Legends for Supplementary Figures
Supplementary Figure 1. Active lever presses during the five days of 2 h cue extinction training
for rats that had previously been given aCSF or PEPA into the RMTg during the modified (15
min) non-cue extinction sessions. Note that no rats received microinjections during the 2 h cued
extinction sessions. A, Active lever presses during the standard (2 h) cue extinction sessions for
the cocaine experiment for rats that had previously received PEPA or aCSF during the modified
non-cue extinction sessions. There was no effect of prior treatment on the lever pressing during
cue extinction. A two-way repeated-measures ANOVA found no effect of group (F(1,10) = 0.5440,
p > 0.05), an effect of time (F(4,40) = 3.368, p < 0.01), and no interaction (F(4,40) = 0.3649, p >
0.05). B, Active lever presses during the standard (2 h) cue extinction sessions for the food
experiment for rats that had previously received PEPA or aCSF during the modified non-cue
extinction sessions. There was no effect of prior treatment on the lever pressing during the cue
extinction. A two-way repeated-measures ANOVA found no effect of group (F(1,15) = 0.1382, p >
0.05), no effect of time (F(4,60) = 1.379, p > 0.05), and no interaction (F(4,60) = 1.262, p > 0.05).
The large variability on day 1 for the aCSF group appears to be due to an outlier in the aCSF
group with a large number of lever presses (489).
Supplementary Figure 2. Extinction learning for cocaine-seeking in rats with microinjections
outside the RMTg. To determine the degree of site specificity for the microinjections, a separate
analysis of the cocaine-seeking data from animals with misplaced cannulae was conducted. A,
Active lever presses (mean ± SEM) during extinction sessions for those rats receiving pre- and
post-training microinjections of PEPA or aCSF for all rats with misplaced cannulae. Days 1-5
were modified (15 min) non-cue extinction sessions with microinjections given immediately
before or after each session, whereas days 6-12 were 2 h in length with no microinjections. A
two-way repeated-measures ANOVA found no significant effect of group (F(1,13) = 0.4157, p >
0.05), a significant effect of time (F(4,52) = 14.12, p < 0.01), and no significant interaction (F(4,52)
= 0.8895, p > 0.05). For the standard (2 h) non-cue extinction sessions on days 6-12, a two-way
repeated-measures ANOVA revealed no significant effect of group (F(1,13) = 0.7112, p > 0.05), a
significant effect of time (F(6,78) = 3.418, p < 0.01), and no significant interaction (F(6,78) =
0.8126, p > 0.05).There was no difference in rate of extinction learning on days 1-5 or retention
of extinction learning on days 6-12. B, Because there were a sufficient number of rats with
posterior misplacements for analysis, we also analyzed that separately and, again, found no effect
of misplacement. Panel B shows the active lever presses (mean ± SEM) for those rats with
injection sites posterior to the RMTg Days 1-5 were 15 min in length with microinjections given
immediately before or after each session, whereas days 6-12 were 2 h in length with no
microinjections. There was no effect of misplaced PEPA microinjections on the extinction
learning or retention for cocaine-seeking. A two-way repeated-measures ANOVA found no
significant effect of group (F(1,5) = 0.2208, p > 0.05), a significant effect of time (F(4,20) = 7.261, p
< 0.01), and no significant interaction (F(4,20) = 2.120, p > 0.05). For the 2 h extinction sessions
on days 6-12, a two-way repeated-measures ANOVA revealed no significant effect of group
(F(1,5) = 0.03545, p > 0.05), no significant effect of time (F(6,30) = 2.211, p > 0.05), and no
significant interaction (F(6,30) = 1.031, p > 0.05).
Supplementary Figure 3. Active and inactive lever pressing during cue-induced reinstatement
tests in rats with BM microinjections outside the RMTg. The analysis was conducted for all
misplacement combined (panels A and B). In addition, due to the within-subjects design, we also
examined the data for each “direction” of misplacement (all remaining panels). BM
microinjections increased lever pressing regardless of the direction of misplacement, relative to
the RMTg, for the microinjection. A and B, Combined active and inactive lever presses
(mean ± SEM), respectively, during cue-induced reinstatement tests for those rats receiving
intra-RMTg microinjections of aCSF or BM regardless of direction of placement error. One-way
ANOVAs of the active and inactive lever presses revealed significant effects (F(2, 39) = 20.47, p <
0.01; and F(2, 39) = 13.98, p < 0.01, respectively). Post-hoc tests in both cases revealed that those
rats that had received BM microinjections had significantly more lever presses during the cueinduced reinstatement test compared to the extinction baseline and those rats that had received
aCSF (p < 0.01 in all cases). C and D, Active and inactive lever presses, respectively, for the rat
with an injection site anterior to the RMTg. E and F, Active and inactive lever presses
(mean ± SEM), respectively, for those rats with injection sites posterior to the RMTg. One-way
ANOVAs of the active and inactive lever presses revealed significant effects (F(2, 15) = 11.06, p <
0.01; and F(2, 15) = 31.90, p < 0.01, respectively). Post-hoc tests in both cases revealed that those
rats that had received BM microinjections had significantly more lever presses during the cueinduced reinstatement test compared to the extinction baseline and those rats that had received
aCSF (p < 0.01 in all cases). G and H, Active and inactive lever presses (mean ± SEM) for those
rats with injection sites with a medial-lateral shift. A one-way ANOVA of the active lever
presses revealed a significant effect (F(2, 3) = 72.10, p < 0.01). Post-hoc tests revealed that those
rats that had received BM microinjections had significantly more active lever presses during the
cue-induced reinstatement test compared to the extinction baseline and those rats that had
received aCSF (p < 0.01 in both cases). A one-way ANOVA of the inactive lever presses was not
significant (F(2, 3) = 0.5309, p > 0.05). I and J, Active and inactive lever presses (mean ± SEM),
respectively, for those rats with injection sites dorsal to the RMTg. A one-way ANOVA of the
active lever presses revealed a significant effect (F(2, 9) = 5.738, p < 0.05). Post-hoc tests revealed
that those rats that had received BM microinjections had significantly more lever presses during
the cue-induced reinstatement test compared to the extinction baseline (p < 0.05). A one-way
ANOVA of the inactive lever presses showed a trend toward a significant effect (F(2, 9) 3.744, p =
0.065). K and L, Active and inactive lever presses for the rat with an injection site ventral to the
RMTg. * p < 0.01 compared to extinction baseline and the aCSF group. # p < 0.05 compared to
extinction baseline.
Supplementary Figure 4. Diagrams showing estimated microinjection termination sites for
incorrect placements, broken down according to the relative direction of missed placement used
in Figure S3. Figures adapted from Paxinos and Watson (2007) and A/P coordinates (in mm) are
given relative to Bregma. It is critical to note that these are estimates and that there is less
confidence regarding the certainty of these termination sites vis-à-vis those of rats whose
termination sites were found to be within the RMTg. In addition, termination sites were
categorized, as in Supplementary Figure 3, based on the primary dimension by the sites were
incorrect. A, Termination site found to be too anterior (n = 1). B, Termination sites found to be
too posterior (n = 6). C, Termination sites found to be shifted too far to the right (n = 2). D,
Termination sites found to be too dorsal (n = 4). E, Termination site found to be too ventral (n =
1).
Supplementary Figure 5. A, Active lever presses for rats grouped based on whether their
missed placements were estimated to be ~0.5 mm or less (proximal) from the border of the
RMTg or greater than 0.5 mm (distal) from the border of the RMTg. A one-way repeatedmeasures ANOVA for the active lever presses for “proximal” rats found a significant effect (F(2,
14)
= 15.14, p < 0.001). Post hoc analysis revealed that those rats receiving BM had significantly
more active lever presses compared to their extinction baseline and aCSF-control group (p <
0.05). A one-way repeated-measures ANOVA for the active lever presses for “distal” rats
revealed a significant effect (F(2, 10) = 5.574, p < 0.05). Post hoc analysis revealed that those rats
receiving BM had significantly more active lever presses compared to the extinction baseline (p
< 0.05) and a trend toward more active lever presses compared to the aCSF control group (p <
0.09). B, Inactive lever presses for rats grouped based on whether their missed placements were
proximal or distal from the border of the RMTg. A one-way repeated-measures ANOVA for the
inactive lever presses for “proximal” rats found a significant effect (F(2, 14) = 34.39, p < 0.001).
Post hoc analysis revealed that those rats receiving BM had significantly more inactive lever
presses compared to their extinction baseline and aCSF-control group (p < 0.05). A one-way
repeated-measures ANOVA for the inactive lever presses for “distal” rats revealed a trend
toward a significant effect (F(2, 10) = 34.39, p < 0.07). Post hoc analysis of the distal rats revealed
no significant effects or trends (p > 0.1 in all cases). C and D, Diagrams showing estimated
microinjection termination sites for incorrect placements, broken down according to the relative
missed placements in panels A and B. Figures adapted from Paxinos and Watson (2007) and A/P
coordinates (in mm) are given relative to Bregma.
* p < 0.01 compared to extinction baseline and the aCSF group. # p < 0.05 compared to
extinction baseline. @ p < 0.09 compared to aCSF group.