INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Interleukin-10 (IL-10) and transforming growth factor beta (TGF-) are anti-inflammatory cytokines that are constitutively expressed in the healthy airway (1). Both IL-10 and TGF- downregulate cellular immunity and allergic inflammation. Diminished secretion of these cytokines could contribute to the development of a milieu in which allergic inflammation and asthma can develop. In contrast to their anti-inflammatory effects, the appearance of IL-10 and TGF- in ongoing allergic inflammation may contribute to disease severity by stimulating IgE synthesis (5) and fibrosis (3, 4), respectively. Polymorphisms in promoter sequences of genes result in abnormal transcriptional regulation and thereby influence the development or severity of disease. We investigated the presence and linkage of polymorphic nucleotides within the IL-10 and TGF- gene promoters to the expression of allergic and asthmatic phenotype. We detected base substitutions as single-stranded conformational polymorphisms (SSCPs). We screened each polymorphism by case-control analysis for linkage to allergy and asthma phenotypes using our data base containing 20 atopic asthmatic and 10 nonasthmatic families.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Subjects were recruited from the asthma risk study which has been performed at the National Jewish Medical and Research Center (6, 7). These families were identified from the presence of a pregnant mother with asthma, and the proband is the child who has been followed since birth (ages 5 to 10 at the time these studies were performed). These were all nuclear families in which we studied both parents, the proband, and siblings. The number of siblings ranged from 1 to 4, although we selected for larger families and most had 2 to 3 siblings. Our phenotyping studies involved the clinical diagnosis of allergic rhinitis or asthma based on medical history and physical examination. Objective parameters included performing pulmonary function tests before and after administration of a bronchodilator, methacholine challenges, skin testing to the Colorado allergen panel, and obtaining blood specimens for total IgE and total eosinophil counts. Specific parameters which we analyzed for these studies included: baseline percent predicted FEV₁, percent reversibility of the FEV₁ after administration of albuterol, log concentration of methacholine producing 20% reduction in the FEV₁ (PD₂₀), the presence of any positive prick skin test, the number of positive skin tests, total IgE, and eosinophil count. Twenty of these families were phenotyped and underwent phlebotomy for DNA extraction using a commercial DNA extraction kit (Gentra Systems, Inc., Research Triangle Park, NC). An additional 10 families were recruited in whom no family member had either allergies or asthma. All patients including children provided their informed consent for this study and the protocol was approved by the investigational review boards of the National Jewish Medical and Research Center and the University of Colorado Health Sciences Center.

Single-Stranded Conformational Polymorphisms (SSCP)

The SSCP technology takes advantage of the different secondary structure and therefore electrophoretic mobility induced in single-stranded DNA by a substitution in as few as a single nucleotide (8, 9). SSCP was performed on the upstream regulatory region of IL-10, TGF-1, TGF-2, and TGF-3. Our approach was to investigate the first 1 kB 5' from the transcription initiation site (TIS) as most base substitutions likely to have important effects on transcriptional regulation would be expected to lie within this region. Promoter/enhancer sequences were analyzed in fragments each ~ 250 bp in length, using primer sequences synthesized on an Applied Biosystems 381A DNA synthesizer (Albany, CA). To perform the SSCP analyses, polymerase chain reaction (PCR) was performed on genomic DNA (1 µg) and internally labeled with [³³]P-dATP. PCR product was diluted in 0.1% sodium dodecyl sulfate (SDS)/10 mM ethylene diaminetetraacetic acid (EDTA) and mixed with formamide dye (95%). The double-stranded DNA (dsDNA) was denatured at 95° C for 5 min, rapidly chilled, and the denatured DNA was loaded on a 4.5% nondenaturing acrylamide gel under four conditions: 4° C and 22° C with and without 5% glycerol. Electrophoresis was performed with constant monitoring of temperature to minimize heating during the electrophoresis. After electrophoresis was completed, the gels were dried and autoradiography performed.

Heteroduplex Analysis

The four combinations of conditions under which SSCP is performed will detect ~ 80 to 90% of all contained mutations. This can be improved to virtually 100% through the additional analysis of PCR products via heteroduplex analysis of double-stranded molecules (10). This is based on the concept that when PCR is performed on two alleles, during the denaturing and reannealing process dsDNA is formed from both identical complementary strands (homoduplexes) but also from the annealing of complementary strands from the two different amplified segments (heteroduplexes). These heteroduplexes may be detected as a mobility shift using the same acrylamide electrophoresis as performed for the SSCP.

DNA Sequencing of Polymorphic Regions

Polymorphic DNA was extracted from the acrylamide, PCR-amplified using the original primers used in the SSCP analysis, and purified from agarose (Qiagen, Inc., Valencia, CA). Automated DNA sequencing was performed using fluorescent dideoxy sequence terminators (Applied Biosystems). Primers internally nested from those used for the SSCP were synthesized. Sequencing reactions were performed in both 5' and 3' directions both to provide corroborative data and to read potentially poorly discernible areas on the sequencer readout. Base exchanges were characterized by Pustell matrix analysis performed with MacVector software.

Association Studies

We performed association analyses to determine influences of the polymorphisms on phenotype. Association studies cannot be performed on related family members. While we screened for the presence of polymorphisms in 144 related individuals derived from 30 families, these association studies were only performed on the 30 unrelated probands from each family and 17 additional probands with and without asthma recruited from the asthma risk study. The presence of a significant linkage to asthma/allergy phenotype supports the hypothesis that these mutations influence the risk for developing these conditions. Association studies compare random samples of unrelated subjects and controls to assess the relative frequencies at which a given allele appears in these two populations. This is the most sensitive indicator of genetic linkages and can generate significant data with fewer individuals. For parametric data with normal distribution (including FEV₁, FEV₁/FVC ratio, percent change in FEV₁ after administration of bronchodilator, and total eosinophil count), statistical analyses were performed using analysis of variance (ANOVA). For the statistical analysis of total serum IgE, data were first converted to log IgE to produce a normal distribution and then analyzed via ANOVA. For nonparametric data (presence of an atopic disease, clinical diagnosis of asthma, presence of positive skin tests) and for parametric data with a non-normal distribution (PD₂₀) statistical analyses were performed utilizing chi-square analyses. These studies were performed to demonstrate the disproportional presence of a given allele in affected individuals. Statistics were performed on a Macintosh computer using JMP 3.1 software.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

SSCP

Our laboratory cloned and sequenced the 5' flanking region of the IL-10 gene to assess the role of its promoter in the development of human allergic diseases. Our approach to sequencing the IL-10 promoter exploited the human IL-10 complementary DNA (cDNA) sequence to perform three cycles of inverse PCR (GenBank accession #: UO6844). The 5' flanking sequences of TGF-1, -2, and -3 were obtained from GenBank. We identified base substitutions as the detection of SSCPs. SSCP demonstrated one polymorphism within the first 1000 bp of the IL-10 open reading frame (ORF) (Figure 1). This polymorphism is located at bp 571 (from the translation start site) and represents a C-to-A base substitution. It is located between consensus sequences for DNA binding by two types of transcription factors: members of the ets family and Sp1 (Figure 2). In our population the polymorphism was present in heterozygous form in 34 of 144 and homozygous form in 13 of 144 subjects (Table 1). Four polymorphisms were identified in the TGF- promoters. These include a base exchange at bp 509 as well as two between 738 and 1038 bp from the first major transcription initiation site of TGF-1 and between ⁺13 and 297 of TGF-2. No polymorphisms were identified in TGF-3. The polymorphism at bp 509 represents a C-to-T nucleotide exchange which is located in an area thought to be a negative regulatory region (11) and creates a YY1 activator consensus sequence (12) (Figure 2). The polymorphism is present in heterozygous form in 77 of 144 subjects and in homozygous form in 16 subjects (Table 1). The two additional polymorphisms which were also identified in the TGF-1 gene were not further characterized or sequenced as they did not associate with any of the phenotypic markers of allergy and asthma. The polymorphism in the TGF-2 promoter is located in a region which may be important in influencing promoter strength as it includes the TATAA box; however, we did not identify any associations with allergy and asthma phenotypes in our kindreds. No additional polymorphisms were identified via heteroduplex analysis.

View larger version (57K): [in this window] [in a new window]   Figure 1.   Single-stranded conformational polymorphisms. Genomic DNA underwent PCR in the presence of primers specific for approximately 250 base pairs of the IL-10 or TGF- promoters, in the additional presence of [33]P-dATP. The dsDNA was heat- denatured and electrophoresed on a 4.5% nondenaturing acrylamide gel at 22° C in the absence of 5% glycerol. (Upper panel  ) SSCP for the IL-10 promoter. One family is shown with one parent homozygous wild type (lane a), one parent homozygous for the base exchange (lane d ), and the two heterozygotic children (lanes b and c). (Lower panel  ) SSCP for the TGF-1 promoter. Unrelated subjects are displayed who demonstrated the patterns for homozygote wild type (lane a), heterozygotes (lanes b-e), and homozygotic for the base exchange (lane f  ). In contrast to the pattern observed for IL-10, the base exchange in the complementary strand of DNA does not produce a conformational change sufficient to produce a motility shift.

View larger version (24K): [in this window] [in a new window]   Figure 2.   Sequence of the IL-10 and TGF-1 polymorphisms. Wild-type and base exchange sequences are shown. The IL-10 polymorphism is present at bp 571 from the open reading frame and represents a C-to-A exchange. It is present between consensus sequences for Sp1 binding and binding by members of the ets family. The TGF-1 polymorphism is present at bp 509 from the transcription initiation site and represents a C-to-T exchange. The presence of this exchange creates a YY1 activation site.

Association Studies

Support for a phenotypic influence for the IL-10 base exchange at bp 571 in asthma and allergy is provided by data derived from our association studies (Table 2). For IL-10, the most significant linkage was observed to total IgE (p < 0.009) with a smaller degree of significance related to the clinical diagnosis of the presence of an atopic disease (allergic rhinitis, asthma, or atopic dermatitis; p = 0.05). The mean level of IgE for subjects homozygous for the polymorphism is 350.0 ± 145.1 as compared with 66.4 ± 56.9 for the homozygous wild types and 322.3 ± 68.4 for heterozygotes. These IgE data suggest a dominant influence of the base exchange. No significant linkages were observed on any of the phenotypic parameters by the TGF-1-1 base exchange with the exception of total IgE (p < 0.01) (Table 2). In contrast to the IL-10 promoter polymorphism, these data suggest a recessive influence. Finally, we assessed the additive influences of each of these markers on total IgE (Table 2). A synergistic influence for the presence of the TGF-1-1 base exchange upon the IL-10 polymorphism is indicated (p < 0.0001). We could not independently assess the influence of the IL-10 exchange on TGF-1 insofar as all of the subjects with the TGF-1 exchange were also polymorphic at the IL-10 promoter locus.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The absence of respiratory inflammation in normal, nonatopic, nonasthmatic subjects is maintained by influences which promote the development of nonresponsiveness toward otherwise benign inhaled bioaerosols. The cytokine milieu of the healthy respiratory tract contributes to this immune nonresponsiveness through the constitutive expression of the anti-inflammatory cytokines IL-10 (1, 2) and TGF- (3, 4). The role of IL-10 as an anti-inflammatory cytokine in allergic inflammation has recently been reviewed (13). IL-10 downregulates IL-4 and IL-5 expression by type 2 T-helper cell (Th2) lymphocytes (14). IL-10 also inhibits accessory cell function (15) and in the presence of IL-10, not only do T lymphocytes not proliferate and produce cytokines in response to allergen, but those cells are rendered irreversibly nonresponsive (tolerant) (16, 17). Support for a modulating role for IL-10 in human allergic diseases is further derived from observations that IL-10 inhibits eosinophil survival (18) and IgE synthesis (19). Administration of IL-10 to sensitized mice abrogates allergen-induced airway inflammation (20). IL-10 knockout mice develop a lethal form of allergic bronchopulmonary aspergillosis with markedly elevated concentrations of interferon gamma (IFN-), IL-4, and IL-5 (21). In contrast to the healthy lung (1, 22), asthma (2) is characterized by diminished IL-10 production, which will contribute to the development of an inflammatory milieu. The inhibitory effects of IL-10 are in contrast to its effect on B lymphocytes where it stimulates proliferation and immunoglobulin secretion (23, 24). The appearance of IL-10 after an allergic immune response has been allowed to develop may contribute to the severity of these responses by stimulating IgE secretion from B lymphocytes that have previously undergone the  isotype switch (5). This potential role for IL-10 in exacerbating established allergic disorders is consistent with the demonstration of enhanced IL-10 secretion in atopic dermatitis (25) and our observations demonstrating IL-10 transcription after an allergen challenge (2).

Similar to IL-10, TGF- mediates a complex mix of pro- and anti-inflammatory activities. TGF- represents a closely related family of peptides of which TGF--1, -2, and -3 are expressed in humans (3). In general, TGF- is an important stimulant of fibrosis, inducing formation of the extracellular matrix and fibrosis. In immunity, however, its predominant functions are anti-inflammatory. TGF-, like IL-10, indirectly inhibits T-cell activation by modulation of antigen-presenting cell function and deactivating macrophages (3, 26). TGF- like IL-10is constitutively expressed in the lungs (4, 27). TGF- knockout mice do not survive secondary to diffuse mononuclear cell infiltration of numerous organs including the lungs (28). TGF- may help prevent the development of allergic inflammation through a capacity to inhibit IgE synthesis and through inhibition of mast cell proliferation. Additionally, TGF- abrogates the survival effects of hematopoietins on eosinophils and thereby induces their apoptosis (29). In contrast to these observations supporting an anti-inflammatory role for TGF-, eosinophils taken from the nasal tissue of subjects with allergic rhinitis as well as from nasal polyposis tissue express increased amounts of TGF-1 gene (30). The secretion of TGF- after an allergic disorder has developed may contribute to fibrosis and the irreversible changes which occur with long-standing asthma (4).

These data suggest a model wherein the healthy lung maintains an appropriate degree of nonresponsiveness to allergens in part through the presence of TGF- and IL-10. Influences that act to diminish production of IL-10 or TGF- will enhance any underlying proinflammatory condition. Asthma will only develop in the additional presence of influences promoting respiratory inflammation and immune responses against allergens. Both IL-10 and TGF- may also positively contribute to the severity and long-term complications of established inflammatory disorders including asthma.

Abnormal regulatory elements within the IL-10 and TGF- genes may contribute to the development of allergies and asthma (or other inflammatory disorders). In the present studies we utilized SSCP to demonstrate the existence of one polymorphism in the IL-10 promoter, three in the TGF-1 promoter, and one in the TGF-2 promoter. Since our initial report of the IL-10 promoter polymorphism (31), this same base exchange has been reported by others (32). While these were fairly frequent base exchanges in our cohort (Table 1), our subjects were recruited on the basis of the presence of an asthmatic proband. Thus, any base exchange that may contribute to the development of allergies and asthma will be overrepresented, and as such, these data do not accurately report the frequency of our polymorphisms in the general population. The frequency distributions of these alleles, while demonstrating an approximation of Hardy-Weinberg distribution, were not in equilibrium for IL-10, TGF-1 and -3. In addition to aberrations induced by the ascertainment of families on the basis of an asthmatic proband, this may also have occurred secondary to the inclusion of related family members. The small size of our cohort may have produced sampling error. Importantly, the data presented by Turner and coworkers (32) demonstrate Hardy-Weinberg distribution for the IL-10 promoter base exchange. As such, it is more likely that sampling issues explain the absence of expected distribution frequencies, rather than migration of these mutations into the population.

The strongest phenotypic association of our polymorphisms was with elevations in total IgE (Table 2) with both the IL-10 and TGF-1 polymorphism, although there was a less impressive linkage to the presence of one of the atopic diseases (allergic rhinitis, asthma, or atopic eczema) with the IL-10 base exchange. Interestingly, synergistic influences were observed with the two polymorphisms on IgE production. Because TGF-1 and IL-10 genes are on different chromosomes (19q and 1q, respectively), these genes clearly will segregate independently. TGF- may inhibit the isotype switch to IgE (33) and as such base exchanges that increase TGF- expression may be predicted to affect total IgE concentrations. As discussed, IL-10 may positively or negatively influence IgE by either inhibiting the  isotype switch through inhibition of IL-4 and IL-13 (13, 19) or by stimulating IgE release from committed B lymphocytes (5), respectively. Our demonstrated association between total IgE and the IL-10 polymorphism therefore cannot predict whether this polymorphism acts to promote or inhibit IL-10 production. No associations were identified between the two further upstream polymorphisms in the TGF-1 promoter nor the base exchange in the TGF-2 promoter with any of our phenotypic markers of allergy and asthma. This does not preclude the possible importance in other kindreds or in other immune and inflammatory disorders. Unfortunately, our sample size was too small and therefore may have lacked the power to detect additional significant associations. It is to be hoped that other groups with larger cohorts of phenotyped individuals can confirm the importance of these linkages and extend them to other phenotypic markers of allergies and asthma. Confirmation of the IgE associations in unrelated cohorts, in fact, is essential in studying the genetics of complex disorders.

The base substitutions we have identified may not have functional significance for the development of allergies or asthma. Our data supporting an association of the polymorphisms with IgE dysregulation, along with the observations derived from the IL-10 and TGF- knockout mice, support the ability of dysregulated transcription of these cytokines to play a causative role in the development of allergic inflammation. In addition, the recent report of a linkage of the base exchange in the IL-10 promoter with severity of systemic lupus erythematosus supports the concept that this polymorphism contributes to the presence or severity of immune disorders (34). There are, however, several alternative explanations. Linkage disequilibrium may be present with the actual responsible polymorphism. More problematic are difficulties with our use of association studies. One argument for performing association analysis is that our specific aim was to identify asthma and allergy "genes" and not genetic markers that link to these conditions. The difficulty with association studies is that they cannot distinguish true linkages from identical patterns which can occur in the presence of population stratification in the distribution of a given allele. Such population heterogeneity has plagued the asthma genetics literature and is responsible for the myriad of studies that have identified significant linkages in one population which are not reproducible by other investigators. In the absence of a surprisingly large influence on phenotype, association studies are the primary mechanism for screening for disease linkages in complex disorders such as asthma. While population stratification may have produced false-positive results in the current study, this will be addressed in part by other asthma geneticists evaluating the role of our polymorphisms in their own unrelated populations. Most importantly, the genetic role for these polymorphisms in asthma, allergy, and other immune diseases will be established by performing functional molecular biological studies. Polymorphisms that influence transcription factor interaction with the gene promoters, modify promoter strength, and thereby modulate gene expression will be established as having phenotypic influences. For example, our current studies have demonstrated an increased binding affinity of Sp1 and one other uncharacterized transcription factor to IL-10 promoters containing C at bp 571, and have shown that this is associated with increased promoter strength (Borish and colleagues, submitted manuscript). Individuals with this form of the IL-10 promoter may therefore demonstrate enhanced synthesis of IL-10, thereby influencing the intensity of their immune responses and producing the associations observed in these studies.