Heme Oxygenase-1-Mediated CD4⁺CD25^high Regulatory T Cells Suppress Allergic Airway Inflammation¹ Free

Heme oxygenase (HO)⁴ is the rate-limiting enzyme in the degradation of heme, converting heme to biliverdin IX, carbon monoxide (CO), and iron. Three isoforms of HO have been fully characterized. HO-2 and HO-3 are constitutive isozymes, whereas HO-1 is induced by a variety of stimuli including hemin and oxidative stress in various types of cells. The exact functional role of HO-1 expression in T cells is not fully understood (1). Recent basic and clinical research data suggest that the HO-1 gene serves as a protective gene, and that overexpression of HO-1 has significant biological consequences such as antioxidative, anti-inflammatory, and antiproliferative effects and modulation of cell cycle activities (2, 3, 4). These protective properties have been observed in various conditions such as asthma, transplantation, and hypertension, whereas the mechanisms of these conditions are quite different, even completely conflicting (2, 3, 4). For example, there is a Th2-specific response in asthma and a Th1-specific response in transplant rejection. Considering the protective role of HO-1 in both situations, it is inferred that HO-1 probably plays a regulatory role in Th1/Th2 balance.

Asthma is a chronic disorder of the airways, and is characterized by reversible airflow obstruction and airway inflammation, persistent airway hyperreactivity (AHR), and airway remodeling. The etiology of asthma is complex and multifactorial. Recent advances have demonstrated that, in the balance of Th1/Th2, Th2 lymphocyte-predominant immune response induces the change of cytokines, adhesion molecules, chemotactic factors, and mediators of inflammation, and plays a key role in the development of airway inflammation. The most recent data have shown the importance of regulatory T cells (Treg) in the development of asthma. The term “Treg cell” refers to cells that actively control or suppress the function of other cells, generally in an inhibitory fashion. The five most promising recent candidates of CD4 regulatory/suppressor T cells are Th3, Tr1, Tr, CD4⁺CD25⁺, and NK T cells. Among these, CD4⁺CD25⁺ regulatory T cells (CD4⁺CD25⁺ Treg) are a “naturally occurring” population of Treg cells (5), and its role in the initiation and orchestration of immune responses is studied widely.

CD4⁺CD25⁺ Treg is a population of CD4⁺ Treg cells produced by the thymus as a functionally mature T cell subpopulation, and constitutes 5∼10% of peripheral CD4⁺ T cells in nonimmunized naive mice. CD4⁺CD25⁺ Treg is the major portion of Treg cells, and plays an important regulatory role in the pathogenesis of allergic diseases and asthma (6, 7). Depletion of CD25⁺ cells results in the development of elevated IgE levels, eosinophilia, various autoimmune diseases, severe eczema, and food allergy (8, 9). In vitro studies show that it is the CD4⁺CD25^high subset, representing ∼2–3% of total CD4 T cells, that exhibits the regulatory characteristics and is identical with the CD4⁺CD25⁺ Treg. Thus, CD4⁺CD25^high cells are mainly CD4⁺CD25⁺ Treg (10, 11). CD4⁺CD25⁺ Treg suppress immune responses mainly through a direct cell-cell contact in a process that is dependent upon signaling via CTLA-4, as well as the secretion of IL-10 and TGF-β (5, 6, 12).

HO-1 has shown anti-inflammatory effects in several models of pulmonary disease including asthma. The characteristics and functions of CD4⁺CD25^high Treg have been well defined in murine and human systems. However, the interaction between HO-1 and CD4⁺CD25^high Treg remains unclear. The purpose of the present study is to illustrate CD4⁺CD25^high Treg modulated immunosuppressive function associated with HO-1 overexpression in the mouse asthmatic model using OVA immunization as well as treatment with hemin or Sn-protoporphyrin (SnPP).

A total of 225 female BALB/c mice (6∼8 wk of age) and 18 B6.129P2-Il10^tm1Cgn/J mice (4∼6 wk of age) was purchased from Shanghai SLAC Laboratory Animal Company. All mice were maintained under specific pathogen-free conditions in our animal facility. BALB/c mice were randomly divided into five groups including OVA, hemin, SnPP, hemin plus SnPP, and control groups. Each group included 15 mice, and three experiments were performed. B6.129P2-Il10^tm1Cgn/J mice were randomly divided into OVA, hemin, and control groups, and each group included six mice. The study has been approved by the Ethics Committee of Ruijin Hospital, Medical School, Shanghai Jiaotong University.

The mouse asthmatic model was established as described previously (13). In all groups except control, mice received i.p. injections of 100 μg (0.2 ml) of OVA (Calbiochem) complexed with alum (Sigma-Aldrich) on days 0 and 14. On days 14, 25, 26, and 27, mice were anesthetized with 0.3 ml, i.p., of pentobarbital (3 mg/ml) in normal saline before receiving intranasally 100 μg of OVA in 0.05 ml of normal saline (days 14) and 50 μg of OVA in 0.05 ml of normal saline (days 25, 26, and 27). Control animals received i.p. saline complexed with alum on days 0 and 14 and an intranasal saline without alum on days 14, 25, 26, and 27.

Mice sensitized and challenged by OVA received an i.p. injection of 75 μmol/kg hemin (Sigma-Aldrich) and of 75 μmol/kg SnPP (Porphyrin Products) on days −2, −1, 12, 13, 23, 24, and 27 in hemin and SnPP groups. Other mice received an i.p. injection of the same dose of hemin plus SnPP on the same days. Hemin and SnPP were dissolved in 0.1 mol/L NaOH and diluted with PBS.

Mice were treated with 75 μmol/kg hemin and SnPP, respectively, i.p. on days 0, 1, 12, 13, 24, and 25 to determine the time course of induction or inhibition of HO-1 expression and HO-1 activity.

Twenty-four hours after the final challenge (day 28), BALB/c mice were anesthetized by i.p. injection of ketamine/xylazine and sacrificed by cervical dislocation. BALF was obtained by the slow injection of ice-cold saline (0.3 ml) into the trachea using a 22-gauge i.v. catheter three times (total, 0.9 ml). This procedure always yielded 80∼90% recovery of the infused fluid. The total number of cells in BALF was counted using a hemacytometer. A differential count was done using cytosmear preparations. The cells were fixed and stained with Wright’s stain. Differential counts of 200 cells were done using standard morphological criteria to identify eosinophils. The absolute number of eosinophils in BALF was then calculated.

For staining with H&E, lungs were inflated and fixed with 10% buffered formalin after BALF cells were collected. Samples were embedded in paraffin, and then sectioned (4 μm), and stained with H&E.

The lungs were fixed in formalin (10%), and embedded in paraffin. Four-micrometer sections were mounted on poly-l-lysine-coated microscope slides. After paraffin removal and hydration, the sections were rinsed in PBS and incubated in blocking buffer (PBS containing 1% BSA and 2% normal goat serum) for 1 h at room temperature. They were then pretreated by heating for Ag retrieval. Subsequently, the slides were incubated with rabbit anti-human HO-1 Ab (1/200; Sigma-Aldrich) overnight at 4°C, and then washed in PBS and incubated with HRP-conjugated goat anti-rabbit IgG Ab (1/1000; DakoCytomation) for 1 h at room temperature. After washing with PBS, localization of HRP was revealed with DAB solution (DakoCytomation).

Flat-bottom 96-well ELISA plates were coated with 100 μg/ml OVA (overnight, 4°C, 0.06 mol/L NaHCO₃/Na₂CO₃ buffer (pH 9.6)) and blocked with 1% BSA (2 h, 37°C). Serum samples (1:10) were incubated (2 h, 37°C), followed by treatment with HRP-conjugated sheep-anti-mouse IgE (1/2000, 2 h, 37°C). Signal was developed with 3,3′5,5′-tetramethylbenzidine substrate (20 min) and stopped with 2 mol/L H₂SO₄. OD was measured with a spectrophotometer. OVA specific-IgE levels were expressed as OD values.

Concentrations of IL-10 and TGF-β in serum supernatants were determined (all from BD Pharmingen) according to the manufacturer’s protocols.

The flow cytometric analysis was performed on peripheral blood. Blood samples were collected in heparin-treated tubes from each subject. The cells were stained with a panel of fluorescently conjugated Abs: anti-CD4-FITC and anti-CD25-PE (BD Pharmingen). For each Ab combination, cells were incubated in the dark at room temperature for 30 min with mAbs at concentrations recommended by the manufacturer. Appropriate isotype controls were performed for each experiment. After incubation, RBC were lysed in the dark at room temperature for 10 min with hemolysin (Sigma-Aldrich). The samples were centrifugated for 5 min (1500 × g). The cells were washed once in PBS, centrifuged for 5 min (1500 × g) again, and then fixed in 1% paraformaldehyde, and counted on a FACScan (Bio-Rad). Analysis was performed with CellQuest software (Bio-Rad). A total of 50,000–100,000 events was acquired.

For RNA isolation, lung tissue was removed from the animal, immediately quick frozen in liquid N₂, and stored at −70°C until use. The frozen lung tissue was then homogenized in TRIzol reagent (Invitrogen Life Technologies), and total RNA was isolated according to the manufacturer’ instructions. Reverse transcription was performed using 40 ng of total RNA in a first-strand cDNA synthesis reaction with Sensiscript RT kit (Qiagen) and oligo(dT) primers (Invitrogen Life Technologies) as recommended by the manufacturer. A total of 1 μl of the resulting RT product was used for PCR amplification with HotStarTaq DNA Polymerase (Qiagen). Primers sequences used for PCR were as follows: Foxp3, 5′-AGG AGA AAG CGG ATA CCA-3′ and 5′-TGT GAG GAC TAC CGA GCC-3′; HO-1, 5′-GGC CCT GGA AGA GGA GAT AG-3′ and 5′-GCT GGA TGT GCT TTT GGT G-3′; HO-2, 5′-GAA GGA AGG GAC CAA GGA AG-3′ and 5′-GTT TTA GGC AGA GGT GGA GAT G-3′; GAPDH, 5′-GTG AAG GTC GGT GTG AAC GG-3′ and 5′-TCA TGA GCC CTT CCA CAA TG-3′. PCR amplifications were performed using the following four-cycle programs: 1) denaturation of cDNA (1 cycle, 95°C for 15 min); 2) amplification (30 cycles, Foxp3: 95°C for 0 min, 94°C for 45 s, and 58°C for 45 s; HO-1 and HO-2: 95°C for 0 min, 94°C for 45 s, and 60°C for 45 s; GAPDH: 95°C for 0 min, 94°C for 45 s, and 61°C for 45 s); 3) melting curve analysis (1 cycle: 95°C for 0 min, 72°C for 1 min, and 95°C for 0 min); and 4) cooling (1 cycle: 72°C for 1 min). PCR products were resolved by 1.5% agarose gel electrophoresis, stained with ethidium bromide, and photographed under UV light. Amplification of the appropriate cDNA resulted in a 170-bp fragment for Foxp3, a 888-bp fragment for HO-1, a 767-bp fragment for HO-2, and a 517-bp fragment for GAPDH. Samples were normalized using GAPDH expression. The results were expressed as a ratio of Foxp3, HO-1, and HO-2 cDNA to cDNA of the constitutively expressed GAPDH gene.

For HO-1 immunoblot, lungs were washed twice with ice-cold PBS and lysed in a buffer containing a broad-spectrum mixture of protease inhibitors consisting of 10 μg/ml aprotinin, 5 mmol/L EDTA, 1 μg/ml leupeptin, 0.7 μg/ml pepstatin A, 1 mmol/L phenylmethanesulfonyl fluoride, and Triton X-100. Protein concentration of lysates was measured by the bicinchoninic acid assay. Samples were separated in a 10% SDS-polyacrylamide gel and then transferred onto a polyvinylidene difluoride membrane. The membranes were incubated for 1.5 h with the anti-HO-1 Ab (1/500 dilution; StressGen) followed by incubation with peroxidase-conjugated goat anti-rabbit IgG Ab (1/10,000 dilution) for 1 h. Labeled protein bands were examined by using a chemiluminescence method according to the manufacturer’s recommendation (Amersham Biosciences). The membranes were then stripped and reprobed with antiactin (1:1000; Sigma-Aldrich) Ab.

HO enzyme activity in the mouse lung was quantified by assessing bilirubin generation with a modification of established techniques. Briefly, lungs were homogenized on ice in 1 vol of 100 mmol/L phosphate buffer with 2 mmol/L MgCl₂ and centrifuged for 15 min at 18,800 × g. The supernatant was used to measure HO activity. The reaction mixture, consisting of 200 μl of sample homogenate, 100 μl of normal liver cytosol (source of biliverdin reductase), 20 mmol/L hemin, and 0.8 mmol/L NADPH (Sigma-Aldrich), was incubated at 37°C for 30 min. The reaction was stopped by placing the mixture on ice, and OD_464∼530 was measured (extinction coefficient, 40 mmol/L/cm for bilirubin) to assess bilirubin generation. Values are expressed as picomoles of bilirubin formed per 0.5 h per milligram of protein. A NADPH-free reaction mixture provided a baseline against which the measured concentrations were determined.

AR was evaluated at 24 h after the last OVA challenge (day 28). Briefly, BALB/c mice were anesthetized by i.p. injection with 30∼50 mg/kg pentobarbital. After anesthesia, tracheostomy was performed using a 22-gauge, stainless-steel, three-way catheter, and an abdominal incision was made to expose the inferior vena cava. Airway pressure and airflow were measured by a pressure transducer and Respiratory Flow Head (MLT1L; ADInstruments) connected to the PowerLab system and were recorded on a strip chart recorder. Acetylcholine (Ach) (50 μg/kg) was injected into the inferior vena cava using a 22-gauge needle. The injections were maintained at a constant rate (over 1 s). AR to Ach was defined as the integrated change in inspiratory airway pressure time index (APTI; cm H₂O · s).

Spleens from each BALB/c mouse group were isolated, and a single-cell suspension was made by passing the spleen through a nylon cell strainer (Falcon; BD Biosciences). After lysis of red cells by ammonium chloride solution, CD4⁺ cells were purified by negative selection (by depletion of CD8a, CD11b, CD45R, CD49b, Ter-119-positive cells) with MACS (CD4⁺CD25⁺ T cell Negative Isolation kit; Miltenyi Biotec). After isolation of CD4⁺ cells, CD25⁺ cells were stained with PE-coupled anti-CD25 mAb and purified after the addition of anti-PE-coupled magnetic beads (Miltenyi Biotec). Starting with 30 × 10⁶ spleen preparations, CD4⁺CD25⁺ cells were obtained with a purity ranging from 90 to 95%. CD4⁺CD25⁻ cells were also collected with a purity ranging from 80 to 90%.

APCs were isolated from normal BALB/c mice splenocytes by plastic adherence. Briefly, single-cell suspensions were prepared and adjusted to a concentration of 5 × 10⁶/ml after lysis of RBCs in NH₄Cl buffer. Cells were plated in 15 × 100 mm petri dishes (5 × 10⁷/dish) and incubated for 1 h at 37°C in an atmosphere of 5% CO₂ and 95% air. Unbound cells were removed by washing twice with warm (37°C) PBS. Adherent cells were detached using a cell scraper, adjusted to 3 × 10⁷/ml, and exposed to 40 Gy gamma irradiation. Irradiated APCs were recounted and adjusted to a concentration of 1 × 10⁶/ml in culture medium.

Cultures were performed in RPMI 1640 supplemented with 10% heat-inactivated FBS, penicillin (100 U/ml), streptomycin (100 μg/ml), 2 mmol/L glutamine, 0.1 mol/L nonessential amino acids, 1 mmol/L sodium pyruvate, and 55 μmol/L 2-ME (all from Invitrogen Life Technologies). To analyze CD4⁺CD25^high T cell-mediated suppression, 2 × 10⁴ CD4⁺CD25⁻ T cells from normal mouse splenocytes (responder) were cultured in 96-well plates with 1 × 10⁵ irradiated APCs, Con A (10 μg/ml), and 2 × 10⁴ CD4⁺CD25⁺ T cells from each group (suppressor) for 72 h at 37°C in the presence of 5% CO₂ and 95% air. For each stimulation assay, CD4⁺CD25⁻ cells, CD4⁺CD25⁺ cells alone, and CD4⁺CD25⁻ plus CD4⁺CD25⁺ cells were assayed at the same time in quadruplicate. Cultures were pulsed with 1 μCi of [³H]thymidine/well for the last 16 h of culture. Proliferative responses in each independent experiment were determined by averaging the cpm of each quadruplicate assay. The relative proliferation of CD4⁺CD25⁻ cells in each of the two cultures was calculated as a ratio, and the proliferation of CD4⁺CD25⁻ cells alone was arbitrarily set at 100%. The assay was performed three times.

The cDNA samples from lung tissue were then subjected to real-time quantitative PCR, performed in 384-well PCR plates (Axgen) with ABI 7900 detector using the SYBR Green fluorescence quantification system (Applied Biosystems) to quantify amplifications. The sequences of the forward (-FW) and reverse (-RV) primers for β-actin, IL-4, and IFN-γ were as follows: β-actin-FW, CTA AGG CCA ACC GTG AAA AG, β-actin-RV, AGC CTG GAT GGC TAC GTA CAT; IL-4-FW, ACA GGA GAA GGG ACG CCA T, IL-4-RV, GAA GCC CTA CAG ACG AGC TCA; IFN-γ-FW, TCT TGA ACG GCA GCT CTG AG, IFN-γ-RV, TGG CGA CAG GTC ATT CAT CA. PCR amplifications were performed in a total volume of 10 μl containing 5 μl of SYBR Green Mix (Applied Biosystems), 200 nm of specific primers, and 2.5 ng of cDNA in each reaction. The threshold for positivity of real-time PCR was determined based on negative controls. Each PCR amplification was performed in duplicate wells under the following conditions: 95°C for 10 min, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. For mRNA analysis, the calculations for determining the relative level of gene expression were made according to the instructions from the User’s Bulletin from Applied Biosystems, by reference to β-actin in the sample, using the cycle threshold (Ct) method. The mean Ct values from duplicate measurements were used to calculate the expression of the target gene with normalization to a housekeeping gene used as internal control (β-actin), using the 2-A Ct formula, also according to the User’s Bulletin.

Data are presented as the mean ± SD. The two-tailed Wilcoxon two-group test was used to test for significant differences between comparisons within individual sample sets; p < 0.05 was considered to indicate statistical significance.

Various parameters were evaluated to assess lung inflammation at 24 h after the final challenge (days 28). Light-microscopic analysis revealed that the inflammatory cells were observed in the peribronchiolar and perivascular areas of sensitized and OVA-exposed BALB/c mice. The number of infiltrated cells, especially eosinophils around the airways and vessels, was significantly increased in OVA group mice, but significantly decreased after administration of hemin. Also, the infiltration of inflammatory cells was moderate in SnPP and hemin plus SnPP mice compared with OVA-sensitized/challenged and hemin-treated mice. Inflammatory cells were absent in control mice. However, the number of inflammatory cells did change in B6.129P2-Il10^tm1Cgn/J mice treated with hemin (Fig. 1).

OVA-sensitized/challenged mice had a 43-fold increase in total cells recovered in BALF compared with control mice. Pretreatment with hemin (hemin group) significantly reduced the total number of cells by 66% in BALF of OVA-sensitized/challenged mice (∗, p < 0.05, hemin vs OVA group) (Fig. 2,A). Meanwhile, the number of eosinophils in BALF from the OVA mice showed the most important proportional increase, being 37-fold higher compared with control mice (#, p < 0.001, control vs OVA). Pretreatment with hemin (hemin group) significantly reduced eosinophil infiltration into BALF by 68% (∗, p < 0.05, hemin vs OVA group) (Fig. 2,B). The numbers of total inflammatory cells and eosinophils in SnPP and in hemin plus SnPP mice were decreased to an intermediate level between OVA and hemin mice (Fig. 2, A and B).

We investigated whether the up-regulation of HO-1 could protect against OVA-sensitized/challenged inflammation. Immunohistochemistry, RT-PCR, and Western blot analysis revealed that the expression of HO-1 mRNA and protein in the lung tissue was significantly higher in both the OVA and the hemin group mice than in controls (∗, p < 0.05) (Figs. 3 and 4,A, Table I). Repeated administration of hemin further increased lung HO-1 expression significantly (∗, p < 0.05), accompanied by an increase in HO-1 activity (#, p < 0.01) (Fig. 5, A and B). Pretreatment with SnPP or hemin plus SnPP also increased the expression of HO-1 mRNA and protein in comparison with OVA group mice (‡, p < 0.01) (Fig. 5,A), but decreased HO-1 activity in the SnPP group (§, p < 0.01) (Fig. 5,B, Table I). There was no significant difference in the enzyme activity between the hemin plus SnPP group and the OVA group (p > 0.05). However, HO-2 mRNA expression was not different across all groups (p > 0.05) (Fig. 4,B, Table I).

To determine the time course of HO-1 expression and HO-1 activity, mice were treated with 75 μmol/kg hemin i.p. on days 0, 1, 12, 13, 24, and 25. Western blot analysis and HO-1 activity detection showed that maximal induction of HO-1 protein and HO-1 activity presented at 24 h (∗, p < 0.05) and began to decline 72 h after hemin treatment (∗, p < 0.05). However, HO-1 protein and HO-1 activity were at lower levels on days 12 and 24 (Fig. 6 A).

The time course of HO-1 expression and HO-1 activity was also determined after administration of SnPP. Mice were treated with 75 μmol/kg SnPP i.p. on days 0, 1, 12, 13, 24, and 25. Western blot analysis and HO-1 activity determination showed that the level of HO-1 expression was increased (∗, p < 0.05), but that the activity of HO-1 was distinctly inhibited at 24 h after SnPP treatment (∗, p < 0.05). HO-1 protein and HO-1 activity were also at lower levels on days 12 and 24 (Fig. 6 B).

These results indicated that it is essential to treat with hemin for induction of HO-1 on days −2, −1, 12, 13, 23, 24, and 27 before OVA sensitization and challenge.

OVA-sensitized/challenged BALB/c mice experienced significant elevations in the levels of serum OVA specific-IgE compared with control mice (‡, p < 0.001, control vs OVA, hemin, SnPP, and hemin plus SnPP groups). Hemin pretreatment reduced the OVA-sensitized/challenged increase in the levels of serum IgE more than OVA, SnPP, and hemin plus SnPP pretreatment (+, p < 0.05). There was no significant difference in serum IgE concentration among OVA, SnPP, and hemin plus SnPP mouse groups (p > 0.05) (Fig. 7,A, Table II). OVA-sensitized/challenged B6.129P2-Il10^tm1Cgn/J mice had the same high level of serum OVA specific-IgE compared with control mice (∗, p < 0.05), but the level of serum IgE in the hemin group was lower than that in the OVA group (+, p < 0.05) (Fig. 7 B).

The percentage of CD4⁺CD25^high Treg in the peripheral blood was measured with flow cytometric analysis. The results show that the percentage of CD4⁺CD25^high Treg was significantly decreased in OVA-sensitized/challenged mice compared with control mice (∗, p < 0.05, except the hemin group). Hemin pretreatment increased the percentage of CD4⁺CD25^high Treg cells compared with OVA mice (+, p < 0.05). Also, there was no significant difference in the percentage of CD4⁺CD25^high Treg cells among OVA, SnPP, and hemin plus SnPP groups (p > 0.05) (Fig. 8,A, Table II). To analyze the effects of HO-1 on function of CD4⁺CD25^high Treg, CD4⁺CD25⁺ cells were isolated and purified from spleen cells by the method of MACS. The percentage of the proliferation inhibition of T effector cells (CD4⁺CD25⁻) was regarded as suppressive capacity of CD4⁺CD25^high Treg. The data shows that suppressive capacity of CD4⁺CD25^high Treg was impaired after OVA challenge in OVA, SnPP, and hemin plus SnPP mice compared with controls (∗, p < 0.05), and there was no difference between the hemin and the control groups. The results show that overexpression of HO-1 by hemin protects cells from impairment of CD4⁺CD25^high Treg function during OVA challenge (Fig. 8,B, Table I). However, the suppressive capacity of CD4⁺CD25^high Treg was significantly impaired in B6.129P2-Il10^tm1Cgn/J mice and was not enhanced after treatment with hemin (Fig. 8 C).

To determine the effect of hemin or SnPP administration on AR, respiratory system pressure was measured via a port in the tracheal cannula and continuously recorded before and during exposure to Ach ∼24 h after the last OVA challenge. The APTI was measured as the AR parameter. APTI to Ach after OVA exposure in the OVA, SnPP, and hemin plus SnPP groups was seen to be significantly greater than the response of control mice. AR in the hemin group was significantly lower than in the OVA, SnPP, or hemin plus SnPP groups (∗, p < 0.05), and there was no significant difference found among OVA, SnPP, and hemin plus SnPP groups (p > 0.05) (Fig. 9).

To examine whether overexpression of HO-1 by hemin administration could reverse the Th-2 bias, the levels of IL-4 and IFN-γ mRNA in the lung were detected by real-time PCR, and the ratio of IFN-γ/IL-4 mRNA as Th-1/Th-2 balance was analyzed. The data show that there was a significantly higher ratio of IFN-γ/IL-4 in the hemin group compared with the OVA group (+, p < 0.05). Otherwise, the ratio was lower in the OVA, SnPP, or hemin plus SnPP group mice compared with controls (∗, p < 0.05) (Fig. 10, Table I).

The expression of Foxp3 mRNA of lung tissue was significantly decreased in OVA-sensitized/challenged mice compared with control mice (∗, p < 0.05, except hemin group). Hemin pretreatment increased the level of Foxp3 mRNA compared with OVA treatment (+, p < 0.05). Also, there was no significant difference in the levels of Foxp3 mRNA among OVA, SnPP, and hemin plus SnPP groups (p > 0.05) (Fig. 11, Table II).

OVA-sensitized/challenged mice experienced significant decreases in the level of serum IL-10 in OVA, SnPP, and hemin plus SnPP groups compared with control mice (∗, p < 0.05). The level of serum IL-10 in the hemin group was higher than that in the OVA group (+, p < 0.05). There was no significant difference in serum IL-10 concentration among OVA, SnPP, or hemin plus SnPP mice (p > 0.05) (Fig. 12,A, Table II).

Although there was a significant difference in serum IL-10 concentration between hemin and OVA groups, serum TGF-β concentration across all groups was similar (p > 0.05) (Fig. 12,B, Table II).

In the present study, we prepared a mouse model of asthma sensitized/challenged by OVA, and examined the role of HO-1-mediated CD4⁺CD25^high Treg cells in suppressing airway inflammation in OVA-sensitized/challenged BALB/c and B6.129P2-Il10^tm1Cgn/J mice treated with hemin and SnPP in vivo. The results showed that the animals sensitized/challenged by OVA had a higher level of serum OVA-specific IgE, increased numbers of inflammatory cells and eosinophils in BALF, and exhibited increased airway eosinophilia and AHR compared with control mice. This animal model had several of the characteristics of asthma. However, HO-1 obviously suppressed airway inflammation, including a decreased eosinophil infiltration around the airway, fewer inflammatory cells and eosinophils in BALF, a lower level of OVA-specific IgE, and attenuated AHR. Furthermore, HO-1 promoted the percentage of serum CD4⁺CD25^high Treg, enhanced their suppressive function and the level of lung Foxp3 mRNA, affected the levels of IL-4 and IFN-γ mRNA, and reversed the Th2 bias after pretreatment with hemin. The results further suggest that IL-10 mediates anti-inflammatory effects of HO-1. Those beneficial effects were reversed by additional pretreatment with SnPP.

Hemin is the major substrate of HO-1. A wide variety of agents, including heavy metals, cytokines, hormones, endotoxin, and heat shock, have been identified as strong inducers of HO-1 expression. The induction of endogenous HO-1 or exogenous administration of HO-1 via inducers and transgene provides protection against tissue injury and oxidative stress in many in vivo and in vitro models (14, 15). Pretreatment with hemin can induce HO-1 expression, and increase HO-1 activity before OVA sensitization and challenge, which persists for 72 h in our studies.

As a protective gene, HO-1 has a remarkable anti-inflammatory role in various diseases. Its precise mechanism is still under study (16, 17, 18). Choi et al. (19) have demonstrated that Foxp3, a marker of CD4⁺CD25⁺ Treg cells, induces expression of HO-1, and HO-1 engages in Foxp3-mediated immune suppression in Jurkat T cells in vitro. These results suggested that HO-1 might be an important effector of Foxp3-mediated immune suppression. Pae et al. (1) and Hori et al. (20) have found that HO-1 is differentially expressed between CD4⁺CD25⁻ and CD4⁺CD25⁺ T cell populations in a manner analogous to human Foxp3 expression. A further study showed that upon anti-CD3/anti-CD28 costimulation, human CD4⁺CD25⁻ T cells could also be induced to express HO-1, a finding that paralleled the expression pattern of human Foxp3 (21). This research also demonstrated that stable transfection of HO-1 into Jurkat T cells was capable of inhibiting cellular proliferation and cytokine production. This observation was blocked in the presence of an HO-1 inhibitor (17, 19). Those findings indicate that there is a close relationship between HO-1 and CD4⁺CD25⁺ Treg cells. However, there was no evidence about the role of HO-1 on CD4⁺CD25⁺ Treg cells in the mouse model of asthma. Our data show that up-regulation of HO-1 by hemin pretreatment cannot only markedly increase the percentage of CD4⁺CD25^high Treg cells in the peripheral blood and the level of Foxp3 mRNA expression in lung tissue, but can also enhance CD4⁺CD25^high Treg suppressive function. This is seen to be accompanied by decreased levels of serum OVA-specific IgE, reversing the Th2 bias with decreased airway eosinophilia, eosinophilia in BALF and airway inflammation, and by attenuated AHR. We therefore speculate that HO-1 may play an anti-inflammatory role by promoting the number of CD4⁺CD25^high Treg cells, and enhancing their function. However, when we pretreated mice with SnPP, an inhibitor of HO-1 activity, we did not observe a higher level of serum OVA-specific IgE, a lower percentage of serum CD4⁺CD25^high Treg and level of lung Foxp3 mRNA, or more severe airway inflammation compared with OVA group mice. Indeed, we found an impaired CD4⁺CD25^high Treg function in this group. HO-1 expression was higher in SnPP and hemin plus SnPP group mice than in OVA group mice. RT-PCR and Western blot analysis revealed that HO-1 was also up-regulated by SnPP administration, but that HO-1 activity was significantly inhibited. These results suggest that SnPP possesses a dual role, a finding that corresponds with the work of Sardana and Kappas (22). Furthermore, these findings illustrate that the remarkable anti-inflammatory role of HO-1 is correlated with its activity, and that the lack of aggravation of inflammation in SnPP mice might be due to maximal OVA-induced alterations that could not be further potentiated by HO-1 inhibition, or to a concomitant up-regulation of antioxidant systems that could compensate for HO-1 inhibition.

Although our study proves that HO-1 can up-regulate CD4⁺CD25^high Treg cells in a mouse model of asthma, the precise role of HO-1 is still under investigation. There is some evidence that HO-1, TGF-β, and IL-10 are interrelated. Both TGF-β and IL-10 can induce the expression of HO-1. In contrast, overexpression of HO-1 can increase the secretion of IL-10, and inhibition of HO-1 activity can decrease the secretion of TGF-β. TGF-β and IL-10 are not only the cytokines, by which CD4⁺CD25⁺ Treg cells exert their immunoregulatory effect, but they can also promote the conversion from CD4⁺CD25⁻ Treg cells to CD4⁺CD25⁺ Treg cells (23, 24, 25). Recent data show that TGF-β and IL-10 have a regulatory role in the immune response. In vitro as well as in vivo studies have evidenced a dual role of TGF-β that functions either as a pro- or anti-inflammatory cytokine on inflammatory cells, participating in the initiation and resolution of inflammatory and immune responses in the airways. TGF-β is also involved in the remodeling of the airway wall (26). IL-10 is an anti-inflammatory cytokine; it can inhibit the release of Th2 profile cytokines and production of TNF-α and IgE. IL-10 can also inhibit the survival of eosinophils and the Ab-induced airway eosinophilia. In addition, patients with asthma had insufficient production of IL-10 (27, 28). Therefore, we speculate that HO-1 may promote the number of CD4⁺CD25^high Treg cells by increasing the secretion of IL-10 and TGF-β. In this study, we determined the level of serum IL-10 and found a higher level of IL-10 in the hemin group compared with that in the OVA group. We also found that the administration of SnPP or hemin plus SnPP increased the level of IL-10 compared with that in the OVA group. Our studies showed further that overexpression of HO-1 by hemin guards against the impairment of CD4⁺CD25^high Treg function during OVA challenge in BALB/c mice. Furthermore, the CD4⁺CD25^high Treg suppressive function did not increase in B6.129P2-Il10^tm1Cgn/J mice after hemin treatment. These results strongly imply that HO-1 is an important intermediary that promotes the secretion of IL-10 to regulate inflammatory responses. Other studies have shown that the anti-inflammatory effects of HO-1 are mediated by CO-dependent down-regulation of synthesis of proinflammatory cytokines (TNF-α, IL-1β, and MIP-1β), while increasing production of the anti-inflammatory cytokine IL-10, and that CO mediates these anti-inflammatory effects via the p38 MAPK pathway (17, 29). CD4⁺CD25⁺ Treg cells rely heavily upon IL-10 for their development (30, 31). But Baechler-Allen et al. (32) has shown that FACS-isolated CD4⁺CD25^high Treg cells and suppressive clones generated from CD4⁺CD25^high Treg cells do not secrete IL-10. However, the level of serum TGF-β was similar across all groups in our study. Numerous studies with both mouse and human CD25⁺ T cells have failed to find a role for TGF-β, and TGF-β regulated Foxp3 expression in both murine and human CD25⁻CD4⁺ T cells, whereas it had little or no effect on Foxp3 expression in CD25⁺CD4⁺ T cells (33, 34, 35). Thus, the potential role of TGF-β in CD25⁺ T cell-mediated suppression remains controversial and deserves careful further study, particularly in view of the potential involvement of TGF-β in suppression in vivo. Indeed, pretreatment with hemin did not enhance the suppressive capacity of CD4⁺CD25⁺ Treg, but attenuated the airway eosinophilia and the levels of serum OVA-specific IgE in B6.129P2-Il10^tm1Cgn/J mice. Therefore, we consider that IL-10 plays a role in promoting the percentage and suppressive function of CD4⁺CD25^high Treg by HO-1, but is probably not a unique anti-inflammatory cytokine.

In conclusion, this study provides evidence that HO-1 has a remarkable protective role in a mouse model of asthma, probably via enhancing the secretion of IL-10 and promoting the number of CD4⁺CD25^high Treg cells.

We thank Sheri M. Skinner (Baylor College of Medicine, Houston, TX) for critical review of the manuscript.

The authors have no financial conflict of interest.