Receptors for Erythropoietin in Mouse and Human Erythroid Cells and Placenta

Blood, Vol 74, No 1 (July). 1989: pp 103-109 103

By Stephen T. Sawyer. Sanford B. Krantz, and Ken-ichi Sawada

High and lower affinity receptors for erythropoietin (EP)were initially identified on a very pure population of EP-responsive erythroblasts obtained from the spleens of miceinfected with anemia strain of Friend virus (FVA). The

structure of the receptor for EP in these cells was deter-

mined to be proteins of 100 and 85 Kd by cross-linking‘25l-EP. In this investigation. studies on the receptors for EPwere extended to other mouse erythroid cells and humanerythroid cells as well as to the placentas of mice and rats.Only lower affinity receptors for EP were detected onerythroblasts purified from the spleens of mice infected

with the polycythemia strain of Friend virus and a murineerythroleukemia cell line. both of which are not responsive

to EP in culture. Internalization of ‘�l-EP was observed inboth groups of cells. The structure of the receptor deter-mined by cross-linking ‘25l-EP was two equally labeledproteins of 100 Kd and 85 Kd molecular mass in all these

S PECIFIC BINDING of erythropoietin (EP) was first

observed in erythroblasts purified from spleens of mice

infected with the anemia strain of Friend virus (FVA).’3

These immature FVA-infected erythroid cells (FVA cells)

respond to physiological levels of EP in culture by pro-

gressing to near erythroid maturation.� ‘251-EP binds to

higher and lower affinity receptors on the surface of these

FVA cells and is subsequently internalized and degraded in

the lysosomal compartment. ‘In our initial investigation,

‘251EP binding was also observed in murine erythroleukemia

(MEL) cells, clone 745, but only to lower affinity receptors.’

In this study, we characterize the internalization of ‘25I-EP in

these MEL cells, which are not responsive to EP in culture,

and the binding and internalization of ‘25l-EP in erythroid

cells purified from the spleens of mice infected with the

polycythemia strain of Friend virus (FVP cells), which

spontaneously differentiate in vitro in the absence of exoge-

nously added EP. The existence of receptors for EP in

membranes prepared from placentas from mice and rats is

also identified.

When ‘251-EP bound to FVA cell membranes was cross-

linked by disuccinimydl suberate (DSS) to the receptor, two

labeled bands corresponding to proteins of molecular weights

of 100 and 85 Kd were observed on sodium dodecyl sulfate

polyacrylamide gel electrophoresis (SDS-PAGE).3 No evi-

dence of disulfide bridges between these two proteins was

found. More recently other investigators have presented

evidence that suggests the existence of additional lower

molecular weight proteins in the receptor7�’#{176} and have sug-

gested that the multiply labeled proteins may be subunits of a

very large complex that is bridged by disulfide bonds.8”#{176}We

have examined this possibility by cross-linking ‘25I-EP to the

receptor in MEL cells, FVP cells, human erythroid colony-

forming cells (CFU-E), and placentas from mice and rats.

The data presented here show a remarkable similarity of the

receptors for EP in human, mouse, and rat tissues and in EP

responsive and nonresponsive erythroid cells. No evidence of

additional subunits of the receptor or a larger complex of

subunits was detected in any of these sources of receptor.

mouse erythroid cells. The structure of the receptor wasfound to be very similar in human erythroid colony formingcells cultured from normal blood. These cells respond to EPwith erythroid maturation and were previously shown tohave high and lower affinity receptors. Placentas from miceand rats were found to have only lower affinity receptorsfor EP. and when placental membranes were cross-linkedto ‘25l-EP, the same 100 Kd and 85 Kd bands were found asseen in mouse and human erythroid cells. The structure ofthe receptor was similar in cells that have high affinityreceptors (FVA-infected and human erythroid colony-forming cells) and nonresponsive erythroid cells and pla-

centa that have lower affinity receptors. but only the cellswith the high affinity receptors respond to the addition ofEP with erythroid maturation.S 1989 by Grune Stratton, Inc.

MATERIALS AND METHODS

Human recombinant EP was purchased from AmGen Biologicals

(Thousand Oaks, CA). Na ‘25I was obtained from Amersham. DSS

and lODO-GEN (l,3,4,6-tetra-chloro-3a, 6a-diphenylglycouril)

were obtained from Pierce (Rockford, IL). Friend virus, pseudotype

SFFVA/FRE cl-3/MuLV (201) originally obtained from W.D.

Hankins (National Institutes of Health [NIH]) and FVP obtainedfrom R. Holdenreid (NIH) were maintained by the passage of

infectious plasma in BALB/c mice. MEL cells, clone 745, were

obtained from W. LeStourgeon, Vanderbilt University.

Cells and plasma membrane preparation. Immature erythroidcells were purified from the spleens of CD2F, mice infected with

FVA or FVP by velocity sedimentation at unit gravity through a

continuous gradient of bovine serum albumin as described previous-

ly.4’5 To prepare plasma membranes from FVA- or FVP-infected

erythroid cells, the total spleen was disrupted to a single cell

suspension and the erythrocytes were lysed by exposure to ISO

mmol/L NH4CI/15 mmol/L Tris HCI, pH 7.65, at 37#{176}Cfor 30

seconds. The NH4CI/Tris solution was diluted fourfold with Iscove’s

modified Dulbecco’s medium and the cells were pelleted by centrifu-

gation at 500 g for I 5 minutes. The pellet was resuspended in

NH4CI/Tris and the procedure was repeated. The cells were then

washed in 105 mmol/L NaCI and 10 mmol/L Tricine, pH 7.4, three

times and resuspended in 10 mmol/L KCI and 10 mmol/L Tricine,

From the Division of Hematology, Department of MedicineVanderbilt University School of Medicine and Veterans Adminis-

tration Medical Center, Nashville,

Submitted August 9, 1988; accepted March 3, 1989.

Supported by grants from the National Institutes of Health,

DK-39781. AM-15555. T32 DK-07186 and VA Medical Research

Funds.Address reprint requests to Stephen T. Sawyer. PhD, Division of

Hematology, Room C-3101. Medical Center North, Vanderbilt

University School ofMedicine. Nashville, TN 37232.

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. section 1734 solely to

indicate this fact.

© I 989 by Grune & Stratton, Inc.

0006-4971/89/7401-0034$3.OO/O

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

104 SAWYER, KRANTZ. AND SAWADA

pH 7.4, containing a mixture of proteinase inhibitors (2 mmol/L

EGTA, 5 mmol/L EDTA, 1 �g leupeptin/mL, 5 mmol/L

ide, 1 mmol/L iodoacetamide, 10 �g/mL tosylamide-2-phenylethl-

chloromethyl ketone, 10 �g/mL p-tosyl-1-arginine methylester, 1

�g/mL n-d-tosyl-l-Tysine chloromethylketone, 2 �zg/mL aprotinin,

I zg/mL pepstatin, and 0.1 mmol/L penylmethyl-sulfonylfluorine).

Plasma membrane fractions of MEL cells, FVP cells, and FVAcells were prepared in identical fashion as described previously3

except that the above described mixture of proteinase inhibitors was

included at every step. Briefly, the cells were swollen in hypotonic

medium, and disrupted with ten strokes of a teflon, motor-driven

homogenizer. The cell debris was discarded after low speed centrifu-

gation, and a crude membrane pellet was obtained after one hour

centrifugation at I 50,000 g. This pellet was resuspended in a solution

containing a final concentration of 40% sucrose, which became the

bottom of a discontinuous sucrose gradient composed of solutions of

35%, 31%, 25%, and 8.5% sucrose. The light membrane fractions at

the 35%/3 1% interface and 3 l%/25% interface were collected as the

plasma membrane fraction.

Placenta membranes were prepared in a similar fashion. Placen-

tas from CD2F, mice were taken at days 14 and 18 of gestation, and

placentas from Sprague Dawley rats were taken at 1 8 to 20 days of

gestation. Placentas were washed in phosphate buffered saline

containing the above mentioned cocktail of protease inhibitors and

finely minced. The minced placenta was then homogenized, centri-

fuged, and crude membranes were fractionated on discontinuous

sucrose gradient exactly as described above for the erythroid cells.

Human CFU-E were obtained as described previously.” Briefly,

erythroid burst forming units were partially purified from normal

human blood. These cells were then cultured in the presence of EP to

the stage where the cells were not producing a significant level of

hemoglobin and were capable of forming colonies of eight to 49 cells

when further cultured in the presence of EP. At this stage the cells

were removed from culture and purified additionally by removing

adherent cells and by Ficoll-Hypaque density centrifugation. Purity

of the cells was >50% and receptors for EP have been found on the

cell surface at about 1 ,000 receptors per cell.’2

lodination of EP. EP was iodinated using IODO-GEN.’ Two

micrograms of lODO-GEN were coated on the walls of a conical

reaction vial. EP (50 units, -.4 �zg protein) and 20 �iCi 251 were

incubated in the reaction vial for five minutes at room temperature

in a final volume of 50 �L of 0.5 mol/L phosphate buffer, pH 7.0,

containing 0.02% Tween 20. After the incubation, the contents of the

reaction vial were transferred to a tube containing 10 mg KI in

phosphate buffered saline, 0.1% bovine serum albumin, and 0.02%

Tween 20; ‘251-EP was separated from the free 1251 by chroma-

tography over a Biogel P6 column. This procedure provided EP with

0.3 to I .0 molecule of 251 per molecule (25 to 75 �Ci/�zg) and with

full biological activity when assayed in these FVA-infected erythroid

cells.4

Binding ‘251-EP. ‘25I-EP was incubated with from 10 to 40 �gprotein of the plasma membrane fraction from FVA cells, FVP cells,

MEL cells, and placentas from mice and rats at 37#{176}Cfor the time

indicated in 100 mmol/L phosphate buffer, pH 7.4, containing I

mmol/L EGTA and 0.1% bovine serum albumin. The binding

mixture was then applied to 0.2 zm Millipore filters (EHWP) and

washed with 10 mL of phosphate buffered saline containing 0.1%

bovine serum albumin. The filters were then counted in a gamma

counter. Nonspecific binding was determined in the presence of 100

to 200 units of EP/mL and was subtracted.

FVA cells, FVP ceils, and MEL cells were washed and resus-

pended in binding medium (Iscove’s modified Dulbecco’s medium)

supplemented to contain 20 mmol/L HEPES, pH 7.4, and 2%

bovine serum albumin. The cells were allowed to stand for one hour

at 37#{176}Cin a 5% CO2 atmosphere before the initiation of binding

studies. For binding at 37#{176}Clabeled EP was added to the cells in thebinding medium at varying concentrations for different times in the

incubator. The concentration of cells was iO� cells/mL or less.

Binding was terminated by sedimenting the cells through dibutyl

phthalate oil (0.5 mL) for one minute in a minifuge (8,000 g). The

tube was frozen at -80#{176}C, and the tip containing the cell pellet was

cut off. Radioactivity in the tip was determined by counting the tip in

a gamma counter. Nonspecific binding of the labeled EP was

determined by adding a 20 to 100-fold excess of unlabeled EP in the

binding assay (100 to 200 units EP/mL).

For binding at 0#{176}C,the protocol was essentially the same as at

37#{176}Cexcept that the cells were cooled for one hour in an ice bath at

4#{176}C,and the incubation was carried out in sealed 1 .5 mL minifuge

tubes in the ice bath.

For determination of the binding affinities by the method of

Scatchard,’3 the cells were incubated for 20 hours at 4#{176}C.This time

is sufficient for the binding to plateau, the cell viability as measured

by trypan blue exclusion to remain >90% viable, and bound EP to be

substantially released. However the binding of ‘251-EP is not totally

reversible at 20 hours (60% of ‘25I-EP bound was released). ‘251-EP

bound to FVA cells was not totally reversible during a reasonable

period at 4#{176}C.Extrapolation of the second order release curve

showed 90% of bound ‘25I-EP was released at 72 hours. However,binding at 4#{176}Cfor this length of time or greater is not feasible since

the cells are not stable and they disintegrate. Nevertheless, binding

at 48 hours at 4#{176}Cdid not affect the binding constant for FVA cells.

Therefore these studies were carried out at 20 hours for conve-

nience.

Determination of surface-bound and internalized EP. An acidwash, which has been used previously to remove surface-bound

ligands,’’4 was used to remove surface-bound labeled EP from the

cell surface. At the indicated times after labeled EP was added to the

cells, the cells were cooled to 0#{176}Cby the addition of a ninefold excess

of coldbinding medium and transferred to an ice bath. After three

washes to remove unbound ligand, the cells were pelleted by

centrifugation in a minifuge for 30 seconds and resuspended in cold

0.5 mmol/L NaCI and to 0.25 mmol/L acetic acid, pH 2.5. Afterthree minutes, the cells were centrifuged through dibutyl phthalate

oil. The tube was frozen, and the tip containing the pellet was cut off

and counted. The aqueous phase also was counted. Parallel experi-

ments were performed with a large excess of unlabeled EP in the

binding assay, and the cells were treated with the high salt, pH 2.5,

wash to determine the distribution of the nonspecifically bound EP

between the acid labile and stable radioactivity. After nonspecific

radioactivity was subtracted, the labeled EP in the acid wash was

considered surface bound, while the labeled EP resistant to the acid

wash was considered internalized.

Cross-linking of251-EP. ‘251-EP was cross-linked to the surface

of human erythroid colony-forming cells with DSS in a similar

fashion as described previously for FVA cells.3 25I-EP was added to10’ cells for 30 minutes at 37#{176}C.DSS was added at 500 zm for 15

minutes at 0#{176}C.The cells were washed three times with 150 mmol/L

Tris-HC1, pH 7.4, to quench the cross-linking reaction and remove

unbound EP. The cells were then extracted for one minute at 0#{176}C

with a solution containing 1.0% Triton X-lOO, 20 mmol/L HEPES,

pH 7.4, containing the mixture of 1 1 protease inhibitors described

above. The disrupted cell preparation was then centrifuged at 1,000

g for one minute to remove nuclei and cell debris; the supernatant

and pellet were both analyzed by SDS-PAGE.

�25��p was bound to receptors on the plasma membrane fraction

and was cross-linked in a similar fashion as described previously.3

Membranes were incubated with 7.5 units of ‘25I-EP per milliliter in

100 mmol/L phosphate buffer, pH 7.4, containing I mmol/L EGTA

and 0.1% bovine serum albumin for 15 minutes at 37#{176}C.The mixture

was transferred to an ice bath, and DSS was added to a final

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

TIME, minutes

‘251-EP Bound, molecules/cell

Fig 1 . Binding of �l-EP to FVP cells and FVA cells. Immatureerythroid cells were purified from mice infected with the polycy-themia and FVA. These cells were then incubated with theindicated concentration of ‘�l-EP for 20 hours at O’C. Nonspecific

binding was determined in the presence of a 20-fold excess ofunlabeled EP. Nonspecific binding was <20% and was subtracted.Data are the mean of triplicates ± SD. At the maximum binding for

FVA cells approximately 5.000 cpm of ‘�l-EP were specificallybound per 1 0 cells. The binding data were plotted by the methodof Scatchard.’3 A, FVA cells, �, FVP cells.

RECEPTORS FOR ERYTHROPOIETIN 105

concentration of 200 �imol/L for 15 minutes. Tris-HCI, pH 8.0, was

added to a final concentration of 1 50 mmol/L, and the membranes

were pelleted by centrifugation at 13,000 g for 30 minutes. The

pellet was resuspended in 10 mmol/L Tris-HCI, pH 7.4, and 100

mmol/L NaC1 and washed twice. The pellet was suspended in

sample buffer, boiled for three minutes and sonicated briefly. The

material was analyzed by SDS-PAGE as described by Laemmli.’5

RESULTS

Binding and internalization of ‘251-EP. FVP cells were

purified from the spleens of mice infected with the polycy- .�

themia strain of the virus. Surface receptors were quanti- �

tated by measuring the binding of ‘25I-EP at 0#{176}Cfor 20 �

hours. Figure 1 shows the amount of bound ‘25I-EP when

increasing concentrations of ‘25I-EP were added. In Fig 1 the

data were plotted by the method of Scatchard.’3 The inter- �

cept reveals a total of 650 surface receptors. The slope o

reveals a single class of receptors having a disassociation �

constant of 700 pmol/L. Binding to FVA cells is also shown

in Fig 1 to illustrate the two classes of ‘251-EP binding sites in

these cells compared with the single class of receptors found

in FVP cells. ‘25I-EP was also bound to FVP cells at 37#{176}C

(Fig 2B) and the internalization of ‘25I-EP by FVP, FVA,

Fig 2. Binding and internalization of ‘251-EP by FVA cells. FVPcells. and MEL cells. 1�l-EP was added to the cells at t = 0 andincubated for the indicated time at 37’C. The surface bound

hormone (#{149})and internalized hormone (0) were determined byextracting surface 12�l-EP from the cell with a pH 2.5. high saltwash as described in the Materials and Methods section.

and MEL cells were compared (Fig 2). FVP cells bind 900

molecules of ‘251-EP after 30 minutes at 37#{176}Cand the

� binding slowly falls to a plateau level of 600 total molecules

L� by 90 minutes (sum ofsurface and internal radioactivity). As:� previously reported,’ FVA cells bind 1,500 molecules at 30� minutes after which the level falls to I ,000 molecules of EP.

c� MEL cells bind 850 molecules of EP at 37#{176}C,which iscompatible with our earlier report of 760 receptors on the cell

surface. A pH 2.5, 0.5 mol/L NaC1 wash was used to strip

away surface bound ‘251-EP to discriminate surface bound

hormone from internalized hormone and degradation prod-

ucts. FVP cells and MEL cells internalized ‘25I-EP in a

similar fashion as FVA cells. As previously reported for FVA

cells, intracellular radioactivity predominated after 30 mm-

utes of incubation with ‘251-EP at 37#{176}C.Additional experi-

ments were carried out to show that the internalized ‘251-EP

was degraded to ‘25I-iodotyrosine (data not shown).

Plasma membrane fractions from FVA cells, FVP cells,

and MEL cells were tested for ‘251-EP binding. Smooth

membrane from mouse and rat placenta were also prepared,

and ‘25I-EP binding was quantitated (Fig 3 and Table 1).

Binding was determined for a range of increasing concentra-

tions of ‘25I-EP and plotted by the method of Scatchard, the

number of receptors was determined from the intercept while

the dissociation constant was determined from the slope. In

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

0 10 20 30 40 50 60l25� � EP BOUND, molecules/pg protein

- l4OkDa- 125 kDa

abcd

± the variance in these determinations.

106 SAWYER, KRANTZ, AND SAWADA

C

a,0

a’a

.%..‘I,0)

‘3

0�

0

E

0z:20

0.Ui‘.4

In

‘I)

0

UiUia:L�.

0z

0

Fig 3. Binding of ‘251-EP to membranes from FVP cells. MELcells. and placentas from mice and rats. Membranes were pre-pared and binding of 125l-EP were carried out as described in theMaterials and Methods section. Binding was performed in tripli-cate and the mean ± SD is shown. Nonspecific binding was carriedout in an excess of unlabeled EP and subtracted. In thesemembranes. nonspecific binding was similar; at the highest con-centration of lThl�EP, this nonspecific binding was 1 5% of total inFVP membranes and increased up to 65% in rat placenta. In the

lower panel. the data were plotted by the method of Scatchard.

FVP. (#{149});MEL. (0); mouse placenta. (D); rat placenta (s).

contrast to FVA cell membranes, which have higher and

lower affinity receptors, the other membranes showed a

single class of receptors with a (Kd) dissociation constant

from 600 pmol/L to 1.0 nmol/L (Table 1). FVA and FVP

cell membranes contained more receptors (72 and 54 recep-

tors/pg protein) than membranes from MEL cells (27

receptors/pg protein), and all the erythroid cell membranes

Table 1 . ‘�I-EP Binding to Membranes From Erythroid Cells and

Placenta

Membrane SourceReceptors/pg

ProteinBinding Affinity

(Kd)

FVA cells 72 80 pmol/L, 600 pmol/L3

FVP cells 54 800 ± 200 pmol/L’

MEL cells 27 900 ± 100 pmol/L

Mouse placenta 1 9 1 .000 ± 400 pmol/L

Rat placenta 9 900 ± 200 pmol/L

Binding affinities hown are the mea n of two to four determinations

contain more receptors than placenta membranes (mouse I 9

and rat 9 receptors/pg protein) when normalized for protein

content. These data are from Fig 3. Other preparations of

membranes had similar numbers of receptors. Inclusion of

additional inhibitors of proteinase activity increased the

receptors/protein ratio by 70% in FVA cells over the previ-

ously reported value.3

Cross-linking of ‘251-EP to receptors. ‘25l-EP bound to

membranes prepared from FVA cells, FVP cells, MEL cells,

and placentas from mice and rats was cross-linked to the

receptor for EP using DSS. As shown by the autoradiograph

in Figs 4 and 5, the ‘251-EP became covalently attached to

two discretely migrating proteins such that the complex

migrated at 140 Kd and I 25 Kd. Subtraction of the molecu-

lar mass of EP lead to the apparent molecular mass of 100

Kd and 85 Kd. These bands were identical in each sample

(FVA cells, FVP cells, MEL cells, and placenta from mouse

and rat). In Fig 4 the SDS-PAGE is run in the absence of

reducing agent. The results are identical to the SDS-PAGE

of the same samples ‘251-EP cross-linked membranes run in

the presence of reducing agent (shown in Fig 5). This

indicates the absence of disulfide bridges between the two

labeled proteins of 100 Kd and 85 Kd in all these erythroid

cells and placentas.

Intact cells were used to identify the receptor for EP in

human erythroid cells. After binding ‘25I-EP to these cells at

37#{176}C,DSS was added to the binding mixture. The cells were

disrupted and both the extract (Fig 6C) and cell debris

fraction (6B) were analyzed by SDS-PAGE. ‘25I-EPO cross-

FVA FVP MEL PLACENTAmouse rat

efghi j

Mr

Fig 4. Cross-linking of 1251-EP to membranes from FVA cells.FVP cells, MEL cells. and placentas from mice and rats. Binding andcross-linking was performed as described in Materials and Meth-ods. Nonspecific binding was done in the presence of 200 U/mLunlabeled EP. The cross-linked membranes were dissolved insample buffer without reducing agents and radioactivity wascounted in a gamma counter. Equal amounts of radioactivity wereloaded (an equal volume of extract of the corresponding nonspe-cific sample was loaded) on a 5% acrylamide gel. The dried gel wassubject to autoradiography. FVA cells. lanes a. b; FVP cells. c. d;MEL cells. e. f; mouse placenta. g. h; rat placenta. i. j. Nonspecificcross-linking is shown in lanes b, d. f. h. and j.

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

RECEPTORS FOR ERVTHROPOIETIN 107

Mr

- 140 kDa- 125 kDa

Mr

k Do

200

-140-130-I--. 25-‘- I 10

92

ABCDE

Fig 5. Cross-linking of ‘25l-EP to membranes from FVA cells.FVP cells. MEL cells. and placentas from mice and rats. Onemilligram of membrane protein was incubated with 5 units of‘�‘l-EP for 1 5 minutes and cross-linked with 0.2 mmol/L DSS. Thesamples were analyzed by SDS-PAGE in the presence of fi-

mercaptoethanol and autoradiography. Lane A. FVA cells; B. FVPcells; C. MEL cells; D. mouse placenta; E. rat placenta.

linked proteins were found to migrate in the gel at a position

corresponding to a molecular mass of 140 Kd and approxi-

mately 130 Kd. While the 140 Kd band is identical to that

found in mouse tissues, the lesser molecular weight band

appears just slightly larger than the corresponding band in

mouse FVA cells (Fig 6D). A minor band that migrates at

1 10 Kd was also seen in the cell extract but not in the cell

debris fraction. The presence of this band was variable from

one experiment to another and may be a proteolytic fragment

of the larger cross-linked bands. The gel pattern was identi-

cal in the presence and absence of reducing agents.

DISCUSSION

Following the appearance of our report of the structure of

the receptor in FVA cells as two proteins of 100 Kd and 85

Kd molecular mass,3 Todakaro et a17 reported that the

receptor in an EP responsive cell line was composed of three

subunits of I I 9, 94, and 63 Kd, which were not connected

through disulfide bridges. In addition, three erythroid cell

lines unresponsive to EP (including clone 745 MEL cells)

were found to have only the 63 Kd band cross-linked to

‘25IEP However, Sasaki et aO#{176}reported identical 100 Kd

and 85 Kd subunits of the receptor for EP in EP-responsive

TSA8 erythroid cells, FVA cells, mouse fetal liver cells, and

spleens from anemic mice. Tojo et al reported the EP

receptor as bands of I 10 and 95 Kd in fetal mouse liver,’6

while Mayeux et al reported two proteins of molecular mass

of 94 and 78 Kd in fetal rat liver8; and Pekonen et al

described two proteins of 85 Kd and 41 Kd in human fetal

liver.9

A B C D E F

68

Fig 6. Cross-linking of ‘�l-EP to human erythroid colony-

forming cells and FVA cells. 1�l-EP was bound to the cells andcross-linked as described in Materials and Methods. Nonspecificbinding was determined in the presence of 100 units of unlabeled

EP. lanes A and F. The cells were extracted with detergent and

centrifuged at 1 .000 g. Cell extract. lanes A. C. D, and F; celldebris. B lanes and E. Human erythroid colony-forming cells. lanesA. B. and C; FVA cells. lanes D. E. and F.

In view of the possibility that different subunits of the

receptor might exist in different animal species or different

cell lines, we compared the structure of receptors for EP in

different erythroid cells. In contrast to the finding of oth-

ers,7”#{176}we found that the structure of the receptor for EP in

the MEL cells consisted of the same 100 Kd and 85 Kd

proteins found in FVA cells and all other tissues tested in this

study. In addition, the rat and human receptor were similar

to the mouse receptor, in contrast to reports of lower molecu-

lar weight receptors.8’9 We suspect that the lower molecular

weight proteins cross-linked to ‘25I-EP in these earlier reports

are proteolytic fragments of the larger molecular weight

proteins. In early experiments we observed a band of 70 Kd

when ‘25I-EP was cross-linked to intact MEL cells and the

cross-linked complex was extracted with Triton X-lOO. In

other studies to be published elsewhere, the receptor for EP

in FVA cells was observed to be degraded to a series of lower

molecular weight protein after cross-linking to ‘251-EP to

intact cells. This could be prevented by the addition of

proteinase inhibitors. While the cross-linked receptor is very

insoluble in nonionic detergents, the degradation fragments

are more easily extracted. Earlier reports on the EP receptor

in MEL and fetal rat liver cells used Triton X-I00 extraction

of cross-linked receptor from intact cells. Likewise, an addi-

tional minor band was seen in the soluble extract of cross-

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

108 SAWYER, KRANTZ. AND SAWADA

linked receptor from human CFU-E but not in the insoluble

fraction (Fig 6). Preparation of plasma membranes from

cells eliminates the need to extract the receptor from the

intact cell and lowers the ratio of receptor to proteinase

activity during the cross-linking reaction.

In contrast to our work and the work of Todokoro et al.,

Sasaki et al showed that the cross-linked bands mostly did

not enter the gel or a minor band was observed at 210 Kd in

the absence of reducing agent during SDS-PAGE.’#{176} Mayeux

et al reported the existence of 94 and 78 Kd proteins

cross-linked to ‘251-EP in fetal rat liver cells and also

suggested that these proteins were subunits of a large molec-

ular weight complex bridged by disulfide bonds.8 In this

study no evidence of disulfide bridging between the 100 and

85 Kd proteins of the EP receptor was detected in any cell or

tissue examined. The observations of large molecular weight

proteins cross-linked to ‘251-EP on SDS-PAGE in the

absence of reducing agents may be due to the incomplete

solubilization of the cross-linked membranes. Mild treat-

ment of ‘251-EP cross-linked to FVA cell membranes and

SDS-PAGE in the absence of reducing agents leads to a

weakly labeled band at 230 Kd and most of the radioactivity

at the top of the running gel and stacking gel. This radioac-

tive material was cut out of an unfixed gel and rerun on

SDS-PAGE in the presence of/3-mercaptoethanol. Only 1%

of this radioactivity was found in the 100 and 85 Kd

cross-linked bands while 99% was noncross-linked ‘251-EP

(data not shown). Mild sonication of the cross-linked mem-

brane in sample buffer for SDS-PAGE containing no reduc-

ing agent converted the higher molecular weight cross-linked

material into the clear 100 and 85 Kd proteins. This strongly

suggested the absence of disulfide bridges between the 100

and 85 Kd proteins.

This study reports the structure of the normal human

erythroid receptor for EP. The human receptor is virtually

identical to the mouse and rat receptor, however a slight

increase in the molecular mass of the lesser protein of the

human receptor compared with mouse and rat may occur.

Other experiments show the very slight difference in migra-

tion of the two proteins of the human receptor compared with

that in FVA cells. We suspect that the lower band of the

receptor for EP may be due to the cleavage of the 100 Kd

band by proteinase. If this is the case, a subtle difference in

the amino acid sequence of the human 100 Kd receptor

protein could lead to a different site of cleavage and therefore

a fragment of slightly larger molecular mass.

It is of interest that the receptor for EP is remarkably

similar in mouse, rat, and human tissue. Furthermore, the

similarity of receptor in erythroid cells and placenta is also of

interest in that the function of the receptor in the placenta is

not proven. This is the first report of the existence of

receptors for EP in placenta. The finding in this laboratory

that ‘25I-EP crossed from maternal circulation to the fetus in

pregnant mice’7 led us to investigate the existence of recep-

tors for EP in the placenta. We think that the placental

receptors for EP may be involved in the transplacental

transfer of ‘251-EP but additional experiments are necessary

to prove the physiological relevance of the transfer of EP to

the fetus and the receptor mediated nature of the transfer.

The nearly identical structure of the receptors in the

normal human erythroid cells and normal placentas com-

pared with the receptors in mouse erythroid cells that

proliferate in response to infection with the Friend virus

indicates that the Friend virus does not stimulate erythroid

development through a gross modification of the structure of

the receptor for EP. The two strains of Friend virus, FVA,

and FVP, lead to different responses when either mice or

bone marrow cells are infected. FVA infection generates

erythroid cells comparable with CFU-E in that the cells are

responsive to EP and respond by undergoing erythroid matu-

ration and growth. In contrast, FVP infection results in a

proliferation of erythroid cells that are apparently undergo-

ing erythroid maturation independently of EP. In culture

these cells do not respond to EP.4”8 It is possible that

autocrine production of EP by FVP cells might be responsi-

ble for this apparent independence of regulation by EP, but

preliminary experiments in our laboratory have failed to

detect production of mRNA coding for EP in FVP cells.

The presence of receptors for EP on MEL and FVP cells

on which EP has no apparent effect suggests that the normal

mechanism by which EP triggers these cells has been altered.

In this regard, we studied internalization of ‘25I-EP by MEL

and FVP cells. ‘251-EP was internalized in these cells in

exactly the same manner as the EP-responsive FVA cells. In

earlier work, the higher affinity receptors on the FVA cells

appeared to be preferentially endocytosed’; however, this

study shows equal endocytosis of higher and lower affinity

receptor since FVP and MEL have only lower affinity

receptor for EP. The presence of higher affinity receptors on

FVA cells and human CFU-E’2 may be responsible for their

capacity to respond to EP with erythroid maturation. How-

ever, the molecular mechanism for this effect is still unknown

since the present study shows that the structure of the

receptor for EP in cells containing either higher and lower

affinity receptor or only affinity receptors is identical in gross

structure.

REFERENCES

1 . Sawyer ST. Krantz SB, Goldwasser E: Binding and receptor

mediated endocytosis of erythropoietin in Friend virus-infected

erythroid cells. J Biol Chem 262:5554, 1987

2. Krantz SB, Goldwasser E: Specific binding oferythropoietin in

spleen cells infected with the anemia strain of Friend virus. Proc

Natl Acad Sci USA 81:7574, 1984

3. Sawyer ST. Krantz SB, Luna J: Identification of the receptor

for erythropoietin by cross-linking to Friend virus-infected erythroid

cells. Proc Natl Acad Sci USA 84:3690, 1987

4. Koury Mi, Sawyer ST, Bondurant MC: Splenic erythroblasts

in anemia-inducing Friend disease: A source of cells for studies oferythropoietin-mediated differentiation. J Cell Physiol 121:526,

1984

5. Sawyer ST, Koury Mi, Bondurant MC: Large-scale procure-

ment of erythropoietin-responsive erythroid cells. Methods Enzymol

147:340, 1987

6. Koury Mi, Bondurant MC, Atkinson JB: Erythropoietin con-

trol of terminal erythroid differentiation: Maintenance of cell viabil-

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

RECEPTORS FOR ERVTHROPOIETIN 109

ity, production of hemoglobin, and development of the erythrocyte

membrane. Blood Cells 13:217, 1987

7. Todokoro K, Kanazawa S, Amanuma H, Ikawa Y: Specific

binding of erythropoietin to its receptor on responsive mouseerythroleukemia cells. Proc Natl Acad Sci USA 84:4126, 1987

8. Mayeux D, Billat C, Jacquot R: The erythropoietin receptor of

rat erythroid progenitor cells. J Biol Chem 262: 1 3985, 1987

9. Pekonen F, Rosenlof K, Rutanen E-M, Fyhrquist F: Erythro-

poietin binding sites in human foetal tissues. Acta Endocriniol

(Copenh) 116:561, 1987

10. Sasaki R, Yanagawa 5, Hitomi K, Chiba H: Characteriza-tion of erythropoietin receptor of murine erythroid cells. Eur JBiochem 168:43, 1987

1 1 . Sawada K, Krantz SB, Kans J, Dessypris EN, Sawyer ST.

Glick AD, Civin CI: Purification ofhuman erythroid colony-forming

units and demonstration of specific binding of erythropoietin. J Clin

Invest 80:347, 1987

12. Sawada K, Krantz SB, Sawyer ST, Civin CI: Quantitation of

specific binding of erythropoietin to human erythroid colony-

forming cells. J Cell Physiol 137:337, 1988

13. Scatchard G: The attraction of proteins for small molecules

and ions. Ann NY Acad Sci 51:660, 1949

I 4. Sawyer ST, Krantz SB: Transferrin receptor number, synthe-

sis, and endocytosis during erythropoietin-induced maturation of

Friend virus-infected erythroid cells. J Biol Chem 261 :91 87, 1986

I 5. Laemmli UK: Cleavage of structural proteins during the

assembly of the head of bacteriophage T4. Nature 227:680, 1970

16. Tojo A, Fukamachi H, Kasuga M, Urabe A, Takaku F:

Identification of erythropoietin receptors on fetal liver erythroid

cells. Biochem Biophys Res Commun 148:443, 1987

17. Koury Mi, Bondurant MC, Graber SE, Sawyer ST: Erythro-

poietin messenger RNA levels in developing mice and transfer of

‘25I-erythropoietin by the placenta. J Clin Invest 82:154, 1988

18. Ruscetti 5, Wolff L: Spleen focus-forming virus: Relationship

of an altered envelope gene to the development of a rapid erythroleu-

kemia. Curr Top Microbiol Immunol 112:21, 1984

For personal use only.on January 3, 2019. by guest www.bloodjournal.orgFrom

1989 74: 103-109  

ST Sawyer, SB Krantz and K Sawada placentaReceptors for erythropoietin in mouse and human erythroid cells and 

http://www.bloodjournal.org/content/74/1/103.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 January 3, 2019. by guest www.bloodjournal.orgFrom