(A) 4h fasting plasma triglyceride levels during the course of the chronic LFD or HFD intervention (Study 1). Differences between LFD and HFD within each genotype were not statistically significant. n=10 per group. (B) Bodyweight changes in WT and Angptl4−/− mice fed LFD or HFD for 19 weeks. (C) Mean daily food intake in WT and Angptl4−/− mice fed LFD or HFD for 19 weeks. (D) Whole animal photographs taken immediately following sacrifice of representative WT and Angptl4−/− mice fed either LFD or HFD for 19 weeks. Pictures on the far right were taken after removal of the peritoneum. Ascites fluid collected from two Angptl4−/− mice fed HFD is shown at the bottom. (E) Photograph of small intestine of WT and Angptl4−/− mouse fed LFD or HFD for 19 weeks. (F) FLPC lipoprotein profiling of ascites fluid of Angptl4−/− mice fed HFD for 19 weeks (n=5). Inset: apoB immunoblot of FPLC fractions from ascites fluid of Angptl4−/− mice or mouse fasting plasma (FP). (G) Plasma endotoxin levels in WT and Angptl4−/− mice fed LFD or HFD for 19 weeks. (G) Lymph vessel area in the jejunum of WT and Angptl4−/− mice fed LFD or HFD for 19 weeks determined after Lyve-1 staining. Grey bars = WT mice, Black bars = Angptl4−/− mice. Error bars represent SEM. * indicates significantly different from corresponding WT mice according to Student’s t-test (P<0.05).

Representative H&E staining of small intestine (A) or mesenteric fat (B) of WT and Angptl4−/− mice fed HFD for 19 weeks (Study 1). Inset: high magnification image of lymphocyte infiltrate (top) or presence of granulocytes (bottom). (C) Sirius red staining of small intestine of WT and Angptl4−/− mice fed HFD for 19 weeks. Arrow indicates inflammatory infiltrate.

(A) Plasma serum amyloid (SAA) levels in WT and Angptl4−/− mice fed LFD or HFD (Study 1). (B) Plasma levels of numerous cytokines in WT and Angptl4−/− fed HFD for 8 weeks. Values are expressed relative to WT. (C) Hepatic mRNA levels of acute phase proteins SAA2, haptoglobin and lipocalin 2, as determined by qPCR. Expression was normalized against 36B4. (D) Hepatic mRNA levels of macrophage marker Cd68. (E) Cd68 immunostaining of liver sections of WT and Angptl4−/− mice fed HFD for 8 weeks. (F) Plasma serum albumin levels. Grey bars = WT mice, Black bars = Angptl4−/− mice. n=4–11 mice per group. Error bars represent SEM. * indicates significantly different from corresponding WT mice according to Student’s t-test (p<0.05).

(A) Kinetics of change in plasma SAA in WT and Angptl4−/− mice fed HFD. n=6–7 per group. (B) Post-prandial endotoxin levels in portal plasma from WT and Angptl4−/− mice fed LFD or HFD for 5 weeks (Study 2). n=6–7 per group. (C) Plasma SAA levels in WT and Angptl4−/− mice fed HFD for 5 weeks while being given oral antibiotics or vehicle (Study 3). n=6–10 per group. (D) Relative abundance of total bacteria expressed per weight of colonic content in Angptl4−/− mice fed HFD and given oral antibiotics or vehicle. *indicates significantly different from corresponding WT mice according to Student’s t-test (p<0.05). (E) Photograph of mesenteric fat of Angptl4−/− mouse fed HFD for 5 weeks (Study 2). Position of a dramatically enlarged lymph node is indicated. (F) Mesenteric lymph node of WT and Angptl4−/− mouse fed HFD for 5 weeks after dissection and removal of adipose tissue. (G) Plasma SAA levels in WT and Angptl4−/− mice fed LFD or different types of HFD for 5 weeks (Study 2). HF-MCT=high fat medium-chain triglycerides, HF-lard=high fat lard-based, HF-palm=high fat palm oil-based, HF-safflower=high fat safflower oil-based. (H) Size of mesenteric lymph nodes in WT and Angptl4−/− mice fed LFD or different types of HFD for 5 weeks. n=6–7 mice per group. Means of HF-palm and HF-lard groups were significantly different from the other groups as determined by One-way ANOVA followed by Tukey post-hoc test. (I) Close correlation between saturated fat content of the diet and the mean plasma concentration of SAA in Angptl4−/− mice after three weeks of feeding. (J) Expression of inflammatory marker genes in MLN of WT and Angptl4−/− mice fed HFD for 24h (Study 4). n=3 per group. Grey bars = WT mice, Black bars = Angptl4−/− mice. Error bars represent SEM.

(A) Low magnification image (200X) of H&E staining of MLN from WT and Angptl4−/− mice fed HFD for 5 weeks (Study 2). Touton giant cells are indicated by yellow arrows. (B) High magnification image (1000X) of MLN with Touton giant cells from Angptl4−/− mice fed HFD for 5 weeks (C) F4/80 immunostaining of MLN with Touton giant cell. Magnification 1000X. (D) Oil red O staining counterstained with Hematoxylin. Magnification 600X. (E) Sudan black staining. Magnification 1000X.

(A) Heat map showing changes in expression of selected genes in Angptl4−/− mouse peritoneal macrophages incubated for 6 hours with chyle (final triglyceride concentration 2 mM) and/or Orlistat (20 µM), or with chyle and/or recombinant Angptl4 (2.5 µg/mL). Expression in untreated macrophages was set at 1. In parallel, expression changes are shown of the same genes in peritoneal macrophages treated for 4 hours with LPS (100ng/mL). All genes induced by chyle by at least 2.5-fold and which are labeled with “cytokine or chemokine activity” or “immune or inflammatory response” by Gene Ontology (Biological Process) or are involved in immunity/inflammation based on literature study were included. Also, SREBP target genes that were suppressed by chyle by >75% were included. (B) Oil red O staining of Angptl4−/− mouse peritoneal macrophages incubated for 6 hours with chyle (triglyceride concentration 2 mM) and increasing concentrations of mouse recombinant Angptl4. Chyle was collected from rats fed palm oil-based HFD. Magnification 200X. Inset: high magnification image of macrophage foam cell. (C) Q-PCR expression of inflammatory genes in Angptl4−/− macrophages treated with chyle and/or Angptl4 (2.5 µg/mL). * significantly different according to Student’s t-test (p<0.05). (D) Q-PCR expression of inflammatory genes in Angptl4−/− macrophages treated with chyle and various pharmacologic inhibitors. SSO=Sulfosuccinimidyl oleate. Differences were evaluated for statistical significance by One-way ANOVA followed by Tukey post-hoc test. * significantly different from control-treated cells (p<0.05). Error bars represent SEM.

(A) Immunoblots of IRE1α, PERK, eIF2α and CHOP using regular gels or Phos-tag gels (IRE1α only) of Angptl4−/− macrophages treated with chyle (triglyceride concentration 2 mM), thapsigargin (100 nM) or tunicamycin (2.5 µg/mL) for 6 hours. “P” represents phosphorylated form. (B) Regular RT-PCR of XBP1 processing revealing spliced and unspliced XBP1 mRNA. (C) Q-PCR expression of ER stress genes. Differences were evaluated for statistical significance by One-way ANOVA followed by Tukey post-hoc test. * significantly different from control-treated cells (p<0.05). (D) Q-PCR expression of selected genes involved in ER stress in Angptl4−/− mouse peritoneal macrophages incubated for 6 hours with chyle and/or recombinant Angptl4 (2.5 µg/mL). * significantly different according to Student’s t-test (p<0.05). (E) Angptl4 mRNA expression in WT mouse peritoneal macrophages treated for 6 hours with chyle (triglyceride concentration 1.3 mM). (F) Angptl4 mRNA expression in WT mouse peritoneal macrophages or human U937 macrophages treated for 6 hours with albumin-bound free fatty acids (500 µM). (G) Q-PCR expression of selected genes in WT mouse peritoneal macrophages treated with albumin-bound free fatty acids (500 µM). Control-treated cells were treated with albumin only. Differences were evaluated for statistical significance by One-way ANOVA followed by Tukey post-hoc test. Bars with different letters are significantly different (p<0.05). Error bars represent SEM.