J Mol Neurosci (2014) 54:796–802 DOI 10.1007/s12031-014-0417-1
The Role of the BDNF Val66Met Polymorphism in Individual Differences in Long-Term Memory Capacity Christian Montag & Andrea Felten & Sebastian Markett & Luise Fischer & Katja Winkel & Andrew Cooper & Martin Reuter
Received: 23 May 2014 / Accepted: 27 August 2014 / Published online: 30 September 2014 # Springer Science+Business Media New York 2014
Abstract The protein brain-derived neurotrophic factor (BDNF) plays an important role in diverse memory processes and is strongly expressed in the hippocampus. The hippocampus itself is a key structure involved in the processing of information from short-term to long-term memory. Due to the putative role of BDNF in memory consolidation, a prominent single nucleotide polymorphism (SNP) on the BDNF gene (BDNF Val66Met) was investigated in the context of long-term memory performance. N=138 students were presented with 40 words from 10 categories, each consisting of eight words such as ‘fruits’ or ‘vehicles’ in a memory recognition task (specifically the Deese-Roediger-McDermott Paradigm). Recognition performance was analyzed 25 min after the initial presentation of the word list and subsequently 1 week after the initial presentation. Overall, individual longterm memory performance immediately after learning the word list (T1) and performance 1 week later (T2) did not differ on the basis of the BDNF SNP, but an interaction effect Christian Montag and Andrea Felten contributed equally to this work. C. Montag Department of Psychology, University of Ulm, Ulm, Germany C. Montag (*) : A. Felten : S. Markett : L. Fischer : K. Winkel : M. Reuter Department of Psychology, University of Bonn, Kaiser-Karl-Ring 9, 53111 Bonn, Germany e-mail:
[email protected] A. Felten : S. Markett : M. Reuter Laboratory of Neurogenetics, University of Bonn, Bonn, Germany A. Felten : S. Markett : M. Reuter Center for Economics and Neuroscience, University of Bonn, Bonn, Germany A. Cooper Department of Psychology, Goldsmiths, University of London, London, UK
of BDNF Val66Met by time-of-recall was found: Carriers of the Met66+ variant showed the strongest decline in hit rate performance over time. Keywords BDNF Val66Met . Long-term memory . Cognition . Individual differences
Introduction Long-term memory (LTM) refers to explicit and implicit memories lasting from days to decades (Polster et al. 1991). Without LTM, we would not be able to build and develop our life story, which is crucial for our identity and helps us to think using segments of time, such as yesterday, today and tomorrow (Squire and Zola-Morgan 1991). Although much is known about long-term potentiation processes in the hippocampus (Bird and Burgess 2008), which are of great importance for the transfer of information from short-term memory (STM) to LTM, less is known about genetic variants explaining individual differences in LTM performance. This is an important question because a twin study by Volk et al. (2006) suggests heritability estimates for individual differences in LTM range between 37 and 55 % (depending on the task under investigation). As already mentioned, it is widely accepted that the hippocampus plays an important role in transferring information from STM to LTM (Bliss and Collingridge 1993). However, the exact function of the hippocampus in the complicated processes of memory formation and retrieval of content from LTM still needs to be elucidated (Bird and Burgess 2008). As the hippocampus plays an important role in consolidation processes (Nadel and Bohbot 2001), a prominent genetic variant of the brain-derived neurotrophic factor (BDNF) gene has been investigated in the present study because BDNF is strongly expressed in this brain area while learning takes place
J Mol Neurosci (2014) 54:796–802
(e.g. Hall et al. 2000). Furthermore, BDNF has been associated with a wide range of processes, including neuronal growth, survival and synaptic plasticity (Groves 2007, Martinowitch et al. 2007). The BDNF Val66Met polymorphism (rs6265) under investigation is located on codon 66 in the 5′ pro domain of the BDNF gene on chromosome 11p14.1. The common rs6265 is a single-base mutation with an adenine instead of a guanine at position 196, resulting in the amino acid substitution Val66Met (Hall et al. 2003). The BDNF Val66Met polymorphism represents one of the most prominent single nucleotide polymorphism (SNPs) in human neuroscience because its functionality has been strongly supported: Carriers of at least one 66Met allele are associated with lower activity dependent BDNF secretion (Egan et al. 2003; Baj et al. 2013). Because of this, one would expect an effect of this polymorphism on circulating BDNF levels. However, there is no widespread agreement on the results of genetic association studies regarding a moderating influence of the BDNF Val66Met SNP on both plasma and brain BDNF levels. Moreover, the evidence from animal studies analyzing the relation between peripheral and central BDNF levels has shown contradictory results, making it more difficult to establish a relationship between the influence of the BDNF SNP in the brain and BDNF levels in peripheral areas (see Montag 2014 for a review). In the Caucasian population, carriers of the 66Met allele are less frequent than carriers of the Val allele and homozygosity for the Met allele is extremely rare (ca.Val66Met=27 % and Met66Met=3 %, e.g. Montag et al. 2010). Therefore, in most experimental studies, for statistical reasons, carriers of at least one Met allele are compared with participants who are homozygous for the Val allele. Evidence from structural imaging has demonstrated that carriers of the 66Met allele show smaller grey matter volume in several areas of the temporal lobe, including the hippocampus, adding to the BDNFhippocampus link (e.g. Pezawas et al. 2004, Montag et al. 2009). An early study by Egan et al. (2003) yielded the first evidence showing that BDNF Val66Met might also influence memory performance in a neuropsychological test. More specifically, their study showed that carriers of the rare Met66Met genotype showed impaired declarative memory, as measured by delayed recall on the revised version of the Wechsler Memory Scale (WMS-R). Since this pioneering work, a large number of studies have investigated the BDNF-memory relationship, pointing towards an association between Met allele carriers and poorer memory performance. For example, in a community of elderly Caucasian individuals, BDNF Val homozygotes performed significantly better compared to both Val/Met heterozygous individuals and Met homozygous individuals on a delayed recall task and an alphabet-coding task (ACT), the latter being a measure of processing speed (Miyajima et al. 2008). Furthermore, assessing cognitive performance with the
797
‘IntegNeuro’ neuro-psychological battery, homozygous Met allele carriers had higher verbal recall errors (Schofield et al. 2009). Ho et al. (2006) found that Met allele carriers showed poorer verbal memory performance, including immediate memory for stories and list learning of unrelated words. Replicating previous findings that Val allele carriers exhibit superior memory for semantically structured, verbally presented stories, Goldberg et al. (2008) showed that Val homozygotes had a significant advantage in recognizing previously presented words compared to Met allele carriers over delays of 30 min and 24 h in a verbal recognition memory task. All of these studies (including the study of Goldberg et al.) tested LTM performance in a relatively short time window after the initial learning test phase. We highlight this point because the present study aimed to investigate the role of BDNF Val66Met in LTM by including a much longer time window between the initial learning of items and the test of LTM. Beyond performance at different time points, a further critical variable in the investigation of LTM is the magnitude of the memory decline between any two given time points. In the present study, we were particularly interested in the course of memory decline. We come back to this point in the hypotheses below. An excellent recent review article by Hong et al. (2011) provides an overview of the studies discussed above in the domains of personality, affective disorders, cognition and dementia. This review also highlights the often heterogeneous results in the BDNF Val66Met research area. Given the focus of the present study on LTM functioning, a recent meta-analysis warrants particular attention. Here, Kambeitz et al. (2012) provided evidence that the 66Met-allele is associated with poorer performance in declarative memory. In contrast, the meta-analytic data on BDNF rs6265 and hippocampus structure/function are more heterogeneous. A recent comment on the meta-analysis by Kambeitz et al. (2012) put forward by Dodds et al. (2013a), pointed out several methodological issues, suggesting that ‘the hippocampal effect is likely to be much smaller than described in the meta-analysis.’ (p. 740). In sum, it has been extensively shown in human and animal studies that BDNF is necessary for memory processing, with implications for encoding, consolidation and retrieval, and more particularly, the BDNF Val66Met polymorphism seems to be associated with several facets of memory functioning, although its exact role is still a matter of debate (Dodds et al. 2013b; Bekinschtein et al. 2007, Lee et al. 2004). The aim of the present study was to investigate the influence of the BDNF Val66Met polymorphism on LTM performance. The present research adds to the literature because we tested LTM performance twice: 25 min after the learning of a word list and then again 1 week later. Therefore, in contrast to most of the earlier BDNF-memory studies, the present study design tests the role of BDNF Val66Met on LTM after a longer period of time has passed since the initial learning experiment.
798
In line with the meta-analysis by Kambeitz et al. (2012), we hypothesized that carriers of the Met66+ variant would be associated with worse LTM performance compared to carriers of the Val66Val genotype. Moreover, in contrast to the studies mentioned above, we administered a memory recognition task which is also able to investigate the influence of BDNF Val66Met on false memories. Using the Deese-RoedigerMcDermott (DRM) paradigm, participants were asked to learn words from categories, where the most typical word from that category was not presented. This strategy helps to implant false memories in later LTM recall. Thus, the present approach has the potential to highlight the role of BDNF Val66Met for individual differences in the distortion of memory processes. We did not specify a directional hypothesis with respect to the role of the BDNF SNP in individual differences in false memory because to our knowledge, no studies have dealt with false memories and BDNF before.
Materials and Methods Participants N=138 healthy participants were invited to a LTM recognition experiment. Twenty-six (18.8 %) of the participants were male and 112 (81.2 %) of the participants were female. This skewed gender distribution is representative of the gender ratio for the psychology students in Bonn/Germany who participated in the present study. Mean age was 22.76 years (SD=5.19; age range, 18–50 years). All participants provided buccal swabs for genotyping the BDNF Val66Met polymorphism (rs6265). Moreover, all participants provided informed written consent. The experiment was approved by the local ethics committee. Experimental Design Participants were instructed to learn a series of words presented on a computer screen. Forty words were presented in total, with each word presented once for 5 s. The order of words presented was randomized between participants. After presentation of the word list, participants had to work through another computer task (measuring the attentional blink, see also Felten et al. 2013), which took approximately 25 min. After this task, participants were presented with a pen and paper questionnaire consisting of 80 words (the 40 words initially presented and 40 new words) representing recall at time 1 (T1). The words presented in the initial word list and the additional words subsequently shown on the questionnaire belonged to a number of different categories, such as fruits, clothes or vehicles. In total, ten categories, consisting of eight words each, were presented on the questionnaire. A prototypical item for each category (such as ‘apple’ in the category
J Mol Neurosci (2014) 54:796–802
‘fruit’) was not presented in the initial word list and was therefore likely to produce false memories. The words used were all rated previously in a doctoral thesis with respect to their familiarity (Stegt 2006). One week (T2) after the initial presentation of the word list, the students were e-mailed with a link to an online version of the word list recognition task. Participants had to tick a box for all words they remembered as present in the initial word list, with no time limit given. The students were not told at the first test session that they would have to participate in the same task 1 week later; they were told that a link with a questionnaire would be sent to them in a week, but were not told what this questionnaire would measure. Due to an error in the online recall task (one item was listed twice), one of the ten categories was excluded from the analyses. The experiment constituted a version of the DRM (e.g. Roediger and McDermott 1995). The DRM paradigm can elicit false memories, remembering something that has not in fact been experienced. This phenomenon is important, e.g. for the reliability of legal testimony. For an extensive review of the DRM paradigm and its applications, see Gallo (2010). The experimental design is also depicted in Fig. 1. Genetic Analysis Genotyping of the BDNF Val66Met single nucleotide polymorphism was performed as described in the study of Montag et al. (2010). Statistical Analyses In order to investigate the effect of the BDNF Val66Met polymorphism on LTM performance, we computed repeated measure ANOVAs with performance at T1 and T2 as the within-subject factor. The time lag between answering the questionnaire (about 1 week, varying only slightly) and LTM performance did not correlate significantly with any of
Fig. 1 After the initial learning of a word list participants were asked to participate in a memory recognition task 25 min and 1 week after presentation of the word list
J Mol Neurosci (2014) 54:796–802
799
the dependent variables and, therefore, was not included as a covariate in the model. Varying time lags between T1 and T2 occurred because not every participant in the study responded immediately to the link sent via e-mail. Dependent variables in the repeated measures ANOVA were hits and false alarms at T1 and T2. Misses and correct rejections were not used as additional dependent variables because they offer no further information (the variable ‘misses’ depends on ‘hits’; the variable ‘correct rejections’ depends on ‘false alarms’); the analysis of hits and false alarms thus adequately covers the research questions in this study. More specifically, hits are described as correctly recognized items in the LTM task. False alarms are items that were recognized, but which did not appear in the initial word list. In addition, we split the false alarms into so-called critical and non-critical false alarm items according to their exemplar status for a category. The category ‘critical false alarm items’ consisted of only 9 items, namely the most exemplary item for each category.
Results Information on the Genotypes The genotype frequencies of BDNF Val66Met were as follows: Val/Val: n=93, Val/Met: n=40 and Met/Met: n=5. Genotype distribution is in Hardy-Weinberg Equilibrium in the total sample (BDNF: χ2 =0.07, df=1, ns) and in both gender groups (male: χ2 =0.44, df=1, ns; female: χ2 =0.69, df=1, ns). Age, Gender and (Long-Term) Memory Performance Age correlated significantly with hits at T1 (rHits =−.28, p=.001) and T2 (rHits =−.21, p=.01). Moreover, independent sample t tests revealed gender effects on hits at both time points T1 (tHits =−3.44, df=136, p=.001) and T2 (tHits = −2.68, df=136, p=.008). Here, women performed better than men. As a consequence, age was inserted as a covariate and gender as a further independent variable in the following
Table 1 Descriptive variables related to memory performance (means and SEMs) T1 (total) T2 (total) T1 (males) T2 (males) T1 (females) T2 (females)
analyses investigating T1 or T2 separately. Given the highly skewed gender distribution of the sample under investigation, the results of the statistical analyses with respect to gender have to be interpreted very cautiously. Despite this and on the basis that it may be useful for further research, we report means and SDs for males and females in Table 1. Due to the small number of males, we refrain from using gender as an independent variable in the repeated measures AN(C)OVA models dealing with BDNF and memory decline. False alarms at T1 or T2 were not significantly correlated with age or influenced by gender. Besides the analyses mentioned above, repeated measure ANOVAs were conducted to investigate the effect of BDNF Val66Met on a putative decline in long-term memory performance. The variable ‘hits at T1 minus hits at T2’ did not correlate with age nor was it influenced by gender. The variable ‘false alarms at T2 minus false alarms at T1’ did not correlate with age and was not influenced by gender. Investigation of the (Long-Term) Memory Variables All variables (hits and false alarms) showed a significant difference across T1 and T2. Repeated measure ANOVAs revealed a significant decline in hits (F(1,137) = 365.17, p<.001; Wilks Lambda=0.27) and a significant increase in false alarms (F(1,137) =152.41, p<.001; Wilks Lambda=0.47). Means and SEMs of these variables for T1 and T2 are presented in Table 1. BDNF and (Long-Term) Memory Variables ANOVAs investigating the influence of BDNF Val66Met on both T1 or T2 separately revealed no significant effects. However, a repeated measure ANOVA revealed an interaction effect between the BDNF 66Met±variant by time point of testing (T1/T2) on the hit rate. Carriers of at least one 66Met allele showed a significantly greater decline in performance from T1 to T2 (F(1,136) =7.77, p=.006, η2 =.05) (Fig. 2). Including age as a covariate in the model only changed the results slightly (F(1,135) =7.55, p=.007, η2 =.05). As hits at T1 are correlated with the decline across T1 and T2 (r=.48, p<.001), we also computed the repeated measures ANOVA with hits at T1 as a covariate. Again, the results changed very
Hits
False alarms
False alarms critical
False alarms non-critical
27.12 (0.53) 20.60 (0.48) 23.46 (1.26) 18.08 (1.09) 27.97 (0.56) 21.27 (0.52)
3.12 (0.20) 6.69 (0.35) 3.77 (0.49) 6.31 (0.77) 2.98 (0.22) 6.80 (0.40)
1.27 (0.09) 2.57 (0.13) 1.27 (0.23) 2.54 (0.27) 1.28 (0.11) 2.58 (0.16)
1.85 (0.15) 4.13 (0.26) 2.50 (0.45) 3.77 (0.59) 1.71 (0.16) 4.22 (0.29)
800
J Mol Neurosci (2014) 54:796–802
Fig. 2 Interaction effect between BDNF 66Met+/- and long-term memory consolidation (depicted are means and SEMs of hits)
little (F(1,135) =7.32, p=.008, η2 =.05). No significant influence of BDNF on false alarms (F(1,136) =0.72, p=.398, η2 =.005), including splitting up this variable into critical and non-critical items, was found. Mean values and number of participants per allelic group are shown in Table 2.
Discussion The present study aimed to further investigate the role of the BDNF Val66Met polymorphism in LTM functioning. More particularly, we focused on a specific facet of LTM, namely the recognition of learned word lists. We did not find a main effect of BDNF Val66Met on any of the variables when investigating hits and false alarms independently at T1 or T2 after the initial presentation of the learned word lists. However, a repeated measure ANOVA showed a significant result in the hypothesized direction: Carriers of the 66Met+ allele showed a stronger decline in LTM performance with respect to the ‘hit rate’-dependent variable. It could be hypothesized that the stronger decline in hit rate performance from T1 to T2 reflects poorer memory consolidation processes in the hippocampus or it may reflect a disturbed encoding process at an earlier stage while the word
list is presented. On a molecular level, the BDNF Val66Met polymorphism is known to influence the activity dependent BDNF secretion (Egan et al. 2003). Carriers of the 66Met allele are associated with lower BDNF secretion, which might result in lower synaptic neuroplasticity in neurons of the hippocampus (Chen et al. 2006). As a consequence of these individual differences in molecular processes, memory consolidation might also be affected. Our interpretation has to be judged cautiously, however, because this causal pathway cannot be tested directly. A combined genetic fMRI study dealing with long-term memory processes tested in the present study could shed some light on this hypothesized causal pathway. Recently, two fMRI studies showed that the 66Met allele is associated with lower hippocampus activation during episodic memory processing, thereby showing that the BDNF polymorphism investigated in this study is indeed closely linked to the hippocampal area (Hariri et al. 2003; Hashimoto et al. 2008). Nevertheless, an fMRI study with the number of participants tested in the current study (i.e. n=138, with a time lag of 1 week between the scan sessions) would be very difficult to achieve. The current study also provides some important insights into the molecular genetic underpinnings of false memories. No influence of the BDNF Val66Met on the performance of the Deese-Roediger-McDermott critical items could be observed. Future studies could use this study design to search for genetic influences on false memories, which could be due to ‘real’ false memories or false memories implanted by social conformity processes (Loftus 1996; Loftus 1997). A recent study by Deuker et al. (2013) provided evidence for a role of the COMT gene in memory and social conformity. Therefore, genetic variations of the COMT gene might be of interest to better understand individual differences in false memories. A strength of the present study is the extended period that elapsed between T1 and T2 of the LTM recognition tasks. Therefore, we are encouraged that key aspects of LTM processes could be investigated with our experimental design. According to Polster et al. (1991), LTM spans memory from days to decades. Following this, our task fits into this time frame of LTM. We would also note that the current sample size is relatively large, considering the experimental design, as discussed above. Nevertheless, we are aware that our present findings need to be replicated by other research groups. We hope that the present findings encourage other researchers to
Table 2 Descriptive statistics for the hit rate (means and SEMs) and number of participants (N) for the allelic configurations BDNF 66Met− and 66Met+ at both time points of recall Time point of recall
BDNF 66Met−
Number
BDNF 66Met+
Number
T1 T2
26.82 (0.642) 21.00 (0.557)
93 93
27.76 (0.961) 19.98 (0.901)
45 45
J Mol Neurosci (2014) 54:796–802
include longer intervals between the initial learning and the LTM test phase in their paradigms, in particular, in relation to BDNF Val66Met. A limitation of the current study is the narrow focus on one facet of LTM performance, namely recognition of learned word lists. Future studies will need to include other items in LTM recognition tasks, such as pictures. Given the influence of BDNF Val66Met on the processing of emotional pictures (e.g. Montag et al. 2008), an experimental manipulation of the emotional content of the presented items could yield interesting results. Another aspect to be mentioned refers to the different facets of LTM, which need to be investigated further (Ericsson and Kintsch 1995). Besides recognition performance in LTM, several other experimental areas, such as free recall performance in LTM, need to be investigated. Finally, we did not ask the participants how ‘sure’ they were about their answers. This could be added in a follow-up study to get a more fine-tuned picture of the influence of molecular genetic variables on memory processes.
References Baj G., Carlino D., Gardossi L. and Tongiorgi E. (2013) Towards a unified biological hypothesis for the BDNF Val66Metassociated memory deficits in humans: a model of impaired dendritic mRNA trafficking. Front. Neurosci., 7, 188.10. 3389/fnins.2013.00188 Bekinschtein P, Cammarota M, Igaz LM, Bevilaqua LR, Izquierdo I, Medina JH (2007) Persistence of long-term memory storage requires a late protein synthesis- and BDNF- dependent phase in the hippocampus. Neuron 53:261–277 Bird CM, Burgess N (2008) The hippocampus and memory: insights from spatial processing. Nat Rev Neurosci 9:182–94 Bliss TV, Collingridge GL (1993) A synaptic model of memory: longterm potentiation in the hippocampus. Nature 361:31–9 Chen ZY, Jing D, Bath KG et al (2006) Genetic variant BDNF (Val66Met) polymorphism alters anxiety-related behavior. Science 314:140–3 Deuker L, Müller AR, Montag C et al (2013) Playing nice: a multimethodological study on the effects of social conformity on memory. Front Hum Neurosci 7:79 Dodds CM, Henson RN, Miller SR, Nathan PJ (2013a) Overestimation of the effects of the BDNF val66met polymorphism on episodic memory-related hippocampal function: a critique of a recent metaanalysis. Neurosci Biobehav Rev 37:739–741 Dodds C, Lawrence P, Maltby K, Skeggs A, Miller S, et al. (2013b) Effects of the BDNF Val66Met polymorphism and met allele load on declarative memory related neural networks. Plos One 8(11):e74133.. Egan MF, Kojima M, Callicott JH et al (2003) The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 112:257–269 Ericsson KA, Kintsch W (1995) Long-term working memory. Psychol Rev 102:211–45 Felten A, Montag C, Kranczioch C, Markett S, Walter NT, Reuter M (2013) The DRD2 C957T polymorphism and the attentional
801 blink—a genetic association study. Eur Neuropsychopharmacol 23:941–947 Gallo DA (2010) False memories and fantastic beliefs: 15 years of the DRM illusion. Mem Cogn 38:833–848 Goldberg TE, Iudicello J, Russo et al (2008) BDNF Val66Met polymorphism significantly affects d’ in verbal recognition memory at short and long delays. Biol Psychol 77:20–4 Groves JO (2007) Is it time to reassess the BDNF hypothesis of depression? Mol Psychiatry 12:1079–1088 Hall J, Thomas KL, Everitt BJ (2000) Rapid and selective induction of BDNF expression in the hippocampus during contextual learning. Nat Neurosci 3:533–535 Hall D, Dhilla A, Charalambous A, Gogos JA, Karayiorgou M (2003) Sequence variants of the brain-derived neurotrophic factor (BDNF) gene are strongly associated with obsessive-compulsive disorder. Am J Hum Genet 73:370–376 Hariri AR, Goldberg TE, Mattay VS et al (2003) Brain-derived neurotrophic factor val66met polymorphism affects human memoryrelated hippocampal activity and predicts memory performance. J Neurosci 23:6690–6694 Hashimoto R, Moriguchi Y, Yamashita F et al (2008) Dose-dependent effect of the Val66Met polymorphism of the brain-derived neurotrophic factor gene on memory-related hippocampal activity. Neurosci Res 61:360–367 Ho BC, Milev P, O’Leary DS, Librant A, Andreasen NC, Wassink TH (2006) Cognitive and magnetic resonance imaging brain morphometric correlates of brain-derived neurotrophic factor Val66Met gene polymorphism in patients with schizophrenia and healthy volunteers. Arch Gen Psychiatry 63:731–40 Hong CJ, Liou YJ, Tsai SJ (2011) Effects of BDNF polymorphisms on brain function and behavior in health and disease. Brain Res Bull 86: 287–297 Kambeitz JP, Bhattacharyya S, Ilankovic LM, Valli I, Collier DA (2012) Effect of BDNF Met66Val-Polymorphism on declarative memory and its neural substrate: a meta-analysis. Neurosci Biobehav Rev 36: 2165–77 Lee JL, Everitt BJ, Thomas KL (2004) Independent cellular processes for hippocampal memory consolidation and reconsolidation. Science 304:839–843 Loftus EF (1996) Memory distortion and false memory creation. Bull Am Acad Psychiatry Law 24:281–95 Loftus EF (1997) Creating false memories. Sci Am 277:70–75 Martinowitch K, Husseini M, Bai L (2007) New insights into BDNF function in depression and anxiety. Nat Neurosci 10: 1089–1093 Miyajima F, Ollier W, Mayes A et al (2008) Brain-derived neurotrophic factor polymorphism Val66Met influences cognitive abilities in the elderly. Genes Brain Behav 7:411–7 Montag C. (2014) The brain derived neurotrophic factor and personality. Adv Biol, Article ID 719723,http://dx.doi.org/10. 1155/2014/719723 Montag C, Reuter M, Newport B, Elger C, Weber B (2008) The BDNF Val66Met polymorphism affects amygdala activity in response to emotional stimuli: evidence from a genetic imaging study. NeuroImage 42:1554–1559 Montag C, Basten U, Stelzel C, Fiebach CJ, Reuter M (2010) The BDNF Val66Met polymorphism and anxiety: support for animal knock-instudies from a genetic association study in humans. Psychiatry Res 179:86–90 Montag C, Weber B, Fliessbach K, Elger C, Reuter, M. (2009). The BDNF Val66Met polymorphism impacts parahippocampal and amygdala volume in healthy humans: incremental support for a genetic risk factor for depression. Psychological medicine, 39(11): 1831–1839. Nadel L, Bohbot V (2001) Consolidation of memory. Hippocampus 11: 56–60
802 Pezawas L, Verchinski BA, Mattay (2004) The brain-derived neurotrophic factor val66met polymorphism and variation in human cortical morphology. J Neurosci 24:10099–10102 Polster MR, Nadel L, Schacter DL (1991) Cognitive neuroscience analyses of memory: a historical perspective. J Cogn Neurosci 3:95–116 Roediger HL, McDermott KB (1995) Creating false memories: remembering words not presented in lists. J Exp Psychol Learn Mem Cogn 21:803–814 Schofield PR, Williams LM, Paul RH et al (2009) Disturbances in selective information processing associated with the
J Mol Neurosci (2014) 54:796–802 BDNF Val66Met polymorphism: evidence from cognition, the P300 and fronto-hippocampal systems. Biol Psychol 80: 176–88 Squire LR, Zola-Morgan S (1991) The medial temporal lobe memory system. Science 253:1380–1386 Stegt S (2006) Die Rolle impliziter assoziativer Reaktionen bei der Entstehung von Pseudoerinnerungen im DRM-Paradigma. Logos, Berlin Volk HE, McDermott KB, Roediger HL 3rd, Todd RD (2006) Genetic influences on free and cued recall in long-term memory tasks. Twin Res Hum Genet 9:623–31