In leukemia, knock-down of the death inducer-obliterator gene would inhibit the proliferation of endothelial cells by inhibiting the expression of CDK6 and CCND1

Cell lines and cell culture

Human umbilical vein endothelial cells (HUVEC) and human myeloid leukemia cell line (K562) were purchased from the American Type Culture Collection (Rockville, MD). The cell lines were maintained using RPMI (Roswell Park Memorial Institute, Buffalo, NY, USA) 1,640 medium (Gibco Co, Waltham, MA, USA) supplemented with 10% FBS at 37 °C in a humidified atmosphere containing 5% CO₂. The HUVEC and K562 cell lines were mixed and co-cultured for 4 days. CCK-8 Cell Counting Kit-8 (CCK-8) was used to determine the cell viability by colorimetric assays at 450 nm (Tang et al., 2019).

For RNA extraction and analysis of cell proliferation and apoptosis, the suspension of K562 cells was removed. Then the HUVEC cells in the two groups (HUVEC vs. HUVEC-K562) were washed by PBS buffer and collected for further experiments. The HUVEC cells were digested with pancreatin before RNA extraction.

Plasmid constructs and transfection

For gene knockdown, the GV115 vector was used in this study, which used the green fluorescent protein as a reporter gene, and the multiple cloning sites were driven by a human U6 promoter. The DIDO was targeted by the shRNA sequences of 5′-GGATGAGACTCATTCAGAA- 3′. The sequence was cloned into the multiple cloning sites by restriction enzyme of AgeI and EcoRI. Plasmid transfection was performed as a former study (Nabzdyk et al., 2011). The cell lines were seeded into 96-well plates. After transfection for 2~3 d, the GFP was observed under a fluorescence microscope. The cells were used for further studies when the cell density in the wells reached 70–90%.

Celigo and MTT assays

Cells were inoculated into the 96-well plates (2,000 cells/well) and three repeats were taken. The cell numbers were measured by Celigo Imaging Cytometer (Nabzdyk et al., 2011) and the numbers were recorded for 5 days. The cell numbers were normalized to the cell numbers on the first day after seeding.

MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay (Moodley et al., 2014) was performed to analyze the proliferation of cells. A 1 mg/mL MTT was added to each well and incubated at 37 °C for 4 h. Then the culture medium was removed and the DMSO (150 μl) was added into each well, and then the plate was shaken for 3 min. The Tecan Infinite M2009PR plate reader was used to measure the absorbance at 490 nm/570 nm.

Cell apoptosis analysis

The cell apoptosis was measured following the manufacturer’s instructions of Annexin-FITC Apoptosis Detection Kit (BD Biosciences, Franklin Lake, NJ, USA). Cells were cultured in a 96-well plate for 3–5 days in the 37 °C incubator, and then the cells were harvested and washed in PBS. Cells were added to 0.5 ml binding buffer and Annexin V-FITC, then the cells were stained in the dark for 15 min at room temperature. Cells stained by Annexin V-FITC were considered apoptotic cells (Pan et al., 2014) which were measured by a BD Accuri™ C6 flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

To analysis the cell apoptosis, Caspase 3/7 enzyme activity was measured by Caspase-Glo® 3/7 Assay (Promega, G8091, Madison, WI, USA). Caspase-Glo 3/7 reagent was added to the sample with a volume ratio of 1:1, and the cells were incubated for another 1 h at 37 °C. The Tecan Infinite M2009PR plate reader was used to detect the luminescence in each well at 490 nm/520 nm (Stennicke & Salvesen, 1997).

Angiogenesis analysis

The serum-free supernatants of tumor cells from different experimental groups were collected and suspended the HUVEC cells to 2 × 10⁴ cells/100 uL. After being cultivated at 37 °C for 4–6 h, the angiogenesis assay was performed by the Celigo instrument.

Microarray processing and data analysis

The samples were hybridized with the GeneChip microarrays (901838; Affymetrix, Santa Clara, CA, USA) to determine gene expression abundance according to the manufacturer’s instructions. The expression profile was preprocessed by the Limma package in Bioconductor.

A robust multiarray averaging algorithm was used to perform background correction, quantile normalization, and probe summarization on the microarray data to obtain a gene expression matrix. The cut-off for the background correction was 20%, and the coefficient of variation was 25%. The Benjamini-Hochberg method was used to correct the significant difference level (FDR). The screening criteria for significantly different genes were: |Fold Change| > 1.5 and FDR < 0.05 (Ritchie et al., 2015). The biological pathways analysis of genes was performed by Ingenuity Pathway Analysis (IPA).

RNA extraction and qRT-PCR analysis

According to the manufacturer’s protocol, the Trizol reagent (Invitrogen, Waltham, MA, USA) was used to extract total RNA from frozen cells. For cDNA synthesis, 1 μg of total RNA was used to synthase the cDNA by the Go Script reverse transcription system (Promega, Maddison, MA, USA). The genes were detected by the SYBR Master Mixture (DRR041B; Takara, Kusatsu, Shiga, Japan) using the LightCycler480 Real-Time PCR system (Roche, Basel, Switzerland). For qRT-PCR, the GAPDH gene was used as endogenous control. The primers sequences and the length of the amplifications were shown in Table S1. The 2^−ΔΔCt method was used to calculate the fold change for gene expression relative to the control.

Protein extraction and Western-blot analysis

Total protein was isolated from cells using protein cell lysis buffer and extracted by centrifugation at 13000 rpm for 20 min at 4 °C. The equal amount of whole cell lysate was separated by SDS-PAGE gel electrophoresis. After the proteins were transferred to the PVDF membranes (Bio-Rad, Hercules, CA, USA), the membranes were blocked by 5% skimmed milk and immunoblotted with the primary antibodies at 4 °C. Then the membranes were blotted with the secondary antibodies at room temperature for 1 h. The following primary antibodies were used: anti-DIDO (1:1000, HPA049904, Sigma), anti-CCND1 (1:500, Cat2978, CST), anti-CDK6 (1:500, Cat3136, CST), anti-GAPDH (1:2000, Sc-32233, Santa Curz). The secondary antibodies were anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Dallas, TX, USA). The Dyne ECL STAR Western Blot Detection kit (Dyne Bio, Seoul, Korea) and a chemiluminescent image system (Fusion Solo system, Villber Lourmat, France) were used to analyze the protein abundance.

Statistical analysis

The data were shown as the mean ± S.D. from three independent replicates. The student’s t-test was performed to analyze the quantitative data. P < 0.05 was considered statistically significant.

GeneChip microarrays analysis of HUVEC and K562-HUVEC co-cultured cell lines

In this study, the human umbilical vein endothelial cells (HUVEC) were used as endothelial cells models in vitro. When the HUVEC cells were co-cultured with the K562 leukemia cell lines for 4 days, the proliferation was inhibited significantly (Fig. S1). Then we analyzed the gene expression changes in HUVEC cells when co-cultured with the K562 leukemia cell lines by GeneChip, to investigate the genes which will inhibit the endothelial cells’ proliferation in leukemia progression.

Compared with HUVEC lines, 398 probes up-regulated expression and 323 probes down-regulated expression in K562 co-cultured HUVEC lines (Fig. S2A). Ingenuity Pathway Analysis (IPA) found that the extracellular signal-regulated kinase 5 (ERK5) signaling was significantly inhibited (Z-score = −2.111) (Fig. S2B). On the other hand, the DEGs were mainly enriched in microtubule dynamics (Z-score = 2.783), migration of brain cancer cell lines (Z-score = 2.549), liver tumor (Z-score = −2.782) and cell death of mononuclear leukocytes (Z-score = −2.561) (Fig. S2C).

Construction of RNAi cell lines and cell proliferation analysis

We selected the first 30 down-regulated expression genes (log2 (change fold) > 1, P < 0.05) for further analysis (Table S2). RNAi lentiviral vectors for these 30 genes were constructed and transfected into HUVEC cells. A total of 22 transgenic cell lines were successfully obtained, including the negative control (NC) and positive control (PC). The cell count results showed that the proliferation of cells was normal in the NC group, and which was significantly inhibited in the PC group. The proliferation folds on the fifth day were 12.09 and 2.12 times higher than those on the first day in the NC and PC groups, respectively. The fold change (FC) of cell count ([FC in NC group on the 5th day compared to which on the 1st day]/[FC in experiment group on the 5th day compared to which on the 1st day]) was used to evaluate the influence of gene RNAi in cell proliferation. The proliferation of shDIDO, shZC3H18, and shSMURF2 cell lines was significantly inhibited, and the change fold was 3.37, 2.54, 2.07, respectively (Fig. S3).

Patients with lower transcript abundance of DIDO showed a better overall survival

To determine the effect of the DIDO, ZC3H18, and SMURF2 in leukemia patients, we analyzed the survival based on their expression status from the public data (http://gepia2.cancer-pku.cn/). As shown in Fig. 1A, the acute myeloid leukemia (AML) patients with the lower DIDO expression level, showed a better overall survival (HR = 1.9; P = 0.025). However, the different expressions of ZC3H18, and SMURF2 did not affect the overall survival in AML patients (Figs. 1B and 1C).

The proliferation of shDIDO cell line is inhibited and the apoptosis is increased

To further investigate the function of the DIDO (Death inducer obliterator) gene in endothelial cells, we analyzed the proliferation and apoptosis of shDIDO cells. Firstly, the expression of DIDO in shDIDO cells was analyzed. qRT-PCR found that the expression level of the DIDO gene at the mRNA level was suppressed in shDIDO cell lines (P < 0.05), and the reduction efficiency reached 95.1% (Fig. 2A). Western-blot detection found that the expression of DIDO protein in the shDIDO cells decreased by four times, compared with the shCtrl group (Fig. 2B).

The proliferation rate of shDIDO cell line was analyzed by Celigo (Figs. 2C and 2D) and MTT (Fig. 2E), and the proliferation rate of the shDIDO cells was significantly decreased. This may indicate that the DIDO gene is significantly related to the proliferation ability of HUVEC cells.

The number of cells in the apoptotic state was detected by Annexin V-APC single staining method, and it was found that apoptosis cells in shDIDO group increased significantly than the HUVEC cells (P < 0.05) after 5 days (Fig. 2G). Additionally, by detecting the activity of caspase, it was found that the activity of caspase3/7 in the shDIDO group was significantly increased. These results indicate that the DIDO gene was significantly related to the apoptosis of HUVEC cells (Fig. 2H).

Due to the importance of angiogenesis in tumor progression, we analyzed the effect of DIDO gene depletion on angiogenesis. The ability of shDIDO cells to form lumens was analyzed to investigate the metastasis ability of tumors. It was found that the area of angiogenesis-related blood vessels in the shDIDO group were 20% less than that in the shCtrl group (P < 0.05) (Fig. 2F), which indicates that the DIDO gene may not associated with HUVEC cells angiogenesis.

GeneChip analysis of shDIDO and shCtrl cell lines

To investigate the biological pathways DIDO involved, the GeneChip expression profiles of shCtrl and shDIDO cell lines were analyzed. It was found that 521 genes in the shDIDO cells were up-regulated and 1,006 genes were down-regulated (Fold Change > 1.5 and FDR < 0.05), compared with shCtrl cells (Fig. 3A, Table S3). IPA analysis found that the ERK/MAPK signaling was significantly inhibited (Z-score = −2.041) (Fig. 3B). Additionally, the functions including morbidity or mortality (Z-score = 6.734) and organismal death (Z-score = 6.709), were significantly activated. The functions including cell viability (Z-score = −5.369), cell survival (Z-score = −5.349) were significantly suppressed (Fig. 3C and Table S4).

The genes down-regulated in shDIDO cells, and which are involved in tumorigenesis and development, were selected for further analysis (Table S5). Among them, 30 probes were identified by qRT-PCR as their expression pattern were similar to the GeneChip (Table S6).

DIDO is a part of the centrosome protein and plays an important role in spindle assembly, so we infer that the DIDO gene may correlate with the cell cycle. We further analyzed the cell cycle regulation genes of Cyclin Dependent Kinase 6 (CDK6) and Cyclin D1 (CCND1). The transcription of CCND1 and CDK6 in the shDIDO cells were 0.364 and 0.404 times of the shCtrl group, respectively. Meanwhile, the western-blot analysis found that the protein expression levels of CCND1 and CDK6 in shDIDO cells were reduced by 81.3% and 58.1%, respectively. However, the DIDO maybe not interact with CDK6 or CCND1 directly, as shown in Fig. 3D. We searched the protein interaction database and did not find the direct interaction of DIDO with CDK6 or CCND1 either. The genes that may directly interact with DIDO were SRSF1, SRPK2, EED, and WWP2 in the interaction network of DIDO analyzed by IPA.