Impaired Response of Polycythemic Mice to Erythropoietin Induced by Protein Starvation Imposed After Hormone Administration

Transcription

1 Impaired Response of Polycythemic Mice to Erythropoietin Induced by Protein Starvation Imposed After Hormone Administration A. C. Barceld," R. M. Alippi,= f? Boyer: M. I. Oliveru," S. M. Mide," J. C. E. Bozzini"gb adepartment of Physiology, University of Buenos Aires, School of Dentistry, and bbio Sidus S. A., Buenos Aires, Argentina; and 'Cardeza Foundation, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Key Words. Erythropoietin Starvation Erythropoiesis Dietary protein Nutrition Abstract. The present study was performed to determine the stage of the erythropoietic pathway which is affected by starvation or protein deprivation and whose manifestation is a depressed response to exogenous erythropoietin (EPO). The response to recombinant human EPO was measured in post-hypoxic polycythemic mice by determination of 59Fe uptake into red cells, spleen and femur and/or erythroid colony forming units (CFU-E) and erythroid precursor cell concentrations in femoral marrow. Experimental mice were either starved or fed one of seven different diets whose protein (casein) content ranged from 0 to 20%. All diets were isocaloric. The response of mice maintained on the standard diet (Purina Lab chow) was taken as the normal one. Starvation during the 48-hour period immediately before EPO injection had no effect on the response to the hormone. Starvation, and protein deprivation to a lesser extent, during the 48-hour period following EPO, on the other hand, significantly reduced the response. There was a progressive increase in the response as the casein content of the diet was increased. A normal response was observed when dietary casein concentration was 10%. These findings indicate that nutritional deprivation or dietary protein alterations during the period immediately following EPO injection in polycythemic mice can have detrimental effects on the erythroid response in a model in which nutritional deprivation was relatively short and acute. They also indicate that the subnormal response is not due to a decreased size of the erythroid progenitor Correspondence: Dr. Carlos E. Bozzini, Chtedra de Fisiologia; Facultad de Odontologia, Universidad de Buenos Aires, M.T de Alvear 2142, (1 122) Buenos Aires, Argentina. Received December 2 1, 1992; provisionally accepted January 25, 1993; accepted for publication March 19, OAlphaMed Press /93/$5.00/0 pool available for differentiation but to deficient rates of differentiation of erythropoietic units. Introduction Fried et al. [l] more than 30 years ago observed that the erythroid response of starved rats to erythropoietin (EPO) decreases as the period of starvation is prolonged. Naets and Wittek [2] found that starvation markedly reduces the response to EPO in both normal and polycythemic rats. Since the disappearance rate of EPO from the plasma compartment was normal and no inhibitor was detectable in the plasma of starved rats, the authors concluded that starvation decreases the EPO-sensitive progenitor cell pool. Studies performed in our laboratory confirmed these findings and extended them to the water-deprived, polycythemic rat [3]. Furthermore, they also showed that the ability of rats to respond normally to EPO is dependent on a continuous dietary intake of proteins at levels which are dependent on their biological values [4]. Fondu et al. [5] also reported that the responsiveness of the bone marrow to EPO is impaired in humans with protein-energy malnutrition. In the studies conducted in our laboratory, the response to EPO was measured by either RBC-59Fe uptake or reticulocyte count. Starvation was started one to two days before EPO injection and continued throughout the experimental period. Therefore, starvation could have depressed the response to EPO either by decreasing the size of the erythroid progenitor cell pool available for differentiation or by inducing a deficient rate of differentiation of STEM CELLS 1993;11:

2 297 Response to Erythropoietin During Protein Starvation erythropoietic units and/or their proliferatiodmaturation activities, depending on whether their negative effect was exerted on an erythropoietic stage preceding or following EPO action. The present study was thus undertaken to determine in polycythemic mice the stage of the erythropoietic pathway which is affected by starvation and whose manifestation is a depressed response to exogenous EPO. The results obtained indicate that starvation does not affect the response when it is imposed during the 48-hour period before EPO injection, while it does when imposed during the immediate 48-hour period following EPO injection. Materials and Methods Adult female CF-1 mice were used in groups of 8-10 unless otherwise indicated, and all data refer to mean *SE of a group. Purina chow (23.4% protein, energy value 4.4 calories/g) was considered as standard diet. Mice were either starved or fed freely one of eight diets. A protein-free diet contained 5.0% oil, vitamins and minerals according to Harper [6] and 89.6% dextrin. Casein was used at seven different levels of concentration: 1, 2, 3, 4, 5, 10 and 20% in the diet. All the diets were isocaloric, and casein was included in the protein-free diet by substituting an equivalent amount of dextrin. The effect of starvation and dietary protein levels on the response to recombinant human EPO (rhu-epo) was investigated in mice made polycythemic by intermittent exposure to hypobaria (PH mice). Mice were exposed to 456 hpa during 18 hlday for three weeks. EPO was injected S.C. on the fourth day after the end of the three-week conditioning period. Radioiron (0.2 pci 59Fe citrate) was injected in the tail vein on the sixth day. Incorporation into red cells was measured 72 h later. Blood was obtained by cardiac puncture, and incorporation was calculated estimating the blood volume to be 8% of body weight. Radioiron incorporation into spleen and femur was determined 6 h after injection. For erythroid colony studies, marrow cells were flushed out of the femur of mice using a 1 ml syringe fitted with a 23-gauge needle. A single cell suspension was prepared by aspirating back and forth through the needle. Erythroid colony forming units (CFU-E) were cultured in methylcellulose with 500 mu EPO by the procedure of Iscove and Sieber [7]. Colonies of more than eight cells were counted after 48 h of incubation. Brush preparations of femoral bone marrow from polycythemic mice at various intervals following EPO injection were prepared and stained with Wright s and Giemsa stains. Erythroid precursors were classified as proerythroblasts and early and late erythroblasts on the basis of their size and stage of maturation. Highly purified rhu-epo was produced in a baby hamster kidney transfected cell line (BHK cells) according to the method of Powell et al. [8]. Data were analyzed by one-way analysis of variance, followed by Newman-Keuls test to compare multiple means. p < 0.05 was considered significant. Results To determine the effect of a 48-hour period of starvation imposed before or after rhu-epo injection on the erythroid response of polycythemic mice, four groups of ten polycythemic mice each were established as follows: 1) Group C = control, noninjected, freely fed mice; 2) Group C-E = control, EPO-injected, freely fed mice; 3) Group A-E = EPO-injected mice which were starved during the 48-hour period preceding EPO injection; and 4) Group E-A = EPO-injected mice which were starved during the 48-hour period following EPO injection. All EPO-injected mice received 1.O U rhu-epo. Average responses to EPO are shown in Figure 1. Radioiron uptake was significantly higher (p c 0.001) in the three groups receiving EPO than in the noninjected, control group irrespective of whether or not they were starved. However, while starvation imposed on mice during the 48-hour period preceding EPO injection did not affect the response (C-E versus A-E, p > 0.05), response was significantly depressed when starvation was imposed during the 48-hour period that immediately followed EPO injection (C-E versus E-A, p c 0.001). Because of the possibility of a delayed response to EPO induced by starvation imposed after hormone injection and in order to determine the optimal time for administration of 59Fe in relation to injection of EPO, 112 polycythemic mice were each injected with 1.0 U of rhu-epo

3 Barcel6 et al I 8- p 10 u m = 5 C C-E - A-E E-A 7- i :: 1 4- a Fig. 1. Effect of a 48-hour period of starvation imposed before or after rhu-epo injection on RBC-59Fe uptake of polycythemic mice: Each bar represents the average of 10 mice. Lines on top of bars indicate 1 SE. C = control, noninjected mice. C-E = Control, EPO-injected, freely fed mice; A-E = EPO-injected mice which were starved during the 48-hour period preceding EPO injection; E-A = EPO-injected mice which were starved during the 48-hour period following EPO injection. four days after removal from the reduced pressure chamber. They were then divided into two equal groups. One group was fed a diet containing 20% casein (control group). The remaining one received a 0% casein diet (experimental group). Both diets were freely offered to animals for the time elapsed between EPO and 59Fe injections. Radioiron was injected i.v. to different groups after various intervals (24 h, 48 h and 72 h), and the 59Fe incorporation to red blood cells was measured 72 h later. Radioiron uptake by both spleen and right femur was determined 6 h after 5pFe injection. The curves in Figure 2 show that maximum values of 5QFe incorporation were obtained in both control and experimental groups when the interval between injection of EPO and injection of 59Fe was 48 h. However, the maximum values for mice fed a protein-free diet after EPO injection were significantly lower (p c 0.05) than those of mice fed a 20% casein-diet. To establish the dose-response relationship for rhu-epo in mice fed diets containing 0% or 20% casein during the 48-hour period following hormone injection, four groups of 20 PH mice each were established. They were injected S.C. with 100,200,400 or 800 mu rhu-epo on the fourth post-exposure day. Half of each group Fig. 2. Time-course of S9Fe incorporation into red cells, spleen and femur after a single EPO injection in polycythemic mice. Injection of EPO given S.C. at To. Radioiron given i.v. on the days after EPO indicated in the abcissa. Radioiron uptake by either red cells or spleen and femur was determined 72 h and 6 h later, respectively. Each point represents the average of eight mice. Solid lines correspond to mice fed a 20% casein diet following EPO. Broken lines correspond to mice fed a 0% casein diet following EPO. Squares, triangles and asterisks indicate uptake by red cells, spleen and femur, respectively. was fed a diet containing 0% protein during the 48 h following EPO injection until a tracer dose of 59Fe was injected. The remaining half was fed a diet containing 20% protein as casein for the same time. A group of 10 mice not receiving rhu-epo was considered as noninjected control. The dose-response relationships are shown in Figure 3. The lower response to EPO in the protein-starved group was not just the result of a right shift of the regression line between response and dose. In fact, the parameters of the curve describing the regression between response and dose in mice fed a 20% protein diet was y = ~ and y = ~ for mice fed a protein-free diet. Both the slope and the intercept values of these two lines were significantly different. The responses of protein-starved mice to 100, 200, 400 and 800 mu rhu-epo were, respectively, 57%, 66%, 46% and 52% of the responses of protein-fed mice. Thus, the average response to EPO of experimental mice was calculated to be 55.3 & 4.2% of the response of control mice (p < 0.01). In order to determine the effect of the protein content of the diet given to polycythemic

4 299 Response to Erythropoietin During Protein Starvation z m 14- g 12- a 8 lo- P *- r8- E 2-0' I 100 ZM) rhu-epo (mv) Fig. 3. Dose-response relationship for rhu-epo in polycythemic mice fed a 20% casein diet (solid squares) or a 0% casein diet (open squares) during the 48-hour period following hormone injection. Both the slope and the intercept values of the two curves were significantly different (see text for details). mice during the 48-hour period after EPO injection on the response to the hormone, nine groups of 10 PH mice each were formed. Six of them were fed a diet that was either protein-free (P-0) or contained 1%, 2%, 5%. 10% or 20% (P-1 to P-20) casein for the 48-hour period between -n a. I I I rhu-epo and 59Fe injections. One group was starved (S), and the remaining two were fed a standard laboratory diet (C-1 and C-2). Average responses to 1.0 U rhu-epo are shown in Figure 4. The response of starved PH mice to rhu-epo was significantly c 0.001) than that of mice fed a protein-free diet. There was a progressive increase in the response as the casein content of the diet increased from 0 to 10%. No further changes in the response were observed by increasing casein content up to 20%. Suppression of CFU-E when mice are made polycythemic has been well documented, as has CFU-E elevation 24 h after injection of EPO [9]. Thus, the present experiment was performed to determine the effect of feeding a protein-free diet on the changes in femoral marrow CFU-E content in response to EPO in polycythemic mice. Figure 5 shows that, compared with normocythemic animals, femoral CFU-E number was drastically reduced in polycythemic mice four days after removal from the altitude chamber. In mice fed the 20% casein diet, a prompt increase in CFU-E to a value 2.4 times greater than the normocythemic value occurred 24 h after EPO injection. This value was significantly (p c 0.001) higher than an increase to 1.06 times the normocythemic value that occurred when EPO was injected into polycythemic mice which Fig. 4. Erythropoietic response of polycythemic mice to I.O unit rhu-epo measured as RBCS9Fe uptake, as function of the casein concentration in the diet. Diets were given during the 48-hour period following hormone injection. Values are mean kse of 10 mice in each group. Different letters on top of bars indicate significant differences. C- 1 = Control, noninjected mice; C-2, S and P-0 to P-20 = rhu-epo-injected (see text for details). Fig. 5. CFU-E content in femoral marrow in response to rhu-epo in polycythemic mice placed on a protein-free diet. Values are mean *SE of five animals. N = normocythemic mice; P = polycythemic mice fed a 20% casein diet (P-20) or a protein-free diet (P-0) during the 24-hour period following injection of 1.0 U rhu-epo. P = noninjected mice on the fifth post-hypoxic day.

5 Barcel6 et al. 300 were fed a protein-free diet for the 24 h following EPO injection. To study the effect of protein starvation on various maturation stages of a cohort of erythroid cells induced in polycythemic mice by single injection of rhu-epo, 16 post-hypoxic polycythemic mice were injected i.v. with 1.0 U of rhu-epo on day 0 (fourth post-hypoxic day). Half of the animals were fed a 20% casein diet, and half of them were put on a 0% casein diet. Four mice of each group were subsequently killed at the intervals shown in the abscissa of Figure 6, and brush preparations of femoral marrow done. Erythroid precursor cells were classified as proerythroblasts and early and late erythroblasts. It was observed that while erythropoiesis was suppressed in the saline-injected control animals, a wave of erythropoiesis passed through the marrow of the mice injected with EPO. The proerythroblasts formed in response to the hormone passed during the following h through the various stages of erythroid development. At every stage, erythroid cells were significantly lower (p < 0.05) in protein-starved than in control mice. Discussion The data reported here indicate that nutritional deprivation or dietary protein alterations during the period immediately following EPO injection in polycythemic mice can have detrimental effects on the erythroid response. This model of nutritional deprivation differs from previous models in that protein (and in several groups, calories) were markedly and acutely restricted and that the period of nutritional deprivation was relatively brief (two days). The data really indicate that acute starvation, and protein deprivation to a lesser extent, effectively inhibit red cell production in post-hypoxic polycythemic mice in which initiation of erythropoiesis was induced by exogenous EPO, thus confirming previous results obtained in rats [2,4]. Unexpected observations were that both feeding conditions depressed the response to EPO when applied during the 48-how post-injection period and were ineffective when applied during the 48-hour pre-injection one. The response to EPO in mice fed an isocaloric, protein-free diet during the 48-hour Fig. 6. Erythroid response to a single injection of EPO in polycythemic mice. Injection of EPO (1.O U) given i.v. on the fourth post-hypoxic day. Pro = proerythroblasts, EE = early erythroblasts, LE = late erythroblasts. Dotted bars = mice fed a 20% casein diet; dashed bars = mice fed a 0% casein diet. Mean ise of four mice. post-injection period, although significantly lower than that found in mice fed a 20% casein-diet, was significantly higher than that of starved mice. This finding suggests that the inhibitory effect of acute starvation on the erythroid responsiveness to EPO is not solely the result of the lack of dietary protein but also the consequence of caloric deprivation. The depressed response to EPO seen in mice fed the protein-free diet allowed us the election of this model instead of that of the starved mouse for further experimentation because of the maintenance of better physical conditions in the protein-starved animals than in the starved ones. The depressed response to EPO found in the mouse fed a protein-free diet was not the result of a delayed response, which could have altered the time of radioiron injection in relation to that of EPO injection to obtain a maximum recording of the response. The time-course of 59Fe incorporation after a single EPO injection was similar in protein-fed and protein-starved mice, although the maximum value for the latter represented only 53.3% of the normal value. In the presence of similar plasma iron levels and assuming similar ratios of iron uptakehemoglobin synthesis in hemoglobin-synthesizing cells, the lower s9fe uptake by femur and spleen would indicate fewer erythroid precursor cells formed in response to EPO in the protein-starved group.

6 30 1 Response to Erythropoietin During Protein Starvation Post-hypoxic polycythemic mice show marked reductions in CFU-E and differentiated erythroid precursor cells on the fourth-day post-hypoxic period [9, 101. After the injection of EPO into these animals, CFU-E and proerythroblast levels increase significantly, peaking 24 h later and returning to initial levels by 48 h [9, 101. Basophilic normoblasts and polychromatic and orthochromatic normoblasts increase significantly 48 h after injection [l 11. In the present experiments, the value for CFU-E in femoral marrow obtained 24 h after EPO injection was significantly lower in the mice fed the protein-free diet after hormone injection than in those fed the 20% protein diet. It thus appears that the CFU-E were less responsive to stimulation by EPO when protein was absent in the diet after EPO injection. The same type of abnormality was noted in the differentiated erythroid precursors. After injection of EPO, the increase in these cells was significantly lower in the protein-starved mice as compared with the normal-fed ones. EPO injection increases erythropoiesis in the polycythemic mouse by inducing proliferation and differentiation of late erythroid progenitor cells into proerythroblasts. Each proerythroblast then gives rise to an erythropoietic unit by the combined effects of proliferation and maturation. The erythropoietic unit will eventually disappear when all precursor cells enter the blood as functioning red cells. On this basis, the results of the different experiments reported here would indicate that: 1) the erythroid progenitor cell pool available for differentiation was not significantly altered by a short (two-day) dietary protein restriction, as evidenced by the normal response to exogenous EPO seen in the polycythemic mice when protein starvation was applied during the 48-hour period prior to EPO injection; 2) dietary protein restriction induced a lower than normal rate of formation of erythropoietic units, as evidenced by a depressed response to EPO observed at the CFU-E and proerythroblast levels when protein starvation was imposed during the 24- or 48-hour period following hormone injection. The hemoglobinization of the erythropoietic units in this situation, however, appeared normal, as evidenced by the normal time-course of radioiron uptake after EPO injection; and 3) a 10% protein content in the diet appears as the lower protein level that allows a normal erythroid responsiveness to EPO when dietary protein is casein, a high biological value protein. The role of protein and other nutritional factors in the regulation of erythropoiesis has been extensively studied [12]. However, the inhibitory effect of protein deprivation on the erythroid response to EPO when imposed after hormone injection has not been previously shown. This observation deserves further study to understand its operating mechanism. References 1 Fried W, Plzak LF, Jacobson LO, Goldwasser E. Studies on erythropoiesis Factors controlling erythropoietin production. Proc SOC Exp Biol Med 1957;94: Naets JP, Wittek M. Effect of starvation on the response to erythropoietin in the rat. Acta Haemat 1974;52: Giglio JM, Alippi RM, Barceld AC, Bozzini CE. Mechanism of the decreased erythropoiesis in the water deprived rat. Brit J Haemat 1979;42: Alippi RM, Giglio JM, Barceld AC, Bozzini CE, Farina R, Rio ME. Influence of dietary protein concentration and quality on response to erythropoietin in the polycythemic rat. Brit J Haemat 1979;43: Fondu P, Haga P, Halvorsen S. The regulation of erythropoiesis in protein-energy malnutrition. Brit J Haemat 1978;38: Harper AE. Amino acid balance and imbalance. I. Dietary level of proteins and amino acids imbalance. J Nutr 1959;68: Iscove NN, Sieber F. Erythroid progenitors in mouse bone marrow detected by macroscopic colony formation in culture. Exp Hematol 1975;3: Powell VS, Berkner KL, Lebo RV, Adamson JW. Human erythropoietin high level expression in stably transfected mammalian cells and chromosome localization. Proc Natl Acad Sci USA 1986;83: Gregory CJ, Tepperman AD, McCulloch EA, Till JE. Erythropoietic progenitors capable of colony formation in culture: response of normal and genetically anemic W m mice to manipulations of the erythron. J Cell Physiol 1974;84: Udupa KB, Reissmann KR. In vivo erythropoietin requirements of regenerating erythroid

7 Barceld et al. 302 progenitors (BFU-E, CFU-E) in bone marrow of mice. Blood 1979;53: Filrnanowicz E, Gurney CW. Studies on erythropoiesis. XVI. Response to a single dose of erythropoietin in polycythemic mouse. J Lab Clin Med 1961;57: Fried W, Anagnostou A. The role of protein and other nutritional factors in the reeulation - of ervthropoiesis. In: Dunn CDR, ed. Current Concepts in Erythropoiesis. John Wiley & Sons Ltd., 1983~