Potential diagnostic role of urinary exosomal microRNAs in immunoglobulin A nephropathy: A case-control study

Table 1 Baseline characteristics of immunoglobulin A nephropathy cases and healthy controls

Characteristics	IgA nephropathy (n = 50)	Healthy controls (n = 50)
Mean age (in years)	36.32 ± 3.06	39.81 ± 4.13
Male:female	3:1	3:1
Mean serum creatinine (mg/dL)	2.26 ± 0.31	0.8 ± 0.24
Mean eGFR (mL/minute/1.73 m2)	45.46 ± 8.49	133.5 ± 53.5
Mean proteinuria (g/day)	2.69 ± 0.64
M1	98%
E1	46%
S1	26%
T1	38%
T2	12%

IgA: Immunoglobulin A; eGFR: Estimated glomerular filtration rate.

Table 2 Baseline characteristics of disease controls

Biopsy	Total number of cases	Mean age (in years)	Sex (male:female)	Mean serum creatinine (mg/dL)	eGFR (mL/minute/1.73 m2)	Proteinuria (grams/day)
Lupus nephritis	8	28.75 ± 5.1	1:3	1.34 ± 0.54	72.75 ± 12.23	1.67 ± 0.62
Diabetic nephropathy	9	47.2 ± 3.2	3:1	3.1 ± 0.43	18.66 ± 5.34	2.65 ± 1.21
Hypertensive nephropathy	9	52 ± 4.3	3:1	3.4 ± 1.1	16.8 ± 4.32	1.89 ± 87
Primary membranous nephropathy	8	39.5 ± 4.3	1:3	1.04 ± 0.56	83.5 ± 11.12	4.85 ± 2.32
Minimal change disease	8	10.25 ± 5.3	2:2	0.37 ± 0.48	165.25 ± 24.23	2.4 ± 1.22
Primary focal segmental glomerulosclerosis	8	18.75 ± 4.6	2:2	0.9 ± 0.63	126 ± 29.97	3.4 ± 1.12

eGFR: Estimated glomerular filtration rate.

Table 3 Least absolute shrinkage and selection operator regression model showing 15 important microRNAs with non-zero coefficient values

Healthy controls		IgAN		Disease controls
	Coefficient		Coefficient		Coefficient
Intercept	-2.277940703	Intercept	3.546568646	Intercept	-1.268627943
hsa-miR-409-5p	0.0763648402	hsa-miR-409-5p	-0.07957252572	hsa-miR-409-5p	0.003207685516
hsa-miR-4532	0.02020578069	hsa-miR-4532	-0.03750367821	hsa-miR-4532	0.01729789752
hsa-miR-320d	0.007678669715	hsa-miR-320d	-0.03569809053	hsa-miR-320d	0.02801942082
hsa-miR-3195	0.02502325322	hsa-miR-3195	-0.04476037204	hsa-miR-3195	0.01973711882
hsa-miR-664b-3p	0.005818441563	hsa-miR-664b-3p	-0.003261823244	hsa-miR-664b-3p	-0.002556618319
hsa-miR-151b	-0.008344659953	hsa-miR-151b	-0.02724506637	hsa-miR-151b	0.03558972632
hsa-miR-101-3p	0.0383626232	hsa-miR-101-3p	-0.008674373141	hsa-miR-101-3p	-0.02968825006
hsa-miR-492	-0.001133029441	hsa-miR-492	-0.0284297396	hsa-miR-492	0.02956276904
hsa-miR-511-5p	0.00200953104	hsa-miR-511-5p	1.32E-06	hsa-miR-511-5p	-0.002010848439
hsa-miR-769-5p	0.03050223486	hsa-miR-769-5p	0.01146667966	hsa-miR-769-5p	-0.04196891452
hsa-miR-499a-3p	0.02096262506	hsa-miR-499a-3p	0.01676192009	hsa-miR-499a-3p	-0.03772454516
hsa-miR-548n	-0.005970531682	hsa-miR-548n	0.01216365301	hsa-miR-548n	-0.006193121324
hsa-miR-654-3p	-0.05616509897	hsa-miR-654-3p	0.02554981477	hsa-miR-654-3p	0.0306152842
hsa-miR-29b-3p	-0.006355112674	hsa-miR-29b-3p	0.004412091123	hsa-miR-29b-3p	0.001943021551
hsa-miR-194-5p	-0.0000242413479	hsa-miR-194-5p	-1.72E-05	hsa-miR-194-5p	4.14E-05

IgAN: Immunoglobulin A nephropathy.

Table 4 MicroRNAs with area under the curve values > 0.75 with a high ability to differentiate immunoglobulin A nephropathy cases from healthy controls

miRNAs	hsa.miR.409.5p	hsa.miR.101.3p	hsa.miR.4532	hsa.miR.654.3p
Category	IgAN vs healthy
Accuracy (95%CI)	0.90 (0.58-0.99)	0.90 (0.58-0.99)	1 (0.71-1)	0.63 (0.30-0.89)
Sensitivity	0.50	0.50	1.00	1.00
Specificity	1.00	1.00	1.00	0.56
PPV	1.00	1.00	1.00	0.33
NPV	0.90	0.90	1.00	1.00
Prevalence	0.18	0.18	0.18	0.18
Detection rate	0.09	0.09	0.18	0.18
Detection prevalence	0.09	0.09	0.18	0.55
AUC (95%CI)	0.75 (0.26-1)	0.75 (0.26-1)	1 (1-1)	0.77 (0.60-0.94)
Kappa	0.62	0.62	1.00	0.31

miRNA: MicroRNA; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; AUC: Area under the receiver operating characteristic curve.

Table 5 microRNAs with area under the curve values > 0.75, with a high ability to differentiate immunoglobulin A nephropathy cases from disease controls

miRNAs	hsa.miR.769.5p	hsa.miR.3195	hsa.miR.4532	hsa.miR.151b
Category	IgAN vs Disease control
Accuracy (95%CI)	0.8 (0.44-0.97)	0.8 (0.44-0.97)	0.8 (0.44-0.97)	0.8 (0.44-0.97)
Sensitivity	0.5	0.5	0.5	0.5
Specificity	1	1	1	1
PPV	1	1	1	1
NPV	0.75	0.75	0.75	0.75
Prevalence	0.4	0.4	0.4	0.4
Detection rate	0.2	0.2	0.2	0.2
Detection prevalence	0.2	0.2	0.2	0.2
AUC (95%CI)	0.75 (0.46-1)	0.75 (0.46-1)	0.75 (0.46-1)	0.75 (0.46-1)
Kappa	0.55	0.55	0.55	0.55

miRNA: MicroRNA; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; AUC: Area under the receiver operating characteristic curve.

Table 6 Least absolute shrinkage and selection operator regression model selected 15 microRNAs that can differentiate immunoglobulin A nephropathy cases from healthy controls

miRNA	hsa.miR.769.3p	hsa.miR.671.3p	hsa.miR.641	hsa.miR.612	hsa.miR.548ar.5p	hsa.miR.548a5p	hsa.miR.513c.3p	hsa.miR.4488	hsa.miR.3158.3p	hsa.miR.30e5p	hsa.miR.297	hsa.miR.222.3p	hsa.miR.219b3p	hsa.miR.1269b	hsa.miR.1268a
Accuracy (95%CI)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.727272727272727 (0.390257440427579-0.939782265827093)	0.818181818181818 (0.482244147639827-0.97716880170004)	0.909090909090909 (0.587220083011617-0.997701027786186)	0.818181818181818 (0.482244147639827-0.97716880170004)
Sensitivity	0	0.5	0	0	0	0	0	0.5	0.5	0	0.5	0	0	0.5	0
Specificity	1	0.7777777778	1	1	0.8888888889	0.8888888889	0.8888888889	0.8888888889	0.8888888889	0.8888888889	0.7777777778	0.8888888889	1	1	1
PPV	NA	0.3333333333	NA	NA	0	0	0	0.5	0.5	0	0.3333333333	0	NA	1	NA
NPV	0.8181818182	0.875	0.8181818182	0.8181818182	0.8	0.8	0.8	0.8888888889	0.8888888889	0.8	0.875	0.8	0.8181818182	0.9	0.8181818182
Prevalence	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818	0.1818181818
Detection rate	0	0.09090909091	0	0	0	0	0	0.09090909091	0.09090909091	0	0.09090909091	0	0	0.09090909091	0
Detection prevalence	0	0.2727272727	0	0	0.09090909091	0.09090909091	0.09090909091	0.1818181818	0.1818181818	0.09090909091	0.2727272727	0.09090909091	0	0.09090909091	0
AUC (95%CI)	0.5 (0.5-0.5)	0.638888888888889 (0.128164113234834-1)	0.5 (0.5-0.5)	0.5 (0.5-0.5)	0.444444444444444 (0.335557556414441-0.553331332474447)	0.444444444444444 (0.335557556414441-0.553331332474447)	0.444444444444444 (0.335557556414441-0.553331332474447)	0.694444444444444 (0.192500691940002-1)	0.694444444444444 (0.192500691940002-1)	0.444444444444444 (0.335557556414441-0.553331332474447)	0.638888888888889 (0.128164113234834-1)	0.444444444444444 (0.335557556414441-0.553331332474447)	0.5 (0.5-0.5)	0.75 (0.260009003864987-1)	0.5 (0.5-0.5)
Kappa	0	0.2325581395	0	0	-0.1379310345	-0.1379310345	-0.1379310345	0.3888888889	0.3888888889	-0.1379310345	0.2325581395	-0.1379310345	0	0.6206896552	0

miRNA: MicroRNA; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; AUC: Area under the receiver operating characteristic curve.

Table 7 Logistic regression statistics showing area under the curve of 11 important least absolute shrinkage and selection operator-selected microRNAs that can differentiate between immunoglobulin A nephropathy cases and disease controls

miRNA	hsa.miR.1224.5p	hsa.miR.1279	hsa.miR.1289	hsa.miR.20a5p.hsa.miR.20b5p	hsa.miR.32.5p	hsa.miR.365b5p	hsa.miR.4787.3p	hsa.miR.520c.3p	hsa.miR.525.3p	hsa.miR.548a3p	hsa.miR.593.3p
Accuracy (95%CI)	0.8 (0.443904537692359-0.974789273673167)	0.8 (0.443904537692359-0.974789273673167)	0.9 (0.554983882971805-0.997471421455538)	0.9 (0.554983882971805-0.997471421455538)	0.9 (0.554983882971805-0.997471421455538)	0.8 (0.443904537692359-0.974789273673167)	0.8 (0.443904537692359-0.974789273673167)	0.8 (0.443904537692359-0.974789273673167)	0.9 (0.554983882971805-0.997471421455538)	1 (0.691502892181239-1)	0.8 (0.443904537692359-0.974789273673167)
Sensitivity	0.75	0.5	0.75	0.75	0.75	0.5	0.5	0.5	0.75	1	0.5
Specificity	0.833333333	1	1	1	1	1	1	1	1	1	1
PPV	0.75	1	1	1	1	1	1	1	1	1	1
NPV	0.833333333	0.75	0.857142857	0.857142857	0.857142857	0.75	0.75	0.75	0.857142857	1	0.75
Prevalence	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4	0.4
Detection rate	0.3	0.2	0.3	0.3	0.3	0.2	0.2	0.2	0.3	0.4	0.2
Detection prevalence	0.4	0.2	0.3	0.3	0.3	0.2	0.2	0.2	0.3	0.4	0.2
AUC (95%CI)	0.791666666666667 (0.497218723153255-1)	0.75 (0.467103566480957-1)	0.875 (0.630004501932493-1)	0.875 (0.630004501932493-1)	0.875 (0.630004501932493-1)	0.75 (0.467103566480957-1)	0.75 (0.467103566480957-1)	0.75 (0.467103566480957-1)	0.875 (0.630004501932493-1)	1 (1-1)	0.75 (0.467103566480957-1)
Kappa	0.583333333	0.545454546	0.782608696	0.782608696	0.782608696	0.545454546	0.545454546	0.545454546	0.782608696	1	0.545454546

miRNA: MicroRNA; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; AUC: Area under the receiver operating characteristic curve.

Table 8 Regulation status of 9 microRNAs within multiple comparisons performed using the nCounter microRNA expression assay

Type of differentiation	miRNAs	IgAN vs HC foldchange	IgAN vs HC P value	DC vs HC foldchange	DC vs HC P value	IgAN vs DC foldchange	IgAN vs DC P value
Differentiate IgAN and disease control	hsa-miR-1289	-1.19	0.05164487	-2.48	0.00001151	2.08	0.00001932
	hsa-miR-20a-5p+hsa-miR-20b-5p	-1.14	0.32179308	2.16	0.00004533	-2.46	0.00000026
	hsa-miR-32-5p	1.11	0.58471572	2.74	0.0002994	-2.48	0.0002138
	hsa-miR-525-3p	-1.13	0.46521613	2.05	0.00166705	-2.31	0.00012489
	hsa-miR-151b	-1.5	0.01793812	1.65	0.00405257	-2.48	0.00000064
	hsa-miR-4532	-2.06	0.00113097	1.17	0.68054599	-2.4	0.04225222
	hsa-miR-3195	-1.78	0.00082084	1.21	0.2488215	-2.14	0.00013383
Differentiate IgAN and healthy control	hsa-miR-4488	-2.65	0.02099009	-1.33	0.52975196	-2	0.05507494
	hsa-miR-3158-3p	-1.79	0.00653545	1.02	0.95068687	-1.82	0.07137911

miRNA: MicroRNA; IgAN: Immunoglobulin A nephropathy; HC: Healthy controls; DC: Disease controls.

Table 9 The combination of these 9 microRNAs has the potential to diagnose immunoglobulin A nephropathy cases with an area under the receiver operating characteristic curve > 0.8¹

Discovery cohort with 9 miRNAs	hsa-miR-4532, hsa-miR-4488, hsa-miR-3158-3p, hsa-miR-151b, hsa-miR-3195, hsa-miR-1289, hsa-miR-20a-5p+hsa-miR-20b-5p, hsa-miR-32-5p, hsa-miR-525-3p
0 = Control, 1 = Disease, 2 = IgAN
Class	Control (0)	Disease (1)	IgAN (2)
Accuracy (95%CI)	0.7692 (0.4619-0.9496)	0.8462 (0.5455-0.9808)	0.7692 (0.4619-0.9496)
Sensitivity	0.33333	0.33333	0.8571
Specificity	0.9	1	0.6667
Positive predictive value	0.5	1	0.75
Negative predictive value	0.81818	0.83333	0.8
Prevalence	0.23077	0.2308	0.5385
Detection rate	0.07692	0.07692	0.4615
Detection prevalence	0.15385	0.07692	0.6154
AUC (95%CI)	0.733 (0.5-0.93)	1 (1-1)	0.88 (0.57-1)
Kappa	0.2642	0.4348	0.5301

¹Validation is presented in Supplementary Table 2, within the external cohort GSE64306.

miRNA: MicroRNA; IgAN: Immunoglobulin A nephropathy; CI: Confidence interval; AUC: Area under the receiver operating characteristic curve.

Table 10 A combination of 2 microRNAs could diagnose immunoglobulin A nephropathy cases with an area under the receiver operating characteristic curve > 0.8

Discovery cohort with 2 microRNAs	hsa-miR-4532, hsa-miR-548a-3p
0 = Healthy controls, 1 = Disease controls, 2 = IgA nephropathy cases
Class	Control (0)	Disease (1)	IgAN (2)
Accuracy (95%CI)	0.7692 (0.4619-0.9496)	0.8462 (0.5455-0.980)	0.8462 (0.5455-0.9808)
Sensitivity	0.6667	1	1
Specificity	1	1	0.8333
PPV	1	1	0.875
NPV	0.9091	1	1
Prevalence	0.2308	0.2308	0.5385
Detection rate	0.1538	0.2308	0.5385
Detection prevalence	0.1538	0.2308	0.6154
AUC (95%CI)	0.8333333 (0.5-1)	1 (1-1)	0.9047619 (0.741-1)
Kappa	0	0.4348	0.6905

miRNA: MicroRNA; IgAN: Immunoglobulin A nephropathy; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; AUC: Area under the receiver operating characteristic curve.

Table 11 Status of the 5 published microRNA markers as observed from our previous study, within 3 comparisons (immunoglobulin A nephropathy vs healthy control, disease control vs healthy control, and immunoglobulin A nephropathy vs disease control)

IgA nephropathy vs healthy control
Probe name	Nephropathy	Control	Ig A nephropathy vs healthy control	P-value of: Nephropathy vs control
hsa-miR-146b-3p	4.07	13.3	-3.27	0.00047341
hsa-miR-221-5p	8.72	14.66	-1.68	0.00512936
hsa-miR-4532	7.01	12.3	-1.76	0.00655229
hsa-miR-599	7.26	16.63	-2.29	0.00649493
hsa-miR-664b-5p	3.85	9.47	-2.46	0.00083272
IgA nephropathy vs disease control
Probe name	Nephropathy	Disease control	IgA nephropathy vs disease control	P-value of: Nephropathy vs disease control
hsa-miR-146b-3p	6.17	8.4	-1.36	0.26787862
hsa-miR-221-5p	11.35	11.84	-1.04	0.73601639
hsa-miR-4532*	8.79	21.09	-2.4	0.04225222
hsa-miR-599	8.66	10.01	-1.16	0.5314607
hsa-miR-664b-5p	5.45	5.32	1.02	0.87722093
Disease control vs healthy control
Probe name	Disease control	Control	Disease control vs healthy control	P-value of: Disease control vs healthy control
hsa-miR-146b-3p	8.4	15.41	-1.83	0.04495602
hsa-miR-221-5p	11.84	18.22	-1.54	0.00787352
hsa-miR-4532	21.09	18.09	1.17	0.68054599
hsa-miR-599	10.01	18.5	-1.85	0.03426054
hsa-miR-664b-5p	5.32	9.86	-1.85	0.0062102

IgA: Immunoglobulin A.

Table 12 Biological relevance and disease associations of the 9 candidate microRNAs¹

miRNA	Ref.	Sample/context reported	Biological relevance/disease association
hsa-miR-4532	Seo et al[33], 2023; Kim et al[34], 2019; Liu et al[35], 2022	Urinary exosomes in diabetic kidney disease; urinary exosomes in kidney transplant recipients	Reported as a urinary exosomal biomarker in diabetic kidney disease, with significantly lower expression in biopsy-proven DKD compared with healthy controls, suggesting an association with chronic kidney injury; also part of a three-miRNA urinary exosomal signature (miR-21-5p, miR-31-5p, miR-4532) that discriminates acute rejection from stable graft function in kidney transplant recipients, indicating relevance to immune-mediated allograft injury; exosomal miR-4532 has been shown to promote endothelial cell injury via SP1 and NF-κB p65 activation, linking it to inflammatory and endothelial pathways relevant to glomerular and vascular damage
hsa-miR-4488	Zhong et al[36], 2021	Plasma-derived exosomes in dermatomyositis-associated interstitial lung disease	Identified among differentially expressed exosomal miRNAs in patients with dermatomyositis-associated interstitial lung disease, implicating roles in systemic autoimmunity and chronic inflammation; although kidney-specific data are not yet available, involvement in autoimmune and inflammatory settings supports potential relevance to immune-mediated glomerular injury
hsa-miR-3158-3p	Gupta et al[37], 2021; Gupta et al[38], 2026	Plasma in cerebral malaria; functional in immune-signaling assays	Elevated plasma miR-3158-3p levels correlate with MRI brain injury and poor outcome in cerebral malaria, a condition characterized by endothelial dysfunction and intense immune activation; functional studies show that miR-3158-3p overexpression downregulates NF-κB expression and modulates immune-related pathways, indicating a role in cytokine and innate immune signaling; no kidney-specific data are currently available, but the link to NF-κB and systemic inflammation is mechanistically compatible with IgAN-related immune activation
hsa-miR-151b	Király et al[39], 2024	Renal cell carcinoma tissue (hsa-miR-15b-5p and other family members studied)	Members of the miR-15b family (e.g., hsa-miR-15b-5p) are significantly downregulated in renal cell carcinoma compared with adjacent normal kidney and correlate inversely with tumor grade, suggesting roles in vascular/angiogenic and extracellular matrix pathways in kidney tissue; however, no kidney- or IgAN-specific functional data are currently available for hsa-miR-151b itself; no kidney/IgAN-specific functional data to date; identified here as a novel, hypothesis-generating diagnostic candidate
hsa-miR-3195	Zhou et al[40], 2021	Serum in Kallmann syndrome	Reported to target an aberrant PROK2 transcript and to modulate PROK2 function in vitro, with low serum expression associated with altered neuroendocrine phenotypes in Kallmann syndrome; while not kidney-specific, PROK2-related signaling can influence vascular and endocrine axes that may intersect with systemic immune and hemodynamic regulation; no kidney/IgAN-specific functional data to date; included as a novel biomarker candidate requiring further mechanistic work
hsa-miR-1289	Srivastava et al[41], 2023	In peripheral blood from patients diagnosed with COVID-19 infection	Cytosolic sulphonation of small molecules; amplification of signal from kinetochores; mitotic spindle checkpoint; no kidney/IgAN-specific functional data to date; identified here as a novel, hypothesis-generating diagnostic candidate
hsa-miR-20a-5p/hsa-miR-20b-5p	Donderski et al[42], 2022	Various kidney and CKD cohorts (miR-20a-5p family often profibrotic/profibrogenic in CKD panels)	miR-20 family members are frequently included among profibrogenic miRNA panels evaluated in CKD, where deregulation correlates with eGFR decline and proteinuria, and they are predicted to target components of TGF-β, cell-cycle, and apoptosis pathways associated with renal fibrosis; although direct IgAN-specific data are limited, their involvement in fibrotic and inflammatory signaling provides a plausible link to chronic glomerular/tubulointerstitial injury
hsa-miR-32-5p	Donderski et al[42], 2022	CKD-related miRNA panels (non-IgAN)	Included in several profiling studies as part of deregulated miRNA sets in chronic kidney disease and proteinuric states, with predicted targets in apoptosis, cell proliferation, and inflammatory pathways; no direct IgAN-specific functional studies, but pathway predictions intersect with NF-κB and TGF-β signaling implicated in glomerulosclerosis and tubulointerstitial fibrosis
hsa-miR-525-3p	Marques et al[43], 2011	In the peripheral circulation of hypertensive kidney disease	RHOF GTPase cycle; signal transduction; RHO GTPase cycle

¹Functional annotations are based on published literature and predicted or experimentally validated targets, and are intended as preliminary, hypothesis-generating information rather than definitive mechanistic assignments for immunoglobulin A nephropathy.

miRNA: MicroRNA; NF-κB: Nuclear factor kappa-light-chain-enhancer of activated B cells; DKD: Diabetic kidney disease; IgAN: Immunoglobulin A nephropathy; PROK2: Prokineticin 2; COVID-19: Coronavirus disease 2019; TGF-β: Transforming growth factor-β; CKD: Chronic kidney disease; RHOF GTPase: Ras homolog family member F.