Expression patterns of RelA and c-mip are associated with different glomerular diseases following anti-VEGF therapy
2013; Elsevier BV; Volume: 85; Issue: 2 Linguagem: Inglês
10.1038/ki.2013.344
ISSN1523-1755
AutoresHassan Izzedine, M. Mangier, Virginie Ory, Shaoyu Zhang, Kélhia Sendeyo, Khedidja Bouachi, Vincent Audard, Christine Péchoux, Jean Charles Soria, Christophe Massard, Rastilav Bahleda, Edward Bourry, David Khayat, A Baumelou, Philippe Lang, Mario Ollero, André Pawlak, Dil Sahali,
Tópico(s)Coagulation, Bradykinin, Polyphosphates, and Angioedema
ResumoRenal toxicity constitutes a dose-limiting side effect of anticancer therapies targeting vascular endothelial growth factor (VEGF). In order to study this further, we followed up 29 patients receiving this treatment, who experienced proteinuria, hypertension, and/or renal insufficiency. Eight developed minimal change nephropathy/focal segmental glomerulopathy (MCN/FSG)–like lesions and 13 developed thrombotic microangiopathy (TMA). Patients receiving receptor tyrosine kinase inhibitors (RTKIs) mainly developed MCN/FSG-like lesions, whereas TMA complicated anti-VEGF therapy. There were no mutations in factor H, factor I, or membrane cofactor protein of the complement alternative pathway, while plasma ADAMTS13 activity persisted and anti-ADAMTS13 antibodies were undetectable in patients with TMA. Glomerular VEGF expression was undetectable in TMA and decreased in MCN/FSG. Glomeruli from patients with TMA displayed a high abundance of RelA in endothelial cells and in the podocyte nuclei, but c-mip was not detected. Conversely, MCN/FSG-like lesions exhibited a high abundance of c-mip, whereas RelA was scarcely detected. RelA binds in vivo to the c-mip promoter and prevents its transcriptional activation, whereas RelA knockdown releases c-mip activation. The RTKI sorafenib inhibited RelA activity, which then promoted c-mip expression. Thus, our results suggest that c-mip and RelA define two distinct types of renal damage associated with VEGF-targeted therapies. Renal toxicity constitutes a dose-limiting side effect of anticancer therapies targeting vascular endothelial growth factor (VEGF). In order to study this further, we followed up 29 patients receiving this treatment, who experienced proteinuria, hypertension, and/or renal insufficiency. Eight developed minimal change nephropathy/focal segmental glomerulopathy (MCN/FSG)–like lesions and 13 developed thrombotic microangiopathy (TMA). Patients receiving receptor tyrosine kinase inhibitors (RTKIs) mainly developed MCN/FSG-like lesions, whereas TMA complicated anti-VEGF therapy. There were no mutations in factor H, factor I, or membrane cofactor protein of the complement alternative pathway, while plasma ADAMTS13 activity persisted and anti-ADAMTS13 antibodies were undetectable in patients with TMA. Glomerular VEGF expression was undetectable in TMA and decreased in MCN/FSG. Glomeruli from patients with TMA displayed a high abundance of RelA in endothelial cells and in the podocyte nuclei, but c-mip was not detected. Conversely, MCN/FSG-like lesions exhibited a high abundance of c-mip, whereas RelA was scarcely detected. RelA binds in vivo to the c-mip promoter and prevents its transcriptional activation, whereas RelA knockdown releases c-mip activation. The RTKI sorafenib inhibited RelA activity, which then promoted c-mip expression. Thus, our results suggest that c-mip and RelA define two distinct types of renal damage associated with VEGF-targeted therapies. In renal glomeruli, podocytes express vascular endothelial growth factor (VEGF), whereas VEGF receptor tyrosine kinases (RTKs) are expressed by both podocytes and glomerular endothelial cells.1.Muller-Deile J. Worthmann K. Saleem M. et al.The balance of autocrine VEGF-A and VEGF-C determines podocyte survival.Am J Physiol Renal Physiol. 2009; 297: F1656-F1667Crossref PubMed Scopus (53) Google Scholar The biological functions of VEGF are mediated by its binding to one of the VEGF receptors (VEGFRs), which include VEGFR1 (Flt-1), VEGFR2 (KDR/Flk-1), and VEGFR-3 (Flt-4). The VEGF family comprises seven members: VEGF-A, -B, -C, -D, -E, and placenta growth factor 1 and 2. VEGF-A (also referred to as VEGF) binds to VEGFR1 and -2, whereas VEGF-C and -D bind to VEGFR2 and VEGFR-3, respectively. VEGFR2 expression has been reported in cultured podocytes.2.Guan F. Villegas G. Teichman J. et al.Autocrine VEGF-A system in podocytes regulates podocin and its interaction with CD2AP.Am J Physiol Renal Physiol. 2006; 291: F422-F428Crossref PubMed Scopus (95) Google Scholar Structurally, RTKs consist of an extracellular ligand–binding domain, a transmembrane region, and an intracellular kinase domain that mediates downstream signal transduction. Upon binding to their ligand, RTKs dimerize and are phosphorylated on their kinase domain, leading to the recruitment of adaptor proteins that trigger intracellular signaling cascades important for processes such as cell proliferation and survival, migration, and metabolism.3.Ivy S.P. Wick J.Y. Kaufman B.M. An overview of small-molecule inhibitors of VEGFR signaling.Nat Rev Clin Oncol. 2009; 6: 569-579Crossref PubMed Scopus (282) Google Scholar Dysregulation of RTK signaling by mutation or by ectopic receptor or ligand overproduction has been implicated in several aspects of tumor progression, including cell proliferation, survival, angiogenesis, and tumor dissemination.4.Kerbel R.S. Tumor angiogenesis.N Engl J Med. 2008; 358: 2039-2049Crossref PubMed Scopus (1892) Google Scholar VEGFR2 is the predominant receptor in angiogenic signaling.5.Maharaj A.S. Saint-Geniez M. Maldonado A.E. et al.Vascular endothelial growth factor localization in the adult.Am J Pathol. 2006; 168: 639-648Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar,6.Neufeld G. Cohen T. Gengrinovitch S. et al.Vascular endothelial growth factor (VEGF) and its receptors.FASEB J. 1999; 13: 9-22Crossref PubMed Scopus (3142) Google Scholar VEGF is upregulated in response to hypoxia, oncogenes, or cytokines, and its expression is associated with poor prognosis in several types of cancer.7.Shweiki D. Itin A. Soffer D. et al.Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis.Nature. 1992; 359: 843-845Crossref PubMed Scopus (4155) Google Scholar,8.Heinrich M.C. Corless C.L. Duensing A. et al.PDGFRA activating mutations in gastrointestinal stromal tumors.Science. 2003; 299: 708-710Crossref PubMed Scopus (1998) Google Scholar Experimental, preclinical, and clinical studies have identified angiogenesis as a key process in the progression of most solid tumors. Thus, inhibition of VEGF and platelet-derived growth factor signaling is predicted to lead to anti-angiogenic effects and prevent the progression of tumors.9.Hurwitz H. Fehrenbacher L. Novotny W. et al.Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.N Engl J Med. 2004; 350: 2335-2342Crossref PubMed Scopus (9167) Google Scholar,10.Waller C.F. Imatinib mesylate.Recent Results Cancer Res. 2010; 184: 3-20Crossref PubMed Scopus (63) Google Scholar Therapeutic approaches targeting the VEGF ligand or RTK inhibitors (RTKIs) have recently been developed. Several antagonists of VEGF signaling are being tested in clinical trials, including bevacizumab (anti-VEGF monoclonal antibody) and RTKIs such as sunitinib, imatinib, and sorafenib.11.Homsi J. Daud A.I. Spectrum of activity and mechanism of action of VEGF/PDGF inhibitors.Cancer Control. 2007; 14: 285-294Crossref PubMed Scopus (120) Google Scholar Although RTKIs are widely used as inhibitors of VEGFRs, they interfere with the activity of other growth factors, such as platelet-derived growth factor receptors, stem cell factor receptor (c-kit), FMS-like tyrosine kinase-3 (Flt-3), b-raf, and Bcl-Abl. Thus, they are commonly named as multitargeted RTKIs and are widely used in medical oncology practice. However, renal complications constitute a dose-limiting side effect of RTKI and anti-VEGF therapies. We report here on a series of 29 patients treated with anti-VEGF and RTKIs, who experienced proteinuria, hypertension, and/or renal insufficiency. Immunomorphological and molecular studies suggest that RelA and c-mip define two separate glomerular damages associated with anti-angiogenic drugs, based on two distinct pathophysiological mechanisms. Baseline patient characteristics are summarized in Table 1. All patients were referred to a nephrology department because of discovery of proteinuria and/or increased serum creatinine following anti-VEGF initiating treatment. Sixteen patients had renal cell carcinoma and received 50mg of sunitinib daily (n=11), 400mg of sorafenib (n=3) or 5mg axitinib (n=1) twice daily, or bevacizumab (10mg/kg/dose) twice monthly (n=2). Thirteen other patients received VEGF Trap (4–6mg/kg/dose) every 3 weeks (for ovarian, adrenal, breast, prostate, esophageal, and rectal cancers) or bevacizumab 10mg/kg/dose twice monthly (for lung, uterine, and colorectal cancers). At the time of kidney biopsy, 20 patients (68,9%) presented with hypertension requiring antihypertensive treatment, and 14 (48,3%) with renal failure defined by a creatinine clearance rate below 60ml/min per 1.73m2. Kidney biopsy was performed 2–12 weeks after the start of anti-VEGF treatment. The average number of glomeruli was 18.2 (range: 9–50). The principal pathological findings were thrombotic microangiopathy (TMA; n=13), minimal change nephropathy/focal segmental glomerulopathy (MCN/FSG)–like syndromes (n=8), acute tubular necrosis (n=3), and one case of each of the following lesions: membranous nephropathy, IgA nephropathy, anti-neutrophil cytoplasmic antibody-negative pauci-immune crescentic glomerulonephritis, diabetic nephropathy, and acute interstitial nephritis. Fourteen patients (56%) died during the study because of cancer progression. Owing to the expected limited survival, managing oncologists were reluctant to repeat the urinary investigations.Table 1Patient characteristics at baselineAll patientsMCN/FSGTMAn29813Men1963Age, years, mean (range)55.2 (20–79)71.5 (37–79)69.5 (20–67)mRCC1773Previous nephrectomy1463Previous radiotherapy and/or IFNα use1254Anti-VEGF agents Bevacizumab (cumulative dose)9-6 (10–240mg/kg) VEGF Trap (cumulative dose)6-5 (12–54mg/kg) Sunitinib (cumulative dose)115 (50–200mg)2 (100mg) Axitinib11 (40mg)- Sorafenib (cumulative dose)32 (800mg)-Renal parameters SBP, mmHg, mean (range)150.0 (110–190)130.0 (110–180)165.0 (120–190) DBP, mmHg, mean (range)95.0 (60–115)83.0 (60–110)100.0 (80–115) Proteinuria, g/day, mean (range)3.50 (0.6–19.5)3.5 (2–5.5)4.11 (0.6–19.5) Edema145 (62.5%)4(31%) Microhematuria122 (28.5%)6 (46%) SCr, mg/dl, mean (range)1.26 (0.70–4.44)0.95 (0.79–1.27)0.86 (0.70–1.45) aMDRD CrCl, ml/min per 1.73m2, mean (range)76.5 (13.7–120)80 (17–102)75 (45–120)Outcome Follow-up duration1 mo–3 years3 years6 mo–2 years Alive1418 SCr, mg/dl1.0 (0.75–2.1)1.200.95 (0.75–2.1)Abbreviations: aMDRD CrCl, creatinine clearance; DBP, diastolic blood pressure; IFNα, interferon-α; MCN/FSG, minimal change nephropathy/focal segmental glomerulopathy; mo, month; mRCC, metastatic renal cell carcinoma; SBP, systolic blood pressure; SCr, serum creatinine; TMA, thrombotic microangiopathy; VEGF, vascular endothelial growth factor. Open table in a new tab Abbreviations: aMDRD CrCl, creatinine clearance; DBP, diastolic blood pressure; IFNα, interferon-α; MCN/FSG, minimal change nephropathy/focal segmental glomerulopathy; mo, month; mRCC, metastatic renal cell carcinoma; SBP, systolic blood pressure; SCr, serum creatinine; TMA, thrombotic microangiopathy; VEGF, vascular endothelial growth factor. Eight patients with a previous nephrectomy for metastatic renal cell carcinoma and/or interferon-α therapy exhibited MCN/FSG-like lesions, defined by nephrotic proteinuria (with or without hypoalbuminemia), whereas light microscopy examination showed either normal glomeruli or focal and segmental glomerulosclerosis (FSGS) lesions. These patients had received sunitinib (5 patients), sorafenib (2 patient), and Axitinib (1 patient). The mean interval between initiation of RTKI therapy and onset of proteinuria was 65±50.5 days (range 14–180). All patients exhibited heavy proteinuria, but only two exhibited hypertension and acute renal failure. Anti-VEGF therapy was discontinued in all patients, and symptomatic treatments such as angiotensin-converting enzyme inhibitors or angiotensin-2 receptor antagonists were started. Six patients died shortly thereafter because of cancer progression. Thirteen patients presented with renal TMA. Platinum derivatives (n=3), gemcitabine (n=1), or radiotherapy (n=1) were administered before anti-VEGF agents. No clinical or biological signs of TMA were observed during gemcitabine treatment, which was given 6 months before the introduction of anti-VEGF medications. The mean interval between initiation of anti-VEGF therapy and the onset of TMA was 5.75±3.6 months (range 2–12). Anti-VEGF agents included bevacizumab (between 2 and 24 doses in six patients), VEGF Trap (after 2, 3, 4, 8, and 9 cycles in five patients, respectively), and sunitinib (after 2 cycles in two patients). All patients showed significant proteinuria (including eight with nephrotic syndrome), whereas eight had hypertension. Two patients developed acute renal failure and six displayed microhematuria. Fifty percent of patients who developed TMA displayed hematological manifestations. Anemia and thrombocytopenia were present in 10 patients, whereas plasma haptoglobin was decreased in 6 patients. Schistocytes were present in nine patients. Renal biopsies revealed TMA in all patients, characterized by glomerular capillary thrombosis associated with mesangiolysis and double contours. All patients exhibited normal complement proteins. No constitutional abnormalities or heterozygous missense mutations were found in factor H, factor I, or membrane cofactor protein—the three major regulatory proteins of the complement alternative pathway. Plasma ADAMTS13 activity was above 20%. Acquired or constitutive anti-ADAMTS13 antibodies were undetectable. Different symptomatic treatments were tried, including plasmapheresis and fresh frozen plasma (seven patients), as well as steroid and antihypertensive drugs. Renal and hematological symptoms were improved 1–6 months after the discontinuation of anti-VEGF agents. Five patients died because of cancer progression. One patient switched to another anti-VEGF agent (from VEGF Trap to bevacizumab) and displayed an absence of proteinuria and stable renal function 2 years later. Another patient continued bevacizumab in association with antihypertensive drugs for 8 months despite persistent proteinuria, but renal function remained stable. The relative abundance of VEGF was reduced in podocytes from patients with MCN/FSG when compared with control kidney tissues, and was undetectable in TMA (Figure 1a), as previously reported.12.Eremina V. Jefferson J.A. Kowalewska J. et al.VEGF inhibition and renal thrombotic microangiopathy.N Engl J Med. 2008; 358: 1129-1136Crossref PubMed Scopus (1159) Google Scholar Quantification of the relative abundance of c-mip, RelA, hypoxia-inducible factor 1α (HIF-1α), and Tie2, as well as their expression pattern, were analyzed in all glomeruli of MCN/FSG/TMA biopsies related to anti-VEGF therapy and compared with idiopathic forms (MCN syndrome (MCNS), FSGS, and TMA) (Figure 1 and Supplementary Figure S1 online). Five biopsies corresponding to each pathological condition were assessed and the data are depicted in Figure 1f. The relative abundance of c-mip was greatly increased in MCN/FSG-like lesions, but no significant difference was observed with idiopathic MCNS and FSGS diseases, whereas it was scarcely, or not, detected in TMA or in control human kidneys (Figure 1b and f). HIF-1α is an oxygen-sensitive transcription factor that is recruited in response to oxygen deprivation. We reasoned that ischemic conditions prevail in TMA and would result in a loss of oxygen delivery to tissues, leading to HIF-1α activation. In biopsies with MCN/FSG-like lesions, HIF-1α expression was increased as compared with control human kidney, but higher abundance was observed in TMA glomeruli without significant difference between idiopathic and VEGF forms (Figure 1c and f). Idiopathic MCNS and FSGS are considered as noninflammatory glomerular diseases.13.Mathieson P.W. Minimal change nephropathy and focal segmental glomerulosclerosis.Semin Immunopathol. 2007; 29: 415-426Crossref PubMed Scopus (46) Google Scholar We investigated whether, unlike in idiopathic forms, inflammation may occur in MCN/FSG-like syndrome. Therefore, we analyzed the expression of RelA, a master nuclear factor-κB (NF-κB) transcription factor, which controls many inflammatory genes.14.Sanz A.B. Sanchez-Nino M.D. Ramos A.M. et al.NF-kappaB in renal inflammation.J Am Soc Nephrol. 2010; 21: 1254-1262Crossref PubMed Scopus (434) Google Scholar The relative abundance of RelA was significantly higher in TMA than in control human kidneys, whereas it was scarcely detected in MCN/FSG-like lesions (Figure 1d and f). Subtle differences in RelA abundance and distribution were observed between anti-VEGF-induced TMA and the idiopathic forms: RelA was uniformly increased in idiopathic TMA glomeruli, but higher abundance was observed in anti-VEGF therapy–related TMA glomeruli (Figure 1f) with a more restricted distribution to some capillary loops (Figure 1d and Supplementary Figure S1 online). On the other hand, RelA abundance was clearly reduced in idiopathic MCNS and FSGS, as well as in MCN/FSG-like lesions, as compared with control human kidneys, without a significant difference between idiopathic forms and anti-VEGF-related lesions (Figure 1f). The relative abundance of Tie2 was significantly increased after anti-VEGF or RTKI therapies (Figure 1e and f). Download .pdf (.11 MB) Help with pdf files Suppplementary Figure S1 Confocal microscopy analysis showed that nephrin expression was significantly reduced and exhibited a granular pattern in MCN/FSG relative to control human kidneys (Figure 2). In contrast, nephrin displayed variable expression and an irregular pattern in TMA glomeruli consisting of a lack of detection in some areas and preservation in others. In control human kidneys, RelA was located on the external side of the specific endothelial cell marker Tie2, showing no colocalization, which suggests that RelA was only expressed in podocytes (Figure 3). The relative abundance of RelA was markedly increased in TMA glomeruli, not only in podocytes but also in endothelial cells, as shown by RelA/nephrin (Figure 2) and RelA/Tie2 (Figure 3) double labeling, respectively. Tie2 immunostaining showed a slight increase in MCN/FSG biopsies compared with control kidneys. Tie2 expression was strongly increased in TMA with a diffuse, irregular pattern within the capillary loops (Figure 3). These results suggest that in MCN/FSG-like syndromes associated with RTKI therapy glomerular lesions affect podocytes almost exclusively, whereas in TMA alterations mainly affect endothelial cells and podocyte injury may be a secondary event. Significant changes were also observed in small arterioles in TMA biopsies, where endothelial cells were swollen and exhibited higher abundance of RelA (Figure 4) when compared with endothelial cells from normal arterioles. Moreover, arteriolar walls were infiltrated by pericytes that exhibited a high amount of RelA. These changes were not observed in control human kidneys.Figure 3Differential expression of RelA and Tie2 in thrombotic microangiopathy (TMA) and minimal change nephropathy (MCN)-like lesions. Confocal microscopy analysis of Tie2 (red) and RelA (green) expression in control glomeruli (Con), MCN-like lesions (MCN), and TMA biopsies. The abundance of RelA and Tie2 is significantly increased in TMA. Tie2 and RelA are colocalized in damaged areas within TMA glomeruli, whereas they are expressed in different cell compartments in control glomeruli. Bars = 10μm.View Large Image Figure ViewerDownload (PPT)Figure 4Overproduction of RelA in endothelials cells in thrombotic microangiopathy (TMA). Confocal microscopy analysis of Tie2 (red) and RelA (green) expression in arterioles of control human kidneys (Con) and TMA biopsies. The abundance of RelA and Tie2 is significantly increased in TMA arterioles, which display a swelling of endothelial cells. The abundance of RelA is also increased in the pericytes of TMA arterioles. Bars = 10μm. The relative abundance of RelA was assessed by quantifying the specific arteriolar fluorescence intensity in three-dimensional stacks of images taken by confocal microscopy and normalized to total arteriolar area. Five samples were analyzed in each condition (Con and TMA). Data represent the mean±s.e.m. (*P<0.05, Mann–Whitney test).View Large Image Figure ViewerDownload (PPT) We studied ultrastructural alterations in glomeruli of patients with bevacizumab-induced TMA and in control kidneys (n=3 each). Transmission electron microscopy analysis showed major alterations in TMA glomeruli, including duplication of glomerular basement membrane, loss of fenestrations, detachment of endothelial cells from original basement membrane, interposition of cells, and marked effacement of visceral epithelial cell foot processes in some areas. Some podocytes exhibited cytoplasm vacuolization, as well as endoplasmic reticulum enlargement and mitochondrial swelling, suggesting an underlying apoptotic process (Figure 5a). To provide accurate determination of RelA increase in bevacizumab-induced TMA, we performed immunogold labeling and quantified the number of RelA gold particles in podocytes and in endothelial glomerular cells. In TMA kidney biopsies, the number of RelA gold particles was strongly increased in nuclear podocytes, as well as in endothelial glomerular cells, as compared with control kidney (P<0.001), whereas no significant difference was detected in the cytoplasm of podocytes (Figure 5b and c). Given the increased abundance of RelA in TMA contrasting with the virtual absence of c-mip, we hypothesized that c-mip expression was repressed at the transcriptional level by NF-κB. NF-κB heterodimers bind to 10-bp κB DNA sites that exhibit the consensus sequence (5′-GGGRNWYYCC-3′), which comprises a constant core and a number of variable nucleotides (R: A or G; N: any nucleotide; W: A or T; Y: C or T). The sequence identified on the human c-mip promoter (c-mip-κB, 5′-GGGGCTGCCC-3′) at position -199 to -214 (+1 corresponds to the transcriptional initiation site) fulfils these criteria (Figure 6a). The mouse κB response element is identical to that of humans, except for the substitution of the T-nucleotide by a C-nucleotide (5′-GGGGCCGCCC-3′). To assess whether NF-κB binds to this sequence in vivo, we performed chromatin immunoprecipitation assay using mouse podocytes. The short region flanking the κB response element was precipitated by the anti-RelA antibody, but not by the rabbit IgG control (Figure 6b). Incubation of nuclear extracts from HEK cells overexpressing RelA with the radiolabeled c-mip κB response element oligonucleotide produced only one specific band shift in electrophoretic mobility shift assay (Figure 6c). The DNA-binding shift appeared to be specific, as it was abolished by co-incubation with a mutated probe and shifted upward in the presence of the anti-RelA antibody. In contrast, the lower band shifts observed with the mutated c-mip-κB seemed to be nonspecific, as they were not altered by preincubation of nuclear extracts with the anti-RelA antibody. This result led us to study the effects of NF-κB on the transcriptional activation of c-mip. The 877-bp full-length sequence containing the entire c-mip proximal promoter was ligated upstream of the luciferase gene and cotransfected with NF-κB p50, p65/RelA, or both, as well as with the empty vector. Protein lysates were prepared 24h after transfection. Luciferase activity driven by the c-mip promoter was strongly reduced in the presence of RelA (Figure 6d). Interestingly, overexpression of p50 alone did not inhibit the transcriptional activity of the c-mip promoter. Moreover, cotransfection of RelA with p50 significantly reduced the inhibition of luciferase activity induced by RelA alone. These results suggest that RelA binds to the c-mip promoter and exerts a powerful inhibitory effect on c-mip transcription. A strong argument supporting this hypothesis came from the study of wild-type and RelA-deficient mouse embryonic fibroblasts (MEFs). Figure 7 shows that c-mip was markedly increased in RelA-deficient MEFs, in the absence of any treatment, whereas it was barely detected in wild-type MEFs.Figure 7Overproduction of c-mip in RelA-deficient cells. Relative expression of RelA and c-mip transcripts on total RNA from wild-type and RelA-deficient mouse embryonic fibroblast (MEF). Bars represent the mean value of five independent experiments with error bars indicating s.e.m. (***P<0.001, Mann–Whitney test).View Large Image Figure ViewerDownload (PPT) We investigated whether overproduction of c-mip in MCN/FSG-like lesions is a secondary molecular event or a direct effect of RTKIs. At first, we tested whether sorafenib affects c-mip expression in wild-type and RelA-deficient MEFs. The abundance of c-mip is remarkably higher in RelA-deficient MEFs as compared with wild-type MEFs (Figure 8a). However, preincubation of cells with sorafenib increased the basal amount of c-mip, both in wild-type and RelA-deficient MEFs. We then treated a podocyte cell line with sorafenib and analyzed c-mip expression by quantitative PCR. The abundance of transcript was significantly increased in podocytes treated with sorafenib when compared with those incubated with the vehicle only (ethanol; Figure 8b). Confocal microscopy analysis showed that sorafenib induces profound disorganization of F-actin cytoskeleton with depletion of stress fibers and abnormal production of enclosed F-actin-rich membrane structures (Figure 8c). As VEGFR is also expressed by lymphocytes, we tested whether sorafenib affects c-mip expression in these immune cells. We purified lymphocytes from normal donors by cell gradient density and incubated them with the same concentration of sorafenib (10μmol/l). Interestingly, c-mip was also increased in normal lymphocytes, suggesting that the induction of c-mip by sorafenib was not restricted to podocytes (Figure 8d). In resting cells, RelA is mostly sequestered in the cytoplasm compartment by its inhibitor IκBα. Stimulation of cells by NF-κB inducers such as cytokines induces phosphorylation of IκBα at serine (Ser) residues 32–36, followed by its ubiquitination and proteasome degradation. RelA released from its inhibitor is phosphorylated and moves into the nucleus, where it promotes transcriptional activation of target genes.15.Nowak D.E. Tian B. Jamaluddin M. et al.RelA Ser276 phosphorylation is required for activation of a subset of NF-kappaB-dependent genes by recruiting cyclin-dependent kinase 9/cyclin T1 complexes.Mol Cell Biol. 2008; 28: 3623-3638Crossref PubMed Scopus (141) Google Scholar Phosphorylation of RelA at serine 276 has a crucial role in NF-κB transcriptional activity.16.Zhong H. Voll R.E. Ghosh S. Phosphorylation of NF-kappa B p65 by PKA stimulates transcriptional activity by promoting a novel bivalent interaction with the coactivator CBP/p300.Mol Cell. 1998; 1: 661-671Abstract Full Text Full Text PDF PubMed Scopus (1023) Google Scholar Inhibition of RelA phosphorylation at serine 276 has been implicated in the transactivation inhibition of NF-κB-dependent genes.17.Arun P. Brown M.S. Ehsanian R. et al.Nuclear NF-kappaB p65 phosphorylation at serine 276 by protein kinase A contributes to the malignant phenotype of head and neck cancer.Clin Cancer Res. 2009; 15: 5974-5984Crossref PubMed Scopus (51) Google Scholar As RelA inhibits the transcriptional activation of c-mip, we checked whether c-mip upregulation in MCN/FSG results from direct inhibition of RelA activity by RTKI therapy. Therefore, we treated the podocyte cell line with sorafenib and analyzed the activation status of RelA and its subcellular localization, as well as the stability of IκBα. In the absence of sorafenib, phospho-ser32/36 IκBα was detected, along with phospho-ser276 RelA, suggesting that NF-κB is constitutively activated in podocytes (Figure 9a). Conversely, Sorafenib blocked RelA phosphorylation, whereas IκBα phosphorylation was significantly reduced. Immunoblotting of the cytoplasm and nuclear podocyte extracts showed that Sorafenib induces accumulation of RelA in the cytoplasm, whereas very low abundance of RelA was detected in nuclear extracts (Figure 9b). In addition, immunoprecipitation experiments showed RelA to be mostly sequestered with IκBα in podocytes incubated with Sorafenib (Figure 9c). Confocal microscopy analysis showed that RelA was diffusely expressed in both the cytoplasm and the nuclear compartment in untreated cells, whereas its expression was mostly restricted to the cytoplasm in sorafenib-treated cells (Figure 9d). These results suggest that sorafenib inhibits NF-κB activity and indirectly promotes c-mip transcriptional activation. Anti-VEGF therapy leads to various glomerular injuries, including MCN/FSG- and TMA-like syndromes. In our group of patients, we show for the first time that MCN/FSG lesions, which are mostly observed following RTKI therapy, are associated with a high abundance of c-mip. In contrast, in TMA resulting from anti-VEGF therapy, c-mip is not detected, whereas RelA is produced at high levels by podocytes and glomerular endothelial cells. We provide evidence that RelA binds to the c-mip promoter in vivo and represses its transcription, whereas RelA knockdown releases this inhibition. These results may account fo
Referência(s)