House dust mites as potential carriers for IgE sensitization to bacterial antigens

Abstract Background IgE reactivity to antigens from Gram‐positive and Gram‐negative bacteria is common in patients suffering from respiratory and skin manifestations of allergy, but the routes and mechanisms of sensitization are not fully understood. The analysis of the genome, transcriptome and microbiome of house dust mites (HDM) has shown that Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) species are abundant bacteria within the HDM microbiome. Therefore, our aim was to investigate whether HDM are carriers of bacterial antigens leading to IgE sensitization in patients suffering from atopic dermatitis. Methods Plasma samples from patients with AD (n = 179) were analysed for IgE reactivity to a comprehensive panel of microarrayed HDM allergen molecules and to S. aureus and E. coli by IgE immunoblotting. Antibodies specific for S. aureus and E. coli antigens were tested for reactivity to nitrocellulose‐blotted extract from purified HDM bodies, and the IgE‐reactive antigens were detected by IgE immunoblot inhibition experiments. IgE antibodies directed to bacterial antigens in HDM were quantified by IgE ImmunoCAP™ inhibition experiments. Results IgE reactivity to bacterial antigens was significantly more frequent in patients with AD sensitized to HDM than in AD patients without HDM sensitization. S. aureus and E. coli antigens were detected in immune‐blotted HDM extract, and the presence of IgE‐reactive antigens in HDM was demonstrated by qualitative and quantitative IgE inhibition experiments. Conclusion House dust mites (HDM) may serve as carriers of bacteria responsible for the induction of IgE sensitization to microbial antigens.


| INTRODUCTION
Already in 1932, Cooke suggested that allergic sensitization to bacteria may play a role in asthma. 1 Since then, IgE sensitization to bacterial antigens has been reported for different manifestations of allergy. Several studies have observed IgE sensitization to bacterial antigens in children suffering from asthma. [2][3][4] In particular, IgE antibodies directed against enterotoxins from Staphylococcus aureus (S. aureus) were found to be associated with severe forms of asthma 5,6 and recently Staphylococcal serine protease-like proteins were suggested as pacemakers of allergic airway reactions to S. aureus. 7 The occurrence of IgE to bacterial antigens was also observed in atopic dermatitis (AD) patients as early as 1981. [8][9][10][11][12] Anti-Staphylococcal IgE has been found in atopic dermatitis patients with or without cutaneous Staphylococcal infection, [8][9][10][11][12] and IgE reactivity against various antigens such as Staphylococcal exotoxins 13 and S. aureus fibronectin-binding protein 14 has been reported. In addition to IgE reactivity against Staphylococcal antigens, IgE reactivity to Escherichia coli (E. coli) antigens and other bacteria of the gut has been detected in about one-third of patients with AD. 15 These findings are unexpected, considering that S. aureus typically induces Th1 or Th17 rather than Th2 immune responses, and nonpathogenic E. coli is associated with immune tolerance rather than with allergic sensitization. 16,17 Staphylococcus aureus (S. aureus) is present in the skin of patients with AD who frequently suffer from S. aureus superinfections, and it has therefore been suggested that sensitization occurs via the skin. 17 By contrast, E. coli and other bacteria (eg, Haemophilus influenza) which have been reported as sources for IgE-reactive antigens 3,4,15 can be found mainly in the gut and in the respiratory tract, suggesting that other routes and mechanisms of allergic sensitization may be important. Recently, the genome, transcriptome and microbiome of house dust mites (HDM) one of the most frequent and potent allergen sources have been reported. 18 More than 50% of atopic individuals have allergic sensitization to HDM, particularly in environments with favourable conditions for mite proliferation, [19][20][21] and sensitization to HDM is common in patients suffering from respiratory and skin manifestations. 20,22,23 Interestingly, the microbiome of Dermatophagoides farinae comprises a variety of Gram-positive and Gram-negative bacteria, among them S. aureus, E. coli and Enterobacteria. 18 The description of the HDM microbiome has contributed to a growing body of evidence highlighting that HDM not only contain allergens but also other components such as bacterial adjuvant compounds which can activate the innate immune system. 24,25 Here, we hypothesized that HDM may have another hitherto unknown role in allergic sensitization. In fact, we investigated whether HDM are carriers of microbial antigens that could cause a Th2 type adaptive immune response to bacterial antigens. To study this hypothesis, several lines of investigations were pursued. First, we studied the occurrence of IgE sensitization to HDM and bacterial antigens from S. aureus and E. coli, in a group of patients with AD from Sweden where sensitization to HDM is expected to be less common, as a result of the cold climate. 26 This allowed us to evaluate whether sensitization to bacterial antigens is more frequent among HDM-sensitized patients. In addition, we investigated the presence of bacterial antigens in extracts prepared from purified HDM bodies. Finally, we were able to demonstrate that HDM contain IgE-reactive bacterial antigens using sera from patients sensi-  28 where they were examined by a dermatologist. Where applicable, the severity of atopic skin disease was determined by the SCORAD and graded as moderate or severe, depending on the extent and intensity of the inflammatory lesions. 29,30 The Swedish AD patients were frequently sensitized to several environmental allergens such as allergens from cat (55%), grass (54%), birch (52%) and dog (37%) as determined using microarrayed allergens. 28 Data on atopic conditions such as asthma and rhinitis were collected by anamnesis (medical history). The study was approved by the regional ethical review board in Stockholm, and written informed consent was obtained from all participants.
For IgE immunoblotting and IgE inhibition experiments requiring large volumes of serum, we included additional samples from wellcharacterized allergic patients (A1-A6) (Table S1) with known IgE reactivity to bacterial antigens. 15   In addition, a rabbit antiserum capable of detecting over 300 low to high molecular weight proteins from E. coli was obtained from Rockland Immunochemicals (Limerick, PA, USA). For purposes of this work, the anti-E. coli antiserum raised from rabbit immunized with E. coli extract is referred to as anti-E. coli.1, and the purchased anti-E. coli antiserum is referred to as anti-E. coli.2. Preimmune serum was not available for the (commercial) anti-E. coli.2.

| Detection and measurement of specific IgE and IgG antibodies
IgG and IgE antibodies specific for S. aureus and E. coli were detected by immunoblotting. 15 Bacterial extracts were added to Laemmli SDS sample buffer, 27 boiled for 5 minutes and separated by SDS-PAGE on a 12.5% SDS polyacrylamide gel. 37 As HDM extracts were already prepared in Laemmli SDS sample buffer, the HDM protein extracts were loaded directly on 12.5% SDS polyacrylamide gels for electrophoretic separation. Aliquots of 50 lg, of each of the extracts were loaded per centimetre of a preparative gel, and a prestained protein ladder (PageRuler TM Plus, Thermo Scientific) was used as a molecular weight marker. The separated proteins were transferred to a nitrocellulose membrane by electroblotting in transfer buffer (25 mmol/L tris base; 190 mmol/L glycine; 20% methanol). 38 Membranes were stained with Ponceau S dye to visualize protein bands and allowed to air dry. Strips were cut from the membranes for immunoblotting. Multiplexing, Thermo Fisher Scientific, Vienna, Austria) using Immu-noCAP ISAC (MeDALL) technology as previously described. [40][41][42] The cut-off to be considered positive for IgE sensitization to any of the individual Der p allergens was set at 0.1 ISU. In addition, IgE specific to D. pteronyssinus extract was quantified using the d1 Immuno-CAP TM (Thermo Fisher Scientific, Uppsala, Sweden) and the cut-off for IgE sensitization was set at 0.1 kU A /L. Total plasma IgE was quantified by the total IgE ImmunoCAP TM (range: 2-5000 kU/L, Thermo Fisher Scientific). Samples with total IgE levels above 5000 kU/L were diluted in ImmunoCAP sample buffer to obtain the actual total IgE levels.

| Immunoblot and ImmunoCAP IgE inhibition assays
In IgE immunoblot inhibition experiments, 1/10 diluted plasma/ serum samples were preincubated overnight (4°C) with 200 lg of E. coli extract or for control purposes, with 0.5% bovine serum albumin and then tested for IgE reactivity to nitrocellulose-blotted HDM or E. coli extract. 39 The ImageJ software 43 was used to analyse and quantify the intensity of bands observed on the autoradiographs.
The reduction in density of bands from samples preincubated with bovine serum albumin compared to samples preincubated with E. coli  bacteria. Sixty-three per cent of the participants who had IgE to S. aureus also had IgE to E. coli and 53% of those who had IgE to E. coli also had IgE to S. aureus. We thus noted that the frequency of IgE sensitization to these bacterial antigens was lower in the Swedish AD patients (ie, 33%) as compared to that observed for Austrian and German AD patients earlier (ie, almost 40%). 15 3.2 | Low prevalence of IgE sensitization to HDM allergens in Swedish AD patients The frequency of IgE sensitization to HDM allergens was 25% among the Swedish AD patients, when measured by the allergen chip (Table 1). 28 Sensitization to HDM allergens was therefore considerably lower in the Swedish AD patients than in the Austrian and German AD patients (ie, more than 50%) who had been analysed for IgE reactivity to bacterial antigens earlier. 15 Age and gender distribution did not differ significantly between Swedish AD participants with HDM allergy and those without HDM allergy (Table S2A). However, HDM-sensitized AD patients had significantly higher levels of total IgE than HDM-nonsensitized AD patients ( Figure S1, Table S2A).

| The frequency of IgE sensitization to bacterial
antigens is significantly higher in patients with AD sensitized to HDM As it has been reported that the microbiome of HDM contains various bacteria, including S. aureus and E. coli, 18 we were wondering whether IgE sensitization to bacterial antigens may be associated with IgE sensitization to HDM. We therefore investigated the frequencies of sensitization to bacterial antigens in AD patients with and without HDM sensitization (Table 1). This analysis showed that 40% and 58% of HDM-sensitized AD patients had IgE antibodies specific for S. aureus and E. coli, respectively, whereas only 15% and 14% of AD patients without HDM sensitization showed IgE reactivity to S. aureus and E. coli. IgE reactivity to S. aureus and/or E. coli was present in 67% of the HDM-sensitized patients but only in 22% of those without HDM sensitization. The odds ratios for having IgE antibodies against E. coli, S. aureus or one of the two were 8.3, 3.8 and 7.2 in the HDM-sensitized vs the HDM-nonsensitized AD patients, respectively ( Table 1). The detailed frequencies of bacterial IgE sensitization among the HDM-sensitized and HDM-nonsensitized Swedish AD patients are depicted in the form of Venn diagrams in Figure 2.
Age and gender distribution did not differ significantly between participants with bacterial IgE and those without bacterial IgE (Table S2B-D).

| IgE sensitization to bacterial antigens but not
to HDM allergens is associated with severe AD Figure 3 shows the proportion of patients with severe AD, history of rhinitis and history of asthma among patients with or without IgE reactivity to HDM allergens and bacterial antigens. Table S2 provides detailed information regarding the numbers, gender and age of the patients in the different groups and concerning significant differences regarding the presence of moderate or severe AD and history of rhinitis and asthma in the groups with HDM and bacterial IgE sensitizations. We found that IgE reactivity to HDM and to S. aureus and/or E. coli was associated with rhinitis. Having IgE against HDM or IgE specifically against E. coli had a relative risk of 1.24 (P = .0306) associated with rhinitis ( Figure 3, Table S2A,D). IgE reactivity to S. aureus and/or E. coli was also associated with severe AD (ie, relative risk 1.68; P = .0359) (Figure 3, Table S2B). In particular, AD patients with IgE antibodies against S. aureus had an almost twofold increase in the relative risk (1.91; P = .0093) associated with severe AD compared to AD patients without IgE antibodies against S. aureus antigens ( Figure 3, Table S2C). Interestingly, we did not find associations between asthma or severe AD and IgE sensitization to HDM allergens in this population ( Figure 3, Table S2A).

| HDM contain bacterial antigens
To search for the presence of bacterial antigens in HDM, we prepared extracts from HDM bodies, separated them by SDS-PAGE and used rabbit antisera raised against S. aureus extract, E. coli extract or E. coli proteins, to detect S. aureus and E. coli antigens by immunoblotting  For control purposes, we tested rabbit sera raised against purified HDM allergens (nDer p 1, rDer p 2, rDer p 5, rDer p 7, rDer p 10,

| HDM contain IgE-reactive bacterial antigens
To detect IgE-reactive bacterial antigens in HDM, we then performed IgE immunoblot inhibition experiments 39 using serum/plasma samples where sufficient volumes were available. We tested whether preadsorption of plasma/sera from patients with AD with E. coli extract can inhibit IgE binding to blotted HDM extract ( Figure 5).

| Quantification of IgE towards bacterial antigens in HDM extracts
The IgE immunoblot inhibition experiments demonstrated that extracts from HDM contain IgE-reactive bacterial antigens. We were therefore interested to quantify the IgE levels towards bacterial antigens by ImmunoCAP TM inhibition experiments. Figure 6 shows the percentages of inhibition of IgE binding, which could be obtained when patients' plasma/sera ( Figure 6,  (Table S1). No reduction in IgE binding was observed in this sample after preincubation with E. coli extract ( Figure 6).

| DISCUSSION
HDM not only represent one of the most potent and frequent allergen sources worldwide, but they also contain factors that alter barrier function, induce pro-inflammatory cytokines and affect IgE responses as well as the innate immune system. 25 Recently, it has been reported that the HDM microbiome contains a variety of Gram-positive and Gram-negative bacteria including S. aureus and E. coli. 18 These bacteria seem to be important as part of the HDM microbiome because mites do not thrive when cultured with antibiotics, 45