Enterocyte K⁺ ion permeability and fluid secretion: missing the correct channel or missing the point?

Julio‐Kalajzic et al. (2018) identify potassium ion channels in the enterocyte that might facilitate fluid secretion. Whilst applauding the undoubted technical expertise in cell biology and genomics that was brought to bear, I find myself at odds with all the conclusions drawn regarding intestinal secretion.

Secretion is fluid movement from behind the enterocyte into the intestinal lumen, supposedly generated by processes within the enterocyte. Chloride ion secretion is the driving force for fluid secretion, with maintenance of membrane potential difference through potassium ion channels conjectured to be essential for secretion. Blockage of the appropriate potassium ion channel could therefore prevent secretory diarrhoeal disease.

Mass transport is the only way secretion can be correctly assessed. Deduction of secretion using Fick's principle with non‐absorbed markers will be false if absorption causes marker loss, as occurs for all markers currently in use. Deductions from ‘unidirectional’ fluxes of chloride isotopes are also false as the correct solution to the mathematics of the Ussing chamber (Lucas, 2005) disallows partition into unidirectional fluxes, since fluxes are reciprocally linked. In vivo experiments that include interstitial volume (Eklund et al. 1985) mislead since increased extracellular volume will resemble but not actually be luminal secretion. Finally, short‐circuit current is not chloride ion, and hence fluid, secretion, although this technique is now the mainstay of investigations into secretion. Its inherent ambiguity is that an increase in current could be an outward chloride ion or an inward sodium ion current. Changes in short‐circuit current show only that normal tissue responds to challenge by enterotoxins or pharmacological agents.

When confronted with evidence from in vivo perfusion experiments, deductions deriving from short‐circuit‐current experiments, however closely and logically argued, fail to be validated. Short‐circuit current changes are coincidental phenomena resulting from interrupted fluid absorption clearly seen when E. coli STa enterotoxin is the perturbing agent. STa causes a large increase in short‐circuit current (Young et al. 1988) in rat jejunum and ileum but does not cause intestinal fluid secretion (Lucas et al. 2005) in vivo. STa toxin will elevate short‐circuit current in isolated cells (Rufo et al. 1997) that is reduced by glibenclamide and clotrimazole. When tested in vivo, these compounds do not reverse STa‐mediated fluid absorption inhibition (Lucas et al. 2011). It is possible to dismiss findings of a mismatch between inhibition of short‐circuit current and lack of effect in vivo since they stem from one laboratory, but while STa does not cause secretion, adverse osmotic forces, E. coli labile toxin and C. difficile toxin A all show net secretion when measured in the same laboratory. The maximum permissible inference from the E. coli experiments is not that there is never fluid secretion; it is that if secretion occurs as with other toxins, it is not explained by fluid secretion directly from the enterocyte and cannot be measured by short‐circuit current measurements.

Is there a similar mismatch between current measurements in vitro and fluid movement in vivo in the paper by Julio‐Kalajzic et al. (2018)? When this concept was tested by examining secretion after exposure to cholera toxin in vivo, the potassium channel mutants failed to show the anticipated reductions in secretion, confirming similar negative findings (Preston et al. 2010) with other mutants. The channel mutant values for cholera‐induced secretion are uncorrelated with the size of secretion that is expected from changes in short‐circuit currents often exceeding 100 μA cm⁻². If secretion is isotonic, Faraday's law indicates that 50 μA cm⁻² equals about 7 μl (cm length)⁻¹ hr⁻¹ of secretion. The colonic short circuit current indicated that a weight to length ratio of 0.125 should arise after 5 hr, with cholera stimulated ileal loops having a ratio of 0.25 gm cm⁻¹, twice what was found. It seems unlikely that short‐circuit current and fluid movement are directly related, but the contemporaneous occurrence of both arises from altered fluid uptake.

The authors conclude that the correct potassium ion channel eludes identification. Yet, some enterotoxins elevate short‐circuit current, fail to instigate fluid secretion and their actions in vivo are unaffected by potassium channel blockers. This is now the pattern emerging with potassium channel mutant studies since short‐circuit current can alter but there is no evidence of fluid secretion being induced or accelerated; indeed, the discussed paper supports the view that aspects of the enterocyte secretion hypothesis fail when tested in vivo. The interesting paper by Julio‐Kalajzic and others inadvertently provides further evidence that the enterocyte secretion model is defective and that secretion is caused by a combination of vasodilatation and increased hydraulic permeability, supplemented by interruption of fluid absorption.

Additional information

Competing interests

None declared.