Abrogation of TGF-β1-induced fibroblast-myofibroblast differentiation by histone deacetylase inhibition

Abstract

Idiopathic pulmonary fibrosis (IPF) is a devastating disease with no known effective pharmacological therapy. The fibroblastic foci of IPF contain activated myofibroblasts that are the major synthesizers of type I collagen. Transforming growth factor (TGF)-β1 promotes differentiation of fibroblasts into myofibroblasts in vitro and in vivo. In the current study, we investigated the molecular link between TGF-β1-mediated myofibroblast differentiation and histone deacetylase (HDAC) activity. Treatment of normal human lung fibroblasts (NHLFs) with the pan-HDAC inhibitor trichostatin A (TSA) inhibited TGF-β1-mediated α-smooth muscle actin (α-SMA) and α₁ type I collagen mRNA induction. TSA also blocked the TGF-β1-driven contractile response in NHLFs. The inhibition of α-SMA expression by TSA was associated with reduced phosphorylation of Akt, and a pharmacological inhibitor of Akt blocked TGF-β1-mediated α-SMA induction in a dose-dependent manner. HDAC4 knockdown was effective in inhibiting TGF-β1-stimulated α-SMA expression as well as the phosphorylation of Akt. Moreover, the inhibitors of protein phosphatase 2A and 1 (PP2A and PP1) rescued the TGF-β1-mediated α-SMA induction from the inhibitory effect of TSA. Together, these data demonstrate that the differentiation of NHLFs to myofibroblasts is HDAC4 dependent and requires phosphorylation of Akt.

idiopathic pulmonary fibrosis (IPF) is a chronic progressive lung disorder of unknown etiology, with a median survival of less than 3 years following diagnosis. The appearance of fibroblastic foci is one pathological hallmark of IPF, and greater profusion of fibroblast foci on histopathology is associated with a worse prognosis (1, 20). The myofibroblast is the key effector for fibrogenesis and is characterized by the expression of α-smooth muscle actin (α-SMA) and increased contractile ability. Transforming growth factor (TGF)-β1 is a potent fibrogenic cytokine and mediates dermal fibroblast transition into myofibroblasts, as well as transition of fibrocytes into myofibroblasts and epithelial-mesenchymal transition (EMT) (18, 19). TGF-β1 activates fibroblasts to express α-SMA by activating the serum response factor (SRF) and SMAD3 and an sp1-like factor (15). Gut-enriched Krüppel-like factor inhibits myofibroblast differentiation by preventing the binding of SMAD3 to the SMAD3-binding element in the α-SMA promoter (17). However, a further regulatory mechanism at the transcriptional level requires clarification. Since overexpression of TGF-β1 in murine lung results in pulmonary fibrosis (28), understanding the mechanisms involved in TGF-β1-mediated fibroblast-myofibroblast transdifferentiation may provide a therapeutic target for pulmonary fibrosis.

The balance of histone acetyltransferase (HAT) and histone deacetylase (HDAC) regulates gene expression and cell functions through modification of core histones or non-histone proteins. The DNA region associated with hyperacetylated histones is usually actively transcribed (2). Deacetylation of histones by HDACs produces compact chromatin that is unfavorable for transcription. Acetylation of non-histone proteins can affect their stability, protein-protein interaction, localization, or DNA binding ability (24). Human HDACs are divided into four classes based on the homology to yeast. HDACs and HDAC4 belong to class II.

HDACs have been shown to be involved in fibrogenesis in various organs. In mice, overexpression of Hop, a nuclear protein associated with HDAC2, results in cardiac interstitial fibrosis, which can be rescued by HDAC inhibitors (21). HDAC4 is required for TGF-β-induced dermal fibroblast-to-myofibroblast transition (12). The inhibition of HDACs by trichostatin A (TSA) abrogated TGF-β-induced collagen1A2 expression in human skin fibroblasts through decreasing the level of transcriptional factor sp1 (11). In this paper, we address whether any specific HDAC is involved in the lung myofibroblast differentiation and what is its downstream target.

The phosphorylation/dephosphorylation of proteins controlled by protein kinases and protein phosphatases (PP) plays a critical role in regulating a variety of cellular processes. PP1 and PP2A belong to serine/threonine phosphatases. The substrate of PP2A includes cellular proteins, viral proteins, and protein kinases. PP2A is capable of dephosphorylating Akt in vitro, and the protein phosphatase inhibitor okadaic acid (OA) has been shown to rescue Akt phosphorylation from PP2A dephosphorylation in 3T3 fibroblasts (3). HDAC1 and HDAC6 associate with PP1, and their knockdown by small interfering RNA (siRNA) results in increased PP1-Akt association (6).

Herein we show that the differentiation of normal human lung fibroblasts (NHLFs) to myofibroblasts is HDAC4 dependent and requires phosphorylation of Akt. The phosphorylation of Akt is necessary for α-SMA expression in response to TGF-β1. Both TSA and HDAC4 knockdown are sufficient to decrease phosphorylation of Akt and block TGF-β1-stimulated α-SMA expression, and the pharmacological inhibition of PP1 and PP2A rescues the α-SMA expression in response to TGF-β1.

MATERIALS AND METHODS

Cell culture, transfection, siRNA experiments, and fluorescent microscopy.

NHLFs were purchased from Lonza (Basel, Switzerland). All NHLFs used in this paper were at passage 6. NHLFs were subcultured in fibroblast growth medium 2 (FGM-2, Lonza). When the cells reached 80% confluence, they were serum starved in fibroblast basal medium 2 (FBM-2, Lonza) with 0.2% BSA overnight. Human recombinant TGF-β1 (R&D, Minneapolis, MN) at 1 ng/ml with and without TSA, at 200 or 500 nM, was used to treat the cells on the next day. For the Akt and protein phosphatase inhibitor assays, NHLFs were treated with the Akt inhibitor 1L6-hydroxymethyl-chiro-inositol-2-(R)-2-O-methyl-3-O-octadecyl-sn-glycerocarbonate (Calbiochem, La Jolla, CA) or the protein phosphatase inhibitors Tautomycin (Calbiochem) or Calyculin A (Sigma, St. Louis, MO) for 1 h before treatment with TGF-β1 and TSA.

HDAC4 siRNA was synthesized by Integrated DNA Technologies (Coralville, IA). HDAC4 siRNA sequences were described previously (12) (HDAC4#1 5′-CGACAGGCCUCGUGUAUGAUU-3′; 5′-UCAUACACGAGGCCUGUCGUU-3′; HDAC4#2 5′-GAAUGUACGACGCCAAAGAUU-3′; 5′-UCUUUGGCGUCGUACAUUCUU-3); negative control siRNA was obtained from Ambion (Austin, TX). The cells were harvested at passage 5 by trypsinization and were spun down for enumeration. The cells were suspended in Nucleofector solution (Amaxa, Gaithersburg, MD) at a concentration of 2 × 10⁶ cells/100 μl. The 100-μl cell suspension was mixed with 4 μl of 50 μM siRNA and moved into an Amaxa-certified cuvette. The program T-16 was used for the transfection. After transfection, cells were cultured in six-well plates in a humidified 37°C 5% CO₂ incubator. Five hours later, the culture medium was changed to FBM-2 with 0.2% BSA, and 12 h later (17 h after the transfection), cells were treated with or without TGF-β1 at a final concentration of 1 ng/ml.

Immunofluorescence.

Cells were fixed with 4% paraformaldehyde in PBS (USB, Cleveland, OH) for 20 min at room temperature in a fume hood for subcellular localization studies. Fixed cells were washed with PBS three times for 5 min and were subjected to block and penetration using PBS with 2% BSA and 0.5% Triton X-100 for 1 h at room temperature. Fixed cells were incubated with monoclonal anti-α-SMA (1:200, Sigma) or anti-phospho-MLC2 (1:200; Cell Signaling, Danvers, MA) in PBS with 2% BSA and 0.5% Triton X-100 for 2 h at room temperature. After three washings using PBS, the cells were incubated with Alexa 488-conjugated goat anti-mouse or Alexa 594-conjugated goat anti-rabbit secondary antibody (1:1,000, Invitrogen, Eugene, OR) for 1 h, which was followed by another three washes with PBS for 5 min. The fixed cells were immersed in drops of Vectashield mounting medium (Vector Laboratories, Burlingame, CA) and covered with a cover glass. Cells were examined with a fluorescent microscope.

Quantitative real-time PCR.

Quantitative real-time PCR (qRT-PCR) analysis was performed using an iCycler device (Bio-Rad Laboratories, Hercules, CA) and SYBR green supermix (Bio-Rad) according to the manufacturer's instructions, along with gene-specific primers. The specific gene's cycle threshold (Ct) values were normalized to the housekeeping gene 36B4 and compared with the control group that was assigned a value of 1 to calculate the relative fold change in expression as previously described (27). The results represented three independent experiments.

Western blot.

Cells were lysed using 1× SDS denaturing buffer and boiled in a water bath for 10 min. An equal quantity of protein for each sample was separated by a SDS-NuPage gel (Invitrogen, Carlsbad, CA). The proteins were then transferred to PVDF membrane (Invitrogen). Membranes were incubated with the following antibodies: α-SMA (Sigma), phosphorate-Akt, total-Akt, α-tubulin, and phospho-myosin light chain 2 (Thr18/Ser19) (Cell Signaling). Blots were rinsed four times with PBST (phosphate-buffered saline, pH 7.4, with 0.05% Tween 20) and incubated with peroxidase-conjugated goat anti-mouse or goat anti-rabbit (Cell Signaling) for 1 h. Blots were then washed four times in PBST and developed using the ECL Plus Western blotting detection system as recommended by the vendor (GE Healthcare, Buckinghamshire, UK). The results are representative of at least two independent experiments.

Type I collagen gel contraction assay.

The contraction assay was done as described previously with minor modification (25). Briefly, 12-well cell culture plates were precoated with 5% BSA/PBS coating solution overnight. On the next day, rat tail type I collagen (BD Biosciences, Bedford, MA) was prepared and mixed with the NHLFs according to the provider's instructions. Briefly, NHLFs in FBM-2 with 0.2% BSA were added to type I collagen at a final concentration of 2 × 10⁵ cells/ml. FBM-2 with 1% BSA and 1 N NaOH was added to the collagen mixture to make a final concentration of 0.2% BSA and 2 mg/ml collagen. The 5% BSA/PBS coating solution was aspirated, and plates were washed twice with PBS. Eight-hundred microliters of the cell-gel mixture was added to each well, and the plates were kept in a 37°C incubator for 30 min before treatment with TGF-β1 or TSA separately or in combination. Gel contraction was assessed as the ratio of the gel surface area to the area of the well. After 48 h, the gel was washed with PBS and treated with trypsin/EDTA for 10 min. The gels were incubated with type I collagenase (5 mg/ml, Invitrogen) at 37°C until the gel was totally digested. The cells were spun down, and an equal number of cells was lysed for Western blotting.

Statistical analysis.

Quantitative data are presented as means ± SD, and statistical analysis was performed using Student's t-test. A P value of <0.05 was considered significant.

RESULTS

TSA abrogates TGF-β1-induced α-SMA and collagen expression.

TGF-β1 is a potent fibrogenic factor, and HDAC inhibitors can modify the gene expression profile. TSA is a highly specific noncompetitive histone deacetylase inhibitor (31) and has no significant toxic effect for NHLFs at a concentration of 500 nM (see Supplementary Fig. 1; supplemental data for this article is available online at the AJP-Lung web site). Here we questioned whether HDAC inhibition could affect the expression of profibrotic genes. We first tested the effect of the pan-HDAC inhibitor TSA on the expression of genes induced by TGF-β1, including α-SMA, h-collagen1, TGF-β1 itself, and connective tissue growth factor (CTGF) using NHLFs. Subconfluent cultures of NHLFs were treated with TGF-β1 alone or with DMSO or TSA simultaneously for 24, 48, and 72 h. The changes in gene expression were analyzed by qRT-PCR at 24 h and Western blot at 48 and 72 h. TSA (200 nM) alone reduced the basal transcription of α-SMA (59%) and h-collagen1 (50%), whereas the transcription of TGF-β1 and CTGF was not altered significantly (+15% and −12.5%, respectively) (Fig. 1). TGF-β1 (1 ng/ml) stimulated the transcription of α-SMA (13-fold), h-collagen1 (10-fold), TGF-β1 (2.43-fold), and CTGF (7.45-fold) at 24 h. When NHLFs were cotreated with TGF-β1 (1 ng/ml) and TSA (500 nM), the transcription of α-SMA (10.5 ± 2.7 vs. 2.4 ± 2.0, means ± SD, n = 3) and h-collagen1 (10.0 ± 2.3 vs. 2.3 ± 1.8) was markedly suppressed (Fig. 1, A and B). In contrast to α-SMA and h-collagen1, the cotreatment of TSA with TGF-β1 stimulated NHLFs to express more TGF-β1 (80% more) and CTGF (106% more) compared with TGF-β1 treatment alone, which indicates that the TGF-β1-stimulated gene expression is not inhibited globally by TSA.

Fig. 1.

Abrogation of transforming growth factor (TGF)-β1-induced α-smooth muscle actin (α-SMA) by histone deacetylase (HDAC) blockade. A–D: subconfluent normal human lung fibroblasts (NHLFs) were cultured in 6-well plates and serum-starved overnight and then treated with TGF-β1 (1 ng/ml) and trichostatin A (TSA) (200 or 500 nM) alone or in combination. The fold change of each transcript in the treated groups over the control group was obtained by setting the values for the control group to 1 after correction for the values of the housekeeping gene 36B4. The relative fold change was presented as means ± SD. *P < 0.05 vs. untreated control. **P < 0.05 vs. TGF-β1+DMSO group. E: α-SMA expression by Western blot. α-Tubulin was used as a loading control. Representative immunoblots from 3 independent experiments are shown. F: the culture condition was similar to A except that the expression of α-SMA was assessed by immunofluorescence (green). The nucleus was counterstained using DAPI (blue). DMSO was the solvent for TSA.

To further test the relationship of HDAC inhibition and fibrogenic gene expression, another global HDAC inhibitor, scriptaid, was used to treat the NHLFs and showed a similar effect on α-SMA, h-collagen1, TGF-β1, and CTGF expression (see Supplementary Fig. 2). The expression of α-SMA at the protein level was measured using Western blot. TGF-β1 treatment resulted in an increase in the expression of α-SMA, which was attenuated with TSA and is consistent with the RT-PCR results (Fig. 1E). These results suggest that HDAC activity is necessary for TGF-β1-mediated α-SMA expression.

To assess the morphological remodeling, NHLFs were treated with TGF-β1 (1 ng/ml) with or without TSA and were then fixed for histological appearance and α-SMA immunofluorescence. TGF-β1 induced a morphological change in NHLFs from a thin and spindle shape to a wide, spread-out morphology that is similar with the morphology of smooth muscle cells cultured in vitro (data not shown). TGF-β1 also promoted α-SMA fiber formation; however, both the myofibroblast morphology and fiber formation failed to occur when the cells were cotreated with TGF-β1 (1 ng/ml) and TSA (500 nM) (Fig. 1F).

Inhibition of HDAC blocks TGF-β1-stimulated contraction.

Myofibroblasts have greater contractile ability than fibroblasts, which may contribute to lung remodeling during lung fibrogenesis. A collagen gel contraction assay was employed to examine whether TSA could decrease the contraction induced by TGF-β1. NHLFs were cultured in a type I collagen gel and treated with TGF-β1 alone or with TSA as mentioned above. After the treatment with TGF-β1 with and without TSA for 48 h, the ratio of the gel surface area to the whole well's area was used to assess contraction. We found that TGF-β1 significantly stimulated collagen gel contraction when compared with untreated control (61% vs. 47%, P = 0.003); however, TSA treatment at the final concentration of 500 nM rescued the gel from contraction mediated by TGF-β1 (control vs. TGF/TSA, 61% vs. 67%) (Fig. 2A).

Fig. 2.

Inhibition of the TGF-β1-induced contractile phenotype by HDAC blockade. A: NHLFs were cultured in a rat tail type 1 collagen gel (2 mg/ml) and treated as shown. The percentage of gel contraction was determined by ratios of gel surface area vs. culture well surface area. The values are presented as means ± SD (n = 3), *P < 0.05 vs. untreated control. ††P < 0.05 vs. TGF-β1+DMSO treatment. B: equal numbers of cells recovered from the collagen gel were lysed using 1× SDS buffer, and an equal amount of protein was loaded to evaluate the expression of α-SMA. C: phosphorylated myosin light chain-2 (p-MLC2) expression was determined using immunoblots after TGF-β1/TSA treatment for 48 h. D: P-MLC2 was visualized by immunofluorescence (red) in NHLFs exposed to TGF-β1 ± TSA. Representative immunoblots from 2 independent experiments are shown.

To establish whether NHLF gel contraction was related with α-SMA expression, NHLFs were recovered from the type I collagen gel after 48 h of treatment. An equal volume of protein from each treatment was used to assess α-SMA expression by Western blot. The results showed that TGF-β1 stimulated α-SMA expression and that TSA blocked the TGF-β1-activated α-SMA expression efficiently in the type I collagen gel (Fig. 2B).

Phosphorylation of myosin light chain 2 (MLC2) is a marker and regulator for cell contraction. Immunostaining and Western blot for the phosphorylation of MLC2 (P-MLC2) was performed to analyze P-MLC2 expression in control and TGF-β1-treated NHLFs with or without TSA (Fig. 2, C and D). TGF-β1 stimulated MLC2 phosphorylation and P-MLC2 fiber formation, whereas cotreatment with TGF-β1 and TSA abolished P-MLC2 fiber formation and decreased the level of P-MLC2 protein expression (Fig. 2, C and D).

Abrogation of fibroblast-to-myofibroblast differentiation is involved in inhibiting the phosphorylation of Akt.

To identify the molecular mechanism by which TSA inhibits fibroblast-to-myofibroblast differentiation, we first evaluated the phosphorylation of SMAD2 and SMAD3, two major TGF-β1 signaling molecules. TGF-β1 activated the phosphorylation of SMAD2 and SMAD3 as expected, but TSA did not block SMAD2 and SMAD3 phosphorylation (Fig. 3A). TSA at 500 nM was sufficient to inhibit the activity of HDACs as indicated by increased acetylation of histone 3 (Fig. 3B). Akt is important for mesenchymal cells to differentiate into smooth muscle cells marked by α-SMA expression (22). We then asked whether the phosphorylation of Akt is necessary for TGF-β1-induced myofibroblast differentiation. TGF-β1 alone increased the phosphorylation of Akt significantly. In contrast, TGF-β1 and TSA cotreatment decreased the phosphorylation of Akt (Fig. 3C).

Fig. 3.

Inhibition of TGF-β1-mediated Akt activation by HDAC blockade. A: after treatment for 30 min, whole cell protein was collected and blotted for phosphorylation of SMAD2 and SMAD3. B: the acetylation of histone 3 (Lys23) was assessed by Western blot after treatment for 12 h. C: the abundance of phosphorylated Akt was determined in NHLFs exposed to TGF-β1 ± TSA for 12 h or pretreated with the Akt inhibitor (Akti) 124005 (phosphatidylinositol analogs) for 1 h and then stimulated with TGF-β1 for 12 h. D: the culture condition was similar to C, and the expression of α-SMA was measured using immunoblots in NHLFs exposed to TGF-β1 with and without 124005. Representative immunoblots from 2 independent experiments are shown.

To further show that Akt is involved in regulating fibroblast-to-myofibroblast differentiation, a phosphatidylinositol ether analog Akt inhibitor (Akti) was used to pretreat NHLFs for 1 h before stimulating the cells with TGF-β1 for 24 h. Western blot results revealed that Akti inhibited the phosphorylation of Akt and α-SMA induction at the protein level in a dose-dependent manner (Fig. 3, C and D). These results suggest that TSA-mediated reduction in P-Akt is enough to inhibit the expression of α-SMA, and the activation of Akt is necessary for TGF-β1-mediated α-SMA expression.

HDAC4 is required for the TGF-β1-induced myofibroblast transformation.

HDAC4 is important for TGF-β1-stimulated α-SMA expression in dermal fibroblasts (12). We tested whether HDAC4 is also important for α-SMA expression in human lung fibroblasts and whether HDAC4 is involved in the regulation of Akt phosphorylation. We used the same siRNA sequences as Glenisson et al. (12) to knock down HDAC4 in our NHLFs following TGF-β1 treatment. The HDAC4 siRNAs knocked down HDAC4 expression at the level of mRNA efficiently (Fig. 4A). The expression of α-SMA at both the protein and mRNA level was also blocked by HDAC4 knockdown with or without TGF-β1 (Fig. 4, B and C). HDAC4 knockdown also blocked TGF-β1-mediated Akt phosphorylation (Fig. 4D), which suggests that Akt phosphorylation is modulated by HDAC4 in the regulation of TGF-β-mediated α-SMA expression (Fig. 4).

Fig. 4.

Requirement of HDAC4 for TGF-β1-mediated α-SMA expression and Akt activation. A: HDAC4 targeting (#1 or #2) siRNA or negative control (NC) siRNA were transiently transfected into NHLFs. The NHLFs were subsequently cultured in FBM-2 with 0.2% BSA in the presence or absence of 1 ng/ml TGF-β1 for 12 h. Real-time PCR amplification was performed to test HDAC4 (A) and α-SMA (B) mRNA levels (n = 3, means ± SD). *P < 0.05 vs. untreated control. **P < 0.05 vs. TGF-β1 alone treated group. C: HDAC4 knockdown and control NHLFs were cultured in the presence or absence of TGF-β1 for 24 h. Western blot analysis was performed using α-SMA and α-tubulin antibodies, and then the membranes were stripped and probed for P-Akt and total-Akt (D). Representative immunoblots from 3 independent experiments are shown.

Pharmacological inhibitor of PP2A and PP1 rescues the expression of α-SMA from HDAC blockade.

Since both PP1 and PP2A can dephosphorylate Akt, we used two potent PP1 and PP2A inhibitors, Calyculin A and Tautomycin (10), to inhibit PP1 and PP2A in NHLFs and examine whether Akt phosphorylation is necessary for induction of SMA in response to TGF-β1.

The Western analysis revealed that the PP1 and PP2A inhibitors Tautomycin (0.05 nM) and Calyculin A (0.5 nM) significantly rescued α-SMA expression (Fig. 5A) from the inhibitory effects of TSA. Moreover, Tautomycin and Calyculin A treatment also rescued Akt phosphorylation (Fig. 5B). Our results suggested that PP1 and PP2A, one or both of them, is/are responsible for the dephosphorylation of Akt and the inhibition of α-SMA expression by TSA.

Fig. 5.

PP1 and PP2A inhibitors rescue α-SMA expression and Akt phosphorylation from the inhibitory effects of TSA. After serum starvation, NHLFs were pretreated with the PP1 and PP2A inhibitors Tautomycin (T) and Calyculin A (C) for 1 h and followed by treatment with TGF-β1 and TSA. Western blot was performed with α-SMA, P-Akt, total-Akt, and α-tubulin antibodies. Relative fold change was calculated by measuring densitometry using ImageJ and normalized to the loading control. All the juxtaposed lanes are from a single gel, but not contiguous, which is showed by the dashed lines. Representative immunoblots from 2 independent experiments are shown.

DISCUSSION

In the present study, we demonstrated a crucial role for HDAC4 in TGF-β1-stimulated lung fibroblast-myofibroblast differentiation via a mechanism regulating the phosphorylation of Akt. Our data revealed that TSA inhibits α-SMA and α₁ type 1 collagen, but not CTGF or TGF-β1, at the transcriptional level. Thus, HDAC inhibition does not appear to block all TGF-β1-mediated cellular events. A correlative functional consequence of using TSA to inhibit TGF-β1-mediated α-SMA synthesis is profound repression of contractility. We also showed that inhibition of PP2A/PP1 could rescue the TGF-β1-mediated induction of α-SMA from HDAC inhibition. Our data suggest that HDAC interferes with protein phosphatase-mediated dephosphorylation of Akt and thereby enhances activation of Akt by TGF-β1.

Myofibroblasts, characterized by increased α-SMA and h-collagen1 expression, are abundant in fibrotic lung suggesting a role in contractility (32). Consistent with this idea, our results showed that TGF-β1 stimulated α-SMA expression and the formation of prominent fibers, which promoted NHLF contractility, as shown by the collagen gel contraction assay (Figs. 1 and 2A). The contraction and α-SMA expression that were stimulated by TGF-β1 were abolished by TSA (Figs. 1E and 2A). We also showed that TSA inhibited expression of the contractile regulator P-MLC2 (Fig. 2, C and D).

Class II HDACs work within both the cytoplasm and the nucleus, and their substrates include both histones and non-histone proteins. HDACs either deacetylate the core histones at the NH₂-terminal tail resulting in a chromatin condensation or work as a part of the transcriptional complex to repress gene expression. An altered balance of HDACs and HATs has been detected in tumors. For example, HDACs fail to disassociate from the transcriptional machinery in response to physiological concentrations of retinoic acid, which leads to acute promyelocyte leukemia. In addition, HDAC2 is overexpressed in human colorectal carcinomas (23, 33). HDACs have emerged as promising targets in drug development for cancer therapy, and several clinical trials have been conducted by using HDAC inhibitors to treat tumors (4, 29). Although we showed that HDAC4 is required for lung fibroblast-myofibroblast differentiation, we cannot exclude the possibility that HDAC4 is cooperating with other HDACs. For example, it has been reported that HDAC4 regulates gene expression by bridging the transcriptional corepressor N-CoR/SMRT-HDAC3 complex, and calcium/calmodulin-dependent protein kinase II-stimulated HDAC5 nuclear exportation is dependent on the HDAC4-HDAC5 heterooligomer formation (5, 9). Here we show that HDAC4 knockdown and the suppression of Akt phosphorylation inhibit lung fibroblast-myofibroblast differentiation, which may provide targets for clinical fibrosis treatment.

The phosphorylation of SMAD2 and SMAD3 is one important pathway to regulate gene transcription in response to TGF-β1. Our results showed the blockade of α-SMA by TSA is not SMAD2 or SMAD3 dependent and is consistent with research in dermal fibroblasts (Fig. 3A) (11).

The α-SMA promoter has several CArG regions recognized by the SRF and one TGF-β control element that is recognized by a non-SMAD unidentified factor, which are involved in enhancing transcription in response to TGF-β (15, 16). HDACs can repress gene transcription by deacetylating histones or be recruited as a part of the transcription complex. TGF-β1 enhances SRF expression in rat fibroblasts (15), and HDAC4 associates with SRF and represses SRF transcriptional potential via its enzymatic activity as well as its ability to antagonize the activity of p300 (7). The phosphorylation of HDAC4 favors its translocation from nucleus to cytoplasm (7). Our results showed that HDAC4 knockdown not only inhibited α-SMA expression but also repressed the phosphorylation of Akt. The PP1 and PP2A inhibitors Tautomycin and Calyculin A rescued TGF-β1-mediated α-SMA expression from the inhibitory effects of TSA. HDAC4 forms a complex with the PP2A holoenzyme to regulate its nuclear entry (26) and thus, one possibility is that HDAC4 and PP2A associate with SRF and are parts of the α-SMA transcriptional complex, and PP2A dephosphorylates HDAC4 to maintain a state of transcriptional inhibition. When HDAC4 is inhibited by TSA or siRNA knockdown, the PP2A may be released and subsequently dephosphorylate Akt; however, further experiments will be required to confirm this hypothesis.

Akt phosphorylation is regulated by kinases and protein phosphatases. TSA can dephosphorylate Akt by releasing PP1 from HDAC-PP1 complexes resulting in an increase in PP1-Akt association (6). HDAC4 is associated with PP2A that regulates HDAC4 nuclear and cytoplasm localization (26). Both PP1 and PP2A have one catalytic subunit and several regulatory units, and the catalytic domains of PP1 and PP2A have 49% identity. Moreover, most of the three-dimensional structure crucial amino acids are conserved between the catalytic subunit of PP1 and PP2A (13). The structural homology of PP1 and PP2A provides both of them with the sensitivity to the structurally variant natural products Tautomycin and Calyculin A. PP1 and PP2A had similar sensitivity to Calyculin A, and Tautomycin was specific for PP1 in the human breast adenocarcinoma cell line (MCF-7) (8); however, PP1 and PP2A showed different sensitivity to Tautomycin and Calyculin A in vitro in another report (14). Due to the high homology between PP1 and PP2A, we did not know the dose that was specific for PP1 or PP2A inhibition in NHLFs. It appears that the sensitivity to Tautomycin or Calyculin A is cell type dependent because all NHLFs were killed when we used the dose reported by Favre et al. (8) and Gupta et al. (14). Although we showed that Tautomycin (0.05 nM) and Calyculin A (0.5 nM) could rescue the α-SMA and phosphorylation of Akt (Fig. 5), further work is needed to specify whether PP1 or PP2A alone, or both of them combined, are involved in the regulatory mechanism of TSA on TGF-β1 activity. Our results showed that TGF-β1 and TSA cotreatment inhibits α-SMA and type I collagen α₁, but not CTGF and TGF-β1 mRNA induction. Consistent with our data, it has been reported that the inhibition of PI-3K/Akt did not affect the expression of CTGF (30). The knockdown or inhibition of HDAC4 represses α-SMA expression, which may involve releasing protein phosphatase 2A/1 from the HDAC complex with subsequent dephosphorylation of Akt.

In summary, we demonstrated that the activation of Akt was indispensable for TGF-β1-mediated lung fibroblast-myofibroblast transition and that either a broad HDAC inhibitor or a specific HDAC4 siRNA can block this Akt-dependent TGF-β1 pathway.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants HL-03901 and HL-083480.