Article Figures & Data
Figures
- Fig. 1.
Down-regulation of TMS1 in fibroblasts overexpressing DNMT1. A, Southern blot prepared from 2μ g of the amplified tester (90SV) and driver (HMT.1E1) cDNA pools used to perform the RDA subtraction. Hybridization shows clone RDA-2.15 is differentially represented in cDNA derived from 90SV and HMT.1E1 cells. B, Northern blot of poly(A)+ selected RNA (2 μg) from 90SV and HMT.1E1 cells hybridized with the TMS1 cDNA probe. The same blot was stripped and rehybridized with a humanβ -actin probe. The exposure times for TMS1 and β-actin were 3 days and 4 h, respectively. C, total RNA isolated from IMR90 human diploid fibroblasts, 90SV fibroblasts, and HMT.1E1 cells was reverse-transcribed (+RT) and amplified with primers specific to the TMS1 transcript (top) or β-actin (bottom). Control reactions in which reverse transcriptase was omitted (−RT) were amplified under the same conditions.
- Fig. 2.
Absence of TMS1 expression correlates with hypermethylation of a CpG island in the TMS1 gene. A, genomic structure of the TMS1 locus. Intron/exon boundaries were determined by comparison of the cDNA and genomic sequences. The TMS1 coding region spans 1.8 kb and three exons (I, II, and III). The 5′ and 3′ untranslated regions are shaded. A CpG island in the 5′ end of the gene is evident from the density of CpG sites relative to GpC (0.88) and the clustering of multiple restriction enzyme sites (Sa, SacII; Ea, EagI; B, BssHII; Sm, SmaI). H, HindIII; E, EcoRI. Closed arrows, primers used for MSP. B, methylation of the TMS1 CpG island was examined by digestion of DNA from the indicated cell line with HindIII alone or HindIII plus the methylation-sensitive restriction enzyme SacII or EagI (all other lanes), followed by Southern blot analysis using a 1.8 kb EcoRI TMS1 genomic fragment as a probe. C, methylation at the TMS1 locus determined by MSP. DNA from IMR90, 90SV, and HMT.1E1 cells was modified with sodium bisulfite, which deaminates unmethylated cytosines to uracil and leaves methylated cytosines unaltered. Bisulfite-modified DNA was used as a template for parallel PCR amplification reactions using primers designed to anneal specifically to unmethylated (U) or methylated (M) DNA. The“ methylated” product was 191 bp and the “unmethylated” product was 196 bp because of differences in the length of the primers.
- Fig. 3.
Methylation-associated silencing of TMS1 CpG island in human breast cancer cells. A, MSP analysis of bisulfite-modified DNA from primary HMECs and two immortal, nontumorigenic breast epithelial cell lines (MCF10A and Hs578Bst) compared with that of nine breast cancer cell lines (MCF-7, T47-D, ZR-75-1, Hs578t, MDA MB231, MDA MB468, SKBR3, CAMA1, and BT-20). Parallel amplification reactions were performed using primers specific to methylated (M) or unmethylated (U) DNA. B, expression of TMS1. Total RNA isolated from the breast epithelial cells and breast cancer cells was reverse transcribed (+RT) and amplified with primers specific to the TMS1 transcript (top) or human β-actin transcript (bottom). Control reactions in which reverse transcriptase was omitted (−RT) were amplified under the same conditions. C and D, demethylation of the TMS1 CpG island restores expression in TMS1-negative breast cancer cells. ZR-75-1, MDA MB231, and Hs578t cells were not treated (None) or treated (5azadC) for 3 days with 0.5 μm 5-aza-2′-deoxycytidine and methylation of the TMS1 CpG island was analyzed by MSP (C), and expression of TMS1 was analyzed by RT-PCR (D).
- Fig. 4.
Methylation of TMS1 in primary breast tissues. Methylation-specific PCR analysis of bisulfite-modified DNA from normal breast tissue from reduction mammoplasty (A), and primary resected breast carcinomas (B). Shown are five representative tumor samples that lacked methylation at the TMS1 CpG island and five representative tumor samples exhibiting hypermethylation of the TMS1 CpG island. Parallel amplification reactions were performed using primers specific to methylated (M) or unmethylated (U) DNA. C, methylation-specific PCR analysis of TMS1 in paired normal-appearing tissue (N) adjacent to breast tumors (T). Shown are 12 representative pairs of a total of 18 analyzed.
- Fig. 5.
TMS1 encodes a proapoptotic CARD protein. A, Northern blots of poly(A)+ RNA isolated from various normal human tissues and human tumor-derived cell lines hybridized to a TMS1 cDNA probe (top) or a human β-actin (bottom) probe. B, amino acid alignment of the TMS1 COOH-terminus with the CARD motif of other apoptotic signaling proteins. Reverse type, ≥50% amino acid identity; gray shading, ≥50% similarity through conserved amino acid substitutions. C, induction of apoptosis by TMS1. The 293 cells were cotransfected with 0.1μ g of pCMVβ-gal and 0.4 μg of vector (pcDNA3.1) or the indicated TMS1 expression construct. Where indicated, zVAD (40 μm) was added immediately after transfection. After 48 h, cells were stained with X-gal and examined by phase contrast microscopy (×400).β -galactosidase-positive cells undergoing apoptosis were distinguished by a more condensed, rounded appearance and showed evidence of detachment from the culture surface (arrows). Data are presented as the percentage ofβ -galactosidase-positive cells exhibiting apoptosis (mean ± SD) from three independent transfection experiments. D, colony-forming ability of breast cancer cells stably transfected with TMS1 Hs578t, MDA MB231, or MCF 7 cells were transfected with the pcDNA3.1 vector or the indicated TMS1 expression construct. After 24 h, cells were plated at clonal density in medium supplemented with G418. Stable G418-resistant colonies remaining after 14 days were fixed in methanol and stained with crystal violet. Data (mean ± SD) represent the number of G418-resistant colonies recovered for each test construct relative to that of the vector. At least three independent transfection experiments were performed.
Volume 60, Issue 22
