Long Noncoding RNA MIATNB Regulates Hyperosmotic Stress-induced Corneal Epithelial Cell Injury by Inhibiting Autophagy in Dry Eye Disease

Abstract Purpose Dry eye disease (DED) has a complex etiology and the roles of long noncoding RNAs (lncRNAs) in its pathophysiology are not completely understood. Autophagy is a self-eating process important for cell survival and homeostasis. The present study explored the role of myocardial infarction-associated transcript neighbor (MIATNB) long non-coding RNA in hyperosmolarity-induced autophagy and apoptosis in human corneal epithelial cell (HCEC)-based model of dry eye disease. Methods In vitro assays were performed with a human SV40 immortalized corneal epithelial cell line. Different concentrations of NaCl were used to create hyperosmolarity. HCECs were cultured in presence of 70–120 mM NaCl for 24 h to create an in vitro model of dry eye. RT-qPCR was performed to assess the expression of dry eye related LC3B, ATG16L, BECN1, ATG1, ATG7, ATG13, ATG5, ATG10, and ATG101 mRNAs and western blot analysis of LC3B and P62 and RFP -GFP-tagged LC3. Flow cytometry and western blot analysis of caspase 3, BCL2 and BAX were performed to detect apoptosis. Chloroquine (CQ) was used to inhibit autophagy pharmacologically. Results Autophagy flux was activated in HCECs subjected to hyperosmotic stress. Hyperosmolarity activated apoptosis and inhibited HCEC migration and autophagy. Hyperosmolarity upregulated MIATNB expression, while MIATNB knockdown inhibited autophagosome degradation and promoted HCEC apoptosis. Under hyperosmolar conditions, MIATNB knockdown also inhibited the degradation of autophagolysosomes and stimulated HCEC apoptosis. Conclusion MIATNB plays a vital role in dry eye pathogenesis and serves as a bridge between autophagy and apoptosis. Targeting MIATNB for DED treatment should be further evaluated.


Introduction
Dry eye disease (DED) is an ocular surface disease that affects millions of individuals worldwide. with a global prevalence ranging from 5% to 34%. DED etiology is complex and multi-factorial, comprising genetic factors, as well as various stresses to the ocular surface, including environmental factors, infections, endogenous stress, and antigens. 1 Overall, DED is characterized by a harmful cycle of ocular surface damage and inflammation. Through the activation of an ocular surface inflammatory cascade, numerous inflammatory mediators are released in tears, causing damage to ocular surface epithelial cells. Moreover, the hyperosmolarity of tears is a key pathological factor. That is, increased tear film osmotic pressure reduces tear film stability, leading to ocular inflammation. 2 Since tears function as an extracellular matrix for corneal, conjunctival epithelial cells, changes in their composition can activate mitogen-activated protein kinase (MAPK) pathways, NF-jB signaling pathways, inflammatory factors (IL-1a, IL-1b, TNF-a), and matrix metalloproteinase (MMP)-9 in the inflammatory cascade, [3][4][5] among others.
Autophagy is a self-eating cellular process that affects the survival, homeostasis, and development of inflammatory cells, including lymphocytes, neutrophils, and macrophages. 6,7 As a cytoplasmic degradation pathway, autophagy protects cells from exogenous hazards, and clears intracellular aggregates and damaged organelles. Dysregulated autophagy affects the development of diseases with an inflammatory component, including infections, autoimmune disorders, cancer, metabolic disorders, neurodegeneration, as well as cardiovascular and liver diseases. 8,9 Moreover, in a DED mouse model, the number of autophagosomes reportedly increased in the lacrimal gland, 10 while damaged mitochondria, stressed endoplasmic reticulum, and increases in autophagy markers were also observed. In fact, autophagy is activated during DED to prevent further damage to acinar cells and maintain the normal function of the lacrimal glands. 11 Meanwhile, Lyu et al. found that calcitriol inhibits apoptosis via autophagy activation in hyperosmotic stress-stimulated corneal epithelial cells. 12 However, the associated upstream mechanisms responsible for regulating autophagy in DED remain unclear.
Long noncoding RNAs (lncRNAs) are RNA molecules of more than 200 nucleotides that are not translated into proteins; however, they impact transcription and translation. 13 Recent developments in lncRNA research have revealed that these molecules are associated with ophthalmological disorders, such as glaucoma and corneal disease. 14 In particular, the lncRNA myocardial infarction-associated transcript neighbor (MIATNB), whose encoding gene is located next to the neighbor of the myocardial infarction-associated transcript (MIAT) gene, impacts the metabolism and function of N6 methyladenosine (m6A) RNA in breast cancer and is significantly correlated with the expression of PD-L1, an immune checkpoint protein. 15 Moreover, the gene encoding lncRNA MIATNB (designated hereafter as MIATNB) is reportedly a susceptibility gene in patients with genetic fever. 16 Furthermore, we recently found that MIAT can inhibit inflammation, 17 whereas the expression of MIATNB is elevated under hyperosmolarity stress; however, its function in DED remains unknown. Hence, the primary aim of the present study was to characterize the mechanism by which MIATNB impacts DED pathogenesis, primarily in the context of apoptosis and autophagy.

RNA interference
When the cells reached 60%-70% confluence, siRNAs targeting MIATNB (GenePharma, Shanghai, China), or negative control (si-NC), were used to transiently knockdown MIATNB in HCECs using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. The si-MIATNB sequences were as follows: Autophagy flux determination GFP-mRFP-LC3 plasmids were transfected using Lipofectamine to detect autophagic flux or co-transfected with lipo3000 and P3000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Cells were passaged and plated in 35 mm-confocal dish at a density of 5 Â 10 4 cells per well. Cells were subjected to hyperosmolar treatment in presence of 90 mM NaCl 24 h after plasmid transfection. Cells were visualized under a fluorescence microscope after they were washed and supplementing with fresh medium. For imaging, the GFP and mRFP fluorescence were monitored by performing confocal scanning-laser microscopy (Nikon, Tokyo, Japan).

RNA isolation and quantitative real-time PCR
Total RNA was extracted from HCECs using TRIzol Reagent (Vazyme Biotech Co., Ltd., Nanjing, China) as directed by the manufacturer. RNA was reverse-transcribed using HiScript V R III Reverse Transcription SuperMix for qPCR (Vazyme, Nanjing, China), and cDNA was used for RT-qPCR with ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). The relative expression of the target genes was normalized to that of the endogenous control GAPDH using the 2 ÀDDCt method. The primers used in this study are listed in Table 1.

Western blot analysis
Western blot analysis was performed as previously described. 17 In brief, radioimmunoprecipitation assay (Solarbio, Beijing, China) lysis buffer containing 0.1% protease inhibitor was used to lyse cells for 30 min. To determine the protein concentration, the Pierce BCA Protein Assay Kit (Thermo Scientific, Shanghai, China) was used. Next, 20-30 lg protein sample was separated via sodium dodecyl sulfate-polyacrylamide gel electrophoresis with 10%, 12%, or 15% resolving gels and transferred to a polyvinylidene difluoride membrane (PVDF; Millipore, USA). The PVDF  GAPDH was used as an internal control to normalize target protein expression.

Cell viability assay
Cell viability was measured using the Cell Counting Kit-8 (CCK-8) assay. Briefly, 100 lL HCEC suspension was seeded in 96-well plates and allowed to adhere. The seeding desnsity was 5000 cells. Subsequently, the cells were treated with different concentrations of NaCl for 24 h. Next, cell viability was determined using the CCK-8 assay kit (Dojindo, Kumamoto, Japan); 10 lL CCK-8 solution was added to each well of the 96-well plate. After 2 h incubation, optical density was detected at 450 nm using a microplate reader (BioTek, VT Lab, USA).

Transwell assay
HCECs were re-suspended in serum-free DMEM and plated into the upper chamber of a Transwell (8 mm pore size; Corning, Beijing, China). Then, 600 lL DMEM containing 10% FBS was added to the lower chamber. The seeding density was 3 Â 10 4 cells/chamber. The cells were washed with phosphate buffer solution (PBS) three times after a 24h treatment under different conditions. Cells that did not invade the upper layer were wiped with cotton swabs, fixed with 4% paraformaldehyde for 1 h, and stained with 0.5% crystal violet solution. The number of purple-stained cells was counted under an inverted microscope; three high power fields were randomly selected for counting cells from each sample, and Image J was used to quantify the data. The experiments were repeated three times.

Flow cytometry
To evaluate HCEC apoptosis, flow cytometry was used. The seeding desnsity was 2 Â 10 4 cells. In brief, cells were stained using the fluorescein isothiocyanate annexin V-PI kit in 1Â binding buffer (Vazyme, Nanjing, Jiangsu, China). The rate of HCEC apoptosis was analyzed using the FlowJo10.6.2 system according to the manufacturer's protocol.

Statistical analysis
SPSS23.0 (Chicago, USA) and GraphPad Prism 8 (California, USA) software were used to analyze the data. All data are shown as the mean ± standard deviation based on at least three replicates. A t-test was used to compare data between two groups, and ANOVA was used to compare measurement data among more than three groups. p < .05 was considered statistically significant.

Autophagy flux is activated under hyperosmolarity in human corneal epithelial cells (HCECs)
LC3, a light-chain protein, is an autophagy marker and primarily functions in the formation of autophagosomes. LC3 precursor molecules are cleaved by ATG4B to form cytoplasmic LC3-I, which is then activated by APG7L/ATG7, transferred to ATG3, and coupled with lipoylethanolamine to form membrane-bound form LC3-II, which can attach to the membranes of autophagosomes, thereby forming the structural protein. 19 To determine whether autophagy flux is activated via hyperosmolarity, we quantified the mRNA expression of LC3B, Beclin1 (BECN1), ATG5, ATG7, ATG10, ATG13, ATG16L, and ATG101, and found them all to be markedly increased following 24 h exposure to hyperosmolar conditions (Figure 1(A) and Supplementary Figure 1). We then quantified the abundance of LC3B and SQSTM1 (P62) proteins and found that LC3B-II was markedly increased, while P62 abundance was decreased following exposure to hyperosmolar conditions, particularly in the presence of 90 mM NaCl (Figure 1(B-D)). Thus, we inferred that the increase in LC3B-II level might have been caused by an increase in autophagosomes following autophagy activation or failure to clear autophagolysosomes. We next transfected the tandem monomeric RFP-GFPtagged LC3 (mRFP-GFP-LC3) reporter to assess autophagy flux ( Figure 1(E,F)). The autolysosomes exhibited red fluorescence with mRFP-GFP-LC3 owing to GFP sensitivity to acidic pH. We further observed a four-fold increase in red puncta within HCECs after 24 h exposure to 90 mM NaCl, indicating the presence of autolysosomes. Collectively, these results demonstrated that exposure to hyperosmolar conditions induced autophagy in HCECs.

Hyperosmolarity activates apoptosis and inhibits HCEC migration
The viability of HCECs was assessed using the CCK-8 assay. Decreased viability was observed in cells exposed to hyperosmolar conditions (Figure 2(A)). Next, to determine whether hyperosmolarity induces apoptosis, we quantified the abundance of apoptosis regulators BCL2 and BAX (Figure 2(B-E)). The BCL2/BAX ratio decreased after exposure to different concentrations of NaCl. Taken together, these findings indicate that hyperosmolar conditions induce apoptosis in HCECs. These results are consistent with those reported earlier. 17 Transwell assay results revealed that the migratory ability of HCECs decreased in a concentration-dependent manner when they were subjected to different osmolarity conditions (0, 70, 90, and 120 mM) for 24 h (Figure 2(F,G)). Hence, increased NaCl concentration caused a gradual decrease in cell viability and migration.
Hyperosmolarity upregulates MIATNB expression, while MIATNB knockdown inhibits autophagosome degradation MIATNB mRNA levels were markedly higher in HCECs subjected to hyperosmolar conditions, particularly after treatment with 90 mM or 120 mM NaCl, than those observed in controls (Figure 3(A)). To explore the effects of MIATNB knockdown on autophagy, we transfected HCECs with MIATNB-specific small interfering RNA (siRNA), while si-NC served as the control. The cells were subsequently exposed to hyperosmolar conditions, and quantitative realtime PCR (RT-qPCR) was performed. The relative expression of MIATNB in the si-MIATNB group decreased by 40% compared to the si-NC group (Figure 3(B)). Moreover, the expression of ATG16L and ATG101 mRNA significantly increased and decreased, respectively, following MIATNB knockdown and exposure to hyperosmolar conditions. Meanwhile, the expression of LC3B, ATG5, BECN1, ATG7,   ATG10, ATG13, and ATG101 mRNA remained unchanged (Figure 3(C) and Supplementary Figure 2).
We then performed western blot analysis to quantify the abundance of LC3B-II, P62, ATG5, and Beclin1 proteins after MIATNB knockdown. The abundance of LC3B-II increased significantly compared to the si-NC control group, whereas that of P62 decreased. However, the abundance of ATG5 and Beclin1 was not affected (Figure 3(D,E)).
To further confirm the effect of MIATNB on autophagy, we assessed autophagy flux using chloroquine (CQ). In acidic lysosomes, CQ raises the pH, thereby effectively inactivating the acidic lysosomes and inhibiting the fusion and degradation of intracellular autophagic lysosomes. 20,21 The abundance of LC3B-II protein did not change after MIATNB knockdown with CQ treatment (Figure 3(F,G)), indicating that MIATNB knockdown impacts autophagosome degradation rather than conversion of LC3B-I to LC3B-II. This result was confirmed with mRFP-GFP-LC3 assays, wherein images showed a one and a half fold increase in yellow puncta within the MIATNB-knockdown group relative to the control group (Figure 3(H,I)). Hence, MIATNB knockdown inhibited autophagosome degradation.

MIATNB knockdown promotes apoptosis
To determine the effect of MIATNB knockdown on apoptosis, we assessed its effect on BCL2 and BAX mRNA and protein levels. The BCL2/BAX mRNA and protein ratio decreased following MIATNB knockdown (Figure 4(A,C)). In contrast, the abundance of caspase 3 protein increased after MIATNB knockdown (Figure 4(B)). These results were supported by the annexin-PI assay, which showed that the number of late apoptotic cells (28.0%) increased compared to the si-NC group (20.3%; Figure 4(D)).
Finally, the migratory ability of HCECs was assessed using Transwell assays. Results show that the number of migrated cells markedly decreased following MIATNB knockdown (Figure 4(E,F)).
Collectively, these findings suggest that MIATNB knockdown promotes HCEC apoptosis and diminishes their migration capacity.

MIATNB knockdown under hyperosmolar conditions inhibits the degradation of autophagolysosomes and stimulates apoptosis
Exposing HCECs to hyperosmolar conditions (90 mM and 100 mM NaCl) led to notable increases in the mRNA expression of LC3B, ATG5, Beclin 1, ATG13, ATG16L, and ATG101 and LC3B protein content in culture supernatants. It also caused an obvious decrease in the protein expression of P62. We chose 90 mM NaCl to emulate hyperosmolar conditions and study apoptosis based on our previous research. 17 We next sought to understand the autophagyand apoptosis-related changes that occur following MIATNB knockdown in HCECs under hyperosmolar conditions. First, we verified that the knockdown efficiency of MIATNB in HCECs exceeded 50% under hyperosmolarity ( Figure 5(A)). We then employed RT-qPCR to assess the expression of LC3B, ATG16L, BECN1, ATG1, ATG7, ATG13, ATG5, and ATG10 mRNA. Results show that the expression levels of ATG5, BECN1, and ATG101 decreased under hyperosmolar conditions following MIATNB knockdown ( Figure 5(B)). Meanwhile, the abundance of LC3BII alone was slightly decreased under these conditions ( Figure 5(C,D)). We then used mRFP-GFP-LC3 to verify autophagic flux and found that the number of red puncta decreased while that of yellow puncta increased two times after MIATNB knockdown under a hyperosmolar environment ( Figure 5(F,H)). Hence, MIATNB knockdown can inhibit autophagolysosome degradation in a hyperosmolar environment.
To examine the effect on apoptosis, we assessed the expression of BCL2, BAX, and caspase 3. Caspase 3 abundance increased while the BCL2/BAX ratio decreased after MIATNB knockdown in HCECs under hyperosmolar conditions (Figure 5(E,G,I)). Hence, MIATNB knockdown promotes apoptosis in a hyperosmolar environment.

CQ promotes HCEC apoptosis and impairs their migration
To determine whether autophagy inhibition influences apoptosis, we confirmed that the BCL2/BAX protein ratio was decreased after CQ administration under hyperosmolar conditions ( Figure 6(A,B)). Moreover, the CCK-8 assay was employed to analyze the rate of cell proliferation after treating HCECs with CQ under hyperosmolar conditions. The proliferation rate was found to increase after the CQ application ( Figure 6(C)). Meanwhile, annexin-PI staining revealed that the rates of late-stage apoptosis were 13.9% and 17.4% in the dimethyl sulfoxide (DMSO) and CQ groups, respectively ( Figure 5(D)). Hence, inhibiting autophagy may have stimulated apoptosis.

Discussion
DED is accompanied by tear film hyperosmolarity and ocular surface inflammation, which is one of the most important pathogenic mechanisms of DED characterized by the infiltration of inflammatory mediators and destruction of the ocular surface epithelium, which directly affects corneal epithelial cells. In response to hyperosmotic stress, autophagy increases due to altered osmolarity on the ocular surface. 12 In the present study, MIATNB expression increased in HCECs subjected to hyperosmotic stress, thereby protecting them from apoptosis via autophagy regulation.
Autophagy, a process of self-phagocytosis, enables cells to maintain intracellular homeostasis. Autophagy consists of several steps, including phagocytic vesicle formation, autophagosome formation, fusion of autophagosomes and lysosomes to form autophagosomes, and degradation of autophagic lysosomes. 22 Autophagy flux is a dynamic process in the cell and interruption of any step in the autophagy pathway prevents process completion. Generally, autophagy activity depends on the loss of P62, an autophagy substrate, and formation of an autophagosome-LC3B. As a result, the expression of P62 protein decreases, while that of LC3B increases upon autophagy activation. 19 Craig et al. concluded that evaporation-induced tear hyperosmosis is the mechanism underlying DED that directly damages the ocular surface and causes inflammation. 23 However, in absence of hyperosmolarity, MIATNB knockdown induces an increase in LC3B-II expression. Regardless of whether this gene was knocked down, the abundance of LC3B-II protein was not altered following CQ treatment. It could thus be inferred that MIATNB knockdown inhibits autophagosome degradation rather than promoting the conversion of LC3B-I to LC3B-II, which was confirmed by fluorescence results (Figure 3). In HCECs treated with CQ after MIATNB knockdown, higher LC3B-II protein levels were observed than in HCECs treated with DMSO. Therefore, MIATNB incompletely inhibited autophagosome degradation. In addition, P62 expression decreases upon interruption of autophagic flow. Hence, a compensatory decrease may occur in the number of autophagosomes and autophagolysosomes. Alternatively, a lag may occur in the soluble and insoluble P62 levels following fluctuation in autophagic flux. When autophagy flux is modulated, the level of LC3B protein changes more rapidly, whereas the autophagy substrate P62 takes longer to degrade. However, hyperosmolarity stress caused LC3B-II protein levels to decrease following MIATNB knockdown. This may have resulted from an increased rate of autophagosome degradation following exposure to a hyperosmolar environment. The functions and regulatory network of apoptosis, the main form of programmed cell death, have been characterized. However, apoptosis is not the only process that determines cell fate. More recently, autophagy, known as type II programmed cell death, has been shown to co-regulate cell death with apoptosis. On the one hand, autophagy inhibits apoptosis, 24 thus functioning as a pro-survival pathway; on the other hand, it can induce cell death or act in concert with apoptosis. 25,26 Indeed, the crosstalk between autophagy and apoptosis allows for better regulation of the execution process. 27 BCL-2 family proteins play a key regulatory role in linking apoptosis and autophagy. 27 In the present study, we found that inhibiting autophagy with CQ promoted apoptosis. Similarly, MIATNB knockdown inhibited autophagy and stimulated apoptosis. This suggests that MIATNB protects HCECs from hyperosmolarity stress via regulating autophagy ( Figure 7).
As the concentration of NaCl increased, the degree of autophagy in HCECs remained relatively stable. Hence, we speculated that under hyperosmolarity stress, autophagy increased to counter ocular surface inflammation, necrotic cells, and damaged proteins. However, this increase was limited and not always dose-related in response to increase in NaCl concentration. Furthermore, limited autophagy could not maintain HCEC homeostasis due to impaired removal of damaged organelles and necrotic components through the lysosomal pathway. This will be investigated in more detail in our subsequent study. In the DED model, autophagy is limited and insufficient to combat pathological injury. Therefore, we next evaluated whether DED prognosis can be improved by regulating autophagy, and especially whether corneal autophagy can be induced through autophagy activators to reduce corneal inflammation and alleviate DED pathology. Panigrahi et al. reported that autophagy activation could alleviate HCEC inflammation and dry eye in a mouse model. 28 Upon autophagy induction in HCECs with autophagy activators, the number of autophagosomes increased, thus improving the clearance of damaged or necrotic organelles from the eye surface to alleviate apoptosis. 28 Meanwhile, in our study, the application of the autophagy inhibitor CQ promoted apoptosis reflected by the small number of autophagosomes that were sufficient to clear the necrotic material. It is, therefore, necessary to discern the optimal level of autophagy in the context of DED.
LncRNAs are critical regulators of various cellular processes, and their expression is closely associated with the development of various diseases, particularly immune-related diseases. 14 In our previous study, we reported that lncRNA MIAT is a critical mediator of HCEC pyroptosis and apoptosis. 17 Meanwhile, MIATNB gene is a "neighboring" gene of MIAT. Although "neighbor" genes are now ignored, they were once considered the most important combination of genes in evolution. In fact, the "crosstalk" between neighboring genes may be mediated by various mechanisms and cisregulatory signals, including RNA splice sites. FSTL1 and the gene encoding microRNA-198 are examples of this, as the gene encoding microRNA-198 resides at the 3 0 UTR of FSTL1. MicroRNA-198 inhibits skin recovery, while FSTL1 promotes damage recovery. Meanwhile, skin damage results in the inhibition of microRNA-198 and induction of FSTL1 to initiate skin repair. 29 In our present and previous studies, we found that MIAT and MIATNB knockdown promotes apoptosis. It is possible that the gene encoding MIATNB regulates its neighboring gene via cis-regulatory actions. Moreover, MIATNB might represent a new potential target for DED treatment.
However, the mechanism underlying mutual regulation between MIAT and MIATNB remains unclear and warrants further investigation. Additionally, the present study is limited by its focus on in vitro model of DED, and did not include an in vivo animal model to confirm the observed results.

Conclusion
Our study provides novel insight to support the accumulating evidence that MIATNB is a critical mediator of HCEC autophagy and apoptosis. Functionally, we discovered that MIATNB might promote the degradation of autophagolysosomes to inhibit apoptosis. These findings can inform the development of a new therapeutic target for DED.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The authors were sponsored by the Natural Science Foundation of Shandong (ZR2019MH115) and the Clinical Medicine þ X Research Project of the Affiliated Hospital of Qingdao University (QDFY þ X202101044). The sponsors or funding organizations had no role in the design or conduct of this research.  Figure 6 mentioned that CQ stimulated apoptosis.

Data availability statement
All data will be freely available to any researcher wishing to use them for non-commercial purposes, without breaching participant confidentiality.