Profiling and annotation of human kidney glomerulus proteome
© Cui et al.; licensee BioMed Central Ltd. 2013
Received: 13 December 2012
Accepted: 2 April 2013
Published: 8 April 2013
The comprehensive analysis of human kidney glomerulus we previously performed using highly purified glomeruli, provided a dataset of 6,686 unique proteins representing 2,966 distinct genes. This dataset, however, contained considerable redundancy resulting from identification criteria under which all the proteins matched with the same set of peptides and its subset were reported as identified proteins. In this study we reanalyzed the raw data using the Mascot search engine and highly stringent criteria in order to select proteins with the highest scores matching peptides with scores exceeding the “Identity Threshold” and one or more unique peptides. This enabled us to exclude proteins with lower scores which only matched the same set of peptides or its subset. This approach provided a high-confidence, non-redundant dataset of identified proteins for extensive profiling, annotation, and comparison with other proteome datasets that can provide biologically relevant knowledge of glomerulus proteome.
Protein identification using the Mascot search engine under highly stringent, computational strategy generated a non-redundant dataset of 1,817 proteins representing 1,478 genes. These proteins were represented by 2-D protein array specifying observed molecular weight and isoelectric point range of identified proteins to demonstrate differences in the observed and calculated physicochemical properties. Characteristics of glomerulus proteome could be illustrated by GO analysis and protein classification. The depth of proteomic analysis was well documented via comparison of the dynamic range of identified proteins with other proteomic analyses of human glomerulus, as well as a high coverage of biologically important pathways. Comparison of glomerulus proteome with human plasma and urine proteomes, provided by comprehensive analysis, suggested the extent and characteristics of proteins contaminated from plasma and excreted into urine, respectively. Among the latter proteins, several were demonstrated to be highly or specifically localized in the glomerulus by cross-reference analysis with the Human Protein Atlas database, and could be biomarker candidates for glomerular injury. Furthermore, comparison of ortholog proteins identified in human and mouse glomeruli suggest some biologically significant differences in glomerulus proteomes between the two species.
A high-confidence, non-redundant dataset of proteins created by comprehensive proteomic analysis could provide a more extensive understanding of human glomerulus proteome and could be useful as a resource for the discovery of biomarkers and disease-relevant proteins.
KeywordsHuman glomerulus proteome Human plasma proteome Human urine proteome Mouse glomerulus proteome Cross-reference analysis Bioinformatics Biomarkers
The glomerulus is the site of plasma filtration and production of primary urine in the kidney. Not only does the glomerulus play a pivotal role in the ultrafiltration of plasma into urine but it is also the locus of kidney diseases which often progress to chronic renal failure. The diagnosis and treatment of glomerular diseases are now based on clinical manifestations, urinary protein excretion level, and the renal pathology of needle biopsy specimens. The cellular and matrix architecture of the glomeruli of biopsy specimens have been mapped in detail, providing a basis for diagnosis, classification, and clinical treatment decisions. In contrast to the detailed information about morphological changes in the glomerulus, the molecular composition of the glomerulus and how it changes with the progress of the diseases are still obscure. Identification and characterization of biomarkers and proteins relevant to the onset and progress of diseases, therefore, are of high priority, as they would allow better disease classification, detection, and prognosis. Proteomic analysis of the glomerulus could be the most straightforward approach toward discovery of disease-related proteins , while most efforts of proteomic analyses have been focused on urine [2, 3].
We have previously analyzed glomeruli purified from kidney cortex with no apparent pathological manifestations in order to compile an in-depth profile of the normal human glomerulus proteome as a resource for clinical research . The large-scale shotgun proteomic analysis provided a dataset of 6,686 identified proteins representing 2,966 distinct genes. The dataset, however, contained considerable redundancy: proteins produced by alternative splicing of primary mRNAs from the same gene and proteins produced from gene families containing highly conservative regions of nucleotide or amino acid sequences are included in the dataset. These proteins were generally identified with lower score and arose from both bioinformatics and biological redundancy because all the proteins matched with the same set of peptides or its subsets were reported as identified proteins. Redundancy is unavoidable in peptide-based protein identification in mass spectrometry, i.e., bottom-up or shotgun analysis. Although we could not exclude the actual presence of these proteins, the redundant dataset obviously contained many ambiguously identified proteins as explained above, and could considerably affect profiling and annotation of proteome under analysis.
In the present study, we reanalyzed raw data obtained in our previous, large-scale proteomic analysis to generate a high-confidence, non-redundant dataset of identified proteins by using the Mascot search engine and highly stringent, computational strategy for extensive profiling and annotation of the normal human glomerulus proteome . The identified proteins were represented by 2-D protein array specifying actual molecular weight (Mw) and isoelectric point (pI) range to demonstrate differences in the observed and calculated physicochemical properties, often lost in shotgun analysis. Characteristics of the glomerulus proteome were illustrated by Gene Ontology (GO) analysis and protein classification with the aid of bioinformatics tools. The depth of proteomic analysis was well documented by comparing the dynamic range of identified proteins with other proteomic analyses targeting the glomerulus. Comparison of the glomerulus proteome with comprehensive analyses of the human plasma and human urine proteomes suggested the extent and characteristics of proteins contaminated from plasma and excreted into urine, respectively. Furthermore, a comparison of ortholog proteins identified in human and mouse glomerulus suggested some biologically significant differences between the two species. The extensive profiling and annotation of the human glomerulus proteome were first conducted in this study and could be a useful resource for the discovery of biomarkers and disease-relevant proteins.
Results and discussion
In our previous study , glomeruli from the kidney cortex of a 68-year-old male patient who had undergone a nephrectomy due to ureter carcinoma were purified to apparent homogeneity by the sieving method. The cortex was histologically normal on the basis of light microscopic observations of PAS- and PAM-stained tissue and no significant deposition of immunoglobulins (IgA, IgG, and IgM) or complement C3 was observed by immunofluorescence microscopy. In the previous study, protein extract from the glomerular preparation was fractionated by two procedures; 1-D SDS-PAGE and 2-D pre-fractionation using solution phase isoelectric focusing (IEF) followed by 1-D-SDS-PAGE. All SDS-PAGE lanes were cut into 15 slices corresponding to a total of 90 slices or fractions, processed by in-gel trypsin digestion, and analyzed by a nanoflow-ESI-ion trap mass spectrometer. Peptides recovered from each gel slice were analyzed in duplicate by two consecutive LC-MS/MS runs followed by two consecutive blank LC-MS/MS runs. The four raw data files generated were merged and identified using the Spectrum Mill search engine (for details, see Additional file 1). In this study, the four raw data files for each of the fractions were converted into Mascot generic files via a Data Analysis software (Agilent) and used for protein identification using the Mascot search engine against the IPI_human protein sequence database (ver. 3.70) to create a dataset of high-confidence, non-redundant identified proteins (see Methods section). In addition, the same raw data files were analyzed by Spectrum Mill using the same version of the IPI_human protein sequence database for convenience of comparison.
Summary of proteins identified by Mascot and Spectrum Mill
A Proteins identified by Spectrum Mill
Number of identified proteins
High confidence a
Low confidence b
Total identified proteins c
Fr. 1 (pH 3–4.6)
Fr. 2 (pH 4.6-5.4)
Fr. 3 (pH 5.4-6.2)
Fr. 4 (pH6.2-7.0)
Fr. 5 (pH 7.0-10.0)
Total distinct proteinsg
Total distinct genesh
B Proteins identified by Mascot
Fr. 1 (pH 3–4.6)
Fr. 2 (pH 4.6-5.4)
Fr. 3 (pH 5.4-6.2)
Fr. 4 (pH6.2-7.0)
Fr. 5 (pH 7.0-10.0)
Total distinct proteinsg
Total distinct genesh
We confirmed successful fractionation with minimal cross-contamination of the solution phase IEF in the first dimension of 2-D pre-fractionation of glomerular proteins . Nevertheless, the difference between observed and calculated pI and Mw was obvious. These observations could possibly be explained by aggregation, degradation or posttranslational modifications (i.e. glycosylation and phosphorylation) of actual proteins in tissue lysate. The solution phase IEF in the first dimension of 2-D pre-fractionation of proteins resulted in extensive condensing of proteins in their pI ranges, which might have contributed to aggregation and/or degradation. The prolonged time required for 2-D pre-fractionation of proteins could also contribute to aggregation as well as degradation of proteins. This result may impose the need for careful consideration on peptide-based targeted proteomics for qualitative and quantitative analysis if protein fractionation is utilized to concentrate target proteins prior to targeted proteomic analysis.
Characterization of glomerulus proteome using bioinformatics tools
All the identified proteins included in the non-redundant, high-confidence dataset consisting of 1,817 unique proteins representing 1,478 unique genes were subjected to bioinformatics analysis based on the structured vocabulary of Gene Ontology (GO) using a PANTHER analytical tool (ver. 7.0) [6, 7]. Subcellular distribution as estimated by analysis with GO Cellular Component vocabulary indicated the highest number of hits to the actin cytoskeleton and a considerably higher number of hits to the intermediate filament cytoskeleton and the extracellular region (a term defining proteins present in the external protective or encapsulating structure outside the plasma membrane including the extracellular matrix) (Additional file 6: Figure S1A). The classification of proteins based on GO Molecular Function vocabulary (Additional file 6: Figure S1B) and Biological Process vocabulary (Additional file 6: Figure S2A) yielded the successful identification of unbiased, widely diverse proteins. Enrichment analysis based on GO Biological Process vocabulary using Cytoscape (version 2.82) with the BinGO plug-in (version 2.42)  further illustrated biological processes in which glomerular proteins are significantly enriched compared to products of whole human genes (p < 0.001, hypergeometric test followed by multiple testing correction using Benjamin and Hochberg false discovery rate correction) (Additional file 6: Figure S2B).
Under-representation or depletion analysis based on GO Biological Process vocabulary using Cytoscape, as described above, indicated significant depletion of proteins involved in “sensory perception”, “regulation of small GTPase mediated signal transduction”, especially in “regulation of Ras protein signal transduction”, and, most notably, in “regulation of transcription” including “regulation of transcription from RNA polymerase II promoter” (Additional file 7).
PANTHER Protein Class analysis is based on PANTHER Molecular Function ontology, which includes commonly used classes of protein functions, many of which are not covered by GO Molecular Function. PANTHER Protein Class analysis again showed that the highest number of proteins was classified as cytoskeletal proteins (Additional file 6: Figure S3).
Depth and coverage of proteomic analysis of human kidney glomerulus
We assessed the depth of our comprehensive analysis of the human kidney proteome by comparison with two high-confidence, non-redundant datasets of proteomic analysis of human and mouse glomeruli. The former was the result of analysis of human glomerulus laser-microdissected from frozen sections of biopsy specimens using conventional HPLC in combination with an LTQ-Orbitrap mass spectrometer , providing identification of more than 400 proteins from 50 glomerular sections. The latter was the result of analysis of laser-microdissected mouse kidney glomerulus by employing a newly developed nanoflow HPLC on a long, smaller internal diameter column coupled with an LC-MS interface, termed “Replay”. This system allows the direct reanalysis of the injected sample without losing signal intensity , providing identification of more than 2,400 proteins from 50 glomerular sections by an LTQ-Orbitrap mass spectrometer.
Comparison of glomerulus proteome with human plasma proteome
Comparison of glomerulus proteome with human urine proteome
Comparison of glomerulus proteome with mouse glomerulus proteome
Waanders et al. reported a comprehensive proteomic analysis of mouse glomerulus prepared by laser microdissection from frozen tissue sections , providing identification of more than 2,400 proteins as described above. In the analysis of mouse glomerulus , MS/MS data were processed by MaxQuant and searched against the mouse IPI database (ver. 3.37), using Mascot (version 2.1). Parameter settings for their search were similar to those of our study, except that peptide and fragment mass tolerance were set at 7 ppm and 0.5 Da, respectively, and proteins were considered identified when at least two peptides were identified, and at least one of which was uniquely assignable to the respective sequence to exclude redundant protein hits. The false discovery rate (FDR) at the peptide level was set to 1%. The dataset of mouse glomerulus proteome, therefore, was highly confident, non-redundant, and almost comparable to our study. In order to compare both proteomes and gain insight into differences between the two, first, we constructed an ortholog table of human and mouse genes and converted the mouse gene symbols to corresponding human ortholog gene symbols. The ortholog table of human and mouse genes was constructed based on an “Evola ortholog list” (version 7.5) in the Human-Invitational Database (H-Inv DB) . To construct an Evola ortholog list in the H-Inv DB, a list of human genes was created via the mapping of all human full-length transcripts to genomic sequences using three nucleotide sequence alignment search engines including BLAT, BLAST and est2genome. For other species, gene loci were similarly predicted through the mapping of all transcripts from DDBJ, RefSeq and Ensemble nucleotide sequence databases to genomic sequences. The Evola ortholog list was then generated if exons overlapped between species in genomic alignment for the longest for both the human side and the other species’ side: the pairs were detected as ortholog by computational analysis if they could be aligned with a length of 50% or more of the human amino acid sequence. Using this approach, a pair of ortholog genes in the two species under comparison, which were generated from an ancestor gene by specification, was selected.
The Human Protein Atlas revealed the uniquely identified proteins in the glomerulus (343 genes) to be localized in the kidney (Additional file 11: Table S1). Among them, 162 proteins were found to be localized in the glomerulus in the Human Protein Atlas immunohistochemistry database. We further inspected the proteins expressed in the glomerulus to find proteins expressed highly or specifically in the glomerulus. We found several glomerular proteins which were also confirmed in the literature to be glomerulus-specific proteins including cadherin-13 (gene CDH13) , RcPTPNS1 or tyrosine-protein phosphatase non-receptor type substrate 1 (gene SIRPA) , isoform 1 of Crumbs homolog 2 (gene CRB2) [20, 21], and isoform 1 of secretory phospholipase A2 receptor (gene PLA2R1)  (Additional file 11: Table S2). It is interesting to note that PLA2R1 was identified as one of the major target antigens in idiopathic membranous nephropathy [23, 24]. Obviously, further experiments are necessary for confirmation, but it might be possible that these proteins are specifically or highly enriched in the human glomerulus in comparison to the mouse glomerulus.
The raw data produced by our previous comprehensive analysis of human glomerulus proteome were reanalyzed using a high-stringency Mascot search engine to create a high-confidence, non-redundant dataset of identified proteins. This approach provided 1,817 unique proteins representing 1,478 genes and allowed extensive profiling of glomerulus proteome using bioinformatics tools and cross-reference analyses with normal human plasma and urine proteomes, and, in addition, with mouse glomerulus proteome. The considerable difference in the observed and calculated pI and Mw of identified proteins was clearly demonstrated. Analysis based on structured vocabulary of GO indicated the successful identification of a wide unbiased diversity of proteins, and underscored significant enrichment of cytoskeletal and extracellular matrix proteins in the glomerulus proteome. A comparison of the dynamic range of our study with two other proteomic analyses of glomerulus demonstrated the considerable depth of our proteomic analysis in terms of dynamic range and coverage. Comparison of glomerulus proteome with normal human plasma and urine proteomes revealed protein contamination from plasma (amounting to 22.1% of proteins identified in the glomerulus) and excretion into urine (23.4%). A cross-reference analysis with the Human Protein Atlas database indicated the excretion of proteins highly or specifically localized in the glomerulus into urine, suggesting their possible clinical use as urinary biomarkers for glomerular injury. A comparison of ortholog proteins identified in human and mouse glomeruli showed considerable similarity but also suggested some biologically significant differences between species.
LC-tandem mass analysis
The workflow for the preparation of human kidney glomeruli, the strategy for comprehensive analysis, and the LC-MS/MS analysis are provided in Additional file 1. Briefly, 2 mg of proteins extracted from purified glomeruli were either separated on 1-D SDS-PAGE (1-D pre-fractionation) or by 2-D pre-fractionation (solution phase IEF followed by 1-D SDS-PAGE). All the lanes of SDS-PAGE gels were cut into 15 slices (90 slices or fractions in total), and subjected to in-gel digestion to produce tryptic peptides. Two replicate LC-MS/MS runs with samples followed by two consecutive LC-MS/MS runs with blanks (0.3% formic acid) were conducted for each sample. The latter two blank runs were included to retrieve and eliminate carryover peptides from the preceding sample runs.
Protein identification using Spectrum Mill
We have previously reported identified proteins using Spectrum Mill (version A.03.12.060) as a search engine against the IPI_human protein sequence database (version 3.18) . In this study, we reanalyzed the same raw data set using a new version of Spectrum Mill (version 03.03.081 SR1a) against the IPI_human protein sequence database (version 3.70) for convenience of comparison with proteins identified using the Mascot search engine against the same version of the IPI-human protein sequence database. The identification strategy and criteria for protein identification were similar to those of the previous report . The protein identification done using Spectrum Mill is summarized in Table 1.
Protein identification using Mascot
For protein identification with the Mascot search engine, the raw data files generated by Spectrum Mill were converted to Mascot generic files (mgf files) by using the built-in script of Data Analysis software (Agilent version 6.1) without grouping. The 4 mgf files corresponding to each of the fractions (2 sample runs and 2 blank runs) were merged, and searched against the IPI_human protein sequence database (version 3.70) using the Mascot search engine (version 2.2.1). Cystein carbamoidmethylation was set as the fixed modification, and other modifications were set as variable modifications including oxidation of methionine, oxidation of histidine and tryptophan, N-terminal glutamine to pyroglutamate, and N-terminal glutamate to pyroglutamate. Peptide and fragment mass tolerance were set at ±2.5 Da, and at ± 0.7 Da, respectively. A maximum of one missed cleavage was allowed. Proteins matched with at least one unique peptide and with peptides of scores above the “identity threshold” were selected to generate a non-redundant, high-confidence dataset of identified proteins. Selection of matched peptides with scores exceeding the identity threshold was performed by using the same value as the significant threshold in the “Ion score or cut-off” parameter setting. The significant threshold was adjusted to give a false discovery rate of less than 1% (0.25 ± 0.24%), which was calculated on the basis of the number of peptide matches against a decoy database. Protein identification is summarized in Table 1 in comparison with data obtained with Spectrum Mill.
Acute kidney injury
Chronic kidney diseases
False discovery rate
- H-Inv DB:
Human Protein Atlas
Mascot generic file
Normalized spectral abundance factor
This study was supported by a Grant-in-Aid for Scientific Research (C) to YY (22590881) from the Japan Society for Promotion of Science, a Grant-in-Aid for Strategic Research Project to TY (500460) from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and a Grant-in-Aid for Diabetic Nephropathy and Nephrosclerosis Research to TY from the Ministry of Health, Labor and Welfare of Japan.
- Yoshida Y, Miyamoto M, Taguchi I, Xu B, Zhang Y, Yaoita E, Fujinaka H, Yamamoto T: Human kidney glomerulus proteome and biomarker discovery of kidney diseases. Proteomics Clin Appl 2008, 2: 420–427. 10.1002/prca.200780016PubMedView Article
- Bonventre JV, Vaidya VS, Schmounder R, Feig P, Dieterle F: Next-generation biomarkers for detecting kidney toxicity. Nat Biotech 2010, 28: 436–440. 10.1038/nbt0510-436View Article
- Decramer S, Gonzalez de Peredo A, Breul B, Mishak H, Monsarrat B, Bascands J-L, Schanstra JP: Urine in clinical proteomics. Mol Cell Proteomics 2008, 7: 1850–1862. 10.1074/mcp.R800001-MCP200PubMedView Article
- Miyamoto M, Yoshida Y, Taguchi I, Nagasaka Y, Tasaki M, Zhang Y, Xu B, Nameta M, Sezaki H, Cuellar LM, Osawa T, Morishita H, Sekiyama S, Yaoita E, Kimura K, Yamamoto T: In-depth proteomic profiling of the normal human kidney glomerulus using two-dimensional protein prefractionation in combination with liquid chromatography-tandem mass spectrometry. J Proteome Res 2007, 6: 3680–3690. 10.1021/pr070203nPubMedView Article
- Kapp EA, Schütz F, Connolly LM, Chakel JA, Meza JE, Miller CA, Fenyo D, Eng JK, Adkins JN, Omenn GS, Simpson RJ: An evaluation, comparison, and accurate benchmarking of several publicly available MS/MS search algorithms: Sensitivity and specificity analysis. Proteomics 2005, 5: 3475–3490. 10.1002/pmic.200500126PubMedView Article
- Mi H, Dong Q, Muruganujan A, Gaudet P, Lewis S, Thomas PD: PANTHER version 7: improved phylogenetic trees, orthologs and collaboration with the Gene Ontology Consortium. Nuc Acids Res 2010, 38: D204-D210. 10.1093/nar/gkp1019View Article
- PANTHER Classification System. [http://www.pantherdb.org/]
- Cytoscape. [http://www.cytoscape.org/]
- Yoshida Y, Nameta M, Kuwano M, Zhang Y, Xu B, Magdeldin S, Cui Z, Fujinaka H, Yaoita E, Tomonaga T, Yamamoto T: Proteomic approach to human kidney glomerulus prepared by laser microdissection from frozen biopsy specimens: Exploration of proteome after removal of blood-derived proteins. Proteomics Clin Appl 2012, 6: 412–417. 10.1002/prca.201200016PubMedView Article
- Waanders LF, Chwalek K, Monetti M, Kumar C, Lammert E, Mann M: Quantitative proteomic analysis of single pancreatic islets. Proc Natl Acad Sci USA 2009, 106: 18902–18907. 10.1073/pnas.0908351106PubMed CentralPubMedView Article
- DAVID. [http://david.abcc.ncifcrf.gov/]
- Paoletti AC, Parmely TJ, Tomomori-Sato C, Sato S, Zhu D, Conaway RC, Conaway JW, Florens L, Washburn MP: Quantitative proteomic analysis of distinct mammalian mediator complexes using normalized spectral abundance factors. Proc Natl Acad Sci USA 2006, 103: 18928–18933. 10.1073/pnas.0606379103PubMed CentralPubMedView Article
- States DJ, Omenn GS, Blackwell TW, Fermin D, Eng J, Speicher DW, Hanash SM: Challenges in deriving high-confidence protein identifications from data gathered by a HUPO plasma proteome collaborative study. Nat Biotech 2006, 24: 333–338. 10.1038/nbt1183View Article
- Anderson NL, Anderson NG: The human plasma proteome. Mol Cell Proteomics 2002, 1: 845–867. 10.1074/mcp.R200007-MCP200PubMedView Article
- Adachi J, Kumar C, Zhang Y, Olsen JV, Mann M: The human urinary proteome contains more than 1500 proteins, including a large proportion of membrane proteins. Genome Biol 2006, 7: R80. 10.1186/gb-2006-7-9-r80PubMed CentralPubMedView Article
- The Human Proteome Atlas. [http://www.proteinatlas.org/]
- Human-Invitational Database (H-Inv DB). [http://h-invitational.jp/hinv/ahg-db/index.jsp]
- Arnemann J, Sultani O, Hasgün D, Coerdt W: T-/H-cadherin (CDH-13): a new marker for differentiating podocytes. Virchows Arch 2006, 448: 160–164. 10.1007/s00428-005-0055-7PubMedView Article
- Willingham SB, Volkmer JP, Gentles AJ, Sahoo D, Dalerba P, Mitra SS, Wang J, Contreras-Trujillo H, Martin R, Cohen JD, Lovelance P, Sceeren FA, Chao MP, Weiskopf K, Tang C, Volkmer AK, Naik TJ, Storm TA, Mosley AR, Edris B, Schmid SM, Sun CK, Chua MS, Murillo O, Rajendran P, Cha AC, Chin RK, Kim D, Adorno M, Raveh T, Tseng D, Jaiswal S, Enger PO, Steinberg GK, Li G, So SK, Majeti R, Harsh GR, van de Rijn M, Teng NN, Sunwoo JB, Alzadeh AA, Clarke MF, Weissman IL: The CD-47-signaling regulatory protein alpha (SIRa) interaction is a therapeutic target for human solid tumours. Proc Natl Acad Sci USA 2012, 109: 6662–6667. 10.1073/pnas.1121623109PubMed CentralPubMedView Article
- van den Hurk JA, Rashbass P, Roepman R, Davis J, Voesenek KE, Arends ML, Zonneveld MN, van Roekel MH, Cameron K, Rohschneider K, Heckenlively JR, Koenekoop RK, Hoyng CB, Cremers FP, den Hollander AI: Characterization of the Crumbs homolog 2 (CRB2) gene and analysis of its role in retinitis pigmentosa and Leber congenital amaurosis. Mol Vis 2005, 11: 263–273.PubMed
- Boroviak T, Rashbass P: The apical polarity determinant Crumbs 2 is a novel regulator of ESC-derived neural progenitors. Stem Cells 2011, 29: 193–205. 10.1002/stem.567PubMedView Article
- Ronco P, Debiec H: Antigen identification in membranous nephropathy moves toward targeted monitoring and new therapy. J Am Soc Nephrol 2010, 21: 564–569. 10.1681/ASN.2009121220PubMedView Article
- Beck LH Jr, Bonegio RGB, Lambeau G, David M, Beck BA, Powell DW, Cummins TD, Klein JB, Salant DJ: M-type phospholipase A2 receptor as target in idiopathic membranous nephropathy. N Eng J Med 2009, 361: 11–21. 10.1056/NEJMoa0810457View Article
- Debiec H, Ronco P: PLA2R autoantibodies and PLA2R glomerular deposits in membranous nephropathy. N Eng J Med 2011, 364: 689–690. 10.1056/NEJMc1011678View Article
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.