Biological responses to PDGF-BB versus PDGF-DDin human mesangial cellsCRC van Roeyen1, T Ostendorf1, B Denecke2, D Bokemeyer3, I Behrmann4, F Strutz5, HS Lichenstein6,WJ LaRochelle6, CE Pena6, A Chaudhuri6 and J Floege1

1Division of Nephrology, RWTH Aachen University, Aachen, Germany; 2Interdisciplinary center of clinical research (IZKF ‘BIOMAT.’),RWTH Aachen University, Aachen, Germany; 3Augusta-Hospital Bochum, Bochum, Germany; 4Institute of Biology and Physiology,University of Luxemburg, Luxemburg; 5Division of Nephrology, University of Gottingen, Gottingen, Germany and 6CuraGen Corporation,Branford, Connecticut, USA

Platelet-derived growth factor (PDGF)-BB and PDGF-DD

mediate mesangial cell proliferation in vitro and in vivo. While

PDGF-BB is a ligand for the PDGF a- and b-receptor chains,

PDGF-DD binds more selectively to the b-chain, suggesting

potential differences in the biological activities. Signal

transduction and regulation of gene expression induced by

PDGF-BB and -DD were compared in primary human

mesangial cells (HMCs), which expressed PDGF a- and

b-receptor subunits. The growth factor concentrations used

were chosen based on their equipotency in inducing HMCs

proliferation and binding to the bb-receptor. Both growth

factors, albeit at different concentrations induced

phosphorylation and activation of extracellular signal-

regulated kinase 1 (ERK1) and ERK2. In addition, PDGFs led to

the phosphorylation and activation of signal transducers and

activators of transcription 1 (STAT1) and STAT3. HMCs

proliferation induced by either PDGF-BB or -DD could be

blocked by signal transduction inhibitors of the

mitogen-activated protein kinase-, Janus kinase (JAK)/STAT-,

or phosphatidyl-inositol 3-kinase pathways. Using a gene

chip array and subsequent verification by real-time reverse

transcriptase (RT)-polymerase chain reaction, we found that

in HMC genes for matrix metalloproteinase 13 (MMP-13) and

MMP-14 and, to a low extent, cytochrome B5 and cathepsin L

were exclusively regulated by PDGF-BB, whereas no exclusive

gene regulation was detected by PDGF-DD. However, at the

protein level, both MMP-13 and -14 were equally induced by

PDGF-BB and -DD. PDGF-BB and -DD effect similar biological

responses in HMCs albeit at different potencies. Rare

apparently differential gene regulation did not result in

different protein expression, suggesting that in HMCs both

PDGFs exert their biological activity almost exclusively via the

PDGF b-receptor.

Kidney International (2006) 69, 1393–1402. doi:10.1038/sj.ki.5000332;

published online 22 March 2006

KEYWORDS: human mesangial cell; platelet-derived growth factor; signal-

ling; gene regulation; matrix metalloproteinase

Platlet-derived growth factor (PDGF)-A and -B are secretedas homo- or heterodimers, bind to and induce PDGFreceptor dimers composed of a- and/or b-chains. Two novelPDGF isoforms, PDGF-C and -D, are released as homo-dimers only.1–3 Whereas PDGF-BB is a ligand for all receptortypes, PDGF-DD binds to the PDGF bb- but not the aa-receptor. PDGF-DD is produced as latent factor. Proteolyticcleavage of a CUB-domain from each chain is then requiredfor activation.4 In addition, PDGF-DD differs from PDGF-BB in that its C-terminus lacks a basic amino-acid sequencethat mediates binding to the extracellular matrix.4

The PDGF receptor possesses tyrosine kinase activity andis autophosphorylated upon ligand binding.5 The receptorthen interacts with several other cytoplasmic proteinscontaining Src homology 2 (SH2) domains, includingphospholipase C (PLC-g), ras guanine 50-triphosphatase-activating protein, phosphatidyl-inositol 3-kinase (PI3K),members of the pp60src family of protein tyrosine kinase,tyrosine phosphatase SHP-2, and members of the signaltransducers and activators of transcription (STAT) family.5

Second messengers include inositol-(1,4,5)-triphosphate anddiacylglycerol, intracellular calcium release, and proteinkinase C (PKC) activation.5

Mesangial cells have repeatedly been shown to expressboth PDGF a- and b-receptor subunits.6–11 PDGF-AA usuallyelicits lower-level and fewer biological effects in mesangialcells in comparison to PDGF-BB.6,11 Nevertheless, mesangialcells are clearly responsive to both growth factor isoforms invitro, demonstrating the functional relevance of both thePDGF a- and b-subunits in these cells. Given that PDGF-DD,like the PDGF B-chain, can bind to the bb- and ab-receptor,their biological activities in the kidney may be similar, albeitnot identical in view of their differential binding to the PDGFaa-receptor. Neither PDGF-B nor -D are normally expressedin the glomerular mesangium, but both are markedly up-regulated during human and experimental nephritis.7,12,13 Inview of the above we asked, whether the effects of PDGF-BB

http://www.kidney-international.org o r i g i n a l a r t i c l e

& 2006 International Society of Nephrology

Received 6 July 2005; revised 21 October 2005; accepted 14 December

2005; published online 22 March 2006

Correspondence: J Floege, Division of Nephrology, University Hospital

Aachen, Pauwelsstr. 30, Aachen D-52057, Germany.

E-mail: juergen.floege@rwth-aachen.de

Kidney International (2006) 69, 1393–1402 1393

and -DD on primary human mesangial cells (HMCs) differin terms of signal transduction and regulation of geneexpression.

RESULTSPDGF-BB and -DD induce proliferation in primary HMCs

In pilot studies, a fourfold higher PDGF-DD concentrationshowed similar biological activity as PDGF-BB (Table 1a),which correlates with a threefold difference in receptoraffinity between the two isoforms.1 Based on these pilotstudies, we initially used 50 ng/ml PDGF-BB and 200 ng/mlPDGF-DD in the experiments (in a later set of experimentsusing different batches of both recombinant PDGF-BB andPDGF-DD, concentrations were changed to 10 and 100 ng/ml, respectively, to again yield equipotency in the mitogenesisassay (Table 1b)).

Using the above concentrations, we first verified that50 ng/ml PDGF-BB and 200 ng/ml PDGF-DD indeed exertedthe same mitogenic effect when studied over a time course ofup to 48 h (Figure 1). Co-stimulation of the cells with theseconcentrations of PDGF-BB and -DD failed to exert anyadditive or synergistic effects (Figure 1).

PDGF-BB and -DD downregulate the expression of the PDGFreceptor b-chain mRNA in HMCs

In unstimulated HMCs mRNA, expression of the a- andb-receptor subunits was detected by real-time polymerasechain reaction (PCR) (Figure 2). If stimulated with PDGF-BBand/or –DD, a significant suppression of the PDGF receptorb-chain mRNA expression level occurred, whereas theexpression of the a-chain remained unchanged (Figure 2).

PDGF-BB and -DD induce similar signaling pathways

Activation by PDGF-BB results in the induction of threemajor signaling pathways, the mitogen-activated proteinkinase (MAPK)-, Janus kinase (JAK)/STAT-, and PI3Kpathways.11,14,15 Until now nothing is known about thesignaling mechanisms of PDGF-DD, which binds moreselectively to the PDGF-receptor b-chain than PDGF-BB.

As demonstrated in Figure 3a, both Western blot analysisfor phoshorylated (i.e. activated) extracellular signal-regu-lated kinase 1/2 (ERK1/2) and a functional ERK1/2 kinase

Table 1 | Cell proliferation measured via BrdU incorporation after stimulation of HMCs with PDGF-BB and/or -DD for 24 h usingtwo different preparations of PDGF-BB and -DD

Growth factor concentration 0 10 ng/ml 50 ng/ml 100 ng/ml 200 ng/ml 400 ng/ml

(a) First batchPDGF-BB 77743 170733 2167129 n.d. n.d.PDGF-DD 77743 54712 118720 181768 2307120

(b) Second batchPDGF-BB 98710 399720 326730 320780 n.d. n.d.PDGF-DD 98710 n.d. 260720 417750 425760 405750

Data (induction of proliferation as % of unstimulated HMCs) are means7s.d. of four independent experiments (n.d.: not determined).BrdU, 5-bromo-20-deoxyuridine; HMCs, human mesangial cells; PDGF, platelet-derived growth factor.

PDGF-BB (50 ng/ml)PDGF-DD (200 ng/ml)PDGF-BB + DD (50+200 ng/ml)

Ind

uct

ion

of

pro

lifer

atio

n in

% o

fu

nst

imu

late

d H

MC

Time of incubation 4 h 8 h 12 h 16 h 24 h 36 h 48 h

100

200

300

400

500

600

700Unstimulated

* *

**

*

*

** *

*P<0.05 vs unstimulated cells

Figure 1 | PDGF-BB and/or -DD induce mesangial cell prolifera-tion. Cell proliferation was measured via BrdU incorporation afterstimulation of HMCs with PDGF-BB and/or -DD. Data are means7s.d.of four independent experiments. In the case of unstimulated controlvalues, means of the absolute BrdU incorporations per time pointwere calculated and set as 100%. Over the course of the 48 h of theexperiment, these means of the absolute BrdU incorporations did notchange significantly (data not shown).

Rel

ativ

e m

RN

A e

xpre

ssio

n

��-chain �-chainPDGF-receptor

PDGF-BB (10 ng/ml)PDGF-DD (100 ng/ml)

PDGF-BB + DD (10+100 ng/ml)

Unstimulated

*P<0.05 vs unstimulated

0.5

1.0

1.5

*

*

*

Figure 2 | Expression of the PDGF receptor chains a and b.Real-time PCR-based expression data normalized to GAPDH mRNA.The gene expression data from unstimulated and PDGF-BB and/or–DD-stimulated HMCs are shown. Data are means7s.d. of eightindependent experiments. *Indicates Po0.05 versus unstimulatedcontrol.

1394 Kidney International (2006) 69, 1393–1402

o r i g i n a l a r t i c l e CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells

assay demonstrated that PDGF-DD, like PDGF-BB, activatesERK1/2, albeit with lesser efficiency at the concentrationsused. By electrophoretic mobility shift assay, we also foundactivated STAT1 and 3 in response to stimulation with eitherPDGF, again with a somewhat weaker response to PDGF-DD(Figure 3b).

The involvement of the MAPK, JAK/STAT, and PI3Ksignaling pathways in mediating the mitogenic properties ofPDGF-BB and -DD was further evaluated in proliferationassays with primary HMC by using specific inhibitors(Figure 4). The PDGF-BB induced proliferation can beinhibited by a pre-incubation with U0126 for 10 min.15

Although pre-incubation times of 30 or 60 min have beendescribed for AG490 and LY294002,16–20 in our hands bothinhibitors prevented PDGF-induced cell proliferation after apre-incubation time of 10 min. Therefore, HMCs were pre-incubated with all inhibitors used for 10 min beforestimulation with PDGF-BB and/or -DD. Blockade of any ofthe three pathways blunted the proliferation induced byeither PDGF-BB or PDGF-DD at all inhibitor concentrationsemployed, indicating that both PDGF isoforms use the samepathways to induce cell proliferation.

Identification of genes differentially regulated by PDGF-BBand -DD in primary HMCs by cDNA microarray analysis

The gene expression profile of growth-arrested HMCs, as wellas PDGF-BB and/or –DD-stimulated HMCs, was analyzedafter 24 h stimulation by Affymetrix gene chip analysis. Bothgrowth factors regulated the mRNA expression of severalgenes involved in cell proliferation like GAP43, RGS-2, IGF-2,LTBP-1, LTBP-2, CDC20, TGFb-1, and PRC1 but no differ-ences between the two PGDFs evolved. Table 2 summarizesthe genes that were classified as being up- or downregulatedby only one PDGF isoform or in the co-stimulation. Eighteengenes and five unspecified transcripts, as expressed sequencetags and hypothetical proteins, were regulated by PDGF-DDbut not by PDGF-BB, whereas 18 genes and 17 unspecifiedtranscripts were regulated by PDGF-BB only. Seventy-eightgenes and 57 unspecified transcripts were specificallyregulated in the co-stimulation of PDGF-BB and -DD. Basedon their potential pathophysiologic relevance, we thenselected 17 genes (five suppressed by PDGF-DD, seveninduced by PDGF-DD, and five induced by PDGF-BB) forsubsequent analysis using real-time reverse transcriptase PCR(RT-PCR) and complementary DNA (cDNA) isolated fromPDGF-BB and/or -DD-stimulated HMCs (n¼ 8) (Figure 5).

Genes suppressed by PDGF-DD. A significant suppressioncould be confirmed for renin–angiotensin system guanylnucleotide releasing protein 1 (RASGRP1), nuclear transcrip-tion factor Y beta (NFYB), transforming growth factor-b-2

PDGF-BB

PDGF-DD

PDGF-BB/-DDSTAT3/3

STAT1/3STAT1/1

STAT3/3

STAT1/3STAT1/1

STAT3/3STAT1/3STAT1/1

0 2 5 10 20 30 minTime of incubation

P-ELK

P-ERK1/2

B D B/D B D B/D B D B/D B D B/D B D B/D B D B/D

0 2 5 10 20 30 minTime of incubation

B D B/D B D B/D B D B/D B D B/D B D B/D B D B/D

0 2 5 10 20 30 minTime of incubation

a

b

Figure 3 | PDGF-BB and/or -DD activate MAPKs ERK1/2 andSTAT1/3. (a) Activation of MAPKs ERK1 and ERK2 was measured afterstimulation of HMCs with PDGF-BB and/or -DD (n ¼ 2) in a Westernblot specific for phosphorylated ERK1/2 and in a kinase assay with theELK protein as a substrate. Both PDGF-BB and -DD phosphorylate andactivate ERK1/2. (b) The activation of STAT 1/3 was measured in anelectro-mobility shift assay with nuclear extracts of HMCs stimulatedwith PDGF-BB and/or -DD (n¼ 3). Both factors activate STAT 1/1homodimers, STAT 1/3 heterodimers, and STAT3/3 homodimers;however, PDGF-BB appeared to be more efficient in this assay.

200

400

600

800

1 000

1 200

Ind

uct

ion

of

pro

lifer

atio

n in

% o

f th

eu

nst

imu

late

d H

MC

10 % FCS –

200 ng/ml PDGF-DD –U0126 (�M) –AG490 (�M) –

50 ng/ml PDGF-BB –

LY294002 (�M) –

+

–––

–

–

–

–––

+

–

–

–5

–

+

–

–

–25

–

+

–

–

––50

+

–

–

––

100

+

–

–

–––

+

5

–

–––

+

20

–

–––

–

5

––

–

25

––

–

–50

–

–

–100

–

–

––

5

–– – – – – – –

––

+ + + + + + +

20

–

5

––

–

25

––

–

–50

–

–

–100

–

–

––

5

–– – – – – –– – – – – –

––

20

Figure 4 | PDGF-BB and -DD use the same intracellular pathwaysfor mediation of the proliferation signal. HMCs proliferation wasmeasured by BrdU incorporation after stimulation with PDGF-BB or-DD. ERK1/2 activation was inhibited by 5 and 25mM U0126, the JAK/STAT pathway was blocked by 50 and 100mM AG490, and signalingthrough PI3K was blocked by 5 and 20 mM LY294002. Incubation withdifferent concentrations of inhibitors for the three major signalingpathways of PDGF-BB prevents the mitogenic effect of PDGF-BB and-DD, indicating that both activate the same pathways. Data aremeans7s.d. of five independent experiments.

Kidney International (2006) 69, 1393–1402 1395

CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells o r i g i n a l a r t i c l e

Table 2 | Identification of genes whose expression is changed significantly (Po0.001) by PDGF-BB and/or -DD

Genbank Acc. Gene

Transcripts suppressed by PDGF-DD onlyNM_005739,2 RAS guanyl releasing protein 1 (calcium and DAG-regulated) (RASGRP1)NM_006166,2 Nuclear transcription factor Y, beta (NFYB)M19154,1 Transforming growth factor-beta-2NM_003062,1 Slit (Drosophila) homolog 3 (SLIT3)M22865,1 Cytochrome b5M31159,1 Growth hormone-dependent insulin-like growth factor-binding protein

Transcripts suppressed by PDGF-BB onlyNM_002510,1 Glycoprotein (transmembrane) nmb (GPNMB)BC002700,1 Similar to keratin 7, clone MGC:3625

Transcripts suppressed by PDGF-BB+PDGF-DD onlyNM_005410,1 Selenoprotein P, plasma, 1 (SEPP1)NM_002667,1 Phospholamban (PLN)NM_005019,1 Phosphodiesterase 1A, calmodulin-dependent (PDE1A)L35594,1 Human autotaxinNM_016201,1 Leman coiled-coil protein (LCCP)NM_005204,1 Mitogen-activated protein kinase kinase kinase 8 (MAP3K8)M18767,1 Human complement subcomponent C1s, alpha- and beta-chainsNM_003494,1 Dysferlin, limb girdle muscular dystrophy 2B (DYSF)NM_002222,1 Inositol 1,4,5-triphosphate receptor, type 1 (ITPR1)NM_001078,1 Vascular cell adhesion molecule 1 (VCAM1)NM_003882,1 WNT1 inducible signaling pathway protein 1 (WISP1)NM_004010,1 Dystrophin (muscular dystrophy, Duchenne and Becker types)NM_000961,1 Prostaglandin I2 (prostacyclin) synthase (PTGIS)NM_005512,1 Glycoprotein A repetitions predominant (GARP)L03203,1 Human peripheral myelin protein 22 (GAS3)M81635,1 Erythrocyte membrane protein mRNA,NM_000104,2 Cytochrome P450, subfamily I (dioxin-inducible), polypeptide 1 (CYP1B1)NM_030786,1 Intermediate filament protein syncoilin (SYNCOILIN)AF078077,1 Growth arrest and DNA-damage-inducible protein GADD45betaNM_002053,1 Guanylate-binding protein 1, interferon-inducible, 67kDa (GBP1)NM_014583,1 LIM and cysteine-rich domains 1 (LMCD1)NM_016277,1 RAB23, member RAS oncogene family (RAB23)NM_001343,1 Disabled (Drosophila) homolog 2 (DAB2)NM_016270,1 Kruppel-like factor (LOC51713)L38969,1 Thrombospondin 3 (THBS3)AF138303,1 Decorin DNM_016147,1 Protein phosphatase methylesterase-1 (PME-1)NM_000861,2 Histamine receptor H1 (HRH1)M60485,1 Human fibroblast growth factor receptorNM_015987,1 Heme-binding protein (HEBP)NM_007216,1 Alpha integrin-binding protein 63 (KIAA1017)NM_001387,1 Dihydropyrimidinase-like 3 (DPYSL3)AF188298,1 Disabled 2 p93 (DAB2)NM_003330,1 Thioredoxin reductase 1 (TXNRD1)NM_000297,1 Polycystic kidney disease 2 (autosomal dominant) (PKD2)NM_018450,1 Uncharacterized bone marrow protein BM029 (BM029)NM_021630,1 PDZ-LIM protein mystique (LOC59346)

Transcripts induced by PDGF-DD onlyNM_001211,2 Budding uninhibited by benzimidazoles 1 (yeast homolog), beta (BUB1B)BC005247,1 Isopentenyl-diphosphate delta isomerase, clone MGC:12281AF003934,1 Prostate differentiation factorNM_005542,1 Insulin induced gene 1 (INSIG1)NM_001673,1 Asparagine synthetase (ASNS)NM_000269,1 Non-metastatic cells 1, protein (NM23A) expressed in (NME1)BC003573,1 Farnesyl-diphosphate farnesyltransferase 1, clone MGC:2200BC001312,1 Protein disulfide isomerase-related protein, clone MGC:5517NM_003887,1 Development and differentiation enhancing factor 2 (DDEF2)U18197,1 ATP: citrate lyaseNM_017567,1 N-acetylglucosamine kinase (HSA242910)NM_001360,1 7-Dehydrocholesterol reductase (DHCR7)

Transcripts induced by PDGF-BB onlyNM_002427,2 Matrix metalloproteinase 13 (collagenase 3) (MMP-13)

1396 Kidney International (2006) 69, 1393–1402

o r i g i n a l a r t i c l e CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells

(TGF-b2), homo sapiens slit homolog 3 (Drosophila) (SLIT3)and cytochrome B5 (Figure 5b). However, in contrast to themicroarray analysis, the gene suppression was not specific forPDGF-DD, but also detected following PDGF-BB and PDGF-BB plus -DD stimulation.

Genes induced by PDGF-DD. None of the seven genesinduced by PDGF-DD only in the microarray experimentcould be verified by RT-PCR (Figure 5c).

Genes induced by PDGF-BB. By RT-PCR, three genes,namely cathepsin L (CTSL), matrix metalloproteinase 13

Genbank Acc. Gene

AF191653,1 Diphosphoinositol polyphosphate phosphohydrolase type 2 beta (NUDT4)NM_004398,2 DEADH (Asp-Glu-Ala-AspHis) box polypeptide 10 (RNA helicase) (DDX10)NM_004995,2 Matrix metalloproteinase 14 (membrane-inserted) (MMP-14)BC003143,1 Dual specificity phosphatase 6, clone MGC:3789NM_021127,1 Phorbol-12-myristate-13-acetate-induced protein 1 (PMAIP1)NM_002712,1 Protein phosphatase 1, regulatory subunit 7 (PPP1R7)AF348514,1 Fetal thymus prothymosin alphaNM_005013,1 nucleobindin 2 (NUCB2)AF306765,1 Junctate mRNAAF318616,1 Synaptojanin 2 mRNANM_005475,1 Lymphocyte adaptor protein (LNK)NM_001912,1 Cathepsin L (CTSL)J03202,1 Laminin B2 chainAB014486,1 mRNA for RA70U67195,1 Tissue inhibitor of metalloproteinase-3

Transcripts induced by PDGF-BB + -DD onlyNM_012323,1 v-maf musculoaponeurotic fibrosarcoma (avian) oncogene family F (MAFF)NM_004267,1 Carbohydrate (chondroitin 6keratan) sulfotransferase 2 (CHST2)NM_006823,1 Protein kinase (cAMP-dependent, catalytic) inhibitor alpha (PKIA)NM_014705,1 KIAA0716 gene product (KIAA0716)NM_001394,2 Dual specificity phosphatase 4 (DUSP4)AY029180,1 Soluble urokinase plasminogen activator receptor precursor (SUPAR)NM_000216,1 Kallmann syndrome 1 sequence (KAL1)NM_001423,1 Epithelial membrane protein 1 (EMP1)NM_004434,1 Echinoderm microtubule-associated protein-like (EMAPL)NM_006579,1 Emopamil-binding protein (sterol isomerase) (EBP)NM_003824,1 Fas (TNFRSF6)-associated via death domain (FADD)NM_005308,1 G protein-coupled receptor kinase 5 (GPRK5)AF031463,1 Phosducin-like protein mRNANM_016020,1 CGI-75 ProteinBC000425,1 Protein disulfide isomerase related protein, clone MGC:8346M73554,1 bcl-1 mRNAAF260566,1 Hepatocyte growth factor-regulated tyrosine kinase substrate, isoform 2 (HRS)AF006011,1 Dishevelled 1 (DVL1) mRNANM_001283,1 Adaptor-related protein complex 1, sigma 1 subunit (AP1S1)U84246,1 Lysosomal sialidaseAF074979,1 Regulator of G protein signaling-Z (RGSZ1) mRNANM_016031,1 Elongation of very long chain fatty acids-like 1 (ELOVL1)NM_016072,1 CGI-141 protein (LOC51026), mRNANM_002004,1 Farnesyl diphosphate synthaseM23612,1 Human GTPase-activating protein (GAP)AF167438,1 Androgen-regulated short-chain dehydrogenasereductase 1 (ARSDR1)BC000478,1 Heat shock 70kDa protein 9B (mortalin-2)L13386,1 (clone 47) Miller–Dieker lissencephaly protein (LIS1)NM_002708,1 Protein phosphatase 1, catalytic subunit, alpha isoform (PPP1CA)AB029156,1 HRP-3NM_016185,1 Hematological and neurological expressed 1 (HN1)NM_002658,1 Plasminogen activator, urokinase (PLAU)NM_005742,1 Protein disulfide isomerase-related protein (P5)AF064238,3 Smoothelin large isoform L2 (SMTN)NM_001069,1 Tubulin, beta polypeptide (TUBB)NM_013417,1 Isoleucine-tRNA synthetase (IARS)BC000498,1 Glutamic-oxaloacetic transaminase 1, soluble (aspartate aminotransferase 1)NM_004181,1 Ubiquitin carboxyl-terminal esterase L1 (ubiquitin thiolesterase) (UCHL1)AF015040,1 NUMB protein (NUMB) mRNAAF257318,1 SH3-containing protein SH3GLB1NM_004447,1 Epidermal growth factor receptor pathway substrate 8 (EPS8)

Table 2 | Continued

Kidney International (2006) 69, 1393–1402 1397

CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells o r i g i n a l a r t i c l e

(MMP-13), and MMP-14, exhibited a significant inductionfollowing stimulation with PDGF-BB, PDGF-BB plus -DD,but not with PDGF-DD alone (Figure 5a). However, in theparticular case of MMP-13, a trend towards increased mRNAexpression after stimulation with PDGF-DD alone wasnotable (Figure 5a).

Thus, all genes that were modulated by PDGF-DD werealso found to be regulated by PDGF-BB in the verificationRT-PCR studies. Vice versa, at least in four cases geneinduction appeared to be specific for PDGF-BB.

In further studies, we focussed on the two genes thatexhibited the most pronounced regulation by PDGF-BB andthe most pronounced differences between PDGF-BB and-DD, that is, MMP-13 and -14.

Stimulation with PDGF-BB and -DD results in the secretion ofpro-MMP-13

MMP-13 showed the highest mRNA induction afterstimulation with PDGF-BB and appeared to be relativelyspecifically induced by PDGF-BB only (Figure 5a). Therefore,we assessed secretion of the MMP-13 protein into the culturesupernatant of PDGF-BB and/or –DD-activated HMCs. Asshown in Figure 6, PDGF-BB resulted in a time- and dose-dependent increase of secreted pro-MMP-13. Unlike themRNA results, PDGF-DD was an equally potent inducer ofMMP-13 release (Figure 6). No synergistic or additive effectsbetween PDGF-BB and -DD were observed at the concentra-tions assayed.

Stimulation of HMCs with PDGF-BB and -DD inducesexpression and activation of the MMP-14 protein

MMP-14 mRNA expression also appeared to be specificallyregulated by PDGF-BB in both the micro-array and the

RT-PCR confirmation experiments. As MMP-14 is notsecreted into the culture supernatant but remains cellassociated, we studied its protein expression by Western blotanalysis using a monoclonal MMP-14 antibody, whichrecognizes inactive as well as active MMP-14. The datashowed that both PDGF-BB and -DD increased levels ofactivated MMP-14 protein in lysates of HMCs (Figure 7). Inaddition, the incubation of HMC with both PDGF-BB and -DD resulted in increased levels of total MMP-14 protein (i.e.inactive plus active MMP-14; Figure 7).

DISCUSSION

In the present study, we focused on potential differences inthe biological functions of PDGF-BB and PDGF-DD inhuman mesangial cells. Both PDGFs are potent mediators ofmesangial cell proliferation in vitro and in vivo. Additionalfunctions of PDGF-BB may be expected by the binding ofPDGF-BB to the PDGF aa-receptor, whereas PDGF-DD canonly activate the PDGF bb- and possibly also the ab-receptor.At the aa-receptor, PDGF-DD failed to compete with thebinding of labeled PDGF-AA.1,2,21,22 We therefore firstverified that our primary HMCs expressed both PDGFreceptor subunits. In these experiments, it was notable thatboth PDGF-BB and -DD induced a negative feedbackregulation of the b-receptor subunit mRNA, as had beendescribed by Oster et al.23 for PDGF-BB, whereas expressionof the a-subunit remained stable after ligand addition.

The most important decision in a study that compares thebiological function of two growth factors is the selection ofagonist concentrations. Bergsten et al.1 demonstrated thatprotease activated PDGF-DD, that is, PDGF-DD devoid ofthe CUB-domain, had a threefold lower binding capacity tothe b-chain of the PDGF receptor than PDGF-BB. Hence, if

Table 3 | Primers for real-time RT-PCR used for the verification of the HMCs cDNA microarray data

Gene Forward primer Reverse primer

ATP:citrate lyase TCAGGAGGGCTTACGGGTG TCTGTGCCAAAGACATGGATGASNS GCATCCGTGGAAATGGTTAAA TCCACATTGTCATAGAGGGCGBUB1B TCAACAGAAGGCTGAACCACTAGA GCCAACAGAGTTTGCCGAGACathepsin L CGGCTTTGTGGACATCCCT CAGAAATGGGCCCCACAGTCytochrome B5 AACAAGCTGGAGGTGACGCT TTGGACATTTCCCTGGCATCDDEF2 TGTCCAATGCTATGGTCCTGC ATAGAGCGCTTTCACCCGCDHCR7 ACATGCTCGGCTCTCGGAC AGGTATAGAGCTGGGCGGCTINSIG1 AGCCCCTACCCCCAACACCT TAGGACCACCCCAACCGAGLaminin B2 CAGACCAACTCCTAGCCCGA GTATCCCGTCCCTTCTTTGCMMP-13 GCAGAGCGCTACCTGAGATCA TCATGGAGCTTGCTGCATTCMMP-14 GGTGAGTCAGGGTTCCCCA TCAGGAACAGAAGGCCGCNFYB AATTCAGAGAGGCTATGAAAGGAGAA CCTCTGTAAGCTCTTCACTTAGTCCANME23A CTACGTTGACCTGAAGGAACCG CGGCCCTGAGTGCATGTATRASGRP1 GGCATCTCCCAGTGGGTACA GAGCCACCTGGATGAACTTGASLIT GCGCGATTTGGAGATCCTTA TTCGGATCTTCGGCATGTGTGF-b2 GGGTGGAAATGGATACACGAA TCCATAAATACGGGCATGCTCTIMP-3 TCCAATTTCGGTTACCCTGG CTGCAGTAGCCGCCCTTCTPDGF-receptor a-chain CCCTTGGTGGCACCCC CGGTACCCACTCTTGATCTTATTGTPDGF-receptor b-chain ATGCCTTACCACATCCGCTC ACATTGCAGGTGTAGGTCCCCGAPDH AGCCACATCGCTCAGACACC GCGCCCAATACGACCAAA

ASNS, asparagine synthetase; BUB1B, budding uninhibited by benzimidazoles 1 (yeast homolog), beta; DDEF2, development and differentiation enhancing factor 2; DHCR7,7-Dehydrocholesterol reductase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; INSIG1, insulin induced gene 1; MMP-13, matrix metalloproteinase 13 (collagenase 3);MMP-14, matrix metalloproteinase 14 (membrane-inserted); NFYB, nuclear transcription factor Y, beta; NME23A, non-metastatic cells 1, protein (NM23A) expressed in (NME1);PDGF, platelet-derived growth factor; RASGRP1, renin–angiotensin system guanyl releasing protein 1; RT-PCR, reverse transcriptase-polymerase chain reaction; SLIT3, Slit(Drosophila) homolog 3; TGF-b2, transforming growth factor-beta-2; TIMP, tissue inhibitor of metalloproteinase-3.

1398 Kidney International (2006) 69, 1393–1402

o r i g i n a l a r t i c l e CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells

based on weight, equal PDGF-BB and -DD concentrationswill result in different activation of the PDGF receptors,pretending different biological functions of the two PDGFisoforms. Since the mitogenic effect of PDGFs on mesangialcells is almost exclusively mediated by the PDGF b-receptorsubunit,11 we based all experiments of the present study onPDGF-BB and -DD concentrations, which induced similarproliferation in HMCs. The validity of this approach wassubstantiated by the virtually identical time course of HMCsgrowth after stimulation with either PDGF-BB or -DD(Figure 1).

Using the optimized concentrations, we next investigatedthe signaling of PDGF-DD in comparison to PDGF-BB.PDGF-BB is able to activate MAPK, JAK/STAT, PI3K, and

phospholipase C pathways.11,15,24 Until now nothing wasknown about the signaling of PDGF-DD. We assessedactivation of ERK1/2, that is, the MAPK pathway, and theactivation of STAT 1/3. Both signaling pathways were foundto be activated by PDGF-DD similarly to PDGF-BB, albeit atlower potency. Proliferation assays using inhibitors for theMAPK-, JAK/STAT-, or PI3K-pathway confirmed theirinvolvement. These findings are similar to a study comparingthe signaling of PDGF-AA and -BB in rat metanephricmesenchymal cells,25 which showed that despite differentreceptor binding, PDGF-AA and -BB stimulated the samesignaling pathways.

Next, we looked for genes differentially regulated byPDGF-BB and -DD by gene chip analysis. Out of 17 genes,which appeared to be regulated by either PDGF-BB or -DDbut not both, only four (MMP-13, MMP-14, cathepsin L, andcytochrome B5) could be verified by real-time RT-PCR.Importantly, using RT-PCR, we were able to show that allgenes regulated by PDGF-DD were also regulated by -BB andno case of an exclusive PDGF-DD-regulated gene wasobserved in this study. This observation is in agreement withthe current model of differential PDGF isoform binding tothe receptors.1,2,21,22 Of the four genes that appeared to beregulated by PDGF-BB only, cathepsin L and cytochrome B5were not pursued further in view of their weak mRNAinduction following stimulation with PDGF-BB.

The gene that showed the strongest induction of mRNAexpression in response to PDGF-BB was MMP-13. Interest-ingly, MMP-13 is a known target gene for STAT1,26

confirming the different activation of the STAT1/3 signallingpathway by PDGF-BB and -DD. This finding is supported by

0.51.01.52.02.53.03.5

Laminin B2 Cathepsin L TIMP-3 MMP14

PDGF-BB (10 ng/ml)

PDGF-DD (100 ng/ml) PDGF-BB+DD (10+100 ng/ml)

10

20

30

40

50

MMP13

Rel

ativ

e m

RN

Aex

pre

ssio

nR

elat

ive

mR

NA

exp

ress

ion

Rel

ativ

e m

RN

Aex

pre

ssio

n

Rel

ativ

e m

RN

Aex

pre

ssio

n

Unstimulated

RASGRP1 NFYB TGF-�2 CytochromeB5

0.51.01.52.02.53.03.5

BUB1B INSIG1 ASNS NME23A DDEF2 DHCR7ATP:citrate lyase

**

*

*

**

*

* *

**

**

**

**

* **

0.4

0.8

1.2

1.6

SLIT 3

*P<0.05 vs unstimulated

Verification of mesangial genesinduced by PDGF-B in the array

Verification of mesangial genessuppressed by PDGF-D in the array

Verification of mesangial genesinduced by PDGF-D in the array

a

b

c

Figure 5 | Verification of the array results. Real-time PCR-basedexpression data normalized to the internal standard GAPDH mRNA.For each gene expression data from unstimulated and PDGF-BB and/or –DD-stimulated HMCs are shown. Abbreviations are: insulin-induced gene 1 (INSIG1), asparagine synthetase (ASNS), non-metastatic cells protein 1 (NME23A), development and differentiationenhancing factor 2 (DDEF2), ATP: citrate lyase, 7-dehydrocholesterolreductase (DHCR7) and BUB1 budding uninhibited by benzimidazoles1 homolog beta (BUB1B). Data are means7s.d. of eight independentexperiments. *Indicates Po0.05 versus unstimulated control.

pg/m

l pro

-MM

P13

in th

e su

pern

atan

t

DMEM 1 10 100 10 100 500 10+100

PDGF-B(ng/ml)

PDGF-D(ng/ml)

PDGF-B + PDGF-D(ng/ml)

24 h stimulation48 h stimulation

*P<0.05 vs DMEM

200

400

600

800

1000

*

**

*

*

*

*

**

**

*

**

Figure 6 | MMP-13 in the supernatant of HMCs stimulated withPDGF-BB and/or -DD. HMCs were stimulated with PDGF-BB and/or-DD for 24 or 48 h. The concentrations of pro-MMP-13 in thesupernatants were measured by ELISA. Data are means7s.d. of threeindependent experiments. *Indicates Po0.05 versus non-stimulatedcells (Dulbecco’s modified Eagle’s medium, DMEM).

Kidney International (2006) 69, 1393–1402 1399

CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells o r i g i n a l a r t i c l e

the regulation of TIMP-3 (a target gene of STAT126) byPDGF-BB (Table 2). In addition, the mRNA expression of theSTAT3 target gene junB27 was specifically induced by PDGF-BB in at least one of the three array experiments. MMP-13(collagenase III) is expressed in pathological situations suchas inflammation or cancer and degrades type I, II, III, and IVcollagen, gelatin, aggrecan, perlecan, the large isoform oftenascin C, fibrillin, and fibrinogen.28 However, when wemeasured pro-MMP-13 protein secretion after stimulation ofHMC with PDGF-BB and/or -DD, both growth factors wereable to induce its secretion. Once again, no absolute PDGF-BB specificity of the biological response could be established.The divergent results of the MMP-13 mRNA analyses andprotein data may result from the failure of a low-gradeinduction of MMP-13 mRNA by PDGF-DD (Figure 5) toreach statistical significance. Alternatively, PDGF-DD mayaffect MMP-13 at the post-transcriptional level, since adiscrepancy between the induction of the MMP-13 mRNAlevels and expression of the protein has also been noted byothers.29

The second gene that appeared to be specifically regulatedby PDGF-BB in vitro was MMP-14. MMP-14 (membranetype 1-MMP) family belongs to the MT-MMP family anddegrades fibrillar collagens, that is, type I, II, and IIIcollagens, as well as gelatin, proteoglycan, fibronectin,vitronectin, type 1 and 5 laminin, and aggrecan.30 Addition-ally, MMP-14 is able to activate pro-MMP2, pro-MMP9, andpro-MMP-13.31 Until now high glucose levels were the onlyknown regulator of MMP-14 mRNA expression in mesangialcells, indicating an involvement in matrix accumulation indiabetic nephropathy.32,33 Interestingly, MMP-14 can mod-ulate PDGF-B binding to the PDGF receptor b and therebyincrease signal transduction in vascular smooth musclecells.34 MMP-14-deficient cells show defective responsestowards PDGF-B with regard to cell proliferation, chemo-taxis, and activation of ERK1/2 and Akt, supporting thenotion that MMP-14 is a mediator of PDGF-BB effects.34

Taken together with our finding of a PDGF-B inducibleMMP-14 expression, this would indicate an positive auto-regulatory loop. Also of relevance to the present data isthe observation that MMP-14 mRNA is overexpressed inthe expanded mesangium of the anti-Thy 1.1 nephritismodel in rats.35 Given that this model is PDGF-B mediated,35

our data provide a mechanistic explanation for thein vivo observations of Hayashi et al.35 In contrast to theapparently PDGF-BB-specific induction of MMP-14mRNA in HMCs, at the protein level, MMP-14, like MMP-13, was equally induced by both PDGF isoforms. This againraises the issue of potential post-translational regulation,since it is known that the expression and activation ofMMP-14 is regulated by proteolytic processing, shedding,endocytosis, degradation, recycling, and formation ofhomodimers.36

In summary, stimulation of HMC with PDGF-DDresulted in qualitatively very similar biological responses incomparison to PDGF-BB. Whether the two PDGF isoformsresult in a quantitatively different activation of individualsignaling pathways remains difficult to prove given theproblem of establishing ‘bioequivalent’ concentrations of theagonists, which contain important structural differences suchas the presence of a latency-inducing CUB domain in PDGF-DD but not -BB. Nevertheless, our gene array data alsopointed to rare instances of genes, where mRNA expressionappeared to be regulated by PDGF-BB only. However,at the protein level, the two most apparently PDGF-BB-specific genes, that is, MMP-13 and -14, were similarlyinduced by PDGF-BB and -DD. It is therefore conceivablethat higher doses of PDGF-DD might have led to a similargene induction as PDGF-BB in the array studies, sincebiological equipotency as defined for proliferation maynot be the same for other biological effects. Taken together,our data suggest that in HMCs PDGF-BB and -DD inducevery similar biological responses, implying that bothPDGF isoforms almost exclusively mediate their actionin the glomerular mesangium via the PDGF b-receptorsubunit.

2

4

6

8

10

12

1

2

3

4

5

6

7

Unstimulated PDGF-B PDGF-D PDGF-B + -D

Unstimulated PDGF-B PDGF-D PDGF-B + -D

Exp

ress

ion

of

acti

veM

MP

-14

pro

tein

in H

MC

rela

tive

to

un

stim

ula

ted

cells

Exp

ress

ion

of

tota

lM

MP

-14

pro

tein

in H

MC

rela

tive

to

un

stim

ula

ted

cells

PDGF-BB (10 ng/ml)PDGF-DD (100 ng/ml) PDGF-BB+DD (10+100 ng/ml)Unstimulated

*

*P<0.05 vs unstimulated

< 50 kDa

< 50 kDa

< 75 kDaActive MMP-14

Inactive MMP-14

Actin

B D NCB+D MU

b

a

Figure 7 | Expression and activation of MMP-14 in of HMCsstimulated with PDGF-BB and/or -DD. (a) HMCs were stimulatedwith PDGF-BB and/or -DD for 24 h. The expression of MMP-14 proteinin the cells was detected by Western blot using a monoclonalantibody, which detects active and inactive MMP-14.U¼ unstimulated cells, B¼ stimulated with PDGF-BB, D¼ stimulatedwith PDGF-DD, BþD¼ stimulated with PDGF-BBþ PDGF-DD,M¼ full range rainbow protein molecular weight marker, NC¼ ne-gative control. (b) Densitometric analysis of the Western blot. Theexpression of active and total MMP-14 was normalized for theexpression of actin. Data are means7s.d. of three independentexperiments. *Indicates Po0.05 versus unstimulated cells.

1400 Kidney International (2006) 69, 1393–1402

o r i g i n a l a r t i c l e CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells

MATERIALS AND METHODSReagents and HMCsRecombinant human PDGF-DD lacking the CUB domain wasproduced as described.2 PDGF-BB was purchased from Sigma-Aldrich (Deisenhofen, Germany). Primary HMCs were obtainedfrom Cambrex (Verviers, Belgium) and were maintained asdescribed.37

Proliferation assaysTo compare the effects of PDGF-BB and -DD, HMCs were seeded in96-well plates (Nunc, Wiesbaden, Germany), grown to subcon-fluence, and growth-arrested for 24 h in modified Czapek Dox broth(MCDB) medium (Sigma-Aldrich). After 24 h, PDGF-DD (200 ng/ml), PDGF-BB (50 ng/ml), or PDGF-DD (200 ng/ml) plus PDGF-BB (50 ng/ml) were added and the cells were incubated for 4 to 48 h.DNA synthesis was determined during the last 4 h by a 5-bromo-20-deoxyuridine (BrdU) incorporation assay according to the manu-facturer’s instructions (Roche, Mannheim, Germany). In otherexperiments, pharmacological inhibitors were added to the medium10 min prior to stimulating HMCs with PDGF-DD (200 ng/ml) orPDGF-BB (50 ng/ml) for 24 h: U0126 (inhibits the ERK-activatingkinases (mitogen-activated protein kinase/ERK kinase 1 andmitogen-activated protein kinase/ERK kinase 2) at a final concen-tration of 5 and 25 mM, AG490 (inhibits the Janus-activated kinase 2)at a final concentration of 50 and 100mM, and LY294002 (inhibitsthe PI3K) at a final concentration of 5 and 20 mM.

Real-time quantitative RT-PCRTotal RNA was isolated from HMCs using the Invisorb Spin Cell-RNA Mini Kit (Invitek, Berlin, Germany). RNA purity determina-tion, cDNA synthesis, and RT-PCR were performed as described.12

The primer sequences are listed in Table 3. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internalstandard.

Western blot analysisSubconfluent HMCs were growth arrested for 24 h in MCDB media.After 24 h, PDGF-DD (200 ng/ml), PDGF-BB (50 ng/ml), or PDGF-DD (200 ng/ml) together with PDGF-BB (50 ng/ml) in RPMI 1640with 100 mg/ml of streptomycin and 100 U/ml penicillin were addedand the cells were incubated for 2–30 min. Subsequently, the cellswere washed in ice-cold phosphate-buffered saline and lysed in700ml Triton X-100 lysis buffer as described.15 After 5 min, cells werescraped with a cell lifter and centrifuged at 41C for 15 min at10 000 g. The soluble cell lysates were separated and blotted asdescribed15 and probed with polyclonal antibodies against thecarboxy-terminal peptide of p42 ERK.15 The primary antibodies(diluted 1:1000) were detected using horseradish peroxidase-conjugated protein A visualized by the Amersham ECL system afterintensive washing of the membrane.

The Western blot analysis for MMP-14 was performed asdescribed above after stimulation of HMCs with PDGF-BB and/or-DD for 24 h. Blots were probed with a monoclonal MT1-MMP(160–173) antibody (Calbiochem, Schwalbach, Germany). Fornormalization, membranes were stripped (2� 30 min incubationin 200 mM glycine, 0.1% SDS, 1% Tween 20, pH 2.2) and reprobedwith a monoclonal anti-actin antibody (clone C-2, Santa Cruz).

Nonradioactive ERK activity assaySoluble cell lysates (500 mg) were blotted and probed as described.15

EMSAFollowing HMCs stimulation (see above), the supernatant wasremoved, nuclear extracts were isolated and probed to a double-stranded mutated SIE oligonucleotide from the c-fos promotor(m67SIE: 50-GAT CCG GGA GGG ATT TAC GGG AAA TGC TG-30)as described.38

Affymetrix gene chip arrayHMCs were cultured in 75 cm2 flasks until subconfluency, growtharrested for 24 h, and subsequently stimulated with PDGF-BB(10 ng/ml), PDGF-DD (100 ng/ml), PDGF-DD (100 ng/ml) togetherwith PDGF-BB (10 ng/ml), or with RPMI 1640 alone in the presenceof 100 mg/ml of streptomycine and 100 U/ml penicillin for 24 h. Thedifference in PDGF concentrations in these experiments ascompared to the experiments described above resulted from theusage of a new PDGF preparation for the array studies. As above, thegrowth factor concentrations used for the chip assay lead toequipotent induction of HMCs proliferation (Table 1). Aftercultivation for 24 h, the cells were lysed and RNA was isolatedusing the RNeasy Mini Kit (Qiagen, Hilden, Germany). RNA samplecollection and generation of biotinylated complementary RNAprobes was carried out essentially as described in the AffymetrixGeneChip Expression Analysis Technical Manual (Affymetrix, SantaClara, CA, USA). Total RNA (10 mg) was reverse transcribed intodouble-stranded cDNA using HPLC-purified T7-(dt) 24 primers(MWG, Ebersberg, Germany) and the Superscript choice cDNAsynthesis system (Invitrogen Corp., Carlsbad, CA, USA). Subse-quently, the purified double-stranded cDNA was used as template tosynthesize biotinylated complementary RNA probes using theBioArray High Yield RNA Transcription Labeling Kit (EnzoDiagnostics, Enzo Life Science Inc., Farmingdale, NY, USA). Eachsample was hybridized to an Affymetrix human Genome U133Amicroarray for 16 h at 451C. Expression values of each probe setwere determined and samples stimulated with PDGF-BB, PDGF-DD, or PDGF-BB plus -DD were compared with unstimulatedsamples using the Affymetrix Microarray Suite 5.0 software. Theintensities across multiple arrays were normalized to a targetintensity of 2500 using global normalization scaling. Three separateexperiments were performed under identical conditions. Genes,whose expression levels were changed significantly (Po0.001) in atleast two of three experiments, were considered to be regulated bythe stimulus.

ELISA for pro-MMP-13Cells (20 000–40 000) were seeded in each well of a 12-well plate,growth arrested for 24 h, and treated with 1, 10, or 100 ng/ml PDGF-BB, 10, 100, or 500 ng/ml PDGF-DD, or 10 ng/ml PDGF-BB plus100 ng/ml PDGF-DD as described. The conditioned media wereharvested at 24 and 48 h and centrifuged at 10000 r.p.m. for 4 min.The supernatants were assayed using to the R&D SystemsQuantikine ELISA kit for the detection of pro-MMP-13.

Statistical analysisAll values are expressed as means7s.d. Statistical significance(defined as Po0.05) was evaluated using analysis of variance andBonferroni t-tests.

ACKNOWLEDGMENTSWe are grateful for expert technical assistance by G Dietzel, G Minartz,A Cosler, and R Lanzmich. This work was supported by the DeutscheForschungsgemeinschaft (SFB 542: project C7) to JF and TO.

Kidney International (2006) 69, 1393–1402 1401

CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells o r i g i n a l a r t i c l e

REFERENCES1. Bergsten E, Uutela M, Li X et al. PDGF-D is a specific, protease-activated

ligand for the PDGF beta-receptor. Nat Cell Biol 2001; 3: 512–516.2. LaRochelle WJ, Jeffers M, McDonald WF et al. PDGF-D, a new protease-

activated growth factor. Nat Cell Biol 2001; 3: 517–521.3. Li X, Ponten A, Aase K et al. PDGF-C is a new protease-activated ligand for

the PDGF alpha-receptor. Nat Cell Biol 2000; 2: 302–309.4. Li X, Eriksson U. Novel PDGF family members: PDGF-C and PDGF-D.

Cytokine Growth Factor Rev 2003; 14: 91–98.5. Heldin CH, Westermark B. Mechanism of action and in vivo role of

platelet-derived growth factor. Physiol Rev 1999; 79: 1283–1316.6. Floege J, Topley N, Hoppe J et al. Mitogenic effect of platelet-derived

growth factor in human glomerular mesangial cells: modulation and/orsuppression by inflammatory cytokines. Clin Exp Immunol 1991; 86:334–341.

7. Iida H, Seifert R, Alpers CE et al. Platelet-derived growth factor (PDGF) andPDGF receptor are induced in mesangial proliferative nephritis in the rat.Proc Natl Acad Sci USA 1991; 88: 6560–6564.

8. Alpers CE, Seifert RA, Hudkins KL et al. PDGF-receptor localizes tomesangial, parietal epithelial, and interstitial cells in human and primatekidneys. Kidney Int 1993; 43: 286–294.

9. Gesualdo L, Di Paolo S, Milani S et al. Expression of platelet-derivedgrowth factor receptors in normal and diseased human kidney. Animmunohistochemistry and in situ hybridization study. J Clin Invest 1994;94: 50–58.

10. Floege J, Hudkins KL, Seifert RA et al. Localization of PDGF alpha-receptorin the developing and mature human kidney. Kidney Int 1997; 51:1140–1150.

11. Choudhury GG, Grandaliano G, Jin DC et al. Activation of PLC and PI 3kinase by PDGF receptor alpha is not sufficient for mitogenesis andmigration in mesangial cells. Kidney Int 2000; 57: 908–917.

12. Ostendorf T, van Roeyen CR, Peterson JD et al. A fully human monoclonalantibody (CR002) identifies PDGF-D as a novel mediator ofmesangioproliferative glomerulonephritis. J Am Soc Nephrol 2003; 14:2237–2247.

13. Gesualdo L, Pinzani M, Floriano JJ et al. Platelet-derived growth factorexpression in mesangial proliferative glomerulonephritis. Lab Invest 1991;65: 160–167.

14. Wang Y, Simonson MS, Pouyssegur J, Dunn MJ. Endothelin rapidlystimulates mitogen-activated protein kinase activity in rat mesangialcells. Biochem J 1992; 287(Part 2): 589–594.

15. Bokemeyer D, Panek D, Kramer HJ et al. In vivo identification of themitogen-activated protein kinase cascade as a central pathogenicpathway in experimental mesangioproliferative glomerulonephritis. J AmSoc Nephrol 2002; 13: 1473–1480.

16. Otero M, Lago R, Lago F et al. Signalling pathway involved in nitric oxidesynthase type II activation in chondrocytes: synergistic effect of leptinwith interleukin-1. Arthritis Res Ther 2005; 7: R581–R591.

17. Zhang Y, Karas M, Zhao H et al. 14—3-3sigma mediation of cell cycleprogression is p53-independent in response to insulin-like growth factor-Ireceptor activation. J Biol Chem 2004; 279: 34353–34360.

18. Tajmir P, Ceddia RB, Li RK et al. Leptin increases cardiomyocytehyperplasia via extracellular signal-regulated kinase- andphosphatidylinositol 3-kinase-dependent signaling pathways.Endocrinology 2004; 145: 1550–1555.

19. Minami T, Abid MR, Zhang J et al. Thrombin stimulation of vascularadhesion molecule-1 in endothelial cells is mediated by protein kinase C(PKC)-delta-NF-kappa B and PKC-zeta-GATA signaling pathways. J BiolChem 2003; 278: 6976–6984.

20. Liu L, Paul A, MacKenzie CJ et al. Nuclear factor kappa B is involved inlipopolysaccharide-stimulated induction of interferon regulatory factor-1and GAS/GAF DNA-binding in human umbilical vein endothelial cells. Br JPharmacol 2001; 134: 1629–1638.

21. Tallquist M, Kazlauskas A. PDGF signaling in cells and mice. CytokineGrowth Factor Rev 2004; 15: 205–213.

22. Fredriksson L, Li H, Eriksson U. The PDGF family: four gene productsform five dimeric isoforms. Cytokine Growth Factor Rev 2004; 15:197–204.

23. Oster SK, Marhin WW, Asker C et al. Myc is an essential negative regulatorof platelet-derived growth factor beta receptor expression. Mol Cell Biol2000; 20: 6768–6778.

24. Choudhury GG, Marra F, Kiyomoto H, Abboud HE. PDGF stimulatestyrosine phosphorylation of JAK 1 protein tyrosine kinase in humanmesangial cells. Kidney Int 1996; 49: 19–25.

25. Ricono JM, Arar M, Choudhury GG, Abboud HE. Effect of platelet-derivedgrowth factor isoforms in rat metanephric mesenchymal cells. Am JPhysiol Renal Physiol 2002; 282: F211–F219.

26. Li WQ, Dehnade F, Zafarullah M. Oncostatin M-induced matrixmetalloproteinase and tissue inhibitor of metalloproteinase-3 genesexpression in chondrocytes requires Janus kinase/STAT signalingpathway. J Immunol 2001; 166: 3491–3498.

27. Kojima H, Nakajima K, Hirano T. IL-6-inducible complexes on an IL-6response element of the junB promoter contain Stat3 and 36 kDaCRE-like site binding protein(s). Oncogene 1996; 12: 547–554.

28. Leeman MF, Curran S, Murray GI. The structure, regulation, and functionof human matrix metalloproteinase-13. Crit Rev Biochem Mol Biol 2002;37: 149–166.

29. Yu Q, Cok SJ, Zeng C, Morrison AR. Translational repression of humanmatrix metalloproteinases-13 by an alternatively spliced form ofT-cell-restricted intracellular antigen-related protein (TIAR). J Biol Chem2003; 278: 1579–1584.

30. Ohuchi E, Imai K, Fujii Y et al. Membrane type 1 matrix metalloproteinasedigests interstitial collagens and other extracellular matrixmacromolecules. J Biol Chem 1997; 272: 2446–2451.

31. Toth M, Chvyrkova I, Bernardo MM et al. Pro-MMP-9 activation by theMT1-MMP/MMP-2 axis and MMP-3: role of TIMP-2 and plasmamembranes. Biochem Biophys Res Commun 2003; 308: 386–395.

32. McLennan SV, Martell SY, Yue DK. High glucose concentration inhibits theexpression of membrane type metalloproteinase by mesangial cells:possible role in mesangium accumulation. Diabetologia 2000; 43:642–648.

33. McLennan SV, Martell SK, Yue DK. Effects of mesangium glycation onmatrix metalloproteinase activities: possible role in diabetic nephropathy.Diabetes 2002; 51: 2612–2618.

34. Lehti K, Allen E, Birkedal-Hansen H et al. An MT1-MMP-PDGFreceptor-beta axis regulates mural cell investment of themicrovasculature. Genes Dev 2005; 19: 979–991.

35. Hayashi K, Osada S, Shofuda K et al. Enhanced expression of membranetype-1 matrix metalloproteinase in mesangial proliferativeglomerulonephritis. J Am Soc Nephrol 1998; 9: 2262–2271.

36. Osenkowski P, Toth M, Fridman R. Processing, shedding, and endocytosisof membrane type 1-matrix metalloproteinase (MT1-MMP). J Cell Physiol2004; 200: 2–10.

37. Floege J, Topley N, Wessel K et al. Monokines and platelet-derived growthfactor modulate prostanoid production in growth arrested, humanmesangial cells. Kidney Int 1990; 37: 859–869.

38. Kube D, Holtick U, Vockerodt M et al. STAT3 is constitutively activated inHodgkin cell lines. Blood 2001; 98: 762–770.

1402 Kidney International (2006) 69, 1393–1402

o r i g i n a l a r t i c l e CRC van Roeyen et al.: Effects of PDGF-BB and -DD on mesangial cells