Reactive oxygen species (ROS) mainly originating from NADPH oxidases have been shown to be involved in the carotid body (CB) oxygen-sensing cascade. For measuring ROS kinetics, type I cells of the mouse CB in an ex vivo preparation were transfected with the ROS sensor construct FRET-HSP33. After 2 days of tissue culture, type I cells expressed FRET-HSP33 as shown by immunohistochemistry. In one population of CBs, 5 min of hypoxia induced a significant and reversible decrease of type I cell ROS levels (n = 9 CBs; P < 0.015), which could be inhibited by 4-(2-aminoethyl)benzensulfonylfluorid (AEBSF), a highly specific inhibitor of the NADPH oxidase subunits p47(phox) and p67(phox). In another population of CBs, however, 5 min of hypoxia induced a significant and reversible increase of ROS levels in type I cells (n = 8 CBs; P < 0.05), which was slightly enhanced by administration of 3 mM AEBSF. These different ROS kinetics seemed to coincide with different mice breeding conditions. Type I cells of both populations showed a typical hypoxia-induced membrane potential (MP) depolarization, which could be inhibited by 3 mM AEBSF. ROS and MP closely followed the hypoxic decrease in CB tissue oxygen as measured with an O2-sensitive dye. We conclude that attenuated p47(phox) subunit activity of the NADPH oxidase under hypoxia is the physiological trigger for type I cell MP depolarization probably due to ROS decrease, whereas the observed ROS increase has no influence on type I cell MP kinetics under hypoxia.

SummaryRecombinant human erythropoietin (rHuEPO) is an effective treatment for anaemia but concerns that it causes disease progression in cancer patients by activation of EPO receptors (EPOR) in tumour tissue have been controversial and have restricted its clinical use. Initial clinical studies were flawed because they used polyclonal antibodies, later shown to lack specificity for EPOR. Moreover, multiple isoforms of EPOR caused by differential splicing have been reported in cancer cell lines at the mRNA level but investigations of these variants and their potential impact on tumour progression, have been hampered by lack of suitable antibodies. The EpoCan consortium seeks to promote improved pathological testing of EPOR, leading to safer clinical use of rHuEPO, by producing well characterized EPOR antibodies. Using novel genetic and traditional peptide immunization protocols, we have produced mouse and rat monoclonal antibodies, and show that several of these specifically recognize EPOR by Western blot, immunoprecipitation, immunofluorescence, flow cytometry and immunohistochemistry in cell lines and clinical material. Widespread availability of these antibodies should enable the research community to gain a better understanding of the role of EPOR in cancer, and eventually to distinguish patients who can be treated safely by rHuEPO from those at increased risk from treatment.

Carotid body (CB) type I cells sense the blood Po2 and generate a nervous signal for stimulating ventilation and circulation when blood oxygen levels decline. Three oxygen-sensing enzyme complexes may be used for this purpose: 1) mitochondrial electron transport chain metabolism, 2) heme oxygenase 2 (HO-2)-generating CO, and/or 3) an NAD(P)H oxidase (NOX). We hypothesize that intracellular redox changes are the link between the sensor and nervous signals. To test this hypothesis type I cell autofluorescence of flavoproteins (Fp) and NAD(P)H within the mouse CB ex vivo was recorded as Fp/(Fp+NAD(P)H) redox ratio. CB type I cell redox ratio transiently declined with the onset of hypoxia. Upon reoxygenation, CB type I cells showed a significantly increased redox ratio. As a control organ, the non-oxygen-sensing sympathetic superior cervical ganglion (SCG) showed a continuously reduced redox ratio upon hypoxia. CN−, diphenyleneiodonium, or reactive oxygen species influenced chemoreceptor discharge (CND) with subsequent loss of O2 sensitivity and inhibited hypoxic Fp reduction only in the CB but not in SCG Fp, indicating a specific role of Fp in the oxygen-sensing process. Hypoxia-induced changes in CB type I cell redox ratio affected peptidyl prolyl isomerase Pin1, which is believed to colocalize with the NADPH oxidase subunit p47phox in the cell membrane to trigger the opening of potassium channels. We postulate that hypoxia-induced changes in the Fp-mediated redox ratio of the CB regulate the Pin1/p47phox tandem to alter type I cell potassium channels and therewith CND.

The protein “amplified in osteosarcoma-9” (OS-9) has been shown previously to interact with the prolyl hydroxylases PHD2 and PHD3. These enzymes initiate oxygen-dependent degradation of the α-subunit of hypoxia-inducible factor (HIF), a transcription factor that adapts cells to insufficient oxygen supply (hypoxia). A new model has been proposed where OS-9 triggers PHD dependent degradation of HIF-α. It was the aim of our study to define the molecular mode of action of OS-9 in the regulation of PHD and HIF activity. Although initial co-immunoprecipitation experiments confirmed physical interaction between OS-9 and PHD2, neither overexpression nor lentiviral inhibition of OS-9 expression affected HIF regulation. Subcellular localization experiments revealed a distinct reticular staining pattern for OS-9 while PHD2 was mainly localized in the cytoplasm. Further cell fractionation experiments and glycosylation tests indicated that OS-9 is a luminal ER protein. In vivo protein interaction analysis by fluorescence resonance energy transfer (FRET) showed no significant physical interaction of overexpressed PHD2-CFP and OS-9-YFP. We conclude that OS-9 plays no direct functional role in HIF degradation since physical interaction of OS-9 with oxygen sensing HIF prolyl hydroxylases cannot occur in vivo due to their different subcellular localization.

Carotid body (CB) type I cell hypoxia-sensing function is assumed to be based on potassium channel inhibition. Subsequent membrane depolarization initiates an intracellular calcium increase followed by transmitter release for excitation of synapses with linked nerve endings. Several reports, however, contradict this generally accepted concept by showing that type I cell oxygen-sensing properties vary significantly depending on the method of their isolation. We report therefore for the first time noninvasive mapping of the oxygen-sensing properties of type I cells within the intact adult mouse CB ex vivo by using multifocal Nipkow disk-based imaging of oxygen-, calcium- and potential-sensitive cellular dyes. Characteristic type I cell clusters were identified in the compact tissue by immunohistochemistry because of their large cell nuclei combined with positive tyrosine hydroxylase staining. The cellular calcium concentrations in these cell clusters either increased or decreased in response to reduced tissue oxygen concentrations. Under control conditions, cellular potential oscillations were uniform at ∼0.02 Hz. Under hypoxia-induced membrane depolarization, these oscillations ceased. Simultaneous increases and decreases in potential of these cell clusters resulted from spontaneous burstlike activities lasting ∼1.5 s. type I cells, identified during the experiments by cluster formation in combination with large cell nuclei, seem to respond to hypoxia with heterogeneous kinetics.