Trypanosoma brucei: attenuation by cortcosteroids of the anemia of infected mice

BALBc/J mice infected with pleomorphic stabilate of Trypanosoma brucei develop a severe anemia. The anemia is initially normoblastic and normocytic but becomes macrocytic due to brisk and sustained reticulocytosis. Red blood cells from uninfected animals are destroyed when transfused into infected animals. Neither extensive intravascular hemolysis nor red cell agglutinins were detected in infected animals, but spherocytes were common in the peripheral circulation. Treatment of mice with either dexamethasone or hydrocortisone greatly retarded the development of the anemia. Comparison of the course of infection in these treated mice to that in normal mice suggests that these corticosteroids attenuate the anemia through their immunosuppressive action.     

Effect of purified recombinant human erythropoietin on anemia in rats with experimental renal failure induced by five-sixth nephrectomy

Recombinant human erythropoietin (rHuEPO) was purified from the conditioned media of Chinese hamster ovary cells with a transfected human erythropoietin gene. We investigated the effects of the rHuEPO in rats with renal anemia induced by partial nephrectomy. Five-sixth nephrectomy resulted in renal failure with anemia. Twenty-five days after the operation plasma urea nitrogen was increased about 2.5 times, and the red blood cell count, hematocrit, and hemoglobin concentration fell to 85% of normal. The reticulocyte count and plasma erythropoietin level did not change such as they do in patients with anemia due to chronic renal failure. Both total red blood cell volume and the plasma iron turnover rate were depressed in five-sixth nephrectomized rats compared with normal rats.The five-sixth nephrectomized rats were injected with rHuEPO (60 IU/kg) intravenously every second day for a total of six injections. After three injections of rHuEPO, circulation volume of total red blood cells was increased from 9.9 ml to 14.6 ml, and the plasma iron turnover rate was increased from 1.03 mg/kg/day to 2.12 mg/kg/day, and the reticulocyte count was also increased. After six injections, a marked increase of the red blood cell count, hematocrit, and hemoglobin concentration were observed. Plasma urea nitrogen and the creatinine levels as indications for renal function did not change after rHuEPO administration in both normal and five-sixth nephrectomized rats.In conclusion rHuEPO has a potent erythropoietic action and it is possible to cure the anemia caused by renal failure.     

KINETICS OF ERYTHROPOIETIN STIMULATED PROLIFERATION OF ERYTHROID PROGENITORS IN THE SPLEEN

The effect of RBC transfusion and erythropoietin (EPO) on the proliferation of immature erythrocyte progenitors was studied in the spleens of RBC transfused, lethally irradiated mice injected with bone marrow. Transfusion decreased expansion of the progenitors and slowed their proliferation: the mean cycle time as measured by per cent labelled mitosis (PLM) on the third day after injection of bone marrow was 10.7 hr in transfused as compared to 5.6 hr in non-transfused mice. One injection of five units of erythropoietin on day 2 decreased the mean cycle time to 7.3 hr in transfused mice and increased expansion of the progenitor cells. The effects of erythropoietin on cell proliferation were prompt: a significant increase of incorporation of ³H-TdR into DNA occurred within 2 hr of injection. Erythroblasts were absent from the spleens of transfused, irradiated bone marrow injected mice; however, erythroblasts appeared by 72 hr and 48 hr following EPO injection either 2 days or 5 days after transplantation respectively. Increased uptake of radioactive iron in spleen after erythropoietin injection preceded the appearance of erythroblasts by 2 and 1 days when erythropoietin was injected either 2 or 5 days after marrow transplantation respectively. The increase in cellular proliferation induced by erythropoietin in transfused irradiated mice injected with bone marrow equivalent to 0.35 femoral shaft was manifested as an increase of the total DNA content in the spleen by 119 μg (11.9 × 10⁶ cells) within 48 hr of injection. The cellular increment produced by EPO injection on day 5 to mice given 0.05 femoral shaft consisted mainly of undifferentiated mononuclear cells, most of which were labelled, with erythroblasts comprising only one quarter of the increment. Erythropoietin inactivated by mild acid hydrolysis failed to increase cellular proliferation.     

Plasmodium berghei: the influence of blood volume changes on the malaria-induced anemia in Balb/C mice

Malaria-induced anemia exceeds that attributable to direct parasite destruction of erythrocytes. Since spleen and liver weights increase significantly, hemodilution may account for part of this excessive anemia. To determine the role of hemodilution in the etiology of anemia, vascular volumes were measured with Evans blue and isotope dilution techniques. The Evans blue dilution technique showed that blood volume increased 20%, in infected Balb/C mice. However, when blood volume was measured with Evans blue bound to protein prior to injection or with iodinated albumin and chromium-labelled red blood cells no significant change was detected. Evidently the permeability of the vasculature and/or erythrocytes of infected mice was increased so that injected Evans blue occupied a space larger than the vascular volume before becoming bound to plasma proteins. We conclude that hemodilution is not involved in the excessive anemia of Plasmodium berghei-infected Balb/C mice.     

Kinetics of antigen specific and non-specific polyclonal B-cell responses during lethal Plasmodium yoelii malaria.

In order to study the kinetics and composition of the polyclonal B-cell activation associated to malaria infection, antigen-specific and non-specific B-cell responses were evaluated in the spleens of mice infected with Plasmodium yoelii 17XL or injected with lysed erythrocytes or plasma from P. yoelii infected mice or with P. falciparum culture supernatants. Spleen/body weight ratio, numbers of nucleated spleen cells and Immunoglobulin-containing and Immunoglobulin-secreting cells increased progressively during the course of infection, in parallel to the parasitaemia. A different pattern of kinetics was observed when anti-sheep red blood cell and anti-trinitrophenylated-sheep red blood cell plaque forming cells response were studied: maximum values were observed at early stages of infection, whereas the number of total Immunoglobulin-containing and Immunoglobulin-secreting cells were not yet altered. Conversely, at the end of infection, when these latter values reached their maximum, the anti-sheep red blood cell and anti-trinitrophenylated-sheep red blood cell specific responses were normal or even infranormal. In mice injected with Plasmodium-derived material, a higher increase in antigen-specific PFC was observed, as compared to the increase of Immunoglobulin-containing and Immunoglobulin-secreting cell numbers. This suggested a "preferential" (antigen-plus mitogen-induced) stimulation of antigen-specific cells rather than a generalized non-specific (mitogen-induced) triggering of B-lymphocytes. On the basis of these and previous results, it is suggested that the polyclonal B-cell activation that takes place during the course of infection appears as a result of successive waves of antigen-specific B-cell activation.     

β-Carotene as a high-potency antioxidant to prevent the formation of phospholipid hydroperoxides in red blood cells of mice

In order to investigate the antioxidant effect of β-carotene in vivo, phospholipid hydroperoxides and β-carotene isomers in red blood cells (RBC), plasma and tissue organelles were quantitatively measured after the oral administration of β-carotene (94.8% all-trans-β-carotene) to mice. Three groups of 24 mice each were fed for 1 week on a semisynthetic diet supplemented with either 0.6% or 3.0% β-carotene/diet or maintained on a control (β-carotene-unsupplemented) diet. The RBC phospholipid hydroperoxides showed a significant decrease followed by an increase of β-carotene intakes; i.e., 201, 16 and 4 pmol of phosphatidylcholine hydroperoxide/ml packed RBC, and 108, 22 and 8 pmol of phosphatidylethanolamine hydroperoxide/ml packed RBC, in the mice given the control diet, 0.6% carotene diet and 3.0% carotene diet, respectively. The RBC β-carotene increased from 14 to 43 pmol/ml packed RBC as followed by the increase of β-carotene intakes. Such a potent antioxidant effect of β-carotene as observed in RBC was not confirmed in the plasma, liver or lungs, although their β-carotene contents increased. The β-carotene ingestion increased the all-trans-β-carotene d and retinol contents in RBC, plasma, liver and lungs, but the α-tocopherol content decreased. In the β-carotene-supplemented (6 g and 30 g/kg diet) mice, cis-β-carotene content was relatively higher in the RBC (25–35% of total β-carotene) than that in plasma, liver and lungs. The present findings indicate that not only does β-carotene act as a potent antioxidant in vivo but also its antioxidant effect is very specific in the RBC phospholipid bilayers rather than in the plasma and other tissue organelles.     

Transfer of Borrelia burgdorferi s.s. infection via blood transfusion in a murine model

Without antibiotic treatment, the Lyme-disease-causing bacterium, Borrelia burgdorferi can be cultured from the peripheral blood of human patients nearly 6 wk post-tick bite. To determine if Lyme disease spirochetes can be transmitted from a spirochetemic donor mouse to a naive recipient during blood transfusion, blood taken from immunocompetent infected mice was transfused into either immunodeficient (SCID) mice, inbred immunocompetent animals (C3H/HeJ), or outbred mice. Nine of 19 (47.7%) immunodeficient mice, 7 of 15 (46.8%) inbred immunocompetent mice, and 6 of 10 (60.0%) outbred mice became infected with B. burgdorferi after transfusion. Our results indicate that it is possible to acquire B. burgdoferi infection via transfused blood in a mouse model of Lyme borreliosis.     

Large unit red cell cryopreservation with hydroxyethyl starch.

Full units of red blood cells frozen with 14% hydroxyethyl starch (HES) yield cell recoveries near 97% and saline stabilities greater than 80%. Potassium leaves the cells during the freeze-thaw cycle and increases the extracellular concentration of this ion to near 35 meq/l. Unwashed cells (those with plasma present) and saline washed cells yield similar results. Storage of the thawed red cells at 4 °C for up to 48 hr causes little change in the cells.Examination by electron microscopy of samples from thawed units reveals some red cells with portions of their membrane missing. We believe this represents damage from the freeze-thaw cycle and also that all free supernatant hemoglobin does not arise from completely lysed cells.     

Prophylactic fibrinolysis through selective dissolution of nascent clots by tPA-carrying erythrocytes

A fibrinolytic agent consisting of a tissue-type plasminogen activator (tPA) coupled to the surface of red blood cells (RBCs) can dissolve nascent clots from within the clot, in a Trojan horse-like strategy, while having minimal effects on preexisting hemostatic clots or extravascular tissue. After intravenous injection, the fibrinolytic activity of RBC-tPA persisted in the bloodstream at least tenfold longer than did that of free tPA. In a model of venous thrombosis induced by intravenously injected fibrin microemboli aggregating in pulmonary vasculature, soluble tPA lysed pulmonary clots lodged before but not after tPA injection, whereas the converse was true for RBC-tPA. Free tPA failed to lyse occlusive carotid thrombosis whether injected before or after vascular trauma, whereas RBC-tPA circulating before, but not injected after, thrombus formation restored blood flow. This RBC-based drug delivery strategy alters the fibrinolytic profile of tPA, permitting prophylactic fibrinolysis.     

Successful Implementation of a Packed Red Blood Cell and Fresh Frozen Plasma Transfusion Protocol in the Surgical Intensive Care Unit

BackgroundBlood product transfusions are associated with increased morbidity and mortality. The purpose of this study was to determine if implementation of a restrictive protocol for packed red blood cell (PRBC) and fresh frozen plasma (FFP) transfusion safely reduces blood product utilization and costs in a surgical intensive care unit (SICU).Results829 total patients were included in the analysis (PRE, n=372; POST, n=457). Despite higher mean age (56 vs. 52 years, p=0.01) and APACHE II scores (12.5 vs. 11.2, p=0.006), mean units transfused per patient were lower for both packed red blood cells (0.7 vs. 1.2, p=0.03) and fresh frozen plasma (0.3 vs. 1.2, p=0.007) in the POST compared to the PRE cohort, respectively. There was no difference in inpatient mortality between the PRE and POST cohorts (7.5% vs. 9.2%, p=0.39). There was a decreased risk of urinary tract infections (OR 0.47, 95%CI 0.28-0.80) in the POST cohort after controlling for age, illness severity and amount of blood products transfused.ConclusionsImplementation of a restrictive transfusion protocol can effectively reduce blood product utilization in critically ill surgical patients with no increase in morbidity or mortality.     

The negative regulation of red cell mass by neocytolysis: physiologic and pathophysiologic manifestations.

We have uncovered a physiologic process which negatively regulates the red cell mass by selectively hemolyzing young circulating red blood cells. This allows fine control of the number of circulating red blood cells under steady-state conditions and relatively rapid adaptation to new environments. Neocytolysis is initiated by a fall in erythropoietin levels, so this hormone remains the major regulator of red cell mass both with anemia and with red cell excess. Physiologic situations in which there is increased neocytolysis include the emergence of newborns from the hypoxic uterine environment and the descent of polycythemic high-altitude dwellers to sea level. The process first became apparent while investigating the mechanism of the anemia that invariably occurs after spaceflight. Astronauts experience acute central plethora on entering microgravity resulting in erythropoietin suppression and neocytolysis, but the reduced blood volume and red cell mass become suddenly maladaptive on re-entry to earth's gravity. The pathologic erythropoietin deficiency of renal disease precipitates neocytolysis, which explains the prolongation of red cell survival consistently resulting from erythropoietin therapy and points to optimally efficient erythropoietin dosing schedules. Implications should extend to a number of other physiologic and pathologic situations including polycythemias, hemolytic anemias, 'blood-doping' by elite athletes, and oxygen therapy. It is likely that erythropoietin influences endothelial cells which in turn signal reticuloendothelial phagocytes to destroy or permit the survival of young red cells marked by surface molecules. Ongoing studies to identify the molecular targets and cytokine intermediaries should facilitate detection, dissection and eventual therapeutic manipulation of the process.     

Dietary-induced hypertrophic--hyperplastic obesity in mice

Metabolically intact NMRI mice and genetically obese NZO mice were fed ad lib. either a high-carbohydrate diet (standard) or a high-fat diet for a period of about 11 (NMRI mice) or 38 (NZO mice) wk. In both strains of mice, body weight increased more in the groups fed the high-fat diet. However, caloric intake by NMRI mice fed the high-fat diet was less than that of the controls. In NMRI mice fed the high-fat diet, epididymal and subcutaneous fat cell volumes increased; when these mice were fed the standard diet, only epididymal fat cell volume increased. Epididymal and subcutaneous fat cell numbers increased only in the group fed the high-fat diet. In NMRI mice fed either diet, the postprandial blood glucose was lower in older animals, but plasma insulin remained unchanged. The glucose tolerance deteriorated insignificantly. In NZO mice fed either diet, epididymal fat cell volumes and fat cell numbers increased. In this strain of mice the postprandial blood glucose and plasma insulin exhibited the strain-specific pattern, independent of the diet. In older animals fed either diet the glucose tolerance decreased.     

Cloning and expression pattern of the Xenopus erythropoietin receptor

Cytokine signaling plays an important role in the survival and differentiation of vertebrate hematopoietic cells. In red blood cells, erythropoietin is a key component of the differentiation program and maintains the homeostasis of the erythroid compartment. In the adult, anemia stimulates high levels of circulating erythropoietin that drives erythropoiesis to restore normal levels of red blood cells in circulation. Erythropoietin activates the erythropoietin receptor on immature red blood cell precursors to promote their survival and differentiation. Although extensively studied in mammalian systems, a complete understanding of the function of the erythropoietin receptor during primitive erythropoiesis has been lacking. To address this problem, we have cloned the Xenopus laevis erythropoietin receptor in order to further understand the development of primitive erythropoiesis. The amphibian erythropoietin receptor shares 33% amino acid sequence identity with the mammalian erythropoietin receptors and contains the conserved extracellular ligand binding and fibronectin domains, the WSXWS motif common to cytokine receptors, and several tyrosine phosphorylation sites located on the intracellular domain of the receptor. Expression of the erythropoietin receptor is first detected by in situ hybridization in the ventral blood island during tailbud stages.     

Factors influencing hematopoietic spleen colony formation in irradiated mice. IV. The effect of erythropoietic stimuli

A variety of erythropoietic stimuli influenced the number of endogenous spleen colonies in irradiated mice and the number of transplantable colony forming cells in the spleen and marrow of unirradiated mice. Bleeding was the most effective stimulus. Bleeding before irradiation resulted in a 30-fold increase in endogenous spleen colonies and in increases in spleen weight, spleen iron and iododeoxyuridine uptake and volume of packed red cells ten days after irradiation. Bleeding unirradiated mice produced a 10-fold increase in the number of transplantable colony forming cells in the spleen and a slight decrease in the total number in the humerus. Bleeding before irradiation resulted in a significant reduction in 30-day post irradiation deaths, an effect abolished by splenectomy. Plasma from bled mice induced an increase in endogenous colonies when injected before irradiation into normal mice. Injection of erythropoietin, testosterone or testosterone plus cobalt induced effects which were, in general, qualitatively similar to those of bleeding, although they were less effective quantitatively. Except for a slight effect induced by ten injections of erythropoietin, post-irradiation stimulation in normal mice proved ineffective. Erythropoietin increased colony numbers and spleen iron uptake when given after irradiation to hypertransfused mice. The results of these studies do not support the concept that the colony forming cell and the erythropoietin sensitive cell are separate entities.     

Glycosphingolipids from rabbit aorta, plasma, and red blood cells: effects of high cholesterol-high fat diets on fatty acid distribution and quantity of glycosphingolipids

Four glycosphingolipids were isolated from rabbit aorta, plasma, and red blood cells. They were identified, by thin-layer chromatography and by quantitative analysis of hexose and fatty acid, as cerebroside, diglycosyl ceramide, triglycosyl ceramide, and globoside. The rabbits had been maintained on a normal diet or on one of three high cholesterol diets for 180 days. The quantities of the glycosphingolipids and their fatty acid distributions were determined, and comparisons were made between the control and experimental animals. Aorta and plasma glycosphingolipids were more affected by the high cholesterol diets than were those from red blood cells. The effects on aorta and plasma glycosphingolipids were similar. The amount of cerebroside was increased in aorta and plasma in all animals in the experimental groups. The amount was also increased in red blood cells in rabbits from two of the experimental groups. The average fatty acid chain length was greater in the lipids from the experimental animals than in those from the control animals for all measured glycosphingolipids from aorta. The average chain length was also greater in cerebrosides from the experimental animals from all three tissues. Probably the most notable differences in the experimental animals were the increased 24:1/24:0 ratios and the increased concentrations of 24:2. These increases occurred in nearly all samples from plasma and aorta, but not in red blood cells. There was also an increase of total unsaturated fatty acids in aorta cerebrosides from the experimental animals. Except for the increase in 24:2, lard generally caused more deviation from normal than did cottonseed oil when the level of cholesterol in the diet was 1%.     

Systemic overexpression of IL-10 induces CD4+CD25+ cell populations in vivo and ameliorates type 1 diabetes in nonobese diabetic mice in a dose-dependent fashion

Early systemic treatment of nonobese diabetic mice with high doses of recombinant adeno-associated virus (rAAV) vector expressing murine IL-10 prevents type 1 diabetes. To determine the therapeutic parameters and immunological mechanisms underlying this observation, female nonobese diabetic mice at 4, 8, and 12 wk of age were given a single i.m. injection of rAAV-murine IL-10 (10(4), 10(6), 10(8), and 10(9) infectious units (IU)), rAAV-vector expressing truncated murine IL-10 fragment (10(9) IU), or saline. Transduction with rAAV-IL-10 at 10(9) IU completely prevented diabetes in all animals injected at all time points, including, surprisingly, 12-wk-old animals. Treatment with 10(8) IU provided no protection in the 12-wk-old injected mice, partial prevention in 8-wk-old mice, and full protection in all animals injected at 4 wk of age. All other treatment groups developed diabetes at a similar rate. The rAAV-IL-10 therapy attenuated pancreatic insulitis, decreased MHC II expression on CD11b+ cells, increased the population of CD11b+ cells, and modulated insulin autoantibody production. Interestingly, rAAV-IL-10 therapy dramatically increased the percentage of CD4+CD25+ regulatory T cells. Adoptive transfer studies suggest that rAAV-IL-10 treatment alters the capacity of splenocytes to impart type 1 diabetes in recipient animals. This study indicates the potential for immunomodulatory gene therapy to prevent autoimmune diseases, including type 1 diabetes, and implicates IL-10 as a molecule capable of increasing the percentages of regulatory cells in vivo.     

Platelet-derived toll-like receptor 4 (tlr-4) is sufficient to promote microvascular thrombosis in endotoxemia

Endotoxin (lipopolysaccharide, LPS) produced by gram-negative bacteria initiates a host of pro-inflammatory effects through Toll-like receptor 4 (TLR-4). We reported previously that LPS enhances microvascular thrombosis in cremaster venules of wild-type mice, but had no effect in mice deficient in TLR-4. Since TLR-4 is expressed on various cell types, the cellular origin of TLR-4 responsible for the LPS-enhanced thrombosis remains undetermined. Platelets are known to express functional TLR-4. Platelet-derived TLR-4 has been suggested to mediate various inflammatory responses in endotoxemia, including production of tumor necrosis factor alpha (TNF-α) and interleukin-1 beta (IL-1β), two cytokines reported to enhance microvascular thrombosis. We determined whether platelet-derived TLR-4 was sufficient to mediate the enhanced thrombosis induced by endotoxin and whether these responses were accompanied by systemic increases in TNF-α and IL-1β. We isolated platelets from wild-type mice and transfused them into either of two strains of TLR-4-deficient mice (C57BL/10ScN and B6.B10ScN-TLR-4(lps-del)/Jth). The mice were then injected with LPS or saline, and the kinetics of thrombosis were studied 4 hours later. Transfusion of wild-type platelets restored responsiveness to LPS in TLR-4-deficient mice with regards to microvascular thrombosis but not to plasma levels of TNF-α or IL-1β. The accelerated rates of microvascular thrombosis induced by platelet transfusions were specific to TLR-4, since isolation and transfusion of platelets derived from TLR-4-deficient donors did not restore responsiveness to LPS. These studies demonstrate that platelet-derived TLR-4 is sufficient to promote microvascular thrombosis in endotoxemia, independent of systemic increases in TNF-α or IL-1β.     

Prolonged administration of a dithiol antioxidant protects against ventricular remodeling due to ischemia-reperfusion in mice

The prolonged production of reactive oxygen species due to ischemia-reperfusion (I/R) is a potential cause of the pathological remodeling that frequently precedes heart failure. We tested the ability of a potent dithiol antioxidant, bucillamine, to protect against the long-term consequences of I/R injury in a murine model of myocardial infarction. After transiently occluding the left anterior descending coronary artery for 30 min, saline or bucillamine (10 microg/g body wt) was injected intravenously as a bolus within the first 5 min of reperfusion. The antioxidant treatment continued with daily subcutaneous injections for 4 wk. There were no differences in infarct sizes between bucillamine- and saline-treated animals. After 4 wk of reperfusion, cardiac hypertrophy was decreased by bucillamine treatment (ventricular weight-to-body weight ratios: I/R + saline, 4.5 +/- 0.2 mg/g vs. I/R + bucillamine, 4.2 +/- 0.1 mg/g; means +/- SE; P < 0.05). Additionally, the hearts of bucillamine-treated mice had improved contractile function (echocardiographic measurement of fractional shortening) relative to saline controls: I/R + saline, 32 +/- 3%, versus I/R + bucillamine, 41 +/- 4% (P < 0.05). Finally, I/R-induced injury in the saline-treated mice was accompanied by a fetal pattern of gene expression determined by ribonuclease protection assay that was consistent with pathological cardiac hypertrophy and remodeling [increased atrial natriuretic peptide, beta-myosin heavy chain (MHC), skeletal alpha-actin; decreased sarco(endo)plasmic reticulum Ca2+ ATPase 2a, and alpha-MHC-to-beta-MHC ratio]. These changes in gene expression were significantly attenuated by bucillamine. Therefore, treatment with a dithiol antioxidant for 4 wk after I/R preserved ventricular function and prevented the abnormal pattern of gene expression associated with pathological cardiac remodeling.     

Plasmodium berghei: uptake of clindamycin and its metabolites by mouse erythrocytes with clindamycin-sensitive and clindamycin-resistant parasites.

Tritiated Clindamycin was used to compare the uptake of Clindamycin in plasma and red cells of mice infected with clindamycin-sensitive or clindamycin-resistant Plasmodium berghei and in uninfected mice. Red cells infected with either sensitive or resistant parasites have a higher concentration of [³H]clindamycin and its active metabolites 1 hr after drug administration than uninfected red blood cells. There was no significant difference in uptake of Clindamycin by red blood cells parasitized by sensitive or resistant parasites. Levels of Clindamycin and its metabolites were consistently higher in red cells than in plasma, both in infected and uninfected mice, but the drug was readily removed by washing red cells with phosphate buffered saline in either case. It is concluded that resistance to Clindamycin is not due to an impaired uptake of the drug by the parasitized red cell as has been shown for chloroquine resistance in P. falciparum and P. berghei.     

Antagonism between interleukin 3 and erythropoietin in mice with azidothymidine-induced anemia and in bone marrow endothelial cells

Azidothymidine (AZT)-induced anemia in mice can be reversed by the administration of IGF-IL-3 (fusion protein of insulin-like growth factor II (IGF II) and interleukin 3). Although interleukin 3 (IL-3) and erythropoietin (EPO) are known to act synergistically on hematopoietic cell proliferation in vitro, injection of IGF-IL-3 and EPO in AZT-treated mice resulted in a reduction of red cells and an increase of plasma EPO levels as compared to animals treated with IGF-IL-3 or EPO alone. We tested the hypothesis that the antagonistic effect of IL-3 and EPO on erythroid cells may be mediated by endothelial cells. Bovine liver erythroid cells were cultured on monolayers of human bone marrow endothelial cells previously treated with EPO and IGF-IL-3. There was a significant reduction of thymidine incorporation into both erythroid and endothelial cells in cultures pre-treated with IGF-IL-3 and EPO. Endothelial cell culture supernatants separated by ultrafiltration and ultracentrifugation from cells treated with EPO and IL-3 significantly reduced thymidine incorporation into erythroid cells as compared to identical fractions obtained from the media of cells cultured with EPO alone. These results suggest that endothelial cells treated simultaneously with EPO and IL-3 have a negative effect on erythroid cell production.