Table 1.
Structure and voltage responses of Nabi1 probes.
Nabi1 is composed of the N-terminus of the Ciona voltage sensitive domain, mKO (1–218 amino acids), the S1-S4 of the voltage sensitive domain, and UKG (1–224 amino acids), followed by a stop codon. Nabi1 probes were evaluated by using 100 ms voltage steps from a -70 mV holding potential to -170 mV, -20 mV, +30 mV, and +130 mV or to -120 mV, -20 mV, +30 mV, +80mV, and +130 mV in transiently transfected HEK293 cells. We tested 3–21 cells for each Nabi1 probe. Averaged optical traces from 16 trials were analyzed. ΔF/F was calculated for 100 mv depolarizations by taking the average of optical signals from the tested cells for each probe. ΔF/Fmax was the largest value of the optical responses observed during any depolarizing voltage step. V1/2 (membrane potential at half maximum ΔF/F) was calculated from optical signals of all tested cells. The time constants were calculated using a double exponential fitting of 1–6 representative donor traces for each probe.
Fig 1.
The structure and optical responses of Nabi1 probes to changes of membrane potential in transiently transfected HEK293 cells.
A. Nabi1 probes contain mKO (orange) and UKG (green) as the FRET pair at flanking regions of the Ciona voltage sensitive domain. B. The sequence of the Ciona voltage sensitive domain (1st to 259th amino acids). The transmembrane domains S1-S4 are underlined. mKO was inserted at one of 6 locations indicated by orange arrowheads. UKG was inserted at one of 8 locations indicated by green arrowheads. C. Representative donor (UKG, green) and acceptor (mKO, orange) signals of 3 selected Nabi1 probes. Traces from single trials of Nabi1.213 (FPs at 84th and 245th amino acids), Nabi1.242 (FPs at 95th and 246th), and Nabi1.244 (FPs at 95th and 245th) are shown without temporal filtering. D. Representative donor (mCitrine, yellow) and acceptor (mKate, red) signals of VSFP Butterfly1.2 [13]. The traces are the average of 16 trials without temporal filtering and shown using scale bars of ΔF/F of 10% and 2%. The same voltage steps as in C were used for VSFP Butterfly1.2.
Fig 2.
Signal sizes of Nabi1 probes in HEK293 cells.
Positions of mKO and UKG are indicated. Grey rectangles indicate the combinations that were not tested. Nabi1 variants Nabi1.82, Nabi1.102, Nabi1.243, and Nabi1.152 (Table 1) are not included in this figure or Figs 3 and 4 (see Table 1). A. ΔF/F of Nabi1 optical responses for a 100 mV depolarization. ΔF/F was calculated by taking the averages of the optical responses for a 100 mv depolarization from 3–21 cells for each probe. Red rectangles indicate Nabi1 probes with ΔF/F of either donor or acceptor signal equal to or greater than 5%, blue rectangles indicate probes with ΔF/F between 4 and 5%, and green rectangles indicate probes with ΔF/F < 4%. Detailed ΔF/F values are listed in Table 1. B. Maximum ΔF/F during voltage steps between -170 mV and +130 mV from a -70 mV holding potential. Red rectangles indicate Nabi1 probes with ΔF/Fmax of donor or acceptor signal equal to or greater than 12%, blue rectangles indicate probes with ΔF/F between 7 and 12%, and green rectangles indicate probes with ΔF/F < 7%. Detailed ΔF/Fmax values are listed in Table 1.
Fig 3.
Time constants for signal activation and decay of the donor signal for Nabi1 probes in response to a 100 mV depolarization.
Optical responses were analyzed using a double exponential function. Averages were taken from signals of 1–5 representative cells. Detailed values are listed in Table 1. Positions of mKO and UKG are indicated. Grey rectangles indicate the combinations that were not tested. A. Fast components of signal activation (τ1on). Red rectangles indicate probes with τ1on faster than or equal to 5 ms. Blue rectangles indicate probes with τ1on equal to or faster than 10 ms but slower than 5 ms. Green rectangles indicate probes with τ1on slower than 10 ms. B. Fast components of signal recovery (τ1off). Same color scale as A. C. The summary of signal activation and recovery kinetics for the slow components of Nabi1. Red rectangles indicate probes with τ2 of signal activation faster than or equal to 50 ms and τ2 of signal decay equal to or faster than 50 ms. Blue rectangles indicate probes with τ2 of activation equal to or faster than 50 ms and τ2 of decay slower than 50 ms. Green rectangles indicate probes with τ2 of activation slower than 50 ms. D. The signal decay kinetics of Nabi1 probes showed an inverse relationship with percent fast components (R square = 0.17, p = 0.01).
Fig 4.
Membrane potential at half maximum ΔF/F (V1/2) of Nabi1 probes in HEK293 cells.
A. V1/2 of Nabi1 probes. Positions of mKO and UKG are indicated. Grey rectangles indicate the combinations that were not tested. V1/2 was estimated from normalized donor ΔF/F of 2–14 measurements for each probe (see Materials and Methods). Detailed values are listed in Table 1. Purple squares indicate probes with V1/2 at more negative values than -40mV, green squares indicate probes with V1/2 between -40mV and -20mV, and yellow squares indicate probes with V1/2 that are more positive than -20 mV. B. The relationship between V1/2 and the time constants of optical responses. The left panel shows the relationship for τ1 of signal activation. There was no significant correlation (R square = 0.04, p = 0.22). The right panel indicates an inverse relationship with τ1 for signal decay. The Nabi1 probes with more negative V1/2 tend to have slower τ1 off values (R square = 0.38, P < 0.01).
Fig 5.
A. Nabi2 has Clover (green) and mRuby2 (red) as the FRET pair. Nabi1 and Nabi2 are also different in the linkers between the Ciona voltage sensing domain (CiVSD) and FP β-cans. The Ciona voltage sensitive domain and the N-terminus are shown in gray. The four transmembrane domains S1-S4 are indicated as black bars. FPs are color-coded as mKO (orange), UKG (dark green), Clover (bright green), and mRuby2 (red). β-cans of the FPs are shown as vertical rectangles. B. Representative donor (Clover; green) and acceptor (mRuby2; red) signals of Nabi2.213, Nabi2.242, and Nabi2.244. All of the traces were from single trials and without temporal filtering. C. Our results using VSFP-CR [20] are shown from single trials using scale bars of ΔF/F of 10% and 5% and without temporal filtering for comparison. The same voltage steps as in B were used for VSFP-CR.
Table 2.
Comparison of Nabi1 and Nabi2 probes.
ΔF/F, time constants, and V1/2 were compared using averaged optical traces of 16 trials in transiently transfected HEK293 cells. ΔF/F and time constants were evaluated from optical responses of the donor to a 100 mV depolarization. We examined 5–20 cells for each probe. V1/2 was evaluated by fitting with a Boltzmann function.
Fig 6.
Comparison of Nabi1 and Nabi2 probes.
A-D. Signal size and time constants of eight Nabi1 and eight Nabi2 probes were averaged and compared by a t test. An asterisk (*) indicates statistical differences (p<0.05). A. ΔF/F of Nabi1 and Nabi2 probes for a 100 mV depolarization in HEK cells. The signals were taken from the averaged signals of Nabi1 or Nabi2 probes. Converting of Nabi1 to Nabi2 probes increased donor signal size, but not acceptor’s. B. Comparison of the time constants of the optical responses of donor to a 100 mV depolarization. Left: τ1 of signal activation and decay. Nabi1 and Nabi2 were not significantly different. Middle: Nabi2 had faster slow component of signal activation compared to Nabi1 (p = 0.03). Signal decay did not change significantly. Right: Replacement of FPs did not change significantly the percent fast component of signal activation, but reduced the percent fast component of signal decay in Nabi2 compared to Nabi1 (p = 0.03). C. Signal vs. voltage for three Nabi1 and Nabi2 probes. ΔF/F values were normalized to the maximum ΔF/F of each Nabi probe (N = 6–8 cells for Nabi1 probes, 14–19 cells for Nabi2 probes) and fit by a Boltzmann function. Left panel: donor signals vs voltage. Right panel: acceptor signals vs voltage. Dotted lines with open symbols are for Nabi1 probes and solid lines with closed symbols are for Nabi2 probes.
Fig 7.
Nabi2.244 responded to action potentials in primary cultured hippocampal neurons.
A. Expression of Nabi2.244 in a neuron, as shown with confocal laser scanning microscopy. A merged image of clover/mRuby2 is shown. The scale bar indicates 10 μm. Nabi2.244 was partially localized in the plasma membrane of soma and processes. Arrowheads point to membrane localization of fluorescent signals. Nabi2.244 was also found in intracellular aggregates. B. Representative donor (green) and acceptor (red) signal traces from a single trial (Fluorescence) responding to a stimulated action potential (Electrode) from a neuron expressing Nabi2.244. Signal traces were smoothed by two passes of a low pass binomial filter. C. A representative donor signal from a single trial (Fluorescence; green) responding to a spontaneous action potential (Electrode; black trace is without filtering and blue trace is with filtering) from a neuron expressing Nabi2.244. A single signal trace is shown smoothed by a low pass Gaussian filter at 50 Hz. D. Representative donor (green) and acceptor (red) signal traces from a single trial during a train of stimulated action potentials (black) from a neuron expressing Nabi2.244. Optical traces from a single trial of a current clamped cell are shown smoothed by 10 passes of a low pass binomial filter. Unfiltered signal traces (purple and dark green) are superimposed. E. An image of a neuron with resting fluorescence intensity (left). The optical signals in D were collected from the soma and the proximal processes, marked with green (right).
Table 3.
Calculated responses of Nabi2 probes for a 5 ms 100 mV depolarization.
Averaged donor traces of 16 trials were obtained from transiently transfected HEK293 cells, fit by a double exponential function, and used to calculate the responses at 5 ms.