Project topics and materials
Get Complete Project Material File(s) Now! »
EEG
Brain-Derived Neurotrophic Factor Val66Met affects Recollection but not Familiarity ERP components of human recognition memory.Thompson, C.S.1,4*, Collins, H. 1, Antia, U.2,4, Shelling, A.N.3,4, Russell, B.R.3,4, Waldie, K.W.1,4, and Kirk, I.J.1,4*.Research Centre for Cognitive Neuroscience, Department of Psychology, University of Auckland, New Zealand.
School of Pharmacy, University of Auckland, Auckland, New Zealand.
Department of Obstetrics & Gynaecology, Faculty of medical and Health Sciences,
University of Auckland, Auckland, New Zealand.
Centre for Brain Research, University of Auckland, New Zealand.To whom correspondence should be addressed. email: chris.thompson@auckland.ac.nz or i.kirk@auckland.ac.nz
Abstract
A single nucleotide polymorphism of the human BDNF gene (Val66Met) may account for much of the variation in human memory performance. However, it is not clear if this polymorphism affects all forms of memory. BDNF is concentrated most heavily in the hippocampus, and therefore would be likely to have a greater effect on hippocampal dependent memory. Recognition memory involves the contribution of two distinct retrieval processes, Recollection and Familiarity. Prior research suggests that Familiarity does not depend on the hippocampus, but Recollection does. Recent evidence has shown Recollection and Familiarity are associated with distinct event-related potentials (ERP): Familiarity with an early-onset effect called the FN400; and Recollection with a later positivity called the late positive component (LPC). The current research investigated whether different BDNF genotypes differ with respect to their generation of the FN400 and the LPC. Using ERP generation in a facial recognition memory paradigm, no genotype differences in FN400 amplitude (evoked when correctly recognising a previously presented face) were found. However, Val/Val individuals generated a significantly more positive LPC when correctly identifying an old face after a period of consolidation of 24 hours. These findings suggest that the Val66Met polymorphism may play an exclusive role for hippocampal dependent memory.
Introduction
Some people have much better memories than others. This variation in memory ability may be due to the variation of a gene that controls secretion of brain-derived neurotrophic factor (BDNF). In humans, a single nucleotide polymorphism (SNP) of the BDNF gene (Val66Met; SNP rs6265) has been shown to influence episodic memory performance, and the degree of task-related hippocampal activation measured by fMRI. Specifically, it has been shown that individuals carrying the Met allele perform significantly worse in a test of episodic memory relative to those homozygous for the Val allele, and have significantly lower levels of neural activation in the hippocampal region(Egan et al., 2003; Hariri et al., 2003). However, it is not known whether this difference in memory seen by genotype holds for all forms of memory. Recognition memory in particular has been shown to be two functionally dissociable systems, known as Familiarity and Recollection (Rugg & Yonelinas, 2003; Yonelinas, 2001, 2002). Familiarity describes a feeling that a particular stimulus may have been experienced before but is devoid of contextual information, whereas Recollection is the recognition of a stimulus as well as the contextual details of the environment accompanying the original presentation of the stimulus (Gardiner, 1988; Jacoby, 1991; Mandler, 1980; Tulving, 1985; Yonelinas, 1999, 2001, 2002; Yonelinas et al., 1996; Yonelinas & Jacoby, 1996). These two systems have been shown through numerous animal, clinical and imaging studies to also be dependent on different neural substrates, namely Recollection being dependent on the hippocampus, whilst Familiarity involving the perirhinal cortex (Vargha-Khadem, Gadian, Watkins, Connelly, Van Paesschen, & Mishkin, 1997; Aggleton & Brown, 1999; Aggleton & Brown, 2006; Yonelinas et al., 2002; Davachi, Mitchell and Wagner, 2003). As BDNF is most strongly concentrated in the hippocampus (Pezawas et al., 2005), it is possible that the BDNF polymorphism affects Recollection forms of memory, but not Familiarity. Thus, this could lead to differences in the neurological brain processes underlying these two forms of memory.Recent evidence has suggested that Familiarity and Recognition are associated with two distinct ERPs. Familiarity is associated with an early frontally distributed ERP called the FN400, whilst Recollection is associated with a later parietal positivity called the LPC (Addante, Ranganath & Yonelinas, 2012; Curran, 2000; Duzel, Vargha-Khadem, Heinze & Mishkin, 2001; Rugg & Curran. 2007). Addante and colleagues (2012) showed that when individuals successfully recognized a stimuli as ‘old’ but were unable to remember contextual information about the stimulus (measured by source accuracy – whether the stimuli came from one condition or another) a clear FN400 ERP was seen. However, when the subject could accurately locate the source of the memory an LPC effect was seen. This suggests the recruitment of contextual elements to the recognition of stimuli is a function of the LPC. As patients with hippocampal damage are able to show an FN400 response but not an LPC response, it is argued that the LPC is dependent on the hippocampus (Duzel et al., 2001).The aim of the current study was to see whether or not the BDNF Val66Met polymorphism played a significant role in modulating the FN400 and LPC components of recognition memory. As the LPC is associated with contextual information, we employed a method where we had subjects study a set of stimuli and return a day later to recall this information. These responses would be compared with individuals responding to stimuli immediately after initial presentation, which would show a FN400. We hypothesized that individuals with a copy of the Met allele would show less of an LPC response, but the polymorphism would have no effect on FN400 response.
Materials and Methods
Genotyping – A full description of the genotyping process can be found in Chapter 2
EEG Procedure
Subjects
An initial cohort was recruited and genotyped. From this cohort, eighteen healthy participants with a mean age of 23.5 years (range 21-29; SD = 2.5 years; 9 females) were selected such that they formed two groups defined by Val66Met genotype (9 Val/Val, and 9 carrying at least one Met allele – this group consisted of 4 Met/Met participants). Some of these participants were individuals from the Chapter 2 study on LTP, but not enough to be able to correlate the results of that study with this study. The cohort did not differ from Hardy-Weinburg equilibrium (χ2(2) = 0.818, p = .664), and there were no gender differences between genotypes (χ2(2) = 0.876, p = .645). Two participants chose not to participate between the first EEG and second EEG, therefore only 16 subjects data was available to be analysed for the second part of the EEG. All participants had normal or corrected-tonormal vision. Subjects gave their informed consent to participate in the study and all experimental procedures were approved by the University of Auckland Human Subjects Ethics Committee. EEG Acquisition A full description of the EEG acquisition parameters can be found in Chapter 2.
Stimuli
Two hundred faces obtained from the FEI Face database (Thomaz & Giraldi, 2010) were used in this experiment. All faces were on white backgrounds and were resized to 480 x 480 pixels using Adobe Photoshop. There was always an equal number of female to male faces in all parts of the experiment. Fifty faces were used in the study phase of each memory experiment (‘old’). During the recall phase participants identified the 50 studied/‘old’ objects from 100 object images (50 ‘old’ and 50 ‘new’). Four sets of faces were counterbalanced between ‘old’ and ‘new’ and counterbalanced between participants so each set were alternated between the Familiarity and Recollection tasks. Subjects were seated 57 cm from the display screen monitor (measured from the computer screen to the participant’s face). Stimuli were presented on a SVGA computer monitor (1024 x 798 pixel resolution; 60Hz refresh rate). Stimulus presentation was controlled using E-Prime v1.1 (Psychology Software Tools). TTL pulses generated via the parallel port of the display computer provided synchronisation of stimulus events with EEG acquisition. Millisecond timing routines for the visual displays and pulse generation were conducted as outlined in the E-Prime User Guide (Psychology Software Tools, Pittsburgh, PA, USA).
Experimental Procedure
The methods for the Familiarity and Recollection experiments are described together. The only methodological differences between the experiments was the time between the study and recall phases. In the Familiarity Task the recall phase came immediately after the study phase. In the Recollection Task the recall phase took place 24 hours after the study phase.Procedure;Two tests of recollection- and familiarity-based recognition memory were employed as measures of overall object recognition memory. In the Recollection Task the study phase and subsequent recall phase were separated by 24 hours, whereas they were consecutive in the Familiarity Task. In the study phase of the Familiarity Task participants were asked to actively put-to-memory 50 faces presented individually with a fixation cross 1500-2500 second jittered delay between each face. After the study phase, participants then had to make ‘old-new’ judgments for 100 faces (50 from the study phase (‘old’) and 50 novel objects (‘new’)). Subjects were advised to use the keyboard to indicate ‘1’ for ‘old’ and ‘2’ for ‘new’. Participants were then removed from the EEG machine and testing equipment, and taken to a separate testing room where they were given 30 minutes to learn another set of 50 faces. These were presented on a computer using Microsoft Office Powerpoint. Participants were advised that they would be required to remember these faces for the recall phase of the Recollection Task, which was to occur the following day. The logic of this was to allow for a deeper processing and encoding of the faces to occur. As the hippocampus is involved in contextual information about a stimulus, we hypothesized that recalling these faces a day later would tap into a different memory network. Twentyfour hours after this study phase, participants completed the ‘old-new’ task to identify the ‘old’ objects they had learned the previous day and the ‘new’ objects.
EEG Analysis
EEG analyses were performed using custom software from The University of Auckland, School of Psychology. EEG readings were then segmented into 1100ms epochs – 100ms before the onset of the stimulus and 1000ms post-stimulus onset. Automatic eyemovement correction was made on all segments according to the methods developed by Jervis and colleagues (Ferree, Luu, Russell, & Tucker, 2001). A priori defined time windows were used based upon established literature of familiarity and recollection-related effects (Addante et al., 2012; Curran, 2000; Rugg & Curran, 2007). As per the literature we selected the following time windows for analysis: 400-600 ms (Figure 3.1) for the Familiarity Task, and 600-800 ms (Figure 3.2) for the Recollection Task. Averages were generated over this time period by analyzing clusters of seven electrodes situated around Cz and Cp5 (Figure 3.3) under the 10-10 system (Luu & Ferree, 2000). The electrodes to identify the FN400 and LPC components were selected for analysis on the basis of previous literature (Addante et al., 2012; Curran, 2000; Rugg & Curran, 2007). FN400 and LPC components were measured for both familiarity and recollection experiments, and amplitude increase was measured by looking at increases for successful ‘old’ responses versus successful ‘new’ responses. Misses and false alarms were removed from the analysis. The difference in component amplitude was taken (‘old’ minus ‘new’) and used for statistical analysis.
Milliseconds
ERP waveforms demonstrating increased positivity seen for successful recognition of ‘old’ stimuli. The blue box shows the time period that the averages were generated from. Data is taken from a representati
Milliseconds
ERP waveforms demonstrating increased positivity seen for successful recognition of ‘old’ stimuli. The blue box denotes the time period for where averages were generated. Data is taken from a representative subject.Approximate location of electrodes used for analysis. A shows the electrodes centred on Cz (under the 10-10 system) and were the electrodes used to identify the FN400 component, whilst B shows the electrodes centred around Cp5 (under the 10-10 system) and were the electrodes used to identify the LPC components.
Statistical Analyses
T-tests showed no differences for FN400 for the recollection experiment, and no differences for LPC for the familiar experiment, and thus were excluded from all further analyses. For all statistical tests, subjects were grouped into two groups, either not having a Met allele (Val/Val individuals) or having a copy of the met allele (Val/Met, & Met/Met). For the FN400 and LPC analysis, 2 x 2 ANOVAs were employed to see whether the amplitude for an ‘old’ and ‘new’ response was different, and whether this difference was higher or lower for one group than the other. An independent t test was employed to see whether or not there were behavioural differences between the two groups. Task performance was calculated as sensitivity (d’), which is the difference between hit rate and false alarm rate. All statistical analyses were run using the computer program SPSS version 17, and an alpha of .05 was used for all analyses. Bonferroni corrections were used for post hoc analyses.
Results
Behavioural Results
On the Familiarity task there were no significant performance differences between Val/Val (M = 1.47, SE = .29) and Met carriers (M = 1.49, SE = .27; t (16) = -.07, p > .05).On the Recollection task there were no significant performance differences between Val/Val (M = 3.36, SE = .18) and Met carriers (M = 2.97, SE = .28; t (13) = 1.15, p > .05). Thus there was no significant performance difference between the genotype groups in either of the facial recognition memory tests.
Recognition memory – Familiarity
Mean values of FN400 (Successful ‘Old’ responses and successful ‘New’ responses) for the two Genotype groups (Val/Val and Met Carrier).The results of the 2x 2 split plot ANOVA (seen in Table 3.1), revealed a significant main effect of Memory on ERP amplitude (F (1, 16) = 11.58, p < .05). The amplitude of the FN400 for ‘Old’ responses (M = 1.31, SE = 0.28) was significantly more positive than ‘New’ responses. (M = .47, SE = 0.34; Figure 3.4). This shows the paradigm successfully elicits a FN400 component. However, there was no significant interaction between Memory and Genotype (F (1, 16) = .04, p > .05) and the main effect of genotype was not significant (p > .05). Thus, the increase in amplitude of the FN400 component seen for successful recognition of an ‘old’ face versus a new ‘face’ does not differ between genotype.
Type of stimuli
The main effect of Memory type on FN400 amplitude. Correctly recognizing an ‘old’ face elicits a significantly more positive FN400 than recognizing a ‘new’ face.
Recognition memory – Recollection
Mean values of LPC (Succesful ‘Old’ responses and successful ‘New’ responses) for the two Genotype groups (Val/Val and Met Carrier).The results of the 2x 2 split plot ANOVA (seen in Table 3.2), revealed a significant main effect of Memory on ERP amplitude (F (1, 14) = 4.63, p < .05). The amplitude of the LPC for ‘old’ responses (M = 1.27, SE = .21) was significantly more positive than ‘new’ responses. (M = .95, SE = .19; p < .05). This shows the paradigm successfully elicits an LPC component. The interaction between Memory and Genotype was also significant (F (1, 16) = .04, p < .05; Figure 3.5). Post hoc analysis revealed for Val/Val individuals, the amplitude of the LPC for ‘old’ responses (M = 1.75, SE = .32) was significantly more positive than ‘new’ responses. (M = .79, SE = .28; p < .05). For Met carriers, the amplitude of the LPC for ‘old’ responses (M = .86, SE = .29) was not significantly more positive than ‘new’ responses. (M = 1.04, SE = .26; p > .05). Lastly, when looking at the amplitude of the LPC for ‘old’ responses, Val/Val individuals were significantly more positive than Met carriers (p < .05). The results show the increase in amplitude of the LPC component seen for successful recognition of an ‘old’ face versus a new ‘face’ differed between genotype.The main effect of genotype was not significant (p > .05).
Genotype
The significant difference in LPC modulation by Genotype. When correctly recognizing an ‘old’ face versus correctly recognizing a ‘new’ face, Val/Val individuals show a more positive LPC component, whereas Met carriers show no significant difference in LPC amplitude. Val/Val’s also show a significantly more positive LPC for ‘old’ responses (asterisks label significances at p <.05).
Discussion
Consistent with previous work (Addante et al., 2012; Duzel et al., 2001) a significant increase in positivity in the FN400 component was found when participants successfully recognized a stimulus as ‘old’. However, the resulting increase in positivity did not differ significantly by genotype. Individuals with a copy of a Met allele showed the same increase in FN400 amplitude as Val/Val individuals. As there was no difference in behavioural scores for Familiarity, this suggests there is no effect of the BDNF polymorphism on this type of memory. In contrast, a significant difference between genotypes for LPC amplitude was found. Relative to Met carriers, Val/Val individuals showed a significantly larger increase in positivity of the LPC component when successfully identifying a stimuli as ‘old’ as opposed to ‘new’. Not only this, but Met carriers did not show a significant difference in LPC amplitudes. This suggests the BDNF polymorphism plays a part in this type of memory. This may be a subtle effect as no significant differences between genotypes was found in behavioural performance on Recollection.It should also be noted that for the Familiarity experiment, no LPC old/new difference was generated, and there was no FN400 modulation generated in the Recollection experiment. This is further support for the body of work suggesting that these paradigms index respectively, the familiarity and recollection forms of recognition memory.As there was no difference in FN400 modulation by genotype in the Familiarity experiment, this strongly suggests that the BDNF Val66Met polymorphism does not play a role in the type of memory. As shown in previous research, the FN400 ERP is associated with Familiarity forms of recognition memory (Addante et al., 2012; Duzel et al., 2001). This type of memory has been shown to be exclusive of the hippocampus (VarghaKhadem, Gadian, Watkins, Connelly, Van Paesschen, & Mishkin, 1997; Aggleton & Brown, 1999; Aggleton & Brown, 2006; Yonelinas et al., 2002; Davachi, Mitchell and Wagner, 2003). Whilst BDNF is widely spread throughout the CNS, it is most highly concentrated in the hippocampus (Pezawas et al., 2004). It is possible that this is the reason the BDNF polymorphism is not having an effect on Familiarity forms of memory.On the other hand, in the Recollection experiment, LPC modulation was significantly influenced by genotype, suggesting that the BDNF Val66Met polymorphism affects this ERP. Previous research has shown the LPC to be associated with Recollection forms of memory, particularly memory that is rich in contextual detail. As mentioned earlier, BDNF is most highly concentrated in the hippocampus (Pezawas et al., 2004). As BDNF plays an important role in long-term potentiation (LTP), and the implication of hippocampal LTP, the BDNF polymorphism may affect hippocampal dependent forms of memory via a disruption to LTP (Lu, 2003; Lu, Christian, & Lu, 2007; Poo, 2001; Egan et al., 2003; (Athos, Impey, Pineda, Chen, & Storm, 2002; Blum, Moore, Adams, & Dash.
Chapter 1: Introduction
1.1 Memory
1.2 Long Term Potentiation
1.3 Non-invasive LTP in humans
1.4 Brain derived neurotrophic factor
1.5 BDNF Val66Met polymorphism
1.6 Hypothesis/Aims
Chapter 2: BDNF Val66Met and LTP
2.1 Abstract
2.2 Introduction
2.3 Materials and Methods
2.4 Results
2.5 Discussion
Chapter 3: BDNF Val66Met and Recognition memory: EEG
3.1 Abstract
3.2 Introduction
3.3 Materials and Methods
3.4 Results
3.5 Discussion
Chapter 4: BDNF Val66Met and Recognition memory: fMRI
4.1 Abstract
4.2 Introduction
4.3 Materials and Methods
4.4 Results
4.5 Discussion
Chapter 5: General Discussion
5.1 BDNF Val66Met polymorphism and LTP
5.2 BDNF Val66Met polymorphism and Recognition memory: ERP evidence
5.3 BDNF Val66Met polymorphism and Recognition memory: fMRI evidence
5.4 More than LTP and Memory? Other considerations involving BDNF
5.5 Limitations, future directions and conclusions References
GET THE COMPLETE PROJECT
Effect of inter-cultural contact on L2 motivation and L2 learning: A processproduct study
