Download Supplementary Figure 1. Generation of IDH2 mutant mice. (A

Supplementary Figure 1. Generation of IDH2 mutant mice. (A) Targeting of the idh2 locus.
The targeting construct was used to target the IDH2 gene in the embryonic stem cell line.
Homologous recombination led to the deletion of a 1.17-kb genomic region containing the
presumed idh2 promoter and ATG start codon. The targeted allele is detected by a probe as
shown, with the expected DNA fragment sizes indicated. Primers from the 3’ end of the Neo
gene and the 3’ end to the 2.5-kb arm of homology yield a PCR product of 0.7 kb from the
recombinant gene. Neo, neomycin resistance gene; a, wild-type sense primer; b, wild-type
antisense primer; c, knock-out sense primer; d, knock-out antisense primer. (B) Genotyping
of idh2 alleles by RT-PCR. Results are shown for WT and IDH2-/- and were obtained using
the PCR primers described previously (6).
Supplementary Figure 2. Physiological effects of IDH2 deficiency on a normal diet. (A)
Organ weight/body weight of major organs from 10-month-old wild-type and IDH2-/- mice
on a normal diet (n = 10). N.S. indicates no significant difference compared with wild-type
mice. Male mice were used for all experiments. (B) Oil Red O-stained sections of visceral
WAT of wild-type and IDH2-/- mice at 10 months of age. Bars in histological sections indicate
50 m.
1
Supplementary Figure 3. Resistance to high-fat diet induced obesity in IDH2-/- mice. (A)
Comparison of daily food intake between high-fat fed wild-type and IDH2-/- mice at 6 months
of age (n = 7). Consumption of food was measured for 7 d consecutively. N.S. indicates no
significant difference compared with wild-type mice. Male mice were used for all
experiments. (B) Gross image of visceral WAT from 6-month-old wild-type and IDH2-/-mice
after 4 weeks of a high-fat diet.
2
Supplementary Figure 4. Reduced lipogenesis and stimulation of thermogenesis in adipose
tissues of IDH2-deficient mice on a normal diet. (A) Quantifications of lipogenesis signal in
sections of visceral WAT from wild-type and IDH2-/- mice. Quantifications of the levels of
p53 and lipogenic protein expression (B), UCP1 and p38 MAPK activation (C), and PGC-1
(D) in visceral WAT of wild-type and IDH2-/- mice. (E) RT-PCR analysis of marker genes in
brown-like adipocytes in visceral WAT of wild-type and IDH2-/- mice. The figure shows
representative data for three independent experiments. The oligonucleotides used for PCR are
listed in Table S1. (F) Quantifications of proteins associated with thermogenesis in sections
of BAT from wild-type and IDH2-/- mice. (G) Quantifications of the levels of proteins related
to p38-mediated UCP1 expression in BAT from wild-type and IDH2-/- mice. Ten images from
randomly selected visual fields were captured from each slide for quantification of
immunohistochemistry. For quantification of immunoblot, the protein levels were normalized
to the actin level. Each value represents the mean ± SD from three independent experiments.
*P < 0.01, between the two genotypes indicated. N.S. indicates no significant difference
compared with wild-type mice. 10-month-old male mice were used for all experiments.
3
Supplementary Figure 4. Mitochondrial contents in adipose tissues of wild-type and IDH2-/mice on a normal diet. Immunohistochemical staining of a mitochondrial marker, VDAC in
visceral WAT of wild-type and IDH2-/-mice. Quantitative results of expression levels of the
mitochondrial marker are shown as a percentage of control. N.S. indicates no significant
difference compared with wild-type mice. 10-month-old male mice were used for the
experiments.
4
5