Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 9.
Published in final edited form as: Tuberculosis (Edinb). 2020 Dec 30;126:102046. doi: 10.1016/j.tube.2020.102046

NIX-mediated mitophagy regulate metabolic reprogramming in phagocytic cells during Mycobacterial infection

Ranjeet Singh Mahla 1,#, Akhilesh Kumar 1,#, Helena Tutil 2, Sreevidhya Tarakkad Krishnaji 3, Bharathwaj Sathyamoorthy 3, Mahdad Noursadeghi 2, Judith Breuer 2,4, Amit Kumar Pandey 5, Himanshu Kumar 1,6,*
PMCID: PMC7619460  EMSID: EMS217698  PMID: 33421909

Abstract

RNASeq analysis of PBMCs from treatment naïve TB patients and healthy controls revealed that M. tuberculosis (Mtb) infection dysregulates several metabolic pathways and upregulates BNIP3L/NIX receptor mediated. mitophagy. Analysis of publicly available transcriptomic data from the NCBI-GEO database indicated that M. bovis (BCG) infection also induces similar rewiring of metabolic and mitophagy pathways. Mtb chronic infection and BCG in-vitro infection both downregulated oxidative phosphorylation and upregulated glycolysis and mitophagy; therefore, we used non-pathogenic mycobacterial species BCG as a model for Mtb infection to gain molecular insights and outcomes of this phenomenon. BCG infection in PBMCs and THP-1 macrophages induce mitophagy and glycolysis, leading to differentiation of naïve macrophage to M1 phenotype. Glucose consumption and lactate production were quantified by NMR, while the mitochondrial mass assessment was performed by mitotracker red uptake assay. Infected macrophages predominantly exhibit M1-phenotype, which is indicated by an increase in M1 specific cytokines (IL-6, TNF-α, and IL-1β) and increased NOS2/ARG1, CD86/CD206 ratio. NIX knockdown abrogates this upregulation of glycolysis, mitophagy, and secretion of pro-inflammatory cytokines in BCG infected cells, indicating that mycobacterial infection-induced immunometabolic changes are executed via NIX mediated mitophagy and are essential for macrophage differentiation and resolution of infection.

Background

Tuberculosis is the leading cause of morbidity and mortality from a single infectious agent. Despite cures’ availability, approximately 4000 people die every day from tuberculosis, imposing a significant threat to global health (1). The disease is caused by an intracellular pathogen, M. tuberculosis (Mtb), that possesses the incredible ability to persist for a long time within host cells by evading the immune system. The early immune response to infection is mediated by innate immune cells, including neutrophils, dendritic cells, and alveolar or tissue macrophages (2). Due to difficulty in elimination of Mtb by macrophages, these cells become the primary cellular niche of Mtb infection, and therefore the progression and outcome of the disease are primarily directed by the early interaction between macrophages and Mtb (3, 4).

Macrophages are the sentinels of tissue homeostasis and are characterized by a high degree of diversity and plasticity. During their routine surveillance, they ingest surrounding material and depending on the microenvironmental cues; macrophages can differentiate into either M1 or M2 phenotypes. The M1 phenotype is portrayed by secretion of high pro-inflammatory cytokines, production of reactive oxygen and nitrogen intermediates, and promotion of Th1 response leading to intense microbicidal activity. In contrast, M2 macrophages produce anti-inflammatory cytokines (IL-10) and are considered to be involved in tissue remodeling and immunoregulation during normal physiological state (5, 6). These extreme heterogeneities of M1 and M2 macrophages result from dynamic transcriptional and metabolic plasticity. M1 macrophages exhibit a metabolic switch from mitochondrial oxidative phosphorylation to aerobic glycolysis indicated by high glycolysis and pentose phosphate pathway flux and low mitochondrial oxidative phosphorylation (OXPHOS) and tri-carboxylic acid pathway (TCA) capabilities.

In contrast, M2 macrophages majorly rely on fatty acid oxidation, OXPHOS, and TCA (7). In TB patients, alveolar lavage, sputum, and bronchoalveolar fluid show elevated levels of IL-10 (8, 9), probably secreted by predominant M2 macrophages in granulomas (10, 11), indicating that Mtb can skew M1/M2 polarization. According to recent studies, Mtb early secretory antigenic target (ESAT-6) first induced M0 to M1 macrophage polarization and finally induced M1 to M2 polarization (10). Region of difference-1 (RD-1), which encodes for ESAT-6, is present in Mtb but absent in M.bovis vaccine strain bacillus calmette-guerin (BCG). However, both Mtb and BCG show metabolic reprogramming in infected macrophages (12). The mechanism of how Mtb modulates immune cell metabolism to facilitate its survival is poorly understood.

Reprogramming of cellular metabolism, in particular, glucose and lipid metabolism of innate immune cells during Mtb infection, regulate differentiation and effector functions (13, 14). Recent studies indicate that programmed mitophagy is essential for differentiation or dedifferentiation of cells, and mitophagy-induced changes in mitochondrial dynamics facilitate metabolic reprogramming (13, 15). Programmed mitophagy is characterized by a rapid and co-ordinated clearance of functional mitochondria, regulated by a series of receptor and adaptor molecules, and has two primary arms; PINK1/Parkin mediated mitophagy and receptor-mediated mitophagy. PINK1/Parkin pathway begins with the accumulation of PINK1 kinase in OMM, PINK1 phosphorylates Parkin (E3 ubiquitin ligase). In turn, Parkin ubiquitinates multiple outer mitochondrial membrane (OMM) proteins and gets identified by adaptor proteins SQSTM1, OPTN, and NDP52, which ultimately leads to the recruitment of LC3 mitophagy receptors.

On the other hand, receptor-mediated mitophagy begins with accumulating specific OMM receptors; NIX/BNIP3L, BNIP3, and FUNDC1. These receptors contain LIR motifs to interact with LC3 molecules that promote direct engulfment of mitochondria. Out of these, HIF-1α-responsive gene NIX is known to regulate glycolysis and fatty acid metabolism (13, 16, 17) and could prove to be a crucial link between mitophagy and metabolic reprogramming during pulmonary tuberculosis.

In this study, we have performed whole-transcriptome profiling of peripheral blood mononuclear cells (PBMCs) isolated from treatment naïve tuberculosis patients and healthy controls to understand the dynamics of activation of innate immune cells, especially the macrophages during the active tuberculosis infection. Both Mtb and BCG show similar metabolic rewiring of cells coupled to the activation of mitophagy. Therefore, we have used non-pathogenic mycobacterium species, BCG as a model for Mtb infection. We have deciphered one of the underlying molecular processes vital for macrophage differentiation and function, which determines the outcome of tuberculosis infection and may be crucial for BCG-mediated protection.

Methods

Ethical Statement

This study was approved by the institutional ethical committee (IEC), Indian Institute of Science Education and Research Bhopal, India. Necessary permissions were provided by the district tuberculosis officer for the collection of blood samples from diagnostically confirmed pulmonary tuberculosis patients at the outpatient department, a district tuberculosis hospital, Bhopal, Madhya Pradesh, India.

Sample Collection

Blood samples were collected at the outpatient department, district tuberculosis hospital Bhopal (Madhya Pradesh, India) from therapy naïve TB patients diagnosed by sputum smear microscopy and chest x-ray examination. The same geographical region reference controls samples collected from asymptomatic healthy controls. All study participants provided written, informed consent per the Declaration of Helsinki ethical principles (18).

PBMCs isolation and cell culture

For in-vitro experiments, we used peripheral blood mononuclear cells (PBMCs) and promonocytic THP-1 (ATCC TIB-202). PBMCs purification was carried out using Histopaque-1077 (Sigma Aldrich; Cat. No. 10771-6×100ML) density gradient, as per manufacturer’s protocol. In brief, equal volume of blood layered on Histopaque-1077 in 15 ml centrifuge tubes and cells were sepereated by density gradiant at 650 rcf for 30 minutes. PBMCs (the opaque layer) transferred into fresh 15 ml centrifuge tubes and washed centrifuged in RPMI twice at 250 rcf for 10 minuted. As per the experiment requirements, purified PBMCs were plated for 2h in tissue culture dishes and non-adherent cells washed and removed. Both PBMCs and THP-1 were cultured in RPMI-1640 (Invitrogen) at 37 °C in a 5% CO2 atmosphere in a humidified cell culture incubator (ThermoFisher Scientific). For growth and nutrition, THP-1 supplemented with 10% fetal bovine serum (FBS), and PBMCs were supplemented with 10% (volume/volume) autologous human serum. Promonocytic THP-1 cells differentiated to M0 macrophages for two days, using 30nM phorbol 12-myristate 13-acetate (PMA; Sigma Aldrich; Cat. No.-P1585), and rested for 24hr.

MTT assay

The PMA differentiated THP1 macrophages (4000 cells/well of 96 well plates) were either left uninfected or infected with BCG (10-MOI). At 2, 4, and 6 days post-infection, cells were processed for viability analysis using colorimetric MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Alfa Aesar by ThermoFisher Scientific L11939) assay. Prior to assay, cells were washed PBS removing cell debris, and treated with MTT (2mg/ml) in RPMI for 1h at 37°C (19). The blue-colored formazan accumulated inside the cell following cleavage of MTT by mitochondrial dehydrogenase of the viable cells were solubilized in 200μL of DMSO, and absorbance was taken at 595nm having background correction at 700nm.

BCG Culture and Infection

M. bovis (BCG; ATCC-35734) cultures maintained to log phase in Middlebrook 7H9 broth (Himedia; M198-500G), supplemented with OADC (Himedia_FD018). BCG culture aliquots sonicated in a water bath sonicator and centrifuged at 5000g. Cell pellets were reconstituted in RPMI-1640 and left idle for 5 minutes; the upper half volume was used for infection to PBMCs and THP-1. The multiplicity of infection (MOI) was calculated from OD600. All infections were performed with 10 MOI of BCG (20).

Glucose uptake assay

Differentiated THP-1 macrophages were either left uninfected or infected with BCG (10 MOI) for 24h in RPMI-1640, supplemented with 2% FBS. Post-infection, cells supplied with 10% serum-containing RPMI-1640. Six days post-infection, cells were washed by PBS and incubated in serum and glucose-free RPMI for 2h. Serum and glucose starved cells were treated with 10μM 2-NBDG (Invitrogen) for 10min at 37 °C (21). Uptake of 2-NBDG was analyzed on FITC channel BD-FACS Aria-II.

MitoTracker Red staining

Six-days post-infection, mitochondria in THP-1 cells in growth medium (10% serum-supplemented RPMI1640) were labeled with 50 nM MitotrackerRed CMXRos (Invitrogen; Cat. No. M7215) for 15 minutes at 37°C (22–24). Mitotracker uptake was analyzed on BD-FACS Aria-II in the PE-Texas-Red channel.

RNA-Seq library preparation and data analysis

Healthy controls (n=4) and treatment naïve tuberculosis patients (n=5) procured PBMCs total RNA was isolated using the RNeasy Mini Kit (QIAGEN; Cat No./ID: 74104) and treated with DNAse. Following rRNA-depletion, samples (RIN score ≥ 7) were processed for library preparation using TruSeq RNA Library Prep Kit v2 following manufacturers’ instructions. Library preperation steps involved RNA fragmentation, quality check (QC), 1st strand synthesis, 2nd Strand synthesis, end-repair, A-tailing, adaptor ligation, and multiplex indexing (in accordance to UCL Genomics inhouse protocol). Following QC, indiced libraries were pooled together and sequenced (pair-end) using Illumina HiSeq2500, with 75X coverage, and the adaptor trimmed read length was constant 41 bp. Fastq file read quality checked using FastQC v0.11.8. Low-quality sequences were trimmed and removed adaptor sequences using Trimmomatic (25). Reads were mapped against the reference human genome HG38 using HISAT2 v2.1.0 (26). Cufflinks (27) software used for transcript assembly and estimation of gene abundances. The DESeq2 R package was used for the analysis of differential expression of genes. PCA plot, volcano plot, and heat maps were generated using R packages. GO analysis was done using WebGeStalt, gene set analysis toolkit (28), and KEGG pathways analysis for DEGs was performed using Enricher (web-based interface) (29).

Quantitative real-time PCR

Total RNA was extracted using TRIzol reagent (Life Technologies). A total of 1 μg of RNA was convereted to cDNA using iScript cDNA synthesis Kit (BIO-RAD) according to manufacturers’ instructions. The relative expression of mRNA expression calculated in reference to housekeeping genes 18S ribosomal rRNA. Reference controls were used for gene expression fold change calculation.Gene-specific primers (Supplementary Table-1) were used to amplify selected amplicons in SYBR green using StepOnePlus Real-Time PCR System (Applied Biosystems). The relative abundance of the transcript was quantified using StepOnePlus software.

NMR Spectroscopy

PBMCs from healthy donors (n=4) seeded at a density of 1 ×106 cells/well in 24 well plates and immediately infected with BCG (10-MOI) for 24h in 2% serum. Post-infection, cells were treated with 100% U-13C6-glucose (2g/L) in 10% serum-supplemented glucose-free RPMI. Samples; infected and controls harvested at 0h, 6h, 12h, 18h, 24h, 36h, 48h, 96h, 144h, and 192h time points, culture supernatants filtered through 0.22 μM syringe filters. Samples lyophilized overnight at -50ºC, 80mbar, reconstituted in 1.5 mL NMR buffer. Reconstituted samples (500 μL) transferred to a 5mm NMR tube, and data acquired at T = 298K on a 700 MHz proton frequency Bruker spectrometer equipped with AVANCE III console and cryogenic triple resonance 1H, 13C, (15N) probe with z-axis gradients. Single-pulse 13C one-dimensional NMR data was acquired using the simple one 90º pulse followed by direct detection on 13C channel with 1H decoupling (30).

The 13C and 1H carriers were set at 100ppm and 3.5ppm, with a spectral width of 236.65ppm (64k points). Recycle delay between scans and acquisition time was set to 10s and 786ms. Data were processed using TopSpin software (Bruker Inc., version-4.0.7). Data was zero-filled to 128 thousand points, apodized using exponential window function (5Hz broadening). Following Fourier transformation, the observed resonances were integrated (sum of all resonances set to 100) and base-line corrected using them. Quantitation of glucose and lactate performed using the integrals obtained for C5 (69.6ppm) and C2 (68.5ppm) carbons for glucose and lactate, respectively. Recycle-delay coupling avoided, and choice of acquisition parameters was set to obtain reliable integration to quantitatively interpret the results (adjudged by getting an anomeric ratio at 298K using C1 carbon position in glucose and combination of all carbons in glucose yielding same integral). The integrals were then converted to percentage concentration using the initial essential value (16.7, i.e., 100/6) of the 0th hour time point as one equivalent of glucose yields two equivalents of lactate, for one to one comparison of glucose and lactate, the concentration of lactate divided by a factor of two.

ELISA

Cytokines profiling from culture superannuants (of BCG infected and uninfected controls) performed according to manufacturer’s (BD, Biosciences, USA) protocol cytokines specific ELISA kits for TNF-α (Cat#-555212), IL-6 (Cat#-555220), and IL-1β (Cat#-557953). Briefly, ELISA plate (Falcon 351172) microwells coated with the capture antibody (1:250 dilution in coating buffer Cat. No. 51-2713KC) 100 μL/well for overnight at 4°C. Microwells washed aspirated thrice with 300 μL/well wash-buffer (BD OptEIA 51-900739), and wells were blocked with 200 μL/well assay diluent for 1h at RT. Microwells washed thrice in wash buffer, samples, and standards in assay diluent 100 μL/well added for 2h at RT. Microwell washed aspirated 5-times with wash buffer, added detection antibody (100 μL/well in assay diluent), and incubated for 1h at RT. Microwell washed 5-times with wash buffer 300 μL/well, added enzyme reagent 100 μL/well for 30 minutes at RT. Microwell washed 7-times with wash buffer 300 μL/well, added substrate reagent A (Cat.No.51-2606KG), and B (Cat.No.51-2606KG) solution (1:1:: volume: volume) 100 μL/well for 10-30 minutes at RT. Before achieving saturation, the calorimetric reaction was stopped using stop solution 50 μL/well, and within 30 minutes, absorbance was read at 450 nm with wavelength correction at 570 nm.

Statistical Analysis

Statistical analysis and data plotting were done using GraphPad Prism 7.0 software. A two-tailed unpaired t-test was performed on data involving two independent experimental groups, while two-way ANOVA was performed on data involving two or more related experimental groups.

Results

Altered gene signature in PBMCs of Tuberculosis patients

We performed global gene expression profiling of peripheral blood mononuclear cells (PBMCs) from healthy individuals and treatment naïve tuberculosis patients (Supplementary Figure 1A) by RNASeq. We observed considerable differences between healthy and disease samples in terms of gene expression, as shown in the density plot of transformed data and principal component plot (Supplementary Figure 1B). Differentially expressed genes (Figure 1A) were subjected to gene ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. KEGG pathway analysis indicated the upregulation of autophagy, mitophagy, toll-like receptor, HIF-1α, TNFα, NF-κB signaling, and glycolytic pathways in TB patients, compared with healthy controls. In contrast, pathways for base excision repair, fatty acid degradation, and alpha-linolenic acid metabolism were upregulated (Figure1B). The upregulated biological processes also indicated autophagic removal of mitochondria and reprogramming of metabolic pathways (Supplementary Figure 1C). Enzymes for glycolysis and pentose phosphate pathway were upregulated, while those associated with fatty acid metabolism, TCA cycle, and oxidative phosphorylation were downregulated (Figure 1C). The expression of selected genes involved in the pathways mentioned above was quantified by qRT-PCR in PBMCs from a separate validation cohort of 22 healthy controls and 22 TB patients (Figure 1D). It was found to be in perfect agreement with quantification using RNASeq (Supplementary Figure 1D). An IFN-γ ELISA was performed using plasma samples collected at the time of PBMCs isolation of study participants in RNASeq (n=9) analysis and qRT-PCR analysis (n=44) to differentiate treatment naïve tuberculosis patients and healthy controls. Active infection of Mtb is associated with an increased circulatory lIFN-γ level, whereas levels of lIFN-γ are low in healthy controls (Supplementary Figure S1E). We then analyzed transcriptomic data obtained from human monocytes derived macrophages infected by BCG, available at NCBI-GEO database (GSE3408). We found that CD14+ monocytes differentiated with GMCSF for 7-days, infected with BCG for 24h leads to metabolic reprogramming in infected cells. The observed similarities reflect that irrespective of pathogenicity, both mycobacterial species cause the upregulation of genes associated with carbon metabolism (Supplimentary Figure 2)

Figure 1. PBMCs from tuberculosis patients indicate metabolic reprogramming and elevated mitophagy.

Figure 1

Open in a new tab

RNASeq analysis was performed to compare the whole transcriptome of PBMCs from TB patients and healthy controls, (A) Volcano plot representing the differential expression of genes; genes upregulated or downregulated in TB patients (n=5) are indicated in red and blue color respectively. (B) Heat maps showing top dysregulated KEGG pathways and (C) genes involved in these pathways between healthy controls (n=4) and TB patients (n=5). (D) qRT-PCR analysis of selected genes in the validation cohort consisted of TB patients (n=22) and healthy controls (n=22).

Mycobacterial infection induces metabolic reprogramming of infected cells

To further validate the results obtained from PBMCs of TB patients, we infected THP-1 cells and PBMCs isolated from healthy individuals with BCG. We found that glycolytic enzymes were upregulated while the TCA cycle enzyme SDHA was not affected in PBMCs (Figure 2A) and THP-1 cells (Supplementary Figure 3A). Gene expression results obtained in THP-1 cells are irrespective of cell death as measured BCG infection does not induce cell death in THP1 cells (Supplementary Figure 3C). Glucose transporter SLC2A was found upregulated in RNASeq data (Figure 1C); therefore, we decided to examine glucose uptake in macrophage cells upon BCG infection. We cultured cells in the presence of 2-NBDG, a fluorescently labeled analog of glucose. As expected, the uptake of 2-NBDG in BCG-infected PBMCs was much higher compared to uninfected cells (Figure 2B).

Figure 2. Metabolic reprogramming during tuberculosis infection.

Figure 2

Open in a new tab

(A) PBMCs from healthy individuals (n=3) were either infected with BCG or left uninfected for the indicated time, and mRNA expression of several glycolysis enzymes was quantified by qRT-PCR; data states fold-change of mRNA expression in BCG infected samples relative to same-day uninfected controls in PBMCs isolated from three different healthy individuals. (B) THP-1 cells were either infected with BCG or left uninfected. On the sixth-day, uptake of 2-NBDG was analyzed by flow cytometry; fluorescence was analyzed in the FITC channel (presented data are representative of three independent experiments). (C) BCG infected and uninfected PBMCs (n=4 healthy donors) were cultured in the presence of 100% U-13C6-labeled glucose for the indicated time; the supernatant was collected and was analyzed by 13C 1D NMR spectroscopy for glucose consumption and lactate production.

Furthermore, to study glucose metabolism’s kinetics in BCG-infected cells, we analyzed glucose and its metabolites by nuclear magnetic resonance (NMR) spectroscopy. From the 13C 1D NMR spectra, it is evident that the lactate that is formed is from the 13C6 source provided. Change in glucose concentration and lactate as a function of time is shown in (Figure 2C). Compared to uninfected cells, BCG-infected cells utilized more glucose and induced glycolysis at a much faster rate indicated by higher 13C-glucose uptake and higher 13C-lactate accumulation in the cell culture supernatant. Interestingly, the consumption of 13C6-glucose is five times higher for the infected cells than the control, so is the production of 13C3-lactate. The infected cells start to show increased consumption from the 12h onwards and reach saturation in four days (~50%). Besides, we also observe the formation of 13C2-alanine (4% of overall concentration) that is being currently further probed in our lab.

Mycobacterial infection skews macrophage maturation towards M1 phenotype by inducing mitophagy

We found that many genes responsible for the regulation of mitophagy were upregulated in TB patients. The mitophagy pathway has two primary arms; PINK1/Parkin mediated mitophagy and receptor-mediated mitophagy (Figure3A). We noticed that TB patients showed upregulation of the receptor-mediated mitophagy genes, and there was no significant effect on PINK1/Parkin pathway genes (Figure 3B). We infected PBMCs and THP-1 cells with BCG and estimated mRNA expression of mitophagy pathway genes at 2, 4, and 6 days post-infection. We found the upregulation of genes associate with receptor-mediated endocytosis LC3A, LC3B, LC3C, NIX, BNIP3. At the same time, there was no significant change in expression of PINK1 and Parkin mRNA in BCG infected PBMCs (Figure 3C) and THP-1 cells (Supplementary Figure 3B). These results confirm that BCG infection upregulates only the receptor-mediated arm of mitophagy. Next, we infected PMA differentiated THP-1 cells with BCG (10 MOI) and labeled mitochondria in live cells using MitoTracker Red. Flow cytometric analysis revealed that cells infected with BCG retain less MitoTracker Red, indicating fewer functional mitochondria than uninfected cells (Figure 3D).

Figure 3. Tuberculosis infection skew receptor-mediated mitophagy in PBMCs.

Figure 3

Open in a new tab

(A) Illustration of two significant mitophagy mechanisms, receptor-mediated mitophagy and adaptor mediated mitophagy. (B) Expression of various mitophagy associated genes in healthy controls (n=4) and TB patients (n=5) estimated using RNASeq. (C) mRNA expression of selected mitophagy associated genes in BCG infected PBMCs at indicated time points, data represented as fold-change of mRNA expression in BCG infected group (n=3), compared with the same day uninfected controls in PBMCs (n=3). (D) THP1 cells were either infected with BCG or left uninfected; after eight days, cells were treated with MitoTracker Red and analyzed using flow cytometry; MitoTracker Red fluorescence quantified in PE Texas red channel (data are representative of three independent experiments). (E) Quantification of TNF-α, IL-1β, and IL-6 by ELISA in culture supernatants from BCG infected or uninfected PBMCs after 2, 4, and 6 days post-infection (n=3 healthy donor) (F) qRT-PCR and ratio of CD86/CD208 and NOS2/ARG1 quantified from mRNA expression of CD86, CD208, NOS2, and ARG1, at 2, 4, and 6 days after BCG infection in infected and uninfected THP-1 cells. (data are representative of three independent experiments)

Interestingly, we noticed that modulation of the pathways, as mentioned above, either upregulation of mitophagy, pentose phosphate pathway and glycolysis or downregulation of fatty acid metabolism, TCA cycle, or oxidative phosphorylation all match the characteristics of M1 macrophages. Next, we asked if this could result from skewing of PMA differentiated M0 macrophages into the pro-inflammatory M1 phenotype because of BCG infection. We analyzed the expression of M1 cytokines released in the culture supernatant by ELISA. We found many folds elevated levels of IL-6, IL-1β, and TNF-α in BCG infected THP-1 cells (Figure 3E). Expression of NOS2, ARG1, CD86, and CD206 was calculated by qRT-PCR; the ratio of NOS2/ARG1 and CD86/CD206 was found increased during BCG infection in THP-1 cells (Figure 3F). The results indicate that M0 macrophages are predominantly differentiating into pro-inflammatory M1 macrophages in response to BCG infection.

NIX is essential for mitophagy and subsequent differentiation of naïve macrophages into M1 phenotype during mycobacterial infection

BCG infected PBMCs, and THP-1 cells showed upregulated expression of several genes associated with receptor-mediated mitophagy; among them, NIX was upregulated the most. Therefore, we decided to investigate the role of NIX in tuberculosis-induced mitophagy. We performed transient knockdown (KD) of NIX using gene-specific siRNA in PBMCs (Figure 4A) and analyzed the expression of genes associated with mitophagy and metabolic pathways at 2- and 4-days post-BCG infection. NIX-KD abolished the induction of LC3A, LC3B, LC3C, and BNIP3L upon BCG infection (Figure 4B). Also, there was no significant change in MitoTracker Red staining (Figure 4C) in BCG infected NIX-KD THP-1 cells compared to uninfected cells, indicating restoration of normal mitophagy following NIX-KD. Furtheremore, the expression of glycolysis pathway genes (LDHA, ENO1, PFKFB3, and GAPDH) decreased in NIX-KD BCG infected cells (Figure 4D) and 2-NBDG (glucose analog) uptake in the NIXsiRNA-BCG infection group was found to be low, compared with the Ctrl SiRN-BCG infection group (Figure 4E). Collectively these results suggest that NIX is required for the upregulated expression glycolysis pathway genes during BCG infection.

Figure 4. Tuberculosis infection induces NIX-dependent mitophagy.

Figure 4

Open in a new tab

(A) NIX knocked down efficacy analysis in PBMCs for gene-specific siRNA, NIX mRNA quantified by qRT-PCR. (B) BCG infection (10 MOI) in siRNA-NIX and Ctrl-siRNA knocked down PBMCs, and fold qRT-PCR quantification of mitophagy pathway genes; LC3A, LC3B, LC3C, and FUNDC1. (C) BCG infection (10 MOI) in siRNA-NIX and Ctrl-siRNA knocked down THP1, Mitotracker Red CMX-Ros uptake analysis by flow cytometry in Texas red channel, at six days post-infection, quantative analysis is represented along with histogram representation. (D) BCG infection (10 MOI) in siRNA-NIX and Ctrl-siRNA knocked down PBMCs, and qRT-PCR quantification (fold change) of glycolysis pathway genes; PFKFB3, ENO1, LDHA, and G6PD. (E) BCG infection (10 MOI) in siRNA-NIX and Ctrl-siRNA knocked down THP1, 2-NBDG uptake analysis by flow cytometry in the FITC channel at 6-day post-infection, quantative analysis is represented along with histogram. (F) ELISA for pro-inflammatory cytokines TNF-α, IL-1β, and IL-6 in the culture supernatants from BCG infected or uninfected PBMCs after 2, 4, and 6 days post-infection in NIX knockdown THP-1 cells. (G) Colony formation unit analysis for BCG; infected siRNA knocked down PBMCs (2, 4, and 6 days post infection), and THP1 (2-days post infection), compared with siRNA control knocked down reference controls.

Next, we asked that if NIX is crucial for mitophagy and glycolysis, does NIX-KD also affect macrophage differentiation in response to BCG infection. We analyzed the expression of M1 cytokines in PBMCs after NIX-KD, followed by BCG infection by ELISA and qRT-PCR, and found that IL-6, IL-1β, and TNF-α were still elevated compared to the uninfected sample. Still, it was significantly less than BCG infected PBMCs (Figure 4F). Additionally, the colony-forming unit (CFU) analysis in PBMCs (2, 4 and 6 days post infection) and THP1 macrophages (2-days post infection) [18] revealed that NIX knockdown negatively affects bacterial clearance (Figure 4G, Figure 4H, Supplementary Figure S4B). These results indicate that NIX-mediated mitophagy is vital for normal macrophage differentiation and clearance of mycobacteria.

Discussion

Immunometabolism or metabolic reprogramming of immune cells has been investigated as a disease signature during mycobacterial infection (31–33). In this study, we have investigated one plausible mechanism of how Mtb and BCG infection induces metabolic reprogramming in immune cells. We performed comparative RNASeq analysis of naïve TB patients and healthy individuals and found that Mtb infection skews cellular metabolism. Next, we saw that BCG infection also leads to similar metabolic reprogramming, suggesting that host cells show similar immune metabolic responses to both BCG and Mtb infection. These results were confirmed by increased uptake of 2-NBDG and 13C6-glucose and a rise in lactate production by BCG infected cells. We also noticed that receptor molecules of the mitophagy pathway (34) were profoundly upregulated, leading to selective elimination of mitochondria indicated by decreased Mitotracker Red staining (23). These results corroborate previous findings that metabolic reprogramming and mitochondrial remodeling are interdependent events (13, 17). We noticed that mitophagy regulator NIX was strongly upregulated upon Mtb and BCG infection, which also regulates glucose and lipid metabolism (13, 16). To determine if NIX plays any crucial physiological role during mycobacterial infection, we performed gene-specific siRNA mediated knockdown. NIX knocked down not only abolish the upregulation of glycolysis and mitophagy but also negatively affects the production of pro-inflammatory cytokines TNF-α, IL-6, IL-1β (13). These cytokines are secreted by M1-macrophages and are necessary for mounting effective immune reaction for bacterial clearance. We noticed that ratios CD86/CD206 and NOS2/ARG1 that increase upon BCG infection and are characteristic of M1 macrophage phenotype (35) also decrease upon NIX knockdown. This indicates that NIX is necessary for the maturation of macrophages, and knockdown of NIX improves mitochondrial mass (23), associated with a reduction in pro-inflammatory cytokines and decreased expression of glycolytic enzymes and mitophagy marker genes, negatively affecting bacterial clearance as shown by CFU analysis (36) in PBMCs and THP1 macrophages. THP1 cells are cancer cells and are inherently activate high levels of glycolysis (a property of cancerous cells) therefore we have performed BCG infection experiments in human PBMCs, which are primary cells. Additionally, we have inferred from our results after comparing infected ‘THP1’ cells with uninfected ‘THP1’ control cells. Therefore, the differences in glycolysis are mainly due to BCG infection.

Noticeably, both BCG and Mtb induce mitophagy-associated metabolic reprogramming and skew macrophage polarization to the M1 phenotype. However, in the case of TB patients, it has been shown that M2 macrophages dominate the TB granuloma microenvironment. This could be because of RD-1 encoded gene ESAT-6, which is secreted by Mtb and is proposed to make a conducive microenvironment for bacteria in and around granuloma. We suggest that both BCG and Mtb initially skew the maturation of macrophages to the M1 phenotype, and contradictory to previous reports, the role of ESAT-6 is dispensable in this process (37). When these pro-inflammatory macrophages reach the infection or granuloma site, the local tissue microenvironment affects M1 macrophages, differentiating them back to the anti-inflammatory M2 phenotype that supports the survival of Mtb and development of the disease. Contrary to our findings Hackett et al. found Mtb infection decrease IL-1β by restriction of PFK-M via microRNA-21 in hMDMs. This couild be because the production of IL-1β is regulated at multiple levels, and mechanisms regulating it could be different in in-vitro differentiated hMDMs and PBMCs.

While several reports suggest that increased glycolysis is an essential process driving host immune response against Mtb infection (8, 38–40), there are also reports that suggests vice-versa phenomenon (12). Whether elevated glycolytic process is a characteristic of Mtb infection is still debatable. For example, Lachmandas et al. show that Mtb infection shift TB patients’ PBMC metabolism towards glycolysis while Cumming et al. have shown that Mtb (h37Rv) infection results in decelerated glycolysis (12). All the mentioned studies have used different cell types and methodologies for measuring glycolytic activity which could be a reason for contrasting results. Our results support findings that Mtb infection induces glycolytic metabolism in patient’s PBMCs supported by RNASeq analysis, mRNA expression of glycolytic pathways enzymes, and mitochondrial mass measurement.

In summary, we have identified that there are similarities in immunometabolism during Mtb infection and ex-vivo BCG infection in PBMCs and THP-1 and shown that NIX-mediated mitophagy is a key player behind this metabolic shift during macrophage differentiation to the M1 phenotype. Previous studies have reported that NIX also plays a crucial in CD8+ T cell memory development, mediates the mitochondrial damage in inflammatory bowel disease (16, 41), and promotes tumor survival in the hypoxic niche (42). Our findings corroborate these reports, further potentiating the claims of NIX as a central player in the innate immune system as it regulates immune response through immuno-metabolism. Our study has some limitations as we have performed RNASeq with whole PBMCs due to the limited amount of blood that could be collected from therapy naïve TB patients. Results could be different from the CD14+ sorted macrophages population. We did not use Mtb pathogenic strains for in-vitro experiments as we do not have an appropriate facility to perform live Mtb infection experiments.

Conclusion

Mycobacteria Mtb or BCG infection strongly induces NIX/BNIP3L, which in turn triggers metabolic reprogramming and differentiation of macrophages from naïve M0 to M1 pro-inflammatory phenotype. NIX triggered mitophagy is central to mount a protective immune response during tuberculosis. As macrophages are the sentinels of innate immunity, first responder, and primary cellular niche for M. tuberculosis, selective regulation of NIX triggered mitophagy can be besieged to development and understanding anti-tuberculosis immune mechanism, further studies are required.

Supplementary Material

Supplementary Materials

Acknowledgments

We thank Dr. Manoj Verma from District Tuberculosis Hospital for providing clinically diagnosed treatment naïve TB patients blood samples, healthy volunteers for donating blood samples, Dr. Ajit Chande from IISER Bhopal for gifting THP1 cells. We also thank Dr. R Williams, Dr. S Roy, and C Williams from UCL for their support in protocol design and quality control measures for RNASeq run.

Funding

This study was supported in parts by the Department of Biotechnology (grant number BT/PR25052/NER/95/983/2017) and intramural IISER Bhopal grant to Himanshu Kumar), Newton Bhabha Ph.D. Placement Program (grant number BT/IN/UK/DBT-BC/2015-16 to Ranjeet Singh Mahla and Judith Breuer) and IISER Bhopal start-up funds (grant number INST/CHM/2016003 to Bharathwaj Sathyamoorthy). For fellowship, R.S.M was supported by IISERB-Ministry of Human Resources Development, A.K. was supported by the Council of Scientific and Industrial Research, India, and DST INSPIRE Faculty Fellowship supported S.T.K.

Declarations

Ethics approval and consent to participate

The IEC, IISER-Bhopal, India approved the study and necessary permissions for blood sample collection provided by the Directorate of Health, MP-Govt.

Consent for publication

This study does not use the patient’s details, photographs, and video content; therefore, consent for publication is not required from study participants.

Competing interests

All authors declare no conflict of interest.

Availability of data and materials

Raw and processed RNA-Seq data files have been submitted to NCBI-GEO. Data files can be accessed with accession number GSE148171.

References

  • 1.Global tuberculosis report 2019. Geneva: World Health Organization; 2019. Licence: CC BY-NC-SA 3.0 IGO. [Google Scholar]
  • 2.Schluger NW, Rom WN. The host immune response to tuberculosis. American journal of respiratory and critical care medicine. 1998;157(3 Pt 1):679–91. doi: 10.1164/ajrccm.157.3.9708002. [DOI] [PubMed] [Google Scholar]
  • 3.Flynn JL, Chan J, Lin PL. Macrophages and control of granulomatous inflammation in tuberculosis. Mucosal Immunol. 2011;4(3):271–8. doi: 10.1038/mi.2011.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Lugo-Villarino G, Hudrisier D, Benard A, Neyrolles O. Emerging trends in the formation and function of tuberculosis granulomas. Front Immunol. 2012;3:405. doi: 10.3389/fimmu.2012.00405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ley K. M1 Means Kill; M2 Means Heal. The Journal of Immunology. 2017;199(7):2191–3. doi: 10.4049/jimmunol.1701135. [DOI] [PubMed] [Google Scholar]
  • 6.Marino S, Cilfone NA, Mattila JT, Linderman JJ, Flynn JL, Kirschner DE. Macrophage polarization drives granuloma outcome during Mycobacterium tuberculosis infection. Infection and immunity. 2015;83(1):324–38. doi: 10.1128/IAI.02494-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Viola A, Munari F, Sanchez-Rodriguez R, Scolaro T, Castegna A. The Metabolic Signature of Macrophage Responses. Front Immunol. 2019;10:1462. doi: 10.3389/fimmu.2019.01462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Huang Z, Luo Q, Guo Y, Chen J, Xiong G, Peng Y, et al. Mycobacterium tuberculosis-Induced Polarization of Human Macrophage Orchestrates the Formation and Development of Tuberculous Granulomas In Vitro. PLoS One. 2015;10(6):e0129744. doi: 10.1371/journal.pone.0129744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Redford PS, Murray PJ, O’Garra A. The role of IL-10 in immune regulation during M. tuberculosis infection. Mucosal immunology. 2011;4(3):261–70. doi: 10.1038/mi.2011.7. [DOI] [PubMed] [Google Scholar]
  • 10.Refai A, Gritli S, Barbouche MR, Essafi M. Mycobacterium tuberculosis Virulent Factor ESAT-6 Drives Macrophage Differentiation Toward the Pro-inflammatory M1 Phenotype and Subsequently Switches It to the Anti-inflammatory M2 Phenotype. Frontiers in cellular and infection microbiology. 2018;8:327. doi: 10.3389/fcimb.2018.00327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.O’Garra A, Redford PS, McNab FW, Bloom CI, Wilkinson RJ, Berry MP. The immune response in tuberculosis. Annual review of immunology. 2013;31:475–527. doi: 10.1146/annurev-immunol-032712-095939. [DOI] [PubMed] [Google Scholar]
  • 12.Cumming BM, Addicott KW, Adamson JH, Steyn AJ. Mycobacterium tuberculosisinduces decelerated bioenergetic metabolism in human macrophages. Elife. 2018;7 doi: 10.7554/eLife.39169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Esteban-Martínez L, Sierra-Filardi E, McGreal RS, Salazar-Roa M, Mariño G, Seco E, et al. Programmed mitophagy is essential for the glycolytic switch during cell differentiation. EMBO J. 2017;36(12):1688–706. doi: 10.15252/embj.201695916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Daniel J, Maamar H, Deb C, Sirakova TD, Kolattukudy PE. Mycobacterium tuberculosis uses host triacylglycerol to accumulate lipid droplets and acquires a dormancy-like phenotype in lipid-loaded macrophages. PLoS Pathog. 2011;7(6):e1002093. doi: 10.1371/journal.ppat.1002093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Naik PP, Birbrair A, Bhutia SK. Mitophagy-driven metabolic switch reprograms stem cell fate. Cell Mol Life Sci. 2019;76(1):27–43. doi: 10.1007/s00018-018-2922-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Gupta SS, Sharp R, Hofferek C, Kuai L, Dorn GW, Wang J, et al. NIX-Mediated Mitophagy Promotes Effector Memory Formation in Antigen-Specific CD8. Cell Rep. 2019;29(7):1862–77.:e7. doi: 10.1016/j.celrep.2019.10.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Esteban-Martínez L, Boya P. BNIP3L/NIX-dependent mitophagy regulates cell differentiation via metabolic reprogramming. Autophagy. 2018;14(5):915–7. doi: 10.1080/15548627.2017.1332567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Association WM. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191–4. doi: 10.1001/jama.2013.281053. [DOI] [PubMed] [Google Scholar]
  • 19.Leinardi R, Pavan C, Yedavally H, Tomatis M, Salvati A, Turci F. Cytotoxicity of fractured quartz on THP-1 human macrophages: role of the membranolytic activity of quartz and phagolysosome destabilization. Arch Toxicol. 2020;94(9):2981–95. doi: 10.1007/s00204-020-02819-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Monahan IM, Betts J, Banerjee DK, Butcher PD. Differential expression of mycobacterial proteins following phagocytosis by macrophages. Microbiology (Reading) 2001;147(Pt 2):459–71. doi: 10.1099/00221287-147-2-459. [DOI] [PubMed] [Google Scholar]
  • 21.Kumari P, Saha I, Narayanan A, Narayanan S, Takaoka A, Kumar NS, et al. Essential role of HCMV deubiquitinase in promoting oncogenesis by targeting anti-viral innate immune signaling pathways. Cell Death Dis. 2017;8(10):e3078. doi: 10.1038/cddis.2017.461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Asalla S, Mohareer K, Banerjee S. Small Molecule Mediated Restoration of Mitochondrial Function Augments Anti-Mycobacterial Activity of Human Macrophages Subjected to Cholesterol Induced Asymptomatic Dyslipidemia. Front Cell Infect Microbiol. 2017;7:439. doi: 10.3389/fcimb.2017.00439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Monteiro LB, Davanzo GG, de Aguiar CF, Moraes-Vieira PMM. Using flow cytometry for mitochondrial assays. MethodsX. 2020;7:100938. doi: 10.1016/j.mex.2020.100938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.de Brito Monteiro L, Davanzo GG, de Aguiar CF, Corrêa da Silva F, Andrade JR, Campos Codo A, et al. M-CSF- and L929-derived macrophages present distinct metabolic profiles with similar inflammatory outcomes. Immunobiology. 2020;225(3):151935. doi: 10.1016/j.imbio.2020.151935. [DOI] [PubMed] [Google Scholar]
  • 25.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20. doi: 10.1093/bioinformatics/btu170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nature methods. 2015;12(4):357–60. doi: 10.1038/nmeth.3317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature protocols. 2012;7(3):562–78. doi: 10.1038/nprot.2012.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Liao Y, Wang J, Jaehnig EJ, Shi Z, Zhang B. WebGestalt 2019: gene set analysis toolkit with revamped UIs and APIs. Nucleic Acids Res. 2019;47(W1):W199–W205. doi: 10.1093/nar/gkz401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90–7. doi: 10.1093/nar/gkw377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sampol D, Ostrofet E, Jobin ML, Raffard G, Sanchez S, Bouchaud V, et al. Glucose and lactate metabolism in the awake and stimulated rat: a (13)C-NMR study. Front Neuroenergetics. 2013;5:5. doi: 10.3389/fnene.2013.00005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Shi L, Salamon H, Eugenin EA, Pine R, Cooper A, Gennaro ML. Infection with Mycobacterium tuberculosis induces the Warburg effect in mouse lungs. Sci Rep. 2015;5:18176. doi: 10.1038/srep18176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Russell SL, Lamprecht DA, Mandizvo T, Jones TT, Naidoo V, Addicott KW, et al. Compromised Metabolic Reprogramming Is an Early Indicator of CD8. Cell Rep. 2019;29(11):3564–79.:e5. doi: 10.1016/j.celrep.2019.11.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Howard NC, Khader SA. Immunometabolism during Mycobacterium tuberculosis Infection. Trends Microbiol. 2020 doi: 10.1016/j.tim.2020.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Gkikas I, Palikaras K, Tavernarakis N. The Role of Mitophagy in Innate Immunity. Front Immunol. 2018;9:1283. doi: 10.3389/fimmu.2018.01283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Bertani FR, Mozetic P, Fioramonti M, Iuliani M, Ribelli G, Pantano F, et al. Classification of M1/M2-polarized human macrophages by label-free hyperspectral reflectance confocal microscopy and multivariate analysis. Sci Rep. 2017;7(1):8965. doi: 10.1038/s41598-017-08121-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Geissmann Q. OpenCFU, a new free and open-source software to count cell colonies and other circular objects. PLoS One. 2013;8(2):e54072. doi: 10.1371/journal.pone.0054072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Refai A, Gritli S, Barbouche MR, Essafi M. Virulent Factor ESAT-6 Drives Macrophage Differentiation Toward the Pro-inflammatory M1 Phenotype and Subsequently Switches It to the Anti-inflammatory M2 Phenotype. Front Cell Infect Microbiol. 2018;8:327. doi: 10.3389/fcimb.2018.00327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lachmandas E, Boutens L, Ratter JM, Hijmans A, Hooiveld GJ, Joosten LA, et al. Microbial stimulation of different Toll-like receptor signalling pathways induces diverse metabolic programmes in human monocytes. Nat Microbiol. 2016;2:16246. doi: 10.1038/nmicrobiol.2016.246. [DOI] [PubMed] [Google Scholar]
  • 39.Lachmandas E, Beigier-Bompadre M, Cheng SC, Kumar V, van Laarhoven A, Wang X, et al. Rewiring cellular metabolism via the AKT/mTOR pathway contributes to host defence against Mycobacterium tuberculosis in human and murine cells. Eur J Immunol. 2016;46(11):2574–86. doi: 10.1002/eji.201546259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Braverman J, Stanley SA. Nitric Oxide Modulates Macrophage Responses to. J Immunol. 2017;199(5):1805–16. doi: 10.4049/jimmunol.1700515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Vincent G, Novak EA, Siow VS, Cunningham KE, Griffith BD, Comerford TE, et al. Nix-Mediated Mitophagy Modulates Mitochondrial Damage During Intestinal Inflammation. Antioxid Redox Signal. 2020;33(1):1–19. doi: 10.1089/ars.2018.7702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Jung J, Zhang Y, Celiku O, Zhang W, Song H, Williams BJ, et al. Mitochondrial NIX Promotes Tumor Survival in the Hypoxic Niche of Glioblastoma. Cancer Res. 2019;79(20):5218–32. doi: 10.1158/0008-5472.CAN-19-0198. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Materials
EMS217698-supplement-Supplementary_Materials.pdf (1.1MB, pdf)

Data Availability Statement

Raw and processed RNA-Seq data files have been submitted to NCBI-GEO. Data files can be accessed with accession number GSE148171.

RESOURCES