The studies reported here were conducted to investigate the transport characteristics of apixaban (1-(4-methoxyphenyl)-7-oxo-6-(4-(2-oxopiperidin-1-yl)phenyl)-4,5,6,7-tetrahydro-1H-pyrazolo [3,4-c]pyridine-3-carboxamide) and to understand the impact of transporters on apixaban distribution and disposition. In human permeability glycoprotein (P-gp)-and breast cancer resistance protein (BCRP)-cDNA-transfected cell monolayers as well as Caco-2 cell monolayers, the apparent efflux ratio of basolateral-to-apical (Pc B-A ) versus apical-to-basolateral permeability (Pc A-B ) of apixaban was >10. The P-gp-and BCRP-facilitated transport of apixaban was concentration-and time-dependent and did not show saturation over a wide range of concentrations (1-100 mM). The efflux transport of apixaban was also demonstrated by the lower mucosal-to-serosal permeability than that of the serosal-tomucosal direction in isolated rat jejunum segments. Apixaban did not inhibit digoxin transport in Caco-2 cells. Ketoconazole decreased the P-gp-mediated apixaban efflux in Caco-2 and the P-gp-cDNA-transfected cell monolayers, but did not affect the apixaban efflux to a meaningful extent in the BCRP-cDNA-transfected cell monolayers. Coincubation of a P-gp inhibitor (ketoconazole or cyclosporin A) and a BCRP inhibitor (Ko134) provided more complete inhibition of apixaban efflux in Caco-2 cells than separate inhibition by individual inhibitors. Naproxen inhibited apixaban efflux in Caco-2 cells but showed only a minimal effect on apixaban transport in the BCRP-transfected cells. Naproxen was the first nonsteroidal antiinflammatory drug that was demonstrated as a weak P-gp inhibitor. These results demonstrate that apixaban is a substrate for efflux transporters P-gp and BCRP, which can help explain its low brain penetration, and low fetal exposures and high milk excretion in rats.

The isolation and characterization of temperature-sensitive mutations in RNA polymerase I from Saccharomyces cerevisiae are described. A plasmid carrying RPA190, the gene encoding the largest subunit of the enzyme, was subjected to in vitro mutagenesis with hydroxylamine. Using a plasmid shuffle screening system, five different plasmids were isolated which conferred a temperature-sensitive phenotype in haploid yeast strains carrying the disrupted chromosomal RPA190 gene. These temperature-sensitive alleles were transferred to the chromosomal RPAJ90 locus for mapping and physiology experiments. Accumulation of RNA was found to be defective in all mutant strains at the nonpermissive temperature. In addition, analysis of pulse-labeled RNA from two mutant strains at 37°C showed that the transcription of rRNA genes was decreased, while that of SS RNA was relatively unaffected. RNA polymerase I was partially purified from several of the mutant strains grown at the nonpermissive temperature and was shown to be deficient when assayed in vitro. Fine-structure mapping and sequencing of the mutant alleles demonstrated that all five mutations were unique. The rpal90-1 and rpal90-5 mutations are tightly clustered in region I (S. S. Broyles and B. Moss, Proc. Natl. Acad. Sci. USA 83:3141-3145, 1986), the putative zinc-binding region that is common to all eucaryotic RNA polymerase large subunits. The rpa190-3 mutation is located between regions III and IV, and a strain carrying it behaves as a mutant that is defective in the synthesis of the enzyme. This mutation lies within a previously unidentified segment of highly conserved amino acid sequence homology that is shared among the largest subunits of eucaryotic nuclear RNA polymerases. Another temperature-sensitive mutation, rpa190-2, creates a UGA nonsense codon.In eucaryotic cells a system of three distinct RNA polymerases has evolved for the expression of nuclear-encoded genes. These enzymes have been extensively studied in vitro; and it is well established that RNA polymerases I, II, and III (or A, B, and C) transcribe rRNA, mRNA, and tRNAs and 5S RNA, respectively (31). While these activities have distinct functional roles in the cell, they are closely related immunologically and in subunit composition (36). Each enzyme contains two large subunits, the larger of which is homologous to the P' subunit of RNA polymerase from Escherichia coli and the smaller of which is homologous to the P subunit (1, 2, 5, 41). However, unlike the E. coli holoenzyme, which has the relatively simple Ora2I33' subunit structure, the eucaryotic RNA polymerases have many more small subunits, none of which appear to be closely related to the a or a subunits of E. coli RNA polymerase by immunological criteria (14). Some of the small subunits are shared among all of the three eucaryotic polymerases, while others are shared only between RNA polymerase I and III. Still others are unique for each enzyme.Of the eucaryotic RNA polymerases, the yeast enzymes are among the best characterized (36). Most of ...

The synthesis of ribosomal proteins (r proteins) under the conditions of greatly reduced RNA synthesis were studied by using a strain of the yeast Saccharomyces cerevisiae in which the production of the largest subunit (RPA190) of RNA polymerase I was controlled by the galactose promoter. Although growth on galactose medium was normal, the strain was unable to sustain growth when shifted to glucose medium. This growth defect was shown to be due to a preferential decrease in RNA synthesis caused by deprivation of RNA polymerase I. Under these conditions, the accumulation of r proteins decreased to match the rRNA synthesis rate. When proteins were pulse-labeled for short periods, no or only a weak decrease was observed in the differential synthesis rate of several r proteins (L5, L39, L29 and/or L28, L27 and/or S21) relative to those of control cells synthesizing RPA190 from the normal promoter. Degradation of these r proteins synthesized in excess was observed during subsequent chase periods. Analysis of the amounts of mRNAs for L3 and L29 and their locations in polysomes also suggested that the synthesis of these proteins relative to other cellular proteins were comparable to those observed in control cells. However, Northern analysis of several r-protein mRNAs revealed that the unspliced precursor mRNA for r-protein L32 accumulated when rRNA synthesis rates were decreased. This result supports the feedback regulation model in which excess L32 protein inhibits the splicing of its own precursor mRNA, as proposed by previous workers (M. D. Dabeva, M. A. Post-Beittenmiller, and J. R. Warner, Proc. Natl. Acad. Sci. USA 83:5854-5857, 1986).In growing cells of both procaryotes and eucaryotes, most of the ribosomal components are present in the mature ribosomes; that is, none of the individual ribosomal components accumulates in significant excess over the others. In Escherichia coli, this balanced accumulation of ribosomal components is largely accomplished by regulating the synthesis rates of ribosomal proteins (r proteins) by posttranscriptional feedback mechanisms as originally proposed from gene-dosage experiments (11; for reviews, see references 29 and 33), although regulation of r-protein synthesis at the transcription step may also be involved for some r-protein operons (28).Studies have also been carried out for a variety of eucaryotic cell systems to examine whether the balanced production of ribosomal components is really due to balanced synthesis and, if so, how this is accomplished. In the case of Saccharomyces cerevisiae, this question has been extensively studied by using gene-dosage experiments similar to those used for E. coli. Cloned genes encoding individual r proteins were introduced into the cell on high-copy-number plasmids, and the synthesis rates of r proteins and the levels of their mRNAs were analyzed (10,18,30,34,42,47). In general, although the r-protein mRNA encoded by the plasmid accumulated to a high level, the accumulation of the r protein itself was not apparently elevated relative to t...

Homothallic switching of yeast mating type (MAT) Recombination intermediates can be studied in detail during synchronously induced switching of MAT (Fig. 1A) and related substrates (10-13). After HO endonuclease cleavage of MAT, the DNA to the right of the double-strand break is converted to a long single-stranded tail by a 5' -* 3' exonuclease that advances by about 1-2 nt/sec (10,12,13). This 3' single-stranded tail apparently invades the intact donor locus, HMLa, and forms heteroduplex DNA. The subsequent elongation of this strand to copy the opposite mating-type (Ya) sequences can be detected by the use of PCR (Fig. 1B) (10).We found that this step (designated Ya-MATdlsg joining) occurs -30 min before the completion of recombination as measured both by a second PCR reaction (MATproximai-Ya joining; Fig. 1C) and by the appearance of the completed product (MATa).Heteroduplex DNA formed during MAT switching has not been detected physically, but its existence can be shown genetically by the appearance of sectored colonies after switching, an event we refer to as PSS (4, 5). For example, switches of the MATa-stk mutation (a T --A mutation at position MAT-Zll) lead either to MATa (normal a mating; Tzii) or matal-stk (sterile; Az,,), depending on the fate ofthe stk mutation (Fig. 1A). By comparing the outcomes of switching in a wild-type strain with a pmsl mutant derivative that is defective in mismatch repair, we showed that nearly all switching events between the mutant Az,, of MATa-stk and the normal Tzll allele at HML involve the formation of heteroduplex DNA formation followed by mismatch repair (5). In a mismatch repair-proficient strain, the Az,, mutation is replaced by Tzll in 85-90%o of all DNA strands, so that 75% of the switches yielded colonies in which all progeny were MATa. Most of the remaining switches exhibit PSS, in which one MATa cell and one matal-stk cell are formed. In contrast, with a pmsl mutant strain, the fraction of switches in which the Az,, mutant is converted to Tzll on both DNA strands decreases from 75% to 15%, with a concomitant increase in MATa/matal-stk sectored colonies (5). These results showed that a mutation only 8 bp from the 3' end of the HO endonuclease cleavage site is usually not removed by exonucleolytic digestion but is included in heteroduplex DNA. Moreover, the repair of the mismatch was shown to be highly preferential; although the information at HMLa is used to convert MAT, reciprocal events, in which the mutation at MAT was introduced into HMLa, are not observed (4, 5).A critical question in understanding exactly how recombination occurs during MAT switching (or in any other homologous recombination event) is, when does mismatch correction occur? As shown in Fig. 2, there are two alternative pathways whereby MATa-stk can switch to MATa. In the first possible pathway (Fig. 2 A, B, and E), heteroduplex DNA is formed between the invading 3'-ended single strand of MAT (stk) DNA ( Fig. 2A) and the resident HMLa locus; then the invading strand is corrected to ...

Alzheimer's disease (AD) is a progressive neurodegenerative disease. Amyloid ␤ (A␤) peptides are hypothesized to cause the initiation and progression of AD based on pathologic data from AD patients, genetic analysis of mutations that cause early onset forms of AD, and preclinical studies. Based on this hypothesis, ␤-site amyloid precursor protein (APP)-cleaving enzyme 1 (BACE1) inhibitors are an attractive therapeutic approach for AD because cleavage of the APP by BACE1 is required to form A␤. In this study, three potent BACE1 inhibitors are characterized. All three inhibitors decrease A␤ formation in cultured cells with IC 50 values less than 10 nM. Analysis of APP C-terminal fragments by immunoblotting and A␤ peptides by mass spectrometry showed that these inhibitors decreased A␤ by inhibiting BACE1. An assay for A␤1-40 in mice was developed and used to show that these BACE1 inhibitors decreased plasma A␤1-40, but not brain A␤1-40, in wild-type mice. Because these BACE1 inhibitors were substrates for P-glycoprotein (P-gp), a member of the ATP-binding cassette superfamily of efflux transporters, these inhibitors were administered to P-gp knockout (KO) mice. These studies showed that all three BACE1 inhibitors decreased brain A␤1-40 in P-gp KO mice, demonstrating that P-gp is a major limitation for development of BACE1 inhibitors to test the amyloid hypothesis. A comparison of plasma A␤1-40 and brain A␤1-40 dose responses for these three compounds revealed differences in relative ED 50 values, indicating that factors other than P-gp can also contribute to poor brain activity by BACE1 inhibitors.Alzheimer's disease (AD) is a progressive neurodegenerative disease that begins with memory loss and progresses to include severe cognitive impairment (Walsh and Selkoe, 2004). Extracellular plaques that consist primarily of amy-J.E.Me. and L.A.T. contributed equally to this work. Article, publication date, and citation information can be found at