WO1993005807A2 - Erythropoietin potentiating agents and methods for their use - Google Patents

Abstract

Methods of potentiating the effect of erythropoietin for growth and differentiation of erythroid precursor cells are disclosed. Also disclosed are pharmaceutical compositions, which are useful in the disclosed methods.

Description

ERYTHROPOIETIN POTENTIATING AGENTS AND METHODS FOR THEIR USE

Background of the Invention

The development of red blood cells from undifferentiated progenitor cells in bone marrow requires the regulatory influence of the glycoprotein hormone erythropoietin (EPO) . EPO triggers differentiation by binding to receptors on the plasma membrane of immature erythroid precursor cells.

There are two principal types of EPO responsive, immature erythroid progenitor cells. Colony-forming unit- erythroid (CFU-E) is a cell that gives rise to a small colony of 8 to 50 mature, hemoglobin-containing erythroblasts in 2 days (mouse) to 7 days (human) . (Axelrad et al. , Properties of Cells that Produce Erythrocytic Colonies In Vitro in Robinson .A. (ed.) Hemopoiesis in Culture, pp. 226-234, U.S. Government Printing Office, Washington, D.C.) Burst-forming unit- erythroid (BFU-E) is a less mature erythroid progenitor cell that gives rise to a large cluster of colonies of mature erythroblasts (several hundred to several thousand cells) in 8 days (mouse) to 14 days (human). (Gregory and Eaves Blood. 49:855-864 (1977)). CFU-E are absolutely dependent on EPO for development in tissue culture, whereas BFU-E can survive for several doublings in the absence of EPO if other hematopoietic growth factors are present. (Sawada et al.. J. Cell Physiol. 142:219-230 (1990)) .

Other hematopoietic growth factors and interleukins, which appear to be important in stimulating erythroid progenitors, have generally been referred to as having burst promoting activity (BPA) , because they stimulate the growth of BFU-E. Among the factors which exhibit erythroid burst stimulatory activity are the ultipoietins, granulocyte-macrophage colony stimulating activity (GM-CSF; D. Metcalf, Science, 229:16 (1985)); T. M. Dexter, Science. 88:1 (1987), interleukin 3 (IL-3; Y.C. Yang et al. - Cell. 47:3 (1986); C . Sieff, J. Clin. Invest._f 79:1549, (1987), platelet derived growth factor (PDGF; N. Dainiak, et al. , J. Clin. Invest.. 71:1206 (1983); F. Delwiche, J. Clin. Invest. , 76:137 (1985), insulin-like growth factors (IGFs; K. Akahane, et al. _ψ Exp. Hematol. _f 15:797 (1987)); N. Dainiak and S. Kreczko, J. Clin. Invest., 76:1237 (1985), erythroid potentiating activity (EPA; C . Westbrook et al.. J. Biol. Chem.. 259:9992 (1984)); E. Niskanen, Blood, 72:806 (1988), and B-lymphocyte derived burst promoting activity, or B-BPA (N. Dainiak and CM. Cohen, Blood. 60:583 (1982); L. Feldman, et al. ■ Proc. Natl. Acad. Sci. USA, 84:6775 (1987); L. Feldman and N. Dainiak Blood. 73.:1814 (1990). In addition to stimulating BFU-E, IL-3, PDGF and B-BPA also stimulate the proliferation of CFU-E.

B-lymphocyte derived burst promoting activity (B-BPA) , a lineage specific regulator of erythroid pro¬ genitors, was first described by Dainiak and Cohen as a component of serum-free lymphocyte conditioned medium (N. Dainiak and CM. Cohen, Blood. 60:583 (1982)). B-BPA is an integral membrane protein of 28 kDa, and its activity recently has been localized to the plasma membrane of normal resting B lymphocytes (L. Feldman, et al. , Proc. Natl. Acad. Sci. USA. 84:6775 (1987); L. Feldman and N. Dainiak, Blood, 73:1814 (1990)). In serum-free culture of normal human bone marrow progenitors, B-BPA is a potent stimulator of BFU-E proliferation, increasing the number of BFU-E derived colonies up to 600% above control while having no apparent effect on the proliferation of granulocyte-macrophage (CFU-GM) , megakaryocyte (CFU-Meg) ■, or mixed (CFU-GEMM) hematopoietic colonies (N. Dainiak et al.. EXP. Hematol.. 18:1073 (1985); L. Feldman, et al.. Proc. Natl. Acad. Sci. USA_f 84:6775 (1987)). B-BPA appears to be biochemically and immunochemically distinct from other growth factors with BFU-E directed growth promoting activities (L. Feldman, et al.. Proc. Natl. Acad. Sci. USA, 84:6775 (1987); L. Feldman and N. Dainiak, Blood. 73:1814 (1990)).

In vertebrates, EPO is synthesized in response to hypoxia. Increased secretion of EPO into the serum is observed after blood loss and other causes of hypoxia such as high-altitudes. Administration of exogenous EPO has proven useful in treating patients suffering from severe anemia, resulting, for example, from chronic renal fail¬ ure. Administration of EPO has advantages over trans¬ fusions, which can transmit infectious diseases including hepatitis and AIDS and cause liver toxicity resulting from iron overload. Administration of EPO is, however, costly.

A method of eliminating the need for exogenous EPO or of reducing the dose of exogenous EPO required by a patient, and therefore decreasing the cost of EPO treat¬ ment, would be very useful.

Summary of the Invention

The present invention relates to Applicants' finding that certain agents potentiate (i.e., increase) the effect of erythropoietin (EPO) on the growth of EPO-responsive erythroid precursor cells and on their differentiation into red blood cells. Such agents are referred to herein as potentiating agents. Thus, the invention comprises, in one embodiment, a method of enhancing, or increasing, the effect of EPO in a vertebrate (particularly humans and other mammals) by administering to the vertebrate an effective amount of a potentiating agent. As a result of administering the potentiating agent, the amount of EPO required for growth and differentiation of erythroid precursor cells is decreased (i.e. less than it would be in the absence of the potentiating agent) .

Since the presence of EPO in an individual promotes the development of red blood cells from undifferentiated progenitor cells, the invention comprises, in another embodiment, a method of promoting erythropoiesis in a vertebrate by administering to the vertebrate an appropriate amount of a potentiating agent.

According to one embodiment of the subject method of promoting erythropoiesis, the potentiating agent is administered alone in a manner such that it potentiates the vertebrate's endogenous supply of EPO. Alternatively, the potentiating agent can be administered alone, followed by administration of exogenous EPO. This method, known as "priming", results in a large amplification of the erythrαpoietin biologic response characterized by an increased percentage of EPO-responsive cells, an increased rate of EPO responsiveness and a profound increase in sensitivity to EPO, as manifested by a left shifted EPO dose-response curve. According to another embodiment of the method of the invention, the potentiating agent is administered in conjunction with EPO administration. Co-administered EPO and potentiating agent synergize to affect the growth and differentiation of erythroid precursor cells. The invention, therefore, also comprises a composition comprising a potentiating agent and erythropoietin.

The presence of EPO in an individual also promotes the proliferation of EPO-responsive erythroid precursor cells. Therefore the invention comprises, in another embodiment, a method of repopulating progenitor cells in bone marrow of a vertebrate, comprising administering to the bone marrow an appropriate amount of a potentiating agent. The potentiating agents can be either biological or chemical in nature. An example of a biological potentiating agent is B-lymphocyte derived burst promoting activity (B-BPA) , which will be known herein as erythroid colony stimulating factor (E-CSF) . E-CSF is a preferred potentiating agent, because it is present naturally in a vertebrate host, suggesting that administration of additional amounts will not result in adverse immune reactions by the host. The invention further relates to methods of obtaining E-CSF.

Preferred chemical potentiating agents are polar/ planar compounds, such as dimethylsulfoxide (DMSO) and hexamethylene bisacetamide (HMBA) and cytodifferentiating chemicals such as sodium butyrate.

A major advantage of using the methods and compo¬ sitions of the subject invention is that erythropoiesis can be induced in an individual with smaller quantities of EPO, which means that less exogenous EPO or no exogenous EPO is required to induce erythropoiesis in an individual, such as in treating a patient suffering from hypoxia, which results for example from severe anemia.

Detailed Description of the Invention

The present invention is based on the finding that contacting erythroid precursor cells with a potentiating agent increases their response to erythropoietin (EPO) . EPO-responsive erythroid precursor cells include early burst forming unit-erythroid (BFU-E) and late colony-forming unit erythroid (CFU-E) .

EPO potentiating agents can be biological or chemical in nature and can be obtained by a variety of methods from a variety of sources. An example of a biological potentiating agent is erythroid potentiating B-lymphocyte derived erythroid burst promoting activity (B-BPA) . B-BPA is a lineage specific but pleiotropic regulator of erythropoiesis. B-BPA has distinct effects on both early (BFU-E) and late (CFU-E) erythroid progenitors from normal human bone marrow. Given its pleiotropic effect, B-BPA is more appropriately called E-CSF, erythroid colony stimulating factor. Analogs and derivatives of E-CSF which have the same biological effects are also useful in the subject invention.

Human E-CSF has been purified from normal peripheral blood B-lymphocytes, as explained in detail in Example 1. In addition to normal B-lymphocytes, other transformed cell types may yield the protein. Two B-cell lines, Raji and Ramos (ATCC designation Nos. CCL86 and CRL1596, respectively) , both derived from patients with Burkitts lymphoma, have been found to yield particularly high levels of the protein. Therefore, these cell lines provide a preferred source of E-CSF.

E-CSF activity has also been identified in medium conditioned by, and on the membranes of, lymphocytes from non-human sources, including mouse, rat, rabbit and bovine. E-CSF can be purified from non-human sources essentially as described for human E-CSF.

Alternatively, E-CSF, can be produced synthetically. For example, E-CSF can be cloned by indirect expression. (Sambrook, J. e al. - eds. Molecular Cloning _f Cold Spring Harbor Laboratory Press (1989)). Alternatively, E-CSF can be cloned by direct expression. For example, a cDNA library can be prepared from B cells or another cell type secreting E-CSF. The library can then be introduced into a suitable expression vector such as CDMB (Seed, B. and A. Aruffo, Proc. Natl. Acad. Sci. USA 84:3365-3369 (1987)). The CDM8 library can be divided into pools and used to transfect COS cells, which are allowed to express E-CSF. The COS culture expressing the factor can be identified for example, by bioassaying the supernatant medium, or by antibody identification of either soluble factor (e.g.. using RIA or ELISA) or membrane bound factor (D'Andrea, A.D. et al.. Cell 57: 277-285 (1989)). The cDNA pool thus identified can then be subdivided and the process repeated until a pure clone of E-CSF cDNA is obtained.

An alternative method of direct expression involves "panning" the transfected COS cells with an anti-E-CSF antibody, thereby enriching for cells containing the cDNA. The panned cells are lysed and the cDNA is harvested and used to transform bacteria where it is amplified. The amplified cDNA is used to transfect cells and the cycle is repeated until a pure cDNA clone is obtained.

The results of experiments, which are explained in detail in Example 1, show that E-CSF acts on both early (BFU-E) and late (CFU-E) erythroid progenitors and that it has different effects on these two classes of progenitors. E-CSF stimulates the proliferation of BFU-E, resulting in an increase in the number of erythroid bursts over a wide range of EPO concentrations. In contrast, E-CSF increases the sensitivity of CFU-E to their primary regulator, EPO. E-CSF potentiates the effect of EPO for growth and differ¬ entiation of erythroid precursor cells, resulting in a left shift in the EPO dose response curve.

The presence of E-CSF plus EPO in contact with erythroid precursor cells results in a synergistic. effect, wherein erythroid colonies produced are dramatically larger and hemoglobinized to a much greater extent than is the case with either growth factor alone. E-CSF alone, in the absence of EPO is able to effect proliferation of erythroleukemia cells. However, the presence of E-CSF and EPO synergize to affect the growth and differentiation of cells.

Medium conditioned by lymphocytes from non-human sources, including mouse, rat, rabbit and bovine, stimulate the proliferation of erythroleukemia cells in liquid culture and the proliferation of Rauscher erythroleukemia cell colonies in serum-free fibrin clot cultures. In combination with EPO, these "animal-derived E-CSF's" support the growth and development of large, well hemoglobinized Rauscher cell colonies in serum-free fibrin clots in a manner indistinguishable from that of human E-CSF.

In addition to E-CSF, some chemical agents have been shown to induce potentiation of EPO on growth and differ¬ entiation of EPO-responsive erythroid precursor cells. For example, polar/planar compounds, such as dimethyl sulfoxide (DMSO) and polymethylene bisacetamides such as hexamethylene bisacetamide (HMBA) (Reuben, R.C et al.. Biochimica et. Biophysica Acta 605:325-346 (1980)); and cytodifferentiating chemicals, such as sodium butyrate are preferred chemical potentiating agents.

The results of studies, which are set forth in detail in Example 2, demonstrate that DMSO potentiates the biologic response of erythroid precursor cells to EPO in three ways: First, the total percentage of EPO responsive cells is increased by DMSO. Second, the rate of response to EPO is increased, i.e., the time required for appearance of maximum numbers of hemoglobinized (Hb⁺) cells is reduced. Third,the EPO sensitivity of cells is increased approximately 20-fold, as manifested by a pronounced leftward shift of the EPO dose-response curve. These changes are accompanied by an almost seven-fold increase in the erythropoietin receptor density.

The effect of DMSO on the erythropoietin response has been compared with the effect of two other chemicals, hexamethylene bisacetamide (HMBA) , another pola /planar compound, and sodium butyrate, a cytodifferentiating compound. The protocol for these experiments is set forth in detail in Example 3.

Ih general, Rauscher cells, a recognized model of erythroid precursor cells, were pretreated (i.e., primed) for 24 hours with 1% DMSO, 2 mM HMBA, or 1 mM sodium butyrate, washed, and then grown in the absence or presence of 10 international units of EPO per ml for 2 days. The percentage of erythropoietin-specific Hb⁺ cells was determined (Table 1) . Hb⁺ cells are differentiated cells and are evidenced by staining with benzidine. Hb⁺ values were determined by subtracting the % Hb⁺ cells detected in replicate cultures pretreated with chemical inducers but incubated in the absence of EPO from that obtained in cultures incubated in the presence of EPO.

Table 1

Effect of pretreatment with polar/planar chemicals on erythropoietin-induced differentiation and receptor number of erythroleukemia cells

EPO-specific Hb⁺ EPO receptors,

The results presented in Table 1 show that Rauscher cells that were grown in the presence of EPO, differentiated. However, pretreatment with a chemical potentiating agent, followed by exposure to EPO, resulted, iii a large amplification of the differentiation induced by EPO. For example, about 40% of the Rauscher erythroid precursor cells that were first primed with HMBA and then contacted with EPO, differentiated. Further, about 50% of the Rauscher cells primed with sodium butyrate and then contacted with EPO, differentiated. Finally, about 60% of the Rauscher cells primed with DMSO and then contacted with EPO, differentiated. The amplification of the differentiation observed in erythroid precursor cells that were primed with a chemical potentiating agent and then contacted with EPO, was accompanied by a proportional increase in the erythropoietin receptor density in these cells.

Table 1 also shows that in identical experiments performed on Friend erythroleukemia cells, which do not hemoglobinize in response to erythropoietin, the presence of chemical potentiating agents do not induce differentiation. A small effect on receptor density was observed in the Friend cells, but it did not approach the magnitude seen with Rauscher cells.

These observations prompted further investigation of the erythropoietin receptor in DMSO- primed cells. The protocol for these experiments is set forth in Example 3. The binding of ¹²⁵I-erythropoietin to receptors on Rauscher cells which were not primed with a chemical potentiating agent was found to be concentration dependent and saturable. A Scatchard analysis revealed two slopes, consistent with two receptor populations. Calculations from this analysis revealed that one Rauscher murine erythroleukemia clone, PAN-4 cells have about 1000 higher-affinity receptors (K_d = 0.8 nM) and about 2000 lower-affinity receptors (K_d = 8.1 nM) per cell. The results for another Rauscher murine erythroleukemia clone, R28 cells, were strikingly similar. This density of receptors and the observation of two affinity populations are similar to those reported on other erythroleukemia cells (Broudy et al. _f Proc. Natl. Acad. Sci. USA- 85:6513-6517 (1988)) and contrast with 350 receptors per cell found on the parent, uncloned Rauscher line.

The results presented in Table 1 show that erythro¬ poietin receptors of Rauscher cells were increased dramatically (i.e., approximately 6 to 7 fold) by DMSO, HMBA and sodium butyrate priming. The results of concentration binding studies showed that at lower ¹⁵-ι- erythropoietin concentrations the binding to DMSO-primed cells was identical to that of unprimed cells. However, binding at higher ¹²⁵-I-erythropoietin concentrations revealed the presence of an apparently new population of high-affinity receptors on DMSO-primed cells. This unexpected observation was made consistently on R28 and PAN-4 cells in several experiments using cells of widely different passage number.

The binding curve of DMSO-primed PAN-4 cells was analyzed by the method of Scatchard. The Scatchard analysis of the portion of the binding curve involving lower concentrations of ¹²⁵I-erythropoietin shows a single slope consistent with 1000 receptors per cell with a K_d=1.0 nM, virtually identical to that of the higher affinity receptor population of non-primed cells (i.e., cells that were not pretreated with DMSO) (K_d=0.8 nM) . In marked contrast, the Scatchard analysis of the DMSO-induced, high-density receptor population reveals a pronounced, convex upward geometry. This finding is considered to be indicative of positive cooperativity, - that is, an increasing affinity of the receptor for erythropoietin with increasing site occupancy (Boeynaems and Du ont, J. Cyclic Nucleic Res. _f 1:123-142 (1975)).

A Hill plot of these data exhibits a coefficient (nH) of 6.75, indicative of marked positive cooperativity among these DMSO-induced erythropoietin receptors. In contrast, the lower affinity noncooperative receptor population found on DMSO-primed cells exhibits a Hill coefficient of nH=0.82. Because of this thermodynamic characteristic, an equilibrium dissociation constant for the DMSO-induced, high-a finity receptors could not be calculated from these data. However, the binding curve revealed saturation to be approached at about 20,000 receptors per cell, a 7-fold increase over nonprimed cells. The association of the appearance of this new high-affinity receptor class with an amplified erythropoietin biologic response (Table 1) strongly suggests that the high-density cooperative receptor class is biologically functional.

One possible mechanism for the up-regulation of the erythropoietin receptor is that potentiating agents increase transcript levels of EPO receptor mRNA. By Northern blot analysis of Friend cell mRNA probed with erythropoietin receptor cDNA following treatment with DMSO for varying periods of time, DMSO priming was found to result in an initial decrease in receptor message followed by an increase of approximately 2- to 3-fold after 24-48 hours. No change was seen in glyceraldehyde-3-phosphate dehydrogenase, a control, housekeeping gene. This observed increase in receptor message as a result of DMSO priming does not appear to account for the magnitude of the receptor up-regulation.

In addition to an induced increase in receptor transcript levels, whether due to an effect on gene transcription or on mRNA processing, other possibilities exist that might explain the increase in receptor binding induced by EPO potentiating agents. Post-translational processing of receptor protein leading to stabilization within the cell and/or increased efficiency of transport to the plasma membrane surface may play a role. Ad¬ ditionally, an EPO potentiating agent may exert a direct effect on the plasma membrane itself, thereby "unmasking" . otherwise "buried" receptor binding units. Utilitv

These results point to a clinical role for EPO potentiating agents in vertebrate animals (i.e., animals having an immune system) , including humans. For example, potentiating agents can be administered to individuals suffering from any disease or condition which results in an inadequate supply of hemoglobin containing red blood cells.. For example, EPO potentiating agents can be administered to patients recovering from chemotherapy or to patients with chronic renal failure.

EPO potentiating agents can be administered alone or in conjunction with EPO to vertebrates to promote erythropoiesis (i.e. the development of red blood cells from undifferentiated progenitor cells, for example, BFU-E and CFU-E) . Administration can be by any route appropriate to the condition being treated. Preferably, administration of compounds containing E-CSF is via a parenteral route, such as by injection into the blood stream.

The compositions of this invention, (i.e. a potentiating agent alone or a potentiating agent in conjunction with EPO) can be employed in admixture with conventional excipients, (i.e., pharmaceutically accept¬ able organic or inorganic carrier substances) . The carrier must be "acceptable" in that it does not inactivate or otherwise adversely affect the EPO potentiating agent and is compatible with the other ingredients of the formulation and not deleterious to the recipient. The formulations may conveniently be presented in unit dosage form and may be prepared by any of the methods well known in the art of pharmacy.

For parenteral administration, particularly suitable are injectable, sterile solutions, preferably oily or aqueous solutions, as well as suspensions, emulsions or implants, including suppositories. Ampoules are convenient unit dosages. It will be appreciated that "an effective amount" of the potentiating agent (i.e., an amount effective to increase the effect of EPO on the growth of erythroid precursor cells and their differentiation into red blood cells will depend, for example, upon the severity of the condition being treated, the route of administration chosen and the amount- of EPO present in the vertebrate (i.e., the vertebrate's endogenous EPO supply). Similarly, an effective amount of EPO will depend, for example, upon the severity of the condition being treated, the route of administration chosen, the concentration of endogenous EPO and the amount of potentiating agent present in the vertebrate and the amount of potentiating agent being administered. The more potentiating agent present in the vertebrate, the less EPO need be administered. Effective amounts can be determined by attending physicians or veterinarians using conventional considerations, (e.g., by means of an appropriate, conventional pharmacological protocol) .

EPO potentiating agents can also be administered to repopulate progenitor cells. For example, at least one potentiating agent can be administered to individuals, who are in need of bone marrow transplants (e.g., patients with aplastic anemia, acute leukemias, recurrent lymphomas or solid tumors) . Prior to receiving a bone marrow transplant, a recipient is prepared by ablating or removing endogenous hematopoietic stem cells. This preparation is usually carried out by total body irradiation or delivery of a high dose of an alkylating agent or other chemotherapeutic, cytotoxic agents (Greenberger, J.S., Br. J. He atol. 62:606-608, 1986; Anklesaria, P. et al.. Proc. Natl. Acad. Sci. USA 84:7681-7685, 1987). Following preparation of the recipient, donor bone marrow cells are injected intra¬ venously (Thomas, E.P., Cancer 49: 1963 (1982)). EPO potentiating agents can be administered in conjunction with the transplanted bone marrow, or can be administered separately (e.g., subsequent to bone marrow transplantation.)

The present invention will now be illustrated by the following Examples, which are not to be seen as limiting in any way.

EXAMPLES

Example 1 Amplification of the Erythropoietin Biologic Response by E-CSF

Testing the Effects of E-CSF on Purified Target Cell Populations of Human and Murine Erythroleukemia Cells

In serum-free culture of human bone marrow pro¬ genitors, E-CSF was found to increase the number of day 12 burst forming unit erythroid tBFU-E) derived erythroid bursts in a concentration dependent manner. In the absence of added E-CSF, control cultures produced a maximum of 11 +/- 1.5 bursts at 2 U EPO/ml. Addition of 100 ul of a 1/50 dilution of E-CSF increased the maximum number of day 12 bursts to 33 +/- 2, while addition of a 1/5 dilution of E-CSF increased the maximum number of bursts to 38 +/- 1. The effect of E-CSF appears saturable, since the ten-fold increase in amount of E-CSF added (from 1/50 to 1/5 dilution of concentrated column-purified E-CSF) resulted in only a small, but significant, increase in the number of day 12 BFU-E.

In addition to increasing the number of BFU-E derived bursts, E-CSF also appears to lower the requirement for EPO during burst formation. To determine the magnitude of this "EPO sparing" effect, E-CSF was administered to early erythroid progenitors over an expanded range of EPO concentrations. Serum-free cultures were established in the presence of 0.00 - 2.0 U EPO/ml, in the absence or presence of a specified amount of E-CSF.

E-CSF was found to significantly increase the number of BFU-E over a wide range of EPO concentrations. At all EPO concentrations greater than 0.1 U/ml, E-CSF signifi¬ cantly stimulates the proliferation of BFU-E, resulting in a two to four fold increase in the number of day 12 erythroid bursts.

In order to determine whether E-CSF also acts on colony forming unit erythroid (CFU-E) , serum-free marrow cultures were examined. E-CSF has a strikingly different effect on more mature erythroid progenitors than the effect on BFU-E. While E-CSF increases the total number of BFU-E derived bursts, E-CSF did not increase the maximum number of day 7 CFU-E derived colonies in culture. In the absence of E-CSF, the number of CFU-E increased with increasing amounts of EPO in culture, reaching a maximum number of CFU-E at 1.0 - 2.0 U EPO/ml.

However, addition of E-CSF to the culture shifted the EPO dose-response curve to the left greater than 10-fold, resulting in a maximum between 0.05 - 0.1 U EPO/ml. Thus, distinct from its stimulatory effect on the proliferation of BFU-E, E-CSF acts on CFU-E and increases their sensitivity to EPO dramatically. E-CSF acts in concert with EPO to stimulate growth and differentiation of CFU-E.

In order to establish a homogeneous target cell population with which to study the growth regulatory properties of E-CSF and to develop an in vitro system wherein the action of E-CSF distinct from that of EPO can be investigated, several human and murine erythroleukemia cell lines were screened for their abilities to respond to this potentiating agent. The effect of human E-CSF on the proliferation of several independently derived clones of Rauscher murine erythroleukemia cells as well as two human cell lines, K562 and HEL was examined. All of these cell lines are EPO independent for proliferation in culture.

Early log phase cells were plated at limiting dilution in the absence or presence of E-CSF, incubated under specified conditions, and colony-containing wells were enumerated from replicate 96-well plates. E-CSF increased the plating efficiency of both murine and human erythroleukemia cells. Proliferation of three Rauscher clones; R404, R28 and EMSIII was increased by 51%, 166% and 50%, respectively; and that of K562 cells was in¬ creased by 52%. E-CSF had no apparent effect on HEL cell growth.

The specificity of E-CSF's effect on the prolif¬ eration of Rauscher erythroleukemia cells was demonstrated by plating R28 cells at limiting dilution in the presence of either 10% E-CSF, 10 - 100 U rmIL-3, or 10 100 U rm GM-CSF/ml. Under conditions where E-CSF markedly increased the plating efficiency of these Rauscher cells, the addition of either IL-3 or GM-CSF was without effect on the proliferation of these cells. The simultaneous addition of E-CSF plus IL-3 or of E-CSF plus GM-CSF provided no proliferative effect above that which was seen with E-CSF alone. The growth promoting activity of E-CSF on these cells is not affected by co-addition of EPO, consistent with the hypothesis that the action of E-CSF on cell proliferation is erythropoietin. independent.

The growth promoting effect of E-CSF on Rauscher erythroleukemia cells was also determined by plating the cells in serum-free fibrin clots (semi-solid medium) in the absence of added growth factor (i.e., "control" conditions) , in the presence of 10% E-CSF, in the presence of 2 U recombinant human erythropoietin/ml (rhEpo/ml) , or in the presence of both E-CSF and EPO. In the absence of any added growth factor, the cells formed very small diffuse colonies in semi-solid medium. The addition of E-CSF to the cultures resulted in colonies that were more compact and significantly larger than the control colonies. In the presence of EPO, the colonies were also somewhat larger than those of the control cultures. In addition, these colonies contained hemoglobinized cells. The presence of E-CSF plus EPO resulted in a synergistic effect, wherein the colonies were dramatically larger and hemoglobinized to a much greater extent than with either growth factor alone. These data indicate that E-CSF alone, in the absence of EPO,. was able to affect the proliferation of erythroleukemia cells, while E-CSF and EPO synergized to affect the growth and differentiation of the cells.

Data from three representative experiments are quantified in the table below (Table 2) . These data indicate that E-SCF alone increases the size of Rauscher cell colonies up to 13-fold while E-CSF plus EPO increases colony size up to 28-fold that of control colonies grown in the absence of added growth factors.

Table 2

EFFECT OF E-CSF ± EPO ON COLONY SIZE Cells/Colony (Mean ± SD)

Rauscher cells were plated in serum-substituted fibrin clot. Size of the resultant colonies was determined by counting individual cells in a gridded field. > 10 colonies were counted for each culture condition in a given experiment.

Collection of Peripheral Blood and Bone Marrow Mononuclear Cells

Lymphopheresis of hematologically normal volunteer donors was performed at the New England Deaconess Hospital Blood Bank. Informed consent and approval of the Institutional Review Board was obtained for each donor. Cells were fractionated by sedimentation in Ficoll-Paque (Pharmacia, Piscataway, NJ) as described (L. Feldman, et al.. Proc. Natl. Acad. Sci. USA. 84:6775 (1987)). Mononuclear cells layering at the interface were washed in minimal essential medium, alpha modification (MEM α; Gibco, Grand Island, NY) and depleted of platelets by the method of R.J. Perper, et al. , J. Lab. Clin. Med. 72:5 (1968) . Monocytes were removed by adherence to plastic tissue culture flasks. These procedures have been demonstrated previously to result in a nonadherent cell population which is greater than 97% lymphocytes by cytochemical and histochemical identification, (L. Feldman and N. Dainiak, Blood. 73:1814 (1990)). In some cases, B-lymphocytes were further enriched by separation of lymphocytes over nylon wool columns (S.A. Eisen, et al. _f Immunol. Commun. , 1:571 (1972)). Nylon wool "adherent" cells were recovered, mixed with AET-treated sheep red blood cells (SRBC) and, following rosette formation, were separated over Ficoll-Paque (A. Saxon, et al. , J. Immunol Methods. 12:285, (1976)). Cells layering at the interface were removed and washed with MEMα. This population of cells was shown to consist of greater than 92% B-lymphocytes (L. Feldman and N. Dainiak Blood, 73:1814 (1990)). Highly purified (greater than 99%) human splenic B-lymphocytes, purified by SRBC rosetting and panning with a panel of monoclonal antibodies, were provided by Dr. Arnold S. Freedman, Dana Farber Cancer Institute, Boston, MA. (A.S. Freedman, et al., Blood. 70:418 (1987)).

Bone marrow was aspirated from the posterior ileac crest of hematologically normal volunteers. Informed consent and approval of the Institutional Review Board were obtained for each donor. Cells were placed in MEMα containing 20 U preservative-free heparin/ml and separated over Ficoll-Paque. Light-density mononuclear cells were washed in MEMα and depleted of monocytes by plastic adherence.

Preparation and Fractionation of Conditioned Medium (CM) and Lymphocyte Plasma Membranes (PM)

Unfractioned lymphocytes, enriched B-cells, or highly purified B-cells were suspended at 5 x 10⁶ cells/ml in MEMα supplemented with L-glutamine, penicillin and streptomycin (all Gibco) and were incubated overnight 14-18 hours) at 37°C, 5% C0₂ humidified air. CM was harvested by centrifugation at 00xg for 10 minutes, and cell-free.CM was fractionated into supernatants (S) and plasma membrane vesicle rich pellets (P) as described previously (L. Feldman, et al.. Proc. Natl. Acad. Sci. USA, 84:6775 (1987)). CM(S) were sterilized through 0.2u filters and stored at 4°C CM(P) were stored at -70°C in phosphate buffered saline (5mM sodium phosphate - 150mM sodium chloride, pH 7.6; PBS).

Lymphocytes recovered from the above CM were washed twice with PBS, lysed by Dounce homogenization, and plasma membranes (PM) were isolated by differential centrifugation and sucrose gradient fractionation by modifications of the methods of L. Feldman, et al. , Proc. Natl. Acad. Sci. USA. 84:6775 (1987); M. Jett, et a , J^. Biol. Chem.. 252:2134 (1977). Purified PM were stored at -70°C in PBS.

Purification of Ervthrocvte Colony Stimulating Factor E-CSF

E-CSF was solubilized from PM and CM(P) by sequential extraction with O.IN NaOH and 30mM octyl yff-D-glucopyranoside (Calbiochem, La Jolla, CA) (L. Feldman, et al. - Proc. Natl. Acad. Sci. USA. 84 : 6775 (1987); L. Feldman and N. Dainiak Blood. 73:1814 (1990)). Protein in solubilized conditioned medium (CM(S)) was concentrated by (NH₄)₂S0₄ fractionation. Solubilized membrane proteins and concentrated proteins from CM(S) were fractionated on Sephacryl S-300 (Pharmacia) , DE-52 DEAE cellulose (Whatman, Clifton, NJ) , and hydroxylapatite (Calbiochem) columns as described to generate "column purified E-CSF", (L. Feldman, et al.. Proc. Natl. Acad. Sci. USA. 84: 6775 (1987)). "Partially purified E-CSF" was designated as that material with E-CSF activity which was recovered following S-300 chromatography. Partially purified or column purified E-CSF also can be electrophoresed on SDS-PAGE and electroeluted to generate "gel purified E-CSF". Alternatively the electrophoresed protein(s) can be transferred to a PVDF membrane, such as Immobilon P (Millipore, Bedford, MA) . The band corresponding to E-CSF can be excised and used either to immunize animals for antibody production and/or to generate the amino acid sequence of the gel purified protein. (Matsudaira, P., Ed Practical Guide to Protein and Peptide Puri ication for Microsequencing Acad. Press (1989)) .

E-CSF also has been solubilized from PM and CM(P) prepared from transformed β-lymphocytes (Raji, Ramos cells) by the same method of sequential extraction described above. Transformed β-lymphocyte cell lines, Ramos and Raji, are maintained in RPMI 1640 - 3% fetal calf serum. Twenty-four hours prior to use, cells are washed extensively in serum-free RMPl 1640 and resuspended in the same serum-free medium for CM production (incubation overnight at 37°C, 5% C0₂ in humidified air) . Harvesting of CM and preparation of PM from the recovered cells is precisely the same as for the normal peripheral blood lymphocyte populations. While E-CSF has not yet been purified further from these cells, it is anticipated that the same chromatographic procedures as were applied to extracts from normal lymphocytes will prove applicable to the Raji and Ramos extracts. However, since the membranes of transformed cells are known to differ from those of normal lymphocytes, it is expected that the presence of additional proteins may alter the behavior of E-CSF during purification. In this event, additional chromatographic steps, including reversed phase HPLC, will be employed for the purification.

E-CSF can be purified from non-human sources essentially as described for human E-CSF. Lymphocytes can be purified from non-human blood by the same procedures used to purify lymphocytes from human lymphopheresis products. In the cases where organs, such as spleens, are . used as a source of lymphocytes, the organs can be minced/disaggregated and single-cell suspensions can be prepared. Lymphocytes then are purified from the single-cell suspensions in the same manner as they are purified from whole blood or blood products. Conditioned media from non-human lymphocytes are harvested and fractionated into supernatants (CM(S)) and membrane-vesicle containing pellets (CM(P)) the same as for human-derived material. Preparation of lymphocyte plasma membranes (PM) and subsequent sequential extraction of E-CSF is identical for human and non-human resources. While E-CSF has not yet been further purified from non-human sources, it is anticipated that the same chromatographic and electrophoretic manipulations can be used for non-human as for human E-CSF. However, since the membranes of these non-human cells may differ from those of normal human cells (e.g., in protein and/or lipid composition) , it is expected that the presence of these differences may alter the behavior of E-CSF during purification. In this event, additional steps can be employed for the purification as needed.

Growth and Maintenance of Erythroleukemia Cell Lines

Clones R404, R28 and EMSIII are independently derived from primary Rauscher murine erythroleukemia line by limiting dilution, (N.J. DeBoth, et al.. Nature, 272:626 (London, 1978) . Rauscher cells, as well as human K256 and HEL cells, were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco) containing 10% heat inactivated (56°C, 30 min) fetal bovine serum (FBS; HyClone, Logan, UT) , (C.B. Lozzio and B.B. Lozzio, Blood. 45:321 (1975); P. Martin and T.H. Papayannopoulou, Science. 216:1233 (1982) . Serum-free Culture of Bone Marrow Progenitors

Adherent cell-depleted, human bone marrow progenitors were cultured in serum-free fibrin clots by modifications of the method of, E. Bruno, et al.. Exp. Hematol. , 16:371 (1988) ; and of the modifications of N. Dainiak, et al.. Hematol. 18:1073 (1985). Cultures contained highly purified recombinant human erythropoietin (rhEPO; 2 x 10⁵ U/mg) (Elanex Pharmaceuticals, Bothell, WA) at 0 - 2.0 U/ml and 10% (vol/vol) column purified or partially purified E-CSF. Control cultures contained 10% (vol/vol) NCTC-109 (GIBCO) in place of E-CSF. Cultures were incubated at 37°C, 6% C0₂ in humidified air and were harvested on day 7 (CFU-E) and day 12 (for BFU-E) . Fixed clots were stained with benzidine and counterstained with hematoxylin. CFU-E containing >8 cells and BFU-E containing >50 cells were counted by light microscopy. Data are expressed as the mean +/- SEM for quadruplicate determinations for each test point. Each experiment was repeated a minimum of three times.

Other Biological Assays for E-CSF

For experiments designed to determine the effect of E-CSF on the proliferation of continuous erythroleukemia cell lines, early log phase cells were diluted in DMEM^' 10% FBS, iii the absence or presence of 10% (vol/vol) E-CSF, and were plated at limiting dilution in replicate Falcon 96-well plates (Becton-Dickinson, Lincoln Park, NJ) . In some experiments, Rauscher cell cultures also contained 10 - 100 U recombinant murine granulocyte acrophage colony stimulating factor (rmGM-CSF)/ml or 10-100 U recombinant murine interleukin 3 (rmIL-3)/ml. rmGM-CSF and rmIL-3 were obtained from Genzyme Corp., Cambridge, MA. The experimental plates were incubated at 37°C, 5% C0₂ in humidified air and were examined under an inverted microscope on days 3, 7 and 10. Wells containing colonies of greater than 4 cells were counted.

In order to examine the effect of E-CSF, alone or in the presence of EPO, on colony formation of Rauscher erythroleukemia cells, the cells were plated in semisolid medium in serum-free fibrin clot cultures. This culture system is analogous to that which was employed for the culture of bone-marrow derived erythroid progenitors. Briefly, cells were diluted in DMEM and plated in the same serum-free medium, supplemented with 10-30% IX synthetic serum (Dainiak, et al. , Exp. Hematol. 18:1073, (1985)) and clotted with thrombin and fibrinogen. In addition, some cultures were supplemented with 10% E-CSF, or with 2U rhEPO/ml, or with both of these growth factors. Cultures were incubated at 37°C, 6% C0₂ in humidified air and were harvested after 6-8 days. Fixed clots were stained with benzidine and counterstained with hematoxylin. Hemoglobinized and nonhemoglobinized colonies were enumerated by light microscopy.

The proliferative effect of E-CSF also can be assayed by measuring the uptake of [³H]thymidine into responsive target cells. Several variations of this assay may be used, depending on the particular target cell employed.

Using phenylhydrazine-treated murine spleen cells (PHZ cells) and the assay developed by Krystal for assay of erythropoietin (Krystal G. Exp. Hematol 11:649, (1983)), the ability of added E-CSF to left-shift the EPO dose response curve can be measured. This is similar to the ability of E-CSF to left-shift the EPO dose response curve for CFU-E derived erythroid colonies in human bone marrow culture.

Derivative from this "Krystal-type" assay is the potential assay, using PHZ cells or enriched CFU-E from thiamphenicol-treated mice (J. Cell Biol 96:386 (1983)), where a specified number of cells are plated in the presence of a single sub-saturating dose of EPO minus or plus E-CSF. Cells plated in the presence of EPO plus E-CSF which potentiate the effect of EPO for CFU-E, will incorporate more [³H]thymidine at a given dose of EPO than cells plated in the absence of E-CSF.

Finally, as described in detail in Example 4, (erythroleukemia) cells, such as murine Rauscher or human K562 cells, which are EPO independent for their in vitro growth, at specified concentrations, are plated in serum- free on other suitable media, in the absence or presence of E-CSF for 12-96 hours and then pulse labeled with [³H]thymidine. This variation of the assay measures the direct proliferative effect of E-CSF on cells in the absence of EPO. Using the variation described herein, we have demonstrated a proliferative effect of E-CSF on Rauscher cells incubated serum-free for 96 hours which is up to 40-fold higher than the proliferation of these cells in the absence of E-CSF.

Protein Determination

Protein was estimated by the method of O.H. Lowry, et al.. J. Biol. Chem.. 193:265 (1951) using bovine serum albumin (BSA) as a standard.

Example 2 Amplification of the Erythropoietin Biologic Response by DMSO

Cell Culture and Induction Studies

R28 was subcloned from a primary Rauscher murine erythroleukemia cell line by limiting dilution. N.J. DeBoth et al.. Nature. 272:626 (1978); A. Hagemeijer, J. Natl. Cancer Inst.. 69:945 (1982). Cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (GIBCO, Grand Island, NY). To induce differentiation, DMSO (Sigma, St. Louis, MO) or highly purified recombinant human EPO (Elanex Pharmaceuticals, Inc. Bothell, WA) was added to the medium. In some experiments, DMSO was "removed" by repeating washing and centrifugation of cells in DMSO-free medium followed by replating in the absence of DMSO. The effect of DMSO treatment on cell proliferation (total number of cells) was examined by performing daily cell counts on replicate cultures grown for specified times in the absence or presence of DMSO. Hemoglobinized cells (Hb⁺) were assayed using benzidine staining. (R.A. Rifkind et al.. In Vitro Aspects of Erythropoiesis. New York, NY, p226 (1978)). Assays were performed by adding 10 ML freshly prepared benzidine reagent containing 0.6% H₂0₂, 0.5 mol/L CH_jCOOH, and 0.2% benzidine dihydrochloride to 50 μL of cells (0.5 to 1 x 10 cell/mL) in culture medium. The proportion of benzidine positive cells (blue cells) was scored out of 200 cells counted and is expressed as "Hb⁺ cells, %". Variation between duplicate counts and repeat experiments was usually less than 5%.

Different clones exhibited characteristic responses to the two inducers in terms of differentiation into erythrocytes, as represented by Hb⁺ cells as percent of total. One subclone, R28, which exhibited a low to moderate response to either DMSO or EPO, was selected for further examination. DMSO and EPO induced hemoglobiniza- tion of R28 to 33% and 8% respectively, after 3 days of exposure to the inducers. However, when the cells were exposed to both inducers simultaneously, a striking synergistic effect was observed. After only 2 days, 70% of the cells were hemoglobinized, while less than 10% of the cells were hemoglobinized when each inducer was used alone for this time period. A similar synergistic effect between DMSO and EPO was observed in two other clones of.Rauscher cells examined. Possibly, DMSO treatment inhibits the growth of some cells, thereby allowing a preferentially EPO-responsive subpopulation to proliferate. DMSO has been shown to exhibit a moderate growth inhibitory effect on some Rauscher cell clones when the agent was present in the culture for 3 days (DeBoth, N.J. et al. _f Nature. 272:626 (1978)). This possibility was tested by establishing three replicate cultures of 150,000 cells/mL and incubating them as follows: culture A, no DMSO; culture B, 24 hours with 1% DMSO followed by 24 hours without DMSO; culture C, 48 hours with 1% DMSO. After incubation, the total cell counts were as follows: culture A = 3.1 x 10⁶ ± 0.5 x 10⁶ cells/mL; culture B = 3.0 x 10⁶ ± 0.5 x 10⁶ cells/mL; culture C = 2.3 x 10⁶ ± 0.3 x 10⁶ cells/mL. These results indicate that the negligible effect of DMSO on growth is insufficient to explain the large increase in EPO response.

To find out whether the simultaneous presence of both inducers was necessary for this synergistic effect, Rauscher cells were pretreated with 1% DMSO for 24 hours, washed twice, and then incubated in the absence (EPO^" ) or presence (EPO⁺) of EPO (10 U/mL) . One day after the DMSO was removed, the EPO^" cells were 28% Hb⁺, whereas the EPO^÷ cells were 56% Hb⁺ (28% EPO-specific Hb⁺ cells) . Two days after DMSO removal, the EPO^" cells were only 6% Hb⁺ while the EPO⁺ cells were 45% Hb⁺ (35% EPO-specific Hb⁺) . This EPO response compares with only 1% Hb⁺ after 48 hours of EPO treatment without DMSO pretreatment. Thus, pretreatment with DMSO followed by its removal is nearly as efficacious in amplifying the EPO response as simultaneous addition of the two inducers. Pretreatment of cells with DMSO followed by its removal before addition of EPO is known as "DMSO priming".

To characterize "DMSO priming" further, cells were pretreated with 1% DMSO for various periods of time. DMSO was removed, and EPO was added to the medium for 24 hours. The percentage of hemoglobinized cells was then assessed. A slight amplification of the EPO response was seen after only 6 hours of priming. A clear effect was observed after 12 hours. Maximal amplification occurred after 48 hours of priming. The dose-response of "DMSO priming" was examined next by pretreating R28 cells with different concentrations of DMSO for 24 hours. After removing DMSO, EPO was added to the cells for 24 hours. An almost linear relationship was observed between the EPO specific hemoglobinization and the concentration of DMSO used for priming.

To further test the effect of DMSO priming on the number of EPO responsive cells, R28 cells were pretreated with 1% DMSO for 48 hours followed by EPO for 48 hours. Fifty-eight percent of the cells were EPO-specific Hb⁺. This contrasts with unprimed cells in which only 2% were Hb⁺ after 2 days of EPO. In addition, the maximal EPO response was seen earlier in the primed cells. A pro¬ nounced shift in the maximal EPO response time from 4 days in unprimed to 2 days in DMSO-primed cells is observed.

The EPO dose-response relationships in DMSO-primed and unprimed cells was examined. There is a marked left-shift of the dose-response curve. The EPO activation constant (K_act) , defined as the concentration of agonist that causes 50% of the maximal response (M.E. Maguire et al.. Adv. Cyclic Nucl. Res.. 8:1, (1977)) is 0.1 U/mL and 2 U/mL for DMSO-primed and unprimed cells, respectively, representing an increased sensitivity of approximately 20-fold. The length of EPO exposure had no significant effect in the K_act in either cell group under each condition tested.

To determine whether the converse was true, i.e., whether "EPO priming" amplified the DMSO response, cells were incubated in the absence or presence of 10 U EPO/mL for 24 hours. The cells were washed and plated in the presence of 1% DMSO. The percent of Hb⁺ cells in each culture was determined after an additional 16 hours and 48 hours of growth in DMSO. The unprimed cultures contained 0% and 23% ± 3% Hb⁺ cells, whereas the EPO-primed cultures contained 3% ± 2% and 26% ± 3% Hb⁺ cells after 16 hours and 48 hours of DMSO treatment, respectively. EPO priming therefore has no significant effect on the DMSO response.

The multiple effects of DMSO priming on the EPO biologic response in Rauscher cells are all consistent with an amplification of the EPO signal transduction pathway by the chemical inducer. To test this possibility, the effect of DMSO priming on the site of initiation of the signal, i.e., the EPO receptor itself, was examined.

The binding of ¹²⁵I-EPO to unprimed Rauscher cells was specific and saturable. A Scatchard analysis (G. Scatchard, Ann. N.Y. Acad. Sci.. 51:660-670 (1949)), of these binding data showed two populations of receptors with equilibrium dissociation constants (Kd) of 1.1 nmol/L and 6.2 nmol/L, respectively. The results indicate approximately 1,200 higher affinity receptors and 2,600 lower affinity receptors, for. a total receptor density of approximately 3,800 per cell. The results of ¹²⁵I-EPO binding to DMSO-primed Rauscher cells were strikingly different. At lower ¹²⁵I-EPO concentrations, the binding curve was virtually identical to that obtained with unprimed cells, a finding that was confirmed repeatedly. As expected, the Scatchard analysis of this portion of the curve was virtually identical to the higher affinity population. However, at higher ¹²⁵I-EPO concentrations, a new high density receptor population was detected, reaching near saturation at about 20,000 receptors/cell. The Scatchard analysis of ¹²⁵I-EPO binding to these new receptors found on DMSO-primed cells showed a pronounced upwardly convex geometry. This result, which was obtained each time in three separate experiments using both this R28 Rauscher cell line as well as another, independently derived line, is considered to be diagnostic of positive cooperativity among receptors. (J.M. Boeynaems, J.E. Dumont, J. Cyclic Nucl Res.. 1:123 (1975)).

Radioiodinated EPO

Recombinant human EPO (Elanex) was labeled using IODOGEN (Pierce) (Fraker, P.J., and Speck, J.C, Biochem. Biophys. Res. Commun. 80:849 (1978); K. Sawada et al. ■ J. Clin. Invest.. 80:357 (1987)) and carrier-free (Amersham; 174 mCi/μg iodine; 664 MBq/μg) . ¹²⁵I-labeled EPO was purified using BioGel P6DG (BioRad) gel filtration in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (Sigma) and 0.02% Tween-20 (Sigma). The specific radioactivity of ¹²⁵I-labeled EPO ranged from 0.1 to 0.4 gram atom/mol for different preparations. Iodinated EPO prepared using this method retained full biologic activity when assayed by an in vitro bioassay. (Krystal, G., EXP. Hematol. 11:649 (1983)).

Binding of Iodinated EPO to Cells

Cells were harvested by centrifugation, washed twice in Dulbecco's PBS, and incubated in DMEM containing 10% FBS, and 0.2% sodium azide (binding medium) for 30 minutes. Triplicate samples of 5 x 10⁶ cells in 200 μh of binding medium containing ¹²⁵I-EPO at specified concentrations were incubated in the absence or presence of 100-fold unlabeled EPO at 37°C for 30 minutes. At the end of the incubation, cells were transferred to microfuge tubes containing 200μL of FBS, centrifugated through the cushion of FBS for 5 minutes in a Beckman Microfuge 12, and frozen at -80°C The tip of the tube containing the cell pellet was cut off. Radioactivity in the cell pellet was measured in a Beckman Gamma 5500 counter. Results were corrected for nonspecific binding of ¹²⁵I-EPO. The nonspecific binding was in all cases less than 10% of the total. The binding data were analyzed by the method of Scatchard. (Scatchard, G., Ann. NY Acad Sci.. 51:660 (1949) . Total receptor number was calculated from specific binding (cpm) by using the following constants: carrier-free ¹²⁵I = 2.2 x 10⁶ Ci/gram atom; Ci = 2.2 x 10¹² dpm; 83% efficiency of Gamma 5500 counter determined experimentally (1 cpm = 1.2 dpm) ; specific activity of ¹²⁵I-EPO used in binding study shown equals 0.12 grams atoms iodine/mol EPO.

Example 3 Amplification of the Erythropoietin Biologic Response by DMSO. HMBA and Sodium Butyrate

Cells

Rauscher murine erythroleukemia were the generous gift of N.J. DeBoth (Erasmus University, Rotterdam) . Clone R28 was derived by limiting dilution from its parent line. Clone PAN-4 was derived by sequential "panning" (Wysocki and Sato, Proc. Natl. Acad. Sci. USA. 75:2844-2848 (1978)) on Petri dishes coated with streptavidin and biotinylated erythropoietin as follows. Petri dishes (Falcon 1024) were coated with 0.01 μg of streptavidin per ml (10 ml) in coating buffer (1.6 g of Na₂-C0₃ and 2.9 g of NaHC0₃ per liter) for 16 hours at 23°C The dishes were washed four times with sterile Dulbecco's PBS. Biotinylated, biologically active recombinant human erythropoietin (rhEPO) (10 μg) in 10 ml of PBS/1% bovine serum albumin was added for 2 hours at 23°C The solution was removed, and the dishes were washed five times with culture medium containing 1% FBS. Ten milliliters of 1 x 10₆ cells per ml in Dulbecco's modified Eagle medium (DMEM)/1% FBS were added to the dish and were incubated for 1 hour at 4°C Nonadherent cells were removed by gentle swirling and aspiration. Adherent cells were removed with a plastic cell scraper, washed twice, and recultured at 37°C for 48 hours. They were then repanned and recultured for a total of four cycles. The resulting cells were designated PAN-4. Friend murine erythroleukemia cells (clone 745) (Friend et al.. Proc. Natl. Acad. Sci. USA. 68:378382 (1971)) were the generous gift of Blanche Alter. Throughout these experiments, R28 and PAN-4 appeared identical by detection of hemoglobin before or after treatment with erythropoietin and/or by staining with benzidine (Rifkind, supra) . DMSO, HMBA and sodium butyrate were purchased from Sigma.

Erythropoietin

Highly purified rhEPO (Powell et. al. _f Proc. Natl. Acad. Sci. USA. 83:6465-6469 (1971)) was obtained from Elanex Pharmaceuticals (specific activity = 200,000 international units/mg) . It was >99% pure by SDA-PAGE (Laemmli^', U.K., Nature London. 227:680-685 (1970)). Erythropoietin was biotinylated with NHS-LC-biotin (sulfomuccinimidyl 6-(biotinamido) hexanoate; Pierce) . After quenching the reaction with 0.1 M glycine and dialysis to remove excess biotin reagent, the specific biological activity of the material was 80% that of the control, nonbiotinylated hormone (Krystal, supra) . - Absorption of the biotinylated rhEPO with immobilized streptavidin followed by bioassay revealed that greater than 75% of the biological activity was absorbed out by streptavidin, thus demonstrating that biotinylated rhEPO was biologically active. Additional studies demonstrated that reaction of biotinylated rhEPO with excess strepta¬ vidin in solution yielded about 50% of the specific biological activity of native erythropoietin, thus confirming that the rhEPO-biotinstreptavidin construct retained its capacity to recognize the erythropoietin receptor and could serve as an affiant in the panning procedure. rhEPO was labeled with ¹²⁵I- using Iodo-Gen (Pierce) as described (Sawyer et al.. Proc. Natl. Acad. Sci. USA. 84:3690-3694 (1987)). The radiolabeled hormone exhibited full biological activity and contained 0.2-0.4 mol of ¹²⁵I per mol of rhEPO.

Receptor binding

Cells were collected by centrifugation, washed twice with PBS, resuspended in DMEM/10% FBS/0.2% sodium azide (binding buffer) , and incubated at 0°C for 30 min to inhibit energy-dependent receptor-mediated endocytosis. Cells (5 x 10⁶) were added to glass test tubes, incubated for 5 minutes at 37°C and then incubated with specified concentrations of ¹²⁵I-erythropoietin in the absence or presence of a 100-fold excess of unlabeled erythropoietin (total volume -= 200 μl) . All determinations were done on triplicate samples. After 30 minutes of incubation, which had been determined in preliminary experiments to achieve equilibrium, each cell suspension was transferred onto a 200 μl cushion of FBS in a polypropylene centrifuge tube. The cells were sedimented by centrifugation at 9000 rp for 3 minutes (Beckman Microfuge) . The tube contents were frozen and the tips were cut off for measurement of bound radioactivity by γ scintillation spectrometry. Specific binding at a given erythropoietin concentration was defined as the difference in bound radioactivity between the mean of samples incubated in the absence or presence of a 100-fold excess of unlabeled erythropoietin. Total Cvtoplasmic RNA Extraction and Northern Blot Analysis.

Cytoplasmic RNA was prepared using guanidinium isothiocynate (Chirgwin, et al.. Biochemistry, 18:5294- 5299 (1979)). Forty micrograms of total RNA was subjected to electrophoresis in 1.2% agarose containing 5.5% formaldehyde and transferred to GeneScreen (DuPont) . The filters were hybridized sequentially with a ³P-labeled synthetic oligomer complementary to nucleotides 256-305 of the Friend murine erythroleukemia cell erythropoietin receptor cDNA (D'Andrea et a ._/ Cell. 57:277-285 (1989)) and with a ³²P-labeled plasmid containing a cDNA of glyceraldehyde- 3-phosphate dehydrogenase, a housekeeping gene (Fort et al.. Nuc. Acids Res.. 13:1431-1442 (1985)). The synthetic oligomer was end-labeled using T4 poly- nucleotide kinase and ( ~³²P)ATP (Richardson, CC, Proc. in Nucleic. Acid Res., p.815, Eds: Cantoni and Davies Harper & Row, N.Y. (1971)). The radiolabeled plasmid probe was generated by nick-translation (Rigby et al.. J. Mol. Biol.. 113:237-242 (1977)). Relative molecular mass of the mRNA that^~hybridized to each of the probes agreed with previously published values.

Example 4 f³H]Thymidine Assay for Cell Proliferation

1. Log phase Rauscher cells are pretreated by incubation in serum-free αMEM medium, or other suitable medium, at a concentration of 2 x 10⁶/ml for 5 hours at 37°C, 5% C0₂ in humidified air.

2. For assay, pretreated Rauscher cells then are diluted to 5 x 10⁴/ml in the specific media. 3. A "bioassay medium stock" is made from αMEM containing 1% deionized bovine serum albumin and 10% "synthetic serum" (see Dainiak et al.. Exp. Hematol. 18:1073, 1985 for composition of synthetic serum).

4. For each experimental point to be tested, a small snap-cap tube is made up containing 0.8 ml bioassay medium, 0.1 ml cells (5 x 10⁴/ml) , and 0.1 ml sample to be tested (i.e., E-SCF containing sample or αMEM for control tubes) .

5. Replicate (typically 3-6) lOOμl aliquots from each control or experimental tube are plated in sterile 96-well tissue culture plates.

6. Plates are incubated for up to 96 hrs. at 37°C, 5% C0₂ in humidified air.

7. At the end of the selected incubation period, cells are pulsed for two hours with [³H]thymidine (typically 25 μl of 20μCi/ml [³H]thymidine in αMEM is added to each well) .

8. Wells are harvested onto glass fiber filters and counted by ?-scintillation spectrometry.

Equivalents

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.

Claims

1. A pharmaceutical composition comprising a therapeutically effective amount of a potentiating agent, a therapeutically effective amount of erythropoietin and a pharmaceutically acceptable carrier, and for example, the potentiating agent is human erythroid colony stimulating factor.

2. A pharmaceutical composition of Claim 1 wherein the potentiating agent is a polar/planar chemical selected from dimethylsulfoxide, hexamethylene bisacetamide and sodium butyrate.

3. A pharmaceutical composition comprising a therapeutically effective amount of erythroid colony stimulating factor or an analog or derivative thereof, and a pharmaceutically acceptable carrier.

4. A pharmaceutical composition of Claim 3, further comprising a therapeutically effective amount of erythropoietin.

5. A pharmaceutical composition of Claim 3 or Claim 4 wherein the erythroid colony stimulating factor is human or animal erythroid colony stimulating factor.

6. Use, for the manufacture of a medicament for administering to a vertebrate, of a potentiating agent in an amount in the medicament effective to decrease the amount of erythropoietin required for growth and/or differentiation of erythroid precursor cells.

7. The use of Claim 6, wherein the potentiating agent is human or animal erythroid colony stimulating factor or an analog or derivative thereof.

8. The use of Claim 6, wherein the potentiating agent is a polar/planar chemical or a cytodifferentiating chemical.

9. Use, for the manufacture of a medicament for administering to a vertebrate, of a potentiating agent in an amount in the medicament effective to increase the red blood cell count.

10. The use of Claim 9, wherein the potentiating agent is human or animal erythroid colony stimulating factor or an analog or derivative thereof, or a polar/planar chemical, or a cytodifferentiating chemical.

11. Use, for the manufacture of treatment agents for promoting erythropoiesis in a vertebrate, of separate effective amounts of a potentiating agent and of erythropoietin.

12. Use, for the manufacture of a medicament for promoting erythropoiesis in a vertebrate, of an effective amount of a potentiating agent and an effective amount of erythropoietin.

13. An erythropoietin potentiating agent in conjunction with transplanted bone marrow for repopulating erythroid progenitor cells in an individual at the site of a bone marrow transplant, said agent for example being erythroid colony stimulating factor or an analog or derivative thereof.

14. An erythropoietin potentiating agent in conjunction with transplanted bone marrow for repopulating erythroid progenitor cells in an individual at the site of a bone marrow transplant, said agent for example being a polar/planar chemical, or a cytodifferentiating chemical.

15. Use of a potentiating agent, said agent for example being erythroid colony stimulating factor or an analog or a derivative thereof, for the manufacture of a medicament for increasing the number of erythropoietin receptors erythropoietin-responsive erythroid precursor cells.

16. An in vitro method of increasing the number of erythropoietin receptors on erythropoietin-responsive erythroid precursor cells, comprising contacting the precursor cells with an effective amount of a potentiating agent, said agent for example being an erythroid colony stimulating factor or an analog or derivative thereof.

17. Use of a potentiating agent for the manufacture of a medicament for increasing the number of erythropoietin-responsive erythroid precursor cells in a vertebrate said agent for example, being an erythroid colony stimulating factor.

18. Use of a potentiating agent, said agent for example being erythroid colony stimulating factor or an analog or derivative thereof, for the manufacture of a medicament for increasing the rate of response or sensitivity erythropoietin-responsive erythroid precursor cells to erythropoietin.

19. An in vitro method of increasing the rate of response or sensitivity of erythropoietin-responsive erythroid precursor cells to erythropoietin, comprising contacting the cells with an effective amount of a potentiating agent, for example erythroid colony stimulating factor.

20. A method of obtaining erythroid colony stimulating factor, comprising the steps of; a. solubilizing proteins form the plasma membrane of β-lymphocytes; and b. purifying E-CSF from the solubilized proteins, for example being the ?-lymphocytes the Raji cell line and the Ramos cell line.

21. The use of Claim 11 or Claim 12, wherein said agent is an erythroid colony stimulating factor or an analog or derivative thereof.

22. A method of assaying for a potentiating agent comprising the steps of: a. plating erythroleukemia cells in serum free fibrin clots in the presence of an agent to be assayed; b. plating erythroleukemia cells in serum free fibrin clots in the presence of both the agent to be tested and erythropoietin; and c. incubating the erythroleukemia cells to obtain fixed clots; d. staining fixed clots with benzidine and hematoxylin; and e. detecting the extent to which hemoglobinization of fixed clots occurs by detecting the degree of benzidine staining, wherein benzidine staining is indicative of hemoglobinization.