A TTX-sensitive local circuit is involved in the expression of PK2 and BDNF circadian rhythms in the mouse suprachiasmatic nucleus Kenkichi Baba, Daisuke Ono, Sato Honma and Ken-ichi Honma Department of Physiology, Hokkaido University Graduate School of Medicine, North 15, West 7, Kita-ku, Sapporo 060-8638, Japan Keywords: cell communication, clock gene, neural network, tissue culture, vasopressin Abstract The roles of a local circuit of electrophysiological activity were examined in the expression of circadian rhythms in the suprachiasmatic nucleus (SCN) of the adult mouse. The neuronal activity of cultured SCN was suppressed with tetrodotoxin (TTX), an Na + channel blocker, and the circadian rhythms in mRNA level were assessed for 13 genes by in situ hybridization. SCN slices were cultured for 3 days and TTX was applied at the peak phase of Per1 expression rhythm. The SCN slices were examined at 4-h intervals up to 32 h after TTX application. The circadian rhythms in the expression of clock genes, Per1, Per2, Bmal1 and Cry1, and of clock-associated genes, Dec1, Dec2, Rev-erba, Rev-erbb and DBP, were not affected by TTX treatment. By contrast, TTX completely abolished the circadian rhythm in the BDNF mRNA level and substantially damped the rhythm in PK2. The circadian rhythm in the AVP mRNA level was not changed significantly by TTX. These findings indicate that input through Na + -channel-dependent electrophysiological activity is not necessary for the expression of the circadian rhythms of clock and clock-associated genes, but necessary for full expression of the circadian rhythms of BDNF and PK2 in the SCN. A TTX-sensitive circuit is involved in the expression of BDNF and PK2 circadian rhythms in the mouse SCN. Introduction The suprachiasmatic nucleus (SCN) is the site of circadian pacemaker in mammals, which is composed of multiple oscillating cells (Reppert & Weaver, 2002). Although the circadian oscillation in each cell seems to be substantially different (Welsh et al., 1995; Yamaguchi et al., 2003), the circadian output from the SCN is quite uniform and stable (Inouye & Kawamura, 1979; Yamazaki et al., 2000). There is a line of evidence that the mutual interaction among the oscillating cells integrates the individual circadian rhythms through humoral as well as neural communications (Yamaguchi et al., 2003; Maywood et al., 2006). However, the precise mechanism of cell interaction is unknown. The SCN pacemaker receives multiple signals from the surround- ings. Conceptually, there are two kinds of input, one from outside the SCN for photic and non-photic entrainment, and the other is the signals generated inside the SCN for mutual communication of oscillating cells. Input signals eventually reach the core oscillating loop consisted of four clock genes [Per(s), Cry(s), Bmal1, Clock] and their protein products. The core oscillating loop is coupled to associated loops: a Rev-erba-ROR-Bmal1 and Dec loops. Output signals from the core loop seem to be mediated by several factors (Honma & Honma, 2003). A CLOCK ⁄ BMAL1 heterodimer is a potential intracellular mediator through E-box activation of other genes. The cellular circadian signals are finally integrated to the circadian rhythm of electrophysiological activity. Tetrodotoxin (TTX), an Na + channel blocker, suppressed the neuronal activity and abolished the circadian rhythms in behaviors, but did not affect the circadian oscillation in the SCN (Schwartz et al., 1987; Shibata & Moore, 1993). Synaptic communication seems to be a major mechanism of circadian signal transmission in the SCN, but humoral factors such as hormones and cytokines are also probably involved in the transmis- sion by means of autocrine and paracrine mechanisms (Cheng et al., 2002; Allen & Earnest, 2004). Arginine vasopression (AVP) is released from the SCN and transmits the circadian signal to other structures in the brain (Kalsbeek et al., 1992; Gomez et al., 1997). Brain-derived neurotrophic factor (BDNF) is secreted from the SCN and suggested to transfer the circadian signals to behavior (Liang et al., 1998; Naert et al., 2006). Prokineticin 2 (PK2) and transforming growth factor a (TGFa) are also released from the SCN, and are involved in the expression of circadian behavioral rhythm (Kramer et al., 2001; Cheng et al., 2002). AVP and PK2 are known to posses the E-box motif (CACGTG) in their promoter regions and are trans- activated by the CLOCK ⁄ BMAL1 complex (Jin et al., 1999; Cheng et al., 2002). Therefore, they seem to be directly regulated by the core oscillating loop. The aim of the present study was to examine the roles of electrophysiological activity in the expression of circadian rhythm of several genes to provide better understanding of the intracellular transmission of circadian signals. The circadian gene expression rhythm would be affected by TTX, if the gene is involved in neither the circadian rhythm generation nor the intracellular transmission of circadian signals. Correspondence: Dr K. Honma, as above. E-mail: kenhonma@med.hokudai.ac.jp Received 30 July 2007, revised 14 November 2007, accepted 17 Decemer 2007 European Journal of Neuroscience, Vol. 27, pp. 909–916, 2008 doi:10.1111/j.1460-9568.2008.06053.x ª The Authors (2008). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd