The Urinary Exosomal miRNA Expression Profile is Predictive of Clinical Response in Lupus Nephritis

Abstract

Data on exosomal-derived urinary miRNAs have identified several miRNAs associated with disease activity and fibrosis formation, but studies on prognosis are lacking. We conducted a qPCR array screening on urinary exosomes from 14 patients with biopsy-proven proliferative lupus glomerulonephritis with a renal outcome of clinical response (n = 7) and non-response (n = 7) following therapy. Validation studies were performed by qRT-PCR in a new lupus nephritis (LN) cohort (responders = 22 and non-responders = 21). Responder patients expressed significantly increased levels of miR-31, miR-107, and miR-135b-5p in urine and renal tissue compared to non-responders. MiR-135b exhibited the best predictive value to discriminate responder patients (area under the curve = 0.783). In vitro studies showed exosome-derived miR-31, miR-107, and miR-135b-5p expression to be mainly produced by tubular renal cells stimulated with inflammatory cytokines (e.g IL1, TNFα, IFNα and IL6). Uptake of urinary exosomes from responders by mesangial cells was superior compared to that from non-responders (90% vs. 50%, p < 0.0001). HIF1A was identified as a potential common target, and low protein levels were found in non-responder renal biopsies. HIF1A inhibition reduced mesangial proliferation and IL-8, CCL2, CCL3, and CXCL1 mesangial cell production and IL-6/VCAM-1 in endothelial cells. Urinary exosomal miR-135b-5p, miR-107, and miR-31 are promising novel markers for clinical outcomes, regulating LN renal recovery by HIF1A inhibition.

Keywords: biomarkers; lupus nephritis; response; therapy; urinary exosomes.

Figure 1

Urinary exosomal miRNA expression profile from lupus nephritis (LN) patients according to their clinical response. Expression of differentially expressed miRNAs was evaluated using quantitative real-time PCR (qRT-PCR) in the screening cohort and in the validation cohort of lupus nephritis patients (responders (n = 29) and non-responders (n = 28)). (A) Volcano plot shows the most differentially and significantly expressed miRNAs found in the screening cohort. Three of these mRNAs were validated (marked in bold). DE: differentially expressed. (B) Paired t-test analysis was done between flare and post-treatment time points in the responder group. Significant differences are shown with the corresponding p-values. (C) MiR-31-5p, miR-107, and miR-135b-5p expression from the LN responder and non-responder groups at flare time or post-treatment time are shown as individual plots. Gene expression was normalized using U6 as endogenous control. Fold change in expression level was calculated using the 2^−ΔΔCt method. p-Values were obtained using Student’s t-test. (D) Correlation of corresponding miRNAs expression levels with proteinuria and activity index at flare time. Spearman’s rank-order correlation was used to obtain r- and p-values.

Figure 2

Differentially expressed microRNAs (miRNAs) according to treatment response. Receiver operator characteristic analysis (ROC) of individual miRNAs to distinguish responder from non-responder patient groups. Analysis from samples obtained at flare time (A) and at post-treatment time (B). The value of the area under the curve (AUC) is shown in each plot.

Figure 3

In situ hybridization of miR-31-5p, miR-135b-5p, and miR-107 in LN renal biopsies. (A) At flare time, renal tissue from lupus nephritis was analyzed to identify miR-31-5p, miR-135b-5p (both in green), and miR-107 (red). Significant differences were found between the responder and non-responder groups. (B) Tubular localization was found for miR-31 and miR-107 in the responder group. (C) No differences in tubular atrophy were observed between the responder and non-responder groups in histology analysis. Scale bar = 20 µm. White arrows marked positive cells. p-Values were obtained using Student’s t-test. NS: not significant, * p < 0.05; ** p < 0.005; *** p < 0.0005.

Figure 4

MiR-31-5p, miR-107, and miR-135b-5p urinary exosome origin and their target cell delivery. (A) Scheme of in vitro experiments performed to evaluate miRNA exosome production in mesangial, endothelial, and epithelial tubular renal cells. (B) After 18 h of stimulation, exosomes were obtained from studied renal cells to quantify their miRNA expression levels. Individual plots from n = 3 experiments. MiR-control is U6 small nuclear RNA. (C) Exosome internalization was measured by immunofluorescence in mesangial, endothelial, and epithelial tubular cells at different times. Blue color (DAPI) labels cell nuclei and red color labels exosomes. Each condition was assayed in triplicate. Error bars represent the means ± SEM from three independent experiments. Scale bar = 20 µm. p-Values were obtained using two-way ANOVA test. ** p < 0.005; *** p < 0.0005; **** p < 0.0001.

Figure 5

Urinary exosomal miRNAs specific for the LN responder group with their target mRNAs. (A) Venn diagram representing overlap of validated targets of miR-31-5p, miR-107, and miR-135b-5p. (B) Hypoxia-inducible factor-1 alpha (HIF1A) protein levels (green) in renal kidney biopsies from responder and non-responder patients. DAPI staining was used to label cell nuclei. Scale bar = 50 µm. * p < 0.05; ** p < 0.005. (C) Luciferase assay showed HIF1A as common target gene for miR-31, miR-107, and miR-135b-5p in mesangial or endothelial primary renal cells (five replicates per group). p-Values were obtained using two-way ANOVA test. ** p < 0.005.

Figure 6

Effect of exosomal miRNAs to mesangial and endothelial renal cells. (A) HIF1A expression levels were analyzed by quantitative real-time in overexpressed or inhibited primary mesangial and endothelial renal cells after interleukin-1 alpha (IL1α), interferon alpha (INFα) or vascular endothelial growth factor (VEGF) stimulation. Overexpression or inhibition was performed for the three studied miRNAs (miR-31-5p, miR-107, and miR-135b-5p). MRC = mesangial renal cells. ERC = endothelial renal cells. Values were normalized using GADPH and 18S. Data are expressed in Log10 scale. *** p < 0.0005, ** p < 0.005, and * p < 0.05. (B) Fold change of proliferation in mesangial renal cells with overexpression or inhibition of the miR-31, miR-107, and miR-135b-5p (over 3miRNAs or Inh 3miRNAS, respectively). Control conditions were obtained with mimic or anti miR-control. *** p < 0.0005. (C) Quantitative real-time RT-PCR analysis shows the relative mRNA levels of CXCL1, CCL2, CCL3, IL6, and IL8 in IL1α- and INFα-stimulated renal mesangial cells (MRCs). Values were normalized using GADPH and 18S. Fold change was calculated over the control condition (mimic miR-control or anti miR-control). *** p < 0.0005, ** p < 0.005 and * p < 0.05. (D) Expression levels of IL6 and VCAM-1 in endothelial renal cells with overexpression or inhibition of miR-31, miR-107, and miR-135b-5p after IL1α and VEGF stimulation. Fold change in expression level was calculated using the 2^−ΔΔCt method and over control condition (mimic miR-control or anti miR-control). p-Values were obtained using two-way ANOVA test. *** p < 0.0005, ** p < 0.005 and * p < 0.05.

Figure 7

Role of patient exosome in renal recovery. Exosomes from LN responder or non-responder patients were incubated with mesangial and endothelial renal cells. After that, gene expression analysis shows a significant downregulation of CXCL1, CCL2, CCL3, IL6, and VCAM-1. Values were normalized using glyceraldehyde 3-phosphate dehydrogenase (GADPH) and 18S ribosomal RNA (18S rRNA). Fold change was calculated over exosomes obtained from healthy donors (HD). p-Values were obtained using two-way ANOVA test. NS: not significant, *** p < 0.0005, ** p < 0.005 and * p < 0.05.

Figure 8

Proposed mechanism for urinary exosomal miRNAs in lupus nephritis renal recovery. Urinary exosomes from LN responder patients with high levels of miR-31-5p, miR-107, and miR-135b-5p are produced mainly in tubular renal cells to be engulfed by endothelial and mesangial renal cells. MiRNAs target HIF1A in endothelial renal cells, reducing the production of IL6 and VCAM-1. In mesangial renal cells, miRNAs target HIF1A, inhibiting mesangial renal cell proliferation and downregulating the production of CXCL1, CCL3, CCL2, and IL6.