Effects of genome architecture and epigenetic factors on susceptibility of promoter CpG islands to aberrant DNA methylation induction

Aberrant DNA methylation is induced at specific promoter CpG islands (CGIs) in contrast with mutations. The specificity is influenced by genome architecture and epigenetic factors, but their relationship is still unknown. In this study, we isolated promoter CGIs susceptible and resistant to aberrant methylation induction during prostate and breast carcinogenesis. The effect of genome architecture was more evident for promoter CGIs susceptible in both of the two tissues than for promoter CGIs susceptible only in one tissue. Multivariate analysis of promoter CGIs with tissue-nonspecific susceptibility showed that genome architecture, namely a remote location from SINE (OR = 5.98; 95% CI = 2.33-15.34) and from LINE (OR = 2.08; 95% CI = 1.03-4.21), was associated with increased susceptibility, independent of epigenetic factors such as the presence of RNA polymerase II (OR = 0.09; 95% CI = 0.02-0.48) and H3K27me3 (OR = 3.28; 95% CI = 1.17-9.21). These results showed that methylation susceptibility of promoter CGIs is determined both by genome architecture and epigenetic factors, independently.     

Suppressed expression of NDRG2 correlates with poor prognosis in pancreatic cancer

Pancreatic cancer is a highly lethal disease with a poor prognosis; the molecular mechanisms of the development of this disease have not yet been fully elucidated. N-myc downstream regulated gene 2 (NDRG2), one of the candidate tumor suppressor genes, is frequently downregulated in pancreatic cancer, but there has been little information regarding its expression in surgically resected pancreatic cancer specimens. We investigated an association between NDRG2 expression and prognosis in 69 primary resected pancreatic cancer specimens by immunohistochemistry and observed a significant association between poor prognosis and NDRG2-negative staining (P = 0.038). Treatment with trichostatin A, a histone deacetylase inhibitor, predominantly up-regulated NDRG2 expression in the NDRG2 low-expressing cell lines (PANC-1, PCI-35, PK-45P, and AsPC-1). In contrast, no increased NDRG2 expression was observed after treatment with 5-aza-2′ deoxycytidine, a DNA demethylating agent, and no hypermethylation was detected in either pancreatic cancer cell lines or surgically resected specimens by methylation specific PCR. Our present results suggest that (1) NDRG2 is functioning as one of the candidate tumor-suppressor genes in pancreatic carcinogenesis, (2) epigenetic mechanisms such as histone modifications play an essential role in NDRG2 silencing, and (3) the expression of NDRG2 is an independent prognostic factor in pancreatic cancer.     

Identification and functional analysis of a novel mutation in the SOX10 gene associated with Waardenburg syndrome type IV

Waardenburg syndrome type IV (WS4) is a rare genetic disorder, characterized by auditory–pigmentary abnormalities and Hirschsprung disease. Mutations of the EDNRB gene, EDN3 gene, or SOX10 gene are responsible for WS4. In the present study, we reported a case of a Chinese patient with clinical features of WS4. In addition, the three genes mentioned above were sequenced in order to identify whether mutations are responsible for the case. We revealed a novel nonsense mutation, c.1063C>T (p.Q355*), in the last coding exon of SOX10. The same mutation was not found in three unaffected family members or 100 unrelated controls. Then, the function and mechanism of the mutation were investigated in vitro. We found both wild-type (WT) and mutant SOX10 p.Q355* were detected at the expected size and their expression levels are equivalent. The mutant protein also localized in the nucleus and retained the DNA-binding activity as WT counterpart; however, it lost its transactivation capability on the MITF promoter and acted as a dominant-negative repressor impairing function of the WT SOX10.     

Mutation and expression analysis of the IDH1, IDH2, DNMT3A, and MYD88 genes in colorectal cancer

Colorectal cancer (CRC) is one of the leading causes of death around the world. Its genetic mechanism was intensively investigated in the past decades with findings of a number of canonical oncogenes and tumor-suppressor genes such as APC, KRAS, and TP53. Recent genome-wide association and sequencing studies have identified a series of promising oncogenes including IDH1, IDH2, DNMT3A, and MYD88 in hematologic malignancies. However, whether these genes are involved in CRC remains unknown. In this study, we screened the hotspot mutations of these four genes in 305 CRC samples from Han Chinese by direct sequencing. mRNA expression levels of these genes were quantified by quantitative real-time PCR (RT-qPCR) in paired cancerous and paracancerous tissues. Association analyses between mRNA expression levels and different cancerous stages were performed. Except for one patient harboring IDH1 mutation p.I99M, we identified no previously reported hotspot mutations in colorectal cancer tissues. mRNA expression levels of IDH1, DNMT3A, and MYD88, but not IDH2, were significantly decreased in the cancerous tissues comparing with the paired paracancerous normal tissues. Taken together, the hotspot mutations of IDH1, IDH2, DNMT3A, and MYD88 gene were absent in CRC. Aberrant mRNA expression of IDH1, DNMT3A, and MYD88 gene might be actively involved in the development of CRC.     

Evolutionary aspects of a unique internal mitochondrial targeting signal in nuclear-migrated rps19 of sugar beet (Beta vulgaris L.)

The endosymbiotic theory postulates that many genes migrated from endosymbionts to the nuclear genomes of their hosts. Some migrated genes lack presequences directing proteins to mitochondria, and their mitochondrial targeting signals appear to be inscribed in the core coding regions as internal targeting signals (ITSs). ITSs may have evolved after sequence transfer to nuclei or ITSs may have pre-existed before sequence transfer. Here, we report the molecular cloning of a sugar beet gene for ribosomal protein S19 (Rps19; the first letter is capitalized when the gene is a nuclear gene). We show that sugar beet Rps19 (BvRps19) is an ITS-type gene. Based on amino-acid sequence comparison, dicotyledonous rps19s (the first letter is lower-cased when the gene is a mitochondrial gene), such as tobacco rps19 (Ntrps19), resemble an ancestral form of BvRps19. We investigated whether differences in amino-acid sequences between BvRps19 and Ntrps19 were involved in ITS evolution. Analyses of the intracellular localization of chimaeric GFP-fusion proteins that were transiently expressed in Welsh onion cells showed that Ntrps19-gfp was not localized in mitochondria. When several BvRps19-type amino acid substitutions, none of which was seen in any other angiosperm rps19, were introduced into Ntrps19-gfp, the modified Ntrps19-gfp became localized in mitochondria, supporting the notion that an ITS in BvRps19 evolved following sequence transfer to nuclei. Not all of these substitutions were seen in other ITS-type Rps19s, suggesting that the ITSs of Rps19 are diverse.     

Excision of the oxidatively formed 5-hydroxyhydantoin and 5-hydroxy-5-methylhydantoin pyrimidine lesions by Escherichia coli and Saccharomyces cerevisiae DNA N-glycosylases

Background

(5R?) and (5S?) diastereomers of 1-[2-deoxy-β-d-erythro-pentofuranosyl]-5-hydroxyhydantoin (5-OH-dHyd) and 1-[2-deoxy-β-d-erythro-pentofuranosyl]-5-hydroxy-5-methylhydantoin (5-OH-5-Me-dHyd) are major oxidation products of 2′-deoxycytidine and thymidine respectively. If not repaired, when present in cellular DNA, these base lesions may be processed by DNA polymerases that induce mutagenic and cell lethality processes.

Methods

Synthetic oligonucleotides that contained a unique 5-hydroxyhydantoin (5-OH-Hyd) or 5-hydroxy-5-methylhydantoin (5-OH-5-Me-Hyd) nucleobase were used as probes for repair studies involving several E. coli, yeast and human purified DNA N-glycosylases. Enzymatic reaction mixtures were analyzed by denaturing polyacrylamide gel electrophoresis after radiolabeling of DNA oligomers or by MALDI-TOF mass spectrometry measurements.

Results

In vitro DNA excision experiments carried out with endo III, endo VIII, Fpg, Ntg1 and Ntg2, show that both base lesions are substrates for these DNA N-glycosylases. The yeast and human Ogg1 proteins (yOgg1 and hOgg1 respectively) and E. coli AlkA were unable to cleave the N-glycosidic bond of the 5-OH-Hyd and 5-OH-5-Me-Hyd lesions. Comparison of the k_cat/K_m ratio reveals that 8-oxo-7,8-dihydroguanine is only a slightly better substrate than 5-OH-Hyd and 5-OH-5-Me-Hyd. The kinetic results obtained with endo III indicate that 5-OH-Hyd and 5-OH-5-Me-Hyd are much better substrates than 5-hydroxycytosine, a well known oxidized pyrimidine substrate for this DNA N-glycosylase.

Conclusions

The present study supports a biological relevance of the base excision repair processes toward the hydantoin lesions, while the removal by the Fpg and endo III proteins are effected at better or comparable rates to that of the removal of 8-oxoGua and 5-OH-Cyt, two established cellular substrates.

General significance

The study provides new insights into the substrate specificity of DNA N-glycosylases involved in the base excision repair of oxidized bases, together with complementary information on the biological role of hydantoin type lesions.     

Genome-wide identification of cytosine-5 DNA methyltransferases and demethylases in Solanum lycopersicum

Recent studies have reported that decreased level of DNA cytosine methylation in the global genome was closely related to the initiation of tomato (Solanum lycopersicum) fruit ripening. However, genome-scale analysis of cytosine-5 DNA methyltransferases (C5-MTases) and demethylases in tomato has not been engaged. In this study, 7 C5-MTases and 3 demethylases were identified in tomato genome, which probably contributed to DNA cytosine methylation level in tomato. The 7 C5-MTases were categorized into 4 subgroups, and the 3 demethylases were classified into 2 subgroups based on phylogenetic analyses. Comprehensive analysis of their structure and genomic localization was also performed in this paper. According to online RNA-seq data, 4 S. lycopersicum C5-MTase (SlC5-MTase) genes (SlMET, SlDRM1L1, SlDRM5, SlMET3L) were expressed higher than others, and one DNA demethylase gene (SlDML) was significantly changed during tomato fruit development and ripening. Furthermore, all these five gene expressions at breaker (BK) stage changed with 1-methylcyclopropene (1-MCP) treatment, indicating that they were regulated by ethylene directly or indirectly in tomato fruit. In addition, subcellular localization analysis indicated that SlDRM1L1 and SlDRM5 located in the nucleus might have responsibility for RNA-directed DNA methylation (RdDM). Collectively, this paper provided a framework for gene discovery and functional characterization of C5-MTases and DNA demethylases in other Solanaceae species.     

Structure and Function of the FeoB G-Domain from Methanococcus jannaschii

FeoB in bacteria and archaea is involved in the uptake of ferrous iron (Fe²⁺), an important cofactor in biological electron transfer and catalysis. Unlike any other known prokaryotic membrane protein, FeoB contains a GTP-binding domain at its N-terminus. We determined high-resolution X-ray structures of the FeoB G-domain from Methanococcus jannaschii with and without bound GDP or Mg²⁺-GppNHp. The G-domain forms the same dimer in all three structures, with the nucleotide-binding pockets at the dimer interface, as in the ATP-binding domain of ABC transporters. The G-domain follows the typical fold of nucleotide-binding proteins, with a β-strand inserted in switch I that becomes partially disordered upon GTP binding. Switch II does not contact the nucleotide directly and does not change its conformation in response to the bound nucleotide. Release of the nucleotide causes a rearrangement of loop L6, which we identified as the G5 region of FeoB. Together with the C-terminal helix, this loop may transmit the information about the nucleotide-bound state from the G-domain to the transmembrane region of FeoB.     

Overexpression of erg20 gene encoding farnesyl pyrophosphate synthase has contrasting effects on activity of enzymes of the dolichyl and sterol branches of mevalonate pathway in Trichoderma reesei

The mevalonate pathway is the most diverse metabolic route resulting in the biosynthesis of at least 30,000 isoprenoid compounds, many of which, such as sterols or dolichols, are indispensable for living cells. In the filamentous fungus Trichoderma of major biotechnological interest isoprenoid metabolites are also involved in the biocontrol processes giving the mevalonate pathway an additional significance. On the other hand, little is known about genes coding for enzymes of the mevalonate pathway in Trichoderma.     

Genetically Derepressed Nucleoplasmic Stellate Protein in Spermatocytes of D. melanogaster Interacts with the Catalytic Subunit of Protein Kinase 2 and Carries Histone-Like Lysine-Methylated Mark

Summary

The X-chromosome-linked clusters of the tandemly repeated testis-specific Stellate genes of Drosophila melanogaster, encoding proteins homologous to the regulatory β-subunit of the protein kinase casein kinase 2 (CK2), are repressed in wild-type males. Derepression of Stellate genes in the absence of the Y chromosome or Y-linked crystal locus (crystal line) causes accumulation of abundant protein crystals in testes and different meiotic abnormalities, which lead to partial or complete male sterility. To understand the cause of abnormalities in chromosome behavior owing to Stellate overexpression, we studied subcellular localization of Stellate proteins by biochemical fractionation and immunostaining of whole testes. We showed that, apart from the known accumulation of Stellate in crystalline form, soluble Stellate was located exclusively in the nucleoplasm, whereas Stellate crystals were located mainly in the cytoplasm. Coimmunoprecipitation experiments revealed that the α-subunit of the protein kinase CK2 (CK2α) was associated with soluble Stellate. Interaction between soluble Stellate and CK2α in the nucleus could lead to modulations in the phosphorylation of nuclear targets of CK2 and abnormalities in the meiotic segregation of chromosomes. We also observed that Stellate underwent lysine methylation and mimicked trimethyl-H3K9 epigenetic modification of histone H3 tail.     

Functional characterization of the Caenorhabditis elegans DNA repair enzyme APN-1

Caenorhabditis elegans possesses two distinct DNA repair enzymes EXO-3 and APN-1 that have been identified by cross-specie complementation analysis of the Saccharomyces cerevisiae apn1Δ apn2Δ tpp1Δ triple mutant deficient in the ability to repair apurinic/apyrimidinc (AP) sites and DNA strand breaks with blocked 3′-ends. While purified EXO-3 directly incises AP sites and removes 3′-blocking groups, such characterization has not been previously reported for APN-1. We recently documented that C. elegans knockdown for apn-1 is unable to maintain integrity of the genome. Despite the presence of EXO-3, the apn-1 knockdown animals are also defective in the division of the P1 blastomere, an observation consistent with the accumulation of unrepaired DNA lesions suggesting a unique role for APN-1 DNA repair functions. Herein, we show that C. elegans APN-1 is stably expressed as GST-fusion protein in S. cerevisiae only when it carries a nuclear localization signal, and with this requirement rescued the DNA repair defects of the S. cerevisiae apn1Δ apn2Δ tpp1Δ triple mutant. We purified the APN-1 from the yeast expression system and established that it displays AP endonuclease and 3′-diesterase activities. In addition, we showed that APN-1 also possesses a 3′- to 5′-exonuclease and the nucleotide incision repair activity. This latter activity is capable of directly incising DNA at the 5′-side of various oxidatively damaged bases, as previously observed for Escherichia coli endonuclease IV and S. cerevisiae Apn1, underscoring the importance of this family of enzymes in removing these types of lesions. Glycine substitution of the conserved amino acid residue Glu261 of APN-1, corresponding to Glu145 involved in coordinating Zn²⁺ ions in the active site pocket of E. coli endonuclease IV, resulted in an inactive variant that lose the ability to rescue the DNA repair defects of S. cerevisiae apn1Δ apn2Δ tpp1Δ mutant. Interestingly, the Glu261Gly variant did not sustain purification and yielded a truncated polypeptide. These data suggest that the Glu261 residue of APN-1 may have a broader role in maintaining the structure of the protein.     

Repair efficiency of (5′S)-8,5′-cyclo-2′-deoxyguanosine and (5′S)-8,5′-cyclo-2′-deoxyadenosine depends on the complementary base

5′-R and 5′-S diastereoisomers of 8,5′-cyclo-2′-deoxyadenosine (cdA) and 8,5′-cyclo-2′-deoxyguanosine (cdG) containing a base-sugar covalent bond are formed by hydroxyl radicals. R-cdA and S-cdA are repaired by nucleotide excision repair (NER) in mammalian cellular extracts. Here, we have examined seven purified base excision repair enzymes for their ability to repair S-cdG or S-cdA. We could not detect either excision or binding of these enzymes on duplex oligonucleotide substrates containing these lesions. However, both lesions were repaired by HeLa cell extracts. Dual incisions by human NER on a 136-mer duplex generated 24–32 bp fragments. The time course of dual incisions were measured in comparison to cis-anti-B[a]P-N²-dG, an excellent substrate for human NER, which showed that cis-anti-B[a]P-N²-dG was repaired more efficiently than S-cdG, which, in turn, was repaired more efficiently than S-cdA. When NER efficiency of S-cdG with different complementary bases was investigated, the wobble pair S-cdG·dT was excised more efficiently than the S-cdG·dC pair that maintains nearly normal Watson-Crick base pairing. But S-cdG·dA mispair with no hydrogen bonds was excised less efficiently than the S-cdG·dC pair. Similar pattern was noted for S-cdA. The S-cdA·dC mispair was excised much more efficiently than the S-cdA·dT pair, whereas the S-cdA·dA pair was excised less efficiently. This result adds to complexity of human NER, which discriminates the damaged base pairs on the basis of multiple criteria.