Deficiency of MIF Accentuates Overloaded Compression-Induced Nucleus Pulposus Cell Oxidative Damage via Depressing Mitophagy

Established studies proved that mechanical compression loading had multiple effects on the biological behavior of the intervertebral disc (IVD). However, the regulating mechanism involved in this process remains unclear. The current study is aimed at exploring the potential bioregulators and signaling pathways involved in the compression-associated biological changes of nucleus pulposus (NP) cells. Tandem mass tag- (TMT-) based quantitative proteomics was exerted to analyze the differentially expressed proteins (DEPs) and signal pathways among the different groups of NP cells cultured under noncompression, low-compression (LC), and high-compression (HC) loading. Eight potential protective bioregulators for the NP cell survival under different compression loading were predicted by the proteomics, among which macrophage migration inhibitory factor (MIF) and oxidative stress-related pathways were selected for further evaluation, due to its similar function in regulating the fate of the cartilage endplate- (CEP-) derived cells. We found that deficiency of MIF accentuates the accumulation of ROS, mitochondrial dysfunction, and senescence of NP cells under overloaded mechanical compression. The potential molecular mechanism involved in this process is related to the mitophagy regulating role of MIF. Our findings provide a better understanding of the regulatory role of mechanical compression on the cellular fate commitment and matrix metabolism of NP, and the potential strategies for treating disc degenerative diseases via using MIF-regulating agents.


Introduction
Intervertebral disc (IVD) degeneration, which can result in low back pain (LBP), instability, and deformity of the spine, has been recognized as the leading cause of degenerative spine disease [1]. The human IVD consists of three compartments: nucleus pulposus (NP), annulus fibrosus (AF), and cartilage endplate (CEP), which connects the adjacent bony vertebral bodies. NP tissue is a type of gelatinous structure, containing collagen fibrils and proteoglycan molecules [2]. NP cell is responsible for the synthesis of the functional extracellular matrix (ECM) components (mainly aggrecan and collagen type II) in the core region of the disc, which firstly reveals apoptosis-and/or senescence-like changes during the process of IVD degeneration [2]. In the degenerative intervertebral disc, the senescent NP cells accumulate and result in weakened proliferation, compromised self-repair, increased organellar dysfunction, and enhanced breakdown of functional ECM [3,4]. Hence, a potential strategy for treating IVD degeneration is to alleviate or retard the apoptosis-and/or senescence-like changes of NP cells by regulating the potential signaling molecules or pathways involved in the degeneration process.
It is commonly held that the overloaded compressive force applied to the IVD is the leading cause of IVD degeneration [4][5][6]. The prior studies reported that the physiological intensity (0.35-0.75 MPa) of compression loading acted as an anabolic factor for the NP and AF cells, owing to the stimulation of proteoglycan synthesis. In contrast, excessive compression (≥1 MPa) aggravated the catabolic metabolism of proteoglycan of NP and AF cells [7,8]. Noticeably, our previous studies also elucidated the function of graded compression loading in controlling NP cell survival and proved that overloaded mechanical compression markedly exacerbated degenerative changes of the NP [5,6,9,10]. Although the molecular mechanism involved in the compression loadingassociated biological behavior changes of NP cell remains unclear, most studies point that the overloaded compression-induced apoptotic and senescent changes of NP cell are related to the intracellular oxidative damage triggered by mechanosensing factors, such as the integrins and cadherins [5,6,[11][12][13]. Therefore, it is meaningful to explore the potential bioregulators and signaling pathways involved in the compression-associated biological changes of NP cell.
Hence, in the present study, we used the human NP cells isolated from a surgical excisional IVD in a patient who suffered lumbar vertebral fracture (LVF) as the investigated subject. The NP cells, seeding in the methacrylamide-modified gelatin (GelMA) hydrogels, commonly used scaffolds for 3D cell culturing, were cultured in our self-developed compression loading bioreactor under the dynamic graded compression loading for 2 weeks, referring to our previous study [5]. After that, we analyzed the apoptosis-and senescencelike changes of the NP cells and determined 5% deformation of the hydrogel as the low-compression (LC) loading and 20% deformation of the hydrogel as the high-compression (HC) loading for the NP cells. We further used tandem mass tag-(TMT-) based quantitative proteomics to analyze the differentially expressed proteins (DEPs) among the different NP cells cultured under noncompression, LC, and HC loading compression. Analysis of the DEPs, Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment predicted eight potential protective bioregulators for the NP cell survival under overloaded compression, among which macrophage migration inhibitory factor (MIF) and oxidative stress-related pathways were further evaluated in vivo and in vitro, due to its similar function in regulating the fate of the human CEP-derived cells [14]. Moreover, the recombinant lentiviral vector construct containing MIF short hairpin RNA (shRNA) was used to establish the MIF knockout (MIF-KO) NP cells to further determine the protective function of MIF in NP cells that suffered overloaded compression.

Material and Methods
2.1. Source of the Human NP Sample. For the better homogeneity of the samples prepared for the proteomics analysis, the experimental NP samples in this study were obtained from a 32-year-old, female patient who suffered lumbar vertebral fracture (LVF) and admitted to the orthopedics department of our hospital. The patient received the discectomy surgery and donated the excisional IVD tissue for experimental research. The patient declared no documented medical history of LBP before suffering the LVF. The preoperative MRI scans of the patient revealed that the excisional IVD was graded in Pfirrmann II (normal or mild degeneration), according to the Pfirrmann classification system (Supplemental Figure 1) [15]. Informed consent was obtained from the patient, and the researching procedure complied with approval from the ethics committee of the Third Affiliated Hospital of Chongqing Medical University.
2.2. Isolation of the Primary Human NP Cells. The primary human NP cells were isolated as follows [5]. The NP tissue was carefully separated from the excisional IVD and washed by phosphate-buffered saline (PBS) five times in the tissue culture dishes (Jet Biofil, China). Then, the tissue was minced into flocculent pieces and digested in 0.2% type II collagenase (Sigma, USA) for 4 hours, following 0.25% trypsin (Gibco, USA) digestion for 30 minutes. The DMEM/F-12 medium (Gibco, USA) containing 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin/streptomycin (Gibco, USA) was used to neutralize the digesting solution. The collected suspension was centrifuged at 1200 rpm for 8-10 minutes. Finally, the collected NP cells were resuspended by the completed culture medium and seeded into flasks (Jet Biofil, China), incubated at 37°C with 5% CO 2 . Culturing media were replaced every four days.

GelMA Hydrogel Synthesis and Preparation.
In brief, 10% (w/v) of gelatin from porcine skin (Sigma, USA) was dissolved in 1x PBS. The methacrylic anhydride (Aladdin, China) was added to the gelatin solution in a dropwise manner when the solution was heated to 60°C. The reaction was allowed to proceed for 4 hours at 60°C before being stopped using a 5x PBS dilution. The resulting solution was dialyzed against distilled water using a 12-14 kDa dialysis tubing (Thermo Fisher, USA) at 60°C for 5-7 days with two water changes per day. The solution was then stored at −20°C for 24 hours and lyophilized for 72 hours before use.
2.4. Establishment of the 3D NP Cell Culturing Network. The NP cells were encapsulated in GelMA hydrogel for 3D culturing and compression loading. The cell encapsulation procedure was performed according to our previous study [5]. In brief, the cells at passage II were trypsinized with 0.25% trypsin (Gibco, USA) for 2 minutes, and the collected cell suspension was then centrifuged at 1200 rpm for 5 minutes. The centrifuged cell pellets were then mixed with the GelMA hydrogel precursor (1 × 10 7 cells/mL). After that, the mixture suspension was pipette into a self-made mold (diameter = 1 cm, thickness = 0:5 cm) and exposed to the UV light (360 ± 5 nm, 850 mW) for about 1 minute at a distance of 8-10 cm to construct the 3D NP cell culturing network. The cell-encapsulated GelMA hydrogels were incubated in the dishes (Jet Biofil, China) for three days before the compression performing regime in the bioreactor. 2.11. Establishment of the Rat Tail IVD Compressed Model. All animal procedures were performed under the approval of the Animal Care and Use Committee at Chongqing Medical University. Twenty 12-week-old female Sprague-Dawley (SD) rats from 500 ± 5 g were used in the study. Radiographs were taken to confirm the rat tail IVD levels and heights under intraperitoneal anesthesia. Rat tail was then fixed with a self-developed compression loading apparatus between the 8th and 10th caudal vertebrae. An axial force from the distal side was exerted to produce a compression loading of 1.3 MPa, according to the previous study [16]. The rats were then divided into three groups based on the compression loading duration: the sham group-unloaded throughout, the 2 weeks group-loaded for 2 weeks, and the 4 weeks group-loaded for 4 weeks.

Histological Hematoxylin and Eosin (HE) Staining.
The treated samples were harvested and fixed with 4% paraformaldehyde, embedded in paraffin, and then cut into 5 μm per section. Then, the sections were stained with hematoxylin and eosin. The stained sections were observed and scanned under an optical microscope (Olympus, Japan).
2.13. Alcian Blue Staining. Alcian blue staining was used to detect the ECM glycosaminoglycan deposition. Briefly, each tissue section was incubated in 0.2% Alcian blue solution before rinsing with deionized water. Then, the stained sections were mounted and observed under an optical microscope (Olympus, Japan).
2.14. MIF Immunohistochemistry (IHC). After rehydration, tissue sections were blocked by goat serum, treated with 3 Oxidative Medicine and Cellular Longevity hyaluronidase (0.8%) for 20 minutes at 37°C, and then incubated with MIF antibody (1 : 100, Abcam, UK) for 60 minutes. After washing in PBS, biotinylated secondary antibody (1 : 100, Dako, Denmark) was applied for 30 minutes, washed in PSB, and treated with avidin-biotin complex reagents. Colour was developed using 3,3-diaminobenzidine reagents (Dako, Denmark), and the sections were counterstained with Harris's hematoxylin. The average optical density (AOD) of five randomly selected visual fields (per immunohistochemical slice) under high magnification (400x) was measured using the ImageJ analysis system.

NP Cell Transfection.
For knockout of the endogenous MIF in NP cells, shRNA targeting MIF (MIF-shRNA (Santa Cruz Biotechnology, Dallas, TX, U.S.A.)) was transfected into cells by using a recombinant lentiviral vector (Gene-Chem, Shanghai, China). For transfection, cells were seeded into a 6-well plate (Jet Biofil, China), incubated for 24 h, and then transfected according to the manufacturer's instructions. In brief, the cells were transfected with lentivirus (multiplicity of infection = 50) for three days. After determining the efficacy of transfection by a fluorescent microscope, the transfected cells were subcultured and seeded in the hydrogels for the subsequent experiments.

Detection of Intracellular Reactive Oxygen Species (ROS)
Content. Intracellular ROS content was detected by fluorescent DCFH-DA molecular probe staining (Beyotime, China). Briefly, 10 mg DCFH-DA molecular probes were added into a 1 mL basic medium without FBS to prepare the working solution. After that, the cells were incubated in a 6-well plate (Jet Biofil, China) with molecular probe working solution at 37°C for 20 minutes. Then, the cells were collected and the fluorescent intensity was determined by a flow cytometer under 488 nm excitation wavelength and 525 nm emission wavelength.
2.17. Determination of Mitochondrial Membrane Potential (ΔΨm). JC-1 fluorescent molecular probe staining (Beyotime, China) was performed to determine the cell ΔΨm, according to the manufacturer's protocol. In brief, after being washed with PBS thrice, 1 mL basic medium was added to each 6-well plate (Jet Biofil, China). Alive cells were immersed in 1 mL of prepared JC-1 molecular probe working solution and incubated for 30 minutes at 37°C. After being washed with precooled JC-1 1x washing buffer thrice, ΔΨm was detected by laser scanning confocal microscopy (LSCM. Zeiss 780, Germany). The JC-1 molecular probes' fluorescence intensity ratio was then quantified by the ImageJ system.

Colocalization of Mitochondria and Autophagosomes.
The intracellular mitophagy initiation was evaluated by the fluorescent colocalization of mitochondria and autophagosomes. NP cells were transiently transfected with GFP-LC3 (Beyotime, China). MitoTracker Red (Beyotime, China) was used to label the mitochondria in cells, which carries a thiol-reactive chloromethyl group that covalently binds to the reduced thiols that present mitochondrial matrix protein.
The numbers of total colocalizing GFP-LC3 puncta per cell (mitophagosomes) were counted using ImageJ.

Transmission Electron Microscopy (TEM) Observation.
TEM is the most reliable approach for monitoring autophagy [17]. The NP cells were collected and immersed in 0.1 M sodium cacodylate buffer at room temperature for 12 hours. The ultrathin 50 nm sections were cut by using an ultramicrotome, stained with 2% (w/v) uranyl acetate and lead citrate, and then visualized and captured with a TEM machine (Hitachi, Japan).
2.20. Statistical Analysis. All data were analyzed using Graph-Pad Prism (version 6.0, GraphPad Software, USA) and presented as mean ± standard deviation with n = 3. The thresholds of fold change (>1.2 or <0.83) and P value < 0.05 were used to identify DEPs. A two-tailed Student's t -test was used to assess the statistical significance of the measurement data (P value < 0.05).

Effects of Graded Mechanical Compression on Cell
Senescence, Viability, and ECM Synthesis of the Human NP Cells Cultured in a Bioreactor. Our previous studies evaluated that overloaded compressive force applied to IVD could lead to disc degeneration [5,6,9,10], whereas proper physiological pressure is beneficial for maintaining the cell viability of NP [18]. Our previous studies verified that mechanical compression-induced hydrogel deformation could effectively imitate the in vivo loading situation of the NP cell or mesenchymal stem cell (MSC) [5,19]. Thus, we cultured the NP cell-encapsulated GelMA hydrogels and imitated the in vivo compression situation using our self-developed compression loading bioreactor (Figure 1(a)). The SA-β-Gal staining was used to determine the degree of NP cell senescence in each group. The result revealed that the proportion of positive cell was significantly increased in the group that exerted HC loading (≥10% deformation), but not in the LC loading group (≤5% deformation) (Figures 1(b) and 1(c)). The result of the fluorescent Live/Dead staining also indicated that the death ratio of the NP cells was markedly enhanced by over 10% gel deformation (Figures 1(d) and 1(e)). However, LC loading did not increase the death ratio of the NP cells (Figures 1(d) and 1(e)). Flow cytometry was utilized to detect the cell apoptosis rate, and the result also indicated that apoptosis of the NP cell was significantly enhanced by HC loading, but not in the LC loading group (Figures 2(a) and 2(b)). Additionally, western blotting results revealed that the senescence-associated markers (P53, P21, and P16) were strongly overexpressed in NP cell that suffered HC loading, which was consistent with the results of SA-β-Gal staining (Figures 2(c) and 2(d)), whereas the synthesis of functional ECM components (aggrecan and collagen type II) was significantly inhibited by HC loading, but not in the LC loading group (Figures 2(c) and 2(d)). Oxidative Medicine and Cellular Longevity the potential functional factors involved in the progress of compression-associated NP cell survival and senescence, the DEPs of the samples, respectively, that suffered nonloading (control), LC loading (5% deformation), and HC loading (20% deformation) were evaluated by TMT-based proteomics analysis (Figure 3). Among the 324 DEPs, 104 proteins were upregulated and 220 proteins were downregulated in the LC loading group compared with the control group ( Figure 4(a), Supplemental Table 1). There were 427 overexpressed proteins and 377 underexpressed proteins in the HC loading group compared with the control group ( Figure 4(a), Supplemental Table 2). Moreover, there were 1109 overexpressed proteins and 779 underexpressed proteins in the HC loading compression group compared with the LC loading group (Figure 4(a), Supplemental Table 3). Depending on the aforementioned experimental results, we speculated that the intersection of the overexpressed proteins in the LC loading group and the underexpressed proteins in the HC loading group might play as protective factors for NP cell survival. In contrast, the intersection of the underexpressed proteins in the LC loading group and the overexpressed proteins in the HC loading group might play as adverse factors for NP cell survival. Thus, we further analyzed the protein clusters and   Oxidative Medicine and Cellular Longevity predicted 8 potential protective factors and 3 potential adverse factors for NP cell survival (Figures 4(b) and 4(c)). Among the 8 potential protective factors, MIF was selected for further study due to its reported role in regulating the fate of the human CEP cells [14]. In addition, the GO and KEGG enrichment analysis indicated that compression loading affected the fate commitment of NP cell via regulating oxidative stress-associated pathways (ATP biosynthetic process and oxidative phosphorylation) (Figures 4(d) and 4(e)).

Effects of Overloaded Compression on Biological
Behavior and MIF Expression of the Rat Tail NP Tissue. To further verify the relationship between the expression of MIF and compression loading-associated NP degeneration, we conducted a rat tail IVD overloaded compression model ( Figure 5(a)) to detect the classification of IVD degeneration and MIF expression referring to the previous study [16]. The MRI T2-weighted image of the compressed rat tail IVD indicated that the NP tissue that revealed the sign of degeneration Relative expression to -actin   Oxidative Medicine and Cellular Longevity

Deficiency of MIF Accentuates Oxidative Stress-Induced
Senescence of the NP Cell under Overloaded Compression via Depressing Mitophagy. As the major organelle for ATP biosynthetic process and oxidative phosphorylation, the mitochondrion is also the main ROS generating and attacking component of cells [21]. Our previous studies confirmed that mitophagy could alleviate the oxidative stress-induced senescence of NP cell via recycling the injured mitochondria [5,22]. In terms of our prior studies and the present GO and KEGG enrichment analysis, we further observed the mitophagy-associated markers in the current study. The western blotting result indicated that both MIF-KO and treatment of HC loading could downregulate the expression of PINK1 and LC3II/I and upregulate the expression of Parkin and P62, which were symbols of the block of mitophagy (Figures 7(a) and 7(b)). Noticeably, knockout of MIF in HC loading-treated NP cells further exacerbated the restraint of mitophagy-related markers (Figures 7(a) and 7(b)). The JC-1 green/red fluorescence ratio is usually used for evaluating the state of the mitochondrion [5]. The JC-1 staining result indicated that knockout of MIF alone did not cause mitochondrial damage, but HC loading aggravated mitochondrial damage in NP cells (Figures 7(c) and 7(d)). Moreover, the MIF-KO NP cells treated with HC loading had the highest JC-1 green/red fluorescence ratio, indicating that deficiency of MIF exacerbated accumulation of the damaged mitochondria induced by the HC loading (Figures 7(c) and 7(d)). To further investigate the mitophagy at various stages, NP cells were transfected with a plasmid encoding GFP-LC3 and incubated with MitoTracker Red to label autophagosomes and mitochondria, respectively. We found that mitochondria in the control group were filamentous and interconnected as a network. Exerting overloaded compression alone enhanced the formation of GFP-LC3 puncta, and a few were costained with mitochondria, which exhibited as short rods or spheres with the network breakdown (Figures 8(a) and 8(c)). Additionally, MIF-KO decreased the amount of colocalizing GFP-LC3 puncta significantly (Figures 8(a) and 8(c)). Furthermore, we observed the amount and morphology of the mitochondria and autophagosomes in the NP cells via TEM. As is shown in Figure 7(e), mitochondria of the NP cells in control and MIF-KO groups maintained the normal morphology and structure, whereas in the HC loadingtreated NP cells, the mitochondria appeared shrunken with enhanced membrane density (Figure 8(b)). In addition, the HC loading-treated cell exhibited more autophagosomes,

11
Oxidative Medicine and Cellular Longevity surrounding double-membrane mitochondrion-like segments, in its cytoplasm (Figure 8(b)). However, the MIF-KO NP cells treated with HC loading did not appear typical autophagosome in its cytoplasm (Figure 8(b)). Thus, the results above indicated that deficiency of MIF did retard the mitophagy of NP cells.

Discussion
Mass studies evaluated that the compressive force applied to the IVD played as a major regulator for the biological behavior of IVD cells, but the particular molecular mechanisms involved in this regulating process are still controversial. Our prior studies elucidated the biological responses of IVD cells to graded mechanical compression loading via establishing the compression-loaded perfusion culture models of IVD cell-encapsulated hydrogels and tissues [5,6,9,10]. In the present study, we analyzed the apoptosisand senescence-like changes of the human NP cells under graded dynamic compression loading and evaluated that LC loading did not lead to NP cell degenerative changes. In contrast, HC loading did cause cell apoptosis, senescence, and breakdown of ECM. For further exploration of potential regulators and molecular mechanisms involved in this process, we exerted TMT-based proteomics, a popular labeling proteomics method in identifying biomarkers and molecular mechanisms for several diseases.
Depending on the results of our current and prior studies, LC loading compression was beneficial for the NP cell survival and ECM homeostasis, whereas HC loading compression aggravated NP cell senescence, apoptosis, and breakdown of ECM [5,6,19]. Hence, we speculated that there might exist protective bioregulators that overexpressed in NP cells under LC loading compression but lost their functions and underexpressed in NP cells that suffered HC loading compression. Based on this hypothesis, we analyzed the DEPs among the NP cells cultured under noncompression, LC, and HC loading compression and predicted eight proteins, which might play as potential protective factors for the NP cell survival under HC loading compression. Among the eight predicted proteins, MIF, a recognized hypoxic stress-associated regulator of innate immunity, was selected for further functional experiments due to its reported similar role in regulating the fate of the human CEP-derived cells [14]. Then, we tested the expression of MIF in both in vitro NP cell 3D-cultured hydrogels and in vivo rat tail IVD compressed models. The results revealed that the expression of MIF was markedly attenuated by HC loading in vitro and overloaded compression in vivo, which were consistent with the proteomics results.
MIF was first identified as a cytokine released from immune or nonimmune cells, including IVD and other types of cartilage cells, which can inhibit the random migration of macrophages and regulate chondroosteogenesis [14,23,24]. Despite its wide distribution, MIF secretion is closely related to multiple stresses, such as hypoxic, oxidative, and mechanical stresses [23][24][25][26]. MIF is thought as the downstream factor of hypoxia-inducible factor-1 (HIF-1), and the HIF-1 response of the cell is a kind of adaptive response that improves the chances of survival under hypoxic and/or oxidative stress by restoring oxidative homeostasis [25]. In the current study, the analysis of GO and KEGG enrichment also indicated that compression loading affected the fate commitment of NP cells via regulating the ATP biosynthetic process and oxidative phosphorylation. As the prominent organelle for ATP biosynthetic process and oxidative phosphorylation,

12
Oxidative Medicine and Cellular Longevity 13 Oxidative Medicine and Cellular Longevity the mitochondrion is also the main targeting organelle for oxidative stress injury [21]. Noticeably, MIF was reported to play an essential role in regulating many repair/senescence-associated genes via activating autophagy or mitophagy, which is a critical antioxidative stress mechanism to restore intracellular oxygen homeostasis [25,27]. Our previous studies have evaluated that overloaded compression could trigger oxidative stress-associated mitochondrial dysfunction and induce NP cell senescence, and mitophagy could alleviate oxidative stress-induced senescent changes of NP cell that suffered overloaded compression via clearance of damaged mitochondria. Therefore, we hypothesized that there might exist a relationship between compressionassociated mitochondrial damage and mitophagy in NP cell.
The mitochondrion is an essential organelle with multiple functions in the maintenance of intracellular environment homeostasis, not only by regulating ATP production but also by regulating several signal cascades that control cell death, differentiation, and senescence [28]. Mitophagy, a particular form of macroautophagy, can specifically target mitochondria for degradation [29,30]. In this process, a series of specific receptors or adaptors that mediated pathways were activated to recruit the autophagic machinery towards dam-aged mitochondria, engulfed and digested by the autophagosomes [29,30]. The aforementioned results of the current study indicated that HC loading could induce oxidative stress injury and senescence of the human NP cells. Simultaneously, HC loading aggravated mitochondrial dysfunction but retarded activity of mitophagy in NP cells. When we knock down the expression of MIF, the activity of mitophagy was further depressed, which substantially exacerbated the mitochondrial dysfunction, oxidative stress injury, and senescence of the NP cells under overloaded compression. Although mass studies confirmed the mitophagy regulating role of MIF in various types of cells, the specific regulating mechanisms involved in this process were still controversial [27]. The established researches identified and elucidated two signal pathways of mitophagy: ubiquitin-(Ub-) dependent and Ub-independent (receptor-mediated mitophagy) [31]. Some study holds that MIF induces mitophagy via BNIP3-mediated pathway, an Ub-independent manner of mitophagy [32], while some other holds that MIF is the regulator of PINK1/Parkin-mediated mitophagy, the most welldefined Ub-mediated pathway of mitophagy [33]. The results of our prior and current studies elucidated that overload compression-associated mitophagy was related to the  14 Oxidative Medicine and Cellular Longevity PINK1/Parkin-mediated macroautophagy [5]. Noticeably, the other series of studies also evaluated that MIF could protect cardiac cells against overload pressure-induced cardiac hypertrophy via activating Parkin-mediated macroautophagy or mitophagy [26,34,35]. Mitophagy-inducing signal cascades are relatively distinct but interconnected [31]. Importantly, mitophagy is dependent on the accumulation of ROS and damaged mitochondria [36]. Deficiency of mitophagy accumulates dysfunctional mitochondria that release ROS into the cytosol and trigger oxidative stress injury [27,31,36]. Thus, this is because deficiency of MIF accentuates the oxidative stress damage and senescence of NP cells that suffered overloaded mechanical compression. There were some limitations in the current study. First, the experimental NP cells in the current study were obtained from the same donor, due to the homogeneity of samples prepared for the proteomics analysis. The collection of clinical patients' IVD samples for the research is ongoing. We  have planned to carry out more advanced high-throughput sequencing, such as single-cell sequencing in future studies, which can conclude larger sample capacity and diminish the deviation caused by individual differences. Second, the mechanism behind ROS-induced MIF release remains elusive, and the current study simply focused on its

16
Oxidative Medicine and Cellular Longevity autophagy-inducing role involved in the progress of mechanical compression-related disc degeneration. Hence, more researches on the MIF-associated biological changes in the pathomechanism of disc degeneration will be exerted in our further studies.

Conclusion
In conclusion, our study reveals that overloaded mechanical compression could induce oxidative stress, mitochondrial dysfunction, and degenerative changes in human NP cells.
In contrast, proper physiological pressure is beneficial for maintaining cell viability and ECM homeostasis of the NP. Moreover, there exists a relationship between MIF expression and mechanical compression stress-associated biological changes of NP cells. The deficiency of MIF accentuates the accumulation of ROS, mitochondrial dysfunction, and senescence under overloaded mechanical compression. The potential molecular mechanism involved in this process is related to the mitophagy-inducing role of MIF ( Figure 9). This work helps us better understand the regulating function of mechanical compression stress in the progress of disc degeneration and provides more substantial proofs for using MIF-alerting agents in treating disc degenerative diseases.

Data Availability
The data used to support the findings of this study are available from the corresponding authors upon request.

Supplementary Materials
Supplementary Figure 1: preoperative MRI image of the patient's lumbar IVDs and the gross view of the separated NP tissues. Supplementary Table 1: annotation detail of the DEPs between the LC loading group and the control group. Supplementary Table 2: annotation detail of the DEPs between the HC loading group and the control group. Supplementary Table 3: annotation detail of the DEPs between the HC loading group and the LC loading group.