RuusuvirtapersName, LipponenpersName, PellinenpersName, PenttonenpersName, AstikainenpersName (2013date). “Auditory Cortical and Hippocampal-System Mismatch Responses to Duration Deviants in Urethane-Anesthetized Rats.” PLoS ONEtitle·j.
Object of study — whether mismatch responsesterm to duration deviantsterm appear in the hippocampal systemterm at the same latency as in auditory cortexterm, or only later and secondarily
Research question — is the received cortico-hippocampal division of labour — MMNterm·def cortical, P3aterm hippocampal — actually supported
Contribution — finds mismatch responsesterm in auditory cortexterm and in three hippocampal sites at the same latency and in the same stimulus condition, which weakens the case for the hippocampus being merely downstream
phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | auditory mismatch responses to duration deviants — “The mismatch negativity (MMN) of event-related potentials (ERPs) reflects this initial stage of auditory change detection.” | |
| claim-type | existence — “We found differential LFPs (mismatch responses) to duration deviant tones in the auditory cortex and in the hippocampal system (CA1, dentate gyrus, and subiculum) in urethane-anesthetized rats.” | |
| method — evidence | — “In the present study, local-field potentials (LFPs) in urethane-anesthetized rats were recorded on the surface of the primary auditory cortex and in three locations of the hippocampal system (dentate gyrus, CA1, and subiculum, Figure 1 ).” | |
| object of study — evidence | — “Our interest was whether mismatch responses, whenever elicited, occupied both the auditory cortex and the hippocampus in a typical latency range of mismatch responses in urethane anesthetized rats or whether mismatch responses were only manifested in the auditory cortex and possibly only followed by later hippocampal activity.” | |
| research question — evidence | — “The evidence for this cortico-hippocampal dichotomy is scarce, however.” | |
| contribution — evidence | — “Most importantly, hippocampal and auditory cortical mismatch responses were found in the same latency range and in the same stimulus condition.” | |
| transfer | from human event-related potential research: the entire oddball paradigm and the deviant-minus-standard subtraction logic, carried into rodent local field potentials with the polarity requirement dropped — “MMN is quantified by arithmetically subtracting ERPs to the standard tone from ERPs to the deviant tone.” | transfers[0] |
| transfer | from human ERP research: the MMN-then-P3a sequence, imported as the prediction the animal data would have to match if the hippocampal response were merely downstream — “This would be analogous to a MMN-P3a sequence in humans that is only observable in full (i.e., also including P3a) when deviant tones of high magnitude attract involuntary switches of attention [1] , [4] .” | transfers[1] |
| mechanism | A long standard-to-deviant interval lets the transient memory trace of the standard decay, so the deviant can no longer be registered as mismatching it, abolishing the response — “Secondly, the long (500-ms) SDI might have led to a high degree of degradation of the memory trace of the standard tone, preventing the detection of the deviant tone as not matching this trace [1] , [7] .” | mechanisms[0] |
| mechanism | A short inter-deviant interval re-activates the mismatch generator too rapidly, diminishing the amplitude of the response it can produce — “First, the short inter-deviant interval (due to the 200-ms SSI) might have diminished mismatch responses by the fast-paced activation of their neural generator (in humans [43] – [45] , in cats [15] ).” | mechanisms[1] |
| mechanism | In the responsive conditions the deviant differed from the standard on two features (duration and inter-stimulus interval) instead of one, making the change easier to detect — “The present pattern of observations may also reflect the fact that the deviant tone was distinguishable from the standard tone by two auditory features (tone duration and inter-stimulus interval) in the former pair of conditions as opposed to solely one feature (tone duration) in the latter pair [1] , [47] .” | mechanisms[2] |
| mechanism | the animal analogue is given a deliberately looser definition than the human one, dropping the polarity requirement so a cross-species comparison is possible at all — “Mismatch responses (i.e., higher-amplitude brain responses to deviant tones than standard tones irrespectively of the polarity of these responses and despite these tones are only passively listened to)” | mechanisms[3] |
| mechanism | that loosening turns out to be load-bearing rather than cautious: the cortical response is positive and the hippocampal one negative in the very same animals — “The responses - positive polarity in the auditory cortical and negative polarity in the hippocampal sites - appeared in the same latency range (between 51.5 and 89 ms post-change) and in all the recording sites, but only in the SSI 500 /SDI 200 -condition.” | mechanisms[4] |
| mechanism | co-occurrence in latency and in condition is the argument against a downstream account — if the hippocampus were merely triggered by cortex it should lag — “This correspondence suggests that the hippocampal responses reflect the initial detection of deviant tones similarly to auditory cortical mismatch responses.” | mechanisms[5] |
| mechanism | duration rather than frequency is chosen as the deviant feature specifically to avoid the refractoriness confound that a tonotopic dimension would introduce — “Sound frequency was not used due to its spatial coding from the cochlea [41] owing to a potential confounding effect of lower levels of refractoriness in afferent pathways activated by the rare deviant tone than the frequently occurring standard tones [1] .” | mechanisms[6] |
| mechanism | two intervals are crossed so that the memory-trace account and the generator-refractoriness account make different predictions across the four cells — “Two standard-to-standard (SSI) and standard-to-deviant (SDI) intervals (200 ms vs. 500 ms) were applied in different combinations to vary the observability of responses resembling MMN (mismatch responses).” | mechanisms[7] |
| mechanism | series order is counterbalanced across animals specifically because anesthetic depth drifts with time on the rig — “was counterbalanced across the animals to control for effects of the duration of stimulus exposure and of the passage of time itself (e.g., through variations in the level of anesthesia)” | mechanisms[8] |
| mechanism | the omnibus ANOVA is used as a licence for the post-hoc comparisons rather than as the result, and the parametric choice is stated as forced — “As there were no suitable non-parametric alternatives, ANOVA for repeated measures with stimulus (standard, deviant) and time-segment (first, second, third) as factors was used to justify post-hoc deviant-standard LFP amplitude comparisons for each time segment separately.” | mechanisms[9] |
| mechanism | electrode positions are verified histologically after the fact and only confirmed placements enter the analysis, so the sample size differs by structure — “For the hippocampal recordings, the electrodes successfully implanted in the targeted areas were applied in the analysis (9 animals for the subiculum, 8 for the CA1 area, and 6 for the dentate gyrus, Figure 1 )” | mechanisms[10] |
| mechanism | the baseline is taken from a window straddling tone onset, so that time zero is the moment of deviance rather than the moment of sound — “baseline corrected against the mean activity in the ±25 ms from tone onset and, hence in the 50-ms time window prior to duration deviance” | mechanisms[11] |
| mechanism | the received dichotomy is attacked on three separate grounds — too little data, an independent hippocampal novelty mechanism, and existing animal evidence — rather than on the new result alone — “The MMN-hippocampus dissociation may, however, be premature for a number of reasons.” | mechanisms[12] |
| critiques | the received cortico-hippocampal dichotomy in which the hippocampus contributes to P3a but not to MMN — “The MMN-hippocampus dissociation may, however, be premature for a number of reasons.” | claim_relations[0] |
| contradicts | intracranial human studies reporting no hippocampal mismatch responses, or only longer-latency novelty responses — “Studies employing intracranial hippocampal recordings in neurological patients have reported no mismatch responses [33] or other, longer-latency differential responses to deviant tones as novel or salient stimuli rather than as changes in standards [34] .” | claim_relations[1] |
| replicates | previous cortical mismatch findings in urethane-anesthetized rats, in polarity and latency — “The polarity and latency range of auditory cortical mismatch responses in the SSI 500 /SDI 200 -condition correspond to previous findings in urethane anesthetized rats [18] , [19] , [22] , [23] .” | claim_relations[2] |
| extends | animal work already reporting hippocampal mismatch responses, some of them specific to change rather than rarity — “Finally, hippocampal mismatch responses have been observed in animals [14] , [16] , and some of these responses have, comparably to MMN [3] , even shown specificity to deviant tones as changes in standard tones [16] .” | claim_relations[3] |
| builds-on | the proposal that the hippocampus responds to rare stimuli by comparison against a model of past repetition — “Second, the hippocampus has been proposed, comparably to MMN [1] , to respond by itself to rare stimuli on the basis of their discrepancy from a neural model of past repetitive stimuli [35] – [38] .” | claim_relations[4] |
| critiques | the state of MMN generator theory generally, as contested and unresolved — “The generators of MMN are largely unknown. Different and partially contradictory explanations have been proposed [8] , [26] – [28] .” | claim_relations[5] |
| sense | mismatch negativity versus mismatch response — the human concept requires negative polarity by name; the animal concept is defined to ignore polarity, and here that difference is load-bearing — “Mismatch responses (i.e., higher-amplitude brain responses to deviant tones than standard tones irrespectively of the polarity of these responses and despite these tones are only passively listened to) have been found in sleeping infants [9] – [11] and in awake, sleeping and anesthetized animals [12] – [23]” | senses[0] |
| sense | deviance as a change in the standard, versus deviance as mere rarity — two accounts the design cannot separate — “Nevertheless, the rarity of the deviant tone as such [26] , [27] also remains a possible explanation for our findings.” | senses[1] |
| sense | time zero — from tone onset versus from the moment of duration change, which fall 25 ms apart and are distinguished throughout the results — “The zero time point is at the moment of change, that is, at 25 ms from tone onset and, thus, at the offset of the 25-ms tone.” | senses[2] |
| sense | deviance as a lengthening versus a shortening of the standard; the two are not equivalent, and only the working conditions included a lengthening — “Mismatch responses in rats have been found to be higher in amplitude to deviant tones as lengthenings than shortenings of standard tones [21] .” | senses[3] |
| sense | change detection versus stimulus-specific adaptation — a response reduction that is feature-specific, held to be a functionally different thing from a mismatch response — “Stimulus-specific adaptation, a decrease in neural responses to a repeated tone which is specific to the physical features of this tone, observed in different stages of the auditory pathway [48] – [51] as functionally distinct from auditory cortical mismatch responses [52] , [53] could provide a useful neurophysiological tool to increase this understanding.” | senses[4] |
| sense | hippocampal response as initial detection versus as a downstream consequence of cortical detection — the two readings the latency correspondence is meant to decide between — “Nevertheless, hippocampal mismatch responses could be claimed to be secondary to and solely triggered by auditory cortical mismatch responses in a limited set of circumstances.” | senses[5] |
| control | the sequential order of the six stimulus series COUNTERBALANCED ACROSS ANIMALS — held — rules out cumulative exposure and the passage of time itself - explicitly including drifting anaesthetic depth - as the source of the difference between conditions, since a fixed order would confound condition with how long the animal had been under — “was counterbalanced across the animals to control for effects of the duration of stimulus exposure and of the passage of time itself (e.g., through variations in the level of anesthesia)” | controls[0] |
| differs in | intracranial human studies reporting no hippocampal mismatch responses, or only longer-latency novelty responses — THE AXIS IS WHAT THE RESPONSE IS TAKEN TO INDEX. Intracranial studies 'reported no mismatch responses [33] or other, longer-latency differential responses to deviant tones as novel or salient stimuli rather than as changes in standards [34].' Two disagreements in one sentence: an absence, and a different reading of a present response. — “Studies employing intracranial hippocampal recordings in neurological patients have reported no mismatch responses [33] or other, longer-latency differential responses to deviant tones as novel or salient stimuli rather than as changes in standards [34] .” — “Studies employing intracranial hippocampal recordings in neurological patients have reported no mismatch responses [33] or other, longer-latency differential responses to deviant tones as novel or salient stimuli rather than as changes in standards” | claim_relations[1] |
| weight | mismatch negativity —operationalised-as→ oddball condition — 0.95 — “MMN is typically experimentally addressed by using the so-called oddball condition.” | concept_edges[2] |
| edge | deviant —contrasted-with→ standard — “In this condition, a (‘deviant’) tone rarely replaces a frequently repeated (‘standard’) tone.” | concept_edges[3] |
| weight | stimulus-specific adaptation —distinct-from→ mismatch responses — 0.9 — “Stimulus-specific adaptation, a decrease in neural responses to a repeated tone which is specific to the physical features of this tone, observed in different stages of the auditory pathway [48] – [51] as functionally distinct from auditory cortical mismatch responses [52] , [53] could provide a useful neurophysiological tool to increase this understanding.” | concept_edges[5] |
| edge | CA1 —part-of→ hippocampal system — “auditory cortical and hippocampal-system (CA1, dentate gyrus, subiculum) local-field potentials were recorded in urethane-anesthetized rats” | concept_edges[6] |
| weight | mismatch responses —operationalised-as→ local-field potentials — 0.95 — “We found differential LFPs (mismatch responses) to duration deviant tones in the auditory cortex and in the hippocampal system (CA1, dentate gyrus, and subiculum) in urethane-anesthetized rats.” | concept_edges[8] |
| outcome (stated by the tagger, no quotation) | control — held | controls[0] |
| for which use | limitation — the two matched-interval conditions specifically: WHY MISMATCH RESPONSES WERE ABSENT PRECISELY IN THE SSI 200/SDI 200- AND SSI 500/SDI 500-CONDITIONS REMAINS TO BE ACCOUNTED FOR — “Therefore, why mismatch responses were absent precisely in the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions remains to be accounted for.” | limitations[0] |
| for which use | limitation — the change-detection reading of the result: THE RARITY OF THE DEVIANT TONE AS SUCH ALSO REMAINS A POSSIBLE EXPLANATION FOR OUR FINDINGS — “Nevertheless, the rarity of the deviant tone as such [26] , [27] also remains a possible explanation for our findings.” | limitations[1] |
| for which use | limitation — the validity of the change-detection account: FUTURE STUDIES WITH APPROPRIATE CONTROL PROCEDURES, SUCH AS DEVIANT-ALONE OR EQUAL-PROBABILITY STIMULUS CONDITIONS, ARE NEEDED TO TEST THE VALIDITY OF THE CHANGE-DETECTION ACCOUNT — “Future studies with appropriate control procedures, such as deviant-alone (e.g. [18] ) or equal-probability (e.g. [19] ) stimulus conditions, are needed to test the validity of the change-detection account.” | limitations[2] |
| for which use | limitation — any claim that the two responses share a mechanism: IT REMAINS TO BE RESOLVED WHETHER THESE RESPONSES OF DIFFERENT ANATOMICAL ORIGINS RECRUIT A SINGLE BUT SPATIALLY DISTRIBUTED NEURAL MECHANISM, SPATIALLY SEPARATE AND INDEPENDENT NEURAL MECHANISMS, OR SPATIALLY SEPARATE BUT INTERCONNECTED ones — “It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms.” | limitations[3] |
| for which use | limitation — the generality of the effect across timing conditions: IN THE REMAINING THREE CONDITIONS, THE FINDINGS WERE NEGATIVE — “In the remaining three conditions, the findings were negative.” | limitations[4] |
| for which use | limitation — the same scope as limitations[0] - the two conditions in which nothing appeared — “Therefore, why mismatch responses were absent precisely in the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions remains to be accounted for.” | limitations[5] |
| for which use | limitation — the same scope as limitations[1] - rarity remains available, so the change-detection reading is not established by this design — “Nevertheless, the rarity of the deviant tone as such [26] , [27] also remains a possible explanation for our findings.” | limitations[6] |
| for which use | limitation — the same scope as limitations[2] - the controls that would settle it were not run — “Future studies with appropriate control procedures, such as deviant-alone (e.g. [18] ) or equal-probability (e.g. [19] ) stimulus conditions, are needed to test the validity of the change-detection account.” | limitations[7] |
| duplicate of | limitation — limitations[3] — “It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms.” | limitations[8] |
| for which use | limitation — the same scope as limitations[3] - one mechanism, two independent, or two interconnected, unresolved — “It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms.” | limitations[8] |
| for which use | limitation — comparisons ACROSS conditions: IN THE SSI 200/SDI 200- AND SSI 500/SDI 500-CONDITIONS, THE STIMULUS ASSIGNMENTS (THE 25-ms TONE AS THE DEVIANT TONE AND THE 75-ms TONE AS THE STANDARD TONE, BUT NOT VICE VERSA) AND NUMBERS (A HALF) OF SWEEPS PER AVERAGED WAVEFORM differed too — “In the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions, the stimulus assignments (the 25-ms tone as the deviant tone and the 75-ms tone as the standard tone, but not vice versa) and numbers (a half) of sweeps per averaged waveform were different from those in the SSI 500 /SDI 200 - and SSI 200 /SDI 500 -conditions.” | limitations[9] |
| for which use | limitation — the conditions where the effect DID appear: THE DEVIANT TONE WAS DISTINGUISHABLE FROM THE STANDARD TONE BY TWO AUDITORY FEATURES (TONE DURATION AND INTER-STIMULUS INTERVAL) IN THE FORMER PAIR OF CONDITIONS — “The present pattern of observations may also reflect the fact that the deviant tone was distinguishable from the standard tone by two auditory features (tone duration and inter-stimulus interval) in the former pair of conditions as opposed to solely one feature (tone duration) in the latter pair [1] , [47] .” | limitations[10] |
| for which use | limitation — every starred pairwise comparison in the paper: * P <0.05 AND ** P <0.01 IN POST-HOC UNCORRECTED COMPARISONS (PAIRED T-TESTS) — “* P <0.05 and ** P <0.01 in post-hoc uncorrected comparisons (paired t-tests).” | limitations[11] |
| replication | contested | |
| limitation | The paper cannot explain why mismatch responses were absent in two of the four timing conditions. — “Therefore, why mismatch responses were absent precisely in the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions remains to be accounted for.” | limitations[0] |
| limitation | Deviant rarity as such, rather than change detection, remains a viable alternative explanation of the responses. — “Nevertheless, the rarity of the deviant tone as such [26] , [27] also remains a possible explanation for our findings.” | limitations[1] |
| limitation | The design lacked deviant-alone and equal-probability control conditions needed to test the change-detection account. — “Future studies with appropriate control procedures, such as deviant-alone (e.g. [18] ) or equal-probability (e.g. [19] ) stimulus conditions, are needed to test the validity of the change-detection account.” | limitations[2] |
| limitation | Whether the cortical and hippocampal responses share one mechanism, or reflect separate ones, is unresolved. — “It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms.” | limitations[3] |
| limitation | three of the four stimulus conditions produced nothing — “In the remaining three conditions, the findings were negative.” | limitations[4] |
| limitation | why the two matched-interval conditions failed is left unexplained, and the explanations offered are labelled tentative — “Therefore, why mismatch responses were absent precisely in the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions remains to be accounted for.” | limitations[5] |
| limitation | the change-detection interpretation is not established by this design — mere rarity remains available — “Nevertheless, the rarity of the deviant tone as such [26] , [27] also remains a possible explanation for our findings.” | limitations[6] |
| limitation | the controls that would settle it were not run — “Future studies with appropriate control procedures, such as deviant-alone (e.g. [18] ) or equal-probability (e.g. [19] ) stimulus conditions, are needed to test the validity of the change-detection account.” | limitations[7] |
| limitation | whether the cortical and hippocampal responses are one mechanism, two independent ones, or two connected ones is undetermined — “It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms.” | limitations[8] |
| limitation | the conditions differ in stimulus assignment and in sweep count as well as in interval, so the comparison across them is not clean — “In the SSI 200 /SDI 200 - and SSI 500 /SDI 500 -conditions, the stimulus assignments (the 25-ms tone as the deviant tone and the 75-ms tone as the standard tone, but not vice versa) and numbers (a half) of sweeps per averaged waveform were different from those in the SSI 500 /SDI 200 - and SSI 200 /SDI 500 -conditions.” | limitations[9] |
| limitation | the deviant differed from the standard on two features rather than one in the conditions that worked, which is itself a candidate explanation of the result — “The present pattern of observations may also reflect the fact that the deviant tone was distinguishable from the standard tone by two auditory features (tone duration and inter-stimulus interval) in the former pair of conditions as opposed to solely one feature (tone duration) in the latter pair [1] , [47] .” | limitations[10] |
| limitation | the pairwise comparisons carry no correction for multiple testing, and this is stated in the figure legend — “* P <0.05 and ** P <0.01 in post-hoc uncorrected comparisons (paired t-tests).” | limitations[11] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | Mismatch responses to duration deviants were recorded not only in the auditory cortex but simultaneously in all three hippocampal-system sites (CA1, dentate gyrus and subiculum) of urethane-anesthetized rats. — “We found differential LFPs (mismatch responses) to duration deviant tones in the auditory cortex and in the hippocampal system (CA1, dentate gyrus, and subiculum) in urethane-anesthetized rats.” | content_tags.findings[0] |
| finding | The cortical and hippocampal responses had opposite polarity (positive in cortex, negative in hippocampus) yet occupied the same 51.5-89 ms post-change latency window and appeared only in the condition combining a 500-ms standard-to-standard interval with a 200-ms standard-to-deviant interval. — “The responses - positive polarity in the auditory cortical and negative polarity in the hippocampal sites - appeared in the same latency range (between 51.5 and 89 ms post-change) and in all the recording sites, but only in the SSI 500 /SDI 200 -condition.” | content_tags.findings[1] |
| finding | Of the four interval combinations tested, only the 500-ms SSI paired with the 200-ms SDI elicited mismatch responses; the other three combinations elicited none. — “Mismatch responses were observed at 51.5–89 ms with the 500-ms SSI coupled with the 200-ms SDI but not with the three remaining combinations.” | content_tags.findings[2] |
| finding | Hippocampal and auditory cortical mismatch responses occurred in the same latency range and the same stimulus condition. — “Most importantly, hippocampal and auditory cortical mismatch responses were found in the same latency range and in the same stimulus condition.” | content_tags.findings[3] |
| finding | The authors conclude that the hippocampus may play a role in the cortical manifestation of the mismatch negativity. — “The findings suggest that the hippocampus may play a role in (cortical) manifestation of MMN.” | content_tags.findings[4] |
| finding | mismatch responses appeared in auditory cortex and all three hippocampal sites, at 51.5–89 ms post-change, but only in one of four conditions — “Mismatch responses were observed at 51.5–89 ms with the 500-ms SSI coupled with the 200-ms SDI but not with the three remaining combinations.” | content_tags.findings[5] |
| finding | the cortical response was positive-going and the hippocampal responses negative-going, in the same animals and the same window — “The responses - positive polarity in the auditory cortical and negative polarity in the hippocampal sites - appeared in the same latency range (between 51.5 and 89 ms post-change) and in all the recording sites, but only in the SSI 500 /SDI 200 -condition.” | content_tags.findings[6] |
| finding | in auditory cortex the effect was a stimulus by time-segment interaction, present in the first segment only — “as reflected by a stimulus (deviant tone, standard tone)×time-segment (first, second, third) -interaction, in the SSI 500/ SDI 200 condition, F(2,18) = 5.13, P = 0.048” | content_tags.findings[7] |
| finding | in this condition the displacement was present in the first and the second time segments — “Post-hoc pairwise comparisons further indicated that, in this condition, the displacement was present (P<0.05) in the first time segment (51.5–89 ms from duration change, P<0.001) and in the second time segment (89.5–127 ms from duration change, P<0.05).” | content_tags.findings[8] |
| finding | the subiculum, like CA1 and dentate gyrus, was displaced towards negative polarity by the deviant tone — “Furthermore, as in the CA1 and DG, the LFPs in the subiculum were displaced towards negative polarity in response to the deviant tone relative to the standard tone, as reflected by a significant stimulus×time-segment -interaction, F(2,16) = 8.18, P = 0.017, in the SSI 500/ SDI 200 condition.” | content_tags.findings[9] |
| finding | no main effect of stimulus in auditory cortex in any condition — “No main effect of stimulus was present in any condition (P≥0.211).” | content_tags.findings[10] |
| finding | in CA1 the tones evoked two negative peaks at about 20 and 40 ms, timed from tone onset rather than from deviance — “In the CA1 area of the hippocampus, the tones evoked a two-peaked negative deflection, the first peak occurring at about 20 ms and the second at about 40 ms from stimulus onset (i.e., not from deviance which occurred at 25 ms after stimulus onset).” | content_tags.findings[11] |
| finding | in dentate gyrus one further condition gave a significant interaction that post-hoc testing reduced to a trend in the last segment — “However, post-hoc comparisons only showed a trend for the deviant-standard difference in the last (127.5 ms −165 ms) time segment (P = 0.09).” | content_tags.findings[12] |
| finding | usable electrodes differed by structure: 9 animals for subiculum, 8 for CA1, 6 for dentate gyrus — “For the hippocampal recordings, the electrodes successfully implanted in the targeted areas were applied in the analysis (9 animals for the subiculum, 8 for the CA1 area, and 6 for the dentate gyrus, Figure 1 )” | content_tags.findings[13] |
| finding | sweep retention was high — never below 67 of 100, averaging 98.7 — “Not less than 67 out of 100 sweeps were included in calculating the average values per stimulus type and condition in any animal, the average number being 98.7 out of 100 sweeps.” | content_tags.findings[14] |
| quantity | 51.5–89 ms — latency window after the duration change in which mismatch responses appeared at every recording site — “Mismatch responses were observed at 51.5–89 ms with the 500-ms SSI coupled with the 200-ms SDI but not with the three remaining combinations.” | content_tags.quantities[0] |
| quantity | Ten — number of rats used, weighing 305-375 g — “Ten Spraque Dawley rats were used in the experiment (weight 305–375 g).” | content_tags.quantities[1] |
| quantity | 1000 — number of tones in each stimulus series, with consecutive deviants separated by at least two standards — “occurred in a series of 1000 tones randomly with the restriction that consecutive deviants were separated by at least two standards.” | content_tags.quantities[2] |
| quantity | 25 vs. 75 ms — the duration difference between deviant and standard tone that constituted the deviance — “The deviant tone differed from the standard tone in duration (25 vs. 75 ms).” | content_tags.quantities[3] |
| quantity | 4000-Hz — frequency of the sinusoidal tones, held identical for deviant and standard — “Sinusoidal 4000-Hz tones of 25 ms and 75 ms in duration, including 5-ms rise and fall times, were used as stimuli.” | content_tags.quantities[4] |
| quantity | 98.7 out of 100 — average number of artifact-free sweeps entering each averaged waveform — “Not less than 67 out of 100 sweeps were included in calculating the average values per stimulus type and condition in any animal, the average number being 98.7 out of 100 sweeps.” | content_tags.quantities[5] |
| quantity | 9 ... 8 ... 6 — number of animals with usable electrodes in subiculum, CA1 and dentate gyrus respectively — “For the hippocampal recordings, the electrodes successfully implanted in the targeted areas were applied in the analysis (9 animals for the subiculum, 8 for the CA1 area, and 6 for the dentate gyrus, Figure 1 ).” | content_tags.quantities[6] |
| quantity | 51.5–89 ms — the post-change window in which mismatch responses appeared at every site — “Mismatch responses were observed at 51.5–89 ms with the 500-ms SSI coupled with the 200-ms SDI but not with the three remaining combinations.” | content_tags.quantities[7] |
| quantity | 25 vs 75 ms — the two tone durations serving as deviant and standard — “The deviant tone differed from the standard tone in duration (25 vs. 75 ms).” | content_tags.quantities[8] |
| quantity | 200 ms vs. 500 ms — the two interval values crossed to make the four conditions — “Two standard-to-standard (SSI) and standard-to-deviant (SDI) intervals (200 ms vs. 500 ms) were applied in different combinations to vary the observability of responses resembling MMN (mismatch responses).” | content_tags.quantities[9] |
| quantity | P = 0.1 and P = 0.9 — deviant and standard probabilities in a series of 1000 tones — “The deviant (P = 0.1) and the standard (P = 0.9) occurred in a series of 1000 tones randomly with the restriction that consecutive deviants were separated by at least two standards.” | content_tags.quantities[10] |
| quantity | ten rats — the sample, 305–375 g — “Ten Spraque Dawley rats were used in the experiment (weight 305–375 g).” | content_tags.quantities[11] |
| quantity | 1.2 g/kg — urethane dose, topped up by pedal withdrawal reflex — “The animals were anaesthetized with urethane (Sigma-Aldrich, St. Louis, MO, USA) i.p. (1.2 g/kg).” | content_tags.quantities[12] |
| quantity | 4000-Hz, 70 dB — tone frequency and sound pressure level — “Sinusoidal 4000-Hz tones of 25 ms and 75 ms in duration, including 5-ms rise and fall times, were used as stimuli.” | content_tags.quantities[13] |
| quantity | 37.5-ms segments — the analysis window width, three of five used, starting 50 ms from change — “For the response amplitude analysis, the response time of 187.5 ms, as calculated from tone onset, was divided into five 37.5-ms segments.” | content_tags.quantities[14] |
| quantity | 150–250 ms — the human MMN latency range — “MMN can be elicited in the latency range of 150–250 ms even by changes perceptually too weak to attract involuntary attention [1] .” | content_tags.quantities[15] |
| quantity | 250–280 ms — the human P3a latency range — “If the auditory changes are of high magnitude, they can capture involuntary attention [4] and elicit a partly hippocampally-contributed P3a of ERPs in the later latency range of 250–280 ms [1] , [5] , [6] .” | content_tags.quantities[16] |
| quantity | 30–150 ms — the reported latency range of cortical mismatch responses in rats — “In awake and anesthetized rats, cortical mismatch responses have been found to occur in the latency range of 30–150 ms [18] , [19] , [21] – [23] (for 25-ms onset latency, see [20] ).” | content_tags.quantities[17] |
| quantity | about 240 ms — when P3a-type responses follow mismatch responses in awake rats — “these responses follow mismatch responses at about 240 ms post-stimulus [39] .” | content_tags.quantities[18] |
| quantity | 98.7 out of 100 — average sweeps retained per average — “the average number being 98.7 out of 100 sweeps” | content_tags.quantities[19] |
| quantity | 0.1–30 Hz — the offline filter band applied before segmentation — “The data were then offline-filtered (0.1–30 Hz, 24 dB/octave, Butterworth Zero Phase filters)” | content_tags.quantities[20] |
| quantity | 35 µm — coronal section thickness for histological verification — “Coronal sections (thickness 35 µm) were cut with a freezing slide microtome.” | content_tags.quantities[21] |
| variable | SDI (independent) — “Two standard-to-standard (SSI) and standard-to-deviant (SDI) intervals (200 ms vs. 500 ms) were applied in different combinations to vary the observability of responses resembling MMN (mismatch responses).” | content_tags.variables[1] |
| variable | sound frequency (control) — “Sound frequency was not used due to its spatial coding from the cochlea [41] owing to a potential confounding effect of lower levels of refractoriness in afferent pathways activated by the rare deviant tone than the frequently occurring standard tones [1] .” | content_tags.variables[3] |
| variable | stimulus (standard, deviant) — “ANOVA for repeated measures with stimulus (standard, deviant) and time-segment (first, second, third) as factors was used to justify post-hoc deviant-standard LFP amplitude comparisons for each time segment separately” | content_tags.variables[5] |
| variable | time-segment — “the mean amplitudes of three consecutive post-stimulus 37.5-ms time segments were calculated from 51.5 ms post-change (76.5 ms post-stimulus) onwards” | content_tags.variables[6] |
| variable | SSI — “Otherwise, the four conditions differed in stimulus-onset-asynchrony (200 ms or 500 ms) between consecutive standards (SSI) and between deviants and their immediately preceding standards (SDI).” | content_tags.variables[7] |
| variable | response amplitude — “Significant differences in response amplitude between the deviant and the standard are marked between the curves.” | content_tags.variables[8] |
| variable | brain location — “We averaged the LFPs for each stimulus type (the deviant tone and the standard tone which immediately preceded the deviant tone) in each brain location, stimulus condition ( Figure 2 ), and animal.” | content_tags.variables[9] |
| definition | mismatch negativity (MMN) — “In humans, the preattentive detection of auditory changes is indexed by an electrical brain response with a fronto-central scalp distribution termed the mismatch negativity (MMN) of event-related potentials (ERPs)” | content_tags.definitions[0] |
| definition | mismatch responses — “Mismatch responses (i.e., higher-amplitude brain responses to deviant tones than standard tones irrespectively of the polarity of these responses and despite these tones are only passively listened to) have been found in sleeping infants” | content_tags.definitions[1] |
| definition | oddball condition — “MMN is typically experimentally addressed by using the so-called oddball condition. In this condition, a (‘deviant’) tone rarely replaces a frequently repeated (‘standard’) tone.” | content_tags.definitions[2] |
| definition | stimulus-specific adaptation — “Stimulus-specific adaptation, a decrease in neural responses to a repeated tone which is specific to the physical features of this tone, observed in different stages of the auditory pathway” | content_tags.definitions[3] |
| definition | Mismatch responses — “Mismatch responses (i.e., higher-amplitude brain responses to deviant tones than standard tones irrespectively of the polarity of these responses and despite these tones are only passively listened to)” | content_tags.definitions[4] |
| definition | oddball condition — “MMN is typically experimentally addressed by using the so-called oddball condition. In this condition, a (‘deviant’) tone rarely replaces a frequently repeated (‘standard’) tone.” | content_tags.definitions[5] |
| definition | MMN — “MMN is quantified by arithmetically subtracting ERPs to the standard tone from ERPs to the deviant tone.” | content_tags.definitions[6] |
| definition | Stimulus-specific adaptation — “Stimulus-specific adaptation, a decrease in neural responses to a repeated tone which is specific to the physical features of this tone” | content_tags.definitions[7] |
| definition | P3a — “The hippocampus is associated with a later generated P3a of ERPs reflecting involuntarily attention switches towards auditory changes that are high in magnitude.” | content_tags.definitions[8] |
| key-terms | mismatch negativity, mismatch responses, P3a, event-related potentials, local-field potentials, oddball condition, deviant, standard, duration deviants, SSI, SDI, auditory cortex, CA1, dentate gyrus, subiculum, hippocampal system, refractoriness, stimulus-specific adaptation, memory trace, urethane, preattentive, involuntary attention, deviant-alone, equal-probability, stimulus-onset-asynchrony, cresyl violet, Greenhouse-Geisser |
| ANOVA F | as written: ANOVA for repeated measures — “As there were no suitable non-parametric alternatives, ANOVA for repeated measures with stimulus (standard, deviant) and time-segment (first, second, third) as factors was used to justify post-hoc deviant-standard LFP amplitude comparisons for each time segment separately.” | content_tags.statistics[0] |
| t-test | as written: paired t test (two-tailed) — “Pairwise deviant-standard comparison for each time segment was made using the paired t test (two-tailed).” | content_tags.statistics[1] |
| Greenhouse-Geisser correction | as written: Greenhouse-Geisser corrected — “Degrees of freedom were Greenhouse-Geisser corrected whenever the sphericity assumption was violated, and corrected P values were reported.” | content_tags.statistics[2] |
| significance level alpha | as written: An alpha level of 0.05 — “An alpha level of 0.05 was used in all analyses.” | content_tags.statistics[3] |
| p-value P | as written: (P<0.05) — “Post-hoc pairwise comparisons further indicated that this displacement was present (P<0.05) in the first time-segment (51.5–89 ms from duration change).” | content_tags.statistics[4] |
| ANOVA for repeated measures | — “ANOVA for repeated measures with stimulus (standard, deviant) and time-segment (first, second, third) as factors was used to justify post-hoc deviant-standard LFP amplitude comparisons for each time segment separately” | content_tags.statistics[5] |
| Greenhouse-Geisser | — “Degrees of freedom were Greenhouse-Geisser corrected whenever the sphericity assumption was violated, and corrected P values were reported.” | content_tags.statistics[6] |
| paired t test (two-tailed) | — “Pairwise deviant-standard comparison for each time segment was made using the paired t test (two-tailed).” | content_tags.statistics[7] |
| alpha level of 0.05 | — “An alpha level of 0.05 was used in all analyses.” | content_tags.statistics[8] |
| post-hoc uncorrected comparisons | — “* P <0.05 and ** P <0.01 in post-hoc uncorrected comparisons (paired t-tests).” | content_tags.statistics[9] |
| stimulus×time-segment -interaction | — “as reflected by a significant stimulus×time-segment -interaction, F(2,14) = 11.34, P = 0.010” | content_tags.statistics[10] |
| method | local field potentials from primary auditory cortex and three hippocampal sites in urethane-anesthetized rats, across four oddball conditions crossing two standard-to-standard and two standard-to-deviant intervals — a second reading DIFFERING from the recorded value; recorded: electrophysiology — “The recordings were made during four oddball conditions with different stimulus-onset asynchronies between consecutive standard tones and between deviants tones and standard tones immediately preceding them adapted from [40] ( Figure 2 ).” | alternatives.method |
| phenomenon | automatic auditory change detection — the brain's response to a rare deviant among repeated standards — a second reading DIFFERING from the recorded value; recorded: auditory mismatch responses to duration deviants — “The human brain preattentively or automatically detects such changes.” | alternatives.phenomenon |
Full text — tagged
The reference list is not part of the text we hold for this paper. Our copy was extracted from the publisher's page and the bibliography did not survive that extraction, so the article below ends before its references. Nothing has been removed by us.
Abstract Any change in the invariant aspects of the auditory environment is of potential importance. The human brain preattentively or automatically detects such changes. The mismatch negativity (MMN)term·def of event-related potentialsterm (ERPs) reflects this initial stage of auditory change detection. The origin of MMNterm·def is held to be cortical. The hippocampus is associated with a later generated P3aterm of ERPs reflecting involuntarily attention switches towards auditory changes that are high in magnitude. The evidence for this cortico-hippocampal dichotomyterm is scarce, however. To shed further light on this issue, auditory cortical and hippocampal-system (CA1term, dentate gyrusterm, subiculumterm) local-field potentialsterm were recorded in urethane-anesthetized rats. A rare tone in duration (deviantterm) was interspersed with a repeated tone (standardterm). Two standard-to-standard (SSIterm) and standard-to-deviant (SDIterm) intervals (200 ms vs. 500 ms) were applied in different combinations to vary the observability of responses resembling MMNterm·def (mismatch responsesterm). Mismatch responsesterm were observed at 51.5–89 ms with the 500-ms SSIterm coupled with the 200-ms SDIterm but not with the three remaining combinations. Most importantly, the responses appeared in both the auditory-cortical and hippocampal locations. The findings suggest that the hippocampus may play a role in (cortical) manifestation of MMNterm·def.
Introduction Any changes in the invariant attributes of the auditory environment are of potential importance for survival. The rapid, effortless, and sensitive detection of auditory changes is, therefore, of crucial importance for the organism.
In humans, the preattentiveterm detection of auditory changes is indexed by an electrical brain response with a fronto-central scalp distribution termed the mismatch negativity (MMN)term·def of event-related potentialsterm (ERPs) [1] – [3] . MMNterm·def can be elicited in the latency range of 150–250 ms even by changes perceptually too weak to attract involuntary attentionterm [1] . If the auditory changes are of high magnitude, they can capture involuntary attentionterm [4] and elicit a partly hippocampally-contributed P3aterm of ERPs in the later latency range of 250–280 ms [1] , [5] , [6] .
MMNterm·def is typically experimentally addressed by using the so-called oddball conditionterm. In this condition, a (‘deviantterm’) tone rarely replaces a frequently repeated (‘standardterm’) tone. MMNterm·def is elicited by the deviantterm tone only if the distinguishing feature(s) of the standardterm relative to the deviantterm are successfully represented neurally, stored in transient auditory memory, and compared to the auditory input by the deviantterm [1] , [7] , [8] . MMNterm·def is quantified by arithmetically subtracting ERPs to the standardterm tone from ERPs to the deviantterm tone. The specificity of MMNterm·def to deviantterm tones as changes in standardterm tones is reflected by the disappearance of these responses by the removal of standardterm tones from the series [3] .
Mismatch responsesterm (i.e., higher-amplitude brain responses to deviantterm tones than standardterm tones irrespectively of the polarity of these responses and despite these tones are only passively listened to) have been found in sleeping infants [9] – [11] and in awake, sleeping and anesthetized animals [12] – [23] (for negative findings in anesthetized rats, see, [24] , [25] ).
The generators of MMNterm·def are largely unknown. Different and partially contradictory explanations have been proposed [8] , [26] – [28] . The general consensus points to the generation of MMNterm·def in the temporal and, subsequently, frontal cortical areas [29] , [30] with a possible contribution from lower levels of the auditory pathway [31] , [32] .
No role in human MMNterm·def generation has been assigned to the hippocampus. Studies employing intracranial hippocampal recordings in neurological patients have reported no mismatch responsesterm [33] or other, longer-latency differential responses to deviantterm tones as novel or salient stimuli rather than as changes in standardsterm [34] .
The MMN-hippocampus dissociation may, however, be premature for a number of reasons. First, too few hippocampal MMNterm·def data have been gathered in humans. Second, the hippocampus has been proposed, comparably to MMNterm·def [1] , to respond by itself to rare stimuli on the basis of their discrepancy from a neural model of past repetitive stimuli [35] – [38] . Finally, hippocampal mismatch responsesterm have been observed in animals [14] , [16] , and some of these responses have, comparably to MMNterm·def [3] , even shown specificity to deviantterm tones as changes in standardterm tones [16] .
Nevertheless, hippocampal mismatch responsesterm could be claimed to be secondary to and solely triggered by auditory cortical mismatch responsesterm in a limited set of circumstances. Namely, these responses could be argued to only emerge to high-magnitude deviantterm tones that capture involuntary attentionterm in contrast to auditory cortical mismatch responsesterm also observable to low-magnitude deviantterm tones close to their behavioral detection threshold. This would be analogous to a MMN-P3a sequence in humans that is only observable in full (i.e., also including P3aterm) when deviantterm tones of high magnitude attract involuntary switches of attention [1] , [4] .
In awake and anesthetized rats, cortical mismatch responsesterm have been found to occur in the latency range of 30–150 ms [18] , [19] , [21] – [23] (for 25-ms onset latency, see [20] ). In urethaneterm anesthetized rats, auditory cortical mismatch responsesterm have, comparable to MMNterm·def [1] , been observed to be specific to deviantterm tones as changes per se in standardterm tones. That is, mismatch responsesterm in these animals have been found to disappear when standardterm tones have either been completely removed from the series [18] or replaced by a set of heterogeneous tones of equal rarity to deviantterm tones [19] . Awake rats also show P3a-type responses to deviantterm tones of high magnitude. Similarly to P3aterm in humans [4] , [6] , these responses follow mismatch responsesterm at about 240 ms post-stimulus [39] .
In the present study, local-field potentialsterm (LFPs) in urethane-anesthetized rats were recorded on the surface of the primary auditory cortexterm and in three locations of the hippocampal systemterm (dentate gyrusterm, CA1term, and subiculumterm, Figure 1 ). The recordings were made during four oddball conditionsterm with different stimulus-onset asynchronies between consecutive standardterm tones and between deviantsterm tones and standardterm tones immediately preceding them adapted from [40] ( Figure 2 ). The deviantterm tone differed from the standardterm tone in duration (25 vs. 75 ms). Sound frequencyterm·var was not used due to its spatial coding from the cochlea [41] owing to a potential confounding effect of lower levels of refractorinessterm in afferent pathways activated by the rare deviantterm tone than the frequently occurring standardterm tones [1] . Early neuronal selectivity to duration in the auditory system (the auditory midbrain) has been demonstrated but through active rather than fatigue-type of passive neural events [42] . We expected that some of the auditory oddball conditionsterm would be more favorable for mismatch responsesterm than others. Our interest was whether mismatch responsesterm, whenever elicited, occupied both the auditory cortexterm and the hippocampus in a typical latency range of mismatch responsesterm in urethaneterm anesthetized rats or whether mismatch responsesterm were only manifested in the auditory cortexterm and possibly only followed by later hippocampal activity.
Figure 1. Electrode locations.
The representative histological cresyl violetterm·key stained sections and a drawing of a coronal section of a rat brain 6.48 mm from bregma illustrate the locations of electrodes in the CA1term radiatum layer (Rad), in the molecular layer of the dentate gyrusterm (MoDG) and in the ventral subiculumterm (VS). Adapted from [54] .
Figure 2. Experimental stimuli. In each of the four stimulus conditions, a duration deviantterm tone (D) was interspersed with a repeated standardterm tone (S).
In the SSIterm 200 /SDIterm 200 and SSIterm 500 /SDIterm 500 -conditions, the duration of the deviantterm was 25 ms of the standardterm 75 ms. In the SSIterm 200 /SDIterm 500 and SSIterm 500 /SDIterm 200 -conditions, each duration was assigned to each stimulus type in separate series. Otherwise, the four conditions differed in stimulus-onset-asynchronyterm·key (200 ms or 500 ms) between consecutive standardsterm (SSIterm) and between deviantsterm and their immediately preceding standardsterm (SDIterm).
Results We averaged the LFPs for each stimulus type (the deviantterm tone and the standardterm tone which immediately preceded the deviantterm tone) in each brain locationterm·var, stimulus condition ( Figure 2 ), and animal. From these averaged LFP waveforms, the mean amplitudes of three consecutive post-stimulus 37.5-ms time segments were calculated from 51.5 ms post-change (76.5 ms post-stimulus) onwards (see also [18] , [19] ).
Auditory Cortexterm In the auditory cortexterm, both stimulus types in all conditions could be seen to evoke a large positive deflection peaking at about 40 ms from stimulus onset (i.e., not from duration deviance). This deflection gradually sloped back to the baseline level, or below, by about 100 ms from stimulus onset ( Figure 3 ).
Figure 3. Electrophysiological results.
LFPs in response to different stimulus types in different stimulus conditions. The zero time point is at the moment of change, that is, at 25 ms from tone onset and, thus, at the offset of the 25-ms tone. The two tones used as the deviantterm and the standardterm are illustrated above the x-axis of the panel for the auditory cortexterm. The three consecutive time segments for calculating average response amplitudesterm·var are illustrated in each panel. Significant differences in response amplitudeterm·var between the deviantterm and the standardterm are marked between the curves. * P <0.05 and ** P <0.01 in post-hoc uncorrected comparisonsterm·stat (paired t-teststerm·stat).
The LFPs were displaced towards positive polarity in response to the deviantterm tone relative to the standardterm tone ( Figure 3 ), as reflected by a stimulus (deviantterm tone, standardterm tone)×time-segmentterm·var (first, second, third) -interaction, in the SSIterm 500/ SDIterm 200 condition, F(2,18) = 5.13, P = 0.048. Post-hoc pairwise comparisons further indicated that this displacement was present (P<0.05) in the first time-segmentterm·var (51.5–89 ms from duration change). In the other conditions, no significant stimulus×time-segmentterm·var interactions were observed (P≥0.532).
No main effect of stimulus was present in any condition (P≥0.211).
CA1term Area of the Hippocampus In the CA1term area of the hippocampus, the tones evoked a two-peaked negative deflection, the first peak occurring at about 20 ms and the second at about 40 ms from stimulus onset (i.e., not from deviance which occurred at 25 ms after stimulus onset). This deflection had returned to or exceeded the baseline level by the end of the response time ( Figure 3 ).
The LFPs were displaced towards negative polarity in response to the deviantterm tone relative to the standardterm tone in the SSIterm 500/ SDIterm 200 condition ( Figure 3 ) as reflected by a significant stimulus×time-segment -interactionterm·stat, F(2,14) = 11.34, P = 0.010. Post-hoc pairwise comparisons further indicated that, in this condition, the displacement was present (P<0.05) in the first time segmentterm·var (51.5–89 ms from duration change, P<0.001) and in the second time segmentterm·var (89.5–127 ms from duration change, P<0.05). In the other conditions, no significant stimulus×time-segmentterm·var interactions (P≥0.202) were observed.
For the SSIterm 500/ SDIterm 200 condition, a main effect of stimulus, F(2,14) = 5.870, P = 0.046 and for the SSIterm 200/ SDIterm 200 condition, a trend towards this effect, F(2,14) = 3.874, P = 0.090, were also observed. This effect was not present (P≥0.440) in the remaining two conditions.
DG of the Hippocampus In the DG area of the hippocampus, the general waveform of the LFPs to the tones themselves closely resembled those observed for the CA1term area. Again, in the averaged waveform, a negative deflection with slight peaks at about 20 ms and 40 ms from stimulus onset (that is, not from duration change) and a gradual return to the baseline level by the end of the response time were observed ( Figure 3 ).
Furthermore, just as in the CA1term, the LFPs in the DG area were displaced towards negative polarity in response to the deviantterm tone relative to the standardterm tone in the SSIterm 500/ SDIterm 200 condition ( Figure 3 ), as reflected by a significant stimulus×time-segment -interactionterm·stat in the SSIterm 500/ SDIterm 200 condition, F(2,10) = 7.50, P = 0.030. Post-hoc pairwise comparisons further indicated that in this condition the displacement was present (P<0.05) in the first (51.5–89 ms from change) time segmentterm·var ( Figure 3 ).
For the SSIterm 200/ SDIterm 500 condition, there was a significant stimulus×time-segmentterm·var interaction, F(2,10) = 11.362, P = 0.015. However, post-hoc comparisons only showed a trend for the deviant-standard difference in the last (127.5 ms −165 ms) time segmentterm·var (P = 0.09). For the SSIterm 200/ SDIterm 200 condition, a trend towards such an interaction (P = 0.085) was present, and for the SSIterm 500/ SDIterm 500 condition, no such interaction was observed (P = 0.772).
No main effect of stimulus was present in any condition (P≥0.279).
Subiculumterm In the subiculumterm, the general waveform of the LFPs to the tones themselves was similar to those observed for the CA1term and DG areas. Again, in the averaged waveform, a negative deflection with slight peaks at about 20 and 40 ms from stimulus onset (that is, not from duration change) had returned to the baseline level by the end of the response time ( Figure 3 ).
Furthermore, as in the CA1term and DG, the LFPs in the subiculumterm were displaced towards negative polarity in response to the deviantterm tone relative to the standardterm tone, as reflected by a significant stimulus×time-segment -interactionterm·stat, F(2,16) = 8.18, P = 0.017, in the SSIterm 500/ SDIterm 200 condition. Post-hoc pairwise comparisons further indicated that in this condition, the displacement was present (P<0.05) in the first (51.5–89 ms from change) time segmentterm·var ( Figure 3 ). No significant stimulus×time-segmentterm·var interaction was observed in other conditions (P≥0.264).
There was only a trend for the main effect of stimulus (P = 0.083) in the SSIterm 500/ SDIterm 200 condition.
Discussion We found differential LFPs (mismatch responsesterm) to duration deviantterm tones in the auditory cortexterm and in the hippocampal systemterm (CA1term, dentate gyrusterm, and subiculumterm) in urethane-anesthetized rats. The responses - positive polarity in the auditory cortical and negative polarity in the hippocampal sites - appeared in the same latency range (between 51.5 and 89 ms post-change) and in all the recording sites, but only in the SSIterm 500 /SDIterm 200 -condition. In the remaining three conditions, the findings were negative.
The polarity and latency range of auditory cortical mismatch responsesterm in the SSIterm 500 /SDIterm 200 -condition correspond to previous findings in urethaneterm anesthetized rats [18] , [19] , [22] , [23] . There are three possible reasons for the absence of mismatch responsesterm in the SSIterm 200 /SDIterm 500 -condition. First, the short inter-deviant interval (due to the 200-ms SSIterm) might have diminished mismatch responsesterm by the fast-paced activation of their neural generator (in humans [43] – [45] , in cats [15] ). Secondly, the long (500-ms) SDIterm might have led to a high degree of degradation of the memory traceterm of the standardterm tone, preventing the detection of the deviantterm tone as not matching this trace [1] , [7] . Third, the long SDIterm relative to the SSIterm might have been a less efficient additional temporal cue (inter-stimulus interval) for the detection of the (stimulus duration) deviantterm tone than the short SDIterm relative to the SSIterm [46] .
In the SSIterm 200 /SDIterm 200 - and SSIterm 500 /SDIterm 500 -conditions, the stimulus assignments (the 25-ms tone as the deviantterm tone and the 75-ms tone as the standardterm tone, but not vice versa) and numbers (a half) of sweeps per averaged waveform were different from those in the SSIterm 500 /SDIterm 200 - and SSIterm 200 /SDIterm 500 -conditions. Therefore, why mismatch responsesterm were absent precisely in the SSIterm 200 /SDIterm 200 - and SSIterm 500 /SDIterm 500 -conditions remains to be accounted for. However, some tentative explanations can be offered.
In the SSIterm 200 /SDIterm 200 -condition, the short inter-deviant interval (due to the 200-ms SSIterm) might have diminished the amplitude of mismatch responsesterm through the fast-paced activation of their neural generator [15] , [43] – [45] while the short decay time of the memory traceterm of the standardterm tone to be compared to the sensory input by the deviantterm tone (due to the 200-ms SDIterm, [1] , [7] ) might not have been sufficient to compensate for this diminution. In the SSIterm 500 /SDIterm 500 -condition, the decay time (the 500-ms SDIterm) might, in turn, have been too long as mismatch responsesterm were observed in the SSIterm 500 /SDIterm 200 -condition (with the same SSIterm but a shorter SDIterm).
Mismatch responsesterm in rats have been found to be higher in amplitude to deviantterm tones as lengthenings than shortenings of standardterm tones [21] . Therefore, mismatch responsesterm in the SSI500/SDI200-condition and a trend for a late differential response in the DG area in the SSI200/SDI500-condition may have reflected the fact that these conditions also involved the deviantterm tone as a lengthening of the standardterm tone as opposed to the SSI200/SDI200- and SSI500/SDI500-conditions which only included the deviantterm tone as a shortening of the standardterm tone. The present pattern of observations may also reflect the fact that the deviantterm tone was distinguishable from the standardterm tone by two auditory features (tone duration and inter-stimulus interval) in the former pair of conditions as opposed to solely one feature (tone duration) in the latter pair [1] , [47] .
Most importantly, hippocampal and auditory cortical mismatch responsesterm were found in the same latency range and in the same stimulus condition. This correspondence suggests that the hippocampal responses reflect the initial detection of deviantterm tones similarly to auditory cortical mismatch responsesterm. It remains to be resolved whether these responses of different anatomical origins recruit a single but spatially distributed neural mechanism, spatially separate and independent neural mechanisms, or spatially separate but interconnected neural mechanisms. The nature of the mechanism(s) should also be further explored. Namely, the mechanisms could, similarly to those for MMNterm·def in humans [1] , [8] , have treated the deviantterm tone as a change in the standardterm tone. Auditory cortical mismatch responsesterm in urethane-anesthetized rats [18] , [19] and hippocampal mismatch responsesterm in awake rabbits [16] to frequency deviantterm tones have been found to reflect such mechanisms. Nevertheless, the rarity of the deviantterm tone as such [26] , [27] also remains a possible explanation for our findings. Future studies with appropriate control procedures, such as deviant-aloneterm (e.g. [18] ) or equal-probabilityterm (e.g. [19] ) stimulus conditions, are needed to test the validity of the change-detection account. There is also need for deeper understanding of the registration and transient memory storage of repetitive standardterm tones at the neural level [1] , [7] . Stimulus-specific adaptationterm, a decrease in neural responses to a repeated tone which is specific to the physical features of this tone, observed in different stages of the auditory pathway [48] – [51] as functionally distinct from auditory cortical mismatch responsesterm [52] , [53] could provide a useful neurophysiological tool to increase this understanding.
To summarize, the brains of urethane-anesthetized rats were found to generate mismatch responsesterm to duration deviantsterm with the same set of stimulus parameters and in the same latency range in the auditory cortexterm and in three locations in the hippocampal systemterm (CA1term, dentate gyrusterm, subiculumterm). The findings prompt the question whether the hippocampal systemterm plays an active role in the auditory cortical manifestation of these responses and hence also possibly of MMNterm·def in humans.
Materials and Methods Ethics Statement The experiments were approved by the Finnish National Animal Experiment Board (Permit code: ESLH-2007-00662), and carried out in accordance with the European Communities Council Directive (86/609/EEC) regarding the care and use of animals for experimental procedures.
Animals and Surgery Ten Spraque Dawley rats were used in the experiment (weight 305–375 g). The animals were housed as groups in cages with water and feed ad libitum. The animals were anaesthetized with urethaneterm (Sigma-Aldrich, St. Louis, MO, USA) i.p. (1.2 g/kg). The level of anesthesia was controlled by pedal withdrawal reflex, and if necessary, extra doses of urethaneterm were given. The animal was positioned in a streotaxic instrument with blunt ear bars which afterwards were removed to allow auditory stimulation. Under lidocaine anesthesia (Lidocain 20%, Orion Pharma, Espoo, Finland) skin and muscle tissue were removed to expose unilaterally a 2×2 mm region over the left primary auditory cortexterm (from bregma anterior posterior (AP): −4.5– (−6.5) mm, dorsoventral (DV) 3–5 mm lateral to the bone edge of the upper skull surface).
A tip of a Teflon-insulated stainless steel wire (diameter 200 µm, A-M Systems, Carlsberg, WA, USA) was positioned on the surface of the dura above the auditory cortexterm on the basis of on-line recorded potentials. In addition, intracranial electrodes were implanted (Formwar® insulated stainless steel wire, diameter 100 µm, California Fine Wire Company Co, Grover Beach, CA, USA). Two electrodes with 400-µm tip separation were lowered to the ventral subiculumterm to coordinates AP: −6.5 mm, ML: 5.5 mm and DV 6.6. Three intermediate (caudal) hippocampal electrodes with 0.6 mm spacing between electrodes were lowered to coordinates AP: 6.0 mm, ML: 4.8, 5.4 and 6.0 mm, and DV: 4.6 mm. Based on histology, electrodes located in the area of interest were selected for further analysis. For the reference electrode, a hole was drilled in the scull over the right side of the cerebellum and a small insulin needle (BD Lo-Dose syringe, USA) was inserted in the cerebellum (AP −10 mm, ML: 2–3 mm and DV: 2 mm) and the animal was grounded by inserting a needle (18G, Terumo, Somerset, NJ, USA) subcutaneously into the neck.
Stimuli and Procedure Sinusoidal 4000-Hz tones of 25 ms and 75 ms in duration, including 5-ms rise and fall times, were used as stimuli. The sound pressure level for each tone was 70 dB with C-weighting (optimized for 40–100 dB measurement), as measured with a sound level meter (type 2235, Bruel & Kjaer, Nærum Denmark), in the location of the animal’s right pinna during the recording.
The deviantterm (P = 0.1) and the standardterm (P = 0.9) occurred in a series of 1000 tones randomly with the restriction that consecutive deviantsterm were separated by at least two standardsterm.
Figure 2 illustrates the four stimulus conditions used in the study.
In two of the stimulus conditions, (SSIterm 200 /SDIterm 200 and SSIterm 500 /SDIterm 500 ), the SSIterm and the SDIterm was either 200 ms or 500 ms, respectively, and the deviantterm always 25 ms and standardterm 75 ms in duration. Each of these conditions comprised one stimulus series.
In the other two stimulus conditions (SSIterm 500 /SDIterm 200 and SSIterm 200 /SDIterm 500 ), the SSIterm and the SDIterm differed. In the SSIterm 500 /SDIterm 200 -condition, the SSIterm was 500 ms and the SDIterm 200 ms. In the and SSIterm 200 /SDIterm 500 -condition, SSIterm was 200 ms and SDIterm 500 ms. Each of these conditions comprised two stimulus series to counterbalance the stimulus assignments (both durations assigned to both stimulus types across the series).
The sequential order of the six series (one per SSIterm 200 /SDIterm 200 - and SSIterm 500 /SDIterm 500 -conditions, respectively, and two per SSIterm 200 /SDIterm 500 - and SSIterm 500 /SDIterm 200 - conditions, respectively) was counterbalanced across the animals to control for effects of the duration of stimulus exposure and of the passage of time itself (e.g., through variations in the level of anesthesia).
The stimulus presentation was controlled by E-prime software (Pittsburg, PA, USA), and the stimuli were delivered from a PC via an active loudspeaker system (Studiopro 3, M-audio, Irwindale, CA, USA). The stimulation was presented via a passive loudspeaker directed towards the right ear of the animal at a distance of 20 cm.
Electrocortical and Hippocampal Recordings After surgery, the right ear bar was removed and recording started. Continuous electrocorticogram and hippocampal EEG were first 10-fold amplified using the MPA8I preamplifier (Multichannelsystems, Reutlingen, Germany), high-pass filtered at 0.1 Hz, 50-fold amplified, and low-pass filtered at 5000 Hz using an FA32I filter amplifier (Multichannelsystems), low-pass filtered at 400 Hz using a CyberAmp 380 filter amplifier (Molecular Devices Corporation), and finally sampled with 16-bit precision at 2 kHz (DigiData 1320A, Molecular Devices Corporation). The data were stored on a computer hard disk using Axoscope 9.0 data acquisition software (Molecular Devices Corporation).
Off-line Analysis The data analyses were performed offline using a Vision Analyzer (Brain Products, Gilching, Germany), Matlab 7.5 (MathWorks Inc., Natick, MA, USA). For the hippocampal recordings, the electrodes successfully implanted in the targeted areas were applied in the analysis (9 animals for the subiculumterm, 8 for the CA1term area, and 6 for the dentate gyrusterm, Figure 1 ).
First, artifacts were removed from the data. Electrocardiogram epochs containing voltage steps larger than 300 µV/ms were deleted for 200 ms before and after the artifact. Visual inspection of the data revealed that no other types of artifacts were present in the data.
The data were then offline-filtered (0.1–30 Hz, 24 dB/octave, Butterworth Zero Phase filters), segmented (for each deviantterm and its immediately preceding standardterm), and baseline corrected against the mean activity in the ±25 ms from tone onset and, hence in the 50-ms time window prior to duration deviance.
Finally, the artifact-free segments were averaged for each animal separately for deviantsterm and immediately preceding standardsterm for each condition. The averaging was also made across two stimulus blocks (with the opposite stimulus assignments) in the SSI500/SSD200 and SSI200/SSD500 conditions and for one stimulus block in the SSI200/SSD200 and SSI500/SSD500 conditions. Not less than 67 out of 100 sweeps were included in calculating the average values per stimulus type and condition in any animal, the average number being 98.7 out of 100 sweeps. For the response amplitudeterm·var analysis, the response time of 187.5 ms, as calculated from tone onset, was divided into five 37.5-ms segments. The last three of these segments (starting at 50 ms from change) which coincided with mismatch-like waveforms observable in the signal (see also previous findings of mismatch responsesterm commencing at about 60 ms post-stimulus in urethaneterm anesthetized rats [18] , [19] ) were included in the statistical analyses.
Histology After recording, the tips of the intracranial electrodes were marked in the tissue by anodal 30-µA 5-s current. The animal was sacrificed by cervical dislocation and the brain was moved from the skull and left for immersion post-fixation for 4 h in 4% paraformaldehyde (PFA) solution and after that in 30% sucrose solution for two days. The brains were stored in −20°C until slicing. Coronal sections (thickness 35 µm) were cut with a freezing slide microtome. The electrode locations were verified from the sections by cresyl violetterm·key staining and the exact locations of the electrode tips were confirmed by microscope observation.
Statistical Analysis The statistical analyses were performed with SPSS for Windows (SPSS Inc., Chicago, IL, USA). As there were no suitable non-parametric alternatives, ANOVA for repeated measuresterm·stat with stimulus (standard, deviant)term·var and time-segmentterm·var (first, second, third) as factors was used to justify post-hoc deviant-standard LFP amplitude comparisons for each time segmentterm·var separately. Degrees of freedom were Greenhouse-Geisser correctedterm·stat whenever the sphericity assumption was violated, and corrected P values were reported. Pairwise deviant-standard comparison for each time segmentterm·var was made using the paired t test (two-tailed)term·stat. An alpha level of 0.05 was used in all analyses.
Acknowledgments The authors wish to thank M.Sc. Mustak Ahmed for assisting in the data collection, Mr. Petri Kinnunen for preparing the stimulus materials, and Mr. Michael Freeman for checking the language of this report.
Funding Statement The study was supported by the Academy of Finland (project number 127595 for PA and 139767 for MP). The supporter had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.