Materials and methods number following potentiation. Further characterization of

T .E. Boyd et al. Brain Research 881 2000 28 –36 29 [23] is the same phenomenon as that monitored over surgery to provide optimal response amplitudes. The weeks and months in the awake animal. resulting mean depths for the white matter stimulating and LTP, like memory, has been shown to be affected by cortical recording electrodes were 3.0 mm and 1.8 mm cholinergic manipulations, but these experiments were ventral to dura, respectively. The electrodes were con- done in in vitro preparations. Cholinergic agonists, for nected to gold-plated male pins that were then inserted into example, enhance LTP induction in both the CAl [5] and a nine-pin miniature connector plug that was mounted onto dentate gyrus [7] regions of the hippocampus as well as in the skull with dental cement and anchored with stainless the visual cortex [6]. The facilitatory action of acetyl- steel screws. One of the screws served as the ground choline on the induction of LTP is most likely the result of electrode. Data acquisition began 10–14 days following enhanced postsynaptic depolarization, increasing the prob- surgery. All experiments were carried out following proper ability, or extent, of NMDA receptor activation [9,32]. guidelines as stipulated by the Canadian Council of Even cholinergic agents by themselves have been shown to Animal Care and approved by the university veterinarian. produce a long-term enhancement of responses in the hippocampus [3] or sensorimotor cortex [27]. In the hippocampus, this cholinergically-induced LTP has been 2.2. Baseline measures and induction of LTP shown to occlude subsequent LTP induction by electrical stimulation [3], suggesting that these two forms of LTP Electrical stimuli were produced by a Grass S88 share a common substrate. stimulator, coupled to photoelectric stimulus isolation Previously in our hands, cholinergic stimulation has units. Recorded signals were fed into Grass Model 12 EEG been shown to promote a long-term depression effect in amplifiers, high pass filtered at 0.3 Hz, low pass filtered at the neocortex of freely moving animals following single- 3 Hz, and digitized 10 KHz with a 12-bit A D converter. session stimulation protocols [37,38]. The following ex- Three sets of field potential input output I O mea- periments were undertaken to characterize the role of sures, spaced at 48 h, were taken to establish baseline. cholinergic neuromodulation in the induction of neocorti- Pulses of increasing intensity were delivered to white cal LTP induced by the spaced and repeated stimulation matter at a frequency of 0.1 Hz. Six 50 ms responses were protocol. Based on the LTP slice literature and behavioural evoked, amplified, digitized and averaged at each of 10 data, we predicted that the cholinergic agonist pilocarpine intensities 16, 32, 64, 100, 160, 250, 500, 795, 1000, 1260 would enhance LTP induction, while the antagonist m A. The evoked field potentials comprised two main scopolamine would attenuate LTP induction. Portions of components: an early, monosynaptic, surface-negative this research have been presented previously in abstract response, and a larger, polysynaptic late response. The form [39]. polysynaptic components were often most evident follow- ing LTP induction. Also evident at latencies coinciding with the early component were at least one, and usually several, population spikes that increased in amplitude and

2. Materials and methods number following potentiation. Further characterization of

these response components can be found in [8,42]. 2.1. Animals and surgery Beginning 24 h after the third baseline I O test, 60 high-frequency trains ITI: 10 s were delivered daily to Fifty-five male Long-Evans hooded rats from the white matter. Trains 24 ms duration consisted of 8 McMaster University Breeding Colonies were used in pulses, with a pulse frequency and duration of 300 Hz and these experiments. At the time of surgery, the animals 0.1 ms, respectively. Pilot work indicated that differential weighed 300–400 g.They were housed individually, main- effects might be obtained depending on whether low or tained on an ad libitum feeding schedule, and kept on a 12 high-intensity trains were used to induce LTP. Conse- h on 12 h off light cycle. quently, two different tetanic pulse intensities were used in Twisted wire bipolar electrodes were prepared from these experiments: 160 mA ‘low-intensity’ and 1260 mA Teflon-coated, stainless-steel wire 120 mm in diameter. ‘high-intensity’. We used fixed intensities, rather than The vertical tip separation for the recording electrode was adjusting intensity according to each animals’ response 1.0 mm, to span the dipole. The tip separation was 0.5 mm profiles, because the baseline response profiles were quite for the stimulating electrode, to span the corpus callosum. consistent across animals. The low intensity 160 mA Following anaesthesia with sodium pentobarbital 65 mg produced a response that was 4560.04 of maximum kg, a bipolar recording electrode was implanted into the amplitude. In addition, the high intensity stimulation primary motor cortex 2.0 mm anterior to Bregma and 4.0 induced a maximal response in all animals. EEG was mm lateral to the midline, and a stimulating electrode was monitored with either an Explorer IIIA digital oscilloscope implanted into white matter the forceps minor corpus Nicollet Instrument Corp., Madison, WI or a Grass callosum, 2.0 mm anterior to Bregma and 2.0 mm lateral to Model 7D polygraph Grass Instrument Co., Quincy, MA the midline. These two electrodes were adjusted during to ensure that epileptiform discharges were not triggered. 30 T 2.3. Cholinergic agonism and antagonism The cholinergic agonist pilocarpine 10–20 mg kg i.p. in saline, Sigma Chemical Co., USA and antagonist scopolamine 15 mg kg i.p. in saline, Sigma Chemical Co., USA were administered in conjunction with high- frequency trains. These drugs were chosen specifically because they are widely used in behavioural tests of memory e.g. [10], and because they have broad-spectrum effects on ACh receptors. Pilocarpine binds to all mus- carinic receptors, but with more affinity to M1. Scopolamine binds to all five muscarinic receptors with equal affinity. On the first day of experimentation, animals were injected with pilocarpine low-intensity group, n55; high- intensity group, n57, scopolamine low-intensity group, n56; high-intensity group, n510 or saline low-intensity group, n56; high-intensity group, n58, followed 15 min Fig. 1. Representative examples of neocortical field potentials evoked by white matter stimulation. The solid line represents an unpotentiated later by the delivery of 60 high-frequency trains. On evoked field potential, while the dashed line represents an evoked field subsequent days, the injection was preceded by an I O test potential following spaced and repeated tetani. Peaks of early and late and followed 15 min later by 60 high-frequency trains. components are identified with arrows. As well, examples of the cursor Additionally, a seventh n57 and eighth n56 group positions used to set the latencies for the peak amplitude measurements received daily pilocarpine or scopolamine injections, re- are drawn. Vertical calibration 1.0 mV; horizontal calibration 10 ms. spectively, preceded immediately by an I O test, but did not receive trains these groups provided drug-only con- trols. These daily LTP induction regimens were continued for 10 and 15 days for the high- and low-intensity groups, field potentials recorded during the last baseline I O test respectively. Following completion of the LTP induction and the I O test following completion of trains. One phase, 2-weekly I O measures were collected to monitor animal each from the SAL High, PILO High and SCOP decay. High groups was discarded from the population spike measures because of artifactual contamination. Once all 2.4. Analyses recordings were complete, animals were perfused and the brains were sliced and stained with Cresyl violet to verify Changes in the field potentials over LTP-induction and electrode placements. In all animals, stimulating and decay sessions were measured by subtracting the final recording electrodes were positioned within their target baseline responses from all other baseline and potentiated structures [36]. responses at a single mid-range I O test intensity that best reflected potentiation. Therefore, all responses, including the first two baseline responses, were standardized to the 2.5. Behavior third baseline. The dominant early and late components were measured separately at fixed latencies corresponding Animals injected with pilocarpine showed walking and to the peak response of the components being analysed crouching behavior, diarrhea and reddening of the sclera see Fig. 1. Control animals were then analysed using the five to seven min after the injection. These side effects means of these latencies from the experimental animals, diminished 1–2 h post injection. When handled 24 h later, because the peaks for the late components were not always the animals appeared normal. Scopolamine animals dis- clear prior to the induction of potentiation. For amplitude played no visible adverse effects during the experiment. measures, repeated measures ANOVAs and post-hoc All animals groomed normally and gained weight through- Tukey HSD tests were calculated for all induction and out the experiment. During the application of LTP trains, decay days. Statistical analyses were conducted using some animals exhibited a pronounced postural deflection at STATISTICA software. Population spikes were measured each train onset. Electrophysiological measures were re- from the largest amplitude response in the I O measures corded, and behavior observed, from animals inside a before and after LTP induction and after the decay phase. dimly lit testing box. The box measured 30.5 cm328 The heights in mV of all population spikes within a cm340.5 cm w3l3h, with a clear Plexiglass door and a response were summed to provide a measure of total spike mirror mounted on the back wall to facilitate behavioral height for that response. These measures were taken from observation. T .E. Boyd et al. Brain Research 881 2000 28 –36 31

3. Results are shown in Fig. 2. While the early component showed a