Hyaluronidase-1 mediates postprandial suppression of hepatic gluconeogenesis

Abstract

Hepatic gluconeogenesis is a critical process that generates glucose from non-carbohydrate precursors during fasting to support vital organs like the brain and red blood cells. Postprandially, this process is rapidly suppressed to allow for glucose storage as glycogen and lipids in the liver. Failure to suppress gluconeogenesis after meals leads to elevated postprandial glucose levels, a key feature of type 2 diabetes. This dynamic switch is regulated by insulin and glucagon, but insulin resistance impairs this regulation. In this study, we identified a novel mechanism involving postprandial circulating hyaluronan (HA) and lysosomal hyaluronidase-1 (HYAL1) that suppresses hepatic gluconeogenesis by rewiring hepatic metabolism and mitochondrial function. Hyal1 knockout (Hyal1 KO) mice exhibited increased gluconeogenesis, while liver-specific Hyal1 overexpression (Liv-Hyal1) mice showed reduced gluconeogenic activity. Transcriptomic analysis revealed minimal changes in liver gene expression due to Hyal1 deletion, but metabolomic profiling demonstrated that Hyal1 overexpression mitigated high-fat diet (HFD)-induced elevations in gluconeogenic pathway metabolites. Mechanistically, HYAL1-mediated HA digestion activates a feedback loop in HA synthesis, repartitioning the cellular uridine diphospho-N-acetyl-D-glucosamine (UDP-GlcNAc) pool. This reduces O-linked N-acetylglucosamine modification (O-GlcNAcylation) of mitochondrial ATP synthase subunits, decreasing ATP production and suppressing gluconeogenesis. Importantly, this pathway remains intact in the livers of HFD-fed, insulin-resistant mice. In summary, our findings reveal a new postprandial mechanism for regulating hepatic gluconeogenesis, highlighting the potential of enhancing postprandial HA levels or hepatic HYAL1 activity as a therapeutic strategy for managing excessive gluconeogenesis in insulin-resistant conditions, such as type 2 diabetes.

Graphical abstract

Keywords

HYAL1 / hyaluronan / hepatic gluconeogenesis / metabolites / mitochondrial function

Introduction

Hyaluronic acid (abbreviated HA), or hyaluronan, is a linear glycosaminoglycan with simple repeating disaccharide units of D-glucuronic acid (GlcA or GlcUA) and N-acetyl-D-glucosamine (GlcNAc) [1, 2]. HA is a significant extracellular matrix component with many critical physiological functions [3]. It is synthesized from UDP-GlcA and UDP-GlcNAc at the plasma membrane by several HA synthases, and the nascent HA chain is extruded through the plasma membrane into the extracellular space [4]. Deletion of the predominant embryonic HA synthase isoform Has2 in mice leads to the lethality of the embryos at mid-gestation (E9.5–10) [5]. Extracellular HA is mostly degraded locally by a family of hyaluronidases (HYALs), and fragmented HA is extracted by lymph and eventually enters the bloodstream [6]. HA is abundant in circulation in a molecular weight range of 100–300 kDa [7], but with an extremely short half-life of 2–3 min, suggesting a rapid turnover in circulation [8]. The liver is the primary organ for clearing circulating HA [9, 10].

Hyaluronidase-1 (HYAL1) and HYAL2 are the two most ubiquitously expressed HYALs. They work together to degrade HA in a stepwise fashion [6], with HYAL2 degrading HA at the cell surface into around 20 kDa fragments, which are then engulfed and delivered to strongly acidic lysosomes for final degradation by HYAL1 and other exoglucosidases [11]. HYAL2 deficiency induces a buildup of very high molecular weight (MW) HA (> 3.10⁶ Da) in the lymph and serum, with severe lymph node distortion [12]. By contrast, HYAL1 deficiency leads to HA overload in the liver and a moderate increase in serum HA concentration without changes to HA MW distribution [12].

HA synthesis and degradation are closely integrated with glucose metabolism in the cell [13]. Synthesis of HA is an energy-consuming process. The major HA synthase, HAS2, is regulated by sirtuin 1 (SIRT1) and the AMP-activated protein kinase (AMPK), two critical energy sensors in the cell, allowing HA synthesis to be synchronized with cellular energy supply states. HA synthesis also depends critically on the size of the cytoplasmic UDP-sugar pool [13]. Glucose metabolism provides energy and precursors through glycolysis for HA biosynthesis; thus, glucose utilization pathways are tightly integrated with HA synthesis. On the other hand, the coordinated action of HYALs, β-glucuronidase, and hexosaminidase ensures the complete degradation of HA, providing GlcA and GlcNAc to the cell. GlcNAc is converted in GlcNAc-6-phosphate (GlcNAc-6-P) by the GlcNAc kinase in the GlcNAc salvage pathway [14, 15] to sustain the synthesis of UDP-GlcNAc for complex glycoconjugates. Vertebrates do not recycle GlcA to GlcA-1-phosphate and then to UDP-GlcA; instead, they convert GlcA into xylulose-5-phosphate through several redox and decar-boxylation reactions [13].

After a meal, circulating HA surges in parallel with the shutting down of gluconeogenesis. However, whether the postprandial HA surge suppresses gluconeogenesis has not been investigated. We showed that the systemic deletion of Hyal1 impaired glucose tolerance in mice treated with HFD, partly due to elevated gluconeogenesis. Liver-specific Hyal1-overexpression mice showed improved glucose tolerance and suppressed gluconeogenesis from pyruvate during the HFD challenge. Measurements of metabolites, the mitochondrial protein O-linked N-acetylglucosamine modification (O-GlcNacylation), and cellular energetics showed an unexpected regulation of hepatocyte gluconeogenesis by HYAL1 via O-GlcNacylation of the mitochondrial ATP synthase subunits. This work links intracellular HA digestion to the regulation of gluconeogenesis in hepatocytes.

Results

Postprandial increase in circulating HA levels

There is a postprandial surge of serum HA levels in human subjects that peaks 1 h after a meal (Fig. 1a). This surge can also be observed in mice (Fig. 1b). Specifically, solid food, including low-fat diet (LFD) and HFD, elicits a quick surge in serum HA levels, whereas PBS and even HA solution fail to elicit such a surge. We reason that this is because HA cannot be absorbed through the intestinal tract, and the increase in serum HA results from intestinal movement and the release of HA from the tissue interstitial space. To test that, we used cisapride, a drug that increases motility in the upper gastrointestinal tract, and loperamide, a drug that inhibits the movement of the gastrointestinal tract. Cisapride treatment, even without solid food, was sufficient to elicit an HA surge similar to LFD or HFD, and treatment of mice with loperamide during feeding with HFD significantly blunted the increase of HA in the circulation (Fig. 1b). Gut-released HA is carried via the inferior vena cava to the portal vein and filtered out through the liver. HYAL1 and HYAL2 are two major enzymes that degrade HA in the liver. Gene expression levels of the two enzymes were not significantly increased after a meal (Supplementary Fig. S1a). Interestingly, the expression levels of HA synthases, especially Has3, were significantly increased 1 h after refeeding (Supplementary Fig. S1b). Given that HYAL1 is the specific enzyme responsible for intracellular HA degradation, we decided to investigate Hyal1 in postprandial HA digestion and metabolic changes.

Hyal1 deficiency increases gluconeogenesis from pyruvate

Systemic deletion of Hyal1 resulted in a significant increase in serum HA levels on both LFD and HFD (Supplementary Fig. S2a). However, there was no further increase of serum HA in Hyal1 KO mice on HFD compared with that on LFD (Supplementary Fig. S2a). Gene expression analysis showed Hyal1 was completely absent from the liver. Deletion of Hyal1 did not lead to a compensatory increase of Hyal2 or Hyal3 in the liver (Supplementary Fig. S2b). Hyal1 was also the only HYAL isoform that upregulated by HFD in the liver (Supplementary Fig. S2b). Deletion of Hyal1 led to a reduction in the expression of Has2 and Has3, the two most abundant HA synthase isoforms in the liver (Supplementary Fig. S2c). In the adipose tissue, HFD increased the expression of Hyal1, Hyal2, and Hyal3. Hyal1 deletion had no effect on Hyal2 or Hyal3 expression (Supplementary Fig. S2d). Similar to what was observed in the liver, Hyal1 deletion led to a reduction in Has1 and Has3 expression, mostly on LFD (Supplementary Fig. S2e), suggesting a common feedback mechanism to reduce HA synthesis after the blocking of HA degradation by Hyal1 deletion.

To understand the role of Hyal1 in the development of metabolic dysfunction, we placed a cohort of male Hyal1 KO mice and their littermate control WT mice on LFD or HFD for 11 weeks. WT mice gained weight (Fig. 2a) and developed hyperglycemia (Fig. 2b) on HFD as compared to LFD. Deletion of Hyal1 did not affect body weight (Fig. 2a) or hyperglycemia (Fig. 2b) relative to WT mice on either LFD or HFD. Hyal1 KO mice on HFD showed higher serum glucose levels at the early timepoints after glucose challenge (Fig. 2c), with similar hyperglycemia-induced insulin levels (Fig. 2d). Hyal1 KO mice fed HFD also showed higher glucose levels during a pyruvate tolerance test (PTT) after overnight fasting, indicating higher gluconeogenesis from pyruvate (Fig. 2e). Sensitivity to intraperito-neally injected insulin did not differ between Hyal1 KO and WT mice (Fig. 2f). By contrast, deletion of Hyal1 did not affect serum lipids (total cholesterol, direct high-density lipoprotein, triglycerides, and non-esterified fatty acids (NEFA)) (Supplementary Fig. S2f). In female mice, Hyal1 KO mice showed no difference in body weight development (Supplementary Fig. S3a) or glycemia (Supplementary Fig. S3b) in comparison to littermate WT mice on LFD. Hyal1 KO females already developed mild glucose intolerance on LFD (Supplementary Fig. S3c) with no difference in glucose-stimulated insulin levels (Supplementary Fig. S3d), which was probably due to increased gluconeogenesis (Supplementary Fig. S3e). Insulin sensitivity was not changed; a ticking up of glucose at 120 min in the Hyal1 KO group might be another indication of higher gluconeogenesis while recovering from hypoglycemia induced by insulin treatment (Supplementary Fig. S3f).

Hepatic Hyal1 overexpression reduces gluconeogenesis in mice on HFD

As the liver is the major gluconeogenic organ [17], we focused on it next. Hepatic Hyal1 expression increased quickly after mice were treated with HFD (Supplementary Fig. S4a). To determine whether this was an adaptation to HFD, we generated a doxycycline-inducible liver-specific Hyal1-overexpression mouse Alb-Cre/Rosa-rtTA/TRE-Hyal1, referred to as LHY TG hereafter. It showed a > 100-fold overexpression of the Hyal1 gene in the liver, no leaky expression in the heart (Supplementary Fig. S4b), and a moderate increase in the HYAL1 protein (Fig. 3a). Hepatic overexpression of Hyal1 did not result in a reduction in hepatic HA content (Fig. 3b) or serum HA levels (Supplementary Fig. S4c); however, they increased postprandial clearance of HA, as they showed a similar spike to control mice but a faster reduction in serum HA levels after a meal (Fig. 3c). LHY TG mice gained similar body weight (Fig. 3d) and showed no difference in glucose tolerance test (GTT) (Fig. 3e) or PTT (Fig. 3f) compared to control mice (Albcre/Rosa-rtTA) on Dox200 LFD. However, when mice were treated with Dox200 HFD, despite similar weight gain (Fig. 3d), the LHY TG mice showed improved GTT (Fig. 3g) and a significant reduction in gluconeogenesis from pyruvate (Fig. 3h), independent of insulin sensitivity (Fig. 3i). In addition, when mice were treated with cisapride to increase gut mobility and circulating HA, the PTT levels in both the control and the LHY TG mice decreased significantly on HFD but not on LFD (Fig. 3f and h). HFD led to severe hepatic steatosis, and Hyal1 overexpression significantly reduced it (Fig. 3j). However, there were no differences in how the mice handled exogenously administered lipids (Fig. 3k). Serum lipid species, including total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, and NEFA, were not affected by Hyal1 overexpression in mice on LFD or HFD (Supplementary Fig. S4d). Interestingly, Hyal1 overexpression markedly upregulated hepatic Has expression in mice on HFD (Fig. 3l). Together with the down-regulation of Has expression observed following Hyal1 deletion in the liver (Supplementary Fig. S2c), these data suggest a reciprocal regulation of HA synthesis by HA degradation in hepatocytes.

The effect of Hyal1 on gluconeogenesis is transcription independent

To understand how Hyal1 might regulate gluconeogenesis in the liver, we performed an RNA sequencing (RNA-seq) of WT and Hyal1 KO livers harvested from mice kept on LFD and HFD. The t-distributed stochastic neighbor embedding (t-SNE) projection of variances among the samples revealed a major transcriptional difference determined by diet but not by genotype (Fig. 4a). Volcano plots comparing Hyal1 KO to WT mice on LFD showed that few genes were significantly changed (Fig. 4b), and a similar lack of significantly altered genes between Hyal1 KO and WT mice was evident for mice treated with HFD (Fig. 4c). Correlation of the significantly changed genes in the Hyal1 KO mice with WT controls on LFD and HFD suggested that few genes showed the same direction of regulation (Fig. 4d). Only two genes showed the same direction of change (downregulation) with LFD and HFD, including Hyal1 itself (Fig. 4e); the other was guanine nucleotide binding protein, alpha transducing activity polypeptide 1 (Gnat1). However, GNAT1 protein levels and reported functions are restricted to the retina [18, 19]. In addition, Gnat1 expression was suppressed in Hyal1-overexpression liver tissue on HFD (Fig. 4f), similar to the results of Hyal1 deletion in the liver, arguing against a possible role of Gnat1 in the Hyal1-mediated metabolic phenotype. A complete list of the detected genes is in Supplementary Table S1.

Hepatic Hyal1 overexpression reduces metabolites in the gluconeogenesis pathway

Because HYAL1 was unlikely to regulate gluconeogenesis by modulating gene expression, we decided to determine whether it directly affected metabolites in the liver. We measured metabolites from gluconeogenesis and the tricarboxylic acid (TCA) cycle pathway in control and Hyal1-overexpressing livers harvested from age-matched male mice treated with LFD or 12-week HFD. All mice were fasted overnight to allow their livers to enter a gluconeogenesis state. HFD significantly increased the pool of glucose/fructose, glucose-6-phosphate (G6P)/fructose-6-phosphate (F6P), and lactate (Supplementary Fig. S5a). The exact identities of glucose and fructose, as well as that of G6P and F6P, could not be distinguished by mass spectrometry. However, as glucose is much more abundant than fructose intracellularly, these data primarily reflected higher glucose concentrations in HFD livers. On LFD, Hyal1 overexpression did not affect metabolites in the gluconeogenesis pathway much except for a mild reduction in glucose (Fig. 5a). On HFD, however, Hyal1-overexpressing liver had most of the gluconeogenesis pathway metabolites reduced (Fig. 5b). HFD feeding increased metabolites in the TCA cycle (Supplementary Fig. S5b), but Hyal1 overexpression did not affect TCA metabolites on LFD and HFD (Fig. 5c and d). A complete list of relative abundances of metabolites measured is given in Supplementary Table S2. Plasma membrane glucose transporters GLUT1 and GLUT2 were not affected by Hyal1 overexpression in the liver (Fig. 5e), excluding the possibility of glucose export as the cause of lower glucose levels in PTT in LHY TG mice. All these data indicate that HYAL1 directly modulates the intracellular metabolite pool in the gluconeogenesis pathway, mostly on HFD (Fig. 5f).

HYAL1-mediated HA degradation reduces mitochondrial ATP production and inhibits gluconeogenesis in hepatocytes

Gluconeogenesis is governed by key enzymes such as pyruvate carboxylase (PC) and phosphoenolpyruvate carboxykinase (PEPCK) [17]. HFD significantly increased PEPCK but not PC levels in the liver (Fig. 6a). However, protein abundance for both enzymes was not affected by Hyal1 overexpression in the liver (Fig. 6a). Because gluconeogenesis is an energy-consuming process, we performed cellular energy measurements to understand how Hyal1 overexpression affects cellular respiration. ATP-coupled oxygen consumption rate (OCR), an indicator of ATP generation, was significantly reduced in the Hyal1 overexpressing AML12 cells (Fig. 6b); a similar pattern was also observed in HepG2 cells (Supplementary Fig. S6). The change in ATP generation is unlikely to be a result of changes in mitochondrial oxidative phosphorylation system (OXPHOS) protein complexes (Fig. 6c). By contrast, digestion of extracellular HA by bovine testis HYAL (BTH) did not affect the OCR values (Fig. 6d). All these data suggest that hepatocyte Hyal1 remodels mitochondrial energetics, which may be responsible for reduced gluconeogenesis.

To evaluate the interactions between HYAL1-mediated HA degradation, metabolic hormones, and HFD-induced insulin resistance in glucose production, we isolated primary hepatocytes from control and LHY TG mice and measured glucose production under different conditions. First, hepatocytes isolated from overnight-fasted mice exhibited increased glucose output only in response to glucagon and dibutyryl-cAMP (db-cAMP), a membrane-permeable analog of cAMP, while HYAL1 showed a trend towards suppressing gluconeogenesis in both conditions. Notably, under low basal gluconeogenesis, insulin did not exert any additional suppression of glucose production (Fig. 6e). In the subsequent experiment, we included db-cAMP in all treatment conditions to mimic glucagon action in the fasting state. Then, we treated either isolated hepatocytes in the vehicle or free fatty acids (FFAs) for 16 h to induce insulin resistance and showed that the suppression of glucose output by insulin was significantly blunted by FFA treatment, but Hyal1 expression still exerted additional inhibition of glucose output beyond the effect of insulin (Fig. 6f).

HYAL1-mediated HA degradation reduces O-GlcNAcylation of ATP synthase subunits and reduces the enzymatic activity of ATP synthase

How does HYAL1-mediated HA degradation affect ATP production? HA degradation releases GlcNAc from the lysosomes. GlcNAc is quickly phosphorylated to reform UDP-GlcNAc via enzymatic reactions, which can then be reused for HA synthesis, O-GlcNAcylation, or other biological pathways. In AML12 cells stably expressing Hyal1, we found that overall cellular O-GlcNAcylation remained unchanged; however, mitochondrial O-GlcNAcylation was reduced (Fig. 7a). This was independent of the levels of O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA), two key enzymes responsible for the dynamic process of O-GlcNAcylation (Fig. 7b). According to the literature, more than 10 proteins in mitochondria can be O-GlcNAcylated [20]. Among them, mitochondrial ATP synthase subunits 5A and 5B are essential components of the ATP synthase complex and play a crucial role in ATP synthesis during cellular respiration [21]. Immunoprecipitation (IP) of mitochondrial O-GlcNAc proteins using the RL2 anti-O-GlcNAcylated protein antibody, followed by detection of specific proteins, showed that the O-GlcNAcylation levels of ATP synthase subunits alpha and beta (ATP5A and ATP5B) were significantly reduced in Hyal1 expressing cells, while the protein abundance was not affected by Hyal1 expression (Fig. 7c).

To understand how mitochondrial protein O-GlcNAcylation was reduced independent of changes in OGT and OGA abundances, metabolites in the amino sugar pathway were quantified in whole AML12 cells, isolated lysosomes, and isolated mitochondria. HA will completely break down into GlcA and GlcNAc. GlcA and its derivative UDP-GlcA were significantly increased in Hyal1-overexpressing AML12 cells (Fig. 7d). The other product, GlcNAc, was not detected, probably due to quick phosphorylation to GlcNAc-6-P by GlcNAc kinase. Consistently with this, we observed elevated GlcNAc-6-P levels in Hyal1-overexpressing AML12 cells (Fig. 7d). Despite a similar increase of GlcA in mitochondria, GlcNAc-6-P was not different, and UDP-GlcNAc, the substrate of OGT and O-GlyNAcylation, decreased in the mitochondria (Fig. 7e). None of those metabolites were different in isolated lysosomes (Fig. 7f), indicating the dynamic nature of lysosomes and quick trafficking of those metabolites. In addition, due to numerous phosphatases and nucleotidases [22], UDP-GlcNAc and UDP-Glc were not detected in the lysosome (Fig. 7f).

The decrease in mitochondrial O-GlcNAcylation corresponded with a reduction in ATP synthase activity (Fig. 7g), aligning with the results previously reported in HeLa cells [23]. All these data point to a repartitioning of the subcellular UDP-GlcNAc that mediates mitochondrial protein O-GlcNAcylation and mitochondrial function, subsequently regulating mitochondrial ATP production, eventually contributing to the suppression of gluconeogenesis (Fig. 7h).

Discussion

This paper studied the role of postprandial surges of circulating HA in hepatic gluconeogenesis. Starting by observing impaired glucose tolerance and elevated gluconeogenesis in Hyal1 KO mice, we pinned down the liver as the primary organ contributing to this metabolic phenotype on the basis of high expression of Hyal1 in the liver, hepatic HA accumulation in Hyal1 KO mice [12], and the fact that the liver is the primary organ for gluconeogenesis [17]. Then, with hepatic Hyal1 overexpression mice, primary hepatocytes, and hepatocyte cell lines, we identified the role of HYAL1-mediated HA degradation in suppressing gluconeogenesis.

Hepatic metabolism is actively regulated by the expression of enzymes catalyzing key reactions in these metabolic pathways. However, Hyal1 deletion did not affect gene expression overall. So, how does HYAL1 regulate gluconeogenesis? In fact, the HA catabolism end product GlcNAc is released from the lysosome and converted into GlcNAc-6-P and UDP-GlcNAc to regulate cellular signaling and metabolism. This is done through subcellular compartmentalization of the UDP-GlcNAc and reduced allocation of UDP-GlcNAc to mitochondria to specifically reduced O-GlcNAcylation levels of ATP5A and ATP5B, resulting in reduced mitochondrial ATP synthase activity and reduced ATP production, which are expected to affect gluconeogenesis.

Compartmentalization of the cells and heterogenous subcellular distribution of metabolites can be achieved with membrane organelles, by phase separation, or by specialized subcellular location of key enzymes, which work together with biophysical limits of diffusion to form a static-appearing concentration gradient within the cell. Our study showed a significant increase in Has expression in livers overexpressing Hyal1 (Fig. 3l) or after refeeding (Supplementary Fig. S1b), suggesting a compensatory HA synthesis at the plasma membrane under those scenarios [13]. This compensatory HA synthesis may drive a flux of cellular UDP-GlcNAc towards the plasma membrane, leading to reduced allocation to mitochondria.

It is essential to point out that mitochondrial protein O-GlcNacylation is a debated topic, especially regarding whether this modification occurs in the mitochondria or before these proteins are imported into the mitochondria, because OGT is localized predominantly in the nucleus and cytoplasm [24, 25], with only a small fraction in the mitochondria. It is possible that mitochondrial OXPHOS complexes (e.g. ATP5A/ATP5B) are synthesized on the endoplasmic reticulum (ER) and undergo O-GlcNacylation before they are imported into the mitochondria. Further, it remains possible that there is a UDP-GlcNAc gradient that is low around the ER, leading to reduced O-GlcNacylation of mitochondrial proteins. It is also likely that both processes occur in parallel and contribute collectively to this postprandial metabolic regulation.

We also observed that the effect of HYAL1 on gluconeogenesis occurred primarily in mice subjected to HFD and that HYAL1 still inhibited glucose production on top of insulin in FFAs-induced insulin-resistant primary hepatocytes. The differential impact can be attributed to the metabolic stress imposed by HFD and exposure to FFAs. In the classical view, postprandial suppression of gluconeogenesis is mediated by the surge of insulin and decrease of glucagon after a meal: insulin suppresses gluconeogenesis directly in the liver and indirectly via other peripheral tissues, such as adipose tissue, by suppressing the release of FFAs and glycerol. With HFD-induced obesity, the fatty liver becomes insulin-resistant and inefficient in these processes, failing to suppress gluconeogenesis and contributing to hyperglycemia. In contrast, HYAL1-HA regulates gluconeogenesis differently: HAYL1 is constantly expressed, and its effect is enhanced by a surge of postprandial blood HA levels, which have the same magnitude for HFD and LFD. At the molecular level, the HYAL1-HA pathway affects ATP production to regulate the same gluconeogenesis pathway that insulin and glucagon regulate. Because the HYAL1-HA pathway is not affected by insulin resistance, it manifests a more significant effect in fatty livers from HFD-treated mice. In addition, ATP levels are lower in those livers than in lean and insulin-sensitive mice due to mitochon-drial dysfunction [26, 27] and increased uncoupling [28], making the livers of HFD mice more vulnerable to decreased ATP and thus causing a more pronounced suppression of gluconeogenesis. All these data support a model that HYAL1-HA, glucagon, and insulin convene on the same metabolic pathway in regulating glucose usage and gluconeogenesis.

Physiologically, gluconeogenesis generates glucose from non-carbohydrate precursors to support critical organs like the brain and red blood cells during fasting and other conditions when dietary glucose is unavailable. It is tightly regulated, and failure to suppress it after a meal leads to high postprandial glucose levels in type 2 diabetic patients [29]. In parallel to insulin action, the direct release of HA from the gastrointestinal tract, which then travels through the portal vein to the liver to be digested by hepatocyte HYAL1 to inhibit gluconeogenesis, offers another level of regulation of postprandial gluconeogenesis, which is especially important in the livers with insulin resistance. A lack of such regulation would lead to increased gluconeogenesis, explaining why the anti-diarrhea medicine loperamide unwantedly increases blood glucose levels in human subjects [30].

In conclusion, we found that HYAL1-mediated HA degradation inhibits gluconeogenesis in parallel to classical insulin action in the liver of HFD-fed mice—a mechanism that could be potentially enhanced to inhibit gluconeogenesis in insulin-resistant patients.

Limitations of the study

While our findings underscore the importance of HYAL1-mediated HA degradation in suppressing gluconeogenesis, the relative contribution of this pathway compared to classical insulin signaling remains unclear. Moreover, our study primarily assessed gluconeogenesis from precursors such as pyruvate, using glucose as the output measure, and did not evaluate glycerol-derived gluconeogenesis, which may be differentially influenced by HYAL1-HA metabolism. Future studies employing more advanced stable isotope tracing techniques to quantify the interplay between insulin, HA, and alternative gluconeogenic substrates could offer a more comprehensive understanding of postprandial glucose regulation.

Materials and methods

Animals

Hyal1^tm1Stn/Mmcd (Hyal1 KO) mice were generated previously [31] and obtained from Mutant Mouse Regional Resource Centers (MMRRC). The mice were backcrossed to a C57Bl/6 genetic background for nine generations before the experiments [32]. TRE-Hyal1 was generated by subcloning mouse cDNA into a pTRE vector (Clontech) with a rabbit β-globin 3'-UTR. Linearized DNA was injected into an oocyte, and the founders were identified through PCR reactions. The founders were then crossed with Adipoq-rtTA mice (Jax #033448) [33] or albumin-Cre transgenic (Jax #003574) [34] and Rosa26-loxP-STOP-loxP-rtTA transgenic (Jax #005572) [35] double transgenic mice to select a founder mouse that could achieve tissue-specific transgene induction after doxycycline treatment and had no leakage of transgene expression into other tissues for the subsequent experiments. All animals were kept on a 12-h light/12-h dark cycle in a temperature-controlled environment. Mice were free to access water and were fed one of the following: a standard chow diet (LFD), a 60% HFD (BioServ, S1850), or a 60% HFD containing 200 mg/kg doxycycline (BioServ, S6223). Mice were genotyped by Transnetyx. The Hyal1 KO study was performed at the University of Namur; the rest of the animal studies were performed at the Baylor College of Medicine. Animal care and experimental protocols were approved by the Institutional Animal Care and Use Committee of the University of Namur and the Baylor College of Medicine.

HA extraction and quantification

The procedure was performed as previously described [16], and the extracted HA was quantified with an ELISA kit (R&D systems, DHYAL0), using a sample digested with BTH overnight as the negative control. The tissue HA content was normalized to the wet weight used for extraction.

Histological analysis

After the mice were euthanized, the tissue was excised immediately, fixed overnight in 10% PBS-buffered formalin, and stored in 50% ethanol. Tissues were sectioned (5 µm), rehydrated, and stained with hematoxylin and eosin (H&E) at the Pathology Core of the Baylor College of Medicine. Microscopic images were taken on a ZEISS Axioscan scanner.

Western blotting

Protein extractions were performed as previously described [36]. Protein abundance was detected using primary antibodies against GLUT1 (Novus Biologicals, NB110-39113SS, 1:1,000 dilution), GLUT2 (Novus Biologicals, NBP2-22218SS, 1:500 dilution), PEPCK/PCK2 (Abclonal, A4466, 1:1,000 dilution), PC (Abclonal, A8980, 1:1,000 dilution), O-GlcNAc (Novus Biologicals, NB300-524SS, 1:1,000 dilution), total OXPHOS (Abcam, ab110413, 1:1,000 dilution), ATP5B (Novusbio, NBP3-15354, 1:1,000 dilution), HSP60 (Novusbio, NBP1-77397SS, 1:1,000 dilution), or β-Actin (Cell Signaling Technology, 3700S, 1:1,000 dilution). Secondary antibodies used included goat anti-mouse, Alexa Fluor Plus 800 (Invitrogen, A32730), and goat anti-rabbit, Alexa Fluor Plus 680 (Invitrogen, A32734), both at 1:10,000 dilutions. Antibody-decorated membranes were then visualized on a Li-Cor Odyssey infrared scanner, and the scanned data were analyzed using Odyssey version 3.0 software.

Metabolic measurements

GTTs were performed as described previously [16]. In brief, the mice were fasted for 4−6 h, after which glucose solution (10 µL/g body weight) was administered orally or by intraperitoneal injection; the final dose of glucose was 1.25 g/kg body weight for the mice on LFD and 0.75 g/kg body weight for the mice on HFD. Blood glucose levels before and 15, 30, 60, and 120 min after the glucose injection were measured using a glucometer. Serum insulin levels were measured using the ALPCO Mouse Insulin ELISA Jumbo kit (cat. number 80-INSMS-E10). PTTs were performed on the overnight fasted mice with or without cisapride (Cayman Chemical Company, 21657) treatment. Basal blood glucose levels were measured in the morning. Cisapride was dissolved in DMSO and then diluted in PBS to 0.05 mg/mL. It was administered by oral gavage at 10 μL/g body weight (0.5 mg/kg body weight) or vehicle (0.5% DMSO in PBS) to mice 1 h before pyruvate injection. Sodium pyruvate was dissolved in water and injected intraperitoneally at a final dose of 1 g/kg body weight. Blood glucose levels before and 15, 30, 60, and 120 min after the glucose injection were measured using a glucometer. Insulin tolerance tests (ITTs) were performed on mice fasted for 4−6 h. Human insulin (Sigma (Roche), 11376497001) was dissolved in saline and injected peritoneally; the final dose was 1 IU/kg body weight. Blood glucose levels before and 15, 30, 60, and 120 min after the glucose injection were measured using a glucometer. Triglyceride clearance tests were performed as described previously [37]. In brief, the mice were fasted overnight, and then 20% intralipid (Sigma, I141-100 mL) was orally gavaged at 15 µL/g body weight. Blood glucose levels before and 1.5, 3, and 6 h after intralipid administration were taken from the tail, and serum triglyceride levels were measured using Infinity triglyceride reagents (Thermo Fisher, TR22421).

Serum lipid measurements

Serum triglycerides were measured using Infinity reagents. Serum NEFA was measured by NEFA-HR assay (Wako, C1057). Total cholesterol and HDL cholesterol were measured using an AF HDL and low-density lipoprotein/very low-density lipoprotein (LDL/VLDL) assay kit (Sigma-Aldrich, MAK331). The measurements were carried out following manufacturer-provided protocols.

Metabolomics

The metabolites were extracted from the mouse liver tissues [38]. The TCA cycle and glycolysis metabolites and their intermediates were separated using a Luna 3 µm NH2 (100 A°) HPLC column, and 5 mmol/L ammonium acetate in water (pH 9.9) and acetonitrile were used as mobile phase solvents. The metabolites were separated through the Agilent 1290 Infinity HPLC system, and the data were acquired using Agilent 6495B Triple Quadrupole mass spectrometry via multiple reaction monitoring (MRM) in negative ionization mode [38, 39]. These data were analyzed using Agilent Mass Hunter quantitation software. The data were normalized with a spiked isotopically labeled internal standard, and relative abundance was plotted.

Quantitative real-time PCR (qRT-PCR)

Using a reverse transcription kit (Bio-Rad), 1 µg RNA was used to transcribe cDNA. qRT-PCR primers were obtained from the Harvard PrimerBank [40]; these are listed in Supplementary Table S3. The messenger RNA levels were calculated using the comparative threshold cycle method, normalized to gene Rps16.

RNA-seq

RNA was isolated from frozen tissues by homogenization in Trizol reagent (Invitrogen, 15596018) as previously described [41]. RNA concentrations were quantified using a NanoDrop Spectrophotometer, and sample integrity was verified using an Agilent 2100 Bioanalyzer (Agilent Technologies). Only samples with RNA integrity number (RIN) values above 8.0 were used for experiments. cDNA libraries were prepared using an Illumina TruSeq RNA sample prep kit. The average size of the library cDNAs was 150 bp (excluding adapters). The integrity and quality of the cDNA libraries were assessed using an Agilent 2100 Bioanalyzer and an ABI StepOne Plus real-time PCR system. RNA-seq was performed by the Novogene. Raw read sequencing quality and adapter contamination were assessed using FastQC v0.11.9. The overall quality was determined to be satisfactory, and raw reads were aligned to the mouse genome index using STAR v2.7.9a. The STAR genome index was created using raw FASTA and annotation files downloaded from the GENCODE portal for mouse genome, build GRCm38, release 23. Alignments were saved in binary format (BAM). Summaries of raw read quality and alignment quality were generated using MultiQC v1.12 [42]. Sample-specific gene expression values were computed as the number of reads aligned per gene using STAR–quantMode GeneCounts. Raw counts were normalized, and genes with an average read count of < 50 across all samples were considered unexpressed and were excluded from the differential analysis. The analysis for differential gene expression was carried out using DESeq2 [43]. A false discovery rate (FDR) cut-off of 0.05 and a fold change cut-off of 20% (−0.263 ≤ log₂(Fold change) ≥ +0.263) were imposed to identify significantly differentially expressed genes. Genes identified in different contrasts were then overlapped to determine the directionality of differential gene expression between the contrasts. The accession number for the RNA-seq data is GEO: GSE220752.

IP

AML12 cells were transfected with plasmids encoding HYAL1 or YFP. Stable transgenic cell lines were established by hygromycin (Thermo Fisher, 10687010) selection.

Direct IP with O-GlcNAc antibody or ATP5B antibody was carried out as follows. O-GlcNAc antibody RL2 (Novusbio, NB300-524) or ATP5B antibody (Novusbio, NBP3-15354) was conjugated to NHS-activated magnetic beads (Thermo Scientific, 88828). The antibody-coupled beads were then incubated with AML12 YPF/HYAL1 mitochondrial protein for 2 h at room temperature, with or without 6.7 mmol N-Acetylglucosamine (Sigma-Aldrich, PHR1432-1G). After incubation, the precipitates were washed with cell lysis buffer to remove non-specifically bound proteins. The bound proteins were then eluted and analyzed with western blot analysis.

Classic IP with ATP5B antibody was carried out as follows. Mitochondrial protein lysates from YFP and Hyal1-overexpressing cells were incubated with ATP5B antibody for 2 h at room temperature. The immune complexes were captured using protein A/G magnetic beads. Western blot analysis was performed to compare the O-GlcNAcylation levels between the YFP and Hyal1 overexpressing mitochondrial protein samples, using RL2 or 9D1.E4 antibodies (Invitrogen, MA1-039).

ATP synthase activity assay

ATP synthase activity was measured using the ATP Synthase Enzyme Activity Microplate Assay Kit (ab109714) following the manufacturer's protocol. Mitochondria were isolated from AML12 cell lines stably expressing mCherry or HYAL1 as previously described [44]. The isolated mitochondria were lysed and diluted to the desired concentration in 4-fold volume of a buffer solution. Then 1/10 volume of the provided detergent was added, mixed thoroughly, and incubated on ice for 30 min. The mixture was centrifuged, and the supernatant was collected for further analysis. For the assay, 50 µL of the sample was added to each well of a 96-well microplate coated with a specific monoclonal antibody, and incubated at room temperature for 3 h to ensure complete enzyme−antibody binding. The plate was then washed twice with buffer solution to remove unbound material. Subsequently, 40 μL of LIPID MIX was added to each well, and the reaction was incubated at room temperature for 45 min. Following this, 200 µL of the detection reagent was added to each well. Absorbance at 340 nm was measured at 30°C using a kinetic program on a microplate reader (BMG Labtech FLUOstar Omega Microplate Reader) over 60−120 min. ATP synthase activity was determined by evaluating the rate of decrease in absorbance at 340 nm over time.

Measurement of cellular energetics

The experiment was carried out on the Agilent Seahorse XFe24 instrument in accordance with the Agilent Seahorse protocol. In detail, HepG2 cells were transfected with indicated constructs (1 µg/well in a 12-well plate using PolyJet DNA transfection reagent (SignaGen)) and seeded into an Agilent Seahorse XF24 cell culture microplate at a density of 60,000 cells per well (Wells A1, B4, C3, and D6 were used as the background, without cells) on the second day. In parallel, an Agilent Seahorse XFe24 extracellular flux assay plate with 1 mL Seahorse XF calibrant solution was equilibrated overnight at 37°C. The next day, a Mito stress test was performed using Palmitate-BSA as described, using an in vitro Seahorse XF Cell Mito Stress Test Kit, with the sequential injection of the following compounds: Port A: oligomycin, with a final concentration of 2 µmol/L; Port B: carbonyl cyanide-p-(trifluoromethoxy)phenylhydrazone (FCCP), with a final concentration of 4 µmol/L; Port C: rotenone, with a final concentration of 1 µmol/L, and antimycin A, with a final concentration of 4 µmol/L. All final concentrations were in the culture medium after injection.

Hepatocyte isolation

Primary hepatocytes were isolated from control and LHY TG mice fasted for 24 h to deplete liver glycogen, following a previously described protocol [45]. In brief, 8- to 10-week-old mice were anesthetized with 100 mg/kg ketamine (Covetrus) and 10 mg/kg xylazine (Rompun), and then perfused at 3 mL/min for 15 min (37°C) with a washing buffer (Hanks balanced salt solution (HBSS) without Ca²⁺/Mg²⁺/phenol red (ThermoFisher, 88284), 25 mmol/L HEPES, and 0.5 mmol/L EDTA, pH 7.4) to remove blood. The liver was then digested (2 mL/min, 10 min, 37°C) with HBSS containing Ca²⁺/Mg²⁺/phenol red (ThermoFisher, 14025076), 25 mmol/L HEPES (Boston Bioproducts, BBH-74), and 25 μg/mL Liberase (Sigma Aldrich, 5401119001). After excision, the liver was placed on ice, ruptured to release cells, filtered (70 μm), and centrifuged (50 g, 2 min, 4°C). Hepatocytes were further purified by Percoll (VWR, 89428-524) gradient centrifugation (200 g, 10 min, 4°C), and then resuspended in Williams' Medium E (Thermo Fisher, 32551020) with 10% fetal bovine serum (Thermo Fisher, 16140071), penicillin (100 UI/mL), and streptomycin (100 μg/mL) (Sigma Aldrich, P4333-100mL). Only isolates with ≥ 90% viability were used. Cells (1.25 × 10⁵/well, 24-well plates) were seeded on collagen-coated plates, incubated for 4 h, and then switched to a fresh medium for further experimentation.

Preparation of 1-mmol/L palmitic acid (PA):oleic acid (OA) (1:2) FFA-BSA complex solution

A 1-mmol/L PA:OA (1:2) FFA-BSA complex was prepared by mixing PA and OA bound to BSA. Specifically, 660 μL of BSA-palmitate complex (5 mmol/L PA and 0.8 mmol/L BSA; Cayman Chemical, 29558) and 2.23 mL of OA-albumin solution (3.0 mmol/L OA and 1.5 mmol/L BSA; Sigma-Aldrich, 03008) were combined. The mixture was diluted to a final volume of 10 mL with fatty acid-free DMEM (approximately 7.11 mL of buffer). The solution was gently heated at 37°C and vortexed to ensure complete mixing. The final solution contained 1 mmol/L total fatty acids (PA:OA = 1:2) and approximately 0.36 mmol/L BSA. Subsequently, the solution was filtered through a 0.22-μm membrane for sterilization. The prepared medium was immediately used to replace the cell culture medium. Equal amounts of BSA were added to the control cells in each experiment.

Assessment of glucose production in primary hepatocytes

Glucose production in primary hepatocytes was evaluated as previously described [46]. In brief, cells were cultured in Williams' Medium E supplemented with 10% fetal bovine serum, followed by three washes with warm PBS to remove residual glucose. Cells were then treated with 100 μmol/L db-cAMP (Sigma Aldrich, D0260-5mg), 100 nmol/L insulin, or 100 nmol/L glucagon (Genscript, RP10772) in glucose-free Dulbecco's Modified Eagle's Medium (Gibco, 11966025) containing gluconeogenic substrates (20 mmol/L sodium lactate (Sigma, 1614308) and 2 mmol/L sodium pyruvate (Sigma, 792500-100G)) and 200 μg/mL sodium hyaluronate (Lifecore Biomedical, HA200K-1, MW = 200 kDa). For the insulin resistance model, cells were pretreated with 1 mmol/L oleate/palmitate (2:1) mixture for 16 h, thoroughly washed with warm PBS, and subsequently subjected to the same stimulation protocol. After 6 h, the cell culture supernatants were collected to measure the glucose content using a Roche glucose assay kit (cat. no. 0716251) and normalized to total protein content. Total glucose production comprised contributions from both glycogenolysis and gluconeogenesis. Glycogenolysis-derived glucose production was measured in the absence of gluconeogenic substrates, while gluconeogenic glucose production was determined as the difference between total glucose output and glycogenolysis.

Statistical analysis

Results are shown as mean ± standard error of the mean (SEM). For experiments with only two groups, Student's t-test was used. For studies with three or more groups, one-way ANOVA was used; for experiments with several groups with a balanced distribution of two factors, a two-way ANOVA test was used. A Sidak test was used for post-hoc analysis of comparisons within subgroups. P values of < 0.05 were considered statistically significant. More details are provided in the figure legends.

References

[1] Zhu Y, Crewe C, Scherer PE. Hyaluronan in adipose tissue: beyond dermal filler and therapeutic carrier. Sci Transl Med 2016; 8: 323ps4.

[2] Zhu Y, Kruglikov IL, Akgul Y et al. Hyaluronan in adipogenesis, adipose tissue physiology and systemic metabolism. Matrix Biol 2018; 78–79: 284-91.

[3] Hascall V, and Esko JD. In: Varki A, Cummings RD, Esko JD, Stanley P, Hart GW, et al. eds. Essentials of Glycobiology. Cold Spring Harbor (NY); 2015: 197-206.

[4] Weigel PH, Hascall VC, Tammi M. Hyaluronan synthases. J Biol Chem 1997; 272: 13997-4000.

[5] Camenisch TD, Spicer AP, Brehm-Gibson T et al. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J Clin Invest 2000; 106: 349-60.

[6] Csoka AB, Frost GI, Stern R. The six hyaluronidase-like genes in the human and mouse genomes. Matrix Biol 2001; 20: 499-508.

[7] Tengblad A, Laurent UB, Lilja K et al. Concentration and relative molecular mass of hyaluronate in lymph and blood. Biochem J 1986; 236: 521-5.

[8] Fraser JR, Laurent TC. Turnover and metabolism of hyaluronan. Ciba Found Symp 1989; 143: 41-53.

[9] Eriksson S, Fraser JR, Laurent TC et al. Endothelial cells are a site of uptake and degradation of hyaluronic acid in the liver. Exp Cell Res 1983; 144: 223-8.

[10] Alston-Smith J, Pertoft H, Laurent TC. Endocytosis of hyaluronan in rat Kupffer cells. Biochem J 1992; 286: 519-26.

[11] Puissant E, Gilis F, Dogne S et al. Subcellular trafficking and activity of Hyal-1 and its processed forms in murine macrophages. Traffic. 2014; 15: 500-15.

[12] Bourguignon V, Flamion B. Respective roles of hyaluronidases 1 and 2 in endogenous hyaluronan turnover. FASEB J 2016; 30: 2108-14.

[13] Caon I, Parnigoni A, Viola M et al. Cell energy metabolism and hyaluronan synthesis. J Histochem Cytochem 2021; 69: 35-47.

[14] Love DC, Hanover JA. The hexosamine signaling pathway: deciphering the “O-GlcNAc code”. Sci STKE 2005; 2005: re13.

[15] MacNicholl AD, Wusteman FS, Winterburn PJ et al. Degradation of [³H]chondroitin 4-sulphate and re-utilization of the [³H] hexosamine component by the isolated perfused rat liver. Biochem J 1980; 186: 279-86.

[16] Zhu Y, Li N, Huang M et al. Adipose tissue hyaluronan production improves systemic glucose homeostasis and primes adipocytes for CL 316,243-stimulated lipolysis. Nat Commun 2021; 12: 4829.

[17] Rui L. Energy metabolism in the liver. Compr Physiol. 2014; 4: 177-97.

[18] Van Dop C, Medynski DC, Apone LM. Nucleotide sequence for a cDNA encoding the alpha subunit of retinal transducin (GNAT1) isolated from the human eye. Nucleic Acids Res 1989; 17: 4887.

[19] Uhlen M, Fagerberg L, Hallstrom BM et al. Proteomics. Tissue-based map of the human proteome. Science 2015; 347: 1260419.

[20] Cao W, Cao J, Huang J et al. Discovery and confirmation of O-GlcNAcylated proteins in rat liver mitochondria by combination of mass spectrometry and immunological methods. PLoS One 2013; 8: e76399.

[21] Acin-Perez R, Beninca C, Fernandez Del Rio L et al. Inhibition of ATP synthase reverse activity restores energy homeostasis in mitochondrial pathologies. EMBO J 2023; 42: e111699.

[22] Kresse H, Glossl J. Glycosaminoglycan degradation. Adv Enzymol Relat Areas Mol Biol 1987; 60: 217-311.

[23] Cha MY, Cho HJ, Kim C et al. Mitochondrial ATP synthase activity is impaired by suppressed O-GlcNAcylation in Alzheimer's disease. Hum Mol Genet 2015; 24: 6492-504.

[24] Kreppel LK, Blomberg MA, Hart GW. Dynamic glycosylation of nuclear and cytosolic proteins. Cloning and characterization of a unique O-GlcNAc transferase with multiple tetratricopeptide repeats. J Biol Chem 1997; 272: 9308-15.

[25] Love DC, Kochan J, Cathey RL et al. Mitochondrial and nucleo-cytoplasmic targeting of O-linked GlcNAc transferase. J Cell Sci 2003; 116: 647-54.

[26] Aharoni-Simon M, Hann-Obercyger M, Pen S et al. Fatty liver is associated with impaired activity of PPARγ-coactivator 1α (PGC1α) and mitochondrial biogenesis in mice. Lab Invest 2011; 91: 1018-28.

[27] Cortez-Pinto H, Chatham J, Chacko VP et al. Alterations in liver ATP homeostasis in human nonalcoholic steatohepatitis: a pilot study. JAMA 1999; 282: 1659-64.

[28] Chavin KD, Yang S, Lin HZ et al. Obesity induces expression of uncoupling protein-2 in hepatocytes and promotes liver ATP depletion. J Biol Chem 1999; 274: 5692-700.

[29] Barroso E, Jurado-Aguilar J, Wahli W et al. Increased hepatic gluconeogenesis and type 2 diabetes mellitus. Trends Endocrinol Metab 2024; 35: 1062-77.

[30] Caldara R, Testori GP, Ferrari C et al. Effect of loperamide, a peripheral opiate agonist, on circulating glucose, free fatty acids, insulin, C-peptide and pituitary hormones in healthy man. Eur J Clin Pharmacol 1981; 21: 185-8.

[31] Martin DC, Atmuri V, Hemming RJ et al. A mouse model of human mucopolysaccharidosis IX exhibits osteoarthritis. Hum Mol Genet 2008; 17: 1904-15.

[32] Dogne S, Rath G, Jouret F et al. Hyaluronidase 1 deficiency preserves endothelial function and glycocalyx integrity in early streptozotocin-induced diabetes. Diabetes 2016; 65: 2742-53.

[33] Sun K, Wernstedt Asterholm I, Kusminski CM et al. Dichotomous effects of VEGF-A on adipose tissue dysfunction. Proc Natl Acad Sci U S A 2012; 109: 5874-9.

[34] Postic C, Shiota M, Niswender KD et al. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic β cell-specific gene knock-outs using Cre recombinase. J Biol Chem 1999; 274: 305-15.

[35] Belteki G, Haigh J, Kabacs N et al. Conditional and inducible transgene expression in mice through the combinatorial use of Cre-mediated recombination and tetracycline induction. Nucleic Acids Res 2005; 33: e51.

[36] Zhu Y, Pereira RO, O'Neill BT et al. Cardiac PI3K-Akt impairs insulin-stimulated glucose uptake independent of mTORC1 and GLUT4 translocation. Mol Endocrinol 2013; 27: 172-84.

[37] Huang M, Mathew N, Zhu Y. Assessing whole-body lipid-handling capacity in mice. J Vis Exp 2020; (165): 10.3791/61927.

[38] Wangler MF, Chao YH, Bayat V et al. Peroxisomal biogenesis is genetically and biochemically linked to carbohydrate metabolism in Drosophila and mouse. PLoS Genet 2017; 13: e1006825.

[39] Gusdon AM, Fu C, Putluri V et al. Early systemic glycolytic shift after aneurysmal subarachnoid hemorrhage is associated with functional outcomes. Neurocrit Care 2022; 37: 724-34.

[40] Wang X, Spandidos A, Wang H et al. PrimerBank: a PCR primer database for quantitative gene expression analysis, 2012 update. Nucleic Acids Res 2012; 40: D1144-9.

[41] Zhu Y, Pires KM, Whitehead KJ et al. Mechanistic target of rapamycin (Mtor) is essential for murine embryonic heart development and growth. PLoS One 2013; 8: e54221.

[42] Ewels P, Magnusson M, Lundin S et al. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 2016; 32: 3047-8.

[43] Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014; 15: 550.

[44] Liao PC, Bergamini C, Fato R et al. Isolation of mitochondria from cells and tissues. Methods Cell Biol 2020; 155: 3-31.

[45] Chen X, Li H, Liu Y et al. Dimethyl sulfoxide inhibits bile acid synthesis in healthy mice but does not protect mice from bile-acid-induced liver damage. Biology (Basel) 2023; 12: 1105.

[46] Liu HY, Collins QF, Xiong Y et al. Prolonged treatment of primary hepatocytes with oleate induces insulin resistance through p38 mitogen-activated protein kinase. J Biol Chem 2007; 282: 14205-12.

RIGHTS & PERMISSIONS

The Author(s) 2025. Published by Oxford University Press on behalf of Higher Education Press.