The renal and intestinal Na+/HCO3- cotransporters are differentially regulated by cAMP-dependent and cholinergic agonists [1, 4–6]. Possible explanations for these findings are differences in their primary structure influencing regulatory properties, or features of the cell type in which they are expressed. In this study, we investigated the secretagogue regulation of the intestinal-pancreatic NBC subtype pNBC1 in HEK293 cells.
The group of Ira Kurtz has thoroughly studied the stoichiometry of NBC1 and found that the renal subtype kNBC1 is transporting 1 Na+ and 3 HCO3- with each cycle, which results in outward-directed transport . Functional inhibition of kNBC1 by cAMP is probably in part due to a stoichiometry shift from 3:1 to 2:1, leading to a switch from export to import. On the other hand, the intestinal subtype pNBC1 serves the uptake of both ions with a 2:1 ratio under physiological conditions , and a stoichiometry change was not reported for the endogenously expressed transporter. However, pNBC1 can be activated via cAMP , which is due to the phosphorylation of a unique consensus phosphorylation site in its N-terminus . As to a structure-function relationship regarding cholinergic stimulation, which increases the transport rate of both subtypes in a PKC-dependent manner [5, 6], the possible relevance of the unique PKC-dependent phosphorylation site in the pNBC1 N-terminus remains to be assessed.
Importantly, cell type- and tissue-specific regulation has been recognized for many ion transporters including Na+/H+-exchange [10, 11], anion exchange , and CFTR . Expression of pNBC1 and kNBC1 is not restricted to the pancreas/intestinal tract and the kidney, respectively, and expression in other organs such as the eye , the gallbladder , and salivary glands  has been reported, but it has not been investigated whether these different cell types modulate transporter regulation. Gross et al. have previously studied the regulation of murine pNBC1 endogenously occurring in pancreatic duct cells and heterologously transfected into a mouse proximal tubule cell line . They found that cAMP causes upregulation of endogenous pNBC1 via one PKA-dependent phosphorylation site not present in the kNBC1 sequence, while changing the stoichiometry from 3:1 to 2:1 and thereby transport direction via another common C-terminal site (Ser982 and Ser1026 for kNBC1 and pNBC1, respectively), the latter resembling kNBC1 regulation. In our experiments, however, we sought to clarify the cell-type dependency of the regulation of the presumed physiological function of pNBC1, which is HCO3- uptake during stimulated anion secretion, and which can be activated by imposing an acid load on the cytosol. In this setting of inward transport, stoichiometry would be expected to be already 2:1 , and there is no evidence that secretagogues can reverse it to 3:1 in either of the NBC1 subtypes. Accordingly, Pedrosa et al. did not find any influence of cAMP stimulation on kNBC1 in renal cells in pHi recovery experiments ; however, it is under debate whether kNBC1 can be a relevant base loader under physiological conditions, a recent study suggesting rather an adaptation of the pNBC1 and kNBC1 expression pattern in renal and submandibular gland epithelia during acid-base disturbances . The differential regulation of SLC4 gene family members in different cell types might be related to possibly specialized roles of the transporters in the respective tissues (e. g. vectorial transport vs. homeostatic functions), but this aspect of NBC function is yet poorly understood.
With respect to the HEK293 cells used in the current study, transfection studies investigating human heart (hhNBC) or renal NBC did not reveal any endogenous expression of the these isoforms by immunoblotting [24, 25]. In their functional experiments, one group reported Na+- and HCO3--dependent pHi recovery after an acid load, which could be due to NHE or NBC , while another group described no recovery in the additional presence of amiloride, arguing against the presence of endogenous NBC . In support of these findings, the recovery rates we measured in untransfected cells in the presence of CO2/HCO3- were very low, and, although not statistically comparable since measured in a different experiment, not higher than in its absence. Therefore, the transfection-independent recovery rate is most likely due to endogenous Na+/H+ exchangers, of which the NHE1 and NHE3 isoforms have previously been detected in HEK293 cells [15, 16]. Since recovery rates were considerably lower in untransfected vs. transfected cells in the presence of CO2/HCO3- (Figure 4B), no attempt was made to pharmacologically inhibit the presumably relatively low endogenous transporter activities in the subsequent experiments. One limitation of the present study is therefore the lack of information on potential effects of carbachol and forskolin on endogenous base loading mechanisms.
The transfection efficiency was relatively low under all conditions we tried, e. g. serum free/serum containing media, lipophilic agents, or calcium precipitation. Furthermore, a suitable transfection system to serve as a control is not available, since the intestinal cell lines Caco-2 and HT-29 show endogenous NBC activities which are not well molecularly characterized [27–29], T84 cells express kNBC1 rather than pNBC1 (B. Bridges, communication), and our attempts to transfect pNBC1 into Caco-2 or T84 cells did not yield sufficient overexpression to achieve unequivocal results. One possible reason for the overall low transfection efficiency could be that NBC increases intracellular pHi above the optimum for cellular metabolism. However, the pHi changes we measured were minor, and the acidification of cytosolic pH by lowering medium pH did not improve NBC abundance (data not shown). More likely, transfection efficiency is limited by the large size of the GFP-pNBC1 transcript. To differentiate between transfected and untransfected cells in this setting, the fluorescence signal of the GFP tag was visualized using confocal microscopy. Since the C-terminal phosphorylation site was characterized as a crucial regulatory sequence which explains the regulation of kNBC1 and pNBC1 in renal cells [7, 8], it appeared reasonable to tag the pNBC1 N-terminus to avoid interference of the tag with this sequence. Nevertheless, we observed an inhibition of NBC after preincubation with forskolin, which cannot be readily explained by the N-terminal GFP tag. However, we cannot exclude that the GFP tag interferes with the binding site for IRBIT close to the N-terminus (see below), which could in part explain our results with carbachol.
The molecular basis for the differential regulation of pNBC in intestinal vs. HEK293 cells remains unknown. The observed acid-induced NBC activity in HEK293 cells under basal conditions and the strong inhibitory effect of forskolin are difficult to reconcile with mere stoichiometry changes as described by Gross et al. , since they would further increase HCO3- import and thus accelerate pHi recovery. Possibly, the signal transduction machinery of the cell types provides an explanation. One could speculate about a possible role of associated proteins, such as NHERF (NHE regulatory factor) family members, IRBIT (IP3R binding protein released with IP3), and carbonic anhydrase. NHERF has been shown to be necessary for cAMP-dependent inhibition of renal NBC; however, no binding, NBC phosphorylation or change in NBC surface expression were observed [30, 31]. As another regulatory protein, IRBIT was found to bind to pNBC1, but not kNBC1, and pNBC1 transfected into X. oocytes only attains functional activity comparable to the one of transfected kNBC1 when IRBIT is co-transfected . IRBIT is expressed in the kidney , while its presence in HEK293 cells and the intestinal tract has not yet been clarified. Furthermore, carbonic anhydrase, of which several isoforms are differentially expressed in the kidney and the gastrointestinal tract , could potentially influence NBC transport capacity by HCO3- generation, but its functional relevance for NBC regulation is highly controversial at present [35–37].