increased cell membrane in pathogenesis ...€¦ · c14-glucose was added to each 1 ml of 35% cell...

Journal of Clinical InvestigationVol. 43, No. 8, 1964

Increased Cell Membrane Permeability in the Pathogenesisof Hereditary Spherocytosis *

HARRYS. JACOB ANDJAMES H. JANDL WITH THE TECHNICAL ASSISTANCE OFSUSANC. BELL AND NANCYEM. FILES

(From the Thorndike Memorial Laboratory and Second and Fonrtlh [Harzvardl Medical Se'r'-ices, Boston City Hospital, and the Department of AMedicine, HarzVard Mledical

School, Boston, lass.)

The hemolytic process in hereditary spherocy-tosis (HS) is presumed to involve an intrinsic de-fect in the structure or metabolism of the red cell(1-3). The nature of the basic cellular defectremains enigmatic, although its manifestations arewell characterized. Distinctive features of theHS red cell are its more spheroidal shape, smallsurface area, and high hemoglobin concentrationand its exceptional proclivity toward splenic se-questration. However, the most singular featureof these cells is their abnormally rapid spheroidalchange and increase in osmotic fragility on sterileincubation in vitro (4-6). This characteristicresponse of HS red cells has provided the mostreliable method for diagnosis of the disorder (7).

Efforts to identify a presumed biochemical le-sion in HS red cells have had little success. Areported diminution in turnover of organic phos-phate esters (8) has not been confirmed (9, 10),nor have abnormalities been found in the concen-trations or kinetics of glycolytic intermediates orin glucose consumption of HS red cells (6, 11,12). More recently, evidence has been presented(13, 14), but later denied (15, 16), that an ab-normality exists in the phospholipids of the redcell membrane in this disease. With incubation,increased loss in lipids from spherocytic mem-branes has also been reported (17, 18).

The general presumption in studies of HS redcells has been that a deficit existed in their energymetabolism. This presumption is difficult to rec-oncile with the findings by Harris and Prankerd(19) and Bertles (20) that the sodium efflux andinflux, respectively, across HS cell membranes are

* Submitted for publication December 9, 1963; acceptedApril 23, 1964.

Supported by U. S. Public Health Service researchgrant HE-07652-01, National Heart Institute, Bethesda,Md.

actually increased, since increased sodium trans-port should be reflected by heightened, rather thandiminished, energy metabolism (21-23).

In view of the inconsistencies noted above, thepresent studies were undertaken to attempt tocharacterize anew the over-all metabolic behaviorof HS red cells and to relate changes in their me-tabolism on erythrostasis in vitro to changes intheir viability in vivo. A preliminary report ofthese findings has been published (24).

Methods

Eight patients, all of whom had been splenectomizedin the past, served as donors of HS red cells. The pa-tients, from five families, demonstrated normal routineblood counts (25), including reticulocyte percentages, atthe time of the studies. The following characteristics,considered typical of the disease, were present in everydonor: a) a congenital hemolytic anemia affecting atleast two family members; b) complete clinical remissionfollowing splenectomy; c) spherocytes demonstrable onperipheral smear with associated increased osmotic fragili-ties of fresh and of incubated blood; and d) abnormaldegrees of autohemolysis following prolonged incubation,partially corrected by adding glucose.

Fresh blood, drawn into heparin, was centrifuged at3,000 rpm, the buffy coat was removed, and the cellswere resuspended in an incubation medium renderedsterile by Seitz filtration. After recentrifugation and asecond removal of any remaining buffy coat, resuspensionin the incubation medium to an approximate 35% cellsuspension was effected. Leukocytes in these washed sus-pensions numbered less than 200 per mm3. The incubationmedium, unless otherwise stated, was a buffered 5%o dia-lyzed human serum albumin 1 solution containing the fol-

1 Normal human serum albumin was obtained fromMassachusetts Public Health Biologic Laboratories as a25% solution. After dialysis against appropriate bufferthe solution was diluted to a final concentration of 5%in buffer. Previous studies (26, 27), especially evidencerecently presented (28), indicate that plasma when in-cubated undergoes alterations in its lipid moieties that

1704

MEMBRANEPERMEABILITY IN HEREDITARYSPHEROCYTOSIS

lowing constituents: K = 6.5 mEq per L, Na = 165 mEqper L, P04 =40 mmoles per L, HCO3=25 mEq per L,and Cl = balance of anions. The medium was isosmolalwith normal plasma by freezing point depression deter-mination. When pertinent, additives to the medium werepresent in the following concentrations: glucose, 0.022 M;ouabain,2 3.5 X 10-' M; sucrose, 0.08 M.

The cell suspension was equilibrated to a pH of ap-proximately 7.4 with a 95% oxygen and 5% carbon di-oxide mixture before incubation. After removal of asample of cells for initial metabolic determinations, theremaining cell suspension was further divided and incu-bated at 370 C in a number of stoppered vessels withoutshaking unless otherwise noted. At various intervalsthereafter one sample was labeled for 30 minutes at 250C with 150 gc of Cr"-labeled sodium chromate.3 Afterfree chromate was removed with a single centrifugation,the cells were suspended in normal saline and were theninjected into serologically compatible normal subjects.Red cell survival and the sites of sequestration werestudied as described previously (29). At the time of in-jection and at intervals thereafter additional samples ofcells were removed from the separate incubation flasksfor metabolic studies. In all experiments the metabolismand survival of HS red cells were compared to those ofred cells from normal subjects incubated at the same timeunder identical conditions.

Duplicate hematocrits determined in Wintrobe tubeswere in agreement to within ± 1%. Ancillary studieswith P`31-labeled albumin demonstrated no significantvariation in plasma trapping during these experiments,as noted previously by others (30). Therefore, no cor-rection for plasma trapping was made. The pH of redcell suspensions was measured in a Beckman model GpH meter at various intervals during incubation. Redcell lactate production by the lactate dehydrogenase4method (31) and glucose consumption by the Somogyi,Nelson method (32) were determined in duplicate as-says from duplicate incubation flasks agitated for 4 hoursat 370 C. Cellular Na gain and K loss during the in-cubation were calculated as described previously (33)from measurements of extracellular and whole-bloodcation levels obtained by internally standardized flamephotometry. The cellular cation concentrations were

produce changes in the shape and fragility of red cells.Albumin media, containing approximately 10-' M pro-tein-bound fatty acid, were used in these studies in orderto circumvent variables that might be associated withsuch plasma alterations.

2 Ouabain U.S.P., obtained from Eli Lilly Company,Indianapolis, Ind., was diluted 10-fold in incubation me-dium before use.

3Available as Rachromate from Abbott Laboratories,North Chicago, Ill.

4Assay reagents produced by C. F. Boehringer andobtained in kit form from California Corp. for Bio-chemical Research, Los Angeles, Calif.

related to the original volume of cells. In control studiesthese values agreed closely with those measured in he-molysates from cell suspensions washed with isotonicMgCI2 (0.1 M). The method of Grunert and Phillips(34), as modified by Beutler (35), was used to assayglutathione (GSH) content in red cells. Red cell osmoticfragility was measured by the method of Emerson andhis associates (4); in incubation mixtures containingsucrose, however, an initial 1: 100 dilution of blood inthe hypotonic salt solutions was used to render negligiblethe osmotic activity contributed by this substance.

Measurements of radioactive C02 derived from cellsuspensions incubated with C14-glucose 5 were made by theuse of incubation bottles containing removable plasticwells of Hyamine, as described by Goodner and Freinkel(36). The incubation was begun with the addition oftracer amounts of the labeled glucose to cells suspendedin the usual 5%7 buffered albumin medium containing 150mg per 100 ml carrier glucose. Approximately 0.1 uc ofC14-glucose was added to each 1 ml of 35% cell suspen-sion. After agitation in a 370 C water bath for 4 hours,0.2 ml of 1 N H2SO4 was added to the red cell suspen-sion through the rubber diaphragmatic top of the incu-bation bottle. After its evolution and subsequent trappingin the plastic well containing Hyamine, C1402 was countedin a methanol-toluene solution of 2,5-diphenyloxazole-p-bis 2- (5-phenyloxazolyl) -benzene 6 in a Packard Tri-Carbliquid scintillation counter with toluene-C"4 as an internalstandard. From a knowledge of the counts of C14-glu-cose added, the total glucose consumption, and the C402counts produced, the percentage of utilized glucose metab-olized to C02 was calculated.

Influx and efflux of sodium in red cells was measuredby using tracer 7 Na". Sodium influx was measured in30% cell suspensions incubated in glucose-containingmedia. Samples of cells, removed at various times, werewashed 3 times in isotonic MgCl2 over a period of 10minutes. Radioactivity of hemolysates derived there-from was measured in a well-type scintillation counter.The radioactivity of the suspending medium was alsomeasured at each interval. Total sodium concentrationsof the hemolysate and suspending medium were obtainedby flame photometry. Calculation of Na influx fromthese data was made by utilizing the formulas derivedby Solomon (37), as modified by Bertles (20). Sodiumefflux was measured after preincubation of red cells withNaa for 2 hours in glucose-containing media. After threewashes and resuspension in a tracer-free medium, a sampleof blood suspension was removed for counting, and the re-maining blood was reincubated. The medium wassampled and counted at various intervals thereafter.Intracellular activity at each interval was calculated by

5 Obtained from New England Nuclear Corp., Bos-ton, Mass.

6iAvailable as Liquifluor, a 25x concentrated solution,from Pilot Chemicals, Inc., Watertown, Mass.

7 Available from Iso-Serve, Inc., Cambridge, Mass.

1705

HARRYS. JACOB ANDJAMES H. JANDL

NORMAL.

0=0BLOOD

v i ! I I I 1

4

2

l- _ *SPLEENa- - -0 LIVER

00

HEREDITARY SPHEROCYTOSIS

BLOOD

SPLEEN

LIVER

2 3 4 5 0 1 2 3 4

DAYS

FIG. 1. COMPARISONOF SURVIVAL (UPPER) AND SITES OF SEQUESTRATION (LOWER) OF NORMALAND HEREDITARY

SPHEROCYTIC (HS) RED BLOODCELLS AFTER GLUCOSEDEPRIVATION. After their sterile incubation at 370 C for 4 hoursin a glucose-free medium, hereditary spherocytes (right) manifest markedly decreased viability in vivo comparedto normal red cells (left). The removal of these nonviable cells is mainly by the spleen (lower right). The ex-

periment depicted is representative of three performed.

the following formula:

[Na24] = [Nall], - (1 - Hct)[Na24]pHct

where [Na24]0 = cellular Na24 activity, Counts per minuteper milliliter red cells; [Na24], = whole cell suspensionNa24 activity, counts per minute per milliliter suspension;[Na24]p = "plasma" Na24 activity, counts per minute per

milliliter medium; and Hct = hematocrit.

Results

Metabolic changes accompanying loss of cellularviability

Effect of glucose deprivation. The survival andsites of sequestration of Cr51-Iabeled I-IS and nor-

mal red cells after their incubation at 370 C with-out agitation for 4 hours in a glucose-free mediumare presented in Figure 1. After an initial phaseof rapid destruction lasting about 24 hours, ap-proximately 75%o of normal cells were viable (up-per left), whereas only 10%o of HS red cells sur-

vived this period (upper right). Coincident withthe removal of hereditary spherocytes from thecirculation a rapid accumulation of radioactivityover the spleen and a slight accumulation over theliver (lower right) occurred. Minimal increasesin organ activity were noted in the recipient re-

ceiving incubated normal red cells (lower left).To rule out unsuspected serologic incompatibility,washed but unincubated hereditary spherocytesfrom the same donor were injected at a later dateinto the same recipient as that in Figure 1. Thirtyper cent of the red cells survived the initial 24-hour period after injection.

The metabolic characteristics of the red cellsdepicted in Figure 1 were studied at the time oftheir injection (Table I). Although destined forrapid splenic removal and destruction, hereditaryspherocytes at this time were able to consume glu-cose even more rapidly than did normal cells.Associated with this increase in glycolysis, HS redcells maintained their levels of reduced GSHmore

30-

60-

40-

20.

aKwI--Mu E

z

cr:

LU.I -

zw0 1

a:a.

> XJ

LL40

Ef <

UO0 a(nE

8 Owcr

Oma.

1706

I.

1-

I ,1 I I r


TABLE I

Effect of glucose deprivation on normal and hereditaryspherocytic (HS) red cells*

HereditaryNormal spherocytosis

Glucose consumption,t 1.66 2.46iumoles/ml cells/hour

GSH, mg/100 ml cells 10 26Na+ gain, mEq/L cells 1.9 6.1K+ loss, mEq/L cells 2.9 3.0

* Cells preincubated at 370 C for 4 hours in glucose-freemedia.

t The survival of these cells in vivo is depicted in Figure 1.Their ability to consume glucose was determined during the4 hours subsequent to this incubation, as described in thetext.

effectively. Despite this "hypermetabolism," how-ever, HS red cells were unable to prevent accumu-lation of intracellular sodium ions; significant in-creases in cellular Na content occurred in HSbut not in normal red cells. Cellular K+ levelsdecreased slightly and equally in the two cell sus-pensions.

Effects of erythrostasis. Duplicate washed sus-pensions of HS red cells in glucose-containing,buffered albumin media were prepared. One ofthese was incubated at 37° C in a Dubnoff meta-bolic shaker at 20 cycles per minute; the otherwas incubated without agitation at the same tem-perature after a 5-minute centrifugation at 2,000rpm. After 3j hours both were labeled with Cr5'and injected into compatible normal recipients.Metabolic determinations were performed on sam-ples of the incubated cells at the time of injection.The results are shown in Figure 2 and Table II.HS red cells, after short periods of stagnant in-cubation (black circles, Figure 2), survived poorlyin the circulation (upper, Figure 2) with 27%oundergoing early sequestration, principally in the

TABLE II

Effect of erythrostasis* on HSred cells

Change in cell Hemolysis inCells volume 0.50% NaCl Naf gain

mEq/L cellsSuspended 0 4.4 2.7Packed +3.6 9.5 12.0

* Red cells were incubated 4 hours in glucose-rich me-dium as described in the text. The above measurementswere made at the time of injection into normal recipients,depicted in Figure 2.

100w0 90.wz= 80.ir0U. 70'0

zw 60

wO- 5C

<0

L&J0

0F

4

3

2

BLOOD

SUSPENDED

PACKED

SPLEENPACKED

SUSPENDED

-o%F-30 60

MINUTES90 120

FIG. 2. EFFECT OF CELL PACKING ON SURVIVAL (UPPER)AND SPLENIC SEQUESTRATION(LOWER) OF INCUBATED HSRED BLOOD CELLS. Compared to identical cells that arepacked by slow centrifugation (black circles), hereditaryspherocytes that are kept suspended by gentle agitation(white circles) during incubation survive better and aretrapped by the spleen more slowly (lower). The ex-periment depicted is representative of two performed.

spleen (lower, Figure 2). Cells that had beengently agitated during their incubation (whitecircles) survived better, with only 10% under-going sequestration. At the time of injection(Table II) the HS red cells that had been sub-jected to stagnation were slightly swollen, os-motically more fragile, and had increased theirintracellular sodium content considerably.

More prolonged incubation was required torender normal cells nonviable under similar con-ditions. Thus, 100%o of normal red cells incu-bated with agitation for 24 hours in glucose-richmedium survived normally (Cr5l-half survival =26 days); in contrast, only 20% of normal cellssurvived beyond the first day in vivo if previ-ously incubated for 24 hours in the packed state.At the time of injection both cell samples werecapable of consuming glucose (agitated = 0.96;spun = 0.24 Mmoles per ml per hour), but onlythe packed, poorly viable cells had gained intra-cellular sodium (8.8 mEq per L cells).

1707


TABLE III

Comparison of rates of glycolysis in normal and HS red cells*

pH

Glucose consumption Lactate production Normal HS

Experiment Normal HS HS/ Normalt Normal HS HS/Normalt Initial Final Initial Final

jsmoles,'ml cells/hour2.10 3.301.94 2.552.25 2.79

Mean 2.10t 3.91

2 2.312.141.96

Mean 2.14t

3 2.903.133.03

Mean 3.01t

4 3.033.09

Mean 3.06t

5 3.486 3.077 1.708 2.009 2.64

10 3.74i1 2.4212 2.5713 2.5714 2.7715 3.6016 2.7917 2.5918 3.1519 2.16

3.682.893.103.472.97

3.733.463.174.504.633.96

/Ccl01 psmoles/ml cells/hour157 3.71 5.32121 3.80 4.71133 2.96 3.94186 Mean 3.49t 5.18

172135145162139

124115105149154131

4.504.454.55

Mean 4.50t

5.895.375.805.455.61

3.23 1064.94 1614.25 1393.80 124

4.014.322.303.203.454.203.373.303.343.044.093.423.123.703.02

115141135160131112139128131110114122120117140

1.53 1.84 1201.32 2.10 1601.55 1.79 116

134 i 3.3

7.77 7.70 7.76 7.707.61 7.637.83 7.49 7.82 7.32

131+ 4.4 pH drop = 0.12 i 0.02pH drop = 0.18 i 0.02

* HS red cells from eight splenectomized patients and normal red cells were washed and simultaneously incubatedwith agitation for 4 hours at 370 C in the same medium. The medium in experiments 1 to 19 was 5%human albuminas a buffered solution confining .040 MP04 and .022 Mglucose (see Methods). Incubations in experiments 20 to 22 were

performed in fresh compatible serum containing .022 Madded glucose.t Experiments 1 to 4 involved two or more normal control blood specimens. Therefore the glycolytic rates of HS

red cells were compared to the mean rate of all normal cells incubated on the same day.

Glvycolytic behavior of hereditarry spherocAtesTo more fully document the increase in glyco-

lytic rate of hereditary spherocytes, as noted above,the consumption of glucose and formation of lac-tate by fresh HS red cells from eight splenecto-mized patients were compared in several experi-

nients with those of red cells f rom numerous

normal subjects. Both cell types were washed andincubated in the same medium under nearly identi-cal conditions of pH and temperature. The resultsof these studies are shown in Table III. Indi-vidual values of glucose consumption and lactate

153135113149

131120129121125

7.80 7.617.76 7.627.77 7.60

7.477.477.47

7.81 7.437.81 7.467.80 7.487.80 7.42

7.477.477.477.477.47

7.57 7.457.54 7.437.58 7.48

7.60 7.527.60 7.57

7.527.567.507.577.517.51

7.437.397.347.397.387.38

7.60 7.527.60 7.377.60 7.507.60 7.51

7.607.587.487.427.407.507.487.507.517.607.627.457.357.627.29

7.517.51

7.357.397.42

7.387.53

7.507.23

202122

Mean i SE

7.607.587.487.477.407.507.477.507.517.607.627.447.357.547.29

7.497.41

7.277.407.38

7.487.46

7.437.17

1708


TABLE IV

Effect of phosphate concentration on glycolysis innormal and HS red cells*

Glucose consumption

1P04] = 1.5 [P04] =40 StimulationRed cells mM mM by high P04

,umoles/ml cells/hour %1 1.32 2.79 2112 1.58 3.07 1943 1.89 2.57 136

Normal 4 1.92 2.68 1405 1.42 2.53 1786 1.15 2.42 2107 1.47 2.83 1918 1.55 2.83 182

Mean ± SE 180 4 10

1 2.10 3.42 1632 1.55 3.46 223

Hereditary 3 2.06 3.30 160spherocytosis 4 2.06 3.30 160

5 1.85 3.36 1826 1.70 3.62 2137 1.79 3.62 203

Mean :1 SE 186 i 10

* This tabulates a series of experiments in which pairedsamples of red cells from normal (4) and HS subjects (2)were incubated with agitation for 4 hours at 370 C inbuffered albumin media (see Methods) of identical pH andcomposition except for the PO, concentration.

production in IIS red cells are compared to thosein the simultaneously incubated normal cells.Hereditary spherocytes manifested a mean in-crease in glucose consumption of 34%o and inlactate production of 31% over normal. Withoutexception, in 22 consecutive experiments, the gly-colytic rate of HS red cells exceeded that of simul-taneously incubated normal cells, as determined byboth glucose consumption and lactate production(p <.001). Reflecting the increased lactate pro-duction in HS red cells, the average fall in pHduring the 4-hour incubation period was slightlygreater (p <.05) in HS red cells (- 0.18 pH

30

25

20

15.

10-

5.

3)-J

-j

w

0

-j

0r

E

0

Na GAIN

o

0

0

88

s

0

30

25- K LOSS

10

5 8i

,J . __ . ._.

4 8 12 16 20 24HOURS

FIG. 3. COMPARISONOF THE EFFECT OF INCUBATION ON

THE INTRACELLULAR NA AND K LEVELS OF HS AND NOR-

MAL RED BLOOD CELLS. Upon incubation in a glucose-containing medium, accumulation of intracellular Na(upper) occurs more rapidly and to a greater degree inhereditary spherocytes (white circles) than in normalred cells incubated simultaneously (black circles). Incontrast, K losses (lower) from both cell types are simi-lar. Each circle represents data from one experiment.

unit) than in normal cells (- 0.12 pH unit), as

shown in Table III.In most of the above studies the concentration

of phosphate in the incubation media was 40 mM,

an amount found by others (38) to promote maxi-mal glycolysis in normal red cells in vitro. At a

TABLE V

Activity of hexose monophosphate (HMP) shunt in normal and HS red cells

No. of Total glucose Conversion Glucose metabolizedCells incubations consumption glucose-I-C14 to C1402 via HMPshunt*

pmoles/ml cells/hour % pg/ml cells/hourNormal 4 2.55 ± 0.10 (SD) 4.29 i 0.07 19.7 0. 9

Up < 0.01Hereditary spherocytosis 4 3.33 ± 0.07 2.86 ± 0.15 17.1 i 1. <

* This value is a minimal approximation of HMPshunt activity, ignoring the contribution by recycling of carbonfragments other than C' through this pathway.

1709

0HARRYS. JACOB AND JAMES H. JANDI,

70-

E 60

t40

NORMAL

0

HEREDITARYSPHEROCYTOSIS

00

lLU

t^ 22

z

-J

HOURS

FIG. 4. COMPARISONOF NA EFFLUX FROMNORMALRED

CELLS (BLACK CIRCLES) AND HEREDITARY SPHEROCYTES

(WHITE CIRCLES). The slopes of the lines derived by themethod of least squares indicate that the rate of tracerNan extrusion from hereditary spherocytes exceeded thatfrom normal red cells by 34%. Each circle and bracketrepresents the mean + 1 SD of data from five experi-ments. Red cells from two HS and three normal donorsare represented.

more physiologic concentration of P04 (1.5 mM)the dichotomy in glycolytic rates between normaland HS red cells remained apparent. As shownin Table IV, glucose consumption by both celltypes was increased by slightly more than 75% in40 mMphosphate as compared with 1.5 mMphos-phate. In addition, glycolysis in HS cells was in-creased above normal in fresh autologous serum

(experiments 20 to 22, Table III) as well as inartificial media of pH levels varying from 6.6 to8.0 (39).

As shown in Table V, the increased over-allrate of glycolysis in hereditary spherocytes is as-

sociated with a significant decrease in both thepercentage and absolute quantity of glucose me-

tabolized by the hexose monophosphate (HMP)shunt. Accordingly, the hypermetabolism of HSred cells is entirely through stimulation of theEmbden-Meyerhof pathway.

Electrolyte metabolism in hereditary sphero-cytes. The decreased survival of incubated

hereditary spherocytes in vivo was noted above tocorrelate with their propensity to accumulate in-tracellular sodium. This was further quantitatedas shown in Figure 3. Gains in intracellular Naby incubated HS red cells (upper, Figure 3; whitecircles) were substantially greater than in normalcells (upper, Figure 3; black circles) at all timeintervals studied. In contrast the loss of K fromboth types of cells was equal (lower, Figure 3).

As shown in Figure 4, the hereditary sphero-cyte, despite its propensity to accumulate intra-cellular Na, was found to actually transfer Nafrom intracellular to extracellular compartmentsat a more rapid rate than normal, confirming thefindings of Harris and Prankerd (19). The ini-tial intracellular Na concentrations in the two celltypes were essentially equal. However, the calcu-lated efflux of Na from normal cells was 3.7 mEqper L cells per hour, whereas that from hereditaryspherocytes was 5.0 mEqper L cells per hour.8

Since efflux of sodium from spherocytes wasactually increased, it follows that sodium accumu-lation by these cells must be secondary to an in-creased leak of sodium into the cells. Evidencefor an abnormal sodium permeability, suggestedpreviously by the tracer Na studies of Bertles(20), was also observed in the present studies.In our hands, in two experiments, the calculatedinflux of sodium into HS cells was 4.6 and 4.8mEqper L, cells per hour, whereas in normal cells,incubated in parallel, sodium influx was 2.3 and2.6 mEqper L cells per hour.9

8 Sodium efflux was determined by multiplying the frac-tional decrease in cellular Na"4 activity per hour by thetotal intracellular sodium concentration. This calcula-tion ignores the back flux of isotope from extracellularto intracellular space, which results in a slight underesti-mation of efflux (40). As evidenced by the acceptablestraight-line fit of the data in Figure 4, this error isnegligible in the short-term experiments depicted.

9 The discrepancy between the measured Na influx andefflux of normal red cells is possibly explained by thefact that cells from different normal donors were usedin these determinations. Also it is acknowledged thatthe calculated sodium fluxes in these studies are sub-ject to the following errors: a) Upon incubation the in-creased lactate production by HS red cells tends to lowerpH slightly more than in normal cells (Table III).This, in turn, might result in a decreased, and henceunderestimation of, Na flux across HS cell membranes(37, 41). b) The calculation of Na influx involves theassumption of a "steady state" of constant cellular so-

1710


The relation of increased sodium transport to in-creased glycolysis in hereditary spherocytes

Effect of cardiac glycosides. An increase inactive transport of sodium out of the cell, neces-

sitated by an increased passive leak of this cationinto the cell, should require heightened energy

expenditure. The energy for active cation trans-port is derived from ATP (42-45), the levels ofwhich are maintained by anaerobic glycolysis (21-23). An ATPase enzymatic reaction (46-48),which may consist of a series of reactions (49,50), facilitates the release of energy from highenergy phosphates for active cation transport;this reaction is partially inhibited by cardiac gly-cosides such as ouabain (51, 47, 48). That thehypermetabolism of the HS red cell is related toits increased sodium turnover is shown in Table

TABLE VI

The effect of ouabain on the glucose consumption ofnormal and HS red cells*

Glucose consumptionOuabain effect

No Ouabain on glucoseRed cells additive 3.5 X10-5 M consumption

pmoles/ml cells/hour %change1 3.03 3.18 + 5.02 3.09 3.31 + 7.13 3.66 3.52 - 3.84 3.78 4.12 + 9.0

Normal 5 3.04 3.16 + 3.96 3.34 3.17 - 5.07 2.31 2.37 + 2.58 2.14 2.47 +15.49 2.42 2.49 + 2.9

Mean + SE + 4.1 2.1(p > 0.50)

1 3.23 2.50 -22.62 4.94 3.38 -31.63 4.25 3.29 -22.64 3.80 3.42 -10.0

Hereditary 5 3.68 2.84 -22.8spherocytosis 6 5.03 3.44 -31.6

7 3.10 3.32 + 7.08 3.47 2.80 -19.39 2.97 2.59 -12.7

Mean SE - 18.5% 4.0(p < 0.01)

* Two samples of each red cell suspension from sevenHSand six normal donors were incubated in parallel withagitation for 4 hours at 370 C. The suspending mediumcontained 5%buffered albumin and 0.022 Mglucose (seeMethods) with or without added ouabain. The initial pHof both samples was identical.

dium concentration during analysis (37, 20). The slightgains in Na content in HS red cells over short periodsof incubation (Figure 3) indicate that this assumptionmay not be strictly true.

TABLE VII

Effect of osmotic swelling on glycolysis of normal red cells*

Glucose consumption

Tonicity No addition Ouabain 3.5 X10-5 M

g/100 ml NaCI pmoles/101" cells/hour0.89 (isotonic) 1.76 1.750.74 1.90 1.580.66 2.12 1.660.61 2.02 1.800.58 2.19 1.54

* Two samples of a washed red cell suspension were in-cubated in parallel with agitation for 4 hours at 370 C.The suspending media with or without added ouabainwere identical in P04 and glucose concentration and pH.The experiment depicted is representative of two experi-ments performed.

VI. Whereas the addition of ouabain to normalred cells produced no significant effect on glycoly-sis (p >.50), its addition to HS red cells led to asignificant decrease in glycolytic rate (p < 0.01).The average decrease of 18.5%o in the nine ex-periments depicted is more than half the meandifference in glycolytic rates between HS and nor-mal red cells (Table III). Indeed, the highly sig-nificant difference in glycolytic rate of untreatednormal and HS red cells became insignificant dur-ing incubation with ouabain (p >.40).

Normal fresh red cells, whose glucose consump-tion was unaffected by ouabain, could be alteredin vitro so as to become ouabain-sensitive. Sus-pension of red cells in hypotonic media producesosmotic swelling and increased membrane perme-ability to a variety of substances (52). Freshnormal red cells were suspended in several non-hemolytic, hypotonic media, and their glucose con-sumption was measured over the ensuing 4 hoursin the presence or absence of ouabain. As seen inTable VII these osmotically swollen cells mani-fested an increased glycolytic rate compared to thesame cells in isotonic medium, and this accelera-tion of glycolysis was blocked by ouabain.

As seen in Figure 5, sodium levels in normalcells (upper left) are not altered by ouabain dur-ing the first 4 hours of incubation in vitro. Inaddition, ouabain failed to decrease the glycolyticrate of fresh normal red cells (Table VI). Incontrast, normal cells injured by incubation for24 hours manifested further increase in intracel-lular sodium concentration when exposed toouabain (Figure 5). In these cells, glucose con-

171 1

HARRYS. JACOB AND JAMES H. JANDL,

sumption was inhibited some 10% by the glycoside, incubated for short periods, gave evidence ofa finding similar to that of Murphy (41). Heredi- utilization of a ouabain-inhibitable sodium trans-tary spherocytes (right, Figure 5), even when port system. In a 4-hour period of incuba-

NORMAL

OUABAIN Oe-ONO OUABAIN *-v

HEREDITARY1 SPHEROCYTOSIS

Na GAIN

IK LOSS

I1

ICHANGEIN CELL VOLUME

16 24 0 EHOURS

FIG. 5. COMPARISONOF THE EFFECT OF OUABAIN ON THE SIZE AND

CATION LEVELS OF INCUBATED NORMALAND HS RED CELLS. Hereditaryspherocytes accumulate larger amounts of Na (right upper) when in-cubated for 4 hours in the presence of ouabain (white circles) thanwhen incubated without ouabain (black circles). In constrast, normalred cells are unaffected by ouabain during this period (left upper).Only with more prolonged incubation of normal cells does ouabainproduce a significantly increased Na gain relative to glycoside-freecontrol cells; this increment is less than in HS red cells at all timeintervals studied. The effect of ouabain on cellular K+ levels (middle)is rapid and identical in both cell types. Cellular swelling (lower)during incubation under these conditions was greater in hereditaryspherocytes than in normal red cells; ouabain magnifies this differ-ence. Incubations were performed in media containing glucose. Theexperiment depicted is representative of five performed.

407

30-

2Co-J.

-Jw(LI 10-

co o4w

a:Ad 30-

wE 201

)f ,C

zw

a.

171;2

K(


tion, intracellular sodium levels increased 3-foldin HS red cells exposed to ouabain (top right,Figure 3). This imposed inhibition of sodiumtransport was associated with a significant de-crease in the glycolytic activity of these cells asseen in Table VI. Further evidence that HS cellsutilize a ouabain-sensitive ATPase system to agreater degree than do normal cells was obtained bycomparing Na24 effluxes in the presence and ab-sence of the glycoside. In normal red cells ouabaininhibited the efflux of 1.8 mEqNa per L cells per

4O0

30~

er'.

LA.)JJw0 A

LL.0CC

w7-

wa.

wEz7z!0

hour, whereas in HS cells, incubated in parallel,ouabain inhibited the efflux of 3.1 mEqNa per Lcells per hour (not shown). As shown in themiddle portion of Figure 5, the effect of ouabainoln K+ loss was identical in both cell types at alltime intervals studied. Red cell swelling (lower,Figure 5), as measured by hematocrit change,paralleled intracellular sodium gain. Thus, dur-ing incubation for 24 hours untreated hereditaryspherocytes (black circles, right lower) increasedtheir volume 40%o more than did normal cells

GLUCOSEMEDIUM

01 NO SUCROSE

1O SUCROSE

NOSUCROSE

)SUCROSE

30. OUABAIN+ GLUCOSEMEDIUM NO SUCROSE

.0- SUCROSE

I0

0 -

.- .- . .

0 4 8 12HOURS

FIG. 6. EFFECT OF EXTRACELLULAR SUCROSE ON NA ACCUMULATIONDURING INCUBATION OF HS RED CELLS. Hereditary spherocytes incubatedfor short periods (4 hours) in a glucose-replete medium (upper) ac-cumulate significant quantities of intracellular Na. Addition of 80 mMsucrose to the suspending medium prevents this accumulation. The pro-tective effect of- sucrose continues over more prolonged periods and isapparent in red cells whose Na pumping mechanism is inhibited as bytheir suspension in a glucose-free medium (middle) or in a glucose me-dium containing 3.5 X 10 M ouabain (lower). The experiment de-picted in the upper portion is representative of ten performed. Themiddle and lower portions depict single experiments.

1713

r_

z

2

I

16 20 24


20-

16-

12-

8-

4-

0 24 48HOURS

FIG. 7. EFFECTS OF EXTRACELLULAR SUCICELL VOLUMEAND AUTOHEMOLYSISOF INCUBCELLS. Over the first 24 hours hereditaryincubated in glucose-free medium swell appraddition of 80 mMsucrose to the medium (upper) inhibits this process throughout theriod of incubation. In addition, by inhibswelling, sucrose prevents the abnormal aut4HS red cells that are incubated for prok

(lower). The levels of autohemolysis ofsuspended in a sucrose medium fall withinrange (not shown). The experiment depi(sentative of two performed.

(black circles, left lower). Ouabainthis difference and after more prolonge(produced increased autohemolysis inwell (not shown).

Effect of sucrose. Reasoning frcthat the permeability of red cells is ihypotonic solutions, an attempt was r

crease the permeability of hereditaryto Na by suspending them in the usua

medium rendered hypertonic by the admMsucrose. This substance is impelosmotically active. The effect of suchon the sodium levels of HS red cellsFigure 6. In actively metabolizing(top portion), sucrose addition (wi

completely prevented the usual sodium gain ob-served over the first 4 hours of incubation. Atthe end of 24 hours, the sucrose-treated red cellshad gained only one-half the quantity of sodium

ItNOSUCROSE as did untreated cells. Normal cells were similarlyprotected from sodium accumulation during pro-

longed incubation (not shown). To elucidate themechanism of this protective effect, spherocyteswere suspended in either glucose-free (middle)or ouabainized media (lower) to inhibit active so-

SUCROSE dium efflux from the incubated red cells. In bothmedia sucrose inhibited cellular sodium gain.

NO SUCROSE These results support the conclusion that this sub-stance produces a diminished influx of sodium intothese red cells, either by diminishing membranepermeability per se or by decreasing the gradientof Na across the red cell membrane. Coincidentwith its inhibition of cellular sodium accumulation,

SUCROSE sucrose diminished markedly the swelling and au-

tohemolysis of HS red cells incubated for pro-

longed periods in the absence of glucose (Figure7). Thus, the presence of sucrose in the incuba-tion medium reduced swelling of HS red cells dur-

ROSEHming the first 24 hours by 10-fold (upper). The3ATEHS RED

hematocrit change recorded with sucrose in Figureyeciably. The 7 compares cell size immediately after sucrose(white circles, addition with that following incubation. The cell> 48-hour pe- shrinkage that ordinarily occurs between 24 andting osmotic 48 hours of incubation, as noted previously by

)nged periods Selwyn and Dacie (6), occurred at a slightlyHS red cells greater rate with added sucrose. Correlated withn the normal its inhibition of cellular swelling, sucrose decreasedacted is repre- the autohemolysis, characteristic of HS cells, by

two-thirds-to within the normal range (lower,

accentuated Figure 7).

d incubationTABLE VIII

HS cells asEffect of sucrose on the metabolism of HS red cells*

)m the factincreased inmade to de-spherocytes

,1 incubationEdition of 80rmeable, but1 suspensionis shown insplierocytes

iite circles)

Medium

No sucrose Sucroset

Glucose consumption, 3.45 3.89Atmoles/ml cells/hour

GSH, mg/100 ml cells 56 55Na gain, mEq/L cells 4.9 0K loss, mEq/L cells 0.7 3.1

* Cells were incubated at 370 C for 4 hours in glucose-containing media either with or without sucrose added to afinal 80 mMconcentration. The survival of these cellsin vivo is depicted in Figure 8.

t All values are expressed per original volume of redcells before their suspension in sucrose.

1714


In addition to its protective effects in vitro, su-crose was found to maintain the viability of incu-bated spherocytes as determined subsequently invivo. As shown in the top portion of Figure 8,HS red cells incubated for 4 hours in vitro sur-vived longer if incubated in the presence of su-crose (white circles) than in its absence (blackcircles). Splenic sequestration of these treatedcells (lower) was correspondingly decreased.Measurements of metabolic activity of these cellsat the time of injection are given in Table VIII.Most significant was the correlation between en-hanced survival of sucrose-suspended HS redcells (Figure 8) and the absence of sodium ac-cumulation in these cells.

Further evidence that the diminished viabilityof incubated HS red cells relates to their pro-pensity to accumulate Na is shown in Table IX.Parallel incubations of HS and normal red cellswere made in media of identical composition ex-cept for their concentrations of Na. A low-Namedium was prepared by replacing sodium chloridewith an equimolar quantity of choline chloride sothat the final extracellular Na concentration was11 mEqper L. This concentration approximatesintracellular Na levels, and thus a net movementof Na into the cells was inhibited. As shown inTable IX, for the first 24 hours of incubationswelling of both normal and HS red cells wascompletely inhibited in this medium; concomitantlyautohemolysis of HS cells after 48 hours wasdiminished.

TABLE IX

Effect of Na concentration of the suspending mediumon the swelling and autohemolysis of red blood cells

HematocritCells Medium* increase Autohemolysis

%/24 hrs %/48 hrsNormal High Na 6.0 0.23

Low Na 0 0.24Hereditary High Na 8.6 5.5

spherocytosis Low Na 0 2.5

* The red cell suspensions were incubated at 370 C inisotonic media of the following cation concentrations:low Na-[Na] = 11 mEq/L, eK] = 4.7 mEq/L, [cho-line] = 134 mM, and pH = 7.3; high Na-[Na] = 145mEq/L, [K] = 4.7 mEq/L, and pH = 7.3. Both mediacontained initially 22 mMglucose. In the absence ofglucose choline produces increased autohemolysis in bothcell types (39). The experiment depicted is representativeof two performed.

0 100-w

W 90.

cr. 80-

0

z

60.

0 .

5C-

> 4,

3--3._ct

o 22-Ji.4

c) 0 I'

owo W

43 a

BLOOD

_~~~SUCROSE

-NO SUCROSE

SPLEEN

NOSUCROSE

_ C)~-OSUCROSE

G-

0 30 60 90MINUTES

120 IS0

FIG. 8. EFFECT OF INCUBATION IN SUCROSE ON THESUBSEQUENTSURVIVAL IN VIVO AND SPLENIC REMOVALOFHS RED CELLS. The survival of hereditary spherocytesinjected after a 4-hour incubation in the presence of 80mMsucrose is significantly greater than that of identicalcells incubated without sucrose (upper). Reflecting thisenhancement in survival, removal of these cells by thespleen occurs at a slower rate (lower). The experimentdepicted is representative of two performed.

Discussion

The present studies suggest that of primary im-portance in the pathogenesis of hereditary sphero-cytosis is an inherent increase in sodium perme-ability of the red cell membrane. The total intra-cellular sodium concentration of HS red cells,sampled directly from the circulation, is normal(6). Presumably, therefore, circulating HS redcells compensate for an increased leak of sodiumby means of active pumping mechanisms. Theexpenditure of energy to support this compensa-tion must come ultimately from glucose utiliza-tion (21-23) and more specifically through an in-creased regeneration of ATP (42-45). Whenthese cells are deprived of glucose for short pe-riods in vitro, they rapidly gain sodium, and theylose viability as judged by their subsequent sur-vival in vivo. If removed from glucose for longerperiods, the intake of sodium (and water) con-tinues, leading to cellular swelling, a further gen-

i715

1HARRYS. JACOB AND JAMES H. JANDI,

eral increase in membrane permeability, anal ulti-mately to atutohemolysis. Although more strikingin the absence of glucose, these changes werefound to occur even in the presence of adequateglucose if HS red cells were allowed to undergopacking. By analogy, red cells subjected to hemo-concentration and erythrostasis in the spleen sinus-oids are probably unable to meet their increasedsodium transport requirements, thereby initiatingthe lethal sequence described above.

The relations of abniormial sodiulml perincabilitvto decreased survival of ihe reditary splicrocytes.A close correlation appears to exist between (de-creased survival of red cells in vivo andl cellularsodium accumulation in vitro. TFhus, sucrose (le-creases the Na influx of red cells (37) and in ourhands inhibited thereby the sodium gain of incu-bated hereditary spherocytes; concomitantly redcell survival in vivo was enhanced. Conversely,red cells that have been exposed to sulfhydryl ini-hibitors were found to manifest both increased ca-tion permeability (33) and an extraordinary lia-bility to splenic sequestration (53, 54). The in-tegrity of cells so altered was also preserved byimpermeable osmotic agents such as sucrose (33).

The fact that excessive autohemolysis occurs inHS cells at a time when red cell volume is dimin-ishing (24 to 48 hours) has led to the view thatgross degeneration or dissolution of the cell mem-brane underlies this phenomenon. That suspen-sion media of moderate sucrose, or low Na, con-centrations prevent the excessive autohemolysisof these cells indicates rather that this hemolysisis principally osmotic in nature. Any decrease,however, in surface area of the incubated cells, asby losses of structural lipids (17, 18) would po-tentiate the lytic effect of osmotic cell swelling.

The mechanisms by which the hereditaryspherocyte compensates for increased membranepermeability. Under optimal conditions, namelyfree suspension in a glucose-containing mediumsuch as circulating plasma, the HS red cell com-pensates well for the increase in sodium influxacross its membrane. The mechanism by whichthis compensation occurs is probably as follows:A tendency to gain intracellular sodium ion by in-creased influx serves to stimulate membrane ATP-ase (55, 56). This in turn facilitates release ofenergy stored as ATP, enhancing thereby the ac-

tixe extrusion of sodium from the cell. Possiblythe reported increased susceptibility of HS cellsto injury by fluoride (57) or by EDTA (58) re-flects the special dependence of these cells on theMg++-requiring ATPase system.

The stimulation of glucose consumption by in-creased sodium tratnsport in the hereditary sphero-cyte. In the present studies glycolytic rates ofred cells were measured over a 4-hour period dur-ing which time the pH of suspensions fell slightly.This fall in pH was generally greater in HS thanin normal red cells ( I'ahle III). A lower pEHinhibits glycolysis (38, 59, 39). Thllus increasedlactate accumulation in [IS cells must have tendedto (limilinish the increased glycolysis actually ob-served. Glycolysis in red cells is also affected byvariations in oxygen tension (59). HS red cellsconsumed glucose more rapidly than normalwhether under 95 % oxygen (experiments 1 to19, Table III) or 20% oxygen (experiments 20to 22, Table III).

Part of the hypermetabolism of HS red cells canbe shown to be intimately related to the increasedsodium turnover manifested by these cells. Thus,the inhibition of membrane ATPase, and hence thedepression of active cation transport, by incuba-tion of red cells in ouabain-containing (51) mediasignificantly reduced the consumption of glucoseby HS but not by normal red cells. With suchtreatment the glycolytic rate of HS cells no longersignificantly differed from that of normal cells.

The increment in glucose consumption necessaryto provide energy for the increased requirementsof sodium transport in HS red cells can be com-puted. The unidirectional sodium flux acrossnormal red cell membranes is roughly 3 mEq perL cells per hour, a figure reported by Solomon(37) and confirmed by Glynn (60). In HS cellssodium flux approximates 5 mEq per L cells perhour as reported by Harris and Prankerd (19)and Bertles (20) and confirmed in the presentstudies. By use of red cell ghost preparations,Sen and Post (61) as well as Glynn (56) havedemonstrated that 3 mEq of sodium is activelytransported for every milliequivalent of phosphateproduced by the ouabain-sensitive ATPase reac-tion. Since the increment in sodium transportedby HS red cells is 2 mEq per L cells per hour,an additional 2 mmole of ATP per L HS cells

I 716

MEMBRANEPERMEABILITY IN HEREDITARYSI'HEROCYTOSIS

per hour must be utilized. This increment wouldbe produced by the metabolism of 1 mmole of glu-cose over the Embden-Meyerhof pathway. Inour hands glucose consumption of normal cellsaverages 2 mmoles per L cells per hour. Anadditional 1 mmole in this consumption wouldrepresent a 16%o increase over that of normal.This figure is in agreement with the observationthat ouabain reduced glycolysis by 18%o in HScells (Table VI). The increase in glycolysis inthe hereditary spherocyte can therefore be inferredto be dependent in large part on membrane ATP-ase activity.

Only about half of the ATPase activity of thered cell membrane is sensitive to ouabain (47, 48).Previous studies of this portion have utilized redcell ghosts (47, 48), or red cells injured by pro-longed storage (47), in each case involving patho-logic accumulations of intracellular sodium. Thatouabain has a minimal effect upon fresh normalred cells, but a marked effect on HS cells, indi-cates that this ouabain-sensitive ATPase is stimu-lated in the former cells by acquired membraneinjury but in the latter cells by an inherent ab-normality in membrane function. Gabrio, Finch,and Huennekens (62) have similarly suggestedfrom unpublished data that ATPase is in a"latent state" in viable cells but becomes more ac-tive as the cell deteriorates.

The products of the reaction catalyzed byATPase, namely ADP and inorganic phosphate(Pi), have both been found to stimulate the con-sumption of glucose in subcellular (63) as well asintact cell systems (64). In addition, both havebeen proposed as regulators of glycolysis in in-tact red cells (38, 65). A direct link, therefore,is suggested between sodium-induced activation ofATPase and hypermetabolism in HS red cells.Thus the increased rate of degradation of ATPstimulated by increased sodium influx providesboth energy for increased cation transport, andby its products stimulates glycolysis as well. Thereports of abnormally rapid ATP depletion withassociated ADP and Pi accumulation in HS redcells incubated in vitro (66) or entrapped insplenic sinusoids (17) are consistent with an in-creased ATPase activity of these cells, as is thefinding (8) of increased inorganic phosphate and

decreased ATP labeling in spherocytes incubatedwith P32. The fact that the Embden-Meyerhofpathway, and not the HMPshunt, was acceleratedin HS cells (Table V) is also consistent with thisexplanation, for the Embden-Meyerhof pathwayis solely responsible for ATP regeneration in thered cell. The high hemoglobin concentration andproportionately low cellular water content of HSred cells (6) remains a puzzling and unexplainedphenomenon. Its possible relationship to en-hanced phosphorolysis of ATP in these cells iscurrently under study.

The role of the spleen. Those US red cellsthat initially escape splenic destruction are never-theless slowed in transit (67) and gradually be-come more spheroidal with repeated passagesthrough the spleen (68). This cumulative splenic"conditioning" (17) of cells repeatedly detainedand released produces a population of markedlyfragile cells destined for permanent trapping anddestruction in the spleen. When the red celltarries in the spleen sinusoids, it is presumablysubjected to a number of stresses that might im-pair metabolic activity. These metabolic stressesinclude: 1) an impaired diffusion of metabolitesand 2) competition for metabolites from sur-rounding tissue (reticuloendothelial) cells. Thefirst effect, which is mimicked by producing eryth-rostasis in vitro, produces rapid damage by de-pleting the red cell of glucose, since the red celllacks appreciable energy stores. In addition, lacticacid will accumulate while glycolysis lasts, loweringthe intracellular pH and thereby inhibiting anaero-bic glycolysis further (59, 38). Indeed, recentstudies (69) have shown that intracellular pH isabnormally low in HS cells before but not aftersplenectomy. Undoubtedly these deleterious proc-esses are greatly speeded when the red cells areclosely approximated to metabolically active tis-sue cells within stagnant sinusoids or cords.

The metabolic stresses of erythrostasis are grad-ually lethal to normal red cells. HS red cells suf-fer double jeopardy by virtue of the fact that theirincreased permeability to sodium necessitates, forcontinued homeostasis, a sustained, acceleratedglycolysis. As with the Red Queen (70), thisred cell must perforce "do all the running it cando, to keep in the same place."

1717


Summary

The effects of erythrostasis on the metabolismof normal and hereditary spherocytic (HS) redcells were correlated with effects on their viabilitywhen transfused into normal subjects. The redcells of splenectomized HS patients lost viabilityrapidly when packed or when deprived of glucose.Surprisingly, HS red cells consumed glucosefaster than normal, through acceleration of theEmbden-Meyerhof pathway, even as they lostviability.

Correlated with their rapid loss of viability, HScells accumulated more sodium than normal dur-ing incubation. Tracer Na24 influx into sphero-cytes was increased roughly two-thirds over thatin normal red cells, and extrusion of Na24 by ac-

tive transport was similarly increased. Inhibitionof active cation transport by ouabain diminishedglycolysis toward normal levels in fresh HS redcells without affecting fresh normal red cells.Contrasted to normal cells, HS red cells rapidlyaccumulated sodium and lost viability on ex-

posure to ouabain.The addition to the suspending medium of im-

permeable, osmotically active substances, such as

sucrose, diminished the influx and accumulation ofsodium by hereditary spherocytes during erythro-stasis. Osmotic swelling was thereby reduced, invivo survival prolonged, and the marked auto-hemolysis characteristic of incubated HS red cellswas prevented. Thus, the autohemolysis of HSred cells is essentially a form of osmotic lysis.

These findings indicate that the hemolysis ofHS red cells is associated with "overwork" ratherthan with a deficiency in energy metabolism. Thefollowing sequence is proposed: 1) The primarydefect of the HS red cell resides in the membrane,which is abnormally permeable to sodium; 2) theheightened influx of sodium stimulates active so-

dium extrusion from the cell by arousing a usuallylatent ATPase system; 3) the resulting increasedbreakdown of ATP provides not only energy forcation transport, but also liberates ADP and in-organic phosphate, which in turn stimulate glycoly-sis. Under optimal conditions, as in the generalcirculation, these compensatory mechanisms suffice.During metabolic stress, as is believed to occur inthe spleen, sodium accumulates intracellularly, and

irreversible changes in membrane permeability andultimately hemolysis ensue.

Acknowledgment

We gratefully acknowledge the cooperation of Dr.Gerald W. Hazard and the advice of Dr. Laurence E.Earley in these studies.

References1. Jandl, J. H. Hereditary spherocytosis in The Meta-

bolic Basis of Inherited Disease, J. B. Stanbury,J. B. Wyngaarden, and D. S. Fredrickson, Eds.New York, McGraw-Hill, 1960, p. 1015.

2. De Gruchy, G. C. Red-cell metabolism: fundamentaland clinical aspects. Aust. Ann. Med. 1960, 9, 237.

3. Carson, P. E., and A. R. Tarlov. Biochemistry ofhemolysis. Ann. Rev. Med. 1962, 13, 105.

4. Emerson, C. P., Jr., S. C. Shen, T. H. Ham, E. M.Fleming, and W. B. Castle. Studies on the de-struction of red blood cells. IX. Quantitative meth-ods for determining the osmotic and mechanicalfragility of red cells in the peripheral blood andsplenic pulp; the mechanism of increased hemoly-sis in hereditary spherocytosis (congenital hemo-lytic jaundice) as related to the functions of thespleen. Arch. intern. Med. 1956, 97, 1.

5. Young, L. E., M. J. Izzo, and R. F. Platzer. Heredi-tary spherocytosis. I. Clinical, hematologic andgenetic features in 28 cases, with particular refer-ence to the osmotic and mechanical fragility of in-cubated erythrocytes. Blood 1951, 6, 1073.

6. Selwyn, J. G., and J. V. Dacie. Autohemolysis andother changes resulting from the incubation invitro of red cells from patients with congenitalhemolytic anemia. Blood 1954, 9, 414.

7. Dacie, J. V. The Haemolytic Anaemias, Congenitaland Acquired: Part I. The Congenital Anaemias,2nd ed. New York, Grune & Stratton, 1960.

8. Prankerd, T. A. J., K. I. Altman, and L. E. Young.Abnormalities of carbohydrate metabolism of redcells in hereditary spherocytosis. J. clin. Invest.1955, 34, 1268.

9. Shafer, A. W. Glycolytic intermediates of erythro-cytes in hereditary spherocytosis (abstract). Clin.Res. 1961, 9, 67.

10. Zipursky, A., D. Mayman, and L. G. Israels. Phos-phate metabolism in erythrocytes of normal hu-mans and of patients with hereditary spherocytosis.Canad. J. Biochem. 1962, 40, 95.

11. Prankerd, T. A. J. Inborn errors of metabolism inred cells of congenital hemolytic anemias. Amer.J. Med. 1957, 22, 724.

12. Dunn, I., K. H. Ibsen, E. L. Coe, A. S. Schneider,and I. M. Weinstein. Erythrocyte carbohydratemetabolism in hereditary spherocytosis. J. clin.Invest. 1963, 42, 1535.

1718


13. Allison, A. C., M. Kates, and A. T. James. Anabnormality of blood lipids in hereditary sphero-cytosis. Brit. med. J. 1960, 2, 1766.

14. Kates, M., A. C. Allison, and A. T. James. Phos-phatides of human blood cells and their role inspherocytosis. Biochim. biophys. Acta (Amst.)1961, 48, 571.

15. De Gier, J., L. L. M. Van Deenen, R. A. Geerdink,K. Punt, and M. C. Verloop. Phosphatide pat-terns of normal, spherocytic and elliptocytic redblood cells. Biochim. biophys. Acta (Amst.) 1961,50, 383.

16. Philips, G. B., and N. S. Roome. Quantitative chro-matographic analysis of the phospholipids of ab-normal human red blood cells. Proc. Soc. exp.Biol. (N. Y.) 1962, 109, 360.

17. Prankerd, T. A. J. Studies on the pathogenesis ofhaemolysis in hereditary spherocytosis. Quart. J.Med. 1960, 29, 199.

18. Reed, C. F., and S. N. Swisher. Abnormalities ofin vitro behavior of structural lipids of red bloodcells from patients with hereditary spherocytosis.J. clin. Invest. 1960, 39, 1019.

19. Harris, E. J., and T. A. J. Prankerd. The rate ofsodium extrusion from human erythrocytes. J.Physiol. (Lond.) 1953, 121, 470.

20. Bertles, J. F. Sodium transport across the surfacemembrane of red blood cells in hereditary sphero-cytosis. J. clin. Invest. 1957, 36, 816.

21. Danowski, T. S. The transfer of potassium acrossthe human blood cell membrane. J. biol. Chem.1941, 139, 693.

22. Harris, J. E. The influence of the metabolism ofhuman erythrocytes on their potassium content.J. biol. Chem. 1941, 141, 579.

23. Maizels, M. Cation control in human erythrocytes.J. Physiol. (Lond.) 1949, 108, 247.

24. Jacob, H. S., and J. H. Jandl. Cellular permeabilityin the pathogenesis of hereditary spherocytosis(abstract). J. clin. Invest. 1963, 42, 944 .

25. Ham, T. H. A Syllabus of Laboratory Examinationsin Clinical Diagnosis. Critical Evaluation ofLaboratory Procedures in the Study of the Pa-tient, revised edition, L. B. Page and P. J. Culver,Eds. Cambridge, Harvard University Press, 1960.

26. Bergenhem, B., and R. Fahraeus. tVber spontaneHaemolysinbildung im Blut. Z. ges. exp. Med.1936, 97, 555.

27. Collier, H. B. and K. M. Wilbur. Lysolecithin andthe antihemolytic value of the blood. J. Lab. clin.Med. 1944, 29, 1123.

28. Murphy, J. R. Erythrocyte metabolism. III. Re-lationship of energy metabolism and serum fac-tors to the osmotic fragility following incubation.J. Lab. clin. Med. 1962, 60, 86.

29. Jandl, J. H., M. S. Greenberg, R. H. Yonemoto, andW. B. Castle. Clinical determination of the sitesof red cell sequestration in hemolytic anemias. J.clin. Invest. 1956, 35, 842.

30. Furth, F. W. Effect of spherocytosis on volume oftrapped plasma in red cell column of capillary andWintrobe hematocrits. J. Lab. clin. Med. 1956,48, 421.

31. Scholz, R., H. Schmitz, Th. Bucher, and J. 0. Lam-pen. tYber die Wirkung von Nystatin auf Backer-hefe. Biochem. Z. 1959, 331, 71.

32. Nelson, N. A photometric adaptation of the Somogyimethod for the determination of glucose. J. biol.Chem. 1944, 153, 375.

33. Jacob, H. S., and J. H. Jandl. Effects of sulfhydrylinhibition on red blood cells. I. Mechanism of he-molysis. J. clin. Invest. 1962, 41, 779.

34. Grunert, R. R., and P. H. Phillips. A modificationof the nitroprusside method of analysis for glu-tathione. Arch. Biochem. 1951, 30, 217.

35. Beutler, E. The glutathione instability of drug-sensitive red cells. A new method for the in vitrodetection of drug sensitivity. J. Lab. clin. Med.1957, 49, 84.

36. Goodner, C. J., and N. Freinkel. Studies of anteriorpituitary tissue in vitro: Effects of insulin and ex-perimental diabetes mellitus upon carbohydratemetabolism. J. clin. Invest. 1961, 40, 261.

37. Solomon, A. K. The permeability of the humanerythrocyte to sodium and potassium. J. gen.Physiol. 1952, 36, 57.

38. Chapman, R. G., M. A. Hennessey, A. M. Walters-dorph, F. M. Huennekens, and B. W. Gabrio.Erythrocyte metabolism. V. Levels of glycolyticenzymes and regulation of glycolysis. J. clin. In-vest. 1962, 41, 1249.

39. Jacob, H. S., and J. H. Jandl. Unpublished observa-tions.

40. Sheppard, C. W., and W. R. Martin. Cation ex-change between cells and plasma of mammalianblood. I. Methods and application to potassiumexchange in human blood. J. gen. Physiol. 1950,33, 703.

41. Murphy, J. R. Erythrocyte metabolism. V. Activecation transport and glycolysis. J. Lab. clin. Med.1963, 61, 567.

42. Ga'rdos, G. Akkumulation der Kaliumionen durchmenschliche Blutkorperchen. Acta physiol. Acad.Sci. hung. 1954, 6, 191.

43. Dunham, E. T. Linkage of active cation transportto ATP utilization. Physiologist 1957, 1, 23.

44. Whittam, R. Potassium movements and ATP in hu-man red cells. J. Physiol. (Lond.) 1958, 140, 479.

45. Hoffman, J. F. Cation transport and structure of thered-cell plasma membrane. Circulation 1962, 26,1201.

46. Skou, J. C. The influence of some cations on anadenosine triphosphatase from peripheral nerves.Biochim. biophys. Acta (Amst.) 1957, 23, 394.

47. Post, R. L., C. R. Merritt, C. R. Kinsolving, andC. D. Albright. Membrane adenosine triphospha-tase as a participant in the active transport of

1719


sodium and potassium in the human erythrocyte.J. biol. Chem. 1960, 235, 1796.

48. Dunham, E. T., and I. M. Glynn. Adenosinetriphos-phatase activity and the active movements of alkalimetal ions. J. Physiol. (Lond.) 1961, 156, 274.

49. Hokin, L. E., and M. R. Hokin. Phosphatidic acidmetabolism and active transport of sodium. Fed.Proc. 1963, 22, 8.

50. Charnock, J. S., and R. L. Post. Evidence of themechanism of ouabain inhibition of cation acti-vated adenosine triphosphatase. Nature (Lond.)1963, 199, 910.

51. Schatzmann, H. J. Herzglykoside als Hemmstoffefiur den aktiven Kalium- undl Natriumtransportdurch die Erythrocytenlmembraii. Helv. physiol.pharmacol. Acta 1953, 11, 346.

52. Ponder, E. Hemolysis and Related Phenomena.New York, Grune & Stratton, 1948.

53. Jacob, H. S., and J. H. Jandl. Effect of membranesulfhydryl deficiency on the shape and survival ofred cells (abstract). Clin. Res. 1961, 9, 162.

54. Jacob, H. S., and J. H. Jandl. Effects of sulfhydrylinhibition on red blood cells. II. Studies in vivo.J. clin. Invest. 1962, 41, 1514.

55. Whittam, R. The asymmetrical stimulation of amembrane adenosine triphosphatase in relation toactive cation transport. Biochem. J. 1962, 84, 110.

56. Glynn, I. M. Activation of adenosinetriphosphataseactivity in a cell membrane by external potassiumand internal sodium. J. Physiol. (Lond.) 1962,160, 18P.

57. Haut, A., G. R. Tudhope, G. E. Cartwright, andM. M. Wintrobe. Studies on the osmotic fragilityof incubated normal and abnormal erythrocytes.J. clin. Invest. 1962, 41, 1766.

58. Simon, E. R. Expanded autohemolysis as an in-vestigative and diagnostic tool in hemolytic dis-orders. Clin. Res. 1963, 11, 200.

59. Murphy, J. R. Erythrocyte metabolism. II. Glucosemetabolism and pathways. J. Lab. clin. Med. 1960,55, 286.

60. Glynn, I. M. The ionic permeability of the red cellmembrane. Progr. biophys. 1957, 8, 241.

61. Sen, A. K., and R. L. Post. Stoichiometry of activeNa+ and K+ transport to energy-rich phosphatebreakdown in human erythrocytes. Fed. Proc.1961, 20, 138.

62. Gabrio, B. W., C. A. Finch, and F. M. Huennekens.Erythrocyte preservation: a topic in molecularbiochemistry. Blood 1956, 11, 103.

63. Lardy, H. A., and H. Wellman. Oxidative phos-phorylations: role of inorganic phosphate andacceptor systems in control of metabolic rates. J.biol. Chem. 1952, 195, 215.

64. Lynen, F., G. Hartmann, K. F. Netter, and A.Schnegraf, in Ciba Foundation Symposium: Reg-ulation of Cell Metabolism, G. E. W. Wolstenholmeand C. M. O'Connor, Eds. Boston, Little, Brown,1959, p. 256.

65. Rose, I. A., J. V. B. Warms, and E. L. O'Connell.Role of inorganic phosphate in stimulating the glu-cose utilization of human red blood cells. Bio-chem. biophys. Res. Commun. 1964, 15, 33.

66. Robinson, M. A., P. B. Loder, and G. C. de Gruchy.Red-cell metabolism in non-spherocytic congenitalhaemolytic anaemia. Brit. J. Haemat. 1961, 7,327.

67. Harris, I. M., J. M. McAlister, and T. A. J. Prank-erd. The relationship of abnormal red cells to thenormal spleen. Clin. Sci. 1957, 16, 223.

68. Griggs, R. C., R. Weisman, Jr., and J. W. Harris.Alterations in osmotic and mechanical fragilityrelated to in vivo erythrocyte aging and splenicsequestration in hereditary spherocytosis. J. clin.Invest. 1960, 39, 89.

69. Bromberg, P. A., J. Theodore, E. D. Robin, andW. N. Jensen. Human erythrocyte pH with spe-cial reference to intracellular acidosis in hereditaryspherocytosis (abstract). J. clin. Invest. 1962, 41,1349.

70. Carroll, L. Through the Looking Glass.

1720

increased cell membrane in pathogenesis ...€¦ · c14-glucose was added to each 1 ml of 35% cell...

Documents