Investigative Ophthalmology & Visual Science, Vol. 29, No. 7, July 1988
Copyright © Association for Research in Vision and Ophthalmology
Characterization of Pigment Epithelial Cell Plasma
Membranes From Normal and Dystrophic Rats
Sharon C. Braunagel,* Daniel T. Organisciak, and Hih-Min Wang
Retinal pigment epithelial cell plasma membranes were isolated from the eyes of normal and RCSdystrophic rats by binding glass microbeads to the intact pigment epithelial cell layer, removal of the
bead-bound cells from the eyes and subsequent sucrose density gradient centrifugation. Plasma membranes were recovered from the gradients in identical yields and characterized by membrane marker
enzymes, lipid analysis and SDS-polyacrylamide gel electrophoresis. Membrane purification by alkaline phosphodiesterase I and 5'nucleotidase activities averaged 8-fold for normal rats and 5.5 for the
dystrophic rats. The ratio of cholesterol per microgram protein indicated 6 to 7-fold purification for
both types of plasma membranes. Na+K+-ATPase in the normal and mutant rat plasma membranes
was purified 5- and 3.5-fold, respectively, but the specific activities of both Na + K + -ATPase and
5'nucleotidase were higher in the dystrophic rat membranes than in normal. Subcellular organelle
contamination was low and relatively uniform in both types of membranes, while opsin contamination
was less than 1%. By electrophoretic analysis the plasma membrane proteins were similar, with 30-40
identifiable bands present in each membrane type. The plasma membranes both contain high levels of
cholesterol, sphingomyelin and phosphatidylcholine and low levels of polyunsaturated fatty acids.
However, the dystrophic rat membranes had significantly hgher levels of docosahexaenoic acid than
normal, and significantly lower levels of arachidonic acid. The differences in these plasma membrane
fatty acids and in the membrane-bound enzymes may affect the ionic balance of the interphotoreceptor
matrix or otherwise contribute to degenerative changes in dystrophic rat photoreceptors. Invest Ophthalmol Vis Sci 29:1066-1075,1988
Effective photoreceptor cell rod outer segment
(ROS) membrane turnover is normally accomplished
by the phagocytic removal of the shed ROS tips by
the retinal pigment epithelium (RPE). However,
phagocytosis, which is light-entrained in normal
rats,1 is impaired in the Royal College of Surgeons
(RCS) strain.2"5 RCS retinal-dystrophic rats have
been studied as a model for human retinal degenerations and to determine the effects of light and eye
pigmentation on ROS development, turnover and
vitamin A transport6"13 (for a review see ref. 14). It is
now firmly established, following studies with allophenic offspring of normal and mutant rats15 and
From the Department of Biochemistry, School of Medicine,
Wright State University, Dayton, Ohio.
* Present address: Department of Biology, Texas A & M University, College Station, Texas.
Supported by NIH grant bY-01959 (Bethesda, Maryland) and
the Stanley Petticrew research fund at Wright State University
(DTO). SCB was the recipient of a Sigma Xi grant-in-aid.
Submitted in part as a partial requirement for the PhD degree,
S. C. B. Wright State University, Dayton, Ohio.
Submitted for publication: September 17, 1987; accepted January 8, 1988.
Reprint requests: Dr. D. T. Organisciak, Department of Biochemistry, Wright State University, School of Medicine, Dayton,
OH 45435.
normal and mutant RPE cells in culture mixed with
ROS,1617 that the genetic defect in the RCS rat resides
in the RPE. Recently, the phenotypic expression of
the endocytotic defect has been further localized to
the RPE plasma membrane, where ROS binding
occurs, although ingestion is impaired.1819
Other abnormalities in the RCS rat RPE plasma
membrane also exist. For example, Na+K+-ATPase,
which is normally localized in the apical membrane
processes,20 redistributes in the plasma membrane
during cellular tight junction breakdown.21 Based on
studies with RPE cells maintained in tissue culture,
Clark and Hall22 demonstrated the incomplete glycosylation of a high molecular weight plasma membrane protein in mutant cells. Alterations in the distribution of plasma membrane cholesterol23'24 and in
the lipid composition of RPE cells from young dystrophic rats25 have also been reported. However, the
RPE plasma membrane has not yet been isolated
from rats and characterized by classical biochemical
methods. Using a glass bead technique originally developed for the isolation of bovine plasma membranes,26 we now report the isolation and partial purification of RPE plasma membranes from normal and
dystrophic rats. Our study shows that the plasma
membranes from RCS rats contain higher activities
of Na+K+-ATPase and 5'nucleotidase than normal,
1066
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
No. 7
RAT RPE CELL PLASMA MEMBRANES / Drounogel er QI.
and lipid abnormalities which may further impair
RPE function in these animals.
Materials and Methods
Animals, Maintenance, Breeding
Long Evans hooded rats weighing 75-100 g were
obtained from Harlan Industries (Indianapolis, IN)
and maintained in a 5 ft cd cyclic light environment
on Purina Rat Chow and water. RCS albino and pigmented rats, originally obtained from W. K. Noell,10
were bred and raised in our animal facility. On postnatal day 7 mothers with litters were moved from the
cyclic light environment to darkness to slow the progression of the retinal degeneration.10" Dystrophic
rats were used as available from the colony, and were
between the ages of 29-34 days. Normals were from
32-44 days of age at the time of use. For most enzyme determinations albino dystrophic rats and pigmented normals were used while both pigmented and
albino mutant rats were available for lipid analysis.
As determined for both the normal and mutant animals eye pigmentation had no effect on the individual determinations, but pigmented eyes were easier to
manipulate during tissue preparation. All animals
were dark-adapted for 18 hr before sacrifice in halothane-saturated chambers and all dissections were
performed in dim red light. The use of animals in this
investigation conformed to the ARVO Resolution on
the Use of Animals in Research.
RPE Plasma Membrane Isolation
Following enucleation in dim red light, the cornea
and lens were removed and the remaining eye cup
inverted over the conical end of a 0.5 ml polyethylene
microfuge tube (5 X 45 mm; Beckman, Inc., Fullerton, CA). The tube, with adherent eye cup, was then
placed into a phosphate-buffered saline solution (pH
7.2) at 4°C, containing 2 mM phenylmethyl-sulfonyl-fluoride, 2.5 mM iodoacetamide and 0.2%
EDTA (PBS-PIE). Although the adhesive forces holding the outer scleral surface to the plastic have not
been studied, the eye tissue-microfuge tube preparation remains intact and floating during the subsequent 30 min period in buffer. This preparation also
provides a convenient way to handle the small rat eye
without disrupting the RPE cell layer. Next, under
ambient light, the plastic tubes were placed over
small glass rods and the retinas gently teased away
from the RPE cell surface. To remove adhering ROS,
the inverted eye cups were then soaked for an additional 2 hr in cold PBS-PIE. Next, the RPE cells were
coated with tris-treated glass microbeads, 1-33 ^m
(Cataphote, Inc., Jackson, MS) by carefully turning
the plastic tubes in a petri dish containing the beads.
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
1067
The exposed surfaces of the glass beads were then
coated with 0.1% silane and the RPE cells, with adherent beads, carefully brushed from the eye cups
into 30 ml of PBS buffer. The RPE cell-bead preparation was then washed with PBS and treated with collagenase as described.26 Following a saline wash, a
whole cell lysate (whole cell) was prepared by treating
the cells with 2 ml ice-cold water. Cytoplasmic contaminants were removed with three saline washes and
the bead-bound RPE plasma membranes were then
suspended in 2 ml of 40% sucrose. To release plasma
membranes from the beads the preparation was vortexed vigorously for 1 min. Next, a discontinuous
gradient was prepared by layering the 40% sucrose
with 1 ml 32%, 0.5 ml 28% sucrose and water to a
final volume of 5 ml. All sucrose solutions were made
in 1 mM imidazole buffer pH 7.4 containing 0.1 mM
MgCb- The preparation was centrifuged at 104,000 g
for 1 hr at 4°C and the plasma membranes collected
from the 28% sucrose layer. The precipitated beadmembrane fraction (beads) was recovered and stored
in liquid nitrogen. The isolated plasma membranes
were then precipitated by mixing with 50 mg fresh
glass beads, diluting the sucrose to less than 10% with
water and centrifuging at 120,000 g for 60 min at
4°C. This last step facilitated the handling of the
small quantities of RPE plasma membranes, which
bind tightly to unsilanized glass. The precipitated
plasma membrane-glass bead pellet was collected and
normally stored in liquid nitrogen until use. In a typical experiment 60 eyes were used for enzyme analysis, while 240-300 eyes were used for the lipid determinations.
Membrane Purity and Composition
To determine plasma membrane purity, the following enzymatic markers were assayed: 5'nucleotidase, alkaline phosphodiesterase I and Na+K+-ATPase. Lysosomal contamination was measured by Nacetyl-j3-D-glucoseaminidase activity. Endoplasmic
reticulum and mitochondrial contamination were
measured by sulfatase C and cytochrome oxidase activity, respectively. RPE membrane marker enzyme
measurements have previously been described in detail.26 All enzyme assays were run at protein concentrations for which a linear response was determined.
With the exception of Na+K+-ATPase, no differences
in enzymatic activity were detected when fresh or
frozen membrane samples were used. As a result,
Na+K+-ATPase determinations were performed on
fresh RPE preparations. To determine ROS contamination opsin was measured with a radioimmunoassay
previously described by Plantner et al.27 Protein was
determined by the method of Lowry et al,28 with bovine serum albumin as the standard.
1068
INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / July 1988
Vol. 29
Table 1. Purification and yield for RPE plasma membranes from normal rats
Specific activity*
(relative specific activity)
Alkaline phosphodiesterase I
fig PNP/fig/hr
n=8
Fraction
Whole cell
0.33 ±0.13
5'Nucleotidase
nmol Pi/iig/hr
n=6
0.81 ±0.29
(1)
Plasma
membrane
Bead
2.70 ± 1.38 [8.2]f
(7.8)*
0.40 ±0.34 [1.2]
(1.2)
(1)
6.55 ±2.49 [8.1]
(8.1)
1.36 ± 1.02 [1.7]
(2.2)
N-Acetyl-fi-D-glucoseaminidase
fig/PNP/ng/hr
n=4
Whole cell
0.74 ± 0.28
(1)
Plasma
membrane
Bead
Sulfatase C
fluorescence
units/ng/hr
n=5
0.08 ±0.05 [0.1]
(0.12)
0.40 ± 0.04 [0.5]
(0.5)
(1)
—
(0.21 ±0.30)
—
(0.73 ± 0.66)
Na+lC-ATPase
nmol Pi/fig/hr
n=8
1.02 ±0.19
Average protein
ng/eye
n = 13
56.7 ± 25.2
(1)
4.92 ± 1.19 [4.8]
(5.0)
1.38 ±0.89 [1.4]
(1.4)
Cytochrome oxidase
nMO-Jvig/min
n=4
0.03 ±0.10
0.6 ± 0.3
9.0 ± 4.4
Opsin
ng/(ig
0.6
(1)
0.001 ±0.003 [0.03]
(0.0002)
0.01 ±0.01 [0.3]
(1.2)
ND§
—
* Results are the mean ± SD for the number n of experiments listed.
t Fold purity in brackets.
i (Relative specific activity) is defined as the total activity in a fraction/
total protein in the fraction. Whole cell activity is set to (1) for each enzyme.
§ ND: Not detectable.
Lipid Analysis
Whole cell, bead and plasma membrane fractions
were extracted with chloroform:methanol 2:1 and
washed according to Folch et al.29 Aliquots of the
lipid extracts were used for the preparation of fatty
acid methyl esters, thin layer chromatographic analysis of lipid composition and lipid phosphorus and
cholesterol determinations. Techniques for the analysis of membrane lipids have been described.121326
Plasma membrane purity was determined from the
ratio of cholesterol to protein in each fraction.
three plasma membrane and subcellular marker enzymes. The combined results of 13 separate membrane preparations for normal rats are contained in
Jable 1. Based on the specific activities for alkaline
phosphodiesterase I'and 5'nucleotidase in the whole
cell fraction, RPE plasma membranes from normal
rats were 8.2- and 8.1-fold purified, respectively.
Plasma membrane Na+K+-ATPase, however, was
only 4.8-fold purified. The relative specific activities
for the same enzymes were very similar: 7.8, 8.1 and
5.0 respectively. This indicates that a true purification of the plasma membranes has occurred. These
membranes contained an average of 11%, 7% and 4%
of the total alkaline phosphodiesterase I, 5'nucleotidase and Na+K+-ATPase activities, respectively. On
the average 0.6 ng protein/eye was recovered in the
plasma membrane fraction, or 1% of the total whole
cell protein. The bead fraction, from which the
plasma membranes were separated, contained about
16% of the protein found in the whole cell. By both
specific activity and relative specific activity, the purity of the bead fraction (row 3) was intermediate
between those of the whole cell and plasma membranes.
Contamination of the plasma membranes by RPE
subcellular organelles or ROS is shown in row 5 of
Table 1. The lysosomal enzyme N-acetyl-j8-D-glucoseaminidase was 10-fold lower in the plasma membranes than in the whole cell fraction. The mito-
Gel Electrophoresis
The basic technique of Laemmli30 was used for
SDS-polyacrylamide gel electrophoresis. A 4% stacking gel was used over a 10-15% gradient gel. Samples
were incubated in 3% SDS, 1% /8-mercaptoethanol,
0.05 M Tris-HCl (pH 6.8) and 15% glycerol for 60
min at 37°C. Following electrophoresis, the proteins
were fixed and stained with 0.2% Coomassie blue.
Gels were photographed and the negatives scanned
with an LKB-2202 Ultrascan laser densitometer
(LKB, Bromma, Sweden).
Results
Membrane Enzyme Marker Analysis
The purity of each plasma membrane preparation
was determined by measuring the activities of two or
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
No. 7
RAT RPE CELL PLASMA MEMBRANES / Drounogel er ol.
1069
Table 2. Purification of dystrophic rat RPE plasma membranes
Specific activity*
(relative specific activity)
Fraction
Whole cell
Plasma
membrane
Bead
Alkaline phosphodiesterase I
tig PNP/fig/hr
n=5
5'Nucleotidase
nmol Pi/ng/hr
n= 4
0.30 ±0.11
(1)
1.51 ±0.42[5.0]f
(5.6)*
0.24 ± 0.08 [0.8]
(0.9)
1.48 ±0.34
(1)
7.80 ± 1.33 [5.3]
(5.4)
2.27 ± 1.06 [1.5]
(1.6)
N-acetyl-fi-D-glucoseaminidase
tig/PNP/ng/hr
n=3
Whole cell
Plasma
membrane
Bead
0.38 ±0.12
(1)
0.08 ± 0.02 [0.2]
(0.22)
0.25 ±0.12 [0.6]
(0.6)
• Results are the mean ± SD for the number n of experiments listed.
t Fold purity in brackets.
% (Relative specific activity) is defined as the total activity in a fraction/
chondrial marker cytochrome oxidase was 30-fold
lower and sulfatase C, an endoplasmic reticulum enzyme, was 5-fold lower than in whole cell. Opsin was
0.6 ng//ig protein in the RPE whole cell fraction, but
was undetectable in the plasma membranes. For
completeness contaminating enzyme activities for
the bead fraction are also included in Table 1 (row 6).
As measured for the plasma membrane markers,
bead values were intermediate between the whole cell
and plasma membrane activities. However, unlike
the plasma membrane enzymes, subcellular marker
activities in the bead fraction were higher than in the
plasma membranes.
Enzyme purification data for the RPE plasma
membranes of RCS-dystrophic rats are given in Table
2. As measured by alkaline phosphodiesterase I and
5'nucleotidase these plasma membranes were 5.0and 5.3-fold purified from the whole cell; 5.6 and 5.4
by relative specific activity (row 2). Na+K+-ATPase
was 3.2- and 3.5-fold higher in the dystrophic rat
plasma membranes by specific and relative specific
activities. Lysosomal contamination of dystrophic rat
plasma membranes was 5-fold reduced from the
whole cell fraction (row 4). Sulfatase C was 10-fold
lower and cytochrome oxidase was undetectable in
the mutant rat RPE plasma membranes. Opsin measured 0.2 ng/jig in the whole cell fraction and 9 ng/^g
protein (0.2 pmol) in the plasma membrane fraction.
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
Sulfatase C
fluorescence
units/fig/hr
n=3
(1)
—
(0.10 ±0.09)
—
(1.08 ±0.37)
Na+K¥-ATPase
nmol Pi/pg/hr
n=4
2.24 ± 0.28
(0
7.26 ±2.11 [3.2]
(3.5)
1.74 ±0.67 [0.8]
(0.8)
Cytochrome oxidase
nMO2/ng/min
n= 1
Average protein
vg/eye
n = JO
55.3 ± 12.6
0.6 ± 0.3
7.3 ± 3.3
Opsin
ng/iig
0.02
(1)
ND§
0.2
0.01 [0.5]
—
—
9.0
total protein in the fraction. Whole cell activity is set to (1) for each enzyme.
§ ND: Not detectable.
The same marker enzyme activities in the dystrophic
rat bead fraction (rows 3 and 6) resembled more
closely those in the whole cell fraction than those in
the plasma membranes. Protein yields for the whole
cell, bead and plasma membrane fractions of the dystrophic rat RPE were nearly identical to the values
found for normal rats (Table 1).
A comparison of the plasma membrane markers
found in Tables 1 and 2 reveals that both 5'nucleotidase and Na+K+-ATPase activities in the dystrophic
rat fractions were higher than normal. For the whole
cell fractions the dystrophic rat values were nearly
two times higher. In the dystrophic rat plasma membranes 5'nucleotidase and Na+K+-ATPase were also
higher than in normal, despite the fact that the mutant rat membranes were not as pure as those from
normal rats. On the other hand, whole cell-alkaline
phosphodiesterase I activities were practically the
same. Alkaline phosphodiesterase in both the plasma
membrane and bead fractions from mutant rats was
lower than normal.
To determine if the higher activity of Na+K+-ATPase in the dystrophic rat RPE was an artefact of isolation or a result of ROS contamination, paired experiments with age-matched normal and RCS rats were
performed (Table 3). In experiments 1 and 2 the
whole cell activities for dystrophic rat RPE were 2.0and 1.7-fold higher than in normal. Experiment 3 is a
1070
Vol. 29
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1988
Table 3. Na+K+-ATPase specific activities in normal and dystrophic rat RPE and ROS (nmol Pi//*g protein/hr*)
Retinal pigment epithelium
Whole cell
Plasma membranes
Rod outer segments
Experiment
Normal
Dystrophic
Normal
Dystrophic
Normal
Dystrophic
1
2
3*
1.14
1.44
2.28
2.36
2.44
3.72 [3.3Jf
4.56 [3.2]
5.90 [2.6]
6.84 [2.9]
15.80 [6.5]
1.26
1.44
* Values are for individual membrane preparations for paired experiments
with normal and dystrophic rat RPE or ROS.
t Fold purification.
i Values obtained from pigmented dystrophic rats. Experiments I and 2
were obtained from albino rats.
result obtained for pigmented dystrophic rats. As
shown, whole cell Na+K+-ATPase activity for the
pigmented and albino dystrophic rats (exps. 1 and 2)
are nearly identical. Although not shown in Table 3,
5'nucleotidase activity in the pigmented dystrophic
rat whole cell fraction was 1.0 and alkaline phosphodiesterase I activity was 0.30. The dystrophic rat RPE
plasma membrane Na+K+-ATPase activity was also
higher than normal. In albino dystrophic rats, the
plasma membrane activities were 5.9 and 6.8 compared to 3.7 and 4.6 for the normals. For the preparation
of pigmented dystrophic rats Na+K+-ATPase activity
was 15.8, giving a 6.5-fold purification from the
whole cell activity. This is in good agreement with the
purification data for pigmented normals in Table 1.
Furthermore, these membranes were 11.9- and 10.2fold purified by 5'nucleotidase and alkaline phosphodiesterase I activities. Thus, for these mutant rat RPE
membranes Na+K+-ATPase purification was 55-64%
of the values for 5'nucleotidase and alkaline phosphodiesterase; this also agrees well with the respective
purifications for the same enzymes in Tables 1 and 2
(59-64%).
As shown by the ROS data in Table 3, the higher
Na + K + -ATPase activity in dystrophic rat RPE
plasma membranes is not from the small (<1%) contamination by opsin-containing ROS. The activities
for young normal and dystrophic rat ROS are similar
and lower than measured in either the whole cell or
plasma membrane fractions from dystrophic rats.
Similarly, Na+K+-ATPase activity in photoreceptor
cell debris from older dystrophic rats was only 2.18
nmol Pi/^g protein/hr. 5'nucleotidase activity in the
ROS of normal and mutant rats was 3.32 and 3.02
nmol Pi/^g/hr; it was 2.16 in the debris from RCS
rats. Taken together, these data suggest that the RPE
membrane enzyme activities in albino and pigmented dystrophic rats are the same, and that the
higher activities found in the plasma membranes of
pigmented animals probably relates to the ease of
tissue preparation rather than to a difference in endogenous activity.
Gel Electrophoresis
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
Figure 1 shows the SDS-gel electrophoretic protein
profiles for the respective normal and dystrophic rat
whole cell (lanes 1 and 2) and plasma membrane
fractions (lanes 3 and 4). The Coomassie-stained gels
are shown in panel A, and the density profiles in
panel B. In lanes 1 and 2 a number of bands ranging
in molecular weights between 97 and 14 kDa are
prominent. A number of higher molecular weight
proteins are not as well stained, but in each case the
same proteins were present in both the normal and
dystrophic rat fractions. In the normal, three low molecular weight proteins (14-17 kDa) were more intensely stained, while the dystrophic rat whole cell
fraction (lane 2) contained a doublet of about 95-97
kDa which stained more intensely than normal. A
similar 95-97 kDa protein, and a protein of about 67
kDa, stained prominently in the dystrophic rat
plasma membrane fraction (lane 4). Otherwise, the
plasma membrane proteins from normal and dystrophic rats were very similar. By Coomassie staining or
densitometry, 30-40 proteins were detectable, ranging in molecular weight from over 200 kDa down to
about 30 kDa. Concanavalin A-conjugated HRP
stained about 20 high molecular weight proteins in
transblots of RPE plasma membranes (data not
shown), confirming the presence of several glycoproteins of over 100 kDa.22
Analysis of Membrane Lipids
The lipid class compositions for the normal and
dystrophic rat RPE whole cell, plasma membrane
and bead fractions are contained in Table 4. Phosphatidylcholine (PC) and phosphatidylethanolamine
(PE) are the major lipids present in the RPE. Together they account for between 70-80% of the total
lipid phosphorous present in all fractions. Comparing
the whole cell lipid profiles, PC was about 45% and
nearly the same, while PE was almost 29% in dystrophic RPE compared to 23% in the normal. Sphingomyelin (SPH) was present in moderate amounts in
the normal and mutant RPE, 14% and 11% respec-
1071
RAT RPE CELL PLASMA MEMDRANE5 / Drounogel er o l
No. 7
B
3 4
Fig. 1. RPE proteins from
normal and dystrophic rat
whole cell (lanes 1 and 2) or
plasma membrane fractions
(lanes 3 and 4) were solubilized with 3% SDS and run
on 10-15% polyacrylamide
gels. Equivalent protein
concentrations were loaded
in lanes 1 and 2 (32 pig) and
in lanes 3 and 4 (20 /ag).
Visualization with Coomassie blue (Panel A); laser scanning densitometry
(Panel B).
94 67 "
43 30 "
20 —
20 '
14 -
practically the same as in the whole cell fractions.
Moderate levels of PE were present in each, but for
the dystrophic rat plasma membranes the decrease
from whole cell PE levels was over 7 mol%. PI was
present in lower concentrations (3.1-3.6%) in both
plasma membrane types, while PS was between 4-7
mol%. Lysophosphatidylcholine (LPC) represented
less than 2% of the lipid phosphorus in all fractions,
while phosphatidylglycerol and phosphatidic acid
each, were less than 1.0% (data not shown). The
molar ratio of cholesterol to lipid phosphorus was
2-3-fold higher in the plasma membranes than in
whole cell, but similar in both the normal and mutant
rat membranes.
The lipids of the normal and mutant rat bead fractions were also similar. However, in comparison to
tively, while both phosphatidylserine (PS) and phosphatidylinositol (PI) were lower (5-6%). The molar
ratio of cholesterol to lipid phosphorus was 0,4-0.3
(normal vs. RCS) in the whole cell fractions.
Based on the ratio of cholesterol/microgram protein present in the two types of plasma membranes
and their respective whole cell values, purification of
the normal rat membranes was 6.7, while the dystrophic rat plasma membranes were 6.2-fold pure. These
purifications are in good agreement with the values
obtained by enzyme marker analysis (Tables 1 and 2).
In most respects the lipid composition of the plasma
membranes resembled those of the RPE cell. As expected for plasma membranes, however, SPH and
cholesterol were higher than in the whole cell fractions. In both types of rat plasma membranes PC was
Table 4. Lipid composition of normal and dystrophic rat RPE (% lipid phosphorus*)
Whole cell
Normal
Dystrophic
(5)
(4)
Lipidf
PE
PS
PC
PI
SPH
23.2
5.8
44.6
5.7
14.2
±2.8
±0.9
±2.9
±0.9
± 1.6
28.9
5.5
46.5
5.0
10.7
Bead
Plasma membranes
± 1.9
±0.7
± 1.9
±0.6
± 1.2
Normal
(4)
25.3
4.4
44.0
3.1
20.4
±2.7
± 1.8
±3.8
±2.2
± 3.4
Dystrophic
Normal
Dystrophic
(2)
(5)
(4)
21.6 [20.4-22.7]
6.8 [4.9-8.1]
46.2 [43.8-48.6]
3.6 [3.4-3.8]
18.5 [17.9-19.1]
20.8 ±3.8
3.5 ± 1.2
53.7 ±3.6
6.1 ±0.7
10.4 ± 1.8
26.3
3.6
55,1
6.0
7.7
±3.0
± 1.4
±3.7
± 1.5
± 1.2
(4)
Chol/PO4
(mol/mol)
Choi/Protein
(nmol/Vg)
0.40 ±0.10
0.30 ± 0.04
0.80 ±0.10
0.85 ± 0.20
0.60 ±0.10
0.36 ± 0.04
0.06 ±0.01
0.08 ±0.01
0.40 ±0.10
0.50 + 0.11
0.04 ± 0.02
0.06 ± 0.04
* Results are the mean ± SD, [range] based on lipid phosphorus
for the number of preparations in parenthesis.
f Abbreviations: PE, phosphatidylethanolamine; PS, phosphati-
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
dylserine; PC, phosphatidylcholine; PI, phosphatidylinositol; SPH,
sphingomyelin, chol, cholesterol; PO4, lipid phosphorus.
1072
Vol. 29
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1988
Table 5. Fatty acid composition of normal and dystrophic rat retinal pigment epithelium* (Mol%)
Whole cell
Fatty Acidf
14:0
16:0
16:1
18:0
18:1
18:2
20:4
22:4
22:5u>6
22:5a>3
22:6
Normal
(5)
3.0
21.9
1.6
22.9
14.6
6.9
19.1
2.2
0.3
0.8
5.9
± 2.5
± 3.1
±0.7
± 2.6
± 1.0
± 0.7
± 1.0
± 0.7
± 0.3
± 0.7
± 1.9
Plasma membranes
Dystrophic
(4)
3.6
29.6
2.3
20.6
9.1
6.3
13.3
1.5
0.2
0.4
13.2
± 1.3
± 0.3
± 0.8
± 0.9
±0.9
± 0.4
±0.7
±0.4
±0.1
± 0.2
± 1.2
Normal
(4)
2.2
33.3
2.8
24.6
14.5
3.9
12.8
1.6
0.1
0.5
2.9
± 1.3
± 1.7
± 0.8
± 2.3
±2.1
± 1.2
±2.4
± 1.8
±0.1
± 0.6
± 0.6
Bead
Dystrophic
(4)
Normal
(5)
Dystrophic
(4)
5.8 ± 1.8
38.4 ± 2.5
4.9 ± 1.4
19.0 ± 1.1
11.6± 1.5
5.7 ± 1.2
5.4 ± 1.4
3.1 ±0.3
—
0.4 ±0.1
6.0 ± 2.4
0.9 ± 0.3
26.4 ± 1.7
1.9 ± 0 . 4
22.5 ± 1.0
16.3 ± 0 . 8
8.0 ± 0.6
17.2 ± 0 . 9
1.3 ± 0 . 7
0.8 ± 0.2
0.6 ± 0.4
3.6 ± 0.3
2.8 ± 1.1
31.5 ± 1.6
2.8 ± 1.2
20.2 ± 0.7
9.6 ± 0.5
6.8 ±0.7
13.7 ±0.6
0.8 ±0.1
0.3 ±0.1
0.5 ± 0.2
11.0± 1.3
• Results are the mean ± SD for the number of separate preparations in
parenthesis.
t Abbreviations stand for: chain length: number of double bonds; 22;5&i6
or w3 identify the first double bond from the methyl end for this pair of equal
chain length fatty acids.
the whole cell and plasma membrane fractions, PC in
the bead fractions was higher, almost 55%, while SPH
was much lower. Overall, the bead fractions were not
purified from the whole cell lipids and the cholesterol:phosphate ratio was only slightly higher than for
the RPE cells.
The fatty acid profiles for the normal and dystrophic rat membrane fractions are contained in Table
5. In the whole cell fractions the major saturated fatty
acids were palmitic (16:0) and stearic (18:0) acids,
which together represented 45-50% of the total. Palmitic acid was almost 8% higher in the dystrophic rat
RPE, while 18:0 was nearly equivalent in the two
whole cell fractions. The major unsaturates were arachidonic acid (20:4), oleic acid (18:1) and docosahexaenoic acid (22:6). Although the levels of 20:4 and
18:1 were lower in dystrophic rat whole cell lipids,
22:6 was two times higher than normal (13.2 vs.
5.9 mol%).
In the RPE plasma membranes of normal rat 16:0
was 33 mol%; it was 38% in the plasma membranes
from dystrophic rats. In both types of plasma membranes 18:0 levels were similar to those found for
whole cell. Arachidonic acid was substantially lower
in both types of membranes, but it was only 5.4 mol%
in the dystrophic rat samples compared to 12.8% in
normal. Similarly, 22:6 was about 50% lower in the
plasma membranes than in the respective whole cell
fractions. In the normal, 22:6 was less than 3%, however, it was 6 mol% in the dystrophic rat RPE plasma
membranes. A comparison of the fatty acid compositions of the two bead fractions shows that they resemble the whole cell fatty acids more closely than those
of the plasma membranes. At the same time, the
levels of many of the major fatty acids (eg, 16:0, 18:0,
20:4, 22:6) in the beads were intermediate between
those of the other two fractions.
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
Discussion
This study shows that the RPE plasma membrane
of RCS rats is characterized by higher than normal
specific activities of the marker enzymes Na + K + ATPase and 5'nucleotidase and by abnormalities in
the contents of the 22:6 and 20:4 fatty acids. The
same enzyme and lipid differences were present in the
whole cell fractions of albino and pigmented RCS
rats, suggesting that the plasma membrane abnormalities are not an artefact of isolation. Several additional lines of evidence also support these findings.
First, not all plasma membrane enzymes were affected to the same extent. Alkaline phosphodiesterase
I activity was the same in the whole cell fractions of
both normal and mutant rats. Despite the different
purifications of the normal and dystrophic rat plasma
membranes (8- and 5-fold, respectively), if phosphodiesterase activities are calculated at equivalent
membrane purities then they would also be very similar. The same applies to the two bead fractions from
which the plasma membranes were isolated (Tables 1
and 2). Second, the protein yields for the various
normal and dystrophic rat membrane fractions were
similar. This indicates that differential membrane
losses probably did not occur during the isolations
and that plasma membrane release from the beads
was nearly equivalent. Based on a consideration of
the protein yields and enzyme recovery data we estimate that between 10-25% of the total theoretical
plasma membrane protein was recovered in our samples. Thus, it is unlikely that a small unrepresentative
fraction of the RPE plasma membranes was isolated.
No. 7
RAT RPE CELL PLASMA MEMBRANES / Braunagel er ol.
Furthermore, contamination by subcellular organelles was relatively low and consistent between the
two RPE plasma membrane types. Finally, Na+K+ATPase and 5'nucleotidase activities in the ROS or
debris from dystrophic rats were lower than in the
RPE fractions (Table 3). Thus, at a contamination of
less than 1% opsin in the dystrophic rat RPE plasma
membranes, an increase in specific activity of these
marker enzymes from ROS would not be expected.
One obvious possibility to explain the elevated activities of plasma membrane enzymes in our dystrophic rat preparations is that they may contain more
apical RPE membranes than normal. Certainly,
judging by the SDS-gel protein profiles in Figure 1,
the more intense staining of bands in the 95-97 kDa
region of the dystrophic rat lanes indicates that these
proteins are present in higher concentrations than
normal. In addition, although we have not further
identified the 97 kDa protein present in our gels, the
catalytic subunit of Na+K+-ATPase has a molecular
weight of 97 kDa.31 However, other evidence argues
against a preferential purification of apical processes
from the dystrophic rat plasma membranes. In both
types of RPE the known apical membrane marker
Na+K+-ATPase20 was purified to about 60% of the
other plasma membrane enzymes. Furthermore, the
activity of 5'nucleotidase, an enzyme that appears to
be localized over both the apical and basal surfaces of
the RPE cell,32 was also elevated in the dystrophic rat
fractions. This suggests that our plasma membranes
contain both the apical and basal components of the
RPE cell surface, and that the adherence of apical
membranes to the glass beads was equivalent and
somewhat greater than for the basolateral membranes. Clearly, measurements of a known basal
membrane marker, such as the receptor for serum
retinol binding protein,33 would allow a better estimate of the apical and basolateral membrane content
of our preparations.
A second possibility is based on the finding that
Na+K+-ATPase redistributes in the RPE plasma
membrane of dystrophic rats during the breakdown
of tight junctional complexes.21 As such Na + K + ATPase would also be expected to be present in basolateral membranes, or in the additional apical plasma
membrane invaginations seen in the RPE of young
mutant rats.21 Changes in the lipid composition or
fatty acids of the RPE membrane might also facilitate
the lateral migration of membrane bound enzymes.
Previous work shows that the binding of filipin and
digitonin by cholesterol23*24 and lipid composition25
are abnormal in young dystrophic rat RPE. In addition, Na+K+-ATPase is more active in fluid membrane regions than in highly ordered regions,34 or in
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
1073
membranes containing high concentrations of cholesterol.35 Na+K+-ATPase also reportedly binds several hundred phospholipid molecules and is stimulated by PS.36 While our data suggest that major differences in the phospholipid head groups and
cholesterol contents of the normal and dystrophic rat
RPE plasma membranes do not exist, the regional
localization of these lipids and their respective fatty
acid compositions is currently unknown.
We did find remarkable differences in two major
polyunsaturated fatty acids in the dystrophic rat RPE
plasma membranes; 22:6 was higher than normal,
while 20:4 was lower than normal (Table 5). These
lipid differences appear to be a result of the lack of
phagocytosis in RCS rats. Based on a consideration of
the opsin contents in the various fractions, a small
amount of extracellular photoreceptor cell debris
may adhere to or have been trapped in the plasma
membranes of the dystrophic rats. Whether this
amount of contamination is sufficient to account for
the fatty acid changes or not, the close association of
the RPE and the photoreceptor debris in RCS animals 641 ' suggests that lipid exchange between the two
membranes can occur. For example, it is known that
debris contains different levels of 22:6, PE and cholesterol than normal rat ROS membranes.12 In addition to the possible effect of lipid exchange (22:6) on
Na+K+-ATPase activity in the RPE plasma membrane, the level of 20:4 was much lower in all membrane fractions from mutant rats. In very young dystrophic rats 20:4 appears to be normal in the RPE,25
while in debris from old dystrophic rats it is elevated.12 Whatever the reason for the loss of 20:4 from
the dystrophic rat RPE, this evidence suggests that
arachidonic acid metabolism in the mutant cells may
be abnormal.
Finally, the elevation of RPE enzyme activity in
RCS rats may be an attempt to compensate for the
primary genetic defect by a process of up regulation
in existing RPE membranes, or by the actual proliferation of the RPE apical microvilli. Up regulation of
Na+K+-ATPase has previously been demonstrated in
the epithelium of kidney cortical collecting tubule37
and in the epithelium of colon.38 Furthermore, Miller
et al39 concluded that the electrogenic nature of
Na+K+-ATPase and its location in the apical RPE
membrane are likely to affect the transport of ions,
fluids and metabolites across the RPE cell layer.
Therefore, an increase in RPE membrane enzyme
activity, for whatever reason, could alter the ionic
milieu of the interphotoreceptor cell matrix. In addition, 5'nucleotidase, through the release of adenosine,
may play a role in the regulation of Na+ ion permeability in ROS.40 Recently, Irons41 has shown that a
1074
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / July 1988
Mn++-dependent pyrimidine 5'nucleotidase redistributes between RPE and ROS during shedding of the
rod tips. Although this enzyme and the 5'nucleotidase
activity measured in our study are not the same, the
accumulation of abnormally high levels of otherwise
normal enzyme activities in dystrophic RPE could
affect these cells as well as the interphotoreceptor
matrix. The added burden of the diffusional barrier
imposed by the accumulation of photoreceptor cell
debris as dystrophic rats age11 may also explain some
of the pathological changes seen in the RPE of older
animals.
The mechanism by which the well described ROS
degeneration occurs in RCS rats remains an open
question. However, Mullen and LaVail's15 observation that patches of normal RPE in chimeric rat eyes
somehow modifies the destructive process in adjacent
ROS facing dystrophic RPE cells lends support to the
idea that a diffusable process is involved in the rod
cell degeneration (see also ref. 14). In this regard it is
interesting that retinol esterification, a process
thought to be associated with the RPE plasma membrane,42 is abnormally low in dystrophic rat RPE.43
Furthermore, light exposure910 and the maintenance
of RCS rats on a vitamin A-deficient diet" each accelerate the destruction of ROS. The extent to which
vitamin A metabolism is associated with the degenerative process in dystrophic rat photoreceptor cells or
RPE remains to be determined.
Key words: retinal pigment epithelium, rat, plasma membranes, retinal dystrophy
Acknowledgments
We thank Dr. James Plantner for the opsin RIA analysis
and Dr. I. M. Leffak for the laser densitometry.
References
1. La Vail MM: Rod outer segment disk shedding in rat retina:
Relationship to cyclic lighting. Science 194:1071, 1976.
2. Herron WL, Riegel W, Myers OE, and Rubin ML: Retinal
dystrophy in the rat—a pigment epithelial disease. Invest Ophthalmol 8:595, 1969.
3 Bok D and Hall MD: The role of the pigment epithelium in the
etiology of inherited retinal dystrophy in the rat. J Cell Biol
49:664, 1971.
4. La Vail MM, Sidman RL, and O'Neil D: Photoreceptor pigment epithelial cell relationships in rats with inherited retinal
degeneration. J Cell Biol 53:185, 1972.
5. La Vail MM: Assignment of retinal dystrophy (rdy) to linkage
group IV of the rat. J Hered 72:294, 1981.
6. Dowling JE and Sidman RL: Inherited retinal dystrophy in the
rat. J Cell Biol 14:73, 1962.
7. Noell WK: Aspects of experimental and hereditary degeneration. In First Symposium on the Biochemistry of the Retina,
Graymore GN, editor. London, Academic Press, 1965, pp.
51-72.
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
Vol. 29
8. Herron WL, Riegel BW, and Rubin ML: Outer segment production and removal in the degenerating retina of the dystrophic rat. Invest Ophthalmol 10:54, 1971.
9. Noell WK: Hereditary retinal degeneration and damage by
light. Estratto dagli Atti del Simposio di Oflatmologia Pediatrica Parma, Italy, 1974, pp. 322-329.
10. La Vail MM and Battelle BA: Influence of eye pigmentation
and light deprivation on inherited retinal dystrophy in the rat.
Exp Eye Res 21:167, 1975.
11. Delmelle M, Noell WK, and Organisciak DT: Hereditary retinal dystrophy in the rat: Rhodopsin, retinol, vitamin A deficiency. Exp Eye Res 21:369, 1975.
12. Organisciak DT and Noell WK: Hereditary retinal dystrophy
in the rat: Lipid composition of debris. Exp Eye Res 22:101,
1976.
13. Organisciak DT, Wang H-M, and Kou AL: Rod outer segment
lipid-opsin ratios in the developing normal and retinal dystrophic rat. Exp Eye Res 34:401, 1982.
14. La Vail MM: The retinal pigment epithelium in mice and rats
with inherited retinal degeneration. In the Retinal Pigment
Epithelium, Zinn KM and Marmor MF, editors. Cambridge,
Harvard University Press, 1979, pp. 357-380.
15. Mullen RJ and La Vail MM: Inherited retinal dystrophy: Primary defect in pigment epithelium determined with experimental rat chimeras. Science 192:799, 1976.
16. Edwards RB and Szamier RB: Defective phagocytosis of isolated rod outer segments by RCS rat retinal pigment epithelium in culture. Science 197:1101, 1977.
17. Goldman AI and O'Brien PJ: Phagocytosis in the retinal pigment epithelium of the RCS rat. Science 201:1023, 1978.
18. Chaitin MH and Hall MO: Defective ingestion of rod outer
segments by cultured dystrophic rat pigment epithelial cells.
Invest Ophthalmol Vis Sci 24:812, 1983.
19. Chaitin MH and Hall MO: The distribution of actin in cultured normal and dystrophic rat pigment epithelial cells during
the phagocytosis of rod outer segments. Invest Ophthalmol Vis
Sci 24:821, 1983.
20. Ostwald TJ and Steinberg RH: Localization of frog retinal
pigment epithelium Na+K+-ATPase. Exp Eye Res 31:351,
1980.
21. Caldwell RB and McLaughlin BJ: Redistribution of Na-KATPase in the dystrophic rat retinal pigment epithelium. J
Neurocytol 13:895, 1984.
22. Clark VM and Hall MO: RPE cell surface proteins in normal
and dystrophic rats. Invest Ophthalmol Vis Sci 27:136, 1986.
23. Caldwell RB and McLaughlin BJ: Freeze-fracture study of filipin binding in photoreceptor outer segments and pigment epithelium of dystrophic and normal retinas. J Comp Neurol
236:523, 1985.
24. Caldwell RB: Filipin and digitonin studies of cell membrane
changes during junction breakdown in the dystrophic rat retinal pigment epithelium. Curr Eye Res 6:515, 1987.
25. Batey DW, Mead JF, and Eckhert CD: Lipids of the retinal
pigment epithelium in RCS dystrophic and normal rats. Exp
Eye Res 43:751, 1986.
26. Braunagel SC, Organisciak DT, and Wang H-M: Isolation of
plasma membranes from the bovine retinal pigment epithelium. Biochim Biophys Acta 426:183, 1985.
27. Plantner JJ, Hara S, and Kean EL: Improved, rapid radioimmunoassay for rhodopsin. Exp Eye Res 35:543, 1982.
28. Lowry OH, Rosenbrough NJ, Fair AL, and Randall RJ: Protein measurement with the Folin phenol reagent. J Biol Chem
193:265, 1951.
29. Folch J, Lees M, and Sloane-Stanley GH: A simple method for
No. 7
30.
31.
32.
33.
34.
35.
36.
I W RPE CELL PLASMA MEMBRANES / Brounogel er QI.
the isolation and purification of total lipids from animal tissues. J Biol Chem 226:497, 1957.
Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680,1970.
Schwartz A, Lindenmayer GE, and Allen JC: The sodium
potassium adenosine triphosphatase: Pharmacological, physiological and biochemical aspects. Pharmacol Rev 27:3, 1975.
Kreutzberg GW and Hussain ST: Cytochemical localization of
5'nucleotidase activity in retinal photoreceptor cells. Neuroscience 11:857, 1984.
Heller J: Interactions of plasma retinol binding protein with its
receptor: Specific binding of bovine and human retinol binding protein to pigment epithelium cells from bovine eyes. J
Biol Chem 250:3613, 1975.
Barnett RE and Palazzotto J: Mechanisms of the effects of lipid
phase transitions on the Na+K+-ATPase and the role of protein
conformational changes. NY Acad Sci 242:67, 1974.
Yeagle P: Cholesterol modulation of (Na+K+)-ATPase hydrolyzing activity in the human erythrocytes. Biochim Biophys
Acta 727:39, 1983.
Shichi H: Properties of biomembranes. In Biochemistry of Vision, Shichi H, editor, Chapter 2. London, Academic Press,
1983, pp. 24-35.
Downloaded From: http://iovs.arvojournals.org/ on 06/15/2017
1075
37. Petty KJ, Kokko JP, and Marver D: Secondary effect of aldosterone on Na+K+-ATPase activity in rabbit cortical collecting
tubule. J Clin Invest 68:1514, 1981.
38. Schultz SG: A cellular model for active sodium absorption by
mammalian colon. Ann Rev Physiol 46:435, 1984.
39. Miller SS, Steinberg RH, and Oakley B: The electrogenic sodium pump of the frog retinal pigment epithelium. J Memb
Biol 44:259, 1978.
40. Bounds D and Brodie AE: Light-sensitive swelling of isolated
frog rod outer segments as an in vitro asay for visual transduction and dark adaptation. J Gen Physiol 66:407, 1975.
41. Irons M: Redistribution of Mn++-dependent pyrimidine 5'-nucleotidase (MDPNase) activity during shedding and phagocytosis. Invest Ophthalmol Vis Sci 28:83, 1987.
42. Ottonello S, Petrucco S, and Maraini G: Vitamin A uptake
from retinol-binding protein in a cell-free system from pigment
epithelial cells of bovine retina. J Biol Chem 262:3975, 1987.
43. Berman ER, Segal N, Photiou S, Rothman H, and FeeneyBurns L: Inherited retinal dystrophy in RCS rats: A deficiency
in vitamin A esterification in pigment epithelium. Nature
293:217, 1981.