ZIP10 drives osteosarcoma proliferation and chemoresistance through ITGA10-mediated activation of the PI3K/AKT pathway

OS tissues were obtained from 64 OS patients at the Eighth Affiliated Hospital, Sun Yat-sen University, Guangdong, China. Of these patients, 52 did not undergo chemotherapy, and 12 underwent chemotherapy (AP, doxorubicin 25 mg/m², cisplatin 100 mg/m²). All specimens were frozen and maintained in liquid nitrogen.

Human OS cell lines (Saos2, U2OS and 143B) were obtained from the American Type Culture Collection (ATCC) Cell Bank and cultured according to the instructions from the ATCC. Starting with the 143B osteosarcoma cell line, the cisplatin-resistant variant was obtained by exposing the parental line to stepwise-increased cisplatin concentrations. This continuous exposure resulted in cells resistant to 3 μM cisplatin (143BR). Mesenchymal stem cells (MSCs) from healthy donors were isolated and cultured as previously described [16]. For stimulation of OS cells, ZnSO₄ (15 μM; Sigma-Aldrich), Zn chelator TPEN (3 μM; TargetMol), cisplatin (0.5 μM; Sigma-Aldrich), CREB inhibitor 666–15 (0.1 μM; MCE), AKT activator SC79 (5 μM; MCE) and AKT inhibitor GSK690693 (10 μM; MCE) were added to the medium.

Total RNA was extracted from OS cells using TRIzol reagent (Invitrogen) and then used for cDNA synthesis as detailed by the manufacturer (TaKaRa). Quantitative analyses were carried out using SYBR Premix Ex Taq II reagents (TaKaRa), with ACTB as the internal reference. Relative expression was calculated based on the 2^-ΔCt method. The primer sequences are listed in Table S1.

Total protein extracts were obtained for protein separation by 10% SDS-PAGE, followed by protein transfer onto PVDF membranes. After treatment with 5% nonfat milk, the membranes were probed with diluted primary antibodies (1:1000) against ACTB (CST; #4970), ZIP10 (SAB; #24824), AKT (CST; #4691), p-AKT (CST; #4060), ERK (CST; #4695), p-ERK (CST; #9101), JNK (CST; #9252), p-JNK (CST; #4671), C-PARP (CST; #5625), C-caspase-3 (CST; #9661), ITGA10 (SAB; #44824), PAK (CST; #2602), p-PAK (CST; #2606), p-FAK (CST; #8556), p-SRC (CST; #12432), CREB (CST; #12432), and p-CREB (CST; #9198) overnight at 4 °C. The membranes were then washed in TBST, followed by incubation with a diluted HRP-labeled secondary antibody (1:5000; CST; #7074 and #7076) for 2 h at room temperature. Signals were examined using an ECL detection system (Bio-Rad). Data were analyzed by ImageJ (version 1.8.0).

The CCK-8 assay was used to assess cell proliferation. OS cells were seeded into 96-well plates at a density of 1 × 10⁴ cells/well and incubated for 24 h at 37 °C in 5% CO₂. After 0, 1, 2, 3, and 4 days of cultivation, 10 μL CCK-8 reagent was added to each well followed by culture for 2 h at 37 °C in 5% CO₂. Absorbance at 450 nm was measured using a microplate reader (Bio-Rad).

Cell proliferation was monitored by EdU staining reagent (Beyotime) following the manufacturer’s instructions. Nuclear detection was carried out using Hoechst 33342 staining, after which the cells were observed by confocal microscopy (Nikon C2).

Cells were treated with/without cisplatin for 48 h, seeded into 6-well plates (5 × 10²/well) and maintained at 37 °C in a 5% CO₂ incubator for 2 weeks. Colonies were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet, and their numbers were counted under a microscope. Three independent experiments were performed in triplicate.

OS cells were seeded into 6-well plates. After culturing for 12 h, 0.5 μM cisplatin was added to each well for 3 days. After treatment, total cells were harvested and suspended at 1 × 10⁶ cells/mL; 5 μL Annexin V and 7-aminoactinomycin D (7-AAD) staining solution was added to 300 μL of the cell suspension. After incubation for 15 min at room temperature in the dark, the stained cells were assayed and quantified using a FACSort Flow Cytometer.

Briefly, to generate ZIP10-knockdown and ZIP10-overexpressing cells, pLKO.1-shZIP10 and pGC-FU-ZIP10 lentivirus plasmids were cotransfected into 293 T cells with psPAX2 and pMD2.G. Viral supernatant fractions were collected at 48 h after transfection followed by infection into Saos2 or 143B cells together with 6 μg/mL polybrene. After overnight infection, the cells were incubated in medium containing 2 μg/mL puromycin for 3 days. To knock down the expression of ITGA10 by siRNA transfection, OS cells in 6-well plates were transiently transfected with 150 pmol siRNA with Lipofectamine 3000 (Invitrogen; #11668–019) for 6 h following the manufacturer’s instructions. The target sequences are as follows: shZIP10#1: 5′-GCG TGA TCT TGG TTC CTA T-3′; shZIP10#2: 5′- GCA TTA GCT GTA GGA ACA A-3′; siITGA10: 5′-CCT GAG AGA AAT TAG AAC T-3′; NC: 5′-TTC TCC GAA CGT GTC ACG T − 3′.

OS cells (5 × 10⁴/well) were precultured overnight in 96-well plates, incubated for 1 h in PBS containing 1 μM FluoZin-3 AM (Life Technologies) and then washed three times with PBS. Fluorescence intensity (494 mm excitation/516 nm emission) was measured using a microplate reader (SpectraMax; Molecular Devices).

ITGA10-CRE was cloned into the pGL3-basic luciferase reporter plasmid. OS cells (2.5 × 10⁴ cells per well) were seeded in triplicate in 24-well plates. After culture for 12 h, the cells were transfected with 200 ng of ITGA10-CRE-luciferase-reporter plasmids. Each transfection included the same amount of Renilla luciferase plasmid, which was used to standardize the transfection efficiency. After incubation for 24 h, PBS, ZnSO4 or 666–15 was added to the medium and incubated for another 24 h. Firefly and Renilla signals were measured using a Dual Luciferase Reporter Assay kit (Promega), and the results are presented as the increase in activation over reporter alone.

ChIP assays were performed according to the manufacturer’s instructions for the SimpleChIP enzymatic ChIP kit (CST). Briefly, a ChIP assay was performed using protein A/G agarose and an anti-p-CREB antibody. The immunoprecipitated DNA was used to amplify DNA fragments via PCR with specific primers. The primer sequences are listed in Table S2.

Xenograft tumors were generated by subcutaneous injection of 143B cells (2 × 10⁶) into the flanks of 6-week-old athymic nude mice (n = 6/group). Tumors were measured using calipers every 3 days. For drug treatment, when the average tumor volume reached approximately 100 mm³, the mice were administered cisplatin (3 mg/kg/2 days), 666–15 (10 mg/kg/day) or GSK690693 (30 mg/kg/day) by intraperitoneal injection. The tumor volume was calculated according to the following formula: tumor volume (mm³) = length × width × width/2. Mice were euthanized at 30 days after cell injection to obtain tumor weights.

Tumor tissues obtained from OS patients or xenografts were embedded in paraffin and subjected to IHC staining with specific antibodies. The immunoreactions were evaluated independently by two pathologists. Briefly, the percentage of positive cells was scored as follows: 0, no positive cells; 1, ≤10% positive cells; 2, 10–50% positive cells; and 3, > 50% positive cells. Staining intensity was scored as follows: 0, no staining; 1, weak staining; 2, moderate staining; and 3, dark staining. The comprehensive score (0, 1, 2, 3, 4, 6, 9) = staining percentage × intensity. ZIP10 expression ≤2 indicates a low level, whereas > 2 indicates a high level.

RNA sequencing of ZIP10 knockdown cells was performed by BGI Tech based on DNBSEQ. In total, 6.8 GB clean data per sample were collected for RNAseq, and the clean reads were aligned to the human genome GRCh38 (Hg38).

Because Zn transporters ZIP/SLC39 have been implicated in tumor progression, we explored their role in OS chemoresistance. To determine whether ZIP/SLC39 expression was induced by cisplatin, we first treated the OS cell lines 143B and Saos-2 with 0.5 μM cisplatin for 24 h, a condition few cells underwent apoptosis (Fig. S1), excluding the effect of cell subpopulation selection by cisplatin. Then, we extracted total RNA for ZIP/SLC39 expression analysis. The results showed that ZIP10 expression in cisplatin-treated OS cells was nearly 3-fold that in untreated cells (Fig. 1a). Elevated ZIP10 expression (nearly 2.5-fold) in cisplatin-treated cells was also confirmed by WB (Fig. 1b and S3a). Both qRT-PCR and WB results indicated that ZIP10 expression increased during cisplatin treatment and reached a peak (nearly 4-fold in mRNA and nearly 3-fold in protein) at 72 h (Fig. S2). Accordingly, ZIP10 expression was higher in postchemotherapy patients (n = 12) than in prechemotherapy patients (n = 12; Fig. 1c, d), indicating that chemotherapy induces ZIP10 expression in OS. Moreover, ZIP10 was overexpressed in 3 OS cell lines tested compared to mesenchymal stem cells (MSCs) (Fig. 1e and S3b), which is a typical control cell line for OS cells. These results indicate that ZIP10 was highly induced by cisplatin treatment and might function in OS carcinogenesis and chemoresistance.

To explore the role of ZIP10 in OS, we generated stable ZIP10-knockdown OS cell lines (shZIP10 decreased by 80%). Knockdown of ZIP10 inhibited proliferation in 143B (Fig. 1f and S3c) and Saos-2 (Fig. S4a) cells and significantly enhanced cisplatin-induced suppression of colony formation (Fig. 1g and S4b). To determine the role of ZIP10 in the tumorigenesis and chemoresistance of OS in vivo, we established an athymic nude mouse model and found inhibition of tumor growth in the shZIP10 group (Fig. 1h). The average tumor weight of the vehicle-treated NC group was approximately 0.52 g, and the average tumor weight of the cisplatin-treated NC group was approximately 0.39 g, a 25% decrease compared to the vehicle-treated NC group. However, the average tumor weight of the cisplatin-treated shZIP10 group was approximately 0.20 g, which was 47% lower than that of the vehicle-treated shZIP10 group (0.38 g). Moreover, we generated stable ZIP10-overexpressing (oeZIP10) 143B cells. As shown in Fig. S5, ZIP10-overexpressing cells exhibited a higher proliferation rate and cisplatin resistance both in vitro and in vivo. These results indicate that ZIP10 positively regulates proliferation and chemoresistance in OS.

We next explored whether ZIP10 contributes to chemoresistance in drug-resistant cell lines. First, we exposed 143B cells to stepwise-increased cisplatin concentrations and generated cells resistant to 3 μM cisplatin (143BR). 143BR showed a proliferation rate and cellular morphology similar to those of 143B cells (Fig. S6a, b). In addition to resistance to cisplatin, 143BR also partially resisted adriamycin (ADR) and methotrexate (MTX) (Fig. S6c), indicating broad chemoresistance. Both qRT-PCR and WB results showed that ZIP10 expression increased by 100% in 143BR (Fig. S6d, e). By knocking down ZIP10 expression by 90% in 143BR cells, we found a decrease in the proliferation rate and cisplatin resistance both in vitro and in vivo (Fig. S7). These results indicated that high ZIP10 expression was associated with chemoresistance in 143BR. Overall, ZIP10 is overexpressed in OS and contributes to proliferation and chemoresistance.

To delineate the functional implications of ZIP10 in OS, we performed transcriptome sequencing to investigate expression changes in ZIP10-knockdown cells. According to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, differentially expressed genes were significantly enriched in gene sets involved in the PI3K/AKT pathway (Fig. 2a). Among them, 22 genes were downregulated (Fig. 2b) and 10 genes were upregulated (Fig. S8) in shZIP10 cells. These results indicate that PI3K/AKT signaling might be inhibited in shZIP10 OS cells. To examine this possibility, we analyzed several chemoresistance-associated signaling pathways in shZIP10 OS cells by WB and found that the protein level of p-AKT was decreased by 75–80% in shZIP10 OS cells but that those of p-ERK and p-JNK were not obviously changed (Fig. 2c and S9). These results indicate that ZIP10 contributes to the PI3K/AKT signaling pathway in OS cells.

To explore whether the ZIP10-mediated PI3K/AKT signaling pathway is associated with OS chemoresistance, we employed flow cytometry (Fig. 2d and S10a) and WB (Fig. 2e and S10b) to assess apoptosis induced by cisplatin. The results showed that the AKT activator SC79 rescued cisplatin-induced apoptosis in shZIP10 143B cells; in contrast, the AKT inhibitor GSK690693 attenuated the chemoresistance induced by ZIP10 overexpression (oeZIP10). Similar results of ZIP10-induced cell proliferation were observed after EdU incorporation (Fig. 2f and S10c). Consistent with these results, the protein level of p-AKT was increased by 200% in 143BR (Fig. S6e). All these results show that ZIP10 promotes cell proliferation and chemoresistance in OS through the PI3K/AKT pathway.

We next explored how ZIP10 activates the PI3K/AKT pathway in OS. We noticed that the expression level of several integrins upstream of the PI3K/AKT pathway was altered in shZIP10 cells (Fig. 3a), which was confirmed by qRT-PCR (Fig. 3b and S11). Among these genes, ITGA10 had a nearly 70% decrease in shZIP10 cells. A previous study showed that ITGA10 activates the RAC/PAK and AKT pathways in myxofibrosarcoma [14], and we thus evaluated the expression of these signaling molecules in OS cells with/without ITGA10 knockdown (ITGA10 siRNA; siITGA10). Knockdown of ITGA10 significantly decreased p-AKT levels by 80–90% while slightly decreasing p-PAK levels but did not change p-FAK and p-SRC levels (Fig. 3c). Similarly, overexpression of ZIP10 increased the p-AKT level by 100–150%, which was attenuated by siITGA10 (Fig. 3d, e). We therefore hypothesize that ZIP10 activates AKT through ITGA10. We next performed CCK-8 assays (Fig. 3f), flow cytometry (Fig. 3g and S12) and WB (Fig. 3h) to further explore whether the ZIP10-ITGA10-p-AKT axis accounts for the observed proliferation and chemoresistance in oeZIP10 OS. The results showed that knockdown of ITGA10 attenuated oeZIP10-induced proliferation and chemoresistance in OS. Consistently, the expression of ITGA10 was increased by 200% in 143BR cells (Fig. S6d, e). Taken together, these results support that ZIP10 activates the PI3K/AKT pathway by promoting ITGA10 expression.

Because Zn transporters ZIP/SLC39 primarily transport Zn from the extracellular to intracellular space, many studies have shown that knockdown of ZIP decreases Zn concentrations in cells [17]. In this study, we confirmed a lower Zn content in shZIP10 OS cells (Fig. 4a) and a higher Zn content in 143BR (Fig. 4b). Intracellular Zn levels were shown to regulate cell proliferation and survival in cancers [6]. We treated OS cells with 0, 5, 10, 15, 30, 50, 100, and 200 μM ZnSO₄ and 0, 1, 3, 5, 10, and 20 μM TPEN (a Zn chelator) to evaluate the role of Zn in proliferation and chemoresistance. The results showed that both 143B and Saos-2 proliferation increased with increasing ZnSO₄ concentrations lower than 15 μM but decreased with increasing ZnSO₄ concentrations higher than 30 μM (Fig. S13a). However, 143B and Saos-2 proliferation decreased with increasing TPEN concentrations (Fig. S13b). Since the human serum Zn concentration is approximately 12–20 μM [18], a decrease in Zn levels might inhibit the proliferation of OS in vivo. Moreover, apoptosis assays showed that 15 μM ZnSO4 increased the chemoresistance of OS cells, while 3 μM TPEN inhibited chemoresistance (Fig. S13c). All these results indicated that an appropriate increase in cellular Zn levels can lead to increased chemoresistance of OS cells.

We next explored whether ITGA10 expression is regulated by the Zn content in OS cells. As revealed by qRT-PCR (Fig. 4c) and WB (Fig. 4d), Zn supplementation rescued the decrease in ITGA10 expression in shZIP10 143B cells, whereas TPEN attenuated the increase in ITGA10 expression induced by ZIP10 overexpression. These results demonstrate that ITGA10 expression is regulated by Zn content in OS.

As previous studies have reported that the Zn content regulates cyclic AMP response element-binding protein (CREB) activation in cancer cells [19, 20], we next explored p-CREB/CREB levels in OS cells. Knockdown of ZIP10 decreased the p-CREB level by 80% in 143B cells, and Zn supplementation rescued this decrease in p-CREB (Fig. 4e). ITGA10 expression and CREB activation are both regulated by the Zn content, and we thus explored their relationship. As shown in Fig. 4f, the CREB inhibitor 666–15 inhibited ITGA10 expression by 65%, whereas knockdown of ITGA10 did not alter the p-CREB level. These results indicate that CREB is upstream of ITGA10. CREB is a basic leucine zipper (bZIP) transcription factor that recognizes the palindromic cAMP response element (CRE) motif 5′-TGA CGT CA-3′ [21]. We used JASPAR to search the ITGA10 promoter, i.e., the − 2000 bp to 0 bp sequence upstream of the transcription start site (TSS), and found candidate CRE sequences (CRE1 and CRE2). To analyze ITGA10 promoter motifs required for Zn-mediated expression, we assessed ITGA10 promoter luciferase constructs with deletions in the CRE1 and/or CRE2 motifs (Fig. 4g). Zn supplement-induced ITGA10 promoter activation was substantially reduced by deletion of the CRE2 motif, although CRE1 deletion decreased ITGA10 promoter activity to a lesser extent (Fig. 4h). Deletion of both CRE motifs repressed ITGA10 promoter activity comparable to CRE2 deletion alone. Direct binding of the ITGA10 promoter by CREB was also monitored by anti-p-CREB ChIP, followed by polymerase chain reaction (PCR) and qRT-PCR for the CRE motifs in the ITGA10 promoter. ChIP revealed enriched p-CREB binding within regions of the putative CRE motifs predicted by JASPAR. Moreover, the binding of p-CREB to CRE1 and CRE2 was increased by ZnSO4 treatment, with CRE2 showing more enrichment than CRE1 (Fig. 4i, j). These results indicate that CREB mediates ITGA10 transcription. Taken together our findings show that ZIP10 increases CREB phosphorylation, leading to upregulation of ITGA10 transcription.

To investigate the role of the ZIP10-ITGA10-p-AKT axis in the regulation of OS cell proliferation and sensitivity to chemotherapy in vivo, we performed xenograft tumor experiments using 143B and 143BR cells.

In 143B cells, tumor growth and chemoresistance were promoted by ZIP10 overexpression, and this promotion was attenuated by the CREB inhibitor 666–15 or AKT inhibitor GSK690693 (Fig. 5a-c). Overexpression of ZIP10 increased the protein levels of ITGA10 and p-AKT; 666–15 reduced ITGA10 and p-AKT, and GSK690693 decreased p-AKT (Fig. 5d). Furthermore, IHC staining results indicated that overexpression of ZIP10 increased Ki67 and p-AKT expression in xenograft tissues with/without cisplatin treatment but that oeZIP10 + 666–15 or oeZIP10 + GSK690693 did not (Fig. 5e and S14). In addition, apoptosis, as analyzed by cleaved caspase 3 (C-caspase-3) staining, was dramatically decreased in cisplatin-treated oeZIP10 xenograft tissues.

For 143BR, knockdown of ZIP10 dramatically inhibited tumor growth and chemoresistance, and this inhibition was rescued by the AKT activator SC79 (Fig. S15a-c). Knockdown of ZIP10 decreased the protein levels of ITGA10 and p-AKT, and SC79 increased p-AKT (Fig. S15d). Furthermore, IHC staining results indicated that knockdown of ZIP10 increased C-caspase-3 levels in xenograft tissues with cisplatin treatment but that shZIP10 + SC79 did not (Fig. S15e).

Taken together, these findings suggest that the activated ZIP10-ITGA10-p-AKT axis is indispensable for cisplatin resistance in OS cells.

To determine whether ZIP10, p-CREB, ITGA10 and p-AKT levels correlate in OS, we collected 52 patient tumor tissue sections without chemotherapy. The association between ZIP10 level and clinical features of patients, including gender, age, anatomical site, Enneking’s stage and AJCC stage, are summarized in Table S3. ZIP10 correlated positively with Enneking’s stage (p = 0.0238) and AJCC stage (p = 0.0111). Representative staining images with high or low levels of ZIP10, p-CREB, ITGA10 and p-AKT expression are shown in Fig. 6a. ZIP10, p-CREB, ITGA10 and p-AKT scores were evaluated by systematically analyzing the IHC staining results (Fig. 6b-f). Among 52 patients, 21 cases of high p-CREB levels were found for all 30 individuals with a high level of ZIP10 staining; 24 of 30 patients with a high level of ZIP10 exhibited an upregulated level of ITGA10 protein, and 26 of 30 patients with a high level of ZIP10 showed an upregulated level of p-AKT protein (Fig. 6g-i). Additionally, increased CREB and AKT activation was accompanied by ITGA10 overexpression (Fig. 6j, k). As expected, statistically significant positive correlations between ZIP10 and p-CREB (Fig. 6l), ZIP10 and ITGA10 (Fig. 6m), ZIP10 and p-AKT (Fig. 6n), ITGA10 and p-CREB (Fig. 6o) and ITGA10 and p-AKT (Fig. 6p) were observed. These findings suggest that ZIP10 leads to ITGA10-mediated AKT activation associated with clinicopathologic features and may contribute to tumorigenesis of OS.