Title: Influence of the nuclear genome on transcription of a maize mitochondrial region associated with male sterility and toxin sensitivity
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00099319/00001
 Material Information
Title: Influence of the nuclear genome on transcription of a maize mitochondrial region associated with male sterility and toxin sensitivity
Physical Description: vi, 118 leaves : ill. ; 28 cm.
Language: English
Creator: Kennell, John Carlyle
Publisher: s.n.
Copyright Date: 1987
Subject: Genomes   ( lcsh )
Plant mitochondria   ( lcsh )
Male sterility in plants   ( lcsh )
Corn   ( lcsh )
Plant Pathology thesis Ph. D
Dissertations, Academic -- Plant Pathology -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Statement of Responsibility: by John Carlyle Kennell.
Thesis: Thesis (Ph. D.)--University of Florida, 1987.
Bibliography: Bibliography: leaves 104-117.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00099319
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 001051125
oclc - 18550097
notis - AFD4343


This item has the following downloads:

influenceofnucle00kenn ( PDF )

Full Text







"It is hardly an exaggeration to say that Nature tells us, in the most

emphatic manner, that she abhors perpetual self-fertilization".

- Final sentence of Charles Darwin's book The Various Contrivances by

which Orchids are Fertilized by Insects".


I thank Daryl Pring for his insightful guidance, constant

enthusiasm, and for providing an atmosphere that stimulates productivity

as well as creativity. His professionalism is an admirable quality that

has hopefully been instilled in me. I would also like to thank the

other members of my committee and other faculty, post-doctoral

researchers and students that have helped in my research and education.

I am particularly grateful to the technical advice given by Al Smith and

Roger Wise as well as the help provided by fellow researchers of 1404

Fifield; Al Fliss, Sandy Fogg, Jeff Mullen and Torbert Rocheford.

Special thanks go to Robin Roffey for putting up with my "thirst for

knowledge" and enabling me to get this part of my life together.

Finally, I thank my family for their continued support and encouragement

for all my interests and efforts.


PREFACE .............................................................ii

ACKNOWLEDGEMENTS................................................... iii

ABSTRACT.............................................................. v


I INTRODUCTION....................................................

Mitochondrial Genomes............................................
Mitochondrial Transcription....................................5
Cytoplasmic Male Sterility.............................. .......11
Host-selective Toxins and T Cytoplasm of Maize.................14
Nuclear-Mitochondrial Interactions......................... ...19


Materials and Methods...........................................25
Results and Discussion.........................................27

TEXAS MALE STERILE CYTOPLASM..................................33

Introduction................................................... 33
Materials and Methods...........................................35
Results ....................................................... 36
Discussion.................................................... 52


Materials and Methods.......................................... 61

V DISCUSSION.....................................................92

LITERATURE CITED....................................................104

BIOGRAPHICAL SKETCH.................................................118

Abstact of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



John Carlyle Kennell

December, 1987

Chairman: Daryl R. Pring
Major Department: Plant Pathology

Maize plants carrying T cytoplasm are male sterile and highly

susceptible to the fungal pathogens Cochliobolus heterostrophus and

Phyllosticta maydis. Mitochondria isolated from T cytoplasm plants are

sensitive to toxins produced by the pathogens. A gene unique to T

mtDNA, urfl3-T, is associated with the phenotypes of cms and toxin

sensitivity. The nuclear genes Rfl and Rf2 restore plants with T

cytoplasm to fertility and decrease the sensitivity of T mitochondria to

the toxins; other unidentified nuclear genes, besides the Rf genes, also

appear to influence the level of toxin sensitivity.

The urfl3-T gene is 5' to the gene ORF25 and adjacent to a 5 kb

duplicated region of the maize mitochondrial genome that also is 5' to

atp6. Transcripts of atp6, urfl3-T and ORF25 from N, T and T-restored

cytoplasms were analyzed in five nuclear backgrounds. Sequences within

the 5 kb repeated region promote transcription of atp6 and the

cytoplasms were analyzed in five nuclear backgrounds. Sequences within

the 5 kb repeated region promote transcription of atp6 and the

cotranscribed genes urfl3-T and ORF25. The steps in transcript

maturation were determined for these genes. Differences between N and T

atp6 transcripts were associated with two small DNA insertions that

apparently create an RNA processing site in the T mitochondrial genome.

The 5' termini of a transcript unique to fertility restored mtRNAs

mapped to position +14 of the urfl3-T gene and was not a suitable

substrate for capping with guanylyl transferase. This transcript is

detected in backgrounds containing the dominant gene Rfl and is

associated with a reduction in the urfl3-T gene product, a 13 kD

protein. The primary role of the RF1 gene may involve aspects of

translation rather than an RNA processing event which creates the

restorer-specific transcript. Nuclear background, separate from

fertlity restoring genes, influenced the synthesis and abundance of

specific urfl3-T/ORF25 transcripts.



Mitochondrial Genomes

Prior to the divergence of the fungal, animal and plant kingdoms, a

single endosymbiotic association between free-living aerobic bacteria

and primitive eukaryotic cells is one hypothesis for the origin of

mitochondria (rev. by Gray and Doolittle, 1982). Despite recognizable

differences, mitochondria from all eukaryotic kingdoms appear to have

originated from a specific subdivision of the purple eubacterial phylum

(Yang et al., 1985) and have maintained their basic structure and

function. The fundamental mechanisms of the electron transport system

and oxidative phosphorylation have been highly conserved throughout the

development and divergence of fungal, animal and plant species (Moore

and Rich, 1980). Plant mitochondria, however, have additional routes

for oxidizing substrates and a terminal oxidase not found in mammalian

mitochodria which may enable plant cells to function under a wide

variety of metabolic and/or environmental conditions (Moore and Rich,


Between 300 to 400 different polypeptides are estimated to be in the

protein component of yeast mitochondria. Of these, approximately 10%

are synthesized in the mitochondria and the remainder are synthesized in

the cytoplasm (Schatz and Mason, 1974). The same basic set of

mitochondrial-synthesized proteins are found in mitochondria of all

eukaryotic species yet the mitochondrial genomes which encode these

proteins vary considerably with regard to size and organization.

Animal mitochondrial genomes range from 16 to 23 kb (Grivell, 1983;

Kessler and Avise, 1985) and exhibit high conservation in gene

organization. The complete nucleotide sequence of the human

mitochondrial genome (Anderson et al., 1981) identified 13 protein

coding regions, two rRNAs and 22 tRNAs (Borst et al., 1984). All

metazoan species studied thus far contain the same 13 proteins although

their order varies slightly (Anderson et al., 1982; Bibb et al., 1981;

Roe et al., 1985). These mitochondrial genomes are very efficient,

containing little non-coding regions between genes (Ojala et al., 1981).

Fungal mitochondrial genomes are generally larger than animal

mitochondrial genomes, ranging from 19 to 176 kb (Clark-Walker, 1980;

Hintz et al., 1985). Besides three additional tRNAs, RNA processing

enzymes encoded within an intron of the cytochrome b gene (cob) gene and

the Fo-ATPase subunit 9 gene found in some species (e.g. yeast), fungal

mitochondrial genomes encode the same set of proteins as animal

mitochondrial genomes (Borst et al., 1984). The majority of the

increased size of the fungal mitochondrial genomes is accounted for by

non-transcribed, A-T rich spacer regions as well as introns (Bernardi,


The size and organization of higher plant (angiosperms)

mitochondrial DNAs (mtDNAs) are quite different than their animal and

fungal counterparts. The size of plant mitochondrial genomes range from

208 (Palmer and Herbon, 1987) to 2500 kb (Ward et al., 1981), and can

vary greatly within a single plant family. For example, an eight-fold

size difference is found within the cucurbit family with sizes of the

watermelon and muskmelon mitochondrial genomes estimated at 330 and 2500

kb, respectively (Ward et al., 1981). The increased size is not the

result of repetitive DNA, as less than 10% of the cucurbit mtDNAs are

repetitive (Ward et al., 1981), nor is it a result of high A-T spacer

regions as is the case in fungal mtDNAs. The base composition of plant

mtDNAs examined to date are consistently between 46 to 51% G-C (Pring

and Lonsdale, 1985). An explanation for the increased size of plant

mitochondrial genomes in comparison to animal and fungal mitochondrial

genomes is still unavailable, but there are some unique aspects of

higher plant mitochondria that may have contributed to maintaining or

expanding their size.

Whereas most animal and fungal mitochondrial genomes are singular

circular molecules, all but one (Brassica hirta; Palmer and Herdon,

1987) of the plant mitochondrial genomes studied thus far exist in

multipartite circular molecules arising through intra- and

intermolecular recombinational events. The complete mtDNA physical maps

for two Brassica species and for maize have been reported (Palmer and

Shields, 1984; Stern and Palmer, 1986; Lonsdale et al., 1984). The

mitochondrial genomes of turnip and spinach exist as three molecules; a

complete or "master" circular molecule and two subcircular molecules

that arise via recombination through repeated DNA regions. For example,

the 218 kb turnip mitochondrial genome contains two direct repeats of

about two kb, spaced 135 and 83 kb from each other. Intramolecular

recombination through these repeats results in subcircular molecules of

135 and 83 kb (Palmer and Shields, 1984). The maize mitochondrial

genome is more complex as it contains at least six repeated regions,

five of which are known to participate in recombination (Lonsdale et

al., 1984). The proposed model of recombination predicts a minimum of

nine subcircular molecules in addition to the 570 kb master circle.

Being typically below 10% of the genome, the repetitive DNA is not

significant for enlarging the genome but for conferring the ability for

intra- and intermolecular recombination which has played a major role in

creating the variability found among plant mitochondrial genomes.

The large size of plant mitochondrial genomes could allow a

substantial increase in its coding capacity. In contrast to the 13

protein coding regions in the animal mitochondrial genomes, isolated

plant mitochondria synthesize at least 20 to 30 polypeptides (Leaver et

al., 1983; Hack and Leaver, 1983). At present, it is known that plant

mitochondrial genomes encode at least four genes not found in animal

mtDNA. These include the 55 rRNA gene (Bonen and Gray, 1980), a

ribosomal binding protein (Bland et al., 1986), alpha-FIATPase (atpA;

Braun and Levings, 1985), and subunit 9 of Fo-ATPase (atp9; Dewey et

al., 1985b). Of these genes, only atp9 has been detected in the

mitochondrial genomes of fungal species (Hensgens et al., 1979).

Measurement of the actual coding capacity of the turnip

mitochondrial genome has shown that there is approximately 60 kb of

abundant, nonoverlapping RNAs (28% of the genome), divided into 24

distinct transcript regions (Makaroff and Palmer, 1987). When less

abundant transcripts are included, the percentage of the total genome

that is transcriptionally active increases to a level similar to that

detected in watermelon (70%; Stern and Newton, 1985) and Brassica napus

(100%; Carlson et al., 1986). These data suggest that plant mtDNAs may

encode a far greater number of proteins than their animal and fungal

counterparts. The synthesis of proteins in isolated mitochondria and

subsequent 2-D gel electrophoresis has detected only 30 polypeptides

(Hack and Leaver, 1983), yet this type of analysis may not be capable of

detecting less abundant polypeptides or those induced by external

factors or by specific plant tissues. It has been reported that

mitochondria isolated from different plant organs and the same organ at

different developmental stages exhibit qualitative and quantitative

differences in the polypeptides they synthesize (Newton and Walbot,

1985). It is also known that certain plant mitochondrial respiratory

mechanisms are active only in specific tissues (Moore and Rich, 1980).

Although it is not commonly believed that plant mitochondrial genomes

encode significantly more proteins than animal and fungal mitochondrial

genomes, there is sufficient evidence indicating that there are at least

a few more genes in plant mitochondrial genomes and that these genomes

are more complex and diversified than the mitochondrial genomes of

animals and fungi.

Mitochondrial Transcription

Transcription of the mammalian mitochondrial genome is relatively

simple and highly coordinated. Transcription is initiated in the

displacement loop region, the site of the heavy strand replication

origin. A consensus sequence 5' CANACC(G)CC(A)AAAGAPyA 3' is present on

both DNA strands in this region and is regarded as a candidate promoter

(Chang and Clayton, 1984). The entire genome is transcribed, from both

strands, ensuring stoichiometric transcription of all genes. The 13

protein coding genes are all located on the heavy strand and most are

directly adjacent to tRNAs. The tRNAs are posttranscriptionally cleaved

from the primary transcript and the intermediate mRNAs are

polyadenylated (Ojala et al., 1981). The relative abundance of the

mitochondrial rRNAs is apparently increased by an attenuation process

occurring on the heavy strand at the 3' end of the 16S rRNA gene,

resulting in the release of transcripts containing both the large and

small rRNAs (Clayton, 1984). The punctuation of the structural genes

with tRNAs and the process of attenuation both help make mammalian

mitochondrial transcription extremely efficient.

Gene order and mode of transcription are much different in fungal

mitochondria than in animal mitochondria. In yeast, genes are scattered

around the genome without apparent order as different orders are found

in different yeast species (Tabak et al., 1983). The genetic

information is almost exclusively coded on one strand and 19 sites of

transcription initiation have been identified (Christianson and

Rabinowitz, 1983). Preceeding the 5' end of yeast primary transcripts

ribosomall and protein coding genes) is the nine nucleotide sequence,

5'T/ATATAAGTA 3', representing positions -8 to +1 (Osinga and Tabak,

1982; Christianson and Rabinowitz, 1983). Most of the yeast

transcriptional units appear to be monogenic, although a few multigenic

transcripts (Christianson and Rabinowitz, 1983) and multiple initiation

sites for a single gene (ATPase subunit 9) have been reported (Edwards

et al., 1983).

To further characterize the yeast mitochondrial promoter region, in

vitro transcription systems have examined the effect of mutations within

and adjacent to the nonanucleotide consensus sequence. Mutations within

the nonanucleotide sequence were found to either completely inhibit

transcription (Biswas and Getz, 1986) or reduce transcriptional

efficiency 80 to 90% (Schinkel et al., 1987). Changes at positions +2,

+3, and +4 also affected transcriptional efficiency (Biswas and Getz,

1986; Wettstein-Edwards et al., 1986) whereas changes at +1 had little

effect (Schinkel et al., 1987). In general, changes in promoter

efficiency could not be correlated to the energy requirements of strand

dissociation or to any other obvious mechanisms. Recently, it has been

shown that efficient in vitro promoter sequences do not necessarily

function well in vivo (Francisci et al., 1987).

The steps involved in posttranscriptional processing of fungal

mitochondrial genes have received a great deal of attention. Unlike

mitochondrial genes in animals, many fungal mitochondrial genes contain

intervening sequences and most genes undergo some sort of

posttranscriptional processing. Processing sites for maturation of

multigenic transcripts have been determined (Osinga et al., 1984) as

well as consensus sequences involved in processing of introns (group I,

Waring and Davies, 1983; group II, van der Veen et al., 1986). Fungal

mitochondrial genes may also be translationally regulated (rev. by Fox,

1986) or undergo posttranslational modification (Pratje et al., 1983).

The genes that control the RNA processing and translational events are

encoded in the nucleus or in the mitochondria, and in some cases, within

the gene they affect (e.g. maturases; rev. by Grivell and Borst, 1982).

A genetic system analogous to mitochondrial genomes exists in

chloroplasts of higher plants. Chloroplast genomes range in size from

120 to 217 kb (Palmer et al., 1987), with most of the size variation

being accounted for by differences in a large inverted repeat (rev. by

Zurawski and Clegg, 1987). The complete nucleotide sequence of two

chloroplast genomes liverwortt, Ohyama et al., 1986; tobacco, Shinozaki

et al., 1986) reveals approximately 50 protein-coding genes, about 30

tRNA genes and four rRNA genes. Chloroplast DNA sequences important in

gene regulation (e.g. promoters, ribosome binding sites, and

transcription termination sites) are similar to those of prokaryotic

genomes, supporting the endosymbiotic theory of the origin of

chloroplasts. However, expression of many genes involves

posttransciptional processing. Furthermore, in the tobacco chloroplast

genome, eight of the protein-coding genes and seven tRNAs contain

introns, many of which have features similar to eukaryotic introns.

This has led to speculation that ancestral photosynthetic prokaryotes

had introns in their genomes which have been retained in chloroplasts

(Shinozaki et al., 1986).

In vitro expression studies have identified two critical regions in

the chloroplast promoter. The first is analogous to the prokaryotic

'-35' consensus region and the second to the '-10', or TATA region

(Link, 1984). It is important to note that although consensus sequences

can be accurately defined, in some prokaryotic systems there appears to

be great tolerance for deviation from the consensus sequence (Siebenlist

et al., 1980).

The complexities of plant mitochondrial genomes have made gene

expression studies difficult and often misleading. Initial transcript

analyses found that the majority of maize mitochondrial genes have very

complex Northern patterns (rev. by Eckenrode and Levings, 1986). Some

of the maize mitochondrial transcript analyses are complicated by the

association of certain genes with repeated regions. For example, atpA

is internal to a 12 kb repeat in N cytoplasm; cytochrome c oxidase,

subunit II (coxll) is immediately 3' to a 2 kb repeat (Dawson et al.,

1986) and the 5' flanking region of atp6 is repeated in T cytoplasm

(Dewey et al., 1986). The location of mitochondrial genes in or

adjacent to repeated regions also has been found in other plant species.

Some of these include the 5S, 18S, and 26S rRNA genes in wheat (Bonen

and Gray, 1980; Falconet et al., 1984), atpA in sorghum and Oenothera

(Dawson et al., 1986), cytochrome c oxidase subunits I and III (coxl and

coxIII) in Oenothera (Heisel et al., 1987), and coxll in soybean

(Grabau, 1987). There also is one report of a duplicate gene, not

associated with a larger repeat (petunia atp9; Rothenberg and Hanson,

1987). It is clear that many plant mitochondrial genes have complex

transcriptional patterns which are unrelated to associations with

repeated regions, yet an equal number exhibit simple, one-transcript

Northern patterns. For example, five of the ten known genes in the

turnip mitochondrial genome have a single major transcript (Makaroff and

Palmer, 1987).

At present, there is only one confirmed example of an intron in

higher plant mitochondrial genomes. The coxlI gene in three monocot

species [maize (Fox and Leaver, 1981), rice (Kao et al., 1984) and wheat

(Bonen et al., 1984)) contains a single intron which is not detected in

dicot species [Oenothera (Heisel and Brennicke, 1983), and pea (Moon et

al., 1985)]. An intron also occurs in the gene for complex I of the

NADH-ubiquinone oxidoreductase detected in tobacco, maize (Bland et al.,

1986) and watermelon (Stern et al., 1986), although it has yet to be

determined if this is a functional gene.

There are a few examples of multigenic transcripts in plant

mitochondrial genomes, such as atp9 and S13-like genes in tobacco (Bland

et al., 1986), urfl3-T and ORF25 genes in T cytoplasm maize (Dewey et

al., 1986), and coxll and initiator met-tRNA in soybean (Grabau, 1987).

However, the majority of plant mitochondrial genes appear to be

independently transcribed. Genetic information is coded on both DNA

strands in plant mitochondrial genomes (Dawson et al., 1986) yet without

a particular order. In the turnip and maize mitochondrial genomes,

genes are located throughout the genome and are not clustered in

functional units (e.g. subunits of the ATPase complexes or cytochrome

oxidases) either on the master circle or subgenomic circles (Makaroff

and Palmer, 1987; Dawson et al., 1986).

Transcriptional initiation sites have not been identified for plant

mitochondrial genomes, although there does appear to be a short sequence

common to the 5' end of some transcripts (Isaac et al., 1985; Hiesel and

Brennicke, 1985; Moon et al., 1985; Young et al., 1986; Rothenberg and

Hanson, 1987). However, none of the plant studies distinguished primary

transcripts from processed transcripts as has been accomplished in other

mitochondrial genomes through the use of a capping enzyme. The capping

enzyme, guanylyl transferase, specifically caps transcripts that retain

a di- or tri-phosphate 5' terminus and has been used to locate

transcript initiation sites in human (Chang and Clayton, 1984) and yeast

(Christianson and Rabinowitz, 1983) mitochondrial genomes as well as the

chloroplast genome in maize (Mullet et al., 1985).

A transcript termination sequence has been proposed for some plant

mitochondrial genes (Schuster et al., 1986). It appears to function in

a manner similar to bacterial transcript terminators by the generation

of a stem-loop structure. There is no evidence of translational control

over the expression of plant mitochondrial genes, but some chloroplast

genes appear to be translationally regulated (Rock et al., 1987).

Cytoplasmic Male Sterility

Mutational analysis is a powerful method for understanding the

functions of a specific gene product. Many mitochondrial mutations are

available for analysis in lower eukaryotes due to the ability to

maintain mutants on non-fermenting carbon sources. Unfortunately, most

mutations in mitochondrial genes of higher eukaryotes appear to be

lethal as few have been documented. A few cytoplasmic traits are known

in higher plants; some involve the chloroplast [tentoxin sensitivity

(Durbin and Uchytil, 1977); triazine resistance (Pfister et al., 1981)]

and others are associated with mitochondria [disease susceptibility

(Hooker et al., 1970), non-chromosomal stripe (Shumway and Bauman, 1967;

Newton and Coe, 1986) and most types of cytoplasmic male sterility (rev.

by Hanson and Conde, 1985)].

Cytoplasmic male sterility (cms) is a cytoplasmically-controlled

alteration in the development of the male gametophyte resulting in

nonfunctional pollen. This phenomenon is usually specific to

microsporogenic and microgametogenic tissues, although abnormalities in

megasporogenesis and megagametogenesis are associated with a few cms

systems (Grun, 1976). In the T-cms system of maize, all plant tissues

are associated with an increased susceptibility to two fungal pathogens

(see below). Cytoplasmic male sterility has wide distribution among

plant families, being reported in 22 different families and in over 150

species (Edwardson, 1970). The prevalence of cms among plant families

suggests that a common mechanism may be involved, however analyses of

cms in major crop species have yet to uncover a simple mechanism

explaining cytoplasmically controlled pollen abortion. Many mechanisms,

or at least many manifestations of a few mechanisms, appear to be


Cytoplasmic male sterility is an economically desirable trait for

the production of hybrid seed. Seeds produced on cms plants are hybrid,

and homogeneous populations of male sterile plants are easier to

maintain with maternally inherited cms lines than with nuclear male

sterile lines. Interspecific crossing has been the most common way of

producing defined cms types (Laser and Lersten, 1972). Described as

alloplasmic cms, the interspecific crosses are believed to create an

incompatibility between the nuclear genome of one species and the

cytoplasm of another which is manifested as cms. The idea that cms

represents a breakdown in a nuclear-cytoplasmic interaction is further

supported by the presence of nuclear genes that "restore" pollen

fertility. The process of fertility restoration was first detected in

maize when the cms condition was being transferred into standard inbred

lines for use in hybrid seed production (Jones et al., 1957). The

progeny of cms lines crossed to certain inbred lines were fertile, even

though they had maintained a "sterile" cytoplasm. The ability to

restore pollen fertility was inherited in a mendelian fashion and

usually controlled by one or two genes. Nuclear-controlled fertility

restoration was subsequently found in most cms systems [40% of cms types

that had arise through interspecific crosses; 80% in

intraspecific-derived cms types (Grun, 1976)].

Restoration of pollen fertility can be monogenically or

polygenically controlled and fertility restoration genes (Rf) have been

found to be recessive or dominant. The Rf genes are classically used to

distinguish specific cms cytoplasms within a species. In maize, for

example, there are three distinct types of cms cytoplasms (T, S, and C),

each distinguished by different Rf genes; T cytoplasm is restored by Rfl

and Rf2, S cytoplasm by Rf3 and C cytoplasm by Rf4 (Duvick, 1965).

Multiple alleles for Rf3 and Rf4 may be present in certain inbred maize

lines (Laughnan and Gabay-Laughnan, 1983).

Cytological studies demonstrated that abnormalities associated with

cms systems can occur at almost every stage of pollen development and in

different anther tissues (rev. by Laser and Lersten, 1972). Some common

cytological abnormalities include 1) tapetal abormalities, 2)

disfunctioning of the stamen vasculature, and 3) failure of callose

dissolution around the developing microspores (rev. by Laser and

Lersten, 1972). An example of the variability in the developmental

stage at which pollen abortion occurs is observed in the differences

between sporophytic and gametophytic restoration. In T-cms and C-cms

systems of maize, restoration is sporophytic while in S-cms, restoration

is gametophytically controlled. Thus, in Rf3rf3 plants carrying S

cytoplasm only 50% of the pollen is functional as restoration is

dictated by the genotype of the individual microspore; the presence of

Rf3-carrying microspores does not appear to affect abortion of

rf3-carrying microspores in the same anther (Buchert, 1961).

The most common defect reported among cms systems is a premature

breakdown of the tapetal cells, a layer of nutritive cells that surround

the developing microspores. At least four monocot genera and nine dicot

genera have cms systems in which the tapetal cells degenerate

prematurely (Laser and Lersten, 1972; Horner, 1977; Peters and Jain,

1987). Tapetal abnormalities are not restricted to cms systems, as they

also occur in nuclear male-sterile mutants of soybean (rev. by Graybosch

and Palmer, 1984). In the T-cms system of maize, ultrastructural

studies have shown that the mitochondria of tapetal cells are frequently

the first organelle to degenerate (Warmke and Lee, 1977). The

mitochondria in the tapetal cells of both N and T-cms maize plants were

found to rapidly increase in number immediately preceding the stage in

which T-cms anthers degenerate. The number of mitochondria within an

individual tapetal cell increased almost 40-fold during

microsporogenesis and microgametogenesis. This observation led to a

proposed mechanism of pollen abortion in which the mitochondria of T

cytoplasm malfunction under the high stress conditions associated with

tapetal cell development (Warmke and Lee, 1978).

Host-selective Toxins and T Cytoplasm of Maize

The T cms system of maize has an additional cytoplasmically

inherited trait, that of disease susceptibility. Maize plants having T

cytoplasm are highly susceptible to the fungal pathogens Cochliobolus

heterostrophus, race T, causal agent of southern corn leaf blight

(Hooker et al., 1970), and Phvllosticta mavdis Amy and Nelson, causal

agent of yellow corn leaf blight (Scheifele et al., 1969). The

association of cms with disease susceptibility has never been separated.

Whereas cms is solely manifested in anther tissues, T cytoplasm plants

are susceptible to the pathogens at all developmental stages, indicating

that host factors involved in this host-parasite interaction are

constituitively expressed. Understanding the molecular basis for

susceptibility of T-cms plants to the pathogens could conceivably

provide important clues to the mechanism of cms, and vice versa.

In the late 1960s, T cytoplasm was the primary maize cms cytoplasm

used for hybrid seed production leading to a virtual monoculture of

plants with T cytoplasm in the U.S. [up to 80% in 1970 (Tatum, 1971)].

This set the stage for the 1970 epiphytotic of the southern corn leaf

blight. The imperfect stage of C. heterostrophus [anamorph: Bipolaris

maydis (Nisik.) Shoemaker], race T, initially colonized maize plants in

Florida and quickly spread throughout the South and parts of the

Midwest. The epidemic caused a 50% loss in the corn crop in the corn

belt states and almost a complete loss in the southern states, resulting

in a total loss of over 1 billion dollars (rev. by Ullstrup, 1972).

Cochliobolus heterostrophus, race T, and Phyllosticta maydis

produce toxins that specifically affect T cytoplasm (Bhullar, et al.,

1975; Yoder, 1973). A toxinless race of C. heterostrophus (race 0) is

able to colonize maize plants and incite disease, although the effects

are less severe and this race exhibits little or no specificity towards

any maize cytoplasm (Hooker et al., 1970). This demonstrates that the

toxin from race T is not required for pathogenicity and serves as a

virulence factor. Sexual crossing studies between race T and race 0

have shown that race differences and presence of the toxin are linked

and segregate from other traits monogenetically (Lim and Hooker, 1971;

Yoder and Gracen, 1975).

There is some controversy over the origin of race T (rev. by Yoder,

1980). It was reportedly isolated from corn leaves in 1955 (Nelson et

al., 1970), yet a later report suggested this could be due to

contamination in the laboratory (Leonard, 1972). Race T isolates from

1970 were found to be predominately of mating type A, whereas race 0

isolates were 50% A and 50% a (Leonard, 1973). Several years later,

mating type A and a were equally distributed in both races (Leonard,

1977). This suggests that race T had a recent origin (1969) due to a

mutation in a single isolate of race 0 having mating type A. Additional

evidence supporting the recent origin of race T comes from studies that

compare its fitness with race 0. When race T was equally mixed with

race 0 and inoculated into maize plants with N cytoplasm, race T

declined dramatically over a two year period. The TOX 1 locus, or

closely linked genes, reduced pathogen fitness of race T compared to

race 0 on N cytoplasm (Klittich and Bronson, 1986).

The host selective toxins from C. heterostrophus, race T (T-toxin),

and P. mavdis (Pm-toxin) directly interfere with mitochondria isolated

from T cytoplasm (Miller and Koeppe, 1971; Comstock et al., 1973). The

toxins inhibit electron transport and stimulate NADH oxidation while

inhibiting malate oxidation (Miller and Koeppe, 1971; Peterson et al.,

1975). The molecular structures of T-toxin and Pm-toxin are very

similar and both are active at concentrations of 10-B to 10-9 M (Suzuki,

et al., 1983). T-toxin is a mixture of several linear polyketols

ranging from C-35 to C-45 in length, with C-39 and C-41 comprising

60-90% (Suzuki et al., 1983). Pm-toxin has 10-15 components, the four

major being linear C-33 and C-35 compounds with beta-ketol functional

groups (Danko et al., 1984). Although their chain length varies, the

toxins have nearly identical spacing of the four sets of oxygen

functions which may be important for binding to the inner mitochondrial

membrane (Danko et al., 1984). The length of T-toxin approximates the

length of a lipid bilayer and space filling models demonstrate that the

toxin molecules could potentially form a cylinder with a hydrophic core,

and could function as an ionophore (Payne et al., 1980). The ionophore

model is further supported by the ability of the T-toxin to form

channels in a planar bilayer membrane system, subsequently altering the

permeability to Ca2*(Holden et al., 1985). Increased membrane

permeablity to Ca2' with the addition of T-toxin was detected in

mitochondria isolated from T cytoplasm but not N cytoplasm plants

(Holden and Sze, 1984).

Plants with T cytoplasm also are sensitive to the insecticide

methomyl (Humaydan and Scott, 1977). Methomyl is a systemic carbamate

insecticide that affects T cytoplasm at concentrations of 10-3 M (Koeppe

et al., 1978). The structure of methomyl bears little resemblance to

T-toxin or Pm-toxin, yet has a similar mode of action on T mitochondria.

Methomyl has also been proposed to form hydrophillic pores in T

mitochondria inner membrane (Klein and Koeppe, 1985).

In an effort to identify binding sites in the inner mitochondrial

membrane, tritium-labelled T-toxin and Pm-toxin analogs were prepared

(Frantzen et al., 1987). The analogs had high biological activity and

specificities identical to the native toxins, but the assay was unable

to detect differences in binding characteristics between mitochondria

from N (normal) and T cytoplasms. The toxins were found to be

lipophillic and bound to mitochondrial membranes in a non-energy

dependent manner.

The molecular characterization of the genetic locus (loci)

involved with toxin sensitivity and cms associated with T cytoplasm has

been approached in various ways. The first approach characterized the

mitochondrial genomes of three cms and one normal (N) cytoplasms of

maize by restriction enzyme analysis (Pring and Levings, 1978).

Restriction fragment length polymorphisms were detected between

cytoplasms yet the extent of these polymorphisms and the apparent size

and complexity of the maize mitochondrial genome precluded

identification of differences that could easily be correlated to the

phenotypic differences between cytoplasms. Analysis of mitochondrial

protein-coding gene expression by labelling proteins in isolated

mitochondria detect far fewer differences between cytoplasms (Forde et

al., 1978; Forde and Leaver, 1980). A protein specific to T cytoplasm

was detected, and the abundance of this protein (13 kD) was reduced upon

fertility restoration (Forde and Leaver, 1980).

Two different approaches proved to be successful at identifying the

DNA region coding for the T-specific 13 kD protein. Dewey and coworkers

chose an approach that assumed the differences in protein expression

between N and T cytoplasm could be detected as transcript differences

(Dewey et al., 1986). They prepared mitochondrial DNA libraries (in the

vector pUC8) from N and T cytoplasms and hybridized these clones with

total radio-labelled mitochondrial RNA from the two cytoplasms. By

sequencing clones that uniquely hybridized to T mitochondrial RNAs they

identified a clone that contained two potential open reading frames

(ORFs). The first ORF (urfl3-T) was found to be unique to T cytoplasm

and could potentially encode a protein of approximately 13 kD in size.

Evidence supporting this ORFs relation to the T-specific 13kD protein

was that its transcripts were affected by restoration, apparently by a

differential processing event (Dewey et al., 1986).

The other approach taken to find the urfl3-T gene was to analyze

tissue-culture derived male fertile, toxin insensitive mutants (Fauron

et al., 1987; Rottman et al., 1987; Wise et al., 1987a). There are no

reports of natural, heritable revertants of T cytoplasm; however, male

fertile and toxin insensitive mutants have been obtained at relatively

high frequencies from plants regenerated from immature embryos in the

presence (Gengenbach et al., 1977; Brettell et al., 1980) or absence

(Brettell et al., 1980; Umbeck and Gengenbach, 1983) of T-toxin.

Restriction enzyme analysis of the fertile, toxin insensitive mutants

revealed a common rearrangement relative to the parental T male-sterile,

toxin sensitive mtDNA restriction pattern, as well as one mutant that

retained the paternal restriction pattern (T-4; Umbeck and Gengenbach,

1983). Sequence comparison of the area of rearrangement in T and T-4

revealed a five bp insertion and a G to A transition located internal to

the urfl3-T gene in the T-4 mutant (Wise et al., 1987a). The insert

places a stop codon in the reading frame which truncates the putative 13

kD protein. Confirmation that the urfl3-T gene codes for the T-specific

13 kD protein was made with antibodies raised to synthetic peptides and

subsequent immunoprecipitation from polypeptides labelled in isolated T

mitochondria (Wise et al., 1987b; Dewey et al., 1987). The 13 kD

protein was not detected in N cytoplasm or the T-4 mutant and was

reduced in abundance in T cytoplasms restored to fertility. Although

both Rfl and Rf2 are required for fertility restoration of T cytoplasm,

Rfl is solely responsible for the reduction of the 13 kD protein (Dewey

et al., 1987). The function of Rf2 in restoration is unclear, as is the

specific role the 13 kD protein plays in T-cms and toxin sensitivity.

Nuclear-Hitochondrial Interactions

Pollen fertility restoring genes compensate for

nuclear-mitochondrial deficiencies or incompatibilities that are

phenotypically expressed during microsporogenesis and

microgametogenesis. In the most thoroughly characterized cms system

(T-cytoplasm), one effect of restoration is a decrease in the abundance

of a specific mitochondrial gene product (13 kD protein: Forde and

Leaver, 1980; Dewey, et al., 1987), apparently by altered RNA processing

(Dewey et al., 1986). Fertility restoration also affects the expression

of specific mitochondrial proteins in certain cms systems of sorghum

(Dixon and Leaver, 1982; Bailey-Serres, et al., 1986). Nuclear gene

effects, apart from the effect of Rf genes, on expression of other

mitochondrial proteins have also been reported in sorghum

(Bailey-Serres, et al., 1986).

The influence of the nuclear genome on mitochondria is also evident

in the range of susceptibility of specific maize inbred lines to C.

heterostrophus, race T, (Watrud et al., 1975; Hallauer and Martinson,

1975; Payne and Yoder, 1978) and sensitivity to T-toxin (Payne and

Yoder, 1978). The Rf genes reduce the sensitivity of T-cms plants

(Watrud et al., 1975) and isolated mitochondria (Barrett and Flavell,

1975) to T-toxin. Reductions up to 50%, as measured by the inhibition

of malate oxidation in isolated mitochondria, have been detected in some

lines (S. J. Danko, J. M. Daly, B. G. Gengenbach, personal

communication); however, it is apparent that nuclear genes besides the

Rf genes may have an influence on this reaction. An evaluation of a

wide range of non-restoring genotypes indicated that both the level of

susceptibility to the pathogen and sensitivity to the toxin were

influenced by the nuclear genome (Payne and Yoder, 1978), although

disease susceptibility and toxin sensitivity did not correlate in every


Certain maize nuclear genomes have been shown to influence the rate

of reversion in the S-cms system of maize as well as the abundance of

the Sl and S2 linear DNA molecules (Laughnan et al., 1981). A

particular nuclear background, M825, was also shown to affect

mitochondrial DNA organization, unrelated to reversion (Escote et al.,

1986). Another maize inbred line, WF9, gives rise to a mitochondrial

mutation ncs (nonchromosomal stripe) at a frequency around 1% (Coe,

1983). The frequency of reversion to fertility in a cms system of

Phaseolus vulgaris is about 15% in one nuclear background and appears to

increase dramatically with the introduction of the Fr gene (fertility

restorer; Mackenzie and Basset, 1987).

There is a growing list of yeast and Neurospora nuclear genes that

regulate some aspect of mitochondrial gene expression. Although no

nuclear genes have yet been identified that directly regulate transcipt

initiation of mitochondrial genes, there are nuclear genes influencing

RNA processing (Dieckman et al., 1984), RNA splicing (Bonitz et al.,

1982; McGraw and Tzagoloff, 1983; Faye and Simon, 1983), translation

(Fox, 1986) and posttranslational modification (Pratje et al., 1983).

As many as six nuclear genes may be specifically involved in the

expression of certain yeast mitochondrial genes.

At this early stage of investigation, plant nuclear-mitochondrial

interactions involve nuclear influences on mitochondrial RNA processing,

genome rearrangement and replication. Considering the size and

complexity of plant mitochondrial genomes, it is likely that there are

at least an equivalent number of nuclear genes involved in mitochondrial

biogenesis and function as there are in other species. The major

direction of the research presented in this dissertation evaluates the


influence of the nuclear genome on maize mitochondrial transcription.

Specifically, it involves the characterization of transcription and

posttranscriptional processing of transcripts from a region of the maize

T cytoplasm mitochondrial genome associated with cms and toxin





There are four major cytoplasms of maize; N (normal) cytoplasm and

three male sterile cytoplasms, S (USDA), C (Charrua), and T (Texas).

The male sterile cytoplasms are formally distinguished by the genetics

of fertility restoration (Laughnan and Gabay-Laughnan, 1983), but also

can be characterized by mtDNA restriction patterns (Pring and Levings,

1978) and gel patterns of proteins labelled in isolated mitochondria

(Forde et al., 1978; Forde and Leaver, 1980). Additionally, plants

having the T-cms cytoplasm are highly susceptible to the fungal

pathogens Cochliobolus heterostrophus, race T, and Phyllosticta maydis

(Chapter I). Mitochondrial restriction analyses have detected

polymorphisms within three of the four types of cytoplasms (N, Levings

and Pring, 1977; McNay et al., 1983; C, Pring et al., 1987b; S, Sisco et

al., 1985); however, these polymorphisms apparently have little effect

on mitochondrially-coded polypeptides within these groups since no

heterogeneties have been reported among polypeptide patterns obtained

from labelling proteins in isolated mitochondria (Forde et al., 1980).

The polypeptide patterns of proteins labelled in isolated

mitochondria have identified a number of polypeptides ranging from 42-88

kD that are unique to S cytoplasm, a 17.5 kD polypeptide unique to C

cytoplasm, and a 13 kD polypeptide unique to T cytoplasm (Forde and

Leaver, 1980). The S-specific and C-specific polypeptides are

unaffected by fertility restoration, whereas the T-specific 13 kD

protein is reduced approximately 67% upon restoration (Forde and Leaver,

1980; Dixon et al., 1982). This reduction is controlled by the nuclear

restorer gene Rfl (Dewey et al., 1987), apparently through differential

RNA processing of its transcripts (Dewey et al., 1986). Differences in

the abundance of specific polypeptides have been noted in different

maize nuclear backgrounds, suggesting the involvement of other nuclear

genes on mitochondrial gene expression (Forde et al., 1978).

Developmental and tissue specific mitochondrial polypetides have also

been reported in maize, indicating that some mitochondrial genes may be

differentially regulated (Newton and Walbot, 1985).

If the mtDNA regions coding for the cytoplasm-specific polypeptides

are associated with regions of mtDNA polymorphism between cytoplasms,

several approaches are available for the isolation of these gene

regions. Transcriptional analyses may be the most appropriate method

since detection of RNA differences is often a direct way of identifying

regions involved in differential gene expression and can at least

simplify analysis of DNA heterogenieties by excluding regions of DNA

polymorphism occurring in non-coding regions. Most methods employed for

the isolation of RNAs unique to a particular developmental stage, tissue

or mutation, involve differential hybridization of RNA to either DNA or

cDNA libraries. For example, the technique of cascade hybridization

developed by Timberlake (1980) for identification of developmentally

specific cDNAs of Aspergillus involved several rounds of hybridation of

excess RNA from sporulating cultures to cDNAs of vegetative cultures,

and vice versa. Nonhybridized cDNAs, unique to either vegetative or

sporulating cultures, were separated from RNA:cDNA hybrids by

hydroxyapatite columns. Another techinque, differential plaque

hybridization (St. John and Davis, 1979), utilizes multiple

nitrocellulose filter replicas of plaque plates produced from lambda

libraries as substrates for cDNA probes. Differential hybridization of

the probes to the plaques is detected by comparison of the resulting

x-ray film exposures.

The relatively small size of the maize mitochondrial genome

(approximately 570 kb) simplifies evaluation of mtRNA differences

between cytoplasms, permitting the application of simpler approaches. A

technique that is a modification of the method used to identify the

ribosomal RNA gene regions of the maize chloroplast (Bedbrook and

Bogorad, 1976) is described in this chapter. This technique involves

the labelling of RNA from different mitochondria and hybridization to

nitrocellulose filters containing endonuclease restricted mtDNA.

Identification of mtDNA regions associated with differential

transcription between N cytoplasm, T-cms cytoplasm and a disease

resistant, male fertile mutant of this cytoplasm, as well as differences

between cell suspension cultures and coleoptile tissue, are presented.

Materials and Methods

Seed Lines

Maize inbred lines A188(N), A188(T) and tissue culture mutant lines

A188(T-4) and A188(T-7) were provided by Dr. Burle Gengenbach,

University of Minnesota. The restored line WF9(T) (RflRfl;Rf2Rf2) was

provided by Dr. Marc Albertson, Pioneer Hi-Bred International Inc. The

Black Mexican inbred line and cell suspension cultures were kindly

provided by Dr. Prem Chourey, USDA-ARS, University of Florida.

Preparation of mtRNA and mtDNA

The preparation of mtRNA was performed as described (Wise et al.,

1987a). The preparation, restriction endonuclease digestion,

electrophoresis and blotting of mtDNA was performed as described (McNay

et al., 1983).

End labelling of mtRNA

End labelling involves the transfer of the gamma phosphate of

gamma-32P dATP to the 5' hydroxyl of a nucleic acid strand, catalyzed by

the enzyme polynucleotide kinase. Hydrolysis of RNA to generate 5'

hydroxyl ends was accomplished by mild alkali treatment. Approximately

1 ug of RNA (1 ug/ul) was incubated with glycine hydrolysis buffer (5 mM

glycine, 100 uM spermidine, 10 um NazEDTA, pH 9.5) in a reaction volume

of 10 ul for 10 min at 900C. Kinase labelling was performed in 15 ul

reaction volume containing 50 mM Tris-HC1, pH 9.5, 10 mM MgCl2, 5 mM

dithiothreitol, 20 to 30 uCi of gamma-32P dATP (7000 Ci/uM; New England

Nuclear) and 4 to 5 units of T4 polynucleotide kinase. The reaction was

incubated at 37C for 30-45 min and run on a Sephadex G-50 column to

separate the unincorporated nucleotides from the labelled RNA.

The Southern blots were prehybridized in 4X SSC, with 100 ug/ml

yeast RNA and 0.1% SDS at 650C for 2 to 3 hours. The labelled RNA was

heated briefly at 700C and hybridized at 650C for 16 hours. The

nitrocellulose membranes were washed two times in 2X SSC, 0.1% SDS for

15 minutes at room temperature then two more times at 0.25X SSC, 0.1%

SDS for 15 min at room temperature. Membranes were dried and exposed to

x-ray film (Kodak).

Results and Discussion

The DNA region coding for the maize chloroplast rRNA genes were

initially identified through isolation and end labelling of the rRNAs

and subsequent probing of Southern blots containing restricted maize

chloroplast DNA (Bedbrook and Bogorad, 1976). In the procedure

described here, total mtRNAs from different cytoplasms were end labelled

and hybridized to identical Southern blots containing restricted mtDNA

from different cytoplasms. All blots contained mtDNAs from the

cytoplasms used as probes, to ensure a DNA substrate was available for

all RNAs, and these mtDNAs were usually digested with a minimum of three

restriction enzymes.

A representative hybridization comparing mtRNA from A188(N)

cytoplasm with that of A188(T-7), a fertile, disease resistant mutant

that was generated through tissue culture (Umbeck and Gengenbach, 1983)

is shown in Figure 2-1. The differences in hybridization patterns

obtained with the two RNA species were most easily visualized in the

lanes containing N and T mtDNA (Fig. 2-1). Differences in BamHI

restriction patterns between N and T mt DNAs are detected in both the

ethidium bromide stained lanes (panel A) and in the hybridization

patterns (panels B-E). The hybridized DNA fragments that are unique to

either the N or T lanes within each panel represent regions of DNA

polymorphism between these mitochondrial genomes that are actively

transcribed (as detected with N and T-7 mtRNAs).

The differences between N and T-7 mtRNAs are detected by comparing

the hybridization pattern of N mtRNA to N and T mtDNA lanes (panel B)

with the hybridization pattern of T-7 mtRNA to N and T mtDNAs (panel C).

The intensity of hybridization to a 3.0 kb BamHI fragment in T mtDNA

RNA: N T7 N T7



Fig. 2-1. Comparison of mtRNAs from N and T7 mitochondrial genomes.
Panel A, ethidium bromide stained agarose gel electrophoresis
pattern of BamHI restricted mt DNAs from N and T cytoplasms
(lanes N and T, respectively) and lambda DNA separately
digested with BamHI and PstI and combined for a marker lane
(M). Panels B and C, identical (sandwich) blots of DNAs
shown in panel A, hybridized with end-labelled mtRNA from N
cytoplasm (panel B) or mtRNA from the fertile, toxin
insensitive revertant mutant T7 (panel C). Panels D and E,
longer exposures of panels B and C, respectively. Black
arrows identify mtDNA bands that show stronger hybridization
intensities with T7 mtRNA compared to N mtRNA. Open arrow
identifies a mtDNA band detected in both N and T mtDNAs that
shows a stronger hybridization intensity with N mtRNA
relative to T7 mtRNA. Lambda markers are labelled in

lane is greater in panel C (dark arrow) than in panel B, indicating a

greater abundance of the RNA fragments that hybridize to this fragment

in T-7 mtRNA compared to N mtRNA. A longer exposure also detects at

least two other bands (dark arrows) that are more intense with T-7 mtRNA

hybridization pattern compared to N mtRNA. All three of these BamHl

fragments appear to be specific to T mtDNA, thus the increased

hybridization intensity may represent a DNA polymorphism between N and T

mt genomes, resulting in non-identical transcripts which unequally

hybridized to this specific DNA fragment, rather than being a reflection

of differential transcription between N and T-7 mitochondria. Another

fragment displayed greater hybridization intensity when hybridized with

N mtRNA compared to T-7 mtRNA (light arrow). This effect was seen in

both N and T mtDNA lanes suggesting it may involve a similar fragment.

The simplest explanation for this effect is that the transcripts

hybridizing to this fragment are more abundant in N mtRNAs and may

represent a difference in rate of transcription or in transcript

stability between cytoplasms.

This technique merely implicates mtDNA regions that may be

differentially transcribed between mt genomes, under these experimental

conditions. Subsequent isolation of the DNA fragments and use in

Northern hybridizations would be necessary to confirm that

transcriptional differences exist. The best application of this

technique may be its ability to find or rule out gross transcriptional

differences between RNAs from two mt genomes. For example, comparison

of mtRNAs isolated from A188(N) and WF9(N) cytoplasms showed that the

A188(N) mtRNA hybridized to a 1.3 kb HindIII fragment from T mtDNA,

while and WF9(N) did not (data not shown). The absence of hybridization

to this band was later found to be related to a 3.5 kb deletion in the

WF9(N) mt genome (see Chapter 3). The technique described in this

chapter also was used to compare mtRNAs from T male-sterile cytoplasm

with the T-4 male fertile, disease resistant mutant. No differences

were detected (data not shown), and further analyses demonstrated that

the only difference detected between these mitochondrial genomes was a

five bp insertion and a G to A transition in the urfl3-T gene, which

truncates the polypeptide but does not affect transcription (Wise et

al., 1987a; Wise et al., 1987b).

It is also noteworthy that the isolation of the urfl3-T gene

involved a technique similar to the differential plaque hybridization;

Dewey et al. (1986) labelled replica nitrocellulose membranes of pUC8

libraries of N and T mt genomes with N and T mt RNA and examined the

clones that were differentially labelled.

Besides evaluating differences between maize cytoplasms, this

differential RNA hybridization technique was used to detect differences

in mitochondrial transcripts from cell suspension cultures and

coleoptile tissue. A representative blot is shown in Figure 2-2.

Mitochondrial DNA isolated from cell suspension cultures was digested

with BamHI, HindIII, and XhoI restriction enzymes, and identical

Southern blots were hybridized with end labelled RNA isolated from

log-phase Black Mexican cell suspension or with Black Mexican coleoptile

mtRNA. Dramatic differences in hybridization patterns were detected

(panel B vs panel C; panels D and E are long exposures of B and C,

respectively; panels Fl and F2 are from another experiment that used

T-cytoplasm mtDNA). The intense bands that are unique to coleoptile

INA: CIll Sit. Celeep.
DNA: A CA ll Ses. CI Cell Sol. Ct Clil Sis. Ct
Euzyme:B/P B H X B H X B 8 H X I

7.2 I


Coll SeS. Celeop. C.S. CIl.
Cell Sis. Coll Sis. T T

D E F1 F2

Fig. 2-2. Comparison of mtRNAs from cell suspension cultures and
coleoptile tissues. Panel A, ethidium bromide stained
agarose gel with electrophoresis pattern of BamHI, HindIII,
and XhoI restricted cell suspension mt DNA; BamHI restricted
maize chloroplast DNA (ct); and lambda DNA restricted
separately with BamHI and PstI and combined. Panels B and C,
identical blots of DNAs shown in panel A hybridized with
end-labelled mtRNA from log-phase cell suspension cultures
(panel B) and 6 day old coleoptile tissue (panel C). Panels
D and E, longer exposures of blots shown in panels B and C,
respectively. Panel F, BamHI digested mtDNA from T cytoplasm
hybridized with cell suspension (Fl) or coleoptile (F2)
mtRNA. Arrowheads identify chloroplast DNA fragments
contaminating the cell suspension mtDNA, hybridizing to
chloroplast RNA from the respective mtRNAs. Open arrows
distinguish mtDNA fragments that are uniquely or more
intensely hybridized with cell suspension mtRNA.


RNA, are primarily from chloroplast contamination. Maize chloroplast

DNA, digested with BamHI, is shown in panels A, B, and C, and the BamHI

"mitochondrial" DNA fragments that comigrate with the chloroplast

hybridized fragments are marked (arrowheads). The cell suspension mtDNA

is contaminated with chloroplast DNA (ct DNA), yet the cell suspension

mtRNA do not appear to be as contaminated with ctRNA as the coleoptile

mtRNA. The most striking result of this analysis is the increase in the

number of DNA bands that hybridize to the cell suspension RNA (panels E

and Fl). Whether this reflects a general increase in transcriptional

level of "minor" mitochondrial transcripts or transcripts from regions

that are not normally transcribed in coleoptile mitochondria remains to

be answered.




The mitochondrial genome of the normal, male-fertile (N) cytoplasm

of maize (Zea mays L.) is estimated to be 570 kb (Lonsdale et al.,

1984). Approximately 42 kb of the maize mitochondrial genome is

represented as six repeated regions that range in size from 1-14 kb;

five of these repeats are involved in intramolecular recombination

(Lonsdale et al., 1984). Recombination through these repeats generates

subgenomic circles from the master circle. The capability for

intramolecular recombination apparently has contributed to the diversity

of cytoplasms in maize, as mtDNA polymorphisms are observed among (Pring

and Levings, 1978) and within (Levings and Pring 1977; McNay et al.,

1983; Pring et al., 1987b; Sisco et al., 1985) groups of cytoplasms.

Recombination events have resulted in new transcriptional units as

inferred from the organization of the gene urfl3-T (Dewey et al., 1986;

Wise et al., 1987a, 1987b). This 345 bp gene is unique to the T male

sterile cytoplasm of maize yet contains 263 bp 87% homologous to a

region 3' to the mitochondrial 26S rDNA, and 58 bp of near perfect

homology to a region interior to 26S rDNA (Dewey et al., 1986). Another

recombination event duplicated a 5 kb DNA region 5' to the gene atp6

resulting in a second configuration located 69 bp 5' to the predicted

translational start sequence of urfl3-T (Dewey et al., 1986; Wise et al.

1987a). Sequences within this repeated region apparently promote

transcription of atp6 and urfl3-T, as well as another open reading

frame, ORF25, located 80 nucleotides 3' to urfl3-T (Dewey et al., 1986).

There are three major sources of cytoplasmic male sterility (cms)

in maize (T, C, and S), each distinguished by the ability of specific

nuclear genes to restore pollen fertility (Laughnan and Gabay-Laughnan,

1983). Fertility restoration of T cytoplasm plants is conditioned by

the dominant nuclear genes Rfl and Rf2. Maize plants carrying the T

cytoplasm are highly susceptible to the fungal pathogens Cochliobolus

heterostrophus Drechsler (Bipolaris maydis) race T, the causal agent of

Southern corn leaf blight, and Phyllosticta maydis Arny and Nelson, the

causal agent of yellow corn leaf blight (Ullstrup, 1972). Both

pathogens produce a series of toxic Beta polyketols which are virulence

determinants (Suzuki et al., 1983; Danko et al., 1984) and specifically

affect T cytoplasm mitochondria (Miller and Koeppe, 1971; Comstock et

al., 1973; Yoder, 1973).

The gene urfl3-T is strongly implicated in playing a role in the

maternally inherited phenotypes of male sterility and toxin sensitivity

in T cytoplasm since it is unique to this cytoplasm and is altered or

deleted in tissue culture derived male fertile and toxin insensitive

mutants (Wise et al., 1987a; Rottman et al., 1987). The gene encodes a

polypeptide of approximately 13 kD (Wise et al., 1987b), the synthesis

of which is reduced in plants restored to fertility (Forde and Leaver,

1980). Nuclear restorer genes appear to function by differentially

processing major urfl3-T transcripts (Dewey et al., 1986).

There is substantial evidence of nuclear background effects on

reaction of maize lines and isolated mitochondria to the fungal toxins.

Mitochondria isolated from different lines in T cytoplasm exhibit

variation in sensitivity to the toxins of C. heterostrophus race T, and

mitochondria isolated from restored T cytoplasm plants are less

sensitive than mitochondria from nonrestored T plants (Barratt and

Flavell, 1975). In another study, sensitivity of malate oxidation to

purified toxins was reduced 50% by fertility restoration in many cases,

and an effect of nuclear background also was evident (S. J. Danko, J. M.

Daly, and B. G. Gengenbach, personal communication). An influence of

nuclear background on susceptibility to race T of C. heterostrophus and

the toxins was demonstrated among 24 lines and single-cross hybrids

(Payne and Yoder, 1978). The lines displayed a range of sensitivity to

the pathogen and to toxins, but disease susceptibility and toxin

sensitivity rankings were not correlated in every comparison.

This study examined the influence of nuclear background on

transcription of urfl3-T and RNA processing events associated with

fertility restoration. Transcripts of atp6, urfl3-T, and ORF25 from N,

T, and T-restored cytoplasms in five inbred lines were mapped with a

series of contiguous DNA clones. DNA sequence and atp6 transcriptional

alterations were detected between N and T cytoplasms. Nuclear

background influenced the abundance of urfl3-T and ORF25 transcripts, as

well as processing of transcripts in a region within ORF25.

Materials and Methods

Seed lines

The maize inbred lines A188(T) (rflrfl;Rf2Rf2), A188(N)

(rfl--;rf2--) and restored WF9(T) (RflRfl;Rf2Rf2) were provided by Dr.

B. Gengenbach. Vf9(T) (rflrfl;rf2rf2) was provided by Pioneer Hi-Bred

International, Inc. The lines A632(T) (rflrfl;----), restored A632(T)

(Rfl--;Rf2--), A632(N) (rfl--;rf2--), W64A(T) (rflrfl;----), restored

W64A(T) (Rfl--;Rf2--), W64A(N) (rfl--;rf2--), C103(T) (rflrfl;----),

restored C103(T) (Rflrfl;Rf2--) and C103(N) (rfl--;rf2--) were provided

by Dr. P. S. Chourey. Other lines used were WF9(N) (rfl--;rf2--) and

the single cross hybrid A188(T) X restored WF9(T).

Preparation and analysis of mtDNA and mtRNA

Isolation of mitochondria, purification of nucleic acids,

endonuclease digestion, electrophoresis, blotting, and hybridization

were performed as previously described (Wise et al., 1987a). The DNA

probes used in Northern hybridizations were obtained from pUC8 clones

subjected to preparatory digestion. The inserts were recovered from

agarose gels with DEAE NA45 paper (Schleicher & Schuell) and ligated

overnight prior to nick translation. If nitrocellulose filters were

reprobed, the previous probe was washed off with boiling 20 mM Tris, pH

8.0 for 20 min.

DNA sequencing:

The dideoxynucleotide chain termination method of Sanger et al.

(1977) was used with the universal primer for clones in the M13 vectors

mpl8 and mpl9 (Yanisch-Perron et al., 1985). The reactions were

radiolabeled with S35 dATP (New England Nuclear) and separated on 6%

denaturing acrylamide wedge gels.


Transcription of atp6

The 5' region of the maize mitochondrial gene atp 6 is repeated in

T cytoplasm from the positions -444 relative to the proposed translation

initiation site (Dewey et al., 1985a) to over -5000 bp (Wise et al.,

1987a; Fig. 3-1). A complex transcript pattern was observed on Northern

blots of T cytoplasm mtRNA with probes from the repeated region, as they

hybridized to transcripts of atp6, and the cotranscribed genes urfl3-T

and ORF25. The pattern observed with mtRNA from normal (N) cytoplasms

was less complex as the probes detected only atp6 transcripts. To

distinguish atp6 transcripts from urfl3-T/ORF25 transcripts in T

cytoplasm, probes containing sequences beginning 22 bp from the 3' edge

of the 5 kb repeat and extending through the atp6 coding region (Fig.

3-1; T-t2, T-tl) were hybridized to RNA from N, T, and T cytoplasm

plants restored to fertility (R) (Fig. 3-2). Major transcripts of 1.8

and 1.6 kb and minor transcripts of 4.1 and 3.6 kb were detected with

the two probes; prolonged exposure also revealed a transcript of 5.8 kb.

The exposure level of the N lanes in Fig. 3-2 was longer than the T or R

lanes to illustrate a transcript difference between N and T or R mtRNAs,

as there appeared to be a transcript migrating at 1.55 kb that was only

found in T and R mtRNAs. Since this transcript could not clearly be

separated from the 1.6 kb transcript detected in N, T and R mtRNAs, it

may reflect a copy number difference of the 1.6 kb transcript rather

than a unique 1.55 kb transcript.

Mitochondrial RNA of N, T and R cytoplasms from five different

nuclear backgrounds were analyzed with probes T-t2 and T-tl, and the

difference between N and T (and R) mtRNAs was consistent among these

particular inbreds. The difference may reflect differential transcript

initiation or processing events between N and T cytoplasms.

Additionally, fertility restoration did not alter atp6 transcription in

T cytoplasm.

Fig. 3-1. Restriction map of the mitochondrial region containing the
genes atp6, urfl3-T and ORF25. Relative position the genes to
the 5 kb repeat in T cytoplasm of maize is shown with solid
vertical line representing the 3' edge of the repeat. Dashed
lines mark the boundaries of three probes used on Northern
blots. X, XhoI; H, HindIII; t, TaqI.

Fig. 3-2. Northern blot analysis of atp6 transcripts. Position of the
clones (T-tl, T-t2) used as probes is designated in Fig. 3-1.
Mitochondrial RNA is from normal (N), Texas male sterile (T),
and nuclear restored Texas male sterile (R) cytoplasms.
Approximate size of transcripts are in kilobases.

tt t t t
v I I

atp 6

*4-5 kb repeat-

X H tt t t t tH H
I I tt I II



0 500bp



J4.1 -

U=1.6- Pma


Differences between N and T mtDNA

A difference between N and T mtDNA was detected in the 5 kb repeat

at positions -572 to -576 and -587 to -590 relative to the proposed

translational start site of atp6. DNA sequence analysis of a TaqI clone

(-466 to -689) revealed that N mtDNAs did not contain 9 bp that were

present in T mtDNAs; the remaining sequences of the 223 bp clone were

identical. The T mtDNA has insertions of 4 and 5 bp in the region shown




The 5 bp insert disrupts an Alul recognition site, thus providing a

diagnostic assay for other cytoplasms. Hybridization of the TaqI clone

to Alul digests of maize mtDNAs revealed that seven N cytoplasms and C

and S cytoplasms do not carry this insert, while it was present in three

T cytoplasms tested.

Transcription of urfl3-T and ORF25

Transcripts of urfl3-T and ORF25 from T and R mitochondria were

mapped using 20 DNA probes to "Northern walk" a region 2 kb 5' to

urfl3-T to over 3 kb 3' to ORF25 (Fig. 3-1). Figure 3-3 shows the

location of selected probes and corresponding Northern blots used in

characterizing the complex transcriptional pattern found in this area of

the T mitochondrial genome. The largest transcript detected was

approximately 3.9 kb (detected by all six probes shown in the upper

panel of Figure 3-3). This transcript extends from a point within a 700

bp Xhol fragment in the 5 kb repeated region (Wise et al., 1987a) to a

point some 200 bp 3' to ORF25 (data not shown). Minor transcripts of


bN N D
S6 C 0 .0 0
SPC , O -p r 10
H CU 0 -
a) 0 -- m 0
_C 0- 0-
O -0 H w E 0 t II -
0.0 r--ao '000
L 44-' C) .1-, N II 0a

*r o r-4 a ) -a u0
c- j000 >-,o 0 -

0- O4-'. 0 00 000.

caH 0 k
i-H I C) WC) a)

10'01CC C00 W0

NdP 0CO) C) 0 a ,C)a
C) p -l a) Id C)

Id .' 0 1 0 O 00 -0 -.

C14 E -01 0) -CO r
i OH -H- -- 1- CO

"= V, N o 6> O X IMI
Hd -C o l r l-iY

Hi nj< -0 -- -P O
0 H 0 eT 41-i e 1- 0 C )

'0 20 m wI -- -> 0 OH r 0 > z
O 0 0 hC Y d E-- H O 101 0

C C7t'd ^ 0 o 1 -'- Ho 5
i tl p ri n41 w PC
-4 ) 0 1 0 -H 00 -r co
Q) N O C)O a, O -I H
* H r 1-4 1- 01 s i

o H- ( ) 0 C
1-11-C) -C) --a-0 IO dO

0, ). ( H- ,0 0 Oi 1 -
S00 01* '. r) C I7 \-Th<
co 'o 0 d0 c a, 0 C) > -A
-h o- C 0arC -4 4E 4 m G

; O = -, p -a O oH 0 C I c
0 P -P -tI- a, 0 *-H U
LO. .;i^ U O (>^




CC I" -






S a




4~ 1-
S- I-





I I /


: ,



4- 0-

I -
_ o


4- m



nC =u


3.0 and 2.3 kb also were detected with the 700 bp Xhol probe and

downstream probes. The 3.0 kb transcript terminates within the first

200 bp of ORF25 (not shown) and the 2.3 kb transcript terminates within

the region represented by the probe T-a2 (positions -505 to -405

relative to the putative translational start of urfl3-T). Probe T-a2

also detected transcripts of 2.1 and 1.9 kb in N mtRNAs.

Probes T-a65 and T-t220 (Fig. 3-3; upper panel) hybridized to the

major transcripts of atp6 and urfl3-T/ORF25. There were two sets of

comigrating transcripts (1.8 and 1.6 kb) that only could be

distinguished from one another with gene-specific probes that do not

contain sequences from the 5 kb repeat. Probe T-a65 (representing

positions -404 to -202) hybridized to major transcripts of 1.8 kb in N,

T, and R, and 2.0 kb only in T and R mtRNAs. Minor transcripts in N

were 4.1 and 3.6 kb, while 3.9 and 3.0 kb transcripts were evident in T

and R mtRNAs. The 1.8 kb transcript is an atp6 transcript (Fig. 3-2)

and urfl3-T-specific probes (Fig. 3-3; T-sal, T-atl) indicated that the

2.0 kb transcript is an urfl3-T/ORF25 transcript and consequently not

present in N mtRNAs. It is likely that the transcript initiation site

(or the precursor RNA processing site) for the 1.8 kb atp6 transcript

also serves to initiate urfl3-T/ORF25 transcription. The difference in

transcript abundance between the 2.0 and 1.8 kb transcripts also was

detected with T-a65.

Probe T-t220 hybridized to major transcripts of 1.8 and 1.6 kb in N

mtRNA (the 1.6 kb transcript was detected with longer exposures) and

2.0, 1.8 and 1.6 kb in T and R mtRNAs (Fig. 3-3, upper panel). The

atp6-specific probes (Figs. 3-1 and 3-2) distinguished the 1.8 and 1.6

kb transcripts as atp6 transcripts. Probes specific to urfl3-T (T-sal,

T-atl) hybridized to a 1.8 kb transcript in T and R mtRNAs. The shape

and migration of this band (arrowhead, T-sal), in comparison with the

1.8 kb transcript detected with T-t220, suggest it represents two

transcripts, of 1.8 and 1.85 kb. If the 1.6 kb atp6 transcript arose by

the processing of the 1.8 kb transcript, the same processing event most

likely would create a 1.8 kb transcript from the 2.0 kb transcript in T

and R mtRNAs. If this scenario is correct, the 1.8 kb transcript

detected in T mtRNA with probe T-t220 represent atp6 and urfl3-T/ORF25


There was a noticable effect of fertility restoration on the 1.6 kb

transcript detected with probe T-t220, resulting in a higher abundance

in R mtRNAs than in T mtRNAs. This subtle difference was detected in

mtRNAs from five different nuclear backgrounds (Fig 3-3; T-t220, lower

panel). The effect of restoration detected with T-t220 was limited to

the region representing positions -202 to -92.

A more obvious effect of restoration (Dewey et al., 1986) was

detected with probes representing map positions +3 relative to urfl3-T

and extending through ORF25. Probes T-sl (+3 to +32; not shown), T-sal

(+31 to +200; Fig. 3-3) and all downstream probes hybridized to a 1.6 kb

transcript in R mtRNA but not T mtRNA. We have designated this

transcript as 1.6'. Another transcript of 1.5 kb was detected with

probes T-sal (with a longer exposure) and T-atl (Fig. 3-3), and other

downstream probes, in both T and R mtRNA. The 1.6' kb transcript was

unique to R mtRNA whereas the 1.5 kb was observed in both T and R


A few weakly hybridizing transcripts were detected in N mtRNA with

T-sal, an urfl3-T-specific probe, which may represent transcripts 3' to

the 26S rDNA gene. It should be noted that the region of urfl3-T that

is homologous to the coding region of 26S rDNA (Dewey et al., 1986) was

not used in any of the Northern hybridizations shown in Figure 3-3.

All ORF25-specific probes and those extending some 200 nucleotides

3' to this gene gave the same pattern on Northern blots of T and R

mtRNA, with regard to the 2.0, 1.85/1.8, 1.6', and 1.5 kb transcripts

(Fig. 3-3; T-al06), as urfl3-T-specific probes. Minor transcripts of

1.0 and 0.8 kb (Fig. 3-3; T-st308), which were first detected with T-a65

and T-t220, respectively, do not extend through ORF25, terminating

within the first 350 nucleotides of the 663 bp ORF. A minor transcript

of 0.6 kb was detected only with urfl3-T-specific probes in R mtRNA

(Fig. 3-3; T-st308). This transcript was first detected with a probe

representing positions of +3 to +32 of urfl3-T (not shown) and

terminated in approximately the same region as the 1.0 and 0.8 kb

transcripts. All restored lines were characterized by the appearance of

the 1.6' kb and 0.6 (upon longer exposure) kb transcripts with the probe

T-st308. An illustration of these transcripts, with putative start and

stop sites, is shown in Figure 3-4. For all transcripts described,

except the 1.85 kb transcript, the predicted size was in close agreement

with their mapped position on the restriction map.

The relationship between the 1.6 kb transcript detected within the

repeat and the 1.6' kb transcript unique to R mtRNA is delineated in the

lower panel of Figure 3-3. Probe T-t220 detected the 1.6 kb atp6

transcript in both T and R mtRNAs as well as what appeared to be a more

abundant 1.6 kb transcript in R mtRNA. If T-t220 detected a higher copy

0 300bp
a t a t t

1.81 ---
1.61-B--- -

a t a t as a t a a a H a

6 URF13-T ORF25

3.- - - - - - -
4------------------------------- -----------------------

2. 0-- -

1.6'. 8

1.0 --- -
l ot------------

- - I --- ,

0.8- - - - - -I

0.6'- -- - - - --I

0.4 -4

Fig. 3-4. Schematic representation of the major and minor transcripts
of the maize mitochondrial genes atp6, urfl3-T and ORF25.
Start and stops are approximated and marked with vertical
lines. Approximate sizes are in kilobases. Transcripts only
detected in nuclear-restored T mtRNA are primed.

number of the 1.6 kb atp6 transcript in R mtRNA relative to T mtRNA, the

atp6-specific probes shown in Fig. 3-2 should also detect this

difference. This was not found in any of the lines analyzed. This

suggests that T-t220 may have detected the restorer-specific 1.6' kb

transcript. Probe T-ts305 (positions -91 to +2) did not hybridize to

the 1.6 kb transcript. Assuming that the 1.6 kb transcript detected

with T-t220 is related to the restorer specific 1.6' kb transcript, this

observation suggests that T-ts305 may be an intron-specific probe and

the 1.6 kb transcript that appears with restoration is the result of an

RNA splicing event.

Nuclear effects on urfl3-T and ORF25 transcripts

The 2.0, 1.85, and 1.8 kb transcripts appear to be the major mature

transcripts of urfl3-T due to their abundance and mapped position. The

relative abundance of these transcripts was found to be affected by

nuclear background with the most noticeable difference detected in the

intensity of the 2.0 kb transcript. This transcript was in greater

abundance in A188(T) mtRNA (Fig. 3-5 a) than in four other inbreds

examined (Fig. 3-5 b). The lower abundance of the 2.0 kb transcript in

the latter inbreds was associated with increased abundance of the

1.8/1.85 kb transcripts.

Three different transcript patterns were detected with

ORF25-specific probes among N cytoplasms of the five lines surveyed.

Figure 3-6 shows a representative of each characteristic pattern

obtained with probe T-al06 (Fig. 3-3 defines the probe location). The

line Wf9(N) displayed transcripts of 3.5, 1.8, and 1.6 kb (Fig. 3-6,

left, a), while A188(N) exhibited transcripts of 3.1, 1.7, and 1.3 kb

Fig. 3-5. Nuclear background effects on urfl3-T/ORF25 transcripts.
A188(T) mtRNA (a) has a greater abundance of the 2.0 kb
transcript relative to the other nuclear genotypes examined
[WF9(T) shown, b] as well as decreased abundance of the 1.8
(1.85) kb transcript. Location of probe T-st308 shown in
Figure 3-3. Transcript sizes in kb.

Fig. 3-6. Effects of nuclear background and cytoplasm on transciption
patterns of ORF25. Left, mitochondrial RNA from Wf9(N) (a),
A188(N) (b), and W64A(N) (c). Right, mitochondrial RNA from
sterile (T) or restored (R) lines Wf9(T) and A188(T); the
restored version of A188(T) is a single cross hybrid of
A188(T) X Wf9(T)RFlRFlRF2RF2. Probes are ORF25-specific
T-al06 (Fig. 3-3) or T-H41 (Fig. 3-1); the two probes
resulted in indistinguishable transcript patterns.
Approximate transcript sizes are indicated in kb.


JI 2.0

a b






n 2.2
hq 1.8
aI ,

a b

1.1 -

a b c

(Fig. 3-6, right, b). The pattern detected in W64A(N) mtRNA (Fig. 3-6,

right, c) was identical to the patterns found in A632(N) and C103(N)

mtRNAs, and had the characteristic Wf9(N) pattern with the addition of a

major transcript of 2.2 kb. The difference in the transcript patterns

between WF9(N) and A188(N) mtRNAs may be attributed to DNA differences

that were detected 3' to ORF25. The WF9(N) mitochondrial genome lacks a

region of over 3.5 kb that is located 3' to ORF25 in A188(N), WF9(T) and

A188(T) mtDNA genomes. Hybridization of HindIII digests of A188(N),

Wf9(N), and A188(T) (Fig. 3-7 a, b, c) mtDNAs with three probes located

3' to ORF25 revealed the extent of the deletion. The probes represented

regions of +1340 to +2330 (T-a22), and approximately, +2510 to +2910

(T-H16), and +2911 to +4343 (T-H15), relative to the initiation codon of

urfl3-T. The probes readily hybridized to their analogous fragments in

A188(N) and A188(T), but did not detect these fragments in Wf9(N) mtDNA.

Minor homology to all three probes was detected in A188(N) and A188(T)

mtDNAs, with a fragment of 3.7 kb evident with T-a22. The next

downstream probe contiguous with T-H15 hybridized to WF9(N) mtDNA,

delineating the extent of this deletion. Transcripts that are initiated

5' to or within ORF25 are terminated in different regions in WF9(N) and

A188(N) cytoplasms which likely explains the difference between A188(N)

and WF9(N) transcript size detected with ORF25-specific probes. The

abundance of ORF25 transcripts varies greatly among N cytoplasms shown

in Figure 3-6, as well as between N and T cytoplasms, perhaps as a

consequence of genomic organization.

The detection of a 1.1 kb transcript with ORF25 probes (T-al06 in

Fig. 3-6) in T mtRNAs was associated with nuclear background. Of five

T-a22 T-H16 T-H15


-b 4 W 1.43

S 0.40
a b c a bc a b c a b c

Fig. 3-7. Deletion in the WF9 mt genome. Ethidium bromide stained gel
of A188(N) (a), WF9(N) (b), and A188(T) (c) mtDNAs digested
with HindIII and three southern blots. Probes are from
regions 3' to ORF25 and represent the approximate positions
+1340 to +2330 (T-a22), +2510 to +2910 (T-H16), and +2511 to
+4343 (T-H15) relative to urfl3-T. Size of the hybridizing
DNA fragments are approximated in kb.

lines examined, restored and sterile W64A(T) and WF9(T) mtRNAs contained

the 1.1 kb transcript (Fig. 3-6, WF9 shown; Fig. 3-3 defines location of

probe T-al06), while sterile C103(T), A632(T) and A188(T) (Fig. 3-6,

A188 shown) did not. The single cross hybrid A188(T) X restored WF9(T)

showed the appearance of the 1.1 kb transcript (Fig. 3-6). The event

that gives rise to this transcript apparently is determined by the WF9

nuclear genome, and is distinct from the combined effect of the restorer

genes Rfl and Rf2.

The 1.1 kb transcript was not detected with a probe that contains

the first 114 bp of the 663 bp ORF25 gene (not shown), which suggests

that it is a processed form of another transcript, or initiated within

the gene. The WF9(T) and A188(T) mtDNAs were examined through this

region with six restriction enzymes and several probes without evidence

of genomic variation which might account for the unusual transcript.


The atp6-urfl3-T-ORF25 region of T cytoplasm mtDNA represents an

interesting complex of two cotranscribed open reading frames

(urfl3-T-ORF25) that apparently share transcription initiation sequences

with atp6 by virtue of the 5 kb repeat. Transcription patterns through

this region are therefore complex. Other maize mtDNA genes apparently

have complex transcriptional patterns, but are simplified with defined

probes. A complex pattern of three major transcripts of 2.5, 3, and 4.5

kb was reported for transcription of atpA when a 2.0 kb probe that spans

the coding region was used (Isaac et al., 1985) but a probe specific to

the coding region identified a probable major mature transcript of 2.6

kb (Braun and Levings, 1985). Similarly, the complex pattern reported

for atp6 in T cytoplasm (Dewey et al., 1985a) was obtained with a probe

that contained part of the 5 kb repeat and consequently detected

transcripts of urfl3-T and ORF25. As shown here, probes specific to

the atp6 coding region clarify the transcript pattern as they identify

two major transcripts of 1.8 and 1.6 kb and three minor transcripts, in

N and T cytoplasms. Even with gene-specific probes, transcript patterns

detected for many higher plant mitochondrial protein encoding genes are

not simple and it is not yet known whether this reflects multiple

initiation sites, multiple termination sites and/or post-transcriptional


The subtle difference in transcription of atp6 between N and T

cytoplasms reflects either a unique transcript in T mtRNA (1.55 kb) or a

greater abundance of the same transcript (1.6 kb). Since the abundance

of the 1.8 kb atD6 transcript and larger transcripts were equivalent

among five inbreds in N and T cytoplasms, it is unlikely that the

difference detected in the 1.6 kb range was a consequence of

differential processing. This suggests there is a unique transcript

(1.55 kb) in T mtRNAs. This subtle difference may be related to

nucleotide sequence differences between N and T mtDNAs 5' to the gene,

since the 4 and 5 bp inserts found in the T mtDNA genome are located in

approximately the same region as the initiation or processing sites of

the 1.6/1.55 kb atp6 transcript.

The 5 bp insert, which has indications of a tandem repeat, was not

found in N, C, or S cytoplasms, suggesting that the event occurred in T

mtDNA. Both copies of the 5 kb repeat in T cytoplasm contain the 4 and

5 bp inserts (Dewey et al., 1986) suggesting the insertion events)

occurred prior to the duplication of the 5 kb region or occurred in one

copy and was subsequently duplicated by copy correction.

Transcription patterns of atp6 and urfl3-T/ORF25 indicate that

major transcripts are initiated (and/or processed) at or near the same

sequences within the 5 kb repeat. This indicates that the atp6 promoter

serves the genes urfl3-T and ORF25. The 3.9, 2.0 and 1.85/1.8 kb

urfl3-T/ORF25 transcripts apparently share the same respective

initiation or processing sites within the repeat as the 3.6, 1.8, and

1.6 kb at6 transcripts. Assuming the 200 nucleotide size difference

between atp6 and urfl3-T/ORF25 transcripts reflects their termination

site in the gene-specific regions, one would expect urfl3-T/ORF25

transcripts of 6.0 and 4.3 kb that correspond to the 5.8 and 4.1 kb atp6

transcripts. Transcripts of 6.0 and 4.3 kb were not detected with

urfl3-T probes. The relative abundance of transcripts from the atp6 and

urfl3-T/ORF25 gene region also was not equivalent, as the urfl3-T

transcripts were detected in greater abundance. This may be related to

the stability of the RNAs or the extent of processing events that each

transcript is subjected to, rather than being a reflection of

differential strength of the promoter in the two presumably identical


The effect of the nuclear restorer genes on urfl3-T/ORF25

transcripts has been described as a differential RNA processing event

that results in the loss of a 1.5 kb transcript and the appearance of

1.6 and 0.6 kb transcripts unique to T cytoplasm lines restored to

fertility (Dewey et al., 1986). The basic effect thought to be

associated with fertility restoration has been observed in lines

carrying the Rflrf2 genotype (Dewey et al., 1987). Lines carrying

Rflrf2 and T cytoplasm are male sterile, yet have the 1.6 kb transcript

and a reduced abundance of the 13 kD urfl3-T gene product. Since these

studies were conducted on somatic cells, Rfl might be considered to be a

constitutively expressed gene. We have confirmed the basic effect of

restoration and extended the observations in five additional maize

inbreds in T cytoplasm. Some anomalies are evident, however, in our

transcriptional analyses. Probes from within the 5 kb repeat that

include positions -202 to -92 relative to urfl3-T, detected an enhanced

abundance of a 1.6 kb transcript in R mtDNAs. The abundance of the 1.6

kb atp6 transcript in T mtRNAs was unchanged by fertility restoration,

suggesting that the enhanced transcript is related to urfl3-T/ORF25. A

probe representing positions -91 to +2 (T-ts305) did not hybridize to

the 1.6 kb transcript, but probes 3' to this region readily detected the

restorer-specific 1.6 kb transcript in R mtDNAs. The possibility of an

intervening sequence is raised by these data, but we have been unable to

confirm this observation by primer extension or nucleotide protection

experiments. The effect of the restorer genes on urfl3-T/ORF25

transcripts can, at least, be described as a unique event (whether 5'

processing or splicing) leading to the appearance of the 1.6 kb

transcript. The 1.5 kb transcript, unaffected by fertility restoration,

is unrelated to processes which produce the 1.6 restorer-specific


The appearance of the unique 1.6 kb transcript characterizes

fertility restoration of T cytoplasm in the five lines examined, yet

subtle nuclear background effects among these lines are evident. The

1.6 kb transcript was not detected with a 95 bp probe terminating at +2

in urfl3-T, suggesting that the 5' terminus of the 1.6 kb transcript may

be within the coding region. This is consistent with the observation

that the 13 kD polypeptide is reduced by fertility restoration (Forde

and Leaver, 1980). This suggests that fertility restoration effectively

reduces the relative amount of the mature message, which appears to be

either the 2.0 and/or 1.85/1.8 kb transcript. The abundance of the

mature transcripts was found to be affected by nuclear background, as

the line A188(T) exhibited a greater abundance of the 2.0 kb transcript

relative to the other lines examined. It remains to be determined if

the variability in abundance of urfl3-T/ORF25 transcripts can be

correlated to quantitative differences in the gene product or to the

range of toxin sensitivity exhibited among maize inbred lines.

A more striking example of nuclear background effects was evident

in the appearance of the 1.1 kb ORF25 transcript in T and R mtRNAs from

the lines W64A and WF9. The 1.1 kb transcript was detected in the

progeny of a cross between a line not having the 1.1 kb transcript

[A188(T)] and restored Wf9(T), indicating nuclear control of its

synthesis. The 1.1 kb transcript is not the mature ORF25 transcript,

since its 5' terminus is internal to ORF25. This nuclear effect appears

to be separate from restoration, and is not associated with any known


It is interesting that the posttranscriptional modifications of

urfl3-T/ORF25 transcripts that are directed by the nuclear genome

apparently are gene-specific and reduce, rather than produce, the mature

message. Gene specific processing events directed by nuclear genes have

been characterized in yeast (Dieckmann et al., 1984; McGraw and

Tzagoloff, 1983; Simon and Faye, 1984), although they are involved in

the process of maturation, rather than the reduction, of transcripts.

The synthesis of the 1.1 kb transcript directed by the WF9 and W64A

nuclear genomes appeared to be specific to T cytoplasm. The ORF25 gene

is believed to be a normal constituent of the maize mitochondria and has

homology to mtDNA of at least five other diverse plant species (Dewey et

al., 1986). The differences in the ORF25 transcript patterns detected

among N cytoplasms was not correlated with the presence or absence of

the 1.1 kb transcript among the analogous T cytoplasms. Thus the

effects of the WF9 or W64A genomes on ORF25 transcripts in T cytoplasm

was not exhibited on transcripts of ORF25 in N cytoplasm. The restorer

genes have been found to only influence transcripts in the urfl3-T

region, a region whose organization is unique to T cytoplasm. The

processing site recognized by the fertility restoration gene Rfl,

however, is in a region that has homology to sequences 3' to the 26S

rRNA gene. It is possible that processing of 26S rRNA transcripts may

also be affected by Rfl.

For atp6, urfl3-T, and ORF25, and perhaps other mitochondrial

genes, it is clear that the transcriptional characteristics should be

considered with regard to both the particular cytoplasm and nuclear

background examined. The recombination events that created the unique

organization of the gene urfl3-T not only makes its transcript analysis

specific to T cytoplasm, but also affects transcription of ORF25.

Variation in ORF25 transcriptional patterns among N cytoplasms is likely

related to genomic differences 3' to the gene, whereas subtle atp6

transcript differences between N and T cytoplasms may be related to

small insertion events in a region 5' to the gene. Finally, nuclear

genes have been shown to influence both abundance and synthesis of

specific mitochondrial transcripts associated with the urfl3-T/ORF25





Cytoplasmic male sterility (cms) is a cytoplasmically inherited

alteration in the development of the male gametophyte resulting in

nonfunctional pollen. This phenomenon is widely distributed among

higher plant species, having been reported in greater than 22 different

families (Edwardson, 1970). In maize, there are three major

male-sterile cytoplasms, S, C and T distinguished by the genetics of

fertility restoration (Laughnan and Gabay-Laughnan, 1983).

Nuclear-encoded fertility restoring genes compensate for

nuclear-cytoplasmic incompatibilities or deficiencies that are

phenotypically expressed during microsporogenesis and/or

microgametogenesis. Plants carrying S and C cytoplasm are restored to

fertility by single dominant genes Rf3 and Rf4, respectively, whereas

two genes, Rfl and Rf2, are necessary to restore fertility to T

cytoplasm plants.

Plants carrying T cytoplasm are also distinguished from normal (N),

S and C cytoplasms by their susceptibility to the fungal pathogens

Cochliobolus heterostrophus, race T, and Phyllosticta maydis (rev. by

Ullstrup, 1972) and sensitivity to the insecticide Methomyl (Humayden

and Scott, 1977). The fungal pathogens produce host-selective toxins

that act as virulence determinants and preferentially affect

mitochondria isolated from T cytoplasm plants (Miller and Koeppe, 1971).

The association of cms and sensitivity to the toxins has not been


There is abundant evidence associating a region of the T

mitochondrial (mt) genome with the traits of cms and toxin sensitivity

(rev. by Pring et al., 1987a). The organization of the mitochondrial

DNA region encoding the gene urfl3-T is unique to T cytoplasm (Dewey et

al., 1986) and the urfl3-T gene product, a 13 kD polypeptide, is only

detected in plants carrying T cytoplasm (Forde et al., 1978; Dewey et

al., 1987; Wise et al., 1987b). Synthesis of the 13 kD protein is

reduced upon fertility restoration (Forde and Leaver, 1980) which is

specifically directed by the nuclear Rfl gene (Dewey et al., 1987). The

primary effect of the Rfl gene is posttranscriptional, altering the

urfl3-T transcriptional pattern by functioning as an RNA processing

enzyme or additional transcription factor (Dewey et al., 1987). Male

fertile, toxin insensitive mutants of T cytoplasm have been obtained

through tissue culture regeneration of plants from immature embyros in

the presence (Gengenbach et al., 1977; Brettell et al., 1980) or absence

(Brettell et al., 1980; Umbeck and Gengenbach, 1983) of the fungal

toxins. The majority of the male fertile, toxin insensitive mutants

have suffered deletions involving the urfl3-T gene (Fauron et al., 1987;

Rottman et al., 1987; Wise et al., 1987a), and do not synthesize the 13

kD protein (Dixon et al., 1982). Another mutant, T-4, retains urfl3-T

sequences but fails to synthesize the 13 kD protein due to a 5 bp

insertion in the coding region that places a premature stop codon in the

correct reading frame (Wise et al., 1987a, 1987b).

The organization of the urfl3-T gene region in the T mitochondrial

genome resulted from numerous recombination events involving

mitochondrial and perhaps chloroplast DNAs (Dewey et al., 1986). The

345 bp coding region of urfl3-T contains 263 bp that has 87% similarity

to the 3' flanking region of the 26S rRNA gene and 58 bp of nearly

perfect similarity to a region interior to the 26S rRNA gene. Another

recombination event duplicated a 5 kb flanking region of the gene coding

for subunit 6 of the ATPase complex (atp6). The 3' edge of this

repeated region is 444 bp and 69 bp from the predicted translational

start codons of atp6 and urfl3-T, respectively (Dewey et al., 1986).

Sequences within the 5 kb repeat apparently promote transcription of

atp6 and urfl3-T, as well as another open reading frame, ORF25, that is

80 bp 3' to, and cotranscribed with, urfl3-T (Dewey et al., 1986).

The atp6 transcripts were mapped from N and T cytoplasms, and the

urfl3-T/ORF25 transcripts from several restoring and non-restoring

nuclear backgrounds of T cytoplasm (Chapter III). Here we demonstrate

that the location of the 5' ends of the major primary and processed

transcripts of atp6 and urfl3-T/ORF25 are at identical positions within

the repeated region. Differences in atp6 transcripts detected between N

and T cytoplasms appear to be related to two small DNA insertions that

apparently create an RNA processing site. The 5' terminus of a

transcript that is unique to mitochondria from fertility restored T

cytoplasm plants was not a suitable substrate for capping with guanylyl

transferase and mapped to position +14 of urfl3-T.

Materials and Methods

Seed lines

The maize inbred lines A188(T) (rflrfl;Rf2Rf2), A188(N)

(rfl--;rf2--) and restored WF9(T) (RflRfl;Rf2Rf2) were provided by Dr.

B. Gengenbach. Wf9(T) (rflrfl;rf2rf2) was provided by Pioneer Hi-Bred

International, Inc. The lines A632(T) (rflrfl;----), restored A632(T)

(Rfl--;Rf2--), A632(N) (rfl--;rf2--), W64A(T) (rflrfl;----), restored

W64A(T) (Rfl--;Rf2--), W64A(N) (rfl--;rf2--), C103(T) (rflrfl;----),

restored C103(T) (Rflrfl;Rf2--) and C103(N) (rfl--;rf2--) were provided

by Dr. P. S. Chourey. Other lines used were WF9(N) (rfl--;rf2--) and

the single cross hybrid A188(T) X restored WF9(T).

RNA Analysis

Total mitochondrial RNA was isolated using guanidine isothiocyanate

and prepared for Northern analysis by denaturation with glyoxal and

eletrophoresis in 1.0% agarose gels as described (Wise et al., 1987a).

The glyoxal gels used for Northern hybridizations shown in Figure 4-2

were electrophoresed for 12 hr at 100 V at 40C. The Northern blots in

Figure 4-9 were from 2.0% agarose gels run at 80 V for 12 hr at 40C.

Hybridizations were conducted at 50 or 550C as described (Wise et al.,

1987a) and washed according to Thomas (1980).

Primer Extension and RNA Sequencing

Oligonucleotides complementary to the DNA sequence

5'-TTGGCTCAACTCTCCGAG-3' (+132 to +149 of urfl3-T) and

5'-GACTAGATGGAGTTCCACTG-3' (-346 to -327 of atp6) were synthesized on an

Applied Biosystems 380 1-A DNA synthesizer and purified by recovery from

a 15% acrylamide 8 M urea gel. Acrylamide gel slices containing the

oligonucleotide primers were crushed in a syringe and eluted overnight

at 370 C with a buffer containing 0.5 M ammonium acetate, 0.01 M

magnesium acetate, 0.1% SDS and 0.1 mM NazEDTA. The eluted fragments

were precipitated by the addition of three volumes of 100% ethanol and

0.1 volume ammonium acetate, followed by a 15 min spin in a

microcentrifuge. Approximately 100 pM of the purified oligonucleotides

were end-labelled using T4 polynucleotide kinase and gamma-32P dATP (see

Chapter II) and 10 to 20 ng were used for primer extension reactions.

The synthetic oligonucleotides were annealed to 15 ug of W64A N, T

or restored T mtRNA in a 10 ul volume containing 50 mM Tris-HC1 (pH

8.3), 75 mM KC1, 10 mM Dithiothreitol and 3mM MgC12. The solution was

initially heated to 650C and allowed to cool to 500C or 420C over 1 to 2

hrs. The primers were extended in 15 ul volumes with 200 units of

M-MLV-reverse transcriptase (or 18 to 20 units of AMV reverse

transcriptase) and a mixture of 0.5 mM dATP, dCTP, dGTP and dTTP (dNTP

mix) at 420C for 45 min. The reactions were stopped by adding 5 ul of

formamide and boiled for 3-5 min prior to loading on 6% acrylamide 8 M

urea wedge gels.

The RNA sequencing reactions were identical to the primer extension

reactions except that approximately 100 ng of labelled primer (18mer)

was annealed to 30 ug of restored W64A mtRNA and split into 4 reactions

containing either 0.5 mM of dideoxy (dd) ATP, ddCTP, ddGTP or ddTTP

along with 1 ul of the dNTP mix. Reverse transcriptase was added and

the reactions were incubated at 50SC to help eliminate premature

termination, as described by Geliebter (1987).

Primary Transcript Capping

Approximately 250 uCi of alpha P32 GTP was dried in a vacuum

evaporator prior to the addition of 20 ug of total mt RNA from either

W64A N, T or restored T cytoplasm. The reaction included 50 mM Tris pH

7.8, 1.25 mM MgCl2, 6 mM KC1, 2.5 mM DTT, 0.1 mM Na2EDTA, 10 units

RNAsin (Promega Biotech.) and 7 units of guanylyl transferase in a total

volume of 20 ul. The reaction was incubated at 370C for 30 min followed

by the addition of another 7 units of guanylyl transferase another 30

min incubation. The reactions were terminated with 50 ul of a solution

containing 0.6 M sodium acetate pH 7.0, 50 mM Na2EDTA, 2% SDS, followed

by two extractions with phenol/chloroform and precipitation with 2.5

volumes of ethanol, 0.1 volume of 8 M ammonium acetate.

Specific DNA clones were restricted and subjected to gel

electrophoresis and the inserts were recovered by binding and extraction

from DEAE paper. In a 20 ul volume, 1 to 3 ug of the DNA inserts were

annealed to the capped RNAs in 80% formamide, heated to 850C for 15 min

and cooled to 49C over a 3 hr period. Single stranded RNAs and DNAs

were digested by addition of 250 units of mung bean nuclease in a 300 ul

volume containing 50 mM NaC1, 10 mM sodium acetate pH 5.0, 1 mM

L-cysteine, 4.5 mM ZnC12 which was incubated at 37C for 1 hr. The

reaction was stopped by addition of 50 ul 8 M ammonium acetate, 0.1 K

Na2EDTA followed by extraction with phenol/chloroform and ethanol


3' Labelling, S-1 Digestion

Restriction fragments having 5' staggered ends were isolated from

agarose gels and approximately 1 ug was incubated at 30C for 30 min in

a 10 ul volume containing 7 mM Tris-HCl pH 7.5, 7 mM MgC12, 50 mM NaCI,

15 uCi of alpha 32P dCTP, 1.5 units of Klenow fragment and 0.5 mM of the

particular dNTPs needed to fill in the sticky end. The labelled

fragments were purified either by Sephadex G-50 columns or by ethanol

precipitation. The fragments used in determining the 3' end of RNA

transcripts were heated to 700C for 10 min and annealed with 10 ug of

total mtRNA by cooling the reaction to 490C over 3 hrs in a 20 ul volume

containing 40 mM PIPES pH 6.4, 1 mM Na2EDTA pH 8.0, 0.4 M NaC1 and 80%

formamide. The reaction was brought up to 300 ul in a buffer containing

0.28 M NaC1 0.05 M sodium acetate pH 4.6, 4.5 mM ZnS04, 300 to 600 units

of S-1 nuclease and incubated at 370C for 45-60 min. The reaction was

terminated by extraction with phenol/chloroform followed by ethanol


DNA Markers

The markers were either pUC8 (Yannish-Perron et al., 1985) cut with

Sau3A or TaqI and 3' labelled as described above or were from DNA

sequencing reactions of M13 clones with known sequences. The dideoxy

chain termination method (Sanger et al., 1977) was used for sequencing

as previously described (Wise et al., 1987a).


The mitochondrial genome regions of T cytoplasm containing the

genes atp6, urfl3-T and ORF25 are shown in Figure 4-1. A region of

approximately 5 kb located 5' to ati6 is duplicated in T cytoplasm and

is 5' to urfl3-T in its other arrangement. Northern analysis has shown

that the transcript patterns for these genes are complex, with atp6

having two major (1.8 and 1.6 kb) and three minor transcripts (Chapter

III) and the cotranscribed genes urfl3-T and ORF25 having as many as

nine distinct transcripts (Dewey et al., 1986; Wise et al., 1987a;

Chapter III). The 5' and 3' termini of the atp6 and urfl3-T/ ORF25

.H 0

C ) *

Q -P


S- c

0-P -

C )M


S0 ',

0 -0 0

o k
P m 0
C) C-
*c! is

0 0 I
. H m



transcripts were approximated from Northern analysis and most of the

major transcripts appear to originate in the repeated region (Chapter

III), suggesting these genes share promoter sequences. To accurately

identify the 5' termini of the major transcripts, primer extension

analysis was conducted rather than S-1 analysis to clearly distinguish

atp6 transcripts from urfl3-T/ORF25 transcripts.

Primer Extension

Synthetic primers were prepared, complementary to gene-specific

regions of atp6 (-327 to -346 relative to the proposed translational

initiation site, Dewey et al., 1985) and urfl3-T (+145 to +128, Dewey et

al., 1986). The urfl3-T primer site is 3' to all major urfl3-T/ORF25

transcripts predicted from Northern analysis, in a region that has

sequence similarity to a region 3' to 26S rRNA gene (the primer has 83%

similarity). The results of primer extension experiments using primers

from gene-specific regions of atp6 (20mer) and urfl3-T (18mer) are shown

in Figure 4-2.

Total mitochondrial RNA from both N and T cytoplasms were analyzed

with the atp6-specific 20mer to determine if there is a transcript

unique to T cytoplasm. Northern analyses detected a difference in atp6

transcript patterns between N and T cytoplasms (Chapter III). In all

nuclear backgrounds examined, T mtRNAs displayed a more abundant (or

unique) transcript migrating at approximately 1.55 to 1.6 kb. Major

extension bands of 241 and 250-256 nucleotides were detected only in the

T mitochondrial RNA lane (Figure 4-2; 20mer-T), indicating a unique

transcriptss, designated as the 1.55 kb transcript. These extension

bands correspond to positions -568 and -577 to -583 of atp6 and span a

five bp insertion that was found to be unique to T mitochondrial DNAs

Fig. 4-2. Primer extension analysis. Primers specific to urfl3-T (18)
and to atp6 (20) were annealed and extended with mtRNA from T
and R, or N and T cytoplasms, respectively. Marker is a from
a DNA sequence reaction and indicates distance of primer
extension in nucleotides; the corresponding transcript is
labeled in kilobases. Unlabelled arrowheads are fragments
that do not correspond to known transcripts. Size of
fragments are listed in Table 4-1.


18 20
2.0 > -





4 1.8

.- 4

S< 1.6



1.6 >

* U


(Chapter III). There are two insertions of four and five bp in the T

mitochondrial genome at positions -572 to -576 and -587 to -590 relative

to the proposed translational start site of atp6. The insertions affect

comparisons between the primer extension fragments from N and T mtRNAs,

reflected by a nine nucleotide difference in fragments which are larger

than 266 nucleotides. Two other major groups of bands are detected in

the T lane at approximately 480/470 and 297/290/279 nucleotides, as well

as three minor bands migrating at 391, 195 and 123 nucleotides

(arrowheads). The mapped positions of the major bands (-807/-797 and

-624/-617/-606, respectively) correlates with the 5' termini of the

major atp6 transcripts approximated by Northern analyses. The intensity

of these bands and their relation to the 5' ends of the major

urfl3-T/ORF25 transcripts (see below) support the assumption that these

bands represent the 5' termini of the 1.8 and 1.6 kb atp6 transcripts.

The DNA sequences at these termini were identical in N and T cytoplasms.

There were numerous primer extension bands obtained with both

primers that did not appear to correlate with transcripts detected in

Northern analyses. The appearance and intensity of many of these bands

were variable under different reaction temperatures and with the reverse

transcriptase employed (M-MLV or AMV). Consequently, only bands that

were consistently observed were identified in Figure 4-2 (arrowheads).

Of these bands, only the -123 nucleotide atp6 band and the -225

nucleotide urfl3-T/ORF25 band were associated with the same site within

the repeat (Table 4-1). There is no definitive evidence that any of the

transcripts associated with atp6, urfl3-T, or ORF25 have intervening

sequences or that any of the major transcripts have unique termination

sites (see below), so a correlation between these minor

Table 4-1. Sequences proximal to primer extension bands.

Band Length site
20mer;18mer atp6;urfl3-T

Transcript (kb)

-807; -432 1.8; 2.0
-797; -422 1.8; 2.0


-624; -249
-617; -242
-606; -231

-568; -193
-577; -202
-579; -204
-581; -

1.6; 1.85
1.6; 1.85
1.6; 1.85


-522 ?

-450; -76 ? ; ?

; +10
; +14

; +43


+ 4








+ denotes mapped transcript termini
underlined sequences are not in N mtDNAs; mapped termini only
detected in T cytoplasm in this region


480; 581+/-5
470; 571+/-5


297; 398+/-3
290; 391+/-3
279; 380+/-3

123; 225

; 141
; 136

; 107


1.6 *




1.5 1 I




Fig. 4-3. Sequence of the 5' terminus of the 1.6 kb transcript specific
to mtRNA from restored cytoplasm. The urfl3-T-specific
primer was extended in the absence (lanes T and R) or
presence of DNA chain terminators (A, C, G and T). All
lanes, except the primer extension of T mtRNA, exhibit a band
that correspond to position +14 relative to the initiation
codon of urfl3-T. Sequences continuing beyond this site
represent extensions of other major transcripts.


primer extension bands and discrete transcripts detected on Northern

hybridizations cannot be made.

Primer extension analysis using the 18mer was conducted with mtRNA

from T cytoplasm and T cytoplasm restored to fertility (R). Identical

extension band patterns were observed with the exception of a 136

nucleotide band detected only in the R lane (Fig. 4-2). This band

corresponds to the restorer-specific 1.6 kb transcript (Dewey et al.,

1986; Chapter III). The 5' ends of the major urfl3-T/ORF25 transcripts

(2.0, 1.85 and 1.8 kb) mapped to approximately the same sequences within

the repeat as the major atp6 transcripts. The DNA sequences proximal to

all the primer extension bands are listed in Table 4-1. The size of the

bands corresponding to the longest urfl3-T/ORF25 transcripts was

difficult to accurately determine and were estimated with different

degrees of accurancy (+/- 3 to 5 nucleotides).

RNA Sequencing

The precise location of the end of the 1.6 kb restorer-specific

transcript was identified by primer extension experiment using the 18mer

synthetic primer and DNA chain terminators (Figure 4-3). The RNA

sequence did not deviate from the DNA sequence (Dewey et al., 1986) and

terminated at position +14 of urfl3-T, in agreement with the primer

extension analysis (Figures 4-2, 4-3). Extension of the RNA sequence

beyond the +14 position represents binding to the higher molecular

transcripts (e.g. 1.8/1.85, 2.0 or 3.9 kb transcripts). Northern

hybridization patterns revealed that the restorer-specific 1.6 kb

transcript may participate in an RNA splicing event (Chapter III). The

RNA sequencing results presented here appear to rule out this

interpretation of the Northern hybridizations.

Primary Transcript Capping

Mitochondrial transcripts originating from initiation, rather than

from RNA processing, retain a triphosphate nucleotide at their 5'

termini and can be specifically labelled with the G-capping enzyme,

guanylyl transferase. Total mitochondrial RNA from N, T and restored T

cytoplasms were labelled with guanylyl transferase and subjected to

nuclease protection analyses using DNA clones T-t231, T-t220, N-t3, and

T-t221 (Fig. 4-1) and mung bean nuclease. Labelled transcripts were

only detected when clones T-t220 or the analogous clone from N

mitochondrial genome, N-t3, were used as protection fragments (Figure

4-4). Another fragment detected in all lanes most likely represents a

highly abundant transcript that is not completely digested with mung

bean nuclease. The size difference of the RNA fragments protected by

probe T-t220 (or N-t3) is approximately 6 or 7 nucleotides within each

lane (lanes 3 and 4) and 9 nucleotides between lanes. The mapped site

for these fragments (estimated by DNA sequencing ladders) is -258/-251

for urfl3-T or -625/-618 for atp6 (in the T mitochondrial genome), which

is within 10 nucleotides of the mapped 5' termini of the 1.85 kb

urfl3-T/ORF25 and 1.6 kb atp6 transcripts (-249/-242/-231 and

-624/-617/-606, respectively). The relative proximity between capped

and primer extension sites and the equivalent spacing between specific

termini sites (7 nucleotides) implicates this mtDNA region as a putative

transcription initiation site. Similar analyses of chloroplast RNAs

(Mullet et al., 1985) showed the capped RNAs do not migrate at an

equivalent rate as the DNA markers.

When correlated with primer extension and Northern data, the

transcript capping experiments identify the 1.85 kb urfl3-T/ORF25 and


t3 t220



i'- 160




Fig. 4-4. Primary transcript capping with guanylyl transferase. Total
mitochondrial RNA from normal (N), male sterile (T), or
restored (R) cytoplasms was capped, protected with specific
DNA clones, and subjected to nuclease digestion. Protection
fragments T-t231 and T-t221 (Fig. 4-1) did not protect any
labeled transcripts, whereas clone T-t220 and its equivalent
clone from normal cytoplasm (N-t3; Fig. 4-1) protected two
transcripts. The size of the protected transcript indicates
they correspond to the 1.85 kb urfl3-T transcript and the 1.6
kb atp6 transcript. The shift of nine nucleotides is due to
a difference between N and T mitochondrial DNA in this
region. Marker lane, from a DNA sequencing reaction, is on
far right with size of three fragments in nucleotides.

1.6 kb atp6 transcripts as primary transcripts (initiated) whereas the

2.0, 1.8 and 1.5 kb urfl3-T/ORF25 transcripts and the 1.8 and 1.55 kb

atp6 transcripts are presumably products of RNA processing events. The

1.6 kb restorer-specific transcript was not found to be a suitable

substrate for capping under our experimental conditions. The DNA

sequences proximal to the atp6/urfl3-T/ORF25 transciption initiation

site are similar to sequences located at the 5' termini of the largest

transcripts from the petunia atp9 gene and maize coxl gene, forming a

short consensus sequence (Table 4-2). The similarities between the 5'

termini of the shortest transcript of these genes and a processing site

described in this report are also included.

The protection clones (T-t231, T-t220, N-t3, and T-t221) were

chosen because they span the 5' termini of the major atp6 and

urfl3-T/ORF25 transcript determined from Northern and primer extension

analyses and they were of optimal size for accurate measurement.

Consequently, it is possible that there are other transcripts that were

capped but escaped our detection (e.g. at or near the TaqI sites common

to these adjacent clones).

A schematic representation of location of transcript initiation and

major RNA processing sites for atp6 and urfl3-T (and ORF25) are shown in

Figure 4-5. The distance (in nucleotides) from the 3' edge of the 5 kb

repeat to the translation start codon of atp6 and urfl3-T, as well as

the regions of urfl3-T having homology to 26S rDNA regions are also

shown. The major events involved in the maturation of atp6 and

urfl3-T/ORF25 transcripts deduced from the primary transcript capping,

primer extension, and Northern analyses are illustrated in Figure 4-6.

Table 4-2. Sequence similarities among 5' termini of plant
mitochondrial gene transcripts.

A. Transcript initiation

Petunia atp9a -253

maize atp6 -635

maize coxIb -161

Young Consensusa A

B. Processing sites

Petunia atp9-l and 2c

maize T-specific atp6

maize coxIb

maize atp6-1.8 kb

maize urfl3-T-l.5 kb

maize urfl3-T-Rfl

C. 3' termini

Rothenberg Consensusc

urfl3-T/ORF25 +1146

maize 26S rRNAd 4020

MI M 11 I l
f i i 1 I l l I1









Young et al., 1986
Isaac et al., 1985
Rothenberg and Hanson, 1987
Dale et al., 1984
denotes mapped transcript termini

0 1 0
0 ,C 0 k Cd

I'd k K
0 + -

S C) r S -z d-
0 e r .0 (: 0 HD 0
Ic:R0 cc 0 0
-O 4. H I H -
0 d Tr w ni H C 04
0 0) 0 C C & d X
(t o o a. i.n o o +> -o
2 0 C -IH C -H O

0H [4' H C )C\f+ 0 -(do 0 o F O P f- 411

HC H CC C) d
N c H u *H *H ( 4H:3;

H (l r- +O G m (1)
) *c)'d c) 4 '
to p -t^ o r o m- Z cd

v0N L 0 C d
-* *0- H [H O) 0 -H
U0 [aCd +
H --- .i P RC o[- ,; C)
PC d -H C )0 C C
.m c hNH C0 -i-O Hw [C ;
na- oC 4 o c 0o 0
C -H [C *H - a-H () 0
[CC Nr ct flP
m oC aC a H c -i-
00 C O 0 [- H -iH +--
LO 4*H H -l-- fErH l 0
0 u O 0) n o
C-H CO-IC C O- C a -- H 0
'd-OtS5 HcCM- )[
EC --I HCO ) 0 -d 0 )
*H[ HC)HI O C -H
C ) O ) d IjC m) UHa
*H C 0 = -t a,! 1- 0
4-H -I- -- I 'd +
*c[r rO O [H H toO d -

-C H 0 r0 K
C ai) C)HC -'t- H IC

r-O O )0 [C c o 0'I
*ri O C U 0 C O N CC

h o C C oH ) C o

c 0t)< rC0c *C ) OC',
Hrl Hf 4,i C 0 0 C) --'H- H

h. = ~ m o







co 0



c) -0

4 0 Id E

$ 0 P )
-4 O d $4
0 -H
C) rd w)

C2 -HC)

M 2. p -$

0 C: .) P,
0 .U 0

SC 0 0
EH 0, H

Co H -4-' q 0

3) C BH -p
n -P a

o c C4 )

S1 Ri H

Cd a E Q

O'C-H -C

SOd -H
S Q 0 c

-H 0 ) ) Cv
H-H CC C: C)

C)- 0

CC -H _C -H -C -

C 0 H 0 C '-

C) 3: -H i --


-) -


U, Ulj
0o o C i uC oOco 10
m cm ( - -


J 0

o o

(D co i( Ui

o 0



3' mapping

The 3' termini of the major urfl3-T/ORF25 transcripts were located

by hybridization of a 3' end-labelled DNA probe representing positions

+716 to +1499, relative to the proposed translational start of ORF25

(positions 2356 to 3139 of TURF2H3, Dewey et al., 1986), to total mtRNA

from N or T cytoplasms and subsequent digestion with S1 nuclease (Figure

4-8). A protected fragment of approximately 440 nucleotides is detected

in the T mtRNA lane, corresponding to position +1155 (2795 of TURF2H3).

There appear to be at least four discrete fragments in this size range,

suggesting termination does not occur at a specific nucleotide. The

urfl3-T/ORF25 transcript termination site had been predicted due to the

potential for an RNA secondary structure similar to bacterial transcript

terminators (Schuster et al., 1986). Comparison of the flanking

sequences of this site and a 3' termini consensus sequence are shown in

Table 4-2.

The 3' terminus of the 26S rRNA gene was determined for reasons

described below. A 1973 bp HindIII (T-tsl; Fig. 4-1) fragment

representing positions 3045 to 5018 of the sequence reported by Dale et

al. (1984) was 3' end-labelled and protected with total RNA from T

cytoplasm (Figure 4-8). A nondiscrete band measuring approximately 972

nucleotides was detected that corresponds to position 4018 (+/- 8

nucleotides) of the DNA sequence (Table 4-2). This site is

approximately 20 nucleotides upstream of the stop site predicted by

homology to E. coli 23 rDNA (Dale et al., 1984). Other protected

fragments of similar size were also detected, but were much less


ml N T

955- -

- 784t-S

26S rRNA
ml T m2

1973 H


A 736


955- .- .

585 *

* 585-


341- S



Fig. 4-7. Mapping the 3' termini of urfl3-T/ORF25 and 26S rRNA
transcripts. A 784 bp Ta'I-Smal fragment (T-tsl, Fig. 4-1)
and a 1973 bp HindIII fragment were labelled and hybridized
to total mtRNA from N or T cytoplasm followed by S-1
digestion. Arrowheads mark the major protected fragments.
Markers are pUC8 digested with Sau3A (ml) or TaqI (m2) and
labelled. Size of fragments are in nucleotides.

Does Restoration Affect 265 rRNA Processing?

The proposed RNA processing sites at positions +14

(restorer-specific) and +43 of urfl3-T are within the region that has

homology to sequences 3' to the 26S rDNA gene. A region of high

homology extends 68 nucleotides upstream of the +14 position (90%

homology) and 52 nucleotides downstream (98% homology; Dewey et al.,

1986). To determine if the +14 or +43 sites are utilized in the

processing of the 265 rRNA transcripts, Northern blots of mtRNA from N,

T and restored T cytoplasms were probed with DNA clones from the 26S

rDNA region of A188(T) genome that represent positions 3881-4049 (A)

4050-4343 (B), and 4344-4422 (C; Figure 4-9). No differences in

transcript patterns between N, T, and restored-T mtRNAs were detected.

Probes B and C hybridized to an abundant message of approximately 3.6

kb, comigrating with the 26S rRNA transcript detected with probe A.

Although the nitrocellulose membranes probed with B and C were exposed

two to three times longer than the membrane probed with A, the

transcript detected with B and C could be the mature 26S rRNA, which

would indicate that the transcript termination site could extend at

least 400 nucleotides downstream from the predicted stop site. The S-1

analysis described above does not support this hypothesis, thus the

transcript detected with probes B and C may represent an unprocessed

form of the the gene. Probes B and C also detect the major

urfl3-T/ORF25 transcripts (2.0, 1.85/1.8, 1.6 and 1.5 kb; Fig. 4-9,

probe C).

Fig. 4-8. Northern analysis of 3' end of 26S rRNA gene. Probes A, B
and C represent the regions shown in the partial restriction
map were hybridized to total mtRNAs from N, T and T-restored
(R) cytoplasms. Blots B and C were expososed for
approximately two to three times longer than blot A. Mapped
3' termini of the major 26S rRNA transcript, predicted in
Fig. 4-7, is shown. The 3.6 kb band detected with probe A
represents the major 26S rRNA transcript, and a less abundant
3.6 kb transcript detected in B and C may represent a
precursor transcript. Hatched region represents region of
similarity to urfl3-T and probe C detects the major
urfl3-T/ORF25 transcripts (2.0, 1.85/1.8, 1.6, and 1.5 kb).

3 lorminus
26S rRNA



S1///////I////1 3.



mm cre 3.6




9i~ ~





The major steps involved in the maturation for the genes atp6,

urfl3-T, and ORF25 transcripts were elucidated through identification of

transcript initiation and processing sites and their relation to the

transcript map for these genes developed from Northern hybridizations

(Chapter III). Although the events involved in the synthesis of a few

minor transcripts, and the authenticity of some primer extension

fragments have not been resolved, there are no major inconsistencies

between the primer extension/capping data and the hybridization

analyses. The size of the major transcripts predicted from the start

and stop sites described in Tables 4-1 and 4-2 were all within 40

nucleotides of the size estimated from Northern analyses. Two ambiguous

primer extension bands (-450 of atp6; -76 of urfl3-T) were detected that

corresponded to the same nucleotide sequence (Table 1), but did not

correlate with Northern data. It is noteworthy that DNA sequences

surrounding this site have great similarity (29 of 30 bp) to a region

that is immediately 5' to the 26S rRNA gene (Dewey et al., 1986).

The major atp6 transcripts detected in mtRNA from N cytoplasm (1.8

and 1.6 kb) result from a processing and an initiation event. The

processed transcript (1.8 kb) is most likely derived from either the 4.1

or the 3.6 kb transcript, since a 2.1 kb transcript is detected upstream

of the processing site (e.g., 1.8 + 2.1=3.9). The relative abundance of

the 1.8 kb transcript is three to five times greater than the abundance

of the 1.6 initiated transcript (Chapter III), thus the major atp6

transcription initiation site is located approximately 2.5 to 3.0 kb

upstream of the predicted translational start. In mtRNA from T

cytoplasm, a processing site apparently has been formed by the addition

of two small DNA insertions of 4 and 5 bp. This processing site

slightly affects the abundance of both the 1.8 and 1.6 kb transcripts,

resulting in a 1.55 kb transcript that is about half as abundant as the

1.8 kb transcript.

The nucleotide sequence of a 3.0 kb region that includes the

urfl3-T gene (Dewey et al., 1986; Chapter III) reveals as many as six

small DNA insertions or deletions, deduced through comparison of

sequence similarities to the DNA progenitor regions. Three of these are

exhibited as perfect 5 bp tandem direct repeats, two as 4 bp repeats

separated by a single bp and one that is a 4 bp insertion with 2 bp in

common with the adjacent 4 bp sequence. Most of these insertions or

deletions resemble classical proposals for frameshift mutations produced

by mechanisms involving misaligned pairing of repeated DNA sequences

(Steisinger et al., 1966). The region of urfl3-T that has similarities

to a sequence 3' to 26S rDNA gene has lost one copy of a tandem 5 bp

repeat and suffered a G to A transition when compared to the 26S rDNA

region. A tissue culture derived fertile, toxin insensitive mutant (T4;

Umbeck and Gengenbach, 1983) was shown to regain the tandem five bp

duplication in its urfl3-T sequence (Wise et al., 1987a), matching the

26S rDNA sequence in this region. This mutant does not synthesize the

13 kD protein because the 5 bp duplication places a stop codon in frame

that truncates the predicted protein to 8.3 kD (Wise et al., 1987a,

1987b). The effect of the small insertions and deletions, presumed to

be caused by errors in replication, or failure to correct these errors,

is illustrated in two ways within a relatively small DNA region of the T

mitochondrial genome; one event apparently creates a new RNA processing

site and the other is involved with a frameshift mutation.

The initiation and processing sites for the urfl3-T/ORF25

transcripts within the 5 kb repeat are the same as those for atp6

transcripts; however, urfl3-T/ORF25 transcripts are subject to as many

as three more events outside of 5 kb repeat. Processed transcripts were

detected in T mtRNA having 5' termini at +43 (1.5 kb transcript) and in

nuclear fertility restoring backgrounds (specifically, those with Rfl;

Dewey et al., 1987) at +14 of urfl3-T (1.6 kb transcript). Both

processing events result in transcripts that do not contain the entire

urfl3-T gene, yet retain ORF25 sequences. Another event, not yet

determined to be the result of processing or initiation, occurs internal

to ORF25 resulting in a 1.1 kb transcript. The detection of this

transcript, like the 1.6 kb transcript, depends on nuclear background

(see Chapter III).

Numerous recombination events were involved in establishing the

complex arrangement of the atp6, urfl3-T and ORF25 genes in T cytoplasm

of maize. An event critical for the transcription of urfl3-T was the

duplication of the atp6 flanking region which placed DNA sequences

involved in transcription initiation 5' to urfl3-T and ORF25. If the

degree of similarity between sequences that make up the chimeric gene

region of urfl3-T and their proposed progenitor sequences (e.g. atp6,

26S rDNA) is indicative of the chronological order of the recombination

events, the duplication of the 5 kb region would be the most recent

event. The regions of urfl3-T similar to sequences internal and 3' to

the 26S rRNA gene have apparently diverged since their duplication as

they are 95% and 85% similar, respectively (Dewey et al., 1986).

All transcripts that are not a suitable substrate for guanylyl

transferase are assumed to be the result of RNA processing events,

although they may have alternative origins. Additional evidence for the

authenticity of the proposed processing sites that result in the 1.8 kb

atp6 and 2.0 kb urfl3-T/ORF25 transcripts and the site for the 1.5 kb

urfl3-T/ORF25 transcript is provided by transcripts 5' to these sites

which are of an appropriate size to be products of a processing events

of a larger transcript. For example, the 0.4 kb transcript detected

immediately upstream of the 1.5 kb urfl3-T/ORF25 transcript is most

likely a product of the cleavage of the 2.0 kb urfl3-T/ORF25 transcript.

The nuclear gene Rfl has been described to act as an RNA processing

enzyme or as an additional transcriptional factor which apparently leads

to a reduction of the 13 kD protein by altering or limiting the

abundance of the urfl3-T transcripts (Dewey et al., 1986, 1987; Chapters

III, IV). There are, however, a number of conflicting observations

concerning this conclusion. First, no transcripts, specific to

fertility restored cytoplasms, are detected 5' of the +14 site of

urfl3-T. Second, the abundance of the 2.0, 1.85, and 1.8 kb

urfl3-T/ORF25 transcripts is only slightly reduced, not nearly as

dramatic as the 67% reduction detected on protein gels with labelled

mitochondrial polypeptides (Dixon et al., 1982) or the even greater

reduction apparent on Western blots (estimated from Dewey et al., 1987).

Finally, although the urfl3-T sequence spanning the +14 site is very

similar to sequences 3' to the 26S rRNA gene, fertility restoring genes

were not found to have an effect on transcripts in this region. An

alternative explanation for the function of the Rfl gene may be that it

affects some aspect of translation, prematurely terminating the

translation of some, but not all, transcripts and secondarily creating a

1.6 kb transcript. This would seemingly be in better agreement with the


analyses of the 13 kD and its reduction upon restoration, as well as the

observation that fertility restoration does not influence 26S rRNA

transcripts. Consequently, although the influence of Rfl gene

effectively reduces the abundance of 13 kD protein, the manner in which

it accomplishes this is not completely clear, nor is the role of Rf2 in

fertility restoration.



The following observations implicate the gene urfl3-T as having a

major role in the phenotypes of cms and toxin sensitivity 1) urfl3-T and

its gene product, the 13 kD protein, are unique to T cytoplasm, 2)

urfl3-T transcripts and the abundance of the 13 kD protein are affected

by fertility restoration, specifically, the Rfl gene, and 3) urfl3-T is

deleted or is truncated in tissue culture derived fertile, toxin

insensitive mutants. Despite these facts, much remains to be answered.

The first major question concerns the toxin sensitivity of T cms

plants restored to fertility. Although these plants are completely

fertile, their sensitivity to the toxins is only slightly moderated, at

best 50%. Fertility restoration has been shown to reduce the abundance

of the 13 kD protein up to 67% (Dixon et al., 1982), determined by

labelling proteins in isolated mitochondria. Western blots, probably a

more accurate method of quantitating proteins in vivo, reveal a greater

reduction in the abundance of the 13 kD protein upon fertility

restoration or with Rfl alone (estimated from data presented in Dewey et

al., 1987). One hypothesis for this apparent separation of the traits

of cms and toxin sensitivity in restored plants is that the decreased

abundance of the 13 kD protein could be sufficient to overcome cms yet

not great enough to affect its ability to serve as a receptor for, or

have some other interaction with, the toxin molecules. In this

scenario, only the complete absence of the 13 kD would confer

insensitivity to the toxins.

The second enigma involves the role of Rf2 in fertility restoration

and toxin sensitivity. The effect of Rf2 on toxin sensitivity can be

determined by assaying the response of isolated T cytoplasm mitochondria

from the proper nuclear backgrounds. Comparison of the toxin response

of T mitochondria from backgrounds having different dominant Rf genes

(e.g. rflrfl;rf2rf2, Rfl--;rf2rf2, rflrfl;Rf2--, and Rfl--;Rf2--) should

discriminate the differences between the individual effects of Rfl and

Rf2. Unfortunately, identifying the influence of Rf2 in fertility

restoration will not be as simple.

The primary influence of the Rfl gene on urfl3-T appears to involve

an RNA processing event that alters the urfl3-T messages (Dewey et al.,

1986, 1987; Chapters III and IV). The relative reduction of mature

urfl3-T messages, determined from Northern analyses (Chapter III), does

not appear to be as great as reduction of the 13 kD protein in restoring

or Rfl containing nuclear backgrounds (Dewey et al., 1987). The mature

transcripts (assumed to be the 2.0, 1.85 and 1.8 kb urfl3-T/ORF25

transcripts) are slightly reduced upon fertility restoration, but they

are present in relatively high abundance when compared with atp6 or

ORF25 transcripts in N mtRNAs. This inconsistency creates some

uncertainty regarding the hypothesis that the Rfl gene product is

directly involved in RNA processing. With the available data, no other

obvious mechanisms for Rfl action are apparent, yet it is conceivable

that Rfl may be related to some aspect of translation and that the

restorer-specific 1.6 kb transcript is a secondary effect of an aborted

translation. This idea is not without precedent, as nuclear genes that

influence the translation of specific mitochondrial genes have been

reported for several genes in yeast (rev. by Fox, 1986).

The loss of the 13 kD protein in the tissue culture derived

fertile, toxin insensitive mutants identifies this protein as having an

involvement in T-cms. The only mtDNA difference detected between one

mutant, T4, and the mtDNA of its progenitor, T cytoplasm, appears to be

a G to A transition and a five bp insertion in the urfl3-T gene (Wise et

al., 1987a). Transcription was unaltered in this mutant and the

insertion created a premature stop in the translational reading frame,

truncating the predicted gene product to 8.3 kD. If no other DNA

regions or polypeptides are affected in this mutant, it would appear

that the 13 kD protein is the causal agent of T-cms. However, this

mutant was derived from and maintained in the nuclear background A188

that is dominant for Rf2 (rflrfl;Rf2Rf2). Until it is demonstrated that

this mutant, and the other tissue culture derived fertile, toxin

insensitive mutants, maintain these characteristics in backgrounds that

are recessive for both Rf genes, other factors besides the 13 kD protein

may play a role in T-cms.

The novel DNA rearrangements in the urfl3-T region of the T

mitochondrial genome not only create the urfl3-T gene but may have

indirect effects on atp6 and ORF25 gene expression. Transcriptional

studies (presented in Chapters III and IV) detected differences between

atp6 transcripts in mtRNAs isolated from N and T cytoplasms, although

these were unrelated to the presence or absence of dominant fertility

restoring genes. Major differences were detected between ORF25

transcripts of N and T mtRNAs and among mtRNAs isolated from N

cytoplasms from different nuclear backgrounds which appear to relate to

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs