- Email: [email protected]
The effects of interstimulus interval on sensory gating and on preattentive auditory memory in the oddball paradigm
Neuroscience Letters 412 (2007) 1–5
The effects of interstimulus interval on sensory gating and on preattentive auditory memory in the oddball paradigm Can magnitude of the sensory gating affect preattentive auditory comparison process? M. Numan Ermutlu a,∗ , Tamer Demiralp b , Sacit Karam¨ursel b a b
Istanbul Bilim University, Faculty of Medicine, Department of Physiology, Turkey Istanbul University, Istanbul Medical Faculty, Department of Physiology, Turkey
Received 12 July 2006; received in revised form 31 August 2006; accepted 1 September 2006
Abstract P50, and mismatch negativity (MMN) are components of event-related potentials (ERP) reflecting sensory gating and preattentive auditory memory, respectively. Interstimulus interval (ISI) is an important determinant of the amplitudes of these components and N1. In the present study the interrelation between stimulus gating and preattentive auditory sensory memory were investigated as a function of ISI in 1.5, 2.5 and 3.5 s in 15 healthy volunteered participants. ISI factor affected the N1 peak amplitude significantly. MMN amplitude in 2.5 s ISI was significantly smaller compared to 1.5 and 3.5 s ISI. ISI X stimuli interaction on P50 amplitude was statistically significant. P50 amplitudes to deviant stimuli in 2.5 s ISI were larger than the P50 amplitudes in other ISIs. P50 difference (P50d) waveform amplitude correlated significantly with MMN amplitude. The results suggest that: (i) auditory sensory gating could affect preattentive auditory sensory memory by supplying input to the comparator mechanism; (ii) 2.5 s ISI is important in displaying the sensory gating and preattentive auditory sensory memory relation. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Sensory gating; P50; N100; Mismatch negativity (MMN); Preattentive sensory processing
Sensory gating can be defined as the ability of the brain to modulate its sensitivity to incoming sensory stimuli. This definition allows the concept of gating to include both the capacities to minimize or stop responding to incoming irrelevant stimuli (gating out) and to respond when a novel stimulus is presented or a change occurs in ongoing stimuli (gating in) [5]. The P50 component of event-related potentials (ERP) has been used to study sensory gating [11,12]. The P50 is a positive component of ERP that appears between 35 and 85 ms following auditory stimulation and reflect mainly sensory gating and early preattentional processing [21]. It is elicited in paired click paradigm, short trains paradigm, and oddball paradigm. In a paired click paradigm, the first stimulus (S1) of the pair is followed by an identical stimulus (S2) a short time ∗
Corresponding author at: Istanbul Bilim University, Faculty of Medicine, Department of Physiology, Buyukdere cad No: 120 Esentepe, 34394 Istanbul, Turkey. Tel.: +90 212 2136486; fax: +90 212 2723461. E-mail addresses: [email protected], [email protected] (M.N. Ermutlu). 0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.09.006
later (e.g., 500 ms). The inhibitory capacity of the brain is then measured as the ratio of the amplitude of the P50 to S2 stimuli to the amplitude of the P50 response to S1 stimuli (S2/S1), or as the mathematical difference between the two responses [30]. In the oddball paradigm, on the other hand, increase in amplitude of P50 to infrequent auditory stimuli was shown to reflect preattentive recognition of novel stimuli or gating in of stimuli while decrease in amplitude of P50 to frequent auditory stimuli reflects gating out of stimuli [7]. In the context of oddball paradigm, sensory gating can be operationally defined as the ratio of the amplitude of the response to the infrequent stimulus divided by amplitude of the response to frequent stimulus [7] and as the arithmetical difference between the amplitudes of responses to deviant (infrequent) stimulus and standard (frequent) stimulus. Higher ratios or differences can either reflect lower amplitudes in response to frequent stimuli, and thus stronger inhibition with repetition, or higher amplitudes in response to infrequent stimuli and, thus, stronger response to rare stimuli [7]. Infrequent auditory stimuli deviating from a repetitive standard sound in some physical feature such as frequency elicit
2
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
the mismatch negativity (MMN) component of the event-related brain potential. The MMN is thought to reflect the outcome of a mismatch process automatically registering the deviation of the current input from the neuronal representation of a repetitive stimulus in sensory memory. MMN reflects preattentive memory and memory comparison process [23]. Dysfunction of sensory gating and preattentive sensory processing has been implicated in the pathophysiology of a number of psychiatric disorders and there are some findings that suggest an interrelation between the two processes as reflected in the mismatch negativity and P50 components of event-related potentials: (i) patients with frontal lesions displayed MMN attenuation and enhanced P50 amplitudes [2]; (ii) MMN and P50 patterns observed in Schizophrenia differs from the patterns observed in normals and in some patients with neurological disorders [20,25]; (iii) P50 amplitude increases and MMN amplitude decreases with aging; (iv) with alcohol consumption, P50 amplitude was significantly augmented and peak latency of MMN was significantly increased [19]. The amplitude of P50 has a tendency to decrease in amplitude when the stimuli were presented with fast rates of repetition [13]. The effects of interstimulus interval (ISI) on ERP components are also important in the oddball paradigm. The MMN and the N1 have different relations with the ISI [24]. ON responses are transient potential changes at the beginning of a stimulus and OFF responses are the potential changes after the end of a continuous stimulus [8]. Both of them are in quite similar shape and amplitude depending on the duration of the stimulus and ISI [18]. Since the mid-latency ERP components of our interest are within 300 ms after stimulus and as MMN was shown to be elicited with long duration stimuli [16], long duration auditory stimuli were used to prevent the possible superimposition of ON and OFF responses. In the present study, a possible interaction between P50, N1 and MMN were investigated in an oddball paradigm by varying the ISI. Fifteen healthy right-handed volunteers (six females and nine males; mean age 34.1 range 25–39) were studied. None of the subjects had a personal or family history of psychiatric or neurological disorders. They also had no personal history of illicit drug use. They were not receiving any medications during the study. The study was conducted in accordance with the Declaration of Helsinki and all procedures were carried out with the adequate understanding and written consent of the participants. The participants were instructed not to take alcohol or any drug for the 48 h preceding recording. Two of the volunteers were smokers and they had not smoked for 2 h before recording. The subjects sat in an electrically shielded and soundproof room. They were asked to read a self-chosen interesting book and to ignore the stimuli given through earphones. The EEG was measured from Fz, F3, F4, and Cz recording sites according to international 10/20 system with reference to linked ear lobes. All data were amplified by using a low-pass filter with cut-off frequency at 70 Hz and with a time constant of 0.3 s. The signals were sampled at a rate of 256 points/s. ERPs were recorded in three sessions.
Three ISI (onset to onset) were used: 1.5, 2.5, and 3.5 s. In all conditions stimulus duration was 500 ms and intensity was 80 dB SPL. The standards were 1050 Hz tones and the deviants were 1250 Hz tones. The rise and fall-times were 10 ms. The probability of the deviant and standard stimuli were equal (20/80) across all three ISI conditions. Deviant stimuli did not occur in direct succession. Artifacts were eliminated by manual off-line selective averaging taking into consideration the EOG recorded from the right eye. The sweep numbers were equalized between the standards and deviants and between the conditions. P50 difference (P50d) and MMN waves were assessed by subtracting the standard responses from the deviant responses. The P50 wave was identified as the most positive peak in the 40–80 ms time window, whereas the mean amplitude and latency of MMN was measured in the time window between 100 and 200 ms. Additionally, P50 and N1 waves were separately measured in standard and deviant ERPs (the latter as the largest negative peak between 80 and 120 ms). Since P50 had negative values in some recordings, deviant minus standard P50 differences were assessed instead of P50 ratios. For the statistical evaluation, a repeated measures ANOVA design with ISI (three levels: 1.5, 2.5, 3.5 s), stimulus (two levels: standard, deviant) and channel (four levels: Fz, F3, F4, Cz) within-subject factors were applied to the peak amplitudes of P50, N1 and peak latency of N1. For MMN and P50d, an ANOVA design with the within-subject factors ISI (three levels: 1.5, 2.5, 3.5 s) and channel (four levels Fz, F3, F4, Cz) were applied to the peak amplitudes of P50d, MMN and peak latencies of MMN. Reduced degrees freedom (Greenhouse-Geisser) was used to counter violations of the sphericity assumption underlying ANOVA with repeated measures. None of the components was significantly affected by channel. ISI factor affected the N1 peak amplitudes significantly (F = 43.37, d.f. = 2.28; p < 0.001). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni) N1 amplitudes increased significantly with increase in ISI. ISI X stimulus interaction on N1 amplitude was significant (F = 32.92, d.f. = 2.28, p < 0.0001). In paired sample t-tests, N1 amplitudes to deviants were significantly larger than N1 amplitudes to standards in 1.5 s ISI (t = 3.99, d.f. = 14; p < 0.002, two-tailed) (Table 1, Fig. 1). ISI was found to have a significant effect on MMN peak amplitudes (F = 7.60, d.f. = 2.28; p < 0.002). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni), peak amplitude in 2.5 s ISI was significantly smaller than peak amplitudes in 1.5 s ISI (p < 0.05) and in 3.5 s ISI (p < 0.05) (Figs. 2 and 3). ISI was found to have no effect on MMN peak latency. P50 was neither significantly affected by ISIs nor by stimuli. However, ISI X stimulus interaction on P50 amplitude was significant (F = 3.53, d.f. = 2.28; p < 0.03). In paired t-tests, amplitude of P50 to deviant stimuli was significantly larger in 2.5 s ISI than amplitudes of P50s to deviant stimuli in 1.5 and 3.5 s ISIs (t = −0.19, d.f. = 14; p < 0.03, t = 2.61, d.f. = 14; p < 0.02).
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
3
Table 1 Mean amplitudes of the ERP components in the three ISIs ERP components (mean ± S.D.)
1.5 s ISI
P50 (deviants) P50 (standards) N1 (deviants) N1 (standards) P50 difference MMN
1.20 V 1.96 V −5.18 V −3.73 V −0.72 V −3.99 V
2.5 s ISI ± ± ± ± ± ±
1.66 1.95 1.193c 2.18 2.86 1.152
2.47 V 1.23 V −7.58 V −7.48 V 1.24 V −1.91 V
3.5 s ISI ± ± ± ± ± ±
1.68a,b 1.79 2.85 2.53 2.28d,e 1.15f
1.40 V 2.31 V −9.43 V −8.55 V −0.91 −4.19
± ± ± ± ± ±
1.35 1.68 3.0 3.45 1.78 2.05
Values indicate the means of the Fz, F3, F4, and Cz recording sites. a Larger than P50 amplitude to deviants in 1.5 s (p < 0.03). b Larger than P50 amplitude to deviants in 3.5 s (p < 0.02). c Larger than N1 amplitude to standards in 1.5 s (p < 0.002). d Larger than P50d amplitude in 1.5 s (p < 0.05). e Larger than P50d amplitude in 3.5 s (p < 0.06). f Smaller than MMN amplitudes in 1.5 and 3.5 s (p < 0.05).
Fig. 3. Mean amplitudes P50 to deviants and standards in 1.5, 2.5 and 3.5 s ISIs at Fz.
Fig. 1. Grand average ERPs of 15 participants to deviants and standards in 1.5, 2.5 and 3.5 s at Fz.
Fig. 2. Grand average difference waveforms of 15 participants obtained by subtracting ERPs to standard tones from those to deviants in 1.5, 2.5 and 3.5 s ISIs. MMN and P50d waveforms are indicated on the figure.
4
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
P50d was significantly affected by ISI (F = 4.25, d.f. = 2.28; p < 0.03, ε = 0.66). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni), P50d peak amplitude in 2.5 s ISI was larger than the peak amplitudes in 1.5 s (p < 0.05) and in 3.5 s ISIs (p < 0.06) (Figs. 2 and 3). Correlation analyses were carried out for MMN and P50d by pooling the data in all ISI conditions. There was a significant correlation between P50d and MMN amplitudes (r = 0.38, p = 0.01). In this study, the increases in the amplitude of the N1 component of ERP as a function of ISI and larger amplitude of N1 to deviants than the amplitude of N1 to standards in the 1.5 s ISI were expected findings. The increase in N1 amplitude with increases in ISI could be function of recovery of complex neuronal circuits generating N1 [24]. However, P50 and MMN amplitudes did not display the same feature as N1. The amplitude of P50 to deviants in 2.5 s ISI was significantly higher than the amplitude of P50 to deviants in the other ISIs. Also the amplitude of P50d was significantly higher and the amplitude of MMN was significantly smaller in 2.5 s ISI than other ISIs. The data suggest that there were increased gating in the 2.5 s ISI in comparison to other ISIs and significant increase in the amplitude of the P50d in the 2.5 s ISI is the result of significant increase in the amplitude of the deviant stimuli (gating in) and decreased amplitude of the standard stimuli (gating out). The absolute amplitude of MMN was significantly smaller in the 2.5 s ISI than 1.5 s, and 3.5 s ISIs. MMN is thought to reflect a neuronal mismatch between an incoming stimulus or the representation of the incoming stimulus in the short term store [29] and representations of those presented over the previous 10 s, that is the long term store. It is reported that the MMN peak amplitudes are not affected by ISI in the 0.8–7.2 s [9] and 1–10 s ranges [4]. However, in Czigler et al. study [9], MMN amplitude in the 2.4 s ISI, although not significant, was smaller than the MMN amplitudes in other ISIs. By considering P50 wave amplitude as “gating” of stimuli, higher P50 response amplitude to deviants in comparison to standards might indicate a biased “gating in” of deviants relative to standards [7]. As a result of gating in of deviant stimuli and gating out of standard stimuli, the coding of the deviant stimuli in the comparison process may compete with the standard stimuli for the long auditory store as if its actual probability increased in the train of stimuli. Increasing the probability of deviants relative to standards was shown to cause decrease in MMN amplitude [28]. In fast dichotic stimulus presentation, MMN elicited by deviant tones was found to be barely discernible in the unattended channel and it was suggested that, when gating of processing in unattended channels at an early sensory level occurs, the input to the comparator mechanism upon which the MMN depends is insufficient [32]. In our data the opposite situation has taken place. That is, there was increased gating in rather than decreased gating to the comparator mechanism. So another explanation could be that, over gating-in of deviant stimuli and gating out of standard stimuli, reflected by increased P50d wave, might have inhibitory effect on the comparison process as a negative feedback mechanism and caused the output of this process, that is MMN, to have a small amplitude. The suggested func-
tion of P50 as a prefrontal control of input to primary auditory regions [22] may also support this explanation. The generator sites of P50 are proposed to be the primary auditory cortex in the planum temporale [27] and the prefrontal cortex [26] as with MMN [15,1] and pedinculopontine nucleus (PPN) [17]. The PPN is closely associated with the lateral dorsal tegmental nucleus both anatomically and functionally. They are cholinergic centers displaying rhythmic discharges and their widespread projections to extrapyramidal, limbic, thalamic and cortical areas via dorsal and ventral tegmental bundles imply significant upstream influence by those two nuclear groups. Those two groups have extensive reciprocal connections with two nearby monoaminergic nuclei, the substantia nigra pars compacta (dopamine) and locus ceruleus (noradrenaline), both of them also demonstrate rhythmic discharges and both of have parallel ascending connections with those of PPN and lateral dorsal tegmental nucleus [3]. The neurotransmitters that are released from the aforementioned nuclei are implicated in sensory gating [12,6,10]. Those brainstem oscillators are strongly influenced by afferent sensory information of sufficient magnitude, will prompt reverberations within the system, which will then be projected to higher cortical centers [3]. PPN stimulation is able to produce activation in the cortex as revealed by decrease in slower rhythms (0–8 Hz), as well as increase in the power of fast rhythms (24–33 Hz) [31]. The increased gating in of deviants and increased gating out of standards suggest increased signal-to noise ratio in the 2.5 s ISI. The effect of PPN stimulation on its targets varies according to the frequency of stimulation (from no response to a prolonged response) [14]. Thus, the oscillatory nature of this system might function optimally with 2.5 s ISI auditory stimulation which might be a candidate for future research. In conclusion our study suggests that auditory sensory gating could affect preattentive auditory memory by supplying input to comparison processes of the preattentive auditory processes. Recording ERP in passive oddball paradigm with 2.5 s ISI is useful to demonstrate the interaction of sensory gating and preattentive auditory processes. References [1] K. Alho, Cerebral generators of mismatch negativity (MMN) and its magnetic counterpart (MMNm) elicited by sound changes, Ear Hear 16 (1996) 38–50. [2] K. Alho, D.L. Woods, A. Algazi, R.T. Knight, R. N¨aa¨ t¨anen, Lesions of frontal cortex diminish auditory mismatch negativity, Electroencephalogr. Clin. Neurophysiol. 91 (1994) 353–362. [3] F.A. Boop, E. Garcia-Rill, R. Dykman, R.D. Skinner, The P50: insights into attention and arousal, Pediatr. Neurosurg. 20 (1994) 57–62. [4] C. B¨ottcher-Gandor, P. Ullsperger, Mismatch negativity in event-related potentials to auditory stimuli as a function of varying interstimulus interval, Psychophysiology 29 (1992) 546–550. [5] N.N. Boutros, A. Belger, D. Campbell, C. D’Souza, J. Krystal, Comparison of four components of sensory gating in schizophrenia and normal subjects: a preliminary report, Psychiatry Res. 88 (1999) 119–130. [6] N.N. Boutros, D. Campbell, I. Pertakis, J. Krystal, M. Caporale, T. Kosten, Cocaine use and the mid-latency auditory evoked responses, Psychiatry Res. 96 (2000) 117–126.
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5 [7] N.N. Boutros, M.W. Torello, B.A. Barker, P.A. Tueting, S.C. Wu, The P50 evoked potential component and mismatch detection in normal volunteers: implications for the study of sensory gating, Psychiatry Res. 57 (1995) 83–88. [8] T.H. Bullock, S. Karam¨ursel, J.Z. Achimowicz, M.C. McClune, C. BasarEroglu, Dynamic properties of human visual evoked and omitted stimulus potentials, Electroencephalogr. Clin. Neurophysiol. 91 (1994) 42–53. [9] Czigler, G. Csibra, A. Csontos, Age and inter-stimulus effects on event related potentials to frequent and infrequent auditory stimuli, Biol. Psychol. 33 (1992) 195–206. [10] M.N. Ermutlu, S. Karam¨ursel, E.U. Ugur, L. Senturk, N. Gokhan, Effects of mild stress on early and late stimulus gating, Psychiatry Res. 136 (2005) 201–208. [11] R.J. Erwin, M. Mawwhinney-Hee, R.C. Gur, R.E. Gur, Midlatency auditory evoked responses in Schizophrenia, Biol. Psychiatry 30 (1991) 430–442. [12] R. Freedman, L.E. Adler, P. Bickford, W. Byerley, H. Coon, C.M. Cullum, J.M. Griffith, J.G. Harris, S. Leonard, C. Miller, M. Myles-Worsley, H.T. Nagamoto, G. Rose, M. Waldo, Schizophrenia and nicotinic receptors, Harv. Rev. Psychiatry 2 (1994) 179–192. [13] H. Fruhstorfer, P. Soveri, T. Jarvilehto, Short-term habituation of the auditory evoked response in man, Electroencephalogr. Clin. Neurophysiol. 28 (1970) 153–161. [14] E. Garcia-Rill, R.D. Skinner, H. Miyazato, Y. Homma, Pedunculopontine stimulation induces prolonged activation of pontine reticular neurons, Neuroscience 104 (2001) 455–465. [15] M. Giard, F. Perrin, J. Pernier, P. Bouchet, Brain generators implicated in processing of auditory stimulus deviance: a topographic event-related potential study, Psychophysiology 27 (1990) 627–640. [16] S. Grimm, E. Schr¨oger, Pre-attentive and attentive processing of temporal and frequency characteristics within long sounds, Cogn. Brain Res. 25 (2005) 711–721. [17] C.L. Harrison, J.S. Buchwald, B.M. Chen, Depth evoked potential and single unit correlates of vertex midlatency auditory responses, Brain Res. 264 (1983) 57–67. [18] U. Isoglu-Alkac, G. Keskindemirci, S. Karam¨ursel, Auditory on- and offresponses adn alpha oscillations in the human EEG, Int. J. Neuroscience. 114 (2004) 879–906. [19] P. J¨aa¨ skelainen, E. Schr¨oger, R. N¨aa¨ t¨anen, Electrophysiological indices of acute effects of ethanol on involuntary attention shifting, Psychopharmacology (Berl.) 141 (1999) 16–21.
5
[20] D.C. Javitt, S. Grochowski, A.-M. Shelly, 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. [21] K. Jerger, C. Biggins, G. Fein, P50 suppression is not affected by attentional manipulation, Biol. Psychiatry 31 (1992) 365–377. [22] R.T. Knight, D. Scabini, D.L. Woods, Prefrontal cortex gating of auditory transmissions in humans, Brain Res. 504 (1989) 338–342. [23] R. N¨aa¨ t¨anen, The role of attention in the auditory information processing as revealed by event-related potentials and other brain measures of cognitive function, Behav. Brain Sci. 13 (1990) 201–288. [24] R. N¨aa¨ t¨anen, 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. [25] H.T. Nagamoto, L.E. Adler, M.C. Waldo, R. Freedman, Sensory gating in schizophrenics and normal controls: effects of changing stimulus interval. Biol. Psychiatry 25 (1989) 549–561. [26] G.F. Potts, J. Dien, A.L. Hartry-Speiser, L.M. McDougal, D.M. Tucker, Dense sensor array topography of the event-related potential to taskrelevant auditory stimuli, Electroencephalogr. Clin. Neurophysiol. 106 (1998) 444–456. [27] M. Reite, P. Teale, J. Zimmermann, K. Davis, J. Whalen, Source location of a 50 ms latency auditory evoked field component, Electroencephalogr. Clin. Neurophysiol. 70 (1988) 490–498. [28] M. Sabri, K.B. Campbell, Effects of sequential and temporal probability deviance occurrence on mismatch negativity, Cogn. Brain Res. 12 (2001) 171–180. [29] E. Schr¨oger, On the detection of auditory deviation: a pre-attentive activation model, Psychophysiology 34 (1997) 245–257. [30] D. Smith, N.N. Boutros, S. Schwartzkopf, Reliability of P50 auditory eventrelated potential indices of sensory gating, Psychophysiology 31 (1994) 495–502. [31] M. Steriade, F. Amzica, A. Nunez, Cholinergic and noradrenergic modulation of the slow (approximately 0.3 Hz) oscillation in neocortical cells, J. Neurophysiol. 70 (1993) 1385–1400. [32] M.G. Woldorff, S.A. Hillyard, C.C. Gallen, S.R. Hampson, F.E. Bloom, Magnetoencephalographic recordings demonstrate attentional modulation of mismatch-related neuronal activity in human auditory cortex, Psychophysiology 35 (1998) 283–292.
The effects of interstimulus interval on sensory gating and on preattentive auditory memory in the oddball paradigm Can magnitude of the sensory gating affect preattentive auditory comparison process? M. Numan Ermutlu a,∗ , Tamer Demiralp b , Sacit Karam¨ursel b a b
Istanbul Bilim University, Faculty of Medicine, Department of Physiology, Turkey Istanbul University, Istanbul Medical Faculty, Department of Physiology, Turkey
Received 12 July 2006; received in revised form 31 August 2006; accepted 1 September 2006
Abstract P50, and mismatch negativity (MMN) are components of event-related potentials (ERP) reflecting sensory gating and preattentive auditory memory, respectively. Interstimulus interval (ISI) is an important determinant of the amplitudes of these components and N1. In the present study the interrelation between stimulus gating and preattentive auditory sensory memory were investigated as a function of ISI in 1.5, 2.5 and 3.5 s in 15 healthy volunteered participants. ISI factor affected the N1 peak amplitude significantly. MMN amplitude in 2.5 s ISI was significantly smaller compared to 1.5 and 3.5 s ISI. ISI X stimuli interaction on P50 amplitude was statistically significant. P50 amplitudes to deviant stimuli in 2.5 s ISI were larger than the P50 amplitudes in other ISIs. P50 difference (P50d) waveform amplitude correlated significantly with MMN amplitude. The results suggest that: (i) auditory sensory gating could affect preattentive auditory sensory memory by supplying input to the comparator mechanism; (ii) 2.5 s ISI is important in displaying the sensory gating and preattentive auditory sensory memory relation. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Sensory gating; P50; N100; Mismatch negativity (MMN); Preattentive sensory processing
Sensory gating can be defined as the ability of the brain to modulate its sensitivity to incoming sensory stimuli. This definition allows the concept of gating to include both the capacities to minimize or stop responding to incoming irrelevant stimuli (gating out) and to respond when a novel stimulus is presented or a change occurs in ongoing stimuli (gating in) [5]. The P50 component of event-related potentials (ERP) has been used to study sensory gating [11,12]. The P50 is a positive component of ERP that appears between 35 and 85 ms following auditory stimulation and reflect mainly sensory gating and early preattentional processing [21]. It is elicited in paired click paradigm, short trains paradigm, and oddball paradigm. In a paired click paradigm, the first stimulus (S1) of the pair is followed by an identical stimulus (S2) a short time ∗
Corresponding author at: Istanbul Bilim University, Faculty of Medicine, Department of Physiology, Buyukdere cad No: 120 Esentepe, 34394 Istanbul, Turkey. Tel.: +90 212 2136486; fax: +90 212 2723461. E-mail addresses: [email protected], [email protected] (M.N. Ermutlu). 0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.09.006
later (e.g., 500 ms). The inhibitory capacity of the brain is then measured as the ratio of the amplitude of the P50 to S2 stimuli to the amplitude of the P50 response to S1 stimuli (S2/S1), or as the mathematical difference between the two responses [30]. In the oddball paradigm, on the other hand, increase in amplitude of P50 to infrequent auditory stimuli was shown to reflect preattentive recognition of novel stimuli or gating in of stimuli while decrease in amplitude of P50 to frequent auditory stimuli reflects gating out of stimuli [7]. In the context of oddball paradigm, sensory gating can be operationally defined as the ratio of the amplitude of the response to the infrequent stimulus divided by amplitude of the response to frequent stimulus [7] and as the arithmetical difference between the amplitudes of responses to deviant (infrequent) stimulus and standard (frequent) stimulus. Higher ratios or differences can either reflect lower amplitudes in response to frequent stimuli, and thus stronger inhibition with repetition, or higher amplitudes in response to infrequent stimuli and, thus, stronger response to rare stimuli [7]. Infrequent auditory stimuli deviating from a repetitive standard sound in some physical feature such as frequency elicit
2
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
the mismatch negativity (MMN) component of the event-related brain potential. The MMN is thought to reflect the outcome of a mismatch process automatically registering the deviation of the current input from the neuronal representation of a repetitive stimulus in sensory memory. MMN reflects preattentive memory and memory comparison process [23]. Dysfunction of sensory gating and preattentive sensory processing has been implicated in the pathophysiology of a number of psychiatric disorders and there are some findings that suggest an interrelation between the two processes as reflected in the mismatch negativity and P50 components of event-related potentials: (i) patients with frontal lesions displayed MMN attenuation and enhanced P50 amplitudes [2]; (ii) MMN and P50 patterns observed in Schizophrenia differs from the patterns observed in normals and in some patients with neurological disorders [20,25]; (iii) P50 amplitude increases and MMN amplitude decreases with aging; (iv) with alcohol consumption, P50 amplitude was significantly augmented and peak latency of MMN was significantly increased [19]. The amplitude of P50 has a tendency to decrease in amplitude when the stimuli were presented with fast rates of repetition [13]. The effects of interstimulus interval (ISI) on ERP components are also important in the oddball paradigm. The MMN and the N1 have different relations with the ISI [24]. ON responses are transient potential changes at the beginning of a stimulus and OFF responses are the potential changes after the end of a continuous stimulus [8]. Both of them are in quite similar shape and amplitude depending on the duration of the stimulus and ISI [18]. Since the mid-latency ERP components of our interest are within 300 ms after stimulus and as MMN was shown to be elicited with long duration stimuli [16], long duration auditory stimuli were used to prevent the possible superimposition of ON and OFF responses. In the present study, a possible interaction between P50, N1 and MMN were investigated in an oddball paradigm by varying the ISI. Fifteen healthy right-handed volunteers (six females and nine males; mean age 34.1 range 25–39) were studied. None of the subjects had a personal or family history of psychiatric or neurological disorders. They also had no personal history of illicit drug use. They were not receiving any medications during the study. The study was conducted in accordance with the Declaration of Helsinki and all procedures were carried out with the adequate understanding and written consent of the participants. The participants were instructed not to take alcohol or any drug for the 48 h preceding recording. Two of the volunteers were smokers and they had not smoked for 2 h before recording. The subjects sat in an electrically shielded and soundproof room. They were asked to read a self-chosen interesting book and to ignore the stimuli given through earphones. The EEG was measured from Fz, F3, F4, and Cz recording sites according to international 10/20 system with reference to linked ear lobes. All data were amplified by using a low-pass filter with cut-off frequency at 70 Hz and with a time constant of 0.3 s. The signals were sampled at a rate of 256 points/s. ERPs were recorded in three sessions.
Three ISI (onset to onset) were used: 1.5, 2.5, and 3.5 s. In all conditions stimulus duration was 500 ms and intensity was 80 dB SPL. The standards were 1050 Hz tones and the deviants were 1250 Hz tones. The rise and fall-times were 10 ms. The probability of the deviant and standard stimuli were equal (20/80) across all three ISI conditions. Deviant stimuli did not occur in direct succession. Artifacts were eliminated by manual off-line selective averaging taking into consideration the EOG recorded from the right eye. The sweep numbers were equalized between the standards and deviants and between the conditions. P50 difference (P50d) and MMN waves were assessed by subtracting the standard responses from the deviant responses. The P50 wave was identified as the most positive peak in the 40–80 ms time window, whereas the mean amplitude and latency of MMN was measured in the time window between 100 and 200 ms. Additionally, P50 and N1 waves were separately measured in standard and deviant ERPs (the latter as the largest negative peak between 80 and 120 ms). Since P50 had negative values in some recordings, deviant minus standard P50 differences were assessed instead of P50 ratios. For the statistical evaluation, a repeated measures ANOVA design with ISI (three levels: 1.5, 2.5, 3.5 s), stimulus (two levels: standard, deviant) and channel (four levels: Fz, F3, F4, Cz) within-subject factors were applied to the peak amplitudes of P50, N1 and peak latency of N1. For MMN and P50d, an ANOVA design with the within-subject factors ISI (three levels: 1.5, 2.5, 3.5 s) and channel (four levels Fz, F3, F4, Cz) were applied to the peak amplitudes of P50d, MMN and peak latencies of MMN. Reduced degrees freedom (Greenhouse-Geisser) was used to counter violations of the sphericity assumption underlying ANOVA with repeated measures. None of the components was significantly affected by channel. ISI factor affected the N1 peak amplitudes significantly (F = 43.37, d.f. = 2.28; p < 0.001). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni) N1 amplitudes increased significantly with increase in ISI. ISI X stimulus interaction on N1 amplitude was significant (F = 32.92, d.f. = 2.28, p < 0.0001). In paired sample t-tests, N1 amplitudes to deviants were significantly larger than N1 amplitudes to standards in 1.5 s ISI (t = 3.99, d.f. = 14; p < 0.002, two-tailed) (Table 1, Fig. 1). ISI was found to have a significant effect on MMN peak amplitudes (F = 7.60, d.f. = 2.28; p < 0.002). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni), peak amplitude in 2.5 s ISI was significantly smaller than peak amplitudes in 1.5 s ISI (p < 0.05) and in 3.5 s ISI (p < 0.05) (Figs. 2 and 3). ISI was found to have no effect on MMN peak latency. P50 was neither significantly affected by ISIs nor by stimuli. However, ISI X stimulus interaction on P50 amplitude was significant (F = 3.53, d.f. = 2.28; p < 0.03). In paired t-tests, amplitude of P50 to deviant stimuli was significantly larger in 2.5 s ISI than amplitudes of P50s to deviant stimuli in 1.5 and 3.5 s ISIs (t = −0.19, d.f. = 14; p < 0.03, t = 2.61, d.f. = 14; p < 0.02).
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
3
Table 1 Mean amplitudes of the ERP components in the three ISIs ERP components (mean ± S.D.)
1.5 s ISI
P50 (deviants) P50 (standards) N1 (deviants) N1 (standards) P50 difference MMN
1.20 V 1.96 V −5.18 V −3.73 V −0.72 V −3.99 V
2.5 s ISI ± ± ± ± ± ±
1.66 1.95 1.193c 2.18 2.86 1.152
2.47 V 1.23 V −7.58 V −7.48 V 1.24 V −1.91 V
3.5 s ISI ± ± ± ± ± ±
1.68a,b 1.79 2.85 2.53 2.28d,e 1.15f
1.40 V 2.31 V −9.43 V −8.55 V −0.91 −4.19
± ± ± ± ± ±
1.35 1.68 3.0 3.45 1.78 2.05
Values indicate the means of the Fz, F3, F4, and Cz recording sites. a Larger than P50 amplitude to deviants in 1.5 s (p < 0.03). b Larger than P50 amplitude to deviants in 3.5 s (p < 0.02). c Larger than N1 amplitude to standards in 1.5 s (p < 0.002). d Larger than P50d amplitude in 1.5 s (p < 0.05). e Larger than P50d amplitude in 3.5 s (p < 0.06). f Smaller than MMN amplitudes in 1.5 and 3.5 s (p < 0.05).
Fig. 3. Mean amplitudes P50 to deviants and standards in 1.5, 2.5 and 3.5 s ISIs at Fz.
Fig. 1. Grand average ERPs of 15 participants to deviants and standards in 1.5, 2.5 and 3.5 s at Fz.
Fig. 2. Grand average difference waveforms of 15 participants obtained by subtracting ERPs to standard tones from those to deviants in 1.5, 2.5 and 3.5 s ISIs. MMN and P50d waveforms are indicated on the figure.
4
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5
P50d was significantly affected by ISI (F = 4.25, d.f. = 2.28; p < 0.03, ε = 0.66). In pair-wise comparisons (adjustments for multiple comparisons: Bonferroni), P50d peak amplitude in 2.5 s ISI was larger than the peak amplitudes in 1.5 s (p < 0.05) and in 3.5 s ISIs (p < 0.06) (Figs. 2 and 3). Correlation analyses were carried out for MMN and P50d by pooling the data in all ISI conditions. There was a significant correlation between P50d and MMN amplitudes (r = 0.38, p = 0.01). In this study, the increases in the amplitude of the N1 component of ERP as a function of ISI and larger amplitude of N1 to deviants than the amplitude of N1 to standards in the 1.5 s ISI were expected findings. The increase in N1 amplitude with increases in ISI could be function of recovery of complex neuronal circuits generating N1 [24]. However, P50 and MMN amplitudes did not display the same feature as N1. The amplitude of P50 to deviants in 2.5 s ISI was significantly higher than the amplitude of P50 to deviants in the other ISIs. Also the amplitude of P50d was significantly higher and the amplitude of MMN was significantly smaller in 2.5 s ISI than other ISIs. The data suggest that there were increased gating in the 2.5 s ISI in comparison to other ISIs and significant increase in the amplitude of the P50d in the 2.5 s ISI is the result of significant increase in the amplitude of the deviant stimuli (gating in) and decreased amplitude of the standard stimuli (gating out). The absolute amplitude of MMN was significantly smaller in the 2.5 s ISI than 1.5 s, and 3.5 s ISIs. MMN is thought to reflect a neuronal mismatch between an incoming stimulus or the representation of the incoming stimulus in the short term store [29] and representations of those presented over the previous 10 s, that is the long term store. It is reported that the MMN peak amplitudes are not affected by ISI in the 0.8–7.2 s [9] and 1–10 s ranges [4]. However, in Czigler et al. study [9], MMN amplitude in the 2.4 s ISI, although not significant, was smaller than the MMN amplitudes in other ISIs. By considering P50 wave amplitude as “gating” of stimuli, higher P50 response amplitude to deviants in comparison to standards might indicate a biased “gating in” of deviants relative to standards [7]. As a result of gating in of deviant stimuli and gating out of standard stimuli, the coding of the deviant stimuli in the comparison process may compete with the standard stimuli for the long auditory store as if its actual probability increased in the train of stimuli. Increasing the probability of deviants relative to standards was shown to cause decrease in MMN amplitude [28]. In fast dichotic stimulus presentation, MMN elicited by deviant tones was found to be barely discernible in the unattended channel and it was suggested that, when gating of processing in unattended channels at an early sensory level occurs, the input to the comparator mechanism upon which the MMN depends is insufficient [32]. In our data the opposite situation has taken place. That is, there was increased gating in rather than decreased gating to the comparator mechanism. So another explanation could be that, over gating-in of deviant stimuli and gating out of standard stimuli, reflected by increased P50d wave, might have inhibitory effect on the comparison process as a negative feedback mechanism and caused the output of this process, that is MMN, to have a small amplitude. The suggested func-
tion of P50 as a prefrontal control of input to primary auditory regions [22] may also support this explanation. The generator sites of P50 are proposed to be the primary auditory cortex in the planum temporale [27] and the prefrontal cortex [26] as with MMN [15,1] and pedinculopontine nucleus (PPN) [17]. The PPN is closely associated with the lateral dorsal tegmental nucleus both anatomically and functionally. They are cholinergic centers displaying rhythmic discharges and their widespread projections to extrapyramidal, limbic, thalamic and cortical areas via dorsal and ventral tegmental bundles imply significant upstream influence by those two nuclear groups. Those two groups have extensive reciprocal connections with two nearby monoaminergic nuclei, the substantia nigra pars compacta (dopamine) and locus ceruleus (noradrenaline), both of them also demonstrate rhythmic discharges and both of have parallel ascending connections with those of PPN and lateral dorsal tegmental nucleus [3]. The neurotransmitters that are released from the aforementioned nuclei are implicated in sensory gating [12,6,10]. Those brainstem oscillators are strongly influenced by afferent sensory information of sufficient magnitude, will prompt reverberations within the system, which will then be projected to higher cortical centers [3]. PPN stimulation is able to produce activation in the cortex as revealed by decrease in slower rhythms (0–8 Hz), as well as increase in the power of fast rhythms (24–33 Hz) [31]. The increased gating in of deviants and increased gating out of standards suggest increased signal-to noise ratio in the 2.5 s ISI. The effect of PPN stimulation on its targets varies according to the frequency of stimulation (from no response to a prolonged response) [14]. Thus, the oscillatory nature of this system might function optimally with 2.5 s ISI auditory stimulation which might be a candidate for future research. In conclusion our study suggests that auditory sensory gating could affect preattentive auditory memory by supplying input to comparison processes of the preattentive auditory processes. Recording ERP in passive oddball paradigm with 2.5 s ISI is useful to demonstrate the interaction of sensory gating and preattentive auditory processes. References [1] K. Alho, Cerebral generators of mismatch negativity (MMN) and its magnetic counterpart (MMNm) elicited by sound changes, Ear Hear 16 (1996) 38–50. [2] K. Alho, D.L. Woods, A. Algazi, R.T. Knight, R. N¨aa¨ t¨anen, Lesions of frontal cortex diminish auditory mismatch negativity, Electroencephalogr. Clin. Neurophysiol. 91 (1994) 353–362. [3] F.A. Boop, E. Garcia-Rill, R. Dykman, R.D. Skinner, The P50: insights into attention and arousal, Pediatr. Neurosurg. 20 (1994) 57–62. [4] C. B¨ottcher-Gandor, P. Ullsperger, Mismatch negativity in event-related potentials to auditory stimuli as a function of varying interstimulus interval, Psychophysiology 29 (1992) 546–550. [5] N.N. Boutros, A. Belger, D. Campbell, C. D’Souza, J. Krystal, Comparison of four components of sensory gating in schizophrenia and normal subjects: a preliminary report, Psychiatry Res. 88 (1999) 119–130. [6] N.N. Boutros, D. Campbell, I. Pertakis, J. Krystal, M. Caporale, T. Kosten, Cocaine use and the mid-latency auditory evoked responses, Psychiatry Res. 96 (2000) 117–126.
M.N. Ermutlu et al. / Neuroscience Letters 412 (2007) 1–5 [7] N.N. Boutros, M.W. Torello, B.A. Barker, P.A. Tueting, S.C. Wu, The P50 evoked potential component and mismatch detection in normal volunteers: implications for the study of sensory gating, Psychiatry Res. 57 (1995) 83–88. [8] T.H. Bullock, S. Karam¨ursel, J.Z. Achimowicz, M.C. McClune, C. BasarEroglu, Dynamic properties of human visual evoked and omitted stimulus potentials, Electroencephalogr. Clin. Neurophysiol. 91 (1994) 42–53. [9] Czigler, G. Csibra, A. Csontos, Age and inter-stimulus effects on event related potentials to frequent and infrequent auditory stimuli, Biol. Psychol. 33 (1992) 195–206. [10] M.N. Ermutlu, S. Karam¨ursel, E.U. Ugur, L. Senturk, N. Gokhan, Effects of mild stress on early and late stimulus gating, Psychiatry Res. 136 (2005) 201–208. [11] R.J. Erwin, M. Mawwhinney-Hee, R.C. Gur, R.E. Gur, Midlatency auditory evoked responses in Schizophrenia, Biol. Psychiatry 30 (1991) 430–442. [12] R. Freedman, L.E. Adler, P. Bickford, W. Byerley, H. Coon, C.M. Cullum, J.M. Griffith, J.G. Harris, S. Leonard, C. Miller, M. Myles-Worsley, H.T. Nagamoto, G. Rose, M. Waldo, Schizophrenia and nicotinic receptors, Harv. Rev. Psychiatry 2 (1994) 179–192. [13] H. Fruhstorfer, P. Soveri, T. Jarvilehto, Short-term habituation of the auditory evoked response in man, Electroencephalogr. Clin. Neurophysiol. 28 (1970) 153–161. [14] E. Garcia-Rill, R.D. Skinner, H. Miyazato, Y. Homma, Pedunculopontine stimulation induces prolonged activation of pontine reticular neurons, Neuroscience 104 (2001) 455–465. [15] M. Giard, F. Perrin, J. Pernier, P. Bouchet, Brain generators implicated in processing of auditory stimulus deviance: a topographic event-related potential study, Psychophysiology 27 (1990) 627–640. [16] S. Grimm, E. Schr¨oger, Pre-attentive and attentive processing of temporal and frequency characteristics within long sounds, Cogn. Brain Res. 25 (2005) 711–721. [17] C.L. Harrison, J.S. Buchwald, B.M. Chen, Depth evoked potential and single unit correlates of vertex midlatency auditory responses, Brain Res. 264 (1983) 57–67. [18] U. Isoglu-Alkac, G. Keskindemirci, S. Karam¨ursel, Auditory on- and offresponses adn alpha oscillations in the human EEG, Int. J. Neuroscience. 114 (2004) 879–906. [19] P. J¨aa¨ skelainen, E. Schr¨oger, R. N¨aa¨ t¨anen, Electrophysiological indices of acute effects of ethanol on involuntary attention shifting, Psychopharmacology (Berl.) 141 (1999) 16–21.
5
[20] D.C. Javitt, S. Grochowski, A.-M. Shelly, 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. [21] K. Jerger, C. Biggins, G. Fein, P50 suppression is not affected by attentional manipulation, Biol. Psychiatry 31 (1992) 365–377. [22] R.T. Knight, D. Scabini, D.L. Woods, Prefrontal cortex gating of auditory transmissions in humans, Brain Res. 504 (1989) 338–342. [23] R. N¨aa¨ t¨anen, The role of attention in the auditory information processing as revealed by event-related potentials and other brain measures of cognitive function, Behav. Brain Sci. 13 (1990) 201–288. [24] R. N¨aa¨ t¨anen, 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. [25] H.T. Nagamoto, L.E. Adler, M.C. Waldo, R. Freedman, Sensory gating in schizophrenics and normal controls: effects of changing stimulus interval. Biol. Psychiatry 25 (1989) 549–561. [26] G.F. Potts, J. Dien, A.L. Hartry-Speiser, L.M. McDougal, D.M. Tucker, Dense sensor array topography of the event-related potential to taskrelevant auditory stimuli, Electroencephalogr. Clin. Neurophysiol. 106 (1998) 444–456. [27] M. Reite, P. Teale, J. Zimmermann, K. Davis, J. Whalen, Source location of a 50 ms latency auditory evoked field component, Electroencephalogr. Clin. Neurophysiol. 70 (1988) 490–498. [28] M. Sabri, K.B. Campbell, Effects of sequential and temporal probability deviance occurrence on mismatch negativity, Cogn. Brain Res. 12 (2001) 171–180. [29] E. Schr¨oger, On the detection of auditory deviation: a pre-attentive activation model, Psychophysiology 34 (1997) 245–257. [30] D. Smith, N.N. Boutros, S. Schwartzkopf, Reliability of P50 auditory eventrelated potential indices of sensory gating, Psychophysiology 31 (1994) 495–502. [31] M. Steriade, F. Amzica, A. Nunez, Cholinergic and noradrenergic modulation of the slow (approximately 0.3 Hz) oscillation in neocortical cells, J. Neurophysiol. 70 (1993) 1385–1400. [32] M.G. Woldorff, S.A. Hillyard, C.C. Gallen, S.R. Hampson, F.E. Bloom, Magnetoencephalographic recordings demonstrate attentional modulation of mismatch-related neuronal activity in human auditory cortex, Psychophysiology 35 (1998) 283–292.