Microarray-based approach identifies microRNAs and their target functional patterns in polycystic kidney disease

Background

MicroRNAs (miRNAs) are known to regulate the expression of key genes relevant to cancer and potentially other diseases [1-3]. MiRNAs are short noncoding RNAs of about 22 nt length that have recently been shown to play important roles in mammalian gene expression [1,3-5]. They induce posttranscriptional gene repression by blocking protein translation (by binding to the 3' UTR of their target genes) or by inducing mRNA degradation, and have the potential to play central roles in physiological and pathological conditions. Recently, it has been shown that miRNAs can also increase translation [6,7]. The physiological conditions of a cell seem to affect the recruitment of regulatory proteins, which can alter the effect of a miRNA.

MiRNAs are transcribed as long primary transcripts (pri-miRNAs), some of them being polycistronic, which are processed in the cell nucleus by an enzyme called Drosha, yielding precursor miRNAs (pre-miRNAs) that exhibit a characteristic stem-loop sequence [8]. These are exported into the cytosol where mature miRNAs are generated by the RNAse Dicer, producing the small single-stranded miRNA [8]. Translational inhibition, which seems to be the major mode of action in animals, is performed by a riboprotein complex called RNA-induced silencing complex (RISC) consisting of the miRNA and proteins of the argonaute family [9,10].

MiRNAs are involved in several cellular processes, including cellular differentiation [11,12], organism development [13,14], and apoptosis [15,16]. While all of these are conserved in metazoans, the number of conserved miRNAs between mammals suggests that there are additional functions only found in vertebrates [4], e.g. controlling hematopoietic differentiation [17]. Recent studies provide growing evidence for the involvement of miRNAs in cancer pathomechanisms [18-21]. However, to date nothing is known regarding miRNAs in the context of Polycystic Kidney Diseases.

Cilia and flagella are ancient, evolutionary conserved organelles that project from cell surfaces to perform diverse biological roles, including whole-cell locomotion, movement of fluid, chemo-, mechano-, and photosensation, and sexual reproduction. The concept of ciliopathies has helped in advancing a unifying theory of cystic kidney diseases [22]. This theory states that the products of all genes that are mutated in cystic kidney diseases in humans, mice, or zebrafish are expressed in primary cilia or centrosomes of renal epithelial cells [22].

There are numerous disorders linked to basal body and/or cilia dysfunction, including polycystic kidney disease (PKD), primary ciliary dyskinesia (PCD), nephronophthisis (NPHP1–9) [23], Senior-Loken syndrome, Joubert syndrome, Meckel syndrome, oral-facial-digital syndrome, Alström syndrome and Bardet-Biedl syndrome [24]. These syndromes are typically associated with one or more of the symptoms like cystic kidneys [25], retinal degeneration and retinitis pigmentosa [16], situs inversus [26,27], anosmia [28], respiratory problems [26], infertility [26], hydrocephalus [29], other ailments like obesity, diabetes, liver fibrosis, hypertension, heart malformations, skeletal anomalies (e.g. polydactyly), cognitive impairment and developmental defects such as exencephaly [16,26].

Although the mechanisms of the cyst formation are not clearly understood, they are postulated to involve improper functioning of several pathways including cell proliferation, apoptosis, cell polarity, and fluid secretion [30]. Woo [31] has described apoptotic cells in glomeruli, cyst walls, and in both cystic and non-cystic tubules of the polycystic kidneys. Apoptotic loss of renal tissue may be associated with the progressive deterioration of renal function that occurs in patients with autosomal dominant polycystic kidney disease (ADPKD). There is evidence that genes involved in the regulation of cell proliferation, such as p53, c-fos, cyclin D1, and c-myc may be involved in the control of apoptosis [32]. Veis and colleagues have shown overexpression of c-myc in human ADPKD in association with increased levels of apoptosis and cell proliferation [33]. It is clear that pathogenesis of PKD is very complicated and involves multiple molecular pathways with overlapping, complementary, or opposing effects. There are several signalling pathways that have been implicated in ciliary function [34]. The dysregulation of mitogen-activated protein kinases (MAPKs) in the cyst epithelium of pcy (polycystic kidney disease) mice, carrying a missense mutation in NPHP3 [35], is a downstream consequence of disturbed renal monocilia function [36]. The proteins, implicated in the formation of renal cysts in tuberous sclerosis, have been found at the ciliary base. They form a complex that inhibits mTOR resulting in retarded cyst formation in rats with PKD [37]. Cilia-mediated signalling acts as a switch between canonical and non-canonical Wnt signalling pathways [34].

Rats or mice have been used as common model systems for the study of PKD. The Hannover rat, Han:Sprague-Dawley (SPRD)-cy rat is an accepted model for human PKD [38,39] and has been efficiently used since more than a decade. It is an autosomal dominant model for PKD resulting in cyst formation and slowly progressive chronic renal failure [39]. In the current investigation, we explore the transcriptional changes that occur in PKD and investigate if these changes could be related to miRNAs. We use a microarray-based approach to profile the transcriptional (mRNA) changes as well as changes in the miRNA expression patterns in PKD. We use the Han:SPRD cy/+ rat model [39] for our current investigation. Our results suggest several miRNAs may be involved in regulating the genetic switches in PKD. Furthermore we describe some newly detected miRNAs in the kidney.

Methods

Results

Discussion

In the current investigation, we explore the possible involvement of miRNAs in PKD. In a well characterized rat model system (Han:SPRD cy/+ rat model; [39]), we used a combinatorial approach involving data mining and microarray analysis, to profile the miRNAs involved in PKD. In parallel to the miRNA expression profiling, we also describe the changes in mRNA transcript patterns in PKD using genome-wide Affymetrix arrays. Whereas, the gene expression arrays e.g. Affymetrix arrays have already established themselves for measuring large-scale differential regulation of mRNAs, microarrays have recently been developed to measure miRNA expression changes accurately [43,62]. We used arrays based on LNA technology [43]. The cluster analysis of the control and PKD miRNA-microarrays showed tight clustering PKD samples (Figure 3). Few studies have appeared showing the use of proteomics based approaches for identifying specific miRNA targets [63-66]. Several of these studies also applied mRNA-expression arrays to derive correlations between the protein- and messenger- turndown. Over-expressing miRNAs resulted in repressed targets, both at protein and messenger levels: how much both processes contribute to down-regulation of targets depends on individual miRNA-mRNA pairs [65]. These studies show that mRNA profiling are indeed valuable tools for studying mRNA-miRNA interaction, although additional information on fine-tuning of proteome-expression may be obtained by including proteomics approaches.

Our study, profiling the changes in the mRNA transcripts in PKD, revealed large-scale changes. Although some transcription factors, such as SP1 [67], JUN and c-myc [68] have been implicated to play a role in PKD, the magnitude of transcriptional changes (>900 genes) suggest involvement of other regulators. Moreover, the regulatory basis of changes in the expression of transcription factors involved in PKD remains poorly understood. On the other hand, miRNAs have emerged recently as key regulators of gene expression in many developmental [69,70] and disease events [71,72]. Our miRNA study shows that changes in gene expression during PKD also involve miRNAs.

When we compared the mRNA and miRNA profiles, differentially regulated in PKD, with Argonaute (a comprehensive database on miRNAs; [45,71]), there were few genes reported as miRNA target like tropomyosin 1, alpha (TPM1) as a target of miR-21, the beta polypeptide of tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein (YWHAB), regulatory subunit 9B of protein phosphatase 1 (PPP1R9B), early growth response 3 (EGR3) and dynamin 1-like (DNM1L) as targets of miR-31, plysia ras-related homolog A2 (RHOA) as targets of miR-217, etc. (Table 3a). But none of these genes had previously been associated with PKD. Over-representation analysis showed the enrichment of five pathways namely, GnRH signaling pathway, MAPK signaling pathway, Long term depression, Calcium signaling pathway and Neuroactive ligand receptor interaction. The comparison of differentially regulated genes/miRNAs to Argonaute database indicates that previously known miRNA-target interactions (Table 3a) could play important roles in PKD.

Parallel profiling of the transcripts as well as miRNAs on the same set of samples gives us insight into potential interactions between miRNAs and differentially expressed targets in PKD. This parallel profiling allowed us to scrutinize for the negative expression patterns for the miRNA-target relation at a given time point (Table 4). A negative relationship of the expression patterns between miRNA and its target mRNA is an important parameter for determining their interactions because miRNA, in general, are regarded as negative regulators of their targets.

Microarray analysis of miRNAs revealed down regulation of 29 out of 30 differentially regulated miRNAs, which corresponded to increased expression of several genes related to pathways upregulated during PKD. To further uncover the functional correlation between differentially expressed mRNAs and miRNAs, we determined functionally related gene sets, or pathways. The differentially regulated mRNAs were associated with 24 functional categories, which included several enriched pathways important to renal diseases. Previous studies of cystic kidneys implicated several pathways thought to contribute to the pathogenesis of renal cysts' formation like mTOR signalling, MAPK (Mitogen-Activated Protein Kinase) signalling, Wnt signalling, and the TGF-β (Transforming Growth Factor-β) pathway.

MAPKs play important roles in the cell by transmitting extracellular signals from the cell membrane to the nucleus [73]. MAPKs are activated by various stimuli, influencing cell proliferation, differentiation, and apoptosis. Aberrant regulation of MAPKs and other signalling pathways has been reported to be consistent with altered regulation of cell proliferation and differentiation observed in renal cystic disease [58]. Sustained activation of MAPKs in kidney epithelial cells inhibits normal epithelial phenotype and formation of adherens junctions [74]. The expression of genes involved in MAPK signalling pathway like MAP3K1, JUN, and MYC was significantly higher in the PKD animals, indicating the activation of MAPK signalling pathway.

Wnt signalling is essential for renal development. Recent research has revealed an unexpected intersection between Wnt signalling and PKD [59]. It has been reported that canonical Wnt signalling seemed mandatory for early renal development, but persistent β-catenin signalling seemed to trigger cyst formation at later developmental stages [59]. Components of the Wnt signalling pathway like FZD2, JUN, MYC, and RHOA were significantly upregulated in our PKD animals.

TGF-β promotes renal cell hypertrophy and stimulates extracellular matrix accumulation in several renal diseases, including diabetic nephropathy [75] and PKD [76]. It activates the inhibitors of the proteases e.g. tissue inhibitors of metalloproteinases and plasminogen activator inhibitor 1 [77]. TGF-β 1, TGF-β 2 and TGF-β 3 were consistently upregulated in PKD-affected animals. The expression of extracellular matrix components, such as collagen triple helix repeat containing 1 (CTHRC1), and fibronectin 1 (FN1) was significantly higher in animals with PKD. One of the integrins, integrin beta 1 (ITGB1) was upregulated. Therefore, there was a clear activation of the TGF-β signaling pathway in the PKD animals with a significant increase of the synthesis of the extracellular matrix components and inhibition of the proteases that digest this matrix. Recently Kato et al. showed that miRNA miR-192 has been associated with TGF-β pathway in diabetic kidney glomeruli [78]. Its implication in PKD has not been shown, though.

We could not find negative co-relations between the predicted targets of the only up-regulated miRNA, miR-21, whose expression was verified by qPCR analysis too. All the predicted targets were up-regulated. It should be noted that alternate action mechanisms for miRNAs also exist. The action of miRNAs may not to be reflected on the level of their target mRNAs as they are believed to block or attenuate further translation of mRNAs to protein. In such conditions, miRNAs will exert their regulatory role on the level of translation. Moreover, a new mode of action where the miRNAs act as positive regulators has also been defined recently [7]. Depending upon the state of a cell, a miRNA can act as positive or negative regulator [7].

The specific cellular pathways that were found to be associated with dysregulated miRNAs or differentially expressed transcripts in PKD can be used to shape some initial hypotheses on how alteration of miRNA expression may be directly involved in the disease. Furthermore, the functional correlation between the differentially expressed mRNAs and miRNAs in PKD revealed a tight posttranscriptional regulation network at both mRNA and protein level.

Conclusion

While this work establishes a multi-tiered approach for investigation of miRNA mode of genetic regulations of PKD, the results await further investigation. PKD is an important disease as 8–10% of all patients on renal replacement therapy including haemodialysis or transplantation, suffer from PKD [79]. The molecular components have just started to become apparent, and we add another layer of regulators, namely miRNAs. We predict that several of the differentially regulated genes are miRNA targets and miR-21, miR-31, miR-128, miR-147 and miR-217 may be the important players in such interaction. It is interesting to note that miR-31 and miR-217 have not been previously reported in kidney. miR-21 has been reported in kidney [45] but its function has yet not been established. It has been speculated in previous studies that a single miRNA could target more than hundred genes and one gene could be the target of several miRNAs [61,80]. Knockout and over-expression studies will provide further insight into the regulatory interaction between these miRNAs and their targets in order to properly understand PKD development and design new therapeutic measures.

Authors' contributions

PP carried out the study, analysed and interpreted the data and drafted the manuscript. BB participated in the design of the study and drafting the manuscript. PKS participated in analysis of data. AB helped in animal experiments. SB helped in animal experiments. HJG participated in revising the manuscript critically for important intellectual content. NG designed the study, participated in drafting the manuscript and has given final approval of the version to be published.

Supplementary Material