- Email: [email protected]
interdeviant interval on the auditory N1 and mismatch negativity in the cat auditory cortex
Cognitive Brain Research 13 (2002) 249–253 www.elsevier.com / locate / bres
Short communication
Effect of deviant probability and interstimulus / interdeviant interval on the auditory N1 and mismatch negativity in the cat auditory cortex ´ Lakatos, Csaba Rajkai, Istvan ´ Ulbert, George Karmos Zsuzsanna Pincze*, Peter Institute for Psychology of the Hungarian Academy of Sciences, P.O. Box 398, Budapest H-1394, Hungary Accepted 25 September 2001
Abstract In passive oddball paradigm the effects of changes in interstimulus / interdeviant interval (ISI; IDI) and deviant probability were investigated on mismatch negativity (MMN), auditory N1 wave and the exogenous P1 component of the auditory event-related potential in the cat. An epidural electrode matrix was chronically implanted over the auditory fields of the neocortex, and the amplitudes of the aforementioned components were measured in the location of their amplitude maxima. Dependence of the MMN both on the ISI and IDI as well as deviant probability was revealed, while the amplitude of the P1 and N1 showed dependence merely on the ISI. This method can be used for separation of the two negative, often overlapping components in the cat. 2002 Elsevier Science B.V. All rights reserved. Theme: Sensory systems Topic: Auditory systems: central physiology Keywords: MMN; Mismatch negativity; N1; Event-related potential; Cat
The auditory N1 wave is a prominent, obligatory, negative event-related potential (ERP) deflection, which can be recorded in humans [19], in macaques [13], and also in the cat [23]. The mismatch negativity (MMN) component follows the N1 of the ERPs elicited by the deviant tones in passive oddball paradigm. Depending on the stimulus parameters these components often overlap. The MMN is regarded as a bioelectric correlate of a result of the mismatch between a sensory memory trace and the incoming stimulus [18]. This wave was shown also in the cat [2]. The N1 wave is exceptionally sensitive to the stimulus rate [5,8,13], with decreasing interstimulus interval (ISI), the N1 amplitude is diminished. The amplitude of MMN is affected by various parameters of the oddball paradigm (deviance, deviant probability, ISI) [10,12,20,31], all of which can alter the strength of the memory trace. In human experiments two different subcomponents of the MMN were identified: a temporal and a frontal one. Most of the authors localised the MMN generation to the supratemporal plane [3,9,26,30,32], additionally Giard et *Corresponding author. Tel.: 136-1-3533-244; fax: 136-1-2692-972. E-mail address: [email protected] (Z. Pincze).
al. [7] suggested the existence of a frontal MMN generator, which was later proved by various scalp current density studies [6,24,28,34]. However, so far no evidence was found in animal research for the frontal MMN subcomponent. In cat, the first positive component of the auditory cortical ERP, elicited earlier than the N1 and MMN, is the exogenous P1, reflecting the thalamocortical input. Its amplitude mainly depends on the physical parameters of the stimulus [29]. In our previous study [23] the temporal sources of the N1 and MMN were topographically separated on the auditory cortical fields of the cat by manipulating the deviance between the standard and deviant stimuli. However, in those experiments the other parameters of the oddball paradigm were kept constant. In order to study the physiological mechanism of MMN and the pharmacological effects exerted upon it, it has to be demonstrated that the MMN of the cat displays similar dependence on stimulus variables as that of humans. Our goal was to investigate the dependence of the MMN, compared to that of the exogenous P1 and N1, on variable ISI, interdeviant interval (IDI) and deviant probability used in passive oddball paradigm, and thus provide further support of the reliability of the cat MMN model.
0926-6410 / 02 / $ – see front matter 2002 Elsevier Science B.V. All rights reserved. PII: S0926-6410( 01 )00105-7
250
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
Auditory ERPs were recorded by chronically implanted epidural electrode matrix, covering the auditory areas, in four adult cats. The electrodes were made of enamelinsulated 0.23 mm diameter stainless steel wire, and were implanted under pentobarbital anaesthesia (40 mg / kg i.p.), using aseptic technique. Holes were drilled above the auditory neocortex and small loops, formed from the 1 mm bared end of the electrode wires, were placed on the dura mater. A stainless steel screw in the bone covering the frontal sinus served as reference. The leads were connected to a miniature multicontact connector and the whole implant was fixed to the skull by dental acrylic. An impression was formed for the bone conductor in the frontal part of the acrylic implant. Acoustic stimuli (tone bursts of 15 ms duration with 5 ms rise and fall times each and 80 dB peSPL intensity) were given by a tone burst generator and delivered by a bone conductor (Oticon, type 10380) to maintain the constant-intensity acoustic input in the freely moving cat [14]. To elicit the frequency MMN the usual passive oddball paradigm was used, varying the deviant probability (2, 5 and 10%) and the ISI (333, 166 and 99 ms), resulting in a total of nine different stimulation conditions. The frequency deviance between the standard and the deviant tones was 100% (1 and 2 kHz). Both members of the tone pairs were presented as standard and also as deviant in two separate blocks in the same experiment (each block contained 50 deviant stimuli); this made it possible to compare the ERPs elicited by identical-frequency standard and deviant tones. Although one experiment was carried out on each animal, two peak amplitude measurements were done for all of the components in each stimulation condition due to the aforementioned paradigm. The brain electrical activity was amplified by a multichannel preamplifier (bandpass: 1 Hz–1 kHz) and digitised at 2000 Hz sampling rate. Because of the implanted intracranial electrodes no artefact suppression was needed. The offline analysis of the data was executed by NeuroScan data processing software. Averaged responses were generated from ERPs elicited by 50 deviants as well as by 50 standards, preceding the deviant ones. Difference waveforms were obtained by subtracting the ERPs to the standards from those to the deviants in order to measure the MMN peak amplitude and MMN area. Statistical analysis [two-way analysis of variance (ANOVA) with between-groups design where probability and ISI served as factors; as well as curve estimation] was performed using SPSS 8.0. Since at certain stimulation conditions (especially at the highest deviant probability and stimulation rate) there were no detectable peaks or deflections in the latency range of P1, N1 or MMN, these cases were excluded from the analysis. The configuration of auditory ERPs recorded in this experiment was similar to that described by Teas and Kiang [29]. The early positive component (P1) had the same amplitude in the responses elicited by the deviant and
standard tones of the passive oddball paradigm (Fig. 1A), and its latency varied between 10.5 and 13 ms depending on the electrode position. The early positive peaks were followed by the N1 component in the responses elicited by the standards, while two nearly overlapping negative waves appeared in the responses elicited by the deviant tones: the shorter-latency N1 (45–58 ms) (Fig. 1A) and the longer-latency MMN (61–75 ms), which can be better seen in the difference curve (Fig. 1B). The P1 and N1 components appeared with highest amplitude in the middle region of the ectosylvian gyrus, which corresponded to the lower region of the AI area; while the MMN had its highest amplitude in the anteroventral part of the AII area [23]. The peak amplitudes of the components were measured on these maximum sites, respectively. To eliminate the disturbing overlapping effect of the MMN, the amplitude of the N1 was measured on ERPs elicited by the standard tones, while the amplitude of the MMN was measured in the difference curves. Measuring the N1 in this way is reasonable, since in some studies [4,10,16] the N1 to both standards and deviants behaved similarly as the parameters were varied (although deviant stimuli elicited slightly larger N1). Difference curves in Fig. 1C display the changes of the MMN depending on the probability and ISI in the nine conditions. The graphs in Fig. 2 delineate the average amplitude values of P1, N1 and MMN of the four cats. Fig. 2A, C and E show the dependence of these components on the ISI while Fig. 2B, D and F delineate the effect of probability. The ANOVA showed a significant rate effect on the P1 (F2, 58 54.019, P,0.05) as well as the N1 (F2, 60 53.558, P,0.05) components. In contrast to the rate effect, the P1 and N1 amplitude did not show any significant changes (P1: F2, 58 51.089, P50.343; N1: F2, 60 50.27, P50.764) as a function of deviant probability (Fig. 2B and D). Fig. 2E and F delineate the average values of MMN as a function of the ISI and probability, respectively. The ANOVA showed both significant probability (F2, 61 5 4.798, P,0.05) and ISI (F2, 61 55.319, P,0.01) effects on MMN. The area of the MMN (curves in Fig. 1C) was also measured as a function of the ISI as well as the probability. This measure showed even more expressed dependence on both the ISI (F2, 61 56.211, P,0.01) and the probability of the deviant stimulus (F2, 61 55.774, P,0.01). Since some human studies [10,12] showed the dependence of MMN amplitude as a function of IDI, we also investigated this phenomenon in the cat. At the curve estimation the logarithmic model proved to be the best fit to our data (R 2 50.92, P,0.001). Fig. 3 clearly shows the log–linear relationship between the two variables. There are several animal studies, which investigate MMN in different brain structures of some species [2,11,15,25]. In these studies the effects of systematically varied deviant probability and stimulus rate have not been
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
251
Fig. 1. Typical auditory ERP waveforms elicited by identical frequency standard and deviant tones (A) and their difference curve (B). The responses were recorded from a representative cat in the case of 10% probability and 333 ms ISI. The arrows show the investigated components and the black areas (B, C) correspond to the MMN area. (C) Grand average of difference waveforms in the nine stimulation conditions.
statistically analysed. Furthermore, it is difficult to compare the results of the experiments because of the different stimulus parameters, method of tone delivering and experimental conditions, like the arousal state of the animal. Beyond these, no animal study has investigated the characteristics of the N1 wave and compared them to that of the MMN. In our present study the effect of deviant probability and the ISI was systematically analysed. Together with the MMN, the change of the P1 and the N1 was investigated too. At a given deviant probability both the P1 and N1 amplitude decreased progressively with increasing the stimulus rate, in a manner similar to that called rate effect in the human experiments [5,8]. Probability of the deviant had no effect on the P1 and N1 waves. These results are reasonable, since the P1 is exogenous and the N1 is also influenced by the physical parameters of the stimulus. MMN reflects the earliest identified cortical response to stimulus deviance in the auditory modality [18]. The incoming deviant stimulus is compared to the existing
memory trace which was formed by the preceding, consecutive standard stimuli. The main determinant of the strength of the acoustic sensory memory trace is the number of standard stimuli presented between the deviants. As the deviant probability decreases the strength of the memory trace increases, leading to an increase in MMN amplitude [12,33]. In our present study the MMN showed systematic changes as a function of the probability which correspond well to the above theory. The aforementioned probability effect refers to the MMN deflection recorded with frontocentral maximum from the human scalp, and not to the separate MMN generators. Using scalp current density mapping Sato et al. [28] investigated the effect of deviant probability on the separate activities of the frontal and temporal MMN generators. They concluded that only the frontal MMN generator was sensitive to the deviant probability, while the temporal one was not affected. Our finding is not consistent with their result, since we found significant probability effect on the cat temporal MMN generator.
252
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
Fig. 2. Charts delineating the means of the peak amplitude of the three investigated components as a function of ISI and deviant probability. (A, C, E) Mean amplitudes of the P1, N1 and MMN (respectively) as a function of ISI, showed separately at the different probabilities. (The asterisk in the title of part E indicates that the ISI effect is probably secondary and it is in reality an IDI effect. See text and Fig. 3). (B, D, F) Mean amplitudes of the P1, N1 and MMN (respectively) as a function of probability, showed separately at the three different ISIs.
There are some human studies, which investigate the ISI dependence of the MMN usually with the aim to determine the duration of the memory trace [1,4,16,21,22,27]. Therefore, in these studies relatively long ISIs were used (1–10 s). However, only Javitt et al. [12] examined systematically the effect of short ISIs (below 0.5 s) on the MMN in humans. Our findings are consistent with their data where similar dependence of the MMN on ISI was found in the range we used. ¨¨ ¨ According to Naatanen [17], if the ISI is made shorter, keeping the probability of the deviants constant, the latter occur with shorter IDIs. When the deviants occur in close succession, the corresponding mismatch processes may weaken, leading to smaller MMN. This conception was experimentally proved by human investigations [10,12]. Thus, in our present study the systematic change of MMN amplitude as a function of ISI is probably due to the change of IDIs and the ISI effect is merely secondary. Together with the finding that the amplitude distribution of the MMN and N1 is different above the auditory areas of the cat, and the MMN amplitude and latency change as a function of the pitch deviance [23], the difference in the ISI and probability effect can be used for identification of the two, often overlapping, components. In conclusion, the change of MMN to the variable stimulus parameters is similar to that of the human one, so the cat frequency MMN model is appropriate for further detailed neurophysiological investigations.
Acknowledgements This work was supported by OTKA (F 021103) and HFSP (RG 0025 / 96).
References
Fig. 3. Amplitude of the MMN as a function of the logarithm of the IDI. The scatterplot shows the data points at the logarithm of different IDIs. The continuous line is the regression line fitted to the MMN amplitude data. Regression equation: MMN amplitude57.841213log IDI.
¨ [1] C. Bottcher-Gandor, P. Ullsperger, Mismatch negativity in eventrelated potentials to auditory stimuli as a function of varying interstimulus interval, Psychophysiology 29 (1992) 547–550. ´ ´ Evoked potential correlates of [2] V. Csepe, G. Karmos, M. Molnar, stimulus deviance during wakefulness and sleep in cat: animal model of mismatch negativity, Electroencephalogr. Clin. Neurophysiol. 66 (1987) 571–578. ´ [3] V. Csepe, C. Pantev, M. Hoke, S. Hampson, B. Ross, Evoked magnetic responses to minor pitch changes: localization of the mismatch field, Electroencephalogr. Clin. Neurophysiol. 84 (1992) 538–548. [4] I. Czigler, G. Csibra, A. Csontos, Age and inter-stimulus interval effects on event-related potentials to frequent and infrequent auditory stimuli, Biol. Psychol. 33 (1992) 195–206. [5] H. Davis, T. Mast, N. Yoshie, S. Zerlin, The slow response of the human cortex to auditory stimuli: Recovery process, Electroencephalogr. Clin. Neurophysiol. 21 (1966) 105–113. [6] L.Y. Deouell, S. Bentin, M.H. Giard, Mismatch negativity in dichotic listening: evidence for interhemispheric differences and multiple generators, Psychophysiology 35 (1998) 355–365. [7] M.H. Giard, F. Perrin, J. Pernier, P. Bouchet, Brain generators
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
implicated in the processing of auditory stimulus deviance: a topographic event-related potential study, Psychophysiology 27 (1990) 627–640. R. Hari, K. Kaila, T. Katila, T. Tuomisto, T. Varpula, Interstimulus interval dependence of the auditory vertex response and its magnetic counterpart: implications for their neural generation, Electroencephalogr. Clin. Neurophysiol. 54 (1982) 561–569. R. Hari, J. Rif, J. Tiihonen, M. Sams, Neuromagnetic mismatch fields to single and paired tones, Electroencephalogr. Clin. Neurophysiol. 82 (1992) 152–154. T. Imada, R. Hari, N. Loveless, L. McEvoy, M. Sams, Determinants of the auditory mismatch response, Electroencephalogr. Clin. Neurophysiol. 87 (1993) 144–153. D.C. Javitt, C.E. Schroeder, M. Steinschneider, J.C. Arezzo, H.G. Vaughan Jr., Demonstration of mismatch negativity in the monkey, Electroencephalogr. Clin. Neurophysiol. 83 (1992) 87–90. D.C. Javitt, S. Grochowski, A. Shelley, W. Ritter, Impaired mismatch negativity (MMN) generation in schizophrenia as a function of stimulus deviance, probability, and interstimulus / interdeviant interval, Electroencephalogr. Clin. Neurophysiol. 108 (1998) 143– 153. D.C. Javitt, M. Jayachandra, R.W. Lindsley, C.M. Specht, C.E. Schroeder, Schizophrenia-like deficits in auditory P1 and N1 refractoriness induced by the psychomimetic agent phencyclidine (PCP), Clin. Neurophysiol. 111 (2000) 833–836. ´ G. Karmos, J. Martin, L. Kellenyi, M. Bauer, Constant intensity sound stimulation with a bone conductor in the freely moving cat, Electroencephalogr. Clin. Neurophysiol. 28 (1970) 637–638. N. Kraus, T. McGee, T. Littman, T. Nicol, C. King, Nonprimary auditory thalamic representation of acoustic change, J. Neurophysiol. 72 (1994) 1270–1277. ¨ ¨¨ ¨ S. Mantysalo, R. Naatanen, Duration of a neural trace of an auditory stimulus as indicated by event-related potentials, Biol. Psychol. 24 (1987) 183–195. ¨¨ ¨ R. Naatanen, P. Paavilainen, K. Alho, K. Reinikainen, M. Sams, Interstimulus interval and the mismatch negativity, in: C. Barber, T. Blum (Eds.), Evoked Potentials III, Butterworth, London, 1987, pp. 392–397. ¨¨ ¨ R. Naatanen, in: Attention and Brain Function, Lawrence Erlbaum Associates, Hillsdale, NJ, 1992, p. 139. ¨¨ ¨ R. Naatanen, T. Picton, The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure, Psychophysiology 24 (1987) 375–425. G. Novak, W. Ritter, H.G. Vaughan Jr., The chronometry of attention-modulated processing and automatic mismatch detection, Psychophysiology 29 (1992) 412–430.
253
¨ J. Partanen, J. Karhu, Mismatch negativi[21] E. Pekkonen, V. Jousmaki, ty area and age-related auditory memory, Electroencephalogr. Clin. Neurophysiol. 87 (1993) 321–325. ¨ M. Kononen, ¨ ¨ [22] E. Pekkonen, V. Jousmaki, K. Reinikainen, J. Partanen, Auditory sensory memory impairment in Alzheimer’s disease: an event-related potential study, NeuroReport 5 (1994) 2537–2540. [23] Zs. Pincze, P. Lakatos, Cs. Rajkai, I. Ulbert, G. Karmos, Separation of mismatch negativity and the N1 wave in the auditory cortex of cat: a topographic study, Clin. Neurophysiol. 112 (2001) 778–784. ¨¨ ¨ [24] T. Rinne, K. Alho, R.J. Ilmoniemi, J. Virtanen, R. Naatanen, Separate time behaviors of the temporal and frontal mismatch negativity sources, Neuroimage 12 (2000) 14–19. [25] T. Ruusuvirta, M. Penttonen, T. Korhonen, Auditory cortical eventrelated potentials to pitch deviances in rats, Neurosci. Lett. 248 (1998) 45–48. ¨ ¨¨ ¨¨ ¨ [26] M. Sams, E. Kaukoranta, M. Hamalainen, R. Naatanen, Cortical activity elicited by changes in auditory stimuli: different sources for the magnetic N100m and mismatch responses, Psychophysiology 28 (1991) 21–29. [27] M. Sams, R. Hari, J. Rif, J. Knuutila, The human auditory sensory memory trace persists about 10 s: neuromagnetic evidence, J. Cogn. Neurosci. 5 (1993) 363–370. [28] Y. Sato, H. Yabe, T. Hiruma, T. Sutoh, N. Shinozaki, T. Nashida, S. Kaneko, The effect of deviant stimulus probability on the human mismatch process, NeuroReport 11 (2000) 3703–3708. [29] D.C. Teas, N.Y.S. Kiang, Evoked responses from the auditory cortex, Exp. Neurol. 10 (1964) 91–119. [30] H. Tiitinen, K. Alho, M. Houtilainen, R.J. Ilmoniemi, J. Simola, R. ¨¨ ¨ Naatanen, Tonotopic auditory cortex and the magnetoencephalographic (MEG) equivalent of the mismatch negativity, Psychophysiology 30 (1993) 537–540. ¨¨ ¨ [31] H. Tiitinen, P. May, K. Reinikainen, R. Naatanen, Attentive novelty detection in humans is governed by pre-attentive sensory memory, Nature 372 (1994) 90–92. [32] I. Winkler, M. Tervaniemi, M. Huotilainen, R. Ilmoniemi, A. ¨ ¨¨ ¨ Ahonen, O. Salonen, C. Standertskjold-Nordenstam, R. Naatanen, From objective to subjective: pitch representation in the human auditory cortex, NeuroReport 6 (1995) 2317–2320. ¨¨ ¨ [33] I. Winkler, G. Karmos, R. Naatanen, Adaptive modeling of the unattended acoustic environment reflected in the mismatch negativity event-related potential, Brain Res. 742 (1996) 239–252. [34] E. Yago, C. Escera, K. Alho, M.H. Giard, Cerebral mechanisms underlying orienting of attention towards auditory frequency changes, NeuroReport 12 (2001) 2583–2587.
Short communication
Effect of deviant probability and interstimulus / interdeviant interval on the auditory N1 and mismatch negativity in the cat auditory cortex ´ Lakatos, Csaba Rajkai, Istvan ´ Ulbert, George Karmos Zsuzsanna Pincze*, Peter Institute for Psychology of the Hungarian Academy of Sciences, P.O. Box 398, Budapest H-1394, Hungary Accepted 25 September 2001
Abstract In passive oddball paradigm the effects of changes in interstimulus / interdeviant interval (ISI; IDI) and deviant probability were investigated on mismatch negativity (MMN), auditory N1 wave and the exogenous P1 component of the auditory event-related potential in the cat. An epidural electrode matrix was chronically implanted over the auditory fields of the neocortex, and the amplitudes of the aforementioned components were measured in the location of their amplitude maxima. Dependence of the MMN both on the ISI and IDI as well as deviant probability was revealed, while the amplitude of the P1 and N1 showed dependence merely on the ISI. This method can be used for separation of the two negative, often overlapping components in the cat. 2002 Elsevier Science B.V. All rights reserved. Theme: Sensory systems Topic: Auditory systems: central physiology Keywords: MMN; Mismatch negativity; N1; Event-related potential; Cat
The auditory N1 wave is a prominent, obligatory, negative event-related potential (ERP) deflection, which can be recorded in humans [19], in macaques [13], and also in the cat [23]. The mismatch negativity (MMN) component follows the N1 of the ERPs elicited by the deviant tones in passive oddball paradigm. Depending on the stimulus parameters these components often overlap. The MMN is regarded as a bioelectric correlate of a result of the mismatch between a sensory memory trace and the incoming stimulus [18]. This wave was shown also in the cat [2]. The N1 wave is exceptionally sensitive to the stimulus rate [5,8,13], with decreasing interstimulus interval (ISI), the N1 amplitude is diminished. The amplitude of MMN is affected by various parameters of the oddball paradigm (deviance, deviant probability, ISI) [10,12,20,31], all of which can alter the strength of the memory trace. In human experiments two different subcomponents of the MMN were identified: a temporal and a frontal one. Most of the authors localised the MMN generation to the supratemporal plane [3,9,26,30,32], additionally Giard et *Corresponding author. Tel.: 136-1-3533-244; fax: 136-1-2692-972. E-mail address: [email protected] (Z. Pincze).
al. [7] suggested the existence of a frontal MMN generator, which was later proved by various scalp current density studies [6,24,28,34]. However, so far no evidence was found in animal research for the frontal MMN subcomponent. In cat, the first positive component of the auditory cortical ERP, elicited earlier than the N1 and MMN, is the exogenous P1, reflecting the thalamocortical input. Its amplitude mainly depends on the physical parameters of the stimulus [29]. In our previous study [23] the temporal sources of the N1 and MMN were topographically separated on the auditory cortical fields of the cat by manipulating the deviance between the standard and deviant stimuli. However, in those experiments the other parameters of the oddball paradigm were kept constant. In order to study the physiological mechanism of MMN and the pharmacological effects exerted upon it, it has to be demonstrated that the MMN of the cat displays similar dependence on stimulus variables as that of humans. Our goal was to investigate the dependence of the MMN, compared to that of the exogenous P1 and N1, on variable ISI, interdeviant interval (IDI) and deviant probability used in passive oddball paradigm, and thus provide further support of the reliability of the cat MMN model.
0926-6410 / 02 / $ – see front matter 2002 Elsevier Science B.V. All rights reserved. PII: S0926-6410( 01 )00105-7
250
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
Auditory ERPs were recorded by chronically implanted epidural electrode matrix, covering the auditory areas, in four adult cats. The electrodes were made of enamelinsulated 0.23 mm diameter stainless steel wire, and were implanted under pentobarbital anaesthesia (40 mg / kg i.p.), using aseptic technique. Holes were drilled above the auditory neocortex and small loops, formed from the 1 mm bared end of the electrode wires, were placed on the dura mater. A stainless steel screw in the bone covering the frontal sinus served as reference. The leads were connected to a miniature multicontact connector and the whole implant was fixed to the skull by dental acrylic. An impression was formed for the bone conductor in the frontal part of the acrylic implant. Acoustic stimuli (tone bursts of 15 ms duration with 5 ms rise and fall times each and 80 dB peSPL intensity) were given by a tone burst generator and delivered by a bone conductor (Oticon, type 10380) to maintain the constant-intensity acoustic input in the freely moving cat [14]. To elicit the frequency MMN the usual passive oddball paradigm was used, varying the deviant probability (2, 5 and 10%) and the ISI (333, 166 and 99 ms), resulting in a total of nine different stimulation conditions. The frequency deviance between the standard and the deviant tones was 100% (1 and 2 kHz). Both members of the tone pairs were presented as standard and also as deviant in two separate blocks in the same experiment (each block contained 50 deviant stimuli); this made it possible to compare the ERPs elicited by identical-frequency standard and deviant tones. Although one experiment was carried out on each animal, two peak amplitude measurements were done for all of the components in each stimulation condition due to the aforementioned paradigm. The brain electrical activity was amplified by a multichannel preamplifier (bandpass: 1 Hz–1 kHz) and digitised at 2000 Hz sampling rate. Because of the implanted intracranial electrodes no artefact suppression was needed. The offline analysis of the data was executed by NeuroScan data processing software. Averaged responses were generated from ERPs elicited by 50 deviants as well as by 50 standards, preceding the deviant ones. Difference waveforms were obtained by subtracting the ERPs to the standards from those to the deviants in order to measure the MMN peak amplitude and MMN area. Statistical analysis [two-way analysis of variance (ANOVA) with between-groups design where probability and ISI served as factors; as well as curve estimation] was performed using SPSS 8.0. Since at certain stimulation conditions (especially at the highest deviant probability and stimulation rate) there were no detectable peaks or deflections in the latency range of P1, N1 or MMN, these cases were excluded from the analysis. The configuration of auditory ERPs recorded in this experiment was similar to that described by Teas and Kiang [29]. The early positive component (P1) had the same amplitude in the responses elicited by the deviant and
standard tones of the passive oddball paradigm (Fig. 1A), and its latency varied between 10.5 and 13 ms depending on the electrode position. The early positive peaks were followed by the N1 component in the responses elicited by the standards, while two nearly overlapping negative waves appeared in the responses elicited by the deviant tones: the shorter-latency N1 (45–58 ms) (Fig. 1A) and the longer-latency MMN (61–75 ms), which can be better seen in the difference curve (Fig. 1B). The P1 and N1 components appeared with highest amplitude in the middle region of the ectosylvian gyrus, which corresponded to the lower region of the AI area; while the MMN had its highest amplitude in the anteroventral part of the AII area [23]. The peak amplitudes of the components were measured on these maximum sites, respectively. To eliminate the disturbing overlapping effect of the MMN, the amplitude of the N1 was measured on ERPs elicited by the standard tones, while the amplitude of the MMN was measured in the difference curves. Measuring the N1 in this way is reasonable, since in some studies [4,10,16] the N1 to both standards and deviants behaved similarly as the parameters were varied (although deviant stimuli elicited slightly larger N1). Difference curves in Fig. 1C display the changes of the MMN depending on the probability and ISI in the nine conditions. The graphs in Fig. 2 delineate the average amplitude values of P1, N1 and MMN of the four cats. Fig. 2A, C and E show the dependence of these components on the ISI while Fig. 2B, D and F delineate the effect of probability. The ANOVA showed a significant rate effect on the P1 (F2, 58 54.019, P,0.05) as well as the N1 (F2, 60 53.558, P,0.05) components. In contrast to the rate effect, the P1 and N1 amplitude did not show any significant changes (P1: F2, 58 51.089, P50.343; N1: F2, 60 50.27, P50.764) as a function of deviant probability (Fig. 2B and D). Fig. 2E and F delineate the average values of MMN as a function of the ISI and probability, respectively. The ANOVA showed both significant probability (F2, 61 5 4.798, P,0.05) and ISI (F2, 61 55.319, P,0.01) effects on MMN. The area of the MMN (curves in Fig. 1C) was also measured as a function of the ISI as well as the probability. This measure showed even more expressed dependence on both the ISI (F2, 61 56.211, P,0.01) and the probability of the deviant stimulus (F2, 61 55.774, P,0.01). Since some human studies [10,12] showed the dependence of MMN amplitude as a function of IDI, we also investigated this phenomenon in the cat. At the curve estimation the logarithmic model proved to be the best fit to our data (R 2 50.92, P,0.001). Fig. 3 clearly shows the log–linear relationship between the two variables. There are several animal studies, which investigate MMN in different brain structures of some species [2,11,15,25]. In these studies the effects of systematically varied deviant probability and stimulus rate have not been
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
251
Fig. 1. Typical auditory ERP waveforms elicited by identical frequency standard and deviant tones (A) and their difference curve (B). The responses were recorded from a representative cat in the case of 10% probability and 333 ms ISI. The arrows show the investigated components and the black areas (B, C) correspond to the MMN area. (C) Grand average of difference waveforms in the nine stimulation conditions.
statistically analysed. Furthermore, it is difficult to compare the results of the experiments because of the different stimulus parameters, method of tone delivering and experimental conditions, like the arousal state of the animal. Beyond these, no animal study has investigated the characteristics of the N1 wave and compared them to that of the MMN. In our present study the effect of deviant probability and the ISI was systematically analysed. Together with the MMN, the change of the P1 and the N1 was investigated too. At a given deviant probability both the P1 and N1 amplitude decreased progressively with increasing the stimulus rate, in a manner similar to that called rate effect in the human experiments [5,8]. Probability of the deviant had no effect on the P1 and N1 waves. These results are reasonable, since the P1 is exogenous and the N1 is also influenced by the physical parameters of the stimulus. MMN reflects the earliest identified cortical response to stimulus deviance in the auditory modality [18]. The incoming deviant stimulus is compared to the existing
memory trace which was formed by the preceding, consecutive standard stimuli. The main determinant of the strength of the acoustic sensory memory trace is the number of standard stimuli presented between the deviants. As the deviant probability decreases the strength of the memory trace increases, leading to an increase in MMN amplitude [12,33]. In our present study the MMN showed systematic changes as a function of the probability which correspond well to the above theory. The aforementioned probability effect refers to the MMN deflection recorded with frontocentral maximum from the human scalp, and not to the separate MMN generators. Using scalp current density mapping Sato et al. [28] investigated the effect of deviant probability on the separate activities of the frontal and temporal MMN generators. They concluded that only the frontal MMN generator was sensitive to the deviant probability, while the temporal one was not affected. Our finding is not consistent with their result, since we found significant probability effect on the cat temporal MMN generator.
252
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
Fig. 2. Charts delineating the means of the peak amplitude of the three investigated components as a function of ISI and deviant probability. (A, C, E) Mean amplitudes of the P1, N1 and MMN (respectively) as a function of ISI, showed separately at the different probabilities. (The asterisk in the title of part E indicates that the ISI effect is probably secondary and it is in reality an IDI effect. See text and Fig. 3). (B, D, F) Mean amplitudes of the P1, N1 and MMN (respectively) as a function of probability, showed separately at the three different ISIs.
There are some human studies, which investigate the ISI dependence of the MMN usually with the aim to determine the duration of the memory trace [1,4,16,21,22,27]. Therefore, in these studies relatively long ISIs were used (1–10 s). However, only Javitt et al. [12] examined systematically the effect of short ISIs (below 0.5 s) on the MMN in humans. Our findings are consistent with their data where similar dependence of the MMN on ISI was found in the range we used. ¨¨ ¨ According to Naatanen [17], if the ISI is made shorter, keeping the probability of the deviants constant, the latter occur with shorter IDIs. When the deviants occur in close succession, the corresponding mismatch processes may weaken, leading to smaller MMN. This conception was experimentally proved by human investigations [10,12]. Thus, in our present study the systematic change of MMN amplitude as a function of ISI is probably due to the change of IDIs and the ISI effect is merely secondary. Together with the finding that the amplitude distribution of the MMN and N1 is different above the auditory areas of the cat, and the MMN amplitude and latency change as a function of the pitch deviance [23], the difference in the ISI and probability effect can be used for identification of the two, often overlapping, components. In conclusion, the change of MMN to the variable stimulus parameters is similar to that of the human one, so the cat frequency MMN model is appropriate for further detailed neurophysiological investigations.
Acknowledgements This work was supported by OTKA (F 021103) and HFSP (RG 0025 / 96).
References
Fig. 3. Amplitude of the MMN as a function of the logarithm of the IDI. The scatterplot shows the data points at the logarithm of different IDIs. The continuous line is the regression line fitted to the MMN amplitude data. Regression equation: MMN amplitude57.841213log IDI.
¨ [1] C. Bottcher-Gandor, P. Ullsperger, Mismatch negativity in eventrelated potentials to auditory stimuli as a function of varying interstimulus interval, Psychophysiology 29 (1992) 547–550. ´ ´ Evoked potential correlates of [2] V. Csepe, G. Karmos, M. Molnar, stimulus deviance during wakefulness and sleep in cat: animal model of mismatch negativity, Electroencephalogr. Clin. Neurophysiol. 66 (1987) 571–578. ´ [3] V. Csepe, C. Pantev, M. Hoke, S. Hampson, B. Ross, Evoked magnetic responses to minor pitch changes: localization of the mismatch field, Electroencephalogr. Clin. Neurophysiol. 84 (1992) 538–548. [4] I. Czigler, G. Csibra, A. Csontos, Age and inter-stimulus interval effects on event-related potentials to frequent and infrequent auditory stimuli, Biol. Psychol. 33 (1992) 195–206. [5] H. Davis, T. Mast, N. Yoshie, S. Zerlin, The slow response of the human cortex to auditory stimuli: Recovery process, Electroencephalogr. Clin. Neurophysiol. 21 (1966) 105–113. [6] L.Y. Deouell, S. Bentin, M.H. Giard, Mismatch negativity in dichotic listening: evidence for interhemispheric differences and multiple generators, Psychophysiology 35 (1998) 355–365. [7] M.H. Giard, F. Perrin, J. Pernier, P. Bouchet, Brain generators
Z. Pincze et al. / Cognitive Brain Research 13 (2002) 249 – 253
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
implicated in the processing of auditory stimulus deviance: a topographic event-related potential study, Psychophysiology 27 (1990) 627–640. R. Hari, K. Kaila, T. Katila, T. Tuomisto, T. Varpula, Interstimulus interval dependence of the auditory vertex response and its magnetic counterpart: implications for their neural generation, Electroencephalogr. Clin. Neurophysiol. 54 (1982) 561–569. R. Hari, J. Rif, J. Tiihonen, M. Sams, Neuromagnetic mismatch fields to single and paired tones, Electroencephalogr. Clin. Neurophysiol. 82 (1992) 152–154. T. Imada, R. Hari, N. Loveless, L. McEvoy, M. Sams, Determinants of the auditory mismatch response, Electroencephalogr. Clin. Neurophysiol. 87 (1993) 144–153. D.C. Javitt, C.E. Schroeder, M. Steinschneider, J.C. Arezzo, H.G. Vaughan Jr., Demonstration of mismatch negativity in the monkey, Electroencephalogr. Clin. Neurophysiol. 83 (1992) 87–90. D.C. Javitt, S. Grochowski, A. Shelley, W. Ritter, Impaired mismatch negativity (MMN) generation in schizophrenia as a function of stimulus deviance, probability, and interstimulus / interdeviant interval, Electroencephalogr. Clin. Neurophysiol. 108 (1998) 143– 153. D.C. Javitt, M. Jayachandra, R.W. Lindsley, C.M. Specht, C.E. Schroeder, Schizophrenia-like deficits in auditory P1 and N1 refractoriness induced by the psychomimetic agent phencyclidine (PCP), Clin. Neurophysiol. 111 (2000) 833–836. ´ G. Karmos, J. Martin, L. Kellenyi, M. Bauer, Constant intensity sound stimulation with a bone conductor in the freely moving cat, Electroencephalogr. Clin. Neurophysiol. 28 (1970) 637–638. N. Kraus, T. McGee, T. Littman, T. Nicol, C. King, Nonprimary auditory thalamic representation of acoustic change, J. Neurophysiol. 72 (1994) 1270–1277. ¨ ¨¨ ¨ S. Mantysalo, R. Naatanen, Duration of a neural trace of an auditory stimulus as indicated by event-related potentials, Biol. Psychol. 24 (1987) 183–195. ¨¨ ¨ R. Naatanen, P. Paavilainen, K. Alho, K. Reinikainen, M. Sams, Interstimulus interval and the mismatch negativity, in: C. Barber, T. Blum (Eds.), Evoked Potentials III, Butterworth, London, 1987, pp. 392–397. ¨¨ ¨ R. Naatanen, in: Attention and Brain Function, Lawrence Erlbaum Associates, Hillsdale, NJ, 1992, p. 139. ¨¨ ¨ R. Naatanen, T. Picton, The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure, Psychophysiology 24 (1987) 375–425. G. Novak, W. Ritter, H.G. Vaughan Jr., The chronometry of attention-modulated processing and automatic mismatch detection, Psychophysiology 29 (1992) 412–430.
253
¨ J. Partanen, J. Karhu, Mismatch negativi[21] E. Pekkonen, V. Jousmaki, ty area and age-related auditory memory, Electroencephalogr. Clin. Neurophysiol. 87 (1993) 321–325. ¨ M. Kononen, ¨ ¨ [22] E. Pekkonen, V. Jousmaki, K. Reinikainen, J. Partanen, Auditory sensory memory impairment in Alzheimer’s disease: an event-related potential study, NeuroReport 5 (1994) 2537–2540. [23] Zs. Pincze, P. Lakatos, Cs. Rajkai, I. Ulbert, G. Karmos, Separation of mismatch negativity and the N1 wave in the auditory cortex of cat: a topographic study, Clin. Neurophysiol. 112 (2001) 778–784. ¨¨ ¨ [24] T. Rinne, K. Alho, R.J. Ilmoniemi, J. Virtanen, R. Naatanen, Separate time behaviors of the temporal and frontal mismatch negativity sources, Neuroimage 12 (2000) 14–19. [25] T. Ruusuvirta, M. Penttonen, T. Korhonen, Auditory cortical eventrelated potentials to pitch deviances in rats, Neurosci. Lett. 248 (1998) 45–48. ¨ ¨¨ ¨¨ ¨ [26] M. Sams, E. Kaukoranta, M. Hamalainen, R. Naatanen, Cortical activity elicited by changes in auditory stimuli: different sources for the magnetic N100m and mismatch responses, Psychophysiology 28 (1991) 21–29. [27] M. Sams, R. Hari, J. Rif, J. Knuutila, The human auditory sensory memory trace persists about 10 s: neuromagnetic evidence, J. Cogn. Neurosci. 5 (1993) 363–370. [28] Y. Sato, H. Yabe, T. Hiruma, T. Sutoh, N. Shinozaki, T. Nashida, S. Kaneko, The effect of deviant stimulus probability on the human mismatch process, NeuroReport 11 (2000) 3703–3708. [29] D.C. Teas, N.Y.S. Kiang, Evoked responses from the auditory cortex, Exp. Neurol. 10 (1964) 91–119. [30] H. Tiitinen, K. Alho, M. Houtilainen, R.J. Ilmoniemi, J. Simola, R. ¨¨ ¨ Naatanen, Tonotopic auditory cortex and the magnetoencephalographic (MEG) equivalent of the mismatch negativity, Psychophysiology 30 (1993) 537–540. ¨¨ ¨ [31] H. Tiitinen, P. May, K. Reinikainen, R. Naatanen, Attentive novelty detection in humans is governed by pre-attentive sensory memory, Nature 372 (1994) 90–92. [32] I. Winkler, M. Tervaniemi, M. Huotilainen, R. Ilmoniemi, A. ¨ ¨¨ ¨ Ahonen, O. Salonen, C. Standertskjold-Nordenstam, R. Naatanen, From objective to subjective: pitch representation in the human auditory cortex, NeuroReport 6 (1995) 2317–2320. ¨¨ ¨ [33] I. Winkler, G. Karmos, R. Naatanen, Adaptive modeling of the unattended acoustic environment reflected in the mismatch negativity event-related potential, Brain Res. 742 (1996) 239–252. [34] E. Yago, C. Escera, K. Alho, M.H. Giard, Cerebral mechanisms underlying orienting of attention towards auditory frequency changes, NeuroReport 12 (2001) 2583–2587.