The wobble hypothesis revisited: uridine-5-oxyacetic acid is critical for reading of G-ending codons

Abstract

According to Crick's wobble hypothesis, tRNAs with uridine at the wobble position (position 34) recognize A- and G-, but not U- or C-ending codons. However, U in the wobble position is almost always modified, and Salmonella enterica tRNAs containing the modified nucleoside uridine-5-oxyacetic acid (cmo(5)U34) at this position are predicted to recognize U- (but not C-) ending codons, in addition to A- and G-ending codons. We have constructed a set of S. enterica mutants with only the cmo(5)U-containing tRNA left to read all four codons in the proline, alanine, valine, and threonine family codon boxes. From the phenotypes of these mutants, we deduce that the proline, alanine, and valine tRNAs containing cmo(5)U read all four codons including the C-ending codons, while the corresponding threonine tRNA does not. A cmoB mutation, leading to cmo(5)U deficiency in tRNA, was introduced. Monitoring A-site selection rates in vivo revealed that the presence of cmo(5)U34 stimulated the reading of CCU and CCC (Pro), GCU (Ala), and GUC (Val) codons. Unexpectedly, cmo(5)U is critical for efficient decoding of G-ending Pro, Ala, and Val codons. Apparently, whereas G34 pairs with U in mRNA, the reverse pairing (U34-G) requires a modification of U34.

FIGURE 1.

The genetic code. The eight codon boxes with shaded background are the family codon boxes, containing four codons encoding one amino acid (fourfold degenerate). The six lighter-shaded boxes contain tRNAs having cmo⁵U as wobble nucleoside. The boxes with white background are the mixed codon boxes. A circle corresponds to a codon read by a tRNA, and a line connecting two or more circles indicates that the same tRNA is able to read those codons. Filled circles indicate codon reading as predicted by the wobble hypothesis (Crick 1966) or the revised wobble rules (Yokoyama et al. 1985). Open circles indicate that those tRNAs are able to read also the C-ending codons (results presented in this study and in Näsvall et al. 2004). Next to the symbol for each tRNA is indicated which wobble nucleoside it contains. The letters within parentheses below the wobble nucleoside in the family boxes for proline, threonine, alanine, and valine indicate the last letter in the name of the genes encoding the corresponding tRNAs (e.g., tRNA^Val _cmo5UAC is encoded by the four genes valT, valU, valX, and valY, and tRNA^Pro _GGG is encoded by the gene proL).

FIGURE 2.

The proposed biosynthetic pathway for the synthesis of cmo⁵U and mcmo⁵U. (Gray arrows) Indicate the link between chorismic acid (or an unknown derivative of it) and different steps in the synthesis of cmo⁵U according to Näsvall et al. (2004). (U) Uridine; (ho⁵U) 5-hydroxyuridine; (mo⁵U) 5-methoxyuridine; (cmo⁵U) uridine-5-oxyacetic acid; (mcmo⁵U) uridine-5-oxyacetic acid methyl ester. (Adapted from Näsvall et al. 2004 and reprinted with permission from the RNA Society ©2004.)

FIGURE 3.

Locations of tRNA genes in the S. enterica genome. (A) Threonine tRNAs. (Upper line) The thrW gene, encoding tRNA^Thr _CGU. (Middle line) The rrnD rRNA operon containing one of the two genes encoding tRNA^Thr _GGU. (Lower line) The tufB operon, containing the gene encoding tRNA^Thr _cmo5UGC as well as the second gene encoding tRNA^Thr _GGU. (B) Valine tRNA genes. (Upper and middle lines) The two tRNA operons containing the four genes encoding tRNA^Val _cmo5UAC. (Lower line) The dicistronic valV, valW operon containing the two genes encoding tRNA^Val _GAC. (C) Alanine tRNAs. (Upper line) The three rRNA operons rrnH, rrnA, and rrnB, containing the genes encoding tRNA^Ala _cmo5UGC, have the same basic organization except an additional tRNA gene (aspU) at the end of rrnH. (Lower line) The alaW, alaX tRNA operon encoding tRNA^Ala _GGC. (D) Proline isoacceptors. (Upper line) The monocistronic proL gene, encoding tRNA^Pro _GGG. (Middle line) The proK gene, encoding tRNA^Pro _CGG. (Lower line) The operon containing the gene encoding tRNA^Pro _cmo5UGG as well as three other tRNA genes. (Black arrows) tRNA genes encoding threonine, valine, alanine, or proline tRNAs; (dark gray arrows) other tRNA genes; (light gray arrows) other genes. The anticodons of the relevant tRNAs are written below the genes. The drawings are not to scale. The asterisk (*) after gltT (which encodes tRNA^Glu _mnm5s2UUC) in A, middle line, indicates that the gene (STM3397) is not named in Salmonella. gltT is the name of an identical gene in the E. coli rrnB operon, which in Salmonella instead contains the genes ileU and alaU (asterisks in C).

FIGURE 4.

Overexpression of tRNA^Val _cmo5UAC restores growth of a ΔvalVW mutant. (A) Growth after 25 h of incubation at 37°C. (Sectors 1–4) Strains carrying pLG339 (vector control); (sectors 5–8) strains carrying p815 (valU valX valY lysV). The chromosomal genotypes are (1,8) LT2 (wt); (2,7) cmoB2<>cat; (3,6) ΔvalVW; (4,5) ΔvalVW cmoB2<>cat. (B) Sectors 3 and 4 of the same plate as in A, but after 44 h at 37°C. No suppressor mutants were apparent in this particular experiment. The relative colony sizes after 15 h of growth were (LT2/pLG339 and cmoB2/pLG339) 1.0 ± 0.03; (LT2/p815) 1.0 ± 0.07; (cmoB2/p815) 1.0 ± 0.03; (ΔvalVW/p815) 0.68 ± 0.04. Colonies of ΔvalVW cmoB2/p815 were visible but still too small to measure, and no colonies were visible of ΔvalVW/pLG339 or ΔvalVW cmoB2/pLG339. After 25 h, colonies of ΔvalVW cmoB2/p815 were ∼30% smaller than ΔvalVW/p815, and after 44 h, (B) colonies of ΔvalVW cmoB2/pLG339 were approximately half the size compared to ΔvalVW/pLG339.

FIGURE 5.

A-site selection rates at valine (GUN) codons. (*) Values in the cmoB2 mutant are significantly different from the control (LT2), as determined by a student's t-test (two sample, equal variance, p < 0.05). All values for the ΔvalVW mutant are significantly different from LT2 (p < 0.005). The values are averages of four experiments, with at least two independent cultures of each strain.

FIGURE 6.

Overexpression of hypo-modified tRNA^Ala _cmo5UGC from plasmid p70 can partially restore growth of a ΔalaXW cmoB2 mutant. (A) Growth of (sector 1) LT2/p70 (wild-type); (sector 2) cmoB2<>kan/p70; (sector 3) ΔalaXW/p70; and (sector 4) ΔalaXW cmoB2<>kan/p70 after 27 h at 37°C on an LA + tetracycline plate. The relative colony diameters (after 16 h) were (LT2/p70) 1.0 ± 0.05; (cmoB2/p70) 0.96 ± 0.04; and (ΔalaXW/p70) 0.92 ± 0.07. The colonies of ΔalaXW cmoB2/p70 were visible but too small to measure. (B) Same as in A, but the strains do not contain any plasmid, and the plate is LA without any antibiotic. The relative colony diameters (after 16 h) were (LT2) 1.0 ± 0.07; (cmoB2) 1.0 ± 0.01; and (ΔalaXW) 0.71 ± 0.04. At the time of the size measurements, no single colonies of the ΔalaXW cmoB2 mutant had appeared, but after 27 h, very tiny colonies (<0.1 mm in diameter) as well as some faster-growing colonies (still too small to be clearly visible in the picture) could be seen. (C) Sector 4 of the plate in B, but after 75 h. The absolute majority of the colonies of the ΔalaXW cmoB2 mutant were still <0.4 mm in diameter, while a few larger colonies ranging in sizes between ∼0.5 and 2 mm were visible.

FIGURE 7.

A-site selection rates at alanine (GCN) codons. The asterisks (*) indicate values from the cmoB2 mutant that are different from the control (LT2), as determined by a student's t-test [two sample, equal variance; (*) p < 0.05, (***) p < 0.0005]. All values from the ΔalaXW mutant are significantly different from LT2 (p < 0.0005). The values are averages from four experiments.

FIGURE 8.

A-site selection rates at proline (CCN) codons. The asterisks (*) indicate values from the cmoB2 mutant that are different from the control (LT2), as determined by a student's t-test [two sample, equal variance; (*) p < 0.05]. All values for the ΔproL proK<>frt and ΔproL proK<>frt cmoB2<>frt mutants are significantly different from LT2 (p < 0.01). The values are averages from at least three experiments. For simplicity, ΔproL proK<>frt is written ΔproKL.

FIGURE 9.

(A) The cmo⁵U34-G(III) base pair seen in the crystal structure solved by Weixlbaumer et al. (2007) (Protein Databank accession code 2UU9). The hydrogen bond between the 2′-OH of U33 and the O5 ether oxygen of cmo⁵U34 is indicated. (B) Keto-enol tautomerization of cmo⁵U. (Left) The keto tautomer of cmo⁵U (black), forming a wobble pair with G (gray) as predicted by Crick. (Right) The enol tautomer of cmo⁵U (black), engaged in a base pair with G (gray). The geometry of the cmo⁵U-G pair is the same as for a Watson–Crick (A-U or G-C) base pair.