PHOSPHOLIPID METABOLISM IN INTACT AND MODIFIED …

PHOSPHOLIPID METABOLISM IN INTACT ANDMODIFIED ERYTHROCYTE MEMBRANES

COLVIN M . REDMAN

From The New York Blood Center, New York 10021

ABSTRACT

Erythrocyte membranes incorporated labeled phosphate from -y-adenosine triphosphate(AT32P) into phosphatidic acid and the polyphosphoinositides . Inositol-3H and palmitate-14C were also incorporated into the phospholipids but a-glycerophosphate 3 2P was not.The incorporation of y-AT32P into phospholipids was increased when the erythrocyteghosts were incubated in hypotonic media which lysed the cells . Lysis had little or no effecton the incorporation of inositol-3H and palmitate-14C into the phospholipids .

If erythrocyte membranes were prepared in 1 mm ethylenediaminetetraacetate (EDTA),instead of 1 mM MgCl2, then the tonicity of the incubating medium did not influence theincorporation of y-AT32P into the phospholipids . Erythrocyte ghosts, prepared by lysisin water, EDTA, or I mm calcium, lead, mercury, zinc, or cadmium, failed to reconstitutewhen placed in isotonic medium, inasmuch as they did not retain potassium against achemical gradient. Ghosts prepared by lysis in 1 mm magnesium, barium, or strontiumcould be reconstituted. Ghosts which failed to reconstitute incorporated more labeledphosphate from y-AT32P into the phospholipids than did intact or reconstituted ghosts .The larger incorporation of labeled phosphate by leaky ghosts was not due to a greaterentrance of y-AT 32 P into those cells.Primaquine phosphate and digitonin, at concentrations which are known to cause cells

to form smaller vesicles or to lyse cells by removing cholesterol, did not increase the incor-poration of labeled phosphate into the phospholipids .

It is suggested that the increased metabolism of phospholipids may be involved in amembrane repair mechanism.

INTRODUCTION

The characteristic ability of the red cell membraneto lyse under a variety of conditions and to re-constitute itself insofar as permeability propertiesare concerned (1, 2) afford a model for the studyof the chemical changes which may occur in amembrane during the processes of membranedamage and repair.

The phospholipids of the human erythrocytesundergo a very slow rate of turnover . It has beendemonstrated, however, that human erythro-cytes can incorporate labeled inorganic phos-phate into phosphatidic acid (3, 4) and also

into the polyphosphoinositides (5), although thelarge majority of phospholipids do not incorporatephosphate. There are some indications that theremay be some interchange of complete phospho-lipid molecules between red cells and plasma,but it is believed that this occurs only to a verylimited extent (6) . The red cell is capable, how-ever, of incorporating unesterified fatty acidsfrom the external medium into phospholipids .It has been calculated that the extent of phospho-lipid fatty acid turnover is approximately 2 %per hr, and that this represents an expenditure

THE JOURNAL OF CELL BIOLOGY . VOLUME 49, 1971 . pages 35-49

35

Dow

nloaded from http://rupress.org/jcb/article-pdf/49/1/35/1069733/35.pdf by guest on 08 April 2022

of about 27o of the energy available from glycoly-sis (7) . Hence, phospholipids, although not soactive metabolically in the erythrocyte membraneas in membranes of other animal cells, are stillin a dynamic metabolic state. This report de-scribes experiments designed to determinewhether changes in phospholipid metabolismaccompany the processes of membrane damageand repair.

MATERIALS AND METHODS

Preparation of Red Cells and

Membrane Fractions

RED CELLS : Fresh, heparinized human bloodwas used . Red cells were prepared by centrifugingthe blood in the cold at 1570 g for 10 min and bywashing the cells three times with cold 0 .15 MNaCl in 0 .01 M Tris-HCI, pH 7.4. Care was takento remove all of the huffy layer above the red cells

MG++ -LYSED MEMBRANES AND RECONSTI-TUTED "GHOSTS" : The washed red cells weremixed with 10 volumes of cold I HIM MgC12 or 1mm MgCl2 in 10 mm : Tris, pH 7.4, and centrifugedat 27,000 g for 10 min. The pellet was washed threetimes with 1 MM MgC12 in 10 mm Tris-HCI, pH 7 .4 .The resulting washed precipitate was resuspendedin 10 Him of Tris-HCI, pH 7 .4. This fraction iscalled the "red cell membranes ." In some experi-ments 15 .5 ideal milliosmolar (imOsM) Na phos-phate, pH 7.4, was used instead of the 1 mM MgC12and 10 mm Tris-HCI, pH 7.4, buffer .

Reconstituted ghosts were prepared by the methodof Whittam and Ager (8) . After the addition of 5-10volumes of 1 MM MgCl2 to the washed red cells,sufficient 3 M NaCl was added to raise the finalconcentration of the hemolysate to 0 .17 M NaCl .The hemolysate was then incubated at 37 °C for30 min to reconstitute the membranes . The recon-stituted ghosts were sedimented at 27,000 g for 10min, washed twice, and resuspended with 0 .15 MNaCI in 10 mm Tris, pH 7.4.EDTA-LYSED MEMBRANES : Washed red cells

were lysed by mixing with 10 volumes of 1 mmEDTA, 10 mm: Tris-HCI, pH 7.4, for 10 min in thecold . The lysed cells were washed three times withthe same EDTA buffer. The centrifuged membraneswere white, showing no traces of attached hemo-globin . The EDTA-lysed membrane preparationswere then resuspended in the appropriate buffer .RECONSTITUTED GHOSTS PREPARED AFTER

LYSIS WITH DIFFERENT CATIONS : Washed redcells were lysed in 10 volumes of a I mm solution ofthe cation to be tested . The lysed cells were resealedby adding sufficient amounts of a 3 M salt solution(3 :1, K/Na) to raise the molarity to 0 .17 . The re-

36

THE JOURNAL OF CELL BIOLOGY . VOLUME 49, 1971

sealed ghosts were washed three times with 0 .17 MNaCl in 0.01 M Tris-HCI, pH 7.4. The packed,washed, reconstituted ghosts were used for deter-mination of potassium in a flame photometer or forincubation with radioactive precursor.

Materials

Palmitic acid- 14C, uniformly labeled (640 mCi/mmole), was purchased from New England NuclearCorp., Boston, Mass . This radioactive fatty acid wasabsorbed on human albumin by the method ofShobet et al . (7) . The final material contained5 X 10 6 cpm of palmitic acid and 23 .5 mg albuminin I ml of solution .y-AT 32P was prepared by the method of Glynn

and Chappell (9), and a-glycero- 32P by the methodof McMurray et al . (10) . Inositol-2- 3H was purchasedfrom New England Nuclear Corp . and had a specificactivity of 3 Ci/mmole . Bacterial alkaline phosphatasewas obtained from Worthington Biochemical Corp .,Freehold, N. J., primaquine phosphate was a giftfrom the Sterling-Winthrop Research Institute inRensselaer, New York, and digitonin was of certifiedgrade from Fisher Scientific Company, Pittsburgh,Pa.

Lipid Extraction and Chromatography

t Phospholipids were extracted from the trichloro-acetic acid (TCA)-precipitated material by usingCHCI3 methanol concentrated HCI (200 :100 :1,v/v) as described by Folch (11) . Paper chromatog-raphy was performed on silica gel-loaded paper witheither phenol-NH3 (15) or diisobutyt ketone-aceticacid-water as solvent (25) .

Buffer mixtures of sodium phosphate to give thedesired ideal milliosmolarity (imOsM) were preparedas described by Dodge et al . (12) . Phosphorus was de-termined by the method of Bartlett (13) . The ex-tracted phospholipids were deacylated as described byDawson (14), and the deacylated materials werechromatographed on Dowex 1-formate (Dow Chemi-cal Co., Midland, Mich .) as previously described (15)or on paper (16) .

RESULTS

Identification of Radioactive Phospholipids

Four radioactive spots were obtained when thechloroform extract, from ghosts incubated withy-AT32P, was chromatographed on silicic acid-impregnated paper, as described by Santiago-Calvo et al . (15) . This agrees with the earlierwork of Hokin and Hokin who identified theselipids as triphosphoinositide, diphosphoinositide,an unknown material, and phosphatidic acid

Dow


(5) . Most of the radioactivity obtained under theconditions used in these experiments occurredin two areas with R f values of 0 .19 (band A) and0.65 (band C) . Minor radioactive areas wereoccasionally noticed at Rf values of 0 .09 and 0.50 .Sometimes the radioactive area at R f 0.65 ap-peared to separate into two spots. Bands A andC were always obtained, and it is these compoundswhich were routinely counted and identified .Fig. 1 shows a typical chromatogram.

The material at R f 0.65 (band C) was elutedfrom the paper and was rechromatographed onsilica gel-loaded paper using the diisobutylketone : acetic acid :water (40 :25 :5) solvent sys-tem as described by Marinetti (25) . It had anR f value of 0 .75, identical with that of phospha-cidic acid . No other radioactive spots were ob-served . The material with Rf 0.19 (band A)streaked in the Marinetti system . This is a prop-erty that the polyphosphoinositides exhibit inthis system .The two major radioactive bands (A and C)

were deacylated with methanolic sodium hy-droxide as described by Dawson (14) . The deacyl-ated material was chromatographed in two dif-ferent paper chromatographic systems and alsoon Dowex 1-formate . The deacylated materialfrom band C chromatographed with a-glycero-phosphate, indicating that the parent lipid wasphosphatidic acid . The deacylated material frombands A and B, on chromatography in the systemdescribed by Desjobert and Patek (16), gave anRf value relative to that of a-glycerylphosphoryl-inositol diphosphate . The minor band withan original R f of 0.50 gave a deacylated productwith an Rf value similar to that of glyceryl-phosphorylinositol phosphate (5) . This indicatesthat the parent phospholipids are triphospho-inositide and diphosphoinositide .

Ghosts prepared by lysis with cadmium incor-porated 32P of y-AT32P mostly into bands A andB, with little or no incorporation into band C .

FIGURE 1 Autoradiogram showing the chromato-tographic separation of the 32P-labeled erythrocytephospholipids . Phospholipids from erythrocyte ghostsincubated with y-AT 32P in Krebs-phosphate bufferwithout saline were chromatographed on silica gel-impregnated paper using phenol-NH3 as solvent asdescribed in Materials and Methods . The identificationof the radioactive bands A and C is given in Results .o is the origin and sf is the solvent front.

COLVIN M. REDMAN Phospholipid Metabolism in Erythrocyte Membrµnes 37

Dow


When the deacylated chloroform extract fromsuch material was chromatographed on Dowex 1-formate as described by Santiago-Calvo et al .(15), two peaks were obtained . One peak waseluted with 0 .4 M ammonium formate in 4 N

formic acid, and the other with 0.8 M ammoniumformate in 4 N formic acid . Previous work hadidentified material with this chromatographicbehavior on Dowex 1-formate as glycerylphos-phorylinositol phosphate and glycerylphosphoryl-inositol diphosphate (15) .

These results indicate that the radioactivephospholipids are phosphatidic acid (band C)and the polyphosphoinositides (band A aad B),and agree with the previous studies of Hokinand Hokin (5) who found these to be the majorradioactive products produced when erythrocyteghosts were incubated with y-AT32P .

Treatment of the Deacylated Material fromBand A with Alkaline Phosphatase

The deacylated material from band A (glyceryl-phosphorylinositol diphosphate), on treatmentwith alkaline phosphatase, liberated more than95% of its radioactivity as inorganic phosphate .Hence, the diesterified phosphate of triphos-phoinositide is not labeled by y-AT32P. Thisindicates that the majority of the radioactivityin triphosphoinositide is obtained by the phos-phorylation of phosphatidylinositol or diphos-phoinositide . A similar finding was obtained byHokin and Hokin (5) .

The Incorporation of 32P of y AT32P into thePhospholipids of ReconstitutedErythrocyte Ghosts

Reconstituted ghosts were used to study themetabolism of membrane phospholipids in re-constituted membranes and in lysed membranes,the latter referring to reconstituted ghosts incu-bated under hypotonie conditions. Phospholipidmetabolism was first measured by following theincorporation of 32P of y-AT32P into the isolatedtotal phospholipids. In the reconstituted mem-branes there was only a small amount of incor-poration, whereas in the lysed membranes therewas a large incorporation of phosphate fromy-AT32P into the phospholipids (Fig . 2) . Sepa-ration of the phospholipids by chromatographyrevealed that about 70-75% of the radioactivity

38

THE JOURNAL OF CELL BIOLOGY • VOLUME 49, 1971

was in phosphatidic acid and that the remainderwas in the polyphosphoinositides .

Incorporation of phosphate into the phospho-lipids only occurs at salt concentrations below0.05 M. At higher salt concentrations the incor-poration is similar to that seen in isotonic medium(Fig . 3 A) . The amount of hemolysis as measuredby the release of hemoglobin from the reconsti-tuted ghosts incubated under these conditionsis shown in Fig . 3 B. The reconstituted ghosts losetheir entrapped hemoglobin at a salt concentra-tion between 0 .05 and 0 .09 M.

The Effect of Varying Salt Concentration onthe Incorporation of 32P from y A T32P intoPhospholipids of Lysed Membranes

The previous experiments were performed withreconstituted ghosts. These are cells which havebeen resealed and have regained their semi-permeable properties . The increased incorpora-tion of phosphate into the phospholipids may bedue to the opening or lysing of the cells. The ex-periments described in this section were donewith cells which were "open" to begin with and

300

100

XI

,%

LYSED [0.012 M]

i

200

x//r/

i

i

r

X

INTACT [0.16M]i

I

10

20

30

40

50

60MINUTES

FIGURE 2 Time course of incorporation of y-AT32Pinto osmotically lysed and intact erythrocyte mem-branes . Reconstituted ghosts were prepared as de-scribed in the text and were incubated with 0 .01mTris-HCI, pH 7.4, 1 MM M902, and 0.1 gmole/ml ofy-AT32P (5.2 X 107 cpm/,ttmole) . One set of ghostswere incubated with the above medium plus 0.16 MNaCl, and the other in hypotonie medium with 0.012 MNaCl. Incubation was carried out at 37 °C .

Dow


A

B50

CLY>- nn 200

PA

F 40°

â i° -°

z 30oE

m3

°O 4 oJ

. 100

2 200xM xaU) m= E0 o

0ä

FIGURE 8 Incorporation of y-AT32P into phospho ipids, and the release of hemoglobin by erythrocyteghosts incubated with varying amounts of salt . Reconstituted ghosts were incubated at 37 °C for 10 minwith varying NaCl concentrations under conditions described in Fig . 2 . The hemoglobin released wasmeasured by reading the optical density of the released material at 500 mµ . The radioactivity in phos-phatidic acid (band C) and triphosphoinositide (band A) was determined as described in the text .

1000ILv

TF- ô> ôF ôâ û.•

0 6004•

na

x

•

É 200IL o

ä

800

400

T

I t0.05

0.10

0.15

0 20

0.30

0.05

0.10

0.15

0.20

0.30NaCI [M]

then were closed or allowed to remain open dur-ing incubation with y-ATS 2P. This was done todetermine whether the phospholipid effect wasdue to either the osmotic opening or the closingof the membrane .In a medium in which the membranes do not

reseal (0.01 M NaCl or I mOsM sodium phos-phate buffer), there was maximum incorporationinto the total phospholipids for about 10-15min. As the salt concentration was raised andthe membranes started to reseal, the amount of

MINUTES

FIGURE 4 Time course of incorporation of y-AT 32P into phospholipids of erythrocyte membranes in-cubated in solutions of different osmolarities . Magnesium-lysed membranes were prepared as describedin Materials and Methods . The membranes were not reconstituted prior to incubation . Experiment Awas performed with 0 .1 M Tris-HCI, pH 7.4, with 1 mat MgCl2 and 0 .1 µmoles y-AT32P/ml (6 X 107cpm/µmole), and experiment B with the same concentrations of Mg++ and y-AT 32P but with varyingconcentration of sodium phosphate buffer at pH 7 .4 .

B

SODIUM PHOSPHATE[m Os M]

incorporation of y-AT32P into phospholipids de-creased (Fig . 4) . The type of salt used to changethe tonicity of the medium was not critical .Fig. 4 shows two experiments, one using NaCland the other using phosphate buffer . Both showsimilar responses to increasing salt concentra-tions. However, when sodium phosphate buf-fer was used instead of NaCl there was lessincorporation into the polyphosphoinositides,with 90% of the incorporation found in phos-phatidic acid .

COLVIN M. REDMAN Phospholipid Metabolism in Erythrocyte Membranes

39

Dow


The Effect of Salt Concentration on Phospho-lipid Metabolism of "Leaky" Membranes

Erythrocyte ghosts prepared by lysing in thepresence of EDTA do not regain their semi-permeable properties with respect to Na+ and K+when put into an isotonic medium . Membranesprepared in this manner always remain leaky .The subsequent addition of Mg++ and Ca++ tocells which were lysed with EDTA fails to repairthis damage (9) . When "leaky" membranes,prepared by lysis in 1 mm EDTA, were incubatedwith y-AT32P, it was found that Mg++ was neces-sary for phosphate incorporation into the phos-pholipids. In the presence of Mg++ there wasactive incorporation and most of the radioactiv-ity was found in the polyphosphoinositides . Un-like cells prepared by lysis in I mu Mg++, theEDTA-lysed cells were not affected by the saltconcentration, and they incorporated y-AT 32Pinto phospholipids equally well in hypotonic orisotonic medium (Fig. 5) .

If Ca++ was introduced into the medium, theradioactivity was found in phosphatidic acidand not in the polyphosphoinositides, but againthere was little or no effect of salt (Fig . 5) . Hence,

40

CL9Q

o} ûF y~ o

U y0•

E EO_O 7a v0•

0

•

é•

UaN do,0= Ma 1ô

E0ä

300

200

100


increased phospholipid metabolism is not due tolowering the salt concentration, but is mani-fested in all leaky cells regardless of the salt con-centration .

The Effect of Cations on the Reconstitution ofErythrocyte Membranes and onPhospholipid Metabolism

The ability to reconstitute an erythrocyte mem-brane is influenced not only by EDTA. The ioniccontent of the lysing medium is of importance .For example, if cells are lysed at very low cal-cium concentrations they are able to retainpotassium against a chemical gradient but theylose this ability if they are lysed at higher calciumconcentration . This is in contrast to cells lysedin magnesium which readily regain their semi-permeable properties when placed in isotonicmedium (17) . Thus, it is possible to prepareerythrocyte ghosts which either are semipermeableto potassium or have become fully permeable orleaky. Experimentally, this was done by lysingthem in the presence of either water, EDTA, or avariety of divalent cations and then resealingthem by raising the salt concentration to 0 .17

FIGURE 5 Effect of tonicity and calcium on the incorporation of y-AT32 P into phospholipids of erythro-cyte membranes prepared in the presence of 1 mm EDTA. EDTA-lysed erythrocyte membranes wereprepared as described in Materials and Methods . They were incubated in Krebs-Ringer phosphate and0.1 µmole y-AT 32P/ml (5 X 10, cpm/µmole) in the presence or absence of calcium and with either iso-tonic NaCl (0 .15 M) or a hypotonie saline solution (0 .03 m) . Experiment A was carried out in the presenceof 2 .58 mm calcium . In experiment B, no calcium was added .

Dow


M with a 5 :1, KCI :NaCI, solution. The resealedcells entrap potassium when they are incubatedat 37 °C for 40 min . The cells were then washedto remove extracellular potassium, and the amountof potassium retained within the cell was used asa measure of the intactness of the reconstitutedmembrane. Reconstituted membranes so pre-pared were incubated with y-AT 32P, and theincorporation of 32P into phospholipids wasmeasured and compared with the ability of thereconstituted cell to retain potassium .

The ability of cells, lysed in different calciumconcentrations, to retain potassium is shown inFig. 6. As the calcium level of the lysing solutionis raised, the membranes fail to reseal and theylose their intracellular potassium. Maximumloss of potassium occurs when cells are lysed in0.5 mm calcium . The incorporation of y-AT32Pinto phospholipid was low in cells prepared in0.25 and 0.5 mm calcium, but it increased incells prepared with higher concentrations of cal-cium. There is some relation between loss of po-tassium from the cell and increased phospholipidmetabolism. This relation is not, however, astrict one . For instance, at 0 .5 mm Ca++ the cellshad completely lost their ability to retain potas-sium but there was little or no increase in phos-pholipid metabolism (Fig . 6) .

0 .5

1 .0

1 .5

2.0MM Ca" USED TO LYSED CELLS

FIGURE 6 Effect of calcium on the reconstitution of erythrocyte ghosts and on phospholipid metabolism .Washed cells were lysed in 10 volumes of a calcium solution with different concentrations as noted . Thelysed membranes were then reconstituted by adding a salt solution to a final concentration of 0 .17 M .The salt solution was 3/1, K/Na. The cells were incubated at 37 °C for 40 min, and then were washedthree times in 0.17 M NaCl in 0 .01 M Tris, pH 7 .4, to remove the extracellular potassium . Some of thewashed cells were kept to determine the amounts of retained potassium as measured by flame photometry,and others were incubated with 7-AT 32P (0 .1 µmole/ml) in isotonic Krebs-Ringer phosphate .

Other divalent cations added to the lysate werealso found to have an influence on the recon-stitution properties of the membranes . Bariumand strontium acted like magnesium, in thatthey allowed the cell to regain its semipermeableproperties to potassium . Other divalent cationssuch as zinc, calcium, lead, and mercury actedlike calcium in affecting the membrane so that itwould not reconstitute . If the cells were lysedin water or in EDTA, they also failed to retainpotassium (Table I) .

Ghosts, prepared by lysis with different cations,were incubated with y-AT32P. Those ghosts whichwere able to retain potassium (lysed in Mg++,Ba++, or Sr++) had the lowest incorporationinto the phospholipids . Ghosts which were leakyto potassium had a higher incorporation . Therewas not, however, a quantitative relation be-tween the amount of potassium lost and theincrease in phospholipid metabolism . Ghostsprepared in the presence of mercury or lead lostas much potassium as those prepared with cad-mium or zinc, but they incorporated less phos-phate into phosphatidic acid and the polyphos-phoinositides (Table IV, experiment 3) . Ingeneral, however, those cells which failed toretain potassium had an increased phospholipidmetabolism .

COLVIN M . REDMAN Phospholipid Metabolism in Erythrocyte Membranes

41

500 -,100

TF> 400 x

N

JWu80 rr- a / m

Q a - 02 '~â 300 /

° W- Z

60 û rn ôäâJ p /

U WL K

O

â 200 X/Y 2

40 W Ny vO= /

yQF

a EnuV

i,X/

OCL

20x

100X I

Dow


Experiment

Cation present during lysis

K+ retained

The ghosts prepared in magnesium or calciumhad a greater percentage of • the incorporatedradioactivity in phosphatidic acid, whereasghosts prepared in water, EDTA, and the otherdivalent cations incorporated 32P mostly intotriphosphoinositide.

Phospholipid Metabolism in Intact andLysed Red Cells and Measurement of theIntracellular Pool of A TP

The intact erythrocyte membrane is imper-meable to ATP. In lysed cells, however, the ATPmay enter the cell. If the incorporation of y-AT 32Pinto the phospholipids occurs only on the insideface of the vesicular membrane, then the in-creased phospholipid metabolism of open cellsmay merely be due to a larger entry of y-AT 32Pinto the vesicle . To determine whether this wasresponsible for the increased phospholipid metab-

TABLE I

The Effect of Various Cations on the Ability of the Erythrocyte Ghost to Retain Potassium and to IncorporateyAT32P into Phospholipids

42


RadioactivityCHC1s extract

Phospholipid radioactivity

PA

Poly PI

The cells were lysed in either 10 volumes of water or I mm of the cation, or EDTA. The cells were re-sealed with 0 .17 M salt containing 5/1 K+/Na+, and were then incubated at 37 °C for 40 min . The resealedcells were washed three times with 0 .01 M Tris, pH 7 .4, with 0.17 M NaCl . The amount of K+ retained inthe cell was determined by flame photometry . The K+ values are the averages of two experiments . Normalred cells contain about 90 uEq of K+/g cells . The resealed cells were also incubated with y-AT 32P asdescribed in Fig . 2 .The CHC13 extract represents the total phospholipid radioactivity . PA, phosphatidic acid ; Poly PI,polyphosphoinositides .

olism, attempts were made to load the erythro-cyte with y-AT32P by lysing the cells in the pres-ence of y-AT32P and then resealing the vesiclewith the radioactive precursor inside . This couldbe accomplished, but the amount of y-ATE 2Pentrapped in the vesicle was too small and tooeasily hydrolyzed to permit measurement oflipid metabolism. Therefore, to determinewhether changes in the intracellular y-AT32Plevels are responsible for the phospholipid effect,experiments were performed with intact redcells, incubated in a medium containing inor-ganic 32P. These cells will allow the entry of inor-ganic 32P and will convert part of it, by glycolysis,into y-AT32P. As a measure of the intracellulary-AT32P pool, the specific activity of the nucleo-tide phosphates within the cell was determined .Cells were incubated in media containing differ-ent amounts of either NaCl or sucrose. Also meas-

tEq/g cell cpm/µg phosphotipid cpm/pg phospholipid

1 Magnesium 48 172 46 35Calcium 10 300 174 52(water) 18 692 28 294(EDTA) 15 412 40 94

2 Magnesium 40 180 50 6Strontium 36 192 23 52(water) 15 639 16 361Zinc 8 1832 107 1012

3 Magnesium 40 28 -Barium 51 40 3 7Strontium 36 54 18 18Lead 5 226 12 52Mercury 5 114 - 24Zinc 8 1760 56 540Cadmium 9 2040 - 926

Dow


TABLE II

Lysis of the Erythrocyte and the Incorporation of 32P into Nucleotide Phosphates and Phosphatidic Acid

Conditions of incubation

Hb lost

ured were the radioactivity of the phospholipidsand the amount of hemoglobin lost .Lysis of the red cell by hypotonic solutions

caused changes in the specific activity of the nu-cleotide phosphates, and these changes differeddepending on the medium in which the cellswere incubated and also on the material used tochange the tonicity of the medium . Lysis of thered cell by decreasing the NaCl concentrationcaused an increase in the specific activity of theintracellular nucleotide phosphates . This in-crease in specific activity when the cell waslysed was more pronounced when the cells wereincubated in Tris buffer, pH 7.4, than whenincubated in Krebs-bicarbonate buffer . In con-trast to the above, lysis of red cells with decreas-ing sucrose concentrations caused a decrease inspecific activity of the nucleotide phosphates(Table II) . The radioactivity in phosphatidic

Radioactivity

1 0.05 os Tris, pH 7 .4, 10-4 NaF5 mm MgC1 2a . NaCI 0 .0125 asb . NaCI 0.0625 asc . NaC1 0.16 as

2 . 0 .01 as Tris, pH 7.4, 5 masMgC12

a. 0 .01 as sucroseb. 0 .03 as sucrosec . 0 .07 as sucrosed. 0 .27 as sucrose

3 . Krebs-bicarbonate buffera. No NaClb. With NaCl (0 .15 as)

Washed red cells were incubated with inorganic 32P in the incubation media described above . Incubationswere carried out at 37 °C for 20 min. Experiments 1 and 2 utilized 6 /tCi 82P/ml of incubation medium,and experiment 3 utilized 10 µCi of 32P per ml .* Conditions under which the red cell is not lysed . The hemoglobin lost was determined spectrophoto-metrically . The ratio is the specific activity of the phospholipid divided by that of the nucleotide phos-phates X 105. The specific activity of nucleotide phosphates was determined by adsorption on charcoaland hydrolysis in 1 N acid for 10 min at 100 °C, as described by Crane and Lipmann (26) . PA, phosphatidicacid .

acid was increased when the cells were lysed .Maximum phospholipid incorporation was no-ticed as soon as hemoglobin was lost from thecell . Complete lysis was not necessary to obtainthis effect . The increased phospholipid effectwas less noticeable when cells were incubatedwith sucrose (Table II) . There was little labelingof phosphatidic acid when the cell remainedintact, although the intracellular nucleotidephosphates were labeled .

To correct for differences in the entry of inor-ganic 32P, or for differences in rates of glycolysisin the cell, the radioactivity found in phospha-tidic acid was corrected to a constant specificactivity of the intracellular nucleotide phosphates .The values are shown in the last column of TableII. In all instances, regardless of the mode oflysis of cells or of the changes in the intracellularATP pool, the corrected radioactivity in phos-


43

Nucleotide P Phosphatidic acid PA X10snucleotide P

cpm /mgcpm/mg P phospholipid P

90 1 .6 X 105 134 84.070 1 .9 X 105 170 89.50 0 .64 X 10 5 11 17 .2*

100 2 .2 X 10 4 28 127 .090 2 .1 X 10 4 41 195 .550 2 .0 X 10 4 42 210 .04 3.8 X 10 4 42 58 .0*

90 2 .8 X 10 5 389 139 .04 2 .0 X 105 53 26 .0*

Dow


phatidic acid was always higher in open cellsthan in intact cells .

Entry of y-AT32P into Reconstituted GhostsPrepared in the Presence of Various Cations

The previous experiments showed that the in-tracellular ATP pool can change with lysis butthat the changes in intracellular ATP cannotaccount for the phospholipid effect noticed.Those experiments were done by incubatingwith inorganic 12P . However, most of the studieswere performed with y-AT 32P in the incubatingmedium, and hence it is important to determinedirectly the amounts of y-AT 32P that entered thecells from the medium. As mentioned before, anattempt to entrap y-AT32P into reconstitutedghosts and then to measure phospholipid metab-olism was unsuccessful. An estimate of the amountof y-AT32P which entered the cell could be ob-tained, however, by incubating intact ghosts(lysed with Mg++) and leaky ghosts (lysed withCa++, Cd++, or water) with y-AT 32P, and byrecovering the cells by centrifugation at 2 °C for5 min at 10,000 g . The amount of y-AT32P inthe pelleted cells was then determined, as wellas the radioactivity in the extracted phospho-lipids . The cells were added to a buffered salinemedium with y-AT32P which was kept at 0°C .The zero time measurement was taken by imme-diately centrifuging after the addition of the cells .In Table III it is seen that at zero time of incuba-tion there was little difference in the levels ofy-AT32P found within the cell. During incuba-

tion at 37 °C the y-AT32P within the cell wasutilized and the levels dropped. Some cellsused more ATP than others . For instance, thecells prepared by lysis in cadmium used lessATP in 30 min than did the other cells. Thegreater incorporation of 32P into the phospho-lipids of leaky cells cannot, however, be ascribedto a greater entry of y-AT32P into those cells,since the intracellular levels of y-AT 32P weresimilar but the phospholipid radioactivity wasgreatly enhanced in leaky cells (Table III) .The cells lysed in magnesium were intact whilethe others were leaky to potassium .

Treatment of Intact Cells with Digitonin andPrimaquine Phosphate

Intact cells were incubated in Krebs-Ringerbicarbonate, with 32P, in the presence and in theabsence of digitonin . Digitonin is known to re-move cholesterol from the membrane and thuscause lysis (29) . Digitonin had no effect on theincorporation of 32P into phospholipids, althoughincorporation into the nucleotide phosphate poolwas slightly enhanced. This experiment againshows that an increased radioactivity in the nu-cleotide phosphate pool is not necessarily reflectedby an increased phospholipid metabolism (TableIV) .Primaquine phosphate, which produces lysis

and resealing into small vesicles (28), caused adecrease in the specific activity of the nucleotidephosphates and an inhibition of the incorporationof 32P into the phospholipids (Table IV) .

TABLE III

Uptake of y-AT32P by Cells Lysed under Different Conditions

The cells were lysed in either water or 1 mm of the cation as described in the text . The cells were washedthoroughly with 0.17 mt NaCl in 0 .01 M Tris, pH 7.4, and then incubated at 37°C in 0.15 M NaCl, 0 .01 M,

Tris, pH 7 .4, with 0.2 Amoles of y-AT 32P per ml (3 .9 X 10 5 cpm/µmole) . At the desired time the reactionwas stopped by centrifuging the cells at 10,000 g for 5 min at 2 °C and by adding cold 5% TCA to thepellet. The y-AT 32P level is the radioactivity absorbed on charcoal from the TCA-soluble portion whichis hydrolysed in 1 N acid at 100 °C after 10 min . All values are corrected to a constant amount of phos-pholipid P .

44


y-AT32P levels in the cell Incorporation of y-ATa2 P into phospholipidsPreparation of ghosts

0 min 15 min 30 min 0 min 15 min 30 min

cpm/pg phospholipid P cpm/pg phospholipid P

Water 32500 8150 6050 620 1660 3720

Magnesium 23000 19500 6100 50 100 50

Calcium 22500 4850 5500 30 270 390

Cadmium 38200 18500 13000 600 2380 3600

Dow


TABLE IV

Effect of Primaquine and Digitonin on the Incorporation of 32P into the Nucleotide Phosphates and Phospholipidsof Intact Red Cells

Whole washed red cells were incubated with 32P (1 µCi/ml) in Krebs-Ringer phosphate at 37 °C. In ex-periment 1 the incubation was carried out for 60 min . In experiment 2 the incubation was as describedabove. The concentration of digitonin was 3 .2 X 10-4 M .The specific activity of the nucleotide phosphates was determined as described in the text . The ratiois the radioactivity of the phospholipids divided by the specific activity of the nucleotide phosphates .

TABLE V

Effect of CTP and Calcium on the Incorporation of y-AT 32P and Inositol- 3H into Membrane Phospholipids

Radioactivity into phospholipids

No salt

Salt

Membranes were incubated in Krebs-Ringer phosphate with or without calcium . Sodium chloride wasadded to some at a final concentration of 0 .15 M .Experiments 1 and 2 were performed with ghosts prepared by lysis in 1 mm MgC12 as described in thetext. The AT32P was 5 X 105 cpm/mole for experiment 1, and 10 5 cpm/µmole for experiment 2 . In bothcases the incubation medium contained 0.28 µmoles/m1 of ATP . Incubation was carried out for 45 minat 37 °C .Experiment 3 was performed with ghosts prepared by lysis in 1 mm EDTA . The inositol-2- 3H used was1 pCi/ml of incubation mixture, with a specific activity of 3 .4 mCi/µmole . PA, phosphatidic acid ; polyPI, polyphosphoinositides . The calcium and CTP concentrations were 5 X 10 -4 M .


45

Experiment Addition Nucleotide P Phospholipids Ratio X 105

I None

cp./mg P cpm/mg P

385 10 .73 .6 X 10 4Primaquine, 3 .6 X 10-3 M 2.2 X 10 4 165 7 .5Primaquine, 1 .6 X 10-2 M 2.0 X 10 4 145 7 .5

2 None 15 min at 37'C 1 .5 X 10 4 80 5 .3Digitonin, 18 min at 37 °C 1 .6 X 10 4 80 5 .0None, 45 min at 37'C 1 .5 X 10 4 140 9 .2Digitonin, 48 min at 37 °C 1 .7 X 10 4 160 9 .3

PA

Poly PI PA

Poly PI

cp./µg P

2850 435 1130 2841760 307 883 1912385 488 968 2891360 232 780 172

346 67 186 28640 78 190 38242 36 182 34

Total Phospholipids Total Phospholipids

358 10528 32

6830 7000137 108

Experiment Radioactive substrate Additions

1 y-AT32P NoneCTPInositolCTP + Inositol

2 y-AT32P NoneCa++Ca++, CTP

3 Inositol- 3H NoneCa++CTPCTP + Ca++

Dow


Effect of CTP and Inositol on the Incorpora-tion of 32P from y-AT32P intoMembrane Phospholipids

Cytidine triphosphate (CTP), although anecessary cofactor for synthesis of the phospho-inositides from phosphatidic acid, did not en-hance the incorporation of y-AT32P into thepolyphosphoinositides, but actually inhibitedthe incorporation into both phosphatidic acidand polyphosphoinositides. The presence of inosi-tol had no effect. Earlier studies by Hokin andHokin showed that CTP inhibited the incorpora-tion of y-AT32P into di- and triphosphoinositide,and had little effect on phosphatidic acid (5) .As seen in previous experiments, incubation ofthe membranes in hypotonic medium caused anincreased incorporation into the phospholipids(Table V). Calcium, added by itself, stimulatedthe incorporation of 32P into phosphatidic acid(Table V, experiment 2) . In EDTA-lysed cellsthe increased incorporation into phosphatidicacid caused by added Ca++ was followed by adecrease in the incorporation into polyphospho-inositides (Fig . 5) .

CTP, however, greatly enhanced the incor-poration of inositol3H into the phospholipids ofcells prepared by lysis in EDTA . Salt had noconsistent effect on the incorporation . Calciumseverely inhibited the incorporation in the pres-ence or absence of CTP. Inositol- 3H was incor-porated into a phospholipid with the same chro-matographic mobility as phosphatidylinositol .There was no detectable incorporation into thepolyphosphoinositides .

Other Phospholipid Metabolic Pathways inthe Erythrocyte Membrane

The erythrocyte is capable of incorporatingfree fatty acids into the phospholipids . Whenpalmitic acid 14C was absorbed to albumin andincubated with washed, intact erythrocytes,rapid incorporation into the triglycerides andthe phospholipids, but mostly into the triglycer-ides, was noted. The cells which were incubatedin hypotonie medium were slightly less activein incorporating palmitate into triglyceridesthan were the cells incubated in complete isotonicKrebs-Ringer. The incorporation of palmitate intotriglyceride leveled at 5 min of incubation in theisotonic medium but fell slightly in the hypotoniemedium. The majority of the radioactivity in the

46


FIGURE 7 Incorporation of 14C-fatty acids into thelipids of intact and lysed erythrocytes . Washed red cellswere incubated in Krebs-Ringer phosphate with orwithout salt . Palmitate-14 C adsorbed to human albumin,prepared as described in the text, was added to the in-cubation mixture . The medium contained 100,000 cpmof palmitate-14C/ml. Incubation was done at 87°C. Theradioactivity in the lipids was determined as describedin Materials and Methods . TG, triglyceride ; PE, phos-phatidylethanolamine; PC, phosphatidylcholine. Theasterisks denote incubation in hypotonie Krebs-Ringerphosphate .

phospholipid was in phosphatidylcholine andphosphatidylethanolamine . With cells incubatedin hypotonic media, it was difficult to measurethe incorporation into phosphatidylethanolaminebecause of the appearance of a material thatstreaked in the chromatographic systems whichwere used. There was a small increased incorpora-tion of palmitate into phosphatidylcholine incells incubated in isotonic saline (Fig. 7) .

The incorporation of a-glycerophosphate32Pinto the phospholipids was also measured. a-glycero-32P (2 X 10 7 cpm/µmole) was incubatedwith red cells in the presence of coenzyme A(CoA), ATP, and added fatty acids, or withstearyl and palmityl CoA . No incorporation intothe phospholipids was detected in either the pres-ence or absence of saline .

In order to determine whether the incorpora-tion of y-AT32P into the phospholipids could bedue to an exchange of phosphate of the phos-pholipids with inorganic phosphate in the media,erythrocyte membranes were incubated withy-AT32P in hypotonie media for 1 hr as de-scribed in Fig. 2. The isolated membranes werethen washed thoroughly with nonradioactivemedium and reincubated at 37 °C for 30 min

Dow


in 0.05 M phosphate buffer in the absence of ATP .There was no change in radioactivity of themembrane phospholipids during the second in-cubation, showing that exchange with inorganicphosphate nad not occurred.

DISCUSSION

The lipid composition of the human erythrocytemembranes has been well documented . Themembranes can be obtained in good yield with-out losing their lipid components (18), and theycan be reconstituted insofar as they can regaintheir semipermeable properties to sodium andpotassium (8, 17). Although the erythrocytemembrane is not actively synthesizing all of itslipids de novo, it is capable of selective synthesisof certain lipids and lipid moieties (6) . Theproperty of the erythrocyte membrane to resealafter lysis and the low metabolic background ofits phospholipids make it valuable as a modelwith which to study the chemical reactions whicha membrane may be undergoing during damageand repair. That phospholipids may be involvedin such repair mechanisms of the membranes issuggested by work from many different labora-tories . For example, there is a relation betweenincreased turnover of certain phospholipids,mainly phosphatidic acid and phosphatidyl-inositol, and secretion (19, 20) . Although someof this effect may be due to other processes, suchas the hormonal action which triggers secretion,or to other aspects of secretion, part of it may bedue to the pinching off and fusion of membraneswhich is thought to accompany the intracellulartransport of proteins . There are other findingswhich suggest that phospholipid metabolismmay accompany membrane rupture and reseal-ing. Leukocytes involved in phagocytosis havean increased metabolism of certain lipids (21, 22),as do HeLa cells treated with digitonin (23),and leukocytes caused to agglutinate by the addi-tion of phytohemagglutinin (24) . An example,that increased phospholipid metabolism of theerythrocyte membrane may accompany mem-brane damage, has been given previously byJacob and Karnovsky (27). They have shown thatred cells suspended in a hypotonic medium thatdid not cause hemolysis incorporated moreinorganic 32P into a "phosphatidylserine" fractionthan did cells suspended in isotonic medium .They also showed that red cells from a patientwith hereditary spherocytosis incorporated more

phosphate into the phospholipids than did nor-mal cells when both were incubated in identicalisotonic medium. These different systems allhave in common the fact that the membraneseither are damaged or are undergoing breakageand resealing.The difference between the amounts of incor-

poration of y-AT32P into the lipids of erythrocyteghosts incubated in hypotonic or in isotonicmedium is pronounced . This difference does notdepend on the type of salt used to change thetonicity of the medium, since the lack of eithersodium chloride, phosphate buffer, or sucrosewill cause increased metabolic activity (Figs.2 and 4, and Table I). This increased metabolismdoes not depend on tonicity, since tonicity hasno effect on ghosts which have been lysed in thepresence of EDTA (Fig. 5) . Also, ghosts preparedby lysis with equal concentrations of differentdivalent cations incorporate different amountsof y-AT32P into the phospholipids, althoughthey were subsequently incubated under thesame conditions of tonicity (Table IV) . Theseresults demonstrate that ghosts or cells which areopen in that they either release hemoglobin orfail to retain potassium, incorporate more y-AT32Pinto the phospholipids than do those cells whichare closed . This may be most easily explained ifopen cells allowed the entry of more substratethan did closed cells, and is of special significancebecause intact erythrocytes are relatively im-permeable to ATP . Two types of experiments wereperformed to determine whether this could ex-plain the increased incorporation found to occurin open cells . One set of experiments measuredthe specific activity of the intracellular pool ofnucleotide phosphates when open and closedcells were incubated with inorganic 32P, and theother experiments measured the actual amountof y-AT32P which entered resealed and leakyghosts (Tables II and III) . Both of these experi-ments indicated that the increased phospholipidmetabolism noted in leaky cells could not beaccounted for by differences in the intracellularlevel of y-AT32P (Tables II and III) .

As a working hypothesis it is postulated thatincreased phospholipid metabolism is associatedwith a repair mechanism of the membrane, andthat osmotic lysis or leakiness, which may resultfrom deleting or exchanging certain divalentcations, occurs by the opening of pores at certainregions along the membrane . It is envisioned


47

Dow


that lipoproteins are situated at these porous re-gions and that they contain phospholipids suchas phosphatidic acid and the polyphosphoinosi-tides . The opening and resealing of the pores onthe membrane may involve a phosphorylationand dephosphorylation of the lipid moieties ofthese lipoproteins .

This scheme takes into account observationsmade in this study. When the membrane isopen due to either osmotic lysis or treatmentwith EDTA, calcium, strontium, or cadmium,there is an increased incorporation of MP ofy-AT32P into the phospholipids. Other phospho-lipid metabolic reactions such as the acylation oflipids (Fig . 6) and the incorporation of inositol-3H (Table V) are not stimulated by increasedleakiness . a-Glycerophosphate is not incorporatedinto phospholipids, and there is no acylation ofphosphatidic acid and the polyphosphoinositides .Hence, the reaction which is stimulated is a

specific one, the phosphorylation of lipid moietiesleading to the formation of phosphatidic acidand polyphosphoinositide . Virtually all of theradioactivity in the polyphosphoinositides ismonoesterified; this coupled with the lack ofincorporation of a-glycerophosphate and fatty

acids into this lipid shows that there is little denovo synthesis but mainly a turnover of the phos-phate moieties .

This hypothesis predicts that there are certainregions in the membrane which are more sus-ceptible to lysis and which contain the functionallipoproteins involved in the opening and closingof "pores." This would explain why addeddigitonin does not stimulate these enzymes. Digi-tonin removes cholesterol (29), and this mayoccur at a different location on the membrane .Primaquine phosphate ruptures and reseals themembrane to give smaller vesicles (28) . Since themembrane is vesicularized or resealed by prima-

quine phosphate and does not remain open,

increased phospholipid metabolism is not ex-

pected, since this only becomes apparent whenthe membrane is open . The incorporation ceases

when the membrane reseals.

The technical assistance of Mr . P. Y. Liu is grate-fully acknowledged . Thanks are due to ProfessorAbraham Mazur of the New York Blood Centerand the City University of New York, and to Dr .Mabel Hokin of the University of Wisconsin formany helpful suggestions during the preparation of

the manuscript ; and also to Dr . John Derrick of theNew York Blood Center for performing the sodiumand potassium determinations .This work was supported by Grants AM13620

from the National Institute of Arthritis and MetabolicDiseases, and HE09011 from the National Heartand Lung Institute .

Received for publication 7 July 1970, and in revised form17 August 1970 .

REFERENCES

TEORELL, J . J. 1952 . J. Gen. Physiol . 35:669 .HOFFMANN, J. F. 1958. J. Gen . Physiol. 42 :9 .PAYSANT-DIAMENT, M ., and J. POLONOVSKI .

1960. Bull. Soc. Chim. Biol . 42:337 .DE GRIER, J ., E. MULDER, I . MULDER, andL. L. M. VAN DEENAN. 1961 . Proc. Kon. Ned.Akad. Wetensch . Ser . B . 64:274.

HOKIN, L. E., and M . R. HOKIN. 1964 . Biochim .Biophys. Acta. 84:563 .

VAN DEENAN, L. L. M., and J . D. GRIER .1964 . In The Red Blood Cell . C. Bishop andD. M. Surgenor, editors. Academic Press Inc .

SHOBET, S . B., D . F. NATHAM, and M . L. KAR-NOVSKY. 1968 . J. Clin . Invest . 47:1096.

WHITTAM, R., and M. E. AGER . 1964 . Biochern .J. 93 :337 .

GLYNN, I. M., and J . B. CHAPPELL. 1964 . Bio-chem . J. 90 :147 .

MCMURRAY, W. C., K. P . STRICKLAND,R. J . RossiTER . 1957 . Biochem . J. 66:634 .

FOLCH, J . 1956. Phosphorus Metabolism . Vol . 2 .H. McElroy and B. Glass, editors . The JohnsHopkins Press, Baltimore, Md .

DODGE, J . T ., C. MITCHELL, and D . J . HANAHAN .1963 . Arch . Biochem . Biophys. 100:119 .

BARTLETT, C. R. 1959 . J. Biol . Chem . 234:466 .DAWSON, R. M. C. 1960 . Biochem . J. 75 :45 .SANTIAGO-CALVO, E ., S . MULE, C . M. REDMAN,M. R. HOKIN, and L . E. HOKIN. 1964 . Bio-chem. Biophys, Acta. 84:550 .

DESJOBERT, A ., and F. PATEK. 1956 . Bull. Soc .Chim. Biol . 38:871 .

HOFFMAN, J. F. 1962 . Circulation . 26:1201 .WAYS, P. and D . J. HANAHAN. 1964 . J. Lipid

Res . 5:318.HOKIN, L. E. and M. R. HOKIN. 1964 . Funkt .

Morphol. Organ . Zelle. Wiss . Konf. 2nd.20. HOKIN, M. R. 1968 . Arch. Biochem. Biophys.

124:271 .21. KARNOVSKY, M . L., and D . F. H . WALLACH .

1961 . J. Biol . Chem. 236:1895 .22. SASTRY, P. S., and L . E. HOKIN. 1966 . J. Biol.

Chem. 241:3354 .

1 .2.3 .

4.

5 .

6 .

7 .

8 .

9 .

10 .

11 .

12 .

13 .14 .15 .

16 .

17 .18 .

19 .

48

THE JOURNAL OF CELL BIOLOGY . VOLUME 49, 1971

and

Dow


23 . TSAO, S . S ., and W. E . CORNATZER . 1967 . Bio-chem . Pharmacol. 16:249.

24 . FISHER, D . B ., and G . C. MUELLER. 1968 . Proc .Nat. Acad . Sci. U.S.A . 60:1396.

25 . MARINETTI, G. V. 1962 . J. Lipid Res. 3 :1 .26 . CRANE, R. K ., and F . LIPMANN . 1953 . J. Biol .

Chem . 201 :235 .

27. JACOB, H . S., and M. L . KARNOVSKY . 1967 .J. Clin . Invest . 46:173 .

28. GINN, F. L ., P . HOCHSTEIN, and B. F . TRUMP .1969. Science (Washington) . 164:843 .

29 . SCHMIDT-THOMé, J . 1942 . Hoppe-Seyler's Z .Physiol . Chem . 275:183 .

COLVIN M. REDMAN Phospholipid Metabolism in Erythrocyte Membranes 49

Dow


Documents

PHOSPHOLIPID METABOLISM IN INTACT AND MODIFIED …