Favism: Effect of Divicine on Rat Erythrocyte Sulfhydryl Status, Hexose Monophosphate Shunt Activity, Morphology, and Membrane Skeletal Proteins

David C. McMillan1,, Laura J. C. Bolchoz and David J. Jollow

Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medical University of South Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425

Received April 3, 2001; accepted May 17, 2001


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Favism is an acute anemic crisis that can occur in susceptible individuals who ingest fava beans. The fava bean pyrimidine aglycone divicine has been identified as a hemotoxic constituent; however, its mechanism of toxicity remains unknown. We have shown recently that divicine can induce a favic-like response in rats and that divicine is directly toxic to rat red cells. In the present study, we have examined the effect of hemotoxic concentrations of divicine on rat erythrocyte sulfhydryl status, hexose monophosphate (HMP) shunt activity, morphology, and membrane skeletal proteins. In vitro exposure of rat red cells to divicine markedly stimulated HMP shunt activity and resulted in depletion of reduced glutathione with concomitant formation of glutathione-protein mixed-disulfides. Examination of divicine-treated red cells by scanning electron microscopy revealed transformation of the cells to an extreme echinocytic morphology. SDS-PAGE and immunoblotting analysis of the membrane skeletal proteins indicated that hemotoxicity was associated with the apparent loss of skeletal protein bands 2.1, 3, and 4.2, and the appearance of membrane-bound hemoglobin. Treatment of divicine-damaged red cells with dithiothreitol reversed the protein changes, which indicated that the observed alterations were due primarily to the formation of disulfide-linked hemoglobin-skeletal protein adducts. The data suggest that oxidative modification of hemoglobin and membrane skeletal proteins by divicine may be key events in the mechanism underlying favism.

Key Words: hemolytic anemia; favism; divicine; rat; glucose-6-phosphate dehydrogenase deficiency; erythrocytes; glutathione.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Favism is a life-threatening hemolytic crisis that can result from the ingestion of fava beans (Vicia faba) by susceptible individuals who have low-activity variants of erythrocytic glucose 6-phosphate dehydrogenase (G6PD). Since G6PD regulates the production of NADPH in the red cell by the hexose monophosphate (HMP) shunt, G6PD-deficient individuals have a decreased capacity to maintain sufficient levels of NADPH in response to an oxidative stress (Beutler, 1978Go). Early studies identified 2 components of fava beans, divicine and isouramil, as the probable causative agents based on their ability to deplete reduced glutathione (GSH) in isolated suspensions of human G6PD-deficient red cells (Mager et al., 1965Go). Divicine and isouramil are not present in fava beans per se, but are aglycones of the biologically inactive fava bean ß-glucosides, vicine, and convicine, respectively.

The mechanism underlying the onset of favism is not yet understood. However, it has been postulated that both pyrimidine aglycones, liberated upon digestion of their parent glucosides (Hegazy and Marquardt, 1984Go), are absorbed into the blood and induce oxidative damage within erythrocytes as a consequence of their redox activity (Chevion et al., 1982Go; Winterbourn and Munday, 1990Go). Studies of red cells withdrawn from patients during early and late stages of favic crises have indicated that GSH depletion and HMP shunt stimulation are key events that precede red cell loss (Gaetani et al., 1979Go), and these biochemical responses are reported to be accompanied by alterations in the morphological appearance of the erythrocytes when viewed under light microscopy (Fischer et al., 1985Go).

We have shown recently that divicine can induce a favic-like response when administered to G6PD-normal rats, and that divicine is directly hemotoxic to rat red cells (McMillan and Jollow, 1999Go). That is, when incubated with 51Cr-labeled red cells in vitro, divicine induces alterations in the cells such that when they are returned to the circulation of isologous rats, the treated cells are rapidly sequestered into the spleen. In view of the importance of sulfhydryl status in the progression of favism and the role of membrane skeletal proteins in the maintenance of normal red cell shape (Mohandas et al., 1983Go), the present studies were undertaken to examine the sulfhydryl status, HMP shunt activity, morphology, and membrane skeletal proteins of rat erythrocytes exposed in vitro to hemotoxic concentrations of divicine. We report that divicine rapidly stimulates HMP shunt activity and depletes GSH in rat erythrocytes with concomitant formation of glutathione-protein mixed-disulfides. These biochemical alterations are associated with profound membrane skeletal protein damage and transformation of the cells to an extreme echinocytic morphology.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals and materials.
Divicine (2,6-diamino-5-hydroxy-4[3H]-pyrimidinone) hemisulfate was synthesized as described previously (McMillan et al., 1993Go). 14C-1-Glucose and Na251CrO4 in sterile saline (1 mCi/ml, pH 8) was purchased from New England Nuclear (Billerica, MA). Rabbit anti-rat hemoglobin, HRP-conjugated goat anti-rabbit IgG, and dithiothreitol (DTT) were purchased from Sigma (St. Louis, MO). All other chemicals and reagents were of the best commercially available grade.

Animals.
Male Sprague-Dawley rats (130–150 g) were purchased from Harlan Laboratories (Indianapolis, IN), and maintained on food and water ad libitum. Animals were acclimated for 1 week to a 12-h light-dark cycle prior to their use.

Red cell incubation conditions.
Red cells were collected from anesthetized rats into heparinized tubes and washed in isotonic phosphate-buffered saline (pH 7.4) supplemented with 10 mM D-glucose (PBSG). The red cells were resuspended to 40% hematocrit and used the same day they were collected. After a 5-min preincubation at 37°C, various concentrations of divicine dissolved in PBSG were added to red cell suspensions (2 ml) and allowed to incubate aerobically for up to 2 h at 37°C.

Analysis of red cell HMP shunt activity and sulfhydryl status.
The evolution of 14CO2 from 14C-1-glucose was used to estimate HMP shunt activity in rat red cell suspensions as described previously (Grossman et al., 1995Go). For determination of sulfhydryl status, aliquots (200 µl) of the red cell incubation mixtures were removed at various intervals after the addition of divicine and assayed for GSH, GSSG, and protein-SSG concentration by HPLC with electrochemical detection as described previously (Jensen et al., 1986Go). The amount of sulfhydryl present in the sample was estimated by comparison of peak height to standards prepared identically to the samples.

Morphological examination of red cells.
Control and divicine-treated red cells were prepared for scanning electron microscopy in a manner similar to that described by Dewar (1982). After incubation, the red cells were washed once in PBSG, and fixed and dehydrated as described previously (Grossman et al., 1992Go). The cells were then cast with a thin coating of carbon and gold, and examined in a JEOL JSM-5410LV scanning electron microscope operating at 10 kV accelerating voltage.

Electrophoretic analysis of membrane skeletal proteins.
Red cell ghosts (unsealed membrane vesicles) were prepared from control and divicine-treated red cells by hypotonic lysis as described previously (Grossman et al., 1992Go). The ghosts were washed repetitively with buffer and solubilized with SDS. Electrophoretic and immunoblotting analysis of the solubilized membrane ghosts was carried out as described previously (McMillan et al., 1995Go). The proteins were resolved on nonreducing, continuous gels (5% monomer and 1.5% bis-acrylamide crosslinker), and the skeletal protein bands were identified according to their migration distance (Fairbanks et al., 1971Go).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Effect of Divicine on Rat Erythrocyte Sulfhydryl Status and HMP Shunt Activity
Previous studies have shown that survival of rat 51Cr-labeled red cells in vivo is reduced dramatically after in vitro exposure of the radiolabeled cells to divicine (McMillan and Jollow, 1999Go). The response to divicine is characterized by a narrow concentration range (1 to 2 mM), with a TC50 of about 1.5 mM. Since it is known that GSH depletion and HMP shunt stimulation occur in red cells of G6PD-deficient patients undergoing a favic crisis (Gaetani et al., 1979Go), it was of interest to examine these parameters in rat red cell suspensions treated with a hemotoxic concentration of divicine. Thus, divicine (1.5 mM) was added to a 40% suspension of rat red cells, and aliquots were removed at various time points and assayed for GSH, GSSG, and protein-SSG. Exposure of rat red cells to divicine resulted in a very rapid decline in cellular GSH (Fig. 1Go), which reached a nadir within 10 min. The loss of GSH was accompanied by a marked increase in protein-SSG formation; GSSG content was very low and remained constant throughout the incubation period. The sum of the sulfhydryl species was constant during the time period of incubation, indicating that the disappearance of GSH was due primarily to the formation of mixed disulfides with soluble protein.



View larger version (22K):
[in this window]
[in a new window]
 
FIG. 1. Effect of divicine on rat erythrocyte sulfhydryl status. Rat red cells were incubated at 37°C in PBSG containing divicine (1.5 mM). At the indicated time points, aliquots were withdrawn and assayed for GSH (open square), GSSG (filled circle), and glutathione-protein mixed-disulfides (filled triangle); sum of the glutathione species (open diamond). Data points are means of duplicate determinations.

 
Addition of divicine to red cells (40%) also induced a significant stimulation of HMP shunt activity, as measured by the accumulation of 14CO2 (derived from 14C-1-glucose) over a 1-h incubation period. Stimulation of HMP shunt activity by divicine was concentration dependent (Fig. 2Go). The stimulatory effect appeared maximal at about 0.75 mM divicine (corresponding to about an 10-fold increase in activity), and then declined modestly at divicine concentrations above 1 mM.



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 2. Effect of divicine on rat erythrocyte HMP shunt activity. Washed red cells were suspended in PBSG and placed in flasks containing a center well. 14C-1-Glucose was added and the flasks sealed. After 1 h of incubation at 37°C, the reaction was terminated by injection with TCA, and the released 14CO2 collected in hyamine hydroxide and counted. The values are means ± SD (n = 3). *Significantly different from control (p < 0.05).

 
Effect of Divicine on Rat Erythrocyte Morphology
Severe alterations in red-cell morphology have been described during the clinical course of favism (Weed and Reed, 1966Go). To investigate whether morphological transformation of rat erythrocytes had occurred due to divicine exposure, 10 µl aliquots of control and divicine-treated erythrocyte suspensions were removed following a 2-h incubation and prepared for scanning electron microscopy. As shown in Fig. 3AGo, control red cells incubated for 2 h at 37°C exhibited the biconcave appearance of normal discocytes. Echinocytic cells were observed on occasion, but represented < 3% of the total cells. In contrast, more than 50% of the red cells exposed to 2 mM divicine for 2 h at 37°C had lost their discocytic morphology and exhibited moderate to severe degrees of echinocytosis (Fig. 3BGo). The cells were characterized by several moderately-sized protuberances that were asymmetrically distributed on the cell surface. These transformations could also be observed to a lesser degree at lower concentrations of divicine (data not shown), and in Giemsa-stained smears viewed at the light microscopic level, which indicated that the changes observed in divicine-treated red cells were not an artifact of the process of preparing the cells for scanning electron microscopy. Crossbonded erythrocytes, such as those reported to occur in divicine-treated rat red cells exposed to hypertonic media (Fischer et al., 1985Go), were notably absent, as were Type III echinocytes and spheroechinocytes, such as those generated by exposure to the classical hemolytic agent, phenylhydrazine.



View larger version (138K):
[in this window]
[in a new window]
 
FIG. 3. Scanning electron micrograph of rat erythrocytes incubated for 2 h at 37°C in (A) vehicle (PBSG) alone, or (B) in PBSG containing 2 mM divicine. Magnificationx4000.

 
Effect of Divicine on Rat Erythrocyte Membrane Skeletal Proteins
Membrane skeletal proteins from control and divicine-treated red cells were separated on SDS-PAGE gels and either stained with Coomassie blue or transferred onto nitrocellulose membranes and immunostained with rabbit polyclonal antibodies to rat hemoglobin. As shown in the Coomassie blue-stained gel (Fig. 4AGo), divicine induced a concentration-dependent alteration in the protein electrophoretic pattern of rat red cells. As compared with the PBSG control (lane 1), red cells exposed to increasing concentrations of divicine (lanes 2–6) showed decreases in bands 1 (spectrin), 2.1 (ankyrin), and 4.2, and an increase in the amount of high molecular weight protein (HMWP) aggregates that did not enter the gel. In addition, divicine treatment induced the appearance of membrane-bound hemoglobin monomer, which can be seen as a new protein band at the lower molecular weight end of the gel.



View larger version (62K):
[in this window]
[in a new window]
 
FIG. 4. Effect of divicine on rat erythrocyte membrane skeletal proteins. Rat erythrocytes were incubated for 1 h at 37°C in PBSG containing the vehicle (lane 1), 0.8 mM (lane 2), 1.0 mM (lane 3), 1.3 mM (lane 4), 1.5 mM (lane 5), and 2.0 mM (lane 6) divicine. The cells were washed, and membrane ghosts were prepared and washed exhaustively to remove unbound hemoglobin. The ghost proteins (32 µg) were then solubilized in SDS and subjected to PAGE. (A) Coomassie blue-stained gel. (B) Immunoblot stained with rabbit polyclonal antibodies to rat hemoglobin.

 
Immunoblot analysis of the resolved skeletal proteins (Fig. 4BGo) revealed the concentration-dependent association of hemoglobin with the membrane. With the exception of a residual amount of hemoglobin monomer (16 kDa), no antibody staining was observed on blotted protein from control cells (lane 1). In contrast, protein bands consistent with formation of membrane-bound hemoglobin monomer (16 kDa), dimer (32 kDa), and tetramer (54 kDa) (Grossman et al., 1992Go) were observed in red cells treated with increasing concentrations of divicine (lanes 2–6). In addition, diffuse hemoglobin antibody staining was observed in the spectrin and band 3 regions.

In view of the observed capacity of divicine to induce the formation of disulfide bonds between GSH and sulfhydryl groups of soluble protein in the red cell (Fig. 1Go), we examined the involvement of intermolecular disulfide bond formation in the alteration of the membrane skeletal proteins. Red cells were incubated with divicine (1.5 mM) for 1 h and then washed and exposed to DTT (5 mM) for an additional 1 h prior to lysis and solubilization. As shown in Fig. 5AGo, exposure of untreated red cells to DTT alone (lane 2) had no effect on the normal electrophoretic pattern. However, post-treatment of divicine-treated red cells with DTT (lane 4) reversed the changes in the major structural proteins and decreased the level of membrane-bound hemoglobin (Fig. 5BGo). Similar results were obtained when cysteamine was substituted for DTT, and when the divicine-treated cells were lysed and solubilized with SDS prior to addition of DTT (data not shown). These data indicate that the skeletal protein alterations induced by divicine were due to the formation of intermolecular disulfide bonds.



View larger version (39K):
[in this window]
[in a new window]
 
FIG. 5. Effect of DTT on divicine-induced membrane skeletal protein alterations in rat erythrocytes. Rat erythrocytes were incubated with the vehicle or divicine (1.5 mM) for 1 h, then washed and incubated with DTT (5 mM) for an additional 1 h before being subjected to SDS-PAGE. (A) Coomassie blue-stained gel. (B) Immunoblot stained with rabbit polyclonal antibodies to rat hemoglobin. Lane 1, control; lane 2, DTT control; lane 3, divicine alone; lane 4, divicine + DTT.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The onset of favism has long been considered to be linked to the oxidative activity of the fava bean aglycones, divicine, and isouramil, within erythrocytes. However, despite several decades of research in the area, the mechanism underlying the favic response remains unclear. Two types of approaches have been used in attempts to resolve the mechanism: (1) examination of red cells withdrawn from favic patients during a hemolytic crisis, and (2) examination of red cells incubated with divicine or isouramil in vitro. Although examination of red cells retrieved from favic patients has provided important information, the major limitation of this approach in regard to understanding the mechanism is that the red cells that are most affected, and hence of greatest interest, are rapidly removed from the circulation by splenic and hepatic sequestration and are thus unavailable for study. The second approach also has had major limitations. Divicine is unstable in the presence of oxygen at physiological pH, and many investigators have chosen to prepare it in situ by enzymatic or chemical hydrolysis of its parent glucoside, vicine. These procedures, however, yield only small amounts of material and, in the case of acid hydrolysis, of chemically undefined material (Pedersen et al., 1988Go). Furthermore, the relevance of divicine-induced changes observed in these studies is questionable because of the lack of correlation with the relevant toxicological end-point, premature splenic sequestration of intact red cells.

We have used a direct synthetic method (Bailey et al., 1982Go) to prepare a stable hemisulfate salt of divicine (McMillan et al., 1993Go). This compound was able to provoke a favic-like response in G6PD-normal rats. That is, divicine administration reduced the survival of previously infused 51Cr-labeled red cells and decreased the hematocrit in a dose-dependent manner (McMillan and Jollow, 1999Go). The decrease in hematocrit was matched by an increase in spleen weight, reflective of the rapid uptake of divicine-damaged red cells. Of importance for the present studies, divicine hemotoxicity could be reproduced in vitro by direct exposure of 51Cr-labeled red cells to divicine before cells were returned to the circulation of isologous rats. Survival of these cells was also reduced in a concentration-dependent manner, and these cells were also found to be removed preferentially by the spleen. These observations indicate that the G6PD-normal rat red cell may be used as a model to examine the relevance of divicine-induced cellular alterations to the hemotoxic response.

The present studies demonstrate that divicine hemotoxicity in the rat is associated with the development of an oxidative stress within the red cell. This oxidative stress response was manifested by depletion of GSH with concomitant formation of glutathione-protein mixed-disulfides (Fig. 1Go), and by stimulation of HMP shunt activity (Fig. 2Go). Under these experimental conditions, alterations in the electrophoretic pattern of membrane skeletal proteins and the appearance of membrane-bound hemoglobin were observed (Fig. 4Go), and the red cells were transformed to an extreme echinocytic morphology (Fig. 3Go). As previously reported, when rat red cells treated in this manner are returned to isologous rats, the cells are rapidly and selectively removed by splenic sequestration (McMillan and Jollow, 1999Go).

The mechanism by which damaged or senescent red cells are removed by the spleen remains poorly understood. The exposure of protein, carbohydrate, and/or lipid epitopes on the external cell surface leading to recognition and ingestion by splenic macrophages have all been proposed to account for the removal of abnormal red cells from the circulation (Bratosin et al., 1998Go). Although the intracellular signal that provokes the appearance of an epitope on the external cell surface is also not known, it has long been postulated that alterations in the skeletal protein assembly that lies on the inner surface of the cell membrane may be responsible for initiating this process. In this regard, a number of studies have suggested a key role for the binding of hemoglobin to cytoskeletal protein in the mechanism underlying normal red cell senescence. For example, Low and colleagues (Turrini et al., 1991Go) showed that hemoglobin binding to the membrane of red cells stimulated the deposition of autologous antibodies and complement, and that these cells were ingested by cultured monocytes.

As noted above, examination of skeletal proteins from divicine-treated red cells revealed marked changes in the electrophoretic pattern. Two major features of this response were the loss of membrane protein bands 1, 2.1, and 4.2, and the appearance of membrane-associated aggregates of hemoglobin monomers. The reversal of these changes by treatment of the cells with DTT (Fig. 5Go) with the liberation of hemoglobin monomer indicated that the hemoglobin addition reactions had occurred as a result of disulfide bond formation. Although the significance of these protein changes is not yet known, they are quite similar to those observed previously in ghosts obtained from rat red cells exposed to dapsone hydroxylamine (Grossman et al., 1992Go), which is the hemolytic metabolite of the primary arylamine drug, dapsone.

The origin of the oxidant stress that converts the sulfhydryl groups of GSH, hemoglobin, and skeletal protein to mixed disulfides in these divicine-treated red cells is not yet known. Reversible oxidation of divicine to a quinone at the expense of molecular oxygen may generate reactive oxygen species (Chevion et al., 1982Go), which have the capacity to react with cellular thiol groups to form reactive thiyl radicals (Bradshaw et al., 1997Go; Wefers and Sies, 1983Go). Hemoglobin thiyl radicals could then attack the membrane proteins, resulting in the formation of hemoglobin-skeletal protein adducts and HMWP aggregates. Autooxidation of divicine in aerobic solution has been shown to generate hydrogen peroxide, and in the presence of GSH, the accumulation of GSSG has been observed (Arese, 1982Go). In our studies, however, GSSG was not detected in divicine-treated red cells, suggesting that the loss of GSH was not due to detoxification of peroxide by GSH peroxidase. Alternatively, mixed disulfide formation could occur as a result of attack by a compound-centered free radical (Winterbourn, 1993Go). In support of this concept, semiquinone radical intermediates of divicine have been detected in acellular systems by EPR spectroscopy (Albano et al., 1984Go; Pedersen et al., 1988Go). Whether or not these species are formed in red cells is not yet clear and warrants further investigation.

Several reports of divicine-induced morphological alterations have appeared in the literature (Beutler, 1978Go; Fischer et al., 1985Go; Weed and Reed, 1966Go). Fischer et al. (1985) described the appearance of a phenomenon referred to as plasma membrane "crossbonding." This altered morphology could be observed in red cells withdrawn from G6PD-deficient individuals during the early stages of a favic crisis, and is characterized by fusion of opposing faces of the internal surface of the plasma membrane, creating 2 compartments, one containing hemoglobin, and one apparently devoid of cellular material. Efforts to reproduce this response in vitro by exposure of rat red cells to divicine was achieved only when the cells were placed in a hypertonic medium (400 milliosmolar) to affect shrinkage of the cells. In the present in vitro incubation studies, intact treated cells with varying degrees of echinocytosis were evident but changes suggestive of crossbonding and cell fragmentation were not observed. Of interest, examination of red cells taken from rats 1 h after administration of a hemotoxic dose of divicine revealed a high proportion of knizocytes (pinched cells; triconcave disc) and occasional echinocytes (McMillan and Jollow, unpublished observations); however, crossbonded red cells were not evident under these in vivo conditions.

In summary, the data presented in this report demonstrate that an oxidative stress response is provoked in rat red cells exposed to concentrations of divicine that commit the cells to premature splenic sequestration. Under these experimental conditions, divicine hemotoxicity in the rat is associated with severe damage to the membrane skeleton, which is in turn manifested by alterations in red cell morphology.


    ACKNOWLEDGMENTS
 
This work was supported by NIH grant DK-47423. The authors wish to thank Jennifer Schulte and Elizabeth Eagleson for their technical assistance in the preparation of this manuscript.


    NOTES
 
1 To whom correspondence should be addressed. Fax: (843) 792–2475. E-mail: mcmilldc{at}musc.edu. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Albano, E., Tomasi, A., Mannuzzu, L., and Arese, P. (1984). Detection of a free radical intermediate from divicine of Vicia faba. Biochem. Pharmacol. 33, 1701–1704.[ISI][Medline]

Arese, P. (1982). Favism—a natural model for the study of hemolytic mechanisms. Rev. Pure Appl. Pharmacol. Sci. 3, 123–183.[Medline]

Bailey, S. W., Weintraub, S. T., Hamilton, S. M., and Ayling, J. E. (1982). Incorporation of molecular oxygen into pyrimidine cofactors by phenylalanine hydroxylase. J. Biol. Chem. 257, 8253–8260.[Abstract/Free Full Text]

Beutler, E. (1978). Hemolytic Anemia in Disorders of Red Cell Metabolism. Plenum Medical, New York.

Bradshaw, T. P., McMillan, D. C., Crouch, R. K., and Jollow, D. J. (1997). Formation of free radicals and protein mixed disulfides in rat red cells exposed to dapsone hydroxylamine. Free Rad. Biol. Med. 22, 1183–1193.[ISI][Medline]

Bratosin, D., Mazurier, J., Tissier, J. P., Estaquier, J., Huart, J. J., Ameisen, J. C., Aminoff, D., and Montreuil, J. (1998). Cellular and molecular mechanisms of senescent erythrocyte phagocytosis by macrophages. A review. Biochimie 80, 173–195.[ISI][Medline]

Chevion, M., Navok, T., Glaser, G., and Mager, J. (1982). The chemistry of favism-inducing compounds. The properties of isouramil and divicine and their reaction with glutathione. Eur. J. Biochem. 127, 405–409.[Abstract]

Dewar, C. K. (1982). Preparation of red cells for scanning and transmission electron microscopy. In Red Cell Membranes, A Methodological Approach (J. C. Ellory and J. D. Young, Eds.), p. 13. Academic Press, New York.

Fairbanks, G., Steck, T. L., and Wallach, D. F. (1971). Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10, 2606–2617.[ISI][Medline]

Fischer, T. M., Meloni, T., Pescarmona, G. P., and Arese, P. (1985). Membrane cross bonding in red cells in favic crisis: A missing link in the mechanism of extravascular haemolysis. Br. J. Haematol. 59, 159–169.[ISI][Medline]

Gaetani, G. F., Mareni, C., Salvidio, E., Galiano, S., Meloni, T., and Arese, P. (1979). Favism: Erythrocyte metabolism during haemolysis and reticulocytosis. Br. J. Haematol. 43, 39–48.[ISI][Medline]

Grossman, S., Budinsky, R., and Jollow, D. (1995). Dapsone-induced hemolytic anemia: Role of glucose-6-phosphate dehydrogenase in the hemolytic response of rat erythrocytes to N-hydroxydapsone. J. Pharmacol. Exp. Ther. 273, 870–877.[Abstract]

Grossman, S. J., Simson, J., and Jollow, D. J. (1992). Dapsone-induced hemolytic anemia: Effect of N-hydroxy dapsone on the sulfhydryl status and membrane proteins of rat erythrocytes. Toxicol. Appl. Pharmacol. 117, 208–217.[ISI][Medline]

Hegazy, M. I., and Marquardt, R. R. (1984). Metabolism of vicine and convicine in rat tissues: Absorption and excretion patterns and sites of hydrolysis. J. Sci. Food Agric. 35, 139–146.[ISI][Medline]

Jensen, C. B., Grossman, S. J., and Jollow, D. J. (1986). Improved method for determination of cellular thiols, disulfides and protein mixed disulfides using HPLC with electrochemical detection. Adv. Exp. Med. Biol. 197, 407–413.[Medline]

Mager, J., Glaser, G., Razin, A., Izak, G., Bien, S., and Noam, M. (1965). Metabolic effects of pyrimidines derived from fava bean glycosides on human erythrocytes deficient in glucose-6-phosphate dehydrogenase. Biochem. Biophys. Res. Commun. 20, 235–240.[ISI][Medline]

McMillan, D. C., and Jollow, D. J. (1999). Favism: Divicine hemotoxicity in the rat. Toxicol. Sci. 51, 310–316.[Abstract]

McMillan, D. C., Schey, K. L., Meier, G. P., and Jollow, D. J. (1993). Chemical analysis and hemolytic activity of the fava bean aglycon divicine. Chem. Res. Toxicol. 6, 439–444.[ISI][Medline]

McMillan, D. C., Simson, J. V., Budinsky, R. A., and Jollow, D. J. (1995). Dapsone-induced hemolytic anemia: Effect of dapsone hydroxylamine on sulfhydryl status, membrane skeletal proteins and morphology of human and rat erythrocytes. J. Pharmacol. Exp. Ther. 274, 540–547.[Abstract]

Mohandas, N., Chasis, J. A., and Shohet, S. B. (1983). The influence of membrane skeleton on red cell deformability, membrane material properties, and shape. Semin. Hematol. 20, 225–242.[ISI][Medline]

Pedersen, J. Z., Musci, G., and Rotilio, G. (1988). Electron spin resonance characterization of the radicals produced by enzymatic or chemical cleavage of vicine. Biochemistry 27, 8534–8536.[ISI]

Turrini, F., Arese, P., Yuan, J., and Low, P. S. (1991). Clustering of integral membrane proteins of the human erythrocyte membrane stimulates autologous IgG binding, complement deposition, and phagocytosis. J. Biol. Chem. 266, 23611–23617.[Abstract/Free Full Text]

Weed, R. I., and Reed, C. F. (1966). Membrane alterations leading to red cell destruction. Amer. J. Med. 41, 681–695.[ISI][Medline]

Wefers, H., and Sies, H. (1983). Oxidation of glutathione by the superoxide radical to the disulfide and the sulfonate yielding singlet oxygen. Eur. J. Biochem. 137, 29–36.[Abstract]

Winterbourn, C. C. (1993). Superoxide as an intracellular radical sink. Free Radic. Biol. Med. 14, 85–90.[ISI][Medline]

Winterbourn, C. C., and Munday, R. (1990). Concerted action of reduced glutathione and superoxide dismutase in preventing redox cycling of dihydroxypyrimidines, and their role in antioxidant defence. Free Radic. Res. Commun. 8, 287–293.[ISI][Medline]