Structural analysis of human GABAA receptors

The human brain is the most powerful biological processor on the planet. It is also extremely complicated. It contains around one-hundred billion brain cells and over a trillion connections between those brain cells which channel the flow of vast quantities of information. Clearly such a complicated system requires rules if it is to operate smoothly. It requires a regulatory system. And the targets of our research, gamma-aminobutyric acid Type-A receptors, termed GABA-A-Rs, are a most important part of this regulatory system.

GABA-A-Rs are proteins expressed at the surface of nerve cells. They exist throughout the brain to allow nerve cells to respond to the neurotransmitter gamma-aminobutyric acid, called GABA. The neurotransmitter GABA acts upon GABA-A-Rs to send a stop signal to nerve cells. This means GABA-A-Rs are the brakes on the brain. They spread calm. They regulate excitement. And so it is no surprise that when these regulators fail the result is heightened brain activity, which leads to a range of debilitating illnesses such as insomnia, anxiety-disorders and epilepsy, disorders that affect tens of millions of people worldwide. Fortunately there are treatments available, most famously benzodiazepines which reinforce the calming influence of GABA-A-Rs. Unfortunately, because GABA-A-Rs are vital to so many different processes in the brain benzodiazepines have side-effects. Furthermore, the many roles that GABA-A-Rs play in brain function are far from fully resolved. For example, GABA-A-Rs are also targets for general anaesthetics and for the brains home-grown anti-stress molecules, the neurosteroids. They are also targets in depression and chronic pain, and are being investigated as therapeutics in schizophrenia, stroke and alcohol addiction. Thus, it is vital to understand as fully as possible how these proteins operate in order to better understand how GABA-A-Rs contribute to brain function.

One way to do this is to analyse the structures of GABA-A-Rs. Obtaining three-dimensional maps of GABA-A-Rs at high (atomic or near-atomic) resolution provides vital information on, for example, how these biological machines respond to the neurotransmitter GABA to internally reorganisation themselves to open a built-in gate that lets ions flow across the membranes of nerve cells. Such maps will also reveal how GABA-A-Rs bind clinically relevant drugs and natural ligands in the human brain. Finally, there are different populations of GABA-A-Rs in the human brain, with each population having unique structures, unique functions, unique expression profiles, and unique pharmacology. So obtaining structural data for each of the distinct GABA-A-Rs populations will unveil their distinguishing features and how these features imbue unique properties. Furthermore, detailed structures permit visualisation of drugs bound to GABA-A-Rs, which will inform on design of improved therapeutics in the future.

However, obtaining high-resolution structures of GABA-A-Rs is extremely challenging. Large quantities of highly pure GABA-A-Rs in a stable form are required, something that is especially difficult because GABA-A-Rs are very delicate proteins. Fortunately in our laboratory we have state-of-the-art technologies for protein production, purification, screening, and structure determination. That is why this year (2014) we were successful in publishing the first ever high resolution structure of a GABA-A-R, revealing data on its mode of operation and on the drug-binding cavities it contains. However, a single structure is only a snap-shot. To gain a better understanding of protein function many alternative structures (snap shots) of these biological machines in action are required, as well as of other subtypes. We believe that by obtaining more GABA-A-R structures in the future we will substantially enhance the understanding of how GABA-A-Rs impact on brain function and inform pharmaceutical enterprise on how to design improved therapeutics.

Building on the efficient protocols we have previously established for milligram-scale eukaryotic expression, purification, and crystallisation of the GABA-A-R beta3 homomer, we aim to crystallise and solve structures of this receptor in complex with physiologically and pharmacologically-relevant ligands. To maximise our chances of success we will perform crystallisation experiments in either apo form or in the presence of high affinity GABA-A-R beta3 nanobodies, which we have already developed (in collaboration with Jan Steyaert, VIB) and characterised, and behave as crystallisation chaperones. Crystallographic information will be validated by functional assays (fluorescence chloride flux activity assays, electrophysiology).

We will also attempt to determine near-atomic (3.5-4.5 Å) resolution structures of GABA-A-R beta3 homomers by single particle free cryo-electron microscopy in the absence and presence of an agonist, to build an activation pathway of secondary structure motions for the GABA-A-R in the resting, activated and desensitised states. To capture the GABA-A-R in the activated state before it enters the desensitised state, we will combine sample treatment with rapid millisecond drug application and plunge-freezing using apparatus available through a collaboration with Nigel Unwin (MRC-LMB, Cambridge).

We have expanded our efficient protein production and purification procedures of monodisperse GABA-A-R beta3 homomers to a range of physiological heteromers and will employ these for crystallisation trials and cryo-electron microscopy analysis. We are fully aware of the technical difficulties associated with human membrane protein structural work, but have succeeded in overcoming such hurdles in the past, and are now in an excellent position to tackle this project, as illustrated by a large body of preliminary data.

GABA-A-Rs are important targets for benzodiazepines, used to treat neurological disorders such as epilepsy, insomnia, anxiety, depression and pain, as well as for general anaesthetics. Thus, our efforts to obtain high-resolution structures of various GABA-A-R subtypes in multiple conformations and bound by ligands will be of interest to anyone interested in rational design of drugs against GABA-A-Rs for development of novel and improved therapeutics. Principally, pharmaceutical companies will be interested in using these structures to understand how current drugs on the market bind GABA-A-Rs, and using this knowledge to rationally design drug analogues that have enhanced binding properties, for example enhanced selectivity against one subtype over another to create a drug with reduced side-effects in patients. Not only will rational drug design generate novel lead molecules for drug development, it will also increase the rate of generation of novel lead compounds because computational screening of 3D GABA-A-R coordinates will help predict suitable leads. Thus our GABA-A-R structures will speed up and improve chances of success, saving companies money in time and resources spent. As a consequence of any such successful research endeavours by pharmaceutical companies, of course patients will then benefit from these new therapeutics. The search for novel drug leads based on GABA-A-R structures and so savings in time and resources for the pharmaceutical industry could begin as soon as we obtain novel structures of interest, within two years of beginning the research proposal. Subsequent benefits to patients will take longer, due to the slow nature of getting drugs to market, perhaps 10 to 20 years from now.

Another area of interest for the public and private sectors in our research will be in the generation of high affinity protein binders (nanobodies) against GABA-A-Rs, which we have and are continuing to produce to aid GABAAR structural work. These proteins have a number of other uses which will be of interest to commercial parties, for example as fluorescent-tagged tools for tracking GABA-A-Rs in living systems. In particular there is an increasing body of literature pointing to the involvement of GABA-A-Rs in types of cancer (e.g. breast) and so nanobodies against relevant GABA-A-R subtypes could be used to deliver toxic payloads, in similar way to how some antibodies are being used. Thus, our research and the nanobody products generated from it may be of interest to commercial enterprises interested in investigating the applicability of these tools for disease areas like cancer. Should they prove useful then they will be of commercial value to companies and offer health benefits to patients. We have already developed nanobodies which may be of interest for investigation in treatment of disease pathways and we will continue to generate nanobodies against a range of GABA-A-R subtypes over the course of the grant. Thus public and private sectors interested in these products will be able to investigate their applicability in disease within the lifetime of this grant.

A third area that we believe will be of interest to the private sector will be in the development of new and improved methods that we have created for producing GABA-A-R structures which can be more widely applied to other membrane protein targets. This will be of interest to any companies involved in offering rational drug design services, who are looking to develop platforms for membrane proteins or who are looking to improve current membrane protein platforms. Our proof of principle novel methodologies that we publish in our research papers of GABA-A-R structures will encourage structure-based-drug-design companies that service the wider drug development industry to take on these methodologies in order to broaden their portfolios and so enhance the services they are offering.