Results ponded to intracellular depolarizations in terms of latency,

54 C point where the membrane potential deviated from the Effective stimulation required that the stimulating elec- resting potential. Amplitude was measured at the first peak trode be placed specifically in the subcortical region e.g., arrows in Fig. 3C and D. For extracellular fast anterior to the hippocampus SC grey area in Fig. 1A; potentials, onset latency could not be measured because mean distance, measured in a straight line between record- the stimulus artifact often lasted into the beginning of the ing and stimulating electrodes, was 1148644 mm that response. Amplitude was measured at the maximum peak contains the fibers of the auditory thalamocortical pathway of the field potential. For intracellular and extracellular [12,43,60,73]. Stimulation of the hippocampus itself or of slow potentials, onset latency was measured at the point the striatum at positions greater than about 1 mm anterior where the slow potential deviated from baseline e.g., Fig. to the hippocampus elicited weak or no response. Thus, SC 1B, Subcortical Extracellular Control. In cases where the stimulation in the present study likely activated potential did not recover to baseline after the fast potential thalamocortical fibers and possibly corticofugal axons, but e.g., Fig. 1B, Intracortical Extracellular Control, onset see below and Discussion. latency was measured at the inflection point after the fast For experiment 1, intracellular data derive from 15 potential. The slow potential’s duration was measured slices and extracellular data derive from 35 slices. In from the onset to the point at which the potential returned intracellular recordings, SC stimulation elicited a fast to baseline. Magnitude was measured as the area under that EPSP followed by a slow, long-lasting depolarization Fig. curve. Variability is expressed as 61 Standard Error of the 1B, intracellular control trace in response to subcortical mean. Statistical comparisons are unpaired t-tests except stimulation. Simultaneous extracellular recordings re- where noted. vealed corresponding fast and slow negative potentials Fig. 1B, extracellular control trace in response to subcorti- cal stimulation. Extracellular negativities generally corres-

3. Results ponded to intracellular depolarizations in terms of latency,

shape, and duration cf. intracellular and extracellular The results are divided into two sets of experiments. The traces in Fig. 1B, implying a common cellular basis. initial experiments were designed to separately activate Although the fast intracellular EPSP to SC stimulation extrinsic subcortical and intrinsic intracortical inputs often consisted of 2 or 3 depolarizing peaks, our aim was leading to AC. We placed one stimulating electrode to examine thalamocortical, monosynaptic responses; subcortically, within the downstream part of the auditory therefore, for quantitative analysis, we measured the first thalamocortical pathway, and a second stimulating elec- peak with the shortest latency and refer to this as the trode within the middle layers of the cortex lateral to the SC-fast potential. The SC-fast potential had a consistent recording electrode Fig. 1A. Recordings were made in onset latency 3.160.3 ms, initial slope 1.760.3 mV ms, layer IV of AC at the site of the maximal field response to peak latency 6.960.7 ms and amplitude 4.260.8 mV subcortical stimulation and the effects of carbachol were and produces a dominant current sink in layers III IV [43]. examined. The results of these first experiments comprise A qualitatively similar potential in AC occurs in response the majority of the dataset in this manuscript. They to stimulation of the MG itself [12,43]; and see Experi- revealed, among other things, a strong differential ment 2 below. These characteristics, along with pharma- cholineric modulation of responses to subcortical vs. cological data presented below, suggest that the SC-fast intracortical inputs to AC. We hypothesize that this dif- potential is a thalamocortical EPSP. Following the SC-fast ferential effect could be due to differences between EPSP was a slower, long-lasting depolarization with fast thalamocortical synapses activated by subcortical stimula- fluctuations and spikes. This will be referred to as the tion and intracortical synapses activated by intracortical SC-slow potential indicated by arrows in Fig. 1B. It had stimulation. However, it is possible that the subcortical a more variable latency to onset 33.863.1 ms, duration stimulus may have activated non-thalamic afferents, so we 501.6638.7 ms and magnitude 2331.86438.0 mV?ms. conducted a second, more limited study, in which the This variability was evident in a given slice from trial to auditory thalamus was stimulated directly, thus decreasing trial, as well as between slices. The temporal characteris- the likelihood of activating non-thalamic cortical afferents. tics and variability of the slow potential indicate that it is The cholinergic modulation of the resulting thalamocorti- polysynaptic see also [43]. cal responses in AC was then compared with that found for The extracellular SC-fast potential had a mean latency to the other inputs. Each of the two sets of experiments will peak of 6.960.3 ms, similar to that of the intracellular be discussed in turn. EPSP. The mean amplitude was 159.1612.9 mV, and both amplitude and latency displayed trial to trial consistency. 3.1. Experiment 1: subcortical vs. intracortical In contrast, the extracellular SC-slow potential had a responses and modulation variable onset 25.363.9 ms, duration 223.7623.8 ms and magnitude area513,21061390 mV ms. This vari- 3.1.1. Responses to subcortical stimulation ability was observed within slices from trial-to-trial, as Subcortical stimulation within the auditory thalamocorti- well as between slices. cal pathway elicited robust responses in AC Fig. 1B. The SC-fast and slow potentials were elicited at differ- C .Y. Hsieh et al. Brain Research 880 2000 51 –64 55 ent stimulus intensities, indicating different thresholds for 10–25 mA elicited only the fast potential while higher generation. In 26 35 slices 74 where SC stimulation intensities elicited both fast and slow potentials. In some elicited extracellular fast and slow potentials, low stimulus slices, still higher intensities caused a reduction of the intensities e.g., 10–25 mA elicited the fast potential alone IC-slow potential, possibly due to the recruitment of whereas higher intensities elicited both potentials. In the intracortical inhibition see Discussion. remaining nine of 35 26 slices, SC stimuli at intensities Pharmacological manipulations indicated glutamate re- up to 100 mA elicited only fast potentials. ceptor involvement in the generation of IC responses, just Pharmacological manipulations revealed that glutamate as for SC responses. IC-slow potentials were almost receptors contribute to the generation of SC-fast and slow completely blocked by APV 50 mM, and subsequent potentials. Bath application of APV 50 mM, an NMDA application of CNQX 20 mM greatly reduced the fast receptor antagonist, reduced the slow potential nearly potentials Fig. 1B. Together, the physiological and completely while having little effect on the fast potential pharmacological data indicated that similar mechanisms Fig. 1B, Subcortical APV. Subsequent application of underlie SC and IC potentials. CNQX 10 mM, an AMPA KA receptor antagonist, reduced the fast potential Fig. 1B, Subcortical APV1 3.1.3. Subcortical and intracortical pathway CNQX. These data suggest that generation of the slow independence potential involves NMDA receptor activity and generation Because the responses elicited by stimulation of SC and of the fast potential involves AMPA KA receptor activity. IC sites had similar characteristics, it was necessary to Note, however, that our previous study [43] demonstrated demonstrate that SC and IC stimulation activated distinct that CNQX alone can also completely reduce the slow afferent pathways. To do this, we used a tetanus protocol potential. Thus, both NMDA and AMPA KA receptors to fatigue one pathway and then determined the response likely contribute to generating the slow potential see to stimulation of the other pathway Fig. 2. For the slice Discussion. in Fig. 2A, both SC and IC stimuli initially elicited extracellular fast and slow potentials Fig. 2A i. Re- 3.1.2. Responses to intracortical stimulation sponses to SC stimulation were then fatigued with a 50 Hz To activate IC afferents, a second stimulating electrode tetanus Fig. 2A ii. During the SC tetanus, a single was placed in the middle cortical layers up to 1 mm mean stimulus pulse delivered to the IC pathway elicited re- distance 620635 mm lateral to the recording electrode sponses similar to control Fig. 2A ii; stimulus artifacts Fig. 1A. Stimulation above or below the middle layers from the tetanus appear in both traces. Fifteen seconds generally elicited weaker responses, suggesting that in- after the SC tetanus, both SC and IC responses recovered tracortical stimuli activated fibers that project horizontally not shown. The protocol was then reversed: the IC within the middle layers. pathway was fatigued with tetanic stimulation, and a single The basic physiological and pharmacological properties pulse delivered to the SC pathway elicited responses of the responses elicited by stimulation of the IC pathway similar to control Fig. 2A iii. Fifteen seconds after the were similar to those described for the SC pathway. In IC tetanus, both SC and IC responses had recovered Fig. intracellular recordings, IC stimulation evoked a fast EPSP 2A iv. Fig. 2B shows the group effect of this procedure followed by a slow, long-duration depolarization Fig. 1B, on the amplitudes of the fast and slow potentials SC Intracortical. As with SC stimulation, the early part of the tetanus n54; IC tetanus n53. Because the tetanus had response included up to three short latency depolarizing little effect on responses to stimulation of the non-tetanized peaks but again, our quantitative analysis involved only the pathway, we conclude that the SC and IC pathways are first peak, which we call the IC-fast potential. The IC-fast largely independent, at least with respect to the mono- potential had a consistent onset latency 4.060.3 ms, synaptic fast potentials this conclusion cannot be extended slope 1.760.4 mV ms, peak latency 8.361.3 ms, to the slow potentials because of their polysynaptic nature, amplitude 3.260.4 mV, and shape, suggesting a mono- see Discussion. Given this demonstration, we could now synaptic response. In contrast, the IC-slow potential that determine the sensitivity of IC- and SC-evoked responses followed arrows in Fig. 1B had a more variable onset to cholinergic modulation. latency 28.263.2 ms, duration 454.0630.0 ms and magnitude 2312.86487.3 mV ms. This variability was 3.1.4. Effects of cholinergic agonist on intracellular seen from trial to trail and between cells, indicating a potentials polysynaptic response. In intracellular recordings, 10–50 mM carbachol gener- Extracellular responses to IC stimulation were negative ally had a suppressive effect on all of the synaptic in polarity and time-locked to the intracellular events Fig. potentials. However, there were significant differences in 1B, Intracortical. The extracellular fast potential had a the degree and manner of this suppression within and peak latency of 5.760.2 ms and amplitude of 132.9612.4 between pathways. First of all, within each pathway, mV, while the slow potential had an onset latency of carbachol reduced the slow potential significantly more 13.962.4 ms, duration of 188.8613.9 ms, and magnitude than the fast potential P’s50.0002; intracellular data from of 12,20061200 mV ms. Low intensity stimulation e.g. 10 and 50 mM carbachol were combined for statistical 56 C to make these comparisons for different doses of carbach- ol, we recorded extracellular potentials while alternating SC and IC stimulation, and applied carbachol at con- centrations ranging from 0.5 to 50 mM. 3.1.5. Dose-dependent effects of carbachol We determined the effects of 0.5, 1, 5, 10, 25 and 50 mM carbachol on extracellular responses to SC and IC stimuli interleaved at 15–30 s intervals. In 11 35 slices, more than one dose was applied to the same slice. Fig. 4 depicts the time course and magnitude of carbachol-in- duced effects at each dose and on each of the four synaptic responses i.e., on SC- and IC-elicited fast and slow potentials. A differential effect on fast vs. slow potentials was first apparent at 5 mM, which reduced the SC- and IC-slow potentials 70.266.9 and 63.768.1, respective- ly P,0.001 but did not significantly affect the fast potentials P.0.05. Higher concentrations of carbachol 10 mM reduced both slow and fast potentials but the slow potentials were always affected more strongly P, 0.05, paired t-test. In addition, reduction of the fast potential took longer to develop than that of the slow potential. This was quantified for the 50 mM effects dose; the maximal reduction of the slow potential occurred 3.0560.58 min before that of the fast potential P,0.001, paired t-test; these differences are not apparent in Fig. 4 because of the variable onset latency of carbachol’s effects across slices. Thus, carbachol reduced the slow potentials at a lower dose, to a greater degree, and more rapidly, than it reduced the fast potentials Figs. 4 and 5. Differential cholinergic modulation of the extracellular SC and IC fast potentials Figs. 4 and 5 was similar to that observed in intracellular experiments Fig. 3C and D. At 10 mM, carbachol reduced both fast potentials sig- Fig. 2. Independence of subcortical and intracortical pathways. A i nificantly P,0.05, but reduced the IC-fast potential to a Single stimuli at arrows elicited control SC- and IC-evoked potentials. greater degree than the SC-fast potential SC reduction5 ii During 50 Hz stimulation to fatigue SC-elicited potentials, a single 6.662.0; IC reduction515.163.7; P,0.05, paired t- pulse delivered to the IC pathway elicited robust potentials. iii Follow- ing recovery from the tetanus not shown, 50 Hz stimulation then test. The differential reduction of fast potentials was fatigued IC-elicited potentials, and a single pulse delivered to the SC enhanced at higher carbachol concentrations Fig. 5. pathway elicited fast and slow potentials similar to control. iv Recovery Since the reduction of synaptic potentials can occur as a of all potentials after tetanic stimulation. B Tetanic stimulation of SC result of membrane depolarization which would reduce n54 slices or IC n53 slices pathways strongly reduced responses to excitatory driving force, we determined the effects of single pulse stimulation of the tetanized pathway P’s,0.01 but not the non-tetanized pathway P’s.0.1. carbachol on membrane potential for two doses. At 10 mM, there was a negligible depolarization of the mem- brane potential mean51.461.4 mV; n56; P.0.3. In tests, but individual dose effects are shown in Table 1. contrast, 50 mM carbachol significantly depolarized the Second, between pathways, carbachol reduced the IC-fast membrane potential mean54.360.9 mV; n58; P,0.01. potential more than the SC-fast potential P,0.02; see However, the carbachol-induced suppression of the synap- Table 1. Both of these differential effects are illustrated in tic responses remained during repolarization of the mem- the examples shown in Fig. 3. Carbachol had no effect on brane potential back to baseline via intracellular current EPSP onset latencies in either pathway P’s.0.3; mean injection n53; data not shown. difference for SC520.0260.02 ms, IC50.0860.08 ms. To determine the subtype of ACh receptors involved in For most of these intracellular recordings, data were the reduction of synaptic responses, we applied the mus- obtained with either SC or IC stimulation, not both. To carinic receptor antagonist atropine. Atropine 0.5 mM obtain within-slice comparisons of carbachol’s effect and blocked the effects of 50 mM carbachol on the evoked C .Y. Hsieh et al. Brain Research 880 2000 51 –64 57 Table 1 Effects of carbachol on intracellular potentials decrease 1S.E. Subcortical Intracortical Fast Slow Fast Slow potential potential potential potential Amplitude Slope Magnitude Amplitude Slope Magnitude 10 mM 15.466.8 12.865.8 94.862.3 36.1613.1 33.3613.0 89.262.7 n56 n56 n56 n54 n54 n53 50 mM 14.765.5 18.567.7 97.961.2 34.069.2 34.1614.9 92.767.2 n55 n55 n55 n54 n54 n53 synaptic responses Fig. 5, indicating the involvement of nated population of thalamocortical synapses. To do this, muscarinic ACh receptors in carbachol’s actions. we developed a fully intact thalamocortical preparation in which direct MG stimulation produced robust cortical 3.2. Experiment 2: cholinergic modulation of responses [12,43]. In such intact slices, we compared the thalamocortical responses effects of 25 mM carbachol on cortical responses evoked by MG stimulation, alternating with SC or IC stimulation Because the subcortical stimulating electrodes in the MG vs. SC, n58; MG vs. IC, n57. above experiments were placed in the auditory Results of these experiments are shown in Fig. 6. As thalamocortical pathway rather than in the MG itself, they before, the slow potentials were suppressed more than the may have also activated cortical afferents originating fast potentials for all pathways P’s,0.0001. Also, as outside the auditory thalamus. Stimulating the MG directly before, both the IC- and SC-evoked fast potentials were would increase the likelihood of activating an uncontami- suppressed by carbachol P’s,0.01 and this suppression Fig. 3. Differential effects of carbachol on synaptic responses. A Carbachol 10 mM strongly reduced the SC-slow potentials both intracellular and extracellular but not the SC-fast potentials. All potentials recovered upon wash. B Similarly, 10 mM carbachol reduced both the IC-fast and slow potentials, with lesser effects on the IC-fast potentials. C and D Higher-resolution views of intracellular records in A and B reveal that carbachol reduced the IC-fast EPSP but not the SC-fast EPSP arrows indicate peak EPSPs. Data in A, C are from the same cell; data from B, D are from another cell. 58 C Fig. 5. Group effects of carbachol on fast and slow potentials, derived from Fig. 4 data. Values are means of carbachol’s maximal effects average of 2 min on fast potential amplitudes and slow potential magnitudes. Atropine, 0.5 mM, n53 blocked all effects of 50 mM carbachol. sponses in either full or partial thalamocortical prepara- Fig. 4. Dose dependence and time course of carbachol’s actions on fast tions. These two cholinergic enhancements contributed to and slow potentials. Fast potential amplitudes solid circles and slow the failure of the average MG data to exhibit significant potential magnitudes open circles were normalized to mean of 20 suppression Fig. 6A i; P50.90. Examples of an responses preceding carbachol 2 min pulse starting at arrow at con- individual enhancement and a more typical effect are centrations of 0.5 1, 5, 10, 25, or 50 mM. Neither 0.5 nor 1 mM shown in Fig. 6A ii and iii. carbachol had any effect; these data were therefore combined. For all concentrations, SC fast potential n57–12, SC slow potential n55–9, IC fast potential n54–14, IC slow potential n53–6.

4. Discussion