Targeting cyclophilin-D by miR-1281 protects human macrophages from Mycobacterium tuberculosis-induced programmed necrosis and apoptosis

Mycobacterium tuberculosis (MTB) infection induces cytotoxicity to host human macrophages. The underlying signaling mechanisms are largely unknown. Here we discovered that MTB infection induced programmed necrosis in human macrophages, causing mitochondrial cyclophilin-D (CypD)-p53-adenine nucleotide translocator type 1 association, mitochondrial depolarization and lactate dehydrogenase medium release. In human macrophages MTB infection-induced programmed necrosis and apoptosis were largely attenuated by CypD inhibition (by cyclosporin A), silencing and knockout, but intensified with ectopic CypD overexpression. Further studies identified microRNA-1281 as a CypD-targeting miRNA. Ectopic overexpression of microRNA-1281 decreased CypD 3’-untranslated region activity and its expression, protecting human macrophages from MTB-induced programmed necrosis and apoptosis. Conversely, microRNA-1281 inhibition in human macrophages, by the anti-sense sequence, increased CypD expression and potentiated MTB-induced cytotoxicity. Importantly, in CypD-KO macrophages miR-1281 overexpression or inhibition was ineffective against MTB infection. Restoring CypD expression, by an untranslated region-depleted CypD construct, reversed miR-1281-induced cytoprotection against MTB in human macrophages. Collectively, these results show that targeting CypD by miR-1281 protects human macrophages from MTB-induced programmed necrosis and apoptosis.


AGING
This will lead to mitochondrial depolarization, mPTP opening and cytochrome C release. It will eventually promote cell necrosis [7-9, 11, 12, 15, 16]. Other studies proposed that the cascade is also important for initiating cell apoptosis, as cytochrome C releases to the cytosol [17][18][19]. The current study tested whether this pathway participated in MTB-induced death of human macrophages.
MicroRNAs (miRNAs) are a large family of endogenous, short (about 22-nt long) and single-strand non-coding RNAs (ncRNAs) [20,21]. By physically binding to the 3′-untranslated region (3′-UTR) of the targeted mRNA, miRNAs will induce degradation of target mRNAs and/or inhibit gene translation [20,21]. Existing literatures have implied that miRNA dysregulation in the host cells (including macrophages) is extremely important in active and latent TB infection [22][23][24][25]. Our previous study has shown that microRNA-579 (miR-579) upregulation mediated MTB-induced macrophage cytotoxicity [26]. Whether CypD is a target of miRNAs and the molecular regulation of CypD in the necrotic machinery of MTB-infected human macrophages remain to be elucidated. The results of the present study will show that microRNA-1281 (miR-1281) is a CypD-targeting miRNA, and miR-1281 protecting human macrophages from MTB-induced programmed necrosis and apoptosis by silencing CypD.

MTB infection induces mPTP opening and programmed necrosis in human macrophages
Understanding the underlying mechanisms of MTBinduced death of macrophages is vital for the control of MTB infection [6,26]. We tested the possible involvement of mPTP in the process. The mitochondrial immunoprecipitation (Mito-IP) assay results, Figure 1A, demonstrated that with MTB infection, p53 immunoprecipitated with mPTP components CypD and ANT1 [8,27,28]. It is known as the initial step for mPTP opening and programmed necrosis [11,13,14,29,30]. The expression levels of CypD, ANT1 and p53 were not significantly changed in human macrophages ( Figure 1A, "Input"). mPTP opening is often followed with mitochondrial depolarization [11,13,14,29,30]. JC-1 assay results, Figure 1B, demonstrated that mitochondrial depolarization occurred in the MTBinfected human macrophages, showing JC-1 green fluorescence accumulation ( Figure 1B). Furthermore, the medium LDH contents were significantly increased in human macrophages with MTB infection ( Figure  1C), indicating programmed necrosis [11,13,14,29,30]. Together, these results suggested that MTB infection induced mPTP opening and programmed necrosis in human macrophages. macrophages were infected with Mycobacterium tuberculosis (MTB) for applied time periods, mitochondrial immunoprecipitation (Mito-IP) assays were carried out to test CypD-ANT1-p53 association in the mitochondria (A, "Mito-IP"), with expression of these proteins examined by Western blotting (A, "Input"); Mitochondrial depolarization was examined by JC-1 dye assay (B); Cell necrosis was tested by medium LDH release assays (C). For JC-1 assays, both JC-1 merged images and JC-1 green fluorescence intensity were presented (same for all Figures). Expression of listed proteins was quantified, normalized to loading controls (A). "C" stands for uninfected control macrophages (same for all Figures). Data were presented as mean ± SD (n=5), and results were normalized to "C". * P <0.05 vs. "C" macrophages. Experiments in this figure were repeated five times with similar results obtained. Bar= 100 μm (B). AGING
We further hypothesized that ectopic overexpression of CypD should facilitate MTB-induced cytotoxicity of human macrophages. The lentiviral CypD expression construct was transduced to human macrophages. Via selection by puromycin the stable cells were established, showing over five-folds CypD mRNA expression (vs. vector control cells, Figure 2H). CypD protein levels were significantly increased as well ( Figure 2I). Importantly, ectopic CypD overexpression potentiated MTB-induced mitochondrial depolarization ( Figure 2J), viability reduction ( Figure 2K) and medium LDH release ( Figure  2L). TUNEL staining results demonstrated that CypD overexpression significantly enhanced MTB-induced apoptosis activation ( Figure 2M). MTB infection did not affect CypD expression in human macrophages ( Figure  2A and 2H). Without MTB infection, CypD inhibition, silencing, KO or overexpression did not affect the functions of human macrophages ( Figure 2C-2G, 2J-2M). These results show that inhibition of the CypD-mPTP pathway largely attenuated MTB-induced death of human macrophages.

miR-1281 overexpression inhibits MTB-induced programmed necrosis and apoptosis in human macrophages
Since miR-1281 targets and downregulates CypD, it would then protect human macrophages from MTBinduced cytotoxicity. The lv-pre-miR-1281-expressing human macrophages (see Figure 3) and control macrophages with non-sense microRNA ("lv-C") were infected with MTB. As shown, miR-1281 overexpression potently inhibited MTB-induced mitochondrial depolarization, or JC-1 green AGING
Data were presented as mean ± SD (n=5), and results were normalized to "C". * P <0.05 vs. "C" treatment in "Pare" macrophages. # P <0.05 vs. MTB treatment in "Pare" macrophages. Experiments in this figure were repeated four times with similar results obtained.
Next, the 3′-UTR-depleted CypD construct was transfected to human macrophages, completely restored CypD mRNA and protein expression in macrophages with lv-pre-miR-1281 ( Figure 6E). As shown lv-pre-miR-128-induced macrophage protection against MTB was completely reversed with re-expression of the 3′-UTR-depleted CypD ( Figure 6F and 6G). Thus, with CypD re-expression MTB-induced viability reduction ( Figure 6F) and cell death ( Figure 6G) were restored even with miR-1281 overexpression. The qPCR assay results, Figure 6H, demonstrated that 3′-UTR-depleted CypD did not alter miR-1281 expression. These results together indicate that CypD should be the important target of miR-1281 in human macrophages.

DISCUSSION
Necrosis is a common form of cell death characterized by cell swelling, plasma membrane fracture and lysis of the intracellular components and cellular organelles. The traditional concept is that necrosis is a form of accidental, unregulated and passive cell death, while apoptosis is the sole form of "programmed cell death"   [10, 37,38]. Yet recent studies have shown that certain necrosis is also programmed and actively regulated [7,10,[37][38][39]. In the present study we show that MTB infection led to programmed necrosis in human macrophages, causing CypD-p53-ANT1 mitochondrial association, mitochondrial depolarization and LDH release (to the medium). Importantly programmed necrosis, together with apoptosis, could be vital for MTB infection-induced cytotoxicity in the human macrophages.
CypD is the prolyl isomerase and the key component forming mPTP, along with ANT1 and the voltagedependent anion channel (VDAC) [13,14,29]. Studies have shown that CypD lies in the center to mediate the pore opening. CypD inhibition or depletion will result in inhibition on mPTP formation and opening [13,14,29]. Since mPTP opening is vital for programmed necrosis, CypD is essential in regulating necrotic cell death pathway [13,14,29]. In the present study we show that CypD is vital for MTB-induced cytotoxicity to human macrophages. MTB infection-induced programmed necrosis and apoptosis were largely attenuated with CypD inhibition (by CsA), silencing (by shRNA) and KO (using CRISPR/Cas9 method), but intensified with ectopic overexpression of CypD. Therefore, targeting CypD-mPTP pathway could be a novel strategy to protect human macrophages from MTB infection-induced cytotoxicity.
Our results imply that miR-1281 inhibited MTBinduced cytotoxicity to the human macrophages. First, lv-pre-miR-1281 largely attenuated programmed necrosis and apoptosis in MTB-infected macrophages. Conversely, miR-1281 inhibition, by antagomiR-1281, protected human macrophages from MTB-induced cytotoxicity. These results imply that miR-1281 offers cytoprotection against MTB infection in human macrophages. Further analyses show that CypD is the primary target gene of miR-1281 in MTB-infected macrophages. Neither miR-1281 overexpression nor miR-1281 inhibition was able to change MTB-induced cytotoxicity in CypD-KO macrophages. Importantly, restoring CypD expression, by the UTR-depleted CypD construct, reversed miR-1281-induced macrophage protection against MTB infection.
Collectively, these results show that targeting CypD by miR-1281 protects human macrophages from MTBinduced programmed necrosis and apoptosis.

Chemicals and reagents
Puromycin, cyclosporin A (CsA), terminal deoxynucleotidyl transferase (TdT)-mediated Dutp nick-end labeling (TUNEL), DAPI and JC-1 dyes were obtained from Sigma-Aldrich (St. Louis, MO). The antibodies were from Cell Signaling Tech (Danvers, MA). From Invitrogen-Thermo Fisher (Shanghai, China) the cell culture reagents, the Trizol reagents and other RNA assay reagents, as well as the cell transfection reagents were obtained. All the sequences, viral constructs and gene products were provided and verified by Shanghai Genechem Co. (Shanghai, China) or otherwise mentioned.

Primary human macrophages.
As described early [26], from the peripheral blood mononuclear cells (PBMCs) of a written-informed consent donor the primary human macrophages were differentiated [42] and cultured under the described protocol [42]. The primary macrophages were always utilized at passage 3-10. The protocols of the present study were approved by the Ethics Committee of Tongji University School of Medicine.

MTB infection
As described early [26], at 2×10 5 cells per well the primary human macrophages were cultured into sixwell plates and then infected with MTB (multiplicity of infection/MOI 10). After 4h the infected macrophages were washed and returned back to the fresh medium.

Mitochondrial Immunoprecipitation (Mito-IP)
As described previously [18], human macrophages with MTB infection were harvested and homogenized by the lysis buffer provided by Dr. Wang at Soochow University [18]. After centrifugation, the supernatants were collected and suspended. The pellets were then re-suspended in the above buffer plus NP-40, forming the mitochondria fraction lysates. The quantified mitochondrial lysates (500 μg per sample) were precleared and incubated with anti-CypD antibody [28,43], with the mitochondrial CypD-p53-ANT1 complex captured by the protein IgG-Sepharose beads (Sigma), and tested by Western blotting.

Quantitative real-time PCR (qPCR)
Total cellular RNA was extracted by the Trizol reagents from MTB-infected macrophages, with the RNA concentrations determined using the NanoDrop system. From each treatment 100 ng total RNA was utilized for the reverse transcription using the described protocol [26]. The detailed procedures for qPCR were described previously [26], with the melt curve analyses performed. Quantification of targeted genes was through the 2 −ΔΔCt method, using GAPDH as the internal control. miR-1281 expression was normalized to U6. From Shanghai Genechem the primers for U6 and GAPDH were obtained, with other primers for miR-1281, CypD and ANT1 listed in Table 1.

Western blotting
The detailed procedures for the Western blotting assay were reported early [26]. In brief, with the applied treatments, 30 μg total lysates (of each lane) were separated by sodium dodecyl sulfate-polyacrylamide gels, thereby transferred to the polyvinylidene difluoride (PVDF) blots (Merck-Millipore). After blocking the blots were incubated with the primary and secondary antibodies, and detected using the enhanced chemiluminescence (ECL) kit (Pierce, Rockford, IL).

Cell viability
Macrophages were plated at 3×10 3 cells per well onto the 96-well tissue-culture plates. Following the indicated treatments the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan) reagent (10 μL in each well) was added. After 2h, the CCK-8 absorbance at 450 nm was tested through a spectrophotometer (Thermo Fisher Scientific, Vantaa, Finland).

Cell necrosis
Cell necrosis was tested through assaying the medium lactate dehydrogenase (LDH) contents by a two-step easy enzymatic reaction LDH kit (Takara, Tokyo, Japan). Medium LDH contents were always normalized to total LDH levels.

TUNEL staining
Following MTB infection, the human macrophages were co-stained with TUNEL and DAPI dyes (Sigma). The apoptotic nuclei percentage (TUNEL/DAPI×100%) was calculated, from at least 500 cells of five random views (1: 100 magnification).

JC-1 assay
As described previously [26], the human macrophages with the indicated treatment were stained with JC-1 (5 μg/mL, for 10-15 min) and washed. JC-1 green fluorescence, indicating mitochondrial depolarization, was tested at 550 nm using the RF-5301 PC fluorescence spectrofluorometer (Shimadzu, Tokyo, Japan). Furthermore, the representative JC-1 fluorescence images were taken, merging the green fluorescence image (at 550 nm) and the corresponding red fluorescence image (at 650 nm).

CypD short hairpin RNA (shRNA)
The CypD shRNA (with the target sequence, CCCG TCCTCTTCCTCCTCCTCCG) lentiviral particles and the control shRNA lentiviral particles were provided by Dr. Xu [46]. Human macrophages were plated onto sixwell plates (in polybrene-containing complete medium), transduced with the applied shRNA lentivirus particles. After 48h, puromycin was added to select stable cells (for 10-12 days), with CypD silencing verified by qPCR and Western blotting assays.

CypD knockout (KO)
The small guide RNA (sgRNA) against human CypD (target DNA sequence, GGCGACTTCACCAACCA CAA) was selected from Dr. Zhang's laboratory (http://crispr.mit.edu/), and inserted into the lentiCRISPR-green fluorescent protein (GFP) plasmid (from Dr. Zhao at Shanghai Jiao Tong University) with the puromycin selection gene. The construct was transfected to the human macrophages by Lipofectamine 2000, with macrophages subjected to FACS-mediated GFP sorting and selected by puromycin (3.0 μg/mL) to achieved stable cells. CypD KO was verified by qPCR and Western blotting assays. Control cells were transfected with the empty vector.

Ectopic CypD over-expression
The CypD expression (with no 3′-UTR region) pSuperpuro-Flag vector, provided by Dr. Xu [46], was transfected to human macrophages by the Lipofectamine 2000 protocol (Invitrogen, Suzhou, China). The macrophages were then selected by puromycin for 10 days to achieve stable cells, with CypD overexpression confirmed by qPCR and Western blotting assays.

Statistical analyses
Data in the present study were shown as mean ± standard deviation (SD). Statistical analyses were carried out by the SPSS 20.0 software (SPSS Co., Chicago, CA), using oneway analysis of variance of post hoc Bonferroni test as comparisons of multiple groups. The Student T Test was utilized for comparison between two groups. Statistically differences were assigned to P < 0.05.

AUTHOR CONTRIBUTIONS
All authors listed carried out the experiments, participated in the design of the study and performed the statistical analysis, conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.

CONFLICTS OF INTEREST
None of the authors has any conflicts of interests to declare.