In Silico Identification of Potential American Cockroach (Periplaneta americana) Allergens.

BACKGROUND: Cockroaches have been recognized as a powerful indoor allergen. Cockroach allergy can be a major factor in serious asthma and nasal allergy. Bioinformatics tools have been developed to identify potential allergens. The present study was conducted to identify potential allergens in Periplaneta americana (Linnaeus).
METHODS: The study focused on the identification of potential allergens among the characterized proteins of P. americana using web-based and publicly available allergen prediction tools that follow the FAO/WHO guidelines for prediction of allergenic proteins. P. americana protein sequences were retrieved from UniProtKB. The sequences obtained were analyzed using AlgPred. The potential allergens obtained were further analyzed by SDAP for confirmation.
RESULTS: Protein sequences (233 cases) of P. americana were obtained from UniProtKB out of which 25 were known allergens. Out of the remaining 208 proteins, 102 potential allergens were predicted by AlgPred. However, only 9 were found to be potential allergens after screening with SDAP. Arginine kinases, RNA polymerase II subunit, parcxpwnx02, peptidylprolyl cis-trans isomerase, hemocyanin subunit type I and type II, homologue of Sarcophaga proteinase and alpha amylase were confirmed to be potential allergens by SDAP.
CONCLUSION: We have identified nine potential allergens in P. americana that may be used as preliminary support for further laboratory experiments.

Introduction

Allergic disorders are among the most common chronic diseases in the developed world and an increasing problem in developing countries (1). Allergy is caused by the exaggerated and harmful response of the immune system to otherwise harmless substances known as allergens. There is overwhelming evidence that indoor domestic allergens play a key role in allergic disease. The primary arthropod allergens associated with allergic disease are house dust mites and cockroaches (2, 3).

Cockroach allergens are one of the strongest risk factors predictive of allergic sensitization and asthma morbidity (4, 5). Cockroach extracts including cast skins, egg shells, and fecal material (6) have been shown to contain several major and minor allergens (7–10). Among the 3,500 known species of cockroaches, the American cockroach (Periplaneta americana) is a frequently encountered cockroach in homes. Allergens with masses ranging from 6 to 120 kDa from P. americana have been identified by various immunochemical techniques (11) and the functional importance of some of these have been determined. Two prominent proteins of 78 and 72 kDa in Per a 3 have been reported to cause T cell proliferation in cockroach allergic patients (12). More recent data indicate that indoor insect allergens, including those of cockroaches, are potent inducers of IL-5 and eotaxin-mediated esophageal eosinophilia (13).

Despite efforts of researchers, our current knowledge about P. americana allergens and their cross-reactivity is still insufficient, at least, partly due to the difficulty involved in purifying cockroach allergens from extracts in significant quantities for detailed characterization. According to Universal Protein Database (UniprotKB) resource, to-date a total of 233 proteins have been identified in the P. Americana of which only 25 are known reported allergens, which suggests that many allergens may still lie unidentified.

The use of bioinformatics tools is becoming increasingly helpful for initially screening of compounds based on existing experimentally validated databases. Algpred is a web server for prediction of allergenic proteins and for mapping IgE epitopes on allergenic proteins with high accuracy (14). SDAP (Structural Database of Allergenic Proteins) is another web server containing information on allergens and can develop correlations that can be used to predict allergenicity of novel proteins and cross-reactivity between allergens (15).

The present study is focused on the identification of potential allergens using these allergen databases and bioinformatics.

Materials and Methods

Amino acid sequences of 233 proteins belonging to P. americana in UniprotKB database were retrieved. Out of these, 25 were known reported allergens. For the remaining 208 proteins, allergenicity test was conducted using AlgPred. The FASTA sequences of proteins were given individually to the server, which presented the results on the basis of the scanning of IgE epitopes, motif-based approach, SVM-based method using amino acid composition of protein, hybrid approach and BLAST search on allergen representative proteins ARPs (14).

The potential allergens obtained through AlgPred screening were further analyzed using SDAP (15). Searches were first carried out according to the FAO/WHO criterion for allergenicity prediction. Proteins that shared more than 35% sequence similarity with an allergen (on a segment of 80 residues) or an identity of at least six contiguous amino acids were screened out. These allergens were further analyzed using the property distance function (PD) method. A protein that passed the initial step and gave a PD score of less than 10 was considered to be potential allergen.

Results

Following screening of the 208 P. americana proteins with AlgPred (excluding very short peptides which the AlgPred was unable to screen), a total of 102 proteins were considered potentially allergenic. Further analysis of these 102 proteins with any of the two FAO/WHO criterion of SDAP indicated that 93 proteins did not fulfill both criteria (Table 1). According to the PD scores obtained 9 proteins were predicted to be potential allergens (Table 2). Alpha-amylase (Fragment), parcxpwnx02, homologue of sarcophaga 26, 29 kDa proteinase, RNA polymerase II largest subunit (fragment), and hemocyanin subunit type I and II were predicted to be potentially significant allergens as they had a PD score of less than 10, whereas the arginine kinases and peptidylprolyl cis-trans isomerase (fragment) passed all the screening steps and showed to have PD scores of less then 3 which indicated their being highly significant potential allergens.

Table 1:

Proteins showing potential allergenicity with only AlgPred analysis

No.	Protein name	Accession Number
1.	Diuretic Hormone	P41538
2.	Corazonin	P11496
3.	Bursicon (Fragments)	P84118
4.	Pyrokinin-6	P82693
5.	Peptide Hormone 4	P82697
6.	Peptide Hormone 3	P82696
7.	Peptide Hormone 2	P82695
8.	Peptide Hormone 1	P82694
9.	Hemolymph Lipopolysacharide	P26305
10.	Trehalase Inhibitor	P19986
11.	Troponin T	Q9XZ71
12.	Sulfakinin 1	P36885
13.	Periviscerokinin-2	P81555
14.	Periviscerokinin 2.2	P84422
15.	Ubiquitin	A1E2I6
16.	Rab5-GTP binding protein	B6ZLL4
17.	Notch protein (Fragment)	B6EBG3
18.	Delta protein	B6EBG2
19.	Odorant-binding protein	B6E9U9
20.	Odorant-binding protein (Fragment)	B6E9U8
21.	Odorant-binding protein	B6E9U7
22.	Cxpwmw03	Q2LIY7
23.	Cxpwmw01	Q1PS51
24.	MRNA, clone: 1. (Fragment)	P92056
25.	MRNA, clone: 2. (Fragment)	P92055
26.	MRNA, clone: 3. (Fragment)	P92054
27.	MRNA, clone: 4. (Fragment)	P92053
28.	MRNA, clone: 5. (Fragment)	P92052
29.	Lectin-related protein (Fragment)	P92050
30.	Lectin-related protein	P92049
31.	Lectin-related protein (Fragment)	P92048
32.	Lectin-related protein (Fragment)	P92047
33.	26-kDa lectin (Fragment)	O76155
34.	Rsp60	O76153
35.	P10	O17447
36.	Elongation factor 1-alpha (Fragment)	O02460
37.	DNA-directed RNA polymerase (Fragment)	O02459
38.	Dynamin (Fragment)	D0UTF4
39.	Methionine aminopeptidase (Fragment)	D0UTA5
40.	Putative uncharacterized protein (Fragment)	D0UT69
41.	AMP deaminase (Fragment)	D0UT23
42.	Pyrimidine biosynthesis (Fragment)	D0USX9
43.	Proteasome subunit (Fragment)	D0USM1
44.	F-box protein (Fragment)	D0US62
45.	Glu-+ pro-tRNA synthetase (Fragment)	D0URW2
46.	Glycogen synthase (Fragment)	D0UR99
47.	Gelsolin (Fragment)	D0UR43
48.	ATP synthase (Fragment)	D0UQZ3
49.	Acetylglucosaminyl-transferase (Fragment)	D0UQT6
50.	Clathrin heavy chain (Fragment)	D0UQ69
51.	Clathrin heavy chain (Fragment)	D0UQ18
52.	Glucosamine phosphate isomerase (Fragment)	D0UPW5
53.	GTP-binding protein (Fragment)	D0UPR4
54.	Syntaxin (Fragment)	D0UPE9
55.	Spliceosome-associated protein (Fragment)	D0UPA1
56.	Signal recognition particle (Fragment)	D0UP45
57.	Pre-mRNA splicing factor (Fragment)	D0UNV4
58.	Alpha-spectrin (Fragment)	D0UND9
59.	Alpha-spectrin (Fragment)	D0UNA0
60.	Alpha-spectrin (Fragment)	D0UN65
61.	Acetyl-CoA carboxylase (Fragment)	D0UN29
62.	domain binding protein (Fragment)	D0UMV2
63.	ATP synthase (Fragment)	D0UM59
64.	polymerase subunit 2 (Fragment)	D0ULU7
65.	RNA helicase (Fragment)	D0ULN9
66.	Protein kinase (Fragment)	D0ULI4
67.	Histone deacetylase (Fragment)	D0ULD3
68.	Gln-tRNA synthetase (Fragment)	D0UL28
69.	Arg methyltransferase (Fragment)	D0UK69
70.	Regenectin	Q9Y098
71.	NADP-dependent isocitrate dehydrogenase (Fragment)	Q9XY39
72.	Putative transcription factor	Q9U0S0
73.	10 kDa LEG regeneration protein (Fragment)	Q9TWV5
74.	Beta-1,4-glucanase 1 (Fragment)	Q9NCF3
75.	Beta-1,4-glucanase 2 (Fragment)	Q9NCF2
76.	40S ribosomal protein S12 (Fragment)	Q8MTJ6
77.	Rab11 (Fragment)	Q8MTJ5
78.	Vitellogenin receptor	Q8MP02
79.	Large conductance calcium activated potassium channel pSlo spliceform 1B (Fragment)	Q8I9V0
80.	Large conductance calcium activated potassium channel pSlo spliceform 4C (Fragment)	Q8I9U6
81.	Large conductance calcium activated potassium channel pSlo spliceform 5B (Fragment)	Q8I9U5
82.	Ryanodine receptor pRyR (Fragment)	Q86LC1
83.	Elongation factor-2 (Fragment)	Q6JU91
84.	RNA polymerase II largest subunit (Fragment)	Q6JU05
85.	TRPgamma cation channel	Q5YJT9
86.	Parcxpwfx01	Q5MBV8
87.	Parcxpwfx02	Q5MBV7
88.	Parcxpwnx01	Q5MBV6
89.	Parcxpwnx03	Q5G5C3
90.	Parcxpwnx04	Q5G5C1
91.	Adipokinetic hormone preproprotein	Q5EY02
92.	Putative uncharacterized protein (Fragment)	Q5EY00
93.	Rieske Fe-S protein (Fragment)	Q5EXZ9

Table 2:

Potential Allergens Predicted by AlgPred and SDAP in Periplaneta Americana

No.	Protein name	Accession Number	AlgPred Analysis	SDAP Analysis
				FAO/WHO Criteria
				Stretches of 6 contiguous amino acid identical to an allergen*	% identity with an allergen over a window of 80 a.as	PD score
1.	Arginine kinase [Periplaneta americana]	D3JUE7	Predicted allergen by SVM-based method and BLAST approach	Present	93.75% with Bomb m1.0101 from a.a number 263 to 342	< 3
2.	Alpha-amylase (Fragment)	D2YVM9	Predicted allergen by SVM-based method	Present	61.25% with Blo t 4.0101 from a.a number 35 to 114	< 10
3.	Homologue of Sarcophaga 26,29kDa proteinase	Q9U914	Predicted allergen by SVM-based method	Present	50.00% with Act c 1 from a.a number 466 to 545	> 3 and < 10
4.	Peptidyl-prolyl cis-trans isomerase (Fragment)	Q9U8K2	Predicted allergen by SVM-based method and BLAST approach	Present	77.50% with Mala s 6	< 3
5.	Parcxpwnx02	Q5MBV5	Predicted allergen by IgE mapping, BLAST and Hybrid approach	Present	42.50% with Act c 1 from a.a number 257 to 336	> 3 to < 10
6.	RNA polymerase II largest subunit (Fragment)	Q5EXZ8	Predicted allergen by SVM-based method	Present	41.25% with Der f 15 from a.a number 16 to 95	> 3 and < 10
7.	Hemocyanin subunit type II	B9W4N8	Predicted allergen by SVM-based method	Present	48.75% with Per a 3.0201 from a.a number 112 to 191	> 3 and < 10
8.	Hemocyanin subunit type I	B9W4N7	Predicted allergen by SVM-based method, BLAST and hybrid approach	Present	65.00% with Per a 3.0201 from a.a number 100 to 179	< 3
9.	Arginine kinase	A1KY39	Predicted allergen by SVM-based method	Present	93.75% with Plo i 1 from a.a number 257 to 336	< 3

present – indicates that there are stretches of 6 contiguous amino acid identical to a known allergen

Of all the predicted allergens, Parcxpwnx02 was positive with IgE mapping searches, and was found to contain an IgE epitope starting from the position 307 of the protein with the sequence LANSWNYDWGDNGY.

Discussion

In silico protein analysis is a well-established technique for assessment of allergenicity and immunological cross-reactivity (16, 17).

We screened 208 P. americana proteins using AlgPred and SDAP bioinformatics tools. Highly significant prediction score (PD<3) were obtained for two arginine kinases (Accession nos. D3JUE7 and A1KY39). Interestingly, another arginine kinase in P. americana (Accession no. B1A7S7) has been documented to be an important allergen in the Thai population (18) that correlates well with our highly significant scores in our study for this enzyme. Arginine kinase isomers have been reported in Caenorhabditis elegans and it has been suggested that tissue restricted expression of isoforms in this family evolved early (19). In addition, highly significant homology (93.75%) was seen between the arginine kinase of P. americana and that of Plodia interpunctella (Indian-meal moth) that also acts as a powerful allergen (16). It is most likely that the two arginine kinases reported in this study are allergens.

The protein, peptidyl-prolyl cis-trans isomerase (accession no. Q9U8K2) also showed significant prediction score (PD<3). Peptidylprolyl cis-trans isomerase belongs to the cyclophilin-type PPIase family and shows significant homology (77.5%) with a cyclophilin allergen from the yeast Malassezia sympodialis (Mala s 6) (20). Cyclophilins constitute a family of proteins involved in many important cellular functions. They have also been identified as a pan-allergen family able to elicit IgE-mediated hypersensitivity reactions (21, 22).

Homologue of Sarcophaga 26 and 29 kDa proteinase (accession no. Q9U914) and parcxpwnx02 (accession no. Q5MBV5) were predicted as potential allergens due to their peptidase property. Both these proteins showed significant identity, 50% and 42.5% respectively, with allergen Act c 1 (or actinidin), a cysteine protease and also a major allergen in kiwi fruit. This 30 kDa acidic protein is present in kiwi fruit in several isoforms that differ in the PI value (23). Homologue of Sarcophaga 26 and 29kDa proteinase appears to eliminate foreign proteins in this insect and is conserved in a wide variety of insects and participates in their defense mechanism (24). German cockroach proteases have been known to play a role as allergens, participating in cleavage of matrix metalloproteinase (MMP-9) thereby remodeling airway passage (25).

Hemocyanin subunit type I and II proteins show 48.75 and 65% identity respectively, with the known cockroach allergen of Per a 3-family. Per a 3 belongs to the most potent allergens (26). This result also becomes noteworthy if we consider that certain other proteins with hemocyanin domains are allergens (27, 28).

Another protein RNA polymerase II (accession no. Q5EXZ8) has shown 41.25% identity with a Der f 15 that is a major canine high molecular weight allergen (29). Alpha-amylase (fragment) (accession no. D2YVM9) was also found to be a potential allergen. Fungal α-amylase is a known dust allergen that is commonly found in bakery, in particularly wheat or flour, products (30). There is a possibility that the α-amylase protein of P. americana is also a dust allergen associated with the cockroach species. Cockroaches are found in flour and it is highly possible that the dust allergen, α-amylase, is transferred from the flour to the cockroaches.

In conclusion, in silico studies are valuable tools for predicting potential proteins should be given priority in allergen research. We have identified 9 proteins of the P. americana that are potential allergens and warrant further studies in this area.

Ethical considerations

Ethical issues (Including plagiarism, Informed Consent, misconduct, data fabrication and/or falsification, double publication and/or submission, redundancy, etc) have been completely observed by the authors.