Plant Molecular Biology 15: 269-280, 1990. © 1990 Kluwer Academic Publishers. Printed in Belgium.

269

Chloroplast DNA variability in the genus Helianthus: restriction analysis and S1 nuclease mapping of DNA-DNA heteroduplexes Pascale Serror, 1 Franfoise Heyraud 2 and Philippe Heizrnann 3, Laboratoire de Biologie Cellulaire, Universit~ Claude-Bernard, Lyon-I, 69622 Villeurbanne cedex, France; 1Present address: INRA, Centre de Recherches de Jouy-en-Josas, Domaine de Vilvert, 78350 Jouy-en-Josas, France; 2present address: Max-Planck-lnstitut far Ziichtungsforschung, Abteilung Schell, Carl yon Linn~ Weg, D-5000 KOln 30, FRG; 3Present address: INRA Reconnaissance Cellulaire et Amdlioration des Plantes, Universit~ Claude-Bernard, Lyon-I, 69622 Villeurbanne cedex, France (* author for correspondence) Received 26 February 1990; accepted in revised form 9 May 1990

Key words: chloroplast DNA variability, Helianthus, restriction endonuclease analysis, S1 nuclease mapping

Abstract

Chloroplast DNA (cpDNA) from 36 wild species of the genus Helianthus has been analysed with three restriction endonucleases (Bam HI, Hind III and Sst I). Out of the 71 restriction sites described on the reference cpDNA (sunflower cpDNA), three insertions/deletions and seven site modifications were detected during the survey of the other cpDNAs. Since restriction mapping showed only a very limited fraction of the DNA variability, we chose to adapt the S 1 nuclease mapping technique to detect fme variations between chloroplast genomes. For this purpose, DNA-DNA heteroduplexes obtained between sunflower and wild-species DNAs were digested by S 1 nuclease and the resulting mismatches were detected by classical endonuclease restriction and hybridization methods. The S 1 nuclease mapping results were confirmed by sequencing one S1 nuclease-sensitive region detected between cultivated sunflower and two perennial wild-type species. As a result of these analyses, it appeared that the combination of restriction mapping and S 1 nuclease mapping might be helpful to differentiate taxonomically close cytoplasms.

Introduction

The sequence divergence of chloroplast DNAs (cpDNAs), as estimated by comparison of their restriction cleavage patterns, has been widely used to trace the phylogenetic relationships between related species, genera or families of higher plants [ 16]. The chloroplast DNA of land plants has indeed a size range (120-180 kb) that

allows a satisfactory resolution of the fragments produced by most six-base recognition enzymes and that provides an informative analysis of this genome. Several different kinds of structural variations have occurred during the evolution of land plant chloroplast genomes: large inversions were fairly rare and can thus be used to differentiate large phylogenetic groups [14]; single-base substitu-

270 tions and small additions/deletions (1 to 100 bp) occurred more frequently and allow discrimination between closely related species or genera [24]. In this work, we have described the variability of the c p D N A between various species or ecotypes of the genus Helianthus; we analysed particularly some ecotypes used in interspecific crosses with H. annuus, to create CMS lines of cultivated sunflower [11], in an attempt to differentiate the various male-sterile cytoplasms. We compared the variations observed by restriction mapping with those detected by an S 1 mapping technique of c p D N A heteroduplexes. The restriction site polymorphism is due mainly to base substitutions at the level of restriction sites, while the S 1 nuclease sensitivity results from mismatches of several base pairs between related sequences [20, 3 ]. The combined techniques permitted us to analyse the relative importance of base substitutions and addition/deletion events. Our results show a certain predominance of small additions, deletions or duplications over single-base substitutions, suggesting that both kinds of polymorphism should be taken into account to differentiate closely related species.

Materials and methods

Plant materials The species, ecotypes and cultivars analysed in this study are listed in Table 1 ; they were mainly obtained from Dr. H. Serieys (INRA station, Montpellier, France), either as seeds grown in Lyon under greenhouse conditions, or as frozen leaves collected in Montpellier.

Chloroplast DNA specific probes A library of Hind III fragments of sunflower chloroplast D N A was constructed in lambda phage and subcloned in plasmid vectors. These probes, reported in Fig. 1, represent 70~o of the sunflower plastome complexity.

Tab& 1. Sources of Helianthus DNA.

Cultivars

Species

Ecotypes

Sunflower (H. annuus)

HA89B HA89A 2 Bolrro 3

Jerusalem artichoke

ID 19

(H. tuberosus)

Annual species

H. H. H. H. H. H. H. H. H.

annuus anomalus argophyllus bolanderi debilis deserticola neglectus niveus petiolaris

H. praecox

Perennial species H. angustifolius H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H. H.

arizonensis californicus decapetalus divaricatus eggertii giganteus glaucophyllus grosseserratus hirsutus laevigatus macrophyllus maximiliani micranthus microcephalus mollis nuttallii occidentalis purnilus resinosus salicifolius silphioides simulans strumosus tomentosus tuberosus

119, 397,519, 521 525 741 667,255,331,671 218 223 201 213

123, 199, 574 200, 674, 737,738 198,219, 220 111,529 594 105,242 100, 551 232 234 552, 553,554 246 237 672 562 107 567,568 106 245,261 230 239, 743 231 227 243 599 262, 531 596 527 287 289, 326

Male-fertile line. 2 Male-sterile line (Leclercq cytoplasm [11]). 3 Hybrid 62 × ZN41 (Leclercq cytoplasm).

271 EcoRI Hincll BamHI

Hindlll

I I

1

10a 15b 10b

10b 15b 10a

\1/

12+14 I

12+14 I

19 14 16

\., SAI/ s //l

IRB

~2',!4

1

14!214"a I

SSC

11 13

BamHI

\' IIIII V#A

0 13 15b 11a 11b ~11 ~ \ IJ /

19 ,.JA 71 11 :111 911' a 2 .,llglltll13blTtllll c li314"b14"a

BamHl

2

I 4 191

IRA

ll13a14 I .l bllo, l s 3

Hindlll

1

,

Hindlll I B.m., s,,,

LSC

Fig. 1. Localization of cpDNA variations in the genus Helianthus on the physical restriction map of sunflower (H. annuus) cpDNA. Polymorphic restriction sites which have been perfectly localized are represented by thick black arrows; the others (which have not been perfectly localized) are represented by thin black arrows. Length mutations are represented by black asterisks. The hatched regions represent the cloned Hind III restriction fragments of sunflower chloroplast DNA which have been used as probes. IRA, IRB: inverted repeat regions; SSC, LSC: small and large single-copyregions.

DNA preparation and restriction analysis

Sunflower chloroplast D N A or total D N A s from wild species were phenol-extracted starting respectively from purified chloroplasts or from frozen leaves. D N A s (2 to 10 #g) were digested to completion, electrophoresed on agarose gels, then transferred onto nylon membranes (Hybond, Amersham) and hybridized with probes labelled in vitro (Boehringer random priming kit) with 32p-dCTP (Amersham).

DNA-DNA heteroduplex S1 nuclease mapping

S 1 nuclease mapping was adapted from the protocol described by Chebloune et al. [3]. 2 to 10 #g of total D N A (used in equimolar quantities to favour the formation of heteroduplexes) were denatured by 0.2 M N a O H for 2 hours at 37 °C in a total volume of 200 #1, neutralized on ice with 2 M HC1 (20#1) and 2 M Tris p H 7.9 (10 #1), then precipitated with ethanol and centrifuged for 20 min in an Eppendorf centrifuge. The pellets were dissolved in 14 #1 of sterile water and 20 #1 ofdeionized formamide. 6 #1 of hybridization mix

(250 mM Tris-HC1 pH 7.9, 250 mM EDTA, 3.75 M NaC1) were added to the D N A solution. Reassociation was carried out for 18 hours at 37 °C. The mixture was then diluted 10 times with S1 nuclease buffer ( 1 2 m M sodium acetate pH 4.5, 200 mM NaC1, 4 m M ZnSO4) and incubated for 1 hour at 37 °C with 10 units of S1 nuclease (BRL) per #g of total DNA. The digestion was stopped by phenol/chloroform extraction and the remaining D N A was ethanol-precipitated using glycogen as carrier. The next steps (endonuclease restriction, agarose gel electrophoresis, Southern transfer, hybridization with 32p-labelled probes and autoradiography) were carried out as mentioned above.

DNA sequencing The Bam HI, chloroplast restriction fragment B14 from H. annuus (cv. Bolero with Leclercq cytoplasm), H. tuberosus and H. decapetalus were subcloned from lambda libraries into M13mpl8 and M13mpl9 vectors [12]. Single-stranded D N A s were sequenced by the dideoxy chain termination method described by Sanger et al. [ 18].

272

Fig. 2. Hybridization ofH4a, H2, H4b, H3, H6 and H7 probes to wild-species and sunflower HA89B total DNAs digested with Barn HI (B), Hind III (H), Sst I (S), Eco RI (E) or Hint II (Hc) restriction endonucleases. Asterisks and plus signs indicate

restriction site polymorphisms and restriction length polymorphisms respectively.

Results Chloroplast restriction polymorphism in the genus Helianthus

Chloroplast DNA restriction fragments were detected among total DNA preparations by hybridization with cloned chloroplast DNA probes from H. annuus. A f'trst estimation of the chloroplast genome variability was performed with Barn HI, Hind III and Sst I digests, since the

physical map of sunflower chloroplast DNA had been previously built with these enzymes (see Fig. 1) and since they generate fragments well suited in size and number for an easy interpretation of the restriction patterns. In addition, the Hind III fragments H3 and H4b were also hybridized on Eco RI blots, and the Hind III fragments H7 on Hinc II blots. The hybridization results are summarized in Table 2 and are illustrated in Fig. 2. This analysis demonstrates the occurrence of a restriction site

273

polymorphism and of a restriction fragment length polymorphism (RFLP), the two kinds of polymorphism classically described in the literature [16]. Restriction site polymorphism events were detected in the case of hybridization with the Hind III fragments H4a, H6, H7 and H6' used as probes: two split subfragments, resulting from the acquisition of a supplementary restriction site, were revealed in some wild species in comparison with cultivated sunflower. The Hind III fragments HI, H2, H4b and H3 used as probes detected only single variant fragments since the supplementary sites were probably not located in the

regions actually complementary with these probes. RFLP events were detected in perennial species only: for instance, the Barn HI fragment B 19 bears an insertion/deletion change of 20 base pairs in all perennial species analysed except the ecotype 287 of 1-1. tomentosus. Similarly the ecotype 594 ofH. arizonensis was found to be the only one among perennial species to show insertion/ deletion changes of about 100 base pairs on the Barn HI fragments B 14 and B8*. Seventy-one restriction sites were analysed in 40 ecotypes belonging to 36 different species. 3 insertions/deletions and 7 site modifications were

274 Table 2. Summary of cpDNA variations detected in the genus Helianthus with sunflower probes.

Probe

Enzyme

Reference pattern 1

Mutated pattern

Mutated ecotypes

Mutation class

H4a

Hind III

H4a (9 kb)

H4a* (8.5 kb) H4a** (0.5 kb)

following ecotypes of annual species: 123-198-201-213 219-255-331-519-521-667-671-674 737-738-741

1

H6

Sst I

$9 (5 kb)

$9" (3.5 kb) $9"* (1.4 kb)

255 of H. bolanderi

1

H7

Hinc II

Hc (2.75 kb)

Hc (2.0 kb) Hc (0.8 kb)

200 of H. petiolaris

1

H6' (or H1 or H2)

Barn HI

B1 (25.5 kb)

BI* (21.5 kb) B8* (3.8 kb)

all perennial ecotypes except 287 and 231

BI* (21.5 kb) B8** (3.9 kb)

594 o f H . anzonens~

1,2

H4b

B a m HI

B3a (9.2 kb)

B3a* (6.2 kb)

667- 331-671 of H. bolanderi

1

H3

Eco RI

E2 (9.5 kb)

E2* (7.5 kb)

667-331-671 of H. bolanderi

1

H3

Sst I

$2 (16 kb)

$2" (13.5 kb)

519 of H. annuus

1

H7

B a m HI

B19 (0.85 kb)

B19* (0.87 kb)

all perennial ecotypes except 287

2

H7

B a m HI

B14 (1.85 kb)

B14* (1.95 kb)

594 of H. arizonensis

2

* Restriction fragment site polymorphism ** Restriction fragment length polymorphism 1Reference pattern concerns the three sunflowers HA89B, HA89A and Bolero.

detected (Fig. 1). These variations were randomly scattered over the entire chloroplast genome, showing no evidence for mutational hot spots. The comparison of several ecotypes of the same species (namely those of H. annuus, H. bolanderi, H. petiolaris, H. praecox, H. angustifolius, H. californicus, H. decapetalus, H. giganteus, H. maximiliani, H. microcephalus, H. nuttallii, H. silphioides and H. tuberosus: see Table 1) demonstrates that most of the structural variations observed for the chloroplast D N A cannot be considered as species-specific markers in the genus Helianthus. For instance, the additional Hind III site on the Hind III fragment H4a occurs in only some ecotypes of the annual species H. annuus, H. petiolaris and H. praecox.

Amongst all of the variations detected in this study, the only modifications which could be species-specific markers are the Bam HI site located on the Hind III fragment H8 and the Eco RI site located on the Hind III fragment H5. Indeed, these sites occur in three ecotypes of the species H. bolanderi (ecotypes 667, 671, 331). However, the ecotype 255 ofH. bolanderi is lacking in these modifications, but this fact could be related to the suspicion that this ecotype might be a natural hybrid bearing an unidentified cytoplasm (H. Serieys, personal communication). The perennial species, on the other hand, show physical modifications (the Barn HI site located on the Bam HI fragment B 1 and the insertion/ deletion on the Bam HI fragment B 19) which dif-

275 ferentiate them from all annual species analysed here, except for the Russian ecotype 287 of H. tomentosus and the ecotype 231 of H. occidentalis (Table 2).

S1 nuclease mapping of DNA-DNA heteroduplexes We concentrated our S 1 nuclease mapping analysis on one of the borders of the 22.5 kb inversion which differentiates the organization of the chloroplast genomes of the Compositae relative to the basic arrangement found in tobacco [ 8, 9, 10 ]. Three wild-type species (H. petiolaris 737; H. tuberosus 289 and H. decapetalus 551) were compared to the Bolero F1 hybrid of cultivated sunflower bearing the CMS cytoplasm originally obtained by Leclercq [ 11 ], using H. petiolaris as a cytoplasm donor female parent.

Bolero/H. petiolaris (737) heteroduplex

Fig. 3. Sl nuclease analysis of the cpDNA H7 region for Bolero/737 (H.petiolaris) and Bolero/289 (H. tuberosus) heteroduplexes and Bolero/Bolero, 737/737 and 289/289 homoduplexes. After S 1 nuclease treatment, the DNAs were digested with Hind III prior to hybridization with the H7 probe. The bands pointed with a black square correspond to

Figure 3 illustrates the results obtained for Bolero/H. petiolaris 737 heteroduplex: the DNA fragments produced by the sequential digestions with S1 nuclease and with Hind III restriction enzyme were hybridized with the Hind III fragment H7. Two fragments of 3.8 and 2.7 kb were detected (lane 'Bol/737'), in addition to the 6.5 kb fragments displayed by the Bolero/Bolero and 737/737 homoduplexes (lanes 'Bol/Bol' and '737/737'): therefore, one can conclude that an S 1 nuclease-sensitive region is located on the heteroduplex Hind III fragment H7, positioned at 3.8 kb from one end and at 2.7 kb from the other end of this fragment (see map Fig. 4E). This sensitive region was further localized on the Barn HI fragment B 14, by digesting the same heteroduplex with S 1 nuclease and Barn HI, and hybridizing the tilters with the Barn HI subfragments B14 and B19 of the Hind III fragment H7: no new fragment was detected by the B19 partial S 1 nuclease digestions on the heteroduplex Bolero/289. In the lower part, the restriction map of the Hind III fragment H7 is shown.

276

Fig. 4. S 1 nuclease mapping of the Bolero/737 (H. petiolaris), Bolero/289 (H. tuberosus) and Bolero/551 (H. decapetalus) heteroduplexes and corresponding homoduplexes in the cpDNA H7 region. Panels A, B, C and D display the results obtained with the four Barn HI subfragments of the Hind III fragment H7 when used as probes. Panel E shows the physical map of S1 nuclease-sensitive regions (arrowheads) deduced from all the experiments carried out. Barn HI, Hind III and Sst I restriction sites are symbolized by B, H and S respectively.

277 probe (Fig. 4B), while a signal migrating at 1.35 kb was detected with the B 14 probe in addition to the 1.85 kb reference of the homoduplexes (Fig. 4C); the 0.5 kb fragment, complementary to the 1.35 kb band, could not be detected in our experimental conditions. These data, together with those obtained with the Hind III digests, enabled us to position the S 1 sensitive region on the Bam HI fragment B 14, and to specify that it is 1.35 kb apart from the Barn HI site separating the Bam HI fragment B14 and the 1.4kb Bam HI-Hind III fragment as shown in Fig. 4E. Two other S1 nuclease-sensitive regions were also determined outside of the Hind III fragment H7, using the other Barn HI subfragments to probe S 1 nuclease and Barn HI or Sst I digests: - the first one, located on the Hind III fragment H5, was detected with the 2.45 kb Hind IIIBam HI fragment: a band migrating at 3.6 kb was detected on S1 nuclease and Barn HI digests of the heteroduplex, in addition to the 5.6 kb Bam HI fragment B7 found in the homoduplex (Fig. 4A, lane 'Bol/737'); - the second sensitive region, located on the opposite side of the Hind III fragment H7 was determined using as probes the Hind III H7 and the 1.4 kb Barn HI-Hind III fragments: the H7 probe revealed two bands at 4.9 and 3.2 kb, while the Barn HI-Hind III probe hybridized a 3.2 kb long fragment on S 1 nuclease and Sst I digests (results not shown). The physical map representing these results is proposed in Fig. 4E.

Bolero/H. tuberosus (289) and Bolero/H. decapetalus (551) heteroduplexes The same experimental procedures were used for the S 1 mapping of the heteroduplexes formed by reannealing total D N A from Bolero in the presence of D N A from H. tuberosus (289) or from H. decapetalus (551). Identical hybridization patterns were obtained for the analyses of the two heteroduplexes (as an example, see Fig. 4C), indicating that within the limits of resolution of the method, the two wild-type perennial species

show the same physical organization of their chloroplast genomes in the studied region. Two S 1 nuclease-sensitive regions were clearly demonstrated: on the 2.45 kb Hind III-Bam HI fragment, by the formation of a fragment 1.7kb long (Fig. 4A, lane 'Bol/289' and map Fig. 4E); on the 1.85 kb Bam HI fragment B14, by the appearance of a 1.6 kb long fragment (Fig. 4c, lanes "Bol/551' and 'Bol/289', and map Fig. 4E). In addition, complementary analyses (patterns not shown) allowed us to demonstrate one S 1sensitive region on the Barn HI fragment B19 (very close to one of the Bam HI sites), and another one to the right of the Hind III fragment H7 (map Fig. 4E). In summary, our modification of the S 1 nuclease mapping technique enabled us to detect three S 1 nuclease-sensitive regions on Bolero/H. petiolaris 737 heteroduplexes, and four S 1 nucleasesensitive regions on Bolero/H. tuberosus 289 and Bolero/H. decapetalus 551 heteroduplexes, in the region of the Hind III fragment H7. To definitely confirm the results obtained with S 1 nuclease mapping, we compared the cultivated sunflower and the perennial wild-type species at the nucleotide sequence level. The S 1-sensitive region mapped on the Bam HI fragment B 14 was chosen for this purpose.

-

-

Sequence characterization of an S1 nuclease-sensitive region The Bam HI chloroplast D N A fragment B 14 was cloned from Bolero, H. tuberosus 289 and H. decapetalus 551. A region of 900 base pairs, carrying the potential S1 nuclease sensitive region, was sequenced for each of the three species. The comparison of these sequences shows several differences between the three ecotypes (Fig. 5): a point mutation substitution in position 202: in sunflower, there is an adenine at this position, altered to a guanine in H. tuberosus and in H. decapetalus ; -

278

Bolero

...... 1

66~TCCAA6~T~UTTACTT~AA6~T~CTT8~TTTAC~A~AAT~66ATTT6ACC~CTT~AAT~TATT8CSCCCTACGT~TA~T6TC~ATT6~C5~C~

..101

~A~A6~A~6~AATAAAAGAA~A~A6AA6AC6~AAATATA~ATT~s~TATATCC~CS~S~TAA~sTTAA~TTAATAAAT~CAAC~TTCAT~ETTTT~T

..201 `~T`~TC~T~TC~T~`~T`~TTc~.~C~TT~T~T~T~T~TN~``~`Cf``~T~T~T~Tc~T`~T`T`~````` H. tuberosus (289) ~ ~ T~'~ H. decapetalus (551) ATAMATAACC6TTATRC~ ! ! ..301

t~' ~'CC'.~t~'''"~I'~6C ~'~'''~" ' ' T.~'"T ' ~ ' '~ " C ~ ' " ' C T ' TTTT,,C,,, TCC,,~C C, , , , ,C,, ,,CT T, , , , , ,A,, TC,, TA,CC,

..401

"~T'~~,TT`T~c~`'r~'~c'~'r`~``~'T~``T~````~`~``~`~````~T~`'~`T~`c~```c~T`~````T~T~ ~~T,,T,,,,,,,,A,C~

H. tuborosus (289)

•.701

~TT6~cTT~AAT~C1~TTTf~A6~T~[C~tT66TA6A~ccCCT~T~66~Aa6~6~t6Cf~[~6~A[CACTc~TAtTCtTC~[T

..801

AT.~T~.T~C`~C8A6A~CATTTs~AAAAA~6ATA~6~8TA6~TATATC~AA~AA~AAACAATTTT6~TA6-~AAA6A~

Fi~. 5. Nucleotide sequence comparison of the first900 base pairs of the Barn HI chloroplast ]Z)NA fragment BI4 from Bolero, H. t~berosus (289) and H. dec~pet~l~s (551): evidence for a perfect 21 base pair duplication in the two wild species. Concerning H. tuberosus and H. decapetalus, only the nucleotides which differ from the Bolero sequence are indicated. The numbering is according to the Bolero sequence. The boxed sequence, starting with an arrow, corresponds to the 5' end of an open reading frame.

-

-

a duplication of 21 base pairs: in sunflower the sequence spanning from positions 235 to 255 is unique while it is duplicated as a perfect direct repeat in the two wild-type species; a point mutation addition/deletion in position 530: an array of 11 thymines occurring in Bolero and in H. decapetalus is increased to 12 thymine residues in the case ofH. tuberosus.

D i s c u s s i o n

Variability detected by restriction mapping

Three 6 base pair recognition enzymes producing a total of 71 cuts were used in this restriction analysis; 7 site modifications and 3 insertions/ deletions were detected among 40 ecotypes of 36 different species of the genus Helianthus. Our data are not numerous enough to assess the phylogenetic relationships between these ecotypes; in addition, the majority of the mutations detected cannot be related with the classification of the species proposed by Schilling and Heiser [19]. They allow however a rough estimation of the

sequence diversity in the genus. Taking into account the number of restriction site modifications detected on Barn HI, Hind III and Sst I digests of the 38 cpDNAs analysed (0 to 2 modifications out of the 71 sites mapped on Bolero cpDNA), one can calculate, according to Brown etal. [2], a sequence divergence ranging from 0.00 % to 0.46 %. This estimation of the variability is consistent with the 0.35% average value reported by Rieseberg et al. [ 17] for the species H. annuus and H. bolanderi. As noticed by these authors, these figures place the chloroplast DNA diversity of the genus Helianthus at an average position, between the high variability found in Brassica (0.3%-2.6%; [15]) or in Clarkia (0.2%-1.6%; [22]) and the low heterogeneity found in Lisianthus (0%-0.3%; [23]) or in Zea (0.03%-0.24%; [5]).

Variability detected by $1 mapping Nature of the Sl-sensitive regions The $1 mapping technique of DNA heteroduplexes was introduced by Shenk et al. [20] who

279 demonstrated that deletions of 32 base pairs in SV40 DNA were detected after S 1 nuclease treatment ofheteroduplexes formed between wild-type and mutant viral genomes. They also suggested that this technique should reveal and map single base changes. In fact, several investigations, performed with synthetic oligonucleotides, with cloned DNA fragments or with heteroduplexes produced between M 13 single-stranded DNA and genomic DNA [6, 21, 7, 3] showed that the mismatches have to be larger than 3 to 5 base pairs to be reproducibly digested by S 1 nuclease. Our own sequence analysis demonstrates that one of the S 1-sensitive regions recorded (the site located on the Bam HI fragment B 14) consists of a duplication of 21 base pairs; several point mutations or single base pair additions in the neighbourhood of this duplication were not detected. The modification of the S 1 mapping technique adapted here involves reannealing of total cellular DNA from two different ecotypes, in contrast to conditions used by others where low complexity DNA mixtures were reassociated (cloned fragments or viral genomes) or where M13 cloned single-stranded fragments in excess were hybridized against total genomic DNA. Owing to the reiteration of the chloroplast DNA sequences, our experimental conditions provided satisfactory hybridization signals with the chloroplast probes after S 1 nuclease treatments. We presuppose that this kind of experiment might be performed with any heterologous chloroplast DNA probe. Independently of the technical details, S1 nuclease mapping reveals additions and deletions but not single-base changes. On the contrary, restriction mapping is generally admitted to detect primarily point mutations affecting the restriction sites [2, 13]. This assumption is however a first approximation since small additions/deletions could affect restriction sites in the same manner as point mutations. Therefore, the combination of these two kinds of analyses provides representative data of the different mutation types which occurred during the recent evolution of chloroplast DNA between closely related species.

High frequency of small addition~deletion events The region corresponding to the Hind III fragment H7 of the chloroplast genome was analysed in the species H. petiolaris 737, H. tuberosus 289 and H. decapetalus 551 by a combination of the two mapping techniques. S1 mapping showed variations, such as the 21 bp duplication located on the Bam HI fragment B 14 in perennial species, which escaped our restriction analysis; other mutations, for example the size increase affecting the Barn HI fragment B19 of perennial species, were detected with both methods, on account of favourable combinations of enzyme, size variation and size of the fragment affected. In all cases, however, the number of modifications detected by the S1 nuclease was considerably greater than that detected by restriction mapping. This was confirmed by an example of intraspecific analysis performed by compairing the chloroplast genomes from H. petiolaris 737 and from the Bolero cultivar: Bolero carries indeed a CMS cytoplasm originating from the species H. petiolaris, associated with a pure nuclear background from 11. annuus [ 11 ]. While only one restriction site among the 71 analysed (Table 2) was found to vary between the two cytoplasms, the S1 mapping procedure could detect 18 different sensitive regions along the heteroduplexes, when probed with our complete collection of cloned sunflower chloroplast fragments (data not shown). These results show that taxonomically close cytoplasms may be differentiated by the S 1 nuclease analysis of their chloroplast genomes, due to the extensive exploration of the sequence performed by S 1 nuclease mapping: in particular, it should be possible to differentiate male-sterile cytoplasms on the basis of their plastome structure, although the chloroplast genomes are probably not directly involved in the mechanisms of male sterility. These results are also consistent with the observation made by Zurawski and Clegg [24], that small additions and deletions occur rapidly during the evolution of chloroplast genomes, in a repetitive manner at a limited number of possible positions. This observation indicates that the relative proportion of single-base mutations and of additions/deletions increases

280 with the phylogenetic distance between the plants c o n s i d e r e d . It t h u s s e e m s p r o b a b l e t h a t a g o o d estimation of the phylogenetic relationships between closely related taxa should be obtained by taking into account both point mutations and small additions/deletions.

P: Physical map and gene localization on sunflower (Helianthus annuus) chloroplast DNA: evidence for an

9.

10.

Acknowledgements 11. W e a r e v e r y g r a t e f u l t o D r H . S e r i e y s for p r o v i d i n g u s t h e Helianthus s e e d s a n d w i l d p l a n t s . W e s h o u l d a l s o like t o t h a n k D r s P. S a v a t i e r a n d Y. C h e b l o u n e f o r s t i m u l a t i n g a n d h e l p f u l d i s c u s s i o n s c o n c e r n i n g d i f f e r e n t a s p e c t s o f this work. This work was supported by a grant from the CNRS-INRA (ATP 'Biologie Molgculaire V r g r t a l e ' ) a n d a grant f r o m t h e M R T ( A l P ' A m r l i o r a t i o n d u T o u r n e s o l ' ) t o D r A . Bervill6 to whom we also address our acknowledgements.

12. 13. 14. 15. 16. 17.

References 1. Bolivar F, Rodriguez PJG, Betlach MC, Heynecker HL, Boyer HW, Crosa JH, Falkow S: Construction and characterization of new cloning vehicles. II. A multipurpose system cloning. Gene 2:95-113 (1977). 2. Brown WM, George M Jr, Wilson AC: Rapid evolution of animal mitochondrial DNA. Proc Natl Acad Sci USA 76:1967-1971 (1979). 3. Chebloune Y, Trabuchet G, Poncet D, Cohen-Solal M, Faure C, Verdier G, Nigon VM: A new method for detection of small modifications in genomic DNA, applied to the human Delta-Beta globin cluster. Eur J Biochem 142:473-480 (1984). 4. Dente L, Cesareni G, Cortese R: pEMBL: a new family of single stranded plasmids. Nucl Acids Res 11: 1645-1655 (1983). 5. Doebley J, Renfroe W, Blanton A: Restriction site variation in the Zea chloroplast genome. Genetics 117: 139-147 (1987). 6. Dodgson JB, Wells RD: Action of single-strand specific nucleases on model DNA heteroduplexes of defined size and sequence. Biochemistry 16:2374-2379 (1977). 7. Gannon F, Jeltsch JM, Perrin F: A detailed comparison of the 5'-end of the ovalbumin gene cloned from chicken oviduct and erythrocyte DNA. Nucl Acids Res 8: 4405-4421 (1980). 8. Heyraud F, Serror P, Kuntz M, Steinmetz A, Heizmann

18. 19. 20.

21. 22. 23.

24.

inversion of a 23.5-kbp segment in the large single copy region. Plant Mol Biol 9:485-496 (1987). Jansen RK, Palmer JD: Chloroplast DNA from lettuce and Barnadesia (Asteraceae): structure, gene localization and characterization of a large inversion. C'urr Genet 11:553-564 (1987). Jansen RK, Palmer JD: A chloroplast DNA inversion marks an ancient evolutionary split in the sunflower family (Asteraceae). Proc Natl Acad Sci USA 84: 5818-5822 (1987). Leelercq P: Une st6rilit6 mgde cytoplasmique chez le tournesol. Ann Amrlior Plantes 19:99-106 (1969). Messing JR: New M13 vectors for cloning. Meth Enzymol 101:20-78 (1983). Nei M, Li W-H: Mathematical model for studing genetic variation in terms of restriction endonucleases. Proc Natl Acad Sci USA 76:5269-5273 (1979). Palmer JD, Thompson WF: Chloroplast DNA rearrangements are more frequent when a large inverted repeat sequence is lost. Cell 29:537-550 (1982). Palmer JD, Shields CR, Cohen DB, Orton TJ: Chloroplast DNA evolution and the origin of amphidiploid Brassica species. Theor Appl Genet 65:181-189 (1983). Palmer JD: Chloroplast DNA evolution and biosystematic uses of chloroplast DNA variation. Am Nat 130: 6-29 (1987). Rieseberg LH, Soltis DE, Palmer JD: A molecular reexamination of introgression between Helianthus annuus and H. bolanderi (Compositae). Evolution 42: 227-238 (1988). Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain terminating inhibitors. Proc Natl Acad Sci USA 74:5463-5467 (1977). Schilling EE, Heiser CB: Infrageneric classification of Helianthus (Compositae). Taxon 30:393-399 (1981). Shenk TE, Rhodes C, Rigby PWJ, Berg P: Biochemical method for mapping mutational alterations in DNA with S 1 nuclease: the location of deletions and temperaturesensitive mutations in Simian Virus 40. Proc Natl Acad Sci USA 72:989-993 (1975). Silber JR, Loeb LA: S1 nuelease does not cleave DNA at single-base mis-matches. Biochim Biophys Acta 656: 256-264 (1981). Sytsma KJ, Gottlieb LD: Chloroplast DNA evolution and phylogenetic relationships in Clarkia sect. Peripetasma (Onagraceae). Evolution 40:1248-1261 (1986). Sytsma KJ, Schall BA: Phylogenetics of the Lisianthus skinneri (Gentianaceae) species complex in Panama utilizing DNA restriction fragment analysis. Evolution 39:594-608 (1985). Zurawski G, Clegg MT: Evolution of higher-plant chloroplast DNA-encoded genes: implications for structurefunction and phylogenetic studies. Ann Rev Plant Physiol 38:391-418 (1987).