DOI: 10.1111/eci.

12585

ORIGINAL ARTICLE

Renal microRNA- and RNA-profiles in progressive

chronic kidney disease
Michael Rudnicki*, Paul Perco†, Barbara D0 haene‡, Johannes Leierer*, Andreas Heinzel†, Irmgard Mu€ hlberger†,
* *,§ ¶ * ‡
Ninella Schweibert , Judith Sunzenauer , Heinz Regele , Andreas Kronbichler , Pieter Mestdagh ,
Jo Vandesompele‡, Bernd Mayer† and Gert Mayer*
*
Department of Internal Medicine IV – Nephrology and Hypertension, Medical University Innsbruck, Innsbruck, Austria,
†
Emergentec Biodevelopment GmbH, Vienna, Austria, ‡Biogazelle, Zwijnaarde, Belgium, §Department of Nephrology, KH
Elisabethinen, Linz, Austria, ¶Institute of Pathology, Medical University Vienna, Vienna, Austria

ABSTRACT
Background MicroRNAs (miRNAs) contribute to chronic kidney disease (CKD) progression via regulating
mRNAs involved in renal homeostasis. However, their association with clinical outcome remains poorly
understood.
Materials and methods We performed miRNA and mRNA expression profiling on renal biopsy sections by
qPCR (miRNA) and microarrays (mRNA) in a discovery (n = 43) and in a validation (n = 29) cohort. miRNAs
differentiating stable and progressive cases were inversely correlated with putative target mRNAs, which were
further characterized by pathway analysis using KEGG pathways.
Results miR-30d, miR-140-3p, miR-532-3p, miR-194, miR-190, miR-204 and miR-206 were downregulated in
progressive cases. These seven miRNAs correlated with upregulated 29 target mRNAs involved in inflammatory
response, cell–cell interaction, apoptosis and intra-cellular signalling. In particular, miR-206 and miR-532-3p were
associated with distinct biological processes via the expression of their target mRNAs: Reduced expression of
miR-206 in progressive disease correlated with the upregulation of target mRNAs participating in inflammatory
pathways (CCL19, CXCL1, IFNAR2, NCK2, PTK2B, PTPRC, RASGRP1 and TNFRSF25). Progressive cases also
showed a lower expression of miR-532-3p and an increased expression of target transcripts involved in apo-
ptosis pathways (MAP3K14, TNFRSF10B/TRAIL-R2, TRADD and TRAF2). In the validation cohort, we confirmed
the decreased expression of miR-206 and miR-532-3p, and the inverse correlation of these miRNAs with the
expression of nine of the 12 target genes. The levels of the identified miRNAs and the target mRNAs correlated
with clinical parameters and histological damage indices.
Conclusions These results suggest the involvement of specific miRNAs and mRNAs in biological pathways
associated with the progression of CKD.
Keywords Biomarker, chronic kidney disease, microarray, miRNA, systems biology, transcriptomics.
Eur J Clin Invest 2016; 46 (3): 213–226

Consequently, transcriptomics studies profiling mRNA

Introduction
Histological hallmarks of progressive chronic kidney disease
(CKD) are an increased degree of tubulointerstitial atrophy
and fibrosis. These pathological changes are preceded and
promoted by events such as infiltration of inflammatory cells,
fibroblast activation and proliferation, excessive production
and deposition of extracellular matrix components, and rar-
efaction of peritubular capillaries [1,2]. At the molecular
level, these processes are triggered by and lead to de novo
intra-renal expression of transcripts representing damaging as
well as protective/regenerative biological pathways [3].
expression have substantially contributed to our under-
standing of renal disease, even though it has been realized
that transcriptional regulation is more complex [4]. Micro-
RNAs (miRNAs) are endogenous, small (20–22 nucleotides)
noncoding RNAs that modulate the abundance of protein-
coding gene transcripts at the post-transcriptional level. Their
activity impacts fundamental cellular processes such as
development, differentiation, apoptosis, ageing and metabo-
lism, and their differential expression has been associated
with the pathophysiology of many diseases [5] including
renal disorders [6–8].

European Journal of Clinical Investigation Vol 46 213

 M. RUDNICKI ET AL.
Although a potential role of several miRNAs has been
described in pathophysiology of human CKD [9–11], data are
limited and sometimes contradictory [12,13]. To our knowl-
edge, no results have been published on large-scale miRNA
and mRNA profiling on the same renal tissue samples. Only
limited information exists on the association of miRNAs and
the respective target mRNAs with histological features and
well-defined clinical parameters in a longitudinal fashion in
human subjects with CKD. The aim of this study was to iden-
tify differentially expressed renal miRNA-profiles contrasting a
stable vs. a progressive course of disease. Using biopsies from
patients with different proteinuric renal diseases allowed us to
identify miRNA-profiles that are associated with the progres-
sion of CKD rather than with a certain diagnosis. After corre-
lation of these miRNAs with the expression of potential target
mRNAs from the same tissue sample and an extensive analysis
of links at the level of molecular pathways, the expression
results were confirmed in an independent validation cohort of
kidney biopsies. Finally, a subset of validated miRNAs and
their target mRNAs were correlated with both structural
alterations in the biopsies and clinical characteristics.
Materials and methods
Kidney biopsies and RNA isolation
For the ‘discovery cohort’, we selected biopsies from 43 patients
with various glomerular diseases. To validate the results from
the discovery cohort, we collected an independent cohort of
biopsy samples from 29 patients (‘validation cohort’). Patients
reaching end-stage renal disease or doubling of serum crea-
tinine during follow-up were defined as ‘progressive’ (n = 20),
all others as ‘stable’ (n = 52). In stable subjects, clinical data
from the last available follow-up visit were recorded. In pro-
gressive subjects who started dialysis during follow-up, the last
serum creatinine and urine protein–creatinine ratio (UPCR)
before initiation of dialysis were recorded. The degree of his-
tological damage was recorded by a pathologist (HR) and
recorded as follows: glomerular sclerosis (absent, present),
tubular atrophy (none, mild, moderate, severe), interstitial
fibrosis (none, mild, moderate, severe), interstitial inflammation
(absent, present), arteriosclerosis (intima fibrosis; none, mild,
moderate, severe) and arteriolosclerosis (arteriolar hyalinosis;
none, mild, moderate, severe) respectively.
Total RNA, including miRNA, was isolated from cryocut
tissue sections using the miRNeasy
Mini Kit (Qiagen, Hilden,
Germany). Quality and concentration of the isolated RNA were
evaluated using a NanoDrop
1000 spectrophotometer
(Thermo Fisher Scientific, Wilmington, DE, USA); all samples
showed an A260/280 ratio of > 160.
The Institutional Review Board of the Medical University of
Innsbruck has accredited the use of surplus material from
214 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
www.ejci-online.com
routine kidney biopsies and anonymized patient data for
research purposes.
miRNA expression profiling
Total RNA was reverse transcribed using the Megaplex RT
stem-loop primer pool (Life Technologies, Carlsbad, CA, USA),
enabling miRNA-specific cDNA synthesis of 755 different hu-
man miRNAs and small RNA controls [14]. In brief, total RNA
was reverse transcribed using the Megaplex RT stem-loop pri-
mer pool A and B (Applied Biosystems, Life Technologies
Corporation, Carlsbad, CA, USA). The cDNA was then
preamplified by means of a 12-cycle PCR reaction with a
miRNA-specific forward primer and universal reverse primer.
Finally, a dilution of pre-amplified miRNA cDNA was used as
input for a 40-cycle qPCR reaction with miRNA-specific
hydrolysis probes and primers (Life Technologies). All reac-
tions were performed on Applied Biosystems 7900 HT using
the gene maximization strategy. Three endogenous small RNA
control targets (RNU44, RNU48 and U6) were included to
check for intra-run and inter-run variation. The Cq values were
subsequently analysed in qbase+ (Biogazelle, Zwijnaarde, Bel-
gium). Cq values above 32 were considered as noise and
excluded [15].
Identification of differentially expressed miRNAs
miRNA expression data were normalized using the modified
global mean method in qbase+ [16]. Only miRNAs expressed in
at least 80% of the samples within each group were included in
the analysis. In the discovery cohort, differentially expressed
miRNAs were identified using two nonparametric tests (Mann–
Whitney test and Rank Products algorithm). Mann–Whitney
P-values were corrected for multiple testing using the Ben-
jamini–Hochberg procedure. The Rank Products algorithm
corrects for multiple testing using permutations (n = 100) and
reports the percentage of false positives (pfp < 005). Differen-
tially expressed miRNAs from both analysis methods were
then combined for the miRNA–mRNA correlation and target
prediction analysis. In the validation cohort, a Mann–Whitney
test was performed, and results with a two-sided P-value < 005
(confidence interval 95%) were defined as being significant.
Transcriptomics
From each biopsy, 400 ng of total RNA was used to generate
Cyanine 3 (Cy3)-labelled cRNA for hybridization to Agilent
oligonucleotide microarrays according to the Agilent One-
Color Microarray protocol (Agilent Technologies, Santa Clara,
CA, USA). Arrays were washed and scanned using an Axon
Gene Pix 4000B scanner (Molecular Devices, Sunnyvale, CA,
USA), and signal intensity data were extracted using Agilent
Feature Extraction software (v10.5.1). The microarray data have
been deposited in NCBI’s Gene Expression Omnibus [17] and

are accessible through GEO Series accession number GSE60861
(http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?
acc=GSE60861). Data preprocessing consisted of (i) excluding
low-abundance features (intensity values < 15 above back-
ground signals), (ii) quantile–quantile data normalization and
(iii) summarization of redundant sequence spots. This step
resulted in a datafile holding 9665 features, which was used for
miRNA–mRNA correlation analysis in the discovery cohort.
Real-time PCR of mRNAs
The validity of the microarray results was tested by reverse
transcription quantitative PCR (RT-qPCR) analysis for four
genes (CCL19, CCL21, LSP1, TIMP3) in six randomly selected
biopsy samples from the discovery cohort. Samples were
reverse transcribed and preamplified using TaqMan
Gene
Expression Assays and the TaqMan
PreAmp Master Mix Kit
(Life Technologies). The preamplified cDNA was analysed in
duplicates on the 7500 Fast Real-Time PCR System (Applied
Biosystems) using the following inventoried TaqMan
Gene
Expression Assays: Hs00171149_m1 (CCL19), Hs00171076_m1
(CCL21), Hs00158886_m1 (LSP1), Hs00165949_m1 (TIMP3).
Cyclophilin A (PPIA, Hs99999904_m1) was used as an
endogenous control, and the expression values of the respective
genes were compared using the delta-ct method. All four genes
showed a significant and positive correlation between
microarray and real-time PCR experiments, thus indicating
validity of the microarray results (R2 from 077 to 094, P from
< 00001–00019, data not shown).
Correlation of miRNAs and mRNAs, and
identification of putative miRNA targets
In the discovery cohort, miRNA/mRNA correlations were
calculated using Spearman0 s rank statistics. Corresponding
P-values were corrected for multiple testing using the Bonfer-
roni procedure. mRNAs that were significantly negatively
correlated with the respective miRNAs were selected and fur-
ther examined using the following miRNA target prediction
algorithms: MIRDB (v. 3.0) [18, 19], TARGETSCAN (v. 5.1) [20],
MICROCOSM (v. 5) [21], DIANA (v. 3.0) [22], RNA22 (v.
August 2007) [23], PITA (v. 6) [24]. An mRNA was defined as a
potential target of the respective miRNA when at least one
database predicted this association. In the validation cohort, we
calculated miRNA/mRNA correlations of selected miRNAs
and their targets using Spearman0 s rank statistics, and a P-value
of < 005 was defined as significant.
Functional analysis
Significantly enriched pathways were identified for the sets of
corresponding mRNAs for each of the significantly regulated
miRNAs using DAVID version 6.7 [Database for Annotation,
Visualization and Integrated Discovery (DAVID)] [25].
MIRNA-RNA PROFILING IN CKD
Pathway images were generated using KEGG database and
KEGG mapping color tool (http://www.genome.jp/kegg/-
tool/map_pathway2.html) [26,27].
Correlation with clinical and histological parameters
The filtered, normalized and log-transformed expression data
of miRNAs and mRNAs were correlated with creatinine and
UPCR at the time of biopsy and at the time of follow-up using
Spearman rank correlation. A P-value of < 005 was defined as
significant. Differences of miRNA and mRNA expression in
different degrees of histological injury were analysed by cal-
culating medians and Mann–Whitney test (glomerulosclerosis
absent/present, interstitial inflammation absent/present and
arteriolosclerosis/hyalinosis absent/present). In the validation
cohort also arteriosclerosis/intima fibrosis was defined as
absent/present due to very low numbers of advanced histo-
logical damage. In all other cases (tubular atrophy, interstitial
fibrosis, arteriosclerosis/intima fibrosis), one-way ANOVA was
calculated and a P-value < 005 was defined as limit of
significance.
Results
The data analysis flow chart is depicted in Fig. 1; a summary of
clinical characteristics of the patients from the discovery and
validation cohorts is presented in Tables 1–3; detailed infor-
mation is shown in supplemental Table 1. In the discovery
cohort, serum creatinine was significantly higher in progressive
than in stable patients at the time of biopsy (27 vs. 16 mg/dL,
P = 0034). Renal function further deteriorated in progressive
subjects during follow-up (creatinine 71 vs. 15 mg/dL,
P < 0001). Proteinuria at last follow-up was also significantly
higher in progressive subjects (UPCR 46 vs. 16 g/g,
P = 0009). The validation cohort was similar to the discovery
cohort, except for significantly lower creatinine values at the
time of biopsy (19 vs. 14 mg/dL, P = 0047). Also proteinuria
at the time of follow-up was significantly lower in the valida-
tion cohort (UPCR 25 vs. 07 g/g, P = 0009). In the validation
cohort, no differences between the progressive and stable
cohorts were present at the time of biopsy regarding clinical
and laboratory parameters, but at the time of last follow-up,
creatinine and proteinuria varied significantly (creatinine 60
vs. 11 mg/dL, P < 0001; UPCR 16 vs. 05 g/g, P = 0039).
There were no differences between both progressive cohorts
regarding gender, age, follow-up time, creatinine and protein-
uria at the time of biopsy and at the time of follow-up. In the
stable cohorts, the gender distribution was significantly differ-
ent, with 58% males in the stable discovery and 41% males in
the stable validation cohort.
Regarding histological damage indices, biopsies from
patients from the discovery cohort with progressive renal
European Journal of Clinical Investigation Vol 46 215
 M. RUDNICKI ET AL.
Figure 1 Analysis flow chart. miRNA and mRNA profiling was
first performed in the discovery cohort. Forty-seven of 55
differentially regulated miRNAs correlated inversely with a total
of 742 mRNAs. Using miRNA target prediction algorithms, we
identified 29 differentially expressed miRNAs that correlated
with 323 mRNA targets. Each group of mRNAs that correlated
with a single miRNA was further analysed using pathway
analysis tools. This step resulted in the correlation of seven
miRNAs, 29 mRNA targets representing 14 biological
pathways. The experiments were repeated for a subset of
miRNAs and target mRNAs in an independent validation cohort
of kidney biopsies. Finally, correlation of miRNA and mRNA
expression levels with laboratory parameters and histology
was performed.
disease showed higher proportions of glomerulosclerosis,
tubular atrophy, interstitial fibrosis, interstitial inflammation,
arteriolosclerosis and arteriosclerosis, when compared with
stable patients of the same cohort. This was also the case in the
validation cohort, except that there were no differences
216 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
www.ejci-online.com
regarding glomerulosclerosis and interstitial inflammation.
When stable cases and progressive cases were compared
between both cohorts, we detected a significantly higher degree
of histological damage in the discovery cohort (tubular atrophy,
interstitial fibrosis, arteriolosclerosis and arteriosclerosis), but
there were no differences in the degree of glomerulosclerosis
and interstitial inflammation.
Differential expression of miRNAs and correlation
with target mRNAs
In a first step, we identified differentially expressed miRNAs
comparing stable and progressive cases from the discovery
cohort. To determine a comprehensive set of differentially
regulated miRNAs between both groups, we accepted miRNAs
that were identified to be differentially regulated by either
Mann–Whitney test or Rank Product Algorithm. In total, 55
miRNAs were identified in the discovery cohort (24 being up-
and 31 being downregulated in the progressive group)
(Table 4). As each miRNA regulates the expression of numer-
ous target mRNAs mostly by degrading mRNA levels, [28] we
focused on those miRNAs that were inversely correlated with
at least one mRNA. Of the 55 differentially regulated miRNAs,
47 significantly and negatively correlated with at least one and
a total of 742 unique mRNA transcripts; miR-33a*, let-7i*, miR-
551b, miR-302a, miR-516a-3b, miR-190b, let-7e* and miR-22*
were not significantly negatively correlated with any mRNA
transcript. We further focused on those mRNAs that were
predicted targets of one of the 47 miRNAs according to at least
one of the miRNA target prediction algorithms (i.e. MIRDB,
TARGETSCAN, MICROCOSM, DIANA, RNA22, PITA)
leaving 29 miRNAs and 323 mRNA targets. For 12 miRNAs,
no information on mRNA targets was available and six
miRNAs did not correlate with any predicted target (Table 4,
Column C).
Pathway analysis
We next hypothesized that pathogenically relevant miRNAs
may regulate entire biological pathways by targeting multiple
pathway-specific mRNAs. Using the DAVID, we performed a
KEGG pathway enrichment analysis separately for target
mRNAs correlating with each specific miRNA starting from the
323 predicted target mRNAs mentioned above. We identified
seven miRNAs that were downregulated in progressive sub-
jects: miR-30d, miR-140-3p, miR-532-3p, miR-194, miR-190,
miR-204 and miR-206. These miRNAs correlated with 29
upregulated target mRNAs, which were associated with 14
biological pathways (Fig. 2). The seven miRNAs and their
corresponding 29 target mRNAs were mainly associated with
pathways involved in inflammation, cell adhesion, apoptosis
and intra-cellular signalling. Interestingly, two of the seven
miRNAs correlated with target mRNAs that were involved in
 MIRNA-RNA PROFILING IN CKD

Table 1 Clinical characteristics of the discovery and the validation cohorts

Discovery cohort Validation cohort
All Stable Progressive All Stable Progressive

n 43 31 12 29 21 8
e e
Gender, n male/female 28/15 18/13 10/2 15/14 8/13 7/1
Age (years) 51  18 49  19 56  11 44  17 41  17 50  16
Follow-up time (months) 43  35 42  33 46  35 50  35 54  37 39  25
Creatinine at BX (mg/dL) 19  14 c
16  13 a
27  13 a
14  06 c
12  05 17  07
Creatinine at FU (mg/dL) 29  31 15  11 b
71  31 b
23  28 11  03 b
60  36b
UPCR at BX (g/g) 44  36 45  37 39  32 37  29 31  26 53  31
UPCR at FU (g/g) 25  31 d
16  24 a
46  37 a
07  11 d
05  10 a
16  10a

BX, biopsy; FU, follow-up; UPCR, urine protein–creatinine ratio.

Stable vs. progressive in the respective cohort: aP < 005, bP < 0001. Discovery vs. validation cohort (all samples): cP < 005, dP < 001. Comparison of stable
cohorts: eP < 005 (chi-square test). There were no significant differences between the two progressive cohorts.

Table 2 Histological diagnosis of subjects included in this study

Discovery cohort Validation cohort
All Stable Progressive All Stable Progressive

Diabetic nephropathy, n 5 1 4 1 1 0
Hypertensive nephrosclerosis, n 9 4 5 0 0 0
Focal-segmental glomerulosclerosis, n 5 3 2 2 1 1
Minimal change disease, n 9 9 0 2 2 0
IgA-nephropathy, n 2 1 1 2 1 1
Membranous nephropathy, n 9 9 0 3 1 2
Lupus nephritis, n 0 0 0 11 10 1
Membranoproliferative GN, n 0 0 0 2 0 2
ANCA vasculitis, n 0 0 0 5 4 1
Other, n 4 4 0 1 1 0

single specific biological processes: reduced expression of miR-
206 significantly correlated with the upregulation of its pre-
dicted target mRNAs CCL19, CXCL1, IFNAR2, NCK2, PTK2B,
PTPRC, RASGRP1 and TNFRSF25 (also known as DR3). These
transcripts all participate in inflammatory pathways such as
natural killer cell-mediated cytotoxicity, chemokine signalling,
leucocyte transendothelial migration, T-cell receptor signalling and
cytokine–cytokine receptor interaction. Lower expression of miR-
532-3p significantly correlated with increased expression of
target mRNAs MAP3K14, TNFRSF10B (also known as TRAIL-
R2), TRADD and TRAF2, that are all involved in apoptosis
pathways. The other miRNAs were associated with more than
one biological pathway entity, but the predominance of
inflammatory pathways was striking (Fig. 2).
Regarding miRNAs that were upregulated in progressive
subjects, we did not identify any significant biological pathway
being enriched with inversely correlated target mRNAs.
Therefore, we focused on the ‘downregulated miRNA–upreg-
ulated mRNA’ pairs.
Validation of differential miRNA expression and
miRNA–target mRNA correlation
We next evaluated the expression of miR-206 and miR-532-3p
and their respective target mRNAs in an independent cohort of
patients with proteinuric nephropathies (i.e. the validation
cohort). Both miRNAs were significantly downregulated in
progressive subjects (Fig. 3). Further, we also analysed the
correlation between the miRNA expression values of miR-206

European Journal of Clinical Investigation Vol 46 217

 M. RUDNICKI ET AL. www.ejci-online.com

Table 3 Degree and distribution of histological damage in the discovery and in the validation cohort
Discovery cohort Intra-cohort Validation cohort Intra-cohort Inter-cohort
All Stable Progressive P-valuea All Stable Progressive P-valuea P-value

Glomerulosclerosis present % of bx 22 13 50 < 00001 11 9 20 n.s. n.s.

Tubular atrophy
None % of bx 40 55 0 < 00001 67 73 40 < 00001 < 005 b,c

Mild % of bx 14 19 0 22 18 40
Moderate % of bx 26 13 58 11 9 20
Severe % of bx 21 13 42 0 0 0
Interstitial fibrosis
None % of bx 28 39 0 < 00001 44 50 20 < 00001 < 005 b,c

Mild % of bx 37 42 3 41 36 60
Moderate % of bx 26 13 58 15 14 20
Severe % of bx 9 6 39 0 0 0
Interstitial Median % 0 0 20 < 00001 0 0 5 n.s. n.s.
inflammation of tissue
Arteriosclerosis (intima fibrosis)
None % of bx 45 59 0 < 00001 56 64 20 < 00001 < 005b,c
Mild % of bx 16 14 22 33 23 80
Moderate % of bx 29 21 56 4 5 0
Severe % of bx 11 7 22 7 9 0
Arteriosclerosis (hyalinosis)
None % of bx 74 84 46 < 00001 85 82 100 < 00001 < 005b,c
Mild % of bx 12 13 9 11 14 0
Moderate % of bx 7 0 27 4 5 0
Severe % of bx 7 3 18 0 0 0

We used chi-square test (glomerulosclerosis), chi-square test for trend (tubular atrophy, interstitial fibrosis, arteriosclerosis/intima fibrosis and arteriosclerosis/
hyalinosis), and Mann–Whitney test (interstitial inflammation), respectively, for statistical analysis. n.s., not significant.
a
Stable and progressive subjects within the respective cohort (intra-cohort comparison).
b
Stable subjects.
c
Progressive subjects between discovery and validation cohorts (inter-cohort comparisons).

and miR-532-3p and the respective target mRNA levels iden-

tified in the discovery cohort. In the validation cohort, miR-206
Association of miRNAs and mRNAs with clinical and
expression correlated negatively and significantly with CCL19,
histological parameters
We analysed the association of miR-206, miR-532-3p, and the
CXCL1, IFNAR2, NCK2, PTK2B, PTPRC and TNFRSF25, while
12 target mRNAs with clinical and histological parameters
no correlation was found with the expression levels of
(Table 6). In the discovery cohort, correlation was strongest
RASGRP1. miR-532-3p correlated negatively and significantly
between these miRNAs/mRNAs and creatinine at the time of
with MAP3K14 and TRAF2, while there was no correlation
biopsy, creatinine at the time of follow-up, and proteinuria at
with the expression of TNFRSF10B and TRADD (Table 5).
the time of follow-up respectively. There were no associa-
In summary, we validated the results from the discovery
tions of miRNA and mRNA levels with proteinuria at the
cohort on the miRNA level for miR-206 and miR-532-3p, and
time of biopsy. The association of rising creatinine and
also for the majority of the predicted target mRNAs, which
elevated levels of proteinuria during follow-up is a clinical
were identified in the discovery cohort.

218 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
 MIRNA-RNA PROFILING IN CKD

Table 4 Differentially expressed miRNAs, and miRNA–mRNA Table 4 Continued

correlation analysis in the discovery cohort
A B C
A B C Log fold # Corr # Corr mRNA
Log fold # Corr # Corr mRNA miRNA change mRNAs targets
miRNA change mRNAs targets
hsa-miR-30e* 078 166 NA
hsa-miR-150 302 81 36
hsa-miR-516a-3p 075 0 NA
hsa-miR-142-5p 281 96 50
hsa-miR-128 072 12 1
hsa-miR-223* 267 92 NA
hsa-miR-193b 072 6 2
hsa-miR-142-3p 227 88 28
hsa-miR-532-3p 072 146 57
hsa-miR-31* 197 84 NA
hsa-miR-194 069 146 37
hsa-miR-31 195 84 40
hsa-miR-192 068 114 26
hsa-miR-183* 190 70 NA
hsa-miR-204 067 131 55
hsa-miR-642a 189 59 NA
hsa-miR-302a 066 0 NA
hsa-miR-629* 181 9 NA
hsa-miR-520d-3p 065 1 0
hsa-miR-223 178 92 38
hsa-miR-192* 063 120 NA
hsa-miR-125b-1* 176 12 NA
hsa-miR-551b 062 0 NA
hsa-miR-135b* 176 62 NA
hsa-miR-140-3p 061 107 34
hsa-miR-224 170 40 23
hsa-miR-378a-5p 058 42 NA
hsa-miR-146b-5p 168 14 3
hsa-miR-190 055 113 24
hsa-miR-96 158 17 8
hsa-miR-206 053 73 31
hsa-miR-21* 156 24 NA
hsa-miR-23B 051 16 3
hsa-miR-33a* 156 0 NA
hsa-miR-601 049 1 1
hsa-miR-511 150 31 13
hsa-miR-184 047 16 6
hsa-let-7i* 148 0 NA
hsa-miR-1305 043 1 0
hsa-miR-155 142 15 7
hsa-miR-216B 035 24 8
hsa-miR-193a-5p 134 67 30 Fifty-five miRNAs were differentially expressed between stable and pro-
hsa-miR-1290 133 2 0 gressive cases; of these, 47 miRNAs correlated inversely and significantly
with at least one mRNA, and of these, 29 miRNAs correlated with at least one
hsa-miR-135b 133 26 13 predicted mRNA target (these miRNAs are shown in bold). Log fold change
values of miRNAs (progressive vs. stable) are shown (column A). Values > 1
hsa-miR-378a-3p 120 3 0 represent upregulation, and values < 1 represent downregulation in pro-
gressive subjects. Number of negatively correlating mRNAs (column B), and
hsa-miR-101 089 1 0 number of predicted mRNA targets (column C) are shown. NA, no informa-
tion on target prediction available.
hsa-miR-29a* 086 1 NA
hsa-miR-27b 082 21 6
hallmark of progressive CKD. Therefore, the correlation of
hsa-miR-744 081 9 5
the identified miRNAs and their respective target mRNAs
hsa-miR-148a 080 65 25 with these clinical parameters further underscores their pos-
hsa-miR-22* 080 0 NA sible involvement in progression of CKD rather than repre-
senting a mere association with cross-sectional parameters.
hsa-miR-29c 080 1 0
These associations were reproduced for > 70% of the
hsa-let-7e* 079 0 NA miRNAs and mRNAs in the validation cohort. The remaining
hsa-miR-190b 078 0 NA results in the validation cohort showed at least the same
direction of correlation (i.e. negative correlation of miR-532-
hsa-miR-30d 078 151 43
3p and a positive correlation of mRNAs with creatinine and

European Journal of Clinical Investigation Vol 46 219

 M. RUDNICKI ET AL.
proteinuria), but these correlations did not reach statistical
significance. These results can be explained in part by vari-
ations in clinical characteristics of the discovery and valida-
tion cohorts, and by the limited numbers of samples included
in this analysis.
Associations with pathohistological changes in the renal
biopsies are summarized in Fig. 4. In line with the results
from pathway analysis, both miRNAs and the 12 selected
target mRNAs correlated significantly with the presence of
interstitial inflammatory infiltrates in both patient
cohorts. Furthermore, most of the target mRNAs correlated
positively with the degree of tubular atrophy and
interstitial fibrosis.
220 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
www.ejci-online.com
Figure 2 Pathway analysis of target
mRNAs correlating with differentially
regulated miRNA. Black squares indicate
miRNA–mRNA target pairs. Open circles
indicate biological pathways that include
mRNA targets from differentially regulated
miRNAs. CA, cell adhesion; A, apoptosis;
SIG, signalling.
Discussion
In the present study, we measured renal miRNA and mRNA
expression in renal biopsies from patients with stable and
progressive CKDs, and the results were analysed using an
integrative analysis approach. We identified a set of 55 miRNAs
being differentially expressed between stable and progressive
subjects. As in most cases miRNAs interfere with the mRNA
target by inducing transcript degradation [28], we focused on
the analysis of inversely correlated miRNA–target mRNA
pairs. Forty-seven of the 55 differentially expressed miRNAs
correlated negatively with at least one mRNA, and 29 of the 55
miRNAs correlated with at least one predicted target mRNA. In

Figure 3 Expression levels of miR-206
and miR-532-3p in the discovery and
validation cohorts. Median values are
shown with Box–Whisker plots including
the 10th and the 90th percentile.
Normalized quantitative Ct-values are
shown. *P = 00010, †P < 0005, ‡P < 005
(Mann–Whitney test).
conditions with a complex phenotype, such as progressive
CKD, an integrative analysis approach may be more appro-
priate to identify pathophysiologically relevant molecular pro-
cesses, rather than an approach focusing on single miRNAs or
transcripts [29]. This integrative approach utilizing multi-level
information from various sources (i.e. miRNA and mRNA
expression data, miRNA–mRNA target prediction, and path-
way analysis) is expected to support the identification of
specific miRNAs that potentially target disease-relevant
mRNAs, thereby altering the activity of whole biological path-
ways. Using such an approach, we specifically concluded
regarding seven of the 55 miRNAs of the discovery cohort that
correlated with 29 upregulated target mRNAs being mainly
associated with pathways involved in inflammation, cell
adhesion, apoptosis and intra-cellular signalling.
These seven miRNAs were all expressed at a lower level in
progressive subjects and include miR-30d, miR-140-3p,
miR-532-3p, miR-194, miR-190, miR-204 and miR-206. Very
limited results have been published on the expression of these
miRNAs in the context of acute and chronic human renal
MIRNA-RNA PROFILING IN CKD
diseases. miRNAs miR-194 and miR-204 were reported to be
specifically expressed in kidney tissue [30], and miR-194 is
preferentially expressed in the renal cortex [31]. TGF-ß1 was
shown to downregulate miR-194 expression in proximal
tubular epithelial cells via a decrease in hepatocyte nuclear
factor [32]. In a recent report on miRNA-profiles in chronic
allograft dysfunction, miR-204 was significantly downregu-
lated in kidney tissue with interstitial fibrosis and tubular
atrophy as compared with normal grafts, which also correlated
with reduced levels of miR-204 in paired urine samples [33].
Furthermore, a reduction of miR-204 expression in epithelial
cells has been shown to be associated with reduced expression
of claudins 10, 16 and 19, suggesting a critical (albeit
indirect) role of this miRNA in maintaining epithelial cell
function [34]. It should be noted that we also detected a
significant decrease of miR-192 expression in progressive
subjects, although this miRNA is not included in the final list of
the seven downregulated miRNAs. Of note, miR-192 and
miR-194 are co-transcribed from the shared precursor primary
miRNA transcript pri-miR-192/194, and both miRNAs show
European Journal of Clinical Investigation Vol 46 221
 M. RUDNICKI ET AL. www.ejci-online.com

Table 5 miRNA–target mRNA correlations in the discovery and Table 6 Relationship of selected miRNA and mRNA levels with
in the validation cohort clinical parameters
Discovery Validation Time of biopsy Time of follow-up
Target mRNA R P R P miRNA/mRNA Age Creatinine UPCR Creatinine UPCR

miR-206 Discovery cohort

CCL19 047 00014 051 00050 miR-206 – 051 – 060 058
CXCL1 046 00021 058 00009 miR-532-3p – 062 – 064 051
IFNAR2 052 00004 062 00004 CCL19 – 064 – 048 042
NCK2 042 00054 053 00032 CXCL1 032 056 – 068 063
PTK2B 037 00158 066 00001 IFNAR2 – 071 – 051 038
PTPRC 053 00002 048 00078 NCK2 – 068 – 079 052
RASGRP1 041 00060 012 05533 PTK2B – 054 – 068 045
TNFRSF25 058 < 00001 045 00136 PTPRC 042 054 – 057 055
miR-532-3p RASGRP1 – 068 – 062 060
MAP3K14 056 < 00001 050 00059 TNFRSF25 033 051 – 078 050
TNFRSF10B 054 00002 033 00813 MAP3K14 – 058 – 060 050
TRADD 047 00014 006 07752 TNFRSF10B – 062 – 061 056
TRAF2 055 00001 055 00019 TRADD – 063 – 057 041

Pearson R and P-values (two-tailed, 95% confidence interval) are shown.

TRAF2 – 066 – 074 043
Non-significant Pearson R (P > 005) are shown in italics. Validation cohort
coordinated regulation [32]. Similar to our results, Krupa et al. miR-206 – 045 – 044 –
identified loss of miR-192 expression in patients with already
miR-532-3p – – – 044 –
progressed diabetic nephropathy (‘late presenters’) as com-
pared with early stages of diabetic nephropathy [12]. In line CCL19 – 046 – 055 064
with our findings, the authors showed also an approximately CXCL1 – 051 – 058 041
two-fold downregulation of miR-194 and miR-204 in pro-
IFNAR2 – 048 – – –
gressed (i.e. advanced) diabetic nephropathy biopsies. To our
knowledge, this is the first report that shows a significant NCK2 – 056 – 044 041
differential expression of miR-30d, miR-140-3p, miR-190, PTK2B – 040 – – –
miR-206 and miR-532-3p being associated with a progressive PTPRC – 040 – 045 045
decline of renal function.
RASGRP1 – – – – –
The integrative analysis approach using negative correlation
of miRNA and target mRNA expression data with pathway TNFRSF25 – 058 – 050 049
analysis of the target mRNAs revealed a particular association MAP3K14 – 048 – 044 045
of two of the seven miRNAs with specific biological pathways:
TNFRSF10B – 054 – 043 042
miR-206 correlated with an upregulation of eight predicted
targets involved in pathways being associated with inflamma- TRADD – – – – 043
tion, and miR-532-3p correlated negatively with four target TRAF2 – 058 – 045 043
transcripts involved in apoptosis.
The miRNA–mRNA pairs mentioned above showed a strik- Only correlation coefficients with a P-value (Spearman) < 005 are shown.
UPCR, urine protein–creatinine ratio.
ing association with the presence of inflammatory infiltrates in
the renal biopsies. Renal inflammation, particularly the influx
of inflammatory cells in the tubulointerstitial compartment, shown) is essential in establishing dendritic cell and T-cell
underlies many CKDs. The interaction of CCL19 and its recruitment and antigen-specific T-cell proliferation [35]. It was
receptor CCR7 (both also upregulated in our cohorts, data not shown that the expression of both CCL19- and CCR7-mRNA in

222 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation

Figure 4 Associations of selected miRNAs and mRNAs with
histological parameters. Black squares: P < 005. For statistical
tests, see Methods section.
a rat model of CKD was associated with the migration and
activation of lymphocytes and antigen-presenting cells [36].
Increased renal expression of CXCL1 in ischaemia–reperfusion
injury (IRI) was associated with marked influx of neutrophils,
suggesting a role of CXCL1 in neutrophil chemotaxis in renal
injury [37]. Treatment with blocking antibodies to CXCL1 or a
MIRNA-RNA PROFILING IN CKD
CXCR2 inhibitor during IRI blocked interstitial neutrophil
inflammation, reduced the degree of histological damage and
increased survival in mice [37,38]. Our study is the first report
to show enhanced renal expression of CCL19, CCR7 and
CXCL1 being associated with progression of CKD.
Diminished expression of miR-206 has been linked to
increased levels of tumour necrosis factor-alpha (TNF-a) in
human inflammatory myopathies, particularly in dermato-
myositis. It was further shown in vitro that TNF-a reduces miR-
206 expression and suppresses myocyte/myotube differentia-
tion [39]. In our discovery and validation cohorts, the majority
(approximately 75%) of TNF-a-associated transcripts such as
TNF-a, TNF-a-induced proteins and members of the TNF
receptor superfamily showed an upregulation in progressive
subjects, indicating an activation of TNF-a-associated inflam-
mation also in our cohorts (data not shown).
To our knowledge, this paper is the first to describe regula-
tion of miR-532-3p expression in progressive human CKD. In
patients with Crohn0 s disease, blood levels of miR-532-3p were
significantly higher as compared with healthy controls [40].
Little is known about the regulation of miR-532-3p expression,
but it has been shown that treatment with nerve growth factor
(NGF) downregulated miR-532-3p expression [41]. NGF is a
promoter of growth and survival also of renal cells [42], and it
has been shown that the expression of NGF is elevated in renal
tissue in various proteinuric kidney diseases [43], in peripheral
blood mononuclear cells from patients with CKD [44], and in
blood from renal transplant recipients [45]. In the discovery
cohort, miR-532-3p downregulation correlated negatively with
upregulation of several target transcripts involved in apoptosis
and inflammation pathways and being associated with the
TNF-a superfamily, such as TNFRSF10B, TRADD and TRAF2.
In addition, we also identified an upregulation of further TNF-
associated transcripts such as TANK (neg. corr. with miR-190),
TNFRSF10D (also known as TRAIL-R4) and TNFSF10 (also
termed TRAIL) (neg. corr. with miR-30d), and TNFRSF25 (neg.
corr. with miR-206). It has been increasingly recognized that
several members of the TNF superfamily (both ligands and
receptors) play a major role in the pathogenesis of kidney
injury, promoting inflammation, apoptosis, and accumulation
of extracellular matrix [46]. The glomerular and tubular
expression of the death ligand TRAIL was upregulated in
human diabetic nephropathy, and the level of expression cor-
related with disease severity [47]. The pro-apoptotic activity of
TRAIL is triggered by binding to a multiple set of receptors,
including TNFRSF10B/TRAIL-R2, which has been shown to be
expressed particularly in renal tubule cells [48]. The activation
of TRADD has been associated with induction of apoptosis,
while the recruitment of TRAF2 leads to the activation of
transcription factors such as nuclear factor-kappa B (NF-jB)
and c-Jun kinase, thus promoting cell survival and
European Journal of Clinical Investigation Vol 46 223
 M. RUDNICKI ET AL.
differentiation and also immune and inflammatory responses
[49]. Zhu et al. [50] showed, in various types of lupus nephritis,
an upregulation of TNF-a, TRADD and TRAF2 in glomeruli,
proximal and distal tubules and in interstitial mononuclear
cells, and the levels of expression correlated positively with
disease severity. Overexpression of TANK (also called I-TRAF)
has been shown to inhibit TRAF2-mediated NF-jB activation,
and presumably also TRAF2-induced apoptosis [51].
TNFRSF25 is induced in human renal tubular epithelial cells in
response to injury [52], and mediates activation of NF-kB and
caspase-3 respectively [53]. In summary, the involvement of
multiple TNF superfamily cytokines in the progression of CKD,
as shown in our work, is in line with published data. However,
in human kidney disease, only limited data have been pub-
lished on the association of renal miRNA expression and the
regulation of the respective TNF target mRNA, or on the effect
of TNF-a on renal miRNA expression respectively [54,55].
This study has several strengths and limitations. Using cry-
ocut material from renal biopsies, we were able to analyse
miRNA- and mRNA-profiles from the same subjects. The results
from the discovery cohort of 43 subjects were largely validated
in an independent cohort of 29 patients, underscoring the
validity of the findings. Both cohorts included patients with
various kinds of proteinuric kidney diseases; thus, the results
from this study may represent general molecular changes
associated with progression of CKD. Furthermore, the mean
follow-up times in the cohorts were 43 and 50 months respec-
tively; thus, the differences in miRNA and mRNA expression
between stable and progressive cases may have potential as
novel biomarkers of CKD progression or represent pathophys-
iologically relevant processes before irreversible damage has
occurred. A major limitation of this study is its descriptive
nature without any in vitro evidence that the identified miRNAs
actually regulate the respective transcripts. Putative targets of
the miRNAs were identified using several established algo-
rithms, but none of these targets has been validated experi-
mentally. Focusing on predicted targets also excludes potential
previously unknown mRNA targets that are not identified by
any of the established algorithms. We did not analyse protein
expression; therefore, we cannot tell if these changes translate
into changes of target proteins. Using a negative miRNA–
mRNA correlation approach excludes more complicated
miRNA regulatory mechanisms, where miRNAs interact with
transcription factors, thus inducing or repressing the expression
of multiple genes that are not a direct target of the respective
miRNA. Finally, we were not able to validate inverse correlation
for RASGRP1 as a target of miR-206, and TNFRSF10B and
TRADD as targets of miR-532-3p respectively. Nevertheless,
TNFRSF10B and TRADD correlated with chronic histological
changes such as interstitial fibrosis and inflammation. These
results suggest the presence of regulators (e.g. inflammatory
224 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
www.ejci-online.com
cytokines) that potentially override the targeting effect of the
respective miRNAs [47,50]. These considerations have to be
taken into account when the results of this study are interpreted
in the complex molecular environment of progressive CKD.
Acknowledgements
We kindly thank Paul K€ onig and Karl Lhotta for establishing
the renal biopsy repository. This study received funding from
the European Union’s Seventh Framework Programme under
grant agreement n° HEALTH-F2-2009-241544.
Address
Department of Internal Medicine IV – Nephrology and Hyper-
tension, Medical University Innsbruck, Anichstrasse 35, 6020
Innsbruck, Austria (M. Rudnicki, J. Leierer, N. Schweibert, J.
Sunzenauer, A. Kronbichler, G. Mayer); Emergentec Biodevel-
opment GmbH, Gersthofer Strasse 29–31, 1180 Vienna, Austria
(P. Perco, A. Heinzel, I. M€uhlberger, B. Mayer); Biogazelle,
Technologiepark-Zwijnaarde 3, 9052 Gent, Belgium (B. D´haene,
P. Mestdagh, J. Vandesompele); Department of Nephrology, KH
Elisabethinen, Fadingerstrasse 1, 4020 Linz, Austria (J. Sun-
zenauer); Medical University Vienna, Institute of Pathology,
Währinger G€ urtel 18–20, 1090 Vienna, Austria (H. Regele).
Correspondence to: Michael Rudnicki, MD, FASN, Priv.-Doz.,
Department of Internal Medicine IV – Nephrology and
Hypertension, Medical University Innsbruck, Anichstrasse 35,
6020 Innsbruck, Austria. Tel.: +43 512 504 81337 or +43 664
5377554; fax: +43 512 504 25857; e-mail: michael.rudnicki@
i-med.ac.at
Received 31 May 2015; accepted 20 December 2015
References
1 Liu Y. Cellular and molecular mechanisms of renal fibrosis. Nat Rev
Nephrol 2011;7:684–96.
2 Mayer G. Capillary rarefaction, hypoxia, VEGF and angiogenesis in
chronic renal disease. Nephrol Dial Transplant 2011;26:1132–7.
3 Ju W, Smith S, Kretzler M. Genomic biomarkers for chronic kidney
disease. Transl Res 2012;159:290–302.
4 Yasuda Y, Cohen CD, Henger A, Kretzler M. Gene expression
profiling analysis in nephrology: towards molecular definition of
renal disease. Clin Exp Nephrol 2006;10:91–8.
5 van Rooij E. The art of microRNA research. Circ Res 2011;108:219–34.
6 Gebeshuber CA, Kornauth C, Dong L, Sierig R, Seibler J, Reiss M
et al. Focal segmental glomerulosclerosis is induced by microRNA-
193a and its downregulation of WT1. Nat Med 2013;19:481–7.
7 Khella HW, Bakhet M, Lichner Z, Romaschin AD, Jewett MA,
Yousef GM. MicroRNAs in kidney disease: an emerging
understanding. Am J Kidney Dis 2013;61:798–808.
8 Deshpande SD, Putta S, Wang M, Lai JY, Bitzer M, Nelson RG et al.
Transforming growth factor-beta-induced cross talk between p53
and a microRNA in the pathogenesis of diabetic nephropathy.
Diabetes 2013;62:3151–62.

9 Wang G, Kwan BC, Lai FM, Choi PC, Chow KM, Li PK et al.
Intrarenal expression of miRNAs in patients with hypertensive
nephrosclerosis. Am J Hypertens 2010;23:78–84.
10 Wang G, Kwan BC, Lai FM, Choi PC, Chow KM, Li PK et al.
Intrarenal expression of microRNAs in patients with IgA
nephropathy. Lab Invest 2010;90:98–103.
11 Wang G, Kwan BC, Lai FM, Chow KM, Kam-Tao Li P, Szeto CC.
Expression of microRNAs in the urinary sediment of patients with
IgA nephropathy. Dis Markers 2010;28:79–86.
12 Krupa A, Jenkins R, Luo DD, Lewis A, Phillips A, Fraser D. Loss of
MicroRNA-192 promotes fibrogenesis in diabetic nephropathy. J Am
Soc Nephrol 2010;21:438–47.
13 Putta S, Lanting L, Sun G, Lawson G, Kato M, Natarajan R.
Inhibiting microRNA-192 ameliorates renal fibrosis in diabetic
nephropathy. J Am Soc Nephrol 2012;23:458–69.
14 Mestdagh P, Feys T, Bernard N, Guenther S, Chen C, Speleman F
et al. High-throughput stem-loop RT-qPCR miRNA expression
profiling using minute amounts of input RNA. Nucleic Acids Res
2008;36:e143.
15 Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J.
qBase relative quantification framework and software for
management and automated analysis of real-time quantitative PCR
data. Genome Biol 2007;8:R19.
16 D’Haene B, Mestdagh P, Hellemans J, Vandesompele J. miRNA
expression profiling: from reference genes to global mean
normalization. Methods Mol Biol 2012;822:261–72.
17 Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI
gene expression and hybridization array data repository. Nucleic
Acids Res 2002;30:207–10.
18 Wang X. miRDB: a microRNA target prediction and functional
annotation database with a wiki interface. RNA 2008;14:1012–7.
19 Wang X, El Naqa IM. Prediction of both conserved and
nonconserved microRNA targets in animals. Bioinformatics
2008;24:325–32.
20 Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often
flanked by adenosines, indicates that thousands of human genes are
microRNA targets. Cell 2005;120:15–20.
21 Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ. miRBase:
tools for microRNA genomics. Nucleic Acids Res 2008;36:D154–8.
22 Megraw M, Sethupathy P, Corda B, Hatzigeorgiou AG. miRGen: a
database for the study of animal microRNA genomic organization
and function. Nucleic Acids Res 2007;35:D149–55.
23 Miranda KC, Huynh T, Tay Y, Ang YS, Tam WL, Thomson AM
et al. A pattern-based method for the identification of MicroRNA
binding sites and their corresponding heteroduplexes. Cell
2006;126:1203–17.
24 Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E. The role of site
accessibility in microRNA target recognition. Nat Genet
2007;39:1278–84.
25 da Huang W, Sherman BT, Lempicki RA. Systematic and integrative
analysis of large gene lists using DAVID bioinformatics resources.
Nat Protoc 2009;4:44–57.
26 Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and
genomes. Nucleic Acids Res 2000;28:27–30.
27 Kanehisa M, Goto S, Sato Y, Kawashima M, Furumichi M, Tanabe
M. Data, information, knowledge and principle: back to metabolism
in KEGG. Nucleic Acids Res 2014;42:D199–205.
28 Guo H, Ingolia NT, Weissman JS, Bartel DP. Mammalian
microRNAs predominantly act to decrease target mRNA levels.
Nature 2010;466:835–40.
MIRNA-RNA PROFILING IN CKD
29 He JC, Chuang PY, Ma’ayan A, Iyengar R. Systems biology of
kidney diseases. Kidney Int 2012;81:22–39.
30 Sun Y, Koo S, White N, Peralta E, Esau C, Dean NM et al.
Development of a micro-array to detect human and mouse
microRNAs and characterization of expression in human organs.
Nucleic Acids Res 2004;32:e188.
31 Tian Z, Greene AS, Pietrusz JL, Matus IR, Liang M. MicroRNA-
target pairs in the rat kidney identified by microRNA microarray,
proteomic, and bioinformatic analysis. Genome Res 2008;
18:404–11.
32 Jenkins RH, Martin J, Phillips AO, Bowen T, Fraser DJ.
Transforming growth factor beta1 represses proximal tubular cell
microRNA-192 expression through decreased hepatocyte nuclear
factor DNA binding. Biochem J 2012;443:407–16.
33 Scian MJ, Maluf DG, Archer KJ, Suh JL, Massey D, Fassnacht RC
et al. Gene expression changes are associated with loss of kidney
graft function and interstitial fibrosis and tubular atrophy: diagnosis
versus prediction. Transplantation 2011;91:657–65.
34 Wang FE, Zhang C, Maminishkis A, Dong L, Zhi C, Li R et al.
MicroRNA-204/211 alters epithelial physiology. FASEB J
2010;24:1552–71.
35 Ziegler E, Gueler F, Rong S, Mengel M, Witzke O, Kribben A et al.
CCL19-IgG prevents allograft rejection by impairment of immune
cell trafficking. J Am Soc Nephrol 2006;17:2521–32.
36 Satirapoj B, Bruhn KW, Nast CC, Wang Y, Dai T, Lapage J et al.
Oxidized low-density lipoprotein antigen transport induces
autoimmunity in the renal tubulointerstitium. Am J Nephrol
2012;35:520–30.
37 Miura M, Fu X, Zhang QW, Remick DG, Fairchild RL.
Neutralization of Gro alpha and macrophage inflammatory protein-
2 attenuates renal ischemia/reperfusion injury. Am J Pathol
2001;159:2137–45.
38 Cugini D, Azzollini N, Gagliardini E, Cassis P, Bertini R, Colotta F
et al. Inhibition of the chemokine receptor CXCR2 prevents kidney
graft function deterioration due to ischemia/reperfusion. Kidney Int
2005;67:1753–61.
39 Georgantas RW, Streicher K, Greenberg SA, Greenlees LM, Zhu W,
Brohawn PZ et al. Inhibition of myogenic microRNAs 1, 133, and
206 by inflammatory cytokines links inflammation and muscle
degeneration in adult inflammatory myopathies. Arthritis Rheumatol
2014;66:1022–33.
40 Paraskevi A, Theodoropoulos G, Papaconstantinou I, Mantzaris G,
Nikiteas N, Gazouli M. Circulating MicroRNA in inflammatory
bowel disease. J Crohn’s Colitis 2012;6:900–4.
41 Hamada N, Fujita Y, Kojima T, Kitamoto A, Akao Y, Nozawa Y et al.
MicroRNA expression profiling of NGF-treated PC12 cells revealed
a critical role for miR-221 in neuronal differentiation. Neurochem Int
2012;60:743–50.
42 Sariola H, Saarma M, Sainio K, Arumae U, Palgi J, Vaahtokari A
et al. Dependence of kidney morphogenesis on the expression of
nerve growth factor receptor. Science 1991;254:571–3.
43 Bonofiglio R, Antonucci MT, Papalia T, Romeo F, Capocasale G,
Caroleo MC et al. Nerve growth factor (NGF) and NGF-receptor
expression in diseased human kidneys. J Nephrol 2007;20:186–95.
44 Antonucci MT, Bonofiglio R, Papalia T, Caruso F, Caroleo MC,
Mancuso D et al. Nerve growth factor and its monocyte receptors
are affected in kidney disease. Nephron Clin Pract 2009;111:c21–8.
45 Gigliotti P, Lofaro D, Leone F, Perri A, Vizza D, Papalia T et al. High
nerve growth factor blood concentration in renal transplantation: a
new prognostic marker? Transplant Proc 2013;45:2654–6.
European Journal of Clinical Investigation Vol 46 225
 M. RUDNICKI ET AL.
46 Sanchez-Nino MD, Benito-Martin A, Goncalves S, Sanz AB, Ucero
AC, Izquierdo MC et al. TNF superfamily: a growing saga of kidney
injury modulators. Mediators Inflamm 2010;2010:pii: 182958.
47 Lorz C, Benito-Martin A, Boucherot A, Ucero AC, Rastaldi MP,
Henger A et al. The death ligand TRAIL in diabetic nephropathy.
J Am Soc Nephrol 2008;19:904–14.
48 Spierings DC, de Vries EG, Vellenga E, van den Heuvel FA,
Koornstra JJ, Wesseling J et al. Tissue distribution of the death
ligand TRAIL and its receptors. J Histochem Cytochem 2004;52:821–31.
49 Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor
superfamilies: integrating mammalian biology. Cell 2001;104:487–
501.
50 Zhu L, Yang X, Ji Y, Chen W, Guan W, Zhou SF et al. Up-regulated
renal expression of TNF-alpha signalling adapter proteins in lupus
glomerulonephritis. Lupus 2009;18:116–27.
51 Rothe M, Xiong J, Shu HB, Williamson K, Goddard A, Goeddel DV. I-
TRAF is a novel TRAF-interacting protein that regulates TRAF-
mediated signal transduction. Proc Natl Acad Sci USA 1996;93:8241–6.
52 Al-Lamki RS, Wang J, Thiru S, Pritchard NR, Bradley JA, Pober JS
et al. Expression of silencer of death domains and death-receptor-3
in normal human kidney and in rejecting renal transplants. Am J
Pathol 2003;163:401–11.
226 ª 2015 Stichting European Society for Clinical Investigation Journal Foundation
www.ejci-online.com
53 Al-Lamki RS, Wang J, Tolkovsky AM, Bradley JA, Griffin JL, Thiru S
et al. TL1A both promotes and protects from renal inflammation and
injury. J Am Soc Nephrol 2008;19:953–60.
54 Zager RA, Johnson AC, Lund S. Uremia impacts renal inflammatory
cytokine gene expression in the setting of experimental acute kidney
injury. Am J Physiol Renal Physiol 2009;297:F961–70.
55 Imaizumi T, Tanaka H, Tajima A, Yokono Y, Matsumiya T, Yoshida
H et al. IFN-gamma and TNF-alpha synergistically induce
microRNA-155 which regulates TAB 2/IP-10 expression in human
mesangial cells. Am J Nephrol 2010;32:462–8.
Supporting Information
Additional Supporting Information may be found in the online
version of this article:
Table S1. Detailed clinical characteristics of the subjects from
the discovery and the validation cohort summarized in Table 1.