The genetic control of susceptibility to tuberculosis in DBA/2J and C57BL/6J mice is complex and influenced by at least four tuberculosis resistance loci (Trl1-Trl4). To further study the Trl3 and Trl4 loci, we have created congenic mouse lines D2.B6-Chr7 and D2.B6-Chr19, in which resistant B6-derived portions of chromosome 7 (Chr.7) and chromosome 19 (Chr.19) overlapping Trl3 and Trl4, respectively, were independently introgressed onto susceptible D2 background. Transfer of B6-derived Trl3 chromosome 7 segment significantly increased resistance of D2 mice, as measured by reduced pulmonary microbial replication at day 70, and increased host survival following aerosol infection. However, transfer of B6-derived chromosome 19 (Trl4) onto D2 mice did not increase resistance by itself and does not improve on the protective effect of chromosome 7. Further study of the protective effect of Trl3 in D2.B6-Chr7 mice indicates that it does not involve modulation of timing or magnitude of Th1 response in the lung, as investigated by measuring the number of Ag-specific, IFN-γ-producing CD4+ and CD8+ T cells. Rather, Trl3 appears to affect the intrinsic ability of activated macrophages to restrict intracellular mycobacterial replication in an NO synthase 2-independent fashion. Microarray experiments involving parental and congenic mouse lines identified a number of genes in the Trl3 interval on chromosome 7 the level of expression of which before infection or in response to Mycobacterium tuberculosis infection is differentially regulated in a parental haplotype-dependent fashion. This gene list represents a valuable entry point for the identification and prioritization of positional candidate genes for the Trl3 effect on chromosome 7.
Tuberculosis (TB),3 caused by Mycobacterium tuberculosis, remains one of the leading causes of morbidity and mortality by an infectious agent worldwide. At present, it is estimated that one-third of the world population is infected with M. tuberculosis (1), with nearly 8 million new cases of active disease per year (2) and 1–1.5 million deaths annually. In contrast, the fact that only a small fraction of individuals infected with M. tuberculosis go on to develop active disease suggests that humans possess robust mechanisms of innate defenses against M. tuberculosis. Such mechanisms of defense can manifest themselves as genetic variants associated with increased susceptibility to TB in humans and in animal models of infection.
Although a large body of literature supports a complex genetic component of predisposition to TB in humans, this phenomenon is very difficult to study in humans, and identifying the genes involved has so far not been possible (3). Conversely, the parallel study of inbred and mutant stocks of mice has shed light on the cell types, and physiological and biochemical pathways involved in innate defenses against TB. This insight has been achieved primarily by the “reverse genetics” approach where the role of individual genes is tested by infecting mice bearing loss-of-function mutations at the corresponding locus (4, 5, 6, 7). Furthermore, inbred strains of mice vary dramatically in their degree of susceptibility to pulmonary TB, and genetic studies of such differences are starting to provide insight into mechanisms of host defense against M. tuberculosis (see Ref. 3 for a recent review). Inbred strains have been classified as highly susceptible (CBA, C3HeB/FeJ, DBA/2, 129SvJ) or highly resistant (C57BL/6J, BALB/c) to i.v. or aerosol infection with M. tuberculosis (8, 9). The extreme susceptibility of C3HeB/FeJ mice to aerosol or i.v. infection with M. tuberculosis has been mapped to a single locus on chromosome 1 designated sst1 (supersusceptibility to tuberculosis 1) (10). Studies in congenic mouse lines have established that sst1-encoded resistance is phenotypically expressed as reduced intracellular microbial replication in macrophages that is associated with increased induction of apoptosis in response to M. tuberculosis infection (11). Positional cloning experiments and studies of macrophages from transgenic mice have shown that differential expression of the Ipr1 (intracellular pathogen resistance 1) gene, coding for the protein Ifi75 (IFN-induced protein 75), is responsible for this effect (11). Ifi75 is a homolog of the human protein SP110 (12, 13), an IFN-regulated nuclear transcriptional regulator regulated by IFN that appears to modulate gene expression and induction of apoptosis in response to M. tuberculosis infection (11). Recently, a number of genetic modifiers of the Ipr1 effect have been characterized on chromosomes 7, 12, 15, and 17 (14), thereby further highlighting the complexity of genetic susceptibility to TB in C3HeB/FeJ. An association of SP110 alleles with TB in humans has been noted in certain populations (15), whereas no association could be detected in others (16). Independently, the interstrain difference in susceptibility to M. tuberculosis of I/St (susceptible) and A/Sn (resistant) strains was mapped to chromosomes 9 and 3 loci, using infection-induced loss of body mass following infection as a phenotypic marker of susceptibility (17, 18). Susceptibility was associated with decreased IFN-γ production in infected lungs, increased influx of neutrophils in lung tissue, and resistance of neutrophils to M. tuberculosis-induced cell death (19). Again, differential intracellular replication of M. tuberculosis in lung macrophages in vitro was correlated with resistance/susceptibility in vivo, highlighting the role of mononuclear and polymorphonuclear phagocytes in host defenses against M. tuberculosis.
We have studied the differential susceptibility of C57BL/6J (resistant) and DBA/2 (susceptible) strains. Susceptibility in DBA/2J is detected following either i.v. or aerosol infection with M. tuberculosis H37Rv, and is associated with reduced control of pulmonary microbial replication, more rapid inflammatory response in the lungs, and reduced time of survival (20, 21). Linkage analysis by whole genome scanning of informative (B6 X D2)F2 mice using survival time (i.v. infection) and pulmonary microbial load (aerosol infection) as phenotypic markers of susceptibility (20, 21) detected four TB-resistant loci Trl1-Trl4 that regulate the extent of pulmonary replication of M. tuberculosis Trl3 (chromosome 7 (Chr.7); LOD score 4.1) and Trl4 (chromosome 19 (Chr.19); LOD 5.6)) (21) or survival following infection (Trl1 (Chr.1; LOD 4.8), Trl2 (Chr.3; LOD 3.9), and Trl3 (Chr.7; LOD 4.7)) (20). All loci account for limited portions of the phenotypic variance, with B6 parental alleles being associated with increased resistance and inherited in a semidominant fashion. In one study, (B6 X D2)F2 mice homozygous for B6 alleles at both Trl3 and Trl4 were found to be as resistant as B6 parents, whereas mice homozygous for D2 alleles at both loci were as susceptible as the D2 parents (21).
Trl3 and Trl4 were prioritized in this study for further investigation based on the observation that 1) Trl3 affects both the rate of pulmonary M. tuberculosis replication and survival to infection, 2) Trl4 has the strongest genetic effect detected in independent genome scans, 3) there is strong additive effect of Trl3 and Trl4, with their combined effect explaining 38% of the phenotypic variance in (B6 X D2)F2 (20, 21). To study the individual contribution of these two loci on host resistance to M. tuberculosis, we have created congenic mouse lines where the B6 chromosomal segments corresponding to Trl3 and Trl4 were introgressed independently or together onto the genetically susceptible background of DBA/2. We have characterized these congenic mouse lines with respect to penetrance of the Trl3 and Trl4 effect on pulmonary microbial replication and host survival. The mechanistic basis of Trl3 protective effect has been studied by following histopathology, identifying the cell types, and biochemical pathways involved, including an analysis of positional candidates on chromosome 7.
Materials and Methods
Inbred, pathogen free, C57BL/6J (B6) and DBA/2J (D2) male mice were purchased from the Trudeau Institute Animal Breeding Facility. Mice were free of common viral pathogens as determined by routine testing performed by the Research Animal Diagnostic and Investigative Laboratory, University of Missouri (Columbia, MO). Trl3 (Chr.7) and Trl4 (Chr.19) congenic mouse strains were generated using a speed-congenic (marker-assisted) approach (22). In this protocol, successive F1 backcross males were genotyped to identify individuals with the least residual donor genomic DNA, and were selected for further backcrossing. In these mice, the chromosome 7 (Trl3) or chromosome 19 (Trl4) segments from B6 strain was introgressed onto the genetic background of the D2 strain. For Trl3, the BXD19 recombinant inbred strain was used as the initial donor for the proximal portion of chromosome 7 (D7Mit178-D7Mit193) (see Fig. 1⇓A), whereas for Trl4 the BXD9 strain was the initial donor of the distal portion of chromosome 19 (D19Mit69-D19Mit137) (see Fig. 1⇓B). Once at the N4 generation, heterozygotes were intercrossed to generate the homozygote congenic lines and also to produce the double congenic line. Genotyping was conducted as previously described (20), using tail genomic DNA and sequence-specific oligonucleotide primers (see supplemental Table S1).4 All experimental procedures involving mice were approved by the Institutional Animal Care and Use Committee of the Trudeau Institute.
Infection with M. tuberculosis
The M. tuberculosis strain H37Rv was obtained from the Trudeau Mycobacterial Culture Collection as a frozen (−70°C) log phase stock dispersed in Proskauer and Beck medium (Difco) containing 0.01% Tween 80. For each experiment, a vial was thawed, subjected to 5-s ultrasound to break up aggregates, and diluted appropriately in PBS containing 0.01% Tween 80. Mice, 8 to 10 wk of age, were inoculated with 102 CFU by the aerosol route in a Middlebrook airborne infection apparatus (Tri Instruments). Bacilli were enumerated in the lungs of infected mice at 30 and 70 days of infection by preparing lung homogenates in PBS containing 0.05% Tween 80 and by plating 10-fold serial dilutions of the homogenates on Middlebrook enriched 7H11 agar (Difco). CFU counts were performed after 3–4 wk of incubation at 37°C, and the data are presented as log10 of total CFU count per lung.
Enumeration of CD4 T cell by flow cytometry and ELISPOT
Mice were euthanized, and their lungs were perfused with PBS containing 10 U/ml heparin to remove intravascular leukocytes. The lungs were then perfused with an enzyme mixture consisting of 150 U/ml collagenase, 0.2 U/ml elastase (Roche Applied Sciences), and 1 mg/ ml DNase (Sigma-Aldrich) in RPMI 1640. The lungs were removed, diced into small fragments, and subjected to further enzyme digestion before being mechanically disrupted to form a single cell suspension (23). Total lung cells were suspended in RPMI 1640-FCS in two 5-ml tubes at 1 × 107/ml, and incubated with brefeldin A (10 μg/ml) for 5 h at 37°C (Epicentre Technologies). They were then stained for flow cytometry with FITC anti-CD3, R-PE anti-CD4, and peridinin chlorophyll protein anti-CD8 mAbs. After fixation overnight in 0.5% paraformaldehyde they were stained for intracellular IFN-γ with allophycocyanin anti-IFN-γ mAb, as previously described (23), and analyzed by FACSCalibur flow cytometer (BD Biosciences) using CellQuest software (BD Biosciences).
Changes in the total number of Ag-specific T cells in the lungs capable of making IFN-γ in response to M. tuberculosis Ags were determined using the ELISPOT assay (Mouse IFN-γ ELISPOT set; BD Biosciences), according to the manufacturer’s instructions and using total lung cells from a pool of mice (n = 4) (23, 24). The M. tuberculosis Ag preparation used to stimulate IFN-γ production was a sonicated extract of a M. tuberculosis culture (23) and an ESAT-6 (1–20) peptide (New England Peptide) (25), which is an early-secreted M. tuberculosis protein.
Lungs were fixed by an intratracheal infusion of 10% formalin followed by 24-h immersion (at 20°C) in the same fixative. The lungs were then dehydrated in serial baths of 70% and 100% ethanol, embedded in paraffin, and sectioned. Immunocytochemistry to detect the NO synthase (NOS)2 enzyme and staining for acid-fast bacteria were performed as previously described (26). Briefly, lung sections were incubated with affinity-purified rabbit anti-mouse NOS2 Ig (primary Ab), biotinylated goat anti-rabbit Ig (secondary Ab), and avidin-coupled biotinylated HRP with diaminobenzidine as the substrate. Sections were then stained for acid-fast bacilli (27) and counterstained with methylene blue. Photomicrography was performed with a Nikon Microphot-Fx microscope fitted with a Spot RT Slider camera (Diagnostic Instruments) using Spot RT Software for image acquisition.
Transcriptional profiling with microarrays
Total lung RNA (three samples per experimental group; 36 in total) from uninfected controls and M. tuberculosisM. tuberculosis-infected (day 30 and day 70) groups for B6, D2, D2.B6-Chr7, and D2.B6-Chr19 mice individually. Data were normalized by the robust multiarray average algorithm, and differential expression was tested by using either pairwise, or two-way ANOVA analysis with independent t tests. Complete microarray data (accession no. E-MEXP-1942) has been deposited in the ArrayExpress database (www.ebi.ac.uk/microarray-as/ae/).
Construction of mouse lines congenic for the TB-resistance loci Trl3 and Trl4
Host resistance loci Trl3 and Trl4 were selected for investigation for the following reasons: 1) Trl3 affects both pulmonary bacterial replication and host survival time following M. tuberculosis infection in independent genome scans of (B6 X D2)F2 mice (20, 21), 2) Trl4 shows the strongest genetic effect in these mice (21), and 3) there is strong additive interactions between Trl3 and Trl4 that explains a large fraction of the phenotypic variance in (B6 X D2)F2 mice (21). To study the independent contribution of Trl3 and Trl4 to the differential susceptibility phenotype of B6 vs D2, and to initiate the identification of the cellular and biochemical pathways underlying both genetic effects, we have used a marker-assisted strategy to construct congenic lines in which the chromosome 7 (Trl3) and chromosome 19 (Trl4) segments of B6 resistant background were independently introgressed onto the D2 susceptible background. For chromosome 7, the proximal segment delineated by markers D7Mit178 to D7Mit193 (Fig. 1⇑A) was transferred to create D2.B6-Chr7, whereas for chromosome 19 the distal chromosomal segment delineated by markers D19Mit69 to D19Mit137 was transferred to create D2.B6.Chr19 (Fig. 1⇑B). To increase the efficiency of the breeding process, B6 chromosomal segments to be introgressed onto D2 were donated by independent BXD recombinant congenic strains (that already contain a 50:50 ratio of B6 vs D2 genome) BXD9 (Trl4) and BXD19 (Trl3). Heterozygotes produced by serial backcrossing were then intercrossed to generate homozygotes at each locus. Independent Trl3 and Trl4 congenic lines were then intercrossed to generate the Trl3:Trl4 double congenic line.
Effect of chromosomes 7 and 19 on susceptibility to pulmonary infection with M. tuberculosis
To evaluate the effect of the transferred chromosome 7 (Trl3) and chromosome 19 (Trl4) B6 segments on the susceptibility of D2 mice to infection with M. tuberculosis, D2.B6-Chr7 and D2.B6-Chr19 singly congenic lines together with C57BL/6 (B6) and DBA/2 (D2) controls were infected with 102 CFU of virulent M. tuberculosis H37Rv by the aerosol route. At day 30 and day 70 postinfection, pulmonary bacterial load was determined. Similar bacterial loads were measured in the lungs of the four mouse strains at day 30 (Fig. 2⇓), a time point of the infection where the innate difference between B6 and D2 in TB susceptibility is not yet detectable. In the subsequent genetically controlled phase of infection (day 70), increased pulmonary replication of M. tuberculosis had occurred in D2 mice, resulting in a 16-fold increase in CFU from day 30 to 70. By contrast, the infection was held stationary in lungs of B6 mice over the same period (Fig. 2⇓), in agreement with previous reports (21). In these experiments, the D2.B6-Chr19 line showed microbial pulmonary loads (days 30 and 70) very similar to those of D2 mice, whereas a significant 7- to 8-fold reduction in lung CFU was measured in D2.B6-Chr7 mice compared with D2 controls (Fig. 2⇓). The double congenic line for the B6-derived chromosomes 7 and 19 segments was also tested in the same setting, and showed CFU counts at day 70 that were undistinguishable from those seen in D2.B6-Chr7 mice (data not shown). Therefore, although transfer of the B6-derived Trl3-containing chromosome 7 segment significantly reduces pulmonary replication of M. tuberculosis in otherwise permissive D2 mice, transfer of chromosome 19 (Trl4) does not enhance resistance by itself and does not improve on the protective effect of chromosome 7.
The effect of the transferred chromosomes 7 and 19 B6-derived segments on host survival from M. tuberculosis infection was next determined (Fig. 3⇓). Following aerosol infection, susceptible D2 mice showed a mean survival time (MST) of ∼100 days, with all animals succumbing to infection by day 107. In contrast, resistant B6 mice all survived beyond the 220-day observation period. The MST of the D2.B6-Chr19 line was 89 days, a value similar to that of D2 controls. In contrast, D2.B6-Chr7 mice displayed a significant increase in survival time over D2, revealing an MST of 155 days with animals surviving up to day 199. Therefore, the protective effect of B6-derived chromosome 7 (Trl3) initially detected by the reduction in pulmonary microbial replication caused an increase in host survival time. These two phenotypes are also known to be influenced by Trl3 in informative (B6 X D2)F2 mice (20, 21). Conversely, the protective effect of B6 alleles at Trl4 noted in the same (B6 X D2)F2 mice (21) is lost upon transfer of chromosome 19 onto the D2 background.
Effect of Trl3 and Trl4 on T cell-mediated immune response in M. tuberculosis-infected lungs
To initiate studies on the mechanistic basis of increased resistance noted in the D2.B6-Chr.7 congenics (Trl3), we investigated the effect of this locus on the type and extent of Th1 response during M. tuberculosis infection. In these studies, B6 and D2 control mice together with D2.B6-Chr7 and D2.B6-Chr19 congenics were infected with M. tuberculosis and the number of IFN-γ-producing CD4+ and CD8+ T cells were enumerated in the lungs, 10, 20, 30, and 50 days following aerosol infection (Fig. 4⇓). In B6-resistant controls, there was a very robust accumulation of IFN-γ-producing CD4+ T cells (Fig. 4⇓A) and to a lower extent CD8+ T cells (Fig. 4⇓B). This T cell influx started at day 20, and peaked at day 30, a time point at which B6 mice start to restrict pulmonary bacterial replication. This cellular response subsided by day 50 in B6 mice. By contrast, no such rapid T cell influx of IFN-γ-producing T cells was seen in susceptible D2 controls. Rather, there was a slow accumulation of both CD4+ and CD8+ T cells in infected lungs of D2 mice that peaked at day 50 and with cell numbers four times lower than peak numbers observed in control B6 at day 30 postinfection. Conversely, the D2.B6-Chr7 congenic mice displayed kinetics of IFN-γ+ CD4+ and CD8+ T cell accumulation in the lung that were indistinguishable from the numbers seen in D2 susceptible controls (Fig. 4⇓, A and B). As expected, D2.B6-Chr19 congenics showed T cells recruitment kinetics similar to D2 controls.
The effect of Trl3 on the production of Ag-specific IFN-γ-secreting T cells in the lung was also investigated by ELISPOT assay, using either a crude M. tuberculosis sonicate or an ESAT-6 (1–20) peptide (a major M. tuberculosis Ag) as antigenic stimuli (Fig. 4⇑C). In agreement with results reported in Fig. 4⇑, A and B, for the number of IFN-γ+ CD4+ and CD8+ T cells in the lungs of infected mice, there was a robust response to both ESAT-6 and M. tuberculosis sonicate in the lungs of control B6 mice that peaked at day 30 and diminished by day 50. In contrast, susceptible D2 controls lacked the peak at day 30, but instead showed a more progressive response that peaked at day 50, the last time point investigated. The response of D2 mice to the M. tuberculosis sonicate was most robust (compared with ESAT-6), and at day 50 the number of IFN-γ-secreting T cells in response to the sonicated extract was similar to the number measured in B6 (Fig. 4⇑C). Again, the number of Ag-specific IFN-γ-secreting T cells detected in both chromosomes 7 and 19 congenic lines was similar to that detected in the D2 controls. Together these results suggest that transfer of B6-derived Trl3 alleles in D2.B6-Chr7 congenics increases resistance to pulmonary TB by a mechanism that is unlinked to the ability of B6 mice to rapidly recruit IFN-γ+ CD4+ and CD8+ T cells to the lung.
Effect of Trl3 on histopathology of M. tuberculosis infected lungs
The lung histopathology of D2 and B6 mice following aerosol challenge with M. tuberculosis (at the time D2 mice begin to succumb from infection (day 90 postinfection)) has been previously described (28). In the present study, lung lesions observed at day 70 postinfection in parental and congenic strains were similar in size, number, and cellular composition (Fig. 5⇓, A–D), although in some cases the alveoli in lesions of D2 and D2.B6-Chr19 mice were beginning to show presence of neutrophils. The striking difference between the lesions of susceptible and resistant mice, however, was in the level of infection of individual macrophages (Fig. 5⇓, E–H), with those in the lesions of D2 (Fig. 5⇓H) and D2.B6-Chr19 (Fig. 5⇓G) mice containing much larger numbers of acid-fast bacilli. This finding suggests that the higher pulmonary load of M. tuberculosis detected in D2 and D2.B6-Chr19 mice may be explained by a much higher level of intracellular replication in individual macrophage, as opposed to a larger number of infected macrophages.
Because expression of the bactericidal enzyme NOS2 is a good indicator of IFN-γ-triggered macrophage activation, NOS2 protein expression was also evaluated in these M. tuberculosis-infected lungs by immunocytochemistry at day 70 postinfection (Fig. 5⇑). Despite important interstrain differences in pulmonary bacterial loads measured at day 70, comparable levels of NOS2 staining was seen in macrophages of B6 and D2 mice and the two congenic strains (Fig. 5⇑). Additional transcriptional profiling studies also showed similar levels of IFN-γ and NOS2 mRNA expression in lungs from all strains following 30 (see supplemental Fig. S1A) and 70 (see supplemental Fig. S1B) days of M. tuberculosis infection.4 This demonstration suggests that neither the B6 vs D2 interstrain difference in susceptibility to M. tuberculosis, nor the enhanced resistance associated with transfer of B6 alleles at Trl3 in D2.B6-Chr7 congenics are linked to differences in activation of the NO-based anti-M. tuberculosis defense.
Effect of Trl3 on transcript profiles in the lungs of M. tuberculosis-infected mice
To gain insight into the host cell types, cellular and molecular pathways possibly involved in the differential permissiveness to pulmonary replication of M. tuberculosis, we conducted transcript profiling studies on M. tuberculosis-infected lungs from congenic and parental strains. We were particularly interested in identifying groups of transcripts associated with increased resistance to M. tuberculosis infection of B6 mice and the D2.B6-Chr7 line. That list consists in the overlap between the lists commonly expressed in response to infection between resistant B6 mice and the D2.B6-Chr7, but that show a significant difference in modulation when compared with infected susceptible D2 mice.
In these experiments, B6 and D2 mice, as well as the D2.B6-Chr19 and D2.B6-Chr7 congenic lines, were infected with M. tuberculosis4 All experimental groups revealed a similar median expression value and a similar distribution of gene expression intensity, indicating homogeneity of the data set. A similar analysis of the three replicates for individual experimental group also showed homogeneity of the data within each group, with no outliers (data not shown).
In a first set of analyses, gene expression data obtained at days 30 and 70 postinfection were compared with that of day 0 (day 30 vs day 0; day 70 vs day 0) for each strain to establish M. tuberculosis-induced strain-specific ratios (pairwise analysis). These ratios were then used to determine gene expression profiles similarities in all mouse strains. Statistical analysis (p < 0.01 by t test; fold change doubled or more) revealed that infection with M. tuberculosis had a dramatic and potent effect on the profile of gene expression of all mouse strains. Interstrain comparison revealed that 1232 and 1876 genes were commonly regulated in all mouse strains in response to M. tuberculosis infection and this level at day 30 and 70, respectively. Of those, a total of 1136 transcripts were modulated in response to infection at both time points and in all mouse strains (see supplemental Fig. S2B).4 Additional analysis of the top 100 genes showing the greatest degree of M. tuberculosis-induced modulation, showed that 30 genes (of these 100) were commonly regulated in all mouse strains at both time points (see supplemental Table S2).4 This gene list represents a fingerprint of the anti-TB transcriptional response following aerosol infection and includes genes such as Ifnγ, Stat1, Irg1, or Iigp1 as well as chemokine ligands Cxcl9, Cxcl10, Ccl8 and chemokine receptor Cxcr6. In contrast, seven and six transcripts were specifically modulated for the day-30 and day-70 time points, respectively, revealing potential stage-specific aspects of host response to pulmonary TB (see supplemental Table S2).4
In a second stage, transcript profiles associated with differential susceptibility to M. tuberculosis in resistant B6 and D2.B6-Chr7 lines vs susceptible D2 mice ((B6 vs D2) and (D2.B6-Chr7 vs D2)) were investigated using a two-way ANOVA analysis of the day-30 data set. We focused our analysis on day 30 because it is at this time point that resistant B6 mice start to restrict pulmonary bacterial replication, and hence are likely to express genes contributing to protection. Likewise, B6, D2, and D2.B6-Chr7 show similar pulmonary microbial loads at day 30 (as opposed to day 70) (Fig. 2⇑), thereby minimizing the effect of bacterial burden on identification of quantitative differences in expression associated with resistance to infection. Gene lists were generated by comparing M. tuberculosis-induced changes in transcript profiles (expression at day 30 postinfection vs day 0) in both resistant strains vs susceptible strain (either B6 or D2.B6-Chr7 compared with D2). The comparison of B6 and D2.B6-Chr7 to susceptible D2 mice revealed a total of 1109 and 1611 genes showing a statistically significant difference in regulation in response to infection (p < 0.05 by t test; fold change twice as much or more), respectively, including a subset of 284 genes commonly regulated between B6 and D2.B6-Chr7. We have selected a subset of 12 such transcripts for validation by semiquantitative RT-PCR (see supplemental Fig. S3).4 These experiments showed a validation success rate of ∼60%. Most nonvalidated transcripts gave poor amplification results and were expressed at very low levels in lungs of all mouse strains tested. A gene ontology report on these 284 (221 annotated) genes revealed that 42 (19% of annotated genes) of them were associated with the immune system, representing the most abundant group of genes. By removing genes in duplicate and genes for which the expression ratio in D2.B6-Chr7 congenics did not follow the expression kinetic observed in B6 when compared with D2, 32 genes (Irf7, Pglyrp1, Bcl3, Cd22, and Coro1a are located on chromosome 7) were identified (Table I⇓). A few chemokine ligand genes are present, including Cxcl5, for which differential expression in B6 and D2 mice has been previously reported (29). Moreover, 16 of the 221 annotated genes commonly regulated in B6 and D2.B6-Chr7 mice map to chromosome 7 (Table II⇓), including seven (Rps9, Pglyrp1, Apoc1, Bcl3, Cd22, Saa1, and Saa2) that are located within the boundaries of the chromosome 7 congenic segment introgressed in D2.B6-Chr7 and that contain the Trl3 locus. These genes constitute strong positional candidates for Trl3, including the B cell-specific regulatory gene Cd22 that maps directly under the Trl3 linkage peak (Fig. 6⇓A). The intrinsic and inducible expression profiles for the seven genes were quantitatively similar in B6 and D2.B6-Chr7 when compared with D2 mice. More precisely, gene expression in B6 and D2.B6-Chr7 was similarly higher (Rps9, Pglyrp1, and Cd22) or lower (Apoc1, Bcl3, Saa1, and Saa2) than in D2 mice. A representative example is the case of genes Cd22 and Rps9, in which there was robust induction expression of these genes in response to M. tuberculosis in B6 and D2.B6-Chr7 mice at day 30, whereas susceptible D2 mice did not show induction of expression of these two transcripts (Fig. 6⇓B), even 70 days postinfection (data not shown). Therefore, the dramatic expression of Cd22 in response to M. tuberculosis infection in the lungs of resistant strains compared with susceptible D2 mice represents a candidate expression quantitative trait locus (QTL) for Trl3.
The complex genetic control of differential susceptibility of B6 (resistant) and D2 (susceptible) mice to pulmonary TB was investigated by genome scanning and led to the mapping of four QTL, on chromosomes 1 (Trl1), 3 (Trl2), 7 (Trl3), and 19 (Trl4) (20, 21). The Trl3 and Trl4 QTL were selected for further study based in part on strength of genetic linkage (Trl4), and penetrance of the genetic determinant on different phenotypic measures of susceptibility (Trl3). Congenic lines were created in which the B6-derived, resistance-associated, chromosome 7 (Trl3; D2.B6-Chr7) and chromosome 19 (Trl4; D2.B6-Chr19) segments were independently introgressed onto susceptible D2. Phenotyping these lines for susceptibility to pulmonary TB validated the protective effect of B6-derived chromosome 7 (Trl3) both on pulmonary microbial replication and host survival, whereas transfer of the B6-derived Trl4 region of D2.B6-Chr19 did not appreciably improve resistance of susceptible D2 mice. Furthermore, although an additive effect of homozygosity for B6-derived resistant alleles at Trl3 and Trl4 on pulmonary bacterial loads had been detected in (B6 X D2)F2 mice (21), such an additive effect was not detected in chromosome 7/19 doubly congenic mice that were phenotypically indistinguishable from the D2.B6-Chr7 Trl3 congenic line. These results suggest that, although the Trl3 QTL can function in an “autonomous” fashion, the Trl4 genetic effect involves additional genetic interactions with loci that remain unknown but that were not included with the transfer of Trl4 to the D2 genetic background of the D2.B6-Chr19 congenic line.
CD4+ Th1 cells with the aid of CD8+ T cells play the major role in immunity to TB (4, 30, 31). In resistant mice, an orchestrated series of innate immune pathways (including IFN-γ and IL-12 production) resulting in a Th1-dominant adaptive immune pathways are activated following phagocytosis of M. tuberculosis, and culminate in the formation of granulomas at the sites of infection. Granuloma formation is essential for restriction of bacterial replication for prevention of dissemination of the infection. Therefore, the effect of the Trl3 locus on magnitude of Th1 response was measured. After a 10-day delay, resistant B6 mice displayed robust accumulation in the lungs of IFN-γ-producing CD4+ (predominantly) and CD8+ T cells, and of T cells capable of producing IFN-γ in response to both ESAT-6 and M. tuberculosis sonicate (ELISPOT). The response of CD4 and CD8 T cells peaked at day 30, and was concomitant with the restriction of pulmonary bacterial replication in the lungs of these mice. In contrast, D2 mice generated a Th1 cell response of much lower magnitude that continued until day 50 when the experiment was terminated. This effect suggests that a weaker response to mounting mycobacterial load may contribute importantly to the susceptibility of D2 animals. However, examination of the T cell response in D2.B6-Chr7 congenic mice showed that it was similar to that seen in D2, both with respect to kinetics of accumulation of IFN-γ-producing CD4/CD8 T cells in the lungs and total number that accumulated. These results strongly argue that the mechanism by which Trl3 improves host defense against TB does not involve major alterations in the magnitude or timing of the Th1 response as measured in this study.
In the murine M. tuberculosis infection model, the Th1 response leads to macrophage activation via the secretion of IFN-γ and other Th1 cytokines, and macrophages activation is evidenced by acquisition of an activation transcriptome (32) that includes induction of inducible NOS (NOS2) that enables these cells to generate NO and its metabolites that function to inhibit further bacterial growth (33). Histological examination of lung sections showed that many more acid-fast bacilli were present in individual macrophages in lung lesions of susceptible D2 and D2.B6-Chr19 mice than lung lesions in B6 and D2.B6-Chr7 mice, suggesting that Trl3-mediated resistance is determined by an enhanced capacity of individual macrophages to inhibit M. tuberculosis growth. Despite this observation, macrophages in susceptible mice (D2 and D2.B6-Chr19) stain as densely for NOS2 as those in the resistant B6 and intermediately resistant D2.B6-Chr7 mice. Therefore, macrophages appeared activated in all mouse strains in response to Th1 immunity, but yet varied in their ability to inhibit M. tuberculosis growth in a Trl3-dependent fashion. These results suggest that Trl3 affects the macrophage response to M. tuberculosis in a NOS2-independent fashion. These findings are reminiscent of those of Yan et al. (34) who studied the TB susceptibility controlled by sst1. They observed that sst1 affected neither activation of Th1 cytokine-producing T lymphocytes, nor their migration to the lungs, but controlled instead an inducible NOS-independent mechanism of innate immunity (34) linked to the ability of macrophages to undergo apoptosis in response to M. tuberculosis infection. Although both our study and Yan et al. (34) appear to share a parallel NOS2-independent mechanism, we failed to detect any differences in apoptosis between B6, D2, D2.B6-Chr7, and D2.B6-Chr19 mice, as evaluated by TUNEL assay on M. tuberculosis-infected lung sections (data not shown).
Infection of B6 and D2 mice with M. tuberculosis results in massive and complex changes in transcript profiles in the infected lungs. The acquisition of a subset of these transcript signatures (expression QTL) in the Trl3 congenic line, caused by the transfer of the chromosome 7 segment overlapping this locus, may give clues as to the mechanisms or genes associated with the partial gain of resistance seen on the D2.B6-Chr7 congenic line. Such signatures (expression QTL) may be detected by pairwise comparisons of transcript profiles induced by infection in both B6 and D2.B6-Chr7 mouse lines but distinct from the parental D2 line. With the microarray platform and statistical analysis implemented in our study, we have been able to detect chromosome-specific expression QTL signatures independently transferred in either the chromosome 7 or chromosome 19 congenic lines. Briefly, pairwise analysis (p < 0.01 by t test; fold change twice or more) comparing B6 to D2 mice at day 0 revealed 565 differentially expressed genes, including 46 (8.1%) genes on chromosome 7 and 17 (3.0%) genes on chromosome 19 (see supplemental Table S3). 4 As expected from their D2 genetic background, similar comparisons of transcript profiles from uninfected lungs of D2.B6-Chr7 and D2.B6-Chr19 mice to D2 revealed lower numbers of differentially expressed genes. In D2.B6-Chr19 mice, 18 of the 104 (17.3%) regulated transcripts mapped to chromosome 19, whereas in D2.B6-Chr7 mice, 24 of the 69 (34.8%) regulated transcripts mapped to chromosome 7, representing a considerable enrichment for transcripts mapping to congenic segments introgressed in these strains (see supplemental Table S3). 4 Similar results were obtained when the analysis was conducted with RNA from infected lungs. Pairwise comparison of transcript profiles from day-30 infected lungs (D2 vs D2.B6-Chr7 and D2 vs D2.B6-Chr19) showed that 37.0% and 37.9% of the differentially regulated transcripts mapped to chromosome 7 and chromosome 19 for D2.B6-Chr7 and D2.B6-Chr19 congenics, respectively (see supplemental Table S3). 4 Of interest, 52.2% (day 0) and 53.3% (day 30) of transcripts coregulated between B6 and D2.B6-Chr19 mapped to chromosome 19, whereas 84.2% (day 0) and 78.6% (day 30) of those coregulated in B6 and D2.B6-Chr7 mapped to chromosome 7 (see supplemental Fig. S4), 4 again showing a dramatic enrichment for cis-regulated genes mapping to the respective congenic segment. The identity (see supplemental Table S4) and physical position of the annotated chromosome 7 genes showing coregulation in B6 and D2.B6-Chr7 mice at day 0 (see supplemental Fig. S5A) or day 30 (see supplemental Fig. S5B) are described. 4
Of the transcripts commonly regulated in B6 and D2.B6-Chr7 (in response to infection) when compared with D2 mice (ANOVA two-by-two interaction), several belong to the ontology class “immune system process.” Moreover, seven of these transcripts commonly regulated in B6 and D2.B6-Chr7 map within the boundaries of the chromosome 7 congenic segment transferred in D2.B6-Chr7 mice, including Cd22. Cd22 is up-regulated in response to M. tuberculosis infection at day 30 in the lungs of resistant B6 and intermediately resistant D2.B6-Chr7 but is not up-regulated in susceptible D2 (Fig. 6⇑B). CD22 belongs to the SIGLEC family of lectins (35). CD22 functions as an inhibitory receptor for BCR signaling, and has been suggested to play a role in preventing overactivation of the immune system, acting as a molecular switch controlling the fate of Ag-stimulated B cells, and whether they undergo apoptosis or proliferation (36). Although little is known about the role played by B cells in TB pathogenesis, recent studies have suggested that they may play a previously unappreciated role in local immunity (37, 38). A recent study revealed that upon aerosol infection with 100 CFU of M. tuberculosis Erdman, B cell-deficient mice display increased pulmonary bacterial loads, and exacerbated immunopathology associated with elevated pulmonary recruitment of neutrophils (39), which is also the case in D2 mice following long-term exposure to the same pathogen (28). In these studies, increased susceptibility of B cell-deficient mice was not linked to a diminished capacity to produce IFN-γ in the lungs (39). Additional experimentation will be required to determine whether CD22 and its known role in controlling B cell hypoactivity or hyperactivity (40, 41) may play a role in host defense against TB in general and in the Trl3-mediated effect in particular.
Although mapping far away from the congenic segment containing the Trl3 locus, coronin 1a (also known as TACO) is an additional excellent candidate for Trl3. It is an actin-binding protein recruited to the membrane of the maturing phagosome, including those containing mycobacteria. Its modulation has strong consequences on the maturation of bacteria-containing phagosomes in macrophages (42, 43). We believe differential expression of coronin 1a in D2 vs B6 and D2.B6-Chr7 congenics may be caused by a trans-effect caused by genes within the congenic segment. The same would apply to other differentially regulated genes mapping outside the Trl3 congenic segment on chromosome 7 (Table II⇑ and Fig. 6⇑A) or mapping to other chromosomes (Table I⇑). Therefore, although differential expression of coronin 1a may contribute to differential susceptibility to TB, we believe it may be an effect secondary to the primary effect that must be caused by genes within the congenic segment.
Although the gene responsible for the Trl3 effect remains unknown, it is interesting to note that chromosome 7 has been previously detected as modifying the sst1-controlled interstrain difference in susceptibility of C57BL/6J (resistant) and C3HeB/FeJ (hypersusceptible) mice strains to pulmonary TB (14). In congenic lines constructed between these two strains, transfer of the sst1r or sst1s alleles on C3HeB/FeJ (C3H.B6sst1r) and B6 (B6.C3Hsst1s) genetic backgrounds, respectively, only partly confers resistance or susceptibility, suggesting that other genetic loci are responsible for the phenotype of parental strains. Such genetic modifiers were investigated by whole genome scan in (C3H.B6sst1r X B6)F2 mice infected with M. tuberculosis (105, i.v. route). Four additional loci mapping to chromosomes 7 (proximal, LOD 4.8), 12 (distal, LOD 6.6), 15 (distal, LOD 4.6), and 17 (proximal, LOD 5.5) were shown to influence survival to infection. Although the relationship between Trl3 and the chromosome 7 locus mapped by Yan et al. (14) remains to be validated, the maximum peak of linkages defining both loci show significant overlap, suggesting that they may represent the same genetic effect acting in different strain combinations.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by Grant AI035237 (to P.G.) and Grant AI069161 (to R.J.N.) from National Institute of Allergy and Infectious Diseases, National Institutes of Health. P.G. is a James McGill Professor of Biochemistry and a Distinguished Scientist of the Canadian Institutes of Health Research. J.-F.M. is supported by a fellowship from the Fonds de Recherche en Santé du Québec.
↵2 Address correspondence and reprint requests to Dr. Philippe Gros, Department of Biochemistry, McGill University, Bellini Building, 3649 Promenade Sir William Osler, Room 370, Montréal, Québec H3G 0B1, Canada. E-mail address:
↵3 Abbreviations used in this paper: TB, tuberculosis; MST, mean survival time; NOS, NO synthase; QTL, quantitative trait locus.
↵4 The online version of this article contains supplemental material.
- Received June 26, 2008.
- Accepted January 9, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.