Primerjava tridimenzionalnih proteinskih struktur alergena breze Bet v 1 in navzkrižno reaktivnih alergenov iz hrane s pomočjo izbranih bioinformacijskih orodij

Mlakar, Maša (avtor), Tavčar, Eva (mentor), Vidak, Marko (komentor)

Navzkrižna reaktivnost alergenov, pri kateri sekundarni alergen veže specifična protitelesa IgE organizma senzibiliziranega na primarni alergen, predstavlja klinično pomemben pojav, zlasti pri sindromu navzkrižne alergije med pelodom in hrano. Čeprav Svetovna zdravstvena organizacija (SZO) za napovedovanje takšne reaktivnosti priporoča analizo lokalne poravnave primarnih zaporedij, je nizka sekvenčna podobnost nekaterih klinično potrjenih navzkrižno reaktivnih alergenov (npr. Act d 11, Vig r 6) nakazala na pomen terciarne strukture. Zanimalo nas je, ali bi primerjava 3D struktur pokazala višjo stopnjo strukturne podobnosti. Prav tako smo želeli preveriti kakovost računalniško napovedanih struktur, saj je znanih bistveno več primarnih zaporedij proteinov kot tridimenzionalnih struktur, ki so eksperimentalno določene. V raziskavi smo s tremi izooblikami glavnega alergena breze (Bet v 1.0101, 1.0106, 1.0107) kot referenco analizirali 81 navzkrižno reaktivnih prehranskih alergenov. Z orodjem COMPASS smo izvedli poravnavo celotnih primarnih zaporedij, nato pa z DALI primerjali njihove eksperimentalno določene (PDBe) in napovedane (AlphaFold) terciarne strukture. Rezultati so pokazali: a. Pri vseh analiziranih alergenih je bilo ujemanje 3D strukture višje kot ujemanje primarne strukture. Povprečna razlika je znašala 30,0 odstotnih točk pri napovedanih in 22,6 odstotnih točk pri eksperimentalno določenih strukturah. b. Znatno višje ujemanje tridimenzionalne strukture smo opazili tudi pri alergenih z zelo nizkim ujemanjem zaporedja. Alergen Act d 11 (ujemanje zaporedja 22,4 %) je izkazal 63,9 % ujemanje pri napovedani in 59,7 % pri eksperimentalno določeni strukturi. c. Ugotovili smo tudi, da je AlphaFold različno dobro napovedal 3D strukture (pri 62,5% alergenov je Q >0,8, vendar so primerjave z eksperimentalnimi strukturami razkrile odstopanja (npr. Act d 8.0101: Q=0,57)). Na podlagi rezultatov sklepamo, da bi vključitev bioinformacijske analize 3D struktur v smernice SZO lahko izboljšala zanesljivost napovedi navzkrižne reaktivnosti med alergeni, kljub omejitvam napovednih orodij.

• Jezik: Slovenski jezik
• Ključne besede: navzkrižna reaktivnost, bioinformatika, tridimenzionalna struktura, Bet v 1, AlphaFold, DALI
• Vrsta gradiva: Magistrsko delo/naloga
• Organizacija: FFA - Fakulteta za farmacijo
• Leto izida: 2025
• PID: 20.500.12556/RUL-171420
• Datum objave v RUL: 26.08.2025

Sekundarni jezik

• Jezik: Angleški jezik
• Naslov: Comparison of the three-dimensional protein structures of the Bet v 1 birch allergen and cross-reactive food allergens using selected bioinformatics tools

Izvleček

Allergen cross-reactivity, in which a secondary allergen binds specific IgE antibodies of an organism sensitized to a primary allergen, represents a clinically important phenomenon, especially in pollen-food allergy syndrome. Although the World Health Organization (WHO) recommends predicting such reactivity based on local alignment of primary sequences, the low sequence identity of some clinically confirmed cross-reactive allergens (e.g. Act d 11, Vir r 6) has indicated the importance of tertiary structure. We aimed to investigate whether the comparison of 3D structures would reveal a higher degree of structural similarity. We also aimed to assess the quality of computationally predicted structures, as the number of known protein sequences is significantly higher than that of three-dimensional structures, which are experimentally determined. In this study we, using three isoforms of the major birch pollen allergen (Bet v 1.0101, 1.0106, 1.0107) as references, analyzed 81 cross-reactive food allergens. We used the COMPASS tool to perform sequence alignment of whole primary sequence, followed by DALI comparison of their experimentally determined (PDBe) and predicted (AlphaFold) tertiary structures. The results showed: a. In all analyzed allergens 3D structural similarity was higher than sequence identity. The average difference was 30.0 percentage points for predicted and 22.6 for experimentally determined structures. b. Significantly higher 3D structural similarity was observed even in allergens with very low sequence identity. Allergen Act d 11 (22.4% sequence identity) exhibited 63.9% similarity for the predicted and 59.7% for the experimental structure. c. We also found that AlphaFold predicted 3D structures with varying accuracy (62.5 % of allergens had Q>0.8, however comparison with experimental structures revealed discrepancies (e.g. Act d 8.0101: Q=0.57)). Based on these results, we conclude that incorporating 3D structure-based bioinformatic analysis into WHO guidelines could enhance the reliability of predicting allergen cross-reactivity, despite the limitations of current prediction tools.

• Ključne besede: cross-reactivity, bioinformatics, three-dimensional structure, Bet v 1, AlphaFold, DALI