Figure 3.Comparisons of the slippery features of phe-arg dicodon sites. (A) and (B) show the pre-shift (at left) and post-shift (right) anticodon:message complexes for UUU CGG and UUC CGU, respectively. Watson-Crick base pairs are indicated by vertical lines; the G:U wobble pair is marked with a colon; and the C:A mispair is denoted with an `x'. Codon pair UUUCGG contains shifty features 1, 3 and 4 from Figure 1: The pre-shift complex contains a G:U wobble pair, which may facilitate loss of the initial frame (feature 3); the post-shift complex is stabilized by three Watson-Crick base pairs (feature 4); and, because the tRNA that reads the CGG codon is very rare, there may be a greater pause during which realignment can occur (feature 1). In contrast, codon pair UUCCGU has none of these shifty features: The pre-shift complex includes no wobble base pairs; the post-shift complex includes a C:A mispair, which can destabilize the complex (21); and, because the CGU codon is read by an abundant tRNA, there may be little time available for tRNA:message realignment.

Frameshift frequencies of the phenylalanine-arginine codon pairs

The four codon pairs were placed at the RF2 programmed frameshift site in RF2:lacZ fusions (Fig. 2 ). The [beta]-galactosidase activities and percent frameshift frequencies are listed in Table 1 (the alleles with the RF2 prefix). The observed frameshift frequencies are completely consistent with the predictions. Dicodon UUU CGG, which has all of the shift-prone features, is ~18-fold more frameshift-prone than UUC CGU, which has none of the features. In addition, the two alleles that have only some of the shift-prone features (UUU CGU and UUC CGG) have intermediate frameshift frequencies. And, because those intermediate values are essentially equivalent, the reduction in slipperiness by the UUU to UUC mutation, and the reduction in the pause by the CGG to CGU mutation, have equal effects RF2asy1on the RF2 programmed frameshift.

Table 1 . Frameshifting frequencies of the UUU CGG/U constructs

Strain	Frameshift allele	[beta]-gal activity	% Frameshifting
MY600 (pRS20)	WT	10 530	-
MY600(pRS31)	RF2/UUU CGG	555	5.3
MY600(pRS32)	RF2/UUU CGU	123	1.2
MY600(pRS33)	RF2/UUC CGG	131	1.2
MY600(pRS34)	RF2/UUC CGU	31	0.3
MY600(pRS21)	NSD/UUU CGG	129	1.2
MY600(pRS22)	NSD/UUU CGU	95	0.9
MY600(pRS23)	NSD/UUC CGG	3.5	0.03
MY600(pRS24)	NSD/UUC CGU	3.2	0.03

[beta]-gal values are averages of from between four and six assays; standard errors of the mean are <10%. Percent frameshifting values are relative to the WT control.

We also assayed the same codon pairs within very short frameshift windows that lack any Shine-Dalgarno-like element (Fig. 2 ). Here, even without the Shine-Dalgarno the alleles with UUU at the first codon frameshift at frequencies that are ~2 orders of magnitude higher than the average frameshift frequency (42 ,43 ). In contrast, neither allele with UUC frameshifts significantly. These data are in general agreement with the earlier observations of Fu and Parker (24 ), who found that UUU Ynn, but not UUC Ynn, codon pairs were frameshift-prone. Thus, a slippery UUU at the first codon appears to be essential for frameshifting in `normal' contexts. However, in contrast to the RF2 context, whether or not the second codon is read by a rare tRNA is relatively unimportant in the normal context (compare NSD/UUU CGG and NSD/UUU CGU). This suggests that in the normal context the pause prior to aa-tRNA selection at the second codon may be less important than it is at the RF2 programmed frameshift site.

Frameshifting with altered translational apparatus

At some sites frameshift frequencies can be affected by mutations and other changes in the translational apparatus. We wished to determine whether such changes would differentially affect the RF2 and NSD alleles. Fu and Parker (24 ) noticed that as cultures enter the stationary phase frameshifting increases at the UUU UAC site at the 5' end of argI. Gallant and coworkers (44 ) also observe that frameshift frequencies can increase in stationary phase cultures. The changes in cell physiology that occur during stationary phase are not fully defined, and perhaps several changes could affect ribosomal kinetics and accuracy.

One change that may be important is the undermodification of tRNA, which can be induced by various stressful physiological conditions (45 ). The slippery tRNA^Phe has the 2-methylthio and 6-isopentenyl modifications at adenosine 37 (ms2i6A in E.coli and ms2io6A in Salmonella typhimurium), which is immediately 3' of the anticodon. Formation of these modifications is dependent on the activity of the miaA gene (46 ). Absence of the miaA modifications decreases translational efficiency, probably because the modifications stabilize the tRNA:message complex (reviewed in ref. 25 ). Therefore, it is plausible that tRNA^Phe might be more likely to slip on the message if it lacks the miaA modifications. To determine whether modification of base 37 and/or stationary growth differentially affects frameshifting by the RF2 and NSD alleles, we compared frameshift frequencies in isogenic miaA+/- S.typhimurium strains in both midlog and stationary phase (overnight) cultures. The S.typhimurium stains were the kind gifts of Dr Glenn Björk.

We measured the steady state [beta]-galactosidase activities of the RF2/UUU CGG and NSD/UUU CGG alleles during expression in isogenic miaA+/- strains in both the midlog and the stationary culture conditions (Table 2 ). Assays made from log phase cultures at various densities <= 1.6 * 10⁹ cells/ml are similar to within ~10% of each other, and the averages of eight such assays are reported in Table 2 . Similarly, assays made from overnight cultures (densities >= 4 * 10⁹ cells/ml) are also consistent, and the averages of those assays are also included in the table. The miaA mutation increases frameshifting of the RF2-based construct by nearly 2-fold (1.8- and 1.5-fold for low and high cell density, respectively). The NSD allele shows slightly greater increases (2.3- and 2.5-fold, for low and high cell density, respectively). These increases suggest that the miaA modifications reduce the inherent tendency of this tRNA to frameshift. Furthermore, the modifications may be slightly more important in the NSD context (the standard errors of these ratios are ~15%). We address the effect of the miaA1 mutation on frameshifting in the Discussion.

Table 2 . Frameshifting increases in an miaA strain at low and high cell densities

Strain	Frameshift allele	miaA allele	Cell density
			Low	High
GT522(pRS31)	RF2/UUU CGG	miaA+	4.5	14.2
GT523(pRS31)	RF2/UUU CGG	miaA1	8.0	21.4
GT522(pRS21)	NSD/UUU CGG	miaA+	1.25	4.8
GT523(pRS21)	NSD/UUU CGG	miaA1	2.9	11.8

Reported are % frameshifting relative to the corresponding Salmonella host expressing WT lacZ from pRS20. Low cell density refers to cultures with viable cell concentrations between 2 * 10⁸ and 1.6 * 10⁹ cells/ml. High cell density refers to cultures with viable cell concentrations between 4 * 10⁹ and 6.5 * 10⁹ cells/ml.

All of the stationary cultures show increased frameshifting. For the RF2-based allele, frameshifting increases ~3-fold (2.8-fold for miaA+ and 3.2-fold for miaA1). The NSD alleles is affected ~4-fold (3.8-fold for miaA+ and 4.1-fold for miaA1). Thus, with borderline significance the NSD allele appears to be slightly more sensitive to the stationary culture condition. Finally, because the stationary culture condition increases frameshifting equally for the miaA+ and miaA1 strains, the stationary-induced increase is not simply due to an inefficient miaA modification. We have not further explored the molecular and physiological bases for stationary-induced frameshifting.

We also studied the effect of a restrictive rpsL (streptomycin resistance) mutation on frameshifting at UUU Ynn. Restrictive rpsL mutations inhibit or `restrict' aminoacyl-tRNA selection (47 -50 ). Sipley and Goldman (12 ) and we (51 ) have shown that restrictive rpsL mutations can increase frameshifting for certain RF2/lacZ fusions, presumably because restricted aa-tRNA selection at the second codon increases the time available for tRNA:message realignment. We also observed that the response to rpsL can differ among RF2 alleles, and rpsL sensitivity correlates with weak codon:anticodon pairing at the second codon (51 ).To determine whether a restrictive rpsL mutation affects frameshifting at UUU Ynn, we assayed constructs from Table 1 in MY599. None of the constructs showed significant increases in frameshift frequency in the rpsL background (not shown), which suggests that instability of the tRNA:message complexes at the second codon is not a factor in frameshifting at these sites. In any case, rpsL does not distinguish the RF2 and NSD alleles.

UUU Ynn dicodons are strongly avoided

It is firmly established that highly expressed genes show strong sequence biases, and sequence bias is thought to contribute to high levels of expression (26 -31 ). Therefore, if UUU Ynn codon pairs are generally frameshift-prone, then such sites should be underrepresented in genes that show strong sequence biases. We searched for the occurrences of UUU Ynn codon pairs in E.coli genes that have high degrees of codon usage bias (40 ). Assuming that codons are randomly distributed, then 122 UUU Ynn codon pairs are expected; however, only 25 such sites occur (see Table 3 ). Assuming a binomial probability, a difference from expectation this great has a chance likelihood of only 5 * 10^-18. We also searched for these sites in a collection of weakly expressed genes (41 ). The UUU Ynn codon pairs are also underrepresented in these genes, but the difference from expectation is of only borderline significance. Interestingly, for both highly and weakly expressed genes, the `missing' UUU Ynn sites are almost exactly matched by `extra' UUC Ynn sites (97 fewer UUU Ynn and 92 more UUC Ynn in the `Hi' set; and 31 fewer UUU Ynn and 33 more UUC Ynn in the `Lo' set). These observations are consistent with there being random distributions of phenylalanine-Ynn dicodons, accompanied by selection for UUU -> UUC mutations at those sites. Further, because this apparent selection is much greater for highly biased genes, this bias may contribute to high levels of expression.

Table 3 . Statistical analyses of the occurrences of UUUY tetranucleotides

Sites	Number of sites	Binomial probability
	Expected	Observed
Hi-UUU Ynn	122	25	5 * 10-18
Hi-UUC YNN	332	424	1.5 * 10-6
Lo-UUU Ynn	169	138	2 * 10-2
Lo-UUC YNN	105	138	2 * 10-3
Hi-vUU UYn	143	146	0.4
Hi-nnU UUC and Hi-nnU UUU Rnn	303	240	5 * 10-4

The prefix `Hi-'refers to genes with high levels of sequence bias; `Lo-' means weakly expressed genes.

A formal alternative, however, is that the scarcity of UUU Ynn codon pairs in highly biased genes is due to some physical property of UUUY tetranucleotides (or the corresponding DNA sequence) that is unrelated to protein synthesis. In this were true, then UUUY tetranucleotides should be rare in all three reading frames. We tested this by searching the highly biased gene set for sites that include the UUUY tetranucleotides phased in the alternate reading frames (e.g., the dicodon gUU UCa includes a UUUY tetranucleotide, but it is phased one nucleotide rightward relative to the translated frame). One complication inherent to studies nucleotide runs, like UUUY, is that certain codon pairs can include them in more than one reading frame. For example, dicodons UUU UYn include UUUY tetranucleotides in both the initial and the rightward phases. Because above we show that UUUY tetranucleotides are rare in the initial frame (above), we were careful to exclude sequences that include UUUY in the initial frame from these control analyses. For example, to study the rightward phase we examined codon pairs of the type AUU UYN plus GUU UYN plus CUU UYN (collectively noted as vUU UYn).

In the rightward phase, vUU UYn dicodons, occur almost exactly as expected for randomly distributed codons (Table 3 ). These data show that the UUUY tetranucleotides are not universally excluded from coding sequences, as would be expected for sequences that have some adverse structural anomaly. Therefore, the extreme scarcity of UUU Ynn codon pairs may be related to a translational property(s) such as frameshifting.

Interestingly, the UUUY tetranucleotides are relatively rare in the leftward reading frame (nnU UUC + nnU UUU Rnn codon combinations, Table 3 ), though the effect is not nearly as strong as it is for the in-phase UUU Ynn dicodon set. Gallant and co-workers have shown that nnU UUC dicodons can be shift-prone into the leftward frame, but in those experiments frameshifting requires starvation for the amino acid specified by the codon followed the phenylalanine codon (52 ) or the stationary culture condition (44 ). Leftward frameshifting also occurs at such sites in several programmed frameshifts (1 ,2 ). Clearly, leftward frameshifting can occur at phenylalanine codons at least under certain growth conditions and/or contexts. Thus, the modest avoidance of the leftward UUUY tetranucleotide might also be related to an inherent shiftiness.

DISCUSSION

We know a great deal about the sequence features that contribute to frameshifting at the prfB (RF2) programmed frameshift site (Fig. 1 ), but we know very little about how those features related to frameshift errors during normal translation. It was previously shown that a naturally occurring UUU UAU codon pair and derivatives that retain a UUU Ynn theme are ~100 times more frameshift-prone than the average message site (24 ). To help understand the mechanism of this apparently errant translation, we analyzed the frameshifting tendencies of variants of the UUU Ynn theme placed within two broader contexts: the RF2 programmed frameshift site, which contains a frameshift-enhancing element (the Shine-Dalgarno-like element, ref. 23 , see Fig. 1 ), and a site that lacks frameshift-enhancing features. Extensive preliminary analyses of the RF2 frameshift enabled the design of a set of four codon pairs that contain various combinations of frameshift-prone features. Our data show that predictions about frameshift tendency in RF2 are partially applicable to a non-programmed slippery site. The disagreements between the RF2 and the normal contexts may result from the effects of the Shine-Dalgarno-like interaction, which occurs only for the RF2-based alleles.

In both contexts, high frequency frameshifting requires that the rephased tRNA:message complex be stable, but the partially-mismatched complex at UUC CGG is relatively frameshift-prone only in the RF2 context (Table 1 ). Together with previous work, these data suggest that the Shine-Dalgarno may be able to help hold the shifted phase following tRNA:message realignment. We previously show that many RF2 alleles frameshift with considerable frequencies despite forming moderately mismatched complexes in the rightward frame (21 ). The intermediate frameshift frequency for RF2-UUC CGG (Table 1 ), which has a single mismatch in the rightward frame (an A:C in the middle position), is consistent with that earlier work. In contrast, the NSD alleles that have UUC as the phenylalanine codon essentially do not frameshift (Table 1 ). The simplest explanation is that those alleles, which can form neither a perfectly realigned tRNA:message complex nor a Shine-Dalgarno interaction, cannot stabilize the rephased complex.

Another effect of the Shine-Dalgarno-like interaction may be to highly elevate frameshifting only when the second codon is available for translation in the A site (i.e., after translocation of peptidyl-tRNA^Phe). An implicit requirement for RF2 autoregulation is that frameshift probability is increased only while the regulatory UGA is available for recognition by RF2 (6 ,7 ,13 ). Most probably, translocation of the message simultaneously places the regulatory codon into the A site and the stimulatory AGGGG to within reach of the ribosomal anti-Shine-Dalgarno. This model makes the programmed frameshift totally dependent on the kinetic competition between Shine-Dalgarno association and translation of the A site codon (7 ). Another implication of this model is that because only post-translocation frameshifting is stimulated by the Shine-Dalgarno, other mechanisms do not contribute significantly to observed [beta]-galactosidase activity.

In contrast, the frameshifting observed for alleles that lack the Shine-Dalgarno could conceivably result from rephasing before, during and/or after translocation. Because the second codon has only a minor effect on frameshifting (compare UUU CGG and UUU CGU in Table 1 ), most frameshifting on the NSD alleles may occur before the second codon becomes available for decoding. In addition, that the normal sites is also slightly more sensitive to the miaA1 mutation and to the stationary phase also supports the notion that frameshifting at the normal site differs mechanistically from that of RF2. Plausible other mechanisms include selecting phe-tRNA^Phe directly into the rightward phase, and/or slippage of the tRNA:message complex prior to or during translocation. Unfortunately, it is not yet possible to distinguish these mechanisms.

The lack of the miaA-dependentmodifications at A37 increases frameshifting for both the RF2 and NSD alleles. Others have observed that the absence of a different modification at position 37 (the lack of the 1-methyl modification at G37) of one or more proline tRNAs is associated with rightward frameshifting (53 ,54 ). There, increased frameshifting is consistent with either increased slippage because of reduced tRNA:message stability and/or out-of-phase pairing between the unmodified G37 and the message in the rightward phase (54 ,55 ). There is abundant evidence that the absence of the miaA modifications decreases tRNA:message stability. For example, in a model system in which tRNAs are paired through complementary anticodons (56 ), complexes are less stable if they include a tRNA that is missing its miaA-dependent modifications (57 ). In addition, unmodified tRNAs are less efficient during nonsense suppression and missense reading in vivo (58 -62 ). It was suggested that the mutation makes tRNA:message complexes less stable because the unmodified adenosine does not stack strongly onto the codon:anticodon complex, and that this relatively weak interaction increases the likelihood that the suppressor will be rejected during ribosomal selection (60 ). This suggestion is supported by the observation that miaA unmodified tRNAs are aggressively proofread and are poor competitors with modified tRNAs during in vitro polypeptide synthesis (63 ). It seems very likely that an miaA induced increased in frameshifting is caused, at least in part, by an instability in the pre-shift complex which leads to an increased probability of tRNA:message slippage.

The reasons for biased codon and context usage are not yet known (26 -31 ), but they are very likely to include selection against sites that are either inefficiently translated and/or error-prone. Our statistical analyses show that UUU Ynn dicodons are extremely rare in the translated phase of E.coli genes that exhibit strong codon usage biases (Table 3 ). This apparent avoidance of UUU Ynn codon pairs is very likely due to one or more translational properties; it seems at least plausible that a strong, inherent frameshifting tendency could be such a property. We previously showed that the CGA codon, which is decoded with an inosine:adenosine wobble base pair, is both extremely rare and inefficiently decoded (51 ). It seems very likely that the avoidance of CGA codons and the scarcity of UUU Ynn codon/contexts are driven, at least in part, by the poor translational qualities of these sequences.

ACKNOWLEDGEMENTS

We are grateful to Ms Jennifer Miller for technical assistance and to Dr Glenn Björk for the kind gift of Salmonella strains. We are also grateful to Drs Jonathan Gallant and Valery Lim for unpublished data and for helpful comments on the manuscript. This work was supported by NIH grant GM52643 to JFC.

REFERENCES

2 Gesteland, R.F., and Atkins, J.F. (1996) Ann. Rev. Biochem. 65, 741-768. MEDLINE Abstract

3 Jacks, T., Madhani, H.D., Masiarz, F.R., and Varmus, H.E. (1988) Cell 55, 447-458. MEDLINE Abstract

4 Weiss, R.B., Dunn, D.M., Atkins, J.F., and Gesteland, R.F. (1987) Cold Spring Harbor Symp. Quant. Biol. 52, 687-693. MEDLINE Abstract

5 Brierley, I., Jenner, A.J., and Inglis, S.C. (1992) J. Mol. Biol. 227, 463-479. MEDLINE Abstract

6 Craigen, W.J., and Caskey, C.T. (1986) Nature 322, 273-275. MEDLINE Abstract

7 Curran, J.F., and Yarus, M. (1988) J. Mol. Biol. 203, 75-83. MEDLINE Abstract

8 Belcourt, M.F., and Farabaugh, P.J. (1990) Cell 62, 339-352. MEDLINE Abstract

9 Curran, J.F., and Yarus, M. (1989) J. Mol. Biol. 209, 65-77. MEDLINE Abstract

10 Tu, C., Tzeng, T.H., and Bruenn, J.A. (1992) Proc. Natl. Acad. Sci. USA 89, 8636-8640. MEDLINE Abstract

11 Tsuchihashi, Z. (1991) Nucleic Acids Res. 19, 2457-2462. MEDLINE Abstract

12 Sipley, J., and Goldman, E. (1993) Proc. Natl. Acad. Sci. USA 90, 2315-2319. MEDLINE Abstract

13 Craigen, W.J., Cook, R.G., Tate, W.P., and Caskey, C.T. (1985) Proc. Natl. Acad. Sci. USA 82, 3616-3620. MEDLINE Abstract

14 Scolnick, E., Tompkins, R., Caskey, C.T., and Nirenberg, M. (1968) Proc. Natl. Acad. Sci. USA 61, 768-774. MEDLINE Abstract

15 Adamski, F.M., Donly, B.C., and Tate, W.P. (1993) Nucleic Acids Res. 21, 5074-5078. MEDLINE Abstract

16 Poole, E.S., Brown, C.M. and Tate, W.P. (1995) EMBO J., 14, 151-158. MEDLINE Abstract

17 Shine, J., and Dalgarno, L. (1974) Proc. Natl. Acad. Sci. USA 71, 1342-1346. MEDLINE Abstract

18 Weiss, R.B., Dunn, D.M., Dahlberg, A.E., Atkins, J.F., and Gesteland, R.F. (1988) EMBO J. 7, 1503-1507. MEDLINE Abstract

19 Larsen, B., Wills, N.M., Gesteland, R.F., and Atkins, J.F. (1994) J. Bacteriol. 176, 6842-6851. MEDLINE Abstract

20 Larsen, B., Peden, J., Matsufuji, S., Matsufuji, T., Brady, K., Maldonaldo, R., Wills, N.M., Fayet, O., Atkins, J.F., and Gesteland, R.F. (1995) Biochem. Cell Biol. 73, 1123-1129. MEDLINE Abstract

21 Curran, J.F. (1993) Nucleic Acids Res. 21, 1837-1843. MEDLINE Abstract

22 Tsuchihashi, Z., and Brown, P.O. (1992) Genes Dev. 6, 511-519. MEDLINE Abstract

23 Weiss, R.B., Dunn, D.M., Shuh, M., Atkins, J.F., and Gesteland, R.F. (1989) New Biologist 1, 159-169. MEDLINE Abstract

24 Fu, C., and Parker, J. (1994) Mol. Gen. Genet. 243, 434-441. MEDLINE Abstract

25 Björk, G.R. (1995) Prog. Nucleic Acids Res. Mol. Biol. 50, 263-338.

26 deBoer, H.A., and Kastelein, R.A. (1986) In Reznikoff, W., and Gold, L. (eds), Maximizing Gene Expression. Butterworths Publishers, Stoneham, MA, pp. 225-285.

27 Andersson, S.G.E., and Kurland, C.G. (1990) Microbiol. Rev. 54, 198-210.

28 Ikemura, T. (1992) In Hatfield, D.L., Lee, B.J., and Pirtle, R.M. (eds), Transfer RNA in Protein Synthesis. CRC Press, pp. 87-111.

29 Buckingham, R.H. (1994) Biochimie 76, 351-354. MEDLINE Abstract

30 Hatfield, G.W., and Gutman, G.A. (1992) In Hatfield, D.L., Lee, B.J., and Pirtle, R.M. (eds), Transfer RNA in Protein Synthesis. CRC Press, Boca Raton, FL, pp. 157-189.

31 Yarus, M., and Curran, J.F. (1992) In Hatfield, D.L., Lee, B.J., and Pirtle, R.M. (eds.), Transfer RNA in Protein Synthesis. CRC Press, Boca Raton, FL, pp. 319-365.

32 Curran, J.F., and Yarus, M. (1986) Proc. Natl. Acad. Sci. USA 83, 6538-6542. MEDLINE Abstract

33 Miller, J.H. and Albertini, A.M. (1983) J. Mol. Biol. 164, 59-71. MEDLINE Abstract

34 Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

35 Ericson, J.U., and Bjork, G.R. (1986) J. Bacteriol. 166, 1013-1021. MEDLINE Abstract

36 Davis, R.W., Botstein, D., and Roth, J.R. (1980) A Manual for Genetic Engineering: Advanced Bacterial Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

37 Vogel, H., and Bonner, O. (1956) J. Biol. Chem. 218, 97-106.

38 Maniatis, T., Fritsch, E.F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

39 Bolivar, F., Rodriguez, R.L., Greene, P.J., Betlach, M.C., Heynecker, H.L., Boyer, H.W., Crosa, J.H., and Falkow, S. (1977) Gene 2, 95-113. MEDLINE Abstract

40 Brown, C.M., Stockwell, P.A., Dalphin, M.E., and Tate, W.P. (1993) Nucleic Acids Res. 22, 3119-3123.

41 Folley, L.S., and Yarus, M. (1989) J. Mol. Biol. 209, 359-378. MEDLINE Abstract

42 Atkins, J.F., Elseviers, D., and Gorini, L. (1972) Proc. Natl. Acad. Sci. USA 69, 1192-1195. MEDLINE Abstract

43 Kurland, C.G. (1979) In Celis, J. and Smith, J.D. (eds), Nonsense Mutations and tRNA Suppressors. Academic Press, N.Y. pp. 95-108.

44 Barak, Z., Gallant, J., Lindsley, D., Kwieciszewki, B., and Heidel, D. (1996) J. Mol. Biol. 263, 140-148. MEDLINE Abstract

45 Björk, G.R. (1987) In Neidhardt, F.C., Ingraham, J.L., Low, K.B., Magasanik, B., Schaechter, M., and Umbarger, H.E. (eds), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. ASM Press, Washington, D.C., pp. 719-731.

46 Connolly, D.M., and Winkler, M.E. (1989) J. Bacteriol. 171, 3233-3246. MEDLINE Abstract

47 Gorini, L. (1971) Nature New Biol. 234, 261-264. MEDLINE Abstract

48 Bohman, K., Ruusala, T., Jelenc, P.C., and Kurland, C.G. (1984) Mol. Gen. Genet. 198, 90-99. MEDLINE Abstract

49 Thompson, R.C., Dix, D.B., Gerson, R.B., and Karim, A. (1981) J. Biol. Chem. 256, 6676-6681. MEDLINE Abstract

50 Bilgin, N., Claesens, F., Pahverk, H., and Eherenberg, M. (1992) J. Mol. Biol. 224, 1011-1027. MEDLINE Abstract

51 Curran, J.F. (1995) Nucleic Acids Res. 23, 683-688. MEDLINE Abstract

52 Gallant, J.A., and Lindsley, D. (1992) J. Mol. Biol. 223, 31-40. MEDLINE Abstract

53 Björk, G.R., Wikström, P.M., and Byström, A.S. (1989) Science 244, 986-989.

54 Hagervall, T.G., Tuohy, T.M.F., Atkins, J.F., and Björk, G.R. (1993) J. Mol. Biol. 232, 756-765. MEDLINE Abstract

55 Lim, V.I. (1997) J. Mol. Biol., 266, 877-890. MEDLINE Abstract

56 Grosjean, H.J., DeHenau, S., and Crothers, D.M. (1978) Proc. Natl. Acad. Sci. USA 75, 610-614. MEDLINE Abstract

57 Vacher, J., Grosjean, H., Houssier, C. and Buckingham, R.H. (1984) J. Mol. Biol. 177, 329-342. MEDLINE Abstract

58 Petrullo, L.A., Gallagher, J.P., and Elseviers, D. (1983) Mol. Gen. Genet. 190, 289-294. MEDLINE Abstract

59 Bouadloun, F., Srichaiyo, T., Isaksson, L.A., and Björk, G.R. (1986) J. Bacteriol. 166, 1022-1027. MEDLINE Abstract

60 Ericson, J., and Björk, G.R. (1991) J. Mol. Biol. 218, 509-516. MEDLINE Abstract

61 Esberg, B., and Björk, G.R. (1995) J. Bacteriol. 177, 1967-1975. MEDLINE Abstract

62 Bjornsson, A., and Isaksson, L. (1993) J. Mol. Biol. 232, 1017-1029. MEDLINE Abstract

63 Diaz, I., and Eherenberg, M. (1991) J. Mol. Biol. 222, 1161-1171. MEDLINE Abstract

^* To whom correspondence should be addressed. Tel:+1 910 759 5956; Fax: +1 910 759 6008; Email: curran@wfu.edu
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				R. LEIPUVIENE and G. R. BJORK A reduced level of charged tRNAArgmnm5UCU triggers the wild-type peptidyl-tRNA to frameshift RNA, May 1, 2005; 11(5): 796 - 807. [Abstract] [Full Text] [PDF]

				F. Ferracci, J. B. Day, H. J. Ezelle, and G. V. Plano Expression of a Functional Secreted YopN-TyeA Hybrid Protein in Yersinia pestis Is the Result of a +1 Translational Frameshift Event J. Bacteriol., August 1, 2004; 186(15): 5160 - 5166. [Abstract] [Full Text] [PDF]

				G. Stahl, S. N. B. Salem, L. Chen, B. Zhao, and P. J. Farabaugh Translational Accuracy during Exponential, Postdiauxic, and Stationary Growth Phases in Saccharomyces cerevisiae Eukaryot. Cell, April 1, 2004; 3(2): 331 - 338. [Abstract] [Full Text] [PDF]

				P. F. Agris Decoding the genome: a modified view Nucleic Acids Res., January 9, 2004; 32(1): 223 - 238. [Abstract] [Full Text] [PDF]

				S. Roos and H. Jonsson A high-molecular-mass cell-surface protein from Lactobacillus reuteri 1063 adheres to mucus components Microbiology, February 1, 2002; 148(2): 433 - 442. [Abstract] [Full Text] [PDF]

				S. D. Hooper and O. G. Berg Gradients in nucleotide and codon usage along Escherichia coli genes Nucleic Acids Res., September 15, 2000; 28(18): 3517 - 3523. [Abstract] [Full Text] [PDF]

				I. P. Ivanov, R. F. Gesteland, and J. F. Atkins SURVEY AND SUMMARY: Antizyme expression: a subversion of triplet decoding, which is remarkably conserved by evolution, is a sensor for an autoregulatory circuit Nucleic Acids Res., September 1, 2000; 28(17): 3185 - 3196. [Abstract] [Full Text] [PDF]

				C. Médigue, M. Rose, A. Viari, and A. Danchin Detecting and Analyzing DNA Sequencing Errors: Toward a Higher Quality of the Bacillus subtilis Genome Sequence Genome Res., November 1, 1999; 9(11): 1116 - 1127. [Abstract] [Full Text]

				B. C. Persson and J. F. Atkins Does Disparate Occurrence of Autoregulatory Programmed Frameshifting in Decoding the Release Factor 2 Gene Reflect an Ancient Origin with Loss in Independent Lineages? J. Bacteriol., July 1, 1998; 180(13): 3462 - 3466. [Abstract] [Full Text]

				Q. Qian, J. F. Curran, and G. R. Björk The Methyl Group of the N6-Methyl-N6-Threonylcarbamoyladenosine in tRNA of Escherichia coli Modestly Improves the Efficiency of the tRNA J. Bacteriol., April 1, 1998; 180(7): 1808 - 1813. [Abstract] [Full Text]

This Article Abstract Print PDF (79K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (22) Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Schwartz, R. Articles by Curran, J. F. Search for Related Content PubMed PubMed Citation Articles by Schwartz, R. Articles by Curran, J. F. Social Bookmarking          What's this?