Human liver cancer cell lines (HepG2 and Hep3B), non-small cell lung cancer cell lines (A549, H1299 and H460) and breast cancer cell lines (MCF-7, MDA-MB-231 and T47D) were purchased from the American Type Culture Collection and cultured in RPMI 1640 (PAA Laboratories GmbH) supplemented with 10% fetal bovine serum (FBS; PAA Laboratories GmbH) and 100 U/ml penicillin-streptomycin (Gibco; Thermo Fisher Scientific, Inc.) at 37°C in a humidified atmosphere with 5% CO₂. The authenticity of all eight cell lines used in the present study was validated by short tandem repeat (STR) DNA profiling, which was performed using PowerPlex 18D system by commercial service (Cosmo Genetech).

D-(+)-glucose, D-(+)-glucosamine hydrochloride, cytochalasin B and phloretin were purchased from MilliporeSigma, while 2-DG and 3-BrPA were purchased from Tokyo Chemical Industry. In the competition experiment between glucose and glucosamine, glucose-free RPMI 1640 medium (MilliporeSigma) was used and cells were pre-incubated for 24 h with the indicated doses of glucose prior to being treated with glucosamine.

Total RNA was isolated from cells using TRIzol^® reagent (Thermo Fisher Scientific, Inc.) and chloroform extraction. First-strand cDNA was synthesized from 5 µg of total RNA using Superscript II reverse transcriptase (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Semiquantitative PCR was performed in a thermal cycler (MyCycler; Bio-Rad Laboratories, Inc.) using gene-specific primer sets and sequences of primers are listed in the supplementary Table SI. The thermocycling conditions were as follows: Initial denaturation at 95°C for 5 min, followed by 25–45 cycles of denaturation at 95°C, annealing at 42–60°C for 30 sec (depending on the target gene), and elongation at 72°C for 1 min, with a final extension at 72°C for 5 min. The reactions were stopped during the exponential phase to ensure accurate comparisons. ACTB was used as a reference gene. PCR products were electrophoresed on 1.2–1.5% agarose gel and visualized using ethidium bromide.

A549 and HepG2 cells (2.5×10⁵ cells/well) were seeded in 6-well plates and transfected with double-stranded small interfering RNA (siRNA). Target-specific siRNAs were designed for knockdown of GLUT1, GLUT2 and HIF1A were from Bioneer Corporation and MBiotech, respectively. A scrambled siRNA (siSCR) was synthesized and validated by Bioneer Corporation. The sense and anti-sense sequences of siRNAs are listed in the Table SII. Following 25 pmole of siRNA transfection for 24 h at 37°C using Lipofectamine^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions, subsequent experiments were conducted. Knockdown of GLUT1, GLUT2 and HIF1A expression were validated by semiquantitative RT-PCR. The thermocycling conditions were as follows: Initial denaturation at 95°C for 5 min, followed by 25–45 cycles of denaturation at 95°C, annealing at 42–60°C for 30 sec (depending on the target gene), and elongation at 72°C for 1 min, with a final extension at 72°C for 5 min. The reactions were stopped during the exponential phase to ensure accurate comparisons. ACTB was used as a reference gene. PCR products were electrophoresed on 1.2–1.5% agarose gel and visualized using ethidium bromide.

Whole cell lysates were prepared using RIPA-B lysis buffer containing 150 mM NaCl, 1 mM EDTA (pH 8.0), 20 mM Tris-Cl (pH 7.4), 1% NP-40, 0.5% sodium deoxycholate, 1% Triton X-100, 1 mM Na₃VO₄, 5 mM NaF, 10% glycerol and protease inhibitor cocktail (Roche Diagnostics GmbH). Protein concentration was determined using BCA protein assay kit (Thermo Fisher Scientific, Inc.). Equivalent amounts of proteins (10–20 µg per lane, depending on the target protein) were separated by electrophoresis on 6, 8, 10 or 12% SDS-PAGE and then transferred to a polyvinylidene fluoride membrane (Pall Corporation). Membranes were blocked with 5% skimmed milk at room temperature for 1 h. The membranes were then incubated overnight at 4°C with primary antibodies against phosphorylated (p-)Akt (1:1,000, cat. no. 9271, Cell Signaling Technology, Inc.), Akt (1:1,000, cat. no. 4691, Cell Signaling Technology, Inc.), a/β-tubulin (1:1,000, cat. no. 2148, Cell Signaling Technology, Inc.), IGF-1R (1:1,000; cat. no. sc-713, Santa Cruz Biotechnology, Inc.) and COX-2 (1:1,000; cat. no. sc-1745, Santa Cruz Biotechnology, Inc.). After washing with Tris-buffered saline containing 0.1% Tween 20, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (1:10,000; cat. no. AbC-5001 for anti-mouse antibody and cat. no. AbC-5003 for anti-rabbit antibody, Abclon) for 1 h at room temperature. Immunoreactive bands were detected using enhanced chemiluminescence (AbSignal; Abclon).

Cells (3.0×10³ cells/well) were seeded into 96-well plates, incubated overnight at 37°C and treated with glucosamine or other agents. Control groups were given dimethyl sulfoxide (DMSO; 0.1% final concentration) or phosphate-buffered saline (PBS) vehicle. Cell viability was measured by MTT assay. An insoluble purple formazan precipitate was solubilized in DMSO and the absorbance was determined by microplate spectrometer at 562 nm.

A549 and HepG2 cells (1.5×10⁵) were plated onto 60×15 mm dishes for 1 day and then cells were treated with glucose (1, 10, 20, or 50 mM) and glucosamine (1, 2, or 5 mM) at 37°C. Glucose was administered for 48 h and glucosamine was co-administered with glucose during the last 24 h of that period. After treatment, cells were washed once with PBS, trypsinized and harvested. All the harvested cells were fixed in 70% ethanol for 1 h at 4°C. After centrifugation for 5 min. at 1,000 × g and 4°C, cell pellets were washed twice with PBS and then stained with PI containing RNase A (40 µg/ml) for 30 min at 4°C in the dark. Total DNA contents were detected by BD FACSCalibur flow cytometer (BD Biosciences), and analyzed using FlowJo 10.10.0 (FlowJo LLC).

A549 and HepG2 cells (1.5×10⁵) were plated onto 60×15 mm dishes for 1 day and after treated with glucosamine for 48 h, cells were harvested and washed twice with PBS on ice. Then, cells were stained using the FITC Annexin V Apoptosis Detection Kit I (BD Pharmingen; BD Biosciences). Cells were resuspended in 1× binding buffer containing 5 µl of fluorescein isothiocyanate (FITC)-Annexin V and 5 µl of PI. After staining, cells were detected by BD FACSCalibur flow cytometer (BD Biosciences) at 488 and 633 nm, and analyzed using FlowJo 10.10.0 (FlowJo LLC). The apoptotic rate was calculated as the percentage of early apoptotic cells and the percentage of late apoptotic cells. All procedures were carried out according to the manufacturer's instructions.

Cells (2.5×10⁵ cells/well) were seeded in 6-well plates and incubated under RPMI 1640 medium containing the indicated concentrations of glucosamine. Glucosamine in harvested culture medium was reacted using D-Glucosamine Assay Kits (K-GAMINE; Megazyme) and optical density was measured with a spectrophotometer at 340 nm. All procedures were carried out according to the manufacturer's instructions.

The cellular glycolytic efficiency could be estimated by measuring the cellular lactate generation rate. Cells (2.5×10⁵ cells/well) in the exponential growth phase were plated in 6-well plates for 1 day and cells were washed and treated with indicated concentrations of glucosamine for an additional 24 h. Aliquots of the culture medium were withdrawn and used for analysis of secreted lactic acid using the YSI 2300D STAT Plus Glucose and Lactate Analyzer (YSI Inc.).

A549 cells were treated with DMSO or 5 mM of glucosamine for 24 h at 37°C and RNA was isolated. From each sample, 500 ng of total RNA was biotinylated and amplified using the Illumina TotalPrep RNA Amplification Kit (Ambion; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. The cRNA yield was measured using RiboGreen RNA quantitation kit (Invitrogen; Thermo Fisher Scientific, Inc. and 750 ng of the cRNA sample was hybridized on a human HT-12 expression bead chip (Illumina, Inc.) profiling 48,804 transcripts per sample. The chips were stained with streptavidin and scanned using an Illumina BeadArray Reader (Illumina, Inc.). BeadStudio v3 (Illumina, Inc.) was used to quantile-normalize the data.

Gene ontology (GO) analysis was performed on genes that were increased or decreased by more than 2-fold in the glucosamine-treated sample compared with the control. The analysis was conducted using the gseGO function in the clusterProfiler package (ver. 4.14.3) (24) in R software (ver. 4.4.1, R Core Team, 2024) (25), based on the Gene Set Enrichment Analysis approach. GO terms which have adjust P-value <0.01 were considered markedly regulated GO. Up- or downregulated GO terms were featured by R software.

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed on differentially expressed genes using the enrichKEGG function from the clusterProfiler package in R software. The result of glycolysis pathway (hsa00010) was visualized using the Pathview package (Release 3.20) (26).

All statistical analyses were conducted using GraphPad Prism 8.0 (Dotmatics) and Microsoft Excel (Microsoft Corporation). For comparisons involving more than three groups, one-way ANOVA was performed, followed by Tukey's post hoc test. For comparisons between two groups, the Student's t-test was employed. P<0.05 was considered to indicate a statistically significant difference.

In 1953, J.H. Quastel and A. Cantero demonstrated, for the first time, that D-glucosamine inhibits the tumor growth in a xenograft mouse model (27). After their findings, a number of studies have sought to discover the mechanism and related molecules in this antiproliferative effect. A number of studies have demonstrated that glucosamine decreases the ATP level in cancer cells and inhibits oncogenes, such as IGF-1R, HIF-1α and COX-2 (8,9,21). However, the mechanism has not been described in detail. Therefore, the present study used a microarray to screen genes that are regulated directly or affected indirectly by glucosamine. Genes induced by glucosamine more than two folds were categorized mainly into pro-apoptosis and cell death (Fig. 1A), while genes downregulated by glucosamine more than two folds mostly belonged to cell cycle and DNA damage responses (Fig. 1B). Next, the present study examined whether the genes involved in carbohydrate metabolism, biomolecules biosynthesis and glycosylation, which pathways are highly associated with glycolysis (28–30), are regulated by glucosamine. Notably, as shown in Table SIII, several genes upregulated by glucosamine more than twofold were involved in these pathways. Also, pathway enrichment of differential genes was used and then the KEGG pathway analysis results viewed to see whether the differential genes are enriched in glycolysis. As a result, differential genes are represented in the glycolysis pathway and their expression is primarily downregulated by glucosamine (Fig. 1C). With all these findings, glucosamine's similar structure to glucose and its reduction of ATP levels in cancer cells (4), as demonstrated by previous studies, it was assumed that glycolysis is the main target pathway of glucosamine.

Other than its NH₂ group, glucosamine has a structure very similar to that of glucose. In fact, several studies have demonstrated that glucosamine competes with glucose in carbohydrate metabolism (31–33). Based on this, whether its antiproliferative effect is linked to this competitive relationship was investigated. A recent study has shown that glucosamine exhibits antiproliferative effects on a liver cancer cell line (34), while our previous research demonstrated similar effects on lung cancer cell lines (9). Therefore, the present study initially conducted experiments on HepG2 liver cancer cell line and A549 lung cancer cell line. Under low level of glucose, glucosamine inhibited proliferation of A549 and HepG2 cancer cells, which anti-proliferative effects of glucosamine were reversed by increasing glucose concentration (Fig. 2A). Glucose-treated cells also showed a dose-dependent recovery of the G0/G1 arrest induced by glucosamine (Fig. 2B). Since previous studies showed that glucosamine inhibits the IGF-1R/Akt pathway (9,35), the present study next investigated the effect of glucose on these target molecules. As expected, IGF-1R and p-Akt expression levels that had been reduced by glucosamine were restored as concentrations of glucose increased (Fig. 2C). To determine whether the competitive effect of glucose and glucosamine occurs during glycolysis in cancer cells, the level of lactate, which is produced from pyruvate during glycolysis was investigated. Glucosamine markedly suppressed lactate production, but this suppression was nearly reversed with a high dose of glucose (Fig. 2D). These findings suggested that the antiproliferative effect of glucosamine was closely related to competition between glucose and glucosamine, which occurs during glycolysis.

Glucosamine enters the cell mostly via GLUT2, along with other isoforms of GLUT (11). Although several studies have examined the kinetics of glucosamine transport, there is no direct evidence on how glucosamine, particularly its antiproliferative effects, influences these kinetics. Thus, it was hypothesized that cancer cells with higher expression levels of GLUT2 were more sensitive to the antiproliferative effect of glucosamine. First, the expression level of GLUT1 and GLUT2 in cancer cell lines originating from various types of cancer including liver cancer, lung cancer and breast cancer was screened. For this purpose, two liver cancer cell lines (HepG2 and Hep3B), three lung cancer cell lines (A549, H1299 and H460) and three breast cancer cell lines (MCF-7, MDA-MB-231 and T47D) were used. Among them, liver cancer cell lines showed higher levels of GLUT2 compared with the others, while most of the cell lines showed similar levels of GLUT1 (Fig. 3A). Then, to assess the correlation between glucosamine uptake and GLUT2 expression levels, a glucosamine consumption assay was performed and the results showed that liver cancer cell lines, which showed highly expressed GLUT2, consumed glucosamine more actively than others (Fig. 3B). Next, whether the expression level of GLUT2 and the rate of glucosamine uptake are related to the antiproliferative effect of glucosamine was investigated. As a result, liver cancer cell lines, which showed higher expressions of GLUT2 and actively consumed glucosamine, were more sensitive to glucosamine than others (Fig. 3C). Along with these results, HepG2 and Hep3B cells showed greater decreases in IGF-1R and p-Akt expression levels in the presence of glucosamine than A549 cells, which were the most sensitive among the remaining cell lines (Fig. 3D). Taken together, these results implied that the antiproliferative effect of glucosamine has a strong relationship with the expression level of GLUT2 and the uptake rate of glucosamine.

Then it was explored whether cancer cells uptake glucosamine through GLUT2, not GLUT1, and whether the uptake of glucosamine is an essential step for its antiproliferative effects. A549 and HepG2 cells were treated with cytochalasin B, a GLUT1-specific inhibitor, or phloretin, a GLUT2-specific inhibitor, in combination with glucosamine, respectively. Cytochalasin B and phloretin concentrations were taken from previously reported studies (36,37). As a result, phloretin recovered glucosamine-induced downregulation of IGF-1R and p-Akt expressions in both cancer cell lines, while cytochalasin B showed no recovery effects in A549 cells and partial recovery effects in HepG2 cells (Fig. 4A). In addition, glucosamine shifted the molecular mass of COX-2 as previously reported (38), which were markedly reversed by phloretin, but not by cytochalasin B in A549 cells (Fig. S1). Following these changes in molecular features, phloretin-mediated GLUT2 inhibition sufficiently rescued A549 and HepG2 cells from glucosamine-induced cell death, while the effects of cytochalasin B were not significant (Fig. 4B). Along with the results using inhibitors, GLUT2-knockdowned HepG2 cells by transfection of siGLUT2 showed recovery of IGF-1R and p-Akt expression levels decreased by glucosamine, while there were no changes in those expressions in GLUT1-knockdowned HepG2 cells (Fig. 4C). The present study also investigated whether treatment with phloretin or siGLUT2, which have an inhibitory effect on glucosamine transportation, could decrease consumption of glucosamine in both cancer cell lines and it was found that glucosamine consumption rate was reduced when GLUT2 was inhibited by phloretin and siGLUT2 (Fig. 4D). These findings demonstrate that the absorption of glucosamine via GLUT2 is the basic and necessary process for the antiproliferative effect of glucosamine.

Although the other six cell lines expressed lower levels of GLUT2 compared with HepG2 and Hep3B cells, there were differences in the antiproliferative effect of glucosamine among those six cell lines (Fig. 3B and C), suggesting that cellular sensitivity to glucosamine is influenced by not only uptake through GLUT2 but also by post-uptake processes. Glucosamine can be phosphorylated by yeast hexokinase (39,40). In addition, yeast hexokinase is similar to rat hexokinase, human hexokinase N-terminal and C-terminal in respect of its sequence and structure amino acids, especially in the active site (41). Thus, it could be hypothesized that glucosamine serves as a substrate for hexokinase in mammalian cells including those of human origin (42–44). Thus, the present study examined the basic expression levels of HK isoforms (HK1-4) in the studied eight cell lines. Most cell lines showed similar levels of HK1 expression and rarely expressed HK3. Notably, the six cell lines showed lower GLUT2 expression and reduced glucosamine consumption rates compared with HepG2 and Hep3B cells, A549 cells exhibited the lowest HK2 expression level compared with the other five cell lines (Fig. 5A). Next, since HK is the first enzyme to be activated during glycolysis and this activation is a rate-limiting step that regulates the speed of glycolysis (45), the differences in glycolytic rates and the extent of glycolysis inhibition by glucosamine was assessed. Notably, the rate of lactate production was most markedly reduced by glucosamine in A549 cells which were the most sensitive to glucosamine after liver cancer cell lines. It was observed that 3-BrPA or 2-DG, which are used as HKII inhibitors (46,47), sufficiently inhibited the proliferation of those six cell lines (Fig. 5C and D). In case of A549 cells, the anti-proliferative effect of 3-BrPA was relatively weak compared with other lung cancer cell lines, H1299 and H460, and this was probably due to the lower level of sodium-coupled monocarboxylate transporter 1 (SMCT1), which is responsible for the transportation of 3-BrPA into cells (48), than other cancer cell lines tested (Fig. 5E). Taken together, these results showed that glycolysis inhibition by glucosamine is negatively related to the level of HKs, especially HKII.

3-BrPA, 2-DG and glucosamine, which are all glycolysis inhibitors, take different pathways to enter cells and work along different mechanisms. For example, 2-DG enters cells via GLUT1 and GLUT4 like glucose and works as a competitor of glucose, which eventually lowers the efficiency of glycolysis (23), whereas glucosamine is transported into cells efficiently via GLUT2 rather than GLUT1 and competes with glucose for HK. By contrast, 3-BrPA is taken up into cells by SMCT1 and directly suppresses the activity of HKII (47,48). Therefore, it was examined whether the antiproliferative effect of glucosamine is elevated when co-treated with 2-DG or 3-BrPA. Co-treatment of 3-BrPA with glucosamine increased the anti-proliferative effects of glucosamine in both A549 and HepG2 cells (Fig. 6A), whereas 2-DG showed a slight combinational effect and no combinational effect in A549 cells and HepG2 cells, respectively (Fig. 6B). Along with these results, glucosamine-induced decreased in IGF-1R and p-Akt expression level were tended to be enhanced by co-treatment of 3-BrPA (Fig. 6C), but not by co-treatment of 2-DG (Fig. 6D). These results suggested that the mechanism of glucosamine inhibiting HK is similar to that of 2-DG and different from that of 3-BrPA and that 3-BrPA may be a suitable HK inhibitor for co-administration with glucosamine.

Glycolysis is mainly regulated by the quantity of glucose taken up into cells and the activity of HK, a rate-limiting step of the glycolysis pathway. Therefore, the expression level of GLUTs and HKs is an important factor in the glycolysis process. HIF-1α, one of the upstream transcriptional factors of GLUTs and HKs, plays an important role in controlling glycolysis (19). Based on this knowledge, it was first investigated whether the expression of GLUTs and HKs changes under hypoxic conditions in A549 and HepG2 cells and whether this is regulated by HIF-1α. As expected, the expression of GLUT1, HK1, HK2 and HK4 markedly increased under hypoxic conditions and it was confirmed that knocking down HIF-1 α markedly reduced their expression again (Fig. 7A). The expression of GLUT2 was constant in the examined conditions and the expression of HK1 was induced by hypoxic stress in HepG2 cells, but not in A549 cells. Next, the effects of hypoxia on the antiproliferative effects of glucosamine was examined. Notably, glucosamine reduced HIF-1α expression under hypoxic conditions in cancer cells (Fig. S2) and it was found that glucosamine-induced apoptosis was diminished under hypoxic conditions and this hypoxia-induced glucosamine resistance were restored by HIF-1α knockdown (Fig. 7B). Furthermore, cell cycle analysis found that the subG1 hypodiploid population, which was increased by glucosamine, was reduced by hypoxic stress and this hypoxic stress-induced change was not observed in HIF-1α depleted cells (Fig. 7C). Along with these results obtained from flow cytometry analysis, MTT assays showed that the anti-proliferation effects of glucosamine were reduced by hypoxic stress and this hypoxia-induced resistance was markedly ablated by HIF-1α downregulation in both A549 and HepG2 cells (Fig. 7D). Lastly, it was found that 3-BrPA, HKII inhibitor, also effectively reversed hypoxia-induced glucosamine resistance in both A549 and HepG2 cells (Fig. 7E). Collectively, these results suggested that glucosamine exerts its antiproliferative activity by inhibiting HKII within the cell and that the induction of HIF-1α and transactivation of its target genes, particularly HKII, are responsible for hypoxia-induced glucosamine resistance (Fig. 8).