Identification of novel susceptibility genes in ozone-induced inflammation in mice

Abstract

Ozone (O₃) remains a prevalent air pollutant and public health concern. Inf2 is a significant quantitative trait locus (QTL) on murine chromosome 17 that contributes to susceptibility to O₃-induced infiltration of polymorphonuclear leukocytes (PMNs) into the lung, but the mechanisms of susceptibility remain unclear. The study objectives were to confirm and restrict Inf2, and to identify and test novel candidate susceptibility gene(s).

Congenic strains of mice that contained overlapping regions of Inf2 and their controls, and mice deficient in either MHC Class II genes or the Tnf cluster were exposed to air or O₃. Lung inflammation and gene expression were assessed.

Inf2 was restricted from 16.42 Mbp to 0.96 Mbp, and bioinformatic analysis identified MHC class II, the Tnf cluster, and other genes in this region that contain potentially informative SNPs between the susceptible and resistant mice. Furthermore, O₃-induced inflammation was significantly reduced in mice deficient in MHC class II genes, or the Tnf cluster genes, compared to wild-type controls. Gene expression differences were also observed in MHC class II and Tnf cluster genes.

This integrative genetic analysis of Inf2 led to identification of novel O₃ susceptibility genes that may provide important, new therapeutic targets in susceptible individuals.

Introduction

The pollutant ozone (O₃) is a highly toxic principal oxidant found in urban environments throughout industrialized cities worldwide. O₃ exposure has been associated with many adverse health effects, such as exacerbation of asthma [1, 2]. Identification of susceptibility genes involved in O₃-induced pulmonary injury may provide critical information for future risk assessment as well as general international health policies. In 2006, an estimated one-third of U.S. individuals were at an increased risk of adverse effects caused by O₃ and 131 million U.S. residents resided in regions that either approached or exceeded the National Ambient Air Quality Standard of 0.08 parts per million (ppm) O₃ [3].

Significant intersubject differences in pulmonary function and inflammatory responses to O₃ suggest that genetic background contributes to O₃ susceptibility in humans and rodents [4–6]. Furthermore, studies in human subjects have suggested that polymorphisms in oxidant defense genes, such as glutathione S transferase M1, and quinone metabolism genes, such as NADPH quinone oxidoreductase 1, associate with differential responsiveness to O₃ [7, 8]. Activating polymorphisms in inflammatory genes, such as tumor necrosis factor alpha (TNF), also enhance susceptibility to O₃ and asthma [9]. A genome-wide linkage analysis of O₃-induced influx of polymorphonuclear leukocytes (PMNs) in an intercross cohort (B6C3F₂) derived from susceptible C57BL/6 (B6) and resistant C3H/HeJ (C3) progenitor mouse strains identified a significant susceptibility QTL on chromosome 17 (inflammation 2; Inf2) [5]. Inf2 (33.73–50.15 Mbp; D17Mit16 – D17Mit10) contains the H-2 locus, including major histocompatibility genes and non-MHC genes such as the pro-inflammatory cytokine Tnf. Pre-treatment of B6 mice with a monoclonal antibody to TNF-α and deletion of TNF-α receptors 1 and 2 [5, 10] significantly attenuated the inflammatory response to O₃ relative to control mice, supporting the importance of Tnf as a candidate gene in this model. However, these studies did not conclusively identify Tnf as the susceptibility gene in Inf2. In addition, Inf2 is located in the most gene dense, polymorphic region of the entire mouse genome, thus making single candidate gene identification difficult [11].

In the current study we sought to confirm the importance of Inf2 in O₃-induced lung inflammation, to reduce Inf2, and to identify candidate susceptibility genes. Using an integrative genomics approach, we utilized congenic mouse strains to limit Inf2 from 16.42 Mbp to 0.96 Mbp, and bioinformatic analysis which identified MHC class II genes and the entire TNF cluster as candidate susceptibility loci. Functional analyses of these genes confirmed novel roles for modulation of the inflammatory response to O₃ exposure.

Materials and Methods

Mouse strains and O₃ exposure

The following male (6-8 wk) congenic mice were used: B10.A- H2^h2/(2R)SgSnJ (2R); B10.A-H2^h4/(4R)SgDvEg (4R), B10.A-H2ⁱ⁵H2-T18^a/(5R)SgSnJ (5R); and C3.SW-H2b/SnJ (C3H-H2^b). The congenic region and haplotype for each of the strains are shown in Table 1 and Figure 1. Control strains for the congenic mice were C57BL/10SnJ (B10), A/WySnJ (A), and C3H/HeSnJ (HeSnJ). The location of the congenic region of the 2R, 4R, and 5R mice with respect to the B10 and A/Wy background strains was identified by genotyping (Malhotra and Travis, data not shown). H2^k haplotypes are O₃-resistant and the H2^b haplotype is O₃-susceptible [5] . Additional strains used for the candidate gene studies were B6.129-H2^dlAb1-Ea/J (H2-^dlAbl-Ea; B6 background strain), B6.129S2-Lta^tm1Dch/J (Lta^−/−, B6 background), and B6.129P2-Ltb/Tnf/Lta^tm1Dvk/J (Ltb/Tnf/Lta^−/− mice, B6 background), C57BL/6J (B6) mice, and C3H/HeJ (C3) mice. B6 are O₃-susceptible and are H2^b, while C3H/HeJ (C3) mice are O₃-resistant and are H2^k. All mice were purchased from Jackson Laboratory (Bar Harbor, ME). All animal use was conducted in facilities accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care and approved by the NIEHS Animal Care and Use Committee and follows the Helsinki convention for the use and care of animals. For air and O₃ exposures, mice were caged in a humidity and temperature-controlled room and provided water and pelleted open-formula rodent diet NIH-07 (Zeigler Brothers, Gardners, PA) ad libitum (see O₃ exposure procedures in Supplement Methods).

Table 1.

Nomenclature and characteristics of the congenic and control strains used for investigation.

		Congenic Region
Strain Name	Abbreviation	Mbp	cM	H2 Haplotype**
Congenic strains
B10.A-H2h2/(2R)SgSnJ	2R	33.74 – 35.34	18.20 – 19.06	H2h2
B10.A-H2h4/(4R)SgDvEgJ	4R	29.09 – 34.38	16.30 – 18.64	H2h4
B10.A-H2i5 H2-T18a/(5R)SgSnJ	5R	34.45 – 36.35	18.70 – 19.14	H2i5
C3.SW-H2b/SnJ	C3H-H2b	33.74 – 43.80	18.20 – 23.30	H2bc
Control strains
C57BL/10SnJ†	B10	- - -	- - -	H2b
A/WySnJ ‡	A	- - -	- - -	H2a
C3H/HeSnJ++	HeSnJ	- - -	- - -	H2k

^*Mbp and cM were determined using the MGI website:http://www.informatics.jax.org/searches/marker and genotyped by Malhotra and Travis, unpublished).

^**MHC H2 Haplotypes according to Appendix 2: MHC H2 haplotypes (www.jax.org/jaxmice/search).

^†background strain for 2R, 4R, and 5R mice ,

^‡donor strain for 2R, 4R, and 5R mice.,

⁺⁺background strain for C3H-H2^b mice.

Figure 1. Schematic of the chromosome 17 congenic segments for 2R, 4R, 5R, and C3H-H2^b mice that were used to reduce inflammation 2 (Inf2).

The congenic segments are shown with respect to Inf2 (on left). The red rectangle represents the reduced Inf2. Mouse genes and human homologues and their chromosomal locations within the reduced Inf2 are shown at right. While most of the genes are in the same order in both species, some are not. These locations were identified using the MGI website:http://www.informatics.jax.org/searches/marker website and NCBI. The yellow highlighted genes are major histocompatibility complex (MHC) class II genes and TNF cluster genes identified as candidate genes using proof of concept experiments (see Figs. 5 and 6). Abbreviations:Ager, advanced glycosylation end product-specific receptor; Agpat1; 1-acylglycerol-3-phosphate O-acyltransferase 1; Aif1, allograft inflammatory factor 1; Atf6b, activating transcription factor 6 beta; Bat5, HLA-B associated transcript 5; Btnl2, butyrophilin-like 2; Btnl1, butyrophilin-like 1; C2, complement component 2; C4a, complement component 4A; C4b, complement component 4B; Cfb, complement factor B; CREBL1, cAMP responsive element binding protein-like 1; Cyp21a1, cytochrome P450, family 21, subfamily a, polypeptide 1; Cyp21a2-ps, cytochrome P450, family 21, subfamily a, polypeptide 2 pseudogene; Ddah, dimethylarginine dimethylaminohydrolase 2;Egfl8, EGF-like domain 8; Ehmt2, euchromatic histone-lysine N-methyltransferase 2; Fkbpl, FK506 binding protein-like; Gpsm3, G-protein signalling modulator 3; H2-Ab1, class II antigen A, beta 1; H2-Aa, class II antigen A, alpha; H2-Eb1, class II antigen E beta; H2-Eb2, class II antigen E beta 2; H2-Ea, class II antigen E alpha; H2-ob, O region beta locus; Hspa1b, heat shock 70kDa protein 1B; Hspa1a, heat shock 70kDa protein 1A; Lsm2, LSM2 homolog; Lta, lymphotoxin a; Ltb, lymphotoxin b; Ly6g6c, lymphocyte antigen 6 complex, locus G6C; Ly6g5c, lymphocyte antigen 6 complex, locus G5C; Msh5, mutS homolog 5;Neu1, neuraminidase1; Notch4, notch gene homolog 4; Pbx2, pre B-cell leukemia transcription factor 2; Ppt2, palmitoyl-protein thioesterase 2; Prrt1, proline-rich transmembrane protein 1; Psmb8, proteosome subunit, beta type 8; Rdbp, RD RNA-binding protein; Rnf5, ring finger protein 5; Slc44a4; solute carrier family 44, member 4; Stk19, serine/threonine kinase 19;Tap-1, transporter 1, ATP-binding cassette, sub-family B (MDR/TAP); Tap-2, transporter 2, ATP-binding cassette, sub-family B (MDR/TAP); Tnf, tumor necrosis factor; Tnxb, tenascin B; Vars, valyl-tRNA synthetase; Zbtb12, zinc finger and BTB domain containing 12. No human homolog exists for Btnl1.

Bronchoalveolar lavage fluid (BALF) analysis

The procedures used for these techniques have been described previously for right lung lavages and inflammatory cell analysis [5, 10, 12].

Total RNA Isolation and Real-Time Quantitative Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)

Mice were sacrificed immediately after air or O₃ exposure, and lungs of each animal were snap-frozen in liquid nitrogen. Total RNA isolation, reverse transcription into cDNA and PCR reaction procedures are described in Supplemental Methods.

Statistics

Data are expressed as group means ± standard error of the mean (S.E.M.). Two-way analysis of variance (ANOVA) was used to evaluate the effects of exposure (air vs O₃) and strain (B10, 2R, 4R, 5R, A, HeSnJ, C3H-H2^b, H2-^{dlAbl-Ea−/−}, H2-^dlAbl-Ea+/+, Lta^−/−, Lta^+/+, Lta/Tnf/Ltb^−/−, Lta/Tnf/Ltb^+/+, B6, C3) on BAL phenotypes and mRNA expression. The figure legends contain the number of mice used per experiment. Student-Newman-Keuls test was used for a posteriori comparisons of means. All analyses were performed using a commercial statistical analysis package (SigmaStat; Jandel Scientific Software, San Rafael, CA). Statistical significance was accepted at P<0.05.

Results

Restriction of Inf2

Responses to O₃ in congenic mice

To confirm the significance of Inf2, C3H-H2^b mice which contain Inf2 from a susceptible mouse (H2^b) on an O₃-resistant H2^K background [HeSnJ; (see Figure 1)] were exposed to air and O₃. Relative to HeSnJ controls, significant increases in mean BAL protein concentration and numbers of macrophages, PMNs, and epithelial cells were found in C3H-H2^b mice after 48 and 72 h exposure to O₃ (Figure 2); no strain differences were observed in any parameter in the air exposed mice. Results thus confirmed the importance of the region of chromosome 17 encompassed by Inf2 in susceptibility to O₃-induced inflammation.

Figure 2. Bronchoalveolar lavage fluid (BALF) inflammatory parameters in the C3H-H2b and HeSnJ mice exposed continuously to 0.3 ppm O₃ or air for 48 or 72 h.

A. Macrophages; B. Polymorphonuclear leukocytes (PMNs); C. Epithelial cells; D. Total BALF protein. Data are presented as means ± SEM (n=3-5 per experimental group). *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from HeSnJ mice (P < 0.05).

To reduce the length of Inf2, O₃-induced inflammatory responses in congenic 2R, 4R, and 5R mice were compared to those of similarly exposed O₃-susceptible background B10 and O₃-resistant donor A strains. Each congenic strain contains a different region of the Inf2 locus from the A strain mouse on a B10 background (see Table 1). O₃ caused significant increases in mean numbers of macrophages and PMNs in B10 but not A mice compared to respective air-exposed animals, and numbers of macrophages and PMNs were greater in B10 mice compared to A mice after O₃ (Figure 3). The mean numbers of PMNs and macrophages in 2R and 5R mice were not different from A strain mice after 48 and 72 h O₃; however, numbers of PMNs and macrophages in 2R and 5R mice were significantly lower compared to B10 mice (Figure 3). In contrast, the numbers of BALF macrophages and PMNs from O₃-exposed 4R mice were not significantly different from those from O₃-exposed B10 mice, and were greater than those from A, 2R, and 5R mice (Figure 3). The mean BALF protein concentrations were significantly reduced in 2R, 4R, and 5R mice compared to B10 mice and significantly greater than A mice after 48 and 72 h O₃. Comparison of the mean numbers of macrophages and PMNs between congenic strains with respective congenic regions thus suggested that the 34.38–35.34 Mbp [18.64–19.06 centimorgan (cM)] region of chromosome 17 accounted for a major portion of the inflammatory response to O₃ in these mice (Figure 1). No differences in mean numbers of BALF epithelial cells were found between the congenic, A, or B10 strains after exposure to O₃ (Figure 3C).

Figure 3. Bronchoalveolar lavage fluid (BALF) inflammatory parameters in B10, 4R, 5R, 2R, and A/Wy mice exposed continuously to 0.3 ppm O₃ or air for 48 or 72 h.

A. Macrophages; B. Polymorphonuclear leukocytes (PMNs); C. Epithelial cells; D. Total BALF protein. Data are presented as means ± SEM (n=3-12 per experimental group). *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from B10 mice (P < 0.05).

Candidate Gene Analysis

Database search for polymorphisms

Genes within the reduced region of Inf2 (34.38–35.34 Mbp) were searched for known exon (non-synonymous), intron (including splice sites) and UTR polymorphisms using the public Mouse Phenome Database [13] (Supplemental Table 1). H2-Aa and H2-Ab1 genes are haplotype H2^k in the 2R and 4R congenic strains, thus excluding these as candidate genes [14]. Additionally, the H2-Eb1 gene is divided between resistant and susceptible haplotypes and therefore remains a candidate gene, as well as H2-Eb2 [15]. The genes that contain known non-synonymous polymorphisms between the B6 (susceptible) and C3 (resistant) mice in the reduced Inf2 in order of location on chromosome 17 are H2-Eb1, H2-Eb2, Btnl1, Notch4, Fkbpl, Crebl1, Tnxb, C4b, C4a, and Stk19. Two genes (Hspa1b and Hspa1a) may have exonic polymorphisms between the B6 and C3, but no details are currently available for the B6 strain at these sites. Genes with intronic polymorphisms between the B6 (susceptible) and C3 (resistant) mice in the reduced Inf2 are: H2-Eb1, H2-Eb2, H2-Ea, Btln2, Btln1, Btln7, Notch4, Ppt2, Crebl1, Tnxb, C4b, Cyp21a1, C4a, Stk19, Cfb, C2, Msh5, Ddah2, Ly6g6c, Bat5, Ly6g5c, Csnk2b, Aif1, Tnf, and Lta (Supplemental Table 1). UTR polymorphisms between the B6 and C3 mice were found in H2-Eb2, Btnl2, Ppt2, C4b, C4a, G6b, Bat5, Ly6g5b, and Tnf (Supplemental Table 1). Information is absent for Ltb and Tnf UTR and intronic polymorphisms in the C3 strain, thus it is possible that more exist between the B6 and C3 strains.

We then asked if mRNA expression differences between B6 and C3 mice existed for several candidate genes within the reduced Inf2, and with either known or possible SNPs (Fig. 4). H2-Eb1 expression was reduced after 24 h of O₃ exposure in the C3 mice, but unchanged in the B6 mice; H2-Eb2 expression was significantly increased only in B6 mice at 24 h. Expression of Hspa1b and Btnl1 was significantly increased after 72 h O₃ in B6 mice only. Tnf and Lta expression was significantly increased in B6 mice after 24 h O₃ compared to no increases observed in C3 mice.

Figure 4. Gene expression for candidate genes identified within the reduced Inf2 region of B6 (susceptible) and C3 (resistant) mice.

Six genes were analyzed using quantitative RT-PCR (H2Eb1, H2Eb2, Hspa1b, Btnl1, Tnf, Lta). Data are presented as means ± SEM (n=3–7 per experimental group) and were determined using the comparative C_T method (see Supplemental Materials and Methods section). Y-axis represents the gene of interest normalized first to 18S followed by determination of the fold-change relative to the average air control value for each strain. *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from B6 mice (P < 0.05).

MHC class II deficient mice

Because of the prevalence of MHC class II genes in Inf2 and the non-synonymous, intronic, and UTR polymorphisms in H2-Eb1 and -Eb2 between the susceptible (B6) and resistant strains (C3) (Supplemental Table 1), we hypothesized that they were candidate susceptibility genes for O₃-induced inflammation. H2Abl-Ea^−/− mice are deficient in MHC class II genes H2-Ab1, -Aa, -Eb1, -Eb2, and -Ea compared to the H2Abl-Ea^+/+ mice. The B6 background of these mice is null at H2-Ea, therefore H2Abl-Ea^−/− mice only test the importance of the H2-Ab1 through H2-Eb2 MHC class II region. Relative to respective air controls, O₃ caused significant (P < 0.05) increases in mean numbers of BALF PMNs and macrophages in H2Abl-Ea^+/+ mice after 24, 48 and 72 h, and after 24 and 48 h O₃ in H2Abl-Ea^−/− mice (Fig. 5A and Table 2). However, significantly greater numbers of PMNs and macrophages were found in H2Abl-Ea^+/+ mice compared to H2Abl-Ea^−/− mice after 24–72 h and 48 and 72 h exposures, respectively. Significant strain and O₃ effects on BAL epithelial cells were found after 48 h O₃ (Table 2). Significant strain differences in BAL protein content were also found after 48 h O₃ (Fig. 5A). mRNA expression of Tnf and Lta was significantly increased in the H2Abl-Ea^+/+ mice after 24 h O₃; however no increase was observed in the H2Abl-Ea^−/− mice (Fig. 5B).

Figure 5. Deficiency in MHC class II genes significantly reduced O₃-induced responses.

A. Bronchoalveolar lavage fluid (BALF) polymorphonuclear leukocytes (PMNs) and total protein in H2Ab1-Ea^−/− and H2Ab1-Ea^+/+ mice exposed continuously to 0.3 ppm O₃ or air for 48 or 72 h. Data are presented as means ± SEM (n=3-10 per experimental group). *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from H2Ab1-Ea^+/+ mice (P < 0.05). B. Gene expression for Lta and Tnf genes in H2Ab1-Ea deficient mice compared to wildtype mice. Data are presented as means ± SEM (n=3-6 per experimental group) and were determined using the comparative C_T method (see Supplemental Materials and Methods section). Y-axis represents either Lta or Tnf normalized first to 18S followed by determination of the fold-change relative to the average air control value for each strain. *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from H2Ab1-Ea^+/+ mice (P < 0.05).

Table 2.

Effect of deletion of H2-Abl-Ea, Lta, or Lta/Tnf/Ltb on macrophages and epithelial cells recovered from BAL following O₃ exposure.

		No. of cells (X 103 ml/BAL)
			O3 exposed
Cell type	Genotype	Air*	24 h	48 h	72 h
Macrophages	H2Abl-Ea+/+	19.6 ± 0.1	33.6 ± 3.9**	48.1 ± 0.4**§	37.9 ± 3.7**§
	H2Abl-Ea −/−	19.0 ±0.1	38.5 ± 6.3**	33.2 ± 0.4**	22.5 ± 1.3
	Lta+/+	12.1 ± 1.2	22.2 ± 3.9**§	29.7 ± 7.7**§	37.9 ± 3.7**
	Lta−/−	13.8 ±1.9	7.6 ± 2.2	21.2 ±2.7	32.0 ± 2.5**
	Lta/Tnf/Ltb+/+	18.3 ± 3.4	26.7 ± 3.0	36.8 ± 3.9**§	38.5 ± 3.4**§
	Lta/Tnf/Ltb−/−	26.6 ± 3.2	32.7 ± 3.7	25.2 ± 3.0	24.4 ± 3.9
Epithelial cells	MHC+/+	1.4 ± 0.3	1.0 ± 0.2	2.6 ± 0.7**§	1.0 ± 0.3
	MHC−/−	0.1 ± 0.2	1.8 ± 0.6	0.9 ± 0.1	0.9 ± 0.2
	Lta+/+	0.7 ± 0.2	0.85 ± 0.3	1.4 ± 0.2	1.0 ± 0.3
	Lta−/−	0.4 ± 0.2	0.5 ± 0.2	1.2 ± 0.3	1.7 ± 0.2
	*Lta/Tnf/Ltb+/+	1.0 ± 0.5	0.6 ± 0.04	1.5 ±0.3§	1.8 ± 1.5
	*Lta/Tnf/Ltb−/−	1.8 ± 1.5	1.0 ± 0.3	0.6 ± 0.05	1.7 ± 0.3

Data are presented as mean ± SEM, n = 3 – 10 mice per treatment group.

^*Air controls shown are for the 72 hr time point except for the Tnf ^+/+ and Tnf ^−/− mice (48hr time point shown). Air controls for each time point were not significantly different from one another.

^**p < 0.05, significantly elevated numbers of cells compared to air controls.

^§p < 0.05, significantly elevated numbers of cells compared to the transgenic strain.

Mice deficient in Lta or the Tnf cluster

Based on the existing evidence to support a role for the TNF cluster in the O₃ response [10], as well as the known intronic and UTR polymorphisms in Tnf and Lta genes between the B6 and HeJ mice, we tested the role of these three genes in the model. Lta^−/−, Lta/Tnf/Ltb^−/−, and wildtype mice were exposed continuously to air or 0.3 ppm O₃ for 24, 48 or 72 h. PMNs were significantly elevated in the Lta^+/+ and Lta/Tnf/Ltb^+/+ mice compared to mice deficient in Lta^−/− or the Lta/Tnf/Ltb cluster following 24 and 48 h O₃ for both strains, and 72 h exposure for Lta/Tnf/Ltb deficient mice compared to air controls (p<0.05; Fig. 6A, B). Mean numbers of BALF macrophages were also significantly elevated in wildtype mice compared to Lta^−/− and Lta/Tnf/Ltb^−/− mice following 24–72 h O₃ compared to air controls (p<0.05; Table 2). Epithelial cell numbers were not different between exposure or strains (Table 2). The BAL protein content was not affected by these genes, except for 72 h exposure in the Lta^−/− mice, supporting the hypothesis that different genes regulate these specific phenotypes [16] (Fig. 6A, B).

Figure 6. O₃-induced responses in mice deficient in Lta or the Tnf cluster (Lta, Tnf, Ltb) genes.

A) BALF PMNs and total protein for mice deficient in Lta and their wildtype controls. Data are presented as means ± SEM (n=3–10 per experimental group). *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from Lta^+/+ mice (P < 0.05). B. BALF inflammatory parameters for the Lta/Tnf/Ltb deficient mice in response to O₃ compared to wildtype mice. Data are presented as means ± SEM (n=3–10 per experimental group). *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from Lta/Tnf/Ltb^+/+ mice (P < 0.05). C) Gene expression for the MHC class II genes (H2-Eb1 and H2-Eb2) in the mice deficient in either Lta or Lta/Tnf/Ltb compared to the wildtype mice. B6 are the wildtype for both strains and the numbers were combined since they were not significantly different from one another. No 24 h samples for Lta/Tnf/Ltb^−/− mice were done for gene expression analysis. Data are presented as means ± SEM (n=3–6 per experimental group) and were determined using the comparative C_T method (see Supplemental Materials and Methods section). Y-axis represents either H2-Eb1 or H2-Eb2 normalized first to 18S followed by determination of the fold-change relative to the average air control value for each strain. *, Significantly different from air exposed mice (P < 0.05). ⁺, Significantly different from wildtype mice (P < 0.05).

MHC class II gene expression was also significantly different between the strains deficient in the TNF cluster. H2-Eb1 mRNA was unchanged in B6 wildtype mice, however, this gene was significantly decreased after 48 h O₃ in the Lta^−/− and Lta/Tnf/Ltb^−/− mice compared to controls (Fig. 6C). H2-Eb2 mRNA was significantly increased in B6 mice after 24 and 48 h O₃, and was significantly greater after O₃ than mRNA expression in Lta^−/− and Lta/Tnf/Ltb^−/− mice at these time points.

Discussion

The overall objective of this investigation was two-fold. We first sought to restrict the length of Inf2 to elucidate candidate susceptibility genes for O₃-induced inflammation. The second objective was to validate the role(s) of identified gene candidates. Significant differences in O₃-induced inflammation between C3.SW-H2b/SJ and C3H/HeSnJ strains confirmed the role of Inf2 inasmuch as the only difference between these two strains is a congenic region of chromosome 17 that encompasses Inf2. Comparison of inflammatory responses in 2R, 4R, and 5R congenic mice with respective background strains further confirmed the importance of Inf2, and more importantly reduced Inf2 to 0.96 Mbp. The reduced Inf2 includes MHC class I, II, and III genes, and non-MHC genes, some of which have previously been identified as candidate genes in other lung injury models [e.g. C4a; [17]. In the current study, gene expression and sequence analyses of the reduced Inf2 suggested that MHC class II genes and the TNF cluster may be important in O₃-induced inflammation, and significantly different phenotypes between O₃-exposed H2Abl-Ea^+/+ and H2Abl-Ea^−/− mice and the mice deficient in Lta alone or the entire Tnf cluster confirmed a role for these genes. To our knowledge, these are the first studies to conclusively demonstrate a role for MHC class II genes and the entire TNF cluster in oxidant-induced lung inflammation and provide evidence supporting a susceptibility “superlocus”.

Genetic association studies have implicated several gene categories, such as pro-inflammatory cytokine genes [9], metabolism genes [18], and innate immunity genes in responses to environmental stimuli in human populations [19]. In inbred mouse models, positional cloning approaches identified QTLs for a number of lung diseases [20]. However, while identification of QTLs is an important initial step to understanding the genetic determinants of disease, QTLs can contain hundreds of genes. It is therefore necessary to refine the disease QTLs to a limited set of genes that can be evaluated [21]. In our integrative genomics approach to reduce Inf2 and identify candidate genes, we used congenic strains that were developed initially for histo-incompatiblity studies by repeated backcrossing of regions of chromosome 17 from the donor strain onto the recipient background strain [22]. One advantage of these strains is that the allelic designations are well characterized and, after phenotyping these strains for their inflammatory response to O₃, it enabled restriction of the previously described Inf2 QTL to a limited chromosomal interval to identify gene candidates.

We queried the Mouse Phenome Database to identify genes located within the narrowed Inf2 that are polymorphic between susceptible B6 and resistant C3 mice. Several of the MHC class II and III genes (and those non-MHC genes located within these regions) have non-synonymous polymorphisms in exons, many have intronic polymorphisms, and several have UTR polymorphisms. However, not all of the genes have been assessed for polymorphisms between B6 and C3 mice (e.g. Hspa1b). It is important to note that functional polymorphisms may also exist in intronic regions, such as with K-ras and mouse lung cancer [23]. Future studies will delineate the functionality of polymorphic regions between the B6 and C3 mice in those genes in which the functionality is unknown, such as Ltb.

Organization of the H2 locus

Genes in the mouse H-2 locus and the homologous human MHC locus are organized in the same relative gene order [24] and span several Mbp. The H2 locus contains class I, class II, and class III genes. Class I and class II genes are important in fulfilling immunologic functions. Class I molecules are expressed by most cells, and present endogenous antigens (cytosolic-derived) to CD8⁺ cytotoxic T cells. Class II molecules (e.g. H2-Eb2, Btnl1) are expressed in antigen presenting cells (APC) such as dendritic cells, and present exogenous, endocytically derived antigens to CD4⁺ helper T-cells [25]. Class III genes are diverse in functionality, are located between the class I and class II genes and include the complement component genes (C2 and C4), heat shock proteins (Hspa1a, Hspa1b, and Hsc70t), and the Tnf cluster.

Candidate gene analysis

Deficiency in MHC class II genes and the TNF cluster decreased the O₃-induced PMN and macrophage phenotypes, similar to those reductions seen in the 2R and 5R mice. Previous studies as well as the data presented here support a clear role for the TNF cluster in these O₃-induced phenotypes. The H2-Ea gene (which is deficient in B6 mice), H2-Ab1, and H2-Aa (as described earlier) can be excluded as candidate susceptibility genes.

Chen et al. [26] demonstrated that inhibition of the CD4⁺ T-cell population using a monoclonal CD4 antibody in vivo greatly reduced the O₃-induced phenotypes (including PMNs, lymphocytes, and epithelial cells) in B6 mice after 72 h continuous O₃ exposure. These studies support our findings in the H2Abl-Ea^−/− mice and suggest that in O₃ susceptible strains, CD4+ T- cells are critical to O₃-induced inflammation and injury. A recent study in human airway monocytes isolated from exposed individuals suggests that O₃ primes the airway monocytes for innate immune responses, increases the capacity of the monocytes to present antigen to the CD4⁺ T-cells, and increases the overall population of antigen presenting cells in the lungs [19].

Candidate susceptibility genes in Inf2 and human disease

The MHC class II genes and the TNF cluster in the reduced Inf2 are all located in close proximity on chromosome 17 in mice and chromosome 6 in humans (Fig. 1, Supplement Table 1). Several of these homologous genes have been implicated in human association studies. For example, polymorphisms in HLA-DRB1 (H2-Eb1), TNF, and LTA have been associated with sarcoidosis [27], a chronic granulomatous disease of unknown etiology, supporting a role for a susceptibility “superlocus”. In addition, TNF is a susceptibility gene for O₃-induced changes in lung function in humans [28]. Thus, it is possible that a cluster of genes in the MHC class II region (H2-Eb1 and H2-Eb2) interacts with the Tnf cluster (Tnf, Lta and Ltb) to promote O₃-induced lung inflammation and injury. In support of this notion, the MHC (specifically HLA genes) has the strongest influence on susceptibility to human autoimmune diseases, but recent evidence suggests that TNF may also be involved in autoimmune diseases [29].

In summary, this integrative genomics investigation utilized congenic mouse lines to narrow the chromosome 17 QTL for susceptibility to O₃-induced inflammation in the mouse. We used sequence analysis to identify candidate genes in the narrowed QTL, which were then tested for proof of concept. The novel role of MHC Class II genes and the TNF cluster in susceptibility to O₃-induced inflammation provides unique insight to the mechanisms of O₃ effects in the lung and may lead to alternative means to prevent oxidant injury to lung tissues.