The effect of acute or repeated stress on the corticotropin releasing factor system in the CRH-IRES-Cre mouse: A validation study

Highlights

• The Crh-IRES-Cre::Ai14 mouse had faithful reporter expression in BNST and PVN.

• The Crh-IRES-Cre::Ai14 mouse had unfaithful reporter expression in PVT and LH.

• Acute stress stimulated Fos expression in PVN CRF neurons.

• Chronic stress followed by a novel acute stressor did not recruit CRF neurons.

Abstract

Corticotropin releasing factor (CRF) is a key component of stress responsivity, modulating related behaviors including anxiety and reward. Difficulties identifying CRF neurons, using traditional approaches including immunohistochemistry, has led to the development of a number of transgenic CRF reporter mice. The Crh-IRES-Cre::Ai14 (tdTomato) reporter mouse is increasing in popularity as a useful tool to assess the localization, connectivity and function of CRF neurons in various stress-related behaviors. However, without proper characterization of reporter expression, the in vivo and in vitro manifestations resulting from the manipulation of these cells must be interpreted with caution. Here we mapped the distribution of tdTomato-expressing CRF cells throughout the rostro-caudal extent of the Crh-IRES-Cre::Ai14 mouse brain. To determine if reporter expression faithfully reproduced native CRF expression, we assessed the colocalization of CRF expression with tdTomato reporter expression across several brain regions. Good concordance was observed in the extended amygdala and paraventricular nucleus of the hypothalamus (PVN), while discrepancies were observed within the lateral hypothalamus and hippocampus. Finally, we examined the activation of CRF neurons in Crh-IRES-Cre::Ai14 mice in response to different types of stressors using Fos immunohistochemistry. Acute psychological (swim) and pharmacological (yohimbine) stress stimulated Fos-protein expression in PVN CRF neurons. Interestingly though, exposure to four daily restraint stress sessions followed by a novel acute stressor did not further recruit CRF neurons across any brain region examined. Our results highlight the importance of thoroughly characterizing reporter mice before use and suggest that acute versus repeated stress may differentially impact the CRF system.

This article is part of the Special Issue entitled ‘Hypothalamic Control of Homeostasis’.

Introduction

Corticotropin releasing factor (CRF) is a key brain mediator of stress-related behaviors. A disruption in homeostasis of the CRF system has been linked to a variety of stress-related psychiatric disorders including depression, anxiety and addiction, all of which involve altered emotional states (Nemeroff et al., 1984, Henckens et al., 2016, Koob and Volkow, 2016). Manipulation of the CRF gene promoter has been accomplished using a variety of technologies, including bacterial artificial chromosome (BAC) transgenics (Alon et al., 2009) and genetic targeting (e.g. knock-in) (Kono et al., 2017), such as IRES-Cre (Taniguchi et al., 2011). The Crh-IRES-Cre mouse has been bred with several reporter lines including Ai3 (GFP) Ai9 and Ai14 (tdTomato), to generate mice with visualizable CRF-expressing cells (Wamsteeker Cusulin et al., 2013, Chen et al., 2015, Hooper and Maguire, 2016, Martin et al., 2010, Peng et al., 2017, Dedic et al., 2018). These neuronal populations are easily available for anatomical and morphological examination as well as electrophysiology and/or optogenetic/chemogenetic manipulation.

A major obstacle in identifying CRF neurons has historically been due to the difficulty in CRF peptide detection by immunohistochemistry using anti-CRF antibodies without additional tools, such as colchicine pre-treatment (Gunn et al., 2013, Borisy and Taylor, 1967). Colchicine is toxic at typically employed doses, leading to physical stress and neuronal activation in stress-related brain regions (Ceccatelli et al., 1989). The recent development of the Crh-IRES-Cre mouse has allowed for examination and manipulation of the CRF system and is emerging as a useful tool to examine the role of the mouse CRF system in stress-associated behaviors (Wamsteeker Cusulin et al., 2013, Dedic et al., 2018, Füzesi et al., 2016, Kondoh et al., 2016, Partridge et al., 2016). Recent studies have examined the validity of reporter expression in several key brain regions of the Crh-IRES-Cre::Ai14 and Crh-IRES-Cre::Ai9 mouse, including the bed nucleus of the stria terminalis (BNST), paraventricular nucleus of the hypothalamus (PVN), hippocampus and central nucleus of the amygdala (CeA) (Wamsteeker Cusulin et al., 2013, Chen et al., 2015, Dedic et al., 2018). Additionally, a recent study has mapped global reporter expression in the Crh-IRES-Cre::Ai3 mouse (Peng et al., 2017). However, some studies suggest that mouse lines crossed to different reporter lines show reporter expression in distinct populations or subpopulations within specific brain regions (Hooper and Maguire, 2016). For example, crossing of the BAC CRF-Cre line to either a tdTomato or mTomato-GFP reporter line led to labelling of anatomically distinct neuronal populations within the hippocampus (Hooper and Maguire, 2016). Therefore, as the Crh-IRES-Cre::Ai14 mouse has emerged as a popular mouse line to assess and manipulate the CRF system, further mapping and validation of expression is required. Thus, we mapped the distribution of tdTomato reporter in the Crh-IRES-Cre::Ai14 mouse and assessed the concordance of CRF protein and reporter expression throughout select brain regions.

Key to the role of CRF in stress responsivity is the hypothalamic–pituitary–adrenal (HPA) axis (Herman and Cullinan, 1997). CRF is released into the periphery from the PVN and directly activates the HPA axis by triggering the release of adrenocorticotropic hormone (ACTH) from the anterior pituitary, which in turn stimulates the synthesis and secretion of glucocorticoids from the adrenal gland (Vale et al., 1981). Central administration of CRF induces anxiety-like behavior in the conflict, social interaction, acoustic startle and elevated plus maze tests (Swerdlow et al., 1986, Dunn and File, 1987, File et al., 1988). Further, both physical and psychological stressors activate CRF PVN neurons (Dayas et al., 2001). CRF is also located outside the PVN to control autonomic and behavioral responses to stress. For example, CRF-like-immunoreactivity is present in the neocortex, medial prefrontal cortex (mPFC), BNST, amygdala, medial septum, hypothalamus, thalamus, cerebellum, midbrain and hindbrain nuclei (Swanson et al., 1983). Additionally, the amygdala is thought to play a primary role in mediating anxiety-like responses (Davis et al., 2003). Regulatory dysfunction in the amygdala can lead to an abnormal pattern of affective responses and the development of recurring emotional disorders such as anxiety, depression and addiction (Gilpin et al., 2015).

Previous research has shown that distinct stressors can differentially modulate CRF expression and neural activation (See McReynolds et al., 2018 for review). For example, chronic unpredictable stress, but not acute restraint stress, increases CRF-like-immunoreactivity within the periventricular nucleus and anterior hypothalamic nucleus (Chappell et al., 1986). Additionally, acute, but not chronic, stress decreases CRF-like-immunoreactivity within the medial preoptic area (MPO) (Chappell et al., 1986). The specific pattern of activation of CRF neurons following acute and repeated psychological and pharmacological stress in mouse models remains uncharacterized. Here we examined the activation of CRF neurons in hypothalamic and extrahypothalamic brain regions following exposure to acute psychological (forced swim) and pharmacological (yohimbine) stressors in mice with and without a history of repeated restraint stress.