Modulation of Transcriptional and Inflammatory Responses in Murine Macrophages by the Mycobacterium tuberculosis Mammalian Cell Entry (Mce) 1 Complex

Abstract

The outcome of many infections depends on the initial interactions between agent and host. Aiming at elucidating the effect of the M. tuberculosis Mce1 protein complex on host transcriptional and immunological responses to infection with M. tuberculosis, RNA from murine macrophages at 15, 30, 60 min, 4 and 10 hrs post-infection with M. tuberculosis H37Rv or Δ-mce1 H37Rv was analyzed by whole-genome microarrays and RT-QPCR. Immunological responses were measured using a 23-plex cytokine assay. Compared to uninfected controls, 524 versus 64 genes were up-regulated by 15 min post H37Rv- and Δ-mce1 H37Rv-infection, respectively. By 15 min post-H37Rv infection, a decline of 17 cytokines combined with up-regulation of Ccl24 (26.5-fold), Clec4a2 (23.2-fold) and Pparγ (10.5-fold) indicated an anti-inflammatory response initiated by IL-13. Down-regulation of Il13ra1 combined with up-regulation of Il12b (30.2-fold), suggested switch to a pro-inflammatory response by 4 hrs post H37Rv-infection. Whereas no significant change in cytokine concentration or transcription was observed during the first hour post Δ-mce1 H37Rv-infection, a significant decline of IL-1b, IL-9, IL-13, Eotaxin and GM-CSF combined with increased transcription of Il12b (25.1-fold) and Inb1 (17.9-fold) by 4 hrs, indicated a pro-inflammatory response. The balance between pro-and anti-inflammatory responses during the early stages of infection may have significant bearing on outcome.

About the Authors:

Ruth Stavrum

* E-mail: [email protected]

Affiliation: Section of Microbiology and Immunology, the Gade Institute, University of Bergen, Bergen, Norway

Anne-Kristin Stavrum

Affiliations Center for Medical Genetics and Molecular Medicine, Haukeland University Hospital, Bergen, Norway, Department of Informatics, University of Bergen, Bergen, Norway

Håvard Valvatne

Current address: Intervet International B.V., Boxmeer, The Netherlands

Affiliation: Section of Microbiology and Immunology, the Gade Institute, University of Bergen, Bergen, Norway

Lee W. Riley

Affiliation: Division of Infectious Diseases and Vaccinology, School of Public Health, University of California, Berkeley, California, United States of America

Elling Ulvestad

Affiliations Section of Microbiology and Immunology, the Gade Institute, University of Bergen, Bergen, Norway, Department of Microbiology, Haukeland University Hospital, Bergen, Norway

Inge Jonassen

Affiliations Department of Informatics, University of Bergen, Bergen, Norway, Computational Biology Unit, Uni Computing, Uni Research AS, Bergen, Norway

Jörg Aßmus

Affiliation: Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway

T. Mark Doherty

Current address: GlaxoSmithKline, Brøndby, Copenhagen, Denmark

Affiliation: Infectious Disease Immunology, Statens Serum Institute, Copenhagen, Denmark

Harleen M. S. Grewal

Affiliations Section of Microbiology and Immunology, the Gade Institute, University of Bergen, Bergen, Norway, Department of Microbiology, Haukeland University Hospital, Bergen, Norway

Introduction

Tuberculosis (TB) is the most common cause of death worldwide caused by a single infectious pathogen, killing approximately 2 million people each year [1]. The causative agent of TB, Mycobacterium tuberculosis, is an intracellular pathogen which has evolved to infect and persist inside the host macrophage, a cell which is specialized for killing intracellular bacteria. M. tuberculosis infection is primarily acquired by invasion across the mucosal surfaces, most commonly by inhalation of the bacteria. Innate immunity is the first line of defense against infections and one of the first events in the innate resistance to intracellular bacterial infection is activation of macrophages. Upon inhalation the bacilli enter the lungs where they bind to the surface of alveolar macrophages, which undergo activation upon stimulation with bacterial secreted products or immune complexes. Activation of the macrophage results in the transcription of a range of macrophage genes. In the majority of cases (∼90%), the cell-mediated immune response to M. tuberculosis results in the formation of a granuloma, which is sufficient to contain the infection and prevent disease. However, in 5–10% of cases the bacilli may evade or subvert the host immune response causing either a latent infection or active disease [2].

The outcome of an M. tuberculosis infection is dependent on both host-specific and pathogen-specific factors [3]. The recognition of M. tuberculosis or mycobacterial products is a crucial step in the initiation of an effective host-response. Previous studies have shown that mycobacterial secretory products trigger cytoskeletal redistribution of the macrophage prior to the adherence of M. tuberculosis [4]. Although several bacterial genes have been reported to be important for the persistence of the M. tuberculosis in a mouse model for TB, such as icl, pca, mprA [5]–[8], little is known about the bacterial factors that influence the outcome of the infection. Structures in the mycobacterial cell envelope are important for the ability of the M. tuberculosis to establish an infection and persist inside its host. A group of proteins termed mammalian cell entry (Mce) 1, encoded by the mce1 operon, localize to the cell wall of the mycobacteria where they form a complex in the cell envelope [7], [9]. There is increasing evidence to suggest that the Mce family of proteins constitute beta-barrel outer membrane proteins responsible for transport across the membrane [10]–[12]. Previous studies suggest that the mce1 operon may act as a lipid importer [13], [14] being involved in lipid biosynthesis and/or lipid degradation [15]. Furthermore, disruption of the mce1 operon leads to alteration of the lipid profile of the mycobacterium and to the accumulation of free mycolic acids on its surface (S. A. Cantrell, personal communication). A previous study in mice infected with either a wild-type M. tuberculosis or with an mce1 mutant showed that the number of colony forming units (cfu) recovered from lungs as early as 2 weeks post-infection with the mutant strain was significantly greater than the cfu recovered from mice infected with the wild-type strain [7]. The same study also demonstrated that the infection with an mce1 mutant resulted in an unusual host cell response with less well-organized granuloma formation.

Given their triple role as host cell, antigen-presenting cell and potential killer of invading mycobacteria, the initial interactions between the cell wall of the pathogen and the host macrophage may be critical in determining the outcome of infection. Little is known about the transcriptional response by the macrophage to an infection with M. tuberculosis earlier than 4 hrs following infection [16], [17], and the transcriptional events potentially responsible for the inability of the host to effectively contain infection by the mce1-deficient M. tuberculosis strain have not been described. The objective of this study was therefore to closely monitor the initial transcriptional and immunologic responses of the macrophage following infection with a wild-type M. tuberculosis H37Rv strain and a Δ-mce1 M. tuberculosis H37Rv strain.

Materials and Methods

Cell-culture conditions

Murine macrophage cell line J774A.1 (MΦ) was purchased from The European Collection of Cell Cultures (ECACC). The cells were maintained using Dulbecco's modified Eagle's medium (DMEM, Lonza, Verviers, Belgium), supplemented with 10% fetal calf serum (FCS), 2 mM glutamine and 100 units/ml penicillin and 100 µg/ml streptomycin in a humidified atmosphere containing 5% CO2. Cells grew with a doubling time of 2 days and were split every 4 days. The last three sub-cultures prior to the infection experiment were performed without the addition of penicillin and streptomycin to the cell culture medium. Prior to infection the cells were allowed to grow to 80% confluence and gently washed twice with 10 ml pre-warmed (37°C) DMEM containing 10% FCS and 2 mM glutamine (complete medium).

Cultivation of mycobacteria

Broth cultures of M. tuberculosis strains H37Rv and Δ-mce1 M. tuberculosis H37Rv [7] were grown in Dubos broth with 10% albumin dextrose complex (ADC) supplement. For the growth of Δ-mce1 M. tuberculosis H37Rv the medium was supplemented with 50 µg/ml Hygromycine for maintaining the plasmid for the mutation. Starter cultures of 10 ml M. tuberculosis H37Rv and Δ-mce1 H37Rv were initially grown for 10 days with gentle shaking until a cloudy suspension was achieved and added to 200 ml fresh Dubos broth/ADC. These cultures were grown to a mid-log phase (OD580 = 0.5). The final cultures before infection were initiated with 20 ml of the mid-log phase suspension of H37Rv and Δ-mce1 H37Rv added to 180 ml of fresh Dubos broth with ADC supplement, and incubated for 7–8 days until an OD of 0.5. To minimize clumping of mycobacteria the cultures were shaken gently during growth, and prior to infection, the bacterial suspensions were vortexed for 1 minute, passed once through a 26G needle followed by 4×15 sec sonication and vortexing again for 1 minute.

Infection and RNA purification

J774A.1 macrophages (∼107 cells) were infected with mid-log phase H37Rv or Δ-mce1 H37Rv at a multiplicity of infection (MOI) of 10. Three biological replicates were included for each time-point for each strain. At 4 hrs post-infection, macrophages were washed three times with pre-warmed DMEM to remove extracellular bacteria. The number of cfu was measured by plating 100 µl onto 7H11 agar from the bacterial suspension prior to infection, and from the cell medium at 60 min and 4 hrs post H37Rv and Δ-mce1 H37Rv-infection and counted 14 days later. After 15 min, 30 min, 60 min, 4 hrs and 10 hrs following infection, the medium was removed and the cells were carefully washed 3 times with pre-warmed complete medium. The monolayers were lysed using 1.2 ml RLT/β-mercaptoethanol (Qiagen Sciences Inc, Germantown, USA). Total RNA was harvested from each flask using the RNAeasy Kit (Qiagen Sciences Inc, Germantown, USA). The concentration was measured using a NanoDrop scanning spectrophotometer (NanoDrop Technologies, Wilmington, USA) and the quality was measured using the Eukaryote Total RNA Nano 6000 assay (Agilent RNA 6000 Nano LabChip Kit) [Agilent Technologies, Santa Clara, USA].

Labelling of RNA

Total RNA was amplified and fluorescently labeled using the Low Input Linear Amplification Kit (Agilent Technologies, Santa Clara, USA) following the manufacturer's description. Each reaction contained 3 µg total RNA and 250 pg of RNA spike-in control. Spike-in mix A was included in the cyanine 3 (Cy-3) reactions and spike-in mix B in the cyanine 5 (Cy-5) reactions. The labeled cRNA was purified using the RNAeasy Kit (Qiagen Sciences Inc, Germantown, USA). Mass yields and specific activities of the labeled cRNA targets were determined by measuring absorbance using the NanoDrop scanning spectrophotometer (NanoDrop Technologies, Wilmington, USA). Quality of the labeled cRNA was further assessed using the mRNA Nano 6000 assay on the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, USA).

Hybridization

A pooled common reference was prepared by mixing 10 µl of total RNA isolated from each biological replicate of uninfected MΦ. Eight-hundred-twenty-five ng of Cy-5 labeled cRNA from M. tuberculosis H37Rv or Δ-mce1 H37Rv infected J774A.1 MΦ was randomly hybridized against 825 ng Cy-3 labeled pooled common reference cRNA onto 4×44K 60-mer oligo whole mouse genome - micro-arrays (Agilent Technologies, Santa Clara, USA) using the Agilent Gene Expression hybridization kit (# 5188–5242) as described in the Two-Color Microarray-Based Gene Expression Analysis v5.5 manual (Agilent Technologies, Santa Clara, USA). The arrays were hybridized at 65°C for 17 hrs/10 rpm. Following hybridization, the arrays were washed using the Gene Expression Wash Buffer 1 at room temperature and 2 at 37°C (Agilent #5188–5325 #5188–5326) following the manufacturer's description. Acetonitrile (Sigma-Aldrich, St. Louis, USA) and Agilent stabilization and drying solution (Agilent Technologies, Santa Clara, USA) were included in the two final steps in the washing procedure (for 1 min and 30 sec respectively). The arrays were scanned using the Agilent's dual laser DNA microarray scanner (part number G2505B). The scans were converted to data files with Agilent's Feature Extraction software (Version 9.1.3.1).

All microarray data are MIAME compliant and fully annotated raw microarray data has been deposited in ArrayExpress (accession number: E-TABM-1170).

Bio-Plex cytokine assay

Cell supernatants from M. tuberculosis H37Rv or Δ-mce1 H37Rv-infected J774A.1 macrophages were harvested from all 3 biological replicates prior to infection and from the 5 time-points post-infection (15 min, 30 min, 60 min, 4 and 10 hours) and immediately stored at −70°C. All cell supernatants were analyzed on a 96-well sterile filter plate using the Bio-Rad Mouse 23-plex cytokine assay (Bio-Rad, CA, USA) according to the manufacturer's instructions. All samples were analysed in duplicate (technical replicates) and the results from the technical replicates were combined. Data preprocessing was performed using the Bio-Plex Manager software (Bio-Rad, CA, USA) and exported into Microsoft Excel for further analysis.

RT-QPCR verification

Selected genes belonging to the two most significantly over-represented biological processes at 15 min post H37Rv-infection were verified by reverse transcription (RT) quantitative (Q)-PCR, of the relative amount of gene expression. The RT-QPCR mixture for genes meeting the predefined criteria and for two genes (Hmbs and Ubc) included as endogenous controls was prepared as follows: 10 µl TaqMan Gene Expression Master Mix (Applied Biosystems, Carlsbad, USA), 1 µl TaqMan Gene Expression Assay (Applied Biosystems, Carlsbad, USA), 8 µl PCR-grade water, and 1 µl (22 ng) template DNA. The thermal cycling protocol was as follows: UDG Incubaction for 2 min at 50°C, AmpliTaq Gold®, UP Enzyme Activation for 10 min at 95°C followed by 40 cycles of 15 sec at 95°C and 1 min at 60°C. The fluorescence signal was measured at the end of each extension step at 60°C.

RT-QPCR amplification and analysis were performed using the ABI 7500 instrument with software version 2.0.3 (Applied Biosystems, Carlsbad, USA) and the relative amount of gene expression was calculated using a pooled common reference as reference sample.

Data Analysis

Microarray data analysis.

Preprocessing and analysis was undertaken using the software J-Express Pro 2.9 [18]. Control spots, as well as all spots that were flagged by Feature Extraction (v. 9.1.3.1) or saturated in both channels were removed from the analysis. Log 2 ratios were calculated between the Cy 5 and Cy 3 signals from the remaining spots. Processed Signal values, which by the Feature Extraction default settings had been background corrected and normalized with respect to dye effects, were chosen to represent the signal intensity values. Technical replicates with the identical hybridization names were combined into a single column using the median of the signals per reporter. Missing values in the dataset, after filtering, were imputed with the method LS impute adaptive [19] as implemented in J-Express. Multiple reporters from the same gene (as defined by GeneName annotation from Agilent) were combined into a single gene profile using the max probe statistic (choosing the highest and probably most reliable signal to represent each sample). The data was divided into two data sub-sets, one for the time-series following H37Rv infection and one for time-series following the H37Rv Δ-mce1 infection. The association between all arrays was analyzed by Correspondence Analysis (CA) [20] and genes differentially expressed post-infection compared to prior to infection for each set of time-series were identified using Rank Product (RP) [21].

Functional classification.

Functional classification of the RP generated lists of genes being differentially expressed at certain time points post H37Rv or Δ-mce1 H37Rv-infection was performed using the Panther Classification System 6.1 [22]. Each of the gene lists were compared to the entire list of genes with detectable expression in at least one of the 36 samples (n = 13990) on the 4×44K 60-mer oligo whole mouse genome - microarrays (Agilent). Statistically over- and underrepresented annotated biological processes were determined by binominal statistics, using the observed number of genes versus the numbers expected by chance within a certain annotation group. Categories meeting the threshold of P-values below ∼10−3 were imported into the TM4 Microarray software Suite Multi Experiment Viewer 3.1 (TMeV, TIGR, US) [23] with one entity per category, where a heat map was created based on the negative log of the P- values for each category.

Results

Bacterial uptake

To assess if a deletion of the mce1 operon would have an effect on the bacterial uptake, by the macrophages, the cfu was determined at 60 min and 4 hrs post H37Rv- and Δ-mce1 H37Rv-infection, in addition to prior to infection. Equal number of cfu of each strain (1.5 ml of 1.33×108 cfu/ml [OD580 = 0.43] for H37Rv and 2.0 ml of 1.25×108 cfu/ml [OD580 = 0.35] for Δ-mce1 H37Rv) was added to a cell culture flask containing 10 ml of pre-warmed complete medium, resulting in a cfu concentration of 2.0×107/ml for H37Rv and 2.5×107/ml for Δ-mce1 H37Rv. The number of cfu recovered by 60 min post-infection was 2.2×106/ml for H37Rv and 1.5×107/ml for Δ-mce1 H37Rv. The cfu recovered at 4 hrs post-infection was 2.3×106/ml for H37Rv and 1.8×106/ml for Δ-mce1 H37Rv.

Microarray analysis

The CA plot (Figure 1) of all samples shows clear distinctions between infection by M. tuberculosis H37Rv [Rv] and Δ-mce1 M. tuberculosis H37Rv [Yk] and between the different time-points post-infection. All samples from the first 3 time-points (15 min, 30 min and 60 min) following Δ-mce1 H37Rv infection, except 1 replicate from the 60-min time-point, grouped together with the uninfected samples, indicating few transcriptional changes within the first hour post Δ-mce1 H37Rv-infection. All samples (except one Rv60 sample) from the first 3 time-points (15 min, 30 min and 60 min) post-infection with the H37Rv (Rv) strain form a separate cluster suggesting that there are clear distinctions with regards to the transcriptional level even after 15 min, and that these changes remain relatively constant during the first hour post H37Rv-infection. For the later time-points (4 hrs and 10 hrs) the CA plot shows that the samples from 4 hrs post Δ-mce1 H37Rv-infection cluster together with the samples from 4- and 10 hrs post H37Rv-infection, whereas the samples from the 10 hrs time-point post Δ-mce1 H37Rv-infection form a separate cluster. The proportion of total chi square statistic, explained by the plot, was 15.6%.

[Figure omitted. See PDF.]

Figure 1. Correspondence analysis (CA) of all samples from 5 time-points post-infection of mouse macrophages with wild-type M. tuberculosis H37Rv (Rv) or M. tuberculosis H37Rv Δ-mce1 (Yk).

Samples from the ‘early’ time-points post H37Rv-infection (15 min 30 min and 60 min) are colored yellow/orange, and samples from the ‘late’ time-points post H37Rv-infection (4 hrs and 10 hrs) are colored red/brown. Samples from the ‘early’ time-points post Δ-mce1-H37Rv infection (15 min, 30 min, and 60 min) are colored light green, and samples from the ‘late’ time-points post Δ-mce1-H37Rv -infection (4 hrs and 10 hrs]) are dark green/black. The uninfected samples (U) are colored purple, and the pooled common reference samples are colored blue.

https://doi.org/10.1371/journal.pone.0026295.g001

An immediate up-regulation of expression of 524 genes (using the RP-test and a false discovery rate [FDR] = 10%) was observed within the first 15 minutes post-infection with the H37Rv strain compared to the uninfected pooled common reference. In contrast, by 15 min post Δ-mce1 H37Rv-infection only 64 (FDR = 10%) genes were up-regulated compared to the uninfected pooled common reference (Table 1). Simultaneously, 505 and 590 genes (FDR = 10%) were down-regulated within the first 15 minutes post-infection with the H37Rv strain or the Δ-mce1 H37Rv strain, respectively, compared to the uninfected pooled common reference. The numbers of both up- and down-regulated genes at the various time-points post H37Rv and Δ-mce1 H37Rv-infection are listed in Table 1.

[Figure omitted. See PDF.]

Table 1. The number of up- and down-regulated genes (using a false discovery rate of 10%) in J774A.1 murine macrophages following infection by the M. tuberculosis H37Rv strain or the M. tuberculosis Δ-mce1 H37Rv strain at different time-points post-infection.

https://doi.org/10.1371/journal.pone.0026295.t001

Out of the 524 genes that were up-regulated by 15 min post H37Rv-infection (FDR = 10%), 32 genes were up-regulated by 5-fold or more compared to the uninfected pooled common reference (Table S1). Except for Hist1h1d, which was continuously expressed throughout the entire Δ-mce1 H37Rv time-course experiment, none of the genes that were up-regulated 5-fold or more by 15 min post H37Rv-infection were transcribed at a similar level the first hour post Δ-mce1 H37Rv-infection. By 4 hrs post Δ-mce1 H37Rv-infection 50% (16 out of 32) of the genes were up-regulated by at least 5-fold compared to the uninfected pooled common reference.

Functional Classification

Gene lists generated by comparing macrophage gene expression levels at each of the different time-points post H37Rv-infection and Δ-mce1 H37Rv-infection (RP lists) to the uninfected pooled common reference were uploaded into the Panther Classification System 6.1. Gene lists generated were: genes up- and down-regulated (FDR 10%) post H37Rv and Δ-mce1 H37Rv-infection, for each time point, compared to the uninfected pooled common reference. The numbers of genes in each of the 20 RP lists are listed in Table 1. Functional classifications of genes that have been mapped by the Panther Classification System 6.1 were performed for all gene lists and the biological processes that were over-represented among each of the lists are shown in Figure 2a (up-regulated genes) and 2b (down-regulated genes).

[Figure omitted. See PDF.]

Figure 2. Over-represented biological processes among up-regulated (A) and down-regulated (B) genes in J774A.1 murine macrophages following infection with M. tuberculosis H37Rv (Rv) or M. tuberculosis Δ-mce1 H37Rv (Yk).

The color intensity indicates the negative log of the p-values, dark values representing highly functional processes significantly over-represented among the genes. The numbers presented on the heat map display the percentage of genes within a gene set that map to a certain term, e.g. 19.7% of the 432 genes up-regulated (A) 15 min post H37Rv infection map to the biological process ‘Response to stimulus’. The first column depicts the overall distribution of a term among the 13, 990 genes with detectable expression in the data set, followed by the gene sets for the 5 time-points post H37Rv-infection and the 5 time-points post Δ-mce1-H37Rv infection.

https://doi.org/10.1371/journal.pone.0026295.g002

Up-regulated genes.

The heat-map displayed in Figure 2a demonstrates significantly over-represented biological processes within the macrophages following infection with the M. tuberculosis H37Rv or the Δ-mce1 M. tuberculosis H37Rv strains compared to the uninfected pooled common reference. The biological processes: ‘Immune system process’ (P-value: 2.81E-09) and ‘Response to stimulus’ (P-value: 4.70E-08) were the two most significantly over-represented biological processes among the up-regulated genes at 15 min post H37Rv-infection comprising 26.6% (115 out of 432) and 19.7% (85 out of 432) of the number of genes in the search list, respectively. In contrast, following Δ-mce1 H37Rv-infection the ‘Immune system process’ and ‘Response to stimulus’ biological processes became significantly over-represented first at 60 min (1.08E-04 and 2.51E-05, respectively) comprising 28.3% (39 out of 138) and 23.2% (32 out of 138) genes, respectively. By 4 and 10 hrs the ‘Immune system process’ process comprised 27.7% (169 out of 611 [P-value: 2.02E-14 ]) and 28.8% (193 out of 670 [P-value: 3.32E-18]), respectively, out of the genes up-regulated post Δ-mce1 H37Rv-infection, whereas the ‘Response to stimulus’ process comprised 21.6% (132 out of 611 [P-value: 1.09E-14]) and 21.5% (144 out of 670 [1.12E-15]), respectively.

Down-regulated genes.

The biological process ‘Cellular process’ (P-value: 1.44E-05) was the most significantly over-represented process among the genes which were down-regulated by 15 min post H37Rv-infection, comprising 47.3% (190 out of 402 genes) [Figure 2b]. In contrast, by 15 min post Δ-mce1 H37Rv-infection the biological process ‘Immune system process’ (P-value: 1.29E-04) was the most significantly over-represented process among the down-regulated genes, comprising 26.8% (110 out of 411) genes. By 10 hrs post H37Rv infection the process ‘Cell-cell signaling’ (P-value: 6.89E-04) was the most significantly over-represented biological process among the down-regulated genes, comprising 12.3% (69 out of 561) of the genes, whereas the biological process ‘Cell cycle’ (P-value: 7.29E-07) was the most significantly over-represented process among genes down-regulated following Δ-mce1 H37Rv-infection, comprising 16.9% (122 out of 724) of the genes (Figure 2b).

Cytokine assay

Of the 23 cytokines analysed, 6 cytokines (IL-2, IL-3, IL-4, IL-17, IFN-γ and KC) were not expressed to a detectable level at any of the time-points investigated.

As measured by the Bio Plex assay, cytokines normally associated with both the classical M1 and the alternative M2 activation of macrophages were present in the cell culture supernatants prior to infection, however, there was a marked reduction in the concentration of all cytokines by 15 min post H37Rv-infection. By 1 hr post H37Rv-infection there was a significant decrease (P<0.001) in the concentration of the cytokines; Rantes (Figure 3A), GM-CSF (Figure 3B), Eotaxin (Figure 3C) IL-13 (Figure 3D), IL-9 (Figure 3E) and IL-1b (Figure 3G). For IL-6 (Figure 3F) a significant reduction in the concentration was observed post H37Rv-infection only, whereas the observed reduction in IL-6 post Δ-mce1 H37Rv-infection, from 607 pg/ml to 271 pg/ml by 15 min post-infection and 258 pg/ml by 10 hrs, was not statistically significant.

[Figure omitted. See PDF.]

Figure 3. Cytokine profiles of cell supernatants secreted from M. tuberculosis H37Rv (white bars) or M. tuberculosis Δ-mce1 H37Rv (grey bars) infected J774A.1 murine macrophages.

The numbers below each column reflect the time-points post-infection: 1; uninfected, 2; 15 min, 3; 30 min, 4; 60 min, 5; 4 hrs, and 6; 10 hrs.

https://doi.org/10.1371/journal.pone.0026295.g003

The concentration in the cell culture supernatant for the cytokines MCP-1, MIP-1a and MIP-1b prior to infection was above the range of the maximum concentration of the standards at 23.1 ng/ml, 10.1 ng/ml and 38.3 ng/ml, respectively. Post H37Rv-infection, the concentration of MCP-1, MIP-1a and MIP-1b in the supernatants decreased continuously throughout the course of the infection to 1.5 ng/ml, 3.1 ng/ml and 2.8 ng/ml, respectively, by 10 hrs post-infection. Post Δ-mce1 H37Rv-infection the concentration of MCP-1 and MIP-1b remained above the range of the maximum concentration of the standards for the 15 min, 30 min and 60 min time-points before a drop in the concentration was observed at 4 hrs (8.6 ng/ml and 13.6 ng/ml, respectively) and 10 hrs (3.0 ng/ml and 7.3 ng/ml, respectively). For MIP-1a the concentration post Δ-mce1 H37Rv-infection remained above the range of the maximum concentration of the standards for the 15 min, 30 min, 60 min, and 4 hrs time-points whereas at 10 hrs post-infection the concentration had dropped to 5.5 ng/ml (Table S2).

For 6 of the cytokines measured [IL-1a, IL-5, IL-10, IL-12(p40) and IL-12(p70)], the concentration in the supernatant dropped to a level which was below the detection limits for the particular assays by 15 min post H37Rv-infection. In contrast, post Δ-mce1 H37Rv-infection the concentration of these cytokines remained similar to the uninfected controls for the 15 min, 30 min and 60 min time-points before a reduction in the concentration was observed at 4 hrs and 10 hrs post-infection. The concentrations for these 6 cytokines at the various time-points post-infection are provided in Table S2.

For the cytokines Rantes, GM-CSF, Eotaxin, IL-13, IL-9, IL-6 and IL-1b (Figure 3A–G) there was no significant mean difference in concentration among the first three time-points (15 min, 30 min and 60 min) and the last two time-points (4 hrs and 10 hrs) for both the H37Rv and Δ-mce1 H37Rv-infections. Thus, the measurements from the Bio Plex assay for the first three time-points (15 min, 30 min and 60 min) for each of the infections (post H37Rv-infection and post Δ-mce1 H37Rv-infection) were combined into one group termed ‘early’ infection, and the measurements for the last two time-points (4 hrs and 10 hrs) were combined into one group termed ‘late’ infection.

A comparison of the groups (‘early’ [15 min, 30 min and 60 min] and ‘late’[4 hrs and 10 hrs]) post H37Rv-infection and post Δ-mce1 H37Rv-infection showed that there was a significant difference in the concentration of cytokines when comparing the post ‘early’ H37Rv-infection vs. the post ‘early’ Δ-mce1 H37Rv-infection groups, and the post ‘late’ H37Rv-infection vs. post ‘early’ Δ-mce1 H37Rv-infection groups for IL-1b, IL-9, IL13, Eotaxin and GM-CSF (all had a P-value of <0.001). However, when comparing the post ‘early’ H37Rv-infection vs. post ‘late’ Δ-mce1 H37Rv-infection groups and the post ‘late’ H37Rv-infection vs. post ‘late’ Δ-mce1 H37Rv-infection groups there were no significant differences in the level of cytokine concentration for any of the cytokines analysed.

RT-QPCR verification

Six genes belonging to the two most significantly over-represented biological processes at the 15 min post H37Rv-infection ‘Immune system response’ (Ifnb1, Il12b, Ccl24, Pparg, and Clec4a2) and ‘Response to stimulus’ (Il13ra1) were selected for verification by RT-QPCR analysis. Statistical analyses showed that there was no significant mean difference in level of transcription among the first three time-points (15 min, 30 min and 60 min) and the last two time-points (4 hrs and 10 hrs) post H37Rv-infection and Δ-mce1 H37Rv-infection, i.e. measurements within the first hour (15 min, 30 min and 60 min) and for the last two time-points (4 hrs and 10 hrs) could be treated as equal, ignoring that the measurements were taken at different time-points. Thus, for the statistical analyses, the RT-QPCR results for the first three time-points (15 min, 30 min and 60 min) for each of the infections (post H37Rv-infection and post Δ-mce1 H37Rv-infection) were combined into one group termed ‘early’ infection, and the RT-QPCR results for the last two time-points (4 hrs and 10 hrs) were combined into one group termed ‘late’ infection.

The relative quantification analysis showed a significant increase in gene transcription by 15 min following H37Rv-infection for the Ccl24 (26.5 fold [P<0.001]), Pparg (10.5 fold [P = 0.003]), Clec4a2 (23.2 fold [P<0.001]) genes (Figure 4E) and the Il12b (12.2 fold [P = 0.002]) gene (Figure 4C). In contrast, following Δ-mce1 H37Rv-infection none of the genes analysed by RT-QPCR were significantly up-regulated compared to the uninfected pooled common reference during the first hour post infection (Figure 4B, 4D, 4F and 4H).

[Figure omitted. See PDF.]

Figure 4. Transcriptional profile of the genes selected for RT-QPCR verification.

Each data point represents the relative quantification of gene expression, by J774A.1 murine macrophages, between the pooled common reference and the different time-points post M. tuberculosis H37Rv (Rv) or M. tuberculosis Δ-mce1 H37Rv (Yk) infection.

https://doi.org/10.1371/journal.pone.0026295.g004

By 4 hrs post H37Rv-infection the Ccl24 and Il12b genes continued to be significantly up-regulated compared to the uninfected pooled common reference (9.4 fold [P = 0.003] and 30.2 fold [P = 0.001], respectively), whereas the Clec4a2 and Pparg genes were down-regulated to a similar level as the uninfected pooled common reference (2.9 fold [P = 0.391] and 1.5 fold [P = 0.019], respectively). The Ifnb1 (Figure 4A) and Il13ra1 (Figure 4D) genes were not significantly differentially expressed compared to the uninfected pooled common reference at any of the time-points following H37Rv-infection.

By 4 hrs post Δ-mce1 H37Rv-infection the transcription of the genes Ifnb1, Il12b and Pparg became significantly up-regulated compared to the uninfected pooled common reference (17.9 fold [P<0.001], 25.1 fold [P<0.001] and 0.7 fold [P = 0.001], respectively), whereas the transcription of Ccl24, Clec4a2 and Il13ra1 remained similar to that of the uninfected pooled common reference throughout out the time-course experiment (Table 2).

[Figure omitted. See PDF.]

Table 2. Relative quantification of genes selected for verification by RT-QPCR at different time-points post-infection with M. tuberculosis H37Rv or M. tuberculosis Δ-mce1-H37Rv.

https://doi.org/10.1371/journal.pone.0026295.t002

A comparison of the groups (‘early’ [15 min, 30 min and 60 min] and ‘late’ [4 hrs and 10 hrs]) post H37Rv-infection and post Δ-mce1 H37Rv-infection showed that there was a significant difference in the level of transcription for all genes analysed by RT-QPCR when comparing the post ‘early’ H37Rv-infection vs. the post ‘early’ Δ-mce1 H37Rv-infection groups (excluding Ifnb1 [P = 0.651]), and the post ‘early’ H37Rv-infection vs. post ‘late’ Δ-mce1 H37Rv-infection (including Ifnb1) [all had a P-value of <0.001]. However, when comparing the post ‘late’ H37Rv-infection vs. post ‘early’ Δ-mce1 H37Rv-infection groups only Il12b and Il13ra1 showed a significant difference in the level of transcription between the two types of infection (P<0.001) and for the post ‘late’ H37Rv-infection vs. post ‘late’ Δ-mce1 H37Rv-infection groups only Ifnb1 (P<0.001) showed a significant difference in the level of transcription between the two types of infection.

Luminex and RT-QPCR data analysis.

Statistical analyses were undertaken to detect differences between groups of measurements. The t-test was applied when comparing groups of 2, whereas ANOVA was applied when comparing >2 groups. For both the RT-QPCR (n = 6) and the cytokine measures (n = 6) multiple testing effects were taken into account, thus adjusting the significance level by the Bonferroni rule using a significance level of 0.05/6 = 0.00833. The computations were done using SPSS 17.

Discussion

The immune response against M. tuberculosis is multifaceted and is further complicated by the dual role of the macrophages; they represent both the primary effector cells for the killing of the bacteria and the primary habitat for bacterial persistence. A number of pathogens modulate the host immune response by the secretion of effector proteins [16], [24], [25] and the modulatory influence by live M. tuberculosis on its cellular host has been demonstrated in several experiments [26]–[29]. The M. tuberculosis Mce1 protein complex, which localizes to the cell envelope of the mycobacteria, is important for infection and persistence [30]. Deletion of the mce1 operon has been shown to result in a hypervirulent M. tuberculosis strain, which is poorly controlled and generates disseminated disease in experimental infections [7]. The present study describes the transcriptional and immunological responses by murine J774A.1 MΦ to the initial encounter with a wild-type M. tuberculosis H37Rv strain and a Δ-mce1 M. tuberculosis H37Rv mutant strain.

Upon stimulation with M. tuberculosis H37Rv the concentration of all cytokines measured declined and in some cases fell below the detection limit of the assay within 15 min post-infection. The up-regulation of the IL-13 receptor, Il13ra1, in combination with the up-regulation of the IL-13 inducible Ccl24 and Pparγ genes (Table 2) by 15 min post H37Rv-infection suggests that an anti-inflammatory transcriptional response was initiated by IL-13 signaling during the initial phase of M. tuberculosis H37Rv infection. Previous studies [31]–[33] have demonstrated the role for IL-13 in inhibiting autophagy in macrophages; a process which suppress intracellular survival of mycobacteria. Surprisingly, upon stimulation with Δ-mce1 H37Rv the concentration of all cytokines measured, except for IL-6, remained similar to that of the uninfected controls during the first hour post-infection (Figure 3). An up-regulation of the transcription of the IL-13 receptor (Il13ra1) was observed by 15 min post Δ-mce1 H37Rv-infection however, a corresponding change in the concentration of IL-13 measured in the supernatant could not be detected during the first hour post Δ-mce1 H37Rv-infection. Although a decline in cytokine concentration was observed by 4 hours post Δ-mce1 H37Rv-infection, the lack of effect on transcription of the IL-13-inducible genes by 4 hrs post Δ-mce1 H37Rv-infection may be explained by a down-regulation of the IL-13 receptor by 4 hrs post Δ-mce1 H37Rv-infection. The genes coding for the majority of cytokines that were measured by the 23-plex luminex assay were, as measured by microarray, not differentially expressed, for any of the time-points investigated. Thus, the observed initial reduction in the cytokine concentration during the first hour post H37Rv-infection may result from post-translational inhibition of the gene transcript, cytokine degradation by secreted mycobacterial proteases, or from an immediate binding of the cytokines to their respective receptors on the macrophage or to receptors on the mycobacteria.

A study by Dunphy et al [34] suggests that one gene (fadD5), which is located within the mce1 operon may be involved in the recycling of mycolic acids. Furthermore, other studies which have been performed on genes with high sequence similarity to genes in the mce1 operon suggest that there is evidence for the involvement of the mce1 operon in fatty acid transport and fatty acid degradation [35], [36]. Disruption of the mechanisms responsible for transport or recycling of fatty acids across the mycobacterial cell envelope may result in an altered lipid composition of the cell envelope. Furthermore, several studies have shown that the mycobacterial cell envelope mycolic acids have a modulatory effect on the host immune response and that the modification of the mycolic acids may have an anti-inflammatory effect on macrophage function [37], [5], [38], [39]. Thus, the lack of an initial transcriptional response by the macrophage to infection by the Δ-mce1 H37Rv strain may be the result of an alteration of the mycolic acid structure or content on the mycobacterial cell envelope caused by a disruption of the fatty acid recycling mechanism coded for by the mce1 operon (S. A. Cantrell, personal communication).

The number of extracellular cfu recovered by 60 min post-infection was ∼10-fold lower for H37Rv than that of Δ-mce1 H37Rv, whereas by 4 hrs post-infection, the number of extracellular cfu recovered was similar for the two strains. Previous studies using RAW264.7 macrophages showed that, by two hours post-infection, there was no difference in the ability of the H37Rv and the Δ-mce1 H37Rv strains to invade the host macrophage [7]. Thus, the lack of macrophage response towards the Δ-mce1 H37Rv strain during the first hour post-infection, as measured by cytokine concentration, may reflect a delay in the phagocytosis of the Δ-mce1 H37Rv strain by the macrophages. Furthermore, it has also been shown that the disruption of the mce1 operon prevented the mutated bacteria from entering into a persistent state resulting in more extensive replication of the bacteria [7] and that the presence of IFN-α/β allowed mycobacteria to grow uncontrollably in monocytes suggesting that secretion of IFN-α/β directly promotes mycobacterial growth [40]. Induction of type I interferons have also been shown to be involved in the reduced Th1-type T-cell response observed in mice infected by the hypervirulent M. tuberculosis HN878 strain, resulting in an increased bacillary load and increased mortality [41]. The increased transcription of Ifnb1 (Figure 4C) and Il1a, Il1b, Cd36, Mmp9 and Mmp12 (data not shown) by 4 hrs post Δ-mce1 H37Rv-infection, in combination with the lack of IL-13 induced immune response support previous findings that, upon an encounter with the Δ-mce1 H37Rv, the macrophage initiates an immunological response which is less able to control bacterial replication leading to more immunopathology than the response induced by the H37Rv wild-type strain. These results differ from those reported previously [7] – an observation which may reflect the fact that host responses in this study were measured in the first few hours after infection whereas the previous report showed results after 1–3 days post-infection. It is clear from the data presented here that gene expression changes dramatically through the early infection period, probably reflecting the evolving response of the host cell and its interaction with the wild type and Mce1-deficient pathogens.

The IL-13 inducible gene Clec4a2, coding for the dendritic cell immunoreceptor (DCIR), is a member of the type II calcium-dependent (C-type) lectin family which efficiently presents internalized antigens to T-cells and selectively inhibits TLR8-mediated IL-12 and TNF-α production and TLR9-mediated IFN-α [42], [43]. The observed lack of up-regulation of transcription of the DCIR receptor post Δ-mce1 H37Rv-infection as compared to post-H37Rv infection, may suggest that the Δ-mce1 H37Rv strain does not effectively present the structures that normally bind to the DCIR receptor, or that the Mce1-mutant is capable of suppressing the induction of the transcription of this inhibitory receptor. In a recent study, Simmons et al [44] showed that mycobacterial lipoproteins signaling through TLR2 inhibited induction of Ifnb1 in dendritic cells, thus contributing to the modulation of the immune response. The observed suppression of transcription Ifnb1 post H37Rv-infection may be explained by the production of immunomodulatory lipoproteins signaling through TLR2 or the up-regulation of transcription of Clec4a2, which may inhibit TLR9-induced induction of Ifnb1. Depending on their degree of virulence, different strains of M. tuberculosis induce different levels of pro- and anti-inflammatory responses [45]. Furthermore, it has been shown that stimulation with either pro-inflammatory or anti-inflammatory cytokines allows the macrophages to switch from one activation state to the other [46] and that reactivation of latent TB disease is correlated with a shift from pro-inflammatory type 1 cytokines to the anti-inflammatory type 2 cytokines [47], while control of TB disease is associated with the opposite pattern [48]. By 4 hrs post H37Rv-infection there appeared to be a shift towards the more protective pro-inflammatory type 1 response, similar to the response observed by 4 hrs post Δ-mce1 H37Rv-infection, indicated by the down-regulation of the IL-13ra1 receptor, Ccl24 and Pparg in combination with an up-regulation of the transcription of the genes Cd36, Mmp9 and Mmp12 and the genes coding for the pro-inflammatory cytokines IL-12 (Il12b) [Table 2] and IL-1 (Il1a and Il1b [data not shown]). The findings from this study indicate that during the initial encounter of the macrophage the Mce1 protein complex may be involved in eliciting an immediate anti-inflammatory immune response by the macrophage which shifts towards a more pro-inflammatory response by about 4 hrs post-infection. It is possible that the Mce1-induced response modulates the subsequent inflammatory response to reduce immunopathology – consistent with the extensive pathology observed in mice infected with Mce1-deficient M. tuberculosis, and with its proposed role as a regulator of latency [7]. Furthermore, the deletion of the mce1 operon results in a different and perturbed immunomodulatory response which may hinder control of bacterial replication, and thereby increased pathology. It would be pertinent to conduct further studies that look more closely at specific bacterial components or molecules that favor the early macrophage shift to more protective host responses.

Supporting Information

[Figure omitted. See PDF.]

Table S1.

Genes induced by the J774A.1 macrophage following H37Rv or Δ-mce1-H37Rv-infection. The values represent the fold-change between each time-point post-infection compared to the uninfected pooled common reference. Only genes induced by at least 5-fold 15 min post H37Rv-infection are shown.

https://doi.org/10.1371/journal.pone.0026295.s001

(DOC)

Table S2.

Cell supernatant cytokines released by the J774A.1 macrophage at the different time-points post-infection with the M. tuberculosis H37Rv strain or the M. tuberculosis Δ-mce1-H37Rv strain. IL-2, IL-3, IL-4, IL-17, IFN-γ and KC were not expressed to a detectable level for any of the time-points.

https://doi.org/10.1371/journal.pone.0026295.s002

(DOC)

Acknowledgments

The authors thank Philip D. Butcher and Sabine Ehrt for valuable contributions towards the planning of the experiments. We thank Karl-Henning Kalland and Kari Rostad for constructive advice, Kristi Øvreås and Kjell Petersen for technical assistance, Haukeland University Hospital for P3-laboratory facilities, the Norwegian Microarray Consortium, Bergen, for microarray facilities, and the FUGE bioinformatics platform for valuable bioinformatics support.

Author Contributions

Conceived and designed the experiments: RS HV HMSG. Performed the experiments: RS HV HMSG. Analyzed the data: RS AKS JA. Contributed reagents/materials/analysis tools: LWR IJ EU. Wrote the paper: RS TMD HMSG.

Citation: Stavrum R, Stavrum A-K, Valvatne H, Riley LW, Ulvestad E, Jonassen I, et al. (2011) Modulation of Transcriptional and Inflammatory Responses in Murine Macrophages by the Mycobacterium tuberculosis Mammalian Cell Entry (Mce) 1 Complex. PLoS ONE6(10): e26295. https://doi.org/10.1371/journal.pone.0026295

References

1. Frieden T, Sterling T, Munsiff S, Watt C, Dye C (2003) Tuberculosis. The Lancet 362: 887–899.T. FriedenT. SterlingS. MunsiffC. WattC. Dye2003Tuberculosis.The Lancet362887899Available: http://linkinghub.elsevier.com/retrieve/pii/S0140673603143334. Available: http://linkinghub.elsevier.com/retrieve/pii/S0140673603143334.

2. Flynn JL, Chan J (2001) Tuberculosis: latency and reactivation. Infection and immunity 69: 4195–201.JL FlynnJ. Chan2001Tuberculosis: latency and reactivation.Infection and immunity694195201Available: http://iai.asm.org. Available: http://iai.asm.org.

3. Malik ANJ, Godfrey-Faussett P (2005) Effects of genetic variability of Mycobacterium tuberculosis strains on the presentation of disease. The Lancet infectious diseases 5: 174–83.ANJ MalikP. Godfrey-Faussett2005Effects of genetic variability of Mycobacterium tuberculosis strains on the presentation of disease.The Lancet infectious diseases517483Available: http://www.ncbi.nlm.nih.gov/pubmed/15766652. Available: http://www.ncbi.nlm.nih.gov/pubmed/15766652.

4. García-Pérez BE, Mondragón-Flores R, Luna-Herrer J (2003) Internalization of Mycobacterium tuberculosis by macropinocytosis in non-phagocytic cells. Microbial Pathogenesis 35: 49–55.BE García-PérezR. Mondragón-FloresJ. Luna-Herrer2003Internalization of Mycobacterium tuberculosis by macropinocytosis in non-phagocytic cells.Microbial Pathogenesis354955Available: http://dx.doi.org/10.1016/S0882-4010(03)00089-5. Available: http://dx.doi.org/10.1016/S0882-4010(03)00089-5.

5. Glickman MS, Cox JS, Jacobs WR (2000) A novel mycolic acid cyclopropane synthetase is required for cording, persistence, and virulence of Mycobacterium tuberculosis. Molecular cell 5: 717–27.MS GlickmanJS CoxWR Jacobs2000A novel mycolic acid cyclopropane synthetase is required for cording, persistence, and virulence of Mycobacterium tuberculosis.Molecular cell571727Available: http://www.ncbi.nlm.nih.gov/pubmed/10882107. Available: http://www.ncbi.nlm.nih.gov/pubmed/10882107.

6. McKinney JD, Höner Zu Bentrup K, Muñoz-Elías EJ, Miczak A, Chen B, et al. (2000) Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406: 735–8.JD McKinneyK. Höner Zu BentrupEJ Muñoz-ElíasA. MiczakB. Chen2000Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase.Nature4067358Available: http://www.ncbi.nlm.nih.gov/pubmed/10963599. Available: http://www.ncbi.nlm.nih.gov/pubmed/10963599.

7. Shimono N, Morici L, Casali N, Cantrell S, Sidders B, et al. (2003) Hypervirulent mutant of Mycobacterium tuberculosis resulting from disruption of the mce1 operon. Proceedings of the National Academy of Sciences of the United States of America 100: 15918–23.N. ShimonoL. MoriciN. CasaliS. CantrellB. Sidders2003Hypervirulent mutant of Mycobacterium tuberculosis resulting from disruption of the mce1 operon.Proceedings of the National Academy of Sciences of the United States of America1001591823Available: http://www.ncbi.nlm.nih.gov/pubmed/14663145. Available: http://www.ncbi.nlm.nih.gov/pubmed/14663145.

8. Zahrt TC, Deretic V (2001) Mycobacterium tuberculosis signal transduction system required for persistent infections. Proceedings of the National Academy of Sciences of the United States of America 98: 12706–11.TC ZahrtV. Deretic2001Mycobacterium tuberculosis signal transduction system required for persistent infections.Proceedings of the National Academy of Sciences of the United States of America981270611Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=60118&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=60118&tool=pmcentrez&rendertype=abstract.

9. Casali N, White AM, Riley LW (2006) Regulation of the Mycobacterium tuberculosis mce1 operon. Journal of bacteriology 188: 441–9.N. CasaliAM WhiteLW Riley2006Regulation of the Mycobacterium tuberculosis mce1 operon.Journal of bacteriology1884419Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1347267&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1347267&tool=pmcentrez&rendertype=abstract.

10. Mah N, Perez-Iratxeta C, Andrade-Navarro MA (2010) Outer membrane pore protein prediction in mycobacteria using genomic comparison. Microbiology (Reading, England) 156: 2506–15.N. MahC. Perez-IratxetaMA Andrade-Navarro2010Outer membrane pore protein prediction in mycobacteria using genomic comparison.Microbiology (Reading, England)156250615Available: http://mic.sgmjournals.org/cgi/content/abstract/156/8/2506. Available: http://mic.sgmjournals.org/cgi/content/abstract/156/8/2506.

11. Pajón R, Yero D, Lage A, Llanes A, Borroto CJ (2006) Computational identification of beta-barrel outer-membrane proteins in Mycobacterium tuberculosis predicted proteomes as putative vaccine candidates. Tuberculosis (Edinburgh, Scotland) 86: 290–302.R. PajónD. YeroA. LageA. LlanesCJ Borroto2006Computational identification of beta-barrel outer-membrane proteins in Mycobacterium tuberculosis predicted proteomes as putative vaccine candidates.Tuberculosis (Edinburgh, Scotland)86290302Available: http://dx.doi.org/10.1016/j.tube.2006.01.005. Available: http://dx.doi.org/10.1016/j.tube.2006.01.005.

12. Song H, Sandie R, Wang Y, Andrade-Navarro MA, Niederweis M (2008) Identification of outer membrane proteins of Mycobacterium tuberculosis. Tuberculosis (Edinburgh, Scotland) 88: 526–44.H. SongR. SandieY. WangMA Andrade-NavarroM. Niederweis2008Identification of outer membrane proteins of Mycobacterium tuberculosis.Tuberculosis (Edinburgh, Scotland)8852644Available: http://dx.doi.org/10.1016/j.tube.2008.02.004. Available: http://dx.doi.org/10.1016/j.tube.2008.02.004.

13. Casali N, Riley LW (2007) A phylogenomic analysis of the Actinomycetales mce operons. BMC genomics 8: 60.N. CasaliLW Riley2007A phylogenomic analysis of the Actinomycetales mce operons.BMC genomics860Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1810536&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1810536&tool=pmcentrez&rendertype=abstract.

14. Joshi SM, Pandey AK, Capite N, Fortune SM, Rubin EJ, et al. (2006) Characterization of mycobacterial virulence genes through genetic interaction mapping. Proceedings of the National Academy of Sciences of the United States of America 103: 11760–5.SM JoshiAK PandeyN. CapiteSM FortuneEJ Rubin2006Characterization of mycobacterial virulence genes through genetic interaction mapping.Proceedings of the National Academy of Sciences of the United States of America103117605Available: http://www.pnas.org/cgi/content/abstract/103/31/11760. Available: http://www.pnas.org/cgi/content/abstract/103/31/11760.

15. Cantrell SA (2005) A structural and functional analysis of the proteins encoded by the Mycobacterium tuberculosis mce1 operon University of California, Berkeley. SA Cantrell2005A structural and functional analysis of the proteins encoded by the Mycobacterium tuberculosis mce1 operon University of California, Berkeley.Available: http://proquest.umi.com/pqdlink?Ver=1&Exp=11-23-2015&FMT=7&DID=982836321&RQT=309&attempt=1&cfc=1. Available: http://proquest.umi.com/pqdlink?Ver=1&Exp=11-23-2015&FMT=7&DID=982836321&RQT=309&attempt=1&cfc=1.

16. Nau GJ, Richmond JFL, Schlesinger A, Jennings EG, Lander ES, et al. (2002) Human macrophage activation programs induced by bacterial pathogens. Proceedings of the National Academy of Sciences of the United States of America 99: 1503–8.GJ NauJFL RichmondA. SchlesingerEG JenningsES Lander2002Human macrophage activation programs induced by bacterial pathogens.Proceedings of the National Academy of Sciences of the United States of America9915038Available: http://www.pnas.org/cgi/content/abstract/99/3/1503. Available: http://www.pnas.org/cgi/content/abstract/99/3/1503.

17. Ragno S, Romano M, Howell S, Pappin DJ, Jenner PJ, et al. (2001) Changes in gene expression in macrophages infected with Mycobacterium tuberculosis: a combined transcriptomic and proteomic approach. Immunology 104: 99–108.S. RagnoM. RomanoS. HowellDJ PappinPJ Jenner2001Changes in gene expression in macrophages infected with Mycobacterium tuberculosis: a combined transcriptomic and proteomic approach.Immunology10499108Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1783284&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1783284&tool=pmcentrez&rendertype=abstract.

18. Dysvik B, Jonassen I (2001) J-Express: exploring gene expression data using Java. Bioinformatics 17: 369–370.B. DysvikI. Jonassen2001J-Express: exploring gene expression data using Java.Bioinformatics17369370Available: http://www.bioinformatics.oupjournals.org/cgi/doi/10.1093/bioinformatics/17.4.369. Available: http://www.bioinformatics.oupjournals.org/cgi/doi/10.1093/bioinformatics/17.4.369.

19. Bø TH, Dysvik B, Jonassen I (2004) LSimpute: accurate estimation of missing values in microarray data with least squares methods. Nucleic acids research 32: e34.TH BøB. DysvikI. Jonassen2004LSimpute: accurate estimation of missing values in microarray data with least squares methods.Nucleic acids research32e34Available: http://www.ncbi.nlm.nih.gov/pubmed/14978222. Available: http://www.ncbi.nlm.nih.gov/pubmed/14978222.

20. Fellenberg K, Hauser NC, Brors B (2001) Correspondence analysis applied to microarray data. Proceedings of the National Academy of Science of the United States of America 98: 10781–10786.K. FellenbergNC HauserB. Brors2001Correspondence analysis applied to microarray data.Proceedings of the National Academy of Science of the United States of America981078110786Available: http://www.pnas.org/cgi/content/abstract/98/19/10781. Available: http://www.pnas.org/cgi/content/abstract/98/19/10781.

21. Breitling R, Armengaud P, Amtmann A, Herzyk P (2004) Rank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments. FEBS letters 573: 83–92.R. BreitlingP. ArmengaudA. AmtmannP. Herzyk2004Rank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments.FEBS letters5738392Available: http://dx.doi.org/10.1016/j.febslet.2004.07.055. Available: http://dx.doi.org/10.1016/j.febslet.2004.07.055.

22. Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, et al. (2003) PANTHER: a library of protein families and subfamilies indexed by function. Genome research 13: 2129–41.PD ThomasMJ CampbellA. KejariwalH. MiB. Karlak2003PANTHER: a library of protein families and subfamilies indexed by function.Genome research13212941Available: http://genome.cshlp.org/cgi/content/abstract/13/9/2129. Available: http://genome.cshlp.org/cgi/content/abstract/13/9/2129.

23. Saeed AI, Sharov V, White J, Li J, Liang W (2003) TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34: 374–378.AI SaeedV. SharovJ. WhiteJ. LiW. Liang2003TM4: a free, open-source system for microarray data management and analysis.Biotechniques34374378Available: http://www.ncbi.nlm.nih.gov/pubmed/12613259/?report=summary. Available: http://www.ncbi.nlm.nih.gov/pubmed/12613259/?report=summary.

24. Finlay BB, McFadden G (2006) Anti-immunology: evasion of the host immune system by bacterial and viral pathogens. Cell 124: 767–82.BB FinlayG. McFadden2006Anti-immunology: evasion of the host immune system by bacterial and viral pathogens.Cell12476782Available: http://dx.doi.org/10.1016/j.cell.2006.01.034. Available: http://dx.doi.org/10.1016/j.cell.2006.01.034.

25. Schwab JH (1975) Suppression of the immune response by microorganisms. Bacteriological reviews 39: 121–43.JH Schwab1975Suppression of the immune response by microorganisms.Bacteriological reviews3912143Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=413897&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=413897&tool=pmcentrez&rendertype=abstract.

26. Beltan E, Horgen L, Rastogi N (2000) Secretion of cytokines by human macrophages upon infection by pathogenic and non-pathogenic mycobacteria. Microbial pathogenesis 28: 313–8.E. BeltanL. HorgenN. Rastogi2000Secretion of cytokines by human macrophages upon infection by pathogenic and non-pathogenic mycobacteria.Microbial pathogenesis283138Available: http://dx.doi.org/10.1006/mpat.1999.0345. Available: http://dx.doi.org/10.1006/mpat.1999.0345.

27. Falcone V, Bassey EB, Toniolo A, Conaldi PG, Collins FM (1994) Differential release of tumor necrosis factor-α from murine peritoneal macrophages stimulated with virulent and avirulent species of mycobacteria. FEMS Immunology and Medical Microbiology 8: 7.V. FalconeEB BasseyA. TonioloPG ConaldiFM Collins1994Differential release of tumor necrosis factor-α from murine peritoneal macrophages stimulated with virulent and avirulent species of mycobacteria.FEMS Immunology and Medical Microbiology87

28. Giacomini E, Iona E, Ferroni L, Miettinen M, Fattorini L, et al. (2001) Infection of human macrophages and dendritic cells with Mycobacterium tuberculosis induces a differential cytokine gene expression that modulates T cell response. Journal of immunology (Baltimore, Md.: 1950) 166: 7033–41.E. GiacominiE. IonaL. FerroniM. MiettinenL. Fattorini2001Infection of human macrophages and dendritic cells with Mycobacterium tuberculosis induces a differential cytokine gene expression that modulates T cell response.Journal of immunology (Baltimore, Md.: 1950)166703341Available: http://www.jimmunol.org/cgi/content/abstract/166/12/7033. Available: http://www.jimmunol.org/cgi/content/abstract/166/12/7033.

29. Stanley SA, Raghavan S, Hwang W, Cox JS (2003) Acute infection and macrophage subversion by Mycobacterium tuberculosis require a specialized secretion system. Proceedings of the National Academy of Sciences of the United States of America 100: 13001–6.SA StanleyS. RaghavanW. HwangJS Cox2003Acute infection and macrophage subversion by Mycobacterium tuberculosis require a specialized secretion system.Proceedings of the National Academy of Sciences of the United States of America100130016Available: http://www.pnas.org/cgi/content/abstract/100/22/13001. Available: http://www.pnas.org/cgi/content/abstract/100/22/13001.

30. Arruda S, Bomfim G, Knights R, Huima-Byron T, Riley L (1993) Cloning of an M. tuberculosis DNA fragment associated with entry and survival inside cells. Science 261: 1454–1457.S. ArrudaG. BomfimR. KnightsT. Huima-ByronL. Riley1993Cloning of an M. tuberculosis DNA fragment associated with entry and survival inside cells.Science26114541457Available: http://www.sciencemag.org/cgi/doi/10.1126/science.8367727. Available: http://www.sciencemag.org/cgi/doi/10.1126/science.8367727.

31. Deretic V (2005) Autophagy in innate and adaptive immunity. Trends in immunology 26: 523–8.V. Deretic2005Autophagy in innate and adaptive immunity.Trends in immunology265238Available: http://dx.doi.org/10.1016/j.it.2005.08.003. Available: http://dx.doi.org/10.1016/j.it.2005.08.003.

32. Wright K (1997) Activation of Phosphatidylinositol 3-Kinase by Interleukin-13. An inhibitory signal for inducible nitric-oxide synthase expression in epithelial cell line HT-29. Journal of Biological Chemistry 272: 12626–12633.K. Wright1997Activation of Phosphatidylinositol 3-Kinase by Interleukin-13. An inhibitory signal for inducible nitric-oxide synthase expression in epithelial cell line HT-29.Journal of Biological Chemistry2721262612633Available: http://www.jbc.org/cgi/content/abstract/272/19/12626. Available: http://www.jbc.org/cgi/content/abstract/272/19/12626.

33. Freeman S, Post FA, Bekker L-G, Harbacheuski R, Steyn LM, et al. (2006) Mycobacterium tuberculosis H37Ra and H37Rv differential growth and cytokine/chemokine induction in murine macrophages in vitro. Journal of interferon & cytokine research: the official journal of the International Society for Interferon and Cytokine Research 26: 27–33.S. FreemanFA PostL-G BekkerR. HarbacheuskiLM Steyn2006Mycobacterium tuberculosis H37Ra and H37Rv differential growth and cytokine/chemokine induction in murine macrophages in vitro.Journal of interferon & cytokine research: the official journal of the International Society for Interferon and Cytokine Research262733Available: http://www.liebertonline.com/doi/abs/10.1089/jir.2006.26.27. Available: http://www.liebertonline.com/doi/abs/10.1089/jir.2006.26.27.

34. Dunphy KY, Senaratne RH, Masuzawa M, Kendall LV, Riley LW (2010) Attenuation of Mycobacterium tuberculosis functionally disrupted in a fatty acyl-coenzyme A synthetase gene fadD5. The Journal of infectious diseases 201: 1232–9.KY DunphyRH SenaratneM. MasuzawaLV KendallLW Riley2010Attenuation of Mycobacterium tuberculosis functionally disrupted in a fatty acyl-coenzyme A synthetase gene fadD5.The Journal of infectious diseases20112329Available: http://www.ncbi.nlm.nih.gov/pubmed/20214478. Available: http://www.ncbi.nlm.nih.gov/pubmed/20214478.

35. Trivedi OA, Arora P, Sridharan V, Tickoo R, Mohanty D, et al. (2004) Enzymic activation and transfer of fatty acids as acyl-adenylates in mycobacteria. Nature 428: 441–5.OA TrivediP. AroraV. SridharanR. TickooD. Mohanty2004Enzymic activation and transfer of fatty acids as acyl-adenylates in mycobacteria.Nature4284415Available: http://dx.doi.org/10.1038/nature02384. Available: http://dx.doi.org/10.1038/nature02384.

36. Dirusso CC, Black PN (2004) Bacterial long chain fatty acid transport: gateway to a fatty acid-responsive signaling system. The Journal of biological chemistry 279: 49563–6.CC DirussoPN Black2004Bacterial long chain fatty acid transport: gateway to a fatty acid-responsive signaling system.The Journal of biological chemistry279495636Available: http://www.ncbi.nlm.nih.gov/pubmed/15347640. Available: http://www.ncbi.nlm.nih.gov/pubmed/15347640.

37. Dubnau E, Chan J, Raynaud C, Mohan VP, Lanéelle M-A, et al. (2002) Oxygenated mycolic acids are necessary for virulence of Mycobacterium tuberculosis in mice. Molecular Microbiology 36: 630–637.E. DubnauJ. ChanC. RaynaudVP MohanM-A Lanéelle2002Oxygenated mycolic acids are necessary for virulence of Mycobacterium tuberculosis in mice.Molecular Microbiology36630637Available: http://doi.wiley.com/10.1046/j.1365-2958.2000.01882.x. Available: http://doi.wiley.com/10.1046/j.1365-2958.2000.01882.x.

38. Korf JE, Pynaert G, Tournoy K, Boonefaes T, Oosterhout AVan, et al. (2006) Macrophage reprogramming by mycolic acid promotes a tolerogenic response in experimental asthma. American journal of respiratory and critical care medicine 174: 152–60.JE KorfG. PynaertK. TournoyT. BoonefaesAVan Oosterhout2006Macrophage reprogramming by mycolic acid promotes a tolerogenic response in experimental asthma.American journal of respiratory and critical care medicine17415260Available: http://ajrccm.atsjournals.org/cgi/content/full/174/2/152. Available: http://ajrccm.atsjournals.org/cgi/content/full/174/2/152.

39. Yuan Y, Zhu Y, Crane DD, Barry CE (1998) The effect of oxygenated mycolic acid composition on cell wall function and macrophage growth in Mycobacterium tuberculosis. Molecular Microbiology 29: 1449–1458.Y. YuanY. ZhuDD CraneCE Barry1998The effect of oxygenated mycolic acid composition on cell wall function and macrophage growth in Mycobacterium tuberculosis.Molecular Microbiology2914491458Available: http://doi.wiley.com/10.1046/j.1365-2958.1998.01026.x. Available: http://doi.wiley.com/10.1046/j.1365-2958.1998.01026.x.

40. Bouchonnet F (2002) Alpha/Beta Interferon Impairs the Ability of Human Macrophages To Control Growth of Mycobacterium bovis BCG. Infection and Immunity 70: 3020–3025.F. Bouchonnet2002Alpha/Beta Interferon Impairs the Ability of Human Macrophages To Control Growth of Mycobacterium bovis BCG.Infection and Immunity7030203025Available: http://iai.asm.org/cgi/content/abstract/70/6/3020. Available: http://iai.asm.org/cgi/content/abstract/70/6/3020.

41. Manca C, Tsenova L, Bergtold A, Freeman S, Tovey M, et al. (2001) Virulence of a Mycobacterium tuberculosis clinical isolate in mice is determined by failure to induce Th1 type immunity and is associated with induction of IFN-alpha/beta. Proceedings of the National Academy of Sciences of the United States of America 98: 5752–7.C. MancaL. TsenovaA. BergtoldS. FreemanM. Tovey2001Virulence of a Mycobacterium tuberculosis clinical isolate in mice is determined by failure to induce Th1 type immunity and is associated with induction of IFN-alpha/beta.Proceedings of the National Academy of Sciences of the United States of America9857527Available: http://www.pnas.org/cgi/content/abstract/98/10/5752. Available: http://www.pnas.org/cgi/content/abstract/98/10/5752.

42. Meyer-Wentrup F, Benitez-Ribas D, Tacken PJ, Punt CJA, Figdor CG, et al. (2008) Targeting DCIR on human plasmacytoid dendritic cells results in antigen presentation and inhibits IFN-alpha production. Blood 111: 4245–53.F. Meyer-WentrupD. Benitez-RibasPJ TackenCJA PuntCG Figdor2008Targeting DCIR on human plasmacytoid dendritic cells results in antigen presentation and inhibits IFN-alpha production.Blood111424553Available: http://bloodjournal.hematologylibrary.org/cgi/content/full/111/8/4245. Available: http://bloodjournal.hematologylibrary.org/cgi/content/full/111/8/4245.

43. Meyer-Wentrup F, Cambi A, Joosten B, Looman MW, Vries IJMde, et al. (2009) DCIR is endocytosed into human dendritic cells and inhibits TLR8-mediated cytokine production. Journal of leukocyte biology 85: 518–25.F. Meyer-WentrupA. CambiB. JoostenMW LoomanIJMde Vries2009DCIR is endocytosed into human dendritic cells and inhibits TLR8-mediated cytokine production.Journal of leukocyte biology8551825Available: http://www.jleukbio.org/cgi/content/abstract/85/3/518. Available: http://www.jleukbio.org/cgi/content/abstract/85/3/518.

44. Simmons DP, Canaday DH, Liu Y, Li Q, Huang A, et al. (2010) Mycobacterium tuberculosis and TLR2 Agonists Inhibit Induction of Type I IFN and Class I MHC Antigen Cross Processing by TLR9. Journal of immunology (Baltimore, Md.: 1950) 185: 2405–2415.DP SimmonsDH CanadayY. LiuQ. LiA. Huang2010Mycobacterium tuberculosis and TLR2 Agonists Inhibit Induction of Type I IFN and Class I MHC Antigen Cross Processing by TLR9.Journal of immunology (Baltimore, Md.: 1950)18524052415Available: http://www.jimmunol.org/cgi/content/abstract/185/4/2405. Available: http://www.jimmunol.org/cgi/content/abstract/185/4/2405.

45. Manca C, Reed MB, Freeman S, Mathema B, Kreiswirth B, et al. (2004) Differential Monocyte Activation Underlies Strain-Specific Mycobacterium tuberculosis Pathogenesis. Society 72: 5511–5514.C. MancaMB ReedS. FreemanB. MathemaB. Kreiswirth2004Differential Monocyte Activation Underlies Strain-Specific Mycobacterium tuberculosis Pathogenesis.Society7255115514

46. Porcheray F, Viaud S, Rimaniol AC, Léone C, Samah B, et al. (2005) Macrophage activation switching: an asset for the resolution of inflammation. Clinical and experimental immunology 142: 481–9.F. PorcherayS. ViaudAC RimaniolC. LéoneB. Samah2005Macrophage activation switching: an asset for the resolution of inflammation.Clinical and experimental immunology1424819Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1809537&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1809537&tool=pmcentrez&rendertype=abstract.

47. Howard AD, Zwilling BS (1999) Reactivation of tuberculosis is associated with a shift from type 1 to type 2 cytokines. Clinical and experimental immunology 115: 428–34.AD HowardBS Zwilling1999Reactivation of tuberculosis is associated with a shift from type 1 to type 2 cytokines.Clinical and experimental immunology11542834Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1905252&tool=pmcentrez&rendertype=abstract. Available: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1905252&tool=pmcentrez&rendertype=abstract.

48. Wassie L, Demissie A, Aseffa A, Abebe M, Yamuah L, et al. (2008) Ex vivo cytokine mRNA levels correlate with changing clinical status of ethiopian TB patients and their contacts over time. PLoS One 3: 1522.L. WassieA. DemissieA. AseffaM. AbebeL. Yamuah2008Ex vivo cytokine mRNA levels correlate with changing clinical status of ethiopian TB patients and their contacts over time.PLoS One31522Available: http://dx.plos.org/10.1371/journal.pone.0001522. Available: http://dx.plos.org/10.1371/journal.pone.0001522.