International Journal of Molecular Sciences Article Survival Bias, Non-Lineal Behavioral and Cortico-Limbic Neuropathological Signatures in 3xTg-AD Mice for Alzheimer’s Disease from Premorbid to Advanced Stages and Compared to Normal Aging Aida Muntsant 1,2 , Maria del Mar Castillo-Ruiz 2 and Lydia Giménez-Llort 1,2, * 1 Department of Psychiatry and Forensic Medicine, School of Medicine, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain; aida.muntsant@uab.cat 2 Institut de Neurociències, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain; mariadelmar.castillo@uab.cat * Correspondence: lidia.gimenez@uab.cat; Tel.: +34-93-581-2378 Abstract: Pre-clinical research in aging is hampered by the scarcity of studies modeling its hetero- geneity and complexity forged by pathophysiological conditions throughout the life cycle and under the sex perspective. In the case of Alzheimer’s disease, the leading cause of dementia in older adults, we recently described in female wildtype and APP23 mice a survival bias and non-linear chronology of behavioral signatures from middle age to long life. Here, we present a comprehen- sive and multidimensional (physical, cognitive, and neuropsychiatric-like symptoms) screening and underlying neuropathological signatures in male and female 3xTg-AD mice at 2, 4, 6, 12, and 16 months of age and compared to their non-transgenic counterparts with gold-standard C57BL/6J background. Most variables studied detected age-related differences, whereas the genotype factor was specific to horizontal and vertical activities, thigmotaxis, coping with stress strategies, working memory, and frailty index. A sex effect was predominantly observed in classical emotional variables Citation: Muntsant, A.; Castillo-Ruiz, and physical status. Sixteen-month-old mice exhibited non-linear age- and genotype-dependent M.d.M.; Giménez-Llort, L. Survival behavioral signatures, with higher heterogeneity in females, and worsened in naturalistically isolated Bias, Non-Lineal Behavioral and males, suggesting distinct compensatory mechanisms and survival bias. The underlying temporal Cortico-Limbic Neuropathological and spatial progression of Aβ and tau pathologies pointed to a relevant cortico-limbic substrate Signatures in 3xTg-AD Mice for roadmap: premorbid intracellular Aβ immunoreactivity and pSer202/pThr205 tau phosphorylation Alzheimer’s Disease from Premorbid to Advanced Stages and Compared in the amygdala and ventral hippocampus, and the entorhinal cortex and ventral hippocampus as the to Normal Aging. Int. J. Mol. Sci. areas most affected by Aβ plaques. Therefore, depicting phenotypic signatures and neuropathological 2023, 24, 13796. https://doi.org/ correlates can be critical to unveiling preventive/therapeutic research and intervention windows and 10.3390/ijms241813796 studying adaptative behaviors and maladaptive responses relevant to psychopathology. Academic Editors: Ryszard Pluta and Keywords: aging; Alzheimer’s disease; heterogeneity; cognitive deficits; neuropsychiatric-like Changjong Moon symptoms; amygdala; ventral hippocampus; animal model; social isolation Received: 26 June 2023 Revised: 11 August 2023 Accepted: 29 August 2023 Published: 7 September 2023 1. Introduction Neurological disorders are recognized as the principal cause of disability-adjusted life years [1]. Alzheimer’s disease (AD) is a heterogenous and multi-factorial neurode- Copyright: © 2023 by the authors. generative disorder caused by the interaction of demographic–environmental and genetic Licensee MDPI, Basel, Switzerland. factors [2]. AD represents the leading causes of dementia, with age-related changes playing This article is an open access article a crucial role [3]. Several studies highlight the heterogenous and complex clinical patterns distributed under the terms and and temporal progression, especially in the late stages of the disease [4,5]. Nowadays, conditions of the Creative Commons there is a high failure rate in developing new disease-modifying treatments, and therapies Attribution (CC BY) license (https:// seem to have limited benefits regarding AD pathogenesis and progression. More precise creativecommons.org/licenses/by/ diagnosis criteria, recognition of AD heterogeneity, and personalized intervention strategies 4.0/). should improve outcomes in basic research and clinical trials [6]. Int. J. Mol. Sci. 2023, 24, 13796. https://doi.org/10.3390/ijms241813796 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2023, 24, 13796 2 of 26 At the clinical level, besides cognitive impairments and memory loss, additional clinical manifestations related to neuropsychiatric symptoms (NPS), also referred to as be- havioral and psychological symptoms of dementia (BPSD), are observed in approximately 90% of patients. These include agitation, aggression (verbal or physical), irritability, anxiety, depression or other disruptive behaviors, sleep disorders, and hallucinations [7,8]. Age, gender, and comorbidities may influence the severity of neuropsychiatric symptoms [9]. Their management is challenging and usually requires pharmacological treatments (i.e., an- tipsychotics) and non-pharmacological intervention [10]. Moreover, NPS are the symptoms most related to disease burden, quality of life, and caregiver burnout [11]. At the neuropathological level, AD is mainly defined by extracellular accumula- tion of amyloid-β (Aβ) plaques and hyperphosphorylated tau proteins forming intra- cellular neurofibrillary tangles (NFTs), accompanied by reactive gliosis and synaptic dysfunction [12–14]. Moreover, although the connections are not yet fully understood, there is growing experimental, clinical, and epidemiological evidence that brain and periphery crosstalk in- teraction may be related to AD development and progression. Underlying systemic disease processes are reflected in AD patients: systemic immunity disorders, cardiovascular disease, hepatic and renal dysfunction, metabolic disorders, blood abnormalities, respiratory and sleep disorders, microbiota disorders, and systemic inflammation [15,16]. AD pathology is characterized by cortical atrophy, particularly in the multimodal association cortices and the structures of the limbic lobe, and the progressive degeneration of neurons in various areas (such as the hippocampus, amygdala, entorhinal, and neocortex) [14,17,18]. Re- garding neuropsychiatric symptoms, the pathological mechanisms are not already known. However, the emotional and anxiety behaviors presented in AD have been associated with metabolic and volumetric alterations in the amygdala [19,20]. On the other hand, several articles have described the heterogeneous function of the rodent hippocampus, with the dorsal hippocampus playing an important role in cognitive function and connected to the brain regions connected to spatial memory; the ventral hippocampus has been related to anxiety- and stress-involved areas, i.e., connections between the amygdala and the ventral hippocampus and prefrontal cortex have been documented with anxiety-like behaviors and fear response in animal models [21–24]. At the translational level, the shorter lifespan of most animal models affords a fleet- footed scenario for long-term monitoring. For this reason, animal model characterization can help to better understand disease mechanisms from morphological to behavioral levels and to evaluate the safety and efficacy of new therapeutic approaches. Experimental geron- tologists highlight the relevance of using aged animals to mimic humans’ complex and multi-factorial aging processes [25,26]. Moreover, the heterogeneity presented in AD pa- tients has also been demonstrated in animal models for the disease, especially at advanced stages [27–29]. Studying the disease and age interaction effects in phenotypic characteristics using AD models can also provide information on normal and pathological aging processes. Among the animal models of AD, we have proposed asymptomatic, prodromal, onset, advanced, and long-term survivors of widely used 3xTg-AD mice. Based on the familial AD mutations PS1/M146V and APPSwe, and harboring tauP301L human transgene, this model progressively develops temporal- and regional-specific neuropathological patterns observed in the human brain of AD patients [30–32]. Cognitive deficits [33,34] and a broad spectrum of NPS-like symptoms have also been described [35,36]. In the initial reports of the 3xTg-AD model, mice first develop intraneuronal Ab at 3–4 months of age, followed by plaque formation at 6 months of age in the cortex and hippocampus, with NFT becom- ing apparent at 12 months of age [30,31]. However, in recent years, several studies have indicated a drift in the phenotype of mice, with males being particularly affected. This fact suggests that the widespread use of these mice has resulted in a generation of sublines with different onset and progression of neuropathology, causing controversy and confusion. Moreover, 3xTg-AD mice exhibit some variability even between littermates [32,37]. Int. J. Mol. Sci. 2023, 24, 13796 3 of 26 Here, we described the age, AD genotype, and sex sensitivity of a behavioral test battery in 3xTg-AD mice of male and female mice at asymptomatic (2 months), prodromal (4 months), onset (6 months), advanced (12 months), and very old (16 months) ages and compared them to age-matched non-transgenic mice with normal aging. Moreover, we assessed the spatial and temporal progression of Aβ pathology and tau hyperphosphoryla- tion and studied the sex-dependent differences. 2. Results 2.1. Strong Contribution of Age but Not Genotype and Sex Effects to Target Variables for AD-Related Phenotype Table 1 summarizes the effects of the main factors: genotype (G), sex (S), and age (A). As shown in most behavioral variables, behavioral performances strongly depended on age. Genotypes differed in their horizontal and vertical activities, thigmotaxis, coping with stress strategies, working memory, and frailty index. Sex effect was consistently observed in urination, a classical emotional variable, and sexual dimorphism in body weight. Also, sex differences were found in the horizontal and vertical activities in the corner and the open-field tests, on risk assessment as in the latency to cross the intersection in the T-maze, and coping strategies as in the distance covered in the CUE, the first paradigm in the MWM, and the frailty index. Interaction effects were also observed, with special mention given to the marble test and mostly involving age × sex interaction effects. Table 1. Genotype, age, and sex factors and interaction effects on behavioral tests and variables from 2 (asymptomatic) to 16 months of age (advanced stages of disease). Statistics: Multivariate general lineal model analysis. G, genotype; A, age; S, sex. * p < 0.05, ** p < 0.01, *** p < 0.001. Behavioral Tests and Variables G A S G×A G×S A×S G×A×S Corner test (CT) Total visited corners (n) * ** ** - - - - Total numbers of rearings (n) - - - - - - * Latency of rearing (s) - - - - - - - Open-field test (OF) Freezing—Latency of first movement (s) - * - - - * - Latency to exit the center (s) - - - - - - - Latency to entering the peripheral ring (s) - *** - - - ** - Latency of rearing (s) *** *** - - - ** - Latency of self-grooming (s) - - - - - * - Total horizontal activity (n crossings) * *** - - - - - Total horizontal activity in the center (n) - * * - - - - Total horizontal activity in the periphery (n) * *** - - - - - Total vertical activity (n rearings) * *** * - - - - Gait analysis—Total number of pauses (n) - *** - - - - - Gait analysis—Mean number of crossings (n) *** * - - - - - Defecation (n) * - - - - - - Urination (%) - - *** - - - - Context and object recognition tests OF2—Freezing—Latency of first movement (s) - *** - - - * - OF2—Latency to exit the center (s) *** *** - - - * - OF2—Latency to entering the peripheral ring(s) *** *** - - - - - OF2—Latency of rearing (s) * ** - - - * - OF2—Latency of self-grooming (s) - - - - - * - OF2—Total horizontal activity (n crossings) * * - - - - - OF2—Total horizontal activity in the center (n) * - - - - - - OF2— Total horizontal in the periphery (n) - * - - - - - OF2—Total vertical activity (n rearings) - * - - - * - OF2—Urination (%) - - *** - - - - Int. J. Mol. Sci. 2023, 24, 13796 4 of 26 Table 1. Cont. Behavioral Tests and Variables G A S G×A G×S A×S G×A×S OR Sample trial—Time exploring object (s) - * - - - - - OR-Sample trial—Time to reach the criteria (s) - - - - - - * OR—Test trial—Object latency (s) * - - - - - ** Marble test (MB) Intact (n) - * - - *** - - Buried (n) - *** - - * - - Spontaneous alternation in the T-Maze test Latency of first movement (s) - *** - * - - - Latency to cross the intersection (s) - * * - - - - Total time to complete the test (s) - - - - - - - Total number of errors (n) * ** - - - - - T-Maze test Latency of first movement (s) * - - - - - - Test trial—Latency to cross the intersection (s) ** - - - - - - Test trial—Time to complete the test (s) ** - - - - - - Morris water maze test (MWM) Escape latency—CUE - - - - - - - Escape latency—PT1 - - - - - - - Escape latency—PT2 - - - - - - - Escape latency—PT3 ** ** - - - - - Escape latency—PT4 * *** - - - - - Distance—CUE - - * - - - - Distance—PT1 - - - - - - - Distance—PT2 - - - - - - - Distance—PT3 - ** - - - - - Distance—PT4 - *** - - - - - Swimming speed—CUE *** *** - ** - - * Swimming speed—PT1 *** ** - - - - - Swimming speed—PT2 ** ** - - - - - Swimming speed—PT3 - *** - - - - - Swimming speed—PT4 * ** - - - - * Probe trial—Opposite quadrant distance (m) - *** - - * - - Probe trial—Target quadrant distance (m) * - - - - - * Probe trial—Latency to platform (s) * - - - - - - Physical status Body weight (g) - *** *** - - * - Frailty index *** *** * - - - - In the corner test, the horizontal locomotor activity (Figure 1A), measured by the number of visited corners, was lower in 3xTg-AD mice than in NTg animals (G*, p = 0.020). However, an age effect pointed to 4- and 12-month-old animals being more active compared to the group of 16-month-old animals (A***, p = 0.001). These differences were mainly observed in males. Sex differences were also observed, as females presented more visited corners (S**, p = 0.003). In the case of vertical activity, no statistical differences were observed in either the total number of rearings (Figure 1B) or the latency to perform the first rearing (Figure 1C). The ethogram describing the temporal sequence of behavioral events of animals in the open-field test (Figure 2A) showed that the actions exhibited had a strong effect on age, together with genotype or age x sex interaction effects. That is, the latency of the first movement (LatM), to enter the peripheral ring (LatP), and to perform the first rearing (LatR) were strongly dependent on age (A*, p = 0.031; A***, p = 0.001 and A***, p < 0.001, respectively). The performance of the first vertical exploration in the periphery was also modulated by genotype (LatR, G***, p < 0.001) and age x sex interaction in almost all the Int. J. Mol. Sci. 2023, 24, 13796 5 of 26 Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 5 o variables studied (A × S, p < 0.05). This effect was also observed in the retest, with higher latencies observed in 3xTg-AD mice and older animals in general (G and A; p < 0.05). CORNER TEST A) G*, A**, S** G*, A*** NTg2m B) NTg2m C) NTg2m Figure 1. NPS-like behaviors in the corner test—Effect of AD genotype, age, and sex on the neophobia Figure 1. NPS-like behaviors in the corner test—Effect of AD genotype, age, and sex on the elicited in a new home cage at 2, 4, 6, 12, and 16 months. 2m-NTg (n = 16, 8 males and 8 females); 2m- phobia elicited in a new home cage at 2, 4, 6, 12, and 16 months. 2m-NTg (n = 16, 8 males an 3xTg-AD (n = 16, 8 males and 8 females); 4m-NTg (n = 16, 8 males and 8 females); 4m-3xTg-AD (n = 15, females); 7 males and 2m-3xTg-AD (n = 16, 8 females); 6m-NTg (n =8 21, males andand 11 males 8 females); 4m-NTg 10 females); (n = 16, 6m-3xTg-AD (n =8 21, males and and 11 males 8 females); 3xTg-AD (n = 15, 7 males and 8 females); 6m-NTg (n = 21, 11 males and 10 females); 10 females); 12m-NTg (n = 17, 9 males and 8 females); 12m-3xTg-AD (n = 18, 9 males and 9 females); 6m-3xTg-A =16m-NTg 21, 11 males(n = 14,and 10 females); 8 males 12m-NTg and 6 females); (n = 17, 9 (n 16m-3xTg-AD males = 18,and 8 females); 9 males 12m-3xTg-AD (n = 1 and 9 females).Visited males and 9 females); 16m-NTg (n = 14, 8 males and 6 females); 16m-3xTg-AD corners (A), number (B), and latency (C) of rearings. Data are expressed by Mean ± SEM. Individual (n = 18, 9 males 9values: females).Visited Circles: NTg corners (A), number mice; Squares: 3xTg-AD(B),mice;and Meanlatency values:(C) ofbars: Blue rearings. males;Data are expressed Red bars: females; by M ±Ice SEM. Individual values: Circles: NTg mice; Squares: 3xTg-AD mice; Mean blue: NTg males; Navy blue: 3xTg-AD males; Pale red: NTg females; Carmine red: 3xTg-AD values: Blue bars: m Red bars: females; Ice blue: NTg males; Navy blue: 3xTg-AD males; Pale red: females. Pointed lines: averaged performance of NTg mice of that sex at 2 months of age is indicatedNTg females; Carm red: 3xTg-ADlevel. as a reference females. Pointed Statistics: lines: Analysis averagedStatistics: of variance: performance of NTg G, genotype; A, mice age; S,ofsex; that * psex at 2 month < 0.05, age ** p