Investigative Ophthalmology & Visual Science, Vol. 29, No. 2, February 1988
Copyright © Association for Research in Vision and Ophthalmology
Bicarbonate Sensitivity of Rabbit Corneal
Endothelium Fluid Pump In Vitro
Michael J. Doughty* and David Mauricef
Stroma-endothelium preparations from rabbit corneas were mounted between two chambers and
incubated with identical media on either side which contained different bicarbonate levels and, in some
experiments, organic (Good's) buffers. Active fluid flow across the preparations was measured by
means of a capillary tube attached to the stroma-side chamber. With media containing 2 to 50 mM
bicarbonate (pH 6.2 to 7.8 in equilibrium with 5% C0 2 -air at 37°C), the fluid pump was constant for at
least 3 hr at a rate of 5 /Ltl/hr cm2 and was not significantly affected by the bicarbonate level. Over the
same range of pH and bicarbonate but supplemented with 50 mM organic buffer, fluid pump was 8
jul/hr cm2 for all bicarbonate concentrations used. Using Ringer solutions supplemented with 50 mM
buffer (pH 6.3 to 8.4) but without added bicarbonate and in equilibrium with air, fluid pump was
observed at approximately 4 /xl/hr cm2 at pH 6.3 and increased to 8 /Ltl/hr cm2 at pH 7.8. In all cases,
fluid pump persisted for at least 5 hr. Invest Ophthalmol Vis Sci 29:216-223, 1988
used to cover a solution bathing the anterior corneal
surface,6'7 the present experiments were undertaken
to check their conclusions. The bicarbonate sensitivity of the endothelial fluid pump (with adequate control of stromal HC0 3 -C0 2 equilibria) was evaluated
by mounting the endothelium-stroma combination
between two chambers containing regularly changed
Ringer's solution and then directly measuring fluid
flow along a capillary. Some of this work has been
published in abstract form.8
The corneal endothelium will thin a swollen cornea or maintain it at its normal thickness by an active
mechanism of deturgescence, which is thought to be
associated with an active fluid pump in the endothelium. The sensitivity of corneal deturgescence to a
lowering of the level of bicarbonate ions in the anterior chamber was noted more than 10 years ago,1'2
and the effect has been confirmed and examined in
some detail.3'4 Recently, it was subjected to a more
thorough analysis, paying particular attention to the
control of C 0 2 - H C 0 3 ~ equilibria in the anterior
chamber.5 In that study, the deturgescence of isolated
rabbit cornea was measured under a specular microscope when the anterior, epithelium-free, surface was
covered with silicone oil after a period of pre-equilibration. However, because of the permeability of the
oil to C 0 2 , it was found that gas equilibria in the
stroma could not be adequately controlled. Since two
previous studies concerning the bicarbonate sensitivity of the endothelial fluid pump (ie, direct measurement of fluid flow across the endothelium) also employed techniques where a layer of silicone oil was
Materials and Methods
Animal care and usage followed the guidelines established in the ARVO Resolution on the Use of Animals in Research. New Zealand White rabbits
(mostly female; 1.6 to 3.0 kg; average external corneal
diameter of 10.75 to 11.75 mm) were killed by decapitation or an intraveneous overdose of Euthanyl®
(M.T.C. Pharmaceuticals, Mississauga, Ontario,
Canada) or T-61® (Hoecht Corp., Sommerville, NJ)
followed by division of the neck blood vessels. The
globes were enucleated, together with their lids and
conjunctiva 2 within 15 min of death and, if they
could not be used immediately, they were stored for
up to 23 hr in sealed, moist containers at 4°C in the
dark, with the lids closed over the cornea, which was
facing downwards.
Immediately before use, the corneal surface of the
enucleated eye was flushed with 150 mM NaCl at
room temperature and the corneal epithelium totally
removed by firm abrasion with a dulled razor blade.
The scraped surface was then thoroughly washed,
From the *School of Optometry, University of Waterloo, Waterloo, Ontario, Canada, and the fOphthalmology Division, Stanford University Medical Center, Stanford, California.
Supported by grants from the School of Optometry, University
of Waterloo, and an operating grant from the Natural Sciences and
Engineering Research Council of Canada to MJD and from the
National Institutes of Health (EY-00431) to DM.
Submitted for publication: May 30, 1986; accepted September
24, 1987.
Reprint requests: Michael J. Doughty, Faculty of Science, School
of Optometry, Waterloo, Ontario, Canada N2L 3G1.
216
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
No. 2
BICARBONATE AND CORNEAL FLUID PUMP / Doughty and Maurice
first with saline and then with the incubation solution
of choice at 37°C. The completion of the epithelial
removal was routinely checked by light microscope
observation of gluraldehyde-fixed sections prepared
after the fluid flow studies were completed.
The stroma-endothelium preparation was
mounted in a chamber similar in essentials to those
used in both fluid pump and solute transport studies
on the rabbit cornea9,10 (Fig. 1). The stromal surface
of the rabbit eye was then sealed to an acrylic plastic
mounting tube (OD, 14 mm; wall thickness, 1.5 mm;
length, 20 mm) by applying a light suction in the tube
and tying the conjunctiva to a groove near its end. To
provide mechanical stability to the preparation, a 100
mesh polypropylene grid had been molded to a 6.25
mm radius of curvature and glued to the end of the
mounting tube. After cutting away the back of the
globe, the vitreous and crystalline lens, iris and ciliary
body were carefully removed after folding a scleral lip
back over the edge of the mounting tube.
A double chamber was designed to stand upright in
a water bath at 37°C with the cornea facing downwards. In the bottom (stromal) half of the chamber,
one fluid connection (PE-100 tubing) was made to its
base and another reached up to the edge of the corneal stroma, where it was cut so as to fit into the angle
of the supporting mesh. This arrangement made for
more efficient replacement of fluid contained in the
chamber and served also to facilitate removal of any
gas bubbles trapped in it.
To assemble the system, the lower tube (L) was
dipped into a reservoir of the incubation solution and
gentle suction applied to the upper tube (U) by means
of a 5 ml syringe. The mounting tube with corneal
preparation attached was inserted into the bottom
chamber and was slowly drawn down to its seating by
the suction. At the same time, the chamber was filled
(—2 ml) by fluid from the reservoir. The assembly
was then completed by carefully clamping the top
onto the bottom chamber while maintaining a slight
suction on the cornea. The top chamber (endothelial
side) was filled with incubation solution and its
opening plugged with a rubber stopper pierced with a
19 gauge syringe needle, into the barrel of which was
inserted a 22 gauge needle. This closure served to
prevent evaporation from the upper chamber and
minimize loss of C 0 2 in the solutions in the upper
chamber. The assembled apparatus was immersed in
a water bath up to the lip of the top chamber. Using
the syringe on line U, approximately 3 ml of incubation solution was drawn from the reservoir through
the bottom chamber and then the line was clamped.
Exit line L was then removed from the reservoir and
connected to a horizontal capillary tube M (PE-100
217
Endothelial
surface
arid on stromal
surface
20 cm
Fig. 1. Schematic diagram of apparatus (assembled) used to assess fluid pump.
tubing), 20 cm below the endothelium. Line U was
then briefly undamped and a few microliters of colored 0.9% NaCl containing 0.5% Photoflow® (Eastman Kodak, Rochester, NY), drawn into line M and
the meniscus position adjusted to be around 10 cm
from its open end. Line U was then clamped again. A
millimeter scale was attached to the benchtop next to
the meniscus.
At any point in the experiments (usually every 60
min), the solution in the lower chamber could be
changed by clamping line M, uncoupling the connection and replacing the free end of line L into the
freshly-replenished reservoir. Fresh solution was then
drawn through the apparatus by the syringe on the
undamped line U without contamination of the bottom chamber with Photoflow. At the same time, the
solution in the top chamber also was replaced by removing the stopper, sucking out its contents with a
Pasteur pipette, quickly refilling the chamber with
fresh solution from the reservoir and replacing the
stopper.
The incubation solutions were prepared and used
as described previously.5 Details of the solutions are
provided in Table 1. The solutions in series 1 and 2
contained various levels (2 to 50 mM) of bicarbonate,
which were exchanged on an isomolar basis for NaCl.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
218
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / February 1988
Table 1. Composition of corneal
incubation solutions
Series
1
2
NaCl (mM)
N a H C 0 3 (mM)*
KC1 (mM)
MgCl2 (mM)
CaCl2 (mM)
Glucose (mM)
Glutathione ( m M ) |
Adenosine (mM)
pH
Good's buffers (mM)§
80-128
2-50
5.5
0.9
1.8
5.5
0.1
1.0
6.2-7.8
0
3
30-78
2-50
5.5
0.9
1.8
5.5
0.1
1.0
6.3-7.9
50
78
zerof
5.5
0.9
1.8
5.5
0.1
1.0
6.0-8.4
50
* All solutions containing added N a H C 0 3 were equilibrated with 5%
C0 2 -air at 37°C to ensure maintenance of bicarbonate levels and pH.
t No special precautions were taken to eliminate C 0 2 and carbonate from
the solutions so that a trace of bicarbonate (0.03 mM) may be present.
% Added as its reduced form but no precautions were taken to maintain
the glutathione in this form.
§ All buffer ingredients were prepared, prior to addition to the corneal
incubation solutions, to contain a titratible equivalent of 50 mM Na (OH)
along with 50 mM buffer concentration.
Thus, with the slight difference in dissociation of the
two salts in aqueous solution, 1112 the osmolarity of
the solutions fell slightly as the bicarbonate level was
raised.5 All solutions were prepared in double distilled water, and used within 8 hr of preparation.
On being gassed to equilibrium with 5% C0 2 -air at
37°C, series 1 solutions exhibited a range of pH
values from 6.2 to 7.8. 513 " 18 Loss of C 0 2 resulted in a
either a rapid or slow increase in pH, depending on
the rate of outgassing,13"19 and careful monitoring of
the pH of the solutions is a convenient way to check
i >
i t
}
the C 0 2 tension. The series 2 solutions were similar
to series 1 but an organic buffer system was added in
order to stabilize the pH. The effectiveness of the
buffer was evaluated by observing the change in pH
that occurred while C 0 2 was allowed to leave these
solutions under standard conditions. The best results
were obtained with the following Good's buffers20,21
at 50 mM concentration: MES (pKa = 5.95; 37°C
adjusted), PIPES (6.65), Bis-Tris-Propane (6.50),
MOPS (7.10), TES (7.16), HEPES (7.30), Tricine
(7.8) and Bicine (8.04). Using a combination of two
buffers with slightly different pK a values enhanced
the control of pH. 21 ' 22 Each buffer was prepared as a
10 X stock in distilled water, its pH adjusted to the
pKa with NaOH and then they were added to the
incubation solutions in that proportion that was
found empirically to give a pH corresponding to that
of any of the bicarbonate solutions when saturated
with 5% C0 2 -air at 37°C. These solutions were then
gassed with 5% C0 2 -air. The final pH was within 0.1
pH unit of the designated values (Fig. 2).
It should be noted that Tris buffer,23 while it has
been used in corneal research,1'24 is not satisfactory
both because it has very limited buffer capacity below
pH 7.5 and because it is rather reactive with C 0 2 and
bicarbonate.15,20'25 It behaved poorly in controlling
the pH in comparison with the buffers listed above.
A third series of solutions (series 3) was also used.
The NaHC0 3 was totally replaced with NaCl and the
solutions equilibrated with air. The pH was set by the
use of the same combinations of buffers used in series
2 thus permitting the assessment of the pH dependence of fluid flow in the nominal absence of bicarbonate and C 0 2 . It is acknowledged that the tissue is
capable of producing C 0 2 4 and that the atmosphere
contains a small quantity of C 0 2 that could produce a
[bicarbonate + C0 2 ] level of 0.03 mM 26 ). No effort
was made during the preparation of solutions to
purge them of atmospheric C 0 2 or dissolved carbonate.
Results
>
7A
Vol. 29
Fluid Flow Measurements
6-6 l
5.4 L
[
j
T
62 I
}
6-o L
o no buffer
• w i t h 50 mM buffer
1
•
0
5
•
>
'
«
1
«
'
10 15 20 25 30 35 40
ADDED BICARBONATE - m M
J
—
45
•
50
Fig. 2. Bicarbonate-pH profiles for corneal incubating solutions
of series 1 and series 2 (with added Good's buffers). All solutions
were equilibrated with 5% C0 2 -air at 37°C.
The chambers were constructed of thick acrylic,
and errors could occur as a result of the uptake of
water by the plastic or of slow thermal expansion of
the chamber27 and its fluid contents. In previous experiments with a similar chamber, capillaries were
attached to both sides of the cornea and the loss of
fluid on one side was seen to be compensated by a
gain on the other, which showed that these errors
were neglible.10'29 With a single capillary system, they
need to be reconsidered.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
No. 2
219
BICARBONATE AND CORNEAL FLUID PUMP / Doughry and Maurice
The chamber was filled with water at room temperature and immersed in the incubator bath while the
temperature at the level of the cornea was measured
with a digital thermometer. It was found to rise halfway to its final value in 10 min and show no appreciable change after 30 min. Therefore, in the experimental situation where the chambers are filled with
37°C solution, temperature changes are not likely to
be a cause of trouble after this time. In other tests, the
cornea was replaced with a 1 mm thick silicone rubber disc and the chamber was assembled and filled in
the normal manner. The movements of the meniscus
were found to be neglible after 60 min. The initial
velocity of the meniscus, which was away from the
chamber, was directly related to the time that the
equipment was allowed to dry between experiments:
it varied from 5 to 37 iA/hr for drying periods of 1 to
21 days. Apparently the volume of the plastic, as it
absorbs water, is greater than the sum of its constituents.
Evaporation, which must be very small through the
needle hole in the stopper, would have a miniscule
effect on the tonicity of the fluid in the upper
chamber and then could only change the meniscus
position indirectly by osmosis across the cornea.
Where a stroma-endothelial preparation is perfused in a similar manner under a specular microscope,5'28 the stroma is seen to swell immediately after
mounting. Since the anterior surface is fixed by the
grid, such swelling would displace the endothelium
into the upper chamber, and this could move the
meniscus as if the cornea was pumping fluid into the
top chamber. However, the extent of swelling is limited, and with a pressure head of 20 cm H 2 0, it is
virtually complete within 60 min, although it may
continue for 90 min (observed in five of 13 preparations) in very low bicarbonate solutions.
For many reasons, therefore, the movements of the
meniscus over the first hour were ignored, and readings were carried out over the next 4 hr.
Effect of Bicarbonate Concentration on Pump With
Bicarbonate Buffer/C0 2
The fluid flow corresponded to a leak for the first
35 to 40 min in most preparations but later reversed
its direction and moved against the applied hydrostatic pressure of 20 cm H 2 0 (Fig. 3). Part of the
initial leak is attributed to the chamber equilibration
(which could persist for as long as 60 min), and the
time of initiation of the net fluid pump activity is
uncertain. However, once established, it continued
for at least 5 hr in all 107 preparations examined. The
slightly longer starting times were generally observed
100
E
E
i
o
in
o
CL
10
tf 60
z:
UJ
I
40
60
1
120
180
MINUTES
240
300
Fig. 3. Typical experimental result for fluid flow measurements
across stroma endothelium preparation. From the net movement
of the fluid meniscus over 1 hr periods, the rate of fluid pump can
be calculated. The flow in the first 60 min is in the leak direction.
Arrows indicate the time of solution changes.
in those preparations incubated with 2 to 10 mM
bicarbonate.
The rate of fluid flow was measured by taking
readings of the meniscus position every 10 min and
then fitting (by linear regression, if necessary) the data
points to a line over the intervals of 60 min between
changes of the chamber solutions. The fluid flow is
expressed as ^1/hr cm 2 of endothelium, taking into
the account the area of exposed endothelium in the
apparatus (1.17 mm 2 ).
In most experiments, changing the solutions in the
chambers every 60 min resulted in good control of
their gas equilibria. This equilibrium was assessed by
measuring the top chamber pH at the end of the 60
min period. The increase in pH was only 0.1 to 0.2
pH units, and often no change was detectable (ie <0.1
pH unit change). However, with longer periods between solution changes, increases of up to 0.5 pH
units were observed.
Changing solutions every 60 min also resulted in a
maintenance of fluid pump activity stable to within
± 10% from hour to hour, at all bicarbonate concentrations tested. A few corneas were studied over a
period of 9-10 hr with 2, 5, 35 and 50 mM bicarbonate and showed the same behavior. With longer intervals between changes (eg 2 hr), the variance from
hour to hour increased to ± 15%. Without such solution changes, the fluid pump declined to zero at a rate
of approximately 30%/hr (see also refs. 29-31).
Using the same bicarbonate-Ringer solutions on
both sides of the preparation, the effect of 11 different
levels of bicarbonate (in equilibrium with 5% C0 2 -air
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
10
8
6
4
\^r
2
0
10
5
JO
15
20
25 30
35
40
45 50
r
(E0
8
6
+_^-^-^—Kj
4
demonstrating that the desired effect of the buffer
addition had been achieved. It is evident, furthermore, that these buffers are not toxic to the endothelium.
At all 11 concentrations of bicarbonate and at all
time periods evaluated, the fluid pump activity was
significantly higher, almost double, than that seen
with solutions in series 1 not containing the buffers.
As with these unsupplemented solutions, no dependence upon bicarbonate concentration was observed
for the fluid pump (Fig. 5).
Effect of pH on Fluid Pump Activity
Without Bicarbonate
In this series, the solutions were similar to those of
series 2 except that no bicarbonate was added and the
solutions were equilibrated with air at 37°C.
2
=*s
0
3
Vol. 29
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / February 1988
220
5
10
15
20
25
30
35
40
45
50
10
Q_
(c)
10
Q
8
6
4
0
5
10 15
ADDED
20
25 30
35 40
BICARBONATE— mM
45
2
50
Fig. 4. Fluid pump activity as a function of the bicarbonate levels
(series 1, without added buffers) in the solutions incubating the
stroma-endothelium preparations in the 2nd (A), 3rd (B) and 4th
(C) hour after starting the experiments. Results are mean values
± SEM for seven to eight preparations incubated under each condition.
0
0
£
10
^
I
6
5
10
15
20
25
30
35
40
45
50
t *r
CL
at 37°C) on the fluid pump was determined (Fig. 4).
The pump rate was found to be essentially independent of bicarbonate concentration over the range of 2
to 50 mM. The only exception was the slower rate at
the lower bicarbonate levels in the second hour.
2
*
°-
2
o
o
B
L
3
<
•
1
<
1
•
0
5
10
15
20
25
•0
5
'
•
•
'
35
40
45
50
10
15
20 25 30
35
40
ADDED
BICARBONATE-mM
45
50
30
'
6
Effect of Bicarbonate Concentration on the Pump
With 50 mM Organic Buffers in the Presence of C0 2
The experiments described above were repeated
but now using the series 2 solutions that were supplemented with a total concentration of 50 mM of the
Good's buffers.
In 56 preparations, the fluid pump started within
40 to 50 min, after an initial leak period, and persisted for at least 5 hr unchanged. With the use of
these buffered solutions, no changes in the pH of the
top chamber were detected after the 60 min period,
4
2
0
L
Fig. 5. Fluid pump activity as a function of the bicarbonate levels
(series 2, containing 50 mM Good's buffers) in the solutions incubating the stroma-endothelium preparations in the 2nd (A), 3rd (B)
and 4th (C) hours after starting the experiments. Results are mean
values ± SEM for four to six preparations incubated under each
condition.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
No. 2
BICARBONATE AND CORNEAL FLUID PUMP / Doughry and Maurice
These results (Fig. 6) reveal that such "bicarbonate-free" solutions support fluid pump activity. However, the fluid pump was rather less predictable when
these solutions were used and in most preparations
(17 out of 21), the fluid pump in the third hour was
higher than that in the second hour. There was also
considerable variability in the fluid pump activity of
different preparations and over time in the same
preparation. While the pump persisted in all preparations for at least 5 hr, the rates observed in each hour
could differ by as much as ±20% from those seen in
the second hour. The data suggest that the pump rate
increases slightly with pH from a value of about 4
lA/hr cm2 at pH 6.0 to about 8 /A/hr cm2 at pH 8.0
and falls at higher pH values.
10
A;
Q_
^
0L
6-0
6-2
6-4
6-6
6-8
7-0
o
o
7-A
7-6
7-8
8-0
8-2
8-4
8-0
8-2
8-4
9 8
Z)
^
6
B
6-2
6-4
6-6
6-8
7-0
7-2
7<
7-6
7-8
PH
Discussion
Several previous studies ' ' ' have shown that
the thickness of the stroma-endothelium preparation
incubated in vitro, between two solutions, changes
from its initial physiological value to a steady state
(equilibrium) value which is dependent upon the hydrostatic pressure in the anterior chamber. The
length of the time required for this stable thickness
value to be achieved depends upon the hydrostatic
pressure and is about 60 min for the 20 cm H 2 0 used
in the present experiments. After this time, the movement of fluid into the top chamber of the apparatus
must correspond to an equal movement of fluid
across the corneal endothelium and out of the bottom
chamber, and is against the hydrostatic pressure gradient. The rate found for this pump, in bicarbonate/
C02-buffered solutions was around 5 iA/hr cm2 which
is within the range found by other workers.
Under the experimental conditions used here, the
fluid pump of the corneal endothelium was found not
to be sensitive to bicarbonate regardless of whether
the organic buffer supplement was present or absent.
It is also noteworthy that, in the presence of these
buffers, the fluid pump activity was almost twice as
high as when the preparations are incubated in
simpler bicarbonate/C02-buffered solutions. Similarly, the pump showed no significant pH sensitivity
over the range tested (6.0 to 8.0); the alternative possibility that the sensitivity to a fall in [HC03~] is exactly compensated for by that to a fall in pH is not
probable. Furthermore, under the very unphysiological situation in which the levels of bicarbonate and
C0 2 are extremely low and the solution pH is controlled by 50 mM organic buffer, fluid pump activity
is maintained over an extended time period. It does
not appear, therefore, that this activity is controlled
by the C0 2 level.
7-2
Q- 10
6-0
5 28 32 33
221
Fig. 6. Fluid pump activity as a function of the pH of the solutions (series 3, without added bicarbonate and equilibrated with
air) in which stroma-endothelium preparations were incubated in
the 2nd (A) and 3rd (B) hours after starting the experiments. Data
points are from single measurements.
Thus, the fluid pump of the endothelium does not
show the marked bicarbonate sensitivity that has
been observed for the phenomenon of corneal deturgescence as measured in the specular microscope
when the cornea is covered with silicone oil.1"6 During the course of the present investigations, measurements of the fluid pump with solution series 1 were
carried out simultaneously with measurements of deturgescence5 using the same solutions in a paired-eye
protocol. The two phenomena show very different
bicarbonate sensitivities: the former changed not at
all and the latter 15-fold. At very low bicarbonate
levels, no deturgescence was observed yet the fluid
pump appeared intact. These results therefore require
reconsideration of the widely accepted view that the
fluid pump in the corneal endothelium is bicarbonate-dependent.34"38
The previous study in which thefluidpump (rather
then deturgescence) had been assessed under several
different levels of bicarbonate, claimed that the "pH
was adjusted to be constant" in all solutions and used
a method in which the stromal side of the preparation
was covered with bicarbonate-Ringer solution overlaid by a layer of silicone oil.6 It must be borne in
mind that of the three variables—C02 pressure, pH
and HCO3" concentration—only two can be independently controlled.13"19 In a second series of experiments, where the effects of incubating stroma-endothelium preparations with or without a single level of
bicarbonate were reported,38 it was noted that "in
HC03~ and C02-free Ringer solution movements (of
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
222
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / February 1988
fluid) generally cease." The solution contained 5.5
mM Na 2 HP0 4 and the pH was not specified.
In the deturgescence experiments, carried out
under a specular microscope, the anterior stroma was
covered with a layer of silicone oil whereas in the
present experiments the stromal side of the preparation was in contact with a large volume of frequently
renewed solution. Carbon dioxide is readily soluble
in silicone oil (Dow Corning literature) as it is in
other organic solvents14,39 and can thus pass across a
layer of silicone oil and be lost from any solution over
which it is placed. The loss of C 0 2 should be accompanied by an increase in pH (which was noted5) and a
shift in the equilibrium to give a lower concentration
of bicarbonate in the solutions immediately underlying the oil. Since the pump was found to be insensitive to pH, H C 0 3 " concentration and C 0 2 , it may be
that a difference in concentration on the two sides of
the endothelium is responsible for the observation of
an apparent bicarbonate sensitivity to endothelial
fluid pump activity. Further studies are required to
test this possibility.
When bicarbonate was present, the addition of an
effective concentration of Good's buffers considerably enhanced the pump. It is perhaps also significant
that when bicarbonate was absent, the pump operated in our experiments when buffers were added but
was largely inhibited in those of Mayes38 where the
solutions were scarcely buffered. In view of the varied
chemical nature of these compounds, their action on
the pump must be a direct response to their buffering
capacity rather than some pharmacological action.
There is only neglible change in pH in the bulk of the
solution when bicarbonate and C 0 2 alone are used,
but it is possible that larger changes take place locally
on the boundary layers that will readily form on either side of the endothelium, and the organic buffers
may be more effective in neutralizing them. It is unknown, at this time, whether the added organic
buffers serve to buffer the bulk solution hydrogen ion
concentration or indirectly serve to buffer H + or C 0 2
equivalents related to cellular pH homeostasis and
responsivity to extracellular or intracellular metabolic changes. In this connection, it can be noted that
the basal surface of the endothelial cells is adjacent to
a natural thick unstirred layer, the corneal stroma,
and thermal gradients in the bottom chamber will be
small. On the other hand, small but appreciable gradients, which might lead to convective stirring, are
present in the upper chamber because its upper surface is not covered by the waterbath, as was confirmed by measurements with a digital thermometer.
Key words: cornea, endothelium, fluid pump, bicarbonate
Vol. 29
References
1. Hodson S: Evidence for a bicarbonate-dependent sodium
pump in corneal endothelium. Exp Eye Res 11:20, 1971.
2. Dickstein S and Maurice DM: The metabolic basis to the fluid
pump in the cornea. J Physiol (Lond) 221:29, 1972.
3. Fischbarg J and Lim JJ: The role of cations, anions and carbonic anhydrase in fluid transport across rabbit corneal endothelium. J Physiol (Lond) 241:647, 1974.
4. Hodson S and Miller F: The bicarbonate ion pump in the
endothelium which regulates the hydration of rabbit cornea. J
Physiol (Lond) 263:653, 1976.
5. Doughty MJ: New observations on bicarbonate-pH effects on
thickness changes of rabbit corneas under silicone oil in vitro.
Am J Optom Physiol Opt 62:879, 1985.
6. Hodson S: The regulation of corneal hydration by a salt pump
requiring the presence of sodium and bicarbonate ions. J Physiol (Lond) 236:271, 1974.
7. Mayes KR and Hodson S: Local osmotic coupling to the active
transendothelial bicarbonate flux in the rabbit cornea. Biochim Biophys Acta 514:286, 1978.
8. Doughty MJ and Maurice DM: Bicarbonate-pH sensitivity of
deturgescence and fluid translocation phenomena in rabbit
corneas. ARVO Abstracts. Invest Ophthalmol Vis Sci
26(Suppl):107, 1985.
9. Kim JH, Green K, Martinez M, and Paton D: Solute permeability of the corneal endothelium and Decemet's membrane.
Exp Eye Res 12:231, 1971.
10. Maurice DM: The location of the fluid pump in the cornea. J
Physiol (Lond) 221:43, 1972.
11. The Merck Index. 10th ed. Rahway, NJ, Merck & Co. Ltd.:
Misc-47, 1983.
12. Wimberley PD, Siggaard-Andersen O, Fogh-Andersen N, and
Boink ABTJ: Are sodium bicarbonate and potassium bicarbonate fully dissociated under physiological conditions? Scand
J Clin Lab Invest 45:7, 1985.
13. Van Slyke DD and Sendroy J: Line charts for graphic calculations by the Henderson-Hasselbach equation, and for calculating plasma carbon dioxide content from whole blood content.
J Biol Chem 79:781, 1929.
14. Quinn EL and Jones CL: Carbon Dioxide. New York, Reinhold Publishing Corp., 1936, pp. 19-145.
15. Edsall JT and Wyman J: Carbon dioxide and carbonic acid. In
Biophysical Chemistry, Vol. 1, New York, Academic Press,
1958, pp. 550-590.
16. Kern DM: The hydration of carbon dioxide. J Chem Ed 37:14,
1960.
17. Roos A and Boron WF: Intracellular pH. Physiol Rev 61:296,
1981.
18. Ashwood ER, Kost G, and Kenny M: Temperature correction
of blood-gas and pH measurements. Clin Chem 28:1877, 1983.
19. Marshall EK: The effect of loss of carbon dioxide on the hydrogen ion concentration of urine. J Biol Chem 51:3, 1922.
20. Good NE, Winget GD, Connolly TN, Izawa S, and Singh
RMM: Hydrogen ion buffers for biological research. Biochemistry 5:467, 1966.
21. Gueffroy DE: A guide for the preparation and use of buffers in
biological systems. La Jolla, CA, Calbiochem, 1975.
22. Perrin DD and Dempsey B: Buffers for pH and Metal Ion
Control. London, Chapman & Hall, 1974.
23. Bates RG: Amine buffers for pH control. Ann NY Acad Sci
92:341, 1961.
24. Lim JJ: Effects of bicarbonate on the potential difference
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017
No. 2
25.
26.
27.
28.
29.
30.
31.
32.
BICARBONATE AND CORNEAL FLUID PUMP / Doughry and Maurice
across the rabbit corneal endothelium. Curr Eye Res 2:529,
1982/1983.
McFarland WN and Norris KS: The control of pH by buffers
in fish transport. California Fish & Game 44:291, 1958.
Woodbury JW: Regulation of pH. In Physiology & Biophysics,
Ruck TC and Patton HD, editors. Philadelphia, W.B.
Saunders, 1965, pp. 899-934.
Gross S: Modern Plastics Encyclopedia. New York, McGrawHill Publications, 1969-1970, pp. 968-969.
Baum JP, Maurice DM, and McCarey BE: The active and
passive transport of water across the corneal endothelium. Exp
Eye Res 39:335, 1984.
Barfort P and Maurice DM: Electrical potential and fluid
transport across the corneal endothelium. Exp Eye Res 19:11,
1974.
Fischbarg JJ, Hofer GL, and Koatz RA: Priming of the fluid
pump by osmotic gradients across the rabbit corneal endothelium. Biochim Biophys Acta 603:198, 1980.
Fischbarg J, Lim JJ, and Bourquet J: Adenosine stimulation of
fluid transport across rabbit corneal endothelium. J Memb
Biol 35:95, 1977.
Mayes KR and Hodson S: Some effects of hydrostatic pressure
33.
34.
35.
36.
37.
38.
39.
223
on corneal hydration during specular microscopy. Exp Eye Res
26:141, 1978.
Ehlers N: Mechanical factors in the maintenance of normal
corneal deturgescence. Acta Ophthalmol (Copenhn) 45:658,
1967.
Maurice DM: The cornea and sclera. In The Eye, 3rd Edn. Vol.
IB. Davson H, editor. New York, Academic Press, 1984, pp.
1-158.
Riley MV: Transport of ions and metabolites across the corneal endothelium. In Cell Biology of the Eye, McDevitt DS,
editor. New York, Academic Press, 1982, pp. 53-95.
Waring GO, Bourne WM, Edelhauser HF, and Kenyon KR:
The corneal endothelium: normal and pathological structure
and function. Ophthalmology 89:531, 1982.
Fischbarg J and Lim J J: Fluid and electrolyte transports across
corneal endothelium. Curr Top Eye Res 4:201, 1984.
Mayes KR: Further evidence for local osmotic coupling in the
rabbit cornea. Biochim Biophys Acta 600:831, 1980.
Gerrard W: Carbon dioxide, nitrous oxide, hydrogen sulfide,
chlorine, sulfur dioxide and carbonyl chloride. In Solubility of
Gasses and Liquids. A Graphical Approach. New York, Plenum Press, 1976, pp. 89-114.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933369/ on 07/31/2017