Functional roles of fructose

Glucose and Fructose Stimulate pTr Cell Proliferation.

We first investigated biological effects of glucose and fructose on proliferation of pTr cells. Treatment of pTr cells with glucose in customized medium containing 5% serum for 48 h and 96 h increased pTr cell numbers by 1.8- fold (P < 0.01) and 3.2-fold (P < 0.01), respectively (Fig. 1A). Similarly, at 48 h and 96 h of fructose treatment, pTr cell numbers increased by 2.7-fold (P < 0.01) and 5.3-fold (P < 0.01), respectively (Fig. 1B). However, there was no evidence that treatment of pTr cells with both glucose and fructose had either additive or synergistic effects (P > 0.10) (Fig. 1C). These results indicate that fructose and glucose stimulate pTr cell proliferation, but the fold-increase was greater for fructose.

Fig. 1.

Dose-dependent effects of glucose (A), fructose (B), and glucose plus fructose (C) on the proliferation of pTr cells. The pTr cells were seeded at 30% confluence in wells and cultured with the indicated doses of glucose and fructose. Cell numbers were determined after 48 h (Left) or 96 h (Right) of incubation, and data are expressed as a percentage relative to nontreated control (100%). All quantitative data are presented as the LSM with overall SEM. See Materials and Methods for a complete description.

Glucose and Fructose Activate MTOR Signaling in pTr Cells.

The pTr cells deprived of glucose and fructose for 6 h after serum starvation were treated with 4 mM glucose or 4 mM fructose. Western blot analyses of whole pTr cell extracts indicated that glucose increased phosphorylated ribosomal protein S6K (p-RPS6K), p-eukaryotic translation initiation factor 4E-binding protein 1 (p-EIF4EBP1), and p-ribosomal protein S6 (p-RPS6) by 2.7-fold (P < 0.01), 2.4-fold (P < 0.01), and 6.1-fold (P < 0.01), respectively, over basal levels within 30 min and that this level of activation was maintained to 120 min after treatment (Fig. 2 A–C). Likewise, fructose increased the abundance of p-RPS6K, p-EIF4EBP1, and p-RPS6 by 2.9-fold (P < 0.01), 2.3-fold (P < 0.01), and 4.2-fold (P < 0.01), respectively, over basal levels within 30 min, and this level of activation was maintained to 120 min after treatment (Fig. 2 D–F). These results indicate that both fructose and glucose activate the MTOR-RPS6K–RPS6 or MTOR–EIF4EBP1 pathways, stimulating proliferation and growth of pTr cells.

Fig. 2.

Analysis of time-dependent phosphorylation of RPS6K (A and D), EIF4EBP1 (B and E), and RPS6 (C and F) in response to glucose (A, B, and C) and fructose (D, E, and F) in pTr cells at 0, 15, 30, 60, and 120 min after treatment. Monolayers of 80% confluent pTr cells were serum starved for 24 h, then deprived of either glucose or fructose for 6 h, and then treated with physiological concentrations of glucose (4 mM) or fructose (4 mM). Blots were imaged to calculate the normalized values as levels of phosphorylated protein relative to total protein. See Materials and Methods for a complete description.

We next performed immunofluorescence analyses to determine cellular locations of p-RPS6K, p-EIF4EBP1, and p-RPS6 in response to treatment of pTr cells with 4 mM glucose or 4 mM fructose. The abundance of p-RPS6K protein increased in the nucleus after treatment with glucose and fructose but not in untreated pTr cells (Fig. 3A). The p-EIF4EBP1 protein was localized predominantly in the nucleus but also, to a much lesser extent, was found in the cytoplasm of pTr cells in response to glucose and fructose (Fig. 3B). The p-RPS6 protein was abundant in the cytoplasm of pTr cells treated with glucose and fructose (Fig. 3C). These results indicate that each effector protein involved in mRNA translation downstream of MTOR was localized in and presumably exerted its functional role in different intracellular compartments of pTr cells to stimulate cell proliferation.

Fig. 3.

Immunofluorescence localization of p-RPS6K (A), p-EIF4EBP1 (B), and p-RPS6 (C) proteins in pTr cells. Immunoreactive p-RPS6K protein was localized predominantly in the nucleus of pTr cells treated with glucose or fructose, p-EIF4EBP1 protein was localized in the cytoplasm and nucleus of pTr cells treated with glucose and fructose, respectively; p-RPS6 protein was localized primarily in the cytoplasm of pTr cells treated with glucose or fructose. (Magnification: 40×.). See Materials and Methods for a complete description.

MTOR Signal Transduction Induced in pTr Cells by Glucose and Fructose Involves PI3K/AKT1 Proteins.

To determine the cell-signaling pathways mediating effects glucose and fructose on PI3K/AKT1, RPS6K, and EIF4EBP1, pTr cells were pretreated with pharmacological inhibitors of either PI3K (25 μM LY294002) or MTOR kinase (25 nM rapamycin) for 2 h. Results indicated that phosphorylation of both RPS6K and EIF4EBP1 in response to glucose was inhibited by both PI3K and MTOR inhibitors (P < 0.01 or P < 0.001, respectively; Fig. 4 A and C). Likewise, induction of phosphorylation of both RPS6K and EIF4EBP1 by fructose was inhibited by both PI3K and MTOR inhibitors (P < 0.01 or P < 0.001, respectively; Fig. 4 B and D). These results suggest that activation of the PI3K–AKT1 pathway by fructose is required for transcriptional activation in the fructose-induced cell signaling cascade from cytoplasm to nucleus to enhance proliferation and growth of pTr cells.

Fig. 4.

Inhibition of RPS6K (A and B) and EIF4EBP1 (C and D) phosphorylation. Monolayers of 80% confluent pTr cells were serum starved for 24 h, deprived of glucose and fructose for 6 h, and then pretreated with 25 μM LY294002 or 25 nM rapamycin for 30 min. After stimulation with 4 mM glucose (A and C) or 4 mM fructose (B and D) for 60 min, total cell lysates were subjected to SDS/PAGE followed by Western blotting. Both rapamycin and LY294002 inhibited increases in p-RPS6K and p-EIF4EBP1. Blots were imaged to calculate the normalized values presented as levels of phosphorylated protein relative to total protein. Asterisks denote an effect of treatment (*P < 0.01; **P < 0.001). All quantitative data are presented as the LSM with overall SEM. See Materials and Methods for a complete description.

MTOR Cell Signaling in Response to Glucose and Fructose Is Mediated by Glutamine-Fructose-6-Phosphate Transaminase 1 Activity.

To examine whether glutamine-fructose-6-phosphate transaminase 1 (GFPT1) is both an upstream effector of MTOR and a nutrient sensor in the hexosamine biosynthesis pathway, we used azaserine (an inhibitor of GFPT1) to assess whether GFPT1 activity is required for glucose and fructose to activate pTr cell proliferation. Azaserine inhibited (P < 0.01) the ability of both glucose (Fig. 5A) and fructose (Fig. 5B) to stimulate proliferation of pTr cells. In particular, the rate of pTr cell proliferation decreased 70–80% with the addition of 1 μM azaserine compared with effects of glucose alone (Fig. 5A) or fructose alone (Fig. 5B). Even in the presence of both fructose and glucose, azaserine inhibited (P < 0.01) proliferation of pTr cells (Fig. 5C). These results indicate that the nutrient-sensing system in pTr cells is more sensitive to fructose than to glucose.

Fig. 5.

Inhibition of GFPT1 activity. pTr cells pretreated for 2 h with the indicated doses of azaserine were stimulated with 4 mM glucose (A), 4 mM fructose (B), or 4 mM glucose and 5 mM fructose (C) or for 96 h. The cell-culture medium was changed every 48 h. Cell numbers were determined after stimulation, and data are expressed as a percentage relative to nontreated control (100%). Asterisks denote an effect of treatment (*P < 0.01). All quantitative data are presented as the LSM with overall SEM. See Materials and Methods for a complete description.

The next experiment used Western blot analyses of whole pTr cell extracts treated with azaserine to determine if inhibition of GFPT1 decreased the abundance of p-RPS6K and p-EIF4EBP1 proteins (Fig. 6 A and B). Interestingly, the phosphorylation of RPS6K and EIF4EBP1 proteins was inhibited more by treatment with the combination of azaserine and LY294002 than by treatment with azaserine alone when pTr cells were stimulated with either glucose or fructose (Fig. 6 C and D). These results suggest that fructose stimulation of the MTOR signaling cascade is affected by both the AKT1- and GFPT1-dependent cell-signaling pathways for stimulation of proliferation and growth of pTr cells.

Fig. 6.

Inhibition of GFPT1 decreased the abundance of RPS6K and EIF4EBP1 phosphorylation. Monolayers of 80% confluent pTr cells were serum starved for 24 h, deprived of glucose and fructose for 6 h, and then were pretreated with 25 μM LY294002 or 25 nM rapamycin for 30 min and azaserine for 2 h to inhibit PI3K, MTOR, and GFPT1, respectively. Panels A and B indicate effects of inhibitors of PI3K, MTOR, and GFPT1 in p-RPS6K. Panels C and D indicate effects of the same respective inhibitors on p-EIF4EBP1. After stimulation, total cell lysates were subjected to SDS/PAGE followed by Western blotting. Blots were imaged to calculate the normalized values presented as levels of phosphorylated protein relative to total protein. Asterisks denote an effect of treatment (*P < 0.01; **P < 0.001). All quantitative data are presented as the LSM with overall SEM. See Materials and Methods for a complete description.

The next step was to perform a GFPT1-knockdown experiment using GFPT1-specific siRNAs. As illustrated in Fig. 7A, GFPT1 protein expression was inhibited by 70, 80, and 81% 48 h after transfection with GFPT1 siRNA at 10, 25, and 50 nM, respectively. Cells transfected with the GFPT1-specific siRNA had less GFPT1 than did naive and mock-treated cells (P < 0.001) or cells transfected with control siRNA (P < 0.001). To verify these results, we performed immunofluorescence analyses and compared the expression patterns of GFPT1 protein in pTr cells cultured in the presence of control and GFPT1 siRNA (Fig. 7B). Immunoreactive GFPT1 protein was most abundant in the nucleus of pTr cells treated with control siRNA and naive cells but was barely detectable in cells transfected with GFPT1 siRNA. When combined with GFPT1 siRNA, glucose inhibited pTr cell proliferation by 60%, and fructose inhibited pTr cell proliferation by 72% (P < 0.01). Also, when both fructose and glucose plus GFPT1 siRNA were used to culture pTr cells, cell proliferation decreased by about 50% (P < 0.01) (Fig. 7 C–E). Collectively, these results demonstrate that stimulation of the MTOR–RPS6K and MTOR–EIF4EBP1 signaling pathways by glucose and fructose is mediated through GFPT1 and the hexosamine pathway.

Fig. 7.

GFPT1 knockdown decreased proliferation of pTr cells. (A) GFPT1 protein levels were measured in the control group (naive, mock, and control siRNA treatment) and in the GFPT1-silenced group (GFPT1 siRNA) in a dose-dependent manner. Blots were imaged to calculate the normalized values presented as levels of GFPT1 protein relative to TUBA protein. (B) Immunofluorescence microscopy detected GFPT1 protein in naive and control siRNA treatments but not in the GFPT1-silenced group. (C–E) Effects of glucose (C), fructose (D), and fructose plus glucose (E) combined with GFPT1 siRNA treatment on proliferation of pTr cells. Asterisks denote an effect of treatment (*P < 0.01). All quantitative data are presented as the LSM with overall SEM. See Materials and Methods for a complete description.

Glucose and Fructose Stimulate Hyaluronic Acid Synthesis.

Hyaluronic acid, a glycosaminoglycan synthesized via the hexosamine biosynthesis pathway, may play an essential role in angiogenesis and other aspects of cell function, particularly in the placenta during early pregnancy (31). Therefore, we determined whether fructose stimulates hyaluronic acid synthesis. As illustrated in Fig. 8A, glucose increased the amount of secreted hyaluronic acid by 6.2-fold at 48 h and by 11.3-fold at 96 h, and this effect was inhibited by azaserine in a dose-dependent manner. Pericellular hyaluronic acid also was increased by glucose treatment and inhibited by azaserine. Similarly, fructose increased secreted hyaluronic acid by 10.5- and 12.0-fold (P < 0.01) and increased pericellular hyaluronic acid by 7.0- and 7.3-fold (P < 0.01) at 48 h and 96 h of culture, respectively, and the effects of fructose were inhibited by azaserine in a dose-dependent manner (Fig. 8B). These results indicate that fructose and glucose stimulate the production of hyaluronic acid via GFPT1 and the hexosamine biosynthesis pathway to stimulate growth and development of the porcine trophoblast and perhaps the embryo/fetus during pregnancy.

Fig. 8.

Glucose and fructose stimulate hyaluronic acid (HA) synthesis and secretion in porcine trophectoderm cells. pTr cells were serum starved for 24 h and pretreated with different concentrations of azaserine for 2 h and then were treated or not treated with glucose (A) or fructose (B) for 48 h (Left) or 96 h (Right). (A) Quantitation of secreted (Upper) and pericellular (Lower) hyaluronic acid by glucose. (B) Quantitation of secreted (Upper) and pericellular (Right) hyaluronic acid by fructose. The amount of secreted and pericellular hyaluronic acid in each extract was measured by ELISA based on the specific interaction of hyaluronic acid with hyaluronic acid-binding protein. Asterisks denote an effect of treatment (*P < 0.01). See Materials and Methods for a complete description.

Cell Culture.

An established mononuclear pTr cell line from day 12 pig conceptuses was used in the present in vitro studies, as described previously (42,43). For each experiment, the design was replicated in three independent experiments.

Proliferation Assay.

The pTr cells (20% confluent) were subcultured in 24-well plates in growth medium until the monolayer reached up to 30% confluence and then were switched to serum- and insulin-free customized medium. After starvation for 24 h, cells were deprived of glucose and fructose for 6 h. Then either glucose or fructose was added to each well (n = 3 wells per treatment) with 5% (vol/vol) serum at the indicated concentration. Cells grown in serum-free specialized medium and specialized medium containing 10% FBS served as negative and positive controls, respectively. Cell numbers were determined as described previously (44–48).

Western Blot Analyses.

Whole-cell extracts and immunoblot assays were prepared and performed as described previously (44). Immunoreactive proteins were detected using rabbit anti-human polyclonal antibodies against p-RPS6 IgG at a 1:1,000 dilution and RPS6 IgG at a 1:2,000 dilution and 10% (wt/wt) SDS/PAGE gel; rabbit anti-human polyclonal p-RPS6K IgG at a 1:1,000 dilution and RPS6K IgG at a 1:2,000 dilution and 10% SDS/PAGE gel; and rabbit anti-human polyclonal p-EIF4EBP1 IgG and EIF4EBP1 IgG, each at a 1:1,000 dilution, and 15% (wt/wt) SDS/PAGE gel. As a loading control, mouse anti–α-tubulin (TUBA) IgG was used after the proteins were detected on the Western blots. All antibodies were from Cell Signaling Technology. Multiple exposures of each Western blot were performed to ensure linearity of chemiluminescent signals.

Immunofluorescence Analyses.

The effects of glucose and fructose on phosphorylation of RPS6K, EIF4EBP1, and RPS6 were determined by immunofluorescence microscopy as previously described (44, 45). Cells probed with rabbit anti-human polyclonal p-RPS6K IgG (Thr421/Ser424) at a 1:200 dilution, rabbit anti-human polyclonal p-RPS6 IgG at a 1:200 dilution, or rabbit anti-rat polyclonal p-EIF4EBP1 IgG at a 1:200 dilution (antibodies were from Cell Signaling Technology) were incubated with goat anti-rabbit IgG Alexa 488 (Chemicon) at a 1:200 dilution for 1 h at room temperature. Cells then were washed and overlaid with Prolong Gold Antifade with DAPI. For each primary antibody, images were captured with identical microscope and detector settings to facilitate comparisons of spatial distribution and fluorescence intensity among treatments.

Target-Specific siRNAs for GFPT1 Knockdown.

For mRNA sequences of porcine GFPT1 ({"type":"entrez-nucleotide","attrs":{"text":"NM_001194979.1","term_id":"303228251","term_text":"NM_001194979.1"}}NM_001194979.1), three potential siRNA target sites were determined using the Invitrogen design program. The most effective target sequence (CAG AUC AUG AAG GGC AAC U) was selected and synthesized. Silencer Negative Control siRNA (Cell Signaling Technology), which does not cause specific degradation of any cellular mRNA, was used as a negative control. Down-regulation of GFPT1 expression was confirmed by Western blotting analyses and immunofluorescence.

Transfection.

The pTr cells were treated with specific GFPT1 siRNA; controls included cells that received naive treatment (no siRNA or Lipofectamine 2000) or mock treatment (Lipofectamine 2000 only). Transfection of siRNA was performed according to the manufacturer’s procedure. To analyze the effects of azaserine on glucose- and fructose-induced proliferation in pTr cells, glucose (4 mM) and/or fructose (4 mM) with azaserine was added to the culture medium 48 h after transfection, and the incubation was continued for another 48 h. Using green fluorescein conjugate control siRNA duplexes (Cell Signaling Technology), we estimated that more than 95% of the cells were transfected successfully.

Quantitation of Hyaluronic Acid.

The amount of hyaluronic acid was measured as described previously (37), with minor modifications. After treatment, cell culture medium was centrifuged at 8,000 × g at 4 °C for 5 min, and the supernatant was assayed to determine secreted hyaluronic acid. Then cells were washed three times with PBS and treated with Trypsin-EDTA (0.25%) solution at 37 °C for 5 min; the reaction was stopped by the addition of culture medium just before cell numbers were determined by using a hemocytometer. The cells then were centrifuged, and the supernatant was used for quantitation of pericellular hyaluronic acid. The cell pellet was digested with proteinase K (120 μg/mL) and centrifuged, and the supernatant was analyzed for intracellular hyaluronic acid. The amount of hyaluronic acid in each different extract was measured by ELISA based on the specific interaction of hyaluronic acid with hyaluronic acid-binding protein as determined using the hyaluronic acid detection kit (R&amp;D Systems). To serve as the negative control, some samples were pretreated with hyaluronidase (2 units/mL at 37 °C for 2 h) before being subjected to the ELISA.

Statistical Analyses.

All quantitative data were subjected to least-squares ANOVA using the general linear models procedures of the Statistical Analysis System (SAS Institute, Inc.). Percent changes in responses were analyzed following log transformation, because percentages are not distributed normally. Data from Western blotting were corrected for differences in sample loading using TUBA data as a covariate. All tests of significance were performed using the appropriate error terms according to the expectation of the mean squares for error. Significance was considered to be P ≤ 0.05. Data are presented as the least-squares mean (LSM) and SEM.

Author Summary

Mammalian conceptuses (i.e., embryo, fetus, and associated placental membranes) undergo growth and differentiation in response to nutrients, such as glucose, that are transported into the pregnant uterus. In ungulates and cetaceans, glucose is converted to fructose by the placenta. Glucose is an energy source for proliferation and growth of mammalian cells, but little is known about the role of fructose in the pregnant uterus, even though it is the most abundant hexose sugar in fetal blood and fluids of ungulates and cetaceans. Because fructose is not metabolized as an energy source by the glycolytic pathway or Krebs cycle, we hypothesized that instead it is used by multiple metabolic pathways critical to pregnancy and to the development of multiple organ systems of the fetus. Here we show that fructose and glucose are metabolized equally by nutrient-sensing pathways that increase cell proliferation, protein synthesis, and production of glycosaminoglycans critical to the growth of the conceptus.

Domestic livestock are invaluable animal models used extensively in research relevant to human and veterinary reproductive medicine (1). Fructose is produced continuously by the placenta, but it is not transported from the conceptus to maternal blood (2). Furthermore, fructose can be used for synthesis of RNA and DNA, for generating reducing equivalents in the form of NADPH H^{, and for the synthesis of neutral lipids and phospholipids in the heart, liver, kidney, brain, and adipose tissue of fetal lambs (}3).

To understand better the role of fructose in the pregnant uterus, we conducted experiments using our established porcine trophectoderm (pTr) cell line in an attempt to unravel the mystery of fructose metabolism. The pTr cells were cultured in the presence of glucose or fructose to determine their respective effects on the activation of mechanistic target of rapamycin (MTOR) cell signaling to effect cell proliferation and protein synthesis and on metabolism via the hexosamine pathway to produce glycosaminoglycans and UDP-N-acetylglucosamine that also can stimulate proliferation and protein synthesis in pTr cells through activation of MTOR cell signaling. Azaserine, an inhibitor of glutamine-6-phosphate (GFPT1), was used to confirm the metabolism of fructose and glucose using the hexosamine metabolic pathway. Results of these experiments indicated a key role for fructose in activating the MTOR cell-signaling pathway to increase proliferation and protein synthesis in pTr cells. The results also supported our hypothesis that fructose is metabolized by GFPT1 to glucosamine-6-phosphate, which enters the hexosamine biosynthesis pathway and thereby leads to synthesis of uridine diphosphate-N-acetylglucosamine, which also can stimulate MTOR cell signaling to enhance proliferation, protein synthesis, and the synthesis of glycosaminoglycans, particularly hyaluronic acid.

In pregnant ewes, the maximum concentration of glucose in allantoic fluid is 1.1 mM between days 35–140 of pregnancy, whereas fructose concentrations range from 11.1–33 mM during the same period (4). Therefore, fructose and glucose exert their maximum effects on cell proliferation well below and above, respectively, their molar concentrations found in allantoic fluid and fetal blood. Unique features of fructose metabolism are the production of fructose by the placenta, sequestration in fetal blood and fluids, and metabolism of fructose along with glutamine via the hexosamine pathway. This process leads to the synthesis of glycosaminoglycans, such as hyaluronic acid, uridine diphosphate-N-acetyl glucosamine, and uridine diphosphate-N-acetyl galactosamine, which are precursors to glycolipids, glycosaminoglycans, and proteoglycans.

Hyaluronic acid and hyaluronidase levels increase in the uterine lumen during pregnancy and stimulate angiogenesis, morphogenesis, and tissue remodeling of the human placenta (5). In addition, uridine diphosphate-N-acetylglucosamine may be involved in a mechanism by which phosphorylation of tuberous sclerosis 2, a GTPase-activating protein, activates MTOR and p70S6K1, a protein kinase downstream of MTOR, to stimulate proliferation of trophectoderm in response to the metabolism of glucose to glucose-6-PO₄, fructose-6-PO₄, and glucosamine-6-PO₄. The results of the present study indicate that fructose metabolism via the hexosamine pathway increases MTOR cell signaling, which increases cell proliferation, protein synthesis, and synthesis of hyaluronic acid. These processes affect angiogenesis and conceptus growth and development during pregnancy (Fig. P1). Thus, our experiments show that the unusual abundance of fructose in fetal blood and fetal fluids plays important physiological roles in metabolic pathways critical to fetal and placental growth and development during gestation.

Fig. P1.

Schematic diagram of the GFPT1-mediated MTOR signaling pathway affected by glucose and fructose in a pTr cell line. Fructose stimulates GFPT1 in the hexosamine biosynthesis pathway and activates MTOR-RPS6K and MTOR-EIF4EBP1 signal transduction cascades, which stimulate pTr cell proliferation and protein synthesis. AKT1, protooncogenic protein kinase Akt; EIF4EBP1, eukaryotic translation initiation factor 4E-binding protein 1; Fru, fructose; Glc, glucose; Fru-1P, fructose-6-phosphate, Fructose-6P; GlcN-6P, N-acetylglucosamine-6-phosphate; GFPT1, glutamine-fructose-6-phosphate transaminase 1; Glc-6P, glucose-6-phosphate; GLUT, glucose/fructose transporter; MAPK/ERK, mitogen-activated protein kinase/extracellular signal-regulated kinase; MTORC1, mechanistic target of rapamycin for cell proliferation and mRNA translation; RPS6, ribosomal protein S6; RPS6K, ribosomal protein S6K; TSC2, tuberous sclerosis 2; UDP-GlcNAC, UDP-N-acetylglucosamine.