Biochimica et Biophysica Acta 1399 (1998) 78^82
Short sequence-paper
7-Methylguanosine at the anticodon wobble position of squid
mitochondrial tRNASer GCU: molecular basis for assignment of
AGA/AGG codons as serine in invertebrate mitochondria1
Kozo Tomita, Takuya Ueda, Kimitsuna Watanabe *
Department of Chemistry and Biotechnology, Graduate School of Engineering, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-8656, Japan
Received 28 May 1998; accepted 9 June 1998
Abstract
In mitochondria of the squid, Loligo bleekeri, both the AGA and AGG codons are considered to correspond to serine
instead of arginine as in the universal genetic code, and its genome encodes a single tRNASer gene with the anticodon GCT.
Therefore, this gene product, tRNASer GCU, should be able to translate all four AGN (N; U, C, A, and G) codons as serine.
To elucidate this recognition mechanism, the tRNASer GCU was isolated from squid liver and its complete nucleotide
sequence determined. The tRNASer GCU was found to possess 7-methylguanosine (m7 G) at the wobble position of the
anticodon. This suggests that in the squid mitochondrial system, tRNASer GCU with the anticodon m7 GCU can recognize
not only the usual serine codons AGU and AGC, but also the unusual serine codons AGA and AGG, as in the case of
starfish mitochondria (Matsuyama et al., J. Biol. Chem. 273 (1988) 3363^3368). ß 1998 Elsevier Science B.V. All rights
reserved.
Keywords: tRNASer GCU; Mitochondrion; Decoding of AGA and AGG codons; 7-Methylguanosine; (Squid, Loligo bleekeri)
Deviation from the universal genetic code is one of
the unique features of animal mitochondrial (mt)
gene expression systems. Some codons in animal mitochondria are found to be located at the hot spots
of changeability [1,2]. Sequence determination of cer-
Abbreviations: m7 G, 7-methylguanosine; 8, pseudouridine;
t A, N6 -threonylcarbamoyladenosine; tRNASer GCU, serine-speci¢c tRNA with anticodon GCU; tRNASer UCU, serine-speci¢c
tRNA with anticodon UCU
* Corresponding author. Fax: +81 (3) 58006950;
E-mail: [email protected]
1
The nucleotide sequence data reported in this paper are
available in the DDBJ/EMBL/GenBank nucleotide sequence
databases under accession No. AB009838.
6
tain animal mt genomes followed by comparison of
the deduced amino acid gene sequences with those of
other species has demonstrated that codons AGA
and AGG, arginine codons in the universal genetic
code, are used di¡erently in di¡erent animal species;
they are stop codons in vertebrates, serine codons in
most invertebrates except fruit £y and mosquito ^
which use only AGA as a serine codon with no
AGG appearing in the mt genome ^ and glycine
codons in tunicates (Refs. in [1] and [2]). The mt
tRNA genes responsible for decoding AGA/AGG
or AGA codons (in addition to the usual AGU/
AGC codons) as serine in invertebrate mitochondria
have an anticodon sequence that is either GCT [2] or
TCT [3], and such tRNA genes exist solely in mt
0167-4781 / 98 / $19.00 ß 1998 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 7 - 4 7 8 1 ( 9 8 ) 0 0 0 9 9 - 2
BBAEXP 91158 24-7-98
K. Tomita et al. / Biochimica et Biophysica Acta 1399 (1998) 78^82
Fig. 1. Nucleotide sequence of squid mt tRNASer GCU shown
as a cloverleaf structure. The numbering of each residue conforms to the proposal of Sprinzl et al. [3]. Puri¢ed
tRNASer GCU was ¢rst analyzed by Donis-Keller's method [12]
and the resulting sequence was found to be identical to the sequence of the corresponding mt tRNA gene except for modi¢ed
residues, which were then determined by the method of Kuchino et al. [13].
genomes. The former anticodon sequence (GCT) occurs in star¢sh, sea urchin, fruit £y, and mosquito,
whereas the latter (TCT) is found in nematode [4]
and honeybee [5]. Since unmodi¢ed U at the wobble
position of the anticodon allows the recognition of
all four cognate codons [6], it is reasonable to assume
that nematode and honeybee mitochondria possess
tRNAsSer whose anticodon is UCU (tRNASer UCU)
so that this tRNA is able to recognize all the AGN
codons as serine; in fact, mt tRNASer UCU of the
nematode Ascaris suum has been shown to have unmodi¢ed U at its anticodon wobble position [7].
However, under the standard wobble rule, G at
the wobble position of an anticodon recognizes
only U and C. Thus, the question arises as to how
tRNASer with the anticodon GCT (at the DNA level)
in invertebrate mt systems decodes all four AGN
codons (in the cases of star¢sh and sea urchin) or
AGU/AGC/AGA codons (in fruit £y and mosquito).
It has been speculated, based on a comparison of
mt tRNASer GCU gene sequences of various organisms, that the unusual decoding capacity of mt
79
tRNASer GCU might correlate with modi¢cation at
the anticodon wobble position, and/or unusual secondary structures of the D arm or the anticodon
stem [1,2,8]. In the present study, we attempted to
address this speculation by investigating the mt genome of another invertebrate, the squid Loligo bleekeri [9].
Previous sequence analysis of about half the mt
genome of this squid suggested that codons AGA/
AGG corresponded to serine in addition to the usual
serine codons AGU/AGC [9]. To elucidate the decoding mechanism of these unusual AGA/AGG serine codons, we sequenced all of the remaining part of
the mt genome (Sasuga et al., in press; Tomota et al.,
unpublished results; DDBJ/EMBL/GenBank accession Nos. Ab005745 and AB000255). From these
sequence data it has been deduced that the whole
genome consists of 17 211 bp and contains 13 protein, two rRNA, and 22 tRNA genes, which is typical of metazoan mt DNAs reported so far [10].
We found a single tRNA gene with the anticodon
GCT (tRNASer GCU), but no tRNA gene with the
anticodon TCT existed in the entire genome, indicating that in the squid mitochondria a single
Fig. 2. Analysis of nucleotide sequence of 5P 32 P-labeled squid
mt tRNASer GCU by Donis-Keller's method [12]. Only the sequencing ladders around its anticodon region are shown. The
nucleotide corresponding to the wobble position of mt
tRNASer GCU is indicated by an arrow, which was resistant to
RNase T1. -E, OH-, T1, U2, PM, and CL denote, no treatment, digestion with alkali, RNaseT1, RNaseU2, RNasePhyM,
and RNaseCL3, respectively.
BBAEXP 91158 24-7-98
80
K. Tomita et al. / Biochimica et Biophysica Acta 1399 (1998) 78^82
Fig. 3. Two-dimensional thin-layer chromatographic (TLC) analysis of squid mt tRNASer GCU with regard to the anticodon region
(from positions 31 to 39) (A), and the nucleotide at the wobble position (B). TLC plates with 5U5 cm (A) and 10U10 cm (B) were
used for the analysis. Two solvent systems were used. The solvent used for the ¢rst dimension was isobutyric acid/concentrated ammonia/water (66:1:33 v/v/v) in both systems and those for the second dimension were 2-propanol/HCl/water (70:15:15 v/v/v) for System
(a) and ammonium sulfate/sodium phosphate 0.1 M pH 6.8/1-propanol (60 g:100 ml:2 ml) for System (b). The arrows indicate the
spot of the modi¢ed nucleotide, pm7 G.
tRNASer GCU reads all the AGN codons as serine, as
is the case in star¢sh and sea urchin mt systems.
To examine this supposition in more detail, we
isolated tRNASer GCU from squid liver by the solid-phase hybridization method [11] and determined
its nucleotide sequence (Fig. 1). Donis-Keller's method [12] showed that the nucleotide at the wobble
position was resistant to RNase T1, indicating the
existence of a modi¢ed nucleotide at that position
(Fig. 2). A modi¢ed nucleotide analysis carried out
by Kuchino's method [13] revealed that the
tRNASer GCU has 7-methylguanosine (m7 G) at the
wobble position of the anticodon (position 34) (Fig.
3), together with two pseudouridines (8) in the anticodon stem (positions 27 and 40) and N6 -threonylcarbamoyladenosine (t6 A) at the position 3P-adjacent
to the anticodon (position 37) (data not shown). The
modi¢ed nucleotide found at the anticodon wobble
position was identi¢ed by comparing its mobility
with that of an authentic sample on two di¡erent
two-dimensional thin-layer chromatograms (data
not shown). Recently, Matsuyama et al. of our laboratory reported the nucleotide sequence of star¢sh
Fig. 4. Inferred codon-anticodon relations between two mitochondrial systems: (A) squid (this study) and star¢sh [14], and
(B) mosquito [17] and fruit £y (K. Tomita et al., unpublished
result; DDBJ/EMBL/GenBank accession No. 009836). In squid
and star¢sh mitochondria, both AGA and AGG are considered
to be codons for serine and the tRNAsSer GCU possess the anticodon m7 GCU. In fruit £y and mosquito mitochondria, AGA
is considered to be a codon for serine, but no AGG appears
within the protein genes; the tRNAsSer GCU possess the anticodon GCU.
BBAEXP 91158 24-7-98
K. Tomita et al. / Biochimica et Biophysica Acta 1399 (1998) 78^82
(Asterina amurensis) mt tRNASer GCU, in which the
anticodon sequence is m7 GCU [14]. Since this is precisely the same as that of the squid mt system described here, it may be reasonably concluded that mt
tRNASer GCU with the anticodon m7 GCU is able to
recognize all four AGN codons.
However, there are other invertebrate groups ^
mosquito [15] and fruit £y [16] are two that have
been identi¢ed ^ in which the AGA codon is used
as serine but no AGG codon appears in the mt genome; nevertheless, the tRNASer gene responsible for
decoding the AGU/AGC/AGA codons has the anticodon GCT. It has already been revealed that mosquito mt tRNASer GCU has unmodi¢ed G at the
anticodon wobble position [17]. We sequenced a fruit
£y (Drosophila melanogaster) counterpart and found
that the nucleotide at the wobble position of the mt
tRNASer GCU was also unmodi¢ed G, and no other
G derivative was found at that position (K. Tomita
et al., unpublished result; DDBJ/EMBL/GenBank
accession No. 009836). From these ¢ndings, it is considered that unmodi¢ed G can recognize not only the
AGU/AGC codons but also AGA because, with respect to invertebrate mitochondrial systems, there
has been no report indicating the existence of a competitor tRNA with an anticodon U*CU (U*; derivatives of U) or AGA/AGG-speci¢c release factor(s)
(RF), which would possibly win in competition with
tRNASer GCU for the recognition of AGA/AGG codons.
Considering all the above observations together, it
can be speculated that modi¢cation of the wobble
residue of mt tRNASer GCU from unmodi¢ed G to
m7 G may expand its decoding capacity from three
(AGU/AGC/AGA) to four (AGN) codons (Fig. 4).
Although tRNAsSer GCU of vertebrate mitochondria,
in which both AGA and AGG are considered to be
stop codons [1,2], also possess the GCU anticodon,
these tRNAs have no modi¢cation at the wobble
position [3]. In this case, the RF would recognize
AGA and AGG codons strongly, so that
tRNAsSer GCU with the wobble position occupied
by unmodi¢ed G could recognize only codons
AGU and AGC and not AGA.
In order to verify whether the m7 GCU anticodon
of tRNASer GCU does actually decode AGA/AGG
codons in addition to AGU/AGC, we recently prepared synthetic Escherichia coli tRNAAla with the
81
anticodon m7 GCU which was chargeable with alanine using a combination of in vitro transcription
[18] and microsurgery techniques [19], and examined
its decoding ability towards AGN codons in an E.
coli in vitro translation system. Preliminary results
indicate that the tRNAAla variant with anticodon
m7 GCU can recognize the codons AGU and AGC,
but not AGA and AGG (our unpublished observation). Therefore, other structural features of mt
tRNASer GCU, such as a truncated D arm and/or
modi¢cation at the position 3P-adjacent to the anticodon, may also need to be taken into consideration,
as discussed previously [8]. There are some reports
indicating that mutation in the D arm induces unusual decoding [20,21]. Furthermore, modi¢cation of
the nucleotide 3P-adjacent to the anticodon (position
37) is known to a¡ect codon-anticodon interaction
[22,23]. The decoding capability of tRNASer GCU toward AGA/AGG codons may be achieved by a combination of these factors. Further study is necessary
for clarifying this.
We thank Mr. T. Kaifu for showing us the unpublished partial sequence of squid mt DNA, Dr. S.
Yokobori (at present, Tokyo University of Pharmacy
and Life Science) for data analysis, Dr. Y. Watanabe
(at present, Dalhousie University, Canada) for valuable discussion, and Mr. H. Nambu (at present,
Banyu Pharmaceutical Co.) for continuous encouragement. This work was supported by a Grant-inAid for Scienti¢c Research on Priority Areas from
the Ministry of Education, Science, Sports and Culture of Japan and by the Human Frontier Science
Program Organization.
References
[1] S. Osawa, T.H. Jukes, K. Watanabe, A. Muto, Microbiol.
Rev. 56 (1992) 229^264.
[2] K. Watanabe, S. Osawa, in: D. So«ll, U. Rajbhandary (Eds.),
tRNA: Structure, Biosynthesis, and Function, American Society for Microbiology, Washington, DC, 1995, pp. 225^250.
[3] M. Sprinzl, C. Steegborn, F. Hubel, S. Steinberg, Nucleic
Acids Res. 24 (1996) 68^72.
[4] R. Okimoto, J.L. MacFarlane, D.O. Clary, D.R. Wolstenholme, Genetics 130 (1992) 471^498.
[5] R.H. Crozier, Y.C. Crozier, Genetics 133 (1993) 97^117.
[6] R.J. Hechman, J. Sarno¡, Y.S. Alzner-DeWeerd, U.L. RajBhandary, Proc. Natl. Acad. Sci. USA 77 (1980) 3159^3163.
BBAEXP 91158 24-7-98
82
K. Tomita et al. / Biochimica et Biophysica Acta 1399 (1998) 78^82
[7] Y. Watanabe, H. Tsurui, T. Ueda, R. Furushima, S. Takamiya, K. Kita, K. Nishikawa, K. Watanabe, J. Biol. Chem.
269 (1994) 22902^22906.
[8] T. Ueda, K. Watanabe, Origin of Life and Evolution of the
Biosphere, 23 (1993) 345^364.
[9] T. Shimayama, H. Himeno, J. Sasuga, S. Yokobori, T.
Ueda, K. Watanabe, Nucleic Acids Res. Symp. Ser. 28
(1990) 73^74.
[10] D.R. Wolstenholme, Int. Rev. Cytol. 141 (1992) 173^216.
[11] H. Tsurui, Y. Kumazawa, R. Sanokawa, Y. Watanabe, T.
Kuroda, K. Watanabe, A. Wada, T. Shirai, Anal. Biochem.
221 (1994) 166^172.
[12] H. Donis-Keller, Nucleic Acids Res. 8 (1980) 3133^3142.
[13] Y. Kuchino, N. Hanyu, S. Nishimura, Methods Enzymol.
155 (1987) 376^396.
[14] S. Matsuyama, T. Ueda, P.F. Crain, J.A. McCloskey, K.
Watanabe, J. Biol. Chem. 273 (1998) 3363^3368.
[15] C.B. Beard, D.M. Hamm, F.H. Collins, Insect Mol. Biol. 2
(1993) 103^124.
[16] D.O. Clary, D.R. Wolstenholme, J. Mol. Evol. 22 (1985)
252^271.
[17] D.T. Dubin, C.C. HsuChen, G.R. Cleaves, K.D. Timko,
J. Mol. Biol. 176 (1984) 251^260.
[18] J.R. Sampson, O.C. Uhlenbeck, Proc. Natl. Acad. Sci. USA
84 (1988) 2185^2188.
[19] T. Ohyama, K. Nishikawa, S. Ikemura, J. Biochem. 99
(1986) 859^866.
[20] D. Hirsch, J. Mol. Biol. 58 (1971) 439^458.
[21] L.A. Raftery, J.A. Bermingham, M. Yarus, J. Mol. Biol. 190
(1986) 513^517.
[22] T.G. Hagevall, J.U. Ericson, K.B. Esberg, G.R. Bjork, Biochim. Biophys. Acta 1050 (1990) 263^266.
[23] J. Weissenbach, H. Grosjean, Eur. J. Biochem. 116 (1981)
207^213.
BBAEXP 91158 24-7-98