Acetylated α-Tubulin Regulated by N-Acetyl-Seryl-Aspartyl-Lysyl-Proline(Ac-SDKP) Exerts the Anti-fibrotic Effect in Rat Lung Fibrosis Induced by Silica

Silicosis is the most serious occupational disease in China. The objective of this study was to screen various proteins related to mechanisms of the pathogenesis of silicosis underlying the anti-fibrotic effect of N-acetyl-seryl-aspartyl-lysyl-proline (Ac-SDKP) using proteomic profile analysis. We also aimed to explore a potential mechanism of acetylated α-tubulin (α-Ac-Tub) regulation by Ac-SDKP. Two-dimensional electrophoresis (2-DE) and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF/TOF MS) were used to assess the different protein expression profiles between control and silicosis rats treated with or without Ac-SDKP. Twenty-nine proteins were identified to be potentially involved in the progression of silicosis and the anti-fibrotic effect of Ac-SDKP. Our current study finds that 1) the lost expression of Ac-Tub-α may be a new mechanism in rat silicosis; 2) treatment of silicotic rats with N-acetyl-Seryl-Aspartyl-Lysyl-Proline (Ac-SDKP) inhibits myofibroblast differentiation and collagen deposition accompanied by stabilizing the expression of α-Ac-Tub in vivo and in vitro, which is related with deacetylase family member 6 (HDAC6) and α-tubulin acetyl transferase (α-TAT1). Our data suggest that α-Ac-Tub regulation by Ac-SDKP may potentially be a new anti-fibrosis mechanism.

Based on the results of our proteomic profile analysis, we found that α -tubulin was down-regulated in lungs with silicosis-induced fibrosis. This down-regulation was reversed by Ac-SDKP treatment, suggesting that α -tubulin potentially has a role in silicosis development and progression. The acetylation of α -tubulin on lysine 40 (K 40) was one of the earliest tubulin post-translational modifications discovered to regulate microtubule inner proteins and intracellular transport. α -Tubulin is deacetylated by histone deacetylase family member 6 (HDAC6) and is acetylated by α -tubulin acetyl transferase (α -TAT1 or Mec-17) 10 . Acetylated α -tubulin (α -Ac-Tub) has been used as a marker of ciliated cells and exhibits an ability to clear mucus from the airway 11 . In this study, the specific expression and location of α -Ac-Tub in a rat model of silicosis were identified, issues that have rarely been addressed in previous studies. In addition, the microtubule system and actin cytoskeleton have been found to participate in the regulation of myofibroblast differentiation, tissue remodeling, and organ fibrosis 12 . In the present study, we explored 1) the differential protein profile related to rats with silicosis that were treated with Ac-SDKP; 2) the expression dynamics of α -Ac-Tub in the rat silicosis model; and 3) the differential expression of α -Ac-Tub regulated by HDAC6 and α -TAT1 in terms of the anti-silicosis effect of Ac-SDKP, particularly the inhibition of myofibroblast differentiation induced by Ang II.

Results
Comparative proteomic analysis of the Ac-SDKP effect on the silicosis rat model. H&E staining results of lungs obtained from rats are shown in Fig. 1a. In the disease model (4 w and 8 w) group, we observed silicosis nodules and interstitial fibrosis areas, and these pathological changes were alleviated in the Ac-SDKP post-and pre-treatment groups. Six samples from each group were analyzed using 2-DE profiling and MALDI-TOF-MS. In total, 2700 proteins (1400 soluble proteins and 1300 insoluble proteins) were identified in the 2-DE gel using ImageMaster 2D Platinum Software version 6.0 (Supplementary Table S1). As shown in Fig. 1, 2-DE patterns representative of rat lung tissues were displayed and distributed over an area of pI 4.0-8.0, with molecular weights of 14-97 kDa. Compared to the control group (4 w and 8 w), the silicosis model group (4 w and 8 w) and the Ac-SDKP treatment group (post-and pre-treatment) showed that seven soluble proteins (Fig. 1b) and 22  insoluble proteins (Fig. 1c) were differentially expressed. The peptide mass fingerprinting (PMF) of all 29 protein spots were successfully obtained, and their Mascot analysis results and differential expression patterns are listed in Supplementary Tables S1-4. Gene ontology (GO) analysis was used to assess the GO distribution of the identified proteins. In the "molecular function" group, the identified proteins that functioned in binding (proteins, ions, GTP, ATP, enzymes and receptors) were ranked at the top of the category. In the "biological process" category, proteins that participated in cell proliferation and apoptosis, or that responded to stimuli, transport, metabolic and signaling pathways, constituted a high proportion of the identified proteins. The above data suggested that relevant functions were important in silicosis fibrosis and the anti-fibrotic effect of Ac-SDKP (Supplementary Table S5).
The protein expression patterns of lamin A (SB2) and HSP60 (SB3) in 2-DE gels were further validated using a Western blot analysis, and the results were similar to those observed in the 2-DE gels. Compared with the control group (4 w and 8 w), the expression of HSP60 increased while lamin A decreased in the silicosis model group (4 w and 8 w). These protein changes were reversed by Ac-SDKP treatment in both the post-treatment and pre-treatment groups compared with the silicosis 8 w group (Supplementary Figure S2 A). We also detected the expression of SB10 (ADRB-Gs protein complex) by immunofluorescence and Western blot. As showed in Supplementary Figure S1, the expression of Gs was decreased in the silicotic group.

The expression profile of α-Ac-Tub in the silicotic rat and in fibroblasts induced by Ang II.
Furthermore, we validated SA4 (α -Tub), another differentially expressed protein related to the anti-fibrotic effect of Ac-SDKP. Unexpectedly, the verification result was not consistent with the 2-DE result. Two different mechanisms can modulate microtubule diversity to adapt to a large variety of cellular functions. One mechanism is the presence of specific tubulin isotypes, and another is post-translational modifications (PTMs), among which acetylation had been previously identified as important for α -tubulin 10 . As shown in Fig. 2a, the expression of α -Ac-Tub was down-regulated in fibroblasts induced by Ang II for 24 h. Co-expression of α -Ac-Tub and α -Tub revealed by immunofluorescence showed that α -Ac-Tub exhibited specific expression in the alveolar walls of the lung tissue. Unlike the diffuse expression of α -Tub, a specific loss of α -Ac-Tub expression was observed in silicosis nodules.
Positive expression of α -Ac-Tub was previously demonstrated in fibroblasts 13,14 , epithelial cells 15,16 , and endothelial cells 17,18 derived from lung tissue in vitro, and all these cell types have been suggested as potential sources of myofibroblasts 19 . To preliminarily determine which cell type positively expressed α -Ac-Tub, we examined the co-expression of α -Ac-Tub with the mesenchymal marker vimentin or with SP-A, a biomarker of alveolar type II cells, using immunofluorescence staining. As shown in Fig. 2b, positive co-expression of α -Ac-Tub with either SP-A or vimentin was observed in the alveolar walls of lung tissue, suggesting that the positive expression of α -Ac-Tub could be detected in alveolar type II and mesenchymal cells. In silicosis nodules, there was sporadic or weak expression of SP-A or α -Ac-Tub, and there was strong expression of vimentin, which was not accompanied by the co-expression of α -Ac-Tub. To eliminate the potential confounding effects of these antibodies, Ac-SDKP stabilized the expression of α-Ac-Tub in a rat silicotic model. To better illustrate the role of interactions between microtubule acetylation and myofibroblast differentiation, we examined the location and expression of α -Ac-Tub and α -SMA in a rat model using IHC. As shown in Fig. 3a, and similar to our previous study 8 , positive staining for α -SMA was observed in vascular vessels and tracheal smooth muscle cells, but not in the interstitial space, in the control group. Specific expression of α -SMA was observed in silicosis nodules and interstitial fibrotic regions in the silicosis model. In contrast, in control animals that received saline treatment without SiO 2 induction, the majority of detected α -Ac-Tub was expressed in the airway epithelium, similar to previous reports 11 ; α -Ac-Tub also appeared in multiple cell types throughout the alveolar cells of the lung, as previously reported 20 . No positive expression of α -Ac-Tub was observed in silicosis nodules or interstitial fibrotic regions, even in the regions with positive α -SMA expression.
The immunostaining results were further confirmed using Western blot analysis (Fig. 3b), which showed that in the silicosis model (4 w and 8 w) group, silica induction increased the expression of α -SMA, which was accompanied by reduced expression of α -Ac-Tub compared with the control group (4 w and 8 w). Taken together, these results demonstrated that myofibroblast differentiation may result in microtubule disruption.

Ac-SDKP attenuated loss expression of α-Ac-Tub in fibroblasts induce by Ang II. Similar to
the effect of TGF-β 1 8 , upon treatment with Ang II for 24 h, rat lung fibroblasts showed morphological changes, including cell widening and strong positive expression of α -SMA accompanied by decreased expression of α -Ac-Tub (Fig. 4). This finding suggested that the fibroblasts differentiated into a myofibroblast-like phenotype with microtubule disruption. Significantly up-regulated levels of α -SMA and col I protein, as well as down-regulated levels of α -Ac-Tub, were observed in response to Ang II treatment. To specifically examine the anti-fibrotic effect of Ac-SDKP, we synthesized mutant Ac-ADKP in which Ser was replaced by Ala, resulting in a non-functional analog of Ac-SDKP. As shown in Supplementary Figure S2, Ac-SDKP strongly inhibited the expression of col I and α -SMA, and it stabilized the expression of α -Ac-Tub induced by Ang II. Ac-SDKP inhibited the expression of HDAC6 in vitro and in vivo. Previous studies have suggested that microtubule deacetylation may be mainly regulated by HDAC6 10 . In the present study, we also found that the effects of Ang II on myofibroblast differentiation and ECM deposition may involve Ang II receptor type 1 (AT1), although there was no evidence of HDAC6 activation via AT1. To test the potential involvement of these mechanisms in the actions of HDAC6, we used pharmacological inhibitors of the corresponding receptors and enzymes. Treatment with valsartan (an AT1 inhibitor), TCS HDAC6 20b (a specific HDAC6 inhibitor 14,15 ) and Y-27632 (a Rho-associated coiled coil-forming protein kinase (ROCK) inhibitor) caused α -Ac-Tub to be redistributed, and it attenuated the up-regulation of α -SMA and col I induced by Ang II. Treatment with Ac-SDKP showed similar results and inhibited the expression of AT1 and p-MYPT (Fig. 4).
Furthermore, we observed the expression and localization of HDAC6 and HSP90 in rats with silicosis. As shown in Fig. 5a, in the rat lung, silicosis nodules had increased expression of HDAC6 and HSP90. In vivo and in vitro, the expression of HDAC6 and HSP90 was up-regulated in the silicosis group and in fibroblasts induced by Ang II. Upon post-or pre-treatment with Ac-SDKP, HDAC6 expression was decreased in rats. Moreover, treatment with Ac-SDKP, valsartan, TCS HDAC6 20b, and Y-27632 alleviated the elevated expression of HDAC6 induced by Ang II in fibroblasts (Fig. 5b). Fig. 6a, positive staining for α -TAT1 was observed in multiple cell types throughout the alveolar cells of the lung. No positive expression of α -TAT1 was observed in silicosis nodules or interstitial fibrotic regions, even in the regions with positive α -SMA expression. The immunostaining results were further confirmed using Western blot analysis, in which pre-treatment and post-treatment with Ac-SDKP reversed the reduced expression of α -TAT1 induced by silica. In addition, treatment with Ac-SDKP and valsartan reversed the reduced expression of α -TAT1 induced by Ang II in fibroblasts. Figure 6b showed that treatment of fibroblasts with α -TAT1siRNA to knockdown α -TAT1 resulted in a decrease of α -TAT1 and α -Ac-Tub, accompanied by increased expression of α -SMA and Col I. In contrast, over-expression of α -TAT1 resulted in an opposite effect. Furthermore, α -TAT1siRNA knockdown partially blocked the inhibitory effect of Ac-SDKP on Col I and α -SMA in fibroblasts induced by Ang II (Fig. 6c). As further shown in Fig. 6c, over-expression of α -TAT1 inhibited the increased expression of Col I and α -SMA induced by Ang II, and co-incubation with Ac-SDKP prolonged the inhibitory effect of pEX-2. These results suggested that over-expression of α -TAT1 was responsive to Ang II and Ac-SDKP.

Discussion
Our group previously reported that Ac-SDKP has a beneficial effect upon silicosis, which involved the attenuation of TGF-β 1 and Ang II signaling, pulmonary fibroblast proliferation, and collagen synthesis via c-Jun N-terminal kinase (JNK) signaling, as well as the regulation of myofibroblast differentiation by serum response factor (SRF) 8,9 . Other groups have also found that Ac-SDKP can inhibit fibrosis in organs including the heart, kidney, liver, and lung. Moreover, Ac-SDKP exerts multiple functions, such as anti-inflammation, anti-apoptosis, anti-fibrosis, and pro-angiogenesis; thus, it may be a candidate target molecule for novel anti-fibrotic drugs 21 . To enhance our ability in identifying potential biomarkers that are affected by Ac-SDKP in a rat silicosis model, we employed 2-DE/MS technologies to identify new biomarkers of silicosis and target proteins of Ac-SDKP. In the present study, we discovered 29 unique proteins whose expression was significantly altered in the silicosis rat model group or the Ac-SDKP treatment group, and whose function has not been previously explored in the organ fibrosis field, particularly in silicosis disease.
The results of the current study revealed a link between microtubule disruption to the formation and progression of silicosis. In fetal lung fibroblasts (FLFs), adult lung fibroblasts (ALFs), and IPF lung fibroblast (ILF) lines, treatment with TGF-β 1 can induce the up-regulation of Col I, myofibroblast differentiation, and cell proliferation, and it can reduce the expression of α -Ac-Tub 18 . In lung tissue, α -Ac-Tub has been established as a marker of ciliated cells in the tracheal epithelium and was sloughed or diminished in repairing trachea after chlorine exposure 22 and diacetyl instillation 17 . In this study, we found strong positive expression of α -Ac-Tub in airway epithelial cells. However, the positive expression was observed in cells of the alveolar wall. More interestingly, the loss of α -Ac-Tub expression was observed in silicosis nodules, which were dramatically occupied by myofibroblasts marked by α -SMA. In lung fibrosis, myofibroblasts can differentiate from fibroblasts, perivascular cells, fibrocytes, and cells derived from the epithelial-mesenchymal transition (EMT) 19 . Our study showed that α -Ac-Tub is co-expressed with alveolus superficial active substance (SPA, a marker of type II alveolar epithelial cells) and vimentin (the marker of mesenchymal cell). This finding suggests that α -Ac-Tub is widely expressed in alveolar epithelial cells, fibroblasts, and endothelial cells, and this expression is lost when the cells differentiate into myofibroblasts. In addition, microtubule polymerization has been identified to play a role in myofibroblast differentiation by controlling actin stress fiber/megakaryoblastic leukemia-1 (MKL1)/SRF signaling, but not Smad-dependent gene transcription 12 . In a previous study, we found that Ac-SDKP could attenuate the myofibroblast differentiation and collagen deposition induced by TGF-β 1/SRF signaling 8 . In the current study, we found that Ac-SDKP counteracted the pro-fibrotic effects of Ang II by up-regulating α -Ac-Tub. These data provide a new mechanism underlying the anti-silicosis effect of Ac-SDKP and its interactive effect with the renin-angiotensin system (RAS).
Moreover, acetylation of α -Tub is a common post-translational modification, which can be inhibited by histone deacetylase family member 6 (HDAC6) and SIRT2 10 . It was demonstrated that HDAC6-dependent deacetylation of α -Tub in a human lung type II epithelial cell line (A549) exposed to TGF-β 1 induced EMT via Smad signaling, and disparate responses to α -Ac-Tub were observed in primary cultures of human pulmonary arterial endothelial cells and fibroblasts stimulated with TGF-β 1 23 . It has also been reported that ROCK increases the activity of HDAC6, followed by a decrease in α -Tub acetylation, and promotes cell proliferation and migration 24,25 . In the present study, we also observed that induction with Ang II resulted in the up-regulation of α -SMA, Col I, and HDAC6, as well as the down-regulation of α -Ac-Tub. Pre-treatment with Ac-SDKP, valsartan, TCS HDAC6 20b, and Y-27632 blocked myofibroblast differentiation, down-regulated the expression of Col I and HDAC6, and up-regulated α -Ac-Tub in cultured fibroblasts. Furthermore, as a classical substrate of HDAC6, HSP 90 was activated by deacetylation of HDAC6, and its inhibitor had an anti-fibrotic effect [26][27][28] . We also found that the expression of HDAC6 and HSP 90 were primarily observed in silicostic lesions and co-expressed with α -SMA. Furthermore, treatment with Ac-SDKP inhibited the up-regulation of HSP 90 in silicotic rats and in fibroblast induced by Ang II. These results suggest that Ac-SDKP has a potential role on HDAC6 in lung fibrosis through inhibiting the activity of HSP 90 and stabilization of the expression of α -Ac-Tub.
However, HDAC6 is not entirely selective for α -Ac-Tub and has deacetylase-independent functions. In contrast, acetyl transferase α -TAT1 specifically acetylates α -Tub on lysine 40 10 . In cultured α-TAT1 − /− fibroblasts, elevated proliferation was related to deficiency of contact inhibition 29 . In the present study, we also found that that siRNA knockdown of α -TAT1 led to increased expression of α -SMA, while over-expression of α -TAT1 revealed an opposite effect. Meanwhile, α -TAT1siRNA knockdown also inhibited the anti-fibrotic effect of Ac-SDKP on myofibroblast differentiation, collagen synthesis and α -Ac-Tub stabilization. In conclusion, the results of our study indicate that Ac-SDKP regulates HDAC6 and α -TAT1 activity, which leads to a marked increase in α -Ac-Tub expression, blocks myofibroblast differentiation and collagen deposition.

Methods
Animals. Animal care and treatment were performed according to the established guidelines. The protocols were approved by the Animal Care Committee of North China University of Science and Technology (2013-038). Specific pathogen-free male Wistar rats (age 3 w; weight 180 ± 10 g) were purchased from Vital River Laboratory Animal Technology Co. Ltd. (SCXY 2009-0004, Beijing, China). The rats were housed in a temperature-controlled facility with a 12-h light/dark cycle and received food and water according to the guidelines established by the North China University of Science and Technology.
Sample preparation for two-dimensional gel electrophoresis (2-DE). Frozen lung tissues (6 per group) were ground in liquid nitrogen and homogenized in deionized lysis buffer A containing 40 mM Tris, 1% w/v DTT, 1% v/v IPG buffer 3-10NL and cocktail protease inhibitor mix (04693116001, Roche), followed by homogenization. Sample homogenates were incubated on ice for 30 min and centrifuged at 40000 g for 30 min at 14 °C. The supernatants were supplied with 2 M thiourea and 7 M urea and labeled "soluble protein".
The remaining precipitates were washed with lysis buffer A twice and re-suspended in lysis buffer B (2 M thiourea, 7 M urea, 40 mM Tris, 1% w/v DTT, 1% v/v IPG buffer 3-10 NL and cocktail protease inhibitor mixture). Next, the homogenates were centrifuged at 40,000 g for 30 min at 14 °C, and the supernatants were labeled "insoluble protein".

2-DE image analysis, and protein identification (MALDI-TOF-MS).
A 2-DE protocol was followed as previously described 30 . Protein samples (1.3 mg, n = 6 per group) were loaded onto Immobiline dry strips (24 cm, pH 3-10 NL, GE Healthcare). Using IPGphor-3 (GE Healthcare), one-dimensional isoelectric focusing (IEF) was performed following the protocol provided by GE Healthcare. After equilibration in a buffer [6 M urea, 50 mM Tris base pH 8.8, 30% (v/v) glycerol, 2% (w/v) SDS (sodium dodecyl sulfate)] supplemented with 1% (w/v) DTT for 15 min, a second incubation was performed in the same buffer supplemented with 2.5% (w/v) iodoacetamide for an additional 15 min. Strips were then loaded onto 12% SDS polyacrylamide gels in an Ettan DALT II system (GE Healthcare). The proteins in the gel were visualized using Coomassie Blue G-250 (Amresco) staining and were scanned and analyzed using ImageMaster 2D Platinum Software version 6.0 (GE Healthcare). Only spots with at least a 1.5-fold change in vol% were considered differentially expressed.
Proteins with significant changes in expression were excised from the gels, washed, dehydrated, and digested with 10 ng/ml trypsin (Promega) in 50 mmol/L NH 4 HCO 3 . The peptide mixtures were deposited on a stainless steel MALDI probe and were dried slowly at ambient temperature. MS was performed using a Bruker Daltonics Autoflex MALDI-TOF-MS. Mass spectra were detected in the reflection mode and were recorded using FlexControl software (version 2.4, Bruker Daltonics) with the default parameters unless otherwise specified.