Fig 2: Gata6 progenitors contribute to SG morphogenesis Bright‐field images of whole‐mount WT and Gata6 cKO epidermis, showing abnormal SG upon loss of Gata6. Quantification of hypotrophic SG in WT, heterozygous (Het), and Gata6 cKO epidermal whole mounts. Data are boxplots with indication of means. Box limits are minimum and maximum values. An average of 165 HF per mouse (from 3 to 4 mice per genotype) was analyzed. *P < 0.05; **P < 0.005; unpaired Student's t‐test.Schematic representation of plasmid constructs used for in utero lentiviral infection. Whole mounts or sections of adult infected epidermis with empty vector (EV) or Gata6‐ires‐GFP (G6OE) lentivirus were stained for GFP and Fasn. In vivo overexpression of Gata6 leads to ectopic Fasn expression in the HF/SG unit. White dotted lines define SG, and yellow dotted lines define a cyst. Note that the cyst is mostly negative for Fasn in agreement with its SD‐like phenotype. White arrows indicate GFP‐positive infected cells. These cells are stained with Fasn only on G6OE expression. H&E‐stained skin from G6OE mice shows a cyst with SG elements in the HF unit (black arrows). Staining for Gata6 (both endogenous and exogenous) shows that Gata6 expression occurs in a limited number of cells (representative image in upper right panel). Bottom left graph shows quantification of the percentage of clones labeled for both Fasn and GFP in the SG, HF, and IFE compartments. Data are means ± SD and originate from 4 EV mice and 8 G6OE mice (average of 11 clones per mouse).Whole‐mount adult epidermis infected with EV or G6OE lentivirus stained with LipidTOX. Ectopic Gata6 expressing cells are not stained with LipidTOX, indicating incomplete sebaceous maturation.Lineage tracing experiments in Gata6EGFPCreERT2:Rosa26‐fl/STOP/fl‐tdTomato (Gata6creER ROSA‐dTom) mice. A single dose of 4OHT was injected into pregnant females at E16.5. Tail skin from pups was collected at P13. Representative example of whole‐mount epidermis showing tdTomato‐labeled cells counterstained with Dapi (top left panel). Right panels show the different Z‐stacks related to this whole mount. Gata6 progeny are mainly found in the upper SG/JZ. Localization of Gata6 progeny (dTomato+) was quantified in 20–26 pilosebaceous units per mouse (N = 4 mice) (bottom left panel). Data are means ± SD. As a control, quantification was also performed on Rosa26‐fl/STOP/fl‐tdTomato (ROSA‐dTom) mice that were treated similarly to Gata6creER ROSA‐dTomato mice.Data information: (A–C) Scale bars: 50 μm. (D) Scale bar: 25 μm.

Fig 3: The expression level of FASN affects the fatty acid content of breast cancer cells. (A-D) FASN expression was significantly decreased following cerulenin and siRNA treatment; in addition, FASN was overexpressed in SK-Br-3 cells (FASN-OE SK-Br-3 cells). Bars represent the means ± SEM (n=3); **P<0.01, ****P<0.0001 vs. control. (E) Heatmap showing the changes in fatty acids in cells following FASN knockdown and overexpression. (F) FASN was inhibited by cerulenin, and the content of 5 fatty acids in cells was decreased; columns, mean (n=3); bars, ± SEM; *P<0.05, **P<0.01 vs. DMSO control. (G) After FASN was silenced by siRNA, the content of 7 fatty acids in cells was decreased; columns, mean (n=3); bars, ± SEM; *P<0.05, **P<0.01, ***P<0.001 vs. negative control. (H) After FASN was overexpressed, the content of 8 fatty acids in cells was increased; columns, mean (n=3); bars, ± SEM; *P<0.05, **P<0.01, ***P<0.001 vs. negative control. (I) Heatmap showing the changes in extracellular fatty acids following FASN knockdown and overexpression. (J) FASN was inhibited by cerulenin and siRNA, and the content of 4 FFAs in the cell medium was reduced; columns, mean (n=3); bars, ± SEM; *P<0.05, **P<0.01 vs. control. (K) After FASN was overexpressed, the content of 3 types of FFAs in the cell medium was increased; columns, mean (n=3); bars, ± SEM; *P<0.05 vs. negative control. FASN, fatty acid synthase; FFAs, free fatty acids.

Fig 4: ROS is induced by FAS inhibition but not responsible for apoptosis(A and B) Cerulenin and orlistat induced ROS generation in MCF-7 and MDA-MB-231 cells. ROS levels were measured after cells were treated with FAS inhibitors as indicated for 2 h. (C and D) Effects of trolox, BHA and NAC on ROS generation in MCF7 and MDA-MB-231 cells induced by cerulenin and orlistat. ROS levels were measured after cells were treated with 10 μM of cerulenin or 100 μg/ml of orlistat in combination with 2 mM of trolox, 10 mM of NAC or 100 μM of BHA for 2 h. (E and F) Effects of trolox, BHA and NAC on apoptosis in MCF-7 and MDA-MB-231 cells induced by cerulenin and orlistat. Apoptosis was measured after cells were treated with 10 μM of cerulenin or 100 μg/ml of orlistat for 48 h. Error bar indicates ± SE (n = 3).*p < 0.05; **p < 0.01 (t-test).

Fig 5: SFN inhibited microtubule-mediated FA synthesis via activating 26S proteasome.A Both A549 and SK-1 cells were treated with 0, 10, 20, and 30 μM SFN for 24 h. Western blotting was used to analyze the accumulation of SREBP1, FASN, ACACA, and ACLY. B Both A549 and SK-1 cells were pretreated with PS-341 (10 μM) for 2 h, then treated with SFN (20 μM) for 24 h. The accumulation of SREBP1, FASN, ACACA, and ACLY was analyzed by western blotting. C After knockdown of SREBP1, western blotting was used to analyze ACACA, FASN, and ACLY accumulation. D GEPIA database was used to analyze the correlation between transcription factor SREBP1 and FASN. If R closer is to “1,” the correlation between the two proteins is higher. Via the String net, we analyzed the interactions among FASN and related partners. E After induction with 20 µM SFN for 24 h, coimmunoprecipitation was employed to detect the interaction between α-tubulin and FASN, ACACA, and ACLY. F Immunofluorescence and confocal microscopy were employed to observe the colocalization of α-tubulin to FASN (scale bar: A549: 75 µm, SK-1: 25 µm), ACACA (scale bar: 25 µm), ACLY (scale bar: 25 µm). G After knockdown of β-tubulin, western blotting was used to analyze ACACA, FASN, and ACLY accumulation. H Determination of total free FA content after induction with 20 µM SFN for 24 h. I After induction with 20 µM SFN for 24 h, mitochondrial phospholipid content was detected by kit. Data were shown as means ± SEM (n ≥ 3). *P ≤ 0.01, **P ≤ 0.001, ***P ≤ 0.0001.