The ionic currents of carotid body type I cells and their possible involvement in the detection of oxygen tension (Po2) in arterial blood are unknown. The electrical properties of these cells were studied with the whole-cell patch clamp technique, and the hypothesis that ionic conductances can be altered by changes in PO2 was tested. The results show that type I cells have voltage-dependent sodium, calcium, and potassium channels. Sodium and calcium currents were unaffected by a decrease in PO2 from 150 to 10 millimeters of mercury, whereas, with the same experimental protocol, potassium currents were reversibly reduced by 25 to 50 percent. The effect of hypoxia was independent of internal adenosine triphosphate and calcium. Thus, ionic conductances, and particularly the O2-sensitive potassium current, play a key role in the transduction mechanism of arterial chemoreceptors.

We desired 02 concentrations. Po2 in the chamber was monitored with an O2-sensing electrode (9). Macroscopic calcium currents were studied using the whole-cell configuration of the patch-clamp technique (3,13) and recorded in isolation after blockade of the voltage-dependent Na+ and K+ channels. Cytosolic [Ca2W] was estimated in unclamped cells loaded with fura-2 by incubation for 10 min at 37°C with saline containing 1 AM fura-2 acetoxymethyl ester. Experiments were performed on an inverted microscope with standard optical components and equipped for epifluorescence and photometry (14). For the two excitation wavelengths, we used the filters shortwave-pass SWP 357 (excitation at =360 nm) and band-pass BP 380 (excitation at 380 nm; bandwidth, 10 nm). Fluorescence from the cells was measured by a dual wavelength photometer. The two output voltage signals from the photometer were digitized and displayed on-line on the screen of a computer in parallel with the estimated [Ca2+] concentration (15). Calibration of the fluorescence signals in terms of [Ca2+] was performed in vitro as described (16). Secretion was monitored in amperometric mode with a glass-sealed 8-,um diameter carbon electrode fabricated as described (17)(18)(19)(20). In most experiments, we used an amplifier built in our laboratory that has in the headstage a Burr-Brown OPA 111 wired as a current-to-voltage converter with a feedback resistor of 500 MQ1. The high-resolution recordings (see below) were obtained with a standard List EPC-7 patch-clamp amplifier. The single events of high resolution, similar to those shown in Fig. 2C, were recorded with a carbon electrode covered with polyethylene, which, as shown before (19), decreases noise and permits the acquisition of signals at a broader bandwidth. We held the carbon fiber at a constant voltage of +950 mV, a potential more positive than the oxidation potential of dopamine, thus assuring the oxidation of dopamine released by the cells (see below). Cyclic voltammograms were obtained by applying voltage ramps from -600 to + 1000 mV (at a rate of 170 V/s) to the carbon-fiber electrode. The signals are characterized by typical reduction (at approximately -400 mV) and oxi-*To whom reprint requests should be addressed. 10208The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

We have monitored cytosolic [Ca 2+] and dopamine release in intact fura-2-loaded glomus cells with microfluorimetry and a polarized carbon fiber electrode. Exposure to low PO 2 produced a rise of cytosolic [Ca 2+] with two distinguishable phases: an initial period (with PO 2 values between 150 and ~70 mm Hg) during which the increase of [Ca 2+] is very small and never exceeds 150-200 nM, and a second phase (with PO 2 below ~70 mm Hg) characterized by a sharp rise of cytosolic [Ca2+]. Secretion occurs once cytosolic [Ca 2+] reaches a threshold value of 180 + 43 nM. The results demonstrate a characteristic relationship between Po 2 and transmitter secretion at the cellular level that is comparable with the relation described for the input (02 tension)-output (afferent neural discharges) variables in the carotid body. Thus, the properties of single glomus cells can explain the sensory functions of the entire organ. In whole-cell, patch-clamped cells, we have found that in addition to O2-sensitive K + channels, there are Ca 2+ channels whose activity is also regulated by PO 2. Ca z+ channel activity is inhibited by hypoxia, although in a strongly voltage-dependent manner. The average hypoxic inhibition of the calcium current is 30% -+ 10% at -20 mVbut only 2% + 2% at +30 mV. The differential inhibition of K + and Ca 2+ channels by hypoxia helps to explain why the secretory response of the cells is displaced toward Po 2 values (below ~70 mm Hg) within the range of those normally existing in arterial blood. These data provide a conceptual framework for understanding the cellular mechanisms of 02 chemotransduction in the carotid body.

The hypothesis that changes in environmental 02 tension (pOi) could affect the ionic conductances of dissociated type I cells of the carotid body was tested. Cells were subjected to whole-cell patch clamp and ionic currents were recorded in a control solution with normal pO 2 (pO~ = 150 mmHg) and 3-5 min after exposure to the same solution with a lower pO,. Na and Ca currents were unaffected by lowering pO, to 10 mmHg, however, in all cells studied (n = 42) exposure to hypoxia produced a reversible reduction of the K current. In 14 cells exposed to a pO 2 of 10 mmHg peak K current amplitude decreased to 35 +_ 8% of the control value. The effect of low pO2 was independent of the internal Ca 2+ concentration and was observed in the absence of internal exogenous nucleotides. Inhibition of K channel activity by hypoxia is a graded phenomenon and in the range between 70 and 120 mmHg, which includes normal pO, values in arterial blood, it is directly correlated with pO 2 levels. Low pO2 appeared to slow down the activation time course of the K current but deactivation kinetics seemed to be unaltered. Type I cells subjected to current clamp generate large Na-and Cadependent action potentials repetitively. Exposure to low pO~ produces a 4-10 mV increase in the action potential amplitude and a faster depolarization rate of pacemaker potentials, which leads to an increase in the firing frequency. Repolarization rate of individual action potentials is, however, unaffected, or slightly increased. The selective inhibition of K channel activity by low pO, is a phenomenon without precedents in the literature that explains the chemoreceptive properties of type I cells. The nature of the interaction of molecular O, with the K channel protein is unknown, however, it is argued that a hemoglobin-like O, sensor, perhaps coupled to a G protein, could be involved.

Contraction of vascular smooth muscle cells (VSMCs) depends on the rise of cytosolic [Ca 2+ ] owing to either Ca 2+ in¯ux through voltage-gated Ca 2+ channels of the plasmalemma or receptor-mediated Ca 2+ release from the sarcoplasmic reticulum (SR). We show that voltage-gated Ca 2+ channels in arterial myocytes mediate fast Ca 2+ release from the SR and contraction without the need of Ca 2+ in¯ux. After sensing membrane depolarization, Ca 2+ channels activate G proteins and the phospholipase C±inositol 1,4,5-trisphosphate (InsP 3 ) pathway. Ca 2+ released through InsP 3 -dependent channels of the SR activates ryanodine receptors to amplify the cytosolic Ca 2+ signal. These observations demonstrate a new mechanism of signaling SR Ca 2+ -release channels and reveal an unexpected function of voltage-gated Ca 2+ channels in arterial myocytes. Our ®ndings may have therapeutic implications as the calcium-channel-induced Ca 2+ release from the SR can be suppressed by Ca 2+ -channel antagonists.