|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
EXTRACELLULAR MATRIX, CELL INTERACTIONS
1Department of Physiology, Faculty of Medicine, Kuwait University, Kuwait; 2Department of Pediatrics, Albert Einstein College of Medicine, New York, New York; and 3Department of Medicine, University of Maryland, Baltimore, Maryland
Submitted 5 February 2006 ; accepted in final form 20 July 2007
| ABSTRACT |
|---|
|
|
|---|
renal; mineral; transport; paracellular; transcellular; cell compartments
In contrast to what occurs in the proximal tubule in vivo, we (2) and others (12) have shown that in monolayers of kidney proximal tubule-like cells in culture there is no adaptive increment in transepithelial Pi transport despite a rise in A and basolateral (B) Na+ gradient-dependent Pi influx into the cells (2, 10). The absence of transepithelial Pi transport in vitro may be the consequence of large paracellular fluxes that may overwhelm and mask any increase in transcellular Pi transport. Indeed, we found that large (60–80% of the Pi flux) fluxes of mannitol occur across monoloyers of OK cells grown on uncoated filters (2).
We have also shown that in OK cells grown on uncoated filters, transcellular Pi fluxes and the intracellular Pi pool were reduced in response to severe Pi deprivation, despite a decrease in paracellular fluxes and increased influx of Pi across A and B cell membranes (1, 2). Effective permeability for cell Pi efflux across B or A cell membranes and the distribution of intracellular Pi among inorganic and organophosphates fractions were unaltered (2). These findings suggest that in severely Pi-deprived OK cells grown on uncoated filters, Pi enters cell compartments where it is not available for transcellular transport. Cell compartments depend on the structure of the cytoskeleton (14), which may be influenced by cell-matrix interactions through kinases and proteins linking the cytoskeleton to the cell membrane (NHERF1-PDZK1-MAP17-Ezrin) (4, 5, 7). Our findings on Pi-depleted (Pi,d) cells stimulated our interest to study transepithelial Pi transport in OK cells grown on a matrix more akin to the natural basement membrane.
| METHODS |
|---|
|
|
|---|
The cells were exposed 16 h to FBS-free Ham F12 solutions containing either low (0.25 mM) or high (2.5 mM) Pi. The cells were then exposed to 2 ml FBS-free Ham-F12 containing 0.25 mM 32Pi and [14C]mannitol in the A or the B compartments and to 2 ml FBS-free Ham F12 containing 0.25 mM unlabeled Pi and 0.25 mM mannitol in the opposite compartment. Aliquots (25 µl) were sampled from the latter compartment (B or A, respectively) after 5 or 15, 30, and 60 min to determine transepithelial 32Pi and [14C]mannitol fluxes, in triplicate wells.
Uptakes of 32Pi from either the A or the B compartments into the OK cell monolayers were measured after 5 min exposure to 0.25 mM 32Pi. Uptake was stopped by aspiration and by three rinsings of the two compartments with ice-cold 100 mM Mg gluconate. The cells were lysed in 2 ml 0.5% Triton X100 in 0.2 N KOH in a plate shaker for 1 h. Aliquots of the lysate were used for Liquid Scintillation counting in 10 ml Aquasol (2). Proteins were measured using the Bio-Rad Coomassie Blue reagent (2) and
-globulin in 0.5% Triton 0.2 N KOH as standard.
Fluorescence microscopy. OK cells were grown to confluency on filter inserts uncoated or coated with Matrigel (1:10 in DMEM-F12), exposed for 16 h to 0.25 or 2.5 mM Pi in DMEM-F12, fixed for 20 min with 3% paraformaldehyde in PBS supplemented with Ca2+-Mg2+, and quenched for 10 min with 20 mM glycine before staining. After being blocked with 10% goat serum in PBS, cells were incubated overnight with COOH-terminal anti-NaPi4 (an isoform of the sodium-phosphate cotransporter) chicken primary antibody and the next day rinsed and incubated with Alexa 488 Fluor-conjugated secondary rabbit antibody (Molecular Probes, Eugene, OR) against chicken IgG/IgY. Samples were mounted in 90% glycerol (Merck, Darmstadt, Germany) and 10% PBS containing 2.5% 1,4-diazabicyclo[2.2.2]octane (Sigma). Imaging was done with a confocal microscope (LSM 410, C. Zeiss, Thornwood, NY) with x63 and x40 oil immersion objectives.
The OK cell line was kindly provided by Drs. Biber and Murer (Zurich, Switzerland). Cell culture supplies, including the PET filter inserts, were obtained from BD, Falcon (Lincoln Park, NJ), or GIBCO-BRL (Grand Island, NY). Radioisotopes were purchased from DuPont New England Nuclear (Boston, MA). Chemicals were procured from Sigma (St. Louis, MO) or Baker (Phillisburg, NJ).The NaPi4 antibody was kindely provided by Dr. Moshe Levy (Denver, CO).
Data were analyzed using paired or unpaired t-tests. A P < 0.05 was considered statistically significant. Values are presented as means ± SE of at least three experiments using in each, triplicate determinations for any given experimental condition.
| RESULTS |
|---|
|
|
|---|
|
|
0.7 (P < 0.01 and P < 0.001, respectively) with Pi depletion. The net transepithelial mannitol flux was from B to A and decreased insignificantly (P > 0.1) with Pi depletion (Table 1). The estimated unidirectional transcellular Pi fluxes (Table 1) were not significantly altered (P > 0.3) by Pi depletion. The net transcellular Pi flux from A to B decreased by half, but the change was not significant (P > 0.4).
|
The largest effects of Matrigel were seen in Pi,d OK cells: Pi,d cells grown on Matrigel-coated compared with uncoated filters had higher fluxes of both Pi and mannitol (Table 1). This was true for the transepithelial unidirectional 32Pi flux from A to B, which was 10.6-fold higher (P < 0.001); for the transepithelial Pi flux from B to A, which was 7.7-fold higher (P < 0.01) and particularly for the net transepithelial Pi flux, which was 33-fold higher (P < 0.01). The paracellular mannitol fluxes (Table 1) were 3.3-fold (A to B, P < 0.01) and 3.7-fold (B to A, P < 0.001) higher and the net mannitol flux from B to A was 6.7-fold larger (P < 0.05) in Pi,d cells grown on Matrigel than in those grown on uncoated filters (Table 1). The estimated unidirectional transcellular Pi fluxes (Table 1) were
35-fold (transcellular A to B, P < 0.01; transcellular B to A, P < 0.05), and the net transcellular Pi flux was 20-fold higher (P < 0.05) in Pi,d OK cells grown on Matrigel than in those grown on uncoated filters. The increase in A to B net transcellular Pi flux was consequent to a larger Matrigel-associated increment in the A to B (102 nmol·mg–1·h–1) than in the B to A (45 nmol·mg–1·h–1) unidirectional Pi fluxes (Table 1).
The paracellular unidirectional fluxes (Table 1), represented smaller percentages of the respective transepithelial Pi fluxes in cells grown on Matrigel than in those grown on uncoated filters. Consequently, the estimated transcellular Pi fluxes represented a larger percentage of the transepithelial Pi fluxes in cells grown on Matrigel than in those grown on uncoated filters (Table 1).
Apical and basolateral Pi influx rates. A and B 32Pi (0.25 mM, 4,000 dpm/µl) influx rates into the cells (C) were higher in Pi,d compared with Pi,r cells (Table 1): In Pi,d cells grown on Matrigel, A to C and B to C Pi influxes were 5- and 3.6-fold higher, respectively, than in those in Pi,r cells (both P < 0.001, Table 1). In Pi,d cells grown on uncoated filters A to C and B to C, Pi influx rates were 1.9- (P < 0.01) and 2-fold (P < 0.05) as high as in Pi,r cells (P < 0.01, Table 1).
In Pi,r cells grown on Matrigel and on uncoated filters and in Pi,d cells grown on Matrigel, the rates of apical Pi uptake into cells were similar in magnitude to the respective estimated transcellular A to B Pi fluxes (Table 1). Consequently, in all these conditions, the apical uptake rates set the pace of the transcellular A to B Pi fluxes. By contrast, in Pi,d cells grown on uncoated filters, the A to C Pi uptake rate was 4-fold higher than the estimated transcellular A to B Pi flux (Table 1). In all conditions, basolateral (B to C) Pi uptake into cells was larger than the estimated respective transcellular B to A Pi fluxes (Table 1). Thus the rate of transcellular B to A Pi flux is not set by the basolateral Pi uptake.
Immunofluorescence. Confocal imaging at x63 exhibited (Fig. 3, XY plane at center of each image) abundance of cells labeled by indirect NaPi4 immunofluorescence with a patchy distribution on the cell surface, in Pi,d cells grown on Matrigel but not on those grown on uncoated filters (Fig. 3). Orthogonal sections in the XZ plane (Fig. 3, top insets) and in the YZ plane (Fig. 3, right inset) confirmed the apical localization of the immunofluorescence in the Matrigel grown Pi-deprived cells and its scarcity in cells grown on uncoated filters. Conventional fluorescence imaging at x40 revealed that indirect NaPi4 inmunofluorescence in Pi,d OK cells was 2.5-fold stronger (P < 0.01) in cells grown on Matrigel than in those grown on uncoated filters (see Fig. 3) In turn, Pi,d cells exhibited stronger fluorescence intensity than Pi,r cells grown on uncoated or on Matrigel-coated filters, which hardly exhibited any inmunofluorescence. The enhanced expression of NaPi4 in Pi,d cells grown on Matrigel than in those grown on uncoated filters is congruent with the higher A to C and A to B Pi fluxes observed in these cells.
|
Cell-matrix (Matrigel) interactions increased unidirectional and net transepithelial and transcellular Pi fluxes, as well as paracellular mannitol fluxes, in Pi,d OK cells. In Pi,r OK cells, matrix interactions reduced unidirectional Pi transepithelial and paracellular mannitol fluxes but did not significantly alter net transepithelial, transcellular Pi, or paracellular fluxes.
| DISCUSSION |
|---|
|
|
|---|
These findings differ from those in cells grown on uncoated filters (2). Under these conditions, the apical Pi uptake also increased with Pi depletion (10), but the change was smaller (2 vs. 5-fold) compared with that in cells growing on Matrigel. Since interactions between cell integrins and matrix can alter the cytoskeleton (7) and interactions between the cytoskeleton and the cell membrane are involved in the upregulation of Pi transport in response to Pi depletion (4), it is not surprising that the magnitude of the upregulatory response was found to be dependent on cell-matrix interactions. The higher Pi fluxes observed in Pi,d OK cells grown on Matrigel correspond with the enhanced expression of NaPi4 cotransporters observed by indirect inmunofluorescence confocal microscopy (Fig. 3).
In Pi,d cells grown on uncoated filters only a fraction (0.25) of Pi entering across the apical cell membrane reached the Pi transport pool and appeared on the B side. The restriction was not at the basolateral cell membrane. Indeed, the unidirectional basolateral Pi efflux (C to B) rate, estimated from the net transcellular Pi transport rate plus the measured basolateral (B to C) Pi uptake rate, was found to exceed by sixfold the magnitude of the transcellular A to B Pi flux . Thus basolateral Pi efflux could not have limited the access of Pi transported across the apical cell membrane to the B compartment. It is more likely that in Pi,d cells grown on uncoated filters, most Pi transported across the apical membrane reaches a Pi pool that is unavailable for transcellular transport, perhaps a subapical pool or another intracellular pool, from which it can backflow to the A compartment or enter other Pi pools. Apparently, changes in cell matrix interactions led to changes in the compartmentalization of the intracellular Pi pools possibly through altered cytoskeletal and membrane-cytoskeletal interactions (7, 14).
Cell-matrix (Matrigel) interactions also led to changes in paracellular (mannitol) fluxes in Pi,r and Pi,d cells. This is consitent with their influence of these interactions on the permeability of tight junctions (13). The effect of cell-matrix interactions on paracellular fluxes was found to depend on the Pi status of the cells. Paracellular fluxes were lower in Pi,r cells (0.7x) but higher (3.5x) in Pi,d cells, when grown on Matrigel compared with uncoated filters. This suggests that signaling initiated by cell-matrix interactions is altered by Pi depletion or repletion. However, the fraction of the transpithelial Pi flux that could be accounted for by paracellular flux in the same direction was smaller (for example, A to B flux, 0.33 vs. 0.68) in cells grown on Matrigel than in cells grown on uncoated filters, independently of their Pi status. This was due to the concomitant increases in transcellular Pi fluxes observed in cells grown on Matrigel (
12x), which were particularly large in Pi,d cells (
35x) That means that changes in transcellular Pi transport have a larger influence on transepithelial Pi transport in cells grown on Matrigel than in those grown on uncoated filters.
The interaction of cells with Matrigel was associated with larger basolateral Pi influx (B to C) than observed in cells grown on uncoated filters. This suggests that Matrigel-cell interactions promote the expression or the activity of yet to be identified Pi transporters at the B membrane, particularly in Pi,d cells. In all cases, however, the basolateral Pi uptake rates exceeded the transcellular B to A Pi fluxes, indicating that this transcellular flux was not limited by basolateral Pi uptake. Indeed except for Pi,d cells on uncoated filters, the transcellular B to A Pi flux was found to nearly equal the magnitude of the apical Pi backflux [estimated as the difference between the measured apical Pi uptake (A to C) rate and the net transcellular Pi flux]. This suggests that in these cells the pace of transcellular B to A Pi flux was set by the rate of Pi efflux across the apical cell membrane. However, in Pi,d cells grown on uncoated filters, the transcellular B to A Pi flux was lower than the estimated apical Pi efflux suggesting that, in this condition, there is entry of basolateral Pi into pools that do not participate in transcellular Pi transport. Such trapping of Pi in pools that do not exchange with the Pi transport pool reflects intracellular Pi compartmentalization, dependent on the nature of cell-matrix interactions in Pi,d cells. That Pi entering via A or B membranes go into different OK cell compartments was already suggested by the finding that adaptive upregulation of Pi transport in response to Pi depletion only occurs when the depletion is at the apical compartment (10).
Because the B to C Pi uptake rate was always larger than the transcellular B to A Pi flux, it is likely that a large fraction of the basolateral Pi uptake recycles across the basolateral membrane. Indeed, we have shown exchange of basolateral for intracellular Pi and transstimulation of basolateral Pi efflux from OK cells by Pi in the basolateral compartment (3). Exchange of basolateral HPO4= for cell H2PO4- results in net Pi transport across the basolateral membrane if the exchange ratio is not 1. In all conditions described here, there was net transfer of Pi to the basolateral compartment, and the estimated basolateral Pi efflux exceeded the basolateral Pi uptake into the cells. The estimated basolateral Pi exchange ratios in Pi,d and Pi,r cells were 1.7 and 1.3 on Matrigel and 1.2 and 1.7 on uncoated filters, respectively. A high exchange ratio may be necessary, but is not sufficient, for net apical to basolateral transepithelial Pi transport, because in Pi replete cells grown on uncoated filters, the ratio was high but net transepithelial Pi transport was the lowest.
A potentially significant caveat is that, in our discussion, paracellular Pi fluxes have been equated with those of mannitol. This implies that Pi and mannitol have the same permeance across the paracellular pathway. Since size and charge differ between these molecules, the assumption may be inaccurate. If mannitol fluxes overestimate paracellular Pi fluxes, then transcellular fluxes may have been underestimated. However, in most conditions studied (except for Pi,d cells on uncoated filters) the transcellular A to B fluxes were found to closely match the magnitude of the Pi influxes measured across the apical cell membranes. This widely accepted characteristic of renal transepithelial Pi transport (6) lends support to our postulate that paracellular permeability to Pi and mannitol are similar. Since there is no reason to believe that paracellular Pi and mannitol permeabilities only differ in Pi,d cells grown on uncoated filters, we surmize that in cells grown on Matrigel and in Pi,r cells grown on uncoated filters, but not in Pi,d cells grown on uncoated filters, transcellular Pi fluxes are limited by apical membrane fluxes rather than by intracellular compartmentalization of Pi. Furthermore, in no condition basolateral membrane fluxes were found to limit trancellular Pi fluxes.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
2. Barac-Nieto M, Alfred M, Spitzer A. Phosphate depletion in opossum kidney cells: apical but not basolateral or transepithelial adaptions of Pi transport. Exp Nephrol 9: 258–264, 2001.[CrossRef][ISI][Medline]
3. Barac-Nieto M, Alfred M, Spitzer A. Basolateral Pi transport in Pi deprived OK cells. Exp Biol Med (Maywood) 227: 626–631, 2002.
4. Hernando N, Deliot N, Gisler SM, Lederer E, Weinman EJ, Biber J, Murer H. PDZ-domain interactions and apical expression of type IIa Na/P(i) cotransporters. Proc Natl Acad Sci USA 99: 11957–11962, 2002.
5. Humphries MJ. Monoclonal antibodies as probes of integrin priming and activation. Biochem Soc Trans 32: 407–411, 2004.[CrossRef][ISI][Medline]
6. Murer H, Hernando N, Forster I, Biber J. Molecular aspects in the regulation of renal inorganic phosphate reabsorption: the type IIa sodium/inorganic phosphate co-transporter as the key player. Curr Opin Nephrol Hypertens 10: 555–561, 2001.[CrossRef][ISI][Medline]
7. Ng T, Parsons M, Hughes WE, Monypenny J, Zicha D, Gautreau A, Arpin M, Gschmeissner S, Verveer PJ, Bastiaens PI, Parker PJ. Ezrin is a downstream effector of trafficking PKC-integrin complexes involved in the control of cell motility. EMBO J 20: 2723–2741, 2001.[CrossRef][ISI][Medline]
8. Quamme GA, Mizgala CL, Wong NLM, Whiting SJ. Effects of intraluminal pH and dietary phosphate on phosphate transport in the proximal convoluted tubule. Am J Physiol Renal Fluid Electrolyte Physiol 249: F759–F768, 1985.
9. Reshkin SJ, Forgo J, Murer H. Functional asymmetry of phosphate transport and its regulation in opossum kidney cells: phosphate transport. Pflügers Arch 416: 554–560, 1990.[CrossRef][ISI][Medline]
10. Reshkin SJ, Forgo J, Biber J, Murer H. Functional asymmetry of phosphate transport and its regulation in opossum kidney cells: phosphate adaptation. Pflügers Arch 419: 256–262, 1991.[CrossRef][ISI][Medline]
11. Saoncella S, Calautti E, Neveu W, Goetinck PF. Syndecan-4 regulates ATF-2 transcriptional activity in a Rac1-dependent manner. J Biol Chem 279: 47172–47176, 2004.
12. Scheinman SJ, Reid R, Coulson R, Jones DB, Ford SM. Transepithelial phosphate transport in rabbit proximal tubule cells adapted to phosphate deprivation. Am J Physiol Cell Physiol 266: C1609–C1618, 1994.
13. Tafazoli F, Holmstrom A, Forsberg A, Magnusson KE. Apically exposed, tight junction-associated beta1-integrins allow binding and YopE-mediated perturbation of epithelial barriers by wild-type Yersinia bacteria. Infect Immun 68: 5335–5343, 2000.
14. Wozniak MA, Modzelewska K, Kwong L, Keely PJ. Focal adhesion regulation of cell behavior. Biochim Biophys Acta 692: 103–19, 2004.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |