Introduction
Morphine is widely used to treat chronic pain clinically, including cancer-related and non-cancer-related pain [
1]. However, chronic morphine treatment will lead to deleterious host innate immunity impairment [
2], which potentiates pathogenic infections such as HIV. Immunosuppressive complications will limit the clinical efficacy of morphine. However, the exact mechanisms involved in morphine-induced immunosuppression, especially in the brain, still remain not fully clarified.
Mitochondria have emerged as critical mediators in the regulation of inflammation and host innate immunity [
3]. Mitochondrion sustains constant fission and fusion to maintain physiological functions and cellular homeostasis. Mitophagy, a selective form of autophagy, eliminates damaged or excess mitochondria in response to cellular stress. During mitophagy, mitophagosomes, the double-membraned vesicles, engulf and encapsulate damaged mitochondria and are shuttled to lysosomes for further proteolytic degradation [
4]. Multiple signaling mediators in innate immunity participate in mitophagy regulation, underlying the crosstalk between mitophagy and host inflammatory response [
5]. Dysregulated mitophagy may orchestrate exuberant inflammasome activation or immunosuppression and various pathogens would be able to manipulate mitophagy facilitating pathogenesis during infection [
6]. Generally, mitophagy is initiated through two major pathways, including the PINK1–Parkin axis and mitophagy receptors which contain an LC3 interacting region (LIR) motif for LC3-decorated autophagosome recruitment [
7,
8]. Classically, the PINK1-Parkin pathway depended on the mitochondrial membrane potential (
Δψm) and functions in an ubiquitin-dependent pattern [
9]. On the other hand, several mitophagy receptors have been identified, among these receptors, NOD-like receptor X1 (NLRX1) is the only mitochondrially localized NOD family receptor protein, which was recently reported to function as a novel mitophagy receptor during
L. monocytogenes infection to evade host killing [
10]. NLRX1 contains a LIR motif for binding with LC3, a prerequisite for mitophagy receptor. However, the proposed mechanisms by which NLRX1 mediates mitophagy remain unknown and require further elucidation. So far, NLRX1 reportedly appears to be a versatile anti-inflammatory regulator during various pathogenic microorganism infection. NLRX1 mediated MAVS signaling and inhibited production of IFNs to facilitate the survival of viruses during mitochondrial antiviral immunity [
11]. Contrastingly, NLRX1-deficient mice tended to be susceptible to LPS-induced septic shock, which resulted from aberrant immune responses [
12]. In summary, NLRX1 seemed to serve as a critical negative regulator in immunity. Besides, intriguing data also revealed that NLRX1 manipulated virus-induced autophagy for host defense [
13]. Taken together, NLRX1-mediated mitophagy may play an important role in pathogen–host cell interactions.
Microglial cells, which serve as the resident phagocytic and immune cells in the brain, have been reported to be pivotal mediators in infectious neuroinflammation in patients receiving morphine treatment [
14]. Aberrant expression of immune-related genes (such as
IL-1β,
IL-6,
IL-18,
TNF-α, and so on) in microglia might inhibit microglial activity and endanger the central nervous system [
15]. Severe or even fatal bacterial sepsis occurs frequently once invading pathogens attack, which might account for the high susceptibility to bacterial infection induced by morphine [
16]. In the present study, we aimed to clarify the role of microglial mitophagy in the induction of cerebral immunosuppression and its regulative mechanisms after morphine treatment, as well as the selective brain regions [
17] where microglial mitophagy mainly served as the immunoregulatory element in morphine-treated mice.
Materials and methods
Reagents
Morphine was provided by Sun Yat-sen Memorial Hospital, which was approved by Guangdong Medical Products Administration. CCCP (10 μM, HY-100941), Mdivi-1 (20 μM, HY-15886) and Bafilomycin A1 (20 nM, HY-100558) were purchased from MedChemExpress. Torin1 (250 μM, SC0245) was purchased from Beyotime. Puromycin (Sigma, P7255) was used to select stably expressed cells.
Cell culture and RNA interference
Primary microglial cells were isolated and cultured according to our previous protocol [
18] and the purity was verified as 95% or higher by immunofluorescence staining with Iba-1 (Additional file
1: Fig. S1A). For primary astrocytes, the cells were cultured in DMEM (HyClone, SH30243.01) containing 10% fetal bovine serum (FBS) (Gibco, 10099–141) and 1% penicillin/streptomycin. After the cells reached confluence (about 7 days), the cells were subcultured. After 30-min pre-adherence, the medium containing nonadherent cells was replaced. Subculture was performed three times every 4 days as suggested [
19]. The purity of astrocyte was confirmed by fluorescence staining with GFAP (Additional file
1: Fig. S1B). BV2 cells and MA cells were purchased from American Type Culture Collection (ATCC). BV2 cells were cultured in DMEM/F12 medium (Gibco, 11330-032) while MA cells in DMEM high glucose. Cells were maintained in complete medium containing 10% FBS and 1% penicillin/streptomycin in a 5% CO
2 incubator at 37 °C. For siRNA transfection, the NC-siRNA and NLRX1-siRNA were purchased from RiboBio Co., Ltd (Guangzhou, China). Transfection was performed using Lipofectamine 3000 reagents (Invitrogen, L3000001) and the sequence targeting NLRX1 was as follows: sense, 5′-GCCACAGAAGCUAUCCAAAdTdT-3′, anti-sense 5′-CGGUGUCUUCGAUAGGUUUdTdT-3′.
DNA, RNA isolation and real-time quantitative PCR (qPCR)
Genomic DNA was isolated using TIANamp Genomic DNA kit (TIANGEN, DP304-03). mtDNA (mitochondrial DNA, the conserved sequence in the D-loop region) copy number was normalized to nuclear DNA (Hbb, β-globin) gene. Total RNA was isolated by TRIzol reagent (Takara, #9109) and then transcribed to cDNA using PrimeScript RT Reagent Kit (Takara, #RR037A). The qPCR was performed in Applied Biosystems QuantStudio 5 (ThermoFisher). The following primers were used:
(Mouse) mt-DNA forward, GCCCATGACCAACATAACTG; (Mouse) mt-DNA reverse, CCTTGACGGCTATGTTGATG; (Mouse) Hbb (β-globin) forward, AGGCAGAGGCAGGCAGAT; (Mouse) Hbb (β-globin) reverse, GGCGGGAGGTTTGAGACA; (Mouse) IL-1β forward, TGCCACCTTTTGACAGTGATG; (Mouse) IL-1β reverse, AAGGTCCACGGGAAAGACAC, (Mouse) IL-6 forward, AGGATACCACTCCCAACAGACCT; (Mouse) IL-6 reverse, CAAGTGCATCATCGTTGTTCATAC, (Mouse) IL-18 forward, ATGCTTTCTGGACTCCTGCC; (Mouse) IL-18 reverse, ATTGTTCCTGGGCCAAGAGG, (Mouse) TNF-α forward, ATGCTTTCTGGACTCCTGCC; (Mouse) TNF-α reverse, ATTGTTCCTGGGCCAAGAGG, (Mouse) iNOS forward, CTTGCCACGGACGAGAC; (Mouse) iNOS reverse, TCATTGTACTCTGAGGGCTGA, (Mouse) NLRX1 forward, ACCTCACCGAGTGGTTTAGC; (Mouse) NLRX1 reverse, TCACGGGGTCAACATGAACTG, (Mouse) GAPDH forward, TGACCTCAACTACATGGTCTACA; (Mouse) GAPDH reverse, CTTCCCATTCTCGGCCTTG, (Mouse) Atp6v0d1 forward, CGCCACATGAGAAACCATGC; (Mouse) Atp6v0d1 reverse, CTCAAAGCTGCCTAGCGGAT, (Mouse) Atp6v0d2 forward, CTGGTTCGAGGATGCAAAGC; (Mouse) Atp6v0d2 reverse, TCCAAGGTCTCACACTGCAC, (Mouse) LAMP1 forward, CCAGAGCGTTCAACATCAGC; (Mouse) LAMP1 reverse, ACAGGCTAGAGCTGGCATTC, (Mouse) LAPTM4A forward, TGCGTTCTTTTTGCCGTCTC; (Mouse) LAPTM4A reverse, GAATCAGCCAGCCCACTTGA.
Co-immunoprecipitation and western blotting
Cells were washed in PBS and lysated in IP lysis buffer (Beyotime, P0027). 500 ug proteins were subjected to immunoprecipitation. 1 ug rabbit anti-NLRX1 antibody (Cell Signaling Technology, 13829 s) was added to lysate for 1 h, followed by 20 μl protein A agarose beads (Santa-Cruz, sc-2003) overnight. The beads were washed by cold PBS and eluted by boiling in 2 × loading buffer. The input and eluted fractions were then subjected to immunoblot analysis. Homophytic IgG was employed as a negative control. NLRX1 was used for equalization for IP.
For western blotting, cells or brains were homogenized in lysis buffer (Beyotime, P0013) with complement of phenylmethylsulfonyl fluoride (Beyotime, ST506). Equal amounts of protein lysates were used for immunoblot analysis. Following primary antibodies were used: HSP60 (Affinity, AF0184), Tim23 (Affinity, DF12052), LC3A/B (Cell Signaling Technology, 4108S), GAPDH (Cell Signaling Technology, 2118S), mTOR (Affinity, AF6308), p-mTOR (Proteintech, 67778-1). Image J was utilized to analyze the densitometry of bands and GAPDH was use as a loading control.
IF/ICC immunofluorescence staining
After cardiac perfusion, brains of mice were removed and fixed in 4% paraformaldehyde solution at 4 °C, following by gradient dehydration in sucrose. Brains were then cut into thick sections (10 µm) by LEICA CM1950. The following primary antibodies were used: NLRX1 Rabbit antibody (Affinity, DF12124), MAP1LC3B Mouse antibody (ABclonal, A17424), Iba1 Goat antibody (Abcam, ab48004), GFAP Mouse antibody (Huabio, EM140707), CD31 Mouse antibody (Abcam, ab222783), NEUN Mouse antibody (Abcam, ab104224). The following secondary antibodies were used: Alexa Fluor 488 AffiniPure Donkey anti-Rabbit IgG (H + L) (Yeasen, 34206ES60, 1:200), Alexa Fluor 647 AffiniPure Donkey Anti-Mouse IgG (H + L) (Yeasen, 34113ES60, 1:100), Cy3-labeled Donkey Anti-Goat IgG(H + L) (Beyotime, A0502, 1:250), Alexa Fluor 555-labeled Donkey Anti-Mouse IgG(H + L) (Beyotime, A0460, 1:500). The images were acquired and analyzed by a confocal microscope (LSM 880 with Airyscan). To elucidate the cellular location of NLRX1 and LC3B in vivo, 11–12 randomly selected fields per group (n = 3 mice) were used to calculate the ratio of NLRX1 + cells in microglia, NLRX1 + cells in LC3B + microglia and the ratio of microglia in NLRX1 + cells.
For ICC, cells were cultured in glass bottom cell cultured dishes (NEST, 801,001) and washed by cold PBS. After fixation by 4% paraformaldehyde solution, permeabilizationwas performed in 0.3% Triton™ X-100 in PBS. After blocking for 1 h, the cells were incubated by primary antibodies as follows: NLRX1 Rabbit antibody (Affinity, DF12124), MAP1LC3B Mouse antibody (ABclonal, A17424), LAMP1 Rabbit antibody (Bioss, bs-1970R), Iba1 Goat antibody (Abcam, ab48004), GFAP Mouse antibody (Huabio, EM140707), Tim23 Rabbit antibody (Affinity, DF12052), HSP60 Rabbit antibody (Affinity, AF0184), TFEB Rabbit antibody (Affinity, AF7015). The secondary antibodies were Alexa Fluor 555-labeled Donkey Anti-Rabbit IgG(H + L) (Beyotime, A0453, 1:500) and Alexa Fluor 488-labeled Goat Anti-Mouse IgG(H + L) (Beyotime, A0428, 1:500), Cy3-labeled Donkey Anti-Goat IgG(H + L) (Beyotime, A0502, 1:250), Alexa Fluor 555-labeled Donkey Anti-Mouse IgG(H + L) (Beyotime, A0460, 1:500). Immunofluorescence images were captured by confocal laser scanning microscope (FV10i, Olympus). The co-localization was analyzed by Image-Pro Plus 6.0 (Media Cybernetics, USA). For the co-localization analysis of NLRX1 and LC3B, Pearson’s correlation coefficient of the single cell from at least 6 randomly selected fields in 3 independent experiments per group were analyzed. For the analysis of TFEB nuclear translocation, 25 randomly selected fields in 3 independent experiments per group were analyzed. For the co-localization analysis of LAMP1 + lysosomes and LC3B + particles, approximately, 25–40 cells from at least 5 randomly selected fields in 3 independent experiments per group were analyzed. The overlap coefficient, lysosomal phagocytosis ratio of LC3B and LC3 + phagosomes per cell were analyzed and plotted.
Adenovirus mCherry‐GFP‐LC3B transfection
BV2 cells in 12-well plates were infected by 40 MOI (multiplicity of infection) adenovirus expressing mCherry‐GFP‐LC3B fusion protein (Beyotime, C3011) and the LC3B-positive autophagosomes were detected by a confocal microscope (LSM 980 with Airyscan). mCherry (red)- and GFP (green)-positive dots meant the aggregation of LC3B-positive autophagosomes. When the autophagosomes were infused by lysosomes, the GFP protein quenched in the acidic environment and mCherry protein was stable. More than 50 cells per group from 3 independent experiments were analyzed.
Electron microscopy assays
Cells were divided into four groups: control group; morphine group; morphine + NC-siRNA group; morphine + NLRX1-siRNA group. Cells were quickly collected and fixed in 2.5% glutaraldehyde at 4 °C overnight. After washing by 0.1 M PBS for three times, samples were fixed in 1% osmic acid and then dehydrated in a series of graded ethanol. The cells were then embedded and 70-nm ultra-thin sections were stained by 3% uranyl acetate–lead citrate. The images were captured by an electron microscope (HITACHI 7800).
JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethyl-benzimidazolyl carbocyanine iodide)
To measure the mitochondrial membrane potential (Δψm), JC-1 (Beyotime, C2006) assay was performed according to manufacturers’ instructions. After incubation with 10 μg/mL JC-1 at 37 °C for 20 min, cells were washed and detected by a flow cytometer (LSR II, BD). The Δψm was revealed by the ratio of red intensity (JC-1 aggregates) to green intensity (JC-1 monomers).
Mitochondrial and lysosomal fluorescent probes and image quantitation
Cells were cultured in glass cultured dishes and incubated with LysoTracker Red DND-99 (Yeasen, 40739ES50) and MitoTracker® Green FM (Yeasen, 40742ES50) according to manufacturers’ instructions for 30 min. Gently replaced the medium with complete medium and captured images immediately by microscope (FV10i, Olympus). The morphologies of mitochondria and lysosomes were subsequently analyzed by Image J. Average size and circularity value of mitochondria were analyzed using a Particle Analysis in Image J from more than 25 cells in 3 independent experiments per group. Form factors of mitochondria were calculated by the reciprocal of circularity value according to a previous research [
19]. The diameters of individual lysosomes (circular or oval shaped) were calculated and plotted as size distribution. As [
19] suggested, lysosomes with diameter between 0.2 and 1 μm were recognized as normal lysosomes while those larger then 1 μm were considered as abnormal or vacuolar lysosomes.
LysoSensor™ Green DND-189 (Yeasen, 40767ES50) was utilized to detect the acidity of lysosomes. Cells were incubated with medium containing 1 uM probe for 30 min at 37 °C. Subsequently, the cells were suspended and detected by flow cytometer (LSR II, BD). The higher intensity indicated the enhanced lysosomal acidity.
Cell viability assays
The Cell Counting Kit-8 (CCK8) assay (DOJINDO, CK04) was performed as manufacturers’ instructions. In brief, cells were pre-treated with morphine or vehicle and subsequent LPS treatment for 6 h, 12 h or 24 h. After washing by PBS, fresh medium containing 10% CCK8 solution was added to cells and incubated for 2 h at 37 °C. The absorbance values were read by a microplate reader (BMG POLARstar Omega) at 450 nm. The cell viability was revealed by the percentage of the optical density value in control group.
Lentiviral silencing of NLRX1
Four shRNAs targeting NLRX1 and scrambled (Scr) shRNA plasmids were purchased from GeneCopoeia and the target sequences were as follows: scrambled, ACAGAAGCGATTGTTGATC; a, GCTGGACCGAAACAAACAACT; b, CCAGAAAGATCCCTTTAATTC; c, GCATCTATACCAGCTTTCTAC; d, GCTGCGCAAATACATGCTTCC. To generate shRNA-expressing lentivirus, 293Ft cells were transfected with 20/3 ug shRNA plasmids, packaging plasmids including 5 ug psPAX2 and 8/3 ug pMD2.G. After 72 h, the supernatant was collected and filtered by a 0.45-μm cell strainer. After concentration (Yeasen, 41101ES50), the biological titer of lentivirus were calculated and the optimal sequence was chosen according to the qPCR analysis of primary microglia (Additional file
1: Fig. S1D).
Experimental animals
All experiments were under protocols approved by the Institutional Animal Care and Use Committee, Sun Yat-Sen University (Approval number, SYSU-IACUC-2018-000182). 6 to 8-week-old, male C57BL/6 mice were house-caged with a 12:12-h light–dark cycle and free to food and water. We made all efforts to minimize pain and suffering of the mice. The mice were randomly divided into groups as follows: (1) Control group; (2) Morphine group (15 mg/kg/day, subcutaneously, 7 d); (3) Lentivirus (Lv)-GFP (green fluorescence protein)-Scr-shRNA (bilaterally, intracerebroventricularly) + Morphine; (4) Lv-GFP-NLRX1-shRNA (bilaterally, intracerebroventricularly) + Morphine; (5) Lv-GFP-Scr-shRNA + morphine + LPS (intraperitoneally, 1 mg/kg, L2880, sigma); (6) Lv-GFP-NLRX1-shRNA + morphine + LPS. The mice were received LPS or saline at the last day and sacrificed at 6 h. The liver, spleen and thymus were weighed and the organ indexes were calculated as follows: organ index = (organ weight (mg)/bodyweight (10 g)) × 100%.
Statistical analysis
The experimental values were expressed by mean ± SEM and the data were obtained from minimum of three repeats independently. Normal distribution test and Levene's test were performed. Student's t-test or Mann–Whitney U test by GraphPad Prism 8 software were employed to indicate statistical significance between two groups. One-way ANOVA or Kruskal–Wallis test were employed to indicate statistical significance among multiple groups. Values at p < 0.05 were considered statistically significant.
Discussion
Morphine-induced immunosuppression has been widely discussed in the past, but the underlying mechanisms remained unclear yet [
16]. In addition to the peripheral immune system, the brain-resident system is also an important susceptible region to pathogen infection after chronic morphine exposure which will cause poor prognosis for patients [
36]. In this study, we demonstrated that NLRX1-mediated insufficient mitophagy facilitated microglial immunosuppression after morphine treatment, which might be responsible for the fragility to invading pathogens.
Firstly, enhanced NLRX1 expression and mitophagy activation were concomitantly observed in microglia treated with morphine for 24 h. Furthermore, because a distinctive LIR motif for LC3 binding was a prerequisite for receptor-mediated mitophagy, we demonstrated the binding of NLRX1 to LC3, serving as a mitophagy receptor. NLRX1-silencing rescued the morphine-induced mitophagy in microglia. Presumably, neither NLRX1 nor mitophagy seemed to function in astrocytes after morphine treatment, which indicated that microglia might be the main inflammatory cell type for morphine-induced mitophagy in the CNS. It was not surprising that there were cell type-specific differences after morphine treatment. It has been demonstrated that astrocytes were resistant to 1 μM morphine induced cytotoxicity but not microglia [
37,
38]. More importantly, morphine tended to protect astrocytes from glutamate-induced apoptosis [
39] and activated the astrocytic μ receptor but not microglial to promote the release of CCL5, which exhibited a neuroprotective property during HIV infection [
40]. Therefore, morphine mediates multiple effects in different cells in CNS. More importantly, there was limited evidence demonstrated the direct role of NLRX1 in the astrocytic inflammatory response. It has been substantiated neurotoxic astrocytes were subsequently induced by activated microglia [
41]. Thus, an alternative explanation assumed that NLRX1 might inhibit microglial activation, resulting in interrupting the generation of neurotoxic astrocytes [
42]. However, future work is required to address whether NLRX1 could directly regulate the astrocytic inflammatory response. The
Δψm remained unchanged and thus the PINK1–Parkin axis might not be responsible for morphine-induced mitophagy, counterintuitively. No detectable change of NLRX1 expression was observed in CCCP-treated BV2 cells. These results indicated that NLRX1-mediated mitophagy was activated by morphine in a specific manner, independently of the PINK1–Parkin pathway.
Secondary, we focused on the mitophagic flux regulated by NLRX1 in morphine-treated BV2 cells. Our results showed that morphine disturbed lysosomal function, including lysosomal acidification, mitophagosome–lysosome fusion, and lysosomal biogenesis. LC3B punctas accumulated in the cytosol which resulted from enhanced LC3-decorated mitophagosomes and insufficient lysosomal degradation. NLRX1-silencing alleviated the lysosomal dysfunction through promotion of lysosomal generation and acidification, but failed to entirely restore the mitophagosome–lysosome fusion. These results implied the NLRX1-mediated mitophagy might be incomplete with lysosomal dysfunction. TFEB is a master modulator of lysosomal catabolic function [
43]. TFEB regulated the lysosomal activity via binding to conserved coordinated lysosomal expression and regulation (CLEAR) motif of targeted genes [
23]. Our results suggested that NLRX1 might suppress the nuclear translocation of TFEB. The association of NLRX1 deficiency and activation of TFEB has also been demonstrated by a previous study [
19]. However, detailed mechanisms remained unknown. mTORC1 has a critical role in suppressing the lysosomal function by inactivating TFEB [
44]. Herein, we supposed the TFEB subcellular localization regulated by NLRX1 might via the mTOR pathway. Our results confirmed our hypotheses and the phosphorylation of mTOR was enhanced in morphine-treated microglia, which could be inhibited by NLRX1 silencing. Recent evidence has demonstrated the association of NLRX1 and mTORC1 activity in lung aging [
45]. Our data shed light into the notion that NLRX1 regulated the lysosomal function might via mTORC1-TFEB signaling. However, further work is required to further demonstrate the interaction of NLRX1 and mTORC1 function.
It has been reported that lysosomal function was essential for maintaining normal innate immunity and pathogen resistance [
46]. The complex interaction between mitochondria and lysosomes is essential for homeostasis of the immune system [
19]. Consistently, we observed the downregulation of inflammatory cytokines (
IL-β,
IL-6,
IL-18,
TNF-α, and
iNOS) in morphine-treated BV2 cells, which indicated immunosuppression after chronic morphine exposure, and this could be rescued by NLRX1-silencing. Hence, this phenomenon supported the viewpoint that NLRX1-mediated incomplete mitophagy led to immunosuppression in morphine-treated microglia. Mdivi-1 was then used as a mitophagy inhibitor to confirm the inflammatory inhibition of mitophagy following morphine treatment. As expected, deficiency of mitophagy significantly increased the expressions of inflammatory cytokines
IL-1β,
IL-6,
IL-18,
TNF-α, and
iNOS in the presence of morphine, while NLRX1-silencing invalidated the effect of Mdivi-1. Therefore, it was confirmed that morphine induced immunosuppression through NLRX1-mediated mitophagy.
Since NLRX1-mediated mitophagy caused by morphine was incomplete, we then aimed to determine whether correcting the lysosomal dysfunction to repair mitophagy could rescue immunosuppression in microglia. As excepted, Torin1 contributed to partial upregulation of inflammatory cytokines (
IL-1β,
TNF-α, and
iNOS), supporting our hypothesis. Unlike CCCP-induced completed mitophagy or bafilomycin A1-induced lysosomal dysfunction alone, morphine was responsible for the extensive inhibition of proinflammatory cytokines and susceptibility to infections in microglia. The coordination of mitochondrial ligands and innate immune sensors, such as TLRs and cGAS/STING, mediated host immune responses to pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs). TLRs signaling pathway has been reported to trigger secretion of inflammatory cytokines, such as
IL-1β,
IL-6, and
TNF-α. The immunogenic capabilities of damaged mitochondria have also been underscored, resulting in excessive mtROS production and accumulated cytosolic mtDNA. NLRP3 inflammasome was subsequently activated and induced accumulation of
IL-1β and
IL-18 [
47]. Inhibiting mitophagy initiation improved immune defense against viruses by enhanced activation of the NLRP3 inflammasome [
48]. Additionally, multiple inflammatory cytokines might serve as downstream factors of
IL-1β, such as
TNF-α,
IL-6, and
iNOS [
49]. Therefore, mitochondria served as platforms of manipulation in integrating complex signals to trigger immune activation [
50]. Lysosomes degraded cytoplasmic constituents including defective mitochondrion for recycling, reconstitution and modification. Insufficient lysosomal function interrupted the mitophagic flux and broke intrinsic immune homeostasis. Additionally, lysosomal maturation served as a critical role in the elimination of invading pathogens and incomplete mitophagy promote intracellular pathogen infection [
51]. Ultimately, incomplete mitophagy potentiated the immune deficiency. Therefore, morphine-treated microglial cells should be more vulnerable to pathogenic challenge such as bacterial LPS. However, the detailed mechanisms underlying how NLRX1-mediated mitophagy facilitated downregulation of inflammatory cytokines required further elucidation.
Thirdly, we successfully generated a mouse model for chronic morphine induced immunosuppression. The suppressed systemic immunity caused by morphine group was observed in peripheral organs by the decline of the spleen index and thymus index. Besides, the inhibition of inflammatory cytokines (IL-1β, IL-6, and iNOS) in the liver and spleen induced by morphine might account for the vulnerability to infections. According to the NLRX1 expression in the ‘HUMAN PROTEIN ATLAS’ (
https://www.proteinatlas.org/ENSG00000160703-NLRX1/brain), we detected the NLRX1 expression and mitophagy in the cortex, striatum, hippocampus, cerebellum, and brainstem to confirm the cellular findings in vivo. We demonstrated that NLRX1-mediated mitophagy was, respectively, enhanced in the cortex, striatum, and cerebellum, after exposure to chronic morphine stimulation. No significant difference was observed in the hippocampus and brainstem between the control and morphine group. Microglial cells, rather than neuron, astroglia or vascular endothelial cells, were then characterized as the main cell type, where NLRX1-mediated mitophagy occurred in brain of morphine-treated mice. Our findings were consistent with previous reports that NLRX1 acted as an enigmatic regulator in immune cells [
52].
Moderate proinflammatory response after infection helps to clean the invading pathogen and therefore transition of M1 microglia to M2 phenotype helps to repair the damage [
53,
54]. Morphine exposure has been demonstrated to disturb the microglial secretion of inflammatory cytokines, such as
TNF-α,
IL-6, and
CCL2/MCP-1 [
55]. The pathological or diseased microglia might lead to the aggravated inflammatory storm or constant presence of inflammation in response to infection [
45,
56,
57]. The immunological exhaustion accounts for the vulnerability of infection and deteriorated inflammatory damage in brain. In addition, microglia in different brain regions displayed diverse characteristics under pathogenic challenge. They have been demonstrated to function as immunoregulatory mediators in the cortex, striatum, and cerebellum [
58,
59]. Herein, it was plausible that NLRX1 functioned as a negative immune regulator and aggravated septic injury in brain of chronic morphine-treated mice, mainly in the cortex, striatum, and cerebellum. There are some issues remained to address. A recent report demonstrated that the brain could modulate adaptive immunity responses directly in immune organs [
60]. In addition, mounting evidence indicate that damage in CNS might contribute to impaired immune system and thus facilitates immunodepression, increasing the risk of infections [
61]. We therefore speculated a possible feedback loops between the brain and the immune system in the mouse after morphine treatment. In addition, whether and how NLRX1-mediated immunodepression in microglia in brain plays a role in peripheral immune system remains unknown.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.