Role of the Lateral Mammillary Nucleus in the Rat Head Direction Circuit

Abstract

We recorded head direction (HD) cells from the lateral mammillary nucleus (LMN) and anterior thalamus (ATN) of freely behaving rats and also made bilateral lesions of LMN while recording HD cells from ATN. We discovered that the tuning functions of LMN HD cells become narrower during contraversive head turns, but not ipsiversive head turns, compared to when the head is not turning. This narrowing effect does not occur for ATN HD cells. We also found that the HD signal in LMN leads that in ATN by about 15-20 ms. When LMN was lesioned bilaterally, HD cells in ATN immediately lost their directional firing properties and never recovered them. Based on these findings, we argue that LMN may be an essential component of an attractor-integrator network that participates in generating the HD signal.

A Continuous Attractor Network Model Without Recurrent Excitation: Maintenance and Integration in the Head Direction Cell System

“…From the neurophysiological point of view, the tuning curves we observe are quantitatively similar to the tuning curves recorded in the rat HD cell system. HD cells are numerous in the LMN (Blair et al, 1998;Stackman and Taube, 1998). In the DTN, the picture is less clear, because few electrophysiological recordings are available in this area Bassett and Taube, 2001b).…”

“…Lesions to DTN disrupt the directional selectivity in ADN (Bassett and Taube, 2001a). Bilateral lesions of LMN abolish the HD signal in ADN (Blair et al, 1998). The directional selectivity of PoSC cells is seriously impaired when lesioning ADN (Goodridge and Taube, 1997), but lesions of the PoSC leave the signal in ADN largely intact (Goodridge and Taube, 1997).…”

“…Although the actual mechanisms underlying the generation of the HD signal are still unclear, anatomical and lesion data suggest that LMN and DTN might constitute an essential sub-circuit of the HD cell system for generating and maintaining the directional signal (Blair et al, 1998;Bassett and Taube, 2001b;Taube and Bassett, 2003). This would yield the following ascending processing scheme: DTN → LMN → ADN → PoSC.…”

“…A further evidence for the above processing pathway is provided by a series of studies that have investigated the temporal properties of HD cells in LMN, ADN, and PoSC. During head turns, LMN neurons tend to anticipate the future head direction by approximately 40-75 ms (Stackman and Taube, 1998;Blair et al, 1998), ADN cells show a smaller anticipatory time delay of about 25 ms (Taube and Muller, 1998;Cho and Sharp, 2001), and PoSC cells tend to encode the current directional heading (Blair and Sharp, 1995). The DTN-LMN circuit seems also well suited to account for the additional property of the HD cell system of integrating the head angular velocity.…”

“…However, Skaggs et al (1995) did not simulate any model system to check the feasibility of this scenario. Blair et al (1998) and Degris et al (2004) realized a neural implementation of this scenario by proposing an attractor-integrator circuit modeling the LMN-DTN system. Finally, Xie et al (2002) studied a two population network of left and right units, and showed that with appropriate connections and linear synapses, the network can integrate the inputs with good accuracy in a large velocity range.…”