当前位置: 首页 > 期刊 > 《生理学报》 > 2005年第15期
编号:11326282
Paired-pulse transcranial magnetic stimulation protocol applied to visual cortex of anaesthetized cat: effects on visually evoked single-uni
http://www.100md.com 《生理学报》 2005年第15期
     1 Department of Neurophysiology, Ruhr-University Bochum, 44780 Bochum, Germany

    Abstract

    In this study, we tested the paired-pulse transcranial magnetic stimulation (ppTMS) protocol – a conditioning stimulus (CS) given at variable intervals prior to a test stimulus (TS) – for visually evoked single-unit activity in cat primary visual cortex. We defined the TS as being supra-threshold when it caused a significant increase or decrease in the visually evoked activity. By systematically varying the interstimulus interval (ISI) between 2 and 30 ms and the strength of CS within the range 15–130% of TS, we found a clear dependence of the ppTMS effect on CS strength but little relation to ISI. The CS effect was strongest with an ISI of 3 ms and steadily declined for longer ISIs. A switch from enhancement of intracortical inhibition at short ISIs (2–5 ms, SICI) to intracortical facilitation (ICF) at longer ISIs (7–30 ms), as demonstrated for human motor cortex, was not evident. Whether the CS caused facilitation or suppression of the TS effect mainly depended on the strength of CS and the polarity of the TS effect: within a range of 60–130% a positive correlation between ppTMS and TS effect was evident, resulting in a stronger facilitation if the TS caused facilitation of visual activity, and more suppression if the TS was suppressive by itself. The correlation inverted when CS was reduced to 15–30%. The ppTMS effect was not simply the sum of the CS and TS effect, it was much smaller at weak CS strength (15–50%) but stronger than the sum of CS and TS effects at CS strength 60–100%. Differences in the physiological state between sensory and motor cortices and the interactions of paired synaptic inputs are discussed as possible reasons for the partly different effects of ppTMS in cat visual cortex and human motor cortex.
, http://www.100md.com
    Introduction

    Paired-pulse transcranial magnetic stimulation (ppTMS) has been shown to be a useful tool for examining the excitability of inhibitory and excitatory interactions in human motor cortex. At present, the most widely used technique is based on a conditioning test design which was originally introduced by Kujirai et al. (1993): a subthreshold conditioning stimulus (CS) is applied prior to a supra-threshold test stimulus (TS). The effect of the conditioning stimulus depends critically on the interstimulus interval (ISI) between CS and TS but also on CS strength. The motor-evoked response (MEP) elicited by the test stimulus is reduced if CS precedes TS by approximately 1–4 ms, while facilitation of the MEP results in ISIs of 7–20 ms (Kujirai et al. 1993; Tokimura et al. 1996; Ziemann et al. 1996; Nakamura et al. 1997; Di Lazzaro et al. 1999a). This phenomenon is commonly referred to as intracortical inhibition (ICI) and facilitation (ICF), since it is thought that either inhibitory or excitatory intracortical connections to the pyramidal tract neurones (PTNs) are activated in an ISI-dependent way (Kujirai et al. 1993; Ziemann et al. 1996). More recently, this kind of inhibition has been termed short-interval intracortical inhibition (SICI) to distinguish it from another kind of cortically mediated inhibition observed with longer (50–200 ms) interstimulus intervals (LICI, Sanger et al. 2001).
, 百拇医药
    A clear SICI is present only if CS is low (60–80% of TS). Increasing the CS strength to that of TS, or combining a supra-threshold CS with a slightly subthreshold TS, also leads to ICF at short ISIs (Ziemann et al. 1998; Di Lazzaro et al. 1999b; Ilic et al. 2002). Based on the assumption that TMS preferentially stimulates horizontally orientated intracortical neuronal elements and related to the finding that ICF generally needs a stronger CS to occur than SICI (> 80%, Kujirai et al. 1993; Ziemann et al. 1996; Chen et al. 1998; Awiszus et al. 1999), Ilic et al. (2002) proposed a model, composed of a short low-threshold inhibitory pathway and a high-threshold polysynaptic excitatory intracortical pathway.
, http://www.100md.com
    Recently, we have shown that it is possible to record cortical single-unit spike activity at the centre peak of the magnetic field induced by a figure-of-8 coil (Moliadze et al. 2003). In a first attempt, we analysed the interaction of single TMS pulses with visually evoked responses in the primary visual cortex of anaesthetized cats by varying the strength of TMS and the interval between magnetic and visual stimulation. In the present study, we systematically tested the effect of pairing a conditioning stimulus of different strength at different intervals with a test stimulus which either facilitates or inhibits visually evoked activity. In this way, we attempted to obtain a better insight into the cellular and network activities elicited by distinct TMS protocols which could be used in future in vivo studies concerned with the effect of TMS on sensory processing.
, http://www.100md.com
    Methods

    General procedures

    All experimental procedures were permitted under the local government guidelines for animal welfare (No. 50.8735/81.6), and in addition conformed to legal requirements in the EU, UK and the US.

    Surgical and electrophysiological procedures were in principle the same as in the previous study (Moliadze et al. 2003). In brief, all surgical procedures allowing the maintenance of the experimental cats in terms of artificial ventilation, relaxation and nutrition during the 5 day recording sessions were performed under deep anaesthesia with a combination of ketamine (20 mg kg–1 I.M., Ketanest, Parke-Davies, Germany) and xylazine (2 mg kg–1 I.M., Rompun, Bayer, Germany). All incisions and pressure points were also locally anaesthetized by xylocaine (2%, Astra Chemicals, Germany). The level of anaesthesia before and during surgery was tested with the toe-pinch and pinna reflex. Absence of motor reactions and stable blood pressure were taken as signs of sufficient anaesthesia. The relaxation and hydration of cats was achieved by continuous infusion of alcuronium chloride (0.15 mg kg–1 h–1, Alloferin 10, Hoffmann-La Roche, Germany) in 1% glucose-Ringer solution through the femoral artery at a rate of 6 ml h–1. The degree of neuromuscular blockade sufficient to avoid active respiration by the animal was assessed by monitoring the time course of expiratory CO2 change during artificial respiration (Datex Normocap 200, Hoyer, Bremen, Germany). The femoral catheter also allowed the measurement of heart rate and arterial blood pressure. Rectal body temperature was measured and kept at about 38.5°C by the aid of a heat blanket. Continuous anaesthesia during recording sessions was guaranteed by artificial respiration with N2O–O2 (70%: 30%) and halothane (0.6–2.5% Fluothane, ICI-Pharma, Germany) through a catheter introduced into the trachea. Adequate ventilation and anaesthesia of the animal was assured via analysis of the spectral composition of the EEG (presence of alpha and delta waves) and via the level and time course of blood pressure, heart rate and end-expiratory CO2. Body temperature, blood pressure, heart rate, end-expiratory CO2, and the inspiratory levels of O2 and N2O were continuously measured and stored by a supervision monitor. For each of these parameters, upper and lower limits could be set for absolute values, rate of change and for the forecast of a critical state by trend analysis. The crossing of one of these limits triggered an alarm in the laboratory which enabled fast-as-possible emergency actions around the clock. EEG recordings were also monitored and stored on hard disk simultaneously with the single-unit recordings.
, 百拇医药
    The optics were corrected for a viewing distance of 56 cm with contact lenses of 5–7D. Atropine sulphate (1%, Atropin-Pos, Ursapharm, Germany) and phenylephrine hydrochloride (5%, Neosynephrin-Pos, Ursapharm, Germany) were applied topically for mydriasis and retraction of the nictitating membranes. Isoptomax (Alcon Pharma, Germany) was topically applied to the cornea to prevent infections. Craniotomies provided access to area 17 of the right hemisphere for single-unit recording, and to area 18 of the left hemisphere for epidural EEG recording via a 0.5 mm silver ball electrode. Recordings were continued as long as the animal could be maintained in a physiological state, with blood pressure above 80 mmHg, end-expiratory CO2 between 3.8 and 5.5%, body temperature around 38.5°C and the EEG showing a normal pattern with episodic changes in spectral composition due to brainstem activity. At the end of each experiment, the animal was deeply anaesthetized by maximal halothane concentration (4 vol.%) and perfused with cold (4°C) Ringer solution followed by 4% paraformaldehyde, to enable further histological studies on the brain tissue.
, 百拇医药
    Recordings and visual stimulation

    Electrophysiological procedures and visual stimulation were also the same as in the previous study. Extracellular recordings of single-unit activity were made within area 17 at the top of the gyrus, corresponding to visual field positions around 5 deg lower and 5 deg lateral to area centralis. Visual stimulation of the receptive fields of individual neurones was achieved by moving an optimally sized and orientated (preferred orientation) bright bar, presented monocularly, forth and back across the receptive field within 3 s. Visual stimuli were generated by a PC-based visual stimulator (‘Leonardo’, Lohmann Research Equipment, Germany) and presented on a 21 inch CRT monitor at a refresh rate of 100 Hz. For further details, see Methods section of the previous paper (Moliadze et al. 2003).
, 百拇医药
    TMS protocols

    Paired magnetic pulses were generated by two MagStim 200 and a BiStim module (The Magstim Company, Whitland, Dyfed, UK) and applied to the occipital cortex of cats via a figure-of-eight coil (2 mm x 70 mm in one plane, monophasic pulse, The Magstim Company). As one difference to the previous study we tested a special version of the 2 mm x 70 mm figure-of-8 coil in two of the four cat experiments. The coil built by the MagStim Company on our request is identical to the conventional 2 mm x 70 mm coil except for a central guiding tube of 5 mm in diameter at the junction of the two coils to enable a microelectrode to be lowered through the coil into the brain at the centre peak of the magnetic field. When using the conventional figure-of-8 coil, the electrode was lowered to the cortex by moving it tangentially to the lower surface of the coil with the coil tilted by 45 deg towards the neck of the cat. The coil was centred above area 17/18 of the right hemisphere with the handle pointing to the left hemisphere, thereby inducing a mediolateral current in the dorsally exposed area 17/18 of the right hemisphere. We could not find any significant differences in the threshold for inducing excitatory or inhibitory effects on visually evoked activity for opposite current directions.
, 百拇医药
    Prior to the paired pulse protocols, for each new recording site we determined the strength of the test stimulus (TS) needed to achieve a reliable change in visually evoked activity. Therefore, the TMS pulse was delivered shortly before (20–40 ms) the onset of a visual response (see Fig. 1A), so that the maximum visually evoked activity was close to 100 ms after TMS, an episode that was found in the previous study (Moliadze et al. 2003) to be most likely to lead to facilitation of visual (and spontaneous) activity, but also inhibition in some cases. As in the former study, trials with three different stimulus conditions were interleaved and each repeated 20 times: one trial with visual stimulation only (moving bar), one trial with combined visual and magnetic stimulation, and one trial with TMS only. Different strengths of TMS were tested and the stimulation strength changing the rate of activity by about 10–30% in two sets of recordings using the same stimulation protocol was taken as the test stimulus (TS) strength for ppTMS. TS strength was within 30–50% of maximal stimulator output. The TS had a facilitatory effect on visual activity in about two-thirds (n = 26) of the 42 cells analysed; in the remainder it was suppressive.
, 百拇医药
    A, peristimulus time histogram (PSTH) showing single-unit activity evoked by an optimally orientated bright bar moving back and forth across the receptive field of the unit as indicated by the drawing below the diagram. TMS was given close to visually enhanced activity (see arrow). B, mean activity within a time window of 500 ms following TMS (time of suprathreshold test stimulus (TS), the second stimulus in the case of ppTMS). Curves 1–3 are: (1) activity after TS only, (2) activity following ppTMS (in this case a conditioning stimulus (CS) of 60% strength of TS, given 4 ms prior to TS), (3) difference between curves 1 and 2 (2 minus 1). Activity rates were calculated as the mean of 20 identical trials. Three different time windows following TMS were analysed (as indicated by the dashed lines): 30–100 ms, 100–200 ms and 200–500 ms. For further analysis, the activity within these time windows was averaged for each condition.
, 百拇医药
    Paired-pulse TMS was performed with varying levels of conditioning stimulus (CS) strength. We included only those cells in the analysis (n = 42) for which we could test at least three different CS strengths, including one or two that were subthreshold and one of about TS strength. In 20 cells we could test five different CS strength ranges. Finally, the data were divided into five CS strength groups: 15–30%, 40–50%, 60–90%, 100% and 110–130% of TS strength. For a given combination of CS and TS strength ppTMS was tested with 2, 3, 4, 5, 6, 7, 8, 10, 12, 15, 20 and 30 ms intervals between CS and TS (ISI). An ISI of 1 ms was not tested because the effect would be dominated by refractoriness of the cell membrane (Fisher et al. 2002). Each ISI, TS alone and visual stimulation alone was tested 20 times and stimulus conditions were presented in an interleaved and quasi-random fashion. In this way we tried to minimize the effects of varying responsiveness of neurones over time. ISI were generated by triggering the two MagStim 200 units separately with the aid of a Spike2 script and the DAC output channels of a CED 1401 interface (Cambridge Electronic Devices, Cambridge, UK). In addition, the protocol included one trial with TS only and one trial without TMS but with visual stimulation only. Each trial lasted 6 s, 3 s with recording and visual stimulation by the moving bar and another 3 s without stimulation to allow the neuronal system to resettle and also to increase the interval between TMS applications.
, http://www.100md.com
    Data analysis

    Quantitative analysis of the ppTMS effect was carried out in the following way: first, the effect of the TS itself was checked by subtracting the activity with visual stimulation only from the activity obtained with combined visual and magnetic stimulation, yielding both the amount and the polarity of change (facilitation or suppression) with TS. Next, the additional effect of the CS was determined – as shown in Fig. 1B – by subtracting the activity evoked by TS + visual stimulation (curve 1) from the activity evoked with ppTMS + visual stimulation (curve 2) resulting in curve 3. This additional change in visual activity caused by the CS (ppTMS) was set in relation to the effect of TS only (e.g. see Fig. 2). In addition, we checked whether the ppTMS effect is the linear sum of the individual effects of CS and TS by subtracting the sum of the changes in visual activity obtained with CS and TS from the change in visual activity achieved with ppTMS (Dpp = vis,ppTMS – (vis,CS + vis,TS)). Since the effects of TMS in the cortical network are likely to change with the delay after the test pulse, e.g. by switching between facilitation and suppression (Moliadze et al. 2003), we calculated mean rate of spike activity in three different time windows following TMS: 30–100 ms, 100–200 ms and 200–500 ms (indicated in Fig. 1B). Time 0–30 ms after onset of TMS was excluded from the analysis because of possible contamination by the TMS artefact. Statistical significance (P < 0.05) of mean changes in activity achieved with ppTMS as compared to TS were determined with Student's paired t test. The paired version of the test was chosen because ppTMS data are coupled to TS control data via the rate of visually evoked activity which varies from cell to cell. Accordingly, the error bars given in Fig. 4 refer to the standard deviation of differences between ppTMS and TS effects.
, 百拇医药
    Scatter plots show the relationship between changes in visual activity resulting from single-pulse TMS (only the test stimulus (TS), abscissa) and additional changes achieved with ppTMS (ordinate). The effect of ppTMS was measured as the difference from the TS effect (ppTMS-elicited activity minus TS-elicited activity). Separate diagrams are shown for each ISI between CS and TS tested (2–30 ms), with CS strength in the range of 60–90% of TS strength. Data points give mean activity in spikes s–1 within 30–100 ms after TMS (time window 1) for 51 measurements. A clear correlation between ppTMS and TS effect is evident for all ISIs. Slopes of the regression lines and Pearson correlation values are given in Fig. 3. Pearson correlation was significant in all cases with < 0.001.
, 百拇医药
    For the 5 different ranges of CS strength analysed, mean change in visual activity caused by ppTMS is plotted versus the ISI used in ppTMS. Mean rates of activity were calculated separately for cases in which visual activity was either facilitated (black curve) or suppressed (grey curve) by the TS alone. Mean change in visual activity caused by TS only is indicated by the large dots (labelled TS at abscissa). Number of measurements averaged for each condition are indicated close to the curves. Potentiation of TS facilitation is present with CS strength of 60–130% and most pronounced around an ISI of 3 ms, while diminution of TS facilitation is found for CS 15–30%. Suppression of visual activity by TS is less enhanced by a conditioning stimulus (grey curve). Data points labelled by asterisks are significantly different from TS values given by the large dots (P < 0.05, one-sided paired t test). Error bars give S.D. of differences between ppTMS and TS effect.
, 百拇医药
    Results

    Forty-two cells recorded in four cats were analysed with the paired-pulse protocol using interstimulus intervals (ISIs) between 2 and 30 ms and three to five different conditioning stimulus (CS) strengths. The TMS test stimulus (TS) induced either facilitation or suppression of visually evoked activity and the effect was variable in strength. In addition, the strength of the visual response itself varied from cell to cell. Cells were recorded primarily from cortical layers 3–6 and included simple and complex types. As in the previous study, we could not find a layer- or cell-type-specific TMS effect and therefore pooled all cells in this analysis.
, 百拇医药
    The experimental conditions for ppTMS in our study are somewhat different from those in human motor cortex. In the motor system, TS (or a single TMS pulse in general) always results in increased motor output, the efferent activity is artificially induced and is the result of the concerted and almost synchronous action of many hundreds or thousands of pyramidal cells and the amplitude of the motor response can be more precisely adjusted by TMS strength. The situation is different in our approach: here, visual activity evoked by a moving bright bar interacts with an artificially evoked volley of activity within the cortical network. Moreover, we do not measure the summed output activity of cortical columns but a single cell's activity within the column. Accordingly, we are confronted with highly variable activities, both in response to visual stimulation and as a result of TMS (see also Moliadze et al. 2003).
, 百拇医药
    In our first attempt, we therefore analysed the relationship between ppTMS and TS effects on visually evoked activity at the single-cell level by plotting the additional change in visual activity resulting from the combination of CS with TS (ppTMS) versus the change in activity caused by the TS alone. Thus, activity with ppTMS minus the activity with TS is plotted on the ordinate of the diagrams in Figs 2 and 3A, and activity with TS minus sole visual activity (without TMS) is plotted along the abscissa. This was done separately for each of the five CS strength ranges (15–30%, 40–50%, 60–90%, 100% and 110–130%) and the 12 ISIs tested, resulting in 60 scatter plots. In Fig. 2, plots are shown only for the CS strength range 60–90%, but for each of the 12 ISIs. The distribution of data points in these diagrams indicates a positive correlation between ppTMS and TS effects: the stronger the TS effect, the stronger also the effect of adding the CS, both for facilitation and for suppression of visual activity by the TS. The distribution of data points did not vary much with changing ISI and this also holds for the other four ranges of CS strength tested. Therefore in Fig. 3A, five different CS strength ranges are compared for only one interstimulus interval (ISI 3 ms). A comparison of the five diagrams in Fig. 3A clearly shows that the correlation between ppTMS effect and TS effect depends on CS strength: for the lowest range of CS (15–30%) the correlation is negative, indicating that the conditioning stimulus had weakened the TS effect, in both facilitation and suppression. With increasing CS strength, the correlation between ppTMS and TS effect became positive and facilitation of the TS effect by the CS increased up to saturation at CS 100–130%. The correlation between the effects of ppTMS and TS was quantitatively analysed by calculating the slope of the regression lines and Pearson's correlation coefficients. For each stimulus condition these two values are given in the diagrams of Fig. 3B by plotting slope and correlation coefficient versus ISI for each of the five CS ranges. The slope of the regression line continuously increases with increasing CS strength but also saturates around 100%. The effect is relatively independent of the ISI; only for high CS strengths (100% and 110–130%) is the slope clearly higher with short ISIs (2–6 ms). Pearson's correlation coefficients are less dependent on the ISI for all cases of CS strength and reach statistical significance in all cases except for the CS range 40–50% for which the slope of the regression lines was very low (about 0.1, < 0.01 for all ISIs in CS range 15–30%, < 0.001 for all ISIs in CS ranges 60–90%, 100% and 110–130%, < 0.05 only for ISI 4 ms in the CS range 40–50%, with values taken from statistic reference tables).
, http://www.100md.com
    A, as in Fig. 2, scatter plots show the relationship between ppTMS- and TS-induced changes in visual activity. Each of the scatter plots shows data for ppTMS with ISI of 3 ms only, but for 5 different ranges of CS strength. Changing the ISI did not affect the correlation as shown in Fig. 2 and panel B of this figure. B, diagrams showing the dependence of slope of the regression line (black curves) and the dependence of Pearson correlation (grey curves) on ISI for each of the 5 ranges of CS strength. Pearson correlation was significant for CS range 15–30% ( < 0.01) and for the range 60–130% ( < 0.001).
, http://www.100md.com
    As a next step, the mean change in visual activity caused by ppTMS with different ISIs in comparison with a single TS was calculated for the five ranges of CS strength. The results are shown in Fig. 4. Since the TS could either facilitate or suppress visual activity and considering the correlation between the ppTMS and TS effect, we separately pooled the data for cases with either TS facilitation (black curve) or TS suppression (grey curve). The mean change in visual activity caused by the TS alone is given by the large dots to the left of ISI 2 ms in the diagrams. Both the mean increase (around 40 spikes s–1) and mean decrease (about 20 spikes s–1) in visual activity are almost identical for the samples with different CS strength range, indicating that activity levels did not vary much between recording sessions, thereby enabling a stable TS reference activity level. For a given strength, the conditioning stimulus resulted either in a weakening or an augmentation of the TS, independent of the ISI between CS and TS. Indications of an ISI-dependent improvement in either intracortical inhibition (ICI, SICI) or facilitation (ICF) as described for human motor cortex could not be found. A significant weakening of facilitation by TS was found only for relatively weak CS, in the range of 15–30% of TS strength. A CS strength of 40–50% had no effect on average, while slightly subthreshold CS (60–90%) and clearly supra-threshold CS (100% and 110–130%) significantly augmented the facilitation of visual activity by the TS. The effect increased with CS strength up to the highest level of 110–130%, was strongest with ISIs of 2–6 ms (peak effect at 3 ms ISI) and declined with longer ISI. The CS effect on the reduction of visual activity by the TS was less clear although significant in most cases. A moderate increase in the suppression of visual activity was found with CS strength 40–50% and 60–90% and a clearly stronger suppression with 110–130% CS strength.
, 百拇医药
    To test whether the changes in visual activity achieved with ppTMS resemble the (linear) sum of the individual effects of the two stimuli (CS + TS), we carried out the following calculation:

    If CS and TS effects linearly sum up to result in the ppTMS effect, Dpp should be zero. This analysis could be performed for 25 cells for which the effect of different CS strengths had been tested before the ppTMS protocol in order to test the threshold strength for TS. Figure 5 compares the mean changes in visual activity achieved with different CS strengths with the changes obtained with TS (always 100%) and ppTMS (ISI 3 ms) in each CS group. In addition, the mean difference Dpp calculated for each cell is shown. As before, cases with either facilitation (n = 16 cells) or suppression (n = 9 cells) of visual activity by the TS were analysed separately. The diagrams demonstrate three effects. First, and as should be expected, the effect of TS is almost identical in each CS group, on average an increase of about 40 spikes s–1 in the case of TS facilitation and a decrease of about 20 spikes s–1 with TS suppression. Second, the change in ppTMS effect with increasing CS parallels that of the CS itself, with about the same slope for both curves (obvious for TS facilitation, less clear for TS suppression). Third, the difference Dpp is not zero and clearly depends on CS strength, at least in the case of facilitation of visual activity with TS: the ppTMS effect is less than the sum of the CS and TS effects for low CS strength (15–50%) but higher for the upper CS strength range (60–100%), replicating the flip from a negative correlation between the ppTMS and the TS effects at low CS strength to a positive correlation for stronger CS as shown in Fig. 3A.
, 百拇医药
    For 25 neurones in which at least 2 CS strength effects have been recorded, the mean effects of single CS of different strength are compared with the mean effects of TS alone, ppTMS with 3 ms and the difference between ppTMS effect and the sum of the individual CS and TS effects (Dpp = vis,ppTMS – (vis,CS + vis,TS)). The upper diagram shows the mean values for those 16 cells in which the TS caused facilitation of visual activity; the lower diagram shows the means of 9 cells with TS suppression.
, 百拇医药
    The data presented so far refer to visual activity within a time window of 30–100 ms following the TMS artefact. Data prior to 30 ms were excluded because of possible contamination with the decaying artefact. Visual inspection of all records revealed that the DC level of background noise had reached control levels 30 ms after TMS onset. The same analyses as documented by Figs 2–4 were also carried out for time windows 100–200 ms and 200–500 ms after TMS. A comparison of the three time windows is given in Fig. 6. As for time window 30–100 ms, we found no dependence of the ppTMS effect on ISIs in the other two time windows and therefore we averaged the data obtained with different ISIs. The strong dependence of facilitation of visual activity on CS intensity and the correlation between the effects of ppTMS and TS evident for time window 30–100 ms, are to a lesser extent still visible in time window 100–200 ms, but statistically significant changes occurred only with CS strength 60–90% and in part with 100%. A ppTMS effect outlasting 200 ms (time window 200–500 ms) was not evident. Interestingly, the ratio of facilitation to suppression of visual activity by the TS changed from 1.6 in time window 1 to 1.2 in time window 2 and finally to 0.9 in the third time window. Thus, facilitation of visual activity prevailed from 30 to 200 ms after TMS while the likelihood of suppression increased thereafter. This finding is consistent with those of the previous study (Moliadze et al. 2003).
, http://www.100md.com
    The data shown in Figs 2–5 were obtained from a time window of 30–100 ms after TMS. The same analyses were carried out for time windows 2 and 3, 100–200 ms and 200–500 ms after TMS, respectively. ppTMS effect on the facilitation (black continuous curve) or suppression (grey continuous curve) obtained with TS, and the characteristics of correlation between ppTMS and TS effect, slope of regression lines (black dashed) and Pearson correlation coefficients (grey dashed) are plotted versus CS strength separately for the 3 different time windows. Data obtained with different ISIs were averaged because of the small quantitative and absent qualitative differences. Thus, curves for time window 30–100 ms give the averaged values of the curves of Figs 3B and 4. Asterisks indicate data points with a significant difference compared to TS control values (black: P < 0.05 with paired t test; < 0.05 for Pearson coefficient).
, 百拇医药
    Discussion

    General findings

    The results of this study reveal that a conditioning stimulus (CS) applied at different intensity and time prior to a supra-threshold TMS test stimulus (TS) modifies the effect of the TS on visually evoked activity in a distinct way. Contrary to the effects of ppTMS described for the human motor cortex, we found no dependence of the ppTMS effect on interstimulus interval (ISI). A moderately subthreshold CS (range 60–90% of TS) caused, on average, an augmentation of the TS effect: either an increased facilitation of visual activity by the TS or strengthened suppression of visual activity if the TS by itself had a suppressive effect. The ppTMS effect, however, varied with the strength of the CS. A facilitation of the TS effect by the CS was found only for CS strength 60–130%. Weaker CS (15–30%) resulted in an ISI-independent reduction of the TS effect, e.g. facilitation of visual activity by the TS was weakened. Furthermore, the ppTMS effect was not equal to the algebraic sum of the individual CS and TS effects. It was lower than the sum of CS and TS effects for low CS strength but stronger for peri-threshold CS.
, 百拇医药
    One aspect of our results was somewhat surprising and needs further discussion: a strengthening of the suppression of visual activity by the TS appeared with CS strength ranges 40–50%, 60–90% and 110–130%, but not with a CS of 100%. Why do we see no change in the amount of inhibition when the conditioning pulse is of the same strength as the test stimulus This phenomenon could be explained as follows: if two identical supra-threshold stimuli are applied one after the other, the same population of neurones will be stimulated but the effect will be different at excitatory and inhibitory synapses. In the likely case of shunting inhibition caused by activation of GABAA receptors, the first inhibitory volley may maximally suppress nearby excitatory postsynaptic potentials and a second volley acting at the same synapse may not add much effect (no temporal summation). A different population of neurones (axons) will be activated if CS and TS differ in strength, leading to a spatially different pattern of inhibition, and additional excitatory inputs may be depressed. On the contrary, excitatory postsynaptic potentials more likely show temporal summation – in part due to NMDA-mediated potentials – and will sum up even when the same inputs are activated by identical stimuli.
, 百拇医药
    Although these effects of ppTMS on single-unit activity in cat visual cortex are different from those observed for motor activity evoked in human M1 with ppTMS, our findings are in accordance with the model proposed by Ilic et al. (2002), which is composed of a low-threshold inhibitory pathway and a high-threshold excitatory pathway. A weak CS might have preferentially activated or pre-depolarized the inhibitory network which might then be potentiated by the following TS prior to the activation of excitatory connections. With increasing CS strength more and more excitatory neurones might be pre-activated. Although most studies using ppTMS were concerned with the human motor system, it is difficult to compare the findings obtained in these studies with those found in cat visual cortex. In human motor cortex, intracortical inhibitory and excitatory activity elicited first by the CS and then by the TS interacts at the corticospinal output neurones and determines the amount of population activity. In the visual cortex, this population activity interacts with afferent sensory input which by itself activates excitatory and inhibitory network activity. The driving effect of the afferent input might have shifted the ratio of excitatory and inhibitory network activity towards excitation. Indeed, we found in this and the previous study (Moliadze et al. 2003) that the combination of TMS with visually evoked activity causes a 2–3 times stronger facilitation of activity than when applied during spontaneous activity, indicating that the TMS pulse mainly pushed subthreshold, visually pre-activated inputs. Thus, the effect of ppTMS in the visual cortex during visual stimulation seems to be more comparable with results obtained in human motor cortex during voluntary contraction of target muscles. Indeed, short interval cortical inhibition (SICI) is reduced during a voluntary contraction (Ridding et al. 1995; Fisher et al. 2002; Roshan et al. 2003), or replaced by cortical facilitation (ICF) (Di Lazzaro et al. 1998; Ilic et al. 2002).
, 百拇医药
    One problem was that we could not analyse the first 30 ms after TMS because of the huge stimulus artefact. With the ‘Cyberamp 380’ amplifier (Axon Instruments, CA, USA) we could reduce but not eliminate the TMS artefact and even an amplifier equipped with a circuit for blanking out artefacts (CED 1902/4, Cambridge Electronic Design, UK) did not work sufficiently for single-unit recordings. Thus, we could not analyse the direct effects of a TMS pulse on the activity of the recorded cells but the intracortically evoked trans-synaptic effects thought to dominate with TMS should last for up to 100 ms due to the duration of EPSPs and interactions of polysynaptic, repetitive inputs.
, 百拇医药
    On the other hand, our data are in accordance with data obtained from human visual cortex by Ray et al. (1998) and Dambeck et al. (2003). Neither reported increased phosphene thresholds with paired-pulse TMS at short ISIs (1–5 ms). The threshold was either stable up to 100 ms ISI and then increased (Ray et al. 1998), or a general facilitation of phosphene sensation was observed at all ISIs when using a CS of 90% (Dambeck et al. 2003). Paired-pulse TMS applied to human parietal cortex revealed an ISI-dependent effect somewhat similar to motor cortex (Oliveri et al. 2000): the threshold for detecting a weak electrical cutaneous stimulus was increased with short ISIs of 1 or 3 ms, but decreased for an interval of 5 ms compared to the situation with a single TS. However, the TS itself caused an increase in detection threshold, and detection with ppTMS of 5 ms ISI was close to control levels without TMS. This result is similar to our finding of increased suppression by the TS with a slightly subthreshold CS.
, 百拇医药
    Possible cellular mechanisms

    What could be the cellular mechanisms that cause the paired-pulse facilitation or suppression of sensory activity As already mentioned above, a stimulus-dependent ratio of excitatory and inhibitory synapses might be activated. However, the monosynaptic or heterosynaptic activation of purely excitatory inputs can also explain these effects. Using in vivo intracellular recordings in anaesthetized cats, Fuentealba et al. (2004) tested the effects of paired-pulse electrical stimulation of thalamic and cortical inputs to neocortical neurones. Heterosynaptic pairing of cortical and thalamic inputs primarily induced a depression of the second response in a pair, while monosynaptic pairing (either cortical or thalamic inputs) also caused facilitation of the conditioned second input, first of all with thalamic pairing. One may speculate that in our study a weak TMS CS might have primarily activated cortical inputs which resulted in a depression of subsequent thalamic inputs while stronger CS caused facilitation by also stimulating the thalamocortical axons in the white matter. The effective time window of paired-pulse interactions in the study of Fuentealba et al. (2004) is sufficiently long to fit not only to the pairing of TMS but also to the interaction with the visual input. A maximum effect was found with an ISI of about 10–20 ms, which is close to the maximum at 3–4 ms in our study. ISIs shorter than 5 ms were not investigated by Fuentealba et al. (2004).
, 百拇医药
    Another possible cellular reason for depression and even facilitation of the second response in a paired-pulse stimulus protocol could be the refractoriness of the neuronal membrane. By applying the paired-pulse electrical stimulation protocol to peripheral nerves Chan et al. (2002) could demonstrate that the evoked volley of axonal population activity showed an ISI dependence very similar to SICI and ICF of human motor cortex: a subthreshold CS applied 1–4 ms prior to the TS reduced the amplitude of the axonal response while it was enhanced with ISIs of 5–20 ms. Chan concluded that the time course of suppression fits the temporal aspects of the relative refractory state of the axonal membrane while response facilitation is due to a following increase in the open probability of voltage-sensitive sodium channels induced by the leading hyperpolarization (rebound). By applying the same stimulus protocol to the H-reflex test, they could further demonstrate that these changes in excitability are even transmitted across an intervening excitatory synapse. Thus, changes in excitability of the axonal membranes may contribute to the effect of conditioning stimuli even for ISIs longer than 1 ms for which synaptic explanations have been postulated.
, 百拇医药
    Effects of anaesthesia

    We have to consider further that activity evoked by stimulating the brain depends on the actual state of the neuronal network which can be affected intrinsically by the brainstem arousal systems, and also by pathological states. For example, stimulus protocols usually inducing the suppression of cortical excitability (1 Hz repetitive TMS) do the opposite during migraine states (Fierro et al. 2003). General anaesthetics affect the state of the cortical network by a spectrum of different actions (for review see Rudolph & Antkowiak, 2004). One prominent action is the potentiation of the GABAA receptor, especially with barbiturates, etomidate and propofol. Although to a somewhat lesser extent, the cortical site of action of the volatile anaesthetic halothane also seems to be the GABAA receptor (Hentschke et al. 2005), while in the spinal cord, potentiation of glycinergic actions by halothane supports immobility and the blockade of transmission of nociceptive information (Rudolph & Antkowiak, 2004). The combined cortical and spinal action of halothane might explain the strongly reduced likelihood of evoking motor potentials by electrically stimulating the human motor cortex (Kawaguchi et al. 1996). At the concentration we used in combination with nitrous oxide, halothane has been found to reduce spontaneous activity in vivo by about 50% (Hentschke et al. 2005). Thalamocortical processing of sensory signals may also be reduced but typical patterns of excitatory and inhibitory responses are still preserved during halothane–nitrous oxide anaesthesia, a reason for choosing this kind of anaesthesia when studying sensory processing in higher mammals. In this and the previous study (Moliadze et al. 2003), we could show that a single TMS pulse can induce both facilitatory and suppressive volleys lasting up to one second or even more, indicating that synaptic transmission is only moderately affected by nitrous oxide–halothane anaesthesia. Nevertheless, the balance between excitatory and inhibitory actions in the cortex may be shifted towards a dominance of inhibitory actions.
, 百拇医药
    Conclusions

    In summary, our results obtained with single-unit recordings during ppTMS of cat visual cortex reveal that the outcome of ppTMS primarily depends on the strength of the conditioning stimulus but less on the interval between conditioning (CS) and test stimulus (TS). At moderately subthreshold and suprathreshold strength (60–130%), CS amplifies the TS effect: facilitation of subsequent visual activity is improved but also the suppression of visual activity induced by the TS is slightly strengthened. A weak CS (15–30%), however, diminishes facilitation of visual activity by the TS. Our findings are in agreement with the idea that inhibitory connections may have a lower activation threshold than excitatory connections. A weak CS would thus pre-activate a relatively higher number of inhibitory synapses as compared to a stronger CS, leading to a weakening of the following TS, while relatively more and more excitatory inputs are activated with increasing CS strength. Our data are also in accordance with the few data obtained in human visual cortex, but more detailed studies of the ppTMS effect on human visual perception are needed to further elucidate the specific nature of the visual system with respect to TMS.
, http://www.100md.com
    References

    Awiszus F, Feistner H, Urbach D & Rostock H (1999). Characterisation of paired-pulse transcranial magnetic stimulation conditions yielding intracortical inhibition or I-wave facilitation using a threshold-hunting paradigm. Exp Brain Res 129, 317–324.

    Chan JHL, Lin CSY, Pierrot-Deseilligny E & Burke D (2002). Excitability changes in human peripheral nerve axons in a paradigm mimicking paired-pulse transcranial magnetic stimulation. J Physiol 542, 951–961.
, http://www.100md.com
    Chen R, Tam A, Butefisch C, Corwell B, Ziemann U, Rothwell JC & Cohen LG (1998). Intracortical inhibition and facilitation in different representations of the human motor cortex. J Neurophysiol 80, 2870–2881.

    Dambeck N, Stock K, Weidemann J, Meister IG, Foltys H & Boroojerdi B (2003). Investigating phosphene elicitation with the paired-pulse paradigm. Proceedings of the 29th Neurobiology Conference and 5th Conference of the German Neuroscience Society, S528.
, 百拇医药
    Di Lazzaro V, Oliviero A, Profice P, Insola A, Mazzone P, Tonali P & Rothwell JC (1999a). Direct recordings of descending volleys after transcranial magnetic and electrical motor cortex stimulation in conscious humans. Electroencephalogr Clin Neurophysiol Suppl 51, 120–126.

    Di Lazzaro V, Restuccia D, Olivero A, Profice P, Ferrara L, Insola A, Mazzone P, Tonali P & Rothwell JC (1998). Effects of voluntary contraction on descending volleys evoked by transcranial magnetic stimulation in conscious humans. J Physiol 508, 625–633.
, 百拇医药
    Di Lazzaro V, Rothwell JC, Oliviero A, Profice P, Insola A, Massone P & Tonali P (1999b). Intracortical origin of the short latency facilitation produced by pairs of threshold magnetic stimuli applied to human motor cortex. Exp Brain Res 129, 494–499.

    Fierro B, Ricci R, Piazza A, Scalia S, Giglia G, Vitello G & Brighina F (2003). 1 Hz rTMS enhances extrastriate cortex activity in migraine – Evidence of a reduced inhibition Neurology 61, 1446–1448.
, 百拇医药
    Fisher R, Nakamura Y, Bestmann S, Rothwell JC & Bostock H (2002). Two phases of intracortical inhibition explored by transcranial magnetic threshold tracking. Exp Brain Res 143, 240–248.

    Fuentealba P, Crochet S, Timofeev I & Steriade M (2004). Synaptic interactions between thalamic and cortical inputs onto cortical neurons in vivo. J Neurophysiol 91, 1990–1998.

    Hentschke H, Schwarz C & Antkowiak B (2005). Neocortex is the major target of sedative concentrations of volatile anaesthetics: strong depression of firing rates and increase of GABAA receptor-mediated inhibition. Eur J Neurosci 21, 93–102.
, http://www.100md.com
    Ilic TV, Meintzschel F, Cleff U, Ruge D, Kessler KR & Ziemann U (2002). Short-interval paired-pulse inhibition and facilitation of human motor cortex: the dimension of stimulus intensity. J Physiol 545, 153–167.

    Kawaguchi M, Shimizu K, Furuya H, Sakamoto T, Ohnishi H & Karasawa J (1996). Effects of isoflurane on motor-evoked potentials induced by direct electrical stimulation of the exposed motor cortex with single, double and triple stimuli in rats. Anaesthesiology 85, 1176–1183.
, http://www.100md.com
    Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD, Ferbert A, Wroe S, Asselman P & Marsden CD (1993). Cortico-cortical inhibition in human motor cortex. J Physiol 471, 501–519.

    Moliadze V, Zhao Y, Eysel U & Funke K (2003). Effect of transcranial magnetic stimulation on single-unit activity in the cat primary visual cortex. J Physiol 553, 665–679.

    Nakamura H, Kitagawa H, Kawaguchi Y & Tsuji I (1997). Intracortical facilitation and inhibition after transcranial magnetic stimulation in conscious humans. J Physiol 498, 817–823.
, 百拇医药
    Oliveri M, Caltagirone C, Filippi MM, Traversa R, Cicinelli P, Pasqualetti P & Rossini PM (2000). Paired transcranial magnetic stimulation protocols reveal a pattern of inhibition and facilitation in the human parietal cortex. J Physiol 529, 461–468.

    Ray PG, Meador KJ, Epstein CM, Loring DW & Day LJ (1998). Magnetic stimulation of visual cortex: factors influencing the perception of phosphenes. J Clin Neurophysiol 15, 351–357.
, 百拇医药
    Ridding MC, Taylor JL & Rothwell JC (1995). The effect of voluntary contraction on cortico-cortical inhibition in human motor cortex. J Physiol 487, 541–548.

    Roshan L, Paradiso GO & Chen R (2003). Two phases of short-interval intracortical inhibition. Exp Brain Res 151, 330–337.

    Rudolph U & Antkowiak B (2004). Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci 5, 709–720.
, 百拇医药
    Sanger TD, Garg RR & Chen R (2001). Interactions between two different inhibitory systems in the human motor cortex. J Physiol 530, 307–317.

    Tokimura H, Ridding MC, Tokimura Y, Amassian VE & Rothwell JC (1996). Short latency facilitation between pairs of threshold magnetic stimuli applied to human motor cortex. Electroencephalogr Clin Neurophysiol 101, 263–272.

    Ziemann U, Rothwell JC & Ridding MC (1996). Interaction between intracortical inhibition and facilitation in human motor cortex. J Physiol 496, 873–881.

    Ziemann U, Tergau F, Wassermann EM, Wisher S, Hilderbandt J & Paulus W (1998). Demonstration of facilitatory I wave interaction in the human motor cortex by paired transcranial magnetic stimulation. J Physiol 511, 181–190., 百拇医药(Vera Moliadze, Dimitrios )