Semaphorin 3B–associated membranous nephropathy is a distinct type of disease predominantly present in pediatric patients

Membranous nephropathy results from subepithelial antigen-antibody complex deposition along the glomerular basement membrane. Although PLA2R, THSD7A, and NELL-1 account for a majority (about 80%) of the target antigens, the target antigen in the remaining cases is not known. Using laser microdissection of PLA2R-negative glomeruli of patients with membranous nephropathy followed by mass spectrometry we identified a unique protein, Semaphorin 3B, in three cases. Mass spectrometry failed to detect Semaphorin-3B in 23 PLA2R-associated cases of membranous nephropathy and 88 controls. Semaphorin 3B in all three cases was localized to granular deposits along the glomerular basement membrane by immunohistochemistry. Next, an additional eight cases of Semaphorin 3B-associated membranous nephropathy were identified in three validation cohorts by immunofluorescence microscopy. In four of 11 cases, kidney biopsy also showed tubular basement membrane deposits of IgG on frozen sections. Confocal microscopy showed that both IgG and Semaphorin 3B co-localized to the glomerular basement membrane. Western blot analysis of five available sera showed reactivity to reduced Semaphorin 3B in four of four patients with active disease and no reactivity in one patient in clinical remission; there was also no reactivity in control sera. Eight of the 11 cases of Semaphorin 3B-associated membranous nephropathy were pediatric cases. Furthermore, in five cases, the disease started at or below the age of two. Thus, Semaphorin 3B-associated membranous nephropathy appears to be a distinct type of disease; more likely to be present in pediatric patients. Membranous nephropathy results from subepithelial antigen-antibody complex deposition along the glomerular basement membrane. Although PLA2R, THSD7A, and NELL-1 account for a majority (about 80%) of the target antigens, the target antigen in the remaining cases is not known. Using laser microdissection of PLA2R-negative glomeruli of patients with membranous nephropathy followed by mass spectrometry we identified a unique protein, Semaphorin 3B, in three cases. Mass spectrometry failed to detect Semaphorin-3B in 23 PLA2R-associated cases of membranous nephropathy and 88 controls. Semaphorin 3B in all three cases was localized to granular deposits along the glomerular basement membrane by immunohistochemistry. Next, an additional eight cases of Semaphorin 3B-associated membranous nephropathy were identified in three validation cohorts by immunofluorescence microscopy. In four of 11 cases, kidney biopsy also showed tubular basement membrane deposits of IgG on frozen sections. Confocal microscopy showed that both IgG and Semaphorin 3B co-localized to the glomerular basement membrane. Western blot analysis of five available sera showed reactivity to reduced Semaphorin 3B in four of four patients with active disease and no reactivity in one patient in clinical remission; there was also no reactivity in control sera. Eight of the 11 cases of Semaphorin 3B-associated membranous nephropathy were pediatric cases. Furthermore, in five cases, the disease started at or below the age of two. Thus, Semaphorin 3B-associated membranous nephropathy appears to be a distinct type of disease; more likely to be present in pediatric patients. Membranous nephropathy (MN) results from antibodies targeting an antigen in the glomerular basement membrane (GBM). MN is typically classified as primary MN, which has no identifiable underlying disease association, and secondary MN, where the MN may be associated with an autoimmune disease, infection, malignancy, and others.1Beck Jr., L.H. Salant D.J. Membranous nephropathy: from models to man.J Clin Invest. 2014; 124: 2307-2314Crossref PubMed Scopus (102) Google Scholar, 2Ronco P. Debiec H. Pathophysiological advances in membranous nephropathy: time for a shift in patient's care.Lancet. 2015; 385: 1983-1992Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar, 3Couser W.G. Primary membranous nephropathy.Clin J Am Soc Nephrol. 2017; 12: 983-997Crossref PubMed Scopus (165) Google Scholar The target antigen in primary MN has been identified as M-type phospholipase A2 receptor (PLA2R), thrombospondin type-1 domain-containing 7A (THSD7A), and the recently described neural epidermal growth factor like-1 protein (NELL-1), respectively.4Tomas N.M. Beck L.H. Meyer-Schwesinger C. et al.Thrombospondin type-1 domain-containing 7A in idiopathic membranous nephropathy.N Engl J Med. 2014; 371: 2277-2287Crossref PubMed Scopus (420) Google Scholar, 5Beck L.H. Bonegio R.G.B. Lambeau G.R. et al.M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy.N Engl J Med. 2009; 361: 11-21Crossref PubMed Scopus (1288) Google Scholar, 6Sethi S. Debiec H. Madden B. et al.Neural epidermal growth factor-like 1 protein (NELL-1) associated membranous nephropathy.Kidney Int. 2020; 97: 163-174Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar Exostosin 1/exostosin 2 (EXT1/EXT2) have also recently been identified as the putative antigen(s) in secondary (autoimmune) MN.7Sethi S. Madden B.J. Debiec H. et al.Exostosin 1/exostosin 2–associated membranous nephropathy.J Am Soc Nephrol. 2019; 30: 1123-1136Crossref PubMed Scopus (69) Google Scholar NELL-1 and EXT1/EXT2 were identified using laser microdissection of PLA2R-negative MN glomeruli followed by mass spectrometry. Using a similar approach, we sought to identify other novel antigen(s) in MN. We detected a unique protein Semaphorin 3B (Sema3B) by tandem mass spectrometry (MS/MS) in the glomeruli of 3 cases (patients 1-3) of MN (Figure 1). The counts ranged from 10 to 42 with an average total spectral count of 23.7 (SD, ±16.5). The average spectral counts of Sema3B were lower than PLA2R (86.1; SD, ±27.5), NELL-1 (63.1; SD, ±21.6), and EXT1/EXT2 (EXT1, 65.3 SD, ±34.6; EXT2, 83.4 SD, ±38.4) in PLA2R-, NELL-1–, and EXT1/EXT2-associated MN, respectively.6Sethi S. Debiec H. Madden B. et al.Neural epidermal growth factor-like 1 protein (NELL-1) associated membranous nephropathy.Kidney Int. 2020; 97: 163-174Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar,7Sethi S. Madden B.J. Debiec H. et al.Exostosin 1/exostosin 2–associated membranous nephropathy.J Am Soc Nephrol. 2019; 30: 1123-1136Crossref PubMed Scopus (69) Google Scholar However, the presence of Sema3B was unique in this subset of MN, and importantly, all control cases including 15 time 0 kidney transplant biopsies, 73 other glomerulopathies, and 23 PLA2R-positive MN cases were negative for Sema3B. The spectral counts of Sema3B in the 3 cases, along with a representative sequence coverage map of Sema3B, are shown in Figure 2. The MS/MS match from 1 case is shown in Supplementary Figure S1. None of the cases showed any spectral counts for THSD7A or NELL-1. Subsequently, we screened over 2500 kidney biopsies, including pediatric cases, from our amyloid laboratory database (over 12,500 data files; both amyloid and nonamyloid cases) and did not find any kidney biopsy that showed any spectral counts for Sema3B. We included both adult and pediatric cases of C3 glomerulopathy. On analysis of other organs, we found trace amounts of Sema3B (1-3 spectral counts) in the heart and liver of 23 cases (over 2000 nonkidney cases were screened for Sema3B). All 4 classes of Igs were detected in Sema3B-associated MN, with average spectral counts of IgG1 25.0 (SD, ±4.0), IgG2 22.7 (SD, ±4.5), IgG3 25.0 (SD, ±9.5), and IgG4 17.0 (SD, ±5.3). We performed immunohistochemical (IHC) staining for Sema3B in all 3 cases positive on MS/MS studies. All cases showed positive (2–3+/3) granular staining for Sema3B along the GBM. Importantly, there was no significant mesangial staining (Figure 3a). There was no staining along Bowman’s capsule, tubular basement membranes, or in vessel walls. The positive Sema3B granular staining mirrored the granular IgG along the GBM seen in each case. All 45 control cases were negative for Sema3B, which included 3 pediatric cases of minimal change disease/focal segmental glomerulosclerosis. Representative negative staining for Sema3B in diabetes, focal segmental glomerulosclerosis, IgA nephropathy, minimal change disease, and PLA2R-associated MN is shown in Figure 3b. Given the results obtained in the Mayo cohort showing 2 adults (patients 1 and 2) and 1 pediatric patient (patient 3) of Sema3B-associated MN, we analyzed 3 validation cohorts, 1 adult (French cohort 1) and 2 pediatric (Italian cohort, French cohort 2) cohorts (Figure 1). Sema3B was detected using immunofluorescence microscopy (IF) studies on paraffin sections after antigen retrieval. Patient 4 of the validation cohort had both IHC and IF studies performed. Two adult cases (patients 4 and 5) of 59 PLA2R-, THSD7A-, EXT1/EXT2-, and NELL-1–negative primary MN were positive for Sema3B. One case (patient 4) was detected by IHC at the Mayo Clinic and was then confirmed by IF studies on paraffin sections at Tenon Hospital (Figure 4a and b). Patient 5 had 3 biopsies performed at the age of 1, 6, and 19 years, showing granular staining for Sema3B along the GBM (biopsy at age 1 and 19 shown in Figure 4c and d and Supplementary Figure S3E and F). It became clear that Sema3B-associated MN was enriched in pediatric patients. After this unique finding, we then screened for Sema3B in a larger cohort of MN in the pediatric age group (Italian cohort and French cohort 2). Four cases (patients 6, 7, 8, and 9) of 43 pediatric cases negative for PLA2R, THSD7A, EXT1/EXT2, and NELL-1 showed bright granular staining for Sema3B along the GBM (Figure 4e–h). Of note, patients 7 and 8 were siblings. Of the 43 cases, 6 had lupus class V MN, including 3 with class III + V, and were all negative for Sema3B. An additional 2 cases (patients 10 and 11) of 16 pediatric cases negative for PLA2R, THSD7A, EXT1/EXT2, and NELL-1 showed the bright granular deposits of Sema3B along the GBM (Figure 4i and j). Of the 16 cases, 9 had lupus class V MN and were all negative for Sema3B. Figure 4 shows the granular GBM staining and absence of mesangial or Bowman capsule staining in all cases. An adult and pediatric case of Sema3B-negative MN is also shown (Figure 4k and l). We performed confocal IF to determine whether Sema3B and IgG co-localized along the GBM (Figure 5). Bright granular staining for Sema3B (green) and IgG (red) was seen along the GBM (Figure 5a and b). Furthermore, superimposition of the 2 signals showed co-localization resulting in a yellow signal (Figures 5c). Laser quantitative analysis (Figure 5d and e) also confirmed the co-localization of Sema3B and IgG, further corroborating that the subepithelial deposits contained both Sema3B and IgG. A second case (patient 6, 2 years old) is shown in Supplementary Figure S2. Western blot analyses in nonreducing and reducing conditions were performed using human Sema3B full-length recombinant proteins to determine the presence of circulating anti-Sema3B antibodies in the available serum of 5 patients (patients 8, and 11) (Figure Four of the 5 patients showed reactivity Sema3B reducing not nonreducing A was seen at the full-length serum of patient was positive for anti-Sema3B were also available for analysis of with of antibodies after Patient was after a was negative by Western blot In sera from patients with conditions including PLA2R-, THSD7A-, and IgA focal segmental and minimal change disease and control did not reactivity with Sema3B analyzed We identified 11 cases of Sema3B-associated MN, 3 from the Mayo Clinic cohort, 2 from the French validation cohort 1, 4 from the validation cohort, and 2 from the French validation cohort 2 1). There were 7 and 4 patients. Of the 11 cases, 8 were pediatric patients years of and 3 were adult patients. Of the Mayo Clinic cohort (patients 1 was a pediatric patient of 4 pediatric PLA2R-negative MN cases for Sema3B, and 2 were adults of remaining primary and secondary PLA2R-negative MN including class V membranous lupus Of the validation cohorts and there were 8 pediatric patients of PLA2R-negative pediatric MN including case and 1 adult patient of PLA2R-negative adult MN patient In patient MN was at the age of 1 (biopsy performed of and biopsies performed 6 and years showed MN. of the pediatric cohorts included 9 patients with lupus MN class In 6 of pediatric cases were Sema3B in Semaphorin membranous protein nephropathy was at the age of 1 (biopsy performed of biopsy in and at age 19 1 and 9 2 to not Membranous nephropathy was at the age of 1 (biopsy performed of biopsy in and at age 19 in a not The average age of pediatric patients at disease was years (SD, and the average age of the 3 adult patients was years (SD, Two had of 1 in patient 10 and idiopathic with positive in patient The average serum was (SD, and was SD, biopsy in all cases showed of MN with microscopy showed granular IgG 2) and Sema3B deposits along the glomerular walls. 4 cases (patients 6, and all years of age at also showed tubular basement membrane staining for IgG in a granular (Figure 7, Supplementary Figure and In patient granular IgG deposits along tubular basement were in the biopsy 1 of However, the tubular basement membrane deposits were negative for Sema3B staining in all biopsies (Figure 7, Supplementary Figure and IgG in 4 cases in the Mayo Clinic cohort and 1 in French cohort 1) showed IgG1 in 3 cases and IgG1 and IgG4 in 1 case. IgG4 staining was in 4 cases of the cohort, and all 4 cases were negative for IgG4 microscopy was performed in all patients patients 4 and It showed glomerular subepithelial deposits in all cases that were as in cases. The deposits seen along tubular basement were confirmed or by microscopy in 3 cases (patients 6, and were in cases as biopsy in Semaphorin membranous basement membrane C3 C3 C3 C3 C3 C3 C3 C3 C3 IgA C3 and tubular in a and tubular 1 and 2 and with minimal at of and Patient 3 (Figures 6 and and and also the was Patient 4 was for 4 with of was 2 of at followed by a at 3 which is after 10 years no anti-Sema3B in Figure Patient 5 was at the age of 1 Figure and failed to to and from age 5 to 7 years was with that after a A second biopsy showed MN 2 with of granular tubular was and to 1 years of the of the patient was at the age of 19 years to of with a kidney biopsy (Figure 5 and Supplementary and was performed. The kidney biopsy showed MN 2 with deposits by IF and microscopy was then One the patient is in antibodies were not detected (Figure Patient 6 2 years of age with was at disease and and had MN. in 6 and to a with a of 7 and 8 are siblings. Patient 7 at age years with of 16 and was with of a after 2 a biopsy was that showed MN. The patient was started on an and was and at the of 4 The patient and has no at Patient 8 at age 2 years with and negative biopsy showed MN. the biopsy showed a IgA C3 of staining on the patient was with and followed by and was and after 2 The a 1) 2 years that showed MN on kidney biopsy that to The patient had a at 9 years of and kidney biopsy one in this confirmed MN. was with with on of and was to with years from was in with on and Patient 9 with of at 2 years of age and was with with after 4 of A biopsy was performed that showed MN. The patient after with and However, each time was and at age 7 years was with was in Patient 10 was to Patient 11 (Figure to kidney disease 3 years, to and MN is the of in On the other is a of in In the has to the of target MN including PLA2R, THSD7A, and putative N.M. Beck L.H. Meyer-Schwesinger C. et al.Thrombospondin type-1 domain-containing 7A in idiopathic membranous nephropathy.N Engl J Med. 2014; 371: 2277-2287Crossref PubMed Scopus (420) Google Scholar, 5Beck L.H. Bonegio R.G.B. Lambeau G.R. et al.M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy.N Engl J Med. 2009; 361: 11-21Crossref PubMed Scopus (1288) Google Scholar, 6Sethi S. Debiec H. Madden B. et al.Neural epidermal growth factor-like 1 protein (NELL-1) associated membranous nephropathy.Kidney Int. 2020; 97: 163-174Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, S. Madden B.J. Debiec H. et al.Exostosin 1/exostosin 2–associated membranous nephropathy.J Am Soc Nephrol. 2019; 30: 1123-1136Crossref PubMed Scopus (69) Google Scholar However, not account for all cases of MN, and in a significant of MN the target antigen is Using a of laser and IHC we a of PLA2R-negative MN to determine whether we identify novel target in this remaining group of MN. MS/MS studies confirmed spectral counts of NELL-1 and EXT1/EXT2 in a cohort of PLA2R-negative S. Debiec H. Madden B. et al.Neural epidermal growth factor-like 1 protein (NELL-1) associated membranous nephropathy.Kidney Int. 2020; 97: 163-174Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar,7Sethi S. Madden B.J. Debiec H. et al.Exostosin 1/exostosin 2–associated membranous nephropathy.J Am Soc Nephrol. 2019; 30: 1123-1136Crossref PubMed Scopus (69) Google Scholar Using the cohort, we then for other proteins that were unique to this group of PLA2R-negative MN. The to identify a putative from over to 2000 proteins detected on MS/MS were the of a unique protein and the absence spectral counts) of the protein in control cases and other cases of MN including both PLA2R-positive and MN. Using this approach, we found spectral counts of a unique protein, Sema3B, in 3 cases of MN. We confirmed the MS/MS by IHC and further our in 2 French cohorts and an Eight of 11 Sema3B-associated MN were pediatric of which 5 patients MN on or the age of 2 The remaining 3 patients MN at the of and Of the 3 adult cases the average age was years, which is lower than the primary MN age Thus, Sema3B-associated MN appears to pediatric patients and In the Mayo Clinic cohort 4 pediatric patients were of which 1 case was positive for Sema3B. However, the French and pediatric cohorts included a of patients and detected Sema3B in 6 of 59 pediatric MN patients 9 patients in each pediatric cohort had lupus class the of Sema3B-associated disease pediatric patients in and was 6 of It is that 5 of 8 patients with MN were years this Sema3B a target antigen in with the 2 patients had an autoimmune disease that included type 1 and One at disease had a immunofluorescence (patient 8 with no or of that was not present in kidney biopsies performed at The kidney biopsy of the glomeruli in Sema3B-associated MN were for MN. Sema3B was present along the subepithelial of the that Sema3B is from the than from the mesangial or there is no mesangial or staining for Sema3B. It is also likely that Sema3B not circulating antigen-antibody are more likely to be present in the of the GBM and have a along the the subepithelial Sema3B staining in our cases. IgG were in 4 cases and IgG1 in 3 cases and IgG4 along with IgG1 and IgG2 in 1 case. In IgG4 staining was in all 4 cases of the cohort, and all 4 cases were negative for to a IgG in Sema3B-associated MN, likely is from PLA2R-associated MN where IgG4 is the IgG tubular basement membrane deposits were present in 4 pediatric cases. All 4 cases had 2 years and showed tubular basement membrane deposits that for IgG. Of note, the tubular basement membrane deposits were negative for Sema3B glomerular deposits were for Sema3B in the sections (Figure 7, Supplementary Figure were and were negative in all cases. Thus, tubular basement membrane deposits in the of MN in a 2 years may a Sema3B-associated MN. The of Sema3B-associated MN in the pediatric age in in and the that in the cohort patients 7 and 8 are and patient 6 has a with MN the of a of the studies are to determine whether in the sequence of the protein or antigen are are a group of and proteins a Semaphorin of by et class 3 proteins and Int. Full Text Full Text PDF PubMed Scopus Google Scholar, and Semaphorin and Scopus Google Scholar, The PubMed Scopus Google Scholar The is the which The were identified as proteins that to then more than have been identified and are classified 8 and to are V, and are and is membrane classes and are In to the are also in other including and The primary are the and of 4 A and of 1 and 2. The of is in and in also have a in by H. for in the Full Text Full Text PDF PubMed Scopus Google Scholar Furthermore, the also a in disease including and and et and as of PubMed Scopus Google Scholar However, the and disease of Semaphorin 3 is a protein with a and a (Figure Semaphorin 3 and have been detected in and tubular et class 3 proteins and Int. Full Text Full Text PDF PubMed Scopus Google et Int. Full Text Full Text PDF PubMed Scopus Google Scholar In has been shown to proteins as and with protein and also et class 3 proteins and Int. Full Text Full Text PDF PubMed Scopus Google Scholar The and of Sema3B in the kidney is not and to the of our of Sema3B has not been in any kidney our results reactivity of Sema3B antibodies reducing that the is by of the Sema3B In conditions in of protein that as It be to determine the of Sema3B to the of change and to determine whether reduced conditions are for the antibodies to Sema3B the of the to It is to that the anti-Sema3B antibodies are to an on the or the detected sequence coverage of of the Sema3B protein (Figure this is a sera in 5 patients and IgG performed in 4 patients. There are also that studies to tubular basement membrane staining for IgG is present in the is the tubular basement membrane staining negative for the antigen in the tubular basement membrane deposits from Sema3B present in the MN the Sema3B-associated MN in pediatric patients with adult antibodies the recombinant protein were detected by Western blot in patients with active disease, and on clinical antibodies were present in MN patients were negative for and antibodies and not of a secondary et in membranous nephropathy and Am Soc Nephrol. PubMed Scopus (165) Google Scholar studies are to whether Sema3B is a antigen or a in Sema3B-associated MN. studies that the Sema3B on and studies that determine the of circulating antibodies to glomerular Sema3B. In Sema3B-associated MN a unique type of primary MN that is more likely to be present in pediatric in patients with years Sema3B-associated MN be to the of MN, including PLA2R-, THSD7A-, and MN. We performed MS/MS in kidney biopsies of cases of PLA2R-negative MN at the Mayo Clinic for analysis by MS/MS and detected the unique protein, Sema3B, in 2 cases. The cases included both primary and secondary membranous cases were also to detect recently novel EXT1/EXT2 and S. Debiec H. Madden B. et al.Neural epidermal growth factor-like 1 protein (NELL-1) associated membranous nephropathy.Kidney Int. 2020; 97: 163-174Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar,7Sethi S. Madden B.J. Debiec H. et al.Exostosin 1/exostosin 2–associated membranous nephropathy.J Am Soc Nephrol. 2019; 30: 1123-1136Crossref PubMed Scopus (69) Google Scholar IHC was then performed to the MS/MS In we screened PLA2R-negative MN cases the negative and 2 positive on by IHC for Sema3B. We detected 1 more positive which was confirmed by than 4 pediatric cases years of the remaining cases were all One biopsies were for Sema3B, of which 59 were from the French cohort 1, 43 from the pediatric cohort, and 16 from the French cohort 2. All cases were PLA2R-, THSD7A-, EXT1/EXT2-, and NELL-1–negative MN. pediatric cohort included 9 lupus class V cases. control cases, we performed MS/MS on cases that included 15 cases of time 0 transplant biopsies, cases of minimal change disease, cases of focal segmental glomerulosclerosis, 7 cases of glomerulosclerosis, 5 cases of IgA nephropathy, and 23 cases of PLA2R-positive MN. control IHC we 45 9 cases of minimal change disease, 9 cases of focal segmental glomerulosclerosis, 4 cases of IgA nephropathy, 7 cases of diabetes, 15 cases of PLA2R-positive MN, and 1 case of time 0 transplant biopsies were in the of and Mayo Clinic for and and IF including PLA2R and microscopy were performed in each case of MN. The clinical was obtained from the The was by the Mayo Clinic each case paraffin sections were obtained and on a membrane laser microdissection and using a the glomeruli were to to case. were with and for MS/MS The were identified by MS/MS using a mass to an All MS/MS were analyzed using and to a human was to and protein were at than by the with protein a and a using et for proteins by tandem mass PubMed Scopus Google Scholar and IHC staining was performed at the using the were at 5 and IHC staining was performed for Sema3B were for using 1 and in for 5 The Sema3B primary was to in and for 15 The was the primary and and was by 10 in and from the this were with were for 5 using and followed by in and is not the with the the was were from the and in for 5 were in of and in 3 of in staining was performed on sections for using target in The Sema3B primary was to in serum and and at with biopsy sections. Next, the were with secondary the IgG antibodies Next, IgG was with the after staining for Sema3B as described were in and with of Sema3B and IgG along the was by confocal microscopy using a and analyzed with Confocal Sema3B recombinant human protein was with nonreducing or reducing and for 10 were to and in were to to and then were with were at 4 with sera from control and antibodies Sema3B Subsequently, were and for 2 at with or IgG, proteins were with All the no We the Mayo Clinic of the Mayo Clinic of and and the Mayo is a of of the for in and the We are to of and of both in for patient We of and Mayo for amyloid database to for Sema3B. We are to the over the years, of the patients with primary MN in the of and and in the Hospital at Tenon and in the of and at and to from the of and to for in We and Confocal Hospital and Tenon Tenon and and all of the Tenon Hospital for in and data and the the and the kidney and and performed the laser microdissection and mass spectrometry. in clinical performed the and for the validation performed the confocal and Western blot and clinical The was and by the with as from the with Supplementary