HAUS6 expression in CRC and noncancerous colorectal tissues was compared using the Oncomine database (www.oncomine.org), as described previously (Shen et al., 2019) (Supplementary Table S1). Correlation of HAUS6 expression with CRC patient survival was performed using the R2 Bioinformatic Platform (http://r2.amc.nl), as described previously (Shen et al., 2019). The threshold of significance was set at p < 0.05.

Both CRC and matched adjacent non-cancerous tissues were collected from 36 CRC patients admitted to the First People’s Hospital Affiliated with Fujian University of Traditional Chinese Medicine (FJTCM). Clinicopathologic characteristics of patients are summarized in Supplementary Table S2. No patients received radio- or chemotherapy prior to surgery. Tissues were snap-frozen and stored in liquid nitrogen for quantitative RT-PCR or processed for immunohistochemistry. Samples were stained with hematoxylin and eosin, and then examined by experienced pathologists, who assigned pathologic tumor-node-metastasis (p-TNM) grades according to International Union Against Cancer guidelines.

In addition, cDNA from 80 primary CRC and 15 adjacent noncancerous colorectal tissues were obtained from the cDNA-HColA095Su01 tissue cDNA array (Shanghai Outdo Biotech, Shanghai, China; Supplementary Table S3). A further 280 primary CRC and 260 adjacent noncancerous colorectal tissues were obtained from the tissue microarray (TMA) (Shanghai Outdo Biotech; Supplementary Table S4).

Cells and tissue samples were lysed with RNAiso Plus reagent (Takara, Beijing, China) and total RNA was extracted according to the manufacturer’s instructions. A total of 1 μg RNA was reverse-transcribed into cDNA using PrimeScript RT reagent kit (Takara, Beijing) and RT-qPCR was performed using SYBR Premix Ex Taq (Takara, Beijing) and the ABI 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA, United States) according to manufacturer protocols. Primer sequences against HAUS6, TP53, CDKN1A and GAPDH are shown in Supplementary Table S5. GAPDH was used as an internal control.

Tissues were processed using routine immunohistochemistry methods as in our previous study (Shen et al., 2019). Tissue specimens collected in our laboratory were fixed, sectioned and stained with primary antibodies against HAUS6 and Ki-67 then secondary biotinylated antibody (Supplementary Table S6), followed by incubation with horseradish peroxidase-conjugated streptavidin (Maixing, Fuzhou, China). Background staining was assessed by omitting the primary antibody. Five randomly selected fields were examined using a Leica microscope (Wetzlar, Germany) at ×40 magnification.

For tissue arrays, slides were incubated with primary antibody against HAUS6 (dilution 1:500, Abcam, United Kingdom) as described (Wang et al., 2014; Shen et al., 2019). Images were captured using a Nano Zoomer 2.0 HT slide scanner (Hamamatsu Photonics, Hamamatsu, Japan) and processed using Nano Zoomer Digital Pathology View 1.6 software. Staining intensity and the percentage of positively stained cells were determined as described (Huang et al., 2016; Shen et al., 2019). Protein expression was calculated by multiplying the intensity and percentage scores together, giving expression scores ranging from 0 to 12.

The median HAUS6 expression in CRC tissues was used as the cutoff to divide tumors into “low expression” or “high expression”. The relationship between HAUS6 expression and overall survival was analyzed using Kaplan-Meier curves and the log-rank test.

Human CRC cell lines HCT116, HCT-8, HT-29, Caco2 and RKO were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Human colon cell line FHC (ATCC No. CRL-1831) was purchased from the American Type Culture Collection (Maryland, United States). Wild-type HCT116 cells (HCT116/p53^+/+) and p53 knockout HCT116 cells (HCT116/p53^−/−) were a gift from Dr. Yao Lin (Fujian Normal University, Fujian, China) and originally obtained from Dr Bert Vogelstein (Johns Hopkins University, Baltimore, MD). All human cell lines have been authenticated using STR (or SNP) profiling within the last 3 years has been included. RKO, HCT116, HCT-8, Wild-type HCT116 cells (HCT116/p53^+/+) and p53 knockout HCT116 cells (HCT116/p53^−/−) cells were maintained in RPMI1640 (Thermo Fisher Scientific); HT-29 cell in M5’A medium (KeyGEN); Caco2 cells in DMEM (Thermo Fisher Scientific); and FHC cells in DMEM supplemented with 10 mM HEPES (for a final conc. of 25 mM), 10 ng/ml cholera toxin, 0.005 mg/ml insulin, 0.005 mg/ml transferring and 100 ng/ml hydrocortisone. All media contained 10% FBS (Thermo Fisher Scientific), 100 U/ml penicillin, and 100 mg/ml streptomycin (Hyclone) at 37°C in a humidified atmosphere of 5% CO₂. Cells were verified using short tandem repeat genotyping and examined for mycoplasma contamination using RT-PCR.

5-FU (Sigma-Aldrich) was dissolved in DMSO and final concentration of 2.5 μM used to treat control and HAUS6 knockdown cells for 24h; DMSO was used as a vehicle control.

Lentiviruses encoding shRNA against HAUS6 or CDKN1A or encoding control shRNA (Supplementary Table S7 for sequences) and enhanced green fluorescent protein (EGFP) under control of the CMV promoter were obtained from Shanghai GeneChem (Shanghai, China). CRC cells were transduced with lentivirus at a multiplicity of infection (MOI) of 10 for 72 h before analysis. To examine effects of HAUS6 over-expression on cell proliferation, HT-29 cells were transduced for 72 h with a lentiviral vector (MOI = 10) encoding full-length human HAUS6 (2,946 nt, GenBank accession NM_017645; Shanghai GeneChem). Transductions were selected for 2 weeks using 1 μg/ml puromycin.

Cell viability was measured every day for 5 days by CCK-8 assay. Briefly, transduced cells were reseeded into 96-well plates at a density of 2,000 cells/well with 100 μl medium/well. CCK-8 dye (10 μl, Cell Counting Kit-8, Abbkine, Wuhan, Hebei, China) was added to each well and cells were incubated for 2 h at 37°C. Cell viability was measured at 450 nm.

Transduced cells were reseeded into 12-well or 6 well plates at a density of 500 or 1,000 cells/well and incubated for 10–12 days at 37°C under 5% CO₂ in a humidified incubator. Colonies were fixed with 4% paraformaldehyde for 20 min and stained with 0.1% of crystal violet for 20 min. Colonies were photographed and the number of colonies counted. Data were normalized relative to control cells.

Transduced cells were collected and fixed with cold 70% ethanol at 4°C overnight. The cells were then washed three times with PBS and incubated with FxCycle PI/RNase Staining Solution (Thermo Fisher Scientific) for 30 min. Fluorescence-activated cell sorting was performed using fluorescence activated cell sorting (FACS) Caliber (Becton Dickinson, San Jose, CA, United States), and ModfitLT version 3.0 (Verity Software House, Topsham, ME, United States) was used to analyze the cell cycle based on DNA content in different phases.

Male pathogen-free athymic nude mice (6–8 weeks old) were purchased from Shanghai Laboratory Animal Co., Ltd. (Shanghai, China) and maintained in a specific pathogen-free (SPF) facility.

Tranduced CRC cells (1 × 10⁶) in 100 µl RPMI-1640 medium containing 50% matrigel were subcutaneously injected into the flank of nude mice (n = 8). Tumor growth was monitored from day 7 after injection on every second day for 19 days. Tumor volume was calculated as (1/2) (length × width²), where length and width refer to the longest longitudinal and transverse diameters.

Tumor growth was also measured in terms of GFP intensity using an IVIS Spectrum whole live-animal imaging system (PerkinElmer, Waltham, MA, United States). Mice were anesthetized with isoflurane and sacrificed by neck dislocation on day 26 after injection, and tumors were dissected and weighed. Collected tissues were processed for immunohistochemistry.

HCT116 cells were transduced with lentivirus encoding shRNA specific for HAUS6 or control shRNA for 72 h. Briefly, total RNA was extracted with Trizol reagent (Thermo Fisher Scientific) and analyzed by CapitalBio (Beijing, China) using the human GeneChip Primeview array (Affymetrix, Santa Clara, CA, United States) according to the manufacturer’s instructions. Microarray images were scanned using GeneChip Scanner 3,000 and analyzed using GeneChip GCOS 1.4 software (Affymetrix). Genes with a p-value < 0.05 and fold change >2 were defined as differentially expressed. The raw microarray data can be found in the NCBI Gene Expression Omnibus (GEO: GSE140326). Enriched pathways in DEGs were identified using KEGG pathway enrichment analysis.

Cells were lysed with RIPA lysis buffer (Thermo Fisher Scientific) and the protein concentration was determined using BCA assay (Thermo Fisher Scientific). The proteins (50 µg) were then separated by SDS-PAGE and transferred onto a nitrocellulose membrane. The membranes were blocked with BSA (Sigma Aldrich) and incubated overnight at 4°C with primary antibody against HAUS6, p53, or p21 (Supplementary Table S6). The membranes were then washed three times with TBST (5 min per wash), and incubated at room temperature for 1 h with horseradish peroxidase-conjugated goat anti-rabbit (CST, United States). The blots were visualized using enhanced chemiluminescence (Thermo Fisher Scientific) according to the manufacturer’s protocols. Band intensities were quantified using ImageLab software. The expression of GAPDH was used as a control. Three independent experiments were performed for each assay.

Statistical analysis was performed using SPSS 20.0 (IBM, Chicago, IL, United States). Data are presented as mean ± SD. Differences between two groups were assessed using the independent or paired Student’s t test, while differences among three or more groups were assessed using one-way ANOVA. Survival differences were assessed using the log-rank test. The correlation between HAUS6 and CDKN1A expression was analyzed using Pearson’s correlation. A p-value of <0.05 was considered statistically significant.

We first compared HAUS6 expression in cDNA samples from 14 matched pairs of primary CRC tumors and non-cancerous tissues from a cDNA array (GEO Submission: GSE113513). HAUS6 expression was higher in CRC tumors than non-cancerous tissues (Figure 1A). Analysis of publicly available tumor expression data from the Oncomine database (http://www.oncomine.com/) also showed higher HAUS6 mRNA levels in CRC tissues than non-cancerous tissues (Supplementary Table S1). Quantitative RT-PCR analysis of HAUS6 expression in clinical samples of 36 CRC tissues and adjacent non-cancerous colorectal tissues from our laboratory and a cDNA tissue array containing 15 pairs of CRC primary tumors and surrounding non-cancerous tissues also showed higher HAUS6 mRNA expression in CRC tissues (Figures 1B,C). Immunohistochemistry of 10 pairs of CRC samples randomly selected from the above mentioned 36 pairs, as well as analysis of a tissue microarray of 260 pairs of CRC samples (Shanghai Outdo Biotech), confirmed that HAUS6 protein levels were higher in CRC primary tumor tissues (Figures 1D,E).

Kaplan-Meier analysis of 80 patients whose primary tumor tissues were sampled on the cDNA array mentioned above did not reveal any difference in patient survival between tumors with low and high HAUS6 mRNA expression (data not shown). In contrast, analysis of data from 194 CRC patients (high expression: n = 79; low expression: n = 115) from the public microarray dataset using the R2 bioinformatic platform showed that high HAUS6 mRNA expression was associated with shorter overall survival (Figure 2A). Similarly, tumors with high HAUS6 protein levels were associated with shorter overall survival in 280 CRC patients (Figure 2B; high expression: n = 36; low expression: n = 244). Representative images of high or low expression of HAUS6 in CRC tissues are shown in Figure 2C.

In cultured RKO and HCT116 cells, which constitutively express relatively high levels of endogenous HAUS6 (Supplementary Figure S1A), we assessed the effects of HAUS6 knockdown on growth and proliferation of CRC cells. Lentiviral transduction with vectors expressing short hairpin (sh)RNAs against HAUS6 down-regulated HAUS6, as expected (Figures 3A,B; Supplementary Figures S1B–C). CRC cells transduced with HAUS6 shRNAs had lower cell viability than those transduced with control RNAs (Figure 3C; Supplementary Figure S1D). Moreover, HAUS6 knockdown also decreased the number of colonies formed by CRC cells (Figure 3D). Cell cycle analysis revealed that HAUS6 knockdown increased the percentage of cells in G0/G1 and decreased the percentage of cells in S phase (Figure 3E). However, HAUS6 knockdown did not affect cell apoptosis in HCT116 cells (Supplementary Figure S2). In contrast, in HT-29 cells which exhibits relative lower expression of HAUS6 (Supplementary Figure S1A), lentiviral transduction with vectors over-expressing HAUS6 up-regulated HAUS6 protein expression, and increased the cell viability of HT-29 cells based on the CCK8 assay (Figures 3F–H).

We next used a xenograft mouse model to assess the effects of HAUS6 knockdown on CRC cell growth and proliferation in vivo. HAUS6 knockdown strongly reduced tumor volume (Figure 4A) and decreased fluorescence of tumor cells (Figure 4B) in nude mice injected with GFP-expressing CRC cells. Final tumor weights were also lower in mice injected with HAUS6 knockdown cells than in those injected with control cells (Figure 4C). Immunohistochemistry of the excised tumors showed that Ki-67 protein expression was decreased in HAUS6 knockdown cells (Figure 4D), suggesting that HAUS6 knockdown suppresses CRC tumor growth by inhibiting cell proliferation.

Microarray analysis of HAUS6 knockdown HCT116 cells revealed a total of 274 differentially expressed genes (DEGs) in CRC (Figures 5A,B; GEO Submission: GSE140326). CDKN1A (encoding p21 protein) was upregulated in HAUS6 knockdown cells (Figure 5C), and this was confirmed in both HCT116 and RKO cells by quantitative RT-PCR (Figure 5D) and western blotting (Figure 5E). Transduction with shRNAs specific for HAUS6 significantly up-regulated p21 protein expression, which were reversed after CDKN1A knockdown in HCT116 cells (Figure 5F). Functionally, HAUS6 knockdown decreased cell viability and colony formation, whereas CDKN1A knockdown had the opposite effect and attenuated the effects of HAUS6 knockdown (Figures 5G,H). HAUS6 knockdown increased the percentage of cells in G0/G1 and reduced the percentage in S phase, which were reversed by CDKN1A knockdown (Figure 5I). Analysis of microarray expression data (GEO Submission: GSE113513) showed that CDKN1A was down-regulated in CRC tissues (Figure 5J) and its expression correlated negatively with HAUS6 expression (Figure 5K).

KEGG pathway enrichment analysis of DEGs in HAUS6 knockdown cells identified the p53 pathway as the one of the most enriched signalling pathways (Figure 6A). Given the key regulatory effects of p53 on cell proliferation and CDKN1A transcription, we next explored the effect of HAUS6 on the p53 pathway. HAUS6 knockdown increased TP53 expression in HCT116 and RKO cells at the levels of mRNA (Figures 6B,C) and protein (Figure 6D). The inhibitory effects of HAUS6 knockdown on cell viability were attenuated by knockout of p53 (Figures 6E,F) or treatment with the p53 inhibitor PFT-α (Figure 6G). Similarly, p53 knockout blocked the inhibitory effects of HAUS6 knockdown on cell survival (Figure 6H) and cell cycle progression (Figures 6I,J). To determine the molecular mechanism by which HAUS6 knockdown regulates p53, we evaluated p53 protein stability in RKO cells expressing wild-type p53. Protein levels of p53 and its downstream p21 decreased over time in sh-Ctrl cells, but remained stable after HAUS6 knockdown (Figure 6K).

We further assessed the effect of HAUS6 on the response of CRC cells to 2.5 μM 5-fluorouracil (5-FU) treatment for 24 h. Transduction of sh-HAUS6 or 5-FU treatment down-regulated HAUS6 and up-regulated p53 and p21 protein expression. These effects of HAUS6 knockdown were further enhanced after combined with 5-FU treatment (Figure 7A). Moreover, HAUS6 knockdown or 5-FU treatment reduced viability of HCT116 cells, and combination of them exhibited greater suppression on cell viability of HCT116 cells at different time points (Figure 7B). Moreover, combining HAUS6 knockdown with 5-FU treatment led to greater suppression of HCT116 colony formation than HAUS6 knockdown or 5-FU treatment alone (Figure 7C), as well as higher percentages of cells in G0/G1 and G2/M and lower percentage in S phase (Figures 7D,E).