Hydroquinone stimulates granulocyte-macrophage progenitor cells in vitro and in vivo.

To investigate whether hydroxylated metabolites of benzene may be responsible for the amplification of granulocyte-macrophage progenitor cells (GM-CFC) observed in mice that inhale benzene, groups of six C57BL6 mice were injected with hydroquinone (HQ) (75 mg/kg) or HQ (50 mg/kg) plus phenol (PHE) (50 mg/kg) twice daily for 11 days. Deviations in blood leukocyte and erythrocyte levels by up to one-third were noted in the treated groups; however, the peripheral blood differential counts were unchanged. Although no changes in bone marrow cellularity were observed in mice treated with HQ, cellularity was decreased by a factor of two in the mice that had received HQ plus PHE. The number of GM-CFC per femur was doubled in both treated groups. In vitro experiments using the murine multipotent hematopoietic progenitor cells FDCP mix also showed a duplication of GM-CFC formation in the presence of HQ at concentrations between 10(-6) M and 10(-10) M. When HQ and PHE were present at equimolar concentrations, significantly increased colony formation was still observed with 10(-12) M of metabolites. The effect was independent of the concentration of GM-colony-stimulating factor used. We suggest that HQ is a major mediator of the stimulatory effect of benzene on GM-CFC in mice. In addition, the in vitro data indicate that a direct effect of GM-CFC is involved.


Introduction
Benzene has been well recognized as a hazardous industrial chemical for many years (1). Exposure to benzene has been associated with increased incidence of blood dyscrasias, including cytopenias, aplastic anemias, myelodysplastic syndromes, and leukemias (2)(3)(4). Damage to the hematopoietic system by benzene has been predominantly linked to the formation and accumulation in bone marrow of hydroxylated derivatives such as hydroquinone (HQ) and catechol (5,6). This was shown to be due to the conversion of phenol (PHE), the major in vivo metabolite of benzene, by myeloid cell-derived myeloperoxidase in the bone marrow (7). Oxidation of HQ to p-benzoquinone (BQ) via a semiquinone intermediate is considered a key step towards the generation of toxic intermediates in bone marrow (8); however, open-ring products such as trans-transmucondialdehyde may also be responsible for cytotoxic damage following benzene exposure (9,10).
Types of permanent DNA damage caused by benzene or its metabolites include structural chromosome aberrations, aneuploidy, sister chromatid exchange, micronuclei (11), and gene mutations (12). Hydroquinone and BQ have been shown to result in the formation of DNA adducts, whose structures have been characterized in part (13). Other studies have looked at short-term effects of hydroxylated benzene metabolites on different cellular compartments of bone marrow, i.e., different types of stroma cells, hematopoietic stem cells, progenitor cells, and maturing cells of different blood cell lineages. Mainly, marrow-adherent cells, including mononuclear phagocytes, have been defined as targets of hematopoietic damage by benzene (14)(15)(16)(17).
Benzene application augments precursor cells of granulocyte-macrophage (GM) lineage in bone marrow of mice during and after chronic inhalation (18). Work by Irons and colleagues (19) demonstrated a synergistic effect of hydroquinone (HQ) on granulocyte-macrophage progenitor cells (GM-CFC) in vitro. Analysis of cytofluorometrically sorted bone marrow cells and plating at low cell numbers demonstrated that this is a direct-i.e., not stroma cellmediated-effect on GM-CFC.
We investigated whether HQ administered in vivo could reproduce the increase in GM-CFC seen in mice that inhaled benzene chronically (18). We used the protocol that had been reported by Eastmond et al. (20) to reproduce, after ip injection of HQ and PHE, the decrease in bone marrow cellularity associated with chronic inhalation of benzene. To determine if this is a direct effect on the progenitor cells, we conducted analogous in vitro experiments. were allowed standard rodent food and water ad libitum throughout the studies. Mice were treated and kept in groups of six. Injections (50-75 mg/kg/day in 0.2 ml solvent, ip) were administered during the day twice daily, with at least a 6-hr interval between injections, following the schedule of Eastmond et al. (20). On day 12 of the trials, blood smears were obtained from the tail vein, animals were killed by cervical dislocation, and bone marrow cells from femora were prepared according to standard methods (21).

Cytology
Blood and bone marrow smears were performed on glass slides, stained with May-Grunwald-Giemsa (Merck, Darmstadt, Germany) and evaluated microscopically. Cell counts were performed in 0.3% acetic acid for white blood cells or in 1% methylene blue in PBS for red cells.

FDCP-mix Cells
Factor-dependent cells Paterson (FDCP)mix cells (22)   throughout the treatment period (25 g). No abnormal behavior was registered, except for some minutes of behavioral irritation after the first dose in animals that received combined HQ and PHE. Blood leukocyte and erythrocyte counts were relatively unchanged between the treated and control groups, except for an anemia in the HQ + PHE group, and a slight leukocytosis in the HQ-treated group, which was not statistically significant (Table 1). Bone marrow cellularity was reduced upon treatment with HQ + PHE, but not in HQ-treated animals ( Table 1). No changes in the blood differential cell count were observed (data not shown). When femur bone marrow cells were assayed for GM-CFC, all mice in the treated groups displayed elevated numbers of GM-CFC per femur in comparison with the levels of the control animals ( Figure 1). No GM-CFC could be grown from spleen cells of the animals. Effect of Hydroquinone on Formation of Granulocyte-Macrophage Colonies in Vtro When HQ was present at noncytotoxic concentrations, the number of GM colonies formed by FDCP-mix cells was enhanced about 2-fold, if either GM-CSF or IL-3 was used as the cytokine driving colony growth (Figure 2A,B). When PHE was present at equimolar concentrations in addition to HQ, a left shift of the dose-response curve of HQ by 2 to 3 logs was observed ( Figure  2C). Addition of PHE to HQ-containing cultures did not result in toxicity at concentrations as high as 106M. Variation of GM-colony stimulating factor (CSF) concentrations at constant levels of HQ did not affect the ability of HQ to elicit enhanced colony formation at both plateau and also non-saturating cytokine concentrations ( Figure 3). This suggests synergistic interaction of HQ and PHE with either GM-CSF or IL-3.

Discussion
We observed an increase in total GM-CFC in the bone marrow of mice treated with HQ or an equimolar mixture of HQ and PHE. Snyder et al (18) reported the presence of increased numbers of GM-CFC in animals that inhaled benzene chronically.
The results of the present study indicate that this effect may be mediated by the metabolite HQ. It cannot be determined, however, if HQ itself, or metabolic products of HQ, such as BQ or trans-transmucondialdehyde, are the moieties that ultimately act on the GM-CFC. A bias also remains with regard to the comparability of the concentrations of the substances reached in vivo between the benzene inhalation protocol and the HQ + PHE injection protocol; this is because direct measurements of bone marrow levels of PHE, HQ, and their metabolites have not been determined for the latter protocol. A relative comparability may be deduced from the fact that the doses of HQ and PHE injected are able to reproduce the bone marrow toxicity that is achieved by inhalation of benzene for the same interval (20).
In previous experiments (R Henschler and CM Heyworth, unpublished data), instead of an expected change of differentiation pattern upon exposure to HQ, we noted an increase in cell numbers in liquid cultures of murine progenitor cell FDCPmix that had been induced to differentiate into the GM lineage. When FCDP-mix cells were subsequently analyzed in colony assays, a profound increase in colony numbers was observed in the presence of HQ. Because colony assays of FDCP-mix cells used as few as 5000 cells/ml, the resulting increase in colony formation in these assays is most likely a direct effect of HQ and GM-CSF on the colony-forming cells and not an indirect effect via bystander cells. These studies demonstrate that FDCP-mix cells retain the capability of primary bone marrow cells to reproduce the synergistic effect of GM-CSF and HQ, or HQ and PHE, on myeloid colony formation, as reported by Irons et al. (19) for unfractionated bone marrow or bone marrow cells enriched for primitive populations. As FDCP-mix cells are multipotential cells, however, and as the primary populations used by Irons et al. (19) also contain substantial numbers of multipotential precursors in addition to GM-CFC, it is possible that HQ and GM-CSF act on a multipotential precursor of GM-CFCs rather than on GM-CFCs themselves. They may induce these "pre-GM-CFCs" to undergo extra cell divisions to yield increased numbers of GM-CFC. Another possibility is that HQ and GM-CSF counteract, for example, an apoptotic mechanism that normally results in cell death of a proportion of potential colony-forming cells. Additional work will be necessary to elucidate the mechanisms underlying the synergism between hematopoietic growth factors and benzene metabolites.