Suppression of Gremlin by RNA interference attenuates glucose-induced fibrogenesis in rat mesangial cells

Background: diabetic nephropathy (DN) is the most common cause of end-stage of renal disease. It is beneficial for us to find effective way to treat the disease. Gremlin is deemed as a key factor in the development of diabetic nephropathy at present. We hypothesized that the pathological changes might be prevented by eliminating Gremlin function in high glucose-induced renal fibrosis. Methods: lentiviral vector targeting Gremlin was applied to inhibit Gremlin expression at a high glucose concentration which simulated diabetic nephropathy in rat mesangial cells. Results: the shRNA vector, designated shGremlin, significantly inhibited Gremlin expression in rat mesangial cells cultured under high glucose conditions. Increase in BMP7 as well as its downstream genes or proteins including phospholated Smad 1/5/8, type IV collagen and fibronectin was observed. Conclusions: our work provides a valuable method to prevent glomerular/renal fibrosis with RNA interference, and also enable development of new therapies that target Gremlin.


Introduction
Diabetic nephropathy (DN) is a serious complication of diabetes and it is also the leading cause of end-stage renal disease [1]. At present, the DN treatments are mainly focus on the control of blood level and blood pressure. It is beneficial for us to find extra strategies to inhibit the development of DN.
At present, gremlin which is one of the bone morphogenetic protein (BMP) antagonists, has deemed as an outstanding regulator of DN [2,3]. Recently, it has been implicated to play great roles in processes such as glomerulosclerosis, tubulointerstitial fibrosis. Several recent reports have proved that gremlin can also modulate BMP activity to affect the kidney morphology during embryonic development, and it will disappear after kidney 3 maturity [4], however, gremlin expression can be induced again by high glucose stimulation. BMP-7 which is considered as a protective factor in DN can increase collagen IV and fibronectin expression. A number of studies show that BMP-7 might play protective roles in cultured renal cells and BMP-7 binding to its receptor can affect Smad1/5/8 phosphorylation and play roles in disease [5].Gremlin is reported to be one of the three BMP-7 antagonists and its increased expression is testified in kidneys of diabetic rats [6].The high expression of Gremlin indicates that it plays great roles in the development of DN [7]. As an important target, gremlin plays an important role in the development of diabetic nephropathy by modulating BMP and Smad1/5/8. Therefore, we hypothesized that the DN pathological changes could be prevented by eliminating Gremlin function in high glucose-induced renal fibrosis. We aim to construct a lentiviral vector targeting Gremlin and examined the effect of Gremlin inhibition in rat mesangial cells (RMCs).

Construction of lentiviral vectors targeting Gremlin
We obtained the RNA silencing sequences targeting Gremlin from the paper listed in reference 8 [8], which was proved to be effective. The sequences of shRNAs against rat Gremlin mRNA (Genbank, accession Gremlin site NM_019282.2) are as follows:  TTA CCA TTG CTT CCG AAG AAG CAA TGG TAA TGG TAG  TC-3';   Antisense strand: 5' -AAA AGA CTA CCA TTA CCA TTG CTT CTT CGG AAG CAA TGG TAA TGG   TAG TC-3'. All the sequences were synthesized and amplified by PCR. The PCR fragments were digested with EcoRI and Xbal, and subcloned into pUC-lentivirus vector, which was performed by Genechem Company of Shanghai, China.

Cell culture and infection
RMCs were purchased from the China Center for Type Culture Collection (CCTCC) and were cultured at 37 °C in an atmosphere of 5% CO 2 in Dulbecco's modified Eagle's medium was also used.

Grouping
After incubation at 37 °C for 24 hr, the cells were examined using fluorescence microscope to ensure ~ 90% transfection efficiency. The RMCs were grouped into 4 groups and cultured for 48hr: 1) 5.   The absorbance was measured at 492 nm using a microplate reader (Model 680, Bio-Rad).
The results were expressed in nanograms per milliliter according to the calibration curve obtained with serial dilutions of a known quantity of sample, and these were then normalized to the β-actin content of the corresponding cells. The procedure was performed in triplicate with each sample.

Statistical analysis
Data are presented as mean ± standard deviation (SD). Statistical analysis was performed by one-way ANOVA with F-test. P value of < 0.05 was considered significant. The data were analyzed with SPSS II for Windows (release 11.0.1J, SPSS Japan Inc.). All experiments were performed at least three times.

Overexpression of Gremlin mRNA
Induction of Gremlin expression was achieved by exposing mesangial cells to 5.5 mM glucose, 10 mM glucose, 20 mM glucose and 30 mM glucose in DMEM media for 24 hr.
Gremlin mRNA expression was determined by qRT-PCR. The expression level of Gremlin increased by about 5.40-fold in 30 mM glucose (Fig. 1).

Overexpression of Gremlin mRNA under exposure to 30mM glucose
Mesangial cells were cultured in DMEM media containing 30mM glucose for 0 hr, 24 hr, 48 hr and 72 hr. Gremlin mRNA expression was determined by qRT-PCR. The expression level 7 of Gremlin increased by about 4.94-fold after 48 hr (Fig. 2).

Efficacy of Infection with lentiviral vector carrying green fluorescein proteinlabled shRNAs
To determine the efficacy of infection with Gremlin and negative control shRNAs (shNC), green fluorescein protein (GFP)-labeled shRNAs were transfected into RMCs. The cells were then evaluated by fluorescence microscopy. Most of the RMCs exhibited strong fluorescence (Fig. 3).

Gremlin shRNA Inhibits Gremlin Expression in RMCs
As seen in Figure 4, Gremlin mRNA expression in the high glucose-treated group was about 3.92-fold greater compared with the normal glucose control group (N). Treatment with Gremlin shRNA significantly inhibited Gremlin expression induced by HG conditions, but the negative control shRNA group was no different than N group. Gremlin shRNA prevented the decrease in phosphorylation of Smad1/5/8 (Fig.5). These results suggest that the protective effects of shRNA-induced inhibition of Gremlin expression on RMCs were, at least partially, mediated by BMP-7.

Transfection with Gremlin shRNA reduces over-expression of FN and Col IV accumulation induced by high glucose
To evaluate the impact of Gremlin inhibition on fibrosis and the underlying molecular mechanisms, cultured RMCs were again infected with Gremlin shRNA or negative control and subjected to stimulation with high glucose. FN and Col IV levels in the culture medium were determined by ELISA (Table 1). Significant accumulation of these fibrotic factors in the culture medium was seen in the HG and HG + shNC groups, while shGremlin infection significantly reduced these changes.

Discussion
The molecular pathogenesis of diabetic nephropathy has not been fully characterized.
Novel molecular targets and mechanisms are still under investigation. Bone morphogenetic proteins active in development include homodimeric members of the TGF-β superfamily of cysteine-knot cytokines [9]. The TGF-β superfamily comprises over twenty BMPs, of which BMP-7 is the most prominent member involved in renal development and disease. In the adult life, BMP-7 is primarily expressed in kidney tubules, as well as glomeruli. Loss of endogenous BMP-7 expression occurs in diabetic rats and is associated with fibrosis [10]. In the streptozotocin-induced diabetes model, BMP-7 was reduced by 50% at 15 weeks and continued to decline further to 10% by 30 weeks [11]. In cultured tubular cells, TGF-β decreased BMP-7 expression, which suggests that a rise in tubular TGF-β levels during the evolution of diabetic nephropathy contributes to the loss of BMP7 and BMP7 type I and II receptors [12]. Morrissey and associates [13] showed that exogenously administered recombinant human (rh) BMP-7 may even resolve, at least partially, glomerular and interstitial fibrosis in experimental diabetic nephropathy. BMP-7 activity in the kidney is also determined by a balance of agonists, such as Kielin/chordinlike protein (KCP) or BMP receptors [14], and antagonists, such as Gremlin, noggin, or uterine sensitization-associated gene-1 (USAG-1) [15] that prevent binding of BMPs with their cognate receptors. Of the three BMP antagonists, only Gremlin increased in kidneys of diabetic rats. We proposed that inhibition of Gremlin may induce therapeutic effects by promoting efficient binding of endogenous BMP-7 with receptors.
Our data demonstrate that expression of Gremlin was closely related with matrix glucose level and with exposure time. Gremlin shRNA significantly suppressed FN, collagen type IV, PAI-1 and MMP-2 accumulation, indicating beneficial effects of Gremlin inhibition in DN.
Based on our data, the expression level of BMP-7 was dramatically reduced by suppression of Gremlin, as well as its downstream signal proteins, phosphorylated Smad 1/5/8.

However, a physical interaction between BMP-7 and Gremlin was demonstrated by
immunoprecipitation, and phosphorylated Smad-5, a marker of BMP-7 activity, was upregulated by Gremlin siRNA transfection [16]. BMPs binding with their receptors prevent Smad signaling, which is revealed by the phosphorylation of Smads. Smad 1, 5 and 8 are receptor-regulated Smads (R-Smads) activated by BMPs [17]. Smad5 was the preferred BMP-7-induced receptor-activated Smad signal in kidney. Loss of BMP-7 signaling activity, as illustrated by decreased Smad 5 protein phosphorylation, was observed in experimental DN [18]. Our results with mesangial cells cultured under high glucose conditions demonstrate a gradual increase in Gremlin protein levels from 0 h to 72 h after HG stimulation and depended on exposure time. Transfection of cells with Gremlin shRNA resulted in significantly increased levels of phosphorylated Smad 1/5/8.
Our data support that Gremlin inhibition in DN leads to the recovery of BMP-7 activity. BMP-7 is involved in ameliorating renal damage due to mesangial proliferation by suppression of mesangial cell mitosis via Smad1/5/8 signaling. BMP-7 is also able to prevent metanephric mesenchymal cells and renal epithelial cells from undergoing apoptosis and then protecting renal function [19]. Evidence suggests that early renal hypertrophy acts as a pacemaker for subsequent irreversible structural changes, such as glomerulosclerosis and tubulointerstitial fibrosis [20]. Maintenance of BMP-7 activity by inhibition of Gremlin expression may result in blockade of extracellular matrix (ECM) accumulation. BMP-7 reduced TGF-β-induced ECM protein accumulation in cultured mesangial cells by maintaining the levels and activity of MMP-2, partially through prevention of TGF-β-dependent upregulation of PAI-1 [20,21]. Our data showed that treatment with Gremlin shRNA resulted in a significant reduction of FN and Col IV level in mesangial cells cultured under HG conditions. Whether Gremlin manifests BMP-7independent effects in the pathogenesis of diabetic nephropathy needs to be determined.
The proliferative activity of mesangial cells is associated with the expression level of Gremlin. It was reported that Gremlin increased DNA synthesis and cell counts and accelerated cell cycle progression of vascular smooth muscle cells (VSMC) via p27(kip1) down-regulation [22]. Gremlin was also overexpressed in various human tumors and widely expressed by cancer-associated stromal cells, and promotes tumor cell proliferation [23], suggesting potential stimulation of proliferation. We supposed that Gremlin might regulate cell growth via a BMP-7-independent pathway. Overexpression of Gremlin in diabetic kidneys suggests its role for the developmental programs in DN. Other developmental genes including FMN1, a gene with a Gremlin transcriptional enhancer should be investigated as well [24]. Knockdown of Gremlin by siRNA plasmid might not affect the expression and function of FMN1. Given the role of FMN1 in gremlin transcriptional regulation, it would be interesting and meaningful to investigate the role of FMN1 in DN.
In summary, in addition to expanding our knowledge of the of diabetic nephropathy pathophysiology, our data should enable development of new therapies that target Gremlin.