K+-Cl- cotransport: 'to be or not to be' oxygen sensitive.

Abstract

Under 20 years ago, K+-Cl− cotransport (Cot) was discovered in red blood cells after hyposmotic cell swelling (Dunham & Ellory, 1981) and cellular thiol modification (Lauf & Theg, 1980), and later shown to be electroneutral (Kaji, 1993; Lauf & Adragna, 1996). Early hypotheses of membrane mechanical stress versus chemical processes (Lauf, 1985) addressed the activation mechanism. Chemically stimulated K+-Cl− Cot is ATP dependent (Lauf, 1985), and volume-stimulated K+-Cl− Cot, exhibiting a lag phase, is inhibited by serine/threonine (ST) protein phosphatase (PP) inhibitors okadaic acid and calyculin. A two state equilibrium model (Jennings & Al-Rohil, 1990) proposed a phosphorylated (A) resting- and a dephosphorylated (B) active-transporter controlled by ST kinases (STPK) and STPP. Low cellular Mg2+ ([Mg2+]i) activates K+-Cl− Cot (Lauf 1985), abolishes a characteristic bell-shaped ‘acid’ flux activation (Brugnara et al. 1989) and reveals an internal pH sensor (Lauf & Adragna, 1998). Two isoforms of K+-Cl− Cot, KCC1 and KCC2, have been expression cloned (Gillen et al. 1996). In sickle cell disease, elevated K+-Cl− Cot is suspected to cause the ‘fast track dehydration’ of reticulocytes homozygous for haemoglobin (HbS; Bookchin et al. 1991). HbS cells have a shorter life span with an age distribution skewed toward younger cells. Hence, elevated K+-Cl− Cot can be considered as an epiphenomenon typical for young cells (Joiner, 1993). K+-Cl− Cot activation by O2 in trout red cells suggested a role in regulation (Nielsen et al. 1992). Deoxygenation-induced inhibition of K+-Cl− Cot in normal red blood cells (HbA) and HbS cells was ascribed to a rise in [Mg2+]i when deoxy-Hb binds 2,3 diphosphoglycerate (Canessa et al. 1987). Gibson et al. (1998) now show in this issue of The Journal of Physiology that, after isotonic equilibration, O2 stimulates K+-Cl− Cot in hyposmotically swollen and acidified or urea-treated normal human red blood cells by almost an order of magnitude, with a PO2 for half-maximal activation of K+ influx (P50) higher than that for Hb oxygenation. This finding, per se, does not exclude an association between K+-Cl− Cot and Hb. The data suggest the presence of an ‘O2 sensor’ which is de-occluded by the chosen interventions. Through this sensor, O2 modifies the swelling/pH- or urea-induced K+-Cl− Cot activity, by either accelerating STPP-mediated dephosphorylation (through O2 derivatives?) or inhibiting STPK like N-ethylmaleimide (Lauf, 1985). The nature of the putative sensor is unknown, and the mechanism could involve enzymes, according to the two state equilibrium model: In swollen or urea-exposed HbS cells, O2 pre-equilibrated in isotonic media, K+-Cl− Cot is fifty times larger than in normal cells in the presence or absence of 150 mmHg O2, and thus can be considered to be already an ‘O2-activated’ B conformer that, however, fails to return to the A state upon deoxygenation. With the caveat of the absence of [Mg2+]i data, the conclusion is that the putative O2 sensor is absent or modified so that K+-Cl− Cot is refractory to O2 removal/addition. Perhaps repeated oxygenation-deoxygenation cycles cause thiol oxidation and hence inactivation of crucial protein kinases (Flatman et al. 1996). Alternatively, the sensor/enzyme is absent as in Src tyrosine kinase-knockout mice with elevated red blood cell K+-Cl− Cot not activated by thiol reagents (De Franceschi et al. 1997). These findings are at variance with reports failing to show O2-induced activation in normal and HbS red blood cells (Canessa et al. 1987). The explanation must lie in the different experimentation and thus may not require ‘re-evaluation of … findings’ (Gibson et al. 1998) but, rather, new approaches. Recently, Joiner et al. (1998) showed that in isosmotically suspended and [Mg2+]i-clamped HbS cells, subsequently N2- or O2-equilibrated, K+-Cl− Cot is 2.5-fold activated by deoxygenation but not by oxygenation. This effect, smaller than in Gibson's study, suggests that O2 is an inhibitor of K+-Cl− Cot. Conceivably, the cells’ O2 sensor was not modified by either cell swelling, acid pH or urea. Thus, in volume-, acid pH- and urea-stressed normal and HbS red cells there is O2 activation and refractoriness, respectively, whereas in isotonic [Mg2+]i-clamped HbS cells there is O2 inhibition of K+-Cl− Cot. However concealed, it appears that indeed there is an O2 sensor. The authors’ finding, and the conclusion of the existence of a putative O2 sensor, are of potential consequence for sickle cell pathophysiology. Commitment to ‘volume suicide’ through KCl and obligatory water efflux in HbS cells may occur in the kidney where low pH and high urea favour K+-Cl− Cot activation not ameliorated, as in normal red cells, by the organ's low PO2 (5-15 mmHg). Fast track dense cells conceivably contribute to the pathogenesis of papillary necrosis, a common pathological feature in HbS disease.