Huh-7 and HepG2 cells were obtained from the American Cell Culture Collection (ATCC, Manassas, VA, USA). They were cultured in DMEM medium (Thermo Fisher Scientific, Waltham, MA, USA, 41965062) supplemented with 10% fetal bovine serum (FBS, Hy-Clone, Logan, UT, USA, SH30910.03), 100 U/mL ampicillin and 100 μg/mL streptomycin (Thermo Fisher Scientific, 15240-062), maintained at standard conditions, 37°C, 5% CO₂ and 90% humidity. Cells were treated with Fructose (Fru, Sigma-Aldrich, St. Louis, MO, USA, F0127) at the physiological concentration (1 mM) [22]. Cell lines were mycoplasma free. Cell line authentication was confirmed using short tandem repeat analysis.

To develop the assays, we used DMEM with high glucose (High Glc, 11 mM, Sigma-Aldrich, St. Louis, MO, USA, G8270), or low glucose (Low Glc, 0.4 mM), and supplemented it with 1 mM Fru with FBS. The cell proliferation and viability assays were carried out using the cell counting kit-8 (CCK-8, Dojindo Lab, Kumamoto, Japan, CK04), following the manufacturer’s instructions, and dimethyl sulfoxide (DMSO) 30% was used as a negative control of viability. On the other hand, to evaluate the importance of the pentose phosphate pathway we inhibited glutathione reductase (GR) enzyme. We treated the cells with 2-acetylamino-3-[4-(2-acetylamino-2-carboxyethylsulfanylthiocarbonylamino) phenylthiocarbamoylsulfanyl] propionic acid (2-AAPA) at 0, 40, and 80 μM for 48 h with High Glc and with/without Fru.

Huh-7 and HepG2 cells were plated using 2 × 10⁴ cells/well in 96-well plates with clear bottoms. Cells were treated with High Glc media with 1 mM of Fru for 48 h before the assay. The day of the assay, cells were treated with DMEM Glc-free media (Thermo Fisher Scientific, 11966025) for 1 h and then incubated with treatment media containing 1-deoxy-1-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]-D-fructose (1-NBDF; Cayman Chemicals, Ann Harbor, MI, USA, 940961-04-6) 100 μM for 1 h at 37°C, 5% CO₂ and 90% atmospheric humidity. Then, the cells were washed twice with PBS, and each well remained with 100 μL of PBS for fluorescence determination. 1-NBDF fluorescence was measured at 475 nm excitation/550 nm emission. Data were normalized against cellular density in each well, determined by Hoechst fluorescence measured at 350 nm excitation/461 nm emission.

After treatments, RNA extraction was performed using TRIzol reagent (Ambion, Inc, Austin, TX, USA, 15596018). A specific RT reaction was established with 2 μg of total RNA using M-MLV RT (Promega, Madison, WI, USA, M170A) and following manufacturer instructions.

The qRT-PCR analysis was performed with a CFX96 Touch (Bio-Rad, Hercules, CA, USA) thermal cycler in a 96-well reaction plate. The 10 μL PCR reaction mix contained 5 μL 2X SYBR Green PCR Master Mix (Bio-Rad, 1725271), 200 nM of each primer, and 1 μL cDNA template. Reactions were incubated for 10 min at 95°C, followed by 40 cycles of 30 s at 95°C and 60 s at specific khk-c and khk-a primers [15] temperatures (57°C). Melting analysis of the PCR products was also conducted to validate the amplification of the product. The expression level of mouse rps18 [23] was used as an internal reference. Relative gene expression level was calculated with the 2^−ΔΔCT method. Primer sequences are listed in Table S1.

HepG2 and Huh-7 were used to evaluate mitochondrial substrate oxidation by carrying out a Mito Fuel Flex Assay (Agilent Technologies, 103260-10) using Agilent Seahorse XFe96 (Agilent Technologies, Santa Clara, CA, USA). Then, the Glycolytic Rate Assay was run using Huh-7 cells using the Agilent Seahorse XFe24 Analyzer (Agilent Technologies) and the Agilent XF Glycolytic Rate Assay Kit (Agilent Technologies, 103344-100). We followed the manufacturer’s instructions for both assays and measured the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR). Mito Fuel Flex Assay reflects the oxidation percentage for each mitochondrial substrate, Glc, glutamine (Gln), or palmitic acid, represented by capacity, dependency, and flexibility. For the second assay, we represented the GlycoPer (glycolysis contribution to media acidification), MitoPER (mitochondrial contribution to media acidification), and PER (proton efflux rate).

1 × 10³ cells were pre-coated with 300 μL agarose per well (6 mL DMEM 20% FBS + 4 mL 2% agarose). Once the agarose had solidified, we seeded the cells dissolved on 0.25% agarose and 20% DMEM. The next day, we applied the treatments. The Fru treatment was applied 24 h after the cells were seeded and left for 14 days. The spheroids were counted and photographed using a Carl Zeiss VERT.A1 microscope (Carl Zeiss VERT.A1, Zeiss, Oberkochen, Germany).

Huh-7 and HepG2 cells were seeded using a 2-well culture insert with a defined cell-free gap for wound healing and migration assays (ibidi, Gräfelfing, Germany, 81176) (30,000 cells per well). After 12 h, cells were previously treated with Low Glc serum-free media for 24 h, and then the media was changed and treated with or without 1 mM of Fru in High Glc or Low Glc serum-free media. We followed the experiment for 24, 48, and 72 h.

The assay was performed by seeding 1 × 10⁶ cells per bottle (Millipore-Sigma, Saint Louis, MO, USA) using DMEM with Glc (Sigma-Aldrich, St. Louis, MO, USA) with FBS. The cells were pre-treated with Fru and left for 72 h. Then, 335,000 cells were re-seeded in a 35 mm plate and treated with Fru: 1 mM and cis-diaminplatin (II)-dichloride (CDDP) (ACCOCIT injectable solution 10 mg/10 mL, Accord Farma, Mexico City, Mexico): 22.11 μM to carry out the test. We followed the experimental model depicted in Figure 1.

After treatments, cells were incubated with PBS-EDTA (5 mM) for 5 min at 37°C and trypsin-EDTA (1X) for 5 min at 37°C. They were then centrifuged at 1,800 rpm for 5 min and resuspended in PBS-heparin to proceed with the protocol. Then, 400 μL of cells were stained with 1 μL of iodine propidium. The assay used a flow cytometer (FACS Aria Fusion, BD, Franklin Lakes, NJ, USA).

Glucose-6-phosphate dehydrogenase (G6PD) activity was measured using 5 mM glucose-6-phosphate (substrate) and 0.4 mM NADP⁺ (coenzyme) in 50 mM Tris-HCl containing and 5 mM MgCl₂ (pH = 8). The assay was performed in a 96-well plate, and the reaction mixtures consisted of 180 µL of buffer, 20 µL of NADP⁺, 20 µL of substrate, and 20 µL of cell lysate. The mixture, excluding the substrate and coenzyme, was incubated for 5 min at room temperature, and the substrate and coenzyme were added just before starting the assay. NADPH production from the reaction was monitored kinetically every 5 min up to 1 h at an absorbance of 340 nm and 37°C.

Cells were seeded in 96-well-well plates. We established two models: (1) to pre-treat the cells with Fru (1 mM) for 48 h, or (2) to treat the cells for 5, 15, 30, and 60 minutes with Fru (1 mM); both conditions were treated in a High Glc media. After that time, we incubated the cells for 15 min with dihydroethidium (DHE, 5 μM) in the dark, at room temperature, to determine superoxide radical by detecting ethidium fluorescence. DHE-derived fluorescence was determined in a multipedal reader (Beckman Coulter DTX 880 Multimode Detector) at excitation and emission wavelengths of 485 nm and 570 nm.

To study the impact of sugar metabolism on tumor formation capacity and therapy resistance, cells were pretreated for 5 days with DMEM High Glc with or without Fru (1 mM). Treated and control cells were inoculated on the extraembryonic chorioallantoic membrane (CAM) of chick embryos on the embryonic developmental day (EDD) 10 following the method described by Gerardo-Ramirez et al., 2019 and Meenen et al., 2022 [24, 25]. Briefly, 10 eggs per group were inoculated with 2 × 10⁶ cells suspended in 50 μL of PBS in at least three independent experiments and incubated at 37.8°C and 60% humidity up to the end of the experiments. Four days after grafting (EDD14), cells were treated with CDDP using 100 μM [26]. At EDD17, grown tumors were excised, measured, weighed, and photographed as described previously by Busch and collaborators, 2015 [27], Figure S1. This in vivo model has been approved by the ethical committee of the Universidad Autónoma Metropolitana-Iztapalapa (UAM-I): CECBS22-02.

All experiments were carried out, at least in triplicate. Data are expressed as the mean ± SEM. The t-test was performed to compare means, two-way ANOVA between treatment groups, and the Tukey-Kramer test. The GraphPad Prism 8.2.1 Software for Mac OsX (GraphPad Software, San Diego, California, USA) was used for data processing. Differences were considered significant at p ≤ 0.05.

We have treated cells with Fru (1 mM) under two different Glc concentrations at different times, as described in the materials and methods section in Figure 1. Under treatments, we observed refracted cells in media without Glc, even in the presence of Fru, after 48 h, suggesting that Glc starvation induces cell death in both conditions (black arrows, Figure S2A and B); however, the cells were able to survive in the combined treatment (Low Glc + Fru) compared to Low Glc alone, suggesting that Glc promotes Fru uptake and cancer cells can switch the substrate to obtain energy.

Next, we determined the relevance of Glc concentration in Fru internalization. In two scenarios, we explored the effect of low and High Glc concentrations in HepG2 and Huh-7 cells. In the first one, we determined Fru uptake in both cell lines with no previous presence of Fru (Figure 2A). Our data indicate that High Glc levels facilitate Fru uptake in both liver cancer cell lines. In the second scenario, cells were previously conditioned with Fru for 48 h, after which the media was changed to the previously established measurement conditions. Figure 2A shows that Huh-7 cells exhibited a 2-fold increase in Fru uptake in the presence of Glc, while there was no significant change in HepG2 cells. These findings suggest that using Fru may elicit an adaptive response, which is dependent on the origin and aggressivity of the cell.

We measured cell proliferation under these conditions to address the impact of Fru uptake. The HepG2 cell line did not proliferate in a High Glc media; however, Huh-7 cells exhibited increased proliferation compared to not treated (NT) cells after 48 h of treatment (Figure 2B and 2C), supporting previous results that suggested Fru metabolism has a differential effect on the differentiation level of transformed cells. Fru-alone treatment improved cell proliferation and survival under Low Glc conditions (Figure 2D and 2E), indicating that Fru could be an energy provider. Fru did not affect cell viability (Figure S2C–F).

Then, we aimed to determine whether proliferation enhancement was associated with fructokinase isoform switching (KHK-C to KHK-A), as previously reported [11]. Fru treatment (1 mM, 48 h) induced significant isoform switching, particularly in the more aggressive Huh-7 cell line, with higher expression in Fru-treated cells and downregulation regarding NT cells (Figure 3A and 3B); however, in HepG2 cells, higher expression of isoform khk-c was still observed with no difference in isoform khk-a expression (Figure 3C and 3D).

Metabolic adaptation is related to the aggressiveness of cancer cells. Considering the earlier results, we investigated if the isoform switching induces metabolic rewiring to enhance proliferation and survival. We analyzed the ECAR and OCR to further explore the metabolic changes by implementing the Mito Fuel Flex Assay and Glycolytic Rate Assay using Seahorse technology.

We aimed to identify the metabolic substrates required for mitochondrial respiration by performing a Mito Fuel Flex Assay. Following treatment with High Glc, both cell lines revealed comparable Glc and fatty acid oxidation capacity, regardless of the presence or absence of Fru. However, analysis of the capacity for Gln oxidation revealed a reduction in the capacity of Huh-7 cells to oxidize this substrate compared to HepG2 cells (Figure 4A). Subsequently, the dependency and flexibility of both cell lines to oxidize Glc, Gln, and fatty acids in the mitochondria were compared. Our findings revealed a significant decrease in fatty acid dependency in Fru-treated HepG2 cells compared to NT (Figure 4B). Approximately 40% of the total mitochondrial energy in both cell lines depended on Glc oxidation with or without Fru supplementation. Huh-7 cells demonstrated a complete dependency for Gln oxidation in both conditions, yet less than 20% of the energy is obtained from this pathway (Figure 4C).

In contrast, HepG2 cells exhibited greater flexibility in utilizing different substrates to meet the energy demands in stressful conditions than Huh-7 cells (Figure 4B). This adaptability may be associated with the observed survival and low proliferation rates in HepG2 cells. These findings indicate that glucose oxidation represents the primary mitochondrial pathway for energy acquisition, and Fru does not induce significant changes at this metabolic level in the cells.

Considering that Fru treatment may elicit differential effects regarding cells’ origin and differentiation status, we pursued further exploration in Huh-7 cells and conducted a Glycolytic Rate Assay. Metabolic alterations were evident in Fru-treated cells, exhibiting a decline in glycolysis and a reduction in the PER relative to the control group (Figure 4D and 4G). We could not observe any change in mitochondrial metabolism (MitoPER and MitoOCR) in Fru-treated cells in either scenario (Figure 4E and 4G) or between treatments (Figure 4F), thereby corroborating the earlier results obtained from the Mito Fuel Flex Assay. These results indicate that Fru impairs glycolysis, which could activate alternative pathways to support survival and proliferation.

It is known that the pentose phosphate pathway (PPP) is critical for anabolic activities in cancer cells. The pathway is critically regulated by the activity of G6PD, the rate-limiting enzyme in the PPP; in fact, the upregulation of G6PD is associated with a poor outcome in liver cancer patients as indicated by the Kaplan-Meier plot (Figure 5A), obtained from the GEPIA2 software [28]. Fru in the medium increased levels of NADPH (Figure 5B), and G6PD achieved Vmax in the shorter time frame (Figure 5C). Moreover, to investigate whether PPP plays a pivotal role in the metabolic effects of Fru on cancer cells, we used the 2-AAPA to inhibit glutathione reductase (GR) activity, one enzyme that uses the NADPH to reduce oxidized glutathione (GSSG); Huh-7 cells treated with Fru exhibited higher viability than NT cells, which is related to the higher proliferation obtained previously. However, cells treated with 2-AAPA demonstrated a 43% reduction in viability compared to NT, confirming that Fru favors the PPP (Figure 5D). Then, we established two models. We followed the previous pre-treatment with Fru for 48 h (model 1) or treated the cells directly for 5, 15, 30, and 60 min with Fru (model 2) and measured the ROS by superoxide anion production. As we expected, short-time treatments did not counteract the production of ROS (model 2, Figure 5E); however, the pre-treatment with Fru agreed with our previous results and decreased superoxide anion content (model 1, Figure 5F), demonstrating that the metabolic rewiring is induced by the monosaccharide.

Metabolic adaptation in cancer cells may promote therapy failure and aggressiveness. First, to test the hypothesis, we pretreated cells with Fru for 72 h and then with CDDP (22 μM) for 48 h. We observed that Fru-treated cells exhibited no difference in morphology regarding NT cells, but CDDP treatment induced evident damage (Figure 6A). Flow cytometry analysis of cell death revealed that Fru treatment induced significant chemoresistance to the CDDP cytotoxicity compared to CDDP treatment alone, according to the dot plots (Figure 6B) and percentage of cell death (Figure 6C). These results strongly suggest that Fru is bringing an aggressive phenotype. To gain more confidence, we assayed some parameters of cellular aggressivity.

Fru treatment significantly increased spheroid number and size in the Huh-7 cell line; spheroids’ phenotype presented less compaction, probably suggesting an increment in cellular scattering (Figure 6D and 6F). Although HepG2 cells presented less spheroid formation due to their nature, we observed similar findings to Huh-7 cells (Figure 6G–6I). To gain evidence of cell migration, suggested by the irregular borders observed in spheroids, we conducted a wound-healing assay demonstrating that Fru treatment enhanced wound closure in both Huh-7 (Figure 6J–6L) and HepG2 cells (Figure S3A–3C) compared to the control group. In addition, we tested the hypothesis that Fru treatment promoted resistance to starvation and evaluated the migratory capacity in both cell lines using Low Glc media. The cells treated with Fru demonstrated enhanced survival and migration capabilities. Conversely, the NT cells exhibited a notable decline in migration and started to detach from the plate after 72 h of treatment (Figure 6K and 6L, Figure S3B and C). These findings demonstrate that liver cancer cells can use Fru metabolism to facilitate tumorigenesis, enhancing their survival potential in response to therapeutic interventions.

To corroborate the in vitro findings, we employed the in vivo model of the chick CAM model. Two days before cell inoculation, Huh-7 cells were treated with Fru. On EDD10, 2 × 10⁶ cells were inoculated. Two days before completing the treatments (EDD15), the tumors were treated with Fru (1 mM), CDDP (100 μM), or a combination of CDDP and Fru.

Tumor growth under Fru treatment exhibits a prominent vascularization regarding the NT (Figure 7A). The size of the tumors is significantly increased in Fru-treated compared with NT and CDDP-induced good toxicity, but this was impaired by Fru treatment (Figure 7B and 7C). The tumor capacity remained consistent across all treatment groups (Figure 7D). However, the Fru-treated cells exhibited a significant tendency for embryo death compared to the NT cells (Figure 7E). As previously shown, the results are associated with the observed differences in cell migration capacity. In addition, Fru promoted the formation of larger tumors compared to NT. In line with our earlier observations, Fru treatment enhanced cell survival in the presence of CDDP. Our results indicate that Fru plays a critical role in the aggressiveness of HCC cells in vitro and in vivo, revealing the underlying causes of therapy failure.