Summary of Scientific Work, Research Results and Publications
He started his medicinal chemistry research as a member of the Scientific Student Association under the supervision of Prof. Gyula Schneider in the field of synthesis of novel steroid derivatives at József Attila University of Sciences.
Prof. Schneider supervised his diploma work, entitled “Investigation of Region- and Stereoselective Reactions of Steroids”. The highly recognized professors of the department at that time (Prof. Mihály Bartók, Prof. Gábor Bernáth, Prof. Árpád Molnár) and the dedicated PhD students (Prof. Ferenc Fülöp (later academic member of HAS), Prof. György Dombi) strongly motivated him to learn and practice organic chemistry of bioactive compounds and medicinal chemistry.
After his graduation in 1981, Prof. Miklós Nádasy, his former lecturer in pesticide chemistry, offered him an assistant research fellow position at the Research Institute for Heavy Industrial Chemistry in Veszprém. There, Dr. András Vas, former PhD student of Prof. Schneider, was his supervisor in the research of novel heterocyclic pesticides. Due to his university training in medicinal chemistry, he soon realized that he was more interested in drug research than in the preparation of highly toxic materials.
He then found an open position at Semmelweis University of Medicine, Department of Pharmaceutical Chemistry. The head of the department and dean of the faculty, Prof. György Szász, accepted his application, and he was able to start his medicinal chemist career in his research group in 1982, supporting the structure–property–activity investigations of the department.
Original Drug Research
The department worked in industrial cooperation with the CHINOIN pharmaceutical company on novel pyrido-pyrimidines (PPs). After consulting with Prof. Zoltán Mészáros and Prof. István Hermecz (heads of research at CHINOIN), he was tasked with preparing analogues of the PPs in order to develop potential cAMP phosphodiesterase IV inhibitors for the therapy of asthma, COPD and allergy.
His supervisor in the laboratory was József Kökösi, known for performing the simplest synthesis of rutaecarpine. Connecting to this polycondensed ring system synthesis his original task was to develop tricyclic quinazoline analogues of the pyrido-pyrimidine hit molecules, and model substances for the university structure–property–relationship (SPR) studies. He prepared series of simpler “ring-opened” model compounds as well for comparison them with the tricycles.
Since the tricyclic compounds were novel and patentable, to protect their IP he prepared his university doctoral (Dr. Mat. Pharm.) thesis from the gas chromatographic investigation and SPR of the ring-opened compounds. He received his doctoral degree, summa cum laude, in 1985 for the thesis “Synthesis and Investigation of Theophylline Analogues”.
This achievement defined his long-term strategic aim: the design of drug-like, potentially active heterocyclic compounds; the development of novel synthetic methods; and the investigation of structure–property–biological activity relationships.
Combinatorial Chemistry
The main profile of the Department of Pharmaceutical Chemistry was qualitative and quantitative drug analysis, so the instrumental background for structure identification was poor. The compounds and intermediates synthesized in the lab were monitored only by thin layer chromatography and were validated by NMR occasionally at CHINOIN. As a consequence of multistep “blind syntheses”, incorrect or mixed products were often obtained. Therefore, simple, “guaranteed” reactions with high yields had to be chosen for reliable compound synthesis.
He applied many classical organic analytical reactions previously taught in the students’ laboratory courses. The only in-house instrumental method was UV spectrophotometry, which fortunately proved useful for detecting the ring closure of quinazolines. Investigating the reactivity of quinazolines, he and his colleagues realized that, besides tricyclic molecules, the ring-opened 2-alkyl compounds could behave as active methylene compounds. This recognition opened the way to aldehyde-condensed (styryl), diazo-coupled (hydrazone) and brominated derivatives, which could be further functionalized, for example through nucleophilic substitution reactions.
In the 1980s, big pharmaceutical companies started building their high-throughput screening (HTS) capacities, creating a high demand for screening libraries. While testing and developing expert softwares at Compudrug Ltd., he used graphical databases containing thousands of compound structures. This led to the idea that tens of thousands of novel structures could be generated by software based on programmed reaction schemes, and that molecules with desired properties could be selected by virtual screening. The next question was whether “real” compounds could also be generated in this way. Relying on his accumulated knowledge in quinazoline chemistry, he concluded that the answer was yes.
He tested the idea of parallel synthesis in the laboratory, and it worked. The owner of Compudrug Ltd., Dr. Ferenc Darvas, commissioned the synthesis of hundreds of compounds per month, and this cooperation continued successfully for six years. Based on these experiences, Dr. Darvas founded Comgenex Inc. (later AMRI Hungary), where the first compound libraries—unsurprisingly—consisted largely of (130K) quinazoline derivatives.
After six years of work in “combinatorial chemistry,” he recognized that finding biologically active hits in randomly synthesized compound libraries would rarely be efficient. Later, big pharmaceutical companies came to the same conclusion, and the high hope combinatorial chemistry bubble gradually declined.
Rational Drug Design and Kinase Inhibitors
Among the many parallel approaches to drug discovery, he chose a biological target–based strategy. Close analogues of heterocycles he had developed, began to appear in the scientific literature as potential anticancer agents. Around the same time, reports were published that protein kinases could serve as key targets in cancer therapy.
He contacted Dr. György Kéri, head of the Peptide Biochemistry Research Group and advisor to SUGEN Inc. (USA). Dr. Kéri’s group worked on the synthesis of small peptides as potential kinase inhibitors, tested in the United States. Initially, Dr. Kéri was not enthusiastic about the proposed heterocycles, since his group aimed to develop peptide-based kinase inhibitors—which were ultimately unsuccessful.
He explained that his quinazolines were built from anthranilic acids (α-amino acids), meaning that the heterocycles could be considered “peptidomimetics”. Oxindoles (core structure of sunitinib/Sutent) could have been interpreted similarly. Dr. Kéri agreed to test 50 compounds. Based on the literature, he had constructed a qualitative structure–activity relationship and selected the test set from his existing compound series. The hit rate was astonishing: 25 compounds showed activity in the kinase assays (around 50% hit rate versus the typical hit rate of >0.01% in random libraries).
These excellent biological results directed his research toward kinase inhibitors with potential applications in the treatment of cancers, bacterial and viral infections, inflammations (e.g. neurogenic inflammation), diabetes and, in general, any proliferative disease.
Due to this success, Gerald McMahon, CEO of SUGEN, visited the group in Hungary in 1994. He disclosed his quinazoline structures and had a long debate with McMahon, who claimed that quinazolines were ATP analogues and therefore could never become drugs (a view echoed later by Prof. István Hermecz academic, who called them “the kiss of death”). They were proven wrong.
The most active compound, coded ŐL-57, a 4-anilinoquinazoline, showed outstanding activity in EGFR kinase assays. It was later identified independently by others and became a widely used reference compound under the code PD-153035 in EGFR assays worldwide. The compound was published by Fry et al. in Science (1994). Follow-up molecules of this quinazoline series including Iressa and Tarceva, which were launched on the pharmaceutical market soon afterwards; today, many other quinazolines are in clinical stages or in development pipelines.
In 1995–1996, he worked as a visiting scientist in the bioanalytical research group of Prof. Cynthia Larive at the University of Kansas, where he performed gradient-enhanced high-resolution NMR spectroscopy. The collaboration resulted in three high impact publications, one of which highlighted the applicability of NMR for qualitative and quantitative characterization of compound libraries.
Meanwhile, one of his quinazolines prepared was published in Cancer Research (Cancer Research 56 (15), 3540-3545, 1996.) and has since become his most cited paper (more than 450 citations).
During 1998 and 1999 he worked in the combinatorial chemistry laboratory of Dr. Richard Houghten in San Diego, within the framework of a US–Hungarian TÉT cooperation. There, he investigated the simplest solid-phase synthesis approach, the “tea-bag” method, combined with microwave activation (using a household microwave oven). He developed a rapid, high-yield (almost quantitative) synthetic method for preparing quinazoline-4-ones, key intermediates of blockbuster drugs such as Viagra, Tarceva and Iressa and their follow-ups. Using his method, the intermediate of Iressa could have been obtained in almost quantitative yield, while the best published method at that time reported only about 20% yield.
He offered this procedure to several pharmaceutical companies without success, and subsequently published it in Current Medicinal Chemistry. Among his synthetic papers, this is his most cited first-authored publication.
His “Rational Drug Design and Development Laboratory” research concept (which failed to receive OTKA grant support four times) was later included in the grant application of the Cooperative Research Centre of Semmelweis University. The application was funded, and with combined governmental and industrial support, the Rational Drug Design Laboratory was established on the second floor of the Hospital of Hungarian Railways.
The aim of this laboratory was to train undergraduate and PhD students in modern synthetic, biochemical and biological methods directly through hands-on laboratory work. He directed the medicinal chemistry research of the Centre from 2002 to 2010. The Rational Drug Design Laboratory produced numerous publications and patents, diploma theses and PhD dissertations. The PhD graduates trained there have become outstanding scientists in Hungary and abroad (at Sanofi, Richter Gedeon, Servier, Institut Pasteur, Institute of Cancer Research (UK), Max Planck Institutes, Cold Spring Harbor Laboratory, KU Leuven, Utrecht University, etc.).
Computational Methods and Expert Systems
His medicinal chemistry work has been connected with computational methods from the very beginning. He started working with personal computers in 1984. In his university doctoral thesis, he performed curve fitting by a hand-hold computer and established acceptable structure–property relationships. From 1986 onwards, particularly with the start of the Compudrug collaboration, he routinely applied computers in both teaching and research.
Starting from basic calculations (molecular weight, pKa, binding constants, curve fitting, equilibria, etc.), even commercially available computers can now be used for molecular modelling as well. Current expert systems can predict physicochemical properties (logP, pKa, metabolic stability, NMR spectra, chemical shifts and coupling constants, stable conformations, etc.) as well as biological activity spectra of virtual compounds that medicinal chemists could only dream of in the 1980s.
Over the past 40 years, he has participated in the development and beta testing of many “knowledge base” expert systems developed by Hungarian and international companies, and he personally knew most of the top developers in this field. These systems included Compudrug’s Pallas (pKalc, ProLogP, Metabolexpert, Eluex), Tripos SYBYL (Prof. Garland Marshall), Dragon (Roberto Todeschini), ACD/Labs Bundle, PASS (Vladimir Poroikov), among others.
He contributed to the development of a ligand based QSAR system having high predictability of biological and physical-chemical properties based on calculated molecular structural properties. The algorithm of 3DNET (later 3DNET4W) was protected by US and EU patent applications, in cooperation with István Kövesdi, who implemented the artificial neural network and the other mathematics tools into the algorithm. This self-learning algorithm was capable of identifying quantitative structure–activity relationships and showed high predictive power, making it a competitor to virtual screening by computational docking. After computational filtering of compound sets by their software, the successful hit rate increased by an order of magnitude. The software is still in use at universities and in pharmaceutical companies.
Several of his prominent students have incorporated modern computational methods into their work. Notably, Dr. Csaba Szántai-Kis and Dr. Ferenc Baska included computational approaches in their theses, and his PhD student, Dr. Marcell Krekó, works on the synthesis of covalently binding kinase inhibitors and has developed a unique computational model for covalent binding, which has already been published.
Recognition and Citations
His research group, composed of graduate and PhD students, was classified as one of the excellent research groups of Semmelweis University in the booklet “Academic Excellence in Biomedical Research” (2008, ISBN 978-963-9879-09-6).
His detailed scientific publication list provides a more comprehensive view of his drug research work. The main results of his research on kinase inhibitors are summarized in a paper listed first in the Hungarian publication list:
Őrfi L. Kinázgátló kismolekulák fejlesztése (Bruckner termi előadás). Magyar Kémikusok Lapja 71: 7–8, pp. 229–235 (2016).
He is proud to be listed in the Excellence Database of the Central Library of Semmelweis University (https://lib.semmelweis.hu/app/excellence?resID=1279).
Based on his validated (MTMT) citation data, he was regularly placed on the Top 100 list of most cited scientists of Semmelweis University (e.g. 83rd at the University and 2nd in the Faculty in 2018; 79th at the University and 3rd in the Faculty in 2017; 65–66th at the University and 3rd in the Faculty in 2016; and 4th in the Faculty among the ranked 50 scientists in 2015).