The use of ferrokinetics in the study of experimental anemia.

Erythropoietic cells in bone marrow are vulnerable to cytotoxic substances. There are three types of erythroid precursors: cells that can take up Fe but do not proliferate (reticulocytes), those that can take up Fe and proliferate (normoblasts and pronormoblasts), and those cells that do not take up Fe but can proliferate and differentiate into the erythroid cell line (ERC and stem cells). Each of these erythroid precursors requires a certain time before they emerge into the peripheral blood as mature red blood cells. By applying our understanding of ferrokinetics associated with erythropoiesis, it was possible to estimate a cytotoxic effect of chemicals on proliferating erythgroid precursors (pronormoblasts) in mice by measuring 24-hr 59Fe uptake in red blood cells 48 hr after treatment with chemicals. The effect of chemicals on pluripotent hemopoietic stem cells in mice was also estimated by measuring 24-hr 59Fe uptake 72 hr after treatment with chemicals. The validity of experimental schemes was tested using cytarabine, methotrexate, vinblastine, cyclophosphamide, and busulfan, which are known to act against specific cell types. Effects on pluripotent hemopoietic stem cells were tested with or without activation of stem cells in G0 into cell cycle. Applications of the 59Fe uptake method in the study of (1) benzene toxicity and (2) effect of pentobarbital on the toxic action of hydroxyurea and cytarabine are described. Proper application of the ferrokinetic characteristics of erythropoietic cells enables the establishment of a methodology which can be used to evaluate potential toxic effects of chemicals on erythroid precursor cells and pluripotent hemopoietic stem cells.


Introduction
Bone marrow, a dynamic system which produces blood cells, is vulnerable to cytotoxic agents. The erythropoietic system retains several characteristics that allow for a rather precise experimental evaluation of the system: the response of early erythroid precursor cells to erythropoietin, the capability of erythroid cells to synthesize hemoglobin, and the existance of a histochemical marker for reticulocytes. By using ferrokinetic features and other developmental characteristics of cells involved in erythropoiesis, it was possible to measure the relative size of a cellular pool of different erythroid precursor cells affected by chemical agents. We believe such a methodology can be used for an assessment of hazard caused by cytotoxic agents on erythropoietic cells including hemopoietic stem cells. This presentation will discuss: ferrokinetics involved in erythropoiesis, rationale in the application of ferrokinetics for evaluating chemically induced erythropoietic toxicity, an experimental scheme and test for the rationale, and application of the 59Fe uptake method for the assessment of potential erythropoietic toxins. 29

Ferrokinetics Involved in Erythropoiesis
The earliest erythropoietic cell that can be recognized morphologically is pronormoblast. This cell is derived from the erythropoietin response cell (ERC), and the ERC is derived from pluripotent hemopoietic stem cells. There is evidence that pluripotent hemopoietic stem cells are the progeny of more primitive hemopoietic stem cells (1)(2)(3)(4). Pronormoblasts differentiate and mature sequentially into normoblasts, reticulocytes, and erythrocytes (Fig. 1). All cells from the ERC to the normoblast are in cell cycle and thus proliferate (5,6) while reticulocytes and erythrocytes do not. In rodents, most pluripotent hemopoietic stem cells (90-95%) are not in cycle; i.e., they are in Go state (7,8), but they can be activated into cell-cycle upon specific physiological stimuli (9)(10)(11). Under normal conditions, only 5-10% of stem cells are in cell cycle, and these constitute the ERC pool. The cells from pronormoblasts to reticulocytes can synthesize hemoglobin (6). Therefore, they can take up 59Fe, but stem cells and ERC do not take up Fe because they do not make hemoglobin. Thus, from the ferrokinetic point of view, the cells involved in erythropoiesis can be classified into three categories: cells that take up Fe but do not proliferate (reticulocytes), cells that take up Fe and proliferate (normoblasts and pronormoblasts), and finally, cells that do not take up Fe but can proliferate and differentiate into the erythroid cell line, i.e., ERC and stem cells. The time required for each of the erythrocyte precursor cells to mature and to be released into the circulating blood has been determined in mice. The turnover time of reticulocytes and the generation times of both pronormoblasts and normoblasts are about 24 hr (12)(13)(14)(15). In these studies, it was decided that the size of the reticulocyte pool would be used as a measure of the effects of chemicals on each of the earlier cell types by administering 59Fe, which rapidly disappears from the circulation, and sampling the peripheral blood for 59Fe uptake 24 hr later, a time at which the label in the blood represents incorporation of the iron into hemoglobin during the period when maturation and release of the cells into the circulation occurs. Therefore, the cytotoxic effect of an agent on normoblasts in mice can be evaluated by giving 59Fe when the affected erythroid precursors mature to the stage of reticulocytes, i.e., 24 hr after treatment with the agent and by measuring the amount of 59Fe in the circulating red blood cells 24 hr after giving 59Fe, a time interval which allows the bone marrow reticulocytes to be released into the blood (Fig. 1). The time when 59Fe is injected as opposed to the time when the blood sample is taken is the time when the size of the reticulocyte compartment is measured in t   this scheme because 59Fe is taken up by reticulocytes at the time of 59Fe injection. By the same token, the effect of agents that damage the pronormoblasts can be reflected by the 24 hr '9Fe uptake value which is measured 48 hr after treatment with the agent.
It has also been shown that it takes 24 hr for the ERC to differentiate into pronormoblasts (12). Regarding stem cells, there is no information on the time required for the cells to differentiate into ERC. However, our experimental data, which will be discussed shortly, suggest that stem cells differentiate into ERC immediately upon physiological stimulus. Thus, we suggest that the effect of agents that inhibit pluripotent hemopoietic stem cells and ERC can be seen when 59Fe is given 72 hr after treatment with the agent and 24 hr 59Fe uptake is measured.

Experimental Scheme and Tests
In order to test the validity of the rationale, three types of cytotoxic agents, for which the mechanism of action is known, were chosen. Methotrexate and cytarabine are S-phase specific cytotoxic agents and can kill only proliferating cells (16)(17)(18)(19)(20)(21)(22), while cyclophosphamide and busulfan are cell-cycle stage nonspecific agents and can kill both proliferating and nonproliferating cells such as hemopoietic stem cells (23)(24)(25)(26). Vinblastine produces both metaphase arrest of mitosis and some damage to hemopoietic stem cells (23,27). The effect of these cytotoxic agents on each compartment of erythroid precursors were determined by the 24-hr 5 Fe uptake measured 24, 48, 72, and 96 hr after injection of a single dose of the agent.
Cytarabine and methotrexate lowered 5 Fe uptake 24 and 48 hr after treatment with the agent, respectively, as expected, but at 72 and 96 hr 59Fe uptake was increased over control values (Fig. 2). An interpretation of the overshoot may be that the ERC pool was immediately replaced by an excessive number of pluripotent stem cells as a compensatory mechanism when cytarabine and methotrexate killed the proliferating ERC, pronormoblasts, and normoblasts (28). An alternative explanation could be that more cell divisions might be inserted in the pronormoblast and normoblast compartments by a shortening of the intermitotic time (29). However, this latter process cannot be the reason for the overshoot because 59Fe is given after the cells of the maturing compartments have already been released into the blood. Thus, the overshoot indicates that the ERC pool was immediately replaced by an excessive number of stem cells. June 1981 Otherwise, a reduction of the 59Fe uptake value should have been observed when it was measured 72 hr after treatment with methotrexate or cytarabine because they kill ERC. Cyclophosphamide and bulsulfan reduced 59Fe uptake at all times, as expected, confirming that the action of cell-cycle stage nonspecific agents on nonproliferating stem cells can be measured by the 59Fe uptake method. Vinblastine reduced the 59Fe uptake at both 24 and 48 hr and also less severely at 72 hr, as expected.
Here again, the 59Fe uptake at 96 hr after vinblastine exceeded the control value, suggesting that the damage on erythropoietic system activated G( stem cells into cell cycle.
The fact that the 24-hr V9Fe uptake measured 72 hr after treatment with cytotoxic agents reflects the effect of the agents on the stem cell compartment is further supported by the following experiments. Three doses of 50 mg/kg of cytarabine were given every 2 hr over a 4-hr period and 24-hr 59Fe uptake was measured 72 hr after the last dose of cytarabine. Under these conditions, the 59Fe uptake was not affected (Table 1). However, when a primary dose of 50 mg/kg of cytarabine was given 12 hr prior to the same series of three doses of 50 mg/kg at 2-hr intervals and 24-hr 59Fe uptake was measured 72 hr after the last dose, a significant reduction of 59Fe uptake by 65% was produced. On the basis of the spleen colony forming unit assay (GFU-S) (30), it has been shown that pluripotent stem cells in the Go state can be induced to synthesize DNA and divide by cytarabine ( Thus, the reduction of 59Fe uptake following a primary dose of cytarabine, but no reduction without the primary dose, further supports the notion that the 24-hr 59Fe uptake measured 72 hr after treatment with cytotoxic agent reflects the effects of agents on nonproliferating pluripotent stem cells.
Application of the 59Fe Uptake Method for the Assessment of Potential Erythropoietic Toxins Erythropoietic Toxicity of Benzene Measured by the 59Fe Uptake Method Effect of a single subcutaneous dose of benzene (88, 440, and 2200 mg/kg as a 50% benzene-olive oil solution) in mice was evaluated by parameters such as erythrocyte 59Fe uptake, total leukocyte counts and hematocrits 24 hr after benzene administration. To determine the 59Fe uptake, mice were given 59Fe (0.5 ,uCi, 20-40 ng of iron) as ferrous citrate 24 hr after benzene treatment and bled 24 hr later. Unless specified, 59Fe was injected IP throughout the study. Blood (0.2 ml) from each mouse was counted in a scintillation well counter and the percentage of the 59Fe taken up into erythrocytes was calculated, assuming a blood volume of 6% of the body weight. Statistical significance was determined by means of Student's t test. Experimental results in Table 2 indicate that the 59Fe uptake measurement is a more sensitive method than total leukocyte counts and hematocrits for evaluating bone marrow activity.
The stage of erythroid cell development sensitive to benzene was evaluated using the 59Fe uptake method. Mice were given single doses of benzene sc and its effect on 59Fe uptake was evaluated after five specific time intervals (Fig. 3). No suppression was found after 1 and 12 hr and also 72 hr, whereas dose-dependent inhibition of the 59Fe uptake was observed 24 hr and 48 hr after treatment with 440 or 2200 mg/kg dose. Thus, the data can be interpreted to suggest that (1) benzene did not interfere with an incorporation of iron into heme, (2) benzene interfered with proliferation of normoblasts and pronormoblasts, and (3)   The effect of toluene on benzene toxicity and metabolism was also studied by using the 59Fe uptake method.
[3H]Benzene was given subcutaneously (SC) at a dose of 440 or 880 mg/kg and the metabolites in urine were measured over a 24 hr period with or without toluene (1720 mg/kg, SC). At 48 hr after giving benzene, 59Fe was administered to determine the degree of benzene toxicity. Table  3 shows that toluene reduced the appearance of benzene metabolites in the urine to 30-44% of controls. Toluene did not affect 59Fe uptake but did alleviate the reduction in 59Fe uptake caused by benzene. Taken together, these results suggest that inhibition of benzene metabolism was accompanied by alleviation or prevention of benzene toxicity and that benzene toxicity was probably   (17) aNumber of animals given in parentheses. bCV = coefficient of variation (standard deviation/mean x 100). cSignificantly lower than the value of control group (p < 0.001).
caused by a toxic metabolite of benzene. The rocytes was calculated assuming a blood volume of implication of benzene metabolites in benzene toxic-6% of the body weight. ity is further described in review articles (33,34).
Sham-operated and partially hepatectomized con-The liver is generally accepted as the major site trol rats showed no difference in 59Fe uptake, of benzene metabolism. Thus, the role of the liver in indicating that partial hepatectomy itself had no producing benzene-induced bone marrow depres-effect on 9Fe uptake. Benzene reduced 59Fe uptake sion was investigated using partially hepatectomto about half of the control values in sham-operated ized rats. Partial hepatectomy (35) resulted in animals (Table 4). No reduction in 59Fe uptake, removal of 75 + 8% (16 rats) of the liver. Sham however, was observed in the benzene-treated operations were performed by making a middle partially hepatectomized rats, showing that partial incision reaching 1-1.5 cm posteriorly from the hepatectomy protected the rat against benzenexiphoid process of the sternum, exposing the me-induced bone marrow depression. On the other dian and left lateral lobes of the liver, and then hand, partial hepatectomy markedly reduced benclosing the area with no liver excision. A single zene metabolism, confirming the requirement for dose of benzene (2200 mg/kg) was injected SC 8 hr the production of a metabolite which mediates after surgery. Generally, labeled [3H]benzene (ap-benzene toxicity. proximately 80,000 dpm/lxmole) was administered to determine benzene metabolites in urine (36 hr E cumulative urinary radioactivity). Total radioactiv-Effect of Pentobarbital on the Action ity in urine was reported as benzene equivalents in of Cytarabine and Hydroxyurea to ,umoles. To determine radioiron utilization, 59Fe Hemopoietic Cells Measured by the 59Fe (1.0 ,uCi per rat) was injected IP 48 hr after Uptake Method benzene administration and sample blood was collected 24 hours after 59Fe administration for analy-Since it has been reported that some hypnotics sis. The percentage of 59Fe incorporated into eryth-protected hemopoietic toxicity caused by x-irradiation Table 3. Effects of toluene on benzene metabolism and on "9Fe uptake in the mouse.
dSignificantly different from both control and toluene groups (p < 0.05).
bThree doses were given at 2 hr intervals. The 24 hr°9Fe uptake was measured 48 hr after the last dose of cytarabine. cTwo or three successive doses of 60 mg/kg to maintain anesthesia for 6 hr. dPentobarbital (60 mg/kg) was given 1 hr prior to cytarabine and anesthesia was maintained for 6 hr by 1 or 2 additional dose(s) of pentobarbital.
eSignificantly higher than the value of cytarabine treated group (p < 0.001). (36)(37)(38) and cytotoxic agents (39), the effect of pentobarbital on the cytotoxic action of cytarabine and hydroxyurea to erythrocytic progenitors (pronormoblasts) and hemopoietic stem cells in cell cycle were investigated using the 59Fe uptake method. Data in Table 5 show that pentobarbital alone had no effect on pronormoblasts, while cytarabine given in three doses of 25, 50, or 100 mg/kg every 2 hr produced dose-dependent damage on pronormoblasts. When pentobarbital was given 1 hr prior to cytarabine and anesthesia was maintained until 1 hr following the last dose of cytarabine, the inhibition of 59Fe uptake by cytarabine was significantly reduced at 50 and 100 mg/kg x 3, indicating that pentobarbital alleviated the toxicity of cytarabine on proliferating erythroid cells. Similar results were also obtained with hydroxyurea (Table 6).
Experimental results of pentobarbital action on the proliferating hemopoietic stem cells challenged with hydroxyurea is shown in Table 7. Here also, pentobarbital alone had no effect on the stem cells. 34 Table 6. Effect of pentobarbital on the action of multiple doses of hydroxyurea on the 59Fe uptake measured 48 hr after treatment.
cTwo or three successive doses of 60 mg/kg to maintain anesthesia for 6 hr. dThree doses were given at 2 hr intervals. The 59Fe uptake was measured 48 hr after the last dose of cytarabine.
'Pentobarbital (60 mg/kg) was given 30 min prior to hydroxyurea and anesthesia was maintained for 6 hr by 1 or 2 additional dose(s) of pentobarbital.
fSignificantly higher than the value of hydroxyurea treated group (p < 0.001).
Environmental Health Perspectives  bPentobarbital anesthesia was maintained for 6 hr by two or three successive doses of 60 mg/kg. "Significantly lower than control values (p < 0.001).
dSignificantly higher than the value of cytarabine treated group (p < 0.001).
Hydroxyurea, which is also a S-phase specific agent (16,40,41), did not affect the 59Fe uptake when the stem cells are in Go. However, when an initial dose of hydroxyurea (2000 mg/kg) was given 12 hr prior to three successive doses of hydroxyTurea (500 mg/kg every 2 hr, schedule A), the 5 Fe uptake value was reduced to 32% of the hydroxyurea treated group given schedule A alone. Several groups of investigators have independently shown that an initial dose of hydroxyurea activated Go stem cells into cell cycle on the basis of the CFU-S assay (11,42,43). Thus, the present data confirms that the S-phase specific agent, hydroxyurea, can only kill the proliferating cells whether they are stem cells or pronormoblasts but it can not kill cells in the Go state such as resting hemopoietic stem cells. When pentobarbital is given 30 min prior to the three successive doses of hydroxyurea following an activation of Go stem cells, the 59Fe uptake value was increased twofold. This protective effect of pentobarbital on proliferating stem cells was also observed when they were insulted with cytarabine ( Table 8). The mechanism by which pentobarbital protects proliferating hemopoietic cells from cytotoxic agents remains to be seen. In summary, the proper application of ferrokinetics and other developmental characteristics of erythropoietic cells enables the establishment of a methodology which can be used to estimate the potential toxic effect of chemicals on erythropoietic cells and pluripotent hemopoietic stem cells.