Dissection of experimental asthma with DNA microarray analysis identifies arginase in asthma pathogenesis

Abstract

Asthma is on the rise despite intense, ongoing research underscoring the need for new scientific inquiry. In an effort to provide unbiased insight into disease pathogenesis, we took an approach involving expression profiling of lung tissue from mice with experimental asthma. Employing asthma models induced by different allergens and protocols, we identified 6.5% of the tested genome whose expression was altered in an asthmatic lung. Notably, two phenotypically similar models of experimental asthma were shown to have distinct transcript profiles. Genes related to metabolism of basic amino acids, specifically the cationic amino acid transporter 2, arginase I, and arginase II, were particularly prominent among the asthma signature genes. In situ hybridization demonstrated marked staining of arginase I, predominantly in submucosal inflammatory lesions. Arginase activity was increased in allergen-challenged lungs, as demonstrated by increased enzyme activity, and increased levels of putrescine, a downstream product. Lung arginase activity and mRNA expression were strongly induced by IL-4 and IL-13, and were differentially dependent on signal transducer and activator of transcription 6. Analysis of patients with asthma supported the importance of this pathway in human disease. Based on the ability of arginase to regulate generation of NO, polyamines, and collagen, these results provide a basis for pharmacologically targeting arginine metabolism in allergic disorders.

Figure 1

Microarray analysis of experimental asthma. In a, the induction of eotaxin-1 in allergen-challenged mice as measured by Northern blot analysis is shown. Total RNA (10 μg) was electrophoresed, transferred, and hybridized with a radiolabeled eotaxin-1 cDNA probe. The location of 18S RNA is shown. Each lane represents a separate mouse. Ethidium bromide (EtBr) staining of the RNA gel is also shown. In b, scatter plots of the average difference of present genes in two representative saline-challenged (left) or OVA-challenged (right) samples are shown. In c, the average difference of present genes in a representative saline-treated sample compared with a representative OVA-treated sample is shown. In d, quantitative analysis of the eotaxin-1 signal for saline- and OVA-treated mice is shown (n = 3 mice each). Error bars represent the SD. In e, the number of genes increased in experimental asthma induced with OVA or A. fumigatus antigen is depicted in a Venn diagram. Data are derived from statistically significant bioinformatic analysis as described in the Methods section. Data are available in supplementary Tables 1–4 (http://www.jci.org/cgi/content/full/111/12/1863/DC1).

Figure 2

Expression of L-arginine metabolizing enzymes. Expression of arginase I (a), arginase II (b), CAT2 (c), and iNOS (e) in ovalbumin (OVA) and A. fumigatus (Asp)-challenged mice as measured by gene chip analysis is shown. The average difference for the hybridization signal after saline (gray bar) and allergen (black bar) challenge is depicted (n = 2 for Aspergillus control group and n = 3 for OVA control group and OVA and Aspergillus experimental groups). Error bars represent the SD. A schematic representation of the L-arginine metabolism pathway is shown in d. Genes not present on the chip are depicted with a white box, genes present but not significantly increased with a gray box, and significantly increased genes with a black box. Abbreviations are AL, argininosuccinate lyase; AS, argininosuccinate synthetase; ODC, ornithine decarboxylase; and OAT, ornithine aminotransferase.

Figure 3

Northern blot and arginase activity analysis. In a, Northern blot analysis of arginase I and arginase II expression after OVA challenge is shown. Time points are as follows: 3H = 1 challenge, 3 hours; 18H = 1 challenge, 18 hours; 2C = 2 challenges, 18 hours. The EtBr–stained gel is also shown. The autoradiograph exposure times were 18 hours and 2 days for arginase I and arginase II, respectively. In b, the expression of arginase I and arginase II after intranasal challenges with A. fumigatus or saline is shown. Sal, saline; Asp, aspergillus. The autoradiograph exposure times were 18 hours and 6 days for arginase I and arginase II, respectively. In a and b, each lane represents a separate mouse. In c, arginase activity in the lungs of saline- and OVA-challenged mice (n = 4 mice and n = 3 mice, respectively) is shown. Arginase activity was measured in lung lysates with the use of the blood urea nitrogen reagent. As a control, arginase activity in the liver was 1522 ± 183 and 1390 ± 78 nmol/min/mg protein for saline- and OVA-challenged mice, respectively. In d, putrescine levels in the whole lungs of saline- and OVA-challenged mice (n = 4 mice and n = 7 mice, respectively) are shown. Putrescine levels were determined by HPLC.

Figure 4

Arginase I mRNA in situ hybridization. The hybridization signal of the arginase I antisense (AS) and sense (S) probes are shown for OVA/alum sensitized mice challenged with two doses of OVA (a–c, e) or saline (d). Tissue was analyzed 18 hours after the second saline or allergen challenge. Bright field (b, e, f) and dark field images (a, c, d) are shown at 100 × (a–d) and 400 × (e, f) original magnification. The dark field signal is white/pink and the bright field signal is black. In the paired dark and bright field photomicrographs (a and b), a peribronchial staining pattern is shown. The hybridization of the AS probe to a subpopulation of isolated inflammatory cells is shown in e. Staining was also observed in isolated large mononuclear cells with abundant cytoplasm, typical for airway macrophages. Examples of such cells stained by in situ (f) and immunohistochemistry against arginase I (g) are shown. Arrows indicate representative positive signal. Representative photomicrographs of four separate mice are shown.

Figure 5

Regulation of arginase by IL-4, IL-13, and STAT6. Northern blot analyses of arginase I and arginase II in IL-4 lung transgenic mice (in the Balb/c background) containing either the wild-type or deleted STAT6 gene (a) and lungs from Balb/c mice treated with IL-13 intranasally (c) are shown. The position of the 18S RNA is shown. Each lane represents a separate mouse. EtBr staining of the RNA gels is also shown. In b, arginase activity in the lungs of saline- and OVA-challenged (n = 4 and n = 3, respectively) wild-type (WT) and STAT6-deficient (STAT6-KO, n = 4 each) mice is shown. Arginase activity was measured in lung lysates with the use of the blood urea nitrogen reagent. In d, a kinetic characterization of IL-13 induced arginase mRNA levels in the lung is shown. Mice (n = 4–10 per group) received one dose of intratracheal IL-13 (10 μg) or PBS. Lung RNA was converted to cDNA and used for PCR analysis of arginase I (Arg I), arginase II (Arg II), or control hypoxanthine phosphoribosyltransferase (HPRT). The lane labeled “control” does not contain cDNA template.

Figure 6

Arginase I protein expression in human asthma. Fiberoptic bronchoscopy of allergic asthmatics and healthy controls was conducted, and BALF was analyzed for arginase I by immunohistochemistry (a). The number of immunopositive cells, expressed as a percentage of total cells, is shown. In b, a representative dark field illumination (blue filter, 200× original magnification) of arginase I mRNA in situ hybridization of cryostat section from an asthmatic biopsy specimen by using a S³⁵-labeled RNA probe is shown. Signal was detected mostly in the inflammatory cells in the mucosa (arrow). Patchy mRNA positive cells were also detected within the epithelium (arrowheads). These results are representative of four asthmatic individuals.