Chaperone mediated autophagy modulates microglia polarization and inflammation via LAMP2A in ischemia induced spinal cord injury

Abstract

Spinal cord injury (SCI)-induced ischemic delayed paralysis is one of the most serious side effects of aneurysms surgeries. Recent studies prove that the activation of autophagy, including macroautophagy and micro-autophagy pathways, occur during SCI-induced brain neuron damage. However, the role of chaperone mediated autophagy (CMA) during SCI remains to be unveiled. In the present work, rat model of delayed paralysis after aneurysms operation and adenovrius induced LAMP2A knockdown in microglia cells were applied in the present work to investigate the involvement of LAMP2A-mediated CMA in the aneurysm operation related SCI and delayed paralysis. The results showed that LAMP2A was upregulated in the SCI procedure, and contributed to neuron death and pro-inflammation perturbation via inducing iNOS⁺ polarization in microgila. We additionally observed that knockdown of LAMP2A resulted in the shift of microglia from iNOS⁺ to ARG1⁺ phenotype, as well as alleviated neuron damage during SCI. Furthermore, the analysis of BBB score, the result of immunohistological staining, and protein detection confirmed the activation of LAMP2A-mediated CMA activation and its interaction with NF-κB signaling, which leads to neuron death and motor function loss. These results prove that LAMP2A-mediated CMA contributes to the upregulation of pro-inflammatory cytokines and results in cell death in neurons during ischemic delayed paralysis via activating NF-κB signaling. Inhibition of LAMP2A promotes neurons survival during ischemic delayed paralysis.

Introduction

Spinal cord injury (SCI) is a challenging complication in various types of aneurysm surgeries. The incidence of SCI included paresis or paralysis that ranges from 2 to 10%.¹ The etiology of ischemia-induced SCI is believed to involve two mechanisms in the pathological development of traumatic delayed paralysis: the primary ischemic injury and a secondary injury due to inflammation induced by the ischemia/reperfusion cycle. Zeidman et al. summarized several possible mechanisms for the secondary injury: hemorrhage within the cord, which leads to the release of toxic excitatory amino acids, accumulation of endogenous opiates, lipid hydrolysis, and free radical release.^2,3 Practical approaches have been employed to minimize the neurotoxic symptoms, including priming and preconditioning of the paraspinal collateral network, intraoperative systemic hypothermia, distal aortic perfusion, motor- and somatosensory-evoked potentials and noninvasive monitoring of blood pressure and reperfusion (left hind limb), near-infrared spectroscopy of the collateral network and spinal cord oxygenation, as well as perioperative and postoperative cerebrospinal fluid drainage.^4,5 It is concluded that the key strategy to alleviate SCI is the reduction of inflammatory damage to neurons.

Recent studies focused on autophagy in the pathogenesis of SCI, since autophagy is one of the most important pathways involved in cell injury upon stress stimulation.^6–9 Autophagy refers to the catabolic process by which the cell recycles its own constituents via autophagosome. To date, studies identified the existence of three types of autophagy, which are macroautophagy, microautophagy and chaperone-mediated autophagy (CMA).^10–12 CMA is a selective form of autophagy that specializes in translocation of individual cargo proteins into the lysosome for degradation. The CMA pathway consists of the following steps: first, the heat shock cognate 70 kDa protein (HSC70) binds to the target protein that subsequently undergoes CMA. After being recognized by KFERQ or KFERQ-like motif, the complex is delivered to the lysosomal surface, where it interacts with the lysosome-associated membrane protein type 2A (LAMP2A) and unfolds, and subsequently translocates to the lumen of the lysosome.^7,10,13 Therefore, CMA plays a key role in maintaining cellular homeostasis, and LAMP2A serves as a key effector of CMA. The expression, localization, and trafficking of LAMP2A are key markers of CMA pathway activation.¹⁴

CMA plays a crucial role in maintaining cellular energy homeostasis, it is characterized by its ability to cope with oxidative stress and hypoxic stress to alleviate protein damage during harmful stimulation^15–17 CMA has been widely recognized in neurodegenerative diseases, where its function is impaired during aging and disease progression. Emerging evidence suggests that alteration in cell cycle dysregulation, e.g. MEF2A and MEF2D, as well as reduced lysosomal activity in neural stem cells.^18–23 Mice deficient in LAMP2A in macrophages exhibited significant intracellular lipid accumulation, suggesting its involvement in the regulation of of lipid metabolism in immune cells.^24–26 IKKβ is reported as a substrate of CMA, and thereby TNF-α expression is found to be correspondingly reduced in microglia.^24–26 However, little is known regarding to the physiological role of CMA in central nervous system (CNS), especially in microglia cells. The activation of microglia with distinct phenotypic changes, including the inflammatory iNOS⁺ and the neuroprotective ARG1⁺ phenotypes was reported.²⁷ Microglia polarization to iNOS⁺, the pro-inflammatory phenotype, exhibits increased pro-inflammatory cytokines and inflammatory receptors and results in cytotoxicity and cell death. While ARG1⁺ phenotype, which plays an immune-protective role, maintains the function of phagocytosis, debris removal and regeneration support.^27–29 iNOS⁺ microglia are characterized by their amoeboid shape, enhanced mobility and production of pro-inflammatory cytokines. Damage-associated molecular patterns (DAMPs) and pathogen-associated molecular patterns (PAMPs) accompanied with Toll-like receptor 4 (TLR4) are activated, and along with pro-inflammatory factors, such as IL-1β, IL-6, IL-23, IL-18, IL-12, TNF-α, ROS, NO MMP9 and MMP3.^30–32 There are three subtypes of protective phenotypes, which are M2a, M2b and M2c. The markers of M2a include Arg-1, Fizz-1, chitinase 3-like 3 and etc. M2b contains markers of IL-10, sphingosine kinase and suppressor of cytokine signaling 3. Both M2a and M2b play essential roles in phagocytosis, tissue repair, and M2a is also involved in the remodeling of extracellular matrix. M2c induces the expression of IL-10 and TGF-β, which contributes to anti-inflammatory effect.^5,30–33

Collectively, CMA is reported to be activated upon cellular stress, while the mechanism remains to be unveiled. In the present work, we aimed to elucidate the role of CMA in the immune system of the brain, in which we focused on the role of LAMP2A-mediated CMA in the brain immune responses upon SCI stimulation.

Materials and methods

Animals and spinal cord ischemia injured model

All animals used in this study were performed according to the principles stated in the Helsinki Declaration. The experiments conducted in the paper were approved by the regional Ethics Committee from Second Xiangya Hospital of Central South University. The aorta cross-clamping rat model was established as previously published.³³ Forty-eight male Sprague-Dawley (SD) rats (Hunan Silaike Jingda Experimental Animal Co., Ltd, China) were randomly divided into control group (ctrl, n = 6), sham group (sham, n = 6) and spinal cord ischemia group (SCI, n = 12), shNC group (SCI + shNC, n = 12) and shLAMP2A group (SCI + shLAMP2A, n = 12). The animals were accommodated at the condition of 22.0 °C, humidity 50%–70%, light/dark cycle 12/12 h.

The delayed paralysis model of rat was established with aortic cross-clamping methods that was described in our previous publication.³³ In brief, rats were anesthetized by inhaling 3% sevoflurane with 100% O₂ as induction, and 2% sevoflurane was provided during the whole process. The descending thoracic aorta was blocked with bulldog clamp and induced spinal cord ischemia injury. The sham group was performed with the whole operation process without bulldog clamp blocking procedure. After the operations, the rats were supported with 50 mg/kg penicillin for 5 days to reduce infection. Bladder massage was performed to support urination for 5 days after operation. Ringer’s Lactate was provided from the next day after operation. Rats had free access to food and water.

To induce LAMP2A knockdown in the microglia cells, 8-week old male SD rats were subjected to tail vein injection of adenovirus of shLAMP2A (Genepharma, China) (1 × 10¹⁰ pfu in a total volume of 500 μL) to induce LAMP2A knockdown, and control adenovirus that carrying a scrambled sequence were used as negative control (shNC). The sequences are included as below:

shLAMP2A

Foward:5′GTACCTCAAGCGCCATCATACTGGATATTCAAGAGATATCCAGTATGATGGCGCTTTTTTTGGAAA 3′;

Reverse:5′AGCTTTTCCAAAAAAAGCGCCATCATACTGGATATCTCTTGAATATCCAGTATGATGGCGCTTGAG 3′

shNC:

Foward:5′GTACCTCAATTTAGCCGATACTGCCTAGTCAAGAGCTAGGCAGTATCGGCTAAATTTTTTTGGAAA 3′

Reverse:5′AGCTTTTCCAAAAAAATTTAGCCGATACTGCCTAGCTCTTGACTAGGCAGTATCGGCTAAATTGAG 3′

Tissue preparation and cell culture

Rat spinal cord microglia cells were isolated from rat SCI models as described before. In brief, the rats 8-week old male SD rats were euthanized and sacrificed by cervical dislocation, the spinal cord and brain tissues were isolated and stored for future experiments. The microglia cells were isolated from the injured site of spinal cord. Briefly, the injured site was isolated and incubated in DMEM/F12 meidium (SH30023, Hyclone, China). The basal membrane and vessels were removed, and the tissue was minced with scissor and forceps. The mixture was then transferred to 1.5 mL tubes and digested with 0.25% trypsin for 12 min at 37 °C. After digestion, the reaction was neutralized with 10% FBS (SH30070.03, Hyclone, China) and penicillin and streptomycin (SV30010, Hyclone, China) supplemented DMEM/F12. Microglia cells were collected by centrifugation, which were then cultured in an incubator at the condition of 5% CO₂ at 37 °C (3131, Thermo Fisher, USA) for 10 days. Then the microglia cells were ready for subsequent treatments.

RNAi and vector transfection

To investigate the function of LAMP2A, shRNA-targeted LAMP2A was used to knock- down the level of LAMP2A in microglia cells. Firstly specific 19 nucleotide sequences were chosen for constructing plasmid that induces LAMP2A knockdown with pSUPER vector. Primer was designed against the sequences 5′-GGAGATGAATTTCACAATAAC-3′ for rat LAMP2A. Before transfection, approximately 4 × 10⁵ cells were seeded to each well of a 6-well plate, transfection of shLAMP2A or an empty control plasmid was performed with PEI (BIOHUB, 78PEI40000) according to the protocol provided by the company. Briefly, plasmid (1.8 μg/4 × 10⁵ cells) and PEI (5.4 μL/4 × 10⁵ cells) were sequentially added to the medium and gently blown and mixed to form a DNA-PEI mixture. The mixture was left for 20 min to allow the DNA and PEI to fully bind. The DNA-PEI mixture was added to the serum-free medium and incubated with cardiomyocytes. After incubation for 3–6 h, an equal volume of double serum-enriched medium was added and transfection was continued for 24 h. The cells were then incubated with the DNA-PEI mixture for 20 min to fully bind the PEI. Subsequently, cells were further incubated for additionally 24 h.

Immunohistochemistry

Excised rat tissues were washed with PBS, fixed in 4% formalin overnight, and embedded in paraffin. Sections were serially cut into 5 μm thickness and subjected to routine histological staining. Briefly, the deparaffinized sections were rehydrated in a series of graded ethanol (100%–70%–50%–30%, 5 min each), followed by antigen retrieval, then the sections were incubated in primary antibody (LAMP2A, ab125068, abcam, USA) overnight at 4 °C. Rat tissue sections were scanned at 100× and 200× objective lens magnification using wide field microscope (Nikon H550S, Japan) added with camera (Nikon DS-Ri2, Japan). The number of LAMP2A positive cells were counted and measured integral optical density (IOD). The positive cells were defined as cells were double staining with hematoxylin (BA4097, Baso, China) staining in the nucleus and LAMP2A. To compare the different treatment groups, at least 3 independent samples were used and measured.

The spinal cord tissue sections were staining with methylene blue (T104232, Aladdin, China). In brief, the Nissl-stain used methylene blue to label active neuron cells in the spinal cord section to reveal the normal neuron with methylene blue which the stains accumulated in the cytoplasm and bind to the DNA, while the wounded neurons missed the staining. Following Nissl staining, the spinal cord tissue sections were scanned at 40× objective lens magnification with Nikon H550S microscope and captured by Nikon DS-Ri2 camera. The staining intensity were measured with positive cell counting. Three independent samples were used for each group in the statistical analysis.

Immunofluorescence

The microglia slides were fixed with 4% paraformaldehyde for 10 min, and permeability treated with 0.5% Trition X-100 for 10 min, then following the immunostainings of Iba1 (ab48004, abcam, USA), iNOS (AF0199, affinity, UK), Arg1 (DF6657, affinity, UK), LAMP1 (#9091, Cell signaling, USA), p-p38 (612288, BD biosciences, USA), NF-κB (ab16502, abcam, USA) with recommended concentrations (1:200) for overnight. The fluoresce conjugated secondary antibodies Alexa Fluor 555 donkey anti rabbit IgG (ab150062, abcam, USA) and Alexa Fluor 488 donkey anti goat IgG (ab150129, abcam, USA) were used at 1:200 concentrations to reveal the targeting proteins expression level and location in the microglia. The slides were scanned with a confocal microscope (LSM710, Zeiss, Germany). To quantify the targeted protein expressions in the microglia, the IOD were measured, and each experiment were repeated at least three times.

The spinal cord tissue sections were fixed and permeability with 4% paraformaldehyde and 0.2% Triton 100× for 10 min each, following by staining with Iba1, Arg1 and iNOS for overnight. The next day, the sections were incubated with Alexa Fluor 555 donkey anti rabbit IgG and Alexa Fluor 488 donkey anti goat IgG (1: 200) at room temperature for 30 min and DAPI were used to staining nucleus. The sections were scanned with a LSM710 confocal microscope, and IOD were measured afterwards. In each experimental group, at least three animals’ tissue samples were used to perform this experiment.

Behavioral scores

For locomotor function evaluation, Basso–Beattie–Bresnahan (BBB) scoring was performed as previously described.³³ The lower extremity motor function score was adjusted with survival time after operation. The observations were conducted with three independent investigators and blinded to the surgeons; the average score was used to compare the difference between groups.

Quantitative real-time-PCR

Total RNA was extract from cells or tissue with RNAiso Plus (Takara, Japan). According to the manufacturer’s protocol, the RNA samples were storage at −80 °C. The reverse transcription PCR was performed with PrimeScript RT reagent Kit from Takara (Japan), and the qPCR experiments were performed with SYBR Premix Ex Taq II kit. All the primers were described in Table 1.

In general, western blotting experiments, the total cellular or tissue protein samples were isolated with RIPA buffer (Sigma, China) supplied with 1x protease inhibitor cocktail (Sigma, China). Following with protein quantification test with BCA kit (Beyotime, China), and all the samples were denatured by boiling for 10 min at 95 °C. 30 μg of total protein of each sample was loaded and followed by SDS-PAGE gel with BOI (Biorad, USA) under 0.3A currency for 1 h. After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were washed with TBST and incubated with primary antibodies at 4 °C overnight. The membranes were incubated with corresponding secondary antibodies for 2 h at room temperature and developed with ECL system (Beyotime, China). Image J software (NIH, U.S.) was used for densitometric analysis.

The lysosome isolation kit (ab2304, Abcam, USA) was employed to isolate lysosome contents from whole cell following manufacturer’s instructions. Briefly, the cell or tissue samples were treated with lysosome isolation buffer, and followed by mix with lysosome enrichment buffer, and further isolated by sequence of different concentrations of lysis buffer. The whole process was performed on ice. Lysosome samples were directly supplied with blue samples buffer and boiled at 95 °C for 5 min and followed by analysis with western blotting.

Both Native-PAGE and SDS-PAGE were used to further examine the activity of LAMP2A. In the experiment of Native-PAGE the proteins are separated according to the net charge, size and the shape of their native structure and the method was performed as study from Bandyopadhyay et al. Briefly, Native-protein sample buffer 4X (Thermo, USA) were directly added to the protein samples and boiled at 95 °C for 5 min, followed by loading to Novex Tris-Glycine gels (Thermo, USA) and electrophoresis at constant voltage of 150 mV, and the LAMP2A containing complexes were detected by immunoblotting using anti-LAMP2A antibody and quantified by densitometry.

RNase intake assay

Bovine pancreatic ribonuclease A (RNase A) was used to measure protein degradation in the lysosome as previous publication.³⁴ In summary, the lysosomal pathway of proteolysis is subject to regulation in that RNase A is taken up and degraded more rapidly when CMA is activated, since it contains a KFERQ-like motif that is directly involved in CMA15. Thus, the microglia cells, which had been transfected with p38 vector or negative control plasmids, were treated with RNase A with or without p38 inhibitor. The protein from treated samples was collected and further analysed by western blot.

ELISA

ELISA kits against Rat IL-1β (CSB-E08055r), IL-6 (CSB-E04640r), TNFα (CSB-E11987r), IL-10 (CSB-E04595r) and TGFβ (CSB-E04727r) were purchased from Cusabio company. According to the protocols, the serum samples and standard proteins were loaded pre-coated plates and three wells were used per sample or standard proteins. Following by capture antibodies incubation, HRP incubation, and color-developing procedure as the protocol suggested, the final data was measured with Multiskan machine from Thermo Fisher, at 450 nm. All the colorimetric data was calculated to concentrations with stand curve and compared to all the experiment groups.

Statistical analysis

All the statistics were performed with SPSS (18.0 version, SPSS, USA) and GraphPad Prism (9.3.0 version, Prism, USA). Data distributions for each analysis were performed to decide further analytic methods (data were not included). Parametric data was expressed as mean + SEM. Differences between two groups were compared by unpaired two-tailed Student’s t-test. ANOVA with multiple comparison (Tukey’s) were used to analysis the comparision among multiple groups. BBB scores were analyzed with 2-way ANOVA repeated measurement with Tukey’s test adjustment.

Results

LAMP2A plays a key role in CMA-mediated inflammation of microgila cells during ischemia-induced SCI

To investigate the involvements of CMA during the spine cord injury scenario, the microglia cells were isolated from the spinal cords of rat with delayed paralysis, and further investigation of LAMP2A activation and inflammation was performed. Firstly, we examined LAMP2A expression in the brain of rat suffering SCI. The results showed that LAMP2A expression was significantly increased in the brain tissues of SCI group compared to sham group (Fig. 1A). The tissue samples from both prefrontal cortex and hippocampus exhibited higher intensity of LAMP2A staining in SCI group than sham group, with no difference between sham and control groups (Fig. 1A). These data suggest that ischemic stress induced autophagy response in the neuron. We next aimed to explore the chaperone medicated autophagy (CMA) in microglia cells that isolated from central nervous system. Immunoblotting showed that LAMP2A protein abundance was significantly increased in both total cell lysate and subcellular lysosome of microglia cells (Fig. 1B and C), indicating upregulation of CMA in microglia cells during the pathogenesis of SCI. Native-PAGE further confirmed that the majority of LAMP2A binds to lysosomal membrane in the form of complexes (Fig. 1D and E).

Microglia cells can shift into iNOS⁺ (which exhibits pro-inflammatory) or ARG1⁺ (which play an anti-inflammatory role) phenotype under the inflammatory perturbation. Therefore, the mRNA level of iNOS, which are markers of pro-inflammation, together with the markers of anti-inflammation, including Arg1 and Ym1/2, were analyzed in the microglia cells isolated from control, sham and SCI groups. The results showed that there was significant elevation of iNOS in microglia cells from SCI group, while Arg1 and Ym1/2 were downregulated (Fig. 1F). Taken together, our results suggest that during SCI, microglia cells are activated in response to CMA signaling and shift towards to a iNOS⁺ phenotype.

LAMP2A-mediated CMA promotes the shift of microgila cells to iNOS⁺ phenotype via during SCI

To investigate the link between CMA and iNOS⁺/ARG1⁺ phenotype shift in microglia cells, we first checked IL-1β, iNOS and LAMP2A expression in microgila cells isolated from rats at different time points after SCI. The result showed that protein abundance of LAMP2A, iNOS and IL-1β exhibit significant increased in a time-dependent manner (Fig. 2A). Subsequently, Lapma2a was knocked down in microglia cells via siRNA transfection (shLAMP2A), and the lentivirus carrying a scrambled sequence was used as negative control (shNC) (Fig. 2B). Then the mRNA level of iNOS, IL-1β, Arg1 and Ym1/2 was evaluated in microglia cells with LAMP2A knockdown driven by siRNA transfection. The result showed that there were significant upregulation of iNOS, IL-1β, Arg1 and Ym1/2 mRNA in microglia cells isolated from SCI rats (Fig. 2C), which was blunted in microglia cells with shLAMP2a (Fig. 2C). In addition, serum concentration of pro-inflammatory cytokines, including IL-1β, IL-6 and TNFα were significantly induced in SCI rats, and these cytokines were suppressed by shLAMP2A (Fig. 2D). On the contrary, IL-10 and TGF-β, two cytokines proved to be anti-inflammatory, were downregulated in in microglia cells of SCI rats, while significantly induced in in microglia cells transfected with shLAMP2A (Fig. 2D). Based on these facts, we hypothesized that in microglia cells, SCI activates CMA, the latter subsequently induces the polarization of microglia cells towards iNOS⁺ phenotype. To further confirm this hypothesis, immunofluorescent staining of Iba1/iNOS and Iba1/Arg1 was performed with microglia cells of ctrl, sham SCI + shNC and SCI + shLAMP2A groups. The results showed that there was a reduction of iNOS and Iba1 intensity in microglia cells transfected with shLAMP2A, while Arg1 was increased (Fig. 2F). Collectively, the results suggest that CMA pathway is activated during ischemia induced spine cord injury and plays a crucial role in microglia mediated inflammation.

p38 MAPK is activated in the lysosome of microglia cells during ischemia-induced SCI

To investigate the role of p38 MAPK in ischemia induced SCI, SB203580, a pharmacological inhibitor of p38 MAPK was applied to treat microgila cells. MEF2D is one of the directly interactive transcription factors of p38 that is indispensable for p38 activation. In addition, MEF2D is protected by ATF2, which is proved to be a substrate for p-p38. Therefore, to identify that p38 MAPK pathway is also activated and involved in regulating CMA activation during SCI, immunoblotting of p-p38α, MEF2D and p-ATF2 was performed with microgila cells isolated from different groups of rats. The result showed that microglia cells from SCI group exhibited a reduction of MEF2D, while there was increased p-p38α and p-ATF2 protein abundance, as well as the ratio of p-p38α/p-38α and p-AFT2/AFT (Fig. 3A and B). On the contrary, application of p38 inhibitor SB203580 blunted the upregulation of p-p38α/p-38α and p-AFT2/AFT induced by SCI (Fig. 3A and B), suggesting the activation of p38 MAPK signaling is activated during SCI. The protein abundance of p-p38α in the lysosome of microglia cells from different groups was detected. Though the total p38α level remained equivalent among all groups, the p-p38α exhibited a significant elevation in the lysosome fractions from microglia cells of SCI group (Fig. 3C and D). Moreover, increased PDI protein level was also observed in the SCI group (Fig. 3C and D). Taken together, these results suggest that p38 MAPK activation is associated with ER stress and CMA activity induced unfolded protein response (UPR).

p38 MAPK regulates LAMP2A-mediated CMA in microglia cells

Wenming Li et al. reported p38 MAPK pathway mediated ER stress-induced CMA via directly phosphorylation of LAMP2A in SN4741 cells.³⁵ The microglia cells with LAMP2A knockdown were used to further investigate the involvement of p38 MAPK pathway in CMA activation. The shLAMP2A function was examined as shown in fig. 4A and B, a clearly reduced protein level of LAMP2A was found in cellular and lysosome. In line with previous report, overexpression of p38 caused increased LAMP2A protein abundance in the lysosome (Fig. 4B), which is blunted by p38 inhibitor treatment (Fig. 4B). The co-localization of p-p38 and LAMP1 (Fig. 4C and D) further verified p38 MAPK pathway activation, as well as p-p38 translocated to lysosome. We additionally observed that SB 203580 treatment blunted the increase of LAMP2A induced by p-p38, while CD63 showed no alteration in the lysosome membrane (Fig. 4E and F). Moreover, protein abundance of LAMP2A in the lysosome matrix was also suppressed by SB203580 (Fig. 4G). Microinjection ribonuclease A (RNase A) assay showed the activation of p38 MAPK signaling during CMA (Fig. 4H). Collectively, these results suggest that p38 pathway regulates CMA activation through LAMP2A in microgila cells.

Knockdown of LAMP2A inhibits the activation of NF-κB signaling during SCI

Classic NF-κB pathway in microglia is accepted for assessing neuron death in the context of injury or disease.^26,36,37 We therefore interrogated the NF-κB pathway activation in the SCI model. Activation of the members involved in NF-κB pathway was observed in microglia cells from SCI group, which were blunted in microglia cells from SCI group with LAMP2A knockdown via shRNA transfection (Fig. 5A and B), and this result was in agreement with the result of immunofluorescent staining of NF-κB in microgila cells isolated from different groups (Fig. 5C and D). Nissl staining with spinal cord of mice from different groups showed that SCI induced increased neuron death in a time-dependent manner, by showing dented nuclear and empty babuls compared to the control group, the normal motor neuron of which exhibited a regular and clear nuclear and cellular membrane shape. However, impaired neuron induced by SCI was blunted by LAMP2A knockdown (Fig. 5E and F). BBB score showed that rats of SCI group exhibited less capacity on limb’s functions compared to the sham group, which was rescued by LAMP2A knockdown (Fig. 5G). We also detected other targets downstream of p38 MAPK, including AKT and STAT3. The result of western blot showed that the was no significant changes of p-AKT and p-STAT3 in microgila cells isolated from SCI models of rats with vehicle and LAMP2A knockdown, suggesting that NF-B signaling is specifically activated by LAMP2A in microgila cells during SCI-induced injury. Taken together, these results suggest that downregulation of LAMP2A inhibits the activation of NF-κB in microgila cells during SCI.

Knockdown of LAMP2A suppresses NF-κB-dependent inflammation during SCI

To testify the involvement of LAMP2A in the NF-κB signaling during SCI in vivo, rats were subjected to the surgery that induced SCI, the spinal cord tissue was isolated at 6 h, 24 h and 48 h after the surgery. Immunoblotting of the tissue samples from different groups showed increased protein abundance of Iba1, iNOS, IκB, p50 and p65 in the spinal cord after SCI, all of which exhibited a time-dependent manner (Fig. 6A–I), suggesting the activation of both NF-κB pathway in the SCI group, which was inhibited by shLAMP2A (Fig. 6A–I). The spinal cord tissue was stained with Iba1/Arg1 or Iba1/iNOS to investigate the activation of microglia cells and CMA at 6 h, 24 h and 48 h after SCI, respectively. The result showed colocalization of Iba1 and Arg1 in the spinal cord (Fig. S1A and B). In addition, there was increased immunofluorescent staining of Iba1 in SCI group in a time-dependent manner, which was blunted by LAMP2A knockdown (Fig. S1A and B). On the contrary, Arg1 exhibited a reduced tendency in the spinal cord of rats from the SCI group, which was significantly rescued by shLAMP2A (Fig. S1A and B). The detection of Iba1 and iNOS also showed colocalization of both proteins in the spinal cord (Fig. S2A and B) that there was increased immunofluorescence of iNOS in the spinal cord of rats in a time-dependent manner, which was inhibited by shLAMP2A (Fig. S2A and B). Furthermore, serum concentration of pro-inflammatory cytokines, including IL-1β, IL-6 and TNFα were significantly induced in SCI rats, were suppressed by shLAMP2A (Fig. 6J). On the contrary, the anti-inflammatory cytokines including IL-10 and TGF-β were downregulated, while both were significantly induced by shLAMP2A (Fig. 6C). Notably, we also detected the involvement of macroautophagy in rat microgila cells subjected to SCI. ELISA assay of cytokines showed that LAMP2A knockdown induced significant decrease of iNOS, IL-1β, IL-6, TGF-β and TNF-α, and LC3 knockdown potentiates the suppression effect of LAMP2A on these genes (Fig. S3A and B). On the contrary, LAMP2A knockdown induced elevation of IL-10, which is strengthened by LC3 knockdown (Fig. S3A and B). Therefore, these results suggest that inhibition of both CMA and macroautophagy has an additive effect on microglia phenotype and neuronal protection during SCI-induce injury. To further confirm the role of LAMP2A in SCI-induced neuron injury, we performed SCI experiments with human iPSC-derived microglia with LAMP2A knockdown via shRNA transfection, and detected inflammatory markers including iNOS, Arg1 and IL-1β. The results showed that the protein abundance of Arg1 increased in a time-dependent manner, while iNOS and IL-1β decrease in a time-dependent manner (Fig. 6K and L), which is in agreement with the result of rat primary microgila cells. Together, these results suggest that downregulation of LAMP2A suppresses inflammation caused by the activation of NF-κB signaling during SCI (Fig. 7).

Discussion

Microglia play an important role in brain defence and repair. Although the activation of CMA in microgila cells upon ischemic SCI stimulation was confirmed by other studies, the consequences are complicated. For instance, Yujung Park and associates reported an elevation of CMA in the traumatic brain injury rat model,⁹ and Hyoichi Handa and associates proved the activation of CMA after SCI.⁷ However, both researches focused on protein misfold and degradation. It is reported that inflammation is also activated in response to dislocated and damaged cellular components during SCI, but the mechanism is unclear. In the present study, the role of LAMP2A was investigated in the progress of SCI. Our results showed that LAMP2A was involved in SCI via regulating neuron death and polarization of microglia cells. We also provided evidence that CMA activation and its interaction with NF-κB and MAPK pathway lead to neuron death and motor function loss. Therefore, our work prove the important role of LAMP2A in the progress of SCI, and provides novel insights into the function of LAMP2A during SCI-induced brain injury.

LAMP2 (lysosomal associated membrane protein 2) is one of the major protein components of the lysosomal membrane. LAMP2A serves as a receptor and channel for transporting cytosolic proteins in a process called chaperone-mediated autophagy (CMA). Our result showed that LAMP2A expression was significantly increased in the damaged neural tissue. The CMA activation is associated with LAMP2A protein level in microglia cells in both total cell lysate and lysosome fraction. We additionally observed that LAMP2A level was positively correlated with ionized calcium binding adaptor molecule 1 (Iba1) intensity. Iba1 is a microglia/macrophage-specific calcium-binding protein, which maintains actin-bundling activity and participates in membrane ruffling and phagocytosis in activated microglia. Therefore, our results suggest a key role of LAMP2A in the activation of microgila cells.³⁰ iNOS is a marker of polarization to be detrimetal to neurons while Arg1 represents benefical polarization for neurons.^38–41 In microgila cells, the iNOS⁺ polarization functions as the pro-inflammatory type, and on the contrary, ARG1⁺ phenotype maintains the anti-inflammatory function. There are several studies testified that CMA activation modulates the phenotype of macrophage polarization, though the consequences were distinct among different tissue.^{25,26,35,42–44} Thus, we were interested in exploring the role of LAMP2A in regulating the polarization of microglia during SCI and the mechanism involved.

After CNS injury, inflammation causes secondary damage and thus exacerbates post-SCI tissue damage and functional loss. These effects result in the release of numerous mediators that exert cytotoxic effects on CNS cells, which then leads todemyelination and neuronal loss in microglia and macrophages.^45,46 During the progression of delayed paralysis and traumatic spinal cord injury, CMA and inflammation are activated in response to removing damaged, misfunctioned components, the process of which is essential for maintaining the homeostasis of the neuron. However, the results of a recent investigation showed that reduced inflammation and immunocyte infiltration in the injured CNS is able to improve neuronal regeneration.⁴⁷ Therefore, both CMA and inflammation consequence are key factors that affect neuron function after SCI. Our data from rat model demonstrates a gradual loss of function due to neuron cell death in the spinal cord, which is in agreement with prvious studies.^7,9,22,35 Thus, promoting ARG1⁺ type of microglia/macrophage in the spinal cord damage serves as a promising therapeutic strategy. In the present work, immunofluorescence double staining and immunoblotting of microglia marker, as well as iNOS⁺/ARG1⁺ marker demonstrated that the polarization to iNOS⁺ phenotype of microglia was significantly inhibited by LAMP2A knockdown both in vitro and in vivo, the ARG1⁺ polarization was not enhanced compare to control or sham group. Therefore, we concluded that LAMP2A-mediated CMA contributes to iNOS⁺ polarization in microglia cells, which is involved in neuron homeostasis after SCI. Notably, we also found that besides CMA, macroautophagy is also involved in SCI-induced brain injury, suggesting complicated signallings are provoked upon SCI-stimulation, which is worthy of further investigation.

In summary, our study provided evidence of the effects of LAMP2A knockdown on neuronal injury, microglia polarization and inflammation after SCI. Our results demonstrate that LAMP2A could be a potential target that regulates microglia polarization and inflammtion through the autophagy-lysosome pathway to attenuate SCI-induced injury in the brain tissue.

Conclusion

LAMP2A-mediated CMA contributes to the upregulation of pro-inflammatory cytokines and results in cell death in neurons during ischemic delayed paralysis via activating NF-κB signaling. Inhibition of LAMP2A promotes neurons survival during ischemic delayed paralysis.

Acknowledgments

Thanks for all the contributors and participants.

Author contributions

Jingjuan Li, Yan Huang and Liang He contributed the central idea, analysed most of the data, and wrote the initial draft of the paper. The remaining authors contributed to refining the ideas, carrying out additional analyses and finalizing this paper.

Funding

This study was supported by the Science and Technology Project of Guangzhou (202201010818). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflicts of interest

There are no conflicts of interest claimed by all the authors.

Data availability

All the data are available up to request, please don’t hesitate to contact the authors.

References

Author notes