Joslin Diabetes Center for the generation of transgenic mice. They also appreciate the technical assistance provided by Helen Shing, Darren Opland, and Ravi Rauniyar.