The mitochondrial genomes of spontaneous orir petite mutants of yeast have rearranged repeat units organized as inverted tandem dimers

61

Gene, 24 (1983) 61-71 Elsevier

GEN 008 17

The mitochondrial genomes of spontaneous units organized as inverted tandem dimers (Saccharomyces

Godeleine

cerevisiae;

Faugeron-Fonty,

excision

April

(Revision

received

(Accepted

recombination;

genome

rearrangements)

Marguerite Mangin, Alain Huyard and Giorgio Bernardi

Laboratoire de GPnktique Moltkdaire, or (I) 336-25-25, ext. 41.01 (Received

sequences;

01-i’ petite mutants of yeast haye rearranged repeat

Institut Jacques Monad, 2 Place Jussieu, 75005 Paris (France) Tel. (I) 329 -58-24

lst, 1983) May 12th, 1983)

May 16th. 1983)

SUMMARY

We have investigated the structure and organization of the mitochondrial genomes of two related ori’ (or-i-rearranged) spontaneous petite mutants of Saccharomyces cereuisiae. In these mutant genomes every repeat unit contains an inverted terminal duplication harboring a second (inverted) ori sequence, and tandem pairs of repeat units alternate with tandem pairs in inverted orientation. We have shown that ori’ genomes are organized as the genomes with inverted repeat units of ethidium petites, and we have clarified the mechanism by which such mutant mitochondrial

INTRODUCTION

Investigations carried out in our laboratory (Bernardi et al., 1970; 1976; 1978; 1980; 1982; Faugeron-Fonty et al., 1979; Marotta et al., 1982; the de Zamaroczy et al., 1983) have clarified molecular basis of the cytoplasmic spontaneous “petite” mutation in the yeast S. cerevisiae. This is

Abbreviations: origin

bp, base pairs;

of replication;

ori+,

ori-,

EtBr,

ethidium

orio, our,

bromide;

ori,

on”, see INTRO-

DUCTION.

0378-l 119/83/$03.00

0 1983 Elsevier Science Publishers

B.V.

bromide (EtBr)-induced genomes arise.

a case of extreme genomic instability, in which segments of the mitochondrial genome units of wild-type cells are excised and tandemly amplified to form the defective genome units of the petite mutants. Excision takes place by an internal recombination process between direct nucleotide repeats present in the abundant noncoding sequences of the wild-type genome units. The high number of potential excision sequences in such sequences accounts for the extremely high rate (about 1% per generation in most laboratory strains) of the spontaneous petite mutation (see Bernardi, 1979; 1982, for two brief reviews). The excised segments of wild-type genomes,

62

which become the repeat units of the mitochondrial genome

units

of spontaneous

petites,

MATERIALS

from different regions of the former and contain, therefore, different sequences where DNA replication starts. In most cases (ori+ petites), the origin of DNA replication is one of the seven canonical ori

of

sequences

Zamaroczy

et al.,

the

wild-type

wise, it may be a partially ori-

petites

surrogate

(Goursot case, the repeat units efficiency

ori

ori sequence

(ori”

1981)

in

or a

sequence)

in

et al., 1982); in the latter contain no intact or partial

sequence.

of replication

(de

1982). Other-

et al.,

of replication

ori‘-’ petites

canonical

deleted

(de Zamaroczy

origin

genome

1981; Bernardi,

In general, decreases

the

strong

but

indirect

in the order ori+

so far;

direct

bio-

chemical evidence has, however, been obtained in our laboratory (G. Baldacci and G. Bernardi, paper in preparation). In the present work, we have studied the rearranged structure and the complex organization of the mitochondrial genomes from yet another class of spontaneous petite mutants, the ori’ (ori rearranged) petites and we have clarified the mechanism

by

which

these

genomes

(a) Yeast strains The parental strain

ous papers, (1979)

S. cerevisiae strain

wild-type

D-243-2B-Rl strain

(called

A; see Faugeron-Fonty

for the properties

strains

were spontaneous

deficient

petite

et al.

of this strain). cytoplasmic

mutants,

derived

Petite

respiratory-

from

wild-type

strain A (see RESULTS, sections a, c). Growth and

culture

conditions

Faugeron-Fonty

was

here, as in our previ-

were

as

media

described

by

et al. (1979).

relative

to ori- to orP. It should be noted that the evidence for ori sequences being origins of DNA replication has been

AND METHODS

can derive

arise.

(b) Mitochondrial

DNA

Mitochondrial DNA was purified by centrifugation in CsCl density gradients, using a method modified from Lang et al. (1977). Restriction enzyme degradations, gel electrophoresis, nick translation of DNA probes, transfer of DNA to nitrocellulose films, and filter hybridization were performed essentially as described by Faugeron-Fonty et al. (1979). The primary structure of DNA was determined (1977).

according

to

Maxam

and

Gilbert

Al-

though ori’ petites are rare among spontaneous petites, their mitochondrial genomes are of special interest because, as shown by these investigations, they are organized as the genomes with inverted repeat units of EtBr-induced petites. This is one of the two major classes of mitochondrial genomes found in EtBr-induced petites, the other consisting of genomes with tandemly arranged repeat units (Bos et al., 1980). Whereas the latter genomes are identical to those of ori+ spontaneous petites, the former have a more complex organization. The present detailed study of ori’ genomes provides, therefore, a new insight into the structure, organization and mechanism of formation of this class of mitochondrial genomes. Furthermore, investigations on subcloning and crosses of ori’ petites have shed light on the correlation between ori sequences and replication of the mitochondrial genome of yeast (see the accompanying paper by Mangin et al., 1983).

RESULTS

(a) The mitochondrial petite mutants

genomes of ori’ spontaneous

The mitochondrial genomes studied here were those of four spontaneous petite mutants. The ori+ petite a-15/3/2 and orir petite a-15/4/1 were derived by subcloning from a-15/3 and a-15/4, two petites endowed with heterogeneous mitochondrial genomes derived in turn from the same parental spontaneous petite a- 15 (Faugeron- Fonty other two petites, orir et al., 1979). The a-15/4/1/23 and ori+ a- 15/4/ l/ 1, were subclones of a-15/4/1 (Mangin et al., 1983) and will be denoted henceforth as a-23 and a- 1. respectively. Petites a-15/3/2 and a-15/4/1 had mitochondrial genomes formed by repeat units about equal in length, which were derived from the same re-

63

gion

of the wild-type

however,

genome;

a suppressivity

the latter

had

the former

of about

a suppressivity

a-15/3/2

had,

SO%, whereas

of about

5% (see

genome

showed

formed

by tandem

taining

an ori sequence

(Fig.

repeat

1) that

this was

units 4200 bp long con(Fig. l), which was local-

Mangin et al., 1983). As a first approximation, suppressivity can be operationally defined as the

ized and oriented on the wild-type genome map, ori (de Zamaroczy et al., 1981). Its left end was

percentage

located

of petites

found

cross with a wild-type pressivity

is basically

tion efficiency

Mangin

o

we know

determined

genome

et al.,

of a

(Bernardi

rightmost

HhaI

parental

1982;

mapping

Tzagoloff,

these

of the

HincII

site and from the

site of the genome

wild-type

H&c11

1980). Its right

1850 bp and 3250 bp, respec-

tively, from a landmark

to that

et al.,

and

end was at about

et al., 1980; de

198 1; Goursot

et al., 1983). Restriction

relative

160 bp after the end of the oxi

at about

gene (Thalenfeld

that sup-

by the replica-

of the petite genome

of the wild-type Zamaroczy

in the progeny

strain;

and

segment

strain

A (Fig.

HhaI

sites correspond

of the

1). Incidentally, to the

I S/3/2

a-23

3OObp a-23

a-l

OH1nf

l Taq 0

Fig.

circular

II

OHha

I

V Hpall

OEco

RI

~Alu

I

A

Hae

III

mMbol

oMbo

II

l

Ava

II

BXbol

@Hind

maps

III

of the repeat

units of four petite

1982) is derived from our results and also from data of Thalenfeld

(Newman

units of petite genomes. This fragment

The HpaII

numbering

sites investigated

(upper designations)

corresponds a-15/4/1

RESULTS, section

c. The deleted

of the repeat

(double-thickness

lines).

to the clockwise

Arrows

and Hue111 (lower designations)

and a-23

repeat

unit of a-l

units are indicated

fragments

is represented

by letters;

and proline

(2120 bp) fragments

indicate

inverted

referred

segments

are numbered

are presented

of

of the

(1980) for its left

to the methionine

(1400 bp) and HaeIII

have been mapped.

of the genome orientation

and Tzagoloff

sites correspond

to that of Fig. 2. Both types of repeat units of a-23

c); some sites on both segment

corresponds

HhaI

et al., 1980). The lengths (in bp) of the HpaII

RESULTS, section

underlined

and the rightmost

I

from the ori5 region

map; see Bernardi,

et al. (1982) for its right end. The HinclI

direction

derived

genome

of a-15/4/1.

(the left-to-right

genomes

map of this region

to in RESULTS, section a, are given. Not all restriction repeat

mitochondrial

A. The restriction

genes, respectively

qHph

@Hlnc

strain

end and of Martin tRNA

I

Thai

1. Restriction

wild-type

I

in the

in the case

(see Fig. 7 and

these sites are referred

by a gap (see Fig. 8). The ori5 sequences

to in are

64

genes

of methionine

tively (Newman established phism

and proline

tRNAs,

respec-

different yeast strains (Bernardi et al., 1975; Prune11 et al., 1977), the map distances given above

et al., 1980). In spite of the well-

restriction

fragment

of the mitochondrial

polymor-

harbored

A

and deduced from work on other strains apply to the case of strain A since restriction maps of the

by

Hoe

HP0 a-i%/1

length

genomes

HP0

A O-IS/
o-is&/~

Hae

A

A O-IS&/I

(2)

( 297 I (A,9 1 (6)

Is ) (I 1

(1) (5)

Fig. 2. (A) Electrophoretic petite mutant

a-15/4/1

pattern

pairs of identical

fragments

DNA

(1230-bp

fragments

inverted

duplication

mitochondrial fragments

as petite fragments, indicated

from

wild-type

gel of the HaeIII genome

(3,lO and 4,9 of Fig. I), another HpaII and 2500-bp

of a-15/4/1

DNA

coincide

on 2% agarose

and its parental

absent

a-15/4/1

HaeIII)

from

to nitrocellulose

with the exception

of a 1400-bp

transfers

wild-type

genome

from

the gel shown

to wild-type HpaII fragment

in Fig. 1, have their right ends in a region not present

to parental

fragments

bands

one (360 bp) to two comigrating

do not correspond

the parental

with those seen in (A). Hybridization

and HpaII restriction

A. Two petite DNA HpaII

fragments

fragments,

(see Fig.

DNA fragments and a 2120-bp

of mitochondrial

1). (B) Hybridization

in panel HaeIII

from to two

(2 and 7 of Fig. 1). Two petite

since they encompass (A). Hybridization

takes place on fragments

in the petite repeat unit). Faintly

DNAs

(770 and 280 bp) correspond

fragment hybridizing

bands

having

(asterisks; fragments

the terminal

of 72P-labelled on petite

the same size

these fragments. are not visible.

65

region

consideration

are extremely

similar.

number

To sum up, petite a-15/3/2

is a “normal”

sponta-

a-15/4/1

neous

under

ori+ petite,

tandem

fashion,

with repeat each

units

arranged

containing

in a

ori

one

overall

fragments

is, therefore, length

se-

the HpuII

of

exception

In contrast, a-15/4/1

the organization

was complex,

vestigations

involving

number

periments obtained

as shown restriction

of enzymes

with

of the repeat

by detailed mapping

and

appropriate

can be summarized

pattern

of the genome

fragments

hybridization

probes.

l-7)

with a

The

showed

wild-type

unit

duplication

7

sequence;

comigrated

with

genome A, with the

fragment with

(fragment

8 of

fragments

its right-end

(comprising

site; (ii) contained, ori

in Fig. 2A,

the left side (HpuII

9 and 10). This duplication

results

as follows. The HpaII

unit of a-15/4/1

to an

of a 1230-bp

of the repeat

terminal

ex-

unit of

of a-15/4/1

Fig. 1) which joins

in-

in the repeat

10 and corresponds

of 4450 bp. As shown

those of the parental

quence.

large

of HpuII

side

HpaII

fragments

(i) followed

the HincII

as mentioned

above,

and (iii) had an inverted

a second

orientation

bands (Fig. 2A). Of these, 5 corresponded to 5 fragments (numbered l-5 in Fig. 1 and 1810 bp in

wise, of the 5 Hue111 fragments

overall length), identical with those forming the right half of a-15/3/2. One HpaII band (360 bp) could be shown to correspond to two comigrating

2A), 3 (l-3 of Fig. 1) were identical to those forming the right half of a-15/3/2; 4 (l-4 of Fig. 1) were identical to those of the parental wild-type

fragments (2 and 7 of Fig. 1); 2 other HpaII bands (770 and 280 bp) corresponded to 2 doublets of

genome, and 1 (2500-bp fragment 5 of Fig. 1) was different from any of the wild-type fragments; this fragment encompassed the terminal inverted du-

relative

identical fragments (3,lO and 4,9, respectively; Fig. 1). Since fragments 3 and 4 flank the HpaII site of GC cluster B (de Zamaroczy et al., 1981) of ori the repeat unit of a-15/4/1 must contain a duplication

harboring

a second ori

o-23.--_ -^

sequence.

of a-15/4/1

(Fig.

plication of a-15/4/1, which obviously does not exist on the parental wild-type genome. Hybridization experiments using labeled mitochondrial

The total

o-23 _--_-

-0-15/4/l -..-

to the first one, as shown in Fig. 1. Like-

DNAs

from

a-15/4/1

(Fig.

2B) and

0 -15/4/l ___I,.-

bp

bP

? 630 25m

2 630 ..__ 2500--

1 330 I200-

----

9 00 -.770-

265 230

:

280--2.50 215 I 80

(A) Fig. 3. (A) Electrophoretic petite a-15/4/1 as an

pattern

and its subclone

ori probe, to restriction

very faint.

(B) on 2% agarose a-23.

fragments

gel of the Hoe111 and

(B) Hybridization

pattern

of (A). The hybridization

Hpcr11restriction

of “P-1abelled

fragments

mitochondrial

band corresponding

of mitochondrial

DNAs

DNA from petite a-I/IR/Zl,

to the 280-bp

HpaII

fragment

of a-15/4/1

from used is

66

from a-l/lR/Zl (Fig. 3; the latter contains more than oril; Gaillard and Bernardi, showed

the following.

hybridized fragments fragments bridizing

The a-15/4/1

as expected of a- 15/4/l

1400-bp

HpaII

fragment

instead

ments

to all HpaII

of petite

(whose lengths

genome

are not visible

the latter case, however, fragment

1979)

DNA probe, and Hoe111

and to the corresponding

of the wild-type fragments

barely

(faintly

the probe hybridized and

a 2120-bp

these parental

in bp are specified

to a

HaeIII

of the 1230-bp and 2500-bp DNA;

hy-

in Fig. 2B); in

INVERSION

frag-

fragments

in Fig. 1) have

their right ends in a region not represented in the petite repeat unit. The a-l/lR/Zl probe indicated that each one of the two largest Hue111 fragments (3 and 5) contained an ori sequence and I

that the same was true for each one of the two pairs of identical HpaII fragments (3,10, 4,9; Fig.

EXCISION

3).

1000 -

3’ 1

(b) The mechanism of the ori’ rearrangement

Fig 4. Possible mechanism of the mitochondrial

The mechanism of formation of the repeat units of a-15/4/1 can be reconstructed as follows (Fig. 4): (i) first, a segment is excised by an internal recombination process between two direct repeats located to the right of the EcoRI and HincII sites, respectively, of the wild-type genome (Fig. 4, a-b); the latter sequence may coincide with sequence L (see below); (ii) this segment is tandemly amplified (Fig. 4b); (iii) one of the repeat units of the

mitochondrial

genome

region. Numbers site clusters; leading

3’ correspond

of the a-15/4/1

sequences

horizontal

which

A in the orj5

arrows

Vertical is excised

HaeIII-Hpcr II

process.

1, 2, 3 sequences,

indicate

the orientation

lines delimit

to become

section b, for further

L

TATCTCCTTCGGGGfTCG*GTCCCCCTC

of ori5

se-

segment

unit of the resulting

sites are indicated

RESULTS,

I’. 2’.

respectively.

the genome

the repeat

Not all restriction

Steps

repeat unit. L and R

used in the inversion

broken

of the

(underlined)

sites are as in Fig. 1. (b-e)

to the inverted

petite genome.

of the repeat unit (a) Segment

and short vertical lines indicate

other restriction

are the inverted Short

of a-15/4/1.

of strain

to the formation

quences.

for the formation

genome

bp

in b-c.

See

details.

primary petite genome undergoes an inversion (as in a- 1; see below) through a crossing-over mechanism

acting

on two inverted

repeats,

L and

R,

located to the right of the HincII and of the ?%a1 sites, respectively (Fig. 4, c-d). The sequence of the L repeat, starting at 69 bp to the right of HincII, was found in data of D. Miller and N. Martin (personal commun.) and that of the R repeat, starting at the rightmost bp of the 7’huI site, in data of M. de Zamaroczy (personal commun.). Both sequences are presented in Fig. 5 and belong in the ori’ family (Goursot et al., 1982); (iv) an excision takes place (Fig. 4d) between two HpuII-Hue111 site clusters, 1 and 3’, which are two ori’ sequences (Goursot et al., 1982) originally located on two subsequent repeat units. Incidentally, sequence 3’ is a sequence also used in the excision of the repeat units of a-l (sequence A2

5’

R

3’

GAGGGGG-C z GAiCCCCGAAGGAGATA

61

ACTCCTTTGGFGT

El2

ACCCCAAAGGAGT

Al

TCdTTCGGGG-TT?C-GGC:CCCGT;GCltGGC%C&GGAACTiT

A2

A~AGTT?~GGG&CCGGC~ACGGG~GC?CGGAACCCCCGAAI~GGA

Fig. 5. Sequences

of inverted

repeats

L and R (see Fig. 4), Bl.

B2 and Al,A2

(see Fig. 8). Restriction

Vertical

lines

single-base

thick

deletions.

to be due to strain

indicate

Mismatches differences.

determined

by D. Miller

Zamaroczy

(R) (unpublished

sites are as in Fig. I.

mismatches, between

horizontal

The L and R sequences

and N. Martin results).

dashes

L and R are likely were

(L) and by M. de

of Fig. 5) and of a-15/3/2. excisions units

It should be noted that

encompassing

are

Zamaroczy

known

in

two other

subsequent petite

genomes

The

et al., 1983); (v) this excision

gradation 6).

The

amplification

of

repeat

with Hind1

produced

stoicheiometry of

(Fig.

units

three bands

of

gels.

(Fig.

A’

5 800

B’

4 OS0

6100

A

4450

B

2800

C

A : B : C was

of bands

by microdensitometry

EtBr-stained

a-15/4/1

leads to

does not occur in the usual tandem as shown by the following results. De-

1 : 2 : 1, as determined negatives

the

B

bp

(de

the formation of the repeat unit of a-l 5/4/l 4e), which is, therefore, a secondary petite. a-l 5/4/l fashion,

o-23

repeat

Band

of

B corre-

sponded to two fragments, whose size (4450 bp) was identical and equal to that of the repeat unit,

6’

2 300

whereas the sum of the sizes of bands A and C (6100 and 2800 bp) corresponded to two repeat units. As shown in Fig. 7, B fragments derive from cuts at HincII sites located on tandem repeat units, whereas A and C fragments derive from cuts at Hind1 sites located on contiguous inverted repeats. The results of Fig. 6 can be explained if the repeat units of a-15/4/1 are arranged in tandem

pairs,

which,

in turn,

are arranged

in an

inverted fashion (Fig. 7), an organization leading to an inverted orientation of all subsequent ori sequences. An alternative explanation of the stoicheiometry of HincII bands involving a very particular combination (possibly in different genome units) of simple inverted repeats with tandem repeats is not compatible with the fact that petite genome units are polydisperse in size, and is not supported by partial HincII digests (not shown) which only reveal the bands expected on the basis of the first model.

Fig. 6. Electrophoretic digests and

its subclone

wild-type

of further genome rearrange-

In the case of the mitochondrial a-23, a petite derived from a- 15/4/

genome of 1 by subclon-

ing, the repeat units were significantly different from those of the parental genome and belonged to two classes (Figs. 1, 3, 6 and 7). Class 1 (3950 bp long) lacked the two small Hue111 and HpaII fragments (1 and 2) forming the left end of the repeat unit of a-15/4/1 (Figs. 1 and 3). In class 2 (4210 bp long), the distance between the AuaII sites a,a’ and the terminal HaeIII-HpaII site clus-

a-23.

from petite

Hue111

experiment,

having

size of repeat

on 0.8% agarose

DNAs

fragments

B were used as molecular

experiments.

fragments

Bands

but other B and

a-15/4/1

of the DNA weight

markers

of in

were used in

B’ correspond

1 and 2 of a-23, correspond

gel of HincII

mutant

fragments

the size of the repeat

units

A+ C and A’+C’

to pairs

of

units and the average respectively.

to two repeat

The sums

units, of the two

classes in the case of a-23.

ters cc’ was larger by 130 bp than that between a,a’ and the nearest HueIII-HpuII site clusters b,b’ on the repeat unit of a-15/4/1 (Figs. 1 and 7); furthermore, the HinfI-TuqI site cluster at 25 bp from shorter

(c) The mechanism ments

strain

this particular other

pattern

of mitochondrial

b,b’ was missing,

and

the repeat

unit

on its left end than that of a-15/4/1.

was The

deletions undergone by repeat units 1 and 2 of a-23 relative to those of the parental petite a-15/4/1 can be accounted for by two excisions taking place on a circular tandem dimer of a-15/4/1 (represented in linearized form in Fig. 7), whereas no excision on a circular inverted dimer of a-15/4/1 can lead to the increased length of a’c’ relative to a’b’. On the basis of the HincII digestion patterns (Fig. 6), the overall arrangement of repeat units in the a-23 genome consisted, as proposed for a-15/4/1, of two tandem repeats followed by two tandem repeats in opposite orientation. Interest-

A

0

c

1

23

a -

-

B

2

Y

c-

--

-A’

1

.------

9

8’

A

B

2

t

1

t

C’

2

t

--

L_

B’

C

? c*

A’

1

__

B’

A

2

P

Lb

-

B

t La

C’

_

B’

A’

b?

(B)

a-

q

48 “,$

15/4/l

AA

I”

A

I

m

6

.

.

a -23 (1) Fig. 7. (A) Mitochondrial upper arrows;

ori

genome

sequences

organization

are indicated

Fig. 6) are denoted

by horizontal

employ

of Fig. 1. Small-case

the symbols

the two different

repeat

of petite mutants

a-15/4/1

by boxes, their orientation

bars. HincII

units present

the repeat units of a-15/4/1

(2)

sites are as in Fig. 1. (B) Restriction

letters indicate

in the genome

to repeat units

and a-23.

by small arrows;

some particular

of a-23;

1 and 2 of a-23

arrows

restriction indicate

are indicated

Orientation HincIl

of repeat

fragments

maps of the repeat sites (see

inverted

RESULTS,

repeats.

below the a-15/4/1

units is given by the

(A, B, C and A’, B’, C’; see units of a-15/4/1

and a-23

section c). (I) and (2) represent

The deletions restriction

supposed

maps (see

to lead from

RESULTS,

section

c).

ingly, the size of B’ fragments corresponded distance of two HincII sites located on units

arranged

in a tandem

fashion

to class 1 and class 2, respectively,

to the repeat

and belonging showing

class 1 and class 2 repeat units are arranged

that in the

(i) the repeat unit of a primary petite (a-15/4/1 a similar one) exhibits two pairs of inverted peats, Al,A2 and Bl,B2 (Fig. 8a); sequences and A2 belong in the oriS sequences, and B2 in the ori” family (Goursot

or reA1

sequences Bl et al., 1982);

order l-2- l-2 (Fig. 7) and not in the order 1-2-2-1. This was confirmed by the sizes of other restric-

these sequences A2 corresponds

are specified in Fig. 5; sequence to the right excision sequence of

tion fragments encompassing repeat units 1 and 2 (such as those produced by MboI, AuaII, HinfI, TaqI and HincII). As in the case of a-15/4/1, partial Hi&I digests were only compatible with the repeat unit arrangement shown in Fig. 7. Another subclone of a- 15/4/ 1, petite a- 1 (Fig. l), was an ori+ secondary petite exhibiting a genome formed by tandem repeat units which presented, however, a deletion and an inversion relative to the wild-type genome. Mapping and sequence analysis of the repeat unit of a-l of the corresponding region of the parental wild-type strain A made possible a precise reconstruction of the process leading to these repeat units (Fig. 8):

a-15/3/2 (de Zamaroczy et al., 1983); sequences Al, Bl and B2 correspond to nucleotides 465-508, 87-99 and 641-653, respectively, of the repeat unit of petite a-10/3 (Goursot et al., 1982); (ii) an inversion takes place between Bl and B2, and changes Al,A2 into two direct repeats (Fig. 8 b, c); (iii) an excision takes place by internal recombination between Al,A2 (Fig. 8d), leading to the formation of the repeat unit of a-l (Fig. 8 e, f). To sum up, the excision process generating the repeat unit of a-l is the classical one (de Zamaroczy et al., 1983) and the special features of the a-l repeat units are simply due to the fact

69

DISCUSSION

The

present

formation

t,

-

A

A

A2 INVERSION

mation

the mitochondrial are

A

A2

were

rare

of repeat

in-

of for-

units

from

of spontaneous

laboratory)

(a-15/4/1

the only

several hundred

1983). The

ori’

and petites

spontaneous

for reasons

and explained al.,

genome

novel

petite

of yeast which we call ori’. These petites

extremely

a-23 h

provide

the mechanism

and the organization

mutants

4

investigations

on the structure,

found

petites studied

among in our

which are now understood

in the following repeat

its subclone

units

paper

(Mangin

of orir petites

et are

characterized by a terminal inverted duplication harboring a second (inverted) ori sequence. The ori’ petites can only arise as secondary petites since the mechanism leading to the formation of the rearranged repeat units involves inversions occurring by internal recombination at inverted

d

EXCISION e

nucleotide

repeats (a point first demonstrated

for the mitochondrial located

on subsequent

genome repeat

here

of yeast) which are units

of a primary

petite genome. We have shown that the repeat units of orir genomes are arranged as tandem dimers in inverted orientation relative to their 200bp

Fig. 8. Possible steps leading of the mitochondrial end of the repeat

to the formation

genome

of petite

unit of a-15/4/1.

Fig. 1. Horizontal

arrows

Restriction

indicate

the relative

the two pairs of inverted

repeats

leading

of the a-l

to the formation

the hybrid

sequences

which involved tion repeat -

generated

Al,A2,

sites are indicated unit of a-

1 correspond

Al,A2

(a) Left

sites are as in orientations

repeat

processes

Not all restricform

of the

and the deleted segments

BI-Al,

and Al-B2

petite shown in (a). See

of Steps

unit. A and B are

by the recombination

in (b, d). (f) Linearized

to segments

a-l.

and Bl,BZ. (b-e)

and Bl,BZ, respectively.

1. The inverted

unit of the primary for further

of the repeat unit

mutant

of a

from the repeat

RESULTS,

section c,

details.

that excision occurs on a petite genome which is already rearranged, in that a DNA segment has been inverted. Since excision involves, on one side, a sequence located on the inverted segment, the final result is that the repeat unit of a-l presents a deletion beside an inverted segment.

nearest neighbors. The amplification mechanism leading to such arrangement cannot be a simple rolling circle mechanism acting on a circular monomer, as in the case of the tandemly arranged repeat units of ori+ petites, but must be a more complex process. Some important points of the latter have been clarified by this work. First of all, the excisions leading to the formation of the two different repeat units of a-23 from those of a- 15/4/l appear to have taken place on tandem dimers of the latter; this indicates that tandem dimers and not monomers or inverted dimers are the initial templates for DNA amplification. Such tandem dimers can only be amplified into inverted tandem dimers. We will not speculate here on the mechanism underlying this process, but we suggest that it is linked with the presence of terminal inverted duplications containing a second inverted ori sequence. This is indicated by the following observations: (i) the removal of the inverted duplication systematically leads to the formation of tandemly amplified ori+ genomes (Mangin et al., 1983); and (ii) the presence of a terminal, nonduplicated inversion not containing an ori se-

70

quence,

as found

conducive

in the genome

to the repeat

genomes, genomes.

but

to the usual

mutation.

tandem

on ori’ genomes

The results

to some new insights Indeed,

with enzymes

of a-

1, is not of ori’

unit arrangement one

just described

ACKNOWLEDGEMENTS

obtained

part of the repeat units

We thank Dr. Nancy Zamaroczy for providing

of a-15/4/1

are identical

quence

(1980) verted

and

by Heyting for

the

repeat

which

a-23,

cut asymmetrito those

et al. (1979) and Bos et al.

mitochondrial

units

genomes

of induced

petites.

with

in-

stoicheiometry

of the HincII bands

bands

studied

by

that such is also the arrangement of repeat units in the induced petite genomes of Bos et al. (1980). A difference between the latter and the orir genomes is that most of the former do not seem, from the limited mapping data available, to contain any canonical ori sequence. We suggest that the role played by such sequences in the amplification of orir genomes is played in those EtBr-induced genomes by the ubiquitous surrogate origins of replication, the ori’ sequences (Goursot et al., 1982).

induced

petite

points lead to some more general about the mitochondrial genome of mutants.

We

Baldacci

for useful discus-

Breton for the artwork

Brient for typing

de se-

and Martine

this manuscript.

REFERENCES

of Fig. 6, or of

from the genomes

Bos et al. (1980), indicates the regular organization of repeat units shown in Fig. 7, and we suggest

The above considerations

data, Giuseppe

sions, Philippe

Martin and Miklos us with unpublished

These pat-

terns were erroneously interpreted by Bos et al. ( 1980) as indicating “random mixed” direct and inverted repeats in those genomes. The 1 : 2 : 1 the equivalent

et al., 1983) account

lead petite

cally in the nonduplicated reported

like HincII,

patterns

paper (Mangin

of ori+

on the EtBr-induced

the restriction

the following

for this difference.

already

know

(de

Zamaroczy et al., 1983) that direct nucleotide repeats, almost always located in the AT spacers and GC clusters of intergenic sequences, are used in the excision of both spontaneous genomes and of induced petite genomes with tandemly arranged repeat units. In other words, no difference can be seen between this class of induced petite genomes and spontaneous petite genomes. Now we see that, although very rare, rearranged genomes of spontaneous petites exist which are identical in the structure and organization of repeat units to the genomes with inverted repeats of EtBr-induced petites. Then the main difference between spontaneous and induced petites consists of the different frequency of such rearranged genomes in these two classes of mutants. Investigations presented in

Bernardi,

G.: The petite

mutation

in yeast.

Trends

Biochem.

Sci. 4 (1979) 197-201. Bernardi,

G.: The origins

genome Bernardi,

of replication

of yeast. Trends G., Baldacci,

Gaillard,

B&hem.

G., Bernardi,

C., Goursot,

sion

origins

land, Amsterdam, G..

G., Colin,

Goursot,

R., Goursot,

Mangin,

M.,

(Eds.),

Bernardi,

A.O.M.

F., Fonty,

Bernardi.

G., Faures,

Mitochondrial plasmic

of yeast.

of Genes

R. and Prunell.

mutation

in Saccharomyc~e.~ B.L.

and

and

Genetics

Stoppani, of

Yeast.

1978, pp. 241-254.

M., Piperno,

G. and

Slonimski,

from respiratory-sufficient

respiratory-deficient

in

1982, pp. 279-289.

M., Horecker,

DNA’s

genome

The

W.J. and Pateman,

Expression

Biochemistry

Press, New York,

C.,

M.:

G., Goursot.

“petite”

in Bacila, (Eds.),

Academic

and

Press, Sydney,

A.: The cytoplasmic cerevisiae,

de Zamaroczy,

A.W., Peacock,

Manipulation

G..

A., Le Van Kim,

of the mitochondrial

Academic

G., Culard,

Y.. Faugeron-Fonty,

R. and

P., Linnane,

Eukaryotes.

and Expres-

Elsevier/North-Hol-

R.. Huyard.

Marotta,

origins of replication J.A.

excision in Kroon,

1980, pp. 21-31.

Baldacci,

in Nagley,

mutation:

and suppressivity, Genome.

G.,

M., Marotta,

C. (Eds.), The Organization

of the Mitochondrial

Bernardi,

A., Mangin.

M.: The petite

replication

A.M. and Saccone,

G., Faugeron-Fonty.

R., Huyard,

R. and de Zamaroczy, sequences,

of the mitochondrial Sci. 7 (1982) 404-408.

mutant

P.P.:

and cyto-

yeast. J. Mol. Biol. 48

(1970) 23-42. Bernardi,

G., Prunell,

A., Fonty,

F.: The mitochondrial tion and the petite

mutation,

A.M. (Eds.). The Genetic Elsevier Biomedical Bernardi,

G., Prunell,

in Saccone.

Function

H. and Strauss.

A. and Kopecka,

genome

Puiseux-Dao,

S. (Ed.),

Relationships.

C. and

of Mitochondrial

Press, Amsterdam,

mitochondrial plasmic

G.. Kopecka,

genome of yeast: organization,

Elsevier.

DNA.

1976, pp. 185- 198. H.: An analysis

of yeast with restriction Molecular

evoluKroon.

Biology

of the

enzymes.

in

of Nucleocyto-

Amsterdam.

1975.

pp.

85-90. Bos. J.L.. Heyting,

C., Van der Horst,

G. and Borst,

P.: The

71

organization

of repeating

yeast petite mutants. de Zamaroczy, Excision

units in mitochondrial

Curr. Genet.

M., Faugeron-Fonty, sequences

DNA from

G.

and

in the mitochondrial

Bernardi,

genome

G.:

of yeast.

M., Marotta.

R., Mangin,

M.. Baldacci.

of replication phenomenon

mitochondrial

of suppressivity.

Nature

G., Culard,

wild-type

yeast

“petite”

mutation.

Gaillard.

mitochondrial

genome

yeast. Mol. Gen. Genet. Goursot,

R., Mangin.

replication mutants Heyting,

of

the

of a spontaneous

sequence

“petite”

M. and Bernardi,

G.: Surrogate

Biochem.

Weijers,

P. and

21s

ribosomal

RNA

I., Thomas,

F.: A simple

of mitochondria

method from

instability

of

origins of

of ori0

petite

Martin,

N.C., Underbrink-Lyon,

cation

P.:

Fine

region

on

yeast

168 (1979) 25 l-277. W.

for the large-scale

microorganisms.

for tRNA

G. and Bernardi,

G.: A region

genome

of yeast.

P. and Attardi,

of a yeast mitochondrial

biosynthesis,

in Slonimski,

G. (Eds.), Mitochondrial

Laboratory,

Cold

A.M. and Gilbert,

DNA.

Spring

locus

P.P.. Borst,

Genes. Cold Spring

Harbor.

NY,

1982, pp.

Newman,

W.: A new method

yeast

Anal.

G.: The ori’ to

for sequencing

Proc. Natl. Acad.

Sci. USA 74 (1977) 560-564.

D., Pham,

Underbrink-Lyon,

H.D.,

K. and Martin,

of tRNA genes in tRNA region II of

mitochondrial

DNA.

Nucl.

Acids

Res.

8 (1980)

F. and Bernardi,

G.: The

5007-5016. A., Kopecka,

mitochondrial evolution. Thalenfeld,

of wild-type

and

Tzagoloff,

membrane

mitochondrial

6173-6180. Communicated

yeast cells, V. Genome

J. Mol. Biol. 110 (1977) 17-52. B.E.

mitochondrial of yeast

H., Strauss,

genome

77 (1977) 110-121.

M., Faugeron-Fonty.

replication.

K. and Miller, D.L.: Identifi-

and characterization

Harbor

Prunell,

Borst,

D.Y., Bandlow,

R. and Bernardi,

in the mitochondrial

N.C.: Characterization

genomes

J.-L.,

G., Doxiadis,

Kaudewitz,

preparation

R., Colin, Y.. Goursot,

Maxam, of the

mutant

is accompanied

genome

263-267.

cytoplasmic

174 (1979) 335-337.

DNA. Mol. Gen. Genet.

B.. Burger,

Mangin,

of

of yeast. EMBO J. 1 (1982) 705-711. C., Talen.

mitochondrial and

G.: The nucleotide

in the mitochondrial

structure Lang,

The spontaneous

R.,

genome

J. Mol. Biol. 134 (1979) 493-537.

C. and Bernardi,

Marotta,

necessary

G., Goursot,

G.: The mitochondrial

cells, VIII.

and the

292 (1981) 75-78.

F., Baldacci.

A. and Bernardi,

G.: The origins genome

yeast petites

in mitochondrial

EMBO J. 1 (1982) 529-534.

G., Goursot,

G. and Bernardi,

of the yeast

Faugeron-Fonty, Prunell,

R., Faugeron-Fonty,

in spontaneous change

Gene, 24 (1983) 73-81. of extreme

Gene 21 (1983) 193-202. de Zamaroczy,

ori+ mutation by a drastic

1 (1980) 233-239.

by J. Carbon.

A.:

Assembly

system. Sequence DNA.

of

the

of the 0x12 gene

J. Biol. Chem. 255 (1980)