Managing Microvascular Complications of Diabetes with MicroRNAs

DNA and proteins have long been viewed as the movers and shakers in genomic studies, with RNA sometimes seen as no more than mere messengers shuttling information between the two. This view, however, dramatically changed when the key roles of microRNAs (miRNAs) in gene expression were gradually disclosed over the last decade.^1,2

miRNAs are short ∼22 nucleotide noncoding RNAs, which constitute a relatively new class of small RNAs that act as post-transcriptional regulators of gene expression. Most miRNAs in animals share a common biogenesis pathway in which miRNAs are transcribed by RNA polymerase II as precursor molecules. These precursors, also known as pri-miRNAs, fold into hairpin structures and are further processed by the endonuclease, Drosha, into pre-miRNAs. The pre-miRNAs are exported from the nucleus to the cytoplasm, where they are cleaved by the endonuclease, Dicer, to yield mature miRNAs. The mature miRNA is then loaded into the RNA induced silencing complex, comprised of the Argonaute family of proteins, where it is able to recognize specific mRNA targets. In general, miRNAs negatively regulate their target mRNAs through Watson-Crick base pairing of nucleotides 2–8 of the miRNA (the seed sequence) with complementary sequences within the target mRNA's open reading frame and 3′ untranslated region.

Dysregulation of a single miRNA can influence an entire signaling network. This is because individual miRNAs have multiple targets and thus can exert robust control over complex biological pathways by targeting multiple interrelated proteins. Indeed, it is becoming increasingly apparent that the aberrant expression of a single miRNA may be causally related to a variety of disease states such as cancer, cardiac diseases, and more recently, kidney diseases. In the kidney, miRNAs play important roles in a variety of pathologic conditions. For example, the consequences of inhibition of miRNAs in the glomerulus was recently examined using conditional deletion of Dicer, the RNAs essential for miRNA biosynthesis, in podocytes by several groups.^3–5 Overall, it was reported that podocyte-specific deletion of Dicer leads to increased proteinuria, podocyte effacement, reduced slit diaphragm protein expression, and ultimately renal failure. As well, a podocyte-specific knockout of Drosha leads to collapsing glomerulopathy,⁶ further establishing the importance of regulated miRNAs in podocytes.

With regard to diabetic nephropathy (DN), Natarajan and colleagues⁷ were the first to report a role for a specific miRNA. Their group reported that miR-192 was upregulated in mesangial cells in vitro and in glomeruli from streptozotocin (STZ)-induced and db/db mouse models of DN. miR-192 has been shown by others to also modulate Smads and fibrogenesis in DN.^8,9 Natarajan and colleagues also convincingly demonstrated that miR-192 targets the E-box repressor Smad-1 interacting protein. More recently, the same group has reported that miR-216a was upregulated by TGF-β in experimental models of DN.¹⁰ Our group has identified miR-93 as a signature miRNA in the diabetic milieu.¹¹ Expression of miR-93 is increased in experimental models of diabetes both in vitro and in vivo. We also identified vascular endothelial growth factor-A (VEGF-A) as a putative target of miR-93 in the kidney. Using transgenic mice containing VEGF-LacZ bicistronic transcripts, inhibition of glomerular miR-93 by peptide-conjugated morpholino oligomers elicits increased expression of VEGF.

In this issue of JASN, Wang et al.¹² provide new insights into the role of the miR-29 family in DN. They report that members of the miR-29 family are downregulated in response to TGF-β stimulation in cultured proximal tubular epithelial cells, podocytes, and mesangial cells. This is accompanied by a concomitant increase in the expression of the validated miR-29 targets, collagens I, III, and IV. The authors report that ectopic overexpression of miR-29 results in increased expression of E-cadherin. Interestingly, a correlative assessment was made between miR-29 expression and treatment of the uninephrectomized STZ-diabetic rats with losartan and fasudil, a Rho kinase inhibitor. The renoprotective effects of fasudil correlate with an increase in levels of miR-29a and miR-29c expression, whereas treatment with losartan correlates with increased miR-29b expression.

The miR-29 family consists of three members (29a, 29b, and 29c) that are encoded by two different loci that give rise to bicistronic precursor miRNAs (miR-29a/b1 and miR-29b2/c). The findings by Wang et al.¹² are consistent with previous reports in which TGF-β signaling has been shown to regulate miR-29 and extracellular matrix proteins in multiple tissues and cell lines.^13–15 However, the findings of this study are different from several other studies, including our own,¹⁶ where we showed that miR-29c expression was increased in the glomeruli of db/db mice, an established model of type 2 diabetes, and in vitro in podocytes and endothelial cells under high glucose conditions. Functionally, we identified Sprouty homolog 1 (Spry1), an inhibitor of Rho kinase, as a novel target of miR-29c. Finally, we found that in vivo knockdown of miR-29c using specific antisense oligonucleotides significantly reduces albuminuria and kidney mesangial matrix accumulation in db/db mice.¹⁶ Wang et al. address these distinctly different patterns of miR-29 expression by calling attention to the different animal models of diabetes used and differences in cell types and stimulation (high glucose versus TGFβ).

There are several aspects of the study by Wang et al. that deserve future consideration. First, the critical experiment needed to establish the functional relevance of low miR-29 levels in vivo in their experimental models was not performed; injecting miR-29 mimics or forced overexpression of miR-29 using a genetic approach in diabetic animals could have unraveled the functional role of miR-29 in the pathogenesis and progression of DN. Second, an interesting finding of this study is that different miR-29 family members display distinct patterns of expression in different experimental settings. For example, miR-29a was not significantly downregulated in human podocytes and in tubular cells after 4 days of TGF-β stimulation, whereas only miR-29a and miR-29c were significantly downregulated in STZ-induced apoE knockout mice. Therefore, it appears as if miR-29 family members have distinctive responses to both the diabetic environment and TGF-β stimulation. The reasons for the selective response of miR-29 family members were not explored. Third, it remains unknown whether the pathogenic effects of miR-29 are only based on their effect on matrix proteins or whether the pathogenic effects of miR-29 stem from their pleiotropic effects and independent or in addition to their effects on extracellular matrix proteins. Finally, a detailed understanding of tissue-specific dysregulation of miR-29 family members in human DN should be investigated because of potential differences in miRNA expression patterns in different species.

One of the principle and primary goals of diabetes management is to delay and/or prevent the development of chronic complications. Despite several prominent biochemical theories on how hyperglycemia contributes to microvascular damage, no conclusive genomic pathway is currently thought to contribute to the development of chronic diabetic microvascular complications. The ability of miRNAs to modulate complex biologic processes by regulating multiple targets represents a new strategy for therapeutic intervention.¹⁷ Indeed, the role of miRNAs in DN and other microvascular complications of diabetes is now at a tipping point, where promising therapeutic strategies modulating miRNAs are being considered. While there are clearly many challenges to the development of miRNA-based therapeutics, the central roles of miRNAs in DN and other microvascular complications of diabetes are rapidly emerging.

DISCLOSURES

None.