Environmental and Endogenous Peroxisome Proliferator-Activated Receptor γ Agonists Induce Bone Marrow B Cell Growth Arrest and Apoptosis: Interactions between Mono(2-ethylhexyl)phthalate, 9-cis-Retinoic Acid, and 15-Deoxy-Δ^12,14-prostaglandin J₂

Reagents

9-Cis-RA was obtained from Biomol (Plymouth Meeting, PA). The PPARγ-specific Ab was purchased from Calbiochem (San Diego, CA). 15d-PGJ₂ was obtained from Cayman Chemical (Ann Arbor, MI). Plasmocin was purchased from InvivoGen (San Diego, CA). Murine rIL-7 was obtained from Research Diagnostics (Flanders, NJ). The NF-κB and p27^kip1-specific Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Troglitazone was obtained from the Sankyo (Tokyo, Japan). Propidium iodide and Protease Inhibitor Cocktail for Mammalian Cells were obtained from Sigma-Aldrich (St. Louis, MO). MEHP was purchased from TCI America (Portland, OR). All other reagents were obtained from Fisher Scientific (Suanee, GA).

Plasmids

The mouse PPARγ expression plasmid pSV-Sport1-PPARγ1 (46) was provided by J. K. Reddy (Northwestern University Medical School, Chicago, IL). The firefly luciferase (Luc) reporter plasmid pHDx3-luc contains three copies of a PPRE derived from the rat enoyl CoA hydratase/3-hydroxyacyl CoA promoter (nt −2956 to −2919) cloned into carbamoyl-phosphate synthetase-Luc reporter (47) and was obtained from Dr. J. Capone (McMaster University, Ontario, Canada). The reporter plasmid pLuc-4A6–880, containing 880 nt of the 5′-flanking DNA of the rabbit CYP4A6 gene cloned into p19Luc was provided by Dr. E. Johnson (The Scripps Research Institute, La Jolla, CA). The Renilla luciferase reporter plasmid pRL-CMV was purchased from Promega (Madison, WI).

Bone marrow B cell isolation

B cells were isolated from the bone marrow of 4- to 6-wk-old C57BL/6 male mice and cultured in RPMI 1640 containing 10% FBS, l-glutamine, penicillin/streptomycin, 2-ME, and 16 ng/ml murine rIL-7 for 5–7 days by the method of Tze et al. (48). Before experimentation, these cells (>80% B220⁺/CD43⁺) were purified further using anti-B220 microbeads and a magnetic column (MACS; Miltenyi Biotec, Auburn, CA). After purification, 95.1 ± 0.5% of the cells were B220⁺/CD43⁺.

Cell lines

The stromal cell-dependent, C57BL/6-derived BU-11 cell line has been characterized previously (49). BU-11 cells represent B cells at the transition between the pro- and early pre-B cell stages as they are CD43⁺/B220⁺/IgM⁻ with rearranged Ig H chain genes (50). BMS2 is a culture dish-adherent, cloned bone marrow stromal cell line that supports BU-11 cell growth (51). Stocks of BU-11 cells were maintained on BMS2 cell monolayers in an equal mixture of DMEM and RPMI 1640 medium with 5% FBS, plasmocin, l-glutamine, and 2-ME. For experiments, BU-11 cells were plated without BMS2 in the presence of murine rIL-7 (16 ng/ml). All cultures were maintained at 37°C in a humidified, 7.5% CO₂ atmosphere. COS-1 cells (American Type Culture Collection, Manassas, VA) were grown in DMEM containing 10% FBS (Sigma-Aldrich) and 50 U/ml penicillin/streptomycin (Invitrogen Life Technologies, Carlsbad, CA). Cell cultures were determined to be mycoplasma negative by PCR (Mycoplasma Detection Kit; American Type Culture Collection).

[³H]Thymidine incorporation

Primary bone marrow B cells or BU-11 cells were cultured in 96-well plates in medium containing 5% FBS and treated with ethanol (vehicle; final concentration, 0.5%), MEHP (10–200 μM), 15d-PGJ₂ (0.5–20 μM), and/or 9-cis-RA (10 μM) for 24 h. ³H-labeled thymidine (1 μCi/well) was added, and the incubation was continued for 24 h. Cells were harvested using a cell harvester (PHD, Cambridge, MA). The [³H]thymidine retained on the filter was detected using a scintillation counter (Beckman Coulter, Fullerton, CA).

Analysis of apoptosis

For propidium iodide (PI) staining, primary bone marrow B cells or BU-11 cells were cultured in 24-well plates in medium containing 5% FBS and were treated with ethanol (vehicle; final concentration, 0.1%), MEHP (50–200 μM), 15d-PGJ₂ (0.5–10 μM), and/or 9-cis-RA (10–20 μM) for 8–24 h. Cells were harvested by gentle pipetting; washed once with cold PBS containing 5% FBS and 10 μM azide; resuspended in 0.25 ml of hypotonic buffer containing 50 μg/ml PI, 1% sodium citrate, and 0.1% Triton X-100; and analyzed on a FACScan flow cytometer (BD Biosciences, Mountain View, CA). For apoptosis studies, cells were monitored with Fl-2 output in the log mode. The percentage of cells undergoing apoptosis was determined to be those with a weaker PI fluorescence than cells in the G₀/G₁ phase of the cell cycle (49).

For determination of DNA fragmentation, BU-11 cells were cultured in T25 flasks in medium containing 5% FBS and treated with ethanol (vehicle; final concentration, 0.1%), MEHP (50–200 μM), 15d-PGJ₂ (2–10 μM), and/or 9-cis-RA (10–20 μM) for 24 h. DNA was isolated and processed as previously described (31). Before loading on a 1.5% agarose gel, the DNA was quantified and treated with 10 μg/ml RNase A at 37°C for 10 min. DNA was visualized using ethidium bromide staining.

Analysis of cell cycle arrest

Cells were treated as described above for 48 h and stained with PI. Stained cells were analyzed by flow cytometry with Fl-2 output in the linear mode.

The expression of p27^Kip1 was determined by Western blotting. BU-11 cells were cultured in T75 flasks in medium containing 5% FBS and treated with ethanol (vehicle; final concentration, 0.1%), MEHP (50–100 μM), 15d-PGJ₂ (0.5–2 μM), and/or 9-cis-RA (10 μM) for 48 h. Whole cell lysates were prepared as described previously (52). Protein concentrations were determined by the Bradford method. Proteins (30 μg) were resolved on 12% gels, transferred to a 0.2-μm nitrocellulose membrane, and incubated with polyclonal p27^Kip1-specific Ab (SC-776). The secondary Ab was HRP-linked goat anti-rabbit IgG (Bio-Rad, Hercules, CA). Immunoreactive bands were visualized with ECL and analyzed on a densitometer (Molecular Dynamics, Sunnyvale, CA). To control for equal protein loading, blots were reprobed with a β-actin-specific Ab (Sigma-Aldrich) and analyzed as described above. An untreated BU-11 cell control was run with each Western blot, and the data from experimental samples were normalized by dividing by the density measured in the untreated control.

Analysis of chemical interactions; additive vs synergistic effects

BU-11 cells were cultured in 96-well plates in medium containing 5% FBS; treated with ethanol (vehicle; final concentration, 0.5%), MEHP (25–100 μM), 15d-PGJ₂ (0.5–2.5 μM), and/or 9-cis-RA (1–10 μM) for 48 h; and assessed for [³H]thymidine incorporation as described above. The method of isoboles was used to analyze combinations of MEHP with 9-cis-RA or 15d-PGJ₂ for synergistic effects (53, 54). The dose of x_a is plotted on one horizontal axis, x_b on the other horizontal axis, and the response on the vertical (z) axis. The dose-response curve for each individual compound lies on the axes. The contours of this surface are called isoboles and denote curves of equal response. The shape of the isoboles provides qualitative information about interactions. Isoboles that are approximately linear with a negative slope suggest an additive effect, using the definition of concentration or Loewe additivity (53, 54). Isoboles that bow toward the origin indicate a greater response than that suggested by additivity, i.e., synergy. The contour plots were created in the R statistical environment (version 1.7.1 for Windows XP; R Foundation, Vienna, Austria; www.r-project.org).

Transient transfections

COS-1 cells were cultured in a 48-well plate and transfected 24 h later using FuGene 6 transfection reagent (Roche, Mannheim, Germany) as previously described (16). The transfection mixture contained 90 ng of luciferase reporter plasmid, 5 ng of PPARγ expression plasmid, and 1 ng of pRL-CMV per well in a volume of 15 μl of DMEM containing 0.3 μl of FuGene 6. Salmon sperm DNA (Stratagene, La Jolla, CA) was added as carrier DNA to give a total of 250 ng of DNA/well. Culture medium was replaced 16–18 h later with serum-free DMEM containing vehicle (DMSO; final concentration, 0.5%), MEHP (20 μM), 15d-PGJ₂ (1–15 μM), or troglitazone (3 μM). After treatment for 24 h, cells were lysed by incubation at 4°C in 200 μl of passive lysis buffer (Promega) for 20 min. Firefly and Renilla luciferase activities were measured in cell lysates using a dual reporter assay system (Promega) and a Monolight 2010 luminometer (Analytical Luminescence Laboratory, San Diego, CA). Luciferase activity values were normalized for transfection efficiency using Renilla luciferase activity assayed in the same cell lysates (firefly luciferase/Renilla luciferase).

EMSA

BU-11 cells were cultured in T75 flasks in medium containing 5% FBS and treated with vehicle (ethanol or DMSO; final concentration, 0.1%), MEHP (100–200 μM), or 15d-PGJ₂ (2.5–5 μM) for 2–24 h. Cells were collected and washed in PBS. Nuclear proteins were extracted as described previously (31).

For determination of PPARγ-DNA binding, a double-stranded oligonucleotide containing the PPRE from the acyl-CoA oxidase gene (5′-GTCGACAGGGGACCAGGACAAAGGTCACGTTCGGGAGTCGAC-3′;the consensus sequences are underlined) was used (44). PPARγ binds to this site with significantly greater affinity than PPARα (55). The DNA probe was end-labeled using T4 polynucleotide kinase (Promega) and [γ-³²P]ATP and was purified using a Centrispin-20 column (Princeton Separations, Adelphia, NJ). EMSAs were performed as follows. The ³²P-labeled DNA (∼0.5 ng, 100,000 cpm) and 4 μg of nuclear protein were combined with buffer (final concentrations: 20 mM HEPES (pH 8), 0.2 mM EDTA, 60 mM sodium chloride, 0.1% Nonidet P-40, 1 mM DTT, and 10% glycerol) and poly(dI-dC) (1 μg) in a final volume of 20 μl. The mixture was incubated at room temperature for 30 min. Polyacrylamide gels (16 cm, 5%) were prerun at 200 V for 30 min. Mixtures were electrophoresed at 200 V for 1.5 h in 0.5× TBE (final concentrations: 44 mM Tris-base (pH 8), 44 mM boric acid, and 0.8 mM EDTA). The gels were dried and exposed to film. The specificity of the shifted bands was determined by including an Ab specific for PPARγ (Calbiochem, La Jolla, CA; 516555) or by including 100× cold oligonucleotides containing the PPRE from the acyl-CoA oxidase gene (see above) or an unrelated site (5′-GAGCCGCAAGTGACTCAGCGCGGATCAATTA-3′).

For determination of NF-κB activation, a double-stranded oligonucleotide containing the NF-κB binding site from the upstream regulatory element of c-myc (5′-GATCCAAGTCCGGGTTTTCCCCAACC-3′; the consensus sequence is underlined) was used (56). The DNA probe was end-labeled and purified as described above. EMSAs were performed as follows. The ³²P-labeled DNA (∼0.5 ng, 50,000 cpm) and 2 μg of nuclear protein were combined with buffer (final concentrations: 10 mM Tris-HCl (pH 7.5), 1 mM EDTA, 100 mM sodium chloride, 0.5 mM magnesium chloride, 1 mM DTT, 20 μg BSA, 10% glycerol, and 0.5% Triton X-100) and poly dI-dC (1 μg) in a final volume of 20 μl. The mixture was incubated at room temperature for 30 min. The gel was run as described above, dried, and exposed to film. For quantification, the gels were analyzed on a PhosphorImager (Molecular Dynamics). The identity of the NF-κB subunits was determined by including Abs specific for p50 (sc-114), p52 (sc-848), Rel A (sc-372), or c-Rel (sc-71).

Statistics

Statistical analyses were performed with StatView (SAS Institute, Cary, NC). Data from a minimum of three experiments are presented as the mean ± SE. Student’s t test and one-factor ANOVAs were used to analyze the data. For the ANOVAs, Dunnett’s or Fisher’s protected least significant difference (PLSD) multiple comparisons tests was used to determine significant differences.

MEHP reduces [³H]thymidine incorporation and induces apoptosis in primary bone marrow B lymphocytes

Previous studies demonstrated that B cells in early stages of development undergo apoptosis when exposed to synthetic PPARγ agonist drugs (30). Given this observation, the presence of a natural PPARγ agonist in the bone marrow (i.e., 15d-PGJ₂), the ability of environmental phthalates to activate PPARγ (13, 14, 15, 16), and the relatively high human exposure to environmental phthalate esters, we investigated the potential for environmental phthalate esters to affect developing B lymphocytes. Models of early bone marrow B cell development were chosen because of the increased sensitivity of developing organs in general to environmental chemicals (49, 57, 58).

Primary bone marrow B cells were exposed to increasing concentrations of MEHP. [³H]thymidine incorporation, a measure of cell proliferation and/or survival, was assessed 48 h later. In primary bone marrow B220⁺/CD43⁺ B cells, MEHP caused a significant reduction in [³H]thymidine incorporation at concentrations as low as 10 μM, whereas higher concentrations caused correspondingly larger decreases in [³H]thymidine incorporation (Fig. 1⇓A). To evaluate the potential for MEHP to induce apoptosis in primary bone marrow B cells, cells were treated at a range of MEHP concentrations (10–150 μM) for 16 h, stained with PI, and analyzed for the percentage of cells with a subG₀/G₁ DNA content. Apoptosis was readily detected by flow cytometry after exposure to MEHP (Fig. 1⇓B). Treatment with doses as low as 25 μM MEHP consistently induced apoptosis in a significant fraction of these B220⁺/CD43⁺ B cells (Fig. 1⇓C). The percentage of apoptotic cells increased with increasing concentrations of MEHP, reaching ∼35% above background at 150 μM. These data are consistent with those obtained with PPARγ agonist drugs and suggest that environmental phthalates are similarly immunotoxic.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

The phthalate metabolite MEHP suppresses [³H]thymidine incorporation (A) and induces apoptosis (B and C) in primary bone marrow B cells. Primary bone marrow B cell populations (>95% B220⁺/CD43⁺) were prepared as described in Materials and Methods. Suspension cultures of primary bone marrow B cells were treated with ethanol (vehicle, <0.5%) or MEHP (10–150 μM) for 48 h ([³H]thymidine; A) or 16 h (PI staining; B and C). [³H]thymidine incorporation and apoptosis (formation of a subG₀/G₁ population after PI staining) were analyzed as described in Materials and Methods. The specific apoptosis was calculated by subtracting the background subG₀/G₁ population from each measurement from cells from that individual mouse. Data are presented as the mean ± SE from cultures derived from five to nine individual mice. ∗, Statistically different from vehicle-treated (p < 0.01, by ANOVA and Dunnett’s test).

MEHP reduces [³H]thymidine incorporation and induces apoptosis and cell cycle arrest in a pro/pre-B cell line

A nontransformed, bone marrow stromal cell-dependent pro/pre-cell line, BU-11, was used to confirm and extend studies on the effects of MEHP on developing B lymphocytes. First, BU-11 cells were exposed to increasing concentrations of MEHP, and [³H]thymidine incorporation was assessed 48 h later to compare the responses of this cell line to those seen with primary B cells. As with primary bone marrow B cells, MEHP caused a significant reduction in [³H]thymidine incorporation in BU-11 pro/pre-B cells (Fig. 2⇓, open histograms), although BU-11 cells were less sensitive to the MEHP-dependent decrease in [³H]thymidine incorporation than primary bone marrow B cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

MEHP suppresses [³H]thymidine incorporation in a nontransformed pro/pre-B cell line (BU-11). Suspension cultures of BU-11 cells were treated with ethanol (vehicle, 0.5%), MEHP (50–200 μM), and/or 9-cis-RA (10 μM) for 48 h. [³H]thymidine incorporation was determined as described in Materials and Methods. Data are presented as the mean ± SE from four experiments. ∗, Statistically different from vehicle-treated (p < 0.01, by ANOVA and Fisher’s PLSD test). +, Statistically different from 9-cis-RA- and MEHP-treated at the same concentration (p < 0.01, by ANOVA and Fisher’s PLSD test).

If inhibition of [³H]thymidine incorporation were due to activation of PPARγ, it would be predicted that simultaneous activation of its dimerization partner, RXRα, would enhance the biological effect of MEHP. For example, cotreatment of cells with PPARγ and RXRα agonists results in enhanced PPRE-dependent transcription (18, 19, 20, 21). Therefore, we tested the effect of MEHP and 9-cis-RA cotreatment on [³H]thymidine incorporation by BU-11 cells. [³H]thymidine incorporation was reduced significantly in BU-11 cells by treatment with 9-cis-RA (10 μM) alone (Fig. 2⇑, first solid histogram). Importantly, cotreatment with 10 μM 9-cis-RA and MEHP potentiated the reduction of [³H]thymidine incorporation by BU-11 cells (Fig. 2⇑, solid histograms) at all MEHP concentrations.

A reduction in [³H]thymidine incorporation can be attributable to a diminution of DNA synthesis (i.e., proliferation) and/or an increase in cell death. To distinguish between these two possibilities, BU-11 cells were stained with PI and analyzed by flow cytometry for the percentage of cells expressing a subG₀/G₁ fluorescence peak (i.e., apoptotic cells) and for the percentage of cells expressing S/G₂/M levels of DNA (i.e., dividing cells).

To evaluate the potential for MEHP to induce apoptosis, BU-11 cells were treated with 50–200 μM MEHP for 24 h. BU-11 cells exhibited a relatively low level of background apoptosis (<7%; Fig. 3⇓, A and B). Treatment with 150 or 200 μM MEHP consistently induced apoptosis in a significant fraction of these pro/pre-B cells (∼20 and ∼40%, respectively). As expected, DNA fragmentation, a hallmark of apoptosis, was readily demonstrated in BU-11 cells after a 24-h exposure to 150–200 μM MEHP (Fig. 3⇓C, lanes 5 and 7). Apoptosis also was confirmed by detection of poly(ADP-ribose) polymerase cleavage, an activity of effecter caspase 3, after exposure to 150–200 μM MEHP (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

MEHP induces apoptosis in pro/pre-B cells. Suspension cultures of BU-11 cells were treated with ethanol (vehicle, 0.1%), MEHP (50–200 μM), and/or 9-cis-RA (10–20 μM) for 24 h. Apoptosis assays were performed as described in Materials and Methods. A, Formation of a subG₀/G₁ population after PI staining. Data are representative of four experiments. B, Quantification of the subG₀/G₁ population. Data are presented as the mean ± SE from four experiments. ∗, Statistically different from vehicle- and 9-cis-RA-treated (p < 0.01, by ANOVA and Fisher’s PLSD test). +, Statistically different from vehicle-, 9-cis-RA-, and MEHP-treated at the same dose (p < 0.01, by ANOVA and Fisher’s PLSD test). C, Formation of DNA ladders. Data are representative of three experiments.

To ascertain the contribution of RXRα to MEHP-induced apoptosis, BU-11 cells were treated with MEHP (50–200 μM) with or without 10 or 20 μM 9-cis-RA for 24 h. Addition of 10 or 20 μM 9-cis-RA to BU-11 cell cultures did not affect BU-11 cell viability (Fig. 3⇑, A and B). In contrast, cotreatment with 20 μM 9-cis-RA and a suboptimal dose of MEHP, e.g., 100 μM, induced a significant level of apoptosis (Fig. 3⇑B). Similarly, the 10 μM 9-cis-RA treatment significantly enhanced apoptosis induced by 150 μM MEHP (Fig. 3⇑, A and B). Cotreatment with 20 μM 9-cis-RA and 150 μM MEHP increased the incidence of apoptosis even more (Fig. 3⇑, A and B). The enhancement of apoptosis by 9-cis-RA in combination with MEHP also is reflected in DNA ladder formation (Fig. 3⇑C, lanes 5 and 6). These data are consistent with a role for the PPARγ-RXRα heterodimer in MEHP-mediated pro/pre-B cell apoptosis signaling.

It was noted that BU-11 cell apoptosis could not be detected after exposure to 100 μM MEHP (Fig. 3⇑, B and C), a concentration at which inhibition of [³H]thymidine incorporation could be demonstrated (Fig. 2⇑). Similarly, in primary bone marrow B cells, [³H]thymidine incorporation was reduced significantly at concentrations below those at which apoptosis was evident (e.g., 10 μM; Fig. 1⇑, A and C). These results suggest that MEHP induces growth arrest rather than apoptosis at low concentrations. To evaluate the potential for MEHP to induce growth arrest and thereby to reduce [³H]thymidine incorporation, BU-11 cells were exposed to low concentrations of MEHP (50–100 μM) with and without 10 μM 9-cis-RA for 48 h.

A significant decrease in the number of cells in S/G₂/M was observed after treatment with 100 μM, but not 50 μM, MEHP (Fig. 4⇓A, open histograms). Therefore, the reduction in [³H]thymidine incorporation at 100 μM MEHP is probably due to growth arrest before or during the DNA synthesis phase. This decrease in the percentage of cycling cells was even more profound upon inclusion of 9-cis-RA (Fig. 4⇓A, solid histograms). Although 10 μM 9-cis-RA or 100 μM MEHP alone reduced the percentage of cycling cells by 7–8%, the effects of treating cells with 10 μM 9-cis-RA together with 100 μM MEHP appeared more than additive (i.e., a 22% reduction in the percentage of cycling cells). Therefore, a PPARγ-RXRα complex appears to participate in growth arrest as it does in apoptosis induction.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Low doses of MEHP induce cell cycle arrest in pro/pre-B cells. Suspension cultures of BU-11 cells were treated with ethanol (vehicle, 0.1%), MEHP (50–100 μM), and/or 9-cis-RA (10 μM) for 48 h. Assays were performed as described in Materials and Methods. A, Cell cycle distribution after PI staining. Data are presented as the mean ± SE from three experiments. ∗, Statistically different from vehicle-treated (p < 0.05, by ANOVA and Dunnett’s test). B, p27^Kip1 expression. Data are presented as the mean ± SE from three experiments. ∗, Statistically different from vehicle-treated (p < 0.05, by ANOVA and Fisher’s PLSD test). +, Significantly different from vehicle- and 9-cis-RA-treated (p < 0.05, by ANOVA and Fisher’s PLSD test). There were no significant differences in β-actin expression.

The p27^Kip1 protein is a cyclin-dependent kinase inhibitor that blocks cyclin E/cyclin-dependent kinase 2, a cyclin that controls passage through G₁ into S phase (59). To begin to assess the mechanisms by which MEHP and 9-cis-RA induce cell cycle arrest, BU-11 cells were exposed to low concentrations of MEHP (50–100 μM) with and without 10 μM 9-cis-RA for 48 h. Whole-cell lysates were examined for p27^Kip1 expression by Western blotting. Treatment with either 100 μM MEHP or 9-cis-RA alone resulted in a modest, but statistically significant, increase in p27^Kip1 (Fig. 4⇑B). As with the other readouts, the effects of 50 μM MEHP added together with 9-cis-RA were more profound than those with either agent alone (4.5-fold; p < 0.05). The results are consistent with the hypothesis that MEHP- and/or 9-cis-RA-mediated growth arrest involves the up-regulation of the cell cycle inhibitor p27^Kip1.

15d-PGJ₂, a natural PPARγ ligand in the bone marrow, reduces [³H]thymidine incorporation by inducing apoptosis and cell cycle arrest in pro/pre-B lymphocytes

15d-PGJ₂ is a PPARγ agonist (44) whose precursor PGD₂ is found in high concentrations in the bone marrow microenvironment (43, 45), suggesting its role in lymphocyte development. As such, it may potentiate the toxic effects of environmental PPARγ agonists. To test this possibility, the ability of 15d-PGJ₂ to affect pro/pre-B cell growth and survival, either alone or in combination with MEHP, was evaluated.

BU-11 cells were exposed to increasing concentrations of 15d-PGJ₂ and 9-cis-RA for 48 h, and [³H]thymidine incorporation was assessed. 15d-PGJ₂ caused a >90% reduction in [³H]thymidine incorporation at concentrations as low as 5 μM (Fig. 5⇓A). [³H]thymidine incorporation was inhibited completely at ≥10 μM 15d-PGJ₂. Similarly, 5 μM 15d-PGJ₂ induced a significant increase in a subG₀/G₁ fluorescence peak after an 8-h exposure (vehicle, 10 ± 1%; 15d-PGJ₂, 62 ± 1%; p < 0.01) and significant DNA fragmentation after a 24-h exposure (Fig. 5⇓B, lane 5). It is important to note that 2 μM 15d-PGJ₂ did not significantly affect either [³H]thymidine incorporation or DNA fragmentation. These results indicate that 15d-PGJ₂ is a potent inhibitor of pro/pre-B cell [³H]thymidine uptake and that this decrease in [³H]thymidine uptake is due, at least at concentrations ≥5 μM, to apoptosis induction.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

15d-PGJ₂, a natural bone marrow-derived PPARγ ligand, inhibits [³H]thymidine incorporation by inducing both apoptosis and cell cycle arrest in pro/pre-B cells. Suspension cultures of BU-11 cells were treated with ethanol (vehicle, <0.5%), 15d-PGJ₂ (0.5–50 μM), and/or 9-cis-RA (10 μM) for 24 h (DNA ladder formation) or 48 h ([³H]thymidine incorporation and p27^kip1 expression). Assays were performed as described in Materials and Methods. A) [³H]thymidine incorporation after treatment with 15d-PGJ₂. Data are presented as the mean ± SE from four experiments. ∗, Statistically different from vehicle-treated (p < 0.01, by ANOVA and Fisher’s PLSD test). B, DNA ladder formation. Data are representative of three experiments. C, [³H]thymidine incorporation after treatment with 15d-PGJ₂ and 9-cis-RA. Data are presented as the mean ± SE from four experiments. +, Statistically different from vehicle-, 9-cis-RA-, and 15d-PGJ₂-treated at the same dose (p < 0.01, by ANOVA and Fisher’s PLSD test). D, p27^Kip1 expression. Data are representative of three experiments. Fold change from untreated cultures: vehicle, 1.1 ± 0.4; 15d-PGJ₂, 3.2 ± 1.0; 9-cis-RA, 9.0 ± 1.4; 15d-PGJ₂ and 9-cis-RA, 18.4 ± 4.0* (∗, statistically different from all other treatments (p < 0.05, by ANOVA and Fisher’s PLSD test)). There were no significant differences in β-actin expression.

Neither 10 μM 9-cis-RA nor 0.5 μM 15d-PGJ₂ had an effect on [³H]thymidine incorporation when either agent was added alone (Fig. 5⇑C). However, a combination of 10 μM 9-cis-RA and 0.5 μM 15d-PGJ₂ significantly reduced [³H]thymidine incorporation (Fig. 5⇑C, solid histograms). Indeed, the combinatorial effect of these respective PPARγ and RXRα agonists was most evident when adding 2 μM 15d-PGJ₂ together with 10 μM 9-cis-RA. However, no increase in DNA ladder formation was seen when combining 2 μM 15d-PGJ₂ and 10 μM 9-cis-RA (Fig. 5⇑B), suggesting that the significant decrease in [³H]thymidine incorporation was probably due to cell cycle arrest. Similarly, although 1–2 μM 15d-PGJ₂ or 10 μM 9-cis-RA alone increased p27^Kip1 expression, the combination of these respective PPARγ and RXRα agonists resulted in a significant effect with regard to p27^Kip1 expression (an 18.4 ± 4.0-fold increase; p < 0.05; Fig. 5⇑D). Collectively, these data indicate that this natural PPARγ ligand induces p27^Kip1 expression concomitant with growth arrest at concentrations lower than those capable of inducing apoptosis (i.e., ≤2 μM) and that this effect is enhanced significantly by coactivation of a PPARγ dimerization partner, i.e., RXRa.

To determine whether 15d-PGJ₂ potentiates the activity of an environmental PPARγ agonist, BU-11 cells were treated with 100 μM MEHP, i.e., a suboptimal concentration that induces a low level of growth arrest (Fig. 4⇑), but no apoptosis (Fig. 3⇑), together with suboptimal concentrations of 15d-PGJ₂ (i.e., 1–2 μM; see Fig. 5⇑). Interestingly, this combination resulted in a significantly greater reduction in [³H]thymidine incorporation than seen with either PPARγ agonist alone (Fig. 6⇓A). The enhanced loss of [³H]thymidine incorporation with 100 μM MEHP and 1–2 μM 15d-PGJ₂ resulted from the induction of apoptosis, as evidenced by a significant increase in a subG₀/G₁ fluorescence peak (Fig. 6⇓B) and an increase in DNA ladder formation (Fig. 6⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Cotreatment of pro/pre-B cells with 15d-PGJ₂ significantly enhances the MEHP-induced suppression of [³H]thymidine incorporation. Suspension cultures of BU-11 were treated with ethanol (vehicle, <0.5%), MEHP (100 μM), and/or 15d-PGJ₂ (1–2 μM) for 24 h (PI staining and DNA ladder formation) or 48 h ([³H]thymine incorporation and p27^kip1 expression). Assays were performed as described in Materials and Methods. A, [³H]thymidine incorporation. B, Formation of a subG₀/G₁ population after PI staining. Data are presented as the mean ± SE from three experiments (A and B). ∗, Statistically different from vehicle-treated (p < 0.01, by ANOVA and Fisher’s PLSD test). ∗∗, Statistically different from vehicle- and MEHP-treated at the same dose (p < 0.01, by ANOVA and Fisher’s PLSD test). C, DNA ladder formation. Data are representative of three experiments. D, p27^Kip1 expression. Data are representative of three experiments. Fold change from naive: vehicle, 1.4 ± 0.1; 1 μM 15d-PGJ₂, 3.7 ± 0.8; 75 μM MEHP, 3.2 ± 0.5; 15d-PGJ₂ and MEHP, 8.2 ± 1.8* (∗, statistically different from all other treatments (p < 0.01, by ANOVA and Fisher’s PLSD test)). There were no significant differences in β-actin expression.

Because apoptosis was induced after cotreatment with 100 μM MEHP and 15d-PGJ₂ (Fig. 6⇑B), a lower MEHP dose, which does not induce apoptosis even in combination with 15d-PGJ₂ (i.e., 75 μM), was used to assess p27^kip1 up-regulation, a marker for growth inhibition, in the presence of MEHP and 15d-PGJ₂. Suboptimal doses of either MEHP (75 μM) or 15d-PGJ₂ (0.5–1.0 μM) did not affect p27^Kip1 expression (Fig. 6⇑D). However, a combination of these PPARγ ligands at these doses significantly increased p27^kip1 expression (Fig. 6⇑D), a result consistent with a synergistic effect on growth inhibition. Considering that concentrations of 15d-PGJ₂ may be as high as the micromolar range in the bone marrow (43, 44, 45), these data suggest that even low exposures to environmental phthalates such as MEHP may have profound effects on bone marrow B cell growth and/or survival.

Analyses of treatment with MEHP and 15d-PGJ₂ or 9-cis-RA; additive vs synergistic effects

Although previous investigators have referred to the interaction of combinations of PPARγ/RXRα ligands as synergistic (18, 19, 20, 21), the methodology used was not sufficient to distinguish true synergy from nonsynergistic, additive effects. The data presented in this study suggested a combinatorial effect between MEHP and 9-cis-RA or 15d-PGJ₂. Therefore, we used the isobole method to determine whether the enhancements in biological responses seen with treatment combinations were additive or synergistic. Contour lines for the response surface of each combination were plotted. These contours represent levels of constant effect in the combination dose space, i.e., isoboles of effect. Isoboles of an asynergistic (i.e., concentration additive) interaction are represented by straight lines with negative slopes. Interactive combinations are manifest as isoboles which curve toward (synergism) or away from (antagonism) the origin (53, 54).

The isoboles of the MEHP plus 15d-PGJ₂ combination are nearly straight lines, indicating an effect that is at least additive effect (Fig. 7⇓A). In contrast, the isoboles of the MEHP plus 9-cis-RA combination are more complex and across most of the concentration range curve toward the origin, indicating a strongly synergistic interaction in this region (Fig. 7⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Isobole analysis of MEHP interactions with 15d-PGJ₂ (A) and 9-cis-RA (B). Suspension cultures of BU-11 cells were treated with ethanol (vehicle, 0.5%), MEHP (25–100 μM), 15d-PGJ₂ (0.5–2.5 μM), and/or 9-cis-RA (1–10 μM) for 48 h. [³H]thymidine incorporation was determined, and contour plots (isobolograms) of the combinatorial effects of MEHP with 15d-PGJ₂ or 9-cis-RA were constructed as described in Materials and Methods. Levels shown are fractional responses of the Vh-Vh cell. Circles indicate locations of data points from three experiments. Contours are linear interpolation between points.

MEHP and 15d-PGJ₂ induce PPARγ-DNA binding and transcriptional trans-activation

MEHP has been shown in mouse embryonic carcinoma cells and in PPARγ-transfected COS cells to be a PPARγ agonist (15, 16). Previously we have shown that PPARγ, along with its dimerization partner, RXRα, are expressed at the pro/pre-B cell stage (31). Furthermore, the synergy between MEHP and 9-cis-RA seen in the present system is consistent with the activation of a PPARγ-RXR dimer. PPARγ-specific EMSAs were performed to confirm MEHP- and 15d-PGJ₂-induced PPARγ-DNA binding in pro/pre-B cells.

Treatment of BU-11 cells with MEHP or 15d-PGJ₂ increased binding to the PPRE sequence from the acyl-CoA oxidase gene (Fig. 8⇓A). The presence of PPARγ in the DNA-binding complex was confirmed by a change in the banding pattern similar to that published previously (31) upon inclusion of a PPARγ-specific Ab (Fig. 8⇓A). In addition, the band could be specifically competed by a 100-fold excess of unlabeled oligonucleotides containing PPREs, but not by DNA of an unrelated sequence. Interestingly, treatment of cells with a combination of suboptimal concentrations of MEHP and 15d-PGJ₂ resulted in PPRE-DNA binding greater than that of either compound alone (Fig. 8⇓A). These results indicate that both MEHP and 15d-PGJ₂ induce PPARγ-DNA binding in these early B lymphocytes.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

MEHP and 15d-PGJ₂ induce PPARγ-DNA binding and transcriptional transactivation. A, EMSA analysis of acyl CoA PPRE oligonucleotide binding. Nuclear extracts were prepared from BU-11 cells treated with ethanol (vehicle, 0.1%), MEHP (200 μM), or 15d-PGJ₂ (5 μM) for 24 h (first four panels) or with MEHP (100 μM) with or without 15d-PGJ₂ (2.5 μM) for 24 h (last panel on the right). EMSAs were performed as described in Materials and Methods. Supershift analyses were performed by including 2 μl of nonspecific rabbit IgG or polyclonal PPARγ-specific Ab. Competition analyses were performed by including a 100-fold excess of nonspecific or specific DNA. Similar results were found using nuclear extracts from 15d-PGJ₂-treated cells. FP, free probe. Data are representative of two or three experiments. B and C, Reporter analysis of PPARγ transactivation. COS-1 cells were transfected with reporters driven by the PPRE from the rat enoyl CoA hydratase/3-hydroxylacyl CoA promoter (B) or from the rabbit CYP4A6 promoter (C) as described in Materials and Methods. Luciferase activity was determined 24 h after treatment with DMSO (vehicle, 0.05%), MEHP (20 μM), 15d-PGJ₂ (1–15 μM), or troglitazone (3 μM) under serum-free conditions. Data represent PPARγ-driven firefly luciferase reporter activity normalized to CMV-driven control Renilla luciferase reporter activity and are presented as the mean ± SE from three experiments. ∗, Statistically different from vehicle-treated (p < 0.01, by ANOVA and Fisher’s PLSD test).

To compare MEHP-induced trans-activation with that induced by 15d-PGJ₂, COS-1 cells were transiently transfected with mouse PPARγ and with the luciferase reporter plasmid pHDx3-luc, containing three copies of a PPRE derived from the rat enoyl CoA hydratase/3-hydroxyacyl CoA promoter (Fig. 8⇑B), or with the luciferase reporter plasmid pLuc-4A6–880, which included 880 nt of the 5′-flanking DNA of the PPARγ-responsive rabbit CYP4A6 gene (Fig. 8⇑C). (Transient transfection of BU-11 cells with reporter constructs was not practical because of the high levels of death and the low transfection efficiency seen in these nontransformed cells after either electroporation or FuGene treatment.)

Treatment of transfected COS-1 cells under serum-free conditions with the potent PPARγ agonist, troglitazone (3 μM), as a positive control resulted in a 15-fold activation of the CoA reporter and a 6-fold activation of the CYP4A6 reporter (Fig. 8⇑, B and C). MEHP (20 μM) stimulated a 7-fold activation of the CoA reporter construct and a 3-fold activation of the CYP4A6 reporter (Fig. 8⇑B). Treatment with 1–5 μM 15d-PGJ₂ resulted in a concentration-dependent increase in activation of the CoA reporter construct, with a 15-fold increase in activation at 5 μm (Fig. 8⇑B). In contrast, 15d-PGJ₂ exhibited weaker activation of the CYP4A6 reporter, requiring 15 μM 15d-PGJ₂ for only a 4-fold increase in activity (Fig. 8⇑C). These results show that although MEHP has a lower potency and efficacy in trans-activating the hydroxylacyl CoA gene than 15-dPGJ₂, MEHP and 15d-PGJ₂ have similar potencies and efficacies in trans-activating the CYP4A6 gene.

Apoptosis induced by MEHP and 15d-PGJ₂ does not result from inhibition of NF-κB activation

Previous studies suggested that PPARγ activation can either enhance (60, 61, 62) or down-regulate (36, 37, 63, 64, 65, 66, 67) NF-κB activity, and that NF-κB down-regulation results in apoptosis (37). To determine the effect of endogenous and environmental PPARγ agonists on NF-κB activation, BU-11 cells were treated with 5 μM 15d-PGJ₂ for 2 h or with 200 μM MEHP for 8 h. Nuclear proteins were extracted and assayed by EMSA for binding to the upstream κB regulatory element of c-myc. Unlike some studies with 15d-PGJ₂ (36, 63), but consistent with our previous studies with PPARγ agonists (31), treatment with 15d-PGJ₂ increased NF-κB-DNA binding 1.7-fold (p < 0.01) at 2 h post-treatment (Fig. 9⇓, A and B). The failure to detect a change in binding of the constitutively active Oct-1 transcription factor with its cognate DNA binding element confirmed the specificity of the NF-κB response (data not shown). Similarly, treatment with MEHP for 8 h resulted in a significant (p < 0.01) 1.5-fold increase in NF-κB-DNA binding.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

MEHP and 15d-PGJ₂ activate NF-κB-DNA binding in pro/pre-B cells. Nuclear extracts were prepared from BU-11 cells treated with ethanol (vehicle, 0.1%) or 15d-PGJ₂ (5 μM) for 2 h or from BU-11 cells treated with DMSO (vehicle, 0.01%) or MEHP (200 μM) for 8 h. Nuclear extracts were analyzed by EMSA for NF-κB oligonucleotide binding as described in Materials and Methods. A, NF-κB activation after treatment with 15d-PGJ₂ or MEHP. Data are representative of three experiments. B, Quantification of NF-κB-DNA binding. Data are presented as the mean ± SE from three experiments. ∗, Statistically different from vehicle-treated (p < 0.01, by Student’s t test). C, Supershift analyses of NF-κB-DNA binding were performed by including 2 μl of polyclonal Ab specific for p50, p65/RelA, or c-Rel. The single arrow indicates the faster migrating p50:p50 complexes, and the double arrows above the dashed line indicate the slower migrating p50:p65 and p50:cRel complexes. Similar results were found using nuclear extracts from 15d-PGJ₂-treated cells (data not shown).

Supershift assays then were used to identify the subunits in the active NF-κB complex. An Ab to p50 supershifted both the faster and slower migrating κB complexes (indicated by single and double arrows, respectively; Fig. 9⇑C), whereas p65/RelA- and c-Rel-specific Abs significantly decreased the slower migrating κB band (double arrows above the dashed line). These results are similar to those obtained with the synthetic PPARγ ligands, ciglitazone and GW7845 (31). In general, these results indicate that, unlike apoptosis induced with other environmental chemicals (52), apoptosis induced in developing B cells by natural or environmental PPARγ ligands is not mediated by NF-κB down-regulation.