Corticotropin-releasing factor (CRF), a hypothalamic peptide that releases adreno-corticotropin (ACTH) from the pituitary anterior lobe, has been reported to be present in various regions of the brain (De Souza et al, 1985; De Souza, 1987; Sakanakaet al , 1987). In the cerebellum, CRF is concentrated in the climbing fibers originating in the inferior olive of the medulla and supplies strong excitatory synapses onto Purkinje cells. The entire course of the olivo-cerebellar climbing fiber system, from the somata of olivary neurons to the climbing fiber terminals, is immunoreactive to CRF (Sakanaka et al, 1987; Sawchenko and Swanson, 1990; Cummings et al, 1994a). On the other hand, mRNA for CRF receptors is shown to be present in Purkinje cells as well as in granule cells (Potter et al, 1994; Chalmers et al, 1995). CRF mRNA expression or immunoreactivity of inferior olive neurons has been demonstrated to increase by experimental manipulations that enhance the activity of these neurons (Cratty and Birkle, 1994; Gabr et al, 1994). In anesthetized rat, CRF is reported to enhance the sensitivity of Purkinje cells to glutamate and aspartate, but reduce the sensitivity to GABA (Bishop, 1990; Bishop and Kerr, 1992; Bishop and King, 1992). It is also reported that CRF reduces the spike-induced afterhyperpolarization in cultured Purkinje cells, presumably due to the closure of Ca 2 + activated K + channels (Fox and Gruol, 1993). These results suggest that CRF may be released from climbing fibers in an activity-dependent fashion and modulate Purkinje cell excitability.One important function of the climbing fiber system is to trigger long-term depression (LTD) of parallel fiber synapses, the other major excitatory inputs to Purkinje cells (Ito et al, 1982; Ekerot and Kano, 1985). Cerebellar LTD is a unique form of synaptic plasticity in the cerebellum that is suggested to be a cellular basis of motor learning such as adaptation of the vestibulo-ocular reflex (Ito, 1984, 1989 for review) and conditioned eyeblink response (Aiba et al, 1994; Chen et al, 1995; Shibuki et al, 1996; Thompson and Krupa, 1994 for review). CRF expression is shown to increase in olivary neurons by sustained optokinetic stimulation, in a condition similar to induced adaptation of vestibulo-ocular reflex (Ito, 1989). These results suggest that CRF might play a role in motor learning, possibly by influencing the induction of cerebellar LTD. Thus, the present study was undertaken to examine whether exogenous CRF can induce modification of parallel fiber synaptic strength.Parasagittal cerebellar slices of 400 μm thickness were prepared from the vermis of 6-7 week-old Wister rats. Slices were kept at room temperature until use in the standard saline of the following composition (mM):NaCl 118; KCl 4.7; CaCl 2 2H 2.5; NaHCO 3 25; KH 2 PO 4 1.2; MgSO 4 7H 2 O 1.2; glucose 11; equilibrated with 95% O 2 + 5% CO 2 gas (pH 7.4). One slice was transferred to a submerged recording chamber that was continuously perfused (2 ml/min) with the standard saline (31-33°C) containing picrotoxin (20 μM) to block inhibition. Two bipolar electrodes made of a pair of platinum-iridium wires were placed on the pial surface and the white matter to stimulate parallel fibers and climbing fibers, respectively. Intradendritic recording was made from Purkinje cells with a sharp glass microelectrode filled with 3 M KCl (resistance, around 80 MΩ). Purkinje cells were penetrated at the inner half of the molecular layer, beneath the parallel fiber stimulation electrode.Stimulation of parallel fibers induced EPSPs whose amplitudes were graded with the stimulus intensity. We stimulated parallel fibers at a low frequency of 0.2 Hz and adjusted the stimulus intensity so as to yield parallel fiber-mediated EPSPs (PF-EPSPs) of 8-15 mV in amplitude. Under this stimulus condition, PF-EPSPs were stable with no sign of progressive depression that has been reported to occur when strong parallel fiber stimulation was repeated (Hartell, 1996). After stable recording of PF-EPSPs for 10 min, CRF (--μM) was bath-applied together with repetitive stimulation of parallel fibers (1 Hz) for 5 min. This caused persistent depression of PF-EPSPs (fig 1A, fig 2; CRF + PF) which lasted longer than 60 min. This depression was analogous to the LTD induced by conjunctive parallel fiber and climbing fiber stimulation (fig 1B, fig 2; LTD). In contrast, repetitive parallel fiber stimulation alone at 1 Hz for 5 min (300 stimuli) did not induce any depression, but a slight potentiation of PF-EPSPs (fig 2; PF alone). Bath application of CRF (0.5 μM) alone without parallel fiber stimulation caused no persistent change in PF-EPSPs (fig 2; CRF alone). Furthermore, CRF (7-14) (1 μM), an inactive analogue of CRF, had no effects on PF-EPSPs when combined with parallel fiber stimulation at 1 Hz for 5 min (fig 2; CRF (7-14)).These results suggest that activation of CRF receptors on Purkinje cells presumably causes persistent depression of PF-EPEPs, when combined with repetitive parallel fiber stimulation. This CRF-induced depression was similar to LTD with regard to the time course (fig 1) and magnitude (fig 2). In our preliminary experiments, CRF-induced depression was mutually occluded with the LTD induced by conjunctive parallel fiber and climbing fiber stimulation. We also observed that CRF-induced depression was abolished by several drugs that have been shown to block LTD. It is therefore likely that both forms of persistent depression share common signal transduction mechanisms. Previous reports have shown that CRF is concentrated in the olivo-cerebellar climbing fiber system (Cummings et al, 1994a) and CRF expression appears to be up-regulated by experimental manipulations to increase the activity of inferior olive neurons. These manipulations included sustained optokinetic stimulation in the rabbit (Barmack and Errico, 1993) and harmaline treatment in the rat (Cummings et al, 1994b). Thus, it is likely that CRF-released from climbing fiber terminals in certain conditions may play a role in the induction of LTD.