Supplementary Table 1. Only Kcnj9 shows differential expression consistent with the phenotype in both congenic (R4)/background (B6) and wildtype/Kcnj9 mutant comparisons.

High/Low Withdrawal Genotype TaqMan Probe (mRNA expression ratio)

R4 congenic/B6 WT/Kcnj9-/-

Wdr42a Mm00463652_m1 0.86 (p =1.5x10-4) 1.18 (NS) Pea15a Mm00440716_g1 1.02 (NS)

Casq1 Mm00486725_m1 5.16 (p =8.0x10-13) 0.93 (NS)

Atp1a4 Mm01290854_g1 20.67 (p =4.0x10-16) 0.88 (NS)

Igsf8 Mm00712984_m1 0.84 (p =5.3x10-4) 1.19 (NS) Atp1a2 Mm00617899_m1 1.00 (NS) 0.98(NS)

Atp1a2 custom 4.13 (p =1.3x10-20) 1.09 (NS)

WT/Kcnj9+/- Kcnj9* custom 1.17 (p =3.4x10-5) 1.96 (p =6.5x10-7)

Kcnj10 Mm00445028_m1 1.31 (p =1.7x10-6) 1.02 (NS) Pigm Mm00452712_s1 1.07 (NS) 1.01 (NS)

Slamf9 Mm00504048_m1 0.93 (NS)

Igsf9 Mm00459672_m1 1.33 (p =1.0x10-5) 0.74 (NS) Tagln2 Mm00724259_m1 0.97 (NS)

Ccdc19 Mm01283396_g1 0.93 (NS)

Vsig8 Mm00624907_m1 6.07 (p =1.0x10-3) 0.19 (p =3.3x10-4)

In null mutant models, there is typical carryover of the embryonic stem (ES) cell-derived genotype (i.e., 129Sv/J in the Kcnj9-/- mice). In order to assess the potential impact of the other confirmed quantitative trait gene (QTG) candidates on the Kcnj9-/- phenotype (reduced withdrawal severity), we compared their expression in Kcnj9 mutant and wildtype (WT) littermates (whole brain). We also tested Pigm, which did not differ in expression between 1 congenic (R4) and background strain (B6) mice, but is one of six (Atp1a4- Atp1a2-Igsf8-Kcnj9-Kcnj10-Pigm) implicated in CNS excitability (Ferraro et al 2007). Data are for N=11-12 adults males in the R4 congenic/B6 comparison, and N=5-8 adult males per group in the WT/Kcnj9 mutant comparisons. Note, because there was no detectable Kcnj9 expression in the null mutants, the expression ratio shown for Kcnj9 is for WT/heterozygote littermates; the remaining expression ratios in this column are for WT/Kcnj9-/- littermates. p<0.0045 is significant after correction for multiple comparisons (i.e., p<0.05/11 Taqman probes tested for knockout analyses). As expected, Kcnj9 expression in WT littermates was significantly higher (and approximately twice) than in Kcnj9-/+ heterozygotes. Only one other gene (Vsig8) differed in expression between Kcnj9-/- and wildtype littermates. *However, only Kcnj9 shows a significant mRNA expression ratio concordant with withdrawal phenotype in both congenic (R4)/background (B6) and wildtype/Kcnj9 mutant comparisons (i.e., higher expression associated with more severe sedative-hypnotic withdrawal). Notably, Kcnj10, Atp1a2, and Atp1a4, which showed marked differential expression in whole brain between R4 congenic and B6 background strain mice did not differ in expression between Kcnj9-/- and wildtype littermates (expression ratios = 0.98, 0.92 and 1.13, respectively; all not significant [NS]). This mitigates the potential confounding effect of linked ES cell-derived genes on the Kcnj9-/- phenotype and supports a causal role for Kcnj9. Furthermore, only one nonsynonymous polymorphism was found in both 129Sv/J vs. D2 and B6 vs. D2 comparisons (rs31557967 in Igsf9, R284G).

Supplement Fig 1. Development of D2.B6 congenic mice with a QTL interval from the B6 donor strain introgressed onto the D2 genetic background

D2 (background) B6 (donor) 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 X

Step 1

11 12 13 14 15 16 17 18 19 X 11 12 13 14 15 16 17 18 19 X F1 1 2 3 4 5 6 7 8 9 10 Step 1. D2 and B6 progenitor strain mice were crossed to produce F1 (D2B6) progeny.

11 12 13 14 15 16 17 18 19 X

Step 2. Male F1 mice were backcrossed to D2 strain mice. Step 2 Progeny were genotyped at ~80 markers across the genome.

Example backcross progeny (each of which is unique) 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

11 1213141516171819 X 11 1213141516171819 X 11 1213141516171819 X

Step 3. Male progeny that retain the QTL interval from the donor strain Step 3 are selected as breeders for the next generation. This process was (repeat in Steps 4-10) repeated in Steps 4-10. After 6 generations of backcrossing, the animals were estimated to be >98% D2 except within the introgressed interval.

Step 11. A final intercross was performed to generate the finished Step 11 congenic strain (donor interval homozygotes).

Finished D2.B6 congenic strain 1 2 3 4 5 6 7 8 9 10 QTL interval

11 12 13 14 15 16 17 18 19 X Supplement Fig 2. Development of interval-specific congenic strains with defined congenic intervals (spanning a chromosome 1 QTL) from the D2 donor introgressed onto the B6 background

B6 B6.D2 (Mtv) 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 X

Step 1

111213141516171819 X 11 12 13 14 15 16 17 18 19 X

F1 (B6.D2 x B6) progeny 1 2 3 4 5 6 7 8 9 10 Step 1. B6.D2 congenic mice were backcrossed to B6 mice to yield F1 (B6.D2 x B6) mice.

11 12 13 14 15 16 17 18 19 X

Step 2. F1 mice were backcrossed to B6 strain mice. Step 2 Individual progeny were genotyped using informative D1Mit and SNP markers within or flanking the congenic interval to identify recombinant mice.

At this point animals recombinants are unique and heterozygous

Step 3. Recombinant mice (founders) were backcrossed Step 3 to B6 mice to expand the genotype (i.e., produce multiple animals with the same recombinant genotype).

Step 4. A final intercross was carried out to generate Step 4 donor region homozygotes (i.e., the finished interval specific congenic strains).

Please see Fig 2A for the recombinations captured in the finished R4, R6, R6, R8, R9 and R12 interval specific congenic strains (donor region homozygotes) Supplement Fig 3. Kcnj9 null mutant on a D2 genetic background D2 Kcnj9 knockout (B6 background) 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Kcnj9 null X Step 1 11 12 13 14 15 16 17 18 19 X 11 12 13 14 15 16 17 18 19 X F1 1 2 3 4 5 6 7 8 9 10 Step 1. Kcnj9 knockout mice developed by Dr. Wickman (B6 background; Torricelli et al., 2002) were crossed to D2 strain mice to produce F1 mice heterozygous for the null mutation (D2B6-Kcnj9 -/-).

11 12 13 14 15 16 17 18 19 X

Step 2. Male F1 mice were backcrossed to D2 strain mice. A PCR- based assay was used to differentiate hemizygotes (-/+) from wildtype Step 2 (+/+) littermates. Hemizygotes were genotyped at ~80 markers across the genome. The three males with the greatest D2 representation were selected as breeders for the next generation and backcrossed to D2 strain mice. This process was repeated in Steps 3-8.

Example backcross progeny (each of which is unique) 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

11 12 13 14 15 16 17 18 19 X 11 12 13 14 15 16 17 18 19 X

Steps 3-8. After 6 generations of backcrossing, the animals were Steps 3-8 estimated to be >98% D2 except within the introgressed interval.

Step 9 Step 9. Final intercross to generate the finished null mutant model.

Finished Kcnj9 null mutant (D2 background) 1 2 3 4 5 6 7 8 9 10

11 12 13 14 15 16 17 18 19 X