The Drosophila fat body is functionally analogous to mammalian white adipose tissue and liver, serving as a buffer to store excess nutrients in fat droplets and glycogen and deploy them to support the animal with fuel during times of fasting (Arrese and Soulages, 2010; Musselman et al., 2013). During larval development, fat body cells halt cell division and dramatically increase in size during the third instar stage (from 72 to 120 hr After Egg Laying (AEL)) (Edgar and Orr-Weaver, 2001; Zheng et al., 2016; Figure 1a–c). As a first step toward understanding the metabolic programs that enable this rapid cell growth, we performed RNA sequencing of the Drosophila fat body across this developmental period. We observed a time-dependent change in mRNAs encoding proteins that have well-characterized roles in supporting cell growth, including components of the insulin and mTORC1 signaling pathways and the Myc transcriptional network, which correlated with increased cell size (Figure 1—figure supplement 1a). Among metabolic genes, we observed modest differences in those that function in amino acid synthesis and fatty acid metabolism (Figure 1—figure supplement 1b). The abundances of mRNAs encoding proteins involved in glycolysis, oxidative phosphorylation (OxPhos), and the TCA cycle were distinctly altered in fat bodies during development (Figure 1—figure supplement 1b). Since pyruvate metabolism is the central node connecting carbohydrate metabolism and the TCA cycle, we studied the abundances of mRNAs that encode proteins that regulate pyruvate metabolism (Figure 1d). We found that the expression of genes that link pyruvate to the TCA cycle was reduced, including Pyk, the Drosophila pyruvate kinase homolog, which converts phosphoenolpyruvate into pyruvate. Mpc1 and Mpc2, which encode the two subunits of the MPC, which transports pyruvate into the mitochondrial matrix; as well as Pdha and Dlat, which encode subunits of PDH complex (Figure 1e). In contrast, some mRNAs encoding proteins that regulate pyruvate abundance were upregulated, such as Pepck and Pepck2, which make phosphoenolpyruvate from oxaloacetate (Figure 1—figure supplement 1c). Based on these data, we hypothesized that the suppression of mitochondrial pyruvate metabolism, which is gated by the action of the MPC, might support the rapid cell growth observed in fat body cells.

To test this hypothesis, we prevented the downregulation of the MPC during larval development (Figure 1e) by ectopically expressing Mpc1 and Mpc2 (termed ‘MPC+’ Figure 1—figure supplement 2a and b). The sustained expression of the MPC (both Mpc1 and Mpc2) throughout the fat body significantly reduced the rate of cell growth compared to a GFP-expressing control (Figure 1—figure supplement 2c). Given the important role of the fat body in controlling organismal growth, we wanted to assess the cell-autonomous effects of MPC expression using mosaic analysis in individual fat body cells (Figure 1—figure supplement 2d; Ito et al., 1997). We generated GFP-labeled clones with expression of the MPC (MPC+), which we confirmed by immunofluorescence (Figure 1—figure supplement 2e, f). The MPC-expressing, GFP-positive cells were significantly smaller in size compared with either mock clones (control) or neighboring GFP-negative cells within the same tissue. Mpc1 knockdown (Mpc1-KD) clones, in contrast, were marginally larger (Figure 1f, g). These results demonstrate that sustained expression of the MPC in developing fat body cells is sufficient to prevent cell growth in a cell-autonomous manner.

If the effects of MPC expression were related to the mitochondrial transport of pyruvate, then limiting the production of pyruvate should mitigate these effects. Therefore, we measured cell size in larvae raised on either a normal (9% sugar) diet or a diet with no added sugars, which limits the production of pyruvate from glucose and fructose. Limiting dietary sugars significantly reduced the size of control clones but increased the size of MPC-expressing clones. Notably, the size of MPC-expressing fat body clones was comparable to that of control clones when larvae were grown in sugar-limited medium, suggesting that limiting pyruvate synthesis abolishes the effect of MPC expression on fat body cell size (Figure 1h, i). Mpc1-KD clones were again larger than control cells, and their size was unaffected by the sugar-limited diet. Inhibiting MPC activity by feeding larvae a normal diet supplemented with the MPC inhibitor UK5099 ameliorated the cell size effects of MPC expression (Figure 1j; Figure 1—figure supplement 2g). These results indicate that pyruvate transport into mitochondria inversely correlates with the size gain in fat body cells. These observations also suggest that the suppression of mitochondrial pyruvate import and metabolism is required for the rapid cell size expansion observed in the fat body during larval development.

The size of a cell is determined by its content of proteins, nucleic acids, and lipids (Björklund, 2019; Lloyd, 2013; Schmoller and Skotheim, 2015). To understand how MPC expression affects the abundance of these macromolecules, we dissected fat bodies from control and fat body-wide MPC-expressing larvae at 120 hr AEL and quantified DNA, RNA, triacylglycerides, and protein. Control and MPC+ fat bodies had equivalent DNA content (Figure 1k; Figure 1—figure supplement 2h) and similar levels of EdU incorporation (Figure 1—figure supplement 2o, p), suggesting that MPC expression does not impact DNA endoreplication. RNA and lipid content were modestly decreased in MPC+ fat bodies compared with control tissues (Figure 1k; Figure 1—figure supplement 2i, j), although the number of lipid droplets was higher with MPC expression (Figure 1—figure supplement 2l–n). In contrast, MPC expression dramatically decreased protein abundance in fat bodies (Figure 1k, Figure 1—figure supplement 2k), and reduced protein synthesis as assessed by staining for the puromycin analog, O-propargyl-puromycin (OPP) (Deliu et al., 2017; Villalobos-Cantor et al., 2023; Figure 1l and m). These data suggest that MPC-mediated mitochondrial pyruvate import decreases protein synthesis, which likely contributes to the reduced size of MPC-expressing cells. Conversely, in developing fly larvae, repression of the MPC and the subsequent decrease in mitochondrial pyruvate appear to provide a metabolic mechanism to support a rapid expansion in cell size.

The best understood mechanisms that govern cell size involve conserved signaling and transcriptional networks (Björklund, 2019; Grewal, 2009; Lang et al., 1998; Lloyd, 2013). For example, the mTORC1 pathway coordinates both extracellular and intracellular growth regulatory signals to dictate the synthesis and degradation of macromolecules, including proteins, lipids, and nucleic acids (Gonzalez and Rallis, 2017; Gonzalez and Rallis, 2017; Saxton and Sabatini, 2017). mTORC1 increases protein synthesis through the phosphorylation of several proteins, including S6 kinase (S6K) and 4EBP1 (Figure 2a). Since MPC-expressing clones were smaller in size and had reduced protein synthesis compared with control clones, we assessed mTORC1 activity in the fat body using our mosaic expression system. As a control, we confirmed that S6k (the Drosophila gene encoding S6K) over-expressing clones were larger in size and had elevated phospho-S6 staining compared with wild-type clones (Figure 2b and c). Surprisingly, MPC+ clones also had elevated phospho-S6 staining (Figure 2b, c), suggesting that despite their small size, these cells have high mTORC1 activity. Overexpression of Rheb, which is an upstream activator of mTORC1, resulted in clones that were larger than wild-type controls and which had increased phospho-4EBP1 staining (Figure 2d, e). Again, even though the MPC+ clones were smaller in size, we observed robust p-4EBP1 staining, indicating that mTORC1 is hyperactive in these cells.

Gene set enrichment analysis of RNA-sequencing data from MPC-expressing fat bodies showed enrichment for signatures associated with pro-growth signaling (Figure 2—figure supplement 1a). We assessed markers of several of these pathways in MPC+ clones. In addition to the mTORC1 pathway, Myc increases cell growth by regulating the transcription of ribosome subunits and biosynthetic metabolic genes. We found that MPC+ clones had an elevated abundance of the Myc protein, both in the cytoplasm and nucleus (Figure 2f, g), suggesting that the growth-promoting Myc transcriptional program is active in these cells. MPC+ clones also had reduced expression of the transcription factor Foxo, consistent with its downregulation by pro-growth signaling pathways (Figure 2—figure supplement 1b). Finally, we stained for phospho-eIF2α to assess the activity of the integrated stress response, which restricts global protein synthesis. Starvation robustly induced the integrated stress response in control clones, but MPC+ clones exhibited no evidence of activation of this pathway under normal growth conditions (Figure 2h, i). Collectively, these data indicate that conventional pro-growth pathways are activated in MPC-expressing clones, which is incongruent with the small size of the clones. This suggests that mitochondrial pyruvate metabolism controls cell size via alternative molecular mechanisms.

We next performed genetic epistasis analysis to better understand the relationship between MPC expression and the mTORC1, PI3K, and Myc pathways. Activation of the PI3K or mTORC1 pathways, via overexpression of PI3K, increased the size of control but not MPC-expressing fat body cells (Figure 2—figure supplement 1c). Similarly, activation of mTORC1 via knockdown of its inhibitor, Tuberous Sclerosis Complex 1 (Tsc1), increased the size of control cells but had no effect on MPC+ cell size (Figure 2—figure supplement 1d). Overexpression of Myc, on the other hand, increased the size of control and MPC+ fat body clones to a similar degree (Figure 2—figure supplement 1e). Knockdown of Myc was sufficient to decrease cell size to a similar extent as MPC expression; however, Myc knockdown had no additional effect on cell size in MPC+ clones (Figure 2—figure supplement 1f). These results suggest that neither mTORC1, PI3K, nor Myc is epistatic to MPC and suggest that MPC likely acts independently of these canonical pathways to regulate the size of fat body cells.

We next performed genetic epistasis analysis to better understand the relationship between MPC expression and the mTORC1, PI3K, and Myc pathways. Activation of the PI3K or mTORC1 pathways, via overexpression of PI3K, increased the size of control but not MPC-expressing fat body cells (Figure 2—figure supplement 1c). Similarly, activation of mTORC1 via knockdown of its inhibitor, Tuberous Sclerosis Complex 1 (Tsc1), increased the size of control cells but had no effect on MPC+ cell size (Figure 2—figure supplement 1d). Overexpression of Myc, on the other hand, increased the size of control and MPC+ fat body clones to a similar degree (Figure 2—figure supplement 1e). Knockdown of Myc was sufficient to decrease cell size to a similar extent as MPC expression; however, Myc knockdown had no additional effect on cell size in MPC+ clones (Figure 2—figure supplement 1f). These results suggest that neither mTORC1, PI3K, nor Myc is epistatic to MPC and suggest that MPC likely acts independently of these canonical pathways to regulate the size of fat body cells.

Since the Drosophila fat body has many features reminiscent of the mammalian liver, we tested whether MPC expression might similarly restrict cell size in cells derived from this tissue. We engineered HepG2 cells, which were originally isolated from a hepatocellular carcinoma, to express epitope-tagged MPC1 and MPC2 in a doxycycline-inducible manner. We observed expression of ectopic MPC1 and MPC2 starting at 4 hr after doxycycline treatment, which increased over the duration of the time course, peaking at 24 hr post-induction (Figure 3a). We stained doxycycline-treated cells with phalloidin and measured cell size using microscopy images taken at 2-hr intervals. Induction of MPC expression coincided with a significant reduction in cell size, which was first apparent 6 hr post-doxycycline treatment and was sustained for the remainder of the time course (Figure 3b). Doxycycline had no effect on the size of control HepG2 cells (EV). Treatment with the MPC inhibitor UK5099 for 24 hr markedly increased the size of control cells and partially reversed the small size phenotype of MPC-expressing cells (Figure 3c). We also assessed cell volume by analyzing a 3D reconstruction of confocal images of phalloidin-stained HepG2 cells and found that the cell volume was lower in MPC+ cells (Figure 3—figure supplement 1a).

We have observed that the physiological consequences of altered pyruvate metabolism are more apparent in cells grown in a three-dimensional culture environment (Schell et al., 2014; Wei et al., 2022). Using our doxycycline-inducible cells, we found that MPC expression resulted in spheroids that were significantly smaller than control spheroids as assessed by microscopy (Figure 3d, e). Cells from MPC-expressing spheroids were also smaller, as shown by flow cytometry (Figure 3f, g). When compared with control (EV) spheroids, expression of MPC did not affect the number of cells per spheroid, the cell cycle phase distribution, or the number of apoptotic or necrotic cells (Figure 3—figure supplement 1b-d). These analyses suggest that the smaller size observed in MPC-expressing spheroids is not due to effects on cell proliferation or cell death.

Consistent with our observations in the Drosophila fat body, protein content was lower in MPC-expressing HepG2 spheroids compared with EV controls, whereas there was no difference in the abundances of DNA, RNA, or triacylglycerides (Figure 3h; Figure 3—figure supplement 1e-h). To directly assess the effect of MPC expression on protein synthesis, we induced the MPC in HepG2 cells and treated the cells acutely with a low concentration of puromycin to label nascent proteins. We found that MPC expression decreased the incorporation of the amino acid analog L-homopropargylglycine (HPG) (Shen et al., 2021; Tom Dieck et al., 2012) into nascent proteins, as assessed by fluorescence microscopy (Figure 3i, j). Protein synthesis was similarly reduced in MPC-expressing HepG2 spheroids (Figure 3k, l). These observations were further supported by the decreased abundance of the short-lived, destabilized GFP (d2GFP) (Li et al., 1998; Pavlova et al., 2020) in MPC-expressing HepG2 spheroids compared with controls (Figure 3m). This reduction in the level of d2GFP was prevented by treatment with the MPC inhibitor UK5099 (Figure 3m). Together, these data suggest that increased transport of pyruvate into mitochondria, mediated by MPC expression, reduces protein synthesis and cell size in both fly and mammalian models.

We have previously shown that loss of the MPC reduces the contribution of glucose and pyruvate to the TCA cycle (Bensard et al., 2020; Cluntun et al., 2021). To investigate how MPC expression impacts carbohydrate metabolism in HepG2 cells, we performed metabolic tracing using ¹³C-glucose (Figure 4a). We observed that MPC expression reduced the labeling fraction of the glycolytic intermediates 3-phosphoglycerate (Figure 4—figure supplement 1a) and pyruvate (Figure 4—figure supplement 1b), as well as alanine (Figure 4—figure supplement 1c), which is derived from pyruvate, all of which suggests a reduction in glycolysis in these cells. Once imported into mitochondria, glycolytic-derived pyruvate has two major fates: conversion to acetyl-CoA by PDH or to oxaloacetate by pyruvate carboxylase (PC). These fates can be differentiated by assessing the abundances of M+2 (derived from PDH) and M+3 (derived from PC) TCA cycle intermediates (Figure 4a). We found that both M+2 and M+3 isotopomers of TCA cycle intermediates were modestly increased following MPC expression (Figure 4b and c; Figure 4—figure supplement 1d–i). The fact that we observe increases in TCA cycle intermediates despite decreased glycolytic labeling in pyruvate suggests that the apparent labeling through PDH and PC might be underestimating the effect on metabolic flux through these enzymes. Thus, it appears that MPC expression increases the activity of both enzymes that utilize pyruvate. Typically, when TCA cycle flux is increased, one observes an increase in activity of the electron transport chain (ETC) to oxidize the resulting NADH. However, MPC expression had no impact on ETC activity as assessed by measuring oxygen consumption (Figure 4d). The implications of elevated NADH production without a concomitant increase in NADH oxidation will be discussed in Figure 5.

To probe the impact of increased flux through the PC and PDH reactions, we conducted genetic epistatic analysis in fat body cells. We found that Drosophila fat body clones in which we over-expressed Pcb (the Drosophila gene encoding pyruvate carboxylase) were significantly smaller than controls and were equivalent in size to MPC+ clones (Figure 4e, f; Figure 4—figure supplement 2a–c). Clones expressing both the MPC and Pcb were even smaller (Figure 4e, f; Figure 4—figure supplement 2a–c). Conversely, knockdown of Pcb (Pcb-KD) in MPC-expressing fat body clones completely rescued the cell size phenotype (Figure 4e, f; Figure 4—figure supplement 2d). Knockdown of Pcb in MPC-expressing HepG2 cells also eliminated the small cell size phenotype (Figure 4—figure supplement 2e). These data suggested that the small size of cells expressing the MPC is likely due to increased flux through the PC reaction. Consistent with this, knockdown of the PDH-E1 (Drosophila gene Pdha) or PDH-E2 (Drosophila gene Dlat) subunits of PDH, which should divert mitochondrial pyruvate, so it is preferentially used by PC, also resulted in smaller fat body cells (Figure 4—figure supplement 2f–k). In contrast, activating PDH by knocking down the inhibitory PDH kinases (mammalian PDKs, Drosophila gene Pdk) (Stacpoole, 2017; Wang et al., 2021), which should promote the flux of pyruvate through PDH and away from PC, rescued cell size in MPC-expressing fat body clones (Figure 4—figure supplement 2j, k). Similarly, in MPC-expressing HepG2 cells, activation of PDH via treatment with the PDK inhibitors DCA and AZD7545 was sufficient to restore cell size (Figure 4—figure supplement 2l), (Stacpoole, 2017; Wang et al., 2021). These analyses suggest that the reduction in cell size observed with MPC expression is due to PC-mediated metabolism of pyruvate.

The product of PC, oxaloacetate, has three major metabolic fates: (1) feeding the TCA cycle via citrate synthase as discussed above, (2) conversion to aspartate by glutamic-oxaloacetic transaminase 2 (GOT2), and (3) conversion to phosphoenolpyruvate by PEPCK2, leading to gluconeogenesis, synthesis of glucose (Jitrapakdee et al., 2008; Kiesel et al., 2021). Sustained expression of the MPC in fat bodies increased the concentration of glucose in this tissue (Figure 4g) as well as glucose and trehalose in larval circulation (known as hemolymph) (Figure 4h), suggesting that MPC expression increases glucose production in fat body cells. Next, we tested whether PEPCK2-mediated gluconeogenesis was elevated in these cells compared with controls. We used ¹³C-lactate to trace the ¹³C-labeling of phosphoenolpyruvate and resultant glucose synthesis via gluconeogenesis (Figure 4i). We found that the relative abundance of M+3 phosphoenolpyruvate was higher in MPC-expressing cells compared with controls (Figure 4j), suggesting increased activity of PEPCK2. The gluconeogenic pathway employs a series of biochemical reactions to convert phosphoenolpyruvate into glucose, which is then excreted from the cell. This production of M+6 glucose from lactate was also higher in MPC-expressing HepG2 cells (Figure 4k). To test whether gluconeogenesis contributes to the small size phenotype of MPC-expressing cells, we knocked down enzymes in the pathway (Drosophila phosphoenol carboxykinase, Pepck2, and fructose bisphosphatase, Fbp) and assessed cell size in Drosophila and HepG2 models. Knockout of PEPCK2 rescued the size of MPC-expressing HepG2 cells grown in spheroids (Figure 4l) and cells (Figure 4—figure supplement 2m). Knockdown of either Pepck2 or Fbp partially rescued the size defect in MPC-expressing Drosophila fat body clones (Figure 4n, o; Figure 4—figure supplement 2n–r), and Pepck2 knockdown also increased protein synthesis in these clones (Figure 4p, q). Collectively, these data suggest that the cell size and protein synthesis phenotypes observed in MPC-expressing cells require PC-mediated gluconeogenesis, and this relationship is found in both the Drosophila fat body and in HepG2 cells.

We were intrigued by the observation that MPC expression exerted its impact on cell size through PC, rather than via increased flux through PDH. We did not observe any differences in the abundances of PDH1a, phosphorylated (Ser-293, inactive) PDH1a, PC, PEPCK2, or G6PC proteins in MPC-expressing HepG2 cells (Figure 5—figure supplement 1a, b). We therefore hypothesized that the shift in pyruvate metabolism in response to MPC expression might be driven by changes in the abundances of enzyme cofactors, metabolite regulators, or cellular redox balance. PC and PDH show reciprocal regulation by such factors: PC utilizes ATP as a substrate and is allosterically activated by acetyl-CoA, whereas PDH is inhibited by ATP, acetyl CoA, and NADH (Chai et al., 2022; Sugden and Holness, 2011; Figure 5—figure supplement 1c). In addition, several reactions in gluconeogenesis require ATP or NADH (Hausler et al., 2006; Siess et al., 1977; Figure 5—figure supplement 1c). Upon induction of MPC expression in HepG2 cells, we observed increased abundance of acetyl-CoA (Figure 5—figure supplement 1d), a higher ATP to ADP ratio (Figure 5—figure supplement 1e), a greater abundance of NADH without an increase in NAD⁺, and thus an increase in the cellular NADH/NAD⁺ ratio (Figure 5—figure supplement 1f–h). To determine how MPC expression impacts the subcellular distribution of redox factors, we separated cytoplasmic and mitochondrial fractions from control or MPC-expressing Drosophila fat bodies and measured the NADH/NAD⁺ ratio. We found that MPC expression increased the NADH/NAD⁺ ratio in both the cytoplasm and mitochondria (Figure 5a). We observed a similar increase in the NADH/NAD⁺ ratio in both fractions in MPC-expressing HepG2 spheroids (Figure 5b). It is important to note that these measurements can be confounded by the very rapid redox reactions that can happen post-harvest. Nonetheless, these combined results suggest that the increased abundances of acetyl-CoA, ATP, and NADH in MPC-expressing cells could promote the rewiring of mitochondrial pyruvate metabolism through PC and support gluconeogenesis.

To test the hypothesis that ATP or NADH concentrations might affect cell size in MPC-expressing Drosophila fat body clones or HepG2 cells, we utilized pharmacological or genetic modulation of these molecules. Treatment with gramicidin, which decreases the ATP to ADP ratio (Xue et al., 2022), did not alter the size of MPC-expressing HepG2 cells (Figure 5—figure supplement 1i). We used several orthogonal approaches to reduce the cellular NADH/NAD⁺ ratio in the MPC-expressing systems and measured their effect on cell size. Co-expression of the Drosophila nicotinamide mononucleotide adenylyl transferase (Nmnat), which increases NAD⁺ biosynthesis, almost normalized the small size phenotype of MPC-expressing clones in the Drosophila fat body (Figure 5c, d). We observed a similar rescue of cell size following expression of the Ciona intestinalis alternate Complex I enzyme NADH dehydrogenase (NDX) (Gospodaryov et al., 2020) or the yeast NADH dehydrogenase (NDI) (Sanz et al., 2010), both of which oxidize NADH to NAD⁺ without concomitant proton translocation and energy capture (Figure 5e, f). Expression of NDI in HepG2 cells also mitigated the effect of MPC expression on spheroid size (Figure 5—figure supplement 1j, k), as did treatment with duroquinone (Merker et al., 2006; Figure 5—figure supplement 1l), which oxidizes NADH to NAD⁺. To extend these investigations to three-dimensional culture, we supplemented the growth medium of MPC-expressing HepG2 spheroids with the NAD⁺ precursors nicotinamide riboside (NR) or nicotinamide mononucleotide (NMN), both of which recovered spheroid size (Figure 5g, h).

Since MPC expression reduced protein synthesis in both Drosophila fat bodies and HepG2 cells, we tested how cellular redox status might contribute to this phenotype in both systems. Expression of NDX, which lowers the NADH/NAD⁺ ratio, increased translation in MPC-expressing Drosophila fat body clones (Figure 5i, j). In HepG2 spheroids, boosting NAD⁺ biosynthesis by supplementing growth media with NR or NMN partially rescued the abundance of destabilized GFP (Figure 5k). These results suggest that the elevated NADH/NAD⁺ ratio in MPC-expressing cells limits protein synthesis and that normalizing that ratio increases protein synthesis and cell size.

Given the reduced protein synthesis observed upon MPC expression, we assessed amino acid concentrations in HepG2 using steady-state metabolomics. We found that MPC expression reduced the abundance of most amino acids (Figure 6a). To test whether the low abundances of amino acids contribute to the smaller size of MPC-expressing cells, we supplemented the growth media of HepG2 spheroids with excess non-essential amino acids (NEAAs)—either two or three times the recommended dilution of a commercially available amino acid cocktail that includes glycine, L-alanine, L-asparagine, L-aspartate, L-glutamate, L-proline, and L-serine. MPC-expressing spheroids grown with excess NEAAs were comparable in size to controls (Figure 6b, c). In parallel, we provided Drosophila larvae with food containing excess (5x) amino acids from 72 to 120 hr AEL, which partially rescued cell size in MPC-expressing fat body clones (Figure 6d, e). To genetically augment intracellular amino acids, we expressed the amino acid importer slimfast (Colombani et al., 2003) in control or MPC-expressing fat body clones and found that it prevented the decrease in cell size (Figure 6f, g).

To determine which amino acid(s) contribute to the cell size effects of MPC expression, we cultured control and MPC-expressing HepG2 cells in media supplemented with an excess of each amino acid from the NEAA cocktail. Treatment with excess glycine, l-alanine, or L-serine had no effect on cell size (Figure 6h). However, the size of MPC-expressing cells was normalized by supplementation with L-aspartate, L-glutamate, or L-proline (Figure 6h), all of which are derived from TCA cycle intermediates. Supplementation with either L-aspartate or L-glutamate also rescued the small size phenotype of MPC-expressing HepG2 spheroids (Figure 6—figure supplement 1a, b). Increasing L-aspartate uptake by over-expressing the aspartate transporter SLC1A3 also recovered size of HepG2 spheroids (Figure 6—figure supplement 1c, d). In addition, treatment with excess (3x) NEAA partially restored the abundance of d2GFP in MPC-expressing HepG2 spheroids (Figure 6—figure supplement 1e), suggesting rescued protein synthesis.

Like glutamate and proline, aspartate is derived from a TCA cycle intermediate, specifically via transamination of oxaloacetate by glutamic-oxaloacetate transaminase 2 (GOT2). Since GOT2 and PEPCK2 both utilize oxaloacetate as a substrate, we hypothesized that knocking down GOT2 might phenocopy MPC expression by driving PEPCK2-mediated conversion of oxaloacetate into phosphoenolpyruvate and thereby suppressing aspartate biosynthesis. Knockdown of Got2 (the Drosophila gene encoding GOT2) reduced cell size in Drosophila fat body clones and phenocopied MPC-expressing cells (Figure 6i, j; Figure 6—figure supplement 1f–h). Similarly, both GOT2 knock-out and MPC-expressing HepG2 cells were smaller than EV (Figure 6—figure supplement 1i). Expression of the MPC in these GOT2 knockdown systems had no significant impact on cell size (Figure 6i, j; Figure 6—figure supplement 1i), suggesting that the effects of MPC expression and GOT2 knockdown act similarly to limit amino acid synthesis and cell size. We next performed the reciprocal experiment by over-expressing Got2 to favor the production of aspartate from oxaloacetate. Got2 expression normalized cell size in both MPC-expressing fat body clones and in HepG2 spheroids (Figure 6i, j; Figure 6—figure supplement 1j, k). The efflux of aspartate from mitochondria into the cytoplasm is a critical component of the malate-aspartate shuttle, which is a major redox shuttle in human cells. To test whether increasing the abundance of aspartate would ameliorate the high NADH/NAD⁺ ratio observed in MPC-expressing cells, we supplemented growth media with exogenous aspartate and assessed cellular redox status. We found that excess aspartate reduced the cellular NADH/NAD⁺ ratio in these cells such that it was comparable with control cells (Figure 6k). We observed similar results when we treated these cells with NMN or UK5099 (Figure 6k). Moreover, all these treatments increased protein synthesis in MPC-expressing cells (Figure 6l, m).

Although HepG2 cells exhibit some features of hepatocytes, they are a transformed, immortalized, and proliferative hepatocellular carcinoma cell line. We wanted to test whether the MPC expression phenotypes that we observed in Drosophila fat bodies and HepG2 cells could be recapitulated in a more physiologically relevant mammalian model. We chose primary rat hepatocytes, which have been used extensively to interrogate hepatocyte cell metabolism and signaling and also have the advantage of being genetically tractable. We expressed MPC1 and MPC2 in cultured primary rat hepatocytes (Figure 7a), and, consistent with our results in other systems, we found that expression of MPC reduced cell size (Figure 7b, c) and decreased protein synthesis (Figure 7d). MPC-expressing primary hepatocytes also had a higher NADH to NAD⁺ ratio in both the cytoplasmic and mitochondrial fractions (Figure 7e). We assessed gluconeogenesis in primary hepatocytes by quantifying glucose in the culture media following incubation with the gluconeogenic precursors, pyruvate, and lactate. Glucose production was higher in MPC-expressing hepatocytes compared with controls (Figure 7f). Treatment with UK5099 eliminated this effect and reduced glucose production to a similar rate in both cells (Figure 7f). These results demonstrate that augmented mitochondrial pyruvate in hepatocytes and related cells in Drosophila drives a metabolic program that results in an increased NADH/NAD⁺ ratio. This scenario results in accelerated gluconeogenesis, decreased protein synthesis, and reduced cell size.

Drosophila melanogaster stocks were maintained at 25°C on semi-defined fly food composed of 20 g agar, 80 g baker’s yeast, 30 g sucrose, 60 g glucose, 0.5 g MgSO₄, 0.5 g CaCl₂, 11 ml tegosept, and 6 ml propionic acid. This was the base medium for all Drosophila experiments, but specific fly food modifications are mentioned in text and figure legends. To induce clones in fat bodies, synchronized eggs were transferred to 29°C for 4 days until dissection. For experiments involving genetic manipulation of all fat body cells, tubGal80^ts20 was used to restrict activity of CG-Gal4 at 18°C for 120 hr (57 hr equivalent time at 25°C) and larvae were shifted to 29°C till dissection at specified time points.

Following fly stocks were procured from Bloomington stock center UAS-MPC1-P2A-MPC2 (28812582), CG-Gal4; tubGal80^ts20, hs-Flp1.22 (BDSC, #77928), Act >CD2>Gal4, UAS-GFP (BDSC, #4413), Act >CD2>Gal4, UAS-RFP (BDSC, #30558), UAS-S6K^STDETE (BDSC, #6914), UAS-Rheb^PA (BDSC, #9689), UAS-Myc^OE (BDSC, #9675), UAS-Myc^RNAi (BDSC, #25783), UAS-PI3K93E.^Excel (BDSC, #8287), UAS-PI3K93E^A2860C (BDSC, #8288), UAS-TSC1^RNAi (BDSC, #31314), UAS-hPC (BDSC, #77928), UAS-PC^RNAi (BDSC, #56883), UAS-PEPCK2^RNAi (BDSC, #36915), UAS-FBPase^RNAi (BDSC, #51871), UAS-Pdk^RNAi (BDSC, #35142), UAS-NMNAT (BDSC, #37002), UAS-CintNDX (BDSC, #93883), UAS-ScerNDI1 (BDSC, #93878), UAS-Slif (BDSC, #52661), UAS-GOT2^RNAi (BDSC, #78778). From VDRC UAS-MPC1^RNAi(KK) and from NIG UAS-Pdha^RNAi, UAS-Dlat^RNAi (PDH E2) were purchased.

Pc and Pepck2 deletion fly stocks were generated using CRISPR–Cas9 as described (Gratz et al., 2014). We targeted specific nucleotide sequences of the genes of interest through homology-directed recombination using two guide RNAs and inserted a dsRed construct that expresses in adult eyes to facilitate the selection of mutant flies. We deleted exon 1 to exon 5 using guide RNAs for Pc^-KO: 5′ guide ATACATTTAAGTCCTAGGC; 3′ guide TCGATTGATCCTGGAAACA. Pepck2^-KO, being a single exon coding sequence, we generated complete deletion by using guides 5′ guide AAAGGGTGCACATCTGTGA; 3′ guide TTTGGGGCGTGGCCTAGAC. The plasmids were injected into y[1] M{GFP[E.3xP3]=vas-Cas9.RFP}ZH-2A w[1118] (BDSC, #55821) fly stock embryos (Bestgene) and one Pc-KO and five Pepck2-KO flies were picked up, confirmed for the dsRed expression in adult eyes and used in subsequent experiments.

To overexpress Got2, Got2 cDNA was amplified from RNA extracted from larval fat bodies using primer 5′ GAATTC ATGAGTAGAACCATTATTATGACGCTTAAGGAC, 3′ CTCGAG CTTGGTAACCTTGTGTATGCTCTCAGCCAGG. The cDNA was then cloned into a pUAST-aatB plasmid using EcoRI and XhoI restriction enzymes, and the construct was injected into pBac{yellow[+]-attp-9A}VK00005 (BDSC, #9725) embryos to obtain the insertion.

Fat body clones were induced by the leaky expression of heat shock flippase 1.22 during embryonic stages, and the 2D size of fat body cells was analyzed at 120 hr after egg laying (AEL) using fluorescence microscopy. Cell size analysis was conducted as reported earlier (Toshniwal et al., 2019). Fat bodies were dissected from larvae at the specified time points in 1X PBS buffer (pH 7.2, Invitrogen, #10010049) and fixed in 8% paraformaldehyde (Sigma-Aldrich, # P6148) for 1 hr at room temperature (RT). The tissues were then washed twice with 0.1% PBT (0.1% Triton X-100 (Sigma-Aldrich, # X100) in 1X PBS buffer) for 10 min each. Subsequently, the tissues were incubated with Rhodamine Phalloidin (Thermo Scientific, # R418) at a 1:400 dilution in 1X PBS buffer for 2 hr at RT. After staining, the tissues were washed once with 0.1% PBT and then with 1X PBS before being mounted in DAPI-supplemented VectaShield (Vector Labs, #H1200). Representative images were captured using a Laser Scanning Confocal Microscope (LSM 880, Carl Zeiss). For cell size analysis, images were captured with a fluorescence microscope (Carl Zeiss, Axio Vision) at 20× magnification, focusing on a plane where all nuclei were in focus. The 2D areas of fat body cells were measured using FIJI software. Cell membranes stained with Rhodamine Phalloidin were traced using the freehand tool, and the area of approximately 20–25 GFP-positive cells from fat bodies collected from about five animals was measured as one replicate. All cell size analyses were conducted in a blinded manner.

Larval fat bodies at 120 hr after egg laying (AEL) at 29°C were dissected in 1X PBS buffer (pH 7.2), fixed in 8% paraformaldehyde for 1 hr at RT, and then washed three times in 0.1% PBT (0.1% Triton X-100 in 1X PBS buffer) for 10 min each. The tissues were blocked in 10% normal goat serum (NGS, Jackson ImmunoResearch Laboratories, #005-000-121) in 0.1% PBT for 1 hr at RT, followed by incubation with primary antibodies overnight at 4°C. Secondary antibody incubation was performed for 2 hr at RT. Following three to four washes in 0.3% PBT (10 min each) at RT, the tissues were mounted in DAPI-supplemented Vectashield (Vector Laboratories, #H1200). Images were captured using a Laser Scanning Confocal Microscope (LSM 880, Carl Zeiss).

For all experiments, confocal images were captured using the same laser power and identical settings, with z-stacks of the dissected tissue taken at 1 μm intervals along the y-axis. The images were analyzed using Fiji software, where a similar number of z-stacks focusing on the nuclei were projected at mean intensities. Using the freehand tool, cell membranes were outlined, and the mean fluorescence intensity (mean gray value) for each cell was recorded. To account for background fluorescence, the average mean gray values of the background (measured from regions of interest [ROIs] of the same size) were subtracted from the recorded mean gray values of the cells. The resulting mean gray values, adjusted for background fluorescence, were then normalized, and the percent normalized mean gray values (with s.d.) of GFP-negative and GFP-positive cells were plotted.

Protein synthesis in Drosophila fat bodies was analyzed using the Click-iT Plus OPP Alexa Fluor 594 kit (Molecular Probes, #C10457). Fat bodies were dissected at 120 hr AEL in Shields and Sang M3 Insect Media (SSM3, Sigma, #S8398) and incubated with 5 mM OPP in SSM3 media for 30 min at RT. Tissues were then washed three times with 1X PBS for 10 min each and fixed in 8% paraformaldehyde for 1 hr at RT. After fixation, the tissues were washed twice with 0.1% PBT (0.1% Triton X-100 in 1X PBS) supplemented with 0.5% bovine serum albumin (BSA) for 10 min each. The OPP was developed for 30 min at RT using the Click-iT reaction mixture, which included 88 μl of OPP Reaction Buffer, 20 μl of Copper (III) Sulfate (component D), 2.5 μl of component B, and 100 μl of Buffer Additive (component E), following the manufacturer’s instructions. The tissues were then washed twice in Reaction Rinse Buffer for 10 min each, followed by two washes in 0.1% PBT supplemented with 0.5% BSA for 10 min each at RT. After a final wash in 1X PBS, the fat bodies were mounted in DAPI-supplemented Vectashield. Images were captured using a confocal microscope (LSM 880, Zeiss).

At 120 hr AEL, 3rd instar larvae were dissected in 1X PBS. The fat bodies were fixed in 8% paraformaldehyde for 1 hr at room temperature. After fixation, tissues were rinsed twice with 1X PBS. They were then incubated for 30 min in a 1:100 dilution of LipidTOX Red (Invitrogen, #H34351) in PBS, followed by two additional rinses with PBS. The tissues were mounted in DAPI-supplemented Vectashield and imaged using a confocal microscope. To quantify lipid droplet size, the diameters of lipid droplets from 35 fat body clones were measured using FIJI software.

At 120 hr AEL, fat bodies were dissected in 1X PBS and incubated with 5 μM 5-ethyl-2′-deoxyuridine (EdU) in 1X PBS for 30 min at room temperature. The tissues were then washed three times with 1X PBS for 10 min each, followed by fixation in 8% paraformaldehyde for 1 hr at room temperature. After fixation, the tissues were washed with 0.1% PBT supplemented with 0.5% bovine serum albumin. EdU detection was performed using Click-iT Plus EdU Alexa Fluor 594 (Molecular Probes, #C10639) for 30 min at room temperature, according to the manufacturer’s instructions. Following development, the fat bodies were mounted in DAPI-supplemented Vectashield. Images were captured using a confocal microscope (LSM 880, Zeiss).

To collect fat body tissues, w¹¹¹⁸ flies were mated at 25°C. The following day, flies were starved for 1 hr, and eggs were collected every 2 hr. The first collection was discarded, and the subsequent two collections were incubated at 25°C. After 20 hr, early hatched 1st instar larvae were discarded, and 1st instar larvae collected over a 2-hr period were kept at 25°C. Fat bodies from 10 male larvae at the specified time points were dissected and preserved in 1X PBS. For RNA isolation from MPC-expressing fat bodies at 120 hr AEL, RNA was extracted and purified using the NucleoSpin RNA kit (Takara Bio USA, Inc, #740955.50) with on-column DNA digestion, as per the manufacturer’s instructions. Four independent samples for each time point were prepared for sequencing.

Library preparation for poly(A)-selected RNAs was carried out using the Illumina RNA TruSeq Stranded mRNA Library Prep Kit with oligo-dT selection. Sequencing was performed using the Illumina NovaSeq Reagent Kit v1.5 (150 bp paired-end reads) at the High-Throughput Genomics Core Facility at the University of Utah. The raw sequencing data were analyzed using the BDGP6.28 genome and gene feature files. Differentially expressed genes were identified using DESeq2 version 1.30.0 with a 5% false discovery rate. RNA quality control, library preparation, and sequencing were performed by the University of Utah Huntsman Cancer Institute High Throughput Genomics and Bioinformatics Shared Resource.

500 ng RNA was used to make cDNA using Superscript II reverse transcriptase (Molecular Probes, #18064-022), dNTP (Molecular Probes, #18427-088) and oligodT (Molecular Probes, #18418012). QPCR analyses were performed on cDNA as described using PowerUp SYBR Green Master Mix (Applied Biosystems, #2828831) on QuantStudio 7 Flex (Applied Biosystems) instrument. Fold changes in transcript level were determined using the ΔΔCt method. Transcript levels were normalized to rp49. Each experiment was performed using four to five independent samples. The following primers were used to do qPCR. A list of primers is provided in Appendix 3.

Glucose concentrations in larval hemolymph and fat bodies were measured as previously described (Ugrankar Banerjee et al., 2023). To isolate hemolymph, 10 third instar larvae were selected from culture tubes, thoroughly washed to remove any food residues, and then dried. Hemolymph was collected by bleeding the larvae on a parafilm strip using Dumont 5 forceps (Fine Science Tools, #11254-20). Two microliters of the colorless hemolymph were transferred to a 96-well plate and mixed with 200 μl of Autokit Glucose reagent (Wako, #997-03001). For measuring intercellular glucose, fat bodies were dissected from 10 larvae per genotype and homogenized in approximately 300 µm of ice-cold 1X PBS using a 29G1/2 syringe. The lysates were inactivated at 70°C for 10 min, then centrifuged at maximum speed at 4°C. Thirty microliters of the lysate were mixed with 170 μl of Autokit Glucose reagent. The plates were incubated at 37°C for 30 min, and absorbance was measured at 600 nm. Glucose concentrations were determined based on the absorbance values recorded for glucose standards. For trehalose measurement from hemolymph, 2 μl of hemolymph was diluted five times, and 8 μl of this diluted hemolymph was incubated with 5 μl 8x porcine kidney trehalase (Sigma) at 37°C overnight. 10 μl of this mix was used for glucose measurement, as described above.

HepG2 cells were purchased from ATCC and maintained in EMEM supplemented with 10% FBS and 1% PenStrep at 37°C in a 5% CO₂ atmosphere. HepG2 verification was provided by ATCC and authenticated by STR profiling. Mycoplasma testing was performed every month, and all cell lines used were mycoplasma-free.

For inducible overexpression of human MPC1 and MPC2, the HA-MPC2-P2A-T2A-MPC1-FLAG sequence was cloned into the pLVX-TetOne-Zeocin vector. Lentiviral particles were generated using Gag-Pol, pMD2.G, and VSVG packaging plasmids.

Viral particles were produced by co-transfecting 293T cells with the respective packaging plasmids using polyethylenimine (PEI, Sigma, #765090) as the transfection reagent at a 3:1 mass ratio of PEI to DNA. The virus-containing medium was collected 48 hr post-transfection, filtered through a 0.45-μm filter, and added to HepG2 cells cultured in normal medium, along with polybrene (Sigma, #P1240000) at a concentration of 10 μg/ml. Transduced cells were selected with 10 μg/ml Zeocin (Gibco, #R25001) for 1 week, and the level of overexpression was assessed by western blotting and presented with respective images.

For PC over-expression, PC coding sequence was amplified from mRNA isolated from HepG2 cells using primers 5′ GAATTC ATGCTGAAGTTCCGAACAGTCCATGGG, 3′ GGATCC CTCGATCTCCAGGATGAGGTCGTCACC and cloned into pLenti-CMV-Blast. Similarly, SLC1A3 cDNA was amplified with primers 5′ GGATCC ATGACTAAAAGCAATGGAGAAGAGC, 3′ CTACATCTTGGTTTCAATGTCGATGG and GOT2 coding sequence was amplified using primers 5′ GGATCC ATGGCCCTGCTGCACTCCGG, 3′ TCTAGA CTTGGTGACCTGGTGAATGGCATGG and cloned in pLenti-CMV-Blast (Addgene, #17486). The coding sequences of NDI (Addgene, #72876) and d2GFP (Addgene, #115665) were also cloned into pLenti-CMV-Blast vector. Cells were selected on 3 μg/ml blasticidin (Gibco, A1113903) for 3 days. To generate knockout cells, the following gRNAs were used. PCg5e-GAAGCCTATCTCATCGGCCG CGG, PCg6e-CGAAGTCCGCTCGCTCAGAG AGG, PEPCK2g2e-ATCTCCACTAAGCACTCGCA GGG, PEPCK2g3e-CATGCGTATTATGACCCGAC TGG, GOT2ga-GAGTGGCCGGGTAAGCTGAGCAG AGG, GOT2gb-GGAGTGGACCCGCTCCGGAACAG TGG. The guides were annealed and clones in lentiCRISPRv2 with blasticidin (Addgene, #83480) resistance using BsmBI.

HepG2 cells were cultured on 12 mm coverslips in a 24-well plate at a low density of 10,000 cells per well in Human Plasma-Like Medium (Gibco, #A4899101) supplemented with 10% FBS (Sigma, F0926) and 1% PenStrep (Thermo, #15140). The next day, treatments were initiated as described in the figure legends, including 1 μg/ml doxycycline (Sigma, #D5207), 10 μM UK5099 (Sigma, PZ0160-5MG), 10 mM glycine (Sigma, #G7403), 5 mM alanine (Sigma, #A7469), 5 mM asparagine (Sigma, #A4159), 5 mM aspartic acid (Sigma, #A7219), 5 mM glutamic acid (Sigma, #G8415), 5 mM proline (Sigma, #P5607), 5 mM serine (Sigma, #S4311), 10 μM AZD7545 (MedChemExpress, #HY-16082), 1 mM dichloroacetate (Sigma, #347795), 100 nM duroquinone (Sigma, #D223204), and 2 nM gramicidin (Sigma, #G5002).

For time-course experiments in Figure 3, cells were fixed at 2, 4, 6, 8, 10, 12, 18, and 24 hr after doxycycline treatment using 4% paraformaldehyde in 1X PBS for 20 min at RT. For all other experiments, cells were fixed 24 hr after doxycycline treatment. Following fixation, cells were washed once with 0.1% PBT (0.1% Triton X-100 in 1X PBS) for 10 min and incubated with Rhodamine Phalloidin at a 1:400 dilution in 1X PBS buffer for 20 min at RT. After a couple of washes in 1X PBS, cells were mounted in DAPI-supplemented VectaShield.

Images were captured using a fluorescence microscope (Carl Zeiss, Axio Vision) focusing on the plane of the cellular nuclei at 20X magnification, where all nuclei were in focus, and the 2D area of HepG2 cells was measured using FIJI software. All cell size analyses were conducted in a blinded manner.

For 3D cell volume analysis, images were captured using a Laser Scanning Confocal Microscope (LSM 880, Carl Zeiss). The red fluorescence signal was used for 3D reconstruction, and cell volume was measured using an ImageJ Macro code in FIJI.

HepG2 cells were cultured in either 2D monolayers or spheroid forms. At the desired time point, cells were treated with 20 μg/ml puromycin (Sigma, #P4512) in Human Plasma-Like Medium (HPLM) for 30 min at 37°C. Following treatment, cells were washed with 1X PBS, and proteins were extracted using 1X RIPA buffer at 4°C. Protein concentrations were quantified, and 15 μg of total protein was separated by SDS-PAGE using standard methods. Proteins were then transferred onto a nitrocellulose membrane, and puromycin-labeled peptides were detected using an anti-puromycin [3RH11] antibody (Kerafast, #EQ0001) followed by incubation with an appropriate secondary antibody.

Protein synthesis in HepG2 cells was also analyzed using the Click-iT HPG Alexa Fluor 594 kit (Molecular Probes, #C10429). HepG2 cells were grown on 12 mm coverslips at a density of 10,000 cells per well and treated with 1 μg/ml doxycycline for 24 hr. Cells were then incubated with 50 μM L-homopropargylglycine (HPG) in methionine-free DMEM (Gibco, #21013024) for 30 min at 37°C. After incubation, cells were washed twice with 1X PBS for 2 min each and fixed in 4% paraformaldehyde for 20 min at room temperature.

For OPP staining, appropriately fixed cells were washed twice with 0.1% PBT (0.1% Triton X-100 in 1X PBS) supplemented with 0.5% bovine serum albumin (BSA) for 10 min each. HPG was detected using the Click-iT reaction mixture, which included 88 μl of OPP Reaction Buffer, 20 μl of copper solution (component D), 2.5 μl of component B, and 100 μl of Buffer Additive (component E), following the manufacturer’s instructions. Cells were washed twice in Reaction Rinse Buffer for 10 min each, followed by two washes in 0.1% PBT supplemented with 0.5% BSA for 10 min each at room temperature. After a final wash in 1X PBS, the coverslips were mounted in DAPI-supplemented Vectashield. Images of the cells were captured using a confocal microscope (LSM 880, Zeiss).

HepG2 cells were cultured in ultra-low attachment 96-well plates (Costar, #7007) at a density of 10,000 cells per well in Human Plasma-Like Medium (HPLM) supplemented with 10% FBS and 1% penicillin–streptomycin. Treatments were applied as specified in the figure legends, including 1 μg/ml doxycycline, 5 mM aspartate, 5 mM glutamate, 10 μM UK5099, MEM non-essential amino acids (Gibco, #11140050), 100 nM nicotinamide riboside (Sigma, #SMB00907), and 10 μM NMN (Sigma, #N3501). Cells were incubated for 6 days at 37°C with CO₂ and O₂. Brightfield images of the spheroids were captured using a Zeiss Axio Observer Z1 microscope, and spheroid size was measured using FIJI software.

To quantify cell numbers, 12 spheroids from each condition (EV or MPC+) were pooled, dissociated by trypsinization, and the number of cells was counted using a CellQuant system (Bio-Rad).

For cell cycle analysis, cells were stained using Vybrant DyeCycle Violet stain (Molecular Probes, #V35003). After pooling and dissociating 12 spheroids from each condition (EV or MPC+), cells were centrifuged and resuspended in 200 μl of 5 μM Vybrant DyeCycle Violet stain in EMEM supplemented with 10% FBS. The staining was performed by incubating the cells at 37°C for 30 min, protected from light. Samples were analyzed using a BD Celesta flow cytometer with ~405 nm excitation and ~440 nm emission. The resulting FCS files were processed in FlowJo software, where the forward scatter of singlets was recorded, and median data were used to plot graphs. Vybrant dye staining was employed to assess the distribution of cells in the G1, S, and G2/M phases.

For apoptosis detection, an Annexin V/PI staining kit (Molecular Probes, #V13241) was used, with 5 μM camptothecin (MedChemExpress, #HY-16560) serving as a positive control. Sixteen spheroids were collected, dissociated with trypsin as described above, and resuspended in 200 μL of 1X annexin-binding buffer. Cells were incubated with 1 μl of Alexa Fluor 488 Annexin V (Component A) and 0.2 μl of 100 μg/ml PI working solution at room temperature for 15 min. Stained cells were then analyzed using a BD Canto flow cytometer, measuring fluorescence emission at 530 and 575 nm (or equivalent) with 488 nm excitation.

For this experiment, biomolecules were extracted and measured from both Drosophila larvae fat bodies and HepG2 cell spheroids. Fat bodies were dissected from 10 male larvae at 120 hr after egg laying (AEL) and collected in 150 μl of 1X PBS. The samples were lysed by performing three freeze–thaw cycles. A total of 24 spheroids were homogenized in 150 μl of radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, and 2 mM EDTA).

Lysates were centrifuged at 7500 × g for 5 min to remove debris. The resulting supernatant was used for the following measurements. Triglyceride measurement: For triglyceride analysis, 30 μl of the supernatant was incubated for 10 min at 75°C. Following this, 10 μl of homogenate from the spheroids or 2 μl from fat bodies was added to 200 μl of Triglycerides Reagent (Thermo Fisher Scientific, #TR22421). The mixture was incubated for 10 min at 37°C in 96-well microplates with gentle shaking. Absorbance was measured at 550 nm using the Synergy Neo2 multimode plate reader (BioTek).

Protein measurement: Protein concentrations were determined by mixing 2 μl of the supernatant with 200 μl of BCA Protein Assay Reagent (Thermo Scientific, #23225). The mixture was incubated for 30 min at 37°C with gentle shaking in 96-well microplates. Absorbance was measured at 560 nm.

RNA and DNA separation: The remaining 100 μl of the supernatant was processed for RNA and DNA separation using TRIzol reagent (1 ml; Invitrogen, # 15596026). To separate RNA, 0.2 ml of chloroform was added to the sample, followed by centrifugation at 12,000 × g for 15 min at 4°C. The RNA in the interphase was purified using 75% ethanol, following standard methods, and quantified using a NanoDrop spectrophotometer. DNA was extracted from the organic phase by adding 100% ethanol, followed by isopropanol precipitation. The resulting pellets were washed with 0.3 M guanidine hydrochloride in 95% ethanol, resuspended in 0.1 M sodium citrate in 10% ethanol (pH 8.5), and washed with 75% ethanol. Finally, the DNA pellets were resuspended in 0.1 ml of 8 mM NaOH by pipetting. The pH was adjusted to 7.2 with HEPES, and the DNA was quantified using a NanoDrop spectrophotometer.

For HepG2 cells, 1 × 10⁶ cells treated with 1 µg/ml dox for 24 hr were scraped in 1.5 ml tube. For NADH/NAD⁺ ratio analysis from spheroids, 18 spheroids were treated with 1 µg/ml dox for 6 days, and then resuspended in 1 ml 1X PBS. Scraped cells or pooled spheroids were centrifuged (13,500 × g, 10 s, 4°C) and resuspended in 250 µl lysis buffer. Fat bodies from 10 male larvae were dissected in 1X PBS and resuspended in the lysis buffer provided with the kit. To separate cytoplasmic and mitochondrial fractions by rapid subcellular fractionation, lysates were centrifuged (13,500 × g, 10 s, 4°C) and the supernatant was collected for the cytosolic fraction, while the remaining pellet contained the mitochondria. Mitochondrial pellet was resuspended in 100 µl lysis buffer.

NADH to NAD⁺ ratios were measured using Amplite Fluorimetric NAD⁺/NADH ratio assay kit (AAT Bioquest, #15263) as directed by the instructions provided. Briefly, 25 µl of cytoplasmic or mitochondrial fractions were mixed with either NADH or NAD⁺ extraction solution. Samples were incubated at 37°C for 15 min. Later, 25 µl of either NAD⁺ or NADH extraction was added, which was followed by incubation with 75 µl mix of NADH sensor buffer and NAD⁺/NADH recycling enzyme for 1 hr at RT. Fluorescence intensity was recorded at 540 nm excitation and 590 nm emission.

HepG2 cells were directly scraped into RIPA supplemented with protease and phosphatase inhibitors (Roche Molecular, #04906845001). For spheroids, 18 spheroids were pooled, washed with 1X PBS, and then incubated in RIPA buffer with the protease and phosphatase inhibitor cocktail. After 45 min on ice with vortexing every 15 min, lysates were centrifuged at 16,000 × g for 15 min at 4°C to remove insoluble material.

Protein concentration was measured using the Bicinchoninic Acid (BCA) protein assay (Thermo Fisher Scientific, 23225). Samples were mixed with 4x sample loading buffer and heated at 95°C for 5 min. Protein samples (15 µg) were separated on SDS–polyacrylamide gel electrophoresis (SDS–PAGE) at 20 mA per gel, transferred onto a 0.45 µm nitrocellulose membrane (GE Healthcare) using the Mini Trans-Blot module (Bio-Rad) at 120 V for approximately 2 hr.

The membrane was blocked with 5% non-fat milk (Serva) in Tris-buffered saline with 0.05% Tween 20 (TBS-T) for 1 hr. It was then incubated overnight with primary antibodies diluted in TBS-T. The next day, the membrane was washed with TBS-T and incubated with fluorophore-conjugated secondary antibodies in TBS-T for 1 hr. Following additional washes with TBS-T, fluorescence was detected using the Odyssey CLx imaging system (LI-COR Biosciences) and analyzed using FIJI software. Antibodies used are listed in the Key Resources Table.

Ten million HepG2 cells with MPC+ and EV control expression were grown in 2D culture and treated with 1 µg/ml dox in HPLM supplemented with 10% HPLM. After 24 hr, the culture medium was collected and quenched with 1:4 volume of 100% methanol. Cells were washed and quenched with 1 ml of 80% methanol in water. Cell lysates were then subjected to three rapid freeze-thaw cycles and then spun at 16,000 × g for 10 min at 4°C. The supernatants were evaporated using a SpeedVac concentrator. Each sample or treatment had 4–5 replicates.

Ten million HepG2 cells with MPC+ and EV control expression were grown in 2D culture and were treated with 1 µg/ml dox in HPLM. After 24 hr, the culture medium was changed to ¹³C-glucose tracing media: glucose-free HPLM supplemented with 4.5 g/l ¹³C glucose and 10% dialyzed FBS. 0, 1, 2, and 4 hr later, cells were washed and quenched with 1 ml of 80% methanol in water. Cells were scraped out in methanol, and lysates were then subjected to three rapid freeze–thaw cycles and then spun at 16,000 × g for 10 min at 4°C. The supernatants were evaporated using a SpeedVac concentrator.

Ten million HepG2 cells with MPC+ and EV control expression were grown in 2D culture and treated with 1 µg/ml dox in DMEM without glucose, without glutamine (Gibco, #A1443001) with 10% dialyzed FBS. After 16 hr, the culture medium was replaced with 6 ml of 1 µg/ml dox in DMEM without glucose, without glutamine, and without FBS. 3 hr later, the culture medium was replaced with 6 ml of 1 µg/ml dox and 20 mM lactate (Sigma, #490040) containing DMEM without glucose or glutamine or FBS. After 4 hr, 300 μl culture medium was collected and quenched with 1:4 volume of 100% methanol. Cells were washed and quenched with 1 ml of 80% methanol in water. Cell lysates were then subjected to three rapid freeze-thaw cycles and then spun at 16,000xg for 10 min at 4°C. The supernatants were evaporated using a SpeedVac concentrator. Each cell type had four to five replicates.

Dried metabolites were derivatized and prepared for Gas chromatography following standard methods. Dried samples were resuspended in 30 μl of anhydrous pyridine with methoxamine hydrochloride (10 mg/ml) and incubated at RT overnight. The next day, the samples were heated at 70°C for 15 min and centrifuged at 16,000 × g for 10 min. The supernatant was transferred to a pre-prepared gas chromatography–mass–mass spectrometry autoinjector vial with 70 μl of N-(tert-butyldimethylsilyl)-N-methyltrifluoroacetamide (MTBSTFA) derivatization reagent. The samples were incubated at 70°C for 1 hr, after which aliquots of 1 μl were injected for analysis. Samples were analyzed using either an Agilent 6890 or 7890 gas chromatograph coupled to an Agilent 5973N or 5975C Mass Selective Detector, respectively. The observed distributions of mass isotopologues of glucose, pyruvate, citrate, succinate, aspartate, glutamate, malate, fumarate, and phosphoenolpyruvate were corrected for natural abundance.

Following standard methods, dried metabolites were resuspended in 100 μl of 0.03% formic acid in analytical-grade water, vortexed, and centrifuged to remove insoluble material. 20 μl of supernatant was collected and injected to AB SCIEX QTRAP 5500 liquid chromatography/triple quadrupole mass spectrometer (Applied Biosystems SCIEX). Chromatogram review and peak area integration were performed using MultiQuant (version 2.1, Applied Biosystems SCIEX). The peak area for acetyl CoA, ADP, ATP, NAD⁺, and NADH was normalized against the total ion count of that sample.

Oxygen consumption rates (OCR) were measured using an XFe96 Extracellular Flux Analyzer (Agilent) according to the manufacturer’s instructions. Cells were plated at a density of 60,000 cells per well in Seahorse microplates (Agilent) and allowed to adhere for 6 hr. MPC was induced with 1 μg/ml doxycycline overnight. Afterward, the cell culture media was removed and replaced with Seahorse assay media, which consisted of DMEM supplemented with 4 mM glutamine. OCR was assessed under basal conditions and following sequential injections of 1 mM glucose, oligomycin (2 μM), FCCP (0.5 μM), and a mix of rotenone plus antimycin A (2 μM each). Immediately after the measurements, cells were lysed in RIPA buffer, and the protein concentration was used to normalize the OCR data.

Rat primary hepatocytes (Lonza, #RWCP01) were thawed and plated in a 24-well plate at a density of 0.16 million cells per well in Hepatocyte Plating Medium (Lonza, #MP100). Four hours post-plating, the medium was replaced with Hepatocyte Basal Medium (HBM) supplemented with BCM SingleQuots (Lonza, #CC-4182), including Bovine Pituitary Extract (BPE), Insulin, Hydrocortisone, Gentamicin/Amphotericin-B (GA), Transferrin, and human Epidermal Growth Factor (hEGF).

Twenty-four hours after plating, the hepatocytes were transfected with 1 μg of either pT3.GFP or pT3.MPC2Flag-P2AT2A-MPC1HA using Lipofectamine 3000 (Invitrogen, #L3000001). Forty-eight hours post-transfection, cells were lysed in 1X RIPA buffer, and Western blot analysis was performed to confirm the overexpression of MPC1 and MPC2.

Forty-eight hours post-transfection, hepatocytes were fixed in 4% paraformaldehyde in 1X PBS. Following several washes with 0.1% PBT, cells were incubated in 5% BSA and stained with anti-MPC1 antibody overnight at 4°C. MPC1 was detected using a secondary anti-rabbit Alexa Fluor antibody. Cells were also stained with Phalloidin Red. Coverslips were mounted in DAPI-supplemented Vectashield, and images were captured using an LSM 880 confocal microscope. Cell area marked by Phalloidin Red was quantified using differential interference contrast filter and analyzed with FIJI software.

For protein synthesis measurement, hepatocytes were incubated with 20 μg/ml puromycin for 30 min. Puromycin-tagged peptides were visualized by immunostaining with rabbit anti-puromycin antibody (1:500) and mouse anti-Flag M2 antibody (1:1000, Sigma, F1800), followed by the appropriate secondary antibodies. Images were captured with the LSM 880 confocal microscope. Puromycin intensity was measured using FIJI, and the percent change in puromycin intensity in Flag-positive cells was compared to Flag-negative cells.

Gluconeogenesis was assessed using lactate and pyruvate as substrates. Forty-eight hours post-transfection, hepatocytes were incubated in low-glucose DMEM with 1 mM sodium pyruvate and 4 mM glutamine without FBS. After 16 hr, hepatocytes were treated with 200 ng glucagon in no-glucose DMEM without FBS for 3 hr. Cells were then cultured in media with 20 mM lactate and 2 mM pyruvate, with and without 10 μM UK5099, for 2 and 4 hr. Glucose levels in the media were measured using the Amplex Red Glucose Assay Kit (Invitrogen, #A22189) according to the manufacturer’s instructions. Glucose production per cell per hour was plotted on a graph.

The compartmentalized NADH to NAD⁺ ratio in hepatocytes was quantified 48 hr post-transfection using the Amplite Fluorimetric NAD⁺/NADH Ratio Assay Kit as described previously.