- Email: [email protected]
Identification of the UUG codon as a translational initiation codon in vivo
9. NoZ. Biol. (1975) 95, 327330
LETTERS TO THE EDITOR
cation of the UUG Codon as a Translationa Initiation Codon in vivo The lac repressor protein was purified from an Escherichia coli strain carrying an amber mutation in the la& gene and the tyrosine-inserting amber suppressor, Su3. Protein sequencing showed a change at position 62 in the repressor polypeptide chain from leuoine to t,yrosine, proving that the amber was derived from ~a.UUG codon at this point in the message. This establishes UUG as an initiation codon in wivo, since it has been previously shown that translational reinitiation can occur at position 62. nitiation of Escherichia coli protein biosynthesis is generally specified by the codon AUG (for a recent review see Haselkorn & Rothman-Denes, 1973). Nucleotide sequence analyses of initiation regions of 13 natural mRNA molecules, including mne -from the closely-related RNA bacteriophages, indicate that all use AUG as the initiation codon (Steitz, 1969 ; Hindley & Staples, 1969; Staples et al., 1971; Gupta et al., 1970; Min Jou et al., 1972; Robertson et al., 1973; Bronson et al., 1973; Maizels, 1974; Muss0 et al., 1974; Pieczenik et al., 1974). The only in v’iivo evidence for a non-AUG initiation codon comes from the overlapping of a nucleotide sequence of a segment of RNA from bacteriophage MS2 with part of the sequence of an initiation gion from the closely-related bacteriophage R17 (Volckaert & Fiers, 1973). The S2 sequence ends in GUG, which corresponds to the AUG initiation codon in the R17 (Steitz, 1969). 19%vitro experiments indicate that N-formylmethionine-dependent initiation of protein synthesis can occur at the codons AUG, GUG, and to a lesser extent at GUA (Ghosh et al., 1967; Clark & Marcker, 1966). Although UUG has not been shown to act as an initiation codon in vitro, the binding of N-formylmethionyl-tRNA to ribosomes is stimulated by the trinucleotides AUG, GUG and UUG (Clark & Marcker, 1966; Ghosh et al., 1967; Dube et al., 1969). The recognition of the UUG codon by the initiator N-formylmethionyl-tRNA suggests that UUG might be capable of initiation of protein synthesis. In our studies of the genetics and protein chemistry of the E. co& EacI gene and its protein product the lac repressor, we have characterized the phenomenon of translational reinitiation in the early part of the lad message (Platt et al., 1972; Gamem et al., 1973; Files et al., 1974). This ia vivo reinitiation of protein synthesis following chain termination at a nonsense codon ocours at a site in the same messenger RNA some distance past the terminator codon. Our results show that there are three inphase translational reinitiation sites within the part of the lad message coding for the lirst 62 amino acids of the kc repressor (Platt et al., 1972; Ganem et al., 1973; et al., 1974). One of these sites (termed 7rh2)is at the first internal in-phase AUG , which codes for methionine residue 42 of the wild-type lac repressor molecule. We have proposed that the other two sites (termed 7rs3and T&I involve non-AUG 327
328
J. G. FILES
ET
BE.
codons which normally code for valine residue 23 and leucine residue 62 of the repressor. We have suggested that the rZ3 initiation codon could be GUG (or possibly QUA) and the r6a could be UUG, since in vitro studies suggest that these codons might be able to initiate protein synthesis iti vivo (Clark & Marcker, 1966; Ghosh et al., 1967; Dube et al., 1969). Because UUG can be converted to the amber codon (UAG) by a single base change, we screened a collection of amber mutants for those with a mutation at position 62 of the lac repressor. Two hundred and ninety-six independently isolated u.v.-induced amber mutations, analyzed by deletion mapping and by pattern of suppression (Coulondre & Miller, unpublished results) with different amber suppressors, were separated into 35 different sites in the lad gene. One of these sites, represented by ten independent isolates, mapped in the region immediately past a site known to specify glutamine residue 60 (Ganem et al., 1973). A suppressed derivative of this amber mutant (XAZ) was prepared by transducing it with Pl phage into a strain containing the tyrosineinserting amber suppressor Su3. Zac repressor was isolated from this strain and subjected to amino acid sequence analysis to determine the position of tyrosine insertion, and thus the position of the amber site. An amino acid sequence comparison of the suppressed derivative of mutant XA2 with wild-type repressor (Fig. 1) shows a change from leucine to tyrosine in position 62. This demonstrates that the internal amber codon in XA2 is at the position of the
50 60 62 -Gly-Lys-Gln-Ser-Leu-Leu-Ite-Gly-Val-Ala-Thr-
(a)
Wild
(b)
re-start
(c)
Su3+derivative
Cd)
m’RNA
type
protein
7r62
of XA2
Met-Leu-lte
-Gin-Ser-Tyr-Leu-iie -
64
66
68
-Gly-Val
-Ala-Thr-
-Gly-Vai
-Ala-Thr-
-GG)(-AA~-CAG-UCG-UUG-UUX-AUX-GGX-----------------
FIG. 1. (a) The amino acid sequence of residues 68 to 68 of wild-type Zac repressor (Canem et al., 1973). (b) The amino-terminal sequence of the ~~~ internal restart protein (Ganem et aE., 1973). (c) The amino acid sequence of residues 60 to 68 of the Su3suppressed amber mutant XA2. Zac repressor isolated from this mutant was digested with trypsin under native conditions (Platt es al., 1973). The trypsin-resistant core was isolated and subjected to automated Edman degmdation (Files & Weber, manuscript in preparation) giving an amino-terminal sequence corresponding to residues 60 to 68. (d) RNA sequence of the ZacI message, deduced from our previous identification of the corresponding codons (Ganem et al., 1973; Miller et al., 1976; Files & Weber, manusoript in preparation) and the identification of the oodon for residue 62 given in this report.
codon normally specifying leucine residue 62. The relatively large number of independent isolates of XA2 indicates that it is a single site mutation. Of the six possible leucine codons, only UUG can mutate to amber (UAG) by a single base change, and
LETTERS
TO THE
339
EDITOR
therefore the leuclne codon specifying residue 62 must be UUG;. Since this leucme codon gives rise to translational reinitiation (Ganem et ccl., 1973), we conclude that UUG can act as an initiation codon in vivo. Since the polypept,ide produced by translational reinitiation at this codon begins with methionine, we further conclude that there is an ila vivo ambiguity in the reading of UUG. The UUG codon can be translated into two different amino acids depending on whether it is recognized during the initiation or elongation of protein synthesis. One of the authors (J. Et. M.) was supported jF.N.3.890.72).
by a grant from the Swiss National
Fund
Biological Laboratories Harvwd Univarsity Cambridge, Mass. 02138, U.S.A.
JAMBS G.FILES
Max Pianck Institute for Biophysical Chemistry Department of Biochemistry Qiittingen, West Germany
KLILUS WEBER
Department of Molecular Wnivorsity of Geneva Geneva, Switzerland
CHRISTINE C!OUL~NDBE JEFFFREYH.MILI,ERP
Received
Biology
6 M.arch 1975
,t To whom reprint requests should be addressed. REFERENCES Bronson,
M. J., Squires, C. & Yanofsky,
C. (1973). Proc.
Nat.
Acad.
SC<., U.S.A.
70,
2335-2339.
Clark., B. F. C. & Marcker, K. A. (1966). J. Mol. Biol. 17, 394-406. Dube, S. K., Rudland, P. S., Clark, B. F. C. & Marcker, K. A. (1969). Gold &wing Ha&or Symp. Qua&. Bid. 34, 161-166. Files, J. G., Weber, K. & Miller, J. H. (1974). Proc. Nat. Acad. Xci., U.S.A. 71, 667-570. Ganem, D., Miller, J. H., Files, J. G., Platt, T. & Weber, K. (1973). PVOC.Nat. Acad. Sci., U.S.A. 70, 3165-3169. Ghosh, H. P., S611,D. & Khorana, H. G. (1967). J. Mol. Biol. 25, 275-298. Gupta, S. L., Chen, J., Schaefer, L., Lengyel, P. & Weissman, S. M. (1970). B&x&n. Biophyya.
Res. Commun.
39, 883-886.
Haselkorn,
R. & Rothman-Denes, L. B. (1973). Annu. Rev. Biochem. 42, 397-438. Eindey, J. & Staples, D. H. (1969). Nature (London), 224, 964967. Maizels, N. M. (1974). Natwe (London), 249, 647-649. her, J. H., Coulondre, C., Schmeissner, U., Schmitz, A. & Lu, P. (1975). Sgmp. 05% Protein-LigawZ Interaction, Walter de Gruyter, Berlin, 238-252. Min Jou, W., Haegeman, G., Ysebaert, M. & Fiers, W. (1972). Nature (London), 237, 8246.
Musso, R. E,, de Crombrugghe, B., Pastan, I., Sklar, J., Yot, P. & Weissman, S. M. (1974). Proc. Nat. Acad. Sci., U.S.A. 71, 4940-4944. Pieczenik, G., Model, P. & Robertson, D. H. (1974). J. Mol. Bid. 90, 191-214. Platt, T., Weber, K., Ganem, D. & Miller, J. H. (1972). Proc. Nat. Acad. Sci., U.S.A. 69, 897-701. Platt, T., Files, J. 6. & Weber, K. (1973). J. Biol. Chewa. 248, 110-121.
J. G. FILES
330
ET AL.
Robertson, 3% D., Barre& B. G., Weith, H. L. & Donelson, J. E. (1973). Nature New Biol. 241, 38-40. Staples, D. H., Him&y, J., Billeter, n/r. A. & Weissmann, C. (1971). Nature New Biol. 234, 202-204. St&z, J. A. (1969). Nature (London), 224, 967-964. Volckaert, G. & Fiers, W. (1973). F’EBX Letters, 35, 91-96. iti Proof: We have recently without the use of a mutagen.
Note Added
as XA2
isolated an amber mutation
at the same site
LETTERS TO THE EDITOR
cation of the UUG Codon as a Translationa Initiation Codon in vivo The lac repressor protein was purified from an Escherichia coli strain carrying an amber mutation in the la& gene and the tyrosine-inserting amber suppressor, Su3. Protein sequencing showed a change at position 62 in the repressor polypeptide chain from leuoine to t,yrosine, proving that the amber was derived from ~a.UUG codon at this point in the message. This establishes UUG as an initiation codon in wivo, since it has been previously shown that translational reinitiation can occur at position 62. nitiation of Escherichia coli protein biosynthesis is generally specified by the codon AUG (for a recent review see Haselkorn & Rothman-Denes, 1973). Nucleotide sequence analyses of initiation regions of 13 natural mRNA molecules, including mne -from the closely-related RNA bacteriophages, indicate that all use AUG as the initiation codon (Steitz, 1969 ; Hindley & Staples, 1969; Staples et al., 1971; Gupta et al., 1970; Min Jou et al., 1972; Robertson et al., 1973; Bronson et al., 1973; Maizels, 1974; Muss0 et al., 1974; Pieczenik et al., 1974). The only in v’iivo evidence for a non-AUG initiation codon comes from the overlapping of a nucleotide sequence of a segment of RNA from bacteriophage MS2 with part of the sequence of an initiation gion from the closely-related bacteriophage R17 (Volckaert & Fiers, 1973). The S2 sequence ends in GUG, which corresponds to the AUG initiation codon in the R17 (Steitz, 1969). 19%vitro experiments indicate that N-formylmethionine-dependent initiation of protein synthesis can occur at the codons AUG, GUG, and to a lesser extent at GUA (Ghosh et al., 1967; Clark & Marcker, 1966). Although UUG has not been shown to act as an initiation codon in vitro, the binding of N-formylmethionyl-tRNA to ribosomes is stimulated by the trinucleotides AUG, GUG and UUG (Clark & Marcker, 1966; Ghosh et al., 1967; Dube et al., 1969). The recognition of the UUG codon by the initiator N-formylmethionyl-tRNA suggests that UUG might be capable of initiation of protein synthesis. In our studies of the genetics and protein chemistry of the E. co& EacI gene and its protein product the lac repressor, we have characterized the phenomenon of translational reinitiation in the early part of the lad message (Platt et al., 1972; Gamem et al., 1973; Files et al., 1974). This ia vivo reinitiation of protein synthesis following chain termination at a nonsense codon ocours at a site in the same messenger RNA some distance past the terminator codon. Our results show that there are three inphase translational reinitiation sites within the part of the lad message coding for the lirst 62 amino acids of the kc repressor (Platt et al., 1972; Ganem et al., 1973; et al., 1974). One of these sites (termed 7rh2)is at the first internal in-phase AUG , which codes for methionine residue 42 of the wild-type lac repressor molecule. We have proposed that the other two sites (termed 7rs3and T&I involve non-AUG 327
328
J. G. FILES
ET
BE.
codons which normally code for valine residue 23 and leucine residue 62 of the repressor. We have suggested that the rZ3 initiation codon could be GUG (or possibly QUA) and the r6a could be UUG, since in vitro studies suggest that these codons might be able to initiate protein synthesis iti vivo (Clark & Marcker, 1966; Ghosh et al., 1967; Dube et al., 1969). Because UUG can be converted to the amber codon (UAG) by a single base change, we screened a collection of amber mutants for those with a mutation at position 62 of the lac repressor. Two hundred and ninety-six independently isolated u.v.-induced amber mutations, analyzed by deletion mapping and by pattern of suppression (Coulondre & Miller, unpublished results) with different amber suppressors, were separated into 35 different sites in the lad gene. One of these sites, represented by ten independent isolates, mapped in the region immediately past a site known to specify glutamine residue 60 (Ganem et al., 1973). A suppressed derivative of this amber mutant (XAZ) was prepared by transducing it with Pl phage into a strain containing the tyrosineinserting amber suppressor Su3. Zac repressor was isolated from this strain and subjected to amino acid sequence analysis to determine the position of tyrosine insertion, and thus the position of the amber site. An amino acid sequence comparison of the suppressed derivative of mutant XA2 with wild-type repressor (Fig. 1) shows a change from leucine to tyrosine in position 62. This demonstrates that the internal amber codon in XA2 is at the position of the
50 60 62 -Gly-Lys-Gln-Ser-Leu-Leu-Ite-Gly-Val-Ala-Thr-
(a)
Wild
(b)
re-start
(c)
Su3+derivative
Cd)
m’RNA
type
protein
7r62
of XA2
Met-Leu-lte
-Gin-Ser-Tyr-Leu-iie -
64
66
68
-Gly-Val
-Ala-Thr-
-Gly-Vai
-Ala-Thr-
-GG)(-AA~-CAG-UCG-UUG-UUX-AUX-GGX-----------------
FIG. 1. (a) The amino acid sequence of residues 68 to 68 of wild-type Zac repressor (Canem et al., 1973). (b) The amino-terminal sequence of the ~~~ internal restart protein (Ganem et aE., 1973). (c) The amino acid sequence of residues 60 to 68 of the Su3suppressed amber mutant XA2. Zac repressor isolated from this mutant was digested with trypsin under native conditions (Platt es al., 1973). The trypsin-resistant core was isolated and subjected to automated Edman degmdation (Files & Weber, manuscript in preparation) giving an amino-terminal sequence corresponding to residues 60 to 68. (d) RNA sequence of the ZacI message, deduced from our previous identification of the corresponding codons (Ganem et al., 1973; Miller et al., 1976; Files & Weber, manusoript in preparation) and the identification of the oodon for residue 62 given in this report.
codon normally specifying leucine residue 62. The relatively large number of independent isolates of XA2 indicates that it is a single site mutation. Of the six possible leucine codons, only UUG can mutate to amber (UAG) by a single base change, and
LETTERS
TO THE
339
EDITOR
therefore the leuclne codon specifying residue 62 must be UUG;. Since this leucme codon gives rise to translational reinitiation (Ganem et ccl., 1973), we conclude that UUG can act as an initiation codon in vivo. Since the polypept,ide produced by translational reinitiation at this codon begins with methionine, we further conclude that there is an ila vivo ambiguity in the reading of UUG. The UUG codon can be translated into two different amino acids depending on whether it is recognized during the initiation or elongation of protein synthesis. One of the authors (J. Et. M.) was supported jF.N.3.890.72).
by a grant from the Swiss National
Fund
Biological Laboratories Harvwd Univarsity Cambridge, Mass. 02138, U.S.A.
JAMBS G.FILES
Max Pianck Institute for Biophysical Chemistry Department of Biochemistry Qiittingen, West Germany
KLILUS WEBER
Department of Molecular Wnivorsity of Geneva Geneva, Switzerland
CHRISTINE C!OUL~NDBE JEFFFREYH.MILI,ERP
Received
Biology
6 M.arch 1975
,t To whom reprint requests should be addressed. REFERENCES Bronson,
M. J., Squires, C. & Yanofsky,
C. (1973). Proc.
Nat.
Acad.
SC<., U.S.A.
70,
2335-2339.
Clark., B. F. C. & Marcker, K. A. (1966). J. Mol. Biol. 17, 394-406. Dube, S. K., Rudland, P. S., Clark, B. F. C. & Marcker, K. A. (1969). Gold &wing Ha&or Symp. Qua&. Bid. 34, 161-166. Files, J. G., Weber, K. & Miller, J. H. (1974). Proc. Nat. Acad. Xci., U.S.A. 71, 667-570. Ganem, D., Miller, J. H., Files, J. G., Platt, T. & Weber, K. (1973). PVOC.Nat. Acad. Sci., U.S.A. 70, 3165-3169. Ghosh, H. P., S611,D. & Khorana, H. G. (1967). J. Mol. Biol. 25, 275-298. Gupta, S. L., Chen, J., Schaefer, L., Lengyel, P. & Weissman, S. M. (1970). B&x&n. Biophyya.
Res. Commun.
39, 883-886.
Haselkorn,
R. & Rothman-Denes, L. B. (1973). Annu. Rev. Biochem. 42, 397-438. Eindey, J. & Staples, D. H. (1969). Nature (London), 224, 964967. Maizels, N. M. (1974). Natwe (London), 249, 647-649. her, J. H., Coulondre, C., Schmeissner, U., Schmitz, A. & Lu, P. (1975). Sgmp. 05% Protein-LigawZ Interaction, Walter de Gruyter, Berlin, 238-252. Min Jou, W., Haegeman, G., Ysebaert, M. & Fiers, W. (1972). Nature (London), 237, 8246.
Musso, R. E,, de Crombrugghe, B., Pastan, I., Sklar, J., Yot, P. & Weissman, S. M. (1974). Proc. Nat. Acad. Sci., U.S.A. 71, 4940-4944. Pieczenik, G., Model, P. & Robertson, D. H. (1974). J. Mol. Bid. 90, 191-214. Platt, T., Weber, K., Ganem, D. & Miller, J. H. (1972). Proc. Nat. Acad. Sci., U.S.A. 69, 897-701. Platt, T., Files, J. 6. & Weber, K. (1973). J. Biol. Chewa. 248, 110-121.
J. G. FILES
330
ET AL.
Robertson, 3% D., Barre& B. G., Weith, H. L. & Donelson, J. E. (1973). Nature New Biol. 241, 38-40. Staples, D. H., Him&y, J., Billeter, n/r. A. & Weissmann, C. (1971). Nature New Biol. 234, 202-204. St&z, J. A. (1969). Nature (London), 224, 967-964. Volckaert, G. & Fiers, W. (1973). F’EBX Letters, 35, 91-96. iti Proof: We have recently without the use of a mutagen.
Note Added
as XA2
isolated an amber mutation
at the same site