The PEMF stimulation system (GHY-III, Airforce Military Medical University (AMMU), Xi’an, China; China Patent no. ZL02224739.4) consisted of a PEMF generator and Helmholtz coils (Figure 1). In cellular experiments, we employed Helmholtz coils with 200 mm diameters and an interval distance of 100 mm. Each coil was composed of 0.8 mm diameter enameled coated copper wire. The PEMF waveform comprised of a pulsed burst (burst width 5 ms; pulse width, 0.2 ms; pulse wait, 0.02 ms; burst wait, 60 ms; pulse rise and fall times: 0.3 µs, 2.0 µs), which was frequented at 15 Hz. This waveform is known to exert a therapeutic effect on nerve injury and osteoporosis, as is evidenced in our prior investigations (Jing et al., 2010; Zhu et al., 2017). A compact 2 Ω resistor was introduced in series connection, accompanying coils. The wave shape and frequency were assessed with an oscilloscope (6,000 series; Agilent Technologies, United States), and a Gaussmeter (Model 455 DSP; Lake Shore Cryotronics) was employed to monitor the precision of PEMF output. NP cells received PEMF stimulation for 4 h/day. In animal experiments, different Helmholtz coils were used, carrying an interval distance of 304, and diameter of 800 mm but the PEMF output was the same in in vitro and in vivo studies, which was confirmed by the oscilloscope and Gaussmeter test.

Our work received ethical approval from the Xijing Hospital, and we received informed consent from all eligible participants. Human intervertebral disc tissue was obtained from patients with lumbar disc herniation (n = 8, age 59.0 ± 0.5 years, range 55–74 years) and fresh NP tissue samples were obtained during the surgery. Samples were washed thrice with normal saline, and cut into 2 mm³, before digestion for 30 min in 0.2% pancreatin. They were then rinsed thrice in PBS and incubated in 0.25% type II collagenase for 4 h, with shaking every half an hour, before terminating digestion. Any undigested tissue was then filtered with a 70-micron filter, before resuspending the digested tissue in DMEM/F12 medium with 10% fetal bovine serum and antibiotics, prior to culture in a 5% CO₂ humidified incubator at 37°C. Cells received PEMF treatment 4 h per day, and samples were collected at 0, 24, 48, and 96 h to test apoptosis rate. In order to examine PEMF-mediated autophagy via SIRT1, cells were pretreated with either 10 mM 3-MA (an autophagy suppressor) or EX-527 (a SIRT1 enzymatic activity suppressor) for 2 h before the first PEMF stimulation and then maintained in the same media for the entire experiment. All experiments were repeated thrice for good measure.

NP cells were seeded to the 96-well plate (density: 5 × 10⁴/ml). In 24, 48, and 96 h timepoint, each well was washed with PBS and then treated using 10-µl of Cell Counting Kit 8 (CCK8; Dojindo, Kumamoto, Japan) for1 h at 37°C. The optical density was then evaluated via a microplate reader (BioTek, United States) at 450 nm absorbance.

Total RNA was extracted from NP cells using the Total RNA Mini Kit (QIAGEN, Germany), following kit directions, then converted to cDNA, prior to qPCR analysis via SYBR Green PCR Master Mix (TAKARA, Japan) and a BioRad CFX96 PCR System (BioRad, Australia). GAPDH served as an endogenous control. The primers used for qRT-PCR reactions were as follows: collagen II (F)5′-CGAGGCAGACAGTACCTTGA-3′, (R) 5′-TGC​TCT​CGA​TCT​GGT​TGT​TC-3’; MMP3 (F) 5′-GCT​CAT​CCT​ACC​CAT​TGC​AT-3′, (R) 5′-GCT​TCC​CTG​TCA​TCT​TCA​GC-3’; GAPDH (F) 5′-GGCACAGTCAAGGCTFAGAATG-3′, (R) 5′-GGT​GGT​GAA​GAC​GCC​AGT​A-3’.

NP cells were thrice rinsed in PBS, before harvest using RIPA buffer and centrifugation at 12,000 g for 15 min. Protein quantification was done with a BCA analysis kit. Proteins were electrophoresized via SDS-PAGE, before transfer to PVDF membranes (Thermo Fisher Scientific). Next, they were blocked using 5% non-fat milk in TBS-Tween (TBST) at room temperature (RT) for 1 h, before exposure to primary antibodies (namely, p62, SIRT1, LC3B, collagen II, MMP-3, and GAPDH (all from abcam and has a dilution of 1:1,000)) at 4°C overnight (O/N). They were next exposed to secondary HRP-linked antibodies, either goat anti-rabbit or goat anti-mouse secondary antibodies (Cwbiotech, Beijing, China) at a 1:2,000 dilution for 1 h, before protein band detection using chemiluminescence (Bio-Rad).

NP cells underwent fixation in 4% paraformaldehyde for 10–15 min, and permeabilization in 0.2% Triton X-100 for 20 min, before blocking in 10% normal goat serum at RT for 1h, and O/N incubation in antibodies against collagen II (1:200) and MMP-3 (1:150) at 4°C. Then, they were thrice washed in PBS before exposure to secondary antibodies (FITC or Cy-3, 1:200) for 1 h at RT, and to DAPI for 10 min. Cells underwent three PBS washes for 15 min after every step. Lastly, they were visualized under a fluorescence microscope and analyzed using the ImageJ software.

A TEM (H-7650; Hitachi, Tokyo, Japan) was employed for calculating autophagic vesicles within NP cells. In short, NP cells underwent O/N fixation in 2.5% glutaraldehyde, followed by a 1 h post-fixation in 2% OsO₄, before receiving gradient dehydration in acetone, followed by embedding, and slicing into 70 nm ultrathin sections using LKB-V ultrathin microtome. Subsequently, they were post-stained with uranyl acetate and lead citrate, and visualized with TEM (H-7650; Hitachi, Tokyo, Japan).

All animal experiments abided by the Institutional Ethical Committee of Fourth Military Medical University guidelines. Fifteen rats were arbitrarily separated into three groups: control (n = 5); needle puncture (n = 5); and needle puncture and PEMF treatment (n = 5). All rats were intraperitoneally anesthetized with pentobarbital solution (1%, 40 mg/kg). As described in previous studies, following tail skin disinfection with iodophor and alcohol, a 20-gauge needle was used to fully puncture the tail disc from Co6-Co8 levels, including two IVDs (Co6-Co7 and Co7-Co8) (Sun et al., 2020; Mei et al., 2021). The needle was then rotated 180 and left for 1 min before removal. Following the operation, all the animals were placed back in their cages and provided with ample food and water. The rats in the experimental group were provided with PEMF treatment (8:00–12:00, 4 h/d) immediately after surgery. Rats were housed in a climate-controlled facility (24 ± 1°C, 35–60% humidity; 12 h light/dark cycle).

MRI images were captured 8 weeks after treatment, using an MRI system (Siemens 3 T Magnetom Trio Tim scanner, Munich, Germany). The IVD degeneration score was computed by two independent spinal surgeons, based on the Pfirrmann grading system, ranging from grade I to grade V.

Upon fixation in 4% paraformaldehyde, the IVD tissue was decalcified, paraffin-embedded, and sliced into 7 μm thick sagittal sections, then stained with Hematoxylin and Eosin (H&E), Safranin O-Fast Green (SOFG), and Sirius Red. The histological scores, including NP morphology and cellularity, AF organization, matrix structure and CEP integrity, were determined by independent researchers who were blinded to the study, as reported previously (Sobajima et al., 2005; Chan et al., 2019).

All data were analyzed using one-way ANOVA from the SPSS 13.0 software (IBM, IL, United States) and expressed as means ± standard deviations (SDs). Tukey’s post hoc and Dunnett’s tests were employed for significance comparisons (GraphPad Prism 7.0; GraphPad Software, CA, United States). p < 0.05 was deemed significant.

To examine the PEMF-mediated effect on NP cell apoptosis, viability and cell morphology, we stimulated NP cells with PEMF 4 h/day for four consecutive days, and monitored cell apoptosis and viability using flow cytometry and CCK-8 assay at 0, 24, 48, and 96 h. Based on our results, PEMF did not exert any obvious cytotoxicity on NP cells (Figures 2A-D and Supplementary Figure 1A). Moreover, phase-contrast microscopy results showed that PEMF didn’t change cell morphology at 96 h timepoint (Supplementary Figure 1B and 1C).

We also assessed the PEMF-mediated regulation of NP cell ECM related genes expression. Following 4 days of treatment, RT-qPCR data revealed that PEMF markedly enhanced collagen II expression, while decreasing Mmp3 levels in NP cells (Figures 2F,G). Results from western blot analysis further corroborated the RT-qPCR data (Figures 2E,H,I). In addition, using immunofluorescence staining and semi-quantitative analysis, we also revealed that PEMF markedly elevated collagen II expression in NP cells (Figures 2J,K). Collectively, these data suggest that PEMF delays the IVD degeneration process in vitro.

To assess whether PEMF regulates NP cell autophagy, we examined LC3B and p62/SQSTM1 levels in NP cells, using western blot. As depicted in Figures 3A–D, after PEMF treatment, the ratio of LC3B-II/LC3B-I increased significantly, whereas, p62 levels dropped significantly. Notably, SIRT1 levels were markedly up-regulated, which is in accordance with autophagic activation. In addition, we evaluated autophagosomes and autophagolysosomes, two key structures associated with autophagy, using TEM. Compared to control cells, the PEMF-treated NP cells displayed more cytoplasmic autophagosomes and autophagolysosomes, thereby confirming autophagic activation in PEMF-treated NP cells (Figures 3E,F).

Next, we examined whether autophagy regulates the PEMF-mediated regulation of ECM related makers expression in NP cells. As illustrated in Figure 4A, our western blot data revealed that PEMF treatment markedly reduced MMP3 levels, while increasing type 2 collagen levels, and this effect was reversed by 3-MA exposure (Figures 4B–E). In addition, we used immunofluorescence to assess type 2 collagen and MMP3 expressions in PEMF treated and un-treated NP cells. Based on our results, type 2 collagen expression increased remarkably, while MMP3 expression decreased after PEMF treatment. Conversely, the autophagy inhibitor 3-MA reversed this phenomenon (Figures 4F–I). Based on these evidences, PEMF regulates degenerative makers expression by inducing autophagic activation.

It is well known that Sirt1 is an upstream regulator of autophagy. Interestingly, we found that PEMF treatment markedly increased Sirt1 expression (Figures 3G,H). We, next, examined the relationship between Sirt1, autophagy, and ECM related genes expression in NP cells. We employed EX-527 as a Sirt1 enzyme activity inhibitor, and revealed that EX-527 had no effect on Sirt1 expression. Based on our results, EX-527 treatment inhibited PEMF-induced autophagy, as evidenced by an increase in the LC3-II/LC3-I ratio and p62 levels in the EX 527 group, compared to the PEMF group (Figures 5A–F). Our TEM results revealed that the number of autophagosomes and autophagolysosomes in the EX-527 treated NP cells was also significantly reduced, compared to the PEMF treated NP cells (Figures 5G,H). Taken together, EX-527 inhibited the PEMF-mediated induction of ECM synthesis in NP cells, as evidenced by the marked decrease in collagen II expression and the simultaneous increase in MMP3 expression. In all, our results suggest that PEMF promotes NP cell ECM synthesis by stimulating the Sirt1-mediated autophagy pathway.

To examine the PEMF-mediated regulation of IVD degeneration in vivo, we established an IVD degeneration model by puncturing the IVD of rat tail. Firstly, we measured the changes in the weight of the rats to monitor whether PEMF would affect other tissues in rats. No significance change was observed in the weight of rats in each group at 2w, 4w and 8w timepoint with or without PEMF stimulation, which indicates that PEMF is safe and has no side effects in rats (Supplementary Figure 2A). We next assessed IVD degeneration via MRI after 8 weeks. Based on our results, the T2-weighted signal intensity of the PEMF rats was stronger, compared to the puncture only rats (Figure 6A). Moreover, the Pfirrmann score grading of IVD degeneration showed that the PEMF treatment group had markedly reduced scores, relative to the puncture-only rats (Figure 6B). Furthermore, using H&E and Safranin-O staining (Figures 6C,D), we revealed that the number of NP tissues in the PEMF treatment group was greater, relative to the puncture-only group, and the integrity of AF and CEP was also significantly better than the puncture-only group. In terms of the histological score, the PEMF-treated rats had markedly reduced scores, compared to the puncture rats (Figure 6F). Additionally, based on the Sirius red staining, the amount of collagen I and collagen III in the PEMF-treated rats were markedly reduced, relative to the puncture rats (Figures 6E,G). Collectively, the above data indicate that PEMF application attenuates the process of IVD degeneration in vivo.

We next examined whether PEMF induces the SIRT1-autophagy network in vivo. Relative to the puncture only group, immunofluorescence staining revealed that LC3B and SIRT1-positive cells were remarkably enhanced in the PEMF treated rats (Figures 7A,B). In addition, based on immunofluorescence staining, Aggrecan and collagen II expressions were also markedly elevated in the PEMF treated rats, relative to the puncture only rats (Figures 6H–J). Taken together, our in vivo investigation confirmed that PEMF preserve healthy ECM, and limit both collagen II and aggrecan degeneration in NP cells via the SIRT1-autophagic network.