Pflfigers Arch (1990) 416: 554-- 560

E PfliJgers Archiv

uropeanJournal of Physiology 9 Springer-Verlag 1990

Functional asymmetry of phosphate transport and its regulation in opossum kidney cells: Phosphate transport Stephan J. Reshkin, Judith Forgo and Heini Murer Department of Physiology,University of Ztirich, Winterthurerstrage 190, CH-8057 Ziirich, Switzerland Received December 19, 1989 / Received after revision March 8 / Accepted March 21, i990

Abstract. The polarized distribution of phosphate (Pi) transport systems in a continuous renal cell line derived from opossum kidney (OK) was measured in monolayers grown on permeant filter support. When cultured on collagencoated nitrocellulose filters, OK cells formed tight, functionally polarized monolayers. Three Pi transport systems were identified in these monolayers: one apical sodium (Na)-dependent system and two systems on the basolateral surface, one Na-dependent and one Na-independent. The apical system was high-affinity (Kin = 0.4 mM P0, lowcapacity (Jmax = 1100 pmol P~/mg protein per minute) with a Na:P~ stoichiometry greater than 1 (n--3) and a high interaction coefficient (K Na = 105 mM Na). On the basolateral surface the Na-independent system comprised about 30 % of the total Pi transport at this surface. Both basolateral systems were of low affinity (Km: Na-independent, 2.6 raM; Na-dependent, 5.2 mM) and high capacity (Jm,x: Na-independent, 2100; Na-dependent, 2400 pmol/mg protein per minute). The basolateral Na-dependent system had a Na~ stoichiometry of i and a relatively low interaction coefficient (K Na = 25 mM Na). Only the basolateral Na-independent system was inhibitable by 4,4'-diisothiocyanostilbene-2,2'disulphonic acid (DIDS). These results are compatible with a net vectorial transcellular transport of P~ from the apical through the basolateral cell surfaces. The presence of a basolateral Na-dependent system may reflect additional metabolic requirements that cannot be met only by apical influx. Taken together, these results demonstrate the ability to grow cell monolayers successfully, displaying polarized transport activities similar to in situ. Key words: OK cells - Monolayer - Permeant support PO4 transport - Polarity - Apical - Basolateral

Offprint requests to: H. Murer

Introduction Kidney proximal tubule is lined by a layer of polarized epithelial cells, the polarized organization reflecting their vectorial function. This intrinsic polarity is particularly expressed in the plasma membrane, which is divided into specialized regions defined by anatomical differences, differences in composition and by functional polarity of enzyme and transport activity (e.g. Simons and Fuller 1985; Murer and Gmaj 1986). Transport polarity is generally organized such that a vectorial transcellular flux (reabsorption) of the solute is realized. The most common type of this polarity is expressed by having a first transport (influx) step at the luminal (apical) membrane that is active and concentrative, and a second diffusive transport (efflux) step at the basolateral membrane (e.g. Murer et al. 1984; Kinne 1986). However, it is possible to have vectorial transepithelial flux with active systems on both membranes when the active systems have very different characteristics and/or transport capacities. Furthermore, special metabolic requirements of the epithelial cells may require additional influx capability on the basolateral membrane that can be used to fulfil these requirements (Schwab et al. 1984a; Reshkin et al. i988; for review see: Murer et al. 1984; Kinne 1986; Murer and Gmaj 1986). Pi homeostasis is maintained to a great degree by vectorial reabsorption in the kidney and some 7 0 - 8 0 % of filtered Pi is reabsorbed in the proximal convoluted tubule. The sodium/phosphate (Na/Pi) co-transporter located in the apical membrane of renal proximal tubular epithelia has been well described in studies with whole tissue, isolated tubules, brush border membrane vesicles (BBMV) from a number of species (for review see Gmaj and Murer 1986) and in cell monolayers grown on either impermeant (Malmstr6m and Murer 1986) or permeant (Rabito 1983) support (for review see Biber 1989). The transport mechanism(s) present on the basolateral side have been studied much less and are much less clearly understood. Available data from basolateral membrane vesicles (BLMV) suggest either a Naindependent pathway sensitive to 4,4'-diisothiocyanostilbene-2,2'-disulphonic acid (DIDS)(Grinstein et al. 1980;

555 Murer 1988) or a N a - d e p e n d e n t pathway ( H a m m e r m a n and Schwab 1984; Schwab et al. 1984a, b). This lack of consensus could be due to the difficulty in obtaining pure preparations (Hagenbuch and M u t e r 1986) or to a natural plasticity of this m e m b r a n e due to function, location in kidney or in the species studied. The objective of this study was to characterize the mechanisms of apical and basolateral Pj transport in monolayer cultures grown on p e r m e a n t support of a renal continuous cell line k n o w n to have apical P~ transport similar to that in vivo when grown on i m p e r m e a n t support. This study is also a necessary first step for further studies on regulatory phen o m e n a of Pi transport in this cell line. I n the following companion paper, the regulation of these transport systems by parathyroid h o r m o n e is examined.

Materials and methods All cell culture supplies were obtained from Amimed (Basel, Switzerland) or Gibco (Basel, Switzerland). Radiolabelled materials were purchased from New England Nuclear (Ztirich, Switzerland) and DIDS from Pierce (Rockford, Ill.) All salts were of analytical grade and purchased from commercial sources.

Cell culture. OK, an established continuous cell culture from opossum kidney (Koyama et al. 1978; Malmstr6m and Murer 1986) which express, in monolayers grown on impermeant support, apical Na/P~ cotransport similar to that measured in (BBMV) isolated from different mammalian kidneys (Gmaj and Murer 1986; Muter and Gmaj 1986; Malmstr6m et al. 1988)were grown in monolayer culture in Dulbecco's modified Eagl's medium (DMEM/I-IAMS F12) (1:1) supplemented with 10% fetal calf serum (FCS), 22 mM NaHCO3, 20 mM 4-(2hydroxyethyl)-l-piperazuneethanesulphonic acid (HEPES), 2 mM Lglutamine, 50 IU/ml penicillin and 50/~g/ml streptomycin in a humidified atmosphere of 5% CO2, 95% air at 37~ Subcultures were prepared by trypsinizing and reseeding at high density (approx. 5 x 10s cells/ml) in either 175 cm2 flasks or on filters (see also below). Monolayers on permeable filter support were grown on MillicetlCM filter inserts (Millipore, 12 mm diameter, 0.45 tzm pore size) coated with a very thin film of rat-tail collagen (R-type, Serva Basel, Switzerland; 0.5 mg/ml in 50% ethanol). Cells from confluent monolayers were trypsinized and seeded at approximately 1-2 • 10s cells/filter in the above growth medium under a 5 % CO2 atmosphere. The medium was changed after 1-2 h of seeding and then cells were refed fresh medium every 12 h. Monolayers just reached confluency in 24 h and transport studies were commenced 36 h after seeding. Transport assays. Transport measurements were performed in uptake media consisting of (in raM): 137 NaC1 [replaced by tetraethylammonium (TMA)-C1for Na-independent transport]; 5.4 KC1; 2.8 CaC12;1.2 MgSO4 and 10 HEPES/Tris pH 7.4. Radioactive substrate was added in concentrations as indicated in the figure legends. For transport assays, growth medium was aspirated and both sides of the monolayer gently rinsed twice in substrate-free TMA uptake solution at 25 ~ Filter insert monolayers were then placed in a 24-well culture plate (Nunclon) for uptake measurements. Substrate-free uptake solution (500 kd) was added to the appropriate filter insert compartment and TMA uptake solution to the opposite compartment. Transport was initiated by mixing 50/A of the same uptake solution containing an 11fold concentration of the desired radioactive substrate to the uptake solution already present in the filter insert compartment. Uptake was stopped and non-specifically bound radioactivity effectively removed (data not shown) by rapid aspiration of the uptake solution and careful rinsing (6x) of the filter and filter insert in an ice-cold isotonic solution containing 100 mM of the cold substrate. Non-specificbinding (blanks) was assessed measuring zero-time uptake by starting uptake, and immediately aspirating the uptake solution and processing the filters as

above. Non-specificbindingwas below 10% of radioactivity associated with any experimental point. Total radioactivity incorporated into the monolayer was measured by liquid scintillation counting of the whole filter insert in I0 ml of Packard 399 scintillation fluid (Dubendorf, Switzerland) in a Kontron counter (ZUrich, Switzerland).

Estimation of influx kinetics and stoichiometry. 32p-orthophosphate influx (3 min) was measured.on each side as a function of either external P~(at 0 or 137 mM Na) or Na [at 0.1 mM (apical) or i mM (basolateral) P~] concentration. Kinetic parameters from the observed data were determined by a curve-fitting procedure in which influx data were computer fitted, using an iterative, non-linear method to either the Michaelis-Menten kinetic equation (P~):J = (Jm~• [S])/(Km + S]) or to a generalized Hill equation (Na): J = (Jmax IS]n)/( g ' q- [S] n) where J is phosphate influx in pmol/mg protein per 3 rain; [S] is external P~or Na concentration in mM; Jmaxis maximal phosphate influx rate; Km is the concentration of P~ that yields one-half Jma• K' is an affinity constant modified to accommodate multisite interactions (interaction coefficient) and n, the Hill coefficient, is an estimate of the number of reactive Na-binding sites. Significanceof differences between estimates of kinetic parameters, calculated by non-linear regression, was determined as described by Motulsky and Ransnas (1987). Protein concentration was measured with the Bio-Rad protein assay (Bio-Rad, Glattbrugg, Switzerland). Statistics. Significance was determined using a two-tailed t-test for paired or unpaired means. A value of P<0.05 was accepted as significant. All values are presented as means +_SE.

Results Impermeability o f confluent cell monolayers It was essential for this study that O K cells growing as monolayers on the Millipore polycarbonate Millicell filters formed a tight monolayer that was impermeable enough to small ions (Pi) that isolated measurements on the two sides of the monolayer could be made. Figure 1 illustrates the diffusive barrier to Pi formed by the monolayer. Monolayers (or blank filters) were incubated in TMA-C1 uptake solution with 0.1raM 32po 4 in either the upper (apical) or lower (basolateral) compartment; a Na-free solution was selected to minimize any transcellular flux c o m p o n e n t (see below). Diffusion was a linear process for at least 10 m i n in all conditions and rates of diffusion in both directions (apical to basal, or basal to apical) were equal within the conditions of bare filters or filters with monolayer. The presence of monolayers resulted in an approximately 20-fold reduction of Pi diffusive flux across the filter which was partially reversed by 30 min incubation of the monolayer in a Ca 2+free buffer [+ ethylenebis (oxorutrilo) tetraacetic and (EGTA)], providing evidence that reduced permeability in the presence of cell monolayers resulted from the formation of tight junctions. W h e n initial Pi concentrations were increased to 0.5 and 2mM the rate of diffusion increased proportionately (data not shown). This experiment demonstrated that the monolayers were tight enough to permit the isolation of transport processes on the two cell surfaces.

556

Time-course and Na-dependence of Pi uptake

25

E

15

t~ e~

E

0 0

10 0 O~ 0

._~

5

I 0 2

4

6

8

10

TIME (min)

Fig. 1. Phosphate permeability in opposum kidney (OK)-cell mono-

layers on Millicelfilter inserts. Monotayerswere incubated in a tetraethylammonium (TMA)-C1 (137 mM) uptake solution containing 0.i mM ~2P-phosphate (P0 in either the apical (diamonds) or basolateral (squares) compartment. The time course of appearance of 3zp~in the opposite compartment was measured for inserts with monolayers (closed symbols), without monolayers (open symbols) and with monolayers that had been preincubated in a Ca-free medium (stars) to reduce monolayer tightness. Each point is the mean of three determinations from separate filters

Localization of the Na/K pump and Na-dependent a-methylglucoside transport Two transport systems known to be localized in intact proximal tubular cells to the separate cell membranes (brush border and basolateral membranes) were used to determine whether the confluent, tight OK-cell monolayers showed the expected polarity; with the Na/K pump facing the lower (basolateral) compartment and the Na-dependent glucose transporter facing the upper (apical) compartment. The Na/ K pump acitivity was obtained by measuring the ouabainsensitive 86Rb (5mM) uptake on the two cell surfaces, and results o f a typical time-course experiment are present in Fig. 2A. The initial rate of ouabain-sensitive S6Rb uptake were 20.5 and 2.6 nmol/mg protein per minute on the basolateral and apical side, respectively. Thus 89% of ouabain-sensitive Rb uptake occurred from the lower compartment of the filter insert, confirming that the basolateral surface this compartment. The Na-dependent glucose transport activity as measured by uptake of [14C]-a-methylglucoside is shown in Fig. 2B. In the present study, as in others (Haggerty et al. 1988), almost no transport activity of this glucose analogue was found on the lower (basolateral) cell surface. In the upper filter-insert compartment (apical surface) the presence of Na (137mM) stimulated the uptake of this compound by 12-fold. The results of these experiments are consistent with the suggestion that OK-cell monolayers displayed "normal" polarity.

The time-course of P~ uptake into OK-cell monolayers from the apical and basolateral cell surfaces is shown in the upper and lower panels of Fig. 3, respectively. Monolayers were incubated in either NaC1 or TMA-C1 uptake solution with 0. lmM 32po 4 for the times indicated. On both sides of the monolayer, transport in the presence or absence of Na was linear for at least 6 rain. Na stimulated the transport rate of Pi on both sides of the monolayer, but to a lower extent on the basolateral side. Na-independent uptake on the apical side amounted to approximately 10% of the rate in the presence of Na, while on the basolateral side the proportion of the Na-independent component was much larger, comprising approximately 30 % of the rate in the presence of Na. Mean initial rates of Na-dependent Pi uptake (four experiments) were 287 pmol/mg protein per minute (apical) and 59 pmol/mg protein per minute (basolateral), thus making, at this Pi concentration, an almost 5-fold greater Nadependent uptake at the apical surface. Due to the linearity of this transport process and to minimize Pi diffusion (Fig. 1), we chose 3 min uptake incubation times for all following experiments.

Saturation kinetics of apical and basolateral Pi transportpathways In view of the lack of knowledge about basolateral transport pathways in the proximal tubule and considering the large differences between the two cell surfaces observed in the time-course experiments, it was now of interest to characterize further the properties of Pi transport in our monolayer system from the apical as well as the basolateral surface. For this, the response of Pi transport at each cell surface to changes in the concentration of Pi and Na was measured. For P~ concentration kinetics, initial influx velocity was determined for 3 min of uptake, with or without Na § (137mM) in the incubation medium. Final Pi concentrations ranged from 0 to 2 mM for the apical compartment and from 0 to 16 mM for the basolateral compartment. In the apical compartment Pi uptake into the monolayer displayed saturation characteristics only in the presence of Na (Fig. 4a) with kinetic constants (Table 1) similar to those reported for in BBMV (Gmaj and Murer, 1986) and for "apical" uptake in cells grown on plastic plates (Malmstr6m et al. 1986; Biber et al. 1988; Quamme et al. 1989). P~ influx into the monolayers from the basolateral side showed saturation kinetics in both the presence (total uptake) and absence (Na-independent uptake) of Na (Fig. 4b), with total uptake in the presence of Na and Na-dependent uptake (total minus Na-independent) having a greater Jmax and Km than uptake in TMA medium (Table 1). The basolateral cell surface thus expressed two saturable Pi transport systems, Na-dependent and Na-independent, each with a higher maximal rate and Km than found on the apical side. Therefore, with respect to P~concentration, OKcell monolayer basolateral P~ transport systems were lowaffinity, high-capacity systems, while the apical transport system had high-affinity, low-capacity characteristics.

557 2000

250

Q A .--.E 6}

200 A

1500

150

r

"6

; 1ooo o.

100 "0

Z CE

50C

50

OC 2

4

TIME

6

1

(rain)

2

3

4

5

TIME (min)

Fig. 2. Transport of either rubidium (Rb) or a-methyl-gtucoside (aMG) in OK-cell monolayers. Apical (diamond) and basolaterat (squares) 86Rb uptake in the absence (open symbols) and presence (closed symbols) of 500 #M ouabain (left). Rb concentration was 5 mM in the NaC1 uptake solution. Apical I4C-aMG uptake in the presence (open symbols) or absence (closed symbols) of Na (right). aMG concentration in the NaC1 and TMA-C1 uptake solutions was 0.1 raM. All points are the mean _+ SE of three determinations

2500 2000 1500

E

1000 500

2500

0

Apical I~

0.5

1

1.5

J 2

2000 "6 1500 8

1000

6

I-4 500

2

0 "~ o

1

2

3

4

5

6

131 0

800

E 2

600

,,~

400

I 4

Basolateral

I 12

I 8 Phosphate

16

(mM)

Fig. 4. Effect of external phosphate (P~) concentration on 32p~uptake into OK-cell monolayers from the apical or basolateral cell surfaces. P~ concentrations ranged from 0 to 2 mM (apical) or from 0 to 16 mM (basolateral) in either isotonic NaC1 (open symbols) or TMA-C1 (closed symbols) uptake solutions. Allpoints are the mean +_of three replicates

200

1

2

3 TIME

4

5

6

(min)

Fig. 3. Apical and basolateral time course of 32p-phosphate (P~) uptake in OK-cell monolayers. Monolayers were incubated in NaC1 (star) or T M A - C t (square) uptake solutions in the presence of 0.1 m M 32p~ for the time periods indicated. Note the different scales on the abscissa. Each point is the m e a n _.+_SE of three determinations

The effect of increasing N a concentration on the initial rate (3 min) of P~ influx was m e a s u r e d over a range of Na concentrations from 0 to 137mM in o r d e r to characterize the Na-P~ interaction properties for the N a - d e p e n d e n t systems present on the two cell surfaces. The results p r e s e n t e d in Fig. 5 and Table 1 indicate that the N a - d e p e n d e n t transp o r t processes occurring on the two cell surfaces have strik-

ingly different Na-Pi interaction characteristics. In the present experimental conditions, the apical m o n o l a y e r surface (Fig. 5A) displayed a sigmoidal function of N a activation of the Pi transporter following the Hill equation (see Materials and methods) that u p o n computer analysis yielded an estimate of three sodium ions participating in Pi entry and an interaction coefficient, K Na, of greater than 100 m M sodium. The basolateral N a - d e p e n d e n t system displayed a hyperbolic relationship of N a concentration and P~ influx consistent with one sodium entering for each p h o s p h a t e and displayed a high affinity for N a ( K Na = 20 mM). The value in each curve of the y-intercept, indicating the rate of u p t a k e in the absence of Na, reflects the relative importance at each cell surface of the N a - i n d e p e n d e n t process; the Na-indep e n d e n t rate was approximately 6 % and 30 % of the maximum N a - d e p e n d e n t rate on the apical and basolateral surfaces, respectively.

558 Table 1. Phosphate kinetic parameters and NzdP~stoichiometryof apicat and basolateral transport

in opossum kidney ceil monolayers Cell Surface

R?'~

JPm.~

n

KN"

jN%.~

Apical Basolateral Total Na-independent Na-dependent

0.37 • 0.05

1100 • 37

3

105 _+ 10

756 • 95

4.8 • 0.4 2.6 • 0.2 5.2 • 0.9

4300 + 200 2100_+ 31 2400 + 200

1

25.4 • 2.2

540 • 13

Values are means • SE. Units of kinetic constants are: K%, mM phosphate; J~.... pmol P~mg protein/rain; n, estimated number of Na entering with each P~; KN", mM sodium; jN%.~, pmol P~mg protein/min

200( Apical 1500

1000

-ff E o3

500

o o.

I ......... 100

t

o

20

40

80

80

I 120

I

140

2500 E o. v "E

Basolateral 2000

only of the basolateral Na-independent Pi transport system (Fig. 6). The degree of inhibition of the basolateral Naindependent system ~ 28 % at 50 #M) was similar to that reported in renal BLMV at the same concentration (Grinstein et al. 1980). At the higher concentration (200/~M) there was additional significant inhibition of both the apical and basolateral Na-dependent systems (Fig. 6), which is in agreement with reports that higher concentrations of DIDS can inhibit co-transport systems as welt as anion-exchange (Garay et al. 1986). The results presented here suggest that the Na-independent Pi carrier in these cells is an anionexchange system and exclusively located on the basolateral cell surface.

o

15o0

Discussion looo

so2 0 0

i

~

l

t

20

40

8o

80

.......

[ ,

too

~

12o

--J

14o

S o d i u m (raM)

Fig. 5. Effect of external sodium (Na) concentration on 32P-phosphate (P0 uptake into OK-cell monolayers from the apical (0.1 mM P0 and the basoIateral (t mM P0 cell surfaces. Uptake media consisted of variable concentrations of NaC1 from 0 to 134 raM, osmoticallybalanced by TMA-CI at: pH 7.4. Each point is the mean -+ SE of three replicates Thus, although Na-dependent transport systems were found on both sides of the polarized OK cell-monolayers, large differences between the kinetic characteristics of the two Na-dependent systems were found as well as differences between the Na-dependent and Na-independent processes.

Localization of

PO 4

transport sensitivity to DIDS

The sensitivity of the three measurable Pi transport systems in OK-cell monolayers to the potent anion-exchange inhibitor, DIDS, was measured. Monolayers were pre-incubated for 10 min with DIDS on the side of the monolayer on which transport measurements were conducted and then transport measured as usual. Application of DIDS to the monolayer resulted in a specific inhibition at the lower concentrations

This report presents the initial data characterizing apical and basolateral Pi transport systems in a renal continuous cell line grown on permeant support. The use of porous growing support permitted free access of the growing monolayer to growth media from both sides and free access to either side of the confluent monolayer to experimental manipulation. This is important, as it has been found that many epithelial cells differentiate more fully, i.e. sort their membrane proteins more completely, when grown on a porous support (Sariban-Sohraby et al. 1983; Handler et al. 1984). The fundamental results of the present study thus concern the formation of "fully" polarized monolayers and the identification and characterization of the polarity of P~transport systems in this cell line with proximal tubular properties. The basic properties of the apical P~ transport system were essentialy identical to those of renal proximal tubular membrane vesicles (Gmaj and Murer 1986) and of monolayers of this cell line grown on impermeant plastic dishes (Matmstr6m et al. 1986; Biber et at. 1988; Quamme et al. t989). This finding is important as it (1) suggests that these cells grown on the collagen-coated f~ters express apical Na/ P~ co-transport as in in vivo epithelium; (2) corroborates previous results from this cell line grown on impermeant support, suggesting that only apical systems are measured in that growth condition/experimental protocol; (3) provides further evidence that these monolayers were differentiated and thus validates data collection on the basotateral surface P~ transport systems.

559

tO

o

e"

8~ 8o

e~

40 4-J tt~ 0 eQ.

20

0

0

f

I

I

I

50

100

150

200

DIDS (pM) Fig. 6. 4,4-Diisothiocyanostilbene2,2 disulphonic acid (DIDS) inhibition of apical Na-dependent (stars), basolateral Na-dependent (cross) and basolateral Na-independent (squares) phosphate (P~)transport in OK-cellmonolayers. Pi concentrationsin the uptake solutions were 0.1 mM and 1 mM P~in the apical and basolateral compartments, respectively. All points are means _+SE of three replicates

The key finding of the present study was that in OKcell monolayers Pi was transported through the basolateral side by two processes; a Na-independent system and a Nadependent system. The Na-independent system at 1 mM phosphate accounted for approximately 30% of total basolateral transport and was characterized by low-affinity (Kin = 2.6 mM Pi), high-capacity (Jm~ = 2100 pmol Pi/mg protein per minute) kinetics (Fig. 4b, Table 1), and inhibition by the anion exchange inhibitor, DIDS, at moderate (50/~M) concentrations (Fig. 6). These features are practically identical to the system described from renal proximal tubular BLMV (Grinstein et al. 1980; Ullrich and Murer 1982; Hammerman and Schwab 1984). The inhibition of only the Na-independent system by low concentrations of DIDS, an anion-exchange inhibitor, suggests that this transport step may be linked to anion exchange and is not, therefore, strictly facilitated diffusion (Grinstein et al. 1980; Ullrich and Murer 1982; Gmaj and Murer 1986). The basolateral Na-dependent Pi transport system accounted for two - thirds of total basolateral transport at i mM phosphate and was also a low affinity (Km = 5.2 mM Pi), high capacity (Jmax = 2400 pmol PJmg protein per minute) system with a slightly lower affinity than the Na-independent system (Fig. 4b, Table 1). Although the kinetic characteristics of the basolateral Na-dependent Pi system presented here are very different from those reported by Schwab et al. (1984a, b) from BLMV of dog kidney, the present data support the concept of a Na-dependent P~ transport process occurring at the basolateral surface. The large differences in Pi concentration kinetics and, for the Na-dependent systems, the Na:Pi stoichiometry between the sets of transport systems on the two cell surfaces is compatible with vectorial transcetlular Pi absorption. The

basolateral Na-independent system had kinetics (Table 1) consistent with a role fitting within the classical concept of basolateral facilitated diffusion (Kinne 1986). The half-saturation constant of this Na-independent system (Kin: 2.6 mM Pi) was similar to values previously reported (Grinstein et al. 1980; Hammerman and Schwab 1984) and is compatible with its suggested efflux role when compared to plasma and/or cytoplasmic Pi values (Sullivan and Grantham 1982). The presence on the basolateral surface of a Na-dependent Pi transport system is still consistent with a polarization of transport systems with respect to vectorial transcellular absorption. The pattern of differences between these two systems for Pi kinetics (Kin: apical, 0.4 mM; basolateral, 5.2 mM) and particularly in stoichiometry (apical: n -- 3, K Na = 105 mM, basolateral: n = 1, K Na = 25 raM) fit the basic requirements presented by Kinne (1986) in his model of vectorial transcellular transport of solutes in epithelium having active systems on both cellular surfaces. The large difference in stoichiometric characteristics is very important stoichastically and energetically in terms of both the rate of transport and the ability to concentrate Pi against a thermochemical gradient (Aronson 1984), giving the apical surface the thermodynamic advantage as an absorptive membrane. Thermodynamic calculations, taking into account the basolateral Na/P~ stoichiometric value (apparent Hill coefficient of 1) measured in the present study together with a membrane potential of ~ - 65 mV cell interior negative, a cell to peritubular Na concentration ratio of 1:10 and a predominantely divalent transported Pi species (Biber 1989) suggest that Pi flux through the basolateral Na-dependent system is at its thermodynamic balance at intracellular Pi concentrations close to that in the basolateral compartment. Although these OK-cell monolayers are organized for the vectorial absorption of Pi from the luminal through contra-luminal surfaces, the question arises to why there would be two basolateral systems for export of Pi from the cell. An important component of intracellular metabolism is an intracellular P~ concentration kept within relatively narrow limits. As the basolateral Na-dependent transport system is in a dynamic equilibrium with the peritubular:cell Pi ratio it might also serve to fulfil metabolic and/or structural nutritive requirements when they cannot be met by apical transport, as was suggested by Schwab et al. (1984). This mechanism would be analogous in function to that proposed by Barfuss et al. (1980) for Na-dependent basolateral glycine transport in rabbit kidney and Reshkin et al. (1988) for basolateral Na-dependent glycine, proline and glutamate transport in fish intestine. Thus, the specific basolateral Na-dependent Pi system may serve to give the cell extra flexibility in adjusting to serum Pi levels. That a basolateral Na-dependent transport system is not universally expressed may reflect differing epithelial needs or function or supply that exist in different parts of an organ or in different species. The presence of a Na-dependent basolaterat Pi transport system could provide a measure of frequency of the necessity for the presence of Pi above that which is luminally available for optimal cellular function. Thus different specific cellular purpose or environment may have served to modify the basolateral transport system(s) available for Pi.

560 I n s u m m a r y , O K cells are able to f o r m tight, f u n c t i o n ally p o l a r i z e d m o n o l a y e r s o n c o l l a g e n - c o a t e d filters that express, in total, three Pi t r a n s p o r t systems o n the two cell surfaces. T h e characteristics of these t r a n s p o r t systems are similar to those r e p o r t e d p r e v i o u s l y a n d are consistent b o t h with the k n o w n role of r e n a l epithelial cells in transceUular Pi r e a b s o r p t i o n a n d with a n a d d i t i o n a l role that could serve to fulfil p o t e n t i a l a d d i t i o n a l m e t a b o l i c r e q u i r e m e n t s .

Acknowledgements. We would like to thank Franqois Wuarin for helpful discussion and Denise Rossi for excellent secretarial assistance. We are indebted to the Roche Research Foundation, Basel, Switzerland, for providing a visting scientist award to S. J. R. and to the Swiss National Science Foundation for financial support (grant 3.854.088 to I-I. M.)

References Aronson PS (1984) Electrochemical driving forces for secondary active transport: energetics and kinetics of Na+-H + exchange and Na +glucose cotransport. In: Blaustein MP, Liberman M: Electrogenetic transport: fundamental principles and physiological implications. Raven Press, New York, pp 49-70 Barfuss DW, Mays JM, Schafer JA (1980) Peritubular uptake and transepithelial transport of glycine in isolated proximal tubules. Am J Physiol 238:F324-F333 Biber J (1989) Cellular aspects of proximal tubular phosphate reabsorption. Kidney Int 36:360-369 Biber J, Forgo J, Murer H (1988) Modulation of Na+-P~ cotransport in opposum cells by extracellular phosphate. Am J Physiol 255: C155-C161 Garay RP, Hannaert PA, Nazaret C, Cragoe EJ (1986) The significance of the relative effects of loop diuretics and antibrain edema agents on the Na, K, C1 cotransport system and the C1/NaCO3 anion exchanger. Naunyn Schmiedebergs Arch Pharmacol 334:202-209 Omaj P, Murer H (1986) Cellular mechanisms of inorganic phosphate transport in kidney. Physiol Rev 66:36-70 Orinstein S, Turner RJ, Silverman M, Rothstein AI (1980) Inorganic anion transport in kidney and intestinal brush border and basolateral membranes. Am J Physiol 238:F452-F460 Hagenbuch B, Murer M (1986) Phosphate transport across the basolateral membrane from rat kidney cortex: sodium-dependence? Pfliigers Arch 407:149-155 Haggerty JG, Agarwal N, Reilly RF, Adelberg EA, Slayman CW (1988) Pharmacologically different Na/H antiporters on the apical and basolateral surfaces of cultured porcine kidney cells (LLCPK1). Proc Natl Acad Sci USA 85:6797-6801 Hammerman MR, Schwab SJ (1984) Phosphate transport in the kidney. Studies with isolated membrane vesicles. In: Bronner F, Pe-

terlik M (eds) Epithelial calcium and phosphate transport: molecular and cellular aspects. Liss~ New York, pp 325-330 Handler JS, Preston AS, Steele RE (1984) Factors affecting the difretention of epithelial transport and responsiveness to hormones. Fed Proc 43:2221-2224 Kinne R (1986) Epithelial transport: The interplay between ionic gradients and cell polarity. In: Alvarado F, van Os C (eds) Ion gradient-coupled transport (INSERM Syrup 26). Elsevier, Amsterdam, pp 255-273 Koyama H, Goodpasture C, Miller MM, Teplitz RL, Riggs ID (1978) Establishment and characterization of a cell line from the American Opposum (Didelphys virginiana). In Vitro 14:239-246 MalmstrSm K, Murer H (1986) Parathyroid hormone inhibits phosphate transport in OK cells but not in LLC-PKI and JTC-12.P3 cells. Am J Physiol 251:C23-C31 Malmstr6m K, Stange G, Mnrer H (1988) Intracellular cascades in the parathyroid-hormone-dependent regulation of Na+/phosphate cotransport in OK cells. Biochem J 251:207--213 Motulsky H J, Ransnas LA (1987) Fitting curves to data using nonlinear regression: a pratical and nonmathematical review. FASEB J i: 365-374 Murer H (1988) Transport of sulfate and phosphate in small intestine and renal proximal tubule: methods and basic properties. Comp Biochem Physiol 90:749-755 Murer H, Gmaj P (1986) Transport studies in plasma membrane vesicles isolated from renal cortex. Kidney Int 30:171-186 Murer H, Biber J, Gmaj P, Stieger B (1984) Cellular mechanisms in epithelial transport: advantages and disadvantages of studies with vesicles. Mol Physiol 6:55-82 Quamme GA, Biber J, Murer H (1989) Sodium-phosphate cotransport in OK-cells: inhibition by PTH and "adaptation" to low phosphate diet. Am J Physiol 257:F967-F973 Rabito CA (1983) Phosphate uptake by a kidney cell line (LLC-PK1). Am J Physiol 245:F22-F31 Reshkin SJ, Vilella S, Cassano G, Ahearn GA, Storelli C (1988) Basolateral amino acid and glucose transport by the intestine of the teleost, Anguilla anguilla. Comp Biochem Physiol 91A: 779788 Sariban-Sohraby S, Burg MB, Turner RJ (1983) Apical sodium uptake in the toad kidney epithelial cell line A6. Am J Physiol 245: C167C171 Schwab SJ, Klahr S, Hammerman MR (1984) Na + gradient-dependent P~ uptake in basolateral membrane vesicles from dog kidney. Am J Physiol 246:F663-F669 Schwab SJ, Klahr S, Hammerman MR (1984) Uptake of P~ in basolateral vesicles after release of unilateral ureteral obstruction. Am J Physiol 247:F543-F547 Simons K, Fuller SD (1985) Cell surface polarity in epithelia. Annu 9 Rev Cell Biol 1:243-288 Sullivan LP, Grantham JJ (1982) Physiology of the kidney, 2"deds. Lea and Febiger, Philadelphia Ullrich KJ, Murer H (1982) Sulphate and phosphate transport in the renal proximal tubule. Philos Trans R Soc Lond [Biol] 229: 549558