Reply to:LeptinIs a Hypoxia-Inducible Gene

Abstract

In their letter, Guerre-Millo et al. ((1)) provide explanation for the different results reported earlier by us ((2)). To discuss the control of leptin production by adipocytes under hypoxic conditions, one should focus on two issues. The first is hypoxia itself. Grosfeld et al. ((3)) reported recently that hypoxia increases leptin gene expression, through activation of the leptin gene promoter, and leptin production by human PAZ6 adipocytes. With consideration to the fact that leptin has a direct ((4)) or synergistic effect ((5)) on angiogenesis and vascular permeability, it is reasonable to consider that leptin is a hypoxia-inducible gene. In this regard, hypoxia can up-regulate leptin production in PAZ6 adipocytes, as well as human trophoblastic cell line (BeWo cells) ((6)). Second, it seems that cell size is an important determinant of leptin expression. Adipocytes can markedly change their own cell size according to lipid (mainly triglyceride) content, which is regulated by a balance between lipogenic and lipolytic functions of these cells, and insulin plays a central role in regulating metabolic pathways associated with energy storage and use of adipocytes. This rapid change in cell size is seldom seen in other cell types. Clinically, leptin may function as part of the signaling pathway from adipose tissue, regulate the size of the body fat depot ((7)), and serve as a useful marker of adiposity, reflecting the total lipid content in humans ((8)). In vitro, small fat cells express less obese (ob) mRNA than large fat cells even when obtained from the same individual ((9)), and exogenously applied tension to the plasma membrane leads to the induction of gene expression in other cell systems ((10)). Therefore, a change in adipocyte cell size serves as an important signal for induction of leptin expression even under hypoxic conditions. Hypoxic stress increases the expression of a variety of genes with products that act in concert, both systematically and at hypoxic sites, to facilitate the supply of metabolic energy. Zelzer et al. ((11)) reported that insulin induces the genes of glycolytic enzymes, glucose transporters Glut-1 and Glut-3, and VEGF in Hep-G2 hepatoma through the hypoxia-inducible factor-1 (HIF-1) pathway, similar to the action of cobalt chloride (CoCl2). Similar data in adipocytes were recently confirmed by Mick et al. ((12)), i.e., VEGF production is up-regulated by insulin and this effect is additive with CoCl2. Taken together, it is conceivable that CoCl2 or desferrioxamine (DFO) increases leptin release from primary rat adipocytes through the HIF-1 pathway ((3)) because leptin secretion is regulated by adipocyte–glucose use in response to insulin ((13)). We speculate that CoCl2 or DFO exerts insulin-like actions through the HIF-1 pathway ((11)) and that these agents induce the products associated with energy influx and storage of adipocytes, including glycolytic enzymes, glucose transporters Glut-1 and Glut-3, VEGF ((12)), and also leptin ((3)). However, the question arises as to why leptin accumulation was not enhanced but rather decreased when primary rat adipocytes were cultured under hypoxia((1),(2)). Hypoxia activates transcription through a mitochondria-dependent signaling process that involves increased reactive oxygen species (ROS), whereas CoCl2 activates transcription by stimulating ROS generation through a mitochondria-independent mechanism ((14)). These different mechanisms of cellular O2 sensing during hypoxia and in response to CoCl2 may be a key factor in explaining the discrepancy in leptin production by primary rat adipocytes in response to hypoxia and chemical hypoxia (CoCl2 and DFO) ((1),(3)). We speculate that during hypoxia, primary rat adipocytes induce several glycolytic enzymes and glucose transporters Glut-1 and Glut-3, through the HIF-1 pathway and facilitate the production of glycolytic adenosine triphosphate (ATP) to compensate for the low-energy state. The adipocytes cannot support the accumulation of triglycerides. Despite up-regulating leptin expression through the HIF-1 pathway, the signal invoked by a change in adipocyte cell size may not be potent enough to up-regulate the pathways that induce leptin expression. Consequently, hypoxia inhibits leptin production by primary rat adipocytes. Then, the question arises as to why hypoxia can stimulate leptin secretion in human PAZ6 brown adipose cells ((3)) and another leptin-producing cell type, the choriocarcinoma BeWo cells ((6)). These cells are a family of cancer cells, i.e., tumor, known to be the most pro-angiogenic. Because of progressive tumor growth, these cells are under persistent hypoxia, which stimulates angiogenesis and serves to relieve hypoxia through reoxygenation of tissues ((15)). If the size of PAZ6 adipocytes cultured under hypoxia does not significantly change relative to that under normoxia—attributable to the immortalized feature of PAZ6 adipocytes ((16)) and/or culture conditions (medium and supplements including insulin), i.e., under the same level of cell-size-dependent signals in the adipocyte—then hypoxia should increase leptin production through the HIF-1 pathway in human PAZ6 adipose cells.