‘

-
..
3
I
l
J
a!
a

“5“.

 

 

 

 

 

 

h.>
it.“ n

 

;
‘i

"-17. .y .

..

 

.
.

—v.
.n—n‘v

.u‘

2....
.

 

.,

., I‘H . J

V36 25

‘H’vh‘iw o r,

I:
3"
:4
p
'7
,
5:
. O

 

 

 

mafia?” '

‘ .1
"4v".- ..
Jar 4 ‘.,,
t"

 

K ‘9
"‘ ”5°35?
z: '5; '

2:,

xi‘ffifi

,
.

[33" 1 5
111.- Eff“:
L

‘5‘
~. . fi‘fi: ~
nfiaanfiam.“
«a ,fimgawr

kw... ._
11' .1629‘; .-
‘ ‘

. . V n
.. 3‘1? 32.
#X— 4v

b“
;.:;.:;L
«4::
21551:

“a
.r_,

‘43::

.. . f: f”— ‘ ..
5:. ..»...:»"...-: :21.- 7 7
-1 mammamq - . ‘1“-
-f .. 1;»: .23.; “v
.
- M

«kn.

- «a?
.m rm'kmrm-
f“:

It

3’33 4" k ‘ fl, =-
A .. A ., ‘ ~ «s. aux, mam
. “L535 ' .‘mcrgez; 'ié ‘1), ; ~19“-
: 1:}; gazgxfi'fl 5-3."; “:1 .2533: 2:5, ‘
EQVy‘" ‘ 3 . ‘ ‘ ~-§,.l\w.' “mm ' -‘
.1 . ' ”1“ "" - hes-41m):

‘ "‘f " ”:7 .1

u

x

u» .73: :33:

‘0‘
- a? n. 1 ;?£
' Q3. 4237::
-: w»

r:

' 'nzu . a;

M 4" tyn- "M -c-

ik "W?’-"‘”
C:

an}

. ..,
.- “l_
~.’:L:‘5,“:

6‘ .-
v .a‘ - ' —-um n

M '5‘" ‘
~ .--- shwm.
”com. v, 1%.“
0' as. wfiflwv.
mp» mtvwmtow‘
u an , m...
u

 

PLASTID INHERITANCE AND CHLOROPLAST DNA REPLICATION IN OENOTHERA

By

Wan-Ling Chiu

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Botany and Plant Pathology

1991

35¢— ,g, 4/ ?

ABSTRACT

PLASTID INHERITANCE AND CHLOROPLAST DNA REPLICATION IN OENOTHERA

BY

WAN‘LING CHIU

Plastids of Oenothera, the evening primrose, can be transmitted to
the progeny from both parents. Previous studies suggested that the ex-
tent of biparental plastid inheritance and the competitive ability of
plastids to multiply are determined primarily by the plastid genome
(plastome) itself, with the nuclear genome having only a minor impact.
These results are in apparent conflict with observations that plastid
multiplication in land plants does not appear to require proteins en-
coded by the plastome. In this dissertation, the impact of the nuclear
background on plastid transmission was reassessed. In crosses with
diverse nuclear and plastome combinations, a good correlation was ob-
served between compatibility of the particular plastome with the nuclear

genome of the progeny.

In order to test the hypothesis that a non-coding function of the
plastome, the origin of chNA replication, may be the plastome component
that influences plastid multiplication, the origins of chloroplast DNA
replication were located in two Oenothera plastomes. Two pairs of dis-
placement loop (D—loop) initiation sites, each lying within the inverted

repeats, were mapped to the Oenothera chNA molecule by electron micro—

scopic analysis of restriction fragments. The starting points of the
two adjacent D-loops are approximately 5 kb apart, bracketing the 168
rRNA gene. No differences in the number and the location of replication
origins were apparent between plastomes with the highest and the lowest

transmission efficiencies.

Restriction analysis of several Oenothera plastomes revealed a
highly variable region in the 168 rRNA-trnI spacer near one of the chNA
replication origins.' DNA sequences of the 168 rRNA-trnI spacer were ob-
tained from several Oenothera plastomes to assess the possibility that
local DNA sequence variation may affect the efficiency of replication
initiation. Multiple copies of repeated sequences that have the poten-
tial to form stable stem—loop structures were found in this region.
These secondary structures may interfere with the movement of the
replication fork. A model based on the slipped strand mispairing of DNA
during replication is proposed to explain the amplification of repeated

sequences in this region.

This dissertation is dedicated in mmory of
my loving aunt Jo-Mei Chiu
for her encouragment and support

throughout my school years

iv

ACKNOWLEDGEMENTS

The past seven and a half years has been quite a journy for me. All of
this was started by a nice invitation letter from my major advisor,
Barbara Sears. I appreciate her and all the other past and present mem-
bers of the Sears laboratory for providing me the freedom and the en-
vironment to explore the facinating world of Oenothera. Special thanks
should go to Drs. Wilfried Stubbe and Franz Schfitz for building the fun-
dation of my research. Without their delicate work on Oenothera. this
dissertation would not have been possible. I would like to thank Drs.
James Hancock, John Ohlrogge and Shauna Somerville for serving as mem-
bers of my guidance committee. I would also like to acknowledge Karen
Klomperans, John Heck-an, Barry Stine, Robert Meeker, Brent Nielsen and
Madeline Nu for their advice in electron microscopy. Finally, I thank
all the members of Chin and Elhai families, especially JAB, "SOB and

LACE for their encouragement and support.

TABLE OF CONTENTS

Page
LIST OF TABLES ...................................................... viii
LIST OF FIGURES ....................................................... ix
GENERAL INTRODUCTION ................................................... 1
CHAPTER 1. PLASTOME—GENOME INTERACTIONS AFFECT PLASTID TRANSMISSION IN
OENOTHERA ................................................... 7
Introduction ....................................................... 7
Material and Methods ............................................... 9
Plant Material .................................................. 9
Genetic Crosses ................................................ 9
Germination and Scoring of the Seedlings ....................... ll
Germination of Oenothera Pollen in vitro...... ...... . ....... ...11
Results. .......................................................... 12

Haploid Chromosome Complex C Facilitates the Transmission of
Plastome Type IV from Pollen ................................. 12
Relative Efficiency of Plastid Transmission Differs in
Reciprocal Crosses That Result in Progeny with Different
Hybrid Genotypes. ............................................ 18
The Relative Tran-ission Efficiencies of Plastome Types 1 and
III from the Pollen are Reversed in Progeny with AB vs. AA

NuclearBackgrounds ..... . .................................... 20
Plastome-Genome Interaction Can Affect Pollen Tube Growth ...... 22
Discussion............................................. ........... 27

CHAPTER 2. ELECTRON MICROSCOPIC LOCALIZATION OF REPLICATION ORIGINS

IN OENOTHERA CHLOROPLAST DNA ............................... 35
Introduction .......................... ........ ................. ...35
Material and Methods .............................................. 37

Plant Material... ................ . ............................. 37
CpDNA Preparation .............................................. 37
Electron Microscopy and Data Analysis. ......................... 38

vi

Results. ........ ...... ............... . ..... ............... ....... .39

Oenothera CpDNA Replication is Initiated through the Formation

of Multiple Displacement Loops.......... ..... . .......... .....39
D-loop Initiation sites are Located in Both Copies of the
Inverted Repeats ...... .. .......... ... .............. . ........ 39
Each Pair of D-loop Initiation Sites is in the Close Vicinity
Of rRNAGeneSO ...... 0.....0...O......OOOOOOOOOOOOOOOO0.0.51
Discussion ............................... ... ..... . ................ 54

CHAPTER 3. PROLIFERATION OF DIRECT REPEATS NEAR THE OENOTHERA

CHLOROPLAST DNA ORIGIN OF REPLICATION ........ . ............ 58
Introduction ...... . ............................................... 58
Material and Methods. ............................................. 59

Plant Material. . . .......... . ...... . ........ . ............... 59
Restriction Analysis and Sequencing ..... ..... ................ .60
Computer Programs Used for Sequence Analysis... ............... .61
Results ........ . ............ . ....... . ...... . ...................... 61
One D-loop Region is Proximal to a Highly Variable Region
of the Oenothera Plastome ............................. . ..... 61
Sequence Comparison of the 16S rRNA-trnI Spacer...... ....... ...64
The Sequence of the Duplicated Region in Oenothera Has the
Potential to Form Stable Hairpin Structures..................68

Most of the Duplicated Sequences in the Spacer Have been
Deleted in the CpDNA of a Strain Derived from the Plastome
Mutator Line. . ..................... ............. ........... 68
Multiple Repeats are not Stable in the Evolution of Oenothera
P1astomes........ ...... ... ..... . .............. ...............73

Discussion.......... ................... . ................. . ...... ..76
comLUSIONOOOOOOOOOO0.0.0....0..........OOOOOOOOOOOOOOOOO ..... 00......85

APPENDIX A. OENOTHERA CHLOROPLAST DNA POLWMORPHISMS ASSOCIATED WITH
PLASTOME MUTATOR ACTIVITY.................. ..... . ..... ....91

APPENDIX B. RELATIONSHIP BETWEEN SEED POSITION AND THE EXTENT OF
BIPARENTAL PLASTID TRANSMISSION.. ..... ... ............. ....97

APPENDIX C. COMPARISON OF TRANSMISSION OF PLASTOMES ID AND IC1.......102

APPENDIX D. ATTEMPT TO LOCATE THE ORIGIN OF DNA REPLICATION IN THE
CHLOROPLAST OF OENOTHERA USING TWO-DIMENSIONAL GEL
ELECTROPHORESIS .......... .... ........ ....... ............. 106

LIST OF REFERENCES............ .............. . ..... . ........ .. ........ 111

vii

Table 1.1.
Table 1.2.

Table 1.3.

Table 1.4.

Table 2.1.

Table B.1.

Table C.1.

LIST OF TABLES
page
Oenothera strains used in the crosses...... .......... . ..... 10

Plastid transmission in crosses producing progeny with AC
genotype ............. . ........................ . ...... . ..... 14

Plastid transmission in AC X AA crosses.. .................. 19

Plastid transmission in crosses producing progeny with
genotype AB ................................................ 21

Distribution of D-loops in chNA of Oenothera plastomes I
and IV.. ............................ . ...................... 50

Frequency of Biparental plastid transmission (BP) obtained
from seeds in upper (U) or lower (L) half of the capsule...99

Frequency of Biparental plastid transmission in crosses

with mutant plastids transmitted by the female parent and
plastome 131 or I° transmitted by the male parent ......... 104

viii

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

LIST OF FIGURES

page

1.1. Frequency of progeny carrying plastids from both parents

in White X Green crosses of Oenothera ..................... 16
1.2. Frequency distribution of the degree of variegation in

biparental progeny ........................................ 24
1.3. Seed capsule development from crosses using APCP-II

pollen .................................................... 26
1.4. Frequency of biparental plastid transmission in Green X

White crosses of Oenothera ................................ 32
2.1. Closed circular molecules of Oenothera chloroplast DNA

(plastome type IV) with D-loops ........................... 41
2.2. Restriction and partial gene map of Oenothera chNA ....... 43
2.3. Electron micrographs of chNA (plastome type I) digested

with Sell ................................................. 45
2.4. Length distribution of chNA fragments containing D-loops

from plastome types I and IV.. ............................ 48
2.5. Proposed alignment of D-loop-containing molecules relative

to the rRNA operon and the small single copy region of

Oenothera chNA ........................................... 53
3.1. Localization of the hypervariable region in the rRNA

operon of Oenothera chNA .................. . .............. 63
3.2. Sequence comparison of the EcoRI-BssHII fragment from

Oenothera hookeri str. Johansen plastome I° (Oj.) and

N. tobaccum (Nt.)....... ........ . ....... .. ...... ..........67
3.3. Secondary structures in the repeated segment a and a? of

Oenothera plastome ID and the equivalent region in

tobacco chNA ............................................. 70
3.4. Restriction map of the variable EcoRI-BssHII fragment in

M plastme I” and 1‘31 . ............................ 72
3.5. Comparison of the variable 168-552; spacer among Qgggthggg

plastomes... ......................... . .................... 75
3.6. The generation of duplication (A) or deletions (B) by

slipped-strand mispairing of direct repeats during
replication of DNA..... ......... . ......................... 79

ix

Figure 3.7. Major steps in the expansion of the 168 rRNA-trnI spacer
of Oenothera plastome ID starting from sequences similar
to other land plant plastomes ..... . ....................... 81

GENERAL INTRODUCTION

Plastids are essential organelles of plant cells. Besides their role in
photosynthesis, plastids function both in photosynthetic and nonphoto-
synthetic plant cells, such as roots, tubers, flowers and fruits, for
the synthesis or storage of starch (amyloplasts), carotenoids (chromo-
plasts), amino acids, fatty acids, and lipids (reviewed by Thomson and

Nhatley 1980).

The many proteins required for the diverse functions of plastids are not
all encoded by the plastid genome (plastome). Judging from genes deduced
frm the complete plastome sequences available frcml three land plants,
Marchentia (Ohyama 35 91. 1986), tobacco (Shinozaki gt 91. 1986), and
rice (Hiratsuka e_t 91. 1989), and frm the small gsmome sizes of
plastomes (120-180 kb; reviewed by Palmer 1985), it is evident that
plastmes encode proteins mainly of the photosynthetic apparatus and the
transcription and translation machinery, with the majority of plastid
proteins being nuclear-encoded. Nuclear gene products not only parti-
cipate in the metabolic processes occurring in the plastid such as
photosynthesis, and amino acid and fatty acid biosynthesis (reviewed by
Kirk and Tilney-Bassett 1978), but also determine the developmental fate
of plastids by regulating plastid gene expression (reviewed by Mullet
1988). The coordination of gene expression occurs in the other direction
as well. Several studies have found that photo-oxidative damage of

chloroplasts can affect the expression of photosynthetic genes in the

nucleus (reviewed by Taylor 1989). Also, some nuclear genes that encode
chloroplast proteins have cis-acting regulatory elements that respond to
the developmental stage of chloroplasts in the same cell (Aoyagi 91,91.
1988, Jefferson 91 91. 1987, Stockhaus 91,91. 1989). Apparently, the
coordinated expression of nuclear- and plastid-encoded components is
necessary to maintain the normal functions of plant cells. This thesis
investigates one vital area of the plastome-genome interaction: the

multiplication and transmission of plastids.

Since new plastids arise only frem pre-existing plastids, mechanisms
must exist to ensure the continuity of plastids in each daughter cell.
Our current understanding of the mechanisms responsible for the
multiplication and segregation of plastids in plant cells remains very
limited. Studies using inhibitors of protein synthesis (Heinhorst 91,91.
1985) and investigations on plastid mutants that lack ribosomes (Nalbot
and Coe 1979, Scott et a1. 1982) suggest that chNA synthesis is
dependent on nuclear-encoded proteins and that plastids can multiply
even in the absence of proteins synthesized in the plastids. By
quantifying the amounts of DNA in individual chloroplasts during leaf
development, several groups concluded that plastid division in land
plants is not tightly coupled to cell division and that chNA synthesis
does not immediately precede plastid division (Lamppa and Bendich 1979,
Lawrence and Possingham 1986). Concordantly, the number of plastids per
cell can increase from 10-15 in.meristematic cells to several hundred in
mature mesophyll cells, and the number of plastomes per plastid can
range from a few in the proplastid to several hundred in chloroplasts of

expanding mesophyll cells prior to the active stage of chloroplast

division (Lawrence and Possingham 1986, Bendich 1987). The multiple
copies of plastomes are organized into several nucleoids that, through
their attachment with the plastid membrane, may serve as basic units for
the segregation of chNA (Kuroiwa 91 91. 1981, Hansmann 91 91. 1985,

Lindbeck 91 a_1. 1987).

The continuity of plastids also depends upon the transmission of
plastids from generation to generation. The inheritance of plastids in
different land plant groups has been observed to be maternal, paternal,
or biparental. Conifers inherit their plastids mainly from the paternal
side (Neale 91 91. 1986, Neale 91 91. 1989, Szmidt 91 91. 1987, Stine 91
91. 1989), however, two smaller groups of gymnosperms, Ginkgoales and
Cycadales, seem to transmit plastids only from the maternal parent
(Nhatley 1982). The majority of angiosperms inherit their plastids only
from the maternal parents, due to the degradation or exclusion of
plastids during male gamete development (reviewed by Hagemann and
Schréder 1989). However, a high frequency of biparental (reviewed by
Kirk and Tilney-Bassett 1978) or paternal plastid transmission (Schumann
and Hancock 1989, Masoud 91,91. 1990, Bobenz 91 91. 1990) has also been
observed in a few angiosperms. Even in some species thought to exhibit
solely maternal plastid inheritance, a low frequency of paternal
plastids has been detected in the progeny (Medgyesy 1986, Schmitz and
Eowallik 1986). Apparently, the barriers that prevent transmission of

plastids from one parent or the other are not absolute.

Genetic studies of plastid transmission in different plant species have

suggested that both the nuclear genomes of the parents and the plastomes

can affect the pattern of plastid inheritance. For example, in Petunia
h rida, a species that exhibits maternal plastid inheritance, a pater-
nal genotype was identified that enables the transmission of paternal
plastids to a low percentage of the progeny (Cornu and Dulieu 1988). In
alfalfa (MediC9go,991119), a species that exhibits a high frequency of
paternal and biparental plastid transmission, a particular maternal
genotype was found to significantly increase the frequency of paternal
plastid transmission (Masoud 91 91. 1990). Similarly, in Pelargonium,
the fraction of progeny inheriting plastids from maternal, paternal, or
both parents is influenced mainly by a nuclear gene on the maternal side
(reviewed by Tilney-Bassett 1988). In contrast, in Oenothera, the
evening primrose, the plastome has a major influence on the frequency of

biparental plastid transmission (Schétz 1974, 1975, Chiu et a1. 1988).

Organelle genetics can be thought of as a question of intracellular
population genetics (Birky 1978). Factors that influence either the
input ratio of plastids from the parents or the dynamics of the plastid
population during embryo development should affect the output of
plastids among the progeny (ie, plastid inheritance). In order to
understand the mechanisms through which nuclear and plastid genomes may
affect plastid transmission, both genetic and molecular approaches were
taken in this dissertation to analyze factors that may influence plastid

multiplication and transmission in Oenothera.

The genus Oenothera is a unique system for studying plastome-genome
interactions in plastid.multiplication and transmission. First,

Oenothera exhibits a high frequency of biparental plastid inheritance.

Second, broad interspecific crosses are possible in this genus (reviewed
by Cleland 1972). Hence, the interactions between different plastomes
and between plastomes and nuclear genomes can be studied with a wide
variety of combinations. Third, due to extensive chromosomal trans-
'locations, many haploid genomes of Oenothera are transmitted as intact
chromosomal complexes. This allows the rapid transfer of plastids into a
desired nuclear background in a few generations (reviewed by Stubbe and
Herrmann 1982). Crosses between some strains of Oenothera result in a
phenomenon known as hybrid variegation (reviewed by Kirk and Tilney-
Bassett 1978) due to incompatibilities between nuclear and chloroplast
genomes. Stubbe (1959) categorized Oenothera plastomes into five basic
types (I, II, III, IV and V) according to their compatibility with the
six major nuclear genome types (AA, AB, BB, BC, AC and CC). Results from
a series of experiments conducted by Schétz (1954, 1974, 1975) suggested
that the five basic plastome types of Oenothera have different abilities
in competing with each other in crosses and that the nuclear genome has
only a minor effect on plastid transmission. The plastome-dependence of
both the efficiency and the onset of plastid.multip1ication after
fertilization was confirmed by more controlled experiments (Chiu _1 _1.
1988), in which the transmission efficiencies of'999919999 plastomes

were compared in a constant nuclear background.

Since plastid multiplication does not appear to require proteins encoded
by the plastome, the observed plastome—dependent variation in Oenothera
plastid multiplication and trssuission raises a paradox: nuclear-
encoded genes are thought to play the most major role in plastid

.multiplication, and yet the plastome has a major influence on the

efficiency of plastid multiplication. This dissertation examines the
hypothesis that the observed plastome-dependent efficiency of plastid
transmission is at least partially dependent on the interaction between
plastid and nuclear genomes. Thus, both nuclear and plastid components
are critical to efficient plastid transmission. Since previous work
confirmed the importance of the plastome in this process (Schétz 1954,
1974, 1975, Chin 91 91. 1988), the first chapter of this thesis
reassesses the degree to which the nuclear genome influences plastid
transmission. Specifically, it examines how the relative efficiency of
plastid transmission in various nuclear backgrounds relates to the

concept of plastome-genome compatibility.

The remainder of this dissertation considers a second hypothesis that a
non-coding function of the plastome, the origin of chNA replication,
may be the plastome component that modulates the efficiency of plastid
multiplication. In the second chapter, the chNA replication origins of
two Oenothera plastomes that have the highest (type I) or the lowest
(type IV) multiplication efficiency were localized by electron micro-
scopy. These investigations examined the possibility that differences in
the number or location of chNA replication origins are responsible for
the observed variations in the efficiency of plastid multiplication. The
third chapter focuses on a highly variable region near one of the
Oenothera chNA replication origins to assess the possibility that local
DNA sequence variation may affect the efficiency of replication
initiation. A model based on the slipped strand mispairing of DNA is
proposed to explain the elevated frequency of DNA length mutation in

this region.

CHAPTER 1

PLASTOME-GENOME INTERACTIONS AFFECT PLASTID TRANSMISSION IN OENOTHERA

Introduction

The majority of angiosperms inherit their plastids only from the
maternal parent. However, in some angiosperms, plastids are transmitted
at a high frequency from both parents (reviewed by Kirk and Tilney-
Bassett 1978) or predominantly from the paternal parent (e.g., Schumann
and Hancock 1989, Masoud 91 91. 1990, Bobenz 91 91. 1990). Genetic
studies of plastid transmission in different plant species have
suggested that plastid inheritance can be affected either by the
parental nuclear genomes (Cornu and Dulien 1988, Tilnebeassett 1988,
Masoud 91 91. 1990) or by the plastid genomes (plastomes) involved in

the crosses (Schiitz 1974, 1975, Chiu 91 _a_1_. 1988).

In Oenothera, plastids can be transmitted from both parents. In some
interspecific crosses of Oenothera, biparental inheritance of the
plastids can be observed as ”hybrid variegation", with plastids from one
parent failing to become fully pigmented in the hybrid nuclear back-
ground. This genetic dysfunction has been termed ”plastome-genome incom-
patibility". The genetic studies of Stubbe (1959) on this subject led
him to categorize the chloroplasts of the subsection Oenothera into five

basic types (I, II, III, IV and V) according to their compatibilities

with the six major diploid nuclear genotypes (AA, BB, CC, AB, AC and

BC).

The five plastome types of Oenothera also differ in their transmission
efficiencies. Schfitz (1954, 1974, 1975) studied the frequencies of
biparental plastid transmission in crosses of a large number of wild-
type Oenothera species with several strains carrying mutant plastids. He
concluded that the five plastome types have differing abilities to
compete with incoming plastomes in crosses and classified the five
plastome types as strong (types I and III), intermediate (type II), or
weak (types IV and V), according to their relative competitive
abilities. The same relative transmission efficiencies were observed
when plastome types I through IV were compared in reciprocal crosses in
a constant nuclear background (Chiu 91 91. 1988). These studies, which
span four decades, are the basis of the widespread belief that the
efficiency of plastid transmission in Oenothera is determined mainly by

the plastids themselves (Kirk and Tilney-Bassett 1978, Stubbe 1989).

Although Schétz (1974) pointed out that the nuclear genome can also
influence plastid inheritance, the actual role of the nuclear genome in
plastid transmission has not been defined in Oenothera. Since both
plastome- and nuclear-encoded gene products are required for trans-
cription and translation within the plastid (reviewed by Sugiura 1989),
it is conceivable that critical processes such as plastid multiplication
and transmission may also rely on the interactions between plasteme and
genome. In this study, crosses were performed to analyze the impact of

both the plastid and nuclear genomes on plastid tran-ission.

Material and Methods

Plant material. Oenothera strains used in this study are listed in Table
1.1 All mutant plastids were maintained as periclinal chimeras in
Oenothera hookeri str. Johansen (genotype AJAJ) that carried wild-type
plastids of plastome type IV, in addition to the mutant plastids (Chiu
91 91. 1988). All the strains heterozygous for chromosome complexes were

provided by Prof. H. Stubbe (University of Dfisseldorf, FRG).

Genetic crosses

Group 1: formation of hybrid genotype AC. AJAJ plants with mutant
plastomes of types I, II, III, and IV were used as the maternal parent
in this set of crosses. Pollen was obtained from the Oenothera hybrid
albicans/percurvans (AFC?) with plastomes of type II, III, or IV,
Oenothera ammophila (AFC?) with plastomes of type II or IV, 91 parvif-
1999 (B'C'-IV) or 91 atrovirens (chf-IV). Due to the pollen lethal
factors carried by the AP, AT, D. and BD genome complexes, the pollen
transmits only the type C chromosome complex (Cleland 1972), and hence,
offspring of these crosses have the genotype AC, with A contributed by

the egg and C by the pollen.

Group 2: formation of hybrid genotype AB. Crosses yielding progeny of

AB-genotype were performed by pollinating Oenothera strains carrying a

10

Table 1.1. Oenothera strains used in the crosses

 

 

 

 

Oenothera Gsmome Genotypeb Plastome Source of
species cgplexes‘ type‘ plastid
O. hookeri str. Joh hJohansenJ'Johansen AJA5 I-S O. hookeri str.

0. hookeri str. Joh hJohansen."Johansen AJAJ
0. hookeri str. Joh hJohansenJ‘Johansen AJAJ

O. hookeri str. Joh hJohansen."Johansen AJAJ

hybrid albi cans $. percurvansc Aa CF I I
hybrid albicans 51-. percurvanss A‘I 09 III
hybrid albi cansg. percurvans 3 A“ Cp I V
O . ammophi 1a rigens 3. percurvans 6 A" CD I I
O . amophi 1a ri gens 3. percurvans 9 Ar 09 I V
O . parviflora augensfi. subcurvans ’8 B‘ C‘ IV
0. atrovirens pingeng. f lectens 6 B"Cf IV
0 . grandi flora ggandi flora. ggandi flora B‘ B‘ I II

J oh

II-€ O. suaveolens

III-r O. grandiflora

IV-oL O. mophila

O. suaveolens
O. grandiflora
O. ammophila
O. suaveolens
O. ammophi 1a
0. parviflora

O. atrovirens

mm

 

' The name of each chrmosome complex follows Cleland (1972). Haploid
genome that can be transmitted only through one of the gametes is

indicated with male or female syfiol.

1’ Each haploid genome is represented by A, B, or C according to the
compatability group to which it belongs (Stubbe 1959). The superscript
letter is an abbreviation of the name of the haploid genome.

c The Roman numerals indicate the plastome type. If followed by a Greek

letter, a plastome mutant,is indicated.

11

type B chromosome complex with pollen from an AJAJ plant carrying a
mutant plastid. Due to the egg lethal factors carried by the Cf and Cs
chromosome complexes (Cleland 1972), crosses using either 91 parviflora
(B‘C‘) or 91 atrovirens (BPCf) as the maternal parent result in progeny

with nuclear backgrounds AiBa or A3BP, respectively.

Germination and scoring of the seedlings. Seeds were placed in water in
a beaker, and were shaken at 100 rpm under constant fluorescent light
until they germinated. The seeds were surface-sterilized daily with a
combination of 20% commercial bleach and 0.1% SDS until roots emerged (3
- 6 days). Due to the strong maternal bias in Oenothera plastid trans-
mission, crosses in which mutant plastids were contributed by the
maternal parents produced mainly white seedlings. For this reason,
germinating seeds from White X Green crosses were placed on agar medium,
as previously described (Chiu 91 91. 1988), while germinating seeds from

Green X White crosses were transferred to soil.

Progeny with plastids from both parents were recognized by the varie-
gation of their cotyledons. In each cross, the number of progeny with
plastids from both parents was recorded, as well as the fraction of both
cotyledons with pigmentation characteristic of the plastids inherited

from the paternal parent.

Germination of Oenothera polls-119 vitro. Mature Oenothera pollen grains
were harvested from mature flower buds that were ready to open on the
same day (judging from the size and color of the flower bud) by

aspirating the pollen into a pasteur pipet. These pollen were sprinkled

12

onto a thin layer of 0.6% agarose, containing 10% sucrose, 0.01% H3303,
and 0.02% CaClz (modified from Corriveau and Coleman 1988) in a petri
plate. After two hours of incubation at room temperature, the extent of
pollen tube growth was recorded by placing a ruler under the petri plate
and estimating the point to point distance between the base and the tip

of several typical pollen tubes in a fairly homogeneous population.

Results

Haploid chm-came cowlex C facilitates the transmission of plastue

type IV fro-[pollen

If the efficiency of plastid transmission is influenced by the relative
compatability between the plastome and the genome, an enhanced
transmission of paternal plastids might be seen when the nuclear
background of the progeny is more compatible with the paternal plastid
than with the maternal plastids. In order to test this hypothesis, the
relative efficiencies of plastid transmission in crosses producing
progeny with AJAJ or AJCp genotypes were compared. Genotype AJAJ is
associated with plastome type I in nature (Stubbe 1964). In the AJAJ
nuclear background, chloroplasts with plastomes I, II, and IV have
normal pigmentation, while those with plastome III are periodically
bleached (Stubbe 1959). Genotype AC is the natural nuclear background of
plastome type IV. The relative compatabilities of the same plastome
types in an AC nuclear background are in the order of IV 1 II > III > I

(Stubbe 1959). Due to plastome-genome incompatibility, AC-I plants were

13

not viable in the field and APCP-III plants produced pollen with poor
vigor. Thus, the analysis of paternal parents with an AC genotype could

only include plastomes II and IV.

In the crosses detailed in Table 1.2, the maternal parent had the
genotype AJAJ and contained mutant plastids, representing one of the
four basic plastome types. Wild-type plastids of plastome type II or IV
were contributed by the pollen. Figure 1.1 shows a comparison of the
crosses in which the paternal parent had an A‘Cp genotype (Panel B and
C) with previous data, obtained using pollen from AJAJ plants (Panel A).
The transmission efficiency of paternal plastome type II in crosses
producing progeny with the genotype AJCp (Panel B and C) is somewhat
higher, but fairly comparable to that obtained from crosses producing
genotype A5A3 (Panel A). But for plastome type IV, crosses producing
progeny with genotype AJCp have a much higher transmission efficiency of
paternal type IV plastome than in the case for crosses producing progeny

with the AJAJ genotype.

The frequencies of biparental plastid tran-ission from crosses
performed in the first field season were higher than those obtained from
the second field season (Table 1.2 and Figure 1.1B and C), but each set
of data is internally consistent in terms of the relative transmission
efficiencies of the different plastome types. Crosses in the first year
were performed later in the season, and their higher frequencies of
biparental plastid transmission may be due to the smaller maternal

flowers produced late in the season (see Discussion).

14

Table 1.2. Plastid transmission in crosses producing progeny with AC

 

 

 

 

 

g9notype
Average
Crosses F1 Transmission area of green
Hybrid tissue in
M91erna1 ,P9ter991 g9go1ype N 59F. 92 progeny
Season 1
Set 1
AJAJ-IS A'CP-II AJCp 96 11.5:3.3 0.18
A5AJ-IIE APCP-II AJCp 84 3.6:2.0 0.17
AJAJ-IIIT' ABCP-II AJCp 138 8.7:2.4 0.15
A5 AJ-IVd A‘CP-II A3 C9 53 22.6:5.7 0.31
Set 2
A3 AJ-IS A'CP-IV A3 C» 132 26513.8 0.29
AJ AJ -IIE A3CP-IV A509 136 5.9:2.0 0.13
AJAJ-IIIT A‘CP-IV AJCP 67 22.415.l 0.11
AJAJ-IVOL APCP-IV AJCp 133 31.1:4.0 0.29
Set 3
AJAJ-IIIF' B‘Cs-IV A40“ 131 6.9:2.2 0.13
AJAJ-IVCK B‘C'-IV AQCF 30 10.015.5 0.17
Set 4
AJAJ-IIII' BPCf-IV AJCf 21 0 -
A3 AJ -IVd ny-IV A3 Cf 19 0 -
Season 2
Set 1
MAJ-I 3 A'CP-II 1va 162 3. 111.4 0. 14
AJAJ-IIE. APCP-II A309 188 2.7:l.2 0.14
AJAé-IIII' APCP-II A509 190 3.7:1.4 0.13
A3 A5 -IV(X A‘CP-II A309 27 7.4:5.0 0.15
Set 2
AJ 113-13 A'CP-IV 1va 176 10.8123 0. 15
AJAJ-IIE. AFCP-IV AJCp 99 4.012.0 0.20
AJAJ-IIIt' APCP-IV AJCp 101 9.913.0 0.18
A3 AJ -IV& A‘CP-IV A309 173 24. ,_3.3 0.24
Set 3
AJAJ-IS A’CP-IV 1va 438 5.911. 1 o. 19
AJAJ-IIE A'CP-IV A$Cp 152 3.3:1.4 0.20
AJAJ-IIII' A'CP-IV AJCp 205 6.81l.8 0.18
A3 A3 -IV(X A’CP-IV A3 09 125 8.012.4 0. 15

 

 

' The biparental (BP) transmission frequencies are presented with their

standard deviations estimated by the normal approximation for the
binomial population (Steel and Torrie 1980).

15

Figure 1.1. Frequency of progeny carrying plastids from both parents in
White X Green crosses of Oenothera. The maternal parents had the
genotype AJAJ and the plastome type indicated on the X-axis. The pollen
carried either plastome II (closed bars) or IV (open bars). (A) Both
paternal and progeny genotypes were AJAJ. (B) and (C) Paternal genotype
was AFCP, and progeny genotype was AJCP. The data for (A) is taken from
Chiu 91 91. (1988). The data for (B) and (C) are from Table 2 (season 1

and 2, respectively).

’ FREQUENCY OF VARIEGTED 'PROGENY (%)

16

 

3°" A PATERNAL PLASTIDS: - u 1:: w

203%

10-~

fidxdfleflfl

 

 

O r
30-» . . . ’7
,1 AJA’anCp-aAle
20+ W
,1
10%L
.4 uLIl—l...
301- s s s ‘
A'AleaneA'Cp
.1
20--
101. '7 ,1
«)
,,L__l.___mfl__l._ .
H "-6 Ill-7 IV-a-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MATERNAL PLASTIDS

17

The relative success of different maternal plastids also varied with
nuclear background. In crosses producing genotype AJAJ (Figure 1.1A),
the transmission success of the maternal plastome was in the order of I-
zeta > II-epsilon 2 III-gamma > IV-alpha, as seen by the increasing
frequency of plastid transmission from the male parent (judged by the
frequency of variegated progeny). However, in crosses producing genotype
AJCP, the relative success of maternal plastid transmission was in the

order of II-epsilon > III—gamma > I-zeta > IV-alpha (Figure l.lB and C).

To distinguish between effects on plastid transmission exerted by the
genotype of the paternal parent as opposed to the genotype of the
progeny, two strains, A9CP-IV and A'CP-IV, were used. These two strains
carry different genotype A complexes (AP vs. A?) but the same C
chromosome complex (09), which is the haploid genome that can be
transmitted through pollen. According to the results from the second
season (Table 1.2, Season 2), pollen plastids are transmitted at a lower
frequency from A’Cp pollen (Set 3) than from AFCp pollen (Set 2). Thus,
the transmission of plastome type IV from pollen carrying the identical
C genome can vary, depending on the genotype of the pollen parent. An
attempt to include AFCP-II in these comparisons failed because plastome-
genome incompatibility resulted in poor vigor of the AFCP-II plants and

poor pollen tube growth.

Crosses with two other strains capable of contributing chromosome
complex C from pollen, B‘C'-IV and chf-IV, were also conducted to allow
a comparison of the impact of different C genomes on plastid

transmission. As shown in Table 1.2, the transmission frequency of

18

plastids from B‘C'-IV pollen (Season 1, Set 3) was lower than that from
A‘CP-IV pollen (Season 1, Set 2). Crosses between AJAJ strains and BPCf
-IV produced very few seeds. As will be discussed later, this may have
been due to limited pollen tube growth. No plastids from the paternal
chf-IV parent were detected in a total of 50 progeny produced from
these crosses (Season 1, Set 4). Despite the difficulty in obtaining
seeds from some crosses, the results from crosses involving pollen
carrying a C genome indicate that the presence of a C genome in the
pollen or progeny does not ensure efficient transmission of plastome
type IV from pollen. Furthermore, the diploid genome of the paternal

parent may also affect the transmission efficiency of pollen plastids.

Relative efficiency of plastid transmission differs in reciprocal

crosses that result in progeny with different hybrid genotypes.

In reciprocal crosses of those performed in season 2 of Table 1.2, the
same AJAJ plants carrying mutant plastomes I through IV were used as
paternal parents while A'CP-II, AFCP-IV and A’CP-IV plants were used as
maternal parents. Due to the egg lethal factors carried by chromosome
complex 0", only the A complexes are transmitted by the maternal
parents. In these crosses, resulting in the hybrid nuclear background
APAJ or ATAJ, maternally— contributed plastome type II (Table 1.3, Set
1) was consistently transmitted at a higher efficiency than maternally-
contributed plastome type IV (Table 1.3, Sets 2 and 3). The
efficiencies of paternal plastid transmission were in the order of I-
zeta 2 III-gamma > II-epsilon > IV-alpha, in all three sets of crosses.

Thus, the relative transmission efficiency was different from that

19

Table 1.3. Plastid transmission in AC X AA crosses‘

 

 

Average
Crosses F1 Transmission area of white
Hybrid tissue in
Maternal Paternal ,4vgenotype N %BPb 111§P progeny
Set 1
A-cv-II MAJ—13 ms 164 75.0134 0.34
APCP-II AJAJ-IIE, A?A5 121 19.013.6 0.14
A'CP-II AJAJ-IIIT' A'Aj 93 61.3:5.0 0.20
A‘CP-II A3 A3 -IVd A‘AJ 195 6.2:1.7 0.12
Set 2
A'CP-IV 1: 14-1 S and 127 82.7134 0.45
APCP-IV AJAJ-IIS. AFAJ 184 69.0:3.4 0.24
AFCP-IV AJAJ-IIIt' A'AJ 201 85.6:2.5 0.30
APCP-IV AJAJ-IV AFAj 120 39.2:4.5 0.14
Set 3
Arcv—Iv AJAJ-I 5 ATA3 235 89412.0 0.45
A'CF-IV AJAJ-IIE A'AJ 187 76.5:3.1 0.29
A’CP~IV AJAJ-IIII' A'AJ 209 85.6:2.4 0.35
A'CP-IV A1 A5 -IVd At A5 196 31.6133 0.25

 

‘ Reciprocal crosses of those listed in Table 1.2, season 2.
b see Table 1.2.

20

observed in reciprocal crosses resulting in progeny with genotype AJCp
but the same as that observed in crosses resulting in genotype AJ'AJ

(Chiu et 91. 1988).

The relative transmission efficiencies of plastomertypes I and III from

the pollen are reversed in progeny with AB vs. AA nuclear backgrounds

Plastome type I is more compatible with an AA nuclear background than is
plastome III, while the reverse is true with an AB nuclear background
(Stubbe, 1959). The crosses shown in Table 1.4 resulted in hybrid
genotype AB. In all three sets of crosses, mutant plastome III-gamma was
transmitted from the pollen with a higher efficiency than was mutant
plastome I-zeta. In crosses resulting in genotype AJBa (Set 1) or AJB‘
(Set 3), the relative efficiency of paternal plastid transmission was in
the order of III-gamma > II-epsilon > I-zeta > IV-alpha, while in
crosses resulting in genotype AJBp (Set 2), the order of I—zeta and II-

epsilon was reversed.

In variegated progeny, the abundance of tissue displaying the paternal
plastid phynotype in individual progeny is an additional parameter that
can be used to assess the competitive multiplication abilities of the
various plastome types. In general, the abundance of paternal plastids
in the cotyledons has a positive correlation with the frequency of
biparental plastid transmission, especially when both values are high
(see Tables 1.2, 1.3 and 1.4). In progeny with an AJBP nuclear
background, the fraction of white tissue was much higher when the pollen

contributed mutant plastome III-gamma than when mutant plastome I-zeta

21

Table 1.4. Plastid transmission in crosses producing progeny with

 

 

 

 

genotyp9 AB
Average
Crosses F1 Transmission area of white
Hybrid tissue in
Maternal P919rnal Genotype N %BP‘ 4§P progeny _
Set 1
sacs-Iv .43 A3 -IS 15 an 218 77. 1:2. 8 0.14
B'C'-IV AJAJ-IIE. AJBa 129 82.2:3.4 0.25
B'C'-IV AJAJ-IIIl' AJBa 108 87.0:3.2 0.25
BaC‘-IV A5AJ-IVCL AJBa 23 53.8:10.4 0.16
Set 2
BPCf-IV 1:14-13 AJBP 196 85. 712.5 0.33
BPCf-IV AJ A3 -II£ AJ 89 285 75.4:2.6 0.37
BPCf-IV AJAJ-Illt’ AJBp 293 91.8:1.6 0.47
BPCf-IV AJ AJ -IVO. A3 B9 162 35213.8 0. 19
Set 3
B‘B‘-III 1514—1 9 AJBI 220 16412.5 0.19
8181—111 AJAJ —112 118‘ 50 24.0160 0. 10
B¢B¢-III AJAJ-IIII' AJBI 156 46.2:4.0 0.17
B‘Bt-III AJAJ-IVOL A381 203 8.412.0 0.11

 

a see Table 1.2.

22

was brought in through the pollen (Figure 1.28). This was true even
though the crosses producing these progeny had very similar frequencies
of biparental transmission (91.8% vs. 85.7% in Table 1.4). The relative
abundance of the two paternal plastids was reversed in progeny with an
AJAP nuclear background (Figure 1.2A), which is more compatible with

plastome I.

Plastome-genome interaction can affect pollen tube growth

When pollen from Oenothera species with small flowers was used to
fertilize strains with large flowers, seed development was always
limited to the upper part of the capsule. This phenomenon was most
likely due to the limited growth capacity of pollen tubes produced by
the small flowered species. However, differences in the extent of
fertilization could also be observed among pollen from strains with the
same genotype but carrying different plastomes. An example of this can
be seen in the seed capsules from AJAJ flowers that had been pollinated
with pollen from A!CP-II or Aficp-IV plants (Figure 1.3). Judging from
the development of the capsule and the location of viable seeds in the
capsules, pollen from the AFCP-II plants could only fertilize eggs in

the upper part of the capsule (Figure 1.33).

The differential capacity for pollen tube growth among Oenothera strains
was examined in gitgg. Most pollen tubes from large flowered AJAJ

strains carrying any of the four plastome types reached a length of 1 cm
after two hours of incubation at room temperature. In the same period of

time, pollen tubes from a strain with medium-sized flowers, Aficp-IV,

23

Figure 1.2. Frequency distribution of the degree of variegation in

biparental progeny.

Crosses were performed between maternal parent (A) AFCP-IV or (B) BPCf
-IV and paternal parent AJAJ-I (closed bars) or AJAJ-III (open bars).
The X-axis indicates the fraction of cotyledon area occupied by the
white plastids transmitted from the pollen parent, with the height of
the bars showing the percentage of progeny in each group. The
frequencies of biparental plastid transmission from these crosses are

listed in Tables 3 and 4 respectively.

FREQUENCY OF VARIEGATED PROGENY (96)

24

 

40»

304-

20 w-

104'-

40»

30¢.

20-

1OJ

 

PATERNAL PLASTIDS: - l-( [:1 Ill-y

 

      

. I
«la

A°ch A'A‘ 419A,

 

 

 

 

 

 

 

1

8de AIM-9MB?

 
 

 

 

 

 

 

 

 

 

II 11

<1/4 <1/3 <1/2 <2/3 <3/4

FRACTION OF COTYLEDON

 

(7)3

 

25

Figure 1.3. Seed capsule development from crosses using APCP-IV or A‘CP
-II pollen. The maternal parent was genotype AJAJ with mutant plastids,
and the paternal parent was either: (A) A‘CP-IV (A) or (B) A'CP-II. The

arrow in each panel indicates a well developed capsule.

 

 

 

27

reached a length of 0.5 cm, while pollen tubes from the same strain
carrying plastome type II (A'CP-II) grew to only 0.2 cm. The latter is a
length similar to that attained by pollen tubes from a small-flowered
strain, AFCP-IV, over the same period of time. Conceivably, the Figure
incompatibility between plastome type II and haploid genome C (Stubbe
1959) reduces the vigor of the pollen tube, although the pollen is

nonetheless fertile.

Discussion

The pattern of plastid inheritance in Oenothera has been thought to be
determined primarily by the plastome, due to different intrinsic rates
of plastid multiplication associated with each plastome type (Schotz
1954, 1974, 1975; Chiu gt 2;. 1988). The present study modifies this
concept and stresses the importance of plastome-genome interactions in
plastid transmission. The influence of both the plastid and nuclear
genomes on the relative efficiencies of plastid transmission is evident
either by comparing the frequency of biparental plastid transmission or
by assessing the relative abundance of paternal plastids in the

variegated progeny.

Although the plastome-genome compatibility is generally measured as the
ability of plastids to become fully pigmented in a given nuclear
background, this phenotype may serve as an indication of how well the
two genetic compartments can cooperate with each other in a broader

sense. Indeed, the relative efficiencies of plastid transmission

28

correlate rather well with the compatibilities of the plastomes with the
nuclear background of the progeny. For example, the relative success of
three maternal plastomes in crosses producing progeny with an AC
genotype (II-epsilon > III-gamma > I-zeta; Table 1.2, Figure 1.1B, C)
fell in the same order as the compatibilities of these plastomes with
the AC nuclear background (Stubbe 1959). Similarly, the relative
transmission efficiencies of paternal plastomes I-zeta and III-gamma
were reversed in crosses producing progeny with AA or AB genotypes, as
might be expected from their relative compatibilities with these nuclear
backgrounds. An extreme example of the influence of plastome-genome
compatibility on plastid transmission was reported by Stubbe (1963), who
recovered progeny carrying only paternal plastids in Oenothera from
crosses producing a hybrid genotype that was severely incompatible with
the maternal plastid. The influence of plastome-genome compatibility on
plastid transmission is independent of the photosynthetic ability of the
plastids, since the correlation between transmission efficiency and
compatibility applies to white mutant plastids as well as wild-type
plastids. The correlation between the relative degree of compatibility
and the relative efficiency of plastid transmission can be explained if
the efficiency of plastid multiplication is dependent at least somewhat

upon the cooperative interactions between plastome and genome.

Schotz (1974) noticed that the competitive ability of a plastome was
highest in its native nuclear background. However, due to two major
limitations in his experiments, Schotz did not observe the broad effect
that plastome-genome interactions can exert on plastid transmission.

First, because of the low viability of the mainly ”white" progeny from

29

"white x Green" crosses (Schotz 1954), only "Green x White" crosses were
performed in his subsequent studies (Schétz 1974, 1975). Since many of
the chromosome complexes of Oenothera are only transmitted through
either the egg or the pollen, reciprocal crosses often produce progeny
of different genotypes. In the absence of reciprocal crosses, the impact
of certain genotypes, especially the native nuclear background of
plastome IV (the "weak" plastome), on plastid transmission was rarely
examined. Second, due to the limitations in the available natural
plastome-genome combinations, the transmission abilities of all plastome
types were not compared simultaneously in the same nuclear background.
In this study, the ability to compare results from reciprocal crosses
and to compare the four Oenothera plastomes simultaneously in several
different nuclear backgrounds made it possible to obtain a more complete
view of the role of both plastid and nuclear genomes in plastid

transmission.

In addition to its influence on the relative efficiency of plastid
multiplication, the compatibility between plastome and genome may also
affect plastid transmission through its influence on other steps in the
process of fertilization, such as pollen tube growth. The growth
capacity of the pollen in gitgg is correlated positively to the size of
the flowers from the pollen donor. However, the growth of the pollen
tubes also can be affected by the plastome type carried by the plant:
pollen from an AFCp strain that carried the less compatible plastome
type II could grow only to a length that was half that of pollen
carrying plastome type IV. The influence of the plastid evidently

extends to pollen tube growth ig vivo, since A'CP-II pollen could only

 

30

fertilize eggs on the upper part of the capsule when used to pollinate
AJAJ, a strain with large flowers (Figure 1.3). Poor pollen tube growth
may also be responsible for the failure of seed production in crosses
between AJAJ and two even less compatible combinations, ArCP-II and
APCP-III. As discussed below, the vigor of pollen tube growth may have a

significant effect on the input of paternal plastids to the zygote.

The impact of the maternal genotype on Oenothera plastid transmission
appears to correlate with the size of the flower. Among the species used
in this study, 9; hookeri str. Johansen (AJAJ) and Q; grandiflora (B‘B‘)
are strains with large flowers. The length of the style in these large-
flowered strains is 7 to 9 cm. 9; ammophila (ArCP), Q; parviflora
(B‘C‘) and Q; atrovirens (BPCf) have small flowers and their styles are
only 2 to 3 cm long. The flower of hybrid strain A'Cp is of medium size
with an average style length of 4.5 cm. In all cases, flower size is not
significantly influenced by the plastome type of the plant. When a
constant pollen source is used in crosses, a much higher frequency of
paternal plastids is detected in the progeny when strains with small
flowers are used as maternal parents, as compared to maternal parents
with large flowers (Table 1.3, 1.4 and Chiu gt 31. 1988). The influence
of maternal flower sizes on plastid transmission is illustrated by the
three sets of crosses shown in Figure 1.4, where pollen was contributed
from AJAJ lines carrying mutant plastids and either AJAJ-IV (large
flower), APCP-IV (medium flower), or ATCP-IV (small flower) was used as
the maternal parent. Progeny from these crosses all have genotype AA and
plastome type IV from the maternal parent. Although biparental plastid

transmission occurred at similar frequencies with plants having small

31

Figure 1.4. Frequency of biparental plastid transmission in Green X
White crosses of Oenothera. The paternal parents had the genotype AJAJ
and the mutant plastome indicated on the X-axis. The maternal parents
carried plastome type IV with AJAJ (large flower), A'Cp (medium flower)
or A’Cp (small flower). Data for crosses involving AJAJ-IV (closed bars)
are from Chiu gt 31. (1988); data for crosses using APCP-IV (hatched

bars) and A’CP-IV (open bars) as maternal parents are from Table 3.

32

MATERNAL PARENTS: - AiAi-N Aan-N 1:1 A'cP-N

nHHH
‘\\\\\\\\\\\ .

 

 

7

 

 

I.

1

 

’ D
T ‘

m 0.0 0 O
8642

GS >Zw60mm th<0mE<>
m0 >Ozm30wmu

J
4.
1

001-

IV-a

Ill—7

PATERNAL PLASTIDS

"-5

It
_

33

and medium flowers as the maternal parent, when crosses involved a
large-flowered maternal parent, a consistently lower frequency of
biparental plastid transmission was observed. A similar observation was
made by Schétz (1974), who reported that a small flower variant of Q;
hookeri allowed the transmission of more paternal plastids than did the
original strain. Conceivably, the differences in plastid heredity could
be due to differences in the input of maternal plastids that correlate
with the size of the flower. According to Schbtz (1954), the strain with
the smallest flowers did have slightly fewer plastids in the egg
(between 14 and 26), but all other strains examined had similar numbers
(between 20 and 30). These results led Schotz to conclude that
differences in the input of maternal plastids should not have a major

impact on plastid transmission.

This study agrees with Schotz, and find it more likely that the input of
paternal plastids has a major impact on the final result of plastid
transmission. For all Oenothera strains examined, 90 to 100% of the
mature pollen generative cells contain plastid DNA aggregates (Corriveau
and Coleman 1990, and personal communication). However, high frequencies
of paternal plastid transmission are achieved only when the maternal
parent is a small-flowered strain or when the paternal parent produces
pollen capable of vigorous growth. The influence of maternal flower size
and the growth capacity of pollen tubes on plastid transmission can be
best explained if plastids degenerate in the generative cell as the
pollen tube extends towards the egg. According to this hypothesis, when
a slow-growing pollen tube progresses through a very long style, the

time between pollination and fertilization is longer, and more plastids

34

in the generative cell would be degraded before fertilization. Plastid
degeneration during pollen development is one of the mechanisms through
which maternal plastid inheritance is achieved in several plant species
(reviewed by Hagemann and Schroder 1989) including Epilobium (Schmitz
and Kowallik 1987), a close relative of Oenothera. In Epilobium,
plastids are inherited almost exclusively from the maternal side, but
biparental transmission can occasionally be detected (Schmitz and

Kowallik 1986).

At least 14% of the angiosperms surveyed contain plastid DNA in their
pollen generative cells (Corriveau and Coleman, 1988) and should be able
to transmit their plastids through the pollen. However, based on this
study and earlier studies of plastid transmission in Oenothera (Schotz
1954, 1974, 1975, Chiu gt al. 1988) and Pelargonium (Tilnebeassett and
Almouslem 1989), the final pattern of plastid inheritance depends not
only on the input of the parental plastids but also on the plastome, the
interaction between plastome and genome, as well as plastome-independent

effects determined by the nuclear genome of the parental plants.

CHAPTER 2
ELECTRON MICROSCOPIC LOCALIZATION OF REPLICATION ORIGINS IN

OENOTHERA CHLOROPLAST DNA

Introduction

Chloroplast DNA (CpDNA) replication in land plants is not coupled to the
synthesis of nuclear DNA. Rather, the timing of chNA amplification is
under developmental control (Lammpa and Bendich 1979, Lawrence and
Possingham 1986). In gigg studies using inhibitors of protein synthesis
(Heinhorst 95 a1. 1985) and studies of mutant chloroplasts completely
lacking ribosomes (Walbot and Coe 1979, Scott gt 9;. 1982) suggest that
all proteins essential for chNA replication and plastid multiplication
are nuclear-encoded. Indeed, a search of sequenced land plant
chloroplast genomes has not revealed any genes similar to those known to
encode proteins involved in DNA replication, with the possible exception
of the gene for single strand DNA binding protein (ssb) (reviewed by

Umesono and Ozeki 1987, Sugiura 1989).

Despite the apparent dispensibility of plastid-encoded proteins in chNA
replication, the plastid genome (plastome) appears to be critical in
determining the efficiency of plastid multiplication, as judged by
genetic studies of differential plastid transmission in the evening

primrose, Oenothera. In crosses of Oenothera, plastids can be

35

36

transmitted from both parents to the progeny (reviewed by Kirk and
Tilney-Bassett 1978). However, the frequency of biparental plastid
inheritance is strongly influenced by the plastome types involved in a
given cross (Schotz 1974, Schotz 1975, Chiu gt g1. 1988). Schotz (1974)
suggested that the five major plastome types of subsection Oenothera
(reviewed by Stubbe 1989) have different rates of multiplication and
ranked them into three categories with regard to their abilities to
compete with each other in crosses: "strong" (plastome types I and III),
"medium” (plastome type II) and "weak" (plastome types IV and V). The
results of comparing plastid transmission in a constant nuclear

background lent support to this concept (Chiu gt gt. 1988).

Since plastid function depends on plastid gene products, it is essential
that chNA replication keep pace with plastid division to ensure that
both daughter plastids receive DNA. With this in mind, we have
considered the possibility that differences at the chNA origin of
replication could be responsible for the observed differences in plastid
multiplication intrinsic to the plastome type. Variation in the
efficiency of mitochondrial DNA replication has been suggested as an
explanation for the differential transmission of mitochondria in baker’s
yeast, Saccharggyces cerevisiae (Piskur 1988, Piskur 1989). The
mitochondrial genome of §; cerevisiae has eight potential DNA
replication origins and mitochondria with some of these origins and the
flanking regions deleted are transmitted less efficiently than wild-type
mitochondria. Since the chNA of the five Qgggtgggg plastome types
differ from each other by a number of insertions or deletions (Gordon gt

a1. 1982), it is possible that some of these length mutations might

37

affect the number of replication origins or the efficiency with which
the origins are recognized by nuclear-encoded proteins and,
consequently, plastid multiplication. As a first step to test this
hypothesis, we have located the origins of chNA replication in two
Oenothera plastomes that differ greatly in their transmission

efficiencies.

Materials and Methods

Plant material. Oenotherg hookgri str. Johansen, carrying its native
plastome type I, or plastome type IV from Q; atrovirens (Chiu gt gt.

1988), was used for this study.

CpDNA preparation. Young leaves (0.1-5 cm long) from actively growing
plants were used for DNA isolations. CpDNA from 30 to 50 g of leaves was
prepared according to the protocol described by Palmer (1982), with some
modifications. Sucrose gradient centrifugation was omitted, and chNA
was separated from nuclear DNA by two successive CsCl gradient
centrifugations in the presence of the flourescent dye bisbenzimide
(Sigma) at 0.01 mg/ml. The dye was removed by two extractions with
isopropanol saturated with 5M NaCl, and the chNA was dialyzed against
three changes of dialysis buffer (10 mM Tris, pH 8.0, 5 mM BDTA, 100 mM
NaCl) at 4°C, using a dialysis apparatus and collodion membranes with an
average retention of 25 kd (Schleicher & Schuell). Purified chNA was
stored at -20°C in small aliquots until immediately before use. CpDNA

was digested with restriction enzymes (GIBCO—BRL) in high salt buffer

38

(React 3) for two hours at room temperature and was then immediately
processed for electron microscopic analysis.

Electron microscopy and data analysis. Spreading of DNA for electron
microscopy was carried out according to Kolodner and Tewari (1975a).
X174 DNA and its RF form (obtained from GIBCO-BRL) were used as internal
standards for single-stranded and double~stranded DNA, respectively.
Grids were shadowed with platinum-palladium (80:20) and reinforced with
a thin layer of carbon. The samples were examined using a Philips 201
electron microscope. The electron micrographs were enlarged and the
contour of each DNA molecule was traced on a digitizing table (Calcomp
91480) connected to an IBM XT PC computer. The length of each DNA
fragment was calculated and converted to kb according to the internal
standards, using software written by the Michigan State University

Computer Center and Jeff Elhai (Michigan State University).

Results

Oenothera chNA replication is initiated through the formation of

multiple displms-nt loops

The preparation procedure yielded a relatively high frequency of intact
chloroplast DNA molecules. In the best preparation, nearly a quarter of
the intact chNM.molecules contained displacement loop (D-loop)
replication intermediates, which result from unidirectional synthesis of
one daughter strand (Kornberg 1980). Frequently, more than one D-loop

was observed on the same DNA molecule. These D-loops lay either within

39

several kb (Figure 2.1A) or quite far apart (est. 30-50 kb) from each
other (Figure 2.lB,C). Very rarely, D-loops could be observed at four
sites on the same molecule, as two distant pairs of D-loops.

D-loop initiation sites are located in both copies of the inverted

repeats

In order to localize the sites of D-loop initiation, chNA was digested
with SalI and PstI immediately prior to processing for the electron
microscope. Digestion of Oenothera CpDNA with Sell or Pstl yields 11 and
12 fragments, respectively (Figure 2.2). Except for minor DNA fragment
length differences, the restriction maps for these two enzymes are the

t l. 1982).

same for plastome types I and IV (Gordon

Digestion with restriction enzymes greatly reduced the frequency with
which D-loops were observed. A total of 54 SalI fragments bearing D-
loops (Figure 2.3) were characterized: 29 from plastome I and 25 frsm
plastome IV. Nine out of these 54 fragments contained two adjacent D-
loops (Figure 2.30). The sizes and numbers of the D-loop containing
fragments are presented in Figure 2.4A. According to the observations
with intact CpDNA, D-loops should be expected in at least two seperate
regions of Oenothera chNA (Figure 2.1B and C). As shown in Figiure 4A,
D-loops were most frequently found in SalI fragments between 15 and 20
kb, but they were also present in some larger SalI fragments (30—40 kb).
However, since the largest expected SalI fragment is around 30 kb
(Gordon t l. 1982), fragments much larger than 30 kb are most likely

products of incomplete SalI digestion. To calculate the observed sizes

40

Figure 2.1. Closed circular molecules of Oenothera chloroplast DNA
(plastome type IV) with D-loops. Arrows indicate the displaced strand.
$11174 DNA and its RF form (M) are used as internal standards for single
stranded and double stranded DNA, respectively. (A) Part of an intact
chNA molecule that contains two adjacent D-loops. Bar, lkb. (B) An
intact chNA molecule with two distant D-loops. Bar, 5kb. (C) Part of an
intact chNA molecule containing two D-loops approximately 30 kb apart.

The D-loop on the left has a short tail resulting from strand migration.
Bar, 5 kb.

41

I‘m
"KW...
‘. 9.9““: 1'
..,_ c. -

‘8»

_"~|'

l‘ ..‘t } '4

.23.?zfififi

1.11.3.5???
' J'v

It
’09

_ ...r . '

.1“; lfiyfilp ' S?
§m8s§ssf 1
0.: ' y "-

'

r

'1 , @386- 3.
' ‘0 ' .

(it .

A
i

. ) a
3 ' ‘ i.
<2, ." _- ‘7 ‘ ..

«yin ,‘ _‘ ’35:..wu
"g .1 11%? .3agg‘;%
A 4‘) PV%I~1“érm/é- “'11
~38 :- 11%;ng
4' .35.... as)?

is“ 8%»

i *7.
4 .1 fig??? ,- «.
Wank“ -5 43:: "

Pa§£§ - {951'
W “5

J _
egg; W1
.. ”ads .,
. . . .
a. ,. . ' ' 1“"
x ’V‘ 3‘
Tali-fiat- , f“. ‘
_ ‘5 . v

flit-gar» ‘9
m I)

.r.
. “ V'
3" 3;?

saw-13%? - f
std-‘3- .. . -

m.
1k" p ’ 1
“13,4. ’93-,»-
44133: 6455“
fill "r 1"

A»

'a }

 

‘*
.1;

on: D -
t? “155%“,
fifikkfis

 

fi

42

Figure 2.2. Restriction and partial gene map of Oenothera chNA.
From Gordon gt gt. 1982 and Herrmann gt gt. 1983. The large inverted
repeats, characteristic of DNA from most higher plant chloroplast
genomes, are outlined by black arcs. SalI and Pst I fragments are

labeled, with the largest fragment in each case designated as A.

43

 

 

st p £130
' J G A \PG’A
atpE D E
psbB
“”3 H \pefBD
rbcLI -
psbAI H L "”58
F
E D
K
C

5F 50” A /
rR J 9511 B rRNA

figL

with

Am
with

100;

Figure 2.3. Electron micrographs of chNA (plastome type I) digested
with Sall.

Arrows point to the displaced strand of each D-loop. (A) 18 kb fragment
with one D-loop, (B) 30 kb fragment with one D-loop, (C) Two adjacent D-

loops on the same fragment. Bar, 1 kb.

45

 

 

46

for D-loop-bearing SalI fragments, those SalI fragments smaller than 22
kb were grouped into one class and those larger than 22 kb the other.
For both plastome types I and IV (Table 2.1), the smaller class of D-
loop-containing fragments had an average size of 18 kb (n=23), and the
larger class had an average size around 32 kb (n=31). These sizes are
comparable to the third largest (Sal C, 18 kb) and the largest (Sal A,
30 kb) SalI fragments, respectively. Sal A and Sal C are two neighboring
fragments spanning most of the inverted repeats and the entire small

single capy region of the Oenothera plastome (Figure 2.2).

When plastome type I DNA was digested with PstI, D-loops were observed
on fragments ranging from 12 to 34 kb (Figure 2.4B). Those fragments
smaller than 20 kb have an average size of 14 kb (n=7), which is the
expected size of the major PstI fragment that overlap Sal C (Figure
2.2). The size distribution of those fragments larger than 20 kb are
quite scattered. However, judging from the position of the D-loops on
these fragments (see below), most of these fragments should represent

the PstI fragments (Pst B) that overlap Sal A (Figure 2.2).

If D-loops are located within the regions spanned by Sal A and Sal C,
double digestion of chNA with PstI and Sell should also yield two
fragments that carry D-loops (Figure 2.2). For both plastome types I and
IV, the double digestions of PstI and Sell resulted in two size groups
of D-loop-containing fragments (Figure 2.40) with average sizes around
17 kb and 36 kb respectively (Table 2.1). Although the same sample gave
the expected patterns for PstI and SalI double digestion when visualized

on an agarose gel, the average sizes of these D-loop-containing

47

Figure 2.4. Length distribution of chNA fragments containing D—loops

from plastome types I and IV.

Fragments generated by digestion with (A) SalI (B) PstI (C) PstI + Sall.
CpDNA fragments of plastome type I are represented by filled bars, while
those of plastome type IV are represented by empty bars. The vertical
axis indicates the number of fragments in each size class observed to

have D-loops.

NUMBER OF FRAGMENTS

48

 

 

 

 

 

 

 

 

A Sal

“ : [H111

'* B Pst

1 .

.L

~» C Pst I Sal

1 I

T : l. U + A

5 1O 15 20 25 30 35 4O

FRAGMENT LENGTH (kb)

 

49

fragments are larger than the lengths of the expected fragments (Table
2.1). Conceivably, complete SalI and PstI double digestions bring D-
loops closer to the end of the fragments and that might have rendered

the D-loops less stable due to strand migration.

To orient the positions of the D-loops on the restricted chNA
fragments, the location of D-loops on each fragment was measured. A
length, d, was defined as the shortest distance from an end of the
fragment to a terminus of the D-loop. It is clear that D-loOps can occur
at two adjacent sites on the same region of the chNA (Figure 2.1A). The
proximal terminus was chosen as the point of measure for the D-loops
closer to the end of the fragment and the distal terminus was chosen for
the more distant D-loops. Each terminus of the D-loops defined in this
way probably coincides with the replication origin for one daughter
strand, according to the model for the initiation of chNA replication
in land plants (Kolodner and Tewari 1975b). In fact, since most of the
D-loops are relatively small in size, the placement of a D-loop was not
affected greatly by the choice of terminus. Two classes of d were
observed and the measurement in these classes were grouped as dh and d2

(Table 2.1).

If the D-loops on fragments Sal A and Sal C lie within the inverted
repeats, then D-loops on both PstI fragments should be 2 kb closer to
the end of the fragments than D-loops on the SalI fragments (Figure
2.4). As shown in Table 2.1, this prediction is fulfilled since db and
d2 from PstI fragments are about 2-3 kb shorter than measurements

derived from the related SalI fragments. The shortening of arm length

50

Table 2.1. Distribution of D-loops in chNA of Oenothera plastomes I and

IV

Enzygg Plastome d1'(n) gg'(n)
SalI I 3.610.6 (6) 8.110.3 (5) 18.010.
IV 4.210.5 (7) 9.110.5 (5) 18.010.
I 6.110.? (13) 11.310.8 (10) 32.511.
IV 6.310.5 (8) 10.910.4 (8) 31.311.
PstI I 1.910.4 (6) 5.610.3 (3) 14.110.
I 3.010.8 (2) 8.210.9 (9) 26.811.
PstI I 3.010.3 (9) 7.010.5 (12) 17.510.
+ IV 3.010.4 (3) 7.2 (1) 16.610.
SalI I 4.210.4 (6) 10.910.7 (6) 36.011.
IV 4.310.6 (5) 9.910.5 (7) 35.811.

 

5 (ll)
8 (12)
2 (18)
6 (13)

6 (7)

4 (12)

5 (19)
3 (4)
0 (10)
0 (10)

Sal
Sal
Sal

Sal

Pst

Pst

Pst
Pst
P+S

P+S

>0

Otg. length (n) Egg, lengthb

17.9
17.8
30.2
30.2

14.2

28.4

14.2
14.2
28.2

28.2

 

'Average distance 1 standard error (in kb) between the terminus (as

defined in text) of D-loop structure and the nearer end of the

fragment.

bLength expected from physical map of Gordon et a1. (1982). Sizes

expressed in megadaltons in the original paper have been converted to

kilobase pairs.

51

coincides well with the position of PstI sites on Sal A and Sal C
(Figure 2.4). This placement of D-loops would further predict that the
shorter distance from a D-loop to the end of a SalI fragment should be
the same for both Sal A or Sal C. However, this distance is about 2 kb
larger in the case of Sal A than 8810 (Table 2.1). This discrepancy can
be explained if many of the larger SalI fragments that contain D-loops
are the products of partial digestion: Sal A joined to the adjacent Sal
K. Consistent with this interpretation, the average length of the larger
SalI fragment is 2 kb greater than the expected length (Table 2.1).
Furthenmore, two distant D-loops are observed to be at least 30 kb apart
on intact chNA molecules (Figure 2.18, C). This distance can be
explained only if the D-loop initiation sites are located toward the
distant ends of the two neighboring SalI fragments within the inverted

repeat.

Each pair of’D—loop initiation sites is in the close vicinity of rRNA

In order to see the positions of D-loop initiation sites relative to
genes on the plastome, most of the Sell or PstI fragments containing D-
1oops were aligned to the end of their short anms (Figure 2.58, C and
D). Fragments that are much longer than the predicted sizes due
presumably to incomplete digestions were not included in this analysis.
Either Sal A or Sal C fragments that were slightly longer and may have
included the neighboring Sal K were aligned in Figure 2.5 to the
appropriate SalI site 2 kb away. In some cases, the original alignment

placed the D-loop initiation sites within the 168 or 23S rRNA genes.

52

Figure 2.5. Proposed alignment of D-loop-containing molecules relative
to the rRNA operon and the small single copy region of Oenothera CpDNA.
(A) Frequency distribution of the D—loop structures, summing data from
plastomes I and IV. Physical map of the Sal A (right) and Sal C (left)
region. The positions of genes within the rRNA operon are indicated with
open boxes, with 168 and 238 indicating the rRNA genes, I designating
ttgt and A designating ttgt. The most likely location for the D-loop
initiation sites are indicated with gttt and gth. (B-D) Scale of lines
indicates length of fragments containing D-loops (indicated as black
boxes): (B) plastome type I DNA digested by PstI, (C) plastome type I

DNA digested by SalI. and (D) plastome type IV DNA digested by SalI.

Type I
Pst I

Type I
Sal I

Type N
Sal l

\r

53

d

woman:
0"
X
0'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\/_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ -
- - 4‘
__I_i J
- __ l
- #-
n #
- #
- J
L
L
A
- 4]
A J
— J
- 4
- Q
- l
- g
- 4‘
— j
- _ I
-
l
L
4+
L J
_ 41
-
i J
- I
L —l
- #1
- #1
- #1
-
_ ‘1

 

 

 

 

54

Since the rRNKA genes are highly concerved, and have never been cited to
function as origins of DNA replication, We realigned these fragments to
alternative restriction sites. Restriction analysis of the cloned 14 kb
Pst F fragment from both plastome types revealed only a 100 bp deletion
in plastome type IV relative to type I (Chapter 3), and therefore, data
from both plastome types were pooled to establish a more precise
localization of D-loop initiation sites. As shown in the D-loop
distribution frequency histogram (Figure 2.5A), two regions appear to be
preferential sites for D-loop initiation on both copies of the inverted
repeat: One (9511) is l - 2 kb upstream from the 168 rRNA gene and the
other (911g) is in the spacer of the rRNA operon near the tRNA-Ile gene.
The distance between the two spacer regions is about 30 kb, which

coincides with the distance between two distant D-loops in Figure 2.10.

Discussion

Both the transmission efficiency of Oenothera plastids and their
competitive multiplication rates are influenced by the plastome type
(Schfitz 1974, Chiu et a1. 1988). Since plastids can multiply even in the
absence of protein synthesis within the plastid (reviewed by Boffey
1985), the different multiplication efficiencies of Oenothera plastids
may be attributable to differences in the non-coding components of the
plastomes, such as the origins of chNA replication. This possibility

inspired the present study.

CpDNA replication in Oenothera was found to be initiated through the

55

formation of D-loops (Figure 2.1). One pair of D-loop initiation sites
was mapped to each copy of the large inverted repeat (Figure 2.2) of the
plastome: one lies 1 - 2 kb upstream from the 168 rRNA gene, while the
other lies in the spacer between the two rRNA genes (Figure 2.5). In
these respects, plastome type I, having the highest transmission
efficiency, is indistinguishable from plastome type IV, having the
lowest efficiency in transmission. Hence, differential plastid
multiplication in Oenothera cannot be explained simply by differences in
the number or placement of functional chNA replication origins present
in the plastome. However, this does not preclude the possibility that
sequence differences at the replication origin could affect the

efficiency of replication initiation.

In order to locate the replication origins, care was taken to preserve
unstable D-loop replication intermediates in the purified DNA. All the
restriction digestions in this study were performed at room temperature
for a limited amount of time to minimize strand migration. This may have
contributed to the frequent appearance of partial digestion products in
the DNA spreads. Despite the measures taken to minimize strand
migration, D-loops were observed far less frequently in digested chNA
than in uncut samples. It is conceivable that D-loops lying near the
ends of fragments or on small fragments may have been lost. However, the
relative locations of D-loops on uncut DNA are consistent with the four

D-loop initiation sites determined on the restricted fragments.

In no other organism besides Oenothera has chNA replication been shown

to proceed by two pairs of D-loops. In maize, chNA fragments that can

56

initiate replication tg vitro have been identified (Gold gt g1. 1987).

Whether these fragments can initiate chNA replication i vivo have not

 

been confirmed by electron microscopy. Pea, the only other land plant in
which the D-loop initiation sites of chNA have been localized, has only
one pair of chNA replication origins near the rRNA operon: one in the
spacer region within the rRNA operon, as in Oenothera, and the other
several kb downstream from the 23S rRNA (Meeker gt gt. 1988). However,
the plastome of pea differs from most other land plants in that it has
only one copy of the rRNA operon (Palmer and Thompson 1981). Since the
rRNA operon generally is found within the large inverted repeat that is
common to the chNA of most land plants (reviewed by Palmer 1985)
including Oenothera, it is likely that the number and the placement of
replication origins in the Oenothera plastome is typical of those in
plastomes of most other land plants. The placement of replication
origins near the rRNA operon may have some functional importance since
several algae also have their chNA replication origins near the rRNA
operon although the exact position differs with respect to rRNA genes
(Holler and Delius 1982, Ravel-Chapuis gt gt. 1982, Weddell gt gt. 1984,

preliminary results in Yamada gt gt. 1986).

The two adjacent D-loop initiation sites in the Oenothera plastome may
represent separate origins for the replication of each strand, as
suggested in studies on chNA replication in pea and maize (Kolodner and
Tewari 1975a, b) and mitochondrial DNA replication in animals (reviewed
by Clayton 1982). It has been proposed that unidirectional synthesis of
a daughter strand from each neighboring origin results in a replication

eye (Cairns structure, Cairns 1963) with two double stranded arms, from

57

which bidirectional DNA synthesis proceeds (Kolodner and Tewari 1975b).
Although no such structures were observed in this study, replication
forks with two double stranded arms were observed in some fragments. In
chNA of the algae Chlamydomonas (Weddell gt gt. 1984), Egglena (Holler
and Delius 1982; Ravel-Chapuis et a1. 1982) and Acetabularia (Santulli
gt gt. 1983), loops with two double stranded arms are the major or only
replication intermediates that have been observed. Apparently, in these
organisms, either the same origin is used for the replication of both
strands or the replication origins for each strand are very close to
each other. It will be necessary to analyze chNA replication tg gtttg
(Wu gt gt. 1986, Meeker gt gt. 1988) to locate the exact sites of
replication initiation in plastids of Oenothera and to reveal further
features of the chNA replication origins that may be common to land

plants and algae.

CHAPTER 3
PROLIFERATION OF DIRECT REPEATS NEAR THE OENOTHERA CHLOROPLAST

DNA ORIGIN OF REPLICATION

Introduction

Plastids can be transmitted from both parents in crosses of Oenothera,
the evening primrose (reviewed by Kirk and Tilney Bassett, 1978),
resulting in progeny in which two different plastid genomes (plastomes)
may coexist in the same cell. This biparental plastid transmission
facilitates the study of plastome-plastome as well as plastome-genome
interactions. Five major types of plastome are found among species of
the subsection Euoenothera (reviewed by Stubbe, 1989), and the relative
efficiency of plastid transmission is strongly influenced by the
plastome types involved in the crosses (Sch:tz, 1974, 1975, Chiu gt gt.
1988).

As a first step towards understanding molecular factors contributed by
the plastome that may affect the efficiency of chloroplast multipli-
cation, origins of Oenothera chloroplast DNA (CpDNA) replication were
localized through electron microscopic analysis of restriction fragments
carrying replication intermediates (Chapter 2). CpDNA replication in
Oenothera is initiated through the formation of displacement loops (D-
1oops). One pair of replication initiation sites was found within each

copy of the large inverted repeat of chNA from both plastome type I,

58

59

which has the highest transmission efficiency and plastome type IV,

which has the lowest. One D-loop initiation site (oriA) lies about 1-2

 

kb upstream from the rRNA operon, and the other (oriB) lies in the

spacer between the 168 and 23S rRNA genes.

In this chapter, restriction fragment length polymorphisms in regions
surrounding the D-loop initiation sites were analyzed to search for
local sequence variations that might affect the efficiency of repli-
cation initiation. A highly variable region was identified in the 16S
rRNA-ttgt spacer near a D-loop initiation site (911;) of Oenothera
plastome. Analysis of this intergenic region revealed multiple duplica-
tions of related sequence motifs. Mechanisms that might be responsible
for the hypervariability in this region and the possible impact of these

length mutations on DNA replication will be discussed.

Material and Methods

Plant material. Wild-type Oenothera plastomes from the following strains
were used in this study: Oenothera hookeri str. Johansen carrying the
Dusseldorf line of plastome type I (ID) (Chiu gt gt. 1990), or plastome
type II, III or IV (Chiu gt gt. 1988); and Oenothera hookeri standard
with its native plastome type I (13“). The plastome from Cornell-l
(Isl), a derivative of ID, was obtained from 91 hookeri str. Johansen
homozygous for the plastome mutator allele (Chiu t l. 1990). Chloro-

plast DNA was isolated according to Chiu gt gt. (1990).

60

Restriction analysis and sequencing. Enzymes for restriction analysis
and cloning were obtained from GIBCO—BRL. SalI fragments that were 18 kb
in size (Sal C, Figure 3.1A) from six Oenothera plastomes were purified
by phenol/chloroform extraction from 0.8% low melting agarose following
gel electrophoresis in TAB buffer (Maniatis gt gt. 1982), and cloned
into the SalI site of pRL498 (Elhai and Walk 1988). The largest EcoRI
fragments (R2, the second largest EcoRI fragment of Oenothera chNA,
approximately 5 kb) in Sal C from six Oenothera plastomes were subcloned
into the EcoRI site of pBR328. Subsequent electrophoresis of restricted

DNA was performed in TBE buffer (Maniatis gt gt. 1982).

EcoRI-BssHII subfragments of R2 (Figure 3.1A) from the two type I
plastomes, ID and 151, were obtained by cutting the two R2 clones
(pOjD-RZ and pOjCl-RZ) with BssHII, filling in the ends with Klenow
fragment of DNA Polymerase I, and then cutting with EcoRI. The fragments
thus obtained were ligated with pICZOH (Marsh gt gt. 1985) that had been
digested with EcoRI and SmaI. Plasmids containing the smaller EcoRI-
BssHII subfragment from R2 of ID and 1°1 were identified according to
the sizes of their inserts and were named p0jD-BBl and pOjCl-BBl
respectively. Deletion derivatives of pOjD-BBl were obtained using
exonuclease III (Henikoff 1984). First, pOjD-B81 was digested with SstI
and EcoRI, cutting in the polylinker region derived from pICZOH. The DNA
was then treated with exonuclease III for different lengths of time,
resulting in unidirectional digestion starting from the EcoRI end, and
the enzyme was inactivated at 70°C fer 15 minutes. The ends of the
resulting fragments were made blunt with mungbean nuclease and rejoined.

Two derivatives were obtained by this method: pOjD—BBZ and pOjD-BE3,

61

carrying 600 bp and 400 bp inserts, respectively. Other deletion
derivatives of pOjD-BEl and p0jCl-BEl were obtained by cutting the

original plasmids with either HincII or BamHI and religating.

The DNA sequences of the inserts in pOjD-BEl and pOjCl-BEl were obtained
using Sequenase (United States Biochemical Corporation) and methods for

double stranded DNA sequencing (Zhang gt gt. 1989). To obtain sequences

from both strands, all the deletion clones of pOjD-BEl and p0jCl-BE1

mentioned above were also used.

Computer programs used for sequence analysis. A computer program,
BDITBASE, for two dimensional comparison of nucleic acid sequences
provided by Niels Nielsen (Purdue University) was used to search for
repeated sequences. Regions capable of forming secondary structures were

found using PC-FOLD (Zuker 1989).

Results

One D—loop region is proximal to a highly variable region of the

teeth—rs plate-e-

The third largest SalI fragment (SalC) from Oenothera plastome types I
and IV contains two D-loop initiation sites (Ch. 2). This fragment was
cloned from six Oenothera plastomes and analyzed after restriction with

EcoRI (Figure 3.1A and data not shown). No length polymorphism was

62

Figure 3.1. Localization of the hypervariable region in the rRNA operon
of Oenothera CpDNA. (A) Physical map of SalC (top) and R2 (bottom)
fragments of the Oenothera plastome. The genes for 168 rRNA and 23S
rRNA, tggt (I) and ttgt (A), are represented with filled boxes. The
region of variable size is represented with a wavy line. BamHI sites
within this region are not shown. The two preferential sites for D-loop
initiation (9511 and 221E) are indicated with dashed lines since their
precise locations are still unknown. (B) Restriction digestion patterns
of the R2 fragments from six Oenothera plastomes. Plasmids, each
containing an R2 fragment inserted into the EcoRI site of pBR328, were
digested with EcoRI + PvuII + XhoI. Lanes contained digested R2 clones
from plastomes I”, IV, I", II, III and 1‘”. Molecular weight markers
(M) consisted of lambda DNA digested with EcoRI and HindIII. Fragments
marked a, b, and c are indicated in the diagram in panel A, and fragment

V is from the pBR328 vector.

63

IDIYIHKIIIIIPM kb

 

 

64

detected for the 4 kb subfragment that contains one D-loop initiation
site (9111) located upstream from the rRNA operon. However, the size of
the BcoRI fragment that lies internal to the rRNA operon varied among
the Oenothera plastomes that were examined. This approximately 5 kb
fragment (R2) contains gttt, a second D-loop initiation site (Figure
3.1A). R2 fragments derived from these Oenothera plastomes were

subcloned into p8R328 and analyzed further.

A conserved PvuII site at the end of the gene for tRNA-Ile (ttgt) and an
XhoI site in the ttgt intron are present in the rRNA operons from
chloroplasts of several land plants (Massenet gt gt. 1987). Restriction
digests showed that these two sites are also present in Oenothera
(Figure 3.1A). Restriction analysis of the cloned R2 fragments showed
that the length variation among six Oenothera plastomes is contained
within the smaller EcoRI/Xhol fragment (fragment b in Figure 3.1).
Further restriction analysis confined the variable region to the
sequences bounded by EcoRI and BssHII (Figure 3.1A and data not shown).
This region covers sequences separating the 168 rRNA gene from ttgt and
is close to gttfi. The fragment from five wild-type Oenothera plastomes
ranged in size from 400 bp to 900 bp. The variable fragment in plastome
191, a variant of ID, is much smaller than that of its wild-type
predecessor and has a size similar to that of plastome type III (Figure

3.18).

Sequence comparison of the 16S rRNA-trnI spacer.

The DNA sequences of the variable 8coRI-8ssHII fragments (Figure 3.1A)

65

from I0 and IC1 were determined. The 1040 bp DNA sequence from 1° is
shown in Figure 3.2 and is compared with the same region from the
tobacco plastome. Overall homology between the two sequences is quite
high (90% sequence identity), excluding two insertions in Oenothera
chNA relative to tobacco: one of 28 bp (positions 388 to 415) and the
other of 516 bp (positions 497-1012). The larger insertion occurs

immediately 5’ to trnI.

Analysis of the 516 bp insertion revealed multiple repeated sequences.
For example, positions 521 to 665 (segment at in Figure 3.2) is a 125 bp
duplication of positions 342 to 520 (segment a in Figure 3.2) with two
deletions of 28 and 24 bp, and two small insertions (lower case letters
in a‘, Figure 3.2). The 28 bp deletion in at makes this region more
similar than a to the equivalent region in the tobacco plastome. The
smaller duplication, b (109 bp) and bi (102 bp), consist of DNA sequence
that can also be found at the junction between a and at (positions 498
to 536) and three segments within at (positions 599 to 612, 599 to 617
and 627 to 643). Interspersed between these two major repeated segments
are several 17 bp direct repeats (symbolized by a long arrow) first
appearing at the end of a. This 17 bp segment is almost identical to
part of the 5’ ttgt exon (15/17 match). The 5’ end of the 17 bp repeats
is usually accompanied by tracks of (ACG)n, which also appear at the 5’

junction between trnI and the spacer in tobacco plastome.

66

Figure 3.2. Sequence comparison of the 8coRI-8ssHII fragment from
Oenothera hookeri str. Johansen plastome ID (Oj.) and N. tobaccum
(Nt.). Tobacco sequences identical to that of Oenothera are not shown.
Base substitutions in tobacco relative to Oenothera are indicated with
capital letters while missing bases in both sequences are represented by
dashes. Two families of large repeated segments in Oenothera are
indicated as a, at, b and bi. The positions and sizes of two deletions
in a! relative to a are indicated byzk, with the number of base pairs
deleted indicated underneath. Long arrows represent a repeating 17 bp
unit. Double lines indicate a repeating 16 bp unit. Wavy lines indicate
the sequences equivalent to the region immediately upstream of ttgt in
tobacco plastome. Horizontal brackets indicate the repeating ACG
element. Lower case letters represent inserted or substituted bases in

the repeated units.

67

33. GAATTCGTTCCCGGGCCTTéTACACACCGCCCGTCACACTATBGGAGCTSGCCATSCEZGAAGTCGTTACCTTAACCACA
t. 0

O O 100 O O I C O
05. AGGAGGGGGATGCCGAAGCGAGGCTTGGTGACTGGAGTGAAGTCGTAACAAGGTAGCCGTACTGGAAGGTGCGGCTGGAT
t - - A -

° lGSrRNA] . . . 200 . . .
01. CACCTCCTTTTCAGGGAGAGCTAATGCTTGTT66GTATTTTGGTCTGACTCTTCCTTTTCA--CCC---AAAACGAGGCC
Nt. T A - a 0A cc AAA A A 00

. . . . . . 300 .
:j. A6CTACATCTGAGTgAAACTTGGAGATGGTAGTCTTCTTTCGTTTCTCGACBGTGAAGTAAGACCAAGCTCATGAGCTTA
t. A C

. . . [a . . . . .
Oj. TTATCCTAGGTCGGAACAAGTCGGTAGGATCCCCTTTTGSACGTCCCCATGCCCTTTCCSCGCG-G-GGTAG-CATGCTC
m. «0 TA 0 n -u -cc TTC-A -u
01. scccsTTchacsc1000000CAtsssssésAAAAAAssAAssAsscseAtssssTTTciCchCTTTTéAcgleggggg
Nt. ------------------------- A A s 0 -~-

. . 500 . a 8* . .

03. 0 scccc-s s 0 c cgetchgATTAGCTcATchtgggATcccCTTTTcsACGTCCCCAIGCCCTTT
Nt. CA I -----------------------------------------------------

. . . . . 600 .
OJ. CCGCGCGGGGTAGCATGGBGGCGAAAAAGGAA tallatsaggaggcTTTGA T 60 GCGG

.... ............... 26? baufiu--- ............................................
. . . . 1*] . ' . ' . 700
Oj.(5&5255255555C60ttagATTAGCTCATTGGTAQfiEEgetgATTAGCTCATTGSTA c c TTAGCTCATTGGTA
"to -"is-""-°"'""'"-""""""'""'"‘""'"°"""°'""""""°""”'f """"
OJ..EEACQACQATTAGCTCgttgétATTGGTAEGAT cc ' c TTGAtalegggsggéAT0Aq535§§15553§959_
Nt. --------------------------------------------------------------------------------
. 000 . b][b* . . . .
OJ.Ԥ;ASCGAGTC c c c qeggegggcgfichTTAGCTCgttggtATTSGTAQEATCCCCTTTTGGACGTTG 0
Mt. ---------------------------------------------------------------------------- 312;;

. . . 900 . b*] . .
03. agcssATaAchAggAscssccgggAceAstsggggbggeggeggAc c c c c TTAGCTCATTGGTA 0 c
Nt. ......................... inlh ............................................ ' ......

. . . . . 1000 . [trnI .
03. tgAITAGCTCATTsstAquTTAGCTCAgTGtTAs tTA CACCCaGTGGGACSGCAC c GGCTATTAGCTCAQTG
Nt. ---------------------------------------------------------------

03. STAG scsccé
m. 4

 

 

 

 

68

The sequence of the duplicated region in Oenothera has the potential to

form stable hairpin structures.

The duplicated segments a and at (Figure 3.2) of Oenothera plastome 1°
contain two regions that can form very stable hairpin structures (Figure
3.3). The stem-loop structure in segment at (base pairs 526 to 578) has
a similar structure in the tobacco plastome (stem 104 in de Lanservin gt
.gt. 1987). The stem-loop structure in the analogous position of segment
Ia (base pairs 346 to 429) is larger by 28 bp, owing to an insertion,
relative to at, that extends the palindrome (Figure 3.2). The smaller
hairpin structure (Figure 3.38) can form in both a (positions 469 to
491) and at (positions 609 to 631), and is very similar to a structure
in tobacco (stem 107 in de Lanservin gt gt. 1987). Besides these two
stable hairpin structures, another 25 bp segment in a (positions 433 to
457) also has the potential to from a stem-loop structure. This segment

is absent in a! due to a 24 bp deletion.

Most of duplicated sequences in the spacer have been deleted in the

chNA of a strain derived from the plastome mutator line.

Plastome ICl arose from ID under the influence of the plastome mutator
gene (Chiu gt gt. 1990). The 604 bp BcoRI-BssHII fragment from ICl has
the identical nucleotide sequence as that of ID except for two large
deletions (Figure 3.4). The first deletion removes the 179 bp segment a
completely, while the second deletion removes 257 bp (positions 712 to
968 including segment b, bt and two units of the 17 bp direct repeat

with accompaning ACG tracks. No base substitutions were found.

69

Figure 3.3. Secondary structures in the repeated segment a and at of
Oenothera plastome ID and the equivalent region in tobacco chNA.

A. The first stem-loop in (l) tobacco GAG = -41.9 Kcal/mole), in (2)
Oenothera segment at (AG = -43.2 Kcal/mole and in (3) Oenothera segment
a.(AG = -76.5 Heal/mole). B. The second stem-loop in (l) tobacco GAG =
-22.2 Kcal/mole) and in (2) Oenothera segment a and a! (AG = -26.9
Heal/mole). The numbers at both ends of the stem-loop structures

correspond to the positions of the bases in Figure 2.

70

C nwlfl TG AA m
nwoo cc 006060 66 GGGCATGGGGGCG AAAAGG
a: __ :2: __ _____________ ______u
TACG 66 660606 CC CCCGTACCCCTGC TTTch
AT TTT AG % I&
G 5 A we
GGGGGCCCGC)
m _ _ _ _ _ _ _ _ _ .1...
GCCCCGGGGG
a...
TA A E 46
6066 CG GOATGGGGGCG AAAAGG)
c __ __ ___________ ______u
600 CC CGTACCCCTGC TTTch
T T A G
T G W
TGGm¢GCTCGC
G__________U
Accccceeece

70

C Aw...“ TG AA m
n.cuGc cc 006060 66 GGGCATGGGGGCG AAAAGG
e: __ :2: __ _____________ ______m
TACG 66 660606 00 CCCGTACCCCTGC TTTch
AT ...TT AG 5 I...
G 2 $3
5 A 6
6666606060)
m _ _ _ _ _ _ _ _ _ m
GCCCCGGGGG
a...
TA A m 46
GCGG GG GOATGGGGGCG AAAAGG)
c __ __ ___________ ______m
600 CC CGTACCCCTGC TTTch
TT A G
T G W
TGGmVGCTCGC
G__________H.
ACCCCCGGGCG

71

Figure 3.4. Restriction map of the variable EcoRI-BssHII fragment in
Oenothera plastomes 1° and IC1. Two major families of repeated segments
are indicated as a and as, b and bi. The symbol indicates the positions
where stem-loop structures may form. Small arrows represent the 17 bp
direct repeat and each dot represents pairs of the 3-bp ACG motif shown
in Figure 2. The 16-bp repeat in Figure 2 is not indicated seperately,

because each copy lies in each of the major repeats, a, at, b and b3.

72

\
S
a?
trnI

 

 

 

QIJ 4
\ I I
rsQQl AMfi 11/ II
a: ’1 n
\%6001 b [IA/y" 4
L w
m. . . ...“ A
min.“ *GQQM 41W uuuuuuuuu u
a x...
\ \
16o Y.
s
\Qoow W
s
6
\Q00 1
0.... ml

73

Multiple repeats are not stable in the evolution of Oenothera plastomes.

Since there is one BamHI site (5’GGATCC3’) at the junction between the
17 bp and the 16 bp repeats (symbolized by an arrow and a double line
respectively in Figure 3.2) in each major repeat unit of plastome I°
(Figure 3.2 and 3.4), the presence or absence of the major repeated
segments in other Oenothera plastomes can be deduced by comparing the
number and position of BamHI sites with that of plastome 1° and IC1
(Figure 3.5A). The results of the restriction analysis are summarized in
Figure 3.58. The five fragments of 1° in Figure 3.5A and B are numbered
according to their increasing sizes. Plastome type IV has a novel
fragment in place of fragments 1 and 3, hence it probably does not
contain segment b. Plastome I"K from Qgpotherg hoogggi std. has a novel
fragment instead of fragments 2 and 3, thus it probably does not carry
segment at. Plastome type II probably lacks segment at and b since a
novel fragment replaces fragments l, 3 and 5. Plastome III has exactly
the same number and sizes of restriction fragments as that of plastome
I01, hence, it probably has only segment 8*. Subsequent sequencing of
plastomes I3“ and 11 (Sears, unpublished results) using primers deduced
from the I° sequence have confirmed the predictions for those two

plastomes.

74

Figure 3.5. Comparison of the variable lSS-trnl spacer among Oenothera

plastomes.

A. Ethidium bromide-stained gel of 1.8% agarose with DNA from 82 clones
of Oenothera chNA digested with EcoRI and BamHI. The five small
fragments from plastome I° are numbered according to their sizes. 8.
Deduced restriction maps of the variable region in R2 fragment. Gaps
indicate regions that are absent in each plastome. Fragment 5 is
bracketed by two BamHI sites with one of them outside of the map (see

Figure 1A).

75

 

 

 

 

 

 

 

 

76

Discussion

A highly variable region of Oenothera chNA was found to be located near
one of the D~1oop initiation sites (911g) in the intergenic region
Legend to Figure 3.5 here

separating 168 rRNA and tggt genes. This variability stands in contrast
to the conservation of this region in other land plant plastomes (de
Lanversin gt gt. 1987, Massenet gt gt. 1987, Markowicz gt gt. 1988). The
size of this intergenic region ranges from approximately 410 to 845 bp
in six Oenothera plastomes, compared to a range of 273 to 310 bp in the
six plastomes of land plants that have been sequenced (Koch gt gt. 1981,
Ohyama gt gt. 1986, Shinozaki gt gt. 1986, de Lanversin gt gt. 19871,
Massenet gt gt. 1987, Hiratsuka gt gt. 1989). The longest of the 168
rRNAetggt spacer region examined (from plastome 1°) is very similar in
sequence to the same region in the tobacco plastome, except for two
additional stretches of DNA. Restriction analysis and DNA sequencing of
this region from five wild-type Oenothera plastomes indicates that it is
rich in repeated sequence motifs and sequences capable of forming stem-
loop structures (Figures 3.2, 3.3, 3.5). It is likely that this region
of the Oenothera plastome evolved frsm an ancestral sequence similar to
that of tobacco and the initial expansion of this intergenic region in

Oenothera probably occurred before the divergence of plastome types.

The major repeat units (a, at, b and bi in Figure 3.2) in the 165 rRNA-
trnI spacer of plastome I° are flanked by numerous short direct repeats.
Slipped-strand mispairing, also known as replication slippage (Moore

1983, Levinson and Gutman 1987, Allgood and Silhavy 1988), has been put

77

forward as a.mechanism to explain the occurrence of DNA length mutations
mediated by short direct repeats in bacterial as well as eukaryotic DNA.
The slipped-strand mispairing model proposes that the resumption of DNA
replication after slippage and mispairing of the daughter strand with a
preceding short homologous sequence duplicates the region between two
short direct repeats (Figure 3.6A). On the other hand, mispairing
between one short direct repeat on the daughter strand with the other
repeat unit ahead of the replication fork would cause the loss of one of

the repeated units and the region between the two (Figure 3.68).

In bacterial DNA, slipped-strand mispairing is facilitated by conditions
that stabilize or create single-stranded DNA, such as the presence of
palindromes or the mutation of single-stranded DNA binding protein
(reviewed by Allgood and Silhavy 1988). The hypervariable region in the
rRNA operon is between the two neighboring D-loop initiation sites (9111
and gttfi in Figure 3.1A) of the Oenothera plastomes. Initiation at
either origin would create a displaced strand of parental DNA between
the two replication origins (Kolodner and Tewari 1975). Indeed, single-
stranded DNA up to 3 kb has been observed in the Oenothera plastome
(Chapter 2). Furthermore, several stem-loop structures that could be
fOrmed in the major repeat units of plastome I° (Figure 3.3) may serve
to bring distant short direct repeats into close proximity (Glickman and
Ripley 1984). Overall, this intergenic region presents itself as an

ideal site for slipped-strand mispairing.

Figure 3.7 shows a plausible sequence of events using slipped-strand

mispairing that may have led to the present structure of the 168 rRNA-

78

Figure 3.6. The generation of duplications (A) or deletions (8) by
slipped-strand mispairing of direct repeats during replication of DNA.
Parallel lines represent complimentary strands of DNA, while shaded
boxes represent short, direct repeats (SDR). In (A), DNA replication
continues after slippage and mispairing of the daughter strand with a
preceding SDR on the parental strand. This duplicates the region between
two SDHs. In (8), DNA replication continues after mispairing between one
SDR on the daughter strand with the other 808 ahead of the replication
fork. This causes the loss of one of the SDRs and the region between the

two.

A

$081 SDRV/

W
if

12
f
1.1%

B
$51 SDRV/
W

——m/flm/
_m\%\

47%

2.

80

Figure 3.7. Major steps in the expansion of the 16S rRNA-ttgt spacer of
Oenothera plastome I° starting from sequences similar to other land
plant plastomes. Symbols used to represent major direct repeats are the
same as those in Figure 2 except for the ACG element, which is
represented by a dot. Each step represents a duplication event mediated
by slipped strand mispairing during DNA replication (Figure 6A).
Sequences of the short direct repeats proposed to have mediated the
mispairing are shown above the horizontal lines with the matched bases
underlined. The numbers indicate the position of these sequences in
Figure 2. Minor addition/ insertion events and base substitutions

required to produce the ultimate sequence of plastome I° are not shown.

81

AGTCGGTAG AGTGGTAG
3&— ——540 ICE- "—7033

(trn I—o

 

 

 

  
 
 
  
    

 

 

 

 

Step1
AGCAGQGG AQAGQCX‘BC
605- .. osz 040
: ”II-9—
s a
Ste 2
p Aeocogsg GGCAQCC
62 :t 89 995
——==———-m~ 1 «‘
: s 1 at (trn l-’
Ste 3
,, " \V" -
I I I!!! . (tn: |—0
Ste 4
p Ac ...-«nee / Ace ...... T Ass
712 _, z,
= .- ,_ A -
p—. 4 u L b “"1 I".
Stap5

-v—
a at e b : haw—4 (trn |—’

82

ttgt region of Oenothera plastome 1°. It is presumed that the ancestral
sequence of this region was similar to that of the tobacco plastome. As
illustrated in Figure 3.7, the first major step is probably the
duplication of the region bracketed by two short direct repeats, one
(5’AGTCGGTAG3’)immediately upstream from segment a and the other
(5’AGTGGTAG3’) within ttgt, due to mispairing of one short direct repeat
on the daughter strand with the preceeding short direct repeat on the
parental strand (Figure 3.6A). In the subsequent step, the region

between 5’AGCAGCGGGC3’ in at and a similar segment 5’AGAGCGCGCB’ in trnI

 

is duplicated. These two duplication events form the precursor of at.
The duplication of the 17 bp motif (symbolized with an arrow) and the
accompanying ACG elements (symbolized as dots) upstream to ttgt (step 3)
may be facilitated by the mispairing of two flanking segments,
5’AGGCCCGC3’ and 5’AGGCACCC3’. This duplication must occur prior to the
fermation of segment b, which is accomplished through the mispairing of
two 17 bp motifs (step 4). Subsequently, segment b itself is duplicated
through a similar process (step 5). Replication slippage may also be
responsible for the deletion of a 24 bp segment in at (Figure 3.2),
since the equivalent region in a can form a stem-loop structure, which

is known to facilitate replication slippage (Glickman and Ripley 1984).

Perhaps because long stretches of repeated sequences tend to be unstable
in the chloroplast genome (reviewed by Palmer, 1985), at least some of
the major repeat units are absent in other wild-type Oenothera plastomes
examined (Figure 3.5). In plastome 1C1, a recent derivative of 1°, most
of the repeated sequences including segments a, b and b* have been

deleted. The changes in 101 are most simply explained by the occurence

83

of two events, each of which resulted in a sizable deletion (Figure
3.4). The 179 bp deletion appears to have resulted from recombination or
slipped-strand mispairing between 46 bp of sequence identity: the region
of segment a and at before the 28 bp deletion. The 257 bp deletion
appears to have resulted from a similar event between 46 bp of sequence
identity: two iterations of the 17 bp motif preceded by two ACG

elements.

Can the observed chNA length variation between the two D-loop
initiation sites affect the efficiency of chNA replication? The
duplicated segment a in Oenothera plastome I° has the potential of
forming two stable stem-loop structures (Figure 3.3). These secondary
structures are very likely since the formation of D-loops would render
this region single stranded. It is known that replication forks tend to
pause at sequences that can form stem-loop structures (Huang and Hearst
1980, Weaver and DePamphilis 1982, Bedinger gt gt. 1989). One of the
deletions in ICl removed segment a and hence two potential stem-loop
structures that could contribute to pausing. In comparing the
transmission abilities of plastomes 1° and IC1, genetic data hinted that
these differences could have some influence (Appendix C). Crosses in
which 1° or I°1 competed against maternal plastomes type I to III did
not reveal a significant difference, however, crosses in which 1° or 101
competed against maternal plastome type IV, I°1 was detected in a higher
percentage of the progeny than was I° (65.8% biparental, n = 144 vs.
36.6% biparental, n = 314). It should also be noted that in the genomic
background of 91 hookeri str. Johansen, the relative transmission

efficiency of the three foreign plastome types (III > II > IV, Chiu gt

84

_t. 1988) is in reverse to the number of the repeat units present in
their 168 rRNAittgt spacer (III < II < IV, Figure 3.5). The effiency of
plastid transmission is influenced by many different factors (Chapter
1), however, and the development of an tg gtttg CpDNA replication system
is necessary to test the predicted pausing sites near the chNA

replication origin and their potential influence on the efficiency of

plastid multiplication.

CONCLUSION
The object of this dissertation was to understand the roles of the
nuclear and plastid genomes in controlling plastid transmission, using
Oenothera as the experimental system. Prior to this work, it was
generally held that the pattern of plastid inheritance in Oenothera is
determined primarily by the plastome, due to different intrinsic rates
of plastid multiplication associated with each plastome type (Schétz
1954, 1974, 1975; Chiu gt gt. 1988). Data presented in Chapter 1
modified this concept and stressed the important role of plastome-genome
interactions in determining the outcome of plastid transmission. In
general, there is a positive correlation between the relative success of
the plastid in transmission and the relative compatibility of the
plastome with the nuclear genome, suggesting that the interactions
between plastid and nuclear genomes can influence plastid transmission
by affecting the efficiency of plastid multiplication. These studies
also found low frequencies of plastid transmission from the paternal
side when the pollen had poor vigor due to an incompatible plastome-
genome combination, indicating that plastome-genome interactions may

affect the input of plastids at fertilization.

Although Schfitz (1954, 1974, 1975) concluded that the plastome had a

predominant effect on plastid transmission, he also noticed that the

competitive ability of a plastome was highest in its native nuclear

85

86

background. In this dissertation, the ability to compare results from
reciprocal crosses and to compare the four Oenothera plastomes
simultaneously in several different nuclear backgrounds made it possible
to obtain a more complete view of the role of both plastid and nuclear

genomes in plastid transmission.

Besides plastome-genome interactions, traits determined mainly by the
parental nuclear genomes, such as the size of the flower and the the
growth capacity of the pollen tube, were also observed to have an effect
on plastid transmission, independent of plastome type. For example, a
much higher frequency of paternal plastids was detected in the progeny
when strains with smaller flowers were used as the maternal parent.
Pollen from small-flowered strains were observed to have low growth
capacity when germinated tg gttgg (Chapter 1). These pollen also showed
a lower efficiency in transmitting plastids in crosses. These findings
suggest that if the time between pollination and fertilization is
lengthened, degeneration of plastids may occur in the pollen generative
cells. This hypothesis is supported by the fact that to some extent a
higher frequency of paternal plastids was detected in plantlets derived
from seeds collected from the upper half of the capsule than in seeds
from the lower half (Appendix B). A direct test for the connection
between the frequency of plastid transmission and the degeneration of
plastids would be to examine the number of chNA aggregates (Corriveau
and Coleman 1988) in the pollen generative cells at different times

after pollen germination in culture.

If generative cell plastids degenerate during pollen tube growth, then

87

Oenothera may differ only slightly from other plant species that exhibit
uniparental maternal plastid inheritance due to plastid degeneration in
the generative cell (Hagemann and Schréder 1989). Hence, the knowledge
we gain from the study of plastid transmission in Oenothera may be
pertinent to plastid transmission in a much larger group of land plants.
Theoretically, those plants that carry plastids in the pollen generative

cells should be able to transmit plastids from the male parent if
favorable conditions for the delivery and the multiplication of paternal
plastids are met. Supporting this view are the findings that even in
tobacco and Bpilobium, plants known to exhibit solely maternal plastid
inheritance, DNA from the paternal plastids can be detected occasionally

in the progeny (Medgyesy gt gt. 1986, Schmitz and Kowallik 1986).

As mentioned previously, the results of Schfitz (1974, 1975) that the
plastome determines the success of plastid transmission led to the the
belief that each Oenothera plastome type has its own intrinsic rate of
multiplication. This concept is in apparent contradiction to the finding
that plastids of barley and maize seem to be able to multiply in the
absence of plastid protein synthesis. These is also evidence that
plastids of tobacco multiply in the presence of an inhibitor of plastid
protein synthesis (reviewed by Boffey 1985). On the other hand, one
should recognize that while plastome-encoded proteins may not be
essential, they nontheless may regulate or increase the rate of plastid

multiplication through their interaction with nuclear gene products.

This dissertation sought to examine whether the plastome-dependent

88

efficiency of plastid multiplication could be modulated through a non-
coding function of the plastome. If one accepts Sch6tz’ hypothesis or
the modified view of plastid transmission suggested in Chapter 1, then
the origin of CpDNA replication is an outstanding candidate for the
plastid component that modulates the efficiency of plastid multipli-
cation. Since initiation of DNA replication is the first step in plastid
multiplication, differences in the replication origins among the
plastome types may influence the efficiency of chNA replication and
plastid multiplication. Chapter 2 mapped the CpDNA replication origins

from Oenothera as a first step towards examining this possibility.

According to electron microscopic data, initiation of chNA replication
in Oenothera proceeds through the formation of D-loops at four separate
sites. These initiation sites are located within the large inverted
repeat region, on both sides of the 168 rRNA gene. An attempt to use
two-dimensional gel electrophoresis to more precisely localize the D-
loop initiation sites was not successful (Appendix D). According to the
available data, there is no difference in the number and the positions
of chNA replications between plastome types I and IV, the two plastomes
with the most distinct transmission properties. Hence, if the origins
are the sites in Oenothera plastomes that vary with respect to DNA

replication, they must differ from each other in a more subtle way.

In mapping the chNA replication origins, a hypervariable region that
differentiated the Oenothera plastomes was noted between the 168 rRNA
and the trnI genes. This hypervariable region was investigated further

(Chapter 3) because of its proximity to two of the four D-loop

89

initiation sites and, hence, the possibility that differences in this
region could account partially for differences in plastid transmission.
Sequencing of this hypervariable region revealed multiple direct repeats
that most likely arose through duplications of part of tggt and the
sequences immediately upstream from ttgt. Some of the repeated regions
contain segments of DNA that are capable of forming stable stem—loop
structures. These secondary structures could potentially block
replication fork movement and hence affect the efficiency of chNA
replication and plastid multiplication. In order to test this
hypothesis, the transmission efficiencies of a wild-type plastome 1 (1°)
and plastome IC1 (Appendix A), a derivative of 1° with a deletion of
most of the direct repeats were compared (Appendix C). In the process of
gathering these data, my other crossing experiments (Chapter 1)
indicated that significant variation in plastid transmission can occur
from year to year and from flower to flower, at different times in the
flowering season. Thus, it is not valid to compare most of the data from
the I° and ICl crosses of Appendix C. Subtle differences may exist
between the two plastomes, but side by side comparisons are necessary to
avoid variations in other factors that may also affect the apparent

efficiency of plastid transmission.

This dissertation began with a simple hypothesis to explain the
molecular mechanisms behind the differential plastid transmission in
Oenothera, but ended by confronting the complex nature that governs
plastid inheritance in higher plants. Basically, the genetic results
suggest that the outcome of plastid transmission depends not only on the

number of plastids in the gametes but also on events that take place

90

before and after fertilization. Furthermore, all of these events rely
heavily on the interactions between the plastid and the nuclear genomes.
Due to the large size of the nuclear genome and the number of genes
involved, it will be easier to investigate the genetic basis for these
crucial interactions from the side of the plastome. Thanks to the recent
development of plastid transformation techniques (Svab gt gt. 1991), it
is now possible to replace genes within the plastid. Experiments using
chNA from one Oenothera plastome type to replace the homologous region
of a different plastome type may be used to localize plastid genes that
are responsible for plastome-genome compatibility. Oenothera, a
classical favorite of cytogeneticists, may play an important role in
helping us to understand the complicated nature of plastome-genome

interactions.

APPENDICES

APPENDIX A. OENOTHERA CHLOROPLAST DNA POLMRPHISMS ASSOCIATED

WITH PLASTCME MUTATOR ACTIVITY

 

Frm Mol. Gen. Genet. (1990) 221:59-64

9]

Mol Gen Genet (I990) 22l : 59—64

NIGE—

C SpringeroVerlag I990

 

Oenothera chloroplast DNA polymorphisms associated

with plastome mutator activity

Was-Ling Chiu, Ellen M. Johnson. Sara A. Kaplam. Kimberly Blacko, Mary B. Sokalsld, Ruth Wolfaon.

asdBarbanRSean

Department of Botany and Plant Pathology. Michigan State University. East Lansing. MI 48824. USA

Sun. Oenothera plants homozygous for a recessive al-
lele at the plastome mutator (pm) locus show non-Mende-
lian mutation frequencies that are lOOO-fold higher than
spontaneous levels. Chloroplast DNA (CpDNA) was iso-
lated from nine mutants and two green isolates of the plas-
tome mutator line. CpDNA restriction patterns were com-
pared to CpDNA from a representative of the progenitor
Johansen strain. and chNAs from all eleven plastome mu-
tator lines show changes of fragment mobility due to dele-
tion events at five discrete regions of the plastome. Most
of the mutants have CpDNA restriction patterns identical
to that of one of the green isolates from the plastome muta-
tor line. and therefore. most of the differences in fragment
length are probably not responsible for the mutant pheno-
types. In contrast to the plastome mutator line. chNA
from several populations of a closely related wild-type Oen-
othera species have few restriction fragment length poly-
morphisms. This suggests that both mutation frequencies
and site-specific CpDNA deletions are elevated in the plas-
tome mutator line. and implicates a defect in the CpDNA
repair or replication machinery.

Key words: Plastome mutator - Oenothera — Chloroplast
DNA polymorphisms

 

Introduction

Chloroplast DNA (CpDNA) coding sequences are highly
conserved (Curtis and Clegg I984: Zurawski and Clegg
1987). which may be due in part to the low frequency of
spontaneous mutation (Michaclis I969; Kutzelnigg and
Stubbe I974). However. comparisons ofthe CpDNA restric-
tion patterns and physical maps from closely related plant
species have revealed the relatively frequent occurrence of
insertion/deletion events (Gordon et al. I982; Salts et al.
I984; Schmitz and Kowallik I986; Doebley et al. I987) or
rearrangements (reviewed by Palmer I985).

A number of higher plant isolates have elevated rates
of plastid genome (plastome) mutations. including Arabi-
dopsis (ha/iana (Redei and Plurad I973). maize (Shumway
and Weier I967: Walbot and Coe I979: Thompson et al.
1983). Petunia (Potrykus I970). and Oenothera (Epp I973).
The plastome mutator activity in the Johansen strain of
0. hookeri is the result of a nuclear gene that. when homo-
zygous for the mutant pm allele (pm pm). induces mutations

Olfprinr requests to; 8.8. Scars

in the plastome (Epp I973). A large collection of plastome
mutator-induced chloroplast mutants has been generated
by several laboratories (Epp I973; Sears 1983; Epp et al.
1987). All mutations isolated thus far were recognized ini-
tially as chlorotic sectors that appeared in pm/prn plants.
In confirmation of the initial findings of Epp (I973). such
plastome mutations are inherited in a non-Mendelian fash-
ion and display an array of phenotypes; independent mu-
tants show varying degrees of chlorosis (Epp I973) and
differ in the extent of plastid development (Epp and Panha-
sarthy I987) and in the accumulation of particular proteins
(Johnson I988). Epp et al. (1987) have recently shown that
the chNAs isolated from different plastome mutator-in-
duced chlorotic sectors are identical in restriction pattern
to CpDNA of the wild-type. when cut with enzymes that
digest CpDNA infrequently. In contrast. our experiments
utilized more frequently cutting enzymes applied to a differ-
ent group of plastome mutator-induced mutants and two
green isolates. and we have found restriction fragment
length polymorphisms (RFLPs). Comparisons of the
CpDNA restriction patterns from isolates of the plastome
mutator (pm) line and from the wild-type 0. hookeri str.
Johansen suggest that the CpDNA alterations in the pm
line result from deletions at several sites that were pre-
viously identified as " hot spots" in the evolution ofchNA
(Gordon et al. 1982; Stein and Hachtel 1986).

Materials and methods

Plum material. Seeds from 0. alum were generously pro-
vided by Prof. W. Stubbe and Dr. W. Dietrich from plants
collected from the Mexican states of Puebla. Mexico (near
the city of Toluca). and Hidalgo (Steiner 1951). catalogued
in the University of Diisseldorf collection as Puebla. To-
luca. and Hidalgo. respectively. The Johansen strain of 0.
hookeri was established from a single rosette collected in
I927 by DA. Johansen (Cleland I935). From I930 it was
maintained in R. Cleland‘s collection by self-pollinations.
and in the l950$ seeds were provided to H. Stinson (Cornell
University. Ithaca. NY). E. Steiner (University of Michi-
gan. Ann Arbor). and W. Stubbe (University of Diisseldorf.
FRG). where the strain was perpetuated by further self-
pollinations (Fig. I). We refer to the plant lines established
at these institutes as the Cornell. Ann Arbor, and Diisseldorf
lines of strain Johansen. The only representatives of the
Cornell line that have been preserved are descended from
an isolate (E-lS-7) derived from ethyl methane sulfonate
(EMS) mutagenesis. and exhibiting the original plastome
mutator activity (Epp I973).

92

Oenothera hooteri sir Jonansen
Indiana University
(RE Cleland)

/Ousseldo\

. _ (W Stubbe) . .
Come" University U. Michigan. Ann Arbor
(H. Stinson) (E. Steiner)

\

EMS malaqenesis
(M. Epp)

E '5 -7 plastome mutator
I
®l®

197635.": (975 field

I980 field 1980 field

i gméingi m5" _pmiil.pmi;; %

(962 field

Fig. I. Pedigree of mutant and wild-type lines descended from the
original Johansen isolate of Oenothera hookeri. An unknown
number of self-pollinations were performed in the propagation of
the Johansen strain at all four institutions. where the stocks were
maintained. Ethyl methane sulfonate (EMS) mutagenesis was used
to produce the plastome mutator isolate (E-I5-7) at Cornell (Epp
I973). Several self-pollinations (indicated by an encircled x) were
performed with progeny of this line to generate more seed prior
to the I976 and I978 field plantings indicated in the pedigree.
In the plastome mutator line. mutants are indicated by pm designa-
tions. while green isolates are indicated as C . and C 3. Different
lines obtained from the same plant have adjoining boxes

 

Both 0. elata and 0. hookeri have an MA nuclear back-
ground and a plastome type 1. according to the classifica-
tion scheme of Stubbe ( I959). and according to Raven et al.
(I979) they should be reclassified as a single species. The
original pm/pm line was inbred for several generations to
produce the plastome mutator seed stocks. which were
maintained by Prof. W. Stubbe after I973. Figurel indi-
cates the pedigree for the plastome mutator-induced mu-
tants (pm7. pm7ss. pm8, pm9. pmll. ptnIZ, pm32. pm35.
pm". and pm43) that were isolated from progeny of further
self-pollinations. All of these mutants have pigment alter-
ations; their phenotypes range from white to relatively
green. but none are photosynthetically competent. To pro-
duce the plant lines used here. flowers from these plants
were emasculated and crossed as the female parent to a
pm‘fprn' line as described by Kutzelnigg and Stubbe
(I974). These crosses accomplished two objectives: (i) seeds
were produced. which could be stored. or used immediately
for the establishment of leaf-tip cultures and (ii) the plastid
mutations were stabilized by placement in a heterozygous
nuclear background.

In order to produce a pm’ ‘pm line containing the same
plastome type in a heterozygous nuclear background for
use as a control. green siblings of plants containing the
pm32. pm35. pm”. and pm43 mutations in the 1982 field

plot were crossed as the female parent with pollen from
0. hookeri strain Johansen carrying plastome type IV. be-
cause plastome IV transmits its plastids to fewer than I°/o
of the progeny (Chiu et al. I988). We refer henceforth to
this green line of 0. hookeri strain Johansen as “C ome/l»!
To obtain leaf tissue for plastid isolations. Cornell-l plants
were grown in a glasshouse in sandy soil.

Another green isolate (Cornell-2) carrying a second
CpDNA variant was recovered from the I980 field season
by crossing the same plant that carried the pm9 mutation
to a plant line with the albicans/percurvans genotype (Cle-
land I972) and plastome type IV; the resulting progeny
have an AC nuclear background. Plants having plastome
typel in combination with the A/C nuclear background
are pale green due to nuclear-plastome incompatibility
(Kutzelnigg and Stubbe 1974) and therefore. it was neces-
sary to grow them as leaf-tip cultures as described by John-
son and Sears (I990). To establish pure mutant cultures.
seeds were surface-sterilized with 40% commercial bleach
and germinated on agar medium lacking plant hormones
(Stubbe and Herrmann I982). The resulting seedlings were
propagated as leaf-tip cultures to allow the plastids to sort
out during vegetative growth. Cultures of pure mutant tis-
sue were maintained by transferring them to fresh medium
every 4 weeks.

Isolation of chloroplast DNA. Chloroplast DNA was pre-
pared as described by Chiu and Sears (I985). Bisbenzimide
(Hoechst 33258. Sigma) was used for the CsCl gradients
at a final concentration of 0.0I mg/ml (after Hudspeth et al.
I980). After buoyant density equilibrium centrifugation. the
upper (chNA) band in the gradient was collected. If the
lower band of nuclear DNA was so broad that it contamin-
ated the upper chNA band during its removal from the
gradient. the DNA band was further purified on a second
gradient under the same conditions. Bisbenzimide was re-
moved by three extractions with NaCl-saturated isopro-
panol. the remaining CpDNA-CsCl mixture was diluted
twofold with water. and precipitated by the addition of
two volumes of ethanol or one volume of isopropanol. The
resulting chNA pellet was washed with 70% ethanol and
dissolved in IO mM TRIS. 0.1 mM EDTA. pH 8.0.

Restriction analysis and Southern blotting. Restriction endo-
nuclease digestions were performed according to the manu-
facturers' specifications (Bethesda Research Laboratories
and Boehringer Mannheim Biochemicals). Fragments were
separated by electrophoresis in 0.8% agarose gels in TAE
buffer (40 mM TRIS. 20 mM sodium acetate. I mM
EDTA. pH 8.0) and visualized with ethidium bromide
staining according to Maniatis et al. (I982). using Hindlll-
digested lambda phage DNA as size markers. Standard pro-
cedures were used for Southern blotting and nick-transla-
tion (Maniatis et al. I982). using 32P-labeled dATP from
New England Nuclear. Dupont. All probes were made from
plasmids having cloned chNA from the Dusseldorf line.

Results
Comparisons ofchNA from plastome mutator lines
and the wild-type O. hookeri str. Johansen

CpDNA was isolated from nine mutants and two green
isolates (C ornell-l and C ornelI-Z) from the plastome muta-

D 7 7.8 C: II l2Ci 32 35 4l437\

 

0 < 9.4 lib
6.7

b < 4.3

C 2.3

d E 2.0

Fig. 2. Clal restriction patterns oI‘chNA from plastome mutants
and green isolates. CpDNA from three green isolates. the Dussel-
tlorf (D). Cornell-l (Cl). and Cornell-2 (C1) lines are compared
to CpDNA from plastome mutator-induced mutants prn7. pm7ss.
pms, pmll. pmll'. pmJZ. pm35. pm4l. and mud. Lanes with
chNAs from the mutants are given the numerical pm-destgnation.
DNA size standards at the right were obtained by Hindlll digestion
of lambda DNA. The letters a—e to the left of the gel indicate
the positions of fragments that vary in mobility

tor line and from the Dusseldorf and Ann Arbor lines of
strain Johansen. No differences were observed in the com-
parison of restriction patterns of the Dusseldorf and Ann

Arbor lines (data not shown). and therefore the Dt‘isseldorf

line has been used to represent the progenitor CpDNA. In
the comparisons of the plastome mutator lines. restriction
fragment length polymorphisms (RFLPs) were observed
when the DNAs were digested with EeoRl. BamHI. and
Clal. Since the Clal digestions revealed the most vanability
(5 RFLPs). this enzyme was used for the subsequent analy-
ses. In Fig.2. five variable Clal fragments are apparent
in the ethidium bromideostained agarose gel. although the
size differences of the larger fragments are more difficult
to discern in this 0.8% agarose gel than in a gel with a
lower percentage of agarose. Comparing the chNAs from
the green isolates. the C ornelI-J line differs from both the
C ornell-l and Dtisseltlorl lines in all five variable fragments.
Four of these five fragments also differ in mobility between
the Cornell-l and Dtisseldnrl'lines. In all cases. the mobility
differences represent variations in fragment size of
50—500 bp.

The chNAs of pm35, pmll. and pm43 have the same
restriction pattern as line Cornell-I. as does pm32. although

93
til

the digest shown in Fig. 2 contains some partial digestion
products for this line. The chNAs of pm7, pink; and
ple are identical to the CpDNA of the Cornell-2 line.
Pm? and pm7ss are two different mutants with quite differ-
ent pigmentation phenotypes that were isolated from a sin-
gle plant. The restriction pattern of mutant pm8 is the same
as that of the Cornell-2 line. except for variable fragment
“".c which is about 300 bp larger than the equivalent frag-
ment in Cornell-2 and has the same mobility as that of
the wild-type Dusseldorf line. Similarly. the restriction pat-
terns of the pin]! mutant resembled that of the Cornell-l
class except that the variable fragment “b" was approxi—
mately 300 bp smaller than the analogous fragment in C or-
nell-l. Among all five variable fragments detected. frag-
ments from the Dfisseldarf line are always the largest.

Localization of the five variable regions
on the physical map of the Oenothera plastome

The five variable Clal fragments were located on the physi-
cal map of Oenothera chNA (Gordon etal. I982) by
Southern hybridimtion (Figs. 3. 4; additional data not
shown). Fragment “a" hybridized to a cloned 5.1 kb EcoRI
fragment from the wild-type Dfisseldorf line. which contains
the entire spacer region between the 165 and 23S rRNA
genes and most of the 23S rRNA (Gordon etal. I982).
This segment of the inverted repeat (see Fig. 3) is found
on the largest and third largest Sall fragments of Oenothera
CpDNA (Sal fragments I and 3 in Fig. 3). The variable re-
gion within fragment a was further determined to be in
the spacer near the I65 rRNA (data not shown). The corre-
sponding fragments in lines Cornell-l and Cornell-2 are
500 bp and 300 bp smaller. respectively. than that of the
Dusseldorf line.

  

Eco - 51kt)
Fig. 3. Phystcal map ofO. hookeri plastome l. showing the location
of the five variable Clal fragments (letters a-e). This map contains
the Pill. It’prtl and Still sites and gene IOCI determined prevrously
(Gordon et al. 1982: Hermann et al. 1983). and shows the position
and sizes offragments produced by BamHI. Clal. and EcoRI diges-
tion. which were used as probes for southern hybridization. The
relative sizes of the major Sit/I fragments are indicated. with I
being the largest. The 2.3 kh Clal fragment is equivalent to frag-
ment c. but its prectsc location within the 5an fragment has not
been determined. In the innermost circle. the inverted repeat rc-
gtons are indicated as Huck um, and the posttion of a large inver-
sion in the phystcal map. relative to spinach. ts indicated by the
wippled urt‘

62

  

g<

A B C

 

94

CiCaDPHT CtCzDPl-l T CICzDPHT CtCzDPHT CICzDPHT

 

D

Flg. 4A—E. Comparison of RFLPs within the Johansen strain of 0. hookeri. and among diITerent populations of 0. elata. CpDNA
from the C ornell-I (C .). C ornell—J (C 1). and Dilute/dorjiD) Wild-type. green lines of strain Johansen and from the Puebla (P). Hidalgo

(H). and Toluca (T) collections of 0. elata were digested with Clal and separated electrophoreticelly. A Stained gel. B—E

Southern

blot sequentially probed with DNA from fragments indicated in Fig. 3. B cloned 5.1 kb EeoRI fragment: C cloned 8 kb BamHI fragment;
D gel-purified 2.5 kb fragment r from a clone of Sale; E cloned 1.6 kb Clal‘BamHl subfragment of the 3 ltb BamHI fragment (Blasko
et al. I988). Because one blot was hybridized several times. carry-over of label can be seen in D. and slightly in E. Some partial
digestion products are apparent in the hybridizations to C I and C 2 chNA

The variable b fragments hybridized to a cloned 8 kb
BamHI fragment (Bum 3b) that overlaps part of fragment
Sit/5 and part of San in the large single copy region. Ac-
cording to our fine mapping. the variable Clal fragment
b is contained completely within the 5:15 fragment. Vati-
able fragments c and d hybridized to 2.5 kb and 2.3 kb
Clal subfragments. from within the Salt: fragment. which
spans the border region between one of the inverted repeats
and the large single copy region. Within this region. frag-
ments 0 and d are in the single copy region very close to
the border of one of the inverted repeats. Their relative
positions within this region have not been precisely deter-
mined. Fragment e hybridizes to the same 13 kb Solo frag-
ment as well as to the Sci/2 fragment. indicating that it
lies within the inverted repeat of the CpDNA. The variable
region e has been further defined by hybridization to a
3kb BamHI fragment described by Blasko et al. (I988).
and mapped to the edge of the inverted repeats.

Comparison of RF LP: ofO. hookeri str. Johansen
and O. elata

We hoped to determine whether the RFLPs observed in
the plastome mutator line represent a normal level of vari-
ability. or indicate that rapid changes were occurring in
the chNA of this line. Thus. we extended our comparisons
to the closely related 0. elata. CpDNAs were isolated from
plants representing three independent collections of 0. elata
from Toluca. Hidalgo. and Puebla. Mexico (Steiner I9SI).

Probes from four variable regions in str. Johansen were
hybridized sequentially to Southern blots of chNA from
the three 0. elata strains. As shown in Fig. 4. CpDNA re-
striction fragments from the three 0. elata strains were
much less vanable than those from the plastome mutator
line. even though they represent plants that are less related
to each other than are the isolates of the Johansen strain
we have investigated here. Furthermore. for every variable
region. the fragments in the Dii.s.\~eldor/'line were more simi-
lar in mobility to those of the elata strains than to those
of the Cornell isolates of the plastome mutator line.

Discussion

A great number of plastome mutator-induced plastid muta-
tions have been collected in the past IO years (Epp et al.
I987). and yet the nature of the pm gene function is still
unknown. The goal of the current study was to investigate
the nature of plastome mutator-induced mutations in order
to elucidate the function of the pm gene.

We have compared RFLPs of chNA from descendants
(two green and nine mutant isolates) of the original pm pm
plant and from two representatives of wild~type 0. hookeri
strain Johansen. from which the plastome mutator line was
derived. Five Clul fragments of the chNA were variable
among these isolates. Among the isolates from the plastome
mutator line. the RFLPs can be classified into two major
types. represented by lines Cornell-l and Cornell-2 and by

95

 

 

63
Table I. Sins of variable fragments a—e in Oenothera hookeri strain Johansen
Plant line' D C . C 3 pm7 pm7ss pmél pm] I pml.’ pm32 pm35 pm” pm“
Fragment”
a 9.0 8.5 8.7 8.7 8.7 8.7 8.5 8.7 8.5 8.5 8.5 8.5
b 4.6 4.5 4.3 4.3 4.3 4.3 4.2 4.3 4.5 4.5 4.5 4.5
c 2.5 2.35 2.2 2.2 2.2 2 5 2.35 2.2 2.35 2.35 2.35 2.35
d 2.25 2.25 2.15 2.15 2.15 2.15 2.25 2.15 2.25 2.25 2 25 2.25
e 1.55 1.50 1.40 1.40 1.40 1.40 1.50 1.40 1.50 1.50 1.50 1.50

 

‘ D. Disseldorf.‘ C I. C omell-l .- C 2. C ornell-J
" Fragment sizes are given in kb

two minor classes. consisting of the mutants pm? and pm! I.
each of which have chNAs differing from that of the
Cornell-2 or Cornell-l line. respectively. by only one frag-
ment. The comparisons with two wild-type lines descended
from the original Johansen isolate and three collections of
0. elata from the wild (Fig. 4) indicate that the CpDNA
variability correlates with the plastome mutator ancestry
of the Cornell strains. Since the isolates from the pm line
invariably have fragments that are smaller than the analo-
gous fragments in the Dusseldorf and Ann Arbor lines. we
conclude that CpDNA deletions have occurred recently
within the plastome mutator line.

The five variable regions have been mapped on the plas-
tome (Fig. 3): two of them (a. e) are inside the inverted
repeat. while the other three (b. c. d) are in the large single
copy region. As indicated in Table 1. the sizes of fragments
a—e are highly variable. with more than two size classes
being observed for each fragment. It is probably not coinci-
dental that all five variable regions overlap with the evolu-
tionary “hot spots" previously identified on the Oenothera
plastome (Gordon et al. 1982; Stein and Hachtel 1986).

Length mutations smaller than 50 bp would not have
been detected in this study due to the limitation of resolu-
tion imposed by agarose gel electrophoresis. and thus we
cannot address the frequency of occurrence of smaller
length alterations. Among the nine mutants analyzed. only
one contained chNA in which a specific deletion (frag-
ment b of pmll) clearly distinguished the mobility of a
fragment from the homologous fragment in any of the green
plant lines. Thus. with the exception of this one fragment.
we conclude that the deletions that result in the RF LPs
are not likely to be responsible for the mutant phenotypes.
Presumably. undetected small insertions. deletions. or base
substitutions are responsible for the chlorotic phenotypes
of the other mutants and. possibly. also of pmll.

A number of mechanisms have been proposed to explain
the elevated mutation rates in pm plants (Redei and Plurad
1973; Epp 1973; Sears 1983). These include defects in the
replication or repair machinery and the suggestion that pm
encodes a repressor of a chloroplast transposable element.
The evidence presented here argues against the involvement
of a transposable element because changes in fragment size
seem to be unidirectional. favoring deletions: also Southern
hybridizations show that no sequences are shared among
the variable fragments and hence no common element could
be involved in the deletions. Furthermore. sequencing of
variable region e (Blasko et a1. 1988) showed that it has
none of the features of known transposable elements (re-
viewed by Charlebois and Doolittle 1989). Rather. the
RFLPs are due to the presence of a variable number of

24 bp direct repeats. Other short repeats in Oenothera
CpDNA have been identified and are probably involved
in deletion/insertion events on an evolutionary scale (Stein
and Hachtel 1988 a. b). We predict that similar short direct
or inverted repeats will be implicated in the other four
chNA hypervariable regions in the plastome mutator line.
Such repeats could be the substrate for plastome mutator
activity. if the pm defect involves a recombination/repair
system or the replication apparatus. Thus. a single mecha-
nism is hypothesized to be responsible for the high fre-
quency of both the chlorotic mutations and the CpDNA
deletions. Determination of chNA sequence changes re-
sponsible for the mutant phenotypes may more clearly im-
plicate one of the above mechanisms.

Acknowledgements. We thank W. Stubbe. E. Steiner. and W. Diet-
rich for providing us with the Oenothera plant lines. and we wish
to acknowledge the encouragement and support of these col-
leagues. as well as that ot' Drs. Reinhold G. Herrmann. Peter West-
hofT. Mel Epp. and Barry Chelm during the course of this work.
We thank Dr. Helmut Bertrand for critical comments on the manu-
script. EM] and WLC were supported by graduate fellowships
from the MSU McKnight photosynthesis training grant. SAK. KB.
MBS. EM]. and RW were supported by NSF Grant DCB-
85002849. This research was also supported by the Michigan State
Agricultural Experiment Station.

References

Blasko K. Kaplan SA. Higgins KG. Wolfson R. Sears BB (1988)
Variation in copy number of a 24-base pair tandem repeat in
the chloroplast DNA of Oenothera hookeri” strain Johansen.
Curr Genet 14:287-292

Charlebois RL. Doolittle WF (1989) Transposable elements and
genome structure in halobacteria. In: Berg DE. Howe MM
(eds) Mobile DNA. Am Soc Microbiol. pp 297-307

Chiu W—L. Sears 88 (1985) Recombination between chloroplast
DNAs does not occur in sexual crosses of Oenothera. Mol Gen
Genet 198:525-528

Chiu W-L. Stubbe W. Sears 88 (1988) Plastid inheritance in Own-
orhera: organelle genome modifies the extent of biparental plas-
tid transmission. Curr Genet 13: 181-189

Cleland RE ( 1935) C yto-taxonomic studies on certain Oenolht’ras
from California. Proc Am Phil Soc 752339—429

Cleland RE (1972) Oenothera. Academic Press. NY. p 370

Curtis SE. Clegg MT (1984) Molecular evolution of chloroplast
DNA sequences. Mol Biol Evol 12291-301

Docblcy JF. Ma DP. Renfroe WT (1987) Insertion deletion muta-
tions in the Zn: chloroplast genome. Curr Genet 11 :617—624

Epp MD ( 1973) Nuclear gene-induced plastome mutations in Oen-
othera hookeri. 1. Genetic analysis. Genetics 751465—483

Epp MD. Parthasarth) MV (1987) Nuclear gene-plastome muta-
tions in Oenothera Inmkeri. ll. Phenotype description with elec-
tron microscopy. Am J Bot 74:143-151

96

64

Epp MD. Erion JL. Sears 88. Stubbe W (1987) The plastome
mutator of Oenothera continues to function as originally de-
scribed. Mol Gen Genet 206:515-518

Gordon KHJ. Crouse EJ. Bohnert HJ. Herrmann RG (1982) Phys-
ical mapping of differences in chloroplast DNA of the five
wild-type plastomes in Oenothera subsection Euoenmhera.
Theor Appl Genet 61 :28l-296

Herrmann RG. Westoff P. Alt J. Winter P. Tittgen J. Bisanz C.
Sears BB. Nelson N. Hurt E. Hauska G. Viebrock A. Sebald
W (1983) Identification and characterization of genes for poly-
peptides of the thylakond membrane. In: Cifferi O. Dure L
III (eds) Structure and function of plant genomes. Plenum. New
York. pp 143—153

Hudspeth MES. Shumard DS. Tatti KH. Grossman Ll (1980) Ra—
pid purification of yeast mitochondrial DNA in high yield. Bio.
chim Biophys Acta 6101221-228

Johnson EM ( 1988) Physiological and molecular characterizations
of plastids from a plastome mutator-induced mutant of Oenoth-
era hookeri strain Johansen. Ph. D. Thesis. Michigan State Uni-
versity

Johnson EM. Sears BB (I990) Membrane composition and physio-
logical activity of plastids from an Oenothera plastome muta-
tor-induced chloroplast mutant. Plant Physiol 91. in press

Kutzelnigg H. Stubbe W (1974) Investigations on plastome mu-
tants in Oenothera. I. General considerations. Sub-Cell Bio-
chem 3:73—89

Maniatis T. Fritsch EF. Sambrook J (1982) Molecular cloning.
A laboratory manual. Cold Spring Harbor Laboratory. Cold
Spring Harbor. New York. p 545

Michaelis P (1969) Uber Plastiden-Restitutionen (Ruckmuta-
tionen). Cytologia 34 (Suppl): 1—115

Palmer JD (1985) Evolution of chloroplast and mitochondrial
DNA in plants and algae. In: Maclntyre RJ (ed) Molecular
evolutionary genetics. Plenum. New York. pp 131—239

Potrykus I (1970) Mutation und Ruckmutation extrachromoso-
male vererbter Plastidenmerkmale von Petunia. Z Pflanzenzuch
63:24—40

Raven PH. Dietrich W. Stubbe W (1979) An outline of the system-
atics of Oenothera subsect. Euoenolhera (Onagraceae). Syst Bot
42242—252

Redei GP. Plurad SB (1973) Hereditary structural alterations of
plastids induced by a nuclear gene in Arabidopsis. Protoplasma
77:361—380

Salts Y. Herrmann RG. Peleg N. Lavi U. lzhar S. Frankel R.
Beckmann JS (I984) Phys‘ical mapping of plastid DNA varia-
tion among eleven .vV'it-un'ana species. Theor Appl Genet
69: 1-14

Schmitz UK. Kowallik KV (1986) Polymorphism and gene ar-
rangement among plastomes of ten Epilohium specxes. Plant
Mol Biol 7:115—127

Sears 88 (1983) Genetics and evolution of the chloroplast. In:
Gustafsson P (ed) Stadler Symposium XV. UniverSIty of M15-
soun Press. Columbia. Mo. pp 119—139

Shumway LK. Weier TE (1967) The chloroplast structure of t’o/up
maize. Am J Bot 54:773—780

Stein J vom. Hachtel W (1986) Chloroplast DNA differences be-
tween two species of Oenothera subsection Mun:ia' location
in the chloroplast genome and relevance to possible interactions
between nuclear and plastid genomes. Theor Appl Genet
73: 141—147

Stein J vom. Hachtel W (l988a) Chloroplast DNA dilTerences in
the genus Oenothera subsection Manna: a short direct repeat
resembling the lambda chromosomal attachment site occurs
as a deletion/insertion within an intron of an NADH-dehy-
drogenase gene. Curr Genet 13: 191-197

Stein J vom. Hachtel W (l988b) Deletions/insertions. short in-
verted repeats. sequences resembling art-lambda. and frame
shift mutated open reading frames are involved in chloroplast
DNA differences in the genus Oenothera subsection Munzi'a.
Mol Gen Genet 213: 513—518

Steiner EE (1951) Phylogenetic relationships of certain races of
Euoenorhera from Mexico and Guatemala. Evolution
5:265-272

Stubbe W (1959) Genetische Analyse des Zusammenwirken von
Genom und Plastom bei Oenothera. Z Bot 48: 191-218

Stubbe W. Herrmann RG (1982) Selection and maintenance of
plastome mutants and interspecific genome/plastome hybrids
from Oenothera. In: Edelman M. Hallick RB. Chua N-I-I (eds)
Methods in chloroplast molecular biology. Elsevier. Amster-
dam. pp 149—165

Thompson D. Walbot V. Coe EH Jr (1983) Plastid development
in iojap and mutator-affected maize plants. Am J Bot
702940—950

Walbot V. C oe EH Jr (1979) Nuclear gene iojap conditions a pro-
grammed change to ribosome-less plastids in Zea mays. Proc
Natl Acad Sci USA 76:2760—2764

Zurawski G. Clegg MT (1987) Evolution of higher plant chloro-
plast DNA-encoded genes: implications for structure-function
and phylogenetic studies. Annu Rev Plant Physiol 38:391-418

Communicated by GR. Fink

Received September I. I989

APPENDIX B. RELATIONSHIP BETWEEN SEED POSITION AND THE EXTENT

OF BIPARENTAL PLASTID TRANSMISSION

Rationale

For all Oenothera strains examined by DAPI-staining, over 908 of the
pollen generative cells contain one or more plastid DNA aggregates
(Corriveau and Coleman 1990; personal communication). One might expect
that these pollen should be able to transmit their plastids to the
progeny. However, results in Chapter 1 indicate that paternal plastids
are not generally transmitted with such a high efficiency. It was noted
that when the parental plastome types are held constant, biprental
transmission of plastids occurs more frequently when the maternal parent
is a small-flowered strain or when the paternal parent produces pollen
with higher growth capacity. These observations indicate that the timing
between pollination and fertilization may be an important factor in
Oenothera plastid transmission. Conceivably, plastids in the pollen
generative cell degenerate as the time increases between pollination and

fertilization.

The purpose of this investigation was to test the hypothesis that there
is loss of plastids in Oenothera during growth of the pollen tube by
examining the relationship between the outcome of plastid transmission

and the position of seeds in the capsule. This experimental approach is

‘97

98

based on the assumption that pollen tubes have to grow longer in order
to fertilize ovules in the lower half of the ovary. If there is
degeneration of plastids during pollen tube growth, pollen reaching the
lower part of the capsule may carry fewer viable plastids, resulting in
a lower frequency of plastid transmission from the paternal parent in
that region. To examine this possibility, for a number of crosses, the
frequency of biparental plastid transmission in seeds obtained from the
upper half of the capsule was compared with that in seeds from the lower
half.

Results and Conclusion

All the Oenothera strains, except fer AJAJ-IC1, used in the crosses
described below are listed in Ch. 1, Table l. MAJ-ICl is a strain of Q;
hookeri str. Johansen carrying plastome 191, which is a derivative of
plastome 1° (Chiu gt_gl. 1990). Mutant plastids used in the crosses were
maintained in 0. hookeri str. Johansen (genotype AJAA), and all the
crosses were made in the 1989 field season. Dried, mature seed capsules
were broken in half and seeds from the upper and the lower halves were

collected separately.

Results from ”Green x White” and ”White x Green" crosses are presented
in Table 8.1, section A and section B, respectively. When strains AFCp
-II, BPCf-IV and AFCP-IV were used as paternal parents, seed development

was limited to the top portion of the capsule, most likely due to

99

Table 8.1. Frequency of biparental plastid transmission (BP) obtained

from seeds in upper (U) or lower (L) half of the capsule.

A. Green x white crosses

 

 

 

 

 

 

 

Esteraal
APCP-IV APCP-II chf-IV A'CP-IV AJAJ-ID
§_§P (g) x Q? (n) x_§P (n) ngf (g) x_§r (9) tb
Paternal
I-b U - 45.3 (53) 73.8(160) 59.4(133) 19.0(142) 0.48
L -- 55.6 (54) 61.7(170) 64.8 (88) 20.3(133)
I-z U 86.9 (61) 69.5 (82) 86.8(159) 85.0(127) 46.4(112) 0.57
L 78.8 (66) 80.5 (82) 81.1 (37) 94.4(108) 46.3(123)
II-e U 72.4 (87) - 78.6(145) 80.2 (91) 2.9(104) 2.61"l
L 66.0 (97) -- 72.1(140) 72.9 (96) 0 (131)
III-g U 80.8 (73) - 90.1(151) 83.3(126) 19.6(107) 2.10‘
L 88.3(128) -- 91.6(142) 89.1 (83) 24.7 (93)
IV-a U 47.6 (63) 4.9(102) - 38.7(111) - 2.70"”I
L 29.8 (57) 7.5 (93) - .gg.4 (85) '-
B. White x Green crosses
Maternal
I-b I-z II-e. III-g IV-a
x 5? (9) 3,9? In), 3.9? (n), x g? (g), x g? (n) t
Paternal
AJAJIC1 U 25.8(128) 22.4 (98) 50.0(104) 37.8(172) 74.7(150) 2.14*
L 17.6(108) 23.1(117) 39.4(127) 21.7 (23) 55.0(229)
A'CPIV U 21.7 (46) 9.1 (33) 4.4 (45) 12.7 (63) 29.9(104) 0.91

L 21.9 (32) 11.2(143) 3.7 (§4)

5.3 (38) 11.8 (17)

'All the strains are as listed in Ch. 1, Table 1. The nuclear background
of the male parents in section A and the female parents in section B is

AJAJ.

”t = (no - pL)/s where a = pu(1 - pu)/nu + pL(1 - pL)/nL. an and pt are
the XBP from the upper and lower halves of the capsule respectively. nu
and nu are the number of the seedlings in each sample. The degrees of

freedom for t is taken to be infinity (Steel and Torrie 1980).

' P < 0.05.
'* P < 0.01.

100

limited pollen tube growth (see discussion in Chapter 1). Hence, no
meaningful comparisons can be made for these crosses and only data from
crosses using AJAJ-ICl and APCP-IV as the paternal parents are presented

in section B.

As shown in Table 3.1, the frequency of biparental plastid transmission
does not differ a great deal between seeds situated in one half of the
capsule compared to the other half. In slightly more than 50% of the
crosses, a higher frequency of biparental plastid transmission was
obtained from the upper half of the capsule. For the purpose of testing
whether there is a relationship between the differences obtained from
two halves of the capsules and the plastome types of the paternal
parents, the data obtained from all crosses with the same paternal
parent were pooled and analyzed by t-tests. According to this analysis,
the frequencies of biparental plastid transmission obtained from the two
halves of the capsule are significantly different in four out of the
seven sets of crosses. In three sets of crosses involving paternal
plastomes II-epsilon, Ill-alpha and I“1 , a higher frequency of biparastal
plastid transmission was obtained from the upper half of the capsule.
However, the reverse was true in crosses that have plastome III-gamma

from the paternal parent.

At first glance, the differences in 88? between the two halves of the

capsule are not as dramatic as one would expect if there is an extensive
loss of plastids during pollen tube development. However, limitations in
the method used for this analysis have to be considered. The lengths of

the styles of Oenothera flowers range from 4 to 10 cm, depending on the

101

species. The ovary, situated at the end of the style, is only 1 em in
length. Thus, the full length of the ovary comprises only a relatively
small fraction of the total distance a pollen tube must extend prior to
fertilization. If, as I postulate, a greater distance implies a greater
time required for growth of the pollen tube, then it follows that the
time required for a pollen tube to grow from one end of the ovary to the
other comprises a small fraction of the total time between pollination
and fertilization. The experiment presented here presumes that
appreciably more degradation of paternal plastids takes place within the
second half of the small duration as compared to the first half. It may
be beyond the power of the experiment to detect such small differences
against biological variability. Furthermore, very rapidly growing pollen
tubes may experience no degradation during their passage through the
style. Conversely, very slowly growing pollen tubes may seldom reach the
base of the ovary at all, which may explain the observation that viable
seeds could be obtained only from the upper part of the capsule in some

crosses (Chapter 1).

Taking into consideration the limitations mentioned above, the results

obtained are consistent with the hypothesis that there is degeneration

of plastids in the generative cell as a pollen tube extends towards the
egg. Furthermore, the extent of plastid degeneration may be greater in

plastomes having lower transmission efficiencies (II-epsilon and IV-

alpha) in the given nuclear background (A~1A«1 ) .

APPENDIX C. COMPARISON OF TRANSMISSION OF PLASTOMES ID AND 151

Rationale

The transmission of plastids in Oenothera can be influenced by multiple
factors contributed by either the nuclear or the plastid genomes
(Chapter 1). One major focus of the research described in this
dissertation was to examine the relationship between the efficiency of
CpDNA replication and the efficiency of plastid transmission. Electron
microscopic examination of chNA replication intermediates did not point
to any difference in the number and the location of the replication
origins between two plastomes with distinct transmission properties
(Chapter 2). However, subsequent analysis indicated that a highly
polymorphic region of chNA lay close to one of the replication origins
(Chapter 3). Conceivably, differences in this region could affect the
functioning of the origin if the sequence contains the recognition sites
for replication proteins or regulatory sequences flanking these sites.
In order to test whether differences in this region could affect the
transmission abilities of plastids, crosses were conducted to compare

two closely related plastomes that differ in this region.

Plastome 161 is derived from plastome ID under the influence of the

plastome mutator activity (Chiu et al. 1990). Four differences in

102

103

restriction fragment length have been detected between these two type I
plastomes. One of the length mutations lies within the 168 rRNAfitggl
spacer, close to one of the members of each pair of replication origin
(Chapter 3). Analysis of the nucleotide sequences show that some
repeated segments capable of forming stable stem-loop structure in the
168 rRNAetggl region of plastome 1° are absent from plastome IC1. Since
stem-loop structures are known to affect the movement of replication
forks, it was of interest to determine if the transmission of plastome

101 is more efficient than that of 1°.

Results and Discussion

In order to compare the transmission efficiencies of plastome ID and
plastome 1C1, crosses were made using the constant nuclear background of
Oenothera hookeri str. Johansen (A9AJ),*with various mutant plastids
(Chapter 1) in the maternal parent and either plastome 1° or I°1 in the
paternal parent. Seeds from these crosses were sterilized and germinated

as described in Chapter 1.

The frequencies of biparental plastid transmission (XBP) from crosses
made in different years are listed separately in Table C.l, due to my
observation that significant differences may occur in plastid
transmission from the same crosses perfbrmed in different years. As
discussed in Chapter 1, seasonal variation in transmission frequency may

be related to variation in the sizes of the maternal flowers. As a

104

Table C.1. Frequency of biparental plastid transmission in crosses with

mutant plastids transmitted by the female parent and plastome I°1 or ID

transmitted by the male parent.

 

Paternal Plastome

 

 

IC 1 ID
Maternal 1988 1989 1984/1985b 1988 1989
Plastome §:§P'(n) spy? (n) xggg (g) x a: (n) x a? 1g)_

 

I-b 23.0:3.2 (174) 22.012.7 (236) 40.915.2 ( 88) - '-

I-z - 22.812.9*(215) 9.7:2.9 (103) - 30.3:4.2*(122)
I-e 21.9:3.0 (192) -— 21.1:3.9 (109) -— -
I-ec 32.3:4.8‘( 93) -- -- 24.114.6’(87) -
II-e - 44.213.3 (231) 26.7:4.0 (120) - -

III-g 24.4:4.6 ( 86) 35.913.4 (195) 21.7:2.9 (203) —- -
IV-a 73.014.1 (115) 62.812.5*(379) 4o.o:4.1 (140) —- 36.612.7+(3l4)

 

'see Table 1.2.

bThese data are taken from Chiu gt_gl. (1988).

cThe nuclear background of this strain is albicans/“Johansen (APAJ).

+The transmission efficiencies of plastomes 191 and 1° can be compared in

these crosses.

105

consequence, meaningful comparisons between plastomes ICl and 1° can be
made only using data obtained from the same year. Three pairs of crosses
meet this criterion. Data from these crosses are marked with a plus in
Table C.1. The transmission efficiency of plastome I° was higher than
101 when the maternal parent contributed plastome I-zeta. The reverse
was true when the maternal parent was an albicans/“Johansen hybrid
strain carrying mutant plastome I-eta. However, the difference between
plastomes 191 and 1° in these two pairs of crosses was not dramatic. A
larger difference between IC1 and 1° was observed when the maternal
parent contributed mutant plastome IV-alpha. In these crosses, plastome
ICl frem the paternal parent was detected in 63* of the progeny while
plastome I° only appeared in 37% of the progeny. Overall, it appears
that the transmssion efficiencies of plastomes ICl and 1° are very
similar. The difference between the two may be visible only when the

frequency of biparental transmission is very high.

APPENDIX D: ATTEMPTS TO LOCATE THE ORIGIN OF DNA REPLICATION IN
THE CHLOROPLAST OF OENOTHERA USING

TWO-DIMENSIONAL GEL ELECTROPHORESIS

Rationale

Replication of chloroplast DNA is an important step in the multipli-
cation of chloroplasts. This dissertation examines the hypothesis that
differences in the structure of origins of CpDNA replication among
various Oenothera plastome types may contribute to the observed
differences in the efficiency of plastid transmission. In Chapter 2, two
pairs of origins of chNA replication were located using electron
microscopy. However, the electron microscopic data can only localize the
D-loop initiation sites within a 1-2 kb region. In order to study the
initiation of chNA replication, it is necessary to identify the D-loop

initiation sites more precisely.

Among alternative methods of’mapping DNA replication origins, the two-
dimensional gel electrophoresis technique developed by Brewer and
Fangman (1987) has been used most frequently. In this method, restricted
DNA fragments are seperated according to their mass in the first

dimension on a low percentage agarose gel at low voltage. In the second

106

107

dimension of gel electrophoresis, DNA fragments are seperated according
to both mass and shape on a high percentage gel at high voltage.
Deviations in mobility occur with restriction fragments derived from DNA
at different stages of replication, and these can be visualized by the
use of radiolabeled probes. Restriction fragments that carry replication
forks will form an arc ("fork arc"), which rises above and then returns
to the line of simple linear frag-nets. Fragments containing replication
bubbles (i.e. origins of replication) will migrate even slower than
those with a simple replication fork and form an additional arc (”bubble
are") above the "fork arc". The position of the origin of replication on
the restriction fragment can be determined according to the relative

positions at which the bubble arc ends.

The two-dimensional gel method has been used successfully to localize
origins of DNA replication in the highly amplified region of yeast
(Linskenns and Huberman 1988), Drosophila (Delidakis and Katatos 1989),
and mammalian chromosomes (Vaughn _e_t__al. 1990). These origins of
replication all form replication eyes with double stranded arms. It is
not known if the two-dimensional gel technique can be used to
characterize origins of replication that form displacement loops, such
as those in the genomes of mitochondria and chloroplasts. This section
describes an initial effort to employ the two-dimensional gel technique

to localize origins of chNA replication in Oenothera.

108

Results and Discussions

Chloroplast DNA from Oenothera hookeri str. Johansen carrying either
plastome I or IV was used for this project. Methods used for the
isolation, purification, and storage of chNA are as described in
Chapter 2. CpDNA was digested with restriction enzymes at room
temperature for two hours immediately before electrophoresis. A protocol
provided by Dr. M. Linskens (Roswell Park Memorial Institute, New York)
was used for two-dimensional gel electrophoresis. In a typical
experiment, the first dimension of electrophoresis was run overnight on
a 0.4x agarose gel (prepared in TAE buffer) at 0.75 to 1.0 V/cm. Both
the gel and the running buffer contained low concentrations of ethidium
bromide (0.1 ug/ml). At the end of the first dimension, the sample lane
was excised, rotated by 90 degrees, and cast on top of a 1.2% agarose
gel (prepared in TBE buffer supplemented with 0.5 ug/ml ethidium
bromide). The second dimension was run at 4 V/cm for 8-10 hours at 4°C.
At the end of the electrophoresis, the DNA was depurinated in 0.25 M RC1
for 10 min and was transferred onto a GeneScreen (NEN/DuPont) meibrane
using buffer containing'0.4 M NaOH and 1 M NaCl. In order to establish
the conditions for detecting origins of CpDNA replication, the 5 kb
EcoRI fragment (R2) that carries one of the D-loop initiation sites
identified by electron microscopy (Chapter 2) was used as a probe for

Southern hybridization.

Three restriction enzymes, EcoRI, ClaI and Sell, were used to digest
Oenothera CpDNA. These enzymes generate fragments with a wide range of

sizes (5 kb, 8kb and 18 kb respectively) that can be recognized by the

109

R2 probe. Since the retardation of replication intermediates on the
second dimension is more obvious with a smaller fragment, one might
expect EcoRI and ClaI to give better results. However, for all three
enzymes tested, only spots representing non-replicating linear fragments
were detected either by X-ray autoradiography or by a beta-scaning

device.

Since the two-dimensional gel technique has never been used for mapping
origins that initiate replication through formation of D-loops, it is
possible that the technique has to be modified to suit the D—loop type
of replication initiation. For example, the mobility of fragments
containing D-loops on the gel may not be as low as the mobility of those
carrying double-stranded replication eyes. However, the fact that even
fork arcs were not detected indicates the quality of CpDNA used in the
experiments was not optimal. There are several ways to improve the
quality of CpDNA. First of all, since replication intermediates tend to
be unstable, it may be necessary to preserve these structures before
restriction digestion. In their electron microscopic study, Meeker gt
2;; (1988) reported that no D-loops could be found on fragments smaller
than 5 kb unless chNA was cross-linked by Trioxalen before digestion
with restriction enzymes. Cross-linking is also necessary to see arcs
formed by double stranded replicating bubbles on two-dimensional gels
(Vaughn gt_gl. 1990). Secondly, the frequency of replicating DNA in the
sample should be increased. Methods such as END-cellulose chromatography
(Levine g£_gl. 1970) have been used to enrich for replicating DNA before
electrophoresis. However, the most efficient way would be to isolate

CpDNA from cells in which chNA is actively replicating. This could be

110

achieved by using suspension culture cells that can be partially

synchronized for CpDNA synthesis (Yasuda et al. 1988).

In conclusion, based on the preliminary work described above, a
significant amount of work will be required before the two-dimensional
gel technique can be adapted to study DNA replication in chloroplasts.
Nevertheless, the procedure would provide valuable information by
allowing a more precise localization of replication origins.
Furthermore, since this method is based on different shapes of the DNA
fragments at different stages of replication (Brewer and Fangman 1987),
it can also be used to locate the terminus of chNA replication

(Kolodner and Tewari 1975b).

LIST OF REFERENCES

LIST OF REFERENCES

Allgood ND, Silhavy TJ (1988) Illegitimate recombination in
bacteria. In Genetic Recombination (Kucherlapati R and Smith GR,
eds), American Society for Microbiology. pp309—330

Bedinger P, Munn M, Alberta BM (1989) Sequence-specific pausing during
in vitro DNA replication on double-stranded DNA templates. J Biol
Chem 264:16880-16886

Birky CW Jr (1978) Transmission genetics of mitochondria and
chloroplasts. Ann Rev Genet 12:471-512.

Boffey SA (1985) The chloroplast division cycle and its relationship to
the cell division cycle. In: The Cell Division Cycle in Plants,
Society for Experimental Biology Seminar Series Vol. 26 (Bryant
JA, Francis D, eds.). Cambridge University Press, Cambridge. pp.
233-246 '

Boblenz K, Nothnagel T, Metzlaff M (1990) Paternal inheritance of
plastids in the genus Daucus. Mol Gen Genet 220:489-491.

Cairns J (1963) The chromosome of Escherichia coli. Cold Spring Harbor
Symp Quant Biol 28:43-46

Chiu Rel, Stubbe W, Sears BB (1988) Plastid inheritance in Oenothera:
organelle genome modifies the extent of biparental plastid
transmission. Curr Genet 13:181-189

Chiu W—L, Johnson EM, Kaplan SA, Blasko K, Sokalski MB, Wolfson R, Sears
BB (1990) Oenothera chloroplast DNA polymorphisms associated with
plastome mutator activity. Mol Gen Genet 221:59—64

Clayton DA (1982) Replication of animal mitochondrial DNA. Cell 28:693-
705

Cleland RE (1972) Oenothera, cytogenetics, and evolution. Academic Press
Inc., London

Cornu A, Dulieu H (1988) Pollen transmission of plastid-DNA under
genotypic control in Petunia hybrids Hort. J Heredity 79:40-44.

Corriveau JL, Coleman AH (1988) Rapid screening method to detect

potential biparental inheritance of plastid DNA and results for
over 200 angiosperm species. Amer J Bot 75: 1443—1458

111

112

Corriveau JL, Coleman AW (1990) Plastid inheritance in Oenothera:

paternal input may influence transmission patterns. Curr Genet
17:327-330

de Lanversin G, Pillay DTN, Jacq B (1987) Sequence studies on the
soybean chloroplast 168-238 rDNA spacer region. Plant M01 Biol 10:
65-82

Delidakis C, Kafatos FC (1989) Amplification enhancers and replication
origins in the autosomal chorion gene cluster of Drosophila. EMBO
J 8:891-901

Elhai J, Wolk CP (1988) A versatile class of positive-selection vectors
based on the nonviability of palindrome-containing plasmids that
allows cloning into long polylinkers. Gene 68: 119-138

Glickman BW and Ripley LS (1984) Structural intermediates of deletion

mutagenesis: a role for palindromic DNA. Proc Natl Acad Sci USA
81:512-516

Gold B, Carrillo N, Tewari KK, Bogorad L (1987) Nucleotide sequence of a
preferred maize chloroplast genome template for in vitro DNA
synthesis. Proc Nat Acad Sci USA 84: 194-198

Gordon KHJ, Grouse EJ, Bohnert HJ, Herrmann RG (1982) Physical mapping
of differences in chloroplast DNA of the five wild-type plastomes
in Oenothera subsection Euoenothera. Theor Appl Genet 61:373-384

Hagemann R, Schroder (1989) The cytological basis of the plastid
inheritance in angiosperms. Protoplasma 152: 57-64

Hansmann P, Falk H, Ronai K, Sitte P (1985) Structure, composition, and
distribution of plastid nucleoids in Narcissus pseudonarcissug.
Plants 164:459-472

Heinhorst S, Cannon GC, Weissbach A (1985) Plastid and nuclear DNA
synthesis are not coupled in suspension cells of Nicotiana
tabacum. Plant Mol Biol 4:3—12

Henikoff S (1984) Unidirectional digestion with exonuclease III creates
targeted breakpoint for DNA sequencing. Gene 28:351-359

Herrmann RG, Westoff P, Alt J, Winter P, Tittgen J, Bisanz C, Sears BB,
Nelson N, Hurt E, Hauska G, Viebrock A, sebald W (1983)
Identification and characterization of genes for polypeptides of
the thylakoid membrane. In: Cifferi O, Dure L III (eds) Structure
and func tion of plant genomes. Plenum, New York, ppl43-153

Hiratsuka J, Shimada H, Whittier R, Ishibashi T, Sakamoto M, Mori M,
Kondo C, Honji Y, Sun C-R, Meng BY, Li Y-Q, Kanno A, Nishizawa Y,
Hirai A, Shinozaki K, Sugiura M (1989) The complete sequence of
the rice (Oriza sativa) chloroplast genome: intermolecular
recombination between distinct tRNA genes accounts for a major
plastid DNA inversion during the evolution of the cereals. Mol Gen
Genet 217: 185-194

113

Huang C—C, Hearst JE (1980) Pauses at positions of secondary structure
during in vitro replication of single-stranded fd bacteriophage
DNA by T4 DNA polymerase. Anal Biochem 103:127-139

Kirk JTO, Tilney-Bassett RAE (1978) The Plastids. Elsevier/North
Holland. pp 960

Koch W, Edwards K, Kbssel H (1981) Sequencing of the 168-238 spacer in a
ribosomal RNA operon of Zea mays chloroplast DNA reveals two split
tRNA genes. Cell 252203-213

Koller B, Delius H (1982) Origin of replication in chloroplast DNA of
Euglena gracilis located close to the region of variable size.
EMBO J 1:995-998

Kolodner R, Tewari KK (1975a) Presence of displacement loops in the
covalently closed circular chloroplast deoxyribonucleic acid from
higher plants. J Biol Chem 250:8840-8847

Kolodner RD, Tewari KK (1975b) Chloroplast DNA from higher plants
replicates by both the Cairns and the rolling circle mechanism.
Nature 256:708—711

Kornberg A (1980) DNA replication. WH Freeman and Company. pp 724

Kuroiwa T, Suzuki T, Ogawa K, Kawano S (1981) The chloroplast nucleus:
distrubution, number, size and shape, and a model for the
multiplication of the chloroplast genome during chloroplast
development. Plant Cell Physiol. 22:381-396

Lammpa GK, Bendich AJ (1979) Changes in chloroplast DNA levels during
development of pea (Pisum sativum). Plant Physiol 64:126-130

Lawrence ME, Possingham JV (1986) Microspectrofluorometric measurement
of chloroplast DNA in dividing and expanding leaf cells of
Spinacia oleracea. Plant Physiol 81:708-710

Levine AJ, Kang HS, Billheimer FE (1970) DNA replication in SV40
infected cells. I. Analysis of replicating SV40 DNA. J Mol Biol
50:549-568

Levinson G, Gutman GA (1987) Slipped-strand mispairing: a major
mechanism for DNA sequence evolution. Mol Biol Evol 4:203-221

Lindbeck AGC, Rose RJ, Lawrence ME, Possingham JV (1987) The role of
chloroplast membranes in the location of chloroplast DNA during
the greening of Phaseolus vulgaria etioplasts. Protoplasma 139:92-
99

Linskenns MHK, Huberman JA (1988) Organization of replication of
ribosomal DNA in Saccharggyces cerevisiae. Mol Cell Biol 8:492?—
4935

Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning. A
Laboratory Manual. Cold Spring Harbor, NY

114

Markowicz Y, Mache R, Loiseaux-De Goer (1988) Sequence of the plastid
rDNA spacer region of the brown alga Pylaiella littoralis (L.)
Kjellm. Evolutionary significance. Plant Mol Biol 10:465-469

Marsh JL, Erfle M, Wykes EJ (1984) The p10 plasmid and phage vectors
with versatile cloning sites for recombinant selection by
insertional inactivation. Gene 32:481-485

Masoud SA, Johson LB, Sorensen EL (1990) High transmission of paternal
plastid DNA in alfalfa plants demonstrated by restriction fragment
polymorphic analysis. Theor Appl Genet 79:49-55

Massenet O, Martinez P, Seyer P, Briat J-F (1987) Sequence organization
of the chloroplast ribosomal spacer of Spinacia oleracea including
the 3’ end of the 168 rRNA and the 5’ end of the 238 rRNA. Plant
Mol Biol 10:53—63

 

Meeker R, Nielsen B, Tewari KK (1988) Localization of replication
origins in pea chloroplast DNA. Mol Cell Biol 8:1216-1223

Moore GP (1983) Slipped-strand mispairing and the evolution of introns.
Trends Biochem Sci 8:411-414

Mullet JE (1988) Chloroplast development and gene expression. Ann Rev
Plant Physiol Plant Mol Biol 39:475-502

Neale DB, Marshall KA, Sederoff RR (1989) Chloroplast and mitochondrial
DNA are paternally inherited in Sequoia sempervirens D Don Endl.
Proc Natl Acad Sci USA 86:9347-9349

Neal DB, Wheeler NC, Allard RW (1986) Paternal inheritance of
chloroplast DNA in Douglas-fir. Can J For Res 16:1152—1154

Ohyama K, Fukuzawa H, Kohchi T, Shirai H, Sano T, Sano S, Umesono K,
Shiki Y, Takeuchi M, Chang Z, Aota S, Inokuchi R, Ozeki H (1986)
Complete nucleotide sequence of liverwort Marchantia polymorpha
chloroplast DNA. Plant Mol Biol Reporter 40:148-175

Palmer JD (1982) Physical and gene mapping of chloroplast DNA from
Atriplex triangularis and Cucumis sativa. Nucl Acids Res 10:1593-
1605

 

Palmer JD (1985) Comparative organization of chloroplast genomes. Ann
Rev Genet 19:325-354

Palmer JD, Thompson WF (1981) Rearrangements in the chloroplast genome
of mung bean and pea. Proc Natl Acad Sci USA 78:5533-5537

Piskur J (1988) Transmission of yeast mitochondrial loci to progeny is
reduced when nearby intergenic regions containing orinep
sequences are deleted. Mol Gen Genet 214:425-432

Piskur J (1989) Transmission of the yeast mitochondrial genome to
progeny: the impact of intergenic sequences. Mol Gen Genet
218:161-168

115

Ravel-Chapuis P, Heizmann P, Bigon V (1982) Electron microscopic
localization of the replication origin of Egglena gracilis
chloroplast DNA. Nature 300:78-81

Santulli A, Casale A, Mazza A (1983) Cairns replicative intermediates in
Acetabularia chloroplast DNA. J Submicrosc Cytol 15:843-847

Schmitz UK, Kowallik KV (1987) Why are plastids maternally inherited in
Epilobium? Plant Sci 53:139-145

Schotz F (1954) Uber Plastidenkonkurrenz bie Oenothera. Planta 43:183—
240

Schbtz F (1974) Untersuchungen fiber die Plastidenkonkurrenz bie
Oenothera. IV Der einfluB des Genoms auf die
Durchsetzungsfahigkeit der Plastiden. Biol Zentralbl 93:41-64

Schotz F (1975) Untersuchungen fiber die Plastidenkonkurrenz bie
Oenothera. V. Die Stabilitét der Konkurrenzfahigkeit bei
Verwendung verschiedenartiger mutierter test Plastiden. Biol
Zentralbl 94:17-26

Schumann CM, Hancock JF (1989) Paternal inheritance of plastids in
Medicago sativa. Theor Appl Genet 78:863-866

Scott NS, Cain P, Possingham V (1982) Plastid DNA levels in albino and
green leaves of the <a1bostrians> mutant of Hordeum vulgare. Z
Pflanzenphysiol 108:187-192

Shinozaki K, Ohme M, Tanaka M, Wakasugi T, Hayshida N, Matsubayasha T,
Zaita N, Chunwongse J, Obokata J, Yamaguchi-Shinozaki K, Ohto C,
Torazawa K, Meng BY, Sugita M, Deno H, Kamogshira T, Yanada K,
Kusuda J,Takaiwa F, Kata A, Tohdoh N, Shimada H, Sugiura M (1986)
The complete nucleotide sequence of the tobacco chloroplast
genome: its gene organization and expression. EMBO J 5:2043-2049

Stine M, Sears BB, Keathley DE (1989) Inheritance of plastids in
interspecific hybrids of blue spruce and white spruce. Theor Appl
Genet 78:768-774

Stockhaus J, Schell J, Willmitzer L (1989) Correlation of the expression
of the nuclear photosynthetic gene ST-LSl with the presence of
chloroplasts. EMBO J 8:2445-2451

Stubbe W (1959) Genetische analyse des Zusammenwirken von Genom und
Plastom bei Oenothera Z Vererbungsl 90:288-298

Stubbe W (1963) Extrem disharmonische Genom-Plastom-Kombinationen und
vfiterliche Plastidenvererbung bie Oenothera. Z Vererbungsl 94:
392-411

Stubbe W (1964) The role of the plastome in the evolution of the genus
Oenothera. Genetics 35:28-33

116

Stubbe W (1989) Oenothera - An ideal system for studying the
interactions of genome and plastome. Plant Mol Biol Report
7:245-257

Stubbe W, Herrmann RG (1982) Selection and maintenance of plastome
mutants and interspecific genome/plastome hybrids from Oenothera.
In: Methods of Chloropalst Molecular Biology (Edelman M, Hallick
R, Chua NH eds). Elsevier Biomecidal Press, Amsterdam. ppl49-165

Sugiura M (1989) The chloroplast chromosomes in land plants. Annu Rev
Cell Biol 5:51-70

Svab Z, Hajdukiewicz P,Maliga P (1990) Stable transformation of plastids
in higher plants. Proc. Natl. Acad. Sci. USA 87:8526-8530

Szmidt AE, Alden T, Hallgren J—E (1987) Paternal inheritance of
chloroplast DNA in Larix. Plant Mol Biol 9:59-64

Taylor WC (1989) Regulatory interactions between nuclear and plastid
genomes. Annu Rev Plant Physiol Plant Mol Biol 40:211-233

Thomson WW, Whatly JM (1980) Development of nongreen plastids. Ann Rev
Plant Physiol 31:375-394

Tilney-Bassett RAE, Almouslem AB (1989) Variation in plastid inheritance
between Pelargonium cultivars and their hybrids. Heredity 63:145-
153

Umesono K, Ozeki H (1987) Chloroplast gene organization in plants. TIG
3:281-287

Waddell J, Wang X-M, Wu M (1984) Electron microscopic localization of
the chloroplast DNA replicative origins in Chlamydomonas
reinhardii. Nucl Acids Res 12:3843-3856

Walbot V and Coe EH (1979) Nuclear gene iojap conditions a programmed
change to ribosome-less plastids in Zea mays. Proc Natl Acad Sci
USA 76:1760-1764

Weaver DT, DePamphilis ML (1982) Specific sequences in native DNA that
arrest synthesis by DNA polymerase alpha. J Biol Chem 257:2075-
2086

Wu M, Lou JK, Chang DY, Chang CH, Nie 20 (1986) Structure and function
of a chloroplast DNA replication origin of Chlamydomonas
reinhardii. Proc Natl Acad Sci USA 83:6761-6765

Yamada T, Shimaji M, Fukuda Y (1986) Characterization of a chNA
sequence from Chlorella ellipsoidea that promotes autonomous
replication in yeast. Plant Mol Biol 6:245-252

Yasuda T, Kuroiwa T, Nagata T (1988) Preferential synthesis of plastid
DNA and increased replication of plastids in cultured tobaco cells
following medium renewal. Planta 174:235-241.

117

Vaughn JP, Dijkwel PA, Hamlin JL (1990) Replication initiates in a broad

zone in the amplified CHO dihydrofolate reductase domain. 61:1075-
1087

Zhang H, Scholl R, Somerville C (1988) Double stranded DNA sequencing as
a choice for DNA sequencing. Nucl Acids Res 16:1220

Zuker M (1989) Computer prediction of RNA structure. Meth Enzymol
180:262-288

i

“@11qu

IES
'l

y),