Fatty acids alter glycerolipid metabolism and induce lipid droplet formation, syncytialisation and cytokine production in human trophoblasts with minimal glucose effect or interaction

Short and medium term interaction of glucose with NEFA metabolism is minimal in human trophoblast cells in culture. Trophoblast cells were established in culture in CMRL complete over 16h for all experiments. In short-term experiments, glucose utilisation (A), fatty acid oxidation (B, C) and fatty acid esterification into (D) triacylglycerol (TG), (E) diacylglycerol (DAG) and (F) phospholipids (PL) were measured in the cells over 2h in KRBH (A) or KRBH/0.5%BSA (B–F) at various glucose and palmitate concentrations as indicated. In medium term experiments, fatty acid esterification into (G) TG, (H) DAG and (I) PL was measured in the cells over 24h in CMRL complete medium containing 0.5% BSA at 5.5 (5.5G) or 15.0 (15G) mM glucose in the presence of 0.2 (0.2FA) or 0.4 (0.4FA) mM palmitate. Due to variation of rates of esterification between placentas, the data were corrected to 100% for the 0.2FA/5G baseline condition. Data are means±SEM of n=12 determinations from 4 separate placentas (A), n=24 determinations from 4 separate placentas (B, C), n=8–9 determinations from 3 separate placentas (D–F) and n=9 determinations from 2 placentas (G–I). FA- fatty acid. Two way ANOVA: p<0.0001 palmitate effect (D, E, G, H, I), p<0.01 palmitate effect (F), p<0.05 glucose effect (H), p<0.01 glucose effect (I). Bonferroni post hoc test: *p<0.05, **p<0.01 and ***p<0.001.

Light microscopic appearance of trophoblast cells in culture, staining of cells with oil red-O to demonstrate lipid and desmoplakin immunofluorescence staining to demonstrate syncytia formation. Trophoblast cultures were established in CMRL complete medium over 16h and then cultured for a further 24–48h in RPMI complete medium containing 0.5% BSA at 1 or 10mM glucose in the presence or absence of 0.25mM NEFA (palmitate:oleate ratio, 1:1). (A-upper panel) Representative photomicrographs of cells from 1 of 7 placentas cultured for 48h in the conditions as indicated (x20). (A-lower panel) Representative photomicrographs of cells from 1 of 4 placentas harvested after 24h culture in the conditions as indicated, fixed and stained with oil red-O as per Material and Methods (x40, inserts x1000). (B) Representative photomicrographs of cells cultured for 48h in 10mM glucose and 0.25mM NEFA (palmitate:oleate ratio, 1:1) (10G+FA) and 10mM glucose without NEFA (10G No FA) and immunostained for desmoplakin (green) together with DAPI staining of nuclei (blue) (x40). Syncytia cells (S) counted if more than one nucleus could be seen within a cell that had complete cell membrane demonstrable by immunostaining. (C) Cell aggregation from photomicrographs of trophoblasts in culture at 48h as shown in the upper panel of (A) scored by a blinded observer. Mainly single cells, score=1; mainly loose small groups, score=2, mainly tight small groups, score=3; mainly tight larger groups, score=4. Culture conditions: 1 (1G) or 10 (10G) mM glucose in the presence or absence of 0.25mM NEFA (palmitate:oleate ratio, 1:1) (+FA or No FA, respectively). (D) Syncytia count per high power field (x40) by blinded observer of desmoplakin immunostained trophoblasts after 48h in culture. Means±SEM of n=7 separate experiments (C) and n=10 randomly selected high power fields per condition from one experiment (D). Kruskal–Wallis test: p<0.0001 (C), p<0.001 (D). Dunn's multiple comparison post hoc test: *p<0.05, **p<0.01.

Electron microscopic (EM) appearance of trophoblast cells established and cultured in conditions as described in Fig. 2 legend. Representative EM photomicrographs from 1 of 2 placentas cultured in (A) 1mM glucose in the absence of NEFA, scale bar=2μm, (B) 10mM glucose in the absence of NEFA, scale bar=2.0μm, insert scale bar=0.5μm, (C) 1mM glucose in the presence of 0.25mM NEFA, scale bar=5μm, insert scale bar=1μm, and (D) 10mM glucose in the presence of 0.25mM NEFA, scale bar=2μm. Nucleus (N), rough endoplasmic reticulum (R), mitochondria (m), glycogen lakes (G), lipid droplets (L) and myelin figures (MF).

Fatty acid partitioning and adipophilin mRNA expression in trophoblast cells is altered by pre-treatment of the cells for 24h with NEFA. Trophoblast cultures were established in CMRL complete medium over 16h. Cells were then cultured for a treatment period of 24h (6h for mRNA analysis) in RPMI complete medium containing 0.5% BSA at 1 (1G) or 10 (10G) mM glucose in the presence or absence of 0.25mM NEFA (palmitate:oleate ratio, 1:1) (+FA or No FA, respectively). Fatty acid esterification into triacylglycerol (TG) (A) diacylglycerol (DAG) (B) and phospholipids (PL) (C), fatty acid oxidation (D) and glycerol release into the media (E) were then measured over 2h in KRBH/0.5%BSA with 0.1mM cold palmitate at 1.0, 5.0 or 15mM glucose and palmitate metabolic tracers as indicated in Materials and Methods. Due to variation of rates of fatty acid oxidation and esterification between placentas, the data were corrected to 100% for the No FA/1.0G treatment condition assessed at 1.0mM glucose (A–D). (F) Expression of adipophilin at the mRNA level corrected for glyceraldehyde-3 phosphate dehydrogenase after 6h in experimental culture conditions. Means±SEM of n=6–8 determinations from 3 separate experiments (A, C), n=5 determinations from 2 separate experiments (B), n=4 determinations from 2 separate experiments (D), n=6 determinations from 2 separate experiments (E) and n=8 determinations from 4 separate experiments (F). Two way ANOVA: p<0.05 24h treatment effect (B), p<0.0001 24h treatment effect (A, C, D, E), p<0.0001 FA effect (F), p<0.01 FA and glucose interaction (F). Bonferroni post hoc test: *p<0.05, **p<0.01 and ***p<0.001.

Cytokine secretion from human trophoblast cells in culture is stimulated by NEFA but not glucose. Trophoblast cultures were established in CMRL complete medium over 16h. Cells were then cultured for a treatment period of 24h in RPMI complete medium containing 0.5% BSA at 1 (1G) or 10 (10G) mM glucose in the presence or absence of 0.25mM fatty acid (palmitate:oleate ratio, 1:1) (+FA or No FA, respectively). The secretion of (A) tumor necrosis factor α (TNFα), (B) interleukin 1β (IL1β), (C) interleukin 6 (IL6), (D) interleukin 10 (IL10) into the incubation medium was measured by cytokine bead array assay at 24h. Scatter dot plots showing means of n=10–12 experiments. Two way ANOVA with placenta matching: p<0.05 FA effect (C, D), p<0.01 FA effect (A, B). Bonferroni post hoc test: *p<0.05.

Models of trophoblast (A) glucose and (B) NEFA metabolism. Blue boxes relate to pathways assessed in the current study. Yellow boxes relate to pathways assessed by others and/or are relevant to the discussion of this work. (A) Glucose metabolic flux via glycolysis is near maximal at 4mM glucose (1), but from previous studies we know that total glucose utilisation of placenta is not maximal until the glucose concentration is greater than 16mM . Elevated glucose in diabetes, therefore, may increase flux into pathways up-stream of glycolysis such as glycogen synthesis (2) and/or the pentose-phosphate pathway (3), but is unlikely to increase flux into mitochondrial oxidation or anaplerosis pathways (down-stream) (4), including to increase malonyl-CoA levels necessary for lipogenesis and to alter fatty acid partitioning via inhibition of fatty acid oxidation. (B) All arms of NEFA partitioning are active in the trophoblast and NEFA are avidly taken up into lipid droplets. Prolonged exposure to elevated NEFA results in lipid metabolism that may protect the fetus from excess supply. Elevated NEFA down-regulates lipoprotein lipase activity [47] such that NEFA entry to the trophoblast from lipoproteins can be curtailed. Elevated NEFA also upregulates NEFA esterification into TG, expression of the lipid droplet protein adipophilin and reduces lipolysis, all of which favours buffering of NEFA into lipid droplets. NEFA minimally down-regulates fatty acid oxidation. Of relevance to diabetes, elevated glucose has minimal effect on the NEFA partitioning pathways due to limited effects on malonyl-CoA levels (panel A), but may attenuate the formation of lipid droplets. The clearance of lipid from trophoblasts by synthesis and secretion of lipoproteins [28] may be of importance but has not been examined in this study. Gluc- glucose; G6P-glucose-6-phosphate; LPL-lipoprotein lipase; LC-CoA-long chain acyl-CoA; MTTP-microsomal triglyceride transfer protein; PA-phosphatidic acid; VLDL-very low density lipoprotein.