Restoring elastin with microRNA-29

Introduction

Elastin (Eln) is a highly distensible protein that enables tissues like lung, skin and large arteries to repeatedly expand and return to their original shape.¹ Tropoelastin is encoded by a single copy gene that is highly expressed in utero and then strongly downregulated after birth.^2,3 Only low maintenance levels of tropoelastin mRNA are found in most elastic tissues in adults.^3,4 Repair of severe injuries, such as third degree burns, produces tissue that is often heavily scarred and severely limited in flexibility, range of motion and tissue function.⁵ Similar scarring is seen after prolonged cancer radiation therapy, in sun-damaged skin⁶ and in systemic sclerosis.^7,8 In the vasculature, elastin haploinsufficient phenotypes include narrowing (stenosis) of the ascending aorta in the perinatal period and aneurysm formation later in life. Overdistention injury of arteries after balloon catheter-based therapies activates repair responses in the adventitia that closely resemble skin wound contractures and can drive lumen narrowing in restenosis.⁹ These aberrant conditions have in common the production of a collagen-rich extracellular matrix (ECM) that is deficient in elastic fibers. A major cause of the deficiency in elastic fiber production is the failure to upregulate tropoelastin gene expression in postnatal tissues subject to the injuries described above. An article by Sessa and colleagues in this issue of the ATVB describes a novel approach to increase elastin production that may change clinical outcomes in conditions such as those described above.¹⁰

Elastin and the Elastic Fiber

The elastic fiber is built up by a stepwise addition of tropoelastin monomers on a microfibrillar scaffold (Figure). The core of a microfibril is composed of the glycoprotein fibrillin.¹¹ Fibrillin is deposited in the extracellular space and once there interacts with a group of microfibril-associated proteins including fibulins, emilins, microfiber-associated glycoproteins, and latent TGF-β binding proteins.^11,12 Cross-linking of tropoelastin monomers is carried out by the lysyl oxidases, a family of copper-dependent amine oxidases.¹³ Once assembled and extensively cross-linked, elastic fibers in tissues are very stable with turnover times measured in decades.^3,12 The stability of the elastic fiber may be the reason for the very low levels of tropoelastin gene expression found in most adult tissues.³ In mice, loss of function mutations in elastin produce dramatic developmental phenotypes, particularly in large elastic arteries.^14,15 Elastin null mice (Eln^−/−) die within a few days after birth from overproliferation of vascular smooth muscle cells (SMCs) that occludes the aortic lumen¹⁴. These mice develop elongated, tortuous arteries with stiff walls that are physiologically non-compliant^{12, 14}. By contrast, Eln^+/− mice are viable and their arteries have thinner walls, smaller lumens, and more layers of SMCs between attenuated elastic lamellae than wild type arteries^12,15. Eln^+/− mice develop higher systemic blood pressures leading to increased circumferential wall stress, but because of the increased number of elastic lamellae the tension per lamellar unit is the same as in wild type mice^12,15.

Figure.

Tropoelastin (TE) production is under control of miR-29. TE mRNA contains seed sequences in the 3’-UTR that are recognized by miR-29. Binding of miR-29 to TE mRNA inhibits TE synthesis. miR-29 inhibitors block this step and thereby increase TE protein production. Once made TE protein is secreted into the extracellular matrix of elastic tissues like lung, skin, and artery wall. TE protein then interacts with elements of the elastic fiber scaffold composed of fibrillin, fibulin, emelin, microfiber associated glycoproteins (MGAPs), lysyl oxidases (LOXs), and latent TGFβ binding proteins (LTBPs). The elastic fiber has unique biomechanical properties that enable tissues to stretch and return to their original shapes^1,11,12. Deficiencies in TE production can produce or exacerbate aneurysms, scars associated with severe burns, tissue fibrosis or systemic sclerosis, and loss of tissue function associated with emphysema and chronic obstructive pulmonary disease. Therapeutic approaches using miR-29 inhibitors may increase TE protein production and prove beneficial for these and related conditions.

Human Phenotypes Associated with Elastin Haploinsufficiency

Williams-Beuren Syndrome (WBS) is a disorder caused by a heterozygous microdeletion of 1.5Mbp of DNA from chromosome 7q11.23.¹⁶ Patients with WBS exhibit a characteristic spectrum of cognitive and cardiovascular disorders among which are stenoses of elastic arteries and hypertension.¹⁶ While the chromosome 7 microdeletion involves 26–28 genes, it is the loss of one copy of the elastin gene that is the major contributor to the cardiovascular phenotypes in WBS individuals. The principle features of WBS arteries are an increase in wall thickness, an increase in the number of layers of smooth muscle cells (SMCs) and elastin, and a focal narrowing (stenosis) of elastic arteries particularly of the ascending aorta. Supravalvular aortic stenosis (SVAS) is an autosomal dominant heterozygous disruption limited to the elastin gene only, either by point mutation, translocation or deletion.^16,17 Patients with SVAS do not exhibit the cognitive, endocrine and metabolic phenotypes common to WBS. However, SVAS patients do demonstrate considerable overlap with WBS patients in elastic artery structural defects and excessive proliferation of arterial SMCs.¹⁸ Collectively the vascular defects observed in SVAS and WBS are called elastin arteriopathies and are conditions resulting from elastin haploinsufficiency.

The MicroRNA-29 Family – Expression and Function

MicroRNAs (miRNAs) are a class of ~22-nucleotide noncoding small RNAs that play essential roles in regulating gene expression by posttranscriptional mechanisms.¹⁹ miRNAs have been shown to play important roles in cardiovascular development, physiology and disease.^20,21,22 The microRNA-29 (miR-29) family consists of three members encoded by two genomic bicistronic clusters.²³ miR-29a and miR-29b1 are coexpressed from one cluster and miR-29b2 and miR-29c are coexpressed from the other cluster. All miR-29 family members are widely expressed and expression levels gradually increase with age. miR-29 family members share a conserved seed region and their predicted targets are enriched in extracellular matrix (ECM) proteins and ECM proteases suggesting physiological roles in morphogenesis and tissue remodeling and pathological roles in aortic aneurysms²⁴ and Marfan’s syndrome.²⁵ Specific ECM targets for miR-29 family members include tropoelastin, collagen type I (α1) (COL1A1), collagen type I (α2) (COL1A2), collagen type III (α1) (COL3A1), and fibrillin-1.²³

Restoring Elastin Expression with miR-29 Antagonists

Using human dermal fibroblasts and vascular SMCs, Zhang et al showed that overexpression of miR-29a/b/c mimics repressed, whereas miR-29 inhibitors increased, tropoelastin mRNA and protein levels in vitro.¹⁰ miRNAs are versatile regulators of gene expression and individual miRNAs are known to target multiple gene products.^19,20,22 In the case of miR-29, conserved seed sequences are found not only in tropoelastin but also in COL1A1, COL3A1 and VEGF-A among others. Indeed, Zhang et al found that overexpression of miR-29 decreased the activity of luciferase reporter plasmids containing 3’UTRs from COL1A1, COL3A1 and VEGF-A. Moreover, incubation of human vascular SMCs in vitro for 15h in the presence of miR-29a inhibitor increased the levels of tropoelastin mRNA by 5-fold yet had little or no effect on COL1A1, COL3A1, fibrillin1, fibrillin2, fibrillin4, emilin1, lysyl oxidase (LOX), LOXL1 or LOXL2 gene expression. The specificity of this response is curious given the inhibition of COL1A1 and COL3A1 3’UTR-luciferase reporters, and previous reports of miR29 inhibitor increasing levels of COL1A1 and COL3A1 in the heart.²³

Of particular interest, the authors then extended the study to include pulmonary artery SMCs obtained from an SVAS patient with an insertion in exon 9 of the tropoelastin gene that causes a frameshift mutation resulting in a premature stop codon in exon 10. This mutation does not alter the level of miR-29 expression by these cells but results in low levels of full-length tropoelastin mRNA arising from one intact allele. Transfection of SVAS-SMCs with a miR-29a inhibitor increased expression levels of tropoelastin mRNA (3.5-fold) with corresponding increases in amounts of tropoelastin protein secreted into the culture medium or associated with the cell surface. Similar experiments were carried out on dermal fibroblasts obtained from two patients with WBS with confirmed haplodeficiency for elastin expression. Treatment of WBS cells with miR-29 inhibitor also increased levels of tropoelastin mRNA and protein secreted into conditioned medium.

While these in vitro results are promising, the functional form of elastin is not the secreted monomer but the heavily cross-linked polymer of elastin incorporated onto a microfiber scaffold in the ECM.^11,12 To address the important question of whether increasing tropoelastin expression per se would result in functional increases in elastic fiber content and distensibility of the ECM, the authors turned to bioengineered vessels. Polyglycolic acid (PGA) scaffolds seeded with human vascular SMCs were grown in a bioreactor for 6 weeks with or without miR-29a inhibitor and then examined by biochemical analysis, electron microscopy (EM) and biaxial mechanical stress tests. In the absence of miR-29a inhibitor there was little evidence of cross-linked elastin fibers in the ECM and only low levels of cell-associated elastin protein detectable by western blots. By contrast, after incubation for 6 weeks in the presence of miR-29a inhibitor, islands of cross-linked elastin appeared in the ECM that significantly increased the distensibility of these PGA bioengineered vessels at low pressures.¹⁰

Remaining Questions Going Forward

The results described by Zhang et al¹⁰ raise the optimistic possibility that miR-29 inhibitor therapy may offer hope to individuals living with elastin haploinsufficient phenotypes where loss of distensibility in arteries, lungs, skin and other tissues impacts overall health and quality of life. However, a number of questions will need to be addressed in future studies to better evaluate this exciting possibility.

1. A major concern with moving forward with miR-29 antagonists to raise ELN production would be inadvertent stimulation of tissue fibrosis in individuals with preexisting conditions. Reducing the functional levels of miR-29 would be predicted to derepress expression of various collagens and produce a profibrotic phenotype. Reductions in miR-29 have already been implicated in fibrosis in liver²⁶, lung²⁷, kidney²⁸ and heart²³.

2. In a previous study, miR29 antagonists produced predicted increases in collagen levels but had no effect on ELN expression in the border zone of infarcted myocardium.²³ These results may point to additional, possibly tissue-specific levels of regulation of miR-29 function in vivo.

3. It still remains unclear if increasing the synthesis and secretion of tropoelastin per se will prove to be sufficient for assembling a complex elastic fiber and incorporating that into the ECM in such a way that it will improve overall tissue flexibility, distensibility and function (Figure). The bioengineered vessel studied by Zhang et al¹⁰ is producing an ECM de novo whereas perinatal and adult tissues have already formed a complex ECM. Some measure of protease-dependent ECM turnover may be a prerequisite for miR29-based therapies to be successful.

4. Likewise, WBS and SVAS patients typically present with symptoms after the disease has progressed to the point of clinical significance. Would increasing elastin expression in the setting of established disease be a sufficient stimulus to repair the existing tissue structure and to restore tissue function?

5. Like all miRNAs, miR-29 has multiple mRNA targets that may present concerns for potential side effects (off-target effects). For example, other predicted targets for miR-29 include matrix metalloproteinase-2, platelet-derived growth factor receptor-β, and krüppel-like factor-4 (http://targetscan.org).

6. In settings of chronic inflammation or repair of acute tissue injury where ECM degrading proteases are abundantly expressed, will sufficient amounts of newly produced tropoelastin escape degradation in the ECM to reach adequate levels to restore tissue function?

7. Finally, it will be necessary to deliver sufficient amounts of miR29 inhibitor to the affected tissue to be an effective therapy. For internal organs including lung and elastic arteries that may present a challenge.

Summary

The effectiveness of miR-29 antagonists to significantly increase the production of tropoelastin in cells from individuals with haploinsufficient phenotypes for elastin suggests a potentially promising way forward to improve the clinical picture for these patients. The beauty of miRNAs as biological regulators is they simultaneously target multiple mRNAs and thereby act to integrate single biological inputs into coordinated changes in expression of multiple gene targets to regulate complex cell phenotypes like migration, differentiation or metabolism. Therein lies a set of potential complications that further studies will be required to sort out to fully evaluate the exciting promise of the results reported by Zhang and colleagues.¹⁰