Glucocorticoid-related molecular signaling pathways regulating hippocampal neurogenesis

Christoph Anacker, PhD1,2,3; Annamaria Cattaneo, PhD4; Alessia Luoni5; Ksenia Musaelyan, MD1; Patricia A. Zunszain, PhD1; Elena Milanesi4; Joanna Rybka, MSc6; Alessandra Berry7, Francesca Cirulli, PhD7; Sandrine Thuret, PhD3; Jack Price, PhD3; Marco Riva, PhD5; Massimo Gennarelli, PhD4,8 and Carmine M. Pariante, MD, PhD1,2

1 King’s College London, Institute of Psychiatry, Department of Psychological Medicine, Section of Perinatal Psychiatry and Stress, Psychiatry and Immunology (SPI-Lab), London, UK

2 National Institute for Health Research, Biomedical Research Centre for Mental Health, Institute of Psychiatry and South London and Maudsley NHS Foundation Trust, London, UK

3 King’s College London, Institute of Psychiatry, Centre for the Cellular Basis of Behaviour (CCBB), London, UK

4 Department of Biomedical Sciences and Biotechnology, University of Brescia, Brescia, Italy

5 Department of Pharmacological and Biomolecular Sciences, University of Milan, Milan,Italy

6 Department of Biochemistry, Nicolaus Copernicus University in Toruń, Bydgoszcz, Poland

7 Department of Cell Biology and Neuroscience, Istituto Superiore di Sanita, Rome, Italy

8 Genetics Unit, Fatebenefratelli, Giovanni di Dio, Brescia, Italy

Supplementary Materials

Supplementary Methods

Cell Culture

The immortalized multipotent, human hippocampal progenitor cell line, HPC03A/07 (provided by ReNeuron Ltd., Surrey, UK), was used for all experiments. HPC03A/07 cells were originally obtained from a 12-week old male fetus and conditionally immortalized with the c-myc-ERTM transgene (Danielian et al, 1993; Littlewood et al, 1995; Pollock et al, 2006). This construct is exclusively responsive to the synthetic steroid 4-hydroxytamoxifen (4-OHT) (Danielian et al, 1993). HPC03A/07 cells proliferate indefinitely in the presence of epidermal growth factor (EGF), fibroblast growth factor (bFGF) and 4-OHT, whereas proliferation is ceased upon their removal (Littlewood et al, 1995; Pollock et al, 2006). During normal expansion, HPC03A/07 cells proliferate with a doubling time of 72 hours (80% confluence) (Johansson et al, 2008). We thus cultured cells for 72 hours in proliferation media containing EGF, bFGF and 4-OHT, and subsequently removed growth factors and 4-OHT for 7 days to induce differentiation.

HPC03A/07 cells were grown in reduced modified media (RMM) consisting of Dulbecco’s Modified Eagle’s Media/ F12 (DMEM:F12, Invitrogen, Paisley, UK) supplemented with 0.03% human albumin (Baxter Healthcare, Compton, UK), 100 µg/ml human apo-transferrin, 16.2 µg/ml human putrescine DiHCl, 5 µg/ml human rec. insulin, 60ng/ml progesterone, 2 mM L-glutamine and 40 ng/ml sodium selenite. To maintain proliferation, 10 ng/ml human bFGF, 20 ng/ml human EGF and 100 nM 4-OHT were added. The cell culture media is free of any glucocorticoids unless cortisol is added as a treatment in the experimental conditions.

Immunocytochemistry

To assess progenitor cell proliferation, BrdU-containing cells were incubated with hydrochloric acid (HCl, 2 N) for 15 min at room temperature, blocking solution for 60 min at room temperature, primary antibody (rat anti-BrdU, Serotec, Oxford, UK. 1:500) at 4°C over night, and secondary antibody (Alexa 488 goat anti-rat, 1:500, Invitrogen) for 2 hours at room temperature. The number of BrdU-positive cells over total Hoechst 33342-positive cells was counted in an unbiased setup with an inverted microscope (IX70, Olympus, Hamburg, Germany) and ImageJ 1.41 software (http://rsbweb.nih.gov). To assess neuronal differentiation after 7 days differentiation, PFA-fixed cells were incubated in blocking solution (10% normal goat serum (NGS), Alpha Diagnostics, San Antonio, Texas) in PBS containing 0.3% Triton-X for 2 hours at room temperature, and with primary antibodies (rabbit anti-Dcx, 1:1000; mouse anti-MAP-2 [HM], 1:500, Abcam, Cambridge, UK; rabbit anti-S100ß, 1:500, DAKO) at 4°C over night. Cells were incubated sequentially in blocking solution for 30 min, secondary antibodies (Alexa 594 goat anti-rabbit; 1:1000; Alexa 488 goat anti-mouse, 1:500, Invitrogen, Paisley, UK) for 1 hour, and Hoechst 33342 dye (0.01 mg/ml, Invitrogen) for 5 min at room temperature. The number of Dcx-, MAP2- and S100ß-positive cells over total Hoechst 33342-positive cells was determined as described above. Expression of synaptic markers in further differentiated neurons (after 5 weeks) was examined by co-staining for synaptophysin (mouse anti-synaptophysin, 1:100, Assaydesigns) and Homer1 (rabbit anti-Homer1, 1:100, Synaptic Systems) with Tuj1 (rabbit anti-TuJ1, 1:500, Sigma, St-Louis, MO) and MAP2 (mouse anti-MAP-2 [HM], 1:500, Abcam, Cambridge, UK), respectively. Developing granule cells were identified by staining for Prospero homeobox protein 1 (Prox1) (mouse anti-Prox1, 1:100, Abcam). Negative controls were incubated with unspecific mouse IgGs (1:500, control for MAP-2), rabbit IgGs (1:500, control for Dcx and S100ß) or rat IgGs (1:500, control for BrdU) in place of the specific primary antibody.

Gene expression analysis

RNA was isolated using RNeasy mini kit (Qiagen, Crawley, UK) according to the manufacturer’s instructions. Samples were treated with DNase (Ambion, Warrington, UK) and RNA quantity was assessed by evaluation of the A260/280 and A260/230 ratios using a Nanodrop spectrometer (NanoDrop Technologies, Wilmington, USA). RNA quality (RIN) was assessed using Agilent Bioanalyzer (Agilent Technologies). Superscript III enzyme (Invitrogen) was used to reverse transcribe 1 µg total RNA. Quantitative Real-Time PCR was performed using HOT FIREPol® EvaGreen® qPCR Mix (Solis BioDyne, Tartu, Estonia) according to the SYBR Green method. PCR cycles consisted of an initial heating step at 95 °C for 15 min to activate the polymerase, 45 PCR cycles were performed. Each cycle consisted of a denaturation step at 95 °C for 30 s, an annealing step at 60 °C for 30 s and an elongation step at 72 °C for 30 s. Primers were designed to amplify 100-150bp regions spanning an exon boundary within the coding sequence of the target gene. GC content of each primer was designed to be 50-55%. For each target primer set, a validation experiment was performed to demonstrate that PCR efficiencies were within the range of 90-100% and approximately equal to the efficiencies of the reference genes. Primer sequences are available upon request.

Each sample was assayed in duplicate and each target gene was normalized to the geometric mean of the three reference genes, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), beta-actin (ACTB); and beta-2-microglobulin (B2M). The Pfaffl Method was used to determine relative target gene expression. Data are expressed as fold change from the vehicle treated control condition.

Affymetrix gene expression microarray and quality control

Gene expression microarray assays were performed using Human Genome U219 Array Strips and Rat Gene 1.1 ST Array Strips on GeneAtlas platform (Affymetrix), following the 3’IVT one cycle labeling and amplification protocol described in the Affymetrix GeneChip Expression Analysis Technical Manuals (http://media.affymetrix.com/support/downloads/ manuals/geneatlas_3ivt_expkit_manual.pdf), and in the GeneAtlas™ WT Expression Kit User Manual. Human Genome U219 Array Strips as well as Rat Gene 1.1 St Array strips are comprised of more than 530,000 probes covering more than 36,000 transcripts and variants, which represent more than 20,000 genes mapped through UniGene or via RefSeq annotation.

To synthesize First-Strand cDNA, 250ng RNA were reverse-transcribed with the Gene Atlas 3’IVT Express Kit or WT Expression Kit (Affymetrix, Santa Clara, CA, USA) using T7 oligo(dT) primer. Second-Strand cDNA synthesis was carried out using DNA polymerase and RNase H to simultaneously degrade RNA and synthesize second-strand cDNA. This step was followed by the in vitro transcription using IVT Labeling Master Mix to generate multiple copies of biotin-modified amplified-RNA (aRNA) from the double-stranded cDNA templates. Subsequently aRNA was purified to remove unincorporated NTPs, salts, enzymes and inorganic phosphate to improve the stability of the biotin-modified aRNA. Labeled aRNA (10ug) was then fragmented and 7.5µg were hybridized onto HGU219 array strips or Rat Gene 1.1 ST Array Strip. The reactions of hybridation, fluidics and imaging were performed on the Affymetrix Gene Atlas instrument according to the manufacturer’s protocol.

Affymetrix CEL files were imported into Partek Genomics Suite version 6.6 for data visualization and statistical testing. Quality control assessment was performed using Partek Genomic Suite 6.6. All samples passed the criteria for hybridization controls, labeling controls and 3’/5’ Metrics. Background correction was conducted using Robust Multi-strip Average (RMA) (Irizarry et al, 2003) to remove noise from auto fluorescence. After background correction, normalization was conducted using Quantiles normalization (Bolstad et al, 2003) to normalize the distribution of probe intensities among different microarray chips. Subsequently, a summarization step was conducted using a linear median polish algorithm (Tukey 1977) to integrate probe intensities in order to compute the expression levels for each gene transcript. Upon data upload, pre-processing of CEL data for the complete data set (total of nine samples; three biological replicates per sample for vehicle, 100nM cortisol and 100µM cortisol) was performed using ANOVA to assess treatment effects. Differential gene expression across treatment was assessed by applying a p-value filter (for treatment) of p<0.05 to the ANOVA results. To investigate the effect of different cortisol concentrations, a three linear contrast was performed (cortisol 100nM versus vehicle; cortisol 100µM versus vehicle; cortisol 100µM versus cortisol 100nM). In this comparison, a maximum filter of p<0.05 and a minimum absolute fold change cut-off of 1.2 was applied. Genes that passed these criteria were used to build up the Venn diagram (Supplementary Figure 5a, Supplementary Table 3).

Supplementary Results and Discussion

HPC03A/07 cultures

As previously described (Anacker et al, 2011), BrdU incorporation for 4 hours during proliferation resulted in 35% BrdU-positive cells in the control condition. Differentiated HPC03A/07 cultures (7days differentiation) consisted of 35% Dcx-positive neuroblasts, 25% MAP2-positive mature neurons, 8% Dcx/MAP2-positive cells and 27% S100ß-positive astrocytes.

MR- and GR-dependent effects of cortisol on differentiation into Dcx-positive neuroblasts

In addition to investigating neuronal differentiation of HPC03A/07 cells into MAP2-positive neurons, as described in the Results, we also wanted to investigate the effects of MR- and GR-activation on differentiation into Dcx-positive neuroblasts. To do this, we first treated cells only during the proliferation phase (3 days), but not during the subsequent differentiation phase (7 days). We used immunocytochemistry to determine changes in differentiation into Dcx-positive neuroblasts (Supplementary Figure 1a).

Low concentrations of cortisol (100nM) decreased the number of Dcx-positive neuroblasts (by -21%, p=0.005, n=3). This effect was counteracted by co-treatment with the MR-antagonist, spironolactone (1µM) (Supplementary Figure 1b). Accordingly, the MR agonist, aldosterone (1µM), also decreased the number of Dcx-positive neuroblasts (by -15%, p=0.048, n=3), and again, this effect was counteracted by spironolactone (1µM) (Supplementary Figure 1b), confirming that this effect is mediated by MR activation.

High concentrations of cortisol (100µM) also decreased the number of Dcx-positive neurons (by -27%, p=0.002, n=3), and this effect was abolished by the GR-antagonist, RU486 (50nM) (Supplementary Figure 1c). Accordingly, treatment with the GR agonist, dexamethasone (1µM), decreased neuronal differentiation into Dcx-positive neuroblasts (by -35%, p<0.0001, n=3), an effect which was also abolished by co-treatment with RU486 (50nM) (Supplementary Figure 1c). These data are in line with the reduction in the number of MAP2-positive neurons upon treatment with cortisol, aldosterone and dexamethasone (Figure 2), and confirm our previous data on the same cell line which had shown dexamethasone-induced reduction in Dcx-positive neuroblasts (Anacker et al, 2011b).

Treatment only during the differentiation phase, but not during the preceding proliferation phase, did not exert any significant effects (Supplementary Figure 1d,e), further supporting the notion that MR- and GR-activation during the proliferation phase is both necessary and sufficient to reduce neuronal differentiation.

MR- and GR-dependent effects of cortisol on neurogenesis and astrogliogenesis: Continuous treatment during the proliferation- and the differentiation-phase

When cells were treated continuously during the proliferation and the differentiation phase, the low concentration of cortisol (100nM) and the MR-agonist, aldosterone (1µM), also decreased the number of MAP2-positive neurons (by -25%, p=0.005, n=3; and by -28%, p=0.01, respectively). Both effects were abolished by co-treatment with spironolactone (1µM) (Supplementary Figure 2a). The high concentration of cortisol (100µM) and the GR-agonist, dexamethasone (1µM), also decreased the number of MAP2-positive neurons (by -29%, p=0.002, n=3; and by -24%, p=0.01, respectively), and both effects were abolished by co-treatment with RU486 (50nM) (Supplementary Figure 2b).

Similar effects were observed when we counted the number of Dcx-positive neuroblasts in this treatment condition. Specifically, the low concentration of cortisol (100nM) and aldosterone (1µM) also decreased the number of Dcx-positive neuroblasts when cells were treated continuously during the proliferation phase and the differentiation phase (by -18%, p=0.042, n=3; and by -16%, p=0.049, respectively), and both effects were abolished by spironolactone (1µM) (Supplementary Figure 2c). The high concentration of cortisol (100µM) and dexamethasone (1µM) also decreased the number of Dcx-positive neuroblasts (by -23%, p=0.013; and by -34%, p=0.0004, respectively), and both effects were abolished by RU486 (50nM) (Supplementary Figure 2d).

Treatment during the proliferation phase and the differentiation phase caused similar effects on astrogliogenesis as treatment during the proliferation phase alone (illustrated in Figure 2f,g). In particular, cortisol (100nM) and aldosterone (1µM), increased the number of S100ß-positive astrocytes (by ~21%, p=0.003; and by ~24%, p=0.004, respectively) and both effects were abolished by spironolactone (1µM) (Supplementary Figure 2e). Cortisol (100µM) and dexamethasone (1µM) did not exert any significant effects on the number of S100β-positive astrocytes compared to the vehicle treated control condition in this treatment condition (Supplementary Figure 2f).

Taken together, these data further confirm that activation of the MR and the GR reduces neuronal differentiation of hippocampal progenitor cells and that activation of the MR shifts cell fate of developing progenitor cells from neurons towards astrocytes (as is shown by increased astrogliogenesis in the presence of decreased neurogenesis and activation of Notch/Hes-signaling in this condition). However, activation of the GR decreases neuronal differentiation but does not redirect hippocampal progenitor development towards astrogliogenesis.

Previous post-mortem studies have reported reduced glia cell numbers in the orbitofrontal cortex, the amygdala and the hippocampus of depressed patients (Rajkowska et al, 1999; Czeh and Lucassen 2007). Although these data seem to contrast our findings, which did not show impaired astrogliogenesis upon GR stimulation, it is important to note that the above mentioned studies did not examine neurogenic regions, such as the dentate gyrus, and have thus likely examined glia cell loss, rather than reduced glia cell birth. It is also possible that, in humans, prolonged exposure to high levels of glucocorticoids (as in depression) inhibits MR-induced astrogliogenesis (as shown by our data), ultimately resulting in reduced numbers of glia cells in the brain.