Klingberglab.se

Developmental Psychology
Dopamine, Working Memory, and Training Induced
Plasticity: Implications for Developmental Research
Stina Söderqvist, Sissela Bergman Nutley, Myriam Peyrard-Janvid, Hans Matsson, Keith
Humphreys, Juha Kere, and Torkel Klingberg Online First Publication, November 21, 2011. doi: 10.1037/a0026179 Söderqvist, S., Bergman Nutley, S., Peyrard-Janvid, M., Matsson, H., Humphreys, K., Kere, J., & Klingberg, T. (2011, November 21). Dopamine, Working Memory, and Training Induced Plasticity: Implications for Developmental Research. Developmental Psychology. Advance online publication. doi: 10.1037/a0026179 2011 American Psychological Association Dopamine, Working Memory, and Training Induced Plasticity: Stina Söderqvist, Sissela Bergman Nutley, Karolinska Institutet and University of Helsinki Cognitive deficits and particularly deficits in working memory (WM) capacity are common featuresin neuropsychiatric disorders. Understanding the underlying mechanisms through which WMcapacity can be improved is therefore of great importance. Several lines of research indicate thatdopamine plays an important role not only in WM function but also for improving WM capacity. Forexample, pharmacological interventions acting on the dopaminergic system, such as methylpheni-date, improve WM performance. In addition, behavioral interventions for improving WM perfor-mance in the form of intensive computerized training have recently been associated with changes indopamine receptor density. These two different means of improving WM performance—pharmacological and behavioral—are thus associated with similar biological mechanisms in thebrain involving dopaminergic systems. This article reviews some of the evidence for the role ofdopamine in WM functioning, in particular concerning the link to WM development and cognitiveplasticity. Novel data are presented showing that variation in the dopamine transporter gene (DAT1)influences improvements in WM and fluid intelligence in preschool-age children following cognitivetraining. Our results emphasize the importance of the role of dopamine in determining cognitiveplasticity.
Keywords: working memory, fluid intelligence, dopamine, plasticity, cognitive training Working memory (WM) is the ability to manipulate and keep performance (Alloway, Gathercole, Kirkwood, & Elliott, 2009; task relevant information in mind for a short period of time. This Gathercole, Brown, & Pickering, 2003; Kane et al., 2004). These is important for reasoning, which typically involves several steps observations emphasize the importance of understanding the basis of planning and execution. WM deficits are commonly observed in of WM function, its development, and plasticity.
several neuropsychiatric disorders occurring during development,such as attention-deficit/hyperactivity disorder (ADHD; Castella- Dopamine and Working Memory
nos & Tannock, 2002; Martinussen, Hayden, Hogg-Johnson, &Tannock, 2005). WM capacity is also strongly associated withgeneral intellectual ability and is a predictor for later academic Evidence From Primates
The link between WM and dopamine has been investigated at a cellular level by studying neurons exhibiting memory fields in theprefrontal cortex (PFC) of monkeys. These neurons are believed to Stina Söderqvist, Sissela Bergman Nutley, and Torkel Klingberg, De- be the cellular basis for visuospatial WM as they are specifically partment of Neuroscience and Stockholm Brain Institute, Karolinska In- active in response to distinct spatial locations of a stimulus and are stitutet, Stockholm, Sweden; Myriam Peyrard-Janvid and Hans Matsson,Department of Biosciences and Nutrition at Novum, Karolinska Institutet, also active during the delay period between stimulus presentation Huddinge, Sweden; Keith Humphreys, Department of Medical Epidemi- and response. Dopamine D1 receptor antagonists enhance the ology and Biostatistics, Karolinska Institutet, Stockholm, Sweden; Juha response of these neurons (Williams & Goldman-Rakic, 1995).
Kere, Department of Biosciences and Nutrition at Novum, Karolinska This effect seems restricted to the D1 receptor, as no effect of a D2 Institutet, Huddinge, Sweden; Science for Life Laboratory Stockholm, agonist was observed in the same study. Effects were also dose Karolinska Instiutet, Stockholm, Sweden; Department of Medical Genet- dependent and specific to the neurons displaying memory field ics, Haartman Institute, University of Helsinki, Helsinki, Finland.
properties. D1 receptor stimulation can also lead to long-term Torkel Klingberg and Sissela Bergman Nutley are both main inventors improvements in WM performance, as observed in rhesus mon- of the non-verbal reasoning training program that has been submitted in an keys with either age-related (Castner & Goldman-Rakic, 2004) or international patent application, financed by Pearson Assessment.
drug-induced (Castner, Williams, & Goldman-Rakic, 2000) cog- Correspondence concerning this article should be addressed to Stina Söderqvist, Karolinska Institutet, Department of Neuroscience, Retzius va¨g nitive impairments. The specific mechanism by which D1 receptor 8, 171 77 Stockholm, Sweden. E-mail: stina.soderqvist@ki.se manipulation acts to enhance visuospatial WM performance has been suggested to occur by spatial tuning of neurons through dopamine D1 receptor binding potential in humans 10 to 30 years decreasing the neurons’ response to nonpreferred directions in a of age (Jucaite, Forssberg, Karlsson, Halldin, & Farde, 2010).
spatial WM task (Vijayraghavan, Wang, Birnbaum, Williams, & Interestingly, no age effects were observed in the posterior parietal Arnsten, 2007). In conclusion, research on primates highlights the cortex for the age range studied, possibly reflecting earlier devel- importance of cortical dopamine D1 receptor activity for WM opmental changes in this region as suggested by Goldman-Rakic and Brown (1982). For other areas, including the frontal, anteriorcingulate, and occipital cortices, an average decrease of 26% in D1 Human Pharmacological Evidence
binding potential was observed during adolescence. This level ofdecrease is comparable with that observed throughout the entire Evidence for dopamine’s involvement in WM functioning in humans stems from pharmacological research, where a distinctionbetween effects of D1 and D2 receptors has also been reported.
Dopamine and Development of WM Capacity
Whereas the D1/D2 receptor agonist pergolide improves perfor-mance on WM tasks, the D2 receptor agonist bromocriptine does The effect of the gene coding for the catechol-O- not, suggesting a predominate effect of D1 over D2 receptors for methyltransferase (COMT) enzyme has been extensively studied WM performance (Mu¨ller, von Cramon, & Pollmann, 1998), con- in the context of both typical and atypical cognitive development.
sistent with the literature on primates. Further studies have dem- The COMT enzyme is important for the degradation of cat- onstrated pergolide’s effects to be dependent on baseline perfor- echolamines, such as dopamine, especially within the neocortex mance, with high-performing individuals benefiting more from the (Matsumoto et al., 2003) and is hypothesized to be of particular treatment than lower performers (Kimberg & D’Esposito, 2003).
importance for tasks relying on PFC functioning. The Val158Met However, there is inconsistent evidence regarding the effect of D2 polymorphism of the COMT gene has been associated with WM receptors and some suggestion that bromocriptine influences WM function in both adults and children, independently and in inter- performance in participants with low baseline WM performance action with the DRD2 gene, coding for the dopamine D2 receptor (Kimberg, D’Esposito, & Farah, 1997). The effect of D2 agonists (Stelzel, Basten, Montag, Reuter, & Fiebach, 2009; Xu et al., on WM has also been suggested to be domain specific, improving 2007). Two studies (Barnett, Heron, Goldman, Jones, & Xu, 2009; performance only on a spatial WM task (Luciana & Collins, 1997).
Wahlstrom et al., 2007) have observed associations between WM Methylphenidate is a psychostimulant drug that is commonly performance and COMT polymorphisms in children and adoles- used to alleviate symptoms of ADHD by blocking reuptake of cents. The effects were found to be curvilinear, with an optimal dopamine and norepinephrine, thus increasing their availability in level of expression being beneficial, but either too much (Val/Val the synapse (Patrick, Caldwell, Ferris, & Breese, 1987; Solanto, genotype) or too little (Met/Met genotype) enzymatic activity 1998). In humans, methylphenidate improves WM performance, having negative effects on performance. Furthermore, a longitudi- specifically in the visuospatial domain (Mehta et al., 2000). Fur- nal study of typically developing children and adolescents dem- thermore, the drug effect also depends on baseline WM capacity, onstrated a developmental dependency of the COMT genotype with greater benefits observed in participants with lower baseline effect on WM performance (Dumontheil et al., 2011). It was performance (Mehta, Goodyer, & Sahakian, 2004; Mehta et al., shown that, whereas the Val-allele tended to be associated with superior performance on a visuospatial WM task in younger ages In summary, human pharmacological studies provide further (6 –10 years), the Met allele was beneficial after the age of 10. This evidence for the importance of the dopamine neurotransmitter is in line with the tendency for adult Met carriers to show better system in WM function. However, the specific roles of task WM performance than Val carriers, although evidence is currently modality, receptor subtypes, and their interaction with subjects’ inconsistent (for a review, see Dickinson & Elvevag, 2009). These baseline WM capacity need further investigation to provide clarity differences in the effect of COMT observed across development to the somewhat conflicting evidence.
are consistent with the developmental changes reported in dopa-mine concentrations discussed earlier, with changes in basal levels Dopamine During Development
of dopamine affecting optimal levels of dopamine degradation(Wahlstrom et al., 2007).
The dopamine system is believed to undergo numerous changes Another genetic variant that has been associated with typically during development. In rhesus monkeys, dopamine concentration developing WM function is a variable number tandem repeat changes during development, with patterns of change being region (VNTR) polymorphism located in the 3-untranslated region of the specific (Goldman-Rakic & Brown, 1981, 1982). Lower dopamine dopamine transporter (DAT1) gene. This polymorphism is believed concentrations are observed in somatosensory and posterior pari- to be involved in the expression of the gene, with higher expres- etal cortices than in the frontal lobe. In these regions, a more rapid sion being associated with the 10-repeat allele in vitro (Fuke et al., decrease in density is also observed, reaching adult levels of 2001; Mill, Asherson, Browes, D’Souza, & Craig, 2002). In chil- concentration already at 5 months of age. In contrast, dopamine dren ages 7 to 12 years, 9/10-repeat heterozygosity has been levels in the prefrontal areas follow a nonlinear developmental associated with better WM performance than 10-repeat homozy- pattern. At birth, levels of dopamine in prefrontal areas are high and similar to adult levels. These levels then decrease significantly A VNTR polymorphism located in exon 3 of the dopamine during the first 6 months of life before once again increasing to receptor 4 gene, DRD4, has been associated with WM perfor- reach adult levels (Goldman-Rakic & Brown, 1982). A recent mance in young children (Froehlich et al., 2007), with carriers of human positron emission tomography (PET) study investigated the 7-repeat allele showing lower performance on a spatial WM task. Furthermore the polymorphism was found to interact with the been related to changes in activity in the caudate nucleus and deteriorating effects of lead levels, such that children not carrying prefrontal and parietal cortices (E. Dahlin, Neely, Larsson, Back- the 7-repeat allele were most severely affected by lead levels and man, & Nyberg, 2008; Olesen, Westerberg, & Klingberg, 2004).
showed worse cognitive performance. Also, activity in the PFC For a review and discussion, see Klingberg (2010).
during performance on an N-back WM task has been associatedwith the 7-repeat allele, with young adults carrying the 7-repeat Dopamine and Cognitive Training
allele showing a larger difference in brain activation as a result oftask difficulty (Herrmann et al., 2007). In summary, the 7-repeat is One study has used PET to investigate the association between associated with poorer WM performance and what can be inter- cognitive training and dopamine D1 and D2 receptor density preted as more ineffective brain activity. This genetic evidence (McNab et al., 2009). Participants, who were all men in early suggests that several genes controlling dopamine levels and sig- adulthood, underwent a 5-week WM training scheme, training for naling have an influential effect on WM during development.
35 min 5 days a week and were scanned using functional and These effects on WM are likely to be complicated by Gene ϫ structural magnetic resonance imaging and PET before and after Gene interactions and factors affecting basal dopamine levels, such training. The functional magnetic resonance imaging was used to identify regions showing WM-related activity. These regionsguided PET analyses, investigating changes of dopamine D1 re- Dopamine and Plasticity
ceptor binding potential in cortical regions and D2 receptor bind-ing potential in subcortical regions. There were no significant Dopamine has been suggested to be important for plasticity by associations between D1 or D2 receptor binding potential and WM enhancing neural sprouting and synaptogenesis (Stroemer, Kent, & performance at baseline, although a trend was observed for cortical Hulsebosch, 1998). In stroke patients, treatment with stimulant D1. Notably, improvements in WM capacity observed after train- medication, which increases dopamine concentrations in the syn- ing were significantly related to changes in D1 receptor binding apse, enhances motor recovery resulting from physiotherapy potential. Furthermore, fitting a quadratic model significantly in- (Scheidtmann, Fries, Muller, & Koenig, 2001; Walker-Batson, creased the variance explained, suggesting an inverted-U-shaped Smith, Curtis, Unwin, & Greenlee, 1995). These effects are long relationship, in line with previous primate and human research lasting, with significant improvements remaining up to 12 months findings discussed earlier. No relation between improvements in WM capacity and D2 receptor binding potential was found. The The effects of behavioral parenting interventions have also been results provide further evidence for a greater importance of the D1 linked to dopaminergic function, with outcomes associated with receptor compared with the D2 receptor not only for WM func- variants of a VNTR in the DRD4 gene (Bakermans-Kranenburg, tioning but also for cognitive plasticity. Note, however, the recent Van, Pijlman, Mesman, & Juffer, 2008). This evidence is of findings by Ba¨ckman et al. (2011), demonstrating changes in D2 particular interest as it suggests an interaction between the dopa- receptor binding potential in the striatum after 5 weeks of updating minergic system and environmental influences on behavioral training. Whether the differences between these two studies result from differences in types of training (purely updating or tasks notinvolving updating) needs further investigation. Differences might Cognitive Training as a Model for Human
also arise from methodological differences. As McNab et al.
Cognitive Plasticity
(2009) performed PET scanning during rest, changes observed inthis study reflect task-independent changes. On the other hand, During the past decade, there has been increasing interest in Ba¨ckman et al. (2011) used PET scanning during task performance improving cognitive functions through targeted training. Many and a ligand sensitive to endogenous DA release. The results might training programs have shifted focus from explicit training, teach- therefore also reflect changes that are task specific, such as dif- ing strategies to improve performance, to implicit training that ferences in behavior during scanning.
involves repetition practice, feedback, and gradually increasing the Brehmer et al. (2009) found additional evidence for involvement cognitive load required to solve the training tasks (Klingberg, of the dopaminergic system in training induced plasticity from 2010). Implicit training of WM has been shown to improve per- genetic analyses in young adults who completed a WM training formance on nontrained WM tasks, reflecting a true increase in program similar to the program used by McNab et al. (2009). A WM capacity (Klingberg et al., 2005; Klingberg, Forssberg, & sample of 29 young adults were genotyped for the DAT1 VNTR Westerberg, 2002). Some studies report improvements in non- and were grouped according to 10-repeat homozygosity or trained cognitive functions associated with WM, such as attention, 9-repeat carriership (heterozygotes or homozygotes for the reading comprehension, mathematical ability, and fluid reasoning 9-repeat allele) while controlling for the COMT Val158Met poly- (K. Dahlin, 2011; Holmes, Gathercole, & Dunning, 2009; Jaeggi, morphism. No significant effects of the DAT1 genotype on base- Buschkuehl, Jonides, & Perrig, 2008; Klingberg et al., 2005; line performance on tasks measuring WM, attention, and fluid Klingberg et al., 2002). WM training can also have positive effects intelligence (Gf) were observed. An initial superior (but nonsig- on symptoms of ADHD and cognitive performance after stroke nificant) performance on the visuospatial WM tasks used in train- (Klingberg et al., 2005; Klingberg et al., 2002; Westerberg et al., ing was observed in 9/10-repeat carriers compared with 10-repeat 2007). This has obvious potential benefits for other clinical pop- carriers. There was a trend for this difference to increase through- ulations as well, and this area of research is currently growing out the training period with 10-repeat carriers increasing their rapidly. WM training is also related to changes in brain activity.
performance more steeply than noncarriers throughout the training For example, improvements in WM observed after training have period. No training related effects were observed for the COMT genotype, nor were there any significant effects of either genotype on the verbal WM tasks being trained. In summary, the effect ofthe DAT1 genotype was apparent across the training period, al- Participants
though not seen at baseline. This pattern of results suggests that the DNA was available from 96 children ages 4.0 to 4.5 years (56 genotype has an effect on susceptibility to training induced im- boys and 40 girls; mean age ϭ 51.2 months, SD ϭ 3.0) who had provements per se (E. Dahlin et al., 2008). It is important to note completed a 5-week training scheme. Participants were recruited that this study was underpowered for analyzing genetic effects that with flyers distributed at preschools in the local Stockholm area tend to be particularly small for complex traits, such as cognitive and by advertisements in the local newspaper and on the lab functions, and because the results were nonsignificant, a replica- website. The children received a small reward (a toy) after com- pletion of each set of 5 days of training. After completion of the In the present study, we investigated the effects of polymor- whole training period (a minimum of 20 completed training days), phisms in five genes involved in the dopaminergic system (Table participants received an additional monetary reward. The study 1) on the effectiveness of WM and nonverbal reasoning (NVR) was approved by the local ethics committee, and informed consent training in preschool children. A training program consisting of was collected from the guardians of all participants (for further NVR tasks was designed to assess the feasibility of improving details, see Bergman Nutley et al., 2011).
fluid intelligence (Gf) (Bergman Nutley et al., 2011). Gf is referredto as the ability to identify patterns and relations and infer rules for Training Program and Procedure
novel problems (Horn & Cattell, 1966). Gf is independent frompreviously learned knowledge, is highly correlated with WM ca- Participants were pseudorandomly assigned (stratifying for sex) pacity, and similar to WM, is a predictor of academic performance to one of four different training programs: WM training, NVR (Alloway & Alloway, 2010; Lynn, Meisenberg, Mikk, & Wil- training, a CB training of WM and NVR, and a placebo training liams, 2007). Deficits in Gf, in combination with deficits in other designed as the CB training but with task difficulty kept constant types of intelligence, are core symptoms of mental retardation, a at the lowest levels throughout the training period. The WM group particularly common mental disorder with a prevalence of around trained with a program developed by Cogmed Systems containing 3% (Roeleveld, Zielhuis, & Gabreels, 1997). It would therefore be seven different versions of visuospatial WM tasks. Training tasks of great potential benefit if this function could be improved with used for the NVR training program were based on three tests from training similar to that used for WM. Bergman Nutley et al. (2011) the Leiter Battery that load highly on Gf: Repeated Patterns, recently showed that Gf could be enhanced in 4-year-old children.
Sequential Order, and Classification (Roid & Miller, 1997). Task Each child was randomly assigned to train WM, to train NVR, or difficulty was carefully assessed to allow for an automatic gener- to train a combination (CB) of the two. For comparison, a placebo ation of multiple items of varying difficulty. The WM, NVR, and group was also included. Training of WM (in both WM and CB CB training programs were adaptive, with level of difficulty au- groups) resulted in significantly improved performance on WM tomatically adjusted according to each child’s performance. Train- transfer tasks (i.e., nontrained tasks) compared with the placebo ing took place in the home and lasted approximately 15 min per group. Training of NVR led to significantly improved performance day 5 days per week until 25 sessions had been performed (Berg- on transfer tasks of Gf, and this group also showed a trend toward transfer between constructs with improvements on a measure of Transfer Tests
visuospatial WM. On the basis of the evidence so far, we inves-tigated here whether these improvements were associated to poly- Gf was measured with the Repeated Patterns, Sequential Orders, morphisms in some genes related to the dopaminergic system.
and Classifications subtests from the Leiter Battery (Roid & Table 1Eleven SNPs From Five Genes Involved in Dopamine-Related Pathways a Alleles of the corresponding single nucleotide polymorphism (SNP) with the minor allele first. b Minor allele frequency (MAF) in our sample set andcorresponding allele frequency in the HapMap CEU data set in parenthesis (NA ϭ no frequency available). c Chromosomal location of the gene basedon ideogram with chromosome number and band.
d SNP position, in base pair (bp), on the respective chromosome (Genome Assembly Build 36.3).
Miller, 1997), Raven’s Coloured Progressive Matrices (Raven, (training or no training) to assess main effect of training, and an 1998), and Block Design from the Wechsler Preschool and Pri- interaction term of genotype and training factor to assess the effect mary Scale of Intelligence–Third Edition (Wechsler, 2004). To of genotype on the influence of training. Person ability was entered assess WM capacity, we used a visuospatial grid task (Bergman as a random effect to account for within-person correlation. The Nutley, Söderqvist, Bryde, Humphreys, & Klingberg, 2009; West- strongest association was found for the Gf latent variable and one erberg, Hirvikoski, Forssberg, & Klingberg, 2004), the Odd One SNP (rs27072, T/C) from the DAT1 gene significantly interacting Out task from the Automated Working Memory Assessment (Al- with the training factor, F(4, 91) ϭ 6.971, p ϭ .01. The T-allele of loway, 2007), and the Word Span test, a verbal WM test similar to rs27072 seemed to be advantageous, as carriers showed a larger the Digit Span Forward subtest from the Wechsler Intelligence training gain than noncarriers on the Gf factor (see Figure 1). Two Scale for Children–Third Edition (Wechsler, 1991) but with unre- additional SNPs from the DAT1 gene also showed significant lated nouns instead of numbers (Thorell & Wåhlstedt, 2006). For associations. One, rs40184, was found to associate with the Gf the measures of WM and Gf, respectively, the three measures were latent factor, F(6, 93) ϭ 3.445, p ϭ .036, whereas another, modeled as independent linear functions of a continuous latent rs3863145, showed significant association with the WM latent factor, F(6, 96) ϭ 3.545, p ϭ .032. No other polymorphismsshowed significant training interaction effects (Table 2), and no Genes and SNPs Studied
genotype effects were observed on baseline performance. All pvalues are uncorrected for multiple comparisons of the 11 SNPs Genetic data for the current sample were available as a subset of tested. Thus, results should be considered preliminary and in need a previous larger study (see Söderqvist et al., 2010). Genotypes of replication in a larger sample set.
from single nucleotide polymorphisms (SNPs) in five genes be-longing to the dopamine pathway were available and were here Discussion
tested for association with training performance. The genes weredopamine receptors D4 (DRD4) and D5 (DRD5), solute carrier Previous literature suggests that dopamine might not be impor- family 6 member 3 (SLC6A3), also known as dopamine transporter tant only for performance on WM tasks and other cognitive tasks (DAT1), dopamine beta-hydroxylase (DBH), and COMT. Table 1 relying on prefrontal and parietal function but also for its plasticity.
summarizes the SNPs genotyped in these genes, their chromo- Computerized cognitive training, such as WM training (Klingberg somal location and base pair positions according to Genome Build et al., 2002, 2005) and reasoning training (Bergman Nutley et al., 36.3. We selected SNP markers available from the HapMap Ge- 2011), could be useful not only for rehabilitation, but also as a nome Browser to analyze genotype–phenotype association using method of studying cognitive plasticity in humans, as highlighted the genes described here. For the dopamine receptors DRD4 and by our study. We investigated the role of genetic polymorphisms DRD5, SNPs were selected so that they tagged the complete gene on the change in cognitive performance resulting from training and regions. The SNPs genotyped are in strong linkage disequilibrium found that polymorphisms of the DAT1 gene are associated with with the previously reported VNTR and Taq1 restriction site training effects. As cognitive training increases in popularity, a polymorphism for the SLC6A3/DAT1 and DBH genes, respec-tively. The VNTR markers reported in previous studies were notused for genotyping as the SNP genotyping technology availableallowed for multiplexing of a large number of SNPs in onereaction, which was the method of choice for screening a largenumber of genes. All samples had a genotyping success rate ofgreater than 95% and a genotype call rate of greater than 80%. Themethods for blood and saliva sampling, genomic DNA extraction,and SNP genotyping are described in (Söderqvist et al., 2010).
Genetic Effect on Transfer
The 11 SNPs were included (separately) as fixed effects in mixed effects models. As the dependent variable we used either thesubject loading on the latent variable for WM (Grid task, Odd OneOut task, and Word Span test) or the subject loading on the latentvariable for Gf (Leiter Battery, Raven’s Coloured ProgressiveMatrices, and Block Design tests). Because of the small number ofparticipants, all active groups (WM, NVR, and CB) were collapsedinto one training group for these analyses. The independent fixedfactors were time (Time 1, before training, or Time 2, after Mean fluid intelligence (Gf) gains per group and genotype. Bars training) to account for change in performance related to baseline, show mean gains on the Gf latent variable for the collapsed trained groups and additive effect of genotype (0, 1, or 2 copies of a specific allele of placebo training (PLT) group and sorted by genotype on the DAT1 (rs27072) each SNP) to assess main effect of genotype, training factor with standard error of the mean. T1 ϭ Time 1; T2 ϭ Time 2.
ing on interactions with other genes and background factors, such Significance Levels (p Values) From Mixed Model Analyses for as age and baseline performance. Future studies should include All SNPs Tested for Main Effect of Genotype on Gf and WM larger samples to allow such interactions to be analyzed. It would Performance as Well as Training ϫ Genotype Interactions also be of interest to investigate the importance of domain in whichthe cognitive training is performed. As discussed earlier, there have been suggestions that dopamine function is of particular importance for visuospatial WM function. In the current study, training included only visuospatial WM tasks. Thus, such a dis- tinction is not possible. Understanding domain-specific interac- tions with dopamine function would provide a good foundation for better understanding how cognitive training can be individualized to best suit people with different baseline capacities and perhaps It has been suggested that some neurodevelopmental disorders can be understood as disorders of learning, rather than a fixed cognitive deficit (Karmiloff-Smith, 1998). The capacity of the brain to adapt and learn in response to environmental influence isconsidered crucial for the development of cognitive functions.
Significant interactions indicate that training effects were associ- Thus, a general learning impairment can lead to more specific ated with genotype. Bold indicates p Ͻ .05 (uncorrected for multiple disruptions later in life. The data presented here suggest that comparisons). SNP ϭ single nucleotide polymorphism; Gf ϭ fluid intel- dopamine is important not only for cognitive performance but perhaps, in particular, for plasticity. The same genotype associatedwith lower plasticity in the present study is also associated with better understanding of the mechanisms underlying its effects is of ADHD (Gizer et al., 2009). Considering the importance of dopamine-related genotypes in many neurodevelopmental disor- The SNP with strongest association in our sample (rs27072) is ders, it will be interesting to further investigate to what extent these located in the 3Јuntranslated region of the gene and has previously effects can be explained through an influence on learning and been implicated in genetic studies of ADHD, with the carriers of plasticity rather than on permanent and fixed functions.
the C/G allele exhibiting higher risk for ADHD (for a review, seeGalili-Weisstub & Segman, 2003; Gizer, Ficks, & Waldman,2009). The same allele (T) found to be advantageous in terms of References
training gains in the present study has also been shown to have a Alloway, T. P. (2007). Automated Working Memory Assessment. Oxford, protective effect for ADHD (Brookes et al., 2006; Feng et al., 2005). Furthermore, for rs27072, allelic imbalance of DAT1 Alloway, T. P., & Alloway, R. G. (2010). Investigating the predictive roles mRNA expression has been observed, with the minor allele (T) of working memory and IQ in academic attainment. Journal of Exper- accounting for increased expression in both in vitro and in vivo imental Child Psychology, 106, 20 –29. doi:10.1016/j.jecp.2009.11.003 studies (Pinsonneault et al., 2011).
Alloway, T. P., Gathercole, S. E., Kirkwood, H., & Elliott, J. (2009). The Dopamine transporter protein removes dopamine from the syn- cognitive and behavioral characteristics of children with low working aptic cleft. Consequently, concentration of dopamine transporter memory. Child Development, 80, 606 – 621. doi:10.1111/j.1467- protein could influence the activity of both D1 and D2 receptors.
The sample size of the current study is considered small for Ba¨ckman, L., Nyberg, L., Soveri, A., Johansson, J., Andersson, M., Dahlin, analyses of genetic effects, and the effects observed do not remain E., . . . Rinne, J. O. (2011, August 5). Effects of working-memorytraining on striatal dopamine release. Science, 333, 718. doi:10.1126/ significant after correcting for multiple comparisons. Thus, the findings need further replication in larger independent samples.
Bakermans-Kranenburg, M. J., Van, I. M. H., Pijlman, F. T., Mesman, J., The low power is also a reason for caution when interpreting the & Juffer, F. (2008). Experimental evidence for differential susceptibility: lack of effects for other genotypes analyzed. It might, for example, Dopamine D4 receptor polymorphism (DRD4 VNTR) moderates inter- seem surprising that no effect was observed for the COMT vention effects on toddlers’ externalizing behavior in a randomized Val158Met polymorphism considering its extensively observed ef- controlled trial. Developmental Psychology, 44, 293–300. doi:10.1037/ fects on WM and other cognitive functions and its importance for degradation of dopamine in the PFC, one of the areas that show Barnett, J. H., Heron, J., Goldman, D., Jones, P. B., & Xu, K. (2009).
training related changes in activity. It is possible that analyzing a Effects of catechol-O-methyltransferase on normal variation in the cog- COMT haplotype would have provided stronger effects, as has nitive function of children. American Journal of Psychiatry, 166, 909 – been previously suggested in studies of pain perception and atten- Bergman Nutley, S., Söderqvist, S., Bryde, S., Humphreys, K., & Kling- tion (Diatchenko et al., 2005) and attention alone in 2-year-old berg, T. (2009). Measuring working memory capacity with greater children (Voelker, Sheese, Rothbart, & Posner, 2009). Currently precision in the lower capacity ranges. Developmental Neuropsychology, available data do not allow for such analyses, and we suggest that 35, 81–95. doi:10.1080/87565640903325741 haplotype analysis of genetic markers in COMT can be addressed Bergman Nutley, S., Söderqvist, S., Bryde, S., Thorell, L. B., Humphreys, in future studies. Furthermore, effects of dopaminergic genes are K., & Klingberg, T. (2011). Gains in fluid intelligence after training likely to follow complicated patterns across development depend- non-verbal reasoning in 4-year-old children: A controlled, randomized study. Developmental Science, 14, 591– 601. doi:10.1111/j.1467- monkeys. Neuroscience, 6, 177–187. doi:10.1016/0306-4522(81) Brehmer, Y., Westerberg, H., Bellander, M., Furth, D., Karlsson, S., & Goldman-Rakic, P. S., & Brown, R. M. (1982). Postnatal development of Backman, L. (2009). Working memory plasticity modulated by dopa- monoamine content and synthesis in the cerebral cortex of rhesus mon- mine transporter genotype. Neuroscience Letters, 467, 117–120. doi: keys. Developmental Brain Research, 256, 339 –349.
Herrmann, M. J., Walter, A., Schreppel, T., Ehlis, A. C., Pauli, P., Lesch, Brookes, K., Xu, X., Chen, W., Zhou, K., Neale, B., Lowe, N., . . .
K. P., & Fallgatter, A. J. (2007). D4 receptor gene variation modulates Johansson, L. (2006). The analysis of 51 genes in DSM–IV combined activation of prefrontal cortex during working memory. European Jour- type attention deficit hyperactivity disorder: Association signals in DRD4, DAT1 and 16 other genes. Molecular Psychiatry, 11, 934 –953.
Holmes, J., Gathercole, S. E., & Dunning, D. L. (2009). Adaptive training Castellanos, F. X., & Tannock, R. (2002). Neuroscience of attention- leads to sustained enhancement of poor working memory in children.
deficit/hyperactivity disorder: The search for endophenotypes. Nature Developmental Science, 12, F9 –F15. doi:10.1111/j.1467-7687.2009 Reviews Neuroscience, 3, 617– 628. doi:10.1038/nrn896nrn896 Castner, S. A., & Goldman-Rakic, P. S. (2004). Enhancement of working Horn, J. L., & Cattell, R. B. (1966). Refinement and test of the theory of memory in aged monkeys by a sensitizing regimen of dopamine D1 fluid and crystallized general intelligences. Journal of Educational Psy- receptor stimulation. Journal of Neuroscience, 24, 1446 –1450. doi: Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Im- Castner, S. A., Williams, G. V., & Goldman-Rakic, P. S. (2000, March 17).
proving fluid intelligence with training on working memory. Proceed- Reversal of antipsychotic-induced working memory deficits by short- ings of the National Academy of Sciences of the United States of term dopamine D1 receptor stimulation. Science, 287, 2020 –2022. doi: America, 105, 6829 – 6833. doi:10.1073/pnas.0801268105 Jucaite, A., Forssberg, H., Karlsson, P., Halldin, C., & Farde, L. (2010).
Dahlin, E., Neely, A. S., Larsson, A., Backman, L., & Nyberg, L. (2008, Age-related reduction in dopamine D1 receptors in the human brain: June 13). Transfer of learning after updating training mediated by the From late childhood to adulthood, a positron emission tomography striatum. Science, 320, 1510 –1512. doi:10.1126/science.1155466 study. Neuroscience, 167, 104 –110. doi:10.1016/j.neuroscience Dahlin, K. (2011). Effects of working memory training on reading in children with special needs. Reading and Writing, 24, 479 – 491. doi: Kane, M. J., Hambrick, D. Z., Tuholski, S. W., Wilhelm, O., Payne, T. W., & Engle, R. W. (2004). The generality of working memory capacity: A Diatchenko, L., Slade, G. D., Nackley, A. G., Bhalang, K., Sigurdsson, A., latent-variable approach to verbal and visuospatial memory span and Belfer, I., . . . Maixner, W. (2005). Genetic basis for individual variations reasoning. Journal of Experimental Psychology: General, 133, 189 – in pain perception and the development of a chronic pain condition.
Human Molecular Genetics, 14, 135–143. doi:10.1093/hmg/ddi013 Karmiloff-Smith, A. (1998). Development itself is the key to understanding Dickinson, D., & Elvevag, B. (2009). Genes, cognition and brain through a COMT developmental disorders. Trends in Cognitive Sciences, 2, 389 –398.
lens. Neuroscience, 164, 72–87. doi:10.1016/j.neuroscience.2009.05.014 Dumontheil, I., Roggeman, C., Ziermans, T., Peyrard-Janvid, M., Matsson, Kimberg, D. Y., & D’Esposito, M. (2003). Cognitive effects of the dopa- H., Kere, J., & Klingberg, T. (2011). Influence of the COMT genotype mine receptor agonist pergolide. Neuropsychologia, 41, 1020 –1027.
on working memory and brain activity changes during development.
Biological Psychiatry, 70, 222–229. doi:10.1016/j.biopsych Kimberg, D. Y., D’Esposito, M., & Farah, M. J. (1997). Effects of bromocriptine on human subjects depend on working memory capacity.
Feng, Y., Wigg, K. G., Makkar, R., Ickowicz, A., Pathare, T., Tannock, NeuroReport, 8, 3581–3585. doi:10.1097/00001756-199711100-00032 R., . . . Barr, C. L. (2005). Sequence variation in the 3Ј-untranslated Klingberg, T. (2010). Training and plasticity of working memory. Trends region of the dopamine transporter gene and attention-deficit hyperac- in Cognitive Sciences, 14, 317–324. doi:10.1016/j.tics.2010.05.002 tivity disorder (ADHD). American Journal of Medical Genetics: Neu- Klingberg, T., Fernell, E., Olesen, P. J., Johnson, M., Gustafsson, P., ropsychiatric Genetics, 139B, 1– 6. doi:10.1002/ajmg.b.30190 Dahlstrom, K., . . . Westerberg, H. (2005). Computerized training of Froehlich, T. E., Lanphear, B. P., Dietrich, K. N., Cory-Slechta, D. A., working memory in children with ADHD—A randomized, controlled Wang, N., & Kahn, R. S. (2007). Interactive effects of a DRD4 poly- trial. Journal of the American Academy of Child and Adolescent Psy- morphism, lead, and sex on executive functions in children. Biological chiatry, 44, 177–186. doi:10.1097/00004583-200502000-00010 Psychiatry, 62, 243–249. doi:10.1016/j.biopsych.2006.09.039 Klingberg, T., Forssberg, H., & Westerberg, H. (2002). Training of work- Fuke, S., Suo, S., Takahashi, N., Koike, H., Sasagawa, N., & Ishiura, S.
ing memory in children with ADHD. Journal of Clinical and Experi- (2001). The VNTR polymorphism of the human dopamine transporter mental Neuropsychology, 24, 781–791. doi:10.1076/jcen.24.6.781.8395 (DAT1) gene affects gene expression. Pharmacogenomics Journal, 1, Luciana, M., & Collins, P. F. (1997). Dopaminergic modulation of working memory for spatial but not object cues in normal humans. Journal of Galili-Weisstub, E., & Segman, R. H. (2003). Attention deficit and hyper- Cognitive Neuroscience, 9, 330 –347. doi:10.1162/jocn.1997.9.3.330 activity disorder: Review of genetic association studies. Israel Journal of Lynn, R., Meisenberg, G., Mikk, J., & Williams, A. (2007). National IQs Psychiatry and Related Sciences, 40, 57– 66.
predict differences in scholastic achievement in 67 countries. Journal of Gathercole, S. E., Brown, L., & Pickering, S. J. (2003). Working memory Biosocial Science, 39, 861– 874. doi:10.1017/S0021932007001964 assessments at school entry as longitudinal predictors of National Cur- Martinussen, R., Hayden, J., Hogg-Johnson, S., & Tannock, R. (2005). A riculum attainment levels. Educational and Child Psychology, 20, 109 – meta-analysis of working memory impairments in children with atten- tion-deficit/hyperactivity disorder. Journal of the American Academy of Gizer, I. R., Ficks, C., & Waldman, I. D. (2009). Candidate gene studies of Child and Adolescent Psychiatry, 44, 377–384. doi:10.1097/ ADHD: A meta-analytic review. Human Genetics, 126, 51–90. doi: Matsumoto, M., Weickert, C. S., Akil, M., Lipska, B. K., Hyde, T. M., Goldman-Rakic, P. S., & Brown, R. M. (1981). Regional changes of Herman, M. M., . . . Weinberger, D. R. (2003). Catechol monoamines in cerebral cortex and subcortical structures of aging rhesus O-methyltransferase mRNA expression in human and rat brain: Evi- dence for a role in cortical neuronal function. Neuroscience, 116, 127– Effects of dopamine-related gene– gene interactions on working memory component processes. European Journal of Neuroscience, 29, 1056 – McNab, F., Varrone, A., Farde, L., Jucaite, A., Bystritsky, P., Forssberg, 1063. doi:10.1111/j.1460-9568.2009.06647.x H., & Klingberg, T. (2009, February 6). Changes in cortical dopamine Stollstorff, M., Foss-Feig, J., Cook, E. H., Jr., Stein, M. A., Gaillard, W. D., D1 receptor binding associated with cognitive training. Science, 323, & Vaidya, C. J. (2010). Neural response to working memory load varies by dopamine transporter genotype in children. NeuroImage, 53, 970 – Mehta, M. A., Goodyer, I. M., & Sahakian, B. J. (2004). Methylphenidate 977. doi:10.1016/j.neuroimage.2009.12.104 improves working memory and set-shifting in AD/HD: Relationships to Stroemer, R. P., Kent, T. A., & Hulsebosch, C. E. (1998). Enhanced baseline memory capacity. Journal of Child Psychology and Psychiatry neocortical neural sprouting, synaptogenesis, and behavioral recovery and Allied Disciplines, 45, 293–305. doi:10.1111/j.1469-7610.2004 with D-amphetamine therapy after neocortical infarction in rats. Stroke, Mehta, M. A., Owen, A. M., Sahakian, B. J., Mavaddat, N., Pickard, J. D., Thorell, L. B., & Wåhlstedt, C. (2006). Executive functioning deficits in & Robbins, T. W. (2000). Methylphenidate enhances working memory relation to symptoms of ADHD and/or ODD in preschool children.
by modulating discrete frontal and parietal lobe regions in the human Infant and Child Development, 15, 503–518. doi:10.1002/icd.475 brain. Journal of Neuroscience, 20, RC65.
Vijayraghavan, S., Wang, M., Birnbaum, S. G., Williams, G. V., & Mill, J., Asherson, P., Browes, C., D’Souza, U., & Craig, I. (2002).
Arnsten, A. F. (2007). Inverted-U dopamine D1 receptor actions on Expression of the dopamine transporter gene is regulated by the 3Ј UTR prefrontal neurons engaged in working memory. Nature Neuroscience, VNTR: Evidence from brain and lymphocytes using quantitative RT- 10, 376 –384. doi:10.1038/nn1846 PCR. American Journal of Medical Genetics, 114, 975–979. doi: Voelker, P., Sheese, B. E., Rothbart, M. K., & Posner, M. I. (2009).
Variations in catechol-O-methyltransferase gene interact with parenting Mu¨ller, U., von Cramon, D. Y., & Pollmann, S. (1998). D1- versus to influence attention in early development. Neuroscience, 164, 121– D2-receptor modulation of visuospatial working memory in humans.
130. doi:10.1016/j.neuroscience.2009.05.059 Journal of Neuroscience, 18, 2720 –2728.
Wahlstrom, D., White, T., Hooper, C. J., Vrshek-Schallhorn, S., Oetting, Olesen, P. J., Westerberg, H., & Klingberg, T. (2004). Increased prefrontal W. S., Brott, M. J., & Luciana, M. (2007). Variations in the catechol and parietal activity after training of working memory. Nature Neuro- O-methyltransferase polymorphism and prefrontally guided behaviors in science, 7, 75–79. doi:10.1038/nn1165 adolescents. Biological Psychiatry, 61, 626 – 632. doi:10.1016/ Patrick, K. S., Caldwell, R. W., Ferris, R. M., & Breese, G. R. (1987).
Pharmacology of the enantiomers of threo-methylphenidate. Journal of Walker-Batson, D., Smith, P., Curtis, S., Unwin, H., & Greenlee, R.
Pharmacology and Experimental Therapeutics, 241, 152–158.
(1995). Amphetamine paired with physical therapy accelerates motor Pinsonneault, J. K., Han, D. D., Burdick, K. E., Kataki, M., Bertolino, A., recovery after stroke. Further evidence. Stroke, 26, 2254 –2259. doi: Malhotra, A. K., . . . Sadee, W. (2011). Dopamine transporter gene variant affecting expression in human brain is associated with bipolar Wechsler, D. (1991). Wechsler Intelligence Scale for Children–Third Edi- disorder. Neuropsychopharmacology, 36, 1644 –1655. doi:10.1038/ tion. New York, NY: Psychological Corporation.
Wechsler, D. (2004). Wechsler Preschool and Primary Scale of Raven, J. C. (1998). Manual for Raven’s Progressive Matrices. Oxford, Intelligence–Third Edition (WPPSI-III). New York, NY: Psychological United Kingdom: Oxford Psychologists Press.
Roeleveld, N., Zielhuis, G. A., & Gabreels, F. (1997). The prevalence of Westerberg, H., Hirvikoski, T., Forssberg, H., & Klingberg, T. (2004).
mental retardation: A critical review of recent literature. Developmental Visuo-spatial working memory span: A sensitive measure of cognitive Medicine and Child Neurology, 39, 125–132. doi:10.1111/j.1469- deficits in children with ADHD. Child Neuropsychology, 10, 155–161.
Roid, G. H., & Miller, L. J. (1997). Leiter International Performance Westerberg, H., Jacobaeus, H., Hirvikoski, T., Clevberger, P., Ostensson, Scale–Revised: Examiner’s manual. Wood Dale, IL: Stoelting.
M. L., Bartfai, A., & Klingberg, T. (2007). Computerized working Scheidtmann, K., Fries, W., Muller, F., & Koenig, E. (2001). Effect of memory training after stroke—A pilot study. Brain Injury, 21, 21–29.
levodopa in combination with physiotherapy on functional motor recov- ery after stroke: A prospective, randomised, double-blind study. Lancet, Williams, G. V., & Goldman-Rakic, P. S. (1995, August 17). Modulation 358, 787–790. doi:10.1016/S0140-6736(01)05966-9 of memory fields by dopamine D1 receptors in prefrontal cortex. Nature, Söderqvist, S., McNab, F., Peyrard-Janvid, M., Matsson, H., Humphreys, 376, 572–575. doi:10.1038/376572a0 K., Kere, J., & Klingberg, T. (2010). The SNAP25 gene is linked to Xu, H., Kellendonk, C. B., Simpson, E. H., Keilp, J. G., Bruder, G. E., working memory capacity and maturation of the posterior cingulate Polan, H. J., . . . Gilliam, T. C. (2007). DRD2 C957T polymorphism cortex during childhood. Biological Psychiatry, 68, 1120 –1125. doi: interacts with the COMT Val158Met polymorphism in human working memory ability. Schizophrenia Research, 90, 104 –107. doi:10.1016/ Solanto, M. V. (1998). Neuropsychopharmacological mechanisms of stim- ulant drug action in attention-deficit hyperactivity disorder: A reviewand integration. Behavioural Brain Research, 94, 127–152. doi:10.1016/ Stelzel, C., Basten, U., Montag, C., Reuter, M., & Fiebach, C. J. (2009).

Source: http://www.klingberglab.se/pub/Soderqvist_2011.pdf