Quantitation of Na/K ATPase Pump Sites in the
Rabbit Corneal Endothelium
Dayle H. Geroski and Henry F. Edelhauser
In these experiments, the binding of 3H • ouabain, a specific inhibitor of Na/K ATPase, was used to
quantitate the density of Na/K ATPase pump sites in the rabbit corneal endothelium. The uptake of
ouabain by the corneal endothelium shows two components: one that saturates at a ouabain concentration
near 2 X 10~7 M (specific binding), and one component that increases linearly with increasing glycoside
concentration (nonspecific uptake). The nonspecific uptake can be accounted for by that ouabain equilibrating with the extracellular space, which, estimated by inulin space, amounts to 13.0 nl/mm2 of
endothelium. The saturable component of endothelial ouabain uptake is displaced by K+ ions, which
is consistent with this fraction being bound to Na/K ATPase. Maximal endothelial ouabain binding
was measured as 20.7 fmoles/mm2 of endothelium, which corresponds to 3.0 X 106 pump sites per
cell. The density of Na/K ATPase pump sites in the rabbit corneal endothelium is comparable to
densities reported for several transporting epithelia. These data are consistent with the known function
of the endothelium in corneal deturgescense and corroborate the importance of Na/K ATPase in
endothelial fluid transport. Invest Ophthalmol Vis Sci 25:1056-1060, 1984
hibition curve for endothelial Na/K ATPase9 Moreover, the effects of ouabain on both corneal deturgescense and Na/K ATPase are both biphasic with low
concentrations of ouabain, enhancing enzyme activity
and stimulating corneal deswelling as described by Anderson et al.10
Electrophysiologic studies have further corroborated
the importance of endothelial Na/K ATPase to pump
function. Fischbarg1' has shown ouabain to inhibit the
transendothelial potential difference, but not to affect
the electrical resistance (barrier function) of the endothelium. Finally, the active sodium flux across the
endothelium has recently been measured by Huff and
Green12 and Lim.13 This net active sodium flux also
is inhibited by ouabain.13
Since it seems quite likely that endothelial pump
function is linked to the activity of Na/K ATPase,
these experiments were undertaken to further quantitate this enzyme in the rabbit corneal endothelium.
In these studies the specific binding of 3 H • ouabain by
the corneal endothelium was investigated to determine
the density of Na/K ATPase "pump sites" in this corneal layer.
Current evidence indicates that pump function of
the corneal endothelium is mediated by the active
transport of sodium and bicarbonate ions across the
endothelium and into the aqueous humor.' As in other
transporting tissue, Na/K ATPase seems to be an essential component of this transport process. Since this
enzyme first was measured in corneal endothelium by
Bonting et al,2 considerable evidence has accumulated,
which implicates it as a necessary component of endothelial fluid transport.
Brown and Hedbys3 and Langham and Kostelnik4
demonstrated that ouabain, a specific inhibitor of Na/
K ATPase,5 prevented corneal temperature reversal of
enucleated eyes. Ouabain also was shown to cause corneal edema when injected into the aqueous humor of
rabbit eyes in vivo.6 Trenberth and Mishima7 have
shown that isolated, ouabain-perfused, rabbit corneas
swell in a dose-dependent manner due to pump-function inhibition. We recently have confirmed these
findings in the human cornea as well.8
The dose-response curve for ouabain-induced corneal swelling is strikingly similar to the ouabain inFrom the Departments of Ophthalmology and Physiology, Medical
College of Wisconsin, Milwaukee, Wisconsin.
Supported in part by National Eye Institute Research Grant EY00933 and Ophthalmic Research Core Grant EY-01931. Dr. Edelhauser is a Research to Prevent Blindness, Olga K. Wiess Scholar.
Submitted for publication: November 10, 1983.
Reprint requests: Dayle H. Geroski, PhD, Department of Physiology, Medical College of Wisconsin, 8701 Watertown Plank Road,
Milwaukee, WI 53226.
Materials and Methods
Ouabain Binding Studies
3
H-ouabain (14.0-18.0 Ci/mmole) was obtained
from New England Nuclear (Boston, MA). The ouabain was supplied in an ethanol-benzene solution which
1056
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933349/ on 06/17/2017
No. 9
1057
N A / K ATPASE PUMP SITES IN CORNEAL ENDOTHELIUM / Geroski and Edelhauser
was evaporated to dryness under nitrogen, and the
isotope was redissolved in deionized water. High performance liquid chromatography (HPLC) was used to
determine the radiochemical purity of the 3 H • ouabain.
A C18-Mm Bondapak (10-/mi particle diameter) column was eluted at a flow rate of 1.0 ml/min (1.0 ml
fractions) using 84% H2O:16% acetonitrile as the mobile phase. Ultraviolet absorbance was monitored at
220 nm. The 3 H • ouabain, which had a retention time
of 7 min in this system, was found consistently to have
a radiochemical purity greater than 95% (Fig. 1).
Eyes of 2-3-kg New Zealand albino rabbits were
enucleated. The corneal epithelium was removed by
scraping with a Gill knife, and the corneas were excised
with a 1-2-mm scleral rim. Each deepithelialized cornea was incubated at 37°C in a shaking waterbath
using 3.5 ml of a K+-free bicarbonate Ringer's incubation solution of the following composition (g/liter):
NaCl: 6.801; CaCl 2 -2H 2 O: 0.153; MgCl2 • 6H 2 O,
0.158; NaH 2 PO 4 :0.103; NaHCO 3 :2.453; and glucose:
0.903. In additional experiments designed to study the
effects of K + on ouabain binding, 16-mM KC1 was
substituted for an equivalent amount of NaCl. Carrierfree 3 H • ouabain was added to the incubation media
in concentrations increasing to 5 X 10~7 M. Concentrations greater than 5 X 10~7 M were achieved by
adding the appropriate amounts of cold ouabain to 5
X 10"7 M carrier-free 3H« ouabain. To estimate extracellular space 1.0 AtCi of 14C inulin (specific activity
3 mCi/g) was added to each of the initial 3.5-ml incubations. In a final group of experiments, 10~4 M
cold ouabain was added to the appropriate concentrations of 3 H • ouabain in order to distinguish specific
from nonspecific binding.
Since initial incubations showed that steady-state
ouabain uptake was achieved in 2.5-3 hr (Fig. 2), a
3-hr incubation period was chosen for subsequent
studies. At the end of this incubation, the corneas were
removed from the incubation medium, briefly drained,
and rinsed three times (10 sec each) in ice-cold, ouabain-free, incubation medium. An 8-mm button of
central cornea then was trephined out, and the endothelium plus Descemets was stripped away carefully
as an intact sheet from the overlying stroma. Each
cornea thus provided one sheet of endothelium and
one sheet of stroma, each of which was dissolved in
1.0 ml of Protosol. The uptake of 3 H and 14C was
measured in each 8-mm diameter sheet using scintillation spectroscopy.
15
25
FRACTION
Fig. 1. High performance liquid chromatograph of 3H • ouabain.
A Cl 8-^m Bondapak column was eluted at 1 ml/min using a mobile
phase of 84% H2O: 16% acetonitrile. UV absorbance was monitored
at 220 nm. Radiochemical purity of 3H« ouabain was consistently
greater than 95%.
a PRO eye bank microscope. Endothelial cells in five
to ten 0.03 mm 2 fields of each cornea were counted.
Mean (±SEM) cell density was determined as 4126
± 1 1 8 cells/mm2 (n = 10 corneas).
The use of animals in these experiments conforms
to the ARVO Resolution on the Use of Animals in
Research.
5.0
4.0
LJ
£ Q)
< O
g e z.o]
1.0(14)
v C
5 3H
3 g4hrs
Endothelial Cell Counts
Corneas of 2-3-kg New Zealand albino rabbits were
excised with a 2-mm scleral rim and placed into a
balanced salt solution (BSS, Alcon)-filled chamber of
20
Fig. 2. The time course of ouabain (10 8 M) and inulin uptake
by the corneal endothelium in potassium-free medium. Values are
the mean ± standard error. The number of measurements at each
time is shown in parentheses.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933349/ on 06/17/2017
1058
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / September 1984
• OK
< 16 ITMK +
000'
100-
/
3~
10
o
OK
+
^
16
1.0
/
I0" 9
IO"8
IO'7
I0" 6
I0" 5
10"*
Ouabain Concentration (moles/liter)
Fig. 3. Endothelial ouabain uptake as a function of glycoside
concentration in potassium-free and 16-mM potassium media. Each
point represents the mean (± standard error) of 5-10 determinations.
Standard errors not shown lie within the boundaries of data point
symbols.
Results
Endothelial ouabain (10~8 M) and inulin uptake as
functions of time are shown in Figure 2. Steady-state
ouabain uptake is achieved in 2.5 hr; therefore, a 3hr incubation period was chosen for subsequent experiments. The uptake of inulin by the corneal endothelium is quite rapid with steady-state being reached
within 1 hr. The inulin space of endothelium plus
Descemet's membrane was determined to be 13.0 ± 0.4
nl/mm2 of endothelium (±SEM, n = 42).
10.000
1000-
3
^
? £
100
< £
o
I0" 9
I0" 8
I0' 7
I0"6
I0'5
Ouabain Concentratration (moles/liter)
Fig. 4. Stromal ouabain uptake as a function of glycoside concentration in potassium-free and 16-mM potassium media. Each
point represents the mean (± standard error) of 5-10 determinations.
Standard errors not shown lie within the boundaries of data point
symbols.
Vol. 25
Figure 3 shows endothelial ouabain uptake as a
function of glycoside concentration. Results are shown
for ouabain uptake in K+-free and in 16-mM K+incubation media. Each data point represents the mean
of 5-10 determinations. Two components of endothelial ouabain uptake are seen clearly in the K+-free
medium. One component saturates at a glycoside concentration near 2 X 10"7 M, while the other increases
linearly with increasing concentration. Baker and Willis14 originally described this pattern of ouabain uptake
and demonstrated that the saturable uptake component
is associated with the specific binding of ouabain to
Na/K ATPase, while the linear component reflects
nonspecific uptake. Figure 3 also shows that potassium
ions displace the saturable component but have no
effect on the linear component. This is consistent with
the known effects of potassium on reducing the affinity
of Na/K ATPase for ouabain.15
For comparison, Figure 4 shows stromal ouabain
uptake at different glycoside concentrations. In this
tissue layer, which consists primarily of extracellular
space, ouabain uptake consists of only one component
that increases linearly with ouabain concentration—
nonspecific uptake. As expected, potassium has no effect on stromal uptake, and all of the glycoside taken
up by this corneal layer can be accounted for by that
entering the extracellular space, which was determined
to be 981 ± 24 nl/mm2 (n = 42). This inulin space is
obtained from the stroma of a deepithelialized cornea
following a 3-hr, in vitro incubation. The stroma is
thus grossly swollen and the inulin space is large.
To determine the contribution of nonspecific binding to total endothelial ouabain uptake, uptake of
3
H • ouabain was measured in the presence of a saturating concentration (10~4 M) of unlabelled glycoside.
This data is shown in Figure 5. Mean (±SEM) values
of ouabain uptake versus concentration are shown as
is the calculated least squares line of bestfit(r = 0.981).
In the presence of 10~4 M cold ouabain, the uptake
of tracer is a linear, nonsaturable function of concentration. Also depicted in Figure 5 is the calculated line
for 3H • ouabain uptake into the inulin space, which
is based on the measured inulin space of 13.0 nl/mm2
endothelium. These data demonstrate that essentially
all of the nonspecific uptake of ouabain is accounted
for by that ouabain taken up by the extracellular space
and that nonspecific binding (ie, nonsaturable uptake
- inulin space uptake) is minimal in the corneal endothelium. Even at 5 X 10~7 M, a concentration at
which nonsaturable ouabain uptake accounts for 32%
(7.6/23.4 to yield a value of 32%) of total uptake, nonspecific binding amounts to only 4.7% of total ouabain
uptake.
Bound ouabain was determined by subtracting tracer
uptake in the presence of 10~4 M unlabelled ouabain
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933349/ on 06/17/2017
N A / K ATPASE PUMP SITES IN CORNEAL ENDOTHELIUM / Geroski and Edelhouser
No. 9
8.0 i
1059
Non-saturable uptake
P--/5.0 (10.5IX
0.50"
Inulin space uptake
±0.4>t
6.0
QIO
0.08
2 4.0
0.06.0.04-
2.0
14.7XI0-8M)
(207
fmoles/mm1)
.....-••'••"
0.02-
0.1
0.2
3
0.3
0.4
0.5
10
H Ouobain concentration (JJLM)
Fig. 5. Nonsaturable endothelial 3 H • ouabain uptake (solid line)
measured in the presence of 10~4 M unlabelled ouabain. Each point
is the mean ± standard error of six determinations. Also shown is
the inulin space 3H • ouabain uptake (broken line), which was calculated from the measured inulin space of 13.0 ± 0.4 nl/mm 2 of
endothelium.
from total tracer uptake in the presence of carrier-free
3
H • ouabain only. A double reciprocal plot was constructed and is shown in Figure 6. Each point represents
the mean of 5-7 determinations. The least squares line
of best fit was calculated (r = 0.999) to determine
maximal binding and the apparent dissociation constant (KD), which were determined to be 20.7 fmoles/
mm 2 endothelium and 4.7 X 10~8 M, respectively.
Discussion
The results of this study demonstrate that ouabain
is bound specifically and with high affinity by the rabbit
corneal endothelium. Since each Na/K ATPase "pump
site" binds one molecule of ouabain,16 determinations
of ouabain binding enable direct quantitation of Na/
K ATPase pump sites. Maximal ouabain binding in
the rabbit corneal endothelium is 20.7 fmoles/mm2 of
endothelium. This corresponds to 1.25 X 10'° pump
sites/mm2. The mean endothelial cell density for the
rabbits used in this study was measured as 4126 cells/
mm 2 of endothelium. Therefore, each endothelial cell
has 3.0 X 106 Na/K ATPase pump sites. Since endothelial Na/K ATPase has been localized histochemically to the lateral cell membrane,17"19 this membrane
region must be capable of accommodating the pump
sites measured in this study. Based on the measured
cell density, the mean anterior (Descemet's surface)
and posterior (aqueous surface) membrane surface
areas for an endothelial cell are 242 ^m 2 each. Assuming that the endothelial cell is a cylinder of 5 /um
height, a minimal estimate of lateral membrane area
can be calculated as 2.76 X 1010 nm 2 /cell. Since each
Na/K ATPase "site" occupies a membrane area of
1000 nm2-20 the lateral membrane area is capable of
20
30
200
I/Ouoboin concentration
Fig. 6. Double reciprocal plot of endothelial ouabain binding versus
glycoside concentration. Ouabain binding was calculated as the difference in uptake in the absence and presence of 10~4 M unlabelled
glycoside. Each point is the mean of six determinations. The least
squares line of best fit was calculated (r = 0.999) and used to estimate
maximal binding (20.7 fmoles/mm2) and the apparent dissociation
constant, KD (4.7 X 10"8 M).
accommodating a minimum of 2.8 X 107 sites per
cell—nearly 10 times the number measured. This calculation neglects the tortuosity of the lateral cell membrane and thus underestimates membrane area. In
reality, many more pump sites could be accommodated
in the lateral membrane than measured; therefore, the
estimated density would appear reasonable.
For comparison to corneal endothelium, Table 1
presents Na/K ATPase pump site densities reported
for several other tissues. The density of Na/K ATPase
pump sites in the corneal endothelium is comparable
with that reported for renal tubule and is of the same
order of magnitude as that determined for choroid
plexus. These comparisons suggest that, on a cellular
basis, the transport capacity of the corneal endothelium
is great and comparable to that of tissues known to
have high rates of active transport. The data of this
study thus provide a new perspective on the relative
magnitude of endothelial transport capacity, which can
Table 1. Comparison of Na/K ATPase pump site
denisites and ouabain-ATPase dissociation
constants reported for several other tissues
to corneal endothelium
Tissue
Human erythrocyte21*
Rabbit renal tubule22
Cat ventricular muscle23
Frog choroid plexus24
Teleost chloride cell25
Rabbit corneal endothelium
Reference number.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933349/ on 06/17/2017
Na/K sites
per cell
260
4.0 X 106
5.2 X 106
10.0 X 106
1.5 X 108
3.0 X 106
KD (M)
2.5
5.1
6.3
8.0
X 10"9
X 10"7
X 10"9
X 10"7
—
4.7 X 10"8
1060
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / September 1984
be difficult to appreciate because of the small mass of
endothelium per cornea.
The high affinity of endothelial Na/K ATPase for
ouabain as estimated by the apparent dissociation constant, K D , suggests that endothelial Na/K ATPase
should be quite sensitive to ouabain inhibition. This
agrees well with the results of ouabain perfusion studies
in both rabbit and human corneas,7-8 which have demonstrated the dose-response relationship of ouabain
concentration to corneal swelling rate.
In these experiments, no stromal ouabain binding
could be measured. This should not be interpreted as
demonstrating the absence of Na/K ATPase in the
stroma. Specific ouabain binding in this corneal layer,
using the present techniques, is masked by a very high
level of nonspecific uptake into the extracellular space
which, in the rabbit cornea, comprises 90% of the
stromal volume.26 Ouabain most certainly is bound
by the stromal keratocytes. The amount bound, however, is immeasurable using the present method, and
all stromal ouabain uptake is accounted for by that
which equilibrates with the extracellular space.
Key words: cornea, corneal endothelium, ouabain, pump
function, sodium-potassium ATPase
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
References
20.
1. Riley M: Transport of ions and metabolites across the corneal
endothelium. In Cell Biology of the Eye, McDevitt D, editor.
New York, Academic Press, 1982, pp. 53-95.
2. Bonting S, Simon K, and Hawkins N: Studies on sodium-potassium-activated adenosine triphosphatase. I. Quantitative distribution in several tissues of the cat. Arch Biochem Biophys
95:416, 1961.
3. Brown S and Hedbys B: The effect of ouabain on the hydration
of the cornea. Invest Ophthalmol 4:216, 1965.
4. Langham M and Kostelnik M: The effect of ouabain on the
hydration and adenosine triphosphatase activity of the cornea.
J Pharm Exp Ther 150:398, 1965.
5. Erdman E: The influence of cardiac glycosides on their receptor.
In Cardiac Glycosides. I: Experimental Pharmacology, Greef K,
editor. New York, Springer-Verlag, 1981, pp. 337-380.
6. Kaye G and Cole J: Electron microscopy: sodium localization
in normal and ouabain treated transporting cells. Science
150:1167, 1965.
7. Trenberth A and Mishima S: The effects of ouabain on the
rabbit corneal endothelium. Invest Ophthalmol 7:44, 1968.
8. Geroski D and Edelhauser H: The effects of ouabain on en-
21.
22.
23.
24.
25.
26.
Vol. 25
dothelial function in human and rabbit corneas. Curr Eye Res
3:331, 1984.
Rogers K: Levels of (Na+K+)-activated and Mg2+-activated
ATPase activity in bovine and feline corneal endothelium and
epithelium. Biochim Biophys Acta 163:50, 1968.
Anderson E, Fischbarg J, and Spector A: Fluid transport, ATP
level and ATPase activities in isolated rabbit corneal endothelium.
Biochim Biophys Acta 307:557, 1973.
Fischbarg J: Active and passive properties of rabbit corneal endothelium. Exp Eye Res 15:615, 1973.
HuffJ and Green K: Demonstration of active sodium transport
across the isolated rabbit corneal endothelium. Curr Eye Res
1:113, 1981.
Lim J: Na + transport across the rabbit corneal endothelium.
Curr Eye Res 1:255, 1981.
Baker P and Willis J: Binding of cardiac glycosides to intact
cells. J Physiol 224:441, 1972.
Matsui H and Schwartz A: Kinetic analysis of ouabain-K + and
Na+ interaction on a Na + , K+-dependent adenosinetriphosphatase
from cardiac tissue. Biochem Biophys Res Commun 25:147,
1966.
Glynn I and Karlish S: The sodium pump. Ann Rev Physiol
37:13, 1975.
Kaye G and Tice L: Studies on the cornea. V. Electron microscopic localization of adenosine triphosphatase activity in the
rabbit cornea in relation to transport. Invest Ophthalmol 5:22,
1966.
Leuenberger P and NovikoffA: Localization of transport adenosine triphosphatase in rat cornea. J Cell Biol 60:721, 1974.
Tervo T and Palkama A: Electron microscopic localization of
adenosine triphosphatase (NaK-ATPase) activity in the rat cornea. Exp Eye Res 21:269, 1975.
Dahl J and Hokin L: The sodium-potassium adenosine-triphosphatase. Ann Rev Biochem 43:347, 1974.
Erdmann E, Krawitz W, and Koch W: Cardiac glycoside receptors
in disease. The number of ouabain binding sites in human erythrocytes is subject to regulation. In Na + + K+-ATPase Structure
and Kinetics, Skou J and Norby J, editors. New York, Academic
Press, 1979, pp. 517-524.
Shaver L and Stirling C: Ouabain binding to renal tubules of
the rabbit. J Cell Biol 76:278, 1978.
Michael L, Schwartz A, and Wallick E: Nature of the transport
adenosine triphosphatase-digitalis complex: XIV. Inotropy and
cardiac glycoside interaction with Na + K + -ATPase of isolated cat
papillary muscles. Mol Pharmacol 16:135, 1979.
Zeuthen T and Wright E: Epithelial potassium transport: Tracer
and electrophysiological studies in choroid plexus. J Membr Biol
60:105, 1981.
Karnaky K, Kinter L, Kinter W, and Stirling C: Teleost chloride
cell II. Autoradiographic localization of gill Na,K-ATPase in
Killifish Fundulus heteroclitus adapted to low and high salinity
environments. J Cell Biol 70:157, 1976.
Kaye G: Stereologic measurement of cell volume fraction of
rabbit corneal stroma. Arch Ophthalmol 82:792, 1969.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933349/ on 06/17/2017