Depression is a long-lasting heterogeneous neuropsychiatric disorder and one of the most common mental diseases, which places a significant economic burden on public health and decreases individuals’ quality of life (Wittchen et al., 2011; Lim et al., 2012; Murray et al., 2013; Whiteford et al., 2013). It affects approximately 4.4% of the world’s population with an incidence rate above the rate of global population growth (Flux and Lowry, 2020). The symptoms include sadness, helplessness, guilt and loss of appetite, sexual desire, and interest in activities that once were pleasurable (for at least 2 weeks) along with recurrent thoughts of suicide (McCarter, 2008; American-Psychiatric-Association [APA], 2013). Because of the unique and complex features in addition to the subjective symptoms of human depression, the generation of valid and insightful depression models for the development of new therapeutic drugs is not straightforward. Still, different rodent models have shown high face validity to human depression and are used in preclinical studies. These depression-like behaviors can be induced in rodents by several means including genetic, environmental, chemical and pharmacological manipulation and brain lesions [for recent reviews, see (Flux and Lowry, 2020; Becker et al., 2021)]. Several tests are usually performed to assess distinct components of depression in rodent models. For example, the tail suspension and forced swim tests measure despair, the splash test measures apathy, while the sucrose preference test evaluates anhedonia (Eltokhi et al., 2018; Becker et al., 2021). Additionally, tests that measure anxiety including open field, elevated plus maze and light/dark compartment are often performed complementary to depression tests for the full characterization of rodent models of neuropsychiatric disorders (Belovicova et al., 2017).

The prevalence and clinical characteristics of depression differ between women and men with women suffering from depression nearly twice as frequently as men during lifetime, independently of race or ethnicity (Weissman and Klerman, 1977; Cyranowski et al., 2000; Andrade et al., 2003; Ford and Erlinger, 2004; Patten et al., 2006; Bromet et al., 2011; Salk et al., 2017). However, male rodent models have traditionally been used in genetic and pharmacological studies of depression. Performing behavioral experiments on male rodents is linked to the assumption of estrous-linked changes in the baseline behavioral activity in females. This concern may require testing female rodents at each of the four stages of the estrous cycle to generate reliable data, which may complicate the experimental design. On the other hand, neglecting the sex difference in depression-related experiments may provoke false interpretations of the results and prevents a successful translation of experimental data into the clinic. Thus, it is essential incorporating female rodents in behavioral, molecular and electrophysiological analyses to decipher the biology behind the sex bias of depression as a prerequisite for improving therapeutics. Unfortunately, so far, rodent models have yielded inconsistent results in different studies and often reported more depression-like symptoms in males than females (Kokras et al., 2009; Dalla et al., 2010; Barkus, 2013; Van den Hove et al., 2013; Najjar et al., 2018; Becker et al., 2021). By standardization of behavioral experiments and by taking into consideration factors that affect behavioral outcomes, the variability in results on sex differences can be likely reduced. Additionally, the sex differences of depression rodent models may also be induced by differences in the general performance and baseline activity of male and female wild-types in depression-related behavioral tests. To test these hypotheses, we investigated the baseline depression-like behaviors of both male and female wild-type mice in three standard behavioral tests, tail suspension, sucrose preference and forced swim tests. We performed our analysis in three inbred strains, C57BL/6N, DBA/2, and FVB/N, well-established to affect the behavioral outcome (McFadyen et al., 2003; Brooks et al., 2005; Moy et al., 2007; Peleh et al., 2019; Eltokhi et al., 2020, 2021), and typically used as background strains in mouse models of neuropsychiatric disorders. Furthermore, although behavioral studies are frequently performed in adult mice taking the advantage of easy handling and complex behaviors, we performed our analysis at two different developmental stages during adolescence since the onset of neuropsychiatric symptoms emerges mainly during adolescence, with more than 50% of adults with neuropsychiatric disorders receiving a diagnosis before 15 years of age (Kim-Cohen et al., 2003; Paus et al., 2008). Because behavioral results are known to be sensitive to small developmental progress during adolescence (Peleh et al., 2019; Eltokhi et al., 2020), we compared the results of these behavioral experiments between these two developmental stages in mice differing only by a few days in age.

Our work indicates that the performance of mice in depression-related behavioral tests during adolescence is mainly strain and age dependent with no obvious effect of sex. This outlines the drawbacks of using an individual strain in genetic and pharmacological studies of depression and highlights the benefits of using adolescent mice in characterizing rodent models of depression, which may reduce inconsistency of results between different laboratories.

The global increased prevalence of depression in women compared to men suggests that the differential risk is highly dependent on biological sex differences rather than race, culture or other potentially confounding social and economic factors (Albert, 2015). As the onset of depression in women peaks in their reproductive years, the increased prevalence of depression may be explained in part by sex hormones. Indeed, the female hormonal fluctuation during puberty, menstruation, pregnancy, and menopause is a trigger for depression (Albert, 2015). The risk of depression increases during the perimenopausal transition (Cohen et al., 2006), with hormone replacement therapy being effective in the prevention of postmenopausal depression in women (Gordon and Girdler, 2014). Androgens seem to have anxiolytic properties whereas estrogens receptors (ER) activation has opposite consequences with ERα having largely anxiogenic-like properties and ERβ serving to generate anxiolytic-like effects (Borrow and Handa, 2017).

Here we evaluated depression-like behaviors in adolescent mice at two different developmental stages. Our previous studies unraveled that sex differences during adolescence are apparent in certain strains and behaviors such as general activity, anxiety and cognitive function (Eltokhi et al., 2020) as also shown for adult mice (Tarantino et al., 2000). Given the divergent susceptibility of males and females to depression, there is an urgent need for tackling the effect of sex in rodents on behavioral tests assessing depression-like behaviors. However, in all the tests commonly used to assess depression-like behaviors in rodents, we did not find any sex difference in C57BL/6N, DBA/2, or FVB/N strains at these two developmental stages during adolescence. Since we investigated the baseline depression-like behaviors only in young mice before puberty corresponding to P42 in mice (Taft et al., 2006; Dutta and Sengupta, 2016), with no severe effect of sex hormones, we interpret the finding that the effect of sex may become apparent only in adult mice. Indeed, a sex difference in the prevalence of patients with major depressive disorder appears mainly after puberty (Bebbington et al., 1998; Merikangas et al., 2010; Mehta et al., 2013). In adult C57BL/6J mice, the baseline immobility durations in the tail suspension tests were lower in males than females (Liu and Gershenfeld, 2001). Thus, the biological maturation occurring during puberty along with the intensification of sex-specific social roles may be a major key of sex differences in depression (Dalla et al., 2010). On the other hand, a more recent piece of evidence suggests that the sex difference in depression begins in childhood and becomes more pronounced during adolescence (Breslau et al., 2017). Irrespective of these analyses in humans, inconsistent results for depression and anxiety-like behaviors have been reported when comparing male and female adolescent rodents to adults (Slawecki, 2005; Doremus et al., 2006; Hefner and Holmes, 2007; Doremus-Fitzwater et al., 2009; Martínez-Mota et al., 2011). Thus, further developmental and longitudinal studies in rodents should be performed to assess the baseline depression in males and females along different developmental stages. Since the onset of puberty is strain dependent (Pintér et al., 2007), a rough dating of puberty of rodent models should be employed using various markers including the vaginal opening, first vaginal cornification, onset of cyclicity in females and balanopreputial separation in males.

Strain differences in behavioral tests assessing neuropsychiatric-like phenotypes in adolescent mice were repeatedly reported by our group (Peleh et al., 2019; Eltokhi et al., 2020, 2021). The contribution of genetic factors specifically in the depression-related behaviors of adult inbred rodents as well as in the efficacy of antidepressant drug treatment has been previously shown (Vaugeois et al., 1997; Liu and Gershenfeld, 2001; Lucki et al., 2001). In our study, the baseline depression-like behaviors differed between different strains. In both tail suspension and forced swim tests, C57BL/6N mice showed the highest immobility duration and the lowest latency to first immobility compared to DBA/2 and FVB/N. Interestingly, FVB/N mice at P26 did not show any immobility in the forced swim test. This behavior can be explained by their severe hyperactivity as previously reported (Eltokhi et al., 2020). These data are consistent with a former study showing adult FVB/NJ mice having the least baseline immobility duration in the forced swim test, DBA/2J being intermediate and C57BL/6J showing the highest immobility duration (Lucki et al., 2001). On the other hand, adolescent FVB/N mice showed the highest sucrose preference index, highlighting the suitability of this test to be performed on mice with FVB/N background. In contrast, DBA/2 mice showed low ability to perform the sucrose preference test and had a high baseline anhedonic level, which may mask the depression behavior in mouse models of neuropsychiatric disorders and should be taken into consideration in designing the experiments and analyzing the data. Nonetheless, DBA/2 can be a good strain to test the effect of anti-depressant drugs. Thus, the choice of the strain should be related to the specific scientific question being asked.

As our previous studies uncovered strong changes in behavioral outcomes including activity, anxiety and social interaction in a small developmental window during adolescence (Eltokhi et al., 2020, 2021), we here extended these findings assessing despair by tail suspension and forced swim tests. Both C57BL/6N and FVB/N mice showed an increase in immobility within 10 days of adolescence. For the sucrose preference test, the effect of the 10-day interval was only apparent in FVB/N mice. Interestingly, DBA/2 mice did not show any difference in the behavioral outcome at both time points, highlighting a direct correlation of the effect of age to specific strains. One limitation of our study is that the same mice were tested in a series of depression-related tests, which may have caused earlier tests to affect subsequent performance in later tests. To mitigate this possibility, we ordered the tests in such a way that the two stressful tests, tail suspension and forced swim tests, were not consecutive but separated by the sucrose preference test that was performed in the mouse homecage. Still, the single housing during the sucrose preference test may have also induced some levels of stress on mice, which may have affected their behaviors in the forced swim test. However, we believe that the possible effect of single housing was not strong as indicated by the similar pattern of behaviors in both tail suspension and forced swim tests at both developmental stages, with C57BL/6N mice showing increased immobility durations in both tests. Our findings suggest that taking the effect of age and strain into account will decrease variabilities and inconsistencies and increase the reproducibility of behavioral results between different laboratories.

Anxiety-related behavioral tests are often performed together with depression tests in mouse models of neuropsychiatric disorders since anxiety usually co-morbids with depression (Cryan and Mombereau, 2004; Millstein and Holmes, 2007; Krishnan and Nestler, 2011; Jung et al., 2014; Belovicova et al., 2017; Sokolowska et al., 2021). Several behavioral tests based on the conflict between competing behaviors of exploring novel environments and avoiding potential threatening situations have been validated to assess anxiety, including the elevated plus maze (Pellow et al., 1985; Lister, 1987), open field (Treit and Fundytus, 1988), and light/dark box (Crawley and Goodwin, 1980; Crawley, 1981; Blumstein and Crawley, 1983), where our previous studies revealed sex differences in the open field and elevated plus maze tests with female C57BL/6N and FVB/N mice being less anxious (Eltokhi et al., 2020). To unravel whether different anxiety paradigms tax distinct aspects of anxiety, we here investigated the sex difference in another anxiety-related test, the hyponeophagia test. Female C57BL/6N and DBA/2 but not FVB/N mice showed higher latency to consume the unfamiliar drink in an unfamiliar arena compared to males, reflecting increased bait shyness as an indication of anxiety. This discrepancy of results of males vs. females in different anxiety tests suggests that a battery of different tests should be used in studies of anxiety-related behaviors to draw the full picture of mouse phenotype (van Gaalen and Steckler, 2000).

Taken together, since depression is a complex disorder with several endophenotypes including despair, anhedonia and apathy, a complete test battery combining the well-established, robust behavioral tests should be employed when testing rodent models of neuropsychiatric disorders or investigating the effect of anti-depressant drugs. Sex, strain and age are suggested to have different effects on distinct behavioral tests. Therefore, performing a complete test battery will provide important information per animal/drug that cannot be covered in separate behavioral studies. Notably, there should be at least a 1-day rest period between each of the tests in the behavioral battery, with some rodent strains requiring even longer durations (Lad et al., 2009). Some paradigms currently used for assessing antidepressant and/or depression-like behaviors in mice are still questionable. For example, in recent years, there is a trend to interpret the transition from swimming to immobility in the forced swim test as a coping mechanism with inescapable stressors, rather than an indication of despair (Molendijk and de Kloet, 2019). However, the majority of researchers still qualify the rodent’s floating response as a depression-like behavior since the persistence of coping with inescapable stressors may indeed enhance the vulnerability to depression (Molendijk and de Kloet, 2019). These different scientific perspectives confirm the necessity of performing a complete test battery in order to draw a full picture of the depression-like behavior.

Finally, we want to touch on the issue of reproducibility of behavioral results in depression-related studies. One potential factor of discrepancy is the slightly different protocols between laboratories. Worth noticing, behavioral tests may cause stress and put unwanted burdens on rodents. To this end, the order of tests on a specific cohort can also play a role in the variabilities between the results of different laboratories. Moreover, different rodent strains and sex dependence of the stress response frequently result in apparent discrepancies in published data. However, our study, exemplified by trials 1 and 2 in latency to drink (Figure 4), demonstrated that consistency of results can be easily achieved by adequate test settings. Furthermore, during adolescence, no sex differences were found. This finding does not exclude an impact of sex in other developmental stages. However, in studies where sex differences are suggested to provoke an additional complication in depression analysis, adolescent mice may open a window for clearly identifying such effects since the general performance and baseline depression activity of male and female wild-type mice are similar.

Thus, strain, age and sex should be taken into account when analyzing neuropsychiatric disorders in mouse models. Without question, optimization and standardization of depression-related tests in rodent models will help in understanding pathophysiological mechanisms and in identifying novel targets for depression treatment.

To our knowledge, this is the first study investigating sex, age and strain effects on the baseline depression-like behaviors. We confirmed that genetic strain differences and even small differences in developmental stage are important determinants of depression-related behavioral outcomes. We suggest using adolescent mice, at least in the three investigated depression-related behavioral tests, to reduce variability and inconsistency between different laboratories. Nonetheless, sex differences in mice still need a thorough evaluation throughout their lifetime. Taken together, our behavioral studies in adolescent mice can be used as a guiding platform for the choice of the most suitable combinations of assays and appropriate strain, age and sex selection in mouse models of neuropsychiatric disorders.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The animal study was reviewed and approved by The animal Ethics Committee of the Government of Baden Württemberg, Regierungspräsidium Karlsruhe, (G-101/16).

AE: conceptualization, data curation, formal analysis, project administration, supervision, validation, visualization, and writing–original draft, review and editing. BK: methodology, data curation, and formal analysis. CP: conceptualization, project administration, supervision, and validation, review and editing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.