All chemical reagents were purchased from Sigma-Aldrich (Merck KGaA). Cell culture media and supplements were obtained from Gibco (Thermo Fisher Scientific, Inc.). HCT116 human colon cancer cells were kindly gifted by Dr B. Vogelstein of the Johns Hopkins University School of Medicine (Baltimore, USA). IEC-6, which is a non-cancerous immortalized rat colon epithelial cell line, was purchased from the Chinese Academy of Sciences Beijing Cell Bank. The cells were maintained in the lab in low-glucose medium.

Cells were cultured in DMEM (low glucose, 5 mM) supplemented with 50 U/ml penicillin, 50 µg/ml streptomycin and 10% (v/v) fetal calf serum, and maintained at 37°C (5% CO₂) in a humidified incubator. Cells in the experimental group were treated with high glucose (15 mM), insulin (5 nM) and palmitic acid (150 µM). Cells treated for 48 h were used for the assessment of the centrosome amplification level; cells treated for 30 h were used for immunoprecipitation and western blot analyses. Palmitic acid, the most common saturated free fatty acid, is frequently used to investigate the biological activities of free fatty acids, particularly their adverse effects (27). Palmitic acid was conjugated to BSA; both reagents were mixed (molar ratio, 3:1) and incubated for 1 h before use.

Cells were seeded on a cover slip into a 6-well plate, at a density of 5×10⁴ cells per well. Following treatment with high glucose, insulin and palmitic acid, the cells were fixed in equal volumes of 100% cold methanol and 100% acetone (6 min; −20°C), washed three times with PBS (10 min each time), incubated with 0.1% Triton X-100 (15 min) and then with 3% BSA (60 min) at room temperature. The cells were subsequently incubated with antibodies against the following: γ-tubulin (cat. no. ab27074; mouse; Abcam; 1:500), ROCK1 (cat. no. 4035; rabbit; Cell Signaling Technology, Inc.; 1:500) and dynactin subunit 2 (DCTN2; cat. no. D122213; rabbit; Sangon Biotech Co., Ltd.; 1:1,000). Cells were washed twice with PBS, and incubated with anti-mouse fluorescent secondary antibody (cat. no. ab150116; Abcam; 1:500) and anti-rabbit fluorescent secondary antibody (cat. no. 4412; Goat; Cell Signaling Technology, Inc.; 1:1,000) in 3% BSA for 60 min at room temperature. Finally, the cells were mounted with mounting medium containing DAPI. Confocal microscopy was performed using a Zeiss LSM880 microscope (Carl Zeiss AG), and image processing was performed using Zen software (3.2; Carl Zeiss AG). The number of centrosome signals in each cell was determined by manually counting 300 cells in each specimen using a fluorescence microscope (magnification, ×1,000). Centrosome amplification was defined as cases with >5% of cells having ≥3 centrosomes per cell.

CoIP was performed under cold conditions (4°C or on ice) and ice-cold solutions were used, except in the sample heating step. A total of 5×10⁶ cells per sample were harvested and lysed in 500 µl precooled CoIP buffer (50 mM Tris/HCl pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% NP-40, 1 mM phenylmethylsulphonyl fluoride, 2 µg/ml leupeptin and 2 µg/ml pepstatin A) under non-denaturing conditions. After 24 h, cells seeded into 10-cm dishes were washed three times with ice cold PBS (5 min each time) and then lysed in 0.5 ml lysis buffer (Sangon Biotech Co., Ltd.) and collected in a 1.5-ml centrifuge tube, which was followed by a 30-min incubation on ice and 14,000 × g centrifugation step for 10 min. The supernatant was harvested and a volume equivalent to 200 µg total cellular protein was incubated with 20 µl Protein G Plus/Protein A agarose suspension (cat. no. IP05; EMD Millipore) for 2 h with gentle shaking. The agarose beads were removed by centrifugation, and the supernatant was collected and incubated overnight with primary antibodies against DCTN2 (cat. no. D122213; Sangon Biotech Co., Ltd.) or ROCK1 (cat. no. 4035; Cell Signaling Technology, Inc.), also with gentle shaking. Subsequently, 30 µl agarose beads were added to each sample, and incubated for 4 h with gentle shaking. Finally, the beads were collected by centrifugation, re-suspended in 2X loading buffer, heated to 100°C and then centrifuged once more (14,000 × g, 1 min). The heated sample was used for western blot analysis for the detection of the target proteins. Up to 40 µl sample was subjected to western blot analysis, as aforementioned. Whole cell lysate (40 µg) was used as the positive control for detection of the target proteins, and IgG was used as the negative control.

Samples from three different CoIP experiments were used for the proteomic identification of binding partners. Experiments were performed on a Q Exactive mass spectrometer linked to an Easy n-liquid chromatography (LC) system (Thermo Fisher Scientific, Inc.). In total, 6 µl per fraction was injected for nano-LC-mass spectrometry (MS)/MS analysis. The peptide mixture (5 µg) was loaded onto the C18-reversed phase column (Thermo Scientific Easy Column, 10 cm long, 75 µm inner diameter, 3 µm resin; Thermo Fisher Scientific, Inc.) in buffer A (0.1% formic acid), and was separated with a linear gradient of buffer B (80% acetonitrile and 0.1% formic acid) at a flow rate of 250 nl/min, which was controlled by an IntelliFlow device over 140 min. MS data were acquired using a data-dependent top 10 method, dynamically selecting the most abundant precursor ions from the survey scan (300–1800 m/z) for high-energy collision-induced dissociation fragmentation. Determination of the target value was based on predictive automatic gain control. The dynamic exclusion duration was 60 sec. Survey scans were acquired at a resolution of 70,000 at 200 m/z, and the resolution for the higher-energy collisional dissociation spectra was set to 17,500 at 200 m/z. The normalized collision energy was 30 eV and the underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 0.1%. The instrument was run with the peptide recognition mode enabled.

To determine the biological and functional properties of the identified proteins, Gene Ontology (GO) analysis was conducted by searching the GO database (http://www.geneontology.org). Functional category analysis was performed with protein2go and go2protein for annotation. Visualization and Integrated Discovery v6.7 (28) was used for functional enrichment analysis of GO terms. A false discovery rate <0.01 was selected as the cut-off criterion.

Cells were lysed in RIPA buffer (Sigma-Aldrich; Merck KGaA). The protein concentration in the samples was determined using a Bradford protein assay (Bio-Rad Laboratories, Inc.). Equal amounts of cell lysate (30 µg/lane) were separated via 10% polyacrylamide gel electrophoresis, and transferred onto PVDF membranes, which were blocked using 5% w/v non-fat milk containing TBS-Tween 20 (TBST; 0.05% v/v) for 1 h at room temperature. The membranes were incubated with antibodies against DCTN2 (cat. no. D122213; Sangon Biotech Co., Ltd., Shanghai, China) and ROCK1 (cat. no. 4035; Cell Signaling Technology, Inc.) overnight at 4°C, washed with TBST, and then incubated with anti-rabbit secondary antibody (cat. no. 7074; Cell Signaling Technology, Inc.) at room temperature for 1 h. β-actin (cat. no. D191047; Sangon Biotech Co., Ltd.; 1:1,000) was detected as a reference protein for equal sample loading. ECL reagents (cat. no. 32106; Thermo Fisher Scientific, Inc.) were used to visualize the protein bands, which were captured on X-ray films. Image Lab 5.0 software (Bio-Rad Laboratories, Inc.) was used for quantitative analysis.

Small interfering (si)RNA technology was used to knockdown protein expression. The pre-designed siRNA oligonucleotides were obtained from Sangon Biotech Co., Ltd.: ROCK1 forward, 5′-UGAUCUUGUAGCUCCCGCAUGUGUC-3′ and reverse, 5′-GACACAUGCGGGAGCUACAAGAUCA-3′; DCTN2 forward, 5′-GCACAAGUGUGGAACACAU-3′ and reverse, 5′-AUGUGUUCCACACUUGUGC-3′ and non-coding siRNA control, forward, 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse 5′-ACGUGACACGUUCGGAGAATT-3′. Cells (5×10⁴ per well) were seeded into a 6-well plate and cultured at 37°C for 24 h, and then were transfected with 200 nM siRNA oligonucleotides using Lipofectamine^® 2000 transfection reagent (Thermo Fisher Scientific, Inc.) at 37°C for 6 h, according to the manufacturer's instructions. Transfection efficiency was evaluated by western blot analysis 24 h after transfection.

RNA expression data of DCTN2 and ROCK1 in normal and colon cancer tissues were acquired in CPTAC from UALCAN (http://ualcan.path.uab.edu). Prognosis correlation data was acquired from the HUMAN PROTEIN ATLAS (https://www.proteinatlas.org). Best value was selected for prognosis analysis. Based on the FPKM value of each gene, patients were classified into two groups and association between prognosis (survival) and gene expression (FPKM) was examined. The best expression cut-off refers the FPKM value that yields maximal difference with regard to survival between the two groups at the lowest log-rank P-value.

Statistical analyses were performed using SPSS version 21 (IBM Corp.), and the data are expressed as the mean ± SD. Differences between two groups were assessed by Student's t-test, and multiple group comparisons were performed using ANOVA followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

HCT116 and IEC-6 cells were treated with high glucose, insulin and palmitic acid, which was found to increase centrosome amplification. Representative images of centrosomes and centrosome amplification in HCT116 and IEC-cells are displayed in Fig. 1A and B, respectively. Compared with the control samples, an increased percentage of centrosome amplification (~4.6- and 3.2-fold) was observed in treated HCT116 (5 vs. ~23%; P<0.01; Fig. 1C) and IEC-6 cells (5 vs. ~16%; P<0.01; Fig. 1D), respectively. The results confirmed that the pathophysiological factors of type 2 diabetes cause centrosome amplification.

Our previous study revealed that treatment with high glucose, insulin and palmitic acid increased the expression level and centrosomal translocation of ROCK1, which promotes centrosome amplification (15). To characterize ROCK1 binding partners, CoIP was performed using a ROCK1 antibody, and binding proteins were identified using MS. In total, 106 proteins were identified exclusively from the treated samples, 193 proteins exclusively from the control samples and 849 proteins from both the control and treated samples (Table SI).

GO clustering analysis identified the enrichment of these proteins in biological processes, molecular functions and cellular components. Within the biological process category, the majority of the proteins were involved in ‘cellular component organization’ or ‘biogenesis and cellular component organization’. For molecular function, the data indicated that most of the proteins were associated with ‘protein binding’, ‘heterocyclic compound binding’ and ‘organic cyclic compound binding’ (Fig. 2A). Moreover, most proteins were located in the ‘cytoplasm’ and ‘cytosol’ (Fig. 2A). The GO term analysis was also used to reveal specific meaningful terms, especially in the treatment condition. Notably, it was found that ‘protein binding’ in molecular function was significantly enriched in all of the identified proteins, and in those found only in the treated samples (Fig. 2B). The differentially expressed proteins that had not been shown to interact with ROCK1 were clustered into three complexes (Fig. 2C). Considering the centrosomal localization of ROCK1, proteins associated with centrosomal localization were investigated (Table I), from which DCTN2 was selected for further investigation. This protein was selected for analysis as it can interact with microtubules, which are associated with the centrosome. These analyses identified various candidate binding partners of ROCK1, which provided the background information for the study that confirmed the binding between ROCK1 and specific partners, signaling for the centrosome amplification.

Immunofluorescent staining was conducted to evaluate the centrosomal localization of DCTN2, as ROCK1 was already shown to be transported to the centrosome following treatment with high glucose, insulin and palmitic acid (15). As presented in Fig. 3A, DCTN2 was localized to only one centrosome in HCT116 cells at the early stage of cell division, and was then distributed between both centrosomes thereafter. DCTN2 also appeared in the spindles, particularly the spindle area closest to the centrosomes. When multipolar division occurred due to centrosome amplification, DCTN2 was present in multipolar spindles. It was also demonstrated that upon treatment, the protein levels of ROCK1 and DCTN2 were upregulated, which peaked at 12 h, and then declined at 24 h (Fig. 3B and C). The results confirm that DCTN2 is indeed a centrosomal protein.

CoIP was performed in combination with western blot analysis to evaluate binding between ROCK1 and DCTN2 in HCT116 and IEC-6 cells, and to investigate whether the experimental treatment enhanced this protein binding. Similar protein binding profiles were obtained from both the HCT116 and IEC-6 cell lines. An anti-ROCK1 antibody was able to pull down DCTN2, and the protein expression level of DCTN2 pulled down by the ROCK1 antibody was increased upon treatment with high glucose, insulin and palmitic acid (Fig. 4A and B). Similarly, the antibody against DCTN2 was able to pull down ROCK1, which was enhanced by the experimental treatment (Fig. 4C and D). The results indicated that high glucose, insulin and palmitic acid enhanced the binding between ROCK1 and DCTN2 in non-cancerous (IEC-6) and cancerous (HCT116) colon cells.

Next, the ROCK1-DCTN2 protein complex was disrupted by knocking down the expression of ROCK1 and DCTN2, and the effect on centrosome amplification in HCT116 cells was investigated. Knockdown of ROCK1 (Fig. 5A and B) or DCTN2 (Fig. 5C and D) partially inhibited the centrosome amplification triggered by the experimental treatment (Fig. 5E and F). Notably, simultaneous knockdown of these two proteins inhibited of centrosome amplification to a large degree (Fig. 5G and H). The results suggested that the binding of these proteins mediated diabetes-associated centrosome amplification.

The expression levels of ROCK1 and DCTN2 in colon cancer and their correlation with prognosis were analyzed using UALCAN and the HUMAN PROTEIN ATLAS. ROCK1 mRNA expression was higher in normal colon tissues compared with cancer tissues (P<0.05). However, ROCK1 expression was not associated with prognosis, and it did not predict the 5-year survival rate (P>0.05; Fig. 6A and B). Furthermore, the expression level of DCTN2 was lower in cancer tissues compared with normal colon tissues (P<0.01), and was not associated with prognosis (P>0.05; Fig. 6C and D). The expression levels of ROCK1 and DCTN2 were also analyzed using Clinical Proteomic Tumor Analysis Consortium, and were also decreased in cancer tissues in comparison with normal tissues (both P<0.001; Fig. 6E and F). It would be interesting to investigate whether binding between ROCK1 and DCTN2, rather than the expression levels, is associated colon cancer prognosis.