INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma is a complex clinical syndrome that probably results from interactions of environmental (1, 2) and genetic (3) influences. Although the degree to which genetic factors contribute to this complex phenotype is unclear, a number of different investigative groups have used genetic linkage techniques to identify human chromosome 5q31 as a region likely to contain genes related to the diagnosis of asthma (7), elevated serum IgE levels (3, 4, 8, 9), and bronchial hyperresponsiveness (3). Among the candidate genes in this region is the gene encoding for human interleukin-4 (IL-4). Genetic variants in the promoter region of the IL-4 gene (10) have been related to elevated levels of serum IgE; this locus has been associated with the diagnosis of asthma in some studies, but not in others (11- 13). Because IL-4 stimulation can influence mast cell responsiveness to IgE-mediated signaling (14) and because genetic variants in the IL-4 promoter have the capacity to modify IL-4 gene transcription, we reasoned that these sequence variants could modify asthma severity. To address this issue, we examined a large population of patients with asthma (ascertained without respect to genetic characteristics) for associations between a genetic variant in the IL-4 promoter region and the FEV₁.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

A cohort of 772 patients participating in a controlled trial of a new asthma medication was recruited from private practice offices and university hospital clinics. To be eligible for inclusion in the trial, patients had to have asthma as defined by American Thoracic Society criteria and had to be free of significant comorbid medical conditions (15). In addition, patients had to have been clinically stable for at least 6 wk while using medium-acting -agonists as their only asthma treatment, could not have been hospitalized for treatment of asthma more than once in the preceding 6 mo, could not have used oral or inhaled steroids for at least 6 wk, could not have used theophylline for at least 1 wk, and had to be between 12 and 70 yr of age. Thus, patients who were not clinically stable while receiving such a regimen were excluded from the sample cohort. At their first visit, patients had to have a FEV₁ (16) between 40 and 80% of predicted normal values after at least 8 h without inhaled -agonist treatment. Premenopausal women who were not using approved methods of contraception or who were pregnant were excluded from the trial, as were patients with a history of cigarette use for more than 10 pack-years in their lifetime and those who had smoked cigarettes at all in the 12 mo prior to the trial's initiation. The FEV₁ in the absence of bronchodilator medications was used as an index of asthmatic airway obstruction (16).

All participants provided EDTA-anticoagulated samples of whole blood and returned them to a central laboratory in a cold pack by overnight courier. On arrival in the laboratory, the specimens underwent DNA extraction (as detailed below) and plasma separation and were stored at 80° C. Written informed consent, including consent for genetic testing, was obtained from each patient and control subject in a form that had been approved by the appropriate institutional review board.

Genotyping

DNA was extracted from blood samples with either a commercial kit (QIAamp Blood Kit; Qiagen, Chatsworth, CA) or an automated nucleic acid purification system (Genepure; Applied Biosystems, Foster City, CA). The C to T mutation 589 base pairs (bp) upstream of the translation start site (Genbank access M23442) in the IL-4 promoter was detected by the loss of a restriction site for the BsmF1 restriction enzyme (13). A 198-bp fragment was amplified from genomic DNA and then digested by the BsmF1 restriction enzyme as described by Walley and Cookson (13). A 5'primer (5'TGGGTAAGGACCTTATGGACC3', nucleotides 683 to 663) and a 3'primer (5'GGTGGCATCTTGGAAACTGT3', nucleotides 486 to 505) were used to amplify a 198-bp fragment from genomic DNA samples. A polymerase chain reaction (PCR) mixture containing a total volume of 15 µl included the following: 50 ng of DNA, 1.5 µl of 10× PCR Buffer (Boehringer Mannheim, Mannheim, Germany), 1.5 µl of a 200 µM solution of deoxynucleoside triphosphates (adenosine, cytidine, thymidine, and guanosine triphosphates), 6 pmol of each primer, and 1 U of Taq polymerase (Promega, Madison, WI). Samples were heated to 94° C for 6 min and then underwent 35 cycles of 94° C for 15 s, 60° C for 30 s, and 72° C for 30 s followed by a final incubation at 72° C for 5 min in a Perkin-Elmer GeneAmp PCR system 9700 (Perkin-Elmer Medical Instruments, Pomona, CA). The restriction enzyme digestion mixture was added directly to the final 15-µl PCR solution to attain a final volume of 25 µl. The digestion mixture included 2 U of BsmF1 (New England Biolabs [NBL], Beverly, MA), 2.0 µl of NBL 10× Buffer no. 4, and 0.2 µl of NBL 100× BSA (10 ng/ml). Restriction digestion was allowed to proceed for a total of 2 h at 37° C. The fragments measuring 198 bp for the (589T allele) and 120 and 78 bp for the (589C allele) were separated on 2% agarose gels and visualized with ethidium bromide. Each reaction was repeated and confirmed by a second independent amplification and restriction enzyme digestion. Genotype assays using the PCR technique followed by BsmF1 restriction digestion were performed and repeated. Samples that were not amplified by PCR or that were not confirmed on repeat PCR/BsmF1 restriction digestion were excluded from the study.

IgE Levels

Total plasma IgE titers were assayed in 678 patients with a commercially available system (UniCAP). The results were analyzed as log₁₀ IgE titers because of the known logarithmic distribution of plasma IgE titers.

Statistical Analysis

All statistical analyses were performed with the SAS statistical package (SAS Statistical Institute, Cary, NC) on a SUN Microsystem. Allele frequencies in white and African American asthmatics were compared in a 2 × 2 table of specific allele counts (C versus T) with Fisher's exact test. Similarly, two-sided Fisher's exact tests were used to compare asthmatic subjects with FEV₁ values that were > 50% or < 50% of predicted; univariate odds ratios with logit confidence intervals were calculated. Hardy-Weinberg equilibrium was assessed with a goodness-of-fit chi square test. Multiple regression analyses were performed with PROC REG in the SAS package. One-way analysis of variance was used to compare FEV₁ (% predicted) and log₁₀ IgE values across genotypes.

To estimate the percentage of variation in FEV₁ (% predicted) attributable to the -589 polymorphism, the measured genotype approach developed by Boerwinkle and associates (17, 18) was applied. Non-IL-4 variance was estimated as the mean square error (MSE) in the analysis of variance of FEV₁ across IL-4 genotypes. The proportion of total variance related to the IL-4 polymorphism was calculated as the following ratio: (IL-4-related variance)/(IL-4-related variance + MSE).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects Enrolled

The baseline characteristics of the 772 subjects are listed in Table 1. Pulmonary function tests were performed by 771 subjects and total plasma IgE titers were assessed in 678 patients. FEV₁ values of > 50% and < 50% were used to categorize high or low lung function levels.

Allele Frequencies

Allele frequencies differed significantly (p = 1 × 10²³) between African American and white asthmatics (Table 2). Of the 682 Caucasians evaluated, 460 (67.4%) were homozygous for the wild-type allele (CC), 195 (28.6%) were heterozygous for the C and T alleles, and 27 (4.0%) were homozygous for the mutant TT allele. Among the 90 African Americans evaluated, 19 (21.1%) were homozygous for the wild-type allele, 44 (48.9%) were heterozygous, and 27 (30.0%) were homozygous for the TT allele. Thus, the frequency of the T allele was 0.183 and 0.544 among whites and African Americans, respectively; the observed distribution of genotypes within each racial group was consistent with Hardy-Weinberg equilibrium. Because the difference in allele frequency between races was highly significant, subsequent analyses of FEV₁ and IgE levels were stratified by race.

Association with FEV₁ Values

Among whites, the presence of the mutant IL-4 promoter allele was associated with a FEV₁ below 50% of predicted (p = 0.013) (Table 3). Multiple regression analysis of FEV₁ in white asthmatics was performed with log₁₀ IgE titer and number of T alleles (0/1/2) as dependent variables. The number of T alleles was significantly associated (p = 0.035) with low FEV₁ values independent of IgE level. A FEV₁ below 50% of predicted was documented for 59.3% of the 27 white patients with a TT genotype but for only 33.9% of the 460 white patients with a CC genotype. A FEV₁ below 50% of predicted was not significantly associated with the mutant allele among African Americans (p = 0.88). Furthermore, the distribution of FEV₁ values among white subjects with a TT genotype was markedly skewed (Figure 1). The estimated variance attributable to this genetic locus in FEV₁ was 0.6% when calculated by the method of Boerwinkle and colleagues (17, 18).

View larger version (27K): [in this window] [in a new window]   Figure 1.   Population density plots for FEV1 in white asthmatics stratified by genotype at the IL-4 promoter locus. Note that the homozygous T/T population has a lower mean and standard deviation than the homozygous C/C population.

Analysis of variance revealed no statistical relationship between the IL-4 promoter genotype and the log₁₀ plasma IgE level. The mean log₁₀ IgE titers by IL-4 promoter genotype are shown in Table 4.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have demonstrated an association between FEV₁ in white individuals, in the absence of chronic controller asthma treatment, and genotype at the C-589T IL-4 promoter locus. To our knowledge, this is the first report of an association between a genetic marker and disease severity in a cohort of patients with asthma. Because we found no association between plasma IgE level and genotype at this locus, our results indicate that genotype at this locus is a marker for a single component of the asthmatic phenotype examined in this study, namely, severity of airflow obstruction as indicated by the FEV₁.

The study cohort consisted of patients who were participating in a clinical trial of a novel asthma treatment. All patients enrolled in the trial had stable asthma, as indicated by the absence of asthma exacerbations during a lack of treatment other than as-needed inhaled medium-acting -agonists. Because the FEV₁ was measured at least 8 h after the use of inhaled -agonists, this value was presumably an acceptable index of asthma severity. The study criteria eliminated subjects whose initial FEV₁ values were < 40% or > 80% of predicted. Thus it is not surprising that the mean FEV₁ of 55.4 ± 12.1% in the entire study cohort indicated a moderate to severe overall level of asthma disease severity. Furthermore, since fewer than 1% of the subjects enrolled in the trial had FEV₁ values > 80% of predicted, very few subjects met the current criteria for mild asthma; it is possible that this genetic locus may not contribute to the variation in the FEV₁ in patients with mild forms of asthma. However, even with this cohort of patients whose range of FEV₁ values was limited, we found a significant association between FEV₁ and genotype at the IL-4 promoter locus. The modest size of our sample of African Americans may have accounted for our inability to demonstrate an association between asthma severity and genotype at the IL-4 promoter locus in this group. Moreover, the FEV₁ values reflected greater asthma severity in the African American cohort (< 50% of predicted in 47.2% of subjects) than in the white cohort (< 50% of predicted 36.4% of subjects).

Among white asthmatics, there was a difference of 4.5 percentage points of predicted in FEV₁ between subjects with the TT and those with the CC genotypes; our data explain 0.6% of the variance in FEV₁ observed in our study cohort. Studies of other complex diseases such as diabetes (19, 20) and hypertension (21), indicate that there may not be major loci that account for a large fraction of the total variance in airway obstruction. Rather, we speculate that it is more likely that each of a number of loci accounts for a small fraction of the total variance in lung function.

We found no association between genotype at the IL-4 promoter locus and IgE level. Although elevated IgE levels are strongly associated with bronchial hyperresponsiveness and atopy (24, 25), they have not been associated with asthma severity as measured by FEV₁. The implication is that the factors responsible for atopy and bronchial responsiveness may be distinct from those associated with FEV₁. Indeed, the lack of association noted by others (13) between the genotype at this locus and the diagnosis of asthma or atopy may be explained by the association of this genotype with factors mediating the severity of the obstructive response rather than its initiation.

We noted a striking difference in allele frequency at this locus between African American and white asthmatics. Because the overall levels of airway obstruction were similar in the two groups despite these differences in allele frequency, there is additional reason to believe that this locus is only one of many that influence the severity of airway obstruction in asthma.

We have identified a significant association between genotype at the IL-4 promoter locus and FEV₁ in Caucasian patients with asthma of moderate severity. This locus accounts for 0.6% of the variance in FEV₁ in this population, with a genotype-attributable difference in FEV₁ of 4.5 percentage points of predicted between subjects homozygous for the wild-type IL-4 promoter and those homozygous for the mutant form. Whether IL-4 itself is the link between asthma severity and genotype at this locus remains unknown. Nevertheless, knowledge of a patient's genotype at this locus serves as an independent index of asthma severity in white patients.