*4.13. Immunohistochemistry Assay*

Liver tissue samples of MCDD- and HFD-fed mice and control mice were collected. Then, the samples were fixed in 4% paraformaldehyde for 24 h. The immunohistochemistry assay was entrusted by Servicebio (Wuhan, China). The detailed processes of this experiment can be referred to in our previous study [12].

#### *4.14. Isolation of Cytosolic and Mitochondrial Fractions*

The cell mitochondria isolation kit (#C3601, Beyotime Biotechnology, Nanjing, China) was utilized to isolate the cytosolic and mitochondrial fractions. Briefly, wash the cells with cold PBS and harvest the cells by trypsin-EDTA solution. Re-wash the cells two times and collect the cells by centrifuge, and then remove the supernatant. Add 1 mL mitochondrial isolation regen<sup>t</sup> (with 1 mM PMSF) and resuspend the cells, and then incubate the suspension in an ice bath. Then the cell suspension was transferred to a glass homogenizer of appropriate size, and the homogenate was about 10–30 times. Centrifuge the cell homogenate at 600× *g*, 4 ◦C for 10 min. Then carefully transfer the supernatant to another centrifugal tube and centrifuge for 10 min at 11,000× *g*, 4 ◦C. The precipitation was the isolated mitochondria. The supernatant collected was then centrifuged for 10 min at 12,000× *g*, 4 ◦C. The supernatant was the cytoplasmic protein without mitochondria.

#### *4.15. Fluorescence Image Analysis*

The ImageJ software was utilized to analyze the number of mitochondria interacting LDs. Briefly, the fluorescence intensity was analyzed along with the dotted line (Figure 4G). If the LD is contacting mitochondria, the signal of Dsred (mitochondria) can be detected between or overlap the signal of BODIPY493/503 (lipid droplet). If the LD does not contact mitochondria, the signal of Dsred (mitochondria) is supposed to be losing between the signal of BODIPY493/503 (lipid droplet).

#### *4.16. Survival Analysis and Normal*/*Cancer Gene Expression Comparison Analysis*

The survival predication was performed by the GEPIA database (http://gepia.cancer-pku.cn/). The prognosis analysis and gene expression analysis were performed according to the construction of the creator of this database [59].
