The NSCLC cell lines, A549, H1975, 95D and H1299 were maintained in Dulbecco's modified Eagle's medium (DMEM; Corning, Corning, NY, USA) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 100 units penicillin/streptomycin (all from Invitrogen, Carlsbad, CA, USA) in 5% CO2 at 37°C in a humidified chamber. Rabbit anti-HURP monoclonal antibody (cat. no. ab107646) was purchased from Abcam (Cambridge, UK). Mouse anti-GAPDH monoclonal antibody (cat. no. sc-32233) and secondary polyclonal antibody, rabbit IgG (cat. no. sc-2004) and mouse IgG (cat. no. sc-2005), were obtained from Santa Cruz Biotechnolgy (Santa Cruz, CA, USA).

Total RNA was isolated from cell lines using Invitrogen™ Trizol reagents (Thermo Fisher Scientific, Inc., Waltham, MA, USA) following the manufacturer's instructions. cDNA was synthesized from 1 µg of total RNA by M-MLV Reverse Transcriptase (Promega, Madison, WI, USA). The primer sequences for HURP and GAPDH are as follows: HURP forward, 5′-AAGTGGGTCGTTATAGACCTGA-3′ and reverse, 5′-TGCTCGAACATCACTCTCGTTAT-3′; GAPDH forward, 5′-TGACTTCAACAGCGACACCCA-3′ and reverse, 5′-CACCCTGTTGCTGTAGCCAAA-3′.

The real-time quantitative polymerase chain reaction (RT-qPCR) analyses were performed with SYBR-Green Master Mixture (Takara, Shiga, Japan) on LightCycler 480 instrument (Roche Diagnostics GmbH, Mannheim, Germany). Each of the 12 µl quantitative PCR mixture contained 6 µl SYBR-Green Master Mixture, 0.6 µl cDNA product, 0.3 µl each of the 5 µM forward and reverse primers, and 5.1 µl RNase-free H2O. All these quantitative PCR experiments were performed in triplicate. The housekeeping gene GAPDH was used as an internal control. The 2^−ΔΔCt (Ct is the threshold cycle) method was used to calculate relative target gene expression.

For the knockdown of human HURP, HURP-specific shRNAs were designed and purchased commercially (GeneChem, Shanghai, China). The target sequences of shRNAs were as follows: HURP-shRNA#1, GCAATGAGAGAGAGAATTA; HURP-shRNA#2, GGATATAAGTACTGAAATG and HURP-shRNA#3, TTGAAAGAGCAGAGAGAGA.

Lentiviral particles were packaged in 293T cells by co-transfecting shRNA vectors, or control shRNA, together with the packaging plasmids.

Briefly, the NSCLC cell lines were lysed with RIPA buffer containing protease inhibitor. Following determination of the protein concentrations using the BCA Protein Assay kit (Beyotime Institute of Biotechnology, Haimen, China), equal amounts of protein per sample (20 µg) were separated on 10% sodium dodecyl sulfate (SDS) polyacrylamide gels for 2 h and transferred to nitrocellulose membranes. The membranes were subsequent blocked with 5% low-fat milk in TBST for 1 h and incubated with primary antibodies overnight at 4°C. The primary antibodies included: anti-HURP (1:2,000; Abcam) and GAPDH (1:2,000; Santa Cruz Biotechnology). This was followed by incubation with secondary antibodies (1:5,000; Santa Cruz Biotechnology) for 1.5 h at room temperature and rinsed four times with TBST for 8 min each. Reactions were visualized using the ECL system (Thermo Fisher Scientific, Waltham, MA, USA).

The cells were seeded in 96-well plates with DMEM containing 10% FBS for 1 to 5 days, followed by the MTT assay. Briefly, 20 µl of 5 mg/ml MTT solution was added and incubated with the cells for 4 h at room temperature. Then the medium was removed and 100 µl dimethyl sulfoxide (DMSO) was added for 5 min to extract the colored product catalyzed by the living cells. The end product was quantified with a spectrophotometer (Tecan Infinite, Mölnda, Sweden) at 490 nm wavelength. The final data were normalized with the OD490 value on day 1.

The NSCLC cells were trypsinized and washed twice with cold phosphate-buffered saline (PBS), and then stained with Annexin V-FITC/PI in the dark for 15 min at 37°C following the manufacturer's instructions. Binding buffer (400 µl) was added to a test tube and the proportion of apoptotic cells was quantified with a flow cytometer (Merck Millipore, Darmstadt, Germany).

Cells under a log phase of growth were trypsinized and washed twice with cold PBS. After centrifugation, the cells were fixed with 100% ice-cold methanol at 4°C overnight. Then the cells were stained with 50 mg/ml propidium iodide (PI) and Rnase staining buffer in PBS for 30 min in the dark at room temperature. BD FACS (Becton Dickinson, San Jose, CA, USA) was used for analyzing cell cycle stage. A total of 10,000 cells were subjected to cell cycle analysis using a flow cytometer (Merck Millipore).

Transwell chambers (8-µm pore size) (Corning, Corning, NY, USA) were used in the cell migration and invasion assays. The cells were suspended in serum-free DMEM. At first, a total of 100 µl DMEM containing 104 cells was plated into the upper chamber of Transwell chambers, and 600 µl DMEM containing 10% FBS was added to the lower chamber. Then the cells were cultured at 37°C with 5% CO2 for 16 h. The cells in the upper surface were removed with cotton swabs. Cells that had adhered to the lower surface were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Images were captured under a microscope (Olympus Corp., Tokyo, Japan) at ×100 magnification. The protocol for the invasion assay was the same, except that the Transwell chambers were precoated with Matrigel (BD Biosciences, San Jose, CA, USA), and the cells were also cultured for 16 h. Each experiment was performed in triplicate, and representative images are shown.

Gene Set Enrichment Analysis (GSEA) was used to compare the gene expression level of each indicated geneset between the high HURP expression group and low HURP expression group. We divided GSE68465 and GSE30219 into two groups according to HURP expression in order to explore the biological process potentially modulated by HURP. The gene sets with normal P-value <0.05 and a false discovery rate (FDR) value <0.25 after 1,000 permutations were considered significantly enriched, and the Molecular Signatures Database was used to download the genesets.

The original array data underwent background correction and were normalized to the base-2 logarithm by Robust Multi-Array Average and Linear Models for Microarray. Then the differentially expressed genes (DEGs) between the low HURP expression group and high HURP expression group were identified. The absolute value of log2-fold change ≥1.5 and P-value <0.001 were considered as the threshold value. To assess the prospective functions of DEGs, we utilized the Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.kegg.jp/kegg/) using the Database for Annotation, Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov/), which provides a comprehensive set of functional annotation tools for investigators to understand biological meaning behind a large list of genes. The Search Tool for the Retrieval of Interacting Genes (STRING, https://string-db.org/) was used to investigate the interaction between these DEGs. The Gene Ontology (GO) enrichment analysis of DEGs was performed by BiNGO plugin for Cytoscape (http://apps.cytoscape.org/apps/bingo. The supplier is Steven Maere (VIB Department of Plant Systems Biology, UGent).

The Human Protein Atlas (HPA, http://www.proteinatlas.org/) is a public database that builds on gene expression data that includes quantitative transcriptomics data and spatial proteomics data (16). It curates histological images of 44 normal human tissues and the 20 most common types of cancer. In addition, HPA also provides functional analyses of proteomes and serves as an advantageous online tool in protein expression analysis and medical diagnosis. The expression of HURP in lung cancer tissues was obtained from the data deposited in the HPA. The expression level of HURP was defined as high, medium or low relative to that of normal lung tissues.

We investigated the prognostic value of HURP among NSCLC patients. The NSCLC mRNA expression data and corresponding clinical data were obtained from Gene Expression Omnibus (GEO) database. The GSE33532 and GSE19188 were gene expression microarrays providing primary tumors and matched distant normal lung tissue or adjacent normal tissue from NSCLC patients. The GSE30219 and GSE68465 were included in the analysis because of complete clinical data and large sample size. We defined the low expression cases as those whose HURP mRNA expression was less than or equal to the median value in order to explore the association between HURP and clinical characteristics. The χ2 test was applied to estimate the relationship between HURP expression and clinical parameters, where appropriate. Overall survival and recurrence-free survival were estimated with the Kaplan-Meier method. Survival differences were further validated by KM-plotter (http://kmplot.com/analysis/index.php?p=background) (17). The expression level of HURP in NSCLC cell lines was detected in Metabolic gEne RApid Visualizer (MERAV, http://merav.wi.mit.edu/SearchByGenes.html) and further validated by RT-qPCR and western blot analysis. Differences between the gene expression levels within different groups of cells were analyzed using ANOVA analysis followed by Tukey's multiple conparison. Each assay was performed three times for averaging replicates, and the unpaired Student's t-test was applied to evaluate the differences between the control and treatment group. All P-values are two-tailed with a significant level at 0.05. The above statistical analyses were performed using SPSS V19.0 (SPSS, Inc., Chicago, IL, USA) and GraphPad Prism (V.6.0) (GraphPad Software, Inc., La Jolla, CA, USA).

At first, we observed HURP expression at the mRNA level of NSCLC patients from GSE33532 and GSE19188. As shown in Fig. 1A, HURP expression was significantly upregulated in tumor tissues when compared with that noted in the distant normal lung tissues. Furthermore, HURP was able to provide high accuracy in regards to NSCLC tissue classification as estimated by ROC curve analysis (Fig. 1B). Therefore, we assessed the Human Protein Atlas database in order to analyze the expression of HURP in lung cancer and their normal counterparts. As shown in Fig. 1C, in accordance with the microarray analysis data, positive immunostaining of HURP in lung cancer tissues was observed (8 positive cases from 12 analyzed tumor samples). Furthermore, we assessed the association between HURP mRNA levels and the clinicopathologic characteristics in two independent cohorts of NSCLC patients from the GEO datasets. We categorized NSCLC patients into two groups based on HURP mRNA levels and defined the low expression cases as whose with HURP mRNA expression less than or equal to the median value. As shown in Table I, higher HURP expression was correlated with higher American Joint Committee on Cancer (AJCC) T stage (P<0.01) and N stage (P<0.01) in GSE30219. There was no significant difference between the high and low HURP expression group in regards to age (P=0.13) distribution. The multivariate analyses revealed that the progression-free survival (PFS) and overall survival (OS) of patients with high expression of HURP were significantly shorter than patients with high expression of HURP after adjusting for age, sex, clinical stage and pathology (PFS: HR=2.71, 95% CI 1.53–4.80, P<0.01; OS: HR=1.75, 95% CI 1.18–2.61, P<0.01) (data not shown).

More high HURP expression patients presented with a higher clinical stage (P<0.01) and poorer differentiated carcinoma (P<0.01; Table II), and the results were similar to those obtained from GSE30219. Although no association between HURP and PFS was observed in the multivariate analyses, the relevance of HURP with OS was significant with adjustment for age, sex and clinical stage (HR=1.46, 95% CI 1.10–1.93, P<0.01). These results suggest that high expression of HURP may cause rapid tumor cell proliferation and a highly aggressive phenotype (data not shown).

The KM-plotter was further utilized to validate the survival differences for all NSCLC patients in 14 microarray datasets with 1928 patients. The results indicated that the first progression time, post progression survival time and overall survival time were significantly different between the high and low HURP expression groups in the NSCLC patients. Higher HURP expression was associated with shorter survival time (Fig. 1D).

At first, we analyzed the expression level of HURP in 95-D, A549, H1299 and NCI-H1975 cell lines. The MERAV database indicated that the expression of HURP was upregulated in all four NSCLC cell lines. Futhermore, RT-qPCR revealed that A549 and H1299 cells expressed higher HURP compared to the 95-D cells (Fig. 2A). To investigate whether HURP is involved in the proliferation of NSCLC cells, we infected H1299 and A549 cells with a lentivirus to silence HURP. Three shRNAs were designed to suppress the expression of HURP and we detected the mRNA and protein levels of HURP in the different groups. The mRNA levels of HURP were significantly reduced in cells infected with shRNA#1 and shRNA#3 (P<0.01). Accordingly, the protein levels of HURP in the shRNA groups were also greatly decreased compared with levels noted in the negative control group (Fig. 2B). shRNA#1 was chosen for the following experiments due to the better efficacy. Expression of green fluorescence protein was observed in each group (Fig. 2C).

To explore the potential cause of HURP in regulating the proliferation of NSCLC cells, MTT assays were conducted to assess cell viability. As shown in Fig. 3A, after a 48-h post infection in control or shRNA groups, the cell viability was markedly decreased in the cells transfected with the HURP-targeting shRNA. The optical density (OD) values at 490 nm of the HURP shRNA H1299 group (0.34±0.01, 0.36±0.01, 0.43±0.01) were significantly lower than those in the negative control group (0.54±0.07, 0.84±0.01 and 1.26±0.01) from day 3 to day 5 (P<0.05). Meanwhile, similar results were found in the A549 cells. The OD values at 490 nm of the HURP shRNA A549 group were 0.36±0.05, 0.38±0.04 and 0.42±0.02 compared to 0.56±0.02, 0.74±0.06, and 0.92±0.03 in the negative control group from day 3 to day 5 (P<0.05).

Cell cycle assays detected by flow cytometry revealed that the cells were mainly distributed in the S stage in the HURP-null group (HURP shRNA H1299 group, 35.09±0.32%; negative control group, 28.83±0.10%, P<0.05). Accordingly, the percentage of HURP shRNA A549 cells in the S phase (39.12±0.43%) was significantly higher than that of the negative control group (32.03±0.25%, P<0.05; Fig. 3B and C). The results suggested that HURP not only regulates cell division through spindle assembly, but also promotes cell proliferation through other means.

We then detected the effects of HURP on apoptosis. However, there was no significant difference in apoptosis rates between the control group and HURP-knockdown group (data not shown). Collectively, HURP may modulate the cell cycle to regulate cell proliferation.

To evaluate whether HURP plays a role in cell migration and invasion, the Transwell assay was performed. A549 and H1299 cells were pretreated with either shNC or shRNA#1. After 16 h of incubation, the cells that transmigrated to the lower chamber were significantly decreased in the shRNA group compared to that noted in the negative control group (P<0.05, Fig. 4A). Furthermore, the Transwell matrix penetration assay was conducted to assess the effects of HURP on cell invasion. As shown in Fig. 4B, after 16-h post infection of shNC or shRNA#1, the cells with downregulated HURP expression demonstrated statistically significant decreased invasion ability in the A549 and H1299 cells (P<0.05). Altogether, HURP modulates NSCLC migration and invasion in vitro.

After elucidating the role of HURP in the modulation of proliferation, cell cycle regulation and migration, we aimed to ascertain other potential functions of HURP in NSCLC cells. In order to corroborate the biological process potentially modulated by HURP, we performed GSEA analysis using the microarray datasets of GSE30219 and GSE68465. The enriched expression of genesets included cell cycle, cell mitosis, cell cycle checkpoints, and mitotic spindle regulation. In particular, HURP was involved in G1/S phase transition, cyclin E-associated events during G1/S transition and M/G1 transition. Furthermore, HURP may participate in activation of NF-κB, P53 independent DNA damage checkpoint and transcription process (Fig. 5).

The DEGs were obtained from GSE30219, GSE68465 and TCGA database, and 247 genes are summarized in total (Fig. 6A). We performed KEGG pathway analysis to clarify the potential biological functions of HURP. The top KEGG pathways enriched for DEGs included cell cycle, p53 signaling pathway, small cell lung cancer, pathways in cancer and viral carcinogenesis (Fig. 6B). In addition, the DEGs were mainly enriched in the GO biological process related to cell cycle regulation (Fig. 6C) and the top 50 significant GO biological process terms are shown in Table III.