Final Report Summary - FATOKUNEUFP7IIF2010 (Blending biophysical and drug discovery platforms to investigate allosterism in G-protein-coupled receptors (GPCRs) and find novel allosteric modulators for neurotherapeutics development)
Description of project objectives
This Marie Curie IIF project sought to develop a number of high-quality fluorescent probes (ligands) to visualise and quantify interactions between ligands and their receptors in order to explore novel pharmacology such as allosterism, which has shown promise as a novel strategy for eliciting neuroprotection, with relevance for neurological and neurodegenerative conditions. These probes were then to be characterised using state-of-the-art pharmacological and imaging approaches and those with superior qualities were to be employed for downstream applications, including investigations of allosterism and allosteric interactions and drug screening. Focus was on three G-protein-coupled receptors (GPCRs): adenosine receptors, metabotropic glutamate receptors and the GABAB receptor, as their modulation is known to be of significant relevance to neurodegeneration.
Account of experimental work performed
We devoted considerable effort to the characterisation of the candidate fluorescent ligands, which were synthesised by our collaborating spin-out company, CellAura Technologies. Studies were conducted in cell cultures engineered to stably or transiently express the receptors of interest, with appropriate control cultures lacking such expression. As these fluorescent ligands are very expensive to design, we first optimised our characterisation assays for robustness and reproducibility using the 5-HT receptor as a prototype, for which we have a range of stable cell lines expressing at least five of its subtypes. Appropriate functional assays were used to assess the quality of nine (9) different fluorescent ligands that are predicted to be agonists or antagonists at the different subtypes of the 5-HT receptor. Of all these, one showed exceptional promise in targeting the 5-HT1A receptor as an antagonist with high affinity and we advanced it to the level of imaging to determine if it also had excellent photophysical properties to make it an ideal fluorescent ligand for probing this receptor subtype using fluorescence-based (imaging) technologies. Interestingly, this ligand also proved excellent for imaging purposes. In fact, we succeeded in using it to develop a fluorescence binding assay for the receptor, which worked extremely well and holds the promise of replacing the use of radioligands in competitive binding experiments.
When we considered the assays in 5-HT receptor-expressing cell lines to be sufficiently robust and reproducible, we turned to the GABAB receptor and sought to characterise fluorescent ligands targeting it. This endeavour, however, presented the greatest of challenges, largely due to the complexity of this receptor. Unlike the 5-HT receptor subtypes, for which it was possible to engineer stable cells for characterisation and imaging assays, the GABAB receptor could only be expressed through transient transfection of its cDNA into cell lines but even so, the transfection mix had to be made up of separate cDNAs of the GABAB1 and the GABAB2 sub-units, in an appropriate stoichiometric ratio (this had to be determined beforehand), as both subunits are required to express a functional GABAB receptor. Upon overcoming the several hurdles, we were eventually able to test two putative fluorescent ligands for this receptor, one of which was a red version while the other was a green version. Both ligands showed antagonist property in functional assays, but it was the green ligand that had corresponding photophysical properties additionally required to approve of such probes as reasonably useful for imaging-based downstream applications. We also tested the applicability of the green ligand in fluorescence binding assays.
Towards the end of the project, we began to explore the binding kinetics of the two characterised fluorescent ligands targeting the 5-HT1A receptor and the GABAB receptor, respectively. This was done using a confocal microscope (LSM 710) coupled to a perfusion system that allows assessment of binding kinetics in single, living cells, a very promising application for the fluorescent compounds, in addition to several other uses to which they can now be put.
Brief scientific description of main results
The most significant result is the identification of two excellent fluorescent probes with superior properties for use in several applications that involve both the qualitative and quantitative assessment of ligand-receptor interactions at the 5-HT1A receptor subtype and the GABAB receptor.
Fluorescent ligand for the 5-HT1A receptor
CA200992, a BODIPY630/650-conjugated derivative of NAN-190 (NAN190-BY630), was shown to be a fluorescent antagonist ligand with high affinity at, and selectivity for, the 5-HT1A receptor. In Chinese Hamster Ovary (CHO) cells stably expressing both this Gi-coupled receptor and a cyclic adenosine monophosphate (cAMP) response element-secreted placental alkaline phosphatase (CRE-SPAP) reporter gene for measurement of changes in adenylate cyclase activity, CA200992 at increasing concentrations (10-8, 10-7, 10-6M) shifted agonist concentration-response curve (agonist-induced decrease in SPAP levels) to the right in a parallel fashion, indicative of competitive antagonism, with a resultant pKB value of 8.49 ± 0.14 which was replicated in a subsequent fluorescence binding assay. In contrast, this fluorescent ligand had no effect on the agonist concentration-response curve in CHO cell lines similarly engineered to express a closely-related 5-HT1B receptor, also Gi-coupled. The other three 5-HT receptor subtypes tested were the Gq-coupled 5-HT2A, 5-HT2B and 5-HT2C receptors, also stably expressed in CHO cells, for which elevated intracellular calcium levels were measured on a Molecular Devices Flexstation. The fluorescent ligand had no effect on the activation of these receptors (agonist-induced increase in intracellular calcium).
For live cell imaging, cells expressing the 5-HT1A receptor or the 5-HT1B receptor or those lacking any 5-HT receptor were stained in the dark with Hoechst 33342 (stains nuclei blue) and then treated with the unlabelled competitor, NAN-190, for 30 min, after which NAN190-BY630 was added for a further 30 min. Cells were imaged on a Molecular Devices ImageXpress Ultra confocal plate reader, with excitation at 635 nm and emission at 685/40 nm for NAN190-BY630 and at 405 nm excitation, 477/60 nm emission for Hoechst.
Incubation of the CHO-5-HT1A cells with NAN190-BY630 (10 nM) revealed clear and intense membrane binding, with low levels of non-specific intracellular sequestration. This binding occurred at a concentration consistent with the 3.2 nM affinity from the functional (SPAP) assay. The membrane binding was demonstrated to be specifically to the 5-HT1A receptor because it was substantially displaced by pre-incubation with the unlabelled competitor, NAN-190 (0.1 - 10 µM), with only a small amount of residual membrane binding. In contrast, no significant membrane binding was observed in either the CHO-5HT1B cells or the CHO cells lacking any 5-HT receptor, in the absence or presence of unlabelled NAN-190.
Fluorescent ligand for the GABAB receptor
CA200935 was validated as a moderate-to-high affinity antagonist ligand for the GABAB receptor (Gi-coupled). To express the receptor, transfection was done in a Human Embryonic Kidney (HEK) cell line expressing a Glo sensor (HEK-Glo cells, Promega) that enables measurement of the levels of cAMP using a bioluminescence assay. Following a significant period of optimisation of various variables (amounts and ratios of cDNAs, duration of transfection before assay, etc.), 30 ng of the GABAB1 subunit was co-transfected with 30 ng of the GABAB2 subunit (ratio 1:1) for 48 h. Receptor expression was confirmed by staining for the two subunits using SNAP-CLIP protein labelling fluorescence technology (both plasmids were available in SNAP- and CLIP-tagged forms – we are very grateful to the laboratory of Prof. Jean-Philippe Pin (Institute of Functional Genomics, CNRS, Montpellier, France) for kindly providing these constructs). In a separate attempt, we transfected the subunits into a HEK cell line and additionally transfected a chimeric G-protein, Gqi9, so as to be able to use measurement of intracellular calcium levels on Flexstation as a functional readout. However, the Glo-sensor assay on PheraStar provided a better platform and was therefore adopted in subsequent investigations.
Both CA200931 (red ligand) and CA200935 (green ligand) showed antagonist property at the receptor, in a manner that suggests the presence of constitutive activity in this receptor. However, when assessed on the Molecular Devices ImageXpress Ultra confocal plate reader, CA200931 displayed non-specific binding that was not displaceable by the unlabelled competitive antagonist CGP 54626 (up to 10 µM) in both GABAB receptor-expressing and non-GABAB receptor-expressing (mock-transfected) HEK-Glo cells, whereas CA200935 showed clear, specific (CGP 54626-displaceable) membrane binding with no significant level of intracellular binding at the GABAB receptor-expressing HEK-Glo cells, but no significant binding at the non-GABAB receptor-expressing (mock-transfected) HEK-Glo cells.
Fluorescent ligands for other receptors
Fluorescent ligands for the adenosine receptors are now available, characterised by other members of the research group. So far, it has not been possible to design fluorescent ligands for the metabotropic glutamate receptors, due to significant similarity (considerable sequence homology) between the subtypes (there are eight of them, grouped in three classes) that makes designing selective ligands difficult.
Potential impact and use (including the socio-economic) of the project
The two fluorescent ligands now fully characterised have an enormous range of applications for which they can be used, some of which possibilities are now being exploited. Their use is capable of enabling the assessment of ligand-receptor interactions in an unprecedented manner, not only qualitatively but also quantitatively, which makes them extremely useful in physiological and pharmacological investigations of receptor- associated molecular and cellular causes or consequences of dysfunction in disease and thus in the identification of relevant treatments. Their versatility is further underlain by a range of fluorescence-based approaches in which they can be potentially employed, including FCS, FRET and FRAP. They will be made available to several investigators in public and private research establishments, including those involved in clarifying disease mechanisms and developing novel therapeutics for the treatment of a wide range of diseases. In particular and as an example, their use will provide detailed mechanistic insights into the off- and on-rate kinetics of drugs targeting GPCRs, an approach that is now essential to our understanding of how and why some drugs exert long-lasting effects, whereas others are short- or intermediate-acting. Knowledge furnished by such investigations will significantly improve drug design and development by reducing the efficacy-related component of the existing high attrition rate that is attributable to the varied binding behaviours of different drugs at their target receptors. In the long run, therefore, the use of these compounds could lead to the development of improved (more efficacious) therapeutics (drugs for clinical use) to prevent or treat debilitating conditions, and this will surely translate to significant socio-economic impact for the wider society by reducing the morbidity and mortality from diseases and their attendant socio-economic consequences.
This Marie Curie IIF project sought to develop a number of high-quality fluorescent probes (ligands) to visualise and quantify interactions between ligands and their receptors in order to explore novel pharmacology such as allosterism, which has shown promise as a novel strategy for eliciting neuroprotection, with relevance for neurological and neurodegenerative conditions. These probes were then to be characterised using state-of-the-art pharmacological and imaging approaches and those with superior qualities were to be employed for downstream applications, including investigations of allosterism and allosteric interactions and drug screening. Focus was on three G-protein-coupled receptors (GPCRs): adenosine receptors, metabotropic glutamate receptors and the GABAB receptor, as their modulation is known to be of significant relevance to neurodegeneration.
Account of experimental work performed
We devoted considerable effort to the characterisation of the candidate fluorescent ligands, which were synthesised by our collaborating spin-out company, CellAura Technologies. Studies were conducted in cell cultures engineered to stably or transiently express the receptors of interest, with appropriate control cultures lacking such expression. As these fluorescent ligands are very expensive to design, we first optimised our characterisation assays for robustness and reproducibility using the 5-HT receptor as a prototype, for which we have a range of stable cell lines expressing at least five of its subtypes. Appropriate functional assays were used to assess the quality of nine (9) different fluorescent ligands that are predicted to be agonists or antagonists at the different subtypes of the 5-HT receptor. Of all these, one showed exceptional promise in targeting the 5-HT1A receptor as an antagonist with high affinity and we advanced it to the level of imaging to determine if it also had excellent photophysical properties to make it an ideal fluorescent ligand for probing this receptor subtype using fluorescence-based (imaging) technologies. Interestingly, this ligand also proved excellent for imaging purposes. In fact, we succeeded in using it to develop a fluorescence binding assay for the receptor, which worked extremely well and holds the promise of replacing the use of radioligands in competitive binding experiments.
When we considered the assays in 5-HT receptor-expressing cell lines to be sufficiently robust and reproducible, we turned to the GABAB receptor and sought to characterise fluorescent ligands targeting it. This endeavour, however, presented the greatest of challenges, largely due to the complexity of this receptor. Unlike the 5-HT receptor subtypes, for which it was possible to engineer stable cells for characterisation and imaging assays, the GABAB receptor could only be expressed through transient transfection of its cDNA into cell lines but even so, the transfection mix had to be made up of separate cDNAs of the GABAB1 and the GABAB2 sub-units, in an appropriate stoichiometric ratio (this had to be determined beforehand), as both subunits are required to express a functional GABAB receptor. Upon overcoming the several hurdles, we were eventually able to test two putative fluorescent ligands for this receptor, one of which was a red version while the other was a green version. Both ligands showed antagonist property in functional assays, but it was the green ligand that had corresponding photophysical properties additionally required to approve of such probes as reasonably useful for imaging-based downstream applications. We also tested the applicability of the green ligand in fluorescence binding assays.
Towards the end of the project, we began to explore the binding kinetics of the two characterised fluorescent ligands targeting the 5-HT1A receptor and the GABAB receptor, respectively. This was done using a confocal microscope (LSM 710) coupled to a perfusion system that allows assessment of binding kinetics in single, living cells, a very promising application for the fluorescent compounds, in addition to several other uses to which they can now be put.
Brief scientific description of main results
The most significant result is the identification of two excellent fluorescent probes with superior properties for use in several applications that involve both the qualitative and quantitative assessment of ligand-receptor interactions at the 5-HT1A receptor subtype and the GABAB receptor.
Fluorescent ligand for the 5-HT1A receptor
CA200992, a BODIPY630/650-conjugated derivative of NAN-190 (NAN190-BY630), was shown to be a fluorescent antagonist ligand with high affinity at, and selectivity for, the 5-HT1A receptor. In Chinese Hamster Ovary (CHO) cells stably expressing both this Gi-coupled receptor and a cyclic adenosine monophosphate (cAMP) response element-secreted placental alkaline phosphatase (CRE-SPAP) reporter gene for measurement of changes in adenylate cyclase activity, CA200992 at increasing concentrations (10-8, 10-7, 10-6M) shifted agonist concentration-response curve (agonist-induced decrease in SPAP levels) to the right in a parallel fashion, indicative of competitive antagonism, with a resultant pKB value of 8.49 ± 0.14 which was replicated in a subsequent fluorescence binding assay. In contrast, this fluorescent ligand had no effect on the agonist concentration-response curve in CHO cell lines similarly engineered to express a closely-related 5-HT1B receptor, also Gi-coupled. The other three 5-HT receptor subtypes tested were the Gq-coupled 5-HT2A, 5-HT2B and 5-HT2C receptors, also stably expressed in CHO cells, for which elevated intracellular calcium levels were measured on a Molecular Devices Flexstation. The fluorescent ligand had no effect on the activation of these receptors (agonist-induced increase in intracellular calcium).
For live cell imaging, cells expressing the 5-HT1A receptor or the 5-HT1B receptor or those lacking any 5-HT receptor were stained in the dark with Hoechst 33342 (stains nuclei blue) and then treated with the unlabelled competitor, NAN-190, for 30 min, after which NAN190-BY630 was added for a further 30 min. Cells were imaged on a Molecular Devices ImageXpress Ultra confocal plate reader, with excitation at 635 nm and emission at 685/40 nm for NAN190-BY630 and at 405 nm excitation, 477/60 nm emission for Hoechst.
Incubation of the CHO-5-HT1A cells with NAN190-BY630 (10 nM) revealed clear and intense membrane binding, with low levels of non-specific intracellular sequestration. This binding occurred at a concentration consistent with the 3.2 nM affinity from the functional (SPAP) assay. The membrane binding was demonstrated to be specifically to the 5-HT1A receptor because it was substantially displaced by pre-incubation with the unlabelled competitor, NAN-190 (0.1 - 10 µM), with only a small amount of residual membrane binding. In contrast, no significant membrane binding was observed in either the CHO-5HT1B cells or the CHO cells lacking any 5-HT receptor, in the absence or presence of unlabelled NAN-190.
Fluorescent ligand for the GABAB receptor
CA200935 was validated as a moderate-to-high affinity antagonist ligand for the GABAB receptor (Gi-coupled). To express the receptor, transfection was done in a Human Embryonic Kidney (HEK) cell line expressing a Glo sensor (HEK-Glo cells, Promega) that enables measurement of the levels of cAMP using a bioluminescence assay. Following a significant period of optimisation of various variables (amounts and ratios of cDNAs, duration of transfection before assay, etc.), 30 ng of the GABAB1 subunit was co-transfected with 30 ng of the GABAB2 subunit (ratio 1:1) for 48 h. Receptor expression was confirmed by staining for the two subunits using SNAP-CLIP protein labelling fluorescence technology (both plasmids were available in SNAP- and CLIP-tagged forms – we are very grateful to the laboratory of Prof. Jean-Philippe Pin (Institute of Functional Genomics, CNRS, Montpellier, France) for kindly providing these constructs). In a separate attempt, we transfected the subunits into a HEK cell line and additionally transfected a chimeric G-protein, Gqi9, so as to be able to use measurement of intracellular calcium levels on Flexstation as a functional readout. However, the Glo-sensor assay on PheraStar provided a better platform and was therefore adopted in subsequent investigations.
Both CA200931 (red ligand) and CA200935 (green ligand) showed antagonist property at the receptor, in a manner that suggests the presence of constitutive activity in this receptor. However, when assessed on the Molecular Devices ImageXpress Ultra confocal plate reader, CA200931 displayed non-specific binding that was not displaceable by the unlabelled competitive antagonist CGP 54626 (up to 10 µM) in both GABAB receptor-expressing and non-GABAB receptor-expressing (mock-transfected) HEK-Glo cells, whereas CA200935 showed clear, specific (CGP 54626-displaceable) membrane binding with no significant level of intracellular binding at the GABAB receptor-expressing HEK-Glo cells, but no significant binding at the non-GABAB receptor-expressing (mock-transfected) HEK-Glo cells.
Fluorescent ligands for other receptors
Fluorescent ligands for the adenosine receptors are now available, characterised by other members of the research group. So far, it has not been possible to design fluorescent ligands for the metabotropic glutamate receptors, due to significant similarity (considerable sequence homology) between the subtypes (there are eight of them, grouped in three classes) that makes designing selective ligands difficult.
Potential impact and use (including the socio-economic) of the project
The two fluorescent ligands now fully characterised have an enormous range of applications for which they can be used, some of which possibilities are now being exploited. Their use is capable of enabling the assessment of ligand-receptor interactions in an unprecedented manner, not only qualitatively but also quantitatively, which makes them extremely useful in physiological and pharmacological investigations of receptor- associated molecular and cellular causes or consequences of dysfunction in disease and thus in the identification of relevant treatments. Their versatility is further underlain by a range of fluorescence-based approaches in which they can be potentially employed, including FCS, FRET and FRAP. They will be made available to several investigators in public and private research establishments, including those involved in clarifying disease mechanisms and developing novel therapeutics for the treatment of a wide range of diseases. In particular and as an example, their use will provide detailed mechanistic insights into the off- and on-rate kinetics of drugs targeting GPCRs, an approach that is now essential to our understanding of how and why some drugs exert long-lasting effects, whereas others are short- or intermediate-acting. Knowledge furnished by such investigations will significantly improve drug design and development by reducing the efficacy-related component of the existing high attrition rate that is attributable to the varied binding behaviours of different drugs at their target receptors. In the long run, therefore, the use of these compounds could lead to the development of improved (more efficacious) therapeutics (drugs for clinical use) to prevent or treat debilitating conditions, and this will surely translate to significant socio-economic impact for the wider society by reducing the morbidity and mortality from diseases and their attendant socio-economic consequences.
