Deshidratacion En Acf

ã

The Hematology Journal (2001) 2, 200 ± 205 2001 The European Haematology Association All rights reserved 1466 ± 4680/01 $15.00 www.nature.com/thj

HYPOTHESIS AND DEBATE

How do sickle cells become dehydrated? Patrick Merciris1 and FrancËoise Giraud*,1 1

Laboratoire des Biomembranes et Messagers Cellulaires, Universite Paris XI-Orsay, France

The Hematology Journal (2001) 2, 200 ± 205 Keywords: sickle cells; hemoglobin F; K/Cl cotransport; KCa channels; serine/threonine phosphatase 2A; protein tyrosine kinase

Introduction In sickle cell anemia, dehydrated erythrocytes (SS RBC) are involved in the propagation of the vasoocclusive process, which results in the most severe complication in sickle cell disease (e.g. stroke). Deoxygenation leads to red cell dehydration, and hence increases the rate of hemoglobin S (HbS) polymer formation, which subsequently sickles the cell and retards RBC transit through the microvasculature. Dehydrated SS RBC exhibit decreased filterability and poor deformability, contributing to acute and chronic vaso-occlusive episodes and organ damage and to early removal from the circulation, uncompensated by erythropoiesis, resulting in hemolytic anemia.1 The hydration state and the hemoglobin F (HbF) content of RBC are known to in¯uence their in vivo survival: SS RBC, and especially denser fractions, have a much shorter lifespan than normal RBC; and, the lifespan of cells with high levels of HbF (F cells), particularly in SS patients, is six to eight weeks, whereas that of non-F cells is only about two weeks.2 Populations of SS RBC are highly heterogeneous with respect to age, density, HbF and cation contents (Figure 1). The densest SS cells are, on average, younger than lesser dense cells, and, in vivo, subpopulations of young cells dehydrate at di€erent rates,3 with a fraction (mostly reticulocytes) being susceptible to rapid dehydration after deoxygenation in vitro. Dehydration is the result of a cationic loss, followed by Cl7 and water leaks to maintain electroneutrality and osmolarity of the cell. Deoxygenation-induced sickling leads to a permeabilization of the SS RBC membrane for mono and divalent cations, resulting in a net movement of Na+, K+, Mg2+ and Ca2+ along their electrochemical gradients, a *Correspondence: F Giraud, Biomembranes et Messagers Cellulaires, CNRS UMR 8619, Bat 440, Universite Paris XI 91405 Orsay Cedex, France; Tel: +33 1 69157644; Fax: +33 169154961 E-mail: [email protected] Received 13 November 2000; accepted 25 January 2001

process named sickling-induced pathway (SIP).4 Stimulation of SIP leads to initially balanced Na+ and K+ ¯uxes, but by increasing the Na+i/K+i ratio, this stimulation results in the activation of the Na/K pump and slow dehydration. The contribution of this mechanism is considered to be minor and cannot account for the extreme K+ depletion of highly dense cells.1 Two K+ transporters have a high capacity for mediating rapid K+ loss and dehydration in RBC: the Ca2+-activated K channels (KCa channels) and the K/Cl cotransport (KCl-cot). KCa channels, if maximally activated, dehydrate normal RBC very rapidly. They are activated in deoxygenated SS RBC, following the stimulation of Ca2+ in¯ux by SIP.3,5 However, as HbF delays and decreases HbS polymerization and hence reduces SIP, KCa-channel activation will exclude most of the F cells.3 KCl-cot is highly active in SS RBC, and particularly in young RBC, where it can promote a rapid dehydration when activated by cell swelling or acid pH.6 Until recently, it was thought that KCl-cot is inhibited by deoxygenation.7,8

Hypotheses and controversies on sickling-induced pathway The nature of SIP and whether it corresponds to the activation of a pre-existing system or to a newly formed pathway are presently unknown. As SIP is sustained during deoxygenation, reversed on reoxygenation and prevented in F cells,3,9 a ®rst hypothesis would suggest a relation to HbS polymerization and morphological sickling.10 The properties of SIP (lack of selectivity for cations, dependence on membrane potential, but insensitivity to inhibitors of all the known cationic transporters) are those of a di€usible pathway or a channel with low selectivity among metal cations.4 Na+ in¯ux mediated by SIP is reduced by external Ca2+,3,11 suggesting a competition between the two cations for the same transporter.

Hypothesis: sickle cell dehydration P Merciris and F Giraud

201

Figure 1 Heterogeneity of density-fractionated SS RBC. SS RBC were separated in 4 or 3 density fractions, with de®ned density limits, by centrifugation on a density gradient preparation (A, B) or in 3 fractions corresponding to the top (low density), the middle (intermediate density) or the bottom (high density) of the density gradient preparation (C), as indicated in the inset in each panel. (A) Distribution of the cells and mean corpuscular Hb concentration (MCHC), percentage of reticulocytes, of irreversibly sickled cells (ISC) and of HbF in each fraction; LD, low density cells; ND, normal density cells; ID, intermediate density cells; HD, high density cells.20 (B) Distribution of the cells and percentage of F-RBC, of T+ cells and of T+F+ cells in each fraction.19 (C) Concentration of Na+, K+ and Mg2+ in top, middle and bottom fraction.3,6,25

Some inhibitors of band 3 anion transport, such as 4,4'-diisothiocyano-2,2' disulfostilbene (DIDS), can reduce SIP, but with a concentration dependence di€ering from that of anion transport.4,12 Moreover, DIDS e€ect on SIP occurs without inhibiting morphological sickling. Therefore, if there is a relationship between sickling and SIP, it would involve di€erent steps, with DIDS inhibiting one of them. The interaction between HbS and the cytoplasmic domain of band 3 might provide a site that is energetically favorable for nucleation,9 and bundles of adjacent

polymers would therefore generate oligomeric aggregates of band 3, functioning as a large ion channel. In favor of this model, deoxy (and denatured) Hb have a much higher anity for band 3 than oxyHb, and clusters of band 3 are found to be locally associated with membrane-bound Heinz bodies in aged and SS RBC. However, to be valid such a mechanism would require that oligomerization of band 3 is suciently rapid to precede the initiation of the cationic permeabilization. In fact, autologous IgG binding on the surface of SS RBC ± which re¯ects the formation of band 3 clusters ± is a slow process, which becomes signi®cant only after at least 16 h of deoxygenation,13 a time scale incompatible with the rapid activation of SIP. A second hypothesis implies that the protrusion of spicule HbS polymers, which causes uncoupling of the membrane from the skeleton,14 would exert a distortion on the membrane, leading to the activation of nonselective mechano-sensitive, or stretch-activated, cation channels.11 Cation permeability, stimulated by stressinduced deformation of normal RBC, exhibits many of the characteristics of SIP: in¯ux of Ca2+ and activation of KCa channels, balanced movements of Na+ and K+ that are inhibited by DIDS; and inhibition of Na+ in¯ux by external Ca2+.15,16 However, SIP and deformation-induced ¯uxes are unlikely to be mediated by the same mechanism, as the concentration dependence of DIDS inhibition is distinct for the two processes. In addition, stress-induced cation transport is more consistent with a pathway formed by rare ¯uctuations in the membrane structure than with the activation of mechano-sensitive channels.16 Isolated spicule vesicles of deoxygenated SS cells contain non-selective cation channels, with characteristics similar to SIP, but how they are activated remains to be determined.17

KCl-cot vs KCa-channel activation: what is the initial trigger for dehydration? The e€ect of deoxygenation on the behavior of a very young and age-de®ned population of SS cells, which still contain transferrin receptors (T+) and either HbF (T+F+) or no HbF (T+F7) has been analysed.18,19 T+ cells dehydrate more than T7 cells with either 4 h continuous deoxygenation or oxy-deoxy (O-D) cycles. Under the former conditions, dehydration proceeds by activation of KCa channels, while under the latter conditions, both KCl-cot and KCa channels are activated. Acid stimulation (pH 7.2) of KCl-cot in moderately dense T+F+ or T+F7 cells promotes, to a similar extent, the formation of intermediate density T+ cells, consistent with a ®rst, HbF-independent, dehydration stage. However, as the circulating densest T+ cells (41.094) contained little or no HbF (Figure 1B), and have more KCl-cot activity (measured after rehydration and low pH activation) than those which remain light, the second stage of dehydration of these young cells, from moderately dense to hyperdense, is The Hematology Journal

Hypothesis: sickle cell dehydration P Merciris and F Giraud

202

considered to be sickling and HbF-dependent and therefore involves KCa-channel activation.18,19 A fraction of cells with normal density (ND; 1.08 ± 1.09 g/mL), enriched in reticulocytes (Figure 1A), generates dense cells of intermediate density (ID; 1.09 ± 1.114 g/mL) and hyperdense (HD; 41.114 g/ mL) cells by independent mechanisms,20 when exposed for 22 h to O-D sickling at pH 6.8, in the presence of Ca2+. Incubations in Cl7, NO37 or EGTA media or with speci®c inhibitors, provide evidence for the role of KCl-cot activation in ID cell formation and of KCa channels in the generation of HD cells. These data are further supported by the marked di€erence in the contribution of F cells to each dehydrated fraction. In the newly formed ID cells, the percentage of F cells is only slightly lower than in the parent ND cells, consistent with the HbF independence of KCl-cot activity.3,19 In contrast, in the newly formed HD cells, the percentage of F cells is much lower than in the parent ND cells, supporting a role for sickling-induced KCa-channel activation mostly in cells with low levels of HbF.3 Although the conclusions derived are convergent with those of others,18,19 the experimental conditions are not physiological (pH 6.8 and long incubations). Figure 2 shows a two-step pathway (sequential activation of KCl-cot and KCa channels) for maximal dehydration of T+F7 cells and a one-step pathway (activation of KCl-cot) for intermediate dehydration of T+F+ cells. This model also accounts for the generation of ID and HD cells, by KCl-cot and KCa channel activation, respectively and for their dependence on HbF.

Figure 2 Mechanisms of SS RBC dehydration induced by acid pH or deoxygenation. Two-step model for maximal dehydration of T+F7 cells (or formation of HD cells from ND cells) and onestep model for intermediate dehydration of T+F+ cells (or formation of ID cells from ND cells). ND, normal density cells; ID, intermediate density cells; HD, high density cells (the density limits are indicated); T+F+ and T+F7, young RBC containing transferrin receptors and HbF or no HbF, respectively; O-D cycles, oxygenation-deoxygenation cycles; deoxy, deoxygenation. The Hematology Journal

Initial activation of KCl-cot as the trigger of dehydration does not support the previously proposed model of reticulocyte dehydration,21 with initial activation of KCa channels, slight acidi®cation and secondary activation of KCl-cot. Such a mechanism is not consistent with the activation of KCl-cot in the absence of external Ca2+ during cyclic deoxygenation18 or during continuous deoxygenation in a fraction of Mg2+-clamped light SS RBC.7 It is neither contradicted nor proved by the ®nding that, upon deoxygenation, rapid dehydration of young SS cells is strictly Ca2+dependent.3 However, it cannot be excluded that, in some light cells (with low HbF content, large SIPinduced Ca2+ in¯ux and high density of KCa channels), activation of the channels could trigger activation of the cotransport.

A role for cyclic vs continuous deoxygenation In the absence of other stimuli such as low pH or cell swelling, deoxygenation promotes dehydration of unfractionated SS cells or T+ cells, through activation of KCl-cot, but only during relatively slow O-D cycling.18,22,23 The increase in intracellular Mg2+ free concentration ([Mg2+]i) induced by deoxygenation, was thought to be responsible for the inhibition of the transporter under continuous deoxygenation.8 Indeed, pharmacological loading of cells with Mg2+ inhibits the activity of KCl-cot.24 Consistent with this hypothesis, after clamping of [Mg2+]i with ionophore A23187, continuous deoxygenation, at physiological pH, stimulates KCl-cot in low-density SS RBC.7 Activation of KCl-cot by deoxygenation might result from a dephosphorylation of the transporter or its regulators7,18 (see below), but would be attenuated by the concomitant rise in [Mg2+]i. Upon reoxygenation, KCl-cot could express its activity, assuming that [Mg2+]i decreases in a shorter time than that required for rephosphorylation. Under fast O-D cycling (30 ± 40 s periods approaching those in vivo), cells would not remain in the deoxy state for a sucient time to allow dephosphorylation and KCl-cot activation upon subsequent reoxygenation.23 However, there is good evidence that this pathway contributes to SS RBC dehydration in vivo, as shown by the bene®cial e€ect of oral administration of magnesium on SS RBC hydration (see below). KClcot might be activated in SS RBC subjected to `slowerthan-normal' O-D cycles due to adherence to postcapillary venules. The role of deoxygenation-induced increase in [Mg2+]i on the inhibition of KCl-cot remains controversial. Such an increase in SS RBC (from 0.4 to 0.7 mM)25 is too small to signi®cantly a€ect the activity of KCl-cot, at least when activated by cell swelling or acid pH.24,26 The activation of the KCl-cot, after [Mg2+]i clamping,7 could result from the removal of other inhibitory divalent cations by the ionophore treatment. Mn2+ is a likely candidate, as it inhibits the transporter activity with a much higher anity than Mg2+.24 Total Mn2+ concentration in RBC is 3.4 mM.

Hypothesis: sickle cell dehydration P Merciris and F Giraud

203

Assuming that the bu€ering capacity of RBC for Mn is the same as that for Mg2+, the free concentration would be about 1 mM, and sucient to promote an inhibition of the transporter, especially under deoxygenated conditions. In RBC of many species (®sh, horse and sheep) and in normal human RBC, O2 exerts an overriding control on the activity of the KCl-cot (see references in Gibson et al.27). In the absence of a suciently high PO2, KClcot is inactive and refractory to stimuli such as low pH or cell swelling. The O2 dependence of KCl-cot is not mediated via changes in [Mg2+]i. The presence of an O2 sensor has been speculated to explain the activation of the transporter. In contrast, in SS RBC, KCl-cot remains capable of responding to these stimuli, even at low PO2.8,27 The di€erent O2 dependence of KCl-cot in SS RBC may be related to the interaction of the abnormal Hb with the transporter,28 to the associated properties of HbS (auto-oxidation, polymerization) or to secondary changes in the activity of kinases/ phosphatases (see below). In contrast to KCl-cot, activation of KCa channels does not require a period of reoxygenation. Indeed, a similar percentage of dehydrated T+ cells (as estimated from the density score) is generated through KCachannel activation, with either 4 h continuous or cyclic deoxygenation at pH 7.4.18 Continuous deoxygenation of ND cells, at pH 6.8, for 22 h promotes HD cell formation, by KCa-channel activation, to a greater extent than O-D cycling.20 Instant state measurement of KCa-channel activation demonstrates that each deoxygenation pulse, either single or repeated, causes a reversible, sustained SIP-induced Ca2+ in¯ux, channel activation and dehydration of 10 ± 45% of SS discocytes (1.095 ± 1.106 g/mL) to dense cells (41.118 g/mL).9 However, continuous deoxygenation causes only one early and limited density shift, whereas frequent O-D cycling generates a progressive increase in the fraction of dehydrated cells. This contradiction with the above mentioned data,18 may arise from di€erences in the experimental conditions or from the lower HbF content of T+ cells, when compared with that of SS discocytes (Figure 1A,B). T+ cells are predicted to be more sensitive to sickling-induced, KCachannel-mediated dehydration, than the discocytes and a quantitative di€erence in the percentage of dehydrated cells, formed by continuous or cyclic deoxygenation, to be less easily detectable.

2+

(PTK) phosphorylation sites. Hence, KCl-cot activity may be regulated by serine/threonine phosphorylation on the transporter itself and by tyrosine phosphorylation on regulatory proteins. Studies with speci®c inhibitors of serine-threonine phosphatases PP1 and PP2A30 (calyculin A and okadaic acid), or with inhibitor of serine-threonine PK and PTK31 (staurosporine), have demonstrated that dephosphorylation of the KCl-cot, or of a regulatory protein, promotes activity, while phosphorylation decreases activity. The targets and the identity of PP and PK involved in the regulation of KCl-cot are not yet clearly characterized and appear to depend on the stimulus. As the activation by staurosporine, or n-ethylmaleimide (NEM), is fully inhibited and reversed by calyculin A, the prevailing model is that staurosporine and NEM can inhibit a kinase that lies upstream to the step catalysed by the calyculinsensitive PP.31 ± 34 PTK of Src family (Fgr and Hck) negatively regulate the transporter in mouse RBC and are likely to be among the staurosporine- and NEMsensitive kinases.34 Consequently, the activation of PP2A (and KCl-cot) by NEM32 would result from NEM-induced inhibition of Src PTK. Mg2+ depletion suppresses the NEM-induced activation of PP2A and of the transporter,32 suggesting that the phosphatase activation depends not only on the inhibition of Src PTK, but also on the activity of another kinase inhibited by removal of Mg2+. This latter kinase might be a PTK of another family (Syk), as tyrphostin B46 (a PTK inhibitor) blocks the transporter stimulation induced by staurosporine or NEM in sheep RBC.33 Therefore, both inhibitory (Src) and activating (Syk) PTK may be involved in regulating the activity of PP2A and of the transporter. Activation of PP2A and of KCl-cot by this mechanism is induced by inhibition of Src (NEM or staurosporine) but requires a basal activity of Syk, because it is suppressed when Syk is inhibited (Mg2+ depletion). Mg2+ depletion (alone), which inhibits PP1, does not stimulate PP2A,32 presumably as it suppresses the activity of both Src 29

Phosphorylation/dephosphorylation: a signal to activate K+ transporters? The major erythroid KCl-cot (KCC1) is a member of the KCC family, including up to now four isoforms, which di€er in their distribution in di€erent tissues and in the presence of putative phosphorylation sites in their N- and C-terminal cytoplasmic domains. The human erythroid KCC1 contains several consensus phosphorylation sites for casein kinase 2 and protein kinase (PK) C, but has no protein tyrosine kinase

Figure 3 Hypothetical model of KCl-cot activation by staurosporine, NEM or Mg2+ depletion in human RBC and by deoxygenation in SS RBC. A represents the inactive form of the transporter and B the active form. PP2A, phosphoserinethreonine phosphatase type 2A; PSK, protein serine-threonine kinase; PTK, protein tyrosine kinase (Src or Syk family). KCl-cot can be activated by inhibition of Src PTK or PSK, or by activation of Syk PTK. The Hematology Journal

Hypothesis: sickle cell dehydration P Merciris and F Giraud

204

and Syk and, hence, would activate the transporter by inhibiting a serine/threonine protein kinase (PSK) which phosphorylates the dephosphorylated KClcot.32,33 A model of regulation of PP2A and KCl-cot in human RBC, integrating the data mentioned above, is presented in Figure 3. Dephosphorylation of the transporter (from state A to B) by PP2A induces its activation and the reverse reaction, catalysed by a PSK, promotes inactivation. The mechanism of KCl-cot activation in deoxygenated SS RBC remains speculative, but could involve protein phosphorylation, as deoxygenation induces both an okadaic acid-sensitive dephosphorylation of membrane proteins,35 presumably corresponding to the activation of PP2A, and a stimulation of PTK.36 More speci®cally, deoxygenation stimulates Syk, but has no e€ect on the Src kinase Lyn (P Merciris, MD HardyDessources and F Giraud, unpublished data). PTK inhibitors (tyrphostins) reduce SIP-induced K+ e‚ux in SS RBC, even in the absence of external Ca2+, suggesting the involvement of PTK on the activity of KCl-cot, KCa channels or both.36 A possible role for Syk in the activation of PP2A and KCl-cot in deoxygenated SS RBC would be consistent with the model in Figure 3. The regulation of KCa channels in RBC may involve both PKA and PKC. Prostaglandin E2 (PGE2) and endothelin-1 (ET-1), whose binding to speci®c receptors triggers respectively PKA and PKC stimulation, activate KCa channels.37,38 PGE2 activation presumably results from a direct PKA-mediated e€ect on the channels (or its regulator), as this metabolite is unable to stimulate Ca2+ in¯ux (B Lakatos and L Varecka, personal communication). ET-1 activates the channels by increasing the maximum velocity and the Ca2+ sensitivity.38 This activation is associated with PKC stimulation, as it is blocked by a PKC inhibitor, suggesting that PKC activation could result in a stimulation of Ca2+ in¯ux. The plasma levels of

PGE2 and ET-1 are elevated during the painful-crisis episodes in SS patients,39 and these physiological stimuli may play an important role in the dehydration of SS cells in the absence of deoxygenation. In deoxygenated SS RBC, KCa-channel activation occurs without PKC stimulation, but pretreatment with a phorbol ester, which activates PKCa, inhibits both SIP and Ca2+ ionophore-induced activation of the channels.40 Under such conditions, phosphorylation by PKC would negatively regulate the activity of KCa channels.

Conclusion Preventing sickle cell dehydration is of potential interest in therapeutic approaches for inhibition of HbS polymerization and ultimately for reducing vasoocclusive crisis in sickle cell patients. Clinical trials in SS patients with oral administration of magnesium or clotrimazole, to inhibit KCl-cot or KCa channels respectively, have produced bene®cial e€ects on RBC hydration state and reduction in the incidence of painful episodes.41,42 Although never investigated, these treatments are likely to contribute to improving the survival of both F and non-F cells, and may a€ect the percentage of HbF. Combination therapy with additional strategies using hydroxyurea or butyrate derivatives, aimed at inhibiting HbS polymerization by increasing HbF production, may also be of bene®t in sickle syndromes.42 Acknowledgements We thank Drs CH Joiner and RS Franco for critically reading the manuscript and Drs L Varecka and B Lakatos for communicating their data. This work was supported by funds from the French National Centre for Scienti®c Research (CNRS, UMR 8619), the University of Paris XI-Orsay, the French Society of Hematology to P Merciris and the Association for Research against Cancer to F Giraud (ARC no9267).

References 1 Bookchin RM, Lew VL. Pathophysiology of sickle cell anemia. Hematology-Oncology Clinics of North America 10: 1241, 1996. 2 Franco RS, Lohmann JM, Silberstein EB, May®eldPratt G, Palascak MB, Nemeth TA, Joiner CH, Weiner M, Rucknagel DL. Time-dependent changes in the density and hemoglobin F content of biotin-labeled sickle cells. Journal of Clinical Investigation 101: 2730, 1998. 3 Bookchin RM, Ortiz OE, Lew VL. Evidence for a direct reticulocyte origin of dense red cells in sickle cell anemia. Journal of Clinical Investigation 87: 113, 1991. 4 Joiner CH. Cation transport and volume regulation in sickle red blood cells. American Journal of Physiology 264: C251, 1993. 5 Etzion Z, Ti€ert T, Bookchin RM, Lew VL. E€ects of deoxygenation on active and passive Ca2+ transport and on the cytoplasmic Ca2+ levels of sickle cell anemia red cells. Journal of Clinical Investigation 92: 2489, 1993. The Hematology Journal

6 Brugnara C, Bunn HF, Tosteston DC. Regulation of erythrocyte cation and water content in sickle cell anemia. Science 232: 388, 1986. 7 Joiner CH, Jiang M, Fathallah H, Giraud F, Franco RS. Deoxygenation of sickle red blood cells stimulates KCl cotransport without a€ecting Na+/H+ exchange. American Journal of Physiology 274: C1466, 1998. 8 Canessa M, Fabry ME, Nagel RL. Deoxygenation inhibits the volume-stimulated, Cl7-dependent K+ e‚ux in SS and young AA cells: A cytosolic Mg2+ modulation. Blood 70: 1861, 1987. 9 Lew VL, Ortiz OE, Bookchin RM. Stochastic nature and red cell population distribution of the sickling-induced Ca2+ permeability. Journal of Clinical Investigation 99: 2727, 1997. 10 Mohandas N, Rossi ME, Clark MR. Association between morphologic distortion of sickle cells and deoxygenation-induced cation permeability increase. Blood 68: 450, 1986.

Hypothesis: sickle cell dehydration P Merciris and F Giraud

205

11 Joiner C, Jiang M, Franco RS. Deoxygenation-induced cation ¯uxes in sickle cells IV: modulation by external calcium. American Journal of Physiology 269: C403, 1995. 12 Rhoda MD, Apovo M, Beuzard Y, Giraud F. Ca2+ permeability in deoxygenated sickle cells. Blood 75: 2453, 1990. 13 Green GA, Kalra VK. Sickling-induced binding of immunoglobulin to sickle erythrocytes. Blood 71: 636, 1988. 14 Liu SC, Derick LH, Zhai S, Palek J. Uncoupling of the spectrin-based skeleton from the lipid bilayer in sickled red cells. Science 252: 574, 1991. 15 Larsen FL, Katz S, Roufogalis BD, Brooks DE. Physiological shear stresses enhance the Ca2+ permeability of human erythrocytes. Nature 294: 667, 1981. 16 Johnson RM. Membrane stress increases cation permeability in red cells. Biophysical Journal 67: 1876, 1994. 17 Jiang M, Joiner CH. Non-selective cation channels in sickle membrane spicules are inhibited by DIDS, a stilbene disulfonate drug which blocks sickling-induced cation ¯uxes in intact cells. Blood 94: 198a, 1999 (abstract). 18 Franco RS, Palascak M, Thompson H, Rucknagel DL, Joiner CH. Dehydration of transferrin receptor-positive sickle retinculocytes during continuous or cyclic deoxygenation: Role of K/Cl cotransport and extracellular calcium. Blood 88: 4359, 1996. 19 Franco RS, Thompson H, Palascak M, Joiner CH. The formation of transferrin receptor-positive sickle reticulocytes with intermediate density is not determined by fetal hemoglobin content. Blood 90: 3195, 1997. 20 Schwartz R, Musto S, Fabry M, Nagel R. Two distinct pathways mediate the formation of intermediate density cells and hyperdense cells from normal density sickle red blood cells. Blood 92: 4844, 1998. 21 Lew VL, Freeman CJ, Ortiz OE, Bookchin RM. A mathematical model of the volume, pH, and ion content regulation in reticulocytes. Application to the pathophysiology of sickle cell dehydration. Journal of Clinical Investigation 87: 100, 1991. 22 Apovo M, Beuzard Y, Galacteros F, Bachir D, Giraud F. The involvement of the Ca-dependent K channel and of the K-Cl cotransport in sickle cell dehydration during cyclic deoxygenation. Biochemical and Biophysical Acta 1225: 255, 1994. 23 McGoron AJ, Joiner CH, Palascak MB, Claussen WJ, Franco RS. Dehydration of mature and immature sickle red blood cells during fast oxygenation/deoxygenation cycles: role of KCl cotransport and extracellular calcium. Blood 95: 2164, 2000. 24 Brugnara C, Tosteson DC. Inhibition of K transport by divalent cations in sickle erythrocytes. Blood 70: 1810, 1987. 25 Ortiz OE, Lew VL, Bookchin RM. Deoxygenation permeabilizes sickle cell anemia red cells to magnesium and reverses its gradient in the dense cells. Journal of Physiology (London) 427: 211, 1990. 26 Ortiz-Carranza O, Adragna NC, Lauf PK. Modulation of K-Cl cotransport in volume-clamped low-K sheep erythrocytes by pH, magnesium, and ATP. American Journal of Physiology 271: C1049, 1996. 27 Gibson JS, Speake PF, Ellory JC. Di€erential oxygen sensitivity of the K+-Cl7-cotransporter in normal and sickle human red blood cells. Journal of Physiology 511: 225, 1998.

28 Vitoux D, Beuzard Y, Brugnara C. The e€ect of hemoglobin A and S on the volume- and pH-dependence of K-Cl cotransport in human erythrocyte ghosts. Journal of Membrane Biology 167: 233, 1999. 29 Pellegrino CM, Rybicki AC, Musto S, Nagel RL, Schwartz RS. Molecular identi®cation and expression of erythroid K : Cl cotransporter in human and mouse erythroleukemic cells. Blood Cells Molecular Disease 24: 31, 1998. 30 Lauf PK, Bauer J, Adragna NC, Fujise H, Zade-Oppen AMM, Hai Ryu K, Delpire E. Erythrocyte KCl cotransport: properties and regulation. American Journal of Physiology 263: C917, 1992. 31 Bize I, Dunham PB. Staurosporine, a protein kinase inhibitor, activates K-Cl cotransport in LK sheep erythrocytes. American Journal of Physiology 266: C759, 1994. 32 Bize I, Guvenc B, Buchbinder G, Brugnara C. Stimulation of human erythrocyte K-Cl cotransport and protein phosphatase type 2A by n-ethylmaleimide: role of intracellular Mg++. Journal of Membrane Biology 177: 159, 2000. 33 Flatman PW, Adragna N, Lauf P. Role of protein kinases in regulating sheep erythrocyte K-Cl cotransport. American Journal of Physiology 271: C255, 1996. 34 De Franceschi L, Fumagalli L, Olivieri O, Corrocher R, Lowell CA, Berton G. De®ciency of Src family kinases Fgr and Hck results in activation of erythrocyte K/Cl cotransport. Journal of Clinical Investigation 99: 220, 1997. 35 Fathallah H, Coezy E, De Neef R, Hardy-Dessources M, Giraud F. Inhibition of deoxygenation-induced membrane protein dephosphorylation and cell dehydration by phorbol esters and okadaic acid in sickle cells. Blood 86: 1999, 1995. 36 Merciris P, Hardy-Dessources MD, Sauvage M, Giraud F. Involvement of deoxygenation-induced increase in tyrosine kinase activity in sickle cell dehydration. European Journal of Physiology 436: 315, 1998. 37 Li Q, Jungmann V, Kiyatkin A, Low PS. Prostaglandin E2 stimulates a Ca2+-dependent K+ channel in human erythrocytes and alters cell volume and ®lterability. Journal of Biological Chemistry 271: 18651, 1996. 38 Rivera A, Rotter MA, Brugnara C. Endothelins activate Ca2+-gated K+ channels via endothelin B receptors in CD-1 mouse erythrocytes. American Journal of Physiology 277: C746, 1999. 39 Graido-Gonzalez E, Doherty JC, Bergreen EW, Organ G, Telfer M, McMillen MA. Plasma endothelin-1, cytokine, and prostaglandin E2 levels in sickle cell disease and acute vaso-occlusive sickle crisis. Blood 92: 2551, 1998. 40 Fathallah H, Sauvage M, Romero JR, Canessa M, Giraud F. E€ects of PKC a activation on Ca2+ pump and K(Ca) channel in deoxygenated sickle cells. American Journal of Physiology 273: C1206, 1997. 41 Beuzard Y, De Franceschi L, Vitoux D, Rouyer-Fessard P, Henri A. Vers les theÂrapeutiques des heÂmoglobinopathies ciblant le globule rouge. HeÂmatologie 5: 60, 1999. 42 Steinberg MH. Sickle Cell Disease: Present and Future Treatment. The American Journal of Medical Sciences 312: 166, 1996.

The Hematology Journal