Urinary Biomarkers Associated With Pathogenic Pathways Reflecting Histologic Findings in Lupus Nephritis

Kazuoto Hiramoto; Shuntaro Saito; Hironari Hanaoka; Jun Kikuchi; Hiroyuki Fukui; Akinori Hashiguchi; Katsuya Suzuki; Tsutomu Takeuchi; Yuko Kaneko

doi:10.1002/art.43017

. 2024 Nov 11;77(3):298–310. doi: 10.1002/art.43017

Urinary Biomarkers Associated With Pathogenic Pathways Reflecting Histologic Findings in Lupus Nephritis

Kazuoto Hiramoto ¹, Shuntaro Saito ¹, Hironari Hanaoka ¹, Jun Kikuchi ¹, Hiroyuki Fukui ¹, Akinori Hashiguchi ¹, Katsuya Suzuki ¹, Tsutomu Takeuchi ², Yuko Kaneko ^1,^✉

PMCID: PMC11865699 PMID: 39317671

Abstract

Objective

There is a pressing need to understand the pathogenesis of histologic findings and identify the biomarkers for predicting the histologic severity in lupus nephritis (LN). This study aimed to identify the pathogenic signal pathway and elucidate urinary biomarkers for predicting the presence or severity of histologic findings in LN.

Methods

Urine samples from patients with biopsy‐proven active LN were screened for 1,305 proteins using an aptamer‐based proteomic assay. The diversity and expansion of individual renal histologic features in LN were quantified to identify the urinary proteins associated with the histologic findings found in each score. Candidate urinary proteins were validated in a validation cohort. Immunohistochemical staining of the renal tissues was performed to clarify the localization of the candidate proteins.

Results

Cluster analysis extracted five histologic subgroups according to their correlations with each histologic finding in LN. Protein groups that correlated with each histologic subgroup revealed a distinct pathogenesis in LN using pathway analyses. Enzyme‐linked immunosorbent assay validation revealed that urinary calgranulin B (S100A9), monocyte chemotactic protein 1 (MCP‐1), and insulin‐like growth factor binding protein 5 (IGFBP‐5) levels could specifically predict the presence and severity of active glomerular lesions, interstitial inflammation, and interstitial fibrosis, respectively. Immunohistochemical staining revealed the localization of these proteins in each lesion.

Conclusion

Renal histologic findings may reflect the different pathogeneses involved in each lesion, and estimating the urinary calgranulin B, MCP‐1, and IGFBP‐5 levels may be useful in predicting the presence and severity of histologic findings in LN.

INTRODUCTION

Systemic lupus erythematosus (SLE) is a common autoimmune disease that affects young female patients and causes a variety of organ disorders. ¹ Half of the patients with SLE develop lupus nephritis (LN), and 10% to 15% of patients with LN develop end‐stage renal disease. ² , ³ LN has six classifications according to its glomerular lesions based on the International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification. However, this classification is not precise enough to evaluate and quantify the accurate progression or expansion of lesions, which leads to the necessity to add activity and chronicity indices defined by the National Institutes of Health (NIH) to ISN/RPS classification. ⁴ In addition, it is important to evaluate chronic tubulointerstitial lesions, which better reflect renal prognosis. ⁵ , ⁶ , ⁷ Therefore, renal biopsy is still the gold standard for the definitive diagnosis and classification of LN, and its severity is determined by various pathologic parameters. ⁸ However, renal biopsy is an invasive procedure that accompanies bleeding in various degrees and is not feasible in cases with associated complications and risk factors. Therefore, the development of noninvasive predictive biomarkers for renal histopathological findings is a great unmet need in clinical practice.

The role of urinary biomarkers in diagnosing LN and predicting relapse has been recognized, and many candidate proteins have been reported for this purpose. ⁹ Recently, proteomic approaches have been used to identify serum and urine proteins as unbiased screening biomarkers in LN. ¹⁰ , ¹¹ , ¹² , ¹³ However, all these studies identified urinary biomarkers by comparing patients with active LN and healthy controls. Hence, these studies did not consider the various differences in histologic characteristics between active and chronic lesions in LN. Recently, the Accelerating Medicines Partnership study reported urine biomarkers associated with each renal histologic finding, according to the NIH activity and chronicity index. ¹⁴ , ¹⁵ However, biomarkers that can predict the progression of individual lesions remain unidentified. This study aimed to identify the pathogenic signal pathway in LN and elucidate urinary biomarkers for predicting the presence or severity of histologic findings of LN by precisely evaluating renal histology.

PATIENTS AND METHODS

Patients and sample preparation

Consecutive patients with biopsy‐proven class III/IV, III/IV+V, or V LN were recruited from Keio University Hospital. Samples from two cohorts were used: (1) a discovery cohort for screening, including patients with LN (n = 24) and diabetic nephropathy (n = 3) diagnosed by renal biopsy, and (2) a validation cohort for enzyme‐linked immunosorbent assay (ELISA) validation, including patients with LN (n = 24). Serum and urine samples were collected within one week before renal biopsy and treatment intensification. Clean‐catch midstream urine samples were collected in sterile containers and refrigerated within one hour after collecting the samples. Blood samples were collected, and the serum was immediately separated by centrifugation. The samples were then aliquoted and stored at −80°C. The clinical data at the time of sample collection were recorded. Informed consent was obtained from all patients, and the study was approved by the Ethics Committee of our institution (Keio University School Hospital, approval number 20140093).

Aptamer‐based screening

Serum and urine samples obtained from patients in the discovery cohort were analyzed by screening with a slow off‐rate modified DNA aptamer–based capture array (SOMAscan; SomaLogic), which is a comprehensive high‐throughput proteomic assay using an Agilent microarray readout that measures 1,305 proteins. ¹⁶ This assay is based on interactions between the aptamers and proteins in a sample. First, aptamer‐coated streptavidin beads were added to the sample to allow the aptamers to bind to the proteins. Next, the bound proteins were tagged with biotin, and the aptamer–protein complexes were released from the streptavidin beads. These aptamer–protein complexes were then recaptured on a second set of streptavidin beads, and aptamers were released from these proteins. Finally, the aptamers were hybridized to complementary sequences on a microarray chip and quantified by fluorescence. Concentrations were measured for 1,305 proteins, and the results were calculated as relative fluorescent units (RFU). RFU is related to the amount of protein in the original sample. These RFU values were standardized by urinary creatinine concentration.

Histologic scoring system

We evaluated renal histologic findings more quantitatively than the ISN/RPS lesion definitions and classification of renal involvement or the NIH activity and chronicity index by developing the original scoring system. ¹⁷ , ¹⁸ The scoring system included a total of 20 renal histologic findings, including 16 glomerular lesions and 4 tubulointerstitial lesions (Supplementary Tables 1 and 2). The renal histology was evaluated using a light microscope. Light microscopy specimens were embedded in paraffin, sectioned, and stained with hematoxylin‐eosin, periodic acid‐Schiff, periodic acid‐silver methenamine, and Masson's trichrome reagent. Each glomerular histologic finding was scored using two methods. (1) For endocapillary hypercellularity and subendothelial deposits, the percentage of lesions in the total glomerular capillary tuft was calculated in the range of 0% to 100%, and the percentage was defined as the score of one glomerulus. The average score of all glomeruli was used as the histologic score of each case. (2) For histologic findings of glomeruli other than those described previously, the presence or absence of lesions was evaluated. The percentage of glomeruli with lesions was calculated in the range of 0% to 100%, and this percentage was used as the histologic score of each case. For tubulointerstitial lesions, the percentage of lesions relative to the whole area of the tubulointerstitium was calculated in the range of 0% to 100%, and the percentage was used as the score of each case. For glomerular lesions, a score was calculated for each glomerulus, and the average of the scores of all glomeruli was used as the histologic score of the case. Histologic scores were assigned by two skilled raters (KH and SS) supervised by a pathologist (AH), and the mean of the two was used for correlation analysis. This scoring system can quantify the diversity and expansion of renal histologic features in LN more precisely than the ISN/RPS classification or the NIH activity and chronicity index because it includes the same or more items as the existing evaluation systems.

Validation studies using ELISA

ELISA tests were performed on candidate urinary proteins that correlated with specific histologic scores in the discovery cohort. The correlation between the ELISA test results and renal histologic scores in the validation cohort was then confirmed. The concentrations of urinary proteins, including calgranulin B (Cloud‐Clone), S100A12 (R&D Systems), high mobility group N1 (HMGN1) (Cloud‐Clone), interleukin‐8 (IL‐8) (R&D Systems), allograft inflammatory factor 1 (AIF‐1) (Cloud‐Clone), superoxide dismutase 1 (SOD1) (RayBiotech), monocyte chemotactic protein 1 (MCP‐1) (R&D Systems), FSTL3 (R&D Systems), and insulin‐like growth factor binding protein 5 (IGFBP‐5) (R&D Systems), were measured using ELISA. Absolute urinary protein levels were determined using standard curves run on each ELISA plate and standardized by the urinary creatinine concentration.

Immunohistochemistry

Immunohistochemical staining of the renal tissues was performed using a Leica BOND MAX Immunostainer (Leica Biosystems) for calgranulin B, MCP‐1, and IGFBP‐5. After deparaffinization, the renal tissues were stained with BOND Epitope Retrieval Solution 1 (AR9961, Leica Biosystems) for calgranulin B and BOND Epitope Retrieval Solution 2 for MCP‐1 and IGFBP‐5 (AR9640, Leica Biosystems). For immunohistochemical staining, the renal tissues were incubated with 1:2,000 diluted mouse monoclonal anti–calgranulin B (TA804091S, OriGene Technologies), 1:100 diluted mouse monoclonal anti–MCP‐1 (MAB 679, R&D Systems), and 1:50 diluted mouse monoclonal anti–IGFBP‐5 (MAB875, R&D Systems) antibodies. BOND Polymer Refine Detection (DS9800; Leica Biosystems) was used as the secondary antibody.

Data analysis

The JMP software version 15 (SAS Institute) was used for statistical analysis. Group comparisons were performed using the Wilcoxon rank sum test. Receiver operating characteristic (ROC) curve analysis was performed to analyze the sensitivity, specificity, and cutoff values of the biomarkers for renal histopathological findings. From the data of the aptamer‐based screening assay and histologic scores, a heatmap of the correlation coefficients was created, and cluster analysis was performed using the Python 3.9 (Python Seaborn package) programming. Spearman's rank correlation and Pearson's correlation coefficients were used for correlation related to aptamer‐based screening–measured and ELISA‐measured data, respectively. There were two reasons for using Spearman's correlation in the screening phase of this study: (1) The measurements by SOMAscan themselves may not reflect precise concentrations and could show considerable variability in results, and (2) ranking was necessary when extracting specific clusters of urinary proteins during cluster analysis. Euclidean distance was used for the distance matrix, and Ward's method was used for hierarchical clustering.

The Enrichr web tool (open source available from https://amp.pharm.mssm.edu/Enrichr) was used to determine the cellular origin of the proteins that correlated with individual clusters of renal lesions. Ingenuity pathway analysis (IPA) was used to ascertain which known pathways were enriched by proteins that were highly correlated with individual clusters of renal lesions. Data were analyzed using IPA (QIAGEN Inc, https://www.qiagen.com/us). Protein–protein interaction (PPI) analysis was performed to identify interactions between individual clusters of renal lesions and highly correlated proteins. The PPI database STRING (available from https://string-db.org/) was used, and clustering was performed with the k‐means method. All data relevant to the study are included in the article or uploaded as online supplementary information.

RESULTS

Cluster analysis with histologic score and aptamer‐based protein screening

Urine and blood samples from 24 patients with LN (class III/IV = 14, III/IV+V = 6, V = 4) were used in the discovery cohort, and urine samples from 24 patients with LN (class III/IV = 10, III/IV+V = 6, V = 8) were used in the validation cohort. The clinical characteristics and renal histologic scores of the discovery, validation, and diabetic nephropathy groups are presented in Supplementary Tables 3 and 4, respectively. The LN classification based on the renal pathology score completely matched the ISN/RPS classification, which was assessed by a pathologist. The two raters have shown good interrater reliability for histologic assessments in the discovery and validation cohorts (κ values of 0.62 to 0.95 for the presence or absence of 20 renal histologic findings).

We first performed a cluster analysis on 20 pathology items according to the correlation between each item and the histologic score. Cluster analysis of the 20 renal histologic findings identified five clusters (Figure 1A). Cluster 1 (extracapillary lesions) included cellular or fibrocellular crescents, fibrinoid necrosis, adhesions, and fibrous crescents. Cluster 2 (endocapillary lesions) included only active lesions within the glomerular tuft, including endocapillary hypercellularity, neutrophil infiltration, subendothelial deposits, and karyorrhexis. Cluster 3 (membranous and mesangial lesions) included basement membrane and mesangial lesions. Cluster 4 (tubulointerstitial lesions) contained all tubulointerstitial lesions but not glomerular lesions. Cluster 5 (other lesions) included two minor lesions: podocyte hypertrophy and collapsed glomeruli.

Cluster analysis with histologic score and aptamer‐based protein screening. (A) Cluster analysis performed on 20 pathology items according to the correlation among each histologic score. (B and C) Extracted UGs with high correlation coefficients for each histologic score in clusters 1, 2, and 4 (UG1, 2, and 4, respectively). (D) Venn diagram of urinary protein numbers in UG1, 2, and 4. GBM, glomerular basement membrane; UG, urinary protein subgroup.

Subsequently, the correlation between the histologic scores and 1,305 urinary proteins was evaluated via cluster analysis for each of the five histologic clusters. For each protein, we calculated the sum of the ranks of correlation coefficients with each histologic score included in the histologic clusters. Then the mean ranks of the correlation coefficients of the proteins for each urinary protein subgroup (UG) were compared, and the UGs with the highest ranks were identified. In clusters 1, 2, and 4, UGs (UG1, 2, and 4, respectively) with high correlation coefficients for each histologic score were identified (Figure 1B and C). UG1, 2, and 4 contained 119, 59, and 85 urinary proteins, respectively (Figure 1D). These UGs shared the same proteins with each other to some extent, although the number of common proteins was less than that of each noncluster overlapping protein between UG1 and UG4 or UG2 and UG4. Notably, UG1 shared more than half of the same proteins with UG2. No urinary protein cluster with high correlation coefficients was extracted in clusters 3 and 5. Hence, we focused on UG1, 2, and 4 in our subsequent analyses. Representative renal histologic findings from clusters 1, 2, and 4 are shown in Supplementary Figure 1. We also created heatmaps of correlation coefficients among each protein group (UG1, 2, and 4) and 20 histologic findings as a supplementary analysis (Supplementary Figure 2).

Cell‐type enrichment analysis

Cell‐type enrichment analysis revealed that different cell types were associated with urinary protein molecules in each subgroup (Figure 2A). UG1 was characterized by an abundance of proteins derived from the vascular endothelial cells and platelets, in addition to monocytes and neutrophils. In general, UG2 proteins are derived from cells involved in innate and acquired immunity, such as monocytes, neutrophils, plasma cells, and plasmacytoid dendritic cells. In contrast, UG4 contained many proteins derived from the mesenchymal stem cell lineages involved in fibrosis pathology, in addition to the expected appearance of monocytes.

IPA

IPA generated common and distinct pathways among the top 10 pathways in each subgroup (Figure 2B). Leukocyte adhesion–associated signals were activated in all three subgroups. In UG1, high mobility group protein B‐cell receptor CD221 signaling was significantly activated compared to that in the other subgroups. Inflammation‐related pathways such as IL‐3 and hypoxia‐inducible factor 1α signaling are dominant in UG2. In UG4, various kinds of IL‐17–related signals and fibrosis‐related pathways, such as fibrosis signaling pathways and pathology in chronic obstructive pulmonary disease, are prevalent. In addition, in the integrated IPA of UG1, 2, and 4, it was confirmed that signaling pathways related to inflammatory cytokines and granulocyte or monocyte function were highly activated (Supplementary Figure 3).

PPI analysis

In UG1, the proteins contributing to chemotactic activity, cell growth and survival, and cytoskeletal dynamics were confirmed. In UG2, the associated proteins contributing to chemotactic activity, cell migration and adhesion, and cell survival and differentiation were identified. In UG4, proteins contributing to chemotactic activity, fibrosis, and cellular homeostasis were identified (Figure 3).

Identification of urinary biomarker protein correlated with individual histologic score

We attempted to identify urinary protein markers specific to the renal histologic findings in each cluster among the proteins included in the three subgroups. Table 1 shows the top 30 proteins of UG1, 2, and 4 with high correlation coefficients between the histologic score and urinary protein concentration. The top 30 proteins are listed according to the most prevalent renal histologic findings in clusters 1, 2, and 4. The top 20 proteins with high correlation coefficients for all 20 pathologic findings are listed in Supplementary Table 5.

Table 1.

The top 30 proteins of UG1, 2, and 4 with high correlation coefficients between histological score and urinary protein concentration^*

UG1					UG2					UG4
Molecule	C/FC crescents	Fibrinoid necrosis	Fibrous crescents	Adhesion	Molecule	Endocapillary hypercellularity	Karyorrhexis	Neutrophil infiltration	Subendothelial deposits	Molecule	Interstitial fibrosis	Tubular atrophy	Interstitial inflammation	Tamm‐Horsfall casts
S100A12	0.634	0.446	0.399	0.354	AIF1	0.891	0.635	0.529	0.369	FSTL3 ^a	0.561	0.489	0.691	0.368
Calgranulin B ^b	0.618	0.459	0.374	0.296	SOD	0.742	0.691	0.512	0.204	MCP‐1	0.548	0.551	0.716	0.421
Calcineurin B a	0.565	0.434	0.424	0.223	6Ckine ^a	0.732	0.616	0.628	0.370	CRIS3	0.485	0.385	0.427	0.332
CEBPB	0.534	0.401	0.386	0.341	MCP‐3	0.725	0.642	0.573	0.262	AN32B	0.451	0.436	0.462	0.383
HMGN1	0.528	0.440	0.418	0.290	HMGN1	0.724	0.698	0.686	0.358	Somatostatin‐28	0.451	0.401	0.338	0.176
cGS‐PDE	0.527	0.396	0.386	0.234	Tropomyosin 4	0.718	0.575	0.455	0.299	b‐NGF ^a	0.450	0.358	0.485	0.121
Apo D	0.508	0.384	0.352	0.282	SP‐D	0.710	0.693	0.553	0.340	VEGF sR2 ^a	0.442	0.345	0.450	−0.006
PKC‐B‐II	0.499	0.404	0.416	0.188	EFNB2	0.709	0.601	0.573	0.292	ARMEL ^a	0.430	0.365	0.396	−0.004
B7‐H2	0.498	0.360	0.346	0.130	GSTP1	0.698	0.565	0.453	0.315	PDK1	0.429	0.353	0.415	0.130
Soggy‐1	0.496	0.414	0.389	0.296	VCAM‐1	0.692	0.613	0.464	0.273	CSRP3 ^a	0.425	0.332	0.446	0.161
HDGR2	0.495	0.440	0.418	0.284	IgM ^a	0.690	0.505	0.452	0.274	STAT6	0.424	0.311	0.456	0.154
PSME1	0.494	0.402	0.380	0.134	IP‐10	0.683	0.629	0.558	0.208	IGFBP‐5 ^a	0.407	0.326	0.292	0.152
Azurocidin	0.493	0.346	0.327	0.300	S100A12	0.683	0.613	0.473	0.287	NSE	0.405	0.305	0.523	0.092
Transgelin‐2	0.489	0.418	0.386	0.266	PLXC1	0.682	0.557	0.471	0.324	C2 ^a	0.400	0.361	0.295	0.196
IL‐8	0.487	0.421	0.316	0.212	Transgelin‐2	0.665	0.625	0.489	0.267	NovH ^a	0.394	0.321	0.354	0.076
Kallikrein 14	0.485	0.414	0.389	0.270	S100A4	0.656	0.560	0.462	0.264	NCC27 ^a	0.392	0.365	0.411	0.051
ABL2	0.484	0.419	0.351	0.348	PKC‐B‐II	0.649	0.694	0.507	0.239	WFKN2 ^a	0.391	0.288	0.375	−0.023
Histone H1.2	0.482	0.404	0.362	0.219	cGS‐PDE	0.649	0.642	0.427	0.266	GHC2	0.388	0.278	0.413	0.010
BMP10	0.481	0.464	0.494	0.312	40S RPSA	0.648	0.558	0.376	0.221	CFC1	0.388	0.249	0.356	0.189
IP‐10	0.477	0.425	0.500	0.345	CD5L ^a	0.642	0.607	0.506	0.255	FAM107B ^a	0.387	0.294	0.459	0.185
OPG	0.476	0.418	0.386	0.216	UCRP	0.637	0.575	0.487	0.189	SPINT2 ^a	0.384	0.272	0.484	0.448
Rb	0.470	0.383	0.404	0.337	PSME1	0.634	0.506	0.452	0.116	Calpastatin	0.381	0.303	0.321	0.039
PDE3A	0.468	0.399	0.387	0.329	GFRa‐1	0.619	0.564	0.375	0.110	ApoM	0.377	0.297	0.304	0.087
HPLN1	0.467	0.470	0.430	0.238	IgA	0.616	0.577	0.381	0.238	HDGR2	0.377	0.287	0.423	0.064
LD78‐β	0.462	0.401	0.369	0.253	CD48 ^a	0.603	0.582	0.518	0.228	BCL2‐like 1 protein ^a	0.374	0.310	0.464	0.023
DYRK3	0.461	0.395	0.371	0.304	Calgranulin B ^b	0.603	0.597	0.381	0.143	COLEC12 ^a	0.374	0.245	0.439	0.085
Leptin ^a	0.459	0.433	0.455	0.321	Histone H1.2	0.592	0.570	0.544	0.158	MPIF‐1 ^a	0.374	0.306	0.271	0.083
TFF1	0.459	0.369	0.416	0.323	ABL2	0.586	0.656	0.522	0.276	Eotaxin	0.373	0.241	0.484	0.196
Ficolin‐3	0.457	0.440	0.464	0.307	Nectin‐like protein 1	0.585	0.546	0.435	0.184	PDE3A	0.368	0.271	0.419	−0.008
MASP3	0.453	0.395	0.371	0.355	NDKA	0.581	0.618	0.452	0.153	MMP‐16 ^a	0.366	0.270	0.424	−0.037

Open in a new tab

Spearman correlation coefficients are indicated in bold and if they showed statistical significance (P < 0.05). AIF1, allograft inflammatory factor 1; b‐NGF, β‐nerve growth factor; C/FC crescents, cellular/fibrocellular crescents; CFC, colony‐forming cell; HMGN1, high mobility group nucleosome binding domain 1; SOD, superoxide dismutase.

^{^a}

The urine proteins are shown with an indication of statistical significance if their urinary concentrations were significantly higher in patients with diabetic nephropathy than in those with LN (P < 0.05).

^{^b}

The urine proteins are shown with an indication of statistical significance if their urinary concentrations were significantly higher in patients with LN than in those with diabetic nephropathy (P < 0.05).

The number of correlated candidate urinary proteins varied according to the renal histology. For active glomerular lesions resulting mainly from inflammatory mechanisms, urine proteins with a correlation of ρ > 0.4 were frequently detected in glomerular lesions, including cellular or fibrocellular crescents in cluster 1 and endocapillary hypercellularity in cluster 2. Meanwhile, no urinary protein was correlated with subendothelial deposition in cluster 2. There were also no positive correlations among urinary proteins that could be detected in basement membrane lesions in cluster 3.

In the analysis of UG1 and UG2, a comparison with diabetic nephropathy was made to exclude proteins that were not specifically increased in the urine, even in nonnephritis pathologies of the glomerulus. Because the scores of interstitial inflammation and interstitial fibrosis correlated with each other (ρ = 0.59, P < 0.01), we focused on proteins that correlated with only one of the histologic scores in the analysis of UG4. To increase the detection rate of ELISA, proteins with an average intensity of <1,000 RFU in 24 cases were excluded. Under these conditions, the top 15 proteins with high correlation coefficients in the most prevalent renal histologic findings in UG1, 2, and 4 were selected as candidate proteins (Supplementary Table 6). Among the urinary proteins in UG1 and UG2, only calgranulin B levels were significantly higher in the LN group than in the diabetic nephropathy group (P = 0.02). In addition to calgranulin B, S100A12, HMGN1, and IL‐8 were selected in UG1 and AIF‐1, SOD1, S100A12, and HMGN1 were selected in UG2 as candidate proteins that did not significantly increase but tended to increase in urinary concentrations in patients with LN compared to those with diabetic nephropathy. In UG4, MCP‐1 (ρ = 0.72, P < 0.001) and FSTL3 (ρ = 0.69, P < 0.001) were selected as candidate proteins correlated with interstitial inflammation scores. Only IGFBP‐5 was extracted as a candidate protein that correlated solely with interstitial fibrosis scores (ρ = 0.41, P = 0.048) and not with interstitial inflammation scores. No protein correlated only with the interstitial inflammation scores.

The urinary concentrations of these nine candidate proteins were measured using ELISA and reanalyzed for correlation with the histologic scores. The results for S100A12, HMGN1, IL‐8, AIF‐1, SOD1, and FSTL3 were less sensitive or did not correlate with the results of the aptamer‐based screening (Supplementary Table 7). High correlations were found among the results obtained by ELISA and aptamer‐based screening for calgranulin B, MCP‐1, and IGFBP‐5 (r > 0.7; Figure 4A). The proteins, measured by ELISA, were then analyzed for correlation with the histologic scores. Urinary calgranulin B levels correlated with the histologic scores of cellular or fibrocellular crescents (r = 0.65, P < 0.001) and endocapillary hypercellularity (r = 0.72, P < 0.001), representative of clusters 1 and 2 (Figure 4B). Urinary calgranulin B levels also correlated with the histologic scores of neutrophil infiltration (r = 0.58, P = 0.003), karyorrhexis (r = 0.68, P < 0.001), and fibrinoid necrosis (r = 0.56, P = 0.005) in cluster 2 (data not shown). MCP‐1 and IGFBP‐5 urinary levels correlated with interstitial inflammation (r = 0.65, P < 0.001) and fibrosis scores (r = 0.47, P = 0.018), respectively, in the 24 patients with LN (Figure 4B). In addition, a stronger correlation was observed for IGFBP‐5 when three cases of diabetic nephropathy with significant interstitial fibrosis were included (r = 0.83, P < 0.001) (data not shown). The correlation between each serum protein concentration detected by aptamer‐based screening and the renal histologic score was also analyzed in the discovery cohort; however, serum calgranulin and cellular or fibrocellular crescents (ρ = 0.26, P = 0.22), endocapillary hypercellularity (ρ = 0.22, P = 0.29), neutrophil infiltration (ρ = 0.07, P = 0.73), karyorrhexis (ρ = 0.03, P = 0.89) or fibrinoid necrosis (ρ = 0.02, P = 0.93), serum MCP‐1 and interstitial inflammation scores (ρ = 0.23, P = 0.27), and serum IGFBP‐5 and interstitial fibrosis scores (ρ = 0.25, P = 0.24) showed no significant correlation with each other (Supplementary File).

ELISA validation of urinary calgranulin B, MCP‐1, and IGFBP‐5 levels in the discovery and validation cohort. (A) Correlation between aptamer screening and ELISA results of urine samples from 24 patients with LN. (B and C) Correlation among urinary calgranulin B, MCP‐1, and IGFBP‐5 levels and the histologic scores of cellular or fibrocellular crescents and endocapillary hypercellularity, interstitial inflammation, and interstitial fibrosis, respectively, in the discovery (B) and validation (C) cohorts. (D) Receiver operating characteristic curves for the presence of cellular or fibrocellular crescents and endocapillary hypercellularity using urinary calgranulin B level, interstitial inflammation using urinary MCP‐1 level, and interstitial fibrosis using urinary IGFBP‐5 level in the validation cohort. (E) Correlation among urinary calgranulin B and MCP‐1 levels and activity index and IGFBP‐5 level and chronicity index in the validation cohort. The plots show the concentrations of each urinary protein. Urinary protein concentrations are standardized using urinary creatinine levels. AUC, area under curve; ELISA, enzyme‐linked immunosorbent assay; IGFBP, insulin‐like growth factor binding protein; MCP, membrane cofactor protein; RFU, relative fluorescence units.

Validation of the three urinary proteins associated with the individual histologic score in the validation cohort

The correlation between ELISA‐measured urinary protein concentrations and each histologic score was confirmed in the validation cohort for calgranulin B, MCP‐1, and IGFBP‐5. Urinary calgranulin B levels correlated with the histologic scores of endocapillary hypercellularity (r = 0.55, P = 0.001) (Figure 4C). However, the validation cohort had fewer cases of cellular or fibrocellular crescents (Supplementary Table 3). Therefore, we could not confirm a significant correlation between the urinary calgranulin B levels and histologic scores (r = 0.41, P = 0.18) (Figure 4C). The urinary MCP‐1 levels correlated with the histologic scores of interstitial inflammation (r = 0.78, P < 0.001) (Figure 4C). In the eight cases in which measurement of the urinary IGFBP‐5 levels was possible, there was a tendency to correlate with the histologic scores of interstitial fibrosis, but no significant difference was observed (r = 0.63, P = 0.09) (Figure 4C). The correlation coefficients among all histologic scores of clusters 1, 2, and 4 and ELISA results of these three urinary protein concentrations in the discovery cohort, validation cohort, and diabetic nephropathy were shown in a heatmap (Supplementary Figure 4).

Using the results of the validation cohort, ROC analysis was performed to determine whether urinary calgranulin B, MCP‐1, and IGFBP‐5 levels could discriminate the presence or absence of each histologic finding (Figure 4D). These three urinary proteins exhibited moderate accuracy: calgranulin B with areas under the curve (AUCs) of 0.78 for cellular or fibrocellular crescents and 0.87 for endocapillary hypercellularity, MCP‐1 with an AUC of 0.84 for interstitial inflammation, and IGFBP‐5 with an AUC of 0.71 for interstitial fibrosis. In addition, it was confirmed in the validation cohort that urinary calgranulin B levels showed a strong correlation with activity index (r = 0.63, P < 0.001), but urinary MCP‐1 levels did not (r = 0.43, P = 0.04). Urinary IGFBP‐5 levels were not significantly correlated with chronicity index (r = 0.44, P = 0.27) (Figure 4E).

Immunohistochemical staining of the renal tissues

Immunohistochemical staining of the renal tissues was performed to clarify the localization of calgranulin B, MCP‐1, and IGFBP‐5. Calgranulin B proteins were localized to the cells infiltrating the glomerulus with endocapillary hypercellularity and cellular or fibrocellular crescents but not to the cells in tubulointerstitial lesions. In addition, glomeruli without these active lesions showed little expression of calgranulin B (Figure 5A–C). In the renal tissues with high histologic scores for interstitial inflammation, MCP‐1 proteins were localized on the tubular epithelial cells but not on the cells infiltrating tubulointerstitial lesions (Figure 5D–F). IGFBP‐5 proteins were localized on the spindle‐shaped fibroblasts present at sites of interstitial fibrosis and normal tubular epithelial cells (Figure 5G–I). Neither MCP‐1 nor IGFBP‐5 was localized in the glomeruli.

DISCUSSION

The previously reported proteome analyses aimed at comparing patients with LN and healthy controls or those with active and inactive LN to identify biomarkers that could distinguish them from each other. ⁹ , ¹⁰ , ¹¹ The present study is novel in that, via a proteomic analysis, we identified urinary biomarkers in patients with LN that correlate with a renal histologic scoring system to precisely reflect the characteristics and progression of individual active and chronic lesions. In this study, pathway analyses revealed the protein and cell‐to‐cell relationships associated with each renal histologic finding. Neutrophils and monocytes are the dominant immune cell type in LN and have been implicated in its pathogenesis. ¹⁹ , ²⁰ This study corroborates previous findings by demonstrating that the activation of granulocytes and monocytes is involved in both glomerular and tubulointerstitial lesions. Furthermore, a stronger association of proteins related to neutrophils and monocytes is observed in active endocapillary and extracapillary lesions (ie, S100A12, calgranulin B, IL‐8, and azurocidin). This finding is consistent with previous reports in LN with active glomerulonephritis. ¹⁹ In contrast, an enhancement of various kinds of IL‐17–related signals was observed in tubulointerstitial lesions. In lupus model mice, IL‐17 increased the expression of molecules such as MCP‐1 in renal tubular epithelial cells (RTECs), thereby promoting the renal infiltration of macrophages. ²¹ Furthermore, IL‐17 stimulation of RTECs increased messenger RNA expression of other chemokines, such as Cxcl1, Cxcl2, and Cxcl8, which attract monocytes and neutrophils. ²² Consequently, under IL‐17 stimulation, RTECs produce mediators that recruit dendritic cells and macrophages, which are crucial sources of transforming growth factor β that promote renal fibrosis, ²³ establishing IL‐17 as a key driver of RTEC‐mediated immunopathogenesis in LN. Thus, these findings support the results obtained from pathway analyses of interstitial inflammation, tubular atrophy, and interstitial fibrosis in this study.

Calgranulin B (S100A9) belongs to the S100 family of proteins and exists as a heterodimer of S100A8 in vivo. S100A8/A9 are mainly expressed in the cytoplasm of neutrophils, monocytes, plasmacytoid dendritic cells, and endothelial cells and play a proinflammatory role in various autoimmune diseases. ²⁴ , ²⁵ Serum and urinary S100A8/A9 concentrations are known to increase in patients with active LN. ²⁶ , ²⁷ , ²⁸ The present study newly showed a correlation between urinary calgranulin B levels and quantified active intraglomerular lesions in the LN. In addition, immunohistochemical staining of the renal tissues from patients with high urinary calgranulin B concentrations showed the localization of calgranulin B in cells infiltrating the glomerulus. These results suggest that urinary calgranulin B levels may reflect local renal pathology, making it more important to measure urinary calgranulin B concentrations.

MCP‐1 (CCL2) is a chemokine secreted by monocytes, macrophages, fibroblasts, and vascular endothelial cells that contributes to monocyte activation and migration via CCR2. ²⁹ MCP‐1 is a useful urinary biomarker for predicting disease activity and renal prognosis in patients with LN. ³⁰ , ³¹ Urinary MCP‐1 has also been reported as a biomarker that can predict inflammatory cell infiltration of the tubulointerstitium in patients with LN. ³² The present study demonstrated a correlation between quantified interstitial inflammation and urinary MCP‐1 concentration, indicating its potential as a detailed predictor of the severity of interstitial inflammation in patients with LN. MCP‐1 was strongly expressed in the tubular epithelial cells in the renal tissues, where interstitial inflammation was more prominent. This indicates that MCP‐1 expression in the tubular epithelial cells promotes inflammatory cell infiltration into the renal interstitium. However, urinary MCP‐1 levels did not significantly correlate with the severity of active glomerular lesions in the present study. This result reinforces the possibility that urinary MCP‐1 levels can specifically predict the severity of interstitial inflammation.

IGFBP‐5 belongs to the IGFBP family and contributes to the regulation of insulin‐like growth factor signaling. It has been reported to be involved in renal diseases, particularly chronic kidney disease (CKD), after its expression was discovered in the renal interstitium of patients with CKD using single‐cell analysis. ³³ , ³⁴ IGFBP‐5 has also been reported to promote lung tissue fibrosis and participate in tissue remodeling. ³⁵ We newly reported that urinary IGFBP‐5 levels can predict the severity of interstitial fibrosis in the renal tissues. Immunohistochemical staining showed IGFBP‐5 staining in the cells that appeared to be fibroblasts in lesions with interstitial fibrosis, suggesting that it is associated with the pathogenesis of fibrosis.

Some urinary proteins, such as vascular cell adhesion molecule 1 and activated leukocyte cell adhesion molecule, have been reported to correlate with activity index in patients with active LN. ³⁶ , ³⁷ Urinary calgranulin B also predicted activity index with as strong a correlation as these proteins in our study. In contrast, urinary levels of MCP‐1 and IGFBP‐5 did not show a strong correlation with activity or chronicity index. This is probably explained by the finding that urinary MCP‐1 and IGFBP‐5 were associated with only one histologic finding of all components in activity or chronicity index, whereas urinary calgranulin B was associated with the major components of histologic findings in activity index.

This study had some limitations. First, our scoring system was not clinically validated, although it comprehensively included histologic findings based on the ISN/RPS classification and the NIH index. Second, the validation failed to confirm the significance of the correlation between urinary calgranulin B levels and the histologic scores of cellular or fibrocellular crescents or urinary IGFBP‐5 levels and the histologic scores of interstitial fibrosis. The former occurred because fewer patients in the validation cohort presented with cellular or fibrocellular crescents. The latter was due to the small number of cases in which IGFBP‐5 levels could be measured. Further validation using urine samples from a larger sample is required. Third, we did not perform analyses of urinary proteins associated with the histologic findings included in the clusters of membranous and mesangial lesions and other lesions. Fourth, the clinical usefulness of urinary calgranulin B, MCP‐1, and IGFBP‐5 levels was not demonstrated. A longitudinal study is warranted to investigate the relationship between these three urinary proteins and disease progression or improvement over time. Fifth, the present study could not include sufficient samples of diabetic nephropathy because renal biopsies are not commonly performed in patients with diabetic nephropathy. The limited samples of diabetic nephropathy could affect the accuracy of candidate protein extraction.

In conclusion, various urinary proteins that correlate with certain renal histologic findings may reflect the different pathogeneses involved in each renal lesion. The extraction and identification of these urinary proteins may enhance clinical management and are expected to be useful in precision medicine.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr Kaneko had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design

Hiramoto, Saito, Hanaoka, Hashiguchi, Suzuki, Takeuchi, Kaneko.

Acquisition of data

Hiramoto, Saito, Hanaoka, Kikuchi, Fukui, Hashiguchi.

Analysis and interpretation of data

Hiramoto, Saito, Hashiguchi.

ROLE OF THE STUDY SPONSOR

Mitsubishi Tanabe Pharma Corporation had no role in the study design or in the collection or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Mitsubishi Tanabe Pharma Corporation.

Supporting information

Disclosure form

ART-77-298-s003.pdf^{(520.5KB, pdf)}

Appendix S1: Supplementary Information.

ART-77-298-s001.docx^{(11.3MB, docx)}

File S1: The correlation between each 1305 serum protein concentration detected by aptamer‐based screening and the renal histological score in the discovery cohort.

ART-77-298-s002.xlsx^{(784.5KB, xlsx)}

ACKNOWLEDGMENTS

We thank Ms. Harumi Kondo for assistance with collecting clinical data. IPA data were provided to the authors with explicit permission from QIAGEN, Inc. We would like to thank Editage (www.editage.jp) for English language editing. We also thank Ryo Takano (former employee of Mitsubishi Tanabe Pharma Corporation), Shinji Kojima, and Fumihiko Miyoshi (Mitsubishi Tanabe Pharma Corporation) for bioinformatics analysis.

Supported by Mitsubishi Tanabe Pharma Corporation.

Drs Hiramoto and Saito contributed equally to this work.

Additional supplementary information cited in this article can be found online in the Supporting Information section (http://onlinelibrary.wiley.com/doi/10.1002/art.43017).

Author disclosures are available at https://onlinelibrary.wiley.com/doi/10.1002/art.43017.

REFERENCES

1. Dörner T, Furie R. Novel paradigms in systemic lupus erythematosus. Lancet 2019;393(10188):2344–2358. [DOI] [PubMed] [Google Scholar]
2. Cameron JS. Lupus nephritis. J Am Soc Nephrol 1999;10(2):413–424. [DOI] [PubMed] [Google Scholar]
3. Borchers AT, Leibushor N, Naguwa SM, et al. Lupus nephritis: a critical review. Autoimmun Rev 2012;12(2):174–194. [DOI] [PubMed] [Google Scholar]
4. Azoicăi T, Belibou IM, Lozneanu L, et al. Large variability of the activity and chronicity indexes within and between histological classes of lupus nephritis. Rom J Morphol Embryol 2017;58(1):73–78. [PubMed] [Google Scholar]
5. Bajema IM, Wilhelmus S, Alpers CE, et al. Revision of the International Society of Nephrology/Renal Pathology Society classification for lupus nephritis: clarification of definitions, and modified National Institutes of Health activity and chronicity indices. Kidney Int 2018;93(4):789–796. [DOI] [PubMed] [Google Scholar]
6. Rijnink EC, Teng YKO, Wilhelmus S, et al. Clinical and histopathologic characteristics associated with renal outcomes in lupus nephritis. Clin J Am Soc Nephrol 2017;12(5):734–743. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Yu F, Wu LH, Tan Y, et al. Tubulointerstitial lesions of patients with lupus nephritis classified by the 2003 International Society of Nephrology and Renal Pathology Society system. Kidney Int 2010;77(9):820–829. [DOI] [PubMed] [Google Scholar]
8. Giannico G, Fogo AB. Lupus nephritis: is the kidney biopsy currently necessary in the management of lupus nephritis? Clin J Am Soc Nephrol 2013;8(1):138–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Aragón CC, Tafúr RA, Suárez‐Avellaneda A, et al. Urinary biomarkers in lupus nephritis. J Transl Autoimmun 2020;3:100042. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Zhang X, Jin M, Wu H, et al. Biomarkers of lupus nephritis determined by serial urine proteomics. Kidney Int 2008;74(6):799–807. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Somparn P, Hirankarn N, Leelahavanichkul A, et al. Urinary proteomics revealed prostaglandin H(2)D‐isomerase, not Zn‐α2‐glycoprotein, as a biomarker for active lupus nephritis. J Proteomics 2012;75(11):3240–3247. [DOI] [PubMed] [Google Scholar]
12. Stanley S, Vanarsa K, Soliman S, et al. Comprehensive aptamer‐based screening identifies a spectrum of urinary biomarkers of lupus nephritis across ethnicities. Nat Commun 2020;11(1):2197. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Vanarsa K, Soomro S, Zhang T, et al. Quantitative planar array screen of 1000 proteins uncovers novel urinary protein biomarkers of lupus nephritis. Ann Rheum Dis 2020;79(10):1349–1361. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Fava A, Rao DA, Mohan C, et al; Accelerating Medicines Partnership in Rheumatoid Arthritis and Systemic Lupus Erythematosus Network. Urine proteomics and renal single‐cell transcriptomics implicate interleukin‐16 in lupus nephritis. Arthritis Rheumatol 2022;74(5):829–839. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Fava A, Buyon J, Magder L, et al; Accelerating Medicines Partnership in RA/SLE network. Urine proteomic signatures of histological class, activity, chronicity, and treatment response in lupus nephritis. JCI Insight 2024;9(2):e172569. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. SomaLogic . SOMAscan Proteomic Assay Technical White Paper. SomaLogic, Inc; 2015.
17. Weening JJ, D'Agati VD, Schwartz MM, et al; International Society of Nephrology Working Group on the Classification of Lupus Nephritis; Renal Pathology Society Working Group on the Classification of Lupus Nephritis. The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int 2004;65(2):521–530. [DOI] [PubMed] [Google Scholar]
18. Austin HA III, Muenz LR, Joyce KM, et al. Diffuse proliferative lupus nephritis: identification of specific pathologic features affecting renal outcome. Kidney Int 1984;25(4):689–695. [DOI] [PubMed] [Google Scholar]
19. Banchereau R, Hong S, Cantarel B, et al. Personalized immunomonitoring uncovers molecular networks that stratify lupus patients. Cell 2016;165(3):551–565. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Olmes G, Büttner‐Herold M, Ferrazzi F, et al. CD163+ M2c‐like macrophages predominate in renal biopsies from patients with lupus nephritis. Arthritis Res Ther 2016;18(1):90. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Ramani K, Biswas PS. Interleukin 17 signaling drives type I interferon induced proliferative crescentic glomerulonephritis in lupus‐prone mice. Clin Immunol 2016;162:31–36. [DOI] [PubMed] [Google Scholar]
22. Ramani K, Pawaria S, Maers K, et al. An essential role of interleukin‐17 receptor signaling in the development of autoimmune glomerulonephritis. J Leukoc Biol 2014;96(3):463–472. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Kassianos AJ, Wang X, Sampangi S, et al. Increased tubulointerstitial recruitment of human CD141(hi) CLEC9A(+) and CD1c(+) myeloid dendritic cell subsets in renal fibrosis and chronic kidney disease. Am J Physiol Renal Physiol 2013;305(10):F1391–F1401. [DOI] [PubMed] [Google Scholar]
24. Sedaghat F, Notopoulos A. S100 protein family and its application in clinical practice. Hippokratia 2008;12(4):198–204. [PMC free article] [PubMed] [Google Scholar]
25. Wang S, Song R, Wang Z, et al. S100A8/A9 in Inflammation. Front Immunol 2018;9:1298. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Turnier JL, Fall N, Thornton S, et al. Urine S100 proteins as potential biomarkers of lupus nephritis activity. Arthritis Res Ther 2017;19(1):242. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Davies JC, Midgley A, Carlsson E, et al. Urine and serum S100A8/A9 and S100A12 associate with active lupus nephritis and may predict response to rituximab treatment. RMD Open 2020;6(2):e001257. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Tantivitayakul P, Benjachat T, Somparn P, et al. Elevated expressions of myeloid‐related proteins‐8 and ‐14 are danger biomarkers for lupus nephritis. Lupus 2016;25(1):38–45. [DOI] [PubMed] [Google Scholar]
29. Singh S, Anshita D, Ravichandiran V. MCP‐1: Function, regulation, and involvement in disease. Int Immunopharmacol 2021;101(Pt B):107598. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Rovin BH, Song H, Birmingham DJ, et al. Urine chemokines as biomarkers of human systemic lupus erythematosus activity. J Am Soc Nephrol 2005;16(2):467–473. [DOI] [PubMed] [Google Scholar]
31. Gulati G, Bennett MR, Abulaban K, et al. Prospective validation of a novel renal activity index of lupus nephritis. Lupus 2017;26(9):927–936. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Zhang X, Nagaraja HN, Nadasdy T, et al. A composite urine biomarker reflects interstitial inflammation in lupus nephritis kidney biopsies. Kidney Int 2012;81(4):401–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Wang S, Chi K, Wu D, et al. Insulin‐like growth factor binding proteins in kidney disease. Front Pharmacol 2021;12:807119. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Park J, Shrestha R, Qiu C, et al. Single‐cell transcriptomics of the mouse kidney reveals potential cellular targets of kidney disease. Science 2018;360(6390):758–763. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Nguyen XX, Muhammad L, Nietert PJ, et al. IGFBP‐5 promotes fibrosis via increasing its own expression and that of other pro‐fibrotic mediators. Front Endocrinol (Lausanne) 2018;9:601. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Singh S, Wu T, Xie C, et al. Urine VCAM‐1 as a marker of renal pathology activity index in lupus nephritis. Arthritis Res Ther 2012;14(4):R164. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Ding H, Lin C, Cai J, et al. Urinary activated leukocyte cell adhesion molecule as a novel biomarker of lupus nephritis histology. Arthritis Res Ther 2020;22(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Disclosure form

ART-77-298-s003.pdf^{(520.5KB, pdf)}

Appendix S1: Supplementary Information.

ART-77-298-s001.docx^{(11.3MB, docx)}

File S1: The correlation between each 1305 serum protein concentration detected by aptamer‐based screening and the renal histological score in the discovery cohort.

ART-77-298-s002.xlsx^{(784.5KB, xlsx)}

[art43017-bib-0001] 1. Dörner T, Furie R. Novel paradigms in systemic lupus erythematosus. Lancet 2019;393(10188):2344–2358. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0002] 2. Cameron JS. Lupus nephritis. J Am Soc Nephrol 1999;10(2):413–424. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0003] 3. Borchers AT, Leibushor N, Naguwa SM, et al. Lupus nephritis: a critical review. Autoimmun Rev 2012;12(2):174–194. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0004] 4. Azoicăi T, Belibou IM, Lozneanu L, et al. Large variability of the activity and chronicity indexes within and between histological classes of lupus nephritis. Rom J Morphol Embryol 2017;58(1):73–78. [PubMed] [Google Scholar]

[art43017-bib-0005] 5. Bajema IM, Wilhelmus S, Alpers CE, et al. Revision of the International Society of Nephrology/Renal Pathology Society classification for lupus nephritis: clarification of definitions, and modified National Institutes of Health activity and chronicity indices. Kidney Int 2018;93(4):789–796. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0006] 6. Rijnink EC, Teng YKO, Wilhelmus S, et al. Clinical and histopathologic characteristics associated with renal outcomes in lupus nephritis. Clin J Am Soc Nephrol 2017;12(5):734–743. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0007] 7. Yu F, Wu LH, Tan Y, et al. Tubulointerstitial lesions of patients with lupus nephritis classified by the 2003 International Society of Nephrology and Renal Pathology Society system. Kidney Int 2010;77(9):820–829. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0008] 8. Giannico G, Fogo AB. Lupus nephritis: is the kidney biopsy currently necessary in the management of lupus nephritis? Clin J Am Soc Nephrol 2013;8(1):138–145. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0009] 9. Aragón CC, Tafúr RA, Suárez‐Avellaneda A, et al. Urinary biomarkers in lupus nephritis. J Transl Autoimmun 2020;3:100042. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0010] 10. Zhang X, Jin M, Wu H, et al. Biomarkers of lupus nephritis determined by serial urine proteomics. Kidney Int 2008;74(6):799–807. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0011] 11. Somparn P, Hirankarn N, Leelahavanichkul A, et al. Urinary proteomics revealed prostaglandin H(2)D‐isomerase, not Zn‐α2‐glycoprotein, as a biomarker for active lupus nephritis. J Proteomics 2012;75(11):3240–3247. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0012] 12. Stanley S, Vanarsa K, Soliman S, et al. Comprehensive aptamer‐based screening identifies a spectrum of urinary biomarkers of lupus nephritis across ethnicities. Nat Commun 2020;11(1):2197. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0013] 13. Vanarsa K, Soomro S, Zhang T, et al. Quantitative planar array screen of 1000 proteins uncovers novel urinary protein biomarkers of lupus nephritis. Ann Rheum Dis 2020;79(10):1349–1361. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0014] 14. Fava A, Rao DA, Mohan C, et al; Accelerating Medicines Partnership in Rheumatoid Arthritis and Systemic Lupus Erythematosus Network. Urine proteomics and renal single‐cell transcriptomics implicate interleukin‐16 in lupus nephritis. Arthritis Rheumatol 2022;74(5):829–839. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0015] 15. Fava A, Buyon J, Magder L, et al; Accelerating Medicines Partnership in RA/SLE network. Urine proteomic signatures of histological class, activity, chronicity, and treatment response in lupus nephritis. JCI Insight 2024;9(2):e172569. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0016] 16. SomaLogic . SOMAscan Proteomic Assay Technical White Paper. SomaLogic, Inc; 2015.

[art43017-bib-0017] 17. Weening JJ, D'Agati VD, Schwartz MM, et al; International Society of Nephrology Working Group on the Classification of Lupus Nephritis; Renal Pathology Society Working Group on the Classification of Lupus Nephritis. The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int 2004;65(2):521–530. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0018] 18. Austin HA III, Muenz LR, Joyce KM, et al. Diffuse proliferative lupus nephritis: identification of specific pathologic features affecting renal outcome. Kidney Int 1984;25(4):689–695. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0019] 19. Banchereau R, Hong S, Cantarel B, et al. Personalized immunomonitoring uncovers molecular networks that stratify lupus patients. Cell 2016;165(3):551–565. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0020] 20. Olmes G, Büttner‐Herold M, Ferrazzi F, et al. CD163+ M2c‐like macrophages predominate in renal biopsies from patients with lupus nephritis. Arthritis Res Ther 2016;18(1):90. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0021] 21. Ramani K, Biswas PS. Interleukin 17 signaling drives type I interferon induced proliferative crescentic glomerulonephritis in lupus‐prone mice. Clin Immunol 2016;162:31–36. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0022] 22. Ramani K, Pawaria S, Maers K, et al. An essential role of interleukin‐17 receptor signaling in the development of autoimmune glomerulonephritis. J Leukoc Biol 2014;96(3):463–472. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0023] 23. Kassianos AJ, Wang X, Sampangi S, et al. Increased tubulointerstitial recruitment of human CD141(hi) CLEC9A(+) and CD1c(+) myeloid dendritic cell subsets in renal fibrosis and chronic kidney disease. Am J Physiol Renal Physiol 2013;305(10):F1391–F1401. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0024] 24. Sedaghat F, Notopoulos A. S100 protein family and its application in clinical practice. Hippokratia 2008;12(4):198–204. [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0025] 25. Wang S, Song R, Wang Z, et al. S100A8/A9 in Inflammation. Front Immunol 2018;9:1298. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0026] 26. Turnier JL, Fall N, Thornton S, et al. Urine S100 proteins as potential biomarkers of lupus nephritis activity. Arthritis Res Ther 2017;19(1):242. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0027] 27. Davies JC, Midgley A, Carlsson E, et al. Urine and serum S100A8/A9 and S100A12 associate with active lupus nephritis and may predict response to rituximab treatment. RMD Open 2020;6(2):e001257. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0028] 28. Tantivitayakul P, Benjachat T, Somparn P, et al. Elevated expressions of myeloid‐related proteins‐8 and ‐14 are danger biomarkers for lupus nephritis. Lupus 2016;25(1):38–45. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0029] 29. Singh S, Anshita D, Ravichandiran V. MCP‐1: Function, regulation, and involvement in disease. Int Immunopharmacol 2021;101(Pt B):107598. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0030] 30. Rovin BH, Song H, Birmingham DJ, et al. Urine chemokines as biomarkers of human systemic lupus erythematosus activity. J Am Soc Nephrol 2005;16(2):467–473. [DOI] [PubMed] [Google Scholar]

[art43017-bib-0031] 31. Gulati G, Bennett MR, Abulaban K, et al. Prospective validation of a novel renal activity index of lupus nephritis. Lupus 2017;26(9):927–936. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0032] 32. Zhang X, Nagaraja HN, Nadasdy T, et al. A composite urine biomarker reflects interstitial inflammation in lupus nephritis kidney biopsies. Kidney Int 2012;81(4):401–406. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0033] 33. Wang S, Chi K, Wu D, et al. Insulin‐like growth factor binding proteins in kidney disease. Front Pharmacol 2021;12:807119. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0034] 34. Park J, Shrestha R, Qiu C, et al. Single‐cell transcriptomics of the mouse kidney reveals potential cellular targets of kidney disease. Science 2018;360(6390):758–763. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0035] 35. Nguyen XX, Muhammad L, Nietert PJ, et al. IGFBP‐5 promotes fibrosis via increasing its own expression and that of other pro‐fibrotic mediators. Front Endocrinol (Lausanne) 2018;9:601. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0036] 36. Singh S, Wu T, Xie C, et al. Urine VCAM‐1 as a marker of renal pathology activity index in lupus nephritis. Arthritis Res Ther 2012;14(4):R164. [DOI] [PMC free article] [PubMed] [Google Scholar]

[art43017-bib-0037] 37. Ding H, Lin C, Cai J, et al. Urinary activated leukocyte cell adhesion molecule as a novel biomarker of lupus nephritis histology. Arthritis Res Ther 2020;22(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Urinary Biomarkers Associated With Pathogenic Pathways Reflecting Histologic Findings in Lupus Nephritis

Kazuoto Hiramoto

Shuntaro Saito

Hironari Hanaoka

Jun Kikuchi

Hiroyuki Fukui

Akinori Hashiguchi

Katsuya Suzuki

Tsutomu Takeuchi

Yuko Kaneko

Abstract

Objective

Methods

Results

Conclusion

INTRODUCTION

PATIENTS AND METHODS

Patients and sample preparation

Aptamer‐based screening

Histologic scoring system

Validation studies using ELISA

Immunohistochemistry

Data analysis

RESULTS

Cluster analysis with histologic score and aptamer‐based protein screening

Figure 1.

Cell‐type enrichment analysis

Figure 2.

IPA

PPI analysis

Figure 3.

Identification of urinary biomarker protein correlated with individual histologic score

Table 1.

Figure 4.

Validation of the three urinary proteins associated with the individual histologic score in the validation cohort

Immunohistochemical staining of the renal tissues

Figure 5.

DISCUSSION

AUTHOR CONTRIBUTIONS

Study conception and design

Acquisition of data

Analysis and interpretation of data

ROLE OF THE STUDY SPONSOR

Supporting information

ACKNOWLEDGMENTS

REFERENCES

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases