Animal Model

Kidneys from wild type (WT; Pkd1 ^+/+ ) and Pkd1-null, designated as Pkd1 ^-/- , mutant littermates were investigated at the embryonic ages 14 (E14.5) and 17 (E17.5). The kidneys were fixed in 4% paraformaldehyde overnight, paraffin-embedded, and stained with hematoxylin and eosin. The sections were visualized with a Nikon Eclipse 80i microscope and photographed with a Qimaging Retiga 2000R Fast 1394 camera using NIS-Elements Basic Research 2.34 software (Micro Video Instruments, Avon, MA).

Total RNA extraction and Microarrays

Total RNA was extracted from embryonic kidneys at, E14.5 and E17.5, using the Qiagen miRNeasy Mini kit (Qiagen, Valencia, CA). Three pairs of kidneys at E14.5 and E17.5 were processed. Each pair was analyzed separately. The integrity and purity of the mRNA samples were assessed prior to hybridization using Bioanalyser 2100 with mRNA Nanochips (Agilent Technologies). RNA hybridization was performed, and gene expression profiles were determined using Illumina Mouse Sentrix 6 version 2 Beadchips. The data extraction was performed by using an Illumina Bead Studio V3.1.3.0 software with the output being raw, non-normalized bead summary values.

The raw data matrix extracted from Beadstudio was uploaded in Bioconductor R version 2.9.1 for downstream analysis. The raw data were read using the beadarray package available through the Bioconductor project. The raw intensities were background adjusted by the Subtract method. The log2 summarized data were quantile normalized. The limma package [26] with empirical Bayes method was used to assess the differentially expressed genes.

The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus (GEO) and are accessible through GEO Series accession number GSE24352 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE24352).

Gene set enrichment analysis and Gene Ontology enrichment analysis

Gene set enrichment analysis (GSEA) (http://www.broad.mit.edu/gsea/) was used to identify potential gene pathways and key transcription factors (TFs) that may modulate cystogenesis (Figure 1). The GSEA C2 database (curated gene sets from BioCarta, KEGG, Signaling Pathway database, Signaling Gateway, Signal Transduction Knowledge Environment, Human Protein Reference database, GenMAPP, Sigma-Aldrich Pathways, Gene Arrays, BioScience Corp., Human Cancer Genome Anatomy Consortium database) that includes well-studied metabolic and signaling pathways and published microarray data sets was used for pathway analysis. The GSEA C3 database containing shared and evolutionarily conserved TF binding motifs defined by the TRANSFAC database was used for TF analysis. The description of each gene set can be found on the GSEA Molecular Signatures Database website: http://www.broad.mit.edu/gsea/msigdb/index.jsp Enrichment of Gene Ontology (GO) categories (Figure 1) in the differentially expressed genes was determined using Bioconductor's GOstats package [27] for both up- and down-regulated genes.

Comparison with other datasets on ADPKD

Two datasets on ADPKD with mutation in Pkd1, Pkd1 ^L3/L3 [28] and human ADPKD [29] were obtained from the GEO database (http://www.ncbi.nlm.nih.gov/geo/; [30]). Raw data were log2 summarized and quantile normalized. The limma package [26] with empirical Bayes method was used to assess the differentially expressed genes between mutants and controls. These were compared with the differentially expressed genes obtained in comparisons using custom written Perl scripts.

miRNA prediction

As shown in Figure 1, a combinatorial strategy was used where target miRNAs were predicted for the differentially expressed genes using four algorithms, TargetScan (http://www.targetscan.org/; [31]), miRanda (http://www.microrna.org/; [32]), miRDB (http://mirdb.org/miRDB/; [33]) and microT (http://diana.cslab.ece.ntua.gr/microT/; [34]). To identify the miRNAs commonly predicted by two or more algorithms, results were intersected (Additional file 1) using custom written Perl scripts (Additional file 2).

Quantitative real-time PCR

Total RNA was extracted from Pkd1 ^-/- and WT kidneys. Three independent biological replicates (different from those used for microarrays) were used and all reactions were run in duplicates for all the expression analysis of genes and miRNAs. For all the quantitative real-time-PCR (qPCR) assays, the ABI PRISM 7300 Sequence Detection System was used. The comparative CT method was used to obtain relative quantitation of genes and miRNAs as per the manufacturer's protocol (Qiagen). For gene expression analysis, cDNA was made using Superscript III reverse transcriptase (Invitrogen). qPCRs were performed, using Power SYBR^® Green PCR Master Mix (Invitrogen). Primer sets for P2rx7, Cpeb3, Hdac9, Sox6, Ltbp1, Calcr, Pitx2, Fgfr3, Fgf10, Adam22, Ddx3y, F2rl2, Grap2, Edil3, Mysm1, Alg6 and Alg8 are available in Additional file 3. 18S rRNA was used as the endogenous control.

miRNA reverse transcription was performed using the TaqMan microRNA Reverse Transcription Kit and miRNA-specific primers (Applied Biosystems, Foster City, CA). Real-time TaqMan miRNA-assays were used to quantify miRNA expression using the TaqMan Master Mix and validated primer/probe sets (Applied Biosystems). miRNA levels were normalized to snoRNA-202. Cycling conditions for TaqMan PCR consisted of an initial incubation at 50°C for 2 min and 95°C for 15 sec and 60°C for 1 min.

Pkd1^-/- mice model and design of experiment

To study the detailed changes in molecular profiles during the progression of ADPKD, we generated and compared gene expression profiles of Pkd1 ^-/- embryonic kidneys and age-matched WT kidneys at two stages, E14.5 and E17.5 (Figure 1; Additional file 4). Pkd1 ^-/- embryos develop rapidly progressive kidney cysts during embryogenesis, with null mutant kidneys showing no cysts at E14.5 and marked cystic changes by E17.5 (Figure 2). Unsupervised hierarchical clustering was able to discriminate WT and mutant kidney samples at both time points (Figure 3). At the same time, the cluster analysis was also able to distinguish changes between E14.5 and E17.5 kidneys (Figure 3). Table 1 shows a summary of number of differentially expressed genes and pathways for all of the comparisons. Genes showing a greater than 2-fold difference in expression between WT and mutant kidneys at E14.5 as well as at E17.5 (empirical Bayes moderated t-statistic, unequal variance, uncorrected p-value ≤0.05), were considered to be differentially expressed whereas in the comparisons of WT kidneys at E14.5 vs E17.5 and Pkd1 ^-/- kidneys at E14.5 vs E17.5, genes with ≥ 2-fold difference in expression at p-value ≤0.05 (corrected for multiple testing by Benjamini-Hochberg method) were considered as differentially expressed. The expression of 454 genes was significantly changed between WT and mutant kidneys at E14.5 (Table 1, Additional file 5) whereas 884 genes were significantly changed at E17.5 between WT and mutant kidneys (Table 1, Additional file 6). The comparison of E14.5 and E17.5 WT kidneys yielded 1189 differentially expressed genes (Additional file 7) whereas 2287 genes were differentially expressed in comparison of Pkd1 ^-/- kidneys at E14.5 vs E17.5 (Table 1, Additional file 8). Comparing differentially expressed genes observed in Pkd1 ^-/- at E14.5 vs E17.5 (2287 genes) and in WT at E14.5 vs E17.5 (1189 genes) identified genes that were specifically changing during development in diseased (Additional file 9) or healthy conditions (Additional file 10) as shown in Figure 4. This comparison also indicated genes that were regulated during aging from E14.5 to E17.5 regardless of the genotype (Figure 4; Additional file 11). These results indicate that maturation accounted for the greatest number of changes in gene expression, more so than the cyst formation. Nevertheless, 1397 genes (Additional file 9) could be identified for which changes in gene expression levels were specific for Pkd1 ^-/- kidneys. Thus, this analysis served to identify a set of genes specifically changing in PKD that can yield insights about the regulation of gene expression during cystogenesis.

Profiling gene expression changes in Pkd1^-/- mutants at E14.5 and E17.5

At E14.5, Pkd1 ^-/- mutants do not exhibit any cyst formation (Figure 2). Therefore gene expression changes at E14.5 may provide an indication of signaling pathways that cause cysts rather than being a consequence of cyst formation. Among the most interesting genes identified by microarrays were P2rx7, Cer1, Frzb, Wnt7b, Dvl3, Gpr34, and Gpr116. These genes are representative of calcium, Wnt and GPCR signaling, and their roles are discussed later.

Pkd1 ^-/- mutant embryos show numerous large renal cysts at E17.5. Microarray analysis showed up-regulation of multiple developmental genes such as Bmp8b, Grem1, Mmp12, Sema3c, Sema4c and Sema6c and transcription factors (TFs)- Gli2, Foxo1, Foxp2, Hoxb8, Pou2f1 and Zbtb32 at E17.5 in Pkd1 ^-/- mutant kidneys compared to WT. Many of these genes are essential for ureteric bud formation, outgrowth, and branching during kidney development. On the other hand, the expression of genes associated with the differentiation of specific nephron segments, such as Umod, Pck1, Fmo2, Slc12a3, Pvalb, and Angpt2 were down-regulated in the null mutants. These data suggest that Pkd1 ^-/- mutants have failed to maintain nephron segment-specific transcriptional signatures. Additionally, we found changes in expression of genes previously related to cyst growth and progression such as components of MAPK and JAK/STAT signaling (these are detailed later in the following sections).

Gene pathway analysis using GO term enrichment and Gene Set Enrichment Analysis

Gene Ontology (GO) and Gene Set Enrichment Analysis (GSEA; http://www.broad.mit.edu/gsea/) are two bioinformatics approaches to identify pathways involved in specific biological processes such as development and disease. GO term enrichment uses association of Gene Ontology terms to genes in a selected gene list. This method discovers what a set of genes may have in common by examining annotations and finding significant shared GO terms whereas GSEA determines whether a known set of genes shows statistically significant, concordant differences between two biological conditions, e.g. WT and Pkd1 ^-/- .

We found that most of the down-regulated gene sets in Pkd1 ^-/- mutants compared to WT at E14.5 and E17.5 represent metabolic pathways (Table 3 and Additional file 13), as suggested by GSEA and GO analysis. In contrast, both the GSEA and GO analysis of up-regulated gene sets in Pkd1 ^-/- mutants compared to WT at E14.5 and E17.5 suggested that Pkd1 ^-/- mutants displayed a rich gene transcriptional profile for kidney development and regeneration, including Wnt, calcium, MAPK and TGFβ signaling (Additional file 13). Among the pathways identified by GO and GSEA were the following:

Comparison with other ADPKD data sets in mouse and human

Transcription factor analysis of Pkd1^-/- kidneys

Computational analysis of genetic regulatory regions of co-expressed genes can furnish additional information on the regulation of a gene set by specific transcription factors (TFs) [29]. Using GSEA, we searched for overrepresented TF promoter binding motifs among differentially expressed genes (Table 1). We defined overrepresented TF binding motifs by a NOM p-value ≤ 0.05. Additional file 17 lists all the dysregulated gene sets with shared TF binding sites for renal development, mitogen-mediated proliferation, cell cycle, epithelial-mesenchymal transition, angiogenesis, and immune/inflammatory response. Several of these enriched transcription factors, such as Gli2 (4.39 fold), Foxo1 (2.51 fold) and Pou2F1 (2.88 fold) were also differentially regulated in our microarray analyses, providing additional evidence for their functional relevance in PKD.

Prediction of miRNA and miRNA-target interactions, and analysis of miRNA expression

Validation of expression by qPCR

Using qPCR assays for the Pkd1 ^-/- and WT samples, we confirmed the microarray results on a subset of genes that were- (i) differentially expressed on microarrays, (ii) known/suggested to be involved in ADPKD, and (iii) targets of miRNAs predicted by two or more tools. These genes were P2rx7, Cpeb3, Hdac9, Sox6, Calcr, Pitx2, Fgfr3, Fgf10, Adam22, Ddx3y, F2rl2, Grap2, Edil3, Mysm1 and Alg6 (Figures 5 and6). Qualitatively, the qPCR validation data agreed with the microarray data (Additional file 20). On the other hand, we found that for certain genes such as Adam22, Grap2 and F2rl2 the fold changes observed on microarrays varied from those obtained by qPCR (Additional file 20). Several factors may be responsible for these variations, among which include the distinct platforms used for microarrays and qPCR, differences in qPCR amplicon/primers and hybridization probes, non-specific and cross-hybridization and issues associated with amplification of mRNA samples [46, 47].

Computational identification of a large number of miRNA-mRNA target interactions suggested that the expression of miRNAs might also change during the progression of PKD. We tested this hypothesis by determining the differential expression of 9 miRNAs (mmu-miR-10a, mmu-miR-30a-5p, mmu-miR-96, mmu-miR-126-5p, mmu-miR-182, mmu-miR-200a, mmu-miR-204, mmu-miR-429, and mmu-miR-488) between WT and Pkd1 ^-/- genotypes at E14.5 and E17.5 (Figures 7 and8). These nine miRNAs are expressed in kidney [48], but have not been previously associated with ADPKD. In our computational analysis, two or more tools predicted them as targets for differentially expressed genes at E14.5 and E17.5. The significance of changes in relative expression of miRNAs among the two groups was tested by Student's t-test, and a cutoff of 1.2 fold [49] and P ≤ 0.05 was used to determine the differential miRNA accumulation.

Among the 9 miRNAs tested at E14.5, only two miRNAs, miR-488 and miR-204, were down-regulated in Pkd1 ^-/- kidneys compared to WT, miR-96 did not change, and the remaining 6 miRNAs were up-regulated in Pkd1 ^-/- kidneys compared to WT (Figure 7). In contrast, at E17.5 miR-96, miR-182 and miR-30a were up-regulated in Pkd1 ^-/- kidneys compared to WT. miR-200a did not show significant variation between the WT and Pkd1 ^-/- kidneys at E17.5 (Figure 8). This indicates that miRNA expression also undergoes reprogramming during PKD. Given that renal cystic- and control- kidneys contained a mixture of cell types, the moderate effects we observed in changes in miRNA levels may in reality be far stronger in specific cell/tissue types.

miRNA-mRNA interactions

We analyzed the functional enrichment of predicted and validated (by qPCR) miRNAs for differentially expressed genes in each comparison in an attempt to uncover the functional meaning among these miRNAs (Figure 9). Genes that were commonly targeted by the differentially expressed miRNAs in Pkd1 ^-/- samples were clustered in 18 biochemical pathways. The overall analysis highlighted that signal transduction pathways such as calcium, VEGF, Notch, and MAPK signaling (Figure 9) may be regulated by miRNA. For example, miR-30a-5p may be involved in histone deactylase inhibitor pathways, apoptosis, calcium and Wnt signaling (Figure 9); miR-10a may be involved in TGF-β and hedgehog signaling; miR-204 may be involved in calcium signaling while miR-488 may be involved in MAPK signaling by targeting Fgfr3 (Figure 9). Thus, we generated a comprehensive atlas of miRNA-target genes and pathways during the progression of PKD. This dataset will serve as a useful resource for future investigations.

Transcriptome changes in PKD

A high level of complexity in the molecular pathobiology was found in Pkd1 ^-/- cystic kidneys. We identified genes that were changed only in Pkd1 ^-/- animals such as Aqp7, Bmp6, Bmp8b, Ddx3y, Kl, Psen1 and Pitx2. This shows changes in expression of developmental and nephron-segment specific genes. Further we identified changes in expression of genes that may be associated with pre-cystic stages leading to cyst formation such as changes in the components of Wnt signaling. We identified changes in gene expression related to cyst growth such as components of MAPK (e.g. Map3k12, Fgf10, Prkcb, Traf2) and TGF-β (e.g. Pitx2) signaling that were up-regulated in the Pkd1 ^-/- animals. The GO term enrichment and GSEA revealed changes in calcium, MAPK, Wnt, TGF-β and JAK/STAT signaling pathways in the Pkd1 ^-/- animals. Additionally, this analysis showed that the class II HDAC, Hdac9 was strongly down-regulated in Pkd1 ^-/- animals. This could effect more global changes in gene expression.

Mitogen-Activated Protein Kinases play central roles in the signal transduction pathways that affect gene expression, cell proliferation, differentiation and apoptosis [50]. Although some studies have suggested that cellular proliferation may be involved in initial cyst formation in PKD, other evidence suggests that abnormal proliferation contributes to cyst growth rather than cyst initiation [51–54]. Our data (Table 3 and Additional file 6), showing a greater increase in MAPK-associated pathways at E17.5 as compared with E14.5, are more consistent with the latter hypothesis that proliferation is more important in cyst growth than in initiation.

Wnt signaling in ADPKD

Previous studies have shown that the activation of canonical Wnt signaling [55] and inhibition of non-canonical Wnt signaling (such as decreased activation of c-jun N-terminal kinase, JNK, and transient changes in intracellular Ca²⁺ concentrations) may play roles in cyst formation in PKD [56]. We found Wnt7b (1.6 fold), Fzd6 (1.3 fold), Dvl3 (1.6 fold) and Lef1 (1.7 fold) were up-regulated while Apc (2.4 fold), an inhibitor of canonical Wnt signaling was down-regulated at E14.5. Although microarray probes failed to detect increased Wnt7a expression, our independent study confirms up-regulation of Wnt7a in Pkd1 ^-/- animals (Qin et al, submitted). Moreover, in a study of Pkd1 ^L3/L3 mouse model Wnt7a and Wnt7b were also found up-regulated in mutant kidneys [28]. The GSEA results also showed enrichment of canonical Wnt signaling for up-regulated genes in Pkd1 ^-/- mutants at E14.5 (Table 3). Canonical Wnt signaling was also up-regulated at E17.5, as determined by both GO analysis and GSEA (Tables 2 and 3, Additional files 12 and 13). Together, these results suggested that there may be an increase in canonical Wnt signaling in Pkd1 ^-/- animals. Additionally, the GO term enrichment analysis showed down-regulation of non-canonical Wnt signaling in Pkd1 ^-/- animals at E14.5 (Table 2, Additional file 12).

In contradistinction to the GO and GSEA analyses suggesting increased Wnt signaling in Pkd1 ^-/- kidneys, we found that two potential inhibitors of Wnt signaling, Cer1 and Frzb (respectively increased 12 fold and 9.7 fold at E14.5), were up-regulated in Pkd1 ^-/- animals at E14.5. Cer1 and Frzb belong to the family of sFRPs (secreted Frizzled Related Proteins) that act as inhibitors of Wnt signaling in animals such as Xenopus and chick by interfering with the binding of Wnt proteins to Frizzled transmembrane receptors [57]. Frzb is a known Wnt antagonist in mammals [57, 58]. However the mouse Cer1 may not inhibit Wnt signaling [59]. At present we are unable to explain the discrepancy between the overall informatic result that canonical Wnt signaling is increased in Pkd1 ^-/- kidneys, and the observed increased expression of Wnt antagonists Cer1 and Frzb. It is possible that Cer1 and/or Frzb have a greater ability to interfere with non-canonical vs. canonical Wnt signaling, and that this will be borne out by future studies.

Calcium signaling in ADPKD

The polycystin-1/2 complex (PC1/PC2) mediates Ca²⁺ entry into the cell in response to mechanical stimulation of the primary cilium [60]. This in turn triggers additional release of Ca²⁺ from intracellular stores. Thus, disruption of the polycystin pathway alters intracellular Ca²⁺ homeostasis [1, 56]. It has been suggested that altered intracellular Ca²⁺ levels increases cAMP levels by stimulating adenylyl cyclases [60]. In turn, increased cAMP paradoxically stimulates MAPK/ERK signaling and promotes renal cystic epithelial proliferation in Pkd1 ^-/- cells [39]. Our results showed that in Pkd1 ^-/- animals at E17.5 an adenylyl cyclase, Adcy7 was up-regulated. Thus, increased gene expression of an adenylyl cyclase may be an additional contributing factor to increased cAMP levels in ADPKD.

Novel regulatory mechanisms in ADPKD

Our study identifies several new nodes of regulation in ADPKD. For example, we found a strong down-regulation of Alg6 (asparagine-linked glycosylation 6 homolog) in the Pkd1 ^-/- genotype at E17.5. ALGs are important components of the glycosylation pathway involving the endoplasmic reticulum (ER) and the Golgi apparatus. Alg6 encodes a member of the glucosyltransferase family that catalyzes the addition of the first glucose residue to the growing lipid-linked oligosaccharide precursor of N-linked glycosylation. Mutations in this gene are associated with congenital disorders of glycosylation type Ic [61]. Indeed, in Pkd1 ^-/- mice, defective glycosylation of protein has been observed, that may have important implications for ADPKD [62, 63]. As Alg6 encodes an ER glucosyltransferase [64], its specific downregulation in the Pkd1 ^-/- genotype indicates metabolic reprogramming in the ER and Golgi complex. Thus, exploring Alg6 function in knock-down studies in kidneys may provide new insights regarding the disease processes underlying ADPKD.

The comparison of the Pkd1 ^-/- mouse model and human ADPKD data sets revealed significant overlaps. Some representative of commonly up-regulated genes include Adcy7 (Calcium signaling), Crebbp (Wnt, Jak-STAT, calcium, TGFβ signaling), Cxcl1 (Jak-STAT signaling), Vegfb (VEGF signaling), Dvl2 (Wnt signaling), Tgif1 and Smurf1 (TGFβ signaling). Representative down-regulated genes that were commonly changed between the two data sets include Umod (loop of Henle marker), Rgs3 (GPCR signaling), Pvalb (distal tubule marker), Pck1 (proximal tubule marker), and Lamc3 (Focal adhesion). Thus, these genes may be used as prognostic markers for ADPKD.

Possible regulatory roles of miRNAs in ADPKD

We examined the possible involvement of miRNAs in ADPKD in the Pkd1 ^-/- mouse model. While several miRNAs can target a single transcript, each miRNA can also have multiple targets [13, 31]. Thus, moderate alterations in miRNA levels can have profound effects on the accumulation of their targets [65, 66]. However, their role in ADPKD, both at the individual as well as the systems levels, is still under investigation, as only three studies have directly demonstrated changes in miRNA expression in ADPKD [18, 22, 23]. We have addressed this problem by systemically estimating the possible miRNAs that may target genes deregulated during disease progression (Additional file 1). We have predicted and verified (by qRT-PCR) reciprocal expression of several miRNAs and their predicted targets in Pkd1 ^-/- animals (Table 6). We observed that miRNAs: miRs-10a, -30a-5p, -96, -126-5p, -182, -200a, -204, -429, and -488; and the miRNA-mRNA interactions such as miR-126-5p-Fgf10, miR-488-Fgfr3, miR-182-Hdac9, miR-204-P2rx7 and miR-96-Sox6 (as shown in Table 6) have not been previously reported in ADPKD.

Signaling pathways are ideal candidates for miRNA-mediated regulation, as their components require finely tuned temporal changes in their expression [65, 67]. miRNAs may affect the responsiveness of cells to signaling molecules such as TGFβ, Wnt, Notch, and epidermal growth factor [67]. Once a miRNA targets an inhibitor of a signaling cascade, it serves as a positive regulator by either amplifying signal strength or duration, or empowering cell responsiveness to otherwise sub-threshold stimuli. For example, miR-126 promotes angiogenesis and vascular integrity [68] by inhibiting the production of natural repressor (SPRED1 and PIK3R2) of VEGF signaling, suggesting that it may serve as an effective target for anti-angiogenic therapies. We found that miR-126-5p may be involved in calcium, EGF, MAPK signaling, and neuroactive ligand-receptor interaction (Figure 9; Additional file 21). Previous studies [13, 65, 67] have shown that a single miRNA can act simultaneously on multiple signaling pathways to coordinate their biological effects and concurrently several miRNAs may regulate single pathway (Figure 9). Moreover, the predicted miRNA:mRNA interactions suggested that the involvement of miRNAs in the signaling pathways through their target mRNAs may lead to the dysregulation of these pathways. For example, the interaction of miR-204:P2rx7 (Figure 10) suggested a possibility that the down-regulation of miR-204 may lead to up-regulation of its target gene, P2rx7 at E17.5 in Pkd1 ^-/- kidneys. P2rx7 is a cell-surface, ligand-gated cation channel and its stimulation leads to alteration in the intracellular Ca²⁺ levels that may further result into dysregulation of calcium signaling. Therefore, by targeting P2rx7, miR-204 may disrupt calcium signaling. Similarly, at E17.5 in Pkd1 ^-/- animals, the up-regulation of Fgfr3 and Fgf10 (components of MAPK signaling) and down-regulation of their target miRNAs- miR-488 and miR-126-5p respectively, may stimulate MAPK signaling and cell proliferation in Pkd1 ^-/- samples.

The functional correlation between the differentially expressed mRNAs and miRNAs as a module in ADPKD revealed a tight post-transcriptional regulatory network at the mRNA level whose alteration might contribute to increased immune response, by either direct miRNA targeting or through secondary proteins. Further progress in the understanding of miRNA activity, the identification of miRNA signatures in different states, and the advancement of miRNA manipulation techniques will be valuable for deciphering the roles of individual miRNAs in PKD. The overall discovery of differentially regulated miRNAs in the different diseases is expected not only to broaden our biological understanding of these diseases, but more importantly, to identify candidate miRNAs as potential targets for future clinical applications.