-
Copyright 0 1984 by the Genetics Society of America
CHLOROPLAST DNA VARIATION IN PEARL MILLET AND RELATED
SPECIES
MICHAEL T. CLEGG,**+ JAMES R. Y. RAWSON* AND KAREN THOMAS*
*Departments .f Botany and tMoleculur and Population Genetics,
University of Georgza, Athens, Georgza 30602
Manuscript received August 2, 1983 Revised copy accepted
November 16, 1983
ABSTRACT
The evolution of specific regions of the chloroplast genome was
studied in five grass species in the genus Pennisetum, including
pearl millet, and one spe- cies from a related genus (Cenchrus).
Three different regions of the chloroplast DNA were investigated.
The first region included a 12-kilobase pair (kbp) EcoRI fragment
containing the 23S, 16s and 5 s ribosomal RNA genes, which is part
of a larger duplicated region of reverse orientation. The second
region was contained in a 21-kbp Sal1 fragment, which spans the
short single-copy sequence separating the two reverse repeat
structures and which overlaps the duplicated copies of the 12-kbp
EcoRI fragment. The third region was a 6-kbp EcoRl fragment located
in the large single-copy region of the chloroplast ge- nome.
Together these regions account for slightly less than 25% of the
chlo- roplast genome. Each of these DNA fragments was cloned and
used as hybrid- ization probes to determine the distribution of
homologous DNA fragments generated by various restriction
endonuclease digests.-A survey of 1 2 geo- graphically diverse
collections of pearl millet showed no indication of chloro- plast
DNA sequence polymorphism, despite moderate levels of
nuclear-encoded enzyme polymorphism. Interspecific and intergeneric
differences were found for restriction endonuclease sites in both
the small and the large single-copy regions of the chloroplast
genome. The reverse repeat structure showed iden- tical restriction
site distributions in all materials surveyed. These results suggest
that the reverse repeat region is differentially conserved during
the evolution of the chloroplast genome.
ECENT advances in molecular biology provide a more detailed view
of R evolutionary processes. T w o types of data are beginning to
accumulate. The first derives from comparative analyses of complete
DNA sequences (MONTGOMERY et al. 1980; BROWN, PRAGER and WILSON
1982; AQUADRO and GREENBURC 1983; BROWN and CLEGG 1983; ZURAWSKI,
CLEGG and BROWN 1984). The second class of data comes from the use
of type I1 restriction endonucleases to fragment DNA molecules at
specific recognition sites. Be- cause restriction endonucleases
cleave DNA molecules if, and only if, a specific permutation of
nucleotides occurs (usually four, five or six nucleotides in
length), the DNA sequences at the termini of the cleavage products
are pre- cisely defined. Restriction endonucleases can thus be used
to detect genetic differences among individuals in a population,
because nucleotide substitutions
Genetics 106: 449-461 March, 1984.
-
450 M. T. CLEGG, J. R. Y. RAWSON AND K . THOMAS
that create, or destroy, restriction sites will alter the
pattern of cleavage prod- ucts. Moreover, events that lead to the
addition, loss or rearrangement of DNA sequences can also be
detected (WYMAN and WHITE 1980).
The utility of restriction analysis for the study of genetic
variation has been firmly established. The most extensive surveys
have concentrated on the DNA sequences coding for the P-globin gene
family in human populations (KAZAZIAN et al. 1983) and on the
mitochondrial genome from a variety of mammalian species (AVISE,
LANSMAN and SHADE 1979; FERRIS, WILSON and BROWN 1981). These
studies show a high level of nucleotide sequence variation; they
also show that the pattern of fragment change contains information
on the se- quence of mutational events that separate different
lineages.
Animal mitochondrial DNAs (mtDNA) have proved to be particularly
useful for the study of sequence variation because these genomes
are small [approx- imately 16 kilobase pairs (kbp)]. Digestion of
mtDNA molecules with a typical six-base restriction endonuclease
will yield from three to six fragments that can be easily separated
by gel electrophoresis. In principle, the rate of nucleo- tide
substitution can be estimated from the fraction of fragments that
comi- grates among pairs of different individuals by a procedure
referred to as the “fragment method” (UPHOLT 1977; NEI and LI
1979). Estimates of rates of nucleotide substitution for mammalian
mtDNAs show a much more rapid rate of evolution than estimated for
nuclear single-copy DNA (BROWN, PRAGER and WILSON 1982).
Studies of plant mtDNA variation are more limited, and the data
that do exist are more difficult to interpret because of the
greater complexity of plant mtDNA (LEVINGS 1983). In addition,
there is some evidence for molecular heterogeneity among mtDNA
molecules within individuals (DALE 198 1). SED- ERHOFF et al.
(1981) and TIMOTHY et al. (1979) have reported mtDNA varia- tion
among different lineages of maize and teosinte. Much of this
variation has been shown to arise from extensive rearrangements of
mtDNA sequences, perhaps due to recombination among different mtDNA
molecules (SEDERHOFF et al. 1981). Consequently, the mode of mtDNA
evolution in plants differs from the mode observed in animals.
Sequence variation has also been reported among chloroplast DNA
mole- cules (cpDNA) in several groups of plants including Zea,
Nicotiana, Lycoper- sicon, Brassica, Triticum, Aegilops, Oenothera
and Hodeum species (TIMOTHY et al. 1979; SCOWCROFT 1979; KUNG, ZHU
and SHEN 1982; PALMER et al. 1983; BOWMAN, BONNARD and DYER 1983;
GORDON et al. 1982; PALMER and ZAMIR 1982; CLEGG, BROWN and
WHITFELD 1984). The chloroplast genome is ap- proximately 130 to
150 kbp in size, so that 25 fragments would be expected following
digestion with many six-base restriction endonucleases. Unambiguous
separation of so large a number of fragments is not possible, and
it is difficult, therefore, to employ the fragment method to
estimate rates of nucleotide substitution. For this reason, and
because we wished to investigate the distri- bution of variant
restriction sites with respect to genome function and struc- ture,
we have studied genetic variation using cloned cpDNA sequences. We
have investigated several species within the genus Pennisetum
including the
-
CHLOROPLAST DNA VARIATION 451
grain crop pearl millet (P. americanum), as well as one species
from the related genus Cenchrus. The resulting data show that
certain cpDNA sequences are highly conserved, whereas other
sequences show variation at or beyond the interspecific level.
MATERIALS AND METHODS
Plant collections: The grass species used in this investigation
are listed in Table 1. In addition to the species listed, 12
samples from within P. americanum were also included in all
comparisons. One of these 12 samples is Tift 23DB, an agronomic
variety in wide use in India and the south- eastern United States.
The remaining 11 samples were obtained from the United States
Depart- ment of Agriculture (USDA) World Collection of pearl
millet. These 11 samples span the mor- phological and geographic
diversity of pearl millet and include entries from India and the
range of Subsaharan Africa.
DNA ~nethods: Procedures for the purification of chloroplast DNA
and total plant DNA (nuclear, chloroplast and probably
mitochondrial) from single-plant samples are given in RAWSON et al.
(1981) and RAWSON, THOMAS and CLECC (1982). Three recombinant DNA
molecules containing various regions of the pearl millet
chloroplast genome were used as hybridization probes. The first
recombinant DNA molecule, Ch9. M (cp.rDNA), contains a 12-kbp EcoRI
fragment carrying the 23S, 16s and 5 s ribosomal RNA genes. The
second cloned molecule was the plasmid pMCS1, which contained a
21-kbp Sal1 fragment. The third was ChS.M(pECPR1) and carried a
6-kbp EcoRI fragment. The cloning and restriction endonuclease
mapping of these cpDNA sequences are described in RAWSON et al.
(1981).
Chloroplast and total DNA preparations were limit digested with
individual restriction endo- nucleases and separated on agarose
gels of varying concentration (0.8, 1.0, 1.5 and 2.0% agarose). DNA
was denatured and eluted from the agarose gels onto strips of
nitrocellulose filter paper (Millipore HA, 0.45 rm) according to
the method of SOUTHERN (1975). The Southern imprints were
hybridized with 82P-labeled DNA that contained cloned pearl millet
cpDNA sequences. Hy- bridization conditions and procedures used to
nick-translate DNA samples are given in RAWSON et al. (1981).
Comparison of autoradiographs prepared from total DNA digests and
from cpDNA digests verified that the fragments homologous to each
cloned DNA sequence employed in this study are located on the
chloroplast genome. One microgram of XDNA, limit digested with
HindIII, was included on each gel as a molecular size standard.
Isozyme procedures: Eight- to 10-day-old seedlings were
extracted in 50 PI of a solution containing 7.6 mM KfHP04, 30 mM
cysteine, 1 mM EDTA, 0.40 M sucrose, 0.20 M Tris, 15 mM sodium
citrate and 6% (w/v) polyvinylpyrrolidone (pH 7.5). In order to
induce alcohol dehydrogenase isozymes, plants were maintained under
flooded conditions for 2 days prior to extraction. Root, leaf and
stem tissues were combined in all assays.
The enzymes assayed included isocitrate dehydrogenase (IDH),
6-phosphogluconate dehydro- genase (GPGD), phosphoglucoisomerase
(PGI), leucine aminopeptidase (LAP), glutamic-oxaloacetic
transaminase (GOT), glutamate dehydrogenase (GDH),
phosphoglucomutase (PGM), malic dehy- drogenase (MDH), alcohol
dehydrogenase (ADH) and esterase (EST). Staining solutions are de-
scribed by BREWER (1970).
Gels of 11% Electrostarch were run for 5 hr at a constant
current of 40 mA at 4" using the following bridge and gel buffer
systems: 0.41 M sodium citrate, pH 7.0 (bridge), and 5 mM
histidine, pH 7.0 (gel), for IDH, GPGD, GDH, PGM, and MDH and 28 mM
LiOH with 0.19 M boric acid (bridge) and 4 mM citric acid, 46 mM
Tris, 3 mM LiOH and 22 mM boric acid (gel) for GOT, PGI, LAP, ADH
and EST.
RESULTS
Chloroplnst DNA restriction fragment analyes: A diagram of the
pearl millet chloroplast genome is given in Figure 1, together with
the map location of the three cloned millet cpDNA sequences used as
hybridization probes. The clone
-
452 M. T. CLEGG, J. R. Y. RAWSON AND K. THOMAS
TABLE 1
Plant species investigated
Species Abbreviation“ Chromosome
no.
P . a inericn n u in P . jlacciduin P . orientale P. p u rpu
rpeuin P . setaceuin C. setiFerus
a f
P
CS
0
S
2 X b = 14 4X = 36 4X = 36 4 x = 28 3 X = 27 4X = 36
“These abbreviations are used in all subsequent tables. X refers
to haploid chromosome number.
CI
FIGURE I.-Schematic diagram of the pearl millet chloroplast
genome. The locations of the cpDNA sequences contained in the
hybridization probes Chg.M(cp.rDNA), pMCSl and Chg. M(pECPR1) are
shown. The reverse repeat regions are represented by the heavy
line.
Ch9. M(cp. rDNA), includes the 1 2-kbp EcoRI cpDNA fragment
coding for the 23S, 16s and 5s rRNA and is located entirely within
a large (approximately 22 kbp) reverse repeat structure (RAWSON et
al. 1981), which is characteristic of most higher plant chloroplast
genomes (BEDBROOK and KOLODNER 1975).
-
CHLOROPLAST DNA VARIATION 453
Thus, the DNA sequences included within the Ch9. M(cp.rDNA)
probe exist twice on the chloroplast genome. The cloned millet
cpDNA sequences con- tained on the plasmid pMCSl overlap the two
identical inverted repeat regions and include 12 kbp of single-copy
DNA, which separates the repeat structures. Detailed restriction
maps of pMCS1 and Ch9.M(cp.rDNA) have been pub- lished (RAWSON et
al. 1981) and are outlined in Figure 2. The map location can be
used to predict the expected number of restriction endonuclease
sites and the expected number of DNA fragments generated by a given
digest. (The EcoRI sites in the small single-copy region spanned by
pMCS1 produce many small fragments and have not been mapped.) It is
thus possible to un- ambiguously recognize restriction endonuclease
site changes within one of the two duplicate regions. The third
hybridization probe, Ch9. M(pECPR l), con- tains a 6-kbp EcoRI DNA
fragment of the millet chloroplast genome and maps in the large
single-copy region of the chloroplast genome as shown in Figure
1.
Variation among sequeizces homologous to C h 9 .M(cp. rDNA): Our
survey of nu- cleotide sequence variation using the probe Ch9.
M(cp. rDNA) gave identical fragment patterns for all 12 lines of
pearl millet and all five related species. A total of three
four-base and seven six-base restriction endonucleases was used in
the survey (Table 2). We can use the restriction map generated with
the six-base enzymes to count the number of nucleotides observed
within these duplicate regions (282 nucleotides). Although the
restriction endonuclease sites for the four-base enzymes have not
been mapped, we estimate that a total of 176 nucleotides was
detected using this class of enzymes (assuming that each observed
band represents two identical fragments). Thus, a total of
approxi-
Sal1 I I I L
BamH1 I I I I 1 1 I 1 1
Pst I I I I I I I I
Xhol I I I I I I 1 I
55 235
55
FIGURE 2.-A restriction map of the cp.rDNAs and the short
single-copy DNA sequence of pearl millet. The cp.rDNA of pearl
millet consists of two inverted repeats, each containing a complete
rRNA gene cluster. The two rDNA sequences are separated by 12 kbp
of single-copy DNA. This region of the cp.DNA of pearl millet is
represented by two recombinant DNA mole- cules: Chg.M(ct.rDNA) and
pMCS1. The EcoRI sites in pMCSl were numerous and thus not
mapped.
-
454 M. T. CLEGG, J. R. Y. RAWSON AND K. THOMAS
TABLE 2
Restriction endonucleases used to digest DIVA preparations
Enzyme
EcoRI Bn iii H I Hind111 PstI Sal1 Xhol HpoI HpaII HwIII
RsaI
Recognition site pMCSl Ch9.
M(pECPR1)
GAATTC GGATCC AAGCTT CTGCAG GTCGAC CTCGAG GTTAAC CCGG GGCC
GTAC
X X X
X X
X
X X X X
X
The probes used with each digest are indicated by an x.
mately 458 nucleotides was included in the recognition sites of
the enzymes used in this study. (In making this calculation, we
have assumed that none of the four-base sites overlaps a six-base
site.)
An estimate of the maximum fraction of nucleotide substitutions
can be made using the following assumptions: First, we assume that
the true fraction is p and that 5% or fewer of samples would have
produced a result as deviant as that observed. When binominal
sampling is assumed, p(0.05) = 0.0065. Second, we make the less
conservative assumption that 50% of samples would produce a result
as deviant as that observed. Then p(0.50) = 0.0015. Of course, the
best estimate based upon the data at hand is p = 0.0, i . e . ,
absolute conservation of the surveyed nucleotides among the
lineages compared.
Vuriatioti among sequences homologous to PMCSl: All 12 samples
from the USDA world collection of pearl millet gave identical
fragment patterns for sequences homologous to pMCS1 when digested
with the seven six-base enzymes listed in Table 2. Comparisons
between species revealed different distributions of fragment
patterns for the enzymes EcoRI, HpnI and BamHI, whereas the re-
maining four enzymes gave identical fragment distributions. Table 3
tabulates the estimated fragment distributions for all digests
observed to vary.
Consider first the digest with BamHI. All species of Pennisetum
have an identical fragment distribution that differs from Cenchrus
in having a 14.1- kbp fragment and lacking a 10.0- and a 5.0-kbp
fragment. The error associated with estimating the size of large
fragments on agarose gels is approximately 10%; hence, the 14.1-kbp
fragment may differ from the 10.0- and 5.0-kbp fragments by a
single BainHI site. Reference to the BainHI map (Figure 2) shows
that the 2.9-kbp band represents the comigration of three
fragments: a duplicate pair of fragments from the reverse repeat
and a fragment from the small single-copy region.
The EcoRI digests produce three different DNA fragment
distributions. P. aineriranuin has a 1.5-kbp fragment absent in the
other distributions, which have instead 0.9- and 0.5-kbp fragments.
We infer that a single EcoRI site separates these fragments. P.
americnnuin, P. purpureum and C. setigvrus all have
-
CHLOROPLAST DNA VARIATION 455 TABLE 3
Chloroplast DNA fraginrnts obserued using pMCS1
EcoRI HpaI BamHI
Frag- Frag- Frag- ment a f,o,s p,Cs ment a,f,o p,s Cs ment
a,f,o,s,p Cs
A 12.0 12.0 12.0 A 10.0 10.0 10.0 A 14.1 B 2.6 2.6 B 7.6 7.6 7.6
B 10.0 C 2.4 c 5.3 C 6.8 6.8 D 2.1 2.1 2.1 D 2.7 2.7 D 5.0 E 1.5 E
2.6 2.6 2.6 E 2.9 2.9 F 1.2 1.2 1.2 F 2.0 2.0 F 1.5 1.5 G 0.9 0.9 H
0.6 0.6 0.6 I 0.5 0.5 J 0.4 0.4 0.4 K 0.3 0.3 0.3
DNA fragments are in kilobase pairs. Letters indicate the
species with the tabulated fragment distribution. See Table 1 for
species identification by letter.
a 2.6-kbp fragment but lack the 2.4-kbp fragment present in the
other species. We have been unable to detect the putative 0.2-kbp
fragment and cannot exclude the possibility that this difference
results from an addition or deletion of a DNA sequence.
Finally, consider the HpaI digest. There are two patterns within
the Penni- setum species. The first pattern shows a 5.3-kbp
fragment absent in P. pur- pureuin and P. setaceuin, which have
fragments of 2.7 and 2.0 kbp instead. The 0.6-kbp difference
between these digests is well beyond the range of measure- ment
error and may reflect multiple site differences or an addition or
deletion event.
All of the variant sites observed using the 21-kbp Sal1 DNA
fragment con- tained in pMCS1 map into the 12-kbp single-copy
region. The DNA sequences located on this DNA fragment that map
within the reverse repeat regions are invariant as expected from
the studies using Ch9. M(cp. rDNA).
Variation f o r sequences hoinologus to Ch9 e(pECPR1): Like the
previous probes, no variation in fragment distribution was observed
among the 12 lines of pearl millet for sequences homologous to the
6-kbp EcoRI DNA fragment in the recombinant DNA probe Ch9.
M(pECPR1). However, variation among species was observed for
digests with EroRI, BainHI and HpaII. Table 4 tabulates fragment
sizes of all observed patterns. Figure 3 shows the fragment distri-
butions observed for EcoRI and HpaIl.
The analysis of DNA fragment pattern differences requires
additional com- ment. For instance, the Pennisetum species exhibit
a 6.0-kbp EcoRI fragment that differs from the 6.9-kbp fragment
observed in C. setigerus. Because the probe containsjust 6 kbp of
millet cpDNA, we infer that loss of an EcoRI site in C. setigerus
bounding the clone from P. ainericanuin differentiates the two
patterns. Likewise single-site differences can account for the
BamHI and HpaIl patterns, except for the 0.29-kbp fragment observed
in C. setigerus.
-
456 M. T. CLEGG, J. R. Y. RAWSON AND K. THOMAS
TABLE 4
Chloroplnst DIVA fragments observed using Ch9. M(pECPR1)
EcuRI HpaII EamHI
Frag- Frag- Frag- cs ment a,f,o,p,s Cs ment a,f,o,s p c s ment
a,f,o,p,s
A 6.9 A 1.6 1.6 1.6 A 3.2 3.2 B 6.0 B 0.8 0.8 0.8 B 2.5 2.5
C 0.7 0.7 0.7 C 1.8 D 0.6 D 1.6 E 0.59 0.59 E 1.1 1.1 F 0.3 0.3
0.3 G 0.29 H 0.2 0.2 0.2
DNA fragments are in kilobase pairs. Letters indicate the
species with the tabulated fragment distribution. See Table 1 for
species identification by letter.
If we assume that all fragment pattern changes are due to
nucleotide sub- stitution, then the number of base substitutions
per base pair @) can be esti- mated by the “fragment method”
(UPHOLT 1977; NEI and Lr 1979). These estimates, which were
obtained from the fraction of shared fragments over all digests
(F), are reported in Table 5. Duplicate fragments, predicted from
the restriction map of the Sol1 fragment in pMCS1, are counted
twice in calculat- ing F. Table 5 shows that, even at the
interspecific level, the fraction of base changes is small. For
example, the best estimate is zero for the comparison between P.
flucciduin and P. orientale. The mean estimates of p over all
pairwise comparisons for each probe are p = 0.0032 k 0.001 for the
SalI fragment in pMCS1 and p = 0.0075 +- 0.005 for the 6-kbp EcoRI
fragment in Ch9. M(pECPR1). The difference between means is further
reduced if the estimates are corrected for the fact that
approximately half of the millet cpDNA se- quences in the SalI
fragment in pMCS1 fall within the invariant reverse repeat region.
Thus, the estimates of p are less than 1% even at the intergeneric
level. These estimates should be interpreted with caution, however,
since we can not ascribe all fragment pattern changes to
single-nucleotide substitutions. Indeed, addition/deletion events
are common in comparative sequence studies of noncoding portions of
the chloroplast genome (ZURAWSKI, CLEGC and BROWN 1984; TAKAIWA and
SUGIURA 1982) and in comparative restricting mapping of chloroplast
genomes (PALMER, SINGH and PILLAY 1983; BOWMAN, BONNARD and DYER
1983; GORDON et al. 1982).
Vurintioli for isozyme loci: The 12 lines sampled from the USDA
World Col- lection of pearl millet were surveyed for variation
using ten enzyme systems (IDH, GPDG, PGI, LAP, GOT, GDH, PGM, MDH,
ADH and EST). Formal genetic analyses have been conducted for ADH
(BANUETT-BOURRILLON and HAGUE 1979) and 6PGD (M. T. CLEGG,
unpublished data). Tentative assign- ments of banding phenotypes to
genetic loci were made for the remaining systems based on patterns
of segregation in population samples. With the use of these
assignments, 15 genetic loci determine the banding phenotypes
ob-
-
CHLOROPLAST DNA VARIATION 457
3 4 5 Kbp -- I- - KbP 1 2
0.3. 0.29
EcoR I Hpsl I FIGURE 3.-Restriction endonuclease fragment
patterns for sequences homologous to Ch9.
M(pECPR1). EmRl digests were separated on 0.8% agarose gels.
Hpnll digests were separated on 2% agarose gels. Fragments were
transferred to nitrocellulose filter paper ( I 7) and were
hybridized to nick-translated probe. Samples run in each lane were:
( I ) P. nwrirnnum, (2) C. srtigtrus, (3) P. nntrricovutn, (4) P.
purpurnrm and (5) C. .w/igrrus.
served for all enzyme systems. Seven loci are polymorphic in one
or more of the 12 lines, yielding a crude estimate of 47% of loci
polymorphic. This esti- mate is comparable to estimates for other
cultivated and wild grass species (HAMRICK, LINHARDT and MIITON
1979). The failure to detect intraspecific variation for cpDNA
sequences in pearl millet is clearly not indicative of low levels
of genetic variability for nuclear sequences translated into
enzymatic proteins.
DISCUSSION
The major results of this investigation are (1) restriction
endonuclease sites within the reverse repeat region are highly
conserved, and (2) changes in restriction fragment patterns do
occur for singletopy regions. The factors
-
458 M. T. CLEGG, J. R. Y . RAWSON AND K . THOMAS
TABLE 5
Matrix of the per nuclrotide number of base substitutions
estimated using the SalI Jragmwt in piMCSl (in boxhectd) nnd EcoRZ
fragment in Ch9.M(pECPRl) (in stub)
Species Spe- cies a f 0 S P cs a
f 0.0
0 0.0
S 0.0
P 0.0049 (0.01 16)
c s 0.0070 (0.0065)
0.0030 (0.0035)
0.0
0.0
0.0049 (0.0116) 0.0070
(0.0065)
0.0030 (0.0035) 0.0
0.0
0.0049 (0.01 16) 0.0070
(0.0065)
0.0064 (0.0051) 0.0017
(0.0026) 0.0017
(0.0026)
0.0049 (0.0116) 0.0070
(0.0065)
0.0036 (0.0038) 0.0038
(0.0039) 0.0038
(0.0039) 0.0037
(0.0039)
0.0116 (0.0114)
0.0036 (0.0038) 0.0038
(0.0039) 0.0038
(0.0039) 0.0037
(0.0039) 0.0022
(0.0030)
The rows and columns are labeled with the first letter of the
species name except for the last row and column, which is denoted
by the first letters of the genus and species names (see Table 1).
Standard errors are given in parentheses.
responsible for the conservation of the DNA sequences within the
reverse repeat region are not completely understood. An important
contributing factor to this conservation arises from the DNA
sequences coding for the chloroplast ribosomal RNA (cp. rDNA),
which are very highly conserved in evolution. We have shown,
through the melting of DNA heteroduplexes formed between the
cp.rDNA of the alga Euglena and the cp-rDNA of P. americanum, that
the cp. rDNA is highly conserved between these phylogenetically
distant organisms (RAWSON et al . 1981). However, cp.rDNA sequences
account for less than 50% of the DNA studied within the reverse
repeat region. Other studies of cpDNA evolution also show
differential conservation of the inverted repeat region (KUNG, ZHU
and SHEN 1982; PALMER, SINGH and PILLAY 1983). How- ever, in a
recent study of three legume species, PALMER et al. (1983) showed
that slow rates of evolution for the cp.rDNA sequences account for
much of the differential conservation associated with the inverted
repeat region.
The reverse repeat structure may act to maintain genetic
homogeneity among these duplicate sequences. For example, it has
been suggested that the reverse orientation of the repeat region
prevents the loss or duplication of genetic material due to
intragenomic recombination events (KOLODNER and TEWARI 1979). Under
this hypothesis the two regions would not accumulate nucleotide
substitutions independently, but, instead, new mutant sites would
be transferred by recombination to both duplicate regions. There is
some evidence from Chlamydomonas to support this suggestion (MEYERS
et al. 1982). In addition, the large and small single-copy regions
of the chloroplast genome of the common bean (Phaseolus vulgaris)
have recently been shown to exist in two orientations with respect
to one another (PALMER 1983), and a similar
-
CHLOROPLAST DNA VARIATION 459
situation appears to exist for the cyanelle genome of Cyanophora
pamdoxa (BOH- NERT and LOFFELHARDT 1982). Two orientations are
consistent with intra- molecular recombination, which may occur
frequently. Nevertheless, different lineages would be expected to
evolve independently, and we would, therefore, expect to see
differences at the interspecific or intergeneric level. Whatever
the cause, the slow rate of genetic change observed for cpDNA
sequences associated with the reverse repeat region opens up the
possibility of using this region to infer genetic relationships
among diverse plant taxa, perhaps up to the family level.
The single-copy regions of cpDNA monitored in this investigation
do vary among species, but the rate of change appears to be much
less than estimated for mammalian mtDNA evolution. Although there
are no good estimates of divergence times among these species,
there is a striking contrast between intraspecific estimates of p =
0.015 for old field mice (ADVISE, LANSMAN and SHADE 1979) and p =
0.03 for pocket gophers (AVISE et al. 1979) with the intergeneric
estimate reported here @ = 0.006). Moreover, the slow rate of cpDNA
evolution does not appear to be correlated with a low level of
genetic variability for nuclear genes. Levels of isozyme
variability are high, and changes in the basic chromosome number
have occurred during the evolution of these species.
We thank WAYNE HANNA for supplying plant materials and for
advice during the course of this project. Supported in part by
grants from the National Science Foundation (DEB-81 18414) and the
United States Department of Agriculture Competitive Grant Program
(80-CRCR-1-0489).
LITERATURE CITED
AQUADRO, C. F. and B. D. GREENBERG, 1983 tion: analysis of
nucleotide sequences from seven individuals. Genetics 103: 287-3
12.
AVISE, J. C., D. GIBLIN-DAVIDSON, J. LAERM, J. C. PATTON and R.
A. LANSMAN, 1979 Mitochondrial DNA clones and matriarchal phylogeny
within and among geographic populations of the pocket gopher,
Geomys pinetis. Proc. Natl. Acad. Sci. USA 76 6694-6698.
T h e use of restriction endonucleases to measure mitochondrial
sequence relatedness in natural populations. I. Population
structure and evolution in the genus Peromyscus. Genetics 92:
279-295.
BANUETT-BOURRILLON, F. and D. R. HAGUE, 1979 Genetic analysis of
alcohol dehydrogenase isozymes in pearl millet (Pennisetum
typhoides). Biochem. Genet. 17: 537-552.
BEDBROOK, J. R. and R. KOLODNER, 1979 T h e structure of
chlorplast DNA. Annu. Rev. Plant Physiol. 30: 593-620.
BOHNERT, H. J. and W. LOFFELHARDT, 1982 Cyanelle DNA from
Cyanophora paradoxa exists in two forms due to intramolecular
recombination. FEBS Lett. 150 403-406.
BOWMAN, C. M., G. BONNARD and T. A. DYER, 1983 Chloroplast DNA
variation between species of Triticum and Aegzlops: location of the
variation on the chloroplast genome and its relevance to the
inheritance and classification of the cytoplasm. Theor. Appl.
Genet. 65: 247-262.
Human mitochondrial DNA variation and evolu-
AVISE, J. C., R. A. LANSMAN and R. 0. SHADE, 1979
BREWER, G. J., 1970
BROWN, A. H. D. and M. T. CLECC, 1983
A n Introduction to Isozyme Technique. Academic Press, New
York.
Analysis of variation in related DNA sequences. In:
Mitochondrial DNA sequences in pri-
Statistical Analysis of D N A Sequences, Edited by B. S. WEIR.
Marcel Dekker, Inc., New York.
mates: tempo and mode of evolution. J. Mol. Evol. 18: 225-239.
BROWN, W. M., E. M. PRACER and A. C. WILSON, 1982
-
460
CLEGG, M. T . , A. H. D. BROWN and P. R. WHITFELD, 1984
M. T. CLEGG, J. R. Y. RAWSON AND K. THOMAS
Chloroplast DNA diversity in wild
Sequence homology among different size classes of plant mtDNAs.
Proc.
Evolutionary tree for apes and humans based on cleavage maps of
mitochondrial DNA. Proc. Natl. Acad. Sci. USA 78: 2432-2436.
Physical mapping of differences in chloroplast DNA of the five
wild-type plastomes in Oenothera subsection Euoenothera. Theor.
Appl. Genet. 61: 373-384.
Relationships between life history characteristics and
electrophoretically detectable genetic variations in plants. Annu.
Rev. Ecol.
DNA poly- morphisms in the human p globin gene cluster. In:
Evolution of Genes and Proteins, Edited by M. NEI and R. K. KOEHN.
Sinauer Associates, Inc., Sunderland, Massachusetts.
KOLODNER, R. and K. K. TEWARI, 1975 T h e molecular size and
conformation of the chloroplast DNA from higher plants. Biochim.
Biophys. Acta 402: 372-390.
KUNG, S. D., Y. S. ZHU and G. F. SHEN, 1982 Nicotiana
chloroplast genome. 111. Chloroplast DNA evolution. Theor. Appl.
Genet. 61: 73-79.
LEVINGS, C. S., 1983
MEYERS, A. M., D. M. GRANT, D. K. RABERT, E. H. HARRIS, J. E.
BOYNTON and N. W. GILLHAM, Mutants of Chlainydomonas reinhardtii
with physical alterations in their chloroplast DNA.
MONTGOMERY, D. L., D. W. LEUNG, M. SMITH, P. SHALIT, G. FAYE and
B. D. HALL, Isolation and sequence for the gene for
iso-2-cytochrome c in Saccharomyces cermisiae.
Mathematical model for studying genetic variation in terms
of
and cultivated barley: implications for genetic conservation.
Genet. Res. In press.
Natl. Acad. Sci. USA 78: 4453-4457. DALE, R. M. K., 1981
FERRIS, S. D., A. C. WILSON and W. M. BROWN, 1981
GORDON, K. H. J., E. J. CROUSE, H. J. BOHNERT and R. G.
HERRMANN, 1982
HAMRICK, J. L., Y. B. LINHART and J. B. MITTON, 1979
Syst. 10: 173-200.
KAZAZIAN, H. H., A. CHAKRAVARTI, S. H. ORKIN and S. E.
ANTONARAKIS, 1983
T h e plant mitochondrial genome and its mutants. Cell 32:
659-661.
1982 Plasmid 7: 133- 15 1 .
1980 Proc. Natl. Acad. Sci. USA 77: 541-545.
NEI, M. and W. H. LI, 1979
PALMER, J. D., 1983
PALMER, J. D., C. R. SHIELDS, D. B. COHEN and T . J. ORTON,
1983
restriction endonucleases. Proc. Natl. Acad. Sci. USA 76:
5269-5273.
Chloroplast DNA exists in two orientations. Nature 301:
92-93.
and the origin of amphidiploid Brassica species. Theor. Appl.
Genet. 65: 181-189.
Structure and sequence evolution of three legume chloroplast
DNAs. Mol. Gen. Genet. 190: 13-19.
Chloroplast DNA evolution and phylogenetic relationships in
Lyco/xmicon. Proc. Natl. Acad. Sci. USA 79: 5006-5010.
A restriction endonuclease map of the ribosomal RNA genes and
short single copy DNA sequence of the pearl millet chloroplast
genome. Gene 16: 11-19.
Purification of total cellular DNA from a single plant. Biochem.
Genet. 2 0 209-219.
Nucleotide polymorphism in chloroplast DNA of Nicotiana debneyi.
Theor. Appl. Genet. 55: 133-137.
Evolution of DNA sequence organization in mitochondrial genomes
of Zea. Proc. Natl. Acad. Sci. USA 78: 5953- 5957.
Detection of specific sequences among DNA fragments separated by
gel electrophoresis. J. Mol. Biol. 98: 503-517.
Chloroplast DNA evolution
PALMER, J. D., G. P. SINCH and D. T. N. PILLAY, 1983
PALMER, J. D. and D. ZAMIR, 1982
RAWSON, J. Y. R., M. T . CLEGG, K. THOMAS, C. RINEHART, and B.
WOOD, 1981
RAWSON, J. R. Y., K. THOMAS and M. T. CLEGG, 1982
SCOWCROFT, W. R., 1979
SEDERHOFF, R. R., C. S. LEVINGS, D. H. TIMOTHY and W. W. L. Hu,
1981
SOUTHERN, E. M., 1975
-
CHLOROPLAST DNA VARIATION 461
Nucleotide sequences of the 16s-23s spacer region in an rRNA
gene cluster from tobacco chloroplast DNA. Nucleic Acids Research
8: 2665-2676.
TIMOTHY, D. H., C. S. LEVINGS, D. R. PRING, M. F. CONDE and J.
L. KERNICKE, 1979 Organelle DNA variation and systematic
relationships in the genus Zea: Teosinte. Proc. Natl. Acad. Sci.
USA 76: 4220-4224.
Estimation of DNA sequence divergence from comparison of
restriction
A highly polymorphic locus in human DNA. Proc. Natl.
ZURAWSKI, G., M. T. CLEGG and A. H. D. BROWN, 1984 T h e nature
of nucleotide sequence
TAKAIWA, F. and M. SUGIURA, 1982
UPHOLT, W. B., 1977 endonuclease digests. Nucleic Acids Res. 4
1257-1265.
Acad. Sci. USA 775754-6758. WYMAN, A. R. and R. WHITE, 1980
divergence between barley and maize chloroplast DNA. Genetics.
In press.
Corresponding editor: J. R. POWELL