INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

USSING CHAMBER STUDIES WITH permanent cell lines of epithelial origin have contributed significantly to our understanding of ion secretory processes and their regulation. These cell lines include T84 (colon; Ref. 5) and Madin-Darby canine kidney (MDCK; Ref. 17) among others. However, no cell line of salivary gland origin, either acinar or ductal, has been reported to be useful in Ussing chamber studies of transcellular ion movements and their regulation by neurotransmitters.

Recently, we reported the establishment of simian virus 40-transformed cell lines of rat parotid (15) and submandibular (14) gland acinar origin. Although not without some inconsistencies, these cell lines exhibit a substantial degree of fidelity to the cell type of origin, including morphological, biochemical, and functional characteristics. Among these is the expression of receptors for salivary gland-relevant neurotransmitters, including norepinephrine, acetylcholine, vasoactive intestinal peptide, and extracellular nucleotides. In this report, we describe experiments indicating the suitability of one of the parotid gland cell lines, Par-C10, in Ussing chamber studies, which revealed the presence in these cells of an anion-dependent short-circuit current (I_sc) that is regulated by neurotransmitters.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell culture. Par-C10 cells (5 × 10⁵) were plated on Falcon cell culture inserts (diameter 2.4 cm, pore size 0.4 µm; Becton Dickinson, Franklin Lakes, NJ) coated with 0.10 mg/ml bovine type I collagen. The cultures were grown to confluence in DMEM-F12 (1:1) containing 2.5% fetal bovine serum (GIBCO BRL, Gaithersburg, MD) and the following supplements: 0.1 µM retinoic acid, 80 ng/ml epidermal growth factor, 2 nM triiodothyronine, 5 mM glutamine, 0.4 µg/ml hydrocortisone, 5 µg/ml insulin, 5 µg/ml transferrin, 5 ng/ml sodium selenite, 50 µg/ml gentamicin, and 8.4 ng/ml cholera toxin. In some experiments, as indicated, the cholera toxin was omitted. Cells were cultured at 37°C in a humidified 95% air-5% CO₂ atmosphere and used at confluence, typically 4-5 days after plating. Cells at passages 50-80 were utilized in these studies, with little change in most of the parameters examined, except as noted in the RESULTS and DISCUSSION sections.

Morphological evaluation. Par-C10 cells were prepared for microscopic evaluation as described previously (15) except that the cells were grown on permeable Falcon cell culture inserts. Briefly, confluent cultures were fixed for 1 h in 2.5% glutaraldehyde in 0.1 M NaPO₄, washed three times, and placed in phosphate buffer containing 0.15 M sucrose. Fixation in OsO₄ and other steps in the preparation of sections for transmission electron microscopy were as described previously (2).

Measurement of I_sc in Par-C10 monolayers. Confluent cultures of Par-C10 cells on permeable supports were mounted in modified Ussing chambers for measurement of agonist-induced changes in I_sc, as described previously for studies with T84 colonic epithelial cells (8). The standard medium for both the apical and basolateral reservoirs (volume 5 ml) was Krebs-Ringer-HCO₃ buffer, pH 7.5, containing (in mM) 118 NaCl, 3 KCl, 1.2 MgSO₄, 25 NaHCO₃, 1.0 CaCl₂, and 10 glucose. HCO₃-free medium, pH 7.5, was buffered with 15 mM HEPES. Cl-free media utilized gluconate as the replacement anion with the Ca²⁺ concentration increased to 4 mM to counteract Ca²⁺ chelation by gluconate. Na⁺-free medium was prepared with N-methyl-D-glucamine or choline. Alterations to the apical or basolateral media for specific experiments are indicated in legends of Figs. 3-7. The Ussing apparatus included a water jacket to maintain the buffer temperature at 37°C, and the medium in both reservoirs was mixed and oxygenated by bubbling with 95% O₂-5% CO₂. I_sc was measured continuously, and transepithelial potential difference was measured intermittently, using a VCC-600 automatic voltage-clamp apparatus (Physiologic Instruments, San Diego, CA) and calomel electrodes connected to the chamber baths with 4% agar-KCl bridges. I_sc and automatic fluid resistance compensation current were applied through Ag-AgCl electrodes connected to chamber baths with 4% agar-KCl bridges. Resistance measurements were made by occasionally clamping the potential difference to a known voltage and measuring the current required to establish the potential. Resistance and conductance were calculated using Ohm's law. I_sc values are expressed as the peak change obtained in response to agonist (µA/cm²) or alternatively as the area under the curve for the first 2 min following agonist addition (arbitrary units) to incorporate differences in the sustained phase of the response.

RT-PCR. Total RNA was prepared from confluent Par-C10 cultures in 100-mm-diameter dishes (passages 60-61) using an RNeasy kit (Qiagen, Chatsworth, CA). Cell lysates were treated with RNase-free DNase. cDNA was synthesized from ~1 µg of Par-C10 total RNA with random hexamer primers using an Advantage RT-for-PCR kit (Clonetech, Palo Alto, CA). The RT-PCR mixture included 2.5 units of Vent(exo) DNA polymerase, 0.125 units of Vent DNA polymerase, 6 mM MgSO₄, 400 µM dNTP, and 40 pmol of each primer in a 50-µl reaction volume. Template and primers were first denatured in a buffer with high salt concentration for 5 min at 95°C. The tubes were moved immediately to 4°C, the enzymes and dNTPs were added, and ultrapure water was used to give the final reaction volume. The PCR cycle consisted of denaturation at 95°C for 1 min, annealing for 1 min (the annealing temperature and number of reaction cycles varied with specific primer sets), and elongation at 72°C for 1 min followed by a 72°C final extension for 5 min. The primer sequences used and details of the RT-PCR reactions are given in Table 1. As a positive control for the P2Y₆ primer set, cDNA was prepared from total RNA isolated from 1321N1 cells heterologously expressing the P2Y₆ receptor and amplified as described above. Aliquots from the PCR reactions were electrophoresed on agarose gels. The remaining PCR products were purified directly from the reaction mixture using a Wizard PCR Preps DNA purification system (Promega, Madison, WI) and sequenced with specific internal primers and fluorescent chain terminator dNTPs on an Applied Biosystems sequencer (Perkin-Elmer, Foster City, CA).

Data analysis. The statistical significance of differences between mean values was determined by unpaired Student's t-test or by one-way ANOVA and Student-Newman-Keuls post hoc test. Differences with P  0.05 were considered significant.

Materials. The following reagents were purchased from the indicated sources: fetal bovine serum, GIBCO BRL; epidermal growth factor, Collaborative Research (Bedford, MA); gentamicin, Fujisawa (Deerfield, IL); glutaraldehyde and OsO₄, EMCorp (Chestnut Hill, MA); DNase, Boehringer-Mannheim (Indianapolis, IN); and Vent and Vent(exo) DNA polymerases, New England Biolabs (Beverly, MA). All other reagents were obtained from Sigma Chemical (St. Louis, MO).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Development of transcellular resistance in Par-C10 cell cultures grown on permeable supports. As shown in Fig. 1, transcellular resistance across Par-C10 cell cultures grown on collagen-coated permeable supports increases as a function of time, approaching a maximum by 4 days. This time course was similar to that for cell proliferation, wherein confluence was typically attained after 3 days in culture. Par-C10 cultures grown on supports not coated with collagen or coated with other matrix components exhibited similar transcellular resistances at confluence, as did another parotid cell line, Par-C5, immortalized by the same technique used for Par-C10 (15). However, our initial investigations with these two parotid cell lines, as well as two immortalized submandibular gland cell lines, SMG-C6 and SMG-C10 (14), revealed that Par-C10 cells typically displayed the highest transcellular resistance values and exhibited the most polarized phenotype. As a result, we focused on the Par-C10 cell line for the studies described in this paper.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Effect of time in culture on transcellular resistance of Par-C10 monolayers. Par-C10 cells were plated at a density of 5 × 105 cells on collagen-coated permeable supports, as described in METHODS. Transcellular resistance values at indicated times in culture were obtained with an EVOM (World Precision Instruments, New Haven, CT) volt-ohmmeter with miniature dual chopstick electrodes. After subtraction of medium resistance (115 ), tissue resistance values were multiplied by effective membrane area (4.52 cm2) and are presented as means ± SE (n = 42).

Morphological evaluation. By transmission electron microscopy, confluent Par-C10 cell cultures consisted of a monolayer of plump cells attached to the collagen layer on the upper surface of the microporous membrane insert (Fig. 2A). The cells contained scattered secretory granules with a substructure that was mostly scanty but occasionally dense (Fig. 2, A and B). None of these was observed to be in the process of exocytosis. Cell processes extended into the pores of the membranes, some all the way to the bottom (Fig. 2, A and C), but did not spread out on the bottom surface. The cells were joined along their apical plasmalemmas by tripartite junctional complexes, consisting of tight and intermediate junctions and several desmosomes (Fig. 2, A and D). A terminal web was associated with the tight junctions (Fig. 2D). Numerous intercellular clefts were observed that had lumina segregated by junctional complexes (Fig. 2, A and D) and thus could be distinguished from ordinary intercellular spaces as secretory canaliculi (22). The plasmalemmas on the apical surfaces and lining the canaliculi were studded with small microvilli. The cytoplasm was rich in free ribosomes and also contained numerous mitochondria and dilated profiles of rough endoplasmic reticulum (Fig. 2, B and D). The only difference observed between the cells cultured with and without cholera toxin was that, in the latter, secretory granules occurred in clusters more frequently.

View larger version (163K): [in this window] [in a new window]   Fig. 2.   Morphological aspects of Par-C10 cells cultured on permeable supports. Shown are transmission electron micrographs of Par-C10 cells cultured with (A, C, and D) or without (B) cholera toxin and processed for morphological evaluation as described in METHODS. Arrowhead, tight junction component of a tripartite junctional complex; arrows, membrane pores containing cell processes; c, secretory canaliculi; g, secretory granules; r, rough endoplasmic reticulum; w, terminal web. Magnifications: A, ×3,900; B, ×13,400; C, ×5,700; D, ×24,800.

Agonist-induced changes in I_sc. We reported previously that Par-C10 cells are responsive, in terms of mobilization of second messengers, to salivary gland-relevant receptor agonists, including the muscarinic cholinergic agonist carbachol, the -adrenergic receptor agonist isoproterenol, the -adrenergic receptor agonist phenylephrine, and the P2 nucleotide receptor agonists ATP and UTP (15). Representative tracings of Par-C10 monolayer I_sc responses to maximally effective concentrations of three Ca²⁺-mobilizing agonists, UTP, carbachol, and phenylephrine, are presented in Fig. 3. As shown, UTP, when applied to the medium bathing the apical side of the monolayer, was the most efficacious of the three agents, whereas carbachol was more effective than phenylephrine when these latter two agents were applied basolaterally. The addition of UTP to the basolateral side or of carbachol or phenylephrine to the apical side was without effect on I_sc. The patterns and magnitudes of response shown in Fig. 3 for these three agonists were consistent across a wide range of passage numbers of Par-C10 cells. Conversely, although initial studies with the cAMP-mobilizing agonist isoproterenol (applied basolaterally) revealed a strong enhancement of I_sc, subsequent experiments revealed little or no I_sc response to this -adrenergic receptor agonist, despite the fact that cAMP production in response to isoproterenol was similar among the various cell preparations used. Thus, although many of the characteristics of Par-C10 cells appear to be maintained, others appear to be more variable.

View larger version (8K): [in this window] [in a new window]   Fig. 3.   Agonist-induced changes in Par-C10 short-circuit current (Isc). Confluent cultures of Par-C10 cells on permeable supports were placed in Ussing chambers and, following equilibration, were exposed to 100 µM UTP, carbachol, or phenylephrine. Isc as a function of time was recorded. In each case, addition of agonist to opposite side of cell monolayer had no effect on Isc.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Concentration dependence of agonist-enhanced Isc. Changes in Par-C10 Isc (Isc) were determined in response to indicated concentrations of ATP () and UTP () applied apically or carbachol () applied basolaterally. Data represent peak increases in Isc and are means ± SE of a minimum of 3 determinations. Associated curves are best fits of cumulative data to sigmoidal dose response (variable slope) model (Prism, Graphpad, San Diego, CA).

The ionic basis of the I_sc and the role of Ca²⁺. As in many other cell types, P2Y₂ nucleotide, muscarinic cholinergic, and ₁-adrenergic receptors in Par-C10 cells are coupled through phospholipase C to increases in the intracellular Ca²⁺ concentration (15), a process that involves both the mobilization of intracellular stores and the influx of extracellular Ca²⁺. We therefore examined the effect of removal of Ca²⁺ from the basolateral and apical media on agonist-induced increases in I_sc. As shown in Fig. 5, the magnitude of the response to apically applied UTP (A) and basolaterally applied carbachol (B), expressed as the area under the curve for the first 2 min following agonist addition, was unaffected when the Ca²⁺ concentration was lowered into the nanomolar range in the medium bathing the apical side of the monolayers. Conversely, lowering the basolateral medium Ca²⁺ concentration dramatically decreased, by 50% or more, the ability of both agonists to produce the sustained increases in I_sc observed in the presence of Ca²⁺. The effect of lowering the Ca²⁺ concentration in both baths simultaneously was not greater than that observed with the decrease in only the apical medium. This finding indicates that the response to both agonists is dependent on the availability of extracellular Ca²⁺ and that this Ca²⁺ is available to the cell only from the basolateral side.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Effect on agonist-stimulated Isc of Ca2+ removal from apical or basolateral incubation buffer. Par-C10 cell monolayers were placed in Ussing chambers and bathed on apical and basolateral sides with buffer containing either 1 mM Ca2+ (+) or 0.2 mM EGTA and no added Ca2+ (), as indicated. Changes in Isc were determined in response to 10 µM UTP applied apically (A) or 100 µM carbachol applied basolaterally (B). Values are expressed as area under curve during first 2 min after agonist addition and are means ± SE of 4 (UTP) or 3 (carbachol) experiments. In each panel, differences in mean values were significant between bars labeled a and b (1-way ANOVA with Student-Newman-Keuls post hoc test; P  0.05).

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Effect on agonist-stimulated Isc of Cl and HCO3 removal from incubation buffer. Par-C10 cell monolayers were placed in Ussing chambers and incubated in HCO3-buffered (A) or HEPES-buffered (B) medium containing 122 mM Cl (solid tracings) or in which Cl was replaced with gluconate (dashed tracings). Apical and basolateral buffer constituents were identical in each case. Changes in Isc in response to 10 µM UTP, applied apically, were then determined. Tracings shown in A and B are representative of a series of 3 experiments, summarized in C with data expressed as area under curve for first 2 min following UTP addition. Differences in mean values were significant between bars labeled a, b, and c (1-way ANOVA with Student-Newman-Keuls post hoc test; P  0.05).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Effect of Cl transport inhibitors on agonist-induced increases in Isc. Par-C10 cell monolayers were placed in Ussing chambers and, after equilibration, were preincubated for 2 min with 500 µM diphenylamine-2-carboxylic acid (DPC), 100 µM DIDS, or 150 µM 5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB) in apical (a) or basolateral (b) medium or with no addition (agonist only). UTP (10 µM; A) or carbachol (100 µM; B) was then added in continued presence of preincubation agent, and changes in Isc were determined. Values represent maximal increase obtained and are expressed as means ± SE of 3 or more experiments. * Significant difference from control cells (agonist only) using Student's unpaired t-test at P  0.05.

RT-PCR analysis for P2Y receptor subtypes and anion transporters in Par-C10 cells. We have confirmed the presence of muscarinic cholinergic receptors in Par-C10 cells with radioligand binding assays and by demonstrating the effectiveness of atropine in blocking carbachol-stimulated Ca²⁺ mobilization and I_sc (data not shown). Because there is no radioligand binding assay or high-affinity selective antagonist for the P2Y₂ receptor, we used RT-PCR detection of the mRNA for the P2Y₂ receptor to support the pharmacological identification (Fig. 4 and Ref. 15) of this subtype in Par-C10 cells. As shown in Fig. 8, RT-PCR with P2Y₂ receptor mRNA-specific primers produced a product of the correct size (778 bp) that was >99% identical to the published rat P2Y₂ receptor sequence. A recent study has indicated that another uridine-nucleotide-preferring P2Y receptor subtype, P2Y₆, is expressed, along with P2Y₂ receptors, in the apical membrane of another epithelial cell type (12). However, RT-PCR with primers specific for the P2Y₆ receptor mRNA amplified no appropriately sized (871 bp) product from cDNA prepared from Par-C10 cell total RNA (Fig. 8), whereas a product of appropriate size and sequence was obtained with cDNA prepared from 1321N1 cells heterologously expressing the P2Y₆ receptor. These results suggest that P2Y₆ receptors are not expressed in Par-C10 cells. A faint smaller band, amplified from Par-C10 cell RNA with the P2Y₆ primers in the presence of RT, is the same size as a more abundant band obtained with these primers in rat aortic smooth muscle cells. The sequence from the smooth muscle cells is unrelated to mRNAs encoding any known receptor proteins (data not shown). As is also shown in Fig. 8, RT-PCR with primers specific for two of the anion-transporting proteins identified previously in salivary gland acinar cells, the cystic fibrosis transmembrane conductance regulator (CFTR) (25) and the ClC-2 Cl channel (1), gave products of the appropriate sizes (352 and 755 bp, respectively) and sequences, suggesting that these anion channels are also expressed in Par-C10 cells.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Presence of P2Y2 receptor and Cl channel mRNAs in Par-C10 cells. Par-C10 cell total RNA was examined by RT-PCR (see METHODS) for presence of mRNA for P2Y2 and P2Y6 subtypes of nucleotide receptors and for ClC-2 and cystic fibrosis transmembrane conductance regulator (CFTR) Cl channels. Effectiveness of P2Y6 primers, which generated no relevant product from Par-C10 mRNA, was confirmed by RT-PCR with total RNA from 1321N1 cells heterologously expressing P2Y6 receptor. Similar to results shown for P2Y2 and P2Y6, no amplification was obtained in absence of RT () when using ClC-2 and CFTR primers (data not shown).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

One of the goals of salivary gland research at the cellular level has been to understand the pathways involved in the ion transport processes essential to saliva formation and modification. The understanding of equivalent processes and their regulation in other epithelia has been facilitated by the availability of permanent cell lines that exhibit sufficient polarization, transcellular resistance, and other features that make them suitable for bioelectric measurements in Ussing chambers. Two particularly useful examples are the T84 (colon; Refs. 5, 11, 23) and MDCK (kidney; Refs. 6, 17, 21) cell lines. The Par-C10 cell line is, to our knowledge, the first cell line of salivary origin that can be reliably utilized in Ussing chamber studies. Although the phenotype of Par-C10 cells is not entirely consistent with that of fully differentiated parotid acinar cells, a number of acinar cell characteristics are exhibited by Par-C10 cells (Ref. 15 and as discussed below).

The morphological observations (Fig. 2) indicate that cultured Par-C10 cells are apically polarized and clearly are acinar, albeit modestly, in terms of cytodifferentiation. In addition, the presence of secretory canaliculi among the cells is evidence of acinar morphodifferentiation, as these structures occur only in the acini of most salivary glands, including the rat parotid gland (16, 22). Here it is important to note that the occasional cells with mitotic figures shared the cited features of acinar differentiation. Thus this cell line is apparently purely acinar in that no ductal or other "stem" cells are seen. The presence of secretory canaliculi may be a consequence of culturing the cells on permeable supports: morphological analysis of Par-C10 cells grown on plastic tissue culture dishes revealed no such structures (15).

The polarized morphology of Par-C10 cells is complemented by the consistency between these cells and other polarized epithelia with respect to apical vs. basolateral distribution of important ion secretory proteins and receptors. The data shown in Figs. 3 and 4 suggest that ₁-adrenergic and muscarinic cholinergic receptor expression is limited to the basolateral aspect of Par-C10 cells, whereas P2Y₂ nucleotide receptor expression is apical, consistent with their distribution in other epithelia (7, 10, 13, 23). In addition, Par-C10 cells appear to express both CFTR and ClC-2 (Fig. 8) and possibly other anion transporting proteins relevant to normal salivary acinar cells (1, 25). The results with anion transport inhibitors (Fig. 7) suggest an exclusively apical distribution for these proteins, consistent with observations made in normal salivary acinar cells. Conversely, Par-C10 I_sc is not diminished by inhibitors of Na⁺-K⁺-2Cl cotransport such as bumetanide (Camden and Turner, unpublished observations), a finding considered more reflective of salivary duct cells than acinar cells (Ref. 24, but see Ref. 9). In addition, as reported previously (13), functional substance P receptors apparently are not expressed in Par-C10 cells, whereas evidence suggests that these receptors are found in normal salivary acinar but not ductal cells (4, 20). Finally, Par-C10 cell monolayers develop high transcellular resistances, similar to values obtained with MDCK (11, 23) and T84 cells (6, 21), and thus do not fit the classical definition of salivary acini as a "leaky" epithelium. Nonetheless, the suitability of Par-C10 cultures for studies of transcellular ion movement and its regulation promises to open new avenues for studying salivary gland secretion.

For the most part, agonist (UTP, carbachol, and phenylephrine) effects in terms of second messenger production (15) and increased I_sc (Figs. 3-7) were found to be reasonably consistent across at least 30 cell passages. One marked exception to this observation was the effect of the -adrenergic receptor agonist isoproterenol on I_sc, which varied from an increase similar to that obtained with maximally effective concentrations of UTP to no response at all, in various Par-C10 cultures (data not shown). This variability was paralleled by similar changes in the response to forskolin, whereas all of the cultures exhibited robust cAMP responses to isoproterenol and forskolin, suggesting that the expression of a component in the pathway downstream of the -adrenergic receptor, G_s, and adenylyl cyclase, possibly CFTR itself, may be altered. The decrease in the responsiveness to cAMP-mobilizing agents is not simply a function of passage number but apparently involves subtle differences in culture conditions or medium constituents, including cholera toxin, which we have used routinely in an attempt to promote a differentiated state of the Par-C10 cells. The basis for the differences in apparent CFTR activity requires investigation.

The above caveats notwithstanding, the Par-C10 parotid acinar cell line exhibits a number of acinar cell-like features as well as the characteristic, unique among salivary cell lines, of sufficient polarization and transcellular resistance to be useful in Ussing chambers. It is hoped that this cell line will prove as beneficial to salivary gland research as similar cell lines have been in studies of other organ systems with important epithelial components.