CAR-Ts redirected against the Thomsen-Friedenreich antigen CD176 mediate specific elimination of malignant cells from leukemia and solid tumors.

Introduction: Chimeric antigen receptor-engineered T cells (CAR-Ts) are investigated in various clinical trials for the treatment of cancer entities beyond hematologic malignancies. A major hurdle is the identification of a target antigen with high expression on the tumor but no expression on healthy cells, since "on-target/off-tumor" cytotoxicity is usually intolerable. Approximately 90% of carcinomas and leukemias are positive for the Thomsen-Friedenreich carbohydrate antigen CD176, which is associated with tumor progression, metastasis and therapy resistance. In contrast, CD176 is not accessible for ligand binding on healthy cells due to prolongation by carbohydrate chains or sialylation. Thus, no "on-target/off-tumor" cytotoxicity and low probability of antigen escape is expected for corresponding CD176-CAR-Ts. Methods: Using the anti-CD176 monoclonal antibody (mAb) Nemod-TF2, the presence of CD176 was evaluated on multiple healthy or cancerous tissues and cells. To target CD176, we generated two different 2nd generation CD176-CAR constructs differing in spacer length. Their specificity for CD176 was tested in reporter cells as well as primary CD8+ T cells upon co-cultivation with CD176+ tumor cell lines as models for CD176+ blood and solid cancer entities, as well as after unmasking CD176 on healthy cells by vibrio cholerae neuraminidase (VCN) treatment. Following that, both CD176-CARs were thoroughly examined for their ability to initiate target-specific T-cell signaling and activation, cytokine release, as well as cytotoxicity. Results: Specific expression of CD176 was detected on primary tumor tissues as well as on cell lines from corresponding blood and solid cancer entities. CD176-CARs mediated T-cell signaling (NF-κB activation) and T-cell activation (CD69, CD137 expression) upon recognition of CD176+ cancer cell lines and unmasked CD176, whereby a short spacer enabled superior target recognition. Importantly, they also released effector molecules (e.g. interferon-γ, granzyme B and perforin), mediated cytotoxicity against CD176+ cancer cells, and maintained functionality upon repetitive antigen stimulation. Here, CD176L-CAR-Ts exhibited slightly higher proliferation and mediator-release capacities. Since both CD176-CAR-Ts did not react towards CD176- control cells, their response proved to be target-specific. Discussion: Genetically engineered CD176-CAR-Ts specifically recognize CD176 which is widely expressed on cancer cells. Since CD176 is masked on most healthy cells, this antigen and the corresponding CAR-Ts represent a promising approach for the treatment of various blood and solid cancers while avoiding "on-target/off-tumor" cytotoxicity.

Mixed-Up Sugars: Glycosyltransferase Cross-Reactivity in Cancerous Tissues and Their Therapeutic Targeting.

The main categories of glycan changes in cancer are: (1) decreased expression of histo-blood group A and/or B antigens and increased Lewis-related antigens, (2) appearance of cryptic antigens, such as Tn and T, (3) emergence of genetically incompatible glycans, such as A antigen expressed in tumors of individuals of group B or O and heterophilic expression of Forssman antigen (FORS1), and (4) appearance of neoglycans. This review focuses on the expression of genetically incompatible A/B/FORS1 antigens in cancer. Several possible molecular mechanisms are exemplified, including missense mutations that alter the sugar specificity of A and B glycosyltransferases (AT and BT, respectively), restoration of the correct codon reading frame of O alleles, and modification of acceptor specificity of AT to synthesize the FORS1 antigen by missense mutations and/or altered splicing. Taking advantage of pre-existing natural immunity, the potential uses of these glycans for immunotherapeutic targeting will also be discussed.

Recent advances on smart glycoconjugate vaccines in infections and cancer.

Vaccination is one of the greatest achievements in biomedical research preventing death and morbidity in many infectious diseases through the induction of pathogen-specific humoral and cellular immune responses. Currently, no effective vaccines are available for pathogens with a highly variable antigenic load, such as the human immunodeficiency virus or to induce cellular T-cell immunity in the fight against cancer. The recent SARS-CoV-2 outbreak has reinforced the relevance of designing smart therapeutic vaccine modalities to ensure public health. Indeed, academic and private companies have ongoing joint efforts to develop novel vaccine prototypes for this virus. Many pathogens are covered by a dense glycan-coat, which form an attractive target for vaccine development. Moreover, many tumor types are characterized by altered glycosylation profiles that are known as "tumor-associated carbohydrate antigens". Unfortunately, glycans do not provoke a vigorous immune response and generally serve as T-cell-independent antigens, not eliciting protective immunoglobulin G responses nor inducing immunological memory. A close and continuous crosstalk between glycochemists and glycoimmunologists is essential for the successful development of efficient immune modulators. It is clear that this is a key point for the discovery of novel approaches, which could significantly improve our understanding of the immune system. In this review, we discuss the latest advancements in development of vaccines against glycan epitopes to gain selective immune responses and to provide an overview on the role of different immunogenic constructs in improving glycovaccine efficacy.

Emerging glyco-based strategies to steer immune responses.

Glycan structures are common posttranslational modifications of proteins, which serve multiple important structural roles (for instance in protein folding), but also are crucial participants in cell-cell communications and in the regulation of immune responses. Through the interaction with glycan-binding receptors, glycans are able to affect the activation status of antigen-presenting cells, leading either to induction of pro-inflammatory responses or to suppression of immunity and instigation of immune tolerance. This unique feature of glycans has attracted the interest and spurred collaborations of glyco-chemists and glyco-immunologists to develop glycan-based tools as potential therapeutic approaches in the fight against diseases such as cancer and autoimmune conditions. In this review, we highlight emerging advances in this field, and in particular, we discuss on how glycan-modified conjugates or glycoengineered cells can be employed as targeting devices to direct tumor antigens to lectin receptors on antigen-presenting cells, like dendritic cells. In addition, we address how glycan-based nanoparticles can act as delivery platforms to enhance immune responses. Finally, we discuss some of the latest developments in glycan-based therapies, including chimeric antigen receptor (CAR)-T cells to achieve targeting of tumor-associated glycan-specific epitopes, as well as the use of glycan moieties to suppress ongoing immune responses, especially in the context of autoimmunity.

ABO blood group A transferase and its codon 69 substitution enzymes synthesize FORS1 antigen of FORS blood group system.

Human histo-blood group A transferase (AT) catalyzes the biosynthesis of oligosaccharide A antigen important in blood transfusion and cell/tissue/organ transplantation. This enzyme may synthesize Forssman antigen (FORS1) of the FORS blood group system when exon 3 or 4 of the AT mRNA is deleted and/or the LeuGlyGly tripeptide at codons 266-268 of AT is replaced by GlyGlyAla. The Met69Ser/Thr substitutions also confer weak Forssman glycolipid synthase (FS) activity. In this study, we prepared the human AT derivative constructs containing any of the 20 amino acids at codon 69 with and without the GlyGlyAla substitution, transfected DNA to newly generated COS1(B3GALNT1 + A4GALT) cells expressing an enhanced level of globoside (Gb4), the FS acceptor substrate, and immunologically examined the FORS1 expression. Our results showed that all those substitution constructs at codon 69 exhibited FS activity. The combination with GlyGlyAla significantly increased the activity. The conserved methionine residue in the ABO, but not GBGT1, gene-encoded proteins may implicate its contribution to the separation of these genes in genetic evolution. Surprisingly, with increased Gb4 availability, the original human AT with the methionine residue at codon 69 was also demonstrated to synthesize FORS1, providing another molecular mechanism of FORS1 appearance in cancer of ordinary FORS1-negative individuals.

Methylation-Sensitive Amplification Length Polymorphism (MS-AFLP) Microarrays for Epigenetic Analysis of Human Genomes.

Somatic, and in a minor scale also germ line, epigenetic aberrations are fundamental to carcinogenesis, cancer progression, and tumor phenotype. DNA methylation is the most extensively studied and arguably the best understood epigenetic mechanisms that become altered in cancer. Both somatic loss of methylation (hypomethylation) and gain of methylation (hypermethylation) are found in the genome of malignant cells. In general, the cancer cell epigenome is globally hypomethylated, while some regions-typically gene-associated CpG islands-become hypermethylated. Given the profound impact that DNA methylation exerts on the transcriptional profile and genomic stability of cancer cells, its characterization is essential to fully understand the complexity of cancer biology, improve tumor classification, and ultimately advance cancer patient management and treatment. A plethora of methods have been devised to analyze and quantify DNA methylation alterations. Several of the early-developed methods relied on the use of methylation-sensitive restriction enzymes, whose activity depends on the methylation status of their recognition sequences. Among these techniques, methylation-sensitive amplification length polymorphism (MS-AFLP) was developed in the early 2000s, and successfully adapted from its original gel electrophoresis fingerprinting format to a microarray format that notably increased its throughput and allowed the quantification of the methylation changes. This array-based platform interrogates over 9500 independent loci putatively amplified by the MS-AFLP technique, corresponding to the NotI sites mapped throughout the human genome.

ABO blood group A transferases catalyze the biosynthesis of FORS blood group FORS1 antigen upon deletion of exon 3 or 4.

Evolutionarily related ABO and GBGT1 genes encode, respectively, A and B glycosyltransferases (AT and BT) and Forssman glycolipid synthase (FS), which catalyze the biosynthesis of A and B, and Forssman (FORS1) oligosaccharide antigens responsible for the ABO and FORS blood group systems. Humans are a Forssman antigen-negative species; however, rare individuals with Apae phenotype express FORS1 on their red blood cells. We previously demonstrated that the replacement of the LeuGlyGly tripeptide sequence at codons 266 to 268 of human AT with GBGT1-encoded FS-specific GlyGlyAla enabled the enzyme to produce FORS1 antigen, although the FS activity was weak. We searched for additional molecular mechanisms that might allow human AT to express FORS1. A variety of derivative expression constructs of human AT were prepared. DNA was transfected into COS1 (B3GALNT1) cells, and cell-surface expression of FORS1 was immunologically monitored. To our surprise, the deletion of exon 3 or 4, but not of exon 2 or 5, of human AT transcripts bestowed moderate FS activity, indicating that the A allele is inherently capable of producing a protein with FS activity. Because RNA splicing is frequently altered in cancer, this mechanism may explain, at least partially, the appearance of FORS1 in human cancer. Furthermore, strong FS activity was attained, in addition to AT and BT activities, by cointroducing 1 of those deletions and the GlyGlyAla substitution, possibly by the synergistic effects of altered intra-Golgi localization/conformation by the former and modified enzyme specificity by the latter.

Murine cell glycolipids customization by modular expression of glycosyltransferases.

Functional analysis of glycolipids has been hampered by their complex nature and combinatorial expression in cells and tissues. We report an efficient and easy method to generate cells with specific glycolipids. In our proof of principle experiments we have demonstrated the customized expression of two relevant glycosphingolipids on murine fibroblasts, stage-specific embryonic antigen 3 (SSEA-3), a marker for stem cells, and Forssman glycolipid, a xenoantigen. Sets of genes encoding glycosyltansferases were transduced by viral infection followed by multi-color cell sorting based on coupled expression of fluorescent proteins.

ABO research in the modern era of genomics.

Research on ABO has advanced significantly in recent years. A database was established to manage the sequence information of an increasing number of novel alleles. Genome sequencings have identified ABO orthologues and paralogues in various organisms and enhanced the knowledge on the evolution of the ABO and related genes. The most prominent advancements include clarification of the association between ABO and different disease processes. For instance, ABO status affects the infectivity of certain strains of Helicobacter pylori and Noroviruses as well as the sequestration and rosetting of red blood cells infected with Plasmodium falciparum. Genome-wide association studies have conclusively linked the ABO locus to pancreatic cancer, venous thromboembolism, and myocardial infarction in the presence of coronary atherosclerosis. These findings suggest ABO's important role in determining an individual's susceptibility to such diseases. Furthermore, our understanding of the structures of A and B transferases and their enzymology has been dramatically improved. ABO has also become a research subject in neurobiology and the preparation of artificial/universal blood and became a topic in the pseudoscience of "blood type diets." With such new progress, it has become evident that ABO is a critical player in the modern era of genomic medicine. This article provides the most up-to-date information regarding ABO genomics.

Rare and frequent promoter methylation, respectively, of TSHZ2 and 3 genes that are both downregulated in expression in breast and prostate cancers.

BACKGROUND: Neoplastic cells harbor both hypomethylated and hypermethylated regions of DNA. Whereas hypomethylation is found mainly in repeat sequences, regional hypermethylation has been linked to the transcriptional silencing of certain tumor suppressor genes. We attempted to search for candidate genes involved in breast/prostate carcinogenesis, using the criteria that they should be expressed in primary cultures of normal breast/prostate epithelial cells but are frequently downregulated in breast/prostate cancer cell lines and that their promoters are hypermethylated. METHODOLOGY/PRINCIPAL FINDINGS: We identified several dozens of candidates among 194 homeobox and related genes using Systematic Multiplex RT-PCR and among 23,000 known genes and 23,000 other expressed sequences in the human genome by DNA microarray hybridization. An additional examination, by real-time qRT-PCR of clinical specimens of breast cancer, further narrowed the list of the candidates. Among them, the most frequently downregulated genes in tumors were NP_775756 and ZNF537, from the homeobox gene search and the genome-wide search, respectively. To our surprise, we later discovered that these genes belong to the same gene family, the 3-member Teashirt family, bearing the new names of TSHZ2 and TSHZ3. We subsequently determined the methylation status of their gene promoters. The TSHZ3 gene promoter was found to be methylated in all the breast/prostate cancer cell lines and some of the breast cancer clinical specimens analyzed. The TSHZ2 gene promoter, on the other hand, was unmethylated except for the MDA-MB-231 breast cancer cell line. The TSHZ1 gene was always expressed, and its promoter was unmethylated in all cases. CONCLUSIONS/SIGNIFICANCE: TSHZ2 and TSHZ3 genes turned out to be the most interesting candidates for novel tumor suppressor genes. Expression of both genes is downregulated. However, differential promoter methylation suggests the existence of distinctive mechanisms of transcriptional inactivation for these genes.

DNA fingerprinting techniques for the analysis of genetic and epigenetic alterations in colorectal cancer.

Genetic somatic alterations are fundamental hallmarks of cancer. In addition to point and other small mutations targeting cancer genes, solid tumors often exhibit aneuploidy as well as multiple chromosomal rearrangements of large fragments of the genome. Whether somatic chromosomal alterations and aneuploidy are a driving force or a mere consequence of tumorigenesis remains controversial. Recently it became apparent that not only genetic but also epigenetic alterations play a major role in carcinogenesis. Epigenetic regulation mechanisms underlie the maintenance of cell identity crucial for development and differentiation. These epigenetic regulatory mechanisms have been found substantially altered during cancer development and progression. In this review, we discuss approaches designed to analyze genetic and epigenetic alterations in colorectal cancer, especially DNA fingerprinting approaches to detect changes in DNA copy number and methylation. DNA fingerprinting techniques, despite their modest throughput, played a pivotal role in significant discoveries in the molecular basis of colorectal cancer. The aim of this review is to revisit the fingerprinting technologies employed and the oncogenic processes that they unveiled.

Identification of genes that exhibit changes in expression on the 8p chromosomal arm by the Systematic Multiplex RT-PCR (SM RT-PCR) and DNA microarray hybridization methods.

Losses of the p-arm of chromosome 8 are frequently observed in breast, prostate, and other types of cancers. Using the Systematic Multiplex RT-PCR (SM RT-PCR) method and the DNA microarray hybridization method, we examined the expression of 273 genes located on the p-arm of chromosome 8 in five breast and three prostate human cancer cell lines. We observed frequent decreases in expression of two dozen genes and increases in expression of several genes on this chromosomal arm. These changes in gene expression of the cell lines were later confirmed by real-time qRT-PCR. Additionally and more importantly, we found that a number of these variations were also observed in the majority of clinical cases of breast cancer we examined. These included downregulation of the MYOM2, NP_859074, NP_001034551, NRG1, PHYIP (PHYHIP), Q7Z2R7, SFRP1, and SOX7 genes, and upregulation of the ESCO2, NP_115712 (GINS4), Q6P464, and TOPK (PBK) genes.

Requirement of RIZ1 for cancer prevention by methyl-balanced diet.

BACKGROUND: The typical Western diet is not balanced in methyl nutrients that regulate the level of the methyl donor S-adenosylmethionine (SAM) and its derivative metabolite S-adenosylhomocysteine (SAH), which in turn may control the activity of certain methyltransferases. Feeding rodents with amino acid defined and methyl-imbalanced diet decreases hepatic SAM and causes liver cancers. RIZ1 (PRDM2 or KMT8) is a tumor suppressor and functions in transcriptional repression by methylating histone H3 lysine 9. METHODOLOGY/PRINCIPAL FINDINGS: Here we show that a methyl-balanced diet conferred additional survival benefits compared to a tumor-inducing methyl-imbalanced diet only in mice with wild type RIZ1 but not in mice deficient in RIZ1. While absence of RIZ1 was tumorigenic in mice fed the balanced diet, its presence did not prevent tumor formation in mice fed the imbalanced diet. Microarray and gene expression analysis showed that, unlike most of its related enzymes, RIZ1 was upregulated by methyl-balanced diet. Methyl-balanced diet did not fully repress oncogenes such as c-Jun in the absence of RIZ1. Higher RIZ1 activity was associated with greater H3 lysine 9 methylation in RIZ1 target genes as shown by chromatin immunoprecipitation analysis. CONCLUSIONS/SIGNIFICANCE: The data identify RIZ1 as a critical target of methyl-balanced diet in cancer prevention. The molecular understanding of dietary carcinogenesis may help people make informed choices on diet, which may greatly reduce the incidence of cancer.

Scanning copy number and gene expression on the 18q21-qter chromosomal region by the systematic multiplex PCR and reverse transcription-PCR methods.

We examined differences in copy number and expression of 127 genes located on the 18q21-qter chromosomal region of the breast and prostate cancer cell lines, using the systematic multiplex PCR and reverse transcription-PCR (SM PCR and SM RT-PCR) methods that we developed. Semi-quantitative data were obtained that were comparable in quality, but not in quantity, to data from DNA microarray hybridization analysis. In the chromosomal region where losses are frequent in breast, prostate, and other cancers, we detected a homozygous deletion of the SMAD4 gene in the MDA-MB-468 breast cancer cell line. We also observed partial or entire loss of expression in genes such as CCBE1, CCDC11, CD226, NP_115536.1, NP_689683.2, RNF152, SERPINB8, and TCF4 in certain breast and/or prostate cancer cell lines. An increase in gene expression was rare, but found with the transcription factor ONECUT2 gene in all of the cancer cell lines examined. Real-time qRT-PCR experiments confirmed these SM RT-PCR results. Further analysis of clinical specimens of breast cancer by real-time qRT-PCR demonstrated that the gene expression of CCBE1, TCF4, NP_115536.1, and NP_689683.2 was downregulated in the majority of clinical cases of breast cancer.

Scanning copy number and gene expression on the 16p13.3-13.2 chromosomal region by the systematic multiplex polymerase chain reaction and reverse transcription-polymerase chain reaction methods.

We developed the systematic multiplex reverse transcription-PCR (SM RT-PCR) method that is distinguishable from other multiplex RT-PCR methods by optimized PCR conditions allowing amplification of sequences that fall within a single exon of genes of similar band intensity using genomic DNA template as a calibration standard. Using an SM RT-PCR system of proto-oncogenes and tumor suppressor genes, we previously showed that the SM RT-PCR system, which was developed for cDNA expression analysis, could also be used for a more exquisite analysis of copy number changes in genomic DNA. Here we report that the SM PCR method semiquantitatively detected less than a two-fold difference in copy number. Furthermore, we also report the results of subchromosomal scanning of copy number and expression using the SM PCR and SM RT-PCR methods. We identified and characterized the novel homozygous deletion that spans over 12-plus genes on 16p13.3-13.2 in the MDA-MB-468 breast cancer cell line.

Systematic multiplex polymerase chain reaction and reverse transcription-polymerase chain reaction analyses of changes in copy number and expression of proto-oncogenes and tumor suppressor genes in cancer tissues and cell lines.

Systematic multiplex reverse transcription-polymerase chain reaction (SM RT-PCR) is distinguishable from other multiplex RT-PCR methods by (i) utilization of primers that amplify sequences that fall within a single exon of the genes, (ii) utilization of genomic DNA as a calibration standard, and (iii) optimized PCR conditions that allow amplification of bands of similar intensity using genomic DNA template. We previously developed the human experimental systems of 68 glycosyltransferase genes, 39 Hox genes, and 26 integrin subunit genes, and analyzed the expression of those genes in human adult tissues. Here we report the establishment of an SM RT-PCR system of proto-oncogenes and tumor suppressor genes and the analysis of gene expression in human cancer tissues and cell lines. We also demonstrate that the SM RT-PCR system, which was developed for cDNA expression analysis, could also be used successfully for more exquisite analysis of copy number changes in genomic DNA. We observed a decrease in band intensity of HRAS, TP73, CDKN2A, and CDKN2B genes in most of the breast and prostate cancer cell lines examined. The decrease in copy number of HRAS proto-oncogene leads us to suspect the presence of tumor suppressor genes in the vicinity of this gene on chromosome 11p15.5.

Gene expression analysis of an integrin family of genes by systematic multiplex reverse transcription-polymerase chain reaction.

We have established the systematic multiplex reverse transcription-polymerase chain reaction (SM RT-PCR) system of 26 members of the integrin family of genes, and used this system to investigate their expression in 25 different kinds of human adult tissues. A hierarchical clustering analysis of the tissue expression data illustrated clustering of functionally related tissues. Although the correlation was weak, bone marrow, thymus, and spleen were clustered, potentially confirming the role of certain integrin molecules in the immune response. We also investigated changes in integrin gene expression in cancer. Several differences were observed between normal and tumor, as well as between normal and cancer cell lines. The SM RT-PCR experiments for the study of alternative splicing showed that most of the integrin genes analyzed exhibited fixed ratios of differentially spliced transcripts probably due to a tissue-independent mechanism of splicing predetermined by the nucleotide sequences around the splicing donor and acceptor sites.

Genetics supersedes epigenetics in colon cancer phenotype.

A CpG island DNA methylator phenotype has been postulated to explain silencing of the hMLH1 DNA mismatch repair gene in cancer of the microsatellite mutator phenotype. To evaluate this model, we analyzed methylation in CpG islands from six mutator and suppressor genes, and thirty random genomic sites, in a panel of colorectal cancers. Tumor-specific somatic hypermethylation was a widespread age-dependent process that followed a normal Gaussian distribution. Because there was no discontinuity in methylation rate, our results challenge the methylator phenotype hypothesis and its hypothetical pathological underlying defect. We also show that the mutator phenotype dominates over the gradual accumulation of DNA hypermethylation in determining the genotypic features that govern the phenotypic peculiarities of colon cancer of the mutator pathway.

Transcription starting from an alternative promoter leads to the expression of the human ABO histo-blood group antigen.

BACKGROUND: Using the 5'-rapid amplification of cDNA ends technique with the ex vivo culture of AC133-CD34+ cells, a transcription start site was recently identified approximately 0.7 kb upstream from the transcription start sites previously determined. The transcripts from the alternative starting exon 1a were demonstrated in the cells of both erythroid and epithelial lineages. Because the nucleotide sequence of exon 1a does not contain an ATG codon, we examined whether transcription starting from exon 1a leads to production of a functional glycosyltransferase. STUDY DESIGN AND METHODS: Stable transfection experiments into the human gastric cancer MKN28 cells were performed using the various A transferase expression plasmids. RESULTS: Large amounts of A antigens were demonstrated on the cells transfected with the A transferase expression plasmid containing the entire cDNA from exon 1a or the 5'-truncated cDNA leading to the production of the N-truncated protein with deletion of the cytoplasmic tail and a portion of the transmembrane domain. However, negligible amounts of A antigens were observed on the cells transfected with the A transferase expression plasmids containing the 5'-truncated cDNA leading to the production of the N-truncated proteins without the cytoplasmic tail and the transmembrane domain. CONCLUSION: This study suggests that a functional A transferase could be produced by the transcription from exon 1a.

Alternative promoter identified between a hypermethylated upstream region of repetitive elements and a CpG island in human ABO histo-blood group genes.

We have studied the expression of human histo-blood group ABO genes during erythroid differentiation, using an ex vivo culture of AC133(-)CD34(+) cells obtained from peripheral blood. 5'-Rapid amplification of cDNA ends analysis of RNA from those cells revealed a novel transcription start site, which appeared to mark an alternative starting exon (1a) comprising 27 bp at the 5'-end of a CpG island in ABO genes. Results from reverse transcription-PCR specific to exon 1a indicated that the cells of both erythroid and epithelial lineages utilize this exon as the transcription starting exon. Transient transfection experiments showed that the region just upstream from the transcription start site possesses promoter activity in a cell type-specific manner when placed 5' adjacent to the reporter luciferase gene. Results from bisulfite genomic sequencing and reverse transcription-PCR analysis indicated that hypermethylation of the distal promoter region correlated with the absence of transcripts containing exon 1a, whereas hypermethylation in the interspersed repeats 5' adjacent to the distal promoter was commonly observed in all of the cell lines examined. These results suggest that a functional alternative promoter is located between the hypermethylated region of repetitive elements and the CpG island in the ABO genes.