SEARCH

SEARCH BY CITATION

Keywords:

  • apoptosis;
  • brain development;
  • cerebral cortex;
  • histamine receptors antagonists;
  • neural precursors;
  • neurogenesis

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Histamine has neurotransmitter/neuromodulator functions in the adult brain, but its role during CNS development has been elusive. We studied histamine effects on proliferation, cell death and differentiation of neuroepithelial stem cells from rat cerebral cortex in vitro. RT-PCR and Western blot experiments showed that proliferating and differentiated cells express histamine H1, H2 and H3 receptors. Treatments with histamine concentrations (100 nM–1 mM) caused significant increases in cell numbers without affecting Nestin expression. Cell proliferation was evaluated by BrdU incorporation; histamine caused a significant increase dependent on H2 receptor activation. Apoptotic cell death during proliferation was significantly decreased at all histamine concentrations, and cell death was promoted in a concentration-dependent manner by histamine in differentiated cells. Immunocytochemistry studies showed that histamine increased 3-fold the number of neurons after differentiation, mainly by activation of H1 receptor, and also significantly decreased the glial (astrocytic) cell proportion, when compared to control conditions. In summary, histamine increases cell number during proliferative conditions, and has a neuronal-differentiating action on neural stem cells, suggesting that the elevated histamine concentration reported during development might play a role in cerebrocortical neurogenesis, by activation of H2 receptors to promote proliferation of neural precursors, and favoring neuronal fate by H1-mediated stimulation.

Abbreviations used
bFGF

basic fibroblast growth factor

BrdU

bromodeoxyuridine

D

differentiated

E

embryonic day

GFAP

glial fibrillar acidic protein

HA

histamine

MAP2

microtubule associated protein 2

ND

non-differentiated

NGS

normal goat serum

NSC

neural stem cells

P

passage

PBS

phosphate buffered saline

TUNEL

terminal deoxyuridine triphosphate nick end labelling

Histamine (HA) is produced, stored, released and metabolized in the brain, filling the criteria for a neurotransmitter/neuromodulator (Schwartz et al. 1991; Hill et al. 1997). In the adult CNS, HA regulates pre- and post-synaptically a variety of functions, such as wakefulness, feeding, drinking, body temperature and motor activity (Schwartz et al. 1979; Knigge and Warberg 1991; Wada et al. 1991; Onodera et al. 1994; Haas and Panula 2003). These HA actions are mediated by the activation of three different histaminergic G protein-coupled receptors named H1R, H2R and H3R, which are widely distributed throughout the CNS (Hill et al. 1997), and have been cloned and characterized by their pharmacology and signal transduction mechanisms (Gantz et al. 1991; Yamashita et al. 1991; Lovenberg et al. 1999; Tardivel-Lacombe et al. 2000). Activation of H1R and H2R excites neurons or potentiates excitatory inputs (Haas and Panula 2003), while activation of H3R causes inhibition of synthesis and release of HA and other neurotransmitters (Clapham and Kilpatrick 1992; Schlicker et al. 1994; Molina-Hernandez et al. 2000, 2001). The affinity of HA for these receptors vary: H1R and H2R are activated at μmolar concentrations of HA (Garbarg and Schwartz 1987; Traiffort et al. 1994), whereas H3R respond to HA in the nM range (Rouleau et al. 2004).

During rat development, HA is one of the first neurotransmitters to be present in CNS, starting at embryonic day (E) 12, and reaching its maximum value at E14–E16, decreasing afterwards 5-fold to adult levels in the prosencephalic area (Vanhala et al. 1994). Between E14 and E18, fibers from transient histaminergic neurons in the mesencephalon can be detected, passing through the ventral tegmental area and within the medial forebrain bundle and the optic tract, reaching the frontal and the parietal cortex at E15, earlier than other monoaminergic systems (Specht et al. 1981; Lidov and Molliver 1982; Auvinen and Panula 1988; Reiner et al. 1988; Vanhala et al. 1994), which coincides with the period where neuronal differentiation is occurring in cerebral cortex (Sauvageot and Stiles 2002). Messenger RNA of H1R and H2R are widely distributed in the developing CNS, whereas H3R is present in spinal cord and mesencephalon, appearing in cerebral cortex at E19 (Kinnunen et al. 1998; Heron et al. 2001; Karlstedt et al. 2001a, 2003). The developmental role of HA in the nervous system, including the cerebral cortex is still unknown (Mezei and Mezei 1978; Happola et al. 1991; Vanhala et al. 1994; Nissinen and Panula 1995; Nissinen et al. 1995).

A correlation of neurogenesis and elevations of HA in cerebral cortex can be proposed from the above mentioned findings. However, a direct approach to test this link has not been reported. Neural stem cells (NSC) are key players in brain development (Temple 2001). This study was designed to establish the role of HA on NSC by exploring in vitro the effect of this biogenic amine on cell proliferation, apoptosis, and differentiation using rat cortical precursor cells from E14. We show that HA is a positive modulator in proliferation/expansion of NSC, and also a factor that promotes neuronal differentiation of neural precursors; in this study we identified the histaminergic receptors responsible for these effects.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Cell culture

In order to obtain multipotent NSC (Johe et al. 1996), E14 embryos were extracted from pregnant Wistar rats and cerebral cortices were dissected in Krebs solution (100 mM NaCl, 2 mM KCl, 0.6 mM KH2PO4, 12 mM NaHCO3, 7 mM glucose, 0.1% phenol red, 0.3% bovine serum albumin and 0.3% magnesium sulfate). The tissue was mechanically dissociated to a single cell suspension. Cells were recovered by centrifugation, resuspended and cultured on plasticware previously treated with 15 μg/mL poly-L-ornithine (Sigma) and 1 μg/mL human fibronectin (Invitrogen) in fully defined N2 medium containing 10 ng/mL basic fibroblast growth factor (bFGF; R & D Systems and Peprotech) as mitogen. Passage (P) of cells was made with 0.1 mM EDTA in phosphate buffered saline (PBS). P2 cells were maintained during 4 days in proliferative control (N2 medium + 10 ng/mL bFGF) and experimental (N2 medium +10 ng/mL bFGF + different concentrations of HA from 100 nM to 1 mM) conditions. For immunocytochemistry and terminal deoxyuridine triphosphate nick end labelling (TUNEL) assays, cells were seeded at 1 × 104/well onto 12 mm coverslips in 24-well plates (Corning) in control and experimental conditions. For crystal violet assays, cells were seeded at the same density without coverslips. Differentiation was promoted by removing bFGF and keeping the cells for 6 days in N2 medium + 200 μM ascorbic acid in the presence or absence of HA. In the case of HA-treated cells, addition of HA was made during proliferation and differentiation phases.

Histamine H1, H2 and H3 receptor antagonists were used to study the effect of 100 μM HA on cell proliferation and differentiation. Chlorpheniramine (Sigma) was used as a H1R antagonist at 1 μM; as a H2R antagonist, we used 30 μM cimetidine (Sigma). Thioperamide (Sigma) at 1 μM was used to block H3R. H1R, H2R or H3R antagonists were added to control and 100 μM HA-treated cells during proliferation. Differentiation was promoted by removing bFGF and keeping the cells for 6 days in N2 medium + 200 μM ascorbic acid in the presence or absence of 100 μM HA, with or without the antagonists.

RNA extraction and RT-PCR

Total RNA was isolated from non-differentiated (ND) and differentiated (D) cultures, using TRIZOL (Invitrogen). For RNA extraction, cells were seeded at 3 × 105 in 6-well plates (Corning). Total RNA (1.0 μg) was reverse transcribed with random hexamers and 2 μL from the RT reaction were used in PCR containing 2 U Taq DNA polymerase (Invitrogen), 20 pmol of specific primers (Sigma), 500 μM deoxynucleoside triphosphates and 1.5 mM MgCl2 (for H1R and H3R) or 2 mM MgCl2 (for H2R). For amplification of cDNA encoding histaminergic receptors, the following reported forward (F) and reverse (R) primer sequences were used: H1R, F:5′-CTTCTACCTCCCCACTTTgCT-3′, R:5′-TTCCCTTTCCCCCTCTTg-3′; H2R, F:5′-TTCTTggACTCCTggTgCTgC-3′, R:5′-CATgCCCCCTCTggTCCC-3′ and for H3R, F:5′-CCAgAACCCCCACCAgATg-3′, R:5′-CCAgCAgAgCCCAAAgATg-3′. The conditions used were as follows for H1R and H3R: denaturalization at 95°C for 15 min, 30 cycles of denaturalization at 95°C for 1 min, annealing at 58°C for 1 min, and elongation at 72°C for 1 min. For H2R: denaturalization at 95°C for 15 min, 30 cycles of denaturalization at 95°C for 1 min, annealing at 62°C for 1 min, and elongation at 72°C for 1 min. Final extension at 74°C for 10 min was terminated by rapid cooling at 4°C. PCR products were analyzed in 2% agarose gel electrophoresis and the size of the reaction products was determined by comparison with molecular weight standards after ethidium bromide staining. As a negative control for PCR amplification, reactions with RNA in the absence of retrotranscription were included. The positive control consisted of RNA extracted from adult rat cerebral cortex, which was used to synthesize cDNA and amplified by PCR as described above.

Representative bands of the amplified PCR products were recovered from gels using the Qiaquick gel extraction kit (Qiagen) according to manufacturer’s instructions. We sequenced these bands at the Molecular Biology Unit in our institute, and confirmed that all bands correspond indeed to histaminergic receptors, in accordance with previous data (Azuma et al. 2003). GenBank accession numbers used to confirm the sequence from mRNA and the expected PCR product sizes are as follows: H1R (292 bp amplification product, primers encompassing nucleotides 593–885), AF387880; H2R (309 bp amplification product, primers encompassing nucleotides 196–505), NM012965; H3RA (393 bp amplification product, primers encompassing nucleotides 720–1113), AY009370; H3RB (297 bp amplification product, primers encompassing nucleotides 720–1017), AY009371; H3RC (249 bp amplification product, primers encompassing nucleotides 1055–1304), BC087707 (Azuma et al. 2003; Bakker 2004). H3RA and H3RB are also know as H3RL and H3RS, respectively.

Electrophoresis and western blot

Assays were performed as described (Diaz et al. 2007). Briefly, cells from proliferative and differentiated cultures, or from adult cerebral cortex were homogenized in lysis buffer supplemented with protease inhibitors (Roche, Germany). Proteins were obtained by centrifugation at 13 800 g at 4°C for 15 min, and quantified by a modified Bradford assay (BioRad, Germany). Proteins (40 μg) were resolved on 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (Amersham Bioscience, USA) which were blocked with 5% non-fat dry milk and incubated overnight with primary antibodies. Pre-stained markers (Invitrogen) were included for size determination. The following antibodies were used: rabitt anti-rat H1R polyoclonal antibody (diluted 1 : 1500, Santa Cruz Biotechnology, USA); goat anti-rat H2R polyclonal antibody (diluted 1 : 1500, Santa Cruz Biotechnology) and rabbit anti-rat H3R polyclonal antibody (1 : 1000, Alpha Diagnostics). Membranes were washed and incubated with corresponding horseradish peroxidase-coupled secondary antibodies (Santa Cruz Biotechnology; diluted 1 : 15 000). Immunoreactive bands were detected using enhanced chemiluminescence method (Amersham) and film exposure. When needed, membranes were stripped for reproving using a commercial solution (Chemicon).

Crystal violet assay

The number of cells was measured by the crystal violet assay, in which optic density is correlated with the amount of viable cells (Bonnekoh et al. 1989). Cells were fixed in 10% formol-PBS, pH 7.4, washed and incubated with a 0.5% solution of crystal violet for 10 min at 21°C. After thorough washing with bi-distillated water, acetic acid (33% vol:vol in H20) was added to elute the dye. The absorbance was determined at 595 nm in a spectrophotometer (Beckman DU650), and measurements were expressed as percent increases in the absorbance with respect to control conditions.

Bromodeoxyuridine (BrdU) incorporation assay

For cell proliferation analysis, cells treated with bFGF were incubated during 3 h with 10 μM 5-bromo-2-deoxyuridine (BrdU, Roche), washed with fresh medium with bFGF and fixed 21 h later. After fixation with 4% paraformaldehyde in PBS, pH = 7.4 for 20 min at 4°C, cells were incubated with 1 N HCl for 30 min at 25°C, and neutralized by washing three times in 0.1 M borate buffer, pH 8.5. Preparations were blocked for 1 h with 0.3% triton X-100 (Sigma) and 10% normal goat serum (NGS, Microlab, Mexico) in PBS. Cells were incubated overnight at 4°C with monoclonal rat anti-BrdU antibody (Accurate) at 1 : 100 dilution in blocking solution without triton. Three washes with 1% bovine serum albumin/PBS were made and then secondary antibody was added (Alexa Fluor 488 goat anti-rat IgG; Molecular Probes) at 1 : 1000 dilution in blocking solution without triton for 1 h at 25°C in the dark, and washed three times with PBS. Nuclei were stained with Hoechst 33258 (1 ng/mL; Sigma). Immunostaining were observed with an epifluorescence microscope (Nikon, Eclipse TE2000-U) and photographed with a Nikon digital camera (DMX1200 F). Negative controls were performed in the absence of primary antibodies and showed no unspecific staining.

Immunocytochemistry

Standard procedures reported before were used (Velasco et al. 2003; Diaz et al. 2007). Cortical cells were fixed at day 4 of proliferation or at day 6 of differentiation with 4% paraformaldehyde in PBS, pH 7.4 for 20 min at 4°C, permeabilized and blocked for 1 h with 0.3% triton X-100 and 10% NGS in PBS. Cells were incubated overnight at 4°C with the following primary antibodies, diluted in PBS containing 10% NGS: rabbit polyclonal anti-β tubulin III (1 : 2000, Babco Covance); rabbit polyclonal anti-glial fibrillary acidic protein (GFAP; 1 : 2000, DAKO); mouse monoclonal antibody anti-microtubule associated protein 2 (MAP2; 1 : 500, Chemicon); mouse monoclonal anti-Nestin (1 : 100; Developmental Studies Hybridoma Bank). Alexa-Fluo 488 anti-rabbit IgG, Alexa 568 anti-mouse IgG were used as secondary antibodies (1 : 500; Molecular Probes) diluted in PBS/10% NGS. Nuclei were counterstained with Hoechst 33258 (1 ng/mL; Sigma). Immunostainings were visualized and photographed as described above. Negative controls made as described previously did not show unspecific staining.

TUNEL assay

Neural stem cells in P2 were plated onto 12 mm coverslips and daily treated with HA, or maintained in control conditions. Cells were fixed with 4% paraformaldehyde, after 4 days of proliferation or 6 days of differentiation, and subsequently washed three times with PBS pH = 7.4. For detection of apoptotic cells, the Terminal deoxy-transferase-mediated deoxyuridine triphosphate nick-end labeling (In Situ Cell Death Detection Kit, Roche Diagnostics) was used to evaluate the effect of HA on cell death. Apoptotic cells were visualized by fluorescence microscopy. To analyze the response of differentiated cells to the used HA concentrations, we used GraphPad software to adjust the data to a sigmoidal curve to calculate the maximum effect and the effective concentration to have 50% of maximal response (EC50).

Cell counting

Cell counts from BrdU, immunocytochemistry and TUNEL experiments were performed from pictures taken with a Nikon digital camera and the Nikon ACT-1 imaging software. Quantification of cells was performed by counting the number of Hoechst stained nuclei (total cells) and the specified markers in at least eight random fields in duplicate, from 3–6 independent experiments.

Statistics

All data are presented as mean ± standard error of mean (S.E.M). One-way anova was performed for statistical analysis, and multiple comparisons between treated and control groups were made using the post-hoc Student-Newman-Keuls test. Differences were considered statistically significant at < 0.05. Graphs and fit adjustment were performed using GraphPad Instat software.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We first studied if cortical NSC express HA receptors by RT-PCR in ND and D cultures without HA treatment. We observed that NSC, at the end of both proliferation and differentiation stages contain mRNA for H1R, H2R and H3R receptors (Fig. 1a). The PCR amplification products were gel-extracted and sequenced. With the resulting sequence for each band, we confirmed that mRNA for histaminergic receptors were present in NSC. H3R was present in three bands that correspond to H3RA, H3RB and H3RC. Western blot analysis of cell extracts at the same time points showed that ND and D cells express H1R, H2R and H3R. For H1R and H2R, a single band was observed, which match the reported molecular weights of these receptors (Smit et al. 1995; Matsuda et al. 2004). In the case of H3R, we identify several bands (Fig. 1b), which have molecular weights similar to those observed by others (Karlstedt et al. 2003).

image

Figure 1.  Neural stem cells from cerebral cortex at P2 express histamine receptors. (a) RT-PCR analysis of H1-, H2- and H3-receptor mRNAs in non-differentiated (ND) and differentiated (D) cells. One μg of total RNA from adult rat cerebral cortex (Ctx), ND or D cortical neural stem cells were subjected to reverse transcription. The obtained cDNAs were amplified for 30 cycles using receptor-specific primers. The expression of gene transcripts for H1-, H2- and H3-receptors was analyzed by electrophoresis on a 2% agarose gel. Rat adult cortex was used as positive control, and RNA without retrotranscription (-RT) yielded no signal. Representative image of five independent experiments. (b) A representative assay of three western blot experiments, performed to detect H1, H2 and H3 receptors in Ctx, ND and D cells. Forty μg of total were resolved on 8% SDS-PAGE and transferred to nitrocellulose membranes, and probed with anti-HA receptors antibodies. Immunoreactive bands were detected by enhanced chemiluminescence.

Download figure to PowerPoint

Effects of histamine on proliferative cells

Effect of HA on Nestin expression in proliferating neural stem cells

Nestin is a component of intermediate filaments used to identify NSC (Johe et al. 1996). To study whether HA was able to modify the phenotype of NSC in terms of Nestin expression, cells were daily treated with different concentrations of HA (100 nM, 1 μM, 10 μM, 100 μM and 1 mM), added in the presence of bFGF. Table 1 shows that HA did not affect the proportion of NSC in culture, since in all HA concentrations used, the values were above 96% of Nestin-positive cells. At the same time, we explored the possibility that HA could promote early differentiation by identifying immunocytochemically the expression of β-tubulin III (an early marker of neurons) or GFAP (astrocyte marker) under proliferative conditions. Our results show only few double-positive cells for Nestin/β-tubulin III and Nestin/GFAP (proportions < 0.75% in all cases).

Table 1.   Effect of histamine on Nestin expression by proliferating cortical neural stem cells
 ControlHistamine
100 nM1 μM10 μM100 μM1 mM
  1. Cells were kept in proliferative conditions (N2 medium + 10 ng/mL bFGF) and treated daily with increasing concentrations of HA during 4 days. Results are expressed as percentages of labeled cells referred to the total cell number detected by Hoechst staining of nuclei. Double immunocytochemistry was made with two combinations of primary antibodies: Nestin/GFAP or Nestin/β-tubulin III. Fluorescent secondary antibodies were added to quantify positive cells. Microphotographs were taken at 40X from eight fields in three independent experiments. Results are expressed as the mean ± SEM. No statistically significant differences were found.

% Nestin97.4 ± 1.296.4 ± 2.698.2 ± 0.798.0 ± 0.597.0 ± 1.896.7 ± 0.2
%Nestin+  GFAP+0.05 ± 0.050.54 ± 0.530.04 ± 0.040.75 ± 0.630.18 ± 0.060.0 ± 0.0
%Nestin+  β-tubulin III+0.22 ± 0.150.13 ± 0.070.15 ± 0.080.28 ± 0.280.22 ± 0.160.23 ± 0.05
Effect of HA on cell number during proliferation

To study HA effect on NSC number, cells were daily treated with increasing concentrations of HA in the presence of 10 ng/mL bFGF during 4 days and NSC number was evaluated with crystal violet assay. HA produced significant increases up to 139% of control at all tested concentrations (Fig. 2a). We also quantified the number of cells present at this stage by counting the Hoechst-stained nuclei in eight random fields. We found significant raises (49–76%) at all HA concentrations used, except 100 nM (21% increase, Fig. 2b).

image

Figure 2.  Histamine increases proliferation of cortical NSC. Cells were kept in N2 medium with 10 ng/mL bFGF and treated daily with increasing concentrations of HA during 4 days. HA effect on cell number was evaluated by crystal violet assay (a) and by cell counting (b) after Hoechst staining of nuclei. For crystal violet, the resulting absorbance in each condition was expressed as percent of the control value (zero HA). Results are means ± SEM from 3–5 experiments. (*)p < 0.05 vs. control condition. (c) Representative micrographs of the antagonic effect of cimetidine on HA-induced increase in BrdU incorporation. Control cells incorporated BrdU (red) in the absence of HA. The proportion of BrdU+ cells augmented 2-fold when 100 μM HA was present in the cultures. This effect was prevented by the H2R antagonist cimetidine. (d) Quantification of BrdU incorporation experiments. HA caused a significant increase in the percentage of positive cells for BrdU. This effect depends on H2R activation, since cimetidine (Cimet.) completely reversed it. Neither chlorpheniramine (Chlorph., H1R antagonist) nor thioperamide (Thioper., H3R blocker) had this effect. To obtain the total number of cells, we quantified the nuclei stained with Hoechst from the combined values of eight fields in duplicate. Results are means ± SEM from 3–6 experiments. (a) p < 0.05 vs. control condition; (b) p < 0.01 vs. 100 μM HA. Scale bar = 50 μm.

Download figure to PowerPoint

To more directly asses HA effects on proliferation, we performed BrdU incorporation experiments with or without 100 μM HA (Fig. 2c). In control conditions, roughly 40% of ND NSC incorporated BrdU. Treatment with 100 μM HA significantly increased to 77% BrdU-positive cells (Fig. 2d). To evaluate if HA receptors were responsible for this increase, we decided to test H1R, H2R and H3R antagonists. Cells with or without 100 μM HA were daily-treated with 1 μM chlorpheniramine (H1R antagonist), 30 μM cimetidine (H2R blocker) or 1 μM thioperamide (H3R antagonist). Our results show that the increase on NSC proliferation induced by HA is due mostly to activation of H2R, since blocking of this receptor completely and significantly reversed the effect of 100 μM HA, obtaining similar values to those seen under control conditions (Fig. 2c and d). Although the H1R antagonist also significantly decreased HA effect on cell proliferation to 63%, the effect of HA remained, since blocking of this receptor resulted in a significantly higher value than control. H3R antagonism did not modify HA effect (Fig. 2d). Treatment with HA receptors antagonists in the absence of HA did not modify control values.

Effect of HA on TUNEL-positive cells

In order to asses whether the increase in proliferation caused by HA was associated with a decrease in programmed cell death, we compared the number of TUNEL-positive cells in controls and HA-treated cells, at the end of the proliferative phase. Cell death under control conditions was 2.8 ± 0.3% of total cells. At all concentrations, HA produced significant decreases in TUNEL positive cell numbers (Fig. 3a), with the maximum effect at 1 μM (59% reduction relative to the control) and the minimum at 100 nM (32% less than control; Fig. 3b). Since activation of H2R was necessary to observe HA-promoting effects on proliferation, we evaluated if blockade of this receptor could also prevent the decrease in cell death observed with 100 μM HA. In this set of experiments (n = 3), the proportion of TUNEL-positive cells were: control: 3.5 ± 0.3%; 30 μM cimetidine: 3.4 ± 0.1%; 100 μM HA: 1.7 ± 0.1% (p ≤ 0.05 relative to control); 100 μM HA ± 30 μM cimetidine: 13.0 ± 0.8% (p ≤ 0.001 vs. both control and 100 μM HA).

image

Figure 3.  Histamine reduces apoptosis in proliferative NSC. Cells exposed to 10 ng/mL bFGF and treated with increasing concentrations of HA showed a significant decrease on TUNEL-positive cells. (a) Representative micrographs of the effect of HA on TUNEL-positive cells (green, arrows), showing the reduction in the proportion of green cells that are positive, relative to the nuclei stained with Hoechst (blue). Scale bar = 50 μm. (b) Quantification of the total TUNEL-positive cells from eight fields in duplicate from three independent experiments. Results are means ± SEM. (*)p < 0.05 and (**)p < 0.001 vs. control.

Download figure to PowerPoint

Effect of HA on differentiated NSC

To promote differentiation of cortical NSC to neuronal and glial lineages, bFGF was excluded from the medium and cells were maintained in N2 medium during 6 days in control cultures, with daily HA treatments at the indicated values in the concentration-response curve, or with daily 100 μM HA in the presence of H1R, H2R or H3R antagonists.

Effect of HA on the number of differentiated cells

The effect of HA on cell number after NSC differentiation was next examined. Cell number estimated by the crystal violet assay showed no differences in any of the experimental condition versus control cells (Fig. 4).

image

Figure 4.  Effect of histamine on cell number after differentiation. Differentiated cells were maintained in N2 medium without bFGF and treated daily with increasing concentrations of HA. After 6 days, HA effect on the final number of cells was estimated by crystal violet assay. Results are means ± SEM from 3–5 experiments. Results are expressed as the percent change compared with control.

Download figure to PowerPoint

Effect of HA on TUNEL positive cells

To evaluate cell death after 6 days of bFGF withdrawal, we performed TUNEL detection at the end of differentiation. Control apoptotic cell death was 22.8 ± 1.3% relative to the total number of nuclei detected by Hoechst. The treatment with increasing concentrations of HA showed a significant dose-dependent increase in the proportion of TUNEL-positive cells at 10 μM, 100 μM and 1 mM HA, with 25%, 43% and 60% increases on TUNEL positive cells, respectively, when compared with control conditions. Best-fit adjustment to a sigmoidal dose-response curve yielded a maximum effect of 35.7 ± 1.1% of TUNEL-positive cells and revealed an EC50 = 13.0 ± 0.2 μM of HA (n = 3; Fig. 5).

image

Figure 5.  Apoptosis is increased by histamine in differentiated cells. Cells treated with increasing concentrations of HA during 6 days showed a significant dose-dependent increase on TUNEL-positive cells. (a) Representative micrographs of the effect of HA on TUNEL positive cells (green, arrows), showing an increase in the proportion of apoptotic cells normalized by the Hoechst (blue)-stained nuclei, when compared with control cultures. Scale bar = 50 μm. (b) Concentration-response curve for HA on TUNEL-positive cells. Data are mean ± SEM from the combined values from cells counted from eight fields in duplicate from three independent experiments. The line is the best-fit estimate to a sigmoid curve. Best fit estimates for maximal effect and EC50 are given in the results section. (*)p < 0.05 and (**)p < 0.001 vs. control, 100 nM and 1 μM HA; (#)p < 0.05 vs. 10 μM HA.

Download figure to PowerPoint

Effect of HA on cell phenotypes after NSC differentiation

To investigate whether HA could modify the differentiation of cortical NSC, we quantified the number of neurons and glial cells by performing double immunodetection of MAP2 (a mature neuron marker) and GFAP (astrocytic marker) in control and HA-treated cultures. HA was able to significantly reduce the number of GFAP-positive cells, and also significantly, to increase MAP2-positive cells after treating cells with concentrations from 1 μM up to 1 mM HA (Fig. 6a and b). The maximum increase in the number of neurons was seen at 1 mM HA, increasing form 7.9% to 23.4% (2.96 times more neurons relative to control). The proportion of GFAP-positive cells in differentiated cultures decreased significantly after incubation with concentrations of 1 μM HA and higher, reducing by about 50% the number of glial cells found in controls. In control conditions, the ratio astrocytes/neurons was 5 and this decreased in a range of 0.8 to 1.3 with HA concentrations above 1 μM (Fig. 6b).

image

Figure 6.  Histamine promotes neuronal differentiation by activation of H1 receptors. After proliferation, cells were kept on differentiating conditions during 6 days and treated daily with HA. (a) Representative micrographs of the labeling for microtubule associated protein 2 (MAP2, a mature neuronal marker shown in red), and glial fibrillar acidic protein (GFAP, an astrocytic marker, in green) and nuclear detection by Hoechst (in blue), showing the effect of HA on the number of neurons and glial cells. Scale bar = 5 μm. (b) Quantification of the total number of MAP2-positive or GFAP-positive cells in duplicate from eight fields taken in three independent experiments. No cells were found to co-express MAP2 and GFAP. Results are means ± SEM expressed as the percent of total cells stained with Hoechst. (*)p < 0.05 and (**)p < 0.001 vs. control and 100 nM HA. (c) Representative micrographs of MAP2 (red), GFAP (green) and nuclear detection by Hoechst (blue), showing the antagonistic effect of 1 μM chlorpheniramine (Chlorph.) on the increase in neuronal number caused by 100 μM HA. Scale bar = 5 μm. (d) Pharmacological analysis of HA action on NSC differentiation. HA-treated cells were incubated with 1 μM of the H1R antagonist Chlorpheniramine (Chlorph.), 30 μM of the H2R antagonist cimetidine (Cimet.) or 1 μM of the H3R blocker thioperamide (Thioper.), and the percentage of the total number of MAP2-positive or GFAP-positive cells was quantified from 3–6 independent experiments. Results are mean ± SEM expressed as the percent of total cells stained with Hoechst. (a) p < 0.01 of MAP2-positive cells vs. control condition. (b) p < 0.01 GFAP-positive cells vs. control condition.

Download figure to PowerPoint

Treatments with H1R, H2R or H3R antagonists showed that the effect of HA on neuronal differentiation of neural progenitors is largely due to H1R activation, since 1 μM chlorpheniramine significantly decreased MAP2-positive cells to values similar to control conditions when cells are co-treated with 100 μM HA (from 29.8% to 13.4%), inducing a significant change on the ratio astrocytes/neurons from 0.7 (100 μM HA) and to a value of 2 (100 μM HA + H1R antagonist). In contrast, 30 μM cimetidine was not able to modify either MAP2 positive cells, nor astrocytes/neurons ratio (0.9) from the values observed with 100 μM HA. Thioperamide did not revert the effect of HA, since 31.8% of cells were MAP2-positive and the resulting ratio was 0.7, when 100 μM HA and this H3R antagonist were present. Interestingly, although the number of neurons is not modified by H2R or H3R antagonists, we found significant decreases relative to control conditions in the number of glial cells when either chlorpheniramine, cimetidine or thioperamide were incubated with 100 μM HA. None of the HA receptor antagonists, modified the number of neurons or astrocytes from control values in the absence of HA (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Neural stem cells are important elements of the developing nervous system that can be isolated and grown in vitro to study the role of a number of factors that might affect proliferation and cell fate. In these cells, we show here that HA induce the following effects: (i) expansion of NSC numbers due to an increase in proliferation caused by activation of H2R; (ii) decreased apoptosis in NSC stimulated with bFGF; (iii) a concentration-dependent induction of TUNEL-positive cells in the differentiation phase, and (iv) higher number of neurons after differentiation of NSC, an effect due to H1R activation.

The biogenic amine HA, which acts as a neurotransmitter/neuromodulator in the adult rat CNS (Schwartz et al. 1991; Hill et al. 1997), is present in high concentrations (five times higher than those found in adult brain) in the prosencephalic area at E14, and these levels remain elevated until E17 (Vanhala et al. 1994) suggesting a role of HA in neurogenesis occurring in that period. Histaminergic receptors must be in place for HA to act. Expression of histaminergic receptors at these stages of development has been reported. Hybridization studies show the distribution of mRNA for HA receptors in different brain regions, but no information about the cell types expressing the receptors is provided (Kinnunen et al. 1998; Heron et al. 2001; Karlstedt et al. 2001b, 2003). It has been shown that the cerebral cortical area expresses H1R at E14 (Kinnunen et al. 1998). Little is known about H2R expression before E15, but after this age it can be clearly detected in the cerebral cortex (Karlstedt et al. 2001a). Messenger RNA for H3R is detected in E19 in the cortical area (Karlstedt et al. 2003). There are no reports showing the expression of HA receptors in NSC. We show in this study that cortical NSC express H1R, H2R and all reported isoforms of H3R before and after differentiation, at the mRNA level. Based on hybridization studies, expression of H1R and H2R was expected in NSC, but the presence of H3R was not anticipated, due to the fact that it is expressed from E19 onwards. H1R and H2R were detected by immunoblot as single bands; however, H3R presented several bands that varied in size between NSC and adult cerebral cortex. Similar differences in abundance and masses have been reported between embryonic and adult brown adipose tissue (Karlstedt et al. 2003). Bands below 35 kDa are not compatible with expected sizes of G-protein coupled receptors, and might result from active H3R degradation processing, as suggested earlier (Karlstedt et al. 2003).

In the present study, we found that HA did not modify NSC identity during proliferation, since a high proportion of cells continue to express the intermediate filament protein Nestin, and therefore, HA caused a significant expansion of NSC. Interestingly, HA could induce the appearance of a few cells positive to differentiated cell markers, although these cells are still Nestin-positive. This suggests that HA could promote premature differentiation under proliferating conditions in a discrete population. It remains to be investigated whether a combination of HA with other signals could promote differentiation, even in the presence of bFGF.

Cell proliferation and survival are regulated by different factors (Kilpatrick and Bartlett 1995; Johe et al. 1996; Weiss et al. 1996; Qian et al. 1997). Survival and proliferation of early embryonic neural precursors are regulated in vitro and in vivo by the mitogenic factors bFGF and epidermal growth factor, among others (Kilpatrick and Bartlett 1995; Qian et al. 1997; Ciccolini and Svendsen 1998; Ortega et al. 1998; Vaccarino et al. 1999). We found here that 100 μM HA was able to increase BrdU incorporation as a measure of cell proliferation in the presence of bFGF. This proliferating action of HA could be explained by second messenger mediated actions of HA. It has been reported that HA is able to increase the intracellular Ca2+concentration ([Ca2+]i) in mouse pluripotent stem cells (Bloemers et al. 1993), and Ca2+ions are known to play an important role in proliferation by interacting with the mitogen-activated protein kinase (Bloemers et al. 1993; Berridge et al. 2000; Yanagida et al. 2004).

Based on HA concentrations that have evident effects in this study, we though that either H1R (Traiffort et al. 1994) or H2R (Garbarg and Schwartz 1987) could be implicated, since these two receptors are activated with micromolar concentrations of HA, while H3R is activated in the nanomolar range (Chen et al. 2003; Rouleau et al. 2004) and suffer desensitization at μM concentrations (Perez-Garcia et al. 1998). We choose 100 μM HA to further study if H1R, H2R or H3R were responsible for the effects seen on NSC proliferation and differentiation caused by HA. The increase of bFGF-induced proliferation caused by HA was due to the activation of H2R, as demonstrated by the reversion of BrdU incorporation when cells were incubated with 100 μM HA + 30 μM cimetidine. Although there is no report of the effect of HA and the pathways that are stimulated by this biogenic amine on NSC, there are evidences that H2 receptor activation is linked to different signaling systems: 1. Stimulation of cAMP formation in brain slices (Al-Gadi and Hill 1985), vascular smooth muscle and neutrophils (Hill 1990). 2. Stimulation of phospholipid methylation in rat mast cells (Tolone et al. 1982). 3. Increases in the slow inward Ca2+current in several models such as guinea pig ventricular myocytes, via cAMP formation (Hill 1990). 4. Inhibition of Cl-mediated K+ conductance in hippocampal pyramidal cells (Haas and Greene 1986). 5. Increases of [Ca2+]i mobilization in a human lymphocytic cell line (HL-60) (Mitsuhashi and Payan 1991). Multiple reports demonstrated that single receptors may be associated with more than one G protein, and thus to multiple intracellular signaling systems (Vallar et al. 1990; Van Sande et al. 1990; Gudermann et al. 1992; Raymond 1995; Arai and Charo 1996). Our results open two possibilities by which HA can be regulating cell proliferation by H2R activation: a) Via phosphoinositide/protein kinase C signal transduction cascade (Del Valle and Gantz 1997) or b) By increasing cAMP, since this cyclic molecule stimulate proliferation in many cell types, an effect that is largely attributed to cross-talk from cAMP and the mitogen-activated protein kinase pathway (Dumaz and Marais 2005). The effect of cimetidine on decreasing cell proliferation is in agreement with a study made by Finn et al., in which this H2R blocker also inhibited proliferation in three out of five glial cell lines (Finn et al. 1996). Pharmacological blockade of HA effects rules out the possibility that this amine could be acting on the polyamine site of the NMDA receptor, because such interaction is not susceptible to be interrupted by histaminergic H1 and H2 receptor antagonists (Bekkers 1993; Vorobjev et al. 1993).

In general, stem cell pools result from the contribution of various factors, i.e. rate of proliferation, time of exponential expansion of cell number, ratio of asymmetric to symmetric cell divisions (Caviness and Takahashi 1995), and apoptotic cell death (Blaschke et al. 1996). In the present study, we measured apoptotic cell death levels, in order to estimate the contribution of this factor on the effect of HA increasing cell number during the proliferation phase. A low proportion of TUNEL-positive cells was found in control NSC cultures. These results are in accordance with a study in cortical stem cells, showing low numbers of apoptotic cells (Chang et al. 2004). Our data show that HA was able to further decrease the proportion of TUNEL-positive NSC in a H2R-dependent manner, suggesting that HA is acting both as a proliferating and an anti-apoptotic factor for NSC in the presence of bFGF. The effect of HA increasing cell number during the proliferation phase was not observed after 6 days of differentiation. This could be due to the concentration-dependent increased in the number of TUNEL-positive cells in HA-treated cells which, together with the HA-induced increase in cell number during proliferation, might account for the similar number of cells found in control and HA-exposed cultures. The increased cell death during differentiation could contribute to the neurogenic effect of HA if glial progenitors/cells are induced to undergo cell death, and this will be certainly interesting to investigate further.

About differentiation, our results show that NSC treated daily with micromolar or low millimolar concentrations of HA during the proliferation and differentiation stages generate more neurons and less GFAP-positive cells. Regulation of cell fate acquisition in the vertebrate CNS is dependent on the stage of development. In the rat cerebral cortex, neurogenesis begins at E12, peaks at E14, and recedes by E17 (Sauvageot et al. 2005). In vivo, neurons are generated first, followed by astrocytes, and later by oligodendrocytes, and this behavior is mimicked by cultured NSC, that first generate neurons and then glial progeny (Qian et al. 2000; Morrow et al. 2001; Panchision and McKay 2002; Sauvageot and Stiles 2002). HA concentration peaks in prosencephalon from E14 to E17 (Vanhala et al. 1994), suggesting a role of this biogenic amine in neuronal differentiation. The marked shift in the proportion of neurons induced by HA is consistent with this idea. Although HA could contribute to neuronal differentiation in vivo, it is important to mention that knockout mice for the HA synthesizing enzyme, histidine decarboxylase, do not show any evident alteration in brain development (Watanabe and Yanai 2001). These results are not necessarily opposed to our findings, since there might be redundant mechanisms for neuronal differentiation in the cerebral cortex.

To study which histaminergic receptor is responsible for the increase on the number of neurons in culture, we performed experiments with H1R, H2R and H3R antagonists and 100 μM HA. Our results show that HA increase neuronal differentiation due to activation of H1R. Activation of this receptor leads to production of IP3 and diacylglycerol, that in turn promote an increase on [Ca2+]i due to activation of IP3 receptors in the endoplasmic reticulum, and the activation of protein kinase C. Calcium release from intracellular stores into the cytosol is a critical component during ontogenesis and contributes particularly to the formation and maintenance of dendritic structures (Lohmann et al. 2002, 2005). Regarding astrocyte production, it is interesting to note that neither chlorpheniramine, nor cimetidine, nor thioperamide were able to revert HA effect on decreasing glial differentiation. Notwithstanding, the overall effect of antagonizing H1R in the presence of HA is to block its neuronal-promoting effect.

There are only a few examples of other neurotransmitters having effects on proliferation and differentiation of NSC, and some of them are conflicting. GABA and glutamate decrease cortical precursors proliferation (LoTurco et al. 1995; Antonopoulos et al. 1997), acetylcholine increases cell proliferation (Ma et al. 2000) and dopamine has been shown to promote (Hoglinger et al. 2004) or inhibit cell proliferation (Kippin et al. 2005) of adult NSC. Regarding neurogenesis, a recent report shows that GABA has an effect promoting this process in NSC derived from adult brain (Tozuka et al. 2005). Thus, HA is the first neurotransmitter showing a positive effect on both NSC proliferation and in the proportion of neurons derived from cortical NSC, that correlate with increased HA levels during neuronal differentiation in the cerebral cortex.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We thank Itzel Escobedo and Griselda Rodriguez for technical help. This work was supported by grants IN226703 and IN224207 from DGAPA, Universidad Nacional Autónoma de México. The Nestin monoclonal antibody developed by Dr. Hockfield was obtained from the Developmental Studies Hybridoma Bank under the auspices of NICHD and maintained by the University of Iowa.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References