TME nutrient stress impair SREBP1 and fatty acid synthesis in PDAC.

(A) Diagram of workflow for the transcriptomic comparison of mPDAC cells grown in TIFM or RPMI. Three pairs of cell lines (mPDAC1, mPDAC2 and mPDAC3) derived into TIFM and RPMI were analyzed with a sample size of n = 3 for each cell line. (B) Plotted normalized enrichment scores of metabolic gene sets from the C2 mSigDB database set. SREBP-related gene sets are highlighted in red. (C) Gene Set Enrichment Analysis (GSEA) of the Reactome regulation of cholesterol biosynthesis by SREBP/SREBF gene set in transcriptomic data of mPDAC cell lines cultured in TIFM or RPMI. (D) Row-scaled heatmap of leading edge genes from the GSEA analysis in (C). Z-scores calculated from trimmed mean of M values (TMM) gene counts for Reactome regulation of cholesterol biosynthesis by SREBP/SREBF genes in mPDAC cells cultured in TIFM versus RPMI. (E) Diagram of workflow for the transcriptomic comparison of mPDAC3 cells grown in RPMI (n = 3) or as orthotopic allograft murine tumors (n = 6). mPDAC3 cells from each condition were isolated by fluorescence-activated cell sorting (FACS) and RNA was isolated for transcriptomic analysis by next generation sequencing. (F) GSEA of the Reactome regulation of cholesterol biosynthesis by SREBP/SREBF gene set using transcriptomic data of mPDAC3 cell lines cultured in RPMI or orthotopically implanted in C57BL/6 mice. (G) Row-scaled heatmap of leading edge genes from the GSEA analysis in (F). Z-scores calculated from trimmed mean of M values (TMM) gene counts for Reactome regulation of cholesterol biosynthesis by SREBP/SREBF genes in mPDAC3 cells cultured in RPMI or isolated from orthotopic tumors. (H) SREBP signatures for human PDAC tumors and PDAC cancer cell lines generated using publicly available transcriptomic data using MERAV. Each point indicates a tumor specimen or cell line. (I) Row normalized heatmap z-scores of human transcriptomic data of genes from the Reactome regulation of cholesterol biosynthesis by SREBP/SREBF gene set. Genes shown are significantly differentially expressed between patient samples and PDAC cell lines with p-value cutoff of p<0.05. (J) Immunoblot analysis of full length and cleaved SREBP1 expression in mPDAC1 and mPDAC3 cell lines grown in TIFM or RPMI treated with 10 µM MG132. (K) Quantification of immunoblot signal of proteolytically processed SREBP1 divided by immunoblot signal of full length SREBP1. This analysis was based on triplicate analysis of each cell line in each condition. (L) Levels of unlabeled (grey) and deuterium labeled (blue, dark grey) palmitate (16:0) in mPDAC cell lines grown in TIFM or RPMI. Labeled palmitate indicates de novo synthesized lipid (n=3). Statistical significance was assessed using two-tailed Welch’s t-test in H, K, L. Statistical significance was assessed using the weighted Kolmogorov–Smirnov statistic in C, F.

TME arginine restriction limits SREBP1 expression and fatty acid synthesis in PDAC cells and tumors.

(A) Immunoblot analysis of full length and cleaved SREBP1 expression in mPDAC1 cells grown in TIFM, RPMI, or TIFM with supplemented arginine (TIFM + arg; 1.19 mM) treated with 10 µM MG132. (B) The fraction of cellular palmitate (16:0) that is deuterium labeled in mPDAC1 cell lines grown in TIFM, RPMI, TIFM + arg or RPMI with TIFM (2.27 µM) levels of arginine (n=3). (C) Immunoblot analysis of full length and mature SREBP1 in the same conditions as (B) and treated with 10 µM MG132. (D) Schematic for generation of orthotopic tumors in Lyz2-Cre and Arg1fl/fl host animals and analysis of tumor fatty acid synthesis and SREBP1 protein levels. (E) Concentrations of arginine in TIF and plasma isolated from animals bearing orthotopic mPDAC3-TIFM tumors. n=6 for Lyz2-Cre+/+; Arg1fl/fl animals and n=5 for Arg1fl/fl littermate controls. (F) (left) Representative immunofluorescent images of mPDAC3-TIFM tumors from Lyz2-Cre+/+; Arg1fl/fl (n =9) and Arg1fl/fl mice (n=9). Cancer cells are YFP labeled. (right) Quantification of mean SREBP1 intensity for YFP positive cancer cells in Lyz2-Cre+/+; Arg1fl/fl animals and Arg1fl/fl littermate controls. (n=3249 cells for each genotype). (G) The fraction of cellular palmitate (16:0) that is deuterium labeled from mPDAC3 cells isolated from tumors from Lyz2-Cre+/+; Arg1fl/fl (n =4) and Arg1fl/fl (n=5) mice. (H) The fraction of plasma palmitate (16:0) that is deuterium labeled from Lyz2-Cre+/+; Arg1fl/fl (n =6) and Arg1fl/fl (n=7) mice. Significance was assessed using the Mann-Whitney U-test for G. Two tailed Welch’s t-test correction was used to assess significance in B, E, F and H.

Arginine restriction prevents PDAC cells from maintaining lipid homeostasis when starved of exogenous lipids.

(A) PUFA/MUFA ratios of all cellular lipids and (B) major classes of lipids species in mPDAC1 cells cultured in TIFM, RPMI or TIFM + arg. (C) Triglyceride (TG) levels in mPDAC1 cells cultured in TIFM, RPMI or TIFM + arg. Plotted data in (C) is row normalized z-scores of total-signal normalized peak areas (n=3). (D) Immunoblot analysis of full length and cleaved SREBP1 expression in mPDAC1 cells grown in TIFM or RPMI with and without starvation of exogenous lipid. (E) Transcriptomic analysis of expression of SREBP1 target genes involved in fatty acid synthesis in mPDAC1 cells cultured with exogenous lipids (+ lipids) or without exogenous lipids (-lipids) in either TIFM or RPMI. Data plotted are log2 fold change of mRNA expression between cultures (n=3). (F) Immunoblot analysis of mPDAC1 full length and cleaved SREBP1 expression in mPDAC1 cells grown in TIFM, RPMI or TIFM + arg with and without exogenous lipid. (G) Transcriptomic analysis of expression of SREBP1 target genes involved in fatty acid synthesis in mPDAC1 cells cultured with exogenous lipids (+ lipids) or without exogenous lipids (-lipids) in either TIFM or TIFM + arg. Data plotted are log2 fold change of mRNA expression between cultures (n=3).(H) Levels of unlabeled and deuterium labeled palmitate (16:0) in mPDAC cell lines grown in TIFM or TIFM + arg (n=3). (I) Principal component analysis of LC-MS measurements of lipid levels of mPDAC1 cells cultured in TIFM or TIFM + arg with or without exogenous lipids (n=3). (J,K) Volcano plots depicting the log2 fold change in levels of lipids in mPDAC1 cells upon starvation of exogenous lipids when cultured in TIFM (J) or TIFM + arg (K). Lipids containing only SFA/MUFA acyl groups are highlighted in red. Lipids containing both an SFA or MUFA and PUFA acyl group are highlighted in blue. Lipids containing only PUFA acyl chains are shown in black. The cut off for statistical significance in this analysis is p<0.01. (L) Quantification of the percent of an indicate lipid class that is altered upon lipid deprivation in (upper) TIFM or (lower) TIFM + arg cultured mPDAC1 cells. (TG: triglycerides; SM: sphingomyelin; PE: phosphatidylethanolamine; PC: phosphatidylcholine; LPE: lysophosphatidylethanolamine; LPC: lysophosphatidylcholine; Hex1Cer: hex-1-ceramides; DG: diacyglycerol; ChE: cholesterol esters; Cer: ceramides) (M) Number of lipid species with indicated composition of acyl groups whose cellular abundance is significantly altered upon lipid withdrawal in (left) TIFM or (right) TIFM + arg cultured mPDAC1 cells. (N) Cellular proliferation rate of mPDAC1, 2 and 3 cell lines grown in TIFM or TIFM + arg upon lipid deprivation. Each point is plotted average of technical triplicates of the assay for each cell line. (O) Immunoblot analysis of SREBP1 expression in mPDAC1 cells transduced with control non-targeting or SREBP1 targeting sgRNAs. (O) Cellular proliferation of mPDAC1 cells analyzed in (P) cultured in TIFM + arg with and without exogenous lipids (n=3). Statistical significance was assessed in all panels by two-tailed Welch’s t-tests.

Arginine restriction sensitizes PDAC cells to exogenous PUFAs.

(A) Cellular peroxidation index of all cellular lipids and (B) major classes of lipids species in mPDAC1 cells cultured in TIFM, RPMI or TIFM + arg (n=3). (C) Cellular proliferation rate of mPDAC1-TIFM or mPDAC1-RPMI cells upon treatment with various PUFAs (n=3). (EPA: eicosapentanoic acid (250 µM); LNA: linolenic acid (250 µM); LLA: linoleic acid (250 µM); α-ESA: alpha-eleostearic acid (6.25 µM); DHA: docosahexaenoic acid (50 µM)). (D) Cellular proliferation rates of mPDAC1 cells cultured in TIFM or TIFM + arg treated with increasing doses of α-ESA with or without ferrostatin-1 (fer-1, 2 µM) (n = 3 per condition). (E) Cellular proliferation rates of mPDAC1 cells cultured in TIFM + arg when transduced with control non-targeting or SREBP1 targeting sgRNAs and treated with α-ESA (6.25 µM) (n=3) (F) Proliferation rates of mPDAC1 cells cultured in TIFM + arg when treated with the SREBP1 inhibitor fatostatin (2.5 µM) and α-ESA (6.25 µM) (n=3). (G) Proliferation rates of mPDAC1 cells cultured in TIFM when treated α-ESA at indicated concentrations with or without supplementation of oleate (MUFA) at 100 µM (n=3). (H) Principal component analysis of LC-MS measurements of lipid levels of mPDAC1 cells cultured in TIFM or TIFM + arg with α-ESA (50 µM) or vehicle (fatty acid free BSA) (n=3). (I) Distribution of 18:3 fatty acyl groups in different complex lipid species. (J) Heatmap of TG and PE species with 18:3 acyl chains in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Plotted values are Z-scores for each lipid species (n=3). (K,L) Heatmaps of MUFA-containing PE and TG species in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Plotted values are Z-scores for each lipid species (n=3). (M,N) Analysis of the number of significantly altered lipid species containing MUFA or 18:3 acyl groups in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Significance cutoff p<0.05. (O,P) Calculated cellular phospholipid peroxidation index (CPI) for TGs (O) and PEs (P) in mPDAC1-TIFM and mPDAC1-TIFM + arg cells treated with α-ESA (50 µM) (n=3). (Q) Cellular proliferation of mPDAC1 cells cultured in TIFM treated with indicated α-ESA concentrations and with or without 2.5 µM DGAT1 inhibitor (PF-04620110) and 2.5 µM DGAT1 inhibitor (PF-06424439) (n=3). Significance assessed in all panels using two-tailed Welch’s t-test.

Arginine deprivation sensitizes PDAC cells and tumors to PUFA enriched oils.

(A) Fatty acid content of tung and safflower oil measured by GC-MS fatty acid methyl ester (FAME) analysis (n=3 technical replicates). (B) Proliferation of mPDAC1 cells cultured in TIFM and TIFM + arg cell lines when treated with BSA-tung oil or fatty acid free BSA with or without ferrostatin-1 (fer-1, 2 µM) (n=3). (C) Diagram of workflow for treatment of mPDAC3-bearing mice with tung or safflower oil orally. (D) Log2 fold change of fatty acid methyl esters (FAMEs) in plasma of mice treated with tung or safflower oil (n=13 tung oil treated mice, n=12 safflower oil treated mice). (E) Tumor weights of mice treated with tung or safflower oil (n=11 safflower oil, n=13 tung oil). (F) Spearman’s correlation analysis of plasma α-ESA levels and PDAC tumor weights of mice treated with tung oil (n=13). (G) PUFA/MUFA ratios of PDAC tumors and other SREBP1 expressing (n=5 tung oil treated mice, n=5 safflower oil treated mice) (H) Diagram of treatment of mPDAC3-tumor bearing Lyz2-Cre+/+; Arg1fl/fl and Lyz2-Cre -/-; Arg1fl/fl littermate control mice with tung or safflower oil. (I) mPDAC3 orthotopic tumor weights from Lyz2-Cre+/+; Arg1fl/fl and Lyz2-Cr -/-; Arg1fl/fl mice post tung or safflower oil treatment (Arg1fl/fl/Safflower n = 12; Arg1fl/fl/Tung n = 12; Lyz2-Cre+/+; Arg1fl/fl/Safflower n = 10; Lyz2-Cre+/+; Arg1fl/fl/Tung n = 15). (J) Cellular peroxidation index (CPI) for TGs in mPDAC3 cells sorted from tumor bearing Lyz2-Cre+/+; Arg1fl/fl and Arg1fl/fl mice given tung or safflower oil (Arg1fl/fl /Safflower n = 5; Arg1fl/fl /Tung n = 6; Lyz2-Cre+/+; Arg1fl/fl /Safflower n = 7; Lyz2-Cre+/+; Arg1fl/fl/Tung n = 6). Statistical significance in E and I was assessed using the Mann-Whitney U test. Statistical significance was assessed using a two-tailed Welch’s t-test correction in D, G, I and J.

PUFA treatment synergizes with other ferroptosis inducers in arginine-restricted PDAC cells.

Proliferation of mPDAC1 cells cultured in RPMI or TIFM when treated with α-ESA (2 µM) of vehicle (fatty acid free BSA) in tandem with (A) deprivation of cystine to fraction present in base media as indicated, (B) Erastin at indicated concentrations, or (C) RSL3 at indicated concentrations. Ferrostatin-1 (fer-1, 2 µM) or vehicle (DMSO) was additionally added.

TME nutrients specifically inactivate SREBP1 and fatty acid synthesis in human PDAC.

(A) Proportion human PDAC and normal pancreas specimens from the Human Protein Atlas131 classified with low, mid, high, or non-detectable (ND) expression of SREBP1 target gene expression. (B) Read counts of fatty acid (SREBP1) or sterol (SREBP2) synthesis genes in TIFM and RPMI cultured mPDAC cells (two experiments plotted, n=3 each, n=6 total). (C) Immunoblot analysis of full length and cleaved SREBP1 expression in Aspc1 human cell lines grown in TIFM or RPMI. (D) The fraction of cellular palmitate (16:0) that is deuterium labeled indicated mPDAC cell lines in TIFM or RPMI (n=3). (E) Levels of unlabeled (grey) and deuterium labeled (blue, dark grey) palmitate (16:0) in human PDAC cell lines grown in TIFM or RPMI. Labeled palmitate indicates de novo synthesized lipid (n=3). (F) The fraction of cellular palmitate (16:0) that is deuterium labeled indicated human PDAC cell lines in TIFM or RPMI (n=3). Significance was tested in A using Fisher’s exact test. Significance was determined using a two-tailed Welch’s t-test for B, D, E and F.

Systematic analysis of TIFM content identifies arginine restriction in TIFM as a regulator of PDAC lipid metabolism.

(A) Concentrations of nutrients in TIFM or in RPMI. Each point represents an individual nutrient in RPMI and in TIFM. (B) Cell proliferation of mPDAC1 cell lines TIFM or RPMI with or without exogenous lipids (n=3). (C) Diagram of nutrient swapping strategy in TIFM. Nutrient pools (e.g. amino acids) are swapped between TIFM and RPMI to determine the nutrients that drive cellular phenotypes such as growth without exogenous lipids. (D) Flow chart of the results of nutrient swap experiments to identify TIFM nutrients that affect the ability of mPDAC1 cells to grow in the absence of exogenous lipids. Briefly, chemicals unique to TIFM when added to RPMI did not prevent mPDAC1 cells from growing in lipid depleted conditions, nor did removing these chemicals from TIFM improve mPDAC1 growth without lipids. This led us to conclude that alterations of levels of a shared nutrient between TIFM and RPMI was responsible for differences in lipid metabolism. We then observed that mPDAC1 cells grown in TIFM containing RPMI amino acid levels were able to grow without lipids, indicating that differences in amino acid availability altered lipid metabolism. We then found that supplementation of arginine in TIFM to RPMI levels alone was sufficient to improve mPDAC1 growth without lipids to levels nearly seen in RPMI (n=3).

SREBP1 immunofluorescent signal reflects the expression of SREBP1 in tissues.

(A) Murine kidney and muscle sections were stained with SREBP1 antibody as performed with murine PDAC samples in Figure 2. (B) Expression levels of SREBP1 mRNA in murine muscle vs kidney from reference 131.

TIFM cultured PDAC cells have more lipid droplets than PDAC cells in standard culture.

(A) Immunofluorescent analysis of YFP-expressing mPDAC3 stained with LipidTox to label lipid droplets. (B) Quantification of lipid droplets per cell area from (A). Significance was assessed in B using a two-tailed Welch’s t-test.

SREBP1 activation is impaired in TIFM-cultured PDAC cells.

Immunoblot analysis of full length and cleaved SREBP1 expression in (A) mPDAC2 and (B) mPDAC3 cells grown in TIFM or RPMI with and without exogenous lipids. (C) Fold change in expression of ELOVL6 mRNA, an SREBP1 target, in mPDAC1-TIFM and mPDAC1-RPMI cells after deprivation of exogenous lipids (n=3).

PDAC cells in RPMI can maintain lipid homeostasis upon starvation of exogenous lipids.

(A) Volcano plots depicting the log2 fold change in levels of lipids in mPDAC1 cells upon starvation of exogenous lipids when cultured in RPMI. Lipids containing only SFA/MUFA acyl groups are highlighted in red. Lipids containing both an SFA or MUFA and PUFA acyl group are highlighted in blue. Lipids containing only PUFA acyl chains are shown in black. The cut off for statistical significance in this analysis is p<0.01. (B) Quantification of the percent of an indicate lipid class that is altered upon lipid deprivation in RPMI cultured mPDAC1 cells. (C) Number of lipid species with indicated composition of acyl groups whose cellular abundance is significantly altered upon lipid withdrawal in RPMI cultured mPDAC1 cells.

TIFM-cultured cells are unable to overexpress mature SREBP1.

(A) Immunoblot analysis of cleaved SREBP1 expression in mPDAC1 cells cultured in TIFM or RPMI after transduction with cleaved SREBP1 cDNA.

Arginine starvation does not sensitive mPDAC cells to GPX4 or xCT inhibitors.

(A-F) Cellular proliferation rates of mPDAC1, mPDAC2, and mPDAC3 cells cultured in TIFM or TIFM + arg upon treatment with RSL3 and erastin at indicated concentrations GPX4 or erastin. Ferrostatin-1 (fer-1, 2 µM) or vehicle (DMSO) was additionally added.

Arginine starvation sensitizes murine and human PDAC cells to exogenous α-ESA.

Cellular proliferation rates of (A) mPDAC2, (B) mPDAC3 and (C) Aspc1 cells cultured in TIFM or TIFM + arg with indicated concentrations of α-ESA. Ferrostatin-1 (fer-1, 2 µM) or vehicle (DMSO) was additionally added.

α-ESA supplementation of RPMI-cultured PDAC cells has minimal effects on the cellular lipidome.

(A) Principal component analysis of LC-MS measurements of lipid levels of mPDAC1 cells cultured in RPMI with α-ESA (50 µM) or vehicle (fatty acid free BSA) (n=3). (B) Heatmap of TG and PE species with 18:3 acyl chains in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Plotted values are Z-scores for each lipid species (n=3). (C) Heatmaps of MUFA-containing PE and TG species in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Plotted values are Z-scores for each lipid species (n=3). (D) Analysis of the number of significantly altered lipid species containing MUFA or 18:3 acyl groups in mPDAC1 cells cultured as indicated upon treatment with α-ESA (50 µM). Significance cutoff p<0.05.

PUFA-rich oils alter the lipidomic profile of mPDAC1-TIFM cells in an arginine dependent manner.

(A) Principal component analysis of LC-MS measurements of lipid levels of mPDAC1 cells cultured in TIFM or TIFM + arg with BSA-tung oil (200 µg/mL) or vehicle (fatty acid free BSA) (n=3). (B) Distribution of 18:3 fatty acyl groups in different complex lipid species in BSA-tung oil (200 µg/mL) or vehicle (fatty acid free BSA) treated mPDAC1 cells cultured in TIFM or TIFM + arg. (C) Heatmap of TG and PE species with 18:3 acyl chains in mPDAC1 cells cultured as indicated upon treatment with BSA-tung oil (200 µg/mL) or vehicle (fatty acid free BSA). Plotted values are Z-scores for each lipid species (n=3). (D) Calculated cellular phospholipid peroxidation index (CPI) for TGs in TIFM and TIFM + arg cultured mPDAC1 cells treated with BSA-tung oil (200 µg/mL) or vehicle (fatty acid free BSA) (n=3). Significance assessed using two-tailed Welch’s t-test.

mPDAC1 cells are not sensitive to FSP1 inhibition.

(A) Proliferation of mPDAC1 cells treated with α-ESA (2 µM) of vehicle (fatty acid free BSA) in tandem with viFSP1 at indicated concentrations. Ferrostatin-1 (fer-1, 2 µM) or vehicle (DMSO) was additionally added.