ABSTRACT
We cloned and sequenced a portion of chloroplast DNA from the hornwort
Anthoceros formosae.
A nucleotide sequence of 7556 bp contained structures similar to those of
ndhK, ndhC
,
trnV
,
trnM
,
atpE
,
atpB
,
rbcL
,
trnR
and
accD.
The arrangement of these was the same as that of other chloroplast DNA. However,
two nonsense codons were located within the putative coding region of
rbcL,
although they were used as putative termination codons of the genes. RNA was
extensively edited in the transcripts of
rbcL
when
cDNA sequences were analyzed. The unusual nonsense codons of TGA and TAA became
CGA and CAA respectively. These are examples of U to C type RNA editing, which
was never been found before in chloroplast mRNA. In general, 13 Cs of genomic
DNA were found as Ts in the cDNA sequence and seven Ts were found as Cs. This
is the first finding of RNA editing on the transcripts of
rbcL
and also in bryophytes. This event had been thought to arise in land plants
after the split of bryophytes. The origin of RNA editing is discussed in
relation to the landing of green plants.
One serious challenge against the central dogma of molecular biology is the
discovery of RNA editing. Genetic information not found in the genomic template
can be transferred into mRNA after transcription. RNA editing was first
discovered in the kinetoplast genetic system of trypanosome (
1
), later in the nuclear encoded mRNA of human apolipoprotein (
2
), and in a number of transcripts encoded by plant mitochondrial DNA (
3
-
7
). These events were found in all major groups of land plants except bryophytes
(
7
). In angiosperm chloroplasts, RNA editing has also been identified (
8
-
18
). In chloroplasts, all RNA editing found so far has been C to U substitutions,
whereas U to C substitutions have also been found in plant mitochondria. It had
been thought that editing arose in early land plants after the split of
bryophytes because no editing has been identified in representatives of green
algae and in liverwort (
7
). However, we found U to C RNA editing as well as C to U in
rbcL
transcripts of hornwort (
Anthoceros formosae
)
chloroplasts.
The following oligonucleotide primers designed from genomic DNA sequence of
A. formosae
were synthesized and obtained from Sawaday Technology (Tokyo, Japan) or
Biologica (Nagoya, Japan):
P1, 5'-AGTAGACTTCGTCCCTGCAAGAGTT;
P2, 5'-TCCTCTCCAGCAACAGGTTCAATGT;
P3, 5'-AACTGGTACATGGACTACTGTTTGG;
P4, 5'-CTACTGTACCTGGATGAATATGATC;
P5, 5'-ACCGACAGACAAAGAAATCATGGTA;
P6, 5'-AAAACGAAAGAGCTGAATTGCAA;
P7, 5'-CCTCCTGTCAAATAATCATGCATTAC.
The positions and orientations of these oligonucleotides are shown in Figure
2
.
Thalli of the hornwort
A.formosae
were incubated at 25oC on 1/2 KnopII-agar medium under continuous fluorescent light. The thalli were
harvested and homogenized in a buffer containing 50 mM Tris-HCl (pH 8.0), 10 mM EDTA, 20% sucrose, 5 mM 2-mercaptoethanol, 0.1% BSA. The homogenate was filtered though
cheese-cloth and unbroken cells were precipitated by centrifugation at 1000
g
for 10 s. The chloroplast-rich fraction was precipitated from the supernatant by centrifugation at
3000
g
for 10 min. Nucleic acids were extracted from the chloroplast-rich fraction as described by Dellaporta
et al
. (
19
). Contaminating RNA was removed from the DNA sample by digestion with RNaseA.
DNA from the chloroplast-rich fraction was partially digested with
Bgl
II and 15-20 kb fragments were electro-eluted from an agarose gel as described (
20
). The fragments were then ligated to the Charomid 9-28 vector (Nippon Gene), packed in phage particles using Gigapack
(Funakoshi), and infected into
Escherichia
coli
DH5[alpha]. 100 colonies grown in the presence of ampicillin were transferred to
Biodyne A membranes (Pall) and hybridized with a
Hin
dIII fragment containing part of the
rbcL
gene of
Angiopteris lygodiifolia
(
21
)
.
The membranes were washed with 0.1* SSC containing 0.1% SDS at 50oC and exposed to Fuji X-ray film. Five positive clones, pCH13, pCH47, pCH48, pCH62 and
pCH79, were identified. Plasmids were isolated by the boiling lysis (
22
) and analyzed by Southern hybridization using the
Hin
dIII fragment described above as the probe.
A
Kpn
I
fragment of 7.5 kb was excised from pCH79 and ligated into pUC18. The resulting
plasmid named pK79 was used for sequencing after subcloning into the pUC18
vector. Plasmids pCH13, pCH47, pCH48 and pCH62 were cut with
Sal
I
and religated. The five resulting plasmids were cut with
Hin
dIII, then 2.9 kb fragments were subcloned into pUC18 and sequenced by dideoxy
chain-termination (
23
) using the BcaBest sequencing kit (Takara) or 7-deaza Sequenase Ver. 2.0 (USB). The products of the sequencing reactions
were applied to a denaturing polyacrylamide gel and exposed to Fuji X-ray film. The resulting sequences were treated with Genetyx software Ver.
7.06 (SDC).
Total cellular nucleic acids were prepared using CTAB (
24
) with a slight modification. Frozen thalli (3 g) of
A.formosae
were disrupted with quartz sand and nucleic acids were extracted in 10 ml of
extraction buffer containing 100 mM Tris-HCl (pH 8.0), 20 mM EDTA, 1.4 M NaCl, 1% cetyltrimethylammonium bromide
(CTAB) and 1% 2-mercaptoethanol at 60oC for 30 min. Total RNA was precipitated by adding LiCl to a final
concentration of 2 M after extraction of chloroform-isoamyl alcohol (24/1,v/v). Contaminating DNA was removed from the RNA
sample by digestion with RNase-free DNase (Boehringer Mannheim).
The total RNA was used to synthesize cDNA using a commercial kit (lst-Strand cDNA Synthesis Kit, Clontech). Total RNA (0.5 [mu]g) and 20 pmol random hexamer primer were annealed at 70oC for 2 min and cooled to 0oC. They were then mixed with 10 pmol of each of four
nucleotide triphosphates (dNTP), 20 U recombinant RNase inhibitor and 200 U M-MLV reverse transcriptase in a total volume of 20 [mu]l. The reaction proceeded at 42oC for 60 min and at 94oC for 5 min, then 80 [mu]l of water was added.
The cDNA was amplified by means of a PCR Amplification Kit (Takara) and the
primer pairs P1-P2, P3-P4 and P5-P6 (described above) by Program Temp Control System PC-700 (Astec). The reaction (100 [mu]l) contained 10 [mu]l of diluted cDNA, 100 pmol of each primer
pair, 20 pmol of each of dNTP and 2.5 U
Taq
DNA polymerase. The thermocycles were 94oC for 2 min, 55oC for 2 min and 72oC for 2 min. Typically, the reaction proceeded though 30 cycles,
followed by 72oC for 7 min. Amplified cDNAs were separated by agarose gel electrophoresis
and extracted from the gel using Gene Clean II Kit (Bio 101). They were then
filled-in with the Klenow fragment (Nippon Gene) and T4 polynucleotide kinase
(Takara) and ligated to pUC18.
E.coli
DH5[alpha] was transformed with the ligated DNA. At least six colonies from each
transformation were picked up and plasmids prepared from them by boiling lysis
(
22
) were sequenced as described above using appropriate primers.
A phylogenetic tree was obtained from amino acid sequences which were deduced
from the genomic DNA and cDNA sequences of
rbcL
genes. The maximum likelihood tree of green plants was selected by PROTML (JTT
model), after comparing all the trees including the best topology among 10 000
random addition heuristic searches by PROTML, the 262 most parsimonious trees
by PAUP using PROTPARS weighing and the five NJ tree topologies using various
means to calculate genetic distances, namely CLUSTALW, PROTDIST and PROTML.
Stop codons except for terminal one in
Megaceros
are substituted by X (unknown residue), and the opening codon T in
Selaginella
and
Isoetes
is accepted as it is.
Five clones (pCH13, pCH47, pCH48, pCH62 and pCH79) assumed to contain
rbcL
were selected from a DNA library of
A.formosae
by colony and Southern hybridization as described in Materials and Methods. A
7.5 kb
Kpn
I
fragment from pCH79 assumed to contain
rbcL
was subcloned. The resulting plasmid pK79 was sequenced. We found that the
sequence contained portions homologous to
ndhK
,
ndhC
,
trnV
,
trnM
,
atpE
,
atpB
,
rbcL
,
trnR
and
accD
in this order (Fig.
1
). The gene arrangement is the same as that of
Marchantia
(
25
) and that of
Nicotiana
(
26
) except for
trnR,
which is not found in
Nicotiana,
suggesting that the sequence is that of chloroplast DNA. However, the nonsense
codons TAA and TGA were identified in the putative coding regions of
rbcL,
though TAA and TGA were used as putative stop codons in
rbcL
and
atpB
, respectively. In addition to these, an ACG codon was found at the position of
putative initiation codon of
atpB
(in preparation). These findings indicate that RNA is edited in these
transcripts.
Typical RNA editing found in chloroplasts is the creation of an initiation codon
by converting ACG to AUG (
8
,
9
,
12
,
15
,
17
). This type of editing is an important control system of organelle gene
expression by the nucleus when the editing system is derived from nuclear
genes, because translation cannot start until the start site is edited (
28
). The situation is similar in
A.formosae,
because the nonsense codons TGA and TAA are located close to the start codon.
Therefore, even if translation starts before editing, it will stop soon after
40 amino acids are synthesized. It has also been observed in maize
chloroplasts, where four C to U editing sites clustered within 150 nucleotides
of 5' terminal region of
rpoB
transcripts (
14
). The unusual initiation codon ACG of the
rbcL
sequences has been found in the lycopodophytes
Selaginella
(L11280) and
Isoetes
(L11054) suggesting the creation of the initiation codon AUG in the
lycopodophytes by RNA editing (
29
).
The RNA editing found in chloroplasts to date is a C to U substitution. However
U to C substitution is found in the hornwort chloroplasts, which should be
found in many other chloroplasts of primitive land plants. The DNA sequence of
the hornwort
Megaceros enigmaticus
(L13481) contains two nonsense codons (TGA) within the putative coding region
suggesting U to C conversion in this plant. The reason why a U to C substitution had not been
identified in chloroplasts may be mainly because such studies have been limited
to higher plants. Alternatively, it cannot be ruled out that a U to C
substitution has been acquired only by anthocerophytes.
RNA editing was thought to arise in land plants after the split of bryophytes,
because no editing in mitochondria had been found either in bryophytes or in
algae (
7
). However, we found extensive RNA editing in the hornwort,
A.formosae.
We believe that RNA editing will be found in the chloroplasts of many other
primitive land plants, because 20 editing sites per 1428 nucleotides in the
rbcL
transcript of the hornwort (this study), eight sites per 480 nucleotides in the
5' region of
chlL
of the fern
Pteridium
(K. Yamada, personal communication), and several editing sites in 10 genes and
the ORF of gymnosperms have disappeared during evolution to the angiosperms
Nicotiana
(T. Wakasugi, personal communication). The same tendency has also been
identified in trypanosomatid mitochondria, where entire genes are edited in the
early diverging branch but where editing is limited to the 5' terminal of editing domains in later separated lineages (
30
). A notion with respect to the origin of the RNA editing process in plant
mitochondria has been proposed. That is the evolutionary appearance of the
event at the level of the first land plants (
7
). The finding of the event in chloroplast of
A.formosae
supports this notion. Since RNA editing has been found in the hornwort, more
studies of this event in chloroplasts of primitive land plants, including green
algae, are important to elucidate the origin of RNA editing.
As a result of RNA editing, 19 amino acids deduced from the mRNA sequence
differed from those predicted from the DNA sequence (Table
1
). This difference, corresponding to 4.0% of the peptide, is comparable with
that between the sequence of this hornwort species and that of the fern,
Angiopteris lygodiifolia
(X58429). We considered that this difference would alter the phylogenetic
position of
A.formosae
on the tree based on amino acid sequences which were deduced from genomic and
cDNA sequences of
rbcL.
Therefore, a tree was constructed (Fig.
4
). However,
A.formosae
based on genomic DNA sequences is a sister to that based on cDNA. The
phylogenetic tree shows that
A.formosae
locates in between
Coleochaete
and pteridophytes, which is in agreement with one of the classical views (
31
) and the molecular data (
3
) that the hornworts are evolutionally distinct from the liverworts and the
mosses. The origin of vascular plants is not clear, however our results suggest
that one of possible origins is hornworts. The fact that editing was found in
A.formosae
but not observed in
Marchantia polymorpha
indicates that the event arose in chloroplasts of the first land plants and
disappeared from liverworts, alternatively it arose in the common ancestor of
hornworts and vascular plants. The DNA sequence of the hornwort
Megaceros enigmaticus
(L13481) contains two nonsense codons within the putative coding region and one of them is located
exactly in the same position as in that of
A.formosae,
suggesting RNA editing in this plant as mentioned by Manhalt (
29
).
M.enigmaticus
should come close to
A.formosae
when the mRNA sequence is analyzed.
We are grateful to Masahiro Sugiura for helpful suggestions. We also thank
Kyouji Yamada and Tatsuya Wakasugi for unpublished data, Kiyomi Wada for
helpful advice and for providing
A.formosae
, and Tatsuya Ishii for sequencing of genomic DNA. This work was supported in
part by a grant-in-aid for scientific research (No. 07640926) from the Ministry of
Education of Japan.
REFERENCES
Return