Bence Szalai

Predicting human olfactory perception from chemical features of odor molecules

How will this molecule smell? We still do not understand what a given substance will smell like. Keller et al. launched an international crowd-sourced competition in which many teams tried to solve how the smell of a molecule will be perceived by humans. The teams were given access to a database of responses from subjects who had sniffed a large number of molecules and been asked to rate each smell across a range of different qualities. The teams were also given a comprehensive list of the physical and chemical features of the molecules smelled. The teams produced algorithms to predict the correspondence between the quality of each smell and a given molecule. The best models that emerged from this challenge could accurately predict how a new molecule would smell. Science, this issue p. 820 Results of a crowdsourcing competition show that it is possible to accurately predict and reverse-engineer the smell of a molecule. It is still not possible to predict whether a given molecule will have a perceived odor or what olfactory percept it will produce. We therefore organized the crowd-sourced DREAM Olfaction Prediction Challenge. Using a large olfactory psychophysical data set, teams developed machine-learning algorithms to predict sensory attributes of molecules based on their chemoinformatic features. The resulting models accurately predicted odor intensity and pleasantness and also successfully predicted 8 among 19 rated semantic descriptors (“garlic,” “fish,” “sweet,” “fruit,” “burnt,” “spices,” “flower,” and “sour”). Regularized linear models performed nearly as well as random forest–based ones, with a predictive accuracy that closely approaches a key theoretical limit. These models help to predict the perceptual qualities of virtually any molecule with high accuracy and also reverse-engineer the smell of a molecule.

Allosteric interactions within the AT1 angiotensin receptor homodimer: Role of the conserved DRY motif

G protein coupled receptor (GPCR) dimerization has a remarkable impact on the diversity of receptor signaling. Allosteric communication between the protomers of the dimer can alter ligand binding, receptor conformation and interactions with different effector proteins. In this study we investigated the allosteric interactions between wild type and mutant protomers of type 1 angiotensin receptor (AT₁R) dimers transiently expressed in CHO cells. In our experimental setup, one protomer of the dimer was selectively stimulated and the β-arrestin2 binding and conformation alteration of the other protomer was followed. The interaction between β-arrestin2 and the non-stimulated protomer was monitored through a bioluminescence resonance energy transfer (BRET) based method. To measure the conformational alterations in the non-stimulated protomer directly, we also used a BRET based intramolecular receptor biosensor, which was created by inserting yellow fluorescent protein (YFP) into the 3rd intracellular loop of AT₁R and fusing Renilla luciferase (RLuc) to its C terminal region. We have detected β-arrestin2 binding, and altered conformation of the non-stimulated protomer. The cooperative ligand binding of the receptor homodimer was also observed by radioligand dissociation experiments. Mutation of the conserved DRY sequence in the activated protomer, which is also required for G protein activation, abolished all the observed allosteric effects. These data suggest that allosteric interactions in the homodimers of AT₁R significantly affect the function of the non-stimulated protomer, and the conserved DRY motif has a crucial role in these interactions.

Mechanisms of angiotensin II-mediated regulation of aldosterone synthase expression in H295R human adrenocortical and rat adrenal glomerulosa cells

In adrenal zona glomerulosa cells angiotensin II (Ang II) is a key regulator of steroidogenesis. Our purpose was to compare the mechanisms of Ang II-induced changes in the expression level of early transcription factors NR4A1 (NGFIB) and NR4A2 (Nurr1) genes, and the CYP11B2 gene encoding aldosterone synthase in H295R human adrenocortical tumor cells and in primary rat adrenal glomerulosa cells. Real-time PCR studies have demonstrated that Ang II increased the expression levels of NR4A1 and NR4A2 in H295R cells within 1 h after stimulation, which persisted up to 6 h; whereas in rat adrenal glomerulosa cells the kinetics of the expression of these genes were more rapid and transient. Ang II also induced prolonged nuclear translocation of Nurr1 and NGFIB proteins in both cell types. Studies using MEK inhibitor (PD98059, 20 microM), protein kinase C inhibitor (BIM1, 3 microM) and calmodulin kinase (CAMK) inhibitor (KN93, 10 microM) revealed that in rat adrenal glomerulosa cells CAMK-mediated mechanisms play a predominant role in the regulation of CYP11B2. In accordance with earlier findings, in H295R cells MEK inhibition increased the expression of NR4A1, NR4A2 and CYP11B2 genes, however, it decreased the Ang II-induced gene expression levels, suggesting that ERK activation has a role in control of expression of these genes. No such mechanism was detected in rat glomerulosa cells. Sar1-Ile4-Ile8-AngII, which can cause G protein-independent ERK activation, also stimulated the expression of CYP11B2 in H295R cells. These data suggest that the previously reported CAMK-mediated stimulation of early transcription factors NGFIB and Nurr1 has a predominant role in Ang II-induced CYP11B2 activation in rat adrenal glomerulosa cells, whereas in H295R cells ERK activation and G protein-independent mechanisms also contribute to this process.

Improved Methodical Approach for Quantitative BRET Analysis of G Protein Coupled Receptor Dimerization

G Protein Coupled Receptors (GPCR) can form dimers or higher ordered oligomers, the process of which can remarkably influence the physiological and pharmacological function of these receptors. Quantitative Bioluminescence Resonance Energy Transfer (qBRET) measurements are the gold standards to prove the direct physical interaction between the protomers of presumed GPCR dimers. For the correct interpretation of these experiments, the expression of the energy donor Renilla luciferase labeled receptor has to be maintained constant, which is hard to achieve in expression systems. To analyze the effects of non-constant donor expression on qBRET curves, we performed Monte Carlo simulations. Our results show that the decrease of donor expression can lead to saturation qBRET curves even if the interaction between donor and acceptor labeled receptors is non-specific leading to false interpretation of the dimerization state. We suggest here a new approach to the analysis of qBRET data, when the BRET ratio is plotted as a function of the acceptor labeled receptor expression at various donor receptor expression levels. With this method, we were able to distinguish between dimerization and non-specific interaction when the results of classical qBRET experiments were ambiguous. The simulation results were confirmed experimentally using rapamycin inducible heterodimerization system. We used this new method to investigate the dimerization of various GPCRs, and our data have confirmed the homodimerization of V2 vasopressin and CaSR calcium sensing receptors, whereas our data argue against the heterodimerization of these receptors with other studied GPCRs, including type I and II angiotensin, β2 adrenergic and CB1 cannabinoid receptors.

Angiotensin type 1A receptor regulates β-arrestin binding of the β2-adrenergic receptor via heterodimerization

Heterodimerization between angiotensin type 1A receptor (AT1R) and β2-adrenergic receptor (β2AR) has been shown to modulate G protein-mediated effects of these receptors. Activation of G protein-coupled receptors (GPCRs) leads to β-arrestin binding, desensitization, internalization and G protein-independent signaling of GPCRs. Our aim was to study the effect of heterodimerization on β-arrestin coupling. We found that β-arrestin binding of β2AR is affected by activation of AT1Rs. Costimulation with angiotensin II and isoproterenol markedly enhanced the interaction between β2AR and β-arrestins, by prolonging the lifespan of β2AR-induced β-arrestin2 clusters at the plasma membrane. While candesartan, a conventional AT1R antagonist, had no effect on the β-arrestin2 binding to β2AR, TRV120023, a β-arrestin biased agonist, enhanced the interaction. These findings reveal a new crosstalk mechanism between AT1R and β2AR, and suggest that enhanced β-arrestin2 binding to β2AR can contribute to the pharmacological effects of biased AT1R agonists.

Community assessment of cancer drug combination screens identifies strategies for synergy prediction

In the last decade advances in genomics, uptake of targeted therapies, and the advent of personalized treatments have fueled a dramatic change in cancer care. However, the effectiveness of most targeted therapies is short lived, as tumors evolve and develop resistance. Combinations of drugs offer the potential to overcome resistance. The space of possible combinations is vast, and significant advances are required to effectively find optimal treatment regimens tailored to a patient’s tumor. DREAM and AstraZeneca hosted a Challenge open to the scientific community aimed at computational prediction of synergistic drug combinations and predictive biomarkers associated to these combinations. We released a data set comprising ~11,500 experimentally tested drug combinations, coupled to deep molecular characterization of the respective 85 cancer cell lines. Among 150 submitted approaches, those that incorporated prior knowledge of putative drug targets showed superior performance predicting drug synergy across independent data. Genomic features of best-performing models revealed putative mechanisms of drug synergy for multiple drugs in combination with PI3K/AKT pathway inhibitors.Abstract The effectiveness of most cancer targeted therapies is short lived since tumors evolve and develop resistance. Combinations of drugs offer the potential to overcome resistance, however the number of possible combinations is vast necessitating data-driven approaches to find optimal treatments tailored to a patient’s tumor. AstraZeneca carried out 11,576 experiments on 910 drug combinations across 85 cancer cell lines, recapitulating in vivo response profiles. These data, the largest openly available screen, were hosted by DREAM alongside deep molecular characterization from the Sanger Institute for a Challenge to computationally predict synergistic drug pairs and associated biomarkers. 160 teams participated to provide the most comprehensive methodological development and subsequent benchmarking to date. Winning methods incorporated prior knowledge of putative drug target interactions. For >60% of drug combinations synergy was reproducibly predicted with an accuracy matching biological replicate experiments, however 20% of drug combinations were poorly predicted by all methods. Genomic rationale for synergy predictions were identified, including antagonism unique to combined PIK3CB/D inhibition with the ADAM17 inhibitor where synergy is seen with other PI3K pathway inhibitors. All data, methods and code are freely available as a resource to the community.

Reverse-engineering human olfactory perception from chemical features of odor molecules

Despite 25 years of progress in understanding the molecular mechanisms of olfaction, it is still not possible to predict whether a given molecule will have a perceived odor, or what olfactory percept it will produce. To address this stimulus-percept problem for olfaction, we organized the crowd-sourced DREAM Olfaction Prediction Challenge. Working from a large olfactory psychophysical dataset, teams developed machine learning algorithms to predict sensory attributes of molecules based on their chemoinformatic features. The resulting models predicted odor intensity and pleasantness with high accuracy, and also successfully predicted eight semantic descriptors (“garlic”, “fish”, “sweet”, “fruit”, “burnt”, “spices”, “flower”, “sour”). Regularized linear models performed nearly as well as random-forest-based approaches, with a predictive accuracy that closely approaches a key theoretical limit. The models presented here make it possible to predict the perceptual qualities of virtually any molecule with an impressive degree of accuracy to reverse-engineer the smell of a molecule. One Sentence Summary Results of a crowdsourcing competition show that it is possible to accurately predict and reverse-engineer the smell of a molecule.

Transactivation within the AT1 angiotensin receptor homodimer: the role of the conserved DRY motif

Methods In this study we examined the possible functional consequence of transactivation within angiotensin AT1 receptor homodimers. To do this, we used the S109Y mutant of the AT1 receptor, in which the binding site of the nonpeptid AT1 receptor antagonist, candesartan, was destroyed. Expressing this mutant receptor together with the wildtype in CHO cells, in the presence of candesartan, we could examine the interaction between them by stimulating the S109Y mutant receptor, and following the activation of the wild-type receptor. To monitor the signaling of the receptor two parameters were measured: the conformational change of the receptor using an intramolecular sensor, and the binding of β-arrestin-2 to the activated receptors. In both cases the highly sensitive method of bioluminescence resonance energy transfer (BRET) was applied. Additionally, we performed a radioactive ligand binding assay to examine the cooperativity between receptor monomers.