Research reportMiddle-aged human apoE4 targeted-replacement mice show retention deficits on a wide range of spatial memory tasks
Introduction
There is widespread agreement that the function of hepatocyte secreted apolipoprotein (apo) E is to maintain lipid homeostasis, however the function of astrocyte secreted apoE remains equivocal. Many theories abound to explain the function of brain apoE, including lipid homeostasis [1], [2], neural regeneration and remodeling [3], [4], immunomodulation and mediating estrogen effects [5], [6], [7], [8]. In humans, three major isoforms of apoE (apoE2, apoE3 and apoE4) result from three allelic variants (ɛ2, ɛ3 and ɛ4) of a single gene on chromosome 19. This APOE genetic polymorphism influences the prevalence of cardiovascular and Alzheimer's disease (AD) ([9], [10], [11], also for reviews [12], [13]). Indeed, apoE4 has been identified as the most significant genetic risk factor for AD and also appears associated with increased cognitive decline in healthy elderly individuals [14], [15], [16]. Memory decline associated with the APOE4 genotype has also been reported in middle-aged populations [17], [18], [19], and seems to be particularly pronounced in women at post-menopausal ages [20], [21], [22].
Our understanding of how apoE influences behavior was first investigated in transgenic apoE animal models which expressed human apoE under the control of either the neuron-specific-enolase (NSE) or glial-fibrillary-acidic-protein (GFAP) promoters. NSE-apoE4 mice showed an age- and sex-dependent spatial learning impairment compared to NSE-apoE3 mice [6], [7], [23]. Compared to their GFAP-apoE3 counterparts, spatial working memory deficits appeared in GFAP-apoE4 males only at an age of 11–14 months, whereas spatial memory retention was already affected in 6-month-old GFAP-apoE4 females [24], [25]. Thus, an isoform-dependent effect of apoE on cognition could be observed regardless of whether the human apoE gene was expressed under the control of the neuronal or the glia-specific promoter. However, these mouse lines show varying levels of transgene expression and/or no physiological distribution of human apoE due to non-physiological promoters. Human apoE targeted-replacement (TR) mice were created using a gene targeting strategy, which allows the physiological expression of human apoE isoforms without disturbing any of the known regulatory sequences [26]. Our initial work assessed the behavioral phenotype of young (4–5 months) apoE TR mice using a battery of behavioral tests to measure cognitive ability [27]. Briefly, only female apoE4-TR mice showed impaired spatial recognition memory compared to apoE3-TR mice. Additionally, apoE4-TR mice showed poor performance during the probe trial in the water-maze reference memory task. These results support the hypothesis that apoE4 has a deleterious effect on spatial cognitive processes.
In this study, we asked whether the early memory deficits in the apoE4-TR mice worsen with age as observed in humans. Thus, we tested 15-month-old apoE-TR mice in the same behavioral paradigms used in the young cohort. Human apoE4 carriers show deficits in recognition memory, memory for location, delayed recall, working memory, new learning and episodic memory [17], [18], [20], [28], [29]. Thus, apoE-TR mice were tested in learning and memory tasks testing similar forms of memory. These included; (i) a spatial object recognition task, (ii) acquisition and delayed memory retention evaluations in a spatial reference memory task, (iii) a delayed matching to place (DMP) task requiring working memory (trial 1 to trial 2), episodic-like memory and a new learning within each daily session [30] and (iv) two avoidance tasks testing long-term memory retention. Since gender-dependent effects of apoE on cognitive abilities have been described in both humans and transgenic apoE mice, we examined both male and female mice [6], [21]. C57BL/6J and apoE deficient (−/−) mice were included in the study to compare expression of mouse apoE and lack of apoE to human apoE isoform specific expression. Since our study of the young cohorts showed smaller differences between genotypes, as seen in humans without disease [20], [21], [31], we applied statistical analyses limited to within-genotype group effects and apoE3-TR/apoE4-TR restricted comparisons in addition to those including the four genotypes.
Section snippets
Origin of the mice
The apoE TR mice were created by replacing the mouse apoE gene with either APOE*3 or APOE*4 genomic fragments of human alleles using mouse embryonic stem cells as described previously [32], [33]. The parental mice of either apoE3-TR or apoE4-TR lines used in this study were backcrossed for eight generations to C57BL/6J. ApoE−/− mice were included as a control to study the absence of apoE, and the C57BL/6J mice were included to compare the expression of mouse apoE to the expression human apoE.
Bodyweight
On day 1, when mice were 15 months old, their weight differed among genotypes (F(3,53) = 8.68, P < 0.0001, see Table 1). The apoE4-TR mice were heavier than those of the other lines (P ≤ 0.004 vs. each other line). Females were much lighter than males over the whole population (F(1,53) = 39.06, P < 0.0001). The weight of apoE4-TR mice was confirmed to be the highest even within each sex (females: F(3,27) = 6.31, P = 0.002; males: F(3,26) = 4.62, P = 0.010). The whole population showed a slight loss of weight
Discussion
Several studies show a direct correlation between the ɛ4 allele and cognitive decline or lower episodic-memory scores in non-demented elderly subjects [14], [15], [16], [17], [18], [19], [29]. We set out to determine whether the apoE4-TR mice modeled the cognitive decline observed in human E4 carriers. The same behavioral tests used in the young cohorts were applied to the 15-month-old mice [27]. Our main findings revealed that the apoE4 mice are impaired in (i) the probe test of the water-maze
Acknowledgements
We are grateful to Dr. Steven M. Paul for his contribution to the present work. This research was supported by Université Louis Pasteur, Centre National de la Recherche Scientifique and grants from the ULP-CNRS-Eli Lilly foundation awarded to C.M. and Association Alsace Alzheimer 68 awarded to A.B.
References (68)
Apolipoprotein E, cholesterol transport and synthesis in sporadic Alzheimer's disease
Neurobiol Aging
(2005)- et al.
Effect of apolipoprotein E deficiency on reactive sprouting in the dentate gyrus of the hippocampus following entorhinal cortex lesion: role of the astroglial response
Exp Neurol
(2005) - et al.
A deficit in astroglial organization causes the impaired reactive sprouting in human apolipoprotein E4 targeted replacement mice
Neurobiol Dis
(2006) - et al.
Apolipoprotein E suppresses glial cell secretion of TNF alpha
J Neuroimmunol
(1997) Apolipoproteins and aging: emerging mechanisms
Ageing Res Rev
(2002)- et al.
Memory performance in healthy elderly without Alzheimer's disease: effects of time and apolipoprotein-E
Neurobiol Aging
(2001) - et al.
Apolipoprotein E variants and cognition in healthy individuals: a critical opinion
Brain Res Rev
(2006) - et al.
Apolipoprotein E epsilon 4 affects new learning in cognitively normal individuals at risk for Alzheimer's disease
Neurobiol Aging
(2003) - et al.
Reduced hippocampal volume in non-demented carriers of the apolipoprotein E epsilon4: relation to chronological age and recognition memory
Neurosci Lett
(2006) - et al.
Gender differences in dementia risk factors
Gend Med
(2007)