PROSTAGLANDINE2 EFFECTS ON CATION FLUX IN SICKLE ERYTHROCITEGHOSTS
I. N. Rabinowitz,P. L. Wolf, S. Berman, N. Shikm, and P. Edwards Departmentof Pathology and ClinicalLaboratory Stanford UniversityMedical Center Stanford,California94305
ACRMX'LRDGEMF,NTS This work was performedunder NIB Contract NBLI 72-2987-B,and the cooperationof Sickle Cell Anemia and Research,Inc, San Francisco, California,is gratefullyacknowledged.
ABSTRACT ProstaglandinE2 has previouslybeen shown to enhance the shape transformationof sickle prone erythrocytes(8) and to reduce the oxygen resaturationof HemoglobinSS within intact sickle cell erythrocytes after deoxygenation(15). In view of the recent importanceattributed to calcium transportin maintainingerythrocyteshape and viabilityGO) and the suggestionthat prostaglandinsmay act via a calcium ionophore mechanism (9) on cell membranes,erythrocyteghosts were prepared followingthe method of Lepke and Passow (12) from normal and sickle cell anemia erythrocytes. These two classes of ghosts are shown to display differingpatterns of sodium and calcium transport,with calcium influx being preferentiallystimulatedby prostaglandinE2 in sickle cell ghosts. It is suggestedthat in hypoxic, stasis conditions in vivo, prostaglandinsmay play a role in accelerating sickling of sickle prone erythrocytesvia stimulationof calcium influx.
Accepted April 10, 1975 PROSTAGLANDINS APRIL
1975
VOL. 9 NO. 4
545
PROSTAGLANDINS
INTRODUCTION The erythrocyteof individualswho are either heteroxygous(AS, "trait") or homoxygous (SS, "disease")for HemoglobinS synthesismay display marked cell deformation,altered cation flux patterns,changes in oxygen dissociationkinetics of hemoglobin,or all three in response to oxygen depletion in the suspensionmedium. Since the initial report of a possible co-factorrole for prostaglandinE2 (PGE2) in the sickling sequence (8), it has been found that PGE2 decreases oxygen resaturationof hemoglobinSS (HbSS)within deoxygenated washed sickle cell erythrocytes(15). Reports of prostaglandineffects on sodium transportare widespread, but remain unsettled in the case of erythrocytes(2). This, and the increasingawarenessof the critical role of calcium in erythrocyte function (10) and, further,the suggestionthat a basic mode of action of prostaglandinsmay be via a calcium ionophoreaction (9), led us to focus our attention to PGE2 effects on sodium and calcium flux in the erythrocytemembrane.
We report data accumulatedon erythrocytescollected from SS volunteers who were not in the hospital;whose erythrocytespresenteda minimum of irreversiblysickled cells, showing less than 10% of cells sickled on wet mount slide at collectiontime (7); who did not take medication for 24 hours prior to collection;and whose cells were used within 24 hours of collection. It should be noted that in obtaining small volumes of blood fran volunteer subjectswith the disease or trait, we endeavoredto perform several differentin vitro experiments with a limited supply of blood. This was the major reason for using ghost cells resealedwith salt solutionsof choice, such that atomic absorptionanalyses for cations would be possible. This strategy for maximal use of generouslysuppliedblood also led to some experimentsbeing performedafter the erythrocyteshad been stored for several hours at 6oC and others without a refrigeration interval. In so far as shape change is concerned,there is evidence indicatingthat refrigerationalters the sickling sequenceof SS erythrocytes(14). The results of the intact cell shape change studies (14) became known to us after several flux experimentswith ghosts had already been run, and since putting the cells "on ice" for variable intervalswas previouslysuch a natural procedure,our records of when this did or did not occur are unreliable. We did not follow possible shape changes of the ghosts during the flux experiments. For the SS samples used in the flux experiments, approximately 60% of the procedureswere carried out before cold storage and 40% after. This may have produced an unknown incrementof variation in the resultswhich would perhaps be apparent in the individualestimated standarddeviations,the largest of which occurred in the calcium flux experiments. With ghost systems, there is also the problem of cell populationswhich are kineticallyinhomogenouswith respect to sodium flux (6) and the general procedure followed to prepare the
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
ghosts was that of Lepke and Passow (12) which results in a recovery of 85-95% ghosts, 90% or more of which are kineticallyhomogenous. Erythrocyteswere extracted from heparinizedblood by centrifugation (2000 g, 4OC, ten minutes) and washed three times with cold isotonic saline. Washed cells were sedimentedone more time and resuspendedin a minimal amount of MFS (2-(N-morpholino) ethanesulfonicacid) buffered isotonic saline, pH 6.0, to form a packed suspension. Aliquots drawn from the suspensionwere transferredto reaction tubes. After five minutes on ice, seven volumes of hemolyzingsolution (4 mM MgS04, 4 mM ATP, 5 mM MES, pH 6.0) were added to each tube. After five more minutes, the hemolyzedcells were sedimentedat 22,000 g, 4OC, ten minutes. The supernatantwas pipetted off and the ghosts suspended in a 4 mM MgC12, 4 mM ATP, solution. Isotonicitywas restoredby adding concentratedsalt solutionbuffered with 4 mM TES (N-tris (hydroxymethyl) methyl-2-aminoethanesulfonicacid) titrated to pli 7.4 with NaOH, such that the final concentrationof salts within the ghosts was approximately.153 N NaCl, 0.014 N KCl. This allowed for sodium determinationsin small numbers of SS erythrocyteghost cells. Resealing was accomplishedby incubatingthe tubes at 37'C with gentle invertingon a Labquake shaker for sixty minutes. Percent resealingwas estimatedby incorporationof 14C sucrose (21) at the resealing step. 14C counting after digestion of cells using the wet oxidationprocedure of Mahin and Lofbery (13) was performedwith "Scintosol-Complete" (Isolab,Inc., Akron, Ohio) in a Nuclear Chicago Unilux II ScintillationCounter. Calculationswere made for 14C incorporation,and correctionfor externallytrappedvolume and cations using lithium in the "stop" solution at each sampling time (11). After resealing,ghosts were sedimentedat 2000 g and resuspended in solutionsof interest. Calcium free solution consisted of .120 N NaCl, .OlO N KCl, .035 N choline chloride, .OOS N TES, pH 7.4, made up in double distilledwater with reagent grade chemicals (no calcium content).5 mMoles CaC12 replaced 5 mMoles of choline-Cl2 in the calcium containingmediun. Ouabain,where indicated,was added to a concentrationof 0.01 mg/ml. Incubationintervalswere ended by pipettingaliquots into an ice cold solution of ,116 N choline chloride, .OSO N LiCl, .OOS M TES, pH 7.4, ("stop"solution). Aliquots were sedimentedat 22,000 g, 4OC, 15 minutes, supernatant removed, and sediment digestedwith three 2 ml. washes of 25% trichloroacetic acid. Resealed cell counts per reaction tube were made using a Coulter Model F Counter, and Na, K, Li, and Ca were assayed with a Perkin-ElmerModel 303 Atomic AbsorptionSpectrometer. The experiments were performedwith ghost cell counts which did not vary by more than 20% from one experimentto the next, and all results were correctedto a common base of lo3 cells/mm3. Experimentsat normal concentratiw of oxygen in buffer solutionswere performed in polycarbonatetubes fitted with silicone rubber stoppers. At reduced oxygen concentration, the tubes were constructedof Dow-CorningSilastic Rubber (No. 500-3, 0.01" thick) cementedwith Dow-CorningSilasticAdhesive, Type A, with silicone rubber stoppers. These tubes were kept just beneath the surface of a 370C water bath and slowly tumbled with a stirringmanifold. The silastic rubber is permeable only to gas. PGE2,generouslyprovided by the Upjohn Co., was dissolved in 70%
APRIL
1975
VOL. 9 NO. 4
547
PROSTAGLANDINS
EtOH, evaporatedto dryness before use, and redissolvedin buffer solutionsof choice for addition to reaction tubes.
RESULTS The prepared ghosts were "red" ghosts containingless than 5% of their __ original hemoglobin,as determinedby conversionto cyanmethemoglobin cycle were not (17). Further iterationsof the hemolysis-resealing carried out, as "white" ghosts can display altered membrane bound romise enzyme activities (4). This procedurewas adopted as a c to maintain maximal membrane bound enzyme activity (e.g. om$ Na -K' ATPase), while simultaneouslyremoving hemoglobinsol-gel transformations from considerationwith respect to the flux results. The mean percent of resealedghosts for four SS plus two AS plus three normal (A) erythrocytesamples was 93.6% _+0.9%. After 165 minutes of incubation, 93.3% + 1.0% remained sealed. The correspondingfigures in the presence of PGE2 were 93.9% + 1.1% and 93.6% + 1.0%. Individualpercent resealingresults showed no tendency for SS or AS ghosts to leak either more or less than AA ghosts with respect to sucrose. The maintenance of sealed ghosts throughoutthe incubationperiod, at least in so far as 14C sucrose leakage is concerned,should be borne in mind when interpreting the calcium results. In Figure 1, the net movement of sodium is followedin both normal and SS ghosts in the presence and absence of transport inhibitors. The increase in normal ghosts cell sodium concentrationduring incubation is reversed in the case of ghosts of SS erythrocytes. No attempt was made to distinguishsodium flux from possiblewater movement (22). Calcium added to the externalmedium and to a larger extent, ouabain, inhibitedboth of these responses. Calcium inhibitionof sodium flux was first noted in ghost systems by Teorell (16). Of particularinterestwas the "enhancement"of sickling in the presence of PGE2 (8) and whether this could be seen in flux activity under simulatedsicklingconditions. To this end, the silastic tubes were incubatedin a 37OC bath, such that as nitrogenwas bubbled into the bath, the oxygen concentrationwithin the tubes decreasedwith time, as seen in Figure 2. The percentageof intact SS erythrocyteswhich sickle as oxygen concentrationis decreased,increasesfrom 2% at 100% oxygen saturationof hemoglobinto 72% at 70% 02 saturationand finally to 100% at 20% saturation (5). Using this deoxygenationprocedure,the same phenomenonof net sodium increaseand loss was observed in normal vs. SS ghosts, and PGE2 is seen to have little significanteffect on net sodium movement (Figure3). In contrast,marked increase in calcium influx into the SS ghosts is shown in Figure 4. In the absence of PGE2, the SS ghosts exhibit an 11% uptake of calcium compared to normal ghosts, and this uptake is increasedto 31%, or slightlymore than 1 equiv. in the presence of PGE2. In these relativelyshort incubationexperiments,we did not attempt to measure possible inactivation of PGE2 or penetrationinto the ghosts (23,24,25). The increased influx of calcium, in the absence of PGE2,hasbeen observedby Eaton et al. (18) for intact sickle cell erythrocytesfollowingdeoxygenation.
548
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
_----5 3 4 -10
\
-15
\ \ -20
-25 45
105 TIME, MINUTES
165
Transport of sodium in normal (curves A,B,C) Fip. 1 and sickle erythrocyte ghosts (curves D,E,F). Curves B and E indicate effect on sodium transport in incubation medium containing calcium. Curves C and F indicate effect of adding ouabain to incubation medium containing no calcium. Curves A and D indicate transport without added ouabain or calcium. Ordinate represents the change (ANa) in mEquiv./L. per lo3 ghosts/mm3. Normal points are the mean of five separate experiments, and sickle cell points the mean of seven different experiments. Bars indicate estimated standard deviation for each point.
APRIL 1975 VOL. 9 NO. 4
549
PROSTAGLANDINS
I.01
, 20
I
40
4
60
1
80
L
,
L
,
I
100
120
140
160
180
MINUTES
Reduction of oxygen concentration (ppm 02) within Fig. 2 silastic tubes filled with cell suspension as nitrogen is bubbled into surrounding 37'C water bath, commencing at zero minutes. Horizontal dotted lines indicate time at which the partial pressure of oxygen in the tubes (~02) reaches arterial (94 mm. Hg) and venous (40 mm. Hg) values. Each triangle represents mean of three separate experiments on normal ghosts. Oxygen concentration measured with a Beckman Model 1008 Oxygen Analyzer, utilizing Clark type electrode.
550
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
+10 -
9 q
-lO-
45
105
165
TIME, MINUTES
Transport of sodium in ghosts in calcium free Fig. 3 medium as ghost suspension is deoxygenated at rate indicated in Figure 2. Curve A represents the mean of nine experiments (five normal plus four AS) without PGE2 and curve B the mean of nine experiments (five normal plus four AS) with PGE2 added (2 x 10S7M) at zero time to a suspension of normal and AS ghosts. Curve C represents the mean of seven experiments without PGE , and curve D the mean of seven experiments with PGE2 a2ded (2 x 10'7M) at zero time to a suspension of sickle cell ghosts. The ordinate represents the change (ANa) in mEquiv./L. per 103 cells/mm3. Bars indicate estimated standard deviation. The values for normal plus AS ghosts have been averaged together for graphical clarity, as their individual values are indistinguishable from each other to within one standard deviation.
APRIL 1975
VOL. 9 NO. 4
551
PROSTAGLANDINS
+1.2
+l.O
+0.8
-0.2
I
1
y
!
I
45
105
185
TIME, MINUTES
Conditions are indicated in Figure 3, except for Fin. 4 addition of calcium to incubation medium. Ordinate represents change (ACa) in calcium within the ghosts expressed as @quiv./L. per lo3 ghosts/mm3. Curve A represents normal ghosts without PGE2 added, and curve B normal ghosts with the addition of PGE2 (2 x 10m7M). Curve C represents sickle cell ghosts without PGE2 added, and curve D in the presence of 2 x lo-TM PGE2. Curves A and B represent the mean values from five normal plus four AS ghost populations. Curves C and D represent the mean values from seven experiments with sickle cell ghosts. Bars indicate estimated standard deviations. Trait and normal ghost values averaged together for reasons given in legend to Figure 3.
552
APRIL 1975 VOL. 9 NO. 4
PROSTAGLANDINS
DISCUSSION The SS erythrocyte membrane displays at least two different cation flux reactivities compared with normal membranes; an altered reactivity to excess sodium at the inner surface of the membrane, and an increased calcium influx under simulated sickling conditions. Gordesky et al. (3) have shown that, although total phospholipid concentration may be the same in normal and SS erythrocyte membranes, the structural arrangements are apparently different. As Whittsm has shown that the Na+-* activated ATPase is asymmetric with respect to sodium and potassium stimulation (19), the comparative sodium flux patterns observed here may reflect the response to sodium inhibition of the enzyme in two different environments. The high internal and nonphysiological sodium concentrations may also explain the reversed sodium flux reported here, compared to that reported by Tosteson (2) using intact cells. The calcium response is maximal at pO2 values approaching venous pO2 (Figures 2,4) where sickling of intact cells would be in excess of 70%, but is partially reversed at even lower oxygen concentrations and is stimuiated by-PGE2, A possible role for-PGE2 in the _in vivo sickling cycle would then be to increase calcium influx under stasis or hypoxic-conditions, thus, increasing membrane rigidity and accelerating the sickling of more cells. The relation of Ca* and the Ca*/ ATP ratio to cell deformability is complex (cf:(lO)) and it remains to be demonstrated whether transitory movement of Cak into the membrane would be sufficient to account for deformability effects in a scheme involving contractile membrane proteins such as spectrin, or whether there must be, in addition, consideration of the several glycolytic enzymes whose activities calcium is known to affect (1). In this regard, it has been shown that the shift to the right of the oxygen dissociation curve of hemoglobin SS within intact erythrocytes in the presence of PGE2 is accompanied by an increase in methemoglobin, suggesting a metabolic impairment of reduced pyridine nucleotide production necessary for reduction of methemoglobin (15).
APRIL
1975
VOL. 9 NO. 4
553
PROSTAGLANDINS
RRFFBENCES 1. Bygrave, F. L. The Ionic Environmentand Metabolic Control, Nature 214:667, 1967. 2. Gardner, J. D. and E. R. Ginzler. Sodium Transportin Human Erythrocytes- Absence of an Effect of ProstaglandinEl, Biochem. Biophy. Res. Comn. 42:1063, 1971. 3. Gordesky,S. E., G. V. Marinetti,and C. B. Segel. Difference in the Reactivityof Phospholipidswith FDNB in Normal RBC, Sickle Cells and RBC Ghosts, Biochem. Biophys. Res. Comm. 47: 1004, 1972. 4. Hanahan, D. J. and J. Ekholm. Changes in ErythrocyteMembranes During Preparation,as Expressedby ATPase Activity, Biochim. Biophys.Acta 255:413, 1972. Ham, and W. B. Castle. 5. Harris, J. W., H. A. Brewster,T. l-l. Studies on the Destructionof Red Blood Cells. X. The Biophysics and Biology of Sickle Cell Disease, Arch. Int. Med. 97:145, 1956. 6. Hoffman, J. L. The Active Transportof Sodium by Ghosts of Human Red Blood Cells, J. Gen. Physiol. 45:837, 1962. 7. Jensen, M, S. B. Shohet, and D. G. Nathan. The Role of Red Cell Energy Metabolism in the Generationof IrreversiblySickled Cells In Vitro, Blood 42:835, 1974. a. Johnson,M., I. Rabinowitz,A. L. Willis, and P. L. Wolf. Detection of ProstaglandinInductionof ErythrocyteSickling, Clin. Chem. 19: 23, 1973. 9. Kirtland,S. J. and H. Baum. ProstaglandinEl May Act as a "Calcium Ionophore", Nature New Biol. 236:47, 1972. 10. LaCelle, P. L. and R. I. Weed. The Contributionof Normal and PathologicErythrocytesto Blood Rheology, Progress in Hematology, Vol. VII, Grune & Stratton,New York, 1971. 11. Lepke, S. and H. Passow. Effects of Fluoride on Potassiumand Sodium Permeabilityof the ErythrocyteMembrane, J. Gen. Physiol. 51: 3658, 1968. 12. Lepke, S. and H. Passow. The Effect of pH at Hemolysison the Reconstitutionof Low Cation Permeabilityin Human Erythrocyte Ghosts, Biochim.Biophys.Acta 255:696, 1972. 13. Mahin, D. T. and R. T. Lofbery. A SimplifiedMethod of Sample Preparationfor Determinationof Tritium, Carbon-14,or Sulfur-35 in Blood or Tissue by Liquid ScintillationCounting, Anal. Biochem. 16:500, 1966.
554
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
I. N., P. L. Wolf,and SBerman. Light Scattering 14. Rabinowitz, Studies of Retardationof Sickling by Aspirin-likeDrugs, Res. Coann.Chem. Pathol. and Pharmacol. 8:417, 1974. 15. Rabinowitz,I. N., P. L. Wolf, N. Shikuma, and S. Berman. ProstaglandinE2 Effects on Oxygen Affinity by Sickle Erythrocytes, Prostaglandins7:309, 1974. 16. Teorell, T. PermeabilityPropertiesof ErythrocyteGhosts, J. Gen. Physiol. 35:669, 1952. 17. Van Kampen, E. J. and W. G. Zijlstra. Standardizationof Method, Clin. Hemoglobinometry. II. The Hemoglobincyanide Chim. Acta 6:538, 1961. S. Swofford,C. E. Kolpin, and 18. Eaton, J. W., T. D. Skelton,I-I. H. S. Jacob. Elevated ErythrocyteCalcium in Sickle Cell Disease, Nature 246:105, 1973. 19.
Whittam, R. and M. E. Ager. Dual Effects of Sodium Ions on an Erythrocyte-Membrane Adenosine Triphosphatase, Biochim. Biophys. Acta 65:383, 1962.
20.
Tosteson,D. C. The Effects of Sickling on Ion Transport. II. Effect of Sickling on Sodium and Cesium Transport, J. Gen. Physiol. 39:55, 1955.
21. Wilbrandt,W. OsmotischeNatur SogenannterNichtosomotische Htlmolysen(Kolloid-Osmotische HYmolyse), Pflug. Arch. Ges. Physiol. 245:22, 1941. 22. Allen, J. E. and C. R. Valeri. Prostaglandinsin Hematology, Arch. Intern.Med. 133:86, 1974. 23. Hensby, C. N. Reductionof ProstaglandinE to Prostaglandin F2c by an Enzyme in Sheep Blood, Brit. J. Pharmacol.50:462P, 1974. 24. Smith, J. B., J. J. Kocsis, C. Ingerman,and M. J. Silver. Inactivationof ProstaglandinAl and ProstaglandinA2 by Human Red Cells, Pharmacologist15:208 (abstr.288), 1973. 25. Shaw, J., W. Gibson, S. Jessup, and P. Ramwell. The Effect of PGEl on Cyclic AMP and Ion Movements in Turkey Erythrocytes, Ann. N.Y. Acad. Sci. 180:241, 1971.
APRIL
1975
VOL. 9 NO. 4
555
I. N. Rabinowitz,P. L. Wolf, S. Berman, N. Shikm, and P. Edwards Departmentof Pathology and ClinicalLaboratory Stanford UniversityMedical Center Stanford,California94305
ACRMX'LRDGEMF,NTS This work was performedunder NIB Contract NBLI 72-2987-B,and the cooperationof Sickle Cell Anemia and Research,Inc, San Francisco, California,is gratefullyacknowledged.
ABSTRACT ProstaglandinE2 has previouslybeen shown to enhance the shape transformationof sickle prone erythrocytes(8) and to reduce the oxygen resaturationof HemoglobinSS within intact sickle cell erythrocytes after deoxygenation(15). In view of the recent importanceattributed to calcium transportin maintainingerythrocyteshape and viabilityGO) and the suggestionthat prostaglandinsmay act via a calcium ionophore mechanism (9) on cell membranes,erythrocyteghosts were prepared followingthe method of Lepke and Passow (12) from normal and sickle cell anemia erythrocytes. These two classes of ghosts are shown to display differingpatterns of sodium and calcium transport,with calcium influx being preferentiallystimulatedby prostaglandinE2 in sickle cell ghosts. It is suggestedthat in hypoxic, stasis conditions in vivo, prostaglandinsmay play a role in accelerating sickling of sickle prone erythrocytesvia stimulationof calcium influx.
Accepted April 10, 1975 PROSTAGLANDINS APRIL
1975
VOL. 9 NO. 4
545
PROSTAGLANDINS
INTRODUCTION The erythrocyteof individualswho are either heteroxygous(AS, "trait") or homoxygous (SS, "disease")for HemoglobinS synthesismay display marked cell deformation,altered cation flux patterns,changes in oxygen dissociationkinetics of hemoglobin,or all three in response to oxygen depletion in the suspensionmedium. Since the initial report of a possible co-factorrole for prostaglandinE2 (PGE2) in the sickling sequence (8), it has been found that PGE2 decreases oxygen resaturationof hemoglobinSS (HbSS)within deoxygenated washed sickle cell erythrocytes(15). Reports of prostaglandineffects on sodium transportare widespread, but remain unsettled in the case of erythrocytes(2). This, and the increasingawarenessof the critical role of calcium in erythrocyte function (10) and, further,the suggestionthat a basic mode of action of prostaglandinsmay be via a calcium ionophoreaction (9), led us to focus our attention to PGE2 effects on sodium and calcium flux in the erythrocytemembrane.
We report data accumulatedon erythrocytescollected from SS volunteers who were not in the hospital;whose erythrocytespresenteda minimum of irreversiblysickled cells, showing less than 10% of cells sickled on wet mount slide at collectiontime (7); who did not take medication for 24 hours prior to collection;and whose cells were used within 24 hours of collection. It should be noted that in obtaining small volumes of blood fran volunteer subjectswith the disease or trait, we endeavoredto perform several differentin vitro experiments with a limited supply of blood. This was the major reason for using ghost cells resealedwith salt solutionsof choice, such that atomic absorptionanalyses for cations would be possible. This strategy for maximal use of generouslysuppliedblood also led to some experimentsbeing performedafter the erythrocyteshad been stored for several hours at 6oC and others without a refrigeration interval. In so far as shape change is concerned,there is evidence indicatingthat refrigerationalters the sickling sequenceof SS erythrocytes(14). The results of the intact cell shape change studies (14) became known to us after several flux experimentswith ghosts had already been run, and since putting the cells "on ice" for variable intervalswas previouslysuch a natural procedure,our records of when this did or did not occur are unreliable. We did not follow possible shape changes of the ghosts during the flux experiments. For the SS samples used in the flux experiments, approximately 60% of the procedureswere carried out before cold storage and 40% after. This may have produced an unknown incrementof variation in the resultswhich would perhaps be apparent in the individualestimated standarddeviations,the largest of which occurred in the calcium flux experiments. With ghost systems, there is also the problem of cell populationswhich are kineticallyinhomogenouswith respect to sodium flux (6) and the general procedure followed to prepare the
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
ghosts was that of Lepke and Passow (12) which results in a recovery of 85-95% ghosts, 90% or more of which are kineticallyhomogenous. Erythrocyteswere extracted from heparinizedblood by centrifugation (2000 g, 4OC, ten minutes) and washed three times with cold isotonic saline. Washed cells were sedimentedone more time and resuspendedin a minimal amount of MFS (2-(N-morpholino) ethanesulfonicacid) buffered isotonic saline, pH 6.0, to form a packed suspension. Aliquots drawn from the suspensionwere transferredto reaction tubes. After five minutes on ice, seven volumes of hemolyzingsolution (4 mM MgS04, 4 mM ATP, 5 mM MES, pH 6.0) were added to each tube. After five more minutes, the hemolyzedcells were sedimentedat 22,000 g, 4OC, ten minutes. The supernatantwas pipetted off and the ghosts suspended in a 4 mM MgC12, 4 mM ATP, solution. Isotonicitywas restoredby adding concentratedsalt solutionbuffered with 4 mM TES (N-tris (hydroxymethyl) methyl-2-aminoethanesulfonicacid) titrated to pli 7.4 with NaOH, such that the final concentrationof salts within the ghosts was approximately.153 N NaCl, 0.014 N KCl. This allowed for sodium determinationsin small numbers of SS erythrocyteghost cells. Resealing was accomplishedby incubatingthe tubes at 37'C with gentle invertingon a Labquake shaker for sixty minutes. Percent resealingwas estimatedby incorporationof 14C sucrose (21) at the resealing step. 14C counting after digestion of cells using the wet oxidationprocedure of Mahin and Lofbery (13) was performedwith "Scintosol-Complete" (Isolab,Inc., Akron, Ohio) in a Nuclear Chicago Unilux II ScintillationCounter. Calculationswere made for 14C incorporation,and correctionfor externallytrappedvolume and cations using lithium in the "stop" solution at each sampling time (11). After resealing,ghosts were sedimentedat 2000 g and resuspended in solutionsof interest. Calcium free solution consisted of .120 N NaCl, .OlO N KCl, .035 N choline chloride, .OOS N TES, pH 7.4, made up in double distilledwater with reagent grade chemicals (no calcium content).5 mMoles CaC12 replaced 5 mMoles of choline-Cl2 in the calcium containingmediun. Ouabain,where indicated,was added to a concentrationof 0.01 mg/ml. Incubationintervalswere ended by pipettingaliquots into an ice cold solution of ,116 N choline chloride, .OSO N LiCl, .OOS M TES, pH 7.4, ("stop"solution). Aliquots were sedimentedat 22,000 g, 4OC, 15 minutes, supernatant removed, and sediment digestedwith three 2 ml. washes of 25% trichloroacetic acid. Resealed cell counts per reaction tube were made using a Coulter Model F Counter, and Na, K, Li, and Ca were assayed with a Perkin-ElmerModel 303 Atomic AbsorptionSpectrometer. The experiments were performedwith ghost cell counts which did not vary by more than 20% from one experimentto the next, and all results were correctedto a common base of lo3 cells/mm3. Experimentsat normal concentratiw of oxygen in buffer solutionswere performed in polycarbonatetubes fitted with silicone rubber stoppers. At reduced oxygen concentration, the tubes were constructedof Dow-CorningSilastic Rubber (No. 500-3, 0.01" thick) cementedwith Dow-CorningSilasticAdhesive, Type A, with silicone rubber stoppers. These tubes were kept just beneath the surface of a 370C water bath and slowly tumbled with a stirringmanifold. The silastic rubber is permeable only to gas. PGE2,generouslyprovided by the Upjohn Co., was dissolved in 70%
APRIL
1975
VOL. 9 NO. 4
547
PROSTAGLANDINS
EtOH, evaporatedto dryness before use, and redissolvedin buffer solutionsof choice for addition to reaction tubes.
RESULTS The prepared ghosts were "red" ghosts containingless than 5% of their __ original hemoglobin,as determinedby conversionto cyanmethemoglobin cycle were not (17). Further iterationsof the hemolysis-resealing carried out, as "white" ghosts can display altered membrane bound romise enzyme activities (4). This procedurewas adopted as a c to maintain maximal membrane bound enzyme activity (e.g. om$ Na -K' ATPase), while simultaneouslyremoving hemoglobinsol-gel transformations from considerationwith respect to the flux results. The mean percent of resealedghosts for four SS plus two AS plus three normal (A) erythrocytesamples was 93.6% _+0.9%. After 165 minutes of incubation, 93.3% + 1.0% remained sealed. The correspondingfigures in the presence of PGE2 were 93.9% + 1.1% and 93.6% + 1.0%. Individualpercent resealingresults showed no tendency for SS or AS ghosts to leak either more or less than AA ghosts with respect to sucrose. The maintenance of sealed ghosts throughoutthe incubationperiod, at least in so far as 14C sucrose leakage is concerned,should be borne in mind when interpreting the calcium results. In Figure 1, the net movement of sodium is followedin both normal and SS ghosts in the presence and absence of transport inhibitors. The increase in normal ghosts cell sodium concentrationduring incubation is reversed in the case of ghosts of SS erythrocytes. No attempt was made to distinguishsodium flux from possiblewater movement (22). Calcium added to the externalmedium and to a larger extent, ouabain, inhibitedboth of these responses. Calcium inhibitionof sodium flux was first noted in ghost systems by Teorell (16). Of particularinterestwas the "enhancement"of sickling in the presence of PGE2 (8) and whether this could be seen in flux activity under simulatedsicklingconditions. To this end, the silastic tubes were incubatedin a 37OC bath, such that as nitrogenwas bubbled into the bath, the oxygen concentrationwithin the tubes decreasedwith time, as seen in Figure 2. The percentageof intact SS erythrocyteswhich sickle as oxygen concentrationis decreased,increasesfrom 2% at 100% oxygen saturationof hemoglobinto 72% at 70% 02 saturationand finally to 100% at 20% saturation (5). Using this deoxygenationprocedure,the same phenomenonof net sodium increaseand loss was observed in normal vs. SS ghosts, and PGE2 is seen to have little significanteffect on net sodium movement (Figure3). In contrast,marked increase in calcium influx into the SS ghosts is shown in Figure 4. In the absence of PGE2, the SS ghosts exhibit an 11% uptake of calcium compared to normal ghosts, and this uptake is increasedto 31%, or slightlymore than 1 equiv. in the presence of PGE2. In these relativelyshort incubationexperiments,we did not attempt to measure possible inactivation of PGE2 or penetrationinto the ghosts (23,24,25). The increased influx of calcium, in the absence of PGE2,hasbeen observedby Eaton et al. (18) for intact sickle cell erythrocytesfollowingdeoxygenation.
548
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
_----5 3 4 -10
\
-15
\ \ -20
-25 45
105 TIME, MINUTES
165
Transport of sodium in normal (curves A,B,C) Fip. 1 and sickle erythrocyte ghosts (curves D,E,F). Curves B and E indicate effect on sodium transport in incubation medium containing calcium. Curves C and F indicate effect of adding ouabain to incubation medium containing no calcium. Curves A and D indicate transport without added ouabain or calcium. Ordinate represents the change (ANa) in mEquiv./L. per lo3 ghosts/mm3. Normal points are the mean of five separate experiments, and sickle cell points the mean of seven different experiments. Bars indicate estimated standard deviation for each point.
APRIL 1975 VOL. 9 NO. 4
549
PROSTAGLANDINS
I.01
, 20
I
40
4
60
1
80
L
,
L
,
I
100
120
140
160
180
MINUTES
Reduction of oxygen concentration (ppm 02) within Fig. 2 silastic tubes filled with cell suspension as nitrogen is bubbled into surrounding 37'C water bath, commencing at zero minutes. Horizontal dotted lines indicate time at which the partial pressure of oxygen in the tubes (~02) reaches arterial (94 mm. Hg) and venous (40 mm. Hg) values. Each triangle represents mean of three separate experiments on normal ghosts. Oxygen concentration measured with a Beckman Model 1008 Oxygen Analyzer, utilizing Clark type electrode.
550
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
+10 -
9 q
-lO-
45
105
165
TIME, MINUTES
Transport of sodium in ghosts in calcium free Fig. 3 medium as ghost suspension is deoxygenated at rate indicated in Figure 2. Curve A represents the mean of nine experiments (five normal plus four AS) without PGE2 and curve B the mean of nine experiments (five normal plus four AS) with PGE2 added (2 x 10S7M) at zero time to a suspension of normal and AS ghosts. Curve C represents the mean of seven experiments without PGE , and curve D the mean of seven experiments with PGE2 a2ded (2 x 10'7M) at zero time to a suspension of sickle cell ghosts. The ordinate represents the change (ANa) in mEquiv./L. per 103 cells/mm3. Bars indicate estimated standard deviation. The values for normal plus AS ghosts have been averaged together for graphical clarity, as their individual values are indistinguishable from each other to within one standard deviation.
APRIL 1975
VOL. 9 NO. 4
551
PROSTAGLANDINS
+1.2
+l.O
+0.8
-0.2
I
1
y
!
I
45
105
185
TIME, MINUTES
Conditions are indicated in Figure 3, except for Fin. 4 addition of calcium to incubation medium. Ordinate represents change (ACa) in calcium within the ghosts expressed as @quiv./L. per lo3 ghosts/mm3. Curve A represents normal ghosts without PGE2 added, and curve B normal ghosts with the addition of PGE2 (2 x 10m7M). Curve C represents sickle cell ghosts without PGE2 added, and curve D in the presence of 2 x lo-TM PGE2. Curves A and B represent the mean values from five normal plus four AS ghost populations. Curves C and D represent the mean values from seven experiments with sickle cell ghosts. Bars indicate estimated standard deviations. Trait and normal ghost values averaged together for reasons given in legend to Figure 3.
552
APRIL 1975 VOL. 9 NO. 4
PROSTAGLANDINS
DISCUSSION The SS erythrocyte membrane displays at least two different cation flux reactivities compared with normal membranes; an altered reactivity to excess sodium at the inner surface of the membrane, and an increased calcium influx under simulated sickling conditions. Gordesky et al. (3) have shown that, although total phospholipid concentration may be the same in normal and SS erythrocyte membranes, the structural arrangements are apparently different. As Whittsm has shown that the Na+-* activated ATPase is asymmetric with respect to sodium and potassium stimulation (19), the comparative sodium flux patterns observed here may reflect the response to sodium inhibition of the enzyme in two different environments. The high internal and nonphysiological sodium concentrations may also explain the reversed sodium flux reported here, compared to that reported by Tosteson (2) using intact cells. The calcium response is maximal at pO2 values approaching venous pO2 (Figures 2,4) where sickling of intact cells would be in excess of 70%, but is partially reversed at even lower oxygen concentrations and is stimuiated by-PGE2, A possible role for-PGE2 in the _in vivo sickling cycle would then be to increase calcium influx under stasis or hypoxic-conditions, thus, increasing membrane rigidity and accelerating the sickling of more cells. The relation of Ca* and the Ca*/ ATP ratio to cell deformability is complex (cf:(lO)) and it remains to be demonstrated whether transitory movement of Cak into the membrane would be sufficient to account for deformability effects in a scheme involving contractile membrane proteins such as spectrin, or whether there must be, in addition, consideration of the several glycolytic enzymes whose activities calcium is known to affect (1). In this regard, it has been shown that the shift to the right of the oxygen dissociation curve of hemoglobin SS within intact erythrocytes in the presence of PGE2 is accompanied by an increase in methemoglobin, suggesting a metabolic impairment of reduced pyridine nucleotide production necessary for reduction of methemoglobin (15).
APRIL
1975
VOL. 9 NO. 4
553
PROSTAGLANDINS
RRFFBENCES 1. Bygrave, F. L. The Ionic Environmentand Metabolic Control, Nature 214:667, 1967. 2. Gardner, J. D. and E. R. Ginzler. Sodium Transportin Human Erythrocytes- Absence of an Effect of ProstaglandinEl, Biochem. Biophy. Res. Comn. 42:1063, 1971. 3. Gordesky,S. E., G. V. Marinetti,and C. B. Segel. Difference in the Reactivityof Phospholipidswith FDNB in Normal RBC, Sickle Cells and RBC Ghosts, Biochem. Biophys. Res. Comm. 47: 1004, 1972. 4. Hanahan, D. J. and J. Ekholm. Changes in ErythrocyteMembranes During Preparation,as Expressedby ATPase Activity, Biochim. Biophys.Acta 255:413, 1972. Ham, and W. B. Castle. 5. Harris, J. W., H. A. Brewster,T. l-l. Studies on the Destructionof Red Blood Cells. X. The Biophysics and Biology of Sickle Cell Disease, Arch. Int. Med. 97:145, 1956. 6. Hoffman, J. L. The Active Transportof Sodium by Ghosts of Human Red Blood Cells, J. Gen. Physiol. 45:837, 1962. 7. Jensen, M, S. B. Shohet, and D. G. Nathan. The Role of Red Cell Energy Metabolism in the Generationof IrreversiblySickled Cells In Vitro, Blood 42:835, 1974. a. Johnson,M., I. Rabinowitz,A. L. Willis, and P. L. Wolf. Detection of ProstaglandinInductionof ErythrocyteSickling, Clin. Chem. 19: 23, 1973. 9. Kirtland,S. J. and H. Baum. ProstaglandinEl May Act as a "Calcium Ionophore", Nature New Biol. 236:47, 1972. 10. LaCelle, P. L. and R. I. Weed. The Contributionof Normal and PathologicErythrocytesto Blood Rheology, Progress in Hematology, Vol. VII, Grune & Stratton,New York, 1971. 11. Lepke, S. and H. Passow. Effects of Fluoride on Potassiumand Sodium Permeabilityof the ErythrocyteMembrane, J. Gen. Physiol. 51: 3658, 1968. 12. Lepke, S. and H. Passow. The Effect of pH at Hemolysison the Reconstitutionof Low Cation Permeabilityin Human Erythrocyte Ghosts, Biochim.Biophys.Acta 255:696, 1972. 13. Mahin, D. T. and R. T. Lofbery. A SimplifiedMethod of Sample Preparationfor Determinationof Tritium, Carbon-14,or Sulfur-35 in Blood or Tissue by Liquid ScintillationCounting, Anal. Biochem. 16:500, 1966.
554
APRIL
1975
VOL. 9 NO. 4
PROSTAGLANDINS
I. N., P. L. Wolf,and SBerman. Light Scattering 14. Rabinowitz, Studies of Retardationof Sickling by Aspirin-likeDrugs, Res. Coann.Chem. Pathol. and Pharmacol. 8:417, 1974. 15. Rabinowitz,I. N., P. L. Wolf, N. Shikuma, and S. Berman. ProstaglandinE2 Effects on Oxygen Affinity by Sickle Erythrocytes, Prostaglandins7:309, 1974. 16. Teorell, T. PermeabilityPropertiesof ErythrocyteGhosts, J. Gen. Physiol. 35:669, 1952. 17. Van Kampen, E. J. and W. G. Zijlstra. Standardizationof Method, Clin. Hemoglobinometry. II. The Hemoglobincyanide Chim. Acta 6:538, 1961. S. Swofford,C. E. Kolpin, and 18. Eaton, J. W., T. D. Skelton,I-I. H. S. Jacob. Elevated ErythrocyteCalcium in Sickle Cell Disease, Nature 246:105, 1973. 19.
Whittam, R. and M. E. Ager. Dual Effects of Sodium Ions on an Erythrocyte-Membrane Adenosine Triphosphatase, Biochim. Biophys. Acta 65:383, 1962.
20.
Tosteson,D. C. The Effects of Sickling on Ion Transport. II. Effect of Sickling on Sodium and Cesium Transport, J. Gen. Physiol. 39:55, 1955.
21. Wilbrandt,W. OsmotischeNatur SogenannterNichtosomotische Htlmolysen(Kolloid-Osmotische HYmolyse), Pflug. Arch. Ges. Physiol. 245:22, 1941. 22. Allen, J. E. and C. R. Valeri. Prostaglandinsin Hematology, Arch. Intern.Med. 133:86, 1974. 23. Hensby, C. N. Reductionof ProstaglandinE to Prostaglandin F2c by an Enzyme in Sheep Blood, Brit. J. Pharmacol.50:462P, 1974. 24. Smith, J. B., J. J. Kocsis, C. Ingerman,and M. J. Silver. Inactivationof ProstaglandinAl and ProstaglandinA2 by Human Red Cells, Pharmacologist15:208 (abstr.288), 1973. 25. Shaw, J., W. Gibson, S. Jessup, and P. Ramwell. The Effect of PGEl on Cyclic AMP and Ion Movements in Turkey Erythrocytes, Ann. N.Y. Acad. Sci. 180:241, 1971.
APRIL
1975
VOL. 9 NO. 4
555
