Exp. Eye Res. (1992) 55, 289-296

Rubidium

Transport

in Cultured Epithelium BRIAN

Department

of Physiology

(Received

Monkey

Retinal

G. KENNEDY

and Biophysics, Northwest Center for Medical Education, School of Medicine, Gary, IN 46408, U.S.A.

Houston

2 July

Pigment

1991 and accepted

in revised

form 28 October

Indiana

University

1997)

Rb+ influx was used to assess Na-K-Cl cotransport and Na,K-ATPase activities in cultured monkey retinal pigment epithelium. Bumetanide-sensitive (Na-K-Cl cotransport-mediated) Rb+ influx exceeds ouabainsensitive (Na,K-ATPase-mediated) Rb+ influx, with these two transporters accounting for w 95 y0 of total Rb+ uptake. Half-maximal inhibition of Rb+ influx by bumetanide is attained at 75 nM bumetanide. The bumetanide-sensitive Rb+ influx depends on both extracellular Na+ and Cl-, and is activated by extracellular Rb’ with a relatively high afiinity. Na-K-Cl cotransport activity is stimulated (2.5fold) by increased extracellular osmolarity. Elevated CAMP content and glycolytic inhibition both depress cotransport activity. Cyanide application, however, had very little effect on Na-K-Cl cotransport activity. Monkey retinal pigment epithelial cells, maintained in culture, provide a system in which the activity and regulation of cation transport mechanisms can be examined. Key words : bumetanide; CAMP; Na-K-Cl cotransport : Na,K-ATPase : ouabain ; volume regulation.

1. Introduction The retinal pigment epithelium (RPE), a single layer of epithelial cells juxtaposed between the retina and choroid, serves a variety of metabolic functions. These cells actively phagocytize rod outer segments and store vitamin A. The epithelium also presents a diffusional barrier, which regulates ion, nutrient and water flow between photoreceptor cells and the choroidal blood supply (Steinberg and Miller, 1979). Many reports have examined the cation transport properties of the RPE (see Steinberg and Miller, 19 79, for a review). The RPE possesses both ouabain-sensitive Na,K-ATPase activity (Miller and Steinberg, 19 77 ; DiMattio, Degnan and Zadunaisky, 1983 ; Tsuboi, Manabe and Iizuka, 1986 ; Jaffe, Burke and Geroski, 1989) and Na-K-Cl cotransport activity (Frambach and Misfeldt, 198 3 ; Wiederholt and Zadunaisky, 1984; Tsuboi, Manabe and Iizuka, 1986; Miller and Edelman, 1990; Adorante and Miller, 1990; Kennedy, 1990; Joseph and Miller, 1991). These transporters perform crucial functions in maintenance of retinal homeostasis. The Na,K-ATPase of the RPE has been shown to regulate [K+] in the subretinal space (Linsenmeier and Steinberg, 1984 : Griff, Shirao and Steinberg, 1985 ; Miller and Edelman, 1990; Joseph and Miller, 1991), to sustain transepithelial fluid transport (Hughes, Miller and Machen, 1984), to catalyse transepithelial salt flux (Miller and Steinberg, 19 7 7 ; DiMattio, Degnan and Zadunaisky, 1983 ; Tsuboi, Manabe and Iizuka, 1986), and to maintain the Na+ electrochemical gradient which drives cotransport activities (Steinberg and Miller, 1979; Edwards, 1977; Ostwald and Steinberg, 1981). Though less well-characterized than the Na,K-ATPase, 0014-4835/92/080289+08 19

$08.00/O

Na-K-Cl cotransport has been implicated in fluid transport (Tsuboi and Pederson. 19 86), transepithelial Cl- flux (Frambach and Misfeldt, 1983: Tsuboi, Manabe and Iizuka, 198 6 ; Miller and Edelman, 1990), control of internal [Cl-] (Wiederholt and Zadunaisky, 1984; Joseph and Miller, 1991), and volume regulation (Adorante and Miller, 1990) in RPE. The present study examines Rb’ influx in cultured RPE isolated from monkey (MRPE). Rb’ is transported, with an affinity equivalent to that for K+, by both the Na,K-ATPase (Robinson and Flashner, 19 79) and the Na-K-Cl cotransport (O’Grady, Palfrey and Field, 198 7). Ouabain-sensitive Rb’ influx is routinely employed as an index of Na,K-ATPase activity (Robinson and Flashner, 19 79) while bumetanidesensitive Rb+ uptake serves as the index for Na-K-Cl cotransport activity (O’Grady, Palfrey and Field, 198 7 ; Haas, 1989). Rb’ influx has therefore ben used to assay Na,K-ATPase and Na-K-Cl cotransport activities in the MRPE cell culture system in order to characterize further the plasma membrane cation transport properties of the RPE.

2. Materials

and Methods

Cell Culture Primary cultures were established from monkey (A4ucaca fuscicularis) eyes obtained from animals employed in a non-related experimental investigation. Eyes were enucleated immediately after death, and then stored in saline at 4°C for up to 12 hr prior to dissection. Cultures were established as described by Yue and Fishman (1985). Briefly, the RPE-Bruch’s membrane-choroid complex was isolated and dis0 1992 Academic Press Limited EER55

290

sected into the - 0.2-cm” patches that were used to initiate cultures. RPE cells migrate out from these explants to form pure cultures (Yue and Fishman, 198 S), which are propagated as attachment-dependent monolayers on tissue culture plastic (25-cm” flasks, Falcon Primaria) at 3 7°C in a 9 5 % air/ 5 % CO, atmosphere. The culture medium was Eagle’s minimum essential medium supplemented with 10% fetal calf serum, 5% donor calf serum, essential and nonessential amino acids, 4 mM glutamine, amphotericin B and gentamycin sulfate. Upon attaining confluence (assessedby visual observation), cells were subcultured by trypsinization. All influx experiments employed confluent monolayers on 12-well culture plates (Corning). The work reported here was carried out with four separate cultures, each establishedfrom an individual animal. Cultures, unless noted, were used between the second and 13th passage. Solutions The standard solution used for uptake experiments is designated Na(Rb) and contained (in mM) NaCl (140) RbCl (S), CaCI, (2), MgCl, (1) glucose (5) Hepes (15) and Tris (8). A Na+-free solution used choline as an isosmotic replacement for Na’. The lowCl- solution (5 mM total Cl-) substituted sodium gluconate for NaCI. The Mg wash solution contained (mM) MgCI, (105). CaCl, (2), Hepes (15) and Tris (7). Solutions were adjusted to 290 mosmol kg-’ with a Wescor vapor pressure osmometer. The pH of all solutions was set to 7.4 at room temperature. Ouabain (Calbiochem) was diluted from a lo-mM stock solution in water to a final concentration of 100 ,UM. Bumetanide (a gift from Hoffman-LaRoche, Nutley, NJ) was diluted from a 20-mM stock solution in ethanol to give a final concentration of 10 ,uM. Stock solutions of isoproterenol (5 mM in water) and isobutylmethyl xanthine (IBMX, 50 mM in ethanol) were made on the day of the experiment. 86Rb as RbCl in water was purchased from Amersham. Other reagents were purchased from Sigma, Whittaker, Intergen, Hazelton and HyClone. Unidirectional @jRbInflux The protocol used in this investigation follows that of Kennedy and Lever (1984) as modified for use with cultured human RPE (Kennedy, 1990). Monolayers on 12-well plates were washed twice with Na(Rb) saline and then incubated for 10 min at 3 7°C in this solution. Uptake (10 min at 3 7°C) was initiated by addition of Na(Rb) saline containing 0.1-0.3 ,uCi mll’ 86Rb.Uptake was terminated by aspiration, followed by three rapid washes with ice-cold Mg wash buffer. As required, ouabain plus or minus bumetanide was present during both the incubation and uptake periods. Radioactivity was quantitated in a Packard Tri-carb liquid scintillation spectrometer, either by

6. G. KENNEDY 25 0 20 = zi $ Q z

15-

z

IO-

-

0/

:

0

/

5 ;/. 0

5

IO

15

20

Trne (mm)

FIG. 1. Time course for sGRbinflux into MRPE. Monolayers were preincubated at 37°C for 10 min in Na(Rb). “6Rb uptake was then determined, as described in Materials and Methods, for the time periods as noted. (Cells were from passage 16.)

Cherenkov radiation or with liquid scintillant (National Diagnostics, Ecoscint). ‘Zero time’ uptake, determined. by exposure of monolayers to 86Rb-labeledsaline followed immediately by the washing procedure described above, was subtracted from total uptake in each experiment. (‘Zero time’ uptake was < 5% of total uptake.) As shown in Fig. 1, Rb+ uptake was linear for 20 min at 37°C. Protein Determination Monolayers used for protein determination were treated in parallel with those used for uptake measurements. After the final Mg wash, monolayers were extracted for > 1 hr in a 2 % sodium carbonate/O.1 N NaOH solution. Protein was assayedin this extract by the method of Lowry et al. (19 5 l), with bovine serum albumin (Sigma, fraction V) as standard. Rb+ uptake was normalized to protein content. 3. Results Rb+ Uptake: Inhibition by Ouabainand Bumetanide Monolayer RPE cultures were establishedfrom four monkeys, and the results characterizing Rb’ influx into these cultures are summarized in Fig. 2. Ouabainsensitive (Na,K-ATPase-mediated) Rb’ influx, 0.469 + 0.033 (mean 4 s.E.)/Am01 Rb+ mg-’ protein hr-‘, comprises - 35 % of total Rb+ influx into these cultures. In comparison, bumetanide-inhibitable Rb+ influx is 0.730 f 0.030 (mean f S.E.) ,umol Rb+ mg-’ protein hr-I. Na-K-Cl cotransport activity thus comprises - 60% of Rb+ influx, a greater fraction than that mediated by Na,K-ATPase activity. Furthermore, the two transport mechanisms together account for approximately 9 5 % of Rb’ uptake in cultured MRPE. As noted in Materials and Methods, all experiments employed a HCO,--free, Hepes-buffered medium. In

RUBIDIUM

TRANSPORT

IN CULTURED

291

MRPE

I.50

T 2 T a

I 00

F B d 5

05c I o-00

I

o-10

,

I

I

I,

0 20 0.30

I

0.40

I

I

050

[Bumetonidel ooc Cotronsport

No, K-ATPose

FIG. 2. Ouabain-sensitive and bumetanide-sensitive Rb+ influxes. “6Rb uptake was determined in the Na(Rb) solution, as described in Materials and Methods. Where appropriate, transport inhibitors (ouabain 100 ym and bumetanide 10 ,uM) were present during both the lo-min incubation and the subsequent IO-min uptake periods. Each determination employed one 12-well plate, which was used to assay either ouabain-sensitive (total minus ouabain) or bumetanidesensitive (ouabain minus ouabain with bumetanide) activity. Values are meansfS.E. from 44 determinations for bumetanide-sensitive uptake, and from seven determinations for ouabain-sensitive uptake. Rb’ influx is almost completely eliminated in the presence of ouabain plus bumetanide. 0, Ouabain : q ( ouabain/bumetanide : q, bumetanide-sensitive : n , total ; q , ouabain-sensitive.

1.50( IF

-i 2 T a

r.ooc 3.

0.50( 3.

0

,

060

(I

,

0.70

,

/

0.80

IO

nw

FIG. 4 Concentration dependence for bumetanide inhibition of *‘jRb influx into MRPE. Uptake was determined as described in Fig. 2. Ouabain (100 pM) and bumetanide, at the concentration indicated, were present in both the incubation and uptake solutions. The curve is a non-linear least-squares fit (Entitter, Elsevier Biosoft) of the data to an equation of the form : uptake = ~V,W,,,)/(K,,, + bumetanide)) + R. K,,, is [bumetanide] for half-maximal inhibition. V, is total bumetanide-sensitive uptake and R is residual uptake. K,,, = 73 f 10 nM, V, = 0.76 f0.03, R =

0.05 + 0.02 pmol Rb+mg-’ protein hr-‘. but statistically significant, decrease in activity was noted as a function of time in culture. Though the effect is not large, it seemsthat Na-K-Cl cotransport activity may decrease as MRPE are maintained and passed in monolayer culture. (Further data, not included in this work, indicate that this downward trend continues in cells maintained up to passage20.) BumetanideSensitivity

.

E aD 2 zi

I

5

IO

I5

Passage

Fro. 3. Bumetanide-sensitive Rb’ influx as a function of time in culture. “Rb influx was determined as described in Fig. 2. Cultures were passed by trypsinization at 1-2-week intervals. All data in this Figure are from a single culture, which was passed 13 times. Values are means If: S.E. The line is the best least squares fit to the data. The negative slope ( - 2 % per passage) is significantly different from 0, P < 0.05 (Woolson, 1987).

cultured human RPE (Kennedy, 1990) 25 mM HCO,did not appreciably alter bumetanide-sensitive Rb’ influx. The data on Na-K-Cl cotransport activity was further analysed. as a function of tie in culture (number of cell passages).During the course of this study, bumetanide-sensitive Rb’ influx was measured in one culture line over 13 passages(Fig. 3). A slight,

To characterize the properties of Na-K-Cl cotransport activity in MRPE, Rb’ influx was assayed as a function of [bumetanide]. The data (Fig. 4) were fitted to a curve based on a single inhibitory site for bumetanide. Bumetanide acts with relatively high affinity: half-maximal inhibition is seenat 73 f 10 nM (mean f s.D.). In this work, 10 ,UM bumetanide is used to define Na-K-Cl cotransport activity. This concentration of bumetanide, while inhibiting > 99% of bumetanide-sensitive Rb+ uptake, is low enough only to minimally perturb other transporters (Haas, 1989). ionic Dependence of Bumetanide-sensitiveRb+ Influx Bumetanide was used, in the experiments summarized in Figs 24, to define the fraction of Rb+ influx that is mediated by Na-K-Cl cotransport activity. If this influx is, in fact, catalysed by the Na-K-Cl cotransporter, then it should require both extracellular Na+ and Cl-. As shown in Fig. 5, bumetanide-sensitive Rb’ influx clearly dependson both Na+ and Cl-. Na+ removal causesbumetanide-sensitive Rb’ influx to fall by + 90%. Moreover, bumetanide-sensitive Rb+ influx is virtually undetectable in a low-cl- solution (5 mM

292

B. G. KENNEDY 0.75

T

0.50 2

L F ir” 2

3.

0.25

0.00

0.10

0.20

0.30

0.40 External

0.50 [Rbl

OKI

0.70

080

J

IO’

m

6. Bumetanide-sensitive Rb’ influx as a function of extracelluar [Rb+]. 86Rb influx was determined as described in Fig. 2, except as noted. The standard Na(Rb) solution (see Materials and Methods) was modified so that [NaCl] was 130 mM. RbCl was varied from 1 to 10 mM, with choline chloride used as the isosmotic replacement (in all solutions pH was 7.4 and osmolarity was 290 mosmol kg-‘). These solutions are designated Na(x Rb) with x = 1,2, 5 or 10 mu. Monolayers were first incubated (37’C, 10 min) in the Na(5 Rb) solution and then washed twice with the desired Na(x Rb) solution. Uptake was then carried out (3 7°C. 5 min) in the same solution, supplemented with - 0.1 &i ml--‘, 86Rb. Ouabain was present in all incubation, wash and uptake solutions. Bumetanide, 1OpM when applied, was also present in the appropriate incubation, wash and uptake solutions. The line is a non-linear least-squares fit (Enzfitter, Elsevier Biosoft) of the data to the Michaelis-Menten equation, with Z$, = 1.3 mM and V, = 0.81 pmol Rb mg-l protein hr-‘. Cells were from passage 17. FIG.

Chol (Rb) Cl

No (Rb)

Cl

5. Ionic dependence

FIG.

No(Rb) Gluconatr

of Rb+ influx

in MRPE. 86Rb

influx was determinedas describedin Fig. 2. Monolayers were exposed to Na(Rb), chol(Rb) or Na(Rb, low Cl-) solution

at 37°C for 20 min (a IO-min

incubation

period

followed by a lo-min uptake period). All incubation and uptake solutions

contained

100

ouabain.

PM

Bumetanide

(10 pM when applied)was alsopresentin both incubation and

uptake

solutions.

All

solutions

were

adjusted

to

290 mosmolkg-‘, pH = 7.4. In the low-cl- solution,sodium gluconatewassubstitutedfor NaCl (final [Cl] was 5 mM). In the Na-free

solution

[chol(Na)],

choline

chloride

was

substitutedfor NaCl. Values are means+ S.E. q . ouabain; n , ouabain/bumetanide; q , bumetanide-sensitive. Cl-). This provides additional evidence that cultured MRPE possessNa-K-Cl cotransport activity. The Na-K-Cl cotransporter in a range of tissuesis stimulated by extracellular K+ or Rb’ (O’Grady, Palfrey and Field, 198 7). The effect of extracellular Rb+ (varied from 1 to 10 mM) on Na-K-Cl cotransport activity in MRPE is illustrated in Fig. 6. Extracellular Rb+ activates the transporter with relatively high affinity (K,,, = 1.3 mM). Uptake is saturated beyond 10 mM Rb’. Physiologically, extracellular [K+] can vary, in the range of 2-5 mM, during light/dark transitions in the intact mammalian retina (Steinberg, Oakley and Niemeyer, 1980; Linsenmeier and Steinberg, 1984). Osmolarity Na-K