1V1SSI.J RETURNING MATERIALS:
Place in book drop to

LJBRARJES remove this checkout from

-:—_ your record. FINES will

 

 

be charged if book is
returned after the date
stamped be10w.

 

 

 

ALTERATIONS IN ERYTHROCYTE MEMBRANE STRUCTURE INDUCED BY
THERAPEUTIC AND PHARMACOLOGICAL AGENTS

By

Denise Lora Mazorow

A DISSERTATION

Submitted to
Michigan State University
in partiaI fquiIIment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biophysics

1984

ABSTRACT

ALTERATIONS IN ERYTHROCYTE MEMBRANE STRUCTURE INDUCED BY
THERAPEUTIC AND PHARMACOLOGICAL AGENTS

By

Denise Lora Mazorow

Many compounds are known to affect erythrocyte membrane physical
structure. In this study, evidence is given, that many of these com-
pounds may modulate the membrane via the cytoskeletal matrix in intact
cells.

Addition of 1 to 1000 prostaglandin molecules per cell were shown
to change membrane structure as detected by electron spin resonance
spectroscopy (ESR). It is unlikely that the membrane changes were the
result of simple membrane lipid-lipid interactions. It is proposed
that prostaglandins have specific receptors inducing the structural
changes. Furthermore, glucose prevented the structural changes induced
by prostaglandins. It is believed that the extrinsic membrane proteins
are linked either directly or indirectly to each other fonning the
cytoskeletal matrix and through specific interactions are bound to.
integral membrane proteins. Thus, changes in cytoskeletal structure

would obviate changes in overall membrane structure.

Chlorpromazine, a drug known to have specific lipid interactions
was added to erythrocyte suspensions. Dose response curves of membrane
structural changes upon addition of chlorpromazine obtained by both ESR
and hypotonic hemolysis studies were in agreement. At 10-5 M,
chlorpromazine induced maximal disordering and maximal protection to
hypotonic hemolysis. Upon further addition of chlorpromazine, it is
believed that all the protein binding sites become saturated and
chlorpromazine begins to enter the lipid matrix with increasing doses,
the drug binds to membrane lipid, until acting as a detergent, it
disrupts the membrane.

Ingestion of aspirin and indomethacin, like addition of chlor-
promazine and prostaglandins also induce membrane structural changes at
therapeutic concentrations as detected with ESR. The degree and time
course of membrane changes from erythrocytes taken from male and female
subjects after ingestion of either aspirin or indomethacin were differ-
ent for men and women, and for women were dependent on the menstrual
cycle. In vitro addition induced no structural changes in erythrocyte
membranes. Therefore, the changes in membrane structure after drug
ingestion must be induced by metabolic products.

Finally, even alterations in the anionic milieu can apparently
induce membrane alterations. Band 3, which is the anion port is
attached to the cytoskeletal matrix. Thus changes in ion content can

also induce membrane structural changes.

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to my major
adivsor Dr. Estelle McGroarty, who has given me adivce and encourage-
ment throughout. I would also like to acknowledge my appreciation to
Dr. Alfred Haug and Dr. Robert Bull for their insight and thoughtful
discussions. Finally, I would like to thank Dr. George Padgett in his
cooperative assistance. These members of my dissertation committee
gave me invaluable direction and I wish to thank them for this and the
review of this work.

Finally I would like to express my sincerest thanks to my Mother,

Father and Grandparents for their encouragement and support.

11

TABLE OF CONTENTS

List Of Tab]es O OOOOOOOOOOOOO 0.00.00.00.00...0.0.0.... ...... 0....

List of Figures ......................... . ......... ........ ......

Chapters

1. Background ................................................
. General Erythrocyte Membrane Structure ................
. Band 3 ................................................
. Prostaglandins (General) ..............................
Prostaglandin Membrane Receptors ......................
Prostaglandin Effects on Erythrocytes .................
. Aspirin ...............................................
. Chlorpromazine Effects on Membrane Structure ..........
Objectives ............................................
Bibliography ...... ....... .....................................

Imfimonw>
C

II. Alterations in the Human Erythrocyte Membrane Induced by
Anions. I. Anions Affect Membrane Physical Structure......

AbStraCt .0OOOOOOOOOOOOOOO0.0.0..OOOOOOOOOOOOOOOOOOOOOOOOOOOOO.

Introduction ..................................................
Materials and Methods .........................................
Hypotonic Hemolysis .......................................
ESR Sepctral Analysis .....................................
Computer Analysis .........................................
Results .......................................................
Anion Perturbation ........................................
Ion Gradient Perturbation .................................
Hypotonic Hemolysis Studies ...............................

Discu551on 00......00......O...00.0.00...OOOOOOOOOCOOOOOOOOO...

Bibliography O...O...0....0.......0...OOOOOOOOOOOOOOOOOOOO...O.

III. Alterations in the Human Erythrocyte Membranes by
Anions. II. Anions Affect Prostaglandin E1 and £2

Receptors O00.0.0.0...0....O...00......OOOOOOOCOOOOCOOOOOO

AbStraCt O.IO...OOOOOOOOOOOOOOOO00.0.00...OOOOOOOOOOOOOOOOOOOOO
IntrOdUCt1on COO...0......0.0...OOOOOOOOOOOOOOOOO...0.00.0.0...

MethOds 0.0.0.0000...IOOOOOOOOOOOOO00......OOOOOOOOOOOOOOOOOOOO

iii

V

v1

61
62
66

ReSUItS 0.00.00.00.000......0.0.0.000000.00.00.000000000.0.0...

Ear1es BUffer 000.000.000.00000000.00.0000.00.00.00.000.00.
Phosphate Buffer 00.000.000.0000000... ....... 0.000.000.0000
HEPES BUffer 0.0.0.0...00.0....00000000.0.000.000.000000000

DiSCUSSion 0000000000000.000000.00.0.00.00.00.00000000.00.00...

Bibliography 0.0000000000000000000.0.0.0.0.0000...0.0.0.0000...

IV. Effects of Aspirin, Indomethacin, and Sodium Salicylate
in Human Erythrocyte Membranes as Detected with Electron
Spin Resonance Spectroscopy ...............................

Abstract ......................................................
Introduction ..................................................
Methods .......................................................
Results .......................................................

DTSCUSSTO" 0.0.00000.000.000000.0.0.0...0000000000.0.00.0..0000

BibITOQraphy 00.00.000.00...0.00.00.00.00...000.000.000.000.000

V. Chlorpromazine Induced Changes in Human Erythrocyte
Membrane StrUCture 00000000000000.0000.00.000000000000000.0

Abstract .................. ........... ......... ..... ...........
Introduction ..................................................
Materials and Methods .........................................
Results .......................................................
Discussion ....................................................
Bibliography ..................................................

VI. Concluding Statements .............. .......................

conCIUSionS 0.000000...0000.00.00.0000.000.00.000000000000.00..

Appendix 0.00.0.0.0.0.0.000.000000000000000000.000.000.00000000

Abstract 0.0000000000000000000000.0...0.000000000000000000.0000

Introduction ..................................................
Material and Methods ..........................................
Preparation of Cells ......................................
Electron-Spin Resonance Spectroscopy ......................
Data Analysis .............................................
ReSUItS 000000000000000000000000000...00000000000000.0000...000
Control Erythrocytes ......................................
Effects of Membrane Stabilizing AGents on Membrane
Structure ...............................................
DTSCUSSTO" O...000.000.00.000.00000000000000.0000...00000000000
conCIUSIOnS 00.00.00...00.00.0000.00.000000000000000000000.0.00

References 0.00.00.0.000.00.00.000000000000000.000.000.00...0.0

iv

108

109
110
111
115
120
123

125
126
130
130
131
133
133

135

.135

135

139
140
143
143

LIST OF TABLES

 

 

Table
Chapter II
1. Effects of buffers on erythrocyte membrane phase
tranSition 0.00000000....00.000.000.00000.00000...0000.00.00.
2. Effects of perturbants in Earles' modified salt on
eWthrocyte phase tranSition 00000000000000000000000000.0000.
Chapter IV
1. Changes in order parameter of erythrocytes drawn from males

and females and suspended in either Earle's salts or
filtered serum containing either aspirin, sodium salicylate,
or indomethaCin 0.0.0.0000.000.000000.0.00.000000000000000.00

44

44

99

LIST OF FIGURES

Elms 35g

10

Chapter I

A schematic of the major erythrocyte membrane components.
The outer monolayer of the membrane contains the
phOSpholipids sphingomyelin (SM) and phosphatidylcholine
(PC). The inner monolayer contains the phospholipids
phosphatidylserine (PS) and phosphatidylethanolamine (PE).
Cholesterol (Ch) is distributed approximately equally in
both monolayers. The intrinsic proteins are the anion
port, Band 3, which is a dimer and Band 7. The
glycoproteins are glycophorin (PAS 1 and 2), which is also
a dimer and PAS 3. Spectrin is shown as a heterodimer
(Bands 1 and 2). The actin oligomer (Band 5) is attached
to the ends of the spectrin heterodimer. Band 4.1
attaches the spectrin-actin complex to the membrane. The
exact placement for this attachment is not known. Ankyrin
also known as band 2.1, attaches spectrin to Band 3. Band
6, the monomeric form of glyceraldehyde-3-phosphate dehy—
drogenase (GPD), is attached to the cytosolic tail of Band

0 .00.00.000...0.0.0..00000.0.0..0000000.00.00.0000000000..0 7

Both the leukotriene pathway and the prostaglandin (PG)
pathway begin with arachidonic acid. Arachidonic acid is
converted to PGGZ by the enzyme cyclo-oxygenase. PGGZ

is further converted to PGHZ. PGHZ can be further
converted to PGan, PGIz (prostacyclin), or thromboxane A2
(TXAZ). This conversion is tissue dependent. PGF a
under the proper conditions can be converted to PGEZ.

An alternate pathway for the oxygenation of arachidonic
acid is via lipoxygenase to the leukotriene pathway. An
example of this oxygenation is the conversion of arachi-
donic acid to 5-hydroperoxyeicosatetraeonic acid (5-HPETE)
to leukotriene A4. The placement of oxygenation and
resulting leukotriene is tissue dependent.................... .12

vi

£12212 Eggg
Chapter II

 

1. Electron spin resonance spectra of hunan erythrocytes
labelled with S-doxyl stearate. The spectra were taken at
the temperatures indicated. Absence of free probe is
indicated by the lack of a major absorption signal at
points indicated by arrows. Symmetry of high and low
field peaks reveals that the probe is in a single
environment. .......... ........ .......... . .......... . ........ 38

2a. The temperature dependent change in the hyperfine split-
ting parameter (2Tu) of erythrocytes labelled with
5-doxyl stearate and suspended in Earles salts. ............. 40

2b. The temperature dependent change in the hyperfine splitt-
ing parameter (2Tu) of erythrocytes labelled with
5-doxyl stearate and suspended in Earles salts containing
10-7 M A23187. ........... . ........... .... ..... ... ....... 41

3. The temperature dependent change in the order parameter
(S) )of erythrocytes labelled with 5- doxyl stearate and
suspended in Earles salts (O-——0) or in Earles salts with
MA23187( (Cl—D). ..................... 42

4. Change in order parameter of cell labelled with 5-doxyl
stearate and suspended in various buffers compared to
cells suspended in filtered serum. Membrane order was
measured at 24°C (a) and 37°C (b). .......................... 46

5. Change in order parameter of cells labelled with 5-doxyl
stearate and suspended in Earles salts containing differ-
ent perturbants compared with cells suspended in Earles
salts alone. Membrane order was measured at 24°C (a) and

37°C (b). 0..00.0..0.0.0.0....0.00000...0000.00.00.0000000000 49

Chapter III

 

1. Percent changes in order parameter (AS%) of human erythro-
cytes gram female volunteers in the presence of 10'8 M
- 10'1 M prostaglandin E (PGEl ), prostaglandin E (PGEZ),
and prostaglandin 16, 16 Dimethyl PGE (di-Me PGE T
suspended in Earles' salts, Earles saIts without a2+, and
Earles salts plus glucose A. at 24°C, and B. at 37°C. ....... .68

2. Percent changes in order parameter (AS%) of human erythro—
cytes from female volunteers in the presence of 10-9 - 10'11
M prostaglandin E1 (PGEl), prostaglandinE

(PGE ), and prostagland1n 16,16 dimethyl PGE (di-Me

PGE I suspended in phosphate buffer, pH 7.0 (POW

7. 0); phosphate buffer, pH 7. 4 (P04 pH 7. 4), phosphate

vii

£19111? 135g;

buffer with 0.5 mM bicarbonate pH 7.0 (P04+ HC03 pH

7.0); phosphate buffer with 0.5 mM bicarbonate, pH 7.4

(PO + HCO3 pH 7.4), HEPES buffer, and HEPES buffer pH 7.4
(HEPES); and HEPES buffer with 11 mM bicarbonate pH 7.4

(HEPES + H603) A. at 24°C and B. at 37°C. ................. 72

Chapter IV

Electron spin resonance spectrum of human erythrocytes

labelled with 5-doxyl stearate. The spectrun was taken at

37°C and the scan range was l00 gauss. Absence of free

probe is indicated by the lack of a major absorption sig-

nal at points indicated by arrows. Symmetry of high and

low field peak reveals that the probe was present in a

single environment in the membrane. ......................... 87

Erythrocytes were drawn from male (open circles) and
female (closed circles) donors. Spectral parameters

2T1] (hyperfine Splitting parameter) and S (order
parameter) were plotted versus temperature (°C). A.
Temperature dependence of membrane structural parameters
of erythrocyte from male and female subjects prior to
ingestion of any drug. B. Temperature dependence of male
and female subjects prior to ingestion of any drug. C.
Temperature dependence of erythrocytes drawn from female
subjects at the beginning of the menstrual period and from
male subjects l hour after ingestion of aspirin. ............ 91

Time related changes in order parameter of erythrocytes

from female subjects at the beginning of the menstrual

cycle (D ), from female subjects at the middle to end of

the menstrual cycle ( ), and from male subjects (I )
following aspirin ingestion. Time related changes in

order parameter of erythrocyte membranes from female sub-

jects at the beginnin of the menstrual cycle (Im ) and

from male subjects ( ) following ingestion of sodiun
salicylate, A. measured at 24°C or B. measured at 37°C. ..... 94

Time related changes in order parameter of erythrocytes

from female subjects at the beginning of the menstrual

cycle (I:] ), from female subjects at the middle to end of

the menstrual cycle (I), and from male subjects (I )
following ingestion of indomethacin, measured either at A. ,
24°C, or at B. 37°C. ........................................ 97

 

Chapter V

The temperature dependent changes in the hyperfine splitt-
ing parameters (2T11) of erythrocytes labelled with

viii

Figure PAGE

5-doxyl stearate suspended in Earles salts (-o-o-) or in
Earles salts with 2 x 10"5 M chlorpromazine

ba-u-L H.n.u.n.u.u.u.u.u.u.u.u.u.n.u.H.” 1M

2. The temperature dependent change in the order parameters
(S) of erythrocytes labelled with 5-doxyl stearate and
suspended in Earles salts (-o-o-) or in Earles' salts with
2 x10'5 M chlorpromazine (-c1-n-). 116

3. Chlorpromazine (-o-o-) or chlorpromazine methiodide
(-o-o-) induced concentration dependent changes in A.
order parameter (S) at 24°C, B. order parameter (S) at
37°C, C. hemolysis in hypotonic Earles buffer at 24°C. ..... . 118

Appendix A

 

1. Electron-spin resonance spectra of human erythrocytes
labE]Ed With 5-dow] Stearate 0000.000000000000000000000.000. 130

2. Hyperfine splitting parameter, 2Tu (Gauss), as a
function of temperature in erythrocytes labeled with
5-doxyl stearate in the absence of perturbants (A), and
in the presence of 5 x 10'4 M propranolol (B) ............... 131

3. The temperature dependence of the order parameter, S, in
erythrocytes labeled with 5-doxyl stearate in the absence
of perturbants (closed circles) and in the presence of
3 x 10'4 M diazepam (open circles) .......................... 132

ix

CHAPTER I

Background

A. General Erythrocyte Membrane Structure

Erythrocytes have been the subject of study since l878 when Herman
Nasse (I) reported that the plasma chloride concentration decreased
when erythrocyte suspensions were exposed to increasing partial pres-
sures of carbon dioxide. Since then the distributions and flux rates
of anions involved with the erythrocyte have been intensively studied.
The chloride-bicarbonate exchange is very rapid with a half time of 0.3
sec. (2, 3). Phosphate and sulfate have much slower rates of exchange
(4-6).

Later, interest was shown in the erythrocyte membrane itself.
Indications are that the phospholipid classes are asymmetrically dis-
tributed in specific halves of the bilayer. This evidence has mostly
been obtained from studies using specific phospholipases (7-ll). Puri-
fied phospholipase added to whole cells converts about 50% of the phos-
phatidycholine (PC) to lysolethicin. All the PC can be hydolyzed by
the same amount of phospholipase A2 when added to leaky ghosts. Two-
thirds of the PC is in the outer monolayer if one calculates the amount
of PC degraded in whole cells versus that degraded in ghosts. Using
purified sphingomylinase in a similar manner, it has been determined
that 80-85% of the sphingomyelin is located in the outer monolayer.
Trinitrobenzenesulfate (TNBS), a probe for amino phospholipids (l2),

does not permeate the erythrocyte membrane (l3). Fluorodinitro—

benzene (FDNB), which easily passes through the membrane (l3, l4), is
also a probe for amino phospholipids. The results using FDNB and TNBS
indicate that 80 to 90% of the total phosphatidylethanolamine (PE) and
phosphatidylserine (PS) are located in the inner monolayer of membrane.
These data indicate that the major phospholipid components in the outer
monolayer are choline containing phospholipid (PC and sphingomyelin)
and the major phospholipid components of the inner monolayer are the
amino phospholipids (PE and PS). Another demonstration of the
asymmetric distribution of the membrane phospholipids is that PC and
sphingomyelin can be exchanged from one membrane to another; whereas PS
and PE can not (ls-l8).

Unfortunately, it is not known how or where cholesterol is located
in the membrane even though approximately half the lipid on a molar
basis is cholesterol (l9). However, it is known that cholesterol is
relatively easy to extract from erythrocyte membranes under conditions
that do not remove any of the other lipids (20-22). Cells with a
cholesterolzphopholipid ratio near 2 have an increased surface area and
become broad, flattened cells. The membrane lipids of these cells have
restricted motion as seen using the fluorescent probe 12-(9-anthroyl)
stearic acid (23).

Like the phospholipids the erythrocyte membrane proteins are also
asymmetric (24). The ghost membrane is composed of 8 major polypeptide
chains when analyzed by sodium dodecyl sulfate polyacrylamide gel elec-
trophoresis (SDS-PAGE) (25). Along with the 8 major polypeptides there
are 4 major glycopeptides as seen by periodic acid-Schiff's reagent
staining (PAS) (26, 27). These four PAS-staining bands represent the

major source of sialic acid found in the hunan red blood cell.

Singer and Nicolson (24) developed the approach of classifying the
proteins as either intrinsic or extrinsic, extrinsic proteins being
those which can be removed from ghost membranes by altering the ionic
strength or the pH of the media. The extrinsic proteins correspond to
the SDS-PAGE protein bands I, 2, 4, 5, and 6. The intrinsic proteins;
therefore, correspond to SDS—PAGE protein bands 3, 7 and PAS protein
bands l, 2, 3, and 4.

The extrinsic proteins, which are approximately 40% of the total,
are believed to lie totally in the cytosol. The major evidence for
this is that none of these polypeptides are degraded when whole cells
are treated with proteolytic enzymes (25, 28, 29). However, when leaky
ghosts are treated with the proteolytic enzymes, each of these proteins
is degraded.

Bands l and 2 are the spectrin polypeptides. Band l (a subunit,
240,000 daltons) and band 2 (3 subunit, 220,000 daltons) can self-
associate to form the spectrin tetramer (a8)2 with a molecular weight
of 920,000 daltons (30). It is now fairly clear that spectrin dimers
and tetramers can form long, filamentous polymers (31).

Band 2.l, now also known as ankyrin, binds one mole per mole spec-
trin heterodimer (32). Ankyrin binds the spectrin heterodimer to the
membrane by also binding to the integral protein band 3 (33-35).

Actin, which is band 5 has a molecular weight of 45,000 daltons.
The spectrin tetramer, arranged by head to tail association of hetero-
dimers (3T), binds actin to the tail ends. The complexes when formed
in the presence of band 4.1, have a higher viscosity and are better
able to survive sedimentation through sucrose than those without band

4.l (30, 36, 37). This complex is extremely sensitive to micromolar

concentrations of calcium (36). Furthermore, it has been found that
Iggly in the presence of band 4.l do spectrin-actin gels become thixo-
tropic (36), and thus may play a role in erythrocyte membrane deform-
ability: Band 4.l along with ankyrin helps to hold the cytoskeletal
matrix to the membrane and both may be important in modulating erythro-
cyte shape changes.

The last of the extrinsic proteins, band 6, is the monomeric form
of glyceraldehyde-3-phosphate dehydrogenase (38, 39). The polypeptide
appears to be associated with the cytoplasmic portion of band 3 (40).

0f the integral membrane proteins band 3 and PAS l are the most
common and both appear in the membrane as dimers (4l). Band 3 appears
to be composed of 5-8% carbohydrate on a dry weight basis (42, 43).

The sugars in this carbohydrate are mostly mannose, galactose, and
N-acetylglucosamine (1:2:2) with trace amounts of fructose and glucose.
PAS-l and PAS-2 are believed to be interconvertible forms of glyco-
phorin-A (44) and represent 75% of the sialoglycopeptide (44). The
portion of the glycophorin/dimer located on the cytoplasmic side of the
membrane is believed to be in close association with spectrin. When
erythrocytes are incubated with antispectrin antibodies, a large shift
in the glycophorin distribution on the cell surface is reported to
occur (45). Little is known of the other integral membrane protein,
band 7.

Other important proteins are known to exist in the erythrocyte
membrane but not in large enough numbers to clearly be seen on gels.
These include the Na-K ATPase and Ca-Mg ATPase. To visualize the

placement of the erythrocyte membrane components see Figure l. Both

Figure 1. A schematic of the major erythrocyte membrane compo-
nents. The outer monolayer of the membrane contains the phospho-
lipids sphingomyelin (SM) and phosphatidylcholine (PC). The inner
monolayer contains the phospholipids phosphatidylserine (PS) and
phosphatidylethanolamine (PE). Cholesterol (Ch) is distributed
approximately equally in both monolayers. The intrinsic proteins
are the anion port, Band 3, which is a dimer and Band 7. The
glyc0proteins are glycophorin (PAS 1 and 2), which is also a dimer
and PAS 3. Spectrin is shown as a heterodimer (Bands 1 and 2).
The actin oligomer (Band 5) is attached to the ends of the spec-
trin heterodimer. Band 4.1 attaches the spectrin-actin complex to
the membrane. The exact placement for this attachment is not
known. Ankyrin, also known as band 2.1, attaches spectrin to Band
3. Band 6, the monomeric fonn of glyceraldehyde-3-phosphate
dehydrogenase (GPD), is attached to the cytosolic tail of Band 3.

 

ME ....s... a...
5.88
253.0

5.2 ....
... N gang. a 33. BB

 

0
.9

ma Mn. :0 mm :0 mm mm. ma :0 mn. ma

j m f m: f 2

:0 .2m :0 2m 2m :0 0n. 2m On. On.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

haz<m

tan— :o.:< ad: m<m m m<1
n 94m . 53.3320 033:0

the lipids and proteins are placed in their relative positions, but the

concentrations are not necessarily correct.

B. BAND 3

Band 3, which comprises 25% of the membrane protein, is believed to
occur as a dimer. It has a molecular weight between 88,000 and l05,000
daltons (25). The function of band 3 in the membrane is, apparently,
to act as a passive port for anion flux. The most compelling evidence
for band 3 involvment in anion transport is that covalent labelling of
band 3 with a variety of compounds, most notably the stilbene deriva-
tives, will inhibit anion flux (46-48). 4,4'-diisothiocyano 2,2'stil-
bene disulfonate (DIDS) was first used by Cabantchik and Rothstein (47,
48). Approximately 300,000 copies of [3H] 0105 were reported bound
to a single erythrocyte and this binding was almost exclusive to band
3. Furthermore, DIDS binding to the membrane inhibited anion exchange.
Experiments in which the disulfonic stilbenes (DIDS and 4-acetamido-4'-
diisothiocyano-2,2'-disulfonic acid stilbene [SITS]) have been incor-
porated into resealed ghosts (49, 50) or have been reacted with
inverted sealed vesicles (51) have not inhibited anion flux or bound to
band 3. Thus, like the asymmetry exhibited in membrane architecture,
0105 and SITS bind only to the external side of the cell.

The model for anion flux that best fits the data is that of a ping-
pong model. This ping-pong model simply proposes that one anion is
translocated before the counterion binds and is transported. The ping-
pong model predicts that if cells are suspended in solutions with dif-
fering internal and external calcium concentrations, the half- satura-
tions concentrations for extracellular calcium flux will be reduced

when the intracellular chloride is reduced (52). The prediction has

proven true. An essential feature of the ping-pong model is that there
are two forms of the unloaded carrier, one facing in and one facing
out. It is the existence of these two forms along with the fact that
anions are generally in equilibrium across the membrane (53, 54) which
accounts for the strange behavior of calcium just described.

Band 3 probably exists as a dimer in the membrane. Band 3 when
extracted in Triton X-l00 is a dimer (35). Almost quantitative cross-
linking of reactive sulfhydryl groups can be accomplished by reducing
crosslinking from the cytoplasmic side (4l, 56). Fluorescent labelling
of cross-linked band 3 doesn't affect its rotational mobility (57).
Counting the nunber of particles seen by freeze-fracture electron
microscopy yields about half the number of band 3 monomers one would
expect (25, 28).

The question has been raised on the importance of integrity of band
3 as a dimer for its function. Evidence suggests that the dimeric
conformation is not needed to allow anion flux. Dixon plots for
several aromatic sulfonic acid inhibitors are linear (59, 60), which
indicates that one inhibitor reacts with one monomer. Further evidence
suggests that one 0105 molecule binds to one band 3 monomer (6l).

A more complete discription of anion exchange can be found in sev-
eral review articles (62, 63). Also of interest in this thesis are how
certain drugs and hormone like compounds interact with the erythrocyte
membrane. The next several sections are devoted to giving some

background information on these compounds and their interactions. '

C. PROSTAGLANDINS (GENERAL)

10

The prostaglandins (PCs) are a group of oxygenated fatty acids
which metabollicaly originate from the fatty acid arachidonic acid.
They are one of the most biologically potent compounds known and can
act either as bioregulators or can take part in pathological states.
Release of arachidonic acid can be affected by hormones directly or
indirectly (64), inflammatory or immunological stimuli (65), calcium
ionophores (66), ultraviolet light (67), tunor promotion agents (68),
and mechanical agitation (69). The arachidonic acid is converted to
the endoperoxide intermediate PGGZ by the enzyme cyclooxygenase (70).
From PGGZ the next step is determined by the tissue under considera-
tion and the tissue enzyme content. For example, platelets will con-
vert the PGGZ primarily to thromboxane A2 (TXAZ), whereas the
aorta will synthesize prostacyclin (P612) (Figure 2). PGE] and
PGE3 are much less common than the other prostaglandins and are
synthesized from eicosanoic acids other than arachidonate.

Cis 5, 8, ll, 14 eicosatetraenoic acid (arachidonic acid) is the
only eicosanoic acid converted enzymatically to active thromboxanes and
prostaglandins. TXAz has a half life in biological tissues of sec-
onds and is the most potent of the known arachidonic acid derivatives.
Its function is to contract the aorta and to trigger platelet aggrega-
tion (71). P612, the natural antagonist of TXAz, relaxes aortic
tissue and prevents the aggregation of platelets (72). The action of
non-steroidal anti-inflammatory drugs regulates the synthesis of all
prostaglandins since they inhibit cyclo-oxygenase activity.

An alternate pathway for oxygenation of arachidonic acid is via the
enzyme lipoxygenase (73). The primary products of this pathway are the
hydroperoxyeicosatetraenoic acids (HPETES). The substitution (5-HPETE,

11

Figure 2. Both the leukotriene pathway and the prostaglandin (PG)
pathway begin with arachidonic acid. Arachidonic acid is con-
verted to PGG by the enzyme cyclo-oxygenase. P002 is further
converted to GHZ. PGH can be further converted to

PGan, PGIZ (prostacycl1n), or thromboxane A2 (TXAZ).

This conversion is tissue dependent. PGFZ under the proper
conditions can be converted to PGE . An aTternate pathway for
the oxygenation of arachidonic acig is via lipoxygenase to the
leukotriene pathway. An example of this oxygenation is the con-
version of arachidonic acid to 5-hydroperoxyeicosatetraeonic acid
(S-HPETE) to leukotriene A4. The placement of oxygenation and
resulting leukotriene is tissue dependent.

12

«8..
z 2

“b
’I

2000

. __, a 8.5..

Ion—
. .. a .. 2.
306034 ‘- 20003
a .3

 

.. «8..
< 0539.3. so
a
z 3000 q
.88 ... (41/? so
3326 28 28253.

.22me l HMO

13

l2-HPETE) depends on the cell type synthesizing the compound. The
class of compounds from this pathway are called leukotrienes (Fig. 2)
(74).

When it became known that indomethacin and aspirin both inhibited
prostaglandin synthesis (75) many laboratories began to study the anti-
inflammatory role of prostaglandins. Glucocorticoids will also sup-
press prostaglandin synthesis (76) through a complex set of cellular
responses. In general, the glucocorticoids will react with a specific
membrane receptor which can alter protein synthesis. This results in

decreased amounts of esterified fatty acid released.

D. PROSTAGLANDIN MEMBRANE RECEPTORS

In 1974, Smigel and Fleisher (77) reported techniques for measuring
and characterizing prostaglandin binding to membrane isolates. They
determined that the PG receptor was in the plasma membrane fraction of
rat liver isolates and that this receptor was of a single type for
binding the E type prostaglandin. Other PGE] binding sites were
studied, most of which were in the plasma membrane fraction (77). PGE
type receptors have been found in non-pregnant human myometrium (78),
skin tissue (79, 80), bovine corpus luteum (8l), human erythrocytes
(82-84), and frog erythrocytes (85).

To demonstrate that the prostaglandins were binding to a protein
receptor, (80), trypsin, pronase, or boiling were used to show a marked
reduction of PGE2 and PGFz binding. A second study showed that
upon exposing rat skin to UV light there was loss of PGE2 binding
capability (79). This loss could be prevented by addition of a protein
sulfhydryl oxidizing agent or by a lipid anti-oxidant. The authors

14

suggested that the UV light initiated reduction of needed disulfide
groups and peroxidation of lipids which are needed in PG-receptor
binding.

Human platelet membranes have been reported to have two distinct
receptors, one for P612 and another for PGDZ (86, 87). It was
shown that binding P612 to platelets induced a large increase in
cAMP. But, if PGE] and PGE2 were added, PGIZ stimulation of cAMP
production was inhibited. Platelets are not the only system in which
PGE can affect enzyme activity or substrate production. Addition of
PGE] to normal or transformed fibroblasts in culture resulted in an
activation of adenyl cyclase (88). In these studies, PGE] binding
was dependent upon calcium.

These results indicate that prostaglandins have specific membrane
receptors. Furthermore, these receptors are likely to be polypeptides

and appear to affect the regulation of other membrane events.

E. PROSTAGLANDIN EFFECTS ON ERYTHROCYTES

Very low doses of prostaglandins (l to lO molecules per cell) have
been reported to dramatically affect erythrocyte membrane strucutre as
well as whole cell properties such as deformability. The prostaglandin
binding to the erythrocyte membrane must induce an amplification. One
to ten molecules of a fatty acid derivative could not induce general-
ized membrane structural changes unless this amplification occurred.

PGEz, when added to hunan erythrocytes, caused a decrease in A
whole cell deformability (82). Kury et al (83), using blood fran only
male subjects repeated the work of Allen and Rasmussen (82). They also

showed that PGEz not only caused a decrease in deformability, but

15

also caused an increase in membrane order parameter (electron spin
resonance study, ESR). Further studies with PGE], a PGEz antago-
nist, showed that PGE] caused an increase in cell deformability and a
decrease in membrane order. The maximal effect for either PGE] or
PGEZ occurred at lO'I‘ M prostaglandin. This was approximately

l0 prostaglandin molecules per cell. A requirement for the membrane
perturbation with either PGE] or PGEZ was the presence of calciun.
Therefore, like the study with cultured fibroblasts, calcium is needed
for PGE] to initiate not only adenylcyclase activity (88) but also to
induce erythrocyte membrane structural changes.

PGE] not only affects lipid membrane bulk properties as indicated
by the studies of Kury et al (83) and Allen and Rasmussen (82), but it
induces protein structural changes. PGE] and PGFZ both reportedly
cause changes in the circular dichroism (CD) absorption spectrum of
washed erythrocyte ghosts between l90 nm and 250 nm (84). These
changes indicate changes in membrane protein a—helical and s-sheet con-
tent. These changes in CD absorption spectra were blocked by the
PGE] antagonists 7-oxa-13-prostynoic acid (89).

PGE] and PGE; not only alter erythrocyte membrane lipid and
protein structure, they also affect hemolysis. PGE] is reported to
protect erythrocytes from hypotonic hemolysis, while PGEZ has the op-
posite effect; i.e., it is lytic (90). This agrees with ESR and de-
formability data already presented. Although PGE1 and PGE2 appear
to protect erythrocytes from mechanical lysis, the degree of proteCtion
seems to be dependent on whether the erythrocytes were drawn from male
or female subjects (84). Erythrocytes drawn from male volunteers were

more easily lysed than cells drawn from female volunteers; however, the

16

degree of protection from lysis given by PGE] was more pronounced for
cells from men. Calcium did not alter PGE affects on lysis. The other
studies presented earlier indicated a calcium requirement for a PG
induced effect (82, 83). Perhaps the prostaglandins allow leakage of
calcium into the cell. Increased internal levels of calcium may induce
specific secondary reactions, which would reflect the need for calcium

in some cases and not in others.

F. ASPIRIN

Both aspirin and indomethacin are cyclo-oxygenase inhibitors and
thus prostaglandin synthesis inhibitors. The effects of indomethacin
and aspirin are apparently ubiquitous, since prostaglandins are
ubiquitous.

Therefore, it is not surprising that aspirin affects not only
platelets (70, 91) and arterial walls (92) but also erythrocyte mem-
brane structure (93). Aspirin added to heparinized whole blood will
become deacetylated. Approximately 80% of the acetate is unbound and
can be dialyzed out. l0% of the aspirin was bound to plasma proteins,
90% of which was to albumin. 0.4l% of the aspirin was bound to the
erythrocyte membrane. The aspirin and/or the labelled acetate presum-
ably passes through the membrane since spectrin can become acetylated.
Along with spectrin, acetylation of 5 or 6 other peptides also has been
reported (93); the acetylation was temperature dependent.

It is also known that men have longer bleeding times than women.
After aspirin ingestion, the bleeding times for both women and men are

increased. However, the increase is larger for men. Interestingly,

 

17

PGEZ will activate the secondary reactions which reduce bleeding
prior to inhibition of prostaglandin synthesis by aspirin.

Even though aspirin has long been known to affect bleeding times of
men more than that of women little work has been done to determine why.
It is known that one hour after aspirin ingestion women have higher
aspirin levels (1.6 ug/ml) in the plasma than do men (1.0 ug/ml)(95).
However, the differential absorption of aspirin between men and women
doesn't afffect platelet synthesis of TXBz. 650 mg aspirin causes a
total inhibition of thromboxane release for both men and women. It was
thought that the differences in bleeding times between men and women
was the result of differing PGF2 circulating in plasma. However, the
PGF2 levels have been shown to be the same for both men and women
(96).

Estradiol and/or estrogen levels may be the cause for the differ-
ences in the effects of aspirin in men and women. Aspirin plasma lev-
els are lower for women taking estrogen based contraceptives. Aspirin
plasma levels and half-life return to normal 3 to 5 months after stop-
ping estrogen contraceptive intake. This indicates that increased
levels of estrogen will ultimately change aspirin absorption (97).
Estradiol (98), when applied to the aorta, enhanced prostacyclin pro-
duction, but when applied to platelets did not affect thromboxane pro-
duction. This increased ratio of PGIleXBz may be one of the rea-
sons why pre-menopausal women have a lower incidence of ischemic heart

disease (99).

18

G. CHLORPROMAZINE EFFECTS ON MEMBRANE STRUCTURE

Chlorpromazine (CPZ) is believed to be able to pass through the
erythrocyte membrane (100-102). Evidence suggests that CPZ can then
bind to the bulk membrane lipids or to membrane proteins on the cyto-
plasmic side (103, 104).

The lipid binding properties of CPZ are fairly well established.
CPZ preferentially binds to PC, PS, and phosphatidyl inositol (105),
which, for the most part, are inner monolayer lipids. The amount of
CPZ bound to dipalmityl phosphatidyl choline liposomes increased dram-
atically above the phase transition at pH 7.4 (106). Also, CPZ
quenches perylene fluorescence. Perylene is known to be situated in
the hydrocarbon region of the membrane (106). High concentrations.
(0.2-1 mM) of CPZ were shown not to affect lipid head group movement of
ghosts (ESR study), but using lZ-doxyl stearate, it was determined that
the molecular motion in the lipid core was increased (107).

CPZ is known to modulate other membrane mediated events. 23 uM CPZ
will change erythrocytes from discocytes to stomatocytes (108, 109) and
eventually to spherostomatocytes (109) inducing a concomitant decrease
(9-10%) in electrophoretic mobility (108) and in suspension viscosity
at high shear rates (110).

CPZ, along with other phenothiazines, binds nonspecifically to
calmodulin (lll). Calmodulin, a calciun modulating protein, is known
to regulate Ca2+-Mg2+ ATPase activity (112-116). 125I-labelled
calmodulin can be derivatized with methyl-4-azidobenzoimidate (117).
This azido calmodulin has been shown to activate the Ca2+- Mg2+
ATPase in the erythrocyte membrane (118) and to photochemically cross-
link with calmodulin binding proteins (118). Binding of this labelled

calmodulin is calcium dependent (118). It has been shown to bind to

19

Ca2+, 1492+ ATPase and to a protein which is 230,000 daltons,
about the same size as spectrin (118).

CPZ has been shown to decrease the Na+, K+-ATPase activity in
erythrocytes (119). If the Na+, K+-ATPase inhibitor ouabain, is
added to cells prior to CPZ, CPZ reportedly does not protect against
lysis (119). Also, the ATP content of CPZ treated cells has been shown
to drop to 10% of that of untreated cells (104), and membrane protein
dephosphorylation of an 180,000 daltron protein and several proteins in
the 60,000-80,000 dalton rage is induced (104).

By binding to calmodulin, CPZ could alter spectrin and/or
Ca2+-Mg2+ ATPase conformation. If CPZ alters spectrin structure it
would modulate the entire cyto-skeletal matrix. If CPZ binds to the
calmodulin controlling Ca2+-Mg2+ ATPase, it would modulate an
important intrinsic membrane protein. At micromolar concentrations,
addition of CPZ induces vast shape changes. This is indicatiave of
vast changes in the cyte-skeletal matrix. As the CPZ concentration in
increased, the binding sites would become saturated and CPZ would begin
to act like a detergent. CPZ displays a biphasic dose dependent curve
of its affects on hypotonic hemolysis (120, 121). As the concentration
of CPZ is increased above 23 uM (maximal protection from lysis), the
drug's calmodulin binding sites may become saturated, and the CPZ may
then enter the lipid matrix. If the dose of CPZ is further increased,
the hypotonic hemolysis studies show hemolysis greater than 100% (121).
This indicates that at high concentrations, CPZ is detergent like,'
since it solubilizes the membrane bound hemoglobin.

In conclusion, at low concentrations CPZ appears to interact with

or affect the protein cytoskeletal matrix to alter erythrocyte shape.

20

As the concentration of CPZ is increased, it becomes detergent-like,

and finally lytic.

H. Objectives

In this thesis presentation, I show that several compounds affect
erythrocyte membrane structure. These changes were observed by two
different techniques. The major technique used was electron spin
resonance (ESR) spectroscopy. I used the fatty acid spin probe 5-doxyl
stearate. This technique (ESR) gives information on the spin probe
mobility and the general environment around the free electron. The
other technique was hypotonic hemolysis. Hypotonic hemolysis is one
procedure used to determine the fragility of erythrocytes. Using
these two techniques, I looked at how various compounds affected
erythrocyte membrane structure on a molecular level (ESR studies) and
on the level of the whole cell (hypotonic hemolysis studies).

The concentration of perturbant needed to induce membrane struc-
tural changes range from picomolar levels for PGE1 and PGE2 up to
the millimolar range for the anionic perturbants such as HEPES and
phosphate. In this thesis presentation I wish to show that alteration
of even "minor" components of the membrane can induce an overall
membrane structural change.

In conclusion, I would like to advance the theory that changes in
membrane lipid order can be instigated by structural changes in the
cytoskeletal matrix. Also, I believe the converse is true; that
changing the conformation of one of the components of the cytoskeletal
matrix will not only affect the rest of the matrix (see fig. 1), but

also the overlying membrane.

10.
11.
12.
13.
14.
15.

16.
17.

BIBLIOGRAPHY

Nasse, H. (1878) Pfugers Arch. Ges. Physiol. Menschen. 16:604.
Dirkin, M.N.J., and Mack H.H. (1931) J. Physiol. (Lond.) 73:349.
Tosteson, D.C. (1959) Acta Physiol. Scand. 46:19.

Brahm, J. (1977) J. Gen. Physiol. 70:283.

Gunn, R.B., Dalmark, M., Tosteson, D.C. and Hieth, J.O. (1973) J.
Gen. Physiol. 61:185.

Funder, J., and Neith, J.0. (1976) J. Physiol. (Lond.) 262:679.

Ibrahim, S.A. and Thompson, R.H.S. (1965) Biochim. Biophys. Acta
99:331.

Gul, S. and Smith, A.D. (1972) Biochim. Biophys. Acta. 288:237.

Low, M.G., Limbrick, A.R., and Finean, J.B. (1973) FEBS Letts.
34:1.

Colley, C.M. Zwaal, R.F.A., Roelofsen, B., and Van Deenan, L.L.M.
(1973) Biochim. Biophys. Acta. 307:74.

Zwaal, R.F.A., Roelofsen, 8., and Colley, C.M. (1973) Biochim.
Biophys. Acta 300:159.

Gordesky, S.E., Marinetti, G.V. and Segel, C.B. (1973) J. Memb.
Biol. 14:229.

Gordesky, S.E., Marinetti, G.V., and Love, R. (1975) J. Memb.
Biol. 20:111.

Marinetti, G.V., Sheeley, 0.3., Baungarten, R., and Love, R.
(1974) Biochem. Biophys. Res. Commun. 59:502.

Sakagami, T., Minar, 0., and Orii, T. (1965) Biochim. Biophys.
Acta. 98:111.

Reed, C.F. (1968) J. Clin. Invest. 74:749.
Rothman, J.E., and Davidowicz, E.A. (1975) Biochemistry 14:2809.

21

18.

19.
20.
21.
22.

23.

24.
25.

26.
27.
28.

29.
30.

310

32.

33.
34.
35.

36.

37.
38.
39.

22

Hellings, J.A., Kamp, A.H., Hirtz, K.N., and Van Deenan, L.L.M.
(1974) Eur. J. Bioch. 47:601.

Nes, N.R. (1974) Lipids. 9:596.
Turner, J.D., and Rouser, G. (1974) Lipids. 9:49.
Murphy, J.R. (1962) J. Lab. Clin. Med. 60:86.

Bruckdorfore, K.R., Demel, R.A., de Gier, J., and Van Deenan,
L.L.M. (1969) Biochim. Biophys. Acta. 183:334.

Vanderkooi, J., Fischkoff, S., Chance, B., and Cooper R.A. (1974)
Biochemistry 13:1589.

Singer, S.J., and Nicolson, G.L. (1972) Science 175:720.

Fairbanks, G., Steck, T.L., and Nallach, D.F.H. (1971)
Biochemistry 10:2606.

Steck, T.L. and Dawson, G. (1974) J. Biol. Chem. 249:2135.
Gahmberg, C.G., Hakamori, S. (1973) J. Biol. Chem. 248:4311.

Bender, N.N., Garan, H., and Berg H.C. (1971) J. Mol. Biol.
58:783.

Triplett, R.B. and Carraway, K.L. (1972) Biochemistry 11:2897.

Ungewickell, E., Bennett, P.M., Calvert, R., Ohanian, V., and
Gratzes, R.B. (1979) Nature 280:811.

Shotten, D.M., Burke, B.E., and Branton, D. (1979) J. Mol. Biol.
131:303.

Morrow, J.S., Speicher, D.M., Knowles, W.J., Hsu, C.J. and
Marchesi, V.T. (1980) Proc. Natl. Acad. Sci. USA 77:6592.

Bennett, V. and Stenbuck, P.J. (1979) Nature 280:468.
Bennett, V. and Stenbuck, P.J. (1980) J. Biol. Chem. 255:6424.

Hargreaves, H.R., Giedd, K.N., Verkleij, A., and Branton, D.
(1980) J. Biol. Chem. 255:11965.

Cohen, C.M. and Korsgren, C. (1980) Biochem. Biophys. Res. Commun.
97:1429. ‘

Fowler, V. and Taylor, D.L. (1980) J. Cell Biol. 85:361.
Tanner, M.J.A., and Grey, H.R. (1971) Biochem. J. 125:1109.
Carraway, K.L. and Shin, B.C. (1972) J. Biol. Chem. 274:2102.

40.
41.
42.
43.
44.

45.
46.

470
48.
49.

50.

51.

52.
53.

54.
55.

56.

57.
58.
59.
60.

61.

23

Kant, J.A., and Steck, T.L. (1973) J. Biol. Chem. 248:8457.
Steck, T.L. (1972) J. Mol. Biol. 166:295.

Tanner, M.J.A. and Boxer, D.H. (1972) Biochem. J. 129:333.
Jenkins, R.E. and Tanner, M.J.A. (1975) Biochem. J. 147:393.

Marton, L.S.E. and Garvin, J.E. (1973) Biochem. Biophys. Res.
Commun. 42:1457.

Nicolson, G.L., and Painter, R.G. (1973) J. Cell Biol. 59:395.

Lepke, S. Fasold, H., Pring, M., and Passaw, H. (1976) J. Membr.
Biol. 29:147.

Cabantchik, Z.I. and Rothstein, A. (1974) J. Membr. Biol. 15:207.
Cabantchik, Z.I. and Rothstein, A. (1974) J. Membr. Biol. 15:227.

Zaki, L., Fasold, H., Schumann, B., and Passow, H. (1975) J. Cell
Physiol. 86:471.

Kaplan, J.H., Serrah, K., Fasold, H., and Passow, H. (1976)
FEBS Letts. 62:182.

Grinstein, S., McCulloch, L. and Rothstein, A. (1979) J. Gen.
Physiol. 73:493.

Gunn, R.B. and Frolich, 0. (1978) J. Supramol. Struct. 2:213.

Sachs, J.R., Knauf, P.A., and Dunham, P.B. (1975) D.M. Surgenor,
ed. 2nd edition 2:613 Ac. Press NY.

Hoffman, J.M. and Laris, P.C. (1974) J. Physiol. (Lond.) 239:519.

Yu, J., Fischman, D.A. and Steck, T.L. (1973) J. Supramol. Struct.
1:233.

Steck, T.L., Ramos, B., and Strapazon, E. (1976) Biochemistry
15:1154.

Nigg, E. and Cherry, R.J. (1979) Nature 277:493.
Steck, T.L. (1974) Methods in Membrane Biochemistry
Barzilay, M. and Cabantchik, Z.I. (1979) Membr. Biochem. 2:255.

Barzilay, M., Ship, 5., and Cabantchik, Z.I. (1979) Membr.
Biochem. 2:227.

Jennings, M.L. and Passow, H. (1979) Biochim. Biophys. Acta.
554:498.

62.
63.
64.
65.

66.

67.

68.
69.

70.

71.

72.

73.

74.
75.
76.
77.

78.

79.
80.

81.

82.

24

Knauf, P.A. (1979) Current Topics in Membrane Transport 12:249.
Gunn, R.B. (1979) Membrane Transport in Biology 2:59.
Hong, 5.1. and Levine, L. (1976) J. Biol. Chem. 251:5814.

Bonney, R.J., Naruns, P., Davies, P., and Humes J.L. (1979)
18:605.

Lapetina, E.G., and Cuatrecasas, P. (1979) Proc. Nat'l. Acad. Sci.
USA 76:121.

Plummer, N.A., Henshy, C.N., Camp, R.D., and Warin, A.P. (1977) J.
Invest. Dermatol. 68:246.

Ohuchi, K. and Levine, L. (1978) Prostaglandins Med. 1:421.

Flower, R.J., and Blackwell, G.J. (1976) Biochem. Pharmacol.
25:285.

Roth, G.J., Stanford, N., and Majercis, P.W. (1975) Proc. Nat'l.
Acad. Sci. USA 72:3073.

Hamberg, M., Svensson, J., and Samuelsson, B. (1975) Proc. Nat'l.
Acad. Sci. USA 72:2994.

Moncada, S., Gryglewski, R., Bunting, 5., and Vane, J.R. (1976)
Nature 263:663.

Hamberg, M. and Samuelsson, B. (1974) Proc. Nat'l. Acad. Sci. USA
71:3400.

Goetzl, E.J. and Sun, F.F. (1979) J. Exp. Med. 1502406.
Smith, J.B. and Willis, A.L. (1971) Nature 231:235.
Flower, R.J. and Blackwell, G.J. (1979) Nature 278:456.

Smigel, M. and Fleischer, S. (1974) Biochim. Biophys. Acta.
332:358.

Crankshaw, D.J., Crankshaw, J., Branda, L.A., and Daniel, E.E.
(1979) Arch. Biochem. Biophys. 198:70.

Lord, J.T. and Ziboh, V.A. (1979) J. Invest. Dermatol. 73:373.

Lord, J.T., Ziboh, V.A., and Warren, S.K. (1978) Endocrinology
102:1300.

Rao, Ch.V. and Harker, C.W. (1978) Biochem. Biophys. Res. Commun.
85:1054.

Allen, J.E. and Rasmassen, H. (1971) Science 174:572.

83.

84.

85.

86.
87.

88.

89.

90.

91.

92.

93.
94.
95.

96.

97.

98.

99.
100.

101.
102.

25

Kury, P.G., Ramwell, P.W. and McConnell, H.M. (1974) Biochem.
Biophys. Res. Commun. 56:478.

Taniguchi, M., Aikawa, M., and Sakagami, T. (1982) J. Biochem.
91:1173.

Lefkowitz, R.J. Mullikin, D., Wood, C.L., Gore, T.B., and
Mulcheizec, C. (1977) J. Biol. Chem. 252:5295.

Miller, O.V. and Gorman, R.R. (1979) J. Pharm. Exp. 210:134.

Schafer, A.I., Cooper, B., O'Hara, D. and Handing, R.I. (1979) J.
Biol. Chem. 254:2914.

Davies, P.J., Chaby, R., Zech, L., Berman, M., Pastan, I. (1980)
Biochim. BiOphys. Acta. 629:282.

Meyers, M.B. and Swislocki, N.I. (1974) Arch. Bioch. Biophy.
164:544. ‘

Rasmussen, H. and Lake, W. (1975) Prostaglandins in Hematology,
187.

Worthington, R.E., and Nakeff, A. (1982) Prostaglandins 23:841.
Dejana, E., Cerletti, C., de Castellarnaw, C., Livio, M., Galetti,
F., Latine, R., and de Gaetano, G. (1981) J. Clin. Invest.
68:1108.

Green, F.A. and Jung, C.Y. (1981) Transfusion 21:55.

Hunt, J.N. and Franz, D.R. (1981) Dig. Dis. Sci. 26:301.

Kelton, J.G., Carter, C.J., Rosenfeld, J., Masicotte-Nolan, and
Hirsh, J. (1981) Thromb. Res. 24:163.

Preston, F.E., Whipps, S., Jackson, C.A., French, A.J., Wyld,
P.J., and Stoddard, C.J. (1981) N. Eng. J. Med. 304:76.

Gupta, K.C., Joshi, J.V., Hazari, K., Pohijani, S.M., and
Satosokar, R.S. (1982) Int. J. Clin. Pharm. 20:511.

Chang-Chang, V., Nakao, J., Neichi, T., Orimo, H. and Murota,
S.-I. (1981) Biochim. Biophys. Acta. 664:291.

Moncada, S., and Vane, J.R. (1979) Fed. Proc. 38:66.

Sheetz, M.P. and Singer, S.J. (1975) Proc. Nat'l. Acad. Sci. USA
71:4457.

Elferink, J.G.R. (1977) Biochem. Pharm. 26:2411.

Tenforde, T.S., Yee, J.P. and Mel, H.C. (1978) Biochim. Biophys.
Acta. 511:152.

103.

104.

105.

106.
107.

108.

109.

110.

111.
112.

113.

114.

115.

116.

117.

118.
119.

120.
121.

26

DiFrancesco, C. and Bickel, M.V. (1977) Chem. Biol. Interactions.
16:335.

Gazitt, Y., Loyter, A., and Ohad, I. (1977) Biochim. Biophys.
Acta 471:361.

Dachary-Prigent, J., Dufourq, I., Lussan, C., and Boisseau, M.
(1979) Thrombosis Res. 14:15.

Romers, J., and Bickel, M.H. (1979) Biochem. Pharmac. 28:799.

09150, T., Iwaki, M., and Mori, K. (1981) Biochim. Biophys. Acta
649:325.

Tenforde, T.S., Yee, J.P. and Mel, H.C. (1978) Biochem. Biophys.
Acta 511:152.

Suda, T., Shimizu, D., Maeda, N., and Shiga, T. (1981) Biochem.
Pharmac. 30:2057.

Suda, T., Maeda, N., Shimizu, 0., Kamitsubo, E., and Shiga, T.
(1982) Biorheology 19:555.

Roufgalis, 8.0. (1981) Biochem. Biophys. Res. Commun. 98:607.

Kobayashi, R., Tawata, M., and Hidaka, H. (1979) Biochem. Biophys.
Res. Commun. 88:1037.

Bond, G.H., and Clough, D.L. (1973) Biochim. Biophys. Acta
323:592.

Farrance, M.L. and Vincenzi, F.F. (1977) Biochim. Biophys. Acta.
471:59.

Graf, E., Verma, A.K., Gorski, J.P., Lopaschuk, G., Niggli, V.,
Zurina, M., Carafoli, E., and Penniston, J.T. (1982) Biochemistry
21:4511.

Waisnan, D.M., Gimble, J.M., Goodman, D.B.P., and Rasmussen, H.
(1981) J. Biol. Chem. 256:409.

Andreason, T.J., Keller, G.H., LaPorte, D.C., Edelman, A.H., and
Storm, D.R. (1981) Proc. Natl. Acad. Sci. USA 78:2782.

Hinds, T.R. and Anreason, T.J. (1981) J. Biol. Chem. 256:7877.

09150, T., Iwaki, M., and Sugiura, M. (1980) Chem. Pharm. Bull.
28:3283.

Olaisen, B., and de, I. (1973) Eur. J. Pharm. 22:112.
Mazorow, D.L., Haug, A., McGroarty, E.J. (1983) Chapter 5:

Chlorpromazine Induced Changes in Human Erythrocyte Membrane
Structure.

CHAPTER II

Title: Alterations in the Human Erythrocyte Membrane Induced by

Anions. I. Anions Affect Membrane Physical Structure.

Authors: Denise L. Mazorow, Alfred Haug, and Estelle J. McGroarty

27

28

ABSTRACT

Fresh whole human erythrocytes were used to study the effects of
ions on the erythrocyte membrane properties. Various anionic agents
affect the human erythrocyte membrane structure as determined using
electron spin resonance (ESR) probing techniques. Membrane order para-
meter was increased with certain anionic agents, the degree of pertur-
bation dependent on the anion and the anion concentration. Control
studies were performed in a physiological bicarbonate buffer.
N-2-hydroxyethyl piperazine-N-Z-ethane sulfonic acid (HEPES) was found
to be the most perturbing of the anions used with phosphate less per-
trubing and sulfate least perturbing. Our results suggest that anions
may effect their changes of membrane order by partially or totally
blocking the anion port. To substantiate this, the anion port blocking
agents 4,4'-diisothiocyano-2,2'—disulfonic acid stilbene and 4-acet-
amido—4'-isothiocyano-2,2'-stilbene disulfonic acid were used. Both
blocking agents affected membrane phase transition temperature. Block-
ing the anion port may be indirectly inducing a change in membrane
structure. The calcium ionophore A23187 was shown to increase membrane
order. This ionophore allows Ca2+ ion influx into the erythrocyte
which presumably induces cross-linking of the phosphatidylserine head
groups and in turn increases membrane order. Addition of dimethylsulf-
oxide induced a large increase in order parameter. Ouabain, a Na-K

ATPase inhibitor decreased the phase transition. Thus, we found that

29

those compounds which affect membrane protein structure appear to

affect membrane lipid structure as well.

30

INTRODUCTION

One of the critical functions of the erythrocyte is to facilitate
02/C02 exchange in the periphery. The passive exchange of anions,
most specifically chloride and bicarbonate across the erythrocyte
plasma membrane is believed to occur via a transmembrane protein known
as band 3 or the anion channel (1). Band 3 represents more than 25% of
the total membrane protein and has a molecular weight between 88,000
and 105,000 daltons (3). Within the membrane this protein apparently
forms a dimer (4). There are five sulfhydryl groups in band 3 which
can be labelled by N-ethylmaleimide (5). The C terminal end, 35,000
daltons by chymotryptic cleavage, is mostly located outside the cell
and has two of the sulfhydryl groups (6). A 9,000 dalton segment of
this 35,000 dalton piece traverses the membrane (6). It is also the
segment which contains the two sulfhydryl groups. The other three
sulfhydryl groups are located in the segment situated on the cyto-
plasmic side of the membrane (5, 7).

Band 3 is reportedly involved in anion transport. Several anion
transport inhibitors including the stilbene derivatives 4-acetamido~
4'-diisothiocyano-2,2'-disulfonic acid stilbene (SITS) and 4,4'-diiso-
thiocyano-2,2'-disulfonic acid stilbene (DIDS) bind to sulfhydryl-
groups of band 3 (8-10). Changes in band 3 structure can be monitored
with DIDS fluorescence (11), and changes in anion flux can be monitored

using 355042- and 3501- (12,13). 0105 binding to band 3

31

has been shown to alter the amount of hemoglobin bound to band 3 (l4)
and to alter band 3 structure (15).

Recent work on the mechanism of anion transport across the erythro-
cyte membrane via band 3 suggests that influx and efflux of anions are
separate events (16), and a ping-pong model best describes the mecha-
nism of anion transport (17, 18). It has been suggested that the
individual units of the band 3 dimer transport anions independently
(19) with the primary event being anion efflux (19).

In this study we show that changing the anionic milieu from that
found in situ alters the erythrocyte membrane structure. Membrane
structural changes were induced by changes in the buffer or by inhibit-

ing ion transport.

32

MATERIALS AND METHODS

Freshly drawn blood was obtained by venapuncture from at least two
healthy individuals for each run. Only non-smokers who had taken no
alcohol or medication for at least 48 hours prior to venapuncture were
used as donors. Anticoagulation was effected by defibrination. Cells
were centrifuged and the buffy coat was removed using standard proce-
dures. The erythrocytes were washed three times in 300 millosmols
(mOSm) NaCl pH 7.4, and packed to a hematocrit of 70 t l. A 1 ml
aliquot of the packed cells was placed into a test tube in which 6 to 8
microliters of a 30 m4 ethanolic solution of 5-doxyl stearate (5-DS)
had been evaporated. The preparation was gently mixed for 2 minutes.
The 5-DS labelled cells were added to 40 ml of the salt solution being
tested and incubated for 10 minutes at 37°C. The salt solutions used
were: a modified Earles' salt solution containing 145 "M NaCl, 5 "M
KCl, 0.7 mM M9504, 1 MM NaH2P04, 2 mM CaClz, and 22 mM NaHC03
at pH 7.4; a phosphate buffer (pH 7.0 and 7.4) containing 145 nM NaCl,
5 mM KCl, 0.7 mM M9504, 1 nM CaClz, 5.5 mM glucose, 3.5 mM NaH2P04 and
1.5 mM NazHPO4; a phosphate buffer with 0.5 mM carbonic acid (PO4/HC03
buffer), pH 7.0 and pH 7.4; a sulfate buffer containing Earles' salts
plus 4 mM sulfate; a N-2-hydroxyethyl piperazine-N-Z-ethane sulfoniC‘
acid (HEPES) buffer containing Earles' salts without NaHC03 and 30 mM
HEPES; an Earles' salts solution with 15 mM HEPES and 11 0M NaHC03
(HEPES/HC03 buffer); filtered serum (proteins larger than 10,000

33

daltons were removed with ultrafiltration using an amicon filter UM 10)
pH 7.4; and an unbuffered 300 mOSm NaCl pH 7.4. All cells were
incubated for 10 minutes at 37°C in the presence of perturbants in the
modified Earles' salt solution. The cells were packed to a hematocrit
of 70 t l and were placed in a quartz cuvette for electron spin
resonance (ESR) analysis. All ESR experiments were carried out with a
Varian century E line ESR spectrometer. An external calibrated
thermistor probe was used to monitor the temperature dependent spectral
parameters over the range of 20°C to 48°C. All temperature dependent

ESR parameters were shown to change reversibly in this range.

Hypotonic Hemolysis.

 

Hypotonic hemolysis studies were performed in both a ph05phate
buffer and an Earles' salt buffer under a variety of conditions. The
solutions were prepared by diluting an isotonic solution (300 mOSm)
with double distilled water to the desired hypotonicities. The solu-
tions studied were: phosphate buffer, pH 7.4, 155 mOSm; Earles' salts
without HC03', 110 mOSm; Earles' salts with and without HCO3', 126 mOSm;
Earles' salts with HC03', 150 mOSm; phosphate buffer without calciun
165 mOSm; and Earles' salts without calcium, 160 mOSm. The osmolarity
of the solutions was determined using an Advanced Instruments osmoter
model 3R. A 10% (v/v) solution of erythrocytes was placed in 5 ml of
hypotonic buffer such that the final concentration of cells was 0.1%.
These cells were incubated for seven minutes in hypotonic buffer and‘
returned to isotonicity by addition of the necessary amount of 4 M

NaCl. The cells were centrifuged and the degree of lysis was

34

determined by the amount of absorbance at 543 nm in the supernatant
solution. 100% lysis was determined by placing the cells in double
distilled water and 0% lysis was determined by reading the degree of
lysis for cells placed in a 300 mOSm solution. Percentage lysis was

taken as the difference between these two values.

ESR Spectral Analysis.

In order to determine the relative fluidity of the erythrocyte mem-
brane in the different buffer systems, the membrane of the intact cell
was labelled with 5-DS. 5-DS incorporates into the lipid phase of the
membrane in such a way that the unpaired electron of the nitroxide rad-
ical is situated close to but shielded from the aqueous phase. Spin
labelled preparations were placed in the resonance cavity of an ESR
spectrometer. In this fashion the erythrocyte membrane preparation was
exposed to microwaves of a constant frequency while held in a linearly
varying magnetic field corresponding to the Spin states of the unpaired
electron. The spectra obtained are displayed as the first derivative
of the microwave absorption plotted in gauss against an increasing
magnetic field. The distance betwen the low field and high field
absorption peaks, the hyperfine splitting parameter (2T11), is a
measure of the largest energy difference between the spin states of the
unpaired electron. With increasing temperatures, intermolecular influ-
ences upon the unpaired electron tend to average, resulting in smaller
values of 2T]; which reflect more fluid environments. 2T1],
therefore is related to the viscosity of the environment for which the
probe is reporting. Large values of 2T1] indicate more fluid

environments.

35

The order parameter, S, measures the deviation of the observed
signal from that of a uniform orientation of the probe. For a uniform-
ly oriented sample S=l; for a random sample S=O. Order parameter
changes of less than 1% are considered to be biologically insignificant

(20, 21).

Computer Analysis.

The data were analyzed by an iterative least squares program (22).
Brifely, a B-spline (23) was used to provide a smooth fit for the ESR
data and points of inflection were used to group data. Regression
lines were calculated for each group and then plotted. This analysis
allowed the determination of break points. Such breaks in the tempera-
ture dependence of 2T1] and S have been correlated with lipid phase
separations or lipid phase transitions from gel to liquid crystalline

lipid states (24).

36

RESULTS

ESR spectra from cells labelled with 5-05 showed little free probe
signal (Figure 1). Spectral symmetry of the low field peak indicates
that the majority of the spin label was located in a single uniform
environment over the temperature range examined (20°C-48°C). The
hyperfine splitting parameter, 2T1], decreased linearly as a func-
tion of increasing temperature (Figure 2a, 2b) with a discontinuity
occurring between 23°C and 39°C depending on the buffer in which the
cells were suspended. Spectra recorded above 12°C penmitted the deter-
mination of the order paramter S. When S was plotted as a function of
temperature, a break similar to that observed with 2T11 was detec-
ted (Figure 3). The transition temperatures measured using S and
21]] were nearly the same and may indicate protein-lipid structural
changes in the erythrocyte membrane.

Our results indicated first, that the presence of specific anions
alter the interactions in the human erythrocyte membrane. Secondly, we
found that the ionic distribution across the erythrocyte membrane
altered lipid-lipid and lipid-protein interactions. Finally we found
that different buffer systems affect the fragility of the erythrocyte

as seen by hypotonic hemolysis.

37

Figure 1. Electron Spin resonance spectra of human erythrocytes
labelled with S-doxyl stearate. The spectra were taken at the
temperatures indicated. Absence of free probe is indicated by the lack
of a major absorption signal at points indicated by arrows. Symmetry
of high and low field peaks reveals that the probe is in a single
environment.

Figure 2a. The temperature dependent change in the hyperfine splitting
parameter (2Tu) of erythrocytes labelled with 5-doxyl stearate and
suspended in Earles salts.

Figure 2b. The temperature dependent change in the hyperfine splitting
parameter (2TH) of erythrocytes labelled with 5-doxyl stearate and
suspended in Earles salts containing 10"7 M A23187.

40

On

.on

8L “gs—wk

ON

 

 

q

$803
.....N

 

 

41

Gov 02w...

ON

 

 

@863
EN

 

 

42

 

76

72

s 1%)“

64

 

 

 

l l l
20 30 40

TEMP (’0)

 

Figure 3. The temperature dependent change in the order parameter (S)
of erythrocytes labelled with 5 -doxyl stearate and suspended in Earles
salts (0—0) or in Earl es salts with 10-7 M A23187(D—D).

43

Anion Perturbation.

 

Filtered serun, which is close to whole serum in ionic composition
was considered as a control solution for resuspending cells. Whole
serum was not used since the serum albunin sequestered most of the spin
probe. Any variations in cell structure from that of cells in filtered
serum were considered to be a perturbation of the erythrocyte membrane.
Since the structure of cells in modified Earles' salts was essentially
indistinguishable from that of cells in the filtered serum (Table l),
the Earles' salts was used as a standard salt solution. The membrane
phase transition of cells in Earle's salts as observed by 2T1] vs.
temp and S vs. temp was at 37°C (Figure 2A, 3). Cells in Earle's salts
and the Earles' salts with glucose were also essentially identical in
their physical properties (Table l). The phase transition temperature
of cells in isotonic NaCl was reduced by several degrees (Table l).
Phosphate buffer at pH 7.0 and pH 7.4 in the presence of bicarbonate
(approx. 0.5 mM) and at pH 7.0 in the absence of calcium induced
Inenbrane perturbation. The phosphate buffer at pH 7.0 with or without
calciun perturbed the erythrocyte membrane. Phosphate buffer, pH 7.0,
slightly ordered the membrane and this was about the same change as
seen in phOSphate buffer at pH 7.4 (Table 1, Figure 4a, 4b). the
presence of bicarbonate in the solution caused significant ordering of
the membrane relative to phosphate buffer alone at pH 7.0 but
disordered slightly at pH 7.4 (Figure 4a, 4b).

30 mM HEPES also significantly ordered the erythrocyte membrane at
room temperature (24°C) and at physiological temperature (37°C) as did
a solution of 15 mM HEPES and 11 mM bicarbonate in Earles' salts. An

44

Table 1

Effects of buffers on erythrocyte membrane phase transitions

 

Phase transition temperature determined using

 

 

Suspending Buffer 2T1] S Average
filtered serwn 37.08 37.4 37.24
NaCl (unbuffered) 33.11 34.38 33.45
Earles' modified salts 38.39 35.77 37.08
Earles w/glucose 35.8 36.22 36.01
P04 pH 7.0 35.0 38.8 36.9
P04 pH 7.4 37.2 36.2 36.7
P04/HC03 pH 7.0 30.40 31.2 30.8
PO4/HCO3 pH 7.4 29.25 30.38 29.82
504 in Earles 33.33 31.24 32.29
HEPES 33.61 32.50 33.06
HEPES/HC03- 31.72 30.53 31.13
Table 2

Effects of perturtants in modified Earle's salts on erythrocyte
membrane phase transitions

 

Phase transition temperature determined using

 

Perturbant 2T1] 5 Average
modified Earles' salt 38.39 35.77 37.08
Ouabain 10-5 M 29.24 32.97 31.11
Ouabain 10-6 M 33.26 37.5 35.38
A23187 10-7 M 32.38 32.15 32.27
A23187 10-8 M 27.45 24.33 25.87
423187 10-9 M 37.88 23.15 30.52
.028 M DMSO 37.27 37.47 37.37
DIDS 10-5 M No No No
transition transition transition

511$ 10-5 M 35.78 35.78 35.78

 

45

Figure 4. Change in order parameter of cell labelled with 5-doxyl
stearate and suspended in various buffers compared to cells suspended
in filtered serum. Membrane order was measured at 24°C (a) and 37°C

0))-

46

.3: v.2... 2. as 3. .... 8...; 2.2.. 38.... ea:
Bub: $8: .8 nouxxoanouix cos .9. .9. .53 8:8 as: 3......

0.01

 

0.0

$m<

0.0

 

0..

 

47

A fit ME:
8: o: no:
ammo: flaw: .8 on. 0.. «on. .8

 

V1.00 ON

$4... .85} Ed 88%

838

.2 team 8.3 .82 85.:

..8-

 

1———1

 

 

l——-l

:4
O

o\om<

 

48

increase of sulfate ions from 0.7 mM to 4 mM in the Earles' salts solu-
tion caused minimal changes in membrane order (Figure 4a, 4b).

The phase transition of membranes from cells in Earles' salts,
Earles' salts with glucose, and phosphate buffer pH 7.0 and pH 7.4 were
not significantly different from that of cells in filtered serum. The
phase transition temperature was lower compared to control when cells
were in unbuffered NaCl, 804 in Earles, HEPES buffers, and the phase
transition temperature was further reduced when bicarbonate was added

to either the phosphate buffer or the HEPES buffer (Table 1).

Ion Gradient Perturbation.

The perturbants ouabain, 4-acetamido-4'-isothiocyano-2,2'-disu1fon-
ic acid stilbene (SITS), 4,4'-diisothiocyano-2,2'-disulfonic acid
stilbene (DIDS), A23187, and dimethylsulfoxide (DMSO) (Table 2, Figure
5a, 5b) directly or indirectly alter the ion environment or membrane
structure of the erythrocyte. Ouabain inhibits the NaT/K+ ATPase
which in turn causes an alteration in internal to external ratios of
Mai/K+ levels. Both DIDS and SITS bind to and inhibit the erythro-
cyte anion port and the Ca2+ ionophore A23187 allows Ca2+ to
enter the erythrocyte. DMSO is known to solubilize membrane protein
components (25).

Ouabain at 10'5 caused slight changes in membrane order at both
24°C and 37°C and a very dramatic shift in the phase transition temper-
ature from 37.2 to 31.1°C. The changes in lipid order induced with
10"6 M ouabain were small and the phase transition temperature was

less affected at the lower temperature.

.Aav 8.4m use Any uoem pa vatsmaas was

cacao wagonsmz .mcopo mupmm mm_cmm cw vmucmamam m__mu cow: uocaaeou
mucencaucma acmcmww_v m:.c.mucou mu_mm mmpcem cw umucmamzm new mumcmmum
pxxouim gu_3 um_Pmnm_ mppmu we cwuosmcmq gouge c. «mango .m «Lamp;

:92 2...: 83.. 3.2 3.9 3.2 a .9 :92
at... 38 .33. 333 58¢ 53¢, 2226 £2.26.

 

 
 

49

ossomsa

 

 

.WN

 

50

[CL
AS’lo

   

 

-oS'-
Ouabain A23|87 A2337 DIDS
IG’M IO'7M IO’M IO‘M

Ouabain A23I87 new srrs
IG‘M Io‘u ouso 05M

51

When the calciun ionophore A23187 was added to erythrocytes at
final concentrations of 10'7 M, 10'3 M, and l0"9 M,
membrane order increased dramatically. At 24°C, A23187 at 10-7 M
induced an increase in membrane order which was greater than 2% (Figure
3). At 37°C, the increase in order was l.7% (Figure 3). The degree of
perturbation induced by A23187 decreased with decreasing concentration.
The shift in phase transition temperature induced by A23l87 was dose
dependent, with the maximal decrease occurring at 10'8 M. However,
at 10'7 M (Fig. 28) and 10'9 M, A23187 still significantly decreased
the phase transition temperature (32.3°C and 30.6°C respectively). The
increase in order induced by 10-7 M A23187 was 0.89% (data not
shown) at 24°C in the absence of Ca2+.

DIDS and SITS are both anion port blocking agents although DIDS is
more specific for anion port binding. The level of DIDS and SITS added
was sufficient to block all available anion port sites in the cell
sample. Although DIDS and SITS at 10'5 M did not induce any appre-
ciable changes in membrane transition temperature, SITS at l0'5 M
did induce a slight increase in membrane order at 24°C. DIDS did not
induce a change in the phase transition temperature, but caused an
elimination of the phase transition. SITS, in contrast, induced a
slight decrease in the phase transition temperature.

Dramatic increases in order parameter of the probe were seen upon
the addition of DMSO to the erythrocyte suspension, but no changes in

phase transition temperature were seen.

52

Hypotonic Hemolysis Studies.

 

The degree of lysis depended on the buffer and ions present. The
presence of bicarbonate in the Earles' salts solution increased the
erythrocytes' fragility. In the absence of bicarbonate 50% lysis
occurred at llO mOsm (pH 8.2) whereas in the presence of bicarbonate
50% lysis occurred at 150 mOsm (pH 7.4). Eleven percent of the cells
suspended in 126 mOsm Earles' salts solution without bicarbonate lysed
while 93% lysed in the same solution containing bicarbonate. In the
phosphate buffer, pH 7.6, 50% lysis was achieved with l55 mOsm phos-
phate. When calcium was deleted from the buffers, the osmolarity of
the buffers needed to be raised by 10 mOsm to maintain 50% lysis, i.e.,
cells in phosphate without calciun were 50% lysed at 165 mOsm and in
Earles' salts (with bicarbonate) without calcium were 50% lysed at l55

mOsm.

53

DISCUSSION

Membrane integrity is vital in normal cell function. We have evi-
dence which suggests that changing the anion environment affects the
binding properties of Prostaglandins E] and E2 (26). We suggest
that agents which alter membrane structure may alter or disrupt the
proper hormone binding (26) and/or enzyme (27) activity within the
erythrocyte membrane. We have noted that both anions and membrane per-
turbants such as anesthetics (24) alter membrane structure and some of
these changes may result from alterations in the ionic environment of
the membrane.

Changes in membrane lipid order reflect changes in lipid domain
structure (27) while alterations in the temperature of the membrane
phase transition may reflect changes in membrane protein-lipid inter-
actions (24). Perturbations in ion transport across the membrane have
been shown to influence membrane order (28), membrane phase transition
temperature (29), and protein conformation (30). Ouabain, acting
extracellularly, (3l) is a specific inhibitor of the Na+/K+ ATPase,
and we have shown that this inhibitor induces a decrease in phase tran-
sition temperature without changing the membrane order. In the pres-
ence of ouabain the normal physiological ratio of K*1n/ Na+out
cannot be maintained due to passive flux (32). Since the Na+-K+
ATPase is closely associated with erythrocyte membrane protein band 3

(33) alterations in the anion transport in the presence of ouabain may

54

occur which are associated with the structure or activity of the

NaT-K+ ATPase. The shift in phase transition induced by the pres-

ence of ouabain may be due to protein conformational changes of the
anion port-Na+-K+ ATPase complex. Conversely, the shift may result
from lipid-lipid interactions induced by the inhibitor which occur as a
result of ion inbalance. A similar shift in phase transition tempera-
ture was seen with the addition of the anion port blocking agent SITS.
SITS induced a small decrease in the phase transition temperature while
causing a small change in membrane order. The blocking agent DIDS
induced no appreciable change in membrane order, but its alteration of
the membrane phase transition was most unusual: no transition was
observed. In all of these cases, upon the addition of specific inhibi-
tors, membrane ion transport was altered and a change in the phase
transition occurred. These alterations in phase transition temperature
may reflect conformational changes in the protein-lipid configuration
which regulates the ionic homeostasis.

The calcium ionophore A23187, when added in the presence of calcium
induced marked alterations in both the phase transition temperature and
in membrane order. Since the increase in membrane order was dependent
on ionophore concentration, it is presumable that higher concentrations
of ionophore allowed more calciun to enter the erythrocyte. The order-
ing could be due to Caz+ induced cross-linking of phosphatidyl-
serine head groups (34), Caz+ cross-linking of proteins forming
heteropolymers (35), or Ca2+ induced cross-linking of band 3 with
spectrin (36). Membrane ordering due to high levels of calcium could
also be due to changes in phospholipid phosphorylation (37). The de-

crease in phase transition temperature in the presence of the ionophore

55

was also dose dependent. This alteration may reflect restructuring of
intrinsic membrane proteins (38).

High concentrations of anions are thought to disrupt the charge
distribution across the erythrocyte membrane (39). This disruption
could occur in two ways. First anions may alter either the lipids
within the inner or outer monolayer of the cell (much as calcium per-
turbs the inner monolayer of the erythrocyte); or secondly, large
anions may "block" the anion port and thus alter the transmembrane
distribution of anions. Chloride ions traverse the membrane several
orders of magnitude faster (40) than either sulfate or phosphate ions;
whereas, phosphate and sulfate have similar rates of flux (40). At
physiological temperatures chloride transport is relatively independent
of pH but sulfate and phosphate flux depends on the pH (40). We have
shown that membrane structural changes occur in the erythrocyte in
phosphate buffer when the pH is changed from pH 7.0 to pH 7.4. In
phosphate buffer the membrane was less ordered at the lower pH.
Changes in anion transport rates have been shown to occur when specific
anions are raised above the physiological concentration. Examples in-
clude chloride inhibition of phosphate (41) and sulfate transport (40,
42); sulfate inhibition of phosphate (43, 44) and chloride transport
(40); chloride and bicarbonate inhibition of chloride transport (44,
45) and phosphate inhibition of sulfate and chloride exchange (46).
The buffer HEPES, also an anion, is known to decrease the stability of
red cell ghosts relative to phosphate buffer (47). This decrease in
stability may be due to alterations of the transmembrane anionic envi-

ronment. In our studies membranes of erythrocytes in HEPES, were more

56

ordered and had a lowered phase transition temperature compared to
cells in Earles' salts.

The decrease in phase transition temperature induced by anion per-
turbation across the membrane may be the result of alteration in the
lipid protein matrix caused by conformational changes in band 3 (30,
24). Although the phosphate buffer, pH 7.4, induced no change in
either membrane order or phase transition, phosphate in the presence of
bicarbonate disordered the membrane relative to membranes in phosphate,
at either pH 7.0 or pH 7.4. This alteration in membrane structure may
relate to the Ca2+.Mg2+ ATPase. It has recently been shown that H2P04
eliminates the 82 transition (29) of band 3, which involves the
cyt0plasmic portion of the protein, detected with microcalorimetry.
This transition is reported to be related to the anion port structure
and is associated with a cyt0plasmic protein (29). Anomalous changes
in band 3 conformation were also reported in the presence of high
phosphate by Ginsburg (30). We saw a reduction in both phase
transition and in lipid order when both phosphate and bicarbonate were
present. Phosphate alone may inactivate the CaZT-Mg2+ ATPase and the
presence of bicarbonate may reactivate the ATPase (48). The presence
of bicarbonate in HEPES buffer may induce the same changes.

Calcium may help to stabilize the erythrocyte membrane. If calcium
is present in the buffer calcium may leak in at the beginning of lysis
and help stabilize the membrane by cross-linking the membrane struc-
tural proteins (e.g. spectrin). He noted that bicarbonate destabilized
the membrane to hypotonic hemolysis. This may be an effect of pH, but
it is possible that destabilization is also related to the bicarbonate

effect on the CaZT-Mg2+ ATPase. Erythrocytes in phosphate buffer

57

buffer and Earles' salts with bicarbonate require approximately the
same osmolarity to achieve 50% lysis.

The erythrocyte membrane appears to exist in a state of flux. The
phase transition temperature was shown to occur at body temperature,
37°C. Thus, the membrane appears to exist in the metastable transition
state of gel + liquid crystalline to liquid crystalline (39). This
state may allow for enhanced transport across the membrane. In addi-
tion, our evidence suggests that bicarbonate destabilizes the red cell
membrane and may further facilitate an ion exchange across the
membrane.

These anions and perturbants all change the ionic milieu of the
erythrocyte membrane from physiological conditions. The induced
changes alter the overall conformation of the erythrocyte membrane
lipids and/or proteins. Alterations in erythrocyte membrane structure
may be amplified by changes in ion transport, in the hormone receptor

structure binding (26) and in membrane enzyme activity (50).

10.
11.
12.

13.

14.

15.

16.
17.

BIBLIOGRAPHY

Knauf, P.A. (l979) Curr. Top. Membr. Transp. 12, 249-363.

Cabantchik, Z.I., Knauf, P.A., and Rothstein, A. (1978) Biochim.
Biophys. Acta 515, 239-302.

Fiarbanks, G., Steck, T.L., and Hallach, D.F.H. (1971)
Biochemistry 10, 2606-2617.

Steck, T.L. (1974) J. Cell Biol. 62, 1-19.
Reithmeier, R.A.F. (1979) J. Biol. Chem. 254, 3054-3060.

Ramjeesingh, M., Garm, A., and Rothstein, A. (19810 J. Bioenerg.
and Biomembr. 13, 411-423.

Rao, A., and Reithmeier, R.A.F. (1979) J. Biol. Chem. 254,
6144-6150.

Snow, J.H., Brandts, J.F, and Low, P.S. (1978) Biochim. Biophys.
Acta 512, 579-591.

Frolich, 0. (1982) J. Membrane Biol. 65, 111-123.
Zaki, L. (1981) Bioch. Biophys. Res. Comm. 99, 243-251.
Davio, S.R., and Low, P.S. (1982) Biochemistry 21, 3585-3593.

Ku, C.P., Jennings, M.L., and Passow, H. (1979) Biochim. Biophys.
Acta 553, 132-141.

Knauf, P.A., Brever, H., McCulloch, L., and Rothstein, A. (1978)
J. Gen. Physiol. 72, 631-649.

Salhany, J.M., Cordes, K.A. and Gaines, E.D. (1980) Biochemistry
19, 1447-1454.

Passow, H., Fasold, H., Gartnes, E.M., Legrum, B., Ruffling, w.
and Zaki, L. (1980) Ann. N.Y.A.S. 341, 351-383.

Jennings, M.L. (1982) J. Gen. Physiol. 79, 169-185.

Rothstein, A., Ramjeesingh, M., Grinstein, S. and Knauf, P.A.
(1980) Ann. N.Y.A.S. 341, 433-443.

58

18.
19.
20.

21.

22.

23.
24.

25.

26.

27.

28.

29.

30.

31.
32.
33.
34.

35.
36.
37.
38.

59

Gunn, R.B. and Frolich, 0. (1980) Ann. N.Y.A.S. 341, 384-393.
Macara, 1.0., and Cantley, L.C. (1981) Biochemistry 20, 5095-5105.

Kury, P.G., Ramwell, P.M., and McConnell, H.M. (l974) Bioch.
Biophys. Res. Commun. 56, 478-483.

Huestis, M.H., and McConnell, H.M. (l974) Bioch. Biophys. Res.
Commun. 57, 726-732.

Brunder, D.C., Coughlin, R.T., and McGroarty, E.J. (1981) Comput.
Biol. Med. 2, 9-15.

Dierckx, P. (1975) J. Comp. Appl. Math. 1, 165-184.

Janoff, S.J., Mazorow, D.L., Coughlin, R.T., Bowdler, A.J., Haug,
A., and McGroarty, E.J. (1981) Am. J. Hematol. 10, 171-179.

Gunn, R.B., Kirk, R.G. (1976) J. Manb. Biol. 27, 265-282.

Mazorow D.L., Haug, A., McGroarty, E.J. (Chapter 3) Alterations in
the Human Erythrocyte Membrane induced by Anions. II. Anions
Affect Prostaglandin E] and E2 Membrane Receptors.

Sackmann, E., Trauble, H., Galla, H.J., and Overath, P. (1973)
Biochemistry 12, 5360-5369.

Hiedemer, T., DiFrancesco, C., and Brodbeck, U. (1979) Eur. j.
Biochem. 102, 59-64.

Snow, J.N., Vincentelli, J., and Brandts, J.F. (1981) Biochim.
Biophys. Acta 642, 418-428.

Ginsburg, H., O'Connor, S.E., and Grisham, C.M. (1981) Eur. J.
Biochem. 114, 533-538.

Schatzmann, H.J. (1953) Helv. Physiol. Pharm. Acta ll, 346-354.
Hieth, J.0. (1970) Acta. Physiol. Scand. 79, 76-87.
Bakker-Grunwald, T. (1981) Biochim. Biophys. Acta 641, 427-43l.
Papahadjopoulos, D. (1978) Cell Surface Reviews, Vol. 6, chapter

14. (Amsterdam: North-Holland Publishing Company) Poste G. and
Nicholson, G.L., eds.

 

Palek, J., and Liu, S.C. (1978) J. Supramolec. Struc. 10, 79-96.
Palek, J., Liu, P.A., and Liu, S.C. (1978) Nature 274, 505-507.
Allan, 0., and Michell, R.H. (1975) Nature 258, 348-349.

Motais, R., Baroein, A., and Baldy, S. (1981) J. Menbr. Biol. 62,
195-206.

39.

40.
41.
42.
43.
44.

45.
46.

47.

48.

49.

50.

6O

Schnell, K. (1974) Untersuchengen zum Mechanismus des
Sulfattransportes durch die Erythrocyten-Membran Thesis.

Gunn, R.B. (1979) Membrane Transport in Biology Vol. II. 81-115.
Deuticke, B. (1967) Pfleugers Arch. 296, 21-38.

Hieth, J.0. (1970) J. Physiol. (Land) 207, 581-609.

Passow, H. (1969) Prog. Biophysics molec. Biol. 19, 423-467.

Gunn, R.B., Dalmark, M., Tosteson, D.C., and Hieth, J.0. (1973) J.
Gen. Physiol. 61, 185-206.

Dalmark, M. (1976) J. Gen. Physiol. 67, 223-234.

Schnell, K.F., Gerhardt, S., and Schoppe-Fredenburg, A. (1977) J.
Membr. Biol. 30, 319-350.

Pinteric, L., Manery, J.P., Chaudry, I.H., and Madapallimattam, G.
(1975) Blood 45, 709-724.

Van Amelsvoort, J.M.M., Van Hoof, P.M.K.B., De Pont, J.J.H.H.M.,
and Bonting, S.L. (1978) Biochim. Biophys. Acta 507, 83-93.

Grant, C.N.M., Nu, S.H.w., and McConnell, H.M. (1974) Biochim.
Biophys. Acta 363, 151-158.

Mazorow, D.L., Haug, A., and McGroaty, E.J. (Chapter 5)
Chlorpromazine and chlorpromazine methiodide effects on the human
erythrocyte membrane.

CHAPTER III

Alterations in the Human Erythrocyte Membrane by Anions.

II. Anions Affect Prostaglandin E1 and E2 Receptors

Authors: D.L. Mazorow, A. Haug, and E.J. McGroarty

61

62

ABSTRACT

Fresh whole human erythrocytes drawn only from female subjects were
used to study the effects of prostaglandin E1 (PGEi), prostaglandin
E2 (PGEz), and the synthetic prostaglandin 16,16 dimethyl PGE]
(di-Me PGE1) on erythrocyte membrane properties. These prostaglan-
dins affect erythrocyte membrane structure as determined using electron
spin resonace probing techniques. The effects of prostaglandins on
erythrocyte membrane structure were examined in several different buf-
fer systems. Each of the prostaglandins induced an increase in mem-
brane order when the erythrocytes were suspended in a physiological
bicarbonate buffer (Earles' ). When calciun was not added to the buffer
the membrane ordering induced by PGE] was not affected, but the mem-
brane ordering induced by PGEZ was eliminated. Neither PGE] nor PGEg
induced significant membrane ordering in an Earles' buffer containing
glucose. Erythrocytes suspended in a phospahte buffer containing both
glucose and calcium at either pH 7.0 or pH 7.4 were not significantly
perturbed by PGE] 0r PGEz. However, if 0.5 mM bicarbonate was not
added to phosphate buffer, pH 7.0, the erythrocytes became dramatically
disordered in the presence of either PGE] 0r PGEz at room temperature.
Erythrocytes added to phosphate buffer pH 7.4 with 0.5 mM bicarbonate
became ordered in the presence of both PGE] and PGE2. Erythrocytes
exposed to prostaglandins in a N-2-hydroxyethy1 piperazine-N-Z-ethane
sulfonic acid buffer (HEPES), pH 7.4, became only minimally disordered

63

in the presence of prostaglandins. However, if 0.5 mM bicarbonate was
added to the HEPES buffer (p 7.4), the erythrocytes became disordered
in the presence of prostaglandins. Thus, we found that prostaglandins
can alter erythrocyte membrane structure and that the degree of

alteration can be modulated by different buffer components.

64

INTRODUCTION

Prostaglandins are a class of compounds derived from arachidonic
acid, which are extremely biologically active and, generally, have a
very short half-life. The techniques to characterize and measure
prostaglandin binding were first developed using rat liver fractions
(1). It was shown that the majority of the prostaglandin binding sites
were located in the plasma membrane and that there existed a single
binding site for the E type prostaglandins. Since then, E type recep-
tors have been found in several different tissues. Prostaglandin E
(PGE) receptors have been reported in the non-pregnant human myometrium
plasma membrane (2), both the plasma membrane and the smooth endoplas-
mic reticulum from both rat and hunan skin (3, 4), the plasma and
nucleus membranes of the bovine corpus luteum (5, 6), human platelet
membrane (7), normal and transformed fibroblasts (8), frog erythrocytes
(9), and human erythrocytes (10-15).

The function of the prostaglandin receptor is known in some of
these tissues, but in others it has not yet been determined. Prosta-
glandins are believed to bind to protein receptors (3-5, 11, 16). In
platelets, addition of P012, resulted in a large increase in cAMP
levels (17, 18). Prostaglandins, likewise, reportedly activate the
adenyl cyclase of normal and transformed fibroblasts (8).

In other tissues, such as the human erythrocyte, the role of pros-

taglandin receptors is not yet clear. In this study, we investigated

65

the erythrocyte membrane structural changes induced by the E type

prostaglandins under a variety of buffer conditions.

66

MATERIALS AND METHODS

The prostaglandins were a gift from Dr. M. Ruwart of the Upjohn
Company. The 5-doxyl stearate (5-DS) was purchased from Molecular
Probes.

The blood was obtained from non-smoking female volunteers and
collected without a tourniquet. The erythrocytes were washed and
packed in the buffers as described earlier (Chapter 2, Methods) and
incubated in either prostaglandin E1 (PGEi), prostaglandin E2
(PGEZ), or the synthetic prostaglandin 16, 16 dimethyl PGE]
(di-Me-PGEi) for 30 minutes at 37°C and repacked to a hematocrit of
70 t 1 (Chapter 2, Methods).

The prostaglandins were initially dissolved at millimolar concen-
trations in 100% ethanol and serially diluted in the appropriate buffer
to the desired concentrations.

All other procedures and techniques are the same as described in

Methods, Chapter 2.

67

RESULTS

EARLES BUFFER

PGE], PGEz and diMe-PGE1 significantly ordered the erythrocyte
membrane at room temperature (24°C). Maximal ordering occurred at
10-9 M for PGE], 10-8 M for PGEz, and 10-10 M for diMe-PGEl. At
37°C PGE] induced significant ordering of the erythrocyte membrane
but PGEZ and diMe-PGE] did not (Figure 1A, 18).

The absence of calcium in the suspending mediun (Earles without
calcium) reduced the membrane structural changes induced by PGE] by
between 30% and 55% at 24°C and by 0% to 50% at 37°C. PGEz induced
no erythrocyte membrane structural changes in the absence of calcium at
24°C. However, the slight changes induced by PGE2 at 37°C were not
affected by the absence of calcium (Figure 1A, 18). Neither PGE] nor
PGEZ induced any membrane structural changes in the presence of

glucose.

PHOSPHATE BUFFER

In phosphate buffer pH 7.0 neither PGE} nor PGEz induced any
significant membrane structural changes in the erythrocyte membrane at
the concentrations studied. PGE] in phosphate buffer, pH 7.4, in-
duced a disordering of the erythrocyte membrane at 24°C, but PGEz in

68

.uasm an .m

ucm .ueem am .< mmooapm mapn mapam mmpcmm wen . «no unocuez mupmm
mmpcmm .mupmm .mm Lam =_ cavemamam Aamue mziwuw Hume Fxsums_u ma

.oH :wucmpmcumoeg can .A moav mm caucmpmmumoca .Afimuav Hm caucapmmumoca
z Nfiiofi . z miofi mo mucwmmca mgu cw mcmwucz—o> wpmsmm soc» mmuxuogsuzum
cuss; we Aam<v cmumsucma emcee cw mmaccgo ucmucma .~ ocsmwm

accusatotam $3 23 3:5 atom
”woe .uoe Joe .5... 62.3... «3.. .2:

..fl. cu. «a. :92? cm. . . om. ca. .W. 2a. a“. :92“. «M. 8.0.. :M. 2m. om. ...”. :92“. am. am.

. v.0:

 

 

 

 

 

 

 

 

 

$m<

 

 

 

 

 

 

 

 

69

33203.25 .0 23 .....u . .....u

 

..m... .u... 3...... .... £233.. 2.... ...".
223332 23.2233332. 33.
..0.o.0. .0 :0. cm. .0. ..M... 0.. ..0. 20. .0. ..0. 20. .0 «.0. .... 0.20. .0...0 ...- 0.2m... 0. .0.

I36:

v.0:

 

 

 

 

 

 

 

. $2

 

 

 

 

 

 

 

 

70

the same buffer induced no structural changes in the erythrocyte
membrane.

When PGE] or PGEZ was added to phosphate buffer, pH 7.0 con-
taining bicarbonate (P04/HC03) a disordering of the erythrocyte
membrane was seen at 24°C with a less significant disordering at 37°C
(Figure 2A,B). When PGE] or PGEz was added to erythrocytes
suspended in phosphate buffer, pH 7.4, with bicarbonate (P04/H603
pH 7.4), the membrane became more ordered at 24°C and less ordered at

37°C.

HEPES BUFFER

 

Erythrocytes were suspended in N-2-hydroxyethyl piperazine-N-Z-
ethane sulfonic acid (HEPES) buffer and mixed with either PGE],
PGEZ, or diMe-PGEi. At 24°C only minimal membrane structural
changes were seen, but at 37°C each of the three prostaglandins induced
membrane disordering. Both 10-9 M PGE] and 10-9 M PGEZ induced a
disordering of the erythrocyte membrane when suspended in a mixture of
HEPES and bicarbonate buffer (HEPES/H003) at both 24°C and 37°C
(Figure 2A,B).

71

Figure 2. Percent changes in order parameters (AS%& of huTanM erythro-
cytes from female volunteers in the presence of 10 - 10
prostaglandin E1(PGE ), prostaglandin E (PGEZ), and prostaglandin 16,
16 dimethyl PGE1 e PGEl) suspended 1n phosphate buffer, pH 7. 0
(P04 pH 7. 0); phosphate buffer, pH 7. 4 (P04 pH 7. 4); phosphate buffer
with 0. 5 mM) bicarbonate pH 7. 0 (P0 + H003 pH 7. 0); phosphate buffer
with 0. 5 mM bicarbonate, pH 7. 4 (P 4 + HC03 pH 7. 4), HEPES buffer,

and HEPES buffer pH 7.4 (HEPES); and HEPES buffer with 11 mM bicarbo-
nate pH 7.4 (HEPES + H003) A. at 24°C and B. at 37°C.

72

n8:
+mm1w: mmmwx
J a
mom
..o. ab. .1 .90.

~moe.moa 90....

 

esxemooz.eoe
«woe Joe
=10_0.b_0.0- ..b—0.0_0b_

 

 

o~ze

no.2.22.

«we.
...0. .0.

 

.moa mod
york. h=.b_

 

ewxe.oe

.moe
20.2.0.

 

 

osze co.
= «woe .moe
:

 

 

 

br޴00.?brhzohx

QT

T?

00

 

o\om<

73

”8..
......

mum!

......

3.2 3.2.8..

«a... ......

 

 

at...
3.... .81.... 78...... 3.... 5a. 8.2.. 6..
«u... ...... ...... ...... an... a... «u... ......
30.20. .0. 20.20. .040. .0. 2.0. .0. ..0..0. 20:0. ...920. .0: ...0r.0. ..9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m..-

e\om<

 

N.

74

DISCUSSION

Results presented here indicate that the amount and degree of
structural changes induced within the erythrocyte membrane by prosta-
glandins is dependent upon the buffer used, the energy level of the
cells (glucose present), the presence of calcium, the pH, and bicarbo-
nate in a buffer.

Several studies have shown that the presence of very small amounts
of prostaglandins affect erythrocyte whole cell properties such as
protection from hemolysis. Kury et al (14) have shown that erythro-
cytes drawn from male volunteers were made less deformable and the
membrane more ordered by PGEZ. The increase in order parameter was
dependent on PGEZ concentration with the maximal increase at
10“] M. Conversely, PGE] was shown to have the opposite effect
(14). It disordered the erythrocyte membrane and increased whole cell
deformability with a maximal affect also at 10"] M. In addition,
their study reported that calcium was needed in the buffer for the
PGE] and PGE2 induced perturbations. Even though PGE] caused an
increase and PGE2 caused a decrease in whole cell deformability, the
relative apparent viscosity (viscosity ratio blood/plasma) of the
erythrocyte was not altered by the presence of either PGE] or PGEZ'
(13) in phosphate buffer, pH 7.0.

It has been shown that the presence of either PGE] or PGEZ

alters the circular dichroism (CD) adsorption spectrum of washed

75

erythrocyte ghosts (14). Both prostaglandins reportedly cause a
decrease in the mean residue ellipticity at 37°C when observed between
250 nm and 190 nm. The CD spectral changes, however, were not biphasic
as a function of prostaglandin concentration.

In other studies erythrocytes in an isotonic solution which were
shaken more than 200 times per minute showed either 6% (cells from
females) or 10% (cells from males) lysis (15). 10'9 M PGE] or
PGEZ reduced lysis of cells from both male and female subjects, but
the degree of protection was greater for erythrocytes drawn from male
subjects. The presence of calciun did not affect protection resulting
with PGE].

Our results show that at room temperature in an Earles buffer
PGE], PGEz, and diMe-PGE; each induce an ordering of the erythro-
cyte membrane. Contrary to the results of Kury et al (14) the absence
of calcium did not dramatically affect the degree of structural altera-
tions induced by PGE], but was needed for the PGEZ induced membrane
changes. PGE] and PGE2 may initiate some calcium leakage which is
protective to cells against lysis. Cells from male subjects are more
susceptible to mechanical lysis and may also be more deformable. The
calciun requiring PGE] induced changes seen by Kury and co-workers
may reflect the use of erythrocytes only from male subjects. One
reason that we did not see a PGE] initiated calcium effect on
erythrocyte structure may be because we used erythrocytes only from
female volunteers. Cells from females may not be as sensitive to a.
calcium leakage as from men.

When we repeated the experiments of Kury et a1 (14), with the

erythrocytes su5pended in phosphate buffer, the membrane structure was

76

not affected by PGE] or PGEZ (glucose present). ‘At pH 7.4, PGE]
induced a decrease in membrane order but PGEZ induced no change.

When either PGE] or PGEz was added to erythrocytes suspended in
phosphate buffer pH 7.0 containing 0.5 mM bicarbonate a dramatic mem-
brane disordering was induced. When the pH of the solution was in-
creased to 7.4, a dramatic membrane ordering was seen which was depend-
ent on the concentration of either PGE] or PGEZ. At 37°C, the mem-
brane structural alterations induced by PGE] or PGE2 are less pro-
nounced in all cases except for cells in phosphate buffer with bicarbo-
nate pH 7.4, where both PGE] and PGE2 induce a slight membrane
disordering. Bicarbonate added to a HEPES buffer system also induced
the same type of changes. The bicarbonate appears to cause an increase
in the membrane structural changes induced by the prostaglandin.

The presence of bicarbonate in either phosphate or HEPES buffers
appears to induce a magnification of the structural changes induced by
prostaglandins. It is possible that the prostaglandin receptor is
situated near the band 3 protein and reflects band 3 conformational
changes. If the presence of prostaglandins helps to regulate transport
processes, the additional presence of bicarbonate may accentuate these
changes.

Band 3 is probably not the only protein involved in the prosta-
glandin induced effect. Changes in CD absorption spectra upon addition
of prostaglandins indicate protein conformational changes. Band 3 is
bound to the cytoskeletal matrix via ankyrin (19-21). So, small
changes in band 3 structure may be reflected as large changes in the
cytoskeletal matrix. This could account for much of the prostaglandin

effect. The cytoskeletal matrix is also extremely sensitive to small

77

fluctuations in calciun levels (22). If PGE] and/or PGEZ induce
calciun leakage, this could be reflected as large conformational
changes in the cytoskeletal structure.

Our data suggests that in an Earles' buffer membrane structural
changes induced by prostaglandins are most pronounced in the presence
of calciun; however, changes induced by PGE] do not appear to depend
on the presence of calciun. Energy depleted cells would be especially
sensitive to prostaglandin induced structural changes if the cytoskele-
tal matrix is involved. It would take energy to restore the cytoskele-
tal matrix to its unperturbed state.

Having repeated the experiments of Kury et a1 (14) we saw no struc-
tural changes induced by either PGE] or PGEz. The major difference
between their methodology and ours was that they used male subjects and
we used female subjects. Taniguchi et al (15) have shown that the
degree of protection given by PGE] from mechanical lysis is more
dramatic for cells drawn from men. Also preliminary data from our
laboratory indicates that aspirin, a prostaglandin synthetase inhibitor
(23), when ingested induces opposite erythrocyte membrane structural
changes in men and women.

In conclusion, we have shown that in an energized cell, neither
PGE] nor PGEZ induce any detectable changes in membrane order. If
bicarbonate is added to the suspending media, a change in membrane
order occurs. Finally, the alterations in membrane order induced by

either PGE] or PGEz may be dependent on the sex of the donor.

IO.

11.

12.
13.

14.

BIBLIOGRAPHY

Smigel, M. and Fleischer, S. (1974) Biochim. Biophys. Acta. 332,
358-373.

Crankshaw, D.J., Crankshaw, J., Branda, L.A., and Daniel, E.E.
(1979) Arch. Biochem. Biophys. 198, 70-77.

Lord, J.T., Ziboh, V.A., and Warren, S.K. (1978) Endocrinology
102, 1300-1308.

Lord, J.T. and Ziboh, V.A. (1979) J. Investigative Dermatology 73,
373-377.

Rao, Ch. V. and Harker, C.W. (1978) Biochem. Biophys. Res. Commun.
85, 1054-1060.

Rao, Ch. V. and Mitra, S. (1979) Biochim. Biophys. Acta 584,
454-466.

McDonald, J.N.D. and Stuart, R.K. (1974) J. Lab. Clin. Med. 84,
111-121.

Davies, P.J.A., Chabay, R., Zech, L., Berman, M., and Pastan, I.
(1980) Biochim. Biophys. Acta 629, 282-291.

Lefkowitz, R.J., Mullikin, D., Wood, C.L., Gore, T.B., and
Mukherjee, C. (1977) J. Biol. Chem. 252, 5295-5303.

Rasmussen, H., and Lake, W. (1975) Prostaglandins and the

Mammalian Erythrocyte in Prostaglandins jn_Hematology. Melvin J.
Silvers, et al. ed. p. 187-202. 1977. Spectrum Pub ishers, Inc.

N.Y.

Meyers, M.B., and Swislocki, N.I. (1974) Arch. Biochem. Biophys.
164, 544-550.

Allen, J.E., and Rasmussen, H. (1971) Science 174, 512-514.

Gerbstadt, H., Hadelt, M., Schuster, R., and Zanker, E. (1978)
Acta Biol. Med. Germ. 37, 907-908.

Kury, P.G., Ramwell, P.W., and McConnell, H.M. (1974) Biochem.
Biophys. Res. Commun. 56, 478-483.

78

15.

16.
17.

18.

19.

20.
21.

22.

23.

79

Taniguchi, M., Aikawa, M., and Sakagami, T. (1982) J. Biochem. 91,
1173-1179.

Johnson, M., and Ramwell, P.W. (1973) Prostaglandins 3, 703-719.

Schafer, A.I., Cooper, 8., O'Hara, D., and Handin, I. (1979) J.
Biol. Chem. 254, 2914-2917.

Miller, 0.V. and Gorman, R.R. (1979) J. Pharmacol. and
Experimental Therapeutics 210, 134-140.

Bennett, V. and Stenbuck, P.J. (1979) J. Biol. Chem. 254,
2533-2541.

Bennett, V., and Stenbuck, P.J. (1979) Nature 280, 468-473.

Shotton, D.M., Burke, B.E., and Branton, D. (1979) J. Mol. Biol.
131, 303-329.

Johnson, R.H. and Kirkwood, D.H. (1978) Biochim. Biophys. Acta
509, 58-66.

Mazorow, D.L., Haug, A., Bull, R., and McGroarty, E.J. (Chapter 4)
Effects of aspirin, indomethocin, and sodium salicylate on human
erythrocyte membranes as detected with electron resonance
spectroscopy.

CHAPTER VI

Title: Effects of aspirin, indomethacin, and sodium salicylate on
hunan erythrocyte membranes as detected with electron spin

resonance spectroscopy.

Authors: D.L. Mazorow, A. Haug, R. Bull, and E.J. McGroarty

80

81

ABSTRACT

Electron spin resonance spectroscopy was used to determine the
structural changes in human erythrocyte membranes prior to and at
intervals following ingestion of either 10 grains acetylsalicylic acid,
10 grains sodium salicylate, or 50 mg indomethacin by both male and
female subjects. Analysis of erythrocytes from female subjects indi-
cated a time-dependent disordering of the membrane over the eight hour
period following aSpirin ingestion while the cells of male subjects
showed a slight membrane ordering over the same time period. Erythro-
cytes drawn from females at the beginning of the menstrual cycle showed
the greatest amount of membrane disordering at one hour following
aspirin ingestion, but by eight hours, the membrane structure had
returned to that of control. The time dependent changes in membrane
structure from females in the middle of the menstrual cycle displayed a
biphasic disordering of the membrane. Ingestion of indomethacin
induced a slight membrane ordering in both male and female subjects
over the times examined. Ingestion of sodiun salicylate by either men
or women did not induce significant changes in erythrocyte membrane
order. Washed erythrocytes when mixed with salicylate, aspirin, or
indomethacin were either identical to control cells or slightly more
ordered. This study suggests that aspirin-induced alterations in

membrane structure may depend upon steroid honmone levels.

82

INTRODUCTION

Ingestion of aspirin (acetylsalicylic acid) reportedly affects many
diverse physiological processes including histamine release (1), plate-
let aggregation (2, 3, 4), broncho constriction (l, 5), renal trans-
plant rejection (6), and renal proteinuria (7). These effects are
likely to be induced by irreversible aspirin inhibition of the enzyme
cyclo-oxygenase, which is involved in the synthesis of the prosta-
glandins and the thromboxanes (8-10). Indomethacin is believed to work
in a manner similar to aspirin, but its inhibition of the cyclo—oxy-
genase is reversible (ll-13). Salicylate is believed to bind revers-
ibly to the cyclo—oxygenase, but unlike aspirin or indomethacin, it
does not inhibit the enzyme (ll-12).

Levels of estradiol and estrogen or other steroid hormones may
modulate the affect of aspirin. Aspirin increases the bleeding time in
men more so than in women (14), and women absorb more aSpirin than men
do (15). The incidence of ischemic heart disease among pre-menopauasal
women is much lower than among men and post-menopausal women (16). Low
dose aspirin therapy has been implicated as an antithrombotic agent
(17,18). Endogenous levels of estrogen are believed to maintain a more
favorable ratio of prostacyclin to thromboxane which prevents platelet

aggregation (l9).

83

Specific low dosages of aSpirin enhance the prostacyclin to throm-
boxane ratio by inhibiting platelet throboxane production with less of
an effect on vasculature prostacyclin production (3, 20-27).

In this study we have analyzed the effects of aspirin, sodium
salicylate, and indomethacin ingestion on the physical structure of the
human erythrocyte membrane. We have shown that the effects of aspirin
are different for men and women, and among women are dependent upon the
menstrual cycle. In addition we analyzed changes in membrane structure
induced by aspirin, sodiun salicylate, and indomethacin addition to
cells suspended either in bicarbonate buffer or in filtered serum. Our
results indicate that differential effects of aspirin on the erythro-
cytes of women may require metabolic reaction or physical association
with serum components and that the effects of aspirin on erythrocyte

membrane structure appear to be dependent upon the menstrual cycle.

84

METHODS

Freshly drawn blood was obtained from healthy non-smoking, con-
senting individuals. The donors had not taken any medication or alco-
hol for at least 48 hours prior to ingestion of aspirin, sodiun salicy-
late, or indomethacin. Female subjects were not taking oral contracep-
tives. Erythrocytes were drawn prior to, 1/2 hour, 1 hour, 2 hours, 5
hours, and 8 hours following ingestion of 10 grains (648 mg) of
aspirin. Cells were also drawn prior to, 1 hour, and 2 hours following
ingestion of 10 grains of sodium salicylate; and prior to, 1 hour, 5
hours, and 8 hours following ingestion of 50 mg indomethacin. Studies
using female subjects were performed both at the beginning and between
the middle and end of their menstrual cycle.

Anticoagulation was effected by defibrination. Cells were centri-
fuged, the buffy coat was removed, and the erythrocytes were washed
three times in 300 im05m NaCl, pH 7.4, and once in an Earle's salt
solution or in filtered serum. The Earle's salts contained 116.4 mM
NaCl, 5.4 mM KCl, 0.8 mM MgS04-7H20, 0.5 mM NaH2P04. 1.8 mM
CaClz, and 26.2 mM NaHC03. Filtered serum‘was prepared by fil-
tering hunan serun through an Amicon filter UM10 (10,000 dalton cut-
off). The cells were packed in fresh Earle's salts or filtered serum
to a hematocrit of 70 t l. A 1.0 ml aliquot of the packed cells was
placed into a test tube which contained 0.2 umoles of the electron spin

resonance (ESR) probe 5-doxyl stearate. For in vitro studies, 22 ug/ml

85

of aspirin, 74 ug/ml of sodium salicylate, or 2 ug/ml of indomethacin
was added to either the Earle's salt solution or the filtered human
serum. 1 ml of packed cells was added to 39 ml of the drug-containing
solutions, incubated at 37°C for 30 minutes and repacked. Packed cells
were placed in a quartz cuvette for ESR probing studies. All ESR
experiments were carried out with a Varian Century line ESR
spectrometer, model E-112, equipped with a variable temperature
controller. An external calibrated thermistor probe was used to
monitor the temperature of the sample.

The data were analyzed by an iterative least squares program (28).
Briefly, a B-spline was used to provide a smooth fit for the ESR data
and points of inflection served for grouping data. Regression lines
were calculated for each group and then plotted. This analysis allowed
the determination of break points in the linear temperature dependent
spectral parameters. Such breaks in the temperature dependence of the
hyperfine Splitting parameter and order parameter have been correlated
with lipid phase separations or lipid phase transitions from gel-to-
liquid crystalline lipid states (28, 29).

The relative fluidity of the erythrocyte membrane from cells drawn
at different times following aspirin, sodium salicylate, or indometha-
cin ingestion was determined by labelling intact cells with 5-doxy1
stearate. S-doxyl stearate incorporates into the lipid phase of the
membrane in such a way that the unpaired electron of the nitroxide
radical is situated close to but shielded fr m the aqueous phase.) Spin
labelled preparations were placed in the resonance cavity of an ESR
spectrometer. In this fashion the erythrocyte membrane preparation was

exposed to microwaves of a constant frequency while held in a linearly

86

varying magnetic field corresponding to the spin stages of the unpaired
electron. The ESR spectra are diSplayed as the first derivative of the
microwave absorption plotted against an increasing magnetic field in
gauss (Figure l). The distance between the low field and the high
field absorption peaks, the hyperfine splitting parameter (2T11),
is a measure of the largest energy difference between the spin states
of the unpaired electron. With increasing temperature, intermolecular
influences upon the unpaired electron tend to average, resulting in
smaller values of 2T1] which reflect more fluid environments.
2T1], therefore, is related to the micro-viscosity of the environ-
ment from which the probe is reporting. Large values of 2T1]
indicate less fluid environments.

The order parameter, S (calculated using 2T1] and ZTL) measures
the deviation of the observed signal from that of a unifonn orientation
of the probe. For a uniformly oriented sample S=l; for a random sample

S=0. Order parameter is an indication of lipid acyl chain orientation.

87

Figure 1. Electron spin resonance spectrum of human erythrocytes
labelled with 5-doxyl stearate. The Spectrum was taken at 37°C and the
scan range was 100 gauss. Absence of free probe is indicated by the
lack of a major absorption signal at points indicated by arrows.
Symmetry of high and low field peak reveals that the probe was present
in a single environment in the membrane.

88

 

0N+

 

-O

4.5

 

 

 

 

 

...—.N

 

0N1

 

 

89

RESULTS

Erythrocyte membrane structure of cells drawn prior to and at
different times following consumption of aspirin, sodium salicylate, or
indomethacin was measured using the ESR spin probe 5 doxyl stearate.
The spectral parameter S, was measured to determine structural changes
in the erythrocyte membrane. The temperature dependence of the
spectral parameters as plotted in Figure 2 were analyzed in terms of
linear components using a linear regression program (28). The changes
in membrane order parameter were compared to that of erythrocytes from
donors prior to ingestion of the drug and the differences were
computed. Spectral order changes of less than 0.5% are not considered
significant.

The probed erythrocytes drawn from both male and female subjects
appeared structurally identical prior to ingestion of any medication
(Figures 2A and B). One hour after ingestion of aspirin, the erythro-
cytes from women at the beginning of the menstrual cycle were disord-
ered to the maximum extent seen after ingestion, and the erythrocytes
from men were maximally ordered at the same time point (Figures 2C and
3). The time dependent changes in erythrocyte membrane structure in
men and women following the ingestion of aspirin is shown in Figures 3A
and 3B. Erythrocyte membranes from men were more ordered following
ingestion of aspirin, and the membranes remained ordered throughout the

time span studied (Figure 3). Membranes of erythrocytes drawn from

90

Figure 2. Erythrocytes were drawn from male (open circles) and female
(closed circles) donors. Spectral parameters 2T1] (hyperfine
splitting parameter) and S (order parameter) were plotted versus
temperature (°C). A. Temperature dependence of membrane structural
parameters of erythrocytes from male and female subjects prior to
ingestion of any drug. 8. Temperature dependence of membrane order of
male and female subjects prior to ingestion of any drug. C. Tempera-
ture dependence of erythrocyte membrane structure of cells drawn from
female subjects at the beginning of the menstrual period and from male
subjects 1 hour after ingestion of aspirin.

 

 

91

 

 

0.. 00 0e mm mm mm .VN 0. N.
. _ . . a . _ . _
. o .0
v . o
.o v
.0 o
0.050. 0
0.05 o

 

 

 

92

 

8

8

8

8‘

 

8

l

 

0 male
0 female

 

 

TC)

16

1
20

24

L J
28 32
Temp (°C)

93

80

 

 

0 male .
A female begmmq’ '
76 menstrual cycle

74-
72—

7o-—

601-

 

 

58

 

56 .
1 1 1 1 1 1 1 1 1

12 16 20 24 23 32 36 4o 44 '48
Temp (°C)

94

 
   

 

 

   

 

L\.\\\\\\\\‘

 

 

 

 

 

241- E

05hr 1hr 2hr 5hr 8hr

Figure 3. Time related changes in order parameter of erythrocytes from
female subjects at the beginning of the menstrual cycle (D ) from
female subjects at the middle to end of the menstrual cycle ( ), and,
from male subjects ( I ) following aspirin ingestion. Time re ated
changes in order parameter of erythrocyte membranes from female sub-
jects at the beginning of the menstrual cycle ( [fl] ) and fran male sub-
jects (§ ) following ingestion of sodiun salicylate, A. measured at
24°C or B. measured at 37°C.

95

08

0.4

A37. 0*

 

 

 

-oer

 

 

 

'16-
Q5hr 111! 2111‘ 5hr

96

women at the end of their menstrual cycle became disordered following
ingestion of aspirin (Figure 3). However, if aspirin was ingested by
the female subjects at the beginning of the menstrual cycle the maximum
membrane disordering occurred at 1 hour with little or no membrane
disordering at 30 minutes and 8 hours (Figure 3).

Following ingestion of indomethacin, erythrocytes from men were
maximally ordered after 1 hour. For women, erythrocytes drawn at the
beginning of the menstrual cycle were maximally ordered in membrane
structure 1 hour following ingestion of indomethacin and remained
ordered over 5 hours after ingestion of indomethacin (Figure 4).
Membranes of erythrocytes from women at midestrous became ordered (at
37°C) by 1 hour after ingestion of indomethacin and remained ordered
for at least 2 hours. In contrast, the membranes of erythrocytes from
men or women drawn either 1 or 2 hours following ingestion of sodium
salicylate were identical to control (Figure 3).

In vitro addition of aspirin to washed erythrocytes resuspended in
either Earle's salt or filtered serun caused no changes in membrane
order when measured at room temperature, but at 37°C, the membranes of
erythrocytes from both men and women were slightly more ordered than
control erythrocytes (Table l). The in vitro addition of indomethacin
to washed cells from either men or women induced no significant changes
in membrane structure. Likewfise, in vitro addition of sodiun salicy-
late induced essentially no changes in cells from women. However, the
membranes of erythrocytes from men when analyzed in an Earle's salt

solution were significantly ordered at 37° (Table l).

97

1.2. r-

 

 

 

 

 

i-EJ-i

 

 

lhr 2hr 5hr 8hr

Figure 4. Time related changes in order parameter of erythrocytes from
female subjects at the beginning of the menstrual cycle ( D ) from
female subjects at the middle to end of the menstrual cycle (Z ), and

from male subjects (I ) following ingestion of indomethacin, measured
either at A. 24°C, or at B. 37°C.

98

|.2 1'

 

 

2”.”"4

 

 

 

-04 -

’O.8 -

'12:-

5hr 8hr

2hr

lhr

Table l.

99

Changes in order parameter of erythrocytes drawn from males

and females and suspended in either Earle's salts or serum ultrafil-
trate containing either aspirin, sodium salicylate, or indomethacin.

Drug Donor
Aspirin female
male

Sodium salicylate female

male

Indomethacin female

male

Incubation
temperature

24°
37°
24°
37°

24°
37°
24°
37°

24°
37°
24°
37°

I
‘60

0000
O O O

O
OOON ONOO 010101“)

Earles
Changes in order(S)%

1+ 1+ 1+ 1+
0 O I O O O
whom—i wood—-
01010101 w—a—a

1+ H- H 1+
COCO 0000 GOOD
0 C

1+ 1+ 1+ 1+
0 O

NNO-‘

010101

filtered serum

0.2

H14- H- 1+

0 O .
mud-d
\INUTUI

H'H-H'H- H-H'HH'

0000 GOOD GOOD
0 O . O O 0 O

“Nd—I wood—-

NUTVL” 01-th

lOO

DISCUSSION

Results presented here indicate no detectable differences in
erythrocyte membrane structure between cells drawn from men and women
prior to aspirin, sodium salicylate, or indomethacin ingestion (Figure
2A and B). After ingestion of aspirin, time related changes in mem-
brane structure occurred in both men and women (Figures 2C, 3A, 3B).
For men the erythrocyte membrane became more ordered than the control.
In contrast, the erythrocyte menbrane of cells from women became less
ordered than that of cells prior to aspirin consunption. For women,
the degree of change appeared to be somewhat dependent upon when in the
menstrual cycle the study was done. As can be seen in Figure 28, 1
hour after ingestion of aspirin, membrane order is dramatically diffe-
rent between women (at the beginning of the menstrual cycle) and men.
For men, the erythrocyte membranes became more ordered than the con-
trol. The differences in effects of aspirin seen between men and women
and among the same women at varying times in the menstrual cycle is
probably caused by differences in amounts of steroid hormones and/or
prostaglandins prior to aspirin ingestion.

Aspirin is known to inhibit prostaglandin synthesis by acetylating
the enzyme cyclo-oxygenase which is involved in one of the first steps
in its synthesis (8). Therefore, following ingestion of a normal dose
of aspirin (648 mg) most if not all prostaglandin synthesis in the
vasculature is inhibited (3, 12, 2-22, 30-34). It is also known that

erythrocytes have receptors for prostaglandins of the E type (35-37,°),

101

and receptors for either estrogen (38) or estradiol (39). It is
possible that either the prostaglandin and the steroid hormone receptor
are related or they interact within the membrane.

Several studies have shown a relationship between estradiol levels
and prostaglandin production. Estradiol is known to increase prosta-
cyclin production in the aorta (l9), and estrogen reportedly increases
prostaglandin Fla (PGFia) production (40). Both uterine estradiol and
PGan concentrations increase during pregnancy (11) as do plasma
estradiol levels (42). These reports indicate a correlation between
estradiol and prostaglandin production. Thus, aspirin and indomethacin
inhibition of prostaglandin synthesis may be modulated by endogenous
levels of estradiol. This may account for differences in erythrocyte
membrane structure following ingestion of aspirin and indomethacin at
different times in the menstrual cycle.

Cyclo-oxygenase, the site of action of these drugs, has an active
site and an allosteric site (11, 12, 31). It is known that pretreat-
ment of the enzyme with either sodium salicylate (ll, 12, 31) or indo-
methacin (ll, 12) will prevent the irreversible inhibition of cyclo-
oxygenase by aspirin. Indomethacin reversibly (12) inhibits the cyclo-
oxygenase. Our studies showed that ingestion of sodiun salicylate did
not significantly alter membrane structure of erythrocytes from either
men or women. The erythrocyte membrane structure has been shown to be
altered by prostaglandins (35-37,°). In vitro addition of sodium
salicylate, aspirin, or indomethacin to cells in a bicarbonate buffer
did not affect membrane structure. Thus, specific enzymatic or physi-

cal interactions must occur with the cells within the vasculature

°Chapter 3

102

following consumption of the drug. Ingestion of aspirin induced drama-
tic membrane changes in the erythrocytes drawn from both men and women
over the 8 hour time span examined. Aspirin as well as indomethacin
are in the blood stream 30 minutes after ingestion (43) and their
effects were followed for several half-lives. Aspirin has a half-life
of 20 to 30 minutes (43) and indomethacin has a half life of 4 hours
(43).

To determine if inhibition of cyclo-oxygenase was the cause for the
membrane structural changes in men and women after aspirin ingestion,
indomethacin and sodium salicylate were given orally to male and female
volunteers. Indomethacin, also a cyclo-oxygenase inhibitor (11, 12,
31), should upon ingestion result in changes in erythrocyte membrane
structure similar to that of aspirin. Sodiun salicylate, which is not
a cyclo-oxygenase inhibitor (11, 12, 31), should not induce erythrocyte
membrane changes.

Indomethacin and aspirin did not induce the same structural changes
in erythrocyte membrane structure, but the membrane changes induced by
indomethacin and aspirin were different for men and women. Other
studies show that indomethacin can modulate the erythrocyte membrane
and estradiol levels in ways different from aspirin; and the differ-
ences observed in membrane order following ingestion of aspirin or
indomethacin may not be solely prostaglandin related. The ability of
these two drugs to protect erythrocytes from heat-induced lysis is
different (44, 45) with only indomethacin having a protective effect.
Also, indomethacin will prevent ovulation of ovariectamized rats if
injected into the anterior hypothalamus while aspirin will not (46).

Indomethacin may inhibit synthesis of membrane active compounds other

103

than prostaglandins. Furthermore, indomethacin may not saturate the
receptors as does aspirin, or aspirin may affect enzymatic systems
other than the cyclo-oxygenase.

Our data indicate that structural changes induced in erythrocytes
by cyclo-oxygenase inhbitors may be modulated by prostaglandin and/or
estradiol levels. Both indomethacin and aspirin induce changes in
erythrocyte structure which in women appear to be menstrual period
modulated. Sodium salicylate, which is not an inhibitor of prostaglan-
din synthesis, induces very little structural change in erythrocyte

membranes.

lO.

BIBLIOGRAPHY

Crocken, J.F.S. 1982. Commentary: Acetylsalicylic acid or
aspirin Can. Med. Assoc. J. 126:611-612.

Waitzman, M.B., H. Kaplan, L. Cornelius, and B. Evatt. 1981.
Aspirin and prostaglandin effects on diabetic dog blood. Fed.
Proc. 40:773.

Preston, F.E., M. Greaves, C.A. Jackson, and C.J. Stoddard. 1982.
Low-dose aspirin inhibits platelet and venous cyclo-oxygenase in
man. Thromb. Res. 27:477-484.

Jouve, R., I. Juhan-Vague, M.F. Aillaud, M.P. Serment-Jouve, and
H. Payan. 1982. Comparison of the effects of aspirin and
indomethacin an aortic atherogenesis atherosclerosis induced in
rabbits. Atherosclerosis 42:319-321.

Spector, S.L., H.G. Morris, and J.C. Selmer. 1981. Clinical
responses and serum prostaglandin levels in aspirin idiosyncrasy:
Effect of aspirin and a new non-steroidal anti-inflammatory agent.
Chest. 80:676-681.

Foegh, M.L., J.F. Winchester, M. Zmudka, G.B. Helfrich, P.W.
Ramwell, and G.E. Schreiner. 1982. Aspirin inhibition of
thromboxane release in thrombosis and renal transplant rejection.
Lancet. Jan 2:48-49.

Kurokawa, H., and K. Sakamoto. 1982. Effects of aspirin
prednisolone and indomethacin on nephrotoxic serun nephritis in
the rat. Br. J. Pharmacol. 75:87-92.

Roth, G.J., N. Stanford, and P.W. Majerus. 1975. Acetylation of
prostaglandin synthase by aspirin. Proc. Natl. Acad. Sci. USA.
72:3073-3076.

Hamberg, M., J. Svensson, T. Wakabayashi and B. Samuelsson. 1974.
Isolation and structure of two prostaglandin endoperoxides that
cause platelet aggregation. Proc. Natl. Acad. Sci. USA. '
71:345-349.

Moncada, S., R. Grylglewski, S. Bunting, J.R. Vane. 1976. An
enzyme isolated from arteries transforms prostaglandin
endoperoxides to an unstable substance that inhibits platelet
aggregation. Nature 263:663-665.

104

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

105

Livio, M., A. Del Maschio, C. Cerletti, and G. de Gaetano. 1982.
Indomethacin prevents the long lasting inhibitory effect of

aspirin on human platelet cyclo-oxygenase activity.
Prostaglandins 23:787-796.

Cerletti, C., M. Liveo, and G. deGaetano. 1981. Non-steroidal
anti-inflammatory drugs react with two sites on platelet
cyclo-oxygenase: Evidence from "in vivo" drug interaction studies
in rats. Biochim. Biophys. Acta 714:122-128.

Stanford, N., G.J. Roth, T.Y. Shen, and P.W. Majerus. 1977. Lack
of covalent modification of prostaglandin synthetase
(cyclo-oxygenase) by indomethacin. Prostaglandins 13:669 .

Young, V.P., A.R. Giles, J. Pater, and W.E.N. Corbett. 1980. Sex
differences in bleeding time and blood loss in normal subjects
folowing aspirin ingestion. Thromb. Res. 20:705-709.

Kelton, J.G., C.J. Carter, J. Rosenfeld, M.P. Massicotte-Nolan,
and J. Hirsh. 1981. Sex-related differences in the efficasy of
acetylsalicylic acid (ASA): The absorption of ASA and its effect
02 cgglagen-induced thromboxane 82 generation. Thromb. Res.

2 :1 - 68.

Final Mortality Statistics, no. 23, U.S. Department of Health,
Education, and Welfare, National Center for Health Statistics,
Washington, D.C. 1971.

Marcus, A.J. 1977. Aspirin and thromboembolism. A possible
dilemma. N. Engl. J. Med. 297:1284-1285.

Buchanan, M.R., E. Dejana, J-P. Cazenave, M. Richardson, J.F.
Mustard, and J. Hirsh. 1980. Differences in inhibition of P012
production by aSpirin in rabbit artery and vein segments. Thromb.
Res. 20:447-460.

Chang, W-C., J. Nakao, T. Neichi, H. Orimo, and S-J. Murota.
1981. Effects of estradiol on the metabolism of arachidonic acid

by aortas and platelets in rats. Biochim. Biophys. Acta.
664:291-297.

Hanley, S.P., S.R. Cockbill, J. Bevan, and S. Heptinstall. 1981.
Differential inhibition by low-dose aspirin of human veinous
prostacyclin synthesis and platelet thromboxane synthesis.
Lancet. May 2:969-971.

Ingerman-Wojenski, C., M.J. Silver, J.B. Smith, M. Nissenbaum and
A.W. Sedar. Prostacyclin production in rabbit arteries in situ:
Inhibition by arachidonic acid-induced endothelial cell damage or
by low-dose aspirin. Prostaglandins 21:655-666.

Bull, H., M. Greaves, and P.A. Castaldi. 1982. Recovery of
prostaglandin synthetic activity in rat platelets and aortic rings
after aspirin. Clin. Exp. Pharmacol. Physiol. 9:532-533.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

106

Mehta, J., and P. Mehta. 1982. Dipyridamole and aspirin in
relation to platelet aggregation and vessal-wall prostacyclin
generation. Am. J. Cardiology 49:941.

Demers, L.M. 1981. Inhibition of prostacyclin and platelet
thromboxane A2 by aspirin. New Eng. J. Med. 304:1173.

Kern, D. 1981. Inhibition of prostacyclin and platelet
thromboxane AZ by aspirin. New Eng. J. Med. 304:1173-1174.

Patrono, C., and Patrignani, P. 1981. Inhibition of prostacyclin
and platelet thromboxane AZ by aspirin. New Eng. J. Med.
304:1174.

Preston, F.E. 1981. Inhibition of prostacyclin and platelet
thromboxane A2 by aspirin. New Eng. J. Med. 304:1174-1175.

Brunder, D.G., Coughlin, R.T., and McGroarty, E.J. 1981. A
program for transition point analysis of experimental data.

Janoff, A.S., D.L. Mazorow, R.T. Coughlin, A.J. Bowdler, A. Haug,
and E.J. McGroarty. 1981. The modification of human erythrocyte
membrane stabilizers: An electron-spin resonance study. Am. J.
Hematol. 10:171-179.

Worthington, R.E., and A. Nakeff. 1982. Aspirin inhibits rat
megakaryocyte thromboxane synthesis. Prostaglandins 23:841-853.

Dejana, C. Cerletti, C. de Castellarnau, M. Livio, F. Galletti, R.
Latini, and G. de Gaetano. 1981. Salicylate-aspirin interaction
in the rat: Evidence that salicylate accumulating during aspirin
administration may protect vascular prostacyclin from
aspirin-induced inhibition. J. Clin. Invest. 68:108-112.

Rao, G.H.R., G.J. Johnson, R.K. Reddy, and J.G. White. 1981.
Rapid return of cyclo-oxygenase active platelets in dogs after a
single dose of aspirin. Prostaglandins 22:761-772.

Buchanan, M.R., J.A. Rischke, J. Hirsh. 1982. Asprin inhibits
platelet function independent of the acetylation of
cyclo-oxygenase. Thromb. Res. 25:363-373.

Patrignani, P. P. Filabozzi, and C. Patrono. 1982. Selective
cumulative inhibition of platelet thromboxane production by
low-dose aspirin in healthy subjects. J. Clin. Invest.
69:1366-1372.

Kury, P.G., P.W. Ramwell, and H.M. McConnell. 1974. The effect of
prostaglandins E] and E2 on the human erythrocyte as monitored
by spin labels. Biochem. BiOphys. Res. Commun. 56:478-483.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

107

Rasmussen, H., and W. Lake. 1977. In Prostaglandins and the
mammalian erythrocyte. ProstaglandTfis in Hematology. Melvin J.
Silver et al, eds. Spectrum Publishers, Inc. New York. 187-202.

Taniguchi, M., M. Aikawa, T. Sakagami. 1982. Effect of
prostaglandin E] and polyphloretion phosphate on hemolysis of
human erythrocytes. J. Biochem. (Japan) 91:1173-1179.

Bates, G.W. and E. Jackson Jr. 1981. Metabolism of cathechol
estrogens by erythrocyte catechol-o methyltransferase. Science
213:1145.

Puca, C.A., and V. Sica. 1981. Identification of specific high
affinity sites for the estradiol receptor in the erythrocyte
cytoskeleton. Biochem. Biophys. Res. Comnun. 1032682-689.

Farley, D.B., and D.E. Van Orden. 1982. Effect of prostocylcin
inhibition by tranylcypromine on uterine 6-keto-PGF a levels
during estrogen hypermia in rats. Prostaglandins 2 :657-674.

Haning, R.V. Jr., L. Choi, A.J. Kiggens, and D.L. Kuzma. 1982.
Effects of prostaglandins, dibutyral CAMP, LHRH, estrogens,
progesterone, and potassium on output of prostaglandin F2“
l3,14-dihydro-lS-keto-prostaglandin F2“, HCG, estradiol, and
progesterone by placental minces. Prostaglandins 24:495-506.

Thaler—Dao, H., M. Saintot, M. Ramonatxo, C. Chavis, and A.
Crastegde Paulet. 1982. Prostaglandin biosynthesis by the rat
uterus during the oestrus cycle, temporal correlation with plasma
oestradiol and progesterone concentrations, identifaction of
6-keto PGFia as the major metabolite. Prostaglandins

23:347-359.

Goodman, L.S., and A. Gilman. 1975. The pharmacological basis of
therapeutics. MacMillan Publ. Co., Inc., New York. 5th edition.
305 326-339, 341-343.

Mizushima, V., S. Sakai, M. Yamaura. 1970. Made of stabilizing
action of non-steroid anti-inflammatory drugs on erythrocyte
membranes. Biochem. Pharmacol. 19:227-234.

Mizushima, Y., and S. Sakai. 1969. Stabilization of erythrocyte
membrane by non-steroid anti-inflammatory drugs. J. Pharm.
Pharmac. 21:327-328.

Linton, E.A., and S.A. Whitehead. 1980. Role of arachidonic acid
metabolites in the hypothalamic control of ovulation in the rat.
Biol. of Reprod. 23:726-732.

CHAPTER V

Chlorpromazine Induced Changes in Human Erythrocyte Manbrane Structure

Authors: D.L. Mazorow, A. Haug, and E.J. McGroarty

108

109

ABSTRACT

Fresh whole human erythrocytes were used to study the effects of
chlorpromazine and chlorpromazine-methiodide on erythrocyte membrane
properties. The effects of these two phenothiazines on membrane
properties were determined using electron spin resonance probing tech-
niques and hypotonic hemolysis. Chlorpromazine induced a biphasic mem-
brane disordering which was maximal at 2.0 x 10"5 M. This is the
same concentration at which chlorpromazine maximally protected erythro-
cytes from hypotonic hemolysis. Chlorpromazine methiodide did not
induce any membrane disordering nor did it protect erythrocytes from
hypotonic hemolysis at any of the concentrations examined. We present

a hypothesis to explain the biphasic effects seen with chlorpromazine.

110

INTRODUCTION

Studies have shown that chlorpromazine (CPZ) protects erythrocytes
from hypotonic hemolysis (1), and induces erythrocyte shape changes
(2-5). The membrane expansion resulting from CPZ addition has been
questioned (6), and CPZ reportedly induces no change in cell volume
(7). The binding of CPZ to the cell is freely reversible (8,9), but
the lower the concentration of plasma, the more CPZ goes into the cell
(8).

To better understand the site of action of CPZ on the erythrocyte
membrane we have investigated the dose dependence of CPZ and its
quanternary ammonium analog chlorpromazine methiodide (CPZ-Mel) in
protecting the erythrocyte from hypotonic hemolysis. In addition, we
have analyzed the dose dependent membrane structurel alterations

induced by CPZ and CPZ-Mel using electron spin resonance spectroscopy.

111

MATERIALS AND METHODS

Freshly drawn blood was obtained from consenting healthy non--
smoking individuals. The donors had not taken any medication or
alcohol for at least 48 hours prior to venapuncture.

Anticoagulation was effected by defibrination. Cells were centri-
fuged, the buffy coat was removed, and the erythrocytes were washed
three times in 300 imOsm NaCl pH 7.4, and once in filtered (Amicon
filter UM 10, 10,000 dalton cutoff) human serum, in Earle's salts (145
mM NaCl, 5 mM KCl, 0.7 mM MgS04, 1 mM NaH2P04, 2 mM CaClz, and
22 mM NaHC03) at pH 7.4, or in Earle's salt solution without calcium.
The cells were packed in Earle's salts, Earle's salts without calcium,
or filtered serun to a hematocrit of 7021. A 1.0 ml aliquot of the
packed cells was placed into a test tube which contained 0.2 uMoles of
5-doxy1 stearate (5-DS). After gentle mixing for one minute the cells
were resuspended into 40 m1 of Earle's salts, Earle's salts without
calcium, or filtered serun containing either 1 x 10'6 M. 2 x
10-5 M, or 5 x 10-5 M cpz (gift of Smith, Kline, and French,

Inc.). Cells were incubated at 37°C for 20 minutes and centrifuged.
Packed cells (hematocrit 7021) were placed in a quartz cuvette for
electron spin resonance (ESR) analysis. All ESR experiments were"
carried out with a Varian century 3 line ESR Spectrometer model E-112,
equipped with a variable temperature controller. An external cali-

brated thermistor probe was used to monitor temperature of the sample.

112

The data were analyzed by an iterative least squares program.
Briefly a B-Spline was used to provide a smooth fit for the ESR data
and points of inflection served for grouping data. Regression lines
were calculated for each group and then plotted. This analysis allowed
the determination of break points._ Such breaks in the temperature
dependence in the hyperfine splitting parameter (2T11) and order
parameter (5) have been correlated with lipid phase separations or
lipid phase transitions from gel-to-liquid crystalline states (11).

Hypotonic hemolysis studies were performed in an Earle's salt solu-
tion at 100 mOsm. The solution was prepared by diluting an isotonic
solution (300 imOsm) with double distilled water to the desired hypo-
tonicity. The osmolarity of the solution was determined with an
Advanced Instruments osmometer model 3R. A 10% (v/v) solution of
erythrocytes was placed in 5 ml of hypotonic buffer such that the final
concentration of cells was 0.1%. These cells were incubated for seven
minutes at room temperature in hypotonic buffer and returned to
isotonicity by addition of the appropriate amount of 4 M NaCl. The
cells were centrifuged and the degree of lysis was measured from the
absorbance at 543 an of the supernatant solution. 100% lysis was
achieved by placing the cells in double distilled water, and 0% lysis
was defined as the degree of lysis of cells suspended in a 300 imOsm
solution. Percentage lysis was taken as the difference between these
two values. Protection from hypotonic hemolysis by CPZ and CPZ-Mel was

analyzed at concentrations between 10'2 M and 10‘7 M.

Figure 1.

113

The temperature dependent changes in the hyperfine

Splitting parameters (2T11) of erythrocytes labelled with 5-doxyl

stearate
2 x 10-5

SUSpended in Ear es salts (-e-e-) or in Earles salts with
M chlorpromazine (-c1 - D-).

 

114

on

8% 12m...
ON
q

 

 

u

q

 

 

 

115

RESULTS

ESR spectra from erythrocytes labelled with 5-DS showed little free
probe signal (data not shown). Spectral symmetry of the low field peak
indicated that the majority of the spin label was located in a uniform
environment over the temperature range examined (-2°C to 48°C). The
hyperfine splitting parameter, 2T1], decreased linearly as a func-
tion of increasing temperature (Figure 1) with discontinuities occur-
ring at 3°C to 5°C and in the range of 31°C to 41°C, depending upon the
concentration of CPZ or CPZ-Mel added. Spectra recorded above 12°C
permitted the determination of the order parameter, S. When S was
plotted as a function of temperature, a break Similar to that observed
with 2T1] was detected (Figure 2). The control erythroctyes were
shown to be more rigid (Figure l) or more ordered (Figure 2) than
erythrocytes suspended in solution containing 2 x 10'5 M CPZ. The
ESR studies showed a maximal decrease in membrane fluidity (2T11)
at 2 x 10'5 M (data not Shown). A similar dose dependent decrease
in membrane order, 5, was also detected (Figure 3A,B). Our results
indicate that similar CPZ dose dependent changes occurred in membrane
structure and protection from hypotonic hemolysis with the maximun pro-
tection from lysis occurring at 1.5 x 10'5 M CPZ (Figure 3C). At
concentrations greater than 10'4 M, CPZ became lytic (Figure 3C).
CPZ-Mel did not protect erythrocytes from hypotonic hemolysis at con-
centrations between 10'7 M and 10'3 M, but neither was it lytic

116

 

8C)

75.—

 

 

555. l 1 1 1 1 1 ' 1
I5 20 25 30 35 40 45 50

TEMP (°C)

 

Figure 2. The temperature dependent change in the order parameters (S)
of erythrocytes labelled with 5-doxyl stearate and suspended in Earles
ialts (-;-o-) or in Earles' salts with 2 x 10' M chlorpromazine
'D-D' o

117

Figure 3. Chlorpromazine (-e-e-) or chlorpromazine methiodide
(-o-o-) induced concentration dependent changes in A. order
parameter (S) at 24°C, B. order parameter (S) at 37°C, C. hemo-
lysis in hypotonic Earles buffer at 24°C.

 

118

 

 

 

 

 

 

 

 

OL—TO
.\°~—°
-08 _ A /
0
AS
'16 -
‘24— .
l J
O O ONO
“03$
0
AS
“LS-P
O
.14—
L l
30 0 3—0—0
20 _ c
°/. HEMOLYSIS ,0,_
o _. .../T
L. 1 11
10" 16° 10" 10“

DRUG CONCENTRATION
(M)

119

these concentrations (Figure 3C). It also did not change membrane
fluidity (data not shown) or membrane order (Figure 3A,B) at any of the
concentrations tested.

Erythrocytes suspended in Earle's salts without calcium containing
10-5 M cpz were labelled with 5.05. At 24°C and 37°C the cells
were dramatically disordered (data not Shown). Cells suspended in
filtered serum and incubated in the presence of CPZ did not appear to

be sensitive to changes in pH between 7.0 and 8.5.

120

DISCUSSION

Our analysis of membrane structure of erythrocytes in the presence
of CPZ suggests that CPZ stabilizes the membrane to lysis by inducing
membrane structural changes. This alteration in membrane structure
does not appear to be calciun dependent (data not shown).

It has been reported that CPZ passes through the erythrocyte mem-
brane and binds to the inner monolayer (3), while the quaternary
ammoniun analog can not penetrate the membrane (9,12). In addition it
has been reported that CPZ preferentially binds phosphatidylcholine,
phosphatidylserine, and phosphatidylinositol (13) phospholipids which
are found in the inner monolayer of the membrane. The amount of CPZ
that can bind to dipalmityl phosphatidylcholine liposomes reportedly
increases dramatically above the phase transition at pH 7.4 (14). CPZ
also has been Shown to quench perylene fluorescence in erythrocytes,
indicating that CPZ must be situated in the hydrocarbon core region of
the membrane where perylene localizes (14). Finally, high concentra-
tions (0.2-l mM) of CPZ reportedly do not affect lipid head group move-
ment in red cell ghosts (ESR spin labelled phospholipid studies) but
molecular motion in the lipid core is increased (15). These studies
support the theory that CPZ is able to interact with the membrane’
lipids, preferably the core region and preferentially with the amino

phospholipids.

121

CPZ, however, may not associate solely with the lipids of the
erythrocyte membrane, but it may also interact with Specific proteins.
CPZ has been Shown to decrease Na-K ATPase activity in red blood cells
(16). In addition, if ouabain, a Na-K ATPase inhibitor, is added to
the cells prior to treatment with CPZ, CPZ reportedly does not protect
against lysis (16). Furthermore, cellular ATP content can be induced
by CPZ to drop to 10% of that in untreated cells, and protein
dephosphorylation is induced (17). Not only does CPZ reduce cellular
ATP content, it also reportedly displaces membrane bound calciun.
Although the data is not conclusive, CPZ may affect glucose transport
(18,21).

Gross cellular changes are induced by CPZ. 23 pH CPZ will change
erythrocytes from discocytes to stomatocytes (9), and eventually to
spherostomatocytes (5) along with a concomitant decrease (9-10%) in
both electrophoretic mobility (9) and suspension viscosity at high
shear rates (5).

Our results support those of Elferink (12) which indicate that
CPZ-Mel does not enter or pass through the erythrocyte membrane.
CPZ-Mel did not affect either hypotonic hemolysis or membrane structure
as seen by ESR spectral analysis. If CPZ-Mel does not pass through the
membrane or if it binds only to the outer monolayer, it will not alter
the membrane the same way as does CPZ.

We propose that at low concentrations CPZ passes through the
membrane and acts on the cytoskeletal matrix. CPZ, along with other
phenothiazines, is known to bind to calmodulin (22,25). Calmodulin is

reported to interact with the cytoskeletal matrix via Spectrin (24-27)

122

and to interact with Ca2+.Mg2+ ATPase and to alter its activity
(22,23). Thus, we propose that CPZ interactions with calmodulin
affects the cyto-skeletal matrix inducing erythrocyte shape changes and
changes in membrane lipid structure. It has been reported that if
ankyrin (band 2.1) is absent from erythrocyte ghosts, CPZ can not
induce shape changes (5). Ankyrin is bound to both band 3 and the
spectrin-actin oligomer (28-30).

At high concentrations, CPZ which is fairly lipophilic, will act as
a detergent. We propose that once the optimal receptor concentration
is surpassed, CPZ will begin to destabilize the membrane (Figures 3A,
B, C) and as the concentration is increased it can even solubilize the
hemoglobin off the lysed membrane (Figure 3C).

In conclusion, ESR and hypotonic hemolysis results suggest that CPZ
affects the erythrocyte membrane perhaps by interactions with the
cyto-Skeletal matrix. CPZ-Mel does not appear to affect membrane
structural properties. This lack of membrane perturbation could be
related to the inability of CPZ-Mel to cross the erythrocyte plasma
membrane. If CPZ does alter membrane structure via the cytoskeletal
matrix, then it iS possible that low dose treatment with CPZ could
ameliorate hemolytic disorders in which membrane rigidification

OCCUY‘S.

IO.

11.

12.
13.

14.
15.

16.

BIBLIOGRAPHY

Olaisen, B., and Oye, I. (1973) Eur. J. Biochem. 22, 112-116.

Suda, T., Maeda, N., Shimizu, D., Kamitsubo, E., and Shiga, T.
(1982) Biorheology 19, 555-565.

Sheetz, M.P., and Singer, S.J. (1974) Proc. Natl. Acad. Sci. USA
71, 4457-4461.

Jimbu, Y., Sato, S. Nakao, T., and Nakao, M. (1982) Biochem.
Biophys. Res. Commun. 104, 1087-1092.

Suda, T., Shimizu, D., Maeda, N., and Shiga, T. (1981) Biochem.
Pharmac. 30, 2057-2064.

Seeman, P. Kwant, W.0., and Sauks, T. (1969) Biochim. Biophys.
Acta 183, 499-511.

Seeman, P., Kwant, W.0., Sauks, T., and Argent, W. (1969) Biochim.
Biophys. Acta 183, 490-498.

Lund, A. (1980) Acta Pharmcaol. Toxicol. 47, 300-304.

Tenforde, T.S., Yee, J.P., and Mel, H.C. (1978) Biochim. Biophys.
Acta 511, 152-162.

Brunder, D.G., Coughlin, R.T., and McGroarty, E.J. (1981) Comput.
Biol. Med. 2, 9-15.

Janoff, A.S., Mazorow, D.L., Coughlin, R.T., Bowdler, A.J., Haug,
A., and McGroarty, E.J. (1981) Am. J. Hematol. 10, 171-179.

Elferink, J.G. (1977) Biochem. Pharmacol. 26, 2411-2416.

Dachary-Prigent, J., Dufourcq, J., Lussan, C., and Boisseau, M.

Romer, J., and Bickel, M.H. (1979) Biochem. Pharmac. 28, 799-805.

Ogiso, T., Iwaki, M., and Mori, K. (1981) Biochim. Biophys. Acta
649, 325-335.

Ogiso, T., Iwaki, M., and Sugiura, M. (1980) Chem. Pharm. Bull.
28, 3283-3290.

123

17.

18.

19.

20.

21.

22.
23.
24.

25.

26.

27.

28.

29.
30.

124

Gazitt, Y., Loyter, A., and Ohad, I. (1977) Biochim. Biophys. Acta.
471, 361-371.

Lacko, L., Wittke, B., and Lacko, I. (1980) Arzneim-Forsch. 30,
1852-1855.

LeFevre, P.G., and Sternberg, D.M. (1982) Biochem. Pharmac. 31,
463-466.

Baker, G.F., and Rogers, H.J. (1972) Biochem. Pharmac. 21,
1871-1878.

Baker, G.F., and Rogers, H.J. (1973) J. Physiol. (Land) 232,
597-608.

Levin, R.M., and Weiss, B. (1976) Mol. Pharmac. 12, 581-589.
Levin, R.M., and Weiss, 8. (1977) Mol. Pharmac. 13, 690-697.

Sobue, K., Muramoto, Y., Fujita, M., and Kakiuchi, S. (1981)
Biochem. Biophys. Res. Commun. 100, 1063-1070.

Kakiuchi, S., and Sobue, S. (1983) Trends in Biochem. Sci. 8,
59-62.

Hinds, T.R., and Andreasen, T.J. (1981) J. Biol. Chem. 256,
7877-7882.

Birchmeier, W., and Singer, S.J. (1977) J. Cell Biol. 73,
647-659.

Bennett, V., and Stenbuck, P.J. (1979) J. Biol. Chem. 254,
2533-2541.

Bennett, V., and Stenbuck, P.J. (1979) Nature 280, 468-473.

Liu, S.-C., and Palek, J. (1979) J. Supramolec. Struct. 10,
97-109.

CHAPTER VI

Concluding Statements

125

126

CONCLUSIONS

We have shown that addition of very small amounts of prostaglandins
(10'9 M to 10'12 M) can induce very large membrane structural
changes in intact erythrocytes. The number of molecules present vary
from 1 per cell (10"12 M) up to 1,000 per cell (10'9 M). It
is unlikely that the changes in membrane structure induced by the
prostaglandins are the result of just membrane lipid interactions. It
is more likely that prostaglandins have specific receptors which induce
the structural changes. In support of this thought, we found that the
presence of glucose prevented the structural changes induced by prosta-
glandin. With glucose present, the cells presumably were able to react
to the presence of prostaglandin and modulate its affects on membrane
structure. Upon addition of bicarbonate to phosphate buffer at either
pH 7.0, or pH 7.4, or to HEPES buffer, red cell membrane structural
changes were enhanced. Bicarbonate, by acting upon band 3, may accen-
tuate membrane structural changes. We believe that the membrane
structural changes are mediated by changes in the cytoskeletal matrix.
It appears that all the extrinsic proteins are linked either directly
or indirectly to each other and through specific interactions the com—
plex is bound to the membrane. Thus, any change in cytoskeletal‘
structure would obviate a change in overall membrane structure.

Since we believe that the cytoskeletal matrix is involved in the

structural changes seen upon addition of prostaglandins, we added

127

chlorpromazine (CPZ), a drug known to have Specific interactions with
lipids, to erythrocyte suspensions. The dose response curve of mem-
brane structural changes upon addition of CPZ obtained both by ESR and
by hypotonic hemolysis studies were in agreement. At 10'5 M. CPZ
induced maximal membrane disordering and maximal protection to hypo-
tonic hemolysis. We believe that upon further addition of CPZ, all the
sites at which CPZ is able to bind proteins became saturated and that
CPZ began to enter the lipid matrix. As the dose of CPZ was increased,
the drug bound to membrane lipid until, acting like a detergent, it
disrupted the membrane.

Prostaglandins and CPZ are not the only compounds which at thera-
peutic concentrations induce membrane structural changes. Upon inges-
tion of aspirin, the erythrocyte membranes from cells of both male and
female subjects underwent structural changes. The degree and time
course of the erythrocyte membrane changes for cells from women were
dependent on when in the menstrual cycle the erythrocytes were drawn.
The in vitro addition of aspirin induced no Significant changes in red
cell structure. Therefore, the changes in erythrocyte structure, seen
upon the ingestion of aspirin must be induced by metabolic products.
Aspirin, a prostaglandin synthetase inhibitor, may indirectly induce
the erythrocyte membrane changes by creating an inbalance in the ratio
of the various prostaglandins or by induction of compounds from the
leukotriene pathway. These changes in membrane structure may be the
result of a receptor directly or indirectly interacting with the
cytoskeletal matrix and thus resulting in an amplification of the
prostaglandin signal. Indomethacin, like aspirin, induced erythrocyte

structural changes which were dependent upon whether the cells were

128

drawn from men or women, and when in the menstrual cycle the cells were
drawn. Indomethacin, also an inhibitor of prostaglandin synthesis,
probably induces membrane structural changes via the same pathway as
does aspirin.

Finally, even an inbalance in the anionic milieu can apparantly
induce membrane alterations. Band 3, which is believed to be the anion
port, is attached to the cytoskeletal matrix via the ankyrin-Spectrin
complex. Thus, changes in anion flux or concentration gradients may
also induce membrane structural changes via interactions with the cyto-
skeletal matrix. HEPES, the largest of the anions studied, induced the
largest structural changes. The changes in membrane structure could be
the result of HEPES blocking anion flux and thus creating an inbalance
in the anionic gradient.

My results suggest that very small amounts of certain classes of
compounds can induce very large membrane structural changes. These
changes are most likely induced by an amplification of protein struc-
tural changes of the cytoskeletal matrix.

The most interesting questions raised by this thesis study are:

(1) Is the cytoskeletal matrix truly sensitive to these compounds and
do changes in the cytoskeletal matrix reflect changes in whole cell
structure?

(2) Are the induced membrane structural changes the result of an
influx of calcium?

(3) IS there a difference in the erythrocyte prostaglandin receptors
of men and women: Or, are the induced changes the result of differing

kinds and levels of steroid hormones.

129

(4) Can administration of the prostaglandins and/or phenathiazines
help regulate cell shape and viability in various hemolytic disorders?
(5) AS the erythrocyte ages, it undergoes various Shape changes. If
these shape changes can be regulated by the compounds discussed here or

other similarly acting agents, can better techniques be developed to

prolong blood shelf life?

APPENDIX

APPENDIX A

THE MODIFICATION OF HUMAN ERYTHROCYTE MEMBRANE STRUCTURE BY MEMBRANE
STABILIZERS: AN ELECTRON-SPIN RESONANCE STUDY*

ABSTRACT

Membrane structure in intact human erythrocytes was analyzed by
electron-spin resonance (ESR) spectroscopy. The spin probes 5-doxyl
stearate and 5-doxyl stearate methyl ester revealed thermally-induced
structural transitions in the membrane at 37°C and 15°C. The addition of
propranolol, diazepam, chlorpromazine, or Pluronic F68 all caused a
decrease in temperature of the upper transition, but did not markedly
alter the temperature of the lower transition. In addition, diazepam
caused a significant decrease in the ordering or packing of the
membrane-lipid acyl chains. It is proposed here that the protection from
hypotonic hemolysis that has been reported in the presence of these drugs
is mediated by a structural rearrangement in the erythrocyte membrane

involving a change in protein-lipid interactions.

 

*Published in American Journal of Hematology (1981) 19, 171-179.

130

131

INTRODUCTION

Present approaches to the treatment of hemolytic disorders are
limited to both in therapeutic principle and in effectiveness. They
depend mainly on the use of agents such as corticosteroids, for which the
mechanism of action is uncertain, or on the prolongation of survival of
damaged cells by Splenectomy. It is axiomatic that hemolysis results
ultimately from structural failure in the erythrocyte membrane: attempts
by Brewer (1) to diminish irreversible sickling by divalent ion
substitution suggest that erythrocyte resistance to hemolysis may be
modified therapeutically. The present study examines the hypothesis that
there are topographical sites in the membrane that are especially
critical with respect to hemolysis and that might be modifiable with
therapeutic effect.

The rapid renoval from the circulation of cells showing prehemolytic
damage mandates an indirect approach to the problem; consequently, we
have done an initial investigation in vitro on the structural
perturbations of the erythrocyte membrane induced by several agents
affecting the mechanical and osmotic resistance of red cells. It is
anticipated that this may ultimately lead to the identification of
susceptible membrane structures that are affected in common in diverse
hemolytic processes.

Osmotic fragility and mechanical deformability of human erythrocytes
are reportedly altered by low concentrations of a variety of
membrane-stabilizing drugs (2). Such drugs have the potential to protect
erythrocytes from hemolysis and to increase deformability in the

microcirculation, provided that ancillary effects in other systems can be

132

held within acceptable limits. There is at present insufficient
understanding of the action of these compounds to define the optimal
molecular form for erythrocyte-related effects. Although protein-lipid
interactions nay be affected, alterations in membrane architecture have
not been characterized in detail. One technique useful in studying
membrane architecture involves incorporating electron-spin resonance
(ESR) probes such as 5-doxyl stearate (50S) into the lipid bilayer. The
unpaired electron of the doxyl group absorbs microwave energy when the
sample is inserted into a magnetic field. The spectrum of the absorption
reflects the structure and fluidity of the membrane in the region of the
probe. Since 50S reportedly localizes in a lipid region that is closely
associated with proteins and devoid of cholesterol in the human
erythrocyte membrane (3,4), electron-spin resonance spectroscopy of
erythrocytes labeled with 505 Should reveal structural information
specifically related to protein-lipid interactions.

In this paper, we report that SDS- and S-doxyl stearate methyl ester '
(5DS-ME)-labeled human erythrocytes exhibit discontinuities in the
temperature dependence of ESR spectral parameters. These
temperature-induced changes in membrane structure correlate with other
reported temperature-dependent phenomena of erythrocytes that are
presuned to reflect membrane structural changes (5-9). We report that a
variety of membrane stabilizing agents affect the temperature dependence
of these measured Spectral parameters. Thus, membrane stabilization may
be the result of a structural rearrangement in the erythrocyte membrane

predominantly involving a change in protein-lipid interactions.

133

MATERIALS AND METHODS

Preparation of Cells

Blood was obtained from healthy individuals by venipuncture;
informed written consent was obtained from each donor. Anticoagulation
was effected by defibrination. Using standard procedures, the buffy coat
was removed and the cells were washed three times in 300 mOsm NaCl. To
renove serum albumin and other potential binding substances of the spin
probe, the reserve serum was filtered using an Amicon ultrafiltration
apparatus and Diaflo ultrafiltration membrane Um10 (molecular exclusion
10,000 daltons). Cells were washed once in serum ultrafiltrate and
packed in fresh ultrafiltrate to a hematocrit of 70 t 2 ml/dl. The cells

were analyzed within 18 hours of collection.

Electron-Spin Resonance Spectroscopy

 

The Spin probes 50S and SOS-ME (Syva Corp., Palo Alto, California)
were dissolved as a 30 mM solution in absolute ethanol. The labeling
procedure used was described previously (10), except that the ethanol was
evaporated prior to the addition of erythrocytes. Drugs were added
following labeling. To standardize drug concentrations with respect to
the stabilizing effect, concentrations were selected that caused
approximately equivalent degrees of submaximal protection against
hypotonic hemolysis (11,12). The final probe concentration was
approximately 0.4 umoles/mg of membrane protein. At this concentration,
505 has been reported to minimally perturb erythrocytes (13) and to

localize in lipid domains in close association with proteins (3,4).

134

All ESR studies were performed by standard methods as described
previously (10). The Spin labels used incorporate into the membrane in
such a manner that the unpaired electron of the doxyl radical is situated
close to, but shielded from, the aqueous phase. A typical spectrun of
erythrocytes labeled with SDS is shown in Figure 1. The distance between
the low-field and high-field microwave absorption peaks, 2T" (the
hyperfine Splitting parameter) reports the rotational mobility of the
probe and therefore the viscosity of the surrounding environment. With
increasing temperatures, intermolecular influences upon the unpaired
electron of the probe result in lower values of 2T" which reflect a
more fluid environment. The hyperfine splitting parameter 2T,I is
therefore directly related to the viscosity of the environment from which
the probe is reporting. High values of 2T" indicate rigid
environments, while low values of 2T" indicate more flexible
environments (14).

In studies reported here, the directly measured parameter, 2T1,
could be determined in SOS-labeled preparations above about 12°C. This
parameter is used along with 2T" to calculate the order parameter S
(14). The order parameter measures the deviation of the observed ESR
Signal from the case of a completely uniform orientation of the probe.
For a uniformly oriented sample, S = 1; for a random sample, S = 0 (14).

The hyperfine coupling constant also calculated from the directly
measured 2T" and 2T1 parameters is considered to reflect local
polarity (15) and thus reflects the position of the probe within the

membrane.

 

135

Data Analysis

The data (2T,I vs. temperature, S vs. temperature) were analyzed
in terms of linear components by an iterative least-squares program to be
described elsewhere (Brunder, D.G., Coughlin, R.T., McGroarty, E., in
press). Briefly, a B-spline (a piece-wise set of polynomials that are
smooth at the points of connection) was used to provide a fit for the ESR
data and points of inflection were used to group data. Regression lines
were calculated for each group and break points were determined. This
analysis has been shown in other membrane systems to permit
characterization of subtle changes (16). All ESR Spectral parameters
changed with temperature reversibly up to 48°C. For each set of data,
blood samples were examined from at least two healthy, nonsmoking

individuals.

RESULTS

Control Erythrocytes

ESR Spectra of SOS-labeled, intact erythrocytes Showed little or no
free probe in the supernatant, indicating that its site is predominantly
in the cellular phase (Figure 1). The Shape of the spectra indicated
that the majority of the fatty acid label was in a single environment
over the temperature interval examined (0-48°C). The hyperfine splitting
parameter, 2T", decreased in a discontinuous fashion as a function
of temperature, and a break point was determined to occur at 37°C (Figure
2a). Spectra recorded at cuvette temperatures above approximately 12°C
allowed the determination of the order parameter, S. When S was plotted

as a function of cuvette temperature, a similar discontinuity was

136

 

 

 

 

 

 

 

h—20900u—tl

 

 

 

 

 

 

Figure 1. Electron-spin resonance of human erythrocytes labeled with
5-doxyl stearate.

137

 

 

 

 

we

 

 

 

Figure 2. Hyperfine Splitting parameter, 2T. (Gauss), as a function
of temperature in erythrocytes labeled with S-doxyl stearaze in the
absence of perturbants (A), and in the presence of S x 10' M
propranolol (B). Arrows indicate the tranSition temperatures.

138

 

 

0.75-1 ~
0
0°
o.7o- ‘
‘1
S O
0.65“ ' '
' e
\ .
0.60- .
O ‘ .
26 5'0 4'0
TEMPERATURE °C
Figure 3. The temperature depence of the order parameter, S, in

erythrocytes labelled with 5-doxyl stearate in the absence of
perturbants (closed circles) and in the presence of 3 x 10-4 M

diazepam (open circles).

 

139

observed (Figure 3). The transition temperature determined with the
order parameter agreed quite well with that determined using 2T" and
presumptively indicates a structural change in the erythrocyte membrane.

A second Spin probe SOS-ME was also used to analyze erythrocyte
membranes. This uncharged probe causes less perturbation of the cells.
Spectra of SDS-ME-labeled cells allowed determination of the hyperfine
coupling constant, which indicated that the probe was in an environment
Similar to that of SOS-labeled cells at temperatures up to approximately
28°C. Above that temperature, the spectra changed in such a manner that
the spectral parameters were difficult to measure.

In spectra of SOS-ME-labeled cells recorded between 0°C and 28°C,
2T" was shown to decrease discontinuously with a break at
approximately 15°C (data not shown). This discontinuity is presumed to
reflect a second structural change in the erythrocyte membrane that
occurs at lower temperatures than the transition detected with 50S.
Therefore, it appears that erythrocyte membranes exhibit two thermotropic

transitions, one at 15°C and a second at about 37°C (Table 1).

Effects of Membrane Stabilizing Agents on Membrane Structure

To further characterize erythrocyte membrane-mediated phenomena, we
analyzed the changes in membrane structure induced by compounds known to
protect human erythrocytes from osmotic hemolysis, propranolol, diazepam,
chlorpromazine (11), and pluronic polyols (polyoxypropylene-poly-oxy-
ethylene condensates) (12). The addition of any of these drugs did not
appreciably alter the position of the Spin probes as determined by the
hyperfine coupling constant. Propranolol (Sigma Chemical Corp.) when

added to 5DS-labe1ed erythrocytes caused a slight decrease in the upper

140

transition temperature as measured using 2T" and 5 (Table 1). In
addition, a low temperature transition could be detected with 508 when
propranolol was added (Figure 2b). The temperature of the lower
transition detected with SDS-ME was not Significantly affected by the
addition of propranolol (Table 1).

Similar studies were carried out with labeled cells in the presence
of diazepam, chlorpromazine, or Pluronic F68. All of these agents
reduced the upper transition temperatures but did not appreciably alter
the lower transition temperature (Table 1). Furthermore, the addition of
diazepam or chlorpromazine (but not Pluronic F68) permitted the detection
of the low-temperature transition in SOS-labeled cells. In addition, the
presence of diazepam was Shown to significantly alter the order parameter
as indicated in Figure 3. The other perturbants induced only slight
changes in the order parameter. The addition of diazepam caused a
decrease in S by as much as 2% at low temperatures; changes greater than

1% in S are regarded as significant (17,18).
DISCUSSION

Numerous investigations employing a variety of techniques have
reported temperature-dependent changes in erythrocyte membranes.
Considering the diversity of approaches utilized, the temperatures at
which changes were found to occur are remarkably similar to those
reported here. Thus quenching of intrinsic tryptophan fluorescence by
spin labels showed discontinuities at 15°C and 35°C (4). Laser-raman
spectroscopy (5) and viscosimetry (6) revealed a discontinuity at

approximately 18°C. A discontinuity in the susceptibility of human

141

at 37°C (8). Finally, the preservation of membrane lipid asymmetry by
Mg+2 upon lysis has been shown to diminish above 18°C and disappear

at about 40°C (9). These data support the contention that temperature-
induced structural transitions occur in erythrocyte membranes and can be
detected using Spin-labeling techniques. Apparently, these transitions
occur in local areas not influenced by cholesterol.

Data presented here indicate that there are two thermally induced
structural transitions in hunan erythrocyte membranes that can be
detected using the ESR probes described above; one of these is found at
15°C, and the other at 37°C. These transitions are presumed to be
associated with changes in the conformation of either membrane proteins,
phospholipids, or both. Two independent studies, one involving quenching
of intrinsic tryptophan fluorescence (2,3), the other involving the
determination of binding affinities (19), support the contention that the
Spin probe 505 is closely associated with erythrocyte membrane protein.
The high-temperature transition demonstrated in this study, therefore,
appears to be associated with protein-lipid interaction.

In all cases, the addition of the membrane stabilizing drugs to the
spin-labeled erythrocytes predominantly affected the high but not the low
transition temperature. Since 50S reportedly localizes in close
proximity with membrane proteins and Since lipids interact more strongly
with proteins at elevated temperatures (20), it appears highly probable
that these agents perturb protein-lipid interactions. In fact,
stabilization of the erythrocytes by other drugs has been reported to
require an intact membrane protein structure (21).

Addition of propranolol, chlorpromazine, and diazepam (but not

Pluronic F68) to erythrocytes allowed for the determination of a

142

erythrocytes to benzyl-lysolecithin has been reported to occur at about
15°C (7). Transfer of phospholipid from hemagglutinating virus of Japan
to erythrocyte membrane was reported to begin at about 19°C and saturate
low-temperature transition using 505 and caused hemolysis at high
concentrations in isotonic saline (data not shown). Since Pluronic F68
did not cause hemolysis at high concentrations, it is probable that its
structural interaction with the membrane has unique features that may be
especially important to its effectiveness as an antihemolytic agent.
Diazepam was Shown to significantly alter membrane order. We are aware
of at least one instance in which a drug-induced decrease in erythrocyte
membrane order (17) can be correlated with a protein conformational
change (22).

Propranolol, diazepam, chlorpromazine, and Pluronic F68 have been
reported to protect erythrocytes against hypotonic hemolysis.
Preliminary evidence (manuscript in preparation) indicates that membrane
structural changes induced by some of these drugs are maximal at
concentrations that cause maximal protection to hypotonic hemolysis.
Thus, it is probable that these compounds affect specific membrane
domains such as sites of lipid-protein interaction and that these domains
are especially significant to cell fragility. The marked reduction of
intact cell osmotic fragility and the increase in mechanical
deformability that is brought about by an elevation in temperature, or by
the presence of perturbants such as the drugs used in this study
(17,23,24), might therefore be mediated by Specific membrane structural
changes. If the relevant changes, as suggested here, are related to
lipid-protein interactions in the erythrocyte membrane, there is the

possibility of developing more highly Specific agents to modify these

143

structures in a controlled fashion. Such agents might find wide
application in a variety of disease states in which red cell destruction

or erythrocyte perfusion in the microcirculation are critical factors.

CONCLUSIONS

The human erythrocyte membrane was Shown by biophysical probing
techniques to undergo structural changes at 15°C and 37°C. Addition of
membrane stabilizing drugs caused a decrease in the high-temperature
transition but did not appreciably alter the structural change detected
at lower temperatures. Furthermore, one of these drugs, diazepam, caused
a significant disordering of the membrane lipids. It is proposed that
these drugs are altering the membrane structure by changing the

lipid-protein interactions within the membrane.

REFERENCES

1. Brewer, G.J. (1979) Perspect Biol. Med. 22, 250.

2. Seeman, P. (1972) Pharmacol Rev. 24, 583.

3. Bieri, V.G. and Wallach, D.F.H. (1975) Biochim. Biophys. Acta 406,
415.

4. Bieri, V.G. and Wallach, D.F.H. (1976) Biochim. Biophys. Acta 443,
198.

5. Verma, s.1>. and Wallach, D.F.H. (1976) Biochim. Biophys. Acta 1111,
307.

6. Zimmer, G.H., Schirmer, H., and Bastian, P. (1975) Biochim. Biophys.
Acta 491, 244, 1975.

11.
12.
13.

14.

15.

16.

17.

18.

19.

20.

144

Weltzien, H.U., Arnold, B., and Kalkoff, H.G. (1976) Biochim.
Biophys. Acta 4§§, 56.

Maeda, T., Asano, A., Ohki, K., Okada, Y., and Ohnishi, S.-I. (1975)
Biochemistry 14, 3736.

Tanaka, K.-I. and Ohnishi, S.-I. Biochim. Biophys. Acta gag, 218.
Janoff, A.S., Haug, A., and McGroarty, E.J. (1979) Biochim. Biophys.
Acta §§§, 56.

Olaisen, B. and Oye, I. (1973) Eur. J. Pharmacol. g2, 112.
Gaehtgens, P. and Benner, K.U. (1973) Eur. J. Physiol. 44;, R1.
Bieri, V.G., Wallach, D.F.H., and Lin, P.S. (1974) Proc. Natl. Acad.
Sci. USA A, 4797.

Griffith, D.H. and Jost, P.C. (1976) In: Spin Labeling: Theory and
Applications, Berliner, L.J. (ed.), p. 454, Academic Press, New
Yam.

Hamilton, C.L. and McConnell, H.M. (1968) In: Structural Chemistry
and Molecular Biology, Rich, A. and Davidson, N. (eds.), p. 115,
W.H. Freeman, San Francisco.

Janoff, A.S., Coughlin, R.T., Racine, F.M., McGroarty, E.J., and
Vary, J.C. (1979) Biochem. Biophys. Res. Commun. g3, 565.

Kury, P.G., Ramwell, P.W., and McConnell, H.M, (1974) Biochem.
Biophys. Res. Commun. §§, 478.

Huestis, W.H. and McConnell, H.M. (1974) Biochem. Biophys. Res.
Commun. §1, 726.

Shiga, T., Suda, T., and Maeda, N. (1977) Biochim. Biophys. Acta
444, 231.

Mateu, L., Caron, F., Luzzati, V., and Billecoeq, A. (1970) Biochim.
Biophys. Acta §g§, 109.

21.

22.

23.
24.

145

Mizushima, Y., Sakai, S., Yamaura, M. (1970) Biochem. Pharmacol. _1_9,
227.

Meyer, M.B. and Swislocki, N.I. (1974) Arch. Biochem. Biophys. _1_6_4_,
544.

Mortensen, E. (1963) Acta Med. Scand. _1_7_3_, 693.

Murphy, J. (1969) J. Lab. Clin. Med. 14, 319.

   

  

RIES

1

 

”111111111111111111111111'1111111111