Retrospective Markov chain Monte Carlo methods for Dirichlet process hierarchical models

Inference for Dirichlet process hierarchical models is typicallyperformed using Markov chain Monte Carlo methods, which canbe roughly categorized into marginal and conditional methods.The former integrate out analytically the infinite-dimensionalcomponent of the hierarchical model and sample from the marginaldistribution of the remaining variables using the Gibbs sampler.Conditional methods impute the Dirichlet process and updateit as a component of the Gibbs sampler. Since this requiresimputation of an infinite-dimensional process, implementationof the conditional method has relied on finite approximations.In this paper, we show how to avoid such approximations by designingtwo novel Markov chain Monte Carlo algorithms which sample fromthe exact posterior distribution of quantities of interest.The approximations are avoided by the new technique of retrospectivesampling. We also show how the algorithms can obtain samplesfrom functionals of the Dirichlet process. The marginal andthe conditional methods are compared and a careful simulationstudy is included, which involves a non-conjugate model, differentdatasets and prior specifications.     

RNA/DNA hybrid binding affinity determines telomerase template-translocation efficiency

Telomerase synthesizes telomeric DNA repeats onto chromosome termini from an intrinsic RNA template. The processive synthesis of DNA repeats relies on a unique, yet poorly understood, mechanism whereby the telomerase RNA template translocates and realigns with the DNA primer after synthesizing each repeat. Here, we provide evidence that binding of the realigned RNA/DNA hybrid by the active site is an essential step for template translocation. Employing a template-free human telomerase system, we demonstrate that the telomerase active site directly binds to RNA/DNA hybrid substrates for DNA polymerization. In telomerase processivity mutants, the template-translocation efficiency correlates with the affinity for the RNA/DNA hybrid substrate. Furthermore, the active site is unoccupied during template translocation as a 5 bp extrinsic RNA/DNA hybrid effectively reduces the processivity of the template-containing telomerase. This suggests that strand separation and template realignment occur outside the active site, preceding the binding of realigned hybrid to the active site. Our results provide new insights into the ancient RNA/DNA hybrid binding ability of telomerase and its role in template translocation.     

A proliferative melanoma cell phenotype is responsive to RAF/MEK inhibition independent of BRAF mutation status

Oncogenic mutations within the MAPK pathway are frequent in melanoma, and targeting of MAPK signaling has yielded spectacular responses in a significant number of patients that last for several months before relapsing. We investigated the effects of two different inhibitors of MAPK signaling in proliferative and invasive melanoma cell cultures with various mutations in the MAPK pathway. Proliferative melanoma cells were more susceptible to pathway inhibition than invasive phenotype cells, irrespective of BRAF mutation status, while invasive phenotype cell response was dependent on BRAF mutation status. Critically, MAPK pathway inhibition of proliferative phenotype cells resulted in acquisition of invasive phenotype characteristics. These results show that melanoma cell phenotype is an important factor in MAPK pathway inhibition response. This suggests that while current therapeutic strategies target proliferative melanoma cells, future approaches should also account for the invasive phenotype population.     

A highly conserved lysine residue in phi29 DNA polymerase is important for correct binding of the templating nucleotide during initiation of phi29 DNA replication

DNA polymerases that initiate replication by protein-priming are able to catalyze terminal protein (TP)-primed initiation, the following transition steps and finally DNA-primed elongation. Therefore, their structures must be able to position sequentially both primers, TP and DNA, at a common binding site. For DNA-templated initiation, these DNA polymerases have to bind the origin of replication as template and TP as primer. It is likely that very precise interactions are required to position both TP and templating nucleotide at the polymerization active site. Such a specificity during TP-priming must rely on specific amino acids that must be evolutionarily conserved in this subfamily of DNA polymerases. By site-directed mutagenesis, we have analyzed the functional significance of Lys392 of phi29 DNA polymerase, immediately adjacent to the Kx3NSxYG motif, and specifically conserved among protein-primed DNA polymerases. During TP-primed initiation, mutations in this residue did not affect untemplated TP-dAMP formation, indicating that the interaction with the initiating nucleotide and TP were not affected, whereas the template-directed initiation activity was severely inhibited. Both mutant DNA polymerases had a wild-type-like (overall) DNA binding activity. We thus infer that residue Lys392 of phi29 DNA polymerase is important for the correct positioning of the templating nucleotide at the polymerization active site, a critical requirement during template-directed TP-priming at phi29 DNA origins. Consequently, mutation of this residue compromised the fidelity of the initiation reaction, not controlled by the 3'-5' exonuclease activity. During DNA-primed polymerization, the mutant polymerases showed a defect in translocation of the template strand. This translocation problem could be the consequence of a more general defect in the stabilization and positioning of a next templating nucleotide at the polymerization active site, during DNA-primed DNA synthesis.     

L-delta-(alpha-Aminoadipoyl)-L-cysteine-D-valine synthetase: production of dipeptides containing valine residue at its C-terminus

L-delta-(alpha-Aminoadipoyl)-L-cysteine-D-valine synthetase (ACVS) has been recently studied as a model enzyme for peptide synthetases. It was found that in the absence of alpha-aminoadipic acid but in the presence of several cysteine analogues it was incorporated into several analogue dipeptides upon incubation of the potential cysteine analogues with ACVS. [(14)C]Cysteine was incorporated into the[(14)C]cysteinyl-valine analogue dipeptides. Notably, [(14)C]valine incorporation in the presence of N-acylated cysteine analogues was observed. The alpha-aminoadipic acid activation site is influential, inhibitory or promotive, on the production of these putative dipeptide products. The production of dipeptide analogues, containing valine or analogues at the C-terminus, leads to the speculation that the biosynthetic direction of ACV could be from the C-terminus to the N-terminus.     

Affinity partitioning of glucose-6-phosphate dehydrogenase and hexokinase in aqueous two-phase systems with free triazine dye ligands

We have established a robust, fully automated analytical method for the determination of indomethacin in rat plasma using a column-switching high-performance chromatographic system. The system consists of a precolumn and an analytical column connected in series via a switching valve. When a 50-microl portion of rat plasma containing a therapeutic level of indomethacin was applied directly to the system, the drug was automatically enriched in the precolumn (TSK BSA-ODS) by on-line solid-phase extraction. After elution of the plasma proteins, the analyte was automatically transferred to the analytical column (Zorbax Eclipse XDBC18) where chromatography was performed using isocratic elution and UV absorption detection at a wavelength of 254 nm. The separation mobile phase consisted of methanol-0.1% phosphoric acid (70:30, v/v) at a flow-rate of 1 ml/min. The calibration line for indomethacin showed good linearity in the range 50-10 000 ng/ml (r>0.999) with the detection quantification of 50 ng/ml (RSD=2.6%). Accuracy ranged from -0.62 to 3.22%, and the within- and between-day precision of the assay was better than 6% across the calibration range. The analytical sensitivity and accuracy of this assay is suitable for characterization of the pharmacokinetics of topical administration of imdomethacin to rats. The method has been successfully used to provide pharmacokinetic data in a large number of diverse pharmaceutical studies.     

Automated column-switching high-performance liquid chromatography system for quantifying N-methyl-D- and -L-aspartate

The occurrence and biological significance of the D-amino acids, N-methyl-D-aspartate (NMDA) and N-methyl-L-aspartate (NMLA), have been recently studied in a variety of living organisms. In this study, we established a highly sensitive and reliable fluorometric HPLC system for determining levels of N-methyl-aspartate (NMA). The system comprises fluorescent derivatization of NMA with 4-fluoro-7-nitro-2,1,3-benzoxadiazole (NBD-F) and two chromatographic steps: one that separates NMA from other primary amino acids in reverse-phase mode and another that enantioseparates NMDA and NMLA in a normal-phase mode. These two steps are linked by an automated column-switching system. A simple pretreatment step with o-phthalaldehyde to remove primary amino acids that can interfere with sensitivity is also described. The detection limit for NMDA is as low as 5fmol and the correlation between peak heights and concentrations between 5fmol and 1pmol is satisfactory (r=0.999). Following sample preparation and separation using the column-switching HPLC system, more than 80% of NMDA was recovered from rat liver homogenates spiked with NMDA. This method was employed to determine the levels of NMDA in tissues from bivalves and the results obtained were consistent with the values reported previously.     

Reiterative template switching: the effect of single-strand homopolymeric DNA on non-template-directed nucleotide addition by DNA polymerase

Template switching occurs when DNA polymerase juxtaposes two discontinuous DNA molecules with 3'-terminally complementary ends generated through non-template-directed nucleotide addition. We examined whether juxtaposition of homopolymeric single-stranded oligonucleotides affects non-templated addition. We hypothesized that if DNA polymerase first juxtaposed the two substrates, then the non-template-directed nucleotide addition of any deoxynucleotide would decrease in the presence of its non-complementary template. For dATP, product formation was unaffected by non-complementary substrates. In contrast, dCTP and dGTP incorporation decreased to varying degrees while dTTP incorporation increased in the presence of oligodeoxythymidine but decreased for other non-complementary homopolymers. Interestingly, the presence of complementary templates strongly influenced the formation of highly periodic products indicative of reiterative template switching. Transient template synapsis was observed and found to be dependent on the non-templated sequence added: 3-4 A:T or 1-2 G:C base pairs were needed for stable synapsis, suggesting that base pairing plays a more important role in the active site of the enzyme than previously thought.     

Inhibitors of protein phosphatase-2A from human brain structures, immunocytological localization and activities towards dephosphorylation of the Alzheimer type hyperphosphorylated tau

Protein phosphatase (PP)-2A, which regulates the phosphorylation of tau, is regulated by two endogenous inhibitor proteins, I(1)(PP2A) and I(2)(PP2A), in mammalian tissues. Here, we report the cloning of I(1)(PP2A) and I(2)(PP2A) from human brain, and show that in PC12 cells and in I(1)(PP2A)-GFP or I(2)(PP2A)-GFP transfected NIH3T3 and human neural progenitor cells, I(1)(PP2A) is localized mostly in the cell cytoplasm and I(2)(PP2A) mostly in the nucleus. The recombinant I(1)(PP-2A) and I(2)(PP-2A) inhibit PP-2A activity towards hyperphosphorylated tau in vitro; the dephosphorylation of the hyperphosphorylated tau at specific sites is selectively inhibited. Overexpression of I(1)(PP2A) as well as I(2)(PP2A) results in tau hyperphosphorylation and degeneration of PC 12 cells.