ONLINE SUPPLEMENTARY MATERIAL

Ultrafast optogenetic control

Lisa A. Gunaydin, Ofer Yizhar, André Berndt, Vikaas S. Sohal,

Karl Deisseroth and Peter Hegemann

Correspondence and requests for materials should be addressed to

P.H (hegemape@rz.hu-berlin.de) or K.D. (deissero@stanford.edu).

Nature Neuroscience: doi:10.1038/nn.2495

Rec

over

y (%

)

100

75

50

25

Time (s)0 20 40 60

2s

ΔI1ΔI2

WTE123A

a

b

c

d

WT ChR2

τ in : 55.5 ms

0.4 sτoff : 9.8 ms

E123A

τ in : 14.7 ms

τoff : 5.5 ms

20 ms

5 ns

Wavelength (nm)

Act

ivity

(a.u

.)

1.0

0.5

0.0450 475 500 525 550 575

WTE123A

WT 9.2 msE123A 4.0 ms

τ off

Supplementary Figure 1. Characterization of an alternative fast mutant: E123A.

Nature Neuroscience: doi:10.1038/nn.2495

Supplementary Figure 1:

Characterization of an alternative fast mutant: E123A.

(a) Voltage clamp recordings of WT ChR2 and E123A in Xenopus oocytes under stimulation

with a 1 s light pulse (blue bars, 500 nm ± 25, 50 mW/cm2) at -100 mV and 100 mM NaCl in the

external solution (pH 7.5). τ-values characterize the mono-exponential transition from peak to

stationary currents (τin) and current kinetics (τoff) after the light was switched off. τoff = 9.8 ± 1.3

ms for WT (n = 10) and τoff = 6.2 ± 1.3 ms for E123A (n = 11; p < 0.005). Vertical black bars: 10

nA. (b) Recovery of peak current in WT ChR2 and E123A upon stimulation with a second light

pulse after a variable dark period, recorded in oocytes. Inset shows the peak recovery of WT

ChR2 after 2s in darkness at -75 mV in standard solution (100 mM NaCl, pH 7.5). The ratio of

ΔI2 to ΔI1 yields the recovery in percent. Values for WT ChR2 (black circles) and E123A (pink

circles) are plotted versus the dark interval between the two flashes; data points were fit by a

mono-exponential decay (colored lines). Peak recovery was τ = 10.7 ± 0.8 s for WT (n = 15) and

τ = 0.9 ± 0.2 s for E123A (n = 4; p < 0.005). (c) Oocyte current recordings; excitation delivered

with a 5 nanosecond laser flash (470 nm, 0.5 mW/cm2, blue arrow) at -100 mV. Photocurrent

decay was fitted mono-exponentially with τ-values shown for WT ChR2 (black) and E123A

(pink), and traces were normalized to highlight differences in decay kinetics. Off kinetics were τ

= 9.2 ± 1.3 ms for WT (n = 9) and 4.0 ± 0.2 ms for E123A (n = 11; p < 0.005). Flash-to-peak

Nature Neuroscience: doi:10.1038/nn.2495

current times were tp = 2.4 ± 0.3 ms for WT (n = 9) and tp = 0.8 ± 0.1 ms for E123A (n = 11; p <

0.005). Vertical bars correspond to 2.5 nA (WT: black; E123A: pink). (d) The action spectrum

of E123A (pink, n = 3) is shown in comparison with spectrum of WT ChR2 (black, n = 6).

Current amplitudes were measured at the different wavelengths, normalized to respective

maximum values, and corrected for fluctuations of flash intensity.

5 Hz

1 ms pulse width 2 ms pulse width 5 ms pulse width

10 Hz

20 Hz

40 Hz

60 Hz

80 Hz

100 Hz

125 Hz

150 Hz

200 Hz

100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0

100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0

100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0100

50

0

% s

ucce

ssfu

l spi

kes

Pulse-train interval1 2 3 4 1 2 3 4

1 2 3 4

Supplementary Figure 2. Temporal stationarity during extended trains.

WT ChR2ChETA

Nature Neuroscience: doi:10.1038/nn.2495

Supplementary Figure 2:

Temporal stationarity during extended trains.

The percentage of successful spikes is plotted over time for WT ChR2 and ChETA, across a

range of light pulse frequency and light pulse width. Note the deterioration over time of WT

ChR2 performance in following light pulses. In contrast, the performance of ChETA (in addition

to showing improved efficacy at high frequencies for >1 ms pulse width) was temporally

stationary, in that performance in response to later light flashes was indistinguishable from that

for earlier light flashes in the train, across all conditions. Trains of 40 light pulses were delivered,

and spike responses were grouped into four intervals corresponding to successive blocks of 10

light pulses (interval 1 = light pulses 1-10, interval 2 = light pulses 11-20, interval 3 = light

pulses 21-30, interval 4 = light pulses 31-40).

Nature Neuroscience: doi:10.1038/nn.2495

ChETA_Suppl_Material_ 1-7-10ChETA_Suppl_Fig1_1-7-10ChETA_Suppl_Fig2_12-30-09ChETA_Suppl_Material_ 1-7-10