Large-scale neural ensemble recording in the brains of freely behaving mice

Abstract

With the availability of sophisticated genetic techniques, the mouse is a valuable mammalian model to study the molecular and cellular basis of cognitive behaviors. However, the small size of mice makes it difficult for a systematic investigation of activity patterns of neural networks in vivo. Here we report the development and construction of a high-density ensemble recording array with up to 128-recording channels that can be formatted as single electrodes, stereotrodes, or tetrodes. This high-density recording array is capable of recording from hundreds of individual neurons simultaneously in the hippocampus of the freely behaving mice. This large-scale in vivo ensemble recording techniques, once coupled with mouse genetics, should be valuable to the study of complex relationship between the genes, neural network, and cognitive behaviors.

Introduction

The arrival of functional genomics era is marked by the ever-increasing need to investigate physiological functions of genes in vivo. As approximately 60–70% of the genes have been estimated to be either brain-specific or highly enriched in the brain, it is important to study the gene function in cognitive behaviors. The rapid development of a series of inducible and region-specific gene knockout (Mack et al., 2001, Shimizu et al., 2000, Tsien et al., 1996a) or more recently, inducible protein knockout techniques (Wang et al., 2003), as well as transgenic methods (Hedou and Mansuy, 2003, Kida et al., 2002) have permitted precise investigations of the relationship between genes and behaviors. For example, series of conditional gene knockout experiments have allowed us to show that the knockout of the NMDA receptor in the CA1 region of the hippocampus impairs the CA1 synaptic plasticity and leads to profound memory deficits (Tsien et al., 1996b). Moreover, genetic enhancement of NMDA receptor coincidence-detection function through the up-regulation of the NR2B subunit in the mouse forebrain can lead to significant enhancement in both learning and memory (Tang et al., 1999, Wong et al., 2002), thereby stringently validating the Hebb's learning rule (Tsien, 2000). Thus, various mouse genetic techniques provide powerful ways to dissect the molecular and genetic mechanisms of cognition in the mammalian species.

One crucial link in our understanding of the relationship between genes and behaviors lies at our ability to measure neural network properties and dynamical patterns associated with genetic and behavioral changes. Over the past several decades, neuroscientists have obtained valuable insights by using EEG to map global brain responses or by recording the activity of one or a few neurons at a time. However, neither approach provides a direct means to investigate the network mechanisms underlying information processing. Encouragingly, in recent years, simultaneous monitoring of activities of many neurons has become more feasible in rats (Gray et al., 1995, Harris et al., 2000, McNaughton et al., 1983, Schmidt, 1999). Since mice are typically only about one tenth to one fifteenth of the body weight of rats (20–30 g versus 300–450 g of body weight), many of the ensemble recording microdrives designed for rats are often too big to be used for the recording in mice. A mouse version of such microdrives has been reported to be able to carry up to 24-channels that can simultaneously record approximately 20–30 individual neurons in the brains of freely behaving mice (McHugh et al., 1996). Here we report the design and construction of a high-density microdrive system which can hold up to 128-channels and allows for a measurement of activities of over two hundreds individual neurons in the brains of freely behaving mice. This high-density ensemble recording array should be a valuable tool in the study of relationships between the genes, neural network, and behaviors.