INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma is the most common chronic childhood disease in developed nations (1). Asthma is closely associated with increased levels of total serum immunoglobulin E (IgE) (2), elevated levels of IgE specific to common aeroallergens (3), and elevated blood eosinophil counts (4). Asthma is also typified by nonspecific increased airway responsiveness (AR) to inhaled agents such as histamine or methacholine (5). There is now substantial epidemiological and molecular evidence that these asthma-associated quantitative phenotypes are likely to be determined in part by genetic factors (3, 6).

Although much progress has been made toward defining the molecular genetics of asthma over the past decade, the intricacy of the numerous genetic and environmental factors involved has made genetic dissection of this and other complex diseases difficult (7). Power estimations strongly suggest that testing for linkage is significantly enhanced by the use of quantitative traits in preference to a binary asthma phenotype; the recurrence risk-ratio (_s, [8]) of physician-diagnosed asthma in the siblings of affected individuals has a value less than 2 (9). Using a binary asthma phenotype, even 500 affected sibling pairs will give less than 50% power to detect linkage to a given marker (9). It is therefore important to examine the genetic basis of asthma-associated physiological traits.

The number of biologically plausible candidate genes that might be involved in the determination of asthma and associated traits is very large (3). To date, research into the molecular genetics of asthma has generally focused on candidate genes with clearly defined roles in the allergic process. Several of these candidate loci have been investigated; most attention has focused on the cluster of cytokine genes on 5q31-33 and the gene for the  chain of the high-affinity receptor for IgE (Fc R1-) on 11q13.

In order to identify genetic factors contributing to asthma susceptibility, we investigated the 5q and 11q chromosomal regions for linkage to quantitative traits associated with asthma. The cytokine cluster in the chromosome 5q31-33 region and the chromosome 11q13 region had been previously linked to asthma phenotypes (10), principally serum total IgE levels. Both regions contain biologically plausible candidate genes (3). Further studies of these regions were required to clarify the existence and nature of possible linkages to asthma phenotypes. As for most investigations of complex diseases, replication of putative linkages is crucial (7), particularly given the controversy generated by the initial 11q13 linkage findings (14). Therefore, this study involved an investigation of microsatellite markers in these regions for linkage to total and specific serum IgE levels, peripheral blood eosinophil counts, and AR in a sample of Western Australian Caucasian nuclear families.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

Family data were collected from two sources. The first consisted of a longitudinal cohort of Caucasian children recruited before birth (15) who were reassessed at 6 yr of age together with their primary relatives (n = 95 families; 442 subjects). Probands from this population were recruited randomly from a suburban general hospital antenatal clinic in Perth, Western Australia and were born between 1987 and 1990. Families participating in the 6-yr follow-up of probands were unselected with regard to asthma status. The second source of data comprised a population of families ascertained from a respiratory outpatient clinic at the tertiary children's hospital in Perth, Western Australia on the basis of a Caucasian proband ages 6 to 15 with current, physician-diagnosed, severe, symptomatic asthma (n = 26 families, 134 subjects). Data from these 121 families were collected contemporaneously by the same investigators, and were combined for the purposes of linkage analysis.

All data collection activities were approved by the ethics committee of the Princess Margaret Hospital for Children. Informed personal and parental consent was obtained for all subjects participating.

Data Collection

Individual and family histories of respiratory symptoms, illnesses, and exposures were assessed by modified American Thoracic Society (16) questionnaires. Adults were administered the questionnaire. Questionnaires relating to children were administered to a parent, generally the mother.

Blood was taken by venipuncture from all consenting subjects (n = 510) for IgE assays and DNA studies. Venous blood was collected into polypropylene centrifuge tubes containing ethylene diamine tetra-acetic acid ([ETDA]; anticoagulant). A full-blood picture was performed on each subject from whom blood was taken; this included an automated white cell count in order to establish a total blood eosinophil count.

IgE Assays

Total serum IgE titers and specific serum IgE titers to whole house dust mite (HDM) (Dermatophagoides pteronyssinus; D1) and mixed grass (GX2) were measured using the Pharmacia FEIA CAP and RAST systems (Pharmacia Diagnostics, Uppsala, Sweden). Standard controls were included in both assays. Total serum IgE levels were expressed in international units per milliliter (IU/ml); specific serum IgE levels were expressed in RAST units. A "combined RAST index" was calculated for each individual as the sum of the RAST scores to HDM and mixed grass.

Airway Responsiveness

Spirometry was performed in the standing position using a hand-held spirometer (Model 61000; Welch Allyn, Skaneateles Falls, NY) calibrated daily with a 3-L syringe. The best FEV₁ was measured according to the guidelines of the American Thoracic Society (17).

Response to histamine challenge was assessed by the Yan rapid method (18). FEV₁ was measured after inhalation of saline and increasing (doubling) doses of histamine delivered from calibrated hand-held nebulizers. At each dose step one forced expiratory maneuver was performed unless it was thought to be technically unsatisfactory. The challenge was continued until either the FEV₁ fell by at least 20% from the postsaline value or until the maximal cumulative dose (7.8 µmol) of histamine was delivered. AR was expressed as the two-point dose-response slope (DRS) of histamine response against percentage fall in FEV₁ (19).

Molecular Analysis

DNA was extracted from whole blood samples by standard phenol- chloroform extraction. The genomic DNA of all individuals was genotyped using the polymerase chain reaction (PCR) for two polymorphic linkage markers in each of the 5q and 11q regions (Table 1). All four markers had been previously linked to atopy (11, 20, 21) or AR (22, 23) in Caucasian populations. For the purposes of linkage analysis, only families where both biological parents were assessed were included in the analysis. Genotyping and phenotyping were performed blind with regard to each other.

PCR reactions were performed in sterile 96-well thermostable microtiter plates (Costar) in 10-µl volumes overlaid with light mineral oil (Sigma, Sigma Chemical Company, St. Louis, MO). The reaction mix included 50 ng of genomic DNA, 0.25 U of DNA polymerase (Biotaq, Bioline, or T_TH+; Biotech International, British Biotech Pharmaceuticals, Oxford, UK), 200 µM of each deoxyribonucleoside triphosphate (dNTP) (Pharmacia Biotech, Bridgewater, NJ), 0.5 µM of each primer, 10 mM TRIS-HCl (pH 8.8), 50 mM KCl, 0.1% Triton X-100 (Bioline) and 1 to 1.5 mM MgCl₂ (as indicated by PCR optimization). Reactions were performed using a Hybaid Omnigene thermocycler (Hybaid, Ashford, UK) or a Corbett FTS-960 thermocycler (Corbet Research, Sydney, Australia) utilizing block control. Amplification conditions for all four markers were 28 cycles of 94° C for 1 min, 55° C for 1 min, 72° C for 45 s.

The oligonucleotide primers used for each microsatellite marker are given in Table 1. The PCR reaction mix for the microsatellite markers FceR1-_CA, D11S480, and D5S393 contained 1.5 mM Mg²⁺; for the microsatellite D5S399 the mix contained 1.0 mM Mg²⁺. Following amplification, fluorescently labeled PCR products were genotyped using a Model 373A automated DNA sequencer (Applied Biosystems, San Jose, CA), as described elsewhere (24).

Statistical Analysis

The principal outcomes of the Haseman-Elston sib-pair linkage analyses reported were the following quantitative traits: total serum IgE titer; combined RAST index to HDM and mixed grass; peripheral blood eosinophil absolute count; and the two-point DRS to histamine challenge. As age and sex were significantly associated with these phenotypes (data not shown), they were adjusted for age and sex using standard linear regression techniques. In order to investigate linkage to noncognate-antigen-specific versus antigen-specific IgE responses (25), total serum level adjusted for age, sex, and the combined RAST index was also analyzed. Total serum IgE levels, the blood eosinophil count, and the DRS were all skewed with a long right-hand tail, and hence were log_e transformed prior to analysis.

Multipoint sib-pair analyses ("interval mapping") (26) based on the Haseman-Elston algorithm were used to perform nonparametric identitity-by-descent (IBD) linkage analyses of quantitative traits to the typed markers. Linkage statistics were only generated by the interval mapping routine at the actual loci typed. Multipoint analysis was undertaken using the recombination fractions between the adjacent marker pairs estimated from regional mapping in a population-based sample of 230 Caucasian Western Australian nuclear families (27) (unpublished data). For the Fc R1-/D11S480 marker pair the recombination fraction () was 0.059 and for the D5S393/D5S399 marker pair  was 0.006. Multilocus sib-pair linkage analysis was performed using the program GAS v2.0 (copyright Alan Young 1995; Oxford University).

The joint nonindependence of multipair sibships was compensated for by adjusting the degrees of freedom for the t-distribution to reflect the effective sample size when assessing the significance of the least-squares fit (28).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Characteristics of Study Population

After excluding those families where only one biological parent was assessed, the dataset consisted of 113 nuclear families (n = 547 subjects). The mean number of subjects per family was 4.8 (range = 4 to 9). There were 321 offspring comprising 372 full sib-pairs. Table 2 summarizes the relevant characteristics of the study population for age, sex, asthma prevalence, and the outcome variables investigated. The sex ratio was balanced.

The observed heterozygosities of the microsatellite markers (Table 1) were consistent with those reported in the Genome Data Base (GDB; http://gdbwww.gdb.org/gdb-bin/wais). All four markers were highly polymorphic, and hence were likely to be informative in linkage analyses.

Sib-Pair Linkage Analysis

Sib-pair analyses (Table 3) gave no evidence of significant linkage of log_e total serum IgE levels to the chromosome 11q13 markers. The chromosome 5q31-33 markers demonstrated some evidence of linkage to log_e total serum IgE levels.

An analysis of linkage to log_e total serum levels adjusted for age, sex, and the combined RAST index was also performed (Table 4). The results were consistent with those of Marsh and coworkers (11, 25); evidence of linkage to the two chromosome 5q markers was greatly strengthened after the adjustment of total IgE levels for the specific IgE levels to common aeroallergens.

Sib-pair analyses suggested significant linkage of the combined RAST index to both chromosomal regions investigated (Table 3). There was no evidence that log_e blood eosinophil counts were significantly linked to the chromosome 11q13 markers; however, evidence of significant linkage to the 5q31- 33 region was found for the marker D5S399 (Table 3). The sib-pair analyses gave no evidence of significant linkage of log_e DRS to histamine to either the chromosome 5q31-33 or 11q13 microsatellite markers (Table 3).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current study suggested that one or more loci predisposing to elevated total and specific serum IgE concentrations and elevated eosinophil counts are present in the 5q31-33 region and that one or more loci predisposing to elevated specific serum IgE concentrations are present in the 11q13 region.

Not all sib-pairs within a population will be informative at a particular locus as the number of alleles shared IBD cannot be calculated unambiguously (e.g., offspring of genotype 12 whose parents are both genotype 12). In recognition of this fact, multilocus sib-pair methods were used (26) in order to extract maximal information from the available pedigrees. This method has the advantage of using information from adjacent markers to infer missing or ambiguous allele sharing (26). While this method is sensitive to undetected errors in the pedigree structure such as nonpaternity and adoption, which may result in spurious indications of linkage, great care was taken when obtaining pedigree information at interview. Simulation studies (26, 29) have suggested that the use of multiple marker information and the joint estimation of IBD probabilities generally result in modest increases in power to detect linkage and substantial decreases in bias of parameter estimates.

Serum IgE Levels

Marsh and coworkers (25) analyzed the linkage of the D5S393 and D5S399 markers to age- and sex-adjusted log_e total IgE levels unadjusted and adjusted for log_e specific IgE level to common allergens in 349 Amish sib-pairs. In their population these markers were closely linked to serum total IgE levels, but not to specific serum IgE levels. Linkage with the D5S393 and D5S399 markers was improved when either total serum IgE levels residualized for specific IgE levels to common aeroallergens or a subset of "nonatopic" sib-pairs (n = 128) who tested IgE antibody-negative to 20 common aeroallergens were analyzed (11, 25).

Similar findings of linkage between polymorphic markers on 5q31-33, including the D5S393 marker, and total serum IgE were made using both logarithm of the odds (LOD) score and sib-pair linkage analysis of 55 Dutch families ascertained through an asthmatic founder (13). Similar to our study, the linkage analysis indicated significant but not strong linkage of the D5S393 marker to log_e total serum IgE level (p = 0.01).

These studies (11, 13, 22) suggested that a major locus influencing total serum IgE levels is present in the 5q31-33 region, and that several linked loci may be influencing asthma-associated phenotypes. However, a study of 230 Australian Caucasian nuclear families (n = 1,020 subjects) failed to detect linkage of several 5q markers to total or specific serum IgE levels (30). A study of 131 Caucasian British families (n = 685 subjects) also found no evidence of linkage, using both parametric and nonparametric methods, to a microsatellite marker within the interleukin-9 (IL-9) gene in the 5q31 region (31).

The results of the current study were consistent with those of both Marsh and coworkers (11, 25) and Meyers and coworkers (13, 21) in finding evidence of significant linkage of D5S393 and D5S399 to total and specific serum IgE levels. Further, the current study, in agreement with the analysis of Marsh and coworkers (25), suggested that the linkage with total serum IgE levels was to one or more loci affecting total IgE in a nonallergen-cognate-specific fashion.

Shirakawa and coworkers (32) reported significant linkage of the Fc R1-_CA marker to "atopy" (defined similarly to Sandford and coworkers [20]) using LOD score methods in four extended Japanese families. Van Herwerden and coworkers (23) reported linkage of the Fc R1-_CA marker with atopy as defined by specific skin reactivity to common allergens (approximating the RAST index) in 106 Australian affected sib-pairs. The p value (p = 0.02) suggested significant but not strong linkage, similar to the current study.

The genome screen performed by Daniels and colleagues (27) confirmed previously reported linkages of the Fc R1-_CA microsatellite marker to both total serum IgE levels and to the sum of skin-test scores to common allergens (a measure closely approximating the combined RAST index [27]); but did not find linkage to markers in the 5q region and these phenotypes. The linkage of the Fc R1-_CA marker was stronger to the skin test index than to total serum IgE levels.

A study of 131 Caucasian British families (n = 685 subjects) found no evidence of linkage to log_e total serum IgE levels adjusted for age and sex, using both parametric and nonparametric methods, to several microsatellite markers, including D11S480, in the 11q region (31, 33). Collée and coworkers also reported no linkage of the Fc R1-_CA marker to atopy (defined similarly to Sandford and coworkers [20]) in 26 Dutch asthmatic sib-pairs (12).

Consistent with the results of previous positive studies (23, 27, 32), the current study found evidence of significant linkage of the Fc R1-_CA marker to specific IgE responses to common allergens.

Blood Eosinophil Counts

Few human linkage studies have investigated blood eosinophil counts as a phenotype. The genome screen performed by Daniels and colleagues (27) found evidence of significant linkage of log_e blood eosinophil count to markers in chromosomes 6 and 7 but not 5 or 11. Our study found evidence of significant linkage of the D5S399 marker to age- and sex-adjusted log_e blood eosinophil counts.

DRS to Histamine

Postma and coworkers (22) found linkage between polymorphic markers on 5q31-33, including the D5S393 marker, and provocative concentration of histamine causing a 20% reduction in FEV₁ (PC₂₀) (analyzed as both a binary and quantitative trait) using affected and Haseman-Elston sib-pair linkage analyses of 84 Dutch two- or three-generation pedigrees (n = 500 subjects) ascertained through an adult founder with bronchial hyperresponsiveness (BHR) to histamine (22). Van Herwerden and coworkers (23) reported linkage of the Fc R1- gene in 53 Australian affected sib-pairs concordant for increased AR to methacholine but without atopy as defined by positive skin reactivity to common allergens. This study did not include a measurement of serum IgE levels, so it is somewhat difficult to compare to other linkage studies of the 11q13 region.

In contrast to both of these studies (22, 23), the genome screen performed by Daniels and colleagues (27) found no evidence of significant linkage to the log_e DRS to methacholine for markers in either the 5q31 or 11q13 regions. The results of the current study, which suggested no significant linkage between the log_e DRS to histamine and the markers investigated, were consistent with these findings. The negative results of the current analysis did not suggest the presence of a common gene of major effect on AR to histamine linked to the chromosomal markers investigated in this population. However, neither a rare gene nor a common gene of small effect could be excluded.

In conclusion, the finding of linkage of both log_e total serum IgE adjusted for the RAST index and the RAST index to the chromosome 5q31-33 region and the RAST index but not log_e total serum IgE to the 11q13 region is consistent with the results of previous segregation (34, 35) and molecular (11) studies, which have also suggested the existence of multiple genes independently regulating total (noncognate-antigen-specific) and specific serum IgE levels. The work of Xu and coworkers (21) has suggested the existence of at least two unlinked loci regulating total serum IgE levels. However, unlike the study of Marsh and coworkers (11, 25), the current study also found evidence of significant linkage between the 5q31- 33 region and specific IgE responses to common allergens. These results are consistent with the existence in the 5q31-33 region of one or more genes regulating specific serum IgE responses and one or more genes regulating total serum IgE levels in a nonallergen-cognate-specific manner. Differences between our results and those of others may be attributable to genetic and environmental differences between study populations, the use of differing statistical methods, and/or differing phenotypic definitions.

The results of this study suggested that both the 5q31-33 and 11q13 chromosomal regions may contain loci regulating physiological traits closely associated with asthma, and are therefore worthy of further molecular genetic investigation. The suggested linkage of several different phenotypes and markers on 5q may reflect the presence of several different genes or the pleiotropic effects of one gene. These results are also consistent with the marked heterogeneity underlying asthma noted by other workers (3).

The large number of candidate genes present within a relatively short genetic region complicates gene mapping in both the 5q31-33 and 11q13 regions. The chromosome 5q31-33 region contains a cluster of cytokine genes (13), including interleukin 4 (IL-4), IL-5, IL-9, IL-13, and their receptors. Other candidate genes in this region include granulocyte-macrophage colony-stimulating factor (GM-CSF), fibroblast growth factor acidic (FGFA), and the ₂-adrenergic receptor gene. The chromosome 11q13 region also contains a number of known candidate genes (20), including the Fc R1- gene and the gene for CC16. Given current technical limitations on mapping to intervals closer than 1cM and the poorly defined physical map of these regions, the precise definition of which genes in these regions are responsible for the reported linkages to total and specific serum IgE levels and to blood eosinophil counts may await completion of high-resolution physical maps of the human genome.