- Email: [email protected]
Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats
Journal Pre-proof Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats Lukas Meyerolbersleben, Christine Winter, Nadine Bernhardt
PII:
S0166-4328(19)31038-1
DOI:
https://doi.org/10.1016/j.bbr.2019.112244
Reference:
BBR 112244
To appear in:
Behavioural Brain Research
Received Date:
4 July 2019
Revised Date:
21 August 2019
Accepted Date:
16 September 2019
Please cite this article as: Meyerolbersleben L, Winter C, Bernhardt N, Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats, Behavioural Brain Research (2019), doi: https://doi.org/10.1016/j.bbr.2019.112244
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.
Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats
Lukas Meyerolbersleben1, Christine Winter1,2, Nadine Bernhardt1* 1Department
of Psychiatry and Psychotherapy, University Hospital Carl Gustav Carus, Technische
2Department
ro of
Universität Dresden, Germany of Psychiatry and Psychotherapy, Charité Universitätsmedizin Berlin, Germany
*Corresponding author:
-p
Nadine Bernhardt, Department of Psychiatry and Psychotherapy, University Hospital Carl Gustav
re
Carus at Technische Universität Dresden, Fetscherstraße 74, 01307 Dresden, Germany.
lP
Email: [email protected], Phone: 0049 351 458 2047
ur na
Highlights
The Sucrose Preference Test can be used to dissociate ‘wanting’ and ‘liking’ within the same behavioural assay.
Hypodopaminergic DAT-tg rats exhibit reduced absolute consumption of sucrose solution, but equal preference for sucrose solution over water. This supports the incentive salience hypothesis of mesolimbic dopamine function by
Jo
showing that reduced dopamine levels lead to reduced ‘wanting’, but do not affect ‘liking’ in rats.
Abstract The predominant theory of mesolimbic dopamine function in recent years has been the incentive salience hypothesis, which describes the role of midbrain dopaminergic circuits as encoding
‘wanting’, but not ‘liking’ of rewards, resulting in a dissociation of the two functions. However, until now, this dissociation was only established in separate behavioural assays. Here, we propose that the Sucrose Preference Test (SPT), which has previously been used to measure ‘liking’, can actually be interpreted in terms of both ‘wanting’ and ‘liking’: while relative preference is a proxy for ‘liking’, we argue that absolute consumption of sucrose is indicative of ‘wanting’. To test this supposition, we conducted the SPT in hypodopaminergic DAT-tg rats. While DAT-tg rats exhibited reduced absolute consumption of sucrose their relative preference for sucrose was unimpaired compared to wildtype rats. When interpreted in terms of the incentive salience hypothesis of DA function, these results indicate that absolute consumption in the SPT depends on ‘wanting’ and that the SPT is a
ro of
valuable instrument for dissociating ‘wanting’ and ‘liking’ within the same behavioural paradigm.
Keywords
-p
Dopamine, motivation, anhedonia, sucrose preference
re
Main text
Dopamine (DA) plays a highly important role in neural coding of reward [1]. Reward coding itself
lP
is usually dissociated into a motivational component (incentive salience or ‘wanting’) and a hedonic component (hedonic impact or ‘liking’)[2]. Whether dopaminergic circuits encode ‘wanting’ or
ur na
‘liking’, or a combination of both, has long been a point of contention [3]. Historically, the (an-) hedonia hypothesis of mesolimbic DA function has been particularly influential [4], stating that DA signalling in the midbrain is crucial for hedonic experience [5]. However, the balance of the evidence has shifted significantly towards DA encoding incentive salience [6] after numerous studies demonstrated that mesolimbic dopaminergic signalling is necessary for ‘wanting’, but not ‘liking’:
Jo
Inducing mesolimbic DA depletion in rodents has shown hedonic reactivity to be unimpaired, while the animals’ ability to exert motivational effort was sharply decreased [2,7–9]. Conversely, elevated DA levels were found to enhance motivation, but not to increase hedonic impact [10–12]. In these studies, ‘liking’ has mostly been measured using orofacial reactions [13], in combination with a separate instrumental task to measure ‘wanting’, e.g. performance and running speed in a T-maze task [2,10,12]. However, using separate behavioural assays introduces confounding factors and decreases direct conceptual comparability. It would therefore be preferable to measure the two constructs within the same behavioural paradigm.
Recently, a modification of the Sucrose Preference Test (SPT) was introduced as a measure of ‘liking’[14,15]. In this version, a preference baseline is first established for which animals choose between two bottles, which both contain only water. For testing, the content of one bottle is then exchanged for sucrose solution. Here, we propose that this SPT can be interpreted in terms of both ‘wanting’ and ‘liking’, allowing for dissociation within the same behavioural assay. We base this on the simple assumption that ‘liking’ determines which reward is consumed, but ‘wanting’ determines how much of it is consumed. Choosing between two bottles requires no significant motivational effort and is therefore reflective of ‘liking’. Conversely, the act of consumption requires motivational effort and is mediated by DA signalling: active consumption of sugar leads to different DA responses in N. accumbens than passive infusion with sucrose [16] and increased accumbal DA levels promotes
ro of
consumption of a reward beyond physiological necessity [17]. Therefore, we argue that absolute consumption, which has hitherto been treated as a confounding factor in the SPT to be eliminated, can actually be considered a measure of ‘wanting’. This means that changes in mesolimbic DA levels should lead to concurrent changes in absolute consumption, while leaving relative preference
-p
unaffected.
To test this, we conducted the SPT in a transgenic rat model, where overexpression of the
re
synaptic dopamine transporter (DAT) leads to a moderate reduction of mesolimbic DA levels [18]. All animal experiments were carried out in accordance with the European Communities Council
lP
Directive of 22nd September 2010 (2010/63/EU) under protocols approved by the local animal ethics committees (LDS, TVV65/2015). Experiments were conducted on 17 male hemizygous DAT-
ur na
tg and 14 control littermates (wildtypes, wt), mean age 82 days (SD = 3). Animals were generated as described elsewhere [18]. Briefly, a construct containing the NSE promoter and murine DAT coding sequence was used for microinjection into the pronucleus of zygotes from Sprague-Dawley (SD) Hanover rats (Janvier labs). Transgenics are maintained on SD-background in a continuous hemizygous x wt offspring breeding scheme (> 20 generations). Genotypes were verified using PCR.
Jo
Animals were raised group housed in polycarbonate cages with standard rat bedding and stainlesssteel wire lids on a 12-h light dark cycle (light on at 06:00 am) with food and water ad libitum. All efforts were made to reduce animal suffering and number of animals used. Animals were separated into individual cages for the experiment. Two 250 ml bottles with sipper caps were filled with water, marked and placed next to each other on one side of the divided wire rack. Animals were weighed and allowed to acclimatize to these conditions for three days. In order to establish a baseline of absolute consumption and bottle preference, consumption was measured for two days with both bottles containing water (baseline days; BD1 and BD2). After that, the water
in one of the bottles was exchanged for a 2%-sucrose solution, and consumption of the two bottles was measured for three further days (testing days; TD1-3). The water and sucrose solution were placed into the bottles at random, so that no single bottle always contained the same liquid. Positions of the bottles were switched daily. Raw values of consumption were calculated as the daily reduction in the weight of each bottle. DA levels indirectly influence bodyweight and metabolic rate, which in turn influences water and food intake [19]. Concurrently, lower bodyweights for DAT-tg rats compared to wt have been reported [18]. For analysis of absolute consumption, we therefore calculated a weight-corrected measure of ml consumed per 100g bodyweight to exclude this confound. On baseline days, we used the sum of water consumption from both bottles. For analysis of relative sucrose preference, an
ro of
index was calculated as: Sucrose Preference = V(sucrose solution)/[V(sucrose solution) + V(water)] [14]. For baseline days, the same formula was used to calculate a simple preference index for one bottle over the other. Weight correction was not necessary for values used to calculate the preference index, as all input values would have been divided by the same weight.
-p
For each dependent variable – sucrose solution consumption, water consumption, and relative sucrose preference – we conducted a 2 (genotype: DAT-tg vs. wt) x 5(time: BD1 vs. BD2 vs. TD1 vs.
re
TD2 vs. TD3) ANOVA with repeated measures and subsequent analysis of linear contrast variables using two-factorial MANOVA. Mauchly’s test indicated violations of sphericity for all three
lP
dependent variables, with χ2(9) = 72.16, p < .001, χ2(9) = 46.30, p < .001 and χ2(9) = 48.01, p < .001 respectively. Therefore, Greenhouse-Geisser correction was applied (estimates of sphericity ε =
ur na
.684, ε = .543, ε = .669, respectively; corrected values marked as p*). A one-tailed t-test confirmed that DAT-tg animals weighed significantly less than wt rats (DATtg: M = 456 g, SD = 52; wt: M = 540 g, SD = 52), t(29) = 4.49, p < .001. For absolute consumption of sucrose solution, the factor time exerted a significant main effect on consumption within groups, F(2.74, 79.34) = 32.21, p < .001, ηp2 = .526. Likewise, there was a main effect for genotype, F(1, 29)
Jo
= 14.69, p = .001, ηp2 = .336, with DAT-tg rats (M = 8.67 ml/100g, SD = 2.94) consuming overall smaller amounts than wt rats (M = 12.18 ml/100g, SD = 1.95). Furthermore, the analysis yielded a significant interaction of time and genotype, F(2.74, 79.34) = 4.77, p* = .005, ηp2 = .141. An analysis of absolute consumption without bodyweight correction yielded the same results (data not shown). Within-group contrasts showed that there was no difference in consumption between BD 1 (M = 7.46 ml/100g, SD = 1.55) and BD2 (M = 7.51 ml/100g, SD = 1.47), F(1, 29) = .12, p = .737, ηp2 = .004. There was an increase in consumption between BDmean (M = 7.49 ml/100g, SD = 1.44) and TDmean (M = 12.10 ml/100g, SD = 4.32), specific to the switch to sucrose, F(1, 29) = 88.05, p < .001, ηp2 = .752.
This increase interacted significantly with genotype, F(1, 29) = 11.61, p = .002, ηp2 = .286, so that from BDs to TDs, consumption increased less in DAT-tg rats than in wt animals (Figure 1). There was also a significant decrease in sucrose consumption from TD 1 to TD2, F(1, 29) = 13.80, p = .001, ηp2 = .322. This effect did not interact with genotype, F(1, 29) = 2.62, p = .116, ηp2 = .083. Between TD2 and TD3, consumption rates did not differ, nor interact with genotype (all Fs < 1). For water consumption, we observed a main effect of time, F(2.17, 62.94) = 603.56, p* < .001, ηp2 = .954, a main effect of genotype, F(1, 29) = 4.42, p = .044, ηp2 = .132, and a significant interaction between time and genotype, F(2.17, 62.94) = 7.72, p* = .002, ηp2 = . 210. Linear contrasts indicated a decrease in water consumption between BDmean (M = 7.49 ml/100g, SD = 1.44) and TDmean (M = .63 ml, SD = .40), F(1, 29) = 1139.21, p < .001, ηp2 = .975 (Figure 1). There was also a significant
ro of
interaction of this decrease with genotype, F(1, 29) = 14.50, p < .001, ηp2 = .333. Linear contrasts between all other time points, as well as their interactions with genotype were non-significant (all Fs < 1).
-p
---insert Figure 1----
re
Regarding sucrose preference, a main effect of time, F(2.68, 77.64) = 64.99, p* < .001, ηp2 = .691, no main effect of genotype, F(1,29) = 3.86, p = .059, ηp2 = .117 and no interaction of time and
lP
genotype, F(2.68, 77.64) = 1.00, p* = .392, ηp2 = .033 was found (Figure 2). Linear contrast analysis revealed a significant difference in preference values between BD1 (M = .648, SD = .237) and BD2 (M
ur na
= .435, SD = .223), F(1, 29) = 15.22, p = .001, ηp2 = .344. This indicates that animals developed a side bias on BDs. However, this effect did not interact with genotype, F(1,29) = .60, p = .444, ηp2 = .020. There was a significant contrast in preference values between BDmean and TDmean, F(1, 29) = 182.71, p < .001, ηp2 = .863, reflective of sucrose preference over water. The contrasts among TDs were nonsignificant (all Fs < 1). Importantly, the between-group factor genotype did not interact with any of
Jo
these contrasts: BDmean and TDmean, F(1, 29) = 1.99, p = .169, ηp2 = .064, TD1 and TD2, F(1, 29) = .50, p = .486, ηp2 = .017, TD2 and TD3, F(1, 29) = .35, p = .559, ηp2 = .012.
---insert Figure 2----
In summary, we found that DAT-tg rats with reduced mesolimbic DA levels exhibit reduced absolute consumption, but equal preference for sucrose compared to wt rats in the SPT. According to the incentive salience hypothesis of mesolimbic DA function [20], this is an indicator of impaired
‘wanting’, but intact ‘liking’. We thereby show that the modern variant of the SPT, which was conceived as a test of anhedonia [14,15], can be used to dissociate ‘wanting’ and ‘liking’ within the same behavioural assay. Concerning absolute consumption, we observed the expected main effects for time, confirming that there was a large increase in consumption specific to the switch from water to sucrose. DAT-tg rats’ overall absolute consumption was significantly lower for both sucrose solution and water compared to controls. Our study thus complements Peciña et al.’s [12] finding that hyperdopaminergic DAT-knockout mice exhibit higher overall food and water intake: we show here that the reverse holds true in hypodopaminergic animals. Crucial to our main hypothesis regarding the connection between ‘wanting’ and absolute consumption, intake in DAT-tg rats increased by a
ro of
lesser amount than in wt rats after switching to sucrose solution. In contrast, water consumption was lower in DAT-tg than in wt rats on baseline days, but dropped to identical levels near zero in both groups on testing days (Figure 1). This amounted to a smaller net decrease in water consumption on TDs for DAT-tg rats and therefore yielded a significant time × genotype - interaction
-p
effect.
Combined with the assumption that decreased DA levels in the midbrain of DAT-tg rats lead to
re
decreased encoding of incentive salience, this provides evidence that ‘wanting’ can indeed be operationalized as absolute consumption in general, and absolute sucrose consumption in
lP
particular. The fact that the pattern of consumption was found irrespective of whether or not absolute consumption was corrected for bodyweight makes it unlikely that differences in metabolic
ur na
rates between genotypes acted as a confounding factor. ‘Wanting’ was not entirely abolished in DAT-tg rats, as evidenced by the fact that their consumption of sucrose solution rose beyond their baseline consumption rate for water. This was to be expected, as DAT-tg rats display only a moderate reduction in DA levels and a partial compensation by upregulation of DRD1/2 [18]. Remaining DA signalling and generally intact sensorimotor functions as demonstrated in prior
Jo
behavioural studies in DAT-tg rats make it unlikely that pure motor deficits underlie reduced consumption rates [18,21]. The dissociation of ‘wanting’ and ‘liking’ is completed by our finding that preference for sucrose
over water is the same in DAT-tg rats as in wt rats, indicating that ‘liking’ is intact in these animals. As predicted, overall preference values differed significantly between time points. Analysis of linear contrasts confirmed that the increase in preference values was specific to the switch to sucrose solution. However, there was no significant main effect for genotype, nor an interaction between genotype and time. Crucially, genotype did not interact with the difference between BDmean and
TDmean. Hypodopaminergic DAT-tg rats thus showed an intact hedonic response, tallying with results from orofacial liking studies [2,10,12]. Interestingly, our results contradict a series of early pharmacological studies which reported decreased sucrose preference, and thus decreased ‘liking’, following the injection of DA antagonists [22–24]. However, these studies used an early form of the sucrose preference test, which failed to establish a preference baseline. Thus, it is possible that these studies confounded the reduction of absolute consumption in DA-deficient animals, which we now show to be a symptom of decreased ‘wanting’, with a non-existent reduction of relative sucrose preference, which we show remains stable in spite of impaired DA signalling. However, inherent discrepancies between pharmacological and genetic models targeting different players of the DA system may also contribute to such
ro of
divergent findings.
The following limitations should be considered. Animals of both genotypes developed a side bias during baseline days. However, the impact of this bias on preference index values used for statistical analysis was mitigated as follows: First, bottle positions were switched daily, so that each bottle
-p
occupied the preferred location on one baseline day. In line with this, preference values reversed themselves on BD2 (see Figure 2). Second, we used the mean over both BDs for the crucial linear
re
contrast between preference values on BDs and TDs, providing a more robust baseline for comparison. We also observed a significant decrease in sucrose consumption between TD1 and TD2.
lP
This occurred in both genotypes, suggesting a common cause (e.g. habituation or satiation). Furthermore, there was a floor effect for water consumption on TDs. However, such a steep
ur na
reduction is expected, with counterbalancing of bottle positions and rats’ innate exploratory behaviour leading to brief exploration and thus consumption of small amounts of water in both genotypes. This floor effect makes it inadvisable to use water consumption directly to calculate the preference index; instead, total consumption should be used [14,15]. Finally, further studies are needed to investigate whether a similar dissociation can be obtained
Jo
using different reward contingencies and whether genotype-specific differences in the microstructure of drinking behaviour correlate with the differences in ‘wanting’ observed here (i.e. number or size of lick clusters [25]).
Acknowledgements This work was supported by the TUD, the Federal Ministry of Education and Research, Germany [Grant 01EE1406A] and the Deutsche Forschungsgemeinschaft [DFG, WI 2140/4-1]. We thank Maike Kristin Lieser and Kristin Wogan for excellent support.
The authors declare no competing interests.
References [1]
P.N. Tobler, C.D. Fiorillo, W. Schultz, Adaptive coding of reward value by dopamine
neurons, Science. 307 (2005) 1642–1645. doi:10.1126/science.1105370. [2]
K.C. Berridge, T.E. Robinson, What is the role of dopamine in reward: hedonic impact,
reward learning, or incentive salience?, Brain Res. Brain Res. Rev. 28 (1998) 309–369.
‘wanting’,
K.C. Berridge, T.E. Robinson, J.W. Aldridge, Dissecting components of reward: ‘liking’, and
learning,
Current
Opinion
in
Pharmacology.
doi:10.1016/j.coph.2008.12.014. [4]
(2009)
65–73.
-p
R.A. Wise, The anhedonia hypothesis: Mark III, Behavioral and Brain Sciences. 8
re
(1985) 178. doi:10.1017/S0140525X00020306. [6]
9
R.A. Wise, Dopamine and reward: the anhedonia hypothesis 30 years on, Neurotox
Res. 14 (2008) 169–183. doi:10.1007/BF03033808. [5]
ro of
[3]
J.D. Salamone, M. Correa, The Mysterious Motivational Functions of Mesolimbic
Dopamine, Neuron. 76 (2012) 470–485. doi:10.1016/j.neuron.2012.10.021. J.A. Parkinson, J.W. Dalley, R.N. Cardinal, A. Bamford, B. Fehnert, G. Lachenal, N.
lP
[7]
Rudarakanchana, K.M. Halkerston, T.W. Robbins, B.J. Everitt, Nucleus accumbens dopamine
ur na
depletion impairs both acquisition and performance of appetitive Pavlovian approach behaviour: implications for mesoaccumbens dopamine function, Behav. Brain Res. 137 (2002) 149–163. [8]
S. Robinson, S.M. Sandstrom, V.H. Denenberg, R.D. Palmiter, Distinguishing Whether
Dopamine Regulates Liking, Wanting, and/or Learning About Rewards., Behavioral Neuroscience. 119 (2005) 5–15. doi:10.1037/0735-7044.119.1.5. M. Darvas, A.M. Wunsch, J.T. Gibbs, R.D. Palmiter, Dopamine dependency for
Jo
[9]
acquisition and performance of Pavlovian conditioned response, Proceedings of the National Academy of Sciences. 111 (2014) 2764–2769. doi:10.1073/pnas.1400332111. [10]
K.S. Smith, K.C. Berridge, J.W. Aldridge, Disentangling pleasure from incentive
salience and learning signals in brain reward circuitry, Proceedings of the National Academy of Sciences. 108 (2011) E255–E264. doi:10.1073/pnas.1101920108. [11]
M.E. Bardgett, M. Depenbrock, N. Downs, M. Points, L. Green, Dopamine modulates
effort-based decision making in rats., Behavioral Neuroscience. 123 (2009) 242–251.
doi:10.1037/a0014625. [12]
S. Peciña, B. Cagniard, K.C. Berridge, J.W. Aldridge, X. Zhuang, Hyperdopaminergic
mutant mice have higher “wanting” but not “liking” for sweet rewards, J. Neurosci. 23 (2003) 9395– 9402. [13]
K.C. Berridge, Measuring hedonic impact in animals and infants: microstructure of
affective taste reactivity patterns, Neurosci Biobehav Rev. 24 (2000) 173–198. [14]
T. Serchov, D. van Calker, K. Biber, Sucrose Preference Test to Measure Anhedonic
Behaviour in Mice, BIO-PROTOCOL. 6 (2016). doi:10.21769/BioProtoc.1958. [15]
M.-Y. Liu, C.-Y. Yin, L.-J. Zhu, X.-H. Zhu, C. Xu, C.-X. Luo, H. Chen, D.-Y. Zhu, Q.-G. Zhou,
Sucrose preference test for measurement of stress-induced anhedonia in mice, Nature Protocols.
[16]
ro of
13 (2018) 1686–1698. doi:10.1038/s41596-018-0011-z.
J.E. McCutcheon, M.F. Roitman, Mode of Sucrose Delivery Alters Reward-Related
Phasic Dopamine Signals in Nucleus Accumbens, ACS Chemical Neuroscience. (2018). doi:10.1021/acschemneuro.8b00262.
A. Hajnal, R. Norgren, Accumbens dopamine mechanisms in sucrose intake, Brain
-p
[17]
Research. 904 (2001) 76–84. doi:10.1016/S0006-8993(01)02451-9.
R. Hadar, H. Edemann-Callesen, C. Reinel, F. Wieske, M. Voget, E. Popova, R. Sohr, Y.
re
[18]
Avchalumov, J. Priller, C. van Riesen, I. Puls, M. Bader, C. Winter, Rats overexpressing the dopamine
lP
transporter display behavioral and neurobiological abnormalities with relevance to repetitive disorders, Scientific Reports. 6 (2016). doi:10.1038/srep39145. L.F. Jakubczak, Food and water intakes of rats as a function of strain, age,
ur na
[19]
temperature, and body weight, Physiol. Behav. 17 (1976) 251–258. [20]
K.C. Berridge, The debate over dopamine’s role in reward: the case for incentive
salience, Psychopharmacology. 191 (2007) 391–431. doi:10.1007/s00213-006-0578-x. [21]
N. Bernhardt, M.K. Lieser, E.-B. Hlusicka, B. Habelt, F. Wieske, H. Edemann-Callesen,
Jo
A. Garthe, C. Winter, Learning deficits in rats overexpressing the dopamine transporter, Scientific Reports. 8 (2018) 14173. doi:10.1038/s41598-018-32608-7. [22]
A. Towell, R. Muscat, P. Willner, Effects of pimozide on sucrose consumption and
preference, Psychopharmacology. 92 (1987). doi:10.1007/BF00177926. [23]
R. Muscat, P. Willner, Effects of dopamine receptor antagonists on sucrose
consumption and preference, Psychopharmacology (Berl.). 99 (1989) 98–102. [24]
S. Hsiao, G.P. Smith, Raclopride reduces sucrose preference in rats, Pharmacol.
Biochem. Behav. 50 (1995) 121–125.
[25] Quarterly
D.M. Dwyer, Licking and Liking: The Assessment of Hedonic Responses in Rodents, Journal
of
Experimental
Psychology.
65
(2012)
371–394.
doi:10.1080/17470218.2011.652969.
ro of
Figure legends
Figure 1: Absolute consumption of water and sucrose solution in ml/100g bodyweight for DAT-tg vs. wt animals across baseline days (BD) and testing days (TD). For baseline days, the sum of water consumption from the two bottles is displayed. Error bars represent ± 1 SE; * represents significant
Jo
ur na
lP
re
-p
main effects of time and genotype (p < .001).
Figure 2: Preference Index values for DAT-tg vs. wt rats. On baseline days (BD), the index reflects a simple bottle preference; on testing days (TD), the index reflects preference of sucrose solution over
water. Preference values above and below 0.5 indicate preference of one bottle over the other.
Jo
ur na
lP
re
-p
ro of
Error bars represent ± 1 SE; * represents the significant main effect of time (p < .001).
PII:
S0166-4328(19)31038-1
DOI:
https://doi.org/10.1016/j.bbr.2019.112244
Reference:
BBR 112244
To appear in:
Behavioural Brain Research
Received Date:
4 July 2019
Revised Date:
21 August 2019
Accepted Date:
16 September 2019
Please cite this article as: Meyerolbersleben L, Winter C, Bernhardt N, Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats, Behavioural Brain Research (2019), doi: https://doi.org/10.1016/j.bbr.2019.112244
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.
Dissociation of wanting and liking in the sucrose preference test in dopamine transporter overexpressing rats
Lukas Meyerolbersleben1, Christine Winter1,2, Nadine Bernhardt1* 1Department
of Psychiatry and Psychotherapy, University Hospital Carl Gustav Carus, Technische
2Department
ro of
Universität Dresden, Germany of Psychiatry and Psychotherapy, Charité Universitätsmedizin Berlin, Germany
*Corresponding author:
-p
Nadine Bernhardt, Department of Psychiatry and Psychotherapy, University Hospital Carl Gustav
re
Carus at Technische Universität Dresden, Fetscherstraße 74, 01307 Dresden, Germany.
lP
Email: [email protected], Phone: 0049 351 458 2047
ur na
Highlights
The Sucrose Preference Test can be used to dissociate ‘wanting’ and ‘liking’ within the same behavioural assay.
Hypodopaminergic DAT-tg rats exhibit reduced absolute consumption of sucrose solution, but equal preference for sucrose solution over water. This supports the incentive salience hypothesis of mesolimbic dopamine function by
Jo
showing that reduced dopamine levels lead to reduced ‘wanting’, but do not affect ‘liking’ in rats.
Abstract The predominant theory of mesolimbic dopamine function in recent years has been the incentive salience hypothesis, which describes the role of midbrain dopaminergic circuits as encoding
‘wanting’, but not ‘liking’ of rewards, resulting in a dissociation of the two functions. However, until now, this dissociation was only established in separate behavioural assays. Here, we propose that the Sucrose Preference Test (SPT), which has previously been used to measure ‘liking’, can actually be interpreted in terms of both ‘wanting’ and ‘liking’: while relative preference is a proxy for ‘liking’, we argue that absolute consumption of sucrose is indicative of ‘wanting’. To test this supposition, we conducted the SPT in hypodopaminergic DAT-tg rats. While DAT-tg rats exhibited reduced absolute consumption of sucrose their relative preference for sucrose was unimpaired compared to wildtype rats. When interpreted in terms of the incentive salience hypothesis of DA function, these results indicate that absolute consumption in the SPT depends on ‘wanting’ and that the SPT is a
ro of
valuable instrument for dissociating ‘wanting’ and ‘liking’ within the same behavioural paradigm.
Keywords
-p
Dopamine, motivation, anhedonia, sucrose preference
re
Main text
Dopamine (DA) plays a highly important role in neural coding of reward [1]. Reward coding itself
lP
is usually dissociated into a motivational component (incentive salience or ‘wanting’) and a hedonic component (hedonic impact or ‘liking’)[2]. Whether dopaminergic circuits encode ‘wanting’ or
ur na
‘liking’, or a combination of both, has long been a point of contention [3]. Historically, the (an-) hedonia hypothesis of mesolimbic DA function has been particularly influential [4], stating that DA signalling in the midbrain is crucial for hedonic experience [5]. However, the balance of the evidence has shifted significantly towards DA encoding incentive salience [6] after numerous studies demonstrated that mesolimbic dopaminergic signalling is necessary for ‘wanting’, but not ‘liking’:
Jo
Inducing mesolimbic DA depletion in rodents has shown hedonic reactivity to be unimpaired, while the animals’ ability to exert motivational effort was sharply decreased [2,7–9]. Conversely, elevated DA levels were found to enhance motivation, but not to increase hedonic impact [10–12]. In these studies, ‘liking’ has mostly been measured using orofacial reactions [13], in combination with a separate instrumental task to measure ‘wanting’, e.g. performance and running speed in a T-maze task [2,10,12]. However, using separate behavioural assays introduces confounding factors and decreases direct conceptual comparability. It would therefore be preferable to measure the two constructs within the same behavioural paradigm.
Recently, a modification of the Sucrose Preference Test (SPT) was introduced as a measure of ‘liking’[14,15]. In this version, a preference baseline is first established for which animals choose between two bottles, which both contain only water. For testing, the content of one bottle is then exchanged for sucrose solution. Here, we propose that this SPT can be interpreted in terms of both ‘wanting’ and ‘liking’, allowing for dissociation within the same behavioural assay. We base this on the simple assumption that ‘liking’ determines which reward is consumed, but ‘wanting’ determines how much of it is consumed. Choosing between two bottles requires no significant motivational effort and is therefore reflective of ‘liking’. Conversely, the act of consumption requires motivational effort and is mediated by DA signalling: active consumption of sugar leads to different DA responses in N. accumbens than passive infusion with sucrose [16] and increased accumbal DA levels promotes
ro of
consumption of a reward beyond physiological necessity [17]. Therefore, we argue that absolute consumption, which has hitherto been treated as a confounding factor in the SPT to be eliminated, can actually be considered a measure of ‘wanting’. This means that changes in mesolimbic DA levels should lead to concurrent changes in absolute consumption, while leaving relative preference
-p
unaffected.
To test this, we conducted the SPT in a transgenic rat model, where overexpression of the
re
synaptic dopamine transporter (DAT) leads to a moderate reduction of mesolimbic DA levels [18]. All animal experiments were carried out in accordance with the European Communities Council
lP
Directive of 22nd September 2010 (2010/63/EU) under protocols approved by the local animal ethics committees (LDS, TVV65/2015). Experiments were conducted on 17 male hemizygous DAT-
ur na
tg and 14 control littermates (wildtypes, wt), mean age 82 days (SD = 3). Animals were generated as described elsewhere [18]. Briefly, a construct containing the NSE promoter and murine DAT coding sequence was used for microinjection into the pronucleus of zygotes from Sprague-Dawley (SD) Hanover rats (Janvier labs). Transgenics are maintained on SD-background in a continuous hemizygous x wt offspring breeding scheme (> 20 generations). Genotypes were verified using PCR.
Jo
Animals were raised group housed in polycarbonate cages with standard rat bedding and stainlesssteel wire lids on a 12-h light dark cycle (light on at 06:00 am) with food and water ad libitum. All efforts were made to reduce animal suffering and number of animals used. Animals were separated into individual cages for the experiment. Two 250 ml bottles with sipper caps were filled with water, marked and placed next to each other on one side of the divided wire rack. Animals were weighed and allowed to acclimatize to these conditions for three days. In order to establish a baseline of absolute consumption and bottle preference, consumption was measured for two days with both bottles containing water (baseline days; BD1 and BD2). After that, the water
in one of the bottles was exchanged for a 2%-sucrose solution, and consumption of the two bottles was measured for three further days (testing days; TD1-3). The water and sucrose solution were placed into the bottles at random, so that no single bottle always contained the same liquid. Positions of the bottles were switched daily. Raw values of consumption were calculated as the daily reduction in the weight of each bottle. DA levels indirectly influence bodyweight and metabolic rate, which in turn influences water and food intake [19]. Concurrently, lower bodyweights for DAT-tg rats compared to wt have been reported [18]. For analysis of absolute consumption, we therefore calculated a weight-corrected measure of ml consumed per 100g bodyweight to exclude this confound. On baseline days, we used the sum of water consumption from both bottles. For analysis of relative sucrose preference, an
ro of
index was calculated as: Sucrose Preference = V(sucrose solution)/[V(sucrose solution) + V(water)] [14]. For baseline days, the same formula was used to calculate a simple preference index for one bottle over the other. Weight correction was not necessary for values used to calculate the preference index, as all input values would have been divided by the same weight.
-p
For each dependent variable – sucrose solution consumption, water consumption, and relative sucrose preference – we conducted a 2 (genotype: DAT-tg vs. wt) x 5(time: BD1 vs. BD2 vs. TD1 vs.
re
TD2 vs. TD3) ANOVA with repeated measures and subsequent analysis of linear contrast variables using two-factorial MANOVA. Mauchly’s test indicated violations of sphericity for all three
lP
dependent variables, with χ2(9) = 72.16, p < .001, χ2(9) = 46.30, p < .001 and χ2(9) = 48.01, p < .001 respectively. Therefore, Greenhouse-Geisser correction was applied (estimates of sphericity ε =
ur na
.684, ε = .543, ε = .669, respectively; corrected values marked as p*). A one-tailed t-test confirmed that DAT-tg animals weighed significantly less than wt rats (DATtg: M = 456 g, SD = 52; wt: M = 540 g, SD = 52), t(29) = 4.49, p < .001. For absolute consumption of sucrose solution, the factor time exerted a significant main effect on consumption within groups, F(2.74, 79.34) = 32.21, p < .001, ηp2 = .526. Likewise, there was a main effect for genotype, F(1, 29)
Jo
= 14.69, p = .001, ηp2 = .336, with DAT-tg rats (M = 8.67 ml/100g, SD = 2.94) consuming overall smaller amounts than wt rats (M = 12.18 ml/100g, SD = 1.95). Furthermore, the analysis yielded a significant interaction of time and genotype, F(2.74, 79.34) = 4.77, p* = .005, ηp2 = .141. An analysis of absolute consumption without bodyweight correction yielded the same results (data not shown). Within-group contrasts showed that there was no difference in consumption between BD 1 (M = 7.46 ml/100g, SD = 1.55) and BD2 (M = 7.51 ml/100g, SD = 1.47), F(1, 29) = .12, p = .737, ηp2 = .004. There was an increase in consumption between BDmean (M = 7.49 ml/100g, SD = 1.44) and TDmean (M = 12.10 ml/100g, SD = 4.32), specific to the switch to sucrose, F(1, 29) = 88.05, p < .001, ηp2 = .752.
This increase interacted significantly with genotype, F(1, 29) = 11.61, p = .002, ηp2 = .286, so that from BDs to TDs, consumption increased less in DAT-tg rats than in wt animals (Figure 1). There was also a significant decrease in sucrose consumption from TD 1 to TD2, F(1, 29) = 13.80, p = .001, ηp2 = .322. This effect did not interact with genotype, F(1, 29) = 2.62, p = .116, ηp2 = .083. Between TD2 and TD3, consumption rates did not differ, nor interact with genotype (all Fs < 1). For water consumption, we observed a main effect of time, F(2.17, 62.94) = 603.56, p* < .001, ηp2 = .954, a main effect of genotype, F(1, 29) = 4.42, p = .044, ηp2 = .132, and a significant interaction between time and genotype, F(2.17, 62.94) = 7.72, p* = .002, ηp2 = . 210. Linear contrasts indicated a decrease in water consumption between BDmean (M = 7.49 ml/100g, SD = 1.44) and TDmean (M = .63 ml, SD = .40), F(1, 29) = 1139.21, p < .001, ηp2 = .975 (Figure 1). There was also a significant
ro of
interaction of this decrease with genotype, F(1, 29) = 14.50, p < .001, ηp2 = .333. Linear contrasts between all other time points, as well as their interactions with genotype were non-significant (all Fs < 1).
-p
---insert Figure 1----
re
Regarding sucrose preference, a main effect of time, F(2.68, 77.64) = 64.99, p* < .001, ηp2 = .691, no main effect of genotype, F(1,29) = 3.86, p = .059, ηp2 = .117 and no interaction of time and
lP
genotype, F(2.68, 77.64) = 1.00, p* = .392, ηp2 = .033 was found (Figure 2). Linear contrast analysis revealed a significant difference in preference values between BD1 (M = .648, SD = .237) and BD2 (M
ur na
= .435, SD = .223), F(1, 29) = 15.22, p = .001, ηp2 = .344. This indicates that animals developed a side bias on BDs. However, this effect did not interact with genotype, F(1,29) = .60, p = .444, ηp2 = .020. There was a significant contrast in preference values between BDmean and TDmean, F(1, 29) = 182.71, p < .001, ηp2 = .863, reflective of sucrose preference over water. The contrasts among TDs were nonsignificant (all Fs < 1). Importantly, the between-group factor genotype did not interact with any of
Jo
these contrasts: BDmean and TDmean, F(1, 29) = 1.99, p = .169, ηp2 = .064, TD1 and TD2, F(1, 29) = .50, p = .486, ηp2 = .017, TD2 and TD3, F(1, 29) = .35, p = .559, ηp2 = .012.
---insert Figure 2----
In summary, we found that DAT-tg rats with reduced mesolimbic DA levels exhibit reduced absolute consumption, but equal preference for sucrose compared to wt rats in the SPT. According to the incentive salience hypothesis of mesolimbic DA function [20], this is an indicator of impaired
‘wanting’, but intact ‘liking’. We thereby show that the modern variant of the SPT, which was conceived as a test of anhedonia [14,15], can be used to dissociate ‘wanting’ and ‘liking’ within the same behavioural assay. Concerning absolute consumption, we observed the expected main effects for time, confirming that there was a large increase in consumption specific to the switch from water to sucrose. DAT-tg rats’ overall absolute consumption was significantly lower for both sucrose solution and water compared to controls. Our study thus complements Peciña et al.’s [12] finding that hyperdopaminergic DAT-knockout mice exhibit higher overall food and water intake: we show here that the reverse holds true in hypodopaminergic animals. Crucial to our main hypothesis regarding the connection between ‘wanting’ and absolute consumption, intake in DAT-tg rats increased by a
ro of
lesser amount than in wt rats after switching to sucrose solution. In contrast, water consumption was lower in DAT-tg than in wt rats on baseline days, but dropped to identical levels near zero in both groups on testing days (Figure 1). This amounted to a smaller net decrease in water consumption on TDs for DAT-tg rats and therefore yielded a significant time × genotype - interaction
-p
effect.
Combined with the assumption that decreased DA levels in the midbrain of DAT-tg rats lead to
re
decreased encoding of incentive salience, this provides evidence that ‘wanting’ can indeed be operationalized as absolute consumption in general, and absolute sucrose consumption in
lP
particular. The fact that the pattern of consumption was found irrespective of whether or not absolute consumption was corrected for bodyweight makes it unlikely that differences in metabolic
ur na
rates between genotypes acted as a confounding factor. ‘Wanting’ was not entirely abolished in DAT-tg rats, as evidenced by the fact that their consumption of sucrose solution rose beyond their baseline consumption rate for water. This was to be expected, as DAT-tg rats display only a moderate reduction in DA levels and a partial compensation by upregulation of DRD1/2 [18]. Remaining DA signalling and generally intact sensorimotor functions as demonstrated in prior
Jo
behavioural studies in DAT-tg rats make it unlikely that pure motor deficits underlie reduced consumption rates [18,21]. The dissociation of ‘wanting’ and ‘liking’ is completed by our finding that preference for sucrose
over water is the same in DAT-tg rats as in wt rats, indicating that ‘liking’ is intact in these animals. As predicted, overall preference values differed significantly between time points. Analysis of linear contrasts confirmed that the increase in preference values was specific to the switch to sucrose solution. However, there was no significant main effect for genotype, nor an interaction between genotype and time. Crucially, genotype did not interact with the difference between BDmean and
TDmean. Hypodopaminergic DAT-tg rats thus showed an intact hedonic response, tallying with results from orofacial liking studies [2,10,12]. Interestingly, our results contradict a series of early pharmacological studies which reported decreased sucrose preference, and thus decreased ‘liking’, following the injection of DA antagonists [22–24]. However, these studies used an early form of the sucrose preference test, which failed to establish a preference baseline. Thus, it is possible that these studies confounded the reduction of absolute consumption in DA-deficient animals, which we now show to be a symptom of decreased ‘wanting’, with a non-existent reduction of relative sucrose preference, which we show remains stable in spite of impaired DA signalling. However, inherent discrepancies between pharmacological and genetic models targeting different players of the DA system may also contribute to such
ro of
divergent findings.
The following limitations should be considered. Animals of both genotypes developed a side bias during baseline days. However, the impact of this bias on preference index values used for statistical analysis was mitigated as follows: First, bottle positions were switched daily, so that each bottle
-p
occupied the preferred location on one baseline day. In line with this, preference values reversed themselves on BD2 (see Figure 2). Second, we used the mean over both BDs for the crucial linear
re
contrast between preference values on BDs and TDs, providing a more robust baseline for comparison. We also observed a significant decrease in sucrose consumption between TD1 and TD2.
lP
This occurred in both genotypes, suggesting a common cause (e.g. habituation or satiation). Furthermore, there was a floor effect for water consumption on TDs. However, such a steep
ur na
reduction is expected, with counterbalancing of bottle positions and rats’ innate exploratory behaviour leading to brief exploration and thus consumption of small amounts of water in both genotypes. This floor effect makes it inadvisable to use water consumption directly to calculate the preference index; instead, total consumption should be used [14,15]. Finally, further studies are needed to investigate whether a similar dissociation can be obtained
Jo
using different reward contingencies and whether genotype-specific differences in the microstructure of drinking behaviour correlate with the differences in ‘wanting’ observed here (i.e. number or size of lick clusters [25]).
Acknowledgements This work was supported by the TUD, the Federal Ministry of Education and Research, Germany [Grant 01EE1406A] and the Deutsche Forschungsgemeinschaft [DFG, WI 2140/4-1]. We thank Maike Kristin Lieser and Kristin Wogan for excellent support.
The authors declare no competing interests.
References [1]
P.N. Tobler, C.D. Fiorillo, W. Schultz, Adaptive coding of reward value by dopamine
neurons, Science. 307 (2005) 1642–1645. doi:10.1126/science.1105370. [2]
K.C. Berridge, T.E. Robinson, What is the role of dopamine in reward: hedonic impact,
reward learning, or incentive salience?, Brain Res. Brain Res. Rev. 28 (1998) 309–369.
‘wanting’,
K.C. Berridge, T.E. Robinson, J.W. Aldridge, Dissecting components of reward: ‘liking’, and
learning,
Current
Opinion
in
Pharmacology.
doi:10.1016/j.coph.2008.12.014. [4]
(2009)
65–73.
-p
R.A. Wise, The anhedonia hypothesis: Mark III, Behavioral and Brain Sciences. 8
re
(1985) 178. doi:10.1017/S0140525X00020306. [6]
9
R.A. Wise, Dopamine and reward: the anhedonia hypothesis 30 years on, Neurotox
Res. 14 (2008) 169–183. doi:10.1007/BF03033808. [5]
ro of
[3]
J.D. Salamone, M. Correa, The Mysterious Motivational Functions of Mesolimbic
Dopamine, Neuron. 76 (2012) 470–485. doi:10.1016/j.neuron.2012.10.021. J.A. Parkinson, J.W. Dalley, R.N. Cardinal, A. Bamford, B. Fehnert, G. Lachenal, N.
lP
[7]
Rudarakanchana, K.M. Halkerston, T.W. Robbins, B.J. Everitt, Nucleus accumbens dopamine
ur na
depletion impairs both acquisition and performance of appetitive Pavlovian approach behaviour: implications for mesoaccumbens dopamine function, Behav. Brain Res. 137 (2002) 149–163. [8]
S. Robinson, S.M. Sandstrom, V.H. Denenberg, R.D. Palmiter, Distinguishing Whether
Dopamine Regulates Liking, Wanting, and/or Learning About Rewards., Behavioral Neuroscience. 119 (2005) 5–15. doi:10.1037/0735-7044.119.1.5. M. Darvas, A.M. Wunsch, J.T. Gibbs, R.D. Palmiter, Dopamine dependency for
Jo
[9]
acquisition and performance of Pavlovian conditioned response, Proceedings of the National Academy of Sciences. 111 (2014) 2764–2769. doi:10.1073/pnas.1400332111. [10]
K.S. Smith, K.C. Berridge, J.W. Aldridge, Disentangling pleasure from incentive
salience and learning signals in brain reward circuitry, Proceedings of the National Academy of Sciences. 108 (2011) E255–E264. doi:10.1073/pnas.1101920108. [11]
M.E. Bardgett, M. Depenbrock, N. Downs, M. Points, L. Green, Dopamine modulates
effort-based decision making in rats., Behavioral Neuroscience. 123 (2009) 242–251.
doi:10.1037/a0014625. [12]
S. Peciña, B. Cagniard, K.C. Berridge, J.W. Aldridge, X. Zhuang, Hyperdopaminergic
mutant mice have higher “wanting” but not “liking” for sweet rewards, J. Neurosci. 23 (2003) 9395– 9402. [13]
K.C. Berridge, Measuring hedonic impact in animals and infants: microstructure of
affective taste reactivity patterns, Neurosci Biobehav Rev. 24 (2000) 173–198. [14]
T. Serchov, D. van Calker, K. Biber, Sucrose Preference Test to Measure Anhedonic
Behaviour in Mice, BIO-PROTOCOL. 6 (2016). doi:10.21769/BioProtoc.1958. [15]
M.-Y. Liu, C.-Y. Yin, L.-J. Zhu, X.-H. Zhu, C. Xu, C.-X. Luo, H. Chen, D.-Y. Zhu, Q.-G. Zhou,
Sucrose preference test for measurement of stress-induced anhedonia in mice, Nature Protocols.
[16]
ro of
13 (2018) 1686–1698. doi:10.1038/s41596-018-0011-z.
J.E. McCutcheon, M.F. Roitman, Mode of Sucrose Delivery Alters Reward-Related
Phasic Dopamine Signals in Nucleus Accumbens, ACS Chemical Neuroscience. (2018). doi:10.1021/acschemneuro.8b00262.
A. Hajnal, R. Norgren, Accumbens dopamine mechanisms in sucrose intake, Brain
-p
[17]
Research. 904 (2001) 76–84. doi:10.1016/S0006-8993(01)02451-9.
R. Hadar, H. Edemann-Callesen, C. Reinel, F. Wieske, M. Voget, E. Popova, R. Sohr, Y.
re
[18]
Avchalumov, J. Priller, C. van Riesen, I. Puls, M. Bader, C. Winter, Rats overexpressing the dopamine
lP
transporter display behavioral and neurobiological abnormalities with relevance to repetitive disorders, Scientific Reports. 6 (2016). doi:10.1038/srep39145. L.F. Jakubczak, Food and water intakes of rats as a function of strain, age,
ur na
[19]
temperature, and body weight, Physiol. Behav. 17 (1976) 251–258. [20]
K.C. Berridge, The debate over dopamine’s role in reward: the case for incentive
salience, Psychopharmacology. 191 (2007) 391–431. doi:10.1007/s00213-006-0578-x. [21]
N. Bernhardt, M.K. Lieser, E.-B. Hlusicka, B. Habelt, F. Wieske, H. Edemann-Callesen,
Jo
A. Garthe, C. Winter, Learning deficits in rats overexpressing the dopamine transporter, Scientific Reports. 8 (2018) 14173. doi:10.1038/s41598-018-32608-7. [22]
A. Towell, R. Muscat, P. Willner, Effects of pimozide on sucrose consumption and
preference, Psychopharmacology. 92 (1987). doi:10.1007/BF00177926. [23]
R. Muscat, P. Willner, Effects of dopamine receptor antagonists on sucrose
consumption and preference, Psychopharmacology (Berl.). 99 (1989) 98–102. [24]
S. Hsiao, G.P. Smith, Raclopride reduces sucrose preference in rats, Pharmacol.
Biochem. Behav. 50 (1995) 121–125.
[25] Quarterly
D.M. Dwyer, Licking and Liking: The Assessment of Hedonic Responses in Rodents, Journal
of
Experimental
Psychology.
65
(2012)
371–394.
doi:10.1080/17470218.2011.652969.
ro of
Figure legends
Figure 1: Absolute consumption of water and sucrose solution in ml/100g bodyweight for DAT-tg vs. wt animals across baseline days (BD) and testing days (TD). For baseline days, the sum of water consumption from the two bottles is displayed. Error bars represent ± 1 SE; * represents significant
Jo
ur na
lP
re
-p
main effects of time and genotype (p < .001).
Figure 2: Preference Index values for DAT-tg vs. wt rats. On baseline days (BD), the index reflects a simple bottle preference; on testing days (TD), the index reflects preference of sucrose solution over
water. Preference values above and below 0.5 indicate preference of one bottle over the other.
Jo
ur na
lP
re
-p
ro of
Error bars represent ± 1 SE; * represents the significant main effect of time (p < .001).