CTGF Inhibits BMP-7 Signaling in Diabetic Nephropathy
Tri Q. Nguyen*,
Peggy Roestenberg*,
Frans A. van Nieuwenhoven*,
Niels Bovenschen*,
Zeke Li,
Leon Xu,
Noelynn Oliver,
Jan Aten,
Jaap A. Joles,
Cecilia Vial||,
Enrique Brandan||,
Karen M. Lyons¶ and
Roel Goldschmeding*
Departments of * Pathology and Nephrology, University Medical Center Utrecht, Utrecht, and Department of Pathology, Academic Medical Center, Amsterdam, Netherlands; FibroGen, Inc., South San Francisco, and ¶ Department of Molecular, Cell, and Developmental Biology, University of California, Los Angeles, California; and || Department of Cell and Molecular Biology, Catholic University of Chile, Santiago, Chile
Correspondence: Dr. Roel Goldschmeding, University Medical Center Utrecht, Department of Pathology, H04.312, Heidelberglaan 100, 3584 CX, Utrecht, Netherlands. Phone: +31-88-755-6559; Fax: +31-30-254-4990; E-mail: R.Goldschmeding{at}umcutrecht.nl
Received for publication November 29, 2007.
Accepted for publication May 21, 2008.
In diabetic nephropathy, connective tissue growth factor (CTGF)is upregulated and bone morphogenetic protein 7 (BMP-7) is downregulated.CTGF is known to inhibit BMP-4, but similar cross-talk betweenBMP-7 and CTGF has not been studied. In this study, it was hypothesizedthat CTGF acts as an inhibitor of BMP-7 signaling activity indiabetic nephropathy. Compared with diabetic wild-type CTGF+/+mice, diabetic CTGF+/– mice had approximately 50% lowerCTGF mRNA and protein, less severe albuminuria, no thickeningof the glomerular basement membrane, and preserved matrix metalloproteinase(MMP) activity. Although the amount of BMP-7 mRNA was similarin the kidneys of diabetic CTGF+/+ and CTGF+/– mice, phosphorylationof the BMP signal transduction protein Smad1/5 and expressionof the BMP target gene Id1 were lower in diabetic CTGF+/+ mice.Moreover, renal Id1 mRNA expression correlated with albuminuria(R = –0.86) and MMP activity (R = 0.76). In normoglycemicmice, intraperitoneal injection of CTGF led to a decrease ofpSmad1/5 in the renal cortex. In cultured renal glomerular andtubulointerstitial cells, CTGF diminished BMP-7 signaling activity,evidenced by lower levels of pSmad1/5, Id1 mRNA, and BMP-responsiveelement–luciferase activity. Co-immunoprecipitation, solid-phasebinding assay, and surface plasmon resonance analysis showedthat CTGF binds BMP-7 with high affinity (Kd approximately 14nM). In conclusion, upregulation of CTGF inhibits BMP-7 signaltransduction in the diabetic kidney and contributes to alteredgene transcription, reduced MMP activity, glomerular basementmembrane thickening, and albuminuria, all of which are hallmarksof diabetic nephropathy.
Connective tissue growth factor (CTGF) is considered an importantfactor in the development of diabetic nephropathy. The renalexpression of CTGF mRNA and protein is upregulated in humanand experimental diabetic nephropathy.1–3 In patientswith diabetic nephropathy, both plasma CTGF levels and urinaryCTGF excretion are increased and correlate with clinical markersof renal disease.4–7 Recently, it was demonstrated thatspecific downregulation of CTGF by antisense oligonucleotidetreatment attenuated albuminuria and mesangial matrix expansionin experimental type 1 and type 2 diabetic nephropathy.8 Theseobservations suggest that CTGF is a critical determinant ofstructural and functional damage in diabetic nephropathy.
Several mechanisms have been proposed for a pathogenic roleof CTGF in diabetic nephropathy. Studies with renal cells demonstratedthat CTGF is involved in diabetes-associated changes such asextracellular matrix synthesis, cell migration, cellular hypertrophy,and epithelial-to-mesenchymal transition.2,9–11 CTGF mightexert these effects in diabetes by modulating the activity ofother growth factors. For example, CTGF is known to enhanceprofibrotic activity of TGF-β1 and IGF-1, which involvesphysical interaction of CTGF with these growth factors.12,13In contrast, binding to CTGF potently antagonizes the signalingactivity of bone morphogenetic protein 4 (BMP-4) in osteogenesisassays and in embryonic patterning.12 In the kidney, the importanceof several BMP, including BMP-4, has been demonstrated mainlyin developmental studies14–16; however, thus far, onlyBMP-7 has also been studied for its contribution as an antifibroticand proregenerative factor in response to injury of the adultkidney.17,18 Renal expression of BMP-7 is progressively decreasedduring the course of human and experimental diabetic nephropathyand in podocytes cultured in high glucose medium,19–21whereas restoration of BMP-7 availability has resulted in preventionor even reversal of functional and structural changes of diabeticnephropathy.22–24 Although it has been hypothesized thatCTGF might inhibit BMP-7,25,26 this has not been addressed experimentally;therefore, we set out to investigate the impact of CTGF on BMP-7signal transduction and target gene expression in experimentaldiabetic nephropathy and in cultured renal cells.
BMP-7 Deficiency in Diabetic Nephropathy Is Accompanied by Decrease of pSmad1/5 and BMP-Target Gene Expression
Induction of diabetes in C57BL6/J mice by intraperitoneal injectionof streptozotocin resulted in characteristic features of diabeticnephropathy, including persistent hyperglycemia; increased glycosylatedhemoglobin levels; proteinuria; structural changes of the kidney;and increased CTGF levels in kidney, urine, and plasma.3 Indiabetic mice, renal cortical expression of BMP-7 mRNA was decreased2.6-fold. This was accompanied by reduced levels of pSmad1/5protein, and also BMP-7 downstream target Id1 was decreased(Figure 1).
Figure 1. Decrease of BMP-7 signal transduction and target gene expression in diabetic nephropathy. Diabetes was induced in C57BL6/J mice by injection of streptozotocin. Renal cortex was harvested 9 wk after injection. Gene expression of BMP-7 and Id1 was evaluated by quantitative PCR, and pSmad1/5 protein level was analyzed by Western blotting. (A through C) In diabetic mice, expression of BMP-7 was decreased 2.6-fold. This was accompanied by reduced levels of pSmad1/5 protein and Id1 mRNA. Data are means ± SD. *P < 0.05.
CTGF+/– Mice Have Lower Levels of CTGF mRNA and Protein
After 17 wk of diabetes, CTGF mRNA and protein expression inrenal cortex and CTGF levels in plasma and urine were increasedin diabetic CTGF+/+ mice as compared with nondiabetic CTGF+/+mice. In diabetic CTGF+/– mice, expression of CTGF mRNAand protein in renal cortex and CTGF levels in plasma were notsignificantly different from nondiabetic mice. Although urinaryCTGF excretion in diabetic CTGF+/– mice seemed to be lowerthan in diabetic CTGF+/+ mice, this difference did not reachstatistical significance. In nondiabetic CTGF+/– mice,CTGF expression seemed slightly lower than in nondiabetic CTGF+/+,but this difference was NS (Figure 2, A through D).
Figure 2. Levels of CTGF mRNA and protein are decreased in CTGF+/– mice. Diabetes was induced in CTGF+/+ and CTGF+/– mice. (A through D) CTGF mRNA and protein in renal cortex and CTGF levels in plasma were two-fold increased in diabetic CTGF+/+ mice, as compared with diabetic CTGF+/– mice. Also urinary CTGF excretion tended to be lower in diabetic CTGF+/– mice than in diabetic CTGF+/+ mice (P = 0.29). (E) CTGF immunohistochemistry showed more prominent glomerular CTGF staining in diabetic CTGF+/+ mice than in diabetic CTGF+/– mice and nondiabetic mice. Data are means ± SD. *P < 0.05 versus control CTGF+/+ mice; **P < 0.05 versus diabetic CTGF+/+ mice.
CTGF immunohistochemistry showed more prominent glomerular CTGFstaining in diabetic CTGF+/+ mice than in diabetic CTGF+/–mice and in nondiabetic mice. No prominent CTGF staining wasobserved in tubuli (Figure 2E).
Diabetic Nephropathy Is Attenuated in CTGF+/– Mice
Albuminuria was increased in diabetic CTGF+/+ mice as comparedwith nondiabetic mice. In diabetic CTGF+/– mice, albuminuriawas significantly less pronounced than in diabetic CTGF+/+ mice(111 ± 39 versus 176 ± 26 mg/g creatinine; P =0.024; Figure 3A).
Figure 3. Diabetic nephropathy is attenuated in CTGF+/– mice. Diabetic nephropathy in CTGF+/+ and CTGF+/– mice was assessed by albuminuria, GBM thickening, and MMP activity. (A) In diabetic CTGF+/– mice, albuminuria was significantly less pronounced than in diabetic CTGF+/+ mice. (B) GBM thickening in diabetic CTGF+/+ mice, as determined by electron microscopy, was absent in diabetic CTGF+/– mice. (C and D) In situ zymography on renal sections (green, MMP activity; red, nuclear counterstain; C) and colorimetric detection in renal lysates (D) showed that MMP activity was decreased in diabetic CTGF+/+ mice. In contrast, MMP activity was preserved in diabetic CTGF+/– mice. Data are means ± SD. *P < 0.05 versus control CTGF+/+ mice; **P < 0.05 versus diabetic CTGF+/+ mice.
Examination of ultrathin sections by electron microscopy showedthat the thickness of the glomerular basement membrane (GBM)in diabetic CTGF+/+ mice was increased compared with controlCTGF+/+ mice; however, no increase in GBM thickness was observedin diabetic CTGF+/– mice as compared with diabetic CTGF+/+mice (154 ± 3.0 versus 174 ± 2.3 nm; P < 0.01;Figure 3B).
In situ zymography showed that gelatinase activity, representingactivity of matrix metalloproteinase 2 (MMP-2) and MMP-9, waslocalized mainly in glomeruli. Activity of gelatinase was decreasedin diabetic CTGF+/+ mice but was preserved in diabetic CTGF+/–mice (Figure 3C). This was confirmed by quantification of gelatinolyticactivity in renal lysates, showing a reduction by approximately40% in diabetic CTGF+/+ mice (P = 0.034) but not in diabeticCTGF+/– mice (Figure 3D).
Preserved BMP Signaling Activity in Diabetic CTGF+/– Mice: Correlation of CTGF Level with Albuminuria and MMP Activity
Renal cortical expression of BMP-7 mRNA was similar in diabeticCTGF+/+ and CTGF+/– mice. Diabetic CTGF+/– micehad relatively preserved pSmad1/5 protein levels and Id1 mRNAexpression as compared with diabetic CTGF+/+ mice, whereas totalSmad5 protein was not different (Figure 4, A through C). Furthermore,Id1 mRNA correlated with albuminuria (R = –0.86, P = 0.011)and with MMP activity in renal lysates (R = 0.76, P = 0.037;Figure 4, D and E).
Figure 4. CTGF inhibits BMP signaling activity in diabetic nephropathy. Renal cortex of diabetic CTGF+/+ and CTGF+/– mice was harvested 17 wk after induction of diabetes. (A) Renal cortical expression of BMP-7 mRNA was similar in diabetic CTGF+/+ and CTGF+/– mice. (B and C) In diabetic CTGF+/– mice, pSmad1/5 protein and Id1 mRNA were higher than in diabetic CTGF+/+ mice, whereas total Smad5 was not different. (D and E) Id1 mRNA correlated with albuminuria and MMP activity. Data are means ± SD. *P < 0.05.
Injection of Recombinant CTGF Impairs Renal Cortical BMP-7 Activity in Nondiabetic Mice
Intraperitoneal injection of recombinant CTGF in nondiabeticmice resulted in transient elevation of plasma CTGF levels (Figure 5A).Although renal cortical expression of BMP-7 and Id1 mRNA wasnot different at the various time points (data not shown), pSmad1/5protein levels were significantly decreased 4 h after injection(P = 0.0083; Figure 5B). Injection with vehicle did not alterrenal pSmad1/5 (Figure 5C).
Figure 5. Injection of CTGF impairs renal cortical BMP-7 activity in nondiabetic mice. Recombinant CTGF or vehicle only was injected intraperitoneally into nondiabetic BALBc mice. (A) CTGF levels in plasma were increased at 10 min after injection with rhCTGF. (B) Renal cortical pSmad1/5 protein levels were significantly decreased 4 h after injection with rhCTGF. (C) Renal cortical pSmad1/5 levels were not different 4 h after vehicle injection. Data are means ± SD. *P < 0.05.
CTGF Inhibits BMP-7 Signal Transduction and Target Gene Expression in Renal Cells
Treatment of rat mesangial cells and HK-2 cells with BMP-7 resultedin increased phosphorylation of Smad1/5 protein and Id1 mRNAexpression. Addition of CTGF partially inhibited phosphorylationof Smad1/5 and reduced expression of Id1 mRNA (Figure 6, A throughD). Also, in mouse podocytes, BMP-7 stimulation resulted ina significant, although less pronounced, increase of Id1 mRNAexpression, which was reduced by addition of CTGF (Figure 6E).
Figure 6. CTGF inhibits BMP-7 signal transduction and target gene expression in renal cells. Effects of CTGF on BMP-7 were studied in rat mesangial cells, mouse podocytes, proximal tubular epithelial cells (HK-2), and renal interstitial fibroblasts (TK173). (A and B) Treatment of rat mesangial cells with BMP-7 resulted in an increase of Smad1/5 phosphorylation and Id1 mRNA expression, both of which were inhibited by co-stimulation with CTGF. (C and D) Treatment of HK-2 cells with BMP-7 resulted in an increase of pSmad1/5 protein and Id1 mRNA, both of which were inhibited by co-stimulation with CTGF. (E) Treatment of podocytes with BMP-7 resulted in increase of Id1 mRNA, which was inhibited by co-stimulation with CTGF. (F) TK173 cells were transfected with the BRE-luciferase construct. Treatment with BMP-7 of cells co-transfected with control vector pCAGGS-lacZ resulted in increase of BRE-luciferase activity, whereas co-transfection with pCAGGS-CTGF inhibited BMP-7–induced luciferase activity. Data are means ± SD. *P < 0.05 versus only 50 ng/ml BMP-7; #P < 0.05 versus p-lacZ with 10 ng/ml rhBMP-7; P < 0.05 versus p-lacZ with 50 ng/ml BMP-7.
CTGF also inhibited activation of a BMP-responsive element (BRE)-luciferaseconstruct, which specifically reports Smad1/5-mediated genetranscription.27 When TK173 cells were co-transfected with BRE-luciferaseand pCAGGS-mCTGF, BMP-7–induced luciferase activity wassignificantly lower than in cells co-transfected with pCAGGS-lacZcontrol vector (Figure 6F).
CTGF Binds BMP-7 with High Affinity
Co-immunoprecipitation experiments showed that captured BMP-7on anti–BMP-7 mAb-coated agarose beads was able to bindCTGF. In the absence of BMP-7, only a weak CTGF band was detected,which is due to known nonspecific binding of CTGF to agarosebeads. Incubation with captured HGF on anti-HGF mAb-coated beadsand with BMP-7–and IgG1-coated beads also did not resultin significant pull-down of CTGF (Figure 7A). Similarly, solid-phasebinding assay showed direct physical interaction between BMP-7and CTGF but not between HGF and CTGF or BSA and CTGF. Increasedbinding between was observed as the concentration of eitherBMP-7 or CTGF was increased (Figure 7B).
Figure 7. CTGF binds directly to BMP-7. Physical interaction between CTGF and BMP-7 was demonstrated by co-immunoprecipitation, solid-phase binding assay, and surface plasmon resonance. (A) Anti–BMP-7 mAb–coated agarose beads were incubated with rhBMP-7 and/or rxCTGF-flag. As controls, beads were coated with anti-HGF and incubated with rhHGF or coated with IgG1 and incubated with rhBMP-7 and/or rxCTGF-flag. Bound proteins were separated by SDS-PAGE and probed with anti-flag. Recombinant CTGF was run as control. (B) Increasing concentrations of rhCTGF were added to microtiter plates coated with rhBMP-7, rhHGF, or 1% BSA. Bound proteins were detected with an AP-conjugated antibody against rhCTGF. (C) Purified CTGF protein was run over BMP-7 sensor chips. Association and dissociation were monitored by a change in the resonance units.
The binding affinity of CTGF for BMP-7 was determined by surfaceplasmon resonance analysis. CTGF displayed time-dependent associationwith immobilized BMP-7 followed by dissociation, which was dose-dependentbecause higher response was observed at higher CTGF concentrations(Figure 7C) and at higher BMP-7 density (data not shown). Thedata showed complex binding behavior in which multiple componentswere involved; therefore, a heterogeneous two-site binding modelwas required to describe accurately the binding behavior. Thisresulted in Kd values describing a high- and low-affinity componentof 14 ± 6 and 316 ± 190 nM for the interactionbetween CTGF and BMP-7 [kon1 = 2.4 ± 0.9 x 105 M/s, koff1= 3.3 ± 0.8 x 10–3, kon2 = 1.3 ± 0.5 x 105,koff2 = 4.0 ± 1.8 x 10–2].
The results of this study reveal how diabetes-induced increaseof CTGF expression contributes to impairment of renal BMP signalingactivity and that this is associated with severity of structuraland functional hallmarks of diabetic nephropathy. To determinehow CTGF expression level might relate to reduced BMP-7 signalingin diabetic nephropathy, we compared pSmad1/5 and Id1 levelsin diabetic CTGF+/+ mice with those in diabetic CTGF+/–mice. The latter mice have lower CTGF expression in kidney,plasma, and urine as compared with diabetic CTGF+/+ mice, butthey have equal TGF-β1 expression (data not shown). Thus,diabetic CTGF+/– mice constitute a unique model to assessthe impact of CTGF level on downstream BMP-7 activity in diabeticnephropathy. BMP-7 mRNA expression in renal cortex was similarin diabetic CTGF+/+ and CTGF+/– mice, indicating thatCTGF level does not influence BMP-7 expression; however, pSmad1/5protein and Id1 mRNA levels were lower in diabetic CTGF+/+ mice,as compared with diabetic CTGF+/– mice. This indicatesthat CTGF might be an important determinant of the diabetes-inducedreduction in signaling activity of residual BMP-7. CTGF is knownalso to modulate BMP-4 and TGF-β1 directly.12 In addition,the level of BMP signaling activity in the kidney is subjectto the influence of other members of the TGF-β superfamily,BMP receptors, and of BMP modulators such as gremlin, noggin,kielin/chordin-like protein, or uterine sensitization-associatedgene 128; therefore, the relative contribution of BMP-7 inhibitionby CTGF remains to be established.
Albuminuria, thickening of the GBM, and decreased activity ofMMP all are hallmarks of human and experimental diabetic nephropathy.29–31Recently, it was shown that renal CTGF protein correlates withGBM thickness and prospective albuminuria in a nonhuman primatemodel of diabetes.32 We observed that these alterations andalso decreased MMP activity were attenuated or absent in diabeticCTGF+/– mice. This indicates that CTGF plays a pathogenicrole in at least these characteristic manifestations of diabeticnephropathy. Similarly, decrease of albuminuria in experimentalmodels of type 1 and type 2 diabetic nephropathy has been observedin mice treated with an anti-CTGF antibody and with a CTGF antisenseoligonucleotide, respectively.8,33 In the latter study, inhibitionof CTGF also resulted in reduction of serum creatinine and inhibitionof mesangial matrix expansion, which were not observed in ourdiabetic CTGF+/– mouse model (data not shown). A possibleexplanation for this discrepancy is that CTGF level derivedfrom a single functional allele might suffice to mediate diabetes-inducedincrease in serum creatinine and matrix accumulation. Treatmentwith antisense oligonucleotides might have lowered CTGF availabilityto levels below those in our diabetic CTGF+/– mice, resultingin more complete protection; however, we cannot directly compareactual reduction of CTGF levels in the different studies, andBMP signaling activity was not addressed after anti-CTGF antibodyor CTGF-antisense oligonucleotide treatment.
The strong correlation of Id1 with both albuminuria and MMPactivity suggests that CTGF-dependent suppression of this targetgene of BMP signaling might be involved directly in the pathogenesisof diabetic nephropathy. Accordingly, diabetic transgenic miceoverexpressing BMP-7 in podocytes and proximal tubular epithelialcells had higher Id1:PAI mRNA ratio and preserved MMP activityand developed less albuminuria, as compared with diabetic wild-typemice.23
As in overexpression of endogenous CTGF in diabetic mice, injectionof recombinant human CTGF into nondiabetic mice also resultedin decrease of pSmad1/5; however, this did not affect levelsof Id1 mRNA. Thus, transient two-fold increase of plasma CTGFin a nondiabetic environment seemed not to be sufficient forinhibition of Id1 mRNA, which might require higher or more sustainedelevation of CTGF or additional diabetes-induced changes.
As for the nature of CTGF–BMP-7 interaction in the kidney,exogenous CTGF protein as well as CTGF transfection inhibitedBMP-7 signaling activity in cultured mesangial cells, podocytes,proximal tubular epithelial cells, and renal interstitial fibroblasts.This was exemplified by inhibition of BMP-7–induced pSmad1/5,Id1, and BRE-luciferase activity. The inhibitory effect of CTGFwas robust, in the sense that it was observed with exogenouslyadded human and Xenopus CTGF, as well as with transfected mouseCTGF.
Direct physical interaction between CTGF and BMP-7 was evidencedby co-immunoprecipitation and in a solid-phase binding assay.Furthermore, surface plasmon resonance analysis demonstratedthat the interaction between CTGF and BMP-7 was complex andconsisted of a high- and a low-affinity component (Kd valuesof 14 and 316 nM, respectively). This not only may result frominherent biologic properties of both CTGF and BMP-7 but alsomay be due to partial blocking of CTGF-interactive sites duringthe immobilization of BMP-7. The high binding affinity of CTGFand BMP-7 was comparable to that described for CTGF and BMP-4,which had a Kd value of 5 nM.12
In conclusion, overexpression of CTGF inhibits BMP-7 signaltransduction in the diabetic kidney and contributes significantlyto altered gene transcription, as well as to reduced MMP activity,and to GBM thickening and albuminuria, which all are hallmarksof diabetic nephropathy.
Animal Experiments
Signaling activity of BMP-7 was studied in diabetic mice.3 Briefly,diabetes was induced in nine 12-wk-old female C57Bl/6J miceby a single intraperitoneal injection of 200 mg/kg streptozotocin(Sigma, St. Louis, MO) in sodium citrate buffer. Six controlanimals were administered an injection of vehicle only. Hyperglycemiawas determined 3 d after injection by measurement of blood glucoselevels. Slow-release insulin pellets (Linshin; Scarborough,Ontario, Canada) were implanted to stabilize the condition ofthe diabetic animals. Mice were killed 9 wk after injection.Renal cortex was harvested by dissecting small caps of the upperand the lower poles. Before homogenization, absence of medullawas checked in frozen sections of the cut surface.
Effects of CTGF level on BMP-7 signaling activity were studiedin diabetic CTGF+/+ and CTGF+/– mice. Outbred male BALBc/129SvCTGF+/– mice, in which exon 1 of one CTGF allele has beenreplaced by a neomycin resistance gene,34 were mated with femaleC57Bl/6J mice. From their first offspring, female CTGF+/–mice and female CTGF+/+ littermates were used for this study.Diabetes was induced in five 16-wk-old CTGF+/– mice andfour CTGF+/+ mice by injection of streptozotocin. Nine controlmice were administered an injection of vehicle only. After 9wk, unilateral nephrectomy was performed on all animals to aggravatethe diabetic nephropathy model. Mice were killed 17 wk afterinduction of diabetes. Albumin levels were determined by sandwichELISA using a goat–anti-mouse albumin antibody (BethylLaboratories, Montgomery, TX). Urinary creatinine excretionwas determined by enzymatic assays (J2L Elitech, Labarthe Inard,France).
Effects of CTGF on BMP-7 signaling activity in renal cortexof normoglycemic mice were studied in 24 BALBc mice, which wereadministered an intraperitoneal injection of rhCTGF (FibroGen,South San Francisco, CA) at a dosage of 20 µg/kg dilutedin 50 mM Tris-HCl buffer containing 800 mM NaCl. Mice were killed0, 5, 10, 15, 30, 60, 120, and 240 min after injection (threemice for each time point). Six control mice were administeredan injection of vehicle only and killed 0 and 240 min afterinjection.
All mice were housed in standard cages in a room with constanttemperature and a 12-h light-dark cycle. Mice were fed a standardpellet laboratory diet and had free access to water. The experimentswere performed with the approval of the Experimental AnimalEthics Committee of the University of Utrecht.
CTGF ELISA
CTGF levels in plasma, urine, and renal lysates were determinedby sandwich ELISA using two distinct specific antibodies (FibroGen),both directed against CTGF.3 The assay detects full-length andN-terminal fragments of CTGF. CTGF levels are expressed as pmol/L.
CTGF Immunohistochemistry
CTGF immunohistochemistry was performed as described previously.3Briefly, antigen retrieval was performed by predigestion withProtease XXIV (Sigma). Sections were incubated with a CTGF-specifichuman mAb (FibroGen), followed by incubation with rabbit–anti-humanIgG (Dako, Glostrup, Denmark) and goat–anti-rabbit Powervision-PO(Klinipath, Duiven, Netherlands). Bound antibody was visualizedwith Nova RED (Vector Laboratories, Burlingame, CA).
Electron Microscopy
Tissue samples were fixed in Karnovsky solution. Upon embedding,samples were rinsed with 0.1 M Na-cacodylate buffer, followedby fixation with 1% osmiumtetroxide, and dehydrated with acetoneand embedded in Epon. Ultrathin sections of 95 nm were cut andmounted on copper one-hole specimen support grids. Sectionswere stained with uranyl acetate and lead citrate to providecontrast. Ultrathin sections were photographed using a transmissionelectron microscope (JEM-1200 EX; JEOL, Peabody, MA). GBM thicknesswas measured in five random glomeruli per mouse at 10 perpendicularcross-sections of GBM per glomerulus at a magnification of x5000and analyzed by computer image analysis (ImageJ; National Institutesof Health, www.rsb.info.nih.gov/ij/).
In Situ Zymography
Glomerular MMP activity was visualized by in situ zymographyand confocal laser-scanning microscopy. Frozen tissue sectionswere incubated with DQTM gelatin from pig skin (Invitrogen,Carlsbad, CA), diluted 1:20 in 50 mM Tris-HCl buffer containing10 mol/L CaCl2, 0.05% Brij 35, and 5 mmol/L PMSF (pH 7.4). Slideswere incubated in a dark humidified chamber at 37°C for19 h. An MMP inhibitor, 1,10-phenanthroline monohydrate (2 µg/ml),was used to verify that the obtained gelatinase activity specificallyrepresented MMP activity. Nuclei were counterstained in redwith propidium iodide.
Gelatinase Activity in Tissue Lysates
From frozen kidneys, 10 sections of 20 µm were cut anddissolved in lysis buffer (50 mM Tris, 150 mM NaCl, and 1% TritonX-100 [pH 7.4]). Gelatinase activity was measured with an EnzChekGelatinase assay kit (Invitrogen). Collagenase was used as standard,and specificity of the gelatinase activity was verified using1,10-phenanthroline monohydrate. Protein concentrations of thelysates were determined colorimetrically and used for normalizationof collagenase activity per lysate.
Cell Culture
The immortalized human proximal tubular epithelial cell lineHK-2,35 the SV40-transformed human renal fibroblast cell lineTK173 (gift of F. Strutz, Göttingen, Germany),36 rat mesangialcells,9 and conditionally immortalized mouse podocytes wereused to study the effects of CTGF on BMP-7 signaling activity.HK-2, TK-173, and rat mesangial cells were maintained in DMEMwith 10% FBS, penicillin, and streptomycin (Invitrogen). Podocytecultures were maintained at 33°C in RPMI with 10% FBS inthe presence of IFN- (R&D Systems, Abingdon, UK). For differentiation,podocytes were plated in collagen I–coated wells at 37°Cand cultured for another 2 wk in RPMI with 10% FBS in the absenceof IFN-. HK-2, rat mesangial cells, and podocytes were seededat a density of 2 x 105 cells per well in six-well plates. Cellswere serum-starved for 24 h, followed by stimulation with andwithout 50 ng/ml rhBMP-7 (R&D Systems) in the presence of0, 1, 50, or 100 ng/ml rhCTGF (FibroGen) or rxCTGF from Xenopuslaevis.12 Cells were harvested after 1 h for Western blot analysisof pSmad1/5, and after 2 h for quantitative PCR.
For transfection experiments, TK173 cells were seeded at 1 x105 cells per well in six-well plates. After overnight culture,cells were washed and transfected with the BRE-luciferase reporterconstruct (provided by P. ten Dijke, Leiden, Netherlands). Cellswere co-transfected with pCAGGS-CTGF or pCAGGS-lacZ. The plasmidpCAGGS-mCTGF was constructed by insertion of mouse CTGF cDNAinto an EcoRI cloning site of the pCAGGS expression vector (providedby J. Miyazaki, Osaka, Japan).37 Transfection was carried outusing Lipofectamine 2000 (Invitrogen) with 2.0 µg of reporterconstruct, 2.0 µg of pCAGGS plasmid, and 40 ng of pRL-TK-Renilla(Promega, Madison, WI) as a control to normalize transfectionefficiency. After 24 h, cells were washed, serum-starved for12 h, and exposed to 10 or 50 ng/ml rhBMP-7 for 24 h. Luciferaseactivity was quantified using the Dual-Luciferase Reporter 1000Assay System (Promega).
Quantitative PCR
Total RNA was extracted from 30 mg of frozen renal cortex orfrom cultured cells using RNeasy columns (Qiagen, Venlo, Netherlands).After cDNA synthesis, expression of BMP-7, Id1, and CTGF mRNAwas assessed by quantitative real-time PCR using TaqMan GeneExpression Assays with predesigned probe and primers (AppliedBiosystems, Foster City, CA). TATA-box binding protein and β-actinwere used as internal reference.
Western Blot Analysis
Cells or sections of renal cortex were homogenized in lysisbuffer (20 mM Tris at pH 7.4, 150 mM NaCl, 1% Triton X-100,0.1% SDS, 10% glycerol, 1 mM EDTA, 1 mM EGTA, 0.5% sodium deoxycholate,50 mM NaF, and 2 mM Na3VO4) containing 5% Protease InhibitorCocktail (Sigma). Protein quantity was determined by BCA proteinassay kit (Pierce, Rockford, IL). Samples were run on 8% SDS-PAGEand transferred onto polyvinylidene difluoride membranes. Afterblocking, membranes were incubated with polyclonal antibodyspecifically directed against pSmad1/5 or total Smad5 (CellSignaling Technology, Beverly, MA) overnight, washed, and incubatedwith horseradish peroxidase–conjugated secondary antibody.For detection, membranes were incubated with SuperSignal WestDura Chemiluminescent Substrate (Pierce). Actin antibody (Sigma)was used on the same blot for loading control. PhosphorylatedSmad1/5 and total Smad5 staining were performed on the sameblots (with in between stripping) to control for possible regulationof total Smad5 expression.
Co-immunoprecipitation
Protein A/G PLUS-Agarose beads (Santa Cruz Biotechnology, SantaCruz, CA) were coated with monoclonal anti–BMP-7 (R&DSystems) and blocked with 1% BSA. Beads were subsequently preincubatedwith rhBMP-7 (R&D Systems) at 4°C. As controls, beadscoated with monoclonal anti-HGF and preincubated with rhHGF(both R&D Systems) or beads coated with IgG1 and preincubatedwith rhBMP-7 were used. The next day, beads were washed andrxCTGF-flag was added for 4 h at 37°C. Beads were washedand resuspended in 20 µl of PBS, and bound proteins wereeluted and denatured in SDS sample buffer and separated underreducing conditions by SDS-PAGE. Membranes were blocked andincubated with monoclonal anti-flag (Sigma). After incubationwith horseradish peroxidase–conjugated rat–anti-mouseantibody, detection was performed as described already.
Solid-Phase Binding Assay
Microtiter plates were coated overnight at 4°C with 0, 25,50, and 100 ng/ml rhBMP-7 or with 150 ng/ml rhHGF. Wells wererinsed and blocked with 1% BSA for 2 h. After washing, a rangeof 0 to 500 ng/ml of rhCTGF (FibroGen) was added, followed byalkaline phosphatase–conjugated mAb against human CTGF(FibroGen). After incubation for 2.5 h at 37°C, plates werewashed and substrate solution containing p-nitrophenyl phosphatewas added. Absorbance was read at 405 nm.
Surface Plasmon Resonance Analysis
Real-time binding experiments were performed on the Biacore2000 (GE Healthcare, Uppsala, Sweden). Carrier-free recombinantBMP-7 (R&D Systems) was immobilized on the CM5 sensor-chipsurface at 66 and 122 fmol/mm2. One control flow channel wasroutinely activated and blocked in the absence of protein. Associationof rhCTGF (5 to 300 nM) was assessed in triplicate in 10 mMHEPES (pH 7.4), 150 mM NaCl, 3 mM EDTA, and 0.005% surfactantP20 for 2 min, at a flow rate of 20 µl/min at 25°C.Dissociation was allowed for 2 min in the same buffer flow.Sensor chips were regenerated using several pulses of 20 mMHEPES (pH 7.4) and 1 M NaCl at a flow rate of 20 µl/min.Data were corrected for both refractive index changes, and associationand dissociation rate constants were determined by nonlinearregression analysis using the BIAevaluation Software 3.1 (GEHealthcare).
Statistical Analysis
Data are presented as means ± SD. Differences betweengroups were analyzed by t test or ANOVA with Bonferroni correctionfor multiple comparisons. Correlations were assessed by linearregression. For all comparisons, P < 0.05 was consideredto be significant (two-tailed).
NO., Z.L., and L.X. are employees of FibroGen, supplier of rhCTGFand anti-CTGF antibodies; R.G. has received research supportgrants and consultancy fees from FibroGen.
Acknowledgments
This work was supported by the Netherlands Organization forScientific Research (Mozaïek grant 017.003.037) and bythe Dutch Kidney Foundation (C05.2144).
We thank C. Vink, R. Verheul, A. Rietdijk, J.W. Leeuwis, andN. Veldhuijzen for technical assistance; J. Miyazaki for providingthe pCAGGS plasmid; F. Strutz for providing the TK173 cell line;and P. ten Dijke for providing the BRE-luciferase construct.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
P.R.'s current affilation is Department of Biochemistry, RadboudUniversity Nijmegen Medical Center, Nijmegan, Netherlands; F.A.v.N.'scurrent affiliation is Department of Physiology, CARIM, MaastrichtUniversity, Maastricht, Netherlands.
Ito Y, Aten J, Bende RJ, Oemar BS, Rabelink TJ, Weening JJ, Goldschmeding R: Expression of connective tissue growth factor in human renal fibrosis.
Kidney Int 53
: 853
–861, 1998[CrossRef][Medline]
Riser BL, deNichilo M, Cortes P, Baker C, Grondin JM, Yee J, Narins RG: Regulation of connective tissue growth factor activity in cultured rat mesangial cells and its expression in experimental diabetic glomerulosclerosis.
J Am Soc Nephrol 11
: 25
–38, 2000[Abstract/Free Full Text]
Roestenberg P, van Nieuwenhoven FA, Joles JA, Trischberger C, Martens PP, Oliver N, Aten J, Hoppener JW, Goldschmeding R: Temporal expression profile and distribution pattern indicate a role of connective tissue growth factor (CTGF/CCN-2) in diabetic nephropathy in mice.
Am J Physiol Renal Physiol 290
: F1344
–F1354, 2006[Abstract/Free Full Text]
Roestenberg P, van Nieuwenhoven FA, Wieten L, Boer P, Diekman T, Tiller AM, Wiersinga WM, Oliver N, Usinger W, Weitz S, Schlingemann RO, Goldschmeding R: Connective tissue growth factor is increased in plasma of type 1 diabetic patients with nephropathy.
Diabetes Care 27
: 1164
–1170, 2004[Abstract/Free Full Text]
Riser BL, Cortes P, deNichilo M, Deshmukh PV, Chahal PS, Mohammed AK, Yee J, Kahkonen D: Urinary CCN2 (CTGF) as a possible predictor of diabetic nephropathy: Preliminary report.
Kidney Int 64
: 451
–458, 2003[CrossRef][Medline]
Nguyen TQ, Tarnow L, Andersen S, Hovind P, Parving HH, Goldschmeding R, van Nieuwenhoven FA: Urinary connective tissue growth factor excretion correlates with clinical markers of renal disease in a large population of type 1 diabetic patients with diabetic nephropathy.
Diabetes Care 29
: 83
–88, 2006[Abstract/Free Full Text]
Gilbert RE, Akdeniz A, Weitz S, Usinger WR, Molineaux C, Jones SE, Langham RG, Jerums G: Urinary connective tissue growth factor excretion in patients with type 1 diabetes and nephropathy.
Diabetes Care 26
: 2632
–2636, 2003[Abstract/Free Full Text]
Guha M, Xu ZG, Tung D, Lanting L, Natarajan R: Specific down-regulation of connective tissue growth factor attenuates progression of nephropathy in mouse models of type 1 and type 2 diabetes.
FASEB J 21
: 3355
–3368, 2007[Abstract/Free Full Text]
Blom IE, van Dijk AJ, Wieten L, Duran K, Ito Y, Kleij L, deNichilo M, Rabelink TJ, Weening JJ, Aten J, Goldschmeding R: In vitro evidence for differential involvement of CTGF, TGFbeta, and PDGF-BB in mesangial response to injury.
Nephrol Dial Transplant 16
: 1139
–1148, 2001[Abstract/Free Full Text]
Abdel-Wahab N, Weston BS, Roberts T, Mason RM: Connective tissue growth factor and regulation of the mesangial cell cycle: Role in cellular hypertrophy.
J Am Soc Nephrol 13
: 2437
–2445, 2002[Abstract/Free Full Text]
Burns WC, Twigg SM, Forbes JM, Pete J, Tikellis C, Thallas-Bonke V, Thomas MC, Cooper ME, Kantharidis P: Connective tissue growth factor plays an important role in advanced glycation end product-induced tubular epithelial-to-mesenchymal transition: Implications for diabetic renal disease.
J Am Soc Nephrol 17
: 2484
–2494, 2006[Abstract/Free Full Text]
Abreu JG, Ketpura NI, Reversade B, De Robertis EM: Connective-tissue growth factor (CTGF) modulates cell signalling by BMP and TGF-beta.
Nat Cell Biol 4
: 599
–604, 2002[Medline]
Lam S, van der Geest RN, Verhagen NA, van Nieuwenhoven FA, Blom IE, Aten J, Goldschmeding R, Daha MR, van Kooten C: Connective tissue growth factor and IGF-I are produced by human renal fibroblasts and cooperate in the induction of collagen production by high glucose.
Diabetes 52
: 2975
–2983, 2003[Abstract/Free Full Text]
Raatikainen-Ahokas A, Hytonen M, Tenhunen A, Sainio K, Sariola H: BMP-4 affects the differentiation of metanephric mesenchyme and reveals an early anterior-posterior axis of the embryonic kidney.
Dev Dyn 217
: 146
–158, 2000[CrossRef][Medline]
Dudley AT, Lyons KM, Robertson EJ: A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye.
Genes Dev 9
: 2795
–2807, 1995[Abstract/Free Full Text]
Godin RE, Robertson EJ, Dudley AT: Role of BMP family members during kidney development.
Int J Dev Biol 43
: 405
–411, 1999[Medline]
Vukicevic S, Basic V, Rogic D, Basic N, Shih MS, Shepard A, Jin D, Dattatreyamurty B, Jones W, Dorai H, Ryan S, Griffiths D, Maliakal J, Jelic M, Pastorcic M, Stavljenic A, Sampath TK: Osteogenic protein-1 (bone morphogenetic protein-7) reduces severity of injury after ischemic acute renal failure in rat.
J Clin Invest 102
: 202
–214, 1998[Medline]
Wang SN, Lapage J, Hirschberg R: Loss of tubular bone morphogenetic protein-7 in diabetic nephropathy.
J Am Soc Nephrol 12
: 2392
–2399, 2001[Abstract/Free Full Text]
De Petris L, Hruska KA, Chiechio S, Liapis H: Bone morphogenetic protein-7 delays podocyte injury due to high glucose.
Nephrol Dial Transplant 22
: 3442
–3450, 2007[Abstract/Free Full Text]
Mitu GM, Wang S, Hirschberg R: BMP7 is a podocyte survival factor and rescues podocytes from diabetic injury.
Am J Physiol Renal Physiol 293
: F1641
–F1648, 2007[Abstract/Free Full Text]
Wang S, Chen Q, Simon TC, Strebeck F, Chaudhary L, Morrissey J, Liapis H, Klahr S, Hruska KA: Bone morphogenic protein-7 (BMP-7), a novel therapy for diabetic nephropathy.
Kidney Int 63
: 2037
–2049, 2003[CrossRef][Medline]
Wang S, de CM, Kopp J, Mitu G, Lapage J, Hirschberg R: Renal bone morphogenetic protein-7 protects against diabetic nephropathy.
J Am Soc Nephrol 17
: 2504
–2512, 2006[Abstract/Free Full Text]
Sugimoto H, Grahovac G, Zeisberg M, Kalluri R: Renal fibrosis and glomerulosclerosis in a new mouse model of diabetic nephropathy and its regression by bone morphogenic protein-7 and advanced glycation end product inhibitors.
Diabetes 56
: 1825
–1833, 2007[CrossRef][Medline]
Neilson EG: Setting a trap for tissue fibrosis.
Nat Med 11
: 373
–374, 2005[CrossRef][Medline]
Wahab NA, Mason RM: A critical look at growth factors and epithelial-to-mesenchymal transition in the adult kidney: Interrelationships between growth factors that regulate EMT in the adult kidney.
Nephron Exp Nephrol 104
: e129
–e134, 2006[CrossRef][Medline]
Korchynskyi O, Ten DP: Identification and functional characterization of distinct critically important bone morphogenetic protein-specific response elements in the Id1 promoter.
J Biol Chem 277
: 4883
–4891, 2002[Abstract/Free Full Text]
Zeisberg M: Bone morphogenic protein-7 and the kidney: Current concepts and open questions.
Nephrol Dial Transplant 21
: 568
–573, 2006[Free Full Text]
Wolf G, Ziyadeh FN: Cellular and molecular mechanisms of proteinuria in diabetic nephropathy.
Nephron Physiol 106
: 26
–31, 2007[CrossRef]
MCLennan SV, Kelly DJ, Cox AJ, Cao Z, Lyons JG, Yue DK, Gilbert RE: Decreased matrix degradation in diabetic nephropathy: Effects of ACE inhibition on the expression and activities of matrix metalloproteinases.
Diabetologia 45
: 268
–275, 2002[CrossRef][Medline]
Del PD, Anglani F, Forino M, Ceol M, Fioretto P, Nosadini R, Baggio B, Gambaro G: Down-regulation of glomerular matrix metalloproteinase-2 gene in human NIDDM.
Diabetologia 40
: 1449
–1454, 1997[CrossRef][Medline]
Thomson SE, McLennan SV, Kirwan PD, Heffernan SJ, Hennessy A, Yue DK, Twigg SM: Renal connective tissue growth factor correlates with glomerular basement membrane thickness and prospective albuminuria in a non-human primate model of diabetes: Possible predictive marker for incipient diabetic nephropathy.
J Diabetes Complications April 15, 2008 [epub ahead of print]
Flyvbjerg A, Khatir D, Jensen LJN, Lomongsod E, Liu DY, Rasch R, Usinger WR: Long-term renal effects of a neutralizing connective tissue growth factor (CTGF)-antibody in obese type 2 diabetic mice [Abstract].
J Am Soc Nephrol 15
: 261A
, 2004[CrossRef]
Ivkovic S, Yoon BS, Popoff SN, Safadi FF, Libuda DE, Stephenson RC, Daluiski A, Lyons KM: Connective tissue growth factor coordinates chondrogenesis and angiogenesis during skeletal development.
Development 130
: 2779
–2791, 2003[Abstract/Free Full Text]
Ryan MJ, Johnson G, Kirk J, Fuerstenberg SM, Zager RA, Torok-Storb B: HK-2: An immortalized proximal tubule epithelial cell line from normal adult human kidney.
Kidney Int 45
: 48
–57, 1994[Medline]
Muller GA, Frank J, Rodemann HP, Engler-Blum G: Human renal fibroblast cell lines (tFKIF and tNKF) are new tools to investigate pathophysiologic mechanisms of renal interstitial fibrosis.
Exp Nephrol 3
: 127
–133, 1995[Medline]
Niwa H, Yamamura K, Miyazaki J: Efficient selection for high-expression transfectants with a novel eukaryotic vector.
Gene 108
: 193
–199, 1991[CrossRef][Medline]
Related Article
This Month's Highlights
J. Am. Soc. Nephrol. 2008 19: A10.
[Full Text][PDF]
This article has been cited by other articles:
J. Loureiro, M. Schilte, A. Aguilera, P. Albar-Vizcaino, M. Ramirez-Huesca, M. L. Perez-Lozano, G. Gonzalez-Mateo, L. S. Aroeira, R. Selgas, L. Mendoza, et al. BMP-7 blocks mesenchymal conversion of mesothelial cells and prevents peritoneal damage induced by dialysis fluid exposure
Nephrol. Dial. Transplant.,
January 18, 2010;
(2010)
gfp618v2.
[Abstract][Full Text][PDF]
T. Turk, J. W. Leeuwis, J. Gray, S. V. Torti, K. M. Lyons, T. Q. Nguyen, and R. Goldschmeding BMP Signaling and Podocyte Markers Are Decreased in Human Diabetic Nephropathy in Association With CTGF Overexpression
J. Histochem. Cytochem.,
July 1, 2009;
57(7):
623 - 631.
[Abstract][Full Text][PDF]
S. A. Roxburgh, J. J. Kattla, S. P. Curran, Y. M. O'Meara, C. A. Pollock, R. Goldschmeding, C. Godson, F. Martin, and D. P. Brazil Allelic Depletion of grem1 Attenuates Diabetic Kidney Disease
Diabetes,
July 1, 2009;
58(7):
1641 - 1650.
[Abstract][Full Text][PDF]
C. A. Droppelmann, J. Gutierrez, C. Vial, and E. Brandan Matrix Metalloproteinase-2-deficient Fibroblasts Exhibit an Alteration in the Fibrotic Response to Connective Tissue Growth Factor/CCN2 because of an Increase in the Levels of Endogenous Fibronectin
J. Biol. Chem.,
May 15, 2009;
284(20):
13551 - 13561.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
(R&D Systems, Abingdon, UK). For differentiation, podocytes were plated in collagen I–coated wells at 37°C and cultured for another 2 wk in RPMI with 10% FBS in the absence of IFN-
