An endogenous vitamin K-dependent mechanism regulates cell proliferation in the brain subventricular stem cell niche

Abstract

Neural stem cells (NSC) persist in the adult mammalian brain, within the subventricular zone (SVZ). The endogenous mechanisms underpinning SVZ stem and progenitor cell proliferation are not fully elucidated. Vitamin K-dependent proteins (VKDPs) are mainly secreted factors that were initially discovered as major regulators of blood coagulation. Warfarin ((S(-)-3-acetonylbenzyl)-4-hydroxycoumarin)), a widespread anticoagulant, is a vitamin K antagonist that inhibits the production of functional VKDP. We demonstrate that the suppression of functional VKDPs production, in vitro, by exposure of SVZ cell cultures to warfarin or, in vivo, by its intracerebroventricular injection to mice, leads to a substantial increase in SVZ cell proliferation. We identify the anticoagulant factors, protein S and its structural homolog Gas6, as the two only VKDPs produced by SVZ cells and describe the expression and activation pattern of their Tyro3, Axl, and Mer tyrosine kinase receptors. Both in vitro and in vivo loss of function studies consisting in either Gas6 gene invalidation or in endogenous protein S neutralization, provided evidence for an important novel regulatory role of these two VKDPs in the SVZ neurogenic niche. Specifically, we show that while a loss of Gas6 leads to a reduction in the numbers of stem-like cells and in olfactory bulb neurogenesis, endogenous protein S inhibits SVZ cell proliferation. Our study opens up new perspectives for investigating further the role of vitamin K, VKDPs, and anticoagulants in NSC biology in health and disease.

Figure 1

Regulation of subventricular zone (SVZ) cell culture growth, proliferation, and apoptosis by warfarin ((S(−)-3-acetonylbenzyl)-4-hydroxycoumarin)). Representative photographs of SVZ neurospheres maintained for 5 days either in serum-free medium (SFM) (A) or in SFM supplemented with 1 μg ml⁻¹ warfarin (B). Scale bar = 200 μm. (C): Growth of SVZ cell cultures maintained for 5 days in either SFM (control) or in SFM supplemented with either 1 μg ml⁻¹ warfarin or with 20 ng ml⁻¹ epidermal growth factor. Percentages of bromodeoxyuridine- (D) or terminal deoxynucleotidyl transferase dUTP nick end labeling-stained (E) nuclei in SVZ cell cultures maintained for 24 hours in SFM (control) or SFM supplemented with 1 μg ml⁻¹ warfarin. (F): Growth of SVZ cell cultures maintained for 5 days in either SFM (control) or SFM supplemented with 1 μg ml⁻¹ warfarin. SVZ cells were derived from newborn or adult rats or mice. Data were obtained from at least three independent experiments each in quadruplicates and are expressed in (C) and (F) as percentages of total viable cell numbers in SFM ± SEM, in (D) and (E) as percentages ± SEM. We used Mann-Whitney test for data statistical analysis. ***, p < .0001; **, p < .01; ns, p = .0570. Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

Figure 2

Intracerebroventricular injection of warfarin ((S(−)-3-acetonylbenzyl)-4-hydroxycoumarin)) stimulates cell proliferation in vivo in the regions surrounding the lateral ventricles. Immunostaining of bromodeoxyuridine (BrdU) (in dark, A–D) in brain sections following a single injection of NaCl (1 μl, 9 mg ml⁻¹) (A, B) or warfarin (1 μl, 100 mg ml⁻¹) (C, D) in the lateral ventricle of mice. Mice were subjected to intraperitoneal injections of BrdU (50 mg kg⁻¹) 68 hours later and sacrificed 72 hours postintracerebroventricular injections. Scale bars = 1 mm (A, C); 200 μm (B, D). In (E) and (F): BrdU-positive cells were quantified within and in the region (region 2) adjacent to the subventricular zone on sections sampled between 1.18 and 0.14 mm anterior to Bregma, on hemispheres contralateral (contra) or ipsilateral (ipsi) to warfarin or NaCl injection. Means ± SEM numbers of BrdU-positive cells per brain section obtained in regions 1 and 2 from six 4-month-old mice per experimental condition were analyzed with one-way analysis of variance (ANOVA) followed by the post hoc Bonferroni test. *, p < .05; **, p < .01; ***, p < .0001. Double immunostaining for BrdU (green) and glial fibrillary acidic protein (GFAP) (red, G and H) or doublecortin (red, J and K) in ipsilateral brain hemispheres of mice injected with NaCl (G, J) or warfarin (H, K). Scale bar = 40 μm. Each labeled cell was examined along the z-axis to ensure proper identification of double labeled cells. Percentages of GFAP-(I) or Dcx-over BrdU (L)-positive cells in ipsilateral brain hemispheres of mice injected with NaCl (white bars) or with warfarin (dark bars). Means ± SEM obtained from six 4 months old mice per experimental condition were analyzed with one-way ANOVA followed by the post hoc Bonferroni test. *, p < .05; **, p < .01; ***, p < .0001. Abbreviations: BrdU, bromodeoxyuridine; DCX, doublecortin; GFAP, glial fibrillary acidic protein; ipsi, ipsilateral.

Figure 3

Subventricular zone (SVZ) cells express the gamma carboxylase and VKOR enzymes and produce active vitamin K-dependent proteins that reverse warfarin ((S(−)-3-acetonylbenzyl)-4-hydroxycoumarin)) effects. (A): Real time-polymerase chain reaction (RT-PCR) detection of the γ-carboxylase and VKOR transcript in SVZ cells. RT-PCR analysis of RNA obtained from reverse transcriptase two different (lanes 1 and 2) SVZ cell cultures. Lane 3 corresponds to DNA molecular weight markers. Lane four depicts RT-PCR products obtained using RNA from tissues (liver) known to express both transcripts (positive control). Lane 5 represents RT-PCR performed in the absence of reverse transcription (negative control). Micrographs representing immunostaining of SVZ cell cultures with either anti-γ-carboxylase (B, in green) or anti-VKOR (C, in green) antibody. Cell nuclei were labeled with TOPRO-3 (blue). Scale bar = 20 μm. (D) Growth of SVZ cell cultures maintained for 5 days in either serum-free medium (SFM; control) or in the presence of 1 μg ml⁻¹ warfarin or 10 μg ml⁻¹ vitamin K1 (vit K) or a combination of both. (E) Growth of SVZ cell cultures maintained for 5 days in SFM (control) or in SFM supplemented with either 1 μg ml⁻¹ warfarin or conditioned media derived from SVZ cell cultures that had been treated for 5 days in either SFM (CM) or 10 μg ml⁻¹ vitamin K1 (vitK-CM) or 1 μg ml⁻¹ warfarin (warfarin-CM) or a combination of the above. Data were obtained from at least three independent experiments each in quadruplicates and are expressed as percentages of total viable cell numbers in SFM ± SEM in (D) or of total viable cells in CM in (E). We performed statistical analysis by one-way analysis of variance followed by the post hoc Bonferroni test for multiple comparisons. *, p < .05; **, p < .01; ***, p < .0001. Abbreviations: CM, conditioned media; VKOR, vitamin K epoxide reductase.

Figure 4

Subventricular zone (SVZ) cells produce Gas6 and protein S in vitro and in vivo. (A) Real time-polymerase chain reaction (RT-PCR) detection of secreted vitamin K-dependent proteins (VKDPs) in SVZ cells cultures. RT-PCR reverse transcriptase analysis of RNA obtained from two different (lanes 1 and 2) SVZ cell cultures using oligonucleotides specific for the indicated VKDP c-DNA or GAPDH used as a control. Lane 3 corresponds to DNA molecular weight markers. Lane 4 depicts RT-PCR products obtained using RNA from tissues known to express the indicated VKDP transcripts (positive control) and which are liver for thrombin, factor VII, IX, and X, protein C, protein S, and choroid plexus for Gas6. Lane 5 represents RT-PCR performed in the absence of reverse transcription (negative control). Western blotting analysis of conditioned media obtained from warfarin ((S(−)-3-acetonylbenzyl)-4-hydroxycoumarin)) (W) or vitamin K1 (K) treated SVZ cells (B-D) using either an anti-γ-carboxyglutamate residues antibody (B) or an anti-Gas6 antibody (C) or an anti-protein S antibody (D). Purified human protein S or murine Gas6 produced by human HEK293 cells, as described in the material and method section were used as positive controls. (E) Mice brain sections from the SVZ were doubleimmunostained using anti-protein S (green) or anti-Gas6 (green) antibody together with anti-nestin (red), anti-glial fibrillary acidic protein (red) or anti-doublecortin antibody. Cell nuclei were stained with DAPI (blue). Scale bar = 60 μm. Brain sections were also stained with anti-protein S (red) and anti-CD31 (green) antibody. Asterisks show protein S localization in blood vessel. Scale bar = 40 μm. Abbreviations: DCX, doublecortin; GAPDH, glyceraldehade-3-phosphate dehydrogenase; GFAP, glial fibrillary acidic protein; LV, lateral ventricle; NSC, neural stem cell.

Figure 5

Subventricular zone (SVZ) cells express Tyro3, Axl, and Mer (TAM) receptors that are activated by Gas6 or protein S. (A) Real time-polymerase chain reaction (RT-PCR) analysis of RNA obtained from two different (lanes 1 and 2) SVZ cell cultures using oligonucleotides reverse transcriptase specific for TAM or GAPDH used as a control. Lane 3 corresponds to DNA molecular weight markers. Lane 4 corresponds to RT-PCR products obtained using RNA from a tissue (testis) known to express TAM receptors (positive control) and lane 5 to RT-PCR performed in the absence of reverse transcription (negative control). (B) Western blotting analysis of 50–100 μg of protein of SVZ cell lysates (neural stem cell) using either anti-Tyro3, anti-Axl, or anti-Mer antibodies. Protein lysates (50–100 μg) from testis were used as positive control. (C): Immunostaining (red) of Axl, Tyro3, or Mer on SVZ mice brain sections. Scale bar = 60 μm. (D, E): Coimmunostaining of Axl (red) or Tyro3 (red) along with nestin or glial fibrillary acidic protein (GFAP) or βIII tubulin (green, D) or GFAP (green, E) and nestin (blue, E). Cell nuclei were stained with DAPI (blue D or cyan in E). Scale bar = 60 μm. (F–H): Western blotting analysis of cell lysates obtained from SVZ cell cultures maintained in serum-free medium (SFM) or in SFM supplemented with 10 μg ml⁻¹ of Gas6, Protein S or both for 8 minutes. Cell lysates of 80 μg were either immunoprecipitated with anti-Tyro3 or anti-Mer antibody and immunoprecipitates analyzed by Western blotting (WB) with antiphosphotyrosine (P-Tyr), anti-Tyro3, and anti-Mer or were directly analyzed by Western blotting using either an anti-phospho-Axl antibody (P-Axl) or an anti-Axl antibody. The bands on the blots were quantified by densitometry and values were presented as means ± SEM. The statistical evaluation of the data was performed with ANOVA. * p < .05, ** p < .01. Abbreviations: GAPDH, glyceraldehade-3-phosphate dehydrogenase; GFAP, glial fibrillary acidic protein.

Figure 6

Gas6 knock out reduces the numbers of subventricular zone (SVZ) cells with stem-like cell properties. (A): Growth of SVZ cell cultures maintained for 5 days in serum-free medium (SFM) (control) or in SFM supplemented with 1 μg ml⁻¹ Gas6. (B): Numbers of primary neurospheres, obtained after incubating SVZ cells derived from wild type (black box) or Gas6^−/− (white box) for 5 days in SFM supplemented with 20 ng ml⁻¹ epidermal growth factor (EGF) are depicted. Neurospheres obtained in each of the conditions were harvested, dissociated as single cells, replated in SFM supplemented with 20 ng ml⁻¹ EGF for 5 additional days and the numbers of secondary neurospheres are also depicted. (C): Numbers of slowly dividing SVZ stem-like cells in wild type or Gas6^−/− mice were obtained by subjecting mice to daily intraperitoneal injections of 50 mg kg⁻¹ bromodeoxyuridine (BrdU) for 5 days and analyzing 24 days later BrdU-positive cells in the SVZ. (D): Numbers of newly generated cells that joined the olfactory bulb in wild-type or Gas6^−/− mice were obtained by subjecting mice to three intraperitoneal injections of 50 mg kg⁻¹ BrdU with a 2-hour interval between each injection during 1 day and analyzing 24 days later, the numbers of BrdU-positive cells within the olfactory bulb. (E): Double immunostaining of BrdU-positive cells with the neuronal marker NeuN in the olfactory bulb, scale bar = 8 μm. (F): Percentages of NeuN (neuronal marker)-positive cells among the BrdU-positive cells in the olfactory bulb in wild-type and Gas6^−/− mice. Data were obtained from at least three experiments (A, B) or six mice (C, D, F) and are depicted as means of percentages of total cell numbers in (A) or numbers of immunoreactive cells per brain section in (C) or density of BrdU-positive cells in (D) or percentages of NeuN-positive cells among BrdU-positive cells in (F). Mann-Whitney test was used for data statistical analysis; ns, p > .05 for (A) and (F); *, p < .05 for (C, D). Abbreviation: BrdU, bromodeoxyuridine.

Figure 7

Protein S inhibits subventricular zone (SVZ) cell proliferation and reverses warfarin ((S(−)-3-acetonylbenzyl)-4-hydroxycoumarin)) effects. (A): Growth of SVZ cell cultures maintained for 5 days in serum-free medium (SFM) (control) or in SFM supplemented with 1 μg ml⁻¹ protein S. (B): Growth of SVZ cell cultures maintained for 5 days in SFM containing 1 μg ml⁻¹ warfarin in absence or presence of 1 μg ml⁻¹ protein S. (C): Growth of SVZ cell cultures maintained for 5 days in SFM containing 10 μg ml⁻¹ vitamin K in the absence or in the presence of 20 μg ml⁻¹ rabbit anti-protein S or an irrelevant rabbit antibody (anti-glial fibrillary acidic protein). (D): Numbers of bromodeoxyuridine (BrdU)-positive cells in the SVZ of 6 months old mice that were subjected to intracerebroventricular injection of anti-protein S or irrelevant antibody (UnAb 1 μl of a 40 μg ml⁻¹ solution). Mice received an intraperitoneal injection of BrdU (50 mg kg⁻¹) 68 hours later and were sacrificed 72 hours postintracerebroventricular injections. Data were derived from four mice injected with irrelevant (Un-Ab) antibody and three with anti-protein S. Data were obtained from at least three independent experiments each in quadruplicates and represent means ± SEM of percentages of total cell numbers in (A), (B), or (C) or of numbers of immunoreactive cells in the SVZ per brain section in (D). Statistical analysis of the data was performed using the Mann-Whitney test in (A), (B), and (D) or the one-way analysis of variance followed by the post hoc Bonferroni test in C. *, p < .05; **, p < .01; ***, p < .001. Abbreviations: PS-Ab, anti-protein S antibody; Un-Ab, unrelated antibody.