In adipocytes, amino acids stimulate the target of rapamycin (TOR) signaling pathway leading to phosphorylation of the translational repressor, eIF-4E binding protein-I (4E-BP1), and ribosomal protein S6. L-leucine is the primary mediator of these effects. The structure-activity relationships of a putative L-leucine recognition site in adipocytes (LeuR(A)) that regulates TOR activity were analyzed by examining the effects of leucine analogues on the rapamycin-sensitive phosphorylation of the translational repressor, eIF-4E binding protein-I (4E-BP1), an index of TOR activity. Several amino acids that are structurally related to leucine strongly stimulated 4E-BP1 phosphorylation at concentrations greater than the EC(50) value for leucine. The order of potency was leucine > norleucine > threo-L-beta-hydroxyleucine approximately Ile > Met approximately Val. Other structural analogues of leucine, such as H-alpha-methyl-D/L-leucine, S-(-)-2-amino-4-pentenoic acid, and 3-amino-4-methylpentanoic acid, possessed only weak agonist activity. However, other leucine-related compounds that are known agonists, antagonists, or ligands of other leucine binding/recognition sites did not affect 4E-BP1 phosphorylation. We conclude from the data that small lipophilic modifications of the leucine R group and alpha-hydrogen may be tolerated for agonist activity; however, leucine analogues with a modified amino group, a modified carboxylic group, charged R groups, or bulkier aliphatic R groups do not seem to possess significant agonist activity. Furthermore, the leucine recognition site that regulates TOR signaling in adipocytes appears to be different from the following: (1) a leucine receptor that regulates macroautophagy in liver, (2) a leucine recognition site that regulates TOR signaling in H4IIE hepatocytes, (3) leucyl tRNA or leucyl tRNA synthetase, (4) the gabapentin-sensitive leucine transaminase, or (5) the system L-amino acid transporter.

In previous studies we have shown that rat adipocytes suspended in Matrigel and placed in primary culture migrate through the gel to form multicellular clusters over a 5- to 6-day period. In the present study, phosphorylation of the insulin-regulated 70-kDa ribosomal protein S6 kinase (p70 S6k ) was observed within 30 min of establishment of adipocytes in primary culture. Two inhibitors of the p70 S6k signaling pathway, rapamycin and LY-294002, greatly reduced phosphorylation of p70 S6k and organization of adipocytes into multicellular clusters. Of all the components of the cell culture medium, amino acids, and in particular a subset of neutral amino acids, were found to promote both phosphorylation of p70 S6k and cluster formation. Lowering the concentrations of amino acids in the medium to levels approximating those in plasma of fasted rats decreased both phosphorylation of p70 S6k and cluster formation. Furthermore, stimulation of p70 S6k phosphorylation by amino acids was prevented by either rapamycin or LY-294002. These findings demonstrate that amino acids stimulate the p70 S6k signaling pathway in adipocytes and imply a role for this pathway in multicellular clustering.

Previous studies indicated that amino acids may activate the protein kinase activity of the target of rapamycin (TOR) and thereby augment and/or mimic the effects of insulin on protein synthesis, p70S6k phosphorylation, and multicellular clustering in adipocytes. To identify the individual amino acids responsible for these effects, the present study focused on the TOR substrate and translational repressor 4E-BP1. A complete mixture of amino acids stimulated the phosphorylation of 4E-BP1, decreasing its association with eukaryotic initiation factor eIF-4E. Studies on subsets of amino acids and individual amino acids showed that l-leucine was the amino acid responsible for most of the effects on 4E-BP1 phosphorylation; however, the presence of other amino acids was required to observe a maximal effect. The stimulatory effect of leucine was stereospecific and not mimicked by other branched chain amino acids but was mimicked by the leucine metabolite α-ketoisocaproate (α-KIC). The effect of α-KIC, but not leucine, was attenuated by the transaminase inhibitor (aminooxy)acetate. The latter result indicates that the effects of α-KIC required its conversion to leucine. Half-maximal stimulation of 4E-BP1 phosphorylation occurred at ∼430 μM; therefore, the response was linear within the range of circulating concentrations of leucine found in various nutritional states.

The role of carbonic anhydrase in de novo lipid synthesis was examined by measuring [1-14C]acetate incorporation into total lipids, fatty acids and non-saponifiable lipids in freshly isolated rat hepatocytes. Two carbonic anhydrase inhibitors, trifluoromethylsulphonamide (TFMS) and ethoxozolamide (ETZ) decreased incorporation of 14C into total lipids. Both fatty acid and non-saponifiable lipid components of the total lipid were inhibited to approximately the same extent by 100 microM TFMS (29 +/- 0.3% and 35 +/- 0.3% of control respectively in replicate studies). However, neither drug significantly affected ATP concentrations or the transport activity of Na+/K(+)-ATPase, two measures of cell viability. To establish the site of this inhibition, water-soluble 14C-labelled metabolites from perchloric acid extracts of the radiolabelled cells were separated by ion-exchange chromatography. TFMS inhibited 14C incorporation into citrate, malate, alpha-oxoglutarate and fumarate, but had no effect on incorporation of 14C into acetoacetate. Since ATP citrate-lyase, the cytosolic enzyme that catalyses the conversion of citrate into acetyl-CoA, catalyses an early rate-limiting step in fatty acid synthesis, levels of cytosolic citrate may be rate controlling for de novo fatty acid and sterol synthesis. Indeed citrate concentrations were significantly reduced to 37 +/- 6% of control in hepatocytes incubated with 100 microM TFMS for 30 min. TFMS also inhibited the incorporation of 14C from [1-14C]pyruvate into malate, citrate and glutamate, but not into lactate. This supports the hypothesis that TFMS inhibits pyruvate carboxylation, i.e. since all of the 14C from [1-14C]pyruvate converted into citric acid cycle intermediates must come via pyruvate carboxylase (i.e. rather than pyruvate dehydrogenase). Our findings indicate a role for carbonic anhydrase in hepatic de novo lipogenesis at the level of pyruvate carboxylation.

Primary rat adipocytes cultured in basement membrane component gels migrated and organized into large, three-dimensional, multicellular clusters. Gross morphological changes seen during this reorganization are described. The rate of cluster formation decreased with age of the rats and was stimulated by insulin in older, but not in younger rats. Echistatin, a disintegrin, partially inhibited the formation of multicellular clusters in a concentration-dependent fashion (50% inhibitory concentration approximately 10 nM). The original extracellular matrix was initially remodeled and eventually destroyed by the time large multicellular clusters were observed. This implied that one or more matrix-degrading protease(s) were being secreted. Adipocyte-conditioned medium was found to contain a divalent cation-sensitive gelatinase activity at approximately 72 and/or approximately 62 kDa. The elution profile of this activity from gelatin-Sepharose 4B was similar to matrix metalloproteinase 2 (MMP-2, a 72-kDa matrixin with a 62-kDa mature form), and the dimethyl sulfoxide eluant from these columns contained MMP-2 immunoreactivity. MMP-2 concentration and activity were greater in conditioned medium from young than from older animals; however, insulin did not affect the amount of MMP-2 in adipocyte-conditioned media. The matrixin inhibitor 1,10-phenanthroline not only blocked gelatinase activity in zymograms but also prevented extracellular matrix remodeling and destruction, as well as adipocyte migration and the formation of cell-cell contacts in adipocyte cultures. These observations are consistent with the hypothesis that the matrixin MMP-2 is secreted by adipocytes. Whereas matrixin activity alone may not be sufficient for the formation of multicellular clusters, the data indicate that it may have a requisite role in this process.

Enhanced phosphorylation of the ribosomal protein s6 kinase, p70(s6k), and the translational repressor, 4E-BP1, are associated with either insulin-induced or amino acid-induced protein synthesis. Hyperphosphorylation of p70(s6k) and 4E-BP1 in response to insulin or amino acids is mediated through the mammalian target of rapamycin (mTOR). In several cell lines, mTOR or its downstream targets can be regulated by phosphatidylinositol (PI) 3-kinase; protein kinases A, B, and C; heterotrimeric G-proteins; a PD98059-sensitive kinase or calcium; as well as by amino acids. Regulation by amino acids appears to involve detection of levels of charged t-RNA or t-RNA synthetase activity and is sensitive to inhibition by amino acid alcohols. In the present article, however, we show that the rapamycin-sensitive regulation of 4E-BP1 and p70(s6k) in freshly isolated rat adipocytes is not inhibited by either L-leucinol or L-histidinol. This finding is in agreement with other recent studies from our laboratory suggesting that the mechanism by which amino acids regulate mTOR in freshly isolated adipocytes may be different than the mechanism found in a number of cell lines. Therefore we investigated the possible role of growth factor-regulated and G-protein-regulated signaling pathways in the rapamycin-sensitive, amino acid alcohol-insensitive actions of amino acids on 4E-BP1 phosphorylation. We found, in contrast to previously published results using 3T3-L1 adipocytes or other cell lines, that the increase in 4E-BP1 phosphorylation promoted by amino acids was insensitive to agents that regulate protein kinase A, mobilize calcium, or inhibit protein kinase C. Furthermore, amino acid-induced 4E-BP1 phosphorylation was not blocked by pertussis toxin nor was it mimicked by the G-protein agonists fluoroaluminate or MAS-7. However, amino acids failed to activate either PI 3-kinase, protein kinase B, or mitogen-activated protein kinase and failed to promote tyrosine phosphorylation of cellular proteins, similar to observations made using cell lines. In summary, amino acids appear to use an amino acid alcohol-insensitive mechanism to regulate mTOR in freshly isolated adipocytes. This mechanism is independent of cell-signaling pathways implicated in the regulation of mTOR or its downstream targets in other cells. Overall, our study emphasizes the need for caution when extending results obtained using established cell lines to the differentiated nondividing cells found in most tissues.