A Controlled Challenge Study on Di(2-ethylhexyl) Phthalate (DEHP) in House Dust and the Immune Response in Human Nasal Mucosa of Allergic Subjects

Background Few studies have yet addressed the effects of di(2-ethylhexyl) phthalate (DEHP) in house dust on human nasal mucosa. Objectives We investigated the effects of house dust containing DEHP on nasal mucosa of healthy and house dust mite (HDM)–allergic subjects in a short-term exposure setting. Methods We challenged 16 healthy and 16 HDM-allergic subjects for 3 hr with house dust at a concentration of 300 μg/m3 containing either low (0.41 mg/g) or high (2.09 mg/g) levels of DEHP. Exposure to filtered air served as control. After exposure, we measured proteins and performed a DNA microarray analysis. Results Nasal exposure to house dust with low or high DEHP had no effect on symptom scores. Healthy subjects had almost no response to inhaled dust, but HDM-allergic subjects showed varied responses: DEHPlow house dust increased eosinophil cationic protein, granulocyte-colony–stimulating factor (G-CSF), interleukin (IL)-5, and IL-6, whereas DEHPhigh house dust decreased G-CSF and IL-6. Furthermore, in healthy subjects, DEHP concentration resulted in 10 differentially expressed genes, whereas 16 genes were differentially expressed in HDM-allergic subjects, among them anti-Müllerian hormone, which was significantly up-regulated after exposure to DEHPhigh house dust compared with exposure to DEHPlow house dust, and fibroblast growth factor 9, IL-6, and transforming growth factor-β1, which were down-regulated. Conclusions Short-term exposure to house dust with high concentrations of DEHP has attenuating effects on human nasal immune response in HDM-allergic subjects, concerning both gene expression and cytokines.

Phthalates are a group of organic esters known as indoor air pollutants. The most common phthalate indoors is di(2-ethylhexyl) phthalate (DEHP), which is widely used as plasticizer (softener), for example, to render polyvinyl chloride (PVC) more flexible. A wide variety of consumer products also contain DEHP, such as flooring and other building materials, household furnishings, clothing, cosmetics and personal care products, lubricants, waxes, cleaning materials, and medical products (Schettler 2006). Other phthalates commonly found in house dust include n-butyl benzyl phthalate (BBzP), di-n-butyl phthalate (DnBP), diisobutyl phthalate (DIBP), and diethyl phthalate (DEP) (Bornehag et al. 2005). Because phthalates are not covalently bound to the plastic matrix, they vaporize directly into the environment, where they can accumulate and adhere to inhalable airborne and sedimented dust particles. In several studies, the median DEHP concentrations in common households ranged from 0.4 to 0.96 mg/g house dust (Bornehag et al. 2005;Butte et al. 2001;Fromme et al. 2004;Kolarik et al. 2008;Pöhner et al. 1997).
Indoor air is an important pathway of human environmental exposure to various phthalates (Latini 2005). One of the potential health concerns of phthalate exposure is the development of asthma and allergies. Emissions from the degradation of PVC flooring materials evoke conjunctival, upper airway, and pulmonary irritations (Jaakkola et al. 1999;Wieslander et al. 1999). In addition, an association between the presence of DEHP in dust and the prevalence of asthma in exposed children has been suggested (Bornehag et al. 2004). Laboratory studies have shown that many phthalate compounds administered to mice by subcutaneous injection or by inhalation exert an adjuvant effect on the immune response to exposure to a coallergen Larsen et al. 2001bLarsen et al. , 2001a. After stimulation with DEHP, murine T cells respond with the enhanced production of interleukin (IL)-4 ( Lee et al. 2004), whereas cultured neutrophils of both humans and rodents exhibit an inflammatory response (Gourlay et al. 2003). However, few human exposure studies have addressed the effect of phthalates in house dust on mucosal airway response in humans. Tuomainen et al. (2006) challenged 10 subjects experimentally to degraded PVC products under controlled conditions and found no differences in the cytokine expression profiles in the nasal lavage fluid between challenge and control exposure, concluding that PVC materials do not evoke immediate asthmatic reactions. Further evidence for direct health effects of inhalational exposure to DEHP in humans is rare. In the present study, we challenged healthy and allergic human subjects with house dust containing low (0.4 mg/g) and high (2.0 mg/g) concentrations of DEHP in a short-term exposure setting. After nasal dust exposure, we analyzed differences of the transcriptional and secretory response of nasal mucosa using an oligonucleotide cDNA microarray and a microspherebased flow cytometric assay.

Materials and Methods
Subjects and eligibility. Sixteen healthy and 16 house dust mite (HDM)-allergic human subjects participated in this exposure study. Twenty subjects were male, and 12 were female, with ages ranging from 22 to 32 (mean, 24) years. Subjects without clinically relevant nasal disorders, except for allergic subjects suffering from HDM allergy, were eligible, as confirmed by rhinologic history and nasal endoscopy. We identified HDMallergic subjects by skin prick test according to the guidelines of the European Academy of Allergy and Clinical Immunology (Dreborg and Frew 1993): a CAP [or RAST (radioallergosorbent test)] class of at least 3 (Phadia AB, Uppsala, Sweden) and a positive nasal or conjunctival provocation test (Riechelmann et al. 2003b). Exclusion criteria were smoking, pregnancy or lactation, allergy to allergens other than HDM, a clinically relevant nasal septal deviation or turbinate hypertrophy, rhinitis medicamentosa, chronic rhinosinusitis, acute rhinosinusitis within the preceding 6 weeks, previous sinus surgery, bronchial hyperreactivity, lung emphysema, and any systemic therapy with corticosteroids or anti-inflammatory drugs or nasal corticosteroid therapy within the preceding 6 weeks. Our study was approved by the ethics committee of the University of Ulm, and each participant gave written informed consent.
Study design. We randomly assigned subjects to four groups exposed to house dust with either low or high concentrations of DEHP ( Figure 1): We challenged eight healthy and eight allergic subjects with DEHP low (0.41 mg/g) house dust, and eight healthy and eight allergic subjects with DEHP high (2.09 mg/g) house dust. We challenged each subject twice in a nose-only exposure setting: once with the house dust in a concentration of 300 µg/m 3 for 3 hr, and once with filtered air (control, 0 µg/m 3 ) for the same time period. The sequences of exposures varied in a random order and were single blinded. We kept a time interval of at least 14 days between the two exposures. We completed visual analogue scales (VASs) immediately before and after exposure. We collected nasal fluid on one nasal side 3 hr after the first exposure, and took a nasal biopsy on the opposite side. After the second exposure, we switched both sides.
Aerosol exposure. We generated aerosols with a modified rotation plate aerosol generator (Small Scale Powder Disperser, model 3433; TSI GmbH, Aachen, Germany) (Riechelmann et al. 2004). We dispersed the dust in conditioned air (35% relative humidity, 21°C). During the exposures, we constantly recorded the accuracy of dust exposures employing a laser particle counter Hiac Royco,Leonberg,Germany).
House dust. We collected the house dust in 42 households using commercial vacuum cleaners. We sieved the content of the vacuum cleaner bags to a fraction of less than 32 µm and then pooled and divided the dust in two fractions. We used one fraction directly for exposure (DEHP low ), and enriched the second fraction with DEHP high in a rotary evaporator Buchi,Flawil,Switzerland). Thus, the two house dust fractions were identical except for their content of DEHP. We measured the concentrations of DEHP and other phthalates by gas chromatography/mass spectrometry (HP G1800A, GCD series II MS; Agilent Technologies, Santa Clara, California, USA) (Butte et al. 2001).
Symptom scores. Before and immediately after each exposure, the subjects estimated their sensation of nasal obstruction, hypersecretion, dryness, itching, and sneezing, mucous membrane burning and unpleasant smell on a 10-cm-long horizontal VAS. We measured the distance from the left border of the scale to the marks made by the subjects in millimeters and calculated the difference after versus before exposure. Positive values indicate a degradation of the appropriate symptom after challenge compared with the control.
Nasal secretions and biopsy. At 3 hr postexposure, an open-cell flexible polyurethane foam sampler of 28 × 18 × 6 mm was placed into one nasal cavity posterior to the mucocutaneous junction under direct visualization and left in place for 10 min. After removal, we extracted the secreted fluid from the sampler by centrifugation and stored it at -20°C (Riechelmann et al. 2003a). On the opposite side, we took a nasal inferior turbinate biopsy under local anesthesia with a Fokkens forceps (Explorent, Tuttlingen, Germany).
DNA microarray analysis. We performed intraindividual differential (control vs. house dust exposure) gene expression analysis from four subjects of each exposure group. We randomly selected subjects for microarray analysis. The cDNA microarray consisted of 1,232 human genes, 10 different extrahuman spiking controls from Arabidopsis and Sinorhizobium genes, and randomized negative controls (i.e., oligonucleotides that do not bind human mRNA) in 300 and in 12 spot quadruples as described previously (Riechelmann et al. 2007). Briefly, we spotted oligomers (Operon Biotechnologies, Cologne, Germany) on Ultra Gaps 2 coated slides (Corning, Schiphol-Rijk, The Netherlands). We isolated total RNA from biopsy specimens using the RNeasy Mini Kit (Qiagen, Cologne, Germany). For spiking controls, we synthesized synthetic mRNA of 10 different Arabidopsis and Sinorhizobium genes on an ABI 394 synthesizer (Purimex, Staufenberg, Germany). We added 0.2-10 pg of this spiking mRNA to the total RNA of control and exposure specimens, resulting in ratios of 1:2 to 1:10. We reverse transcribed mRNA, labeled it using Superscript-2 (Invitrogen, Karlsruhe, Germany) in combination with the 3 DNA Array 350 Kit (Genisphere, Hatfield, PA, USA), and hybridized it with the spotted oligonucleotides at 57°C overnight. We scanned microarrays with a dual-laser microarray scanner (GenePix 4000 B; Axon Instruments, Foster City, CA, USA) recorded with GenePix Pro 4.1 software (Axon Instruments) and carried out further analysis using the platform-independent java application ArrayNorm, version 1.7.2 (Pieler et al. 2004). After background subtraction and Lowess subgrid normalization, we performed replicate handling as we averaged quadrupled spots within a slide. Accordingly, we calculated the log 2 -transformed cyanine 5:cyanine 3 ratio (log 2 -expression ratio) for all genes and controls. We identified differentially expressed genes between DEHP low and DEHP high exposure using Student's t-test with Bonferroni stepdown correction to adjust the critical alpha limit of 0.05. We considered only genes with valid expression values in each biologic replicate for analysis. We annotated genes according the glossary of the Human Genome Organisation (HUGO) (Wain et al. 2004). The full list of array data is available on the server of the Ear, Nose, and Throat Department Ulm (2008).
Statistical analysis. For VASs, we calculated changes before and after exposure with the Wilcoxon signed rank test. For protein concentrations, we calculated and tabulated the median and interquartile range (IQR). We calculated dependent samples using the Wilcoxon signed rank test, and independent samples with the Mann-Whitney U-test. We considered a p-value < 0.05 to be significant. We performed calculations using Systat 10.2 (Systat Software, Point Richmond, CA, USA).

Results
House dust. The concentration of DEHP in the DEHP low house dust fraction was 0.41 mg/g. Enrichment of house dust with DEHP yielded to a 5-fold higher concentration of 2.09 mg/g in the DEHP high fraction. Table 1 lists the amounts of phthalates, major allergens, fungal spores, and endotoxin activity. Pollen from weeds, alder, birch, and yew were identified but not quantified.
Symptom scores. Nasal exposure to DEHP low or DEHP high house dust did not cause significant changes in nasal symptom scores concerning dryness, itching and sneezing, mucous membrane burning, and unpleasant smell compared with filtered air exposure (Table 2). HDM-allergic subjects complained about nasal obstruction and hypersecretion after exposure to DEHP low house dust, but this effect was not significant (p > 0.05).
Cytokines and ECP. Sampling of epithelial lining fluid with a polyurethane foam sampler from all subjects yielded to a median volume of 275 µL (IQR, 180-450 µL). Table 3 outlines the results of protein determination.
Independent from exposure to DEHP low or DEHP high house dust, we found no significant changes in the concentrations of cytokines and ECP in nasal secretions in healthy subjects. Nasal challenge of eight HDM-allergic subjects with DEHP low house dust resulted in significantly elevated median concentrations of ECP (p = 0.01), G-CSF (p = 0.02), IL-5 (p = 0.03), and IL-6 (p = 0.02) in nasal secretions compared with the appropriate control exposure with filtered air (Figure 2). The eight HDMallergic subjects that we challenged with DEHP high house dust showed significantly    lower concentrations of G-CSF and IL-6 in nasal secretions than did the HDM-eight allergic subjects we challenged with DEHP low house dust (G-CSF, p = 0.04; IL-6, p = 0.001) (Figure 3). We found no significant changes for the other cytokines or for ECP. DNA microarray gene expression analysis. After control and dust exposures, we obtained nasal mucosa biopsies using the Fokkens forceps from all participants. The median amount of extracted total RNA from biopsy material was 10.1 µg (IQR, 7.8-13.2 µg). We performed gene expression analysis from four subjects of each exposure group. Because of insufficient hybridization quality, we excluded two microarrays from HDM-allergic subjects from further evaluation.

Discussion
In this study, we investigated the effects of house dust with low and high concentrations of DEHP on nasal mucosa of healthy and HDM-allergic human subjects in a shortterm exposure setting. Corresponding parameters included a transcriptional profile of 1,232 genes and the concentrations of several cytokines and ECP in nasal secretions. We found two overall results: a) Healthy subjects were not significantly affected by the challenge with house dust, and b) HDM-allergic subjects perceived little of the challenge, and low concentrations of DEHP induced a silent inflammation in human nasal mucosa. However, high concentrations of DEHP attenuated the immune response based on protein and mRNA levels.
We collected and pooled house dust from 42 German homes to provide a reasonably representative sample. Coarse particles with diameters > 10 µm are deposited predominantly in the upper airways (Keck et al. 2000). Hence, we used the selected fraction with coarse particles < 32 µm for exposure. A dust concentration of 300 µg/m 3 is equal to that in interiors with human activity (Junker et al. 2000), in offices (Molhave et al. 2004), and in public transportation (Praml and Schierl 2000). Because of its ubiquity, DEHP appears in almost every house dust sample. Dust samples collected in office buildings in Austria showed DEHP concentrations ranging up to 3 mg/g (Hutter et al. 2006). The median concentration of DEHP in dust collected in children's bedrooms in 346 homes in Sweden was 0.7 mg/g (Bornehag et al. 2004). These concentrations resemble the concentrations of 0.41 mg/g and 2.09 mg/g used in our study, which consequently agree with DEHP concentrations naturally occurring in house environments. The content of the major allergens Der p1 and Der f1 in house dust depends of several factors, such as geographic region, seasonal variation, climate conditions, age of the dwelling, and floor covering. In dwellings in the Netherlands, van Strien et al. (1994) found a mean Der p1 concentration of 2.4 µg/g for floor dust from living rooms, but

Concentration (ng/mL)
DEHP low DEHP high * in more than the half of the houses the maximum concentration exceeded 10 µg/g. However, the mean concentration of 0.09 µg/g (range, < 0.01-9.54 µg/g) Der p1 found in European homes within the European Community Respiratory Health Survey II is several times lower (Zock et al. 2006). Differences in allergen contents in dust samples may be caused by different dust collecting methods (Schram-Bijkerk et al. 2006). Our house dust was collected over at least 4 weeks in the whole dwelling with conventional vacuum cleaners, provided by the inhabitants themselves. When we used multilayered vacuum cleaner bags, we also included the fine dust fraction between the unique paper layers. It is possible that the long sampling time and dust extracted from the whole vacuum cleaner bag yielded Der p1 and Der f1 concentrations of 2.0 µg/g each, representing typical concentrations in house dust samples from common households. Endotoxin levels resembled concentrations found in a recent European study (Bouillard et al. 2006). Study subjects' nasal symptom severity is an important outcome measure in the assessment of dust-induced effects. Symptom scores based on VAS are being used increasingly for this purpose. Concerning healthy volunteers, no upper or lower respiratory symptoms were described after challenge with concentrated ambient air particles (23.1-311.1 µg/m 3 ) (Ghio et al. 2000) or urban dust (150 and 500 µg/m 3 ) (Riechelmann et al. 2004). However, upper respiratory symptoms were not associated with plastic wall materials (Jaakkola et al. 2000). As expected, house dust did not affect nasal function in our healthy subjects irrespective of the administered DEHP concentration. Although our HDMallergic subjects had positive nasal or conjunctival provocation tests, nasal challenge with house dust did not result in distinct allergic symptoms. Apparently, the concentration of house dust and the affiliated allergen dose was too low to elicit nasal symptoms. Particularly in persistent allergic rhinitis, affected subjects can habituate to allergen and remain asymptomatic. Among our HDM-allergic subjects, we found a faint but nonsignificant association with increased nasal obstruction and hypersecretion for DEHP low house dust but not for DEHP high house dust. Phthalates and their metabolites have the potential to interact with the immune system. Concentration-dependent effects of some phthalates are described in animal and in vitro studies. In general, lower concentrations of phthalates have shown stimulatory and higher concentrations suppressive properties (Jepsen et al. 2004; Nasal immune response to DEHP in allergic subjects Environmental Health Perspectives • VOLUME 116 | NUMBER 11 | November 2008   , v-erb-b2 erythroblastic leukemia viral oncogene homolog 3 (NM_001982); ESD, esterase D/formylglutathione hydrolase (NM_001984); FGF9, fibroblast growth factor 9 (glia-activating factor) (NM_002010); GNAT1, guanine nucleotide binding protein (G protein), alpha transducing activity polypeptide 1 (NM_000172); HLADPA1, major histocompatibility complex, class II, DP alpha 1 (NM_033554); IL6, IL-6 (NM_000600); ITGB5, integrin, beta 5 (NM_002213); KRT17, keratin 17 (NM_000422); LDHA, lactate dehydrogenase A (NM_005566). NTRK3, neurotrophic tyrosine kinase, receptor, type 3 (NM_002530); RORA, RAR-related orphan receptor A (NM_002943); TGFB1, transforming growth factor, beta 1 (NM_000660).  al. 2002). The effect of DEHP on nasal symptoms of HDM-allergic subjects may be concentration dependent. Potential adverse effects of phthalates arise from their developmental and reproductive toxicity, particularly in males (Howdeshell et al. 2007). Phthalates act as endocrine disruptors because they alter the hormone profiles of both sexes by interfering with estrogen-or androgen-signaling mechanisms (Ma et al. 2006;Sharpe and Irvine 2004). In a human breast cancer cell line, DEHP mimicked estrogen in the inhibition of tamixofen-induced apoptosis , and it acts as an anti-androgen in rats (Andrade et al. 2006). AMH is a member of the transforming growth factor-β (TGFB) family that is implicated in the regression of Müllerian ducts in male fetuses and in the development and function of gonads. Currently, AMH is not suspected to play a major role in endocrine disruption (Sharpe 2006). However, in HDM-allergic subjects, transcription of AMH increased significantly after exposure to DEHP high house dust compared with DEHP low house dust. It remains unknown whether DEHP affects transcription of AMH in nasal mucosa tissue via an androgen or estrogen receptor mechanism or some other pathway. Some more interesting observations include the down-regulation of LDHA and FGF9 in HDM-allergic subjects after exposure to DEHP high house dust. Both genes are involved in sex determination. In the testes of rats, gene expression of LDHA is down-regulated after phthalate exposure in utero (Plummer et al. 2007) and FGF9 plays distinct roles in testis development and male sex determination (Colvin et al. 2001). In fact, the potential adverse effects of DEHP have not yet been elucidated in detail. But DEHP is known to be a modulator of gene expression in human cells (Hokanson et al. 2006). Peroxisome-proliferator-activated receptor-α (PPARα) is a member of the nuclear receptor superfamily. Its activation can result in potential anti-inflammatory effects by repressing nuclear factor-κB (NF-κB) (Mandard et al. 2004). Metabolites of DEHP act as ligands of PPARα (Ward et al. 1998). Thus, DEHP may interact with NF-κB via the PPARα-pathway, thereby down-regulating gene expression.

Fold difference in gene expression
As mentioned above, the effects of some phthalates on immune responses are dose dependent, with adjuvant and suppressive properties in lower and higher concentrations, respectively. Larsen and colleagues demonstrated the adjuvant effect of DEHP after simultaneously administered ovalbumin, indicating that DEHP is a mixed T-helper cell type 1/2 adjuvant , but nonoccupational, realistic levels of DEHP neither revealed an adjuvant effect nor induced allergic lung inflammation in mice ). Our HDM-allergic subjects revealed no significant nasal symptoms after exposure to DEHP high house dust, but the cytokine expression profile changed. We found increases in ECP, G-CSF, IL-5, and/or IL-6 after exposure to DEHP low house dust. After exposure to DEHP high house dust ECP, G-CSF, IL-5, and/or IL-6 were not increased. Additionally, G-CSF and particularly IL-6 decreased significantly after exposure to DEHP high house dust, pointing toward an attenuated immune response at high concentrations of DEHP in house dust. It is not clear from the data in this study whether the applied range of exposure causes a linear or nonlinear attenuation because this study used a three-point design with no exposure (control), DEHP low house dust, and DEHP high house dust. This could be addressed by future studies by having several levels of exposure within a certain range.
As detailed above, a plausible molecular mechanism for the attenuating effects of DEHP is the repression of transcription factor NF-κB. Like DEHP, volatile organic compounds may attenuate the immune response (Wichmann et al. 2005). We found the same diminished release of cytokines after exposure to house dust with high concentrations of DEHP as was found for thalidomide (α-N-phthalimidoglutarimide), a sedative with related chemical structure and teratogenic properties (Rowland et al. 1998). However, thalidomide does not act directly on NF-κB (Rowland et al. 2001). These observations do not exclude a possible association between exposure to DEHP and suppressed gene expression via the NF-κB pathway, because the metabolites of DEHP, and not DEHP itself, interact with NF-κB.
The results of our gene expression analysis indicate an altering effect of DEHP on gene expression in human nasal mucosa. The changes demonstrated for IL-6 on the protein level are consistent with the results of gene expression analysis observed in HDM-allergic subjects: IL-6 was up-regulated after exposure to DEHP low house dust and down-regulated after exposure to DEHP high (Figure 4). We did not detect transforming growth factor-β1 (TGF-β1) protein, but transcription of TGFB1 decreased significantly after exposure of HDM-allergic subjects to DEHP high house dust. TGF-β1 is an anti-inflammatory cytokine with comparatively high concentrations in healthy mucosa but decreased concentrations in allergic inflammation (Benson et al. 2002). Hence, it remains unknown whether the decrease in TGFB1 transcription is related to the fact that DEHP attenuates the immune response or to the fact that TGF-β1 is already decreased in allergy mucosa.

Conclusions
In conclusion, our study provides the first results of short-term exposure to house dust containing low and high concentrations of DEHP on nasal mucosa of healthy and HDM-allergic human subjects. Healthy subjects were not affected by the dust exposures. Low doses of DEHP in house dust elicited a silent inflammation in nasal mucosa of HDM-allergic subjects on both the protein and mRNA level. High concentrations of DEHP in house dust attenuated this inflammation. Our results suggest a stimulating effect at low doses of DEHP and an attenuating effect at high doses. However, exposing adults for 3 hr in a short-term exposure setting is very different from long-term exposure of children to house dust and phthalates in their homes; these two populations may exhibit true differences in immunologic reactions.