International Symposium on "HCO₃^- AND CYSTIC FIBROSIS". San Diego, CA (USA). March 3-5, 2001.

Article in PDF format  -  JOP Home page

Yoshiro Sohma¹, Michael A Gray², Yusuke Imai¹, Barry E Argent²

¹Department of Physiology, Osaka Medical College. Takatsuki, Osaka, Japan. ²Department of Physiological Sciences, University Medical School. Newcastle upon Tyne, United Kingdom

Cystic fibrosis (CF) takes its name from the pathological changes that occur in the pancreas. Cystic fibrosis transmembrane conductance regulator (CFTR) is highly expressed in the pancreatic ductal epithelium and plays a key role in ductal HCO₃^- secretion. In humans, the pancreatic duct secretes near isotonic NaHCO_3. Experimental data suggests that HCO₃^- secretion occurs via apical Cl^-/HCO₃^- exchangers working in parallel with Cl^- channels (CFTR and calcium activated chloride channels, CaCC). Programming the currently available experimental data into our computer model [1, 2] (based on network thermodynamics) shows that while the anion exchanger/Cl^- channel mechanism will produce a relatively large volume of a HCO₃^--rich fluid, it can only raise the luminal HCO₃^- concentration up to about 70 mM. To achieve secretion of about 150 mM NaHCO₃ it is necessary to modulate the properties of the apical membrane transporters as the secreted fluid flows down the ductal system. On the basis of our computer simulations, we propose that HCO₃^- secretion occurs mainly via the exchanger in duct segments near the acini (luminal HCO₃^- concentration up to about 70 mM), but mainly via channels further down the ductal tree (raising luminal HCO₃^- to about 150 mM). We speculate that the switch between these two secretory mechanisms is controlled by a series of luminal signals (e.g. pH, HCO₃^- concentration) acting on the apical membrane transporters in the duct cell.

The pancreatic duct cell is a particularly interesting system in which to study HCO₃^- transport because it secretes HCO₃^- at high concentration. The maximum HCO₃^- concentration depends on the species, and varies between about 70 mM in the rat and 145 mM in cat, dog and humans [3, 4]. On the basis of studies that have been performed largely (but not exclusively) on ducts isolated from the rat and guinea pig pancreas the HCO₃^- secretory model shown in Figure 1 has been proposed [3, 4].

The key problem with this model is how the anion exchanger/Cl^- channel mechanism could secrete near isotonic NaHCO₃. Under these conditions the concentration of Cl^- in the luminal fluid will be very low (about 10 mM). This means that given reasonable values for intracellular [HCO₃^-], the intracellular Cl^- concentration would have to be very low indeed (less than 1 mM) to drive HCO₃^- secretion on an apical Cl^-/ HCO₃^- exchanger. This intuitive assessment is confirmed by simulation data from our computer model, summarised in Figure 2.

The computer simulation shows that the duct cell will secrete a fluid containing about 110 mM HCO₃^-, but only if the luminal HCO₃^- concentration is low (about 25 mM), as might be expected in the most proximal ducts just downstream from the acini (Figure 2). As the concentration of HCO₃^- in the duct lumen rises, the volume and HCO₃^- concentration of the secreted fluid decreases so that the maximal attainable HCO₃^- concentration in the luminal fluid is about 70 mM (Figure 2). The inability of this scheme to establish a luminal HCO₃^- concentration greater than 70 mM results from: i) a slowing of HCO₃^- efflux on the exchanger as luminal HCO₃^- concentration increases and luminal Cl^- concentration decreases (reversal of the exchanger will eventually occur), and ii) reduced HCO₃^- efflux via the conductive pathway as luminal HCO₃^- increases.

Thus while the cellular model shown in Figure 1 might describe how the rat pancreas works (maximum juice [HCO₃^-] about 70 mM), it cannot explain how near isotonic HCO₃^- is secreted by species such as the guinea-pig, cat, dog and humans. It is possible that these species use a different HCO₃^- secretory mechanism at the apical membrane. However, in our view the evidence for this is not strong. Indeed, far from identifying differences between the rat and guinea pig duct cells, recent publications have emphasised the similarities in terms of the key transport elements that are involved in HCO₃^- secretion (see Figure 1).

Using our computer model we have investigated how the transport parameters on the duct cell must be altered for the scheme shown in Figure 1 to secrete near isotonic NaHCO₃. Based on the computer simulations, we propose a new hypothesis for pancreatic ductal HCO₃^- secretion. The hypothesis is summarised in Figure 3. We view the starting point of the HCO₃^- secretory process as the secretion (by the acini) of a small volume of a plasma-like fluid (containing 25 mM HCO₃^-) into the top of the ductal system. Programming the currently available experimental data into the computer model shows that the cell depicted in Figure 3 (top right) will produce a relatively large volume of secretion containing about 110 mM HCO₃^-. We call this cell the proximal cell, predict that it exists in the upper part of the ductal system close to the acini, and note that it represents one end of a functional spectrum within the ductal tree.

As the fluid flows along the first part of the ductal system, the luminal HCO₃^- concentration will rise and the net secretory flux of HCO₃^- from the proximal cell will fall (Figure 3). We propose that as this occurs, the activity of membrane transporters in the proximal cell is progressively modified to those of a cell that can maintain a small net secretory flux of HCO₃^- in the face of near isotonic NaHCO₃ in the lumen (Figure 3). We call this cell the distal cell and note that it represents the other end of the functional spectrum that exists in the ducts. Essentially, the high HCO₃^- secretion produced by the distal cells would dilute out the chloride in the secretions produced by acini and proximal cells in the upper regions of the ductal tree. Compared to proximal cells, the major changes in the transport parameters that allow the distal cells to secrete in the face of a high luminal HCO₃^- concentration are: i) a reduced luminal Cl^- permeability, ii) a reduced activity of the luminal Cl^-/HCO₃^- antiporter, and iii) activation of a basolateral volume-sensitive Cl^- conductance to offset cell swelling due to intracellular Cl^- accumulation. Thus HCO₃^- secretion would occur mainly by the exchangers in duct segments near the acini (proximal cells), but mainly via the channels further down the ductal tree (distal cells).

As yet, there is no experimental evidence for high HCO₃^- concentration or pH inhibiting the anion exchanger on the apical membrane of the duct cell. However, some data does exist supporting other predictions made by the model. For example, we have shown that volume-activated Cl^- channels are present in the duct cell [5], and that an increase in extracellular [HCO₃^-] reduces the CFTR Cl^- conductance of guinea-pig duct cells [6]. The mechanism by which extracellular HCO₃^- inhibits CFTR is unknown. Both the Cl^- and HCO₃^- conductances on the luminal membrane probably reside within the same anion channels (i.e. CFTR and Ca²⁺-activated Cl^- channels). This means that signals from the duct lumen would have to change the characteristics of ion permeation through these channels. This might come about from multi-ion characteristics. In this respect it is interesting to note that CFTR is known to be a multi-ion channel [7]. However, we should be clear that decreasing apical membrane Cl^- permeability does not imply a change in the anion selectivity of the duct cell luminal membrane. The necessity for decreasing apical membrane Cl^- permeability comes only from the requirement for a lower Cl^- as compared to HCO₃^- efflux across the luminal membrane at negative membrane potentials (about -50 mV), and not from a requirement to increase the anion selectivity of the apical membrane to HCO₃^- over Cl^-. Indeed, our patch-clamp study showed that high extracellular [HCO₃^-] did not change the Cl^-:HCO₃^- selectivity ratio of CFTR channels in guinea-pig pancreatic duct cells, although Cl^- efflux at negative membrane potentials was inhibited markedly [6]. It is, therefore, important to measure the ratio of Cl^- to HCO₃^- effluxes across the luminal membrane of the duct cell under defined experimental conditions and to compare those data with values predicted by the model.

Article in PDF format
Look up who cited this article

Key words Cystic Fibrosis; Models, Theoretical; Pancreatic Ducts; Sodium Bicarbonate

Abbreviations CaCC: calcium activated chloride channels; CF: cystic fibrosis; CFTR: cystic fibrosis transmembrane conductance regulator

Acknowledgements Funded by grants from the Cystic Fibrosis Trust and The Medical Research Council.

Correspondence
Barry E. Argent
Department of Physiological Sciences
University Medical School
Framlington Place
Newcastle upon Tyne, NE2 4HH
United Kingdom
Phone: +44-191-222.7009
Fax: +44-191-222.6706
E-mail address: b.e.argent@ncl.ac.uk