Published Ahead of Print 20 July 2012. 10.1128/AEM.01674-12.

2012, 78(19):6829. DOI:Appl. Environ. Microbiol. J. Hansen, Tawanda Gumbo and Nicolai S. C. van OersShashikant Srivastava, Kole T. Roybal, José A. Aínsa, Eric Jennifer L. Eitson, Jennifer J. Medeiros, Ashley R. Hoover, MacrophagesHigh-Level Protein Expression in Infected Mycobacterial Shuttle Vectors Designed for

http://aem.asm.org/content/78/19/6829Updated information and services can be found at:

These include:

SUPPLEMENTAL MATERIAL Supplemental material

REFERENCEShttp://aem.asm.org/content/78/19/6829#ref-list-1at:

This article cites 45 articles, 18 of which can be accessed free

CONTENT ALERTS more»articles cite this article),

Receive: RSS Feeds, eTOCs, free email alerts (when new

http://journals.asm.org/site/misc/reprints.xhtmlInformation about commercial reprint orders: http://journals.asm.org/site/subscriptions/To subscribe to to another ASM Journal go to:

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

Mycobacterial Shuttle Vectors Designed for High-Level ProteinExpression in Infected Macrophages

Jennifer L. Eitson,a Jennifer J. Medeiros,a Ashley R. Hoover,a Shashikant Srivastava,b,c Kole T. Roybal,a José A. Aínsa,d Eric J. Hansen,e

Tawanda Gumbo,b,c and Nicolai S. C. van Oersa,e,f

The Departments of Immunologya and Medicine,b Office of Global Health,c and Departments of Microbiologye and Pediatrics,f The University of Texas SouthwesternMedical Center, Dallas, Texas, USA; and Departamento de Microbiología, Facultad de Medicina, Universidad de Zaragoza, Zaragoza, Spaind

Mycobacterial shuttle vectors contain dual origins of replication for growth in both Escherichia coli and mycobacteria. One suchvector, pSUM36, was re-engineered for high-level protein expression in diverse bacterial species. The modified vector (pSUM-kan-MCS2) enabled green fluorescent protein expression in E. coli, Mycobacterium smegmatis, and M. avium at levels up to 50-fold higher than that detected with the parental vector, which was originally developed with a lacZ� promoter. This high-levelfluorescent protein expression allowed easy visualization of M. smegmatis and M. avium in infected macrophages. The M. tuber-culosis gene esat-6 was cloned in place of the green fluorescence protein gene (gfp) to determine the impact of ESAT-6 on the in-nate inflammatory response. The modified vector (pSUM-kan-MCS2) yielded high levels of ESAT-6 expression in M. smegmatis.The ability of ESAT-6 to suppress innate inflammatory pathways was assayed with a novel macrophage reporter cell line, de-signed with an interleukin-6 (IL-6) promoter-driven GFP cassette. This stable cell line fluoresces in response to diverse mycobac-terial strains and stimuli, such as lipopolysaccharide. M. smegmatis clones expressing high levels of ESAT-6 failed to attenuateIL-6-driven GFP expression. Pure ESAT-6, produced in E. coli, was insufficient to suppress a strong inflammatory response elic-ited by M. smegmatis or lipopolysaccharide, with ESAT-6 itself directly activating the IL-6 pathway. In summary, a pSUM-pro-tein expression vector and a mammalian IL-6 reporter cell line provide new tools for understanding the pathogenic mechanismsdeployed by various mycobacterial species.

Mycobacterium tuberculosis has killed millions of people sincethe beginning of recorded human history (9, 14, 15). Every

year, 12 to 14 million people suffer from active disease, and ofthese 2 million die. For those apparently cured from such syn-dromes as tuberculous (TB) meningitis, more than 60% are likelyto die within 5 years, and close to half of those who survive willdevelop permanent medical problems (32, 38). While therapy forTB has been lifesaving, preventing infections and disease progres-sion would be much more effective. M. avium and M. leprae aretwo additional mycobacterial species of concern for the humanpopulation, with immunocompromised patients being very sus-ceptible to M. avium. There are no broadly effective vaccines formycobacterial infections. An attenuated derivative of M. bovis (thebacillus Calmette-Guérin strain, or BCG) is the most commonlyused vaccine for TB. However, BCG has limited effectiveness innewborns and children, is not protective in adults, and thus is notcurrently recommended in the United States (3). Indeed, BCGdoes not prevent the reactivation of latent TB to clinical disease. Ina mouse study, BCG vaccination enhanced the virulence of themost menacing of M. tuberculosis genotypes in terms of drug re-sistance and transmission, the Beijing strain (26). In a separateguinea pig model, multiple sequential vaccinations with BCG re-duced survival following subsequent challenges with virulent M.tuberculosis (5). With 11 new recombinant vaccines, all but onedepends on boosting with BCG (1, 8, 23). A promising alternativecandidate has emerged with a recombinant M. smegmatis (43).Even with such advances, it is imperative that better tools to studythe pathogenic mechanisms of M. tuberculosis be developed, espe-cially because M. tuberculosis has a large, complex genome com-prised of �4.4 million base pairs with approximately 4,000 openreading frames (12). About 1,500 M. tuberculosis genes have noknown functions.

During infection, M. tuberculosis can exhibit both an activephase and a long-term, poorly understood latency that can lastseveral decades in infected individuals. The latency likely involvesthe suppression of innate inflammatory pathways that are moredefined early in infection. For example, avirulent strains of myco-bacteria induce higher levels of inflammatory cytokines (tumornecrosis factor [TNF], nitric oxide, and interleukin-6 [IL-6]) ininfected macrophages than do pathogenic strains (7). This ispartly due to several M. tuberculosis genes that inhibit innate in-flammatory pathways (12, 42). The best understood involves theESX-1 type VII secretion system, a protein transport system thatreleases multiple proteins across the cell envelope (20, 21, 34, 42).One substrate of this pathway is EspR, a secreted transcriptionfactor that controls M. tuberculosis virulence, partly by function-ing in a negative feedback loop (34). Two additional substratesinclude early-secreted antigenic target protein 6 (ESAT-6) andCFP-10, which exist in a 1:1 heterodimer complex (21, 36).ESAT-6 attenuates Toll-like receptor signaling pathways, result-ing in diminished production of inflammatory cytokines, includ-ing IL-12, IL-6, and TNF (33). The inhibition occurs following adirect binding of ESAT-6 to Toll-like receptor 2 (TLR2), whichsubsequently antagonizes TLR2-, TLR4-, and TLR9-mediated re-

Received 25 May 2012 Accepted 25 June 2012

Published ahead of print 20 July 2012

Address correspondence to Nicolai S. C. van Oers,nicolai.vanoers@utsouthwestern.edu.

Supplemental material for this article may be found at http://aem.asm.org/.

Copyright © 2012, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.01674-12

October 2012 Volume 78 Number 19 Applied and Environmental Microbiology p. 6829–6837 aem.asm.org 6829

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

lease of inflammatory cytokines (33). However, ESAT-6 has alsobeen reported to enhance cytokine production (11). Moreover,ESAT-6 has additional proposed functions, including a lysin andan inflammasome activator (10, 17, 22, 30). Using M. marinuminfections in zebra fish larvae, ESAT-6 enhances mycobacterialspread early in infections by elevating matrix metalloproteinase 9(MMP9) expression in epithelial tissues, which functions as amacrophage chemoattractant (16, 42, 44). Thus, while it is clearthat ESAT-6 functions in early infection stages when M. tubercu-losis is actively replicating, its mechanism of action remains com-plicated.

The objectives of the current study were to develop mycobac-terial vectors that provide high-level protein expression and de-velop novel mammalian reporter cell lines for studying mycobac-terial infections.

MATERIALS AND METHODSBacterial strains, plasmids, and antibodies. The parental pSUM plas-mids were originally developed and described elsewhere (2). pSUM36 wasused in the current studies, with transformations done into chemicallycompetent E. coli (XL1-Blue or HB101) or electrocompetent M. smegma-tis or M. avium (subsp. homimissuis) (ATCC 700084 and 700898, respec-tively; American Type Culture Collection, Manassas, VA). E. coli cellswere grown in Luria-Bertani (LB) broth (Miller’s LB broth; Thermo-Fisher) or plated on LB agar. M. smegmatis and M. avium cells were grownin Difco Middlebrook 7H9 broth supplemented with oleic acid-albumin-dextrose-catalase (OADC) and Tween 80 at a 0.5% (vol/vol) final concen-tration. The cells were plated on Middlebrook 7H10 agar plates supple-mented with 10% OADC. Kanamycin was added at a final concentrationof 10 and 50 �g/ml for M. smegmatis and M. avium, respectively. GenomicDNA from M. tuberculosis H37Rv was obtained from Colorado State Uni-versity under a TB Vaccine Testing and Research Materials Contract withthe National Institutes of Health (N01-AI-40091). The plasmid pMRLB7,obtained under the same contract, contains histidine (His)-taggedESAT-6. The green fluorescent protein (GFP) gene was originally modi-fied for expression in prokaryotes, and this version included an optimizedtranslational initiation region from the plasmid pGreenTIR as describedearlier (29). The GFP gene was cloned from pGreenTIR into the HindIIIsite of the pSUM36 vector (pSUM-lacZ-gfp) (2). The groEL promoter wasamplified from M. tuberculosis genomic DNA using standard PCRs. ABamHI restriction site was included at the 5= end. This fragment wascloned into the pCR2.1 TOPO-TA cloning vector and subsequently iso-lated as a BamHI/XbaI piece that was subcloned into pSUM36 cut withBamHI and NheI to remove the lacZ� promoter. The GFP gene was sub-cloned into the HindIII site of this new pSUM-groEL vector. The HindIIIfragment containing gfp was also cloned in the reverse orientation inpSUM36, resulting in it being expressed off the 3= end of the kanamycincassette (pSUM-kan-MCS1-gfp). This vector was subsequently modifiedto reorient the BamHI and HindIII restriction sites to facilitate directionalcloning and introduce the Shine-Dalgarno sequence in the correct orien-tation (pSUM-kan-MCS2). esat-6, without a His tag, was subcloned intothe HindIII site of this new vector, enabling expression of ESAT-6 to bedriven by the kanamycin cassette (pSUM-kan-MCS2-esat-6). The variouspSUM plasmids were electroporated into chemically competent E. coli orelectrocompetent M. smegmatis and M. avium. The resulting clones weremonitored for GFP expression by flow cytometry on a FACSCalibur (Bec-ton, Dickinson, San Jose, CA). Data were analyzed using FloJo software(Tree Star, Inc., Ashland, OR). When using M. avium cultures, both themycobacterial and mammalian cells were fixed with 4% paraformalde-hyde prior to imaging studies. Bacterial cell extracts were prepared bybead beating, and the protein component was precipitated with the addi-tion of 10% (vol/vol; final concentration) trichloroacetic acid as detailedelsewhere (13). Total protein or immune-precipitated proteins were re-solved on low-molecular-mass protein separation gels as described previ-

ously (37). A monoclonal antibody directed against ESAT-6 was used forWestern immunoblotting experiments according to the manufacturers’instructions (HYB 076-08; Abcam, Cambridge, MA). In certain experi-ments, ESAT-6 was immunoprecipitated from 50 ml of M. smegmatisculture supernatants. After 3 days of culture, the mycobacteria were pel-leted by centrifugation and the supernatant sterilized by filtration(0.22-�m filter). The filtered supernatant was mixed with 50 �l of proteinA gel slurry in the presence of 6 �l of anti-ESAT-6 polyclonal antisera(equal mix of antisera from Pierce [catalog no. PA1-19446] and Cedarlane[catalog no. CLX305AP]; Thermo-Fisher, Inc.; Cedarlane Labs, Burling-ton, NC). After 4 to 5 h of mixing, the antibody (Ab)-protein A beads werecentrifuged at 2,000 rpm for 10 min, followed by 4 washes in 20 mMTris-Cl, pH 7.60, containing 0.15 M NaCl and 1% Triton X-100. Theremaining pellet was boiled in SDS sample buffer and the material re-solved on low-molecular-weight gels.

Recombinant ESAT-6 purification. ESAT-6(His6) was purified as de-scribed previously (40), with the following modifications. Following son-ication, the cleared lysate was applied to a HiTrap immobilized metalaffinity column (GE Biosciences). Prior to elution in 1 M imidazole, thecolumn was washed with 0.5% 3-[N,N-dimethyl(3-myristoylamino-propyl) ammonio] propanesulfonate amidosulfobetaine-14 (ASB-14)and reconstituted in 10 mM Tris-Cl, pH 7.9. This zwitterionic detergent isused to remove contaminating lipopolysaccharide.

Macrophage reporter cell lines. The J774A.1 murine macrophage cellline was obtained from the American Type Culture Collection. These cellswere maintained in Dulbecco’s modified Eagle medium (DMEM) supple-mented with 10% (vol/vol) fetal calf serum, 5 � 10�5 M 2-mercaptoeth-anol, penicillin (100 U/ml final concentration), streptomycin (100 �g/mlfinal concentration), and 10 mM glutathione. The macrophages wereplated at 3 � 105 cells/ml 1 day prior to infection. M. smegmatis or M.avium clones containing the lacZ�-, groEL-, and kan-driven GFP plas-mids, as well as the parental vector, pSUM36, were used to infect theJ774A.1 macrophage cells at various multiplicities of infection (MOI).After 3 h of infection, the cells were washed with DMEM containing 10%(vol/vol) fetal calf serum and 0.05% (vol/vol) Tween 80. Cells were incu-bated with amikacin for 45 min to limit the extracellular growth of themycobacteria. GFP expression was monitored at 24 h postinfection byfluorescence microscopy. The murine Il-6 promoter was amplified fromgenomic DNA (extracted from tail biopsy specimens from C57BL/6 mice)with standard PCRs (4). AseI and AgeI sites were included at the 5= and 3=ends, respectively. The resulting fragment was subcloned into an AseI/AgeI-digested pEGFP-N1 plasmid (Clontech, BD Biosciences, Inc.). Thenew pmIl-6-gfp plasmid was electroporated into J774A.1 macrophages.Individual clones were isolated by G418 selection, with one clone(J774A.1-Il-6-gfp #3) being used for the studies described here. This clonewas maintained in the aforementioned DMEM supplemented with 1mg/ml of G418. These cells were infected with M. smegmatis at MOIs of30:1, 10:1, and 3:1. IL-6 activation was determined 24 h postinfection byfluorescence microscopy. Alternatively, IL-6 activation was monitored byGFP expression using flow cytometry on a FACSCalibur instrument.

Microscopy imaging. A Nikon Eclipse T5100 microscope was used forlight and fluorescence microscopy. Confocal microscopy was done with apDV Deltavision deconvolution microscope equipped with a 100� (1.4numeric aperture; APO) Olympus lens in Cool Snap HQ2 camera withfluorescein isothiocyanate (FITC)/CY5 filter sets for all fluorescence mi-croscopy experiments. A single differential interference contrast (DIC)reference image and fluorescent z-stacks (0.2-�m z-step) were taken foreach cell. Images were background subtracted, and fluorescent channeloverlays were performed in ImageJ (http://rsbweb.nih.gov/ij/). Represen-tative fluorescent z-plane images and movies of the entire z-volume ofeach cell are provided in the supplemental material.

Statistical analyses. For flow cytometry comparisons between differ-ent bacterial clones and strains, an unpaired Student’s t test was used fordetermining the significance. Statistical significance was indicated by one(*, P � 0.005), two (**, P � 0.0005), or three asterisks (***, P � 0.0001).

Eitson et al.

6830 aem.asm.org Applied and Environmental Microbiology

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

RESULTSDevelopment of high-level protein expression shuttle vectors.The pSUM plasmids are a family of shuttle vectors designed forreplication in both E. coli and mycobacteria (2). The vectors con-tain multiple cloning sites, lacZ� for recombinant screening, and akanamycin resistance gene, kan. These vectors have been usefulfor screening genomic libraries (35). We wanted to determinewhether such vectors were suitable for high-level protein expres-sion in both E. coli and distinct mycobacterial strains. The geneencoding a modified GFP, under the control of an optimizedtranslational initiation region, was subcloned into the parentalpSUM36 vector under the control of the lacZ� promoter (pSUM-lacZ-gfp) (Fig. 1A) (29). To determine whether different promotercassettes could increase GFP expression, the groEL promoter,cloned from M. tuberculosis, was substituted for lacZ� (pSUM-groEL-gfp). A third expression vector was developed in which thekanamycin cassette, already part of the original pSUM vector, wasused to express GFP (pSUM-kan-MCS1-gfp). The parental plas-mid, pSUM36, and each of the gfp expression vectors were trans-formed into E. coli. Transformants were grown overnight, and thelevels of GFP expressed by each were determined by flow cytom-etry. E. coli cells containing plasmids with the lacZ-(pSUM-lacZ),groEL-(pSUM-groEL), and kan-(pSUM-kan-MCS1) promotershad statistically significant 3-, 3-, and 9-fold increases in fluores-cence, respectively (mean fluorescence intensity [MFI]), com-pared to control plasmid transformants (P � 0.0005) (Fig. 1B).The overexpression of GFP at different levels had no effect on E.coli growth, as similar CFU/ml were obtained with control andhigh-GFP-expressing clones (Fig. 1C). The same plasmids weresubsequently transformed into M. smegmatis, and the fluores-cence intensity of the mycobacterial clones was compared after 3days of culture (Fig. 1D). There was a 7-, 9-, and 190-fold increasein the MFI, respectively (P � 0.0005) (Fig. 1D). There was noimpact of these high-level protein expression cassettes on the my-cobacterial growth characteristics (Fig. 1E). These results demon-strate that the kanamycin promoter/gene cassette was the mosteffective for GFP expression in both E. coli and M. smegmatis, withthe levels detected in the mycobacterial species being 21-foldhigher than those of standard E. coli strains (Fig. 1B and D). Giventhe success of the kanamycin cassette, the parental pSUM36 vectorwas redesigned to facilitate directional cloning. The HindIII andBamHI restriction sites were flipped, and the optimized transla-tion initiation site was cloned at the 3= end of the kanamycincassette (pSUM-kan-MCS2). While this new cloning vector re-sulted in GFP expression in E. coli equivalent to that with thepSUM-kan-MCS1 plasmid, it provided a 2-fold higher level offluorescence in M. smegmatis than that achieved with pSUM-kan-MCS1 (Fig. 1B and D). The experiments were repeated in M.avium. A 500-fold increase in fluorescence was detected over thecontrol vector-transformed M. avium clones, with limited impacton mycobacterial growth (Fig. 1F and G).

GFP-expressing mycobacteria are detected in infected mac-rophages. Previous studies have shown that mycobacterial strainsexpressing GFP can be visualized in infected macrophages (18,25). To determine whether the pSUM-modified vectors were suit-able for detecting mycobacteria in cell lines, macrophages wereinfected at an MOI of 30:1 and the fluorescence levels assayed 24 hlater with fluorescence microscopy. M. smegmatis containing theparental vector could not be detected in the infected cells

(Fig. 2A). The mycobacterial subclones expressing GFP from thelacZ� and groEL promoters were evident as small, faint green ba-cilli in the macrophages (Fig. 2A). M. smegmatis clones expressingGFP under the control of the kanamycin cassette (pSUM-kan-MCS1-gfp) yielded highly visible mycobacteria within the macro-phages. The pSUM-kan-MCS2-gfp construct provided the mostfluorescently intense mycobacteria. To determine whether the dif-ferent levels of GFP expression affected virulence and host cellsurvival, the cells were incubated with a LIVE/DEAD blue fluores-cent reactive dye, a dye that permeates compromised cellularmembranes, resulting in an intense fluorescent stain. Macro-phages cultured in the presence of 0.1% NaN3 exhibited an intenseblue fluorescence under UV light, consistent with the loss of mem-brane integrity. In contrast, the macrophages exhibited a normalviability, even after infection with the different mycobacterialclones, independent of GFP expression (Fig. 2A, middle). Bright-field images revealed evenly distributed macrophage monolayersthat were unaffected by the type of expression vector used (Fig. 2A,lower). These experiments indicated that the pSUM vectors withthe kanamycin cassette coupled to an optimized translational ini-tiation site were ideal for achieving high-level protein expressionin infected cells. To determine if these vectors were suitable forother mycobacterial species, M. avium was transfected withpSUM-kan-MCS2-gfp. The resulting clones were used to infectmacrophages at MOIs of 3:1, 10:1, and 30:1 (Fig. 2B). As with M.smegmatis, the M. avium-GFP clones were highly visible in theinfected macrophages (Fig. 2B, upper). The infected macrophagesexhibited similar viabilities at the different MOIs, as determinedby LIVE/DEAD staining and bright-field images (Fig. 2B, middleand lower). M. marinum clones expressing GFP have a reducedvirulence in zebra fish infection models (31). To determine if theM. smegmatis GFP-overexpressing clones were attenuated in themacrophages, the CFU/ml were calculated after lysing the infectedmacrophages. While there were variations in the CFU/ml calcu-lated for the different M. smegmatis clones, we did not detect sta-tistically significant differences between the clones expressing lowand high levels of GFP (Fig. 2C). Given our ability to achievehigh-level GFP expression in mycobacteria, we next assessedwhether these would be suitable for detecting the location of my-cobacteria in infected macrophages. Confocal microscopy analy-ses of the infected macrophages revealed distinct mycobacterialmorphologies between M. smegmatis and M. avium. M. smegmatiscells were seen as extended filaments that overlapped with thephagolysosome, as detected by LAMP staining. This pattern wasdistinct from the M. avium cells, which appeared clustered in dis-crete, rounded foci that again were in proximity to lysosomalgranules (LAMP) (Fig. 3A versus B). The differences between M.avium and M. smegmatis were clearly revealed with multiple z-plane stacks (see the movies in the supplemental material).

High-level ESAT-6 expression is insufficient to suppress in-nate inflammatory pathways in macrophages. To further exam-ine the usefulness of the modified pSUM vectors for studyinghigh-level protein expression, the gfp gene was replaced withesat-6 (Fig. 1A). ESAT-6 is normally secreted via the ESX-1 secre-tion system and is proposed to inhibit Toll-like receptor signaling(33, 34, 42). To compare the expression levels of ESAT-6, extractswere prepared from E. coli and M. smegmatis clones containingeither the parental vector or the pSUM-kan-MCS2-esat-6 expres-sion vector (Fig. 4A). M. smegmatis containing esat-6 was grown in7H9 medium under culture conditions that prevent ESAT-6 se-

Mycobacterial Expression Vectors for Infection Studies

October 2012 Volume 78 Number 19 aem.asm.org 6831

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

FIG 1 Mycobacterial shuttle vectors with distinct promoter cassettes provide high-level protein expression. (A) The original pSUM36 vector and its modifiedderivatives are shown. In the parental vector, gfp was cloned downstream of the lacZ� promoter (pSUM-lacZ-gfp). In a second construct, the groEL promoterreplaced lacZ� (pSUM-groEL-gfp). The gfp construct was also introduced in an opposite orientation, leading to its regulation by the existing kanamycin cassette(pSUM-kan-MCS1-gfp). This version of the vector was modified to create a vector with HindIII and BamHI cloning sites (pSUM-kan-MCS2). This vector wasalso used for expressing esat-6. Restriction sites included the following: HindIII, H; PstI, P; SalI, S; XbaI, X; BamHI, B; SmaI, Sm; SacI, Sa; EcoRI, E; KpnI, K; andNdeI, N. (B) GFP expression in E. coli clones transformed with the indicated plasmids was monitored by flow cytometry. The average mean fluorescenceintensities of 5 distinct E. coli clones were plotted. (C) The growth characteristics of the bacteria were determined by measuring the CFU/ml after overnightculture. (D) GFP expression in the various M. smegmatis transformants is shown. The mean fluorescence intensities of 5 distinct M. smegmatis clones is plotted.(E) The growth of the mycobacterial clones was determined by measuring the CFU/ml after 3 days in culture. (F) The fluorescence levels of M. avium clonescontaining either the parental vector or the GFP-expressing cassette are shown, with the average mean fluorescence intensities of 3 distinct M. avium clonesplotted. (G) The CFU/ml were determined from the M. avium cultures after 7 days of growth. Statistically significant differences were calculated using standardpaired t tests. *, P � 0.005; **, P � 0.0005; ***, P � 0.0001.

Eitson et al.

6832 aem.asm.org Applied and Environmental Microbiology

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

cretion that would enable us to detect the protein in mycobacteriallysates (13). Western immunoblotting confirmed that ESAT-6was specifically expressed at high levels in the M. smegmatis clonescontaining the enhanced protein expression plasmid (Fig. 4A,lanes 6 and 7). ESAT-6 was almost undetectable in the E. coliextracts, consistent with the lower levels of GFP expression in E.

coli versus the mycobacterial strains (Fig. 4A, lanes 3 and 4). Eventhough M. smegmatis was grown in medium that does not allowESAT-6 secretion, it was unknown whether ESAT-6 would be re-leased with the high-level protein expression vector (13). To de-termine whether ESAT-6 was released into the medium, M. smeg-matis cultures were harvested after 3 days of culture. A polyclonal

FIG 2 M. smegmatis and M. avium clones expressing high levels of GFP are easily visualized in infected macrophages. (A) M. smegmatis transformants containingdifferent GFP expression plasmids were used to infect macrophages at a multiplicity of infection (MOI) of 30:1. (Upper) The macrophages were washed toremove extracellular mycobacteria, cultured for 24 h, and analyzed by fluorescence microscopy. The M. smegmatis clones included the pSUM36 parent vector,pSUM-lacZ-gfp, pSUM-groEL-gfp, pSUM-kan-MCS1-gfp, and pSUM-kan-MCS2-gfp, as indicated. In the middle panel, the fluorescence intensity of the cellsshown in the upper panel was monitored using LIVE/DEAD cell staining. A dark blue stain revealed dead/dying cells. This was evident with the dark blue patternsseen in cells treated with NaN3. (Lower) The macrophage cells were detected with visible light. Magnification, �40. (B) M. avium transformants containingpSUM-kan-MCS2-gfp were used to infect macrophages at MOIs of 3:1, 10:1, and 30:1. The upper panel illustrates the fluorescently labeled mycobacteria. Thefluorescent intensity of the cells was monitored using LIVE/DEAD cell staining in the middle panel. In the lower panel, the macrophage cells were detected withvisible light. (C) The viability of the M. smegmatis clones containing the pSUM-empty vector or the indicated GFP expression cassettes was determined bymeasuring the CFU/ml 24 h after infection in the macrophages.

Mycobacterial Expression Vectors for Infection Studies

October 2012 Volume 78 Number 19 aem.asm.org 6833

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

antisera was used to precipitate ESAT-6 from the clarified andfiltered culture supernatant. This polyclonal antisera precipitatedan exogenous source of pure ESAT-6, produced in E. coli, that wasadded directly into 1.0 and 50 ml of lysis buffer, as determined bysubsequent Western immunoblotting (Fig. 4B, lanes 1 to 2). Incontrast, no endogenously produced ESAT-6 was precipitatedfrom the clarified M. smegmatis culture supernatants (Fig. 4B, lane3). To confirm that the culture supernatant did not contain aprotease activity that might have degraded ESAT-6, a separateimmunoprecipitation was performed in which pure ESAT-6 wasspiked into the culture supernatant. This exogenously addedESAT-6 was detected in the precipitate (Fig. 4B, lane 4). Takentogether, these findings indicate that the modified pSUM vectorsprovided for high-level protein expression in diverse mycobacte-rial species.

Development of an IL-6-driven GFP macrophage reportercell line responsive to inflammatory stimuli. Macrophages se-crete inflammatory cytokines, such as IL-1, IL-6, IL-12, andTNF-�, in response to bacterial infections and/or bacterial prod-ucts such as lipopolysaccharide (LPS). To monitor innate cell ac-tivation, a murine Il-6 promoter (600 bp) was cloned upstream ofGFP, replacing the standard cytomegalovirus promoter (Fig. 4C)(4). This construct was used to generate stable macrophage celllines (J774A.1) expressing GFP under the control of the IL-6 pro-moter. One subclone (J774A.1-Il-6-gfp #3) exhibited a �7-fold

FIG 3 M. smegmatis and M. avium clones expressing high levels of GFP exhibitdistinct morphologies in infected macrophages. (A) M. smegmatis transformantscontaining different GFP expression plasmids were used to infect macrophages ata multiplicity of infection (MOI) of 30:1. The macrophages were washed to re-move extracellular mycobacteria and then cultured for 24 h. The cells were fixedand stained with antibodies that detect lysosomal granules (LAMP staining). Thecells were analyzed by confocal microscopy, with green detecting GFP expressionand red identifying lysosomal granules. (B) M. avium transformants containingpSUM-kan-MCS2-gfp were used to infect macrophages at MOIs of 30:1. Confocalmicroscopy was used to image the cells, with green detecting GFP expression andred identifying lysosomal granules. The bar represents a 5 �M length.

FIG 4 Mammalian IL-6 reporter cell line responsive to innate inflammatory stimuli. (A) ESAT-6 protein expression in E. coli (lanes 2 to 4) and M. smegmatis(lanes 5 to 7) transformed with pSUM-empty vector (lanes 2 and 5) or pSUM-kan-MCS2-esat-6 (lanes 3, 4, 6, and 7) was determined by Western immunoblot-ting, with pure ESAT-6 used as a positive control (lane 1). (B) ESAT-6 produced in mycobacteria with a high-level protein expression vector is not secreted.Culture supernatant from an ESAT-6-expressing M. smegmatis clone was clarified and filtered. ESAT-6 was immunoprecipitated from the culture supernatantusing polyclonal antisera directed against ESAT-6. Pure ESAT-6 was used as a positive control (lanes 1 and 2), with ESAT-6 being more dilute in lane 2 (50 versus1 ml). ESAT-6 was immunoprecipated from 50 ml of culture supernatant of an M. smegmatis clone overexpressing ESAT-6 intracellularly (lane 3). As a positivecontrol, pure ESAT-6 was added to 50 ml of culture supernatant with the same M. smegmatis clone overexpressing ESAT-6 intracellularly, and this wasprecipitated with anti-ESAT-6 antisera (lane 4). The immunoprecipitates were washed and processed for immunoblotting with the MAb. (C) The murine IL-6promoter, with the indicated transcription factor binding sites, was cloned upstream of a mammalian eGFP expression vector. (D) The J774A.1-Il-6-gfp reportercell line was cultured in media in the absence or presence of 2.5 ng/ml of LPS or infected with M. smegmatis at MOIs of 3:1, 10:1, and 30:1. Twenty-four hourspostinfection, the cells were imaged for fluorescence (upper row) and visible light with a 40� objective.

Eitson et al.

6834 aem.asm.org Applied and Environmental Microbiology

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

increase in GFP expression after the cells were cultured for 24 h inthe presence of LPS. A dose-response curve revealed that the cellswere sensitive to LPS concentrations as low as 1 ng/ml (data notshown). These J774A.1 Il-6 reporter cells were infected with M.smegmatis at MOIs of 3:1, 10:1, and 30:1 (Fig. 4D). GFP was de-tected in the macrophages 24 h postinfection at all three MOIs,with the optimal signal being detected at an MOI of 30:1. Suchexperiments established the usefulness of this reporter cell line tomonitor an innate inflammatory response upon exposure to my-cobacteria.

Previous work has shown that pathogenic mycobacteria cause

an attenuated inflammatory response compared to nonpatho-genic strains (7). This is partly attributed to the release of ESAT-6by strains such as M. tuberculosis (33). M. smegmatis clones thatexpressed high levels of ESAT-6 were tested for their ability toinhibit the IL-6 promoter-driven GFP expression in J774A.1 celllines. Macrophages were infected at MOIs of 3:1, 10:1, and 30:1with either the M. smegmatis parental cells or the clones expressingESAT-6. The levels of GFP, as measured by flow cytometry, wereindistinguishable at each MOI tested, regardless of whether themacrophages were infected with M. smegmatis clones lacking orcontaining the kan-MCS2-esat-6 expression vector (Fig. 5A).

FIG 5 ESAT-6 activates innate inflammatory responses in macrophages through the IL-6-driven inflammatory pathway. (A) M. smegmatis clones expressingesat-6 were used to infect the IL-6 macrophage reporter cell line at MOIs of 3:1, 10:1, and 30:1. The mean fluorescence intensities of the reporter cells weredetermined by flow cytometry and compared to those of cells infected with the M. smegmatis clones containing the parental vector (pSUM36) as well asuninfected cells. There was no significant difference in the mean fluorescence intensity between control and ESAT-6-expressing clones compared at each MOI.(B) The J774A.1-Il-6-gfp reporter cell line was cultured in media in the absence or presence of 2.5 ng/ml of LPS or infected with M. smegmatis at MOI of 30:1. PureESAT-6 (12 �g/ml) was added at the initiation of the infection and after each wash when the mycobacteria were used to infect the macrophages, as indicated. (C)The mean fluorescence intensity in the macrophages was determined from at least 3 independent infections. (D) ESAT-6 (12 �g/ml starting material) was seriallydiluted (1/3) and added to the J774A.1-Il-6-gfp reporter cell line at the indicated concentrations. After 24 h, the cells were analyzed by flow cytometry for meanfluorescence intensity, determined from 3 independent samples per ESAT-6 dilution. ***, P � 0.0001.

Mycobacterial Expression Vectors for Infection Studies

October 2012 Volume 78 Number 19 aem.asm.org 6835

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

These experiments suggested one of two possibilities. First,ESAT-6 needed to be secreted to suppress innate inflammatorypathways. Second, ESAT-6 does not inhibit Il-6 promoter activity.To address these possibilities, pure ESAT-6 was added to the Il-6-gfp macrophage reporter cell line, either alone or with LPS or M.smegmatis (Fig. 5B). Unexpectedly, pure ESAT-6 directly acti-vated the Il-6-gfp reporter cell line, as measured by the increasedfluorescence intensity (Fig. 5B, upper). Moreover, the activationof the IL-6 reporter cell line was stronger in the LPS-stimulatedcultures when ESAT-6 was supplemented. The macrophage cellline was clearly activated, as revealed by increased forward andside scatter (Fig. 5B, lower). In fact, the fluorescence levels werehigher in the presence of pure ESAT-6 than with LPS (Fig. 5C).The ability of ESAT-6 to induce this strong inflammatory re-sponse was reduced when combined with M. smegmatis, matchingthe GFP expression when the macrophages were exposed to themycobacteria. We speculate that this is due to ESAT-6 sequestra-tion by the mycobacteria. To examine if the activation of the IL-6pathway correlated with ESAT-6 levels, a dose-response assay wasundertaken. Starting at 12 �g/ml of pure ESAT-6, a serial dilutionindicated that levels as low as 50 ng/ml induced GFP expression byaround 25-fold (Fig. 5D). These experiments indicate that ESAT-6can stimulate certain components of the innate inflammatory re-sponse.

DISCUSSION

The genomic sequence of M. tuberculosis was completed in 1998,revealing a complex genetic organization with approximately4,000 open reading frames (12). However, relatively few geneproducts have been identified that directly attenuate innate im-mune cell functions (6, 19, 24, 27, 28, 33, 34, 39, 42, 45). Severaldistinct approaches have been used to identify putative virulencefactors, including transposon mutagenesis of M. tuberculosis, iso-lation and characterization of regions of difference (RD), andstudies of eukaryotic-like proteins encoded by the M. tuberculosisgenome. This has led to the identification of a few virulence fac-tors. However, none can fully explain the ability of M. tuberculosisto maintain a latent infection in infected humans without overtlyactivating the innate inflammatory response. For example, thesuppressive effects of ESAT-6, a well-known virulence factor, re-quire �g quantities in in vitro systems, and some studies havesuggested that ESAT-6 can increase inflammatory responses (11,30, 33). One limitation of these studies is a system amenable tohigh-level protein expression in infected cells. We re-engineeredthe mycobacterial shuttle vector pSUM36 to provide high-levelprotein expression in bacterial cultures and in infected macro-phages. By using the kanamycin cassette already present inpSUM36 and incorporating a Shine-Dalgarno sequence down-stream of this cassette, a 50-fold improvement in GFP expressionwas achieved compared to that obtained with the original lacZ�promoter. The pSUM-kan vector was suitable for expressing GFPand ESAT-6 in E. coli and the mycobacterial strains M. smegmatisand M. avium. The levels of protein expression were consistentlyhigher in the mycobacterial strains. This enabled us to readilydetect the GFP-expressing mycobacteria within macrophages. Aninteresting difference in bacterial growth/morphology was uncov-ered with high-resolution confocal microscopy. The M. aviumclones appeared as discrete foci within the macrophages. This wasin marked contrast to the M. smegmatis clones, which exhibited afilamentous pattern that resembled spider webs within the cell.

The basis for these differences is under investigation. The highlevels of GFP expressed by the mycobacteria did not attenuate thegrowth or virulence of the bacteria. This is in contrast to a recentreport that M. marinum virulence is reduced in a zebra fish infec-tion model when GFP is expressed to high levels (31). The differ-ences may relate to our use of macrophages in in vitro assays com-pared to the in vivo infections in the latter study.

A second useful tool developed in the current study was a mac-rophage reporter cell line responsive to bacterial lipopolysaccha-ride and M. smegmatis infections. This stable cell line provides a40-fold range of GFP expression in every cell infected. Anotherreport described the use of immunoreporter macrophage cell lines(6). In that case, the macrophages only provided transient re-porter expression that had a wide fluorescence distribution. Thecurrent cell line provides sensitive, uniform GFP expression in allcells in response to mycobacterial infections. We tested the M.smegmatis clones overexpressing ESAT-6 to determine howESAT-6 suppresses innate inflammatory pathways. Importantly,this intracellular form of ESAT-6 did not suppress an innate in-flammatory response, as measured via the activation of the IL-6promoter through either LPS or M. smegmatis. Since a previousreport suggested that ESAT-6 must be released in a secreted formto attenuate Toll-like receptor signaling, we tested pure ESAT-6 inour IL-6 reporter assay (33). ESAT-6 was purified using estab-lished procedures that eliminate contaminating LPS. Surprisingly,ESAT-6 strongly activated the IL-6-dependent inflammatorypathway, as indicated by a 40-fold increase in GFP expression. Ourdata are consistent with another published report that ESAT-6induces a strong IL-6 and tumor growth factor response (11). Themechanism by which ESAT-6 attenuates Toll-like receptor signal-ing, as reported elsewhere, requires further clarification but is notsupported by our data. It is interesting that the combination ofpure ESAT-6 with M. smegmatis caused GFP expression that waslower than that with just ESAT-6 alone. This could be due toESAT-6 sequestration by M. smegmatis, although further experi-ments are required to address this possibility. In summary, ournew protein expression vectors and reporter cell line provide ad-ditional tools for uncovering the pathogenic mechanisms of di-verse mycobacterial strains. For example, this modified vectorprovides investigators with an excellent tool for testing differentdrugs for the eradication of M. avium and/or M. tuberculosis frominfected cells (41). In addition, such a vector could enhance immuneresponses to specific mycobacterial proteins, which may further im-prove the efficacy of the recently developed M. smegmatis vaccines(43).

ACKNOWLEDGMENTS

We thank Maria Labandeira-Rey and Dana Dodd for their helpful sugges-tions, critiques, and reagents. We also thank Angela Mobley for assistancewith flow cytometry using mycobacterial species. We appreciate the helpof Amie Torres, Sarah Gonzales, and Sara Ireland in testing the mycobac-terial plasmids. The ESAT clones were obtained from Colorado State Uni-versity and are currently provided by BEI Resources.

This work was supported in part by High Risk/High Impact grantsfrom the University of Texas Southwestern Medical Center to N.S.C.V.O.and National Institutes of Health/National Institute of General MedicalSciences Director New Innovator Award 1 DP2 OD001886 to T.G.

REFERENCES1. Aagaard C, et al. 2011. A multistage tuberculosis vaccine that confers

efficient protection before and after exposure. Nat. Med. 17:189 –194.

Eitson et al.

6836 aem.asm.org Applied and Environmental Microbiology

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from

2. Ainsa JA, Martin C, Cabeza M, De la Cruz F, Mendiola MV. 1996.Construction of a family of Mycobacterium/Escherichia coli shuttle vec-tors derived from pAL5000 and pACYC184: their use for cloning an anti-biotic-resistance gene from Mycobacterium fortuitum. Gene 176:23–26.

3. Andersen P, Doherty TM. 2005. The success and failure of BCG—implications for a novel tuberculosis vaccine. Nat. Rev. Microbiol. 3:656 –662.

4. Baccam M, Woo SY, Vinson C, Bishop GA. 2003. CD40-mediatedtranscriptional regulation of the IL-6 gene in B lymphocytes: involvementof NF-kappa B, AP-1, and C/EBP. J. Immunol. 170:3099 –3108.

5. Basaraba RJ, Izzo AA, Brandt L, Orme IM. 2006. Decreased survival ofguinea pigs infected with Mycobacterium tuberculosis after multiple BCGvaccinations. Vaccine 24:280 –286.

6. Beaulieu AM, et al. 2010. Genome-wide screen for Mycobacterium tu-berculosis genes that regulate host immunity. PLoS One 5:e15120. doi:10.1371/journal.pone.0015120.

7. Beltan E, Horgen L, Rastogi N. 2000. Secretion of cytokines by humanmacrophages upon infection by pathogenic and non-pathogenic myco-bacteria. Microb. Pathog. 28:313–318.

8. Bertholet S, et al. 2010. A defined tuberculosis vaccine candidate boostsBCG and protects against multidrug-resistant Mycobacterium tuberculo-sis. Sci. Transl. Med. 2:53ra74. doi:10.1126/scitranslmed.3001094.

9. Brosch R, et al. 2002. A new evolutionary scenario for the Mycobacteriumtuberculosis complex. Proc. Natl. Acad. Sci. U. S. A. 99:3684 –3689.

10. Carlsson F, et al. 2010. Host-detrimental role of Esx-1-mediated inflam-masome activation in mycobacterial infection. PLoS Pathog. 6:e1000895.doi:10.1371/journal.ppat.1000895.

11. Chatterjee S, et al. 2011. Early secreted antigen ESAT-6 of Mycobacte-rium tuberculosis promotes protective T helper 17 cell responses in atoll-like receptor-2-dependent manner. PLoS Pathog. 7:e1002378. doi:10.1371/journal.ppat.1002378.

12. Cole ST, et al. 1998. Deciphering the biology of Mycobacterium tuber-culosis from the complete genome sequence. Nature 393:537–544.

13. Converse SE, Cox JS. 2005. A protein secretion pathway critical forMycobacterium tuberculosis virulence is conserved and functional in My-cobacterium smegmatis. J. Bacteriol. 187:1238 –1245.

14. Corbett EL, et al. 2003. The growing burden of tuberculosis: global trendsand interactions with the HIV epidemic. Arch. Intern. Med. 163:1009 –1021.

15. Cosma CL, Sherman DR, Ramakrishnan L. 2003. The secret lives of thepathogenic mycobacteria. Annu. Rev. Microbiol. 57:641– 676.

16. Davis JM, Ramakrishnan L. 2009. The role of the granuloma in expan-sion and dissemination of early tuberculous infection. Cell 136:37– 49.

17. de Jonge MI, et al. 2007. ESAT-6 from Mycobacterium tuberculosisdissociates from its putative chaperone CFP-10 under acidic conditionsand exhibits membrane-lysing activity. J. Bacteriol. 189:6028 – 6034.

18. Dhandayuthapani S, et al. 1995. Green fluorescent protein as a markerfor gene expression and cell biology of mycobacterial interactions withmacrophages. Mol. Microbiol. 17:901–912.

19. Dussurget O, et al. 2001. Role of Mycobacterium tuberculosis copper-zinc superoxide dismutase. Infect. Immun. 69:529 –533.

20. Fortune SM, et al. 2005. Mutually dependent secretion of proteins re-quired for mycobacterial virulence. Proc. Natl. Acad. Sci. U. S. A. 102:10676 –10681.

21. Guinn KM, et al. 2004. Individual RD1-region genes are required forexport of ESAT-6/CFP-10 and for virulence of Mycobacterium tubercu-losis. Mol. Microbiol. 51:359 –370.

22. Hsu T, et al. 2003. The primary mechanism of attenuation of bacillusCalmette Guerin is a loss of secreted lytic function required for invasion oflung interstitial tissue. Proc. Natl. Acad. Sci. USA 100:12420 –12425.

23. Kaufmann SH. 2010. Future vaccination strategies against tuberculosis:thinking outside the Box. Immunity 33:567–577.

24. Koul A, et al. 2000. Cloning and characterization of secretory tyrosinephosphatases of Mycobacterium tuberculosis. J. Bacteriol. 182:5425–5432.

25. Kremer L, Baulard A, Estaquier J, Poulain-Godefroy O, Locht C. 1995.Green fluorescent protein as a new expression marker in mycobacteria.Mol. Microbiol. 17:913–922.

26. Lopez B, et al. 2003. A marked difference in pathogenesis and immuneresponse induced by different Mycobacterium tuberculosis genotypes.Clin. Exp. Immunol. 133:30 –37.

27. Malik ZA, et al. 2003. Cutting edge: Mycobacterium tuberculosis blocksCa2� signaling and phagosome maturation in human macrophages viaspecific inhibition of sphingosine kinase. J. Immunol. 170:2811–2815.

28. Manca C, Paul S, Barry CE, III, Freedman VH, Kaplan G. 1999.Mycobacterium tuberculosis catalase and peroxidase activities and resis-tance to oxidative killing in human monocytes in vitro. Infect. Immun.67:74 –79.

29. Miller WG, Lindow SE. 1997. An improved GFP cloning cassette de-signed for prokaryotic transcriptional fusions. Gene 191:149 –153.

30. Mishra BB, et al. 2010. Mycobacterium tuberculosis protein ESAT-6 is apotent activator of the NLRP3/ASC inflammasome. Cell. Microbiol. 12:1046 –1063.

31. Mutoji KN, Ennis DG. 2012. Expression of common fluorescent report-ers may modulate virulence for Mycobacterium marinum: dramatic at-tenuation results from Gfp over-expression. Comp. Biochem. Physiol. CToxicol. Pharmacol. 155:39 – 48.

32. Pasipanodya JG, et al. 2007. Pulmonary impairment after tuberculosis.Chest 131:1817–1824.

33. Pathak SK, et al. 2007. Direct extracellular interaction between the earlysecreted antigen ESAT-6 of Mycobacterium tuberculosis and TLR2 inhib-its TLR signaling in macrophages. Nat. Immunol. 8:610 – 618.

34. Raghavan S, Manzanillo P, Chan K, Dovey C, Cox JS. 2008. Secretedtranscription factor controls Mycobacterium tuberculosis virulence. Na-ture 454:717–721.

35. Ramon-Garcia S, Otal I, Martin C, Gomez-Lus R, Ainsa JA. 2006. Novelstreptomycin resistance gene from Mycobacterium fortuitum. Antimi-crob. Agents Chemother. 50:3920 –3922.

36. Renshaw PS, et al. 2002. Conclusive evidence that the major T-cell anti-gens of the Mycobacterium tuberculosis complex ESAT-6 and CFP-10form a tight, 1:1 complex and characterization of the structural propertiesof ESAT-6, CFP-10, and the ESAT-6-CFP-10 complex. Implications forpathogenesis and virulence. J. Biol. Chem. 277:21598 –21603.

37. Schagger H, von Jagow G. 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in therange from 1 to 100 kDa. Anal. Biochem. 166:368 –379.

38. Shaw JE, Pasipanodya JG, Gumbo T. 2010. Meningeal tuberculosis: highlong-term mortality despite standard therapy. Medicine 89:189 –195.

39. Singh R, et al. 2003. Disruption of mptpB impairs the ability of Myco-bacterium tuberculosis to survive in guinea pigs. Mol. Microbiol. 50:751–762.

40. Smith J, et al. 2008. Evidence for pore formation in host cell membranesby ESX-1-secreted ESAT-6 and its role in Mycobacterium marinum es-cape from the vacuole. Infect. Immun. 76:5478 –5487.

41. Srivastava S, et al. 2010. Efflux-pump-derived multiple drug resistance toethambutol monotherapy in Mycobacterium tuberculosis and the phar-macokinetics and pharmacodynamics of ethambutol. J. Infect. Dis. 201:1225–1231.

42. Stanley SA, Raghavan S, Hwang WW, Cox JS. 2003. Acute infection andmacrophage subversion by Mycobacterium tuberculosis require a special-ized secretion system. Proc. Natl. Acad. Sci. U. S. A. 100:13001–13006.

43. Sweeney KA, et al. 2011. A recombinant Mycobacterium smegmatisinduces potent bactericidal immunity against Mycobacterium tuberculo-sis. Nat. Med. 17:1261–1268.

44. Volkman HE, et al. 2010. Tuberculous granuloma induction via interac-tion of a bacterial secreted protein with host epithelium. Science 327:466 –469.

45. Walburger A, et al. 2004. Protein kinase G from pathogenic mycobacteriapromotes survival within macrophages. Science 304:1800 –1804.

Mycobacterial Expression Vectors for Infection Studies

October 2012 Volume 78 Number 19 aem.asm.org 6837

on October 3, 2012 by U

NIV

TE

XA

S S

W M

ED

CT

R 904

http://aem.asm

.org/D

ownloaded from