American Journal of Hematology 41 :147-150 (1992)
Dyserythropoiesis in Iron-DeficiencyAnemia: Ultrastructural Reassessment Maria Rozman, Teresa Masat, Evarist Feliu, and Ciril Rozman Postgraduate School of Hematology “Farreras Valenti”, Hospital Clinic, University of Barcelona, Barcelona, Spain
Iron deficiency is usually included among the causes of acquired dyserythropoiesis. This concept was derived mainly from light microscopic studies. To reassess such a notion at ultrastructural level, a transmission electron microscopic evaluation of bone marrow was performed in seven patients with iron-deficiency anemia. In contrast to the widely accepted concept, derived from light microscopic studies, only a small proportion (2-4%, not different from controls) of erythroblasts displayed some of the features of nuclear dyserythropoiesis. On the contrary, when examining the cytoplasm, we found a significantly increased number of void ropheocytotic vesicles in the majority of late erythroblasts as compared to controls (P < 0.001). This feature may be considered as an ultrastructural marker of iron deficiency and is consistent with the present knowledge on transferrin-mediateddelivery of iron to the cell. 0 1992 Wiiey-Liss, tnc. Key words: erythropoiesis, transmission electron microscopy, erythroblasts
confirmed iron-deficiency anemia. The most relevant clinical and laboratory data concerning these patients can The term dyserythropoiesis defines a morphological be found in Table I. In addition to anemia, low plasma and functional disorder leading to intramedullarydeath of iron concentration and decreased bone marrow iron, a a proportion of the early and/or maturing erythroblasts secondary thrombocytosis was found in all patients. [l]. This phenomenon may occur in a wide range of These abnormalities subsided with iron therapy in all diseases, congenital and acquired, and is recognized by a cases. Seven bone marrow transplantation donors were number of characteristic cytologic abnormalities. Dysused as controls. Their ages ranged from 8 to 59 years, erythropoietic changes visible by light microscopy are and all had normal hemoglobin concentration and serum better defined by transmission electron microscopy iron level. (TEM) , whereas some dyserythropoietic features can be seen only at the ultrastructural level. TEM Iron deficiency is usually included among the known Aspirated sternal marrow from each patient was placed causes of acquired dyserythropoiesis [ 11. The degree and directly into cold 2% glutaraldehyde in 0.1 M phosphate type of iron deficiency associated dyserythropoietic phe- buffer. Marrow particles were then separated and fixed in nomena have been studied by light microscopy [2,3]. In fresh 2% glutaraldehyde for 2 hr at 4”C, rinsed in 0.1 M addition, some TEM descriptions of erythropoietic cells phosphate buffer with sucrose, postfixed in osmium in iron deficiency have been performed [4-61. However, tetroxide, dehydrated in increasing series of acetone, and systematic ultrastructural evaluation of dyserythropoiesis embedded in an epoxy resin (Durcupan). The sections in this condition has not yet been carried out. The purpose were cut on a Reichert’s ultramicrotome (Ultracut E) with of the present study was to analyze the ultrastructure of a diamond knife, mounted on copper grids, stained with erythropoietic cells in iron-deficiency anemia and eval- uranyl acetate and lead citrate, and examined in a Jeol uate the degree of dyserythropoiesis in this condition by 100-B transmission electron microscope at 80 kV. In TEM. INTRODUCTION
MATERIAL AND METHODS Patients After obtaining informed consent, a sternal bone marrow aspirate was performed in seven patients with o 1992 Wiley-Liss, Inc.
Received for publication April 18, 1990; accepted May 16, 1991. Address reprint requests to Ciril Rozman, MD, Postgraduate School of Hematology, Hospital Clinic, Villarroel 170,08036 Barcelona, Spain.
148
Rozman et al. TABLE 1. Clinical and Laboratory Features of Patients With Iron-Deficiency Anemia
Patient
Age (years)/sex
Source of bleeding
Hb (gldl)
Platelets (XlO’iliter)
Plasma iron (pg/dl)
Bone marrow iron
1
45lF 37lF 50lF 43lF 361F 45lF 63lF
Gastrointestinal Hypermenorrhea Gastrointestinal Hypermenorrhea Hypermenonhea Hypermenorrhea Gastrointestinal
6.8 7 8.4 8.2 8.2 10.4 6.4
585 63 1 623 528 484 440 652
13 11 22 10 23 31 34
Decreased Decreased Decreased Decreased Decreased Decreased Decreased
2 3 4 5 6 7
TABLE II. Frequency of Nuclear Dysewthromietic Features and Number of RoDheocvtotic Vesicles
Iron-deficiency anemia Case No.
Control
Percent of abnormal nuclei
Mean number (SEM) of ropheocytotic vesicles
Percent of abnormal nuclei
Mean number (SEM) of ropheocytotic vesicles
2 2 4 4 3 5 4 3.43
32.72 (2.35) 26.96 (2.51) 21.60 (1.75) 27.28 (2.39) 23.6 (2.38) 18.16 (1.72) 28.32 (2.45) 25.52 (2.22)
1 4 3 3 2 3 4 2.86
8.78 (1.34) 6.44 (0.76) 13.36 (1.03) 17.38 (1.39) 18.12 (1.39) 20.36 (1.91) 5.04 (0.82) 12.78 (1.23)
1 2 3 4 5 6 7 Mean
each case, 100 sections corresponding to red cell precursors were carefully examined with the microscope. This examination was blinded, since it was performed without knowledge of whether the sample belonged to the patients or control group. For the quantitative evaluation of ropheocytotic vesicles, in each patient and control, 25 micrographs corresponding to late erythroblasts were obtained randomly. In each section, the number of ropheocytotic vesicles and invaginations was counted, and the mean number of two blind counts was obtained. Statistical Analysis
Although the proportion is slightly higher in the patients with respect to the control group, the difference is not significant (x2 = 0.37; P = 0.54). Conversely, there is a marked increase of ropheocytotic vesicles in iron deficiency (Figs. 1,2) as compared with the controls. In 175 sections of the former group, a mean of 25.5% (SD 11.89) vesicles were observed compared with a mean of 12.78 (SD 8.49), a difference that is highly significant (t = 11.53, P = 8.9E-16). Moreover, six of seven patients with iron deficiency anemia showed a mean number of ropheocytotic vesicles higher than 20, whereas this occurred in only one of seven control (P = 0.029, Fisher exact test).
Different means were compared by two-sample t statistics with separate variances, which is appropriate whenever the population variances are not assumed to be DISCUSSION equal [7]. The contingency tables were examined by x2 and Fisher exact tests. In this study two important findings were observed: 1) the scarcity of dyserythropoietic features involving nucleus and 2) a markedly increased number of void RESULTS ropheocytotic vesicles in the cytoplasm of late erythroIn Table 11, the findings of nuclear and cytoplasmic blasts. To the present, a systematic ultrastructural evalexamination both in iron-deficiency anemia and in con- uation of dyserythropoiesis in iron-deficiency anemia has trols can be seen. In all cases of both groups, a small not been performed. In a TEM study Liu, Tsui, and percentage of late erythroblasts dysplayed some of the Wang [5] examined the bone marrow of four patients features considered as characteristic of nuclear dyseryth- with iron deficiency and detected several dyserythropoiropoiesis, including binuclearity , enlarged nuclear pores, etic features, such as nucleuskytoplasm maturation asyn“spongy” nuclei, and dilatation of perinuclear cisterna. chronism, widening of perinuclear cisterna, increased
Dyserythropoiesis in Iron-Deficiency Anemia
Fig. 1. A great number of vesicles can be observed in the cytoplasm of a late erythroblast. TEM ~18,000.
phagocytosis of red cell precursors by macrophages, and others; nevertheless, no quantitative analysis was performed. In a light microscopic study of 10 patients with iron deficiency, Hill et al. [2] detected a mean of 20.1% (range 5.146.2%) erythroblasts with nuclear features of dyserythropoiesis, such as karyorrhexis, nuclear budding or fragmentation, multinuclearity, and nuclear bridging. Moreover, in this study, the degree of dyserythropoiesis provided a reliable indication of the severity of iron deficiency, since it was positively correlated with both serum iron and serum iron saturation. These findings were confirmed using the same methodology as Omran and Hussein [3] used in 20 patients, 10 bilharzial and 10 nonbilharzial, with iron deficiency. In our study, the percentage of red cell precursors with dyserythropoietic features involving nucleus was much lower, ranging between 2% and 4%. Moreover, this percentage was not significantly different from that observed in controls. One explanation for this discrepancy may be the difference in severity of anemia in these series of patients. In the Hill et al. [2] series, the degree of dyserythropoiesis was positively correlated not only with serum iron level and plasma iron concentration but also with blood Hb level (r = 0.75, P = 0.013). Moreover, the Hill et al. patients showed a trend to a more severe anemia (seven of 10 had less than 6.8 gr/dl Hb) than patients in the present series (only one of seven had less than 6.8 gr/dl Hb) (P = 0.07, Fisher exact test). On the other hand, it may be speculated that some dyserythropoietic features can be more easily detected by light microscopy than by TEM, since in the latter the cell section might not cross the abnormality. Among the features analyzed by Hill et al. [2], this speculation may be true for the nuclear bridging; however, in our experience [8-151, this is relatively rare in acquired dyserythropoiesis and is more frequent in the congenital
149
Fig. 2. The cytoplasmic vesicles of the erythroblast are void of ferritin particles. TEM ~37,000.
dyserythropoietic anemia type I. The remaining features analyzed by Hill et al. [2], specially karyorrhexis and nuclear budding and/or fragmentation, are more reliably demonstrated by TEM, since artifacts readily appear in marrow smears. This may lead to an overestimation of dyserythropoiesis in light microscopic studies. On the other hand, the lack of nuclear dyserythropoietic features in iron deficiency is not surprising, since DNA is not involved as a primary abnormality in this situation. A small percentage of dyserythropoietic nuclei in controls is consistent with the minor degree of ineffective erythropoiesis existing normally. The most striking finding of our study was the discovery of a markedly increased number of void ropheocytotic vesicles in the cytoplasm of erythroblasts. This feature has been previously mentioned in at least two studies [4,6] and may be considered as an ultrastructural marker of iron deficiency. According to present knowledge, iron uptake by erythroid cells takes place by two mechanisms: 1) a pinocytosis-like process called ropheocytosis and 2) through surface transferrin receptors. It is increasingly evident that both mechanisms are related. Since the study by Morgan and Aplleton [ 161, increasing evidence has accumulated suggesting that the delivery of iron by transferrin to erythroid cells involves an intracelMar cycle. After the interaction between transferrin and membrane receptors, the cell membrane invaginates into a vesicle that enters the cytoplasm. There the iron is liberated from the transferrin-receptor complex [ 171. After the iron has been released, receptors are returned to the cell membrane and transferrin to the plasma. In addition, shedding of transferrin receptor to the plasma takes place, both in soluble and in vesicular form [ 181. On the other hand, it has been shown [19] that in iron deficiency, there is a significant increase in the number of erythroblastic surface transferrin receptors, which may be responsible for the increase in transferrin-mediated iron
150
Rozman et al.
uptake and its incorporation into heme and, thus, the increased number of void ropheocytotic vesicles. ACKNOWLEDGMENTS
The technical expertise of Mrs. Montserrat Fabregues and the secretarial skills of Miss Ma JosC Shnchez are greatly appreciated. This study was supported in part by grant 91/213 from Fondo de Investigaciones Sanitarias de la Seguridad Social. REFERENCES 1. Lewis SM, Verwilghen RL: Dyserythropoiesis: definition, diagnosis
2. 3. 4.
5.
6.
7.
and assessment. In Lewis SM, Verwilghen RL (eds): “Dyserythropoiesis.” London: Academic Press, 1977, p 3. Hill RS, Pettit JE, Tattersall MHN, Kiley N, Lewis SM: Iron deficiency and dyserythropoiesis. Br J Haematol 23507, 1972. Oman SA, Hussein AT: Iron deficiency and dyserythropoiesis in bilharzial and non-bilharzial patients. Egypt J Bilharziosis 5 : 1, 1978. Bessis M, Breton Gorius J: L’Elot erythroblastique et la rhophiocytose dans l’anhmie femprive. Rev Hematol 15:233, 1960. Liu L, Tsui YE, Wang LY: Ultrastructural study of bone marrow cell series of patients with iron-deficiency anaemia. Chung-Hua-NeiKO-Tsa-Chih 25:61 I , 1986. Shu-Ling Q, Chong-li Y, Wen-wei H, Yue-lai S, Xiao-li H: Pathological changes in red cell series. Electron microscopic study on bone marrow cells of some blood disorders. Chin Med J 100:34, 1987. Dixon WJ, Brown MB, Engelman L, Hill MA, Jennrich RI: BMDP Statistical Software Manual. Berkeley: University of California Press, 1988, p 535.
8. Rozman C, Woessner S, Ribas-Mundb M, San Miguel JG, VivesCorrons JL, Vives-Puiggrbs J, Nomdedeu J: Congenital dyserythropoietic anaemia type 11. Clinical and ultrastructural study. Acta Haematol 52:312, 1974. 9. Rozman C, Woessner S: Cytoplasmic connections between erythroblasts in megaloblastic anaemia. Acta Haematol 56: 10, 1976. 10. Rozman C, Feliu E, Grafiena A, Bruguks R, Woessner S, VivesCorrons, JL: Transient dyserythropoiesis in repopulated human bone marrow following transplantation: an ultrastructural study. Br J Haematol 50:63, 1982. 11. Woessner S, Bruguks R, Rozman C: Aspectos ultraestructurales de la diseritropoyesis en las anemias refractarias. Sangre 21:670, 1976. 12. Woessner S, Pardo P, Lafuente R, Feliu E, Vives JL, Sans Sabrafin J: Congenital dyserythropoietic anemia type I. Report of a case. Blut 42:47, 1981. 13. Feliu E, Rozman C: Aspectos ultraestructurales de la dishemopoyesis en 10s sindromes mielodisplasicos adquiridos. Sangre 30:65 1, 1985. 14. Femindez-Rafiada JM, Olmeda F, Escudero A, Feliu E , Rozman C: Congenital dyserythropoietic anaemia type 111 (CDA 111). Haematological and ultrastrutural study. Abstract VI Meeting of the Intemational Society of Haematology , European and African Division, Athens, 1981 p 122. 15. Rozman C, Woessner S, Feliu E, Lafuente R, Berga L1: “Ultraestructura Celular en Hematologia.” Barcelona: Editorial Salvat, 1990. 16. Morgan EH, Aplleton TC: Autoradiographic localization of lZ5Ilabeled transferrin rabbit reticulocytes. Naure 223: 1371, 1979. 17. Karin M, Mintz B: Receptor-mediated endocytosis of transfemn in developmentally totipotent mouse teratocarcinoma stem cells. J Biol Chem 256:3245, 1981. 18. Chitambar CR, Noble NA: Shedding of transfenin receptor (TFR) from reticulocytes during maturation: Soluble TFR is derived from TFR shed in vesicles. Blood 7[Suppl 1]:27a, 1990. 19. Muta K, Nishimura J, Ideguchi H, Umemura T, Ibayashi H: Erythroblasts transferrin receptors and transfenin kinetics in iron deficiency and various anemias. Am J Hematol 25:155, 1987.
Dyserythropoiesis in Iron-DeficiencyAnemia: Ultrastructural Reassessment Maria Rozman, Teresa Masat, Evarist Feliu, and Ciril Rozman Postgraduate School of Hematology “Farreras Valenti”, Hospital Clinic, University of Barcelona, Barcelona, Spain
Iron deficiency is usually included among the causes of acquired dyserythropoiesis. This concept was derived mainly from light microscopic studies. To reassess such a notion at ultrastructural level, a transmission electron microscopic evaluation of bone marrow was performed in seven patients with iron-deficiency anemia. In contrast to the widely accepted concept, derived from light microscopic studies, only a small proportion (2-4%, not different from controls) of erythroblasts displayed some of the features of nuclear dyserythropoiesis. On the contrary, when examining the cytoplasm, we found a significantly increased number of void ropheocytotic vesicles in the majority of late erythroblasts as compared to controls (P < 0.001). This feature may be considered as an ultrastructural marker of iron deficiency and is consistent with the present knowledge on transferrin-mediateddelivery of iron to the cell. 0 1992 Wiiey-Liss, tnc. Key words: erythropoiesis, transmission electron microscopy, erythroblasts
confirmed iron-deficiency anemia. The most relevant clinical and laboratory data concerning these patients can The term dyserythropoiesis defines a morphological be found in Table I. In addition to anemia, low plasma and functional disorder leading to intramedullarydeath of iron concentration and decreased bone marrow iron, a a proportion of the early and/or maturing erythroblasts secondary thrombocytosis was found in all patients. [l]. This phenomenon may occur in a wide range of These abnormalities subsided with iron therapy in all diseases, congenital and acquired, and is recognized by a cases. Seven bone marrow transplantation donors were number of characteristic cytologic abnormalities. Dysused as controls. Their ages ranged from 8 to 59 years, erythropoietic changes visible by light microscopy are and all had normal hemoglobin concentration and serum better defined by transmission electron microscopy iron level. (TEM) , whereas some dyserythropoietic features can be seen only at the ultrastructural level. TEM Iron deficiency is usually included among the known Aspirated sternal marrow from each patient was placed causes of acquired dyserythropoiesis [ 11. The degree and directly into cold 2% glutaraldehyde in 0.1 M phosphate type of iron deficiency associated dyserythropoietic phe- buffer. Marrow particles were then separated and fixed in nomena have been studied by light microscopy [2,3]. In fresh 2% glutaraldehyde for 2 hr at 4”C, rinsed in 0.1 M addition, some TEM descriptions of erythropoietic cells phosphate buffer with sucrose, postfixed in osmium in iron deficiency have been performed [4-61. However, tetroxide, dehydrated in increasing series of acetone, and systematic ultrastructural evaluation of dyserythropoiesis embedded in an epoxy resin (Durcupan). The sections in this condition has not yet been carried out. The purpose were cut on a Reichert’s ultramicrotome (Ultracut E) with of the present study was to analyze the ultrastructure of a diamond knife, mounted on copper grids, stained with erythropoietic cells in iron-deficiency anemia and eval- uranyl acetate and lead citrate, and examined in a Jeol uate the degree of dyserythropoiesis in this condition by 100-B transmission electron microscope at 80 kV. In TEM. INTRODUCTION
MATERIAL AND METHODS Patients After obtaining informed consent, a sternal bone marrow aspirate was performed in seven patients with o 1992 Wiley-Liss, Inc.
Received for publication April 18, 1990; accepted May 16, 1991. Address reprint requests to Ciril Rozman, MD, Postgraduate School of Hematology, Hospital Clinic, Villarroel 170,08036 Barcelona, Spain.
148
Rozman et al. TABLE 1. Clinical and Laboratory Features of Patients With Iron-Deficiency Anemia
Patient
Age (years)/sex
Source of bleeding
Hb (gldl)
Platelets (XlO’iliter)
Plasma iron (pg/dl)
Bone marrow iron
1
45lF 37lF 50lF 43lF 361F 45lF 63lF
Gastrointestinal Hypermenorrhea Gastrointestinal Hypermenorrhea Hypermenonhea Hypermenorrhea Gastrointestinal
6.8 7 8.4 8.2 8.2 10.4 6.4
585 63 1 623 528 484 440 652
13 11 22 10 23 31 34
Decreased Decreased Decreased Decreased Decreased Decreased Decreased
2 3 4 5 6 7
TABLE II. Frequency of Nuclear Dysewthromietic Features and Number of RoDheocvtotic Vesicles
Iron-deficiency anemia Case No.
Control
Percent of abnormal nuclei
Mean number (SEM) of ropheocytotic vesicles
Percent of abnormal nuclei
Mean number (SEM) of ropheocytotic vesicles
2 2 4 4 3 5 4 3.43
32.72 (2.35) 26.96 (2.51) 21.60 (1.75) 27.28 (2.39) 23.6 (2.38) 18.16 (1.72) 28.32 (2.45) 25.52 (2.22)
1 4 3 3 2 3 4 2.86
8.78 (1.34) 6.44 (0.76) 13.36 (1.03) 17.38 (1.39) 18.12 (1.39) 20.36 (1.91) 5.04 (0.82) 12.78 (1.23)
1 2 3 4 5 6 7 Mean
each case, 100 sections corresponding to red cell precursors were carefully examined with the microscope. This examination was blinded, since it was performed without knowledge of whether the sample belonged to the patients or control group. For the quantitative evaluation of ropheocytotic vesicles, in each patient and control, 25 micrographs corresponding to late erythroblasts were obtained randomly. In each section, the number of ropheocytotic vesicles and invaginations was counted, and the mean number of two blind counts was obtained. Statistical Analysis
Although the proportion is slightly higher in the patients with respect to the control group, the difference is not significant (x2 = 0.37; P = 0.54). Conversely, there is a marked increase of ropheocytotic vesicles in iron deficiency (Figs. 1,2) as compared with the controls. In 175 sections of the former group, a mean of 25.5% (SD 11.89) vesicles were observed compared with a mean of 12.78 (SD 8.49), a difference that is highly significant (t = 11.53, P = 8.9E-16). Moreover, six of seven patients with iron deficiency anemia showed a mean number of ropheocytotic vesicles higher than 20, whereas this occurred in only one of seven control (P = 0.029, Fisher exact test).
Different means were compared by two-sample t statistics with separate variances, which is appropriate whenever the population variances are not assumed to be DISCUSSION equal [7]. The contingency tables were examined by x2 and Fisher exact tests. In this study two important findings were observed: 1) the scarcity of dyserythropoietic features involving nucleus and 2) a markedly increased number of void RESULTS ropheocytotic vesicles in the cytoplasm of late erythroIn Table 11, the findings of nuclear and cytoplasmic blasts. To the present, a systematic ultrastructural evalexamination both in iron-deficiency anemia and in con- uation of dyserythropoiesis in iron-deficiency anemia has trols can be seen. In all cases of both groups, a small not been performed. In a TEM study Liu, Tsui, and percentage of late erythroblasts dysplayed some of the Wang [5] examined the bone marrow of four patients features considered as characteristic of nuclear dyseryth- with iron deficiency and detected several dyserythropoiropoiesis, including binuclearity , enlarged nuclear pores, etic features, such as nucleuskytoplasm maturation asyn“spongy” nuclei, and dilatation of perinuclear cisterna. chronism, widening of perinuclear cisterna, increased
Dyserythropoiesis in Iron-Deficiency Anemia
Fig. 1. A great number of vesicles can be observed in the cytoplasm of a late erythroblast. TEM ~18,000.
phagocytosis of red cell precursors by macrophages, and others; nevertheless, no quantitative analysis was performed. In a light microscopic study of 10 patients with iron deficiency, Hill et al. [2] detected a mean of 20.1% (range 5.146.2%) erythroblasts with nuclear features of dyserythropoiesis, such as karyorrhexis, nuclear budding or fragmentation, multinuclearity, and nuclear bridging. Moreover, in this study, the degree of dyserythropoiesis provided a reliable indication of the severity of iron deficiency, since it was positively correlated with both serum iron and serum iron saturation. These findings were confirmed using the same methodology as Omran and Hussein [3] used in 20 patients, 10 bilharzial and 10 nonbilharzial, with iron deficiency. In our study, the percentage of red cell precursors with dyserythropoietic features involving nucleus was much lower, ranging between 2% and 4%. Moreover, this percentage was not significantly different from that observed in controls. One explanation for this discrepancy may be the difference in severity of anemia in these series of patients. In the Hill et al. [2] series, the degree of dyserythropoiesis was positively correlated not only with serum iron level and plasma iron concentration but also with blood Hb level (r = 0.75, P = 0.013). Moreover, the Hill et al. patients showed a trend to a more severe anemia (seven of 10 had less than 6.8 gr/dl Hb) than patients in the present series (only one of seven had less than 6.8 gr/dl Hb) (P = 0.07, Fisher exact test). On the other hand, it may be speculated that some dyserythropoietic features can be more easily detected by light microscopy than by TEM, since in the latter the cell section might not cross the abnormality. Among the features analyzed by Hill et al. [2], this speculation may be true for the nuclear bridging; however, in our experience [8-151, this is relatively rare in acquired dyserythropoiesis and is more frequent in the congenital
149
Fig. 2. The cytoplasmic vesicles of the erythroblast are void of ferritin particles. TEM ~37,000.
dyserythropoietic anemia type I. The remaining features analyzed by Hill et al. [2], specially karyorrhexis and nuclear budding and/or fragmentation, are more reliably demonstrated by TEM, since artifacts readily appear in marrow smears. This may lead to an overestimation of dyserythropoiesis in light microscopic studies. On the other hand, the lack of nuclear dyserythropoietic features in iron deficiency is not surprising, since DNA is not involved as a primary abnormality in this situation. A small percentage of dyserythropoietic nuclei in controls is consistent with the minor degree of ineffective erythropoiesis existing normally. The most striking finding of our study was the discovery of a markedly increased number of void ropheocytotic vesicles in the cytoplasm of erythroblasts. This feature has been previously mentioned in at least two studies [4,6] and may be considered as an ultrastructural marker of iron deficiency. According to present knowledge, iron uptake by erythroid cells takes place by two mechanisms: 1) a pinocytosis-like process called ropheocytosis and 2) through surface transferrin receptors. It is increasingly evident that both mechanisms are related. Since the study by Morgan and Aplleton [ 161, increasing evidence has accumulated suggesting that the delivery of iron by transferrin to erythroid cells involves an intracelMar cycle. After the interaction between transferrin and membrane receptors, the cell membrane invaginates into a vesicle that enters the cytoplasm. There the iron is liberated from the transferrin-receptor complex [ 171. After the iron has been released, receptors are returned to the cell membrane and transferrin to the plasma. In addition, shedding of transferrin receptor to the plasma takes place, both in soluble and in vesicular form [ 181. On the other hand, it has been shown [19] that in iron deficiency, there is a significant increase in the number of erythroblastic surface transferrin receptors, which may be responsible for the increase in transferrin-mediated iron
150
Rozman et al.
uptake and its incorporation into heme and, thus, the increased number of void ropheocytotic vesicles. ACKNOWLEDGMENTS
The technical expertise of Mrs. Montserrat Fabregues and the secretarial skills of Miss Ma JosC Shnchez are greatly appreciated. This study was supported in part by grant 91/213 from Fondo de Investigaciones Sanitarias de la Seguridad Social. REFERENCES 1. Lewis SM, Verwilghen RL: Dyserythropoiesis: definition, diagnosis
2. 3. 4.
5.
6.
7.
and assessment. In Lewis SM, Verwilghen RL (eds): “Dyserythropoiesis.” London: Academic Press, 1977, p 3. Hill RS, Pettit JE, Tattersall MHN, Kiley N, Lewis SM: Iron deficiency and dyserythropoiesis. Br J Haematol 23507, 1972. Oman SA, Hussein AT: Iron deficiency and dyserythropoiesis in bilharzial and non-bilharzial patients. Egypt J Bilharziosis 5 : 1, 1978. Bessis M, Breton Gorius J: L’Elot erythroblastique et la rhophiocytose dans l’anhmie femprive. Rev Hematol 15:233, 1960. Liu L, Tsui YE, Wang LY: Ultrastructural study of bone marrow cell series of patients with iron-deficiency anaemia. Chung-Hua-NeiKO-Tsa-Chih 25:61 I , 1986. Shu-Ling Q, Chong-li Y, Wen-wei H, Yue-lai S, Xiao-li H: Pathological changes in red cell series. Electron microscopic study on bone marrow cells of some blood disorders. Chin Med J 100:34, 1987. Dixon WJ, Brown MB, Engelman L, Hill MA, Jennrich RI: BMDP Statistical Software Manual. Berkeley: University of California Press, 1988, p 535.
8. Rozman C, Woessner S, Ribas-Mundb M, San Miguel JG, VivesCorrons JL, Vives-Puiggrbs J, Nomdedeu J: Congenital dyserythropoietic anaemia type 11. Clinical and ultrastructural study. Acta Haematol 52:312, 1974. 9. Rozman C, Woessner S: Cytoplasmic connections between erythroblasts in megaloblastic anaemia. Acta Haematol 56: 10, 1976. 10. Rozman C, Feliu E, Grafiena A, Bruguks R, Woessner S, VivesCorrons, JL: Transient dyserythropoiesis in repopulated human bone marrow following transplantation: an ultrastructural study. Br J Haematol 50:63, 1982. 11. Woessner S, Bruguks R, Rozman C: Aspectos ultraestructurales de la diseritropoyesis en las anemias refractarias. Sangre 21:670, 1976. 12. Woessner S, Pardo P, Lafuente R, Feliu E, Vives JL, Sans Sabrafin J: Congenital dyserythropoietic anemia type I. Report of a case. Blut 42:47, 1981. 13. Feliu E, Rozman C: Aspectos ultraestructurales de la dishemopoyesis en 10s sindromes mielodisplasicos adquiridos. Sangre 30:65 1, 1985. 14. Femindez-Rafiada JM, Olmeda F, Escudero A, Feliu E , Rozman C: Congenital dyserythropoietic anaemia type 111 (CDA 111). Haematological and ultrastrutural study. Abstract VI Meeting of the Intemational Society of Haematology , European and African Division, Athens, 1981 p 122. 15. Rozman C, Woessner S, Feliu E, Lafuente R, Berga L1: “Ultraestructura Celular en Hematologia.” Barcelona: Editorial Salvat, 1990. 16. Morgan EH, Aplleton TC: Autoradiographic localization of lZ5Ilabeled transferrin rabbit reticulocytes. Naure 223: 1371, 1979. 17. Karin M, Mintz B: Receptor-mediated endocytosis of transfemn in developmentally totipotent mouse teratocarcinoma stem cells. J Biol Chem 256:3245, 1981. 18. Chitambar CR, Noble NA: Shedding of transfenin receptor (TFR) from reticulocytes during maturation: Soluble TFR is derived from TFR shed in vesicles. Blood 7[Suppl 1]:27a, 1990. 19. Muta K, Nishimura J, Ideguchi H, Umemura T, Ibayashi H: Erythroblasts transferrin receptors and transfenin kinetics in iron deficiency and various anemias. Am J Hematol 25:155, 1987.
