* Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; and Department of Medicine, The First Affiliated Hospital, Nanjing Medical University, Nanjing, China
Correspondence: Dr. Youhua Liu, Department of Pathology, University of Pittsburgh, S-405 Biomedical Science Tower, 200 Lothrop Street, Pittsburgh, PA 15261. Phone: 412-648-8253; Fax: 412-648-1916; E-mail: liuy{at}upmc.edu
Received for publication January 30, 2007.
Accepted for publication May 2, 2007.
Hepatocyte growth factor (HGF) is a potent antifibrotic cytokinethat antagonizes the TGF-1/Smad signaling in diverse types ofkidney cells by different mechanisms. HGF is shown to induceSmad co-repressor Sloan-Kettering Institute proto-oncogene–relatednovel gene, non–Alu-containing (SnoN) expression in proximaltubular epithelial cells (HKC-8) but not in glomerular mesangialcells and interstitial fibroblasts. This study investigatedthe molecular mechanisms underlying the cell type–specificinduction of SnoN by HGF. Treatment of HKC-8 cells with actinomycinD completely abolished HGF-mediated SnoN induction, suggestingits dependence on gene transcription. Although HGF activatedseveral signal pathways in HKC-8 cells, blockade of extracellularsignal–regulated kinase-1 and -2 (Erk-1/2) activationbut not Akt and p38 mitogen-activated protein kinase abolishedSnoN induction. HGF rapidly activated both upstream and downstreamsignaling of Erk-1/2, which led to the activation of the cAMPresponse element–binding protein (CREB). In the promoterregion of human SnoN gene, two cAMP response elements were locatedin close proximity to Sp1 sites. Chromatin immunoprecipitationassay revealed that activated CREB and Sp1 bound to their cognatecis-acting elements in SnoN promoter in response to HGF stimulation.Ectopic expression of wild-type CREB promoted SnoN expression,whereas dominant negative mutant CREB abrogated SnoN inductionby HGF. Likewise, chemical blockade of Sp1 binding abolishedHGF-mediated SnoN induction. Furthermore, HGF selectively inducedCREB phosphorylation in HKC-8 cells but not in mesangial cellsand fibroblasts. In vivo, administration of HGF gene inducedrenal Erk-1/2 phosphorylation, CREB activation, and SnoN expressionin obstructive nephropathy. Collectively, these results suggestthat CREB activation, in concert with Sp1, constitutes a molecularswitch that confers the cell type–specific induction ofSnoN in response to HGF stimulation.
Hepatocyte growth factor (HGF) is a unique endogenous proteinthat possesses potent antifibrotic potential.1,2 Extensive studiesin the past several years have demonstrated that administrationof exogenous HGF or its gene effectively ameliorates kidneydysfunction and fibrotic lesions in a wide variety of animalmodels of chronic kidney diseases.3–9 Conversely, blockadeof HGF signaling with neutralizing antibody markedly exacerbatesthe kidney function and accelerates the progression of renalfibrosis in different chronic kidney disease models.10,11 Experimentalevidence indicates that HGF is able to antagonize directly theprofibrotic TGF-1/Smad signaling in diverse types of kidneycells by different mechanisms.1 In kidney tubular epithelialcells, HGF has been shown to block the epithelial-to-mesenchymaltransition induced by TGF-1.12,13 This action of HGF is primarilymediated by its ability to upregulate Sloan-Kettering Instituteproto-oncogene (Ski)-related novel gene, non–Alu-containing(SnoN).14 Consistently, HGF induces SnoN expression in vivo,and ectopic expression of SnoN impedes TGF-1 action in tubularepithelial cells.14
SnoN belongs to a family of proteins that are known as Smadtranscriptional co-repressors, which include Ski, SnoN, andTG-interacting factor.15–17 Through various mechanisms,SnoN can efficiently block TGF-1/Smad signaling. Copious studieshave shown that SnoN can interact physically with Smad proteinsin the nuclei and sequester the ability of Smads to transactivategene expression.14,15 Furthermore, SnoN is instrumental in recruitingother transcriptional co-repressors, such as the nuclear hormonereceptor co-repressor and mammalian Sin3 ortholog, and preventsSmads from binding to the transcriptional co-activator p300/CBP.15,16In normal kidney, SnoN is highly expressed,18 which is presumablynecessary for keeping the TGF-1/Smad signaling tightly constrained.However, the abundance of SnoN protein is progressively reducedin the fibrotic kidney, primarily through an enhanced ubiquitin-dependentproteasomal degradation.19,20 Therefore, TGF-1/Smad signalingin the pathologic state is dramatically amplified by virtueof losing a crucial negative control mechanism. In this context,restoration of SnoN expression by HGF in the fibrotic kidneycould be a primary pathway leading to inhibition of the profibroticTGF-1/Smad signaling.
SnoN induction by HGF seems to occur in a cell type–specificmanner. In glomerular mesangial cells and interstitial fibroblasts,HGF is unable to induce SnoN gene expression, although it competentlyblocks TGF-1/Smad signaling via other mechanisms.21,22 Theseobservations indicate that HGF seems to elicit its antifibroticeffects in different kidney cells through distinctive modesof action. However, the molecular mechanism that governs thecell type–specific induction of SnoN by HGF in tubularepithelial cells remains completely unknown.
In this study, we show that induction of SnoN by HGF is operatedat the gene transcriptional level through an extracellular signal–regulatedkinase-1 and -2 (Erk-1/2)-dependent pathway. HGF seems to triggera cascade of signal transduction, which leads to the activationof the cAMP response element (CRE)-binding protein (CREB), atranscription factor that is selectively activated in tubularepithelial cells. Our data suggest that CREB activation, togetherwith Sp1, plays an essential role in mediating the cell type–specificinduction of SnoN.
SnoN Induction by HGF Occurs at the Gene Transcriptional Level Figure 1, A and B, shows that HGF specifically induced SnoNmRNA expression in human proximal tubular epithelial cells (HKC-8),consistent with a previous report.14 To understand the mechanismthat governs SnoN induction by HGF in tubular epithelial cells,we first determined whether SnoN induction occurs at the genetranscriptional or posttranscriptional level. To this end, HKC-8cells were incubated with HGF in the absence or presence ofactinomycin D, an inhibitor of gene transcription. As shownin Figure 1, C and D, in the absence of gene transcription aftertreatment with actinomycin D, HGF failed to induce SnoN mRNAand protein expression. In contrast, HGF was able to induceSnoN expression in HKC-8 cells when the gene transcription wasintact. Therefore, it seems clear that SnoN induction by HGFoccurs at the gene transcriptional level.
Figure 1. Induction of Sloan-Kettering Institute proto-oncogene–related novel gene, non–Alu-containing (SnoN) expression by hepatocyte growth factor (HGF) occurs at the gene transcriptional level. (A and B) Semiquantitative reverse transcriptase–PCR (RT-PCR) analyses demonstrated SnoN mRNA induction by HGF. Human proximal tubular epithelial (HKC-8) cells were treated with 20 ng/ml HGF for various periods of time as indicated. (A) Representative RT-PCR result. (B) Graphic presentation of the relative abundance of SnoN mRNA at various periods of time as indicated after normalization with -actin. Data are means ± SEM of three experiments. *P < 0.05 versus controls; **P < 0.001 versus controls. (C and D) RT-PCR and Western blot analyses demonstrated that the induction of SnoN expression by HGF occurred at the gene transcriptional level. HKC-8 cells were incubated in the absence or presence of actinomycin D (5 µg/ml) for 1 h, followed by treatment with or without HGF (20 ng/ml) for 3 h (C) or 24 h (D). Total cellular RNA was subjected to RT-PCR (C), whereas whole-cell lysates were immunoblotted with antibodies against SnoN and actin (D).
SnoN Induction Depends on the Activation of Erk-1/2 Signaling
We next sought to determine which signal pathway triggered byHGF is responsible for mediating SnoN induction. As shown inFigure 2, HGF activated multiple signal pathways in tubularepithelial cells. When HKC-8 cells were treated with HGF forvarious periods of time, Erk-1/2 mitogen-activated protein kinase(MAPK) was rapidly phosphorylated and activated, as demonstratedby phospho-specific antibody (Figure 2A). Similarly, HGF rapidlyinduced Akt kinase (Figure 2B) and p38 MAPK phosphorylation(Figure 2C) in HKC-8 cells. We found that blockade of Erk-1/2activation with PD98059 abolished SnoN induction by HGF (Figure 2E).However, blockade of Akt kinase and p38 MAPK activation hadno or little effect on SnoN induction. Of note, these chemicalinhibitors at the concentrations used were able to block Aktand p38 MAPK activation in HKC-8 cells, as previously reported.23,24These results suggest that HGF induces SnoN expression primarilyby an Erk-1/2 MAPK-dependent pathway in tubular epithelial cells.
Figure 2. Induction of SnoN by HGF depends on the extracellular signal–regulated kinase-1 and -2 (Erk-1/2) signaling. (A through C) HGF triggered various signaling pathways in tubular epithelial cells. HKC-8 cells were treated with 20 ng/ml HGF for various periods of time as indicated. Whole-cell lysates were immunoblotted with antibodies against either phospho-specific or total Erk-1/2 (A), Akt (B), and p38 mitogen-activated protein kinase (MAPK; C). (D and E) Blockade of Erk-1/2 activation abolished SnoN induction by HGF. HKC-8 cells were pretreated with either various chemical inhibitors or vehicle (DMSO) for 30 min, followed by incubation in the absence or presence of 20 ng/ml HGF for 1 (D) and 24 h (E). Special inhibitors for phosphatidylinositol 3-kinase (10 nM wortmannin), Mek1 (10 µM PD98059), and p38 MAPK (20 µM SC68376) were used. Whole-cell lysates were immunoblotted with antibodies against phospho-specific or total Erk-1/2, SnoN, and actin.
Consistent with the data on SnoN induction, we found that Erk-1/2activation was also essential for HGF to elicit its anti–TGF-1action. As shown in Figure 3, HGF abolished TGF-1–inducedE-cadherin suppression and -smooth muscle actin induction inHKC-8 cells; however, blockade of Erk-1/2 activation with PD98059completely restored the profibrotic TGF-1 action in tubularepithelial cells. Under the same conditions, blockade of Aktor p38 MAPK signaling had no or little effect on the abilityof HGF to antagonize TGF-1 action.
Figure 3. Blockade of Erk-1/2 activation abrogates the HGF-mediated inhibition of TGF-1 action in tubular epithelial cells. (A) HKC-8 cells were pretreated with either various chemical inhibitors or vehicle (DMSO) for 30 min, followed by incubation without or with 2 ng/ml TGF-1 and/or 20 ng/ml HGF for 48 h as indicated. Whole-cell lysates were immunoblotted with antibodies against E-cadherin, -smooth muscle actin (-SMA), and actin. (B through I) Immunofluorescence staining for E-cadherin and -SMA in HKC-8 cells after various treatments as indicated. T, TGF-1; H, HGF; PD, PD98059.
HGF Activates CREB and Induces Its Binding to SnoN Promoter Figure 4 shows that both upstream and downstream signaling mediatorsof Erk-1/2 were activated by HGF in tubular epithelial cells.HGF was able to induce a rapid phosphorylation of Raf and Mek1,the upstream kinase of Erk-1/2 in HKC-8 cells. Similarly, thedownstream effector kinase of Erk-1/2, p90RSK, was quickly activated.This led to the phosphorylation and activation of CREB, a transcriptionfactor that is activated by Erk-1/2–p90RSK signaling.25As demonstrated in Figure 4, phospho-specific CREB but not totalcellular CREB was dramatically increased shortly after HGF stimulationin HKC-8 cells. The band below phospho-CREB was likely the phosphorylatedactivating transcription factor-1, a CREB-related protein thatis also recognized by this antibody.
Figure 4. HGF activates both upstream and downstream signaling mediators of Erk-1/2 in tubular epithelial cells. HKC-8 cells were incubated with 20 ng/ml HGF for various periods of time as indicated, and the cell lysates were immunoblotted with antibodies against phosphorylated Raf, phosphorylated Mek1, phosphorylated ant total Erk-1/2, phosphorylated p90RSK, phosphorylated and total cAMP response element (CRE)-binding protein (CREB), and -tubulin. The band below phospho-CREB was likely the phospho–activating transcription factor-1 (ATF-1), a CREB-related protein that is also recognized by this antibody.
To establish the relevance of CREB activation to SnoN induction,we analyzed the promoter region of human SnoN gene. As shownin Figure 5A, there were two putative cAMP response elements(CRE) in the SnoN promoter region. Chromatin immunoprecipitation(ChIP) assay revealed that HGF induced phosphorylated CREB bindingto that region in the SnoN promoter in a time-dependent manner(Figure 5B). Quantitative analysis showed that HGF increasedthe interaction between phospho-CREB and SnoN promoter by approximately14-fold at 30 min after stimulation (Figure 5C).
Figure 5. HGF induces the binding of CREB to its cognate element in the SnoN gene promoter. (A) Partial sequence of human SnoN gene promoter region. Highlighted are two putative CRE and two Sp1 binding sites in this region. (B and C) Chromatin immunoprecipitation (ChIP) assay revealed that HGF induced phosphorylated CREB binding to SnoN promoter. HKC-8 cells were treated with 20 ng/ml HGF for various periods of time as indicated and subjected to ChIP assay. Shown are a representative ChIP assay (B) and the quantitative data of three independent experiments (C). * P < 0.05.
CREB Activation Is Necessary for Mediating SnoN Induction
To examine the importance of CREB activation in SnoN expression,we sought to investigate whether overexpression of CREB mediatesSnoN induction in tubular epithelial cells. To this end, HKC-8cells were transiently transfected with wild-type CREB expressionvector (pCMV-wt-CREB), followed by incubation with HGF for 24h. As shown in Figure 6A, ectopic expression of CREB inducedSnoN expression in HKC-8 cells in the absence of HGF (Figure 6A,lane 1 versus lane 3). SnoN expression was further increasedin the CREB-transfected cells after HGF stimulation (Figure 6A,lane 3 versus lane 4).
Figure 6. CREB is necessary for mediating SnoN induction in tubular epithelial cells. (A) Ectopic expression of wild-type CREB induced SnoN expression. HKC-8 cells were transiently transfected with empty vector pcDNA3 or wild-type CREB expression vector (pCMV-wt-CREB) as indicated, followed by incubation with 20 ng/ml HGF for 24 h. Whole-cell lysates were immunoblotted with antibodies against SnoN and -tubulin. (B) Overexpression of mutant CREB abolished SnoN induction by HGF. HKC-8 cells were transfected with either wild-type (pCMV-wt-CREB) or mutant CREB expression vector (pCMV-dn-CREB), followed by incubation with 20 ng/ml HGF for 24 h. Mutant CREB abrogated SnoN induction in a dominant negative manner. (C) Immunofluorescence staining for SnoN expression in tubular epithelial cells. HKC-8 cells were transfected with either pCMV-wt-CREB or pCMV-dn-CREB expression vectors, followed by incubation with or without HGF (20 ng/ml) for 24 h. Cell nuclear staining was carried out with 4',6-diamidino-2-phenylindole, HCl.
Figure 6B shows that CREB activation was also required for HGF-mediatedSnoN induction in tubular epithelial cells. HKC-8 cells weretransfected with dominant negative mutant CREB expression vector(pCMV-dn-CREB) that expresses a variant of the human CREB proteinthat contains a serine-to-alanine mutation, which blocks phosphorylationof CREB, thereby preventing transcription in a dominant negativemanner. We found that overexpression of mutant CREB completelyabolished SnoN induction by HGF in HKC-8 cells (Figure 6B),suggesting that CREB activation is essential for SnoN inductionin tubular epithelial cells. Similar results were obtained whenusing immunofluorescence staining for SnoN (Figure 6C).
HGF Selectively Activates CREB in Tubular Epithelial Cells
Because HGF induces SnoN expression in a cell type–specificmanner, we next determined whether there is a variation in CREBactivation by HGF in different types of kidney cells. Therefore,HKC-8 cells, human mesangial cells (HMC), and rat kidney interstitialfibroblast (NRK-49F) cells were treated with HGF for variousperiods of time, and CREB activation was assessed. As shownin Figure 7, HGF induced a dramatic phosphorylation and activationof CREB in HKC-8 cells but not in HMC and NRK-49F cells. However,Erk-1/2 phosphorylation was detected in all three cell typesafter HGF stimulation (Figure 7). These data suggest that thereis a strong correlation between CREB activation and SnoN inductionin different kidney cells.
Figure 7. HGF induces CREB phosphorylation and activation in a cell type–specific manner. HKC-8 cells, human mesangial cells (HMC), and rat kidney interstitial fibroblasts (NRK-49F) cells were treated with 20 ng/ml HGF for various periods of time as indicated. Whole-cell lysates were immunoblotted with antibodies against phosphorylated CREB, phosphorylated Erk-1/2, and actin. Despite Erk-1/2 activation in all three cell types, significant CREB phosphorylation was found only in HKC-8 cells.
Sp1 Is Indispensable for Mediating the Cell Type–Specific Induction of SnoN by HGF
Sequence analysis also showed that there were two Sp1 bindingsites in the proximity of CRE in the SnoN promoter (Figure 5A).This prompted us to examine whether Sp1 transcription factorparticipates in SnoN regulation. ChIP assay demonstrated thatSp1 could constitutively bind to its cognate sites in the SnoNpromoter region (Figure 8A). Furthermore, HGF substantiallyaugmented the interaction between Sp1 and SnoN promoter in atime-dependent manner (Figure 8, A and B). To test the functionalityof this Sp1 binding, we examined the effect of chemical blockadeof Sp1 binding on SnoN expression. Therefore, HKC-8 cells werepretreated with mithramycin A, a potent inhibitor of Sp1 binding,26–28followed by incubation with HGF for 24 h. As shown in Figure 8,C and D, blockade of Sp1 binding suppressed SnoN expressionat basal condition and completely abolished SnoN induction byHGF. Because Sp1 is predominantly expressed in epithelial cellsin the kidney,29 it seems that Sp1 also plays a critical rolein mediating the cell type–specific induction of SnoNby HGF.
Figure 8. HGF promotes Sp1 binding to SnoN promoter, and blockade of Sp1 binding abolishes SnoN induction by HGF in tubular epithelial cells. (A and B) ChIP assay revealed that HGF promoted Sp1 binding to its cognate sites in the SnoN promoter. HKC-8 cells were treated with 20 ng/ml HGF for various periods of time as indicated and subjected to ChIP assay. Shown are a representative ChIP assay (A) and the quantitative data of three independent experiments (B). *P < 0.01. (C and D) Blockade of Sp1 binding abolished SnoN induction by HGF. HKC-8 cells were pretreated with 10–7 M mithramycin A for 16 h, followed by incubation with 20 ng/ml HGF for 24 h. (C) Representative Western blot showing SnoN abundance after various treatments. (D) Graphic presentation of relative abundance of SnoN after normalization with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in different groups. Relative levels (fold induction over control) are means ± SEM of three experiments. *P < 0.001 versus normal control; P < 0.001 versus HGF group.
HGF Induces Erk-1/2 and CREB Activation In Vivo
Earlier studies showed that HGF can restore SnoN expressionthat was repressed in the obstructed kidney in vivo.14 To explorethe mechanism by which HGF induces SnoN expression in vivo,we examined the Erk-1/2 and CREB activation in the obstructedkidney after delivery of HGF gene. As shown in Figure 9, A andB, administration of HGF gene induced Erk-1/2 phosphorylationin the obstructed kidney after unilateral ureteral obstruction(UUO). The relative abundance of phosphorylated Erk-1/2 in theobstructed kidney after HGF administration was increased byapproximately 12-fold over the controls. Similarly, phospho-CREBwas markedly induced by HGF and primarily localized in the nucleiof tubular epithelia (Figure 9C), a pattern that closely resemblesSnoN expression in this model.14 To confirm further the relevanceof CREB activation to SnoN induction, we used a double immunostainingapproach to examine the co-localization of phosphorylated CREBand SnoN. As shown in Figure 9D, co-localization of phosphorylatedCREB and SnoN was clearly observable in the nuclei of the obstructedkidney after administration of HGF gene. Therefore, it seemsthat HGF also induces SnoN expression in an Erk-1/2–CREB–dependentpathway in vivo.
Figure 9. HGF induces Erk-1/2 and CREB phosphorylation and SnoN expression in vivo. (A and B) Administration of HGF gene induced Erk-1/2 phosphorylation in the obstructed kidney. Whole-tissue homogenates from the obstructed kidney of different groups as indicated were immunoblotted with antibodies against phosphorylated and total Erk-1/2. Numbers (1, 2, and 3) indicate each individual animal in a given group. Relative abundance of phosphorylated Erk-1/2 after normalization with total Erk-1/2 in each group is shown in B. *P < 0.05 versus the pcDNA3 control. (C) Immunohistochemical staining showed that HGF induced CREB phosphorylation in the obstructed kidney. Phosphorylated CREB was predominantly localized in the nuclei of renal tubules. (D) Double immunofluorescence staining for phosphorylated CREB (red) and SnoN (green) in the obstructed kidneys. Co-localization of phospho-CREB and SnoN (arrowheads) was evident in the obstructed kidney after HGF administration. Nuclear staining in kidney tissue was carried out with4',6-diamidino-2-phenylindole, HCl.
HGF, a potent endogenous antifibrotic cytokine that is capableof antagonizing the profibrotic TGF-1 signaling, specificallyinduces SnoN mRNA and protein expression in kidney proximaltubular epithelial cells14 but not in glomerular mesangial cellsand interstitial fibroblasts.21,22 However, the molecular basisfor this cell type–specific induction was completely unknown.The results presented herein demonstrate that this cell type–specificinduction of SnoN is mediated primarily by selective activationof CREB transcription factor in the tubular epithelial cellsafter HGF stimulation. HGF activates Erk-1/2 MAPK signalingthat involves a cascade of signal transduction, which leadsto phosphorylation and activation of CREB. Phosphorylated CREBthen binds to the cis-acting CRE in the promoter region of humanSnoN gene and presumably trans-activates SnoN transcription.Consistently, overexpression of CREB induces SnoN expressionin tubular epithelial cells, and a dominant negative mutantCREB abrogates SnoN induction by HGF. These results establishthat selective CREB activation may play a crucial role in conferringthe cell type–specific induction of SnoN in tubular epithelialcells. Our studies provide significant insights into understandingthe mechanism underlying SnoN transcriptional regulation inkidney cells.
CREB is a ubiquitously expressed nuclear transcription factorthat contains the basic region/leucine zipper domain. It isoriginally isolated as a gene regulatory protein that bindsto the CRE site of the somatostatin gene promoter region inresponse to an increase in the cellular cAMP level.25,30 CREBforms a homodimer through its leucine zipper domain and bindsto DNA through its basic region and is involved in regulatingthe expression of a wide range of genes that are important tocell proliferation, differentiation, adaptation, and survival.25,30,31Activation of CREB requires phosphorylation of the serine residueat 133,32 which is necessary and sufficient for transcriptionalactivation of target genes. This study demonstrates that HGFactivates SnoN expression by an Erk-1/2–CREB–dependentpathway and suggests that SnoN is a direct target of CREB inkidney cells. Because SnoN is intimately implicated in the regulationof TGF-1/Smad signaling, this suggests that CREB may be an importanttranscriptional regulator that is capable of controlling thefibrogenic responses of tubular epithelial cells after chronicinjury. Of interest, a recent report indicated that CREB activationalso plays a critical role in mediating Erk-1/2–inducedtubular epithelial cell survival after oxidant stress.33 Therefore,it is conceivable that CREB activation may be vital for mediatingthe beneficial effects of HGF by virtue of its ability to induceSnoN expression and to promote tubular epithelial cell survival.Consistently, attenuation of renal fibrosis and induction ofSnoN after administration of HGF gene in vivo are closely correlatedwith the activation of Erk-1/2 and CREB in obstructive nephropathy(Figure 9).12,14
CREB could be activated by a wide range of extracellular stimulithrough distinct signaling pathways.25,34 However, it seemsthat HGF-mediated CREB activation in tubular epithelial cellsis mainly dependent on Erk-1/2 signaling, because blockade ofErk-1/2 activation completely abolishes SnoN induction. Of note,TGF-1 itself does not activate Erk-1/2 signaling in HKC-8 cells.14Although p38 MAPK and Akt may potentially stimulate CREB phosphorylation,25blockade of p38 MAPK and Akt signaling displays little effecton SnoN induction by HGF (Figure 2). p90RSK, one of the CREBkinases downstream of Erk-1/2,35,36 is activated and may bedirectly responsible for CREB activation after HGF stimulation(Figure 4). The importance of CREB activation in mediating SnoNinduction by HGF is corroborated by several lines of evidence.First, CREB is rapidly activated in tubular epithelial cellsin response to HGF stimulation. Second, SnoN gene promoter harborstwo CRE and phosphorylated CREB binds to the CRE-containingregion of SnoN promoter. Third, ectopic expression of CREB inducesSnoN expression and dominant negative CREB abrogates SnoN inductionby HGF. Fourth, CREB is selectively activated in tubular epithelialcells but not in mesangial cells and fibroblasts after HGF stimulation,suggesting an intrinsic coupling of SnoN induction to CREB activation.Finally, SnoN induction by HGF in vivo is associated with Erk-1/2and CREB activation in obstructive nephropathy. Therefore, itseems plausible that selective CREB activation after HGF stimulationis indispensable for conferring the cell type–specificinduction of SnoN.
Besides CREB, HGF induction of SnoN is completely dependenton proper binding and function of a ubiquitous transcriptionfactor Sp1, as chemical ablation of Sp1 binding abrogates HGF-inducedSnoN expression in tubular epithelial cells (Figure 8). Sp1is the prototype of a family of proteins that consists of fourmembers with distinct expression pattern and diverse functionsin different types of cells.37 The abundance of Sp1 proteindisplays tremendous variations in different types of cells,29albeit ubiquitously expressed. In kidney, proximal tubular epithelialcells express a high level of Sp1, whereas glomerular mesangialcells and interstitial fibroblasts barely produce this protein.29These observations establish a tight correlation between SnoNinducibility by HGF and the levels of Sp1 in various types ofkidney cells. Hence, cellular Sp1 level probably is anotherkey molecular determinant for SnoN induction in diverse typesof kidney cells in response to HGF stimulation. In support ofthis, there are numerous Sp1 sites in SnoN promoter, and Sp1binds to SnoN promoter after HGF stimulation, as demonstratedby ChIP assay.
Cooperation between CREB and Sp1 may be operative in mediatingSnoN induction in response to HGF stimulation. Sp1 sites arelocated in close proximity to CRE in SnoN promoter, suggestinga physical availability for Sp1 and CREB to interact. Of note,this type of cooperation between Sp1 and CREB for gene regulationis not without precedent. In the promoters of the Na-K-ATPase1-subunit, a synergy between Sp1 and CREB for gene activationis documented.38 Hence, cooperation between Sp1 and CREB mayrepresent a general regulatory mechanism for conferring inductionof many genes. It should be stressed that although Sp1 is aubiquitous protein that regulates the constitutive expressionof many genes, it also participates in controlling either suppressiveor inducible expression of a single gene under different circumstances,often through interacting with different partners.26,39 Forinstance, Sp1 mediates HGF receptor (c-met) suppression afteroxidative stress by interacting with Egr-1, leading to its sequestration.40Sp1 is also crucial for c-met induction triggered by TGF-1 throughinteracting with Smad proteins.28 The involvement of both Sp1and CREB in SnoN induction is also highlighted by their bindingto the cognate sites in SnoN promoter in a HGF-dependent manner.
SnoN is an imperative transcriptional co-repressor that makesthe TGF-1/Smad signaling tightly controlled in normal conditions.Given its eminent role in constraining TGF-1 signaling, SnoNexpression is stringently regulated at both transcriptionaland posttranscriptional levels. Earlier studies demonstratedthat SnoN protein is progressively reduced in obstructive nephropathy,18and such downregulation of SnoN is primarily mediated by anenhanced ubiquitin-dependent degradation.19,20 However, as shownin this study, SnoN induction by HGF occurs at the transcriptionallevel and is mediated by transcription factors CREB and Sp1.Although more studies are needed, these observations have seta foundation for better understanding of the mechanism thatgoverns the regulation of SnoN under different circumstances.
Cell Culture and Treatment
Human proximal tubular epithelial cells (HKC-8) were culturedin DMEM and Ham's F12 medium (1:1) supplemented with 5% FBS(Invitrogen, Carlsbad, CA), as described previously.12,41 HMCwere purchased from ScienCell Research Laboratories (San Diego,CA).42 Normal rat kidney interstitial fibroblast cells (NRK-49F)were obtained from American Type Culture Collection (Manassas,VA). HMC and NRK-49F cells were maintained in DMEM/F12 mediumsupplemented with 10% FBS. Cells were typically serum-starvedfor 16 h, followed by incubation with HGF for various periodsof time as indicated. In some experiments, cells were pretreatedwith either various inhibitors at given concentrations or vehicle(0.1% DMSO) 0.5 h before incubation with different cytokines.Human recombinant HGF and TGF-1 were obtained from R&D Systems(Minneapolis, MN). PD98059 (Mek1 inhibitor), wortmannin (phosphatidylinositol3-kinase inhibitor), and SC-68376 (p38 MAPK inhibitor) werepurchased from Calbiochem (La Jolla, CA). Actinomycin D andmithramycin A were purchased from Sigma (St. Louis, MO). Forchemical blockade of Sp1 binding, HKC-8 cells were pretreatedwith mithramycin A at 10–7 M for 16 h.28
Animal Model and HGF Treatment
Animal studies were approved by the Institutional Animal Careand Use Committee at the University of Pittsburgh and carriedout as described previously.43 Briefly, male CD-1 mice thatweighed 18 to 22 g were obtained from Harlan Sprague-Dawley(Indianapolis, IN). They were housed in the animal facilitiesof the University of Pittsburgh Medical Center, with free accessto food and water. UUO was performed using an established procedure.12On the day of surgery, mice received a single intravenous injectionof naked HGF expression plasmid (pCMV-HGF) or empty vector pcDNA3at 1 mg/kg body wt. At day 7 after UUO, mice were killed andthe kidneys were removed. One part of the kidneys was frozenimmediately in OCT compound for cryosection to perform immunofluorescencestudies. The remaining kidneys were snap-frozen in liquid nitrogenand used for Western blot analyses.
RNA Isolation and Reverse Transcriptase–PCR
Total RNA isolation, reverse transcription of the RNA, and PCRamplification were performed as described previously.44 Briefly,the first-strand cDNA synthesis was carried out by using a ReverseTranscription System kit according to the instructions of themanufacturer (Promega, Madison, WI). PCR amplification was performedusing HotStar TaqMaster Mix Kit (Qiagen, Valencia, CA). Theprimer sequences were as follow: SnoN 5'-TTTCTGCCTCTTCCATCACC-3'(sense) and 5'-GACTTGGGGCAAACAGAGTC-3' (antisense) and -actin,5'-TCAAGATCATTGCTCCTCCTGAGC-3' (sense) and 5'-TGCTGTCACCTTCACCGTTCCAGT-3'(antisense). Relative levels of SnoN mRNA (fold induction overthe controls) were calculated after normalization with housekeepinggene -actin.
Western Blot Analysis
Detection of protein expression by Western blot was carriedout according to the established protocols described previously.12The primary antibodies used were as follows: Anti-SnoN (sc-9595),anti-Sp1 (sc-59), and anti-actin (sc-1616) (Santa Cruz Biotechnology,Santa Cruz, CA); anti–-smooth muscle actin (clone 1A4)and anti–-tubulin (T-9026l Sigma); anti-glyceraldehyde-3-phosphatedehydrogenase (Ambion, Austin, TX); anti–E-cadherin (clone36; BD Biosciences, San Diego, CA); antibodies against phospho-specific(Thr202/Tyr204) and total Erk-1/2, phospho-specific (Ser473)and total Akt, phospho-specific (Thr183/Tyr182) and total p38MAPK, phospho-specific (Ser133) and total CREB, phospho-specificRaf (Ser338), phospho-specific Mek1/2 (Ser217/221), and phospho-specificp90 ribosomal protein S6 kinase (p90RSK; Ser380; Cell SignalingTechnology, Beverly, MA). Quantitative analysis of Western blotdata was performed by measurement of the intensity of the bandsignals with the use of National Institutes of Health Imageanalysis software.
Immunofluorescence and Immunohistochemical Staining
Indirect immunofluorescence staining was performed using anestablished procedure.45 Briefly, HKC-8 cells cultured on coverslips,or kidney cryosections at 5-µm thickness were washed twicewith cold PBS and fixed with cold methanol:acetone (1:1) for10 min at –20°C. After three extensive washes withPBS containing 0.5% BSA, the cells were blocked with 20% normaldonkey serum in PBS buffer for 30 min at room temperature andthen incubated with specific primary antibodies described previously.For visualization of the primary antibodies, slides were stainedwith cyanine Cy2- or Cy3-conjugated secondary antibodies (JacksonImmunoResearch Laboratories, West Grove, PA). Cells were double-stainedwith 4',6-diamidino-2-phenylindole, HCl to visualize the nuclei.As a negative control, the primary antibody was replaced withnon-IgG, and no staining occurred. For immunohistochemical stainingof kidney sections, paraffin-embedded slides were stained withanti–phospho-CREB antibody using the Vector M.O.M. immunodetectionkit by the protocol specified by the manufacturer (Vector Laboratories,Burlingame, CA). Stained slides were mounted with anti-fademounting medium (Vector Laboratories) and viewed with a NikonEclipse E600 Epi-fluorescence microscope (Melville, NY).
ChIP Assay
ChIP assay was performed to analyze in vivo interactions oftranscription factors and their cognate cis-acting elementsin SnoN promoter. This assay was essentially carried out accordingto the protocols specified by the manufacturer (ChIP assay kit;Upstate, Charlottesville, VA). Briefly, after treatment withHGF for various periods of time, HKC-8 cells were cross-linkedwith 1% formaldehyde for 10 min, washed twice with PBS containingprotease inhibitors and scraped, and then resuspended in SDSlysis buffer containing protease inhibitors. The chromatin solutionwas sonicated. After centrifugation, the supernatant was diluted10-fold in ChIP dilution buffer and precleared with proteinA–agarose containing salmon sperm DNA for 1 h. The antibodiesagainst phospho-specific CREB or Sp1 were added and incubatedat 4°C overnight, followed by incubation with protein A–agarosefor 1 h. The precipitates were washed, and chromatin complexeswere eluted. After reversal of the cross-linking at 65°Cfor 4 h, the DNA was purified, and 1 µl of input controlor ChIP samples were used as a template for PCR using the primersets for the SnoN promoter regions (from –1838 to –1639)containing two putative CRE and two Sp1 sites. The sequencesof primers used for ChIP assay were as follows: Forward 5'-GGTCGGGGCCAGAGG-3'and reverse 5'-CGGACCGGGTCGAGA-3'. The relative abundance ofthe PCR amplified product was calculated after normalizationwith the intensities of input DNA.
Plasmid Transfection
For transient transfection, HKC-8 cells were seeded in six-wellplates at 5 x 105 cells/well. The cells were then transfectedwith either human wild-type CREB (pCMV-wt-CREB) or dominantnegative mutant CREB (pCMV-dn-CREB) expression vectors (ClontechLaboratories, Mountain View, CA) using Lipofectamine 2000 reagentaccording to the instructions specified by the manufacturer(Invitrogen). The pCMV-dn-CREB vector expresses a mutant variantof the human CREB protein that contains a serine-to-alaninemutation corresponding to amino acid residue 133.32 The emptypcDNA3 vector was also transfected as a negative control. Aftertransfection for 24 h, the cells were incubated with 20 ng/mlHGF for an additional 24 h and then subjected to Western blotanalyses and immunofluorescence staining.
Statistical Analyses
Statistical analysis was performed using SigmaStat software(Jandel Scientific Software, San Rafael, CA). Comparisons betweengroups were made using one-way ANOVA, followed by the Student-Newman-Keulstest. P < 0.05 was considered significant.
Liu Y: Hepatocyte growth factor in kidney fibrosis: Therapeutic potential and mechanisms of action.
Am J Physiol Renal Physiol 287
: F7
–F16, 2004[Abstract/Free Full Text]
Mizuno S, Nakamura T: Suppressions of chronic glomerular injuries and TGF-beta 1 production by HGF in attenuation of murine diabetic nephropathy.
Am J Physiol Renal Physiol 286
: F134
–F143, 2004[Abstract/Free Full Text]
Azuma H, Takahara S, Matsumoto K, Ichimaru N, Wang JD, Moriyama T, Waaga AM, Kitamura M, Otsuki Y, Okuyama A, Katsuoka Y, Chandraker A, Sayegh MH, Nakamura T: Hepatocyte growth factor prevents the development of chronic allograft nephropathy in rats.
J Am Soc Nephrol 12
: 1280
–1292, 2001[Abstract/Free Full Text]
Mizuno S, Kurosawa T, Matsumoto K, Mizuno-Horikawa Y, Okamoto M, Nakamura T: Hepatocyte growth factor prevents renal fibrosis and dysfunction in a mouse model of chronic renal disease.
J Clin Invest 101
: 1827
–1834, 1998[Medline]
Herrero-Fresneda I, Torras J, Franquesa M, Vidal A, Cruzado JM, Lloberas N, Fillat C, Grinyo JM: HGF gene therapy attenuates renal allograft scarring by preventing the profibrotic inflammatory-induced mechanisms.
Kidney Int 70
: 265
–274, 2006[CrossRef][Medline]
Dworkin LD, Gong R, Tolbert E, Centracchio J, Yano N, Zanabi AR, Esparza A, Rifai A: Hepatocyte growth factor ameliorates progression of interstitial fibrosis in rats with established renal injury.
Kidney Int 65
: 409
–419, 2004[CrossRef][Medline]
Dai C, Yang J, Bastacky S, Xia J, Li Y, Liu Y: Intravenous administration of hepatocyte growth factor gene ameliorates diabetic nephropathy in mice.
J Am Soc Nephrol 15
: 2637
–2647, 2004[Abstract/Free Full Text]
Cruzado JM, Lloberas N, Torras J, Riera M, Fillat C, Herrero-Fresneda I, Aran JM, Alperovich G, Vidal A, Grinyo JM: Regression of advanced diabetic nephropathy by hepatocyte growth factor gene therapy in rats.
Diabetes 53
: 1119
–1127, 2004[Abstract/Free Full Text]
Mizuno S, Matsumoto K, Kurosawa T, Mizuno-Horikawa Y, Nakamura T: Reciprocal balance of hepatocyte growth factor and transforming growth factor-beta1 in renal fibrosis in mice.
Kidney Int 57
: 937
–948, 2000[Medline]
Liu Y, Rajur K, Tolbert E, Dworkin LD: Endogenous hepatocyte growth factor ameliorates chronic renal injury by activating matrix degradation pathways.
Kidney Int 58
: 2028
–2043, 2000[CrossRef][Medline]
Yang J, Liu Y: Blockage of tubular epithelial to myofibroblast transition by hepatocyte growth factor prevents renal interstitial fibrosis.
J Am Soc Nephrol 13
: 96
–107, 2002[Abstract/Free Full Text]
Liu Y: Epithelial to mesenchymal transition in renal fibrogenesis: Pathologic significance, molecular mechanism, and therapeutic intervention.
J Am Soc Nephrol 15
: 1
–12, 2004[Abstract/Free Full Text]
Yang J, Dai C, Liu Y: A novel mechanism by which hepatocyte growth factor blocks tubular epithelial to mesenchymal transition.
J Am Soc Nephrol 16
: 68
–78, 2005[Abstract/Free Full Text]
Luo K: Ski and SnoN: Negative regulators of TGF-beta signaling.
Curr Opin Genet Dev 14
: 65
–70, 2004[CrossRef][Medline]
Liu X, Sun Y, Weinberg RA, Lodish HF: Ski/Sno and TGF-beta signaling.
Cytokine Growth Factor Rev 12
: 1
–8, 2001[CrossRef][Medline]
Bottinger EP, Bitzer M: TGF-beta1 signaling in renal disease.
J Am Soc Nephrol 13
: 2600
–2610, 2002[Abstract/Free Full Text]
Yang J, Zhang X, Li Y, Liu Y: Downregulation of Smad transcriptional corepressors SnoN and Ski in the fibrotic kidney: An amplification mechanism for TGF-beta1 signaling.
J Am Soc Nephrol 14
: 3167
–3177, 2003[Abstract/Free Full Text]
Tan R, Zhang J, Tan X, Zhang X, Yang J, Liu Y: Downregulation of SnoN expression in obstructive nephropathy is mediated by an enhanced ubiquitin-dependent degradation.
J Am Soc Nephrol 17
: 2781
–2791, 2006[Abstract/Free Full Text]
Fukasawa H, Yamamoto T, Togawa A, Ohashi N, Fujigaki Y, Oda T, Uchida C, Kitagawa K, Hattori T, Suzuki S, Kitagawa M, Hishida A: Ubiquitin-dependent degradation of SnoN and Ski is increased in renal fibrosis induced by obstructive injury.
Kidney Int 69
: 1733
–1740, 2006[CrossRef][Medline]
Yang J, Dai C, Liu Y: Hepatocyte growth factor suppresses renal interstitial myofibroblast activation and intercepts Smad signal transduction.
Am J Pathol 163
: 621
–632, 2003[Abstract/Free Full Text]
Dai C, Liu Y: Hepatocyte growth factor antagonizes the profibrotic action of TGF-beta1 in mesangial cells by stabilizing Smad transcriptional corepressor TGIF.
J Am Soc Nephrol 15
: 1402
–1412, 2004[Abstract/Free Full Text]
Liu Y: Hepatocyte growth factor promotes renal epithelial cell survival by dual mechanisms.
Am J Physiol 277
: F624
–F633, 1999[Medline]
Dai C, Yang J, Liu Y: Transforming growth factor-beta1 potentiates renal tubular epithelial cell death by a mechanism independent of Smad signaling.
J Biol Chem 278
: 12537
–12545, 2003[Abstract/Free Full Text]
Ichiki T: Role of cAMP response element binding protein in cardiovascular remodeling: Good, bad, or both?
Arterioscler Thromb Vasc Biol 26
: 449
–455, 2006[Abstract/Free Full Text]
Poncelet AC, Schnaper HW: Sp1 and Smad proteins cooperate to mediate transforming growth factor-beta 1-induced alpha 2(I) collagen expression in human glomerular mesangial cells.
J Biol Chem 276
: 6983
–6992, 2001[Abstract/Free Full Text]
Ray R, Snyder RC, Thomas S, Koller CA, Miller DM: Mithramycin blocks protein binding and function of the SV40 early promoter.
J Clin Invest 83
: 2003
–2007, 1989[Medline]
Zhang X, Yang J, Li Y, Liu Y: Both Sp1 and Smad participate in mediating TGF-beta1-induced HGF receptor expression in renal epithelial cells.
Am J Physiol Renal Physiol 288
: F16
–F26, 2005[Abstract/Free Full Text]
Zhang X, Li Y, Dai C, Yang J, Mundel P, Liu Y: Sp1 and Sp3 transcription factors synergistically regulate HGF receptor gene expression in kidney.
Am J Physiol Renal Physiol 284
: F82
–F94, 2003[Abstract/Free Full Text]
Conkright MD, Montminy M: CREB: The unindicted cancer co-conspirator.
Trends Cell Biol 15
: 457
–459, 2005[CrossRef][Medline]
Gonzalez GA, Montminy MR: Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133.
Cell 59
: 675
–680, 1989[CrossRef][Medline]
Arany I, Megyesi JK, Reusch JE, Safirstein RL: CREB mediates ERK-induced survival of mouse renal tubular cells after oxidant stress.
Kidney Int 68
: 1573
–1582, 2005[CrossRef][Medline]
Deak M, Clifton AD, Lucocq LM, Alessi DR: Mitogen- and stress-activated protein kinase-1 (MSK1) is directly activated by MAPK and SAPK2/p38, and may mediate activation of CREB.
EMBO J 17
: 4426
–4441, 1998[CrossRef][Medline]
Bohm M, Moellmann G, Cheng E, Alvarez-Franco M, Wagner S, Sassone-Corsi P, Halaban R: Identification of p90RSK as the probable CREB-Ser133 kinase in human melanocytes.
Cell Growth Differ 6
: 291
–302, 1995[Abstract]
Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K, Cobb MH: Mitogen-activated protein (MAP) kinase pathways: Regulation and physiological functions.
Endocr Rev 22
: 153
–183, 2001[Abstract/Free Full Text]
Suske G: The Sp-family of transcription factors.
Gene 238
: 291
–300, 1999[CrossRef][Medline]
Matlhagela K, Taub M: Regulation of the Na-K-ATPase beta(1)-subunit promoter by multiple prostaglandin-responsive elements.
Am J Physiol Renal Physiol 291
: F635
–F646, 2006[Abstract/Free Full Text]
Jungert K, Buck A, Buchholz M, Wagner M, Adler G, Gress TM, Ellenrieder V: Smad-Sp1 complexes mediate TGFbeta-induced early transcription of oncogenic Smad7 in pancreatic cancer cells.
Carcinogenesis 27
: 2392
–2401, 2006[Abstract/Free Full Text]
Zhang X, Liu Y: Suppression of HGF receptor gene expression by oxidative stress is mediated through the interplay between Sp1 and Egr-1.
Am J Physiol Renal Physiol 284
: F1216
–F1225, 2003[Abstract/Free Full Text]
Racusen LC, Monteil C, Sgrignoli A, Lucskay M, Marouillat S, Rhim JG, Morin JP: Cell lines with extended in vitro growth potential from human renal proximal tubule: characterization, response to inducers, and comparison with established cell lines.
J Lab Clin Med 129
: 318
–329, 1997[CrossRef][Medline]
Li Y, Wen X, Spataro BC, Hu K, Dai C, Liu Y: Hepatocyte growth factor is a downstream effector that mediates the antifibrotic action of peroxisome proliferator-activated receptor-gamma agonists.
J Am Soc Nephrol 17
: 54
–65, 2006[Abstract/Free Full Text]
Yang J, Dai C, Liu Y: Hepatocyte growth factor gene therapy and angiotensin II blockade synergistically attenuate renal interstitial fibrosis in mice.
J Am Soc Nephrol 13
: 2464
–2477, 2002[Abstract/Free Full Text]
Yang J, Shultz RW, Mars WM, Wegner RE, Li Y, Dai C, Nejak K, Liu Y: Disruption of tissue-type plasminogen activator gene in mice reduces renal interstitial fibrosis in obstructive nephropathy.
J Clin Invest 110
: 1525
–1538, 2002[CrossRef][Medline]
Yang J, Liu Y: Dissection of key events in tubular epithelial to myofibroblast transition and its implications in renal interstitial fibrosis.
Am J Pathol 159
: 1465
–1475, 2001[Abstract/Free Full Text]
This article has been cited by other articles:
R. Tan, W. He, X. Lin, L. P. Kiss, and Y. Liu Smad ubiquitination regulatory factor-2 in the fibrotic kidney: regulation, target specificity, and functional implication
Am J Physiol Renal Physiol,
May 1, 2008;
294(5):
F1076 - F1083.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
1/Smad signaling in diverse types of kidney cells by different mechanisms. HGF is shown to induce Smad co-repressor Sloan-Kettering Institute proto-oncogene–related novel gene, non–Alu-containing (SnoN) expression in proximal tubular epithelial cells (HKC-8) but not in glomerular mesangial cells and interstitial fibroblasts. This study investigated the molecular mechanisms underlying the cell type–specific induction of SnoN by HGF. Treatment of HKC-8 cells with actinomycin D completely abolished HGF-mediated SnoN induction, suggesting its dependence on gene transcription. Although HGF activated several signal pathways in HKC-8 cells, blockade of extracellular signal–regulated kinase-1 and -2 (Erk-1/2) activation but not Akt and p38 mitogen-activated protein kinase abolished SnoN induction. HGF rapidly activated both upstream and downstream signaling of Erk-1/2, which led to the activation of the cAMP response element–binding protein (CREB). In the promoter region of human SnoN gene, two cAMP response elements were located in close proximity to Sp1 sites. Chromatin immunoprecipitation assay revealed that activated CREB and Sp1 bound to their cognate cis-acting elements in SnoN promoter in response to HGF stimulation. Ectopic expression of wild-type CREB promoted SnoN expression, whereas dominant negative mutant CREB abrogated SnoN induction by HGF. Likewise, chemical blockade of Sp1 binding abolished HGF-mediated SnoN induction. Furthermore, HGF selectively induced CREB phosphorylation in HKC-8 cells but not in mesangial cells and fibroblasts. In vivo, administration of HGF gene induced renal Erk-1/2 phosphorylation, CREB activation, and SnoN expression in obstructive nephropathy. Collectively, these results suggest that CREB activation, in concert with Sp1, constitutes a molecular switch that confers the cell type–specific induction of SnoN in response to HGF stimulation. 
-smooth muscle actin induction in HKC-8 cells; however, blockade of Erk-1/2 activation with PD98059 completely restored the profibrotic TGF-
