6'-J. Membrane Biol 63, 71 75 (1981)

Sulfatide Role rn the Sodium Pump

Fernando Zambrano, Miguet Morales, Nelson Fuentes, and Mireya RojasDepartamento de Biologia, Facultad de Ciencias, Universidad de Chile, Casilla 653, Santiago, Chile

The Journa l o l

Membrane Biology

In the present paper we show that active sodiumefflux in human red cells after sulfatide hydrolysisis reduced. This inhibit ion is of similar magnitudeto that found in intact human red blood cells in theabsence of external Potassium.

Materials and Methods

Red blood cells isolated from heparinized fresh human blood,washed three t imes in isotonic sal ine solut ion, were loaded wi tht tNa. The red cel l loading, measurements of 22Na loss, and deter-mination of internal concentration of sodium ions of the loadedcells were carried out according to methods described previously

t3l.22Na-loaded red cel ls of f inal hematocr i t of I % were incubatedfor 10.20.40 and 60 min at 37 "C in Ringer 's solut ion, potassium-free Ringer 's solut ion, 100 mu potassium Ringer 's solut ion ' andRinger 's solut ion wi th ouabain (70 ru) . The same procedure wasrepeated. adding arylsulfatase or microdispersed sulfatide in themedium. The Ringer 's solut ion contained ( in mu): 10, KCl; 145,NaC l : 1 . MgC lz ; l . CaC l2 : and 2 .5 ' phospha te bu f f e r ( pH 7 4 ) 'In the polassium-free Rlnger 's solut ion or 100 mlvt potassiumRinger 's solut ion. the NaCl concentrat ion was adjusted to maintainthe same ionic st rength as that in normal Ringer 's solut ion. Sol idglucose u,as added to each incubation medium used at a finalconcentration of I 1 mv. The arylsullatase used [4] showed an en-richment of 1700-fold over the homogenate and a specific activityof 30 rmol of p-nitrocatechol sulfate hydrolyzed per min and permg protein. It is necessary to point out that the analysis of mito-chondrial phospholipids treated with 1 20 units of arylsulfatase bythin layer chromatography did not show any kind of lysocom-pound. Also, there is no protease contamination since rat kidneymitochondrial fraction crystalline bovine serum albumin fails toshow protein breakdown after 60 min of incubation. Furthermore,mitochondrial enzymes like DPNH cytochrome c reductase andcytochrome c oxidase were not inhibited by the highest amountof arylsulfatase used in the assay.

Loss of 22Na of red blood cel1s in Ringer's solution in thepresence of 0.30, 0.60, and 1 20 units of the enzyme (a unit isdefined as the amount of enzyme which hydrolyzes 1 pmol ofp-nitrocatechol sullate per min at 37 'C, pH 5.4) per ml of mediumwas measured. The highest amount of the enzyme used in theexperiments was also added to the mediums containing ouabainor sulfatide micelles.

Sulfatides were obtained by extracting rat brain with 20 vo-lumes of chloroform/methanol (2:1, vol/vol) at room temperature[14]. The residues were collected on a sintered glass, filtered, and

0022-2631 181/0063-0071 $01.00@) 1981 Springer-Verlag New York Inc.

Summary. Sodium efflux was studied in 22Na-loaded

red blood cells in the presence of arylsulfatase, anenzyme that specifically hydrolyzes sulfatide. Sodiumefflux was inhibited in proportion to the amount ofarylsulfatase present. Maximum inhibition was al-most as high as the efflux obtained in medium withK+ absent. At maximum inhibit ion 83.2%o of theslrlfatide content of the fragmented red blood cellmembranes was hydrolyzed and ouabain-sensitive(Na+ +K+)-ATPase activity was inhibited by 100%.Sodium efflux, sulfatide content, and (Na"+K*)-ATPase activity were unaffected with arylsulfatasein the presence of a high concentration of sulfatide.These results indicate that sulfatide plays a specificrole in sodium and potassium ion transport. Theyalso suggest that most sulfatide is iocalized externallyin the red blood cell membrane.

Key words: Sulfatide, function, red cells, sodiumtransport

Analyses of the sulfatide content in rectal glands ofspiny dogfish U0l, bovine kidney medulla [9], saltglands of either duck or herring gull [11], and ofhuman erythrocyte membranes [6] suggest that sulfa-tide may act as a l ipid requirement for (Na- *K*)-ATPase activity. In a previous paper [4] we showedthat ouabain-sensitive ATPase activity in frog skinduring larval development is fully inhibited. We alsoestablished a correlation between (Na*+K*)-ATPase activity, sulfatide content, and sodium flux,indicating that sulfatide must be involved in the sodi-um pump mechanism. The sulfatide cofactor model[8] postulates affinity between sulfatide and potassiumions. As the affinity site for potassium on the externalside of the membrane would be lacking when sulfatideis absent, total inhibition of potassium influx wouldbe expected.

72

eo

g6

I.Ec

'Eo

z

o

EL

q

ooo=o

o

;:EEo

oz

5

L

Fig. l. The effect of arylsulfatase on the 22Na loss from freshhuman red blood cells. The cells were incubated for up to t hra t37 "C . ( o ) R inge r ' s so l u t i on ; ( o ) l 00mu K R inge r ' s so l u t i on ;(r ) Ringer 's solut ion wi th ouabain (70 pu); ( r ) Ringer,s solut ionplus ary lsul fatase: ( l ) 0.30 uni ts/ml ; (B) 0.0 uni ts/ml ; (C) 1.20units/ml. The correlation coefficients ranged from 0.9g to 0.99

re-extracted with 10 vol chloroform/methanol. The two extractswere combined, evaporated to a smal l volume, and f reed of nonl i -pd contaminants by chromatography on Sephadex G-25 [15] .

The l ip ids, e luted wi th chloroform/methanol (19:1, vol /vol )saturated with water, were evaporated by means of a flash evapora_tor under reduced pressure to a moist residue and resuspendedin a smal l volume of chloroform. Neutral l ip id, cerebroside, andsulfatide were separated from the total lipid extract by silicic acidchromatography column (Unis i l , Clarkson Chemical Co., Wi l_liamsport, Pa.), using as eluents chloroform, chloroform/acetone(1:1, vol /vol ) , and acetone, respect ively [15] . The acetone_elutedsulfatide fraction was evaporated under nitrogen, and the moistresidue was resuspended in a srall volume of chloroform/methanol(2: l , vol /vol ) and stored af -20"C. The sul fat ide thus obtainedwas microdispersed in aqueous medium by ultrasonic irradiation[1]. To clarify it, the microdispersion was spun down at 20,000 rpmin the N"40 Spinco rotor for 30 min. Sulfatide determinations werecarried out as previously described [12], using bovine sulfatide(Applied Science Laboratories) as a standard.

(Nar+K+)-ATPase act iv i ty at 37 "C was tesred [4] in thefragmented membranes [5] of the loaded red cells incubated for60 min in different mediums. To determine the sulfatide break_down, sulfatide was extracted and determined in the loaded redcells incubated for 20, 40 and 60 min with 0.30 to 1.20 unit ofarylsulfatase in Ringer's solution by the method mentioned above.The protein concentration was obtained using crystalline bovineserum albumin as a standard 113] .

Radioactivity was measured in 1 ml of the supernatant dis_solved in 10 ml of a Tr i ton X-1O0-Toluene l iquid scint i l la t ion mix-ture (333 ml Tr i ton X-100, 666 ml Toluene, 4.0 g ppO and 0.05 gPOPOP) in a Philips Specrromerer, p.W. 5403.

For each set of conditions the plots of 22Na loss against timewere extrapolated to zero time to estimate the extracellular 22Napresent at the beginning of the incubation period. The fractionsof the initial intracellular radioactivity that remained inside thecells at each time were calculated and semilog plotted against time.Apparent rate constants were taken as the slope of the straight

F. Zambrano et al.: Sulfatide Role in Sodium Transnorr

Fig. 2. The effect of sulfatide on the 22Na loss from fresh humanred blood cells. The cells were incubated for up to I hr at 37 "C.(o) Ringer 's solut ion; Ringer 's solut ion plus sul fat ide: ( t ) 2.5 tg lm l ; ( n ) 5 .0 pg /m l ; ( o ) 10 ,0 r g /m l l ( r ) R inge r ' s so l u t i on -a r y l su l f a -tase 1.2 uni ts/ml p lus sul fat ides: ( l ) 5.0 Lg/ml; (B) 10.0 rg/ml ;(C) 20.0 "Lg/ml

l ines obtained. These t ,a lues and the mean Na content values wereused to calculate efflux.

Results

All f igures show the fraction of 22Na remaining inred cells, plotted as a function of t ime. The slopesof the curves give the apparent rate constants of theeffluxes. Figure I represents the loss of 22Na inRinger's solution. 100 mu K Ringer's solution, oua-bain, and in Ringer's solution with different amountsof arylsulfatase. This figure shows that as the amountof arylsulfatase increases, the slope of the 22Na losscurve diminishes and approaches the slope of thecurve for 22Na loss in K-Free Ringer's solution(Fig. 3). The 22Na loss curves for cells in Ringer'ssolution with microdispersed sulfatide with and with-out 1.20 units of arylsulfatase, respectively, are shownin Fig. 2. The different amounts of sulfatides usedapparently do not change the slope of the curves.On the other hand, the action of arylsulfatase gradual-ly diminishes when the sulfatide concentration in-creases in the medium. The slope of the 22Na losscurve for 20.0 tg of sulfatide is higher than that ob-tained in Ringer's solution alone.

Figure 3 shows the curve in K-Free medium with1.20 units of arylsulfatase, and the curves for a medi-um with 1.20 units of arylsulfatase in Ringer's solu-tion or 100 mlr K Ringer's solution, as well as the22Na loss curves in K-free medium with and withoutouabain. It is evident that the slope of the 22Na loss

oo

a

E

Eo

z

o

.9

L

F, Zambrano et al.: Sulfatide Role in Sodium Transport

Fig.3. The ef fect of 1.20 uni ts/ml ary lsul fatase on the 22Na lossfrom fresh human red blood cells. The cells were incubated forup to t hr at 3 l "C. (o) K-f ree Ringer 's solut ion; ( l ) 100 mrrK Ringer's solution plus I.2 units arylsulfatase; K-free Ringer'ssolut ion wi th: (e) ouabain; ( r ) 1.2 uni t ary lsul fatase; 1.2 uni tsary lsul fatase plus ouabain in: ( r ) 10mu K Ringer- 's solut ion:(r ) 100 mrr l K Ringer 's solut ion

curve with ouabain in Ringer's solution or 100 mlrK Ringer's solution is unaffected by 1.20 units ofarylsulfatase. Moreover, the slopes of the curves inthe K-free medium with or without 1.20 units of aryl-sulfatase are very similar.

Effluxes calculated from the slopes of the 22Naloss curves (figures) under the studied conditlons areshown in Table l. The highest amount of arylsulfatasein Ringer's solution caused a reduction of abolf 5Jo/oin the sodium efflux, which is close to the reductionobserved in the K-free medium (52%). Ir may alsobe seen that with 1.20 units of arylsulfatase, the Na+effluxes from red cells in K-free Ringer's solutionand Ringer's solution with or without ouabain werenot significantly different. On the other hand, in nor-

Table 1. Effecr of arylsulfatase on sodium efflux in red blood cells

I J

mal Ringer's solution, 0.30 and 0.60 units of arylsulfa-tase reduce the sodium efflux to about 13 and 43oA,respectively. Again, the highest amount of arylsulfa-tase added reduced the active sodium effluxes definedas the ouabain-sensitive sodium efflux to abouf. l4oA,while the reduction observed in the potassium-freemedium is of the order of 72'A. The sodium effluxesof red cells in Ringer's solution in the presence ofsulfatide microdispersed with or without the additionof 1.20 units of arylsulfatase are shown in Table 2.The addition of 2.5 to 10.0 rg of sulfatide reducesthe sodium efflux 1 to 4%, respectively. Theseamounts of sulfatide moreover produced 1.60 to1.10% of hemolysis. Table 2 also shows that the inhi-bit ion induced by 1.20 units of arylsulfatase (Table 1)is diminished in the presence of sulfatide. Thus, 10 pgof sulfatide in the medium decreases the inhibitionlevel of 610/o of the sodium efflux caused by the aryl-sulfatase to about 20o/o, and with 20 pg of sulfatidethe sodium efflux is 2lo/o higher than the efflux inRinger's solution alone, even with the etzyme present.It should be pointed out that when arylsulfatase isused in the medium in a concentration higher than1.20 units, 360/o of the cells were hemolyzed after60 min after incubation. Under these conditions theefflux appears to diffuse freely.

Table 3 summarizes the sulfatide breakdown mea-sured in the fragmented membrane isolated from theloaded red cells incubated for 20, 40 and 60 min inRinger's solution with 1.20 units of arylsulfatase, re-spectively (i.e., from the same red cells from whichthe 22Na loss curves were obtained), and the (Na+ fK*)-ATPase activity after 60 min incubation. The re-sults show a typical enzymatic reduction. They alsoindicate that even the highest sulfatide breakdownlevel (83.2%) does not affect the ouabain-insensitiveATPase. On the other hand, the ouabain-sensitiveATPase is reduced by 31 and 51% with the use of

T i m e ( m i n )

Aryl-sulfatase(units/ml)

Ringer's Inhibitionsolut ion (%)

Ouabain Inhib i t ion(70 pu) (" / " )

K-l'ree Inhibitionmedium (%)

100 mv Kmedium

Na el'flux (mmol/liter cells/hr)

Active Na efflux(mmo1/liter cells hr)

None0.300.601 . 2 0None0.300.601.20

3.342 . 9 12.201 . 4 5

2.442.011 . 3 00.55

12.8734.1356.59

17.6246.7277.46

0.90

0 . 8 50 .00

0.00

13 .05

7 4 . 5 5

r 00.00

1 00.00

1 . 5 9

l . 5 30.69

0.3

52.39

54.L91 l .72

7 4 . 1 8

1 7 /

1 . 6 12.44

0.7 |Na effluxes in red cells suspended in different media expressed as mmol/liter cells/hr. The values were calculated from the slopesof 22Na loss curves (figures) and represent the mean of 4 experiments. The reproductibility of the estimated effluxes was better than+ 0.03 mmol/ l i ter cel ls /hr .

74

Table 2. Effect of sulfatide on sodium efflux in red blood cells

F. Zambrano et al.: Sulfatide Role in Sodium Transport

Arylsulfatase(units/m1)

Sulfatide(pe/ml)

Ringer'ssolution

Inhibition(%)

Ouabain(70 pu)

Hemolysis rangeof values(%)

Na efflux(mmol/liter cells/hr)

Active Na efflux(mmol/liter cells/hr)

NoneNoneNone1 .20Noner .201 .20

NoneNoneNone1 .20None| . 20r . 20

None2.55.05 .0

10.010.020.0None2 . 55 .05 .0

1 0 . 01 0 . 020.0

0.000.903 . 8 9

60.783.59

20.360.00

1..224 . 1 5

82.984.92

27.8'70.00

N . D .0.00-1.601.80-2.50N.D.5 .40 7 .10N.D.N.D.

3 . 3 43 . 3 13 .211 . 3 1

2.663 . 8 6

0 .90

2.442.412 . 3 10.412.321 . 7 62.96

100.00

Na effluxes of red cells in the presence of sulfatide microdispersion. Data are the mean values of 4 experiments expressed as mmol/litercells/hr calculated from the sloped of 22Na loss curves (figures).

N.D.:not detected.

Arylsulfatase Total Ouabain-ATPase insensitive

ATPase

% Sulfatidedegradation afterincubation (min)

Table 3. Effect of arylsulfatase on the loaded erythrocyte mem-branes

erage of ten experiments). This value remains un-changed after 60 min incubation at 37 "C with orwithout 20 pg of sulfatide in Ringer's solution.

Discussion

Interest has been focused on the role of sulfatidein the sodium pump mechanism since red blood cellmembranes have been shown to contain sulfatide,[6]and sulfatide hydrolysis known to inhibit ouabain-sensitive (Na+ +K*)-ATPase activity [a]. The presentresearch compares normal sodium efflux with thatobtained in the presence of different amounts of aryl-sulfatase.

The demonstration that membranes of loaded redcells treated with arylsulfatase show sulfatide break-down, that cells are not hemolyzed when sodium ef-flux is reduced by about one third, and that the addi-tion of sulfatide is able to protect the cells from en-zyme actton, seem to indicate that sulfatides are locat-ed externally in the red blood cell membrane andare involved in the sodium pump.

The breakdown of 83.2% of sulfatide that pro-duced the highest reduction of the efflux seems toindicate that not all the sulfatide content functionsin the sodium pump. The use of more than 1.20 unitsof enzyme hemolyses over 36oA of the cells, and the22Na loss appears to occur by free diffusion. Theseresults seem to confirm that sulfatide functions inmaintaining the integrity of the membrane.

One-for-one sodium exchange and reduction insodium efflux by about a third [7] has been shown

None0.30 unit0.60 unit1.20 unit1.20 unit plus

20 pg sulfatide

1 . 3 91.261 . 1 70.931 . 6 7

0.940.950 . 9 50.930.96

N.D. N.D. N.D.21.5 40.1 58.927.2 46.6 69.232.6 57.r 83.2N.D. N.D. N.D.

The loaded red cells incubated for 20, 40 and 60 min af 37 "Cin different mediums were fragmented. Half the final volume of60 min was used to measure the ATPase activites. The other, 20and 40 min, were extracted with chloroform/methanol (2:1, vollvol). In the isolated glycolipid fraction, sulfatide amount was mea-sured.N.D.:not detected. Activities, expressed as pmol P/mg proteinhr 1. The sulfatide content of cell suspension was 7.75 -Lg/ml. Dataare the average of 4 experiments.

0.30 and 0.60 units, respectively. Total inhibit ion ofouabain-sensitive ATPase is obtained with a 83.2oof sulfatide breakdown.

No breakdown of the structurally present sulfatidein the membrane was observed with 1.20 units ofarylsulfatase in the incubation medium including20 ltg of sulfatide even after 60 min of incubation.Under these conditions ouabain-sensitive ATPase isincreased by 53'h. The sulfatide content of 1 ml cellsuspension with a hematocrit of lo/o was 7 .l 5 pg (av-

F. Zambrano et al.: Sulfatide Role in Sodium Transport

[3] in the absence of external potassium. Similar re-duction is obtained in the presence of arylsulfatase.

In ATPase crude preparations treated with aryl-.

sulfatase, preliminary experiments show that thefuphosphorylation as well as the ouabain binding are

blocked. On the other hand, it has been reported

75

ization ol a new photoalfinity derivative of ouabain: Labelingof the large polypeptide and of a proteolipid component ofthe Na, K-ATPase*. Biochemistry. 17:3667 3676

t 3. Garrahan, P.J., G1ynn, I.M. 1967. The behavior of the sodiumpump in red ells in the absence of the external potassium.J. Physol. (London) 1922159 114

4. Gonzi iez, M., Morales, M., Zambrano, F.1979. Sul fat ide con-

l2lthat the ouabain binding site of the ATPase may *,lI tent and (Na*+K*)-ATPase activity of skin and gill duringinvorve a proteolipid component of the enzyme, b; fl' t:::;t,::;",iff;:;t,\if ;f::*f:;nCatvptocephatetta

caudi-it is unknown if sulfatide is present in this component. s. rtunut uo, D.J., Ekholm, J.E. 1974. The preparation of red

Thecofactor site model [8] together with our sodi- cell ghosts (membranes). 1z: Methods in Enzymology. S. Flei-um efflux findings in human red blood cells might scher and L. Parker, editors. Vo1. 31, parr A, p. 168-172. Aca-suggest that sulfatides in the sodium pump are in-

-

d_emic Pess, New York

voived in speciric K+ binding or in the ariinity site 'filliliJ;,,i"%5,ii[1,1il.I;f ;,i,1TJ'llilft,j1,: lt##l:on the outside of the red cell membrane. Further and their relation to sodium-potassium dependent adenosinepotassium influx experiments are needed to distin- triphosphatase. J. Biochem. 33:813-819guish between these two possibilities, since the evi- 7. Hoffman, J.F., Kregenow, F M. 1966. The characterizationence at present is too iragmentary ror any clear :.,iJffi:?f:t;::;.\::\Zi5#:H,processesinredbloodchoice' 8. Karlsson, K.A. 1916. Aspects on structure and function of

sphingolipids in cell surface membranes. 1n; Structure of Bio-

Conclusions logical Membranes. S. Abrahamson and O. Pascher, editors.p.245-276. University of Goteborg, Goteborg, Sweden

rhe resutts reported in this work sussest the rouowin s e SXliilli;,"kJ":T,:',':i;;j,3 ;j,lli';""t i _131,i" l,ljconclusionsl pu pilla. Biochim. Biophys. Acta 316:317-335

a) Sulfatide is a lipid requirement for the sodium 10. Karlsson, K.A., Samuelsson, B.8., Steen, G.O. 1974. The lipidpump of the red blood cell. composition of the salt (rectal) gland of spiny dogfish. Biochtm.

b) The majority of the sulfatide content is located ,, !,'?o.hr'' Acta337:356-376in rhe exrernat monolayer or the red blooJ i.i-',"o': " fflillft"X i;jffil'Ilil l;l;;lll';';f,?;,}'tti;,ffj,ifli1brane. tase activity of the salt (nasal) gland of eider duck and herrrngThe authors wish to thank Dr. p. Garrahan for his herpful sugges-

t]l);rkt.t)2r']irs;]nhatides in sodium ion ansport' Eut' J'tions. The competent technical assistance of Mrs. M.I. Navarrete | 12.Kean, E.L. 1968. Rapid, sensitive spectrophotometric methodand Mr. E. Gonzlez is also acknowledged. for quantitative determination of sulfatides. J. Lipid. Res.

This work was supported by the joint Program of Comisin 92319-32:'Nacional de Investigacin Cientf ica y Tecnolgica de Chile, | 13. Lo*ry,O.H.,Rosebrough,N.J.,Farr,A.L.,Randall ,J.R. 1951.CONICYT-National Science Foundation, Servicio de Desarrollo protein measufement with the Folin phenol reagent. J. Biol.Cientfico, Artistico y Cooperacin Internacional de la Universidad Chem. 193:265-2j5de Chile (Grant ff 8952-8123) and PNUD/UNESCO RLA Pro- 14. Rouser, G., Fleischer, S. 1967. Isolation, characterization, andgram 076/006 (Granf ft12). determination of polar lipids of mitochondria. ir: Methods

in Enzymology. R.W. Estabrook and M.E. Pullman, editors.Vol. 10, p. 385-406. Academic Press, New York

ReferenCes 15. Rouser, G., Kritchevsky, G., Yamamoto, A. 1967. Columnchromatographic and associated procedures for separation and

1 . Fleischer, S., Fleischer, B. 1967 . Removal and binding of polar deteimination of phosphatides and glycolipids. 1r : Chromatog-lipids in mitochondrial and other membrane systems. ln. Meth- raphy Analysis. G.V. Marinetti, editor. pp. 99-120. Marcelods in Enzymology. R.W. Estabrook and M.E. Pullman, edi- Dekker, New york

tors. Vol. 10, p.406433. Academic Press, New York'[ 2. Forbush,8.. I I I . , Kaplan, J.H., Hoffman, J.F. 1978. Character- Received 22December 1980; revised 29 Apri l 1981