Mrna Biogenesis in Tuberculosis Evidence for Postinitiation

Evidence for Postinitiation Regulation of mRNA Biogenesis in Tuberculosis Hugh Salamon, Yaming Qiao, Jeff C. Cheng, Ken D. Yamaguchi, Patricia Soteropoulos, Michael Weiden, Maria This information is current as Laura Gennaro and Richard Pine of October 1, 2021. J Immunol 2013; 190:2747-2755; Prepublished online 1 February 2013; doi: 10.4049/jimmunol.1202185 http://www.jimmunol.org/content/190/6/2747 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2013/02/01/jimmunol.120218 Material 5.DC1

References This article cites 61 articles, 26 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/190/6/2747.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on October 1, 2021

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Evidence for Postinitiation Regulation of mRNA Biogenesis in Tuberculosis

Hugh Salamon,* Yaming Qiao,† Jeff C. Cheng,‡,1 Ken D. Yamaguchi,* Patricia Soteropoulos,† Michael Weiden,x Maria Laura Gennaro,†,{ and Richard Pine†,{

Mycobacterium tuberculosis infection alters macrophage gene expression and macrophage response to IFN-g, a critical host defense cytokine. However, regulation of these changes is poorly understood. We report discordance of changes in nascent transcript and total nuclear RNA abundance for the transcription factors STAT1 and IRF1, together with lack of effect on their RNA half-lives, in human THP-1 cells infected with M. tuberculosis and stimulated with IFN-g. The results indicate that negative postinitiation regulation of mRNA biogenesis limits the expression of these factors, which mediate host defense against M. tuberculosis through the cellular response to IFN-g. Consistent with the results for STAT1 and IRF1, transcriptome analysis reveals downregulation of postinitiation mRNA biogenesis processes and pathways by infection, with and without IFN-g stimulation. Clinical relevance for Downloaded from regulation of postinitiation mRNA biogenesis is demonstrated by studies of donor samples showing that postinitiation mRNA biogenesis pathways are repressed in latent tuberculosis infection compared with cured disease and in active tuberculosis compared with ongoing treatment or with latent tuberculosis. For active disease and latent infection donors from two populations (London, U.K., and The Gambia), each analyzed using a different platform, pathway-related gene expression differences were highly correlated, demonstrating substantial speciﬁcity in the effect. Collectively, the molecular and bioinformatic analyses point toward downregulation of postinitiation mRNA biogenesis pathways as a means by which M. tuberculosis infection limits expression of http://www.jimmunol.org/ immunologically essential transcription factors. Thus, negative regulation of postinitiation mRNA biogenesis can constrain the macrophage response to infection and overall host defense against tuberculosis. The Journal of Immunology, 2013, 190: 2747–2755.

orbidity and mortality from tuberculosis are extremely latent infection; reviewed in 5). Determining effects of M. tu- high, with ∼8 million new cases of active disease per berculosis infection on host macrophage gene expression and M year and ∼2 million deaths (1); in the absence of ef- relating those effects to differences in gene expression between fective treatment, mortality is ∼50% (2). Increased failure of anti- individuals who maintain latent tuberculosis infection (LTBI) and those who develop pulmonary tuberculosis (PTB) should facilitate

tuberculous chemotherapy (3, 4) with the emergence of multidrug by guest on October 1, 2021 and extensively drug-resistant strains of Mycobacterium tubercu- efforts to apply host immune pressure against tuberculosis. losis has added urgency to the goal of developing effective vac- The interaction between macrophages infected with M. tuber- cines and immunotherapies. Macrophages are the immune cells culosis and the host immune mediator IFN-g is a major determi- predominantly targeted by M. tuberculosis. Bacterial replication nant of host response to M. tuberculosis (6). The host transcription occurs in macrophages at two points in the immunologic life cycle factors STAT1 and IRF1 are essential mediators of the response of tuberculosis: during the innate immune response to infection to IFN-g and of host defense against M. tuberculosis (7–10). In (before adaptive immunity forces a transition to latent infection) humans, mutations in STAT1 confer susceptibility to normally and during reactivation (when adaptive immunity fails to maintain nonpathogenic mycobacterial infections (11, 12); in mice, a deficiency of STAT1 or IRF1 abolishes immune control of M. tuberculosis growth, which leads to a fatal fulminant infection *Knowledge Synthesis, Inc., Berkeley, CA 94716; †Public Health Research Institute Center, New Jersey Medical School, University of Medicine and Dentistry of New rather than a chronic illness with slow disease progression (13, Jersey, Newark, NJ 07103; ‡Department of Biological Sciences, New Jersey Institute x 14). The consequences of deficiencies in these transcription of Technology, Newark, NJ 07102; Division of Pulmonary and Critical Care Med- factors emphasize that their regulation is essential for an effec- icine, Department of Medicine, New York University School of Medicine, New York, NY 10016; and {Department of Medicine, New Jersey Medical School, University of tive host response to M. tuberculosis. Moreover, both are induced Medicine and Dentistry of New Jersey, Newark, NJ 07103 by M. tuberculosis infection and by IFN-g stimulation (15–21). 1Current address: Early Clinical Development Statistics, Merck Research Laborato- IFN-g induction of STAT1 and IRF1 and M. tuberculosis in- ries, Rahway, NJ. duction of IRF1 are attributable at least in part to increased Received for publication August 7, 2012. Accepted for publication January 1, 2013. transcription. However, little is known about whether mecha- This work was supported by National Institutes of Health Grants AI37877 and nisms other than regulation of transcription initiation control their HL68517 (to R.P.) and HL106788 (to M.L.G. and R.P.) and by the Foundation of the University of Medicine and Dentistry of New Jersey (to R.P.). expression, or any other transcriptome changes, with or without IFN-g stimulation in cells infected with M. tuberculosis. Address correspondence and reprint requests to Dr. Richard Pine, New Jersey Med- ical School–University of Medicine and Dentistry of New Jersey, 225 Warren Street, Postinitiation steps in mRNA biogenesis (59 end capping, elon- Newark, NJ 07103. E-mail: [email protected] gation, splicing, 39 end cleavage and polyadenylation) allow reg- The online version of this article contains supplemental material. ulation of gene expression in addition to control of transcription Abbreviations used in this article: CERNO, Coincident Extreme Ranks in Numerical initiation (22–25). Alternative splicing and polyadenylation can Observations; GO, gene ontology; LTBI, latent tuberculosis infection; PTB, pulmo- determine tissue-specific or signal-mediated levels of transcript nary tuberculosis. isoform expression (reviewed in Refs. 26, 27). For example, in Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 patients with chronic granulomatous disease, increased levels of www.jimmunol.org/cgi/doi/10.4049/jimmunol.1202185 2748 REGULATION OF mRNA BIOGENESIS IN TUBERCULOSIS functional NADPH (phagocyte) oxidase as a response to IFN-g Quantitative RT-PCR therapy result from mutations that alter CYBB gene exon usage Reverse transcription reactions, quantitative PCR reactions using molecular (28, 29). In other examples, stimulation of TLRs by bacteria- beacons for amplicon detection, and calculation of target abundance were derived ligands alters transcripts through effects on alternative performed as described previously for IRF1. The results were normalized to splicing and polyadenylation (30–36). With M. tuberculosis in- the level of GAPDH exon 9 for the respective samples (21). Mock reverse fection, alternatively spliced transcripts of IL12Rb are produced transcription samples for each RNA preparation demonstrated negligible DNA contamination. The specificity of assays for introns and exon junc- (37). Even without alternative mRNA processing, changing the tions was confirmed with genomic DNA and cDNA templates. Data are rate of a single processing event can control gene expression from four to six independent experiments; not all genes were assayed in level, as demonstrated for glucocorticoid-mediated repression of some experiments. RNA half-life was calculated as the average of values gonadotropin-releasing hormone expression through inhibition determined from exponential decay curves for individual experiments (n = 2–4). Error bars represent 6 SEM (for n . 2) or the range of values (for of pre-mRNA splicing (38). Thus, postinitiation regulation of n = 2). Statistical significance for data sets with n $ 3 was taken as p , mRNA biogenesis might be an important host response to M. 0.05 based on a two-tailed Student t test. tuberculosis infection. Transcriptome analysis In this study, we characterized expression of genes responsive to M. tuberculosis infection and IFN-g stimulation and analyzed Gene expression profiles for THP-1 cells were determined for four replicate transcriptome data to better understand the basis for their regulated experiments that each included all four experimental conditions (control, M. tuberculosis infection, IFN-g stimulation, infection and stimulation) using expression. Data from in vitro infection of THP-1 cells indicated Affymetrix Human Genome U133A 2.0 Gene Chips, following the ven- that negative postinitiation regulation of mRNA biogenesis, su- dor’s protocols for cRNA preparation, labeling, hybridization, and scan- perimposed on IFN-g–stimulated activation of transcription, limits ning. Quantification of probe expression levels was based on the “preferred increases in STAT1 and IRF1 gene expression. Analysis of tran- methods” of Choe et al. (44). This analysis included Microarray Suite 5 Downloaded from (Affymetrix) for background correction, global scale to 500 for probe-level scriptome data demonstrated that downregulation of postinitiation normalization, perfect match minus mismatch for hybridization specificity mRNA biogenesis pathways occurs with in vitro infection and correction, robust multiarray analysis’ median polish for gene expression distinguishes individuals who develop PTB from those who main- summary calculation, GAPDH expression for gene expression normaliza- tain LTBI. tion, and Loess normalization for intensity skew correction. The use of GAPDH as an internal standard for expression level normalization was em-

pirically validated based on 60 different segments of the GAPDH gene. The http://www.jimmunol.org/ Materials and Methods signal varied among the probes, but was comparable under all conditions for each one (data not shown). These data are publicly available (Gene Expression Cell growth and infections Omnibus accession number GSE17477, http://www.ncbi.nlm.nih.gov/geo/). All manipulations with viable M. tuberculosis were performed under bio- For gene ontology (GO) annotation analysis of gene expression, we first safety level 3 containment. M. tuberculosis TN913, a prevalent, drug- identified regulated genes based on statistically significant differential ex- sensitive clinical isolate of the C strain from the 1990–1994 New York pression in comparisons of gene expression between each perturbation and City tuberculosis outbreak (39), was obtained from the Public Health Re- uninfected, unstimulated cells. A t test with a dynamic threshold (Cyber-T) search Institute Tuberculosis Center. The human monocytic cell line THP-1 (45) was used to identify differentially expressed genes (p , 0.05 or p , was obtained from American Type Culture Collection. The bacteria and 0.01). A more stringent threshold for differential expression was set using cells were maintained, and THP-1 cells were differentiated and infected, as a permutation-based implementation to calculate false discovery rate (q , described previously (17). Differentiated THP-1 cells model human alve- 0.1). Fisher exact test was used to assess the statistical significance (p , by guest on October 1, 2021 olar macrophages, as judged by a variety of criteria (17, 20, 40, 41). Three 0.05) for the proportion of regulated genes having particular combinations days after infection, infected cells and parallel cultures of uninfected cells of annotations taken from the GO database (http://www.geneontology.org/) were left untreated or were treated for 2 h with IFN-g (Peprotech, Rocky (46) among downregulated genes in comparison with the overall proportion Hill, NJ) at 1 ng/ml. In some experiments, actinomycin D (Calbiochem, La of genes having those annotations among those probed. Jolla, CA) was then added to a concentration of 10 mg/ml, and cells were The transcriptomes of samples from clinical studies were obtained from harvested at various times thereafter to determine transcript half-life. The Gene Expression Omnibus—GEO, GSE19491 (47), GSE19439 (48), and titer of the inoculum on the day of infection and the presence of intracellular GSE11199 (49). The first two studies included comparison of whole blood bacteria on the day of harvest were confirmed by plating to determine CFUs. from LTBI and PTB donors. The third study included comparison of monocyte-derived macrophages from LTBI and cured PTB donors. The Cell fractionation and recovery of RNA publishing authors’ normalized expression values were used. For pathway analysis of data from in vitro infection and from the clinical All steps were performed at 0–4˚C. Cells were scraped from flasks and studies, the significance of gene sets defined by Reactome pathway annota- 3 collected by centrifugation at 400 g for 5 min. To isolate total cellular tions (http://www.reactome.org) (50, 51) was evaluated using the Coincident RNA, cell pellets were extracted immediately. Alternatively, nuclear and Extreme Ranks in Numerical Observations (CERNO) method (52). The cytoplasmic fractions were prepared as described previously (20). Cyto- CERNO test uses the rank order of significance for differential expression of plasmic RNA was extracted, and the nuclei were suspended in nuclear run- the genes in a pathway to determine significance for the pathway annotation on buffer containing 0.1% NP-40 and lacking nucleotide triphosphates. without a threshold on expression differences for individual genes. A nested Nuclei then were used in the nuclear run-on assay (below) or were further CERNO testing routine was applied (53). Benjamini-Hochberg false dis- purified by sedimentation through a cushion of 60% glycerol in the same covery rates (54) were calculated for the entire collection of gene sets, but for buffer (42) before extraction of RNA. RNA was recovered using TRI- each sample comparison separately. Eighteen postinitiation mRNA biogen- Reagent (Molecular Resource Center) as recommended by the vendor. esis Reactome gene sets were identified with at least three representative Nuclear run-on assay genes on the Affymetrix Human Genome U133A 2.0 and the Agilent-014850 Whole Human Genome Microarray. A one-sided Student t test was per- Nascent transcript abundance was determined as described previously (21). formed to test whether individual genes were downregulated. Statistical Signal from hybridization to pGem1 provided a negative control for spec- analysis methods not otherwise cited were performed in the R programming ificity and for background correction. Signal from hybridization to human environment (55). GAPDH provided a positive control and an internal standard for normalization. Probes for GAPDH, IRF1, ISG15, and STAT1 were described Results previously (15, 16, 20, 43). All genes are denoted by Human Genome Or- Differential induction of nascent transcripts and total nuclear ganization official gene symbols. Data are from six independent experiments. Quantification of nascent transcript abundance is shown as the RNA for both STAT1 and IRF1 average of fold induction relative to expression in uninfected, unstimulated Comparing induction at different points in the course of gene ex- cells. Fold induction is used to allow comparison among genes and among experimental conditions. Error bars for each perturbation (M. tuberculosis pression will reveal the presence of regulated steps. To set a baseline infection, IFN-g stimulation, or both) represent 6 SEM. Statistical signif- for identifying postinitiation regulation, we first measured the levels icance is taken as p , 0.05 based on a two-tailed Student t test. of nascent transcripts for STAT1 and IRF1, two central regulators of The Journal of Immunology 2749 response to M. tuberculosis infection and to IFN-g stimulation. M. of ISG15 because it is induced by M. tuberculosis, but not by IFN-g tuberculosis infection and IFN-g stimulation each induced ex- (15, 20). In contrast to STAT1 and IRF1, the level of ISG15 nascent pression of the two genes (Fig. 1A, 1B). When infected cells were transcripts was similar with and without IFN-g stimulation of in- also stimulated with IFN-g, the increases in nascent transcripts fected cells (Fig. 1A, 1B). These data demonstrate specific syner- for STAT1 and IRF1 were 2–3-fold greater than the sum of the gistic induction of STAT1 and IRF1 nascent transcripts with IFN-g responses to each perturbation alone; therefore, a synergistic re- stimulation of M. tuberculosis-infected cells. sponse occurs. To test for specificity in the effects of M. tubercu- If regulation were only at the level of transcription initiation, the losis infection on the response to IFN-g, we measured transcription synergistic induction of nascent transcripts for STAT1 and IRF1 would lead to comparable increases in the nuclear and cytoplasmic pools of the corresponding mRNAs. This was the case with M. tuberculosis infection alone, because the levels of STAT1, IRF1, and ISG15 increased in total nuclear RNA as much as in nascent transcripts (compare Fig. 1B to 1C; the ratio is shown in Fig. 1D). IFN- g–mediated induction of STAT1 in total nuclear RNA and in nascent transcripts was also similar. However, induction of IRF1 in total nuclear RNA was 2.5-fold greater than in nascent transcripts, indicating that induction of nascent transcripts and additional positive regulation determine the expression of IRF1 in response to IFN-g stimulation. In sharp contrast, when cells were infected with M. tuberculosis and stimulated with IFN-g, the ratio of the total nuclear Downloaded from RNA to nascent transcript for both STAT1 and IRF1 was one fourth the ratio in uninfected cells stimulated with IFN-g (Fig. 1B–D). ISG15 showed less of an effect. These results suggest that expression of STAT1 and IRF1 is subject to gene-specific negative regulation that limits their induction by IFN-g in infected cells. Comparing induction of polyA+ mRNA to nascent transcripts also revealed the http://www.jimmunol.org/ limit on expression of STAT1 and IRF1 in M. tuberculosis–infected cells stimulated with IFN-g (data not shown). The reduced induction of STAT1 and IRF1 associated with IFN-g stimulation of infected cells was specific for M. tuberculosis, because it was not observed in cells infected with the nonpathogenic Mycobacterium bovis BCG (Supplemental Fig. 1). Thus, in cells infected with M. tuberculosis, regulation in addition to initiation of transcription controls IFN-g– stimulated STAT1 and IRF1 expression. by guest on October 1, 2021 Postinitiation regulation of response to M. tuberculosis and IFN-g The differentials between induction in total nuclear RNA and in nascent transcripts for both STAT1 and IRF1 seen with IFN-g stimulation of M. tuberculosis–infected cells could result from increased RNA turnover, negative postinitiation regulation of mRNA biogenesis, or both. We first addressed the turnover hypothesis by measuring the half-life of exon and intron sequences for STAT1 and IRF1 in total nuclear RNA. We found no significant difference in turnover under any condition (Fig. 2), suggesting that neither

FIGURE 1. Effects of M. tuberculosis infection and IFN-g stimulation on nascent and total nuclear RNA. THP-1 cells were differentiated and infected with M. tuberculosis or stimulated with IFNg, or both. Next, nascent transcripts were measured using the nuclear run-on assay, or total nuclear RNA was extracted and quantified by quantitative RT-PCR. (A) The hybridization results for nuclear run-on assays were imaged and quantified using a phosphorimager. The figure is a composite of noncon- tiguous portions from a single phosphorimager exposure that included the membranes for all four conditions in a representative experiment. (B) Nascent transcript measurement is shown as average fold induction 6 SEM for six replicate experiments. (C) Total nuclear RNA abundance is shown as average fold induction 6 SEM for four to six replicate experiments as for (B). (D) The ratios of the abundance of total nuclear RNA FIGURE 2. Effects of M. tuberculosis infection and IFN-g stimulation relative to the level of nascent transcripts were calculated from the aver- on STAT1 and IRF1 exon and intron half-life in total nuclear RNA. STAT1 ages for each. Error bars represent 6 SEM. Statistically significant dif- and IRF1 transcripts were assayed, and the half-life of each target region ferences (p , 0.05) are indicated compared with control (†), compared was calculated. The averages are shown with the range or SEM (see with M. tuberculosis (*), and compared with IFN-g (^). Materials and Methods). 2750 REGULATION OF mRNA BIOGENESIS IN TUBERCULOSIS degradation nor nuclear export were affected. We next determined regulated splice would alter the ratio. For STAT1, the ratios of the that the induction of poly-A+ nuclear RNA was proportional to the spliced exon 2:3 junction compared with intron 2 and of the spliced induction of total nuclear RNA for both STAT1 and IRF1 (Sup- exon 22:23 junction compared with intron 22 were similar in plemental Fig. 2), suggesting that cleavage and polyadenylation were uninfected, unstimulated cells, and they decreased similarly with not regulated for either gene. Moreover, the half-lives of the total M. tuberculosis infection, IFN-g stimulation, and stimulation of cellular poly-A+ RNAs were comparable in uninfected and infected infected cells (Fig. 4A, 4B). In contrast, analysis of IRF1 splices cells with and without IFN-g stimulation for each of the two genes showed that the ratios between the spliced exon 1:2, 4:5, and 9:10 (Supplemental Fig. 3), indicating that the overall turnover of those transcripts is also not regulated in response to these perturbations. Because regulation of neither turnover nor final maturation explains differences between induction of nascent transcripts and total nuclear RNA, we interpret the data as indicating that STAT1 and IRF1 expression is limited at a postinitiation step in mRNA biogenesis (i.e., before a mature mRNA transcript is produced). Control of transcript elongation is one mechanism for postinitiation regulation of mRNA biogenesis. Therefore, we considered whether the observed limited induction of STAT1 and IRF1 in M. tuberculosis–infected cells stimulated with IFN-g was caused by downregulation of elongation. When we assayed introns throughout the STAT1 and in IRF1 transcripts, each target ex- Downloaded from hibited the induction caused by M. tuberculosis infection, IFN-g stimulation, or both, that was shown by the exon measurements (compare Fig. 1C and Fig. 3, left panel). Little or no change occurred in the ratio of STAT1 intron 2 to intron 22 or in the ratio of IRF1 intron 1 to intron 9 under any condition (Fig. 3, right panel). Thus, regulated elongation contributes to neither positive postini- http://www.jimmunol.org/ tiation regulation of IRF1 by IFN-g nor negative postinitiation regulation of STAT1 and IRF1 in cells infected with M. tuberculosis and stimulated with IFN-g. To assess the possibility that regulated splicing accounts for the observed differences between induction of nascent and total nuclear RNA for STAT1 and for IRF1, we measured the abundance of exon-exon junctions and of the respective introns for both transcripts. Because the ratio of spliced to unspliced transcripts reflects by guest on October 1, 2021 the rate of each splice, a change in gene expression owing to a

FIGURE 4. Effects of M. tuberculosis infection and IFN-g stimulation on spliced and unspliced STAT1 and IRF1 transcripts. Exon junction and intron sequences in total nuclear STAT1 and IRF1 transcripts were FIGURE 3. Effects of M. tuberculosis infection and IFN-g stimulation assayed. The average fold-induction of each target region is shown. Error on STAT1 and IRF1 transcript elongation. Intron sequences in total nuclear bars represent 6 SEM for each average and for the ratios. (A) STAT1 exon STAT1 and IRF1 transcripts were assayed. The average fold-induction of 2-3 junction and intron 2. (B) STAT1 exon 22-23 junction and intron 22. each target region is shown. Error bars represent 6 SEM for each average and (C) IRF1 exon 1-2 junction and intron 1. (D) IRF1 exon 4-5 junction and for the ratios. (A) STAT1 intron 2 and intron 22. (B)IRF1intron1andintron9. intron 4. (E) IRF1 exon 9-10 junction and intron 9. Statistically signiﬁcant Statistically signiﬁcant differences (p , 0.05) are indicated compared with differences (p , 0.05) are indicated compared with control (†), compared control (†), compared with M. tuberculosis (*), and compared with IFN-g (^). with M. tuberculosis (*), and compared with IFN-g (^). The Journal of Immunology 2751

stimulated with IFN-g suggests that the processes and pathways involved may be downregulated. To test this possibility, we analyzed gene expression profiles (Gene Expression Omnibus accession number GSE17477) for differentiated THP-1 cells that were uninfected or infected with M. tuberculosis with and without IFN-g stimulation. We found that among all regulated genes the number induced was much greater than the number repressed under all conditions tested (Fig. 5A). The same result was obtained in IFN-g–treated cells with genes annotated for mRNA processing in GO (46) (Fig 5B). In contrast, more of those genes were repressed than induced in cells that were infected with M. tuberculosis with and without IFN-g stimulation (Fig. 5B). Thus, the effect of M. tuberculosis infection was different for expression of genes annotated for mRNA processing than for the overall transcriptome. Moreover, mRNA processing genes were overrep- FIGURE 5. M. tuberculosis infection and IFN-g stimulation of infected resented among all the downregulated genes when various strin- cells preferentially downregulates genes annotated for mRNA processing. gency criteria were used for defining regulated genes (see The percentage of (A) all genes probed and (B) genes having a GO an- Materials and Methods). Among the 14 mRNA processing genes notation for mRNA processing or a descendant that was upregulated or deemed regulated under the more stringent criteria, four were downregulated is shown for THP-1 cells that were infected with M. tu- responsive to infection alone, and the other 10 were downregu- Downloaded from berculosis, stimulated with IFN-g, and infected then stimulated, as indi- lated only in cells that were infected and stimulated with IFN-g cated. Among downregulated genes defined by a two-sided Cyber-T p , (Table I). Thus, the transcriptome analysis demonstrated a statis- 0.05 for cells infected with M. tuberculosis or infected with M. tuberculosis then stimulated with IFN-g, each compared with uninfected cells, tically significant downregulation of mRNA processing genes genes annotated for mRNA processing or a descendant were overrepre- consistent with limited induction of STAT1 and IRF1 expression sented (Fisher exact test, p , 0.0001). owing to postinitiation downregulation when cells infected with M. tuberculosis are stimulated with IFN-g. http://www.jimmunol.org/ junctions and the respective introns differed from each other in Analogous results were obtained when the transcriptome results uninfected, unstimulated cells (Fig. 4C–E). This result suggests were analyzed using a different annotation knowledge base focused that individual splices occur at different rates. The characteristic on pathways (Reactome; http://www.reactome.org) (50, 51) and ratio of each splice was changed little by M. tuberculosis infection a different statistical approach (CERNO, see Materials and alone, but each ratio increased in uninfected and infected cells Methods) (52, 53) (Fig. 6 and Supplemental Fig. 4). Of 18 Re- stimulated with IFN-g. The increase in the level of spliced relative actome pathways pertaining to postinitiation mRNA biogenesis, to unspliced transcripts could contribute to the upregulation of 16 were significantly downregulated in infected cells (Fig. 6A and Supplemental Fig. 4A). Significant downregulation occurred for IRF1 expression by IFN-g stimulation of uninfected and infected by guest on October 1, 2021 cells. However, the data do not support a role for regulated splic- 17 of these pathways in cells that were infected and stimulated ing in the negative postinitiation regulation that limits induction of with IFN-g compared with cells that were only IFN-g stimulated STAT1 and IRF1 in infected cells stimulated with IFN-g. (Fig. 6B and Supplemental Fig. 4B). The analyses of transcriptome databases based on annotations for process and for path- Regulated expression of genes that mediate postinitiation ways, taken together with the results for expression of STAT1 and mRNA biogenesis IRF1, identify negative postinitiation regulation of mRNA biogen- The finding that induction of STAT1 and IRF1 is limited by esis as a factor in an essential immunologic response to IFN-g negative postinitiation regulation in M. tuberculosis–infected cells stimulation of M. tuberculosis–infected cells.

Table I. mRNA processing genes downregulated by M. tuberculosis infection with or without IFN-g stimulation

Gene Symbola Regulationb Biologic Processc HNRPA0 M. tuberculosis 6 IFN-g Nuclear mRNA splicing, via spliceosome LSM5 M. tuberculosis 6 IFN-g Spliceosome assembly NUDT21 M. tuberculosis 6 IFN-g mRNA 39 end formation SFRS11 M. tuberculosis 6 IFN-g mRNA splicing SNRNP25 M. tuberculosis + IFN-g mRNA splicing CPSF1 M. tuberculosis + IFN-g mRNA splicing, mRNA 39 end formation GEMIN4 M. tuberculosis + IFN-g Spliceosome assembly LSM4 M. tuberculosis + IFN-g Spliceosome assembly NHP2L1 M. tuberculosis + IFN-g Spliceosome assembly RBM8A M. tuberculosis + IFN-g Exon junction complex formation SF1 M. tuberculosis + IFN-g Pre-mRNA 39 splice site recognition, spliceosome assembly SFRS14 M. tuberculosis + IFN-g mRNA splicing SNRPN M. tuberculosis + IFN-g snRNP assembly TARDBP M. tuberculosis + IFN-g Alternative mRNA splicing aHuman Genome Organization Gene Nomenclature Committee official symbol. bPerturbations that caused downregulation: M. tuberculosis 6 IFN-g (downregulated by infection with or without IFN-g stimulation); M. tuberculosis + IFN-g (downregulated only by infection and IFN-g stimulation). cBiological process annotations from gene ontology. 2752 REGULATION OF mRNA BIOGENESIS IN TUBERCULOSIS Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021 FIGURE 6. Postinitiation mRNA biogenesis Reactome pathways and genes are downregulated by M. tuberculosis infection of THP-1 cells. Eighteen pathways were tested. The pathways shown are significantly different for the indicated comparisons (A, B) based on a false discovery rate criterion of 0.05 for the CERNO test results. The pathway tests were based on the rank order of differential expression for all genes in a pathway (among all transcript measurements of all named genes targeted by the gene expression platform). The top matrix depicts the membership of genes (columns) in pathways (rows) with red squares. The rows and columns of the top matrix are sorted to bring together similar pathway membership patterns. Genes shown are members of one or more pathways that exhibited differential expression for the indicated comparison (one-sided Student t test to assess gene downregulation; p , 0.05 unless noted otherwise). The lower matrix is a heat map of expression for each of the significantly regulated genes in each of the samples that were compared. The heat map shows gradations from higher to lower expression as yellow to blue. The decrease in expression (blue) is evident for infected cells. Gene symbols are shown with the NCBI Gene IDs. (A) Sixteen pathways were downregulated when comparing infected THP-1 cells with uninfected cells. (B) Seventeen pathways were downregulated when comparing THP-1 cells infected with M. tuberculosis and stimulated with IFN-g to uninfected, IFN-g– stimulated THP-1 cells.

Pulmonary tuberculosis is associated with downregulation an effect of in vitro infection is also identified as a difference of postinitiation mRNA biogenesis pathways associated with infection/disease status in vivo. To investigate the clinical relevance of the above results from We next considered the relationship between the expression of in vitro infection of macrophage-like THP1 cells, we analyzed individual genes and the disease-stage–specific effects on post- transcriptome data obtained from clinical samples. In two studies initiation mRNA biogenesis pathways. Although the data from the comparing whole blood from LTBI and PTB donors (47, 48), two studies of PTB and LTBI donors come from different patients the analysis revealed that PTB was associated with significant in different geographic regions measured on different gene ex- downregulation of postinitiation mRNA biogenesis pathways pression platforms (Affymetrix versus Agilent), differential ex- (Supplemental Fig. 4C, 4D). Moreover, PTB donors also exhibited pression of individual transcripts was significantly similar in the this downregulation in longitudinal comparisons with gene ex- two studies (Fig. 7). The significance of the correlation provides pression data from samples collected 2 and 12 mo after initiation evidence that PTB is associated with specific changes in the ex- of antituberculous treatment (Supplemental Fig. 4E, 4F). In a pression of particular genes that mediate postinitiation mRNA study comparing monocyte-derived macrophages from LTBI biogenesis. Moreover, querying for mRNA postinitiation biogen- donors and cured PTB donors (49), significant downregulation for esis genes differentially expressed at a nominal p , 0.05 level 10 of the 18 postinitiation mRNA biogenesis pathways was as- revealed nine genes common across sample comparisons in four sociated with LTBI (Supplemental Fig. 4G). Thus, the negative independent studies involving three distinct biologic systems postinitiation regulation of mRNA biogenesis described above as (THP-1 cells infected in vitro, two studies of whole blood com- The Journal of Immunology 2753

negative postinitiation regulation in vitro limits the expression of these genes induced by IFN-g stimulation of infected cells. In agreement with the molecular data interpretation, analysis of biologic process and pathway annotations identified significant effects on genes annotated for postinitiation mRNA biogenesis among genes downregulated by M. tuberculosis infection with or without IFN-g stimulation in vitro. Moreover, analysis of publicly available transcriptome data from clinical studies identified reduced expression of postinitiation mRNA biogenesis pathways with PTB compared with LTBI, PTB compared with treated PTB, and LTBI compared with cured PTB. These data demonstrate that M. tuberculosis infection licenses novel responses to IFN-g that limit induction of STAT1 and IRF1 through negative postinitiation regulation. This limit to induction constitutes a new constraint on IFN-g–induced gene expression in macrophages. Moreover, the limit on induction of STAT1 and IRF1, and perhaps other genes, may be associated with clinically relevant host response to M. tuberculosis, particularly PTB, inasmuch as postinitiation mRNA biogenesis pathways are downregulated in PTB compared with LTBI or to treated PTB, and in LTBI compared with cured PTB. Downloaded from The results from both in vitro infection and clinical samples FIGURE 7. Two comparisons of PTB to LBTI establish specificity in help elucidate the consequences of M. tuberculosis infection for the differential expression of postinitiation mRNA biogenesis genes. Transcript measurements from two independent studies in two geographic postinitiation regulation of mRNA biogenesis. For in vitro in- regions that implemented different gene expression microarray platforms fection, the comparison of different transcript pools for STAT1, were compared at the level of differential expression. Differential ex- IRF1 and several other immune response genes (e.g., CCL2, pression was measured as the difference in mean expression between PTB CXCL10, FCGR1, and MX1; Y. Qiao and R. Pine, unpublished http://www.jimmunol.org/ and LTBI. For the data from each study, the difference in mean expression observations) provided evidence that negative postinitiation re- was normalized to the respective SE in the estimate of the difference to gulation does not alter expression uniformly, because these provide a T-statistic for each probe or probe set. Positive values indicate genes exhibited various ratios for induction in total nuclear RNA higher expression in PTB disease relative to LTBI. T-statistics calculated relative to nascent transcripts. Moreover, the transcriptome an- for differential expression of individual genes in each study demonstrated alyses indicated the involvement of a specific subset of signifi- correlation of the direction and magnitude of differential expression of the cantly downregulated genes associated with postinitiation mRNA same genes measured in the two studies. The linear regression trend (black line) is highly significant (p , 1 3 10212). The significance provides biogenesis. Nine genes were found common to the in vitro and in vivo results, and the two comparisons of PTB to LTBI tran- evidence that the pathway level effect arises from particular changes in by guest on October 1, 2021 expression of specific differentially regulated genes. scriptomes from blood demonstrated similar extents of differential expression for individual transcripts. Thus, the consequences of M. tuberculosis infection for postinitiation regulation of mRNA paring LTBI and PTB donors from different regions using dif- biogenesis are gene specific. ferent platforms, and monocyte-derived macrophages from LTBI The negative postinitiation regulation that limits STAT1 and IRF1 and cured donors): RNPS1, HNRNPA0, HNRNPUL1, COBRA1, expression is not merely an inhibition of a positive response, for at GTF2H4, CPSF1, NHP2L1, SNRNP70, TCEB1. The existence of least two reasons. First, IFN-g stimulation caused an increase in the a core set of regulated genes and the correlation for differential ratio of spliced to unspliced IRF1 nuclear RNA that was similar in gene expression in data from independent studies of PTB and uninfected and infected cells, indicating that positive postinitiation LTBI donors provide strong support for the clinical relevance regulation of IRF1 by IFN-g still occurs in cells infected with M. of decreased gene expression related to postinitiation mRNA tuberculosis. The positive effect could be related to the presence biogenesis. of exonic splicing enhancer sequences in the IRF1 transcript (J.C. Cheng and R. Pine, unpublished observations), strongly Discussion suggesting that IFN-g mediates posttranslational regulation of When M. tuberculosis infects macrophages, complex changes SR family proteins that act at such sites. Second, IFN-g stimu- occur in host gene expression. Infection increases host production lation did not cause positive postinitiation regulation of STAT1 of IFN-g, which then alters macrophage gene expression. Infec- expression, yet STAT1 is subject to the negative regulation. tion also affects the macrophage response to IFN-g. To understand Taken together, the results reveal the simultaneous occurrence of the consequences for regulation of host cell gene expression, we a positive postinitiation response to IFN-g and a distinct negative characterized the induction of STAT1 and IRF1, transcription postinitiation response for these two genes in cells that are also factors that are key to the cellular IFN-g response and to host infected with M. tuberculosis. defense against the pathogen, and we analyzed transcriptome data To our knowledge, the case reported above for STAT1 and IRF1 obtained from in vitro and in vivo infection. These studies re- is the first example for the immune response in which downreg- vealed negative postinitiation regulation of mRNA biogenesis as ulation of genes in postinitiation mRNA biogenesis pathways is a novel way in which M. tuberculosis alters macrophage gene associated with negative postinitiation regulation of target gene expression. With IFN-g stimulation of cells infected with M. tu- expression. A parallel occurs in the neuroendocrine system, where berculosis, but not with IFN-g stimulation alone or with infection a glucocorticoid-mediated decrease in expression of the mRNA alone, less induction of STAT1 and IRF1 transcripts occurred in processing factor NOVA1 decreases the splicing of gonadotropin- total nuclear RNA than in nascent RNA, and no difference oc- releasing hormone pre-mRNA to reduce the level of the mature curred in transcript turnover. These results strongly suggest that mRNA in the hypothalamus (38). The negative postinitiation 2754 REGULATION OF mRNA BIOGENESIS IN TUBERCULOSIS regulation identified by our study is layered over synergistic 9. Kamijo, R., H. Harada, T. Matsuyama, M. Bosland, J. Gerecitano, D. Shapiro, J. Le, S. I. Koh, T. Kimura, S. J. Green, et al. 1994. Requirement for transcription transcriptional activation of STAT1 and IRF1 expression as a re- factor IRF-1 in NO synthase induction in macrophages. Science 263: 1612–1615. sponse to M. tuberculosis infection and IFN-g stimulation. Lim- 10. Martin, E., C. Nathan, and Q.-W. Xie. 1994. Role of interferon regulatory factor iting the consequences of synergistic transcriptional activation can 1 in induction of nitric oxide synthase. J. Exp. Med. 180: 977–984. 11. Chapgier, A., R. F. Wynn, E. Jouanguy, O. Filipe-Santos, S. Zhang, J. Feinberg, account for the comparable induction of IRF1 mRNA in unin- K. Hawkins, J. L. Casanova, and P. D. Arkwright. 2006. Human complete Stat-1 fected and M. tuberculosis–infected cells stimulated with IFN-g deficiency is associated with defective type I and II IFN responses in vitro but (21, 56). Many studies have examined changes in macrophage immunity to some low virulence viruses in vivo. J. Immunol. 176: 5078–5083. 12. Dupuis, S., E. Jouanguy, S. Al-Hajjar, C. Fieschi, I. Z. Al-Mohsen, S. Al- gene expression caused by M. tuberculosis and other mycobacteria Jumaah, K. Yang, A. Chapgier, C. Eidenschenk, P. Eid, et al. 2003. Impaired (17, 19, 20, 56–64). Effects of infection were described for RNA response to interferon-alpha/beta and lethal viral disease in human STAT1 de- stability, total RNA abundance, nascent transcript levels, promoter ficiency. Nat. Genet. 33: 388–391. 13. Sugawara, I., H. Yamada, and S. Mizuno. 2004. STAT1 knockout mice are activity, and epigenetic modifications, but changes in nascent and highly susceptible to pulmonary mycobacterial infection. Tohoku J. Exp. Med. total transcript abundance were not compared. In this study, com- 202: 41–50. parison of target gene expression at different levels, such as total 14. Yamada, H., S. Mizuno, and I. Sugawara. 2002. Interferon regulatory factor 1 in mycobacterial infection. Microbiol. Immunol. 46: 751–760. nuclear RNA relative to nascent transcripts or spliced relative to 15. Pine, R., T. Decker, D. S. Kessler, D. E. Levy, and J. E. Darnell, Jr. 1990. Pu- unspliced, allowed detection of postinitiation regulation. rification and cloning of interferon-stimulated gene factor 2 (ISGF2): ISGF2 (IRF-1) can bind to the promoters of both beta interferon- and interferon- This report opens new avenues to investigate the ability of M. stimulated genes but is not a primary transcriptional activator of either. Mol. tuberculosis to persist or grow despite ongoing macrophage re- Cell. Biol. 10: 2448–2457. sponse to IFN-g. Future studies will further define the mechanisms 16. Pine, R., A. Canova, and C. Schindler. 1994. Tyrosine phosphorylated p91 binds to a single element in the ISGF2/IRF-1 promoter to mediate induction by IFN a that affect host defense by limiting expression of STAT1 and and IFN g, and is likely to autoregulate the p91 gene. EMBO J. 13: 158–167. Downloaded from IRF1. For example, detailed characterization of RNA polymerase 17. Weiden, M., N. Tanaka, Y. Qiao, B. Y. Zhao, Y. Honda, K. Nakata, A. Canova, II distribution on the STAT1 and IRF1 genes might discover local D. E. Levy, W. N. Rom, and R. Pine. 2000. Differentiation of monocytes to macrophages switches the Mycobacterium tuberculosis effect on HIV-1 repli- fluctuations in elongation that were not discovered by comparing cation from stimulation to inhibition: modulation of interferon response and the abundance of one upstream and one downstream intron in total CCAAT/enhancer binding protein b expression. J. Immunol. 165: 2028–2039. nuclear RNA for each transcript. Regulation of 39 transcript 18. Ting, L. M., A. C. Kim, A. Cattamanchi, and J. D. Ernst. 1999. Mycobacterium tuberculosis inhibits IFN-gamma transcriptional responses without inhibiting cleavage and polyadenylation also needs to be addressed. While activation of STAT1. J. Immunol. 163: 3898–3906. characterizing mechanisms for the newly discovered negative 19. Qiao, Y., S. Prabhakar, E. M. Coccia, M. Weiden, A. Canova, E. Giacomini, and http://www.jimmunol.org/ R. Pine. 2002. Host defense responses to infection by Mycobacterium tubercu- regulation that limits macrophage response to IFN-g, it will be losis. Induction of IRF-1 and a serine protease inhibitor. J. Biol. Chem. 277: important to determine whether manipulating mRNA processing 22377–22385. can restore an unconstrained macrophage response to IFN-g and 20. Prabhakar, S., Y. Qiao, Y. Hoshino, M. Weiden, A. Canova, E. Giacomini, E. Coccia, and R. Pine. 2003. Inhibition of response to alpha interferon by thereby impede M. tuberculosis infection. Moreover, the discovery Mycobacterium tuberculosis. Infect. Immun. 71: 2487–2497. that PTB downregulates postinitiation mRNA biogenesis path- 21. Qiao, Y., S. Prabhakar, A. Canova, Y. Hoshino, M. Weiden, and R. Pine. 2004. ways indicates a need to better understand how that effect influ- Posttranscriptional inhibition of gene expression by Mycobacterium tuberculosis offsets transcriptional synergism with IFN-gamma and posttranscriptional up- ences the course of infection. regulation by IFN-gamma. J. Immunol. 172: 2935–2943. 22. Topisirovic, I., Y. V. Svitkin, N. Sonenberg, and A. J. Shatkin. 2011. Cap and cap-binding proteins in the control of gene expression. Wiley Interdiscip Rev by guest on October 1, 2021 Acknowledgments RNA 2: 277–298. We thank Karl Drlica for critically reading the manuscript, Anthony Galante, 23. Weake, V. M., and J. L. Workman. 2010. Inducible gene expression: diverse regulatory mechanisms. Nat. Rev. Genet. 11: 426–437. Olga Tkachenko, and Antony Canova for technical assistance, Yoshihiko 24. de Nadal, E., G. Ammerer, and F. Posas. 2011. Controlling gene expression in Hoshino for assistance with nuclear run-on assays, and Barry Kreiswirth response to stress. Nat. Rev. Genet. 12: 833–845. for providing M. tuberculosis TN913. 25. Sims, R. J., III, R. Belotserkovskaya, and D. Reinberg. 2004. Elongation by RNA polymerase II: the short and long of it. Genes Dev. 18: 2437–2468. 26. Shin, C., and J. L. Manley. 2004. Cell signalling and the control of pre-mRNA Disclosures splicing. Nat. Rev. Mol. Cell Biol. 5: 727–738. 27. Lutz, C. S. 2008. Alternative polyadenylation: a twist on mRNA 39 end for- The authors have no financial conflicts of interest. mation. ACS Chem. Biol. 3: 609–617. 28. Condino-Neto, A., and P. E. Newburger. 2000. Interferon-gamma improves splicing efficiency of CYBB gene transcripts in an interferon-responsive variant References of chronic granulomatous disease due to a splice site consensus region mutation. Blood 95: 3548–3554. 1. Corbett, E. L., C. J. Watt, N. Walker, D. Maher, B. G. Williams, M. C. Raviglione, 29. Ishibashi, F., T. Mizukami, S. Kanegasaki, L. Motoda, R. Kakinuma, F. Endo, and C. Dye. 2003. The growing burden of tuberculosis: global trends and inter- and H. Nunoi. 2001. Improved superoxide-generating ability by interferon g due actions with the HIV epidemic. Arch. Intern. Med. 163: 1009–1021. to splicing pattern change of transcripts in neutrophils from patients with a splice 2. Enarson, D. A., C.-Y. Chiang, and J. F. Murray. 2004. Global epidemiology of site mutation in CYBB gene. Blood 98: 436–441. tuberculosis. In Tuberculosis, 2nd Ed. W. N. Rom, and S. M. Garay, eds. Lip- 30. Leeman, J. R., and T. D. Gilmore. 2008. Alternative splicing in the NF-kappaB pincott Williams & Wilkins, New York, p. 13–29. signaling pathway. Gene 423: 97–107. 3. Kim, H. R., S. S. Hwang, H. J. Kim, S. M. Lee, C. G. Yoo, Y. W. Kim, S. K. Han, 31. Jaresova´, I., D. Rozkova´,R.Spı´sek, A. Janda, J. Bra´zova´, and A. Sediva´. 2007. Y. S. Shim, and J. J. Yim. 2007. Impact of extensive drug resistance on treatment Kinetics of Toll-like receptor-4 splice variants expression in lipopolysaccharide- outcomes in non-HIV-infected patients with multidrug-resistant tuberculosis. stimulated antigen presenting cells of healthy donors and patients with cystic Clin. Infect. Dis. 45: 1290–1295. fibrosis. Microbes Infect. 9: 1359–1367. 4. Sotgiu, G., G. Ferrara, A. Matteelli, M. D. Richardson, R. Centis, S. Ruesch- 32. Leung, E., J. Hong, A. Fraser, and G. W. Krissansen. 2007. Splicing of NOD2 Gerdes, O. Toungoussova, J. P. Zellweger, A. Spanevello, D. Cirillo, et al. 2009. (CARD15) RNA transcripts. Mol. Immunol. 44: 284–294. Epidemiology and clinical management of XDR-TB: a systematic review by 33. Kobayashi, K., L. D. Hernandez, J. E. Galań, C. A. Janeway, Jr., R. Medzhitov, TBNET. Eur. Respir. J. 33: 871–881. and R. A. Flavell. 2002. IRAK-M is a negative regulator of Toll-like receptor 5. Ernst, J. D. 2012. The immunological life cycle of tuberculosis. Nat. Rev. signaling. Cell 110: 191–202. Immunol. 12: 581–591. 34. Janssens, S., K. Burns, J. Tschopp, and R. Beyaert. 2002. Regulation of 6. Chan, J., and J. Flynn. 2004. The immunological aspects of latency in tuber- interleukin-1- and lipopolysaccharide-induced NF-kappaB activation by alter- culosis. Clin. Immunol. 110: 2–12. native splicing of MyD88. Curr. Biol. 12: 467–471. 7. Durbin, J. E., R. Hackenmiller, M. C. Simon, and D. E. Levy. 1996. Targeted 35. Hardy, M. P., and L. A. O’Neill. 2004. The murine IRAK2 gene encodes four disruption of the mouse Stat1 gene results in compromised innate immunity to alternatively spliced isoforms, two of which are inhibitory. J. Biol. Chem. 279: viral disease. Cell 84: 443–450. 27699–27708. 8. Meraz, M. A., J. M. White, K. C. F. Sheehan, E. A. Bach, S. J. Rodig, 36. Shell, S. A., C. Hesse, S. M. Morris, Jr., and C. Milcarek. 2005. Elevated levels A. S. Dighe, D. H. Kaplan, J. K. Riley, A. C. Greenlund, D. Campbell, et al. of the 64-kDa cleavage stimulatory factor (CstF-64) in lipopolysaccharide- 1996. Targeted disruption of the Stat1 gene in mice reveals unexpected physi- stimulated macrophages influence gene expression and induce alternative poly ologic specificity in the JAK-STAT signaling pathway. Cell 84: 431–442. (A) site selection. J. Biol. Chem. 280: 39950–39961. The Journal of Immunology 2755

37. Robinson, R. T., S. A. Khader, C. A. Martino, J. J. Fountain, M. Teixeira-Coelho, tions, pathways and biological processes. Nucleic Acids Res. 39(Database issue): J. E. Pearl, S. T. Smiley, G. M. Winslow, D. L. Woodland, M. J. Walter, et al. D691–D697. 2010. Mycobacterium tuberculosis infection induces il12rb1 splicing to generate 52. Yamaguchi, K. D., D. L. Ruderman, E. Croze, T. C. Wagner, S. Velichko, a novel IL-12Rbeta1 isoform that enhances DC migration. J. Exp. Med. 207: A. T. Reder, and H. Salamon. 2008. IFN-beta-regulated genes show abnormal 591–605. expression in therapy-naı¨ve relapsing-remitting MS mononuclear cells: gene 38. Park, E., M. S. Lee, S. M. Baik, E. B. Cho, G. H. Son, J. Y. Seong, K. H. Lee, and expression analysis employing all reported protein-protein interactions. J. Neu- K. Kim. 2009. Nova-1 mediates glucocorticoid-induced inhibition of pre-mRNA roimmunol. 195: 116–120. splicing of gonadotropin-releasing hormone transcripts. J. Biol. Chem. 284: 53. Croze, E., K. D. Yamaguchi, V. Knappertz, A. T. Reder, and H. Salamon. 2012. 12792–12800. Interferon-beta-1b-induced short- and long-term signatures of treatment activity 39. Friedman, C. R., G. C. Quinn, B. N. Kreiswirth, D. C. Perlman, N. Salomon, in multiple sclerosis. Pharmacogenomics J. 10.1038/tpj.2012.27. N. Schluger, M. Lutfey, J. Berger, N. Poltoratskaia, and L. W. Riley. 1997. 54. Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate: Widespread dissemination of a drug-susceptible strain of Mycobacterium tu- a practical and powerful approach to multiple testing. J. R. Stat. Soc., B 57: 289– berculosis. J. Infect. Dis. 176: 478–484. 300. 40. Riendeau, C. J., and H. Kornfeld. 2003. THP-1 cell apoptosis in response to 55. R Development Core Team. 2010. A language and environment for statistical Mycobacterial infection. Infect. Immun. 71: 254–259. computing. R Foundation for Statistical Computing, Vienna, Austria. 41. Sly, L. M., M. Lopez, W. M. Nauseef, and N. E. Reiner. 2001. 1alpha,25- 56. Kincaid, E. Z., and J. D. Ernst. 2003. Mycobacterium tuberculosis exerts gene- Dihydroxyvitamin D3-induced monocyte antimycobacterial activity is regu- selective inhibition of transcriptional responses to IFN-gamma without inhibiting lated by phosphatidylinositol 3-kinase and mediated by the NADPH-dependent STAT1 function. J. Immunol. 171: 2042–2049. phagocyte oxidase. J. Biol. Chem. 276: 35482–35493. 57. Ehrt, S., D. Schnappinger, S. Bekiranov, J. Drenkow, S. Shi, T. R. Gingeras, 42. Boulikas, T. 1988. At least 60 ADP-ribosylated variant histones are present in T. Gaasterland, G. Schoolnik, and C. Nathan. 2001. Reprogramming of the nuclei from dimethylsulfate-treated and untreated cells. EMBO J. 7: 57–67. macrophage transcriptome in response to interferon-gamma and Mycobacterium 43. Pine, R., D. E. Levy, N. Reich, and J. E. Darnell, Jr. 1988. Transcriptional tuberculosis: signaling roles of nitric oxide synthase-2 and phagocyte oxidase. J. stimulation by CaPO4-DNA precipitates. Nucleic Acids Res. 16: 1371–1378. Exp. Med. 194: 1123–1140. 44. Choe, S. E., M. Boutros, A. M. Michelson, G. M. Church, and M. S. Halfon. 58. Gehring, A. J., R. E. Rojas, D. H. Canaday, D. L. Lakey, C. V. Harding, and 2005. Preferred analysis methods for Affymetrix GeneChips revealed by W. H. Boom. 2003. The Mycobacterium tuberculosis 19-kilodalton lipoprotein a wholly deﬁned control dataset. Genome Biol. 6: R16. inhibits gamma interferon-regulated HLA-DR and Fc gamma R1 on human 45. Baldi, P., and A. D. Long. 2001. A Bayesian framework for the analysis of macrophages through Toll-like receptor 2. Infect. Immun. 71: 4487–4497. Downloaded from microarray expression data: regularized t -test and statistical inferences of gene 59. Alvarez, G. R., B. S. Zwilling, and W. P. Lafuse. 2003. Mycobacterium avium changes. Bioinformatics 17: 509–519. inhibition of IFN-gamma signaling in mouse macrophages: Toll-like receptor 2 46. Ashburner, M., C. A. Ball, J. A. Blake, D. Botstein, H. Butler, J. M. Cherry, stimulation increases expression of dominant-negative STAT1 beta by mRNA A. P. Davis, K. Dolinski, S. S. Dwight, J. T. Eppig, et al; The Gene Ontology stabilization. J. Immunol. 171: 6766–6773. Consortium. 2000. Gene ontology: tool for the uniﬁcation of biology. Nat. Genet. 60. Shi, S., C. Nathan, D. Schnappinger, J. Drenkow, M. Fuortes, E. Block, A. Ding, 25: 25–29. T. R. Gingeras, G. Schoolnik, S. Akira, et al. 2003. MyD88 primes macrophages 47. Berry, M. P., C. M. Graham, F. W. McNab, Z. Xu, S. A. Bloch, T. Oni, for full-scale activation by interferon-gamma yet mediates few responses to

K. A. Wilkinson, R. Banchereau, J. Skinner, R. J. Wilkinson, et al. 2010. An Mycobacterium tuberculosis. J. Exp. Med. 198: 987–997. http://www.jimmunol.org/ interferon-inducible neutrophil-driven blood transcriptional signature in human 61. Pai, R. K., M. E. Pennini, A. A. Tobian, D. H. Canaday, W. H. Boom, and tuberculosis. Nature 466: 973–977. C. V. Harding. 2004. Prolonged toll-like receptor signaling by Mycobacterium 48. Maertzdorf, J., M. Ota, D. Repsilber, H. J. Mollenkopf, J. Weiner, P. C. Hill, and tuberculosis and its 19-kilodalton lipoprotein inhibits gamma interferon-induced S. H. Kaufmann. 2011. Functional correlations of pathogenesis-driven gene regulation of selected genes in macrophages. Infect. Immun. 72: 6603–6614. expression signatures in tuberculosis. PLoS ONE 6: e26938. 62. Wang, Y., H. M. Curry, B. S. Zwilling, and W. P. Lafuse. 2005. Mycobacteria 49. Thuong, N. T., S. J. Dunstan, T. T. Chau, V. Thorsson, C. P. Simmons, inhibition of IFN-gamma induced HLA-DR gene expression by up-regulating N. T. Quyen, G. E. Thwaites, N. Thi Ngoc Lan, M. Hibberd, Y. Y. Teo, et al. histone deacetylation at the promoter region in human THP-1 monocytic cells. J. 2008. Identiﬁcation of tuberculosis susceptibility genes with human macrophage Immunol. 174: 5687–5694. gene expression proﬁles. PLoS Pathog. 4: e1000229. 63. Sow, F. B., G. R. Alvarez, R. P. Gross, A. R. Satoskar, L. S. Schlesinger, 50. Joshi-Tope, G., M. Gillespie, I. Vastrik, P. D’Eustachio, E. Schmidt, B. de Bono, B. S. Zwilling, and W. P. Lafuse. 2009. Role of STAT1, NF-kappaB, and C/ B. Jassal, G. R. Gopinath, G. R. Wu, L. Matthews, et al. 2005. Reactome: EBPbeta in the macrophage transcriptional regulation of hepcidin by myco-

a knowledgebase of biological pathways. Nucleic Acids Res. 33(Database issue): bacterial infection and IFN-g. J. Leukoc. Biol. 86: 1247–1258. by guest on October 1, 2021 D428–D432. 64. Benson, S. A., and J. D. Ernst. 2009. TLR2-dependent inhibition of macrophage 51. Croft, D., G. O’Kelly, G. Wu, R. Haw, M. Gillespie, L. Matthews, M. Caudy, responses to IFN-gamma is mediated by distinct, gene-speciﬁc mechanisms. P. Garapati, G. Gopinath, B. Jassal, et al. 2011. Reactome: a database of reac- PLoS ONE 4: e6329.