Figure 1.
(A) Representative phosphor images showing LOH for the markers D8S1819, D8S277 and D8S1798 flanking the MCPH1 locus. B & T denote constitutive blood/normal oral tissue and tumor DNA respectively. Arrows indicate the loss of alleles in tumor DNA. (B) Diagrammatic representation of LOH data from 81 matched blood/normal oral tissue and tumor DNA samples using three microsatellite markers. Tumor samples are arranged according to their T classification (T1 to T4) or ED (epithelial dysplasia). Abbreviations: NI, non-informative; IN, informative; MSI, microsatellite instability; and, L, LOH.
Figure 2.
(A) Sequencing chromatograms from the normal and tumor oral tissues of the pt# 110 with a nonsense mutation c.1561G>T (p.Glu521X) in a homozygous state. The LOH study has shown deletion of one allele in this tumor (Figure S1 in File S2), suggesting that both alleles are non-functional due to either deletion or mutation. Arrows mark the sites of the mutation in chromatograms. The codon affected by the mutation is underlined. Schematic representations of wild-type and mutant MCPH1 proteins are shown below the chromatograms. The wild-type protein is 835 amino acids long and harbors three BRCT domains, a NLS (nuclear localization signal) and a CIIBR (condensin II binding region). The truncated mutant protein is predicted to be 520 amino acids long with the loss of two C-terminal BRCT domains. (B) Sequencing chromatograms from SCC084 and SCC131 cells with the mutations c.321delA (p.Lys107fsX39) in exon 4 and c.1402delA (p.Thr468fsX32) in exon 8, respectively, in a heterozygous state. Schematic representations of the mutant proteins are shown below the chromatograms. Red bars in protein structure represent abnormal amino acids.
Figure 3.
Methylation of CpG islands in the MCPH1 promoter.
(A) Schematic representation of the MCPH1 promoter with two CpG islands. The vertical lines represent the CpG sites. The solid and open horizontal arrows represent primers to amplify CpGI and CpGII islands respectively. Sites for Bst UI and Aci I in CpGI and CpGII, respectively, are marked by filled vertical arrowheads. The numbers represent nucleotide positions with respect to the TSS. (B) Representative agarose gel images of COBRA for CpGI (upper panel) and CpGII (lower panel). Note the absence of methylation of CpGI in tumor samples 95T and 150T, and the methylation of CpGII in tumor samples 80T and 116T. (C) Schematic representation of bisulfite treated genomic DNA sequence of CpGII in normal and tumor tissues from patient numbers 80, 116, 177 and 202. Each row represents a sequenced TA clone. The filled and unfilled squares represent methylated and unmethylated CpGs respectively. Note the methylation of tumor samples and non-methylation of their corresponding normal oral tissues. (D) Representative agarose gel images of COBRA data for CpGI (upper panel) and CpGII (lower panel) in cell lines. None of the cell lines show CpGI methylation, whereas CpGII shows methylation in SCC084 cells only. (E) Bisulfite sequencing of CpGII in SCC084 cells before and after the AZA (2′-deoxy-5-azacytidines) treatment. The CpG sites in CpGII show methylation in DMSO (vehicle control) treated DNA, whereas, as expected, methylation is lost after AZA treatment. Abbreviations: N, normal; T, tumor; PD, positive control (ASPM fragment); PU, positive control undigested; UN, undigested CpG island I or II; and, NB1 or NB2, peripheral blood DNA from unrelated normal individuals. Numbers represent patient numbers.
Figure 4.
MCPH1 overexpression in stable clones.
Representative images of Western blots of the lysates from stable clones generated with either the pcDNA3.1(+)/MCPH1/myc-His construct or the empty vector pcDNA3.1(+)/myc-His after G418 selection in KB cells. ß–actin was used as a loading control. Abbreviations: KB, parental cell line; V2 and V4, stable clones with the vector only; and, B1, B4, B9 and B10, MCPH1 overexpressing stable clones.
Figure 5.
Tumor suppressor activity of MCPH1.
(A) The analysis of cell proliferation in KB (parental cell line), empty vector stable clones (V2 and V4) and MCPH1 overexpressing stable clones (B1 and B9) by the BrdU assay. The graph represents the OD (optical density) measurment at 450 nm. Note the B1 and B9 clones incorporated significantly lesser BrdU than the V2, V4 and KB cells. The values are the mean±SD of three independent experiments. Abbreviations: ns, no statistical significance; and, *** indicates p<0.001. (B) Anchorage-independent growth of KB, V2, V4, B1 and B9 cells in soft agar. Note MCPH1 overexpression reduces colony growth of B1 and B9 clones as compared to KB, V2 and V4 cells. (C) The in vivo tumorigenic potential of V2, V4, B1 and B9 cells in nude mice. The representative images of the nude mice with tumors (upper panel) and the harvested tumors (lower panel) are shown. Note that the overexpression of MCPH1 reduces tumor growth in nude mice. The arrows mark the tumors. (D) Tumor volumes at different time points. The values given are the mean±SD from five animals. * indicates p<0.05.
Figure 6.
Increased apoptosis and decreased invasion in MCPH1 overexpressing cells.
(A) The analysis of sub-G1 populations (a measure of cell death) in PI stained KB, V2, V4, B1 and B9 cells by flow cytometry. The graph represents one of the three separate experiments. Note higher sub-G1 populations in MCPH1 overexpressing B1 and B9 cells in comparison to KB, V2 and V4 cells. The mean values (%) ± SDs for sub-G1 populations in different cells are as follows: KB, 5.07±0.63; V2, 2.06±0.71; V4, 2.40±0.48; B1, 16.16±0.57; and B9, 22.64±0.45. (B) Analysis of apoptosis by in vitro quantitation of CASP3 activity. Note significantly increased CASP3 activity in B1 and B9 cells in comparison to KB, V2 and V4 cells. The values shown are mean±SD of three separate experiments. (C) Analysis of cell invasion in KB cells, and V2, V4, B1 and B9 clones. Note MCPH1 overexpression reduced invasiveness in B1 and B9 cells as compared to KB, V2 and V4 cells. (D) The quantitative representation of the cell invasion assay data. The values are the mean±SD of the number of invaded cells counted in four random microscopic fields. Abbreviations: ns, statistically not significant; **p<0.005; and, ***p<0.001.
Figure 7.
miR-27a targets both the seed regions of MCPH1 cloned together.
The luciferase reporter assay of different constructs of the full-length 3′-UTR of MCPH1 in KB cells. Note a significantly reduced luciferase activity in cells co-transfected with pMIR-Report-3′-UTR-S and pcDNA3/pre-miR-27a/EGFP (0.2 or 0.4 µg) in comparison to those with pMIR-Report or pMIR-Report-3′-UTR-AS, suggesting miR-27a targets MCPH1. Note that both the seed regions in the 3′-UTR of MCPH1 are important for the interaction with miR-27a as a significant increase in the luciferase activity was observed in cells transfected with pMIR-Report-3′-UTR-M1F, pMIR-Report-3′-UTR-M2F or pMIR-Report-3′-UTR-MF along with pcDNA3/pre-miR-27a/EGFP in comparison to cells transfected with pMIR-Report-3′-UTR-S and pcDNA3/pre-miR-27a/EGFP. Abbreviations: ns, statistically not significant; and, ***p<0.001.
Figure 8.
miR-27a targets both the seed regions of MCPH1 cloned separately.
(A) The luciferase reporter assay of different constructs harboring seed region 1 (SDR1) of MCPH1 3′-UTR in KB cells. Note a significantly reduced luciferase activity in cells co-transfected with pMIR-Report-3′-UTR-S1 and pcDNA3/pre-miR-27a/EGFP (0.2 or 0.4 µg) in comparison to those with pMIR-Report, suggesting miR-27a targets SDR1 of MCPH1. As expected, no significant difference in luciferase activity was observed in cells co-transfected with pMIR-Report-3′-UTR-AS1 or pMIR-Report-3′-UTR-M1 with pcDNA3/pre-miR-27a/EGFP (0.2 or 0.4 µg) in comparison to those with pMIR-Report. (B) The luciferase reporter assay of different constructs harboring seed region 2 (SDR2) of MCPH1 3′-UTR in KB cells. As with SDR1, miR-27a also targets SDR2.
Figure 9.
miR-27a negatively regulates MCPH1.
(A) miR-27a negatively regulates MCPH1 level in KB cells. Representative images of the correlative expression of MCPH1 protein level after the transient transfection of the pcDNA3/pre-miR-27a/EGFP construct in KB cells. Note the reduced expression of MCPH1 upon overexpression (4 µg) of pcDNA3/pre-miR-27a/EGFP. Mock lane represents transfection with the empty vector pcDNA3/EGFP. 5S rRNA and ß–actin were loading controls for RT-PCR and Western blotting respectively. (B) The correlative expression analysis of miR-27a and MCPH1 in 10 paired OSCC samples. Note a negative correlation between the levels of miR-27a and MCPH1 in six matched samples: 63, 68, 109, 155, 183 and 191. However, no correlation between the levels of miR-27a and MCPH1 was observed in three matched samples: 62, 92 and 140. In the remaining one matched sample (pt# 128), the level of miR-27a was downregulated in tumor, but the level of MCPH1 was unchanged in the tumor tissue. 5S rRNA and ß–actin were loading controls for RT-PCR and Western blotting respectively. Abbreviations: N, normal oral tissue; and, T, tumor oral tissue. The numbers refer to patient numbers.