PMC

Structures of the compounds from the NCI DTP library that passed the counter-screen. (A) NSC218351, isoquinoline-1-carboxylic acid. (B) NSC92218, 7-amino-5-iodo-8-quinolinol.

Analogs of NSC218351 (A) and NSC92218 (B) that were tested at various concentrations in the dual luciferase assay to measure their effects on initiation at UUG and AUG codons.

Overview of the dual luciferase assay and screen. (A) Schematic of the coding region of the plasmid used for the dual luciferase assay. (P) Promoter; (T) terminator. The ADH promoter and HIS terminator were used to produce Renilla luciferase mRNA and the GPD promoter and CYC terminator were used for firefly luciferase mRNA. (B) Flow-chart for identifying compounds that alter the fidelity of start codon recognition.

Normalized Fluc(AUG) (squares) and Fluc(UUG) (circles) expression in WT yeast (BY4741; closed symbols) and a strain deficient in drug efflux pumps (YRP1: snq2Δ, pdr5Δ, erg6Δ; open symbols) treated with compounds NSC218351 (A) and NSC92218 (B). Points for YRP1 are averages of data from three separate transformants ± average deviation. Data were analyzed as in , and are plotted with BY4741 data from , for comparison.

Both compounds increase translation of a luciferase reporter fused to the small, endogenous uORF from PRE2 beginning with a UUG codon. (A) Schematic of the reporter. As a control, the uORF was fused out of frame from the luciferase coding region. BY4741 expressing the reporter was treated with NSC218351 (B) or NSC92218 (C), and Fluc activity was measured. Closed squares are the in-frame reporters, and open squares are the out-of-frame controls. Fluc activity with DMSO alone was used to normalize the activity with compound. Points are averages of two independent experiments ± average deviation.

Both compounds increase the growth of Sui⁻ strain sui1-1 (eIF1 D83G) on media lacking histidine. Compound or solvent-soaked paper strips were placed on agar plates. Yeast were spotted onto the plate as four rows increasingly distant from the paper strip and grown for 4 d on SC-His and 2 d on SC. Two rows from one representative experiment are shown (the 50 mM isoquinoline data are from a separate experiment). The results were consistent in all rows and in three independent experiments.

The effect of Sui⁻ mutants and NSC218351 and NSC92218 on all near-cognate start codons in the dual luciferase assay. (A) The FlucXXX/FlucAUG ratio was measured for several Sui⁻ mutants, where XXX is the start codon that varies from AUG by 1 bp. This ratio was normalized to the same ratio in the wild-type control [(Fxxx/Faug)_Sui-/(Fxxx/Faug)_wt] to illustrate the magnitude of the effect of the Sui⁻ mutations on initiation at each codon relative to a wild-type strain. See for strain details. (B,C) Fluc expression (as in ) from reporters with different near-cognate start codons in cells treated with NSC218351 or NSC92218 (the black X is AUG, changes to position 1 of the start codon are in red, changes to the second position are in green, and changes to the third position are in blue).

Effect of NSC218351 and NSC92218 on start site selection in strains of yeast with altered levels of eIF 1, eIF 1A, and eIF 5. (A) FlucUUG/RlucAUG expression ratio and (B) FlucAUG/RlucAUG expression ratio, in the presence of DMSO (white bars) or compounds (NSC218351, gray bars; NSC92218, black bars), was measured in strain H3984 (hc eIF1) and compared with wild-type (BY4741) and haplo-insufficient diploids for eIFs 1 (+/sui1Δ), 1A (+/tif11Δ), and 5 (+/tif5Δ), and compared with diploid wild-type BY4743 (+/+). H3984 data are an average of two independent experiments. Error bars represent average deviation. The concentration of NSC218351 was 60 μM and NSC92218 was 3.8 μM. Haplo-insufficiency results are averages of data from two separate transformants. The concentration of NSC218351 was 33 μM, and NSC92218 was 7 μM. FlucAUG/RlucAUG expression ratio with DMSO alone was used to normalize the Fluc/Rluc ratio for both FlucAUG and FlucUUG for each strain (normalized FlucAUG/RlucAUG with DMSO alone is 1 for each strain). Concentration dependence of the effect of NSC218351 and NSC92218 on Fluc expression from UUG (squares), AUG (circles) in hc eIF1 (closed symbols), and wild-type (open symbols) yeast is shown in C and D. Data, from two independent experiments were analyzed as in .

The effects of NSC218351 and NSC92218 in the dual luciferase assay. (A,C) BY4741 expressing Fluc with either an AUG or UUG start codon was treated with various concentrations of NSC218351 and NSC92218, respectively, and normalized to the internal Rluc control with an AUG start codon. The Fluc/Rluc ratio from each sample was then normalized to the appropriate DMSO-only control, so that each ratio equals 1 in the absence of compound (normalized FlucUUG, red circles; normalized FlucAUG, blue squares). Points are the averages of at least seven independent experiments ±SE. (B,D) Raw luciferase activity values with increasing concentrations of NSC218351 and NSC92218. Red symbols indicate luciferase values from pFuugRaug plasmid, and blue symbols indicate luciferase values from pFaugRaug counter-screening plasmid (Fluc, squares; Rluc, circles). Luciferase activites are normalized to the DMSO control values in each experiment, and the normalized values of at least six independent experiments are averaged (± average deviation). (E) Relative levels of Fluc/Rluc mRNA measured using RT-q–PCR, from strain BY4741 expressing FlucUUG and RlucAUG treated with 2 μM NSC92218 or 50 μM NSC218351. The Rluc mRNA levels were used to normalize the Fluc mRNA levels, and the Fluc/Rluc ratio of the DMSO sample was used to normalize the samples treated with compounds. Data are the averages of duplicate samples.

Evaluation of the dual luciferase assay under screening conditions. (A) Increasing volumes of yeast (strain BY4741) culture expressing FlucUUG (●) and RlucAUG (■) from plasmid pFuugRaug were added to lysis buffer, and the resulting luciferase activities measured. Points are fit to a straight line with R > 0.99 for both Renilla and firefly luciferase activities. (B) BY4741 transformed with pFuugRaug was grown with various concentrations of cycloheximide, and the luciferase activities were measured after 4 h (●, FlucUUG; ■, RlucAUG). (C) BY4741 with pFuugRaug, as above, as well as with pFaugRaug, was grown with cycloheximide and the UUG/AUG (■) and AUG/AUG (●) ratios were measured. The Fluc/Rluc ratio of the solvent-only control was used to normalize the treated samples so that both ratios (UUG/AUG and AUG/AUG) equal 1 in the absence of drug. (D) Effect of DMSO on the Fluc/Rluc ratio for AUG/AUG (white bars) and UUG/AUG (filled bars). BY4741 with pFuugRaug or pFaugRaug were grown with 3% DMSO under screening conditions. The effect of DMSO on the ratio was controlled for by normalization with the DMSO-only control during the screening analysis. (E) The FlucUUG/FlucAUG ratios were measured using a wild-type (TD76-8D, closed circles) and a Sui⁻ strain (301-4D, open squares). This data set was used to calculate the Z′ factor (see Materials and Methods).