Breast cancer is the most common cancer in women (Ferlay et al., 2015), with nearly 70 % of patients having estrogen receptor-positive (ER+) cancers. Women suffering from ER+ breast cancers are commonly treated with aromatase inhibitors, which, by inhibiting the activity of the enzyme aromatase [2] encoded by the CYP19 gene, prevent the synthesis of 17β-estradiol (Burstein et al., 2014) from testosterone. Aromatase inhibitors, such as letrozole, are usually prescribed as daily oral treatment over 3 to 5 years (American Cancer, 2023). However, these drugs have long been associated with CNS-related side effects that impair quality of life, including hot flushes (Rand et al., 2011), cognitive deficits (Bayer et al., 2015; Collins et al., 2009; Jenkins et al., 2008; Schmidt et al., 2000), insomnia (Desai et al., 2013), and depression (Ganz et al., 2016). Importantly, these side effects can become so severe that a significant proportion (∼15 %) of women discontinues this potentially life-saving treatment (Chirgwin et al., 2016). Thus, there is an urgent need to better understand the mechanism of action of letrozole and introduce novel therapies that can mitigate these CNS-related effects and, thereby, improve the quality of life of breast cancer patients treated with aromatase inhibitors to increase treatment compliance.

The CNS-related side effects of aromatase inhibitors are thought to be due to lack of E2 produced by the ovaries and de novo in different brain regions, including those associated with learning and memory, regulation of sleep, depression, and anxiety (e.g., hippocampus, cerebral cortex, amygdala, and raphe nuclei) (Roselli et al., 1985; Stoffel-Wagner et al., 1999). Indeed, the de novo synthesis of E2 in the hippocampus induced by aromatase results in higher E2 concentrations (8.4 nM) in the hippocampus than in the peripheral blood (0.014 nM) (Hojo et al., 2004).

The hippocampus has a central role for memory formation and consolidation in the mammalian brain, and hippocampal functions are highly supported by E2 (Mukai et al., 2010). Dendritic spine and synaptic densities are regulated by the estrus cycle (being the highest in proestrus) (Woolley and McEwen, 1992) and treatment of ovariectomized (OVX) animals with E2. Estrogen increases: (i) synaptic protein levels (such as spinophilin, a marker of dendritic spines) (Kaidah et al., 2016; Li et al., 2004; Tuscher et al., 2016; Woolley et al., 1990; Woolley et al., 1996; Xing et al., 2018), (ii) phosphorylation of Cofilin (p-Cofilin stabilizes the spine cytoskeleton) (Kretz et al., 2004; Vierk et al., 2012; Zhou et al., 2004), and (iii) downregulates steroid receptor coactivator-1 (SRC-1) expression (Bian et al., 2014; Liu et al., 2015) and the expression of the excitatory synapse marker PSD95 (Liu et al., 2015) each involved in the regulation of hippocampal synaptic plasticity and spatial memory (Bian et al., 2014; Liu et al., 2015). Synapse and dendritic spine densities greatly contribute to synaptic plasticity and to learning and memory behaviors (Dominguez-Iturza et al., 2016; Meng et al., 2011). Hippocampus-dependent spatial reference memory fluctuates with the estrous cycle (Frick and Berger-Sweeney, 2001) and is improved by E2 on OVX animals (Kaidah et al., 2016; Li et al., 2004). Earlier studies demonstrated that intrahippocampal administration of letrozole impaired spatial working memory in intact and OVX rats (Marbouti et al., 2020), and non-spatial memory in OVX mice (Tuscher et al., 2016b). Similarly, inhibition of hippocampal aromatization decreases spatial learning and memory in male zebra finches (Bailey et al., 2017; Bailey and Saldanha, 2015). The negative effect of letrozole on cognition appears to be a result of suppressed synaptic plasticity, because inhibition of E2 synthesis by letrozole applied to primary hippocampal cultures or microinjected into the hippocampus of OVX mice reduces the number of dendritic spine synapses (Fester et al., 2012; Hojo et al., 2004; Kretz et al., 2004; Prange-Kiel et al., 2006).

In contrast to intrahippocampal injections, systemic administration of letrozole that is clinically relevant shows inconsistent effect on spatial learning and memory in mice and rats, ranging from impairment (Liu et al., 2019; Zameer and Vohora, 2017) or no effect (Meng et al., 2011; Vierk et al., 2015) in mice, and to improvement in rats (Alejandre-Gomez et al., 2007; Aydin et al., 2008). In addition to differences in the response to letrozole between rats and mice, there are differences among mouse strains (Cestari et al., 1999; Ciamei et al., 2000; Lipartiti et al., 1993) that can affect the effects of letrozole on cognitive performance. Inbred and outbred strains differ from one another in genetic makeup, in neurochemistry (Ingram and Corfman, 1980), synaptic plasticity, and learning and memory (Nguyen, 2006). For example, several studies have demonstrated strain-dependent effects of GABA-receptor agonists and antagonists in a number of behavioral tests (Castellano et al., 1993; Jacobson and Cryan, 2005; Puglisi-Allegra et al., 1981; Simler et al., 1982). Therefore, comparison of different mouse strains provides valuable information about the importance of genetic background in behavioral and pharmacodynamic phenotypes. To date, C57Bl mice were the most often used strain in studies evaluating effect of letrozole on spatial learning (Liu et al., 2015; Meng et al., 2011; Vierk et al., 2015). Since our goal was to evaluate the effect of prolonged exposure to letrozole on cognitive functions in an animal model used in breast cancer research (Orlandella et al., 2021), the present study used female BALB/c mice. The nude phenotype of BALB/c mice is the most frequently used animal model in breast cancer research (Orlandella et al., 2021), because they lack the thymus and are deficient in T cells, which permits the xenograft to grow in them (Hunt et al., 2000). The BALB/c mice are significantly different from C57BL mice in their emotionality (Brinks et al., 2007; Razzoli et al., 2011; Shoji and Miyakawa, 2019) and neurochemistry (Bach et al., 2011). Therefore, their evaluation may provide additional information about effects of letrozole on cognitive behavior. Women suffering from breast cancer, due to the nature of the disease, have emotional instability, higher prevalence of anxiety and depression (Busby et al., 2018; Bui et al., 2020; Maass et al., 2015; Trevino et al., 2020; Wang et al., 2020). The BALB/c mice also have high levels of anxiety (Belzung and Griebel, 2001; Sartori et al., 2011) and therefore can mimic emotional conditions of humans diagnosed with cancer.

The goal of the current study was to find a cognitive domain that was affected by a prolonged exposure to letrozole. Since it has been reported that low doses of letrozole impair spatial learning and memory of intact female and male C57BL mice (Liu et al., 2019; Zhao et al., 2018), the primary focus of the current study was to evaluate the effect of similar doses of letrozole on the Morris water maze (MWM) learning and memory in ovary intact female BALB/c mice. In addition, the same mice were assessed in different spatial memory tests such as Y-maze and object location tests and in non-spatial locomotor habituation test. The rational of studying hippocampus-dependent spatial memory is also supported by the fact that in humans letrozole therapy has been reported to impair hippocampus-dependent episodic memory (Bayer et al., 2015). To target the effect of letrozole on recognition memory, another set of mice was tested in novel object recognition (NOR) test. Since the effects of systemic exposure to letrozole on NOR has not been reported, a broader range of letrozole doses was tested in this study compared to the MWM study. In addition, the effects of this dose range were evaluated in the Y-maze. Because the primary target of letrozole is to block estrogen synthesis, the effects of various doses of letrozole were also compared with OVX, which is deprived form ovarian estrogen. However, there are major differences in the endocrine milieu of intact and OVX groups. In intact animals, estrogen synthesis is not totally blocked by letrozole due to the high levels of aromatase in the ovaries. We hypothesize that the doses of letrozole used (0.1–1.0 mg/kg) are sufficient to block estrogen synthesis in the brain. If OVX and intact letrozole-treated animals behave differently in the tests aimed at evaluating cognition, the difference should be due to estrogen whose synthesis in the ovaries is not completely inhibited by the doses of letrozole used in these studies and to the elevated levels of FSH and androgens in ovary intact animals, although other, yet unknown factor(s) cannot be ruled out.

Data presented here support the hypothesis that similarly to humans, BALB/c mice exposed to letrozole develop learning and memory deficits depending on dose ranges and the specific tasks. Based on the results of this study, the BALB/c mice emerge as a preclinical model for future investigation of countermeasures against side effects of the letrozole therapy.