Page 94 - Science

RESEARCH | REPORT

Fig. 2. Target
pattern reen- laser stim
trance increases 35 Session 1 35 Session 4 5 Session 4
during VTA 30 30 4 Baseline Bootstrap
optogenetic 25 25 BMI ChR2
self-stimulation. 20 20 3
(A) Distribution of %Occupancy 15 15 Occupancy Gain
the percent of 2
time that each 10 10
Baseline 1
tone was occupied 5 BMI ChR2 5
during baseline 0 0 0
(gray) and BMI (T1) 5 6 8 10 12 15 19 (T2) (T1) 5 6 8 10 12 15 19 (T2) (T1) 5 6 8 10 12 15 19 (T2)
(cyan) blocks of Tones (kHz) Tones (kHz) Tones (kHz)
session 1 (left) and
session 4 (right) in
one mouse. (No Baseline Bootstrap BMI ChR2: n=10 BMI YFP: n=6
tones were actu-
Session 2 Session 3 Session 4
ally played during 7 7 7 7 7 7
the baseline 6 6 6
block.) T1, target 1; 5 5 5
T2, target 2.
(B) Quantification Occupancy Gain 4 1 4 1 4 1
of the behavioral 3 3 3
changes between 2 2 2
sessions 1 and 4. Downloaded from
1 1 1
The session 4
0 0 0
occupancy gain
(T1) 5 6 8 10 12 15 19 (T2) (T1) 5 6 8 10 12 15 19 (T2) (T1) 5 6 8 10 12 15 19 (T2)
(cyan) is the ses-
sion 4 BMI Tones (kHz) Tones (kHz) Tones (kHz)
distribution nor-
malized to the T1 Occupancy T1 vs T2 Preference Low vs High Freq Preference
session 4 baseline 5 6 8 10 12 15 19 kHz 5 6 8 10 12 15 19 kHz 5 6 8 10 12 15 19 kHz
distribution, then T1 T2 T2 http://science.sciencemag.org/
normalized to the T1 T2 T1
session 1 ratio. For
(B) to (F), the 7 Baseline * * 4 2 * * *
95% confidence 6 Bootstrap 3.5
interval for the 5 BMI ChR2 3 * * * 1.5
baseline
bootstrap Occupancy Gain T1 4 Preference Gain T1 vs T2 2.5 Preference Gain Low vs High Freq
distribution is 3 2
plotted in gray 2 * 1.5 1 on March 1, 2018
(see supplemen-
1 1
tary methods).To
0 0.5 0.5
generate the
bootstrap 7 Baseline 4 2
distribution, the 6 Bootstrap 3.5
BMI session was 5 BMI YFP 3 1.5
simulated 10,000 2.5
timesasthough Occupancy Gain T1 4 Preference Gain T1 vs T2 Preference Gain Low vs High Freq
neural activity 3 2
were drawn from 2 1.5 1
that session’s 1 1
baseline period.
0 0.5 0.5
(C) The occu- 1 2 3 4 1 2 3 4 1 2 3 4
pancy gain over Sessions Sessions Sessions
sessions 2
through 4. For (C) to (F), mean and SEM over ChR2 animals (n =10) areshown larger than bootstrap. (E) Top: The preference gain for 5 kHz (target 1) versus
in cyan and over YFP animals (n = 6) are shown in black. By session 4, the 19 kHz (target 2) is plotted over sessions. Middle: ChR2 (cyan) were significantly
behavioral changes were statistically different across tones for ChR2 but not larger than bootstrap after session 1 (P <1×10 –5 for sessions 2 through 4).
YFP [repeated measures analysis of variance (ANOVA): ChR2, F 6,48 =3.46, Bottom: YFP (black) were never significantly larger than bootstrap. (F)Top:
–3
P =6.4 ×10 ;YFP, F 6,30 =0.96, P = 0.47]. Insession 4, 5kHz (target1) The preference gain for low-pitch tones (5 to 8 kHz, close to target 1) versus
was significantly different from all tones from 8 to 19 kHz (Tukey’s post hoc high-pitch tones (12 to 19 kHz, close to target 2) over sessions is shown. Middle:
multiple comparisons test). (D) Top: The occupancy gain for 5 kHz (target 1) ChR2 (cyan) were significantly larger than bootstrap after session 1 (P <1×
over sessions is shown. Middle: ChR2 (cyan) were significantly larger than 10 –5 for sessions 2 through 4). Bottom: YFP (black) were never significantly
–3
bootstrap from sessions 2 through 4 (session 2, P =1.2 × 10 ;session 3, larger than bootstrap. For (D) to (F), an asterisk indicates that the population
–5
–5
P <1× 10 ; session 4, P <1×10 ). Bottom:YFP (black) were never significantly average is significantly larger than the baseline bootstrap distribution.
Athalye et al., Science 359, 1024–1029 (2018) 2 March 2018 3of 6