Epithelial recognition and elimination against aberrant cells.

Epithelial cells, which are non-immune cells, not only function as a physical defence barrier but also continuously monitor and eliminate aberrant epithelial cells in their vicinity. In other words, it has become evident that epithelial cells possess immune cell-like functions. In fact, recent research has revealed that epithelial cells recognise the Major Histocompatibility Complex I (MHC-I) of aberrant cells as a mechanism for surveillance. This cellular defence mechanism of epithelial cells probably detects aberrant cells more promptly than the conventional immune response, making it a novel and primary biological defence. Furthermore, there is the potential for this new immune-like biological defence mechanism to establish innovative treatment for disease prevention, leading to increasing anticipation for its future medical applications. In this review, we aim to summarise the recognition and attack mechanisms of aberrant cells by epithelial cells in mammals, with a particular focus on the field of cancer. Additionally, we discuss the potential therapeutic applications of epithelial cell-based defence against cancer, including novel prophylactic treatment methods based on molecular mechanisms.

Epithelial cells remove precancerous cells by cell competition via MHC class I-LILRB3 interaction.

Epithelial cells have an ability termed 'cell competition', which is an immune surveillance-like function that extrudes precancerous cells from the epithelial layer, leading to apoptosis and clearance. However, it remains unclear how epithelial cells recognize and extrude transformed cells. Here, we discovered that a PirB family protein, leukocyte immunoglobulin-like receptor B3 (LILRB3), which is expressed on non-transformed epithelial cells, recognizes major histocompatibility complex class I (MHC class I) that is highly expressed on transformed cells. MHC class I interaction with LILRB3 expressed on normal epithelial cells triggers an SHP2-ROCK2 pathway that generates a mechanical force to extrude transformed cells. Removal of transformed cells occurs independently of natural killer (NK) cell or CD8+ cytotoxic T cell-mediated activity. This is a new mechanism in that the immunological ligand-receptor system generates a mechanical force in non-immune epithelial cells to extrude precancerous cells in the same epithelial layer.

Periostin antisense oligonucleotide suppresses bleomycin-induced formation of a lung premetastatic niche for melanoma.

Metastasis is the leading cause of cancer death. A tumor-supportive microenvironment, or premetastatic niche, at potential secondary tumor sites plays an important role in metastasis, especially in tumor cell colonization. Although a fibrotic milieu is known to promote tumorigenesis and metastasis, the underlying molecular contributors to this effect have remained unclear. Here we show that periostin, a component of the extracellular matrix that functions in tissue remodeling, has a key role in formation of a fibrotic environment that promotes tumor metastatic colonization. We found that periostin was widely expressed in fibrotic lesions of mice with bleomycin-induced lung fibrosis, and that up-regulation of periostin expression coincided with activation of myofibroblasts positive for α-smooth muscle actin. We established a lung metastasis model for B16 murine melanoma cells and showed that metastatic colonization of the lung by these cells was markedly promoted by bleomycin-induced lung fibrosis. Inhibition of periostin expression by giving an intratracheal antisense oligonucleotide targeting periostin mRNA was found to suppress bleomycin-induced lung fibrosis and thereby to attenuate metastatic colonization of the lung by melanoma cells. Our results indicate that periostin is a key player in the development of bleomycin-induced fibrosis and consequent enhancement of tumor cell colonization in the lung. Our results therefore implicate periostin as a potential target for prevention or treatment of lung metastasis.

Mutagenic consequences of cytosine alterations site-specifically embedded in the human genome.

INTRODUCTION: Cytosine residues in CpG dinucleotides often undergo various types of modification, such as methylation, deamination, and halogenation. These types of modifications can be pro-mutagenic and can contribute to the formation of mutational hotspots in cells. To analyze mutations induced by DNA modifications in the human genome, we recently developed a system for tracing DNA adducts in targeted mutagenesis (TATAM). In this system, a modified/damaged base is site-specifically introduced into intron 4 of thymidine kinase genes in human lymphoblastoid cells. To further the understanding of the mutagenesis of cytosine modification, we directly introduced different types of altered cytosine residues into the genome and investigated their genomic consequences using the TATAM system. FINDINGS: In the genome, the pairing of thymine and 5-bromouracil with guanine, resulting from the deamination of 5-methylcytosine and 5-bromocytosine, respectively, was highly pro-mutagenic compared with the pairing of uracil with guanine, resulting from the deamination of cytosine residues. CONCLUSIONS: The deamination of 5-methylcytosine and 5-bromocytosine rather than that of normal cytosine dramatically enhances the mutagenic potential in the human genome.

Xeroderma Pigmentosum Group A Suppresses Mutagenesis Caused by Clustered Oxidative DNA Adducts in the Human Genome.

Clustered DNA damage is defined as multiple sites of DNA damage within one or two helical turns of the duplex DNA. This complex damage is often formed by exposure of the genome to ionizing radiation and is difficult to repair. The mutagenic potential and repair mechanisms of clustered DNA damage in human cells remain to be elucidated. In this study, we investigated the involvement of nucleotide excision repair (NER) in clustered oxidative DNA adducts. To identify the in vivo protective roles of NER, we established a human cell line lacking the NER gene xeroderma pigmentosum group A (XPA). XPA knockout (KO) cells were generated from TSCER122 cells derived from the human lymphoblastoid TK6 cell line. To analyze the mutagenic events in DNA adducts in vivo, we previously employed a system of tracing DNA adducts in the targeted mutagenesis (TATAM), in which DNA adducts were site-specifically introduced into intron 4 of thymidine kinase genes. Using the TATAM system, one or two tandem 7,8-dihydro-8-oxoguanine (8-oxoG) adducts were introduced into the genomes of TSCER122 or XPA KO cells. In XPA KO cells, the proportion of mutants induced by a single 8-oxoG (7.6%) was comparable with that in TSCER122 cells (8.1%). In contrast, the lack of XPA significantly enhanced the mutant proportion of tandem 8-oxoG in the transcribed strand (12%) compared with that in TSCER122 cells (7.4%) but not in the non-transcribed strand (12% and 11% in XPA KO and TSCER122 cells, respectively). By sequencing the tandem 8-oxoG-integrated loci in the transcribed strand, we found that the proportion of tandem mutations was markedly increased in XPA KO cells. These results indicate that NER is involved in repairing clustered DNA adducts in the transcribed strand in vivo.

Mechanism of induction of binucleated cells by multiwalled carbon nanotubes as revealed by live-cell imaging analysis.

INTRODUCTION: Asbestos-induced formation of mesothelioma has been attributed to phenotypic and morphological changes in cells caused by polyploidization and aneuploidization, and multiwalled carbon nanotubes (MWCNTs) are suspected to have similar adverse effects due to the similarity in their physical form. MWCNTs and crocidolite, a kind of asbestos, show similar genotoxicity characteristics in vitro, including induction of binucleated cells. We here focused on the mechanisms underlying polyploidization during cell division on exposure to MWCNTs and conducted confocal live-cell imaging analysis using MDA-435 human breast cancer cells in which chromosomes and centromeres were visualized using fluorescent proteins. FINDINGS: During anaphase, relatively short MWCNT fibers (approximately 5 µm) migrated rapidly to either of the daughter cells, whereas some long MWCNT fibers (approximately 20 µm) remained inside the contractile ring and induced the formation of binucleated cells through impairment of cytokinesis. This toxicity mechanism has also been observed with crocidolite. CONCLUSIONS: Our findings indicate that the mechanism of polyploidization by MWCNTs is very similar to that observed with crocidolite.

Tracing the fates of site-specifically introduced DNA adducts in the human genome.

We developed a system for tracing DNA adducts in targeted mutagenesis (TATAM) and investigated the prevalence and types of consequent mutations. Targeted mutagenesis methods site-specifically replace endogenous DNA bases with bases carrying synthetic adducts using targeting vectors. The TATAM system was enabled by introduction of site-specific DNA double strand breaks (DSB), which strongly enhanced targeting efficiency through homologous recombination (HR), and a new polymerase chain reaction-based technique, which gives high yields of the target vectors carrying DNA adducts. Human lymphoblastoid TSCER122 cells are compound heterozygous for the thymidine kinase gene (TK-/-), and have a homing endonuclease I-SceI site in intron 4 of the TK gene. The TATAM system enabled targeting of the TK- allele with the I-SceI site using a synthetic TK+ allele containing an 8-oxo-7,8-dihydroguanine (8-oxoG) adduct, a typical product of oxidative DNA damage. The targeted clones (TK+/-) were then isolated by drug selection. Site-specific HR for DSB induced by I-SceI improved targeted integration of the synthetic allele by five orders of magnitude (from 10(-7) to 10(-2)). Subsequent analyses of approximately 800 target clones revealed that 8-oxoG was restored to G in 86% clones, probably reflecting base excision repair or translesion synthesis without mutation. Lesions of the remaining clones (14%) were associated with mutations. The mutation spectrum corresponded closely with that of oxidative DNA damage inducers reported, in which G:C to T:A transversions (5.9%) were predominant. Over-expression of MutY homologs in cells, which prevents G:C to T:A transversions by removing 8-oxoG:A mispairing, significantly decreased the frequency of mutations to 2.6%, indicating that the 8-oxoG adducts introduced by the TATAM system are processed in the same manner as those generated by oxidative DNA damage.

Evaluation of in vivo genotoxicity induced by N-ethyl-N-nitrosourea, benzo[a]pyrene, and 4-nitroquinoline-1-oxide in the Pig-a and gpt assays.

The recently developed Pig-a mutation assay is based on flow cytometric enumeration of glycosylphosphatidylinositol (GPI) anchor-deficient red blood cells caused by a forward mutation in the Pig-a gene. Because the assay can be conducted in nontransgenic animals and the mutations accumulate with repeat dosing, we believe that the Pig-a assay could be integrated into repeat-dose toxicology studies and provides an alternative to transgenic rodent (TGR) mutation assays. The capacity and characteristics of the Pig-a assay relative to TGR mutation assays, however, are unclear. Here, using transgenic gpt delta mice, we compared the in vivo genotoxicity of single oral doses of N-ethyl-N-nitrosourea (ENU, 40 mg/kg), benzo[a]pyrene (BP, 100 and 200 mg/kg), and 4-nitroquinoline-1-oxide (4NQO, 50 mg/kg) in the Pig-a (peripheral blood) and gpt (bone marrow and liver) gene mutation assays. Pig-a assays were conducted at 2, 4, and 7 weeks after the treatment, while gpt assays were conducted on tissues collected at the 7-week terminal sacrifice. ENU increased both Pig-a and gpt mutant frequencies (MFs) at all sampling times, and BP increased MFs in both assays but the Pig-a MFs peaked at 2 weeks and then decreased. Although 4NQO increased gpt MFs in the liver, only weak, nonsignificant increases (two- or threefold above control) were detected in the bone marrow in both the Pig-a and the gpt assay. These findings suggest that further studies are needed to elucidate the kinetics of the Pig-a mutation assay in order to use it as an alternative to the TGR mutation assay.

Miscoding properties of 8-chloro-2'-deoxyguanosine, a hypochlorous acid-induced DNA adduct, catalysed by human DNA polymerases.

Many chronic inflammatory conditions are associated with an increased risk of cancer development. At the site of inflammation, cellular DNA is damaged by hypochlorous acid (HOCl), a potent oxidant generated by myeloperoxidase. 8-Chloro-2'-deoxyguanosine (8-Cl-dG) is a major DNA adduct formed by HOCl and has been detected from the liver DNA and urine of rats administered lipopolysaccharide in an inflammation model. Thus, the 8-Cl-dG lesion may be associated with the carcinogenesis of inflamed tissues. In this study, we explored the miscoding properties of the 8-Cl-dG adduct generated by human DNA polymerases (pols). Site-specifically modified oligodeoxynucleotide containing a single 8-Cl-dG was prepared and used as a template in primer extension reactions catalysed by human pol α, ĸ or η. Primer extension reactions catalysed by pol α and ĸ in the presence of all four dNTPs were slightly retarded at the 8-Cl-dG site, while pol η readily bypassed the lesion. The fully extended products were analysed to quantify the miscoding frequency and specificity of 8-Cl-dG using two-phased polyacrylamide gel electrophoresis (PAGE). During the primer extension reaction in the presence of four dNTPs, pol ĸ promoted one-base deletion (6.4%), accompanied by the misincorporation of 2'-deoxyguanosine monophosphate (5.5%), dAMP (3.7%), and dTMP (3.5%) opposite the lesion. Pol α and η, on the other hand, exclusively incorporated dCMP opposite the lesion. The steady-state kinetic studies supported the results obtained from the two-phased PAGE assay. These results indicate that 8-Cl-dG is a mutagenic lesion; the miscoding frequency and specificity varies depending on the DNA polymerase used. Thus, HOCl-induced 8-Cl-dG adduct may be involved in inflammation-driven carcinogenesis.