Introduction

Corticotropin-releasing hormone (CRH)—also designated as corticotropin-releasing factor (CRF)—is important in coordinating the neuroendocrine, autonomic, behavioral and immunological responses to various stressful stimuli.1 Besides its function as the major physiological regulator of the hypothalamic–pituitary–adrenocortical (HPA) system, CRH is capable of modulating a wide range of behaviors, including anxiety-related behavior, arousal, sensory information processing, learning and memory as well as locomotor activity.2, 3, 4 Most behavioral effects of CRH are attributed to extrahypothalamic neuronal circuits including neocortical, limbic and brainstem structures where CRH functions as a neuromodulator. Dysregulation of the CRH system and accompanying chronically elevated levels of CRH are implicated in human stress-related and affective disorders, including anxiety disorders and major depression.5, 6, 7 In this line, elevated levels of CRH in the cerebrospinal fluid,5 increased numbers of CRH and CRH/arginine vasopressin-expressing neurons and elevated CRH mRNA levels in the paraventricular nucleus of the hypothalamus (PVN),8 as well as decreased CRH binding sites in the frontal cortex have been demonstrated in depressed patients. In addition, animal studies involving central application of CRH revealed phenotypic alterations reminiscent of symptoms observed in affected subjects.4 Moreover, CRH receptor antagonists are capable of attenuating the behavioral consequences of stress, underscoring the role of endogenous CRH in mediating many stress-induced behaviors. Finally, clinical trials have demonstrated the efficacy of a selective CRH type 1 receptor (CRH-R1) antagonist in treating depressed patients.9

Unlike classical antidepressants that modulate the monoaminergic system, CRH receptor antagonists have not produced comparable results in antidepressant screening paradigms in rodents, such as the forced swim test (FST) or the tail suspension test.10 In these tests animals were studied under basal conditions. However, according to the hypothesis by Hökfelt and Lundberg,11 neuropeptides, unlike monoamines, are only secreted at substantial amounts under pathological conditions or severe stress. Accordingly, anxiolytic-like activities of CRH-R1 antagonists such as DMP696 and R121919 are best observed in animals that have increased levels of CRH, and are hyperresponsive or more susceptible to stress.12 These findings are in agreement with a blunted stress hormone response in healthy humans pretreated with a CRH-R1 antagonist.13 To study the effects of central CRH hyperactivity in an animal model, CRH transgenic mouse lines were established expressing CRH either under the control of the broadly active metallothionein (CRH-Tg) or the central nervous system (CNS)-restricted Thy-1.2 (CRH-OE2122) promoter.14, 15 In both cases, unrestricted CRH overexpression resulted in elevated adrenocorticotropin (ACTH) and corticosterone levels accompanied by symptoms of Cushing-like syndrome, complicating the interpretation of stress-related behavioral results. To circumvent these problems we developed a mouse model that permits the overexpression of CRH without producing marked neuroendocrine disturbances under basal conditions. Combining the knock-in of a single copy of the murine Crh cDNA into the ROSA26 (R26) locus16 with the Cre/loxP system has enabled us to overexpress CRH in a spatio-temporally regulated fashion at different dosages. We activated CRH expression in the entire brain and in specific types of neurons within the forebrain using Nes-cre,17 Camk2a-cre18 and Dlx-cre.19 We demonstrated that this approach is an effective way to study the behavioral and neuroendocrine effects of spatially confined CRH overexpression in mice.

Materials and methods

Targeting vector

The targeting vector was based on pROSA26-1 bearing 5.5-kb homology to the murine ROSA26 locus (R26) and a diphtheria toxin (DTA) expression cassette.20 It was constructed by introducing the following components into the unique XbaI site of pROSA26-1: adenovirus splice acceptor (SA), loxP, PGK-Neo—including PGK polyadenylation sequence (pA) and two copies of the SV40 pA (PGK-Neo-3 × pA), loxP, IRES-LacZ and bovine growth hormone (bGH) pA (from 5′ to 3′). The SA and bGH pA were subcloned from pSAβgeo.21 The IRES-LacZ was isolated and modified from ETLpA-/LTNL.22 The loxP-flanked PGK-Neo-3 × pA cassette was amplified by PCR from genomic tail DNA of R26 reporter mice.20 The DTA cassette was inverted and concomitantly an SwaI site was introduced for linearization. The murine Crh cDNA was inserted in a unique PacI site between the second loxP site and the IRES-LacZ cassette. External probes used for identification of homologous recombination events were amplified by PCR from genomic DNA and cloned using the TOPO TA cloning kit (Invitrogen, Karlsruhe, Germany). 5′-Probe: forward 5′-GCG-AGA-CTC-GAG-TTA-GGC-3′ and reverse 5′-GCG-GCC-GCC-GCC-CGC-CTG-CG-3′ (150 bp); 3′-probe: forward 5′-GTT-GAG-CCA-CTG-AGA-ATG-G-3′ and reverse 5′-GAA-ACT-ACA-ACC-ATT-GTT-CAT-3′ (662 bp).

Generation of conditional CRH overexpressing mice

The linearized targeting vector was electroporated into TBV2 embryonic stem (ES) cells (129S2). Mutant ES cell clones were identified by Southern blot analysis of genomic ES cell DNA digested with EcoRV or ApaI using the external 5′- or 3′-probe respectively. Mutant ES cells were used to generate chimeric mice by blastocyst injection. Germ-line transmission of the modified R26 allele (R26flopCrh floxed stop) was confirmed in offspring from male chimeras bred to wild-type C57BL/6J mice. For the conditional, CNS-restricted overexpression of CRH (CRH-COE-Nes), obtained R26+/flopCrh mice were crossed to transgenic nestin (Nes)-cre mice.17 Resulting heterozygous R26+/flopCrh and R26+/flopCrh Nes-cre F1 animals were intercrossed to obtain in the F2 generation animals of the desired genotypes: R26+/+ (CRH-COEwt-Nes), R26flopCrh/flopCrh (CRH-COEcon-Nes), R26+/flopCrh Nes-cre (CRH-COEhet-Nes) and R26flopCrh/flopCrh Nes-cre (CRH-COEhom-Nes). For the forebrain-restricted overexpression of CRH in principal neurons (CRH-COE-Cam), R26+/flopCrh mice were crossed to transgenic Camk2a-cre mice.18 As above, R26+/+ (CRH-COEwt-Cam), R26flopCrh/flopCrh (CRH-COEcon-Cam), R26+/flopCrh Camk2a-cre (CRH-COEhet-Cam) and R26flopCrh/flopCrh Camk2a-cre (CRH-COEhom-Cam) animals were obtained in the F2 generation. CRH-COE-Dlx mice overexpressing CRH in GABAergic neurons of the forebrain were accordingly generated using Dlx5/6 (Dlx)-cre mice.19

Genotyping was performed by PCR using primers: ROSA-1, 5′-AAA-GTC-GCT-CTG-AGT-TGT-TAT-3′; ROSA-2, 5′-GCG-AAG-AGT-TTG-TCC-TCA-ACC-3′ and ROSA-4, 5′-GGA-GCG-GGA-GAA-ATG-GAT-ATG-3′. Standard PCR conditions resulted in a 398-bp wild-type and a 320-bp mutant PCR product. The presence of Nes-, Camk2a- and Dlx-cre was evaluated using primers CRE-F, 5′-GAT-CGC-TGC-CAG-GAT-ATA-CG-3′ and CRE-R 5′-AAT-CGC-CAT-CTT-CCA-GCA-G-3′ resulting in a PCR product of 574 bp. Genotypes were confirmed by Southern blot analysis of EcoRV-digested tail DNA using the 5′-probe and a Cre-recombinase-specific probe.23 The efficiency of Nes-cre-mediated excision of the transcriptional terminator sequence was demonstrated by Southern blot analysis of EcoRV-digested genomic DNA prepared from cortex, hippocampus, thalamus, cerebellum, tail and liver using the external 5′-probe. Mice used for this study were kept on a mixed 129S2/Sv × C57BL/6J background.

X-Gal staining

Animals (2- to 3-month old; n=4–5 per genotype) were killed by an overdose of isoflurane and transcardially perfused with 4% paraformaldehyde, 2 mM MgSO4 and 5 mM EGTA. Subsequent X-Gal staining was performed on free floating 50- or 100-μm thick vibratome sections or on intact organs as previously described.22

In situ hybridization

Mice (10-week old) were killed in the morning (10:00 am) by an overdose of isoflurane. For quantification of immediate early genes (IEGs) c-fos and zif268 animals were either killed under basal conditions or subjected to 10 min of forced swimming 30 min before killing. Brains were carefully removed and immediately shock-frozen on dry ice. Frozen brains were cut on a cryostat in 20-μm thick sections. For quantitative in situ hybridization cryostat sections of CRH-COEcon and CRH-COEhom brains were mounted side by side on SuperFrost Plus slides (Menzel GmbH, Braunschweig, Germany). This procedure allowed for parallel in situ hybridization of sections under identical conditions assuring meaningful quantification and comparison of hybridization signals. All sections were processed for in situ hybridization according to a modified version of the procedure described by Dagerlind et al.24 The following riboprobes were used: CRH, nucleotides 1306–1661 of GenBank accession no. AY128673; endogenous CRH (3′-UTR), nucleotides 1816–2107 of GenBank accession no. AY128673; CRH-R1, nucleotides 1570–2273 of GenBank accession no. NM_007762; c-fos, nucleotides 608–978 of GenBank accession no. NM_010234; zif268, nucleotides 245–786 of GenBank accession no. NM_007913 and LacZ, nucleotides 192–569 of GenBank accession no. U46489. Specific riboprobes were generated by PCR applying T7 and T3 or SP6 primers using plasmids containing above-mentioned cDNAs as templates. Antisense and sense cRNA probes were transcribed from 200 ng of respective PCR product and directly used as a template for the synthesis of radiolabeled transcripts by in-vitro transcription with 35S-UTP (Amersham Biosciences, Piscataway, NJ, USA) using T7 and T3 RNA polymerase (Roche, Penzberg, Germany), respectively. After 20 min of DNase I (Roche) treatment, the probes were purified by the RNeasy Clean up protocol (Qiagen, Hilden, Germany) and measured in a scintillation counter. For hybridization, sections were pretreated and prehybridized as previously described.24 Subsequently, they were hybridized overnight with a probe concentration of 7 × 1 06 c.p.m. ml−1 at 57 °C and washed at 64 °C in 0.1 × saline sodium citrate (SSC) and 0.1 mM dithiothreitol. The hybridized slides were dipped in autoradiographic emulsion (type NTB2; Eastman Kodak, Rochester, NY, USA), developed after 3–6 weeks and counterstained with cresyl violet.

For quantification, autoradiographs were digitized and relative levels of mRNA were determined by computer-assisted optical densitometry (ImageJ; http://rsb.info.nih.gov/ij/). For in situ hybridizations routinely three different exposure times were applied to assure that the signals to be quantified were in the linear range.

CRH radioimmunoassay

Mice (2- to 3-month old) were killed by cervical dislocation at 10:00 am. Brains were carefully removed and used in total or further dissected for selective preparation of cortex, hippocampus and thalamus. The CRH-specific radioimmunoassay (RIA) on tissue homogenates was performed after prior extraction as previously described.25

Endocrine analyses

Two weeks before the experiments, 3- to 5-month-old animals were separated and singly housed with a 12:12 h light:dark schedule (lights off at 07:00 pm). All experiments and data analyses were performed separately for male and female animals. To determine the basal hormone plasma levels, mice were left undisturbed throughout the night before the experiment. Blood sampling was performed in the early morning (07:30–09:30 am) and afternoon (04:30–05:30 pm) by collecting trunk blood from animals rapidly decapitated under light isoflurane anesthesia or by incision of the tail, with the time from first handling of the animal to completion of bleeding not exceeding 45 s. For evaluation of the endocrine response to stress, we collected blood samples immediately after and 30 min after 10-min restraint stress, for which animals were placed in a 50-ml conical tube with the bottom removed. Stress experiments were performed in the morning (07:30–10:00 am). Plasma corticosterone and ACTH concentrations were measured in duplicate by commercially available RIA kits (ICN Biomedicals, Irvine, CA, USA).

Subjects for behavioral testing

Mice were singly housed 2 weeks before experiments under standard laboratory conditions (22±1 °C, 55±5% humidity) with food and water ad libitum under a 12:12 h inverted light:dark schedule (lights off at 09:00 am). Age of tested animals ranged between 3 and 6 months. Animal experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of the Government of Bavaria. Experiments were performed during the dark, active phase of the animals between 01:00 pm and 06:00 pm hours under red light conditions, unless otherwise stated. Animals' behavior in the tail suspension and FST was analyzed online by trained observers who were blind to treatment and genotype. If not stated otherwise, male mice were used for the experiments. Experiments to assess the behavioral consequences of stress-regulated activation of endogenous CRH were performed on previously described CRH-R1 knockout mice,26 which were kept on a mixed 129/Ola × CD1 background.

Open field

Animals were tested under red light in an open field (26 × 26 × 38 cm high) made of white floor and clear plastic walls, and equipped with infrared photocell sensors. Testing lasted for 30 min. Distance travelled and rearings were measured using the Tru Scan Software version 1.1 A (Coulbourn Instruments, Allentown, PA, USA) (sampling frequency, 4 Hz).

Forced swim test

Each mouse was placed in a 5-l glass beaker (height, 23.5 cm; diameter, 16.5 cm) containing 15 cm of water at 25±1 °C or 32±1 °C for 6 min. The water was changed between subjects. During each trial, floating (immobility) and struggling time was scored by pressing preset keys on a computer keyboard, using a customized freeware software (EVENTLOG). The resulting λ-channel ethogram was further processed by customized software (Winrat version 2.31; Heinz Barthelmes, MPI Munich, Germany). A mouse was judged floating when it stopped any movements except those that were necessary to keep its head above water. Vigorous swimming movements involving all four limbs of the mouse with the front paws breaking the surface of the water, usually at the walls of the beaker, were regarded as struggling. The test was carried out on 2 consecutive days.

For the analysis of the influence of restraint stress experience on FST behavior, animals were subjected to 1 h of restraint stress in a 50-ml conical tube 4 h before exposure to forced swimming according to a previously established protocol for stress-induced activation of CRH-R1.27

Tail suspension test

Animals were suspended by the end of their tail with adhesive tape to a steel bar that was 35 cm above the floor. Each session lasted 6 min and was videotaped. The duration of immobility was scored by a trained observer who was blind to the animals' genotype, using EVENTLOG software. Mice were considered immobile only when they hung passively without moving the limbs or the head.

Pharmacology

All drugs were freshly prepared in a volume of 10 ml kg−1. DMP696 (Bristol-Myers Squibb, Munich, Germany) was suspended in a 0.9% saline solution containing 5% dimethyl sulfoxide, 5% polyethylene glycol 400 and 10 μl of Tween 80 per 1.5 ml (all chemicals from Sigma-Aldrich, Steinheim, Germany), and injected i.p. 1 h before the FST on days 1 and 2. For the analysis of DMP696 influences on stress-induced ACTH levels, mice were decapitated immediately after the FST on day 2.

Para-chlorophenylalanine methyl ester (PCPA) and α-methyl-para-tyrosine methyl ester (AMPT) (all drugs from Sigma-Aldrich, Steinheim, Germany) were suspended in a 0.9% saline solution containing 1% dimethyl sulfoxide and administered i.p. PCPA (250 mg per kg) was administered twice daily (every 12 h) for 3 days with the last dose given 18 h before the FST.28 AMPT (200 mg per kg) was administered as a single dose 4 h before the FST.28

Neurochemical analysis

Mice treated with AMPT and PCPA were killed directly after the FST. Brains were freshly dissected on ice and hippocampi removed, weighed and stored at −80 °C until analysis. Hippocampal tissue samples were diluted 20-fold w/v with HCl (0.1 M) in an ice bath and homogenized by sonication (40 s, 75% duty cycle, 3.5 micro tip limit; Branson sonifier 250, Sonic Power Company, Danbury, CT, USA). Homogenized tissue was then ultracentrifuged (35 000 g, 20 min, 4 °C; Beckman L-60 Ultracentrifuge, Munich, Germany), and the resulting aqueous layer filtered (0.22 μm × 13 mm, polyvinylidene fluoride, Millex Filters; Millipore, Bedford, TX, USA). Filtrate was aliquoted in 20 μl samples into separate Eppendorf tubes and stored at −80 °C until further analysis.

Determination of serotonin (5-hydroxytryptamine), 5-hydroxyindoleacetic acid, 3,4-dihydrophenylacetic acid and 4-hydroxy-3-methoxyphenylacetic acid was performed by reverse-phase high-performance liquid chromatography with electrochemical detection as previously described.29 Hippocampal filtrates (20 μl stock solutions) were diluted and 50 μl were automatically injected by a CMA 200 refrigerated autosampler (CMA Microdialysis AB, Stockholm, Sweden). The mobile phase consisted of 93% phosphate buffer (0.1 M NaH2PO4, 1 mM sodium octanesulphonic acid, 10 mM NaCl and 0.5 mM Na2-EDTA) and 7% acetonitrile, and the pH was adjusted to 4.0 with o-phosphoric acid (all chemicals were from Merck, Darmstadt, Germany). For noradrenaline (NA) and dopamine (DA) tissue quantification, hippocampal filtrates (20 μl stock solutions) were diluted and concentration was determined by a radioenzymatic assay as previously described. This assay involves COMT-catalyzed O-methylation using [3H]S-adenosylmethionine as methyl donor and separation of the resulting [3H]normetanephrine by thin-layer chromatography (TLC).

Statistical analysis

Data were analyzed for multiple comparisons using one-, two- or three-way analyses of variance (ANOVAs) for repeated measures where appropriate, followed by post-hoc Newman–Keuls multiple comparison test. For two-group comparisons unpaired Student's t-test was used. Differences were considered statistically significant when P<0.05. Data are presented as mean±s.e.m.

Results

Generation of mutant mice conditionally overexpressing CRH in the CNS (CRH-COE-Nes)

We used homologous recombination in ES cells to target the ubiquitously expressed ROSA26 (R26) locus with a single copy of the murine Crh cDNA preceded by a loxP-flanked (floxed) transcriptional ‘Stop’ sequence (Figures 1a–c). As previously reported,20 mice homozygous for the modified R26 allele (R26flopCrh/flopCrh floxed Stop), which is Cre-recombinase-sensitive, were indistinguishable from wild-type littermates, behaviorally (data not shown), as well as with respect to endogenous CRH mRNA (Figures 2e and f) and protein levels (Figure 3a). Homozygous R26flopCrh/flopCrh mice were crossed to transgenic Nes-cre mice17 allowing for a CNS-restricted overexpression of CRH in double transgenic animals (CRH-COE-Nes). In this mouse line Cre expression is controlled by the nestin promoter and neural enhancer, which drive cre expression in neuronal and glial precursors as early as embryonic day 10.5.30 In the F2 generation we obtained R26+/+, R26flopCrh/flopCrhR26+/flopCrh Nes-cre and R26flopCrh/flopCrh Nes-cre mice (Figure 1d), which we will refer to as CRH-COEwt-Nes, CRH-COEcon-Nes, CRH-COEhet-Nes and CRH-COEhom-Nes, respectively. On the genomic level, Cre-mediated deletion of the transcriptional terminator sequence was observed only in the CNS, but not in peripheral organs of CRH-COEhom-Nes mice (Figure 1e).

Figure 1
figure 1

Generation of mice overexpressing corticotropin-releasing hormone (CRH) restricted to the central nervous system (CNS). (a) Strategy for conditional, Cre-mediated expression of CRH from the R26 locus. Partial restriction maps of wild-type R26 locus, targeting vector, recombined R26flopCrh allele and activated R26Crh allele (WT, wild-type fragment; MT, mutant fragment following homologous recombination; DEL, deletion fragment resulting from Cre-mediated excision of the ‘Stop’ cassette; A, ApaI; E, EcoRV; S, SwaI; S, splice acceptor; loxP sites are indicated as black arrowheads). (b) Southern blot analysis of wild-type and targeted embryonic stem (ES) cell clones. The R26 5′-probe was hybridized to EcoRV-digested genomic ES cell DNA. The targeted allele was indicated by the presence of an additional mutant 4.1-kb fragment. (c) The R26 3′-probe was hybridized to ApaI-digested DNA from the same ES cell clones confirming homologous recombination by detection of an additional mutant fragment at 11.4 kb. (d) Southern blot analysis of EcoRV-digested tail DNA of CRH-COE-Nes mice simultaneously hybridized with the 5′-probe and a Cre-recombinase-specific probe. The hybridizing fragments obtained correspond to the indicated genotypes. (e) Southern blot analysis of EcoRV-digested genomic DNA from various tissues of a CRH-COEhom-Nes animal hybridized with the 5′-probe, showing the extent of Cre-mediated deletion of the transcriptional terminator sequence as indicated by the presence of an additional 5.2-kb fragment.

Full size image
Figure 2
figure 2

Verification of the central nervous system (CNS)-restricted overexpression of corticotropin-releasing hormone (CRH) in CRH-COE-Nes mice. Intact organs of (a) control, (b) heterozygous and (c) homozygous CRH-COE-Nes mice were stained overnight with X-Gal. (left: brain with spinal cord; right, from top: liver, lung, heart, spleen, testis with epididymis; inlay: adrenal gland (left) and pituitary (right). Note the background staining of the epididymis. Intense blue staining in the brain and spinal cord of CRH-COEhet- and CRH-COEhom-Nes mice reflects the brain-specific overexpression of the IRES-LacZ reporter gene. No staining was observed in peripheral organs of CRH-COE-Nes mice. Determination of IRES-LacZ reporter gene expression by X-Gal staining on coronal sections of (d) CRH-COEcon-, (g) CRH-COEhet- and (j) CRH-COEhom-Nes mice. Crh overexpression was demonstrated by in situ hybridization using a specific radiolabeled riboprobe detecting both endogenous and exogenous CRH expression. Representative dark-field photomicrographs of coronal and sagittal brain sections of CRH-COE-Nes mice are depicted. (e, f) CRH-COEcon-Nes mice display the characteristic, heterogeneous Crh expression throughout the entire CNS with strong expression in the paraventricular nucleus (PVN) of the hypothalamus, central nucleus of the amygdala (CeA), bed nucleus of stria terminalis (BST), olfactory bulb (OB) and nuclei of the brain stem. In addition, CRH-expressing neurons are found scattered within the cortex (CX) and hippocampus (HIP). In (h, i) CRH-COEhet- and (k, l) CRH-COEhom-Nes mice exogenous CRH is expressed throughout the brain corresponding to the pattern of IRES-LacZ reporter gene expression. The level of exogenous Crh mRNA expression is gene dosage-dependent as demonstrated by the stronger in situ hybridization signals detected in CRH-COEhom- versus CRH-COEhet-Nes animals (ac, anterior commissure).

Full size image
Figure 3
figure 3

Corticotropin-releasing hormone (CRH) overexpression from the R26 locus results in a gene dosage-dependent increase of CRH protein content in the (a) entire brain, (b) cortex, (c) hippocampus, (d) thalamus and (e) cerebellum of male CRH-COEhet-Nes (het) and CRH-COEhom-Nes (hom) mice in comparison to CRH-COEcon-Nes (con) mice (n=5–12). The CRH content in the brain of CRH-COEcon-Nes mice is indistinguishable from wild-type CRH-COEwt-Nes (wt) mice. CRH content is given as pg mg−1 tissue wet weight. *P<0.05, **P<0.01, ***P<0.001.

Full size image

Verification of the CNS-restricted overexpression of CRH in CRH-COE-Nes mice

To assess the spatial distribution of exogenous CRH expression in detail, we made use of the introduced IRES-LacZ reporter gene, which is co-activated upon Cre-mediated excision of the transcriptional terminator sequence (Figure 1a). X-Gal staining of intact organs revealed the absence of any specific staining in CRH-COEcon-Nes mice (Figure 2a). CRH-COEhet-Nes and CRH-COEhom-Nes mice exhibited an intense staining in the brain and spinal cord whereas peripheral organs were devoid of LacZ-dependent staining. Increased staining intensities in the intact brain and spinal cord as well as on coronal brain sections of CRH-COEhom-Nes compared to CRH-COEhet-Nes animals reflected the assumed gene dosage effect (Figures 2b, c, g and j). The conditional overexpression of CRH was verified by in situ hybridization using a CRH-specific riboprobe. In CRH-COEcon-Nes mice, endogenous CRH expression was detected heterogeneously throughout the entire CNS as previously described31 (Figures 2e and f). In CRH-COEhet- and CRH-COEhom-Nes mice, the pattern of CRH induction paralleled the activation of the IRES-LacZ reporter gene as demonstrated by X-Gal staining. Expression of exogenous CRH was detected at varying levels throughout the brain and attributed to the CNS-wide expression of Nes-cre and to the ubiquitous activity of the R26 locus. CRH mRNA was detected at highest levels in the olfactory bulb, cortex and hippocampus, again in a gene dosage-dependent manner (Figures 2h, i, k and l). Moreover, we confirmed the increased CRH peptide content in the entire brain and various brain areas of CRH-COEhet-Nes and CRH-COEhom-Nes mice using a CRH-specific RIA (Figures 3a–e). No difference in CRH peptide content was observed between CRH-COEwt- and CRH-COEcon-Nes mice (Figure 3a).

Expression of endogenous CRH and CRH-R1 is altered in the brain of CRH-COE-Nes mice

To explore alterations in the expression of endogenous CRH and CRH-R1 in response to exogenous CRH expressed from the R26 locus, we used in situ hybridization. We observed a strong decrease of endogenous CRH mRNA levels in the PVN (Supplementary Figure 1c), in all areas of the hippocampus proper (CA1-CA3) and in the dentate gyrus (Supplementary Figure 1a) of CRH-COEhom-Nes mice compared to control littermates. However, in the CeA (Supplementary Figure 1b) endogenous CRH mRNA levels were significantly increased. CRH-COEhom-Nes mice exhibited increased CRH-R1 mRNA levels in the hippocampus and dentate gyrus (Supplementary Figure 1d). Surprisingly, CRH-R1 mRNA levels in the hippocampus were only increased in the CA2 and CA3, however not in CA1 (Supplementary Figure 1d). In addition, CRH-R1 expression was significantly increased in the basolateral amygdala of homozygous CRH overexpressing mice (Supplementary Figure 1e).

HPA axis of male but not female CRH-COE-Nes mice is hypersensitive to stress

Basal plasma ACTH and corticosterone levels did not differ significantly between male CRH-COEcon-, CRH-COEhet- and CRH-COEhom-Nes mice over the circadian cycle, neither at the diurnal trough, nor at the diurnal peak (Supplementary Figures 2a and c). To examine the response of the HPA axis to stress, animals were subjected to 10 min of restraint stress in the morning and killed either directly or 30 min after the end of the stressor (Supplementary Figures 2b and d). Restraint stress resulted in significantly elevated ACTH (F2,32=7.5; P<0.01) and corticosterone (F2,32=4.10; P<0.05) levels in male CRH-COEhet- and CRH-COEhom-Nes mice compared to CRH-COEcon-Nes littermates. In CRH-COEhom-Nes mice, corticosterone levels remained significantly elevated 30 min after stress compared to CRH-COEcon-Nes and CRH-COEhet-Nes mice (F2,27=4.10; P<0.05).

Similar to males, female CRH-COEhom-Nes animals showed no difference in basal ACTH and corticosterone levels compared to CRH-COEcon-Nes mice (Supplementary Figures 2e and f). However, in contrast to males, female CRH-COEhom-Nes mice also did not show elevated ACTH and corticosterone secretion in response to restraint stress in comparison with CRH-COEcon-Nes mice (Supplementary Figures 2e and f).

CNS-restricted overexpression of CRH results in decreased immobility in the FST and the tail suspension test

In the open field, CRH-COEhom mice showed no difference in locomotor activity compared to CRH-COEcon littermates (Figure 4a). CRH-COEhom mice did show increased exploratory behaviors compared to CRH-COEcon littermates as indicated by an elevated number of vertical movements (rearings) (t18=2.14, P<0.05; Figure 4a). To examine stress-coping behaviors, CRH-COE-Nes mice were exposed to the FST. In male mice, CRH overexpression resulted in a gene dosage-dependent decrease in floating in the FST both on day 1 (F2,30=20.0; P<0.001) and on day 2 (F2,30=15.0; P<0.001) with both CRH-COEhom-Nes and CRH-COEhet-Nes mice floating significantly less than CRH-COEcon-Nes littermates (Figure 4b). Furthermore, CRH overexpression resulted in a dose-dependent increase in struggling both on day 1 (F2,30=3.4; P<0.05) and on day 2 (F2,30=4.3; P<0.05) with CRH-COEhom-Nes mice struggling generally more than CRH-COEcon-Nes littermates (data not shown). Female CRH-COEhom-Nes mice showed essentially the same phenotype in the FST as males: they floated significantly less than their CRH-COEcon-Nes littermates both on days 1 and 2 (Supplementary Figure 3a).

Figure 4
figure 4

Corticotropin-releasing hormone (CRH) overexpression leads to an increase in active stress-coping behavior in antidepressant screening paradigms. (a) During a 30 min exposure to an open field, male CRH-COEhom-Nes (hom) mice showed no difference in total distance moved compared to CRH-COEcon-Nes (con) littermates (left panel). However, CRH-COEhom mice showed increased vertical movements (rearings; right panel, n=8–12 animals per genotype). (b) In the forced swim test (FST; 25 °C) CRH overexpression resulted in a dose-dependent decrease in floating both on day 1 (d1) and on day 2 (d2) with both male CRH-COEhet- (het) and CRH-COEhom-Nes (hom) mice floating significantly less than their male CRH-COEcon-Nes (con) littermates (n=9–15 animals per genotype). (c) In the tail suspension test male CRH-COEhom-Nes mice showed significantly decreased immobility compared to male CRH-COEcon- and CRH-COEhet-Nes littermates. Data were collapsed from two independent experiments, which revealed essentially the same results, resulting in n=27–30 mice per genotype and protocol. (d) DMP696 (applied at 10 and 50 mg per kg, i.p. 1 h before testing on days 1 and 2) dose-dependently increased floating behavior of CRH-COEhet-Nes mice compared to vehicle-treated controls (0 mg per kg) both on day 1 (d1) and on day 2 (d2) (n=12–13 animals per group). (e) DMP696 was able to attenuate the swim stress-induced increase of adrenocorticotropin (ACTH) on day 2 in CRH-COEhet mice at doses of 10 and 50 mg per kg (n=5–6 animals per group). (f) Previous exposure to restraint stress resulted in decreased floating behavior in the forced swim test in CRH type 1 receptor (CRH-R1) wild-type (CRH-R1+/+) mice compared to unstressed controls and stressed CRH-R1 knockout (CRH-R1−/−) animals on day 1. Restraint stress failed to significantly affect floating behavior of CRH-R1−/− animals (n=13–16 animals per group). *P<0.05, **P<0.01, ***P<0.001.

Full size image

Exposure to cold water (25 °C) in the FST leads to fast body cooling in mice. In order to exclude that the phenotype of CRH-COEhet- and CRH-COEhom-Nes mice in the FST did not solely derive from a faster reduction of body temperature compared to CRH-COEcon-Nes littermates, we repeated the FST at 32 °C water temperature. Also at 32 °C, CRH-COEhom-Nes mice floated significantly less than CRH-COEcon-Nes littermates (Supplementary Figure 3b).

To further corroborate the finding that CRH promotes a reduction of immobility in the FST, we subjected male animals to the tail suspension test (TST).32 Again, CRH-COEhom-Nes mice displayed significantly decreased immobility compared to CRH-COEcon-Nes and CRH-COEhet-Nes littermates (F2,83=8.9; P<0.001), no difference was observed between CRH-COEhet-Nes and CRH-COEcon-Nes animals (Figure 4c).

Effects of CRH overexpression on FST behavior can be reversed by the selective CRH-R1 antagonist DMP696

To ascertain that the behavioral as well as the neuroendocrine consequences of CRH overexpression rather relate to acute effects of CRH than to long-term changes in brain physiology caused by the life-long overexpression of CRH, we treated male CRH-COEhet mice with the selective CRH-R1 antagonist DMP696.33 Antagonist treatment dose-dependently reversed the FST phenotype by increasing floating time both on day 1 (F2,35=4.4; P<0.05; Figure 4d) and on day 2 (F2,35=4.4; P<0.05; Figure 4d). Furthermore, DMP696 attenuated the swim stress-induced hyperactivation of the HPA axis (F2,14=10.6; P<0.01; Figure 4e).

Conditional CRH overexpression mimics the behavioral consequences of stress-induced activation of the endogenous CRH system

To assess whether overexpression of exogenous CRH mimics the behavioral consequences of stress-regulated activation of endogenous CRH, we analyzed FST behavior of CRH-R1 knockout (CRH-R1−/−) mice26 and wild-type littermates (CRH-R1+/+) without or with prior 1-h restraint stress. In accordance with our previous findings,34, 35 ablation of CRH-R1 failed to affect floating behavior in naive animals, thus arguing against a general involvement of endogenous CRH in FST behavior. However, prior restraint stress resulted in a decrease in floating in CRH-R1+/+ mice compared to nonstressed CRH-R1+/+ controls (t27=3.7, P<0.01) that was similar to the behavior of CRH-COEhet-Nes and CRH-COEhom-Nes mice (Figure 4f). The floating response was also significantly smaller than that of restraint stressed CRH-R1−/− mice (t29=3.1, P<0.01), which remained unaffected by the stressor (t27=1.2, P<0.243). These results imply that the CRH system must be activated by prior stressor exposure before endogenous CRH may modulate behavior in the FST via CRH-R1.

Forebrain-restricted overexpression of CRH in principal neurons or GABAergic interneurons does not affect stress-coping behavior in the FST

In order to specify the brain regions and neurochemical substrates involved in the immobility-reducing effect of exogenous CRH, we exploited the properties of the conditional R26flopCrh allele. Breeding CRH-COEcon mice to Camk2a-cre18 and Dlx-cre19 mice restricted the CRH overexpression to principal neurons or GABAergic interneurons respectively, of the anterior forebrain including limbic brain structures. The Dlx5/6 promoter drives cre expression as early as embryonic day 10.5 (JL Rubenstein and M Ekker, unpublished data), whereas the Camk2a promoter drives cre expression from around postnatal day 15.18 Analogous to CRH-COE-Nes mice we obtained desired control (CRH-COEcon-Cam/-Dlx), heterozygous (CRH-COEhet-Cam/-Dlx) and homozygous (CRH-COEhom-Cam/-Dlx) CRH-COE-Cam/-Dlx mice, respectively in the F2 generation. In situ hybridization confirmed the forebrain-restricted expression of exogenous CRH in both overexpressing mouse lines (Figures 5a, c and e). The function of the HPA axis of homozygous CRH-COE-Cam/-Dlx mice was not significantly altered compared to control mice, neither under basal nor under stress conditions (Supplementary Figure 4).

Figure 5
figure 5

Forebrain-restricted overexpression of corticotropin-releasing hormone (CRH) does not affect forced swimming behavior. In situ hybridization using a LacZ-specific riboprobe, which detects the CRH-LacZ fusion transcript, confirmed exogenous CRH expression in the forebrain of CRH-COEhom-Cam and CRH-COEhom-Dlx mice. (a) Sagittal section of a CRH-COEhom-Cam mouse brain shows strong overexpression of CRH in the olfactory bulb, all cortical layers, hippocampus and striatum. (b) In the forced swim test floating behavior on days 1 (d1) and 2 (d2) did not differ significantly between male CRH-COEcon-Cam (con) and CRH-COEhom-Cam (hom) mice (n=10–13 animals per genotype). (c) Sagittal section of a CRH-COEhom-Dlx mouse brain shows strong overexpression of CRH in the olfactory bulb, striatum, reticular nucleus, cortical layers and hippocampus. (d) In the forced swim test floating behavior on days 1 (d1) and 2 (d2) did not differ significantly between male CRH-COEcon-Dlx (con) and CRH-COEhom-Dlx (hom) mice (n=11 animals per genotype). (e) Sagittal section of a CRH-COEwt-Cam mouse brain demonstrating background in situ hybridization signals of the LacZ-specific riboprobe. Sections of CRH-COEwt-Dlx mice exhibited identical background hybridization signals (data not shown).

Full size image

To examine whether forebrain-restricted expression of CRH is sufficient to recapitulate the immobility-reducing effect of exogenous CRH expressed throughout the CNS in CRH-COE-Nes mice, male CRH-COEhom-Cam and CRH-COEhom-Dlx mice were exposed to the FST. Forebrain-restricted CRH overexpression failed to significantly affect floating or struggling behavior (data not shown) in homozygous CRH-COE-Cam (Figure 5b) or CRH-COE-Dlx (Figure 5d) mice compared to control littermates, both on days 1 and 2.

Active stress-coping behavior of CRH overexpressing mice depends on catecholaminergic transmission

CRH-COE-Cam as well as CRH-COE-Dlx mice suggest an involvement of more caudal brain nuclei within the mid/hind brain including monoaminergic cell populations overexpressing CRH, which promote the reduced immobility in the FST and TST. In an attempt to understand the neurochemical mechanisms underlying the behavior of CRH-COE-Nes mice in the FST, animals were pretreated with either the tryptophan hydroxylase inhibitor PCPA or the tyrosine hydroxylase inhibitor AMPT before testing in the FST. PCPA pretreatment reduced hippocampal serotonin levels by 85% in CRH-COEcon-Nes mice and by 71% in CRH-COEhom-Nes mice (data not shown). Catecholamine levels were not affected. Blockade of tryptophane hydroxylase failed to significantly affect floating behavior of CRH-COEcon-Nes and CRH-COEhom-Nes mice in the FST (Figure 6a; statistics not shown). In accordance with our previous findings in naive animals (compare Figure 4b), vehicle-treated CRH-COEhom-Nes mice floated less than vehicle-treated CRH-COEcon-Nes littermates.

Figure 6
figure 6

Active stress-coping behavior of CRH-COEhom-Nes mice is partly mediated by increased catecholaminergic neurotransmission, which originates from corticotropin-releasing hormone (CRH)-mediated hyper-activation of the locus coeruleus. (a) Pretreatment of male CRH-COE-Nes mice with the tryptophane hydroxylase inhibitor Para-chlorophenylalanine methyl ester (PCPA) did not affect floating behavior of male CRH-COEcon- or CRH-COEhom-Nes mice in the forced swim test (25 °C). (b) Inhibiting tyrosine hydroxylase by α-methyl-para-tyrosine methyl ester (AMPT) pretreatment led to a significant increase in floating behavior of CRH-COEhom-Nes mice, whereas floating behavior of CRH-COEcon-Nes remained unaffected (n=7–10 animals per group). (c–g) Forced swim stress induced a significantly stronger expression of c-fos and zif268 mRNA in the locus coeruleus of CRH-COEhom-Nes (hom) mice than in CRH-COEcon-Nes (con) mice. Representative dark-field photomicrographs of in situ hybridizations of coronal brain sections depicting (c) c-fos and (e) zif268 mRNA expression in the locus coeruleus of male CRH-COEcon-Nes (con) and CRH-COEhom-Nes (hom) mice. Quantification of respective (d) c-fos and (f) zif268 mRNA expression in the locus coeruleus as determined by in situ hybridization. (g) Quantification of zif268 mRNA expression in the locus coeruleus of female CRH-COE-Nes mice as determined by in situ hybridization. *P<0.05, **P<0.01, ***P<0.001.

Full size image

AMPT pretreatment reduced hippocampal NA levels by 35% in CRH-COEcon-Nes mice and by 43% in CRH-COEhom-Nes mice (data not shown). DA levels were reduced by 38% in CRH-COEcon-Nes mice and by 47% in CRH-COEhom-Nes mice. Serotonin levels were not affected. Blockade of tyrosine hydroxylase failed to affect floating behavior of CRH-COEcon-Nes mice, but led to a significant increase of floating in CRH-COEhom-Nes mice (t13=2.5, P<0.05; Figure 6b), suggesting that decreased floating in CRH-COEhom-Nes mice is at least partly mediated by increased catecholaminergic neurotransmission in these animals. Vehicle-treated CRH-COEhom-Nes mice floated less than vehicle-treated CRH-COEcon-Nes littermates, according to our previous findings in naive animals (compare Figure 4b).

CRH overexpression enhances FST-mediated activation of the locus coeruleus

To determine to what extent CRH overexpression activates catecholaminergic neurons, we analyzed by in situ hybridization the transcript levels of IEGs c-fos and zif268 with focus on the locus coeruleus (LC), nucleus of the solitary tract, ventral tegmental area (VTA), substantia nigra (SN) and dorsal raphe nucleus (DRN). As previously demonstrated in the rat, expression of c-fos and zif268 in these nuclei was undetectable under basal conditions (data not shown).36 However, 30 min post forced swim stress, a marked increase of c-fos and zif268 expression was detected in the LC (Figures 6c and e) but not in the VTA and SN or DRN (data not shown). Quantification of the signals revealed a stronger increase of c-fos (1.35-fold; t6=4.24, P<0.01) and zif268 (1.55-fold; t6=6.15, P<0.001) transcript levels in male CRH-COEhom-Nes mice than in CRH-COEcon-Nes mice (Figures 6d and f), suggesting an enhanced stress-dependent activation of the LC due to CRH overexpression.

To assess whether LC hyperactivation is independent of enhanced HPA axis reactivity, we also analyzed zif268 expression in female CRH-COEhom-Nes mice that do not show HPA axis hypersensitivity (compare Supplementary Figures 2e and f). Similar to males, female CRH-COEhom-Nes mice showed a significantly stronger increase of zif268 expression (1.30-fold; t10=3.12, P<0.05) in the LC compared to CRH-COEcon-Nes mice 30 min post forced swim stress (Figure 6g).

Discussion

To genetically dissect the impact of excess CRH in distinct brain regions and neuronal cell populations on stress-coping behavior, we used a knock-in approach to generate a mouse model that allowed CRH overexpression at different levels in a spatially restricted manner. Superior to standard transgenesis this conditional mouse model provides the opportunity to generate and compare different CRH overexpressing mouse lines—as demonstrated by breeding to Nes-, Camk2a- and Dlx-cre mice—avoiding common uncertainties of transgene production such as copy number or site of transgene insertion. Although the pattern of CRH overexpression exclusively depends on the spatial and/or temporal properties of the introduced Cre recombinase, the transcriptional control via R26 guarantees for identical expression levels.

This approach enabled us to specifically investigate the CNS effects of different dosages of CRH in CRH-COE mice without affecting the peripheral CRH system or the circadian HPA axis regulation under basal conditions. Nevertheless, chronic overexpression of exogenous CRH activates compensatory mechanisms affecting the expression levels of endogenous CRH and CRH-R1 in a brain region-specific manner as demonstrated in CRH-COE-Nes mice. Alterations of endogenous CRH and CRH-R1 levels will mutually interfere with existing regulatory circuits and, in concert with effects of exogenous CRH expression, add another layer of complexity. For instance, expression of CRH in the PVN of CRH-COEhom-Nes is significantly decreased under basal conditions and thereby probably sensitizes or upregulates CRH-R1 in pituitary corticotrophs. As a consequence, basal ACTH and corticosterone plasma levels of male CRH-COEhet- and CRH-COEhom-Nes mice are indistinguishable from those of control littermates, whereas the HPA axis of these animals is hyperreactive in response to stress. Interestingly, female CRH-COEhom-Nes mice displayed no such stress-dependent HPA axis hyperreactivity, supporting previous observations of gender differences in biological functions of the endogenous CRH system.37 In CRH-COEhom-Nes mice we observed increased expression of CRH in the CeA as well as of CRH-R1 in the BLA and hippocampus. In contrast, Thy-1-driven overexpression in projection neurons of CRH-OE2122 mice results in a rather uniform downregulation of CRH-R1 in several brain nuclei.38 However, it is of notice that chronically increased corticosterone levels15 might here dominate the effects on CRH-R1 expression compared to transgene CRH expression.

Male CRH-COE-Nes mice exhibited a marked gene dosage-dependent reduction of immobility in the FST and TST, which is not due to excessive stress hormone secretion, as female CRH-COE-Nes mice, which displayed normal HPA axis reactivity, showed similar behavioral alterations in the FST. Accordingly, neither CRH-R1 deletion35 nor adrenalectomy resulted in altered forced swimming behavior.39 In line with the observations in our chronic model of CRH excess, also the acute intracerebroventricular (i.c.v.) injection of CRH or cortagine, a potent CRH-R1 agonist, as well as site-directed injection of CRH into the LC, decreases immobility in the FST in rats and mice.40, 41, 42 Furthermore, CRH-Tg mice also showed reduced immobility in the FST.43 Interestingly, CRH has also been demonstrated to elicit antidepressant-like effects in a differential reinforcement of low-rate schedule in rats.44

The Porsolt FST is highly predictive for clinically effective antidepressants targeting the monoaminergic system, however, the construct of reduced immobility is not understood.45 Immobility is thought to reflect either a failure of persistence in escape-directed behavior (that is, behavioral despair) or the development of passive behavior that disengages the animal from active forms of coping with stressful stimuli.46 However, others have suggested that immobility could also reflect an adaptive mechanism to conserve energy.47 Besides, the FST certainly also involves a strong component of arousal, alertness and stress-coping behavior. Therefore, decreased immobility in CRH-COE-Nes mice might rather reflect enhanced responsiveness to a stressful situation resulting in increased active coping behavior than antidepressant-like behavior. CRH has been demonstrated to influence neuronal connectivity in the developing hippocampus.48 Hence, CRH overexpression starting from early embryogenesis might have caused adaptive changes of neuronal physiology. However, the attenuation of active coping behavior and of HPA axis hyperreactivity by pretreatment of CRH-COEhet-Nes mice with the CRH-R1-selective antagonist DMP696 suggests that these phenotypes are an acute consequence of CRH overexpression.33 Considering the observation that CRH-R1 antagonists exhibit highest efficacy in rodents that are hyperresponsive to stress and exhibit increased levels of CRH,12, 39 CRH-COE-Nes mice may constitute a mouse model with strong and predictable responsiveness to CRH-R1 antagonists.

To assure that the FST phenotype of CRH-COE-Nes mice is not of artificial nature, arising from excessive ectopic expression of CRH in the brain, we elucidated the role of the endogenous CRH system in FST behavior. Under basal, nonactivated conditions the endogenous CRH system does not influence FST behavior as neither the genetic disruption nor the pharmacological blockade of CRH-R1 affected FST behavior.10, 34, 35 However, we could demonstrate that stress-mediated activation of the endogenous CRH system before the FST decreased immobility in control mice, but not in CRH-R1 knockout mice. These findings suggest that ectopic overexpression of CRH in CRH-COE-Nes mice mimics the behavioral consequences of stress-mediated activation of the endogenous CRH system, and that CRH-R1-dependent signaling pathways promote the increase in active coping behavior.

The forebrain-restricted disruption of CRH-R1 in CRH-R1lox/lox Camk2a-cre mice decreases anxiety-related behavior,23 suggesting the involvement of CRH/CRH-R1-dependent pathways in structures of the anterior forebrain including those of the limbic system in emotionality control. In order to investigate whether these structures are also causally related to altered FST behavior, we spatially restricted CRH overexpression using Camk2a-cre mice. Similar to CRH-OE2122 mice,15 CRH overexpression in the forebrain did not result in the reduced immobility observed in CRH-COE-Nes (this study) and CRH-Tg mice,14, 43 presumably owing to the overexpression in principal neurons only. To date the cellular and subcellular localization of CRH and its receptors including receptor-mediated effects at synapses are not well understood. In the hippocampus, endogenous CRH expression has been assigned to GABAergic interneurons from where CRH is released in the cause of acute physiological stress.49 CRH release excites pyramidal cells, which express CRH-R1 postsynaptically on their dendritic spines. However, using Dlx-cre mice to direct CRH overexpression to GABAergic interneurons, in order to model more closely endogenous expression sites in the forebrain, failed to affect FST behavior. The absence of the FST phenotype in forebrain-specific CRH overexpressing mouse lines argues against volume transmission throughout the brain as mediating CRH effects, but clearly favors a mechanism involving specific synaptic release. Furthermore, our results suggest that CRH overexpression in more caudal brain nuclei of the mid/hind brain could promote the effects of CRH on stress-coping behavior.

In the brain stem of mice and rats CRH-R1 is expressed in serotonergic neurons of the median and DR50 and in dopaminergic neurons of the SN and VTA.51 CRH is expressed in noradrenergic neurons of the LC,52 and i.c.v. injection of CRH has been demonstrated to induce a strong Fos immunoreactivity in the LC indicating the activation of CRH-R-dependent signaling pathways.52 Accordingly, CRH is able to potentiate noradrenergic,40, 53 dopaminergic54 and serotonergic55 neurotransmission. Here we could demonstrate that the pharmacological blockade of catecholamine synthesis by AMPT, but not of serotonin synthesis by PCPA, could partially reverse the phenotype of CRH-COEhom-Nes mice. Therefore, it is likely that CRH overexpression activates the endogenous catecholaminergic system similarly as antidepressants, which would explain increased arousal in an open field and the reduced immobility in the FST and TST.

Quantification of c-fos and zif268 expression in response to forced swim stress further identifies the hyperactivation of noradrenergic neurons of the LC in CRH-COE-Nes mice. Intracoerulear microinfusion of CRH in rats has demonstrated that CRH can serve as an excitatory neurotransmitter in the LC56 resulting in enhanced NA release in LC projection areas.57 CRH–NA interactions not only occur in the LC, where CRH activates the LC, but also at the projections of the forebrain noradrenergic system, where NA stimulates CRH release.58 For instance, stress induces NA release in the PVN and thereby stimulates secretion of CRH.59 This feed-forward mechanism could be a causal factor to the hyperreactivity of the HPA axis upon stress.

In conclusion, we have created a new, highly flexible transgenic mouse model, which can help in dissecting the contribution of CRH-sensitive pathways involved in the transition from physiological to pathological stress responses, which are thought to underlie the etiology of affective and anxiety disorders.6 This animal model is also suited for validating drug candidates targeting the central CRH system.