Cross-talk between miR-29 and Transforming Growth Factor-Betas in Trabecular Meshwork Cells

Abstract

Purpose.: To investigate the interactions between microRNA-29 (miR-29), a negative regulator of extracellular matrix (ECM), and transforming growth factors (TGF)β-1 and TGFβ-2.

Methods.: Changes in expression of the miR-29 family were analyzed by quantitative-PCR (Q-PCR) after treatment with TGFβ1 and TGFβ2 (1 ng/mL). TGFβ1 and TGFβ2 were evaluated at gene expression and protein levels by Q-PCR and ELISA, respectively, in human trabecular meshwork (HTM) cells transfected with miR-29b or scramble control. TGFβ1 promoter activity was analyzed using an adenovirus with the reporter SEAP. The effects of miR-29b and TGFβ2 on ECM gene expression were evaluated in cells transfected with miR-29b or scramble control and treated with TGFβ2, and the expression of ECM genes was analyzed by Q-PCR.

Results.: TGFβ2 but not TGFβ1, downregulated the three members of the miR-29 family. Overexpression of miR-29b antagonized the effects of TGFβ2 on the expression of several ECM components. MiR-29b decreased the expression of TGFβ1 at the promoter, transcript, and protein levels but had only a minor effect on the expression of active TGFβ2. The inhibition of TGFβ1 by miR-29b was partially recovered after co-transfection with a plasmid-expressing bone morphogenetic protein 1.

Conclusions.: Results showed some level of crosstalk between TGFβs and miR-29. Specifically, the downregulation of miR-29 by TGFβ2 contributed to the induction of several ECM components by this cytokine in TM cells. This observation, together with the inhibitory effects of miR-29b on the expression of TGFβ1, suggests that the miR-29 family could play an important role in modulating TGFβs on the outflow pathway.

Glaucoma is second only to cataracts as the greatest blinding disorder worldwide, and it is estimated that it will affect 60.5 million people by 2010. ¹ The main risk factor for primary open-angle glaucoma (POAG) is elevated intraocular pressure (IOP), presumably due to changes in the conventional outflow pathway (trabecular meshwork/Schlemm's canal). Alterations in the composition of the extracellular matrix (ECM) of the trabecular meshwork (TM) are known to be associated with POAG and are believed to play an important role in the abnormal increase in outflow resistance that leads to elevated IOP in this disease. ^2,3  

The transforming growth factor beta (TGFβ) subfamily of cytokines includes three isoforms, TGFβ1, β2, and β3, which regulate a wide range of essential cellular activities, including proliferation, differentiation, and ECM dynamics in many cell types. ^{4 –6} Because of their pivotal role in cell regulation, alteration of TGFβs is a characteristic of many diseases and pathologies, ^{7 –10} including glaucoma. TGFβ2 has been found to be elevated in the aqueous humor of glaucoma patients, ^11,12 and elevation of TGFβ1 has been associated with pseudoexfoliative glaucoma. ¹³ Experiments in vitro and in vivo have shown that TGFβs might play an important role in the pathogenesis of the TM in glaucoma. For instance, TM cells treated with TGFβ2 showed senescence-associated changes ¹⁴ and increased synthesis of several ECM components. ^15,16 In addition, perfusion of human anterior segments with TGFβ2 resulted in increased ECM and IOP, ¹⁷ and overexpression of the active form of TGFβ2 increased IOP and reduced outflow facility in mice and rats. ¹⁸ Similarly, TGFβ1 overexpression changed the morphology of the anterior segment of rat eyes and affected IOP. ¹⁹  

The miR-29 family of microRNAs (miRNAs) is composed of three highly similar ortologs (miR-29a, miR-29b, and miR-29c) that share identical seed sequences. The members of the miR-29 family are known to repress posttranscriptional expression of several mRNAs that encode proteins involved in fibrosis, including multiple ECM components such as collagens, fibrillins, and elastin. ^{20 –23} We have previously demonstrated that miR-29b negatively regulates the expression of genes involved in ECM synthesis and deposition in TM cells and that downregulation of miR-29b, under chronic oxidative stress conditions, contributes to an increase in expression of multiple ECM components. ²³ Some recent reports have suggested that downregulation of members of the miR-29 family by TGFβs might contribute to the fibrogenic effects of these cytokines. TGFβ1 has been implicated in the reduction of the levels of miR-29a observed in fibroblasts from systemic sclerosis patients, which is believed to contribute to increased expression of multiple collagen genes targeted by miR-29a. ²⁴ Similarly, TGFβ1 has been shown to induce a significant downregulation of miR-29a in proximal tubule cells leading to an increase in collagen IV. ²⁵ However, our knowledge about the interactions between miR-29 and TGFβs is still very limited. Therefore, we investigated the potential effects of TGFβ1 and TGFβ2 on the expression of the miR-29 family and evaluated whether alterations in miR-29 expression might contribute to the effects mediated by these cytokines on the expression of ECM genes in human TM cells. In addition, we analyzed whether miR-29 can, in turn, affect the expression of TGFβ1 and TGFβ2. 

Materials and Methods

Cell Culture and TGFβ Treatment

HTM cell cultures were generated from cadaver eyes, with no history of eye disease, within 48 hours post mortem, as previously reported. ²⁶ All procedures involving human tissue were conducted in accordance with the tenets of the Declaration of Helsinski. Cell cultures were maintained at 37°C in 5% CO₂ in media (low-glucose Dulbecco's Modified Eagle Medium with l-glutamine, 110 mg/mL sodium pyruvate, 10% fetal bovine serum, 100 μM non-essential amino acids, 100 units/mL penicillin,100 μg/mL streptomicyn sulfate, and 0.25 μg/mL amphotericin B; all reagents were obtained from Invitrogen, Carlsbad, CA). For TGFβ1 and TGFβ2 (Sigma Aldrich, St. Louis, MO) treatment, the cells were serum starved for 24 hours and treated with 1 ng/mL TGFβ1 or TGFβ2 for 24 hours. 

Transfections

HTM cells were plated 24 hours before transfection and transfected between 50% and 70% of confluence using reagent (lipofectamine 2000; Invitrogen), following the manufacturer's instructions. In brief, for transfection of cells in a 12-well plate, 40 picomoles mirna or mirna plus plasmids (Dharmacon, Chicago, IL) and 1 μL lipofectamine were diluted in 50 μL reduced serum medium (OPtiMem I; Invitrogen) each, incubated for 5 minutes at room temperature (RT), and then lipofectamine and mirna/plasmids were combined and incubated further for 20 minutes at RT and added to the cells in media without antibiotics. Cells were incubated overnight at 37°C in 5% CO₂ and changed to complete media after that. Cells were co-transfected with plasmids expressing a bone morphogenetic protein 1 (BMP1) open reading frame (Origene, Rockville, MD) or green fluorescent protein (GFP; 0.3 μg). The efficiency of the transfection with miRnas or plasmid was confirmed by quantitative-PCR (Q-PCR). 

RNA Isolation and Q-PCR

Total RNA was isolated using one of two extraction methods (RNeasy kit; Qiagen, Valencia, CA; or Trizol; Invitrogen) according to the manufacturers' instructions. RNA yields were measured using fluorescent dye (RiboGreen; Invitrogen). First-strand cDNA was synthesized from total RNA (500 ng) by reverse transcription using reverse transcriptase (oligodT and SuperScript II; Invitrogen) according to the manufacturer's instructions. Q-PCR reactions were performed in 20 μL mixture containing 1 μL of the cDNA preparation (1X iQ SYBR Green Supermix; Bio-Rad, Hercules, CA), using the following PCR parameters: 95°C for 5 minutes followed by 50 cycles of 95°C for 15 seconds, 65°C for 15 seconds, and 72°C for 15 seconds. Here β-actin or GADPH were used as an internal standard of mRNA expression. Primers were designed using online available software (Primer 3′ Input Software, http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi), ²⁷ and the annealing temperatures were determined from the above mentioned software and by Q-PCR using a thermal gradient in a single experiment with the same template for all primers (temperatures ranging from 55°C to 65°C; a CFX96 System and CFX Manager Sofware; Bio-Rad). The absence of nonspecific products was confirmed by both the analysis of the melt curves and by electrophoresis (3% Super AcrylAgarose gels; DNA Technologies, Gaithersburg, MD). The primers used for Q-PCR amplification are shown in Table 1. MicroRNAs were extracted using an miRNA isolation kit (RT² qPCR-Grade; SABiosciences, Frederick, MD). MiRNAs cDNA (25 ng) were amplified using a microRNA reverse transcription kit (TaqMan; Applied Biosystems, Foster City, CA) and specific primers for miR-29a, miR-29b, miR-29c, and U6B as a standard (all from Applied Biosystems). Q-PCR products were amplified following the manufacturer's instructions (TaqMan Universal PCR Master Mix; Applied Biosystems). The fluorescence threshold value (C_t ) was calculated using commercially available system software (iCycle; Bio-Rad). The results were expressed as mean value ± SE in three independent experiments. 

Table 1.

 

Primers Used for Q-PCR Amplification

Gene Symbol	Forward 5′–3′	Reverse 5′–3′
COL1A1	AGCCAGCAGATCGAGAACAT	TCTTGTCCTTGGGGTTCTTG
COL1A2	TGCAAGAACAGCATTGCATAC	GGCAGGCGAGATGGCTTATTTGTT
COL5A1	GGCTGTGCTACCAAGAAAGG	GAGGTCACGAGGTTGCTCT
LAMC1	AATGAAGCCAAGAAGCAGGA	ATGGACAGCAGCAGAGGAGT
SPARC	CCGGGACTTCGAGAAGAACT	CTCATCCAGGGCAATGTACT
TGFβ1	GTCCTCGAGCTCCATGGCGCTCTTCGTG	GTAAAGCTTCAAGCTAATGCTTCATCCT
TGFβ2	AGGGCGGCCGCCTGCAGCGCGAGAGGA	GGATATCTTTAGCTGCATTTGCAAGACT
CTGF	CCTGGTCCAGACCACAGAGT	TGGAGATTTTGGGAGTACGG
BMP1	GTGTGGCCCGATGGGGTCAT	CCCGCAAGGTCGATAGGTGAA
GAPDH	TCGACAGTCAGCCGCATCTTCTTT	ACCAAATCCGTTGACTCCGACCTT
ACTB	CCTCGCCTTTGCCGATCCG	GCCGGAGCCGTTGTCGACG

TGFβs Measurement

TGFβ1 and TGFβ2 were measured (Quantikine Human TGFβ1 and Human TGFβ2, respectively; R&D Systems, Minneapolis, MN) following the manufacturer's instructions. These are “sandwich” enzyme-linked immunoassays that measure activated TGFβ1 and TGFβ2. 

Promoter Activity Assay

The adenovirus AdTGFβ1 containing the TGFβ1 promoter region and the reporter SEAP is described elsewhere. ²⁸ Activation of TGFβ1 promoter after transfection with miR-29b or scramble was quantified by the amount of SEAP released to the culture medium (Great EscAPeTM SEAP chemiluminescence kit 2; Clontech, Mountain View, CA) following the manufacturer's instructions. 

Results

Effects of TGFβ1 and TGFβ2 on the Expression of miR-29

To evaluate the effects of TGFβ1 and TGFβ2 on the expression of the miR-29 family, three independent HTM cell lines were treated with either TGFβ1 or TGFβ2 (1 ng/mL), and the expression of miR-29a, miR-29b, and miR-29c was analyzed by Q-PCR. TGFβ1 did not significantly affect the expression of miR-29b, increased the expression of miR-29a, and showed variable effects on the expression of miR-29c. On the other hand, TGFβ2 significantly and consistently decreased the expression of all three miRNAs in the three cell lines analyzed (Fig. 1). 

Figure 1.

 

Effects of TGFβs on miR-29b expression. Three HTM cell lines were treated with TGFβ1 or TGFβ2 (1 ng/mL), and the expression of miR-29a, -b, and -c was analyzed by Q-PCR. The figures represent the relative expression of the fold change between cells treated with either TGFβ1 or TGFβ2 compared to controls (not treated cells). Bars represent SD from three different experiments. *P ≤ 0.05, **P ≤ 0.01.

Effects of miR-29b on the Induction of ECM Genes by TGFβ2

To analyze the potential relevance of the downregulation of miR-29 mediated by TGFβ2 on the induction of ECM-related genes by this cytokine, HTM cells were transfected with miR-29b or scramble control and split, and half of the cells were treated with TGFβ2 (1 ng/mL) for 24 hours. To ensure that lipofectamine transfection was capable of delivering enough levels of miR-29 mimic, the efficiency of transfection was analyzed by Q-PCR in three HTM cell lines transfected with scramble or miR29. Transfection with miR-29 mimic resulted in an average fold increase of 1917 (P ≤ 0.01) in the presence of miR29 compared to the controls transfected with scramble microRNA. Transfection with miR-29b mimic resulted in significant downregulation of COL1A1, COL1A2, LAMC-1, SPARC, and COL5A1. CTGF was strongly upregulated by TGFβ2, and its upregulation was partially inhibited by miR-29b in only one cell line. Transfection of miR-29b significantly prevented the upregulation of collagens LAMC-1 and SPARC induced by TGFβ2 when compared with cells transfected with scramble and treated with TGFβ2 (Fig. 2). 

Figure 2.

 

MiR-29b antagonized the effects of TGFβ2 on ECM genes. HTM cell lines transfected with miR-29b mimic or scramble control were split, and half of the cells were treated with TGFβ2 (1 ng/mL). The expression of COL1A1, COL1A2, LAMC-1, SPARC, CTGF, and COL5A1 was analyzed by Q-PCR. The figure represents the relative expression of the fold change between cells transfected with scramble control and treated with TGFβ2 and cells transfected with miR-29b and treated TGFβ2, both compared to scramble control without treatment. Bars represent SD from three different experiments. *P ≤ 0.05, **P ≤ 0.01.

Effects of miR-29b on the Expression of TGFβ1 and TGFβ2

To analyze the effects of miR-29 on the expression of TGFβ1 and TGFβ2, three independent HMT cell lines were transfected with either miR-29b mimic or scramble control. Expression of activated TGFβ1 and TGFβ2 proteins were quantified by ELISA, and changes in transcripts expression were analyzed by Q-PCR. Cells transfected with miR-29b showed a significant reduction in the levels of TGFβ1 protein, 40% on average (Fig. 3A), and also downregulated mRNA levels (Fig. 3B) in three HTM primary cell lines. To study the effects of miR-29b on TGFβ1 promoter, we infected three HTM cell lines with adenovirus expressing secreted luciferase under the control of the TGFβ1 promoter. MiR-29b significantly reduced the luciferase expression driven by the TGFβ1 promoter compared to controls (Fig. 3C). Mir-29b decreased TGFβ2 mRNA in all three tested cell lines (Fig. 4A) and showed a significant, but small, decrease in TGFβ2 at the protein level in two out of three cell lines transfected with miR-29b compared to the scramble control (Fig. 4B). 

Figure 3.

 

Effects of miR-29b on TGFβ1. The protein, mRNA, and promoter levels of TGFβ1 were evaluated after transfection with miR-29b in three HTM cell lines. (A) Amount of activated TGFβ1 on the supernatant of cells transfected with miR-29b compared to cells transfected with scramble control. (B) Relative expression of the fold change in TGFβ1 between cells transfected with miR-29b compared to scramble control. (C) Percentage of SEAP activity of cells infected with adenovirus containing the TGFβ1 promoter region and the reporter SEAP (25 pfu) and transfected with miR-29b compared to cells infected with the same virus and transfected with scramble control. Bars represent SD from three different experiments. *P ≤ 0.05, **P ≤ 0.01.

Figure 4.

 

Effects of miR-29b on TGFβ2. The protein and mRNA levels of TGFβ2 were evaluated after transfection with miR-29b in three HTM cell lines. (A) Amount of activated TGFβ2 on the supernatant of cells transfected with miR-29b compared to cells transfected with scramble control. (B) Relative expression of the fold change in TGFβ2 between cells transfected with miR-29b compared to scramble control. Bars represent SD from three different experiments. *P ≤ 0.05, **P ≤ 0.01.

Effects of BMP1 Overexpression on the Inhibition of TGFβ1 by miR-29b

Since we have previously identified BMP1, a known activator of TGFβ1, ²⁹ as a direct target of miR-29b, ²³ we investigated its potential involvement on the inhibition of TGFβ1 mediated by miR-29. Three HTM cell lines co-transfected with (1) miR-29b and BMP1 ORF plasmid, (2) miR-29b and a plasmid-expressing GFP, or (3) scramble control and a plasmid-expressing GFP were analyzed for active TGFβ1 protein expression by ELISA. Expression of BMP1 completely prevented the downregulation of TGFβ1 induced by miR-29 in two cell lines and significantly decreased the level of TGFβ1 downregulation in the third analyzed cell line (Fig. 5). The overexpression of BMP1 was confirmed by Q-PCR in cells transfected with miR29b and BMP1 compared to miR-29b and GFP (HTM-1, -2, and -3 showed 3.58, 2.53, and 1.83 folds, respectively; P ≤ 0.01). 

Figure 5.

 

BMP1 expression partially recovers TGFβ1 protein levels. HTM cells were transfected with miR-29b or scramble and plasmids expressing BMP1 ORF or GFP and the amount of active TGFβ1 protein was analyzed by ELISA. Bars represent SD from three different experiments. *P ≤ 0.05, **P ≤ 0.01.

Discussion

Our results showed that although TGFβ1 did not alter miR-29b and miR-29c and upregulated miR-29a, TGFβ2 significantly downregulated all members of the miR-29 family (29a, -b, and -c) at concentrations closer to those found in aqueous humor. ^11,30 Although there are no reports about the effects of TGFβ2 on miR-29 expression from other cell types, the lack of effects of TGFβ1 on miR-29 expression contrasts with the results reported in human fibroblasts ²⁴ and proximal tubule cells. ²⁵ However, the inhibition of miR-29 by TGFβ1 in these two cell types was observed at a concentration 10 times higher than those used in our experiments. In contrast, TGFβ1 used at a concentration similar to that of our experimental model did not alter the expression of miR-29b in stellate cells. ³¹ These discrepancies could be the result of cell type specific responses or the different concentrations of TGFβ1 used to treat the cells. Since our experiments were conducted with a concentration of TGFβ1 higher than those believed to be present in the aqueous humor, ^13,32,33 our results suggest that, at physiologic concentrations, TGFβ1 might not significantly affect the expression of miR-29. However, the downregulation induced by TGFβ2 in all members of the miR-29 family suggests that changes in expression of these miRNAs might contribute to the upregulation of ECM genes induced by TGFβ2 in TM cells. 

We have previously shown that miR-29b regulates the expression of various genes involved in ECM metabolism in HTM cells. ²³ It is believed that all members of the miR-29 family target a similar set of genes because of their strong sequence similarities and identical seed sequences. The upregulation of ECM genes induced by TGFβ2 in our experiments showed a relatively high degree of variability among the three HTM cell lines analyzed, suggesting that a high level of interindividual variability in the cellular responses to TGFβ2 may exist in human populations. However, in spite of such high levels of variability, overexpression of miR-29b consistently inhibited the upregulation of COL1A1, COL1A2, COL5A1, LAMC1, and SPARC induced by TGFβ2. These results suggest that downregulation of miR-29 and the subsequent derepression of genes regulated by this family may indeed play an important role in the upregulation of ECM genes induced by TGFβ2. 

The induction of ECM components by TGFβ2 in optic nerve astrocytes ³⁴ and trabecular meshwork ³⁵ cell cultures has been reported to be mediated by its downstream mediator CTGF. However, miR-29b had only a minor effect on the upregulation of CTGF induced by TGFβ2 in one cell line and had no significant effect in the other two cell lines analyzed. Therefore, the effects of miR-29b appear to be independent of CTGF and are more likely to be mediated by direct targeting of their 3′UTRs and downregulation of transcription factors such as SP1. ²³  

In addition to the effects of TGFβs on miR-29 expression, we investigated whether miR-29 could, in turn, affect the expression of TGFβ1 and TGFβ2. Although miR-29b had only a minor effect on TGFβ2 protein, our results showed a significant and consistent downregulation of TGFβ1 at protein, transcript, and promoter levels. We have previously reported the targeting of BMP1 by miR-29b in HTM cells, ²³ which is one of the known regulators of TGFβ1 protein activation. BMP1 has been shown to activate latent TGFβ1 by MMP2-dependent cleavage of latent TGFβ binding protein 1 and to form an amplification loop with TGFβ1. ²⁹ Consistent with this role in TGFβ1 activation, overexpression of BMP1, lacking the 3′UTR that contains the binding site for miR-29, partially prevented the decrease in active TGFβ1 induced by miR-29 in HTM cells. These results support a contributing role of BMP1 targeting the regulation of TGFβ1 expression by miR-29, but they do not explain the observed decrease in promoter activity and downregulation of TGFβ1 transcript induced by miR-29b. MiR-29b decreased the activity of the TGFβ1 promoter by mechanisms yet to be defined. There are few examples of miRNAs that regulate gene expression at the promoter level. MiR-520b downregulated MHC class I–related chain A at the 3′UTR and promoter levels, ³⁶ and miR-373 has been shown to target the promoters of E-cadherin and CSDC2 genes. ³⁷ Therefore, we used two available online tools (RegRNA software [http://regrna.mbc.nctu.edu.tw]; and miRBase [http://www.mirbase.org/search.shtml]) ^38,39 to search for putative target sequences in the promoter region of TGFβ1. However, no putative binding sites for the miR-29 family were found (data not shown). Therefore, the downregulation of TGFβ1 promoter may be mediated by alterations of other genes involved in the regulation of the TGFβ1 promoter activity. The combined effects of miR-29b on the inhibition of TGFβ1 activation through BMP1 targeting and the repression of the transcriptional activity of the TGFβ1 promoter suggest that miR-29 regulates TGFβ1 through redundant and potentially additive mechanisms. 

Although our experiments were conducted in vitro and only short-term effects were evaluated, the results indicate that interactions between the TGFβ family of cytokines and the miR-29 family of miRNAs might contribute to the modulation of ECM synthesis in TM cells. Specifically, the downregulation of the miR-29 family by TGFβ2 and the subsequent derepression of genes targeted by this family of miRNAs appear to be an important regulatory event that contributes to the upregulation of several ECM components induced by TGFβ2. This observation, together with the inhibitory effects of miR-29b on the expression of TGFβ1, suggests that the miR-29 family could play an important role in modulating the pathogenic effects of TGFβs on the outflow pathway in glaucoma. 

Footnotes

 Supported by National Institutes of Health Grants NEI EY01894, NEI EY016228, and NEI EY05722, and Research to Prevent Blindness.

Footnotes

 Disclosure: C. Luna, None; G. Li, None; J. Qiu, None; D.L. Epstein, None; P. Gonzalez, None