Fig 1.
The ecATeam3.10 sensor detects extracellular ATP.
(A) Schematic of ecAT3.10 design in which the ATeam3.10 FRET-based ATP sensor is displayed on the cell surface via a PDGFR transmembrane anchor. (B) Fluorescence intensity in the FRET channel increases upon wash-in of 100 μM ATP while the CFP donor intensity decreases. As expected, the YFP direct acceptor channel does not show an ATP-dependent change. Cells were imaged under continuous perfusion. (C) Representative widefield fluorescence images for the cells analyzed in (D-E). The first panel in the YFP channel shows morphology, the subsequent panels are false colored to show the change in the normalized FRET/CFP pixel-by-pixel ratio signal. Scale bar is 20μm. (D) The average FRET/CFP emission ratio, which we refer to as the ratio signal, shows a robust increase of 0.27 ± 0.01 (n = 41 cells). Values and solid line traces are cell means, and errors and error bars are standard errors of the means. (E) To account for drift, a linear baseline was fitted, and the ratio signal was normalized on a cell-by-cell basis (individual cells, gray), showing an average fold change of 1.127 ± 0.005 (population average, red).
Fig 2.
Surface-localized ecAT3.10 responds to the addition of extracellular ATP.
High magnification confocal microscopy demonstrates that the ecAT3.10 ratio signal response to extracellular ATP occurs exclusively at the membrane. (A) Representative confocal fluorescence intensity images in the CFP and FRET channels show strong membrane localized signal and some intracellular puncta. Scale bar is 20 μm. (B) Pixel-by-pixel FRET/CFP ratio images before (left) and after addition of 100 μM ATP (right) illustrates a response at the membrane only. (C-E) Line profile analysis for the two example cells shown in (A). A 5-pixel width line region of interest was drawn through the membrane and peak areas of intracellular fluorescence. (C) The fluorescence intensity line profiles exhibit peaks at the membrane and also from intracellular locations that are likely ER/Golgi in origin. The line profiles from the ratio images shown in (B) before (D) and after (E) the addition of extracellular ATP clearly show that the ecAT3.10 ratio signal increases at the membrane and not from the intracellular sites. (F) After 3 minutes, 100 μM ATP was added. The time course shows that the membrane-localized ratio signals increase, but the signals from intracellular regions do not change (mean ± sem for three experiments with n = 32 cells, n = 60 cells, and n = 46 cells).
Fig 3.
Characterization of ATP affinity and response kinetics.
ecATeam3.10 exhibits a concentration-dependent change in FRET/CFP ratio signal. Live Neuro2A cells expressing ecAT3.10 were imaged at room temperature under continuous perfusion, during which increasing concentrations of extracellular ATP were washed in a stepwise fashion. (A) A representative time course from one experiment with four cells. Each grey trace represents the response from an individual cell in which the regions of interest were drawn around the membrane. The orange trace is the mean response. (B) Summary of ATP dose-response data from five independent experiments, each represented its own symbols. The orange squares represent data from (A), and the dashed red curve is fit to the mean of all experiments. In four of the five experiments, sufficient data was available for independent fitting to a Hill equation (mean ± sem, n = 4 experiments: K = 12 ± 5 μM, Ratiomax/Ratiomin = 1.23 ± 0.01, n = 1.4 ± 0.2). A global fit of data from all five experiments was also performed for comparison. (fitting mean ± sem, n = 5 experiments: K = 11 ± 2 μM, Ratiomax/Ratiomin = 1.28 ± 0.01, n = 0.9 ± 0.1). (C-D) ecAT3.10 responds to an increase in extracellular ATP within seconds. (C) Cells were imaged during fast perfusion wash in of 100 μM ATP, and the sensor exhibited an increase in FRET within seconds (n = 27 cells, 3 experiments; two-exponential fit: τfast = 13 ± 1 sec, τslow = 172 ± 18 sec). (D) An expanded view of the initial response shown in (C). Red, FRET fluorescence channel intensity. Yellow, YFP fluorescence channel intensity. Dotted black curve, two-exponential fit.
Fig 4.
ecATeam3.10 detects ATP hydrolysis by ectonucleotidases.
Live Neuro2A cells expressing ecAT3.10 were imaged under non-perfusion, static bath conditions. (A) The addition of 10 μM ATP did not elicit a response (black) unless the cells were pre-treated with the ectonucleotidase inhibitor ARL67156 at 100 μM (red). Apyrase addition degrades extracellular ATP confirming a reversible ATP-specific sensor response. (B) Ectonucleotidase inhibition by ARL67156 pre-treatment also potentiated responses when 30 μM ATP was added. Vehicle, black. ARL67156, red. (A) and (B), n = 6, 2 independent investigators. (C) ARL67156 does not directly affect the ATeam3.10 sensor response to ATP.
Fig 5.
ecAT3.10 detects nucleotide-stimulated release of ATP from cells.
(A) A biphasic response was apparent on average when 100 μM ATP was added at time = 30 minutes to the static bath of ecAT3.10-expressing Neuro2A cells. The decrease in signal at time = 45 minutes is due to ARL67156-sensitive ectonucleotidase activity, and the subsequent increase in signal after time = 55 minutes reports a secondary release of endogenous ATP from cells. (B) Pretreatment with 100 μM ARL67156 abrogates the transient decrease in ectonucleotidase activity. (C) Addition of 100 μM ADP also stimulate a release of ATP from cells, which was potentiated by pretreatment with ARL67156 (D). Apyrase addition degrades extracellular ATP confirming a reversible ATP-specific sensor response. Solid lines show average responses from ecAT3.10, and dashed traces show average responses from the negative control ecATYEMK sensor (n = 3).
Fig 6.
ecATeam3.10 detects ADP-stimulated release of endogenous ATP from live Neuro2A cells.
(A) ecAT3.10-expressing Neuro2A cells were imaged under continuous perfusion conditions to demonstrate that ecAT3.10 has greater sensitivity to ATP compared to ADP. (B) Under static bath conditions, paired measurements by ecAT3.10 real-time imaging and endpoint luciferase assays demonstrate that ADP stimulates ATP release. In (A) and (B) grey traces are individual cells, and the bold trace is the cell mean. Arrows in (B) and (C) indicate the addition of ADP and apyrase, sequentially. In (C), time points represent samples from the bath solution taken during the ecAT3.10 imaging experiment shown in (B).
Fig 7.
ADP-stimulated release of ATP release from Neuro2A cells may be mediated in part by purinergic pathways.
(A) Average time course of ecAT3.10 ratio signal (baseline normalized, n = 6 experiments) from Neuro2A cells that were imaged under static bath conditions. 30 μM ADP was added to stimulate release of ATP. Pretreatment with 100 μM ARL67156 (red) potentiated the response in comparison to vehicle treatment (black). Treatment with 3.3 μM suramin (green) attenuated extracellular ATP levels compared to vehicle (black), and treatment with suramin in the presence of ARL67156 (blue) attenuated the release of ATP compared to ARL67156 alone (red). Apyrase addition degrades extracellular ATP confirming a reversible ATP-specific sensor response. (B) Summary of cell averaged peak responses for individual experiments. Peak responses: vehicle, 1.06 ± 0.01; ARL67156, 1.13 ± 0.01; suramin, 1.027 ± 0.002; ARL67156 plus suramin, 1.048 ± 0.006, mean ± sem. t-test, *p = 0.004, **p = 0.02, ***p = 0.0004. n = 6, 2 independent investigators.