Integrating image analysis and quantitative modeling for a holistic view of GPCR ligand binding dynamics

Abstract

Oletatavasti on tulnud kõnekäänd “mõru pilli alla neelama” ravimite kõige tuntumast kõrvaltoimest, mõrust maitsest. Paljudel ravimitel on aga tõsisemad kõrvaltoimed ning tervele hulgale haigustele polegi ravi. Erinevad teadusharud panustavad ravimite väljatöötamisse ja olemasolevate edasiarendamisse. Ravimi ülesanne on muuta retseptorkontrollitud protsesse organismis, et tekiks raviv või sümptomeid leevendav efekt. Üks olulisemaid ravimite märklaudu on G-valk seotud retseptorid. Inimeses on ligi 800 erinevat G-valk seotud retseptorit, mis reguleerivad erinevaid funktsioone nagu nägemine ja südame toimimine. Senini ei ole suudetud retseptorsüsteemide kõiki eripärasid lahti muukida. Doktoritöös arendati välja spektroskoopial ja mikroskoopial põhinevad katsesüsteemid, et jälgida ravimimolekulide retseptoritele seostumist nii lipiidsetes nanoosakestes kui ka elusates rakkudes. Näiteks õnnestus täieliku sisepeegeldusmikroskoobiga jälgida üksikute fluorestseeruvate ravimimolekulide seostumist retseptorile nanoosakeste pinnal. Kuna meetodid toodavad sadade gigabaitide kaupa andmeid, loodi Tarkvara Aparecium, mis koos sügavõppemudelitega suudab sellest andmehulgast olulise info välja sõeluda. Katsesüsteemide abil loodi kineetilised mudeleid, mis suudavad dünaamiliselt kirjeldada ja ennustada ravimainete retseptorile seostumist. Näiteks leiti, et muskariinsed M2 atsetüülkoliini retseptorid paiknevad membraanides enamasti paarikaupa ning pärast ravimimolekuli kinnitumisel ühele retseptorile lukustab teine ravimimolekul esimese kinni. Selle teadmise abil saaks disainida pikema toimeajaga või hoopis uute omadustega ravimeid. Edaspidi võiks mudeleid ja meetodeid kasutades läbi sõeluda suure hulga molekule ja leida need, millel on need huvitavad omadused olemas.Presumably, the idiom "to swallow a bitter pill" has emerged from the fact that many medicines have a bitter taste. However, many drugs have more serious side effects, and there is no cure for a number of diseases. Various scientific disciplines contribute to the development of drugs and the improvement of existing ones. The purpose of a drug is to modify receptor-controlled processes in the body to achieve a therapeutic or symptom-relieving effect. One of the most important targets of drugs are thr G-protein-coupled receptors. In humans, there are nearly 800 different G-protein-coupled receptors that regulate various functions such as vision and heart function. So far, it has not been possible to fully decipher all the intricacies of receptor systems. In this doctoral thesis, experimental systems based on spectroscopy and microscopy were developed to monitor the binding of drug molecules to receptors in lipid nanoparticles and living cells. For example, using total internal reflection microscopy, the binding of individual fluorescent drug molecules to receptors on the surface of nanoparticles could be observed. Since these methods produce vast amounts of data, the Aparecium Software, together with deep learning models, was created to extract essential information from this data set. By using the data from experimental systems, kinetic models were developed that can dynamically describe and predict the binding of drugs to receptors. For instance, it was discovered that muscarinic M2 acetylcholine receptors are mostly located in the membranes as dimers, and once one drug molecule binds to one receptor monomer, it locks the other drug molecule onto the first one. With this knowledge, it would be possible to design drugs with longer therapeutic effect or design drugs with entirely new properties. In the future, the developed models and methods could be used to screen a large numbers of molecules and identify those that possess these interesting properties.