High glucose promotes cisplatin chemoresistance in MDA‑MB‑231 breast cancer derived cells through changes in gene expression and multiple signaling pathways

Introduction

The development of cancer is promoted by age, overweight, obesity, type 2 diabetes mellitus (T2DM) and metabolic syndrome. T2DM is associated with a significant number of cancers, such as those of the breast, pancreas, liver, colorectal, renal and reproductive systems. Diabetic patients with cancer have increased mortality compared with cancerous non-diabetic individuals (1).

Hyperglycemia, transient hyperinsulinemia, chronic inflammation and oxidative stress promote the activation of signaling mechanisms that favor the initiation, development and progression of cancer during diabetes (1,2). The increased plasma levels of insulin-like growth factors, enhanced production of inflammatory cytokines (IL-6, TNFα and IL1β), low expression of certain adipokines (adiponectin), and aberrant protein glycosylation in T2DM organisms induce changes in metabolic and signaling pathways (2). These changes help cancer cells adapt to a hyperglycemic environment and promote the proliferation and acquisition of a migratory and chemoresistant phenotype (3). The increased proliferative activity of tumor cells depends on a continuous supply of biomolecules such as amino acids, carbohydrates, nucleotides and fatty acids capable of providing the energy requirements of the cell, favoring the synthesis of DNA, RNA, proteins and lipids at a higher rate in diabetic patients (4).

The malignant cells that survive primary treatment evolve into a population of resistant clones, leading to cancer progression and patient death (5). Mortality from breast cancer (BC) is higher in women with diabetes compared with women without it, since diabetes can exacerbate the onset and development of cancer by activating various cellular processes such as proliferation, migration and epithelial-mesenchymal transition (EMT) (6). Additionally, variations in the expression and activity of several drug-metabolizing enzymes, altered drug membrane transport, and evasion of apoptosis play a critical role in drug resistance (2).

It was found that a high concentration of glucose promotes an invasive/metastatic phenotype, inducing proliferation, migration, EMT and cell invasion in MDA-MB-231 cells, which was associated with increased expression of plasminogen activator urokinase-type (uPA), its receptor (uPAR), and type 1 inhibitor (PAI-1), leading to the activation of plasminogen activation system and AKT signaling (4,6). EMT is one of the mechanisms to enhance migratory and invasive properties of tumor cells, together with cell survival, invasiveness and chemotherapy drug resistance (7).

Elevated blood glucose levels regulate the chemoresistance in BC (2,3,8). Hyperglycemia and diabetes promote the sensitivity of BC cells or breast tumors to some DNA repair inhibitors (8). On the other hand, the development of resistance to cisplatin and other drugs by BC cells is frequently mediated by the prevention of apoptosis, increased EMT, DNA repair, Wnt and PI3K pathways, since cisplatin leads to apoptosis through increased reactive oxygen species, which leads to DNA damage (9). However, the genes and signaling pathways implicated in the development of chemoresistant and aggressive BC phenotype by high glucose (HG) in diabetic patients have not been fully explored.

It was aimed to detect the action of HG on the chemoresistance to cisplatin and changes in gene expression in MDA-MB-231 BC-derived cells. In addition, to analyze the genes, metabolic and signaling pathways that can mediate the promotion of cisplatin resistance and a migratory phenotype by HG.

Materials and methods

Cell culture

MDA-MB-231 BC cells were acquired from the American Type Culture Collection (ATCC) and cultured in Dulbecco's Modified Eagle's Medium/Ham's Nutrient Mixture F12 (DMEM-F12; MilliporeSigma). The medium was supplemented with 10% fetal bovine serum (FBS; Biowest) and the following antibiotics: penicillin 100 U/ml and streptomycin 100 µg/ml (Thermo Fisher Scientific, Inc.). The cells were authenticated by the 17 short tandem repeat profiling (ATCC Cell Line Authentication Service) in August 2020. The RNA extraction and microarray were processed in 2019, and the corresponding microarray series GSE136277 was made public in August 2019. Most of the experiments were concluded in 2020. Cultures were routinely tested for mycoplasma contamination using a PCR-based detection kit (cat. no. ab289834; Abcam). For propagation, cells were seeded and cultured in monolayers in 75-cm2 tissue culture flasks. For some experiments, cells cultured in 6.0-cm-diameter culture plates to 60-80% confluence were cultured for 48 h in the following conditions: i) Normal glucose (NG), 5.6 mmol/l; ii) 30 mmol/l (HG); iii) HG + 60 µmol/l cisplatin (MilliporeSigma); and iv) NG + 60 µmol/l cisplatin. The cells were detached with 0.25% trypsin-EDTA solution (Biowest).

Supplementation with ~5.6 mmol/l D-glucose approximates normal blood sugar levels, whereas concentrations ~10 mmol/l and above mimic pre-diabetic and diabetic conditions, respectively. In most studies, HG conditions use 25-30 mmol/l glucose. In previous studies, it was found that 30 mmol/l glucose produced the most evident and reproducible effects on MDA-MB-231 cells (4,6). Furthermore, interstitial glucose can be similar to blood glucose, considering that interstitial glucose concentrations in skeletal muscle of patients with T2DM during a glucose tolerance test were similar to blood glucose levels (10).

MTT assay

The MTT method is a quantitative colorimetric assay that measures the metabolic activity of cells, which is proportional to changes in the cell populations. This assay was used to determine the changes in the cell population and the median lethal concentration (LC50) of cisplatin under standard culture (17.5 mmol/l glucose), NG and HG conditions, and using an osmotic control of glucose 5.6 mmol/l plus mannitol 19.4 mmol/l (Mannitol). A total of 5,000 cells per well were seeded in 96-well plates and exposed to different concentrations of cisplatin (15-90 µmol/l). The culture medium was replaced with 200 µl serum-free DMEM and 50 µl of MTT solution (5 mg/ml) after 48 h of incubation at 37˚C. Then, the cells were incubated for an additional 4 h, the medium was removed, and 100 µl of dimethyl sulfoxide was added to dissolve the formazan crystals. The plate was read at 570 nm wavelength using a microplate reader iMark (Bio-Rad Laboratories, Inc.).

Western blotting

The cells were washed with PBS and lysed in RIPA Buffer (cat. no. 89900; Thermo Fisher Scientific, Inc.), containing a protease inhibitor cocktail (cat. no. I3786; MilliporeSigma) to prevent protein degradation. Proteins were quantified by the Bradford assay, and the cell lysates were stored at -70˚C. Equal amounts of protein (100 µg per well) were mixed with Laemmli sample buffer and separated by 10% SDS-polyacrylamide gel electrophoresis. The transfer was realized in a PVDF membrane (0.8 mm/cm2, 2 h), using Tris buffer 25 mmol/l, glycine 192 mmol/l and methanol 20%. Membranes were then blocked with 5% BSA in TBS for 1 h at room temperature to prevent non-specific binding. Subsequently, membranes were incubated overnight at 4˚C with primary anti-caspase-3 antibody (1:500; cat. no. sc-7272; Santa Cruz Biotechnology, Inc.) in TBS-T (0.1% Tween-20) containing 1% BSA. Antibody binding was detected using an HRP-conjugated secondary antibody (1:5,000; cat. no. 7076S; Cell Signaling Technology, Inc.), incubated for 1 h at room temperature, followed by visualization with an enhanced chemiluminescence kit (Clarity Western ECL substrate, Bio-Rad Laboratories, Inc.), according to the manufacturer's instructions. The images were obtained using the C-Digit Blot Scanner (LI-COR Biosciences). The bands were quantified by densitometry analysis using the Image Studio 4.0 software (LI-COR Biosciences).

RNA extraction

Total RNA was extracted using TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. RNA quality was verified by 2% agarose gel electrophoresis, followed by ethidium bromide staining and visualization under UV light and quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Inc.).

Hybridization and microarray analysis

A total of 10 µg of total RNA obtained from cells cultured in NG and HG media was used for cDNA synthesis and labeling with the Superscript II kit (Invitrogen; Thermo Fisher Scientific, Inc.), incorporating dUTP-Cy5 for HG samples and dUTP-Cy3 for NG (control) samples. The absorbance was measured at 655 nm for Cy5 and 555 nm for Cy3. Fluorophore-labeled cDNAs were hybridized in similar quantities to the Human 10K 50-mer oligonucleotide collection (MWG Biotech Oligo Bio Sets).

Microarray images were acquired and quantified using a ScanArray 4000 scanner with QuantArray software (Packard BioChips; PerkinElmer, Inc.), and data were analyzed using Array-Pro Analyzer software (Media Cybernetics, Inc.). Background correction, normalization, intensity filtering, replicate analysis, and the selection of differentially expressed genes were performed using genArise, a free program developed by the Computing Unit of the Cellular Physiology Institute, National Autonomous University of Mexico (http://www.ifc.unam.mx/genarise/). The data generated in the present study were deposited in Gene Expression Omnibus (GEO) platform under accession number GSE136277 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE136277). Differentially expressed genes were identified using the following statistical cutoffs: Z-score ≥2 or ≤-2 for biological significance, and a false discovery rate (FDR) ≤0.05 to account for multiple testing. These thresholds were applied from the outset of the analysis to ensure the robustness of the candidate genes included in subsequent evaluations.

Reverese transcription-quantitative PCR (RT-qPCR)

The microarray analysis was validated by RT-qPCR for frizzled 3 (FZD3) and tetraspanin 1 (TSPAN1). Total RNA (1 µg) was reverse-transcribed using the SuperScript™ III First-Strand Synthesis System (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol (50˚C for 60 min, followed by enzyme inactivation at 70˚C for 15 min). Alternatively, when individual components were used, reverse transcription was performed in a 20 µl reaction mixture containing 200 U M-MLV reverse transcriptase, 1X first-strand buffer, 0.5 mM dNTP mix (Thermo Fisher Scientific, Inc.), 25 µg/ml oligo(dT) primers (Invitrogen; Thermo Fisher Scientific, Inc.), and 40 U RNase inhibitor, incubated at 37˚C for 60 min with enzyme inactivation at 95˚C for 15 min. The resulting cDNA was amplified using the Maxima SYBR Green/ROX qPCR Master Mix (Fermentas; Thermo Fisher Scientific, Inc.) on a StepOnePlus™ Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.).

The resulting cDNA was amplified using the Maxima SYBR Green/ROX qPCR Master Mix (Fermentas; Thermo Fisher Scientific, Inc.) in a StepOnePlus™ Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.) with the following thermocycling conditions: Initial activation at 95˚C for 15 min; 40 cycles of 95˚C for 15 sec and annealing at the primer-specific melting temperature for 1 min; followed by a final extension at 72˚C for 10 min. Expression of the selected genes was normalized to hypoxanthine-guanine phosphoribosyltransferase (HPRT). The primer sequences used were as follows: FZD3 forward, 5'-CATGGAGATGTTTGGTGTTCCTT-3' and reverse, 5'-AAGTCGAGGATATGGCTCATCAC-3'; TSPAN1 forward, 5'-TGGGCTGCTATGGTGCTA-3' and reverse, 5'-TGCAGGTTTCATTGGCTGT-3'; and HPRT forward, 5'-CATTATGCTGAGGATTTGGAAAGG-3' and reverse, 5'-CTTGAGCACACAGAGGGCTACA-3'. Relative expression levels were calculated using the 2-ΔΔCq method (11).

HPRT was used as a housekeeping gene because its expression has been considered moderately stable in BC cell lines, including MDA-MB-231 cells (12). Under hyperglycemic conditions, Liu et al (13) identified the HPRT gene as one of the more reliable reference genes in retinal pigment epithelial cells, and even HPRT expression was more stable than ACTB and GAPDH genes. Additionally, no changes in the expression of this gene under HG were found in our microarray study.

Immunohistochemical (IHC) expression of TSPAN-1 and FZD3 in BC tissue

IHC expression of TSPAN1 and FZD3 in BC tissue. To investigate the protein expression patterns of TSPAN1 and FZD3, representative IHC images of ductal breast carcinoma were analyzed from 50 samples available in the Human Protein Atlas database (https://www.proteinatlas.org/), corresponding to women aged 30-37 years. Adjacent normal breast tissue and adenoma samples were used as controls.

Gene analysis and Gene Ontology (GO)

The analysis of genes differentially regulated by HG and their associated functions, including GO categories (cellular components, molecular functions and biological processes), as well as biological pathways, was performed using Gene Set Enrichment Analysis (GSEA). For this purpose, the WEB-based GEne SeT AnaLysis Toolkit (WebGestalt; http://www.webgestalt.org/) and WikiPathways Cancer were employed. Gene sets were considered significant at an FDR ≤0.05. In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (https://www.genome.jp/kegg/pathway.html) was used to determine the interactions of differentially expressed genes within metabolic and signaling pathways.

Functional association network and selection of genes involved in chemoresistance

All the overexpressed and downregulated genes associated with cancer were uploaded into the program STRING (https://string-db.org/), which predicts the protein-protein interaction (PPI) network. The confidence score for displaying interaction was 0.9. The formatted PPI networks were uploaded and visualized with the Cytoscape software (version 3.10.1) (14). Due to the complexity of this network, the CytoHubba plugin was used to select the 50 nodes with the highest degree score, and their interaction network was constructed (15). Furthermore, the overexpressed genes with at least 2-fold change and their involvement in chemoresistance, according to scientific studies (16-19), were selected, and their possible participation in HG-dependent chemoresistance was considered.

Kaplan-Meier (KM) plot analysis

Overall survival (OS) analysis was performed for the key genes YBX3 (also known as CSDA), amphiregulin (AREG), amyloid precursor protein (APP) and N-cadherin (CDH2) using the KM Plotter database (20); (https://kmplot.com) to assess the prognostic value of the genes associated with cisplatin resistance in BC. A cohort of 153 patients with TNBC who were split into high and low expression groups was selected, considering the median expression of the gene. OS plots were created using data from BC samples negative for estrogen, progesterone and HER2 receptors, as detected by IHC and microarray. The hazard ratio (HR) with 95% confidence intervals (CIs) and log-rank P-values were calculated.

Coexpression analysis

To evaluate whether the selected genes are coexpressed in a cohort of patients with BC and the potential molecular and regulatory interactions between proteins codified by them, the coexpression of AREG, FZD3, APP and CDH2 with YBX3 was performed using R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl) on TCGA Breast Invasive Carcinoma (BRCA) datasets (1,097 samples) with R2 internal identifier ps_avgpres_tcgabrca1097_tcgars. Pearson correlation coefficients and associated P-values were calculated, with scatter plots and linear regression used for visualization.

Statistical analysis

All experiments were performed in at least three independent replicates. Data are expressed as the mean ± SD. One-way ANOVA followed by Tukey's multiple comparison test was used for MTT, western blotting and RT-qPCR analyses, and performed using GraphPad Prism 8.02 software (Dotmatics). P≤0.05 was considered to indicate a statistically significant difference. Microarray and GSEA analyses were considered significant with fold-change ≥1.5 and FDR≤0.05, respectively. Co-expression analyses of AREG, FZD3, APP and CDH2 with YBX3 were performed using the R2 Genomics platform https://hgserver1.amc.nl/cgi-bin/r2/main.cgi, and survival analyses were conducted using the KM Plotter, with HRs, 95% CIs and log-rank P-values. P≤0.05 was considered to indicate a statistically significant difference.

Results

Previously, it was found that HG increases cell population growth, migration, invasion and EMT in MDA-MB-231 cells (4). In the present study, the effect of HG was analyzed on cisplatin resistance, cell death and changes in gene expression. Additionally, it was analyzed how the change in gene expression is associated with a chemoresistant and aggressive phenotype.

Hyperglycemia-induced resistance to cisplatin

To investigate whether HG induces drug resistance, the LC50 of cisplatin for MDA-MB-231 cells was determined under standard culture conditions (glucose 17.5 mmol/l), NG, NG with mannitol (osmotic control) and HG. The osmotic control group showed no difference compared with the NG group. Cell viability progressively decreased as cisplatin concentration increased from 15 to 90 µmol/l in a concentration-dependent manner, with LC50 values of 60, 15 and 75 µmol/l for standard, NG and HG conditions, respectively. These findings indicated that cells cultured under HG and standard conditions were more resistant to cisplatin compared with those cultured under NG. Thus, in the present study, cisplatin resistance was directly proportional to glucose concentration (Fig. 1A and B).

Figure 1 HG attenuates the effect of cisplatin on the cell population and prevents apoptosis. (A and B) MDA-MB-231 cells were treated for 48 h with CCP. The growth of the cell population was evaluated by the MTT assay in standard, (NG), HG, and mannitol osmotic control (NG Mannitol) culture conditions. *P≤0.01 and **P≤0.001 vs. the respective NG for each cisplatin concentration. No significant differences were observed between the NG and mannitol-treated groups. (C-E) Additionally, the activation of caspase 3 was evaluated by western blotting and densitometry in lysates of cells cultured for 48 h, in NG and HG conditions, and treated with cisplatin (NG/CCP and HG/CCP). The intensities of the bands were calculated after background subtraction, and the sum of the bands (total caspase 3) and active caspase 3 were normalized to (D) β-actin and (E) procaspase 3, respectively. Data are presented as the mean ± SD of three independent experiments. *P≤0.05 and ***P≤0.001 vs. NG, and + P≤0.001 vs. HG. HG, high glucose; NG, normal glucose; CCP, cisplatin.

Hyperglycemia prevents the apoptosis induced by cisplatin

Apoptosis induced by cisplatin under normoglycemic and hyperglycemic conditions was evaluated in MDA-MB-231 cells by western blot analysis of caspase-3 activation. The caspase 3 antibody recognizes two band, one of higher molecular weight (MW) corresponding with the procapase 3, and other of lower MW corresponding to the cleaved/active caspase 3 (Fig. 1C). Both NG and HG groups exhibited similar levels of total caspase 3 independently of the cisplatin treatment (Fig. 1D). Whereas, cisplatin induced a significant increase in the proportion of active caspase 3 in both NG (P≤0.001) and HG (P≤0.001) conditions (Fig. 1E). Notably, the NG and NG/cisplatin groups displayed higher levels of active caspase-3 than the corresponding HG and HG/cisplatin groups (Fig. 1E). These results indicated that HG attenuated cisplatin-induced apoptosis, thereby promoting cisplatin chemoresistance.

Overall changes in gene expression

Considering an expression change equal to or greater than 1.5-fold, 1,099 genes were differentially expressed. Among these, 457 were overexpressed, while 642 were downregulated under the HG condition. The top 15 overexpressed and downregulated genes in the HG microenvironment are presented in Table I. All differentially expressed genes are listed in Tables SI and SII.

Table I Top 15 upregulated and downregulated genes in a hyperglycemic environment.

Table I

Top 15 upregulated and downregulated genes in a hyperglycemic environment.

Upregulated
ID	Genes	Z score	Full name
NM_032741	AGPAT1	4.152511	1-Acylglycerol-3-phosphate O-acyltransferase 1
NM_003043	SLC6A6	4.058222	Taurine transporter
NM_005727	TSPAN-1	3.811546	Tetraspanin 1
NM_004189	SOX14	3.648091	SRY-box 14
NM_021625	TRPV4	3.609894	Transient receptor potential cation channel subfamily V member 4
NM_016176	SDF4	3.445534	Stromal cell derived factor 4
NM_006213	PHKG1	3.320789	Phosphorylase kinase catalytic subunit gamma 1
NM_020649	CBX8	3.251316	Chromobox 8
NM_001252	TNFSF7	3.129304	CD70 molecule
NM_005919	MEF2B	3.041269	BORCS8-MEF2B readthrough
NM_002600	PDE4B	3.035975	Phosphodiesterase 4B
NM_002150	HPD	2.983852	4-Hydroxyphenylpyruvate dioxygenase
NM_014610	GANAB	2.950495	Glucosidase II alpha subunit
NM_017412	FZD3	2.854118	Frizzled class receptor 3
Downregulated
NM_006544	SEC10L1	-4.523611	Exocyst complex component 5
NM_012261	LAMP5	-4.45501	Lysosomal associated membrane protein family member 5
NM_012070	ATRN	-4.333672	Attractin
NM_004185	WNT2B	-3.908675	Wnt family member 2B
NM_001697	ATP5O	-3.87772	ATP synthase, H+ transporting, mitochondrial F1 complex, O subunit
NM_053004	FLJ21125	-3.814458	G protein subunit beta 1 like
NM_003692	TMEFF1	-3.744839	Transmembrane protein with EGF like and two follistatin like domains 1
NM_006866	LILRB1	-3.717488	Leukocyte immunoglobulin like receptor A2
NM_003072	SMARCA4	-3.709549	SWI/SNF related, matrix associated, actin-dependent regulator of chromatin, subfamily A, member 4
NM_021176	G6PC2	-3.584369	Glucose-6-phosphatase catalytic subunit 2
NM_004937	CTNS	-3.50955	Cystinosin, lysosomal cystine transporter
NM_004208	PDCD8	-3.506222	Apoptosis-inducing factor, mitochondrion-associated, 1
NM_013335	GMPPA	-3.505725	GDP-mannose pyrophosphorylase A
NM_017882	CLN6	-3.495366	Ceroid-lipofuscinosis, neuronal 6, late infantile, variant
NM_022444	SLC13A1	-3.435806	Solute carrier family 13 (sodium/sulfate symporters), member 1

GSEA

Microarray data were subjected to functional enrichment analysis using WEB-Gestalt and the WikiPathways database. In the cellular component analysis, an activation of genes that encode protein constituents of the ribosomes, plasma membrane, cell junctions and endoplasmic reticulum was found (Fig. 2A). This was also reflected in the molecular function analysis, where changes in structural constituents of ribosomes, sodium symporter activity and ion transmembrane transport were identified. In addition, changes in hydrolase activity, calcium ion binding, kinase binding and protein-containing complex binding were observed (Fig. 2B). According to the enrichment analysis, the following signaling pathways were enriched in a hyperglycemic microenvironment: Wnt/β-catenin, angiogenesis, TP53 network, NF-κB survival signaling, apoptosis, TGF-beta receptor signaling and ATM-dependent DNA damage response (Fig. 2C).

Figure 2 Gene enrichment analysis. (A and B) Gene Ontology terms of the significantly upregulated and downregulated genes for (A) the cellular components and (B) molecular function categories. (C) Wikipathways gene enrichment analysis in a hyperglycemia microenvironment.

Gene validation

The expression of TSPAN1 and FZD3 was evaluated by RT-qPCR to validate the microarray analysis. Both genes are associated with an aggressive BC phenotype and showed some of the highest Z-scores (3.81 for TSPAN1 and 2.85 for FZD3). In particular, FZD3, a protein of the Wnt/β-catenin pathway, exhibited the highest Z-score. RT-qPCR analysis confirmed that FZD3 expression under HG conditions was 3.81-fold higher compared with NG (Fig. 3). Similarly, TSPAN1 expression under HG conditions was 2.26-fold higher than under NG conditions, consistent with the microarray results (Fig. 3). These findings are also supported by data from the Human Protein Atlas, which shows higher expression of TSPAN1 and FZD3 proteins in BC tissues compared with healthy mammary tissue (Fig. 3).

Figure 3 Expression of FZD3 and TSPAN1 in breast cancer by IHC and RT-qPCR. (A) Expression of FZD3 and TSPAN1 was analyzed by IHC using images obtained from the Human Protein Atlas. Brown staining indicates the presence of FZD3 and TSPAN1, showing higher expression in ductal carcinoma compared with control groups, including adenoma and normal breast tissue. (B) MDA-MB-231 cells were cultured in NG and HG. Total RNA was isolated 48 h after treatment, and mRNA was evaluated by RT-qPCR. These results are consistent with the microarray data. Media ± standard deviation of three experiments with two replicates each. ***P≤0.001 vs. NG. IHC, immunochemistry; RT-qPCR, reverse transcription-quantitative PCR; HG, high glucose; NG, normal glucose.

Hyperglycemia alters genes associated with metabolism

According to the aforementioned analysis with the KEGG pathways program, it was found that a high number of genes of transmembrane transporters and ribosomal proteins were differentially expressed in HG conditions (Table II). Additionally, several genes associated with other metabolic processes were found, including: i) Synthesis of DNA, RNA and proteins; and ii) purine, pyrimidine and lipid metabolism. Noteworthy, the upregulation of some genes of electron chain transport or ECT (COX41, COX61, COX7B and NDUFB3) indicated an increase in the electron flux through this chain. However, subunits ATP5O and ATP5D of the ATP synthase were downregulated, suggesting the uncoupling between the ECT and ATP synthesis.

Table II Kyoto Encyclopedia of Genes and Genomes pathways associated with genes regulated by hyperglycemia.

Table II

Kyoto Encyclopedia of Genes and Genomes pathways associated with genes regulated by hyperglycemia.

Metabolic or signaling pathway	Total genes	Genes upregulated	Genes downregulated
Transmembrane transport	27	SLC6A6, SLC44A4, SLC35F2, SLC16A8, SLC13A2, SLC6A20, SLC23A2, SLC25A19, SLC39A7, SLC22A18AS, SLC25A5	SLC13A1, SLC22A7, SLC26A3, SLC7A7, SLC2A8, SLC35B3, SLC6A12, SLC1A7, SLC39A6, SLC39A4, SLC35C2, SLC38A3, SLC16A10, SLC44A2, SLC25A17, SLC28A2
Cytokine-cytokine receptor interaction	24	CXCL1, CXCR3, PDGFRA, IL9R, IL21R, IL1RA, EPOR, IL17RA, IL17RB, TNFSF7, CCL20, IL32	TNIP2, CCR9, CXCR2, CXCR4, CXCR6, CCRL1, TGFBR2, TGFBR1, CXCL6, IFNA8, IL4, IL17C
Regulation of cytoskeleton	24	CHRM4, FGFR1, PDGFRA, VAV3, CFL, ITGB4, RASAL-1, RIN1, CDC42BPA, RAB9A, RHOB	PIK3C2B, PIK3R5, CHRM1, FGF20, FGFR2, ACTN3, VCL, VAV2, ILK, ARHGDIB, RDX, RRAS, RASSF7
Ribosomal proteins (extra and intramitochondrial)	20	RPS13, RPS17, RPS21, RPS3A, RPS8, RPS7, RPL23A, RPL13A, RPL15, RPL9, MRPS18B, MRPS7, MRPL15, MRLP17, MRPL49	MRPS31, MRPL11, RPL23, MRPL33, MRPL42
Purine, pyrimidine metabolism/nucleic acid synthesis	22	XDH, POLA2, POLG, GUCY2F, PDE1B, PAPSS2, PDE2A, UPB1, PDE4B, ADCY7, NME1, ECGF, NOC3L, ATIC	CDA, NT5E, AK3, AMPD2, AMPD3, ADA2, MTAP, DNPH1
MAPK	20	FGFR1, PDGFRA, MKNK, DUSP4, RIN1, MAPK8IP1, RASGRF1	PPP3C, FGF20, TGFBR2, MAP2K3, MAPKAPK5, MEF2C, BLK, DUSP5, TGFBR1, FGFR2, RRAS, RAPGEF2, RASSF7
Ras	19	FGFR1, PDGFRA, TEK, RIN1, RASGRF1, SPEC1, RASAL-1, RIN1	PIK3C2B, PIK3R5, FGF20, GNB1, FGFR2, ANGPT1, RRAS, RALA, RAPGEF2, RASSF7, PLAAT1
Apoptosis	19	TRAF1, CASP10, XIAP, BCL2L2, BIRC2, RHOB, PERP	PTPN13, PIK3C2B, PIK3R5, PDCD8, PDCD5, BIRC3, BIK, DAPK1, CCAR2, BNIP1, TNIP2, RASSF7
Wnt β-catenin	17	FZD3, FZD4, LPR8, DVL2, WNT1, LEF1, DKKL1, RHOB, PPP2R3B	WNT2B, TGFBR2, BMP2, TGFBR1, ID1, PPP3CA, TBL1, CAMK2B
Focal adhesion	17	PDGFRA, XIAP, VAV3, LAMC2, ITGB4, PPP2R5B, PPP2R5C, RASGRF1	PIK3C2B, PIK3R5, THBS2S, ACTN3, VCL, BLK, VAV2, BIRC2, ILK
Glycerophospholipid metabolism	16	AGPAT1, MBOAT7, CEPT1, GLA, PNLIPRP2, ACHE, LYPLA1, PLA2G2A	AGPAT4, CDIPT, CHPT1, PCYT1B, PLA2G2D, TAZ, PAFAH1B3, PLAAT1
Jak-STAT	14	PIM1, PIAS1, IL9R, IL21R, EPOR, PTPN6	PIK3C2B, PIK3R5, BLK, SOCS5, IFNA8, IL4, IL17C, GFAP
Oxidative phosphorylation	10	COX4I1, COX6A1, COX7B, NDUFB3	ATP5O, ATP5D, COX7A1, NDUFA13, NDUFB4, E2IG2

Hyperglycemia alters genes associated with apoptosis and cell survival

On the other hand, it was found that HG induces increased expression of some genes that promote cell survival (TNFSF7 and TRAF1) (21) or prevent apoptosis (XIAP and BCL2L2) (22,23). Additionally, HG leads to low expression of proapoptotic factors (BIK, PDCD8, PDCD5, BNIP1, DAPK1 and CCAR2) (24-26). These data indicated that HG prevents BC cell death and promotes cell survival under stress conditions, such as the presence of chemotherapeutic agents, such as cisplatin (Fig. 1).

Signal pathways associated with the adaptation to hyperglycemia

The KEGG pathways program indicates that Wnt-β-catenin, Jak/STAT and Ras-MAPK are involved in the action of HG, together with processes such as cytokine-cytokine receptor interaction, regulation of actin cytoskeleton and focal adhesion (Table II). These results and the PPI analysis suggested that the Wnt-β-catenin signaling has a significant role in the action of HG.

Selection of hub genes and PPI

In the PPI network of overexpressed genes constructed by Cytoscape, a total of 303 nodes and 834 edges were arranged (data not shown); further analysis by cytoHubba based on this network revealed a network with 50 principal nodes (Fig. 4). Between the highest degree score hubs, genes for ribosomal proteins were found, such as RPL13a, RPL23a, RPS3a1, RPS13, RPS8, RPS21, RPL6, RPL15, RPS7, MRPS7 and RPL9 with a high degree of interaction shown in red (Fig. 3). Interacting with these nodes genes coding for transcription and protein synthesis (PABPC1, EIF4G2, EIF5B, EEF1G and HNRNPA1), potassium channels (KCNG2, KCNG1 and KCNS1) and metabolism (COX4I1, COX6A1 and NDUFB3) were found. Other important hubs were associated with the Wnt/canonical pathway, such as WNT-1, DVL2 and LEF1. Other proteins related to cancer progression were also represented (APP, AREG and CDH2). It is noteworthy that ribosomal proteins were overrepresented and had the highest degree scores (red and orange in Fig. 4). Because of this, nodes of important proteins had lower degree scores (orange and yellow in Fig. 4). The PPI network according to the enrichment analysis indicates the relevance of Wnt/β-catenin (Fig. 2C).

Figure 4 Major 50 hub genes interaction. These genes were selected using the String network of protein-protein interaction, considering those overexpressed with a fold change >1.5, followed by hub selection using Cytohubba. The nodes with major degree scores are shown in a decreasing scale from red to yellow. Data show gene hubs such as BIRC2, DVL-2, WNT1 and ATM, which have regulatory functions associated with the WNT/β-catenin signaling pathways, apoptosis, and ATM-dependent DNA damage repair.

Prognostic value of Hub genes YBX3, AREG, APP and CDH2

Considering the hub genes with the highest number of interactions obtained from the CytoHubba analysis and their prognostic evaluation using the KM-plot platform, YBX3 or CSDA (HR=2.04; 95% CI, 1.01-4.01; P=0.041), AREG (HR=2.57; 95% CI, 1.26-5.24; P=0.0073), APP (HR=1.73; 95% CI, 0.87-3.44; P=0.11) and CDH2 (HR=1.74; 95% CI, 0.88-3.46; P=0.11) were associated with poor prognosis in diabetic patients with BC (Fig. 5). These data demonstrate that high expression of YBX3 and AREG is significantly associated with reduced survival probability, whereas APP and CDH2 did not reach statistical significance. Nevertheless, elevated YBX3 and AREG levels may serve as potential biomarkers for identifying diabetic patients with BC with a worse prognosis.

Figure 5 Prognostic biomarkers in hyperglycemia. (A-D) KM survival curves were performed using the KM-plotter platform, considering the mRNA expression levels of (A) CSDA (YBX3), (B) AREG, (C) APP and (D) CDH2 for patients with breast cancer, negative to estrogen, progesterone, and HER2 receptors for IHC and microarrays. KM, Kaplan-Meier; HR, hazard ratio.

Coexpression of hub genes YBX3, AREG, APP and CDH2

The coexpression analysis was performed in a cohort of 1,097 patients diagnosed with breast invasive carcinoma, aged 26 to 90 years, with heterogeneous clinical characteristics (Fig. 6), including estrogen receptor status (ER+ 71.7%, ER- 22%), progesterone receptor status (PR+ 63.7%, PR- 31.4%), HER2 status (HER2+ 14%, HER2- 51.3%) and metastatic condition (M0 82.6%, MX 14.8% and M1 2%). The analysis revealed that APP and AREG expression were significantly correlated with YBX3 (R=0.373 and R=0.119, respectively), indicating that higher YBX3 expression is associated with increased APP and AREG expression. A positive but weak correlation was observed between YBX3 and FZD3 (R=0.139). Importantly, CDH2 and YBX3 also exhibited a weak but statistically significant correlation (R=0.131; P=1.29x10-5).

Figure 6 Correlation analysis of gene hubs. (A-D) Correlation analysis of gene expression was performed using the R2 Genomics platform (http://r2.amc.nl) in a cohort of 1,097 patients diagnosed with breast invasive carcinoma (TCGA), aged 26 to 90 years, with heterogeneous hormone receptor status (ER+ 71.7%, ER- 22%, PR+ 63.7%, PR- 31.4%, HER2+ 14%, and HER2- 51.3%), and metastatic condition (M0 82.6%, MX 14.8% and M1 2%). Scatter plots show the relationships between (A) APP and YBX3, (B) YBX3 and FZD3, (C) AREG and YBX3, and (D) CDH2 and YBX3. Each plot includes a regression line, the Pearson correlation coefficient (R), and the statistical significance (P-value). TCGA, The Cancer Genome Atlas.

Discussion

Hyperglycemia is one of the most important factors considered a promoter of cancer in T2DM. In the present study, it was detected that HG induces cisplatin resistance and changes in gene expression in MDA-MB-231 cells. These changes indicate a variety of alterations in metabolic and signaling pathways that may be related to preventing cell death and increased chemoresistance associated with an aggressive phenotype. Additionally, some genes (AREG and YBX3) were selected as possible biomarkers of chemoresistance and prognostic indicators. Furthermore, it was confirmed that hyperglycemia increases the migratory phenotype in these cells.

HG induces resistance to cisplatin. The prometastatic action of HG on MDA-MB-231 cells has been demonstrated. In our previous studies, HG promoted the growth of the cell population, EMT, clonogenicity, cell migration and cell invasion compared with NG and the control of osmolality (4,6). It was also described that T2DM promoted the progression of BC in mice (27). In the present study, the lower IC50 for cisplatin in cell cultures with HG indicated the induction of cisplatin resistance by HG, corroborating that HG is a promoter of chemoresistance (2,3,8). Chemoresistance is associated with an invasive phenotype, and both are dependent on the prevention of apoptosis and the induction of EMT, cancer stemness, oncogenes, tumor suppressors, and metabolic and cell signaling reprogramming, involving transporter pumps, mitochondrial alterations, DNA repair and autophagy (28).

Prevention of cisplatin-induced apoptosis by HG. To confirm if the prevention of the cytotoxic effect of cisplatin by HG is due to the prevention of cell death, caspase-3 was assessed. Caspase-3 is involved in the final stages of apoptosis, participating in the amplification of intracellular apoptotic signals as a consequence of the caspase cascade induced by both intrinsic and extrinsic signaling (26). In the present study, it was found that HG prevents caspase-3 activation and, therefore, apoptosis, which explains the increased chemoresistance to cisplatin (Fig. 1). HG induces apoptosis in normal cells; however, in cancer cells, HG prevents a variety of proapoptotic mechanisms and promotes their proliferation and invasiveness (4,6,28). Additionally, the microarray analysis indicated that HG prevents the cytotoxic action of cisplatin through the increased expression of genes related to cell survival. (TNFSF7 and TRAF1) (Table II) (21), and inhibition of apoptosis (XIAP and BCL2L2) (Table II) (22,23), together with diminished expression of proapoptotic factors (BIK, PDCD8, PDCD5, BNIP1, DAPK1 and CCAR2) (24,25). The overexpression of BCL2L2 has been associated with an aggressive and radioresistant BC phenotype (22).

Genes involved in metabolism are associated with chemoresistance. Diabetes and hyperglycemia are frequently associated with metabolic disorders. The data suggest that cancer metabolism is intimately linked to drug resistance. The enrichment analysis indicates the differential expression of numerous genes for components of membranes and ribosomes in HG conditions. The first is related to metabolite and ion transporters, and suggests the reprogramming of solute interchange (ions and nutrients) through the cell membranes. The second could be associated with the increased synthesis of proteins.

The reprogramming of membrane transport is suggested because of changes in the expression of 27 solute carrier transporters (SLC) and 10 ATP-dependent transporters in the HG condition (Tables II, SI and SII). The overexpression of SLC that controls the entry of essential amino acids, monocarboxylates, vitamins, glucose, nucleotides and organic cations indicates an increased flux of nutrients, favoring the biosynthetic and energetic demands of cancer cells. The SLC may be novel therapeutic targets and predictive markers of chemoresistance. Among the transporter-coding genes upregulated in the HG condition, SLC6A6 (4.05 fold) and SLC39A7 (1.57-fold) have been related to multidrug and cisplatin resistance in colorectal and non-small cell lung cancers (16,29).

The dysregulation of genes coding for ribosomal proteins promotes BC metastasis. Among them, RPS13 was upregulated more than 2-fold during HG treatment and promoted drug resistance in gastric cancer cells, suppressing apoptosis (17).

Recently, changes in lipid metabolism have been recognized as important mediators of drug resistance. The gene coding for 1-acylglycerol-3-phosphate O-acyltransferase 1 (AGPAT1) was the most highly overexpressed in HG. It is highly expressed in breast tumor tissues, and it has been proposed as a component of a gene signature for prognosis in BC (30). More studies are required to determine its role in chemoresistance under HG conditions; some members of the AGPAT gene family are upregulated in multiple cancers and are associated with drug resistance and poor prognosis (31).

Crosstalk between different oncogenic signaling pathways in hyperglycemia. The metabolic alterations and changes in gene expression induced by HG suggest the implication of multiple signaling pathways that favor an aggressive phenotype and chemoresistance in MDA-MB-231 cells. RT-qPCR confirmed the overexpression of TSPAN1 and FZD3 genes, and several studies suggested their role in cancer progression and chemoresistance. TSPAN-1 promotes the proliferation and development of a migratory-invasive phenotype through the induction of EMT and activation of the PI3K/Akt signaling (32). In addition, the suppression of TSPAN1 expression prevented BC tumor growth in mice (18). These data are in agreement with the high expression of TSPAN1 and FZD3 proteins in breast ductal carcinoma in women of reproductive age (Fig. 3).

The finding that FZD3 and other members of the Wnt canonical signaling [WNT1, FZD-4, disheveled 2 (DVL2), lymphoid enhancer factor 1 (LEF1) and LRP8] were overexpressed in a hyperglycemic microenvironment indicates the participation of Wnt signaling in the promotion of an aggressive and chemoresistant phenotype by hyperglycemia, which was supported by the enrichment and the PPI analyses. The altered expression of genes of the Wnt pathway has been related to cancer progression and poor prognosis (33,34). WNT1 can induce transformation of the mouse mammalian cell line C57MG (34). The expression of WNT1 and LEF1 is upregulated in human breast carcinoma (35). Whereas the upregulation of FZD4 in glioma cells is correlated with resistance to drug-induced cell death, partly due to the activation of Wnt/β-catenin and downregulation of proapoptotic genes (CASP3 and PDCD4) (36). The silencing of LRP8 by interference RNA attenuates Wnt/β-catenin, decreases the metastatic potential, and sensitizes BC cells to chemotherapy (37). As a consequence of the activation of the Wnt pathway, LEF-1 and T cell factor (TCF) form a complex and activate the expression of genes such as MDR1, C-MYC, MET, MMP7, C-JUN and CCND1, contributing to carcinogenesis and tumor progression in several cancers (33,35). In MCF7 BC cells, LEF1 induces resistance to docetaxel (DTX), and the inhibition of LEF1 with quercetin reestablishes sensitivity to DTX (38). LEF1 is a transcription factor that controls gene expression dependent or independent of Wnt/β-catenin signaling and increases the expression of EMT effectors, such as snail, vimentin and N-cadherin; the gene that encodes N-cadherin, CDH2, was also found to be overexpressed in the HG condition (Table SI). The potential role of the Wnt pathway in the promotion of BC under hyperglycemic conditions requires further studies.

The non-canonical signaling of Wnt can also be implicated in the promotion of BC by hyperglycemia. Two genes that encode CDC42 effectors were overexpressed by 2-fold in HG conditions, including CDC42 binding protein kinase alpha and CDC42 small effector 1. The binding of WNT1 with FZD-3 induces the activation of CDC42 in a β-catenin independent signaling, and CDC42 induces the reorganization of the cytoskeleton and cell migration (39).

As the PPI indicates, the ATM-dependent DNA-repairing pathway is of special interest, considering that platinum salts are DNA-damaging agents. ATM improves the ability to repair double-strand DNA breaks and can lead to increased resistance to DNA-damaging chemotherapeutic agents, such as platinum salts (40).

Potential genes associated with chemoresistance induced by hyperglycemia

Through the evaluation of hub genes with a greater number of interactions (Fig. 4), the survival of TNBC patients using the KM plotter (Fig. 5) and coexpression analysis (Fig. 6), the authors propose YBX3 as a central regulator positively correlated with AREG, APP and FZD3, suggesting that these genes may participate in a signaling network that promotes proliferation, invasion and therapy resistance under hyperglycemic conditions. In particular, YBX3 and AREG showed significant prognostic value, as their high expression was associated with reduced survival in patients with BC, highlighting their potential as biomarkers of poor outcome (Fig. 5).

By contrast, APP and CDH2 did not reach statistical significance in survival analyses, but coexpression studies demonstrated that both genes are positively correlated with YBX3 (Fig. 6). This finding suggests that APP and CDH2 may not act as independent prognostic markers, but rather as cooperative elements within YBX3-driven regulatory networks, thereby contributing indirectly to tumor progression and aggressiveness. Importantly, recent studies have shown that YBX3 is overexpressed in multiple cancer types and contributes to chemoresistance by regulating transcriptional and translational programs involved in EMT, proliferation, and survival pathways (19,41,42). Moreover, YBX3 regulates the expression of growth factors and ligands such as AREG, creating a positive feedback loop that amplifies growth factor signaling, enhances EMT, and promotes chemoresistance (43).

AREG is the primary EGFR ligand in MDA-MB-231 cells, and its binding to EGFR promotes BC growth and metastasis (44). APP overexpression drives EMT, proliferation and migration/invasion in BC cells, and it has been proposed as a risk factor for lymph node metastasis and poor prognosis in non-luminal BCs, including TNBC (45). CDH2 is a mesenchymal marker induced by EMT that facilitates migration and invasion, and its high expression correlates with poor prognosis in invasive micropapillary BC (46,47).

Taken together, these findings suggest that while AREG and CDH2 emerge as potential prognostic biomarkers, APP and CDH2 may act primarily as modulators of YBX3-driven transcriptional networks that support tumor progression rather than serving as independent prognostic indicators. Nevertheless, despite these associations, the current evidence remains insufficient to establish these hub genes as reliable biomarkers, underscoring the need for further validation in larger patient cohorts and in-depth functional studies.

It must be considered that an increase in the mRNA level of a gene is not always followed by an increase in the codified protein, and the protein generally realizes a function in metabolism, secretion, or gene expression regulation. This situation limits the conclusions of the estimations of the OS based on mRNA values, as our data were derived from the KM plot analysis.

Therapeutic implications and resistance bypass strategies. The present results provide a foundation for therapeutic strategies aimed at overcoming hyperglycemia-induced chemoresistance in TNBC. Previous studies have shown that inhibitors or interfering RNAs targeting YBX3 increase chemosensitivity, highlighting a promising approach for TNBC (19,42). Other potential therapeutic targets include metabolic reprogramming and transporter activity, considering the SLC6A6, SLC39A7 and RPS13 genes (16,17,29), which are overexpressed in HG and aforementioned. Pharmacological inhibition of these transporters, or modulation of glucose and lipid metabolism, could sensitize TNBC cells to chemotherapy. Combined strategies that target metabolic and oncogenic signaling pathways may prove particularly effective. The inhibition of the Wnt/β-catenin pathway or ATM-mediated DNA repair, in combination with chemotherapy, could synergistically reduce tumors. These findings support a framework for personalized combinatorial therapies in diabetic patients with BC, integrating metabolic modulators, inhibitors of YBX3-regulated networks, and conventional chemotherapeutics to overcome resistance and improve clinical outcomes.

Limitations of the present study. The present study has potential limitations. It was based on the analysis of the changes induced by HG in an invasive BC cell line, using an in vitro experimental and bioinformatics approach. Although the MDA-MB-231 cell line has been widely used as a model of invasive BC, it does not represent the genetic heterogeneity of invasive breast tumors. Therefore, the current findings need to be validated in other experimental models and clinical cohorts. However, the present study provides important insights and allows us to better understand the behavior of this type of cancer under HG conditions, as well as to identify potential signaling pathways and genes associated with chemoresistance and prognosis in patients with T2DM.

Although the use of HPRT as a housekeeping gene is considered unreliable for comparing the gene expression between cancerous and non-cancerous tissues (48), this gene has been recommended in hyperglycemic conditions (13), and its expression is moderately stable in BC cell lines (12), although in hyperglycemic conditions, it has been validated in other types of cells. Therefore, the use of an additional reference gene is recommended to further consolidate the overexpression of TSPAN and FZD3 genes.

The present findings and bibliographic references suggest that TSPAN1 and FZD3 play an important role in the induction of chemoresistant and invasive phenotypes by HG (18,32). However, gene knockdown experiments are required to better understand their roles.

The clinical implications were partially validated using public patient datasets and informatics platforms; however, these platforms are generally based on non-diabetic patients, and the estimation of the effect of selected overexpressed genes on OS was therefore performed in non-diabetic conditions. These results should be interpreted with caution, highlighting the need for further validation in diabetic patient cohorts.

Additionally, both diabetes and cancer are complex processes involving multiple genetic and environmental contexts; therefore, further studies are required to validate the clinical implications of our findings. These include validation at both protein (western blotting and IHC) and mRNA (RT-qPCR) levels of overexpressed genes (FZD3 and TSPAN1), as well as of potential genes associated with chemoresistance and prognosis (AREG and YBX3) in breast tumor tissues of diabetic patients.

In addition, changes in gene expression and the induction of cisplatin chemoresistance by diabetes in a mouse model of syngeneic BC are currently studied, and patient-derived organoids can be used to analyze cisplatin action under HG conditions in human tissues. Finally, the correlation of the selected genes and clinical outcomes should be studied in other patient cohorts, preferably in diabetic patients with BC.

In conclusion, these data indicate that alterations in gene expression induced by hyperglycemia lead to changes in the metabolism of cancer cells, affecting a variety of processes that promote a more aggressive phenotype through the activation of Wnt/β-catenin, NF-κB survival, angiogenesis and PI3K/AKT signaling, contributing to enhanced proliferation, migration, invasion and inhibition of apoptosis. Some potential biomarkers of cisplatin resistance under hyperglycemia could be AREG and YBX3.

Supplementary Material

Upregulated genes in MDA-MB-231 cells supplemented with glucose 30 mmo/l.

Downregulated genes in MDA-MB-231 cells supplemented with glucose 30 mmo/l.

Acknowledgements

The authors would like to thank Ms Lorena Chávez, Dr José Luis Santillán, Mr Simón Guzmán and Dr Jorge Ramírez for performing the microarray analysis at the Instituto de Fisiología Celular, National Autonomous University of Mexico. The authors thank student Mr Jael Gutiérrez Savedra for his technical assistance, and Dr Leticia Moreno Fierros for her technical advice [both are affiliated to the National Autonomous University of Mexico (UNAM), FES-Iztacala].

Funding

Funding: The present study was supported by the Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT), Dirección General de Apoyo al Personal Académico (DGAPA), Universidad Nacional Autónoma de México (UNAM), (grant nos. IN223121 and IN220024), by the Financing for Research of Women Scientists (grant no. FICDTEM-2023-56) and by the Fiscal Resources Program for Research of the National Institute of Pediatrics (grant no. 2020/016).

Availability of data and materials

The data generated in the present study may be found in the Gene Expression Omnibus under accession number GSE136277 or at the following URL: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE136277. The data generated in the present study are included in the figures and/or tables of this article.

Authors' contributions

ARVR, MGMH and LABG contributed to the conception and design of the study. ARVR, MGMH, LAFL, MÁVF, RREG and JRPB performed material preparation, data collection and analysis. ARVR and LABG wrote the first draft of the manuscript. All authors commented on previous versions of the manuscript. All authors read and approved the final version of the manuscript. ARVR and LABG confirm the authenticity of all the raw data.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References