Proximal events in 7,12-dimethylbenz[a]anthracene-induced, stromal cell-dependent bone marrow B cell apoptosis: stromal cell-B cell communication and apoptosis signaling

Abstract

Intercellular communication is an essential process in stimulating lymphocyte development and in activating and shaping an immune response. B cell development requires cell-to-cell contact with and cytokine production by bone marrow stromal cells. However, this intimate relationship also may be responsible for the transfer of death-inducing molecules to the B cells. 7,12-Dimethylbenz[a]anthracene (DMBA), a prototypical polycyclic aromatic hydrocarbon, activates caspase-3 in pro/pre-B cells in a bone marrow stromal cell-dependent manner, resulting in apoptosis. These studies were designed to examine the hypothesis that an intrinsic apoptotic pathway is activated by DMBA and that the ultimate death signal is a DMBA metabolite generated by the stromal cells and transferred to the B cells. Although a loss of mitochondrial membrane potential did not occur in the DMBA/stromal cell-induced pathway, cytochrome c release was stimulated in B cells. Caspase-9 was activated, and formation of the apoptosome was required to support apoptosis, as demonstrated by the suppression of death in Apaf-1(fog) mutant pro-B cells. Investigation of signaling upstream of the mitochondria demonstrated an essential role for p53. Furthermore, DMBA-3,4-dihydrodiol-1,2-epoxide, a DNA-reactive metabolite of DMBA, was sufficient to upregulate p53, induce caspase-9 cleavage, and initiate B cell apoptosis in the absence of stromal cells, suggesting that production of this metabolite by the stromal cells and transfer to the B cells are proximal events in triggering apoptosis. Indeed, we provide evidence that metabolite transfer from bone marrow stromal cells occurs through membrane exchange, which may represent a novel communication mechanism between developing B cells and stromal cells.

Fig. 1

DMBA induces caspase-independent mitochondrial cytochrome c release, but not membrane potential loss, in B cells co-cultured with bone marrow stromal cells. (A) BU-11/BMS2 co-cultures were treated with vehicle (Vh, 0.1% DMSO) or DMBA (1 μM), and BU-11 cells were harvested after 2–8 hr. Cytochrome c release was analyzed by immunoblotting of cytoplasmic extracts from digitonin-permeabilized cells. Data are representative of 3 experiments. (B) BU-11/BMS2 co-cultures were treated with Vh or DMBA (1 μM) or BU-11 cultures were treated with Vh or etoposide (0.1 μg/ml) (inset) for 4–10 hrs. BU-11 cells were analyzed for mitochondrial membrane potential loss by JC-1 staining followed by flow cytometry. Data are are presented as means + SE from at least three experiments. (C) BU-11/BMS2 co-cultures were pre-treated with Vh or the pan-caspase inhibitor VAD-FMK (30 μM) for 30 min prior to treatment with Vh or DMBA (1 μM) for 8 hr. Cytochrome c release was analyzed as above. Data are representative of 3 experiments. *Statistically greater than Vh treated (p<0.05, ANOVA, Tukey-Kramer).

Fig. 2

The apoptosome is essential for DMBA/stromal cell-induced B cell apoptosis. BU-11/BMS2 co-cultures were treated with Vh or DMBA (1 μM), and BU-11 cells were harvested after 2–10 hr. (A) Total proteins were extracted and analyzed for active caspase-9 fragments and for β-actin by immunoblotting. Data are representative of 4 experiments. (B) Cytosolic proteins were extracted, and caspase-9-like activity was measured using p-nitroaniline-conjugated LEHD substrate. Data were quantified as the average fold increase in caspase-9-like activity relative to the activity in untreated cells and presented as means ±SE from 4 experiments. (C) Primary proB cells isolatedfrom mice either homozygous or heterozygous for the Apaf-1^fogmutation, were treated in co-culture with BMS2 with Vh or DMBA (1 μM). Pro-B cells were analyzed for apoptosis by PI staining 16 hr after treatment. The percentage of death measured in naïve cell populationswas subtracted prior to analysis. Data are presented as means + SE from primary pro-B cells prepared from 5–6 individual mice. *Statistically different from Vh (p<0.05, ANOVA, Dunnett’s).* *Statistically greater than all other groups(p<0.05, ANOVA, T ukey-Kramer).

Fig. 3

Contribution of p53 to DMBA/stromal cell-induced B cell apoptosis. (A) BU-11/BMS2 co-cultures were treated with Vh or DMBA (1 μM), and BU-11 cells were harvested after 2–10 hr. Total proteins were extracted and analyzed for p53 and for β-actin by immunoblotting. Data are representative of 3 experiments. (B, C) Primary pro-B cells isolated from wildtype or p53 mutant mice were treated in co-culture with BMS2 cells with Vh or DMBA (1 μM) for 2–16 hr. (B) Pro-B cells were analyzed for apoptosis by PI staining 16 hr after treatment. The percentage of death measured in naïve cells was subtracted prior to analysis. Data are presented as means + SE from primary pro-B cells prepared from 3–4 individual mice. (C) Total proteins were extracted from pro-B cells and analyzed for cleaved caspase-3 and for β-actin by immunoblotting. A caspase-3 positive control was included on each gel to enable comparison between blots. Data are representative of experiments using primary pro-B cells from 3 individual mice. *Statistically greater than all other treatment groups (p<0.05, ANOVA, Tukey-Kramer). **Statistically less than wildtype DMBA treated (p<0.05, ANOVA, Tukey -Kramer).

Fig. 4

A terminal metabolite of DMBA induces stromal cell-independent B cell apoptosis. BU-11 cell suspension cultures were treated with Vh or DMBA-DE (1 μM) for the indicated times. (A) BU-11 cells were analyzed for apoptosis by PI staining. The percentage of death measured in naïve cells was subtracted prior to analysis. Data are presented as means + SE from at least 3 experiments. (B) Total proteins were extracted from BU-11 cells and analyzed for p53 and β-actin by immunoblotting. Data are representative of at least 3 experiments. (C) Primary pro-B cells isolated from wildtype orp53 mutant mice were treated with Vh or DMBA-DE (0.01 μM). Pro-B cells were analyzed for apoptosis by PI staining 16 hr after treatment. The percentage of death measured in naïve cells was subtracted prior to analysis. Data are presented as means + SE from primary pro-B cells prepared from 4 individual mice. *Statistically different from Vh (p<0.05, ANOVA, Dunnett’s).

Fig. 5

Cell contact is required for DMBA/stromal cell-induced apoptosis, potentially facilitating metabolite transfer from BMS2 to BU-11 cells. (A) BMS2 cells were cultured in plates as usual or in 3 μm pore size transwells. BU-11 cells were added directly to BMS2 monolayers or to transwells suspended over BMS2 monolayers. Co-cultures were treated with Vh or DMBA (1 μM) for 16 hr. BU-11 cells were analyzed for apoptosis by PI staining. Data are presented as means + SE from 3 experiments. (B-D) BMS2 cultures were treated with 10 μM DMBA for 30 min, washed, incubated for 1 hr, then washed again. BU-11 cells then were centrifuged onto the BMS2 cell layer and cultures immediately analyzed by two-photon microscopy for DMBA/metabolite fluorescence transfer. (B) False color image of BU-11 cells in contact with the BMS2 cell monolayer overlayed on the BMS2 cell image. (C) Image of BU-11 cells in contact with the BMS2 cell monolayer. (D) BU-11 cells not in contact with the BMS2 cell monolayer. Representative data from one of three similar experiments are presented. BU-11 cells in contact with BMS2 cells tended to contain a higher level of fluorescence than those no it contact with BMS2 (p<0.06). (E, F) BMS2 cultures were treated with Vh or DMBA (10 μM) for 30 min, washed extensively, incubated for 5 hr, and washed again. BU-11 cells then were added directly to the BMS2 cell layer or to a 1 μm pore size transwell insert above the BMS2 cells. BU-11 cells were analyzed at 18 hr by flow cytometry for fluorescence at 450 nm. (E) Representative fluorescence histogram of BU-11 cells cultured on naïve (filled) or DMBA-treated (open) BMS2 cells. (F) DMBA/metabolite fluorescence expressed as percent of naïve fluorescence. Data are presented as the means + SE from 3 experiments. *Statistically greater than other treatment groups (p<0.05, ANOVA, Tukey-Kramer). **Statistically different from naïve and from DMBA in transwell groups (p < 0.05, ANOVA, Tukey-Kramer).

Fig. 6

B lineage cells take up membrane lipid from bone marrow stromal cells in a contact-dependent manner. BMS2 cells were left unstained or were labeled with Vybrant DiO plasma membrane dye according to the manufacturer’s protocol. (A) BU-11 cells were added directly to unstained or stained adherent BMS2 cells (Contact) or in a 3 μm pore size transwell insert over the BMS2 cells (Transwell) and harvested after 16 hr. In addition, naïve BU-11 cells were resuspended for 20 min in 16 hr supernatants from unstained or DiO-stained BMS2 cells (Supernatant). DiO uptake was analyzed by flow cytometry in live-gated cells. Inset: The filled gray histogram representsdata from BU -11 cells cultured in contact with unstained BMS2 cells, the open gray lined histogram represents data from BU-11 cells cultured in transwells over DiO-stained BMS2 cells, and the open black lined histogram represents data from BU-11 cells cultured in contact with DiO-stained BMS2 cells. (B) BU-11 cells were added directly to unstained or stained adherent BMS2 cells, harvested after 0.5–2 hrs, and analyzed by flow cytometry. (C, D) Primary pro-B cells, splenic B cells, or splenic T cells from wildtype C57Bl/6 mice were added to unstained or DiO-stained BMS2 cells directly, and lymphocytes were assayed 16 hr later by flow cytometry. (C) Representative histograms of lymphocytes on unstained (filled) or DiO-stained (open) BMS2 cells. (D) Quantification of DiO positive lymphocytes. Data are presented as means + SE from 3–4 experiments or from 3–5 experiments with cells from individual mice. *Statistically different from lymphocytes on unstained BMS2 cell controls (p< 0.05, ANOVA, Tukey-Kramer).