Neural Substrates Underlying Human Delay and Trace Eyeblink Conditioning

Neural substrates underlying human delay and trace eyeblink conditioning

Dominic T. Cheng†‡, John F. Disterhoft§, John M. Power¶, Deborah A. Ellis†, and John E. Desmond†

†Department of Neurology, Division of Cognitive Neuroscience, Johns Hopkins University School of Medicine, 1620 McElderry Street, Reed Hall East 2, Baltimore, MD 21205; §Department of Physiology, Institute for Neuroscience, Northwestern University Feinberg School of Medicine, 303 East Chicago Avenue, Chicago, IL 60611; and ¶Queensland Brain Institute, Building 79, University of Queensland, St. Lucia QLD 4072, Australia

Edited by Larry R. Squire, Veterans Affairs Medical Center, San Diego, CA, and approved April 11, 2008 (received for review January 14, 2008) Classical conditioning paradigms, such as trace conditioning, in Previous functional neuroimaging investigations (20–26) which a silent period elapses between the offset of the conditioned [positron emission tomography and functional magnetic reso- stimulus (CS) and the delivery of the unconditioned stimulus (US), nance imaging (fMRI)] of eyeblink conditioning have used only and delay conditioning, in which the CS and US coterminate, are delay procedures to explore associative learning-related changes widely used to study the neural substrates of associative learning. in the brain, whereas trace eyeblink conditioning has been However, there are significant gaps in our knowledge of the neural largely ignored. Because the hippocampus has been shown to be systems underlying conditioning in humans. For example, evidence critical for trace EBCC (9–19) and significant hippocampal from animal and human patient research suggests that the hip- activity has also been reported during delay EBCC (21, 22, 24, pocampus plays a critical role during trace eyeblink conditioning, 27, 28), the question of whether hippocampal responding during but there is no evidence to date in humans that the hippocampus trace EBCC differs significantly from delay EBCC remains. is active during trace eyeblink conditioning or is differentially To address this question, the present EBCC work compared responsive to delay and trace paradigms. The present work pro- fMRI brain activation during delay and trace conditioning by vides a direct comparison of the neural correlates of human delay using a within-subjects design. Concurrent measurements of and trace eyeblink conditioning by using functional MRI. Behav- eyeblink responses were recorded by using infrared light deliv- ioral results showed that humans can learn both delay and trace ered through a fiber-optic probe (29). Participants received conditioning in parallel. Comparable delay and trace activation was pseudoconditioning and acquisition blocks while watching a measured in the cerebellum, whereas greater hippocampal activity silent movie. Pseudoconditioning consisted of four delay-alone, was detected during trace compared with delay conditioning. four trace-alone, and four airpuff-alone blocks, whereas acqui- These findings further support the position that the cerebellum is sition consisted of alternating delay and trace blocks (16 of each). involved in both delay and trace eyeblink conditioning whereas Each block consisted of nine trials and lasted 27 sec. Participants the hippocampus is critical for trace eyeblink conditioning. These were instructed to attend to the movie and were informed that results also suggest that the neural circuitry supporting delay and the study was designed to understand how their knowledge about trace eyeblink classical conditioning in humans and laboratory the movie would be affected by distracting sounds and airpuffs animals may be functionally similar. to the left eye. After conditioning, awareness of the CS–US relationship and movie knowledge were assessed by a postex- cerebellum ͉ functional MRI ͉ hippocampus perimental questionnaire (1). Behavioral and imaging data were parsed into early and late acquisition periods to assess time- lassical conditioning has been widely used as a model system dependent changes as a function of learning. Cto study the neural substrates of associative learning. One Results form of classical conditioning, eyeblink classical conditioning (EBCC), has received considerable attention by the neuro- Behavioral Results. Learning-related changes in behavior were assessed by repeated-measures ANOVA on both % CRs and science community. In EBCC, repeated pairings of a neutral peak response latencies. Changes in participant’s % CRs devel- conditioned stimulus (CS), such as a tone, and an unconditioned oped over the course of the experiment as demonstrated by a stimulus (US), such as an airpuff to the eye, eventually result in time main effect for delay (F ϭ 21.20, P ϭ 0.001) and trace the CS alone eliciting conditioned responses (CRs), suggesting 1,10 (F ϭ 6.21, P ϭ 0.032) conditioning (Fig. 1d). Post hoc tests that an association between the CS and US has been learned. 1,10 revealed that both early (t ϭ 3.84, P ϭ 0.003) and late (t ϭ Delay and trace conditioning are two different procedures that 10 10 4.60, P ϭ 0.001) delay % CRs were greater than % CRs vary in timing. In delay conditioning, the CS and US coterminate measured during pseudoconditioning. Early trace % CRs (t ϭ whereas in trace conditioning, a silent period (called the trace 10 1.75, P ϭ 0.111) did not differ from pseudoconditioning, but interval) elapses between offset of the CS and delivery of the US. participants showed significantly greater late trace % CRs (t ϭ This minor variation in temporal contiguity has significant 10 2.49, P ϭ 0.032). These behavioral findings suggest that partic- consequences on behavioral performance in both healthy indi- ipants successfully learned the CS–US relationship for both viduals and medial temporal lobe amnesiacs. For example, delay and trace stimuli. hippocampal patients show normal acquisition of delay condi- As an additional index of conditioning, peak response tioning but cannot learn trace conditioning (1–4). Considerable evidence from animal and human research suggests that the hippocampus plays a critical role during trace Author contributions: D.T.C., J.F.D., and J.E.D. designed research; D.T.C., D.A.E., and J.E.D. conditioning whereas the cerebellum is involved in both delay performed research; D.T.C., J.M.P., and J.E.D. analyzed data; and D.T.C., J.F.D., J.M.P., and trace conditioning (5–8). Many experimental techniques, D.A.E., and J.E.D. wrote the paper. such as electrophysiological recordings, lesion procedures, and The authors declare no conflict of interest. genetic manipulations, have been used in laboratory animals to This article is a PNAS Direct Submission. test the idea that the hippocampus is essential for successful trace ‡To whom correspondence should be addressed. E-mail: [email protected]. EBCC (9–19). However, the role of the human hippocampus This article contains supporting information online at www.pnas.org/cgi/content/full/ during trace EBCC has been mainly addressed through behav- 0800374105/DCSupplemental. ioral studies of healthy individuals and amnesic patients (1, 2, 4). © 2008 by The National Academy of Sciences of the USA

8108–8113 ͉ PNAS ͉ June 10, 2008 ͉ vol. 105 ͉ no. 23 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800374105 Downloaded by guest on October 1, 2021 ϭ ϭ Delay Conditioning Trace Conditioning (t10 0.588, P 0.570). There did not appear to be a significant a b ϭ 850 ms 250 ms correlation between delay awareness and delay % CRs (r 0.313, P ϭ 0.35). However, there was a positive trend (r ϭ 0.567, CS CS 500 ms P ϭ 0.069) between trace awareness and trace % CRs during the final four blocks. Participants answered 9.4 movie questions (of US US 100 ms 100 ms 10) correctly.

c Delay Block Trace Block Imaging Results. To investigate whether neural activity during trace conditioning differs from delay conditioning, we separated the acquisition period into early (first eight blocks) and late (last CS eight blocks) acquisition and contrasted activation between US delay and trace conditioning. Anatomically defined regions of interest based on probability maps defined by a cytoarchitectonic d n.s. n.s. n.s. n.s. e * ** ** ** mapping study (30) were used to sample brain activity in medial 180 45 temporal lobe (MTL) regions on an individual subject level. The Latency o

40 160 development of associative learning-related changes in these

f structures was assessed by comparing late acquisition with early 35 140 Peak

% CR acquisition for both delay and trace conditioning. Overall re- 30 Res 120

pons sponding in these structures during the acquisition of delay and 25

e trace conditioning is shown [supporting information (SI) Fig. 100 20 S1]. Bilateral MTL regions showed similar response profiles, but Pseudo- Early Late Pseudo- Early Late conditioning Acquisition Acquisition conditioning Acquisition Acquisition only the right MTL showed significantly greater trace-related activity compared with delay (t10 ϭ 2.44, P ϭ 0.035). Contralat- eral hippocampal activation to the trained eye is consistent with Delay Conditioning Trace Conditioning expectations based on animal work (17). Given the overall significance of the right MTL, we further Fig. 1. Study design and behavioral results. (a–c) Temporal relationship explored activations restricted to specific subregions of this area. between the CS and US in both delay and trace conditioning. (a) Delay CSs These subregions were determined by a cytoarchitectonic mapping PSYCHOLOGY lasted 850 ms and coterminated with a 100-ms US presentation. (b) Trace CSs lasted 250 ms and were followed by a 500-ms trace interval before a 100-ms study (30) and differentiated between the cornu ammonis (CA), the US presentation. (c) During acquisition, participants received 16 delay and 16 subiculum (SUB), the fascia dentata (FD), the hippocampal– trace blocks in an alternating fashion (nine trials per block). (d and e) Behav- amygdaloid transition area (HATA), the perirhinal (PRh) and ioral data showing learning-related changes over the course of acquisition. (d) entorhinal cortex (EC) (30). Of all of these subregions comprising Behavioral performance expressed as % CR. Relative to pseudoconditioning, the right MTL, only the FD showed significant differential respond- participants increased their % CRs to delay CSs during early and late acquisi- ing between the development of delay and trace conditioning [mean tion whereas greater % CRs for trace CSs were evident during late acquisition. (t10 ϭ 2.25, P ϭ 0.048) peak (t10 ϭ 2.25, P ϭ 0.049)] (Fig. 2 a and No signiﬁcant behavioral differences were seen between delay and trace CSs b). Activity during trace conditioning increased over the course of NEUROSCIENCE within each acquisition phase. (e) Behavioral performance expressed as la- the experiment, whereas a relative decrease was detected as delay tency of peak response. A signiﬁcant increase in response latencies was observed for both delay and trace conditioning during early and late acqui- conditioning progressed (Fig. 2a). Furthermore, a direct compar- sition (i.e., during training, the peak response shifted temporally closer to US ison of late delay and late trace acquisition revealed significantly ϭ ϭ presentation). greater trace-related activity (t10 4.59, P 0.001) (Fig. 2c). Importantly, there were no differences in behavioral CR expression between these two forms of conditioning during the late stages of latencies were calculated, and they further supported the idea acquisition, suggesting that increased responding in the FD was not that subjects acquired both delay and trace conditioning (Fig. performance-related. 1e). Learning-related changes developed over the course of the Similar analysis procedures were conducted to determine experiment as a repeated-measures ANOVA revealed a time whether responding in the cerebellum differentiates between main effect for delay (F1,10 ϭ 38.44, P ϭ 0.0001) and trace delay and trace conditioning. Contrasts between delay and trace (F1,10 ϭ 28.57, P ϭ 0.0003) conditioning. Post hoc tests showed conditioning blocks vs. a baseline period showed a common that, relative to pseudoconditioning, both early delay (t10 ϭ region of functional activation in the left cerebellar lobule HVI 5.55, P ϭ 0.0002) and late delay (t10 ϭ 6.20, P ϭ 0.0001) peak (Fig. 3). Mean and peak ␤ values showed significantly higher responses occurred closer to the time of US presentation. responding during delay (mean: t10 ϭ 3.62, P ϭ 0.0047; peak: Comparable results were seen for early (t10 ϭ 4.68, P ϭ 0.001) t10 ϭ 6.73, P ϭ 0.00005) and trace (mean: t10 ϭ 2.32, P ϭ 0.043; and late (t10 ϭ 5.35, P ϭ 0.0003) trace peak response latencies. peak: t10 ϭ 5.75, P ϭ 0.0002) conditioning compared with These findings indicate that participants produced more ap- baseline. However, delay- and trace-related activations did not propriately timed CRs during acquisition relative to differ from each other (mean: t10 ϭ 0.83, P ϭ 0.427; peak: t10 ϭ pseudoconditioning. 0.24, P ϭ 0.813), suggesting that this region is involved in both Direct comparisons between delay and trace conditioning delay and trace conditioning but does not discriminate between trials showed no significant changes in acquisition between these these two forms of learning. two forms of learning. Differences in % CRs (t10 ϭ 1.96, P ϭ 0.079) and latency measures (t10 ϭ 0.14, P ϭ 0.895) during early Discussion delay and trace trials were not significant. Also, % CRs (t10 ϭ The neural basis underlying EBCC has been well characterized 1.53, P ϭ 0.156) and latency measures (t10 ϭ 1.13, P ϭ 0.283) did by using a variety of animals and techniques and continues to be not differ between delay and trace trials during late acquisition. one of the most studied forms of associative learning. The These data suggest that participants learned both conditioning cerebellar cortex and deep nuclei have been linked to an animal’s paradigms equally quickly. ability to learn successfully both delay and trace EBCC (5–8), Awareness of the CS–US relationship was measured by a whereas additional structures, such as the hippocampus, are true/false postexperimental questionnaire (1). Participants cor- recruited during trace conditioning (9–19). The present work rectly answered the same number of delay and trace questions extends our current understanding of this neural circuit by using

Cheng et al. PNAS ͉ June 10, 2008 ͉ vol. 105 ͉ no. 23 ͉ 8109 Downloaded by guest on October 1, 2021 Right Hippocampus

a 0.2 * b 0.8 * c 0.6 * 0.7 0.5 0.1 0.6 0.4 0.0 0.5 0.3 -values -values -values 0.4 0.2 ∆ β ∆ β ∆ β -0.1 0.3 0.1 -0.2 0.2 0.0 Late Acquisition Late Acquisition Late Trace minus Late Delay minus minus (peak) Early Acquisition (mean) Early Acquisition (peak)

Delay Trace

LR-29 +23 -4

Fig. 2. Delay and trace activity specific to a subregion of the MTL (center of mass Talairach coordinates: ϩ23, Ϫ29, Ϫ4). Brain maps illustrate the right fascia dentata (FD) (30) overlayed onto high-resolution T1-weighted MPRAGE images. (a) Similar to overall MTL findings, the right FD showed significant differential responding between delay and trace conditioning. (b) Peak responses within this region also resulted in significantly greater trace-related activity. (c) Although behavioral performance between late delay and late trace conditioning was equivalent (Fig. 1 d and e), a direct comparison of FD responding between these two forms of learning revealed significantly greater FD activity during late trace conditioning, suggesting that differential responding in this area may not be performance-related.

fMRI to compare directly brain activity during human delay and culating two different indices of conditioning: % CR and latency trace eyeblink conditioning within the same acquisition session. of peak response. A significantly greater number of delay and Behavioral results showed that humans were capable of learning trace conditioned responses were measured during acquisition both delay and trace conditioning in parallel. Brain imaging data relative to pseudoconditioning, and the time of maximal eyelid outlined the important role of the human cerebellum during closure (peak response) shifted closer to US presentation over delay and trace EBCC and the critical role of the hippocampus the course of the experiment (Fig. 1 d and e). Both of these during trace EBCC. findings indicated that participants produced more adaptive eyeblink CRs during acquisition. Comparable levels of delay and Behavioral CRs. Participants were able to learn both delay and trace conditioned responses were measured throughout acqui- trace conditioning in parallel, which was demonstrated by cal- sition, suggesting that participants learned both conditioning paradigms equally quickly. Left Cerebellum (Lobule HVI) In the current work, identical interstimulus intervals (ISIs) for both trace and delay trial types were used to equate the stimulus **** conditions for fMRI contrasts, and as a result, distinguishing 0.4 1.0 between delay and trace CRs on the basis of temporal topog- 0.8 0.3 raphy of the CR was not possible. However, evidence that 0.6 separate learning processes occurred include differential pat- 0.2 -values -values 0.4 terns of brain activity between these two forms of learning and the positive relationship between awareness and trace CRs. Δ β

Δ β 0.1 0.2 Furthermore, the fact that delay and trace stimuli were suffi- 0.0 0.0 ciently dissimilar (pure tone vs. white noise) such that all mean peak participants reported hearing two different auditory stimuli Delay minus Baseline Trace minus Baseline suggests that CRs were specific to each trial type. Concurrent delay and trace conditioning has also been successfully demonstrated by using fear conditioning procedures (31), but the present fMRI study uses eyeblink conditioning procedures. This general approach allowed us to assess common and unique brain responses during the parallel acquisition of both delay and trace conditioning. Although significant learning-related changes were measured, LR-57 -24 -23 the level of conditioning inside the scanner was less robust than what is typically seen in human behavioral conditioning studies Fig. 3. Delay and trace activity observed within the left cerebellum (lobule (5). This effect likely reflects the distractions that accompany the HVI; center of mass Talairach coordinates: Ϫ24, Ϫ57, Ϫ23). Mean and peak response measures show signiﬁcantly greater activity for both delay and trace MRI environment (e.g., radiofrequency pulse; gradient noise). conditioning relative to baseline. Unlike hippocampal regions, differences Other factors such as pseudoconditioning (potentially producing between delay and trace conditioning were not found in this cerebellar ROI, latent inhibition) and a conservative CR time window (250 ms suggesting that this area does not discriminate between these two forms of pre-US) may have also contributed to the level of conditioning learning. observed in the current work.

8110 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800374105 Cheng et al. Downloaded by guest on October 1, 2021 Cerebellum. Significantly greater activation in the left (ipsilateral support this form of learning. Although both delay and trace to the trained eye) cerebellum (lobule HVI) was detected during conditioning rely on cerebellar deep nuclei to produce CRs, both delay and trace conditioning (Fig. 3). The present results trace conditioning depends on an intact hippocampus to provide are consistent with findings from previous functional neuroim- inputs to the deep nuclei via pontine afferents (8). aging studies on delay conditioning, which reported learning- Animal neurophysiology studies showed trace-related activity related changes in the cerebellum/lobule HVI (20–26). Evidence in recordings made from the CA region (14–18), whereas the from animal and human studies has shown that the cerebellar current work observed significant responding in the human FD. cortex and deep nuclei are involved in both delay and trace Although the anatomical regions of interest (ROIs) were based eyeblink conditioning (6–8). An event-related fMRI study of on precise measurements from a cytoarchitectonic mapping single-cue delay conditioning in the rabbit reported decreased study (30), the spatial resolution of our functional voxels and the blood-oxygenation-level-dependent (BOLD) responses in lobule close proximity between the FD and CA make it difficult to rule HVI of the cerebellar cortex during early and asymptotic learn- out any involvement of the CA region. Higher spatial resolution ing, with no changes visible during the unpaired pseudocondi- functional images of the human hippocampus may provide a tioning trials that preceded learning, although the BOLD re- better method to distinguish between structures within the sponse reduced in area and became more localized as learning hippocampus proper. proceeded (26). The decreased BOLD response was interpreted By simply considering the temporal profile of hippocampal as consistent with the fact that reduced drive from cerebellar activity during delay and trace conditioning, it is tempting to cortical Purkinje cells onto cerebellar deep nuclei would be conclude that activity in this region is simply related to the required to initiate CR performance controlled from the cere- production of CRs. However, a direct comparison between late bellar dentate–interpositus as training proceeded. In the present delay and trace learning revealed greater hippocampal activity work, we compared the BOLD response during late compared during trace learning (Fig. 2c). If trace-related activity in the with early training trials and observed no significant differences. hippocampus was strictly a function of CR expression, differen- Although the early vs. late training contrast was not significant tial responding in the FD should not have been observed in this for the cerebellar cortex in the present work, we did find contrast because delay and trace CRs were comparable during significantly greater cerebellar activity relative to baseline for this stage of acquisition (Fig. 1 d and e). These findings suggest both delay and trace conditioning. However, differential re- that increased activity in the hippocampus during trace learning sponding between delay and trace conditioning was not detected, could not simply be performance-related. Rather, the hippocam- suggesting that this region does not discriminate between these pus may be recruited during the production of delay and trace PSYCHOLOGY two forms of learning. CRs and may be critically relied on during the associative processes mediating trace conditioning. Hippocampus. Contrasting early and late acquisition periods The current findings are also consistent with conclusions allowed us to investigate the development of hippocampal drawn from animal lesion and human amnesic studies. The activity as a function of delay and trace learning. Greater acquisition of delay conditioning does not appear to be affected hippocampal responding was detected in the later stages of trace by lesions to the hippocampus (33–35), suggesting that this conditioning (Fig. 2a), suggesting that hippocampal involvement structure is not a part of the minimal essential circuitry required gradually increased over the course of trace learning. This for delay learning (36, 37). However, trace conditioning is NEUROSCIENCE response pattern, which closely parallels the development of severely impaired by lesions to the hippocampus (9–12), high- subject’s behavioral trace % CRs, is in agreement with data lighting this region’s importance during this form of learning. showing that greater activity in rabbit hippocampal cell record- Measuring greater hippocampal activity during trace than delay ings was detected on the same day that significant trace CRs conditioning in the current work further emphasizes the critical began to develop (16). The fact that the largest BOLD response role of the hippocampus during trace conditioning. was seen in the hippocampus contralateral to the trained eye late A limited number of studies have used functional neuroim- in acquisition is consistent with the pattern of changes in aging to study the role of the hippocampus during trace condi- single-unit responses seen in rabbit hippocampus (17), where the tioning. A magnetencephalogram (MEG) study reported greater larger firing rate increases were seen in the hippocampus con- hippocampal activity related to trace but not delay eyeblink tralateral compared with ipsilateral to the trained eye as conditioning (38). However, subcortical localization with MEG conditioning became well established. These findings are also is difficult because it lacks the anatomical resolution that MRI consistent with a rodent study (18) that reports hippocampal provides. Two studies used fear conditioning procedures and single-unit activity during the late stages of trace acquisition. In fMRI to report trace-related hippocampal activity (31, 39). In contrast to the late developing increase in trace-related hip- particular, results from a similar within-subjects fear condition- pocampal activity, a late decrease in hippocampal responding ing design used online US expectancy ratings to show that during delay conditioning was observed (Fig. 2a). This temporal participants rapidly acquired the trace CS–US relationship. profile suggests relatively greater hippocampal activity (mean ␤ Interestingly, this early explicit learning was accompanied by values: ϩ0.08 Ϯ 0.04) during early delay conditioning, which also differences in hippocampal response magnitude related to the corresponds to the development of delay CRs. The delay-related accuracy of predicting US presentation during trace condition- hippocampal activity also appears to be in line with animal ing. These fear conditioning results and the current findings electrophysiological evidence. These studies report that hip- suggest that the hippocampus is active around the time of pocampal cells respond very rapidly during the early stages of successful trace conditioning. delay learning, correspond strongly with behavioral CRs, and Several theories regarding the specific role of the hippocam- attenuate with continued training (18, 27, 28, 32). pus during trace conditioning have been proposed. For example, The pattern of hippocampal activation in the present work is the hippocampus may be needed to overcome stimulus discon- consistent with the interpretation that trace conditioning re- tiguity, to time CRs accurately, or to distinguish between the cruits additional forebrain structures, such as the hippocampus, interstimulus interval and trace interval (10, 17, 40). Another to bridge the temporal gap between the CS and US, whereas theory is that trace conditioning is an associative learning task delay eyeblink conditioning relies only on circuitry within the dependent on awareness or declarative memory, which requires cerebellum. Initial hippocampal responding during delay con- the hippocampus (1, 41–44). One other possibility is that the ditioning suggests an early involvement of the hippocampus that hippocampus is more active during complicated and difficult eventually subsides because the cerebellum is sufficient to forms of classical conditioning (e.g., reversal and trace) (13).

Cheng et al. PNAS ͉ June 10, 2008 ͉ vol. 105 ͉ no. 23 ͉ 8111 Downloaded by guest on October 1, 2021 Unfortunately, the design of the current work limits our ability delay contrast), suggesting that this activity was not strictly to test several of these theories. We were not able to rule out performance-related. These results suggest that the hippocam- certain theories (10, 17) because the short ISIs and rapid pus may be critically relied on during the associative processes presentations of conditioning trials in a block design prevented mediating trace conditioning. us from disentangling the hemodynamic response to the CS, trace interval, and US. Materials and Methods Longer trace intervals may require a greater dependence on Participants. Eleven healthy, right-handed volunteers (23.6 Ϯ 0.8 years; six an intact hippocampus, but shorter trace intervals have also been females, five males) were compensated $20 per hour for their participation. shown to have a deleterious effect on patients with hippocampal All procedures were approved by the Institutional Review Boards for human damage. Using a 500-ms trace interval, one study (4) showed that subject research at the Johns Hopkins University School of Medicine. MTL amnesiacs were impaired relative to control subjects, suggesting that the MTL is involved during trace conditioning Behavioral Apparatus. Stimulus presentation and behavioral data acquisition were controlled and analyzed by using a laptop computer interfaced to with a 500-ms trace period. Given these findings and the DT9834 data acquisition module (Data Translations) running custom software numerous studies that emphasize the importance of declarative developed under LabView version 7.1 (National Instruments). Presentation of knowledge/awareness during trace conditioning (1, 2, 41–44), it auditory stimuli was achieved by using pneumatic headphones (MRA, Inc.). A is reasonable to predict hippocampal activity even during shorter movie (Charlie Chaplin’s The Gold Rush) was projected onto a back- trace intervals. illuminated screen and viewed through an adjustable mirror attached to the head coil. Custom modifications to standard laboratory safety goggles were Awareness. The idea that awareness of CS–US relationships is made so that the end of a polyethylene tube could be attached (Nalgene) for critical for successful trace learning has been debated (45–47). airpuff delivery and also to accommodate a MRI-compatible infrared sensor All participants were able to differentiate between delay and for recording eyeblinks. A fiber-optic probe (RoMack, Inc.) measured the reflectance of infrared light from the left eye (29), and airpuff delivery was trace tones, allowing us to present separate postexperimental controlled by a solenoid valve (Asco). questionnaires for each stimulus. Performance on these questionnaires was not significantly correlated with their behavioral Imaging Apparatus. Whole-brain functional imaging was performed on a CRs for either delay or trace conditioning. A closer examination 3-tesla MRI scanner (Phillips) by using a T2*-weighted gradient echo planar showed a positive trend (P ϭ 0.069) between the number of imaging (EPI) pulse sequence. Contiguous 7-mm oblique (perpendicular to the correctly answered questions on the trace questionnaire and axis of the hippocampus) slices were collected (TR, 2,000 ms; TE,30.0 ms; FOV, eyeblink CRs recorded at the end of acquisition. This finding is 24 cm; flip angle, 70°) in a series of 810 sequential images (for a total of 1,620 in agreement with previous work (48) showing that awareness sec). Structural images were acquired by using a T1-weighted magnetization- was significantly correlated with single-cue trace (r ϭ 0.49) but prepared rapid acquisition gradient echo (MPRAGE) pulse sequence. not single-cue delay conditioning. In fact, the relationship between awareness and trace conditioning may even be stronger in Stimuli. Delay and trace CSs were counterbalanced between a binaural ϭ ϭ 1,200-Hz tone and white noise (90 dB). The delay stimulus lasted 850 ms and the present work (r 0.60) than in the previous study (48) (r coterminated with a 100-ms corneal airpuff (4 psi) to the left eye (Fig. 1a). The 0.49) but simply did not reach statistical significance because of ϭ trace stimulus lasted 250 ms and was followed by a 500-ms trace period before a small sample size (n 11). The present behavioral findings are airpuff presentation (Fig. 1b). Trials were grouped into blocks: nine trials per in agreement with several studies reporting that awareness is block, 2 sec per trial, 3-sec intertrial interval, such that each block lasted 27 sec necessary for trace learning (1, 2, 41–44). (Fig. 1c). Pseudoconditioning was seamlessly followed by acquisition blocks. Although a participant’s hippocampal activity was not corre- Pseudoconditioning consisted of alternating four delay-alone, four trace- lated with awareness, it should be noted that this finding does not alone, and four airpuff-alone blocks, whereas acquisition consisted of alter- discount the idea that awareness is necessary for trace condi- nating 16 delay and 16 trace blocks. Presentation order was counterbalanced tioning. There are several procedural differences between the among subjects. current article and previous reports supporting the hippocampus–awareness–trace conditioning theory. Perhaps most impor- Procedure. Once participants were fitted with the safety goggles and placed in the magnet, they were instructed to watch and pay attention to the silent tantly, participants in prior studies (1, 41–43) received only trace movie while distracting sounds and airpuffs were presented. They were told conditioning, whereas participants in the current work received that their movie knowledge would be tested after the experiment. Further, both delay and trace trial types. It has been shown that learning they were informed that this study investigated the effects that distracting delay conditioning can affect the development of trace condi- sounds and airpuffs have on their ability to remember details about the movie. tioning (13, 49), which leaves open the possibility that awareness After conditioning, a movie quiz and postexperimental questionnaire assess- of one trial type could influence awareness of the other. Com- ing awareness of the CS–US contingencies were administered. parisons between the current results and previous awareness– trace conditioning papers should be made with these procedural Behavioral Eyeblink Analysis. The criterion for a response to be considered a CR differences in mind. is that the difference between the maximum and minimum responses in a 250 ms pre-US time window must exceed four times the standard deviation of the Conclusions mean of the baseline period (250 ms before CS presentation). A time window of 250 ms pre-US presentation was selected to minimize the inclusion of This work combines fMRI and a within-subjects design of voluntary and ␣ responses as CRs (50). Performance was expressed as % CR. delay/trace eyeblink conditioning to show that participants can Peak response latencies relative to CS onset during this 250-ms time window learn both delay and trace conditioning in parallel. Significant were also calculated as an additional behavioral measure of conditioning. cerebellar activity to both delay and trace conditioning was Repeated-measures ANOVA were performed on these two variables. observed, but differential responding between these two trial types was not. These findings support the view that this region Imaging Analysis. Structural and functional imaging preprocessing and statis- is involved in both delay and trace conditioning but does not tical analyses were performed with Statistical Parametric Mapping (SPM2) discriminate between the two. Our hippocampal results confirm software (Wellcome Department of Cognitive Neurology, London, U.K.). EPI and extend previous animal lesion, human amnesic, and elec- functional images were realigned and resliced correcting for minor motion artifacts and MPRAGE structural images were coregistered to the mean mo- trophysiology data by showing greater hippocampal activity tion-corrected functional image for each subject. Both functional and struc- during trace than delay conditioning in humans. The timing of tural images were transformed into standard stereotaxic space according to this activity appeared to coincide with the development of delay the Montreal Neurological Institute (MNI) protocol, and the functional images and trace CRs. However, differential hippocampal responding were smoothed with a Gaussian filter (full-width half-maximum 5 mm). The was measured when CR expression was comparable (trace vs. general linear model was used to estimate individual subject activations, and

8112 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800374105 Cheng et al. Downloaded by guest on October 1, 2021 a random-effects analysis was conducted for all subjects. Estimated mean and the MTL were analyzed separately, and a direct comparison between late peak ␤ weights were used as a measure of brain activation. A Talairach delay acquisition and late trace acquisition was conducted to test the idea that transformation was performed, and these coordinates are reported in the the hippocampus is critically important for trace conditioning. figures and text (51, 52). Similar imaging analyses were conducted to determine whether respond- Overwhelming evidence from laboratory animal studies and the human ing in the cerebellum differentiates between delay and trace conditioning. lesion literature identify regions in the MTL as critical for learning trace Because of the known involvement of cerebellar lobule HVI in EBCC from conditioning. Structural ROIs based on probabilistic maps of the MTL region animal and human literature (6–8), a union of activation from both averaged trace and delay conditioning SPM {Z} maps (based on a delay/trace minus reported in a cytoarchitectonic mapping study (30) and made available as an pre-CS baseline contrast), thresholded at P Ͻ 0.025, within the anatomical SPM toolbox were used to sample neural activity on an individual subject level. boundaries of left HVI (53) was made to create an appropriate cerebellar ROI. The MTL region consisted of the cornu ammonis, the subiculum, the fascia This ROI was used to sample cerebellar activity on an individual subject level dentata, the hippocampal–amygdaloid transition area, the perirhinal, and to investigate the idea that this region responds equally to both delay and entorhinal cortex (30). Contrasts were designed to explore the nature of MTL trace conditioning. activity during both delay and trace conditioning. The development of MTL responding was investigated by separating the acquisition period into early ACKNOWLEDGMENTS. We thank M. M. Oh and C. Weiss for technical assis- (first eight blocks) and late (last eight blocks) acquisition and contrasting these tance. This work was supported by National Institutes of Health Grant R01 two periods during delay and trace conditioning. Specific subregions within AG021501 (to J.E.D.).

1. Clark RE, Squire LR (1998) Classical conditioning and brain systems: The role of 28. Berger TW, Thompson RF (1978) Neuronal plasticity in the limbic system during classical awareness. Science 280:77–81. conditioning of the rabbit nictitating membrane response. I. The hippocampus. Brain 2. Knuttinen M-G, Power JM, Preston AR, Disterhoft JF (2001) Awareness in classical Res 145:323–346. differential eyeblink conditioning in young and aging humans. Behav Neurosci 29. Miller MJ, Li L, Weiss C, Disterhoft JF, Wyrwicz AM (2005) A ﬁber optic-based system for 115:747–757. behavioral eyeblink measurement in a MRI environment. J Neurosci Methods 141:83– 3. Gabrieli JDE, et al. (1995) Intact delay-eyeblink classical conditioning in amnesia. Behav 87. Neurosci 109:819–827. 30. Amunts K, et al. (2005) Cytoarchitectonic mapping of the human amygdala, hippocam- 4. McGlinchey-Berroth R, Carrillo MC, Gabrieli JDE, Brawn CM, Disterhoft JF (1997) pal region and entorhinal cortex: Intersubject variability and probability maps. Anat Impaired trace eyeblink conditioning in bilateral, medial temporal lobe amnesia. Embryol 210:343–352. Behav Neurosci 111:873–882. 31. Knight DC, Cheng DT, Smith CN, Stein EA, Helmstetter FJ (2004) Neural substrates 5. Disterhoft JF, et al. (2003) in Neuropsychology of Memory, eds Squire LR, Schachter D mediating human delay and trace fear conditioning. J Neurosci 24:218–228. (Guilford, New York), pp 97–113. 32. Sears LL, Steinmetz JE (1990) Acquisition of classically conditioned-related activity in 6. Christian KM, Thompson RF (2003) Neural substrates of eyeblink conditioning: Acqui- the hippocampus is affected by lesions of the cerebellar interpositus nucleus. Behav

sition and retention. Learn Mem 11:427–455. Neurosci 104:681–692. PSYCHOLOGY 7. Gerwig M, Kolb FP, Timmann D (2007) The involvement of the human cerebellum in 33. Schmaltz LW, Theios J (1972) Acquisition and extinction of a classically conditioned eyeblink conditioning. Cerebellum 6:38–57. response in hippocampectomized rabbits (Oryctolagus cuniculus). J Comp Physiol 8. Woodruff-Pak DS, Disterhoft JF (2008) Where is the trace in trace conditioning? Trends Psychol 79:328–333. Neurosci 31:105–112. 34. Solomon PR, Moore JW (1975) Latent inhibition and stimulus generalization of the 9. Solomon PR, Vander Schaaf ER, Thompson RF, Weisz DJ (1986) Hippocampus and trace classically conditioned nictitating membrane response in rabbits (Oryctolagus cunic- conditioning of the rabbit’s classically conditioned nictitating membrane response. ulus) after dorsal hippocampal ablation. J Comp Physiol Psychol 89:1192–1203. Behav Neurosci 100:729–744. 35. Akase E, Alkon DL, Disterhoft JF (1989) Hippocampal lesions impair memory of short- 10. James GO, Hardiman MJ, Yeo CH (1987) Hippocampal lesions and trace conditioning in delay conditioned eye blink in rabbits. Behav Neurosci 103:935–943. the rabbit. Behav Brain Res 23:109–116. 36. Desmond JE, Moore JW (1982) A brain stem region essential for the classically condi-

11. Moyer JR, Deyo RA, Disterhoft JF (1990) Hippocampectomy disrupts trace eye-blink tioned but not unconditioned nictitating membrane response. Physiol Behav 28:1029– NEUROSCIENCE conditioning in rabbits. Behav Neurosci 104:243–252. 1033. 12. Weiss C, Bouwmeester H, Power JM, Disterhoft JF (1999) Hippocampal lesions prevent 37. Mauk MD, Thompson RF (1987) Retention of classically conditioned eyelid responses trace eyeblink conditioning in the freely moving rat. Behav Brain Res 99:123–132. following acute decerebration. Brain Res 403:89–95. 13. Beylin AV, et al. (2001) The role of the hippocampus in trace conditioning: Temporal 38. Kirsch P, et al. (2003) Cerebellar and hippocampal activation during eyeblink condi- discontinuity or task difficulty? Neurobiol Learn Mem 76:447–461. tioning depends on the experimental paradigm: A MEG study. Neural Plast 10:291–301. 14. Moyer JR, Thompson LT, Disterhoft JF (1996) Trace eyeblink conditioning increases CA1 39. Bu¨chel C, Dolan RJ, Armony JL, Friston KJ (1999) Amygdala–hippocampal involvement excitability in a transient and learning-specific manner. J Neurosci 16:5536–5546. in human aversive trace conditioning revealed through event-related functional mag- 15. Weiss C, Kronforst-Collins MA, Disterhoft JF (1996) Activity of hippocampal pyramidal netic resonance imaging. J Neurosci 19:10869–10876. neurons during trace eyeblink conditioning. Hippocampus 6:192–209. 40. Bolles RC, Collier AC, Bouton ME, Marlin NA (1978) Some tricks for ameliorating the 16. McEchron MD, Disterhoft JF (1997) Sequence of single neuron changes in CA1 hip- trace-conditioning deficit. Bull Psychonom Soc 11:403–406. pocampus of rabbits during acquisition of trace eyeblink conditioned responses. 41. Clark RE, Squire LR (1999) Human eyeblink classical conditioning: Effects of manipu- J Neurophysiol 78:1030–1044. lating awareness of the stimulus contingencies. Psychol Sci 10:14–18. 17. Weible AP, O’Reilly J-A, Weiss C, Disterhoft JF (2006) Comparisons of dorsal and ventral 42. Manns JR, Clark RE, Squire LR (2000) Parallel acquisition of awareness and trace hippocampus cornu ammonis region 1 pyramidal neuron activity during trace eye- eyeblink classical conditioning. Learn Mem 7:267–272. blink conditioning in the rabbit. Neuroscience 141:1123–1137. 43. Manns JR, Clark RE, Squire LR (2000) Awareness predicts the magnitude of single-cue 18. Green JT, Arenos JD (2007) Hippocampal and cerebellar single-unit activity during trace eyeblink conditioning. Hippocampus 10:181–186. delay and trace eyeblink conditioning in the rat. Neurobiol Learn Mem 87:269–284. 44. Knight DC, Nguyen HT, Bandettini PA (2006) The role of awareness in delay and trace 19. Tseng W, Guan R, Disterhoft JF, Weiss C (2004) Trace eyeblink conditioning is hip- fear conditioning in humans. Cogn Affect Behav Neurosci 6:157–162. pocampally dependent in mice. Hippocampus 14:58–65. 45. LaBar KS, Disterhoft JF (1998) Conditioning, awareness, and the hippocampus. Hip- 20. Molchan SE, Sunderland T, McIntosh AR, Herscovitch P, Schreurs BG (1994) A functional pocampus 8:620–626. anatomical study of associative learning in humans. Proc Natl Acad Sci USA 91:8122– 46. Lovibond PF, Shanks DR (2002) The role of awareness in Pavlovian conditioning: 8126. Empirical evidence and theoretical implications. J Exp Psychol Anim Behav Process 21. Logan CG, Grafton ST (1995) Functional anatomy of human eyeblink conditioning 28:3–26. determined with cerebral glucose metabolism and positron-emission tomography. 47. Manns JR, Clark RE, Squire LR (2002) Standard delay eyeblink conditioning is indepen- Proc Natl Acad Sci USA 92:7500–7504. dent of awareness. J Exp Psychol Anim Behav Process 28:32–37. 22. Blaxton TA, et al. (1996) Functional mapping of human learning: A positron emission 48. Manns JR, Clark RE, Squire LR (2001) Single-cue delay eyeblink conditioning is unre- tomography activation study of eyeblink conditioning. J Neurosci 16:4032–4040. lated to awareness. Cogn Affect Behav Neurosci 1:192–198. 23. Schreurs BG, et al. (1997) Lateralization and behavioral correlation of changes in 49. Woodruff-Pak DS (1993) Eyeblink classical conditioning in HM: Delay and trace para- regional cerebral blood flow with classical conditioning of the human eyeblink re- digms. Behav Neurosci 107:911–925. sponse. J Neurophysiol 77:2153–2163. 50. Gormezano I (1966) in Experimental Methods and Instrumentation in Psychology,ed 24. Ramnani N, Toni I, Josephs O, Ashburner J, Passingham RE (2000) Learning- and Sidowski JB (McGraw–Hill, New York), pp 385–420. expectation-related changes in the human brain during motor learning. J Neuro- 51. Talairach J, Tournoux PA (1988) Co-Planar Stereotaxic Atlas of the Human Brain: physiol 84:3026–3035. 3-Dimensional Proportional System: An Approach to Cerebral Imaging (Thieme, New 25. Knuttinen M-G, et al. (2002) Electromyography as a recording system for eyeblink York). conditioning with functional magnetic resonance imaging. Neuroimage 17:977–987. 52. Lancaster JL, et al. (2007) Bias between MNI and Talairach coordinates analyzed using 26. Miller MJ, et al. (2003) fMRI of the conscious rabbit during unilateral classical eyeblink the ICBM-152 brain template. Hum Brain Map 28:1194–1205. conditioning reveals bilateral cerebellar activation. J Neurosci 23:11753–11758. 53. Tzouri-Mazoyer N, et al. (2002) Automated anatomical labeling of activations in SPM 27. Berger TW, Alger B, Thompson RF (1976) Neural substrate of classical conditioning in using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. the hippocampus. Science 192:483–485. Neuroimage 15:273–289.

Cheng et al. PNAS ͉ June 10, 2008 ͉ vol. 105 ͉ no. 23 ͉ 8113 Downloaded by guest on October 1, 2021