The cervical cancer cell lines CaSki and HeLa were maintained in the Molecular Genetics Laboratory (Central South University, Changsha, China). Cervical cancer cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium containing 10% fetal bovine serum (FBS; both from Gibco; Thermo Fisher Scientific, Inc.) with 5% CO₂ at 37°C. The CD38‑overexpressing cell lines (CaSki-CD38 and HeLa-CD38) and the control cell lines (CaSki-vector and HeLa-vector) were transfected with the CD38-overexpressing plasmid pEGFP-N1-CD38 and the control plasmid pEGFP-N1 using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol, as described previously (30,31). The transfected cell lines were grown in RPMI-1640 supplemented with 10% FBS and G418 (500 µg/ml). Following 3 weeks, the stable cells were used for subsequent experiments.

Protein extraction and digestion were performed as previously described (32). Briefly, cells were lysed using a protein extraction buffer consisting of 50 nM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDA and sodium orthovanadate, sodium fluoride, EDTA, leupeptin supplemented with 1X halt protease inhibitor cocktail (CWBio) and 1X halt phosphatase inhibitor cocktail (BestBio). The protein concentration was estimated using the bicinchoninic acid (BCA) method using a Micro BCA™ Protein assay kit (Thermo Fisher Scientific, Inc.). Then, cell lysates (50 µg of each sample) were loaded onto 10% SDS-PAGE gels and separated electrophoretically at 80 volts for 40 min and then 120 volts for 60 min (PowerPac Universal; Bio-Rad Laboratories, Inc.). Protein bands were visualized using Coomassie brilliant blue G-250 (Sigma-Aldrich; Merck KGaA) and excised from the gel into eight slices. The proteins were destained using 15 mM K₄Fe(CN)₆, and then 50 mM sodium thiosulfate and 1.25 µg trypsin (1:20 enzyme/substrate ratio) were added to each slice and in‑gel‑digestion was performed at 37°C overnight for ~16 h. The generated peptides were extracted by sonication (15 min, with ice cooling) from the gel pieces in ~20 µl 50% acetonitrile in 0.1% formic acid, twice. Following extraction, the peptides were dried using vacuum centrifugation at room temperature for 30 min to ensure the complete removal of acetonitrile and reconstituted in 20 µl 0.1% formic acid.

LC-MS/MS analysis was performed as previously described (32). Briefly, peptides were diluted using 0.1% formic acid. Then, peptides were pre-concentrated. The analytes were transferred to the analytical column and separated using a binary system. The mass spectrometer was operated in the data-dependent mode. Normalized collision energy was set to 35% and an isolation width of 2 m/z was chosen.

Protein identification and quantification were performed as previously described (32,33). Briefly, proteins were identified using Proteome Discoverer 1.4 software (Thermo Fisher Scientific, Inc.). Thermo raw files were imported and used to conduct a search of the UniProt proteomes-Homo sapiens database (UP000005640; https://www.uniprot.org/taxonomy/960). For database searches, mass tolerances were set to 10 ppm and 0.8 Da for precursor and fragment ions, respectively. Peptides identified with false discovery rates <1% (q-value <0.01) were discarded. A common contaminants database was also included for quality control. Proteins that met the following criteria were considered differentially expressed proteins: i) proteins had ≥2 peptides with ≥95% confidence; ii) proteins were considered downregulated when the protein levels demonstrated an average fold‑change ≤0.5 in the LC‑MS/MS analyses; and iii) proteins were considered upregulated when the protein levels demonstrated an average fold‑change ≥2 in the LC-MS/MS analyses (33).

Cells were trypsinized and inoculated into 6-well cell culture plates (1×10⁶ cells/well), and then incubated at 37°C in the presence of 5% CO₂. Following 24-48 h, when the density of cells had reached ~80%, the culture medium supernatant was transferred into a new centrifuge tube and centrifuged at 1,000 × g for 5 min at 4°C to remove insoluble materials. The supernatant was transferred into specific tubes to detect the glucose concentration using an ADVIA 1650 automatic biochemical analyzer (Siemens AG). Meanwhile, the cells in the culture plate were trypsinized, washed with cold 1× PBS solution and placed in microcentrifuge tubes. Following ultrasonication on ice for 3 min, the samples were centrifuged at 12,000 × g for 20 min at 4°C. The supernatant was transferred into specific tubes to detect the intracellular glucose concentration using the ADVIA 1650 automatic biochemical analyzer (Siemens AG). All values were normalized on the basis of cell number.

The cellular ATP concentration was measured using an ATP detection assay kit (catalog no. S0026; Beyotime Institute of Biotechnology), according to the manufacturer's protocol. Cells were trypsinized, seeded into 6-well cell culture plates (1×10⁶ cells/well) and incubated at 37°C in the presence of 5% CO₂. Following 24 h, when the cell density reached ~80%, the culture medium was aspirated from the plate and 200 µl ATP lysis buffer was added into each well. The lysis buffer and cells were pipetted repeatedly to homogenize the cells on ice for 5 min. The cell lysate was transferred into new pre-cooled tubes and centrifuged at 12,000 × g for 15 min at 4°C. The supernatant was transferred into new pre-cold tubes and kept on ice. The ATP standard buffer and ATP detection buffer were prepared. The ATP detection buffer was added into the wells of a 96-well plate (100 µl/well) and kept at room temperature for 5 min. The samples or ATP standard buffer were then added into the wells (20 µl/well) and incubated in the dark for 5 min at room temperature. The luminescence emitted by the samples was detected using a Paradigm Detection Platform (Beckman Coulter, Inc.). The concentration of cellular ATP (nmol/mg) was calculated based on a standard curve. All values were normalized to the protein concentration.

The lactate concentration was detected using a Lactate assay kit (catalog no. MAK064; Sigma-Aldrich; Merck KGaA), according to the manufacturer's protocol. Briefly, cells were trypsinized and seeded into 6-well cell culture plates (1×10⁶ cells/well). The cells were incubated at 37°C in the presence of 5% CO₂. Following 24-48 h, when the cell density reached ~80%, the culture medium supernatant was transferred into new centrifuge tubes, and centrifuged at 12,000 × g for 20 min at 4°C to remove insoluble materials. The supernatant was transferred into 10 kDa cut‑off spin filters to deproteinize the samples to remove enzymes that may consume lactate. The filtrate was transferred into new tubes and kept on ice to detect the lactate concentration in the medium. Meanwhile, cells in the culture plate were digested and washed with cold 1× PBS solution and the resuspended cells were placed in microcentrifuge tubes. Following ultrasonication on ice for 3 min, the samples were centrifuged at 12,000 × g for 20 min at 4°C. The supernatant was transferred into 10 kDa cut‑off spin filters to deproteinize the samples. The filtrate was transferred into new tubes and kept on ice to detect the intracellular lactate concentration. Then, 50 µl samples were placed into the wells of a 96-well plate, 50 µl Master Reaction mix was added to each sample, and the plate was incubated in the dark for 30 min at room temperature. The absorbance value at 570 nm was detected with a microplate reader and analyzed with SoftMAx Pro 6.4 (Beckman Coulter). The lactate concentration was calculated based on a standard curve. All values were normalized on the basis of cell number.

Western blotting was performed as described previously (30,31). The cells were trypsinized and then lysed using RIPA lysate buffer (CWBio). Samples were centrifuged at 12,000 × g for 15 min at 4°C to remove insoluble materials. The protein concentration estimated using the BCA method, and each protein sample (50 µg) was electrophoresed using a 10% SDS-PAGE gel at 80 volt for 40 min and then at 120 volt for 60 min (PowerPac Universal; Bio-Rad Laboratories, Inc.). The separated proteins were transferred onto a polyvinylidene fluoride membrane (HyClone; GE Healthcare Life Sciences) at 100 volt for 90 min and then the membranes were blocked for 1-2 h using 5% non-fat milk dissolved in PBS-0.1% Tween 20. The membranes were then incubated overnight with primary antibodies at 4°C. The following primary antibodies were used: Anti-CD38 (catalog no. YM0122; ImmunoWay Biotechnology Company), anti-3-phos-phoinositide dependent protein kinase 1 (catalog no. YT3645; ImmunoWay Biotechnology Company), anti-PI3K P110 (catalog no. YT3709; ImmunoWay Biotechnology Company), anti-AKT (catalog no. YT0173; ImmunoWay Biotechnology Company), anti-phosphorylated (p)-AKT T308 (catalog no. YP0590; ImmunoWay Biotechnology Company), anti-mTOR (catalog no. YT2913; ImmunoWay Biotechnology Company), anti-p-mTOR (catalog no. YP0716; ImmunoWay Biotechnology Company), anti-lactate dehydrogenase A (LDH-A; catalog no. 19987-1-AP; ProteinTech Group, Inc.) and anti-ATP synthase peripheral stalk subunit D (ATP5H; catalog no. YT0406; ImmunoWay Biotechnology Company). These antibodies were stored at ‑20°C, and the dilution used for all antibodies was 1:1,000. The membranes were then washed and incubated with the following horseradish peroxidase‑conjugated secondary antibodies for 1 h at 37°C: Anti-rabbit secondary antibody (catalog no. sc-2004; Santa Cruz Biotechnology. Inc.; 1:3,000) and anti-mouse secondary antibody (catalog no. sc-2005; Santa Cruz Biotechnology. Inc.; 1:3,000). The immunoreactive protein bands were then visualized using ECL luminescent liquid (EMD Millipore) and analyzed using Molecular Imager^® Gel Dox XR System (Bio-Rad Laboratories). Rabbit anti-GAPDH (catalog no. 2118; Cell Signaling Technology, Inc.; 1:5,000), rabbit anti-β-actin (catalog no. T0022; Affinity Biosciences; 1:5,000) and rabbit anti-β‑tubulin antibody (catalog no. T0028; Affinity Biosciences; 1:5,000) were used as controls. ImageJ (v 1.51; National Institutes of Health) was used to quantitatively analyze the results.

Cervical cancer cells were transfected with the empty vector and the CD38 overexpression vector and then seeded at a density of 800 cells/well in a 6-well plate. Following adherence, the cells in the wells were treated with 0.1 µM DMSO and 0.1 µM rapamycin (catalog no. S1039; Selleck Chemicals) at 37°C for 24 h. When the clone was visible to the naked eye, the culture was terminated. The culture medium was discarded and the cells were washed three times with PBS buffer. The cells were treated with paraformaldehyde for 30 min at room temperature, washed with 1× PBS, stained with 0.1% crystal violet for 30 min at room temperature and washed with tap water. Images were then obtained and the cells were counted.

A Hoechst33342/propidium iodide (PI) double staining kit (catalog no. bb-4131; Bestbio) was used to detect cell apoptosis according to the recommended protocol. The CD38 overexpression vector and the blank vector were transfected into CaSki and HeLa cells. The cells were seeded in 6‑well plates and cultured at 37°C with 5% CO₂. After adhering to the wall, the cells were treated with DMSO and rapamycin. Following 24 h, the cells were digested with 0.25% trypsin, collected, and washed twice with PBS. The cells were resuspended in 0.5-1.0 ml staining buffer, and Hoechst33342 staining solution A (5-10 µl) was then added. Following gentle mixing, the cells were incubated at 4°C in the dark for 10 min, and 5-10 µl PI was then added. Staining solution B was added (5-10 µl), the samples were gently mixed, and then incubated at 4°C in dark for 5‑10 min. Finally, the cells were washed with PBS, resuspended in PBS and detected using flow cytometry. Analysis was performed using Summit v5.2 software (Beckman Coulter, Inc.).

Data are presented as the mean ± standard deviation of three independent experiments. Differences in parametric variables were analyzed using ANOVA followed by Student-Newman-Keuls post hoc test, and differences in quantitative variables between groups were analyzed by Kruskal-Wallis followed by Bonferroni's test. Statistical analyses were performed with the EPI Info software (version 3.2.2; www.CDC.gov/epiinfo/). P<0.05 was considered to indicate a statistically significant difference.

Our previous study demonstrated that CD38 expression is upregulated in cervical cancer. In vitro experiments revealed that CD38 could inhibit the apoptosis of cervical cancer cells and promote their proliferation. The previous results also demonstrated that CD38 affects the intracellular calcium concentration, mitochondrial membrane potential and intracellular reactive oxygen species in cervical cancer cells (30). Therefore, it was speculated that CD38 may be associated with the metabolism of cervical cancer cells.

To investigate if CD38 affects the metabolism of cervical cancer cells, LC-MS/MS technology was first used to screen and identify differentially abundant proteins between CD38-overexpressing cervical cancer cells and the control groups. Proteins were identified using Proteome Discoverer 1.4 software and analyzed by the UniProtKB/Swiss-Prot database. It was identified that 573 proteins were downregulated in CaSki-CD38 cells compared with the control cells and 274 proteins were downregulated in HeLa-CD38 cells compared with the control cells. Among these proteins, 40 proteins were consistently downregulated in both CaSki-CD38 and HeLa-CD38 cells. These 40 overlapping downregulated proteins are listed in Table I. By contrast, 357 proteins were found to be upregulated in CaSki-CD38 cells and 359 proteins were upregulated in HeLa-CD38 cells compared with the control cells. Among these, 36 proteins were consistently upregulated both in CaSki-CD38 and HeLa-CD38 cells. These 36 overlapping upregulated proteins are listed in Table II.

The results of the LC-MS/MS analysis revealed that ATP5H-2, cytochrome B5 type B (CYB5B), cytochrome C1 (CYC1) and other protein molecules were upregulated in CD38-overexpressing cervical cancer cells, while BRI3 binding protein (BRI3BP) and other protein molecules were downregulated. To further analyze the biological processes and signaling pathways that involve these differentially abundant proteins, GO and KEGG analyses were performed. As presented in Fig. 1, the differentially abundant proteins were annotated to be involved in the following biological processes: 'Glucose catabolic process to pyruvate', 'oxidation-reduction process', 'glucose metabolic process', 'nucleoside monophosphate metabolic process', 'glycosyl compound metabolic process', 'NAD/NADH metabolic process' and 'ATP/ADP metabolic process'. This indicates that these proteins are predominantly involved in glycolytic pathways, oxidative phosphorylation and NAD/NADH metabolic processes. KEGG analysis demonstrated that CD38 overexpression was associated with 'pyruvate metabolism', 'glycolysis/gluconeogenesis', 'metabolic pathways' and 'fatty acid elongation'. Thus, it can be suggested that CD38 is involved in regulating cellular energy metabolism in cervical cancer cells.

To test the hypothesis that that CD38 is involved in regulating cellular energy metabolism in cervical cancer cells, the effects of CD38 over-expression on glucose uptake, intracellular ATP production and intracellular lactic acid concentration in cervical cancer cells were investigated. The glucose concentration in the culture medium was significantly decreased in CaSki‑CD38 and HeLa-CD38 cells compared with vector-transfected cells (P=0.0002 and P<0.0001, respectively; Fig. 2A). The intracellular glucose concentration showed no significant difference between CD38-overexpressing cervical cancer cells and the control groups (Fig. 2B). It was speculated that, to a certain extent, CD38 promotes glucose consumption in cervical cancer cells. The lactate concentration in the culture medium was significantly increased in CaSki‑CD38 and HeLa‑CD38 cells compared with vector-transfected cells (P<0.0001 and P=0.0021, respectively; Fig. 2C). The intracellular lactate concentration was also significantly increased in CaSki‑CD38 and HeLa-CD38 compared with vector-transfected cells (P<0.0001 and P=0.019; Fig. 2D).

Further research revealed that CD38 overexpression significantly increased the intracellular ATP concentration in CaSki cells (P=0.001; Fig. 2E) and HeLa cells (P=0.003; Fig. 2F). These results suggest that CD38 is involved in regulating the glucose concentration and can lead to an accumulation of lactate in cervical cancer cells.

To investigate the possible effects and mechanism of CD38 on the energy metabolism in cervical cancer cells, key molecules of the PI3K/AKT/mTOR signaling pathway were investigated in CaSki and HeLa cells. In CaSki cells, CD38 overexpression significantly increased the expression levels of PI3K, PDK1, ATP5H and LDHA. CD38 overexpression also significantly increased the ratios of p-AKT/AKT and p-mTOR/mTOR (Fig. 3A). These results suggest that CD38 had a significant effect on the phosphorylation of AKT and mTOR. To confirm that CD38 is involved in the regulation of energy metabolism in cervical cancer cells via the PI3K/AKT/mTOR signaling pathway, the cervical cancer cells were treated with rapamycin. Following treatment with DMSO, the expression levels of LDHA and ATP5H were significantly higher in CaSki‑CD38 cells compared with the control CaSki-vector cells. In addition, the p-mTOR/mTOR ratio was significantly higher in CaSki-CD38 cells compared with the control cells following treatment with DMSO. After treatment with rapamycin, the expression levels of LDHA and ATP5H were significantly lower in CaSki-CD38 cells compared with the control CaSki-vector cells. In addition, the p-mTOR/mTOR ratio was significantly decreased (Fig. 3B). These results suggest that rapamycin can block the effects of CD38 on LDHA, ATP5H and mTOR. ATP5H and LDH-A affect the concentration of ATP and lactic acid in CaSki cells (34). To determine whether rapamycin alters the effects of CD38 on ATP and lactic acid in CaSki cells, the concentration of ATP and lactic acid were measured in rapamycin-treated cells. The results demonstrated that rapamycin could significantly reverse the effects of CD38 overexpression on ATP and lactic acid levels in CaSki cells (Fig. 3C and D).

Furthermore, the protein levels of mTOR, p-mTOR, LDH-A and ATP5H were detected in HeLa cells. In HeLa cells, CD38 overexpression significantly increased the levels of p‑mTOR and LDHA, but no difference was observed for total mTOR or ATP5H. However, CD38 significantly increased the ratio of p‑mTOR/total mTOR, indicating that CD38 has a significant effect on the phosphorylation of mTOR (Fig. 4A). Similarly, following treatment with DMSO, the expression levels of LDHA and ATP5H were significantly higher in HeLa‑CD38 cells compared with the control HeLa-vector cells. In addition, the ratio of p-mTOR/mTOR was increased in HeLa-CD38 cells. Following treatment with rapamycin, the expression levels of LDHA and ATP5H in HeLa‑CD38 cells were significantly lower compared with the control HeLa-vector cells. Furthermore, the p‑mTOR/mTOR ratio was significantly lower in HeLa‑CD38 cells compared with HeLa-vector cells following treatment with rapamycin (Fig. 4B). These results were consistent with the results in CaSki cells. The effects of rapamycin on the ATP and lactic acid levels in HeLa-CD38 cells were consistent with those observed in CaSki-CD38 cells (Fig. 4C and D). Therefore, the mTOR signaling pathway may serve a role in the effects of CD38 on the energy metabolism of cervical cancer cells.

The levels of lactic acid and ATP metabolites are closely associated with cell proliferation. To determine whether CD38 promotes the proliferation of cervical cancer cells by activating the PI3K/AKT/mTOR signaling pathway, cells were treated with rapamycin and the effect of CD38 overexpression on cell proliferation was detected using clone formation experiments. Following CD38 overexpression, the number of clones of CaSki cells treated with DMSO was significant increased, whereas following treatment with rapamycin the number of clones of CaSki‑CD38 cells was significantly decreased compared with those treated with DMSO (Fig. 5A). Similar results were obtained in HeLa cells (Fig. 5B). This indicates that CD38 affects the proliferation of cervical cancer cells via the PI3K/AKT/mTOR signaling pathway.

Furthermore, apoptosis of rapamycin and DMSO-treated CaSki-CD38 and HeLa-CD38 cells was assessed. It was identified that CD38 inhibited the apoptosis of DMSO‑treated cervical cancer cells (P<0.01), and compared with the CD38-overexpressed cells treated with DMSO, the apoptosis rate of CD38-overexpressed cells treated with rapamycin increased (P<0.05; Fig. 6). This indicates that the effects of CD38 on the apoptosis of cervical cancer cells may occur via the PI3K/AKT/mTOR signaling pathway (Fig. 7).