5'-DMT-dI-Suc-CPG (deoxyInosine CPG); 1000 Å

Multiplex bisulfite-PCR sequencing is a convenient and scalable method for the quantitative determination of the methylation state of target DNA regions. A challenge of this application is the presence of CpGs in the same region where primers are being placed. A common solution to the presence of CpGs within a primer-binding region is to substitute a base degeneracy at the cytosine position. However, the efficacy of different substitutions and the extent to which bias towards methylated or unmethylated templates may occur has never been evaluated in bisulfite multiplex sequencing applications. In response, we examined the performance of four different primer substitutions at the cytosine position of CpG's contained within the PCR primers. In this study, deoxyinosine-, 5-nitroindole-, mixed-base primers and primers with an abasic site were evaluated across a series of methylated controls. Primers that contained mixed- or deoxyinosine- base modifications performed most robustly. Mixed-base primers were further selected to determine the conditions that induce bias towards methylated templates. This identified an optimized set of conditions where the methylated state of bisulfite DNA templates can be accurately assessed using mixed-base primers, and expands the scope of bisulfite resequencing assays when working with challenging templates.

Synthetic phosphorothioate oligodeoxynucleotides (ODN) bearing unmethylated CpG motifs can mimic the immune-stimulatory effects of bacterial DNA and are recognized by Toll-like receptor 9 (TLR9). Past studies have demonstrated that nucleotide modifications at positions at or near the CpG dinucleotides can severely affect immune modulation. However, the effect of nucleotide modifications to stimulate human leukocytes and the mechanism by which chemically modified CpG ODN induce this stimulation are not well understood. We investigated the effects of CpG deoxyguanosine substitutions on the signaling mediated by human TLR9 transfected into nonresponsive cells. ODN incorporating most of these substitutions stimulated detectable TLR9-dependent signaling, but this was markedly weaker than that induced by an unmodified CpG ODN. One of the most active ODN tested contained deoxyinosine for deoxyguanosine substitutions (CpI ODN), but its relative activity to induce cytokine secretion on mouse cells was much weaker than on human cells. The activity was dependent on TLR9, as splenocytes from mice genetically deficient in TLR9 did not respond to CpI ODN stimulation. It is surprising that CpI ODN were nearly as strong as CpG ODN for induction of human B cell stimulation but were inferior to CpG ODN in their ability to induce T helper cell type 1 effects. These data indicate that certain deoxyguanosine substitutions in CpG dinucleotides are tolerated to stimulate a TLR9-mediated immune response, but this response is insufficient to induce optimal interferon-alpha-mediated effects, which depend on the presence of an unmodified CpG dinucleotide. These studies provide a structure-activity relationship for TLR9 agonist compounds with diverse immune effects.

Acetaldehyde (AA), occurring widely in the human environment, is a mutagen and carcinogen. AA can react with DNA to form AA-DNA adducts. Several types of adducts, including an interstrand cross-link 3-(2-deoxyribos-1-yl)-5,6,7,8-tetrahydro-8-(N2-deoxyguanosyl)-6-methylpyrimido[1,2-a]purine-10(3H)one (7), have been previously characterized by our laboratory. We hypothesize that cross-link 7 may be involved in determining the mutagenic and carcinogenic properties of AA. To address this question, the double-stranded oligonucleotide 13, bearing cross-link 7, was synthesized in a sequence appropriate for mutagenicity studies in human cells. Oligonucleotide 9, containing 2-fluoro-O6-(trimethylsilylethyl)deoxyinosine (dIno), was reacted with 4-amino-1,2-pentanediol, followed by treatment with NaIO4. The resulting oligonucleotide 11 containing the 1,N2-propano-deoxyguanosine (dGuo) 5 was incubated with the complementary oligonucleotide 12 to give the desired cross-link 13, which was characterized by negative-ion electrospray ionization quadrupole time-of-flight mass spectrometry (ESI-Q-TOF-MS) and enzymatic hydrolysis to cross-link 7. The formation of cross-link 13 at 5'-CpG-3' was confirmed by incubation of 11 with [15N5]12 containing a 5'-Cp[15N5]G-3' sequence. The formation of cross-link 13 was reversible. Therefore, oligonucleotide 24 containing the irreversible analogue of cross-link 7, 1,3-bis(2'-deoxyguanos-N2-yl)butane, was synthesized for use as a control in the mutagenicity studies. Oligonucleotide 21 was reacted with 1,3-diaminobutane dihydrochloride, followed by incubation with the complementary oligonucleotide 23, to give 24. To determine the optimum distance and orientation for cross-link formation, six oligonucleotides, containing 5 at the i + 1, i + 2, and i + 3 or the i - 1, i - 2, and i - 3 positions relative to dGuo in the complementary strand, were 5'-end labeled with [gamma-32P]ATP, followed by incubation of each labeled oligonulceotide with its complementary strand and then analysis by denaturing polyacrylamide gel electrophoresis. Only the oligonucleotide containing 5'-Cp5-3' formed the cross-link with the complementary 5'-CpG-3' sequence. The results of this study confirm the structure of an AA-derived DNA cross-link, supply characterized cross-link-containing oligonucleotides for mutagenicity studies, and provide information on the optimum distance and orientation for cross-link formation.