The inhibitive effects of proteasome inhibitor MG-132 on pterygium fibroblasts in vitro and the potential key regulators involved

Dandan Zhao a,b, Hongxia Zhao b, Yang He b, Yang Yang b, Yan Du b, Meixia Zhang a

Abstract

This study aimed to determine whether MG-132 as a proteasome inhibitor can effectively hinder pterygium progression, and to screen out potential regulators involved in MG-132 mediated process. Human pterygium fibroblasts (HPFs) were derived from pterygium tissues from 5 patients. Cell proliferation was examined by MTT, cell cycle and apoptosis were detected by flow cytometry. The overgrowth pterygium tissues were characterized by H&E staining and IHC compared with normal tissues. Differential mRNA expression with MG-132 treatment was determined by RNA sequencing and analyzed by GO and KEGG pathways. The expression levels of Nrf2, MCPIP1, CDKN1B and XBP1, four genes closely associated with pterygium, were detected by RT-qPCR and western blotting. MG-132 dose-dependently inhibited the growth of HPFs, induced G2/M phase arrest of cell cycle at a certain dose, and also caused cell apoptosis, with the levels of cleaved caspase3, cleaved PARP, Bax and p21 increased. Ki-67 and Bcl-2 were highly expressed while Bax was decreased in pterygium tissues. Total 7199 differentially expressed genes (DEGs) were identified, including HSPA family most significantly increased, and AL590428.1, AL122125.1 and lincRNAs such as FGF14-AS2 decreased. The up-regulated DEGs were mainly enriched in RNA degradation pathway, while down-regulated DEGs were related to the regulation of cell cycle. The expressions of Nrf2 and MCPIP1 were significantly increased, while XBP1 and CDKN1B were decreased. In conclusion, MG-132 inhibited the proliferation and induced apoptosis of HPFs in vitro with 7199 DEGs participated in, which may provide a useful reference for the exploitation of MG-132 in treating pterygium.

Keywords:
Pterygium
Fibroblasts
MG-132
Cell proliferation Cell apoptosis
RNA sequencing

1. Introduction

Pterygium is a type of uncontrolled overgrowth conjunctive tissue that locates over the sclera [1]. Due to its fibrovascular characterization, it can cause remarkable damage to visual function in advanced cases by triggering redness and irritation [2]. As the incidence of pterygium can reach to 31% in certain areas, the aggressive clinical behavior of pterygium is considered as a threaten issue [3]. One of the recent studies has suggested the progression of pterygium requires cell proliferation and invasion, fairly similar to the cancer metastases [1]. It has been identified that genes responsible for cell cycle, cell proliferation and apoptosis, such as tumor inhibitor p53 and c-Myc, are abnormally downregulated in pterygium samples [4–7]. However, effective therapeutic strategies to pterygium require further exploration, since the exact etiology of pterygium is not fully understood. At present, the treatment of pterygium mostly depends on surgery or drug combined surgery, while the treatment effect is not ideal and easy to relapse after surgery [8]. Therefore, it is very important to study how to inhibit the proliferation of HPFs after surgery to prevent the recurrence and treatment of pterygium fibroblasts.
Ubiquitin-proteasome pathway (UPP) plays a vital role in cellular processes regulation, including cell proliferation and apoptosis [9]. UPP can selectively degrade short-lived proteins, in which way mediates most intracellular proteolysis [9]. Specifically, proteasome lay in the nucleus and cytoplasm of all eukaryotic cells recognizes ubiquitinated proteins, then hydrolyzes and degrades the substrates of ubiquitinated protein, including transcription factors, cell cycle regulators and pro- apoptotic proteins. MG-132 (carbobenzoxy-Leu-Leu-leucinal) is a peptide aldehyde inhibitor derived from Chinese natural plant [10]. It is reported that MG-132 is capable of blocking the proteolytic activity of proteasome complex (such as 20S and 26S proteasome) by covalently binding to their active sites. Therefore, MG-132, via blocking UPP to prevent ubiquitin–proteasome from degradation [11,12], is able to inhibit the growth of tumor cells by leading to the cell cycle arrest and triggering apoptosis [10]. Although MG-132 is considered as a kind of promising antitumor drug [13], the effect of MG-132 on the progression of pterygium has not been reported at present according to our investigation.
In this study, we employed human pterygium fibroblasts (HPFs) derived from pterygium tissue to investigate the effect of MG-132 on proliferation and apoptosis of HPFs in vitro. Further, RNA sequencing and bioinformatical analysis were carried out to elucidate the possible molecular mechanism by which MG-132 downregulated the HPFs growth.

2. Methods

2.1. Patients and tissue samples

Primary human pterygium tissue samples were obtained from Yan’An Hospital of Kunming City (Yunnan, China). All the surgical resected tissues were provided by 5 patients (2 males and 3 females) and control conjunctiva tissues were provided by 5 patients with retinal detachment (2 males and 3 females). The study protocol was approved by the institutional review board of Yan’An Hospital of Kunming City (Yunnan, China), and the study was performed as per the principles of the Declaration of Helsinki. Written informed consents were obtained from all the sample donors according to the International Ethical Guidelines.

2.2. Cell cultures

Human pterygium fibroblasts (HPFs) were isolated from the caudomedial part of harvested pterygium tissues according to the published method [14]. In brief, the fresh pterygium tissues were washed 3 times with sterile Hank’s balanced salt solution (Gibco, Carlsbad, CA, USA) and cut into pieces (1.0 mm3). Several small tissue pieces were sucked and distributed evenly on the bottom of the culture flask. The culture flask was turned over and placed in a 37 ◦C cell incubator for 2 h. Afterwards, a high-glucose Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS and 1% double antibody was added to cover the tissue with small pieces, and then placed in an incubator for culture. The medium was changed every 3 days, and when the cells reached 80% confluence, they were subcultured and continued. After being digested using 0.125% trypsin-EDTA and 0.1% collagenase (Thermo Fisher, Waltham, MA, USA), the primary HPFs were extracted and then cultured in the medium at 37 ◦C in a humidified incubator with 5% CO2. The DMEM was supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin/streptomycin (Gibco, Carlsbad, CA, USA). After primary monolayer was obtained, cells were detached and sub-cultured at a split ratio of 1:2. Finally, cells at a density of 1 × 105/mL were seeded in cell bottles and incubated to the mid-log phase before the subsequent experiments.
For cell identification, cells were fixed in ice-cold methanol for 5 min and then in acetone for 1 min at 220 ◦C, followed by a brief wash. Cells were permeabilized in a 0.1% Triton X-100 solution (Solarbio, Beijing, China) for 1 h at 37 ◦C. Then, the slides were stained with mouse anti- vimentin antibody (clone V9, 1:100, Santa Cruz, CA, USA) for 18 h at 4 ◦C. After that, cells were incubated with Goat Anti-Rabbit IgG H&L (Alexa Fluor® 555, ab150078, 1:500, Abcam, Cambridge, MA, USA) for 45 min at 25 ◦C. Nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI, 1 mg/mL in 0.1 M PBS) for 10 min. Finally, images were captured by a light microscope (Olympus, Tokyo, Japan).

2.3. Cell viability assay

Cell viability was detected using a Cell Counting Kit-8 (CCK-8, Beyotime Biotechnology, Shanghai, China) as per the manufacturer’s protocols [15]. HPFs (1 × 104 cells/well) were seeded into 96-well plates and incubated for 24 h. The cells were treated with MG-132 (Sigma, St. Louis, MO, USA) at a final working concentration of 0, 1, 2.5, 5, 10, 15 μmol/L, respectively. Cell viability was assessed at 12, 24, 48, and 72 h with MG-132treatment at different doses. The absorbance (A) at 450 nm was measured using an automatic microplate reader (Bio- Rad, Richmond, CA, USA).

2.4. Cell cycle and apoptosis analysis

The cells were treated with MG-132 for 48 h at a final working concentration of 0, 1, 2.5, 5, 10, 15 μmol/L, and then were fixed with 70% pre-cold alcohol at 4 ◦C overnight. Subsequently, the cell pellet was treated with 100 mg/L RNase (DNA Reagent Kit, BD Biosciences, CA, USA) for 30 min and was double stained by Annexin-V-FITC and PI (Sigma, St. Louis, MO, USA) in dark for another 30 min. The pretreated samples in each group were set at 1 × 106 density, and immediately applied onto a flow cytometer (FACStar, BD Biosciences, CA, USA). The rate of cell apoptosis and the cell cycle were analyzed using FlowJo software (Version 10.5.3; FlowJo LLC, Ashland, OR, USA). At least 20,000 events were acquired per analysis.

2.5. Western blot analysis

The total cellular protein was extracted from treated cells and quantified using a BCA Kit (Beyotime, Haimen, China). Then, the protein samples were subjected to a 10% polyacrylamide gel for separation. Subsequently, proteins were transferred onto a polyvinylidene fluoride (PVDF, Thermo Fisher, CA, USA) membranes, followed by incubation with primary antibodies including anti-Ki-67 (ab16667, 1:1000), anti- PCNA (ab92552, 1:1000), anti-CDK1 (ab32094, 1:1000), anti-cyclin D1 (ab16663, 1:1000), anti-p21 (ab218311, 1:1000), anti-cleaved cas- 3 (ab2302, 1:1000), anti-cleaved PARP (ab32064, 1:1000), anti-Bcl-2 (ab185002, 1:1000), anti-Bax (ab32503, 1:1000) and anti-GAPDH (ab181602, 1:1000) antibodies. The membranes were then incubated with the secondary antibody goat anti-rabbit IgG (1:1000) for 1 h at room temperature. The used antibodies were purchased from Abcam (Cambridge, MA, USA). The density of bands was normalized by using GAPDH as an internal reference, and measured by ImageJ software (Rawak Software Inc., Stuttgart, Germany).

2.6. Hematoxylin-eosin (H&E) staining

The pterygium tissues and control conjunctiva tissues were fixed overnight with 4% paraformaldehyde at 4 ◦C. After that, tissues were cut into 5 μm slices and stained with H&E solution (Nanjing Jiancheng, Jiangsu, China) as per the manufacturers’ instructions [16]. Finally, slices were mounted with neutral resins and images were captured by a light microscope (Olympus, Tokyo, Japan).

2.7. Immunohistochemistry (IHC)

The IHC assay was conducted as per the standard method published elsewhere [17]. The pterygium tissues and control conjunctiva tissues were prepared into paraffin tissue sections of about 5 μm, followed by a series of instructions including dewaxing, hydration, and blocking. The primary antibodies anti-Ki-67 (ab245113, Abcam; 1:1000), anti-Bax (ab32503, Abcam; 1:1000) and anti-Bcl-2 (ab185002, Abcam; 1:1000) were added and incubated at 37 ◦C for 30 min. After that, Goat Anti- Mouse Anti-Rabbit IgG/IgM H&L (HRP polymer) (ab2891, Abcam, 1:1000) was blocked for 20 min. After that, DAB chromogenic solution (Solarbio, Beijing, China) was added to the tissues and observed under a microscope, and the staining was terminated with water at appropriate time, and the staining time was not more than 5 min. A total of 100 μL of hematoxylin was added to stain the nucleus, and then placed in water for 3 min to stop. Hydrochloric acid: alcohol (1:99) solution and ammonia water are sequentially added for cleaning, and then washed with water until the dyeing is uniform. Finally, the slices were washed in 75% ethanol, 95% ethanol, 100% ethanol and xylene III/II/I in sequence, then sealed and observed by fluorescence microscope (Olympus, Tokyo, Japan).

2.8. RNA sequencing and bioinformatics

The total mRNA was extracted from HPFs with or without MG-132 treatment, followed by purity detection. Two cell lines in each group were employed. The analysis of differential mRNA expression was accomplished by Novogene Co., Ltd. (Beijing, China). The functions of differential mRNA were analyzed through bioinformatics methods including GO analysis and KEGG pathway analysis [18].

2.9. Real-time reverse transcription quantitative PCR (RT-qPCR)

Total mRNA in HPFs was extracted using the RNA Extraction kit (Life Technologies, MA, USA) in accordance to the manufacturer’s instructions. The extracted total mRNA was added into the reverse transcription system to synthesize the complementary DNA (cDNA). Subsequently, the level of relative mRNA was detected using SYBR qPCR Master Mix (Thermofisher, Beijing, China). A total 20 μL of StepOne Real-Time PCR system was subjected to the following thermal conditions: 5 min at 95 ◦C for pre-denaturation, followed by 30 cycles of 10 s at 95 ◦C, 30 s at 60 ◦C, and 30 s at 72 ◦C for extension. The specific primers used in RT-qPCR were as follows: Nrf2, F: 5′-GTGGCTGCTCAGAATTGCAG-3′, R: 5′-CATTGCCATCTCTTGTTTGCTG-3′; MCPIP1, F: 5′- GGTGTGCTATGACGACA GATTC-3′, R: 5′-CTTCTTACGCAGGAAGTTGTCC-3′; CDKN1B, F: 5′-CCTACAGGGGATTGTGTTTTG-3′, R: 5′-CTGGAGTTGAGTAACGAGCTG-3′; XBP1, F: 5′-CAGCACTCAGACTACGTGC-3′, R: 5′-GGTCCTTCTGGGTAGACCTC-3′. The gene GAPDH was employed as an internal reference and the relative mRNA level of target genes were calculated using the 2− ΔΔCt method.

2.10. Statistical analysis

All data represented at least 3 independent repeats and were presented as the mean ± SD by GraphPad Prism 7.0 analysis (software, CA, USA). The comparisons among multiple groups were performed using one-way analysis of variance (ANOVA) followed by Tukey’s test. The comparison between two groups was examined by Student’s t-test. P values of less than 0.05 or 0.01 were considered significant.

3. Results

3.1. Effects of MG-132 on cell growth and cell cycle distributions in HPFs

During cell culture, the adherent cell population that spread on the surface of the plastic culture vessels was not homogeneous but made up mainly cells that showed spindle-like morphology at first. After culturing for several generations, we identified the cells with Vimentin staining (Fig. S1). We found that Vimentin-positive cells all presented spindle-like cell morphology, which was the typical characteristic of fibroblast cells. Thus, these extracted and purified cell population were defined as fibroblast cells. As shown in Fig. 1A, MG-132 at relatively low doses significantly reduced the viability of HPFs beginning after 48 h in a concentration-dependent manner, whereas MG-132 at 15 μmol/L remarkably decreased cell viability after treatment for 12 h (P < 0.01). When the working dose was less than 5 μmol/L, though MG-132 played an inhibitory role, the HPFs still exerted growing trend. However, HPFs stopped growing within 72 h and even underwent severe reduction at the concentration of 5.0 μmol/L and 15 μmol/L, respectively. The suppression effect reached a maximum at 72 h, with inhibition rates of 22.08%, 33.53%, 38.42%, 40.33%, 50.72% and 68.62% at concentrations of 1.0, 2.5, 5.0, 10.0, and 15.0 μmol/L, respectively. Then, we examined the effect of MG-132 on cell cycle distributions in HPFs (Fig. 1B–C). MG-132 at high doses (5.0, 10.0, and 15.0 μmol/L) significantly decreased the cells in G1 phase while induced a G2/M phase arrest of the cell cycle in HPFs (P < 0.05). Treatment with 1 μmol/ L MG-132 made no evident change on cell cycle distribution, while 2.5 μmol/L MG-132 induced an S phase arrest. As western blot assay indicated (Fig. 1D), the protein expression of Ki-67, PCNA, CDK1, cyclin D1 significantly decreased with the working concentration of MG-132 increased (P < 0.05). The protein level of p21 presented the contrary trend. When at 15 μmol/L, the level of p21 was promoted to more than 1.5-fold of that in control group without MG-132.

3.2. Effects of MG-132 on cell apoptosis in HPFs

The result of flow cytometer detection (Fig. 2A–B) showed that the rate of late cell apoptosis was increased as the dose of MG-132 increased. Consistent with the cell viability result, MG-132 treatment for 48 h exhibited significant pro-apoptotic effect on HPFs, with apoptotic rates of 5.25%, 15.61%, 18.39%, 30.49%, 35.79%, 42.51% at the corresponding concentrations of 0, 1.0, 2.5, 5.0, 10.0, and 15.0 μmol/L. The levels of apoptosis related proteins cleaved caspase 3, cleaved PARP and Bax, as well as the calculated ratio of Bax/Bcl-2, were significantly promoted in a dose-dependent manner with treatment of MG-132 (Fig. 2C).

3.3. The morphology and expression of proliferation- and apoptosis- related genes in pterygium tissues

The H&E staining result showed that pterygia tissues were characterized by the covering of excessive fibrovascular proliferations and connective tissues (black oval) (Fig. 3A). In the shallow layer of the matrix, there was a transparent area of amorphous material and rough granular acidophilic substances, and there was rough fibrous tissue in the deep matrix of the pterygium (Fig. 3A). In control conjunctiva tissues, the cells were arranged neatly, the matrix was loose connective tissue, the collagen fibers were arranged in parallel, and fibroblasts could be seen between them, and a small amount of neutrophils and capillaries were scattered (Fig. 3A). Compared to the control conjunctiva tissues, the pterygia tissue presented highly branching vascular stem and looser connective tissue. Fibrinoid change in unique lesions could be observed in pterygia tissues (black arrow, Fig. 3A). As shown in Fig. 3B, Bcl-2 and Bax could be seen distributed in the surface of cell membrane and nuclear membrane, while Ki-67 mainly located in the nucleus. The results of IHC presented that Ki-67 and Bcl-2 were highly expressed while Bax was decreased in pterygium tissues, compared to the control conjunctiva tissues (red arrow, Fig. 3B).

3.4. RNA-Seq transcriptional profiles of HPFs in response to MG-132 stimulation

HPFs cells isolated from the tissues treated with or without MG-132 at 5 μmol/L for 48 h were set as MG-132 group and untreated (control) group, respectively. Through bioinformatical analysis, 7199 differentially expressed genes (DEGs) were identified, including 3913 upregulated genes and 3286 downregulated genes. Overall gene-based differential expression in HPFs with MG-132 was presented as a volcano plot (Fig. 4A) and a corresponding heatmap that omitted gene name (Fig. 4B). The full text of DEGs was listed in Table S1. Specifically, we noticed that the expressions of HSPA family and some lincRNAs (DNAJB5-AS1, Lnc-NTMT1-2 and Lnc-CTH-8) were the most significantly increased, while the expressions of other lincRNAs and antisense RNAs were most decreased in response of MG-132 stimulation. The expressions of HSPA7, HSPA6 and HSPA1A were promoted to 14.7-fold and 13.0-fold and 8.7-fold, respectively. Certain antisense RNAs including AL590428.1, AL122125.1 and lincRNAs such as FGF14-AS2 were reduced to more than 6-fold with MG-132 treatment. The gene MFNG was the most downregulated mRNA with a reduction of 7-fold.

3.5. GO analysis and KEGG pathway analysis of the differentially expressed mRNAs

The GO analysis (Fig. 5A) showed that both the upregulated and downregulated DEGs were significantly enriched in biological processes (BP), cellular components (CC) and molecular functions (MF). As for BP, the upregulated DEGs were mainly enriched in ribonucleoprotein complex biogenesis, mRNA catabolic process and translational initiation and the downregulated DEGs were enriched in extracellular matrix organization, positive regulation of cell cycle process and regulation of cell cycle phase transition. With regards to CC, the upregulated DEGs were enriched in proteasome complex while downregulated DEGs were focus on nuclear chromosome part. The enrichment of all the up- and down- regulated DEGs was identified using KEGG analysis separately (Fig. 5B). Interestingly, the downregulated DEGs were enriched in metabolism and metabolic pathways while upregulated DEGs were concentrated on RNA degradation pathway.

3.6. MG-132 upregulated the expression of Nrf2, MCPIP1, and downregulated CDKN1B and XBP1

The RT-qPCR and western blot method were employed to detected four genes of Nrf2, MCPIP1, CDKN1B and XBP1 that were reported associated in pterygium progression [19–22]. As shown in Fig. 6A, the mRNA expressions of Nrf2 and MCPIP1 were significantly increased, while XBP1 and CDKN1B were decreased with 5 μmol/L MG-132 treatment at 48 h (P < 0.05), which was highly consistent with the result of western blot (Fig. 6B). However, no significant changes in Nrf2 and MCPIP1 expression were found and XBP1 was significantly increased in the RNA-seq expression profile. The RT-qPCR verification and western blot results were not consistent with RNA-Seq profiling results, which may be caused by the small sample size in RNA- Sequencing.

4. Discussion

MG-132 is an aldehyde peptide proteasome inhibitor that reversibly inhibits the activity of the proteasome when combined with the proteasome. It can inhibit the function of the proteasome, making it unable to degrade the target protein related to cell proliferation, differentiation and apoptosis, block UPP and promote apoptosis [11]. Previous studies had found that MG-132 had a significant inhibitory effect on the proliferation, migration and invasion of many types of cancer cells, and could also induce their apoptosis [23,24]. Given that pterygium was a fibrous vascular proliferative disease with tumor-like features [25], we explored the effect of MG-132 on the growth and cycle distribution of HPFs. As expected, MG-132 was able to significantly inhibit the viability of HPFs in a concentration-dependent manner. Clinical studies had shown that the occurrence of pterygium may be associated with the abnormality of the normal cell cycle [26]. Our results showed that high doses of MG-132 could significantly reduce HPFs in the G1 phase, make the G2/M phase of the cell cycle stagnating, and significantly change the expression of proteins closely related to abnormal cell proliferation. The results of further apoptosis tests showed that MG-132 could increase the apoptosis of HPFs in a concentration-dependent manner, and can significantly increase the expression of apoptosis-promoting proteins. These results indicated that MG-132 could not only inhibit the proliferation of HPFs like tumor cells, but also induce the apoptosis of HPFs.
The proliferation and apoptosis of HPFs are closely related to pterygium development and postoperative recurrence [27]. Ki-67 is an antigen related to cell division and proliferation in the nucleus and an important indicator reflecting cell proliferation [28]. Bcl-2 and Bax have an important effect on cell apoptosis, Bax and Bcl-2 can form heterodimers to induce apoptosis, and Bax itself has a pro-apoptotic effect [29]. The results of cell and histological analysis indicated that MG-132 significantly promoted the protein expression of Bax and Bcl-2 in HPFs in a dose-dependent manner, while significantly inhibited the expression of Ki-67, thus reducing the abnormal proliferation and promoting apoptosis of HPFs.
The pathogenesis of pterygium is complicated. With the widespread application of transcriptome sequencing technology, the exploration of the corresponding pathogenesis of pterygium can be further explored [30]. Based on this, we investigated the effect of MG-132 on HPFs transcriptome changes. The results showed that the expressions of 3913 mRNAs were up-regulated and 3286 mRNAs were down-regulated significantly. Among them, the mRNA expressions of HSPA family and some certain lincRNAs were the most significant up-regulated, while the mRNA expressions of another some lncRNAs and antisense RNAs were the most significant down-regulated.
Heat shock protein family A (HSPA, also known as HSP 70), a heat shock protein with a molecular weight of approximately 70 kDa, is an important member of the heat shock protein family. The HSP70 family is highly induced in stress cells of almost all organisms, and has the function of protecting cells and resisting stress. RNA-Seq profiling results of DEGs showed that the mRNA levels of various including HSPA7, HSPA6 and HSPA1A in MG-132-treated HPFs were significantly increased, whereas these HSPAs were critical participants in protein homeostasis, especially in protein folding, depolymerization and degradation [31], and this could also be reflected in the results of GO terms that the upregulated DEGs were mainly enriched in ribonucleoprotein complex biogenesis and proteasome complex. It was reported that the expression of HSP70 gene in cells was related to the inhibition of monocyte toxicity [32]. In addition, HSP70 was characterized by the function of regulating cell proliferation [33]. Since the biological behavior of HPFs was similar to tumor cells, which suggested that MG- 132 may exert anti-tumor activity by increasing the expression of HSP70, thereby inducing apoptosis of HPFs. HSPA1A was characterized by the stress-induced expression and was the main stress-induced expression member of the HSP70 family [34], the stress response was also one of the typical characteristics of cancer [35]. Moreover, some GO terms of HSPA1 revealed it was closely related to transcription corepressor activity and RNA binding, which was also reflected in our GO terms analysis. HSPA6 was not expressed in most tissues, but was abundantly expressed in blood and immune cells, and HSP70 1A has been reported as a target for anti-cancer drugs in Jurkat cells [36,37]. Compared with other HSP70 family members, HSPA6 had stronger regulation and inducibility. It was reported that MG-132 was an effective inducer of HSPA6 and efficiently induced the expression of HSPA6. Our results also showed that the mRNA expression of HSPA6 could be significantly up-regulated after treatment with MG-132. There were few reports about HSPA7, only some of its GO terms were found, such as heat shock protein binding, suggesting HSPA7 may act as a helper to activate the expression of other HSPA family proteins (https://www.genecards. org/Search/Keyword?queryString=HSPA7). Combined with our results, MG-132 could activate HSPA family to exert its anti-fibrotic effect. Since there are few reports on the specific mechanism of HSPA family regulating fibrosis, however, their specific mechanism is still unclear.
Long chain non-coding RNAs (lncRNAs) are a type of non-protein transcripts larger than 200 nt in size [38], it can be divided into different types according to the relative positions of their coding sequences and protein genes. If their coding region is located between the two protein-coding genes, they are called intergenic lncRNAs (lincRNAs). Some lincRNAs with the most significant up-regulated expression after MG-132 treatment included DNAJB5-AS1, lnc-NTMT1-2 and lnc- CTH-8. There is almost no literature about these three lincRNAs with only GO terms annotated that they were associated with neuromodulation drugs, and we found for the first time that MG-132 caused an increase in the expression of these lincRNAs. The lincRNA with the most significant down-regulation of mRNA expression was FGF14-AS2. lincRNA FGF14-AS2 is associated with breast cancer progression and prognosis, and studies have reported that FGF14-AS2 was significantly down-regulated in breast cancer [39]. Previous studies have shown that some lincRNAs could activate fibrosis in tissues or cells. For example, lincRNA TINCR promoted excessive proliferation and inflammation of skin fibroblasts [40], and lincRNA Gm4419 increased proinflammatory production of cytokines and fibrosis biomarkers. It should be noted that not all lincRNAs were pro-fibrotic [41], our results only showed that MG-132 may inhibit some fibrosis- or tumor-related lincRNAs in HPFs to exert anti-fibrosis effect.
In addition, our RNA-Seq profiling results showed that the treatment with MG-132 significantly reduced the expression of some antisense RNAs including AL590428.1 (lnc-SLC17A5-1) and AL122125.1(FRMD6- AS1) in HPFs. There are almost no reports about the role of lnc- SLC17A5-1 in the disease. As for FRMD6-AS1, studies have reported that his family gene-FRMD6-AS2 had a significant inhibitory effect on endometrial cancer of the uterus. We discovered for the first time that MG-132 treatment could reduce the expression of lnc-SLC17A5-1 and FRMD6-AS1 in HPFs. Some antisense RNAs may also play important roles in the fibrosis of living cells. For instance, some antisense lincRNAs could regulate the expression of some fibrotic transmembrane receptors [42]. Combined with our results, extensive researches on the mechanism of lnc-SLC17A5-1 and FRMD6-AS1 influencing pterygium fibrosis can be conducted in the future.
The RNA-seq expression profile results showed that the mRNA expression of CDKN1B was decreased and XBP1 was significantly increased. CDKN1B can mediate cell adhesion and promote cell differentiation [20], and is a related gene that regulates cell cycle [43]. From our results, MG-132 may activate the polymorphism of CDKN-1B gene and raised its expression [44]. XBP1 is the main regulatory factor in the endoplasmic reticulum stress response, which can affect the growth and apoptosis of cells by affecting multiple signaling pathways [21]. Clinical studies revealed that XBP1 was highly expressed in a variety of malignant tumor cells, and could induce neovascularization of malignant tumors through various mechanisms and inhibit tumor-related immunity, resulting in the rapid proliferation of malignant tumor cells [45]. Our RT-qPCR verification and western blot results were not consistent with RNA-Seq profiling results, which may be caused by the inaccuracy in RNA-Seq profiling results. Nrf2 is a key transcription factor in anti- oxidative stress, and oxidative stress plays an important role in apoptosis [19]; MCPIP1 is an important RNA-binding protein that can induce apoptosis [22]. Although no significant changes in Nrf2 and MCPIP1 expression were found in the expression profile, we also investigated the protein expression of them based on these reports. The results of western blotting showed that the protein expressions of Nrf2 and MCPIP1 were significantly increased. Previous studies have shown that Nrf2 had a protective effect on lung fibrosis damage caused by toxoid [46], whereas MCPIP1 was a tumor suppressor gene that inhibited the vitality of cancer cells and regulate the proliferation of cancer cells [47]. These results indicated that Nrf2, MCPIP1, CDKN1B and XBP1 were involved in the formation of pterygium, which was closely related to the development of MG-132 regulating pterygium.
In summary, the effect of MG-132 on pterygium was examined in detail, and the RNA-seq expression profile was comprehensively analyzed. MG-132 could significantly ameliorate the pterygium, which may be related to inhibiting the proliferation of HPFs, increasing the apoptosis of HPFs cells, and regulating the mRNA expression profiles (such as HSPA and certain lincRNAs and antisense RNAs) and some protein expressions of Nrf2, MCPIP1, CDKN1B and XBP1 related to cell proliferation and apoptosis. Therefore, MG-132 may be used as a potential drug to prevent the recurrence and treatment of pterygium.

References

[1] T. Liu, Y. Liu, L. Xie, X. He, J. Bai, Progress in the pathogenesis of pterygium, Curr. Eye Res. 38 (12) (2013) 1191–1197, https://doi.org/10.3109/ 02713683.2013.823212.
[2] E.T. Detorakis, D.A. Spandidos, Pathogenetic mechanisms and treatment options for ophthalmic pterygium: trends and perspectives (review), Int. J. Mol. Med. 23 (4) (2009) 439–447, https://doi.org/10.3892/ijmm_00000149.
[3] N. Li, T. Wang, R. Wang, X. Duan, Tear film instability and Meibomian gland dysfunction correlate with the pterygium size and thickness pre- and postexcision in patients with pterygium, J. Ophthalmol. 2019 (2019) 5935239, https://doi.org/ 10.1155/2019/5935239.
[4] S. Tsironi, E. Ioachim, M. Machera, M. Aspiotis, N. Agnantis, K. Psillas, Immunohistochemical HLA-DR antigen expression with lymphocyte subsets and proliferative activity in pterygium, In Vivo. 16 (5) (2002) 299–306.
[5] Y. Ueda, S. Kanazawa, T. Kitaoka, Y. Dake, A. Ohira, A.M. Ouertani, et al., Immunohistochemical study of p53, p21 and PCNA in pterygium, Acta Histochem. 103 (2) (2001) 159–165, https://doi.org/10.1078/0065-1281-00584.
[6] K. Liang, Z. Jiang, B.Q. Ding, P. Cheng, D.K. Huang, L.M. Tao, Expression of cell proliferation and apoptosis biomarkers in pterygia and normal conjunctiva, Mol. Vis. 17 (2011) 1687–1693.
[7] E.T. Detorakis, A. Zaravinos, D.A. Spandidos, Growth factor expression in ophthalmic pterygia and normal conjunctiva, Int. J. Mol. Med. 25 (4) (2010) 513–516, https://doi.org/10.3892/ijmm_00000371.
[8] A. Kormanovski, F. Parra, A. Jarillo-Luna, E. Lara-Padilla, J. Pacheco-Yepez, ´ R. Campos-Rodriguez, Oxidant/antioxidant state in tissue of prymary and recurrent pterygium, BMC Ophthalmol. 14 (2014) 149, https://doi.org/10.1186/1471-2415- 14-149.
[9] N. Guo, Z. Peng, MG132, a proteasome inhibitor, induces apoptosis in tumor cells, Asia-Pacific Journal of Clinical Oncology 9 (1) (2013) 6–11, https://doi.org/ 10.1111/j.1743-7563.2012.01535.x.
[10] Y.H. Han, W.H. Park, MG132, a proteasome inhibitor decreased the growth of Calu-6 lung cancer cells via apoptosis and GSH depletion, Toxicol. in Vitro 24 (4) (2010) 1237–1242, https://doi.org/10.1016/j.tiv.2010.02.005.
[11] A.F. Kisselev, A.L. Goldberg, Proteasome inhibitors: from research tools to drug candidates, Chem. Biol. 8 (8) (2001) 739–758, https://doi.org/10.1016/s1074- 5521(01)00056-4.
[12] L. Zhang, H. Tang, Y. Kou, R. Li, Y. Zheng, Q. Wang, et al., MG132-mediated inhibition of the ubiquitin-proteasome pathway ameliorates cancer cachexia, J. Cancer Res. Clin. Oncol. 139 (7) (2013) 1105–1115, https://doi.org/10.1007/ s00432-013-1412-6.
[13] L. Skalniak, M. Dziendziel, J. Jura, MCPIP1 contributes to the toxicity of proteasome inhibitor MG-132 in HeLa cells by the inhibition of NF-κB, Mol. Cell. Biochem. 395 (1–2) (2014) 253–263, https://doi.org/10.1007/s11010-014-2134- z.
[14] L. Kria, A. Ohira, T. Amemiya, Growth factors in cultured pterygium fibroblasts: immunohistochemical and ELISA analysis, Graefes Arch. Clin. Exp. Ophthalmol. 236 (9) (1998) 702–708, https://doi.org/10.1007/s004170050144.
[15] M.Y. Sun, Y.N. Song, M. Zhang, C.Y. Zhang, L.J. Zhang, H. Zhang, Ginsenoside Rg3 inhibits the migration and invasion of liver cancer cells by increasing the protein expression of ARHGAP9, Oncol. Lett. 17 (1) (2019) 965–973, https://doi.org/ 10.3892/ol.2018.9701.
[16] J.H. Jung, I.G. Kang, H.E. Cha, S.H. Choe, S.T. Kim, Effect of Asian sand dust on mucin production in NCI-H292 cells and allergic murine model, Otolaryngol. Head Neck Surg. 146 (6) (2012) 887–894, https://doi.org/10.1177/ 0194599812439011.
[17] Y. Zhou, Y. Zhou, K. Wang, T. Li, M. Zhang, Y. Yang, et al., ROCK2 confers acquired gemcitabine resistance in pancreatic cancer cells by upregulating transcription factor ZEB1, Cancers (Basel) 11 (12) (2019), https://doi.org/10.3390/ cancers11121881.
[18] L. Geng, X. Xu, H. Zhang, C. Chen, Y. Hou, G. Yao, et al., Comprehensive expression profile of long non-coding RNAs in peripheral blood mononuclear cells from patients with neuropsychiatric systemic lupus erythematosus, Annals of Translational Medicine 8 (6) (2020) 349, https://doi.org/10.21037/ atm.2020.03.25.
[19] Z. Feng, W. Zheng, Q. Tang, L. Cheng, H. Li, W. Ni, et al., Fludarabine inhibits STAT1-mediated up-regulation of caspase-3 expression in dexamethasone-induced osteoblasts apoptosis and slows the progression of steroid-induced avascular necrosis of the femoral head in rats, Apoptosis: an International Journal on Programmed Cell Death. 22 (8) (2017) 1001–1012, https://doi.org/10.1007/ s10495-017-1383-1.
[20] S. Ogino, T. Kawasaki, A. Ogawa, G.J. Kirkner, M. Loda, C.S. Fuchs, Cytoplasmic localization of p27 (cyclin-dependent kinase inhibitor 1B/KIP1) in colorectal cancer: inverse correlations with nuclear p27 loss, microsatellite instability, and CpG island methylator phenotype, Hum. Pathol. 38 (4) (2007) 585–592, https:// doi.org/10.1016/j.humpath.2006.09.014.
[21] H. Yoshida, M. Oku, M. Suzuki, K. Mori, pXBP1(U) encoded in AZD5305 XBP1 pre-mRNA negatively regulates unfolded protein response activator pXBP1(S) in mammalian ER stress response, J. Cell Biol. 172 (4) (2006) 565–575, https://doi.org/10.1083/ jcb.200508145.
[22] W. Lu, H. Ning, L. Gu, H. Peng, Q. Wang, R. Hou, et al., MCPIP1 selectively destabilizes transcripts associated with an antiapoptotic gene expression program in breast cancer cells that can elicit complete tumor regression, Cancer Res. 76 (6) (2016) 1429–1440, https://doi.org/10.1158/0008-5472.Can-15-1115.
[23] Y.M. Ning, D. Suzman, V.E. Maher, L. Zhang, S. Tang, T. Ricks, et al., FDA approval summary: atezolizumab for the treatment of patients with progressive advanced urothelial carcinoma after platinum-containing chemotherapy, Oncologist. 22 (6) (2017) 743–749, https://doi.org/10.1634/theoncologist.2017-0087.
[24] A.V. Balar, D. Castellano, P.H. Odonnell, P. Grivas, J. Vuky, T. Powles, et al., Pembrolizumab as First-line Therapy in Cisplatin-ineligible Advanced Urothelial Cancer: Results From the total KEYNOTE-052 Study Population 35, 2017 (284-).
[25] M.T. Perra, C. Maxia, A. Corbu, L. Minerba, P. Demurtas, R. Colombari, et al., Oxidative stress in pterygium: relationship between p53 and 8- hydroxydeoxyguanosine, Mol. Vis. 12 (2006) 1136–1142.
[26] D. Cao, W.K. Chu, T.K. Ng, Y.W.Y. Yip, A.L. Young, C.P. Pang, et al., Cellular proliferation and migration of human pterygium cells: mitomycin versus small- molecule inhibitors, Cornea. 37 (6) (2018) 760–766, https://doi.org/10.1097/ ico.0000000000001569.
[27] K.W. Kim, S.H. Park, S.H. Lee, J.C. Kim, Upregulated stromal cell-derived factor 1 (SDF-1) expression and its interaction with CXCR4 contribute to the pathogenesis of severe pterygia, Invest. Ophthalmol. Vis. Sci. 54 (12) (2013) 7198–7206, https://doi.org/10.1167/iovs.13-13044.
[28] A.S. Tan, J.P.S. Yeong, C.P.T. Lai, C.H.C. Ong, B. Lee, J.C.T. Lim, et al., The role of Ki-67 in Asian triple negative breast cancers: a novel combinatory panel approach, Virchows Arch. 475 (6) (2019) 709–725, https://doi.org/10.1007/s00428-019- 02635-4.
[29] H. Cao, Y. Feng, L. Chen, C. Yu, Lobaplatin inhibits prostate cancer proliferation and migration through regulation of BCL2 and BAX, Dose-Response 17 (2) (2019), https://doi.org/10.1177/1559325819850981, 1559325819850981.
[30] M. Morgan, C. Much, M. DiGiacomo, C. Azzi, I. Ivanova, D.M. Vitsios, et al., mRNA 3′ uridylation and poly(A) tail length sculpt the mammalian maternal transcriptome, Nature. 548 (7667) (2017) 347–351, https://doi.org/10.1038/ nature23318.
[31] E.R. Zuiderweg, L.E. Hightower, J.E. Gestwicki, The remarkable multivalency of the Hsp70 chaperones, Cell Stress Chaperones 22 (2) (2017) 173–189, https://doi. org/10.1007/s12192-017-0776-y.
[32] P.K. Srivastava, Heat shock proteins in immune response to cancer: the fourth paradigm, Experientia. 50 (11− 12) (1994) 1054–1060, https://doi.org/10.1007/ bf01923461.
[33] S. Zhang, J. Wei, J. Li, W. Zhang, L. Yang, C. Xun, Immunohistochemical study of heat shock protein 70 in mouse endometrium during early pregnancy, Chinese Journal of Anatomy 25 (1) (2002) 21–24.
[34] M.E. Murphy, The HSP70 family and cancer, Carcinogenesis. 34 (6) (2013) 1181–1188, https://doi.org/10.1093/carcin/bgt111.
[35] K.Y. Chang, C.T. Huang, T.I. Hsu, C.C. Hsu, J.J. Liu, C.K. Chuang, et al., Stress stimuli induce cancer-stemness gene expression via Sp1 activation leading to therapeutic resistance in glioblastoma, Biochem. Biophys. Res. Commun. 493 (1) (2017) 14–19, https://doi.org/10.1016/j.bbrc.2017.09.095.
[36] A.I. Su, T. Wiltshire, S. Batalov, H. Lapp, K.A. Ching, D. Block, et al., A gene atlas of the mouse and human protein-encoding transcriptomes, Proc. Natl. Acad. Sci. U. S. A. 101 (16) (2004) 6062–6067, https://doi.org/10.1073/pnas.0400782101.
[37] F. Dal Piaz, R. Cotugno, L. Lepore, A. Vassallo, N. Malafronte, G. Lauro, et al., Chemical proteomics reveals HSP70 1A as a target for the anticancer diterpene oridonin in Jurkat cells, J. Proteome 82 (2013) 14–26, https://doi.org/10.1016/j. jprot.2013.01.030.
[38] W. Liu, C. Ding, Roles of LncRNAs in viral infections, Front. Cell. Infect. Microbiol. 7 (2017) 205, https://doi.org/10.3389/fcimb.2017.00205.
[39] F. Yang, Y.H. Liu, S.Y. Dong, R.M. Ma, A. Bhandari, X.H. Zhang, et al., A novel long non-coding RNA FGF14-AS2 is correlated with progression and prognosis in breast cancer, Biochem. Biophys. Res. Commun. 470 (3) (2016) 479–483, https://doi. org/10.1016/j.bbrc.2016.01.147.
[40] G. Qin, Y. Song, Y. Guo, Y. Sun, W. Zeng, LincRNA TINCR facilitates excessive proliferation and inflammation in post-burn skin fibroblasts by directly binding with SND1 protein and inducing SND1-mediated TGF-β1 expression, Biochem. Biophys. Res. Commun. 509 (4) (2019) 903–910, https://doi.org/10.1016/j. bbrc.2019.01.013.
[41] Z. Yang, S. Jiang, J. Shang, Y. Jiang, Y. Dai, B. Xu, et al., LncRNA: shedding light on mechanisms and opportunities in fibrosis and aging, Ageing Res. Rev. 52 (2019) 17–31, https://doi.org/10.1016/j.arr.2019.04.001.
[42] S.M. Saayman, A. Ackley, J. Burdach, M. Clemson, D.C. Gruenert, K. Tachikawa, et al., Long non-coding RNA BGas regulates the cystic fibrosis transmembrane conductance regulator, Molecular Therapy: the Journal of the American Society of Gene Therapy. 24 (8) (2016) 1351–1357, https://doi.org/10.1038/mt.2016.112.
[43] F. Papadopoulou, E. Efthimiou, Thyroid cancer after external or internal ionizing irradiation, Hellenic Journal of Nuclear Medicine. 12 (3) (2009) 266–270.
[44] I. Landa, C. Montero-Conde, D. Malanga, S. De Gisi, G. Pita, L.J. Leandro-García, et al., Allelic variant at − 79 (C>T) in CDKN1B (p27Kip1) confers an increased risk of thyroid cancer and alters mRNA levels, Endocr. Relat. Cancer 17 (2) (2010) 317–328, https://doi.org/10.1677/erc-09-0016.
[45] C. Hetz, E. Chevet, H.P. Harding, Targeting the unfolded protein response in disease, Nat. Rev. Drug Discov. 12 (9) (2013) 703–719, https://doi.org/10.1038/ nrd3976.
[46] A. Boutten, D. Goven, E. Artaud-Macari, M. Bonay, [Protective role of Nrf2 in the lungs against oxidative airway diseases], Medecine Sciences: M/S. 27 (11) (2011) 966–972, https://doi.org/10.1051/medsci/20112711012.
[47] E. Boratyn, I. Nowak, I. Horwacik, M. Durbas, A. Mistarz, M. Kukla, et al., Monocyte chemoattractant protein-induced protein 1 overexpression modulates transcriptome, including microRNA, in human neuroblastoma cells, J. Cell. Biochem. 117 (3) (2016) 694–707, https://doi.org/10.1002/jcb.25354.