Download Is Interlineage Recombination Responsible for

Is Interlineage Recombination Responsible for Low Divergence of
Mitochondrial nad3 Genes in Mytilus galloprovincialis?
Artur Burzyński and Beata Śmietanka
Department of Genetics and Marine Biotechnology, Polish Academy of Sciences, Institute of Oceanology, Sopot, Poland
The existence of mtDNA recombination in animals has been confirmed by several case studies. Still, for Mytilus mussels
possessing two divergent mitochondrial genomes (M and F), which can recombine, no recombination between coding
sequences of highly diverged M and F genomes has been shown. Based on the full sequences of both genomes, it has
been suggested that particularly low divergence observed within the mitochondrial nad3 gene of the Mytilus
galloprovincialis mussel may be caused by its exceptionally low evolutionary rate. Here, we contribute a new pair of
mitochondrial genomes typical for M. galloprovincialis and show that this low divergence is not a sign of evolutionary
conservation but is rather caused by the acquisition of an F-related sequence by the published M genome of
M. galloprovincialis. The most likely scenario for this apparent mtDNA-coding region recombination case is an assembly
artifact.
Introduction
Mytilus mussels have an unusual system of mitochondrial inheritance: In addition to the classic, maternal route
(F genome), the males pass another mtDNA (M genome) to
sons (Skibinski et al. 1994; Zouros et al. 1994). Because
these two genomes can be quite divergent, Mytilus mussels
became the mitochondrial recombination hunting ground.
Indeed, several different variants of the fragment of the
cox3 gene have been cloned from a gonadal tissue of an
atypical Mytilus galloprovincialis male bearing two similar
(;3% divergence) M and F genomes. Some of them
showed a recombination signature (Ladoukakis and Zouros
2001). Haplotypes with mosaic M–F control region sequences are widely known in Mytilus (Burzyński et al. 2003,
2006; Rawson 2005; Venetis et al. 2007; Filipowicz
et al. 2008). Only analyses of entire mtDNA genomes
may answer the question if other parts of the molecule also
recombine in Mytilus.
Mizi et al. (2005) described full sequences of both M
and F genomes of M. galloprovincialis. One of the most
interesting findings was that the nad3 gene showed the lowest divergence among all protein-coding genes (0.127). The
explanation given for this fact was ‘‘the low rate of evolution of nad3’’ in Mytilus (p 962). The hypothesis that nad3
may contain the origin of replication of the lagging strand
(OL) was given as a possible cause for this conservation.
Several mitochondrial (mt) genomes of Mytilus mussels have been published, including the sequence of F
(Boore et al. 2004) and M (Breton et al. 2006) genomes
of Mytilus edulis. To test the ‘‘conserved nad3’’ hypothesis,
the alignment of four genomes—both haplotypes of
M. galloprovincialis and both haplotypes of M. edulis—has
been created. The sliding window plots of M–F divergence
are presented in figure 1A. Contrary to the description given
by Mizi et al. (2005), it is not the ‘‘nad3 plus the adjacent
100 bp of UR4’’ (p 937) that is responsible for the lowest
divergence region in this plot. This region is located at the
very beginning of nad3. Importantly, the relevant comparison of M. edulis genomes (fig. 1A, gray line) does not show
Key words: DUI, mtDNA recombination, assembly errors,
mitogenomics.
E-mail: [email protected].
Mol. Biol. Evol. 26(7):1441–1445. 2009
doi:10.1093/molbev/msp085
Advance Access publication April 22, 2009
Ó The Author 2009. Published by Oxford University Press on behalf of
the Society for Molecular Biology and Evolution. All rights reserved.
For permissions, please e-mail: [email protected]
this. This is the only region of the genome for which there is
such a difference in M–F divergences between M. edulis and
M. galloprovincialis. It argues against the ‘‘conserved nad3’’
hypothesis predicting that the nad3 gene should exhibit low
divergence also in M. edulis, whereas the overall M–F distance of nad3 is 0.27 in M. edulis (for the most conservative
cox1, the distance is 0.19). The most likely position of OL
(Rodakis et al. 2007) is outside the region (fig. 1, arrow)
so its influence on the divergence anomaly is unlikely.
A closer look at the region in question (fig. 1B) shows
that the distance approaches zero within the anomaly. Such
a pattern could be the result of an M–F recombination event
in which the M genome acquired a fragment of the F
genome. To test this hypothesis, a set of recombination detection algorithms has been applied to the four-genome
alignment. All programs conclusively confirmed recombination (table 1). The positions of the most probable breakpoints are marked with asterisks in figure 1C. Taking these
data at the face value, it must be concluded that the fragment
of the M genome, approximately 150 bp long, is derived
from the F genome in M. galloprovincialis. This would
be the first case of a mosaic mitochondrial protein-coding
gene resulting from homologous recombination between
highly divergent (.20%) genomes in animals.
To check whether this is a typical feature of
M. galloprovincialis, we sequenced another pair of representative M and F genomes from this species. The descriptive statistics for the newly sequenced genomes are
presented in table 2. They differ from the comparable data
presented by Mizi et al. (2005) in five genes: cox1, nad1,
cob, nad6, and nad3. The differences in length of nad1 and
cox1 are merely the result of different annotations chosen
for these genes. Should the same annotation convention
have been used for the sequences of Mizi et al. (2005),
these two differences would disappear. The differences
in length of cob as well as both length and Ka/Ks differences in nad6 result from the fact that the 3# parts of these
two genes differ in their reading frames. These apparent
frameshifts have been discussed in detail by Zbawicka
et al. (2007). Most importantly, the new data do not support particularly low divergence of the nad3: The distance
is 0.215, only slightly lower than in M. edulis. The nad3
region was examined in detail, following the approach described above for the original data. Figure 2 shows that
newly sequenced genomes do not reveal any nad3 divergence anomaly. Recombination detection programs also
1442 Burzyński and Śmietanka
FIG. 1.—Comparison of M and F genomes of Mytilus edulis (gray) and Mytilus galloprovincialis (black). Arrow points at the location of OL
(Rodakis et al. 2007). (A) M–F nucleotide diversity has been calculated in a 150-bp window moving along the alignment in 50-bp steps and plotted at
each window midpoint. The positions of all protein-coding genes and two rRNA genes are shown above the plot. (B) The same as in (A) except the plot
is limited to the region from nad2 to cox1, with a smaller window of 75 bp and in 1-bp steps. The positions of two relevant cloned fragments are shown
by lines above the protein-coding genes. The position of the single direct PCR product is shown below the protein-coding genes. (C) The details of the
alignment over the region from 8374 to 8702 in (A) and (B). Sequences are cited by their GenBank accession numbers. AY484747 and AY497292
represent F haplotypes AY823623 and AY363687 represent M haplotypes. The nucleotides in bold support the F origin of the first M sequence. The
underlined nucleotides support a regular M–F dichotomy. Asterisks over the alignment mark the most probable recombination breakpoints.
The positions of genes are indicated under the alignment, the recognition sites for restriction enzymes used for cloning are boxed. The positions and the
sequences of two primers: ND3-f1 and ND3-r used to obtain the PCR product linking cloned fragments are shown under the alignment.
do not indicate mosaic fragments in the newly sequenced
mt genomes—neither in nad3 nor in any other coding
sequence. The recombination signal in the original M
sequence of Mizi et al. (2005) became even stronger when
all six sequences were taken into account (table 1). This was
a result of the closer relationship between the newly
Table 1
Statistical Support (Bonferroni—Corrected Average
P Value) for the Recombination in the nad3 of Mytilus
galloprovincialis Based on Two Alignments Including Four or
Six Genomes
Program
RDP
GENECONV
BootScan
MaxChi
Chimaera
SiScan
LARD
3Seq
Six
1.37
1.44
7.34
9.13
8.17
2.00
3.73
2.00
10
10
10
10
10
10
10
10
Four
30
24
31
09
09
12
29
14
6.17
4.26
6.49
1.89
6.38
3.93
5.77
1.95
10
10
10
10
10
10
10
10
21
18
21
08
09
16
29
14
sequenced M genome and the nonrecombinant majority
of the original sequence.
Recombinant sequences of supposedly mitochondrial
origin are not rare in GenBank. Still, many of them may
rather represent all kinds of artifacts. According to Piganeau
et al. (2004), much care should be taken to avoid errors, and
even greater care should be exercised before announcing
recombination. Following this recommendation, it is reasonable to ask: Is it possible that the case discussed here
does not a represent genuine recombinant mitochondrial sequence? Mizi et al. (2005) used a three-step procedure to
obtain the sequence. It involved two long-range polymerase
chain reactions (PCR), digestion of obtained products with
restriction enzymes and cloning of the resulting fragments
followed by sequencing of the obtained clones. The final
assembly was facilitated by sequences obtained from additional PCR products spanning clone junctions. The span of
obtained clones is presented in figure 1B, and the relevant
restriction sites are shown in figure 1C. Apparently, the cut
sites are flanked by potential recombination breakpoints.
Therefore, the whole ‘‘recombined’’ fragment must have
been obtained in the PCR with primers ND3-f1 and
nad3 Recombination in Mytilus
1443
Table 2
Length, Base Composition, and Sequence Divergence of Newly Sequenced Genomes
mtDNA
Gene/Region
Noncoding
VD1
CD
VD2
Whole CR
tRNA
All tRNA
rRNA
rrnaL
rrnaS
All rRNA
Protein
atp6
cox1
cox2
cox3
cob
nad1
nad2
nad3
nad4
nad4L
nad5
nad6
All CDS
Complete
Base Composition (%)
Divergence (SE)
Type
Length
T
C
A
G
K
Ks
Ka
F
M
F
M
F
M
F
M
690
491
355
341
149
83
1,194
915
28.3
28.9
33.0
33.4
14.8
10.8
28.0
29.0
15.4
16.5
14.1
13.8
12.1
14.5
14.6
15.3
26.4
35.4
36.6
36.4
47.0
43.4
32.0
36.5
30.0
19.1
16.3
16.4
26.2
31.3
25.5
19.2
0.446 (0.042)
NA
NA
0.046 (0.011)
NA
NA
0.106 (0.041)
NA
NA
0.232 (0.018)
NA
NA
F
M
1,517
1,518
34.2
34.5
13.8
14.0
32.6
32.9
19.4
18.6
0.120 (0.009)
NA
NA
F
M
F
M
F
M
1,244
1,243
947
949
2,191
2,192
33.8
31.5
31.4
30.1
32.7
30.9
13.0
14.9
13.4
14.0
13.2
14.5
32.3
33.6
32.5
34.8
32.4
34.1
20.9
20.0
22.7
21.1
21.7
20.4
0.166 (0.013)
NA
NA
0.134 (0.012)
NA
NA
0.151 (0.009)
NA
NA
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
717
717
1,656
1,740
729
729
936
936
1,308
1,311
1,079
1,067
948
948
351
351
1,308
1,308
282
282
1,707
1,689
465
465
11,486
11,543
16,780
16,639
38.4
37.4
34.3
33.2
33.6
34.6
35.0
34.9
35.7
36.2
36.6
37.3
31.9
35.7
38.5
38.8
33.3
35.1
39.1
38.4
35.6
36.2
40.0
39.4
35.3
35.8
34.2
34.5
14.8
14.1
15.2
16.6
15.6
16.1
16.4
17.0
17.4
18.3
15.6
14.1
13.9
12.0
12.4
12.9
15.5
13.5
12.9
14.3
14.6
14.2
8.4
10.0
15.0
14.9
14.5
14.6
22.1
24.8
26.7
27.4
27.0
27.4
23.4
24.1
24.4
24.9
22.8
23.3
27.1
27.9
21.6
20.7
25.7
26.9
24.7
25.1
26.5
26.9
24.9
27.7
25.1
26.0
27.5
28.6
24.6
23.7
23.8
22.8
23.8
21.9
25.2
23.9
22.5
20.6
25.0
25.4
27.2
24.4
27.6
27.6
25.5
24.5
23.3
22.2
23.4
22.7
26.6
22.9
24.6
23.3
23.8
22.2
0.263 (0.023)
0.884 (0.115)
0.081 (0.014)
0.200 (0.012)
0.835 (0.071)
0.029 (0.006)
0.237 (0.023)
0.924 (0.135)
0.061 (0.012)
0.237 (0.017)
0.831 (0.090)
0.070 (0.013)
0.239 (0.016)
0.757 (0.073)
0.086 (0.011)
0.302 (0.022)
0.895 (0.095)
0.123 (0.015)
0.350 (0.024)
0.996 (0.116)
0.162 (0.019)
0.215 (0.028)
0.707 (0.128)
0.061 (0.017)
0.295 (0.019)
0.965 (0.091)
0.099 (0.012)
0.318 (0.044)
1.083 (0.255)
0.107 (0.025)
0.273 (0.016)
0.759 (0.062)
0.114 (0.012)
0.311 (0.033)
0.776 (0.120)
0.147 (0.024)
0.264 (0.006)
0.852 (0.031)
0.091 (0.004)
0.224 (0.004)
NA
NA
ND3-r (fig. 1C). Notably, both primers show a bias toward
F-like sequences, particularly in the 3# part. Mizi et al.
(2005) used sperm as the source of DNA in the hope to
avoid contamination with the F genome. However, it is difficult, if not impossible, to obtain DNA in which sensitive
PCR of defined specificity could not detect both genomes—even if one of them is present only in a minuscule
quantity. Because the primers used for PCR show bias toward the F genomes, the advantage of using sperm DNA in
an effort to avoid F contamination could have been easily
offset by PCR specificity. Therefore, it is possible that PCR
amplified the fragment of the F genome and such sequences
were then assembled with the sequences from the cloned M
genome fragments leading to the observed mosaic structure.
The GenBank database contains sequences from several
M. galloprovincialis expressed sequence tag libraries (Venier
et al. 2003; Tanguy et al. 2008). It has been checked for the
presence of nad3 sequences returning nine reasonably long
sequences, eight of F-type, one of M-type, and no mosaic
sequences. Therefore, even if the presented interpretation is
wrong and this is the true case of mtDNA recombination, it
would then be most likely limited to this single individual.
Methods
The selection of individual genomes for sequencing
was as follows: Two M and three F major phylogenetic
clades have been identified among M. galloprovincialis–
M. edulis mt haplotypes (Śmietanka et al. 2009). Of the
1444 Burzyński and Śmietanka
lated in MEGA4 (Tamura et al. 2007), following the
procedure used by Mizi et al. (2005).
Acknowledgments
This work was partially funded by grant no. N303
418336 from the Polish Ministry of Science to A.B.
Research was done at Polish Academy of Sciences, Institute
of Oceanology, Department of Genetics and Marine
Biotechnology.
Literature Cited
FIG. 2.—Comparison of newly sequenced Mytilus galloprovincialis
M and F genomes over the area shown in figure 1B (gray). The span of
overlapping, sequenced PCR products is shown by lines above the
protein-coding genes. The comparison of genomes sequenced by Mizi
et al. (2005) is shown in black.
two M clades, one was associated with Mediterranean–
Black Sea M. galloprovincialis. One male individual carrying the haplotype representative for this clade has been
chosen for sequencing of the entire M genome (ORI27).
Likewise, a female carrying the haplotype typical for
M. galloprovincialis F clade was chosen for sequencing
of the entire F genome (AZO20). DNA was isolated from
the mantle tissue of ripe animals, as described previously
(Śmietanka et al. 2009). The sequencing strategy involved
two steps: a long-range PCR (LR-PCR) and a set of rePCRs in which the product of the LR-PCR was highly diluted (1:1,000) and used as a substrate. PCR products were
then sequenced directly, as described previously (Zbawicka
et al. 2007). The nearly complete M genome (16,432 bp)
was amplified in one reaction with the following, highly
specific primers: MGL 5#-ACGCTTAGATTCCTTGCCATTGCC-3# and MGLR 5#-GAAACTATCTCCAACTATCTGCGTATATTCCTG-3#. The remaining 264 bp located
between LR-PCR primers was determined by sequencing
the M-specific AB40-AB20 PCR product (Burzyński
et al. 2006; Filipowicz et al. 2008). The F genome was amplified in two overlapping LR-PCRs (10,782 and 7,709 bp,
1,737-bp overlap) with universal primers AB23–AB50R
and AB44–AB33: 5#-AATTGCGCTGTTATCCCTAGAGTAGC-3# (Burzyński et al. 2006; Zbawicka et al.
2007). The small fragment (164 bp) from between LRPCR primers was obtained from PCR product AB40CBM2 (Burzyński et al. 2006). The annotation of mt
sequences followed the procedure described by Zbawicka
et al. (2007). Obtained sequences have been deposited
in GenBank (FJ890849–50). Sliding window plots of
M–F divergences were obtained using DnaSP version
4.50 (Rozas et al. 2003). The following programs were used
for detecting recombination: RDP (Martin and Rybicki
2000), MaxChi (Maynard Smith 1992), LARD (Holmes
et al. 1999), GENECONV (Padidam et al. 1999), SiScan
(Gibbs et al. 2000), Chimaera (Posada and Crandall
2001), Bootscan (Martin et al. 2005), and 3Seq (Boni
et al. 2007) under RDP version 3.34 with default parameters
(Martin et al. 2005). Genetic diversity indices were calcu-
Boni MF, Posada D, Feldman MW. 2007. An exact nonparametric method for inferring mosaic structure in sequence
triplets. Genetics. 176:1035–1047.
Boore JL, Medina M, Rosenberg LA. 2004. Complete sequences
of the highly rearranged molluscan mitochondrial genomes of
the scaphopod Graptacme eborea and the bivalve Mytilus
edulis. Mol Biol Evol. 21:1492–1503.
Breton S, Burger G, Stewart DT, Blier PU. 2006. Comparative
analysis of gender-associated complete mitochondrial genomes in marine mussels (Mytilus spp.). Genetics. 172:
1107–1119.
Burzyński A, Zbawicka M, Skibinski DOF, Wenne R. 2003.
Evidence for recombination of mtDNA in the marine mussel
Mytilus trossulus from the Baltic. Mol Biol Evol. 20:
388–392.
Burzyński A, Zbawicka M, Skibinski DOF, Wenne R. 2006.
Doubly uniparental inheritance is associated with high
polymorphism for rearranged and recombinant control region
haplotypes in Baltic Mytilus trossulus. Genetics. 174:
1081–1094.
Filipowicz M, Burzyński A, Śmietanka B, Wenne R. 2008.
Recombination in mitochondrial DNA of European mussels
Mytilus. J Mol Evol. 67:377–388.
Gibbs MJ, Armstrong JS, Gibbs AJ. 2000. Sister-scanning:
a Monte Carlo procedure for assessing signals in recombinant
sequences. Bioinformatics. 16:573–582.
Holmes E, Worobey M, Rambaut A. 1999. Phylogenetic
evidence for recombination in dengue virus. Mol Biol Evol.
16:405–409.
Ladoukakis ED, Zouros E. 2001. Direct evidence for homologous recombination in mussel (Mytilus galloprovincialis)
mitochondrial DNA. Mol Biol Evol. 18:1168–1175.
Martin DP, Posada D, Crandall KA, Williamson C. 2005. A
modified bootscan algorithm for automated identification of
recombinant sequences and recombination breakpoints. AIDS
Res Hum Retroviruses. 21:98–102.
Martin DP, Rybicki E. 2000. RDP: detection of recombination
amongst aligned sequences. Bioinformatics. 16:562–563.
Martin DP, Williamson C, Posada D. 2005. RDP2: recombination detection and analysis from sequence alignments.
Bioinformatics. 21:260–262.
Maynard Smith J. 1992. Analyzing the mosaic structure of genes.
J Mol Evol. 34:126–129.
Mizi A, Zouros E, Moschonas N, Rodakis GC. 2005. The
complete maternal and paternal mitochondrial genomes of the
Mediterranean mussel Mytilus galloprovincialis: implications
for the doubly uniparental inheritance mode of mtDNA. Mol
Biol Evol. 22:952–967.
Padidam M, Sawyer S, Fauquet CM. 1999. Possible emergence
of new geminiviruses by frequent recombination. Virology.
265:218–225.
nad3 Recombination in Mytilus
Piganeau G, Gardner M, Eyre-Walker A. 2004. A broad survey
of recombination in animal mitochondria. Mol Biol Evol.
21:2319–2325.
Posada D, Crandall KA. 2001. Evaluation of methods for
detecting recombination from DNA sequences: computer simulations. Proc Natl Acad Sci USA. 98:13757–
13762.
Rawson P. 2005. Nonhomologous recombination between the
large unassigned region of the male and female mitochondrial
genomes in the mussel, Mytilus trossulus. J Mol Evol.
61:717–732.
Rodakis GC, Cao L, Mizi A, Kenchington ELR, Zouros E. 2007.
Nucleotide content gradients in maternally and paternally
inherited mitochondrial genomes of the mussel Mytilus. J Mol
Evol. 65:124–136.
Rozas J, Sánchez-DelBarrio JC, Messeguer X, Rozas R. 2003.
DnaSP, DNA polymorphism analyses by the coalescent and
other methods. Bioinformatics. 19:2496–2497.
Skibinski DOF, Gallagher C, Beynon CM. 1994. Mitochondrial
DNA inheritance. Nature. 368:817–818.
Śmietanka B, Burzyński A, Wenne R. 2009. Molecular
population genetics of male and female mitochondrial
genomes in European mussels Mytilus. Mar Biol (Berl).
156:913–925.
1445
Tamura K, Dudley J, Nei M, Kumar S. 2007. MEGA4: molecular
Evolutionary Genetics Analysis (MEGA) software version
4.0. Mol Biol Evol. 24:1596–1599.
Tanguy A, Bierne N, Saavedra C, et al. (25 co-authors). 2008.
Increasing genomic information in bivalves through new EST
collections in four species: development of new genetic
markers for environmental studies and genome evolution.
Gene. 408:27–36.
Venetis C, Theologidis I, Zouros E, Rodakis GC. 2007. A
mitochondrial genome with a reversed transmission route in
the Mediterranean mussel Mytilus galloprovincialis. Gene.
406:79–90.
Venier P, Pallavicini A, De Nardi B, Lanfranchi G. 2003.
Towards a catalogue of genes transcribed in multiple tissues
of Mytilus galloprovincialis. Gene. 314:29–40.
Zbawicka M, Burzyński A, Wenne R. 2007. Complete sequences
of mitochondrial genomes from the Baltic mussel Mytilus
trossulus. Gene. 406:191–198.
Zouros E, Ball AO, Saavedra C, Freeman KR. 1994. Mitochondrial DNA inheritance. Nature. 368:818.
Richard Thomas, Associate Editor
Accepted April 14, 2009