You are here

Article Text

Article menu

Download PDFPDF

Colonic cancer
Human Sgo1 downregulation leads to chromosomal instability in colorectal cancer
  1. M Iwaizumi1,2,
  2. K Shinmura1,
  3. H Mori1,
  4. H Yamada1,
  5. M Suzuki1,
  6. Y Kitayama1,3,
  7. H Igarashi1,
  8. T Nakamura4,
  9. H Suzuki5,
  10. Y Watanabe6,
  11. A Hishida2,
  12. M Ikuma2,
  13. H Sugimura1
  1. 1
    First Department of Pathology, Hamamatsu University School of Medicine, Higashi-ku, Hamamatsu, Shizuoka, Japan
  2. 2
    First Department of Medicine, Hamamatsu University School of Medicine, Higashi-ku, Hamamatsu, Shizuoka, Japan
  3. 3
    Department of Pathology, International University of Health and Welfare, Mita Hospital, Minato-ku, Tokyo, Japan
  4. 4
    Second Department of Surgery, Hamamatsu University School of Medicine, Higashi-ku, Hamamatsu, Shizuoka, Japan
  5. 5
    Department of Gene Therapy, Institute of DNA Medicine, Jikei University School of Medicine, Minato-ku, Tokyo, Japan
  6. 6
    Laboratory of Chromosome Dynamics, Institute of Molecular and Cellular Biosciences, University of Tokyo, Bunkyo-ku, Tokyo, Japan
  1. Dr H Sugimura, First Department of Pathology, Hamamatsu University School of Medicine, 1-20-1 Handayama, Higashi-ku, Hamamatsu, Shizuoka, 431-3192, Japan; hsugimur{at}hama-med.ac.jp

Abstract

Background and aims: Chromosomal instability (CIN) is recognised as a hallmark of cancer and is caused by a spindle assembly checkpoint disorder or chromosome mis-segregation during mitosis. Although the recent identification of human shugoshin (hSgo1), an important player in proper chromosome segregation, has suggested the involvement of hSgo1 in colorectal tumourigenesis, little is known about how it is involved. The aim of this study was to obtain information about the status of hSgo1 in human colorectal cancer.

Method and results: Among the 46 colorectal cancer cases, hSgo1 mRNA expression was decreased in the tumour tissue in comparison with the corresponding normal tissue (p = 0.032). Human Sgo1-downregulated tumours (tumour to normal mucosa ratio<0.5) had preferential location on the left side large bowel rather than on the right side (p = 0.012), and a higher variation of centromere numbers revealed by fluorescence in situ hybridisation (FISH). To assess the effects of hSgo1 downregulation, hSgo1 knockdown was performed by transfecting the diploid HCT116 cell line with a short hairpin RNA expression vector. hSgo1 knockdown cells proliferated slowly because of both G2/M arrest and apoptosis (p<0.001), and markers of CIN in the form of aneuploidy (p<0.001) and micronuclei (p<0.005) were later observed in hSgo1 knockdown cells. Increased centrosome amplification (p<0.05), the presence of binucleated cells and mitotic catastrophes were also noted in hSgo1 knockdown cells.

Conclusions: These findings suggest that hSgo1-downregulated colorectal cancers have a clinicopathological character of CIN, and hSgo1 downregulation leads to CIN in colorectal cancer cells.

https://doi.org/10.1136/gut.2008.149468

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    In colorectal cancer progression, two forms of genomic instability have been identified: microsatellite instability (MIN) and chromosomal instability (CIN).1 2 MIN is generally due to inactivation of mismatch repair (MMR) genes such as MSH2 and MLH1,3 and about 15% of sporadic colorectal cancers show some form of MIN.4 On the other hand, about 85% of colorectal cancers display recurrent and tumour-specific chromosome alterations.5 The identification of specific patterns of chromosome gains and losses during tumour progression is consistent with the idea that CIN is an aetiological factor in colorectal cancer, but the definitive molecular basis of CIN remains elusive. Some of the culprits for chromosome alterations in general may be mutations of spindle assembly checkpoint genes. Moreover, previous data have suggested that chromosome mis-segregation during mitosis is the main cause of aneuploidy in cancer cells, and that it contributes to oncogenesis.6

    In 2004, shugoshin (Sgo) protein was identified in budding and fission yeasts as a protector of centromeric cohesion in meiosis.7 Sgo is conserved among eukaryotic organisms including vertebrates. One of the two human Sgo proteins (hSgo1 and hSgo2), hSgo1 (also called shugoshin-like 1 (SGOL1)) has been shown to be essential to accurate chromosome segregation during both mitosis and meiosis.8 9 Coordinated chromosome segregation also depends on the function of Polo-like kinase and Aurora B kinase,10 11 which are overexpressed in human tumours,12–14 and of BUB1,15 which is downregulated in colorectal cancer.16 These findings concerning hSgo1 and the other chromosome segregation-regulating proteins suggest that hSgo1 may also participate in the pathogenesis of CIN in tumours, but the mechanism has never been investigated in detail.

    In this study, we searched for the clinicopathological significance of hSgo1 alterations in human colorectal cancer cases and analysed the effects of hSgo1 alterations in a human cell line.

    MATERIALS AND METHODS

    Tissue samples and cell lines

    Tissue from cases of primary colorectal cancer was surgically resected at the Hamamatsu University School of Medicine, University Hospital. The resected tissue was excised within 30 min after the operation from tumour tissue and corresponding non-tumour tissue >5 cm from the tumour. The colorectal cell line used in this study was HCT116.17 The cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and penicillin–streptomycin (100 μg/ml and 100 U/ml, respectively), and they were maintained at 37°C in a 5% CO2/95% air incubator. The study design was approved by the Institutional Review Board (IRB) of Hamamatsu University School of Medicine.

    Microdissection

    Frozen sections of the colorectal cancer tissue were stained with H&E. Cancerous cell clusters in stained sections were examined and microdissected under an inverted microscope (Olympus IX71; Olympus, Tokyo, Japan) equipped with a microdissection device (MicroDissector; Eppendorf AG, Hamburg, Germany). Areas of interest were microdissected with an ultrasonically oscillating needle, and the tissue particles obtained were aspirated into a pipette tip with a piezo-driven micropipette (fig 1A). The microdissected materials were used for DNA or RNA extraction.

    Figure 1
    Figure 1 Loss of heterozygosity (LOH) analysis at the hSgo1 gene locus on 3p24 in colorectal cancer samples. (A) Microdissection of colon cancer tissue in H&E-stained sections. The histological appearance of colon cancer is shown. The middle panel is a higher magnification of the boxed area in the top panel. The bottom panel shows the section after that in the middle panel and microdissection under an inverted microscope (Olympus IX71) equipped with a microdissection device (MicroDissector). Scale bars: 100 μm. (B–D) Detection of LOH with polymorphic markers (B, c.416+39_42delGAAA; C, c.504A>T; D, c.1461C>T) by PCR-SSCP (single strand conformational polymorphism) and direct sequencing. Tumour, tumour tissue; normal, adjacent normal tissue. (E) Summary of informative colorectal cancer samples with LOH at the hSgo1 gene locus on 3p24. Filled circles, informative and retention of heterozygosity; open circles, informative and LOH; NI, non-informative.

    DNA isolation, PCR followed by direct sequencing, PCR-single strand conformation polymorphism (SSCP)

    Genomic DNA was isolated by standard procedures.18 For mutation and loss of heterozygosity (LOH) analysis, we performed PCR-SSCP using 10 primer sets displayed in table 1, and the PCR products that produced an abnormally shifted band in the SSCP analysis were directly sequenced according to the method described previously.19

    View this table:
    Table 1 Primers used for PCR amplification of hSgo1 exons

    RNA extraction and quantitative real-time reverse transcription (qRT)-PCR

    Total RNA was extracted from tumours and normal tissues by using the commercially available ISOGEN kit (Nippongene, Tokyo, Japan) according to the manufacturer’s instructions. qRT-PCR was performed as previously described.20 cDNA was synthesised from total RNA with a SuperScript II First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, California, USA). The cDNA was used as a template to perform qRT-PCR of hSgo1 cDNA with a QuantiTect SYBR Green PCR kit (QIAGEN, Hilden, Germany). The primers used were 5'-GAC CCC AAT AGT GAT GAC AGC-3' and 5'-GAA ATG ATT CTC CTT GTC CTG G-3' for the hSgo1 transcript, and 5'-AGG TGA AGG TCG GAG TCA AC-3' and 5'-CCA TGG GTG GAA TCA TAT TG-3' for the transcript of the control housekeeping gene, the gene that encodes glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The relative amounts of the hSgo1 transcript in the tumour tissue (T) and the normal tissue (N) were standardised according to those of the GAPDH transcript. The tumour to normal mucosa (T/N) ratio was calculated in each case by using the standardised relative amounts of hSgo1 transcript in the tumour tissue and normal tissue.

    Protein extraction and western blotting

    Protein was extracted from H&E-stained frozen sections of tumour tissue in which 90% of the tissue was occupied by clusters of cancer cells. The frozen tissue and cultured cells were solubilised in lysis buffer (10 mM N-2-hydroxyethylpiperazine-N′-ethanesulfonic acid (HEPES)-KOH (pH 7.5), Nonidet P-40, 1 mM dithiothreitol (DTT), 0.5 mM EDTA, 1.5 mg/ml phenylmethylsulfonyl fluoride (PMSF)) with an ultrasonic processor (Hielscher, Teltow, Germany) on ice and then centrifuged (100 000 g) for 15 min at 4°C. The supernatant was mixed with 2× SDS sample buffer (125 mmol/l Tris–HCl (pH 6.8), 4% SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.2% bromophenol blue) adjusted to 2.5 μg of protein/μl and heated for 10 min at 98°C. The proteins mixed with 2× SDS sumple buffer were separated by electrophoresis on 12.5% SDS–polyacrylamide gels, and the proteins in the gels were transferred to polyvinylidene fluoride membranes (GE Healthcare Bio-Sciences, Fairfield, California, USA) in a transfer apparatus (Atto, Tokyo, Japan). The membranes were blocked with 1% skim milk and 0.1% Tween in phosphate-buffered saline (PBS). Monoclonal anti-hSgo1 antibody (13G1) raised by Dr Suzuki of Jikei University21 and diluted 1:500 in blocking buffer, antiphospho-histone H3 (Ser10) antibody (Millipore, Billerica, Massachusettts, USA) diluted 1:500 in bocking buffer, and mouse anti-β-tubulin antibody (GE Healthcare Bio-Sciences) diluted 1:1000 in blocking buffer were used as the primary antibodies. Sheep anti-mouse immunoglobulin G–horseradish peroxidase (HRP)-linked F(ab′)2 fragment diluted 1:5000 in blocking buffer was used as the secondary antibody. Antigens on the membrane were detected with enhanced chemiluminescence detection reagents (GE Healthcare Bio-Sciences).

    Immunostaining of hSgo1 and Ki-67

    Tissues were fixed in 10% formalin. The details of the immunostaining procedure have been described elsewhere.22 In brief, after incubation for 1 h with the anti-hSgo1 mouse monoclonal antibody (13G1) diluted 1:50 or anti-Ki-67 mouse monoclonal antibody (MIB1, DAKO, Copenhagen, Denmark) diluted 1:50, polymer reagent (dextran backbone to which a large number of peroxidase molecules and secondary antibody molecules have been coupled) (ChemoMate ENVISION kit; DAKO) was added. Colorisation development was performed with 0.2 mg/ml 3-3'-diaminobenzidine tetrahydrochloride (DAB) (DAKO) and 0.3% H2O2. Haematoxylin was used for nuclear counterstaining.

    Fluorescence in situ hybridisation (FISH)

    For FISH analysis, the microwave-assisted FISH protocol was performed as described previously.23 The centromere enumeration probes (CEPs) were used to identify the numbers of chromosomes 3, 12, 17 and 18 (CEP3, 12, 17 and 18; Vysis, Abbott Park, Illinois, USA, respectively). The number of signals per cell was counted in >200 cell nuclei. The modal signal number was defined as the number of the chromosomes most prevalent in tumour cells, and an abnormal signal number was defined as any number of signals other than the modal number.

    Construction of the short hairpin RNA (shRNA) expression vector, and transfection

    The hSgo1 shRNA expression vector was constructed by inserting shRNA sequences targeting hSgo1 (target sequence: 5'-GTG AAA GAA GCC CAA GAT A-3') into pSilencer 2.1-U6 puro (Applied Biosystems, Foster City, California, USA). The negative control pSilencer 2.1-U6 puro vector expresses a shRNA whose sequence is not found in the human genome databases (target sequence: 5'-GTC AGG CTA TCG CGT ATC G-3'). HCT116 cells cultured in 100 mm diameter dishes were transiently transfected with 10 µg of shRNA expression vector by means of Lipofectamine 2000 (Invitrogen). Transfected cells were selected by culturing in 0.75 μg/ml puromycin-containing culture medium for 48 h before use.

    Cell proliferation assay

    A cell counting Kit-8 (DOJINDO, Kumamoto, Japan) was used to assess cell viability.24 A colony formation assay was used to analyse long-term cell proliferation. HCT116 cells (1×105) transfected with the negative control shRNA expression vector (Control shRNA) or hSgo1 shRNA expression vector (hSgo1 shRNA) were seeded in 0.75 μg/ml puromycin-containing culture medium and incubated for 14 days. The colonies obtained were stained with Giemsa solution (Merck, Darmstadt, Germany) and counted.

    Fluorescence-activated cell sorting (FACS)

    FACS analysis was performed for HCT116 cells transfected with the negative control shRNA expression vector (Control shRNA) or hSgo1 shRNA expression vector (hSgo1 shRNA) as described previously.25 26 In brief, the cells were washed in PBS, incubated in PBS for 10 min, and then trypsinised and fixed in 80% ethanol. The fixed cells were incubated for 30 min with normal goat serum diluted at 1:30 in PBS, and then for 2 h at room temperature in the presence of mouse monoclonal antibody to human cyclin B1 (BD Pharmingen, San Diego, California, USA) that had been diluted to 1:30 in PBS. The cells were washed and incubated for an hour with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG antibody diluted at 1:30 in PBS at room temperature. The cells were washed again, resusupended in 50 μg/ml propidium iodide and analysed by using a Beckman Coulter Epics XL (Beckman Coulter, Fullerton, California, USA). Single cell suspensions were prepared from paraffin-embedded material as described previously.27

    Time-lapse imaging

    Cells maintained in culture medium in a 35 mm dish were placed in an incubation chamber equipped with a time-lapse imaging system (BioStation IM; Nikon, Tokyo, Japan). Phase contrast images were captured at 10 min intervals and then digitised with BioStation IM viewer software.

    Indirect immunofluorescence analysis

    Indirect immunofluorescence analysis was performed as described elsewhere.28 Cells were probed with anti-γ-tubulin monoclonal antibody (GTU88; Sigma-Aldrich, St Louis, Missouri, USA) or anti-claudin-4 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, California, USA). Indirect immunofluorescence labelling of γ-tubulin or claudin-4 was performed with Alexa Fluor 594-conjugated goat anti-mouse secondary antibody (Molecular Probes, Eugene, Oregon, USA) or FITC-conjugated mouse anti-goat secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, Pennsylvania, USA), and the nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, California, USA).

    RESULTS

    Expression of hSgo1 mRNA and protein was downregulated in the colorectal cancer cases

    Among the 30 colorectal cancer cases tested, five hSgo1 genetic polymorphisms were detected (table 2) and 31.2% of informative cases had LOH at the hSgo1 locus on 3p24, but there was no mutation in the other allele (fig 1). To analyse expression of hSgo1, qRT-PCR was used in 46 colorectal cancer cases (table 3). In 22 of the 46 cases (47.8%) hSgo1 was significantly downregulated (T/N ratio <0.5) in the colorectal cancer tissue in comparison with the expression level in the paired normal mucosa (fig 2A). A statistical analysis revealed that hSgo1 was significantly downregulated in colorectal cancers compared with the corresponding normal tissues (p = 0.032 in Wilcoxon matched pairs test). Western blotting was then performed in 15 colorectal cancer cases with 13G1 monoclonal antibody to hSgo1, which recognised both full-length hSgo1 protein and proteolytic cleavage fragments.21 The results showed that expression of the full-length protein product of hSgo1 (72 kDa) was decreased in 40% (6/15) (fig 2B). We also performed immunohistochemical staining of tissue from the colorectal cancer cases with the antibody used for western blotting. In the non-tumour colonic mucosa, hSgo1 expression was localised in the nuclei of the cells at the bottom and the middle of the crypts, where the Ki-67-positive cells were located, implying that the hSgo1-positive area overlapped with the active proliferative zone of the crypt. In addition, hSgo1 expression decreased gradually toward the top of the normal colonic crypts (fig 2C). In contrast, no gradient of hSgo1 expression was observed in the case with a T/N ratio of hSgo1 mRNA expression of 0.97, and diffuse hSgo1 expression was observed in the tumour cells (fig 2C, panels e,f). The results for the level of hSgo1 expression detected by immunohistochemical staining reflected the level of hSgo1 mRNA expression (fig 2C, panels i,j).

    Figure 2
    Figure 2 Human shugoshin (hSgo1) expression at the transcription and protein level in colorectal cancer. (A) Comparison of hSgo1 mRNA levels between tumour tissue and adjacent normal tissue among 46 cases by real-time quantitative reverse transcription-PCR. Relative hSgo1 mRNA expression (tumour/normal (T/N)) in each case is indicated by a column, and the cases are numbered in order of decreasing relative hSgo1 mRNA expression. Columns under the solid line and under the broken line have hSgo1 T/N<1.0 and hSgo1 T/N<0.5, respectively. Differences between the standardised hSgo1 mRNA level in tumour tissue and adjacent normal tissue were statistically analysed by the Wilcoxon matched pairs test, and the p value was 0.032. (B) Western blotting of hSgo1. (C) Immunostaining of tumour tissue and adjacent normal tissue with anti-hSgo1 antibody and anti-Ki-67 antibody. Scale bars: 100 µm. Panels b, d, f, h, j and l are high-power magnifications of panels a, c, e, g, i and k, respectively.
    View this table:
    Table 2 Summary of the human shugoshin genetic polymorphisms in 30 colorectal cancer cases
    View this table:
    Table 3 Profile of cases in which human shugoshin (hSgo1) expression was analysed by quantitative real-time reverse transcription-PCR and comparison of clinicopathological parameters between the group in which the tumour to normal mucosa (T/N) ratio of hSgo1 mRNA expression was ⩾0.5 and the group in which it was <0.5

    Clinicopathological characteristics of cases of colorectal cancers in which hSgo1 was downregulated

    To characterise further the cases of colorectal cancer in which hSgo1 was downregulated, we compared the clinicopathological parameters of those of the 46 colorectal cancer cases in which the T/N ratio of hSgo1 mRNA expression was <0.5 (hSgo1 T/N<0.5) and in which it was 0.5 or more (hSgo1T/N⩾0.5) (table 3). The tumours in the hSgo1 T/N<0.5 group tended to be located on the left side of the large bowel (descending colon, sigmoid colon and rectum p = 0.012), especially in the rectum (p = 0.040), rather than in the other regions of the large bowel (fig 3A).

    Figure 3
    Figure 3 Comparison between tumour location, mean centromeric number and centromeric number distribution in the group with a tumour to normal mucosa ratio of human shugoshin (hSgo1) mRNA expression of 0.5 or more (hSgo1 T/N⩾0.5) and the group in which it was less than 0.5 (hSgo1 T/N<0.5). (A) Comparison between tumour location in the two groups. Open circles, hSgo1 T/N⩾0.5 cases; filled circles, hSgo1 T/N<0.5 cases. The percentages of left and right side colorectal cancer cases and of rectal and non-rectal cancer cases were statistically compared in the two groups. p Values were calculated by the χ2 test. (B) Mean signal number and signal number variations represent different characteristics of tumour cells. The cells in the left and right upper panels have the same mean signal numbers (both have a mean number of two), but they have different signal variation. The cells in the right lower panel have greater signal variation than those in the left lower panel. Original magnifications: ×600. (C and D) Comparison between the mean signal number per cell (C) and the SD of the signal number per case (D) in the two groups. Filled circles, mean (C) or SD (D) value in each case. Open circles, average mean (C) or SD (D) value in each group. Centromeric probes for chromosomes 3, 12, 17 and 18 were used. The number of signals per cell was counted in >200 cell nuclei. p Values were calculated by Student t test. (E) Frequency distribution histograms of cellular DNA content. The arrowhead points to the population with near diploid aneuploidy in the cases of hSgo1 T/N = 0.28.

    To assess aneuploidy in the 46 cases we performed a FISH analysis with a centromeric probe specific for chromosomes 3, 12, 17 and 18 (CEP3, 12, 17 and 18). Aneuploidy was assessed on the basis of the mean signal number and the signal number distribution which was calculated as the “SD of the signal number” in each case (fig 3B). The difference between mean signal number in the hSgo1 T/N<0.5 group and hSgo1T/N⩾0.5 group was not significant (fig 3C), but the signal number distribution in the hSgo1 T/N<0.5 group appeared to be wider, because there were higher SD values in the hSgo1 T/N<0.5 group, especially for chromosomes 3 and 18 (chromosome 3, p = 0.045; chromosome 18, 0.042; fig 3D). FACS analysis of the cellular DNA content of paraffin-embedded colorectal cancer cells showed aneuploidy when hSgo1 T/N was <0.5, as opposed to diploidy when the hSgo1 T/N was ⩾0.5 (fig 3E). These results suggested a more unstable chromosome numerical abnormality, which is a clinicopathological character of CIN, in colorectal cancers in which hSgo1 is downregulated.

    Knockdown of hSgo1 in HCT116 leads to accumulation of mitotic or aneuploid cells

    Based on the results of the hSgo1 expression analysis in a clinicopathological context, we next tested the in vitro effect of hSgo1 depletion by transfecting a hSgo1 shRNA expression vector that contains a puromycin resistance gene (hSgo1 shRNA) into the HCT116 cell line, which is well known for its microsatellite instability due to a MMR defect. The hSgo1 shRNA expression vector reduced the hSgo1 protein expression level in HCT116 cells to 14%, and hSgo1 knockdown cells expressed a higher level of phospho-histone H3 than the control cells (fig 4A). A WST-8 assay and a colony formation assay showed that the hSgo1 shRNA-transfected cells proliferated more slowly than the control HCT116 cells (fig 4B,C). To scrutinise the mechanism whereby the hSgo1 knockdown cells stopped their proliferation, we performed FACS analysis of cells that had asynchronously grown for 5 days after puromycin selection. The percentage of the sub-G1 population (apoptotic cell population) in the hSgo1 knockdown cells was higher than in the control cells (9.1% (0.29%) vs 4.4% (0.9%), p = 0.002; fig 4D), and the hSgo1 knockdown phenotype tended to have more cells in the G2/M phase (4N DNA content) (27.4% (3.7%) vs 11.6% (2.8%), p = 0.009; fig 4D) and more cells with >4N DNA content (9.4% (1.9%) vs (2.4% (0.7%), p = 0.008; fig 4D) than the control cells. The expression of cyclin B1 in the population of hSgo1 knockdown cells with a 4N DNA content was as high as in the control cell population with a 4N DNA content (fig 4E), suggesting that hSgo1 knockdown affected mitosis. These results indicate that knockdown of hSgo1 leads not only to apoptotic cell death but to G2/M arrest or aneuploidy/polyploidy.

    Figure 4
    Figure 4 Growth properties of human shugoshin (hSgo1) knockdown cells. (A) Western blot showing the efficiency of hSgo1 knockdown by transfection of HCT116 cells with hSgo1 short hairpin RNA (shRNA) in comparison with negative control shRNA. hSgo1 knockdown cells expressed a higher level of phospho-histone H3 than the control cells, suggesting accumulation of hSgo1 knockdown cells in mitosis. (B) WST-8 assay. Cells (5000 cells/well) of each population were seeded in 100 μl of culture medium in the wells of three 96-well plastic plates, and one plate each was incubated in a 5% CO2/95% air incubator for 24, 48 and 72 h. After incubation, 10 μl of WST-8 reagent solution was added to each well, and the plate was incubated for an additional 4 h in the 5% CO2/95% air incubator. Cell viability was measured by scanning with a microplate reader at 450 nm. This experiment was performed three times. (C) Colony formation assay of puromycin-resistant colonies of cells transfected with the control shRNA expression vector or the hSgo1 shRNA expression vector. Colonies were counted 14 days post-transfection. This experiment was performed three times. (D) Fluorescence-activated cell sorting (FACS) analysis of the DNA content profile of asynchronously growing cells 5 days after transfection with the control shRNA expression vector or hSgo1 shRNA expression vector. Numbers show the mean (SD) of three independent experiments. (E) FACS analysis of DNA and cyclin B1, which is highly expressed in the G2/M phase. Cyclin B1 was highly expressed in the 4N DNA population both of control cells and of hSgo1 knockdown cells. The black line represents the upper boundary of cyclin B1 fluorescence as a control. Blue dot plot, a cell with a 2N DNA content; red dot plot: a cell with a 4N DNA content.

    Knockdown of hSgo1 leads to chromosomal instability

    A FISH analysis was performed to assess the development of aneuploidy/polyploidy in hSgo1 shRNA cells. We used CEP12 and 17 to count the chromosome number. A higher percentage of cells with an abnormal signal number was observed among the hSgo1 shRNA cells than the control shRNA cells or parental cells (fig 5A). Cells with abnormal signal numbers, especially four signals, were more frequent in the hSgo1 shRNA cells than in the control shRNA cells or the parental cells (table 4A, p<0.0001). The cytological features such as nucleoplasmic bridges, micronuclei and nuclear blebs are well known signatures of CIN.29 hSgo1 shRNA induced micronuclei in the HCT116 cells (3.9% of parental cells, 2.4% of control shRNA cells and 9.8% of hSgo1 shRNA cells, p<0.005) (table 4B, fig 5B). This finding also indicates that hSgo1 shRNA led to CIN.

    Figure 5
    Figure 5 Observation of chromosomal instability under human shugoshin (hSgo1)-depleted conditions. (A) Fluorescent in situ hybridisation (FISH) analysis with centromeric probes for chromosomes 12 (red) and 17 (green). The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). The numbers of chromosome 12 centromere signals and chromosome 17 centromere signals in cells with any abnormal signals in these chromosomes are shown as a fraction, with the former number in red and the latter number in green. hSgo1 knockdown cells (right panels) showed aneuploidy or polyploidy, in contrast to the parental cells (left upper panel) and control short hairpin RNA (shRNA) phenotype (left lower panel). (B) Detection of micronuclei (white and yellow arrowheads). The nuclei were stained with DAPI (blue). The FISH analysis (right lower panel) detected a centromeric probe (red/chromosome 12) in a micronucleus (yellow arrowhead). Original magnifications: ×600.
    View this table:
    Table 4 Distribution of numerical chromosome variability (A) and frequency of micronuclei (B) in human shugoshin (hSgo1) knockdown cells

    Time-lapse imaging of HCT116 cells transfected with hSgo1 shRNA

    We thought that observation for a longer time than in any previous studies would enable a more thorough description of the effect of hSgo1 shRNA, thus time-lapse microscopy was used to monitor cells through mitosis. Figure 6A(a) shows that control shRNA cells initiated mitosis and underwent normal cytokinesis. On the other hand, the hSgo1 shRNA cells initially remained in a prolonged period of mitotic arrest (a period in which the cells assumed a spherical shape), after which some hSgo1 shRNA cells underwent an aberrant cell division into three daughter cells (fig 6A(b)), some were binucleated cells, probably due to cytokinesis defects30 (fig 6A(c)), and others exited the mitotic phase without undergoing cell division (which may correspond to “mitotic slippage” or “adaptation” of the cells31) (fig 6A(d)). Most cells shrank immediately after mitosis and died within a day (mitotic apoptosis or catastrophe) (fig 6A(e)). The indirect immunofluorescence analysis performed using antibodies to γ-tubulin, a core component of the centrosome, revealed a higher percentage of hSgo1 shRNA cells containing more than three centrosomes than among the control cells (fig 6B, C). The indirect immunofluorescence analysis showed that 5% of the hSgo1 shRNA cells were binucleate, but that none of the control shRNA or parental cells were (fig 6D). These results suggest that hSgo1 knockdown of colorectal cancer cells is accompanied by the emergence of aberrant cell division, centrosome amplification and mitotic slippage, as well as mitotic catastrophes.

    Figure 6
    Figure 6 Centrosome amplification, binucleated cells and mitotic slippage accompanied human shugoshin (hSgo1) knockdown in HCT116 cells. (A) Time-lapse microscopy of the control short hairpin RNA (shRNA) phenotype (a) and the hSgo1 shRNA phenotype (b–e) 5 days post-transfection. Images were captured at regular intervals over the course of the experiment, and selected images are shown. (b) Aberrant cell division into three daughter cells. (c) A binucleated cell. Yellow arrowheads point to the nuclei in a binucleated cell. (d) After a prolonged period of mitotic arrest (up to 310 min, the cell continues to have a spherical shape), the cell (yellow arrows) exits the mitotic phase without undergoing cell division. (e) Mitotic catastrophe. The cells enter the mitotic phase and become apoptotic. Original magnification, ×10. (B) Detection of centrosome amplification in the hSgo1 shRNA phenotype. Cells were stained with anti-γ-tubulin monoclonal antibody (red) and counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). (C) Percentages of cells containing more than three centrosomes in interphase and the mitotic phase. p Values were calculated by Student t test. Columns represent the means of three independent experiments; bars, SD. (D) Detection of binucleated cell in the hSgo1 shRNA phenotype. Cells were stained with anti-claudin-4 monoclonal antibody (green) and counterstained with DAPI (blue). Original magnification: ×600.

    DISCUSSION

    We have investigated the clinicopathological significance of hSgo1 in human colorectal cancer. The results showed (1) that hSgo1 was downregulated at the transcriptional and protein level in human colorectal cancers; (2) that the tumours of hSgo1 downregulation preferentially located on the left side large bowel and tended to have a wider centromecic numerical distribution in each tumour; and (3) that knockdown of hSgo1 in colorectal cancer cell lines leads to CIN. This is the first study on the role of hSgo1 in gastrointestinal cancer.

    Genetic alterations of hSgo1 in human carcinogenesis have never been extensively elucidated. In terms of expression at the transcriptional level, hSgo1 has been reported to be a breast cancer-specific message,32 but no follow-up characterisations have been performed. We found a lower level of hSgo1 expression at the transcription and protein level in colorectal cancer tissue than in normal colorectal mucosa, suggesting a key role for hSgo1 in colorectal carcinogenesis. In the same way as in a previous study of another gene in which Kloth et al concluded that there was no association between the tumour necrosis factor α (TNFα) mRNA expression level and LOH at the TNFα locus,33 no association between LOH at the hSgo1 locus and hSgo1 mRNA expression level was found in our study, suggesting that hSgo1 mRNA expression is controlled by another factor, such as transcriptional or post-transcriptional activity, and not by LOH. Suzuki et al reported that a western blot analysis of human cell lines yielded proteolytic cleavage fragments as well as full-length hSgo1 protein.21 In interphase Xenopus egg extracts, Xenopus Sgo has been found to be ubiquitinated and degraded by the anaphase-promoting complex in vitro.8 hSgo1 may be degraded in the same manner, although the precise mechanism is unknown.

    Various biological differences between adenocarcinomas on the right side and left side of the large bowel have been reported34—for example that sporadic CIN tumours tend to be located on the left side.35 36 The results of our study showing predominant left side location of hSgo1-downregulated tumours is consistent with these reports, and thus hSgo1 downregulation may be related to CIN. Although our FISH analysis of 46 colorectal cancer cases did not show any significant difference between mean signal number in the hSgo1 T/N<0.5 group and the hSgo1 T/N⩾0.5 group, the standard deviation of the signal number in each case was higher in the hSgo1 T/N<0.5 group than in the hSgo1 T/N⩾0.5 group, a finding that may reflect unstable chromosome number status (near diploid aneuploidy) in hSgo1 T/N<0.5 tumours. To assess CIN rigorously, our counts included CEP3 and 18, which were found to be relatively stable, especially in early stage gastric cancer, in the previous study.23 The results of the precise assessment of chromosome numerical status supported the clinical view of the CIN character of hSgo1-downregulated colorectal cancers.

    Recently, it was reported that hSgo1 knockdown causes premature chromatid separation in vivo9 but that overexpression of hSgo1 did not alter the DNA content according to a FACS analysis.21 In view of these hSgo1 knockdown effects, it is conceivable that hSgo1 knockdown leads to CIN. However, it should be noted that all the previous studies observed and analysed the effects of hSgo1 alteration only over a relatively short period (within the period of “one” cell cycle or mitotic phase). Thus, the long-term observation of the hSgo1 knockdown cells for several days to a week was essential to obtain further insights into the role of hSgo1 in tumour cells. By extending our observation period to a week, we were able track the sequence of a wider variety of molecular alterations caused by hSgo1 downregulation.

    At first, the low proliferation rate in hSgo1 knockdown cells seemed to suggest that hSgo1 authentically suppresses tumour progression, but the subsequent FACS analysis and time-lapse imaging analysis revealed that hSgo1 knockdown cells were blocked in the mitotic phase, which is consistent with the observation and previous reports.8 9 Moreover, there was a larger cell population with a >4N DNA content, implying that hSgo1 knockdown leads to a partial mitotic block and escape into the next cycle as aneuploid or polyploid cells or proceed to mitotic catastrophe. The high frequency of aneuploidy/polyploidy observed in the FISH analysis of hSgo1 knockdown cells validated this conclusion. Several previous reports have described a similar phenomenon as the effect of alterations. Sotillo et al reported that cells arrest in G2/M when MAD2 is overexpressed and escape from mitotic block,25 and Hernando et al recently reported that MAD2 overexpression in IMR90 primary fibroblasts did not lead to a permanent block but to escape into the next cycle and generation of cells with >4N content.37 In these lines, we expected that hSgo1 dysfunction may induce CIN through G2/M arrest. Actually, the FISH analysis showed a higher percentage of tetraploid cells than that of other aneuploid/polyploid cells in the hSgo1 shRNA group. Tetraploid cells can arise through mitotic slippage or cytokinesis failure, which result in binucleated cells.31 Moreover, supernumerical centrosomes are a consequence of tetraploid formation,38 as we observed in hSgo1 knockdown cells. By observing tetraploid cells, centrosome amplification and binucleated cells, we succeeded in observing the process of diploid colon cancer cells developing CIN as a result of knockdown of hSgo1 in real time. In view of the above circumstances, “cell growth suppression” by hSgo1 downregulation at first glance may not simply be the same as “tumour suppression”, but it is an important step on the path to aneuploidy.

    In conclusion, the results of this study showed that the hSgo1-downregulated colorectal cancer has a clinicopathological character of CIN and that hSgo1 downregulation leads to CIN in human colorectal cancer cells. These results may be of help in determining the molecular mechanism of the CIN pathway in colorectal cancer progression.

    Acknowledgments

    We thank Mrs Nagura, Dr Takashi Sakurai and Dr Susumu Terakawa for their technical assistance.

    REFERENCES

    1. ↵
      1. Lengauer C,
      2. Kinzler KW,
      3. Vogelstein B
      . Genetic instabilities in human cancers. Nature 1998;396:643–9.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Kinzler KW,
      2. Vogelstein B
      . Lessons from hereditary colorectal cancer. Cell 1996;87:159–70.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Lynch HT,
      2. de la Chapelle A
      . Hereditary colorectal cancer. N Engl J Med 2003;348:919–32.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Rajagopalan H,
      2. Lengauer C
      . Aneuploidy and cancer. Nature 2004;432:338–41.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Grady WM
      . Genomic instability and colon cancer. Cancer Metastasis Rev 2004;23:11–27.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Zou H,
      2. McGarry TJ,
      3. Bernal T,
      4. et al
      . Identification of a vertebrate sister-chromatid separation inhibitor involved in transformation and tumorigenesis. Science 1999;285:418–22.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Kitajima TS,
      2. Kawashima SA,
      3. Watanabe Y
      . The conserved kinetochore protein shugoshin protects centromeric cohesion during meiosis. Nature 2004;427:510–7.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Salic A,
      2. Waters JC,
      3. Mitchison TJ
      . Vertebrate shugoshin links sister centromere cohesion and kinetochore microtubule stability in mitosis. Cell 2004;118:567–78.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. McGuinness BE,
      2. Hirota T,
      3. Kudo NR,
      4. et al
      . Shugoshin prevents dissociation of cohesin from centromeres during mitosis in vertebrate cells. PLoS Biol 2005;3:e86.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Losada A,
      2. Hirano M,
      3. Hirano T
      . Cohesin release is required for sister chromatid resolution, but not for condensin-mediated compaction, at the onset of mitosis. Genes Dev 2002;16:3004–16.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Sumara I,
      2. Vorlaufer E,
      3. Stukenberg PT,
      4. et al
      . The dissociation of cohesin from chromosomes in prophase is regulated by Polo-like kinase. Mol Cell 2002;9:515–25.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Holtrich U,
      2. Wolf G,
      3. Brauninger A,
      4. et al
      . Induction and down-regulation of PLK, a human serine/threonine kinase expressed in proliferating cells and tumors. Proc Natl Acad Sci USA 1994;91:1736–40.
      OpenUrlAbstract/FREE Full Text
      1. Eckerdt F,
      2. Yuan J,
      3. Strebhardt K
      . Polo-like kinases and oncogenesis. Oncogene 2005;24:267–76.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Smith SL,
      2. Bowers NL,
      3. Betticher DC,
      4. et al
      . Overexpression of aurora B kinase (AURKB) in primary non-small cell lung carcinoma is frequent, generally driven from one allele, and correlates with the level of genetic instability. Br J Cancer 2005;93:719–29.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Kitajima TS,
      2. Hauf S,
      3. Ohsugi M,
      4. et al
      . Human Bub1 defines the persistent cohesion site along the mitotic chromosome by affecting Shugoshin localization. Curr Biol 2005;15:353–9.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Shichiri M,
      2. Yoshinaga K,
      3. Hisatomi H,
      4. et al
      . Genetic and epigenetic inactivation of mitotic checkpoint genes hBUB1 and hBUBR1 and their relationship to survival. Cancer Res 2002;62:13–7.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Wang J,
      2. Kataoka H,
      3. Suzuki M,
      4. et al
      . Downregulation of EphA7 by hypermethylation in colorectal cancer. Oncogene 2005;24:5637–47.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Guo RJ,
      2. Wang Y,
      3. Kaneko E,
      4. et al
      . Analyses of mutation and loss of heterozygosity of coding sequences of the entire transforming growth factor beta type II receptor gene in sporadic human gastric cancer. Carcinogenesis 1998;19:1539–44.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Tao H,
      2. Shinmura K,
      3. Hanaoka T,
      4. et al
      . A novel splice-site variant of the base excision repair gene MYH is associated with production of an aberrant mRNA transcript encoding a truncated MYH protein not localized in the nucleus. Carcinogenesis 2004;25:1859–66.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Tomonaga T,
      2. Matsushita K,
      3. Ishibashi M,
      4. et al
      . Centromere protein H is up-regulated in primary human colorectal cancer and its overexpression induces aneuploidy. Cancer Res 2005;65:4683–9.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Suzuki H,
      2. Akiyama N,
      3. Tsuji M,
      4. et al
      . Human Shugoshin mediates kinetochore-driven formation of kinetochore microtubules. Cell Cycle 2006;5:1094–101.
      OpenUrlPubMed
    22. ↵
      1. Kawasaki Y,
      2. Monden T,
      3. Morimoto H,
      4. et al
      . Immunohistochemical study of p53 expression in microwave-fixed, paraffin-embedded sections of colorectal carcinoma and adenoma. Am J Clin Pathol 1992;97:244–9.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Kitayama Y,
      2. Igarashi H,
      3. Sugimura H
      . Different vulnerability among chromosomes to numerical instability in gastric carcinogenesis: stage-dependent analysis by FISH with the use of microwave irradiation. Clin Cancer Res 2000;6:3139–46.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Jiang Y,
      2. Ahn EY,
      3. Ryu SH,
      4. et al
      . Cytotoxicity of psammaplin A from a two-sponge association may correlate with the inhibition of DNA replication. BMC Cancer 2004;4:70.
      OpenUrlCrossRefPubMed
    25. ↵
      1. Sotillo R,
      2. Hernando E,
      3. Diaz-Rodriguez E,
      4. et al
      . Mad2 overexpression promotes aneuploidy and tumorigenesis in mice. Cancer Cell 2007;11:9–23.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Kakino S,
      2. Sasaki K,
      3. Kurose A,
      4. et al
      . Intracellular localization of cyclin B1 during the cell cycle in glioma cells. Cytometry 1996;24:49–54.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Hedley DW,
      2. Friedlander ML,
      3. Taylor IW,
      4. et al
      . Method for analysis of cellular DNA content of paraffin-embedded pathological material using flow cytometry. J Histochem Cytochem 1983;31:1333–5.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Yamada H,
      2. Shinmura K,
      3. Okudela K,
      4. et al
      . Identification and characterization of a novel germ line p53 mutation in familial gastric cancer in the Japanese population. Carcinogenesis 2007;28:2013–8.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Camps J,
      2. Ponsa I,
      3. Ribas M,
      4. et al
      . Comprehensive measurement of chromosomal instability in cancer cells: combination of fluorescence in situ hybridization and cytokinesis-block micronucleus assay. FASEB J 2005;19:828–30.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Trachana V,
      2. van Wely KH,
      3. Guerrero AA,
      4. et al
      . Dido disruption leads to centrosome amplification and mitotic checkpoint defects compromising chromosome stability. Proc Natl Acad Sci USA 2007;104:2691–6.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Ganem NJ,
      2. Storchova Z,
      3. Pellman D
      . Tetraploidy, aneuploidy and cancer. Curr Opin Genet Dev 2007;17:157–62.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Scanlan MJ,
      2. Gout I,
      3. Gordon CM,
      4. et al
      . Humoral immunity to human breast cancer: antigen definition and quantitative analysis of mRNA expression. Cancer Immun 2001;1:4.
      OpenUrlPubMed
    33. ↵
      1. Kloth JN,
      2. Gorter A,
      3. ter Haar N,
      4. et al
      . Lack of TNFalpha mRNA expression in cervical cancer is not associated with loss of heterozygosity at 6p21.3, inactivating mutations or promoter methylation. Mol Immunol 2008;45:152–9.
      OpenUrlCrossRefPubMed
    34. ↵
      1. Birkenkamp-Demtroder K,
      2. Olesen SH,
      3. Sorensen FB,
      4. et al
      . Differential gene expression in colon cancer of the caecum versus the sigmoid and rectosigmoid. Gut 2005;54:374–84.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Iacopetta B
      . Are there two sides to colorectal cancer? Int J Cancer 2002;101:403–8.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Lindblom A
      . Different mechanisms in the tumorigenesis of proximal and distal colon cancers. Curr Opin Oncol 2001;13:63–9.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Hernando E,
      2. Nahle Z,
      3. Juan G,
      4. et al
      . Rb inactivation promotes genomic instability by uncoupling cell cycle progression from mitotic control. Nature 2004;430:797–802.
      OpenUrlCrossRefPubMed
    38. ↵
      1. Nigg EA
      . Centrosome aberrations: cause or consequence of cancer progression? Nat Rev Cancer 2002;2:815–25.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • See Commentary, p 163

    • Funding: Supported in part by a Grant-in-Aid from the Ministry of Health, Labor and Welfare for the 2nd-term Comprehensive 10-Year Strategy for Cancer Control (15-22, 19-19), by a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan on Priority Area (18014009), a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (17590629) and the 21st century COE programme “Medical Photonics”, and by the Smoking Research Foundation and Aichi Cancer Research Foundation.

    • Competing interests: None.

    • Ethics approval: The study design was approved by the Institutional Review Board (IRB) of Hamamatsu University School of Medicine.

    Linked Articles