Download Volume regulation of non-pigmented cells from ciliary

Volume Regulation of Non-Pigmenfed Cells
From Ciliary Epithelium
Nasser A. Farahbakhsh and Gordon L. Fain
We describe a new method for investigating ion and water transport in ciliary epithelium. A single
ciliary process from the rabbit is isolated, placed in a chamber, and rapidly perfused with physiological
Ringer. If this process is then viewed at its edge with a light microscope using Hoffman modulation
contrast optics, it is possible to record the image of a single layer of non-pigmented epithelial cells.
When these cells are exposed to hypotonic Ringer, they swell and then reduce their volume by
extruding salt and water. We have used the rate of swelling to calculate the hydraulic conductivity of
the non-pigmented cells. These measurements show that the aqueous humor could be secreted by the
generation of a modest osmotic gradient across the non-pigmented cell basal membrane. We have also
used this preparation to investigate the mechanism of the decrease in cell volume after swelling in
hypotonic medium. This volume regulatory decrease can occur at a rate as large as the rate of normal
fluid transport during the secretion of the aqueous. It appears to be caused in part by active Na
extrusion, since it is partially inhibited by exposure to ouabain. In addition, there appears also to be a
contribution to the regulatory response from a swelling-induced K efflux, since the rate of volume
regulation can be altered by perfusion with Ba2+ and by changing the extracellular K concentration.
Invest Ophthalmol Vis Sci 28:934-944,1987
preparation is complicated, since there are four
membranes between the two sides of the epithelium,
rather than two. Futhermore, the pigmented and
non-pigmented cell layers are electrically connected
to one another 1 ' by an extensive network of gap
junctions,12 so that membrane potential changes in
the two layers cannot be readily distinguished.
In an attempt to circumvent some of these difficulties, we have developed a new preparation for the
study of ion and water transport in ciliary epithelium,
similar in concept to the one used by Berggren13 to
measure time-dependent shrinkage of whole ciliary
processes. In our experiments, we isolate and perfuse
a single ciliary process and measure at high magnification the change in width of the non-pigmented cell
layer during exposure to media of altered tonicity.
From these measurements it is possible to estimate
the hydraulic conductivity of the non-pigmented cell
membrane and to deduce some of the properties of
ion movements in these cells. A preliminary report
of this work was presented at the annual meeting
of the Association for Research in Vision and Ophthalmology.14
The aqueous humor is produced by an epithelial
layer which covers the surface of the ciliary body at
the inner margin of the eye. The physiology of this
epithelium has not been extensively studied, in part
because of its complicated morphology. The ciliary
epithelium consists of two cell layers1: a pigmented
cell layer, whose basal membrane lies adjacent to the
fenestrated capillaries of the stroma; and a non-pigmented cell layer, whose basal membrane faces the
lumen of the posterior chamber. The non-pigmented
layer is probably primarily responsible for the secretion of the aqueous (see for example ref. 2). A further
complication of this tissue is that it occurs for the
most part as a highly convoluted and invaginated
layer, most of whose surface area is within the tortuous folds of the ciliary processes of the pars plicata.
Despite numerous attempts,3"10 it is unclear whether
this tissue can be satisfactorily mounted in an Ussing
chamber with access to both serosal and luminal surfaces. Even if this were possible, the interpretation of
potential and resistance measurements from such a
From the Department of Ophthalmology, Jules Stein Eye Institute, UCLA School of Medicine, Los Angeles, California.
Supported in part by NIH grants EY 01844 and EY 00331 and
by a Research to Prevent Blindness Inc. Manpower Award to GLF,
and by NIH postdoctoral traineeship NS 07101 to NAF.
Submitted for publication: September 9, 1986.
Reprint requests: Nasser A. Farahbakhsh, PhD, Department of
Ophthalmology, Jules Stein Eye Institute, UCLA School of Medicine, Los Angeles, CA 90024.
Materials and Methods
The dissection and mounting of the tissue is diagrammed in Figure 1. Two-kilogram pigmented rabbits were anaesthetized by a lethal injection (10-15
cc) of a 30% solution of chloral hydrate into the peri-
934
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
No. 6
VOLUME REGULATION OF NON-PIGMENTED CELLS / Forohbokhsh and Fain
toneum. The dissection was begun after all reaction
of the animal to mechanical stimuli (pinching of the
ear) had stopped, and after breathing had become
shallow and irregular. This usually occurred a few
minutes before death. The eye was quickly removed
and hemisected and the animal immediately sacrificed. These procedures were in accordance with the
ARVO Resolution on the Use of Animals in Research. The anterior segment of the eye was placed
with the cornea facing downwards on a petri dish
whose bottom was covered with a 5 mm layer of
Sylgard 184 (Dow Corning Corp., Midland, MI). The
petri dish rested on a slide warmer and contained the
normal balanced salt solution (BSS, Table 1, solution
A) saturated with 95% O2/5% CO2 at 37°C. The anterior segment of the eye was fastened at its margin to
the Sylgard with insect pins, and the lens and lens
capsule were carefully removed. A single (usually secondary) ciliary process was excised from the posterior
side of the pars plicata (Fig. 1A) and gently transferred with a Pasteur pipette to the experimental
chamber. The chamber was a 35-mm plastic petri
dish, specially modified by cutting out a hole about
14 mm in diameter in its bottom and cementing a
22-mm microscope glass coverslip under the hole
with paraffin. This greatly improved the optics of the
chamber. The dish wasfilledwith oxygenated BSS at
37°C, and the ciliary process was placed on its side in
the center of the chamber. In order to improve the
mechanical stability of the preparation, a 1-mm-wide
sliver of glass coverslip was placed on top of the ciliary process leaving only the edge of the process uncovered (Fig. 1B). A plastic insert was placed into the
petri dish to provide an input and output for the
perfusion system (Fig. 1C) and reduce the effective
volume. The chamber volume was 0.5 ml.
The tissue was perfused with BSS at a rate of 0.5
ml/sec. Bottles containing the solutions were placed
in a constant temperature water bath, and the temperature of this bath was adjusted so that the temperature of the Ringer when it reached the chamber was
37°C ± 1°C. The design of the perfusion system was
similar to that of Hodgkin et al.15 The selection of the
solution bathing the tissue could be made by using a
solenoid which activated a remote-controlled pneumatic valve. This valve was in turn connected to a
Hamilton four-way flow valve, which was constructed so that as one solution was flowing to the
chamber, another was passing to a waste container.
Since the Hamilton valve was placed near the
chamber, the dead space in the perfusion system was
only about 200 n\, equivalent to a dead time of about
400 msec.
We give the composition of the solutions used in
these experiments in Table 1. All solutions were satu-
935
Iris
Secondary
Processes
Ora Serrata
Cut Edge
C
Inflow
Outflow
0.5 ml Volume
\
7\
Plexiglass
Perfusion
. Insert
Plastic Petri Dish
Glass Coverslip
Water Immersion
Microscope Objective
Fig. 1. Schematic view of dissection and preparation. (A) Pieshaped section of posterior surface of rabbit anterior segment after
removal of lens showing anatomy of ciliary body in rabbit. Area in
box shows approximate region of secondary process removed for
experimentation. (B) Single ciliary process laid on its side and
covered with a sliver of coverslip. Note single layer of non-pigmented cells at the edge of the process (cells not to scale). (C)
Cut-away view of chamber. Preparation with sliver of coverslip can
be seen just above microscope objective and glass coverslip. Inflow
and outflow refer to inlets and outlets for perfusion system.
rated with 95% O2/5% CO2 and were at pH 7.1 Although our solutions contained less Ca2+, PO42~, and
HCO3~ and were at a lower pH than rabbit aqueous
humor,16 these alterations produced no noticeable
change in the swelling and volume recovery and
greatly decreased the likelihood of divalent ion precipitation from our solutions. Solutions containing
Ba2+ were made simply by adding 2 mM BaCl2. This
increased the measured osmolarity by 5-6 mosm. For
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / June 1987
936
Vol. 28
Table 1. Solution composition
Solution
A
B
C
D
E
F
G
H
I
J
NaCl
126
30
78
98
30
78
98
30
30
26.2
KCl
3.82
3.82
3.82
3.82
3.82
3.82
3.82
3.82
3.82
7.6
NaHCOs
12
12
12
12
12
12
12
12
12
12
CaCh
0.311
0.311
0.311
0.311
0.311
0.311
0.311
0.311
0.311
0.311
MgSO4
Mannitol
Glucose
Osmolarity
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
_
5
5
5
5
5
5
5
5
5
5
268
89
187
219
268
266
268
172
212
89
—
179
93
59
86
120
—
All solutions contained 0.1% phenol red and were saturated with 95% O 2 /5% CO 2 (pH 7.1).
these solutions, we found it necessary to replace the
MgSO4 with MgCl2 to prevent precipitation of
BaSO4. Substitution of MgSO4 with MgCl2 had by
itself no effect on swelling or the time course of volume regulation. In some experiments, ouabain
(Sigma Chemical Co., St. Louis, MO) was added to
the solutions at a concentration of 100 nM.
We placed the chamber containing the ciliary process on the stage of a Zeiss inverted microscope (Invertoscope D, Carl Zeiss, Oberkochen, West Germany) and focused on the very edge of the tissue with
a 40X Zeiss water immersion objective modified for
Hoffman modulation contrast optics (Fig. 1C). The
microscope was fitted with a trinocular tube, so that
the image could be projected onto a Hitachi
HV-625U TV camera (Hitachi, Denshi, Japan). The
image was displayed on a TV monitor and recorded
on video tape. After each experiment was completed,
the video tape was replayed, and the width of the
non-pigmented cell layer was measured directly from
the screen of the TV monitor. These measurements
were calibrated (in micrometers) using a stae micrometer. Since the field of the water-immersion objective was not entirely flat, and since the TV camera
introduced some geometric distortion, the measured
width of an object in the field could vary by as much
as 5% when this object was moved from one side of
the field to the center. For this reason, we attempted
to keep the tissue within the center of the field of the
objective, even when the tissue was swelling.
Data were analyzed on an IBM AT computer
(IBM, Boca Raton, FL). The rate of increase of the
non-pigmented cell width during swelling was estimated by fitting a straight line to the cell width increase during its rising phase (from 0.1 to 0.9 of the
total width increase). The rate of volume regulatory
decrease was calculated by fitting an exponential
function to the width of the non-pigmented cell layer
after the width had reached its maximum value (ie,
during the decay phase). This function was of the
form:
w0
w=
(1)
where w was the cell width at time t, to was the time at
which the width began to decay, w^ was the cell
width at long times after exposure to hypotonic solution, A was the total change in cell width during volume regulation (ie, difference between the maximum
width and the steady state width in hypotonic solution, w^) and T was the time constant of volume
regulatory decrease. The initial rate of regulatory decrease could be calculated as -dw/dt at t = to and can
be shown to be equal to A/T.
Results
When the isolated ciliary process was perfused with
hypotonic Ringer, the volume of the non-pigmented
cells quickly enlarged, and the cells then slowly regulated their volume back towards its initial value. The
series of photographs in Figure 2 show the results of a
typical experiment, beginning with Figure 2A in normal BSS (Table 1, solution A). At the edge of the
tissue, there is an optically bright band which we believe is formed by one or at most two layers of nonpigmented epithelial cells. This can be seen more
clearly in Figure 2B, which shows the samefieldas in
Figure 2A but 23 sec after perfusing with a Ringer
(solution B) having an osmolarity 33% of that of normal BSS. The cells have swollen to twice their normal
width. An especially clear outline of a single cell is
indicated by the arrow. In Figures 2C-E, we show the
samefieldstill in hypotonic Ringer but 1 min, 3 min,
and 5 min, 14 sec after the solution change. There is a
slow decrease in the width of the cells (or volume
regulatory decrease), produced presumably by the
cells extruding salt and water. When the solution was
changed back to normal BSS, the volume of the cells
decreased to below that observed initially (Figure
2F), probably because of a net loss of salt during the
volume regulatory response. The cells assumed a
darkened, crenated appearance, which we believe is
caused by a collapsing and shrinking together of the
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
No. 6
VOLUME REGULATION OF NON-PIGMENTED CELL5 / Forohbokhsh and Fain
937
Fig. 2. Swelling and volume regulation of non-pigmented cell layer in hypotonic medium. All pictures are from the same microscope field
taken during the same experiment and are shown at an identical magnification (640X) (A) normal Ringer.(B, C, D, E) at 23 sec, 1 min, 3 min,
and 5 min, 14 sec after changing to the 33% hypotonic Ringer (solution B). Cells are perfused for 8 min in this Ringer and then returned to
normal Ringer. (F) taken 1 min 30 sec after change back into normal Ringer. Arrow in (B) shows a clearly identifiable single cell within the
non-pigmented layer.
basal and lateral interdigitations. When the cells were
perfused with normal BSS for an additional 10-15
min, they recovered their normal size and appearance, presumably by taking up salt and water by a
mechanism we have not as yet investigated.
The time course of the changes in the width of the
non-pigmented cell layer in response to hypotonic
shock is shown in Figure 3. The squares show the
response in a 33% hypotonic Ringer as in Figure 2.
The ordinant gives the width of the non-pigmented
cell layer in /um and the abscissa gives time in min.
Note the rapid swelling and slower volume regulatory
decrease. At 8 min, the width was only slightly greater
than at the beginning of the experiment. Upon the
return to normal Ringer there was a rapid undershoot, followed by a recovery to the initial volume.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
938
INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / June 1987
T i me
[mini
Fig. 3. Time course of change in width of non-pigmented cell
layer in response to hypotonic shock. Ordinate is width of non-pigmented cell layer in /im. Abscissa is time in min. Squares show
response to 33% hypotonic Ringer (solution B), and circles to 70%
hypotonic Ringer (solution C). For both squares and circles, the
beginning of solution change to hypotonic medium was positioned
at time zero and hypotonic Ringer was perfused for 8 min. Perfusate was then changed back to normal Ringer. Both squares and
circles are from same preparation.
The circles show a similar experiment on the same
tissue but with a 70% hypotonic Ringer (solution C).
The swelling was smaller, but the general features of
the response were similar. If the ciliary body was exposed to hyper-tonic solution, the non-pigmented
cells shrank and showed a volume regulatory increase
(not shown). This response was more difficult to
measure because of the smaller change in cell width,
and we have not as yet studied it in detail.
We have used experiments like those of Figures 2
and 3 to perform two different kinds of measurements. On the one hand, the rate of volume increase
during cell swelling was used to estimate the hydraulic conductivity (or osmotic permeability) of the nonpigmented epithelial cells. On the other, we have used
the rate of volume regulatory decrease and the effects
of drugs and ion concentration changes on this response to investigate the mechanisms of ion and
water transport in ciliary epithelium.
Hydraulic Conductivity of Non-Pigmented
Epithelial Cells
In the absence of an applied hydrostatic pressure
gradient, the flow of water across a semipermeable
membrane is proportional to the osmotic concentration difference according to:
J v = -Lpff s RTAc s
(2)
where Jv is the rate of water flow, Lp is the hydraulic
conductivity, R is the gas constant, T is the tempera-
Vol. 28
ture, and Acs is the difference in solute concentration
across the membrane.1718 The negative sign indicates
that water flows from low to high solute concentration. The constant <rs in (2) is the reflection coefficient
of the solute, which is a measure of solute permeability through the membrane. For an ideal (ie, impermeant) solute, <rs = 1.
It would be possible to use (2) together with our
measurements of the rate of cell volume increase to
calculate the hydraulic conductivity of the non-pigmented cell plasma membrane, provided the reflection coefficient of the solute was known. In the experiments of Figures 2 and 3, the osmotic gradient
was produced by changing the Na concentration.
Since the reflection coefficient of Na for the non-pigmented cells is not known, it was necessary for us to
repeat these experiments using a different solute,
mannitol, which is generally impermeant.19 We first
perfused the ciliary process with one of solutions E, F,
or G. These solutions resembled the hypotonic solutions B, C, and D (see Table 1), except that sufficient
mannitol was added to bring the osmolarity up to the
same value as in the normal BSS. We then produced
the osmotic shock by removing the mannitol; that is,
solution E was changed to B, F to C, or G to D. For
the 70% and 82% hypotonic shocks, in some experiments we first perfused with solution E and then
changed to solution H or I. This method had the
virtue of keeping the Na concentration in all three
sets of hypotonic solutions at the same level. However, in practice we found that the rate and extent of
cell swelling was not significantly altered by differences in Na concentration, suggesting that the permeability of the cells to Na may be rather low.
The results of these experiments are given in Table
2. For the calculations of hydraulic conductivity, we
have assumed that the basal membrane of the nonpigmented cells is smooth; that is, we have made no
attempt to take into account the membrane area of
the basal and lateral interdigitations, both because
careful measurements are not available for their total
membrane area, and because measurements of hydraulic conductivity from the whole ciliary body have
also assumed a smooth basal membrane.4'58 We show
Lp for three sets of experiments with 82%, 70%, and
33% hypontonic Ringers. Although our data seem to
show a progressive increase in Lp as the osmotic gradient is increased, the standard deviations are large
and the differences in means are significant only between the 82% and 33% data (t-test, P < 0.006). In
addition to Lp, we also give the osmotic permeability
(POs), which is equal to LpRT/Vw- where Vw is the
partial molar volume of water.18 As we explain below
(see Discussion), the values for Lp and Pos in Table 2
probably represent lower limits.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
No. 6
VOLUME REGULATION OF NON-PIGMENTED CELLS / Farohbokhsh o n d Foin
From the hydraulic conductivity and the total surface area of the ciliary body,20 it was possible to calculate a lower limit for the total volume flow from the
non-pigmented cells per milliosmole of osmotic gradient. Values for volume flow calculated from the
results of our experiments are given in the sixth row
of Table 2. From these values, we also calculated an
upper limit for the amplitude of the standing gradient21 which would be required across the basal
membrane of the non-pigmented cells for each fi\/
min of aqueous flow from the ciliary body.
Effects of Ouabain and Ba2+ on Rate of
Volume Regulation
Since during the volume regulatory decrease, the
non-pigmented cells extrude salt and water as they
are thought to do during the secretion of aqueous
humor,22 it seemed possible that the mechanism of
volume regulation might be similar to the mechanism of aqueous formation. Aqueous inflow has been
shown to be inhibited by ouabain2023 and vanadate,24
which are known to block the Na+-K+ ATPase. For
this reason, we investigated the effect of ouabain on
volume regulation.
In these experiments (see Fig. 4), wefirstrecorded
response of the non-pigmented cells to a 33% hypotonic solution (solution B—see Fig. 4, squares). The
time course of the regulatory response wasfittedwith
an exponential function (1), as described in the
methods. The regulatory response in Figure 4 had a
best-fitting time constant of 4.8 min and an initial
rate of volume decrease of 3.9 /xm/min. The average
time constant for the regulatory response calculated
for all our experiments with 33% hypotonic solution
was 3.3 ± 0.4 min (mean ± SE, N = 51), and the
average initial rate of volume decrease was 8.0 ± 1.0
jum/min. It is of some interest that the mean initial
rate of volumeflowfrom the cells was equivalent to a
rate of fluid movement from the epithelium as a
whole of 4.6 ix\/mm, which is as large or larger than
the normal rate of aqueous inflow in rabbit.22
We next perfused for 10 min with normal BSS
containing 100 nM ouabain. In some cases (as in Fig.
4), this produced a slight shrinkage of the width of the
non-pigmented epithelial cell layer (notice the decrease in the value of the width at time zero for the
circles), but this was not a consistent finding. We then
re-exposed the non-pigmented cells to the 33% hypotonic solution (Fig. 4, circles). In some experiments
this solution also contained 100 nM ouabain, but in
others the ouabain was left out with no difference in
the results. Exposure to ouabain produced a significant inhibition in volume regulation. The time constant of volume regulatory decrease increased to 14.7
939
Table 2. Hydraulic conductivity of
non-pigmented cells
700% to
82%
Osmolarity
change
(mosmol)
N
Rate of swelling
Omi/min)
Lp (/im/sec-osmol)
PM (/im/sec)
Total volume flow
from epithelium
(Ml/minmosmol)
Gradient for 1 \i\l
min (mosmol)
49
14
4.6 ±2.3
1.56 ±0.77
86 ± 4 3
0.0574 ±0.027
18.6
100% to
70%
81
9
100% to
33%
179
63
10.8 ±5.1
2.23 ±1.04
123 ±57
25.5 ±11.0
2.37 ±1.02
131 ±56
0.076 ±0.036
0.081 ±0.035
13.1
12.3
Values given are means plus and minus standard deviations. N gives the number of
separate experiments for each determination. The rate of swelling of the non-pigmented
cell layer was calculated by fitting a straight line to the rising phase of the width increase
(see Materials and Methods). Lp was calculated from the rate of swelling and the osmolarity change, assuming that the cross-sectional area of the non-pigmented cells remained
fixed. The area of the basal (lumen facing) membrane of the non-pigmented cell was
assumed to be the same as the cross-sectional area. PM was calculated by dividing Lp by
1.87 X 10~2 osmol" 1 ." The total volumeflowfrom the epithelium was estimated from the
total area of the ciliary body (5.72 cm2)20 and the rate of swelling. The final row gives the
amplitude of the osmotic gradient across the basal membrane of the non-pigmented cells
required for each pl/min of aqueous inflow.
± 6.7 min (mean ± SE, N = 7), and the initial rate of
volume recovery decreased to 1.8 ± 0.4 ^m/min. The
effect of ouabain could not be reversed at this ouabain concentration even after as much as 1 hr of
perfusion with normal Ringer."
T i me
[mini
Fig. 4. Effect of ouabain on volume regulation. Squares and
circles show change in width of non-pigmented cell layer from
same preparation in response to a 33% hypotonic Ringer (solution
B) before and immediately after a 10 min exposure to 100 nM
ouabain. Ordinate is width of non-pigmented cell layer in nm, and
abscissa is time in min. For both squares and circles, the beginning
of the solution change to the hypotonic Ringer was positioned at
time zero, and the solution was changed back to normal Ringer
after 8 min.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
940
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1987
0
4
8
T i me
12
16
20
[mini
Fig. 5. Effect of Barium on volume regulation. Change in width
of non-pigmented cell layer from same preparation for exposure to
three hypotonic solution. For all three, the beginning of the solution change to the hypotonic Ringer was positioned at time zero,
and the solution was changed back to normal Ringer (solution A)
after 8 min. Squares show first solution change to 33% hypotonic
Ringer without Ba (solution B), circles show second solution
change 20 min later to 33% hypotonic Ringer with 2 mM Ba, and
triangles show third solution change 20 min after second to 33%
hypotonic Ringer without Ba (solution B).
The inhibition of volume regulation by ouabain
suggests that some part of the volume regulatory decrease may be the result of active extrusion of Na+
(presumably as NaCl or NaHCO3). However, in
12
Time
16
20
[mini
Fig. 6. Combined effect of ouabain and Ba. Change in width of
non-pigmented cell layer from same preparation for exposure to
33% hypotonic solution. For all three, the beginning of the solution
change to the hypotonic Ringer was positioned at time zero, and
the solution was changed back to normal Ringer after 8 min.
Squares showfirstsolution change to 33% hypotonic Ringer (solution B) in the absence of ouabain and Ba. Circles show second
solution change to 33% hypotonic Ringer (solution B) after 10 min
treatment with 100 /xM ouabin in normal Ringer. Triangles show
third solution change 20 min after second to 33% hypotonic Ringer
containing 100 nM ouabain and 2 mM BaCl2.
Vol. 28
many cells,25 including those of epithelia,26 volume
regulatory decreases are produced by swelling-induced increases in K efflux. In order to test for a
possible role for K efflux in the regulatory response of
the non-pigmented cells, we perfused ciliary processes with hypotonic media containing Ba2+. Ba2+
has been shown to block K+ conductances in nerve27
and epithelium28"30 and has recently been postulated
to block the K+ conductance of ciliary epithelial
cells.31
The effects of Ba2+ on the swelling and volume
regulatory decrease of the non-pigmented cells are
illustrated in Figure 5. We first measured the change
in width of the non-pigmented cells in response to a
33% hypotonic medium without Ba2+ (solution B) as
a control (Fig. 5, squares). Afterwards, the perfusate
was changed back to normal Ringer to permit the
width of the cells to return to their initial value. We
then changed the perfusate to a 33% hypotonic
Ringer to which we had added 2 mM BaCl2 (Fig. 5,
circles). The increase in the width of the non-pigmented cells was larger, and the rate of the volume
regulatory decrease was substantially reduced. The
time constant for the regulatory response increased to
11.4 ± 2.0 min (N = 3), and the initial rate of volume
recovery decreased to 2.1 ± 0.4 ^m/min. After returning the perfusate to normal Ringer, we re-exposed the tissue to the hypotonic solution without
Ba2+ (Fig. 5, triangles). The effect of Ba2+ was completely reversible.
The results of Figures 4 and 5 show that either the
Na+-K+ ATPase inhibitor ouabain or the K+ channel
blocker Ba2+ can reduce the rate of volume regulatory
decrease. An even greater effect on the volume regulatory response is seen if both Ba2+ and ouabain are
added together, as shown in Figure 6. The tissue was
first exposed to a 33% hypotonic Ringer (squares) and
it was then perfused with ouabain for 10 min. As we
have previously shown, ouabain greatly reduces the
rate of volume regulation, and the response to a 33%
hypotonic Ringer (solution B) after ouabain, shown
as the circles in Figure 6, resembled the time course of
volume decrease in Figure 4. The ciliary process was
then perfused with a 33% hypotonic Ringer containing 2 mM BaCl2 (Fig. 6, triangles). This solution also
contained 100 ^M ouabain. Ba2+ after ouabain exposure produced a nearly complete block of volume
regulation. Similar though somewhat smaller effects
were seen in three other experiments.
Effects of Elevated K+ on Rate of Volume Regulation
Since the effects of ouabain and Ba2+ seemed to be
additive, our results suggested that the volume regulatory decrease may have two components: one pro-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
No. 6
941
VOLUME REGULATION OF NON-PIGMENTED CELLS / Forohbokhsh and Fain
duced by active Na+ extrusion, and the other, by a
swelling-induced increase in K permeability. There
is, however, an alternative interpretation. It would be
possible that volume regulation was produced only
by K efflux and that Ba2+ reduced this efflux by
blocking the K conductance. Ouabain, on the other
hand, might produce its effect simply by decreasing
the intracellular K concentration and therefore the
gradient for K exit. Ouabain has been shown to produce a rapid depolarization of the ciliary epithelial
cell membrane potential11-31 presumably, at least in
part, as the result of a decrease in internal K. Reduction of the K gradient would reduce the efflux of K
during the volume regulatory decrease.
In order to test this alternative hypothesis, we examined the effect of changes in the extracellular K
concentration on the regulatory responses. We reasoned as follows. A 10 min exposure to 100 nM ouabain (as in Figs. 4 and 6) has been shown to decrease
the resting potential of the ciliary epithelial cells by a
mean value of 15 mV (T. C. Chu and O. A. Candia,
personal communication). Since the membrane potential of the ciliary epithelial cells is determined for
the most part by the K equilibrium potential," this
ouabain exposure would be expected to produce an
approximately two-fold decrease in K gradient. To
examine the effect of such a change in K gradient on
the volume regulatory decrease, we exposed the ciliary epithelial cells in the absence of ouabain to twice
the normal extracellular K.
This experiment was done in the following way (see
Fig. 7A). We first measured the change in width of
the non-pigmented cells in response to a 33% hypotonic solution as a control (Fig. 7A, squares). The
response was then measured to a 33% solution for
which the extracellular K+ was doubled (from 3.82 to
7.6 mM) by substituting for some of the Na (solution
J) (Fig. 7A, circles). Finally, we remeasured the control response to the 33% solution containing the normal K concentration (Fig. 7A, triangles). The results
in Figure 7 are representative of five experiments in
which we doubled the extracellular K concentration.
In these experiments, the mean time constant (r) for
the regulatory response was 2.8 ± 0.7 min in normal
Ringer and 2.4 ± 0.7 in high K (mean ± SE). Corresponding values for the initial rate of volume regulation were 10.6 ± 3.4 and 16.1 ± 6.7 ^m/min, respectively. These are not significantly different.
The results of these experiments show that there is
no significant effect on the time constant and rate of
the volume regulatory decrease produced by doubling the extracellular K concentration. This result
may seem surprising, because if part of the volume
regulatory decrease were produced by K efflux, a decrease in the K gradient should be expected to have
0
A
8
T i me
0
A
8
T i me
12
16
20
16
20
[mini
12
[mini
Fig. 7. (A) Volume regulation in elevated K. Change in width of
non-pigmented cell layer from same preparation for exposure to
three hypotonic solutions. For all three, the beginning of the solution change to the hypotonic Ringer was positioned at time zero,
and the solution was changed back to normal Ringer after 8 min.
Squares show first solution change to 33% hypotonic Ringer with
normal K (solution B), circles show second solution change 20 min
later to 33% hypotonic Ringer with twice normal K (solution J),
and triangles show third solution change 20 min after second to
33% hypotonic Ringer with normal K (solution B). (B) Volume
regulation in elevated K after ouabain exposure. All responses are
from same preparation, exposed at the beginning of the experiment
for 10 min to 100 nM ouabain. Protocol for experiment and format
of figure are identical to those in (A). Three sets of symbols show
change in width of non-pigmented cell layer for exposure to 33%
hypotonic Ringer,first(squares) containing normal K (solution B),
then 20 min later (circles) containing twice normal K (solution J),
and finally another 20 min later (triangles) containing normal K
again. Solutions did not contain ouabain. Note that squares and
triangles are responses to the identical 33% hypotonic Ringer solution, but squares were recorded immediately after ouabain exposure and triangles 40 min later.
some effect. A possible resolution emerges from the
experiment of Figure 7B, where we used exactly the
same protocol as in Figure 7A except that the ciliary
process was first perfused for 10 min with Ringer
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
942
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1987
containing 100 nM ouabain. Doubling the external K
concentration then produced a clear decrease in the
volume regulatory response. If the volume regulatory
decrease consists of two components, a ouabain-sensitive Na+ extrusion and a swelling-induced increase
in K+ efflux, then doubling the K concentration in
the absence of ouabain (Fig. 7 A) may fail to alter the
rate of volume regulatory decrease because it affects
these mechanisms in opposite directions. That is, increasing extracellular K may accelerate. Na-K pumping32 but decrease K efflux. These effects may cancel,
leaving the net rate of volume decrease unchanged.
Discussion
We have described a new method for investigating
ion and water transport of the non-pigmented cells of
the ciliary epithelium, a method based upon measurements of the width of the layer of cells at the edge
of an isolated ciliary process. The advantages of this
method are that the preparation of the tissue is simple
and straightforward, and the results in most cases are
readily reproducible. We believe this method may be
of considerable use in the study of the physiology and
pharmacology of the ciliary body. Our method could
also be used (in conjunction with pH- or Ca-sensitive
dyes) to measure proton transport and changes in free
Ca concentration in the non-pigmented cells.
Our method has, however, some important limitations. For one, we have in general no way of distinguishing the properties of the apical and basal membranes of the non-pigmented cells. Furthermore,
since the two epithelithal cell layers are tightly coupled through an extensive network of gap junctions," 12 we cannot in general distinguish the properties of membranes of the pigmented and non-pigmented layers. However, in the experiments we have
reported here, we believe that the basal membrane of
the non-pigmented cells (the membrane facing the
posterior chamber) is directly responsible for swelling
and volume regulation. In the experiments where we
measure the rate of swelling of the cells to estimate
the hydraulic conductivity (Table 2), the basal membrane on the non-pigmented cells is likely to be the
primary locus of water influx, both because the area
of the basal membrane is considerably larger than the
apical membrane of the non-pigmented cells or either
membrane of the pigmented cells, and because the
basal side of the non-pigmented cells is directly exposed to the bathing medium. During a solution
change, the change in the composition of the medium
would occur rapidly at the mucosal (luminal) surface
of the ciliary process but much more slowly at the
serosal. Although salts and water could conceivably
pass from one side of the epithelium to the other
Vol. 28
through the junctional complex, a change in the
composition of the bulk phase at the serosal surface
would take place only if there were considerable diffusion into the fine capillaries which invest the ciliary
process.33 The rate of this diffusion is likely to be
considerably slower than the rate of cell swelling
which we have observed. Ouabain-sensitive transport
can also be preliminarily attributed to the basal
membrane, since the Na-K ATPase is concentrated
here.34"36 The location of the K permeability cannot
as yet be determined with certainty; however, the
speed of the Ba2+ effect on swelling and volume regulation (see Figs. 5 and 6) suggests that the K permeability may also be on the non-pigmented basal
membrane.
Another limitation of our method is that we measure the width of the layer of non-pigmented cells
rather than the volume of single cells. In practice in
any given experiment, we always measure the width
at the same place in the field during cell swelling and
volume regulation, so that we are probably measuring changes in the long axis of a single cell. However,
it is clear from our recordings that cells in hypotonic
media swell in cross-sectional area as well as in
length. Our measurements of the rate of cell swelling
are therefore likely to be underestimates. An indication of the extent of this error is given by the following considerations. In the absence of solute movement into or out of the non-pigmented cells, the extent of volume swelling is given by a modified
Boyle-van't Hoff relationship:
TT(V - b) = 7r°(v° - b)
(3)
0
where TT and v° are the osmotic pressure and volume
in isotonic Ringer, TT and v are the osmotic pressure
and volume in the hypotonic medium, and b is the
"non-solvent" volume, the part of the cell volume
which is not osmotically active.18 Although b is not
known for the non-pigmented cells, in other cells b is
of the order of 10-20% of v0.18 The total volume
increase for a 33% hypotonic shock should therefore
be of the order of 160-180%. Since in our experiments the width of the non-pigmented cells approximately doubles in this solution, measurement of
width alone is likely to underestimate the total volume change by the order of 40%. We hope in the
future to be able to improve the optics of our apparatus and use image analysis to measure changes in
total volume for the non-pigmented cells.
Hydraulic Conductivity
Our estimate of the hydraulic conductivity of the
non-pigmented cells is nearly an order of magnitude
smaller than for epithelial cells of the gall bladder19'37
but is likely to be too low; first, because we do not
measure the total volume of the cells; second, because
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
No. 6
VOLUME REGULATION OF NON-PIGMENTED CELLS / Forahbokhsh and Foin
some K efflux may occur as the tissue swells (cf. Fig.
5); and third, because our measurements may be limited by unstirred layers adjacent to the basal membrane.38 We have observed (unpublished) that the
rate of cell swelling increases with the rate offlowof
our perfusion system. It therefore seems possible that
we would obtain larger values for Lp and Pos with
even faster flow rates or an even smaller chamber.
The swelling of the cells does not seem to be limited
by the presence of the basal lamina,3940 since when
this was partially removed by 10 min incubation in 1
mg/ml collagenase (Worthington, type I, Worthington Diagnostic Systems, Freehold, NJ), there was no
significant increase in the swelling rate (unpublished
observation). Treatment with collagenase also had no
effect on the volume regulatory decrease or its inhibition by ouabain. However, the basal lamina did actually seem to have been removed by this treatment,
since the basal membrane of cells exposed to hypotonic solutions showed large numbers of 1-4 /urn-diameter spherical evaginations after collagenase, apparently as the result of the swelling and eversion of
the basal interdigitations. These gradually disappeared as the cell volume decreased during the regulatory response.
It is of some interest to compare our measurements
of Lp and Pos for the non-pigmented cells with the
hydraulic conductivity of the whole ciliary body,
measured by applying hydrostatic pressure differences across the tissue mounted in an Ussing
chamber. These measurements give a Pos of 109 /xm/
sec8 to 4600 Mm/sec.45 Since this method of measuring the hydraulic conductivity is likely to overestimate the true value, as the result of tissue damage or
water flow at the edges of the preparation8 and pressure-induced changes in osmotic permeability41; and
since our measurements of Pos (86-131 /im/sec) are
likely to be underestimates, it seems possible that the
rate of waterflowacross the whole of the ciliary body
is determined by the rate of flow across the membranes of the non-pigmented epithelial cells. Since
the capillaries of the ciliary body are fenestrated,1
there are no significant impediments for water movement between the blood vessels and tight junctions of
the junctional complex. The nearness of our estimate
for the osmotic permeability of the non-pigmented
epithelial cells to those for the ciliary body as a whole
suggests that the major route for water flow in the
ciliary epithelium may be trans-cellular rather than
paracellular. A similar conclusion has been reached
for the gall bladder.19
Our estimates of hydraulic permeability show that
the standing gradient of osmolarity across the basal
membrane need be at most 50-80 msom to account
for a flow of 4.5 ix\/min of aqueous humor, a typical
value for the normal inflow rate in rabbit.22 This is
943
equivalent to a 25-40 mM concentration difference
of NaCl. This estimate is independent of the value we
use for the surface area of the basal membrane. It is,
however, inversely proportional to Lp. Since we believe Lp to be underestimated, the actual value of the
concentration gradient required to support aqueous
secretion is probably considerably less than given in
Table 2. The basal membrane of the non-pigmented
cells is highly infolded,42'43 and it seems likely that
active secretion of salt into the spaces between the
lateral and basal interdigitations could produce a sufficiently large standing gradient to account for the
formation of most if not all of the aqueous humor.22
Mechanism of Volume Regulation
Our experiments provide evidence for two mechanisms for the volume regulatory decrease of non-pigmented epithelial cells: a ouabain-sensitive Na extrusion via the Na-K ATPase, and a Ba2+-sensitive increase in K efflux. We have shown that a 10 min
exposure of the ciliary body to 100 /iM ouabain reduces the rate of the volume regulatory decrease (Fig.
4). The potassium channel blocker Ba2+ also reduces
the rate of volume regulation. These effects are to
some degree additive, since both ouabain and Ba2+
together inhibit volume regulation more than either
alone (Fig. 6). This suggests that the components of
the response which are blocked by ouabain and Ba2+
represent at least partially independent mechanisms
of volume regulatory decrease by the non-pigmented
cells.
An alternative interpretation of our experiments is
that the volume regulatory decrease is produced entirely by K efflux, and that Ba2+ and ouabain reduce
this efflux in different ways: Ba2+ by blocking the
membrane K permeability, and ouabain by reducing
the K concentration gradient. We have attempted to
exclude this possibility by showing that decreases in
K gradient similar to those which we estimate to be
produced by ouabain exposure have no significant
effect on volume regulation (Fig. 7A). These experiments are not conclusive, since their interpretation
depends to some extent upon our assumption that a
10 min ouabain exposure decreases the K gradient by
a factor of no more than two. Though doubling external K does not significantly alter the rate of volume regulation, we have shown in other experiments
(unpublished) that larger increases in extracellular K
have pronounced effects on volume decrease, and a
complete substitution of K for Na in the hypotonic
Ringer not only eliminates volume regulation but
also causes the non-pigmented cells actually to swell.
In order to quantify separately the roles of K+ and
Na+ in the volume regulatory response, it will be necessary to measure the fluxes of these ions during
changes in cell volume, for example with ion-sensi-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017
944
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1987
tive electrodes. Experiments of this kind are in preparation.
Key words: Ciliary body, epithelium, non-pigmented cells,
volume regulation, hydraulic conductivity, aqueous
humor, glaucoma, Na-K ATPase, ouabain.
Acknowledgments
The authors thank J. A. Fox and R. H. Steinberg for
useful discussion.
References
1. Raviola G: Morphological aspects of aqueous humor production. In The Structure of the Eye, Hollyfield JG, editor. Amsterdam, Elsevier North Holland, 1982, pp. 331-340.
2. Okisaka S, Kuwabara T, and Rapoport SI: Effect of hyperosmotic agents on the ciliary epithelium and trabecular meshwork. Invest Ophthalmol 15:617, 1976.
3. Cole DF: Transport across the isolated ciliary body of ox and
rabbit. Br J Ophthalmol 46:577, 1962.
4. Green K and Pederson JE: Contribution of secretion and filtration to aqueous humor formation. Am J Physiol 222:1218,
1972.
5. Green K and Pederson JE: Aqueous humor formation. Exp
Eye Res 16:273, 1973.
6. Watanabe T and Saito Y: Characteristics of ion transport
across the isolated ciliary epithelium of the toad as studied by
clinical measurements. Exp Eye Res 27:215, 1978.
7. Kishida K, Sasabe T, Manabe R, and Otori T: Electrical characteristics of the isolated rabbit ciliary body. Jpn J Ophthalmol
25:407, 1981.
8. Brodwall J and Fischbarg T: The hydraulic conductivity of
rabbit ciliary epithelium in vitro. Exp Eye Res 34:121, 1982.
9. Burstein NL, Fischbarg J, Liebovitch L, and Cole DF: Electrical potential, resistance, and fluid secretion across isolated ciliary body. Exp Eye Res 39:771, 1984.
10. Krupin T, Reinach PS, Candia OA, and Podos SM: Transepithelial electrical measurements on the isolated rabbit iris-ciliary body. Exp Eye Res 38:115, 1984.
11. Green K, Bountra C, Georgiou P, and House CR: An electrophysiologic study of rabbit ciliary epithelium. Invest Ophthalmol Vis Sci 26:371, 1985.
12. Raviola G and Raviola R: Intercellular junctions in the ciliary
epithelium. Invest Ophthalmol Vis Sci 17:958, 1978.
13. Berggren L: Direct observation of secretory pumping in vitro
of the rabbit eye ciliary processes: Influence of ion milieu and
carbonic anhydrase inhibition. Invest Ophthalmol 3:266,
1964.
14. Farahbakhsh NA and Fain GL: Active water extrusion by
non-pigmented ciliary epithelium cells in vitro. ARVO Abstracts. Invest Ophthalmol Vis Sci 27(Suppl):177, 1986.
15. Hodgkin AL, McNaughton PA, Nunn BJ, and Yau K-W: Effect of ions on retinal rods from Bufo marinus. J Physiol
350:649, 1984.
16. Davson H: The intraocularfluids.In The Eye, Vol. 1, Vegetative Physiology and Biochemistry, Davson H, editor, 2nd edition. New York, Academic Press, 1969, pp. 67-186.
17. Katchalsky A and Curran PF: Non-equilibrium Thermodynamics in Biophysics. Cambridge, MA, Harvard University
Press, 1967, pp. 113-132.
18. House CR: Water Transport in Cells and Tissues. London,
Edward Arnold, 1974, pp. 36-76.
19. Persson B and Spring KR: Gallbladder epithelial cell hydraulic
water permeability and volume regulation. J Gen Physiol
79:481, 1982.
Vol. 28
20. Cole DF: Aqueous humor formation. Doc Ophthalmol
21:116,1966.
21. Diamond JM and Bossert WH: Standing gradient osmotic
flow: A mechanism for coupling water and solute transport in
epithelia. J Gen Physiol 50:2061, 1967.
22. Cole DF: Ocularfluids.In The Eye, Vol. la, Vegetative Physiology and Biochemistry, Davson H, editor, 3rd edition. New
York, Academic Press, 1984, pp. 269-390.
23. Bonting SL and Becker B: Studies on sodium-potassium activated adenosine triphosphatase. Invest Ophthalmol 3:523,
1964.
24. Becker B: Vanadate and aqueous humor dynamics. Proctor
Lecture. Invest Ophthalmol Vis Sci 19:1056, 1980.
25. Siebens AW: Cellular volume control. In The Kidney: Physiology and Pathophysiology, Seldin DW and Giebisch G, editors.
New York, Raven Press, 1985, pp. 91-115.
26. Larson M and Spring KR: Volume regulation by Necturus
gallbladder: Basolateral KC1 exit. J Membr Biol 81:219, 1984.
27. Hille B: Ionic Channels of Excitable Membranes. Sunderland,
MA, Sinauer Associates, 1984.
28. Garcia-Diaz JF, Nagel W, and Essig A: Voltage-dependent K
conductance at the apical membrane of Necturus gallbladder.
Biophys J 43:269, 1983.
29. van Driessche W and Zeiske W: Ionic channels in epithelial
cell membranes. Physiol Rev 65:833, 1985.
30. Griff ER, Shirao Y, and Steinberg RH: Ba2+ unmasks K+ modulation of the Na+-K+ pump in the frog retinal pigment epithelium. J Gen Physiol 86:853, 1985.
31. Candia OA, Iizuka S, and Chu T-C: Intracellular recordings in
the isolated ciliary epithelium. ARVO Abstracts. Invest Ophthalmol Vis Sci 27(Suppl): 178, 1986.
32. DeWeer P: Cellular sodium-potassium transport. In The Kidney: Physiology and Pathophysiology, Seldin DW and Giebisch G, editors. New York, Raven Press, 1985, pp. 31-48.
33. Morrison JC and Van Buskirk EM: Comparative microvascular anatomy of mammalian ciliary processes. ARVO Abstracts.
Invest Ophthalmol Vis Sci 27(Suppl):163, 1986.
34. Shiose Y and Sears M: Fine structural localization of nucleoside phosphatase activity in the ciliary epithelium of albino
rabbits. Invest Ophthalmol 5:152, 1966.
35. Uusitalo R and Palkama A: Localization of sodium-potassium
stimulated adenosine triphosphatase activity in the rabbit ciliary body using light and electron microscopy. Ann Med Exp
Fenn 48:84, 1970.
36. Fain GL, Usukura J, and Bok D: Localization of Na-K ATPase
in rabbit ciliary body using 3H-ouabain autoradiography.
ARVO Abstracts. Invest Ophthalmol Vis Sci 28(Suppl):67,
1987.
37. Reuss L: Changes in cell volume measured with an electrophysiologic technique. Proc Natl Acad Sci USA 82:6014, 1985.
38. Diamond JM: Osmotic waterflowin leaky epithelia. J Membr
Biol 51:195, 1979.
39. Paillard M, Leviel F, and Gardin J-P: Regulation of cell volume on separated renal tubules incubated with hypotonic medium. Am J Physiol 236:F226, 1979.
40. Linshaw MA and Grantham JJ: Effect of collagenase and ouabain on renal cell volume in hypotonic media. Am J Physiol
238:F491, 1980.
41. van Os CH, Wiedner G, and Wright EM: Volumeflowsacross
gallbladder epithelium induced by small hydrostatic and osmotic gradients. J Membr Biol 49:1, 1979.
42. Kozart DM: Light and electron microscope study of regional
morphologic differences in the ciliary body in the rabbit. Invest
Ophthalmol 7:15, 1968.
43. Weingeist TA: The structure of the developing and adult ciliary complex of the rabbit eye: A gross, light, and electron
microscopic study. Doc Ophthalmol 28:205, 1970.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933365/ on 06/16/2017