Download Mitochondrial Gene Diversity Associated with the atp9 Stop Codon

Journal of Heredity 2012:103(3):418–425
doi:10.1093/jhered/esr142
Advance Access publication February 15, 2012
Ó The American Genetic Association. 2012. All rights reserved.
For permissions, please email: [email protected].
Mitochondrial Gene Diversity
Associated with the atp9 Stop Codon in
Natural Populations of Wild Carrot
(Daucus carota ssp. carota)
JENNIFER R. MANDEL, EDWARD V. MC ASSEY, KATHERINE M. ROLAND,
AND
DAVID E. MCC AULEY
From the Department of Plant Biology, University of Georgia, Athens, GA (Mandel and McAssey); the School for Science
and Math, Vanderbilt University, Nashville, TN (Roland); and the Department of Biological Sciences, Vanderbilt University,
Nashville, TN 37235 (McCauley).
Address correspondence to Dr Jennifer R. Mandel at the address above, or e-mail: [email protected].
Data deposited at Dryad: http://dx.doi.org/10.5061/dryad.3jt640b7
Abstract
Mitochondrial genomes extracted from the wild populations of Daucus carota have been used as a genetic resource by
breeders of cultivated carrot, yet little is known concerning the extent of their diversity in nature. Of special interest is an
SNP in the putative stop codon of the mitochondrial gene atp9 that has been associated previously with male-sterile and
male-fertile phenotypic variants. In this study, either the sequence or PCR/RFLP genotypes were obtained from the
mitochondrial genes atp1, atp9, and cox1 found in D. carota individuals collected from 24 populations in the eastern United
States. More than half of the 128 individuals surveyed had a CAA or AAA, rather than TAA, genotype at the position
usually thought to function as an atp9 stop codon in this species. We also found no evidence for mitochondrial RNA editing
(Cytosine to Uridine) of the CAA stop codon in either floral or leaf tissue. Evidence for intragenic recombination, as
opposed to the more common intergenic recombination in plant mitochondrial genomes, in our data set is presented. Indel
and SNP variants elsewhere in atp9, and in the other 2 genes surveyed, were nonrandomly associated with the 3 atp9 stop
codon variants, though further analysis suggested that multilocus genotypic diversity had been enhanced by recombination.
Overall the mitochondrial genetic diversity was only modestly structured among populations with an FST of 0.34.
Key words: crop, cultivated, heteroplasmy, Queen Anne’s Lace, recombination, RNA editing
The magnitude of intraspecific sequence diversity of the
mitochondrial genome has not been widely characterized in
natural populations of plants. This may be due to the fact
that several features of plant mitochondrial genomes once
thought to be universal, or nearly universal, would suggest
that such diversity would be very low. First, until recently, it
was thought that the average rate of nucleotide substitution
in plant mitochondrial genomes was considerably lower than
that of plant chloroplast or nuclear genomes, as evidenced
by the high level of sequence conservation observed in
between-species comparisons (Wolfe et al. 1987; Palmer
1992). This high level of conservation often seen in betweenspecies comparisons would not predict much intraspecific
diversity. The second feature of the mitochondrial genome
that would limit diversity is maternal inheritance, which
would reduce the likelihood that recombination between
mitochondrial genomes would create novel genotypes. This
418
is because maternal inheritance of the mitochondrial genome
would limit the degree of within-individual genetic diversity
(heteroplasmy) and thus the opportunity for divergent
genotypes to participate in recombination events.
Recently, exceptions to both of these putatively common
features of the mitochondrial genome have been found. For
example, unusually high mitochondrial gene diversity has
been found in several plant genera, including Silene, Plantago,
and Pelargonium (Cho et al. 2004; Mower et al. 2007)
suggesting that the substitution rates need not be low for all
mitochondrial genes in all plant species. Furthermore,
evidence for occasional biparental inheritance of the mitochondrial genome has been found in Silene vulgaris (Pearl et al.
2009; Bentley et al. 2010). These observations are coincident
with the observations in both S. vulgaris (Houliston and Olson
2006; Barr et al. 2007; McCauley and Ellis 2008) and its
relative S. acalis (Städler and Delph 2002; Touzet and Delph
Mandel et al. Mitochondrial Genetic Diversity in Wild Carrot
2009) of considerable intraspecific sequence diversity in
several mitochondrial genes and an evidence for intragenic
and intergenic recombination. However, despite these recent
findings the extent of intraspecific mitochondrial gene
diversity remains unexplored in most plant taxa, as does the
functional significance of any such diversity.
One reason that characterizing plant mitochondrial
genetic diversity may be important is that mitochondrial
gene variants could be useful markers in breeding efforts,
especially when identifying breeding lines harboring cytoplasmic male sterility (CMS), a phenomenon that appears to
involve the mitochondrial genome in almost all documented
cases (Schnable and Wise 1998; Hanson and Bentolila 2004).
One example of this is that a number of mitochondrial
genetic markers have been developed for use in breeding
programs of the domesticated carrot, Daucus carota ssp. sativa
(Wolyn and Chahal 1998; Szklarczyk et al. 2000; Bach et al.
2002). The wild progenitor of the domesticated carrot is
D. carota ssp. carota, also known as wild carrot or Queen
Anne’s Lace. Wild carrot has a gynodioecious mating system
in which the flowers of individuals may be either hermaphroditic or male sterile (Ronfort et al. 1995). Some
male-sterile phenotypes from both European and North
American wild populations have been incorporated into the
breeding programs for domesticated carrot (Wolyn and
Chahal 1998; Robison and Wolyn 2002), including the
brown anther form (anthers are brown and shriveled) and
the petaloid form (anthers develop as petals). These
phenotypes are thought to be cases of CMS traced to the
mitochondrial genome (Ronfort et al. 1995). Consequently,
there has been an effort to identify mitochondrial genetic
markers that distinguish male-sterile and fertile cultivars of
D. carota, including those incorporated from wild populations (e.g., Pingitore et al. 1989; Ronfort et al. 1995;
Nakajima et al. 1999; Szklarczyk et al. 2000; Bach et al. 2002;
Robison and Wolyn 2006). Thus, it is clear that mitochondrial gene diversity exists in natural populations of D. carota
ssp. carota. However, there has been no systematic study of the
level of mitochondrial genetic diversity in natural populations
of D. carota, as would be useful when considering the broader
question of how diverse mitochondrial genes are in natural
populations of plants.
Of particular interest among the suite of mitochondrial
markers developed by carrot breeders is a T/C transition
SNP in the mitochondrial gene atp9 that would seem to
covert a putative stop codon TAA (in D. carota atp9 is 229
nucleotides in length), to a CAA codon that may then code
for Glutamine. The transcript in these CAA individuals
would presumably extend 39 more nucleotides to the next
possible stop codon in the sequence, TGA (Szklarczyk et al.
2000; Bach et al. 2002). Some molecular and genetic
evidence in domesticated carrot supports the hypothesis for
atp9 involvement in the CMS petaloid phenotype; however,
the evidence for this gene as the cause of this phenotype is
not yet conclusive (Szklarczyk et al. 2000). One possibility
with regard to the Daucus atp9 gene is that the CAA triplet
could be converted to a TAA via RNA editing such that its
stop function is retained, as suggested by Szklarczyk et al.
(2000). A form of RNA editing, which occurs with
a relatively high frequency in plant mitochondrial genomes,
involves a posttranscriptional process whereby Cytosine is
changed to Uridine via deamination (Araya et al. 1998).
Studies of both rice and tobacco have demonstrated that
unedited versions of the mitochondrial transcript atp9 show
reduced or no male fertility where edited transcripts show
normal male floral structure development. It should be
noted that D. carota genomes that differ with regard to the
T/C SNP in atp9 also differ consistently with regard to
a number of other aspects of the sequence of atp9, as well as
in other mitochondrial genes (Bach et al. 2002).
Interestingly, although the majority of the coding sequence
for atp9 is conserved across angiosperms, neither the stop
codon triplet sequence nor the stop codon position is as
highly conserved. For example, the length of the Arabidopsis
atp9 coding sequence is 258 nucleotides with a TGA stop
codon (Giege and Brennicke 1999), whereas in Vitis, it is 225
nucleotides in length with TGA (Goremykin et al. 2009), and
in maize, the stop codon is TAA at 225 nucleotides
(Gallagher et al. 2002). Even within the Asteraceae family,
Helianthus and Carthamus differ in having TGA at 251 and
TAA at 261 nucleotides, respectively (Recipon 1990; Kalinati
et al. 2008). Thus, if the 39 bp extension hypothesized for
the CAA genotype D. carota individuals is correct, an atp9
transcript size polymorphism occurs within wild carrot that
rivals the largest difference in atp9 transcript size seen across
other angiosperm species.
Studies of atp9 in wild carrot have focused largely on
the relatively few mitochondrial genomes that have been
incorporated into domesticated carrot cultivars. Thus, the
level of TAA/CAA polymorphism in wild populations of
D. carota ssp. carota is not known, nor how consistent the
association between the alternate forms of this SNP and the
other mitochondrial markers that distinguish the previously
studied fertile and petaloid breeding lines is in natural
populations. Here, we report the mitochondrial genotypes
of a number of D. carota ssp. carota individuals collected
from 24 wild populations in the eastern United States. Using
both DNA sequencing and PCR/RFLP methods, we report
on the relative frequency and distribution across populations
of the TAA/CAA atp9 polymorphism, as well as the
association between this SNP and the additional variation in
mitochondrial genes atp9, atp1, and cox1. Of particular note
is the discovery of a relatively common third atp9 stop
codon variant (AAA) that would not be subject to the usual
mitochondrial RNA editing. We also genotyped the RNA/
cDNA of CAA individuals in order to assay for any evidence
of RNA editing at the stop codon polymorphism.
In addition to providing frequency and distribution
information about the stop codon polymorphism and its
association with other genetic markers in the natural populations of D. carota, the results presented here can be used
to address another aspect of the evolution of the mitochondrial genome. First, in keeping with the approach used
in recent inferences about the recombination in the mitochondrial genome of the gynodioecious plants S. acalis and
S. vulgaris, we utilize multilocus sequence data collected here
419
Journal of Heredity 2012:103(3)
Table 1 Location of Daucus carota populations sampled for this
study, along with the number of individuals sampled from each
Population
State
County
Sample size
wsh
bet
bp
br
chat
church
hom
knx
phip
rad
shlby
shute
nel
sa
ox
wil
sm
ocean
beav
bwl
cream
gg
grav
ut
GA
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
TN
VA
VA
OH
OH
OH
NJ
NY
NY
NY
NY
NY
NY
Clarke
Davidson
Cumberland
Davidson
Hamilton
Davidson
Cumberland
Knox
Cumberland
Davidson
Davidson
Sumner
Nelson
Roanoke
Lake
Cuyahoga
Cuyahoga
Cape May
Delaware
Delaware
Delaware
Delaware
Schoharie
Delaware
5
6
2
3
3
7
5
4
8
3
8
3
7
2
9
2
6
4
7
7
6
3
10
8
to test the hypothesis that recombination in the mitochondrial genome of D. carota can enhance genetic diversity.
Evidence for recombination would be of general interest for
the study of plant mitochondrial genome evolution and also
help to explain the consistency (or lack of such) in the
association between mitochondrial genetic markers and the
male-sterile phenotypes.
Materials and Methods
Daucus carota ssp. carota is abundant along roadsides,
abandoned fields, and pastures across much of North
America. For this study, the leaf material or, in a few cases,
the seeds were collected from 128 wild D. carota individuals
from 24 sites across eastern United States (Table 1). All sites
were .1 km from one another. Individuals within sites were
collected ,50 m from one another. Genomic DNA was
extracted from leaf tissue using an Applied Biosystems 6100
Nucleic Acid PrepStation and the associated protocols
(Foster City, CA).
Sequencing Genotyping Strategy
We first genotyped a subset of the total sample by randomly
choosing 2–3 D. carota individuals per population and
sequencing 3 of their mitochondrial genes: atp1, atp9, and
cox1. Variable flanking regions downstream of the coding
region in atp9 and upstream of orf in atp1 were also
sequenced for these individuals. Primers for the atp9
sequences were obtained from Bach et al. (2002), who
420
developed these markers to differentiate the mitochondrial
genomes of fertile and male-sterile (petaloid) forms of
D. carota. For atp1, we used 2 sets of primers to obtain
sequence information for the 5# flanking region and the orf
of the gene (from Bach et al. (2002) and Bowe et al. (2000),
respectively). Finally, the primers used to amplify the orf of
the cox1 gene were also obtained from Bowe et al. (2000).
In the flanking regions of atp9 and atp1, Bach et al.
(2002) demonstrated that fertile individuals contained direct
repeats of 42 or 71 bp, respectively, which were absent in
the petaloid form. In their study, these markers differentiated the fertile (N cytoplasm) and male sterile (Sp
cytoplasm) in all lines characterized for atp1 and atp9.
PCR amplifications were conducted in 50 lL reactions that
contained ;5 ng of the whole genomic DNA, 2 units of Taq
polymerase, 125 nM of each primer, 72 mM tricine, 120 mM
KCl, and 4.8 mM MgCl2. The PCR conditions were as follows:
3 min at 95 °C; 10 cycles of 30 s at 94 °C, 30 s at 65 °C, and
45 s at 72 °C, annealing temperature decreasing to 55 °C by
1 °C per cycle, followed by 30 cycles of 30 s at 94 °C, 30 s at
55 °C, 45 s at 72 °C, followed by 20 min at 72 °C.
PCR product cleanup was performed by the Vanderbilt
University or the University of Arizona Sequencing Facilities
using the 96-well Millipore PCR purification blocks on a
Biomek FX robot. Sequencing was performed using the
BigDye Terminator chemistry (Applied Biosystems) and
electrophoresed on ABI’s 3730xl DNA Analyzer (Applied
Biosystems). Sequences were viewed using the Sequencher
Software (Gene Codes, Ann Arbor, MI), and sequences
were aligned, with the aid of published D. carota sequences
(Szklarczyk et al. 2000; Bach et al. 2002), using both the
Sequencher and the ClustalX (Larkin et al. 2007). Representative sequences can be found in GenBank: accession
numbers JQ246964-JQ246994. Estimates of nucleotide
diversity (calculated as p on a per site basis) were made
using the software program DnaSp v5 (Librado and Rozas
2009). We also generated a haplotype network by defining
haplotypes and calculating the minimum-spanning tree
among haplotypes in Arlequin v 3.5 (Excoffier and Lischer
2010). The distance matrix output from Arlequin was used
in Hapstar (Teacher and Griffiths 2010) to draw the
minimum spanning tree. The tree was then modified using
Microsoft (Redmond, Washington) drawing tools to color
code haplotypes based upon their atp9 stop codon polymorphism (blue 5 TAA, red 5 CAA, and green 5 AAA).
We utilized several published methods to test for recombination in our data set. The first method was a parsimonybased approach for detecting recombination implemented in
the program RECOMP (Ruths and Nakhleh 2006). This
method was developed for evaluating the interspecific
recombination in bacteria, and thus utilizing this method in
our study is a new application of this approach. The basic
principle for this test is that when recombination occurs,
different regions of a sequence alignment will demonstrate
different underlying trees. The atp1, atp9, and cox1 genes
were concatenated for the test for recombination. We
utilized a sliding window of 300 bp, a step size of 100 bp,
and the parsimony score difference function (a robust
Mandel et al. Mitochondrial Genetic Diversity in Wild Carrot
measure that performs well despite the mutational hotspots)
as recommended by the authors.
To further test for recombination in our data set, we
used 3 additional programs (SiScan, 3Seq, and MaxChi) as
implemented in the software package RDP3 (Martin et al.
2010). The SiScan, or ‘‘Sister-Scanning,’’ method utilizes a
Monte Carlo randomization to examine the putative recombination in all sets of triplets and compares with a fourth
sequence that is either randomly generated or chosen from
the set of sequences presented. We ran this program with
settings of both a randomly generated sequence and an
option of choosing a sequence from the set. The window size
was set to 300 bp with the steps set at 100 bp each. For more
details, see Gibbs et al. (2000). The 3Seq method also scans
triplets, where each sequence in a triplet is in turn queried to
determine if it could potentially be the recombinant of the
other 2 sequences in the triplet (Boni et al. 2007). Finally, the
MaxChi test, developed by Smith (1992) was used to examine
sequence pairs along a variable sliding window. In this
method, MaxChi examines sequence pairs and attempts to
identify the recombination breakpoints in an alignment by
searching for significant differences in the proportions of
variable and nonvariable polymorphic positions in adjacent
regions of sequence. The fraction of variable sites per
window was set to 0.1.
In all tests implemented in RDP3, a step-down
Bonferroni correction was utilized. Following the authors’
recommendation, redundant and/or highly similar genotypes were removed from the data set, and the auto mask
function was used for the exploratory search for recombination signals. A preliminary scan for recombination
was performed, and then putative recombination events were
inspected manually by examining MaxChi plots, recombination breakpoint matrices, and neighbor joining trees, all
following the authors’ recommended steps for executing the
programs in RDP3 and evaluating the results.
PCR/RFLP Genotyping Strategy
In order to increase our sample size with regard to the atp9
stop-codon position and the atp1 and atp9 repeat markers,
we conducted a second round of genotyping using the same
PCR amplification methods but followed by restriction
digests and fragment size analysis of the resulting product,
rather than sequencing. These PCR/RFLP methods were
first used to regenotype 30 individuals whose genotype was
already known from the sequence data, in order to verify
their accuracy, and then applied to individuals of unknown
genotype. The PCR/RFLP protocols were as follows.
In addition to the atp9 stop codon, SNP of C or T (CAA
or TAA associated with the petaloid and fertile floral
morphotypes, respectively), identified by Szklarczyk et al.
(2000) and Bach et al. (2002), we identified individuals from
our sequencing panel with AAA at this codon position. We,
therefore, designed a restriction digest assay targeting this
polymorphism in our sample as follows. The restriction enzyme DraI cuts TAA individuals, ApoI cuts CAA individuals,
and the AAA individuals will remain uncut by either enzyme.
We assayed the 42-bp direct repeat in atp9 by PCR
amplification and then the digestion of the product with the
restriction endonuclease MseI (New England BioLabs, MA)
following the manufacturer’s suggested protocol. MseI
recognizes a restriction site within atp9 that is common to
all individuals. One of the resulting fragments varies in size
according to the presence or the absence of the direct repeat
and is more reliably scored on a 2% agarose gel than is the
presence or the absence of the repeat when considering the
migration patterns of the larger uncut amplicons. We
examined the distribution of atp9 genotypes (on the full data
set of 128 individuals) among the 24 populations, by
calculating Wright’s FST and 95% confidence limits (jackknifing across populations), using the method for haploid
data of Weir (1996). The presence of the 71-bp direct repeat
at the atp1 locus was also assayed by visualization of PCR
product on 1% agarose gels where the presence or the absence of the 71-bp direct repeat was easily scored. All
restriction digests were performed in 10-ll total volume
digestions using 5-ll PCR product.
RNA Extraction, cDNA Synthesis, and Assay for
RNA Editing
If RNA editing occurred at the stop-codon position,
individuals who were CAA at the genomic DNA level
should have been TAA at the cDNA level. In order to assay
this, the total RNA was extracted from the leaf tissue of
16 individuals (5 TAA, 5 CAA, and 6 AAA plants) and the
floral tissue from the one representative in each of the 3
stop codon types using the RNeasy Kit from Qiagen
(Valencia, CA) following the manufacturer’s suggested
protocol. cDNA was synthesized from the RNA extractions
with the Qiagen QuantiTech Reverse Transcription protocol following the manufacturer’s suggested protocol. We
assayed for editing in the non-CAA individuals to serve as
a control (i.e., AAA should not be edited into a TAA).
cDNA was amplified via PCR using the atp9 specific
primers, and the restriction digests were performed as
before, to assay the identity of the atp9 stop codon position
(TAA, CAA, or AAA). We then sequenced the cDNA PCR
products from a sample of these individuals using the same
sequencing protocol as described above and examined the
sequences for any additional occurrences of editing by
comparing them to the genomic DNA sequence for these
individuals.
Results
Atp9 Stop Codon Polymorphism
A total of 128 individuals from 24 populations were
characterized for the atp9 stop codon and the atp9 and atp1
repeat polymorphisms. Of these, 42 genotypes were
extracted from the atp9, atp1, and cox1 sequence data and
86 obtained using the PCR/RFLP method. Among the 128
individuals, 62 (48%) were the atp9 TAA codon genotype
thought to function as a stop and 35 (27%) were the CAA
421
Journal of Heredity 2012:103(3)
Table 2 Seven observed multilocus mitochondrial genotypes
found in a sample of 128 Daucus carota individuals based on the
sequence of the atp9 stop codon, the presence/the absence
(Y/N) of a 42 bp repeat associated with atp9 and a 71 bp repeat
associated with atp1
Atp9 codon
TAA
TAA
TAA
CAA
CAA
AAA
AAA
5 Other genotypes
Total
Atp1
repeat
Atp9
repeat
Observed
(expected)
Y
Y
N
N
Y
N
Y
Y
N
N
N
N
N
N
47 (13.0)
10 (22.1)
5 (16.6)
28 (9.4)
7 (12.4)
22 (8.7)
9 (11.4)
0 (34.3)
128
The number expected, given a random association of alleles across the loci,
is listed in parentheses. Note that in 14 of the TAA individuals carrying the
atp1-associated repeat, a third copy of the repeat was also present.
alternate codon identified earlier (Szklarczyk et al. 2000;
Bach et al. 2002). In addition, 31 (24%) had an AAA triplet
at the atp9 stop codon position, a variant not seen previously.
Overall, the CAA and AAA variants were geographically
widespread and found in 12 and 11 of the 24 populations
sampled, respectively. TAA individuals were found in 19
of the 24 populations surveyed, including 5 populations
containing only TAA individuals. The distribution of atp9
genotypes (128 individuals) among populations yielded
a value of Wright’s FST equal to 0.343 with 95% confidence
limits of 0.181 and 0.505, as calculated for the haploid data
by the method of Weir (1996).
Seven unique multilocus genotypes (considering all 3 first
position nucleotides, ‘‘T,’’ ‘‘C,’’ and ‘‘A’’) were observed
among these 128 individuals, as listed in Table 2. Note that
the 12 multilocus combinations are possible given the occurrence of 3 alleles at the stop codon and the presence or the
absence of the direct repeat at both atp1 and atp9. Inspection
of Table 2 reveals that the observed genotypic proportions
deviate strongly and significantly from the expectations based
on the random association of alleles across the loci (Gind. 5
185.1, degrees of freedom 5 7, P , 0.001; the 5 genotypes
not observed were pooled as one class). The TAA and CAA
stop codons are strongly associated with the specific atp1- and
atp9-associated repeat presence (Y)/absence (N) genotypes
reported previously by Bach et al. (2002) (YY and NN,
respectively). The AAA codon is strongly associated with the
NN motif. However, the associations are not universal. For
example, 15 TAA individuals are not associated with the
presence of the atp9 repeat, 7 CAA individuals are associated
with the atp1 repeat, and 9 AAA individuals are associated
with the atp1 repeat. In fact, when only the ‘‘T’’ and ‘‘C’’ stop
codon alleles are considered along with the presence or the
absence of the atp1 repeat, all 4 possible combinations were
observed (TY, TN, CY, and CN). The same is true when the
‘‘T’’ and ‘‘A’’ alleles are considered with the atp1 repeat
variants. This would be considered evidence for recombination under the ‘‘4 gamete rule’’ of Hudson and Kaplan (1985).
422
Figure 1. Minimum-spanning haplotype network representing
20 unique haplotypes based upon the sequence data from
42 individuals. Names and numbers of individuals are given in
Supplementary Table 2. Colors correspond to the atp9 stop codon
polymorphism: blue 5 TAA, red 5 CAA, and green 5 AAA.
Black tick marks represent unsampled haplotypes and the size of
the circle represents the number of that haplotype sampled.
Additional Sequence Diversity
The 42 atp1 and atp9 sequences, as well as the associated
cox1 sequences, were inspected for additional variation and
the association of that variation with the atp9 stop codon
type. All variable sites for the 42 concatenated atp1, atp9, and
cox1 sequences are presented in Supplementary Table 1. We
observed numerous SNPs and indels (insertion/deletion)
within the atp9 gene, which were not previously characterized in studies of cultivated carrot (Szklarczyk et al. 2000;
Bach et al. 2002). Two SNPs occurred in the coding region
just before the normal stop codon position at 229 bp. In the
atp1 gene, we observed 3 previously uncharacterized SNPs
downstream from the direct repeat, one being in the coding
region. Finally, we identified 8 SNPs within the orf of the
cox1 gene. At the cox1 gene, all TAA individuals were
identical in sequence. Nucleotide diversity measurements (p)
were 0.01404 for atp9, 0.00069 for atp1, and 0.00264 for
cox1. The haplotype network, color coded by the atp9 stop
codon, revealed that TAA individuals all grouped together
and suggests a possible scenario of multiple origins of the
AAA haplotype (Figure 1). Names and numbers of the
Mandel et al. Mitochondrial Genetic Diversity in Wild Carrot
Figure 2. Results of the parsimony-based method for detection of recombination along a sliding window of the sequence
alignment as implemented in the RECOMP software program.
individuals that are demonstrated in the figure can be found
in Supplementary Table 2.
Evidence for Recombination in the Sequence Data Set
The RECOMP program identified 2 putative recombination
breakpoints in our data set, one at approximately 1600 and
another around 2200 as evidenced by the peaks at these
points for the parsimony score difference (see Figure 2).
The recombination analyses SiScan, 3Seq, and MaxChi
implemented in RDP3 also detected significant evidence for
recombination within our data set (Table 3). The SiScan and
3Seq methods identified one unique recombination event,
whereas the MaxChi test also identified this event and
demonstrated significant evidence for an additional unique
recombination event. Putative breakpoints identified by the
RDP3-implemented analyses were in agreement with those
identified from the RECOMP analysis. The P values were
obtained utilizing the step-down Bonferroni correction.
Assay for RNA Editing
We found no instances of RNA editing in the atp9 stop codon
position in leaf or floral tissues when cDNA was genotyped
with PCR/restriction enzymes or by sequencing. In other
words, no CAA individuals appeared to change to TAA when
genotyped in cDNA. We also found no instances of editing in
other sites within the atp9 PCR amplicons.
Table 3 Three methods, implemented in the RDP3 program,
which detected significant evidence for the recombination in the
sequence dataset
Method
SiScan
3Seq
MaxChi
Number unique
events
1
1
2
P value
5.30 10
2.99 10
4.67 10
P value
6
3
3
—
—
7.52 10
3
Significance values are based upon the step-down Bonferroni correction.
For more details, see text.
Discussion
Daucus carota joins the growing list of plant species in which
considerable mitochondrial gene diversity has been found.
Many of the variants previously known to distinguish breeding
lines of domesticated carrots were found to be widespread
in natural populations, as were a number of previously undescribed SNPs. For example, although we found that roughly
half of the individuals sampled in this study contained the
normal atp9 stop codon, TAA, the balance consisted of
individuals carrying the previously identified CAA or a newly
identified codon, AAA. This SNP warrants special consideration because of its potential functional significance.
The atp9 mRNA transcript containing the CAA codon
has the potential to be edited back to the normal stop codon
via mitochondrial RNA editing (commonly C to U edits;
Araya et al. 1998). However, we found no evidence for
RNA editing of CAA, or AAA, individuals for atp9 in leaf or
floral tissue. The AAA codon would likely not be changed
because this type of editing is not known to occur. As noted
by Bach et al. (2002), the next likely atp9 stop codon is
a TGA triplet that occurs 39 bp downstream, potentially
yielding a product that is 13 amino acids larger. Thus, it may
be that the atp9 T/C/A SNP results in D. carota populations
that are highly polymorphic in the atp9 transcript size.
The question arises as to what the functional significance
(if any) of this might be and what evolutionary forces
maintain this polymorphism. In cultivated carrot, the atp9
codon TAA is associated with the fertile flower phenotype,
whereas the CAA is associated with the petaloid phenotype,
and it has been speculated that if atp9 CAA codons are not
edited, the resulting difference in transcript size could influence the flower development (Szklarczyk et al. 2000; Bach
et al. 2002). If this is true, the AAA variant could have the
same consequence as CAA. We encountered no petaloid
flower phenotypes when collecting samples. However, because many of the individuals we sampled were not yet in
full flower and others had gone to seed, we cannot yet make
423
Journal of Heredity 2012:103(3)
a meaningful statement regarding any association between
the atp9 genotype and the floral phenotype in our data set.
However, nuclear genes which restore fertility have been
identified in wild carrot (Wolyn and Chahal 1998), as well as
for cultivated carrot (Morelock 1974; Stein and Nothnagel
1995), and thus, the occurrence of restoration in natural
populations could be quite common.
The value calculated for FST of the atp9 stop codon
polymorphism indicates substantial population structure
but is quite low when compared with the other maternally
inherited markers in the angiosperm species. Petit et al.
(2005) reviewed the levels of genetic differentiation of the
maternally inherited DNA for 124 angiosperm species and
found the mean to be 0.637 ± 0.002 standard error. The
finding of low FST for a mitochondrial marker in our study
suggests that the seeds of D. carota may move much more
efficiently than the other angiosperm taxa, and/or mitochondrial DNA may occasionally move via pollen, that
is, paternal leakage (see below). Interestingly, inspection of
the haplotype network suggests the possibility of multiple
origins of the atp9 stop codon; however, a data set with
more sequence information is needed to confirm or reject
this possibility.
We also found evidence for genic recombination in our
data set and thus note that the association between the atp9
stop codon type and the presence or the absence of atp1and atp9-associated repeats previously described by Bach
et al. (2002) is not absolute in the natural populations of
D. carota. Recombination in the mitochondrial genome generally occurs across intergenic repeats (e.g., Alverson et al.
2011); however, previous evidence for the intragenic
recombination has been reported in other gynodioecious
species (Städler and Delph 2002; McCauley et al. 2005).
Whereas the majority of TAA individuals do carry both
repeats, others carry just the atp1 repeat or neither repeat.
The evidence for recombination suggests that any statistical
association between the markers (linkage disequilibrium)
would decay over time. This finding certainly has implications for the use of mitochondrial markers in carrot
breeding.
The evidence for recombination also suggests the
possibility of mitochondrial paternal leakage and resulting
heteroplasmy in D. carota. Recombination would be most
likely to create novel genotypic combinations when it occurs
between divergent genomes. The cooccurrence of divergent
mitochondrial genomes within the same cell (heteroplasmy)
can be a result of biparental inheritance of the mitochondrial
genome. Recent studies of S. vulgaris have provided a direct
evidence of paternal leakage of the mitochondrial genome as
well as the mitochondrial heteroplasmy (Welch et al. 2006;
Pearl et al. 2009; Bentley et al. 2010). It has been suggested
that this heteroplasmy then provides conditions favorable
for the generation of the multilocus genotypic diversity used
by McCauley and Ellis (2008) to infer mitochondrial genome
recombination in S. vulgaris.
Additional indirect evidence for paternal leakage comes
from an unpublished study (Roland KM, Skains SN, and
McCauley DE, unpublished data) in which many of the
424
same D. carota individuals characterized here for mtDNA
were also characterized for an indel polymorphism identified
in the chloroplast genome. When the joint cpDNA and atp9
mtDNA genotypes are considered for these individuals,
there is evidence for a reassociation between the 2 genomes
using the same 4 gamete rule applied above to the pairs
of mitochondrial genes (McCauley and Ellis 2008) but
considering the cpDNA/mtDNA genotypic combinations.
Evidence for reassociation between the 2 organellar genomes suggests that one or both must undergo at least an
occasional paternal leakage (Houliston and Olson 2006). It
would be interesting, then, to investigate the possibility of
the mitochondrial paternal leakage and the heteroplasmy in
D. carota directly through controlled crosses of parents with
differing mitochondrial haplotypes.
Supplementary Material
Supplementary material can be found at http://www.jhered.
oxfordjournals.org/.
Funding
This work was supported by the National Science
Foundation award (0621867) to D.E.M.
Acknowledgments
We would like to thank Stephanie Pearl for assistance in collecting samples
and helpful comments. We would also like to thank Michael McKain,
Aaron Richardson, Mark Chapman, and Jonathan Corbi for valuable
comments on an earlier version of this manuscript. We thank Dan Sloan for
beneficial discussion of this project. Finally, we would like to thank
Samantha Skains for lab work and assistance.
References
Alverson AJ, Zhuo S, Rice DW, Sloan DB, Palmer JD. 2011. The
mitochondrial genome of the legume Vigna radiata and the analysis of
recombination across short mitochondrial repeats. PLoS One. 6(1):e16404.
Araya A, Zabaleta E, Blanc B, Begu D, Hernould M, Mouras A, Litvak S.
1998. RNA editing in plant mitochondria, cytoplasmic male sterility and
plant breeding. Electronic J Biotechnol. 1:1–9.
Bach IC, Oleason A, Simon PW. 2002. PCR-based markers to differentiate the
mitochondrial genomes of petaloid and male fertile carrot (Daucus carota L.).
Euphytica. 127:353–365.
Barr CM, Keller SR, Ingvarsson PK, Sloan DB, Taylor DR. 2007. Variation
in mutation rate and polymorphism among mitochondrial genes of Silene
vulgaris. Mol Biol Evol. 24:1783–1791.
Bentley KE, Mandel JR, McCauley DE. 2010. Paternal leakage and
heteroplasmy of mitochondrial genomes in Silene vulgaris: evidence from
experimental crosses. Genetics. 185:961–968.
Boni MF, Posada D, Feldman MW. 2007. An exact nonparametric method
for inferring mosaic structure in sequence triplets. Genetics.
176:1035–1047.
Bowe LM, Coat G, de Pamphilis CW. 2000. Phylogeny of seed plants based
on all three genomic compartments: extant gymnosperms are monophyletic
Mandel et al. Mitochondrial Genetic Diversity in Wild Carrot
and Gnetales closest relatives are conifers. Proc Natl Acad Sci U S A.
97:4092–4097.
Cho Y, Mower JP, Qiu YL, Palmer JD. 2004. Mitochondrial substitution
rates are extraordinarily elevated and variable in a genus of flowering plants.
Proc Natl Acad Sci U S A. 101:17741–17746.
Excoffier L, Lischer HEL. 2010. Arlequin suite ver 3.5: a new series of
programs to perform population genetics analyses under Linux and
Windows. Mol Ecol Resour. 10:564–567.
Gallagher LJ, Betz SK, Chase CD. 2002. Mitochondrial RNA editing
truncates a chimeric open reading frame associated with S male-sterility in
maize. Curr Genet. 42:179–184.
Gibbs MJ, Armstrong JS, Gibbs AJ. 2000. Sister-scanning: a Monte Carlo
procedure for assessing signals in recombinant sequences. Bioinformatics.
16:573–582.
Giege P, Brennicke A. 1999. RNA editing in Arabidopsis mitochondria
effects 441 C to U changes in ORFs. Proc Natl Acad Sci U S A.
96:15324–15329.
Goremykin VV, Salamini F, Velasco R, Viola R. 2009. Mitochondrial DNA
of Vitisvinifera and the issue of rampant horizontal gene transfer. Mol Biol
Evol. 26:99–110.
Hanson MR, Bentolila S. 2004. Interactions of mitochondrial and nuclear
genes that affect male gametophyte development. Plant Cell. 16:S154–S169.
Houliston GJ, Olson MS. 2006. Nonneutral evolution of organelle genes in
Silene vulgaris. Genetics. 174:1983–1994.
Hudson RR, Kaplan NL. 1985. Statistical properties of the number of
recombination events in the history of a sample of DNA sequences.
Genetics. 111:147–164.
Kalinati YN, Dinesh Kumar V, Reddy SS. 2008. RNA editing in NAD3 and
ATP9 transcripts of safflower (Carthamus tinctorius). Int J Integr Biol.
3:143–149.
Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA,
McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, et al. 2007. Clustal
W and Clustal X version 2.0. Bioinformatics. 23:2947–2948.
Librado P, Rozas J. 2009. DnaSP v5: a software for comprehensive analysis
of DNA polymorphism data. Bioinformatics. 25:1451–1452.
Martin DP, Lemey P, Lott M, Moulton V, Posada D, Lefeuvre P. 2010.
RDP3: a flexible and fast computer program for analyzing recombination.
Bioinformatics. 26:2462–2463.
McCauley DE, Bailey MF, Sherman NA, Darnell MZ. 2005. Evidence for
paternal transmission and heteroplasmy in the mitochondrial genome of
Silene vulgaris, a gynodioecious plant. Heredity. 95:50–58.
McCauley DE, Ellis JR. 2008. Recombination and linkage disequilibrium
among mitochondrial genes in structured populations of the gynodioecious
plant Silene vulgaris. Evolution. 62:823–832.
Morelock TS. 1974. Influence of cytoplasmic source on expression of male
sterility in carrot (Daucus carota L.) [thesis]. [Fayetteville (AR)]: University of
Arkansas.
Mower JP, Touzet P, Gummow JS, Delph LF, Palmer JD. 2007. Extensive
variation in synonymous substitution rates in mitochondrial genes of seed
plants. BMC Evol Biol. 7:135.
Nakajima Y, Yamamoto T, Muranaka T, Oeda K. 1999. Genetic variation
of petaloid male-sterile cytoplasm of carrots revealed by sequence-tagged
sites (STSs). Theor Appl Genet. 99:837–843.
Palmer JD. 1992. Mitochondrial DNA in plant systematic: applications and
limitations. In: Soltis PS, Soltis DE, Doyle JJ, editors. Molecular systematic
of plants. New York: Chapman and Hall. p. 36–49.
Pearl SA, Welch ME, McCauley DE. 2009. Mitochondrial heteroplasmy and
paternal leakage in natural populations of Silene vulgaris, a gynodioecious
plant. Mol Biol Evol. 26:537–545.
Petit RJ, Duminil J, Fineshci S, Hampe A, Salvini D, Vendramin GG. 2005.
Comparative organization of chloroplast, mitochondrial and nuclear
diversity in plant populations. Mol Ecol. 14:689–701.
Pingitore M, Matthews B, Bottino PJ. 1989. Analysis of the mitochondrial
genome of Daucus carota with male sterile and male fertile cytoplasm.
J Hered. 80:143–145.
Recipon H. 1990. The sequence of the sunflower mitochondrial ATPase
subunit 9 gene. Nucleic Acids Res. 6:1644.
Robison MM, Wolyn DJ. 2006. Petaloid-type cms in carrot is not associated
with expression of atp8 (orfB). Theor Appl Genetics. 112:1496–1502.
Robison WW, Wolyn DJ. 2002. Complex organization of the mitochondrial
genome of petaloid CMS carrot. Mol Genet Genomics. 268:232–239.
Ronfort J, Saumitou–Laprade P, Cugen J, Couvet D. 1995. Mitochondrial
DNA diversity and male sterility in natural populations of Daucus carota ssp
carota. Theor Appl Genet. 91:150–159.
Ruths D, Nakhleh L. 2006. RECOMP: a parsimony-based method for
detecting recombination. In: Proceedings of the 4th Asia Pacific Bioinformatics Conference; 2006 Feb 13–16; Taipei, Taiwan. London: Imperial
College Press.
Schnable PS, Wise RP. 1998. The molecular basis of cytoplasmic male
sterility. Trends Plant Sci. 3:175–180.
Smith MJ. 1992. Analyzing the mosaic structure of genes. J Mol Evol.
34:126–129.
Städler T, Delph LF. 2002. Ancient mitochondrial haplotypes and evidence
for intragenic recombination in a gynodioecious plant. Proc Natl Acad Sci
U S A. 99:11730–11735.
Stein M, Nothnagel Th. 1995. Some remarks on carrot breeding (Daucus
carota sativus Hoffm.). Plant Breed. 114:1–11.
Szklarczyk M, Oczkowski M, Augustyniak H, Borner T, Linke B, Michalik B.
2000. Organisation and expression of mitochondrial atp9 genes from CMS
and fertile carrots. Theor Appl Genet. 100:263–270.
Teacher AGF, Griffiths DJ. 2010. HapStar: automated haplotype network
layout and visualization. Mol Ecol Resour. 11:151–153.
Touzet P, Delph LF. 2009. The effect of breeding system on polymorphism
in mitochondrial genes of Silene. Genetics. 181:631–644.
Weir BS. 1996. Genetic data analysis II. Sunderland (MA): Sinauer
Associates.
Welch ME, Darnell MZ, McCauley DE. 2006. Variable populations within
variable populations: quantifying mitochondrial heteroplasmy in natural
populations of the gynodioecious plant Silene vulgaris. Genetics.
174:829–837.
Wolfe KH, Li WH, Sharp PM. 1987. Rates of nucleotide substitution vary
greatly among plant mitochondrial, chloroplast, and nuclear DNAs. Proc
Natl Acad Sci U S A. 84:9054–9058.
Wolyn DJ, Chahal A. 1998. Nuclear and cytoplasmic interactions for
petaloid male-sterile accessions of wild carrot (Daucus carota L.). J Am Soc
Hortic Sci. 123:849–853.
Received October 12, 2011; Revised October 12, 2011;
Accepted November 18, 2011
Corresponding Editor: Kenneth Olsen
425