|Home | About | Journals | Submit | Contact Us | Français|
Alpha-tocopherol transfer protein (ATTP) null mice (ATTP−/−) have a systemic alpha-tocopherol (AT) deficiency, with their lung AT levels being < 10% of those in AT-replete ATTP+/+ mice when fed a standard rodent chow diet. ATTP+/+ and ATTP−/− mice (4 wk old male mice, n = 16 per group) were fed a standard diet (35 IU AT/kg diet) for 8–12 wk, exposed 6 h/day for 3 days to either to O3 (0.5 ppm) or filtered air, then sacrificed. No significant differences in plasma or lung AT concentrations were observed in response to this level of O3 exposure. Lung genomic responses of the lungs to O3 were determined using Affymetrix 430A 2.0 arrays containing over 22,600 probe sets representing 14,000 well-characterized mouse genes. As compared with filtered air exposure, O3 exposure resulted in 99 genes being differentially expressed in ATTP−/− mice, as compared to 52 differentially expressed genes in ATTP+/+ mice. The data revealed an O3-induced upregulation of genes related to cell proliferation/DNA repair and inflammatory-immune responses in both ATTP+/+ and ATTP−/− mice, with the expression of 22 genes being common to both, whereas 30 and 77 genes were unique to ATTP+/+ and ATTP−/− mice, respectively. The expressions of O3 sensitive genes—Timp1, Areg, Birc5 and Tnc—were seen to be further modulated by AT status. The present study reveals AT modulation of adaptive response of lung genome to O3 exposure.
Ozone (O3), the principal component of photochemical smog and a powerful highly reactive hydrophobic oxidant, forms peroxides, aldehydes, and lipid ozonation products (LOP) as a result of unsaturated fatty acid oxidation in biological systems (Pryor, 1992, 1994; Pryor et al., 1995). Upper respiratory tract lining fluids (RTLF) come into direct contact with inhaled O3, absorbing and detoxifying significant portions of inhaled O3, thereby reducing amounts of the inhaled toxin reaching more vulnerable peripheral gas exchange regions of the lung (Halliwell and Cross, 1994; Langford et al., 1995). Epidemiological studies have provided evidences that O3 exposures are related to adverse health effects, including exacerbations of respiratory-tract diseases (Tager et al., 2005) and increased mortality (Thurston and Ito, 2001). Human supplementation with antioxidants, such as α-tocopherol (AT), has been reported to ameliorate decrements in O3-induced lung injury in some (Samet et al., 2001; Sienra-Monge et al., 2004), but not all (Mudway et al., 2006), studies.
α-Tocopherol (AT), a lipid-soluble dietary antioxidant, is the most abundant and biologically active form of vitamin E (VE) (Schneider, 2005). VE is absorbed into the lymphatics from the intestine and initially transported in chylomicrons to the systemic circulation. Following uptake of the circulating VE-containing chylomicrons and their remnants by the liver, AT is preferentially bound to the hepatic AT transfer protein (ATTP), incorporated into lipoproteins, and secreted into the circulation (Traber and Arai, 1999). ATTP mRNA has been detected in low levels in other tissues, including rat brain, spleen, lung, and kidney (Hosomi et al., 1998). ATTP−/− mice, therefore, have low plasma and extrahepatic tissue AT concentrations (approximately 2–20% of littermate ATTP+/+ mice), while liver concentrations remain approximately 40% of those of their ATTP+/+ littermates (Leonard et al., 2002).
Previously, we reported genome-wide changes in mRNAs from livers and brain cortex of ATTP−/− mice using high-density oligonucleotide arrays. Our findings included enhanced gene expression related to cell proliferation pathways and oxidant stress in liver and repression of specific genes that determine synaptic plasticity and neuronal development in brain cortex (Gohil et al., 2003b). In a later report our results also suggested age-related multi-organ deregulation of gene networks in ATTP−/− as compared with ATTP+/+ mice (Gohil et al., 2004). We have also reported modest increases in lung and liver heme oxygenase 1 (HO-1) and inducible nitric oxide synthase (iNOS) expression in lipopolysaccharide (LPS)-injected (ip) ATTP−/− mice as compared with ATTP+/+ mice (Schock et al., 2004). Recently we reported dysregulated expression of immune- and lipid metabolism-related genes in ATTP−/− heart tissue and dysregulated airway allergic inflammatory responses in ovalbumin-sensitized ATTP−/− mice (Lim et al., 2008; Vasu et al., 2007).
O3 exposure has been used to explore transcriptomic responses of the lungs to oxidative stress. When C57BL6/J mice were exposed to O3 (1 ppm for 3 consecutive nights, 8 h/night) followed by microarray analysis of lung mRNA, we observed induction of the expression of serum amyloid A3 gene and of DNA- and cell cycle-related genes, as well as repression of xenobiotic and cytoskeletal genes, as compared with lungs from air-breathing mice (Gohil et al., 2003a). These findings suggest that various defensive mechanisms are activated to protect lungs from O3 exposure in mice.
AT deficiency to the extent obtained by ATTP−/− mice would be uncommon in humans except for those with severe fat malabsorption or rare families exhibiting ATTP deficiency (Alex et al., 2000; Jeppesen et al., 2000). However, by manipulating systemic AT, a major lipophilic antioxidant, we would predict a robust “antioxidant action,” which would be further augmented by O3 (Traber and Atkinson, 2007). On the other hand, AT may have biological activities beyond its antioxidant action (Azzi, 2007; Brigelius-Flohe and Davies, 2007). ATTP−/− mice represent a unique strain where we can study the effect of AT on biological systems by genetically (via deletion of ATTP gene) manipulating systemic AT levels without any dietary AT regulations. The present study was carried out to analyze the genome-wide effect of O3 in AT-deficient lungs secondary to ATTP deletion in ATTP−/− mice and AT-sufficient mice (ATTP+/+) as compared to their air-breathing ATTP−/− and ATTP+/+ littermate controls. Some of the results in this article have been previously reported in abstract form (Vasu et al., 2006a, 2006b).
The protocols for the care and use of animals were approved by the Institutional Care and Use Committee at the University of California, Davis. Male C57BL/6 mice with a deletion of ATTP gene (ATTP−/−) and wild-type mice (ATTP+/+) were used from our colonies. The ATTP−/− mice in C57BL6/J genetic background were obtained by backcrossing the original colony (Terasawa et al., 2000) of mixed (50% C57BL6/J and 50% 129/SvJae) mice heterozygous for the deletion (ATTP+/-) with C57BL6/J (ATTP+/+) mice. The mice were housed in polycarbonate cages in a room maintained at 21–23°C and 60–70% humidity on a 12-h light/dark schedule with ad libitum access to water and food. ATTP−/− mice were bred by mating heterozygous (ATTP+/-) male and female mice, and the offspring were genotyped using specific primers for ATTP and confirmed by Western blot analysis using antibody against ATTP protein in liver (Terasawa et al., 2000). After weaning, the offspring were fed diets containing 35 IU dl-tocopheryl acetate/kg diet (USB Corporation, Cleveland, OH).
At 12–14 wk, ATTP+/+ or ATTP−/− mice were exposed either to filtered air or to 0.5 ppm O3 for 6 h/day for 3 consecutive days (n = 8 per group), with body weights being measured before and after exposure. The O3 was produced from O2 by electrical corona arc discharge (sander Ozonizer model IV, Eltze, Germany). The O2–O3 mixture (~95% O2, 5% O3) was mixed with ambient air and allowed to flow into a Teflon-lined exposure chamber, with the O3 concentration in chamber being adjusted to 0.5 ppm and continuously monitored by an O3 detector (Dasibi model 1003-AH, Glendale, CA) (Valacchi et al., 2004). Exposure to filtered air was done in similar exposure chambers except that filtered airflow was released into the chamber at flow rates similar to the O3 flow. Temperature and humidity were monitored during exposures (25–28°C and 45–55%, respectively).
All the animals were sacrificed immediately after the third 6-h exposure by intraperitoneal injection of beuthanasia (120 mg/kg body weight) and bronchoalveolar lavage (BAL) performed in half of the animals (n = 4 per group). Blood was collected by cardiac puncture and lung (dissected free of trachea, large airways, and blood vessels), and liver tissues were rapidly obtained and stored at −80°C until further processing.
Plasma, liver, and lung AT analyses were determined using high-pressure liquid chromatography (HPLC) with electrochemical detection (Podda et al., 1996).
Plasma nitrate/nitrite were measured as an indicator of NO production through reduction with acidified vanadium(III) using the Sievers NO analyzer (Sievers, Boulder, CO) (Van Der Vliet et al., 2000).
Animals were sacrificed, tracheal cannula inserted, and bronchoalveolar lavage (BAL) performed three times via the trachea using approximately 1.0 ml of ice-cold phosphate-buffered saline (PBS, pH 7.4) and recovered by gentle manual aspiration. The recovered BAL fluid was centrifuged, the cell pellet washed twice, and finally resuspended in 1 ml PBS. The total cell count was performed in a Burker chamber (Novolab, Geraardsbergen, Belgium).
To assess alveolar-capillary membrane protein permeability, total protein from cell-free BAL fluid was measured using a commercially available protein bioassay kit (Bio-Rad Laboratories, Hercules, CA).
Total RNA from ~30 mg unlavaged lung tissue was extracted by RNeasy Minikit (Qiagen, Valencia, CA) according to the manufacturer's protocol. Total RNA purity and quantity were determined spectrophotometrically (A260/A280 ratio being close to 2.0 for pure RNA). Four replicates, each from a mouse lung, from each group (n = 4 per group), were further processed as follows for gene chip analysis.
The methods used to prepare the eukaryotic target preparation were followed according to the manufacturer's protocol (Affymetrix, Inc.) for expression arrays. An aliquot (10 μg) of RNA solution was used for preparation of one-cycle cDNA synthesis (first-strand and second-strand cDNA synthesis) followed by cleanup of double-stranded cDNA and synthesis of biotin-labeled cRNA. The biotin-labeled cRNA (40 μg) was used for fragmenting for target preparation. Fragmented cRNA samples were hybridized to high density oligonucleotide Affymetrix Mouse 430A 2.0 GeneChips (Santa Clara, CA) which contains 22,600 probe sets representing transcripts and variants from over 14,000 well-characterized mouse genes. For hybridization, gene expression was assessed using one chip per mouse lung.
The scanned images of hybridization signals were analyzed with the Affymetrix GeneChip Operating Software (GCOS) including the GeneChip Scanner 3000 High-Resolution Scanning Patch and DNA-D chip analyzer (dChip), a software package implementing model-based expression analysis of oligonucleotide arrays at http://www.dchip.org (Li and Hung Wong, 2001). The absolute mRNA expression (present or absent) and differential mRNA expression data were obtained from GCOS. When the p value for detection signal was < .049 (range of p value .0002–.049), the expression of the mRNA was classified as present (P). All mRNAs with the p value for detection > .05 were considered absent (A). The signal intensities for transcripts classified as present ranged from 5 to 7000 U.
The entire data of ATTP+/+, ATTP−/−, air/O3-exposed groups of mice were subjected to analysis. CEL files generated in GCOS were analyzed using dChip (version 1.0) by model-based expression (perfect match/mismatch; PM/MM) analysis to compute expression values. Hybridization data from O3-exposed lungs were treated as “Test” data and those for control, air-exposed lungs as “Baseline” data. The generated gene list was filtered representing the following criteria: (i) data that showed either an “increase” or a “decrease” of ≥ 2-fold expression comparing air-exposed ATTP−/− group of mice versus ATTP+/+ groups; (ii) O3-exposed ATTP+/+/ATTP−/− groups of mice versus air-exposed ATTP+/+/ATTP−/− groups greater than 100 expression (E) value difference in either direction (i.e., Etest − Econtrol > 100 or Econtrol − Etest > 100); (iii) p for testing was set at < .05, where p is based on the t distribution used to assess the relative magnitude of the expression value difference. The D-Chip software program calculates mean and standard error of the mean (SEM) for each mRNA in a baseline (e.g., air) sample and compares the mean and SEM for the same mRNA in test sample (e.g., O3). The program then gives a p value for the difference and magnitude of the difference. The data are then edited to select only the genes that show ≥ 2-fold and p value < .05. Our previous experience (Gohil et al., 2003a, 2003b, 2007; Oommen et al., 2007; Vasu et al., 2007, 2009) has shown that we can confirm “all” the changes that are ≥ 2-fold, by independent quantitative real-time polymerase chain reaction (PCR) assay using a sample size of 4–6 mice per group. However, the entire data from this study will be submitted online at Array Express (http://www.ebi.ac.uk/arrayexpress). Hierarchial clustering or “heat map” of the resulting differentially regulated gene list was analyzed by dChip software.
Total RNA was extracted from lung as described for GeneChip analysis and cDNA was synthesized from 5 μg total RNA and reverse transcribed to obtain cDNA. Gene-specific primers (Table 1) for selected genes were designed using Primer Express 2.0 software (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primer was used to normalize for gene expression level. qRT-PCR was carried out with SYBR Green I Master Mix (Applied Biosystems) reagent. The PCR results were analyzed using the ABI Prism 7700 sequence detector (PE Applied Biosystems, Foster City, CA). The 2−ΔΔCT method (Livak and Schmittgen, 2001) was used to calculate relative changes in gene expression determined from qRT-PCR experiments (Applied Biosystems User Bulletin No. 2 [P/N4303859]). The threshold cycle, Ct, which correlates inversely with the target mRNA levels, was measured as the cycle number at which the SYBR Green emission increases above a preset threshold level. The specific mRNA transcripts were expressed as fold difference in the expression of the specific mRNAs in RNA samples from the lungs of ATTP+/+/ATTP−/− mice exposed to O3 compared to those in the respective lungs of mice exposed to filtered air.
Statistical evaluation of analytical data was done by Student's t test using the statistical software GraphPad Prism 4.0 (San Diego, CA). In all comparisons p < .05 was considered as significant. Results are expressed as mean ± SEM. Tissue AT concentrations were evaluated by ANOVA after log transformation of data, and Tukey's multiple comparison test was employed for pairwise comparisons.
The plasma, liver, or lung AT levels in ATTP−/− mice were 10-to 20-fold lower than those of ATTP+/+ mice (Figures 1–3). Though O3 exposure (0.5 ppm, 6 h/day for 3 days) caused a significant decrease (p < .01) in liver AT concentrations of ATTP−/− mice as compared to air-exposed ATTP−/− mice, no significant changes were observed in liver AT levels of ATTP+/+ mice (Figure 2). Plasma and lung AT levels were unaffected by O3 exposure in both genotypes (Figures 1 and and3,3, respectively).
No significant changes in plasma nitrate/nitrite levels were observed in ATTP+/+/ATTP−/− mice exposed to O3 as compared to their respective air-breathing controls (data not shown).
Significantly (p < .001) increased levels of total BAL cell counts and cell-free BAL protein levels were observed in response to O3 exposure in both groups of mice (Figures 4 and and5).5). Notably, the total BAL cell count doubled in response to O3 exposure in ATTP−/− (p < .001) as compared with ATTP+/+ mice (Figure 4).
The global gene expression profile of unlavaged lung tissue of ATTP+/+/ATTP−/− mice exposed to O3 (0.5 ppm, 6 h/day for 3 days) were analyzed and compared to their air-breathing littermate ATTP+/+/ATTP−/− controls. The microarray analysis by mouse 430 A2.0 arrays or gene chips (4 replicates per group) detected ~15,000 probe sets (~9000 genes) in both O3-exposed ATTP+/+ and ATTP−/− lung tissues. O3 exposure resulted in the selective modulation of 52 and 99 genes in the lungs of ATTP+/+ and ATTP−/− mice, respectively. The expression of 22 genes was common to both (Table 2), of which the expressions of 4 genes (Timp1, Areg, Birc5, Tnc) were seen to be further modulated by lung AT status. The expression of 30 and 77 genes were unique to ATTP+/+ and ATTP−/− mice, respectively (Figure 6). Figure 7 shows the “heat map” for the entire set of differentially regulated genes. The coefficient of variation for Chip–Chip, i.e., same sample on four different chips, was <2%. As reported earlier by us (Oommen et al., 2007), mouse-to-mouse variation depends on the activity of the gene and can vary from 5 to 41%. Our selection criterion for a significant change was empirically set at ≥ 100%.
Comparing lung mRNA expression in air-breathing mice, only two genes, Prrx1 and Hspb1, were downregulated, ~4- and 3-fold, respectively, in ATTP−/− compared with ATTP+/+ mice (Table 3). It must be noted that as compared to our earlier published reports (Oommen et al., 2007; Vasu et al., 2007), in this study we have used more stringent criteria for selection of genes that are sensitive to AT deficiency and/or O3. The primary aim was to obtain the most sensitive genes that have high statistical probability to be “real change” and possibly have biological significance. However, the entire data set from this study will be published online at Array Express (http://www.ebi.ac.uk/arrayexpress).
Table 4A shows functional groups of selected genes that were upregulated in ATTP+/+ mice in response to O3 exposure. Most of these genes were related to inflammatory-immune responses and cell proliferation or DNA repair. Some of these included FK506 binding protein 5 (Fkbp5), metallothionein 2 (Mt2), chemokine (C-X-C motif) ligand 2 (Cxcl2), prostaglandin-endoperoxide synthase 1 (Ptgs1), cell division cycle 2 homolog A (Cdc2a), and cell division cycle 6 homolog (Cdc6). The expression of transcription factors such as period homolog 1 (Per1), nuclear factor interleukin 3 regulated (Nfil3), and activating transcription factor 3 (Atf3) was also seen to be upregulated. Interestingly, only seven genes were found to be downregulated in ATTP+/+ mice in response to O3 exposure as compared to their air-breathing controls (Table 4B). These included immunoglobulin heavy chain (Igh-VJ558), N-acylsphingosine amidohydrolase 3-like (Asah3l), hepatitis A virus cellular receptor 1 (Havcr1), casein kappa (Csnk), membrane-spanning 4-domains, subfamily A, member 4C (Ms4a4c), WNT1 inducible signaling pathway protein 2 (Wisp2), and protocadherin 18 (Pcdh18), of which Igh-VJ558 showed a 36-fold downregulation. The entire list of differentially expressed genes is shown in Appendix I.
The lung gene expression of ATTP−/− mice when exposed to O3 showed 99 genes differentially expressed as compared to their air-breathing ATTP−/− controls. This was approximately 50% more than that observed in differential comparison of ATTP+/+ lungs exposed to O3 and respective air-breathing controls (Figure 6). The functional classification of these upregulated genes resulted in six groups: cell proliferation- or DNA repair-related genes (40%), inflammatory-immune related genes (27%), transcription-related genes (7%), cytoskeleton-related genes (3%), lipid metabolism-related genes (2%), and ion transport-related genes (1%) (Table 5A). Seven genes were found to be downregulated in lungs of ATTP−/− mice exposed to O3 as compared to their air-breathing controls (Table 5B). These included membrane-spanning 4-domains, subfamily A, member 4c (Ms4a4c) (−3.5-fold), complement receptor 2 (Cr2) (−3.3–fold), interferon-induced transmembrane protein 6 (Ifitm6) (−3.3-fold), CD19 antigen (Cd19) (−3.1-fold), ATPase, Na+/K+-transporting, alpha 3 polypeptide (Atp1a3) (−3.0-fold), nuclear receptor subfamily 1, group D, member 1 (Nr1d1) (−3.0-fold), and insulin-like growth factor binding protein 3 (Igfbp3) (−2.9-fold). The entire list of differentially expressed genes is shown in Appendix II.
We selected 14 genes from the microarray data to confirm independently by qRT-PCR. The selected genes included upregulated (Gclc, Saa3, Marco, Timp1, Areg, Check1, Cxcl2, Fkbp5, Tnc, Ccna2, Mbd1) and downregulated (Nr1d1, Hspb1, Igfbp3) genes. The fold-change data analyzed by qRT-PCR were similar to or lower than the microarray data and was highly correlated (R2 = .8076) (Figure 8). The higher degree of consistency between the qRT-PCR data and oligonucleotide microarray data raised confidence in the gene expression data obtained and the validity of the experiments.
ATTP−/− mice represent a unique strain in which one can study the effect of AT on biological systems by genetically (via deletion of ATTP gene) manipulating systemic AT levels. The experimental design utilized ATTP−/− mice in which extremely low levels of lung AT were achieved secondary to ATTP deletion. The present study was carried out to explore the potential of the major nutritional lipophilic antioxidant, AT, on the adaptive response of lung transcriptomes to O3, the most important photochemical oxidant. Genome-wide effect of O3 were analyzed in ATTP−/− mice and AT-sufficient mice (ATTP+/+) as compared to their respective air-exposed ATTP−/− and ATTP+/+ littermate controls.
The AT levels in plasma, liver, and lung tissues of ATTP−/− mice were much lower as compared to their respective levels in ATTP+/+ littermate controls, as reported earlier by us and by other workers (Gohil et al., 2003b, 2004; Leonard et al., 2002; Lim et al., 2008; Schock et al., 2004; Terasawa et al., 2000). We observed an increase in total BAL cell count and protein levels in O3-exposed mice, confirming O3-induced inductions of lung injury (Guth et al., 1986), but we did not find any significant change in plasma or lung AT levels in O3-exposed ATTP+/+ or ATTP−/− mice (0.5 ppm, 6 h/day for 3 days) as compared to their respective air-exposed littermate controls. However, a significant O3-induced drop was observed in liver AT levels in ATTP−/− mice as compared to air-exposed ATTP−/− mice (Figure 2). A possible explanation would be an increase in circulatory lipid hydroperoxides as a result of AT deficiency in O3-exposed ATTP−/− mice as observed by other workers in O3-exposed AT deficient rats (Sato et al., 1980). In a study published in 1980, rats fed vitamin E-deficient diet or vitamin E-supplemented diet were exposed to 0.3 ppm O3 for 3 h daily, 5 days/wk, for 7 mo (Sato et al., 1980). The investigators observed that serum vitamin E concentration decreased following O3 exposure in the E-supplemented group, but remained unaffected in the E-depleted group. Interestingly, they observed that liver tissues showed more damaging affect as compared to lung tissues that were directly exposed to O3, as reflected by biochemical and morphological observations, although no clear explanation could be put forth (Sato et al., 1980). In another study, in a burn and inhalation injury model in sheep, Traber and workers observed nearly 2000 nmol alpha-tocopherol depletion from the liver, as compared to only 1000 nmol alpha-tocopherol loss in lungs, over 48 h after exposure to burn and inhalation injury (Traber et al., 2007). The loss of hepatic alpha-tocopherol was suggested to be caused by lipid hydroperoxides, which are delivered to the liver by lipoproteins (Traber et al., 2007).
We had earlier reported a decrease in plasma and lung AT levels (23 and 20%, respectively) in O3-exposed SKH-1 hairless mice (0.8 ppm, 6 h/day for 6 days) as compared to their air-breathing controls (Valacchi et al., 2004). Elsayed and workers have previously reported an increase in lung AT content after O3 exposure (0.5 ppm, 5 days) in Long-Evans rats fed an AT-supplemented diet, suggesting mobilization of AT to lungs as a response to “oxidative stress” or as a result of O3-induced inflammatory responses with leakage of blood constituents into lung tissues (Elsayed et al., 1990). In another study, no significant changes were found in plasma AT levels in mature female rabbits exposed to 0.1 to 0.6 ppm of O3 for 3 h (Canada et al., 1987). These observations suggest the degree of oxidative stress caused by O3 exposure, potentially dependent upon the duration of exposure, the levels of antioxidant (both enzymatic and nonenzymatic), the magnitude of the inflammatory responses, and/or the species/strain of animals may all be involved in modulating lung AT levels.
Comparing the differential expression of lung genes in air-breathing mice, only two genes, paired related homeobox 1 (Prrx1) and heat-shock protein 1 (Hspb1), were downregulated, by 4.2- and 3.2-fold, in ATTP−/− lungs (Table 3). Prrx1 (also known as Prx1 or Mhox) is a transcription factor associated with differentiation and vascularization of vessel walls (Bergwerff et al., 1998). In lungs, Prrx1 has been localized to the differentiating endothelial cells and Prrx1-/- mice have been reported to show abnormal lung vascularization (Ihida-Stansbury et al., 2004). Hspb1 (also known as Hsp27 or Hsp25) belongs to the family of small heat-shock proteins that are stimulated by heat shock or disturbed cell environment or inflammation. Hspb1 was also suggested to be anti-apoptotic effector acting by sequestering cytochrome c released from mitochondria (Bruey et al., 2000). Sinha et al. reported that AT deficiency leads to reduced Hspb1 expression in lung type II pneumocytes and thus promotes apoptosis (Sinha et al., 2002). Reduced Hspb1 expression was assumed to be dependent upon the protein kinase C (PKC) pathway, as AT is known to modulate PKC activity (Mahoney and Azzi, 1988). In type II pneumocytes its deficiency has been reported to strongly increase PKC activity (Sabat et al., 2001). The present study further supports a role for AT in Hspb1 expression.
Of note, we were surprised to observe that only two genes were differentially expressed in ATTP−/− lungs as compared to lungs of respective littermate ATTP+/+ controls. We were expecting a “significant number” of differentially expressed genes, particularly those related to “oxidant responsive” genes. Interestingly, we previously have not observed any significant increase in isoprostanes in lungs from ATPP−/− mice (unpublished data, from the laboratory of the late Dr. J. Morrow) as compared to ATTP+/+ controls. Similar unexpected observations were recently reported by other workers in brain tissue of ATTP−/− mice (Cuddihy et al., 2008).
Of the expression of 22 genes in lungs common to both ATTP+/+ and ATTP−/− mice exposed to O3, the expression of 4 genes [tissue inhibitor of metalloproteinase 1 (Timp1), amphiregulin (Areg), baculoviral IAP-repeat containing 5 (Birc5), and Tenascin C (Tnc)] was seen to be further modulated by lung AT status (Table 2). Timp1, Areg, Birc5, and Tnc were induced 4.1-, 5.9-, 4.7-, and 4.4-fold in ATTP+/+ mice, whereas they were induced by 13.8-, 8.2-, 8-, and 8-fold, respectively, in ATTP−/− mice. TIMPs are regulators of matrix metalloproteinase activity and their levels are known to be increased in lung injury (Kim et al., 2005). They have also been touted as a physiological regulator of normal airway repair (Chen et al., 2008). Wang et al. have reported a consistent increase in Timp1 mRNA expression in rat type I-like lung pneumocytes exposed to 100 ppb (0.1 ppm) O3 for 1 h, thus suggesting the importance of this lung cell type in alveolar injury repair (Wang et al., 2006). Similarly, the expression of amphiregulin (Areg) was induced 8.2-fold in O3-exposed ATTP−/− mice, as compared to 5.9-fold in O3-exposed ATTP+/+ mice. Areg, a member of epidermal growth factor, is known to regulate cell proliferation (Schuger et al., 1996). Shao and workers have reported that Areg can be activated by prostaglandin E(2) (PGE2), a major product of cyclooxygenase (cox) via cAMP/protein kinase A pathways in colon cancer cells (Shao et al., 2003).
Of note, in the present study, the mRNA expression of prostaglandin-endoperoxide synthase 1 and 2 (Ptgs1 and 2/Cox-1 and 2) was seen to be induced by 2.5- and 3.8-fold, respectively, in O3-exposed ATTP−/− mice. Interestingly, only Ptgs1 was seen to be induced (2.4-fold) in O3-exposed ATTP+/+ mice. AT supplementation has been reported to reverse age-associated increase in murine macrophage PGE2 production and COX activity, but had no effect on COX 1 and 2 protein expression or COX 2 mRNA expression (Wu et al., 1998). The authors suggested that AT may be exerting its effect posttranslationally by inhibiting COX activity (Wu et al., 1998). Hence our microarray data suggest that O3 exposure to lung tissues with low AT levels (ATTP−/− mice) may increase Ptgs1 and -2 expressions, leading to increased expression of Areg, which in turn may modulate cell proliferation mechanisms. Similarly, the increased expression of tenascin C (Tnc), an extracellular matrix component, in lung tissues of O3-exposed ATTP−/− mice (8-fold) as compared to lung tissues of O3-exposed ATTP+/+ mice (4.4-fold) further consolidates our hypothesis of modulated lung injury pathways in O3-exposed ATTP−/− mice as a result of low AT status. Tnc is associated with cell detachment, migration, proliferation, and tissue remodeling (Chiquet-Ehrismann, 1991). Potter-Perigo and workers have shown an increased expression of Tnc in O3-exposed (0.3 ppm, 3 h) primate nasal epithelial cells (harvested after 24 h postexposure) as a result of epithelial cell detachment and disruption (Potter-Perigo et al., 1998).
The increased expression of Birc5 (also known as survivin), a member of the inhibitor of apoptosis protein family (IAP) (Salvesen and Duckett, 2002), in lungs of O3-exposed ATTP-/- mice (8-fold) represents another marker of inflammation (Altznauer et al., 2004). Overexpression of Birc5 had also been reported in proliferating cells and its expression had been described to be cell-cycle dependent (Altznauer et al., 2004). Thus the present data reveal O3-sensitive genes that may be further modulated by AT levels in lung tissues. It must also be noted that the absence of any dose response or time course of O3 exposure in the present study (one of the limitation of the study) renders it difficult to discern direct O3-specific changes from those of O3 plus the resulting inflammatory-immune system activation-induced changes. Similarly, another limitation of the current study includes the difficulty in ascribing noted changes to changes in a specific lung cell population. However, it should also be recognized that the GeneChip assay used in this study is very sensitive with a large dynamic range for detections (signal intensity of 10–10,000 units) (Oommen et al., 2007). Hence by pairwise comparison of two treatment groups, some of the large changes in even a small population of lung cells should be detected.
Comparative analysis of lung gene expression in ATTP−/− mice exposed to O3 as compared to their respective air-breathing ATTP−/− control mice showed similar trends in upregulation of inflammatory-immune related and cell differentiation/DNA repair/DNA synthesis related genes as observed in lungs of ATTP+/+ mice exposed to O3. The changes in gene expression in lungs of ATTP−/− mice in response to O3 were significantly higher, as seen by the differential expressions of greater than 50%, likely due to AT deficiency as a result of ATTP gene deletion (Figure 6).
We have earlier reported an induction of lung mRNA expression of serum amyloid A3 (Saa3), a main acute-phase protein in mice exposed to 1 ppm O3 for 3 consecutive nights (8 h/night) (Gohil et al., 2003a). In the current study we did not see any change in Saa3 expression in ATTP+/+ mice exposed to O3. However, there was an 18.7-fold increase in Saa3 expression in ATTP−/− mice exposed to O3 (Table 5A). Saa3 has been shown to induce extracellular matrix-degrading enzymes such as collagenase and matrix metalloproteinases (Mitchell et al., 1991; Vallon et al., 2001). The expression of macrophage receptor with collagenous structure (Marco), class A scavenger receptor, was upregulated by 6.1-fold in lungs of O3-exposed ATTP−/− mice. Its expression was reported to be restricted to subsets of macrophages active in scavenging pathogenic microorganisms (Kraal et al., 2000). Dahl and workers recently reported increased mRNA expression of Marco in lungs in response to O3 exposure in wild-type mice and at the same time reported increased lung inflammation and injury in MARCO−/− mice instilled (intratracheally) by pro-inflammatory O3-generated lipids: 5β,6β-epoxycholesterol (β-epoxide) and 1-palmitoyl-2-(9′-oxo-nonanoyl)-glycerophosphocholine (PON-GPC) (Dahl et al., 2007). The unavailability of time-course observation in the present study makes it difficult to explain the modest increase (6.1-fold in lungs of O3-exposed ATTP−/− mice) in MARCO mRNA expression as compared to reports by Dahl and workers (Dahl et al., 2007) in a different cohort of mice.
Several genes related to cell proliferation markers and regulators, such as cyclin A2 (Ccna2), cyclin B1 (Ccnb1), Cyclin B2 (Ccnb2), Chek1, Tnfrsf12a, Gadd45g, and Top2a, were also upregulated in O3-exposed ATTP−/− mice. Alpha-fetoprotein (Afp), reported to promote cell proliferation of NIH 3T3 cells by induction of the PKA-cAMP pathway (Li et al., 2002), was also upregulated.
As compared to respective air-breathing controls, in lungs of ATTP+/+ mice exposed to O3, chemokine (C-X-C motif) ligand 2 (Cxcl2), also known as macrophage inflammatory protein 2 (Mip-2), was induced by 5-fold. Cxcl2 has been reported to be secreted from alveolar macrophages, neutrophils, type II epithelial cells, and lung fibroblasts (Driscoll et al., 1993a) and is believed to play an important role in O3-induced inflammatory responses (Driscoll et al., 1993b). Tumor necrosis factor receptor superfamily, member 12a (Tnfrsf12a) (also known as TWEAK-R or Fn14), predominantly found in macrophages and a well-known modulator of cell proliferation, differentiation, and apoptosis (Girgenrath et al., 2006), was induced by 2.8-fold. Similarly, cell cycle-related genes such as cell division cycle 2 homolog (Cdc2), cyclin A2 (Ccna2), survivin (baculoviral IAP repeat-containing 5, Birc5), mitotic arrest deficient, homolog (MAD2), cell cycle division 20 homolog (Cdc20), cell division cycle 6 homolog (Cdc6), topoisomerase DNA II alpha (Top2A), and other related genes were also induced.
Transcription factors such as period homolog 1 (Per1), nuclear factor, interleukin 3 regulated (Nfil3), and activating transcription factor 3 (Atf3) were upregulated by 3.2-, 3.2-, and 2.9-fold, respectively, in lungs of O3-exposed ATTP+/+ mice (Table 4A). Per1, implicated in regulating circadian rhythm, has been reported to be elevated in the paraventricular nucleus of mice as a result of stress-induced disturbance of circadian corticosterone secretion (Takahashi et al., 2001). The expression of the “clock genes” including Per1 has been reported to be present in larynx, trachea, bronchus, and lung of mice (Bando et al., 2007) and has also been reported to regulate cell cycle by interacting with components of the cell cycle checkpoint system activating checkpoint kinase 1 and 2 (Chek1 & 2) in response to DNA damage (Gery et al., 2006). Interestingly, Chek1 and another key regulator of cell cycle progression, growth arrest and DNA-damage-inducible 45 gamma (Gadd45g), were seen to be induced by 4.3- and 3.0-fold in lungs of O3-exposed ATTP+/+ mice. Kondratov and Antoch have reported that Gadd45g showed circadian patterns of recognition at mRNA or protein levels (Kondratov and Antoch, 2007). Nfil3 (also known as E4BP4), a mammalian basic leucine zipper (bZIP) transcription factor found primarily in T cells (Cowell, 2002) and induced by interleukin (IL)-3 (Ikushima et al., 1997), has been reported to regulate inflammatory processes and apoptosis and also involves circadian clock mechanisms (Cowell, 2002). Atf3, a member of the ATF/CREB family of transcription factors, is a stress-inducible gene induced by a variety of stress signals and also induced by ischemia and hypoxia (Ameri et al., 2007; Chen et al., 1996; Hai et al., 1999). It is also reported to affect cell cycle progression and apoptosis (Lu et al., 2006).
Nr1d1 (Nuclear receptor subfamily 1, group D member 1), also known as Rev-erb alpha, is seen to be downregulated (3-fold) in lungs of O3-exposed ATTP−/− mice (Table 5B). It is a member of the “orphan” nuclear receptor family (Chawla and Lazar, 1993). Nr1d1 has also been reported to display circadian rhythmic expression by regulating or being regulated by other “clock genes” of circadian circuitry. It is also known to regulate various metabolic pathways (Duez and Staels, 2008; Torra et al., 2000). Though it is not very clear how O3 may directly/indirectly influence the expression of genes involved in circadian pathways, studies indicate that oxidative stress might play an important role in their transcriptional regulation (Igarashi et al., 2007).
One of the limitations of the study was nonavailability of data for O3-modulated lung gene expression profiles and/or lung injury at different time intervals (van Bree et al., 2001). It must be noted that our transcriptomic data were in lungs exposed to O3 for 3 days, a period in which the lung epithelial cell proliferation and alveolar wound repair mechanisms and inflammatory responses are simultaneously occurring (Dormans et al., 1999; Savov et al., 2004). The present study did not characterize lipid ozonation products that may have directly or indirectly modulated gene expression as a result of increased lipid peroxidation due to AT deficiency (Kafoury et al., 2007). Other workers have provided evidence that O3-induced inflammatory responses may be maximally present on the third day of acute O3 exposure (Dormans et al., 1999; Savov et al., 2004). Bioinformatic analyses of several genes modulated in the present study reveals involvement of nuclear factor (NF)-κB and related pathways (Bing et al., 2000; Cavin et al., 2004; Girgenrath et al., 2006; Gohil et al., 2003a; Huffman et al., 2002; Morante et al., 2005). The induction of the catalytic subunit of glutamate-cysteine ligase, Gclc, a rate-limiting enzyme in de novo reduced glutathione (GSH) synthesis, by 2.9-fold in O3-exposed ATTP−/− mice (Table 5A) also suggests the involvement of a transcription factor of the cap'n'collar-basic leucine zipper proteins (CNC-bZIP), Nrf2 (Wild et al., 1999). Thus, the present data also suggest possible highly complex cross-talk mechanisms involved between the “stress-responsive” transcription factors such as Nrf2, NF-κB, and Atf3. The present study also identifies four O3-sensitive lung genes (Timp1, Areg, Birc5, and Tnc) that appear to be further modulated by AT status. However, it must also be noted that in the present study we exposed the mice to O3 (0.5 ppm) 6 h/day for 3 days and thus the gene expression pattern observed represents changes due to O3, O3-induced injury-inflammation-repair processes, and/or adaptive responses to O3 exposure, thus significantly limiting mechanistic interpretations regarding respiratory tissue overall pathobiological responses. However, it can be reasonably concluded that lung genomic responses to O3 are modulated in ATTP-/- mice compared to their littermate ATTP+/+ mice.
This research was partly funded by grants from National Institute of Health Sciences (ES011895) and United States Department of Agriculture (35200-13456).
|Description||Affymetrix probe set ID||Fold change||p Value|
|FK506 binding protein 5, Fkbp5||1448231_at||10.9||.046047|
|Metallothionein 2, Mt2||1428942_at||5.7||.039341|
|Cell division cycle 20 homolog (S. cerevisiae), Cdc20||1439394_x_at||5.7||.01147|
|Glucosaminyl (N-acetyl) transferase 2, I-branching enzyme, Gcnt2||1451733_at||5.4||.006715|
|Chemokine (C-X-C motif) ligand 2, Cxcl2||1449984_at||5.3||.018505|
|Cell division cycle associated 1, Cdca1||1430811_a_at||5.1||.016262|
|C-type lectin domain family 4, member n, Clec4n||1425951_a_at||5.0||.003182|
|Similar to histone 2a, MGC73635||1438009_at||4.9||.002595|
|Mitogen-activated protein kinase kinase kinase 6, Map3k6||1449901_a_at||4.8||.004217|
|Ribonucleotide reductase M2, Rrm2||1416120_at||4.8||.010085|
|Baculoviral IAP repeat-containing 5, Birc5||1424278_a_at||4.7||.011477|
|A disintegrin-like and metallopeptidase (reprolysin type) with thrombospondin type 1 motif, 4, Adamts4||1452595_at||4.5||.010648|
|Tenascin C, Tnc||1416342_at||4.4||.005287|
|Cell division cycle 2 homolog A (S. pombe), Cdc2a||1448314_at||4.4||.001511|
|Checkpoint kinase 1 homolog (S. pombe), Chek1||1450677_at||4.3||.028655|
|Tissue inhibitor of metalloproteinase 1, Timp1||1460227_at||4.1||.003086|
|Similar to thrombospondin 1, LOC640441||1450377_at||4.0||.003209|
|Lipin 2, Lpin2||1460290_at||4.0||.010188|
|CEA-related cell adhesion molecule 1, Ceacam1||1452532_x_at||4.0||.026812|
|Cell division cycle 6 homolog (S. cerevisiae), Cdc6||1417019_a_at||3.8||.019495|
|Serine (or cysteine) peptidase inhibitor, clade E, member 1, Serpine1||1419149_at||3.5||.009815|
|Ubiquitin-conjugating enzyme E2C, Ube2c||1452954_at||3.5||.003733|
|Natriuretic peptide receptor 3, Npr3||1448024_at||3.4||.006901|
|Iduronate 2-sulfatase, Ids||1450166_at||3.4||.008805|
|Thrombospondin 1, Thbs1||1460302_at||3.3||.013611|
|Polo-like kinase 4 (Drosophila), Plk4||1426580_at||3.3||.00878|
|Rac GTPase-activating protein 1, Racgap1||1451358_a_at||3.2||.010008|
|Period homolog 1 (Drosophila), Per1||1449851_at||3.2||.002077|
|Nuclear factor, interleukin 3, regulated, Nfil3||1418932_at||3.2||.002181|
|Coagulation factor III, F3||1417408_at||3.0||.006349|
|Growth arrest and DNA-damage-inducible 45 gamma, Gadd45g||1453851_a_at||3.0||.022487|
|Activating transcription factor 3, Atf3||1449363_at||2.9||.009167|
|Transmembrane protein 37, Tmem37||1417611_at||2.9||.003164|
|Tumor necrosis factor receptor superfamily, member 12a, Tnfrsf12a||1418571_at||2.8||.006708|
|Ribonucleotide reductase M1, Rrm1||1448127_at||2.7||.004612|
|Basic leucine zipper and W2 domains 1, Bzw1||1450846_at||2.7||.011171|
|Prostaglandin-endoperoxide synthase 1, Ptgs1||1423414_at||2.4||.002984|
|Immunoglobulin heavy chain (J558 family), Igh-VJ558||1427839_at||−36.2||.030697|
|N-Acylsphingosine amidohydrolase 3-like, Asah3l||1421496_at||−6.2||.001439|
|Hepatitis A virus cellular receptor 1, Havcr1||1450364_a_at||−5.8||.022315|
|Casein kappa, Csnk||1419735_at||−4.7||.015088|
|Membrane-spanning 4-domains, subfamily A, member 4C, Ms4a4c||1450291_s_at||−4.4||.017093|
|WNT1 inducible signaling pathway protein 2, Wisp2||1419015_at||−3.3||.013676|
|Protocadherin 18, Pcdh18||1422889_at||−2.9||.00596|
|Description||Affymetrix Probe set ID||Fold change||p Value|
|Serum amyloid A 3, Saa3||1450826_a_at||18.7||.008959|
|Tissue inhibitor of metalloproteinase 1, Timp1||1460227_at||13.8||.001442|
|Bromodomain and WD repeat domain containing 3, Brwd3||1441182_at||10.1||.004669|
|Resistin like alpha, Retnla||1449015_at||8.4||.000305|
|Tenascin C, Tnc||1416342_at||8.0||.000555|
|Baculoviral IAP repeat-containing 5, Birc5||1424278_a_at||8.0||.001819|
|Cyclin A2, Ccna2||1417910_at||7.2||.000661|
|Kinesin family member 22, Kif22||1437716_x_at||6.3||.005116|
|Metallothionein 2, Mt2||1428942_at||6.2||.005926|
|Macrophage receptor with collagenous structure, Marco||1449498_at||6.1||.006801|
|Similar to histone 2a, MGC73635||1438009_at||6.1||.02903|
|Sperm associated antigen 5, Spag5||1433893_s_at||6.0||.016256|
|Lipocalin 2, Lcn2||1427747_a_at||5.9||.01288|
|Cell division cycle 2 homolog A (S. pombe), Cdc2a||1448314_at||5.9||.000422|
|Cholesterol 25-hydroxylase, Ch25h||1449227_at||5.7||.034847|
|Leucine-rich alpha-2-glycoprotein 1, Lrg1||1417290_at||5.6||.016171|
|Solute carrier family 26, member 4, Slc26a4||1419725_at||5.5||.005928|
|Shc SH2-domain binding protein 1, Shcbp1||1416299_at||5.3||.014011|
|Cell division cycle associated 3, Cdca3||1452040_a_at||5.3||.002641|
|Cyclin B1, Ccnb1||1419943_s_at||5.2||.00567|
|Ribonucleotide reductase M2, Rrm2||1416120_at||5.1||.019074|
|Antigen identified by monoclonal antibody Ki 67, Mki67||1426817_at||5.1||.000155|
|Centrosomal protein 55, Cep55||1452242_at||5.0||.000317|
|Cyclin B2, Ccnb2||1450920_at||4.8||.000411|
|Ubiquitin-conjugating enzyme E2C, Ube2c||1452954_at||4.8||.025162|
|A disintegrin-like and metallopeptidase (reprolysin type) with thrombospondin type 1 motif, 4, Adamts4||1452595_at||4.8||.008109|
|Thymidine kinase 1, Tk1||1416258_at||4.6||.00906|
|RAD51 homolog (S. cerevisiae), Rad51||1418281_at||4.6||.000758|
|Spindle pole body component 25 homolog (S. cerevisiae), Spbc25||1424118_a_at||4.6||.000531|
|Helicase, lymphoid specific, Hells||1417541_at||4.5||.010822|
|Topoisomerase (DNA) II alpha, Top2a||1454694_a_at||4.4||.000008|
|C-type lectin domain family 4, member n, Clec4n||1425951_a_at||4.0||.03387|
|Cell division cycle associated 1, Cdca1||1430811_a_at||4.0||.000807|
|Inter alpha-trypsin inhibitor, heavy chain 4, Itih4||1431808_a_at||3.9||.018824|
|Prostaglandin-endoperoxide synthase 2, Ptgs2||1417262_at||3.8||.004319|
|Suppressor of cytokine signaling 3, Socs3||1455899_x_at||3.8||.030528|
|Cyclin B1, related sequence 1, Ccnb1-rs1||1448205_at||3.7||.006735|
|Cell division cycle 20 homolog (S. cerevisiae), Cdc20||1416664_at||3.7||.007289|
|Ect2 oncogene, Ect2||1419513_a_at||3.7||.001832|
|Membrane targeting (tandem) C2 domain containing 1, Mtac2d1||1439045_x_at||3.7||.001114|
|Leucine zipper protein 5, Luzp5||1417926_at||3.6||.000143|
|Transforming, acidic coiled-coil containing protein 3, Tacc3||1455834_x_at||3.6||.00117|
|ASF1 anti-silencing function 1 homolog B (S. cerevisiae), Asf1b||1423714_at||3.6||.003282|
|CDC28 protein kinase 1b, Cks1b||1416698_a_at||3.6||.00004|
|Tubulin, alpha 6, Tuba6||1416128_at||3.6||.006892|
|Tumor necrosis factor (ligand) superfamily, member 9, Tnfsf9||1422924_at||3.5||.003302|
|Fidgetin-like 1, Fignl1||1422430_at||3.4||.000638|
|Polymerase (DNA directed), epsilon, Pole||1448650_a_at||3.4||.005435|
|Methyl-CpG binding domain protein 1, Mbd1||1430838_x_at||3.4||.009822|
|Rac GTPase-activating protein 1, Racgap1||1451358_a_at||3.4||.002712|
|Ubiquitin-like, containing PHD and RING finger domains, 1, Uhrf1||1415810_at||3.4||.006189|
|Tumor necrosis factor receptor superfamily, member 12a, Tnfrsf12a||1418571_at||3.2||.007977|
|Arginase type II, Arg2||1418847_at||3.4||.021006|
|Secreted phosphoprotein 1, Spp1||1449254_at||3.3||.002481|
|Checkpoint kinase 1 homolog (S. pombe), Chek1||1449708_s_at||3.3||.003636|
|PDZ binding kinase, Pbk||1448627_s_at||3.3||.000095|
|Calcitonin/calcitonin-related polypeptide, alpha, Calca||1452004_at||3.3||.009521|
|Aldo-keto reductase family 1, member B8, Akr1b8||1448894_at||3.3||.000588|
|Serine (or cysteine) peptidase inhibitor, clade A, member 3N, Serpina3n||1419100_at||3.2||.000509|
|Kinesin family member 23, Kif23||1455990_at||3.2||.000845|
|Protein regulator of cytokinesis 1, Prc1||1423775_s_at||3.1||.001146|
|Anillin, actin binding protein (scraps homolog, Drosophila), Anln||1433543_at||3.1||.000536|
|Serine (or cysteine) peptidase inhibitor, clade E, member 1, Serpine1||1419149_at||3.0||.003037|
|Coagulation factor VII, F7||1419321 at||3.0||.000498|
|Myelin and lymphocyte protein, T-cell differentiation protein, Mal||1417275_at||3.0||.000093|
|Thrombospondin 1, Thbs1||1460302_at||3.0||.01315|
|Cytidine 5′-triphosphate synthase, Ctps||1416563_at||3.0||.006285|
|Restin-like 2, Rsnl2||1427278_at||3.0||.00517|
|ERBB receptor feedback inhibitor 1, Errfi1||1416129_at||3.0||.01848|
|Glutamate-cysteine ligase, catalytic subunit, Gclc||1424296_at||2.9||.001408|
|Budding uninhibited by benzimidazoles 1 homolog (S. cerevisiae), Bub1||1424046_at||2.9||.00066|
|GINS complex subunit 4 (Sld5 homolog), Gins4||1452598_at||2.9||.001311|
|Cathepsin K, Ctsk||1450652_at||2.8||.002077|
|Nucleolar and spindle associated protein 1, Nusap1||1416309_at||2.8||.00123|
|Early growth response 2, Egr2||1427683_at||2.8||.000942|
|Transcription factor 19, Tcf19||1423809_at||2.6||.001691|
|Growth arrest and DNA-damage-inducible 45 gamma, Gadd45g||1453851_a_at||2.7||.010569|
|Prostaglandin-endoperoxide synthase 1, Ptgs1||1436448_a_at||2.5||.000417|
|MAD2 (mitotic arrest deficient, homolog)-like 1 (yeast), Mad2l1||1422460_at||2.5||.002757|
|Karyopherin (importin) alpha 2, Kpna2||1415860_at||2.4||.000024|
|Membrane-spanning 4-domains, subfamily A, member 4C, Ms4a4c||1420671_x_at||−3.5||.034934|
|Complement receptor 2, Cr2||1425289_a_at||−3.4||.003537|
|Interferon induced transmembrane protein 6, Ifitm6||1440865_at||−3.3||.034152|
|CD19 antigen, Cd19||1450570_a_at||−3.1||.000444|
|ATPase, Na+/K+ transporting, alpha 3 polypeptide, Atp1a3||1427481_a_at||−3.0||.005875|
|Nuclear receptor subfamily 1, group D, member 1, Nr1d1||1426464_at||−3.0||.005349|
|Insulin-like growth factor binding protein 3, Igfbp3||1423062_at||−2.9||.001285|
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.