As a graduate student, I imposed acute intracellular acid loads and found that the sudden pH_i decrease is followed by a recovery—the first demonstration of pH_i regulation. Later work showed that this recovery is mediated by a Na⁺-driven Cl-HCO₃ exchanger (NDCBE)—the first documented pH_i regulator. As a fellow, we identified the electrogenic Na/HCO₃ cotransporter (NBCe1), which not only regulates pH_i but also mediates HCO₃⁻ movement of across many epithelia.

Since cloning the cDNA encoding NBCe1, our group has studied the molecular mechanism of Na⁺-coupled HCO₃⁻ transporters (NCBTs)—members of the SLC4 family of transport proteins. For example, emerging data indicate that NBCe1 and NDCBE—expressed in Xenopus oocytes—transport CO₃⁼ rather than HCO₃⁻. Carbonic anhydrases do not change transport rate but stabilize surface pH. Chimera making shows that extracellular loop #4 is critical for determining electroneutrality vs electrogenicity. Certain splice variants have autoinhibitory or autostimulatory domains. The large cytoplasmic N terminus of NBCe1 binds to and thereby activates the transmembrane domain. Finally, we are producing large quantities of NBC-related proteins to study their molecular biophysical properties.

While studying isolated, perfused proximal tubules (PTs), we noticed that adding CO₂/HCO₃⁻ to the PT's basolateral (BL or blood) side causes pH_i to increase, reflecting stimulated H⁺ extrusion across the apical membrane (facing lumen). To elucidate the mechanism, we invented out-of-equilibrium (OOE) CO₂/HCO₃⁻ solutions, which allow us to vary [CO₂]_BL, [HCO₃⁻]_BL, and pH_BL—one at a time. We found that isolated increaes in [CO₂]_BL or isolated decreases in [HCO₃⁻]_BL stimulate H⁺ secretion into the tubule lumen (J_H) . Surprisingly, acute changes in pH_BL have no effect on J_H! Thus, the tubule regulates blood pH by sensing not pH but the two major blood buffers. 

The signal-transduction system linking basolateral CO₂ and HCO₃⁻ to apical H⁺ extrusion involves: (1) Apical AT₁ (angiotensin II, ANG II) receptors and ANG II that the tubule secretes into the lumen. (2) Tyrosine phosphorylation. Inhibitors of ErbB receptor tyrosine kinases block responses to basolateral CO₂ or HCO₃⁻. The same is true for knockout of receptor protein tyrosine phosphatase γ (RPTPγ), in which the extracellular ligand-binding region is an inactive carbonic-anhydrase-like domain (CALD). We are testing the hypothesis that the CALD is the sensor for extracellular CO₂ and HCO₃⁻. We are also using biochemical approaches to explore signal transduction.

While studying pH_i regulation in gastric parietal cells we found that the apical membrane has no detectable permeability to either CO₂ or NH₃—the first documented gas-impermeable membrane. This led us to rethink “Overton’s rule” and to describe the first gas channel aquaporin 1 (AQP1), which is not only permeable to H₂O but also CO₂. Others later showed that AQP1 is permeable to NH₃, and that the rhesus (Rh) proteins are permeable to CO₂ and NH₃.

We developed an approach for using a large pH microelectrode to monitor the extracellular surface pH (pH_S) on Xenopus oocytes exposed to CO₂ or NH₃. Initial pH_S transients are indices of CO₂ or NH₃ permeability. Studying a wide range of AQPs and Rh proteins, we see that each has a characteristic CO₂/NH₃ permeability ratio—the first example of gas selectivity. Preliminary work suggests that NH₃ moves only through AQP or Rh monomers, whereas CO₂ generally prefers the central pore of the oligomer. Preliminary stopped-flow work with erythrocytes (RBCs) indicates that certain inhibitor combinations can block 75% of O₂ efflux from RBCs—the first evidence for O₂ channels.