Spot the Difference: How Do Brains Tell one Place from Another?
25 Jul, 2007 05:11 pm
Our brains are continuously bombarded by shifting patterns of sensory inputs evoked by our environment. Even subtle changes in these patterns need to be noticed, and recent work has highlighted the importance of a specific protein in a specific subset of neurons in rapidly encoding these changes.
Pattern separation would intuitively occur if multiple episodes were encoded and processed as multiple patterns of activity in the nervous system, each one a unique neuronal representation. As such, the sights, sounds and smells of each dental surgery you encounter would at some level activate a unique subset of your 100 billion neurons. How is this possible given the many similarities between surgeries, the same antiseptic smell and ominous sound of drilling? As we move from one place to another, how does our nervous system deal with the demands of rapidly and efficiently disentangling the incoming patterns of sensory information?
The hippocampus is a central brain structure that plays key roles in encoding, storing and recognising contextual information. In particular, electrophysiological recordings in rodents, primates and even humans reveal that the activity of neurons throughout the hippocampus is acutely dependent upon location. The firing rate of individual hippocampal ‘place cells’ is highest when the animal is in that neuron’s ‘place field’, and coactive networks of place cells collectively encode location (the hippocampus of taxi drivers has been shown to become very active as they imagine driving around London, for example). These place cells respond to the convergent sensory inputs feeding in to the hippocampus from other brain structures. Perhaps, then, the hippocampus contributes to disentangling all this information, generating a unique neuronal activity code for each context. By recording the activities of ensembles of hippocampal neurons in different contexts, we can gain insight into the nature of the neuronal code and begin to understand how the anatomy and physiology of neuronal networks may relate to distinct function roles. Furthermore, by using the latest genetic technologies to manipulate gene expression in specific subsets of hippocampal neurons, we can begin to unravel the molecular recipes that define the region’s properties.
We generated a line of mice ‘knockout’ mice lacking a single protein in the principal excitatory neurons of the dentate gyrus (a specific anatomical subfield of the hippocampus). The protein (technically known as the NR1 subunit of the NMDA receptor) would normally form part of a chemical-sensing signalling complex found at the synaptic junctions between neurons. We chose to knock-out this particular protein because it is known to be essential in regulating the strength of cross-talk between connected neurons, shaping their activity in an experience-dependent manner. We then tested these mice in an experiment analogous to the ‘bad trip to the dentist’ described above: when they were placed in one chamber (the clumsy dentist’s surgery), they received a mild electric foot shock. The shock was not powerful enough to do any damage, but did startle the mice because it was so unexpected. The mice were also introduced to a different chamber with altered shape, lighting and smell, where they did not receive a foot shock. As expected, normal mice learned to be suspicious of the shock chamber (which they show by freezing still) whilst relaxing in the ‘safe’ chamber. In contrast, genetically-altered mice from our knockout line were less able to distinguish between the two chambers, and tended to just freeze in both. This experiment (in conjunction with some important controls to show that our knockout mice do not just freeze all the time) therefore showed that this particular protein in these particular dentate gyrus neurons allows mice spot subtle differences between spatial contexts, i.e. to pattern separate.
Might the impaired pattern separation in the knockout mice result from an impaired neuronal encoding of spatial context? To test this, we recorded from hippocampal neuronal ensembles as both normal and knockout mice explored recording chambers of different colours and shapes. Sure enough, we found that patterns of hippocampal activity in normal mice tended to be quite different in different contexts; presumably these different neuronal codes reflect or enable the fact that normal mice can accurately recognise changes in their environments. However, the genetically altered animals showed more overlapping patterns of hippocampal activity in the different recording chambers. The neuronal recording data thereby mirrored the behavioural data, showing that without a fully-functional dentate gyrus, mice became less efficient at rapid and accurate contextual pattern separation.
By using this powerful combination of genetics, electrophysiology and behaviour, we are able to link molecular, synaptic, neuronal network and systems-level analyses, shedding light on fundamental brain mechanisms. This study exemplifies how the exquisite fine-tuning of neuronal connectivity and activity enables processes like pattern separation to occur rapidly and efficiently (helping us to avoid dangerous dentists). But brains can go awry, for example during post-traumatic stress disorder and senility. Research like this therefore remains essential if we are to truly understand the nature of brain function and dysfunction in order to develop vital future therapies.
McHugh T.J., et al, (2007) Dentate Gyrus NMDA Receptors Mediate Rapid Pattern Separation in the Hippocampal Network. Science 317: 94-99
Updated July, 26