Proteasome activator PA200 regulates myofibroblast differentiation

The proteasome is essential for the selective degradation of most cellular proteins and is fine-tuned according to cellular needs. Proteasome activators serve as building blocks to adjust protein turnover in cell growth and differentiation. Understanding the cellular function of proteasome activation in more detail offers a new strategy for therapeutic targeting of proteasomal protein breakdown in disease. The role of the proteasome activator PA200 in cell function and its regulation in disease is unknown. In this study, we investigated the function of PA200 in myofibroblast differentiation and fibrotic tissue remodeling. PA200 was upregulated in hyperplastic basal cells and myofibroblasts of fibrotic lungs from patients with idiopathic pulmonary fibrosis. Increased expression of PA200 and enhanced formation of PA200-proteasome complexes was also evident in experimental fibrosis of the lung and kidney in vivo and in activated primary human myofibroblasts of the lung in vitro. Transient silencing and overexpression revealed that PA200 functions as a negative regulator of myofibroblast differentiation of human but not mouse cells. Our data thus suggest an unexpected and important role for PA200 in adjusting myofibroblast activation in response to pro-fibrotic stimuli, which fails in idiopathic pulmonary fibrosis.

not fully understood yet 7 . Proteasome activators are proposed to function as building blocks that are recruited to the 20S core particle to allow adaptive regulation of proteasome function according to cellular needs 7,[9][10][11] .
The proteasome activator PA200 is a 200 kDa, monomeric protein expressed in both the cytosolic and nuclear compartment. It is composed of HEAT repeats forming a dome-like cap that binds to the 20S core particle but also attaches to 26S proteasome complexes forming so-called hybrid proteasomes [12][13][14] . PA200 has been shown to be involved in DNA repair and degradation of acetylated histones [15][16][17] . Its exact molecular function as well as its regulation in disease or upon certain cellular stimuli, however, is still elusive.
We recently observed that activation and differentiation of pro-fibrotic myofibroblasts in fibrotic tissue remodeling of the lung requires the increased assembly and activity of 26S proteasomes 18 . Regulation of alternative proteasome complexes has not been investigated in this context so far. The present study focuses on the role of PA200 in myofibroblast differentiation and fibrotic tissue remodeling.

Results
PA200 is upregulated in idiopathic pulmonary fibrosis. PA200 protein levels were analyzed in explanted lung tissues from IPF patients and controls by Western blotting and immunohistochemistry. We first validated the specificity of the most commonly used anti-PA200 antibody targeted against amino acids 1620-1634 of the human protein. Using PA200-silenced human lung cells and testis tissue of PA200 −/− mice, we observed that this antibody unspecifically recognizes a protein species of 160 kDa, which has previously been proposed as a PA200 isoform 12 (Supplementary Fig. S1A-C). Further screening of antibodies identified other commercially available anti-PA200 antibodies that specifically detected PA200 by Western blotting or immunohistochemistry ( Supplementary Fig. S1D-F). When we applied these antibodies for Western blot analysis of human lung tissues, we observed that expression of PA200 was significantly upregulated in lungs of IPF patients compared to donors (Fig. 1A). Immunohistochemistry analysis identified increased expression of PA200 in well-known IPF driver cells, i.e. α smooth muscle actin (αSMA)-positive myofibroblasts 3 and keratin 5 (KRT5)-positive hyperplastic bronchial basal cells 19 (Fig. 1B and Supplementary Fig. S2). In contrast, PA200 was expressed only at low basal levels in normal donor lungs.
Increased formation of PA200-proteasome complexes in experimental fibrosis of lung and kidney. We investigated the formation of alternative PA200-proteasome complexes in fibrotic remodeling of the lung using bleomycin-induced lung fibrosis in mice 18,20 . Similar to IPF tissue, we observed significant upregulation of PA200 protein levels in fibrotically remodeled lungs 14 days after bleomycin instillation ( Fig. 2A). Furthermore, native gel analysis and subsequent immunoblotting revealed that PA200 was effectively recruited to the 20S proteasome to form alternative PA200-proteasome complexes (Fig. 2B). We also noted pronounced induction of PA200 expression as well as formation of PA200-proteasome complexes in an experimental model of kidney fibrosis (Fig. 2C,D). Fibrotic remodeling of the kidney was provoked by feeding mice with an oxalate-rich diet and confirmed by Masson-Trichrome staining for collagens ( Supplementary Fig. S3A) and elevated protein levels of αSMA 21 (Fig. 2C). Fibrotic kidney remodeling also involved the marked induction of 26S proteasome activity as shown previously for fibrotic remodeling of the lung 18 ( Fig. 2D and Supplementary Fig. S3B). Taken together, these data suggest that increased formation of alternative PA200-proteasome complexes is an inherent feature of fibrotic tissue remodeling.

PA200 does not protect from bleomycin-induced lung fibrosis in mice.
To assess the causal contribution of PA200 to fibrotic tissue remodeling, we induced lung fibrosis in PA200 −/− mice by a single instillation of bleomycin into the lungs and analyzed lung function and histology in comparison to wild type littermates. Development of lung fibrosis was not grossly altered in PA200 −/− mice as assessed by lung function parameters such as resistance and compliance, profibrotic marker gene expression, and hematoxylin & eosin staining ( Fig. 3A-C). These data indicate that deficiency of PA200 does not have a major impact on bleomycin-induced lung fibrosis in mice.
Induction of PA200-proteasome complexes in myofibroblast differentiation. Despite these negative in vivo data, we continued to analyze the function of PA200 in primary human cells of the lung, focusing on the two main profibrotic cell types with elevated PA200 expression in IPF tissue, i.e. bronchial basal cells and myofibroblasts. We first analyzed PA200 expression in primary bronchial basal cells isolated from lung organ donors and compared it to primary bronchial epithelial cells after 28 days of in vitro differentiation at air-liquid interface 22 . Expression of PA200 was almost twofold higher on transcript and protein level in undifferentiated basal cells compared to fully differentiated bronchial epithelial cells (Fig. 4A,B), which corresponded well to the elevated levels of PA200 in hyperplastic basal cells in IPF. Next, we investigated regulation of PA200 in primary human lung fibroblasts (phLF) in response to transforming growth factor (TGF)-β1 as the major profibrotic cytokine 23 . Treatment of phLF with TGF-β1 for 48 h increased expression of the myofibroblast marker αSMA and of the extracellular matrix proteins fibronectin (FN) and collagen1α1 (COL1A1) (Fig. 4C). Of note, TGF-β1 specifically upregulated expression of PA200 on the transcript and protein level (Fig. 4D,E). In contrast, other subunits of the 26S proteasome, such as the 19S regulator subunit RPT5 (gene name: PSMC3), the 20S subunits α1-7 and the 20S catalytic subunit β5 (gene name: PSMB5), were not regulated (Fig. 4D,E). Moreover, not only the expression but also the formation of alternative PA200-proteasome complexes was augmented upon myofibroblast differentiation (Fig. 4F). PA200 was found to be mainly associated with 20S proteasome complexes and to a lesser extent with the 26S proteasome forming hybrid complexes. These data demonstrate that PA200 is upregulated by the profibrotic stimulus TGF-β1 in the process of myofibroblast differentiation. PA200 is a negative regulator of myofibroblast activation. To dissect the functional role of PA200 in fibroblasts, we silenced PA200 in phLF. To our surprise, PA200 strongly induced myofibroblast differentiation www.nature.com/scientificreports www.nature.com/scientificreports/ as demonstrated by increased expression of the proliferation marker cyclin D1 and the myofibroblast markers αSMA and TGF-β1 over time (Fig. 5A). Protein levels of αSMA, proliferation markers cyclin D1, and proliferating cell nuclear antigen (PCNA) were also significantly upregulated 72 h after PA200 silencing (Fig. 5B). Accordingly, PA200 depletion resulted in enhanced fibroblast proliferation as shown by increased cell count (Fig. 5C). In contrast, an only 2-fold overexpression of PA200 in phLF for 24 h led to a significant reduction in cell proliferation (Fig. 5D,E). We were puzzled about this strong phenotype of PA200 silencing in human lung fibroblasts in light of the lack of phenotype for PA200 −/− mice in bleomycin-induced lung fibrosis and speculated www.nature.com/scientificreports www.nature.com/scientificreports/ that there might be species-specific differences. To investigate this hypothesis further, we analyzed primary mouse lung fibroblasts (pmLF) isolated from PA200 −/− mice for myofibroblast marker gene expression compared to pmLF isolated from wildtype littermates. Of note, deficiency of PA200 in these mouse lung fibroblasts did not result in any change of αSMA and collagen1α1 gene expression (Fig. 5F). In order to investigate if chronic depletion of PA200 leads to a compensatory effect in knockout mice and thereby conceals the function of PA200 as a negative regulator of myofibroblast differentiation we also performed siRNA-mediated PA200 silencing experiments in pmLF isolated from wildtype mice ( Supplementary Fig. S6). In contrast to our results obtained with primary human lung fibroblasts, we did not observe any effect of PA200 silencing on profibrotic marker gene expression in murine lung fibroblasts. These data point towards a species-specific effect for PA200 in myodifferentiation of human fibroblasts. PA200 regulates myofibroblast activation beyond TGF-β1 signaling. The induction of PA200 by TGF-β1 on the one hand and its role as a negative regulator of myofibroblast differentiation on the other can be reconciled if assuming that PA200 functions in a negative feedback loop to limit TGF-β1-induced myofibroblast activation. In silico analysis of the human PA200 promoter for conserved binding sites of the TGF-β1 responsive SMAD transcription factors, however, did not reveal any evidence for direct activation of the human PA200 promoter by TGF-β1 (data not shown). This is in line with the delayed upregulation of PA200 after 48 h (Fig. 4D,E) and suggests that PA200 is not directly regulated by TGF-β1. To corroborate this finding, we www.nature.com/scientificreports www.nature.com/scientificreports/ co-treated PA200-silenced phLF with TGF-β1 and assayed myofibroblast activation. Of note, co-treatment of PA200-silenced cells with TGF-β1 further stimulated myofibroblast marker gene expression when compared to PA200 expressing controls ( Fig. 6A and Supplementary Fig. S4A). Moreover, cellular proliferation was significantly increased upon TGF-β1 stimulation in PA200-depleted cells when compared to control siRNA-transfected cells (Fig. 6B). In contrast, PA200 overexpression prevented TGF-β1-induced cell proliferation confirming its function as a negative regulator of myofibroblast proliferation (Fig. 6C and Supplementary Fig. S4B). The additive effect of TGF-β1 treatment and PA200 silencing suggests that PA200 regulates myofibroblast activation to some extent independently of the TGF-β1 signaling pathway to limit myofibroblast differentiation.

Discussion
In the present study, we observed enhanced formation of alternative PA200-proteasome complexes in fibrotic tissue remodeling of lung and kidney and localized increased PA200 expression to hyperplastic basal cells and myofibroblasts in human IPF tissues. In accordance with this observation, PA200 expression was higher in undifferentiated basal cells compared to in vitro-differentiated bronchial epithelial cells and PA200-proteasome complexes were upregulated upon TGF-β1-induced myofibroblast differentiation. Unexpectedly, PA200 acted as a negative regulator of myofibroblast differentiation and partly independent of TGF-β1. This effect was specific for primary human cells and not observed in PA200 −/− mice. Hence, we propose that in primary human lung www.nature.com/scientificreports www.nature.com/scientificreports/ fibroblasts PA200 adjusts myofibroblast activation in response to pro-fibrotic stimuli, which fails in idiopathic pulmonary fibrosis.
Our immunohistochemical staining of IPF lungs revealed upregulation of PA200 in driver cells of fibrotic tissue remodeling 3,19 . For this analysis, we used anti-PA200 antibodies that had been validated for their specific detection of PA200 using siRNA-mediated gene silencing in cells and tissue from PA200 −/− mice. We would like to stress our validation effort as the most commonly used, commercially available anti-PA200 antibody turned out to unspecifically recognize proteins not related to PA200, which questions some of the published data 12 . PA200 was upregulated in hyperplastic basal cells overlying fibroblast foci. These KRT5-positive basal cells were recently described as a characteristic feature of IPF tissues 19,24 . We observed downregulation of PA200 in the course of basal cell differentiation in vitro indicating that elevated levels of PA200 are a specific feature of basal cells compared to differentiated bronchial epithelial cells. Expression of PA200 was also increased in myofibroblasts which drive excessive formation of connective tissue in tissue fibrosis 25,26 . Moreover, PA200 was upregulated by the pro-fibrotic cytokine TGF-β1 in primary human lung fibroblasts resulting in augmented formation of PA200-proteasome complexes in differentiated myofibroblasts. These results indicated that alternative PA200-proteasome complexes may promote myofibroblast activation and fibrotic tissue remodeling. Silencing and overexpression of PA200 in phLF, however, revealed that PA200 negatively regulated myofibroblast differentiation. These data suggest that PA200-proteasome complexes serve to limit myofibroblast differentiation similar to other components of the ubiquitin-proteasome system 27 . We speculate that this negative regulation of myofibroblast activation fails in fibrotic lungs of IPF patients and that PA200 is upregulated in a frustrated attempt to limit myofibroblast activation. The additive effect of PA200 silencing and TGF-β1 treatment on myofibroblast activation indicates that PA200 regulates myofibroblast differentiation partly independent of TGF-β1 signaling pathways. In contrast to the acute silencing of PA200 in primary human fibroblasts, we neither observed increased myofibroblast differentiation in chronically PA200-depleted or acutely PA200-silenced primary mouse lung fibroblasts nor any effect on bleomycin-induced lung fibrosis in PA200 −/− mice suggesting that PA200 function is possibly species-specific or somehow compensated in these knockout mice as observed in several other knockout strains 28 . In line with this observation, PA200 −/− mice lack any phenotypic defects except for their male infertility 29 . Even when crossed with p53 −/− mice, which exhibit increased tumor development, PA200 deficiency did not give rise to any additional phenotype 29 . Moreover, embryonic stem cells isolated from PA200 −/− mice did not show any altered sensitivity to bleomycin-induced DNA damage compared to wildtype cells 29 . PA200 has originally been described as an activator of the caspase-like activity of the proteasome by in vitro assays using PA200 purified from bovine testis 12 . A recent publication on recombinant human PA200-proteasome complexes, however, unambigiously shows activation of the trypsin-like activity of the proteasome 30 . PA200 −/− mice, however, do not show any obvious alterations in proteasome activity neither in the lung nor in testis (data not shown). Accordingly, PA200 has also been suggested to serve as a proteasome adaptor protein based on its structural composition and its rather small pore facilitating the entry of substrates for degradation 7,31,32 . PA200 is mainly composed of HEAT repeats, a repetitive array of amphiphilic alpha helices, which allow highly flexible and elastic conformational changes when interacting with different binding partners 33 . HEAT repeats facilitate protein-protein interactions and proteins harboring this structural motif are proposed to execute anchoring or adaptor functions 34 . Interestingly, the structurally related 200 kDa protein ECM29, which is also composed of multiple HEAT repeats and interacts with the 26S proteasome, was also proposed to function as an adaptor protein shuttling proteasomes to molecular motors and endosomes 35,36 . Adaptor proteins act as important temporal and spatial mediators of signaling cascades by transferring signals to specific protein complexes or subcellular compartments 37,38 . It is thus tempting to speculate that PA200 recruits proteasomes to distinct cellular  compartments and thereby indirectly regulates proteasomal protein degradation. We have previously shown that the activity of the 26S proteasome is essential for myofibroblast differentiation 18 . As TGF-β1 promoted the formation of PA200-containing 26S proteasome complexes in phLF these complexes might be required for fine-tuning TGF-β1-induced myofibroblast differentiation. This may also involve chromatin regulation as PA200 has been shown to be recruited to chromatin upon DNA stress and to be involved in the degradation of acetylated histones 14,16 .
In conclusion, we here report an unexpected regulation of myofibroblast activation by PA200 and for the first time observe its regulation in disease. Using validated antibodies our study indicates the limited use of PA200 −/− mice and contributes to an increased understanding of PA200 regulation and function.

Materials and Methods
All methods used in this study were carried out in accordance with the relevant guidelines and regulations.
Human IPF and donor lung tissues. Human lung tissues from 13 IPF patients and 9 organ donors were used for mRNA/protein expression analyses including immunohistochemical studies and have previously been described 18 . Additionally, lung tissue samples of 4 donors (mean age ± SD: 46.5 ± 4.80 years; 3 females, 1 male) and from histologically normal areas of surgical lung specimens from 2 patients undergoing resective surgery for benign or malignant tumors (66.00 ± 8.48 years; 2 females) were used for isolation of normal human lung fibroblasts. All lung tissue samples were collected within the European IPF registry (eurIPFreg) and provided by the UGMLC Giessen Biobank (member of the DZL Platform Biobanking). The study protocol was approved by the Ethics Committee of the Justus-Liebig-University Giessen (No. 111/08 and 58/15), and informed consent was obtained in written form from each subject. All methods were carried out in accordance with the relevant guidelines and regulations. Bleomycin-induced lung fibrosis. All animal studies conducted for this study were performed in accordance with international guidelines and were approved by the regional government of Upper Bavaria (animal approval file number 55.2-1-54-2532-114-2016). Lung fibrosis was induced in pathogen-free mice at the age of 10-12 weeks by instillation of a single dose of bleomycin (2 or 3 U/kg body weight depending on the batch) (Sigma Aldrich) to the lungs of age-and sex-matched mice as previously reported 18 .
Oxalate-induced kidney fibrosis. Seven to eight week old female C57Bl6N mice were fed either a calcium-free control diet or oxalate-rich diet (prepared by mixing 50 µmol/g sodium oxalate in calcium-free control diet) for 21 days to induce fibrotic remodeling of the kidneys as previously reported 21 . Isolation and culture of primary human lung fibroblasts. Primary human lung fibroblasts (phLF) from non-diseased control lungs were isolated and cultured as described previously 39 . phLF between passage 3-5 were used for experiments. phLF were synchronized in starvation medium supplemented with 1% FBS 24 h prior to treatment with 5 ng/mL TGF-β1 (240-B-002, R&D Systems) for 24 h or 48 h. For analysis of proliferation, cells were harvested with trypsin, mixed with trypan blue (Sigma-Aldrich) and counted using a Neubauer counting chamber.

Isolation and differentiation of primary human basal cells. Primary human bronchial epithelial cells
were obtained from Lonza (Wokingham, UK) or isolated from airway tissue resected from patients undergoing lung surgery as described previously 22 . Isolated basal cells were in vitro differentiated for 28 days to a fully reconstituted epithelium, as described previously 22  PA200 overexpression. phLF were transfected with a PA200 pcDNA3.1 expression vector by forward transfection using the Lipofectamine LTX reagent (Thermo Fisher Scientific). 24 h after seeding phLF, the transfection mix was prepared in OptiMEM according to the manufacturer's protocol and added dropwise to cells. Medium was replaced with culture medium after 5 h and cells were harvested after 24 h hours. For TGF-β1 treatment, cell culture medium was replaced by 1% FBS starvation medium supplemented with 5 ng/mL TGF-β1 24 h after transfection and cells were harvested 24 h later. Native gel analysis. For generation of native protein extracts frozen tissues were homogenized twice for 30 s at 3000 rpm using a Mikro-Dismembrator S (Sartorius) prior to lysis under non-denaturing conditions. Lysis of tissue powder or cell pellets under native conditions in TSDG buffer and native gel analysis was performed as described before 40 .
Immunohistochemistry. Immunohistochemistry of paraffin-embedded IPF and donor lung tissues was performed as previously reported [18]. For staining of PA200 in paraffin-embedded murine testis tissues, sections were deparaffinized and permeabilized in methanol/hydrogen peroxide (80%/1.8% (v/v)) for 20 min. Heat-induced antigen retrieval was performed in citrate buffer pH 6 using a decloaking chamber (Biocare Medical) and unspecific binding sites were blocked with Rodent Block M (Biocare Medical) for 30 min. Slides were incubated with primary antibody diluted in Antibody Diluent (DAKO) for 1 h at RT. Afterwards, slides were incubated with MACH 2 Rabbit AP-Polymer (Biocare Medical) for 30 min at RT and Vulcan Fast Red AP substrate solution (Biocare Medical) for 10 min. Nuclei were stained with hematoxylin and slides were imaged using the MIRAX scanning system (Zeiss).

Statistical analysis.
Statistical analysis was performed with GraphPad Prism 5 and 7. Applied statistical tests are indicated in figure legends. P-values < 0.05 were considered statistically significant with *P < 0.05, **P < 0.01 and ***P < 0.001. Experiments with primary human lung fibroblasts were always performed in biological replicates with cells from different organ donors.
Other experimental procedures are described in the Supplementary Information.

Data availability
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files). Materials, data and associated protocols will be made available to readers without undue qualifications in material transfer agreements.