A molecular mechanism for aberrant CFTR-dependent HCO(3)(-) transport in cystic fibrosis

Abstract

Aberrant HCO(3)(-) transport is a hallmark of cystic fibrosis (CF) and is associated with aberrant Cl(-)-dependent HCO(3)(-) transport by the cystic fibrosis transmembrane conductance regulator (CFTR). We show here that HCO(3)(-) current by CFTR cannot account for CFTR-activated HCO(3)(-) transport and that CFTR does not activate AE1-AE4. In contrast, CFTR markedly activates Cl(-) and OH(-)/HCO(3)(-) transport by members of the SLC26 family DRA, SLC26A6 and pendrin. Most notably, the SLC26s are electrogenic transporters with isoform-specific stoichiometries. DRA activity occurred at a Cl(-)/HCO(3)(-) ratio > or =2. SLC26A6 activity is voltage regulated and occurred at HCO(3)(-)/Cl(-) > or =2. The physiological significance of these findings is demonstrated by interaction of CFTR and DRA in the mouse pancreas and an altered activation of DRA by the R117H and G551D mutants of CFTR. These findings provide a molecular mechanism for epithelial HCO(3)(-) transport (one SLC26 transporter-electrogenic transport; two SLC26 transporters with opposite stoichiometry in the same membrane domain-electroneutral transport), the CF-associated aberrant HCO(3)(-) transport, and reveal a new function of CFTR with clinical implications for CF and congenital chloride diarrhea.

Fig. 1. Cl^– and HCO₃^– transport in CFTR-expressing HEK 293 cells. (A) pH_i measurement: CFTR-transfected cells were incubated in HCO₃^–-buffered solutions as indicated. After stabilization of pH_i, the cells were stimulated with forskolin and transiently exposed to a Cl^–-free medium. (B) [Cl^–]_i measurements: cells expressing CFTR and loaded with MQAE were perfused with HEPES-buffered solutions to measure Cl^–/NO₃^– exchange and then in HCO₃^–-buffered solutions to measure Cl^–/HCO₃^– exchange. For measurement of Cl^–/NO₃^– exchange, Cl^– was replaced with NO₃^–. (C) Flux ratio: to calculate the flux ratios, the rates of [Cl^–]_i changes due to incubating the same cells with NO₃^– (HEPES, first portion of trace in B) or Cl^–-free gluconate solution (HCO₃^–, second portion of trace in B) were calculated and used to obtain the HCO₃^–/NO₃^– ratio. The average is from nine similar experiments. The Cl^–/HCO₃^– ratio was calculated from measurement of pH_i (A) and accounting for the buffer capacity, and the measurement of Cl^– transport with MQAE, as in the second portion of (B). The number of pH_i and [Cl^–]_i experiments averaged is five and nine, respectively.

Fig. 2. Macroscopic Cl^– and HCO₃^– currents in Xenopus oocytes and HEK 293 cells expressing CFTR. (A) Oocytes were injected with water (controls) or cRNA coding for CFTR. Water-injected oocytes (n = 15) never showed forskolin-activated Cl^– current, and all cRNA injected oocytes (n = 10) showed such a current. Oocytes were stimulated with 10 µM forskolin to activate CFTR. After the current reached steady state, 94 mM Cl^– was replaced with gluconate and then 25 mM gluconate was replaced with 25 mM HCO₃^–. The duration of incubation with HCO₃^– is marked by the gray bar. The average of Cl^– and HCO₃^– currents from 10 experiments is given in the figure and the text. (B) Cl^– and HCO₃^– currents in HEK 293 cells. After establishing the whole-cell configuration, HEK 293 cells transfected with CFTR were perfused with a HEPES-buffered solution containing 150 mM NMDG-Cl. The cells were then stimulated with forskolin before (n = 6) or after (n = 5) exposure to 3.5 mM Cl^–. After the current stabilized, the cells were incubated with solutions buffered with 25 mM HCO₃^– while maintaining constant external [Cl^–]. After stabilization of the current, HCO₃^– was removed by perfusing with HEPES-buffered solutions. The currents mediated by CFTR recorded under each condition were averaged and are presented as the means ± SEM of the number of experiments listed in parentheses.

Fig. 3. CFTR does not activate the SLC4 exchangers. The figure shows only experiments performed with AE1; a similar lack of activation by CFTR of AE2, AE3 and AE4 was observed. HEK 293 cells were transfected with GFP (A; n >10), AE1 (B and C; n = 8), CFTR (D; n >10) or AE1 and CFTR (E; n = 6). Cl^–/OH^– and Cl^–/HCO₃^– exchange were measured by exposing the cells to Cl^–-free media in solutions buffered with HEPES or HCO₃^–, respectively. In (A), the lack of measurable Cl^–/OH^– and minimal Cl^–/HCO₃^– exchange activity in control cells is shown. It can be seen in (B) that AE1 has DIDS-inhibitable Cl^–/OH^– exchange activity and in (C) that AE1 activity is not stimulated by forskolin. In (D), it is shown that activated CFTR alone does not activate a large Cl^–/OH^– exchange, and in (E) that neither AE1-mediated Cl^–/OH^– nor Cl^–/HCO₃^– exchange are activated by CFTR.

Fig. 4. Electrogenic Cl^– and HCO₃^– transport by DRA. Control oocytes (A and B) or those expressing DRA (C–H) were used to measure MP (A and C), current (B and D), the rate of current changes (E and F) or I–V relationships (G and H). As indicated by the bars, the oocytes were incubated in Cl^–-free solutions in the absence (open bars) or presence (gray bar) of HCO₃^–. Where indicated in (B) and (D), the oocytes were exposed to 0.1 mM DIDS. Current was measured at a holding MP of –30 mV and stepping to ±60 mV every 10 s for 0.5 s. The I–V plots were recorded after stabilization of the MP following solution changes by holding the MP at 0 mV for 500 ms between steps and stepping from –100 to +80 for 250 ms at 10 mV intervals. The plots in (G) and (H) are typical of six to eight separate experiments.

Fig. 5. Electrogenic Cl^– and HCO₃^– transport by SLC26A6. Oocytes expressing SLC26A6 were used to measure MP (A and C), current (B and C) or I–V relationships (D–H). The I–V protocol (D) was the same as that in Figure 4. As indicated by the bars, the oocytes were incubated in Cl^–-free or Na⁺-free solutions in the absence or presence of HCO₃^–. Current was measured at a holding MP of –30 mV. (C) The effect of HCO₃^– on the rates of changes in MP (left) and current (right) in an expanded time scale. Small arrows next to traces in (G–I) point to tail currents.

Fig. 6. Activation of DRA by CFTR. Cells were transfected with 50 ng (A) or 0.5 µg of DRA (B), or 0.5 µg of DRA and 1.5 µg of CFTR (C and D). Cells were perfused with HEPES-buffered solutions to evaluate Cl^–/OH^– exchange and then with HCO₃^–-buffered solutions to measure Cl^–/HCO₃^– exchange. Cl^–/OH^– and Cl^–/HCO₃^– exchange were estimated from the changes in pH_i due to the removal and addition of Cl^–, as indicated by the bars in each experiment. Where indicated, the cells were also stimulated with forskolin. Note that stimulation of CFTR was obligatory for activation of DRA (C and D). The first portion of the experiments in (C) and (D) is illustrated in (E). (F) A summary of the results of 11 (DRA only) and 17 (DRA + CFTR) experiments. (G–J) The effect of CFTR on Cl^–/OH^– exchange by SLC26A6 and PDS, respectively. The average of four experiments under each condition is presented.

Fig. 7. Active DRA is essential for CFTR stimulated Cl^–/OH^– exchange. HEK 293 cells were transfected with wild-type or the indicated mutants of DRA and used for immunolocalization. (A) Control with cells transfected with GFP. Note that wild-type (B), L489R (C) and I668–669ins (D) were expressed in the plasma membrane, whereas ΔV310 (E) was retained in the endoplasmic reticulum and is used as a control. When expressed alone or together with CFTR, none of the mutants showed any Cl^–/OH^– exchange activity.

Fig. 8. Interaction between CFTR and mutants with DRA in vitro and in vivo. HEK 293 cells were transfected with low levels of DRA and the indicated CFTR constructs. Cells from each group were stimulated before measurement of Cl^–/OH^– exchange. All cells were transfected with 0.2 µg of DRA. In (A) and (B), the cells were transfected with 0.5 µg of CFTR, in (C) with 0.5 µg of CFTR(R117H) and in (D) with 0.5 µg of CFTR(G551D). (E) Summary of the results of four to seven experiments under each condition. In (F), pancreatic extract was used to immunoprecipitate DRA and the immunoprecipitate was blotted for CFTR.