Human pancreatic cancer cell lines including BXPC3, PANC-1, and ASPC-1 as well as human normal pancreatic cells HPDE6-C7 were obtained from the Jiangsu Center for the Collaboration and Innovation of Cancer Biotherapy. PANC-1, ASPC-1, and HPDE6-C7 were cultured in DMEM (sigma, D5796) with 10% fetal bovine serum (FBS), and 1% penicillin, streptomycin, and BXPC3 were cultured in RPMI-1640 medium (sigma, R8758) with 10% FBS and 1% penicillin and streptomycin at 37°C, 5% CO₂.

We designed two small guide RNAs for knocking out ITGB5 by the CHOPCHOP web tool (20):

sgITGB5-1: 5′-CACCGCTTTCTCCTACACGGCACCG-3′

sgITGB5-2: 5′-CACCGCTGGCGAACCTGTAGCTGGA-3′

293T cells growing on a 6-cm dish were firstly transfected with 4 μg pCDH-ITGB5 or Lenticrispr-ITGB5 or control vector (pCDH, Lenticrispr V2), 3 μg PSPAX2, 1 μg PMD2G, and Lipofectamine 2000 for obtaining lentivirus expressing ITGB5. Forty-eight hours after transfection, the culture medium supernatant was centrifuged to remove cell debris at 1,000 rpm, 5 min for infection. PANC1 or BXPC3 cells seeded on a 6-cm dish with 60% cell density were infected by the above supernatant, and then stable ITGB5 knockout cell lines (ITGB5-KO) or ITGB5 overexpression cell lines (ITGB5-OE) with puromycin for 3 days were selected. Western blot analysis was used to evaluate the ITGB5 expression efficiency.

ITGB5 was obtained from Abcam (ab184312, Abcam) and Novus Biologicals (NBP1-88117, Novus Biologicals) for Western blot and immunohistochemistry (IHC) studies, respectively. MEK1/2 (9154T) and γ-H₂AX (9718S) were from Cell Signaling Technology (CST). Ki67 (A11005), ERK, and pERK (A11116) were from ABclonal. KU70 (10723-1-AP), KU80 (16389-1-AP), RAD51 (14961-1-AP), ATM (27156-1-AP), CHEK2 (13954-1-AP), Phospho-CHEK2 (Thr68) (29012-1-AP), PAX6 (12323-1-AP), STAT1 (10144-2-AP), GAPDH (10494-1-AP), and β-actin (23660-1-AP) were from Proteintech. DNA-PKcs (BS1092) was from Bioworld. HRP-Goat Anti-Rabbit IgG (VA001) and HRP-Goat Anti-Mouse IgG (VA002) were obtained from Vicmed.

IHC was conducted as previously described (21). Simply, we used the PV-9000 two-step immunohistochemistry kit and DAB color reagent (ZSBIO) to detect antigens in human PAAD tissue microarrays (TMA) (HPanA 150Su01, Shanghai Outdo Biotech, China) and mouse tissue samples according to the manufacturer’s instruction.

CCK-8 and colony formation assays were used to evaluate the influence of ITGB5 on the tumor cells’ growth rate. ITGB5-KO, ITGB5-OE, and control BXPC3 and PANC-1 cells were seeded into 96-well plates (1,000 cells per well). The changes in cell number were detected with a CCK-8 assay kit continuously for 4 days. The absorbance at 450 nm (OD₄₅₀) of cells was measured by a microplate reader. Furthermore, cell proliferation ability was further detected by colony formation assays. The equal number of cancer cells was seeded in six-well plates with appropriate density and cultured for 10–14 days at 37°C, 5% CO₂. The medium was replaced every 3 days. In radiation sensitization assays, cancer cells with appropriate cell intensity seeded in six-well plates were one-off irradiated with 0, 2, 4, 6, and 8 Gy, respectively, at a dose rate of 300 cGy/min. All cells were irradiated with 6-MeV X-ray, and the source-surface distance was 100 cm. Besides, the surface of plates was covered with a 1-cm-thick tissue equivalent filler. After radiation, cells were cultured until the visible colonies formed about 10~14 days later. Then, cell clones were fixed with 4% paraformaldehyde (Vicmed, China) and stained with crystal violet (Beyotime, China). The number of colonies was counted by ImageJ software. Plating efficiency (PE) was calculated by the following formula: (the number of colonies/the number of cells plated) × 100%. Survival fraction (SF) was calculated by the following formula: (PE of experimental group/PE of control group) × 100% (22, 23). The relevant parameters such as α, β, D0, Dq, and sensitization enhancement ratio (SER_D0 = D0 of experimental group/D0 of control group) and surviving curves were obtained according to the single-hit multitarget model (SF = 1-(1-e^-D/D0)^N, Dq = D0 × lnN) and linear–quadratic model (SF = e^-αD-βD2) by GraphPad Prism 8.0 software. Each experiment was repeated three times.

Transwell assay was used to assess the cell invasion and migration ability. The cell density was adjusted to 2.5 × 10⁵ cells/ml in serum-free medium, and 200 µl was added to the upper chamber with or without Matrigel. Eight hundred microliters of medium with 20% FBS was added to the lower chamber. Cells on the membrane filters were fixed and stained with crystal violet. The number of invasive cells was photographed using a bright microscope and then calculated. Each experiment was repeated three times.

Western blot (WB) was used to evaluate the effect of ITGB5 knockout or overexpression efficiency and irradiation (IR) on the expression of DNA damage proteins or other signal pathway proteins in cancer cells with different ITGB5 expression levels. For evaluating expression efficiency, the adherent cancer cells seeded in six-well plates were directly harvested for WB; for expression detection of other proteins, the adherent cancer cells seeded in six-well plates were irradiated with 6 Gy at a dose rate of 300 cGy/min. At different time points after irradiation, cells were harvested for WB as in a previous study (22). Briefly, membranes were incubated with primary antibodies at 4°C overnight and with secondary antibodies for 1 h at room temperature. Then, membranes were washed three times and then imaged with Chemiluminescence Image System. The band intensity was analyzed by ImageJ software. Each experiment was repeated three times.

Cell-cycle assay was used to examine the effect of ITGB5 expression on pancreatic cancer cell-cycle distribution. Briefly, cancer cells (ITGB5-KO, ITGB-OE, and control BXPC3 cells) were seeded in 6-cm plates with appropriate density. All cells were irradiated with 6 Gy at dose rate of 300 cGy/min. Twenty-four hours after irradiation, all cells were directly harvested using a cell cycle assay kit (KeyGEN BioTECH, KGA512) according to vendors’ instructions and analyzed by flow cytometry. Each experiment was repeated three times.

All experiments in vivo were conducted ethically on the basis of the Jiangsu Council on Animal Care. Four- to 6-week-old male BALB/c nude mice (GemPharmatech, China) were adaptively reared at the SPF level Laboratory Animal Center for 1 week before the experiment began. The distal right or left lower extremities of mice were subcutaneously injected with 5 × 10⁶ BXPC3 or PANC-1 cells suspended in 200 μl medium without FBS. Tumor volumes were measured with a caliper every 2 days and calculated by the following formulation: V = 0.5 × a × b² (a is the larger diameter and b is the smaller diameter). Thirty days after subcutaneous neoplasia, mice were euthanized by overdose of carbon dioxide. Subcutaneous tumors were dissected completely and weighted, which were then processed for further detection of ITGB5 and Ki67 expression by IHC. Tumor growth curves were obtained by GraphPad Prism 8.0 software.

The expression level of ITGB5 in different types of tumors was obtained from the GEPIA website (http://gepia.cancerpku.cn/) (24). Gene expression data of PAAD were downloaded and processed from TCGA-PAAD (https://portal.gdc.cancer.gov/) and GTEx (https://xenabrowser.net/datapages/) databases as in our previous study (25). Simply, a total of 171 normal pancreatic samples (167 from GTEx and four from TCGA) and 178 PAAD samples were used to analyze the ITGB5 expression difference. According to the clinical information downloaded from TCGA-PAAD, the samples with incomplete follow-up information were deleted and, ultimately, 173 PAAD samples were used for survival analysis and clinical correlation analysis. Kaplan–Meier (KM) survival curves were used to evaluate the influence of ITGB5 on overall survival (OS) by the “survminer” package in R (V.4.0.2) software. The cutoff value of ITGB5 expression was its median value. Wilcox test was used for clinical correlation analysis of ITGB5 expression in R (V.4.0.2) software.

All statistical analyses were conducted by SPSS 25. Data were presented as means ± standard deviation (SD). The statistical significance difference was compared by two-sample t-test or one-way ANOVA. The X ² test was done to analyze the association between ITGB5 expression and clinical features of PAAD. Kaplan–Meier (K-M) analysis was used to assess survival curves, and log-rank test was used to compare survival curves among groups. p < 0.05 was considered significant.

Firstly, we analyzed ITGB5 expression in cancers by the GEPIA database ( Figure 2A ). ITGB5 is upregulated in numerous cancer types including PAAD (N_tumor = 179, N_normal = 171) compared to normal samples. Then, we analyzed the ITGB5 expression difference between PAAD samples and normal pancreatic samples based on TCGA and GTEx databases; ITGB5 is significantly upregulated in PAAD ( Figure 2B ). KM survival analysis revealed that ITGB5 overexpression is highly bound up with poor prognosis ( Figure 2C ). Concerning the clinical correlation ( Figure 2D ), ITGB5 expression was clearly upregulated in patients with a T3&T4 stage or distant metastases compared to patients with T1&T2 stage or non-distant metastases, which highlighted the invasive and metastatic potential of ITGB5 expression in pancreatic cancer. Lastly, we examined ITGB5 protein expression in TMA containing 90 PAAD samples and 60 normal pancreatic samples. As shown in Figures 2E, F , consistent with the results from the database analysis, ITGB5 is highly regulated in PAAD samples. We further analyzed the relationship of ITGB5 with prognosis and clinical characteristics in PAAD. Samples with incomplete clinical factors were rejected, and a total of 87 PAAD samples were included in the analysis. Clinical characteristics of 90 PAAD samples are summarized in Table 1 , of which 39 samples were 60 years old or younger and 48 samples were 60 years old or older. Fifty-six were men and 31 were women. Concerning pancreatic cancer infiltration, 31 samples were coupled with perineural invasion, 52 samples were coupled with lymph node metastasis, and 50 samples were located in the head of pancreas. Most of the PAAD samples had earlier clinical stages, 27 samples were stage I, 59 samples were stage II, and the rest of the samples were stage IV. ITGB5 expression was related with tumor size, lymph node metastasis, and clinical stage significantly (p < 0.05). Besides this, ITGB5 overexpression in PAAD predicted unsatisfactory prognosis ( Figure 2G ). In conclusion, ITGB5 is not only highly regulated in PAAD compared to normal pancreas but also related with poor prognosis, which can be a potential prognostic marker in PAAD.

To investigate the role of ITGB5 in PAAD, we established ITGB5 knockout or overexpression pancreatic cancer cell lines. Firstly, we examined the expression of ITGB5 in three pancreatic cancer cell lines (BXPC3, PANC-1, and ASPC1) and one normal pancreatic cell line (HPDE6-C7). As shown in Figure 3A , compared to HPDE6-C7, ITGB5 expression was obviously upregulated in BXPC3, PANC-1, and ASPC1. Considering that the expression of ITGB5 was highest in PANC-1 and lowest in BXPC3, we used PANC-1 and BXPC3 to construct stable ITGB5 knockout or overexpression cells ( Figures 3B, C ). CCK-8 results showed that compared to control cells (Lenticrispr V2 or pCDH), knocking out ITGB5 expression in PANC-1 or BXPC3 (ITGB5-KO-1, ITGB5-KO-2) obviously inhibited cell proliferation; on the contrary, upregulated ITGB5 expression in PANC-1 or BXPC3 (ITGB5-OE) obviously promoted cell proliferation ( Figures 4A, B ). Colony formation assays reached the same conclusion ( Figures 4C-F ). Transwell assay showed that ITGB5 overexpression promoted migration and invasion of PAAD cells, which was oppositely inhibited by the silence of ITGB5 ( Figure 5 ).

Based on results in vitro, we further investigated ITGB5 function in vivo by constructing xenograft tumors. Considering that the knockout efficiency of sgITGB5-1 is superior to sgITGB5-2, we used ITGB5-KO constructed by sgITGB5-1 for experiment in vivo. ITGB5-KO and control cells were subcutaneously injected in the distal right and left lower extremities of one mouse, respectively. As shown in Figures 6A, B , ITGB5 depletion significantly inhibited xenograft tumor growth in both PANC-1 and BXPC3 by recording tumor volume and tumor weight. The expression of proliferating protein Ki67 in xenograft tumors was further detected by IHC, which was consistent with ITGB5 expression ( Figure 6C ). When ITGB5 was downregulated, Ki67 protein expression was inhibited, which further supported the results of in vitro experiments that ITGB5 depletion inhibited tumor proliferation.

To investigate the influence of ITGB5 on radiation sensitivity, ITGB5-KO, ITGB5-OE, and control cells were irradiated with 0, 2, 4, 6, and 8 Gy. According to colony formation assay and survival curves generated using survival fractions at different radiation doses in Figures 7A-F , ITGB5 depletion inhibited the number of colonies and survival curves became steeper ( Figures 7G, H ). Sensitization enhancement ratios (SER_D0) of ITGB5-KO and ITGB5-OE in PANC-1 cells were 1.404 and 0.862, respectively, which in BXPC3 were 1.341 and 0.864 ( Table 2 ). ITGB5 overexpression in pancreatic cancer cells promoted radio-resistance.

We then investigated the possible mechanism by which ITGB5 induces enhanced radio-resistance in pancreatic cancer cells. Firstly, we tested whether ITGB5 overexpression induced more DNA double-strand breaks (DSBs). γ-H₂AX, formed at DSBs damage sites, is used as a marker of DNA DSBs (26). WB was further used to detect the γ-H₂AX protein expression of ITGB5-KO, ITGB5-OE, and control BXPC3 cells at different time points (0, 2, 4, 6, 8, 24 h) after irradiation. As shown in the results in Figures 8A, B , γ-H₂AX protein expression increased 2 h after irradiation in the control group and basically dropped to normal levels after 6–8 h of irradiation. However, in the group of ITGB5-KO BXPC3 cells, γ-H2AX expression increased obviously and was still at a high expression level 24 h after irradiation; on the contrary, in the group of ITGB5-OE cells, γ-H₂AX expression was significantly lower than that in the control group. A semiquantitative statistical analysis of the gray level of WB strips showed that 0, 2, 4, 6, 8, and 24 h after irradiation, the expression level of γ-H₂AX in ITGB5-KO cells was significantly higher than that in the control group, and the difference was statistically significant. On the contrary, the expression level of γ-H₂AX in ITGB5-OE cells was significantly lower than that in the control group.

Secondly, we examined DNA damage repair protein expression ( Figure 8C ). DNA-PK, ATM (ataxia telangiectasia mutated), and ATR (ataxia telangiectasia and Rad3 related) belong to the phosphatidylinositol 3-kinase-related kinase family (PI3K). DNA-PK and ATM were mainly activated by DSBs, while ATR was mainly activated by single-strand breaks (SSBs). Here, we focused on the effect of ITGB5 on the expression changes of DNA-PK and ATM targeting at DSB repair. The results showed that Rad51, DNA-PKcs, and KU70/80 protein expression levels were significantly increased in the ITGB5-OE BXPC3 cells and decreased in ITGB5-KO BXPC3 cells compared with control cells (p < 0.05). After irradiation, the expression differences of Rad51, DNA-PKcs, and KU70/80 protein between ITGB5-OE or ITGB5-KO BXPC3 cells and control cells were more obvious (p < 0.01). Another ATM-CHEK2 signaling pathway, ATM, CHEK2, and phospho-CHEK2 (pCHEK2) were highly upregulated in the group of ITGB5-OE. Abnormally increased ITGB5 promoted DNA damage repair by activating DNA-PK and ATM, which then enhanced radio-resistance.

The activation of DNA damage repair proteins may induce cell-cycle arrest. We further examined cell-cycle distribution after irradiation. As shown in Figure 9 , in group ITGB5-OE, the proportion of G2/M phase cells was significantly increased and, adversely, the proportion of G2/M phase cells was significantly decreased in group ITGB5-KO.

Lastly, we examined protein expression related with the MEK/ERK signaling pathway ( Figure 10 ). MEK and ERK protein expression showed no significance between ITGB5-KO or ITGB5-OE BXPC3 cells and control cells both before and after irradiation (p > 0.05). However, compared with the control cells, the protein expression levels of p-MEK and p-ERK were increased in the ITGB5-OE BXPC3 cells and significantly decreased in the ITGB5-KO BXPC3 cells, and the differences were more significant after irradiation (p < 0.05). Besides, in the ITGB5-OE group, PAX6 and STAT1 expression downstream of the MEK/ERK signaling pathway was also significantly increased before and after irradiation. These results suggested that ITGB5 promoted DNA damage repair by activating the MEK/ERK pathway and then enhanced radio-resistance.