Fig. 6: The specific oxygen level increased the frequency of calcium oscillations.

A section of the symmetric oxygen gradient was acquired using stage scanning fluorescence microscopy, encompassing a 0–20% gradient as characterized earlier. a The resultant raw intensity image has 13,156 × 2556 pixels from the collage, or 33 Megapixels total per image. Artifacts visible in these images were removed prior to analysis. Collage images were acquired at 1-min intervals for 46 min. Only time points corresponding to glucose boluses were used, giving 24 points from 13 to 36 min. The initial overshoot at 12 min was removed to reduce artifacts (see Supplementary Data S3). b A per-pixel time-based Fourier transform was applied, and the magnitude profile was weighed at the center of gravity, giving a nominal frequency value for each pixel. The result was replotted to show this frequency versus underlying oxygen concentrations, where an intense region approximately 12% oxygen was observed (visible banding seen in collage was due to image stitching). c The pixels were averaged and flattened vertically to show a frequency profile versus oxygen concentrations. The resultant profile showed a peak oscillation at 12% oxygen. This oxygen level is neither hypoxic nor normoxic in the standard cell culture sense. d A conventional dual oscillator model for calcium includes both electrical and metabolic components. Feedback in the PFK portion of glycolysis is viewed as the center of metabolic oscillation, whereas calcium channel feedback is viewed as the center for electrical oscillations. Oxygen, due to its effect on the mitochondria, could be a key factor in both oscillator components