Sodium, Potassium, Two Chloride Cotransport in Corneal ... - IOVS

1 downloads 0 Views 1MB Size Report
Sodium, Potassium, Two Chloride Cotransport in. Corneal Endothelium: Characterization and Possible. Role in Volume Regulation and Fluid Transport. 1'5. 2'4.
Sodium, Potassium, Two Chloride Cotransport in Corneal Endothelium: Characterization and Possible Role in Volume Regulation and Fluid Transport F. P.J. Diecke,11'5'5 Z Zhu,22'4'4 F. Kang,2 K. Kuang,2 andj. Fischbarg1'2 search for membrane transporter proteins that could contribute to volume regulation and fluid transport by corneal endothelium. As an initial step, the authors have focused on Na+-K+-2C1~ cotransporters. PURPOSE. TO

Bovine corneal endothelial cells were cultured to confluence. 86Rubidium was used as a tracer for K+ uptake determinations; uptake values were normalized per milligram of cell protein.

METHODS.

Three components of K+ uptake were characterized: ouabain (1 mM) sensitive, bumetanide (0.1 mM) sensitive, and ouabain-bumetanide insensitive. Both the ouabain-sensitive and bumetanide-sensitive components increased in the presence of 26.2 mM HCO^"; 0.5 mM 4,4'diisothiocyanato-stilbene-2,2'-disulfonic acid abolished this increase. The bumetanide-sensitive component was completely inhibited in the absence of Na+ or Cl~. This component was increased 33% by a 33% hypertonic solution and was decreased 38% by a 33% hypotonic solution. The protein kinase C activator phorbol 12-myristate 13-acetate decreased the activity of the cotransporter, whereas forskolin, in the presence of isobutylmethylxanthine, decreased it. Calyculin A (100 nM), an inhibitor of phosphatases 1 and 2a, produced a large (97%) activation of this component. RESULTS.

These results provided for the first time conclusive evidence for the presence of a Na+-K+-2C1~ cotransporter in corneal endothelium and of its possible involvement in volumeregulatory processes in these cells. Given the uptake values reported here, such cotransporter could contribute significantly to electrolyte transport and hence to fluid transport across this preparation. (Invest Ophthalmol Vis Sci. 1998;39:104-110) CONCLUSIONS.

T

he major function of the corneal endothelium is the pumping offluidfrom stroma to aqueous humor at rates approximating 4 to 5 ju-l/cm2 per hour to maintain constant hydration of the cornea. In addition, corneal endothelial cells are capable of rapid regulation of their volume in response to anisotonic challenges.1 The electrolyte transport mechanisms that drive the fluid transport are not well understood. Similarly, the mechanisms for regulatory volume decrease and increase have not been defined. Current hypotheses are based on bicarbonate cotransport mechanisms of varied stoichiometries located in the apical membrane2'3 or basolateral membrane.4 However, it has also been reported that chloride is required forfluidtransport by the corneal endothelium,5 suggesting that chloride transport mechanisms may be involved.

From the Departments of 'Physiology and Cellular Biophysics, and Ophthalmology, College of Physicians and Surgeons, Columbia University, New York, New York. 'Present address: Department of Pharmacology and Physiology, UMDNJ-New Jersey Medical School, Newark, New Jersey. ^Present address: Department of Ophthalmology, Shanghai Second Medical University, Shanghai, Peoples Republic of China. Supported by National Institutes of Health grant EY06178, in part by Research to Prevent Blindness, and by the Shanghai Municipal Government. Submitted for publication May 23, 1997; revised September 3, 1997; accepted September 25, 1997. Proprietary interest category: N. Reprint requests: F. P. J. Diecke, Department of Pharmacology and Physiology, UMDNJ-New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103-2714. 2

104

In the present study we have investigated whether cultured bovine corneal endothelial cells (CBCECs) possess a Na+-K+-2C1~ cotransporter and have attempted to assess its contribution to fluid transport and to volume regulation. Na+K+-2C1~ cotransporters have now been identified in many mammalian and nonmammalian tissues. They represent electroneutral transport systems with a presumed stoichiometry of lNa:lK:2Cl driven by the sum of the gradients of sodium, potassium, and chlorine. Two isoforms of the transport protein have been identified so far. One of these has been demonstrated to operate in chloride-secreting epithelia such as the shark rectal gland6 and human colon,7 in which it is confined to the basolateral membrane, and is therefore considered the secretory isoform of the transporter. The second isoform appears to be expressed only in the thick ascending limb of Henle's loop in the kidney,8 is located in the apical membrane, and represents the absorptive isoform. Using rubidium 86 to trace potassium uptake, we have established that a significant component of the potassium uptake by CBCEC has the characteristics of a Na+-K+-2CF cotransporter. In addition, we have investigated factors that can modulate the activity of the cotransporter.

MATERIALS AND METHODS

Cell Culture Bovine eyes were obtained from a local abattoir. On arrival, each eye was briefly wiped with alcohol for sterilization. The cornea was removed, placed in a shallow hemispherical holder Investigative Ophthalmology & Visual Science, January 1998, Vol. 39, No. 1 Copyright © Association for Research in Vision and Ophthalmology

Na + -K + -2C1

IOVS, January 1998, Vol. 39, No. 1

Cotransport in Corneal EndotheUal Cells

105

TABLE 1. Composition of Ringer's Solutions (in millimolar)

Buffer NaCl KC1 NaHCO, MgSO4 * CaCl2 KH2PO4 HEPES Glucose Sucrose

CO2-HCO3

HEPES

HEPES-HCO,

129.5 3.8 0.0 0.78 1.7 1.0 20.0

104.4 3.8 26.2 0.78 1.7 1.0 20.0

119-4 3.8 26.2 0.78 1.7 1.0 0.0

6.9

6.9

0.0

0.0

Hyperosmotic

Hyposmotic

54.4 3.8 26.2 0.78 1.7 1.0 20.0

54.4 3.8 26.2 0.78 1.7 1.0 20.0

26.2 0.78 1.7 1.0 20.0

6.9

6.9

6.9

6.9

0.0

200.0

0.0

100.0

(endothelium up), and washed with a calcium-free physiological buffer solution. The endothelium was then covered with 1 ml of calcium- and magnesium-free saline solution containing 0.25% trypsin and 0.02% ethylenediaminetetraacetic acid, and the holder was placed in an incubator for 5 to 10 minutes. The solution was then changed to Dulbecco's modified Eagle's medium (Life Technologies, Gaithersburg, MD), and the endothelium was gently rubbed with a plastic rod to loosen the cells. The cells were aspirated and plated in 25-cm2 culture flasks (Falcon; Becton Dickinson, Franklin Lakes, NJ) with 3 ml of Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum, 2 ng/ml basic fibroblast growth factor, and an antibiotic mixture containing penicillin and streptomycin. Cells were maintained in a CO 2 incubator and provided with new culture medium every 3 days. Cells reached confluence in 5 to 7 days, after which they were subcultured. Culture medium was aspirated, and 1.5 ml of trypsin-EDTA solution was added. The cells were then observed under a microscope; when they began to round, the trypsin medium was removed and fresh Dulbecco's modified Eagle's medium containing 10% v/v fetal bovine serum was added. Cells were loosened from the flask by gentle trituration by cyclic aspiration into and emptying of a pipette. The cell suspension was divided into three fresh flasks (25 cm 2 , as above) and grown to confluence. For the experiments, one flask of cells was subcultured into up to 16 four-well culture dishes (Nunc; Naperville, IL).

Solutions and Chemicals Three types of Ringer's solution with different buffers were used, as given in Table 1. One of these solutions was buffered with Hepes and a second solution, also vised as control solution, was buffered with Hepes and HCO^ at ambient CO 2 partial pressure. The third solution contained NaHCO 3 equilibrated with an air-5% CO 2 gas mixture. In addition, we used sodium- and chloride-free modifications of the control solution and solutions in which the osmolarity was increased with sucrose or decreased by removal of sodium chloride. Sodium-free solutions were prepared using yV-methyl-i>glucamine as a substitute for NaCl, and choline bicarbonate was used as a NaHCO 3 replacement. Chloridefree solutions were prepared with methanesulfonate as replacement. All solutions had a pH between 7.35 and 7.43 and an osmolarity of 295 — 5 mOsm. All salts for the Ringer's solutions were obtained from Sigma (St. Louis, MO). In addition, ouabain, bumetanide, phorbol 12-myristate 13-acetate (PMA), and 4,4'-diisothiocyanato-stilbene-2,2'-disulfonic acid (DIDS) were obtained from Sigma. Calyculin A,

54.4

Isosmotic

3.8

forskolin, isobutylmethylxanthine, and 8-bromoguanosine3',5'-cyclophosphate were obtained from Research Biochemicals International (Natick, MA).

Uptake Measurements Potassium uptake in control and test solutions was determined by measuring the uptake of 86 Rb as a tracer for potassium. Cells were grown in 16-mm four-well culture dishes and reached confluence in 3 days. For the experiments, cells were preincubated for 1 hour in control solution (HEPES-HCO^, Table 1) and then incubated for 10-minute periods in control or test solutions containing 2 jtxCi/ml of 86 Rb. For the determinations in the presence of DIDS, forskolin, calyculin A, and 8-bromoguanosine-3',5'-cyclophosphate, cells were preincubated for an additional period of 15 minutes in control solution containing the agent. The volume of medium per well was 0.5 ml in all procedures. The incubation was terminated by aspirating the incubation solution and washing the cells rapidly with an ice-cold 0.1 M MgCl2 stop solution. The 86 Rb taken up by the cells was leached by exposing the cells for two 30-minute periods to 500 jul of 5% trichloroacetic acid per well. The 86 Rb was then counted in a liquid scintillation counter (model Rack Beta 1219; LKB). Aliquots of 10 /xl of incubation solution diluted with 1 ml of 5% trichloroacetic acid solution were counted to obtain the specific activity for 86 Rb. Protein was extracted from the cell residues remaining in the wells in 500 /xl of a solution containing 0.5 NaOH and 5% sodium dodecyl sulfate. Protein content per well was then assayed with the Lowry method 9 with bovine serum albumin (Sigma) as standard. The potassium uptake per 10-minute period and millgram of protein (/ kin ) was then computed from the 86 Rb uptake with an equation adapted from Gelehrter et al. l() : /kin = (c.p.m./well)[K+]/{(c.p.m./jal)(mg protein/well)} where cpm indicates counts per minute, [K + ] is the potassium concentration in the incubation medium in nanomoles per microliter, and cpm//xl represents the activity of the incubation medium.

Data Analysis Data from an individual well constituted one experiment. The results are expressed as means ± SEM. For two sets of data points we used an unpaired, two-tailed /-test, whereas Dunnett's post-test was used to compare three ore more data sets.

106

Diecke et al.

IOVS, January 1998, Vol. 39, No. 1 200

600 150u 100-

[K] i n = 598.6 ± 12.2 nmol/mg protein k = 0.031 ± 0.0009 min"1

0

50 100 INCUBATION TIME (minutes)

FIGURE 1.

Potassium uptake of cultured bovine corneal endothelial cells (CBCECs) as a function of time (n = 8) in HepesHCC»3~ medium. The data were fit to an exponential association equation to extrapolate the saturating intracellular potassium concentration [K]in (dashed line) and the initial rate of uptake k (dotted line).

RESULTS

Components of Potassium Uptake: BumetanideSensitive Component Figure 1 shows the potassium uptake of CBCECs in HepesHCO^~ solution as a function of time. The data points were fit to a single exponential buildup equation. Uptake was quite rapid and proceeded with a rate coefficient of 0.031 minute""1 or 3-1% per minute. The potassium uptake became saturated at 598 nmol/mg protein. At a cell water space of 4.1 /Ltl/mg protein (unpublished data), this would correspond to an intracellular potassium concentration of 148 mmol/kg H2O or 118 mmol/kg cell mass. As shown in Figure 2, potassium uptake consisted of three components. The largest component, representing approximately 50% of the total uptake, was inhibited by ouabain. A second component of approximately 45% was inhibited by bumetanide at a half-maximally effective concentration (^ O5 ) of 51 nM (Fig. 3). A small third component of approximately 5% of total uptake was insensitive to ouabain and bumetanide. Potassium uptake was significantly affected by the presence of HCO^". Figure 2 shows that total potassium uptake in Hepes-HCO^" Ringer's solution was increased 50% over that observed in Hepes-buffered Ringer's solution. Both of the major components of potassium uptake were increased in media containing HCO^"—the ouabain-sensitive component increased by 44% and the bumetanide-sensitive component by 71%. Neither the total potassium uptake nor the size of each of the K+ uptake components differed significantly in CO2-HCO^-buffered solutions from the corresponding uptake values in Hepes-HCO^ Ringer's solution. The HCO^-dependent potassium uptake was completely blocked with 0.5 mM DIDS. Figure 4 compares the effect of DIDS on the bumetanide-sensitive potassium uptake in HepesHCOj"-buffered Ringer's solution with that observed in a solu-

HEPES

HEPES-HCO3

CO 2 -HCO 3

FIGURE 2. Potassium uptake and the effects of bumetanide

(0.1 mM) and ouabain (1 mM) in bicarbonate-free (Hepes) and bicarbonate-containing Ringer's solutions (Hepes-HCO3 and CO2-HCO3); N=4for all groups. The pH of all solutions was between 7.35 and 7.45. tion buffered with Hepes only. Figure 5 shows the effect of DIDS on the ouabain-sensitive K+ uptake. DIDS produced a significant inhibition of both the bumetanide and ouabainsensitive potassium uptakes in Hepes-HCOJ buffer but was without effect on the uptake in Hepes buffer. Moreover, the uptake in Hepes-HCO^ buffer plus DIDs was virtually identical with that in Hepes buffer without DIDS, suggesting that DIDS blocked the HCO^-sensitive fraction of the bumetanide-sensitive component of the potassium uptake.

Effect of Sodium-Free and Chloride-Free Solutions The observation of a potassium uptake component with a high sensitivity to bumetanide suggests the presence of a Na+-K+-

J=J bs *(l-[B]/([B]+K 1/2 ))+J is K1/2=0.051 uM

0.0 0.5 1.0 BUMETANIDE CONCENTRATION (uM) FIGURE 3. Dose-response curve for bumetanide in Hepes-

HCO3 Ringer's solution containing 1 mM ouabain. The data were fit to a standard hyperbolic expression describing binding of the inhibitor as indicated at the top; the dotted line indicates the ouabain and bumetanide-insensitive component of potassium uptake./ bs and/ is indicate the bumetanide-sensitive, and ouabain and bumetanide-insensitive uptakes, respectively. TV = 4 for all groups.

Na+-K+-2C1~ Cotransport in Cornea! Endothelial Cells

rOVS, January 1998, Vol. 39, No. 1

200

100

LJ

107

| Z>

^CONTROL

^ 0 . 5 mM DIDS

I I CONTROL 0.1 mM BUMETANIDE 1.0 mM OUABAIN BUMETANIDE + OUABAIN

O L_

Q.

LJJ

>

50

co

il

CONTROL

LLJ

21 Z) CD

HEPES

HEPES-HCO3

FIGURE 4. The effect of 0.5 mM 4,4'-diisothiocyanato-stilbene-

2,2'-disulfonic acid (DIDS) on the bumetanide-sensitive potassium uptake in bicarbonate-free and bicarbonate-containing solutions. N = 4 for all groups.

2C1 cotransporter in CBCECs. Such cotransporter is driven by the sum of the chemical gradients of the three ions and has been reported to require an ordered binding sequence of the ions involved.1 '"' 3 We have examined the effects of sodiumand chloride-free solutions on the bumetanide-sensitive potassium uptake to obtain further support for the presence of the cotransporter in CBCECs. NaCl in Ringer's solution was replaced with ./V-methyl-D-glucamine chloride, whereas NaHCO3 was replaced with choline-HCO3 to obtain sodium-free solutions. To obtain chloride-free solution we replaced NaCl and KC1 with their respective methanesulfonate salts and replaced CaCl2 with calcium gluconate. As shown in Figure 6, the

Ld

I

I CONTROL

(^8 0.5 mM DIDS

, +

^

"o l_

^

QL

LJJ

CH E

>

50 Ld O CO r"1

z: o

2