Impact of Targeted Deletion of the Circadian Clock Gene Bmal1 in Excitatory Forebrain Neurons on Adult Neurogenesis and Olfactory Function

As conventional Bmal1-/- mice show impaired adult neurogenesis in the dentate gyrus presumably as a consequence of oxidative stress [9], we analyzed oxidative stress in the conditional Bmal1 fKo mice using the oxidative stress marker, 8-hydroxydeoxyguanosine (8-OH(d)g). Perinuclear 8-OH(d)g-immunoreaction (IR) was not significantly different between wildtype mice (Bmal1 WT) (n = 4) and Bmal1 fKo (n = 4) mice in the DG (p = 0.1), hilus (p = 0.4), and CA1 (p = 0.1) (Figure 1). However, in the CA3 subregion 8-OH(d)g-IR was significantly higher in Bmal1 fKO as compared to Bmal1 WT (p = 0.03) (Figure 1). This shows a selective subregional increase in oxidative stress upon forebrain-specific neuronal Bmal1 deletion.

Conventional Bmal1-/- mice and NestinCRE⁺;Bmal1^f/f mice show increased activation of astrocytes indicated by the increased expression of GFAP [25]. Thus, we analyzed GFAP expression in the hippocampus of Bmal1 WT (n = 4) and Bmal1 fKO (n = 4) mice by immunofluorescence and immunoblot. Forebrain specific deletion of Bmal1 did not lead to the activation of astrocytes in the hippocampus, as GFAP-IR (Figure 2a) and relative GFAP expression (Figure 2b) were not significantly different between Bmal1 WT and Bmal1 fKO mice (GFAP-IR, p = 0.4; relative GFAP-expression, p = 0.9).

Proliferation and distribution of NPCs in the dentate gyrus were analyzed by Bromodeoxyuridine (BrdU) assay. One day after the last BrdU injection, BrdU+ cells were arranged in clusters or distributed sporadically, mainly in the subgranular zone (SGZ) of both genotypes (Figure 3a). The number of proliferating BrdU+ cells did not differ between Bmal1 WT (3950+/−691) and Bmal1 fKO (3132+/−225) mice (p = 0.3, n = 5 per genotype).

For the further analysis of the proliferating cells, BrdU was co-labeled with the marker for neuronal precursors and migrating neuroblasts, doublecortin (DCX). The percentage of BrdU/DCX-co-labeled cells was not different between Bmal1 WT (44.2+/−5.4) and Bmal1 fKO (49.6+/−2.7) mice (p = 0.6, n = 4 per genotype) (Figure 3b). This shows that specific Bmal1-deficiency in the forebrain neurons does not affect the proliferation of neuroblasts. Moreover, the subregional spatial distribution of the proliferating BrdU+ cells in the SGZ and the subdivisions of the granular cell layer of the dentate gyrus was not different between the two genotypes (Figure 3c).

For the analysis of the effect of forebrain specific Bmal1 deletion on the survival of adult-born cells in the hippocampus, BrdU+ cells were analyzed four weeks after the last BrdU injection. The number of BrdU+ cells in the DG was comparable between Bmal1 WT (750+/−117) and Bmal1 fKO (715+/−105) mice (p = 0.9, n = 5 per genotype) (Figure 4a). Thus, forebrain specific Bmal1 deletion does not affect the survival of adult-born cells. For further analysis of the survival of adult-born mature neurons, the percentage of cells co-labeled for BrdU and neuronal marker NeuN was determined four weeks after the last BrdU injection. In both genotypes, most of the BrdU+ cells expressed NeuN. The percentage of BrdU/NeuN+ cells was comparable between Bmal1 WT (81.6+/−4.2) and Bmal1 fKO (86.9+/−4.7) mice (p = 0.5, n = 4 per genotype) (Figure 4b). Consistently, there was no significant difference in the total DG volume between Bmal1 WT (0.36+/−0.04 mm³) and Bmal1 fKO (0.36+/−0.03 mm³) mice (p = 0.9, n = 3 per genotype). Taking these findings together, we presume that the survival and fate decisions of NPCs in the hippocampus were not affected by forebrain specific deletion of Bmal1.

As conventional Bmal1-deficiency affects NPC migration in the RMS and the OB presumably, as a consequence of oxidative stress [10], we analyzed 8-OH(d)g-IR in the RMS and the OB of Bmal1 fKO mice (n = 4 per genotype). 8-OH(d)g-IR was not different in both the RMS (p = 0.7) and the GCL (p = 0.2) (Figure 5). However, the MCL showed enhanced 8-OH(d)g-IR in Bmal1 fKO as compared to Bmal1 WT mice (p = 0.03) (Figure 5). This is consistent with the impact of forebrain specific Bmal1-deletion on oxidative stress levels in selective brain subregions.

The proliferation of NPCs in the SVZ and the RMS was analyzed by BrdU assay one day after the last BrdU injection (n = 5 per genotype). The NPC proliferation in the neurogenic niches of the SVZ and the RMS was not affected by forebrain specific deletion of Bmal1, as there was no significant difference in the number of BrdU+ cells between Bmal1 WT and Bmal1 fKO mice in the SVZ (p = 0.5), in the vertical limb of the RMS (p = 0.4) or in the horizontal limb of the RMS (p = 0.6) (Figure 6).

The number of BrdU+ cells reaching the GCL and the GL of the OB one day after the last BrdU injection was not significantly different between Bmal1 WT and Bmal1 fKO mice (p = 0.8, n = 5 per genotype) (Figure 7). Moreover, both the thickness of the RMS, defined by the DCX+ migrating neuroblasts (p = 0.8, n = 5 per genotype), and the dense network of astrocytes surrounding the RMS, defined by GFAP-IR (p = 0.07), were not different between Bmal1 WT and Bmal1 fKO mice (Figure S1).

Consistently, the cytoarchitecture of the olfactory bulb in cresyl violet stained sections appeared not different between Bmal1 WT and Bmal1 fKO mice (Figure S2, Table 1).

In both genotypes, Reelin+ cells could not be found in the RMS (not shown) but in the mitral cell layer (MCL) and the external plexiform layer (EPL) of the OB (Figure 8). In the EPL, there was no difference in the number of Reelin+ between Bmal1 WT (n = 5) and Bmal1 fKO (n = 4) mice (p = 0.66) (Figure 8). However, in MCL the number of Reelin+ cells was significantly lower in Bmal1 WT (n = 5) as compared to Bmal1 fKO (n = 4) mice (p = 0.03) (Figure 8). Interestingly, in the OB of Bmal1 WT mice, Reelin-Ir showed a co-localization with Bmal1 (Figure S3).

The buried food test was used to test the olfactory function. The latency to find the hidden food was comparable in Bmal1 WT (43.7+/−11.0 sec) and Bmal1 fKO (56.7+/−16.3 sec) mice (p = 0.5, n = 14 per genotype) (Figure 9).

All animal experiments were approved by the local government, North Rhine-Westphalia State Agency for Nature, Environment and Consumer Protection, Germany (AZ: 84-02.04.2012.A102, 84-02.04.2014.A314) in accordance to international guidelines on the ethical use of experimental animals [43]. All efforts were exerted to decrease the number and suffering of animals.

In Camk2a: Cre transgenic mice (B6.Cg-Tg(Camk2a-cre)T29-1Stl/J, Jackson Laboratory, Bar Harbor, ME, USA) Cre recombinase is expressed under the Camk2a promoter. Camk2a is expressed most abundantly in forebrain mature neurons [44]. This mouse line was crossed with transgenic mice in which Bmal1 is flanked by loxP sites (B6.129S4(Cg)-Arntl^tm1Weit/J, Jackson Laboratory, Bar Harbor, ME, USA), allowing recombination between the loxP sites and inactivation of the Bmal1 gene only in the/cells in which Cre recombinase is expressed [44]. Thus, in the offspring (Bmal1 fKO) Bmal1 is selectively deleted in forebrain mature neurons [11]. The B6.129S4(Cg)-Arntltm1Weit/J mice were used as Bmal1 WT mice. Mice were kept for breeding at the local animal facility. Adult male (3–6 months old) mice were used for experiments. Prior to experiments, mice were housed in standard cages in a temperature-controlled environment under 12 h light and 12 h darkness (LD) condition, lights on at 6:00 am. Mice had free access to food and water. Genotype was confirmed by PCR as previously described [11]. Cell type-specific Bmal1 deletion was validated by immunohistochemistry using an anti-Bmal1 antibody (Figure S4). The integrity of the circadian system was validated by spontaneous locomotor activity rhythms in LD and constant darkness (DD) (Figure S5). Consistent with previous findings [11], Bmal1-immunoreaction was reduced in the cerebral cortex (Figure S4a) but not in the SCN (Figure S4b) and the entrainment as well as circadian rhythms of spontaneous locomotor activity were not affected in Bmal1 fKO mice (Figure S5).

Mice were injected with the S-phase marker BrdU (Roche, Basel, Switzerland) at a dose of 100 mg/kg twice daily, at the beginning and the end of the light phase (ZT2 and ZT12, respectively) for three consecutive days. To study NPCs proliferation, one group of mice (n = 5 per genotype) was sacrificed on the next day after the last BrdU administration. To study NPC survival and neuronal differentiation, a second group of mice (n = 5 per genotype) was sacrificed 28 days after the last BrdU administration.

Animals were deeply anesthetized using Ketamine/Xylazine (100 mg/10 mg/kg body weight, respectively). Mice were perfused with 0.9% NaCl transcardially followed by 4% paraformaldehyde using a Ministar Peristaltic Pump (World Precision Instruments, Sarasota, FL, USA). Brains were removed from the skull, post-fixed in 4% paraformaldehyde for 24 h followed by cryoprotection in 20% then 30% sucrose, each for 24 h. Brains were divided into two sagittal halves. One half was sectioned through the entire rostro-caudal extent of the brain into six 40 μm free-floating coronal parallel sections. The other half was embedded in optimal cutting temperature compound (OCT) and cut into six 20 µm sagittal parallel sections. Sectioning was performed using a cryostat (Leica CM, Wetzlar, Germany).

After rinsing with phosphate-buffered saline (PBS), sections were incubated with 0.6% H₂O₂ for 30 min at room temperature (RT) and then washed in PBS. DNA was denatured using 2N HCl for 30 min at 37 °C, followed by rinsing with 0.1 M boric acid for 10 min at RT. After rinsing with PBS, sections were incubated in 10% normal goat serum in PBS-0.2% Triton for 1h at RT to quench unspecific binding of the secondary antibody. Sections were incubated with rat monoclonal anti-BrdU antibody (1:800, Serotec, oxford, UK), mouse monoclonal anti-8 hydroxy 2′ deoxyguanosine (8-OH(d)G) antibody (1:250, QED Bioscience, San Diego, CA, USA) or rabbit anti-Bmal1 primary antibody (1:1000, generous gift from Prof. D. Weaver) overnight at 4 °C. Sections were incubated with biotinylated goat anti-rat, anti-mouse or anti-rabbit IgG (1:500, Vector Laboratories, Burlingame, CA, USA), respectively, for 1h at RT followed by incubation with VECTASTAIN® Elite® ABC solution (Vector Laboratories, Burlingame, CA, USA) for 1h at RT. Then, sections were rinsed and incubated with 0.05% 3, 3′-Diaminobenzidine (SIGMA-ALDRICH, St. Louis, MO, USA) for 5 min. Sections stained with using an anti-BrdU antibody were counter-stained with Cresyl violet. Staining against Bmal1 was done without DNA denaturation. Sections were mounted and dehydrated. Slides were cover-slipped using Depex (SERVA Electrophoresis, Heidelberg, Germany).

Coronal sections of the hippocampal region were used for double immunofluorescence labeling with BrdU/doublecortin (DCX), or BrdU/NeuN. DNA was denatured as described above. Sections were incubated with a mixture of anti-BrdU (1:500, AbD Serotec, Kidlington, UK) with rabbit polyclonal anti-DCX (1:1000, Abcam, Cambridge, UK) or polyclonal rabbit anti-NeuN (1:1000, Millipore-Chemicon, Burlington, MA, USA). Sections were then rinsed in PBS, followed by incubation with a mixture of the secondary antibodies Alexa Fluor 488 goat anti-rat IgG (1:500, Molecular Probes, Eugene, OR, USA) and Alexa Fluor 568 goat anti-rabbit IgG (1:500, Molecular Probes, Eugene, OR, USA) for 1h at RT. Sections were rinsed, and nuclei were counter-stained with NucBlue (Molecular probes, Eugene, OR, USA). Sections were mounted and dehydrated. Slides were cover-slipped using Vectashield Hard Set anti-fade reagent (Vector Laboratories, Burlingame, CA, USA) and stored in darkness at 4 °C. For staining against Reelin, sagittal sections were incubated with mouse monoclonal anti-Reelin (1:2000, Abcam, Cambridge, UK) and then with Alexa Fluor 488 goat anti-mouse IgG (1:500, Molecular Probes, Eugene, OR, USA) as secondary antibody.

The animal genotype was obscured to the investigator. The camera settings (exposure time, photo interval, haze reduction condition) were kept identical during images acquisition and processing in all samples. Image acquisition, processing, and analysis of co-localization were performed by BZ-II analyzer software (Keyence Corporation, Osaka, Japan). Immunohistochemically stained sections were analyzed using the bright field mode on a KEYENCE BZ 900E microscope (Keyence Corporation, Osaka, Japan) by a 40X objective.

In one out of six series of coronal sections containing the hippocampus, BrdU-immunopositive (+) cells in the DG were analyzed. The granular cell layer of DG was subdivided into inner, middle and outer thirds while the subgranular zone (SGZ) was considered as a two-nucleus-wide region along the inner border of the granular cell layer of DG towards the hilus. Cells were ranked in regard to their location in the SGZ or granular cell layer of DG and counted manually. The resulting cell number was multiplied by six to estimate the number of BrdU+ cells in the entire hippocampus.

In sagittal sections, BrdU+ cells were analyzed in the SVZ, RMS, and OB. The RMS was divided into two subregions: the vertical limb (VL) and the horizontal limb (HL). The granular cell layer (GCL) and the glomerular layer (GL) of the OB were analyzed separately. Reelin+ cells were analyzed in the mitral cell layer (MCL) and the external plexiform layer (EPL) of the OB. Cells were counted in a delineated area, and the mean cell density in each mouse was expressed as the number of cells/mm².

Mean intensity of 8-OH(d)G-IR in the hippocampus (DG, CA1, CA3), SVZ, RMS (VL, HL), and OB (GCL, GL, MCL) was determined. Quantification of IR was performed using Image J software (available online: http://rsbweb.nih.gov/ij↗). Background staining in cell-free neuropil was used to define the threshold. The area of positively 8-OH(d)g-IR above the threshold was expressed as the percentage of the total area.

Cytoarchitecture of the OB was determined based on cresyl violet staining; the width of the layers was measured as described previously [45].

Fluorescent signals were analyzed using a 40X objective and respective filters of the KEYENCE BZ 900E (Keyence Corporation, Osaka, Japan). Fifty BrdU+ cells were analyzed for co-localization with DCX, while twenty BrdU+ cells were analyzed for co-localization with NeuN per animal.

For volume analysis, one out of six coronal (DG) and sagittal (OB) series, respectively, was stained with cresyl violet. The boundaries of the DG and the OB were outlined, and the area was multiplied by six and section thickness (40 μm or 20 µm; respectively) to calculate the total volume according to the Cavalieri principle as described before [9]. Volumes were displayed in mm³.

Bmal1 WT and Bmal1 fKO mice 16 weeks old (n = 5 per genotype) were killed with isoflurane. The hippocampus was carefully dissected out under a dissection microscope. Tissue was homogenized and lysed in RIPA buffer (Thermo Scientific, Waltham, MA, USA) with 1% Halt Protease Inhibitor Cocktail (Thermo Scientific, Waltham, MA, USA). Protein concentration was determined using the BCA kit (Thermo Scientific, Waltham, MA, USA).

PAGE and Immunoblot were performed using Novex XCell Sure Lock Electrophoresis and Blot module (Thermo Scientific, Waltham, MA, USA) according to manufacturer’s instructions and Invitrolon™ PVDF membranes (45 µm pores, Life technologies, Carlsbad, CA, USA).

The membranes were washed in TBS-Tween (TBST) and incubated with the blocking solution (TBST containing 5% fat-free milk powder, NEB, Ipswich, MA, USA) for 1 h. Membranes were incubated with rabbit polyclonal anti-GFAP (1:80000, DAKO, Glostrup, Denmark) for 12 h at 10 °C. After washing, membranes were incubated with secondary HRP-conjugated goat anti-rabbit IgG (1:40000, Dianova, Hamburg, Germany) for 1 h. After washing, immunoreactive bands were visualized using Super Signal West ECL (Thermo Scientific, Waltham, MA, USA) by Molecular Imager^® ChemiDoc™XRS (BioRad, Hercules, CA, USA). The intensity of the respective immunoreactive bands was normalized against b-Actin using densitometric analyses using ImageJ software.

For analysis of spontaneous locomotor activity, Bmal1 fKO (n = 5) and Bmal1 WT (n = 5) mice were individually housed in standard cages with free access to food and water. Spontaneous locomotor activity was recorded using infrared movement detectors linked to a monitoring system (Mouse-E-Motion) (infra.e.motion, Hamburg, Germany). For the analysis of entrainment of spontaneous locomotor activity, the mice were kept in a 12h light/12 h dark cycle (LD) (light on at 06.00, light off at 18.00) for two weeks. For the analysis of a circadian rhythm in spontaneous locomotor activity, the mice were kept under constant darkness (DD) for two weeks. Locomotor activity was measured continuously and recorded in 10 min intervals. For activity profile analysis MatLab R12 based Clocklab software (Actimetrics, Wilmette, IL, USA) was used. Rhythm stability of the period length in LD and in DD was determined by Qp-analysis [6]. The chronotype of mice was determined by the “median of activity” (MoA). The MoA is the time-point on the timescale at which the mouse has achieved 50% of its daily activity. The standard deviation of the MoA (SDevMoA) gives an additional measure of rhythm instability [46].

For assessment of olfactory function in Bmal1 fKO and Bmal1 WT (n = 14 for each genotype), a buried food test was performed as previously described [47,48]. During the habituation period of two weeks, the mice received one hedonistic piece of food (Froot Loops, Kellogg’s, Battle Creek, MI, USA) daily. Before the test phase, the mice were food-deprived for 18 h. The mice were tested individually between 9:00–11:00 am. The mouse was allowed to acclimate for 5 min in the test cage (clean cage filled with a 3-cm depth of bedding), then transferred temporarily into a holding cage. The mouse was transferred again to the test cage after a single Froot Loop was completely hidden 1 cm beneath the surface in one corner of the cage. The latency to find the Froot Loop was recorded as soon as the mouse touched the bedding until it held the Froot Loop with the paws or when the mouse started to eat it. Mice that couldn’t find the food within 300 s were excluded from the analysis (one Bmal1 fKO mouse).

Statistical analysis was performed using Graph Pad Prism software. Mann-Whitney-U test, unless stated otherwise, was used to determine differences between groups. p-value < 0.05 was considered statistically significant. Values are presented as mean ± SEM.