Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Factor analysis of attentional set-shifting performance in young and aged mice

Factor analysis of attentional set-shifting performance in young and aged mice Background: Executive dysfunction may play a major role in cognitive decline with aging because frontal lobe structures are particularly vulnerable to advancing age. Lesion studies in rats and mice have suggested that intradimensional shifts (IDSs), extradimensional shifts (EDSs), and reversal learning are mediated by the anterior cingulate cortex, the medial prefrontal cortex, and the orbitofrontal cortex, respectively. We hypothesized that the latent structure of cognitive performance would reflect functional localization in the brain and would be altered by aging. Methods: Young (4 months, n = 16) and aged (23 months, n = 18) C57BL/6N mice performed an attentional set- shifting task (ASST) that evaluates simple discrimination (SD), compound discrimination (CD), IDS, EDS, and reversal learning. The performance data were subjected to an exploratory factor analysis to extract the latent structures of ASST performance in young and aged mice. Results: The factor analysis extracted two- and three-factor models. In the two-factor model, the factor associated with SD and CD was clearly separated from the factor associated with the rest of the ASST stages in the young mice only. In the three-factor model, the SD and CD loaded on distinct factors. The three-factor model also showed a separation of factors associated with IDS, EDS, and CD reversal. However, the other reversal learning variables, ID reversal and ED reversal, had somewhat inconsistent factor loadings. Conclusions: The separation of performance factors in aged mice was less clear than in young mice, which suggests that aged mice utilize neuronal networks more broadly for specific cognitive functions. The result that the factors associated with SD and CD were separated in the three-factor model may suggest that the introduction of an irrelevant or distracting dimension results in the use of a new/orthogonal strategy for better discrimination. Background advancing age [4-7]. Executive control supports adaptation Aging causes declines in cognitive functions. For example, of behavioral responses according to the specific context older adults exhibit attentional deficits that can impact and requirements of varying situations. The Wisconsin their everyday lives [1] and they generally take longer to Card Sorting Test (WCST) has been widely used to assess process information than younger adults [2]. Impairment executive function in humans [8]. Subjects are required to of memory performance is one of the most noticeable adapt behavioral responses to choose the “correct” stimu- changes in aging; however, not all types of memory decline lus array based on sudden rule changes across multiple uniformly. Episodic memory is commonly impaired in the modalities. elderly, while implicit and semantic memory remains rela- A modification of the WCST, the intradimensional/ tively intact [3]. Executive dysfunction is likely to be a extradimensional (ID/ED) task, has been used to specifi- major contributor to the cognitive deficits that are cally assess attentional set-shifting abilities [9]. In this task, observed with aging because the frontal lobes that mediate subjects may initially learn that the color red is the main executive functions are particularly vulnerable to rule for discriminating between stimuli, and they must form an attentional set using color as the predictive stimu- lus dimension. The rule is then suddenly changed such * Correspondence: tanaka-s@sophia.ac.jp Department of Information & Communication Sciences, Sophia University, that red color no longer predicts the correct array, but Tokyo, Japan instead the color blue. This change is an “intradimensional Full list of author information is available at the end of the article © 2011 Tanaka et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 2 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 shift” (IDS). However at another point in the task, the rule domains (discrimination learning, reversal learning, and is changed such that the stimulus of number is the rele- set shifting) that are mediated by distinct brain regions, vant dimension for discrimination, and color no longer we hypothesized that performances in these domains has any predictive value. Thus subjects must rapidly shift would be dependent on distinct factors and that aging their attentional set to a new dimension (i.e. an extra- would modify the latent structure of ASST performance. dimensional shift, EDS), requiring both inhibition of the The preliminary results of this study have been presented in abstract form [24]. old rule (color) and acquisition of the new rule (numbers). The EDS is a core component of the WCST, and the main testing procedures of the ID/ED task were taken from the Methods Cambridge Neuropsychological Test Automated Battery Subjects and apparatus (CANTAB) [10,11]. This studyanalyzedthe data from apreviouslyreported Rat and, more recently, mouse versions of the atten- study [15]. The animals were either young (5 months, n = tional set-shifting task (ASST) have been developed 16) or aged (24 months, n = 18) male C57BL/6N mice [12,13]. Like the human versions, the rodent ASST has from the NIA aged rodent colony (Charles River Labora- multiple rule-shifting stages (i.e., IDS and EDS), simple tories, San Diego, California, USA). The experiment and compound discrimination, and reversal learning, started with 27 young and 23 aged mice. While 11 young usually using visual, tactile, and odor stimuli as the stimu- and 5 aged mice were excluded due to time constraints lus modalities. Just as in humans, animals need signifi- beyond experimental control, 16 young and 18 aged mice cantly larger numbers of trials to reach the EDS criterion completed all the stages of the ASST and were subjected compared with the IDS criterion, validating the task as a to the analyses. The test apparatus was an adapted perspex measure of set-shifting [12,13]. Again, similarly to aged cage (30 × 18 × 12 cm) (Figure 1). Two digging bowls humans, aged rats show impairments in EDS performance separated by a clear plastic panel were placed in each [14]. A recent ASST study with young and aged mice reported that, while the number of trials needed to reach the criterion did not differ between the young and aged groups, the EDS performance (measured by mean correct latency) was significantly longer and exhibited larger varia- bility in the aged animals [15]. Thus, aged animals appeared to sacrifice the speed of performance in order to maintain the accuracy of performance. Attentional set-shifting ability is impaired in patients with localized excisions of the frontal lobes, normal elderly controls [9], and patients with schizophrenia [16]. Patients with localized excisions of the frontal lobes have been shown to be selectively impaired in EDS but not IDS. In contrast, both temporal lobe and amygdalo-hippocampectomy patients were unimpaired in either EDS or IDS [9]. Other lesion studies in humans and non-human primates have confirmed that EDS requires an intact dorsolateral prefrontal cortex (DLPFC) [17,18]. Lesion analyses using rats and mice have suggested that IDS, EDS, and reversal learning are mediated by the anterior cingulate cortex; the medial prefrontal cortex (mPFC), the homolog of the primate DLPFC; and the orbitofrontal cortex (OFC), respectively [12,19-23]. These studies demonstrated consistency between humans and rodents in terms of the neural cir- cuits required for attentional set shifting. We investigated whether the data structure of ASST performance would reflect these functional specificities, and if these distinctions would be altered in aged animals. Here, we report the results of an exploratory factor analy- Figure 1 ThetestapparatususedinthemouseASSTstudy sis performed on a dataset taken from a study by Young (Young et al. 2010). et al. [15]. Because the ASST assesses different cognitive Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 3 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 quarter section. The bowls were placed on platforms (11 × needed to reach the criterion and by the mean correct 5 cm) that, in conjunction with odors, were used as cues latency. to guide the selections by the mice. Analysis Task We subjected the data to three statistical tests: Student’st- The ASST is comprised of 7 stages: simple discrimina- test; correlation analysis; and exploratory factor analysis. tion (SD); compound discrimination (CD); CD reversal Theseanalysesusedonlytheperformancescoresmea- (CDR); IDS; ID reversal (IDR); EDS; and ED reversal sured by the number of trials needed to reach the criter- (EDR) (Table 1). The SD stage required the animals to ion. The exploratory factor analysis used the observed distinguish the target stimulus from the irrelevant sti- scores of all seven stages of the ASST, assuming that two mulus, both being presented within a single dimension or three factors would be extracted. A four-factor model (e.g., odor). A second, irrelevant dimension (e.g., plat- was not successful. To obtain factor loadings, we used a form) was introduced in the CD stage and the mice promax rotation, a method of oblique rotation that allows were still required to discriminate between the two sti- factors to be correlated with each other. The promax rota- muli used in the SD task. In the CDR stage, the salien- tion was preferable to other rotation methods because the cies of the original stimuli were reversed with the assumption of interfactor correlations was reasonable and, relevant stimulus within the same dimension. The IDS more importantly, oblique rotation could reduce cross- stage required the mice to discriminate the target stimu- loadings [25]. The analyses were performed using R (The lus from the irrelevant stimulus in the unchanged rele- R Project for Statistical Computing), a software environ- vant dimension; however, novel stimulus sets were ment for statistical computing and graphics. introduced for both the relevant and the irrelevant dimensions. The IDR stage reversed the saliencies of the Results IDS stimuli. The EDS stage changed the relevant dimen- Correlations sion (e.g., from odor to platform). In the EDR task, the The correlation matrices of ASST performances in target and irrelevant stimuli were again reversed. Perfor- young and aged mice are listed in Table 2. The correla- mance was measured both by the number of trials tions of SD with other ASST variables are shown in Table 1 Descriptions of stages within the Attentional Set-Shifting Task Stage Description Dimensions Exemplar combinations Relevant Irrelevant Correct Incorrect Simple Discrimination Two stimuli are presented within one dimension (e.g., odor): one stimulus is Odor – O1 O2 (SD) the target and the other is irrelevant. Compound A second dimension (e.g., platform) is introduced but is irrelevant because the O1/P1 O2/P1 Discrimination (CD) subject is still required to discriminate between the two original stimuli. Odor Platform O1/P2 O2/P2 Compound The saliencies of the original stimuli are reversed: the target stimulus is now O2/P1 O1/P1 Discrimination irrelevant, while the irrelevant stimulus becomes the target. Reversal (CDR) Odor Platform O2/P2 O1/P2 Intradimensional Shift Novel stimuli are introduced for both dimensions. The target dimension (e.g., O3/P3 O4/P3 (IDS) odor) remains constant. Odor Platform O3/P4 O4/P4 Intradimensional The saliency of the novel stimuli is reversed: the target stimulus is now O4/P3 O3/P3 Reversal (IDR) irrelevant, while the irrelevant stimulus becomes the target. Odor Platform O4/P4 O3/P4 Extradimensional Shift Novel stimuli are introduced for both dimensions, and the target dimension is P5/O5 P6/O5 (EDS) now changed (e.g., from odor to platform). Platform Odor P5/O6 P6/O6 Extradimensional The saliency of the novel stimuli in the new target dimension is reversed: the P6/O5 P5/O5 Reversal (EDR) target stimulus is now irrelevant while the irrelevant becomes the target. Platform Odor P6/O6 P5/O6 One half of the mice received odor as the initial relevant dimension, while the other half received platform. Odors were ground ginger, nutmeg, garlic, coriander, thyme, and cinnamon. Platforms were sandpaper, wood, neoprene, metal wire, tile, and a scrubber. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 4 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Table 2 Pearson’s correlations for ASST performance Table 3 Factor loadings of the ASST stages for the two- factor models with promax rotation Young SD CD CDR IDS IDR EDS EDR Young Factor1 Factor2 Aged Factor1 Factor2 SD 1 0.41 -0.48 -0.17 -0.13 0.03 0.01 SD 1.006 SD 0.708 -0.277 CD 0.41 1 -0.19 0.17 -0.26 -0.13 -0.08 CD 0.409 CD 0.283 0.550 CDR -0.48 -0.19 1 0.15 -0.05 0.23 0.08 CDR -0.475 CDR -0.602 0 IDS -0.17 0.17 0.15 1 -0.10 0.07 0.48 IDS 0.475 IDS -0.156 1.021 IDR -0.13 -0.26 -0.05 -0.10 1 0 -0.30 IDR -0.310 IDR 0.308 0.506 EDS 0.03 -0.13 0.23 0.07 0 1 0.19 EDS 0.194 EDS -0.640 0 EDR 0.01 -0.08 0.08 0.48 -0.30 0.19 1 EDR 1.006 EDR -0.343 0.551 Aged SD CD CDR IDS IDR EDS EDR Cutoff = 0.15. SD 1 -0.05 -0.45 -0.22 0.16 -0.48 -0.18 CD -0.05 1 -0.37 0.56 0.59 -0.23 -0.02 CDR -0.45 -0.37 1 -0.04 -0.06 0.34 0.22 factor. Because the IDS was the first shift in the ASST, IDS -0.22 0.56 -0.04 1 0.52 -0.13 0.52 the second factor is termed the “shifting factor.” IDR 0.16 0.59 -0.06 0.52 1 -0.37 0.18 EDS -0.48 -0.23 0.34 -0.13 -0.37 1 0.04 Three-factor model EDR -0.18 -0.02 0.22 0.52 0.18 0.04 1 The three-factor model is shown in Table 4 and Figure 4. In this model, SD and CD loaded on distinct factors in Figure 2. Both the young and the aged groups showed both young and aged mice. Because the difference significant negative correlations between SD and CDR. between SD and CD tasks was the absence or presence of The young group, but not the aged group, showed a distractors, discrimination in a distracting environment positive correlation between SD and CD. The aged can be functionally differentiated from discrimination group, but not the young group, showed a negative cor- without distractors. Therefore, the first two factors of relation between SD and EDS. this three-factor model are termed the “discrimination factor” and the “undistracted-discrimination factor.” The Two-factor model third factor was similar to the second factor in the two- The extracted two-factor model is shown in Table 3 and factor model in that both IDS and EDR had high factor Figure 3. In young mice, SD, CD, and CDR had high loadings in both young and aged mice. EDS did not have factor loadings on the first factor, which is termed the consistent loadings between the groups. We term the “discrimination factor.” The remaining stages (IDS, IDR, third factor the “shifting factor” because IDS was the first EDS, and EDR) had high factor loadings on the second shifting stage in the ASST. As with the two-factor model, factor. The aged mice did not show such a clear separa- the factor loading profiles were less clear in aged mice tion, but the IDS had the highest loading on the second than in young mice. Distractibility Previous studieshaveshown that thepresenceof dis- tractors increases error rates during cognitive tasks in aged subjects [26,27]. We estimated the aging effects on distractibility by subtracting SD from CD, to approxi- mate the increased number of trials needed to learn the discrimination when distracters are present. The aged group had a higher mean value of “CD minus SD” (mean: young = -5.56, aged = 0.28), with a trend toward a significant effect of age (p = 0.065, one-tailed t-test). Discussion The two-factor model clearly showed a separation of performance during discrimination stages (SD, CD, and CDR) from the remaining stages (IDS, IDR, EDS, and EDR) in young mice only. CDR had a negative factor loading, whereas SD had a positive factor loading in Figure 2 Pearson’s correlations of SD with other ASST both groups. The correlation analysis also confirmed variables. negative correlations between SD and CDR (Figure 2). Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 5 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Figure 3 Factor loading profiles across the ASST stages for the two-factor models with promax rotation. A1and A2:Two factorsin young mice. The cumulative variance is 0.41 and p = 0.90. B1 and B2: Two factors in aged mice. The cumulative variance is 0.51 and p = 0.38. Table 4 Factor loadings of the ASST stages for the three-factor models with promax rotation Young Factor1 Factor2 Factor3 Aged Factor1 Factor2 Factor3 SD 0.991 SD 1.066 -0.271 CD 0.926 CD -0.229 1.102 CDR -0.469 CDR -0.351 -0.376 0.251 IDS -0.221 0.539 IDS 0.405 0.675 IDR -0.361 IDR 0.165 0.474 0.310 EDS -0.219 0.169 EDS -0.511 EDR -0.476 1.012 EDR -0.270 0.811 Cutoff = 0.15. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 6 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Figure 4 Factor loading profiles across the ASST stages for the three-factor models with promax rotation. A1, A2 and A3: Three factors in young mice. The cumulative variance is 0.56 and p = 0.78. B1, B2 and B3: Three factors in aged mice. The cumulative variance is 0.69 and p = 0.55. This negative relationship is interesting because the CDR in both groups (young = -0.477; aged = -0.455), CDR involves discrimination but is also the first reversal which suggests that mice with better SD performed learning stage in the ASST. The negative correlation worse in CDR, irrespective of age. between SD and CDR indicates that mice with high SD The three-factor model further separated CD from SD, performance performed poorly with reversal learning. which suggests that the addition of irrelevant stimuli There might be a trade-off relationship between simple required a new strategy for discrimination. This separa- discrimination and reversal learning. Therefore, the “dis- tion was shown in both groups of mice. Notably, the crimination factor,” on which the SD, CD, and CDR factor loadings showed that the separation of CD from load, might also include a tendency for “perseveration.” SD was enhanced in aged mice when compared with The reciprocal SD-CDR relationship observed in our young mice (Table 4). This clearer separation may be mouse study is consistent with a similar study using due to increased distractibility in aged mice. The degree ASST with young and aged rats [14]. Those researchers of distractibility was estimated by calculating the differ- used 4- to 5-month-old and 27- to 28-month-old male ence between SD and CD. Higher values of “CD minus Long-Evans rats (n = 26) in a similar ASST study, but SD” appeared to be indicative of higher distractibility. no CDR or EDR stage was employed. In the rat study, The aged group had a higher mean value of “CD minus IDR was the first reversal task, while CDR was the first SD,” with a trend toward significance between the two in the mouse study that we analyzed. There was a nega- groups. This finding indicates that, unlike young mice, tive correlation between SD and IDR in the rat study, a aged mice were affected by the addition of an irrelevant stimulus. These data are consistent with human studies, finding that is similar to our study. Their results differ from ours in that the correlation was negative only in indicating that aged mice have higher distractibility or the young rats (young = -0.728; aged = -0.155). In con- lower robustness of executive functioning when com- trast, we found a negative correlation between SD and pared with young mice [28-30]. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 7 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Because IDS, EDS, and reversal learning may be subjects who were performing at comparable levels. mediated by distinct brain regions [12,19-21,23], we This result suggests that alternate networks are being expected that these variables would load on distinct fac- recruited in older subjects [36-39] and that the elderly tors. IDS, EDS, and CDR did indeed load on distinct fac- use a different strategy when performing a cognitive tors in both groups. The separation of these factors is task [39-42]. Potential neurochemical differences that meaningful because CDR was the first reversal learning might account for alterations in aging performance in ASST are numerous. In young animals, dopamine sig- stage in the ASST. However, the factor loadings of EDS naling is critical for EDS performance. For example, were fairly low for all factors extracted. It is unclear why amphetamine-sensitized rats or rats that acquired CDR, IDR, and EDR did not load on a single factor; there- fore, the extracted three-factor model had unexpected fac- amphetamine self-administration exhibited impaired tor loadings. There are several possible reasons for these EDS but not IDS performance [43-45]. Treatment with results. First, the mice may not have formed an attentional tolcapone, a catechol-O-methyltransferase inhibitor that set; however, this explanation is unlikely because both the increases dopamine and norepinephrine levels in the young and the aged mice showed significant increases in frontal cortex, improved EDS performance in rats [46]. the number of trials needed to reach the criterion for the This latter study suggests that catecholamine transmis- EDS stage when compared with the IDS stage, regardless sion in the rat mPFC is critically involved in set-shifting of the initial perceptual dimension [15]. This finding sug- functions. A recent study in rats has suggested that gests the successful formation of an attentional set in this remodeling of mesocortical dopaminergic fibers is group. Second, despite the hypothesized functional locali- involved in age-associated cognitive decline [47]. Other zation in the brain, these regions likely interact to perform studies have also demonstrated a specific involvement of the ASST stages. It has been suggested that cognitive pro- noradrenergic transmission in ASST performance cessing is mediated by a network [31-34]. The network [48-51], while serotonin modulation may affect reversal involved in ASST performance would include the mPFC, learning components in this task but not EDS perfor- OFC, and other regions, and interaction among the mance [52]. Interestingly, administration of naltrexone involved regions in the network might have blurred the (2 mg/kg, i.p.), an opioid antagonist, has been shown to separation of factors. The capability of factor analysis to reverse aging-related deficits in EDS [53], suggesting extract a functional structure in such a distributed net- that altered endorphin signaling plays a role in aging- work is an interesting issue that needs to be addressed. induced decline in set shifting. Research on the complex underpinnings of age-related deficits in attentional set Third, a small sample size (n = 16 for young mice and n = shifting is in its infancy; however, these data suggest 18 for aged mice) was used in this analysis. The same ana- lysis with a larger sample size should be performed to that there are a number of potential systems that could determine whether the sample size was a limiting factor. It influence aging effects on this task. would be interesting to see whether a factor analysis with a larger sample size would result in a better model with Conclusion more consistent factor loadings. Finally, there is evidence The exploratory factor analysis applied to the mouse in rats and mice that reversal learning improves over ASST data extracted two- and three-factor models. The repeated exposure, with fewer trials required at EDR vs. two-factor model clearly separated discrimination perfor- CDR [15,23,35]. This ‘learning to reverse’ might incorpo- mance (SD and CD) from the remaining ASST stages in rate other neuroanatomical structures that are not specific young mice. This tendency was obscured in the aged to reversal learning, hence the lack of commonality in mice, suggesting that they utilize a less selective network reversal loading. in the brain for cognitive functions. The three-factor In our study, the overall deterioration in the cognitive model further separated CD from SD. The result that the function of aged mice was not marked. However, while factors associated with SD and CD were separated in the the number of trials needed to reach the criterion was three-factor model may suggest that the introduction of not different between groups for EDS, aged mice exhib- an irrelevant or distracting dimension results in the use ited significantly longer mean correct latencies during of a new/orthogonal strategy for better discrimination. this task (p < 0.05) [15]. This result indicates that the This model, however, did not show consistent separation aged mice have reduced processing speed only during of the factors associated with IDS, EDS, and reversal the EDS stage. Thus, when compared with young mice, learning in either young or aged mice, contrary to the aged mice might require more time to perform the EDS hypothesis that these functions are mediated by distinct (i.e. speed-accuracy trade off) [15]. Magnetic resonance brain regions. Whether the inconsistent separation of fac- imaging studies in humans performing an executive tors was due to possible interaction among the brain function task reported a broader activation of brain regions or to the small sample size needs to be clarified regions in older subjects when compared with younger by a future analysis using a larger sample size. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 8 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 15. Young JW, Powell SB, Geyer MA, Jeste DV, Risbrough VB: The mouse Acknowledgements attentional set-shifting task: A method for assaying successful cognitive The original ASST mouse study was funded by Stein Institute of Research on aging? Cognitive, Affective, & Behavioral Neuroscience 2010, 10:243-251, Aging (VBR and JWY) and MH052885 (MAG). This analytical study was PMID: 20498348. supported by the Human Informatics Research Center at Sophia University 16. Jazbec S, Pantelis C, Robbins T, Weickert T, Weinberger DR, Goldberg TE: (ST). Intra-dimensional/extra-dimensional set-shifting performance in schizophrenia: Impact of distractors. Schizophrenia Research 2007, Author details 89:339-349, PMID: 17055703. Department of Information & Communication Sciences, Sophia University, 17. Dias R, Robbins TW, Roberts AC: Dissociable forms of inhibitory control Tokyo, Japan. Department of Psychiatry, University of California, San Diego, within prefrontal cortex with an analog of the Wisconsin Card Sort Test: USA. restriction to novel situations and independence from “on-line” processing. J Neurosci 1997, 17(23):9285-9297, PMID: 9364074. Authors’ contributions 18. Rogers RD, Andrews TC, Grasby PM, Brooks DJ, Robbins TW: Contrasting ST carried out the whole analysis. JWY, JEG, MAG, and VBR carried out the cortical and subcortical activations produced by attentional-set shifting original animal experiment. All authors read and approved the final and reversal learning in humans. J Cogn Neurosci 2000, 12(1):142-162, manuscript. PMID: 10769312. 19. Ng C-W, Noblejas MI, Rodefer JS, Smith CB, Poremba A: Double Competing interests Dissociation of Attentional Resources: Prefrontal Versus Cingulate The authors declare that they have no competing interests. Cortices. J Neurosci 2007, 27:12123-12131, PMID:17989278. 20. Bissonette GB, Martins GJ, Franz TM, Harper ES, Schoenbaum G, Powell EM: Received: 1 December 2010 Accepted: 12 August 2011 Double Dissociation of the Effects of Medial and Orbital Prefrontal Published: 12 August 2011 Cortical Lesions on Attentional and Affective Shifts in Mice. J Neurosci 2008, 28:11124-11130, PMID: 18971455. References 21. Ragozzino ME: The contribution of the medial prefrontal cortex, 1. Glisky EL: Changes in cognitive function in human aging. In Brain aging: orbitofrontal cortex, and dorsomedial striatum to behavioral flexibility. models, methods, and mechanisms. Edited by: Riddle DR. Boca Raton: CRC Ann N Y Acad Sci 2007, 1121:355-375, PMID: 17698989. Press; 2007:. 22. Dalley JW, Cardinal RN, Robbins TW: Prefrontal executive and cognitive 2. Verhaeghen P, Cerella J: Everything we know about aging and response functions in rodents: neural and neurochemical substrates. Neurosci time. In Handbook of cognitive aging: interdisciplinary perspectives. Edited by: Biobehav Rev 2004, 28(7):771-784, PMID: 15555683. Hofer SM, Alwin DF. Los Angeles: Sage Publications; 2008:134-150. 23. McAlonan K, Brown VJ: Orbital prefrontal cortex mediates reversal 3. Old SR, Naveh-Benjamin M: Age-related changes in memory. In Handbook learning and not attentional set shifting in the rat. Behav Brain Res 2003, of cognitive aging: interdisciplinary perspectives. Edited by: Hofer SM, Alwin 146:97-103, PMID: 14643463. DF. Los Angeles: Sage Publications; 2008:151-167. 24. Tanaka S, Risbrough V, Young J, Gresack J, Geyer M: Factor analysis of 4. Raz N, Rodrigue KM: Differential aging of the brain: Patterns, cognitive attentional set-shifting task performance of aged and young mice. correlates and modifiers. Neuroscience & Biobehavioral Reviews 2006, Society for Neuroscience Abstracts 2009, 580.023. 30:730-748, PMID: 16919333. 25. Brown TA: Confirmatory factor analysis for applied research New York: The 5. Grady CL: Cognitive Neuroscience of Aging. Ann N Y Acad Sci 2008, Guilford Press; 2006. 1124:127-144, PMID: 18400928. 26. Chao L, Knight R: Prefrontal deficits in attention and inhibitory control 6. Luszcz M, Lane A: Executive function in cognitive, neuropsychological, with aging. Cereb Cortex 1997, 7:63-69, PMID: 9023433. and clinical aging. In Handbook of cognitive aging: interdisciplinary 27. Lustig C, Hasher L, Tonev ST: Distraction as a determinant of processing perspectives. Edited by: Hofer SM, Alwin DF. Los Angeles: Sage Publications; speed. Psychonomic Bulletin & Review 2006, 13:619-625, PMID: 17201361. 2008:193-206. 28. Clapp WC, Gazzaley A: Distinct mechanisms for the impact of distraction 7. Moss MB, Moore TL, Schettler SP, Killiany R, Rosene D: Successful vs. and interruption on working memory in aging. Neurobiology of Aging Unsuccessful Aging in the Rhesus Monkey. In Brain Aging: Models, 2010, PMID: 20144492. Methods, and Mechanisms. Edited by: Riddle DR. Boca Raton (FL): CRC Press; 29. Healey MK, Campbell KL, Hasher L: Cognitive aging and increased 2007:. distractibility: Costs and potential benefits. Progress in Brain Research 2008, 8. Eling P, Derckx K, Maes R: On the historical and conceptual background 169:353-363, PMID: 18394486. of the Wisconsin Card Sorting Test. Brain and Cognition 2008, 67:247-253, 30. Machado L, Devine A, Wyatt N: Distractibility with advancing age and PMID: 18328609. Parkinson’s disease. Neuropsychologia 2009, 47:1756-1764, PMID: 9. Owen AM, Roberts AC, Polkey CE, Sahakian BJ, Robbins TW: Extra- dimensional versus intra-dimensional set shifting performance following 31. Bullmore E, Sporns O: Complex brain networks: graph theoretical analysis frontal lobe excisions, temporal lobe excisions or amygdalo- of structural and functional systems. Nat Rev Neurosci 2009, 10:186-198, hippocampectomy in man. Neuropsychologia 1991, 29:993-1006, PMID: PMID: 19190637. 32. Colom R, Karama S, Jung RE, Haier RJ: Human intelligence and brain 10. Downes JJ, Roberts AC, Sahakian BJ, Evenden JL, Morris RG, Robbins TW: networks. Dialogues Clin Neurosci 2010, 12:489-501, PMID:21319494. Impaired extra-dimensional shift performance in medicated and 33. Green JJ, Doesburg SM, Ward LM, McDonald JJ: Electrical Neuroimaging of unmedicated Parkinson’s disease: evidence for a specific attentional Voluntary Audiospatial Attention: Evidence for a Supramodal Attention dysfunction. Neuropsychologia 1989, 27:1329-1343, PMID: 2615934. Control Network. J Neurosci 2011, 31:3560-3564, PMID: 21389212. 11. Robbins TW, James M, Owen AM, Sahakian BJ, McInnes L, Rabbitt P: 34. Naghavi HR, Nyberg L: Common fronto-parietal activity in attention, Cambridge Neuropsychological Test Automated Battery (CANTAB): a memory, and consciousness: Shared demands on integration? factor analytic study of a large sample of normal elderly volunteers. Consciousness and Cognition 2005, 14:390-425, PMID:15950889. Dementia 1994, 5:266-281, PMID: 7951684. 35. Boulougouris V, Dalley JW, Robbins TW: Effects of orbitofrontal, infralimbic 12. Birrell JM, Brown VJ: Medial frontal cortex mediates perceptual and prelimbic cortical lesions on serial spatial reversal learning in the attentional set shifting in the rat. J Neurosci 2000, 20:4320-4324, PMID: rat. Behav Brain Res 2007, 179:219-228, PMID:17337305. 36. Milham MP, Erickson KI, Banich MT, Kramer AF, Webb A, Wszalek T, 13. Garner JP, Thogerson CM, Wurbel H, Murray JD, Mench JA: Animal Cohen NJ: Attentional Control in the Aging Brain: Insights from an fMRI neuropsychology: Validation of the Intra-Dimensional Extra-Dimensional Study of the Stroop Task. Brain and Cognition 2002, 49:277-296, PMID: set shifting task for mice. Behav Brain Res 2006, 173:53-61, PMID: 37. Grady CL, Protzner AB, Kovacevic N, Strother SC, Afshin-Pour B, 14. Barense MD, Fox MT, Baxter MG: Aged Rats Are Impaired on an Wojtowicz M, Anderson JAE, Churchill N, McIntosh AR: A Multivariate Attentional Set-Shifting Task Sensitive to Medial Frontal Cortex Damage Analysis of Age-Related Differences in Default Mode and Task-Positive in Young Rats. Learning & Memory 2002, 9:191-201, PMID: 12177232. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 9 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Networks across Multiple Cognitive Domains. Cereb Cortex 2009, 20:1432-1447, PMID: 19789183. 38. Prakash RS, Erickson KI, Colcombe SJ, Kim JS, Voss MW, Kramer AF: Age- related differences in the involvement of the prefrontal cortex in attentional control. Brain and Cognition 2009, 71:328-335, PMID: 19699019. 39. Venkatraman VK, Aizenstein H, Guralnik J, Newman AB, Glynn NW, Taylor C, Studenski S, Launer L, Pahor M, Williamson J, Rosano C: Executive control function, brain activation and white matter hyperintensities in older adults. NeuroImage 2010, 49:3436-3442, PMID: 19922803. 40. Grady CL, McIntosh AR, Beig S, Keightley ML, Burian H, Black SE: Evidence from Functional Neuroimaging of a Compensatory Prefrontal Network in Alzheimer’s Disease. J Neurosci 2003, 23:986-993, PMID: 12574428. 41. Reuter-Lorenz PA: New visions of the aging mind and brain. Trends in Cognitive Sciences 2002, 6:394-400, PMID: 12200182. 42. Reuter-Lorenz PA, Lustig C: Brain aging: reorganizing discoveries about the aging mind. Current Opinion in Neurobiology 2005, 15:245-251, PMID: 43. Featherstone RE, Rizos Z, Kapur S, Fletcher PJ: A sensitizing regimen of amphetamine that disrupts attentional set-shifting does not disrupt working or long-term memory. Behav Brain Res 2008, 189:170-179, PMID: 44. Fletcher P, Tenn C, Rizos Z, Lovic V, Kapur S: Sensitization to amphetamine, but not PCP, impairs attentional set shifting: reversal by a D1 receptor agonist injected into the medial prefrontal cortex. Psychopharmacology 2005, 183:190-200, PMID: 16220338. 45. Parsegian A, Glen WB, Lavin A, See RE: Methamphetamine Self- Administration Produces Attentional Set-Shifting Deficits and Alters Prefrontal Cortical Neurophysiology in Rats. Biological Psychiatry 2011, 69:253-259, PMID: 21051037. 46. Tunbridge EM, Bannerman DM, Sharp T, Harrison PJ: Catechol-o- methyltransferase inhibition improves set-shifting performance and elevates stimulated dopamine release in the rat prefrontal cortex. J Neurosci 2004, 24(23):5331-5335, PMID: 15190105. 47. Allard S, Gosein V, Cuello AC, Ribeiro-da-Silva A: Changes with aging in the dopaminergic and noradrenergic innervation of rat neocortex. Neurobiology of Aging 2010, PMID: 20096955. 48. Lapiz MDS, Morilak DA: Noradrenergic modulation of cognitive function in rat medial prefrontal cortex as measured by attentional set shifting capability. Neuroscience 2006, 137:1039-1049, PMID: 16298081. 49. Lapiz MDS, Bondi CO, Morilak DA: Chronic Treatment with Desipramine Improves Cognitive Performance of Rats in an Attentional Set-Shifting Test. Neuropsychopharmacology 2007, 32:1000-1010, PMID: 17077810. 50. McGaughy J, Ross RS, Eichenbaum H: Noradrenergic, but not cholinergic, deafferentation of prefrontal cortex impairs attentional set-shifting. Neuroscience 2008, 153:63-71, PMID: 18355972. 51. Newman L, Darling J, McGaughy J: Atomoxetine reverses attentional deficits produced by noradrenergic deafferentation of medial prefrontal cortex. Psychopharmacology 2008, 200:39-50, PMID: 18568443. 52. Clarke HF, Walker SC, Crofts HS, Dalley JW, Robbins TW, Roberts AC: Prefrontal serotonin depletion affects reversal learning but not attentional set shifting. J Neurosci 2005, 25(2):532-538, PMID: 15647499. 53. Rodefer JS, Nguyen TN: Naltrexone reverses age-induced cognitive deficits in rats. Neurobiology of Aging 2008, 29:309-313, PMID:17098330. doi:10.1186/1744-9081-7-33 Cite this article as: Tanaka et al.: Factor analysis of attentional set- shifting performance in young and aged mice. Behavioral and Brain Functions 2011 7:33. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral and Brain Functions Springer Journals

Factor analysis of attentional set-shifting performance in young and aged mice

Download PDF
9 pages

Loading next page...
 
/lp/springer-journals/factor-analysis-of-attentional-set-shifting-performance-in-young-and-YBXJwTcAAA

References (60)

Publisher
Springer Journals
Copyright
Copyright © 2011 by Tanaka et al; licensee BioMed Central Ltd.
Subject
Biomedicine; Neurosciences; Neurology; Behavioral Therapy; Psychiatry
eISSN
1744-9081
DOI
10.1186/1744-9081-7-33
pmid
21838910
Publisher site
See Article on Publisher Site

Abstract

Background: Executive dysfunction may play a major role in cognitive decline with aging because frontal lobe structures are particularly vulnerable to advancing age. Lesion studies in rats and mice have suggested that intradimensional shifts (IDSs), extradimensional shifts (EDSs), and reversal learning are mediated by the anterior cingulate cortex, the medial prefrontal cortex, and the orbitofrontal cortex, respectively. We hypothesized that the latent structure of cognitive performance would reflect functional localization in the brain and would be altered by aging. Methods: Young (4 months, n = 16) and aged (23 months, n = 18) C57BL/6N mice performed an attentional set- shifting task (ASST) that evaluates simple discrimination (SD), compound discrimination (CD), IDS, EDS, and reversal learning. The performance data were subjected to an exploratory factor analysis to extract the latent structures of ASST performance in young and aged mice. Results: The factor analysis extracted two- and three-factor models. In the two-factor model, the factor associated with SD and CD was clearly separated from the factor associated with the rest of the ASST stages in the young mice only. In the three-factor model, the SD and CD loaded on distinct factors. The three-factor model also showed a separation of factors associated with IDS, EDS, and CD reversal. However, the other reversal learning variables, ID reversal and ED reversal, had somewhat inconsistent factor loadings. Conclusions: The separation of performance factors in aged mice was less clear than in young mice, which suggests that aged mice utilize neuronal networks more broadly for specific cognitive functions. The result that the factors associated with SD and CD were separated in the three-factor model may suggest that the introduction of an irrelevant or distracting dimension results in the use of a new/orthogonal strategy for better discrimination. Background advancing age [4-7]. Executive control supports adaptation Aging causes declines in cognitive functions. For example, of behavioral responses according to the specific context older adults exhibit attentional deficits that can impact and requirements of varying situations. The Wisconsin their everyday lives [1] and they generally take longer to Card Sorting Test (WCST) has been widely used to assess process information than younger adults [2]. Impairment executive function in humans [8]. Subjects are required to of memory performance is one of the most noticeable adapt behavioral responses to choose the “correct” stimu- changes in aging; however, not all types of memory decline lus array based on sudden rule changes across multiple uniformly. Episodic memory is commonly impaired in the modalities. elderly, while implicit and semantic memory remains rela- A modification of the WCST, the intradimensional/ tively intact [3]. Executive dysfunction is likely to be a extradimensional (ID/ED) task, has been used to specifi- major contributor to the cognitive deficits that are cally assess attentional set-shifting abilities [9]. In this task, observed with aging because the frontal lobes that mediate subjects may initially learn that the color red is the main executive functions are particularly vulnerable to rule for discriminating between stimuli, and they must form an attentional set using color as the predictive stimu- lus dimension. The rule is then suddenly changed such * Correspondence: tanaka-s@sophia.ac.jp Department of Information & Communication Sciences, Sophia University, that red color no longer predicts the correct array, but Tokyo, Japan instead the color blue. This change is an “intradimensional Full list of author information is available at the end of the article © 2011 Tanaka et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 2 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 shift” (IDS). However at another point in the task, the rule domains (discrimination learning, reversal learning, and is changed such that the stimulus of number is the rele- set shifting) that are mediated by distinct brain regions, vant dimension for discrimination, and color no longer we hypothesized that performances in these domains has any predictive value. Thus subjects must rapidly shift would be dependent on distinct factors and that aging their attentional set to a new dimension (i.e. an extra- would modify the latent structure of ASST performance. dimensional shift, EDS), requiring both inhibition of the The preliminary results of this study have been presented in abstract form [24]. old rule (color) and acquisition of the new rule (numbers). The EDS is a core component of the WCST, and the main testing procedures of the ID/ED task were taken from the Methods Cambridge Neuropsychological Test Automated Battery Subjects and apparatus (CANTAB) [10,11]. This studyanalyzedthe data from apreviouslyreported Rat and, more recently, mouse versions of the atten- study [15]. The animals were either young (5 months, n = tional set-shifting task (ASST) have been developed 16) or aged (24 months, n = 18) male C57BL/6N mice [12,13]. Like the human versions, the rodent ASST has from the NIA aged rodent colony (Charles River Labora- multiple rule-shifting stages (i.e., IDS and EDS), simple tories, San Diego, California, USA). The experiment and compound discrimination, and reversal learning, started with 27 young and 23 aged mice. While 11 young usually using visual, tactile, and odor stimuli as the stimu- and 5 aged mice were excluded due to time constraints lus modalities. Just as in humans, animals need signifi- beyond experimental control, 16 young and 18 aged mice cantly larger numbers of trials to reach the EDS criterion completed all the stages of the ASST and were subjected compared with the IDS criterion, validating the task as a to the analyses. The test apparatus was an adapted perspex measure of set-shifting [12,13]. Again, similarly to aged cage (30 × 18 × 12 cm) (Figure 1). Two digging bowls humans, aged rats show impairments in EDS performance separated by a clear plastic panel were placed in each [14]. A recent ASST study with young and aged mice reported that, while the number of trials needed to reach the criterion did not differ between the young and aged groups, the EDS performance (measured by mean correct latency) was significantly longer and exhibited larger varia- bility in the aged animals [15]. Thus, aged animals appeared to sacrifice the speed of performance in order to maintain the accuracy of performance. Attentional set-shifting ability is impaired in patients with localized excisions of the frontal lobes, normal elderly controls [9], and patients with schizophrenia [16]. Patients with localized excisions of the frontal lobes have been shown to be selectively impaired in EDS but not IDS. In contrast, both temporal lobe and amygdalo-hippocampectomy patients were unimpaired in either EDS or IDS [9]. Other lesion studies in humans and non-human primates have confirmed that EDS requires an intact dorsolateral prefrontal cortex (DLPFC) [17,18]. Lesion analyses using rats and mice have suggested that IDS, EDS, and reversal learning are mediated by the anterior cingulate cortex; the medial prefrontal cortex (mPFC), the homolog of the primate DLPFC; and the orbitofrontal cortex (OFC), respectively [12,19-23]. These studies demonstrated consistency between humans and rodents in terms of the neural cir- cuits required for attentional set shifting. We investigated whether the data structure of ASST performance would reflect these functional specificities, and if these distinctions would be altered in aged animals. Here, we report the results of an exploratory factor analy- Figure 1 ThetestapparatususedinthemouseASSTstudy sis performed on a dataset taken from a study by Young (Young et al. 2010). et al. [15]. Because the ASST assesses different cognitive Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 3 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 quarter section. The bowls were placed on platforms (11 × needed to reach the criterion and by the mean correct 5 cm) that, in conjunction with odors, were used as cues latency. to guide the selections by the mice. Analysis Task We subjected the data to three statistical tests: Student’st- The ASST is comprised of 7 stages: simple discrimina- test; correlation analysis; and exploratory factor analysis. tion (SD); compound discrimination (CD); CD reversal Theseanalysesusedonlytheperformancescoresmea- (CDR); IDS; ID reversal (IDR); EDS; and ED reversal sured by the number of trials needed to reach the criter- (EDR) (Table 1). The SD stage required the animals to ion. The exploratory factor analysis used the observed distinguish the target stimulus from the irrelevant sti- scores of all seven stages of the ASST, assuming that two mulus, both being presented within a single dimension or three factors would be extracted. A four-factor model (e.g., odor). A second, irrelevant dimension (e.g., plat- was not successful. To obtain factor loadings, we used a form) was introduced in the CD stage and the mice promax rotation, a method of oblique rotation that allows were still required to discriminate between the two sti- factors to be correlated with each other. The promax rota- muli used in the SD task. In the CDR stage, the salien- tion was preferable to other rotation methods because the cies of the original stimuli were reversed with the assumption of interfactor correlations was reasonable and, relevant stimulus within the same dimension. The IDS more importantly, oblique rotation could reduce cross- stage required the mice to discriminate the target stimu- loadings [25]. The analyses were performed using R (The lus from the irrelevant stimulus in the unchanged rele- R Project for Statistical Computing), a software environ- vant dimension; however, novel stimulus sets were ment for statistical computing and graphics. introduced for both the relevant and the irrelevant dimensions. The IDR stage reversed the saliencies of the Results IDS stimuli. The EDS stage changed the relevant dimen- Correlations sion (e.g., from odor to platform). In the EDR task, the The correlation matrices of ASST performances in target and irrelevant stimuli were again reversed. Perfor- young and aged mice are listed in Table 2. The correla- mance was measured both by the number of trials tions of SD with other ASST variables are shown in Table 1 Descriptions of stages within the Attentional Set-Shifting Task Stage Description Dimensions Exemplar combinations Relevant Irrelevant Correct Incorrect Simple Discrimination Two stimuli are presented within one dimension (e.g., odor): one stimulus is Odor – O1 O2 (SD) the target and the other is irrelevant. Compound A second dimension (e.g., platform) is introduced but is irrelevant because the O1/P1 O2/P1 Discrimination (CD) subject is still required to discriminate between the two original stimuli. Odor Platform O1/P2 O2/P2 Compound The saliencies of the original stimuli are reversed: the target stimulus is now O2/P1 O1/P1 Discrimination irrelevant, while the irrelevant stimulus becomes the target. Reversal (CDR) Odor Platform O2/P2 O1/P2 Intradimensional Shift Novel stimuli are introduced for both dimensions. The target dimension (e.g., O3/P3 O4/P3 (IDS) odor) remains constant. Odor Platform O3/P4 O4/P4 Intradimensional The saliency of the novel stimuli is reversed: the target stimulus is now O4/P3 O3/P3 Reversal (IDR) irrelevant, while the irrelevant stimulus becomes the target. Odor Platform O4/P4 O3/P4 Extradimensional Shift Novel stimuli are introduced for both dimensions, and the target dimension is P5/O5 P6/O5 (EDS) now changed (e.g., from odor to platform). Platform Odor P5/O6 P6/O6 Extradimensional The saliency of the novel stimuli in the new target dimension is reversed: the P6/O5 P5/O5 Reversal (EDR) target stimulus is now irrelevant while the irrelevant becomes the target. Platform Odor P6/O6 P5/O6 One half of the mice received odor as the initial relevant dimension, while the other half received platform. Odors were ground ginger, nutmeg, garlic, coriander, thyme, and cinnamon. Platforms were sandpaper, wood, neoprene, metal wire, tile, and a scrubber. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 4 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Table 2 Pearson’s correlations for ASST performance Table 3 Factor loadings of the ASST stages for the two- factor models with promax rotation Young SD CD CDR IDS IDR EDS EDR Young Factor1 Factor2 Aged Factor1 Factor2 SD 1 0.41 -0.48 -0.17 -0.13 0.03 0.01 SD 1.006 SD 0.708 -0.277 CD 0.41 1 -0.19 0.17 -0.26 -0.13 -0.08 CD 0.409 CD 0.283 0.550 CDR -0.48 -0.19 1 0.15 -0.05 0.23 0.08 CDR -0.475 CDR -0.602 0 IDS -0.17 0.17 0.15 1 -0.10 0.07 0.48 IDS 0.475 IDS -0.156 1.021 IDR -0.13 -0.26 -0.05 -0.10 1 0 -0.30 IDR -0.310 IDR 0.308 0.506 EDS 0.03 -0.13 0.23 0.07 0 1 0.19 EDS 0.194 EDS -0.640 0 EDR 0.01 -0.08 0.08 0.48 -0.30 0.19 1 EDR 1.006 EDR -0.343 0.551 Aged SD CD CDR IDS IDR EDS EDR Cutoff = 0.15. SD 1 -0.05 -0.45 -0.22 0.16 -0.48 -0.18 CD -0.05 1 -0.37 0.56 0.59 -0.23 -0.02 CDR -0.45 -0.37 1 -0.04 -0.06 0.34 0.22 factor. Because the IDS was the first shift in the ASST, IDS -0.22 0.56 -0.04 1 0.52 -0.13 0.52 the second factor is termed the “shifting factor.” IDR 0.16 0.59 -0.06 0.52 1 -0.37 0.18 EDS -0.48 -0.23 0.34 -0.13 -0.37 1 0.04 Three-factor model EDR -0.18 -0.02 0.22 0.52 0.18 0.04 1 The three-factor model is shown in Table 4 and Figure 4. In this model, SD and CD loaded on distinct factors in Figure 2. Both the young and the aged groups showed both young and aged mice. Because the difference significant negative correlations between SD and CDR. between SD and CD tasks was the absence or presence of The young group, but not the aged group, showed a distractors, discrimination in a distracting environment positive correlation between SD and CD. The aged can be functionally differentiated from discrimination group, but not the young group, showed a negative cor- without distractors. Therefore, the first two factors of relation between SD and EDS. this three-factor model are termed the “discrimination factor” and the “undistracted-discrimination factor.” The Two-factor model third factor was similar to the second factor in the two- The extracted two-factor model is shown in Table 3 and factor model in that both IDS and EDR had high factor Figure 3. In young mice, SD, CD, and CDR had high loadings in both young and aged mice. EDS did not have factor loadings on the first factor, which is termed the consistent loadings between the groups. We term the “discrimination factor.” The remaining stages (IDS, IDR, third factor the “shifting factor” because IDS was the first EDS, and EDR) had high factor loadings on the second shifting stage in the ASST. As with the two-factor model, factor. The aged mice did not show such a clear separa- the factor loading profiles were less clear in aged mice tion, but the IDS had the highest loading on the second than in young mice. Distractibility Previous studieshaveshown that thepresenceof dis- tractors increases error rates during cognitive tasks in aged subjects [26,27]. We estimated the aging effects on distractibility by subtracting SD from CD, to approxi- mate the increased number of trials needed to learn the discrimination when distracters are present. The aged group had a higher mean value of “CD minus SD” (mean: young = -5.56, aged = 0.28), with a trend toward a significant effect of age (p = 0.065, one-tailed t-test). Discussion The two-factor model clearly showed a separation of performance during discrimination stages (SD, CD, and CDR) from the remaining stages (IDS, IDR, EDS, and EDR) in young mice only. CDR had a negative factor loading, whereas SD had a positive factor loading in Figure 2 Pearson’s correlations of SD with other ASST both groups. The correlation analysis also confirmed variables. negative correlations between SD and CDR (Figure 2). Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 5 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Figure 3 Factor loading profiles across the ASST stages for the two-factor models with promax rotation. A1and A2:Two factorsin young mice. The cumulative variance is 0.41 and p = 0.90. B1 and B2: Two factors in aged mice. The cumulative variance is 0.51 and p = 0.38. Table 4 Factor loadings of the ASST stages for the three-factor models with promax rotation Young Factor1 Factor2 Factor3 Aged Factor1 Factor2 Factor3 SD 0.991 SD 1.066 -0.271 CD 0.926 CD -0.229 1.102 CDR -0.469 CDR -0.351 -0.376 0.251 IDS -0.221 0.539 IDS 0.405 0.675 IDR -0.361 IDR 0.165 0.474 0.310 EDS -0.219 0.169 EDS -0.511 EDR -0.476 1.012 EDR -0.270 0.811 Cutoff = 0.15. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 6 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Figure 4 Factor loading profiles across the ASST stages for the three-factor models with promax rotation. A1, A2 and A3: Three factors in young mice. The cumulative variance is 0.56 and p = 0.78. B1, B2 and B3: Three factors in aged mice. The cumulative variance is 0.69 and p = 0.55. This negative relationship is interesting because the CDR in both groups (young = -0.477; aged = -0.455), CDR involves discrimination but is also the first reversal which suggests that mice with better SD performed learning stage in the ASST. The negative correlation worse in CDR, irrespective of age. between SD and CDR indicates that mice with high SD The three-factor model further separated CD from SD, performance performed poorly with reversal learning. which suggests that the addition of irrelevant stimuli There might be a trade-off relationship between simple required a new strategy for discrimination. This separa- discrimination and reversal learning. Therefore, the “dis- tion was shown in both groups of mice. Notably, the crimination factor,” on which the SD, CD, and CDR factor loadings showed that the separation of CD from load, might also include a tendency for “perseveration.” SD was enhanced in aged mice when compared with The reciprocal SD-CDR relationship observed in our young mice (Table 4). This clearer separation may be mouse study is consistent with a similar study using due to increased distractibility in aged mice. The degree ASST with young and aged rats [14]. Those researchers of distractibility was estimated by calculating the differ- used 4- to 5-month-old and 27- to 28-month-old male ence between SD and CD. Higher values of “CD minus Long-Evans rats (n = 26) in a similar ASST study, but SD” appeared to be indicative of higher distractibility. no CDR or EDR stage was employed. In the rat study, The aged group had a higher mean value of “CD minus IDR was the first reversal task, while CDR was the first SD,” with a trend toward significance between the two in the mouse study that we analyzed. There was a nega- groups. This finding indicates that, unlike young mice, tive correlation between SD and IDR in the rat study, a aged mice were affected by the addition of an irrelevant stimulus. These data are consistent with human studies, finding that is similar to our study. Their results differ from ours in that the correlation was negative only in indicating that aged mice have higher distractibility or the young rats (young = -0.728; aged = -0.155). In con- lower robustness of executive functioning when com- trast, we found a negative correlation between SD and pared with young mice [28-30]. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 7 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Because IDS, EDS, and reversal learning may be subjects who were performing at comparable levels. mediated by distinct brain regions [12,19-21,23], we This result suggests that alternate networks are being expected that these variables would load on distinct fac- recruited in older subjects [36-39] and that the elderly tors. IDS, EDS, and CDR did indeed load on distinct fac- use a different strategy when performing a cognitive tors in both groups. The separation of these factors is task [39-42]. Potential neurochemical differences that meaningful because CDR was the first reversal learning might account for alterations in aging performance in ASST are numerous. In young animals, dopamine sig- stage in the ASST. However, the factor loadings of EDS naling is critical for EDS performance. For example, were fairly low for all factors extracted. It is unclear why amphetamine-sensitized rats or rats that acquired CDR, IDR, and EDR did not load on a single factor; there- fore, the extracted three-factor model had unexpected fac- amphetamine self-administration exhibited impaired tor loadings. There are several possible reasons for these EDS but not IDS performance [43-45]. Treatment with results. First, the mice may not have formed an attentional tolcapone, a catechol-O-methyltransferase inhibitor that set; however, this explanation is unlikely because both the increases dopamine and norepinephrine levels in the young and the aged mice showed significant increases in frontal cortex, improved EDS performance in rats [46]. the number of trials needed to reach the criterion for the This latter study suggests that catecholamine transmis- EDS stage when compared with the IDS stage, regardless sion in the rat mPFC is critically involved in set-shifting of the initial perceptual dimension [15]. This finding sug- functions. A recent study in rats has suggested that gests the successful formation of an attentional set in this remodeling of mesocortical dopaminergic fibers is group. Second, despite the hypothesized functional locali- involved in age-associated cognitive decline [47]. Other zation in the brain, these regions likely interact to perform studies have also demonstrated a specific involvement of the ASST stages. It has been suggested that cognitive pro- noradrenergic transmission in ASST performance cessing is mediated by a network [31-34]. The network [48-51], while serotonin modulation may affect reversal involved in ASST performance would include the mPFC, learning components in this task but not EDS perfor- OFC, and other regions, and interaction among the mance [52]. Interestingly, administration of naltrexone involved regions in the network might have blurred the (2 mg/kg, i.p.), an opioid antagonist, has been shown to separation of factors. The capability of factor analysis to reverse aging-related deficits in EDS [53], suggesting extract a functional structure in such a distributed net- that altered endorphin signaling plays a role in aging- work is an interesting issue that needs to be addressed. induced decline in set shifting. Research on the complex underpinnings of age-related deficits in attentional set Third, a small sample size (n = 16 for young mice and n = shifting is in its infancy; however, these data suggest 18 for aged mice) was used in this analysis. The same ana- lysis with a larger sample size should be performed to that there are a number of potential systems that could determine whether the sample size was a limiting factor. It influence aging effects on this task. would be interesting to see whether a factor analysis with a larger sample size would result in a better model with Conclusion more consistent factor loadings. Finally, there is evidence The exploratory factor analysis applied to the mouse in rats and mice that reversal learning improves over ASST data extracted two- and three-factor models. The repeated exposure, with fewer trials required at EDR vs. two-factor model clearly separated discrimination perfor- CDR [15,23,35]. This ‘learning to reverse’ might incorpo- mance (SD and CD) from the remaining ASST stages in rate other neuroanatomical structures that are not specific young mice. This tendency was obscured in the aged to reversal learning, hence the lack of commonality in mice, suggesting that they utilize a less selective network reversal loading. in the brain for cognitive functions. The three-factor In our study, the overall deterioration in the cognitive model further separated CD from SD. The result that the function of aged mice was not marked. However, while factors associated with SD and CD were separated in the the number of trials needed to reach the criterion was three-factor model may suggest that the introduction of not different between groups for EDS, aged mice exhib- an irrelevant or distracting dimension results in the use ited significantly longer mean correct latencies during of a new/orthogonal strategy for better discrimination. this task (p < 0.05) [15]. This result indicates that the This model, however, did not show consistent separation aged mice have reduced processing speed only during of the factors associated with IDS, EDS, and reversal the EDS stage. Thus, when compared with young mice, learning in either young or aged mice, contrary to the aged mice might require more time to perform the EDS hypothesis that these functions are mediated by distinct (i.e. speed-accuracy trade off) [15]. Magnetic resonance brain regions. Whether the inconsistent separation of fac- imaging studies in humans performing an executive tors was due to possible interaction among the brain function task reported a broader activation of brain regions or to the small sample size needs to be clarified regions in older subjects when compared with younger by a future analysis using a larger sample size. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 8 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 15. Young JW, Powell SB, Geyer MA, Jeste DV, Risbrough VB: The mouse Acknowledgements attentional set-shifting task: A method for assaying successful cognitive The original ASST mouse study was funded by Stein Institute of Research on aging? Cognitive, Affective, & Behavioral Neuroscience 2010, 10:243-251, Aging (VBR and JWY) and MH052885 (MAG). This analytical study was PMID: 20498348. supported by the Human Informatics Research Center at Sophia University 16. Jazbec S, Pantelis C, Robbins T, Weickert T, Weinberger DR, Goldberg TE: (ST). Intra-dimensional/extra-dimensional set-shifting performance in schizophrenia: Impact of distractors. Schizophrenia Research 2007, Author details 89:339-349, PMID: 17055703. Department of Information & Communication Sciences, Sophia University, 17. Dias R, Robbins TW, Roberts AC: Dissociable forms of inhibitory control Tokyo, Japan. Department of Psychiatry, University of California, San Diego, within prefrontal cortex with an analog of the Wisconsin Card Sort Test: USA. restriction to novel situations and independence from “on-line” processing. J Neurosci 1997, 17(23):9285-9297, PMID: 9364074. Authors’ contributions 18. Rogers RD, Andrews TC, Grasby PM, Brooks DJ, Robbins TW: Contrasting ST carried out the whole analysis. JWY, JEG, MAG, and VBR carried out the cortical and subcortical activations produced by attentional-set shifting original animal experiment. All authors read and approved the final and reversal learning in humans. J Cogn Neurosci 2000, 12(1):142-162, manuscript. PMID: 10769312. 19. Ng C-W, Noblejas MI, Rodefer JS, Smith CB, Poremba A: Double Competing interests Dissociation of Attentional Resources: Prefrontal Versus Cingulate The authors declare that they have no competing interests. Cortices. J Neurosci 2007, 27:12123-12131, PMID:17989278. 20. Bissonette GB, Martins GJ, Franz TM, Harper ES, Schoenbaum G, Powell EM: Received: 1 December 2010 Accepted: 12 August 2011 Double Dissociation of the Effects of Medial and Orbital Prefrontal Published: 12 August 2011 Cortical Lesions on Attentional and Affective Shifts in Mice. J Neurosci 2008, 28:11124-11130, PMID: 18971455. References 21. Ragozzino ME: The contribution of the medial prefrontal cortex, 1. Glisky EL: Changes in cognitive function in human aging. In Brain aging: orbitofrontal cortex, and dorsomedial striatum to behavioral flexibility. models, methods, and mechanisms. Edited by: Riddle DR. Boca Raton: CRC Ann N Y Acad Sci 2007, 1121:355-375, PMID: 17698989. Press; 2007:. 22. Dalley JW, Cardinal RN, Robbins TW: Prefrontal executive and cognitive 2. Verhaeghen P, Cerella J: Everything we know about aging and response functions in rodents: neural and neurochemical substrates. Neurosci time. In Handbook of cognitive aging: interdisciplinary perspectives. Edited by: Biobehav Rev 2004, 28(7):771-784, PMID: 15555683. Hofer SM, Alwin DF. Los Angeles: Sage Publications; 2008:134-150. 23. McAlonan K, Brown VJ: Orbital prefrontal cortex mediates reversal 3. Old SR, Naveh-Benjamin M: Age-related changes in memory. In Handbook learning and not attentional set shifting in the rat. Behav Brain Res 2003, of cognitive aging: interdisciplinary perspectives. Edited by: Hofer SM, Alwin 146:97-103, PMID: 14643463. DF. Los Angeles: Sage Publications; 2008:151-167. 24. Tanaka S, Risbrough V, Young J, Gresack J, Geyer M: Factor analysis of 4. Raz N, Rodrigue KM: Differential aging of the brain: Patterns, cognitive attentional set-shifting task performance of aged and young mice. correlates and modifiers. Neuroscience & Biobehavioral Reviews 2006, Society for Neuroscience Abstracts 2009, 580.023. 30:730-748, PMID: 16919333. 25. Brown TA: Confirmatory factor analysis for applied research New York: The 5. Grady CL: Cognitive Neuroscience of Aging. Ann N Y Acad Sci 2008, Guilford Press; 2006. 1124:127-144, PMID: 18400928. 26. Chao L, Knight R: Prefrontal deficits in attention and inhibitory control 6. Luszcz M, Lane A: Executive function in cognitive, neuropsychological, with aging. Cereb Cortex 1997, 7:63-69, PMID: 9023433. and clinical aging. In Handbook of cognitive aging: interdisciplinary 27. Lustig C, Hasher L, Tonev ST: Distraction as a determinant of processing perspectives. Edited by: Hofer SM, Alwin DF. Los Angeles: Sage Publications; speed. Psychonomic Bulletin & Review 2006, 13:619-625, PMID: 17201361. 2008:193-206. 28. Clapp WC, Gazzaley A: Distinct mechanisms for the impact of distraction 7. Moss MB, Moore TL, Schettler SP, Killiany R, Rosene D: Successful vs. and interruption on working memory in aging. Neurobiology of Aging Unsuccessful Aging in the Rhesus Monkey. In Brain Aging: Models, 2010, PMID: 20144492. Methods, and Mechanisms. Edited by: Riddle DR. Boca Raton (FL): CRC Press; 29. Healey MK, Campbell KL, Hasher L: Cognitive aging and increased 2007:. distractibility: Costs and potential benefits. Progress in Brain Research 2008, 8. Eling P, Derckx K, Maes R: On the historical and conceptual background 169:353-363, PMID: 18394486. of the Wisconsin Card Sorting Test. Brain and Cognition 2008, 67:247-253, 30. Machado L, Devine A, Wyatt N: Distractibility with advancing age and PMID: 18328609. Parkinson’s disease. Neuropsychologia 2009, 47:1756-1764, PMID: 9. Owen AM, Roberts AC, Polkey CE, Sahakian BJ, Robbins TW: Extra- dimensional versus intra-dimensional set shifting performance following 31. Bullmore E, Sporns O: Complex brain networks: graph theoretical analysis frontal lobe excisions, temporal lobe excisions or amygdalo- of structural and functional systems. Nat Rev Neurosci 2009, 10:186-198, hippocampectomy in man. Neuropsychologia 1991, 29:993-1006, PMID: PMID: 19190637. 32. Colom R, Karama S, Jung RE, Haier RJ: Human intelligence and brain 10. Downes JJ, Roberts AC, Sahakian BJ, Evenden JL, Morris RG, Robbins TW: networks. Dialogues Clin Neurosci 2010, 12:489-501, PMID:21319494. Impaired extra-dimensional shift performance in medicated and 33. Green JJ, Doesburg SM, Ward LM, McDonald JJ: Electrical Neuroimaging of unmedicated Parkinson’s disease: evidence for a specific attentional Voluntary Audiospatial Attention: Evidence for a Supramodal Attention dysfunction. Neuropsychologia 1989, 27:1329-1343, PMID: 2615934. Control Network. J Neurosci 2011, 31:3560-3564, PMID: 21389212. 11. Robbins TW, James M, Owen AM, Sahakian BJ, McInnes L, Rabbitt P: 34. Naghavi HR, Nyberg L: Common fronto-parietal activity in attention, Cambridge Neuropsychological Test Automated Battery (CANTAB): a memory, and consciousness: Shared demands on integration? factor analytic study of a large sample of normal elderly volunteers. Consciousness and Cognition 2005, 14:390-425, PMID:15950889. Dementia 1994, 5:266-281, PMID: 7951684. 35. Boulougouris V, Dalley JW, Robbins TW: Effects of orbitofrontal, infralimbic 12. Birrell JM, Brown VJ: Medial frontal cortex mediates perceptual and prelimbic cortical lesions on serial spatial reversal learning in the attentional set shifting in the rat. J Neurosci 2000, 20:4320-4324, PMID: rat. Behav Brain Res 2007, 179:219-228, PMID:17337305. 36. Milham MP, Erickson KI, Banich MT, Kramer AF, Webb A, Wszalek T, 13. Garner JP, Thogerson CM, Wurbel H, Murray JD, Mench JA: Animal Cohen NJ: Attentional Control in the Aging Brain: Insights from an fMRI neuropsychology: Validation of the Intra-Dimensional Extra-Dimensional Study of the Stroop Task. Brain and Cognition 2002, 49:277-296, PMID: set shifting task for mice. Behav Brain Res 2006, 173:53-61, PMID: 37. Grady CL, Protzner AB, Kovacevic N, Strother SC, Afshin-Pour B, 14. Barense MD, Fox MT, Baxter MG: Aged Rats Are Impaired on an Wojtowicz M, Anderson JAE, Churchill N, McIntosh AR: A Multivariate Attentional Set-Shifting Task Sensitive to Medial Frontal Cortex Damage Analysis of Age-Related Differences in Default Mode and Task-Positive in Young Rats. Learning & Memory 2002, 9:191-201, PMID: 12177232. Tanaka et al. Behavioral and Brain Functions 2011, 7:33 Page 9 of 9 http://www.behavioralandbrainfunctions.com/content/7/1/33 Networks across Multiple Cognitive Domains. Cereb Cortex 2009, 20:1432-1447, PMID: 19789183. 38. Prakash RS, Erickson KI, Colcombe SJ, Kim JS, Voss MW, Kramer AF: Age- related differences in the involvement of the prefrontal cortex in attentional control. Brain and Cognition 2009, 71:328-335, PMID: 19699019. 39. Venkatraman VK, Aizenstein H, Guralnik J, Newman AB, Glynn NW, Taylor C, Studenski S, Launer L, Pahor M, Williamson J, Rosano C: Executive control function, brain activation and white matter hyperintensities in older adults. NeuroImage 2010, 49:3436-3442, PMID: 19922803. 40. Grady CL, McIntosh AR, Beig S, Keightley ML, Burian H, Black SE: Evidence from Functional Neuroimaging of a Compensatory Prefrontal Network in Alzheimer’s Disease. J Neurosci 2003, 23:986-993, PMID: 12574428. 41. Reuter-Lorenz PA: New visions of the aging mind and brain. Trends in Cognitive Sciences 2002, 6:394-400, PMID: 12200182. 42. Reuter-Lorenz PA, Lustig C: Brain aging: reorganizing discoveries about the aging mind. Current Opinion in Neurobiology 2005, 15:245-251, PMID: 43. Featherstone RE, Rizos Z, Kapur S, Fletcher PJ: A sensitizing regimen of amphetamine that disrupts attentional set-shifting does not disrupt working or long-term memory. Behav Brain Res 2008, 189:170-179, PMID: 44. Fletcher P, Tenn C, Rizos Z, Lovic V, Kapur S: Sensitization to amphetamine, but not PCP, impairs attentional set shifting: reversal by a D1 receptor agonist injected into the medial prefrontal cortex. Psychopharmacology 2005, 183:190-200, PMID: 16220338. 45. Parsegian A, Glen WB, Lavin A, See RE: Methamphetamine Self- Administration Produces Attentional Set-Shifting Deficits and Alters Prefrontal Cortical Neurophysiology in Rats. Biological Psychiatry 2011, 69:253-259, PMID: 21051037. 46. Tunbridge EM, Bannerman DM, Sharp T, Harrison PJ: Catechol-o- methyltransferase inhibition improves set-shifting performance and elevates stimulated dopamine release in the rat prefrontal cortex. J Neurosci 2004, 24(23):5331-5335, PMID: 15190105. 47. Allard S, Gosein V, Cuello AC, Ribeiro-da-Silva A: Changes with aging in the dopaminergic and noradrenergic innervation of rat neocortex. Neurobiology of Aging 2010, PMID: 20096955. 48. Lapiz MDS, Morilak DA: Noradrenergic modulation of cognitive function in rat medial prefrontal cortex as measured by attentional set shifting capability. Neuroscience 2006, 137:1039-1049, PMID: 16298081. 49. Lapiz MDS, Bondi CO, Morilak DA: Chronic Treatment with Desipramine Improves Cognitive Performance of Rats in an Attentional Set-Shifting Test. Neuropsychopharmacology 2007, 32:1000-1010, PMID: 17077810. 50. McGaughy J, Ross RS, Eichenbaum H: Noradrenergic, but not cholinergic, deafferentation of prefrontal cortex impairs attentional set-shifting. Neuroscience 2008, 153:63-71, PMID: 18355972. 51. Newman L, Darling J, McGaughy J: Atomoxetine reverses attentional deficits produced by noradrenergic deafferentation of medial prefrontal cortex. Psychopharmacology 2008, 200:39-50, PMID: 18568443. 52. Clarke HF, Walker SC, Crofts HS, Dalley JW, Robbins TW, Roberts AC: Prefrontal serotonin depletion affects reversal learning but not attentional set shifting. J Neurosci 2005, 25(2):532-538, PMID: 15647499. 53. Rodefer JS, Nguyen TN: Naltrexone reverses age-induced cognitive deficits in rats. Neurobiology of Aging 2008, 29:309-313, PMID:17098330. doi:10.1186/1744-9081-7-33 Cite this article as: Tanaka et al.: Factor analysis of attentional set- shifting performance in young and aged mice. Behavioral and Brain Functions 2011 7:33. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit

Journal

Behavioral and Brain Functions – Springer Journals

Published: Aug 12, 2011

There are no references for this article.