20.500.12556/RUL-186181 Uporaba molekulskega modeliranja za analizo vezave derivatov peptida GSYWYQVWF na podrazrede človeških IgG Use of molecular modelling for binding analysis of GSYWYQVWF peptide derivatives to human IgG subtypes Imunoglobulini G (IgG) so najpogostejši razred protiteles v človeškem telesu in imajo osrednjo vlogo pri humoralnem imunskem odzivu. V medicini se uporabljajo kot učinkovine za zdravljenje rakavih, avtoimunskih, infekcijskih in drugih bolezni. Njihova regija Fc, sestavljena iz domen CH2 in CH3, opravlja efektorsko funkcijo ter je odgovorna za vezavo efektorskih receptorjev in molekul komplementa. Hkrati ta predel predstavlja vezavno mesto za bakterijska Ig-vezavna proteina, stafilokokni protein A (SpA) in streptokokni protein G, ki ju tradicionalno uporabljamo pri afinitetni kromatografiji za čiščenje in izolacijo protiteles. Zaradi tveganja za izpiranje ligandov, imunogenosti in visokih proizvodnih stroškov proteinskih afinitetnih ligandov so kot alternativo razvili krajše in stabilnejše peptidne ligande. Mednje spadajo peptid min19Fc (GSYWYQVWF) in njegovi derivati, ki so jih pridobili s tehnologijo predstavitve na bakteriofagu. Ker je za poglobljeno razumevanje osnov njihove vezave na regijo Fc IgG potreben vpogled na atomski ravni, vse pogosteje uporabljamo metode molekulskega modeliranja. To so napovedovanje strukture proteinov z modeli umetne inteligence (UI; npr. AlphaFold), molekulsko sidranje in simulacije molekulske dinamike (MD). Te metode omogočajo vpogled v vezavne interakcije na nivoju, ki jih z eksperimentalnimi metodami težko dosežemo. V sklopu magistrske naloge smo z računalniškimi pristopi molekulskega modeliranja proučevali vezavo peptida min19Fc in njegovih derivatov na podrazrede človeškega IgG (IgG1–IgG4). Vezavno mesto in vezavne konformacije peptidov smo napovedali z modeloma UI AlphaFold3 in AlphaFold2-Multimer v3, z molekulskim sidranjem s postopkom Peptide Docking v uporabniškem vmesniku Maestro ter jih dodatno ovrednotili s simulacijami MD v orodju Desmond. Rezultate smo primerjali z eksperimentalnimi podatki iz predhodnih raziskav in z eksperimentalno določenimi strukturami sorodnih ligandov regije Fc. Na podlagi rezultatov modelov UI smo pokazali, da se peptidi vežejo na stik domen CH2 in CH3 fragmenta Fc, pri čemer se njihovo vezavno mesto delno prekriva z vezavnim mestom SpA. Z molekulskim sidranjem in MD smo določili ključne interakcije ter minimalni vezavni motiv. Z analizo substitucij izbranih aminokislin smo pokazali, da substitucija V7F poveča vezavno afiniteto, a zmanjša specifičnost vezave, medtem ko negativna kontrola s substitucijo W4F potrjuje pomen aromatskega motiva YWY. Peptid min19Fc izkazuje najvišjo afiniteto do IgG1, sledijo IgG3, IgG2 in IgG4, kar sovpada z eksperimentalnimi podatki. Pridobljena spoznanja predstavljajo izhodišče za nadaljnji racionalni razvoj afinitetnih peptidnih ligandov z izboljšano afiniteto in selektivnostjo. Immunoglobulins G (IgG) are the most abundant class of antibodies in the human body and play a central role in the humoral immune response. In medicine, they are used as therapeutic agents for the treatment of cancer, autoimmune, infectious, and other diseases. Their Fc region, composed of the CH2 and CH3 domains, mediates effector functions and is responsible for binding effector receptors and complement molecules. At the same time, this region serves as a binding site for bacterial Ig-binding proteins, staphylococcal Protein A (SpA) and streptococcal Protein G, which are traditionally used in affinity chromatography for the purification and isolation of antibodies. Due to the risk of ligand leaching, its immunogenicity, and the high production costs of protein-based affinity ligands, shorter and more stable peptide ligands have been developed as an alternative. These include the peptide min19Fc (GSYWYQVWF) and its derivatives, which were obtained using phage display technology. Since an in-depth understanding of the basis of their binding to the Fc region of IgG requires insight at the atomic level, molecular modelling methods are increasingly used. These include protein structure prediction with artificial intelligence (AI) models (e.g., AlphaFold), molecular docking, and molecular dynamics (MD) simulations. These methods provide insight into binding interactions at a level that is difficult to achieve with experimental methods alone. In this master's thesis, we used molecular modelling approaches to investigate the binding of the peptide min19Fc and its derivatives to human IgG subclasses (IgG1–IgG4). The binding site and binding conformations of the peptides were predicted using co-folding in AlphaFold3 and AlphaFold2-Multimer v3, molecular docking using the Peptide Docking workflow in Maestro, and further evaluated through MD simulations in Desmond. The results were compared with prior experimental data and with experimentally determined structures of related Fc-region ligands. Based on the results obtained using the AI models, we showed that the peptides bind at the hinge between the CH2 and CH3 domains of the Fc fragment, where their binding site partially overlaps with that of SpA. Through molecular docking and MD, we determined the key interactions and the minimal binding motif. Substitution analysis showed that the V7F substitution increases binding affinity but reduces binding specificity, while the negative control with the W4F substitution confirms the importance of the aromatic YWY motif. The peptide min19Fc exhibits the highest affinity for IgG1, followed by IgG3, IgG2, and IgG4, which is consistent with experimental data. These findings provide a starting point for the further rational development of affinity peptide ligands with improved affinity and selectivity. IgG molekulsko sidranje molekulska dinamika peptidni ligandi vezavna konformacija binding conformation IgG molecular docking molecular dynamics peptide ligand true false false [T. Meterc] Slovenski jezik Angleški jezik Magistrsko delo/naloga 2026-08-28 08:46:03 2026-08-28 08:46:17 2026-09-02 08:25:30 0000-00-00 00:00:00 2026 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 1970-01-01 577.27:57.087(043.2) 127708 289550851 33001.pdf 33001.pdf 1 AC6B478D0B1FE76F874E13081D9938DE 42f0c6a4e7536e5a5642a40fd1dd740de3d8cabdaa40e98f13f7d6b7e1d377fd c3c50e8e-a2ab-11f1-9b0d-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=245294 Fakulteta za farmacijo 0 0 0