Development of novel FP-based probes for live-cell imaging of nitric oxide dynamics

Abstract

Nitric oxide () is a free radical with a wide range of biological effects, but practically impossible to visualize in single cells. Here we report the development of novel multicoloured fluorescent quenching-based probes by fusing a bacteria-derived -binding domain close to distinct fluorescent protein variants. These genetically encoded probes, referred to as geNOps, provide a selective, specific and real-time read-out of cellular dynamics and, hence, open a new era of bioimaging. The combination of geNOps with a Ca(2+) sensor allowed us to visualize and Ca(2+) signals simultaneously in single endothelial cells. Moreover, targeting of the probes was used to detect signals within mitochondria. The geNOps are useful new tools to further investigate and understand the complex patterns of signalling on the single (sub)cellular level.

Figure 1. Fusion of the bacterial -binding GAF domain to fluorescent proteins, resulting in differently coloured fluorescent quenching-based probes, the geNOps.

(a) Predicted three-dimensional structure of geNOps. (b) Schematic overview of differently coloured geNOps. (c) Average curves (mean±s.e.m.) over time of normalized delta fluorescence signals in % of the differently coloured geNOps signals in response to 10 μM NOC-7 (n=10 for C-geNOp cyan curve; n=12 for M-geNOp light green curve; n=11 for G-geNOp dark-green curve; n=13 for Y-geNOp yellow curve; n=9 for O-geNOp orange curve). Experiments were performed using HeLa cells. (d) Representative pseudo-coloured images of a HeLa cell expressing O-geNOp before cell treatment (0 s), upon addition of 10 μM NOC-7 (5 and 12 s) and upon the removal of NOC-7 (87 and 175 s). Scale bar, 10 μm. See also Supplementary Video 1. (e) Fluorescence intensity change in % versus time of a single HeLa cell expressing C-geNOp in response to different concentrations of NOC-7. (f) Concentration response curves showing the effects of different NOC-7 concentrations on fluorescence intensities of the differently coloured geNOps that were expressed in HeLa cells. Points represent average values±s.e.m.; n=5 for C-geNOp, n=8-11 for M-geNOp; n=5–6 for G-geNOp; n=3 for Y-geNOp; n=3–5 for O-geNOp. (g) Representative curves showing fluorescence over time of wild-type C-geNOp and C-geNOp^mut upon addition of 10 μM NOC-7 to HeLa cells. Statistics are shown in Supplementary Fig. 7. (h) Bars representing maximal delta fluorescence signals±s.e.m. of G-geNOp expressed in HeLa cells in response to 10 μM NOC-7 (, green column, n=26), 100 μM of the CO-releasing compound CORM-3 (CO, n=16), 100 μM KO₂ (O₂, n=12) or 100 μM peroxynitrite (ONOO⁻, n=7). *P<0.05 versus control using the unpaired t-test.

Figure 2. The properties of geNOps remain unaffected upon mitochondria targeting.

(a) Confocal images of HeLa cells expressing either mtC-geNOp (left image) or mtG-geNOp (right image). Scale bar, 10 μm. (b) Normalized average curves±s.e.m. of mtC-geNOp (left panel) and mtG-geNOp (right panel) signals with (n=4 for mtC-geNOp; n=7 for mtG-geNOp) and without (n=5 for mtC-geNOp; n=4 for mtG-geNOp) iron(II)/vitamin C pretreatment. Experiments were performed using HeLa cells. (c) Representative original curve showing fluorescence over time of mtG-geNOp expressed in HeLa cells in response to consecutive applications of 3 mM SNP (n=3). (d) Concentration response curves showing the effects of different NOC-7 concentrations on fluorescence intensities of either mtC-geNOp (left panel, cyan curve, n=4) versus C-geNOp (left panel, grey curve, for n see Fig. 1f) or mtG-geNOp (right panel, green curve, n=6) versus G-geNOp (right panel grey curve, for n see Fig. 1f). Experiments were performed using HeLa cells. Points represent average values±s.e.m.

Figure 3. Imaging of cellular dynamics with geNOps in response to different -liberating molecules.

(a) Representative single HeLa cell dynamics in response to 1 μM NOC-7, 1 mM SNP or 1 μM PROLI NONOate. Cells expressing C-geNOp were imaged. Inverted curves (1−F/F0 in %) are shown. Average curves with s.e.m. are shown in Supplementary Fig. 15. (b) Scatter dot plot showing maximal single-cell C-geNOp signals in response to 10 μM NOC-7 on different dishes. White column represents the normalized average±s.e.m. C-geNOp signal of all single HeLa cells (n=67). (c) Intracellular dynamics of a single HeLa cell expressing C-geNOp in response to different concentrations of S-NO-HSA (curve is inverted). (d) Respective ΔF_Intensity mean values±s.e.m. are blotted against concentrations that are released by 1, 3, 10 and 30 μM S-NO-HSA (n=6). released by S-NO-HSA was quantified using a porphyrinic nanosensor (for details see Supplementary Fig. 16 and methods).

Figure 4. Live-cell imaging of signals and cell functions in primary cardiomyocytes and glioblastoma cells using geNOps.

(a) Curves represent representative simultaneous recordings of cellular Ca²⁺ (black ratio curve) and (red inverted curve) signals over time of a single fura-2/am-loaded embryonic ventricular cardiomyocyte expressing G-geNOp. The cell was treated with 30 μM S-NO-HSA in the presence of extracellular Ca²⁺ using a perfusion system (n=4). (b) Representative recordings of cellular dynamics (red inverted curve, n=4) of human glioblastoma cells (U87-MG cells) expressing C-geNOp. Cells were either treated with a mixture of 10 μM PROLI NONOate and 10 μM NOC-7 (red curve) or remained untreated (control cell, black curve). (c) Cell velocity of glioblastoma cells in μm min⁻¹ extracted from the X/Y positions over time of a control cell (black columns) and a cell treated with donors as indicated in b and d. (d) Graphs represent X/Y positions of glioblastoma cells over time as indicated in b and c.

Figure 5. Live-cell imaging of Ca²⁺-triggered production in signals endothelial cells.

(a) Single endothelial cell (EA.hy926 cells) responses upon cell treatment with different concentrations of histamine (right panel, 0.3 μM; 1.0 μM; 3.0 μM; 100 μM histamine, inverted curves are shown) in the absence of extracellular Ca²⁺. For the concentration response curve (right panel), cells expressing C-geNOp were stimulated with 0.1 μM (n=6), 0.3 μM (n=6), 1.0 μM (n=7), 3.0 μM (n=7), 10.0 μM (n=7) or 100.0 μM (n=12) histamine, yielding an effector concentration for half-maximum response of 1.4 (0.8–2.5) μM. Red points represent average values±s.e.m. (b) Cellular dynamics of EA.hy926 cells expressing C-geNOp. Cells were stimulated with 100 μM histamine in Ca²⁺ containing buffer for 9 min under control conditions (black inverted curve, n=4) or during stimulation, 1 mM L-NAME was added (red inverted curve, n=9). (c) Columns represent maximal G-geNOps signals±s.e.m. in response to either 10 (red columns) or 100 μM (black columns) histamine under control conditions (n=5 for both histamine concentrations) and in the presence of the NOS inhibitor (1 mM; n=10 for both histamine concentrations). *P<0.05 versus control (10 μM histamine); #P<0.05 versus control (100 μM histamine). P values were calculated using unpaired t-test. (d) Average curves over time (right panel) and statistics of the maximal cytosolic increase (columns representing average values±s.e.m. in the left panel) in EA.hy926 cells in response to 30 μM histamine (black curve, black column, n=16), 30 μM ATP (red curve and red column, n=20) or 1 μM thapsigargin (blue curve, blue column, n=15). Endothelial cells expressing C-geNOps were used *P<0.05 versus histamine/ATP using unpaired t-test. (e) Curves represent simultaneous recordings of cellular Ca²⁺ (black ratio curve) and (red inverted curve) signals over time of a single fura-2/am-loaded endothelial cell expressing O-geNOp as shown in Supplementary Fig. 21. The cell was stimulated with 100 μM histamine in the presence of extracellular Ca²⁺. (f) Simultaneous recordings of cellular Ca²⁺ (black ratio curve) and (red inverted curve) signals over time of a single fura-2/am-loaded endothelial cell expressing G-geNOp. During imaging, the cell was treated with 1 μM ionomycin in the absence (1 mM EGTA) and presence of 2 mM Ca²⁺.

Figure 6. Visualization of signals within mitochondria of signals endothelial cells.

(a) Average curves±s.e.m. showing mitochondrial signals measured with mtG-geNOp expressed in EA.hy926 cells (green curve, n=7) and respective signals obtained with mtG-geNOp^mut (red curve, n=7). Cells were treated first with 100 μM ATP, then with 1 mM L-NAME and subsequently with 10 μM NOC-7. (b) Average curves±s.e.m. showing cytsolic signals measured with G-geNOp expressed in EA.hy926 cells (green curve, n=5). As shown in a, cells were treated first with 100 μM ATP, then with 1 mM L-NAME and subsequently with 10 μM NOC-7. (c) Columns represent maximal average values of curves shown in a and b. *P<0.05 versus basal. #P<0.05 versus +ATP. P values were calculated using unpaired t-test. (d) Confocal images of endothelial cells expressing both mtC-geNOp (left image) and cytosolic G-geNOp (middle image). Scale bar, 10 μm. (e) Representative simultaneous recordings of mtC-geNOp (grey curve) and cytosolic G-geNOp (green curve) signals over time in a single EA.hy926 cell in response to 100 μM ATP.