Small interfering RNA inhibition of SPARC attenuates the profibrotic effect of transforming growth factor β1 in cultured normal human fibroblasts

Abstract

Objective

SPARC (secreted protein, acidic and rich in cysteine), or osteonectin, is a matricellular protein. Recently, it was observed to be overexpressed in fibroblasts obtained from the skin of patients with scleroderma, as well as in different tissues from patients with several other fibrotic disorders. Moreover, a genetic polymorphism in SPARC has been associated with susceptibility to scleroderma. Transforming growth factor β1 (TGFβ1) is a profibrotic cytokine that stimulates excessive collagen production in patients with scleroderma or other fibrotic diseases. The purpose of this study was to examine whether specific inhibition of SPARC can influence the expression of type I collagen and ameliorate the profibrotic activity of TGFβ1 on normal human fibroblasts.

Methods

Fibroblasts obtained from the skin of 4 healthy individuals were cultured and transfected with SPARC small interfering RNA (siRNA). TGFβ was used as a fibrosis stimulus in cultured fibroblasts. Real-time quantitative reverse transcriptase–polymerase chain reaction and Western blotting were used to measure transcription and protein levels of SPARC and type I collagen, respectively.

Results

The fibroblasts transfected with SPARC siRNA showed decreased expression of both SPARC and type I collagen. Exogenous TGFβ1 induced increased expression of both SPARC and type I collagen in cultured normal human fibroblasts, but this response was significantly blunted in the fibroblasts transfected with SPARC siRNA.

Conclusion

TGFβ1 can induce increased expression of both SPARC and type I collagen. Specific inhibition of SPARC led to decreased expression of type I collagen and attenuated the profibrotic effect of TGFβ1 in cultured normal human fibroblasts. Use of siRNA to silence SPARC represents a potential therapeutic approach to fibrotic disorders such as scleroderma.

Tissue fibrosis occurs in a wide range of disease conditions resulting from excessive deposition of extracellular matrix (ECM). The ECM contains 3 major components: structural proteins and proteoglycans, matricellular proteins, and growth factors (1). Although structural proteins such as collagens are most prominently increased in fibrotic tissues, matricellular proteins and growth factors are believed to be the major players in the maintenance of homeostasis in the ECM (1).

SPARC (secreted protein, acidic and rich in cysteine), which is also referred to as osteonectin or BM-40, is a matricellular protein in the ECM that has multiple biologic functions. It participates in the modulation of cell–matrix interactions, cell adhesion, wound repair, and angiogenesis (1–3). Recently, accumulating evidence has suggested that SPARC may play an important role in fibrosis. Previous studies have shown increased expression of SPARC in many fibrotic disorders, including scleroderma (systemic sclerosis [SSc]), pulmonary fibrosis, renal interstitial fibrosis, hepatic cirrhosis, and atherosclerotic vascular lesions (4–8). Although the physiologic and pathologic significance of these observations is not clear, SPARC has shown the ability to stimulate the transforming growth factor β (TGFβ) signaling system through a TGFβ receptor– and Smad2/3-dependent pathway (9).

In animal models, SPARC-null mice displayed a diminished amount of pulmonary fibrosis compared with control mice after exposure to the profibrotic drug bleomycin (10). SPARC-null mesangial cells displayed reduced expression of both TGFβ1 and type I collagen, while the addition of SPARC to these cells induced early production of TGFβ1 and later augmentation of type I collagen (11). These observations suggest that SPARC regulates collagen expression that may be associated with its influence on TGFβ1 expression in mesangial cells. Type I collagen genes are stimulated by TGFβ1 through their promoters (12, 13), while any direct interactions between type I collagen and SPARC have not been defined.

In studies of SSc, a devastating human fibrotic disease that demonstrates overexpression of SPARC in dermal fibroblasts, our previously reported data suggested that a specific single-nucleotide polymorphism at the 3′-untranslated region of the SPARC gene is associated with quantitative expression of SPARC, as well as genetic susceptibility to this fibrosing disease (8). Therefore, increased amounts of SPARC may be harmful to the local environment in living tissues, presumably through dysregulation of the ECM. All of these factors make SPARC an attractive candidate to be targeted in the treatment of fibrotic conditions.

TGFβ is a multifunctional cytokine that controls proliferation, differentiation, and other functions in many types of cells. TGFβ has been widely accepted as a profibrotic cytokine that not only is found in fibrotic tissues of certain disease states but that also can convert normal tissues into a fibrotic phenotype (14, 15). TGFβ1 mediates the formation of ECM by stimulation of the synthesis of components such as collagen, suppression of matrix metalloproteinases, and induction of tissue inhibitors of these enzymes (16).

Considering the regulatory roles of both SPARC and TGFβ in the ECM, altered expression of one may affect the expression and functions of the other and may ultimately influence the synthesis of structural components of the ECM, such as type I collagen. This mutually regulatory relationship could provide the basis for exploring therapeutic strategies in the treatment of fibrotic diseases. The purpose of the in vitro studies described here was to examine whether specific inhibition of SPARC with small interfering RNA (siRNA) can influence the expression of type I collagen and ameliorate the profibrotic activity of TGFβ1 on normal skin fibroblasts.

MATERIALS AND METHODS

Dermal fibroblast culture.

A 2-mm punch biopsy specimen of skin was obtained from 4 normal individuals. All subjects provided written consent, and the study was approved by the Institutional Review Committee for the Protection of Human Subjects. Four cultured fibroblast strains were established by mincing tissues and placing them into 60-mm culture dishes, which were then maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum (FCS) and an antibiotic–antimycotic cocktail. Early-passage (passages 3–5) fibroblast strains were placed at a density of 1.5 × 10⁵ cells in 25-cm² flasks and grown until 80% confluency for the transfection studies.

Preparation and transfection of SPARC siRNA.

Two 21-base double-stranded RNAs (termed SPARC 262 and SPARC 402) were synthesized and purified by Qiagen (Valencia, CA). The target sequences of these 2 siRNA were 5′-AAAATCCCTGCCAGAACCACC-3′ for SPARC 262 and 5′-AACAAGACCTTCGACTCTTCC-3′for SPARC 402. A fluorescein-labeled nonsilencing control siRNA (catalog no. 1022079; Qiagen) was used for detection of transfection efficiency and silencing control. The culture medium (DMEM) in each culture dish was replaced with Opti-MEM I (Invitrogen, Carlsbad, CA) without FCS. The fibroblasts were transfected with SPARC siRNA and nonsilencing siRNA, using Metafectene (Biontex, Munich, Germany) in a concentration of 16 μg/ml. After 24 hours, the culture medium was replaced with normal DMEM with or without human TGFβ (10 μg/ml) (R&D Systems, Minneapolis, MN). The cells transfected with siRNA were examined at different time points (after 24, 48, and 72 hours of transfection) and used for RNA and protein expression assays. In addition, the cells transfected with nonsilencing siRNA were examined by fluorescence microscopy for estimation of transfection efficiency.

Immunostaining.

Transfected or untransfected fibroblasts were grown in the culture media as described above. After reaching 90% confluency, the cells were washed with phosphate buffered saline (PBS) and fixed with 100% methanol at 4°C for 2 minutes. The cells were washed with PBS again and incubated with a rabbit anti-human SPARC first antibody, followed by incubation with Texas Red–labeled anti-rabbit antibody. DAPI (4′,6-diamidino-2-phenylindole) was used for nuclear counterstaining. The images of fibroblasts with fluorescence-labeled proteins were acquired by fluorescence microscopy, using the Eclipse TE2000-4 microscope (Nikon, Melville, NY).

Quantitative reverse-transcriptase–polymerase chain reaction (RT-PCR).

Quantitative real-time RT-PCR was performed using the ABI Prism 7900 Sequence Detector System (Applied Biosystems, Foster City, CA). The specific primers and probes for each gene were from the Assays-on-Demand product line (Applied Biosystems). As described previously (17), total RNA from each sample was extracted from the cultured fibroblasts described above using TRIzol reagent (Invitrogen) and treated with DNase I (Ambion, Austin, TX). Complementary DNA (cDNA) was synthesized using SuperScript II reverse transcriptase (Invitrogen). Synthesized cDNA were mixed with primers/probes in 2× TaqMan universal PCR buffer and then assayed on an ABI Prism 7900 sequence detector. The data obtained from the assays were analyzed with SDS 2.1 software (Applied Biosystems). The amount of total RNA in each sample was normalized against 18S recombinant RNA transcript levels.

Western blot analysis.

The cellular lysates extracted from cultured fibroblasts were used for protein assays. The protein concentration was determined by a spectrophotometer. Equal amounts of protein from each sample were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Resolved proteins were transferred onto nitrocellulose membranes and incubated with a 1:1,000 dilution of primary antibody, mouse anti-human SPARC (Biodesign International, Kennebunk, ME), or rabbit anti-human type I collagen (LF-41) (18). Mouse anti-human annexin V (Alexis Biochemicals, San Diego, CA) was used as an internal control. The secondary antibody was a peroxidase-conjugated anti-mouse or anti-rabbit IgG. Specific proteins were detected by chemiluminescence using an ECL system (Amersham, Piscataway, NJ). The intensity of the bands was quantified using ImageQuant software (Molecular Dynamics, Sunnyvale, CA).

RESULTS

Transfection and SPARC inhibition.

The fibroblasts transfected with fluorescein-labeled nonsilencing siRNA showed 60–85% transfection efficiency by direct cell counting under fluorescence microscopy (Figure 1A). Immunostaining for SPARC protein showed clear inhibition in SPARC siRNA–transfected fibroblasts compared with either untransfected control or nonsilencing siRNA–transfected fibroblasts (Figure 1B). After 24-, 48-, and 72-hour transfections, total RNAs and proteins were extracted from the fibroblasts. Quantitative RT-PCR analysis showed that the fibroblasts transfected with SPARC 402 demonstrated consistent inhibition of SPARC gene expression after 72 hours, while at the 24-hour and 48-hour time points, clear inhibition of SPARC was not observed. At the 72-hour time point, fibroblasts transfected with SPARC 402 showed a 48% reduction in SPARC transcription levels compared with fibroblasts transfected with nonsilencing siRNA–transfected fibroblasts (P = 0.003 by paired t-test), while the fibroblasts transfected with SPARC 262 and nonsilencing siRNA showed 10% and 3% reductions, respectively, compared with fibroblasts in normal culture media (P > 0.05 by paired t-test) (Figure 2A). Therefore, SPARC 402 was used for the transfections in other assays.

Figure 1. A, Transfection efficiency of cultured human fibroblasts transfected with fluorescence-labeled nonsilencing small interfering RNA (siRNA). B, Immunofluorescence staining for SPARC (secreted protein, acidic and rich in cysteine) in cultured fibroblasts transfected with SPARC siRNA or nonsilencing siRNA, and in untransfected control fibroblasts.

Figure 2. Quantitative reverse transcriptase–polymerase chain reaction analysis. A, Comparison of transcription levels of SPARC (secreted protein, acidic and rich in cysteine) in cultured fibroblasts with different transfections. B, Comparison of transcription levels of SPARC and COL1A2 in fibroblasts under different conditions. Bars show the mean and SD. ∗ = P = 0.003 versus fibroblasts transfected with nonsilencing small interfering RNA (siRNA). ∗∗ = P = 0.043 versus fibroblasts transfected with nonsilencing siRNA plus transforming growth factor β1 (TGFβ1); † = P = 0.05 versus fibroblasts transfected with nonsilencing siRNA; †† = P = 0.005 versus fibroblasts transfected with nonsilencing siRNA plus TGFβ, by Student's t-test.

Type I collagen expression and TGFβ stimulation.

As measured by real-time RT-PCR, COL1A2 and SPARC showed parallel expression in SPARC 402–transfected fibroblasts. In untransfected and nonsilencing siRNA–transfected fibroblasts stimulated with TGFβ1, an average 2.83-fold and 2-fold increase in SPARC, respectively, was observed, and a 1.71-fold increase in COL1A2 RNA transcript was observed in both (Figure 2B). In SPARC 402 siRNA–transfected fibroblasts, expression of both SPARC and COL1A2 was more resistant to the effects of TGFβ1 (Figure 2B). Western blots also showed that expression of both SPARC and type I collagen proteins was decreased in SPARC 402–transfected fibroblasts, as well as in TGFβ1-stimulated SPARC 402–transfected fibroblasts (Figures 3A and B).

Figure 3. A, Western blot analysis of type I collagen and SPARC (secreted protein, acidic and rich in cysteine) in fibroblasts under different conditions. Lane 1, Control fibroblasts; lane 2, control fibroblasts in response to transforming growth factor β1 (TGFβ1); lane 3, SPARC small interfering RNA (siRNA)–transfected fibroblasts; lane 4, SPARC siRNA–transfected fibroblasts in response to TGFβ1; lane 5, nonsilencing siRNA–transfected fibroblasts; lane 6, nonsilencing siRNA–transfected fibroblasts in response to TGFβ1. Annexin V was used as an internal control. B, Densitometric analysis of Western blots. Values are the mean and SD of 4 assays. ∗ = P = 0.012 versus fibroblasts transfected with nonsilencing siRNA; ∗∗ = P = 0.013 versus fibroblasts transfected with nonsilencing siRNA plus TGFβ1; † = P = 0.005 versus fibroblasts transfected with nonsilencing siRNA; †† = P = 0.003 versus fibroblasts transfected with nonsilencing siRNA plus TGFβ1, by Student's t-test.

DISCUSSION

In fibrotic diseases such as SSc, fibroblasts are activated to produce excessive amounts of ECM components such as collagen (19). Numerous studies have shown that ECM biosynthesis and deposition are regulated by matricellular proteins and growth factors and by alterations in cell–ECM interactions that are accompanied by reorganization of the cytoskeletal network (20). SPARC and TGFβ represent a matricellular protein and a growth factor, respectively, that are 2 major regulators of the ECM.

In studies aimed at finding ways to ameliorate the fibrotic process, many profibrotic stimuli, such as TGFβ and other cytokines, have been intensely investigated. Few studies, however, have investigated how matricellular proteins, another group of important ECM regulators, influence the effects of these cytokines, modulate cellular functions, and ultimately prevent fibrosis. SPARC is a matricellular glycoprotein with multiple cellular and extracellular functions. SPARC and TGFβ have been shown to maintain homeostasis of the ECM in a cooperative and mutually regulatory manner. Results of recent studies suggest that SPARC may regulate the TGFβ signaling system through a TGFβ receptor– and Smad2/3-dependent pathway (9).

Currently, application of double-stranded siRNA to induce RNA silencing in cells has been widely accepted in many studies of gene functions (21). Gene-specific siRNA has been demonstrated to be able to facilitate the degradation of homologous RNA, resulting in corresponding gene silencing (22). Using siRNA to disrupt clinically important genes, such as p53, Ras (V12), CD4, and CD25, has exemplified potential therapeutic applications of siRNA (21).

In this study, we designed a SPARC sequence–specific siRNA that showed inhibition of SPARC gene expression in normal cultured human fibroblasts. Interestingly, SPARC “knock-down” fibroblasts with transient transfection of siRNA also showed reduction of type I collagen expression. This indicates that an important relationship between SPARC and type I collagen may be present in the cells and also supports the previous observation in mouse mesangial cells that SPARC-null mice displayed diminished expression of type I collagen (11). The finding that the addition of exogenous TGFβ1 to normal fibroblast cultures induced increased expression of both SPARC and type I collagen indicates that a fibrogenic stimulus impairs cellular and/or ECM homeostasis involving both matricellular and structural proteins. In contrast to the normal fibroblasts, the SPARC “knock-down” fibroblasts showed a diminished response of both SPARC and type I collagen to TGFβ1 stimulation. These observations indicate that controlling SPARC expression in cultured fibroblasts may influence the expression of collagen and attenuate the fibrotic activity of TGFβ1. This finding is of potential interest, especially given the knowledge that TGFβ is a major player in many forms of tissue fibrosis. Regulation of SPARC expression in tissue may optimize cellular and/or extracellular conditions to prevent the stimulation of fibrosis. Therefore, it will be interesting to pursue the studies of SPARC siRNA in certain fibrotic tissues, such as skin from patients with scleroderma, in the future.

Although this study represents SPARC inhibition in cultured fibroblasts only, and the precise mechanisms involved in the process need to be further explored, it represents a first attempt to use SPARC siRNA in the regulation of a major ECM component in culture conditions. Application of SPARC siRNA in controlling cellular and/or extracellular response to fibrotic stimulus may provide insight into therapeutic strategies for treating fibrotic diseases.