As announced in my ‘Phoenix rising’ blog (Schreiber 2015), the second edition will focus on novel insights from molecular pharmacology that allow to modulate activity of known receptors. The main attractiveness of such an ‘allosteric’ approach is that normalisation of protein activity in the disease state might lead to a better therapeutic window. I am grateful to Byron DeLabarre for encouraging me to write this blog and for his great input.
Allosteric vs. orthosteric approaches
The way that ligands interact and modulate activity of proteins is far more complex that initially thought. This makes not only for more interesting science and Nobel prizes (DeLaBarre, 2015), but also offers great potential for the discovery of novel drugs. Initially, drug discovery efforts focused almost exclusively on modulation of the orthosteric site. That is, the binding site for the endogenous agonist on the protein/receptor. Since these sites are often conserved between different receptors, it has been challenging to obtain drugs with selectivity over other receptors. In contrast, allosteric binding sites are less conserved. These sites are “non overlapping and spatially distinct from, but conformationally linked to, the orthosteric site” (Christapoulus et al., 2014). To explore the progress that has been made with allosteric approaches in the last 10-15 years, let’s look in more detail at the glutamatergic field, taking one example for two classes of proteins that have historically been intensively targeted in CNS drug discovery by orthosteric approaches: Ligand-Gated Ion Channels (LGICs) and G-Protein Coupled Receptors (GPCRs). Allosteric targeting of membrane embedded proteins has been challenging since large regions of the proteins are not accessible to pharmaceutical interaction. That makes the recent progress in the GPCR field even more remarkable. Although the current focus on LGICs and GPCRs represents just a small snapshot from a thriving field, there are perhaps lessons to be learned that are more widely applicable, especially since substantial efforts have been put in these examples.
Allosteric modulation of LGICs: NMDA receptors
Designing orthosteric ligands for glutamate receptors has been difficult and most small molecular weight (SMW) agonists and antagonists are typically amino acid derivatives with limited oral bioavailability and brain permeability. Although much effort has been put on addressing the problems associated with orthosteric receptor modulation, success has been limited. Allosteric receptor modulation has emerged as an attractive alternative to overcoming many of the inherent challenges of orthosteric target-centered approaches (for a review, see Menniti et al., 2013). Indeed, one of the strengths of allostery is that it allows to move away from the small polar type molecules.
Structure of an NMDAR. The GluN1 subunits are represented in grey and the GluN2 subunits in blue (from Zhu and Paoletti, 2013). The NTDs bind modulators such as ifenprodil; the ABDs bind glycine (or D-serine) in GluN1 and glutamate in GluN2, and the TMD contains the ion channel pore. Abbreviations: NTD, N-terminal domain; ABD, agonist-binding domain; TMD, transmembrane domain; CTD, C-terminal domain.
The excitatory neurotransmitter glutamate binds to (amongst others) ionotropic NMDA (N-Methyl-D-aspartate) receptors. This receptor is rather unique as it requires two co-agonists, the amino acids glutamate and glycine, to activate the receptor. NMDARs assemble as hetero-tetramers associating two GluN1 subunits and two GluN2 subunits, of which there are four subtypes (GluN2A–2D). Receptors with different GluN2 subunits are differentially localized to different brain circuits and in the subsynaptic space, representing distinct drug targets. For example, the GluN2B subunit is highly expressed in the adult cortex, hippocampus, thalamus, and striatum, forebrain regions that support higher cognitive and emotional functions (Menniti et al., 2013).
There are several modulatory binding sites at this receptor that can be targeted by allosteric drugs. First generation modulators like ifenprodil target the GluN2B site at the N-terminal domain (NTD; Zhu and Paoletti 2013). Recently, modulators have become available that act at the agonist-binding domain (ABD), such as the NR2A negative allosteric modulator, TCN-201, and the transmembrane domain (TMD), such as the GluN2A and GluN2B positive allosteric modulator, pregnenolone sulfate (for a recent review, see Zhu and Paoletti, 2015). However, to my knowledge, so far none of the allosteric drugs that target these modulatory sites have been succesful in the clinic.
The recently described oxysterol binding site for neuroactive steroids offers a novel allosteric approach for the NMDA receptor. The major brain-derived cholesterol metabolite 24(S)-hydroxycholesterol (24(S)-HC) is a very potent, direct, and selective positive allosteric modulator at this site. Its mechanism that does not overlap that of other allosteric modulators. Its actions are mimicked by novel synthetic oxysterols such as SGE-201 and SGE-301 (Lisenbradt et al., 2013). Previous efforts that targeted the orthosteric binding site have all failed, often due to mechanism-based neurotoxicity. Targetting the oxysterol site could offer the opportunity for novel, first-in-class, NMDA-PAM approach to treat CNS disorders such as Alzheimer’s disease and schizophrenia without neurotoxic side effects.
Allosteric modulation of GPCRs: metabotropic glutamate receptors
Most allosteric approaches for targeting mGluRs (there are 8 receptors in total) – and perhaps for allosteric CNS GPCR drug discovery overall – have been with the mGluR5 receptors. More than 15 companies, including Roche, Novartis, Merck, Pfizer, Lilly, Addex and GlaxoSmithKline had mGlu5 NAM drug discovery programs (Scharf et al., 2015). An area of focus has been Fragile X Syndrome (FXS), a genetic disorder that is often accompanied by autism. The strong scientific rationale for treating FXS with mGluR5 NAMs was mainly developed by Mark Bear (Kreuger and Bear 2011) who co-founded a company (Seaside Therapeutics), that developed their own mGluR5 NAMs (which are now discontinued). At least two mGluR5 NAMs were tested in late clinical development and failed: basimglurant (aka RG-7090/RO4917523) from Roche, and mavoglurant (AFG056) from Novartis. This was a hard hit for the field in light of the enthusiasm that the strong science generated. The question is whether these are ‘false negative’ or ‘good negative’ data. The advocates of the former point to limitations in efficacy outcome measures, the heterogenic nature of FXS, the age of the patients, dose selection and too much reliance on a single animal model as possible reasons for the failure of these drugs (Scharf et al. 2015; Mullard 2015). Others, myself included, believe that some sort of efficacy signal should have been observed in these studies and that it might be time to move on to other therapeutic approaches. Meanwhile, this class of molecules is pursued for other indications, notably L-DOPA induced dyskinesia in Parkinson’s patients. Although movoglurant eventually failed, there appears still life in this field, as dipraglurant (ADX-48621) from Addex is claimed to still be in clinical development for LID.
Where does this leave us with regard to allosteric approaches in CNS drug discovery? This is a dynamic and rapidly developing field which is bound to make further significant progress in years to come. For example, the recent advances in structural biology (>50 structures published in 2014/2015) of GPCRs will enable greater definition and exploitation of allosteric approaches for these important biological drug targets. Eventually, structures of all the human GPCRs may become available (http://gpcrconsortium.org; http://gpcr.usc.edu). In the meanwhile, models of yet to be determined structures provide a deeper pool of experimental evidence to draw upon. Therefore, I expect that allosteric approches will lead to the discovery of novel drugs that will be valuable for patients. But at this stage, I find it more difficult to judge what impact this field may have on finding more efficacious CNS drugs.
Especially the mGluR5 NAMs illustrate that allosteric drug discovery can indeed succesfully find selective drugs. However, this in and by itself is not enough to develop efficacious drugs. Even for rare genetic conditions such as FXS that are thought to give us the best shot at goal because 1) we often have a much better understanding of the pathophysiology of these disorders; 2) we have access to transgenic rodent models with good validity, and 3) a strong potential exists for successfully translation of drug findings between rodents and human. The reasons for these negative findings with allosteric approaches are not clear. Of course the complexity of CNS disorders puts a high hurdle on finding an efficacious SMW mono-therapy per se. Alternatively, polypharmacy or even ‘target-agnostic’ (aka phenotypical) approaches may hold more promise. This will be the topic of my next blog. Stay tuned!
Schreiber (2015). Trends in CNS drug discovery: Phoenix rising. http://suadeo-consulting.com/index.php/blog/trends-in-cns-drug-discovery-phoenix-rising
DeLaBarre (2015). Of Llama Blood & Nobel Prizes: How We Came to Understand GPCR Structure. http://theconsultingbiochemist.com/2015/05/gpcrs_nanobodies/
Christapoulus et al. (2014). International Union of Basic and Clinical Pharmacology. XC. Multisite Pharmacology: Recommendations for the Nomenclature of Receptor Allosterism and Allosteric Ligands. Pharmacological Reviews 66:918–947.
Linsenbradt et al. (2014). Different oxysterols have opposing actions at N-methyl-D-aspartate receptors. Neuropharmacology 85: 232-42.
Menitti et al. (2013). Allosteric modulators for the treatment of schizophrenia: targeting glutamatergic networks. Current Topics in Medicinal Chemistry 13:26-54.
Zhu and Paoletti (2015). Allosteric modulators of NMDA receptors: multiple sites and mechanisms.Current Opinion in Pharmacology 20: 14-23.
Scharf et al. (2015). Metabotropic glutamate receptor 5 as drug target for Fragile X syndrome. Current Opinion in Pharmacology 20: 24-34.
Mullard (2015). Fragile X disappointments upset autism ambitions. Nature Reviews Drug Discovery 14: 151-3.
Krueger and Bear (2011). Toward Fulfilling the Promise of Molecular Medicine in Fragile X Syndrome. Annual Review of Medicine 62:411-19.