Proteomic profiling of commercial dust mite skin prick test solutions and allergy vaccines from India.

BACKGROUND: Skin prick test (SPT) solutions and allergy vaccines (AVs) are crucial tools for diagnosis and therapy of allergies. It was the aim of this study to corroborate the content of products for diagnosis and treatment of dust mite allergies that are produced and sold in India.
METHODS: SDS-PAGE, immunoblots and high-resolution mass spectrometric analysis was performed with 16 house dust mite (HDM) SPT solutions and AVs from 3 Indian manufacturers. Authority-approved European SPT solutions and in-house extracts were used as references.
RESULTS: From the 5 Indian Dermatophagoides pteronyssinus products, none contained proteins from this source. Instead, 1 sample contained Dermatophagoides farinae and human serum proteins, 4 products contained allergens from the storage mite Suidasia medanensis, allergens from the legume Cicer arietinum (chickpea), and proteins from baker's yeast. From 4 Indian D. farinae-labeled products, 2 contained human serum proteins and a limited number of D. farinae allergens. Two contained only Suidasia, Cicer, and yeast proteins. In contrast, the European authority-approved D. pteronyssinus and D. farinae SPT solutions that were used as reference in this study, contained exclusively proteins of the respective species and covered the expected allergen spectra. The Blomia tropicalis sample contained no Blomia allergens at all, but consisted exclusively of Suidasia, Cicer, and yeast proteins. All 6 HDM samples consisted of human serum proteins and limited amounts of D. farinae allergens.
CONCLUSIONS: All commercial Indian SPT solutions and AVs analyzed in this study are not suitable for dust mite allergy diagnosis and therapy, as they contain either no, or only a limited number of, HDM allergens. In addition, their use could lead to misdiagnosis since some of them contain allergens from other sources, including the storage mite Suidasia, chickpea, as well as baker's yeast. Further, their application might be harmful to patients, as some products contain large amounts of proteins of human origin. Analysis of European SPT solutions, on the other hand, confirmed their suitability for dust mite allergy diagnosis.

Introduction

The “allergy epidemic” has become a major burden to public health, affecting more than 150 million people in Europe and approximately 25–30% of the world's population.1, 2, 3 Particularly, in developing countries they are suspected to rise even further over the next few years. Skin prick testing (SPT) currently represents the most commonly used method applied by clinicians to verify sensitization of IgE-mediated allergic diseases. SPT is generally considered a sensitive, relatively specific, and minimally invasive procedure, in which patients are tested against diverse allergenic sources through the introduction of natural allergen extracts into the skin using a standard skin prick test lancet. A resulting wheal-and-flare reaction provides a visual and objective indication of sensitivity., However, the accuracy of such tests is highly dependent on the quality of the employed extract solutions and thus on the underlying production processes, handling, and storage conditions. To guarantee batch-to-batch consistency and reproducibility of diagnostic tests, standardization of allergen extracts is essential. In this respect, initiatives such as the Global Allergy and Asthma European Network (GA2LEN), and the European Union CREATE project network, have significantly contributed for standardization of allergenic products in Europe and the United States, which are controlled by their respective authorities such as, the Paul Ehrlich Institute in Germany, and the US Food and Drug Administration (FDA) via its Center for Biologics Evaluation and Research (CBER)., Their protocols and guidelines regulate standardization procedures including the selection of source materials and the quality control of extracts according to: i) their major allergen contents, ii) allergen references/standards, and iii) their allergenic potency. Furthermore, the guidelines clearly state that extracts have to be free of other allergenic source materials, as it could lead to false-positive results.

In other countries, allergen extracts do not necessarily follow all the above-mentioned guidelines and special requirements. For example in India, Drug Controller General of India (DCGI) and their associated agencies are responsible for the manufacture and approval of allergen preparations, which generally relies only on Good Manufacture Practice (GMP) and allergen units.

House dust mites (HDMs) and related allergens are major allergen sources worldwide. They potently trigger innate and adaptive immune responses, which typically manifest as allergic rhinitis and bronchial asthma. The most dominant genera of house dust mites are Dermatophagoides (species: D. pteronyssinus and D. farinae) followed by Blomia (species: B. tropicalis), which globally affect 65 to 130 million allergic individuals.12, 13, 14 Reported HDM sensitization rates differ widely in various regions reaching values higher than 20% in Europe, 12% in India, and up to 40% in North America., This variability has been attributed to diverse factors such as geographical locations, quality of raw material and diagnostic tests used, socio-economic conditions, and ethnicities. Dermatophagoides not only represents the most common but also the best studied HDM genus. The WHO/IUIS Allergen Nomenclature Subcommittee database has currently 30 D. pteronyssinus (European HDM) and 36 D. farinae (American HDM) allergens acknowledged. Among them, cysteine protease allergens are classified as Group 1 allergens and comprise the major allergens Der p 1 and Der f 1. Group 2 includes MD-2 like lipid-binding proteins, such as the major allergens Der p 2 and Der f 2. Group 1 and Group 2 allergens account for 50–80% of the total IgE binding measured for HDM extracts. For B. tropicalis, which is abundant mostly in tropical and subtropical regions, 14 allergens have been identified and listed in the WHO/IUIS Allergen database. Among them, Blo t 5 is considered an important major allergen responsible for 60% of the IgE binding measured in the whole extract.,

In India, dust mites are among the most important sources of inhalant allergens.,, However, reliable sensitization rates for HDM in India have been hampered by the use of different allergen extracts and different testing protocols. In fact, a recent review by Dey et al, focusing on HDM allergy in India, pointed out the difficulties in comparing SPT results from different centers and studies and highlighted the urgent need for standardization of reagents and procedures in allergy diagnosis and treatment within the country. In this respect, a study evaluating the potency of SPT allergen solutions used for diagnosis from 2 different Indian manufacturers revealed significantly lower mean wheal sizes compared to FDA-approved HDM extracts. Taken together, these observations strongly suggest that there might be a high variability in the allergen content of extracts manufactured under Indian guidelines and regulations implemented in other countries. As variability of SPT solutions was shown to significantly impact the outcome of diagnostic tests, standardization of allergenic products from different manufacturers around the world is of utmost importance to avoid misdiagnosis of allergic conditions.

To evaluate the quality of SPT mite solutions available on the Indian markets, we characterized HDM solutions on a molecular basis. SPT solutions and allergen vaccines (AVs) from 3 Indian manufacturers that were labeled with "D. pteronyssinus", "D. farinae", "B. tropicalis", and"H" were analyzed by mass spectrometry, gel electrophoresis, and immunoblots using patients’ sera and allergen-specific monoclonal antibodies. Detailed knowledge on the allergen content and thus potency of SPT solutions and AVs is essential for reliable diagnostics and therapeutic approaches.

Materials and methods

Samples

All samples were purchased by collaborating physicians in India from local vendors and shipped to Europe under temperature-controlled conditions at 4 °C. Upon arrival at the mass spectrometric facilities of the University of Salzburg, all sample vials were originally sealed and intact. Samples were stored at 4 °C according to the manufacturers’ recommendations. The manufacturers were coded as follows:

M1 - Creative Diagnostic Medicare Pvt. Ltd. (Navi Mumbai, India).

M2 - All Cure Pharma Pvt. Ltd. (Bahadurgarh, India).

M3 - Alcit Pvt. Ltd (Delhi, India).

All samples were "ready-to-use" for physicians and contained glycerol and phenol. The samples were coded to indicate the manufacturer and the corresponding allergenic extract (eg, M1-DP, M2-DP, etc.). A full description of the samples is given in Table 1.

Table 1

List of Skin prick test (SPT) solutions and allergy vaccines (AV) analyzed in this study.

Allergen source (according to label)	Type	Manufacturer	Batch Number on vial	Sample code
D. pteronyssinus	SPT	M1	AD0677	M1-DP
D. pteronyssinus	SPT	M2	084143a	M2-DPa
D. pteronyssinus	SPT	M2	084143a	M2-DPb
D. pteronyssinus	SPT	M3	A-080225	M3-DP
D. farinae	SPT	M1	AD0682	M1-DF
D. farinae	SPT	M2	084143a	M2-DF
D. farinae	SPT	M3	080225	M3-DF
House Dust Mite	SPT	M2	042743	M2-HDM
B. tropicalis	SPT	M1	AD0679	M1-BT
D. farinae	AV Dose 4	M1	T0783/0435/2018	M1-DF-AV4
D. pteronyssinus	AV Dose 4	M1	T0784/0435/2018	M1-DP-AV4
House Dust Mite	AV Dose 1	M2	Not availableb	M2-HDM-AV1
House Dust Mite	AV Dose 2	M2	Not availableb	M2-HDM-AV2
House Dust Mite	AV Dose 3	M2	Not availableb	M2-HDM-AV3
House Dust Mite	AV Dose 2	M3	Not availableb	M3-HDM-AV2
House Dust Mite	AV Dose 3	M3	Not availableb	M3-HDM-AV3

These D. farinae and D. pteronyssinus products were labeled by manufacturer 2 with identical batch numbers.

Batch numbers were not given on the vials.

As reference for D. pteronyssinus and D. farinae, authority-approved SPT solutions from Allergopharma (Reinbek, Germany) were used (ref-DP, ref-DF). As reference for B. tropicalis, an in-house produced B. tropicalis extract was used (ref-BT).

Sample preparation

Protein concentrations of SPT solutions were determined with the Pierce 660 nm Protein Assay Reagent (Thermo Fisher Scientific, Rockford, USA). For SDS-PAGE, proteins were precipitated with trichloroacetic acid (TCA). To 100 μL SPT solution, 10 μL 6.1 N TCA were added. After 3 h at −20 °C, precipitated material was centrifuged and washed twice with cold acetone.

For mass spectrometric analyses, SPT solutions were prepared and digested with trypsin using 2 different approaches. TCA-precipitated material from 100 μL SPT solution was resuspended in Extraction buffer 1 of the ProteoExtract All-in-One Trypsin Digestion Kit (Merck Millipore, Billerica, USA) and proteins were reduced, alkylated, and digested using this kit following the manufacturer's instructions. Resulting peptides were desalted with 10 μl ZipTips C18 (Merck Millipore). In a second approach, samples were digested without prior precipitation. 50 μL SPT solution was mixed with 200 μl Extraction buffer 2 of the ProteoExtract All-in-One Trypsin Digestion Kit and 200 μL 100 mM ammonium bicarbonate. Samples were then reduced with dithiothreitol (DTT) at 60 °C for 1 h, alkylated with iodoacetamide for 1 h at room temperature (RT) and digested with MS-grade trypsin at 37 °C (all reagents from Pierce/Thermo Fisher Scientific). After 20 h, detergents present in Extraction buffer 2 were precipitated by adding 500 μL 0.5% trifluoroacetic acid (TFA). After centrifugation, peptides were enriched and desalted using 100 μl Pierce C18 Tips (Thermo Fisher Scientific). Allergy vaccines were digested similarly, except that volumes were adapted to compensate for their low protein concentrations: 500 μL sample, 450 μl Extraction buffer 2, 450 μL 100 mM ammonium bicarbonate.

SDS-PAGE

Dried TCA pellets were resuspended in SDS sample buffer (200 mM Tris, 8% SDS, 0.4% bromophenol blue, 40% glycerol, 400 mM β-mercaptoethanol, pH 6.75) and proteins were separated using a 15% polyacrylamide gel. Sample volumes were adjusted to compensate for the different protein concentrations of the samples. Proteins were visualized with the SilverXpress staining kit (Thermo Fisher Scientific, Waltham, MA, USA).

Immunoblots

Following SDS PAGE and blotting, nitrocellulose membranes (Whatman, Maidstone, UK) were placed into blocking buffer containing 25 mM Tris/Cl, pH 7.5, 150 mM NaCl, 0.5% (v/v) Tween, 0.5% (w/v) BSA, 0.05% (w/v) NaN3 and left shaking for 1 h at RT. Blots were then either incubated with 1:10 diluted serum pools of house dust mite allergic individuals, with different dilutions of mouse monoclonal IgG antibodies (anti-Der f 1 1:1000, anti-Der p 1 1:1000, non-commercial products provided by Allergopharma, Germany) or with blocking buffer alone (for secondary antibody controls) and left shaking overnight at 4 °C. After several washing steps with blocking buffer, bound allergen-specific human IgE or mouse IgG antibodies were detected using AP-conjugated monoclonal anti-human IgE (BD Biosciences, Franklin Lakes, NJ, USA; diluted 1:4000 in blocking buffer) or monoclonal anti-mouse IgG + IgM antibodies (Jackson ImmunoResearch Europe Ltd, Cambridgeshire, United Kingdom; diluted 1:5000 in blocking buffer), respectively. Following 2 h of incubation and additional washing steps with blocking buffer, membranes were equilibrated for 5 min with AP buffer containing 100 mM Tris/Cl, pH 9.5, 100 mM NaCl, 5 mM MgCl2. For detection, a staining solution consisting of 10 mL AP buffer, 30 μL BCIP (stock: 50 mg/ml in dimethylformamide) and 30 μL NBT (stock: 100 mg/ml in 70% dimethylformamide, 30% sterile water) was used. Membranes were washed with distilled water and air-dried.

Mass spectrometry

Desalted peptides resulting from the tryptic digests were separated by reverse-phase nano-HPLC (Dionex Ultimate 3000, Thermo Fisher Scientific, Bremen, Germany), directly coupled via nano electrospray to a Q Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific). The column (Acclaim PepMap RSLC C18, 75 μm × 15 cm, Dionex, Thermo Fisher Scientific) was developed with an acetonitrile gradient (Solvent A: 0.1% (v/v) formic acid; solvent B: 0.1% (v/v) formic acid/90% (v/v) acetonitrile; 5–45% B in 120 min) at a flow rate of 300 nl/min at 55 °C. Capillary voltage of the nano spray was 2.5 kV. Lock mass calibration was used for highest accuracy. Peptide fragmentation/identification was done with a top 12 method and a normalized fragmentation energy at 27%.

Data analysis and protein identification

Survey and fragment spectra were analyzed/sequenced with PEAKS Studio 8.5 and X+, respectively (Bioinformatics Solutions, Waterloo, Canada). Primary searches were done against the complete UniProt database (release April 2019). Based on the results obtained in the primary search, searches were repeated with specific subsets of this database in a multi-round approach. Settings for searches: trypsin (semi-specific), 0.1% FDR for UniProt, −10lgP≥35 for refined searches. Fixed modification: carbamidomethylation (C), variable modifications: deamidation (NQ), oxidation (M).

Data presented here were based on 4 to 6 independent experiments for each sample. In order to compare the results of the different MS analyses, the data were normalized by calculating the sum of the ion count of all fragment spectra with a de novo sequencing score higher than 50% and using this value as normalization factor for individual peptides. Normalized values were then used to determine means and standard errors with Prism 5 (Graphpad, USA).

A flow chart of sample treatment and analysis is shown in Fig. 1.

Fig. 1

Overview of sample preparation and analysis by mass spectrometry (LC-MS/MS) and gel electrophoresis/immunoblot.

Results

Protein concentrations

In general, complex and glycerinated samples with very low protein content prepared from natural sources, such as allergy vaccines (AVs) and skin prick test (SPT) solutions, are difficult to analyze by standard photometric protein determination methods because of the low protein content and the possible interference of non-proteinaceous compounds found in natural extracts. We therefore aimed to determine, whether mass spectrometry (MS) could be used as a primary tool to estimate the protein concentration of such samples. For this purpose, an appropriate amount of sample (up to 1 mL) was digested with trypsin. Resulting peptides were enriched with C18 tips, fragmented by LC-MS/MS and analyzed with a de novo sequence analysis program. Signal intensities of all de novo sequenced peptides with a quality score ≥50 were summarized and compared to values obtained by a classical photometric assay (Supplementary Fig. 1A and B). Although mass spectrometry cannot provide absolute quantification under the chosen experimental conditions, it produced data comparable to the photometric assay, thus providing a useful tool for quality control of the samples. As shown in Supplementary Fig. 1, the Indian SPT solutions contained between 40 and 80 ng/μL total protein, which was comparable to the authority-approved references. Due to their low content, protein concentrations in the AVs could not be determined using photometric assays. Nevertheless, from the intensity values of the de novo sequenced peptides, a protein concentration between 0.05 and 5 ng/μL could be estimated. Determined protein concentrations were in accordance to the doses given by the manufacturers for each AV vial (Supplementary Fig. 1C, Table 1).

SDS-PAGE and immunoblotting

The comparison of the Indian SPT solutions with authority-approved references (D. pteronyssinus, D. farinae) and the in-house reference (B. tropicalis) by SDS-PAGE indicated significant differences between the samples. As shown in Fig. 2, the pattern of protein bands showed no similarity between the 3 manufacturers and the reference samples. These differences were further confirmed by immunoblot analyses using a serum pool obtained from European mite allergic patients and specific monoclonal antibodies (moAB) raised against Der p 1 and Der f 1. Compared to the references, positive immune reactions were weak (D. pteronyssinus) or missing altogether (D. farinae, B. tropicalis and HDM). No signals were observed with the D. pteronyssinus SPT solutions, while the reference showed clear reactivity with anti-Der p 1 antibody. Interestingly, Der f 1 moAB produced a strong signal in DF SPT solutions of manufacturers M2 and M3, although no significant IgE reactivity was observed with patients’ serum pool. A possible explanation for this result could be that conformational IgE-binding epitopes identified in Der p 1 and Der f 1 are partially destroyed under SDS-PAGE/blotting conditions, while binding of moABs may not be affected by such denaturing conditions. To compensate for the low amounts of allergens in the SPT solutions (as determined by mass spectrometry, see below), higher protein amounts of these samples were used for the moAB immunoblots. Therefore, the signals of the Der f 1 moAB were significantly stronger with M2-DF, M3-DT and M2-HDM than with the reference sample. For this reason, gels and immunoblots depicted in Fig. 2 provide only qualitative but not quantitative comparisons.

Fig. 2

SDS-PAGE and immunoblotting of SPT solutions from D. pteronyssinus (A), D. farinae and house dust mite (HDM) (B) and B. tropicalis (C). For optimal detection of allergens in immunoblots, different total protein concentrations of the references and SPT solutions were loaded onto the gels. Therefore, the immunoblots depicted in this figure represent qualitative and not quantitative comparisons.

Mass spectrometry-based proteomics

To further investigate the composition of the Indian SPT and AV products, we performed extensive MS analyses of these samples in comparison to European reference products. SPT solutions were trypsin-digested both after TCA precipitation and in solution without prior precipitation. Due to their low protein content, AVs were digested in solution only. Resulting peptides were analyzed by LC-MS/MS. Quantitative data presented here were based on signal intensities of identified peptides. The mean of 4–6 independent experiments is shown. Special care was taken to unequivocally distinguish between proteins of closely related species, like D. pteronyssinus and D. farinae. To achieve this goal and avoid false-positive hits, database searches were performed with the multi-round option of PEAKS: for the initial search, the complete UniProt library was used. This first round provided an overview of proteins present and gave a first hint on possible contaminations. The following rounds were based on the preliminary results from the first pass, using sequence subsets of single organisms only. After each search round, peptides with positive matches were excluded from the following searches. The remaining non-matched peptides were then searched against a new sequence set, and so forth. Applying this search method, the identification of unique/diagnostic peptides was greatly facilitated, and hence the identification of false-positive hits was avoided. As example, the differentiation of Der p 2 and Der f 2 in the samples ref-DP and M3-DP is shown (Supplementary Fig. 2). Both proteins share 128 of 146 amino acids (88% identity). In the initial search with the complete database, both Der p 2 and Der f 2 were listed as positive hits. After the multi-round search with D. farinae sequences in the first and D. pteronyssinus sequences in the second round, on the other hand, it became evident that sample M3-DP did not contain Der p 2, but only Der f 2, as 5 peptides were matched to Der f 2, but no peptides to Der p 2. In the reference sample, on the other hand, only Der p 2, but no Der f 2 specific peptides were identified, indicating the exclusive presence of Der p 2. These results are summarized in Supplemental Fig. 2. Detailed results for all samples obtained with this method are described below.

D.pteronyssinus samples

Mass spectrometric analyses showed that the content of all commercial D. pteronyssinus-labeled samples did not match their description and confirmed the results obtained by SDS-PAGE and immunoblot experiments. No D. pteronyssinus proteins were detected in the SPT solutions of manufacturers M1, M2 and M3, and in the AV of manufacturer M1 (Fig. 3). Instead, samples M1-DP, M2-DPa, M2-DPb and M1-DP-AV4 contained proteins from the storage mite Suidasia medanensis (S. medanensis), the legume Cicer arietinum (C. arietinum, chickpea) and the baker's yeast Saccharomyces cerevisiae (S. cerevisiae). For instance, 255 S. medanensis peptides (assigned 12 proteins out of 16 UniProt entries) were detected in sample M1-DP. Also, 2177 C. arietinum peptides (assigned to 330 proteins) and 1497 yeast peptides (assigned to 486 proteins) were identified with high confidence in this sample. In this context, it should be mentioned that both S. medanensis and C. arietinum are allergen sources. Allergome lists 10 potential allergens from S. medanensis, 8 of which were identified in M1-DP. Similarly, 6 potential allergens are listed for C. arietinum, 3 of which were found in this D. pteronyssinus sample. Results from M2-DPa, M2-DPb and M1-DP-AV4 were similar. In M3-DP, no S. medanensis, C. arietinum and yeast proteins were detected. Instead, this sample contained D. farinae proteins (226 peptides assigned to 73 proteins) and a substantial amount of human serum proteins (2391 peptides, 641 proteins). In contrast, the authority-approved SPT solution (ref-DP) contained exclusively D. pteronyssinus proteins (1254 peptides assigned to 57 proteins). Fig. 3 shows the quantitative comparison of proteins/allergens identified in the samples and reference. A detailed analysis of all allergens detected in these samples is given in Supplemental Fig. 3. Allergens Der p 26, 28, 29, 30, 31, and 37 were also identified in the reference, but not included in this figure, as their protein sequences are not listed in public databases.

Fig. 3

Allergen and total protein content of the D. pteronyssinus products compared to the reference. Panel A shows signal intensities of peptides obtained in multi-round database searches. Panel B shows the relative quantitative comparison of the allergen content as heat map. Red indicates high allergen content, green indicated low/no content of the indicated allergen. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

D.farinae samples

Similar to the D. pteronyssinus samples of manufacturer M1, the SPT solution M1-DF and allergy vaccine M1-DF-AV4 contained S. medanensis, C. arietinum and yeast proteins/allergens No D. farinae proteins could be detected (Fig. 4). For manufacturers M2 and M3, both M2-DF and M3-DF samples contained mostly human serum proteins and a limited number of D. farinae proteins (Fig. 4). In sample M2-DF, 251 D. farinae peptides (39 proteins) and in M3-DF, 224 D. farinae peptides (43 proteins) were identified. This is significantly less than in the reference ref-DF, where 1302 D. farinae peptides (assigned to 83 proteins) were identified. The number and relative amounts of D. farinae allergens was also reduced in these products, when compared to the reference (Supplementary Fig. 4). Whereas the reference contained a total of 30 D. farinae allergens, in M2-DF and M3-DF only 15 allergens could be identified. In M2-DF and M3-DF, both Der f 1 and Der f 2 were present, but only in reduced amounts (Der f 1: 16% and 41% of the reference, respectively, Der f 2: 1% and 0.7% of the reference, respectively) (see summarizing Fig. 7). The following allergens were not detectable in M2-DF and M3-DF, although present in the reference: Der f 4, 5, 7, 8, 15, 16, 20, 26, 28, 31, 32 and 33. Der f 18 and 21 were present in M3-DF, but missing in M2-DF. On the other hand, Der f 25, 29 and 34 were present in M2-DF, but missing in M3-DF. It is interesting to note that the reference sample contained a significant amount of non-allergenic D. farina proteins, mostly common household enzymes. As a result, the intensity ratio of total Der f proteins to Der f allergens was close to 2:1. In samples M2-DF and M3-DF, on the other hand, non-allergenic proteins are nearly completely absent (data not shown), resulting in a total Protein to allergen ratio of 1:1 (Fig. 4A). Human serum proteins constituted the majority of proteins in samples M2-DF (1542 peptides, 479 proteins/protein families) and M3-DF (1775 peptides, 668 proteins/protein families). Abundant human proteins identified in the D. farinae samples are listed in Supplementary Table 1.

Fig. 4

Allergen and total protein content of the D. farinae products compared to the reference. For details, see the legend of Fig. 3.

Fig. 7

Direct quantitative comparison of selected important allergens in SPT-solutions and allergy vaccines with the corresponding references. Signal intensities are shown in linear scale.

B.tropicalis sample

Only 1 B. tropicalis SPT solution was available for this study. As shown in Fig. 5 and Supplementary Fig. 5, M1-BT did not contain any identifiable B. tropicalis proteins/allergens. The constituents of this sample, similar to the D. pteronyssinus and D. farinae products of this manufacturer, were proteins from S. medanensis (236 peptides, assigned to 13 proteins), C. arietinum (358 peptides, 124 proteins) and baker's yeast (287 peptides, 106 proteins). In addition, nearly 8000 peptides without a sequence equivalent in UniProt were detected in M1-BT. This might be due to the fact that only a limited number of sequences of S. medanensis and C. arietinum proteins are presently available. A total of 7 S. medanensis allergens were identified. Cic a 1 was the only identified chickpea allergen, albeit in minute amounts. By contrast, 44 B. tropicalis proteins (1401 peptides) and 11 B. tropicalis allergens (988 peptides) were identified in ref-BT.

Fig. 5

Allergen and total protein content of the B. tropicalis sample compared to the reference. For details, see the legend of Fig. 3.

House dust mites (HDM) samples

"House dust mites" generally refers to mites of the species D. pteronyssinus, D. farinae, B. tropicalis, and Euroglyphus maynei. Consequently, refined database searches concentrated on proteins of these mites. In all 6 samples labeled "House Dust Mites" (M2-HDM, M2-HDM-AV1, M2-HDM-AV2, M2-HDM-AV3, M3-HDM-AV2, M3-HDM-AV3), no D. pteronyssinus, B. tropicalis or E. maynei proteins were identified. The only mite proteins detected were from D. farinae (Fig. 6). As with the D. farinae samples, the allergen composition of all samples significantly differed from the reference, Der f 1 and Der f 3 being the only allergens present in all samples. Other allergens (e.g. Der f 4, 5, 7, 8) were not detectable (Fig. 6B and Supplementary Fig. 6). The major constituents of all commercial HDM SPT solutions and AVs were human serum proteins (Fig. 6A).

Fig. 6

Protein and allergen content of the house dust mite (HDM) products (A) and comparison to the D. pteronyssinus, D. farinae and B. tropicalis references (B). For details, see the legend of Fig. 3.

Fig. 7 summarizes the main results of this study for selected mite allergens. The amounts of 3 important D. pteronyssinus, D. farinae, and B. tropicalis allergens detected in the corresponding samples are shown and compared to the references. Der p 1, Der p 2 and Der p 5 were not detected in any of the samples, just as Blo t 1, Blo t 2, Blo t 5 and Der f 5. All these allergens were present in the references. Der f 2 was present, albeit in very small amounts, in 2 D. farinae and 1 HDM SPT solution, as well as in 3 HDM AVs. Although in lower amounts than in the reference, Der f 1 was present in 2 D. farinae SPT solutions (M2-DF, M3-DF) and in 2 HDM AVs (M2-HDM-AV3, M3-HDM-AV3).

Discussion

Allergic sensitization to HDM allergens is a major risk factor for allergic rhinitis and asthma,, with more than 50% of asthmatic patients being allergic to either D. pteronyssinus or D. farinae, or to both species. Thus, accurate diagnosis based on information about exposure to allergen sources combined with appropriate tests to confirm sensitization is critical for the management and therapy of HDM allergy. However, studies focusing on HDM prevalence and quality of diagnostic tools reported wide variability on the allergen composition of commercial HDM extracts used for in vivo diagnosis through SPT.,, The lack of comparability between allergen extracts, particularly for HDM, is a well-known problem. Variability of these diagnostic/therapeutic products seems to be linked to: (i) cultivation conditions employed for different mite species, and (ii) materials used for the extract preparation (eg, mite bodies, fecal particles, mite bodies plus fecal particles) and their degree of purity.30, 31, 32 Together, these 2 factors tremendously impact the protein composition and concentrations of the extracts. Further, insufficient standardization of these extracts, which in most cases is based solely on “in-house assays”, likely accounts for the large differences among allergenic products from various manufacturers.

In our study, we uncovered additional problems with HDM SPT solutions and AVs commercialized by 3 different Indian manufacturers, namely a discrepancy between the products' labels and their contents. None of the 5 D. pteronyssinus-labeled products contained D. pteronyssinus proteins. Instead, 1 sample contained D. farinae and human serum proteins, and the remaining 4 products contained allergens from the storage mite S. medanensis and from chickpea, as well as proteins from baker's yeast. From 4 D. farinae-labeled products, 2 contained human serum proteins plus a limited number of D. farinae allergens and the other 2 contained only S. medanensis, chickpea and yeast proteins. The B. tropicalis sample contained no B. tropicalis allergens at all, but consisted exclusively of S. medanensis, chickpea and yeast proteins. All 6 HDM-labeled samples consisted of human serum proteins and limited amounts of D. farinae allergens. These alarming findings demonstrate that the challenge in standardization and quality control of allergenic products should not be overlooked and underestimated.

Although it is certainly difficult to explain these results, it could be speculated that the divergent contents of these products could have originated from the production of the mite source material. In general, the production of HDM extracts involves cultivation and growth of mites in large quantities under specified and controlled conditions using appropriate culture media. In the past, HDM were cultured using human skin scales, which are their natural food source. Later on, alternative culture media consisting of animal dander, dog food, rodent chow, or fungal cultures were developed. Presently, manufacturers use a mix of different food sources (eg, soybean powder, wheat germ) or yeast supplemented with vitamins or amino acids to culture mites. Thus, it is conceivable that yeast and chickpeas were used in the cultivation media for the HDM products analyzed here. Independent of the type of culture media used for the mite cultivation, the FDA sets that foreign material should not exceed 1% of the raw preparation, whereas the European Pharmacopoeia indicates that foreign species should not be present in the raw material. For approval of allergen products, the Indian Pharmacopoeia Commission (IPC) indicates that: (i) contamination by foreign species/strains should be avoided, (ii) the composition of culture media must be justified, and (iii) purification procedures should be designed to minimize the presence of potential irritants and non-allergenic components. However, the IPC Guidelines are very general and do not give precise specifications for allergen products. The complete absence of D. pteronyssinus proteins (allergenic and non-allergenic) indicates that the respective SPT and AV products contained only culture media components contaminated by other mite species such as D. farinae and Suidasia medanensis. Because mites belonging to the genus Suidasia are common in house dust and extensive IgE cross reactivity with D. farinae and B. tropicalis have been reported, the D. pteronyssinus-labeled products would not identify genuinely D. pteronyssinus-sensitized patients. Similarly, the B. tropicalis-labeled products would fail to diagnose B. tropicalis-sensitization and would give false positive tests for patients sensitized to chickpeas and/or S. medanensis.

In the process of manufacturing allergen extracts from raw materials, it is considered appropriate to include measures to minimize allergen degradation (eg, extraction time, temperature, pH) and to maximize stability (eg, aqueous formulation containing 50% glycerol or 0.01–0.05% human serum albumin; inclusion of preservatives such as phenol). Of note, the inclusion of human or animal products in the culture medium or extracts needs to be justified and appropriate measures (eg, γ-radiation) should be employed to ensure that they do not contain any potential pathogens. Our data revealed that both D. pteronyssinus- and D. farinae-labeled products contained either no mite proteins or very low amounts of D. farinae allergens. Instead, large amounts of human serum proteins made up the bulk of the proteinaceous material in these samples. It remains unclear whether human proteins were used in the culture media or as stabilizer in the extract formulation. In addition to serum albumin, we detected high amounts of several other proteins found in human serum (eg, immunoglobulins, S100 proteins, transferrin, lactoferrin, complement, macroglobulin) supporting the notion that human serum was included in the manufacturing process. In contrast, European SPT products used as reference in this study covered the respective allergen spectra and lacked contaminations.

Conclusions

The value of extract-based mite products for diagnosis and therapy depends on several factors, including the selection of the mite species, appropriate cultivation, and the manufacturing procedures used to process the raw material and to prepare the mite extracts. Thus, characterization and standardization of such extracts is the difficult but inescapable task for diagnosis and therapy of allergic diseases caused by mites. Our investigation demonstrated the value of mass spectrometry to fingerprint complex allergen extracts. The use of non-standardized diagnostic and therapeutic products, such as the ones analyzed here, could lead not only to misdiagnosis of HDM allergy, due to contamination by allergens from other sources, but also their application may be even harmful to patients, as some of these products contain large amounts of proteins of human origin.

Ethics approval

The use of dust mite patients’ sera was approved by the ethics committee of the University of Vienna (Ethics permission number EK1263_2014 to BB).

Authors contributions

SH: Made substantial contributions to design of experiments, acquisition of data and drafting the manuscript.

GG: Made substantial contributions to the design of experiments and drafting the manuscript.

BB: Made substantial contributions to the design of experiments.

FF: Made substantial contributions to the design of experiments and drafting the manuscript.

PB: Made substantial contributions to the design of experiments, acquisition and analysis of proteomic data and drafting the manuscript.

Availability of data

Mass spectrometry proteomics data of one representative dataset per sample have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD021383 and 10.6019/PXD021383.

Funding

This study was partially supported by a research grant of to FF and PB.

Confirmation of unpublished work

The authors declare that this manuscript is original, has not been published before, is not currently being considered for publication elsewhere, and has not been posted to a preprint server.

Consent for publication

All authors agreed to the publication of the work.

Declaration of competing interest

S. Huber reports personal fees from Allergopharma GmbH & Co. KG during the conduct of the study. F. Ferreira is a member of Scientific Advisory Boards (HAL Allergy, NL; AllergenOnline, USA) and has been supported by the Austrian Science Fund (FWF). The rest of the authors declare that they have no relevant conflict of interest in relation to the methods or materials employed in this study.