*3.3. Behavioral Assessment for Physical, Emotional, and Cognitive Phenotypes*

At 14 months of age, a comprehensive screening of three main behavioral dimensions and functions, namely, physical, emotional, and cognitive, was performed using a battery of seven tests as previously described [37].

## 3.3.1. Physical Phenotype

Genotype differences were found in body weight (Figure 1C), which was increased in 3xTg-AD mice (Genotype (G), F(1,32) = 5.204; *p* = 0.030), an effect that was more clearly observed among females (post hoc test, *p* < 0.05).

Frailty score (Figure 1D) was increased in 3xTg-AD mice (G, F(1,32)= 12.052, *p* = 0.002) with a statistically significant two-fold increase in male mice as compared to NTg counterparts (post hoc test, *p* = 0.001), while NTg and 3xTg-AD females exhibited similar frailty scores (GxS; F (1,32) = 6.136; *p*= 0.020).

3.3.2. Neuropsychiatric symptoms (NPS)-like phenotype and cognitive impairment under different anxiogenic conditions

In the corner test (CT) for neophobia (Figure 2A), NTg females exhibited increased behavior, as measured by a higher number of visited corners and faster onset of rearing than NTg males (Student *t*-test, *p* < 0.05). In the 3xTg-AD mice, the behavior was slightly increased compared to male NTg response but did not reach statistical significance. Overall, the sex difference was shown in the visited corners (S, F(1,36) = 8.032, *p* = 0.008). Genotype per sex interaction effects in the variables for vertical exploratory behavior indicated the consistent results between male and female 3xTg-AD mice in this regard, while in the NTg mice, sex differences were shown (GxS, F(1,36) = 4.267, *p* = 0.047).

In the open-field (OF) test (Figure 2B), male and female NTg mice behaved quite similarly, as shown by the time course and total counts of their horizontal and vertical activities. In contrast, male 3xTg-AD mice exhibited sustained horizontal activity during the test and higher total counts than their NTg counterparts (OF1, repeated measures ANOVA (RMA); crossings: time × genotype × sex, F(1,36) < 0.001, *p* = 0.003). Female 3xTg-AD behaved like NTg mice, with a drop of activity from the first to the second minute of the test. These patterns resulted in statistically significant genotype x sex interaction effects (OF1 min 5, F(1,36) = 17.187, *p* = 0.001). When central and peripheral activity were distinguished, male 3xTg-AD mice showed an increased number of peripheral crossings in the third, fourth, and fifth minutes of the test (not shown) and, as a result, also on the total number, as compared to NTg animals and 3xTg-AD females.

In the repeated corner test (24 h later), all the groups showed lower levels of activity (paired *t*-test, *p* < 0.05). The genotype × sex factors interaction effects shown on the first day were also found here.

In the repeated open-field test (24 h later), the time course of horizontal and vertical activity was also dependent on the genotype and sex or both (OF2, RMA crossings: TxG, TxS and RMA rearing: TxG, TxS, TxGxS, all Fs (1,36) > 4.251.000, *p* < 0.047). NTg mice performed similar total activity levels than the precedent day. During the first minute of the repeated open-field test, performance did not differ from that shown in their first experience in the test (OF21 vs. OF11, *n.s*. in all the groups). Here, genotype effects on the number of crossings and rearing in the first minute of the test (F(1,36) > 4.584, *p* < 0.05) indicated higher performances in 3xTg-AD mice than NTg mice. Still, male 3xTg-AD reduced their total activity to control levels.

A GxS effect was observed when we calculated the difference between the crossings performed in the first minute of the test on day2 (OF21) and those in the last minute of day 1 (OF15) (F(1,36) = 10.889, *p* = 0.002), with only female 3xTg-AD mice differing from their NTg counterparts (*p* < 0.05).

In the dark–light box (DLB) test (Figure 2C), 3xTg-AD mice exhibited a disinhibitory behavior, as shown by the increased number of crossings in the lit area (data not shown, G, F(1,35)= 5.186, *p* = 0.03) and increased total time spent into it (G, F(1,35) = 4.387, *p* = 0.044), as compared to the NTg genotype. Emotionality in 3xTg-AD mice was also increased, as they spent more time grooming (G, F(1,35) > 4.919, *p =* 0.034).

**Figure 2.** Mental health: Neuropsychiatric-like phenotype and cognitive impairment under different anxiogenic conditions. (**A**) Two-day corner test, (**B**) two-day open-field test, (**C**) dark–light box, (**D**) marble test. Results are the mean ± SEM. Bars illustrate the genotype groups, as indicated in the abscissae. Symbols are used to illustrate individual values of males (black, left panel) and females (red, right panel). In green: the no-survivors. Genotype (G) and sex (S) effects and GxS interaction were analyzed by 2 × 2 factorial ANOVA analysis, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001 (above line). Time (T) factor (day-by-day) was analyzed by repeated measures ANOVA, <sup>T</sup> *p* < 0.05, TT *p* < 0.01, vs. the corresponding day 1 results. Student *t*-test comparisons: \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001 vs. the corresponding NTg group; # *p* < 0.05, ### *p* < 0.001 vs. the corresponding male group.

In the marble (MB) test (Figure 2D), 44.4% (4/9) of the marbles were buried on average by the groups except in the group of 3xTg-AD females that buried only 22.2% (2/9), resulting in a sex difference not found in NTg genotype (GxS, F (1,34) = 4.254, *p =* 0.048). A sex effect was also observed when we measured the marbles left intact (not shown, S, F (1,34) = 4.501, *p* = 0.042).

Two different paradigms were carried out in a T-shaped maze, as depicted in Figure 3A; all the groups included animals that failed to complete the T-maze spontaneous alternation test (latency, 300 s) and the forced memory test (latency, 600 s). On average, all the groups showed similar latencies in the ethogram of behaviors exhibited in the T-maze spontaneous alternation (TMSA) test (turning, reaching the intersection, crossing the intersection with four paw criteria, completing the test) (G, S, all Fs (1,34) < 0.231; *p* > 0.05). The most representative of these latencies—that of achieving "test completion criteria"—is illustrated. The number of spatial working memory errors (revisiting an explored area) were recorded in those animals able to initiate the task. No statistically significant differences were observed between groups in spatial alternation.

In the second paradigm for working memory in the T-maze (TM) test (Figure 3B), all the groups needed the same time to reach the acquisition criteria in the forced trial (G, S, all Fs(1,33) < 0.708; *p* > 0.05). In the recall trial, considering the animals that completed the test (n = 25; nine NTg males; four NTg females; seven 3xTg-AD males and five 3xTg-AD females), sex differences, clearer among NTg, were found (S, F (1,25) = 10.063, *p* = 0.005), with males investing shorter times than females to reach the exploration criteria. Despite the small number of animals, an increased number of errors in working memory (revisiting an explored area) was noted in females (S, F (1,19) = 34.135, *p* < 0.001).

In the Morris water maze (MWM) (Figure 3C), genotype differences were found in the navigation speed, reaching statistically significant differences in the cue learning paradigm. Thus, 3xTg-AD mice were swimming slower than NTg counterparts (CUE: G, F (1,32) = 4.437, *p* = 0.044), an effect that was more clearly shown in females (*p* < 0.05). Therefore, the distance covered to reach the platform was used to illustrate all the paradigms' performances. Genotype x sex interaction effects in the cue and the place tasks (PT) indicate that males' and females' performances were dependent on the genotype. Here, sex differences in the cue learning were shown in NTg mice, but not 3xTg-AD mice. Female NTg covered more distance than NTg males, whereas both sexes performed equally in the 3xTg-AD genotype. At the end of the four trial sessions of the cue learning task, all the groups could reach the platform in 20 s and cover 2 to 3 m.

In the two daily sessions of the place learning task, where the platform was hidden and located in a reversed position, animals exhibited a genotype effect in the distance covered to find the new location of the platform (PT11: G, F(1,32) = 6.228, *p* = 0.019). NTg mice were faster (shorter latency, not shown) and covered less distance to find the hidden platform. After that, the performances between NTg and 3xTg-AD mice differed in some trials. Genotype effect was also observed in the last trial of the second day; in this case, as in the mean distance of place task, two 3xTg-AD male mice performed less distance to arrive at the platform (PT24: G, F (1,32) = 4.964, *p* = 0.034; meanPT2, G, F(1,32) = 5.926, *p* = 0.022). The sex effect was observed in PT23, with females covering more distance to find the platform (PT23: S, F(1,32) = 4.716, *p* = 0.039).

In summary, different survival and behavioral signatures were found in these cohorts. Namely, (1) physical: An early mortality window of the female sex found enhanced in the AD-genotype and increased frailty only in male 3xTg-AD mice. (2) Neuropsychiatric-like: increased and persistent neophobia in female 3xTg-AD mice, a hyperactive pattern of male 3xTg-AD mice, and disinhibitory behavior in male and female 3xTg-AD mice. (3) Worse long-term memory of female 3xTg-AD in the open-field test; overall bad performances of all the animals in the mazes, with worse performance and working memory of female 3xTg-AD mice, a slower swimming speed of female 3xTg-AD mice, and paradoxical performances of male 3xg-AD mice, probably related to emotional and physical comorbidities. As discussed in previous work, these sex-dependent effects point at the relevance of the sex-specific analysis of AD disease. The results also illustrate the relevance of controlling for frailty and mortality rates to discriminate against the confounding factors (synopsis; Table 1).

**Figure 3.** Mental health: cognitive impairment (**A**) Spontaneous alternation T-maze, (**B**) T-maze, (**C**) Morris water maze. Results are the mean ± SEM. Bars illustrate the genotype groups, as indicated in the abscissae. Symbols are used to illustrate individual values of males (black, left panel) and females (red, right panel). In green: the no-survivors. Genotype (G) and sex (S) effects and GxS interaction were analyzed by 2 × 2 factorial ANOVA analysis, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001 (above line). Student *t*-test comparisons: \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001 vs. the corresponding NTg group; # *p* < 0.05, ### *p* < 0.001 vs. the corresponding male group.


**Table 1.** Synopsys of main genotype and sex effects in the physical condition and behavioral phenotype.

Domains studied, tests used (CT, corner test; OF, open-field test; OF2, open-field test day 2; DLB, dark–light box test and MWM, Morris water maze). Statistics: genotype and sex effect; between groups differences, \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001 and related figures.
