Time-Dependent Proteomic iTRAQ Analysis of

Time-Dependent Proteomic iTRAQ Analysis of Nasal Lavage of Hairdressers Challenged by Persulfate Monica H. Kåredal, Harriet Mo ¨ rtstedt, Marina C. Jeppsson, Kerstin Kronholm Diab, Jørn Nielsen, Bo A. G. Jo ¨ nsson, and Christian H. Lindh* Division of Occupational and Environmental Medicine, Department of Laboratory Medicine, Lund University, SE-221 85 Lund, Sweden Received May 12, 2010

Hairdressers are frequently exposed to bleaching powder containing persulfates, a group of compounds that may induce hypersensitivity in the airways. The mechanism causing this reaction is not clear. The aim of this study was to identify changes in the nasal lavage fluid proteome after challenge with potassium persulfate in hairdressers with bleaching powder-associated rhinitis. Furthermore, we aimed to compare their response to that of hairdressers without nasal symptoms, and atopic subjects with pollen-associated nasal symptoms. To study the pathogenesis of persulfate-associated rhinitis, the response in protein expression from the upper airway was assessed by time-dependent proteomic expression analysis of nasal lavage fluids. Samples were prepared by pooling nasal lavage fluids from the groups at different time points after challenge. Samples were depleted of high-abundant proteins, labeled with iTRAQ and analyzed by online 2D-nanoLC-MS/MS. Differences in the protein pattern between the three groups were observed. Most proteins with differentially expressed levels were involved in pathways of lipid transportation and antimicrobial activities. The major finding was increased abundance of apolipoprotein A-1, 20 min postchallenge, detected solely in the group of symptomatic hairdressers. Our results suggest there may be differences between the mechanisms responsible for the rhinitis in the symptomatic and atopic group. Keywords: biomarkers • challenge of human subjects • iTRAQ • 2D-LC-MS/MS • proteomics • rhinitis

Introduction Hairdressers are exposed to reactive chemicals through the use of hair care products such as bleaching agents, permanent wave solutions, and hair dyes. Negative health effects are common in this profession. Epidemiological studies of female hairdressers have shown increased risks of developing symptoms from the upper and lower airways, such as asthma, chronic bronchitis, and rhinitis.1-3 Bleaching powders containing persulfates are often mentioned as cause of the airways symptoms in hairdressers.2 Several studies have shown that persulfates can elicit symptoms in hairdressers with bleaching powder-associated airways symptoms.4,5 Although the clinical picture is that of a hypersensitivity reaction, serum IgE antibodies against persulfates have only been found on rare occasions.6 A newly published study by Diab et al. reports that persulfates elicited nasal symptoms in a group of hairdressers with bleaching powder-associated rhinitis, as well as in hay fever patients, probably through a T helper 1 (Th1) cell activation.7 Differences in gene expression levels of certain cytokines were noticed between the two groups.8 * To whom correspondence should be addressed. Christian Lindh, Associate Professor, Division of Occupational and Environmental Medicine, Lund University, SE-221 85 Lund, Sweden. Phone: +46-46-173819. Fax: +4646-143702. E-mail: [email protected].

5620 Journal of Proteome Research 2010, 9, 5620–5628 Published on Web 09/04/2010

Proteomic-based studies of nasal lavage fluids and mucus have been used to assess the response from the airways and study the effects of chemicals,9,10 as well as the pathogenesis of rhinitis11 and chronic rhinosinusitis,12,13 among other diseases. Proteomic profiling of biological samples may identify changes in protein patterns, revealing the pathways or systems that are activated and potentially associated with the disease processes. Biomarkers reflecting changes early in the disease progression are important in preventive work, while biomarkers later in the progression are important as diagnostic markers. To date, the pathogenesis of bleaching powder-associated rhinitis is incomplete and tools for risk assessment are consequently lacking. Mass spectrometry (MS)-based proteomics are widely used today. In combination with the amine-specific isobaric tags for relative and absolute quantification (iTRAQ),14 they are a reliable way to identify and quantify proteomes. Using a fourplex iTRAQ reagent set, concurrent analysis of four samples can be performed. In this study, pooled nasal lavage fluids from human subjects challenged with potassium persulfate were analyzed. The nasal lavage fluids were obtained from a study that examined the effects of the persulfate challenge on both the nasal mucosa and immune cells.7 The aim of this study was to measure changes in the nasal proteome of hairdressers challenged with persulfates to eluci10.1021/pr100436a

 2010 American Chemical Society

Nasal Lavage of Hairdressers Challenged by Persulfate

research articles

Figure 1. Overview of the experimental procedure of sample preparation, labeling, and analysis. Nasal lavage fluids were pooled within each time point (at 0 and 20 min, and 5 h) and group (symptomatic, asymptomatic, and atopic groups) and depleted for six high-abundant proteins. Aliquots of each pool were combined to get a reference sample (“Ref”) containing all detectable proteins and used to enable comparison between the three groups. All samples were then reduced, alkylated, and trypsinized before iTRAQ labeling using a four-plex reagent set for each group. After labeling, the samples within each group were combined in a new test tube and analyzed using online 2D-nanoLC-MS/MS.

date the pathophysiological mechanism behind persulfateassociated rhinitis.

Materials and Methods Subjects. Samples were obtained from the study by Diab et al., where female subjects were challenged with potassium persulfate.7 Briefly, three groups of subjects were studied. They were: female hairdressers with work-related rhinitis mainly linked to bleaching powder-containing persulfates (symptomatics, N ) 15), female hairdressers without work-related nasal symptoms (asymptomatics, N ) 14), and atopic females with pollen-associated rhinitis, but without prior work-related exposure to persulfates (atopics, N ) 12). The reason for choosing the atopics as a control group was to investigate if the persulphate induced a specific or a nonspecific reaction as further discussed by Diab et al.7 Challenge. Nasal lavage fluids were collected twice prior to challenge by flushing the nasal cavity with 15 mL × 3 of an isotonic saline solution. The first lavage was performed to wash the cavity and the second constituted the baseline sample. The challenge was performed by spraying a 0.001% solution of potassium persulfate into the nasal cavity, followed, 20 min later, by a 0.01% solution.7 Nasal lavage fluids were collected 20 min, 2 h, and 5 h after the last challenge. All samples were frozen at -70 °C. All samples used in this study had been previously thawed. No blood contamination was visually observed in the samples. Ethics. The study was approved by the Ethical Committee at the Medical Faculty at Lund University and the participants gave written, informed consent. Samples. To obtain groups as distinct as possible, in an attempt to eliminate confounding effects, the inclusion criteria

in this study were more stringent than in the original study. Only samples from nonsmokers in all three groups were included. Furthermore, in the symptomatic group, we included only nonatopics and subjects with a symptom score change >3 (N ) 6). In the asymptomatic group, nonatopic subjects with a symptom score change 2.0 (confidence >99%) were further evaluated. To evaluate the expression pattern at the time points for each of the three groups, iTRAQ ratios were normalized to the associated 0 min time point. In addition, iTRAQ ratios were also normalized to the reference sample to make it possible to compare the ratios between the groups. The calculated bias correction factors (the median value of all ratios) were used to normalize the ratios to compensate for systematic errors. Sample Preparation for Verification Analysis by Scheduled Selected Reaction Monitoring (sSRM)-LC-MS/MS. Nasal lavage fluids were pooled for each group (symptomatics, asymptomatics, and atopics) and time point (0 min, 20 min, and 5 h), resulting in nine pooled samples. Within each pooled sample, equal amounts of protein from each subject were added to give a total amount of 100 µg. The samples were desalted, reduced, alkylated, and trypsin-digested, as described above. However, the samples were not depleted of high-abundant proteins in order to reduce inaccuracies. The samples were evaporated to


research articles

dryness and dissolved in 50 µL of 40% methanol with 0.1% formic acid before analysis.

cates were iTRAQ-labeled as follows: 10 µg: iTRAQ reagent 114; 20 µg: iTRAQ reagent 115; 20 µg: iTRAQ reagent 116; and 50 µg: iTRAQ reagent 117. The labeled samples were mixed, resulting in a 1:2:2:5 ratio. The samples were analyzed by 1DnanoLC-MS/MS and evaluated using Protein Pilot 3.0 software. All values were transformed into the natural logarithm before mean ratios were calculated.

Verification of Proteins with Significantly Changed iTRAQ Ratios. A targeted LC-MS/MS method based on sSRM experiments to relatively quantify 22 proteins (identified by significantly changed iTRAQ ratios) was developed. The SRM peptides were selected from Protein Pilot data by identifying clear and highly intense y- and b-ions. Analysis was performed using a linear ion trap quadrupole MS, QTRAP 5500 system (Applied Biosystems, Foster City, CA) coupled to a Prominence UFLCXR HPLC system from Shimadzu (Columbia, MD). A list of evaluated sSRMs and peptide sequences is supplied as Supporting Information in Table S3. Peptides were separated at a flow rate of 500 µL/min using a linear gradient over 10 min from 3 to 35% B and then over 1 min from 35 to 95% (solvent A: 0.08% formic acid in ultrapure water and solvent B: 0.08% formic acid in acetonitrile) on a C18 column (2.1 × 50 mm, 4 µm; Genesis, Grace Davison, Lokeren, Belgium). The MS settings were: positive mode, sSRM detection window 60 s, cycle time 1 s, ionspray voltage 5500 V, and temperature 500 °C. Expected retention time for each transition as well as collision energy was specified in the method. The acquired data was evaluated using Analyst 1.5.1 (Applied Biosystems, Foster City, CA). For all proteins, relative quantification was achieved by relating mean peak areas of duplicate analyses for each transition of the 20 min and 5 h time points to the 0 time point. Protein ratios were then calculated based on the geometric means of individual peptide ratios. Quantification of Apolipoprotein A-1 using sSRM-LC-MS/ MS. Tryptic peptides of apolipoprotein A-1 (>90% purity, Sigma Aldrich, St. Louis, MO) were analyzed with the QSTAR mass spectrometer using offline nanoelectrospray tips in MS and MS/MS mode to identify exact masses and fragmentation patterns (y- and b-series ions). Candidate SRM peptides were evaluated and two peptides best representing apolipoprotein A-1 were selected and included with two transitions each in the sSRM-LC-MS/MS method described above. Also, a standard curve for apolipoprotein A-1 was generated by serial dilution (in 40% methanol with 0.1% formic acid) of a tryptic digest of 50 µg reduced and alkylated protein. Concentrations of the standards were in the range of 2.5-0.019 µg/mL. Concentrations of apolipoprotein A-1 in samples were determined by evaluating the acquired data using Analyst 1.5.1. Protein Loss During Desalting. To evaluate the recovery of proteins after desalting, total protein concentration in nasal lavage fluids was determined before and after the desalting procedure. Normal nasal lavage fluids were pooled from twelve different samples and diluted to 25 mL with 0.9% NaCl. Ten aliquots of 2 mL each were evaporated to dryness, diluted with 450 µL water, and then desalted using Microcon centrifugal filters. The samples were centrifuged at 13 000× g for 1 h, the waste was discarded, 350 µL of water was added, and then the samples were centrifuged as before. The spin column was removed and placed upside down in a new Eppendorf tube. The samples were spun briefly for ∼10 s. The protein concentration was determined using a BCA protein assay kit. Reproducibility. To assess the reproducibility of the method, four aliquots from the same six-protein mixture (Applied Biosystems, Foster City, CA) were prepared in parallel. To four aliquots containing 10, 20, 20, and 50 µg of protein, 10 mL of 0.9% saline solution (Fresenius Kabi, Halden, Norway) was added. The samples were then treated according to the method procedure (protein depletion was not performed). The repli-

Results Proteomic Expression Analysis of Pooled Nasal Lavage Fluids. Proteins in nasal lavage samples were identified and relatively quantified with iTRAQ-2D-nanoLC-MS/MS, followed by database searches. The study resulted in identification of 70 proteins in the symptomatic group, 75 in the group without symptoms, and 81 in the atopic group (for a complete list, see the Supporting Information, Table S4). A total of 99 protein identifications were made; and 56 of these proteins were identified in all three groups. Out of the total number of protein identifications, nine, eleven, and 13 proteins had significantly (p < 0.05) changed levels in the sample from the symptomatic, asymptomatic, and atopic group, respectively (Table 1). Some of the proteins were found exclusively in one of the study groups. For a typical iTRAQ spectrum, see Supporting Information, S5. The results revealed some differences between the symptomatic, asymptomatic, and atopic groups. Major findings were the upregulated apolipoprotein A-1 and the twice downregulated lactotransferrin at 20 min after challenge in the symptomatic group, both levels returning to base level at 5 h. No changed levels of apolipoprotein A-1 were found for the two other groups. Two other proteins with increased levels at 20 min, detected in both the symptomatic and the atopic group, were alpha-amylase 1 and mucin-5B. A protein with a more pronounced increase at 5 h that was identified in the asymptomatic group was extracellular glycoprotein lacritin. Lipocalin-1 was detected in increased amounts at 5 h, but this was consistent across all groups. Protein ratios relative to the reference sample showed differences between the groups. For example, it was shown that levels of the alpha-2-macroglobulin were higher in the symptomatic than in the asymptomatic group, while the level in the atopic group was in between the other two groups at all three time points. Also, alpha-amylase 1 was higher in the symptomatic group compared to the asymptomatic group. CystatinSN was higher in the symptomatic and the atopic groups compared to the asymptomatics, while the opposite was found for IgGFc-binding protein and sulfhydryl oxidase 1. The results for all protein ratios relative to the reference sample are provided as Supporting Information (Table S6). Verification of Protein Changes by sSRM-LC-MS/MS. A mass spectrometric method was developed to verify the findings of the iTRAQ results. Proteins with significantly different levels in at least one of the groups at one of the time points were included by analyzing tryptic digests of pooled, undepleted nasal lavage fluids using an sSRM-LC-MS/MS method with at least two selected transitions per peptide. Mean ratios for all proteins are presented in Table 2. The significant result of increased levels of apolipoprotein A-1 (5.1) found in the symptomatic group was verified by the sSRM-LC-MS/MS analysis (4.5). The agreement between ratios of remaining samples regarding apolipoprotein A-1 varied to some extent but, on the other hand, none of these iTRAQ ratios were significant (see Supporting Information, S7, for an extracted Journal of Proteome Research • Vol. 9, No. 11, 2010 5623

research articles

Kåredal et al. a

Table 1. Significantly Differentially Expressed Proteins in at Least One Group at One Time Point

symptomatics: mean ratio relative to 0 time point accession

protein name

scoreb

# peptc

20 min (95%CI)d

5 h (95%CI)d

P61626 P02788 P31025 P04745 Q9HC84 P61769 P01036 P12273 P02647 Q96DA0 Q9UGM3 Q9GZZ8 P01023 P06702 O75556 P08118 P00450 P02774 O95968 P80188 P31151 P08246

Lysozyme C Lactotransferrin Lipocalin-1 Alpha-amylase 1 Mucin-5B Beta-2-microglobulin Cystatin-S Prolactin-inducible protein Apolipoprotein A-I Uncharacterized protein UNQ773/PRO1567 Deleted in malignant brain tumors 1 protein Extracellular glycoprotein lacritin Alpha-2-macroglobulin Protein S100-A9 Mammaglobin-B Beta-microseminoprotein Ceruloplasmin Vitamin D-binding protein Secretoglobin family 1D member 1 Neutrophil gelatinase-associated lipocalin Protein S100-A7 Leukocyte elastase

86.9 116.3 54.2 30.9 29.1 8.2 6.0 19.4 18.8 13.4 21.5 12.7 17.5 8.5 9.7 7.0 12.7 7.7 8.3 6.1

132 110 79 18 17 17 13 12 11 11 11 9 9 9 9 6 6 4 4 3

0.81 (0.70;0.93) 0.48 (0.38;0.60) 1.0 (0.83;1.2) 1.6 (1.3;1.9) 1.7 (1.3;2.2) 0.85 (0.35;2.0) 1.7 (1.2;2.3) 0.91 (0.69;1.2) 5.1 (2.7;9.7) 0.71 (0.64;0.80) 1.1 (0.54;2.3) 2.5 (1.3;5.0) 1.0 (0.10;10) 0.80 (0.58;1.1) 1.2 (0.73;1.9) 1.0 (0.68;1.6) 0.98 (0.77;1.2) 1.0 (0.55;1.8)

0.97 (0.89;1.1) 1.1 (0.94;1.2) 1.5 (1.2;1.8) 2.3 (1.9;2.8) 1.4 (1.1;1.7) 1.1 (0.72;1.6) 1.3 (0.86;1.9) 0.78 (0.51;1.2) 1.0 (0.84;1.2) 1.2 (1.0;1.3) 1.4 (0.79;2.4) 1.3 (0.99;1.6) 0.93 (0.07;13) 0.80 (0.35;1.8) 1.0 (0.60;1.8) 0.81 (0.45;1.5) 0.81 (0.49;1.3) 1.2 (0.07;22)

1.0 (0.67;1.5)

0.95 (0.23;4.0)

accession

protein name

scoreb

# peptc

20 min (95%CI)d

5 h (95%CI)d



87.3 136.0 50.7 10.3 19.8 9.6 17.6 17.1 8.3 14.2 22.6 11.9 14.4 13.8 18.3 11.5 14.4 10.4 9.8 6.2 4.0 5.1

167 112 73 5 9 20 14 8 4 12 12 8 7 12 16 8 7 6 9 3 2 2

0.94 (0.83;1.1) 0.93 (0.85;1.0) 0.95 (0.82;1.1) 1.1 (0.72;1.7) 0.87 (0.59;1.3) 0.73 (0.57;0.93) 1.5 (0.79;3.0) 1.1 (0.96;1.3) 1.0 (0.80;1.3) 1.0 (0.79;1.3) 1.3 (0.90;1.8) 2.2 (0.89;5.7) 1.2 (0.72;2.0) 0.54 (0.15;2.0) 2.1 (0.85;5.4) 0.97 (0.83;1.1) 1.0 (0.66;1.6) 1.2

0.94 (0.78;1.1) 1.2 (1.1;1.2) 1.6 (1.2;2.0) 0.96 (0.45;2.0) 0.61 (0.47;0.79) 0.82 (0.71;0.95) 2.2 (1.1;4.7) 1.7 (1.4;2.0) 1.4 (0.96;2.0) 1.4 (1.1;1.7) 1.9 (0.95;4.0) 6.6 (1.1;40) 0.77 (0.61;0.97) 0.71 (0.40;1.3) 4.7 (0.92;24) 0.74 (0.56;0.97) 0.96 (0.77;1.2) 0.96 (0.23;3.9)

0.74 (0.006;92) 1.0 (0.18;6.1) 0.54

1.1 (0.006;176) 0.57 (0.34;0.94) 0.33

accession

protein name

scoreb

# peptc

20 min (95%CI)d

5 h (95%CI)d



98.0 114.8 42.7 26.0 27.0 12.6 16.9 15.0 10.3 16.3 26.7 13.0 16.9 10.1 12.7 11.7 16.1 11.1 11.1 8.6

155 97 56 16 12 22 13 10 5 11 14 8 9 8 9 7 9 7 5 4

1.2 (0.97;1.4) 0.94 (0.81;1.1) 1.0 (0.78;1.3) 5.1 (3.5;7.6) 1.5 (1.1;2.0) 1.2 (0.50;2.9) 1.3 0.97 (0.68;1.4) 1.1 (0.64;2.0) 0.95 (0.77;1.2) 1.6 (0.90;2.8) 1.3 (1.0;1.7) 1.1 (0.90;1.4) 0.84 (0.72;0.98) 0.91 (0.34;2.4) 1.1 (0.52;2.3) 0.84 (0.70;1.0) 1.2 (1.1;1.3) 1.2 (1.0;1.4) 0.76 (0.56;1.0)

1.2 (1.0;1.3) 1.1 (0.98;1.2) 1.3 (1.0;1.5) 0.48 (0.35;0.67) 0.81 (0.56;1.2) 0.98 (0.80;1.2) 1.3 1.4 (0.88;2.2) 1.2 (0.40;3.7) 1.1 (0.87;1.5) 1.4 (1.1;1.8) 3.2 (1.9;5.3) 0.81 (0.62;1.1) 0.89 (0.65;1.2) 2.1 (1.2;3.7) 0.90 (0.56;1.5) 0.85 (0.78;0.93) 1.3 (0.77;2.2) 2.4 (0.43;13) 0.67 (0.52;0.87)

4.3

1

0.84 (0.20;3.5)

0.41 (0.35;0.47)

asymptomatics: mean ratio relative to 0 time point

atopics: mean ratio relative to 0 time point

a Samples were analyzed using online 2D-nanoLC-MS/MS and the acquired data processed using Protein Pilot 3.0. Proteins were identified and relatively quantified using the peak areas of the iTRAQ signature ions. iTRAQ ratios are expressed relative to the 0 time point. Bold figures indicate significant changes (p-value < 0.05). b Unused protein score. c Number of peptides with 95% confidence used for protein identification. d Calculated 95% confidence interval for each mean ratio.

5624

Journal of Proteome Research • Vol. 9, No. 11, 2010

research articles

Nasal Lavage of Hairdressers Challenged by Persulfate a

Table 2. Summary of the sSRM-LC-MS/MS Verification Analysis of Protein Ratios symptomatics: mean ratio relative to 0 time point

asymptomatics: mean ratio relative to 0 time point

atopics: mean ratio relative to 0 time point

5h 20 min 5h 20 min 5h 20 min (iTRAQ-ratio)b (iTRAQ-ratio)b (iTRAQ-ratio)b (iTRAQ-ratio)b (iTRAQ-ratio)b (iTRAQ-ratio)b

accession

protein name



1.2 (0.81) 0.79 (0.48) 0.93 (1.0) 2.4 (1.6) 1.3 (1.7) n.d.c 2.4 (1.7) 1.3 (0.91) 4.5 (5.1) 0.37 (0.71) 2.4 (1.1) 2.4 (2.5) 1.7 (1.0) 1.3 (0.80) 1.4 (1.2) 1.0 (1.0) 1.8 (0.98) 1.1 (1.0) n.d.c 1.2 (1.0) 0.84 n.d.c

0.50 (0.97) 0.63 (1.1) 0.77 (1.5) 2.0 (2.3) 0.75 (1.4) n.d.c 1.1 (1.3) 0.56 (0.78) 0.68 (1.0) 1.1 (1.2) 0.68 (1.4) 0.66 (1.3) 0.36 (0.93) 0.48 (0.80) 1.0 (1.0) 0.62 (0.81) 0.47 (0.81) 0.75 (1.2) n.d.c 0.76 (0.95) 0.88 n.d.c

0.62 (0.94) 2.3 (0.93) 0.33 (0.95) 0.35 (1.1) 0.21 (0.87) n.d.c 0.96 (1.5) 0.29 (1.1) 1.7 (1.0) 0.32 (1.0) 0.38 (1.3) 0.70 (2.2) 0.28 (1.2) 0.43 (0.54) 5.6 (2.1) 1.0 (0.97) 0.33 (1.0) 0.41 (1.2) n.d.c 1.0 (0.74) 1.0 (1.0) n.d.c

1.9 (0.94) 8.7 (1.2) 1.7 (1.6) 0.93 (0.96) 0.68 (0.61) n.d.c 2.7 (2.2) 1.3 (1.7) 3.0 (1.4) 1.5 (1.4) 2.0 (1.9) 4.2 (6.6) 0.81 (0.77) 0.65 (0.71) 10 (4.7) 2.5 (0.74) 1.2 (0.96) 1.6 (0.96) n.d.c 1.4 (1.1) 1.4 (0.57) n.d.c

0.88 (1.2) 1.3 (0.94) 0.96 (1.0) 5.9 (5.1) 0.65 (1.5) n.d.c 1.2 (1.3) 0.95 (0.97) 0.81 (1.1) 0.88 (0.95) 1.1 (1.6) 0.96 (1.3) 0.78 (1.1) 0.85 (0.84) 3.7 (0.91) 1.2 (1.1) 0.71 (0.84) 0.48 (1.2) n.d.c 0.71 (0.76) 0.98 n.d.c

0.87 (1.2) 1.5 (1.1) 1.2 (1.3) 0.18 (0.48) 0.73 (0.81) n.d.c 1.3 (1.3) 0.93 (1.4) 0.86 (1.2) 0.76 (1.1) 0.94 (1.4) 1.5 (3.2) 0.48 (0.81) 0.89 (0.89) 6.1 (2.1) 1.1 (0.90) 0.64 (0.85) 1.3 (1.3) n.d.c 1.8 (0.67) 1.6 n.d.c

a Tryptic digests of pooled, undepleted nasal lavage fluids were analyzed. Mean ratios of all proteins at 20 min and 5 h were presented relative to the 0 time point. iTRAQ ratios are presented in brackets. b Significantly (p < 0.05) changed iTRAQ ratios are given in bold. c n.d. ) not detected.

Table 3. Amount of Apolipoprotein in Pooled Nasal Lavage Fluids from Each Time Point and Groupa amount of apolipoprotein A-1 (ng) time point

symptomatics

asymptomatics

atopics

0 min 20 min 5h

20 90 14

4 7 12

13 11 12

a

Each pooled sample contained 100 µg of total protein.

chromatogram of apolipoprotein A-1). Out of 39 significant iTRAQ ratios, 24 were confirmed by the sSRM-LC-MS/MS analysis; eleven were not confirmed. The remaining four ratios were not determined because the corresponding proteins (beta2-microglobulin (two significant ratios), secretoglobin family 1D member 1, and leucocyte elastase) could not be detected. Quantification of Apolipoprotein A-1. A linear standard curve of apolipoprotein A-1 was obtained. Two peptides with two transitions each were evaluated by measuring the peak area. The quantified amount of apolipoprotein A-1 in the pooled samples was determined and is presented in Table 3. The results are in agreement with the iTRAC analysis. Moreover, the apolipoprotein A-1 levels in the asymptomatic hairdressers seem to be somewhat lower than in the other groups. Reproducibility Test of Protein Content After Desalting. The recovery of proteins during desalting steps was 70% on average, with a coefficient of variation (CV) of 4%. Overall Reproducibility. Evaluation of the mean iTRAQ ratios of the trypsin-digested six-protein solution, combined in a 1:2:2:5 ratio and analyzed using 1D-nanoLC-MSMS, gave an average of 1.7, 1.6, and 4.7, respectively, suggesting that the sample preparation procedure has acceptable reproducibility. Also, the bias correction used during data analysis compensates in part for the loss of accuracy.

Discussion This is the first study showing changes in the nasal proteome after challenge with potassium persulfate in hairdressers with bleaching powder-associated rhinitis. The group was compared to hairdressers without nasal symptoms, who did not react clinically to the challenge, and to pollen-allergic subjects with rhinitis, who reacted to a lesser extent.7 Several differences in relative protein expression between the three groups were identified. However, the major findings were a clear increase in apolipoprotein A-1 and a reduction in lactotransferrin in the symptomatic hairdresser group only. The results may provide information about the pathogenesis of persulfate-induced rhinitis and identify interesting proteins that should be evaluated as potential biomarkers of response or disease. Almost all of the identified proteins with changed expression, according to our criteria, were either linked to inflammation or involved in host defense, such as antimicrobial activities, in lipid transportation or in tissue repair. Apolipoprotein A-1 was significantly increased 5-fold in the symptomatic group only. This protein probably takes part in host defense. Apolipoprotein A-1 constitutes the major protein in high density lipoproteins (HDLs), which mediates reversed cholesterol transport but also has antioxidant, anti-inflammatory, and antithrombotic properties.15 Apolipoprotein A-1 has been proven to exhibit antiinflammatory properties by inhibiting activated monocytes from producing tumor necrosis factor alpha (TNF-R) and interleukin-1 beta (IL-1β).16 Apolipoprotein A-1/HDL has also been shown to have a positive association with asthma, allergies, and atopy.17,18 Apolipoprotein A-1 is probably in part responsible for the antioxidant properties of HDL because it can bind to, and reduce, oxidized lipids (of the low density lipoproteins, LDLs). Several methionine groups in the sequence can undergo oxidation19 and methionine oxidations have been suggested Journal of Proteome Research • Vol. 9, No. 11, 2010 5625

research articles

Kåredal et al. 20

to play a role during antioxidant events. It has been shown that exposure of proteins in vitro to persulfates may oxidize methionine and tryptophan residues in the proteins.21,22 Thus, the increased expression of apolipoprotein A-1 may be a protection against oxidation of proteins and other endogenous substances. On the other hand, this alone does not explain why increased expression was seen in the symptomatic group only. There must be some further reason for the upregulation in the symptomatic hairdressers, but this remains to be established. The higher level of apolipoprotein A-1 already at the 0 time point may be a result of previous exposure since the group constitutes occupationally persulfate exposed subjects. It is unlikely that the 5-fold increase in apolipoprotein A-1, detected in the symptomatic group, would be a result of plasma exudation or plasma leaking into the nasal mucosa since no increase in expression levels for plasma proteins such as alpha2-macroglobulin, a marker for plasma exudation, was found at 20 min. On the other hand, the levels of alpha-2-macroglobulin were constantly higher in the symptomatics, which may indicate a higher degree of plasma exudation/leakage in this group. It cannot be ruled out that this group could be more sensitive to the lavage procedure. In contrast to apolipoprotein A-1, lactotransferrin was decreased 20 min after challenge in the symptomatic hairdresser group only. Lactotransferrin is an iron-binding transport protein with antimicrobial properties23 which probably take part in the host defense. It is a product of glandular processes in the nasal mucosa.23 Interestingly, in agreement with the present study, lactotransferrin in serum was recently found to be inversely associated with allergic rhinitis caused by mites.24 A number of other proteins with antimicrobial activities were detected with significantly changed expression in one or several of the study groups. Lysozyme C showed decreased levels at 20 min in the symptomatic group and a minor increase at 5 h in the atopic group. This is an antimicrobial protein, produced and secreted by glands, that degrades bacterial cell walls and can be found in nasal secretions.23 In accordance with our results, lysozyme was present at lower levels in the nasal mucus of patients with chronic rhinosinusitis compared to healthy controls.12 It has also been associated with allergic rhinitis.25 Protein S100-A7, with a decreased level after 5 h in the asymptomatic group, has antimicrobial properties and has been shown to activate neutrophils to produce cytokines.26 Changes in lipocalin-1 were consistent between the groups, with unchanged levels at 20 min followed by increased levels at 5 h in all groups, suggesting that the upregulation was associated with the exposure situation or a daily variation rather than with the symptoms. Lipocalin-1 is a protein that binds to lipophilic molecules such as fatty acids and lipid peroxidation products (e.g., isoprostane and arachidonic acid).27 Mucin-5B was present at increased levels in the symptomatic group at 20 min and 5 h, as well as in the atopic group at 20 min, while a decreased level was seen in the asymptomatic group at 5 h. This suggests that the increased levels were associated with the symptoms. This high-molecular-weight protein is a gel-forming mucin and is secreted from the lacrimal gland28 as well as from the small airways.29 In general, excessive mucus production is commonly seen in respiratory disease30 and increased levels of the mucins 5B have been found in persons suffering from chronic rhinosinusitis.31 Extracellular glycoprotein lacritin was increased in the symptomatic and atopic groups at 20 min. At 5 h, this protein 5626


was found at even higher levels in the asymptomatic and atopic groups. It was most increased in the asymptomatic group, suggesting a protective role for this protein. Lacritin is secreted by lacrimal and salivary glands32 and has been suggested to stimulate epithelial proliferation.33 It is possible that the salivary glands were stimulated by persulfate exposure and secreted lacritin, explaining the increased levels. It is also possible that it was released to start proliferation of the epithelial layer that was possibly damaged by exposure. Increased levels of mammaglobin-B were observed at 5 h in the atopic group. No change was seen in the symptomatic group. Mammaglobin-B belongs to the uteroglobin gene family expressing proteins which have anti-inflammatory properties.34 Increased expression levels of mammaglobin messenger ribonucleic acid (mRNA) have been found in mucosal nasal tissue of patients with nasal polyps and allergic rhinitis, compared to patients with allergic rhinitis only.35 In this study, alpha-amylase 1 was upregulated in the symptomatic group at 20 min as well as 5 h and in the atopic group at 20 min. However, at 5 h, it was highly downregulated in the atopic group. For the asymptomatic group, no changes were detected. This is a salivary protein and could be a contamination of the samples. This protein was also found to be upregulated in nasal mucus in a study of patients with chronic sinusitis.12 Cystatin-S was upregulated in the symptomatic group at 20 min and in the asymptomatic group at 5 h. By contrast, in another study, lower levels of this protein were found in nasal fluids from persons suffering from seasonal allergic rhinitis compared to controls.11 The levels of cystatin-SN in the asymptomatic group were lower than in the other two groups. Cystatin-S and cystatin-SN are proteinase inhibitors, which could suggest that the need for inhibition of proteases was highest in the two groups with symptoms. It has been reported that the total protein concentration decreased to 60% of the initial amount 30 min after a nasal lavage.36 In our study, the total protein levels also decreased after the baseline lavage. However, 20 min after the 0 time point there was no decrease in total protein on average. An sSRM-LC-MS/MS method was applied in this study to confirm the iTRAQ ratios. This method was applied because we wanted to use a multiplexed method, enabling analysis of many proteins simultaneously. The method used was highly selective and sensitive. With this method, the increased level of apolipoprotein A-1 in the sample from the symptomatic group at 20 min was confirmed. Moreover, most of the other iTRAQ ratios were confirmed; however, for some proteins, the LC-MS/MS analyses deviated from the iTRAQ ratios. There may be several reasons for this. The verification by sSRMLC-MS/MS was performed on undepleted nasal lavage samples and this is probably the major reason why all significant iTRAQ results could not be confirmed. The dissimilarities between iTRAQ and sSRM ratios could also be due to differences within the trypsination step, which could be addressed in the future by addition of internal standards. Another reason could be that some of the significant proteins were low-abundance proteins only identified with a few peptides and therefore quantified at a lower confidence level. Persulfates may initiate the formation of reactive intermediates, for example, reactive oxygen species (ROS), that can modify lipids or proteins. Moreover, they can themselves modify macromolecules through oxidation. As previously mentioned, in vitro studies of persulfate-exposed proteins have


research articles

shown that persulfate oxidizes methionine and tryptophan residues.21 To improve the accuracy of the protein identification process and quantification, these potential modifications were included in the database searches. Proteins in nasal lavage fluids are supposedly extracellular proteins since the cells are removed by centrifugation. However, some intracellular proteins may be present due to lyzed cell walls, etc. Intracellular proteins may be present in the nasal cavity due to cell disruption caused by the method and/or cellular events as a result of the exposure. If the flushing of the nasal cavity during collection of nasal lavage is too vigorous or if the mucous membrane is fragile, blood vessels may be ruptured, resulting in blood leakage. This would contaminate the nasal lavage sample with red blood cells and plasma proteins. However, no hemoglobin was detected in our samples from the symptomatic and asymptomatic groups, suggesting that no addition of proteins had occurred due to nose bleeding. In the atopic group, the beta unit of hemoglobin was identified with two peptides (unused protein score 3.7), which implies a small contamination by nose bleeding. However, this contribution would probably have a limited impact on the results. In this study, we analyzed pooled samples, thus generating data of biological averages, in order to evaluate the effects on the group level. Information relating to biological variation is therefore lost. Theoretically, a detected upregulation may come from only one of the samples, thus not representing the group. However, a fairly high concentration would be required if one sample would have such impact on the mean. The changed levels of several proteins detected exclusively in the symptomatic group suggest that this group did indeed react with mechanisms different from the atopic and asymptomatic groups. Diab et al. showed that the symptomatic hairdresser and atopic groups developed significantly more symptoms after the persulfate challenge compared to the asymptomatic group.7 No specific IgE antibodies were found and it was suggested that the reaction was driven by Th1 cell activation. Jo¨nsson et al. found an altered gene expression of cytokines when analyzing mRNA in nasal lavage from the same subjects.8 They detected an increase in IL-5 in the symptomatic as well as the atopic group, suggesting eosinophilic activation. In the latter group, an increase in IL-13 expression was also seen. In this study, we found differences between the symptomatic and atopic groups, which indicate that, although members of both groups reacted to the challenge, the mechanism that underlies the onset of symptoms may be different. This study deals with the pathogenesis of persulfate-induced rhinitis, which is not very common outside the hairdressing profession. However, many other low molecular weight chemicals seem to induce airways diseases with a pathogenesis different from the classical type-I allergy with specific IgE antibodies. It is possible that these airways diseases all share the same pathogenesis as persulfate-induced rhinitis, but this remains to be tested in future studies.

need to be confirmed in future studies of proteomic analyses of the individual nasal lavage fluids. Abbreviations: BCA, bicinchoninic acid; CI, confidence interval; 2D-nanoLC-MS/MS, two-dimensional nanoliquid chromatography coupled to tandem mass spectrometry; HDL, high density lipoprotein; HPLC, high-performance liquid chromatography; IDA, information-dependent acquisition; IgA, IgE, IgG, etc., immunoglobulin A, E, G, etc.; IgGFc, immunoglobulin G fragment complement; IL, interleukin; iTRAQ, isobaric tags for relative and absolute quantification; LC, liquid chromatography; LDL, low density lipoprotein; MS, mass spectrometry; MS/MS, tandem mass spectrometry; ROS, reactive oxygen species; RP, reversed phase; SCX, strong cation exchange; sSRM, scheduled selected reaction monitoring; Th1, T helper 1; TNFR, tumor necrosis factor-alpha.

Conclusion Proteomic methods may serve as a powerful tool to investigate pathogenesis of persulfate-induced rhinitis. Our results suggest that proteins with various functions such as host defense and lipid metabolism were potentially involved in the pathogenesis. Different mechanisms may be responsible for persulfate-induced rhinitis in hairdressers with bleaching powder exposure and pollen-allergic subjects. These results

Acknowledgment. This work was supported by the Swedish Council for Working Life and Social Research (Grant No. 2005-0687), the AFA Foundation, and Lund University, Sweden. Supporting Information Available: Supplementary files S1-S7. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Leino, T.; Tammilehto, L.; Luukkonen, R.; Nordman, H. Self reported respiratory symptoms and diseases among hairdressers. Occup. Environ. Med. 1997, 54, 452–455. (2) Albin, M.; Rylander, L.; Mikoczy, Z.; Lillienberg, L.; Dahlman Ho¨glund, A.; Brisman, J.; Tore´n, K.; Meding, B.; Kronholm Diab, K.; Nielsen, J. Incidence of asthma in female Swedish hairdressers. Occup. Environ. Med. 2002, 59, 119–123. (3) Brisman, J.; Albin, M.; Rylander, L.; Mikoczy, Z.; Lillienberg, L.; Ho¨glund, A. D.; Tore´n, K.; Meding, B.; Diab, K. K.; Nielsen, J. The incidence of respiratory symptoms in female Swedish hairdressers. Am. J. Ind. Med. 2003, 44, 673–678. (4) Mun ˜ oz, X.; Cruz, M. J.; Orriols, R.; Torres, F.; Espuga, M.; Morell, F. Validation of specific inhalation challenge for the diagnosis of occupational asthma due to persulphate salts. Occup. Environ. Med. 2004, 61, 861–866. (5) Moscato, G.; Pignatti, P.; Yacoub, M. R.; Romano, C.; Spezia, S.; Perfetti, L. Occupational asthma and occupational rhinitis in hairdressers. Chest 2005, 128, 3590–3598. (6) Aalto-Korte, K.; Ma¨kinen-Kiljunen, S. Specific immunoglobulin E in patients with immediate persulfate hypersensitivity. Contact Dermatitis 2003, 49, 22–25. (7) Diab, K. K.; Truedsson, L.; Albin, M.; Nielsen, J. Persulfate challenge in female hairdressers with nasal hyperreactivity suggests immune cell. but no IgE reaction. Int. Arch. Occup. Environ. Health 2009, 82, 771–777. (8) Jo¨nsson, L. S.; Broberg, K.; Paulsson, K.; Kronholm Diab, K. Nielsen, J Gene expression in nasal lavage from hairdressers exposed to persulfate. Int. Arch. Occup. Environ. Health 2009, 82, 1261–1266. (9) Lindahl, M.; Irander, K.; Tagesson, C.; Ståhlbom, B. Nasal lavage fluid and proteomics as means to identify the effects of the irritating epoxy chemical dimethylbenzylamine. Biomarkers 2004, 9, 56–70. (10) Johannesson, G.; Lindh, C.; Nielsen, J.; Bjo¨rk, B.; Rosqvist, S.; Jo¨nsson, B. A. In vivo conjugation of nasal lavage proteins by hexahydrophthalic anhydride. Toxicol. Appl. Pharmacol. 2004, 194, 69–78. (11) Ghafouri, B.; Irander, K.; Lindbom, J.; Tagesson, C.; Lindahl, M. Comparative proteomics of nasal fluid in seasonal allergic rhinitis. J. Proteome Res. 2006, 5, 330–338. (12) Tewfik, M. A.; Latterich, M.; DiFalco, M. R.; Samaha, M. Proteomics of nasal mucus in chronic rhinosinusitis. Am. J. Rhinol. 2007, 21, 680–685. (13) Benson, L. M.; Mason, C. J.; Friedman, O.; Kita, H.; Bergen, H. R., 3rd; Plager, D. A. Extensive fractionation and identification of proteins within nasal lavage fluids from allergic rhinitis and asthmatic chronic rhinosinusitis patients. J. Sep. Sci. 2009, 32, 44– 56. (14) Ross, P. L.; Huang, Y. N.; Marchese, J. N.; Williamson, B.; Parker, K.; Hattan, S.; Khainovski, N.; Pillai, S.; Dey, S.; Daniels, S.;

Journal of Proteome Research • Vol. 9, No. 11, 2010 5627

research articles

(15) (16)

(17)

(18)

(19)

(20) (21) (22)

(23) (24)

(25)

5628

Purkayastha, S.; Juhasz, P.; Martin, S.; Bartlet-Jones, M.; He, F.; Jacobson, A.; Pappin, D. J. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell. Proteomics 2004, 3, 1154–1169. Ansell, B. J.; Watson, K. E.; Fogelman, A. M.; Navab, M.; Fonarow, G. C. High-density lipoprotein function recent advances. J. Am. Coll. Cardiol. 2005, 46, 1792–1798. Hyka, N.; Dayer, J. M.; Modoux, C.; Kohno, T.; Edwards, C. K., 3rd; Roux-Lombard, P.; Burger, D. Apolipoprotein A-I inhibits the production of interleukin-1beta and tumor necrosis factor-alpha by blocking contact-mediated activation of monocytes by T lymphocytes. Blood 2001, 97, 2381–2389. Scha¨fer, T.; Ruhdorfer, S.; Weigl, L.; Wessner, D.; Heinrich, J.; Do¨ring, A.; Wichmann, H. E.; Ring, J. Intake of unsaturated fatty acids and HDL cholesterol levels are associated with manifestations of atopy in adults. Clin. Exp. Allergy 2003, 33, 1360–1367. Nagel, G.; Weiland, S. K.; Rapp, K.; Link, B.; Zoellner, I.; Koenig, W. Association of Apolipoproteins with Symptoms of Asthma and Atopy among Schoolchildren. Int. Arch. Allergy Immunol. 2009, 149, 259–266. Sigalov, A. B.; Stern, L. J. Oxidation of methionine residues affects the structure and stability of apolipoprotein A-I in reconstituted high density lipoprotein particles. Chem. Phys. Lipids 2001, 113, 133–46. Stadtman, E. R.; Moskovitz, J.; Levine, R. L. Oxidation of Methionine Residues of Proteins: Biological Consequences. Antioxid. Redox Signaling 2003, 5, 577–582. Hird, F. J.; Yates, J. R. The oxidation of protein thiol groups by iodate, bromate and persulphate. Biochem. J. 1961, 80, 612–616. Mo¨rtstedt, H.; Jeppsson, M. C.; Jo¨nsson, B. A. G.; Kåredal, M. H.; Lindh, C. H. Identification of oxidative modification sites after in vitro incubation of human hemoglobin and human serum albumin with potassium persulfate. Rapid Commun. Mass Spectrom. 2010, Submitted. Raphael, G. D.; Jeney, E. V.; Baraniuk, J. N.; Kim, I.; Meredith, S. D.; Kaliner, M. A. Pathophysiology of rhinitis. Lactoferrin and lysozyme in nasal secretions. J. Clin. Invest. 1989, 84, 1528–1535. Choi, G. S.; Shin, S. Y.; Kim, J. H.; Lee, H. Y.; Palikhe, N. S.; Ye, Y. M.; Kim, S. H.; Park, H. S. Serum lactoferrin level as a serologic biomarker for allergic rhinitis. Clin. Exp. Allergy 2010, 40, 403– 410. Kalfa, V. C.; Spector, S. L.; Ganz, T.; Cole, A. M. Lysozyme levels in the nasal secretions of patients with perennial allergic rhinitis and recurrent sinusitis. Ann. Allergy Asthma Immunol. 2004, 93, 288–292.


Kåredal et al. (26) Zheng, Y.; Niyonsaba, F.; Ushio, H.; Ikeda, S.; Nagaoka, I.; Okumura, K.; Ogawa, H. Microbicidal protein psoriasin is a multifunctional modulator of neutrophil activation. Immunology 2008, 124, 357–367. (27) Wojnar, P.; Dirnhofer, S.; Ladurner, P.; Berger, P.; Redl, B. Human lipocalin-1, a physiological scavenger of lipophilic compounds, is produced by corticotrophs of the pituitary gland. J. Histochem. Cytochem. 2002, 50, 433–435. (28) Paulsen, F. Cell and molecular biology of human lacrimal gland and nasolacrimal duct mucins. Int. Rev. Cytol. 2006, 249, 229–279. (29) Evans, C. M.; Kim, K.; Tuvim, M. J.; Dickey, B. F. Mucus hypersecretion in asthma: causes and effects. Curr. Opin. Pulm. Med. 2009, 15, 4–11. (30) Kettle, R.; Simmons, J.; Schindler, F.; Jones, P.; Dicker, T.; Dubois, G.; Giddings, J.; Van Heeke, G.; Jones, C. E. Regulation of neuregulin 1beta1-induced MUC5AC and MUC5B expression in human airway epithelium. Am. J. Respir. Cell Mol. Biol. 2010, 42, 472–481. (31) Viswanathan, H.; Brownlee, I. A.; Pearson, J. P.; Carrie, S. MUC5B secretion is up-regulated in sinusitis compared with controls. Am. J. Rhinol. 2006, 20, 554–557. (32) Sanghi, S.; Kumar, R.; Lumsden, A.; Dickinson, D.; Klepeis, V.; Trinkaus-Randall, V.; Frierson, H. F., Jr.; Laurie, G. W. cDNA and genomic cloning of lacritin. a novel secretion enhancing factor from the human lacrimal gland. J. Mol. Biol. 2001, 310, 127–139. (33) Wang, J.; Wang, N.; Xie, J.; Walton, S. C.; McKown, R. L.; Raab, R. W.; Ma, P.; Beck, S. L.; Coffman, G. L.; Hussaini, I. M.; Laurie, G. W. Restricted epithelial proliferation by lacritin via PKCalphadependent NFAT and mTOR pathways. J. Cell. Biol. 2006, 174, 689– 700. (34) Becker, R. M.; Darrow, C.; Zimonjic, D. B.; Popescu, N. C.; Watson, M. A.; Flemingv, T. P. Identification of mammaglobin B, a novel member of the uteroglobin gene family. Genomics 1998, 54, 70– 78. (35) Fritz, S. B.; Terrell, J. E.; Conner, E. R.; Kukowska-Latallo, J. F.; Baker, J. R. Nasal mucosal gene expression in patients with allergic rhinitis with and without nasal polyps. J. Allergy Clin. Immunol. 2003, 112, 1057–1063. (36) Riechelmann, H.; Deutschle, T.; Friemel, E.; Gross, H. J.; Bachem, M. Biological markers in nasal secretions. Eur. Respir. J. 2003, 21, 600–605.

PR100436A

Time-Dependent Proteomic iTRAQ Analysis of

Recommend Documents