The Foxo Family: Partners in Crime or Silent Heroes

High forkhead box protein 01 (Foxo1) activity in liver increases gluconeogenesis and hepatic glucose output; thus, Foxo1 is accused of contributing to the pathology of diabetes (for review, see Ref. 1). Until recently, other closely related members of the forkhead/winged-helix family of transcriptional regulators, Foxo3 and Foxo4, have received little attention because of negligible knockout phenotypes in mice (2, 3). In the current issue of Endocrinology, Zhang et al. (4) reveal new metabolic functions for these seemingly redundant family members through analysis of overlapping, liver-specific deletions of the three Foxo proteins. This elegant new study illustrates the increasing complexity of the transcriptional networks controlling metabolism and emphasizes the importance of considering these interactions when devising new therapeutic strategies. The Fox transcriptional regulators (including Foxa, Foxo, Foxm, and Foxl proteins) have well-known roles in liver organogenesis (for review, see Ref. 5). However, in terms of liver metabolism, Foxo1 has gained the most notoriety due to its potent regulation of glucose homeostasis. Since the discovery of Foxo1 as a key regulator of gluconeogenic gene expression, there has been great interest in developing therapies to block hepatic Foxo1 activity in diabetic patients to reduce glycemia. Toward this end, Zhang et al. (4) show that a liver-specific knockout of Foxo1 can reduce blood glucose levels, and that concurrent reduction of Foxo3 further improves glycemia and whole-body insulin sensitivity. These particular data are not surprising, as another recent study reports similar findings (6). Although they differ in their conclusions concerning the role of Foxo4 in glycemia, these papers and others provide strong evidence that reducing Foxo activity in liver may indeed be useful to treat diabetes and metabolic syndrome. However, pertinent information contradicting this dogma is later revealed as Zhang et al. (4) further characterize the knockout metabolic phenotypes. Zhang et al. observe that loss of both Foxo1 and Foxo3 also results in significant hypertriglyceridemia and hypercholesterolemia due to increased hepatic lipid secretion and mild steatosis. These new data not only reveal previously unrecognized metabolic functions for the mostly ignored Foxo3 but also demonstrate that Foxo-regulated metabolic networks may be far more complex than originally thought. By dissecting both global and target gene expression, Zhang et al. (4) suggest that Foxo1 and Foxo3 synergistically suppress key lipogenic pathways, in addition to enhancing genes controlling glucose metabolism. This has significant therapeutic relevance, as it implies that targeting Foxo proteins for the treatment of hyperglycemia may also have unforeseen negative consequences on cardiovascular health. The metabolic outcome of reducing Foxo expression in diabetic models has been of great interest for a number of years (see Table 1 for summary). Thus, evidence of hyperlipidemia in mice on a Foxo-null background can be found in other reports. An independent triple Foxo1/3/4 hepatic Foxo-null mouse line also displays hypertriglyceridemia and increased lipogenesis, but only after exposure to a high-fat diet (7). Unfortunately, this earlier study presents data only for the triple knockout and does not address the relative contributions of individual Foxo proteins. In contrast, Zhang et al. (4) systematically analyze the detailed phenotype of single- and double-knockout mice, shedding light on the role of each family member in liver metabolism. Interestingly, the new study does not include detailed phenotypic or expression data for the triple knockout, with the authors citing lack of an additive effect. These recent reports provide strong evidence that hepatic Foxo family members regulate lipid metabolism. However, whether this is achieved in a coordinated or redundant fashion still remains unclear, as in vitro experiments exploring a synergistic relationship between Foxo1 and Foxo3 on liver gene expression are less convincing. Phenotypic comparison of energy homeostasis in hepatic Foxo-null mouse models HFD, High-fat diet; STZ, streptozotocin; WT, wild type; —, no significant change; NR, not reported. With age. db/db Mouse insulin resistance due to obesity. Phenotypic comparison of energy homeostasis in hepatic Foxo-null mouse models HFD, High-fat diet; STZ, streptozotocin; WT, wild type; —, no significant change; NR, not reported. With age. db/db Mouse insulin resistance due to obesity. Of considerable note, hyperlipidemia caused by Foxo loss is most apparent on insulin-depleted or insulin-resistant backgrounds (Table 1). Insulin potently decreases the activity of all three Foxo proteins (for review, see Ref. 8), and reciprocally, both Foxo1 and Foxo3 feed back to augment insulin activity (9, 10). Thus, direct effects on hepatic insulin signaling may play a more complex and underappreciated role in the phenotype of these mice. Along these lines, a recent study by Haeusler et al. (11) demonstrates that loss of Foxo1 alone increases triglyceridemia and cholesterolemia, but only when insulin is removed from the equation. Paradoxically, the opposite phenotype is observed when hepatic Foxo1 is eliminated in mice lacking the insulin receptor (12). In a model of severe hepatic insulin resistance (hepatic IRS1/2 knockout mice), additional loss of hepatic Foxo1 increases circulating lipids, but only in older mice (12). Complete loss of insulin action is generally not applicable to the typical insulin-resistant type 2 diabetic making these previous studies difficult to interpret in terms of human disease. In contrast, obesity and diets high in fat and sugar are enough to exacerbate hyperlipidemia in double or triple Foxo-null mice (4, 7). Thus, when insulin resistance is diet induced, loss of more than one Foxo is required to observe hyperlipidemia. Taken together, it's clear that the activities of these foxo proteins influence each other and that insulin signaling is involved, but the mechanism by which these transcriptional regulators in liver so dramatically impact whole-body metabolism still remains a mystery. The significant effects on glucose tolerance and lipid metabolism, and relatively modest direct effects on gluconeogenic and lipogenic gene expression, point a finger toward possible inter-organ cross talk. Clues from expression analysis in Foxo-null models suggest that dysregulatation of hormones secreted from liver (hepatokines) may provide the answer. Zhang et al. (4) show that fibroblast growth factor 21 (FGF21) mRNA levels are significantly higher in Foxo3 knockout mice. Loss of Foxo1 also increases circulating FGF21 in mice when insulin is depleted (11) or signaling blocked (12). Interestingly, FGF21 stimulates lipolysis in peripheral tissues to increase circulating free fatty acids and significantly improves insulin sensitivity in mice (for review see Ref. 13), which may explain improved glucose tolerance in Foxo null-mice. This is but one possibility, but it is tempting to speculate that the dramatic metabolic effects of hepatic Foxo loss in vivo may actually be a consequence of indirect changes in adipose or muscle biology. Therefore, the question remains as to whether the observed metabolic effects are primarily due to changes in hepatic energy metabolism or the result of cross talk between liver and other organ systems. There is little doubt that Foxo1 is a predominant regulator of glycemia. However, the impact of Foxo on hepatic lipid homeostasis remains controversial due to an abundance of conflicting data from both gain- and loss-of-function in vivo models (9, 14), leaving more questions than answers at this point. It is clear that we need more information about the activity of these factors at a molecular level, including how they are regulated and their transcriptional targets, before we can understand the complex metabolic phenotypes of the knockout mice. Although the data presented by Zhang et al. (4) and others suggest a cooperative effort between the Foxo family members, there is little direct evidence indicating whether these proteins work in concert on the same targets or independently regulate different gene sets. Moreover, it is still unclear whether the effects on gene expression patterns are actually due to direct transcriptional regulation or indirect effects on other cellular pathways (i.e. insulin signaling). Future studies will benefit from global screening methods, such as chromatin immunoprecipitation sequencing technology and chromatin state maps, to reveal detailed gene networks regulated by this family. Foxo activity and diabetes generally have a negative association. Zhang et al. (4) now highlight that Foxo proteins quietly keep triglyceride and cholesterol levels in check and may actually possess hidden metabolic benefits. This work opens up doors for new therapeutic strategies to target this family. However, warning flags from mouse models should emphasize caution when targeting the activity of this complex transcriptional network for treatment of metabolic disease. Disclosure Summary: The authors have nothing to disclose. Fibroblast growth factor 21 forkhead box protein 01.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.