If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
a R.T., T.Y., and Y.T. contributed equally to this work.
T. Yurube
Correspondence
Address correspondence and reprint requests to: T. Yurube, Department of Orthopaedic Surgery, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe, 650-0017, Japan. Tel: 81-78-382-5985; Fax: 81-78-351-6944.
In the largest avascular low-nutrient intervertebral disc, resident cells would utilize autophagy, a stress-response survival mechanism by self-digestion and recycling wastes. Our goal was to elucidate the involvement of autophagy in disc homeostasis through RNA interference of autophagy-related gene 5 (Atg5).
Design
In vitro, small interfering RNAs (siRNAs) targeting autophagy-essential Atg5 were transfected into rat disc cells. Cell viability with levels of autophagy including Atg5 expression, apoptosis, and senescence was assessed under serum starvation and/or pro-inflammatory interleukin-1 beta (IL-1β) stimulation. In vivo, time-course autophagic flux was monitored following Alexa Fluor® 555-labeled Atg5-siRNA injection into rat tail discs. Furthermore, 24-h temporary static compression-induced disruption of Atg5 siRNA-injected discs was observed by radiography, histomorphology, and immunofluorescence.
Results
In disc cells, three different Atg5 siRNAs consistently suppressed autophagy with Atg5 protein knockdown (mean 44.4% [95% confidence interval: −51.7, −37.1], 51.5% [−80.5, −22.5], 62.3% [−96.6, −28.2]). Then, Atg5 knockdown reduced cell viability through apoptosis and senescence not in serum-supplemented medium (93.6% [−0.8, 21.4]) but in serum-deprived medium (66.4% [−29.8, −8.6]) further with IL-1β (44.5% [−36.9, −23.5]). In disc tissues, immunofluorescence detected intradiscal signals for the labeled siRNA even at 56-d post-injection. Immunoblotting found 56-d autophagy suppression with prolonged Atg5 knockdown (33.2% [−52.8, −5.3]). With compression, Atg5 siRNA-injected discs presented radiographic height loss ([−43.9, −0.8]), histological damage ([−5.5, −0.2]), and immunofluorescent apoptosis ([2.2, 22.2]) and senescence ([4.1, 19.9]) induction compared to control siRNA-injected discs at 56 d.
Conclusions
This loss-of-function study suggests Atg5-dependent autophagy-mediated anti-apoptosis and anti-senescence. Autophagy could be a molecular therapeutic target for degenerative disc disease.
. The intervertebral disc is the largest avascular low-nutrient organ in the body, comprising the central nucleus pulposus (NP), surrounding annulus fibrosus (AF), and cartilaginous endplates
The human lumbar intervertebral disc: evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration.
The human lumbar intervertebral disc: evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration.
. Therefore, there are possible links between apoptosis/senescence and notochordal cell disappearance/chondrocyte-like cell proliferation in the pathogenesis of disc degeneration.
Autophagy in the intervertebral disc has received increasing attention
. Autophagy, an intracellular clearance mechanism in which cells degrade and recycle their proteins and organelles, is involved in maintaining cell homeostasis under stressors including starvation and inflammation
. Under typical physiological conditions, Atg protein expressions are relatively low and basal autophagy maintains cell homeostasis by decomposing waste products
. Meanwhile, under stress conditions, Atg proteins are activated, resulting in the formation and maturation of autophagosomes from phagophores, which capture damaged organelles, misfolded proteins, and impaired mitochondria in induced autophagy
. The light chain 3 (LC3), localized in the autophagosome membrane, is converted from cytosolic LC3-I to phosphatidylethanolamine-conjugated LC3-II by autophagy induction. Therefore, LC3-II is used as a marker of autophagosome formation and maturation
Phosphotyrosine-independent binding of a 62-kDa protein to the src homology 2 (SH2) domain of p56lck and its regulation by phosphorylation of Ser-59 in the lck unique N-terminal region.
. The p62/SQSTM1 and p62/SQSTM1-bound polyubiquitinated proteins become incorporated into completed autophagosomes and are degraded in autolysosomes, which inversely correlate with autophagosome degradation
Phosphotyrosine-independent binding of a 62-kDa protein to the src homology 2 (SH2) domain of p56lck and its regulation by phosphorylation of Ser-59 in the lck unique N-terminal region.
Fig. 1Schematic illustration of disc cellular autophagy and in-vitro and in-vivo experimental regimens. (A) Schematic illustration of autophagic flux. Under stress conditions, mTOR is suppressed, which initiates autophagy through the activation of autophagy-related genes. The Atg5 is an autophagy-essential factor for phagophore elongation and autophagosome formation, which is driven by LC3-II. The p62/SQSTM1 and p62/SQSTM1-bound polyubiquitinated proteins become incorporated into the completed autophagosome. It fuses with the lysosome to form the autolysosome, where the enclosed contents are degraded and recycled. To suppress autophagy, post-transcriptional gene silencing using siRNAs against Atg5 was used. (B) Schematic illustration of the in-vitro study design and sample number. First passage, ∼80%-confluent, monolayer disc NP cells from 12-week-old male rats (n = 20) were reverse transfected with siRNAs targeting Atg5 or non-specific control in DMEM with 10% FBS under 2% O2 for 24 h and analyzed by WB for the phenotype, Atg5 knockdown, and autophagy (n = 4); by cell count and CCK-8 for cell viability after 24-h serum deprivation in DMEM with 0% or 10% FBS (n = 4); by cell count, CCK-8 for cell viability, WB for matrix anabolism, apoptosis, and senescence, staining for apoptosis and senescence, and immunofluorescence for autophagy, apoptosis, and senescence after 24-h serum withdrawal and pro-inflammatory stimulation in DMEM with 0% FBS and 10-ng/ml IL-1β (each n = 4). (C) Schematic illustration of the in-vivo siRNA transfection study design and sample number. In rat tails (n = 18), control siRNA was injected using a 33-gauge needle into C8–C9 and C11–C12 discs, while Alexa Fluor® 555-labeled Atg5 siRNA was injected into C9–C10 and C12–C13 discs. At 2–56 d, tissues of C8–C9 control siRNA-injected and C9–C10 Atg5 siRNA-injected discs and C11–C12 control siRNA-injected and C12–C13 Atg5 siRNA-injected discs were acquired and analyzed by immunofluorescence for Alexa Fluor® 555-labeled Atg5-siRNA transfection and WB for Atg5 knockdown and autophagy, respectively (n = 6/time point). (D) Schematic illustration of the in-vivo temporary static compression study design and sample number. In rat tails (n = 24), an Ilizarov-type apparatus with springs was affixed between C8 and C10. Control siRNA was injected into C8–C9 and C11–C12 discs, whereas Atg5 siRNA was injected into C9–C10 and C12–C13 discs. Then, 1.3-MPa axial force was applied to C8–C9 and C9–C10 discs for 24 h and subsequently released. At 0–56 d, tissues of C8–C9 control siRNA-injected loaded, C9–C10 Atg5 siRNA-injected loaded, C11–C12 control siRNA-injected unloaded, and C12–C13 Atg5 siRNA-injected unloaded discs were acquired following radiography for the height and analyzed by histomorphological safranin-O staining and immunofluorescence for autophagy, apoptosis, and senescence (n = 6/time point).
, autophagy involvement in intervertebral disc degeneration has not been fully clarified. Pharmacological inhibition would be simple to elucidate the autophagy function; however, its uncontaminated intracellular signaling blockade is difficult, leading to possible confounding effects
. Because of the disc's anatomical, biochemical, and nutritional characteristics, we hypothesized that autophagy would serve for the maintenance of disc homeostasis. To clarify the autophagy involvement, we designed an in-vitro and in-vivo loss-of-function study of autophagy using the RNA interference (RNAi) technique against autophagy-essential Atg5 in rat disc cells and tissues.
Materials and methods
Ethics statement
All experimental procedures were performed in accordance with the Institutional Animal Care and Use Committee (P140609 and P190709) at Kobe University Graduate School of Medicine.
Sixty-two 12-week-old male Sprague–Dawley rats (mean 476.6 g [95% confidence interval (CI): 471.7, 481.5]) purchased from CLEA Japan (Tokyo, Japan) were randomly applied to in-vitro and in-vivo experiments without exclusions. Coccygeal (C) discs from 20 rats (475.4 g [464.2, 486.5]) were dissected into the NP and AF after euthanasia. Rat disc NP tissues were digested in 1% penicillin/streptomycin-supplemented Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) and 0.114% collagenase type 2 for 1 h at 37°C. Isolated disc NP cells were grown to ∼80% confluence as a monolayer in DMEM with 10% FBS under 2% O2 at 37°C
Selective interference of mTORC1/RAPTOR protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism with Akt and autophagy induction.
Pharmacological inhibition of mTORC1 but not mTORC2 protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism through Akt and autophagy induction.
Inhibition of autophagy at different stages by ATG5 knockdown and chloroquine supplementation enhances consistent human disc cellular apoptosis and senescence induction rather than extracellular matrix catabolism.
The density of randomly distributed seeding cells (n = 20) was 5.0 × 103/well (96-well plate) for viability, 1.5 × 105/well (6-well plate) for protein extraction, and 1.2 × 104/well (8-well chamber) for staining. In respective experiments, 4 cell samples from 4 different animals (each n = 4) were tested based on not a priori sample-size calculation but literature
Selective interference of mTORC1/RAPTOR protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism with Akt and autophagy induction.
Pharmacological inhibition of mTORC1 but not mTORC2 protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism through Akt and autophagy induction.
Inhibition of autophagy at different stages by ATG5 knockdown and chloroquine supplementation enhances consistent human disc cellular apoptosis and senescence induction rather than extracellular matrix catabolism.
After 24-h RNAi treatment in 10% FBS-supplemented DMEM, cells were applied to Western blotting (WB) for disc NP phenotype, Atg5 knockdown, and autophagy. Additionally, to simulate clinically relevant disease conditions of serum deprivation and/or inflammation, cells were cultured for 24 h in DMEM with 10% FBS, with serum-free 0% FBS, or with 0% FBS and 10-ng/ml interleukin-1 beta (IL-1β), a pro-inflammatory cytokine linked to the pathogenesis and severity of disc degeneration
. Then, cell viability was assessed using the Cell Counting Kit-8 (CCK-8). Cell count was performed. Immunoblotting for matrix anabolism, apoptosis, and senescence, apoptotic terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, senescence-associated beta-galactosidase (SA-β-gal) staining, and immunofluorescence for autophagy, apoptosis, and senescence were conducted [Fig. 1(B)].
Animals and surgical procedures
Forty-two rats (477.1 g [471.3, 482.9]) were used. To confirm in-vivo transfection in rat tails (n = 18), control and Alexa Fluor® 555-labeled Atg5 small interfering RNA (siRNA)–Invivofectamine™ 3.0 reagent complexes were injected into C8–C9 and C11–C12 discs and C9–C10 and C12–C13 discs, respectively. Under general anesthesia, 2-μl solution was injected using a 33-gauge needle at the disc center through a 5-mm longitudinal skin incision
. At 2–56 d, C8–C9 control siRNA-injected and C9–C10 Atg5 siRNA-injected disc and C11–C12 control siRNA-injected and C12–C13 Atg5 siRNA-injected disc tissues were acquired and analyzed by immunofluorescence for Alexa Fluor® 555-labeled Atg5-siRNA transfection and WB for Atg5 knockdown and autophagy, respectively [Fig. 1(C)].
A rat tail temporary static compression model was used to reproduce mechanical stress-induced disc degenerative changes
Rat tail static compression model mimics extracellular matrix metabolic imbalances of matrix metalloproteinases, aggrecanases, and tissue inhibitors of metalloproteinases in intervertebral disc degeneration.
A rat tail temporary static compression model reproduces different stages of intervertebral disc degeneration with decreased notochordal cell phenotype.
. Under general anesthesia, an Ilizarov-type apparatus with springs was attached between C8 and C10 vertebrae of rat tails (n = 24). Control and Atg5 siRNA–Invivofectamine™ 3.0 reagent complexes were injected into C8–C9 and C11–C12 discs and C9–C10 and C12–C13 discs, respectively. Then, 1.3-MPa axial force, corresponding to a disc loading force produced by lifting a moderate weight in the human lumbar spine
, was applied to C8–C9 and C9–C10 discs for 24 h and subsequently released. At 0–56 d, C8–C9 control siRNA-injected loaded, C9–C10 Atg5 siRNA-injected loaded, C11–C12 control siRNA-injected unloaded, and C12–C13 Atg5 siRNA-injected unloaded disc tissues were acquired after radiographic imaging for the height and analyzed by histomorphological safranin-O staining and immunofluorescence for autophagy, apoptosis, and senescence [Fig. 1(D)].
Sample size (n = 6/time point) was based on literature
Rat tail static compression model mimics extracellular matrix metabolic imbalances of matrix metalloproteinases, aggrecanases, and tissue inhibitors of metalloproteinases in intervertebral disc degeneration.
A rat tail temporary static compression model reproduces different stages of intervertebral disc degeneration with decreased notochordal cell phenotype.
but not a priori sample-size calculation. Rats were fed separately in a specific pathogen-free housing cage with freely available food and water. The room had a controlled 12-h light/dark cycle, temperature (23 ± 2°C), and humidity (55 ± 5%). Humane endpoints, e.g., ≥20% weight loss and behavioral changes, were determined.
RNAi
The RNAi was performed using small interfering RNAs (siRNAs) to knockdown Atg5 with the reverse transfection method, allowing high transfection efficiency
. In vitro, three different Atg5 siRNAs [Supplemental Table 2] were used to exclude off-target effects. A non-targeting siRNA was used as a negative control. Cells in 10% FBS-supplemented DMEM were added to each siRNA with Lipofectamine™ RNAiMAX transfection reagent diluted in Opti-Minimal Essential Medium I and then cultured for 24 h. Applied amounts of siRNAs were 60 (6-well plate), 4.8 (8-well chamber), and 2 (96-well plate) pmol/well.
In vivo, Atg5 or control siRNA with Invivofectamine™ 3.0 reagent (final 1.5-pmol/l concentration) was prepared. The Alexa Fluor® 555-labeled Atg5 siRNA was used to assess successful tissue transfection.
Cell viability assay
In vitro, cell viability was assessed by CCK-8 dehydrogenase activity, the absorbance of which (450 nm) was measured using the Model 680 microplate reader. In addition, images were photographed with the BZ-X700 microscope. The number of adherent cells was counted in duplicated four random low-power fields ( × 100) (LPFs) using the ImageJ software (https://imagej.nih.gov/ij/).
Protein extraction, sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis (PAGE), and WB
Cells were scraped off on ice in 3-(N-morpholino)propanesulfonic acid buffer containing protease and phosphatase inhibitors. Harvested tissues were homogenized using the MS-100R bead-beating disrupter for 30 s twice at 4°C in the T-PER™ tissue protein extraction reagent with protease and phosphatase inhibitors. Soluble proteins were collected after 20,000-×g centrifugation for 15 min at 4°C. Samples were stored at −80°C. Protein concentration was determined by the bicinchoninic acid assay.
Equal 30-μg amounts of protein were mixed with the SDS–PAGE sample buffer, boiled for 5 min, and resolved on a 7.5–15.0% polyacrylamide gel. Separated proteins in the tris(hydroxymethyl)aminomethane–glycine–SDS buffer system were transblotted to a polyvinylidene difluoride membrane and probed with primary antibodies for 12 h at 4°C (1:200–1:1,000 dilution) followed by secondary antibodies (1:400 dilution) for 1 h at room temperature. Signals were visualized by enhanced chemiluminescence. Images were obtained using the Chemilumino analyzer LAS-3000 mini. Band intensity was quantified using ImageJ.
In vitro and in vivo, WB was designed to analyze intracellular expression of disc NP notochord-related brachyury and CD24
The human lumbar intervertebral disc: evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration.
in total cell or tissue protein extracts. Protein expression was normalized to loading control tubulin and shown as the relative percentage of control.
Paraffin-embedded tissue preparation
In vivo, functional rat caudal spinal units (vertebral body–disc–vertebral body) were obtained after euthanasia, 1-d fixed en-bloc with 4% paraformaldehyde, 7-d decalcified in 10% ethylenediaminetetraacetic acid, embedded in paraffin, and cut mid-sagittal into 7-μm sections for histomorphology and immunofluorescence.
TUNEL staining
In vitro, cells were fixed with 4% paraformaldehyde for 10 min and applied to fluorescein-labeled TUNEL staining for apoptotic fragmented DNA detection with 4’,6-diamidino-2-phenylindole (DAPI) for counterstaining
. In vivo, apoptotic TUNEL positivity in tissue sections was examined. Under the BZ-X700 microscope and ImageJ, the percentage of TUNEL-positive cells was calculated relative to the total number of DAPI-positive cells in duplicated four random LPFs.
SA-β-gal staining
In vitro, senescent cells were identified by cytochemical SA-β-gal staining at pH 6.0
. The percentage of SA-β-gal-positive cells was calculated in duplicated four random LPFs.
Immunofluorescence
In vitro and in vivo, multi-color immunofluorescence was performed to understand disc cellular relationship between autophagy, apoptosis, and senescence. After antigen-retrieval, permeabilization, and blocking, fixed cells and disc tissue sections were incubated with autophagic Atg5
for 12 h at 4°C and subsequently with Alexa Fluor® 568 and 647 secondary antibodies (1:200 dilution) and DAPI for 1 h at room temperature. The percentage of Atg5-positive cells, TUNEL-positive cells, and p16/INK4a-positive cells was calculated relative to DAPI-positive cells in duplicated four random LPFs.
In vivo, immunofluorescence for Alexa Fluor® 555-labeled Atg5 siRNA and DAPI was performed to disclose the efficacy and working period of intradiscal transfection. The Alexa Fluor® 555 signals are detectable only under the presence of Atg5 siRNAs incorporated in cells.
Radiography
In vivo, lateral radiographs were taken using a VPX-30E system and IXFR film (exposure time, 40 s; distance, 40 cm; current, 3 mA; voltage, 35 kV). Disc height was measured using ImageJ twice at a 1-week interval by each of two investigators blinded to this study, normalized to adjacent vertebral body heights as the disc height index (DHI), shown as the percent of preoperative DHI (%DHI = [postoperative DHI/preoperative DHI] × 100), and further normalized to the intact disc as the normalized %DHI (normalized %DHI = [experimental %DHI/intact %DHI] × 100)
A novel rabbit model of mild, reproducible disc degeneration by an anulus needle puncture: correlation between the degree of disc injury and radiological and histological appearances of disc degeneration.
In vivo, safranin-O, fast green, and hematoxylin staining was performed to demonstrate disc tissue morphological disruption. Histopathological grade, from 0 (non-degenerated) to 16 (severely degenerated) for NP morphology, NP cellularity, NP–AF border, AF morphology, and endplate
Development of a standardized histopathology scoring system for intervertebral disc degeneration in rat models: an initiative of the ORS spine section.
, was assessed in duplicate twice at a 1-week interval by each of two blinded investigators, and the median scores were used for evaluation.
Statistical analysis
Data are presented as the mean [95% CI: lower limit, upper limit] in the text and dot and box plots in the graphs. With the normality assumption, one-way repeated measured analysis of variance (ANOVA), two-way ANOVA, and three-way ANOVA with the Tukey–Kramer post-hoc test were used to assess effects of “treatment”, “experimental condition”, and “time” on in-vitro cell viability, cell count, WB, TUNEL staining, SA-β-gal staining, and immunofluorescence, on in-vivo WB, and on in-vivo radiography, histomorphology, and immunofluorescence, respectively. The intra-class correlation coefficient was calculated to determine the intra-observer and inter-observer reliability for radiography and histology. Statistical analysis was performed using IBM SPSS Statistics 23.0 (IBM, Armonk, NY).
Results
In-vitro Atg5 knockdown suppresses autophagy and promotes apoptosis and senescence in rat disc NP cells
First, notochordal marker expression was evaluated to validate rat disc NP phenotype. In WB, protein extracts from tested samples all showed positive expression of notochord-related brachyury and CD24 [Fig. 2(A)]. Then, we assessed whether Atg5 RNAi using three different siRNAs could effectively knockdown the corresponding protein. Significant decrease in Atg5 expression was seen following every Atg5-siRNA transfection (sequence 1, mean 44.4% knockdown [95% CI: −51.7, −37.1]; sequence 2, 51.5% knockdown [−80.5, −22.5]; sequence 3, 62.3% knockdown [−96.6, −28.2]) [Fig. 2(A)]. Furthermore, we assessed Atg5-RNAi effects on autophagy. Immunoblotting demonstrated that Atg5-siRNA treatment decreased LC3-II (sequence 1, 62.1% [−64.5, −11.3]; sequence 2, 64.0% [−62.6, −9.5]; sequence 3, 48.3% [−78.3, −25.2]) and increased p62/SQSTM1 (sequence 1, 170.5% [33.7, 107.3]; sequence 2, 169.4% [32.6, 106.2]; sequence 3, 188.0% [51.2, 124.8]) [Fig. 2(A)], both indicating successful Atg5 knockdown-mediated autophagy suppression. In subsequent experiments, the sequence-3 Atg5 siRNA with the highest knockdown efficiency was used.
Fig. 2In-vitro Atg5 knockdown suppresses autophagy and promotes apoptosis and senescence in rat disc NP cells. (A) WB for phenotypic brachyury and CD24, autophagic Atg5, LC3, and p62/SQSTM1, and loading control tubulin of total protein extracts from rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h. Changes in relative protein expression of Atg5, LC3-II, and p62/SQSTM1 normalized to tubulin are shown. (B) Changes in CCK-8-based viability, counting number, and morphological appearance of rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h followed by in DMEM with 10% FBS, with 0% FBS, or with 0% FBS and 10-ng/ml IL-1β for 24 h. Cell count was performed in respective four random LPFs of duplicates. (C) WB for phenotypic brachyury and CD24, anabolic aggrecan and Col2a1, apoptotic PARP, cleaved PARP, and cleaved caspase-9, senescent p53, p21/CIP1, and p16/INK4a, and loading control tubulin of total protein extracts from rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h followed by in DMEM with 0% FBS and 10-ng/ml IL-1β for 24 h. Changes in relative protein expression of aggrecan, Col2a1, PARP, cleaved PARP, cleaved caspase-9, p53, p21/CIP1, and p16/INK4a normalized to tubulin are shown. In (A), (B), and (C), data are presented with dot and box plots (n = 4). One-way repeated measures ANOVA with the Tukey–Kramer post-hoc test was used. Immunoblots and cellular images shown are representative of experiments with similar results.
Second, to understand Atg5 RNAi-modified disc cellular physiology, we assessed cell viability by CCK-8 and counted cell number. These were not significantly different between the groups treated by control and Atg5 siRNAs in 10% FBS-supplemented DMEM (CCK-8, 100.0% vs 93.6% [−0.8, 21.4]; cell number, 132.3/LPF vs 127.3/LPF [−16.5, 6.5]) but significantly decreased with Atg5 knockdown under autophagy-requiring conditions in serum-free DMEM (CCK-8, 77.4% vs 66.4% [−29.8, −8.6]; cell number, 84.8/LPF vs 71.8/LPF [−24.5, −1.5]) and further with IL-1β supplementation (CCK-8, 56.7% vs 44.5% [−36.9, −23.5]; cell number, 49.5/LPF vs 33.0/LPF [−28.0, −5.0]), indicating Atg5 knockdown-dependent and serum withdrawal and inflammation-dependent decrease in cell number and viability [Fig. 2(B)].
Third, we assessed Atg5-RNAi effects on disc cellular matrix anabolism, apoptosis, and senescence. Pro-inflammatory IL-1β stimulation resulted in downregulation of anabolic aggrecan ([−49.8, −16.6]) and Col2a1 ([−53.8, −27.6]) protein expression, which was further enhanced by Atg5 RNAi (Col2a1, [−26.1, −0.3]). Then, IL-1β stimulation induced significant downregulation of apoptosis-related PARP ([−31.4, −7.5]) and significant upregulation of apoptotic cleaved PARP ([17.3, 69.2]) and cleaved caspase-9 ([12.5, 52.6]) and senescent p21/CIP1 ([19.7, 103.1]) and p16/INK4a ([11.5, 34.6]) expression. Furthermore, Atg5 knockdown additionally increased IL-1β-induced markers of apoptosis and senescence compared with Atg5 siRNA-treated IL-1β-unstimulated cells (PARP, [−39.7, −15.7]; cleaved PARP, [12.3, 64.3]; cleaved caspase-9, [22.3, 62.4]; p53, [38.1, 156.8]; p21/CIP1, [24.2, 107.6]; p16/INK4a, [14.8, 37.9]) and control siRNA-treated IL-1β-stimulated cells (cleaved PARP, [6.2, 58.1]; cleaved caspase-9, [0.5, 40.6]; p53, [10.1, 128.8]; p16/INK4a, [5.7, 28.8]) [Fig. 2(C)]. These data suggest accelerated apoptosis and senescence by Atg5 knockdown under stressful serum deprivation and inflammation.
In-vitro Atg5 knockdown increases the incidence of apoptosis and senescence in rat disc NP cells
In-vitro Atg5-RNAi effects on disc cellular apoptosis and senescence were further evaluated by TUNEL staining, SA-β-gal staining, and multi-color immunofluorescence. The percentage of TUNEL-positive cells increased following IL-1β stimulation ([95% CI: 15.8, 33.2]), amplified by Atg5 RNAi ([0.4, 22.9]) [Fig. 3(A)]. Also, SA-β-gal-positive cell percentage increased with IL-1β supplementation ([13.3, 29.5]), exaggerated by Atg5 RNAi ([1.6, 17.8]) [Fig. 3(B)]. Then, Atg5 RNAi significantly decreased immunopositivity for Atg5 in IL-1β-unstimulated ([−62.0, −40.7]) and IL-1β-stimulated ([−57.2, −35.9]) cells, which contrastingly increased the percentage of TUNEL-positive ([1.1, 29.2]) and p16/INK4a-positive ([0.1, 15.8]) cells under IL-1β stimulation. However, immunopositivity for apoptotic TUNEL and senescent p16/INK4a remained overlapped partially (3.7–17.0%) [Fig. 3(C)]. These results support promoted apoptosis and senescence by Atg5 knockdown.
Fig. 3In-vitro Atg5 knockdown increases the incidence of apoptosis and senescence in rat disc NP cells. (A) Immunofluorescence for apoptotic TUNEL (green), nuclear DAPI (blue), and merged signals of rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h followed by in DMEM with 0% FBS and 10-ng/ml IL-1β for 24 h. Changes in the percentage of TUNEL-positive cells relative to DAPI-positive cells are shown. (B) SA-β-gal staining of rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h followed by in DMEM with 0% FBS and 10-ng/ml IL-1β for 24 h. Changes in the percentage of SA-β-gal-positive cells relative to total cells are shown. (C) Immunofluorescence for autophagic Atg5 (red), apoptotic TUNEL (green), senescent p16/INK4a (purple), nuclear DAPI (blue), and merged signals of rat disc NP cells after Atg5 or control siRNA transfection in DMEM with 10% FBS for 24 h followed by in DMEM with 0% FBS and 10-ng/ml IL-1β for 24 h. Changes in the percentage of Atg5-positive, TUNEL-positive, p16/INK4a-positive, and TUNEL and p16/INK4a-co-positive cells relative to DAPI-positive cells are shown. In (A), (B), and (C), cell count was performed in respective four random LPFs of duplicates. Data are presented with dot and box plots (n = 4). One-way repeated measures ANOVA with the Tukey–Kramer post-hoc test was used. Immunofluorescent and cytochemical images shown are representative of experiments with similar results.
In-vivo Atg5 knockdown facilitates prolonged suppression of autophagy in rat disc NP tissues
Based on the in-vitro disc cytotoxic effects of Atg5 knockdown, in-vivo intradiscal gene-silencing experiments using Atg5 RNAi were designed. All rats underwent surgery well and gained body weight throughout the experiment (mean 584.7 g, [95% CI: 577.6, 591.9] at 56 d). All springs maintained their compressive length and fully recovered after release, indicating sustained axial loading. There were no signs of infection, skin necrosis, or neurological problems.
To corroborate local siRNA introduction into rat tail discs and uptake by NP cells, Alexa Fluor® 555-labeled Atg5 siRNA was used. Successful administration was confirmed by radiography upon injecting a contrast agent [Fig. 4(A)]. Technically in this system, only siRNAs incorporated into the cytoplasm are detectable. Immunofluorescence showed the red signal for labeled Atg5 siRNA in the disc NP-cell cytoplasm 2–56 d after injection, supporting successful Atg5-siRNA transfection into disc NP cells and a long-term 56-d maintenance [Fig. 4(B)]. Immunoblotting displayed sustained Atg5 protein downregulation in Atg5 siRNA-injected discs (2 d, 45.0% knockdown [−66.3, −19.9]; 28 d, 36.9% knockdown [−56.4, −10.0]; 56 d, 33.2% knockdown [−52.8, −5.3]) [Fig. 4(C)]. Furthermore, Atg5 RNAi-mediated prolonged autophagy suppression was observed with decreased LC3-II (2 d, 61.4% [−59.0, −14.6]; 28 d, 64.4% [−54.1, −9.8]; 56 d, 60.6% [−56.9, −12.5]) and increased p62/SQSTM1 (2 d, 139.1% [16.9 65.0]; 28 d, 131.9% [11.5, 59.6]; 56 d, 124.2% [4.8, 52.9]) [Fig. 4(C)]. These findings indicate extended intradiscal autophagy suppression after Atg5 knockdown.
Fig. 4In-vivo Atg5 knockdown facilitates prolonged suppression of autophagy in rat disc NP tissues. (A) Confirmation with radiography of successful intradiscal 33-gauge needle insertion and 2-μl contrast injection. (B) Immunofluorescence for Alexa Fluor® 555-labeled autophagic Atg5 siRNA (red), nuclear DAPI (blue), and merged signals of rat tail discs at 2, 28, and 56 d after Alexa Fluor® 555-labeled Atg5 or control siRNA injection. White rectangles indicate the disc NP area shown as the enlarged merged images. (C) WB for autophagic Atg5, LC3, and p62/SQSTM1 and loading control tubulin of total protein extracts from rat tail disc NP tissues at 2, 28, and 56 d after Atg5 or control siRNA injection. Changes in relative protein expression of Atg5, LC3-II, and p62/SQSTM1 normalized to tubulin are shown. In (C), data are presented with dot and box plots (n = 6). Two-way ANOVA with the Tukey–Kramer post-hoc test was used. Immunofluorescent images and immunoblots shown are representative of experiments with similar results.
In-vivo Atg5 knockdown advances radiographic and histomorphological disc disruption under temporary static compression in rats
In-vivo Atg5-RNAi effects on disc tissue disruption were evaluated under temporary static compression. Radiographic %DHI analysis exhibited the intra-observer reliability of 0.949–0.978 and inter-observer reliability of 0.901, indicating an acceptable reproducibility. In Atg5 siRNA-injected loaded discs, %DHI was significantly lower compared with Atg5 siRNA-injected unloaded discs (7 d, [95% CI: −23.1, −5.3]; 28 d, [−41.5, −15.5]; 56 d, [−65.4, −20.6]) and control siRNA-injected loaded discs (7 d, [−19.3, −1.4]; 28 d, [−36.3, −10.2]; 56 d, [−43.9, −0.8]), demonstrating the progressive loss between 7 d ([−49.8, −26.0]) and 28 d ([−33.6, −9.7]) vs 56 d [Fig. 5(A)]. These findings suggest an involvement of Atg5-dependent autophagy in maintaining disc height against mechanical loading.
Fig. 5In-vivo Atg5 knockdown advances radiographic and histomorphological disc disruption under temporary static compression in rats. (A) Lateral radiographs of rat tail C11–C12 control siRNA-injected unloaded (white arrow), C12–C13 Atg5 siRNA-injected unloaded (light-gray arrow), C8–C9 control siRNA-injected loaded (intermediate-gray arrow), and C9–C10 Atg5 siRNA-injected loaded (dark-gray arrow) discs at 0, 7, 28, and 56 d under temporary static compression at 1.3 MPa for 24 h. Changes in the height of control siRNA-injected unloaded, Atg5 siRNA-injected unloaded, control siRNA-injected loaded, and Atg5 siRNA-injected loaded discs are shown. (B) Safranin-O staining of rat tail C11–C12 control siRNA-injected unloaded, C12–C13 Atg5 siRNA-injected unloaded, C8–C9 control siRNA-injected loaded, and C9–C10 Atg5 siRNA-injected loaded discs at 0, 7, 28, and 56 d under temporary static compression at 1.3 MPa for 24 h. Changes in the histopathological grade of control siRNA-injected unloaded, Atg5 siRNA-injected unloaded, control siRNA-injected loaded, and Atg5 siRNA-injected loaded discs are shown. In (A) and (B), data are presented with dot and box plots (n = 6). Three-way ANOVA with the Tukey–Kramer post-hoc test was used. Radiographs and histomorphological images shown are representative of experiments with similar results.
Histological grading illustrated the intra-observer reliability of 0.943–0.968 and inter-observer reliability of 0.916, suggesting an appropriate reproducibility. Safranin-O staining presented significantly higher degenerative scores in Atg5 siRNA-injected loaded discs than Atg5 siRNA-injected unloaded discs (28 d, [−7.7, −2.6]; 56 d, [−10.3, −5.0]) and control siRNA-injected loaded discs (56 d, [−5.5, −0.2]). While Atg5 siRNA-injected unloaded discs showed no significant histopathological changes to control siRNA-injected unloaded discs, NP-cell clustering, matrix condensation, and AF disorganization were remarkable in Atg5 siRNA-injected loaded discs at 56 d [Fig. 5(B), Supplemental Table 3], suggesting a contribution of Atg5-dependent autophagy to disc tissue maintenance under mechanical stress.
In-vivo Atg5 knockdown enhances immunofluorescent apoptosis and senescence under temporary static compression in rats
We finally performed multi-color immunofluorescence for autophagic Atg5, apoptotic TUNEL, and senescent p16/INK4a in this model. Immunopositivity for Atg5 was significantly lower in Atg5 siRNA-injected discs than control siRNA-injected ones under unloaded (7 d, [95% CI: −56.6, −28.4]; 28 d, [−46.3, −33.1]; 56 d, [−58.2, −16.8]) and loaded conditions (7 d, [−55.6, −30.3]; 28 d, [−61.1, −33.5]; 56 d, [−56.7, −18.9]) In Atg5 siRNA-injected loaded discs, the percentage of TUNEL-positive (7 d, [27.8, 45.8]; 28 d, [34.4, 54.9]; 56 d, [27.0, 45.9]) and p16/INK4a-positive (7 d, [23.9, 39.1]; 28 d, [35.3, 50.6]; 56 d, [27.6, 43.5]) cells was both significantly higher than Atg5 siRNA-injected unloaded discs and control siRNA-injected loaded discs at 56 d (TUNEL, [2.2, 22.2]; p16/INK4a, [4.1, 19.9]). In Atg5 siRNA-injected unloaded discs, the percentage of TUNEL-positive and p16/INK4a-positive cells was not significantly different from control siRNA-injected unloaded discs. Similar to the in-vitro finding, in-vivo co-immunopositivity for apoptotic TUNEL and senescent p16/INK4a was relatively limited (7 d, 1.6–9.8%; 28 d, 2.2–15.0%; 56 d, 3.7–17.0%) [Fig. 6]. Collectively, the observed findings support anti-apoptosis and anti-senescence as a consistent role of Atg5-dependent autophagy in intervertebral disc homeostasis.
Fig. 6In-vivo Atg5 knockdown enhances immunofluorescent apoptosis and senescence under temporary static compression in rats. Immunofluorescence for autophagic Atg5 (red), apoptotic TUNEL (green), senescent p16/INK4a (purple), nuclear DAPI (blue), and merged signals of rat tail C11–C12 control siRNA-injected unloaded, C12–C13 Atg5 siRNA-injected unloaded, C8–C9 control siRNA-injected loaded, and C9–C10 Atg5 siRNA-injected loaded discs at 0, 7, 28, and 56 d under temporary static compression at 1.3 MPa for 24 h. Changes in the percentage of Atg5-positive, TUNEL-positive, p16/INK4a-positive, and TUNEL and p16/INK4a-co-positive cells relative to DAPI-positive cells of control siRNA-injected unloaded, Atg5 siRNA-injected unloaded, control siRNA-injected loaded, and Atg5 siRNA-injected loaded discs are shown. Cell count was performed in respective four random LPFs of duplicates. Data are presented with dot and box plots (n = 6). Three-way ANOVA with the Tukey–Kramer post-hoc test was used. Immunofluorescent images shown are representative of experiments with similar results.
This is the first study showing progressive disc degenerative changes induced by in-vitro and in-vivo Atg5 RNAi-mediated autophagy suppression. In vitro, successful autophagy suppression was confirmed by introducing siRNA targeting Atg5, an essential autophagy factor, into rat disc NP cells. This is consistent with human disc NP-cell findings, undergoing Atg5 downregulation
Inhibition of autophagy at different stages by ATG5 knockdown and chloroquine supplementation enhances consistent human disc cellular apoptosis and senescence induction rather than extracellular matrix catabolism.
. Following Atg5 RNAi, apoptosis and senescence rates increased in cells under serum deprivation, further promoting after pro-inflammatory IL-1β stimulation. These anti-apoptotic and anti-senescent findings of Atg5-dependent autophagy in rat disc NP cells are consistent with human articular chondrocytes exhibiting enhanced apoptosis and senescence by Atg5 RNAi
Selective interference of mTORC1/RAPTOR protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism with Akt and autophagy induction.
Pharmacological inhibition of mTORC1 but not mTORC2 protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism through Akt and autophagy induction.
, in human disc NP cells. Selective mTORC1 inhibition activated autophagy with anti-apoptosis and anti-senescence. However, Akt, located upstream of mTORC1 to promote cell survival by suppressing apoptosis
. Therefore, mTORC1 inhibition-mediated cytoprotection cannot completely exclude Akt activation effects. This study shows that specific autophagy suppression through Atg5 knockdown increases apoptosis and senescence in rat disc NP cells, suggesting an anti-apoptotic and anti-senescent contribution of Atg5-dependent autophagy to the homeostasis maintenance.
In vivo, Atg5 siRNA was found within the disc NP through 56-d post-injection. The observed knockdown efficiency of Atg5 siRNA locally administered in rat disc NP tissues is comparable with that in other tissues, e.g., brain, eye, and skin (10–75%)
. Furthermore, 56-d intradiscal autophagy suppression was confirmed, suggesting a long-term stability and capability of siRNA to block intradiscal protein expression without being metabolized. In other organs, local siRNA administration effects are maintained as short as 1–16-d
. This notable feature could be due to the anatomical (avascular, encapsulated) and biological (highly differentiated, slow proliferated) characteristics of the disc
. Thus, we regard siRNAs as a potent tool for local gene therapy in disc disease. Since high effectiveness and long-term efficacy are advantages of gene therapy
, single administration might provide sufficient therapeutic effects in future clinical practice.
The observed mechanically induced radiographic, histomorphological, and immunofluorescent disc disruption was further accelerated by Atg5 knockdown in this animal model
Rat tail static compression model mimics extracellular matrix metabolic imbalances of matrix metalloproteinases, aggrecanases, and tissue inhibitors of metalloproteinases in intervertebral disc degeneration.
A rat tail temporary static compression model reproduces different stages of intervertebral disc degeneration with decreased notochordal cell phenotype.
. The lowest disc height, highest degeneration grade, and most prominent apoptosis and senescence induction were observed in Atg5 siRNA-injected loaded discs. Excessive disc NP-cell apoptosis and senescence can lead to disc degeneration through various signaling pathways in response to such stressors as aging, inflammation, nutrient deprivation, and mechanical overloading
Inhibition of autophagy at different stages by ATG5 knockdown and chloroquine supplementation enhances consistent human disc cellular apoptosis and senescence induction rather than extracellular matrix catabolism.
Recombinant human SIRT1 protects against nutrient deprivation-induced mitochondrial apoptosis through autophagy induction in human intervertebral disc nucleus pulposus cells.
autophagy. Thus, roles of autophagy in disc degeneration are still controversial. The present autophagy-inhibition study through Atg5 RNAi, consistently presenting the in-vitro and in-vivo progression of disc degenerative changes with enhanced apoptosis and senescence, suggests the involvement and importance of Atg5-dependent autophagy in disc cell and tissue homeostasis, consequently the primary role of which is to protect against apoptosis and senescence.
This study has several limitations. The disc disruption mechanism in the used rat tail model is not exactly the same as that in human aging. Rats retain disc-NP notochordal cells throughout their lives unlike adult humans, which could differently respond to mechanical loading
. The applied 24-h 1.3-MPa temporary static compression is not physiological, although demonstrating subsequent disc degenerative changes with downregulated aggrecan and Col2a1 gene expression
A rat tail temporary static compression model reproduces different stages of intervertebral disc degeneration with decreased notochordal cell phenotype.
. Then, this study did not evaluate autophagy induction and/or Atg gene overexpression. Further investigation of autophagy activation is warranted to clarify the therapeutic potential for disc degeneration.
In conclusion, the in-vitro and in-vivo rat intervertebral disc RNAi-mediated loss-of-function study of Atg5 suggests that autophagy maintains homeostasis through cytoprotective effects against apoptosis and senescence. Autophagy could be a new molecular therapeutic target for degenerative disc disease.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
The authors have no competing interests to declare.
Role of the funding source
This work was supported by the Japan Society for the Promotion of Science KAKENHI Grant Numbers JP26893151, JP15H03033, JP15K10406, JP16K20051, and JP21K09323 and a Grant of Japan Orthopaedics and Traumatology Research Foundation, Inc. Number 312. The study sponsors had no involvement in this study.
Acknowledgements
The authors thank Mses. Kyoko Tanaka, Maya Yasuda, and Minako Nagata (Department of Orthopaedic Surgery, Kobe University Graduate School of Medicine, Kobe, Japan) for their technical assistance. We also thank Joseph Iacona, Ph.D. from Edanz Group (https://en-author-services.edanz.com/ac) for editing a draft of this manuscript.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
The human lumbar intervertebral disc: evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration.
Phosphotyrosine-independent binding of a 62-kDa protein to the src homology 2 (SH2) domain of p56lck and its regulation by phosphorylation of Ser-59 in the lck unique N-terminal region.
Selective interference of mTORC1/RAPTOR protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism with Akt and autophagy induction.
Pharmacological inhibition of mTORC1 but not mTORC2 protects against human disc cellular apoptosis, senescence, and extracellular matrix catabolism through Akt and autophagy induction.
Inhibition of autophagy at different stages by ATG5 knockdown and chloroquine supplementation enhances consistent human disc cellular apoptosis and senescence induction rather than extracellular matrix catabolism.
Rat tail static compression model mimics extracellular matrix metabolic imbalances of matrix metalloproteinases, aggrecanases, and tissue inhibitors of metalloproteinases in intervertebral disc degeneration.
A rat tail temporary static compression model reproduces different stages of intervertebral disc degeneration with decreased notochordal cell phenotype.
A novel rabbit model of mild, reproducible disc degeneration by an anulus needle puncture: correlation between the degree of disc injury and radiological and histological appearances of disc degeneration.
Development of a standardized histopathology scoring system for intervertebral disc degeneration in rat models: an initiative of the ORS spine section.
Recombinant human SIRT1 protects against nutrient deprivation-induced mitochondrial apoptosis through autophagy induction in human intervertebral disc nucleus pulposus cells.
01005-0&id=gr1.jpg){kind=link}
01005-0&id=gr2.jpg){kind=link}
01005-0&id=gr3.jpg){kind=link}
01005-0&id=gr4.jpg){kind=link}
01005-0&id=gr5.jpg){kind=link}
01005-0&id=gr6.jpg){kind=link}