Yoshihiro Jinno

Haloperidol treatment induces tissue- and sex-specific changes in DNA methylation: a control study using rats

BackgroundWe previously found that there is a subtle difference in the global methylation state of blood leukocyte DNA between male subjects with and without schizophrenia. The aim of the current study was to determine whether this difference was a primary effect of the disease state, or a secondary effect of antipsychotics administered to these patients.MethodsWe examined the methyl cytosine (mC) content of DNA from the leukocytes, brain, and liver of rats using high performance liquid chromatography. A total of 40 male and female rats received for 21 days daily injection of haloperidol or vehicle solution alone.ResultsIn control rats injected with buffer only, there was a sex-dependent difference in mC content in leukocyte DNA (male > female; P = 0.028, n = 10), similar to our previous observations in human peripheral leukocytes. No difference in mC content between the sexes was observed in the brain or liver in buffer-treated animals. Haloperidol treatment slightly decreased the mC content of leukocytes in male rats, but unexpectedly, increased the mC content of leukocytes in females. We observed a trend toward a higher level of mC in the liver in both sexes following haloperidol treatment, compared to buffer-treated animals. In contrast, haloperidol treatment resulted in a decrease in mC content in the brain in females, and this difference was statistically significant (P = 0.026).ConclusionThese results indicate that haloperidol can affect DNA methylation states in the brain, as well as in certain other tissues, and raise the possibility that antipsychotic drugs play a role in the observed disparity in mC content in male subjects with and without schizophrenia.

Characterization of the gene encoding human pituitary-specific transcription factor, Pit-1

Pit-1 is a pituitary-specific transcription factor that binds to and transactivates promoters of growth hormone- and prolactin-encoding genes. A chromosomal gene related to human Pit-1 isolated from human gene libraries was over 14 kb long and split into six exons. All of the splice donor and acceptor sites conformed to the GT/AG rule. The gene was mapped to human chromosome region 3p11.

Chromosome-band-specific painting : chromosome in situ suppression hybridization using PCR products from a microdissected chromosome band as a probe pool

SummaryWe describe a chromosome-band-specific painting method that involves (1) microdissection of the chromosome, chromosomal region or band, (2) amplification of a variety of chromosome/region/band-specific DNA fragments with the polymerase chain reaction (PCR), and (3) chromosome in situ suppression hybridization (CISS) with the direct use of the PCR products as a probe pool. With this method, it was possible 1) to paint an entire X or Y chromosome, a distal one-fourth of 2q, and only a band at 8q24.1, 2) to identify the origin of a minute marker chromosome in a mentally retarded patient, 3) to detect an X;Y translocation in another patient, and 4) to identify one human chromosome 2 in a human-mouse hybrid cell line. This method allows us to identify not only structural chromosome abnormalities at the band level, but also the origin of cytogenetically unidentifiable marker chromosomes. It will also be useful in studies of evolutionary cytogenetics.

The origin of cytologically unidentifiable chromosome abnormalities: six cases ascertained by targeted chromosome-band painting.

De novo chromosome structural abnormalities cannot always be diagnosed by the use of standard cytogenetic techniques. We applied a previously developed chromosome-band-specific painting method to the diagnosis of such rearrangements. The diagnostic procedures consisted of microdissection of an aberrant chromosomal region of a given patient, polymerase chain reaction (PCR) amplification of the dissected chromosomal DNA, and subsequent competitive fluorescence in situ hybridization (FISH) using the PCR products as a probe pool on metaphase chromosomes from the patient and/or a karyotypically normal person. With this strategy, we studied 6 de novo rearrangements (6p+, 6q+, 9p+, 17p+, +mar, and +mar) in 6 patients. These rearrangements had been seen by conventional banding but their origin could not be identified. In all 6 patients, we successfully ascertained the origin. Using an aberrant region-specific probe pool, FISH signals appeared on both the aberrant region and a region of another specific chromosome pair. A reverse probe pool that was generated through the microdissection of normal chromosomes at a candidate region for the origin of the aberration hybridized with both the aberrant and the candidate regions. We thus diagnosed one patient with 17p+ as having trisomy for 14q32-qter, one with 9p+ as having trisomy for 12pter-p12, one with 6q+ as having a tandem duplication (trisomy) of a 6q23-q25 segment, one with 6p+ as having a tandem duplication (trisomy) of a 6p23-q21.3 segment, one with a supernumerary metacentric marker chromosome as having tetrasomy for 18pter-cen, and the last with an additional small marker chromosome as having trisomy for 18p11.1 (or p11.2)-q11.2. The present targeted chromosome-band-painting method provides the simple and rapid preparation of a probe pool for region-specific FISH, and is useful for the diagnosis of chromosome abnormalities of unknown origin.

The Human ASCL2 Gene Escaping Genomic Imprinting and Its Expression Pattern

The mouse achaete-scute homolog-2 gene (Ascl2 or Mash2) encodes a transcription factor playing a role in the development of the trophoblast. The Ascl2 is an imprinted gene with maternal expression and assigned to an imprinting gene cluster region (ICR) at a distal region of mouse chromosome 7. We previously isolated a phage clone carrying the human homolog, ASCL2, and mapped it to human chromosome 11p15.5, a human ICR. In the present study, we demonstrate the expression patterns of the human ASCL2 in the fetus at a stage between first and second trimesters and in the placental tissues. In addition, it has been shown that the human ASCL2 gene escapes genomic imprinting.

Human endogenous retroviruses with transcriptional potential in the brain

AbstractGenetic studies of neuropsychiatric disorders have often produced conflicting results, which might partly result from the involvement of epigenetic modifications. We intended to explore the possible implication of DNA methylation and human endogenous retroviruses (HERVs) in neuropsychiatric disorders. In the present study, we identified two HERV loci that are expected to retain the transcriptional activity in the brain. One was located on chromosome 1q21-q22 and the other on 22q12. Interestingly, these regions were overlapped with or included in those of schizophrenia-susceptible loci, SCZD9 and SCZD4, respectively. Particularly, the HERV on 22q12 was located in the opposite direction 4 kb downstream of the Synapsin III gene. These HERV loci could afford clear targets for methylation and expression analyses in postmortem brains of patients with psychiatric disorders such as schizophrenia. In addition, we confirmed our previous finding that only a few of particular HERV-K loci were activated among a number of highly homologous loci in teratocarcinoma cell lines. These activated loci included ones common to all teratocarcinoma cell lines analyzed and depending on their male or female origin.

Isolation and localization of an IDDMK1,2-22-related human endogenous retroviral gene, and identification of a CA repeat marker at its locus.

AbstractWe intended to confirm genetically the involvement of the IDDMK1,2-22 gene in the pathogenesis of insulin-dependent diabetes mellitus (IDDM). For this purpose, we isolated a human endogenous retrovirus gene, possibly corresponding to IDDMK1,2-22. The isolated gene showed 99.8% and 99.7% homologies in nucleotide sequences to a part of the env region and of the 3′-LTR region, respectively, compared to those of IDDMK1,2-22 deposited in GenBank. The gene also showed a close relation to HERV-K18, of which the 3′-LTR sequence gave 99.5% homology. It seemed likely that these genes represented the same single gene. The newly isolated gene was present in the first intron of the CD48 gene and was located on chromosome 1q21.2–q22. A CA repeat marker was found approximately 20kb upstream from the 5′-end of the 5′-LTR of the gene.

Methylation imprinting of H19 and SNRPN genes in human benign ovarian teratomas.

In humans, studies of female germ cells are very limited by ethics. The current study investigated the usefulness of benign ovarian teratomas as a substitute for ova in analyses of imprinted genes. Twenty-five human benign ovarian teratomas were typed with 45 microsatellite DNA markers and classified according to their genotypic features. Two oppositely imprinted genes, H19 and SNRPN, were then chosen for analysis of their methylation states in these tumors. These analyses revealed that benign ovarian teratomas consist of a mixture of genetically and epigenetically heterogeneous cell populations. In contrast to previous reports, we could document only one case rising from germ cells by meiosis-II nondisjunction. H19 and SNRPN were methylated in individual teratomas to various degrees, ranging from normal somatic cell to expected ovum levels. The allele with residual methylation of H19 was consistent with that methylated in the patients blood DNA, thus being of paternal origin. Degrees of H19 hypomethylation and SNRPN hypermethylation increased as the cellular origin of the tumors advanced in oogenesis and were closely correlated in individual teratomas. These results could be best explained by the assumption that the primary imprinting is a progressively organized process and suggest that the establishment of primary imprints on different genes might be mechanistically linked, even when those genes are oppositely imprinted.

Promoter-specific insulin-like growth factor 2 gene imprinting in human fetal liver and hepatoblastoma.

Insulin‐like growth factor 2 (IGF2) gene imprinting has been demonstrated to be promoter‐specific, in that expression from the P1 promoter is biallelic whereas that from the P2–P4 promoters is monoallelic. In the present study, in order to investigate IGF2 gene imprinting status at the cellular level, allelic analysis was performed of IGF2 gene expression transcribed from the P1 and P3 promoters, using reverse transcription polymerase chain reaction (RT‐PCR) on human fetal liver and hepatoblastoma. In situ hybridization was also undertaken, to obtain information about the cellular localization of transcripts expressed from the P1 and P3 promoters. The results indicated that transcripts expressed from the P1 and P3 promoters co‐localized in the same fetal or neoplastic hepatocytes. These data should provide information regarding the molecular basis of genomic imprinting, suggesting that an imprint recognized for the differential expression may be strictly local and localized downstream of the IGF2 P1 promoter.

Tissue specificity of methylation and expression of human genes coding for neuropeptides and their receptors, and of a human endogenous retrovirus K family

AbstractThe purpose of the present study was to understand the tissue specificity of DNA methylation and the relationship between methylation and expression of genes with essential roles in neurodevelopment and brain function. We chose dopamine receptor genes (DRD1 and DRD2), NCAM, and COMT as examples of genes with CpG islands around the promoter region, and serotonin receptor genes (HTR2A and HTR3A), HCRT, and DRD3 as genes without CpG islands. Methylation states were investigated in fetal brain, fetal liver, placenta, and in adult peripheral leukocytes from three individuals by Southern blot and bisulfite-modified DNA sequencing. A repetitive sequence, human endogenous retrovirus (HERV)-K was also examined. All genes examined were almost completely unmethylated in brains. The genes with CpG islands were unmethylated regardless of their expression state. In contrast, genes without CpG islands showed various methylation patterns, which did not necessarily reflect the transcriptional activity of the genes. Most HERV-K loci were methylated, but some loci showed relatively low methylation in the placenta and liver. Interestingly, we found inter-individual differences in methylation levels in HTR2A and HCRT in the placenta and in some loci of HERV-K in the placenta and liver. The sample with the lowest methylation levels in the two unique genes showed higher methylation of HERV-K loci than the other samples. These results provide detailed information about the methylation states of the genes analyzed and evidence for inter-individual variations in methylation in both unique and repetitive sequences.