ORIGINAL RESEARCH published: 24 September 2019 doi: 10.3389/fphar.2019.01061 Impact of Chronic Risperidone Use on Behavior and Survival of 3xTg-AD Mice Model of Alzheimer’s Disease and Mice With Normal Aging Virginia Torres-Lista 1,2, Secundí López-Pousa 3 and Lydia Giménez-Llort 1,2* 1 Medical Psychology Unit, Department of Psychiatry and Forensic Medicine, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain, 2 Institut de Neurociències, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain, 3 Research Unit and UVaMiD (Memory and Dementia Assessment Unit), Institut d’Assistència Sanitaria, Salt, Spain Psychosis and/or aggression are common problems in dementia, and when severe or persistent, cause considerable patient distress and disability, caregiver stress, and early institutionalization. In 2005, the Food and Drug Administration (FDA) determined that atypical antipsychotics were associated with a significantly greater mortality risk compared Edited by: to placebo, which prompted the addition of an FDA black-box warning. The American Yukihiro Ohno, College of Neuropsychopharmacology (ACNP) White Paper, 2008, reviewed this issue Osaka University of Pharmaceutical Sciences, Japan and made clinical and research recommendations regarding the use of antipsychotics Reviewed by: in dementia patients with psychosis and/or agitation. Increased mortality risk has also Yukio Ago, been described in cerebrovascular adverse events in elderly users of antipsychotics. In Osaka University, Japan Karolina Pytka, the present work, at the translational level, we used male 3xTg-AD mice (PS1M146V, Jagiellonian University, Poland APPSwe, tauP301L) at advanced stages of the disease reported to have worse survival *Correspondence: than females, to study the behavioral effects of a low chronic dose of risperidone (0.1 mg/ Lydia Giménez-Llort kg, s.c., 90 days, from 13 to 16 months of age) and its impact on long-term survival, as lidia.gimenez@uab.cat compared to mice with normal aging. Animals were behaviorally assessed for cognitive Specialty section: and BPSD (behavioral and psychological symptoms of dementia)-like symptoms in This article was submitted to naturalistic and experimental conditions (open-field test, T-maze, social interaction, Morris Neuropharmacology, a section of the journal water maze, and marble test) before and after treatment. Weight, basal glucose levels, Frontiers in Pharmacology and IPGTT (i.p. glucose tolerance test) were also recorded. Neophobia in the corner Received: 28 May 2019 test was used for behavioral monitoring. Survival curves were recorded throughout the Accepted: 20 August 2019 Published: 24 September 2019 experiment until natural death. The benefits of risperidone were limited, both at cognitive Citation: and BPSD-like level, and mostly restricted to burying, agitation/vibrating tail, and other Torres-Lista V, López-Pousa S and social behaviors. However, the work warns about a clear early mortality risk window Giménez-Llort L (2019) Impact of during the treatment and long-lasting impact on survival. Reduced life expectancy and Chronic Risperidone Use on Behavior and Survival of 3xTg-AD Mice Model life span were observed in the 3xTg-AD mice, but total lifespan (36 months) recorded in of Alzheimer’s Disease and Mice C57BL/6 × 129Sv counterparts with normal aging was also truncated to 28 months in With Normal Aging Front. Pharmacol. 10:1061. those with treatment. Sarcopenia at time of death was found in all groups, but was more doi: 10.3389/fphar.2019.01061 severe in wild-type animals treated with risperidone. Therefore, the 3xTg-AD mice and Frontiers in Pharmacology | www.frontiersin.org 1 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics their non-transgenic counterparts can be useful to delimitate critical time windows and for studying the physio-pathogenic factors and underlying causal events involved in this topic of considerable public health significance. Keywords: comorbidities, antipsychotics, risk factors, mortality, aging, memory, neuropsychiatric symptoms, social behavior INTRODUCTION in long-term therapy, whereas for agitation symptoms, the results are not well established. In addition, there’s an increase Behavioral changes and neuropsychiatric symptoms (NPS), also in the concerns about adverse outcomes with these treatments, known as ’’behavioral and psychological symptoms of dementia including strokes and death (Ballard et al., 2009). (BPSD),” may occasionally signal the onset of Alzheimer’s disease In April 2005, the FDA issued the following warning for all (AD) (Reisberg et al., 1987). They are present in 90% of patients as atypical antipsychotics based on their evaluations: “The FDA the disease progresses in a neurodegenerative process which is faster informed health professionals and the public about the increased and more severe in males in spite of similar incidence of AD among risk of mortality in elderly patients who received atypical sexes or women showing higher incidence with increasing age antipsychotic drugs for the treatment of dementia-related (Lapane et al., 2001; Pike, 2017; Rezanni et al., 2019). The prevalence psychosis. Analyses of 17 placebo-controlled trials involving 5,377 of delusions in patients with AD is between 9% and 63% and that elderly patients with conduct disorders associated with dementia of hallucinations is between 4% and 41%; the rate of aggression is revealed a risk of death in patients treated with the drug between between 11% and 46% and agitation has an even higher prevalence 1.6 and 1.7 times that observed in patients treated with placebo. rate among 20–80% (Jeste et al., 2008). Most of these symptoms The mortality rate in the drug-treated patients was approximately diminish the quality of life of the patients and, particularly, the 4.5% compared to a rate of approximately 2.6% in the placebo psychosis is associated with a rapid cognitive deterioration (reviewed group. Although the causes of death were varied, most of the deaths by Cummings, 2000 and Kalman et al., 2008). The cognitive decline appeared to be cardiovascular (for example, heart failure, sudden characteristic of dementia is also associated with an increase in social death) or infectious (pneumonia)”. Based on this analysis, the FDA vulnerability in humans that sometimes leads to death (Andrew and requested manufacturers of the atypical antipsychotic drugs to Rockwood, 2010). Therefore, these neuropsychiatric symptoms include information about this risk in the package leaflet of the present diverse clinical implications in patients, as is the specific case drug (Jeste et al., 2008). These alerts brought a diversity of opinions of psychosis, agitation, and aggression, which increase the burden of in the scientific community. Some authors considered the warnings disease, also resulting a strong cause of distress among the familiar about atypical antipsychotics as alarming and potentially harmful and professional caregivers (Tan et al., 2005). for patients with dementia, while others were concerned that there During the last two decades, the most recent atypical was no clear evidence to support a greater benefit in relation to antipsychotic drugs that have been approved by the Food and atypical antipsychotics compared to conventional antipsychotics Drug Administration of the US Department of Health and Human (Trifiro et al., 2009). In a previous work at our UVaMiD neurology Services (US FDA) are risperidone in 1993, olanzapine in 1996, unit, we studied the mortality risk in AD patients at advanced ages quetiapine in 1997, ziprasidone in 2001, and aripiprazole in 2002. of the disease who received risperidone therapy, but we could not These atypical antipsychotics have replaced the first-generation find any relationship to metabolic syndrome nor history of heart antipsychotics such as haloperidol and thioridazine (Schneider et disease (Vilalta-Franch et al., 2008). al., 2005 and Jeste et al., 2008). Atypical antipsychotics are used This diversity of opinions can be observed in studies of as the first line of pharmacological approach for the treatment circadian cycle disorders where patients with dementia after of neuropsychiatric symptoms in AD (Ballard et al., 2009). In a 12-week treatment at a dose of 1.49 mg/day of risperidone the present work, we focused on risperidone, one of the most reported improvements in total sleep hours, waking hours in bed, used atypical antipsychotics and co-administered with different insomnia, and other variables related to sleep (Durán et al., 2005). types of drugs. At the pharmacological level, risperidone is a Other studies have indicated that in elderly patients with dementia selective monoaminergic antagonist, which has a high affinity low doses of risperidone 0.5 and 1 mg were well tolerated and with serotonergic receptors 5-HT2 and dopaminergic D2 and were associated with reductions in BPSD, in particular, agitation, binds also to α1 adrenergic receptors and with lower affinity aggression, irritability, delusions, sleep disturbances, anxiety, to histaminergic H1 and α2 adrenergic receptors. It has been and phobias. Despite efficacy in the reduction of various adverse approved by the FDA for the treatment of schizophrenia (positive symptoms, risperidone and olanzapine should not be used and negative symptoms), bipolar disorders, and autism. It is also routinely for the treatment of aggression and psychosis in patients used in dementia, depression, obsessive-compulsive disorders, with dementia (Onor et al., 2007). Other studies have indicated personality disorders, and attention deficit hyperactivity disorder that risperidone and olanzapine increase the risk of mortality in (Katz et al., 2007; Shekelle et al., 2007 and Rodriguez-Antona elderly patients with dementia with an increased risk of the latter et al., 2009). This drug has a moderate but significant effect on over the former (Vilalta-Franch et al., 2008). Also, mortality risks short-term treatment (> 6–12 weeks) for aggression but is limited are increased with high doses of atypical antipsychotics and the Frontiers in Pharmacology | www.frontiersin.org 2 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics causes of mortality are cerebrovascular accidents, respiratory engineered at the University of California Irvine, as previously diseases, and circulatory disorders (Huybrechts et al., 2012) described (Oddo et al., 2003b). Briefly, two independent compared with people who received placebo (Trifiro et al., 2009). transgenes (encoding human APPSwe and human tauP301L, Cerebrovascular risks (CVA) are especially observed during both under control of the mouse Thy1.2 regulatory element) were the first weeks of treatment; this risk decreases with time and co-injected into single-cell embryos harvested from homozygous normalizes after 3 months of treatment (Kleijer et al., 2009). mutant PS1M146V knock-in (PS1KI) mice. The PS1 knock-in Currently, atypical antipsychotics continue to be used under strict mice were originally generated as a hybrid C57BL/6 x 129Sv. supervision and monitoring in some hospitals and/or geriatrics Forty-six 12-month-old 3xTg-AD mice (n = 23) and for the BPSD. This is even though older people are more sensitive C57BL/6 x 129Sv (n = 23) wildtype mice (from now, referred as to their side effects than young and middle-aged adults, in part non-transgenic mice, NTg) from litters of a breeding program by the interaction of changes caused by age and pharmacological established in our laboratory at the Medical Psychology Unit, sensitivity to antipsychotic treatments (Salzman et al., 2008). Universitat Autònoma de Barcelona, were used in this study. All Due to the variety of results and the ethical impossibility of the animals were housed three to four per cage and maintained conducting new clinical studies in humans, it is important to (Makrolon, 35 × 35 × 25 cm) under standard laboratory model the pharmacological responses of antipsychotics in animal conditions (12 h light/dark, cycle starting at 8:00h, food and models of the disease and to study their effects in the BPSD. At water available ad libitum, 22 ± 2ºC, 50–60% humidity). The the translational level, the triple transgenic mouse for AD hosts circadian activity was recorded during one whole light/dark (LD) human transgenes PS1/M146V, APPswe, and tau P301L (Oddo period, and the rest of the tests from 9:00h to 13:00h. et al., 2003a). These rodents uniquely mimic various symptoms This study was carried out in accordance with the of the disease in a temporal and neuroanatomical pattern similar recommendations of ARRIVE guidelines developed by the NC3Rs to that observed in humans (Belfiore et al., 2019). The onset of (Kilkenny et al., 2010) and the Spanish legislation on “Protection symptoms has been established between 4 and 6 months of age of Animals Used for Experimental and Other Scientific Purposes” and involves electrophysiological deficits (in LTP, long term and the European Communities Council Directive (2010/63/ potentiation and fEPSP, field excitatory postsynaptic potential) EU) on this subject. The protocol CEEAH 2481/DMAH 8700 at the hippocampal level, learning and memory problems, entitled “Risk factors and preventive/therapeutical strategies in cholinergic deficiencies, and emotional disturbances. However, Alzheimer’s disease: studies in triple-transgenic 3xTg-AD mice” at these ages, the brains of the animals only show presence of was approved by Departament de Medi Ambient i Habitatge, intraneuronal immunoreactivity of Aβ (Kitazawa et al., 2005; Generalitat de Catalunya. Giménez-Llort et al., 2007; Oddo et al., 2003a; Oddo et al., 2003b). After 12 months of age, the neuropathological profile finds its parallelism with the advanced stages of the disease in humans, Experimental Design and Risperidone with the characteristic deposits of Aβ and neurofibrillary tangles Treatment of tau protein (Oddo et al., 2003a; Oddo et al., 2003b). We have A longitudinal study divided into successive phases including previously shown increased mortality in male 3xTg-AD mice as a “before–after treatment” design was performed. The study compared to females, ranging from 33% (Giménez-Llort et al., started at 12 months of age; that in the 3xTg-AD mice has been 2008) to 100% at 15 months of age (García-Mesa et al., 2012) extensively reported mimicking neuropathological hallmarks and its relation to impaired neuroimmunoendocrine system of the disease (Belfiore et al., 2019) and that in the NTg mice (Giménez-Llort et al., 2014). The increased impact of AD (faster, (C57BL/6 x 129Sv genetic background) corresponds to middle more severe) of male sex has recently also been reported in the age. The sample of NTg mice was segregated into two groups human patient (Pike, 2017; Rezanni et al., 2019). according to the activity levels exhibited in the corner test The present study aims to model in 3xTg-AD mice the (CT) for neophobia and the open-field (OF) test, to be used vulnerability that leads to an increase in mortality observed in as controls that will be treated with saline (NTg mice with low patients with AD chronically treated with atypical antipsychotics motor activity) or risperidone (NTg mice with high motor such as risperidone. Before we can address this modeling, we activity), respectively. defined the starting phenotype of the subject of study. To this Risperidone was used at a dose of 0.1 mg/kg equivalent end, the animals were evaluated in a battery of tests for measuring to that administered in patients with AD and used in most exploratory activity, anxiety, learning and memory, burying of experimental work performed in rodents (Bruins Slot et al., objects, and social behavior. An assessment of the basal state 2005). The chronic administration, subcutaneous for 3 months of glucose and the tolerance response to it when administered from 13 to 16 months of age, rotated three injection sites (the intraperitoneally was also made. neck and the two flanks). First, we characterized the basal phenotype (phase 1, weeks 1–6, phenotype “before treatment”; animals at 12 months of MATERIALS AND METHODS age). As in the case of geriatric patients, the treatment regimen was initiated with a lower dose of 0.05 mg/kg (phase 2, low dose Animals and follow-up tests, week 7; animals at 13 months of age). After Homozygous triple-transgenic 3xTg-AD mice harboring human 7 days, the final dose of 0.1 mg/kg was started and behavioral PS1/M146V, APPSwe, and tauP301L transgenes were genetically effects assessed (phase 3, treatment and behavioral effects Frontiers in Pharmacology | www.frontiersin.org 3 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics “after treatment”, weeks 8–16; animals until 15 months of age). Social Interaction Test (SIT) Thereafter, treatment followed without behavior and completed Behavioral signatures of social dysfunction in 3xTg-AD mice the total period of 3 months of subcutaneous treatment (phase 4, were assessed by means of the SIT (File and Hyde, 1978) as only treatment, until 16 months of age). From that moment and recently described (Torres-Lista and Giménez-Llort, 2019). A until the end of their days, the variables of weight and survival dyad of two unfamiliar mice of the same genotype and sex were were recorded continuously with a weekly or daily cadence, introduced in a standard home cage and video recorded for 5 respectively (phase 5, from 16 to 36 months of age). min. Behaviors were classified into social (social investigation, Four experimental groups were studied and are plotted in the aggression, vibrant tail) and non-social (exploring, digging, before/after graphs of the figures as follows: NTg mice (s) (NTg self-grooming) interactions. We also scored the total number of mice that will receive or have received saline, n = 12), NTg mice episodes and their total duration. (r) (NTg mice that will receive or have received risperidone, n = 11), 3xTg-AD mice (s) (3xTg-AD mice that will receive or have received saline, n = 12), and 3xTg-AD (r) (3xTg-AD mice that Morris Water Maze (MWM) will receive or have received risperidone, n = 11). A 5-day place learning task for short- and long-term spatial reference memory (four trial sessions per day, with trials spaced 30 min apart) was followed 2 h 30 min later by a probe trial Behavioral Assessments (removal of the platform) for short-term memory in the MWM Behavioral assessment consisted in a battery of naturalistic (Morris, 1981; Morris, 1984). Mice were trained to locate a and experimental conditions (Giménez-Llort et al., 2007). hidden platform (7-cm diameter, 1 cm below the water surface) Neophobia in the CT was used for behavioral monitoring in a circular pool for mice (Intex Recreation Corp., Long Beach, through the treatment. CA, United States; 91-cm diameter, 40-cm height, 25°C opaque water), located in a completely black painted 6-m2 test room. Corner Test (CT) Mice that failed to find the platform within 60 s were placed Animals were individually placed in the center of a clean standard on it for 10 s, the same period as was allowed for the successful home cage, filled with wood shave bedding. Number of corners animals. White geometric figures, one hung on each wall of the visited were recorded during 30 s (Belzung and Le Pape, 1994). room, were used as external visual clues. Behavior was evaluated Latency to realize the first rearing, and the number of rearings by direct observation and analysis of videotape-recorded images. were also registered (Giménez-Llort et al., 2007). Variables of time (escape latency) and quadrant preference and entries were analyzed in all the trials of the tasks. The escape latency was readily measured with a stopwatch by an observer Open Field Test (OF) unaware of the animal’s genotype and confirmed during the Immediately after the CT, mice were placed in the center of an subsequent video-tracking analysis. In the probe trial, the time open field (homemade woodwork, white box, 50 × 50 × 20 cm) spent and number of entries in each of the four quadrants were and observed for 5 min (Hall and Ballachey, 1932). The also measured retrospectively by means of the automated video- ethogram, described by the temporal profile of the following tracking analysis. sequence of behavioral events, was recorded: duration of freezing behavior, latency to leave the central square and that of entering the peripheral ring, as well as latency and total duration of self- Marble Burying Test (MB) grooming behavior. Horizontal (crossings of 10 × 10 cm squares) The procedure for MB was adopted with minor modifications and vertical (rearings with a wall support) locomotor activities from that originally described by Broekkamp et al. (1986). were also measured. Bizarre behaviors observed in this test were Mice were placed individually in a standard home cage also measured according to the previous reported criterion containing six glass marbles (1 × 1 × 1 cm) evenly spaced (Baeta-Corral and Giménez-Llort, 2014). During the tests, making a triangle (three rows of three, two, and one marble defecation boli and urination were also recorded as measures of per row only in the left area of the cage) on a 5-cm-thick individual differences in emotionality (Hall, 1934). layer of sawdust. The mice were left in the cage with marbles for a 30-min period after which the test was terminated by removing the mice and counting the number of marbles: intact T-Maze (TM) (untouched), rotated or at least half buried by sawdust, and Working memory was assessed by means of a spontaneous buried (completely hidden) as previously described (Torres- alternation task (Douglas, 1966) in a black TM. The apparatus Lista et al., 2015) consisted of a woodwork, three arms of 30 × 5 × 20 cm connected by a 5 × 5 × 20 cm intersection. The animal was placed inside the “vertical” arm of the maze with its head facing the end wall, Body Weight (BW) and Basal Glucose and it was allowed to explore the maze during a maximum of 3 Levels (G) min. Freezing behavior (latency to move), the latency to reach Throughout the experimental process, evaluation of weight and the intersection, the total time invested to explore the three arms survival was continuously monitored until the natural end of of the maze, and the number of errors (revisiting an arm) were the life of the animals. The blood samples were taken from an recorded. Defecation boli and urination were also noted. incision made at the tip of the tail. Frontiers in Pharmacology | www.frontiersin.org 4 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics Survival Curve from weeks 8 to 12, with all the [F(1,43) < 25.961, P < 0.001] The survival curves were obtained with the percentage of animals and interaction effect genotype × treatment in week 10 [F(1,43) = that were maintained throughout the experimental procedure. 5.975, P < 0.05] (Figure 1C). Statistics Open Field Test (OF) Statistical analysis were performed using SPSS 17.0 software. The delayed ethogram (sequence of behavioral events), the reduced The results are expressed as means ± SEM or percentage. A 2×2 horizontal and vertical activities, and the increased emotionality factorial design with multivariate general lineal model analysis behavior (urination) exhibited in the OF described an increased evaluated genotype (G) and treatment (T) effects, followed by anxious-like profile in the 3xTg-AD mice. After treatment, genotype post hoc Tukey B test. In the Morris water maze, the factor ‘day and treatment effects were found as detailed in Figure 2. In the (D)’ was included when appropiate. Student’s t-test was used to NTg mice treated with saline, the latencies of “first movement” compare two independent groups. The comparisons for related (freezing behavior) (t = 2.353, gl 11, P < 0.05) and “grooming” (t = samples were made with the paired t-test. Survival curve was 2.935, gl 11, P < 0.05) were delayed as compared before treatment analyzed with Kaplan-Meier test. The correlations between (Figures 2A, B). In addition, there was a decrease in the “total survival and the different variables studied were evaluated with distance traveled” (t = 4.674, gl 11, P < 0.01) (Figure 2E). In the Pearson’s correlation. In all the tests, P < 0.05 was considered contrast, the sequence of behaviors [“first movement”, “leaving the statistically significant. center,” “entering into the periphery,” all t(10) < 14,317, P < 0.001; “grooming,” t = 3,681, gl 10, P < 0.01] was faster in those receiving risperidone. A decrease in the “total number of rearings” (t = Results 2.948, gl 10, P < 0.05) was observed (Figure 2D). In the 3xTg-AD We confirmed that the sample of 3xTg-AD mice studied mice treated with saline and risperidone, the “latency to enter the exhibited cognitive deficits in the MWM, mimicking the periphery” was advanced in time (t = 3.236, gl 11, P < 0.01 and cognitive hallmark of AD. However, only data of animals that t = 4.071, gl 10, P < 0.01, respectively). Likewise, 3xTg-AD (r) mice could be included in the “before–after” analysis were considered showed a decrease in the horizontal activity “distance traveled” (animals dying in phases 1, 2, and 3 were excluded). Statistics of (t = 4.104, gl 10, P < 0.01) (Figure 2E) and in the vertical activity genotype effects for the different behavioral tests and variables “total number of rearings” (t = 2.512, gl 10, P <0.05) indicating studied in 3xTg-AD and NTg mice at 12 months of age (week differences between the untreated phase and the treatment phase 1, basal, but without the segregation for the treatment they will at 0.1 mg/kg (Figure 2D). When the four groups of treated animals receive) are cited in the text and depicted in Supplementary were compared with each other, a genotype effect [F(1,43) = Table 1. Behavioral correlates with lifespan in animals treated 20.950, P < 0.001], treatment effect [F(1,43) = 6.802, P < 0.05], with saline or risperidone are also indicated. Figures 1–12 depict and interaction effect “genotype × treatment” [F(1,43) = 13.161, the effects of chronic risperidone on these behaviors, weight, P < 0.01] was detected in the variable “total distance traveled.” and survival curves of animals. Finally, Table 1 details the This was due to the fact that the NTg (r) mice performed greater behavioral correlates with lifespan in male NTg and 3xTg-AD horizontal exploratory activity with respect to the group of NTg mice chronically treated with saline or risperidone. (s) mice, while the 3xTg-AD (r) mice performed less horizontal activity with respect to the 3xTg-AD mice (s) (Figure 2E). In addition, in the variables of vertical activity, it was observed that Corner Test (CT) the “latency of the first rearing” showed treatment effect [F(1,43) = Increased neophobia exhibited by 3xTg-AD mice before 4.376, P < 0.05]. Similarly, the “total number of rearings” indicated the treatment, as shown by reduced number of corners and genotype effect [F(1,43) = 16.431, P < 0.001], treatment [F(1,43) = rearings and increased latency of rearing, was not ameliorated 5.594, P <0.001], and interaction “genotype × treatment” [F(1,43) = by risperidone (Figure 1). The repeated CT allowed to observe 4.410, P < 0.05] (Figures 2C, D). The variable “grooming latency” a reduction of the number of corners through the treatment as showed genotype effect [F(1,43) = 26.950, P < 0.001] because time compared to basal levels, in all groups [Figure 1A, NTg (s), t = was delayed in 3xTg-AD mice relative to the other two NTg groups. 3.015, gl 11, P < 0.05; NTg (r), t = 8.517, gl 10, P < 0.001; 3xTg- Regarding the variable “total number of grooming,” a “treatment AD(s), t = 2,620, gl 11, P < 0.05; 3xTg-AD(r), t = 2.776, gl 10, P < × genotype” interaction effect was observed [F(1,43) = 8.718, 0.05]. Likewise, a decrease in vertical activity was observed in P <0.01] due to a decrease in behavior in the groups of 3xTg-AD NTg (r) mice (t = 2.637, gl 10, P < 0.05) and 3xTg-AD (s) mice mice with respect to the other two NTg groups. In the “urine,” a (t = 2.327, gl 11, P < 0.05) (Figure 1C). When the four groups genotype effect was observed [F(1,43) = 11.133, P < 0.01] due to were compared, genotype effects were still shown in the ”number higher incidence of urine in the 3xTg-AD groups with respect to of corners visited” in weeks 11 [F(1,43) = 6.623, P < 0.05] and 12 the NTg groups (Figure 2F). [F(1,43) = 15,503, P < 0.001]. The latency of rearing was more sensitive to the genotype effect as it was observed from weeks 9 to 12 [all F(1,43) < 14,450, P < 0.001] and showed a treatment effect T-Maze (TM) in week 9 [F(1,43) = 4.495, P < 0.05] and interaction “genotype × Behavioral patterns exhibited in the TM differing between treatment” in week 10 [F(1,43) = 7.151, P < 0.05] (Figure 1B). In genotypes were referred to emotionality behavior (urination), the variable “number of rearings,” genotype effect was observed which was increased in the 3xTg-AD mice. As shown in Frontiers in Pharmacology | www.frontiersin.org 5 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 1 | Effects of chronic risperidone assessed in the corner test. Results in the corner test before (basal) and through different weeks of treatment. (A) Number of corners, (B) latency of first rearing, and (C) number of rearings. ANOVA 2×2, genotype effect (G), genotype × treatment interaction (G×T), ***p < 0.001, **p < 0.01, and *p < 0.05. Post hoc Tukey B test *p < 0.05 vs. all other experimental groups; ap < 0.05 vs. different genotype but the same treatment; and cp < 0.05 vs. different genotype and different treatment. t -test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs. no treatment, &&&p < 0.001 and &p < 0.05. Figure 3, before–after differences were only shown in the P < 0.05] because the groups of 3xTg-AD mice performed a groups of 3xTg-AD mice. Animals treated with saline showed greater number of episodes with respect to the two NTg groups. an increase in the number of defecation boli (t = -4.486, gl 11, “Genotype × treatment” interaction effects were shown in the P < 0.01), while the 3xTg-AD (r) group needed longer “time to “number of fecal boli” [F(1,43) = 4.930, P < 0.05] and “total complete the maze” (t = -2.578, gl 10, P < 0.05). When the four number of urine” [F(1,43) = 11.381, P < 0.01]. groups of animals were compared with each other, the “latency to cross the intersection” indicated genotype effects [F(1,43) = 5.668, P < 0.05] due to a delay in the 3xTg-AD mice with respect Social Interaction Test (SIT) to the two groups of NTg mice. The genotype effect was also In this test, exhibition of social (Figure 4) and non-social evidenced in the variable “exploratory activity” [F(1,43) = 4.496, (Figure 5) behaviors were analyzed. “Face/body contact,” Frontiers in Pharmacology | www.frontiersin.org 6 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 2 | Effects of chronic risperidone assessed in the open field test. Results in the open field test before (basal) and during treatment. (A) Latencies before treatment; (B) latencies after treatment; (C) number of rearings; (D) total number of episodes of rearing and self-grooming; (E) total distance covered; (F) total number of episodes of defecation and urine. ANOVA 2×2, genotype effect (G), treatment (T) and genotype × treatment interaction (G×T), ***p < 0.001, **p < 0.01, and *p < 0.05. Post hoc Tukey B test ap < 0.05 vs. different genotype but the same treatment and cp < 0.05 vs. different genotype and different treatment. Effect t-test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs. before treatment &&&p < 0.001, &&p < 0.01, and &p < 0.05. “anogenital contact,” “vibrant tail,” and “aggressive contact” P <0.05) (Figure 4C). The 3xTg-AD (r) mice showed statistical presented distinct latencies of appearance and number of episodes differences in the appearance of the “face/body contact” behavior depending on the genotype and were modified by treatment. Before (advancement, t = 2.754, gl 10, P < 0.05) and “vibrant tail” (delay, treatment, reduced body/face interaction, horizontal and vertical t = -4.643, gl 10, P < 0.01) (Figure 4A). These differences were exploratory behaviors but increased vibrating tail and digging also reflected in the duration and in the number of episodes, since were phenotypic characteristics of 3xTg-AD mice (Figure 4B). the animals spent more time in the “face/body contact” behavior After treatment, in NTg (s) mice, the face/body contact behavior (t = 2.443, gl 10, P < 0.05) and it increased the number of episodes showed a decrease in the “total number of episodes” (t = 2.217, “face/body” and decreased the number of “vibrant tail” (t = 2.372, gl 11, P < 0.05) (Figure 4C). In the NTg (r) mice, it was observed gl 10, P < 0.05) (Figures 4B, C). When comparing the four groups that the “face/body” latency was delayed in time (t = 3.006, gl 10, with each other, the variable “face/body contact latency” showed P < 0.05) and the “vibrant tail” latency was advanced (t = 3.738, a “genotype × treatment” interaction effect [F(1,43) = 8.4428, gl 10, P < 0.01) (Figure 4A). In addition, there was a decrease P < 0.01], which was also evidenced in the variable total number in the contact time “face/body” (t = 4.107, gl 10, P < 0.01) and of episodes of “face/body contact” [F(1.43) = 9.199, P < 0.01] “anogenital” (t = 3.546, gl 10, P < 0.01) and an increase in “vibrant (Figures 4A, B). tail” (t = -4.282, gl 10, P < 0.01) (Figure 4B). Likewise, there was In the “latency of anogenital contact,” genotype effects were a decrease in the number of episodes in “face/body” (t = 5.255, observed [F(1,43) = 0.818, P < 0.01) with a delay in the groups gl 10, P < 0.001), “anogenital” (t = 3.975, gl 10, P <0.01) and an of 3xTg-AD mice as compared to the other two NTg groups increase in “vibrant tail” (t = -4.282, gl 10, P <0.01) (Figure 4C). (Figure 4A). Conversely, “total anogenital contact time” showed The mice 3xTg-AD (s) presented a decrease in the duration genotype [F(1,43) = 6.766, P < 0.05], treatment [F(1,43) = 5.855, “anogenital contact” (t = 2.520, gl 11, P < 0.05) (Figure 4B) and P < 0.05], and interaction “genotype × treatment” [F(1,43) = number of episodes of the “vibrant tail” behavior (t = 2.480, gl 11, 4.535, P < 0.05] effects (Figure 4B). The genotype effect was also Frontiers in Pharmacology | www.frontiersin.org 7 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 3 | Effects of chronic risperidone assessed in the T-Maze before (basal) and during the treatment. (A) Latencies (s), (B) time to complete the maze (s), and (C) total number of defecations and urine. ANOVA 2×2, genotype effect (G), genotype × treatment (GxT), ***p < 0.001 and *p < 0.05. Post hoc Tukey B test ap < 0.05 vs. different genotype but the same treatment and cp < 0.05 vs. different genotype and different treatment. t-Test effect for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs. no treatment, &p < 0.05. evidenced in the total number of “anogenital contact” episodes differences were observed in any of the three variables studied [F(1,43) = 9.264, P < 0.01], treatment [F(1,43) = 7.509, P < 0.01], for this behavior (Figures 4A, B, C). and interaction [F(1,43) = 5.017, P < 0.05] (Figure 4C). This Figures 5A and B illustrate the pattern of non-social behaviors was due to the fact that the NTg (s) mice had greater contact before and after the treatment. In the NTg mice, changes were duration and number of episodes with respect to the other three only observed in those animals treated with the antipsychotic, experimental groups. In the “tail vibration latency,” “genotype × with a reduced number of rearings (t = 5.519, gl 10, P < 0.001) treatment” interaction effect was observed [F(1,43) = 5.396, P < and visited corners (t = 2.679, gl 10, P < 0.001) as compared to 0.05] (Figure 4A). Likewise, the variable “total time” presented basal levels (Figure 5B). In the 3xTg-AD (s) mice, the “latency of genotype effect [F(1,43) = 7.836, P < 0.01] as well as the variable rearing” was delayed in time (t = 2.205, gl 11, P < 0.05) resulting “number of episodes” [F(1,43) = 6.949, P < 0.05] because groups in a decrease of the “number of rearings” (t = 3.116, gl 11, P < of 3xTg-AD mice invested more time and presented an increase 0.01). This was also evidenced in the 3xTg-AD (r) mice where the of episodes with respect to the other two NTg groups (Figures “latency of rearing” was delayed in time (t = -2.323, gl 10, P < 0.05), 4B, C). In the aggressive contact component, no significant the “number of rearings” decreased (t = 5.723, gl 10, P < 0.001) and Frontiers in Pharmacology | www.frontiersin.org 8 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 4 | Effects of chronic risperidone in the social behaviors assessed in the social interaction test before (basal) and during the treatment. (A) Latencies (s). ANOVA 2×2, genotype effect (G), genotype × treatment interaction (G×T), **p < 0.01 and *p < 0.05. Post hoc Tukey B test *p < 0.05. vs. all other experimental groups; ap < 0.05 vs. different genotype but the same treatment and cp < 0.05 vs. different genotype and different treatment. t-test effect for paired data, treatment effect (T): treatment with a dose of 0.1 mg/kg vs. no treatment, &&p < 0.01 and &p < 0.05. (B) Total time and (C) total number of episodes. ANOVA 2×2, genotype effect (G) treatment effect (T) and genotype × treatment interaction effect (G×T), ***p<0.001 **p < 0.01 and *p < 0.05. Post hoc Tukey B test *p < 0.05. vs. all other experimental groups; at p < 0.05 vs. different genotype but the same treatment; bp < 0.05 vs. different treatment but the same genotype and cp < 0.05 vs. different genotype and different treatment. t-test effect for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs. no treatment, &&&p < 0.001, &&p < 0.01, and &p < 0.05. Frontiers in Pharmacology | www.frontiersin.org 9 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 5 | Effects of chronic risperidone in the non-social behaviors assessed in the social interaction test before (basal) and during the treatment. (A) Latencies (s) and (B) total number. ANOVA 2×2, genotype effect (G), treatment (T), ***p < 0.001 and *p < 0.05. Post hoc Tukey B test ap < 0.05 vs. different genotype but the same treatment and cp < 0.05 vs. different genotype and different treatment. Effect t-test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs. untreated &&&p < 0.001, &&p < 0.01, and &p < 0.05. also the “number of visited corners” reduced (t = 3.264, gl 10, P < treatment, statistical significant differences were only shown on 0.01) (Figures 5A, B). When comparing the four experimental day 1, with a “trial × genotype” [F(3,144) = 2.664, P < 0.05] groups with each other, these results were noted as genotype and “trial × genotype × treatment” [F(3,144) = 3.766, P < 0.05] effect in the “latency of the first rearing” [F(1,43) = 24,813, P < effects. No differences in the acquisition curves were found 0.001], “total number of rearings” [F(1,43) = 67.857, P < 0.001], among the four treated groups. When performance on this first and “total number of corners” [F(1,43) = 89.604, P < 0.001] due day after treatment was compared to the last performance before to the reduced activity of 3xTg-AD mice with respect to the two treatment, only NTg (s) mice showed worsen performance groups of NTg mice. The treatment effect was observed only in (t = -2.239, gl 11, P < 0.05). As compared with their respective the duration of the burying, which was reduced by risperidone acquisition curves before treatment, all groups of mice were [F(1,43) = 4.185, P < 0.05] in both genotypes. No effects of the faster finding the platform in this re-test. NTg (s) showed genotype or treatment were observed in the “grooming” variable. statistical differences on days 1, 2, and 5 (all t > 2.484, gl 11, P < 0.05), NTg (r) on days 1 and 2 (both t > 1.866, gl 10, P < 0.01), 3xTg-AD (s) from days 1 to 4 (all t > 1.950, gl 11, P < 0.01), and Morris Water Maze (MWM) 3xTg-AD (r) on all the days (all t > 2.096, gl 11, P < 0.01). Figure 6 depicts the before and after trial-by-trial (Figures 6A, The second paradigm consisted of the removal of the B) and day-by-day (Figure 6C) place learning task acquisition platform. Before treatments, NTg groups showed the highest curves as well as the platform preference shown in the probe number of entries in the trained quadrant, where the platform trial to assess memory after treatments (Figure 6D). The trial- was previously located, with respect to the other quadrants [NTg by-trial performance before treatment indicated that on each (s), F(3,47) = 13.48, P < 0.001; NTg (r), F(3,43) = 9.62, P < 0.001]. day, except for day 3, there was an effect of the “trial” [day 1, The number of entries in the trained platform was significantly F(3,144) = 9.914, P < 0.001; day 2, F(3,144) = 4.082, P < 0.01; lower in the 3xTg-AD groups. The detailed analysis showed that day 4, F(3,144) = 7.242, P < 0.001; day 5, F(3,144) = 4.934, P < preference was also shown for the right quadrant [3xTg-AD 0.01]. However, this basal performance in the 3xTg-AD mice (s), F(3,47) = 8.80, P < 0.001; 3xTg-AD (r), F(3,43) = 8.14, P < resulted in a worse day-by-day acquisition curve with higher 0.001] indicating lower focused search strategies in these mice. latencies as compared to NTg mice, on days 3, 4, and 5. After After treatment, the NTg (s) mice showed an increase in the Frontiers in Pharmacology | www.frontiersin.org 10 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 6 | Effects of chronic risperidone in the Morris water maze before (basal) and during the treatment genotype effect (G). Latencies (s) evolution before (A) and after (B) the treatment, evolution of learning and memory, trial by trial for each of the 5 days. Repeated measures ANOVA, 2×2×5, day effect (D), day × genotype effect (DxG), and interaction effect (D×G×T), *p < 0.05, **p < 0.01, and ***p < 0.001. Post hoc Tukey B test ap < 0.05 vs. different genotype but the same treatment and cp < 0.05 vs. different genotype and different treatment. t-Test for paired data, treatment effect: trial 1 of day 1 with treatment vs. trial 4 of day 5 without treatment, &p < 0.05. (C) Latency (s), evolution of learning and memory day by day. Repeated measures ANOVA 2×2×5, ns and post- hoc Tukey B test, ns (D) total number of entries in the quadrant: platform (P), right (Ar), opposite (O), and left (Al). ANOVA 2×2, genotype × treatment interaction effect (G×T), ***p < 0.001. t-Test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs no treatment, &&p < 0.01 and &p < 0.05. ANOVA and post hoc Tukey B test interaction of the three levels #p < 0.001. number of entries in the “trained quadrant” with respect to the (s) mice, in the marble number variables “intact” (t = 5.348, gl other quadrants [F(3,47) = 18.91, P < 0.001]. In contrast, in the 11, P < 0.001) and “buried” (t = -3.824, gl 11, P < 0.001) because other treated groups, an increase in entries was observed in both after the treatment, there was a decrease in the number of intact the quadrant of the platform and the right quadrant [NTg (r), marbles and an increase in the number of marbles buried. In F(3,43) = 31.18, P < 0.001; 3xTg-AD (s), F(3,47) = 18.91, P < the 3xTg-AD (s) mice, an increase in the number of “changed 0.001; 3xTg-AD (r) [F(3,43) = 15.27, P < 0.001]. Thus, the variable position” marble was observed (t = -2.601, gl 11, P < 0.05) and “number of entries in the right quadrant” indicated “genotype × a decrease in “buried” marble (t = 3.348, gl 11, P < 0.01). In the treatment” interaction effect [F(1,43) = 19.401, P < 0.001]. As groups of animals treated with risperidone NTg (r) and 3xTg-AD compared to their respective performances before treatment, (r), no changes were observed. Treatment effects were shown in statistical differences were only found in the NTg (r) mice, as a the “number of intact marble” [F(1,43) = 7.802, P < 0.01] as an decrease in the number of entries in the opposite (t = 3.203, gl 10, increase in its number in the groups treated with risperidone P < 0.01) and left (t = 3.909, gl 10, P < 0.01) quadrant. compared to those treated with saline. Conversely, the “number of buried marbles” also indicated treatment effect [F(1,43) = 8.573, P < 0.01] with a reduction as compared to those treated Marble Burying Test (MB) with saline, although this decrease is only significant in the case As illustrated in Figure 7, before treatment, 3xTg-AD mice of NTg (r) animals. Furthermore, in this variable, the interaction showed an increased pattern of interaction with marbles, with effect ”genotype × treatment” was observed [F(1,43) = 6.114, P < a higher number of marbles that changed position and were 0.05] due to a decrease in the number of marbles buried in the buried than NTg counterparts. After treatment NTg (s) mice NTg (r) mice and an increase in the 3xTg-AD (r) mice. the “intact” marbles were the least frequent compared to the marbles that changed position and were buried [F(2,35) = 19.51, P < 0.001]. Oppositely, in NTg (r) mice, an increase in the Baseline State of Blood Glucose number of marbles “intact” and a decrease in “buried” marbles Basal state of glucose levels before treatment did not differ [F(2,32) = 13.05, P < 0.001] was observed. In the two groups of between 3xTg-AD and NTg mice (Figure 8). However, after treated 3xTg-AD mice, no significant changes were observed treatment, NTg (r) mice showed a marked decrease (t = 7.611, between the three levels of interaction with marbles. In each gl 11, P < 0.001). This led to a ”genotype × treatment” effect group, before and after comparisons indicated differences: NTg [F(1,43) = 9.441, P < 0.01]. Frontiers in Pharmacology | www.frontiersin.org 11 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 7 | Effects of chronic risperidone in the marble burying test before (basal) and during the treatment. Number of marbles. t-Test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs no treatment, &&&p < 0.001, &&p < 0.01, and &p < 0.05. ANOVA and post hoc Tukey B test interaction of the three levels #p < 0.001. ANOVA 2×2, (T), treatment effect and (G×T), genotype × treatment interaction, ***p < 0.001, **p < 0.01, and *p < 0.05. Post hoc Tukey B test bp < 0.05 vs. different treatment but the same genotype and cp < 0.05 vs. different genotype and different treatment. FIGURE 8 | Effects of chronic risperidone in basal state of glucose before and during the treatment. t-Test for paired data, treatment effect: treatment with a dose of 0.1 mg/kg vs no treatment, &&&p < 0.001. ANOVA 2×2, interaction effect genotype × treatment (G×T), **p < 0.01. Post hoc Tukey B test bp < 0.05 vs. different treatment but the same genotype. Weight was lost in week 6, just before treatment started (NTg mice, t = The body weight (BW) of animals was monitored throughout 4.035, gl 23, P < 0.01; 3xTg-AD mice, t = 6.234, gl 27, P < 0.001). the experimental procedures (Figure 9). At the beginning During the first week of treatment (risperidone, 0.05 mg/kg), all of the experiments, weight of 3xTg-AD mice was higher than the groups lost weight [NTg (r) mice, t = 2.965, gl 10, P < 0.05; that of controls (t = -2.213, gl 50, P < 0.05), but differences were 3xTg-AD (s), t = 2.718, gl 11, P < 0.05; 3xTg-AD (r), t = 4.633, attenuated through the experiments. Statistical significance gl 10, P < 0.01] except NTg mice treated with saline. At 10 weeks Frontiers in Pharmacology | www.frontiersin.org 12 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 9 | Effects of chronic risperidone in the body weights of the animals before (basal), during the treatment, and immediately after death Phase 1 Basal. Student’s t-test; *p < 0.05. Phase 2. Start of treatment with risperidone at a dose of 0.05 mg/kg. Weights. t-Test for paired data, treatment effect: treatment with a dose of 0. 05 mg/kg vs untreated, &&p < 0.01 &p < 0.05 for Phase 3. Treatment with risperidone at a dose of 0.1 mg/kg. t-Test for paired data, treatment effect: treatment with a dose of 0. 1 ml/kg vs untreated &&p < 0.01 &p < 0.05 Phase 4. Immediately after death vs. the weights at the beginning of the longitudinal study. t-Test for paired data &&p < 0.01, and &p < 0.05. FIGURE 10 | Effects of chronic risperidone in the total mean survival. Mean ± SEM. Survival curves based on the age of the animals (in days). Each line represents the cumulative survival of the animals of each experimental group [NTg (s) and 3xTg-AD (s)] Log-rank test p < 0.001. Survival curve based on the age of the animals (in days). Each line represents the cumulative survival of the animals of each experimental group [NTg (s), NTg (r), 3xTg-AD (s) and 3xTg-AD (r)]. Log-rank-test ###p < 0.001. Frontiers in Pharmacology | www.frontiersin.org 13 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics months in 3xTg-AD but in 8 months in NTg mice). Survival curves showed that until 14 months of age, all groups exhibited 100% survival. Thereafter, only NTg (s) mice maintained its survival intact until 19 months, that is, 5 months longer than the other groups. When comparing the four groups, their mean survival at 15 months of age showed a genotype [F(1,45) = 4.968, P < 0.05] and “genotype × treatment” [F(1,45) = 4.968, P < 0.05] interaction effects. Thus, although onset of mortality window was 15 months for NTg (r), 3xTg-AD (s), and 3xTg-AD (r), the different slopes in their survival curves indicated different severity levels: NTg (r), survival of 91.61%, 3xTg-AD (s), survival of 85.71%, 3xTg-AD (r), 78.48% (Figure 11). Average life expectancy (50% survival) differed among the four groups [F(3,28) = 262.25, P < 0.001]. In NTg mice, the average life expectancy was 24 months, whereas in the two groups of 3xTg-AD mice, it was 18 months. Likewise, when 25% of the survival was analyzed, statistically significant differences with respect to the group of NTg (s) mice [F(3,12) = 137.41, P < 0.001] were found. The Kaplan-Meier test showed significant differences between 3xTg-AD (s) and NTg (s) (Log rank = 7.218, gl 1, P < 0.01) and also among the four experimental groups (Log rank = 22.833, gl 1, P < 0.001). Table 1 summarizes the meaningful correlation analysis of lifespan with the behavioral phenotype before and after treatment with risperidone for each one of the four experimental groups, while Figure 12 illustrates the correlations with the whole sample of 46 animals. As shown, in most tests, freezing behavior and activity to complete the task were strongly (P < 0.01) correlated with lifespan in NTg (s). In the NTg (r), it was mostly related to the elicitation of emotionality (grooming), vertical exploratory activity (rearing), and agonistic behaviors (body face, inversely with vibrating tail). Poorest behavioral correlates were found in 3xTg-AD mice, where survival FIGURE 11 | Effects of chronic risperidone in the total mean survival. Mean ± was inversely correlated to a poor marble burying, scarce SEM. ANOVA and post hoc Tukey B test ***p<0.001, ap < 0.05 vs. same and slow exploratory activity, and presence of vibrating treatment but different genotype, bp < 0.05 vs. different treatment but the tail. Survival in 3xTg-AD (r) was negatively correlated with same genotype and cp < 0.05 vs. different treatment different genotype. delay in the elicitation of vertical exploratory activity and number of groomings. Correlation with cognitive task was only shown in NTg(s), while the tests and/or variables with of treatment (risperidone, 0.1 mg/kg), the loss of weight was predictive value in the whole sample were those related to the only observable in the 3xTg-AD mice [3xTg-AD (s), t = 2.312, BPSD-like phenotype. gl 11, P < 0.05; 3xTg-AD (r), t = 3.478, gl 11, P < 0.01]. The four experimental groups showed reduced “weight immediately after death” as compared to before treatment [NTg (s), t = 2.209, gl DISCUSSION 11, P < 0.05; NTg (r), t = 5.382, gl 10, P < 0.0 01; 3xTg-AD (s), t = 5.721, gl 11, P < 0.001; 3xTg-AD (r), t = 3.937, gl 10, P < The main objective of the present work was to model in 0.01]. When the four groups of animals were compared with each 3xTg-AD mice the increased mortality risk induced by the other, no differences were found among them at either of both chronic administration of the atypical antipsychotic risperidone time points. shown in patients with AD. This is the first study that considers this objective despite the imperative need to have an animal modeling this vulnerability, an issue that cannot be addressed Survival Curve in clinical studies. Figure 10 illustrates the survival curves and lifespan, which Experimental design. First, the battery of behavioral tests differed among the groups: NTg (s): 1.078 days (36 months), NTg allowed us to confirm the AD phenotype of the animals (Giménez- (r): 874 days (28 months), 3xTg-AD (s): 768 days (24 months), Llort et al., 2007). Thus, we verified the existence of cognitive 3xTg-AD (r): 660 days (22 months), with a genotype effect in and BPSD-like behaviors in this initial sample of middle-aged mean lifespan. Lifespan was reduced by AD genotype [F(1,48) = 3xTg-AD mice, an age that has extensively been described 24.812, P < 0.001] but also by risperidone treatment (shortened 2 to mimic advanced neuropathological stages of the disease Frontiers in Pharmacology | www.frontiersin.org 14 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics FIGURE 12 | Behavioral correlates for lifespan considering all the sample of animals. Pearson’s correlations between behavioral variables and lifespan: *p < 0.05, **p < 0.01. Frontiers in Pharmacology | www.frontiersin.org 15 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics TABLE 1 | Behavioral correlates with lifespan in male NTg and 3xTg-AD mice chronically treated with saline or risperidone. Behavioral test and variable Lifespan (days) NTg(s) NTg(r) 3xTg-AD(s) 3xTg-AD(r) Before treatment. Corner Test Vertical activity (latency, s) 0.684* n.s. n.s. n.s. Vertical activity (number) -0.614* n.s. n.s. n.s. Open field test Initial movement (latency of freezing, s) 0.715** n.s. n.s. n.s. Exist of the center (latency, s) 0.63** n.s. n.s. n.s. Vertical activity (latency, s) n.s. n.s. n.s. -0.606* Self-grooming (latency, s) n.s. n.s. n.s. n.s. Vertical activity (3 min) n.s. n.s. n.s. n.s. Vertical activity (5 min) n.s. n.s. n.s. n.s. Self-grooming (number) n.s. n.s. -0.657* T-maze test Initial movement (latency of freezing, s) 0.721** n.s. n.s. n.s. Complete the test (total time, s) 0.751** n.s. n.s. n.s. Social Interaction Test Non-social interactions n.s. Self-grooming (latency, s) n.s. -0.716** n.s. n.s. Marble interaction test Intact (number of marbles) n.s. n.s. -0.707** n.s. Morris water maze Day 4 (mean latency, s) 0.684* n.s. n.s. n.s. After treatment. Corner Test Vertical activity (latency, s) n.s. n.s. 0.633** n.s. Vertical activity (number) n.s. 0.594* -0.734** n.s. Open field test Initial movement (latency of freezing, s) 0.724** n.s. n.s. n.s. Exist of the center (latency, s) 0.734** n.s. n.s. n.s. Entrance to the periphery (latency, s) 0.713** n.s. n.s. n.s. T-maze test Initial movement (latency of freezing, s) n.s. n.s. -0.687** n.s. Complete the test (total time, s) n.s. -0.653* n.s. n.s. Social Interaction Test Social interactions Body/face (latency, s) n.s. -0.599* n.s. n.s. Vibrating tail (latency, s) n.s. 0.611* n.s. n.s. Vibrating tail (total time) n.s. -0.611* n.s. n.s. Vibrating tail (total no. of episodes) n.s. 0.611* n.s. n.s. Pearson’s correlations, *P < 0.05, **P < 0.01. n.s., non significant. (Belfiore et al., 2019). Here it is important to note that although on age-matched NTg mice but also as compared to their own the cognitive deficits in the MWM were confirmed, the inclusion phenotype before treatment. Risperidone was used at a dose of criteria in the “before–after treatment” analysis excluded the 0.1 mg/kg equivalent to those administered in patients with AD animals that died during the period of behavioral assessments. and used in most experimental work performed in rodents. As in In AD, treatments with atypical antipsychotics are effective in the case of geriatric patients, the treatment regimen was initiated controlling anger, agitation, aggression, and symptoms more with a lower dose of 0.05 mg/kg. typical of the psychotic spectrum such as hallucinations, paranoia, In the longitudinal study, the baseline characterization was and delusions, while cognitive symptoms, quality of life, and performed at 12 months of age and the administration of the attention do not improve with these treatments (Kalman et al., treatment started at 13 months, which in both cases corresponds 2008). In view of the pharmacological action of the antipsychotic to advanced neuropathological stages of the disease. The number treatment, it is therefore interesting to note that the animals of animals used (n = 12–14 for each group) was adequate for before treatment had cognitive deficits but also an increase in the behavioral studies according to the guidelines on the use of stereotyped behavior of marble burying (usually used to assess genetically mutated animals and taking into account that at the efficacy of antipsychotics) and alterations in social behavior the middle age, the individual variability increases due to age. (also modifiable with antipsychotics). Once the phenotype was However, for the study of long-term survival curves, it would verified, we evaluated the response of 3xTg-AD mice to chronic have been advisable to use a much greater number, between 40 treatment with risperidone as compared to the effects exerted and 50 animals. This and other works of the literature are limited Frontiers in Pharmacology | www.frontiersin.org 16 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics experimentally to use a reduced number of animals by the Risperidone enhanced the behavioral differences but did so difficulty of obtaining the experimental subjects. differently depending on the genotype. Although the differences Corner test. In the modeling of the behavioral effects between the two groups of NTg animals before the use of drugs induced by the atypical antipsychotic risperidone, the behavior may be potentiating this contrast of NTg (r) mice, we have of neophobia was evaluated by the CT. At the beginning of verified that this group, but not the one that would be treated the treatment, the 3xTg-AD mice showed a greater behavioral with saline, is the one that shows the most standard values inhibition when confronting novelty, presenting higher levels of described for the NTg male animals of our colonies. Therefore, neophobia with respect to the NTg mice, in the three variables of in any case, the results of NTg mice treated with saline are the test. Although this response was independent of treatment, downward biased. it was observed that in NTg (r) mice, the dose of 0.05 mg/kg About the effect of risperidone, studies with other mouse of risperidone induced a significant increase in vertical activity models for other pathologies, such as the ICR mouse, have found with respect to the other three experimental groups. In the that a low dose of risperidone (0.01 mg/kg, i.p.) decreases the following week (week 8, dose of 0.1 mg/kg), an attenuation of the freezing or freezing behavior, and at high doses (0.04 mg/kg, i.p.), differences between genotypes was observed, now only detectable spontaneous motor activity is reduced. However, when in the rearing variable. In fact, in general, during the rest of co-administered with other drugs, it loses effectiveness in its the weeks, vertical behavior was the most sensitive to indicate anxiolytic effect (Miyamoto et al., 2004). In male ddY mice the effects of the factors studied. Repeatedly, the predominant of 5 weeks of age, risperidone at a dose of 0.1 mg/kg inhibits factor was genotype, corroborating the neophobia described in methamphetamine-induced hyperlocomotion and at several our laboratory in 3xTg-AD mice (Giménez-Llort et al., 2006; other doses attenuates that induced by MK-801 through Giménez-Llort et al., 2007; Giménez-Llort et al., 2008 and blockade of the 5-HT2A/2C receptor (Uchida et al., 2009). In Giménez-Llort et al., 2010), and in a few cases, treatment effect BTBR mice at doses of 0.125 and 0.25 mg/kg i.p., risperidone was observed. Therefore, chronic treatment with risperidone did decreases exploratory activity (total distance traveled) in the not modify the neophobic response of the 3xTg-AD mice, and if OF (Silverman et al., 2010). The data confirm the dopaminergic any effect had, it was sporadically in NTg animals. action of the drug modulating motor and motivational functions Open-field test. In agreement with the increased neophobia (Rinaldi et al., 2007) and exerting anxiolytic properties. in the CT, in the OF to assess exploratory activity and anxiety, On the other hand, the literature also suggests that atypical AD genotype differences were also observed with reduction of antipsychotics such as risperidone could be useful in attenuating vertical and horizontal activity, temporal delay of the grooming stereotyped behaviors and not just locomotor activity. In the behavior, and increased presence of urine. This predominant open field, grooming behavior also showed genotype and effect of the genotype factor corroborates the results obtained genotype interaction effects by treatment. Grooming is an in the open field in the pretreatment phase and the CTs. Both activity of daily life linked to hygiene of animals but is also the corner and open-field tests converged to show the emotional part of copying with stress strategies, and can be developed as and anxious-like profile of 3xTg-AD mice, in agreement with our a stereotyped behavior that gives and looks easily modified previous studies describing the appearance of these behaviors in anxious situations (Kalueff and Tuohimaa, 2004a, Kalueff (Giménez-Llort et al., 2007; Giménez-Llort et al., 2008; Giménez- and Tuohimaa, 2004b). Furthermore, the grooming is used Llort et al., 2010; Giménez-Llort et al., 2018; García-Mesa et al., experimentally as a tool to measure behavioral stimulation of 2011; García-Mesa et al., 2012) since the early stages to advanced D1 receptors (Rimondini et al., 1998) so that risperidone, as stages of the disease (Giménez-Llort et al., 2006). regards its antagonist action D1, could induce a decrease in this The activity variables, both horizontal and vertical, also behavior. In this respect, in the 3xTg-AD mice, it was observed showed treatment effect per se and their effects were influenced that the grooming behavior appeared later in time with respect by the interaction with the genotype. Thus, most of the effects to its NTg control, thus corroborating the genotypic differences of risperidone were observed again in NTg (r) animals as an in the anxious-like profile of these animals described in our increase with respect to the other three experimental groups. laboratory (Giménez-Llort et al., 2006). In these previous studies This seems to be a pattern similar to that observed in the CT, but with old male and female 3xTg-AD mice, the grooming behavior in both horizontal and vertical vectors of locomotor activity, with appears later than in the NTg mice and, in turn, aging reduces the NTg (r) group being different from all the others. the duration (Giménez-Llort et al., 2008). Regarding the effect of When comparing the data at the longitudinal level, it can be the treatment, the effects of interaction with the genotype are due seen that in the treatment phase, there is an attenuation of all the to a more frequent grooming behavior in the NTg (r) group of responses that could be due to the simple fact of the repetition of mice, although this selective difference replicates the existing one the test. Also, it could be attributed to the repeated manipulation or before treatment. In fact, work on strains of standard mice such handling of animals. In both cases, these are some of the limitations as Swiss albino mice indicates that risperidone, at doses of 0.1 inherent to the use of behavioral batteries and are enhanced in and 0.2 mg/kg, decreases the number of grooming (De Oliveira longitudinal studies and chronic treatments. These limitations are et al., 2008). Similarly, in the BTBR mouse model for autism, they difficult to avoid but can be mitigated by reducing the number of have also observed that at doses of 0.125 and 0.25 mg/kg i.p. they tests and the use of tests with variables with convergent validity, as reduce grooming by 40% to 50% with respect to their control, we have done in our study, and the selection of guidelines for oral although the reduction in behavior can be confused with levels administration or subcutaneous pumps. of sedation (Silverman et al., 2010). Frontiers in Pharmacology | www.frontiersin.org 17 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics Social interaction test. The SIT was found the most complete with anticholinesterase galantamine (0.3 mg/kg) and with 0.1 mg/ and sensitive tests to evaluate the deficits of social behavior kg of risperidone (Wang et al., 2007). Risperidone also corrects the among 3xTg-AD mice in comparison with those observed aggression induced by social isolation in male ddY mice (Uchida among NTg mice of the same age and sex (Torres-Lista and et al., 2009) and the attack behavior in male albino mice, although Giménez-Llort, 2019). Here, in the basal characterization before it is not exempt from side effects at the level of motor behavior treatment, we replicated the results but, this time, evaluated (Rodríguez-Arias et al., 1998). In our animal model, we observed within a behavioral battery. The 3xTg-AD mice showed results spontaneous increase aggressiveness in 3xTg-AD mice1. However, similar to those of the NTg animals in the most common social in this study, the presence of this behavior in mice was very low or interaction behavior such as body-face contact. However, the nonexistent, since it is spontaneous aggressiveness and not induced/ genotype effects indicated less anogenital contact, greater vibrant enhanced by isolation (animals start from a daily social condition). tail, and absence of aggressiveness in groups of 3xTg-AD mice Even so, the aggressiveness observed in the NTg (s) mice is corrected with respect to what was observed among NTg mice. Besides, in the group treated with risperidone NTg (r). In the case of the effect of chronic treatment with risperidone was evaluated 3xTg-AD mice, it is likely that1 the behavior of “vibrant tail” acts as in these behaviors since atypical antipsychotics are often chosen a dissipating mechanism that temporarily attenuates the response to treat BPSD symptoms, such as psychosis, aggression, and of directed aggressiveness, with no “aggressiveness” observed. agitation. Risperidone exerted effects decreasing the anogenital T-maze. The spontaneous alternation in the TM is a paradigm behavior and in the case of the 3xTg-AD mice, it significantly mostly used to evaluate working memory, which also includes corrected the preponderance of the vibrant behavior recently aspects such as exploratory activity and emotionality/anxiety. It described as a characteristic of the social interaction pattern of is a test based on the possibility that the rodent chooses one of female 3xTg-AD mice (Torres-Lista and Giménez-Llort, 2019). the arms arranged in a T-shape. Many brain areas, such as the The effects of risperidone interacted with genotype in several hippocampus, septum, prefrontal cortex, and the basal forebrain, variables of social behaviors studied, in such a way that the as well as several neurotransmitters, such as dopamine and effects were more intense in the NTg genotype where risperidone norepinephrine, are involved in the implicit working memory decreased social behaviors, bringing them to levels equivalent to involved in the performance of the test (Zhang et al., 2004; those 3xTg-AD groups. Deacon and Rawlins, 2006). Variables such as latency to cross the Among the non-social behaviors that appear interspersed intersection of the maze allow, in addition, to assess the copying during the development of social interactions, the effects of with stress strategies of animals, this being a variable that correlates the genotype were observed as a decrease in the horizontal with a worse neuroimmunoendocrine function, indicators of and vertical exploratory activity in the groups of 3xTg-AD accelerated aging in mice, and premature death (Guayerbas mice with respect to the NTg groups. In fact, the behavior et al., 2001). In our study, no differences were observed in the of vibrant tail is associated with an immobility of the animal, number of execution errors, but the efficiency to complete the so that the highest levels of this behavior in the 3xTg-AD mice different phases of the test was diminished in 3xTg-AD mice and justify that the exploratory activity, which usually is already lower slightly affected by risperidone. Thus, in phase 2 of the study, the in these animals, be significantly reduced. genotype effects were observed in the emotionality as an increase The effects of risperidone were observed only in the variable in the presence of urine with respect to the NTg groups. Later, in number of burials, where the drug decreased this behavior, phase 3, the variable latency to intersection and time to complete considered mimicking psychotic type stereotypies or obsessive- the maze denoted the deficiencies of the 3xTg-AD. compulsive type anxiety. The improvement in this variable The effects of the treatment were only observed in phase in both genotypes and the improvement also of the behavior 2, in which the groups treated with risperidone at the dose of vibrant tail in the 3xTg-AD mice suggest that risperidone of 0.05 mg/kg needed more time to complete the exploration mainly exerted an antipsychotic/anxiolytic management effect, of the maze with respect to the groups that received saline. modifying the features of the anxious trait and/or anxious states, In addition, in the 3xTg-AD (r) mice, the time to complete in both genotypes, respectively. the maze was also higher compared to that needed before While the effects of atypical antipsychotics in exploratory treatment. The effects of the interaction between genotype and activity and emotionality are discussed by the scientific community treatment were observed in phase 3 at the emotional level in as mainly due to their dopaminergic actions, it is considered that the both bowel movements and urine. treatment with risperidone could modify the pattern of the social In our previous studies, we have observed that the latency behavior thanks to its profile as a 5-HT2 antagonist, since these to cross the intersection is a variable that reflects changes and serotonergic receptors have been implicated in social interaction subsequent deficiencies in the stress management strategies of behavior (File and Seth, 2003). In a comparative study in APP/ animals and is related to the deterioration of the homeostasis London mice, treatment with risperidone decreased its aggressive of the neuroimmunoendocrine system, due to the age itself behavior and did so consistently to that exerted by 8-OH-DPAT and, above all, in the presence of transgenes (Giménez-Llort and buspirone, two serotonergic agonists (Moechars et al., 1998). et al., 2010; Giménez-Llort et al., 2012). In 6-month-old female In other models for AD, such as the APP23 mice, treatment with 3xTg-AD mice, these changes are reflected as a greater speed to risperidone also attenuates the aggressive behavior of animals (Vloeberghs et al., 2008) whereas in psychosis models, the 1Ramírez-Boix, P., and Giménez-Llort, L. (submitted). Modeling BPSD in rodents: deterioration of social interaction induced by PCP can be improved relevance of spontaneous aggressiveness and chronic mild social stress. Frontiers in Pharmacology | www.frontiersin.org 18 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics reach the intersection of the TM with respect to the controls in per se but rather changes in search strategies. Thus, as in the a typical flight behavior, while at later ages, the strategy fight- case of working memory, risperidone did not affect the results or-flight chosen to combat the acute stressful situation is the decisively since the animals continued to distinguish what had petrifaction (Giménez-Llort et al., 2010, Giménez-Llort et al., been the position of the platform. It should also be noted that the 2012). These observations agree with those of the reference parity in the results of this test is surely conditioned by the fact laboratory that described this functional relationship in a of the previous knowledge of the paradigms since the genotype longitudinal study with female OF-1 Swiss mice. Thus, De la differences that were observed in the first experience in the water Fuente’s laboratory showed that the animals that spend more maze no longer exist among the saline groups. The latencies to time in the TM show high levels of emotionality/anxiety and reach the platform in the first trials were very low, equivalent have a less competent immune system with respect to those that to the third day of testing before treatment, supporting this explore more quickly. In addition, the animals that showed lower interpretation. Here, it is important to remind that the analysis performance in the TM showed a reduction in survival compared is based on a “before–after design” on censured data, that is, to those that obtained better results (Guayerbas et al., 2001). excluding the animals that died during phases 1, 2, and 3. Since The decrease of spontaneous alternation in the TM has also the performance of the initial sample of animals was in agreement been observed in other transgenic models of AD such as Tg2576 with the cognitive deficits described in this animal model at 12 (Lalonde et al., 2003) and APP/PS1 (Tempier et al., 2013) months of age, the present results in the before–after analysis mice, in animal models for schizophrenia as STOP-null mice suggest a mortality bias, with exclusion of animals with worse (Delotterie et al., 2010), or pharmacological models by selective life prognostic as determinant to the deficits in the overall group blockade of dopamine D1 and D2 receptors in the pre-limbic performance. Also, this suggests that the effects of risperidone region of the prefrontal cortex (Rinaldi et al., 2007). The effects illustrated here were those exerted in the “less worse” animals, as of antipsychotics differ depending on the model, the drug, and those that died during the behavioral assessments were excluded. the duration of treatment. Thus, in APP/PS1 mice, the chronic In some cases, in this mouse model, it has also been observed treatment with Quetiapine for 7 and/or 10 months normalized that the treatments, according to their pharmacological the anxiety-like behavior observed in the maze, minimized actions, selectively affect different aspects of the learning and memory deterioration, and decreased Aβ plaques in the brain. memory process. This is the case of the study with 10-month- In the STOP-null mice, treatment with risperidone only induced old 3xTg-AD mice in both sexes, where it was found that the a trend to reduce the spontaneous alternation in the Y-maze chronic treatment of 5 months with paroxetine improved the (Delotterie et al., 2010). Other work in the radial maze has shown deficit of the space navigation in both males and females, without beneficial effects of olanzapine and clozapine on the memory affecting the speed of swimming or the distance traveled, which impairment of male BALB/c mice, although they were not suggests a conservation of cognitive functions (Nelson et al., exempt from some extrapyramidal effects (Mutlu et al., 2012). 2007). Likewise, studies with 12-month-old female 3xTg-AD Together with the CT and the open field, the results in the TM mice treated with melatonin found that the treatment improved provided new data to confirm the emotional/anxious-like profile learning retention from platform position (García-Mesa et al., of the 3xTg-AD mice and their worse capacity to cope with 2012). In APP/PS1 mice treated with quetiapine, the continuous stress; however, it did not show problems in working memory administration of 4 to 7 months of the antipsychotic decreased the as measured by spontaneous alternation. The chronic treatment number of plaques of Aβ in the cortex and in the hippocampus with risperidone, in any case, worsened this AD-profile and of the animals and reduced memory loss, also attenuating the the exploratory efficiency of NTg animals, probably due to the anxiety-like behavior (He et al., 2009). On the other hand, it has extrapyramidal effects. Still, being an atypical antipsychotic and been described that olanzapine does not affect the processes of administered in a low dose regime, it is considered that these side acquisition, consolidation, or recovery in the MWM test (Hou effects are much lower than in other classic antipsychotics. et al., 2006). However, the same work shows that clozapine Morris water maze. To evaluate short- and long-term spatial and haloperidol appeared to affect the acquisition process and reference learning and memory, the MWM was used (Morris, consolidation and induced a deterioration in spatial learning 1981; Morris, 1984). All groups showed the same acquisition (Hou et al., 2006). curve in either the “trial-by-trial” or the “day-by-day” analysis In fact, the effects of antipsychotics on learning and memory for short-term and long-term learning and memory, respectively. in processes that occur properly with psychosis are controversial, Likewise, in the probe trial with removal of the platform, all the since there are studies in rats that have indicated that the groups distinguished the quadrant of the platform with respect classic antipsychotic haloperidol and the atypical antipsychotic to the other quadrants, although the risperidone treatment risperidone, at certain doses, affect cognitive processes (Didriksen decreased the selective search of the platform incorporating one et al., 2007), while clozapine and sertindole were effective in the of the adjacent quadrants in the preference. The preference of treatment of psychosis without producing detrimental effects this adjacent area indicates less focused goal-directed swimming on cognition (Mutlu et al., 2012). It is more than probable that strategies (Baeta-Corral and Giménez-Llort, 2015). The these discrepancies in the pharmacological actions are due to interaction between genotype and treatment in the preferences the differences they present in the profile of pharmacological of the two adjacent quadrants indicated that the effects of selectivity by different neurotransmission systems such as risperidone affect the two genotypes differently, although in dopamine and serotonin. In this respect, risperidone is classified general these and the other data do not indicate deficiencies as a “qualitatively atypical” antipsychotic agent with a relatively low Frontiers in Pharmacology | www.frontiersin.org 19 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics incidence of extrapyramidal effects when given at low doses that abnormalities, and development of type 2 diabetes. Given these have a serotonergic antagonist action higher than dopaminergic. serious health risks, FDA requested that antipsychotics of second Marble burying test. The burial test for marbles, which is generation such as clozapine, olanzapine, risperidone, quetiapine, used for the detection of new antidepressants, anxiolytics, and ziprasidone, and aripiprazole should be labeled to indicate that antipsychotics (Njung’e and Handley, 1991; Bruins Slot et al., 2005 they increase the risk of developing diabetes (Cope et al., 2009 and Kaurav et al., 2012), was found sensitive to detect alterations and Savoy et al., 2010). In the case of increased vulnerability to in the 3xTg-AD mice (Torres-Lista et al., 2015). The response higher mortality risk in AD induced by atypical antipsychotics, patterns were clearly different, since the level of interaction of the metabolic syndrome has been hypothesized initially as one of the NTg mice with the marbles results in half of the objects being the possible risk factors. Antipsychotics bind with high affinity changed of position, while the other two quarters remained to a wide variety of neurotransmitter and transporter receptors intact or have been buried. In the 3xTg-AD mice, the burying that could be involved in various metabolic effects; however, the behavior is enhanced so that more than half of the objects appear underlying mechanisms are not entirely clear. buried at the end of the test, a quarter changed position, and only In the basal study, we evaluated baseline fasting glucose levels the remaining 10% remained intact. as well as ip glucose tolerance test (IPGTT) and we repeated the The marble burying behavior of marbles was studied again to measurement of basal glucose levels, this time without fasting, evaluate the effect of the antipsychotic. However, it is interesting after chronic treatment with risperidone. The glucose tolerance to observe previously, in the results of the groups treated with test was ignored because it also involves fasting in animals added saline, that the behavioral pattern described above for one and to the normal fast in the light phase and the consequent loss of another genotype has been significantly modified by the protocol weight that, at these ages, could affect the survival curves of the of chronic manipulation. In the NTg mice, the interaction with animals. the objects increased, enhancing the burying of the marbles, As in the basal characterization phase, no genotype differences which are now quantitatively equivalent to the marbles that were observed during the treatment phase but genotype × changed position. In contrast, in 3xTg-AD mice, chronic saline treatment interaction effects. These effects resulted, however, treatment reduced the interaction with the objects, significantly from a significant decrease in glucose levels in the NTg (r) mice decreasing the number of marbles buried in favor of the marbles compared to those in the NTg (s) animals or when the glycemic that changed position. Taken together, these effects reduce values of the NTg (r) were compared longitudinally with those genotype differences. obtained before treatment. The results also agree with studies When all the factors are evaluated together, the observation of carried out in our laboratory with 3xTg-AD animals at 6 months the absence of genotype effect is ratified. Risperidone increased the of age in which we observed, in both sexes, that the basal levels number of marbles intact and decreased the number of marbles of plasma glucose with and without fasting are normal, with the buried with respect to the groups administered with saline, so, homeostasis of glucose being the one that is compromised in in general, risperidone reduced the interaction with objects. 3xTg-AD mice, although with aging, the genotype differences These results corroborate those observed in the burying/digging are lost (Giménez-Llort et al., 2010). Studies in male C57BL/6 behavior during the SITs. Converging, differences were also found mice administered with risperidone showed a slight increase in between genotypes with greater burying behavior in the 3xTg-AD blood glucose levels, but only at a low dose of the drug (Dwyer mice, the genotype differences disappeared when studied during and Donohoe, 2003), while other more recent studies, with the treatment, and risperidone decreased burying behavior in males FVB/N mice, observed a significant reduction of plasma the two groups. Although the burying/digging behavior that is glucose (-30%) through the induction of insulin release (Savoy evaluated during the SIT is not a behavior with respect to any et al., 2010). object, the behavior is similar to that which is measured, in a more Body weight curves. The results on weight in patients with AD directed way, in the marble burying test. treated with antipsychotics are somewhat contradictory since there Investigations with other mouse models such as male ICR are studies that indicate that the administration of risperidone, treated with risperidone at a dose of 1 mg/kg, p.o. have observed olanzapine, or quetiapine at low doses is not associated with that risperidone treatment is effective in reducing the number weight gain (Rondanelli et al., 2006). However, other research of buried marbles, although it also reduces locomotor activity indicates that the use of olanzapine and quetiapine significantly (Matsushita et al., 2005). This effect is also shown in male BTBR increases weight in women and modifies cholesterol levels (Zheng mice (Gould et al., 2011) and in NIH Swiss male mice (Li et al., et al., 2009). At the experimental level, weight measurements are 2006). Likewise, in male NMRI mice, risperidone, at a dose of a good indicator of the health status of the animal. In addition, in 0.16–0.63 mg/kg, significantly reduces marble behavior and this study, we can evaluate the possible effect of chronic treatment locomotor activity, suggesting that the antagonist action of the with risperidone in this variable, in addition to the fact that weight drug 5-HT2A receptor may contribute to the effectiveness of the is required to administer the correct dose of treatment. burying of marbles and having anxiolytic effect (Bruins Slot et al., For these reasons, it was decided to make a longitudinal 2005). According to these results, it can be said that risperidone, follow-up of the weight from the beginning of the experimental at a low dose without cataleptic effect, can effectively reduce the process to the end of the mice’s life. According to our own data, number of buried marbles (Bardin et al., 2006). higher BW was found in the 3xTg-AD mice (Giménez-Llort Glucose levels. Antipsychotic drugs can cause a variety of et al., 2007), and as they advanced in age, these differences were metabolic problems such as weight gain, hyperglycemia, lipid attenuated. Risperidone produced a weight loss only in 3xTg-AD Frontiers in Pharmacology | www.frontiersin.org 20 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics mice, while groups of NTg mice maintained their BW. In the final The total life expectancy, the mean life expectancy, and the stages of the animal’s life, an acute weight loss was observed and life expectancy with a lifespan of 25% showed the predominant this, immediately after the death, was found to be diminished in effects of the genotype, but also the effect of risperidone in the the four experimental groups, although the process of cachexia range of ages below. Therefore, although the survival curves in the and the dehydration intrinsic to the death process make the data mice treated with risperidone and saline showed similar slopes in to lose accuracy. Previous studies with 6-month-old 3xTg-AD the four experimental groups, the initial vulnerability and from mice (males and females) have observed a decrease in the the mean life expectancy determined important changes in the weight curve only in the group of females in potentially stressful maximum survival recorded. In NTg mice, this maximum was situations (Giménez-Llort et al., 2010). This could be because reduced by 8 months with the risperidone treatment (from 36 they are more sensitive to repeated handling or the exposure to 28 months), and in groups of 3xTg-AD mice, the maximum of behavioral tests; however, it cannot be stated to what extent survival was reduced by 2 months (from 24 to 22 months). The data weight loss can be associated with emotional reactivity. suggest, in addition, that risperidone exerted deleterious effects in Several studies suggest that weight reduction may be NTg animals that equaled them in vulnerability to animals with associated with decreased activity but depending on the doses 3xTg-AD genotype. This deleterious effect of risperidone in NTg administered. Some authors have shown a relationship of the animals was mainly observable between 15 and 17 months of age, neuroleptic effects of olanzapine, ziprasidone, and risperidone although it reappeared in the age ranges corresponding to their with the decrease in food intake, the reduction of body mass, and half-life expectancy (24 and 25 months, respectively) and the motor activity since these changes are only present in wild-type difference persisted until the end of the curve. mice but not in D2R knockout mice (Yoon et al., 2010). However, The behavioral correlates for survival pointed out variables treatment with risperidone 4 mg/kg p.o. in female C57BL/6J mice and tests related to the BPSD-like phenotype, in agreement with of 12 weeks of age induces an increase in food consumption, the the neuroimmunoendocrine hypothesis (Giménez-Llort et al., corresponding weight gain, an increase in body temperature 2014) pointing at sex-specific neuroimmunoendocrine aging in during the light phase, and a reduction in the activity during the 3xTg-AD mice and its relation with longevity (Giménez-Llort dark phase (Cope et al., 2009). et al., 2008). The crosstalk refers to the oxi-inflamaging theory Life expectancy and survival. Treatment with atypical developed in PAM and NPAM mice where the divergence of antipsychotics may double the risk of mortality in elderly patients the speed of the aging process (premature in PAM and normal with dementia, although this risk is similar to or even lower than in NPAM) was correlated with immune function, oxidative antipsychotics. The increase in risk is faster in antipsychotic stress, and worse/better copying with stress strategies in the TM treatment and does not seem to be related either to the dose or to (Guayerbas et al., 2001). While neophobia and worse long-term the type of substance (Vilalta-Franch et al., 2008). Recent studies memory were found to be correlated with reduced survival in have indicated that patients with dementia and prescription female NTg and 3xTg-AD mice (Torres-Lista et al., 2017), it was antipsychotics such as thioridazine, chlorpromazine, haloperidol, surprising that memory variables did not correlate with survival in trifluoperazine, and risperidone to treat neuropsychiatric the present work, probably due to the mortality bias, as discussed symptoms increased the long term risk. The most pronounced above. Thus, this was probably because the impairment was less difference will be between 24 and 36 months after the start of pronounced in the current male sample, as correlation was only treatment (Ballard et al., 2009; Kales et al., 2012). In addition, found with the level of optimization of the performance (day 4) on other mental pathologies such as bipolar disorder in the in the water maze. Interestingly, the effects of risperidone on elderly patient, risperidone has a higher incidence of mortality social behaviors in the NTg mice were the ones to correlate with associated as compared to quetiapine (Bhalerao et al., 2012). survival, with faster body/face interaction and delayed elicitation Our longitudinal study replicates the increased mortality of vibrating tail being correlated with longer longevity of NTg described in male 3xTg-AD mice as compared to age-matched (r) animals, but also when the whole sample of animals was NTg mice by our laboratory (Giménez-Llort et al., 2008) and considered. This is important to note, since, to our knowledge, demonstrates that this animal can model the vulnerability to it is the first time that social behavior is being related to lifespan. increased mortality risk associated with risperidone in the human Risperidone is mainly metabolized in the liver (Mannens et al., patients. Thus, the group of NTg (s) mice maintained their survival 1993 and Zhou et al., 2006) by the enzyme cytochrome P-450 intact until 19 months of age, while the start of mortality was 2D6 (CYP2D6), which has more than 20 genetic polymorphisms during 15 months of age for the groups of NTg (r), 3xTg-AD (s), (Leysen et al., 1988; Mannens et al., 1993 and Matsubara et al., and 3xTg-AD (r). In previous studies with 15-month-old males 2007). Whether biological aspects related to the individual 3xTg-AD and NTg mice, the increase in mortality, evaluated from variability in drug metabolism either by polymorphisms or 6 to 15 months of age, was 40%. In the present study, at the same by hepatomegaly (Marchese et al., 2014) explain the increased age, the percentage was 15% since survival was recorded from 12 mortality risk remains to be elucidated. months and, therefore, mortalities that occurred earlier (i.e., from In summary, in this longitudinal study, the effects of the 6 months of age) were not counted (Giménez-Llort et al., 2008). chronic administration of risperidone on cognition and Increased mortality in 10–20-month-old male 3xTg-AD mice BPSD-like (motor, NPS, emotional, and social) symptoms as compared to B6129F2 wild-type controls has also been found were evaluated in male 3xTg-AD and NTg mice, as well as the by other laboratories and shown to be accompanied by elevated short- and long-term impact on their survival. It was observed frailty scores (Kane et al., 2018). that the factors “genotype,” “treatment,” and “genotype × Frontiers in Pharmacology | www.frontiersin.org 21 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics treatment” interaction effects were present in most of the ETHICS STATEMENT behaviors studied, such as decreased neophobia, and improved the development of exploratory activity in the open field in NTg The animal study was reviewed and approved by the protocol mice but had a weak effect on these variables in the 3xTg-AD CEEAH 2481/ DMAH 8700 entitled “Risk factors and preventive/ mice model. Risperidone was effective in reducing the therapeutical strategies in Alzheimer’s disease: studies in triple- number of marbles buried in the groups of mice that received transgenic 3xTg-AD mice” and was approved by Departament de antipsychotics with respect to the groups that were given Medi Ambient i Habitatge, Generalitat de Catalunya. saline. The study replicated, in 3xTg-AD mice, the increased mortality risk associated to risperidone observed at a clinical level in humans. Thus, despite the fact that risperidone allowed AUTHOR CONTRIBUTIONS to modify the alterations shown by the 3xTg-AD mice in their interaction with congeners (social behavior) or environment Development of the concept and study design: LG-L. Financial (burying behavior of marbles), it exerted negative effects resources: LG-L and SL-P. Treatment: LG-L and VT-L. Behavioral by reducing the vertical and horizontal exploratory activity studies and data collection: VT-L. Data analysis: VT-L. Data in the tests. What is more relevant, chronic treatment with interpretation: VT-L and LG-L. Scientific discussions: VT-L, SL-P, risperidone severely compromised the life expectancy of the LG-L. Drafting manuscript: VT-L and LG-L. Critical revision animals already since the beginning of treatment as shown by of manuscript: VT-L, SL-P, LG-L. Approving final version of early mortality windows. The impact of this mortality on the manuscript: VT-L, SL-P and LG-L. analysis of “before–after treatment” effects on the censured data is relevant to note since it further suggests that the reported effects are those exerted in the “less worse” animals, FUNDING that is, those that are able to survive during the period of behavioral assessment. In the same way, risperidone exerted The work received support from Instituto de Salud Carlos III, ISC3 especially severe effects in the old NTg animals, also modeling, PI10/00283, Spain; Research Agreement LG-L UAB and SL-P, at the experimental level, the window of early vulnerability UVaMiD, 2017-SGR-1468 and UAB2019-GE260408. VT-L received described for the adverse effects of cerebrovascular risk in a predoctoral grant Fundació La Marató de TV3 2010/062930. the elderly treated with atypical antipsychotics. The results support the awareness on the use of risperidone and the associated increased risk of mortality in AD, which suggests ACKNOWLEDGMENTS the relevance of the dosage (dose and treatment period) and that this treatment has to be used in short-term schedules to We thank Helga Rivas and Xenia Planas for their technical address the symptoms that cause morbidity and pain in the assistance. The animals used in the present study come from patient and to diminish the potential of self-hurt. The decision the colony of homozygous 3xTg-AD and wild-type NTg mice to use atypical antipsychotics should be based on the patient’s established by Dr. Lydia Giménez-Llort at the Universitat medical history and both the benefit and the risk of treatment. Autònoma de Barcelona, Spain, from progenitors kindly provided At the translational level, the 3xTg-AD mice model and their by Prof. Frank M. LaFerla, Department of Neurobiology and NTg counterparts can be useful to delimitate critical time Behavior, University of California Irvine, CA, USA. windows and for studying the physio-pathogenic factors and underlying causal events involved in this topic of considerable public health significance. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar.2019.01061/ DATA AVAILABILITY full#supplementary-material The raw data supporting the conclusions of this manuscript will TABLE S1 | Genotype effects in the behavioral phenotype before treatment. be made available by the authors, without undue reservation, to Statistics: Student t-test, 3xTg-AD vs. NTg (n=46), *p<.05, **p<.01, any qualified researcher. ***p<0.001. REFERENCES Baeta-Corral, R., and Giménez-Llort, L. (2014). Bizarre behaviors and risk assessment in 3xTg-AD mice at early stages of the disease. Behav. Brain Res. Andrew, M. K., and Rockwood, K. (2010). Social vulnerability predicts cognitive 258, 97–105. doi: 10.1016/j.bbr.2013.10.017 decline in a prospective cohort of older Canadians. Alzheimer’s and Dementia Ballard, C., Hanney, M. L., Theodoulou, M., Douglas, S., McShane, R., 6, 319–325.e1. doi: 10.1016/j.jalz.2009.11.001 Kossakowski, K., et al. (2009). The dementia antipsychotic withdrawal trial Baeta-Corral, R., and Giménez-Llort, L. (2015). Persistent hyperactivity and (DART-AD): Long-term follow-up of a randomised placebo-controlled trial. distinctive strategy features in the Morris water maze in 3xTg-AD mice at Lancet Neurol. 8, 151–157. doi: 10.1016/S1474-4422(08)70295-3 advanced stages of disease. Behav. Neurosci. 129, 129–137. doi: 10.1037/ Bardin, L., Kleven, M. S., Barret-Grevoz, C., Depoortere, R., and Newman- bne0000027 Tancredi, A. (2006). Antipsychotic-like vs cataleptogenic actions in mice of Frontiers in Pharmacology | www.frontiersin.org 22 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics novel antipsychotics having D2 antagonist and 5-HT1A agonist properties. Giménez-Llort, L., Maté, I., Manassra, R., Vida, C., and De la Fuente, M. (2012). Neuropsychopharmacology 31, 1869–1879. doi: 10.1038/sj.npp.1300940 Peripheral immune system and neuroimmune communication impairment in Belfiore, R., Rodin, A., Ferreira, E., Velazquez, R., Branca, C., Caccamo, A., a mouse model of Alzheimer’s disease. Ann. N. Y. Acad. Sci. 1262, 74–84. doi: et al. (2019). Temporal and regional progression of Alzheimer’s disease-like 10.1111/j.1749-6632.2012.06639.x pathology in 3xTg-AD mice. Aging Cell. 18, e12873. doi: 10.1111/acel.12873 Giménez-Llort, L., Garcia, Y., Buccieri, K., Revilla, S., Sunol, C., Cristofol, R., et Belzung, C., and Le Pape, G. (1994). Comparison of different behavioral test al. (2010). Gender-specific neuroimmunoendocrine response to treadmill situations used in psychopharmacology for measurement of anxiety. Physiol. exercise in 3xTg-AD mice. Int. J. Alzheimers Dis. 128354, 1–17. doi: Behav. 56, 623–628. doi: 10.1016/0031-9384(94)90311-5 10.4061/2010/128354 Bhalerao, S., Seyfried, L. S., Kim, H. M., Chiang, C., Kavanagh, J., and Kales, H. C. Giménez-Llort, L., Arranz, L., Maté, I., and De la Fuente, M. (2008). Gender-specific (2012). Mortality risk with the use of atypical antipsychotics in later-life bipolar neuroimmunoendocrine aging in a triple-transgenic3xTgAD mouse model for disorder. J. Geriatr. Psychiatry Neurol. 25, 29–36. doi: 10.1177/0891988712436687 Alzheimer’s disease and its relation with longevity. Neuroimmunomodulation Broekkamp, C. L., Rijk, H. W., Joly-Gelouin, D., and Lloyd, K. L. (1986). Major 15, 331–343. doi: 10.1159/000156475 tranquillizers can be distinguished from minor tranquillizers on the basis Giménez-Llort, L., Blázquez, G., Cañete, T., Johansson, B., Oddo, S., Tobeña, A., of effects on marble burying and swim-induced grooming in mice. Eur. J. et al. (2007). Modeling behavioral and neuronal symptoms of Alzheimer’s Pharmacol. 126, 223–229. doi: 10.1016/0014-2999(86)90051-8 disease in mice: a role for intraneuronal amyloid. Neurosci. Biobehav. Rev. 31, Bruins Slot, L. A., Kleven, M. S., and Newman-Tancredi, A. (2005). Effects of 125–147. doi: 10.1016/j.neubiorev.2006.07.007 novel antipsychotics with mixed D(2) antagonist/5-HT(1A) agonist properties Giménez-Llort, L., Blázquez, G., Cañete, T., Rosa, R., Vivó, M., Oddo, S., et al on PCP-induced social interaction deficits in the rat. Neuropharmacology 49, (2006). “Modeling neuropsychiatric symptoms of Alzheimer’s disease dementia 996–1006. doi: 10.1016/j.neuropharm.2005.05.013 in 3xTg-AD mice,” in Alzheimer’s Disease: new advances. Eds. K. Iqbal, B. Cope, M. B., Li, X., Jumbo-Lucioni, P., DiCostanzo, C. A., Jamison, W. G., Winblad, and J. Avila (Pianoro: Medimond SRL), 513–516. Kesterson, R. A., et al. (2009). Risperidone alters food intake, core body Guayerbas, N., Catalán, M., Víctor, V. M., Miquel, J., and De la Fuente, M. (2001). temperature, and locomotor activity in mice. Physiol. Behav. 96, 457–463. doi: Relation of behaviour and macrophage function to life span in a murine model 10.1016/j.physbeh.2008.11.011 of premature immunosenescence. Behav. Brain Res. 134, 41–48. doi: 10.1016/ Cummings, J. L. (2000). Cognitive and behavioral heterogeneity in Alzheimer’s S0166-4328(01)00449-1 disease: Seeking the neurobiological basis. Neurobiol. Aging 21, 845–861. doi: Gould, G. G., Hensler, J. G., Burke, T. F., Benno, R. H., Onaivi, E. S., and Daws, L. C. 10.1016/S0197-4580(00)00183-4 (2011). Density and function of central serotonin (5-HT) transporters, 5-HT1A Deacon, R. M., and Rawlins, J. N. (2006). T-maze alternation in the rodent. Nat. and 5-HT2A receptors, and effects of their targeting on BTBR T+tf/J mouse social Protoc. 1, 7–12. doi: 10.1038/nprot.2006.2 behavior. J. Neurochem. 116, 291–303. doi: 10.1111/j.1471-4159.2010.07104.x Delotterie, D., Ruiz, G., Brocard, J., Schweitzer, A., Roucard, C., Roche, Y., et al. Hall, C.S., Ballachey, E.L. (1932). “A study of the rat’s behavior in a field: a (2010). Chronic administration of atypical antipsychotics improves behavioral contribution to method in comparative psychology”. in University of California and synaptic defects of STOP null mice. Psychopharmacol. (Berl) 208, 131–141. Publications in Psychology. (Berkeley: Univ. of California Press), 1–12. doi: 10.1007/s00213-009-1712-3 Hall, C.S. (1934). Emotional behavior in the rat. I. Defecation and urination as De Oliveira, M., Matias, P., Ribeiro, J. E., De Sousa, R., Camelo, E., De Freitas, R., measures of individual differences in emotionality. J. Comp. Psychol. 18, 385– et al. (2008). Activities of the antipsychotic drugs haloperidol and risperidone 403. doi: 10.1037/h0071444 on behavioural effects Induced by ketamine in mice. Sci. Pharm. 76, 673–687. He, J., Luo, H., Yan, B., Yu, Y., Wang, H., Wei, Z., et al. (2009). Beneficial effects doi: 10.3797/scipharm.0810-11 of quetiapine in a transgenic mouse model of Alzheimer’s disease. Neurobiol. Didriksen, M., Skarsfeldt, T., and Arnt, J. (2007). Reversal of PCP-induced Aging 30, 1205–1216. doi: 10.1016/j.neurobiolaging.2007.11.001 learning and memory deficits in the Morris’ water maze by sertindole and Hou, Y., Wu, C. F., Yang, J. Y., and Guo, T. (2006). Differential effects of other antipsychotics. Psychopharmacol. (Berl) 193, 225–233. doi: 10.1007/ haloperidol, clozapine and olanzapine on learning and memory functions s00213-007-0774-3 in mice. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 30, 1486–1495. doi: Douglas, R. J. (1966). Cues for spontaneous alternation. J. Comp. Physiol. Psychol. 10.1016/j.pnpbp.2006.06.001 62, 171–183. doi: 10.1037/h0023668 Huybrechts, K. F., Gerhard, T., Crystal, S., Olfson, M., Avorn, J., Levin, R., et al. Durán, J. C., Greenspan, A., Diago, J. I., Gallego, R., and Martinez, G. (2005). (2012). Differential risk of death in older residents in nursing homes prescribed Evaluation of risperidone in the treatment of behavioral and psychological specific antipsychotic drugs: Population based cohort study. BMJ 344, e977. symptoms and sleep disturbances associated with dementia. Int. Psychogeriatr. doi: 10.1136/bmj.e977 17, 591–604. doi: 10.1017/S104161020500219X Jeste, D. V., Blazer, D., Casey, D., Meeks, T., Salzman, C., Schneider, L., et al. Dwyer, D. S., and Donohoe, D. (2003). Induction of hyperglycemia in mice with (2008). ACNP white paper: Update on use of antipsychotic drugs in elderly atypical antipsychotic drugs that inhibit glucose uptake. Pharmacol. Biochem. persons with dementia. Neuropsychopharmacology 33, 957–970. doi: 10.1038/ Behav. 75 (2), 255–260. doi: 10.1016/S0091-3057(03)00079-0 sj.npp.1301492 File, S. E., and Hyde, J. R. (1978). Can social interaction be used to measure Kalman, J., Kalman, S., and Pakaski, M. (2008). Recognition and treatment of anxiety? Br. J. Pharmacol. 62, 19–24. doi: 10.1111/j.1476-5381.1978.tb07001.x behavioral and psychological symptoms of dementias: Lessons from the File, S., and Seth, P. (2003). A review of 25 years of the social interaction test. Eur. CATIE-AD study. Neuropsychopharmacol. Hung. 10, 233–249. J. Pharmacol. 463, 35–53. doi: 10.1016/S0014-2999(03)01273-1 Kales, H. C., Kim, H. M., Zivin, K., Valenstein, M., Seyfried, L. S., Chiang, C., García-Mesa, Y., Giménez-Llort, L., López, L. C., Venegas, C., Cristòfol, R., et al. (2012). Risk of mortality among individual antipsychotics in patients with Escames, G., et al. (2012). Melatonin plus physical exercise are highly dementia. Am. J. Psychiatry 169, 71–79. doi: 10.1176/appi.ajp.2011.11030347 neuroprotective in the 3xTg-AD mouse. Neurobiol. Aging 33, 1124.e13–1124. Kalueff, A. V., and Tuohimaa, P. (2004b). Grooming analysis algorithm for e29. doi: 10.1016/j.neurobiolaging.2011.11.016 neurobehavioural stress research. Brain Res. Protoc. 13, 151–158. doi: 10.1016/j. García-Mesa, Y., Lopez-Ramos, J. C., Gimenez-Llort, L., Revilla, S., Guerra, R., brainresprot.2004.04.002 Gruart, A., et al. (2011). Physical exercise protects against Alzheimer’s disease in Kalueff, A. V., and Tuohimaa, P. (2004a). Contrasting grooming phenotypes 3xTg-AD mice. J. Alzheimers Dis. 24, 421–454. doi: 10.3233/JAD-2011-101635 in C57Bl/6 and 129S1/SvImJ mice. Brain Res. 1028, 75–82. doi: 10.1016/j. Giménez-Llort, L., Guitart-Masip, M., Tobeña, A., Fernández-Teruel, A., and brainres.2004.09.001 Johansson, B. (2018). Distinct phenotypes of spontaneous activity and Kane, A. E., Shin, S., Wong, A. A., Fertan, E., Faustova, N. S., Howlett, S. E., et al. induction of amphetamine sensitization in inbred Roman high- and low- (2018). Sex differences in healthspan predict lifespan in the 3xTg-AD mouse avoidance rats: vulnerability and protection. Neurosci. Lett. 23, 92–98. doi: model of Alzheimer’s disease. Front. Aging Neurosci. 10, 172. doi: 10.3389/ 10.1016/j.neulet.2018.03.011 fnagi.2018.00172 Giménez-Llort, L., Torres-Lista, V., and De la Fuente, M. (2014). Crosstalk between Kaurav, B. P., Wanjari, M. M., Chandekar, A., Chauhan, N. S., and Upmanyu, N. behavior and immune system during the prodromal stages of Alzheimer’s disease. (2012). Influence of withania somnifera on obsessive compulsive disorder in Curr. Pharm. Des. 20, 4723–4732. doi: 10.2174/1381612820666140130205500 mice. Asian Pac. J. Trop. Med. 5, 380–384. doi: 10.1016/S1995-7645(12)60063-7 Frontiers in Pharmacology | www.frontiersin.org 23 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics Katz, I., de Deyn, P. P., Mintzer, J., Greenspan, A., Zhu, Y., and Brodaty, H. Oddo, S., Caccamo, A., Kitazawa, M., Tseng, B. P., and LaFerla, F. M. (2003a). (2007). The efficacy and safety of risperidone in the treatment of psychosis Amyloid deposition precedes tangle formation in a triple transgenic model of Alzheimer’s disease and mixed dementia: a meta-analysis of 4 placebo- of Alzheimer’s disease. Neurobiol. Aging 24, 1063–1070. doi: 10.1016/j. controlled clinical trials. Int. J. Geriatr. Psychiatry 22, 475–484. doi: 10.1002/ neurobiolaging.2003.08.012 gps.1792 Oddo, S., Caccamo, A., Shepherd, J. D., Murphy, M. P., Golde, T. E., Kayed, R., Kilkenny, C., Browne, W. J., Cuthill, I. C., Emerson, M., and Altman, D. G. et al. (2003b). Triple-transgenic model of Alzheimer’s disease with plaques and (2010). Improving bioscience research reporting: the ARRIVE guidelines tangles: intracellular Aβ and synaptic dysfunction. Neuron 39, 409–421. doi: for reporting animal research. PLOS Biol. 8, e1000412. doi: 10.1371/journal. 10.1016/S0896-6273(03)00434-3 pbio.1000412 Onor, M. L., Saina, M., Trevisiol, M., Cristante, T., and Aguglia, E. (2007). Clinical Kitazawa, M., Oddo, S., Yamasaki, T. R., Green, K. N., and LaFerla, F. M. (2005). experience with risperidone in the treatment of behavioral and psychological Lipopolysaccharide-induced inflammation exacerbates tau pathology by a cyclin- symptoms of dementia. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 31, 205– dependent kinase 5-mediated pathway in a transgenic model of Alzheimer’s 209. doi: 10.1016/j.pnpbp.2006.09.001 disease. J. Neurosci. 25, 8843–8853. doi: 10.1523/JNEUROSCI.2868-05.2005 Pike, C. J. (2017). Sex and the development of Alzheimer’s disease. J Neurosci. Res. Kleijer, B. C., van Marum, R. J., Egberts, A. C., Jansen, P. A., Knol, W., and Heerdink, 95, 671–680. doi: 10.1002/jnr.23827 E. R. (2009). Risk of cerebrovascular events in elderly users of antipsychotics. Reisberg, B., Borenstein, J., Salob, S. P., Ferris, S. H., Franssen, E., and Georgotas, J. Psychopharmacol. 23, 909–914. doi: 10.1177/0269881108093583 A. (1987). Behavioral symptoms in Alzheimer’s disease: phenomenology and Lalonde, R., Lewis, T. L., Strazielle, C., Kim, H., and Fukuchi, K. (2003). Transgenic treatment. J. Clin. Psychiatry 48 (Suppl.), 9–15. doi: 10.1037/t13385-000 mice expressing the βAPP695SWE mutation: effects on exploratory activity, Rezzani, R., Franco, C., and Rodella, L. F. (2019). Sex differences of brain and anxiety, and motor coordination. Brain Res. 977, 38–45. doi: 10.1016/ their implications for personalized therapy. Pharmacol. Res. 141, 429–442. doi: S0006-8993(03)02694-5 10.1016/j.phrs.2019.01.030 Lapane, K. L., Gambassi, G., Landi, F., Sgadari, A., Mor, V., and Bernabei, R. (2001). Rimondini, R., Ferré, S., Giménez-Llort, L., Ögren, S. O., and Fuxe, K. (1998). Gender differences in predictors of mortality in nursing home residents with Differential effects of selective adenosine A1 and A2A receptor agonists AD. Neurology 56, 650–654. doi: 10.1212/WNL.56.5.650 on dopamine receptor agonist-induced behavioural responses in rats. Eur. Leysen, J. E., Gommeren, W., Eens, A., de Chaffoy de Courcelles, D., Stoof, J. C., and J. Pharmacol. 347, 153–158. doi: 10.1016/S0014-2999(98)00107-1 Janssen, P. A. (1988). Biochemical profile of risperidone, a new antipsychotic. Rinaldi, A., Mandillo, S., Oliverio, A., and Mele, A. (2007). D1 and D2 receptor J. Pharmacol. Exp. Ther. 247, 661–670. antagonist injections in the prefrontal cortex selectively impair spatial Li, X., Morrow, D., and Witkin, J. M. (2006). Decreases in nestlet shredding of mice learning in mice. Neuropsychopharmacology 32, 309–319. doi: 10.1038/ by serotonin uptake inhibitors: comparison with marble burying. Life Sci. 78, sj.npp.1301176 1933–1939. doi: 10.1016/j.lfs.2005.08.002 Rodríguez-Antona, C., Gurwitz, D., de Leon, J., Llerena, A., Kirchheiner, J., et al. Mannens, G., Huang, M. L., Meuldermans, W., Hendrickx, J., Woestenborghs, R., (2009). CYP2D6 genotyping for psychiatric patients treated with risperidone: and Heykants, J. (1993). Absorption, metabolism, and excretion of risperidone considerations for cost-effectiveness studies. Pharmacogenomics 10, 685–699. in humans. Drug Metab. Dispos. 21, 1134–1141. doi: 10.2217/pgs.09.15 Marchese, M., Cowan, D., Head, E., Ma, D., Karimi, K., Ashthorpe, V., et al. (2014). Rodríguez-Arias, M., Miñarro, J., Aguilar, M. A., Pinazo, J., and Simón, Autoimmune manifestations in the 3xTg-AD model of Alzheimer’s disease. V. M. (1998). Effects of risperidone and SCH 23390 on isolation-induced J Alzheimers Dis. 39, 191–210. doi: 10.3233/JAD-131490 aggression in male mice. Eur Neuropsychopharmacol. 8, 95–103. doi: 10.1016/ Matsubara, Y., Ohnuma, T., Shibata, N., Takahashi, T., and Arai, H. (2007). S0924-977X(97)00051-5 Plasma concentration of risperidone showed a correlation with daily dose, Rondanelli, M., Sarra, S., Antoniello, N., Mansi, V., Govoni, S., Falvo, F., et al. but not with age nor cytochrome P-450 2D6 gene polymorphisms in Japanese (2006). No effect of atypical antipsychotic drugs on weight gain and risk of schizophrenics. Psychogeriatrics 2, 284–288. doi: 10.1111/j.1479-8301.2002. developing type II diabetes or lipid abnormalities among nursing home elderly tb00042.x patients with Alzheimer’s disease. Minerva Med. 97, 147–151. Matsushita, M., Egashira, N., Harada, S., Okuno, R., Mishima, K., Iwasaki, K., Salzman, C., Jeste, D. V., Meyer, R. E., Cohen-Mansfield, J., Cummings, J., et al. (2005). Perospirone, a novel antipsychotic drug, inhibits marble-burying Grossberg, G. T., et al. (2008). Elderly patients with dementia-related symptoms behavior via 5-HT1A receptor in mice: implications for obsessive-compulsive of severe agitation and aggression: consensus statement on treatment options, disorder. J. Pharmacol. Sci. 99, 154–159. doi: 10.1254/jphs.FP0050144 clinical trials methodology, and policy. J. Clin. Psychiatry 69, 889–898. doi: Miyamoto, J., Tsuji, M., Takeda, H., Ohzeki, M., Nawa, H., and Matsumiya, 10.4088/JCP.v69n0602 T. (2004). Characterization of the anxiolytic-like effects of fluvoxamine, Savoy, Y. E., Ashton, M. A., Miller, M. W., Nedza, F. M., Spracklin, D. K., Hawthorn, milnacipran and risperidone in mice using the conditioned fear stress M. H., et al. (2010). Differential effects of various typical and atypical paradigm. Eur. J. Pharmacol. 504, 97–103. doi: 10.1016/j.ejphar.2004.09.043 antipsychotics on plasma glucose and insulin levels in the mouse: Evidence for Moechars, D., Gilis, M., Kuiperi, C., Laenen, I., and Van Leuven, F. the involvement of sympathetic regulation. Schizophr. Bull. 36, 410–418. doi: (1998). Aggressive behaviour in transgenic mice expressing APP 10.1093/schbul/sbn104 is alleviated by serotonergic drugs. Neuroreport 9, 3561–3564. doi: Schneider, L. S., Dagerman, K. S., and Insel, P. (2005). Risk of death with 10.1097/00001756-199811160-00004 atypical antipsychotic drug treatment for dementia: meta-analysis of Morris, R. G. M. (1981). Spatial localization does not require the presence of local randomized placebo-controlled trials. JAMA 294, 1934–1943. doi: 10.1001/ cues. Learn. Motiv. 12, 239–260. doi: 10.1016/0023-9690(81)90020-5 jama.294.15.1934 Morris, R. (1984). Developments of a water-maze procedure for studying Shekelle, P., Maglione, M., Bagley, S., Suttorp, M., Mojica, W. A., Carter, J., et al. spatial learning in the rat. J Neurosci. Methods 11, 47–60. doi: (2007). Efficacy and comparative effectiveness of off-label use of atypical 10.1016/0165-0270(84)90007-4 antipsychotics [Internet]. AHRQ Comp. Eff. Rev. 6, 1–82. Report No.: Mutlu, O., Celikyurt, I. K., Ulak, G., Tanyeri, P., Akar, F. Y., and Erden, F. (2012). Effects 07-EHC003-EF. of olanzapine and clozapine on radial maze performance in naive and MK-801- Silverman, J. L., Tolu, S. S., Barkan, C. L., and Crawley, J. N. (2010). Repetitive treated mice. Arzneimittelforschung 62, 4–8. doi: 10.1055/s-0031-1291360 self-grooming behavior in the BTBR mouse model of autism is blocked by Nelson, R. L., Guo, Z., Halagappa, V. M., Pearson, M., Gray, A. J., Matsuoka, Y., et al. the mGluR5 antagonist MPEP. Neuropsychopharmacology 35, 976–989. doi: (2007). Prophylactic treatment with paroxetine ameliorates behavioral deficits 10.1038/npp.2009.201 and retards the development of amyloid and tau pathologies in 3xTgAD mice. Tan, L. L., Wong, H. B., and Allen, H. (2005). The impact of neuropsychiatric Exp. Neurol. 205, 166–176. doi: 10.1016/j.expneurol.2007.01.037 symptoms of dementia on distress in family and professional caregivers in Njung’e, K., and Handley, S. L. (1991). Evaluation of marble-burying behavior Singapore. Int. Psychogeriatr. 17, 253–263. doi: 10.1017/S1041610205001523 as a model of anxiety. Pharmacol. Biochem. Behav. 38, 63–67. doi: Tempier, A., He, J., Zhu, S., Zhang, R., Kong, L., Tan, Q., et al. (2013). 10.1016/0091-3057(91)90590-X Quetiapine modulates conditioned anxiety and alternation behavior Frontiers in Pharmacology | www.frontiersin.org 24 September 2019 | Volume 10 | Article 1061 Torres-Lista et al. Early Mortality Risk by Antipsychotics in Alzheimer’s transgenic mice. Curr. Alzheimer Res. 10, 199–206. doi: of social interaction in phencyclidine-treated mice as a schizophrenic 10.2174/1567205011310020010 animal model. Neuropharmacology 52, 1179–1187. doi: 10.1016/j. Torres-Lista, V., and Giménez-Llort, L. (2019). Vibrating tail, digging, body/ neuropharm.2006.12.007 face interaction and barbering: sex-dependent behavioral signatures Yoon, S., Noh, J. S., Choi, S., and Baik, J. (2010). Effects of atypical antipsychotic of social dysfunction in 3xTg-AD mice at advanced stages of disease as drugs on body weight and food intake in dopamine D2 receptor knockout compared to mice with normal aging. J. Alzheimer’s Dis. (JAD) 69 (4), 969– mice. Biochem. Biophys. Res. Commun. 393, 235–241. doi: 10.1016/j. 977. doi: 10.3233/JAD-190253 bbrc.2010.01.108 Torres-Lista, V., De la Fuente, M., and Giménez-Llort, L. (2017). Survival curves Zhang, K., Grady, C. J., Tsapakis, E. M., Andersen, S. L., Tarazi, F. I., and and behavioral profiles of female 3xTg-AD mice surviving to 18-months of age Baldessarini, R. J. (2004). Regulation of working memory by dopamine D4 receptor as compared to mice with normal aging. J. Alzheimers Dis. Rep. 1, 47–57. doi: in rats. Neuropsychopharmacology 29, 1648–1655. doi: 10.1038/sj.npp.1300491 10.3233/ADR-170011 Zheng, L., Mack, W. J., Dagerman, K. S., Hsiao, J. K., Lebowitz, B. D., Lyketsos, Torres-Lista, V., López-Pousa, S., and Giménez-Llort, L. (2015). Marble-burying is C. G., et al. (2009). Metabolic changes associated with second-generation enhanced in 3xTg-AD mice, can be reversed by risperidone and it is modulable antipsychotic use in Alzheimer’s disease patients: the CATIE-AD study. Am. J. by handling. Behav. Processes. 116, 69–74. doi: 10.1016/j.beproc.2015.05.001 Psychiatry 166, 583–590. doi: 10.1176/appi.ajp.2008.08081218 Trifiro, G., Spina, E., and Gambassi, G. (2009). Use of antipsychotics in elderly Zhou, Z. L., Li, X., Peng, H. Y., Yu, X. Y., Yang, M., Su, F. L., et al. (2006). Multiple patients with dementia: do atypical and conventional agents have a similar dose pharmacokinetics of risperidone and 9-hydroxyrisperidone in Chinese safety profile? Pharmacol. Res. 59, 1–12. doi: 10.1016/j.phrs.2008.09.017 female patients with schizophrenia. Acta Pharmacol. Sin. 27, 381–386. doi: Uchida, N., Egashira, N., Iwasaki, K., Ishibashi, A., Tashiro, R., Nogami, A., 10.1111/j.1745-7254.2006.00256.x et al. (2009). Yokukansan inhibits social isolation-induced aggression and methamphetamine-induced hyperlocomotion in rodents. Biol. Pharm. Bull. Conflict of Interest Statement: The authors declare that the research was 32, 372–375. doi: 10.1248/bpb.32.372 conducted in the absence of any commercial or financial relationships that could Vilalta-Franch, J., Lopez-Pousa, S., Garre-Olmo, J., Turon-Estrada, A., and Pericot- be construed as a potential conflict of interest. Nierga, I. (2008). Mortality rates in patients with Alzheimer’s disease treated with atypical neuroleptic drugs. Rev. Neurol. 46, 129–134. doi: 10.33588/rn.4603.2007382 Copyright © 2019 Torres-Lista, López-Pousa and Giménez-Llort. This is an open- Vloeberghs, E., Coen, K., Van Dam, D., and De Deyn, P. P. (2008). Validation of access article distributed under the terms of the Creative Commons Attribution the APP23 transgenic mouse model of Alzheimer’s disease through evaluation License (CC BY). The use, distribution or reproduction in other forums is permitted, of risperidone treatment on aggressive behaviour. Arzneimittelforschung 58, provided the original author(s) and the copyright owner(s) are credited and that the 265–268. doi: 10.1055/s-0031-1296505 original publication in this journal is cited, in accordance with accepted academic Wang, D., Noda, Y., Zhou, Y., Nitta, A., Furukawa, H., and Nabeshima, T. practice. No use, distribution or reproduction is permitted which does not comply with (2007). Synergistic effect of galantamine with risperidone on impairment these terms. Frontiers in Pharmacology | www.frontiersin.org 25 September 2019 | Volume 10 | Article 1061