1008 Blood, Vol 68. No 5 (November), 1986: 1008-1014

Studies on the Effect of Vanadate on Endocytosis and Shape Changes

in Human Red Blood Cells and Ghosts

By S.L. Schrier, Irene Junga, and Lisa Ma

When amphipathic cationic drugs are added to intact

human RBCs. the RBCs first undergo a stomatocytic shape

change and then. if relatively large amounts of drug are

added and if the metabolic state of the RBC is appropriate.

endocytic vacuoles form. Vanadate has a structural similar-

ity to the transition state of phosphate. which presumably

accounts for its ability to inhibit phosphohydrolases.

although other actions of vanadate have been described.

Vanadate inhibited three forms of drug-induced endocyto-

sis in intact RBCs despite the fact that the three drugs

chosen (primaquine. chlorpromazine, and vinblastine) are

known to have differing requirements for RBC ATP. Vana-

date also inhibited the stomatocytic shape change pro-

duced by primaquine, chlorpromazine, and vinblastine, but

W HEN AMPHIPATHIC anions and cations are added

to RBCs in vitro, they produce, respectively, echino-

cytosis and stomatocytosis.’ Somewhat similar shape altera-

tions also occur under certain circumstances in vivo.2 Stoma-

tocytosis can be extended in vitro to proceed to the formation

of endocytic vacuoles, which are inside-out portions of the

RBC plasma membrane that form preferentially in the

invaginated portion of the cup.3 The irreversible loss of

membrane surface area into endocytic vacuoles reduces the

surface area/volume ratio’ and results in the formation of a

spherostomatocyte. Continuation of the process results in

RBC lysis.3’4 A role of RBC metabolism in this in vitro

process was deduced from the observation that glucose

deprivation could prevent primaquine-induced hemolysis.4

Some forms of drug-induced endocytosis (ie, that produced

by chlorpromazine) can occur in ATP-depleted RBCs, an

observation that tended to uncouple the relationship between

RBC metabolism and drug-induced RBC shape change.5

Subsequently, it was noted that most forms of drug-induced

stomatocytosis and endocytosis in vitro take place more

readily in glucose-supplemented, ATP-replete RBCs.6 It has

been suggested that ATP has an important but poorly

understood role in controlling intact RBC shape.7 On the

basis of experiments performed with ghosts, a shape-mediat-

ing role of Mg-ATP8 acting through the membrane Mg’�-

ATPase has been proposed.9

These observations led us to suspect that there is a role for

ATP and perhaps ATP hydrolysis in drug-induced stomato-

From the Division of Hematology. Stanford University MedicalCenter, Calif

Supported by US Public Health Service grant No. AM13682.Submitted Jan 13, /986; accepted May 27. /986.

Address reprint requests to Dr Stanley L. Schrier, Division of

Hematology. Stanford University Medical Center, Stanford, CA

94305.

The publication costs ofthis article were defrayed in part by page

charge payment. This article must therefore be hereby marked

“advertisement” in accordance with 18 U.S.C. §1734 solely to

indicate this fact.

.�) I 986 by Grune & Stratton. Inc.

0006-4971/86/6805-0005$03.00/0

not the stomatocytosis produced by low pH. Vanadate had

no effect on RBC echinocytosis produced by lysophosphati-

dylcholine. In studying endocytosis in hypotonic, leaky.

“white” ghosts. we discovered that vanadate inhibited

only the endocytosis produced by Mg-ATP and not the

endocytosis produced by manipulations that directly attack

the cytoskeletal proteins. These findings suggest that ATP

hydrolysis has a role in some forms of amphipathic cation-

induced stomatocytosis and endocytosis in intact RBCs. In

addition, studies in ghosts support the idea that Mg-ATP

does indeed produce “energized” endocytosis dependent

on utilization or hydrolysis of ATP.

a 1986 by Grune & Stratton, Inc.

cytosis and the subsequent endocytosis in intact RBCs. An

opportunity for evaluating the potential role of ATP hydroly-

sis in RBC stomatocytosis and endocytosis was provided by

recent observations on the properties of the several forms of

vanadium. Ionic vanadium, under certain circumstances, has

a structural similarity to the transition state of phosphate,

thereby perhaps accounting for its ability to inhibit phospho-

hydrolases at micromolar concentrations.’#{176}” Vanadate inhi-

bition of RBC membrane Na�, K�-ATPase’2 and Ca2�,

Mg2�-ATPase’3”4 has been reported, whereas incubation of

vanadate with intact RBCs has inhibited the sodium pump in

a manner consistent with inhibition of the pump-related

RBC membrane Na�, K�-ATPase.’2 Therefore, we tested a

hypothesis that ATP hydrolysis was important in drug-

induced stomatocytosis and endocytosis in intact human

RBCs by studying the effect of vanadate on RBC shape

change and endocytosis. In addition to inhibiting the phos-

phohydrolases, vanadate can inhibit phosphatases, stimulate

adenyl cyclase’5 and, at very high concentrations, can pro-

duce RBC metabolic alterations’6”7 and even cause shape

changes.’7 Therefore, experiments using vanadate must be

carefully controlled and interpreted.

Use of vanadate also offered us the opportunity of testing a

hypothesis on endocytosis produced in white ghosts.”9 We

and others had identified two forms of endocytosis in white

ghosts. One form, called “nonenergized” endocytosis, is

produced by hypotonicity, EDTA, or trypsin, all of which

appear to act by stripping away the cytoskeleton.’9 The other

form is produced by addition of Mg-ATP and has been called

“energized � although the mechanism of ener-

gization has not been identified. If vanadate blocked Mg-

ATP-induced energized endocytosis, it would mean that

ATP hydrolysis was probably required.

MATERIALS AND METHODS

All reagents used were reagent grade or better. Sodium ortho-vanadate was obtained from Fisher Scientific Company, and its

concentration was verified.�#{176}The solutions were adjusted to pH 7.5,

and only solutions that were clear and devoid of color were used.”Carrier-free 32Pi was obtained from New England Nuclear Corp(Boston). Lysophosphatidylcholine was obtained from Sigma Co, St

Louis. Irypsin was obtained from the Worthington Chemical Co,

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

VANADATE, RBC, AND GHOST SHAPE CHANGE 1009

Freehold, NJ. RBC membrane actin was purified as previously

described.22 Silicon 550 Fluid was obtained from Dow Corning.

[G-’H] Vinblastine sulfate-3H 11.4 Ci/mmol was obtained fromAmersham International, Amersham, England. Chlorpromazine

HCI (benzene ring-’H) 39.1 Ci/mmol was obtained from New

England Nuclear. Protosol and Biofluor were also obtained from

New England Nuclear.

Freshly drawn, heparinized venous blood was obtained from

normal human donors according to a protocol approved by the

Stanford University Committee on Human Experimentation. Theblood was centrifuged at room temperature, and the plasma and

buffy coat were removed, after which the RBCs were washed once

with 15- to 20-fold volumes of 0.15 mol/L of NaC1 and then three

times with either cold Hanks’ solution or phosphate-buffered saline

(PBS) pH 7.4 containing 2 mg glucose/mL.

Ghost preparation. White ghosts were made exactly as previ-

ously described, using three washes of 5 mmol/L of phosphate buffer

pH 8.0 at a ratio of 40 vol oflysis solution to I vol of packed RBCs.2’The protein concentration was determined by the method of Lowry,

using five times recrystalized bovine serum albumin (BSA) as

standard.

A TPase assays. The several different ATPase activities dis-

played by ghosts were measured exactly as previously described

using ‘y ‘2P-ATP as substrate.22 The ‘2Pi generated as a result of

ATPase action was quantitatively recovered in the isobutanol-benzene extract.2’ Four ATPases were measured. All assays con-

tamed -100 �tg of membrane protein, 2.5 mmol/L of Mg2�, and 1.5

mmol/L of ‘y-’2P-ATP, which defined the Mg2�-ATPase. For theNa�, K�-ATPase, 30 mmol/L of K� and 100 mmol/L of Na� were

added. For the Ca2�, Mg2�-ATPase, 0.12 mmol/L of Ca2� was

added. For the actin-activated ATPase, 100 zg of RBC membraneactin was added to 100 �g of RBC membrane protein.

RBC endocytosis. Endocytosis was induced in intact RBCs by

adding primaquine, chlorpromazine, and vinblastine as described

previously.”6’24 To test the effect of vanadate on RBC endocytosis, it

was preincubated with RBCs suspended in plasma or Hanks’ solu-

tion for 30 to 60 minutes at 37 #{176}Cto allow passage of vanadate into

the RBC interior.’2”5 At the end of the preincubation period, theendocytosis-inducing drug was added, and the timing of the incuba-

tion was begun and continued at 37 #{176}Cfor the length of time

indicated in the text and Tables I through 6. The extent of

endocytosis was monitored by phase microscopy and by the radioiso-

topic method, which is based on the trapping of 57Co-vitamin B,2

within endocytic vacuoles.’6’24 Preliminary experiments indicated

that vanadate, at the concentrations used (5 to 100 izmol/L),produced neither RBC shape changes nor endocytosis, nor did it

interfere with the binding of the 57Co-vitamin B,2-transcobolamin

complex to the RBC surface. AlP levels were measured by thecoupled hexokinase-G6PD assay25 in intact RBCs before and afterincubation with the endocytosis-inducing drugs alone or in combina-

tion with vanadate. It was important to determine whether the

inhibitory effects of vanadate were reversible. Therefore, RBCs as

described above were incubated without and with vanadate at 37 #{176}C.

Then, relying on the published information of vanadate effiux,’5 wewashed the RBCs twice with a 20-fold volume of Iris-buffered salinepH 7.2 containing 1% BSA. We then incubated them overnight at

37 #{176}Cat a hematocrit of 10 in Hanks’ solution containing 10%autologous plasma, 5 mmol/L of glucose, and penicillin 200 U/mLand streptomycin 200 �zg/mL. In the morning, the RBCs were

washed again three times with Hanks’ solution and were then

studied as above for their capacity to undergo cationic amphipath-

induced RBC shape change and endocytosis.

RBC shape changes. To produce echinocytosis, intact washedRBCs were incubated in PBS-glucose (2 mg/mL) at 2% hematocrit

with 6 �g/mL of lysophosphatidylcholine for up to 30 minutes at

37 #{176}Cwithout or with the addition of vanadate. Samples of 20 �zL

were removed at given time intervals from 5 to 30 minutes, added to

200 �zL of 1% glutaraldehyde in PBS, and then examined by phase

microscopy. The proportion of RBCs of different morphologies was

recorded using the nomenclature and classification of Bessis.2 The

effect of vanadate on shape change reversibility was then measured

by removing the suspending media and washing the RBC pellet

serially with either 1% BSA in PBS, or 1% BSA plus 30 .tmol/L of

vanadate in PBS. Morphologic assessment by phase microscopy was

made of the RBCs after each wash. Selected samples were alsophotographed by Nomarski interference microscopy. The effect of

vanadate on induction of stomatocytosis was also studied. The

amphipathic cations, primaquine, chlorpromazine, and vinblastine,

were used to produce stomatocytosis; the incubations were carried

out as described for endocytosis. Acid-pH was also reported by

Deuticke’ to induce stomatocytosis. Therefore, washed intact RBCs

at a hematocrit of 2% were incubated in PBS pH 7.4 as a control, andthen with both 0.1 mol/L of P04 buffer pH 5.2 or citrate buffer 0.1

mol/L pH 5.2 at 37 #{176}Cfor 30 minutes without or with concurrent

addition of 30 �zmol/L of vanadate. Samples were obtained for

morphologic analysis. Reversal of shape change was then performed

by washing either with 1% BSA in PBS pH 7.4 or 1% BSA plus 30

�mol/L ofvanadate in PBS pH 7.4.Drug entry into RBCs. The effect of vanadate on the uptake of

chlorpromazine and vinblastine was measured by a method previ-

ously used, with a single modification.26 Instead of diluting and

washing the RBC pellet after incubation, we layered the samples on

top of a I-mI silicon barrier in a I .5-mL Eppendorf centrifuge tube

and immediately centrifuged them at room temperature for 2

minutes at full speed. The clear supernatant formed a distinctly

visible band above the barrier, and the RBC pellet was densely

packed at the bottom of the tube. The top supernatant fraction was

carefully removed and added to a scintillation vial. The RBC pellet

was harvested by cutting off the tip of the tube and placing it in a

scintillation vial. One milliliter of Protosol/ethanol ( 1 :2) was added

to all vials, which were digested at 60 #{176}Cfor 90 minutes and thenallowed to cool, after which 0.3 mL of 30% H2O2 was added and the

samples were stored overnight at room temperature. Ten milliliters

of Biofluor was added, and the vials were shaken; then 0.5 mL of 0.5N HCI was added. The mixture was shaken, allowed to cool for 2

hours, and then counted by liquid scintillation spectrometry in a

Beckman LS 230 counter. The specific activity of endocytosis-inducing drugs was adjusted by adding radioisotopes to “cold” drugs

to provide a stock standard solution. The chlorpromazine specific

activity used was I . 13 x 106 cpm/izmol, and the vinblastine specific

activity was 3.46 x l0� cpm/�zmol.

Ghost endocytosis. Endocytosis in white ghosts was induced by

three methods.’9 Addition of 20 ng/mL of trypsin or 0.1 mmol/L of

EDTA produces endocytic vacuoles presumably by stripping off

segments of the cytoskeletal proteins, ie, nonenergized endocytosis.

Conversely, addition of 3 mmol/L of Mg-ATP to leaky ghosts

suspended in 50 mmol/L of TES pH 7.5 produces energized

endocytosis.” The extent ofendocytosis was measured semiquantita-

tively by observing coded ghosts samples under phase microscopy

and quantitatively by the acetylcholinesterase method originallydescribed by Jarrett and Penniston.27 In all reported experiments,

the semiquantitative morphologic assessment was in excellent agree-

ment with the results of the acetylcholinesterase assay. Vanadate

had no effect on the acetylcholinesterase assay (data not shown). Todetermine if the effect of vanadate was reversible, ghosts incubated

without and with vanadate were washed three times in 50 mmol/L of

TES and then restudied for their capacity to undergo Mg-AlP-

induced endocytosis.

Radiophosphorylation studies ofghosts during endocytosis. Inattempts to explore the molecular mechanism by which Mg-ATP

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

1010 SCHRIER ET AL

might cause ghost endocytosis, RBCs were radiolabeled by incubat-

ing them under sterile conditions for 20 hours with 100 to 500 zCi of32P/mL packed RBCs.2’ Ghosts were then prepared and incubatedas above with 50 mmol/L of TES buffer, and 3.0 mmol/L of

Mg-ATP without and with the addition of 30 MmoI/L of vanadate to

block endocytosis. Following incubation at 37 #{176}Cfor a time suffi-cient to cause extensive endocytosis in the samples free of vanadate,

ghosts were separated and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on slab gels usingthe Laemmli system as previously described.22 The gels were ana-lyzed by densitometry, bands were cut out and weighed, and the

proportion of total proteins was recorded. The extent of radiophos-phorylation of the membrane proteins was then assessed semiquanti-

tatively by 32P-radioautography and quantitively by slicing out the

relevant bands, dissolving them in 600 �tL of 30% H2O2, addingaquasol, and measuring the radioactivity by liquid scintillationspectrometry. From the radioisotopic data and protein concentra-

tions, the 32P specific activity of the individual phosphorylated

proteins could be calculated.

RESULTS

A TPase assays. The ATPase assays were carried out in

duplicate as described in the Materials and Methods section,

and the samples were incubated 60 mm at 37 #{176}Cboth

without vanadate and with the addition of 2 and 20 smol/L

of vanadate at the start of the incubation period. The results

(Table I) showed that all the ATPases were strongly inhib-

ited by vanadate at 20 �zmol/L and that Mg2�-ATPase

showed the greatest sensitivity, being inhibited 53% at 2

�mol/L of vanadate.

Ghost endocytosis and the effect of vanadate. Experi-

ments were designed to test the effect of vanadate on

energized endocytosis as produced by addition of 3 mmol/L

of Mg-ATP, or on nonenergized endocytosis produced by

incubation of ghosts with 20 ng/mL of trypsin or 0.!

mmol/L of EDTA. Incubation was carried out (as described

in the Materials and Methods section) at 37 #{176}C,with sam-

pies observed at 0, 1 5, 30, 45, and 60 minutes. In each case, a

sample with buffer alone and another with an appropriate

concentration of vanadate used were incubated in parallel.

Acetylcholinesterase results were correlated with indepen-

dent coded morphologic observations. Vanadate in concen-

trations of 30 �tmol/L had no inhibitory effect on endocytosis

produced by either EDTA or trypsin. In contrast, it pro-

foundly inhibited Mg-ATP endocytosis (Table 2A). This

effect of vanadate was completely reversible. Ghosts incu-

bated with 30 �imoi/L of vanadate were washed three times

with 50 mmol/L ofTES, after which their ability to undergo

Mg-ATP-induced endocytosis was unimpaired as compared

with values of ghost suspensions incubated in parallel but

without prior vanadate addition (Table 2B). Ghosts were

Table 1 . Effect of Vanadate on RBC Membrane ATPases

VanadateConcentration

(�&mo1/L) Mg”-ATPase

Na’ .K�-ATPase

Ca”.

Mg”-ATPase AA-ATPase

2.0 53 ± 17(3) 26(1) 12 ± 13(3) 19 ± 7(3)

20.0 75±0.6(3) 77(1) 51 ± 14(2) 69±6(4)

Number of experiments in parentheses. Values are expressed as

percent of inhibition of the control value; mean ± SD.

Table 2. Effect of Vanadate on Endocyt osis in Ghosts

Time of incubation

No. ofat 37 ‘C

Experi-Endocytosis Inducer ments 30 Mm 60 Mm

A

Mg-Al? 5 97±6 95±7

Trypsin 7 0 0

EDTA 3 0 0

B

Reversal of vanadate inhibition

Mg-ATP 2 0 0

Results are expressed as percentage of inhibition ± SD produced by 30

zmol/L of vanadate.

then prepared from RBCs that had been radiophosphory-

lated. Aliquots were taken for analyses at 0, 60, and 1S0

minutes of incubation at 37 #{176}C.No differences in the amount

and distribution of the membrane proteins were detectable

by densitometry; nor were there detectable differences in the

phosphorylation patterns of the membrane proteins among

four samples containing: buffer only; 30 �zmol/L of vanadate;

3 mmol/L of Mg-ATP, and 3 mmol/L of Mg-ATP plus 30

zmol/L ofvanadate (data not shown).

Effect of vanadate on drug-induced endocytosis. Addi-

tion of up to 100 �mol/L of vanadate and incubation for up

to 60 minutes at 37 #{176}Chad no effect on RBC shape or on the

binding of the indicator 57Co-vitamin B,2-transcobalamin

complex to the RBC outer surface (data not shown). Vana-

date was allowed to enter RBCs during a 60-minute preincu-

bation at 37 #{176}C.Endocytosis produced by all three drugs

(Table 3) was inhibited by vanadate. The magnitude of the

vanadate effect can be appreciated from a single experiment,

in which primaquine endocytosis trapped 1,150 cpm of

57Co/10’#{176}RBCs but preincubation with 30 �smol/L of vana-

date reduced that to 250 cpm/lO’#{176} RBCs. In the same

experiment, control values for chlorpromazine endocytosis

were 881 cpm/lO’#{176} RBCs, falling to 313 cpm/l0’#{176} RBCs on

addition of 30 �zmol/L of vanadate. Parallel values for

Table 3. Vanadate Inhibition of Drug-Induced Endocytosis in

RBCs

Drug

No. of

Experiments

Concentrationof Vanadate

(�tmol/L)

Inhibition (%)of Endocytosis

Mean ±5D

Primaquine

3 mmol/L 4

7.5

15.0

30.0

53±21

75 ± 19

95±11

Chlorpromazine

1 mmol/L 6

7.5

15.0

30.0

36±29

64 ± 27

90 ± 17

Vinblastine

0.6mmol/L 6

7.5

15.030.0

10±15

29 ± 26

77 ± 18

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

Fig 1 . Nomarski interference microscopy. original magnifica-

tion 900x; current magnification 432x. (A) Control RBCs incu-bated without primaquine or vanadate for 45 minutes at 37 ‘C. Theaddition of 100 �zmol/L of vanadate during a 30-minute preincuba-tion at 37 ‘C resulted in the same discocytic picture. (B) RBCswere incubated with 4 mmol/L of primaquine for 45 minutes at

37 ‘C. Note the extensive and uniform conversion to spherostoma-tocytes I.’ (C) RBCs were preincubated with 100 gzmol/L of vana-date for 30 minutes at 37 ‘C before 4 mmol/L of primaquine was

added; the reaction was then continued for an additional 45minutes at 37 ‘C. Note the persistence of the discocytic shape inessentially all RBCs.

VANADATE, RBC. AND GHOST SHAPE CHANGE 1011

Values are zmol ATP/mL packed RBCs.

vinbiastine were I ,78 , falling to 619. Phase microscopy

revealed that >50% of RBCs had one or more vacuoles in

control cationic amphipath-treated samples, whereas after

vanadate < 1 0% of the RBCs contained vacuoles except in

the vinblastine sample, in which -20% of the RBCs con-

tamed vacuoles. Primaquine endocytosis seemed most sus-

ceptible, with an I5� of -7.5 jsmol/L of vanadate.

Because it is known that reduction in ATP levels can

impair drug-induced endocytosis, particularly primaquine

endocytosis,6 six experiments were performed with six dif-

ferent donors to measure ATP levels in RBCs exposed to

endocytosis-inducing drugs without or with the addition of

vanadate. Table 4 shows the results of a single experiment

and indicates that no change in AlP levels occurred as a

consequence of incubation with amphipathic cations or

vanadate.

Effect of vanadate on RBC shape change. Because our

experimental design called for coded morphologic assess-

ment of RBC shapes during drug-induced endocytosis, it

became quickly apparent that vanadate not only blocked

endocytosis but also blocked the acquisition of the spherosto-

matocytic shape change that invariably precedes endocytosis

(Fig I). Figure 1 shows the results with primaquine; how-

ever, vinblastine and chlorpromazine stomatocytosis were

similarly blocked. Primaquine and vinblastine produce the

stomatocytic shape change over a period of minutes, whereas

the chlorpromazine shape change is virtually instantaneous.

Nevertheless, preincubation with vanadate blocked the

chlorpromazine-induced stomatocytosis (data not shown) as

effectively as it blocked primaquine and vinblastine stomato-

cytosis. Higher concentrations of primaquine (4 to 6

mmol/L) could partially overcome the shape inhibition of 30

jsmol/L ofvanadate (data not shown) but not of 100 �tmol/L

of vanadate.

The next question was whether vanadate could block other

forms of shape change. Acid pH produces stomatocytes I and

!I.�2 We therefore incubated RBCs as described in the

Materials and Methods section without or with vanadate in

isotonic citrate or phosphate buffer at pH 5.2. At intervals,

RBCs were removed, fixed with glutaraldehyde, and exam-

med under phase microscopy. In two such experiments,

vanadate had no effect on the acquisition of the stomatocytic

shape produced by acidic buffers (data not shown). The

shape change produced by acid pH was not readily reversed

by four 60-fold volume washes in 1% BSA dissolved in PBS

pH 7.4. Therefore, the role of vanadate on the reversal of the

acid pH stomatocytic shape change could not be assessed.

Echinocytosis was induced as described in the Materials

Table 4. RBC ATP levels

Value

Initial RBC ATP value 1.2

Following incubation with

Vanadate 30 zmol/L and:

Nodrugadded 1.3

Primaquine 3 mmol/L 1.3

Chlorpromazine 1 mmol/L 1.9

Vinblastine 0.6 mmol/L 1.6

A p C,

and Methods section by adding 6 �tg/mL of lysophosphati-

dylcholine, which produces echinocytes II and III. Addition

of 30 �imol/L ofvanadate had no effect on the appearance or

extent of this shape change. The reversibility of this shape

change was then studied by washing RBCs in 1% BSA in

PBS; neither preincubation with vanadate nor addition of

vanadate to the wash solution interfered with the prompt

reversal of the echinocytic shape change to the discocytic

shape.

Reversibility of vanadate effect on RBC stomatocytosis

and endocytosis. To determine if the effect of vanadate

was reversible, RBCs that had been pretreated with 30

�tmol/L of vanadate were extensively washed and then

incubated overnight under conditions calculated to enhance

vanadate effiux. The results of one of two such experiments

indicated (Table 5) that the effect ofvanadate on endocytosis

was reversible. The effects of vanadate on inhibition of

stomatocytosis were also completely reversible (data not

shown).

Effect of vanadate on drug entry into RBCs. It was

possible that micromolar amounts of vanadate could block

uptake of amphipathic drug into RBCs. Therefore, the effect

of vanadate on drug uptake was measured using identical

Table 5. Reversibility of Vanadate Inhibiti on of RBC Endocytosis

Fresh RBCs

Treated With

Drug Vanadate 30 zmol/L (%)

Vanadate-Treated

(30 �tmol/L) RBCs.

Washed and Incubated

Overnight 1%)

Primaquine

3 mmol/L 76

Chlorpromazine

1 mmol/L 31

5

0

Values are percentage of inhibition.

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

1012 SCHRIER ET AL

Table 6. Effect of Vanadate on Drug Uptake Into RBCs

Time ofIncubation Vanadate Concentration of Drug Concentration of Drug Recovery

Drug Concentration at 37 ‘C (mm) Addition in ABCs’ in 5upernatant� of Drug (%)

Vinblastine 0.5 mmol/L 45 0

30zmol/L

0.52

0.51

0.39

0.38

83

83

Chlorpromazine 1 .0 mmol/L 1 5 0 0.76 0.79 79

30 �zmol/L 0.7 1 0.84 81

�zmol/mL of packed RBCs.

ti.�mol/mL of supernatant.

concentrations and ratios ofdrug, RBCs, buffer, and plasma,

and the same time intervals we used to study endocytosis.

Tritiated radioisotopes ofvinblastine and chlorpromazine are

available and we have previously studied their uptake into

RBCs. The recoveries of radioisotopic drug were in the order

of 80%. Vanadate produced no difference in RBC uptake of

either chlorpromazine or vinblastine. One of the two experi-

ments performed is shown in Table 6.

DISCUSSION

The chemistry of vanadium2’ and the oxyvanadium com-

pounds’5’2’ has been the object ofconsiderable study because

of their potential physiological role. At nanomolar to micro-

molar concentrations and at physiologic pH, the predomi-

nant species is pentavalent metavanadate anion (VO,).’5’2’

The metavanadate may resemble the transition state of

phosphate, thus allowing it to insert into phosphate subsites

of some ATP binding sites. Perhaps it is by this mechanism

that metavanadate inhibits the RBC membrane Na�, K�-

ATPase,”’2 and Ca2�, Mg2�-ATPase,’3”4 as well as various

phosphohydrolases in kidney, muscle, nerve, and liver.’5

These studies generally have been carried out on disrupted

cell fractions. Vanadate can also inhibit the Na�-K� pump

when added to intact RBCs, but somewhat higher concentra-

tions are required.’2’29 One explanation is that vanadate

transverses the RBC membrane as the metavanadate anion;

once inside the RBCs, however, substantial amounts of

metavanadate are reduced to the tetravalent vanadyl cation

(VO2�),’521”9 a form that is less inhibitory to the Na�,

K�-ATPase and that may also bind to cellular constituents.29

In addition to its activity against ATPases, vanadate can also

inhibit phosphatases, stimulate adenyl cyclase, and inhibit

ATP-dependent proteases.’5”#{176} Of importance to our study is

the fact that vanadate at the micromolar concentrations used

(2 to 100 �mol/L) has no effect on RBC shape, deformabili-

ty, osmotic fragility, or metabolism,’6”7 whereas it can affect

all these modalities at considerably higher concentrations.

We initially confirmed reports that micromolar concentra-

tions of vanadate inhibited four human RBC membrane

ATPases: Mg2�-ATPase, Na�, K�-ATPase, Ca2�, Mg’�-

ATPase, and actin-activated ATPase (Table 1). The mem-

brane preparation used for these measurements is essentially

the same as the leaky hypotonic ghosts used to study endocy-

tosis.

In ghosts suspended in 50 mmol/L of TES, endocytosis

can be produced by 3 mmol/L of Mg-ATP (energized

endocytosis)” and by attack on the cytoskeleton produced by

trypsin or EDTA (nonenergized endocytosis).’9 Vanadate

blocked only Mg-ATP endocytosis (Table 2), thereby sug-

gesting that hydrolysis of ATP is the important factor in

energized endocytosis. Fairbanks and colleagues9 noted that

vanadate blocked the discocytic shape transformation pro-

duced by isotonically sealing Mg-ATP within ghosts and

suggested that vanadate inhibited Mg2�-ATPase, which

could be a shape-change-mediating enzyme. The evidence

linking Mg2�-ATPase to ghost shape change is based on a

parallelism between the concentration of vanadate that

inhibits Mg’�-ATPase and the concentration of vanadate

that inhibits Mg-ATP-induced ghost discocytosis.9 Using a

similar form of analysis, we suggested that in a somewhat

different preparation-resealed red ghosts-endocytosis

requires the Ca2�-induced activation of , Mg2 � -ATPase,

and that in the absence of Ca2� or in the presence of

inhibitors of Ca’�, Mg’�-ATPase red ghost endocytosis was

inhibited.” Hayashi and Penniston used a series of aikylating

agents and noted a strong correlation between inhibition of

ATPase and inhibition of endocytosis in ghosts. They pro-

posed that inhibition at a single site accounted for both

actions.’2 In trying to determine which of the ATPases might

be involved, they noted concurrent inhibition of ghost endo-

cytosis and a low-affinity Ca2�-stimulated Mg2�-ATPase.33

It is unlikely, however, that there is any free Ca2� in these

thoroughly washed white ghosts. Thus, the Mg2�-ATPase as

well as the actin-activated ATPase are candidates for the

mechanisms by which Mg-ATP could energize, and vana-

date block, ghost endocytosis. The prompt reversibility of the

vanadate effect is consistent with the reversibility of vana-

date inhibition of RBC membrane ATPases’5 and against the

idea that vanadate caused the oxidation of sulfhydryl

groups.2’ Attempts to identify a membrane protein dephos-

phorylated during Mg-ATP-induced endocytosis were

unsuccessful. The methods used, however, were suitable only

to detect dephosphorylation that might have occurred in f.�

spectrin, protein 2.1 , 3, or 4.1 , but not the RBC membrane

ATPases. In addition to ATPase action, the hydrolysis of

Mg-ATP in ghost endocytosis may be produced by protein

kinases, by enzymes involved in lipid phosphorylation’4 of

ghost membranes, or by as yet unknown mechanisms.’5

To study endocytosis in RBCs, a very different system

than leaky ghosts, we preincubated intact RBCs with vana-

date at concentrations reported to inhibit the Na�-K�

pump” and presumably the pump-related Na�, K�-ATPase;

we observed a dose-dependent inhibition of drug-induced

endocytosis (Table 3). It is highly unlikely that the RBC

membrane Na�, K�-ATPase was involved in drug-induced

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

VANADATE, RBC, AND GHOST SHAPE CHANGE 1013

endocytosis since neither ouabain addition’ nor substitution

ofcholine for Na� or K�24 have an effect on such endocytosis.

We have previously shown that primaquine-induced endocy-

tosis is absolutely dependent on active glycolysis and mainte-

nance of ATP levels, whereas vinblastine and chlorproma-

zinc endocytosis are reduced -60% but not obliterated by

RBC ATP depletion.6 Preincubation of RBCs with vanadate

did not reduce RBC ATP levels (Table 4). Because ATP was

not depleted and micromolar additions of vanadate inhibited

endocytosis, it is highly likely that it is not ATP in itself, but

ATP hydrolysis that is involved in primaquine endocytosis to

a great degree and in chlorpromazine and vinblastine endo-

cytosis to a lesser extent.

While we were monitoring endocytosis morphologically, it

quickly became apparent that the stomatocytosis (Fig 1) that

invariably precedes the endocytosis was blocked by vanadate.

We confirmed the fact that vanadate at the concentrations

used (2 to 100 zmol/L) did not cause echinocytosis.”

Therefore, its action in blocking the stomatocytic shape

change was not to produce a conflicting or neutralizing shape

alteration. We then addressed the question of whether

vanadate blocked all forms of stomatocytosis or only those

produced by cationic amphipaths. The stomatocytosis pro-

duced by pH 5 buffers was not inhibited by vanadate. The

next question was whether vanadate would inhibit all forms

of amphipath-induced shape change. Vanadate did not

inhibit the echinocytosis produced by incubation of intact

RBCs with lysophosphatydlcholine.

The use of vanadate indicates that all forms of stomatocy-

tosis do not proceed by identical mechanisms because acid

pH stomatocytosis is not blocked by vanadate but cationic

amphipath-induced stomatocytosis is. The initial description

of the bilayer couple hypothesis suggested that amphipath-

produced stomatocytosis and echinocytosis were, in a sense,

mirror images, with the charge of the amphipath passively

determining its localization in the lipid bilayer and thus the

site of membrane expansion.36 Our studies show, however,

that micromolar amounts of vanadate (metavanadate and

vanadyl)28 block the RBC stomatocytic shape change pro-

duced by cationic amphipaths but not lysophosphatydlcho-

line-induced echinocytosis. A possible explanation for this

difference comes from recent data indicating that spin-

labeled phospholipids may assume their transbiiayer assy-

metric distribution in resealed RBC ghosts by a mechanism

using Mg-ATP and blocked by 1 to 2 smol/L of vanadate.37

In intact RBCs, Mg-ATP is required to translocate amino-

phospholipids to the inner half of the bilayer.” Thus, vana-

date could block Mg-ATP-mediated translocation of

cationic amphipaths to the inner half of the bilayer. Against

this proposal are the data in Table 6 indicating that vanadate

does not inhibit the passage of such drugs through the

membrane, during which time they would have access to the

inner half of the bilayer. Mg-ATP-induced transiocation of

drug would only be involved if such drugs passed through the

membrane in sealed channels and if their subsequent access

from cytosol to the inner half of the bilayer was blocked by

the cytoskeleton.

These studies indicate that it is not the presence of ATP

but ATP hydrolysis, reversibly blocked by vanadate, that is

required for certain forms of stomatocytosis and endocytosis

in RBCs and endocytosis in ghosts. It is not proven that RBC

and ghost endocytosis proceed by identical mechanisms, but

the vanadate inhibition of both, at similar concentrations,

suggests that the two are related. The pathways by which the

hydrolysis of ATP leads to these morphologic changes must

be determined.

REFERENCES

1 . Deuticke B: Transformation and restoration of biconcaveshape of human erythrocytes induced by amphiphilic agents and

changes ofionic environment. Biochim Biophys Acta 163:494, 1968

2. Bessis M: Red cell shapes. An illustrated classification and its

rationale. Nouv Rev Fr Hematol 12:721, 1972

3. Ben-Bassat I, Bensch KG, Schrier SL: Drug-induced erythro-cyte membrane internalization. J Clin Invest 5 1 : 1833, 1972

4. George JN, O’Brien RL, Pollack 5, Crosby WH: Studies of in

vitro primaquine hemolysis: Substrate requirement for erythrocyte

membrane damage. J Clin Invest 45:1280, 1966

5. Feo C, Mohandas N: Clarification of role of AlP in red cellmorphology and function. Nature 265:166, 1977

6. Schrier SL, Junga I, Krueger J, Johnson M: Requirements of

drug-induced endocytosis by intact human erythrocytes. Blood Cells

4:339, 19787. Nakao M, Nakao I, Yamazoe 5: Adenosine triphosphate and

maintenance of shape of the human red cells. Nature I 87:946, 1960

8. Sheetz MP, Singer SJ: On the mechanism of AlP-inducedshape changes in human erythrocyte membranes. J Cell Biol 73:638,

1977

9. Fairbanks G, Patel VP, Dino JE, Carter DP: Vanadate inhibi-

tion of the MgATP-dependent shape change of human erythrocyte

ghosts. J Cell Biol 95:254a, 1982 (abstr)

10. Macara IG: Vanadium-an element in search of a role.

Trends Biochem Sci 5:92, 1980

I I . Norby JG: Ligand interactions with the substrate site of Na,

K-ATPase: nucleotides, vanadate, and phosphorylation. Curr TopicsMembranelransport 19:281, 1983

12. Cantley LC Jr. Resh MD, Guidotti G: Vanadate inhibits the

red cell (Na�, K�) ATPase from the cytoplasmic side. Nature

272:552, 1978

I 3. Rossi JPFC, Garrahan PJ, Rega AF: Vanadate inhibition of

active Ca2� transport across human red cell membranes. Biochim

Biophys Acta 648:145, 198114. Bond GH, Hudgins PM: Inhibition of red cell Ca2�-ATPase

by vanadate. Biochim Biophys Acta 600:78 1 , I 98015. Heinz A, Rubinson KA, Grantham JJ: The transport and

accumulation of oxyvanadium compounds in human erythrocytes in

vitro. J Lab Clin Med 100:593, 1982

16. Ninfali P, Accorsi A, Fazi A, Palma F, Fornaini G: Vanadate

affects glucose metabolism of human erythrocytes. Arch Biochem

Biophys 226:441, 1983

17. Vives-Corrons JL, Jou JM, Ester A, Ibars M, Carreras Jr

Bartrons R, Gliment F, Grisola 5: Vanadate increases oxygenaffinity and affects enzyme activities and membrane properties of

erythrocytes. Biochem Biophys Res Commun 103:1 11, 1981

I 8. Penniston JT: Endocytosis by erythrocyte ghosts; dependenceupon ATP hydrolysis. Arch Biochim Biophys 153:410, 1972

19. Hardy B, Schrier SL: The role of spectrin in erythrocyte

ghosts endocytosis. Biochem Biophys Res Commun 81 :1 1 53, 1978

20. Goodno CC: Inhibition of myosin AlPase by vanadate ion.Proc NatI Acad Sci USA 76:2620, 1979

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

1014 SCHRIER ET AL

AlP-dependent degradation of proteins in reticulocytes withoutaffecting ubiquitin conjugation. J Biol Chem 259:2803, 1984

21. Macara L, Kustin K, Cantley Jr LC: Glutathione reduces

cytoplasmic vanadate: Mechanism and physiological implications.

Biochim Biophys Acta 629:95, I 980

22. Schrier SL, Hardy B, Junga I, Ma L: Actin-activated

ATPase in human red cell membranes. Blood 58:953, 1981

23. Schrier SL, Giberman E, Danon D, Katchalski E: Studies on

ATPase in sheared micro vesicles of human erythrocyte membranes.Biochim Biophys Acta 196:263, 1970

24. Schrier SL, Junga I, Seeger M: The mechanism of drug-induced erythrocyte vacuole formation. J Lab Clin Med 83:215,

I 97425. Beutler E: Red Cell Metabolism: A Manual of Biochemical

Methods (ed 2). Orlando, Fla, Grune & Stratton, 1975, p 10

26. Schrier SL, Junga I: Entry and distribution of chlorproma-zinc and vinblastine into human erythrocytes during endocytosis.

Proc Soc Exp Biol Med 168: 1 59, 1981

27. Jarrett HW, Penniston JT: A new assay for endocytosis in

erythrocyte ghosts based on loss of acetylcholinesterase activity.Biochim Biophys Acta 448:314, 1976

28. Rubinson KA: Concerning the form of biochemically active

vanadium. Proc R Soc Lond 212:65, 198129. Cantley LC Jr. Aisen P: The fate ofcytoplasmic vanadium. J

Biol Chem 254:1781, 197930. Tanaka K, Waxman L, Goldberg AL: Vanadate inhibits the

31. Schrier SL, Bensch KG, Johnson M, Junga I: Energized

endocytosis in human erythrocyte ghosts. J Clin Invest 56:8, 1975

32. Hayashi H, Penniston iT: Energy-dependent endocytosis in

erythrocyte ghosts. Arch Biochem Biophys I 59:563, 1973

33. Jarrett HW, Reid TB, Penniston JT: Concurrent inhibition of

the low-affinity Ca2�-stimulated ATPase and MgATP-dependentendocytosis in erythrocyte ghosts by naphthylmaleimide and carbon-

cyicyanide-m-chlorophenylhydrazone. Arch Biochem Biophys183:498, 1977

34. Fairbanks G, Patel VP, Dino JE: Biochemistry of AlP-dependent red cell membrane shape change. Scand J Clin Lab Invest41:139, 1981

35. Jinbu Y, Nakao M, Otsuka M, Sato 5: Two steps in AlP-

dependent shape change of human erythrocyte ghosts. Biochem

Biophys Res Commun 1 12:384, 1983

36. Sheetz M, Singer SJ: Biological membranes as bilayer cou-pies. A molecular mechanism of drug-erythrocyte interactions. Proc

NatI Acad Sci USA 71:4457, 1974

37. Seigneuret M, Devaux PF: AlP-dependent asymmetric dis-tribution of spin-labeled phospholipids in the erythrocyte membrane:

Relation to shape change. Proc NatI Acad Sci USA 8 1 :3751, 1984

38. Daleke DL, Huestis WH: Incorporation and translocation of

aminophospholipids in human erythrocytes. Biochemistry 24:5406,

I985

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom

1986 68: 1008-1014  

SL Schrier, I Junga and L Ma human red blood cells and ghostsStudies on the effect of vanadate on endocytosis and shape changes in 

http://www.bloodjournal.org/content/68/5/1008.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of

For personal use only.on December 14, 2018. by guest www.bloodjournal.orgFrom