The majority of cell-based studies are performed under atmospheric oxygen (~18kPa), even though most tissues in vivo, including the brain, experience only ~4–8kPa O2, defined as physioxia (Keeley & Mann, 2019). This mismatch has influenced experimental design and our understanding of cellular physiology, with emerging evidence confirming cellular responses, proteome and metabolic profiles differ significantly in cells adapted to in vivo-like O2 conditions (~4–8kPa, physioxia) compared to hyperoxic atmospheric O2 environments.
To characterise the effects of physiological O2 on neuronal and brain endothelial physiology, we assessed ion channel function in differentiated mouse neuroblastoma N2a cells, primary embryonic mouse cortical neurons and human brain microvascular endothelial cells (hCMEC/D3) following long-term (4-5 days) adaptation to physiological (5kPa) and hyperoxic (18kPa) environments.
N2a cells were cultured in DMEM+GlutaMAX (Gibco) and differentiated (1% foetal bovine serum (FBS, Sigma Aldrich), 20µM retinoic acid (RA) for 7-10 days) and subjected to reduced serum and RA for 7-10 days prior to experimentation. Primary embryonic cultures were derived from brain cortices of C57BL/6J mice embryos (E15-17, Charles River UK; Envigo) following Schedule-1 procedure and cultured in DMEM+GlutaMAX supplemented with 10% FBS. hCMEC/D3 cells were cultured in 1% rat-tail collagen I coated-substrates (Sigma) in EBM phenol-red free basal cell media (Lonza, Switzerland) supplemented with EGM-2MV growth factors.
Electrophysiological recordings were made using a Nanion Port-a-Patch (Nanion Technologies) or MultiClamp 700B amplifier and HEKA Patchmaster. Immunocytochemistry and proliferation assays were performed at various timepoints. Physioxia and was achieved by culturing cells for 4-5 days at 5kPa O2. Unpaired Student’s t-test and ANOVA were used where appropriate with P<0.05 considered significant.
Under physioxic conditions, all three cell types showed altered electrophysiological phenotypes with markedly enhanced outward K+ currents which were sensitive to the Kv3/BK channel blocker TEA (3mM). Following partial current inhibition in hCMEC/D3 cells by TEA (@50 mV, 5kPa and 18kPa: TEA: 62±5 % P=0.0381 versus 79±8 % P=0.3723), the remaining K+ currents showed differential sensitivity to the SK and IK channel blockers 50µM TRAM-34 and 100nM apamin, respectively, under both O2 adaptations. At 5kPa adaptations, N2a cells showed comparable TEA sensitivities (@50 mV, 12% versus 18%), whereas primary neurons showed larger TEA sensitivity (@50 mV, 31% versus 14%), Both cell types exhibited larger voltage-activated Na+ currents under physioxia and activation voltages for both, voltage-gated Kv and Nav currents in N2a and primary neurons displayed left-shifted values.
Current-evoked neuronal excitability and action potential (AP) amplitudes were unaffected under physioxia, however, AP half-width were strongly reduced in N2a cells mirroring the enhanced activity of Kv currents under physioxia. Passive membrane properties were partially altered in physioxia, resulting in lower membrane time-constants (tm) and membrane resistances (Rm) but unchanged cell capacitance (Cm). Furthermore, N2a cells showed an increased proportion of NeuN-positive cells and enhanced axonal outgrowth following physiological oxygen adaptations without effects on cell proliferation.
Our data provide new evidence that adaptation of cells to physioxia alters neuronal and endothelial ion channel characteristics, reinforcing the need for investigating the effects of physiologically oxygen levels on the redox signalling in cell models to improve clinical translation.