of February 10, 2018.This information is current as Ligand-Receptor Pair

A Case of a Heterologous and NoncanonicalModulate Foamy Biogenesis in Granulomas:

TR4Acid and Macrophage Nuclear Receptor Keto-MycolicMycobacterium tuberculosis

Thakur, Shanmugam Mayilraj and Pawan GuptaSomavarapu, Ashish Arora, Raman Parkesh, Krishan GopalMahajan, Sandeep Dave, Ankita Saini, Arun Kumar Hedwin Kitdorlang Dkhar, Ravikanth Nanduri, Sahil

http://www.jimmunol.org/content/193/1/295doi: 10.4049/jimmunol.14000922014;

2014; 193:295-305; Prepublished online 6 JuneJ Immunol 

MaterialSupplementary

2.DCSupplementalhttp://www.jimmunol.org/content/suppl/2014/06/06/jimmunol.140009

        average*  

4 weeks from acceptance to publicationSpeedy Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

?The JIWhy

Referenceshttp://www.jimmunol.org/content/193/1/295.full#ref-list-1

, 23 of which you can access for free at: cites 75 articlesThis article

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2014 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on February 10, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Mycobacterium tuberculosis Keto-Mycolic Acid andMacrophage Nuclear Receptor TR4 Modulate FoamyBiogenesis in Granulomas: A Case of a Heterologousand Noncanonical Ligand-Receptor Pair

Hedwin Kitdorlang Dkhar,* Ravikanth Nanduri,* Sahil Mahajan,* Sandeep Dave,*

Ankita Saini,* Arun Kumar Somavarapu,* Ashish Arora,† Raman Parkesh,*

Krishan Gopal Thakur,* Shanmugam Mayilraj,* and Pawan Gupta*

The cell wall of Mycobacterium tuberculosis is configured of bioactive lipid classes that are essential for virulence and potentially

involved in the formation of foamy macrophages (FMs) and granulomas. Our recent work established crosstalk between

M. tuberculosis cell wall lipids and the host lipid-sensing nuclear receptor TR4. In this study, we have characterized, identified, and

adopted a heterologous ligand keto-mycolic acid from among M. tuberculosis lipid repertoire for the host orphan NR TR4.

Crosstalk between cell wall lipids and TR4 was analyzed by transactivation and promoter reporter assays. Mycolic acid (MA)

was found to transactivate TR4 significantly compared with other cell wall lipids. Among the MA, the oxygenated form, keto-MA,

was responsible for transactivation, and the identity was validated by TR4 binding assays followed by TLC and nuclear magnetic

resonance. Isothermal titration calorimetry revealed that keto-MA binding to TR4 is energetically favorable. This keto-MA–TR4

axis seems to be essential to this oxygenated MA induction of FMs and granuloma formation as evaluated by in vitro and in vivo

model of granuloma formation. TR4 binding with keto-MA features a unique association of host nuclear receptor with a bacterial

lipid and adds to the presently known ligand repertoire beyond dietary lipids. Pharmacologic modulation of this heterologous axis

may hold promise as an adjunct therapy to frontline tuberculosis drugs. The Journal of Immunology, 2014, 193: 295–305.

The global death toll resulting from tuberculosis (TB)remains enormous. The threat of the disease has worsenedbecause of increased reported cases of multidrug-resistant

TB and the coinfection with HIV. In most developed countries, thedecline in TB mortality in the last century can be attributed to thebetter nutritional habits, improved living conditions, and an im-proved health system providing treatment for the infected indi-viduals (1). In addition, along with the World Health Organizationdirectly observed treatment, short course (DOTS) program, thedevelopment of better diagnostics and vaccines by researchers in

the late 1990s has also contributed to the decline of TB. A freshemphasis on unraveling the molecular events that occur duringhost-pathogen interaction could be the basis of new technologiesfor detecting and treating TB (2). The complete genome sequenceof Mycobacterium tuberculosis reveals that 30% of the genes codefor enzymes involved in lipid synthesis or metabolism, much ofwhich occurs within host macrophages (3, 4). In addition toabundance, other features make mycobacterial lipids paradigmaticmediators in supporting M. tuberculosis survival within macro-phages either by subverting host microbicidal functions or byaugmenting pathogenesis.Earlier reports on TB have clearly described the trafficking of

mycobacterial lipids inside the infected cells. It is noteworthy thatmycobacteria cell wall lipids disperse or shed upon association withthe host cell or an appropriate environment within the macrophage(5). M. tuberculosis lipids have been profiled (3), and several ofthese mycobacterial lipids found across species have been re-ported to be involved in the immunosuppression of PBMC pro-liferation (6), modulation of cytokines (7, 8), and exacerbation ofthe granulomatous response (9, 10). It has been reported thatM. tuberculosis lipids contribute to the virulence of the bacteria byinactivating host bactericidal enzymes and that tissue-specific lo-calization of the lipids leads to increased pathogenicity (11, 12).An in vivo study was performed in mice to examine the effect ofmycolic acid (MA), the main component of the envelope, onvirulence (13). In addition, a subunit vaccine formulation withM. tuberculosis lipids was proposed from a study in guinea pig (14).M. tuberculosis survives within foamy macrophages during la-

tency and starvation by utilizing lipids as their primary carbon source(15). Current evidence strongly suggests a functional role for MA,the most abundant lipid component of the M. tuberculosis cell wall,

*Council of Scientific and Industrial Research–Institute of Microbial Technology,Chandigarh 160036, India; and †Council of Scientific and Industrial Research–CentralDrug Research Institute, Lucknow 226031, India

Received for publication January 13, 2014. Accepted for publication May 5, 2014.

This work was supported by the Department of Biotechnology-India project BT/01/IYBA/2009 and the Council of Scientific and Industrial Research 12th Plan Networkproject Bugs to Drugs, Biodiscovery (BSC0211, BSC0120) and Open Source DrugDiscovery to P.G.

Address correspondence and reprint requests to: Dr. Pawan Gupta, Institute ofMicrobial Technology, Sector 39-A, Chandigarh 160036, India. E-mail address:pawan@imtech.res.in

The online version of this article contains supplemental material.

Abbreviations used in this article: AG, arabinogalactan; AIL, acetone-insoluble lipid;ASL, acetone-soluble lipid; 1D, one-dimensional; 2D, two-dimensional; DOT, di-rectly observed treatment, short course; FM, foamy macrophage; hMDM, humanmonocyte–derived macrophage; IMTECH, Institute of Microbial Technology; ITC,isothermal titration calorimetry; LAM, lipoarabinomannan; LBD, ligand bindingdomain; LM, lipomannan; MA, mycolic acid; m-AGp, mycolylarabinogalactan–peptidoglycan covalent complex; NMR, nuclear magnetic resonance; PIM, phospha-tidylinositol mannoside 1,2; PIM 6, phosphatidylinositol mannosides 6; PG,peptidoglycan; shRNA, short hairpin RNA; TB, tuberculosis; TDM, trehalosedimycolate.

Copyright� 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1400092

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

as an inducer of foamy macrophages during granuloma formation(16). InM. tuberculosis,MAs are classified as a-MA, methoxy-MA,and keto-MA (17). These different classes of MAs are not only usedas signatures in the identification of different mycobacterial speciesand strains but also seem to define the virulence property of thestrain (18). Although it has been known for some time that MA andthe arabinogalactan layer of the M. tuberculosis cell wall are cova-lently associated (19–21), recently the occurrence of free MA inbiofilm matrix formation (22, 23) has been reported. Pathogenicstrains such as M. tuberculosis (24) and M. avium (25) produce theoxygenated MA class, which is absent in the nonpathogenic strains,such as M. smegmatis (26). Therefore, it is tempting to hypothesizeabout and investigate the role of the oxygenated MA class in thesurvival and pathogenesis of M. tuberculosis.Within the host macrophage, in addition to membrane receptors

involved in the phagocytosis of the pathogen, TLRs and somenuclear receptors are considered to be important lipid sensorsas well as integrators of lipid signaling, and they have been shownto be integral elements of foamy biogenesis (27–31). Testicularreceptor 4 (TR4), peroxisome proliferator-activated receptor (PPAR),and liver X receptor (LXR) function as fatty acid sensors and cansense an array of complex lipids (28, 31–33). TR4 is widely ac-cepted as having a role in macrophage foamy biogenesis (28, 31),and it has been suggested that PPARg is involved in modulating lipidbiogenesis (34). On the other hand, LXRa has the opposite role ofinhibiting foam cell formation (35). We recently reported thecrosstalk ofM. tuberculosis lipids with TR4, which leads to bacterialsurvival, although the identity of the heterologous ligand was un-known.To determine the exact identity of this heterologous ligand for

the lipid-sensing nuclear receptor (LSNR) TR4, we first screenedfor crosstalk between TR4 and various M. tuberculosis cell wallcomponents. Transactivation and promoter reporter assays revealeda ligand-like behavior of MA with TR4. MA extraction and isola-tion were performed to distinguish which of the distinct classes ofMA was involved in the transactivation. Transactivation and pro-moter reporter assays confirmed the response of TR4 to keto-MA. ATR4 binding assay followed by one-dimensional (1D) and two-dimensional (2D) TLC corroborated the identity of keto-MA. Iso-thermal titration calorimetry (ITC), circular dichroism and nuclearmagnetic resonance (NMR) analysis further confirmed a stable andhigh-affinity binding between keto-MA and TR4. We have shownthat the heterologous ligand-receptor pair, keto-MA–TR4 crosstalkmodulates foamy biogenesis and causes granuloma induction. Wehave reported keto-MA as the heterologous ligand for the orphannuclear receptor, TR4.

Materials and MethodsBacterial strain and culture conditions

Mycobacteria strains used were M. tuberculosis (H37Ra and H37Rv),M. smegmatis, and M. phlei and were obtained from Institute of MedicalTechnology (IMTECH)-Microbial Type Culture Collection Chandigarh orNational Institute for Research in Tuberculosis, Chennai, India. They werecultured in 7H9 broth medium containing 5% glycerol and 0.05% Tween 80.OADC (10%) was added as a supplement, and the culture was incubated at 37˚Con a shaker. Log phase cultures were spun down at 6000 rpm at room tem-perature. Cell pellets were stored at 280˚C and lyophilized into powder form.

M. tuberculosis lipids source, isolation, and extraction

Acetone-soluble lipids (ASLs), acetone-insoluble lipids (AILs), trehalosedimycolate (TDM), sulfolipids-1, phosphatidylinositol mannoside 1,2 (PIM),phosphatidylinositol mannosides 6 (PIM 6), lipoarabinomannan (LAM),lipomannan (LM), arabinogalactan (AG), peptidoglycan (PG), and mycoly-larabinogalactan–peptidoglycan covalent complex (m-AGP) were obtainedfrom BEI Resources, the National Institute of Allergy and Infectious Dis-eases, and the National Institutes of Health.

MAwas freshly extracted by two different methods, the first of which wasacid methanolysis (24). Dried cell mass (150 mg) was added to methanol(2.5 ml), toluene (2.5 ml), and concentrated sulfuric acid (0.1 ml) in a10-ml polytetrafluoroethylene screw-capped tube. The contents were mixedthoroughly and kept at 75˚C for 12–16 h. After methanolysis, the reactionmixture was cooled to room temperature. MA was extracted by addinghexane (2 ml) to the mixture with vigorous shaking and centrifuging toseparate the solvent. The upper phase containing the MAwas transferred toa fresh tube, and the volume was reduced to 100 ml on a rotatory evapo-rator. Samples were stored at 4˚C until further use.

The second method for MA extraction was based on saponification (18)followed by methyl ester derivatization with trimethylsilyldiazomethane.Dried cell mass was dissolved in 15% aqueous tetrabutylammonium hy-droxide and heated at 70˚C overnight. The acidified hydrolysate wasextracted with diethyl ether and separated by centrifugation. The washedresidue was further hydrolyzed by the addition of tetrabutylammoniumhydroxide until no spot of MAwas observed on a TLC plate. The extractedMA was run through a silica gel column that was eluted first with hexanefollowed by a series of diethyl ether:hexane mixtures (6:100, 1:9, and 3:7,v/v). TLC on each mixture was run to check for MA. The crude MAfraction obtained was added to trimethylsilyldiazomethane and incubatedfor 48–72 h on a rotary shaker. The mixture was loaded onto a silica gelcolumn and eluted first with hexane followed by diethyl ether:hexane (3:7,v/v) to obtain the methyl mycolate fraction.

Thin layer chromatography

Extracted crude MA was loaded onto a thin layer plate (Merck), and thechromatogram was analyzed with 1D TLC with the solvent system pe-troleum ether:diethyl ether (85:15, v/v) (24). MA spots were visualized onthe plate after spraying with phosphomolybdic acid (Aldrich) and charredat 100˚C for 10–15 min. For 2D TLC, a triple development with petroleumether/acetone (95:5, v/v) in one direction was followed by a single de-velopment at 90˚C to the first direction with toluene/acetone (97:3, v/v), asdescribed previously (24, 36). The different spots of MA were designateda-MA, methoxy-MA, and keto-MA based on the functional groups, asreported previously (37). For purification of each class of MA, the crudeMA was run on a preparative TLC plate, and spots were scraped off theplates and extracted with diethyl ether.

Cell culture, stimulation, and infection

PBMCs were isolated from fresh-drawn blood by Ficoll-Hypaque gradientcentrifugation (Sigma-Aldrich). PBMCs were cultured and obtained asdescribed previously (28). Cells were incubated for 3 h at 37˚C and thenwashed with PBS to remove nonadherent cells. Human monocyte–derivedmacrophages (hMDMs) were obtained by culturing the cells with M-CSF 7(50 ng/ml) for 7 d.

ATHP-1 stable knockdown cell line (shTR4) was generated and culturedas described in our previous study (28). For the foamy macrophage study orin vitro M. tuberculosis survival assay, hMDMs, THP-1 macrophage cells,or both were stimulated with different classes of MA along with M. tu-berculosis infection at a multiplicity of infection 1:5. Lipid droplets wereenumerated by Nile Red or Oil Red O staining. Mycobacterial viabilitydetermination was done by CFU assay and flow cytometry. Briefly,hMDMs were infected with M. tuberculosis H37Rv or M. smegmatis(multiplicity of infection 1:5) along with treatment of the cells with MA orsolvent control. After infection, cells were allowed to phagocytose for 1–4 h,and any nonphagocytosed bacteria were cleared by washing with PBS orincomplete medium and incubating with repletion medium for 24–48 h.Macrophages were then solubilized with 0.06% SDS, and bacterial sus-pensions were used with the Live/Dead BacLight Bacterial Viability andCounting kit per the manufacturer’s instructions (Invitrogen). The per-centages of live and dead bacteria were determined by flow cytometry (BDFACSCalibur) after staining with SYTO 9 and propidium iodide. For CFUdeterminations, the bacterial suspensions were serially diluted after mac-rophage solubilization, 50 ml of each sample were plated, and CFUs werecounted. Final calculations included the dilution factor and the volume ofdiluted sample used for plating.

Luciferase reporter assay

The luciferase reporter assay was performed as described (28). COS-1 cellswere plated in 24-well plates, and at 70% confluency the media waschanged to Opti-MEM, and the cells were transfected with pBIND humanTR4, pG5 for the transactivation assay and with pFLAG human TR4,pGL3-CD36 for the promoter reporter assay. In another experiment, pGL3-CD36 and pBIND plasmids were transfected to control and TR4 andPPARg stably knocked down THP-1 cell lines. Firefly luciferase was usedas reporter, and Renilla luciferase was used to check transfection efficiency

296 Keto-MA–TR4 AXIS IN M. TUBERCULOSIS–INDUCED FOAMY Mf

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

and for normalization. Normalized luciferase activities (relative light units)were plotted as the average (6SD) of triplicate samples from typicalexperiments (SigmaPlot).

Lipid body enumeration by microscopy

hMDMs or THP-1 macrophages were cultured on a coverslip and thenstimulated with MA or infected with M. tuberculosis as detailed above, orboth. The hMDMs were fixed with 4% paraformaldehyde, and intracellularlipids were stained with Nile Red (3.3 mg/ml) for 15 min. DAPI was usedto stain the nuclei. Images were acquired by Nikon AIR confocal micro-scope using a 603 objective. The THP-1 macrophages were fixed andstained with 0.5% Oil Red O for 30 min at room temperature, and coun-terstained with hematoxylin for 5 min to stain nuclei. Cells were rinsedwith PBS, mounted on glass slides, and imaged. Lipid bodies were enu-merated from 10–15 fields containing 20 consecutive macrophages on eachslide by light microscopy using a 403 objective lens.

Determination of intracellular lipid bodies by flow cytometry

To check for intracellular lipid content after MA stimulation or infectionwith M. tuberculosis as described above, cells were stained with Nile Red,a lipophilic dye that labels fat accumulation within the cytosol. After 48 hof stimulation or infection, adhered cells were washed twice with PBS andtrypsinized for detachment from the plate. Following centrifugation at1500 rpm for 5 min, the cell pellets were resuspended in 1 ml PBS with1 mg/ml Nile Red and incubated for 15–20 min at room temperature. NileRed fluorescence was measured by flow cytometry on a Becton DickinsonFACSCalibur system vide FL2 emission channel. Data were acquired andanalyzed with CellQuest Pro software from BD Biosciences.

Virus production

Replication-deficient adenovirus was produced by transfection into the293A cell line of either adenoviral construct pAd/BLOCK-iT-DEST car-rying short hairpin RNA (shRNA) for LacZ or shRNA specific for murine orhuman TR4, according to the manufacturer’s instructions (Invitrogen).Supernatants were collected when 80% cytopathic effect was seen andprocessed according to the manufacturer’s protocol (Invitrogen). Viralparticles were stored at 280˚C and thawed at the time of transduction. Thesilencing efficiency of these constructs was monitored by RT-PCR analy-sis. Mouse cell line RAW or human cell line THP-1 was used to check forknockdown efficiency.

In vitro granuloma formation

One hundred micrograms TDM (Sigma), total MA, or keto-MAwere mixedwith Polybead Polystyrene Microspheres inert beads suitable for micros-copy applications and with capability of passive adsorption from Poly-sciences. The mixture was rotated overnight at room temperature to allowuniform and homogeneous coating of the beads. The efficiency of lipidcoating on the beads was evaluated. Approximately 100 coated beads wereincubated with 106 PBMCs in a 12-well plate in a total volume of 1 ml forup to 8 d at 37˚C in a 5% CO2 atmosphere. Alternatively, 24 h beforeincubation with beads, PBMCs were transduced with adenoviral super-natant having control shRNA or with shRNATR4. The cellular aggregateswere observed on an Olympus IX51 (Olympus) using cellSens imagingsoftware.

Animal study and ethics

C57BL/6 mice (male, 6–8 wk old) were procured from the animal facilityat the IMTECH and were housed at the Biosafety Level 3 facility of theInstitute. The animals were maintained with proper food and water andmonitored regularly. All experiments were approved (IAEC/11/18: Study ofNuclear Receptors in Tuberculosis Granuloma) by the Institutional AnimalEthics Committee of the Institute of Microbial Technology and performedaccording to the National Regulatory Guideline issued by the Committee forthe Purpose of Supervision of Experiments on Animals (No. 55/1999/CPCSEA), Ministry of Environment and Forest, Government of India.

Aerosol infection and determination of Mycobacterium burden

Experimental mice were injected in the tail vein and aerosol challenged(with a nebulization system Glas-Col; Terre Haute, IN) with control ade-novirus (shRNA LacZ) or carrying the shRNATR4 in the BSL3 (IMTECH).The lungs were dissected, and RT-PCR was performed to check the ex-pression of TR4 in alveolar macrophages at different time points. Afteradenovirus treatment, the mice were exposed to aerosol inhalation ofM. tuberculosis H37Rv (∼100 CFU per lung, as per the standardized doseobserved after day 1 of infection) using the inhalation exposure system.

The bug burden in the experimental animals was determined at differenttime point after aerosol challenge (15, 30, and 60 d). Animals were kept inseparate chambers, and the exposure setting was set for 45 min (nebuliza-tion setup) with a standardized CFU count. Mycobacterium burden per lungwas determined by serial diluting the total lung homogenates and platingthem on Middlebrook 7H11 plates supplemented with OADC enrichment.Plates were monitored weekly, CFU count was observed within 3–4 wk ofincubation at 37˚C, and the total bug load of the lung tissue was recorded.

Histopathology and body weight analysis

Control mice and TR4-knockdown mice infected with M. tuberculosisH37Rv or M. smegmatis were sacrificed after 60 d postinfection, and lungtissue was dissected and fixed in buffered 10% formalin. Histologic sec-tions of the lungs were stained with H&E to check for in vivo granulomaformation. Images were captured on Olympus CH 20L and ocular mi-crometer disc was used for counting the area. In addition, the body weightsof other groups of control and TR4-knockdown mice were monitoredweekly for 20 wk after aerosol exposure.

Isothermal titration calorimetry

Isothermal titration calorimetry experiments were performed with a MicrocalVP-ITC microcalorimeter. Full-length human TR4 protein (hTR4) was pu-rified and extensively dialyzed against the buffer used in ITC, 20mMTrisHCl(pH 7.4), 250 mM NaCl, and 10% glycerol. The same buffer was used fortitration. The stock of ligand was prepared in diethyl ether, and the ligandconcentrations used were in 15–20 M excess of TR4. The ligand was dis-solved in titration buffer, and an equivalent amount of buffer was added tothe protein. Titrations consisted of 45 injections of 6 ml separated by 120 s.The titrations were performed with 3–6-mM concentrations of protein and45–60-mM concentrations of ligand. Data was analyzed with Origin 7software to determine the binding constant (Ka), apparent stoichiometry (N),and change in enthalpy (DH) and entropy (DS).

Structural modeling and docking studies

The human TR4 ligand binding domain (LBD) active-form homologymodel was generated with a Swiss model server by using human RXRaLBD (Brookhaven Protein Data Bank ID: 1FM6) as a template (38). The3D structural coordinates of g-linoleic acid and keto-MA were generatedby a Pro-DRG2 server. Docking was performed with AutoDock 4 (39).Figures were created using Pymol.

Circular dichroism

Full-length hTR4 was purified and extensively dialyzed against buffer20 mM Tris HCl (pH 7.4), 250 mMNaCl, and 10% glycerol. Measurementswere recorded with a JASCO circular dichroism instrument. For temper-ature melting experiments, a gradient of 20–90˚C was used with incrementsof 1˚C. The change in circular dichroism signal was measured at 223 nM.Protein concentration was 0.2 mg/ml. Experiments were performed withhTR4 in the presence and absence of keto-MA.

Ethics statement (human PBMC isolation)

The project was approved by the Ethics Committee of the GovernmentMedical College and Hospital (GMCH), Sector 32, Chandigarh, India(GMCH/TA-1 [19]/2011/Agenda Number 2) and the Ethics and BiosafetyCommittee of the Institute of Microbial Technology (IMTECH), Sector 39A,Chandigarh, India (01/2011/IMT/IEC-Blood; 12/2010IMT/IBSC). The studywas conducted strictly in accordance with the Ethical Guidelines for Bio-medical Research on Human Subjects by the Central Ethics Committee onHuman Research, Indian Council of Medical Research-2000 and those ascontained in the Declaration of Helsinki. Each subject was provided withwritten information about the study, and written consent on the consent formwas obtained from each healthy volunteer before his or her induction in thestudy in the language (English, Hindi, and Punjabi) familiar to them.

Statistics

The results are expressed as mean 6 SD, and SigmaPlot and SPSS wereused for statistical analysis unless otherwise mentioned.

ResultsM. tuberculosis keto-MA is the heterologous ligand formacrophage nuclear receptor TR4

To determine whether M. tuberculosis cell wall components area possible source of heterologous ligands for TR4, a GAL4-based

The Journal of Immunology 297

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

FIGURE 1. M. tuberculosis keto-mycolic acid binds to macrophage nuclear receptor TR4. (A) Gal4-based dual-reporter assay was performed in COS-1 cells.

When M. tuberculosis H37Rv cell wall components (TDM, sulfolipids-1, PIM, PIM 6, LAM, LM, AG, PG, and m-AGP) were evaluated for their ability to

transactivate Gal4-TR4, MA was found to cause significant transactivation. Linoleic acid was used as a positive control. (B) Total MA extracted from

M. tuberculosis H37Rv and H37Ra,M. smegmatis, andM. phlei was confirmed with 1D TLC.M. smegmatis contains only a and a’-MA,M. phlei contains a-MA

and keto-MA, whereas H37Rv and H37Ra contains all three forms, a-MA, methoxy-MA, and keto-MA. (C) Total MA from each of these species was evaluated

for their ability to transactivate Gal4-TR4. Total MA from M. tuberculosis H37Rv and H37Ra and M. phlei showed an increased transactivation of TR4 as

compared with that of M. smegmatis, which lacks the oxygenated form of MA. (D) Individual MA from M. tuberculosis H37Rv was separated by 1D and 2D

TLC (Supplemental Fig. 1A), and spots corresponding to a-MA, methoxy-MA, and keto-MAwere scraped and purified. (E) a-MA, methoxy-MA, and keto-MA

from M. tuberculosis H37Rv were evaluated for their ability to transactivate TR4 by Gal4-based dual reporter and (F) CD36 (a target gene of TR4) gene

promoter reporter assays. For transactivation or promoter reporter assay, 1 mg/ml of individual lipids was used. Positive control linoleic acid was used at 1 mM.

(G) TR4-coated beads were incubated along with total MA followed by repeated washing. Precipitated beads were solvent extracted to release the bound MA,

which was run on analytical TLC plates for 1D and (H) 2D TLC. The spot of the bound MA corresponds to that of the keto-MA in the standard total MA from

H37Rv. (I) hMDMs with control or TR4 knockdown (shRNA) background were either stimulated with keto-MA or left unstimulated and then infected with

eitherM. smegmatis (for 24 h) orM. tuberculosis H37Rv for 48 h. Intracellular bacilli load in the infected macrophages was monitored by counting the number

of colonies (i.e., CFU) in 7H11 plates. Interestingly, keto-MAwas able to rescueM. smegmatis for survival in control but not TR4 knockdown background. Data

are representative of three independent experiments with similar results and are shown as mean 6 SD of the indicated number of experiments. *p , 0.05, with

linoleic acid control in reporter assay or with vehicle control and as otherwise depicted for CFU assay.

298 Keto-MA–TR4 AXIS IN M. TUBERCULOSIS–INDUCED FOAMY Mf

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

transactivation reporter assay was performed in COS1 cells (Fig.1A). Interestingly, the crude MA fraction significantly trans-activated TR4 compared with TDM, sulfolipids-1, PIM, PIM 6,LAM, LM, AG, PG, and m-AGP etc of M. tuberculosis H37Rv.To gain insight into the fraction or type of MA responsible for thetransactivation of TR4, total MA was extracted from M. tuber-culosis H37Rv and nonpathogenic strains, including M. tubercu-losis H37Ra, M. smegmatis, and M. phlei, and examined with 1Dand 2D TLC (Fig.1B, Supplemental Fig. 1A). The total MA ofboth M. tuberculosis H37Rv and H37Ra contained a-MA,methoxy-MA, and keto-MA, whereas the total MA of M. smeg-matis lacked keto-MA and methoxy-MA and that of M. phleilacked methoxy-MA (18, 37). Interestingly, a 10–14-fold trans-activation of GAL4-TR4 was observed with the total MA fromM. tuberculosis H37Rv/H37Ra andM. phlei; however, no significanttransactivation was observed with the total MA from M. smegmatis(Fig. 1C). These reporter assay observations suggest the possibleinvolvement of keto-MA in the modulation of TR4.The separation and characterization of total MA from M. tuber-

culosis H37Rv was performed with 1D and 2D TLC as described(24, 37), and the spots corresponding to a-MA, methoxy-MA, andketo-MA, were scraped and further extracted for analysis in theGAL4-based transactivation assay. The results of these assays con-firm the involvement of keto-MA in the modulation of TR4(Fig. 1D, 1E). A gene promoter reporter assay was also performedon TR4 target gene CD36 to have a direct measure of the MA–TR4crosstalk in modulating promoter activity (Fig. 1F) (31). Keto-MAwas observed to have the highest induction of TR4, which is similarto the results of the transactivation assay. Linoleic acid was used asa positive control in all the assays.To clarify further the class of MA that can function as a ligand

for TR4, we used a conventional bead-based pull-down approachin which bead-bound TR4 was allowed to bind in solution tothe total MA extract from M. tuberculosis H37Rv containing allthree classes of MA. Precipitated beads were solvent extracted toseparate the bound MA from the TR4. This MA fraction wasevaluated and analyzed with 1D and 2D TLC, which confirmedthat the identity of the bound MAwas keto-MA (Fig. 1G, 1H). Wealso confirmed of the pull-down fraction to be keto-MA by ana-lyzing the TLC run on hexane:diethyl ether with similar elutionprofile as that of petroleum ether:diethyl ether or dichloromethaneelution system in which spot III (methoxy-MA) and IV (keto-MA) merge because of similar mobility (37) (SupplementalFig. 1B). We were also able to rule out methoxy-MA in thepull down fraction by NMR characterization (SupplementalFig. 1C).To assess the role of keto-MA as a virulence factor in Myco-

bacterium survival, we examined the fate of M. smegmatis andM. tuberculosis H37Rv/H37Ra in hMDMs in control and TR4knockdown background (95% efficiency). Macrophages were infec-ted in the presence and absence of keto-MA. At 48 h after infection,intracellular bacteria was isolated and their viability was determinedby CFU assay (Fig.1I). Using an alternative approach, bacteria werestained with green fluorescent SYTO9, which stains both live anddead cells, and red fluorescent propidium iodide, which penetratesonly bacteria with ruptured membranes. Flow cytometry was per-formed and the percentage of dead bacteria was calculated(Supplemental Fig.1D). Interestingly, the addition of keto-MArescued M. smegmatis from clearance by the host comparedwith the M. smegmatis-infected sample without added keto-MA.The keto-MA–mediated increase in bug survival was abrogated inTR4 knockdown background.The above evidence firmly support the hypothesis that M. tu-

berculosis H37Rv keto-MA crosstalks with host NR, TR4. We

were then set to unravel the mechanism by which the keto-MAcrosstalk with host TR4 augments the survival of M. tuberculosis.

M. tuberculosis keto-MA induces foamy macrophages via hostnuclear receptor TR4

Formation of foamy macrophages (FMs) has been convincinglyimplicated in mycobacterial persistence, because the lipids withinthe granulomas in the host serve as a carbon source for the pathogenduring starvation (15). It has been shown that oxygenated MAfrom M. tuberculosis plays a vital role in FM generation frommacrophages (16), although the underlying mechanism has notbeen elucidated. To our knowledge, the present study is the first toestablish a link between keto-MA and the host nuclear receptorTR4 that leads to FMs, as observed with Nile Red and Oil Red Ostaining (Fig.2, Supplemental Fig. 2). Because TR4 has recentlybeen reported as a lipid sensor and has a role in foamy biogenesis(28, 31), we hypothesized that the underlying mechanism of theketo-MA induction of FMs could largely be attributed to its abilityto crosstalk with the host NR TR4. hMDMs were stained withNile Red after incubation with MA, M. tuberculosis H37Rv, orM. smegmatis and evaluated by confocal microscopy and flowcytometry (Fig. 2). As expected, cells infected with M. tuber-culosis H37Rv but not M. smegmatis differentiated into FMs.Strikingly however, M. tuberculosis H37Rv induction of FMsin TR4 knockdown cells (95% efficiency) was significantly re-duced. The addition of keto-MAwas effective in inducing FMs inhMDMs, which was significantly reduced in a TR4 knockdownbackground. We also evaluated the induction of FMs in THP-1macrophages by Oil Red O staining (Supplemental Fig. 2A, 2B).A similar pattern was observed. FM induction was maximum incontrol THP-1 cells infected or induced with M. tuberculosisH37Rv (∼70% of the infected cells) or keto-MA (∼70% of thetotal cells compared with control), followed by methoxy-MA(∼10%), and a-MA (∼5%). THP-1 macrophages from cells witha stable knockdown of TR4 (95% efficiency) had less induction ofFMs after stimulation with M. tuberculosis H37Rv (∼25% of theinfected cells) or keto-MA (∼25% of total cells compared withcontrol), whereas methoxy-MA (∼11%) and a-MA (∼4%) addi-tion showed no significant change. Our data suggest a definitemechanism by which M. tuberculosis H37Rv keto-MA interactswith the host NR TR4 to ensure pathogen survival by inducingFMs, which behave as a nutritional niche for M. tuberculosis andcontribute to granuloma formation.

Keto-MA crosstalk with TR4 functions in M. tuberculosissurvival and induction of tuberculous granuloma in vitro andin vivo

Granuloma formation, which is a hallmark of tuberculosis infec-tion, involves recruitment and formation of FMs that subsequentlyare surrounded by lymphocytes (40). Our observations thus farindicate a strong role for keto-MA in FM formation via crosstalkwith TR4. To further define the role of MA, and in particular keto-MA, in granuloma induction, keto-MA–coated beads and totalMA-coated beads were incubated with human PBMCs, and theformation of in vitro granulomas was monitored for 8 d as de-scribed earlier (41). Uncoated beads were used as a negativecontrol as they have earlier been shown to have minimal recruit-ment of the cells. TDM-coated beads, which are known to inducegranulomas, were used as a positive control. Total MA- and keto-MA–coated beads induced the recruitment of cells and theformation of a cellular monolayer starting from day 1. The mono-layer continued growing, and granuloma-like structures were seenat day 5 (Fig. 3A). We further evaluated the role of the crosstalkbetween keto-MA and the host NR TR4 in modulating granuloma

The Journal of Immunology 299

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

formation. The knockdown of TR4 in human PBMCs was foundto decrease migration of the cells around keto-MA–coated beads(Fig. 3B, Supplemental Fig. 3A). In keeping with its reported role,TDM was a potent inducer of granulomatous response in allbackgrounds, which is understandable because alone it fails totransactivate TR4 (Fig.1A, Supplemental Fig. 3B).To assess specifically the role of the host NR TR4 in M. tuber-

culosis H37Rv survival in vivo, an adeno-based knockdown of TR4in mice was attempted through delivery by tail vein injection andaerosols. The observed TR4 knockdown of 50–60% (SupplementalFig. 3C) in alveolar macrophages corresponds with the reduced

survival ofM. tuberculosis H37Rv determined at different times 15,30, and 60 d after infection. (Fig. 3C). Histopathology of the lungdissected 60 d after infection from TR4 knockdown mice comparedwith lung from control mice showed a reduction in the size of thedeveloping follicular granulomas as visualized by H&E staining(Fig. 3D). This reduction in the size of granulomas is seeminglya consequence of increased clearance of M. tuberculosis H37Rv(Fig. 3C, 3D). In addition, when monitored over a period of 20 wk,the control mice had prevailing disease symptoms of laziness andslow movement, particularly at later time points, with a slightchange in the body weight as compared with the TR4 knockdown

FIGURE 2. Keto-MA–TR4 axis induces FM phenotype in hMDMs. (A) Control or (B) TR4 knockdown hMDMs were infected with either M. tuber-

culosis H37Rv or M. smegmatis or stimulated with keto-MA for 48 h and stained with Nile Red. The extent of FMs was assessed by fluorescence mi-

croscopy (original magnification 360) and flow cytometry. Data are representative of three independent experiments with similar results.

300 Keto-MA–TR4 AXIS IN M. TUBERCULOSIS–INDUCED FOAMY Mf

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

mice (Supplemental Fig. 3D). Lung histopathology of the miceinfected with M. smegmatis appears normal without any sign ofgranuloma formation (Supplemental Fig. 3E, 3F). However,when keto-MA from M. tuberculosis H37Rv was administeredalong with M. smegmatis infection, there was multiple well-formed granuloma structure. Interestingly, the coadministrationin the TR4 knockdown mice, had a reduced number of granulomaformation, which clearly explains the importance of TR4 for theketo-MA mediated granuloma formation.

Noncanonical binding of keto-MA to host NR TR4

Docking and ITC studies were performed to demonstrate the ob-served binding of keto-MA to TR4. We performed in silico rigiddocking studies by making a grid around the predicted and modeledTR4 ligand binding domain (LBD: 349–615) and docking keto-MA. Interestingly, whereas linoleic acid, a known ligand of TR4,showed a negative binding energy that falls in the range observedwith other TR4 known ligands (26.54 to 23.92; data not shown)and is suggestive of a stable complex formation, keto-MA showeda positive binding energy (Fig. 4A). ITC titration experiments inwhich the sample cell contained full-length human TR4 (1–615)and the syringe contained keto-MA revealed a sigmoidal titrationwith an analyzed N value of ∼1.59, which indicates that keto-MA

is probably binding to TR4 at two binding sites (Fig. 4B). Com-plete saturation was not obtained, which has also been reported forheme-Reverb binding (42). The process was energetically favorable.ITC titration experiments performed with a-MA and methoxy-MAdid not show any binding pattern (Fig.4B, Supplemental Fig. 4A).Docking studies and ITC experiments suggest that keto-MA acts asa heterologous ligand with presumably a noncanonical binding to aTR4 full-length dimer. Circular dichroism experiments were per-formed to investigate the effect of keto-MA binding on the thermalstability of hTR4. Interestingly, keto-MA increased the meltingtemperature (Tm) of hTR4 full-length from 62˚C to 65˚C (Fig. 4C).Finally, the dose-dependent activation of hTR4 by keto-MAwas observed by transactivation reporter assay (Fig. 4D), whichfurther supports our claim on the specificity of the binding of keto-MA to hTR4. Our data lead to the conclusion that keto-MA func-tions as a virulent factor for M. tuberculosis by hijacking the hostNR TR4 function. This report proposes the noncanonical binding ofa host orphan receptor, TR4, with a heterologous ligand, keto-MA(Fig. 5).

DiscussionM. tuberculosis lipids are known to be efficiently shared and se-creted within the infected macrophage and into the extracellular

FIGURE 3. Keto-MA induces of granu-

loma formation vide host NR TR4. (A)

PBMCs were incubated with beads that

were uncoated or coated with TDM, total

MA, or keto-MA, and granuloma generation

was monitored. Representative images of

cellular aggregation at day 5 are shown

(original magnification3200, objective 203).

(B) PBMCs were infected ex vivo with ad-

enoviral constructs encoding control shRNA

(LacZ) or shRNA specific for TR4 for 24 h

before being incubated with uncoated beads

or beads coated with keto-MA, and granu-

loma generation was monitored. Represen-

tative images of cellular aggregation at day 5

are shown (original magnification 3200). (C)

Adeno-based TR4 knockdown was performed

in mice via tail vein injection and aerosol

challenge. TR4 expression was checked in

lungs and liver, which showed a 50–60% re-

duction from the control. Following an aerosol

challenge with M. tuberculosis H37Rv, the

CFU count was monitored to check the

M. tuberculosis survival in the TR4 knock-

down mice versus the control mice. (D) A

histopathologic study of the lung (60 d

postinfection) was performed to determine

the granuloma profiles in TR4 knockdown

and control mice (original magnification

320, 3100). The arrowhead indicates the

representative granuloma area (counted)

visible because of the positive stain with

H&E. Data are representative of three in-

dependent experiments with similar results.

Error bars indicate mean 6 SD. *p , 0.05,

indicating significant differences from the

control.

The Journal of Immunology 301

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

milieu (5, 43). Among the secreted lipids, the cell-wall lipid TDMand LM have already been characterized as bioactive componentsduring granuloma formation and differentiation of macrophagesinto multinucleated giant cells, respectively (44–47). Recently, therole of oxygenated forms of MA during foam-cell formation washighlighted (16). To our knowledge, the present study is the first toprovide mechanistic insight into keto-MA–induced FM, whichinvolves crosstalk with the host LSNR TR4. TR4 reporter assayswith cell wall components and MA from different Mycobacteriumspecies as well as biochemical (TLC) and biophysical (ITC,NMR) characterization of the TR4-bound MA fraction identifiedketo-MA as a heterologous ligand that shows a noncanonicalbinding to the host NR TR4. This crosstalk provides insight intothe previously observed role of oxygenated MA in the differen-tiation of macrophages to FMs and in granuloma formation (16).The relevance of keto-MA in FM formation is emphasized by thefact that M. smegmatis, which lacks keto-MA, induces only ∼5%of the FM formation seen with M. tuberculosis. In comparison,transformed M. smegmatis that carries the hma gene coding forketo-MA induced up to ∼60% FM (16). Similarly, our studyreveals the importance of the crosstalk between the cell wallcomponent keto-MA of M. tuberculosis and the host factor TR4.Previously, TDM was shown to be recognized by TLR2 and thescavenger receptor MARCO (macrophage receptor with a collag-enous structure) (48), whereas LM modulation was TLR2 de-pendent (47).TR4 has been shown unequivocally to induce FMs partly by

regulating oxidized LDL receptor CD36 (28, 31). Our previous

study also suggested that crosstalk between host TR4 and mem-bers of the M. tuberculosis lipid repertoire is a prerequisite for theinduction of FM in M. tuberculosis, although the exact identityof the heterologous ligand was lacking (28). The present studyestablishes the host–pathogen axis as keto-MA–TR4, in which thehost factor TR4 is indispensable for keto-MA–induced FM. Thiswas evident in the experiment with hMDMs with a TR4 knock-down background that were treated with keto-MA, stained withNile red and Oil Red O, and monitored with imaging and flowcytometry (Fig. 2, Supplemental Fig. 2). FMs are the sites thatsustain the pathogen within the granuloma core (16, 45). Given theinvolvement of TR4 in the modulation of FM, it also possiblyplays a role during granuloma formation. As FMs appear amongM. tuberculosis/PPD/TDM granuloma, there might be a resem-blance in tuberculous granuloma structure arising from thesedeterminants and keto-MA. Pathogen-induced FMs have beenreported in cases of Chlamydia (49) and toxoplasma infections(50) and in diseases such as atherosclerosis (51). It is tempting tospeculate that the lipid-metabolite-host LSNR cellular deregula-tion observed in TB and atherosclerosis (28, 52) could be genericto other infections and diseases.The dynamics of aggregation of immune cells during TB in-

fection leads to granuloma formation (53). This involves changesin macrophage morphology, differentiation, and polarization.Mature macrophages appear epithelial (54); however, they mayform multinucleated giant cells (55) or FMs with increased lipidbodies (40). Although oxygenated MA fromM. tuberculosis H37Rvhas been shown to modulate granuloma formation, the interacting

FIGURE 4. Stable but noncanonical bind-

ing of keto-MA to TR4. (A) Ligand binding

pocket of human TR4 LBD active confor-

mation was generated based on an active

conformation structure of RXRa (1FM6). A

known TR4 ligand, linoleic acid or keto-MA,

was docked into the ligand-binding pocket of

hTR4 LBD. A negative docking score indi-

cated a stable complex formation with lino-

leic acid; however, a positive score was seen

with keto-MA. (B) ITC experiment repre-

senting the titration of keto-MA with TR4

full-length protein. The titration contained

3 mM TR4 in the sample cell and 45 mM

keto-MA in the ligand syringe. Background

(titration of keto-MA into buffer) was sub-

tracted from the data points from titrating

keto-MA with TR4 full-length protein, and

curve fitting was used. The solid line repre-

sents the best nonlinear fit. (C) Far-UV ther-

mal melts of apoprotein and keto-MA bound

form of hTR4. The Tm value for the apo form

hTR4 was ∼62˚C, and that of keto-MA bound

form was ∼65˚C. (D) A dose-dependent TR4

transactivation reporter assay using different

concentration of keto-MA. Data are represen-

tative of three independent experiments with

similar results. Error bars indicate mean6 SD.

302 Keto-MA–TR4 AXIS IN M. TUBERCULOSIS–INDUCED FOAMY Mf

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

partners from the host were unknown. Experiments in the currentstudy involving an in vitro model of granuloma formation (16, 41,47) validate the presence of the keto-MA–TR4 axis in the gran-uloma, as seen from the increased number of cells bound to thebeads coated with keto-MA, similar to the number with thepositive control TDM (Fig.3A). Furthermore, it is clear that TR4is indispensable to keto-MA–induced granuloma formation(Fig. 3B). In vivo infection studies in mice showed significantabatement in bacterial burden and consequent reduction in size ofdeveloping follicular granuloma in lungs of mice with a TR4knockdown background. Taken together, these observations con-firm that LSNR TR4 functions as a host factor augmenting thesurvival ofM. tuberculosis by impeding its clearance and inducinggranuloma. Our previous report (28) showed that TR4 polarizesmacrophages toward the less microbicidal and immunomodula-tory M2 phenotype with a consequential increase in FMs, blocksphagolysosome maturation, and reduces levels of ROS/NO, and asresult aids in the survival of M. tuberculosis. These observations,as well as the effects of other unknown interactions by keto-MA–TR4, expose the vulnerability of the host. Greater comprehensiveinsight into the downstream physiology and crosstalk is requiredfor a more concerted intervention.The roles and relevance of several host molecules that are

manipulated by the pathogen are underestimated and thus farinadequately addressed. Because of emergent threats of multidrug-resistant and extensively drug-resistant strains, it is exigent toidentify host factors playing combatants and cohorts and to assesstheir amenability to pharmacologic modulation. A novel designwould be a combinatorial therapy that, while targeting the path-ogen with frontline drugs, also targets the host cohorts or em-powers the host combatants. Several host factors have recentlybeen implicated in M. tuberculosis survival and clearance. Thesepathogen-mediated factors include prostanoid receptor EP2 (56),protein kinase R (57), genes involved in the calcium and cysteineprotease pathways (58), the NOD-like receptor family, pyrindomain containing 3 (NLRP3) inflammasome (59), and severalhost-dependent survival factors screened by siRNA (60). Ourrecent work along with other reports has shown that PPARg and

TR4 provoke the anti-inflammatory and pro–M. tuberculosisproperties of macrophages (28, 32, 34). In addition, differentgroups have also reported other nuclear receptors such as vitaminD receptors, Rev-erba, and LXRa anti-infectives functions againstM. tuberculosis (27, 28, 61–65).NRs in particular have been shown to modulate host effectors

that are imperative for clearance or survival of intracellularmicroparasites Listeria monocytogenes, Salmonella typhimurium,Leishmania donovani, and M. tuberculosis (27, 32, 34, 66–69).Compared with members of a signaling network, NRs are func-tionally less pleiotropic, less redundant, and more amenable topharmacologic modulation and, as such, better therapeutic targets.Given their physiologic importance in human disease, not sur-prisingly NRs have been reported to be modulated by variousagonists and antagonists composed of both biologically andpharmacologically synthesized drugs, a few of which have alreadybeen approved by the U.S. Food and Drug Administration (70–74). However, the use of such drugs has not been evaluated inother heterologous contexts, such as infection, and the present TBdrug regimens primarily target pathogens (75, 76). We have ex-tensively studied and identified keto-MA as a heterologous spe-cific ligand for TR4, which was also confirmed by gene promoterreporter assay in the PPARg and TR4 knockdown backgroundsand the failure of PPARg to pull down keto-MA (Fig.4D,Supplemental Fig. 4C). Design of inhibitory analogs to this het-erologous ligand-receptor binding and evaluation of their efficacytogether with frontline TB drugs holds promise as a combinatorialadjunct therapy.

AcknowledgmentsWe thank Dr. B.N. Datta for histopathology; Ishwinder Kaur for assistance

with TLC; Deepak Bhatt, Anjali Koundal, and Dr. Mohan Pal, laboratory

members, and volunteer donors for their help; Dr. Girish Sahni for support;

and Council of Scientific and Industrial Research-IMTECH for providing

facilities.

DisclosuresThe authors have no financial conflicts of interest.

FIGURE 5. A schematic representation of the

mechanistic crosstalk between M. tuberculosis keto-

MA and TR4, and consequent FM and granuloma

formation. Upon infection, macrophages phagocytosed

M. tuberculosis through the scavenger and complement

receptors. Once inside the host, M. tuberculosis is

retained within the phagosomes, and it secretes its cell

wall components. Thereafter, the secreted components,

mainly lipids and mycolic acid, are trafficked from the

phagosome into the cytosol. Various host factors are

present as sensors and effectors in the cellular milieu.

We identified the nuclear receptor TR4 binding to the

secreted mycolic acid (keto-MA). Activation of the

receptor by ligand binding leads to downstream regu-

lation of gene involved in lipid biogenesis through its

binding to its target gene having the TR4 response el-

ement. CD36, a scavenger receptor known to be in-

volved in lipid uptake is modulated by TR4 and leads to

increased lipid droplets formation and FMs. In addi-

tion, in vitro and in vivo model studies reveal that keto-

MA–TR4 crosstalk consequently leads to an increase in

granuloma formation.

The Journal of Immunology 303

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

References1. Lienhardt, C., P. Glaziou, M. Uplekar, K. Lonnroth, H. Getahun, and

M. Raviglione. 2012. Global tuberculosis control: lessons learnt and futureprospects. Nat. Rev. Microbiol. 10: 407–416.

2. WHO. 2013. Global Tuberculosis Control. World Health Organization, Geneva,Switzerland.

3. Cole, S. T., R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris,S. V. Gordon, K. Eiglmeier, S. Gas, C. E. Barry, III, et al. 1998. Deciphering thebiology of Mycobacterium tuberculosis from the complete genome sequence.Nature 393: 537–544.

4. Fontan, P., V. Aris, S. Ghanny, P. Soteropoulos, and I. Smith. 2008. Globaltranscriptional profile of Mycobacterium tuberculosis during THP-1 humanmacrophage infection. Infect. Immun. 76: 717–725.

5. Beatty, W. L., E. R. Rhoades, H. J. Ullrich, D. Chatterjee, J. E. Heuser, andD. G. Russell. 2000. Trafficking and release of mycobacterial lipids frominfected macrophages. Traffic 1: 235–247.

6. Barrow, W. W., J. P. de Sousa, T. L. Davis, E. L. Wright, M. Bachelet, andN. Rastogi. 1993. Immunomodulation of human peripheral blood mononuclearcell functions by defined lipid fractions of Mycobacterium avium. Infect. Immun.61: 5286–5293.

7. Moreno, C., J. Taverne, A. Mehlert, C. A. Bate, R. J. Brealey, A. Meager,G. A. Rook, and J. H. Playfair. 1989. Lipoarabinomannan from Mycobacteriumtuberculosis induces the production of tumour necrosis factor from human andmurine macrophages. Clin. Exp. Immunol. 76: 240–245.

8. Rocha-Ramırez, L. M., I. Estrada-Garcıa, L. M. Lopez-Marın, E. Segura-Salinas,P. Mendez-Aragon, D. Van Soolingen, R. Torres-Gonzalez, R. Chacon-Salinas,S. Estrada-Parra, C. Maldonado-Bernal, et al. 2008. Mycobacterium tuberculosislipids regulate cytokines, TLR-2/4 and MHC class II expression in humanmacrophages. Tuberculosis (Edinb.) 88: 212–220.

9. Bekierkunst, A., I. S. Levij, E. Yarkoni, E. Vilkas, A. Adam, and E. Lederer.1969. Granuloma formation induced in mice by chemically defined mycobac-terial fractions. J. Bacteriol. 100: 95–102.

10. Reggiardo, Z., and A. K. Shamsuddin. 1976. Granulomagenic activity of sero-logically active glycolipids from Mycobacterium bovis BCG. Infect. Immun. 14:1369–1374.

11. Cox, J. S., B. Chen, M. McNeil, and W. R. Jacobs, Jr. 1999. Complex lipiddetermines tissue-specific replication of Mycobacterium tuberculosis in mice.Nature 402: 79–83.

12. Goren, M. B., O. Brokl, and W. B. Schaefer. 1974. Lipids of putative relevance tovirulence in Mycobacterium tuberculosis: correlation of virulence with elabo-ration of sulfatides and strongly acidic lipids. Infect. Immun. 9: 142–149.

13. Dubnau, E., J. Chan, C. Raynaud, V. P. Mohan, M. A. Laneelle, K. Yu,A. Quemard, I. Smith, and M. Daffe. 2000. Oxygenated mycolic acids arenecessary for virulence of Mycobacterium tuberculosis in mice. Mol. Microbiol.36: 630–637.

14. Dascher, C. C., K. Hiromatsu, X. Xiong, C. Morehouse, G. Watts, G. Liu,D. N. McMurray, K. P. LeClair, S. A. Porcelli, and M. B. Brenner. 2003. Im-munization with a mycobacterial lipid vaccine improves pulmonary pathology inthe guinea pig model of tuberculosis. Int. Immunol. 15: 915–925.

15. Ehrt, S., and D. Schnappinger. 2007. Mycobacterium tuberculosis virulence:lipids inside and out. Nat. Med. 13: 284–285.

16. Peyron, P., J. Vaubourgeix, Y. Poquet, F. Levillain, C. Botanch, F. Bardou,M. Daffe, J. F. Emile, B. Marchou, P. J. Cardona, et al. 2008. Foamy macro-phages from tuberculous patients’ granulomas constitute a nutrient-rich reservoirfor M. tuberculosis persistence. PLoS Pathog. 4: e1000204.

17. Watanabe, M., Y. Aoyagi, H. Mitome, T. Fujita, H. Naoki, M. Ridell, andD. E. Minnikin. 2002. Location of functional groups in mycobacterial mer-omycolate chains; the recognition of new structural principles in mycolic acids.Microbiology 148: 1881–1902.

18. Watanabe, M., Y. Aoyagi, M. Ridell, and D. E. Minnikin. 2001. Separation andcharacterization of individual mycolic acids in representative mycobacteria.Microbiology 147: 1825–1837.

19. Brennan, P. J. 2003. Structure, function, and biogenesis of the cell wall of My-cobacterium tuberculosis. Tuberculosis (Edinb.) 83: 91–97.

20. Karakousis, P. C., W. R. Bishai, and S. E. Dorman. 2004. Mycobacterium tu-berculosis cell envelope lipids and the host immune response. Cell. Microbiol. 6:105–116.

21. Ryll, R., Y. Kumazawa, and I. Yano. 2001. Immunological properties of treha-lose dimycolate (cord factor) and other mycolic acid-containing glycolipids—a review. Microbiol. Immunol. 45: 801–811.

22. Ojha, A. K., A. D. Baughn, D. Sambandan, T. Hsu, X. Trivelli, Y. Guerardel,A. Alahari, L. Kremer, W. R. Jacobs, Jr., and G. F. Hatfull. 2008. Growth ofMycobacterium tuberculosis biofilms containing free mycolic acids and har-bouring drug-tolerant bacteria. Mol. Microbiol. 69: 164–174.

23. Ojha, A. K., X. Trivelli, Y. Guerardel, L. Kremer, and G. F. Hatfull. 2010. En-zymatic hydrolysis of trehalose dimycolate releases free mycolic acids duringmycobacterial growth in biofilms. J. Biol. Chem. 285: 17380–17389.

24. Minnikin, D. E., L. Alshamaony, and M. Goodfellow. 1975. Differentiationof Mycobacterium, Nocardia, and related taxa by thin-layer chromato-graphic analysis of whole-organism methanolysates. J. Gen. Microbiol. 88:200–204.

25. Levy-Frebault, V., K. S. Goh, and H. L. David. 1986. Mycolic acid analysis forclinical identification ofMycobacterium avium and related mycobacteria. J. Clin.Microbiol. 24: 835–839.

26. Wong, M. Y., P. A. Steck, and G. R. Gray. 1979. The major mycolic acids ofMycobacterium smegmatis. Characterization of their homologous series. J. Biol.Chem. 254: 5734–5740.

27. Liu, P. T., S. Stenger, H. Li, L. Wenzel, B. H. Tan, S. R. Krutzik, M. T. Ochoa,J. Schauber, K. Wu, C. Meinken, et al. 2006. Toll-like receptor triggering ofa vitamin D-mediated human antimicrobial response. Science 311: 1770–1773.

28. Mahajan, S., H. K. Dkhar, V. Chandra, S. Dave, R. Nanduri, A. K. Janmeja,J. N. Agrewala, and P. Gupta. 2012. Mycobacterium tuberculosis modulatesmacrophage lipid-sensing nuclear receptors PPARg and TR4 for survival. J.Immunol. 188: 5593–5603.

29. Sasindran, S. J., and J. B. Torrelles. 2011. Mycobacterium Tuberculosis Infectionand Inflammation: what is Beneficial for the Host and for the Bacterium? Front.Microbiol. 2: 2.

30. Varga, T., and L. Nagy. 2008. Nuclear receptors, transcription factors linkinglipid metabolism and immunity: the case of peroxisome proliferator-activatedreceptor gamma. Eur. J. Clin. Invest. 38: 695–707.

31. Xie, S., Y. F. Lee, E. Kim, L. M. Chen, J. Ni, L. Y. Fang, S. Liu, S. J. Lin, J. Abe,B. Berk, et al. 2009. TR4 nuclear receptor functions as a fatty acid sensor tomodulate CD36 expression and foam cell formation. Proc. Natl. Acad. Sci. USA106: 13353–13358.

32. Rajaram, M. V., M. N. Brooks, J. D. Morris, J. B. Torrelles, A. K. Azad, andL. S. Schlesinger. 2010. Mycobacterium tuberculosis activates human macro-phage peroxisome proliferator-activated receptor gamma linking mannose re-ceptor recognition to regulation of immune responses. J. Immunol. 185: 929–942.

33. Tsai, N. P., M. Huq, P. Gupta, K. Yamamoto, H. Kagechika, and L. N. Wei. 2009.Activation of testicular orphan receptor 4 by fatty acids. Biochim. Biophys. Acta1789: 734–740.

34. Almeida, P. E., A. R. Silva, C. M. Maya-Monteiro, D. Torocsik, H. D’Avila,B. Dezso, K. G. Magalhaes, H. C. Castro-Faria-Neto, L. Nagy, and P. T. Bozza.2009. Mycobacterium bovis bacillus Calmette-Guerin infection induces TLR2-dependent peroxisome proliferator-activated receptor gamma expression andactivation: functions in inflammation, lipid metabolism, and pathogenesis. J.Immunol. 183: 1337–1345.

35. Sorrentino, R., S. Morello, S. Chen, E. Bonavita, and A. Pinto. 2010. The ac-tivation of liver X receptors inhibits toll-like receptor-9-induced foam cell for-mation. J. Cell. Physiol. 223: 158–167.

36. Butler, W. R., M. M. Floyd, J. M. Brown, S. R. Toney, M. I. Daneshvar,R. C. Cooksey, J. Carr, A. G. Steigerwalt, and N. Charles. 2005. Novel mycolicacid-containing bacteria in the family Segniliparaceae fam. nov., including thegenus Segniliparus gen. nov., with descriptions of Segniliparus rotundus sp. nov.and Segniliparus rugosus sp. nov. Int. J. Syst. Evol. Microbiol. 55: 1615–1624.

37. Secanella-Fandos, S., M. Luquin, M. Perez-Trujillo, and E. Julian. 2011.Revisited mycolic acid pattern of Mycobacterium confluentis using thin-layerchromatography. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 879: 2821–2826.

38. Arnold, K., L. Bordoli, J. Kopp, and T. Schwede. 2006. The SWISS-MODELworkspace: a web-based environment for protein structure homology modelling.Bioinformatics 22: 195–201.

39. Morris, G. M., R. Huey, W. Lindstrom, M. F. Sanner, R. K. Belew, D. S. Goodsell,and A. J. Olson. 2009. AutoDock4 and AutoDockTools4: Automated dockingwith selective receptor flexibility. J. Comput. Chem. 30: 2785–2791.

40. Russell, D. G., P. J. Cardona, M. J. Kim, S. Allain, and F. Altare. 2009. Foamymacrophages and the progression of the human tuberculosis granuloma. Nat.Immunol. 10: 943–948.

41. Puissegur, M. P., C. Botanch, J. L. Duteyrat, G. Delsol, C. Caratero, andF. Altare. 2004. An in vitro dual model of mycobacterial granulomas to inves-tigate the molecular interactions between mycobacteria and human host cells.Cell. Microbiol. 6: 423–433.

42. Raghuram, S., K. R. Stayrook, P. Huang, P. M. Rogers, A. K. Nosie,D. B. McClure, L. L. Burris, S. Khorasanizadeh, T. P. Burris, and F. Rastinejad.2007. Identification of heme as the ligand for the orphan nuclear receptors REV-ERBalpha and REV-ERBbeta. Nat. Struct. Mol. Biol. 14: 1207–1213.

43. Beatty, W. L., H. J. Ullrich, and D. G. Russell. 2001. Mycobacterial surfacemoieties are released from infected macrophages by a constitutive exocyticevent. Eur. J. Cell Biol. 80: 31–40.

44. Geisel, R. E., K. Sakamoto, D. G. Russell, and E. R. Rhoades. 2005. In vivoactivity of released cell wall lipids of Mycobacterium bovis bacillus Calmette-Guerin is due principally to trehalose mycolates. J. Immunol. 174: 5007–5015.

45. Hunter, R. L., C. Jagannath, and J. K. Actor. 2007. Pathology of postprimarytuberculosis in humans and mice: contradiction of long-held beliefs. Tubercu-losis (Edinb.) 87: 267–278.

46. Hunter, R. L., M. Olsen, C. Jagannath, and J. K. Actor. 2006. Trehalose 6,69-dimycolate and lipid in the pathogenesis of caseating granulomas of tuberculosisin mice. Am. J. Pathol. 168: 1249–1261.

47. Puissegur, M. P., G. Lay, M. Gilleron, L. Botella, J. Nigou, H. Marrakchi,B. Mari, J. L. Duteyrat, Y. Guerardel, L. Kremer, et al. 2007. Mycobacteriallipomannan induces granuloma macrophage fusion via a TLR2-dependent,ADAM9- and beta1 integrin-mediated pathway. J. Immunol. 178: 3161–3169.

48. Bowdish, D. M., K. Sakamoto, M. J. Kim, M. Kroos, S. Mukhopadhyay,C. A. Leifer, K. Tryggvason, S. Gordon, and D. G. Russell. 2009. MARCO,TLR2, and CD14 are required for macrophage cytokine responses to myco-bacterial trehalose dimycolate and Mycobacterium tuberculosis. PLoS Pathog. 5:e1000474.

49. Kalayoglu, M. V., and G. I. Byrne. 1998. Induction of macrophage foam cellformation by Chlamydia pneumoniae. J. Infect. Dis. 177: 725–729.

50. Portugal, L. R., L. R. Fernandes, V. S. Pietra Pedroso, H. C. Santiago,R. T. Gazzinelli, and J. I. Alvarez-Leite. 2008. Influence of low-density lipo-protein (LDL) receptor on lipid composition, inflammation and parasitism duringToxoplasma gondii infection. Microbes Infect. 10: 276–284.

304 Keto-MA–TR4 AXIS IN M. TUBERCULOSIS–INDUCED FOAMY Mf

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

51. Galkina, E., and K. Ley. 2009. Immune and inflammatory mechanisms of ath-erosclerosis (*). Annu. Rev. Immunol. 27: 165–197.

52. Prieur, X., T. Roszer, and M. Ricote. 2010. Lipotoxicity in macrophages: evi-dence from diseases associated with the metabolic syndrome. Biochim. Biophys.Acta 1801: 327–337.

53. Ramakrishnan, L. 2012. Revisiting the role of the granuloma in tuberculosis.Nat. Rev. Immunol. 12: 352–366.

54. Adams, D. O. 1974. The structure of mononuclear phagocytes differentiatingin vivo. I. Sequential fine and histologic studies of the effect of BacillusCalmette-Guerin (BCG). Am. J. Pathol. 76: 17–48.

55. Helming, L., and S. Gordon. 2007. The molecular basis of macrophage fusion.Immunobiology 212: 785–793.

56. Kaul, V., D. Bhattacharya, Y. Singh, L. Van Kaer, M. Peters-Golden,W. R. Bishai, and G. Das. 2012. An important role of prostanoid receptor EP2 inhost resistance to Mycobacterium tuberculosis infection in mice. J. Infect. Dis.206: 1816–1825.

57. Wu, K., J. Koo, X. Jiang, R. Chen, S. N. Cohen, and C. Nathan. 2012. Improvedcontrol of tuberculosis and activation of macrophages in mice lacking proteinkinase R. PLoS ONE 7: e30512.

58. Singhal, J., N. Agrawal, M. Vashishta, N. G. Priya, B. K. Tiwari, Y. Singh,R. Raman, and K. Natarajan. 2012. Suppression of dendritic cell-mediatedresponses by genes in calcium and cysteine protease pathways during Myco-bacterium tuberculosis infection. J. Biol. Chem. 287: 11108–11121.

59. Wong, K. W., and W. R. Jacobs, Jr. 2011. Critical role for NLRP3 in necroticdeath triggered by Mycobacterium tuberculosis. Cell. Microbiol. 13: 1371–1384.

60. Jayaswal, S., M. A. Kamal, R. Dua, S. Gupta, T. Majumdar, G. Das, D. Kumar,and K. V. Rao. 2010. Identification of host-dependent survival factors for in-tracellular Mycobacterium tuberculosis through an siRNA screen. PLoS Pathog.6: e1000839.

61. Chandra, V., S. Mahajan, A. Saini, H. K. Dkhar, R. Nanduri, E. B. Raj,A. Kumar, and P. Gupta. 2013. Human IL10 gene repression by Rev-erbaameliorates Mycobacterium tuberculosis clearance. J. Biol. Chem. 288: 10692–10702.

62. Hewison, M. 2011. Antibacterial effects of vitamin D. Nat. Rev. Endocrinol. 7:337–345.

63. Khoo, A. L., L. Y. Chai, H. J. Koenen, M. Oosting, A. Steinmeyer, U. Zuegel,I. Joosten, M. G. Netea, and A. J. van der Ven. 2011. Vitamin D(3) down-regulates proinflammatory cytokine response to Mycobacterium tuberculosisthrough pattern recognition receptors while inducing protective cathelicidinproduction. Cytokine 55: 294–300.

64. Yuk, J. M., D. M. Shin, H. M. Lee, C. S. Yang, H. S. Jin, K. K. Kim, Z. W. Lee,S. H. Lee, J. M. Kim, and E. K. Jo. 2009. Vitamin D3 induces autophagy inhuman monocytes/macrophages via cathelicidin. Cell Host Microbe 6: 231–243.

65. Fabri, M., S. Stenger, D. M. Shin, J. M. Yuk, P. T. Liu, S. Realegeno, H. M. Lee,S. R. Krutzik, M. Schenk, P. A. Sieling, R. Teles, D. Montoya, S. S. Iyer, H. Bruns,D. M. Lewinsohn, B. W. Hollis, M. Hewison, J. S. Adams, A. Steinmeyer, U. Zugel,G. Cheng, E. K. Jo, B. R. Bloom, and R. L. Modlin. 2011. Vitamin D is required forIFN-gamma-mediated antimicrobial activity of human macrophages. Sci. Transl.Med. 3: 104ra102.

66. Chan, M. M., N. Adapala, and C. Chen. 2012. Peroxisome Proliferator-ActivatedReceptor-g-Mediated Polarization of Macrophages in Leishmania Infection.PPAR Res. 2012: 796235.

67. Joseph, S. B., M. N. Bradley, A. Castrillo, K. W. Bruhn, P. A. Mak, L. Pei,J. Hogenesch, R. M. O’connell, G. Cheng, E. Saez, et al. 2004. LXR-dependentgene expression is important for macrophage survival and the innate immuneresponse. Cell 119: 299–309.

68. Korf, H., S. Vander Beken, M. Romano, K. R. Steffensen, B. Stijlemans,J. A. Gustafsson, J. Grooten, and K. Huygen. 2009. Liver X receptors contributeto the protective immune response against Mycobacterium tuberculosis in mice.J. Clin. Invest. 119: 1626–1637.

69. Valledor, A. F., L. C. Hsu, S. Ogawa, D. Sawka-Verhelle, M. Karin, andC. K. Glass. 2004. Activation of liver X receptors and retinoid X receptorsprevents bacterial-induced macrophage apoptosis. Proc. Natl. Acad. Sci. USA101: 17813–17818.

70. Bouhlel, M. A., B. Derudas, E. Rigamonti, R. Dievart, J. Brozek, S. Haulon,C. Zawadzki, B. Jude, G. Torpier, N. Marx, et al. 2007. PPARgamma activationprimes human monocytes into alternative M2 macrophages with anti-inflammatory properties. Cell Metab. 6: 137–143.

71. Li, L., S. M. Dial, M. Schmelz, M. A. Rennels, and N. M. Ampel. 2005. Cellularimmune suppressor activity resides in lymphocyte cell clusters adjacent togranulomata in human coccidioidomycosis. Infect. Immun. 73: 3923–3928.

72. Rigamonti, E., G. Chinetti-Gbaguidi, and B. Staels. 2008. Regulation of mac-rophage functions by PPAR-alpha, PPAR-gamma, and LXRs in mice and men.Arterioscler. Thromb. Vasc. Biol. 28: 1050–1059.

73. Sladek, F. M. 2003. Nuclear receptors as drug targets: new developments incoregulators, orphan receptors and major therapeutic areas. Expert Opin. Ther.Targets 7: 679–684.

74. Szanto, A., and T. Roszer. 2008. Nuclear receptors in macrophages: a link be-tween metabolism and inflammation. FEBS Lett. 582: 106–116.

75. Grzegorzewicz, A. E., J. Kordulakova, V. Jones, S. E. Born, J. M. Belardinelli,A. Vaquie, V. A. Gundi, J. Madacki, N. Slama, F. Laval, et al. 2012. A commonmechanism of inhibition of the Mycobacterium tuberculosis mycolic acid bio-synthetic pathway by isoxyl and thiacetazone. J. Biol. Chem. 287: 38434–38441.

76. Hartkoorn, R. C., C. Sala, J. Neres, F. Pojer, S. Magnet, R. Mukherjee, S. Uplekar,S. Boy-Rottger, K. H. Altmann, and S. T. Cole. 2012. Towards a new tuberculosisdrug: pyridomycin - nature’s isoniazid. EMBO Mol. Med. 4: 1032–1042.

The Journal of Immunology 305

by guest on February 10, 2018http://w

ww

.jimm

unol.org/D

ownloaded from