Supplementary material
Detailed methods for conducting alcohol exposure in the home-cage
Alcohol exposure in the home-cage was initiated after single-housed rats attained
an average weight of 300 gm. On Monday, Wednesday and Friday of each week, rats
were weighed, and then a pre-weighed 100-ml graduated cylinder containing 15%
ethanol together with a pre-weighed 400-ml plastic bottle containing water were placed
onto the home-cage. Both containers were corked using rubber stoppers with metal sipper
tubes that contained ball bearings to prevent spillage. Sipper tubes extended into the
home-cage and provided rats with access to the liquids contained therein. After 24-hrs,
both bottles were removed and weighed. The water bottle was placed back onto the
home-cage and the ethanol-containing graduated cylinder was replaced with a preweighed 100-ml graduated cylinder containing water.
In order to control for spillage, a graduated cylinder and plastic bottle containing
ethanol and water, respectively, were placed on two empty home-cages and weighed
before and after each 24-hr session. Spillage was calculated as an average between the
two control cages, and subtracted from daily intake values. The left and right placement
of all bottles on the home-cage was alternated daily to reduce the impact or development
of impact of side preferences.
Ethanol intake was calculated by expressing the amount of ethanol consumed
(gm) as a function of body weight (kg). The distribution of ethanol in solution as a
function of its density was accounted for by multiplying the grams of ethanol consumed
by 0.1185.
Subjects with low ethanol intake values (g/kg ≤ 0.50, averaged across sessions 6
and 7 for experiment 1; sessions 13 and 14 for Exp. 2a) were given a sweetened ethanol
solution for 3 sessions in order to increase consumption. This solution was prepared by
dissolving sucrose (2%, w/v) in 15% ethanol. The different time-points for providing
access to sweetened ethanol were a function of experimenter error; typically access to
sweetened ethanol is restricted to sessions 6-7. The significant interactions between
Session and Vendor observed in Experiment 2 but not in Experiment 1 might be
attributable to these differences in when sweetened ethanol was made available. To
explore this possibility we analyzed ethanol intake (g/kg and ethanol preference)
averaged across the last 4 sessions of this phase as a function of Vendor (Harlan, Charles
River) and Experiment (Experiment 1, Experiment 2) and found no main effects of
Experiment or significant Vendor x Experiment interactions. This analysis indicates that
ethanol consumption was comparable at the end of both experiments, and suggests that
the time-point at which sweetened ethanol was made available had no significant impact
on the final outcomes achieved during this phase.
In Experiment 1, 15 rats (Charles River n=13; Harlan n=2) were given access to
the sweetened ethanol solution. In experiment 2a, 12 rats (Charles River n=8; Harlan
n=4) required the sweetened ethanol solution. It is notable that in both experiments more
Charles River rats required access to sweetened ethanol, compared to Harlan rats.
a
Experiment 1
80
b
Experiment 2a
80
60
Total Port Entries
Total Port Entries
Charles River
Harlan
40
20
60
40
20
0
0
1
2
3
Session
4
5
1
2
3
4
5
Session
Supplementary Fig. 1 Mean (± SEM) total port-entries made across 5 sessions of
exposure to a second, non-alcohol context in Charles River (filled symbols) and Harlan
(open symbols) rats in a Experiment 1 and b Experiment 2a. During each session rats
were exposed to a context that differed from the context in which they had previously
received Pavlovian discrimination training. Neither cues nor ethanol were delivered
during these sessions. In Experiment 1 there was a trend for total port-entries to decrease
across session [Session, F(4,112)=3.06, p=0.058), with no differences as a function of
vendor, or Session x Vendor interaction. In Experiment 2 total port-entries decreased
across session [Session, F(4,108)=6.93, p<0.001], with no main effect of Vendor or
Session x Vendor interaction.
a
60
b
Charles River
Baseline
Test
Harlan
50
Total Port Entries
50
Total Port Entries
60
40
30
20
40
30
20
10
10
0
0
0
3.33
10
SCH 23390 (µg/kg)
0
3.33
10
SCH 23390 (µg/kg)
Supplementary Fig. 2 Mean (± SEM) total port-entries averaged across the last 2
sessions of exposure to the non-alcohol context (open bars) and at test (filled bars)
following saline or SCH 23390 (3.33 μg/kg or 10 μg/kg) injections in a Charles River
and b Harlan rats. ANOVA conducted on data from test sessions revealed that rats from
Harlan made fewer total port-entries than rats from Charles River [Vendor, F(1,28)=6.26,
p<0.05]. When analyzed separately, there was a main effect of Dose in rats from each
vendor [Charles River, F(2,30)=4.45, p<0.05; Harlan, F(2,26)=8.38, p<0.01]. Specific
dose comparisons within each vendor were made using paired-samples t-tests with a
modified Bonferroni correction to control for multiple comparisons (α=0.03). In Charles
River rats, total port-entries were significantly lower following injection with
10 μg/kg of SCH 23390, compared to saline [t(15)=2.96, p=0.010] and the 3.33 μg/kg
dose [t(15)=2.69, p=0.017]. There was no difference between saline and 3.33 μg/kg of
SCH 23390. In Harlan rats, total port-entries were significantly reduced following 10
μg/kg of SCH 23390 [t(13)=4.76, p<0.001] compared to saline. There was no difference
between saline and 3.33 μg/kg or the 3.33 μg/kg and 10 μg/kg doses. Within each vendor,
baseline responding in the non-alcohol context was compared to total port-entries at test
following 10 μg/kg of SCH 23390. These comparisons were not different for either
Charles River or Harlan rats suggesting that the high dose of SCH 23390 did not reduce
total port-entries relative to pre-treatment levels.
80
b
Charles River
80
70
70
60
60
Total Port Entries
Total Port Entries
a
50
40
30
20
10
Baseline
Test
Harlan
50
40
30
20
10
0
0
0
5
Eticlopride (µg/kg)
10
0
5
10
Eticlopride (µg/kg)
Supplementary Fig. 3 Mean (± SEM) total port-entries averaged across the last 2
sessions of exposure to the non-alcohol context (open bars) and at test (filled bars)
following saline or eticlopride (5 μg/kg or 10 μg/kg) infusions in a Charles River and b
Harlan rats. ANOVA comparisons on test data revealed no main effects of Dose or
Vendor and no Dose x Vendor interaction. Within each vendor, paired samples t-tests
were conducted to compare total port-entries at test following 10 μg/kg of eticlopride,
relative to baseline responding in the non-alcohol context. Charles River rats responded
significantly more at test relative to baseline [t(15)=-3.47, p<0.01], whereas this
comparison approached statistical significance for Harlan rats [t(12)=-2.02, p=0.066]