http://www.jhltonline.org

Connective tissue growth factor (CTGF/CCN2):A protagonist in cardiac allograft vasculopathydevelopment?Malena P. Pantou, PhD,a Athanasios Manginas, MD,b Peter A. Alivizatos, MD,c andDimitrios Degiannis, MD, PhDa

From the aMolecular Immunopathology and Histocompatibility Laboratory, the bFirst Department of Cardiology, and the cFirst

Department of Cardiac Surgery & Thoracic Transplantation Unit, Onassis Cardiac Surgery Center, Athens, Greece.

BACKGROUND: Connective tissue growth factor (CTGF) has been reported to be upregulated inexperimental models of chronic cardiac allograft rejection. We investigated the contribution ofCTGF to the development of cardiac allograft vasculopathy (CAV), a surrogate marker for chronicrejection.METHODS: This prospective study included 72 adult heart allograft recipients. Genotyping of thers6918698 polymorphism was performed by sequence-specific primer polymerase chain reaction(PCR). CTGF protein levels were measured in serum. CTGF messenger RNA (mRNA) from myocar-dial biopsy specimens was quantified by quantitative real-time PCR.RESULTS: Recipient genotype was associated with the development of CAV (p � 0.014) and thecarriers of the C allele (CC and CG genotype) were high-risk recipients for the development of CAV(odds ratio, 3.30; 95% confidence interval, 1.12–9.74; p � 0.044). Serum CTGF protein levels couldnot be associated with the presence of the C allele but were significantly lower in the patients that haddeveloped CAV (p � 0.038). This was attributed to the addition of everolimus to their immunosup-pression scheme. Myocardial relative CTGF mRNA expression was estimated to be approximatelytwice as much in the CAV patients than in the patients without CAV (p � 0.013).CONCLUSIONS: The important role of CTGF during the development of CAV in heart transplantationwas supported by the association of CAV with the recipient CTGF-945 CC/CG genotypes. The CAVpatients, who were all receiving everolimus treatment, displayed elevated myocardial CTGF mRNAtranscription levels, while everolimus has been observed to reduce serum CTGF protein levels.J Heart Lung Transplant 2012;31:881–7© 2012 International Society for Heart and Lung Transplantation. All rights reserved.

KEYWORDS:cardiac allograftvasculopathy;CCN2;CTGF;everolimus;SNP;serum;cardiac allografttransplantation

In cardiac transplantation, allografts that undergo chronicrejection manifest a progressive loss of function, interstitialfibrosis, and the occlusion of luminal structures. In detail,graft arterial vessels are characterized by gradual sub-endo-thelial tissue development, a process that is referred to ascardiac allograft vasculopathy (CAV).1 Hence, CAV andinterstitial fibrosis are commonly viewed as surrogate mark-ers for chronic rejection, and the changes observed during

Reprint requests: Malena P. Pantou, PhD, Molecular Immunopathologyand Histocompatibility Laboratory, Onassis Cardiac Surgery Center, 356Sygrou Ave, Athens TK 176 74, Greece. Telephone: 0030-210-949-3023.Fax: 0030-210-949-3018.

E-mail address: pantou@ocsc.gr

1053-2498/$ -see front matter © 2012 International Society for Heart and Lungdoi:10.1016/j.healun.2012.02.034

the development of this pathologic condition resemblechronic tissue remodeling and/or wound repair processesthat follow tissue injury.2

CTGF, which has recently risen as a potent mediator ofthese phenomena,3 is a matricellular protein member of theCCN (abbreviation for “connective-tissue-growth-factor,cysteine-rich-protein-61, and nephroblastoma-over-expressed”)family. Like all other members of this family, CTGF displays amodular protein structure. The gene expressing CTGF con-sists of 5 exons and encodes for a 38-kDa cysteine-richmolecule that is secreted in the extracellular matrix (ECM).In the promoter region of the gene, a Smad-binding site and

a signal transducer and activator of transcription (STAT) 3

Transplantation. All rights reserved.

882 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

response element have been shown to be important in theregulation of CTGF expression.4–6

Accentuation of the transcription factor Smad3 or CTGFin the arterial wall has been proven to stimulate intimalhyperplasia, one of the main conversions observed duringarterial remodeling in chronic rejection. In detail, transform-ing growth factor-� (TGF-�) has been reported to activatethe transcription factor Smad3 in medial smooth musclecells, causing them to secrete CTGF, which in turn stimu-lates adventitial fibroblasts to migrate, proliferate, producecollagen, and transform into myofibroblasts.6 In addition,CTGF expression has been shown to be induced in exper-imental models of chronic cardiac allograft rejection.7–9

A polymorphism located in the promoter region of thegene (rs6918698), 945 bp upstream from the ATG startcodon, has been reported to induce its transcriptionalactivity possibly due to a shift of the binding transcrip-tion factors.10 In detail, the presence of cytosine at thecorresponding position increases the affinity of the regionfor the specificity protein (Sp)1 transcription factor andenhances the transcriptional activity of the gene.10 InBrazilian and Chinese fishermen infected with Schisto-somes, rs6918698CC has been associated with hepaticfibrosis, but this association could not be replicated for 2other groups of patients, leading to the assumption thatthe rs6918698 polymorphism may be in linkage disequi-librium with some other causal single nucleotide poly-morphism (SNP).11

Our aim was to explore the contribution of CTGF to thedevelopment of CAV, one of the main processes that char-acterize the progression of chronic rejection to patients whohad undergone cardiac allograft transplantation. In this di-rection we explored for a possible association between thers6918698 polymorphism and the development of CAV.Furthermore, we attempted to assess whether the presenceof the studied SNP and/or the onset of CAV could conse-quently influence CTGF expression. Quantification ofCTGF expression has been attempted through the detectionof serum protein levels along with the relative quantificationof myocardial transcribed CTGF messenger RNA (mRNA)levels.

Materials and methods

The Ethics Committee of the Onassis Cardiac Surgery Centerapproved this study.

Patients and samples

This prospective study included 72 adults who received heartallografts from 1996 to 2009 at the Onassis Cardiac SurgeryCenter. As part of our standard protocol, all heart transplant pa-tients received anti-thymocyte globulin as induction therapy andstatins at moderate doses, with routine evaluation for myopathiccomplications. Extraction of donor DNA was performed fromwhole blood samples at the time of transplantation and stored forsubsequent analyses. For the assessment of CAV, patients under-

went coronary angiography and intravascular ultrasound (IVUS)

imaging during their routine hospital visits. The surveillance CAVprotocol briefly included coronary angiography and IVUS imagingat 1 month, 1 year, and yearly thereafter, or sooner if clinicallyindicated in case of cardiac events or abnormal echocardiographicindices.

Patients were included in the CAV group if newly detected (notpreviously observed) maximal intimal thickness in the proximalleft anterior descending artery exceeded 0.5 mm and/or if angio-graphic findings substantiated the development of CAV. At thatpoint, patients were switched to an immunosuppression regimenthat comprised everolimus in addition to a reduced dose of cal-cineurin inhibitor according to the therapeutic protocol followed inour hospital.

Samples of whole blood and serum, along with routine trans-plant surveillance biopsy specimens were collected during routineperiodic examination of the recipients at least 1 year after trans-plantation. The median time points (months after transplantation)of sampling between patients with and without CAV (No-CAV)were comparable, [61 (range, 13–167) vs 64 (range, 13–165) forserum samples and 65 (range, 12–167) vs 59 (range, 23–165) forbiopsy samples].

Assessment of the “everolimus” effect on serum CTGF lev-els was performed by comparing CTGF levels in serum samplesfrom 10 patients immediately after diagnosis of CAV and be-fore everolimus administration with CTGF levels in samplesobtained from the same patients at a median of 2 months (range,1– 6 months) after the addition of everolimus in the immuno-suppression scheme.

CTGF polymorphism genotyping

A polymorphism corresponding to the position �945 from theATG start codon of the gene (rs6918698) was analyzed. Theanalysis encompassed all recipients and all but 3 donors whoseDNA had not been extracted and stored during transplantation.Genotyping was performed with the use of a sequence-specificprimer-polymerase chain reaction (PCR) assay, as described byFonseca et al.10 The DNA used as template was extracted fromwhole blood samples with the QIAamp DNA Blood mini kit(Qiagen, Valencia, CA).

Measurement of CTGF protein

Plasma levels of CTGF protein for the study patients were mea-sured using a commercially available sandwich enzyme-linkedimmunosorbent assay (ELISA) according to the manufacturer’sinstructions (USCN Life Science Inc, Wuhan, China). In 10 pa-tients, CTGF levels were determined twice —before and after theswitch of their immunosuppression regimen. The lower limit ofdetection of the method was 8.2 pg/ml, and all samples wereanalyzed simultaneously. Seven samples that displayed CTGFserum levels above the upper limit of detection (2,098.4 pg/ml),even after their dilution by 10 times, were assigned the maximumvalue of 20,984 pg/ml.

RNA extraction and real-time PCR analysis

Total RNA was isolated from transplant biopsies using theRNeasy Fibrous Tissue Mini kit (Qiagen). The RNA sampleswere separated in 1% agarose gels containing ethidium bromideand their quality was determined by the visibility of 18S and

28S RNA bands. Total RNA of high quality was processed to

883Pantou et al. Connective Tissue Growth Factor and CAV

complimentary DNA (cDNA) with the use of an oligo-dTprimer provided in the Transcription First Strand cDNA syn-thesis kit (Roche Diagnostics, Mannheim, Germany) accordingto the manufacturer’s instructions. The primer pair used foramplification of the human CTGF gene was 5=-CGACTGGAA-GACACGTTTGG-3= and 5=-AGGCTTGGAGATTTTGGGAG-3=. As an internal standard, a fragment of �-actin was amplifiedwith the use of the primer pair 5=-GATCATTGCTCCTCCT-GAGC-3= and 5=-ACTCCTGCTTGCTGATCCAC-3=. The primerswere designed to overlap exon junctions to avoid the amplification ofDNA.

Amplification reactions were performed in a 20-�l volumeof the LightCycler FastStart DNA Master SYBR Green I mix-ture (Roche Diagnostics) with the addition of primers andMgCl2 at a final concentration of 0.5 �mol/liter and 2 mmol/liter, respectively. All reactions were performed in triplicate ina LightCycler 2.0 with the following thermal cycling condi-tions: 95°C for 10 minutes, 45 cycles of 95°C for 10 seconds,58°C for 10 seconds, and 72°C for 20 seconds. To confirm thespecificity of amplification PCR products were subjected to amelting curve analysis and electrophoresis in a 4% agarosegel.

Statistical analyses

Values are expressed as the mean � standard deviation ormedian with interquartile range. Statistical analyses were per-formed using SPSS 17.0 software (SPSS Inc, Chicago, IL). Alltests were 2-tailed, and differences at values of p � 0.05 wereconsidered significant. The Fisher’s exact test was used toestimate the association of the genotyped SNP with the devel-opment of CAV. The difference of the plasma CTGF concen-tration among different groups was tested by the Kruskal-Wallisor Mann-Whitney test as appropriate. The significance of thedifferent serum CTGF serum protein levels before and after theaddition of everolimus to the immunosuppression regimen ofthe patients that developed CAV was estimated by the Wil-coxon rank sum test. Correlation of serum CTGF concentrationand CTGF mRNA expression in the myocardial biopsy speci-mens was tested by the Kendall’s �. The difference of CTGFmRNA expression between recipients with and without CAVwas assessed by the Mann-Whitney test.

Table 1 Genotype Distributions in Donors and Recipients AccRecipient Follow-up

Genotype rs6918698CAVn (%)

Recipient GG 6 (20)CG 17 (57)CC 7 (23)CC�CG vs GG 24 (80)Total 30 (100)

Donor GG 10 (33)CG 9 (30)CC 11 (37)CC�CG vs GG 20 (67)Total 30 (100)

CAV, cardiac allograft vasculopathy; CI, confidence interval; OR, odd

aAssessed by Fisher’s exact test.

Results

Clinical characteristics of recipients

The study included 72 recipient-donor pairs (recipients: 58males, 14 females) with a mean age at transplantation of38.28 � 13.66 years (range, 14–63 years). Follow-up hadan average duration of 71.9 � 45.4 months. Transplantationwas performed on a background of dilated (48 patients),ischemic (16 patients), hypertrophic (4 patients), restrictive(2 patients), or valvular (1 patient) cardiomyopathy. Onerecipient was originally diagnosed with amyloidosis. Dys-lipidemia was present in 24 patients, diabetes in 10, andmild hypertension in 4 at the time of transplantation. Therewas no significant difference between the distribution of thepatients with dyslipidemia, diabetes, and hypertension to theCAV and No-CAV groups.

Proximal left anterior descending artery intimal thick-ness was significantly different between CAV (30 patients)and No-CAV (42 patients) groups (0.9 � 0.4 vs. 0.3 � 0.1respectively; p � 0.001). Everolimus was added to the CAVgroup at a median of 29.5 months (range, 16.5–54.5months) after transplantation. Two recipients died, 1 patientunderwent retransplantation after loss of graft function dueto development of CAV, and the recipient originally diag-nosed with amyloidosis subsequently required kidney trans-plantation.

Association of CTGF polymorphism with thedevelopment of CAV

In our analysis, a polymorphism corresponding to the pro-moter region of the CTGF gene (rs6918698) was studied.The genotype distribution of CTGF was not in accordancewith Hardy-Weinberg equilibrium (p � 0.001). The asso-ciation of the genotyped SNP (CTGF �945C/G) with thedetection of CAV is presented in Table 1. Recipient geno-type was associated with the development of CAV (p �0.014), and recipient genotypes, including the C allele (CC

to the Occurrence of Cardiac Allograft Vasculopathy During

No CAVOR (95% CI) p-valuean (%)

19 (45) 0.01410 (24)13 (31)23 (55) 3.3 (1.1–9.7) 0.04442 (100)12 (31) �0.99913 (33)14 (36)27 (69) 0.9 (0.3–2.5) �0.99939 (100)

ording

s ratio.

884 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

and CG genotype), were high-risk genotypes for CAV de-velopment (odds ratio [OR], 3.30, 95% confidence interval[CI], 1.12–9.74; p � 0.044). These promising data, alongwith the observation that the deposition of ECM is mainlymediated by donor fibroblasts and cardiomyocytes,12 led usto extend our study to include donor genotypes. However,our findings did not support any association between the Callele in the donor genotype and the development of CAV(p � 0.999).

Serum CTGF concentration in relation to thedevelopment of CAV

Serum CTGF concentration was measured in all study pa-tients because the detected C allele of the rs6918698 poly-morphism has been shown in vitro and in vivo to change thebalance of the basic transcriptional activity of the CTGFgene toward enhanced transcription.10 Nevertheless, no as-sociation could be established between the presence of the Callele in the recipient and/or donor genotype and serumCTGF protein levels (data not shown). On the contrary anddespite the association of the C allele with the developmentof CAV, patients who had developed CAV displayed sig-nificantly lower serum CTGF protein levels than patientswho did not exhibit CAV (306.31 [interquartile range(IQR), 158.51–1,262.63] pg/ml vs 747.06 [IQR, 344.70–2,662.13] pg/ml, p � 0.038; Figure 1).

We hypothesized that this controversy could be partiallyattributed to the switch of the pharmacologic regimen in allpatients who developed CAV (ie, the addition of everolimusto their immunosuppression therapy). To investigate thishypothesis, we examined serum samples from 10 patientsbefore and after the addition of everolimus to their immu-nosuppression regimen. CTGF serum protein levels were

Figure 1 Box and whisker plot displays the serum concentra-tions of connective tissue growth factor (CTGF) protein accordingto the occurrence of cardiac allograft vasculopathy (CAV). Thehorizontal line in the middle of each box indicates the median, thetop and bottom borders of the box mark the 75th and 25th per-centiles, respectively, the whiskers mark the 10th and 90th per-centiles, and the 5th and 95th percentiles are marked as circleddots. The difference of the concentration in each group was tested

by the Mann-Whitney test.

significantly higher before the administration of everolimus(471.90 [IQR, 268.15–1,005.44] pg/ml vs 195.77 [IQR,141.99–514.94] pg/ml, p � 0.009), suggesting an activerole of everolimus to the regulation of the CTGF expression(Figure 2). Consequently, no correlation could be estab-lished between the extent of CAV defined by IVUS imagingand the serum CTGF concentration (data not shown).

Elevated intragraft CTGF expression is associatedwith CAV

To study the expression pattern of the CTGF gene in myo-cardial biopsy specimens, levels of CTGF mRNA werequantified by real-time PCR in 14 patients who had notdeveloped CAV and 9 patients who had developed CAV.Expression levels were calculated as a ratio between CTGFand the reference gene �-actin to correct for the variation inthe amount of RNA. Relative expression levels were deter-mined after the normalization of results against the medianCTGF expression level in biopsy specimens from patientswho had not developed CAV. CTGF mRNA expression wasestimated to be approximately twice as much in the CAVpatients than in the No-CAV patients (2.18 [IQR, 1.22–5.44] vs 1 [IQR, 0.57–1.55], p � 0.013; Figure 3), suggest-ing an important role of this protein during the development

Figure 2 Serum levels of connective tissue growth factor(CTGF) measured by enzyme-linked immunosorbent assay in 10cardiac allograft recipients were significantly reduced after theaddition of everolimus to the immunosuppression regimen. Statis-tical significance (p � 0.009) was estimated by the Wilcoxon ranksum test.

of this fibrotic disorder.

885Pantou et al. Connective Tissue Growth Factor and CAV

No correlation could be established between serumCTGF protein levels and CTGF mRNA transcription in themyocardial biopsy specimens, undermining the merit ofserum CTGF as a potential biomarker for the developmentof chronic rejection. Similarly, CTGF mRNA levels did notcorrelate with the severity of CAV as quantified by themaximum intimal thickness (data not shown).

Discussion

CAV remains of primary concern in heart transplant recip-ients, having distinct features from atherosclerosis,13 andalthough studies have focused on the driving mechanisms ofCAV development, these mechanisms are poorly under-stood and probably involve immunologic and non-immuno-logic triggers.14 We investigated for a possible link betweenthe development of CAV, one of the main processes thatcharacterize the phenotype of chronic rejection in patientswho had undergone cardiac allograft transplantation, andCTGF, a major pro-fibrotic factor that frequently acts down-stream of TGF-�–mediated fibrogenic pathways.

We focused on a polymorphism (rs6918698) participat-ing in a Sp-factor binding site located in the promoter regionof CTGF. Our data support an association between thers6918698 polymorphism in the genotype of heart allograftrecipients and the development of CAV (p � 0.014),whereas recipients carrying the C allele, which has beenassociated with enhanced transcriptional activity, werehigh-risk genotypes for development of CAV. These find-ings come in agreement with data from experimental mod-els of chronic cardiac allograft rejection that support anincreased CTGF expression during the development of this

Figure 3 Box and whisker plot displays connective tissuegrowth factor (CTGF) expression according to the occurrence ofcardiac allograft vasculopathy (CAV). The horizontal line in themiddle of each box indicates the median, the top and bottomborders of the box mark the 75th and 25th percentiles, respec-tively, and the whiskers mark the 10th and 90th percentiles. Thedifference of CTGF mRNA expression in each group was tested bythe Mann-Whitney test.

pathologic condition.7–9

Cardiac fibroblasts are one of the most important celltypes that produce and maintain myocardial ECM bal-ance,12 and therefore, the excessive deposition of ECM inthe setting of cardiac allograft chronic rejection might bedonor derived. Our study was extended to include donorgenotypes, but no association could be established betweenthe polymorphism and the development of CAV. In thisregard, CTGF protein produced by recipient infiltratingcells could induce, in a “feed forward” process, the expres-sion of CTGF by donor cardiomyocytes as has been ob-served in vitro.15 In support of this hypothesis, it has beenfound in sex-mismatched transplantations that 3% to 15% ofthe neointimal smooth muscle cells are recipient derived.16

Furthermore and in addition to cardiac fibroblasts and car-diac myocytes that have been reported to produce CTGF,��-T cells also express CTGF mRNA and produce CTGFprotein.17,18 Data from the heart and kidney of experimentalmodels showed that CTGF expression colocalizes veryclosely with infiltrated leucocytes.7,19

Serum CTGF has been proposed as a potential biomarkerof cardiac dysfunction and its concentration was positivelycorrelated with New York Heart Association heart failurefunctional class.20,21 CTGF serum levels were also linked tothe degree of liver fibrosis of patients infected with hepatitisC virus, even though no association could be establishedbetween the genotypes and the alleles for 6 polymorphisms(including rs6918698) and the severity of hepatic fibrosis.22

In our cohort, serum CTGF levels were not associated to therecipient or donor genotype, despite data supporting thelinkage of the C allele with enhanced transcriptional activityof the gene.10 Furthermore and surprisingly, patients whohad developed CAV displayed significantly lower serumCTGF protein levels than patients who did not have CAV.

Because the expression of CTGF has been reported to beaffected by various pharmacologic agents like 3-hydroxy-3 methyl-glutaryl-coenzyme A reductase inhibitors (st-atins),23 angiotensin II type 1 receptor antagonists (cande-sartan).24 and mammalian target of rapamycin (mTOR) in-hibitors (rapamycin),25 our hypothesis was that the additionof everolimus to the immunosuppression regimen of pa-tients who presented CAV might account for the puzzlingresults. In defense of this hypothesis, serum samples takenfrom 10 patients displayed significantly lower proteinCTGF levels after the addition of everolimus (Figure 2).

No correlation could be detected between serum CTGFprotein levels and myocardial CTGF mRNA levels. Be-sides, CTGF is a matricellular protein that exerts its actionat a local level, and therefore, it is conceivable that itspresence in the serum does not quantitatively correlate withits presence at different tissues. Furthermore, the exact ef-fect of the administration of everolimus seems to be contextand tissue dependent because CTGF mRNA expression hasbeen reported to be induced by the administration ofeverolimus in the heart and kidney of transgenic rats har-boring human renin and angiotensinogen genes,19 whereasthe administration of sirolimus, another mTOR inhibitor, in

renal transplant recipients with biopsy-proven chronic allo-

886 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

graft nephropathy revealed that sirolimus attenuates renalCTGF expression.26

Regardless of everolimus, several groups have reporteddata from experimental models indicating that CTGF isupregulated and functions as a downstream mediator offibrosis in chronic cardiac allograft rejection.7–9 In the hu-man setting of our study, the relative CTGF mRNA expres-sion was estimated to be approximately twice as much in theCAV patients than in the No-CAV patients (Figure 3),although CTGF mRNA transcription did not correlate withthe severity of CAV as quantified by the maximum intimalthickness (data not shown).

Because all of the CAV patients in our study werereceiving everolimus at the time of biopsy, it is difficult todiscern the relative contribution of everolimus and CAV onthe mRNA expression that was observed. In this regardthere are 3 distinct possibilities:

1. Everolimus, similarly to what was observed in serum,reduces CTGF myocardial expression. In that case, theobserved elevated levels of CTGF mRNA in the CAVgroup would have been more intense had not the patientsreceived everolimus.

2. Everolimus induces myocardial CTGF expression, sim-ilarly to what was observed by Finckenberg et al19 intheir transgenic rats harboring human renin and angio-tensinogen genes. In that case the observed elevatedlevels of CTGF mRNA in the CAV patients would bepartially everolimus driven.

3. Everolimus does not affect myocardial CTGF expres-sion, and the observed levels can be attributed to thedevelopment of CAV alone.

Our data do not permit us to distinguish between thesepossibilities because the Hospital Ethics Committee limitedthe number of myocardial biopsy specimens we could ob-tain to those that were part of the routine protocol andprevented us from acquiring additional specimens from pa-tients at the onset of CAV and before the administration ofeverolimus.

Connective tissue growth factor has been proposed to bea useful prognostic marker because its levels in variousbiologic fluids, such as blood plasma, urine, and vitreous,were found to correlate with the presence or the progressionof various fibrotic diseases.27 In the setting of cardiac allo-graft transplantation, serum protein levels of CTGF werereduced by the administration of everolimus and thereforecould not be associated with the development of CAV.Nevertheless, myocardial CTGF mRNA levels were ele-vated in patients who developed CAV, pointing to an activerole of this protein during the progression of chronic allo-graft rejection in accordance with extensive data from ani-mal models7–9 that support the contribution of CTGF to thedevelopment of cardiac fibrosis. In the human setting, how-ever, the pathogenesis of CAV is still obscure and complex;therefore, it remains possible that the CTGF concentrationdifferences observed in our study constitute an epiphenom-

enon.

The role of CTGF was further supported by the associ-ation of the rs6918698 polymorphism with the developmentof CAV, a polymorphism that has been reported to enhancethe transcriptional activity of the CTGF gene. This associ-ation awaits confirmation by a study of significantly largernumber of patients. The exact mechanism for this associa-tion remains to be resolved as the presence of the mutatedallele could either constitute a causal parameter or alterna-tively could represent an accelerator of the progression ofthis pathologic condition.

Disclosure statementThis work was supported by a research grant from the CentralHellenic Health Council (KESY), Ministry of Health.

None of the authors has a financial relationship with a com-mercial entity that has an interest in the subject of the presentedmanuscript or other conflicts of interest to disclose.

References

1. Libby P, Pober JS. Chronic rejection. Immunity 2001;14:387-97.2. Orosz CG, Pelletier RP. Chronic remodeling pathology in grafts. Curr

Opin Immunol 1997;9:676-80.3. Shi-Wen X, Leask A, Abraham D. Regulation and function of con-

nective tissue growth factor/CCN2 in tissue repair, scarring and fibro-sis. Cytokine Growth Factor Rev 2008;19:133-44.

4. Holmes A, Abraham DJ, Sa S, Shiwen X, Black CM, Leask A. CTGFand SMADs, maintenance of scleroderma phenotype is independent ofSMAD signaling. J Biol Chem 2001;276:10594-601.

5. Okada H, Kikuta T, Inoue T et al. Dexamethasone induces connectivetissue growth factor expression in renal tubular epithelial cells in amouse strain-specific manner. Am J Pathol 2006;168:737-47.

6. Kundi R, Hollenbeck ST, Yamanouchi D, et al. Arterial genetransfer of the TGF-b signalling protein Smad3 induces adaptiveremodelling following angioplasty: a role for CTGF. CardiovascRes 2009;84:326-35.

7. Csencsits K, Wood SC, Lu G, et al. Transforming growth factorbeta-induced connective tissue growth factor and chronic allograftrejection. Am J Transplant 2006;6:959-66.

8. Yuan YC, Xia ZK, Mu JJ, Zhang QC, Yin BL. Increased connectivetissue growth factor expression in a rat model of chronic heart allograftrejection. J Formos Med Assoc 2009;108:240-6.

9. Booth AJ, Csencsits-Smith K, Wood SC, Lu G, Lipson KE, BishopDK. Connective tissue growth factor promotes fibrosis downstream ofTGF� and IL-6 in chronic cardiac allograft rejection. Am J Transplant2010;10:220-30.

10. Fonseca C, Lindahl GE, Ponticos M, et al. A polymorphism in theCTGF promoter region associated with systemic sclerosis. N EnglJ Med 2007;357:1210-20.

11. Dessein A, Chevillard C, Arnaud V, et al. Variants of CTGF areassociated with hepatic fibrosis in Chinese, Sudanese, and Braziliansinfected with Schistosomes. J Exp Med 2009;206:2321-8.

12. Brown RD, Ambler SK, Mitchell MD, Long CS. The cardiac fibro-blast: therapeutic target in myocardial remodeling and failure. AnnuRev Pharmacol Toxicol 2005;45:657-87.

13. Rahmani M, Cruz RP, Granville DJ, McManus BM. Allograft vascu-lopathy versus atherosclerosis. Circ Res 2006;99:801-15.

14. Hasegawa T, Iwanaga K, Hultquist DE, Liao H, Visovatti SH,Pinsky DJ. Suppression of nitrosative and oxidative stress to reducecardiac allograft vasculopathy. Am J Physiol Heart Circ Physiol

2009;H1007-16.

887Pantou et al. Connective Tissue Growth Factor and CAV

15. Wang X, McLennan SV, Allen TJ, Twigg SM. Regulation of pro-inflammatory and pro-fibrotic factors by CCN2/CTGF in H9c2 cardi-omyocytes. J Cell Commun Signal 2010;4:15-23.

16. Suzuki J, Isobe M, Morishita R, Nagai R. Characteristics of chronicrejection in heart transplantation: important elements of pathogenesisand future treatments. Circ J 2010;74:233-9.

17. Chen MM, Lam A, Abraham JA, Schreiner GF, Joly AH. CTGFexpression is induced by TGF-� in cardiac fibroblasts and cardiacmyocytes: a potential role in heart fibrosis. J Mol Cell Cardiol 2000;32:1805-19.

18. Workalemahu G, Foerster M, Kroegel C, Braun RK. Human gammadelta-T lymphocytes express and synthesize connective tissue growthfactor: effect of IL-15 and TGF-beta 1 and comparison with alphabeta-T lymphocytes. J Immunol 2003;170:153-7.

19. Finckenberg P, Inkinen K, Ahonen J, et al. Angiotensin II inducesconnective tissue growth factor gene expression via calcineurin-de-pendent pathways. Am J Pathol 2003;163:355-66.

20. Koitabashi N, Arai M, Niwano K, et al. Plasma connective tissuegrowth factor is a novel potential biomarker of cardiac dysfunction inpatients with chronic heart failure. Eur J Heart Fail 2008;10:373-9.

21. Bergestuen DS, Gravning J, Haugaa KH, et al. Plasma CCN2/connec-tive tissue growth factor is associated with right ventricular dysfunc-

tion in patients with neuroendocrine tumors. BMC Cancer 2010;10:6.

22. Kovalenko E, Tacke F, Gressner OA, et al. Validation of connectivetissue growth factor (CTGF/CCN2) and its gene polymorphism asnoninvasive biomarkers for the assessment of liver fibrosis. J ViralHepat 2009;16:612-20.

23. Goppelt-Struebe M, Hahn A, Iwanciw D, Rehm M, Banas B. Regu-lation of connective tissue growth factor (ccn2;ctgf) gene expression inhuman mesangial cells: modulation by HMG CoA reductase inhibitors(statins). Mol Pathol 2001;54:176-9.

24. Guo P, Nishiyama A, Rahman M, et al. Contribution of reactiveoxygen species to the pathogenesis of left ventricular failure in Dahlsalt-sensitive hypertensive rats: effects of angiotensin II blockade.J Hypertens 2006;24:1097-104.

25. Khoo YT, Ong CT, Mukhopadhyay A, et al. Upregulation of secretoryconnective tissue growth factor (CTGF) in keratinocyte-fibroblast co-culture contributes to keloid pathogenesis. J Cell Physiol 2006;208:336-43.

26. Liu M, Zhang W, Gu M, et al. Protective effects of sirolimus byattenuating connective tissue growth factor expression in humanchronic allograft nephropathy. Transplant Proc 2007;39:1410-5.

27. Daniels A, van Bilsen M, Goldschmeding R, van der Vusse GJ, vanNieuwenhoven FA. Connective tissue growth factor and cardiac fibro-

sis. Acta Physiol (Oxf) 2009;195:321-38.