Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Airway inflammation is the hallmark of asthma (1). Although numerous cell types, including T helper (Th) 2 lymphocytes, monocytes, macrophages, and epithelial cells, contribute to this inflammation, there is a large body of evidence implicating the eosinophil as the major effector of airway damage and dysfunction (1). Severity of disease correlates with the number and the state of maturation and activation of eosinophils detected in the bronchial mucosa and in bronchoalveolar lavage (BAL) fluid (BALF) of asthmatics (2). There is also a correlation between the resolution of eosinophilia and remission of asthma symptoms (1, 3). Eotaxin, a member of the CC family of chemokines (4), is a potent eosinophil attractant and can induce eosinophil superoxide and leukotriene production as well as granule release (5). In the lung, eotaxin is produced by resident cells (epithelial, endothelial, and smooth-muscle cells) as well as by inflammatory cells (macrophages, lymphocytes, and eosinophils themselves) (8). Eotaxin binds specifically and exclusively to CC-chemokine receptor (CCR)-3, a G protein- coupled, seven-transmembrane receptor present on eosinophils, basophils, and Th2 lymphocytes (9, 10).

Expression of eotaxin messenger RNA (mRNA) and protein is markedly upregulated after allergen challenge in several animal species (5, 11). Furthermore, administration of eotaxin antibodies markedly attenuates allergen-induced airway inflammation (11, 12).

There is also evidence of the importance of eotaxin in human asthma. Eotaxin and CCR-3 mRNA and protein are significantly elevated in bronchial biopsies from asthmatic compared with normal control subjects (13, 14). Levels of eotaxin are increased in sputum after allergen challenge (15) and serum levels of eotaxin are elevated during acute asthma exacerbations (16).

Although eotaxin is important for eosinophil recruitment in the early phase of the late asthmatic response, it does not appear to account for the ongoing eosinophil recruitment to the airways after allergen challenge. In asthmatic patients, levels of eotaxin mRNA and protein in BALF peak within 4 h but are lower 24 h after challenge (17). It is therefore likely that alternative mechanisms are operative for maintenance or persistence of eosinophilic inflammation in the later stages of allergen-induced inflammation and during prolonged clinical asthma exacerbations.

Recently, two new cytokines with properties similar to eotaxin (termed "eotaxin-2" [18] and "eotaxin-3" [19]) have been identified in humans. Eotaxin-2 is only 39% homologous to eotaxin and is located on chromosome 7q11.23 (as compared with chromosome 17q21.1 for eotaxin). The eotaxin-3 gene lies close to the eotaxin-2 gene on chromosome 7 but shares only 33% homology with it. Eotaxin is more potent than both eotaxin-2 and -3 in inducing calcium mobilization as well as chemotactic activity (19). Like eotaxin, these chemokines also bind specifically to the CCR-3 receptor.

Using a quantitative polymerase chain reaction (PCR) method, we have studied the expression of mRNA for eotaxin, eotaxin-2, and eotaxin-3 in the airways of nonatopic, nonasthmatic control subjects; stable mild asthmatic patients; and asthmatic subjects 24 h after allergen challenge. Although both eotaxin and eotaxin-2 mRNA were upregulated in asthmatic subjects, only eotaxin-3 was significantly upregulated 24 h after allergen challenge. Our data suggest that eotaxin-3 and not eotaxin may be an important mediator of late-phase eosinophil recruitment to the airways of asthmatic patients.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects

Three separate groups of 10 nonsmoking subjects were included in the study: mild asthmatics, mild asthmatics 24 h after allergen challenge, and normal nonatopic control subjects (healthy volunteers). Approval for performance of the study was obtained from the Hospital Ethics Committee. Asthmatic subjects were all considered to have mild disease requiring only the intermittent use of -agonist therapy and not requiring inhaled or oral corticosteroid therapy over the preceding 6 mo. All asthmatic subjects were positive to one or more common allergen skin-prick tests. All normal subjects had a provocative concentration of methacholine producing a 20% fall in forced expiratory volume at 1 s (FEV₁) (provocative concentration causing a 20% fall in FEV₁ [PC₂₀])  16 mg/ml with asthmatics having a PC₂₀ < 8 mg/ml. Allergen challenge subjects all demonstrated both an early and a late asthmatic response as detailed under challenge protocol.

Fiber-Optic Bronchoscopy

Bronchoscopy was performed according to published guidelines (20). Subjects undergoing bronchoscopy were required to fast from the midnight before the procedure. In allergen-challenge subjects, bronchoscopy was performed 24 h after challenge. All subjects were pretreated with nebulized albuterol (2.5 mg) and midazolam (5 to 10 mg intravenously). Oxygen (3 liters/min) was administered via nasal prongs throughout the procedure and oxygen saturation was monitored with a digital oximeter. After the administration of local anesthesia to the upper airways and larynx, a fiber-optic bronchoscope (Olympus BF10 [Olympus, Tokyo, Japan] or Pentax-FB15P [Pentax, Hamburg, Germany]) was passed through the nasal passages into the trachea. BAL was performed in the right middle lobe using four successive aliquots of 60 ml prewarmed 0.9% saline. Three or four endobronchial biopsies were taken from segmental and subsegmental airways of the right lower and upper lobe using a size 19 cupped forceps, and were placed in guanidium thiocyanate (Sigma, St. Louis, MO) and stored at 70°C for RNA extraction.

BALF cells were spun (500 × g; 10 min) and washed twice with Hanks' buffered salt solution. Cytospins were prepared and stained with May-Grunwald stain for differential cell counts. Cell viability was assessed using the trypan blue exclusion method. Aliquots of 5 × 10⁵ viable cells were placed in guanidinium thiocyanate and kept at 70°C for later RNA extraction.

Allergen Challenge Protocol

Subjects were challenged with the allergen that caused the biggest wheal on skin-prick tests. Allergen inhalation tests were performed using a handheld nebulizer (Respirgard-II; Marquest, Englewood, CO) with a flow rate of 5 liters/min. Freeze-dried allergen extracts (Aquagen SA; Aquagen, Nottingham, UK) were diluted to give final concentrations of 200, 1,000, 2,500, 5,000, 12,500, 25,000, and 50,000 IU/ml. The initial dose for the allergen inhalation test was 200 IU/ml, and FEV₁ was measured 5 and 10 min after each dose. FEV₁ was measured by dry wedge spirometer (Vitallograph, Buckingham, UK). Serially increasing doses of allergen were inhaled, and the cumulative dosage resulting in a 15% reduction in FEV₁ at 10 min was recorded. FEV₁ was measured at 10, 20, 30, and 45 min and 1 h; then at 30-min intervals until 3 h; and then at 4, 5, 6, 7, 8, 9, 10, 12, 16, 21, and 27 h. An early asthmatic response was defined as a > 15% fall in FEV₁ (postsaline FEV₁) at 10 min after antigen challenge and late asthmatic response as a > 15% fall in FEV₁ from baseline (post-saline FEV₁) between 4 and 10 h after challenge.

Quantitative PCR

BALF cells and endobronchial biopsies from all subjects were subjected to RNA extraction and reverse transcription performed using 200 U Moloney murine leukemia virus-reverse transcriptase (GIBCO BRL, Paisley, UK); 1 mM each of deoxyadenosine triphosphate, deoxycytidine triphosphate, deoxyguanidine triphosphate, and deoxythymidine triphosphate (Promega, Madison, WI); oligo deoxythymidine-15 primer (0.4 µg); ribonuclease inhibitor (Promega); 5 mM MgCl₂; 50 mM KCl; 10 mM Tris-HCl (pH 9); and 0.1% Triton X-100 in a total volume of 40 µl. Oligo-dT and RNA were heated for 10 min and placed on ice. The remaining ingredients were then added and the samples were incubated at 42°C for 90 min, followed by 10 min at 80°C. The complementary DNA (cDNA) was then diluted to a final volume of 200 µl in water and 10 µl was used for subsequent PCR reactions.

PCR was performed using 7.5 pmol of forward and reverse primers, deoxyribonucleotide triphosphate mix at a final concentration of 0.2 mM each, 1.5 U Taq polymerase (Promega), 1.5 mM MgCl₂, 50 mM KCl, 10 mM Tris-HCl (pH 9), and 0.1% Triton X-100 in a final volume of 30 µl.

Primer sequences were as follows:

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH): forward, 5'-GTTGCCATCAATGACCCCTTC-3'; reverse, 5'-CAT GTGGGCCATGAGGTCCAC-3' (903 base pairs [bp]).

Eotaxin: forward, 5'-CTCGCTGGGCCAGCTTCTGTC-3'; reverse, 5'-GGCTTTGGAGTTGGAGATTTTTGG-3' (227 bp).

Eotaxin-2: forward, 5'-CACATCATCCCTACGGGCTCT-3'; reverse, 5'-GGTTGCCAGGATATCTCTGGACAGGG-3' (288 bp).

Eotaxin-3: forward, 5'-GGAACTGCCACACGTGGGAGT GAC-3'; reverse, 5'-CTCTGGGAGGAAACACCCTCTCC-3' (354 bp).

PCR was carried out using a multiwell thermocycler (MJ Research, Inc., Waltertown, MA). The number of cycles used was determined in each case by performing a cycle profile study using a mix containing equal amounts of all sample DNAs and determining amounts of product obtained in two-cycle increments between 20 and 46 cycles. The number of cycles for subsequent PCR was determined according to the exponential part of the cycle curve and varied between 28 and 38 cycles for different reactions. Initial denaturation was performed at 94°C for 3 min followed by 28 to 38 cycles of denaturation at 94°C for 45 s, annealing at 60°C for 45 s, and extension at 72°C for 90 s. Final extension was at 72°C for 10 min.

To quantify PCR, cDNA obtained from normal lung tissue was subjected to PCR using specific primers for GAPDH, eotaxin, eotaxin-2, and eotaxin-3. PCR products were electrophoresed, removed from the gel, cleaned, and cloned into a pGEM-5Z vector using the pGEM-T vector system (Promega). Cloned products were sequenced to confirm authenticity. Bacteria containing each insert were grown and the inserted fragment of DNA was excised from the plasmid using restriction enzymes Apa I and Sac I (Fermentas, Vilnius, Lithuania) and quantified using a spectrophotometer (Beckman DU 530; Beckman Instruments, Fullerton, CA).

To plot a standard curve, serial half-log dilutions ranging from 10³ to 10¹² µg of starting DNA obtained from the plasmid were made.

All standards and samples were subjected to PCR simultaneously and in duplicate. After PCR, 5 µl of product was dot-blotted onto Hybond-N nylon membranes (Amersham, Bucks, UK), denatured (0.5 M NaOH and 1.5 M NaCl), neutralized (0.5 M Tris-base and 3 M NaCl, pH 7.5), and crosslinked to the membrane. Filters were then hybridized with [³²P]-labeled probes specific for eotaxin, eotaxin-2, eotaxin-3, or GAPDH, as appropriate, generated with a random primer labeling kit (Amersham). Prehybridization and hybridization buffer contained 10% dextran sulfate, 6× sodium chloride citrate (SSC), 0.5% sodium dodecyl sulfate (SDS), 5 mM ethylenediaminetetraacetic acid, 0.2% sodium pyrophosphate, 0.1 mg/ml salmon sperm DNA, and 10× Denhardt's solution, and was performed at 65°C. After hybridization, filters were rinsed twice in 3× SSC/0.1% SDS, then washed in 0.1× SSC/0.1% SDS at 65°C for 40 min and counted on a -counter. In addition, PCR products were electrophoresed on 1% agarose gel, denatured, neutralized, and probed as described earlier. These filters were then subjected to autoradiography.

A standard curve was constructed relating final radioactivity counts from the known standards to amount of starting DNA. PCR products from all subjects were then plotted on the standard curve and the amount of starting cDNA (in picograms) was determined for each (Figure 1). Amount of housekeeping gene (GAPDH) was determined in the same fashion and results are expressed as the ratio of eotaxin, eotaxin-2, or eotaxin-3 cDNA (in picograms) to GAPDH cDNA (in micrograms). Subjects in whom radioactivity counts from PCR products fell outside the range obtained from the standard curve for either target gene or GAPDH were excluded from analysis for that particular gene product.

View larger version (75K): [in this window] [in a new window]   Figure 1.   A representative standard curve used to quantitate PCR products. Serial dilutions of DNA were subjected to PCR simultaneously with DNA samples obtained from study subjects. The standard DNA samples were transferred to nylon membranes and hybridized with a 32P-labeled radioactive probe. (A) A representative agarose-gel showing progressive dilutions of eotaxin-3 PCR products (from 107.5 to 1013 µg). (B) Southern blot of samples from the same gel as in A. (C) A standard curve of amount of starting DNA plotted against radioactive counts from serial dilution samples (known standards).

Data Analysis

Median values and 95% confidence intervals (CIs) are given for each group. Data was analyzed using nonparametric analysis of variance statistics (Kruskal-Wallis) followed by Mann-Whitney analysis to determine differences within groups. P < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Demographic data, baseline FEV₁ values, and PC₂₀ in response to methacholine are listed in Table 1. No differences were observed between groups aside from PC₂₀ values in normal controls.

Differential cell counts from BALF cells are listed in Table 2. Eosinophil numbers were significantly greater in asthmatic than in normal control subjects, and significantly higher in asthmatics after allergen challenge as compared with baseline asthmatic subjects.

Eotaxin

Values obtained for eotaxin mRNA expression in bronchial biopsies and BALF cells from the three subject groups are illustrated in Figure 2.

View larger version (11K): [in this window] [in a new window]   Figure 2.   Analysis of eotaxin mRNA in biopsies (A) and in BALF cells (B) of normal control subjects, asthmatic subjects, and asthmatic subjects 24 h after allergen challenge. Values are expressed as the ratio of eotaxin mRNA (pg) to GAPDH mRNA (µg) (log values). Amount of starting cDNA was evaluated as explained in MATERIALS AND METHODS and Figure 1. Horizontal bar represents median value for all subjects. *P < 0.05 as compared with normal control subjects.

Biopsies. High constitutive levels of eotaxin mRNA were found in endobronchial biopsies obtained from normal control subjects. The median log value for eotaxin mRNA expressed relative to the housekeeping gene GAPDH was 3.18 pg/µg (95% CI: 2.27 to 3.79).

Levels were significantly higher in asthmatic subjects (median log 4.37 pg/µg, 95% CI: 3.97 to 4.65; P = 0.003). There was no further increase in eotaxin mRNA levels after allergen challenge (median log 3.81 pg/µg, 95% CI: 3.19 to 4.29); in fact, values were lower than those found in subjects with mild asthma, although this difference was not statistically significant.

BAL. Relatively high levels of eotaxin mRNA were obtained from BALF cells in all groups, with no significant differences between controls, asthmatics, and asthmatics after challenge.

Eotaxin-2

Values obtained for eotaxin-2 mRNA expression in bronchial biopsies and BALF cells from the three subject groups are illustrated in Figure 3.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Analysis of eotaxin-2 mRNA in biopsies (A) and in BALF cells (B) of normal control subjects, asthmatic subjects, and asthmatic subjects 24 h after allergen challenge. Values are expressed as the ratio of eotaxin-2 mRNA (pg) to GAPDH mRNA (µg) (log values). Amount of starting cDNA was evaluated as explained in MATERIALS AND METHODS and Figure 1. Horizontal bar represents median value for all subjects. *P < 0.05 as compared with normal control subjects.

Biopsies. Low levels of eotaxin-2 were found in biopsies of control subjects (median log 0.82 pg/µg, 95% CI: 0.08 to 1.72) with significantly elevated levels in asthmatic subjects (2.97 pg/µg, 95% CI: 1.97 to 3.45, P = 0.006) and in asthmatics after allergen challenge (2.49 pg/µg, 95% CI: 1.36 to 3.38).

BAL. Relatively high levels of eotaxin-2 were observed in BALF cells in all groups, with no significant differences between groups.

Eotaxin-3

Values obtained for eotaxin-3 mRNA expression in bronchial biopsies and BALF cells from the three subject groups are illustrated in Figure 4.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Analysis of eotaxin-3 mRNA in biopsies (A) and in BALF cells (B) of normal control subjects, asthmatic subjects, and asthmatic subjects 24 h after allergen challenge. Values are expressed as the ratio of eotaxin-3 mRNA (pg) to GAPDH mRNA (µg) (log values). Amount of starting cDNA was evaluated as explained in MATERIALS AND METHODS and Figure 1. Horizontal bar represents median value for all subjects. **P < 0.05 as compared with asthmatic subjects.

Biopsies. In sharp contrast to the findings with eotaxin and eotaxin-2, eotaxin-3 values were not significantly elevated in asthmatic as compared with normal subjects (median log 3.14 pg/µg, 95% CI: 1.62 to 3.53 versus 1.93 pg/µg, 95% CI: 0.74 to 3.92). However, values in asthmatic subjects after allergen challenge were significantly increased above those in asthmatic subjects (4.62 pg/µg, 95% CI: 3.05 to 6.23, P = 0.036) and over those in control subjects (P = 0.046). This corresponds to a more-than-100-times increase in mRNA expression compared with values in asthmatic subjects.

BAL. Values for eotaxin-3 in BALF cells were extremely low in all groups, with no significant difference in values obtained between normal, asthmatic, and allergen-challenge subjects.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Using quantitative PCR, we found that expression of eotaxin and eotaxin-2 mRNA is upregulated in endobronchial biopsies of asthmatic as compared with normal control subjects. Neither eotaxin nor eotaxin-2 was further increased 24 h after allergen challenge. In contrast, eotaxin-3 mRNA was not increased in asthmatic subjects before challenge but was increased by approximately 100 times 24 h after allergen challenge. For BAL, we found no significant differences between patient groups in the expression of eotaxin, eotaxin-2, and eotaxin-3 mRNA.

Eotaxin-3 was first identified by Kitaura and colleagues by a gene screen technique and maps to chromosome 7q11.23 (19). It is produced by endothelial cells after stimulation with interleukin (IL)-4 and IL-13 but not with tumor necrosis factor (TNF)-, IL-1, or interferon- (21). No data have been published to date regarding the cellular source of this cytokine in the lung. In vitro activity of eotaxin-3 as determined using calcium mobilization and chemotaxis assays is intermediate between that of eotaxin (the most potent) and eotaxin-2 (19). Eotaxin-3 binds specifically to the CCR-3 receptor, as do eotaxin and eotaxin-2.

Increased expression of eotaxin after allergen challenge has been clearly demonstrated in several animal models, including guinea pigs, mice, and monkeys (5, 22), as well as in humans (15, 17). Eotaxin has also been shown to be increased in exacerbations of asthma (16).

The importance of eotaxin for recruitment of eosinophils to the airways in vivo has been shown from studies in mice in which administration of eotaxin antibodies attenuates allergen-induced eosinophilic airway inflammation by 60 to 70%, although it does not abrogate the inflammation entirely (12, 23). Studies using eotaxin-deficient mice have shown a similar reduction in airway eosinophilia (24), although other investigators found no change in airway eosinophilia (25). Eotaxin may also be important for development of airway hyperreactivity, with studies demonstrating a 50% reduction in ovalbumin (OVA)-induced hyperreactivity in mice treated with anti-eotaxin antibodies (12).

Studies in vitro demonstrate that upregulation of eotaxin expression occurs rapidlywithin 1 h after TNF- stimulation (26). In mice, eotaxin mRNA induction occurs within 1 h after challenge, peaking at 3 h, followed by a rapid fall-off already seen 6 h after challenge (23). The kinetics of eotaxin expression seem, however, to vary between species and even between strains (22, 27). In asthmatic subjects after allergen challenge, levels of eotaxin mRNA and protein in BALF peak within 4 h and are lower 24 h after challenge (17). These findings are concordant with our results in that, although eotaxin mRNA was increased in asthmatic subjects, no further increase was observed 24 h after allergen challenge.

Thus, eotaxin expression seems to occur rapidly after challenge but declines within a short time period thereafter. Eotaxin-deficient mice sensitized and then challenged with OVA showed a 70% reduction in BALF eosinophil count 18 h after challenge as compared with wild-type mice, but no difference in eosinophil number was seen 48 h after challenge (24). Eotaxin-deficient mice immunized subcutaneously and then injected intracorneally with the nematode parasite Onchocerca volvulus demonstrated a markedly attenuated early eosinophilic inflammatory response (24 h). In contrast, delayed eosinophilic inflammation (8 d) was comparable to that seen in wild-type mice (24). It is therefore unlikely that eotaxin accounts for the ongoing eosinophil recruitment to the airways after allergen challenge. Rather, alternative mechanisms are likely to be operative for maintenance or persistence of eosinophilic inflammation in the later stages of allergen-induced inflammation and possibly also during prolonged exacerbations of asthma. Several potential candidate molecules may account for late-phase eosinophil recruitment. Ying and associates (28) found a correlation between 24-h tissue eosinophilia and eotaxin-2 expression in allergen-induced cutaneous responses. Ying and colleagues (13) also showed increased eotaxin-2 in atopic and nonatopic asthmatic subjects, but they did not perform allergen challenge on their patients. Although we also observed increased eotaxin-2 mRNA expression in asthmatic subjects before challenge, no significant change was seen after allergen challenge.

The striking upregulation of eotaxin-3 that we found 24 h after allergen challenge (up to 100 times!) makes this cytokine a likely candidate to account for prolonged or late-phase eosinophil recruitment. It is possible that other chemoattractants, such as monocyte-derived chemokine, may also play a role in late-phase eosinophil recruitment via non-CCR-3-mediated pathways (29).

Because our study was limited to the determination of mRNA expression, we do not know how long functional eotaxin protein persists in the airways after challenge. However, the decline in eotaxin protein at 24 h described by Brown and coworkers (17) and the low mRNA levels we observed at this time point make it unlikely that eotaxin protein levels remain significantly elevated beyond 36 h after challenge.

From our study, we cannot be certain that the marked upregulation of eotaxin-3 mRNA after challenge is accompanied by a parallel increase in functional eotaxin-3 protein, although we assume this to be the case. Thus, further studies using immunohistochemistry would be of value in confirming eotaxin-3 protein expression and allowing cellular localization of this chemokine in the airways.

Nevertheless, our findings are very striking and, despite the limitations we mention, strongly suggest that eotaxin-3 but not eotaxin is important for persistent eosinophil recruitment in the late phase of allergen challenge. Eotaxin-3 exerts its effect by binding to the same receptor as eotaxin CCR-3. Our findings have important implications in that development of therapeutic approaches to attenuate eosinophil recruitment to the airways would preferably be targeted at blocking the CCR-3 receptor rather than a specific chemokine, such as eotaxin.

In conclusion, we found that expression of eotaxin-3 but not eotaxin mRNA is markedly upregulated in the airways of asthmatic patients 24 h after allergen challenge. Our study further supports the contention (12) that airway inflammation in asthma results from coordinated expression of several chemotactic cytokines working at different time points.