Introduction

Atopy is a complex trait characterized by predisposition to hyperproduction of immunoglobulin E (IgE) against common environmental allergens. It is a major risk factor for the development of allergic diseases such as asthma, rhinitis, and dermatitis. The susceptibility to atopic diseases has an important hereditary component. In the past 12 years, a number of genome-wide studies of atopy in humans identified several IgE-controlling loci and genes on different chromosomes (reviewed in Ober and Hoffjan 2006; Vercelli 2008). Some of these genes or loci were found to regulate susceptibility to atopic diseases in several populations, whereas others have been detected in one or a few populations only (reviewed in Ober and Hoffjan 2006). These controlling loci in most cases contain a number of genes, and it is not known which of them is responsible for the observed effects. It was suggested that in different populations, different genes from the same chromosomal region may be involved in control of susceptibility (reviewed in Ober and Hoffjan 2006; Zhang et al. 2008).

The most often detected linkages were to chromosomal regions 5q (Xu et al. 2000; Yokouchi et al. 2000, 2002; Haagerup et al. 2002; Koppelman et al. 2002), 6p (Daniels et al. 1996; Wjst et al. 1999; Haagerup et al. 2002; Fereira et al. 2005), 7p (Daniels et al. 1996; Laitinen et al. 2001; Shugart et al. 2001; Altmüller et al. 2005), 7q (Xu et al. 2000; Koppelman et al. 2002; Altmüller et al. 2005), 11q (Daniels et al. 1996; Shugart et al. 2001; Altmüller et al. 2005), 12q (Xu et al. 2000; Koppelman et al. 2002; Yokouchi et al. 2002), and 16q (Daniels et al. 1996; Ober et al. 2000; Kurz et al. 2005). Positional cloning indicated six genes at loci 2q14 (DPP10—dipeptidyl serine protease; Allen et al. 2003), 2q33 (CTLA4—cytotoxic T-lymphocyte-associated-4 gene; Howard et al. 2002), 5q32-33 (PCDH1—protocadherin-1; Whittaker 2003), 7p14.3 (GPRA—G-protein-coupled receptor; Laitinen et al. 2004), 13q14 (PHF11—PHD finger protein 11; Zhang et al. 2003), and 20p13 (ADAM33—Zn-dependent metalloproteinase; van Eerdewegh et al. 2002) predisposing for atopy or atopy-associated traits. Genome-wide association mapping led to the identification of the genes ORMDL3 (an endoplasmic reticulum membrane protein) at locus 17q21 (Moffatt et al. 2007) and CHI3L1 (chitinase 3-like 1) at locus 1q32.1 (Ober et al. 2008) that contribute to the risk of asthma.

Despite this remarkable progress in the identification of genes controlling atopy and asthma in humans, the complete elucidation of its genetics is hindered by many factors including sample size, genetic heterogeneity of human populations, gene interactions, low frequency and/or incomplete penentrance of trait-controlling alleles, and a high variability of environmental factors (Lander and Schork 1994).

Some limitations of human genetic studies could be overcome by the use of mouse models. The availability of genetically homogenous mouse strains and possibility of testing large numbers of F2 and backcross mice in a controlled environment that reduces the phenotypic variance makes the mouse a useful model for study of the complex traits in human (Lipoldová and Demant 2006). Once the genetic regions of interest have been identified in mouse, the high level of synteny between many mouse and human chromosomal segments allows predicting their locations in human (DeBry and Seldin 1996). A mouse model has been successfully applied for the identification of human homologues of mouse asthma genes Tim1 and Tim3 in Tapr (an airway hyperreactivity regulatory) locus at chromosomes 5q33.2 and 5q33.3, respectively (McIntire et al. 2001).

In our previous genome-wide search performed in mouse, we described nine genetic loci on chromosomes 1, 2, 3, 4, 5, 8, 10, 16, and 18 that control IgE level (Lipoldová et al. 2000; Badalová et al. 2002). This search did not target certain genes throughout the genome, but a set of genetic regions of total length of about 360 Mb. These regions are distributed randomly in the tested recombinant congenic (RC) strains. It must be emphasized that this occurred without any previous selection and without any prior knowledge about the gene content of these regions. The description of the principle of construction of RC strains is described in Demant and Hart (1986) and Lipoldová and Demant (2006). Subsequently, we used the homology between genetic maps of mouse and human to identify the corresponding orthologous regions on human chromosomes and found that the loci Lmr3, Lmr5, Lmr8, Lmr10, Lmr11, Lmr13, and Lmr14 (Lipoldová et al. 2000; Badalová et al. 2002) are located in the regions homologous with the human chromosomal segments known to control serum IgE in human atopic diseases (Wjst et al. 1999; Dizier et al. 2000; Xu at al. 2000, 2001; Yokouchi et al. 2000, 2002; Koppelman et al. 2002), indicating a likely relatedness of IgE-controlling genes in the two species. However, for two loci (Lmr9 and Lmr12; Badalová et al. 2002) described by us, the homologous human regions have not been connected with atopy. These two loci may point to hitherto undetected human genes that are relevant for atopy. As Lmr12 maps to a broad segment, we used for further study the locus Lmr9, which is rather precisely mapped to a segment with the most likely length of 3.58 Mb and maximal possible length of 9.32 Mb on chromosome 4 in the strain CcS-20. The mice homozygous for BALB/c (high IgE responder) and STS/A (low IgE responder) alleles at this locus differed 1.6 times in IgE level (corrected P value 0.00313; Badalová et al. 2002). In the orthologous region on human chromosome 8q12, we selected three short tandem repeat (STR) markers, D8S1828 (56.96 Mb), D8S285 (57.22 Mb), and D8S1816 (57.52 Mb; Fig. 1), and tested their non-parametric linkage and association (quantitative and discrete traits, QTDT) with inhalant and food atopy and with levels of total IgE and specific IgEs to 20 inhalant and five food allergens in the 67 Czech atopic nuclear families comprising 276 subjects. In order to define other loci that may control IgE in Czech atopic patients, we also selected markers in additional 16 candidate chromosomal regions for linkage and association testing with the atopic phenotypes (Table 1). These additional regions were previously shown to be linked or associated with atopy and/or asthma (Daniels et al. 1996; Wjst et al. 1999; Ober et al. 2000; Xu et al. 2000; Yokouchi et al. 2000, 2002; Laitinen et al. 2001; Shugart at al. 2001; Haagerup at al. 2002; Koppelman et al. 2002; van Eerdewegh et al. 2002; Allen et al. 2003; Altmüller et al. 2005; Ferreira at al. 2005; Kurz et al. 2005).

Fig. 1
figure 1

Conserved synteny between the mouse and human genome regions. The region on recombinant mouse chromosome 4 (Lmr9 locus) is homologous to the region on human chromosome 8q12. In the mouse chromosomal map, the segments of BALB/c origin, where the Lmr9 locus is excluded, is marked open. The segment of STS origin, containing the Lmr9 gene, is marked closed. The gray segment indicates region of undetermined origin. The most likely and the maximal lengths of Lmr9 locus in mouse are 3.58 and 9.32 Mb, respectively. The 530 Kb region shown in detail encompasses genes in the close vicinity of the marker D4Mit149. The position of the marker is given as 0 cM because it has not yet been separated from the centromere by recombination (http://www.informatics.jax.org/, August 25, 2008). The markers in the segment on human chromosome 8 were selected so that one is located in the center of the ±5 Mb region orthologous to Lmr9 (D8S285) and the two other markers are approximately equidistant centromerically (D8S1828, 260 kb) and telomerically (D8S1816, distance 300 kb; NCBI Homology Maps page—http://www.ncbi.nlm.nih.gov/projects/homology/maps/)

Full size image
Table 1 List of markers tested for linkage and association with asthma, rhinitis and dermatitis, total IgE, and specific IgE to 20 inhalant and five food allergens
Full size table

Hence, in the present study, instead of performing total genome scans in the analyzed families, we tested only the markers in the chromosomal regions that were either previously shown to define linkages in humans or representative markers in the human chromosomal region that is homologous to the mouse Lmr9 locus that controls IgE levels (Table 1).

Materials and methods

Subjects and families

Nuclear families from the Czech Republic (67 families, n = 276) originated from Prague (47 families, n = 192) and from towns Ústí nad Labem, Teplice and Most (13 families, n = 55) and Trutnov (7 families, n = 29), all located within less than 120 km from Prague. The families were collected through probands registered in local clinics as patients with a medical history of atopic disease. All available families were recruited into the study with the exception of 8% of families, which declined to participate for different reasons. The probands were not ascertained for another disorder.

The 67 nuclear atopic families contained 276 subjects, 138 of whom were offspring (Table 2, Electronic supplementary material Table 1). There was no stated relatedness between families. The mean and median age of the parents were 43.4 and 42.0 years, respectively, and that of the offspring were 15.9 and 14.0, respectively. The sex ratio of the offspring was 0.53:0.47 (male to female). The percentage of the allergen-sensitized subjects among parents was 84.8% and among offspring 87.7%.

Table 2 Characteristics of group of 276 participants from 67 families
Full size table

All participants (offspring and spouses) have undergone clinical examination under the protocol approved by the Ethical Committee of the Third Faculty of Medicine, Charles University, Prague, Czech Republic. A full explanation of the study design was given to all participants, and subsequently, a written consent was obtained. Clinical specialists performed a structured interview with each participant (or his/her guardians), verified or newly established the diagnosis of asthma, rhinitis, dermatitis, conjunctivitis and/or urticaria according to EAACI guidelines (Johansson et al. 2001), and completed a questionnaire about the disease manifestations and smoking status.

Estimation of total and specific IgE levels

The collection of blood samples was conducted from February 1999 to February 2001. No blood samples were collected during the summer months. The sera were stored at −70°C before use. The total IgE level was estimated in by CAP-FEIA (Pharmacia, Uppsala, Sweden). Specific IgE was measured by the in vitro test system EUROLINE (EUROIMMUN, Medizinische Labordiagnostika GmbH, Lübeck, Germany) according to the instructions of the manufacturer. In this system, allergen extracts were used for the detection of specific IgEs. The lowest threshold of detection was 0.35 kU/l. We have tested 20 inhalant and five food allergens. Sensitization to moulds (m6—Alternaria alternata, m3—Aspergillus fumigatus, m2—Cladosporium herbatum, m1—Penicillium notatum), animals (e3—horse, e2—dog, e1—cat), mites (d2—Dermatophagoides farinae, d1—Dermatophagoides pteronyssinus), weeds (w9—Plantago lanceolata, w6—Artemisia vulgaris, w1—Ambrosia elatior), trees (t7—Quercus alba, t4—Corylus avellana, t3—Betula verrucosa, t2—Alnus incana), and grasses (g12—Secale cereale, g6—Phleum pratense, g3—Dactylis glomerata, g1—Anthoxantum odoratum) was measured by inhalation test system. We also measured reactivity to celery (f85), potato (f35), almond (f20), hazelnut (f17), and rice (f9). The inhalant and food atopy were defined as sensitization to at least one of the tested inhalant and food allergens, respectively.

Genetic markers

For the analysis, 19 STR markers located at different chromosomes/chromosomal regions were selected (Table 1) from the National Centre for Biotechnology Information (NCBI) database (http://www.ncbi.nih.gov). The STR markers are characterized in the Marshfield genetic map and show high heterozygosity. All markers, with the exception of the markers on chromosome 8 (D8S1828, D8S285, and D8S1816), are located in atopy candidate regions previously described in genome-wide studies of atopy and/or asthma (Daniels et al. 1996; Wjst et al. 1999; Ober et al. 2000; Xu et al. 2000; Yokouchi et al. 2000, 2002; Laitinen et al. 2001; Shugart et al. 2001; Haagerup et al. 2002; Koppelman et al. 2002; Altmüller et al. 2005; Ferreira et al. 2005; Kurz et al. 2005) by other groups. Markers selected in regions 2q14.3 (Allen et al. 2003), 5q31.1 (Koppelman et al. 2002), 5q33.3 (Yokouchi et al. 2000), 6p21 (Wjst et al. 1999), 7p14.1 (Daniels et al. 1996, Laitinen et al. 2001; Shugart et al. 2001), 7q22.1 (Xu et al. 2000), 12q21.31 (Nickel et al. 1997), 16q23.2 (Ober et al. 2000), and 19q13.32 (Venanzi et al. 2001) were exactly those that exhibited linkage or, in region 11q12 (Adra et al. 1999), were located in close proximity. Chromosome 12q harbors multiple genetic loci related to asthma and asthma-related phenotypes including atopy, distinct peaks of linkage being observed in different populations (Raby et al. 2003). Markers in regions 12q13 and 12q24 were selected in positions that would enable to test presence of these linkages in the studied population. Similar approach was used in the selection of marker in 16q21 in the vicinity of linkages detected by (Daniels et al. 1996) and (Kurz et al. 2005). In tests of 13q14 and 20p13, we selected the nearest STR marker to the genes PHF11 (Zhang et al. 2003) and ADAM33 (van Eerdewegh et al. 2002), respectively.

Markers on chromosome 8q12 (D8S1828, D8S285, and D8S1816) were chosen on the basis of our previous whole genome search for IgE-controlling loci in mouse (Badalová et al. 2002). The mouse locus Lmr9 (represented by marker D4Mit149) was mapped to the centromeric part of the mouse chromosome 4 with the most likely and maximal lengths 3.58 and 9.32 Mb, respectively, and was shown to have linkage with IgE level (Badalová et al. 2002). The region homologous to the mouse Lmr9 is located on human chromosome 8q12 (NCBI database), and the markers D8S1828, D8S285, and D8S1816 were chosen for the search for IgE-controlling loci in human. There is no LD between the regions carrying the three STRs (UCSC Genome Browser Assembly March 2006).

Genotyping

The primer sequences were obtained from the NCBI database. We used Cy5 carbocyanine dye 5′-end-labeled forward primers and unlabeled reverse primers synthesized by Generi-Biotech s.r.o. (Hradec Králové, Czech Republic) or Sigma-Genosys, (Steinheim, Germany). DNA was amplified in a 10-μl polymerase chain reaction (PCR) reaction with 10 pmol/μl of forward and reverse primer, 0.2 mM concentration of each dNTP, 1.5 or 2.0 mM MgCl2 (optimized for each STR), 50 mM KCl, 20 mM Tris–HCl (pH 8.4), and 0.1 U of Taq polymerase, recombinant (GIBCO, Grand Island, NY, USA) and 5 ng/μl of template DNA. The PCR reaction was performed on 0.2 ml non-skirted 96-well U-bottom microtiter plate (ABgene, Epsom, UK) by MJ Research Thermal Cycler PTC 100 model 96 (MJ Research, Watertown, MA, USA). The universal program was used for DNA amplification: an initial hot start 5 min at 94°C, followed by 39 cycles of 94°C for 20 s for denaturing, 55°C for 20 s for annealing, 74°C for 20 s for elongation, and finally 10 min at 72°C for final extension. PCR products (0.25 μl) were separated by CEQ™ 8800 Genetic Analysis System (Beckman Coulter, Fullerton, CA, USA). All inconclusive genotypes were excluded (less than 2.2% for each marker).

Statistical analysis

The statistical analysis included all family members (also probands) regardless of affected status. Two different approaches were used for statistical analysis of data. The first approach included non-parametric linkage analysis for co-segregation of a chromosomal region and a trait of interest (qualitative and quantitative). The analysis is based on the calculation of LOD score using the linear model of Kong and Cox (1997). This method allows using small nuclear families and calculation of linkage without assuming the normal distribution of the studied trait. We used the Whittemore and Halpern NPL pair statistics (Whittemore and Halpern 1994) to test for allele sharing among affected individuals. The computer program MERLIN version 1.0.0-© 2000–2005 (Abecasis et al. 2002) was used for the calculation of identical-by-descent, allele frequencies (across all individuals), and LOD scores.

The second approach used association analysis for QTDT. The general model of QTDT described by Abecasis et al. (2000a, b) is applicable to the analysis of quantitative or discrete traits in nuclear families of any size and optionally uses parental phenotypes. We used the orthogonal model (Abecasis et al. 2002) to perform the association analysis of the markers as well as their allelic variants with total and specific IgEs. Calculation was conducted by QTDT program version 2.4.6-© 1998–2004 (Abecasis et al. 2000b). Permutation framework (100,000 permutations) provided by QTDT program was used to obtain global P values. These were subsequently corrected for multiple testing by Bonferroni correction for number of tested markers and numbers of alleles of the tested markers.

Sex and age were chosen as covariates in both non-parametric linkage and linkage disequilibrium analysis of the total and specific IgEs and inhalant and food atopy. Correlation between phenotypes (sensitization to different allergens) was estimated by the nonparametric Spearman’s correlation analysis using STATISTICA for Windows version 8.0 (StatSoft 1984–2008, Tulsa, OK, USA).

Results

8q12 is the human genetic homologue of the IgE-controling mouse locus Lmr9

Marker D8S285, located in the human homologue of Lmr9, showed a suggestive linkage with IgE to P. lanceolata (w9) allergen (LOD = 2.42, P = 0.0004) and with a composite phenotype–inhalant allergic sensitization (LOD = 2.11, P = 0.0009; Table 3). The locus 8q12 has not been previously reported in connection with atopy in humans. Potential linkage with LOD > 1 was also suggested to specific IgE against moulds (A. fumigatus—m3 and C. herbatum—m2), animal origin allergens (dog—e2 and cat—e1), mites (D. farinae—d2 and D. pteronyssinus—d1), A. elatior (w1), and potato (f35) allergens (Table 3). The association QTDT analysis of the D8S285 with IgE to P. lanceolata (w9) revealed association with the marker (P = 0.0317) and with the certain alleles of the marker (allele 112 bp P = 0.0056 and allele 114 bp P = 0.0184), supporting the linkage and suggesting that a gene controlling atopy in humans is localized close to D8S285. However, we did not find the significant association after adjusting for number of comparisons by Bonferroni correction.

Table 3 Specific IgE-controlling loci in the Czech atopic families
Full size table

We found also some evidence of linkage to the markers D8S1828 and D8S116 on chromosome 8q12 that are flanking D8S285. D8S1828 that is located 260 Kb centromerically from the marker D8S285 exhibited potential linkage with alder allergens (A. incana—t2; LOD = 1.16, P = 0.011) and with allergen of cultivated rye (S. cereale—g12; LOD = 1.11, P = 0.012). The marker D8S116, located 300 Kb telomerically from the marker D8S285, showed a weak linkage with allergens of cultivated rye (LOD = 0.68, P = 0.04).

Testing of previously reported atopy-controlling regions in the Czech population

We also tested STR markers at the human chromosomal regions that were previously described in genome-wide studies to control atopy in order to determine whether these genetic loci influence IgE level in the Czech population (Table 1).

Two loci on chromosomes 13q14 and 5q33.3 showed the strongest linkages with IgE to plant allergens (Table 3). We found a suggestive linkage of marker D13S165 (Zhang et al. 2003) with IgE to A. elatior (w1) allergen (LOD = 2.74, P = 0.0002). We also detected a suggestive linkage for marker D5S1507 (Yokouchi et al. 2000) with IgE to S. cereale (g12; LOD = 2.11, P = 0.0009). Finally, markers on chromosomes 7p14.1 (D7S2250; Daniels et al. 1996; Laitinen et al. 2001; Shugart et al. 2001) and 12q13 (D12S1298; Barnes et al. 1996) were found by QTDT assay to be significantly associated with P. lanceolata (w9; marker D7S2250, corrected P = 0.026; allele 147bp of marker D7S2250, corrected P = 0.034) and A. vulgaris (w6; allele 199bp of marker D12S1298, corrected P = 0.043), respectively.

Thus, two markers (D13S165 and D5D1507) were found to have suggestive linkages, and markers D7S2250 and D12S1298 were significantly associated with a number of specific IgEs (Table 3). These data support the results that have been previously published by others (see references in Table 1 and in “Introduction”).

There was also some evidence for linkage (LOD > 1) with a number of inhalant and food allergens to the markers D5S816 (5q31.1), D12S1059 (12q22), D16S3253 (16q21), and D20S473 (20p13; Table 3). We did not find any association with and any linkage exceeding the level of a LOD > 1 to the markers D2S308 (2q14.3), D6S291 (6p21), D7S821 (7q22.1), D11S2006 (11q12), D12S379 (12q21.31), D12S1282 (12q24.31), D16S539 (16q23.2), and D19S601 (19q13.32).

No significant linkage was found with asthma, rhinitis, dermatitis, and total IgE. Markers D19S601 (LOD = 0.70, P = 0.04), D16S539 (LOD = 0.69, P = 0.04), D7S2250 (LOD = 0.83, P = 0.03), and D8S285 (LOD = 0.38, P = 0.09) showed the highest LOD score with asthma, rhinitis, dermatitis, and total IgE, respectively.

Discussion

Although several genome-wide linkage studies of IgE-controlling loci in humans were conducted, our data for the first time indicate a locus on chromosome 8q12 that could influence development of atopy. The finding of this linkage shows the precision and predictive power of mouse models in investigation of the complex traits in humans.

There are no obvious candidate genes in 8q12 chromosomal region. In the near proximity of the marker D8S285 are localized two oncogenes: MOS (V-MOS Moloney murine sarcoma viral oncogene homolog) and PLAG1 (Pleiomorphic adenoma gene 1). MOS exerts many cellular functions; however, its described impact on B cell functions is limited to B cell malignancies caused by chromosomal translocations of this chromosomal segment (Kirsch et al. 1982). PLAG1 encodes a developmentally regulated, SUMOylated, and phosphorylated zinc finger transcription factor which recognizes a specific bipartite DNA consensus sequence regulating expression of a spectrum of target genes (Van Dyck et al. 2007). One of the target genes of PLAG1 is insulin-like growth factor-2 (IGF2; Van Dyck et al. 2007), which has pleiotropic functions in immunity. It was shown that Igf2−/− mice had decreased numbers of B220+ dendritic cells in spleen (Hansenne et al. 2006). Adoptive transfer of dendritic B220− cells from allergic mice induces specific immunoglobulin E antibody against food allergens in naïve recipients (Chambers et al. 2004), thus showing possible pathway how could PLAG1 influence atopy.

Another promising target for future research of this locus is LYN kinase gene (LYN), which is mapped near marker D8S1828 and 150 Kb proximally of the marker D8S285. The Src tyrosine kinase Lyn is an important modulator in the high affinity receptor for IgE (FcɛRI) signaling (reviewed in Rivera and Olivera 2007). Lyn-deficient mice exhibit increased serum levels of IgE, increased numbers of mast cells, increased expression of FcɛRI on mast cells, and other allergy-associated traits (Odom et al. 2004). Although the etiology of the allergy-like phenotype of Lyn deficiency is not completely understood, the regulatory role of Lyn kinase in the development of allergy is strongly indicated by these results. Further, dense SNP coverage, fine mapping, expression studies, and re-sequencing of the 8q12 region are required to define the gene(s) affecting disease susceptibility.

In pooled groups of Caucasian families recruited from Minnesota and from 11 clinical centers in Europe, Australia, and USA, a suggestive linkage was detected with total serum IgE with a peak of linkage near the marker D8S2324 (94.08 cM—Marshfield; Webb et al. 2007). This marker is located outside the <9 Mb region homologous to mouse IgE-controlling locus Lmr9 (Fig. 1), and also 23 cM distally, and therefore is distinct from the locus at 8q12 described here. These data suggest that two loci controlling human IgE might be localized on chromosome 8q.

Linkage of locus 13q14 with atopy was initially detected in atopic families from Busselton, Western Australia (Daniels et al. 1996). Subsequent association mapping using dense SNP map of this region postulated PDH finger protein 11 gene (PHF11) as a gene predisposing to atopy (Zhang et al. 2003). In our study, the marker D13S165 that maps within 1 Kb from PHF11 gene is linked to IgE to A. elatior (w1; Table 3), suggesting the effect of this gene on IgE regulation in the Czech atopic patients.

Chromosomal region 5q31-33 seems to be one of the most attractive targets for the investigation of the IgE control in humans. This region encompasses a cluster of pro-inflammatory cytokine genes [IL-4, IL-5, IL-9, IL-13, IRF-1 (interferon releasing factor 1) and CSF-1R (receptor for colony stimulating factor 1], the protein products of which are directly involved in immune regulation. The linkage to this locus was corroborated in multiple studies (Marsh et al. 1994; Ober et al. 2000; Xu et al. 2000; Haagerup et al. 2002; Koppelman et al. 2002; Yokouchi et al. 2002). This locus also contains genes TIM1/HAVCR/KIM1 (5q33.2) and TIM3 (5q33.3) that were shown to control asthma and airway hyperreactivity (McIntire et al. 2001). In the present study, a marker on chromosome 5q33.3 (D5S1507) had a suggestive linkage with IgE to allergens of S. cereale (g12) and showed a potential linkage to allergens of Phleum pretense (g6), D. glomerata (g3), and A. odoratum (g1; Table 3). All grass sensitization phenotypes (g12, g6, g3, and g1) showed very high positive correlation with each other (R = 0.937–0.988) and therefore showed very similar LOD scores with D5S1507.

QTDT analysis also revealed two atopy-associated markers on chromosomes 7p14.1 (D7S2250) and 12q13 (D12S1298) that did not show genetic linkage with atopy phenotype. Linkage and QTDT association analysis [Kong and Cox linear model (Kong and Cox 1997) and Abecasis orthogonal model (Abecasis et al. 2000a, b), respectively] exploit two different aspects of genetic information. Thus, the results of association and linkage analysis may, but need not, coincide (Wills-Karp and Ewart 2004).

Remarkably, none of the loci showed a significant linkage or association with total IgE. This might be partly due to a high overall sensitization of Czech atopic families that reached nearly 90% in both parents and offspring. However, we observed that various types of the specific IgE were controlled by different genetic loci. Thus, sensitization to different allergens seems to be determined by different genes. This might also partly explain differences in results obtained by different laboratories that postulated loci controlling total IgE in humans (Hoffjan and Ober 2002).

We did not find any linkage or association with asthma, rhinitis, and dermatitis. This is probably due to the low number of affected sib pairs with asthma and dermatitis in the tested sample (please see details in “Materials and methods”, Table 2). Moreover, asthma (Hoffjan and Ober 2002), and probably also rhinitis and dermatitis, may comprise groups of several disorders. In the mouse model, the various components of the pathogenetic pathway of allergic asthma are under separate genetic control (Piavaux et al. 2007); this may explain why in this and other similar studies, asthma is not necessarily linked to the genes that control one out of the multiple sets of pathogenetic components. The analysis of less complex and more exactly defined phenotypes such as levels of total and specific IgE is therefore an important part of genetics of atopic diseases.

The present work demonstrates the power of the genome-wide screening in mice in finding new loci determining complex traits such as IgE levels in humans. Using this approach, a new IgE-controlling locus has been identified on chromosome 8q12 that influences the sensitization to a number of allergens. Our data also confirm the role of the previously reported loci 13q14, 5q33.3, 7p14.1, and 12q13 in control of IgE and development of atopy.