Regulation of mitochondrial gene expression during early development of Xenopus laevis

MISSING IMAGE

Material Information

Title:
Regulation of mitochondrial gene expression during early development of Xenopus laevis
Physical Description:
xi, 274 leaves : ill. ; 29 cm.
Language:
English
Creator:
Ammini, Chandramohan V
Publication Date:

Subjects

Subjects / Keywords:
Molecular Genetics and Microbiology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Molecular Genetics and Microbiology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 248-273).
Statement of Responsibility:
by Chandramohan V. Ammini.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 027672258
oclc - 37274513
System ID:
AA00012971:00001


This item is only available as the following downloads:


Full Text










REGULATION OF MITOCHONDRIAL GENE EXPRESSION
DURING EARLY DEVELOPMENT OF XENOPUS LAEVIS













By

CHANDRAMOHAN V. AMMINI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1997














ACKNOWLEDGEMENTS


I thank Dr. William W. Hauswirth, my mentor and supervisor, for

his constant support in both academic and personal matters. He gave

me immense freedom in the conduct of experiments and stood by me

during difficult times. I knew I could always count on him. I would

also like to thank all the past and present colleagues in the lab, who

have made life in the lab enjoyable and friendly.

I would like to thank the members of my advisory committee,

Drs. Al Lewin, Maury Swanson, and Tom Rowe, for the various

constructive suggestions they have made during the course of the

investigation and for taking the time to review various documents,

including this one. I also appreciate the input from all the other

faculty and students in the department, which has helped in my

personal growth. I also thank the other personnel in the Department,

especially Joyce Conners and Brad Moore, for their efficient help.

My special thanks go to Hyma, my wife, who has stood by me

during all times, good and bad, and made this all possible.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS

LIST OF TABLES . .

LIST OF FIGURES . .


ABSTRACT .


CHAPTERS


1 INTRODUCTION .
General Introduction. .
Mitnhnnrdrial DNA


Mitochondrial DNA Replication
Mitochondrial Transcription .
Regulation of Mitochondrial Gene


ix
. ix






4
6


Expression .' 15


Focus of Dissertation .

2 MATERIALS AND METHODS .

Frogs . .
DNA Primers . .
Laboratory Fertilization of Eggs.. .
De-jellying of Eggs and Embryos .
Isolation of Mitochondria from Eggs and Embryos. .
Isolation of Mitochondria from Ovary Tissue .
Dimethyl Sulfate Footprinting of Mitochondrial DNA. .
Steady State RNA Analysis. .
Mitochondrial Run On Transcription. .
Measurement of Mitochondrial Adenine Nucleotides
Electrophoretic Mobility Shift Assays .


25
25
25
27
28
29
30
48
56
61
63


: : :


. .








3 STEADY STATE ANALYSIS OF MITOCHONDRIAL
RNA LEVELS . 68

Introduction . 68
Results . 72
Discussion ........ 92

4 ANALYSIS OF PROTEIN-MITOCHONDRIAL DNA
INTERACTIONS AT CRITICAL CIS-ELEMENTS 1 01

Introduction . .. 101
Results. . 115
Discussion . 159

5 MITOCHONDRIAL TRANSCRIPTION RATE ANALYSIS. 178

Introduction . .. 178
Results. . 8 1
Discussion ...... 2 1 2

6 SUMMARY AND PERSPECTIVES ... 230

LIST OF REFERENCES ... 248

BIOGRAPHICAL SKETCH. . 274














LIST OF TABLES


3 -1 Steady state levels of mitochondrial RNA in total egg or
embryonic RNA during early development
of Xenopus laevis ..... 77

3-2 Mitochondrial mRNA steady state levels in isolated
mitochondria from distinct developmental
stages of X. laevis. .. 82

3-3 Steady state levels of mitochondrial RNA (normalized
to mtDNA) during early development of X. laevis 89

5-1 Promoter-wise and overall transcription rates per unit
mitochondrial genome during early development
of Xenopus laevis. . 190

5-2 Mitochondrial adenine nucleotide levels during early
development of Xenopus laevis. 210

5-3 Total adenine nucelotide levels in unfertilized eggs and
early developmental stages of Xenopus laevis 211














LIST OF FIGURES


2-1 Experiment to determine the optimal number of cycles
of linear amplification .. ... .. 43

2-2 Experiment comparing 3 different thermostable DNA
polymerases for accurate amplification of the
footprint ladder .. 45

3-1 Steady state levels of mitochondrial RNA during early
development of Xenopus laevis .. 75

3-2 Steady state levels of mitochondrial RNA isolated from
purified mitochondria during early development
of Xenopus laevis ..... 81

3-3 Steady state levels of mitochondrial RNA per unit
mitochondrial genome during early development
of Xenopus laevis .... 87

3-4 Steady state levels of mitochondrial mRNAs per unit
mitochondrial genome during early development
of Xenopus laevis . 91

3-5 Revised view of mitochondrial gene regulation during
early development of Xenopus laevis 99

4-1 General scheme for in vivo DNA footprinting. 117

4-2 In organello footprint of the promoter region. 122

4-3 In organello footprinting of the D-loop upstream region 124









4-4 Schematic view of protein-DNA interactions detected by
in organello footprinting of the D-loop upstream region 128

4-5 In vivo footprinting of the H-strand of D-loop upstream
region during early development of X. laevis 132

4-6 In vivo footprinting of the L-strand of D-loop upstream
region during early development of X. laevis. 134

4-7 In vivo footprinting of the L-strand of the promoter
region during early development of X. laevis. .. 137

4-8 In organello footprint of the mTERF region
in ovary mtDNA .. 140

4-9 Summary of footprints on the L- and H-strand of the
mTERF region of ovary mtDNA ... 142

4-10 In vivo footprinting of the L-strand of the mTERF region
during early development of X. laevis 146

4-11 In vivo footprinting of the H-strand of the mTERF region
during early development of X. laevis 148

4-12 Gel mobility shift assay for mTERF protein. 152

4-13 Comparison of the relative abundance of mTERF in
distinct developmental stages 156

5-1 Size fractionation of mitochondrial run-on RNA. 183

5-2 Hybridization target panel for strand-specific
detection of run-on transcripts 1 86

5-3 Mitochondrial run-on transcription in early
developmental stages .......... 189

5-4 Promoter-wise and overall transcription rates
during early development .. .. .. 192








5-5 Comparison of mitochondrial transcription rates and
steady state levels of mitochondrial transcripts during
early development of X. laevis 196

5-6 Effect of addition of metabolic intermediates on the
transcription rate of 20 hr stage mitochondria 200

5-7 Effect of addition of metabolic inhibitors on the
transcription rate of 20 hr stage mitochondria .. 204














Abstract 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

REGULATION OF MITOCHONDRIAL GENE EXPRESSION
DURING EARLY DEVELOPMENT OF XENOPUS LAEVIS

By

Chandramohan V. Ammini

May, 1997

Chairperson: William W. Hauswirth
Major Department: Molecular Genetics and Microbiology


Mitochondrial transcript levels are actively regulated during early

development of Xenopus laevis, whereas mitochondrial DNA (mtDNA)

levels are constant during this time, providing the unique

opportunity to study mitochondrial transcriptional regulation

decoupled from mtDNA replication. The main focus of this

dissertation was to gain further insights into the molecular

mechanisms of mitochondrial gene regulation in this system.

Measurement of steady state transcript levels per unit

mitochondrial genome showed that transcript levels were basal up to








gastrulation, with a coordinate induction thereafter. Mitochondrial

run-on transcription assays using isolated organelles from distinct

developmental stages showed that transcription was the primary

determinant of differences in transcript abundance. To test whether

transcription was regulated by the changing stoichiometries of the

Xenopus mitochondrial transcription factor A (xl-mtTFA), in vivo

footprinting of mtDNA was carried out. Although footprint assays

confirmed that the mtTFA binds the D-loop region in a phased

manner, no significant differences were detected in the occupancy of

xl-mtTFA during regulated transcription, showing that regulation

was not by changing xl-mtTFA binding to mtDNA. Mitochondrial

transcriptional activity showed a positive correlation with the

metabolic status of mitochondria, as indicated by mitochondrial ATP

measurements and run-on transcription assays in the presence of

metabolic intermediates. This suggests that transcriptional

regulation may be sensitive to the mitochondrial metabolic status.

Footprinting and band shift assays confirmed the presence of the

mitochondrial termination factor (mTERF) in Xenopus mitochondria.

The mode of footprinting suggested a localized distortion of DNA

structure, consistent with in vitro reports of DNA bending by this








protein. However, the binding site was equivalently occupied during

regulated transcription, and band shift assays showed equivalent

protein concentrations during this time, suggesting that mTERF is a

constitutive DNA-bending protein, which remains bound to its

cognate site at all times, posing a physical block to the elongating

RNA polymerase.














CHAPTER 1
INTRODUCTION


General Introduction


Mitochondria are essential organelles of eukaryotic cells which

are largely maternally inherited and play a vital role in cellular

physiology. Their principal function in the cell is production of

adenosine triphosphate (ATP) by the highly efficient process of

oxidative phosphorylation, but they also contribute to the

biosynthesis of macromolecules such as pyrimidines, amino acids,

phospholipids, nucleotides, folate coenzymes, heme, urea, and many

other metabolites (Attardi and Schatz, 1988).

A feature that sets them apart from other organelles is their small

but essential DNA genome, which is distinct from nuclear DNA. Each

human cell has hundreds of mitochondria and thousands of mtDNAs.

The presence of an extranuclear genome necessitates the presence of

biosynthetic machinery for the transcription and replication of

mitochondrial DNA, and mitochondrial ribosomes for the translation

of mitochondrial mRNAs. Mitochondrial biogenesis involves a major

contribution of gene products (more than 300) encoded in the






2

nuclear genome, and a minor, though essential, contribution of gene

products (less than 40) encoded in the mitochondrial genome

(Attardi and Schatz, 1988). Sophisticated mechanisms ensure the

import and intramitochondrial sorting of cytoplasmically synthesized

proteins, phospholipids, and probably a few RNAs. Hence,

mitochondrial biogenesis is complex and under multiple layers of

regulation (Attardi and Schatz, 1988; Chomyn and Attardi, 1992).

The linkage of chronic degenerative human diseases to

mitochondrial dysfunction has resulted in a dramatic upswing of

scientific interest in the biogenesis and regulation of this complex

organelle. Mitochondrial encephalomyopathies are a diverse group

of disorders that result from the structural, biochemical, or genetic

derangement of mitochondria. Therefore, they are manifested

primarily in organs requiring high energy levels, like the brain,

heart, muscle, kidney and endocrine glands. Since these

encephalomyopathies share systemic features, they are named after

the symptoms they induce, like MELAS (Myoclonic Epilepsy, Lactic

Acidosis and Stroke-like episodes), MERRF (Myoclonic Epilepsy with

Ragged-Red Fibers), LHON (Leber's heriditary optic neuropathy), KSS

(Kearns-Sayre syndrome), etc. (Wallace, 1993).

The physiological basis of all these disorders is that mitochondria

have to produce a certain amount of energy continually.






3

Theoretically, a defect in any one of the three to four hundred plus

genes (nuclear- or mitochondrial-encoded) involved in

mitochondrial biogenesis can weaken mitochondrial function, leading

ultimately to chronic degenerative disease. If the mutation is in

mtDNA, an intracellular mixture of normal and mutant mtDNAs is

created- a phenomenon called heteroplasmy (Wallace, 1993).

Replicative segregation of heteroplasmic genotypes towards

homoplasmic types leads to rapid fixation of the mutations (Laipis et

al., 1988; Yoneda et al., 1992). Cellular model systems, where human

cells lacking mtDNA are repopulated with exogenous mitochondria

from patients with mitochondrial myopathies (King and Attardi,

1988; 1989; Chomyn et al., 1991) have provided elegant tools to

study the role of these mtDNA mutations in cellular physiology.

These studies show that even 10-15% of wild type mtDNA, when

present in the same organelle with 85-90% mutant mtDNA, has a

protective effect against protein synthesis and respiration defects,

(Yoneda et al., 1994; Attardi et al., 1995). In human disease, a

similar mechanism is presumed to induce disease when the genetic

threshold is crossed. This explains how a mtDNA mutation can be

present at birth, yet only be clinically manifest in the adult or

primarily affect only certain tissues. In higher plants, mitochondrial

dysfunction, resulting from rearrangements in mtDNA, or






4

incompatibility between mitochondrial and nuclear genetic

backgrounds, often result in cytoplasmic male sterility, causing

infertile pollen (Levings and Brown, 1989).


Mitochondrial DNA


Mitochondrial DNA has the same basic role in all eukaryotes that

contain it: it encodes rRNA and tRNA components of a mitochondrial

protein synthesizing system, which translate a limited number of

mtDNA-encoded mRNAs that code for some of the subunits of key

enzyme systems involved in oxidative phosphorylation. There is

extraordinary variation in the size, structure, organization, and mode

of expression of the mtDNA. They include the gigantic 570-kb circles

of maize (Lonsdale et al., 1984), 46-kb linear molecules of

Tetrahymena (Borst, 1980), 74-85 kb of Saccharomyces cerevisiae

(Zamaroczy and Bernardi, 1986), the extraordinarily complicated

maxicircles of Trypanosoma brucei (Benne, 1985; Simpson, 1987),

and the extremely compact vertebrate mtDNA (16-20 kb). The

existence and structure of vertebrate mtDNA was established more

that 25 years ago. In all vertebrate systems studied to date, mtDNA

is a closed circular duplex species of approximately 16-18 kb, which

in many cases contains a triple-stranded region called the

Displacement loop or D-loop (ranging from 0.5-1.7 kb in different







5

species) at a specific location in the genome. Early studies on the

mode of mtDNA replication revealed that each strand of the duplex

could be distinguished on the basis of G+T base composition: a bias in

composition results in different buoyant densities of each strand

('heavy' or H-strand and 'light' or L-strand) in alkaline cesium

chloride gradients.

The most striking feature of vertebrate mtDNA is its extremely

compact gene organization which allows it to encode information for

37 known genes. The major coding strand (the H-strand) contains 28

genes: two rRNAs, 14 tRNAs, and 12 proteins. The L-strand encodes

8 tRNAs and one protein. The 13 proteins include seven subunits of

NADH dehydrogenase (ND 1, 2, 3, 4, 4L, 5, 6), three subunits of

cytochrome c oxidase (COX 1, 2, 3), two subunits of ATP synthase

(ATPase 6, 8), and the apocytochrome b (Cyt b). In most cases, these

genes abut each other or are separated by only a few nucleotides,

and there are no introns, the only noncoding stretch being in the D-

loop control region housing the transcription and replication

elements. Almost all reading frames lack significant nontranslated

flanking regions (Montoya et al., 1981), and most of them lack even a

complete termination codon, ending instead with either a T or TA

following the last sense codon. In these cases, completion of the






6

termination codon TAA occurs post-transcriptionally by

polyadenylation of the mRNAs (Ojala et al., 1981).

Mitochondrial DNA Replication


The mechanism of vertebrate mtDNA replication has been studied

primarily in cultures of human and mouse cells, yeast, rat liver, and

is understood in considerable detail. In 1973, Berk and Clayton,

using mouse L cell lines lacking the cytoplasmic thymidine kinase

but retaining the mitochondrial enzyme, which allowed selective

radiolabeling of mtDNA, showed that mtDNA replication continues

throughout the cell cycle, doubling at about the same rate as the cell,

with mtDNA molecules being selected at random for replication. This

phenomenon forms the basis of replicative segregation of

heteroplasmic genotypes.

There are two separate and distinct origins of replication for each

strand. The origin of H-strand synthesis (OH) is located within the

non-coding D-loop region of the genome (Robberson and Clayton,

1972; Brown and Vinograd, 1974; Robberson et al., 1974; Tapper and

Clayton, 1981), whereas the origin of L-strand synthesis (OL) is

nested within a cluster of five tRNA genes located several kilobases

away from the On (Martens and Clayton, 1979). A commitment to

mtDNA replication begins by the initiation of H-strand synthesis that






7

results in strand elongation over the entire length of the genome.

Initiation of L-strand synthesis begins only after the 0, is exposed as

a single-stranded template. Asynchronous replication of each strand

then completes replication, the region between the two origins being

replicated last.

The critical elements of the mammalian H-strand replication

origin include the L-strand promoter and three downstream

conserved sequence blocks (CSB I, II, III). The involvement of RNA

primers in the initiation of H-strand synthesis was shown by the

presence of alkali-sensitive 5' ends in D-loop DNA (Gillum and

Clayton, 1979; Brennicke and Clayton, 1981). The RNA primer was

shown to originate from the L-strand promoter (Chang and Clayton,

1985; Chang et al., 1985), with a sharp transition from RNA to DNA

occurring within a series of highly conserved sequence blocks (CSB

1, 2, 3), in a region of mtDNA which is otherwise highly divergent

across species (Walberg and Clayton, 1981). A search for an

endonucleolytic activity with the ability to recognize features of the

CSBs and process nascent RNA transcripts led to the discovery of

mouse RNAase MRP (mitochondrial RNA processing) (Chang and

Clayton, 1987a). This was found to be a ribonucleoprotein, with a

136-nucleotide RNA species (Chang and Clayton, 1987b), with a 10

nucleotide sequence complementary to CSB II (Chang and Clayton,








1989). The involvement of this factor in creating replication primers

has been questioned by a subcellular and suborganellar fractionation

study which failed to detect RNase MRP RNA in mitochondria (Kiss

and Filipowicz, 1992). However, a recent analysis of the RNase MRP

RNA distribution by direct in situ hybridization localized a definable

portion of intracellular RNase MRP to the mitochondria of mouse cells

(Li et al, 1994).

CSB II (also called GC cluster C in yeast), is GC rich and is the most

conserved element in mtDNA replication origins from Saccharomyces

cerevisiae to humans. Xu and Clayton (1995) found a stable and

persistent RNA-DNA hybrid formation beginning at the 5' end of CSB

II during in vitro transcription with templates carrying yeast

replication origins. CSB II was the necessary and sufficient nucleic

acid element for establishing stable hybrids, with the efficiency of

hybrid formation decreasing with increasing distance between CSB II

and the promoter. Once made, the RNA strands of these RNA-DNA

hybrids could be elongated efficiently by mtDNA polymerase. These

results suggest that after transcription initiation, the next critical

step for initiating mtDNA replication is the formation of a stable

RNA-DNA hybrid. Xu and Clayton (1996) also found the formation of

a similar RNA-DNA hybrid (R-loop) in the human system. Efficient

hybrid formation was again dependent on the GC-rich CSB II






9

element. Interestingly, efficient hybrid formation was influenced by

sequences 5' to the hybrid, including the CSB III element. In

structural studies of this RNA-DNA hybrid, Lee and Clayton (1996)

found the R-loop to be exceptionally stable, with extensive

intrastrand folding, suggesting that the RNA strand in the RNA-DNA

is organized into a specific conformation, which may have important

implications for RNA processing in vivo. These newly discovered

features of H-strand priming increase the analogy between mtDNA

replication and ColEl plasmid replication. In the ColEl plasmid origin

of replication, a stretch of six rG residues in the primer RNA,

hybridized to a distant downstream region of template strand dCs at

the ColEl origin, is of fundamental importance in forming primer 3'

ends by stabilizing an RNA-DNA hybrid duplex site recognized by

RNase H (Masukata and Tomizawa, 1990).

After priming occurs in the CSB region, DNA replication either

proceeds to completion or terminates prematurely 500 to 1800

nucleotides (depending on the vertebrate species) downstream

(Gillum and Clayton, 1978; Dunon-Bluteau and Brun, 1987), creating

the triple DNA structure known as the Displacement loop (D-loop),

one of the enduring mysteries of mitochondrial molecular biology

(Kasamatsu et al., 1971; Brown and Vinograd, 1974; Clayton, 1982).

The in vivo role of the D-loop is not understood in any system.






10

Bogenhagen and Clayton (1978) studied the kinetics of D-loop DNA

(7S DNA) formation by pulse-chase experiments in Mouse L-cells.

The half-life of pulse-labeled 7S DNA is approximately 70 minutes, a

turnover so rapid that at least 95% of this DNA pool is lost without

acting as primers for H-strand synthesis, suggesting that the primary

role of D-loop DNA is not replication priming.

After successful H-strand synthesis proceeds approximately two-

thirds of the way around the genome, the L-strand origin, 0L is

exposed (Martens and Clayton, 1979). This origin is located in a 30

to 33 base noncoding region within a cluster of five tRNA genes,

forming a stable stem-loop structure highly conserved in most

vertebrates (Martens and Clayton, 1979; Tapper and Clayton, 1981;

Wong and Clayton, 1985a). A primase activity, distinct from

mitochondrial RNA polymerase, has been identified which recognizes

0, and synthesizes 9-12 base RNA primers, which are then used by

DNA polymerase-yfor L-strand synthesis (Wong and Clayton, 1985b).

Both the conserved stem-loop and 5' flanking sequences are

important for L-strand initiation, with five nucleotides at the base of

this stem loop the site of RNA to DNA transition (Hixson et al., 1986).






11

Mitochondrial Transcription


The coding capacity of the mitochondrial genome is relatively

constant among different organisms (except plant mtDNA, which

encodes more proteins); however, genome size, structure, and mode

of expression are variable. Vertebrate transcription is usually

achieved by symmetrical, almost genome-length polycistronic

transcription (Aloni and Attardi, 1971; Murphy et al., 1975) from

two overlapping divergent promoters in the D-loop region, following

by processing to generate unit-length mRNAs. In contrast, yeast

genomes have at least 20 different promoters, each driving the

expression of shorter polycistronic RNAs (Costanzo and Fox, 1990),

with differences in promoter strength and attenuation of

transcription determining transcript levels (Mueller and Getz, 1986).

Most plant genomes also display multiple promoters (Mulligan et al.,

1988; Levings and Brown, 1989), but differential transcription is

achieved by selective amplification of subgenomic size fragments,

resulting in increased mRNA levels as a consequence of increase in

gene dosage (Muise and Hauswirth, 1995).

Vertebrate mitochondria have only two major promoters, the L-

strand promoter (LSP) and the H-strand promoter (HSP) for

transcription of each strand (Montoya et al., 1982; Chang and Clayton,






12

1984). The existence of a third promoter, specific for transcription of

the rRNA genes, has been proposed (Montoya et al., 1983; Chomyn

and Attardi, 1992) on the basis of SI nuclease protection studies. In

humans, the LSP and HSP, the two bi-directional promoters, are in

close proximity but do not overlap. These promoters require two

distinct sequence elements for activity (Chang and Clayton, 1984;

Hixson and Clayton, 1985; Topper and Clayton, 1989), one

surrounding the initiation site and the other located just upstream of

the start site, required for activation by the human transcription

factor, h-mtTFA (Fisher and Clayton, 1985). Bi-directional

transcription initiation is also seen in chicken (L'Abbe er al., 1991)

and Xenopus laevis (Bogenhagen and Yoza, 1986). In Xenopus, there

are two bi-directional promoter elements, each consisting of two

overlapping octamer sequences (analogous to the nonanucleotide

promoter of yeast) located on opposite DNA strands (Bogenhagen et

al., 1986; Bogenhagen and Yoza, 1986; Bogenhagen and Romanelli,

1988).

The search for protein factors which activate mitochondrial

transcription led to the identification and isolation of the human

mitochondrial transcription factor, h-mtTFA (Fisher and Clayton,

1985). Molecular cloning of this protein from human (Parisi and

Clayton, 1991), yeast (Diffley and Stillman, 1991), and Xenopus laevis






13

(Antoschechkin and Bogenhagen, 1995) revealed homology to the

high mobility group (HMG) of proteins. Another protein, called

mtTFB or the specificity factor in yeast mitochondria, has been

isolated and cloned (Schinkel et al., 1987), and has amino acid

homologies to bacterial o factors (Jang and Jaehning, 1991), but

operates by a distinct mechanism (Shadel and Clayton, 1995).

Recently, Antoschechkin and Bogenhagen (1995) isolated a similar

protein from Xenopus laevis mitochondria, the xl-mtTFB. This is the

first report of the existence of such a protein in metazoan

mitochondria. Mitochondrial transcription factor A (mtTFA) appears

to be the primary transcription factor in human mitochondria,

absolutely required from high levels of specific transcription

initiation in vitro. In contrast, sc-mtTFB appears to be the primary

transcription factor in the yeast system, with the sc-mtTFA protein

playing only a minor role in overall transcription initiation efficiency

(Shadel and Clayton, 1993). In Xenopus laevis, xl-mtTFB appears to

be the basal transcription factor, with xl-mtTFA required for

transcriptional activation (Antoschechkin and Bogenhagen, 1995).

Hence, the role of sc-mtTFA is fundamentally different from its

homologs in vertebrates. In vitro binding studies with purified

mtTFA from yeast, human, and Xenopus mitochondria display






14

generalized binding of this protein to any DNA, but phased or

periodic binding to the DNA of mitochondrial regulatory regions

(Diffley and Stillman, 1991; 1992; Fisher et al., 1992; Antoschechkin

and Bogenhagen, 1995). This led to the suggestion that the in vivo

role of this protein might be in packaging of mtDNA regulatory

elements into a specific structural conformation, a suggestion which

has been confirmed for cow (Ghivizzani et al., 1993), human

(Ghivizzani et al., 1994), and Xenopus (this study) mitochondria by in

organello and in vivo footprinting (Madsen et al., 1996; Ammini et

al., 1996).

General processing of the polycistronic transcripts to generate

individual mRNAs is not fully understood yet. The tRNA genes are

interspersed among the H-strand and L-strand encoded protein

genes, with some of them abutting the rRNA or mRNA sequences.

This peculiar arrangement led to the proposal of the tRNA

punctuation model of RNA processing in mitochondria (Ojala et al.,

1981), implying the existence of a very precise processing machinery

which recognizes the tRNA sequences with an RNAse P-like activity,

and the process is believed to occur co-transcriptionally. The

existence of an RNAse P activity in mitochondria of HeLa cells has

been reported (Doersen et al., 1985), and a 5'- and 3'-tRNA





15

processing activity has been identified and characterized in rat liver

mitochondria (Manam and Van Tuyle, 1987).


Regulation of Mitochondrial Gene Expression


Mitochondrial biogenesis requires the coordination of both nuclear-

and mitochondrial-encoded mitochondrial genes. How this

coordination is achieved to meet the dynamic energy needs of the

cell is still one of the central questions in mitochondrial molecular

biology. The aggregate of the experiments done in this field suggests

that there is no one answer. The mode of regulation seems to be as

varied as the plethora of regulatory events involved in mitochondrial

biogenesis. More is known about the regulation of genes encoded on

mtDNA than those on nuclear-DNA. Mitochondrial gene expression is

regulated by both transcriptional and post-transcriptional

mechanisms. The levels of the ribosomal RNAs seem to be controlled

by a transcription termination mechanism in vertebrate

mitochondria, as the two rRNAs have a 15- to 60-fold higher rate of

synthesis than mRNAs (Gelfand and Attardi, 1981). After initiation

of transcription from the HSP, transcription is terminated

downstream of the ribosomal genes by DNA-binding protein-

mediated termination at a tridecamer sequence in the tRNAle" gene

(Christianson and Clayton, 1986; 1988). The protein factor,





16

mitochondrial transcription termination factor (mTERF), has been

isolated from human mitochondria (Kruse et al., 1989; Hess et al.,

1991) and shown to cause sequence-specific termination of

transcription in vitro. Regulation of mRNA levels has been shown to

operate at the level of differential transcription in some cases. In

yeast, the relative strengths of the multiple promoters, which vary

over a 20-fold range, coupled with transcriptional attenuation

(Mueller and Getz, 1986) of the polycistronic units, have been shown

to dictate transcript levels in mitochondrial run-on assays. Thyroid

hormone is a physiological effector of mammalian mitochondrial

biosynthesis, influencing respiration, cytochrome content, protein

synthesis and increasing mitochondrial RNA levels (Kadenbach et al.,

1964). Mutvei et al. (1989) showed by run-on transcription assays

that the increased mitochondrial transcript levels in rat liver during

thyroid hormone treatment were due to increased mitochondrial

transcription. Many central regulatory events of hormone action are

being increasingly traced to their effects on mitochondrial gene

expression by increasing mitochondrial transcript levels. Androgen

treatment of mice (Cornwall et al., 1992), and corticotropin treatment

of bovine cultured cells (Raikhinstein and Hanukoglu, 1993) both

cause the induction of mitochondrial RNAs, as shown by subtractive

hybridization screens, although the exact mechanism of regulation,






17

whether transcriptional or post-transcriptionally regulated RNA

stability, is not known. In a differential screen to isolate cDNAs for

mRNAs whose levels are suppressed by interferon treatment, many

mitochondrial mRNAs were picked up (Lou et al., 1994). These

results were confirmed by Northern blot analyses and were found to

be at the level of transcription rather than stability of these RNAs,

and was dependent on protein synthesis. Similar screens have

traced increased expression of mitochondrial-encoded genes in

skeletal muscle of humans with diabetes mellitus (Antonetti et al.,

1995), despite decreases in mtDNA copy number, suggesting that the

increased mitochondrial gene expression may contribute to the

increase in mitochondrial respiration observed in uncontrolled

diabetes.

Control at the level of RNA stability is clearly seen for the tRNAs

(King and Attardi, 1993). The tRNA genes belong to three distinct

transcription units, which exhibit widely different rates of

transcription. Thus, in HeLa cells, the rDNA transcription unit is

transcribed -25 times more frequently than the mRNA genes

downstream of the mTERF element, while the L-strand transcription

unit is transcribed 10-16 times more frequently than the H-strand

transcription unit (Chomyn and Attardi, 1992; King and Attardi,

1993). Despite these wide differences in transcription rates of the





18

tRNA genes, the steady state levels of the various tRNAs are more or

less similar. Pulse labeling experiments have indicated that this

uniformity may not be due to differences in the metabolic stability of

the mature tRNAs (half-lives longer than 24 h), but may be due to

elimination of excess tRNA transcripts before they enter the mature

pool (Chomyn and Attardi, 1992; King and Attardi, 1993).

Regulation of gene expression can also be pre-transcriptional, at

the level of gene dosage. The number of mtDNA genomes per cell is

variable and has been estimated as >1000 in mouse L-cell fibroblasts

and >8000 in cultured human cells (Bogenhagen and Clayton, 1974;

Robin and Wong, 1988). In some cases, mtDNA is amplified to

extraordinarily high numbers, as in Xenopus laevis oocytes (-108 per

oocyte) due to the large number (107) of organelles per cell (Chase

and Dawid, 1972; Webb and Smith, 1977). Highly oxidative tissues

such as heart muscle contain a significantly greater number of

mtDNA genomes than do the more glycolytic type II skeletal muscles

(Annex and Williams, 1990). Therefore, mtDNA is not just a static

repository of genetic information but is under dynamic control. A

correlation between gene expression and mtDNA copy number has

been reported in a few cases. Williams et al. (1986) studied the

regulation of mitochondrial gene expression during long-term

electrical stimulation of rabbit skeletal muscle, and found the higher





19

oxidative capacity induced by this protocol to also elevate Cyt b

mRNA, rRNA, and mtDNA levels, suggesting that the regulation was

at the level of mtDNA copy number (Williams, 1986). Amplification

of subgenomic fragments as a mechanism for control of

mitochondrial transcript levels has been reported for maize (Muise

and Hauswirth, 1995).


Focus Of Dissertation


The embryogenesis of vertebrate species like Xenopus laevis and

mouse provide invaluable natural model systems for understanding

some of the intricacies of mitochondrial gene regulation. They

illustrate the existence of general control mechanisms for mtDNA

replication and transcription that are not evident from experiments

with undifferentiated cells growing in culture (Attardi and Schatz,

1988). In both species, mtDNA replication occurs during oogenesis

and stops at maturation, whereas transcription of mtDNA is activated

some time after fertilization, long before the resumption of mtDNA

synthesis. In the mouse, the mitochondrial genome is inactive in the

egg and the 2-cell embryo, after which a high rate of mitochondrial

transcription was detected (Piko and Taylor, 1987). The relative

period of basal transcription in the X. laevis system is longer,

allowing more experimental malleability. The other advantages of





20

choosing Xenopus embryos as a model system for this study are

manifold: this vertebrate embryo is accessible in adequate numbers

throughout the year, its biology has been well studied, efficient

synchronous fertilization can be achieved in the laboratory ex vivo, a

considerable amount of molecular studies on mitochondria and

mtDNA metabolism have been done, and the particular features of

mitochondrial biogenesis and nucleic acid metabolism during early

development make it extremely amenable for fulfilling the aims of

this study. The disadvantage of this system is the non-availability of

Xenopus laevis inbred lines, making it essential to utilize eggs and

embryos from a single female frog within one experiment to avoid

sequence heterogeneity.

Oogenesis in Xenopus laevis is generally characterized by the

synthesis and storage of macromolecules like yolk protein (Wahli et

al., 1981), histones (Woodland, 1980), DNA polymerases (Zierler et

al., 1985), and organelles like ribosomes (Bozzoni et al., 1982;

Pierandrei-Amaldi et al., 1982) and mitochondria (Dawid, 1966;

Chase and Dawid, 1972), to enable rapid divisions during early

cleavage of the embryo without fresh biogenesis of these

components. The stockpiling of mitochondria (-107 organelles per

mature oocyte) also necessitates the dramatic amplification of mtDNA

(Webb and Smith, 1977), which accounts for 99 % of the total oocyte






21

DNA. These mitochondria are partitioned into cleaving cells during

early embryogenesis, with production of new mitochondria occurring

only after two days of fertilization (Chase and Dawid, 1972).

Therefore, active mtDNA replication does not occur in the fully

developed oocyte and in the developing embryos until relatively late

in development (Chase and Dawid, 1972; Webb and Smith, 1977).

However, early studies suggest that transcription of mtDNA is

activated soon after fertilization, long before active synthesis of

mtDNA starts, suggesting a decoupling of mitochondrial transcription

and replication (Chase and Dawid, 1972; Young and Zimmerman,

1973; Dawid et al.. 1985). Meziane et al. (1989) reported further

proof for active regulation of mitochondrial gene expression by

performing Northern analyses of mitochondrial transcripts during

early development. They found high levels of transcripts in the

unfertilized eggs and post-neurula stages (after I day post-

fertilization), with 5- to 10-fold lower levels between 6 and 24 hours

of development. Based on these results, they suggested, contrary to

the findings of the earlier groups, that a fertilization induced

shutdown of mitochondrial transcription occurs in X. laevis, with

active transcription resuming only late during embryogenesis.

However, all four groups agree that resumption of active

mitochondrial transcription appears to be decoupled from active





22

mtDNA replication, emphasizing the value of this natural model for

studying the developmental regulation of mitochondrial gene

expression. Insights gained in this system should be more or less

generally applicable to mammalian mitochondria, as the size,

structure and organization of Xenopus mtDNA are extremely similar

to that mammals, especially human (Roe et al., 1985; Dunon-Bluteau

et al., 1985; Cairns and Bogenhagen, 1986). Additional similarities

can be noted in the similarity of organization of the transcriptional

promoters and mode of transcription from the LSP and HSP

(Bogenhagen and Yoza, 1986; Bogenhagen el al., 1986; Bogenhagen

and Romanelli, 1988), presence of the transcription factors mtTFA

and mtTFB (Antoschechkin and Bogenhagen, 1995), and the presence

of the tridecamer mTERF element (with only one base change from

the human equivalent) at the 16S rRNA/tRNAL'" boundary.

The main focus of this research project was to utilize this natural

model to gain further insights into the molecular mechanisms of the

developmental control of mitochondrial gene expression. However, it

was initially essential to clarify the confusion in the literature

regarding the transcriptional status of the mitochondria during

fertilization and early embryogenesis. To achieve this aim, the steady

state levels of mitochondrial transcripts during early development

were measured by Northern analyses and the value of this system





23

re-established. Then the molecular mechanisms responsible for

these rapid changes in mitochondrial transcript levels were

investigated by two main approaches, mitochondrial run-on

transcription and in vivo footprinting of mtDNA during early

development of X. Iaevis. The mitochondrial run-on transcription

assay was developed based on modifications of the protocol of Gaines

et al. (1987) to measure the transcription rates of mitochondria

isolated from unfertilized eggs and distinct embryonic stages (with

basal and enhanced steady state levels). These assays were aimed at

determining whether the primary determinant of the differential

transcript levels was at a transcriptional or post-transcriptional level.

It is known that changes in mtDNA are not the main regulatory

event, as the levels of mtDNA do not change during the first two days

of development (Chase and Dawid, 1972).

The second approach was to conduct an extensive in vivo DMS

(dimethyl sulfate) footprint analysis of the known regulatory

elements of transcription during the developmental stages of basal

and activated mitochondrial transcription, so that changes in protein-

DNA interactions could be correlated with transcriptional regulation.

In order to perform this comparative analysis, it was initially

essential to establish a baseline in vivo footprint pattern for Xenopus

mtDNA, as no previous studies are available for comparison. This





24

was done by performing in organello DMS footprinting of ovary

mtDNA. An in vivo DMS footprinting protocol based on DMS

treatment of intact eggs and developing embryos was then

developed (Ammini et al., 1996), based on modifications of the

somatic tissue in i" ;..'.* protocols of Ghivizzani et al. (1993). The

regions footprinted were the promoter region, the region between

the promoters and origin of H-strand replication, and the mTERF

element. In addition, band shift assays were designed to look for the

presence of the mTERF protein in Xenopus mitochondria, as its

presence in X. laevis mitochondria is suspected but not confirmed.

The results of these various experiments, along with my

interpretation of the results and predictions about possible

mechanisms for the developmental regulation of mitochondrial gene

expression, are presented in the following chapters.












CHAPTER 2
MATERIALS AND METHODS


Frogs


Sexually mature male and female Xenopus laevis frogs were

purchased from Xenopus I, Ann Arbor, Michigan, and maintained in

separate tanks using standard procedures (Wu and Gerhart, 1991).


DNA Primers


Primers were synthesized by the DNA synthesis core, ICBR,

University of Florida, and gel purified where necessary using

standard procedures (Maniatis et al., 1989).


Laboratory Fertilization of Eggs


Fertilized embryos were obtained using the in vitro fertilization

procedure of Hollinger and Corton (1980). The female frogs were

injected with 700-1000 U HCG (Sigma CG-5; reconstituted in water at

1000 U/ml and stored at 40C for 3-4 weeks) into the peritoneal

cavity 10 hrs before eggs were required and placed into white

buckets with about 2 liters of Ix OR-2 (10x OR-2 is 825 mM NaCI, 25

25





26

mM KCI, 10 mM CaCl2, 10 mM MgCl2, 10 mM Na2HPO4, 50 mM

HEPES, pH to 7.8 with NaOH. The divalent and monovalent ions were

made up as separate stocks and diluted just before use). After

spontaneous egg laying begins (usually 12-18 hrs later), eggs were

gently squeezed from the female onto 100 x15 mm petridishes

containing -20 ml of 1 x OR-2 in batches of 400-500 eggs. It was

found that eggs stored in Ix OR-2 could be successfully fertilized

within 4 hrs of laying. Therefore, fertilization could be done in a

staggered fashion in batches of 500-1000 over the course of 8 hrs,

during which time the frog was allowed to lay its eggs at its own

pace in fresh lx OR-2 buffer. In this way, a high yield of fertilized

embryos could be obtained (3000-6000 embryos compared to only

<1500 with the quick squeeze protocol).

Sperm was obtained by dissecting the testes from a male frog

(anesthetized on ice) and gently macerating it in 1.5 X OR-2 to

release the sperm. Sperm motility was observed under a light

microscope after diluting the saline to 0.5 X OR-2. The eggs were

then drained completely of excess IX OR-2, 5-10 drops of the sperm

suspension added to each dish of eggs, and 30-50 ml of IX F-l was

quickly added to dilute the sperm suspension (O1x F-l is 412.5 mM

NaCI, 12.5 mM KCI, 2.5 mM CaCl2, 0.625mM MgCl2, 5 mM Na2HP04,






27

25 mM HEPES, pH to 7.8 with NaOH. The divalent and monovalent

ions were made up as separate stocks and diluted just before use).

After re-placing the lid, the petridishes were gently swirled to

evenly distribute the sperm and also to spread out the eggs into a

monolayer, and then left undisturbed. Fertilization can be monitored

within 30 minutes, as fertilized eggs undergo the cortical reaction

and the animal pole rotates to top. Unfertilized and damaged eggs

can then be culled. During the first 8 hours of growth, the Ix F-I

buffer was replaced with 0.2x F-I buffer in the petridishes.

Afterwards, the whole petridish was immersed in a large dish filled

with 2-3 liters of 0.2x F-l buffer. This step is very critical for

obtaining good survival of the embryos. The embryos were allowed

to hatch in this volume of buffer without further changes.


De-jellying of Eggs and Embryos


Eggs and embryos were de-jellied by gently swirling in 2%

Cysteine solution [Cysteine, free base; made using Ix OR-2 within 1

hour of use and titrated to pH 7.8 with NaOH (4.5 ml 1 M NaOH per

100 ml)] for 3-4 minutes, followed by repeated washing in Ix OR-2

(for eggs) or 0.2x F-l (for embryos) to remove traces of the cysteine

solution. The de-jellied eggs or embryos were then gently placed

into a fresh petridish of Ix OR-2 (eggs) or 0.2x F-l (embryos) for





28

sorting and counting using a broad-tipped 10 ml pipet. The eggs and

early embryos are especially fragile after de-jellying and should be

handled only using broad-tipped pipets.


Isolation of Mitochondria from Eggs and Embryos


Mitochondria were purified from eggs and embryos with

modifications of the protocol used by McKee et al. (1990) for the

isolation of heart mitochondria. This protocol was adopted both due

its ease of application and the replacement of EDTA with EGTA, which

is especially important in run-on and band-shift experiments, where

chelation of Mg"' must be avoided.

De-jellied eggs and embryos were floated in MSE buffer (40 mis

for 200-300 eggs or embryos) and homogenized in a glass dounce

homegenizer with a type B teflon pestle with 7-10 strokes. The

homogenate was spun at 1,500 x g for 10 min. in a JA20 rotor; the

supernatant was re-spun as above and the second supernatant spun

at -9,000 x g for 10 min in a JA20 rotor; the pellet was collected,

resuspended in 1 ml of MSE and spun in an Eppendorf tube at 9,000

x g for 2 min. This step was repeated 2-3 times until the

supernatant cleared up to remove traces of fat. The final pellet was

used as such in the experiments involving run-on transcription or

mitochondrial DNA isolation in the footprinting experiments. In






29

experiments where further purification of mitochondria was

necessary (e.g. ATP measurement studies), mitochondria were

purified in a scaled-down sucrose gradient in a TLS-55 rotor. This

gradient was pre-formed with 0.8 ml each of 51% sucrose and 34%

sucrose, and the mitochondrial pellet was resuspended in 20%

sucrose and layered on top in a volume of 0.6 ml. The gradient was

spun at 37,000 rpm in a TLS-55 rotor in a TL100 centrifuge

(Beckman) for 10 min. The mitochondrial band at the 34%/51%

sucrose interface was collected, diluted with an equal volume of MSE

buffer and re-pelleted in an Eppendorf centrifuge at 9,000 x g for 2

min.

Isolation of Mitochondria from Ovary Tissue


Isolation of mitochondria from ovaries of adult frogs was done

according to the protocol used for purification of bovine brain

mitochondria by Hehman and Hauswirth (1992), with the

modifications described below. Whole ovaries were dissected from

anesthetized female frogs and dropped into chilled MSB-Ca-' buffer

(about 100 ml/ovary). The tissue was minced and, homegenized in a

glass homegenizer by three strokes of a motor driven, teflon coated

pestle. Then EDTA was added to a final concentration of 20mM. This

was then passed through three layers of cheese cloth and centrifuged






30

in a JAlO rotor at 1,500 x g for 10 min; the supernatant was collected

and recentrifuged as above. The supernatant from the second spin

was centrifuged in a JAO1 rotor at -17,700 x g for 45 min. The

pellets were gently resuspended in 10 ml of MSB-EDTA per 40 ml of

discarded supernatant and re-spun at -23,000 x g in a JA17 rotor for

20 min. Each pellet was brought up in 10 ml of 20% sucrose-TE

buffer and layered on top of a sucrose step gradient (10 mis each of

51% and 34% sucrose-TE). The gradients were then spun in a

Beckman SW27 rotor at 22,000 rpm for 20 min. The mitochondrial

band at the 34%/51% sucrose interface was carefully collected with a

Pasteur pipet, diluted with an equal volume of MSB-EDTA and

pelleted at -23,000 x g in a JA17 rotor. The pellets were then

processed for in organello methylation or quick-frozen in a dry

ice/ethanol bath and stored at -700C.


Dimethyl Sulfate Footprinting of Mitochondrial DNA


Oligonucleotides


WH352. 21 base oligonucleotide ACT CAA ACC.TCC ACT ATT

GAC. Corresponds to mitochondrial L-strand sequence 17498-17518

(Roe elt al.. 1985).






31

WH353. 21 base oligonucleotide TAA CCA TAA AAT ATG CTA

AGT. Corresponds to mitochondrial H-strand sequence 1725-1705 in

the sequence correction by Dunon-Bluteau et al., (1985), and

incorporating the changes listed in Cairns and Bogenhagen (1986).

WH354. 21 base oligonucleotide CCC ATG TTA TAC ATT TTT

GTA. Corresponds to mitochondrial L-strand sequence 2008-2028

(Roe et al., 1985).

WH355. 21 base oligonucleotide CCT TTA TGC TTT CGG AGC

TTT. Corresponds to H-strand sequence 3832-3812 (Dunon-Bluteau

et al., 1985).

WH356. 21 base oligonucleotide AAT CTG TTG TGA CCT TAA

CCT. Corresponds to L-strand sequence 3073-3093 (Dunon-Bluteau

et al.. 1985).

WH357. 21 base oligonucleotide CTT CAT GTT TTT TTT TTT TCT.

Corresponds to L-strand sequence 3187-3207 (Dunon-Bluteau et al.,

1985).

WH358. 21 base oligonucleotide GTG GAT AAT AAT AGG AAA

GCT. Corresponds to H-strand sequence 3387-3367 (Dunon-Bluteau

et al., 1985).

WH359. 21 base oligonucleotide AGG GCT AGC TAG CGT GGC

AGA. Corresponds to L-strand sequence 4717-4737 (Roe et al.,

1985).






32

WH360. 21 base oligonucleotide TGG GGC CTT TAC GGT GTT

GTA. Corresponds to H-strand 4925-4905 (Roe et al., 1985).


Plasmid Clones


The three clones described below were used to make strand-

specific riboprobes for probing Southern blots in the in organello

footprinting experiments. All the clones were constructed as

described below except where indicated. The DNA of the

corresponding region was PCR amplified with the indicated primers.

For this, the primers were first 5'-end phosphorylated with g-'"P

using T4 polynucleotide kinase (US Biochemicals) according to

manufacturer's instructions. The kinased primers were then used to

amplify the mtDNA fragments with either Taq DNA polymerase (BRL)

or Pfu DNA polymerase (Stratagene). Then the PCR products were

gel purified (where necessary) and ligated with HincII-linearized pBS

(+/-) phagemid (Stratagene) overnight at room temperature. The

linearized vector was 5'-end dephosphorylated with Shrimp Alkaline

Phosphatase (USB), deactivated at 650C for 15 min. and purified after

organic extractions. Different dilutions of the ligations were used to

transform DH5a cells (BRL) according to the supplier's protocols and

plated on LB plates containing ampicillin, IPTG, and X-gal (Maniatis

et al., 1989) White colonies were screened for the presence of the






33

insert as a single copy by digesting with Hind III and BamHI (BRL)

which flank the Hinci site. Positive clones were amplified in a large

scale culture and DNA isolated and purified by either CsCI gradients

(Maniatis et al., 1989) or Wizard maxi-prep columns (Promega

Corporation).

pBSLSP. This plasmid (3405 bp) contains the frog mitochondrial

sequence 3631-3832 in the corrected sequence of Dunon-Bluteau et

al. (1985). This stretch of DNA (amplified with primers WH354 &

WH355) is cloned in the reverse orientation and contains all the

promoter elements in the D-loop region (including LSP 1 & 2 and HSP

1 & 2). Strand-specific riboprobes made from this clone were used

to probe in organello footprint Southern blots of the promoter region

using the Avall site at 3836 as reference.

pBSCSB. This plasmid (3404 bp) contains the frog mitochondrial

sequence 3187-3387 in the corrected sequence of Dunon-Bluteau et

al. (1985). This stretch of DNA (amplified with primers WH357 &

WH 358) is cloned in the forward orientation and contains the two D-

loop 5' start sites, and the conserved sequence box 1 (CSB 1).

Strand-specific riboprobes made from this clone were used to probe

in organello footprint Southern blots of the CSB region using the

HaeIII site at 3187 as reference.





34

pBSmTERF. This plasmid (3412 bp) contains the frog

mitochondrial sequence 4717-4925 (Roe et al., 1985). This stretch of

DNA (amplified with primers WH359 & WH360) is cloned in the

reverse orientation and contains the terminal end of the 16S rRNA

gene, the entire tRNAI- gene carrying the mTERF element, and the

initial sequences of the NDI gene. Strand-specific riboprobes made

from this clone were used to probe in organello footprint Southern

blots of the mTERF region using the SspI site at 4927 as reference.


Riboprobes


The DNA clones were linearized with HindIII for transcription

with T7 RNA polymerase, and with BamHI for transcription with T3

RNA polymerase. For transcription reactions, 1 ug of linearized DNA

was used in a standard analytical scale transcription reaction in the

supplier's buffer (Stratagene), with 100 uCi of a32P-UTP

(3,00OCi/mmol). The whole transcription reaction was loaded on a

4% non-denaturing polyacrylamide gel and the full-length RNA

product band located with a very short exposure (30 sec) on X-ray

film. This band was cut out, macerated in a 15 ml centrifuge tube

and added directly to the hybridization solution.





35

In Organello Methylation of Mitochondrial DNA


After the final centrifugation step, the pelleted mitochondria (-2

ml) were resuspended in an equal volume of phosphate-buffered

saline (PBS: 140 mM NaCI, 2.5 mM KC1, 10 mM Na2HP04, 1.5 mM

KH2PO4, pH 7.5) and 400 ul aliquots were transferred into Eppendorf

tubes and placed at room temperature. Then Dimethyl Sulfate (DMS)

was added in concentrations of 0.125%, 0.25% and 0.5%, vortexed

briefly to mix well and incubated at room temperature for 3 minutes.

Then 1 ml of ice-cold PBS was added to dilute out the DMS and

mitochondria pelleted immediately by centrifuging for 1 min at

10,000 x g. PBS washes were repeated twice to remove all traces of

DMS. Then 400 ul of mitochondrial lysis buffer was added and

mtDNA extracted as described below. The methylation reaction of

0.125% for 3 minutes was found to give satisfactory in organello

reactions which matched the reaction of 0.125% for 1.5 minutes using

purified mitochondrial DNA.


In Vivo Methylation of Eggs and Embryos


De-i.1i..J eggs or embryos Nwere drained of buffer, refloated in

PBS containing 0.125 % DMS and gently swirled for 3 minutes. The

DMS solution was rapidly emptied after 3 minutes and the eggs or





36

embryos repeatedly washed with buffer to remove traces of DMS.

Mitochondria are then isolated by differential centrifugation as

described above, except that banding on sucrose gradients was not

performed. The mitochondrial pellets were quick-frozen and stored

at -800C, or mtDNA was extracted immediately as described below.

Typically, one mitochondrial pellet from a batch of 400 eggs or

embryos yielded sufficient mtDNA to perform 10-12 primer

extension reactions. Similar patterns of methylation in control DNA

were achieved by 0.125% DMS treatment for 30-45 sec.


Purification of Mitochondrial DNA for Footprinting


The mitochondrial pellet was resuspended in 400 ul of lysis

buffer and SDS added to 1%. The mitochondrial lysate was then

extracted twice each with phenol, phenol:chloroform:iso amyl alcohol

(25:24:1), and chloroform and then ethanol precipitated with 2.5

volumes ethanol. For in c ... '. footprinting using Southern

blotting, the mtDNA samples were digested with restriction enzymes

and a 10% aliquot checked on a 1% agarose gel for complete reaction.

The control DNA samples were first methylated as described below

and then used for restriction digestion in order to avoid artifactual

methylation patterns arising on linearized DNA. Restriction digestion

of the DNA samples was not performed in the primer extension






37

protocol as the sequencing reference is provided by the site of

hybridization of the primer. On an average, about 0.8 to 1 ug mtDNA

is recovered from 300-400 eggs, assuming -4 ng mtDNA/egg (Chase

and Dawid, 1972; Webb and Smith, 1977).


In Vitro Methylation of Purified Mitochondrial DNA


Mitochondrial DNA isolated from control mitochondrial pellets

(untreated with DMS either in -* .. lih.' or in vivo) was resuspended

in 200 ul of 50 mM sodium cacodylate (pH 7.0). DMS titrations were

then done for short reaction times (30 sec. to 1.5 min), as necessary

to match the respective extent of in organello or in vivo reaction.

Then 50 ul of chilled DMS stop buffer (1.5 M sodium acetate (pH 7.4),

1 M 2-mercaptoethanol) was added and the mtDNA samples ethanol

precipitated.


Piperidine Cleavage of Methylated DNA


To cleave the mtDNA at methylated residues, the DNA pellets

were resuspended in 200 ul of 10% piperidine and incubated at 900C

for 30 min. Then the DNA was again ethanol precipitated in the

presence of 2.5 M ammonium acetate, resuspended in 200 ul of

water and dried in a speed-vac to remove all traces of piperidine.






38

Samples were resuspended in 20 ul of formamide loading dye (for

Southern blotting) or TE (for primer extension reactions).


Southern Blot Detection of In Organello Footprints


As a preliminary experiment to optimize loading intensities, equal

amounts of test and control mtDNA samples from above (1/20th to

1/10th of each sample) were loaded into adjacent lanes of a 6%

polyacrylamide, 7.5M urea sequencing gel and electrophoresed under

standard sequencing gel conditions. Based on the hybridization

signals on the Southern blots of these gels, loading volumes for each

stage were decided by a combination of visual assessment and

Phosphorimager (Molecular Dynamics) quantitation.

DNA transfer to membrane was by vacuum blotting using the

method of Lopez et al. (1993). Briefly, following electrophoresis, a

precut piece of nylon transfer membrane (Hybond-N+ from

Amersham, Chicago. IL) was placed directly on the gel surface and

allowed to draw the buffer for 5 min. The sandwich was then placed

in a standard gel-drying apparatus, membrane side down, on a single

sheet of Whatman 3mm paper (Whatman Paper Ltd., Clifton, NJ),

covered with plastic wrap, and the DNA transferred to the membrane

by applying vacuum (without heat) for 45 min to 1 hour. After

transfer, DNA was crosslinked to the membrane on a transilluminator





39

at full power for 3-5 minutes. Pre-hybridization and hybridization

were done using the protocol of Church and Gilbert (1984). To

prehybridize the membrane, 30 ml of hybridization solution (7% SDS,

1% BSA, ImM EDTA, and 0.5 M sodium phosphate buffer (0.5 M with

respect to Na', pH 7.2) was added to the rolled-up membrane in a

rolling hybridization bottle (Robbins Scientific) and placed in a

hybridization oven (Biometra, Tampa, FL) for 30 min at 650C. After

prehybridization, the solution was replaced with 10 ml of

hybridization solution containing the crushed polyacrylamide gel

slice containing the labeled probe, and hybridized overnight at 650C.

The hybridized blots were washed three times with 100 ml of wash

buffer (40 mM sodium phosphate buffer, pH 7.2, ImM EDTA, 1% SDS)

at 600C for 30 min each in the hybridization bottles, followed by a

20 min wash in 500 ml of buffer in a Tupperware tray, and exposed

to X-ray film.


Linear Amplification of Footprints by Primer Extension


The most important factor for obtaining accurate and

reproducible sequences is optimal selection of genomic primers. In

theory, a primer can be as close as 10-30 nucleotides to the domain

of interest. However, in practice, I have found that primers located





40

about 100 nucleotides away from the region to be analyzed work

best in providing reproducible genomic ladders which are essentially

free of low molecular weight background. In general, primers which

fall within 70-150 bp of the target region were used successfully.

Some of the primers used in the primer extension footprinting

experiments have been described above for the creation of the three

plasmid clones used for production of riboprobes. Additional

primers which were used in the primer extension footprinting

experiments are described below.

WH 476. 21 base oligonucleotide GGT CAG TTT CTA TCT ATG

AAG. Corresponds to the L-strand sequence 4582-4602 (Roe et al.,

1985).

WH 479. 21 base oligonucleotide TTC TAT ACA AGG CTA ACA

GTC. Corresponds to the H-strand sequence 3582-3602 (Dunon-

Bluteau et al., 1985).

WH 480. 21 base oligonucleotide CGC TAA ACC CCC CTA CCC CCC.

Corresponds to the L-strand sequence 3425-3445 (Dunon-Bluteau et

al., 1985).

End labeling of oligonucleotide primers. Gel purified

primers were 5' end-labeled to a high specific activity with g- 2P

ATP (ICN, 7000 Ci/millimole) and a T4 Polynucleotide Kinase kit (USB

Corporation-Amersham Life Science) in a standard 50 ul reaction.






41

Free label was removed by two sequential ethanol precipitations in

the presence of 2.5 M ammonium acetate.

Optimization of Primer extension reaction conditions.

Preliminary experiments indicated that about one tenth of the

piperidine cleaved DNA sample (about 50-100 ng) was sufficient for

obtaining a useable primer extension signal. Increasing primer

concentration beyond 1 picomole per reaction did not further

enhance signal intensity and led to higher background. The results

of an experiment designed to determine the number of linear

amplification cycles necessary to obtain reproducible results are

shown in Figure 2-1. Amplifying for 10-15 cycles provided adequate

signal intensity, whereas non-specific bands began to appear with a

higher number of cycles (compare lanes 3 and 4 with lanes 6 and 7).

It is important to note that some of the non-specific bands appearing

with higher cycles are 1-2 nucleotides longer than the authentic

bands, suggesting that the Taq DNA polymerase was adding non-

templated nucleotides to the 3' end of duplex DNA (Clark, 1988)

made in earlier cycles. Thermostable polymerases having a 3'-5'

proofreading ability, such as Vent DNA polymerase or Pfu DNA

polymerase. have been used to alleviate this problem in ligation-

mediated PCR (LMPCR) footprinting protocols (Garrity and Wold,



















Figure 2-1. Experiment to determine the optimal number
of cycles of linear amplification. Mitochondrial DNA from X.
laevis ovaries which had been processed for in organello footprinting
(as described for Figures 4-2 and 4-3) was used as a template for
primer extension with "'P-end labeled primer WH 480 as described
in Materials and Methods. The samples were denatured at 940C for 5
min, followed by independent cycling of each reaction for 1, 5, 10,
15, 20, 30, and 40 cycles (lanes 1-7) with the following profile: 940C
for 1 min, 60'C for 1 min, and 720C for 5 min. The reactions were
then phenol extracted, ethanol precipitated, run on 6% sequencing
gels and autoradiographed. Note the appearance of additional bands
which are longer by 1-2 nucleotide, especially in the samples cycled
for 30 and 40 cycles (lanes 6 and 7). Cycling for 15 cycles was found
to be optimal with respect to both signal intensity and fidelity of
reproduction of the footprint ladder.













PCR cycles (#) 1 5 10 15 20 3 0 4 u
Lane (#) 1 2 3 4 5 6 7


C -- 2127


LSP2


- S gLf- 2039

--. --1998




*--1938


N-- VW-1901
l-1895
M11-;I s

















Figure 2-2. Experiment comparing 3 different thermostable
DNA polymerases for accurate amplification of the footprint
ladder. The performance of two thermostable polymerases, Vent
DNA polymerase (NEB) and Pfu DNA polymerase (Stratagene), which
possess 3'-5' proofreading ability in addition to the polymerase
activity, were compared with Taq DNA polymerase (GIBCO BRL),
which does not possess the 3'-5' proofreading function. DNA
template and primer were the same as that described for Figure 2-1.
Each 100-ul reaction was assembled in the manufacturer's
recommended buffer, with 2 mM MgClI, 1 pMol 'P-end labeled
primer, the indicated dNTP concentrations, and 2-2.5 units of the
respective enzyme. The samples were denatured at 940C for 5 min,
then cycled for 5 cycles with the profile: 940C for 2 min, 650c for I
min. 720C for 5 min, followed by 10 cycles with the profile: 94'C for
2 min. 680 for 30 sec, 720C for 3 min. The reactions were then
phenol extracted, ethanol precipitated, run on 6% sequencing gels
and autoradiographed. Note that only Taq DNA polymerase gives
reproducible banding patterns at a wide range of dNTP
concentrations.











DNA polymerase Vent Taq Pfui
dNTP (uM) 20 100 200 20 0 40 60 0 100 100


*- PSB-------r-----2085

.2 0 8 5



LSP2 E...-

-- S ---M---2026


-1998



-1977


h 3lfH-B "-~l -1946

gab -L B-1938

V-U







-1914


CsRBL


- I


-I


1 J 4 -') 0 / N 1 ) 0


----


rc-... OY Cikr ---- -- c- ---
Cr
L;fL~LI

t2' alc--- "i-
ti' ~~llrrrr~r






46

1992). The results of an experiment with Pfu DNA polymerase and

Vent DNA polymerase in a PEF reaction using the manufacturer's

suggested buffers for primer extension are shown in Figure 2-2.

Neither polymerase primer extended accurately in this system,

compared to the reproducible pattern produced by Taq DNA

polymerase at a wide range of dNTP concentrations (compare lanes

4-8 with lanes 1-3 and lanes 9 & 10). The manufacturers suggest

lowering the annealing temperature of primers for these

polymerases due to the lower ionic ircn-iJh of the buffers. However,

these and a variety of other conditions were tried with these

polymerases, but without success (data not shown). It remains

probable that these polymerases could be adapted to this system, but

only after extensive optimization of buffer and PCR conditions, as has

been noted for Vent DNA polymerase in the LMPCR protocols (Garrity

and Wold, 1992; Hornstra and Yang, 1993). However, the need for a

reliable thermostable polymerase which requires minimal buffer

optimizations is ably fulfilled by Taq DNA polymerase, and I have

used this polymerase in all the other experiments described. For Taq

DNA polymerase, a final dNTP concentration of 20 uM was chosen for

specificity and reproducibility (Figure 2-2, lanes 4-8 and data not

shown). Lowering dNTP concentration below this level sacrificed

both fidelity and signal intensity. Most current protocols for primer






47

extension up to 300 or 400 bases recommend extension times of 10

minutes to allow completion of all initiated primers (Garrity and

Wold, 1992). However, an extension time of 3 minutes was found to

be sufficient for a number of primer-template combinations without

compromising fidelity or sensitivity. Based on these considerations, a

typical PEF reaction consisted of:

50-100 ng piperidine cleaved DNA

1 picomole "2P-5' end-labeled primer

IX Taq polymerase reaction buffer (BRL)

2 mM MgC1,

20 uM each of dATP, TTP, dGTP, dCTP

2.5 U thermostable DNA polymerase (BRL)

The reaction was assembled on ice and cycled in a thermal cycler

as follows : denaturing at 940C for 5 minutes, followed by 10-15

cycles with the profile: denaturing at 94 oC for 1 minute, annealing at

45-650C (based on primer T,) for 1 minute, and primer extension at

720C for 3 minutes. The tubes were then immediately chilled on ice,

10 ug of carrier tRNA added, and ethanol precipitated in the

presence of 2.5 M ammonium acetate. An organic extraction prior to

DNA precipitation was usually not necessary. The precipitate was

resuspended in 5 ul of sequencing dye, and 1.5 ul of each reaction





48

resolved in a 6% polyacrylamide-7 M urea gel as a preliminary test

to optimize signal intensities. Using visual inspection or

Phosphorimager quantitation, the signals in each sample were

equalized in a final gel and autoradiographed to obtain the footprint

pattern. Autoradiography at room temperature for 12-48 hours was

usually sufficient to generate adequate signal intensity.


Steady State RNA Analysis


Oligonucleotides


WH 513. 20 base oligonucleotide GTA TAG GCG ATA GAA CAA

TC. Corresponds to the L-strand sequence 3174-3194 in the 16S

gene (Roe et al., 1985).

WH 514. 21 base oligonucleotide CTA CAT TGA TAT AGA GAG

AGG. Corresponds to the H-strand sequence 4655-4675 in the 16S

gene.

WH 515. 20 base oligonucleotide CTC CCA CCA TAT TGA CTT CG.

Corresponds to the L-strand sequence 15680-15699 in the ND6 gene.

WH 516. 20 base oligonucleotide CCT TCT CCT TTT TAT GCT GC.

Corresponds to the H-strand sequence 16092-16110 in the ND6 gene.






49

WH 529. 21 base oligonucleotide CAG CAG ACA CAT CTA TAG

CCT. Corresponds the L-strand sequence 16418-16438 in the Cyt b

gene.

WH 530. 20 base oligonucleotide GTG TAT CTG CAA CTA GGG

CT. Corresponds to the H-strand sequence 17226-17245 in the Cyt b

gene

WH 534. 20 base oligonucleotide ACT TCA CTT CCA CGA CCA

TA. Corresponds to the L-strand sequence 9168-9187 in the Cox II

gene.

WH 535. 20 base oligonucleotide CTA GTA TTG ATG AAG ATC

AG. Corresponds to the H-strand sequence 9771-9790 in the Cox II

gene.

WH 536. 20 base oligonucleotide GAT CCT CCT AGC AGT AGC

AT. Corresponds to the L-strand sequence 4849-4868 in the NDI

gene.

WH 537. 20 base oligonucleotide GTT ATG GCT AGT GTG ATT

GG. Corresponds to the H-strand sequence 5696-5715 in the NDI

gene.

WH 538. 21 base oligonucleotide CTC ACT TCT CAA ACT ACC

TTA. Corresponds to the L-strand sequence 12409-12429 in the ND4

gene.





50

WH 539. 20 base oligonucleotide TGT GTT CTC GGG TAT GTG TA.

Corresponds to the H-strand sequence 13553-13572 in the ND4 gene.

WH 581. 38 base oligonucleotide GTT GGT TGG TTT CGA GGC

CAT GAT TAC TGT TGC CAA TC. Corresponds to the H-strand

sequence 10359-10396 in the ATPase 6 gene. This oligonucleotide

was 5'-end labeled and used as a hybridization probe for detecting

ATPase 6 mRNA in Northern blots.

WH 583. 36 base oligonucleotide GGC GGG TTT GGC AAG AAG

TGG TGA GGT TTA GCG AGG. Corresponds to the H-strand sequence

2634-2669 in the 12S rRNA gene. This oligonucleotide was 5'-end

labeled and used as a hybridization probe for detecting 12 S rRNA in

Northern blots.


Plasmid Clones


The cloning strategy was identical to the one used for the

construction of the footprinting clones described earlier. All the

clones described below were derived from pBS (+/-) phagemid

(Stratagene), with the PCR-amplified gene-specific fragments cloned

into the HinclI site.

pBS16S. This plasmid (4.7 kb) contains the 16 S rRNA sequence

information from 3174-4675 (amplified with oligos WH513 & WH

514). The insert is cloned in the reverse orientation. Therefore, T3






51

RNA polymerase makes the H-strand sense of cRNA whereas T7 RNA

polymerase makes the L-strand sense.

pBSND1. This plasmid (4.066 kb) contains the ND1 gene

sequence information from 4849-5715 (amplified with oligos WH536

& WH537). The insert is cloned in the reverse orientation.

Therefore, the strand sense of the cRNAs are same as above.

pBSCOX II. This plasmid (3.822 kb) contains the Cox II gene

sequence information from 9168-9790 (amplified with oligos WH534

& WH535). The insert is cloned in the reverse orientation.

Therefore, the strand sense of the cRNAs are same as that of pBS16S.

pBSND4. This plasmid (4.362 kb) contains the ND4 gene

sequence information from 12409-13571 (amplified with oligos

WH538 & WH539). The insert is cloned in the forward orientation.

Therefore, T3 RNA polymerase makes the L-strand sense of cRNA

whereas T7 RNA polymerase makes the H-strand sense.

pBSCYTb. This plasmid (4.028 kb) contains the Cytb gene

sequence information from 16417-17245 (amplified by oligos

WH529 & WH530). The insert is cloned in the forward orientation.

Therefore, the strand sense of the cRNAs are same as that of pBSND4.

pBSND6. This plasmid (3.63 kb) contains the ND6 gene

sequence information from 16110-15680 (amplified with oligos WH

516 & WH 515). The insert is cloned in the forward orientation with







52

reference to the linear sequence, but cloned in the reverse

orientation with reference to the gene. Since ND6 is transcribed by

the L-strand promoter, the T3 promoter is still utilized for

synthesizing an antisense probe for the ND6 transcript.


Total RNA Extraction from Eggs and Embryos


Batches of 100 eggs or staged embryos were frozen at -700C and

used for total RNA extraction. The frozen eggs or embryos were

thawed in 10 ml of lysis buffer (4 M Guanidine thiocyanate, 100 mM

sodium acetate, pH 5.0, 5 mM EDTA) and incubated for 20 minutes at

room temperature to facilitate dissociation of nucleoprotein

complexes. The homogenate was then extracted 3-4 times with

phenol:chloroform:isoamyl alcohol (25:24:1, pH 7.0) in 50 ml

centrifuge tubes (Corning or Fisher) to remove pigment, excess

protein and cell debris. Phase separation was achieved by spinning

in a JA-10 rotor at 3,000 x g. The clarified supernatant was then

loaded as a 4.0 ml CsCl solution in SW 41 polyallomer ultracentrifuge

tubes and spun at 38,000 rpm for 24 hrs at 40C. After the spin, the

liquid phase was removed, the clear RNA pellet at the bottom of the

tube rinsed once with 80% ethanol, air-dried and resuspended in TES

buffer (10mM Tris, pH7.4, 1 mM EDTA, 0.1% SDS). The RNA






53

concentrations were then measured spectroscopically and their

A260/Ao0 ratios used as a measure of purity. Typically, ratios more

than 1.9 were obtained. If not, the samples were phenol extracted

again and ethanol precipitated.

As an alternative protocol, TriZol reagent (BRL) was also tried.

Though RNA free of DNA was obtained, abundant amounts of

polysaccharides always co-purified with the RNA, in spite of

repeated phenol extractions and ethanol precipitations. The modified

method reported by Chomczynski and Mackey (1995) for the

removal of polysaccharides from RNA was tried, but polysaccharides

were not removed adequately. Since polysaccharides interfere with

spectroscopic readings, the former method was followed in all the

experiments reported for total RNA levels.


Isolation of Total Mitochondrial Nucleic Acids


For isolating total mitochondrial RNA, crude mitochondrial pellets

were lyzed in TriZol reagent (BRL) and RNA extracted according to

the manufacturer's protocols. Co-purification of polysaccharides with

the RNA preparation was not a problem when isolated mitochondria

were used.

For experiments where the steady state levels of mitochondrial

RNA were normalized to mitochondrial DNA, it was essential to






54

isolate both DNA and RNA in the same solution. For this, a

modification of the protocol of Nelson and Krawetz (1992) by

guanidine thiocyanate/isobutyl alcohol fractionation was tried, in

which the mitochondria were lysed in 4M guanidine thiocyanate,

extracted with phenol:chloroform:isoamyl alcohol (25:24:1, pH 7.0),

and the nucleic acids precipitated with isobutyl alcohol. Although

this method was used successfully, the yields of nucleic acids were

very poor. An alternative, more successful protocol was identical to

the protocol for isolation of mtDNA detailed in the footprinting

section, except that pH 7.0 was maintained in all solutions to

facilitate the simultaneous isolation of RNA and DNA.


Northern Analysis of Mitochondrial RNA


The RNA samples were denatured in 50% formamide, Ix MOPS

buffer (20 mM MOPS, pH 7.0, 5 mM sodium acetate, 0.1 mM EDTA),

0.22 M formaldehyde, 40 ug/ml ethidium bromide by heating at

650C for 15 min. The samples were then cooled on ice, loading buffer

added (10 x loading buffer is 1 mM EDTA, pH 8.0, 0.25%

bromophenol blue, 0.25% xylene cyanol, 50% glycerol), and

electrophoresed in 1.2 % formaldehyde agarose gels in a horizontal

gel apparatus (Model H5, BRL) at 80-100 volts. After






55

electrophoresis, the RNA samples were transferred to nylon N' by

downward transfer with lOx SSC (20x SSC is 3M NaCI, 0.3M sodium

citrate, pH 7.0). The blots were then air-dried and uv-crosslinked by

the autocrosslink program in Stratalinker (Stratagene).

The blots were pre-hybridized in 6x SSPE (20x SSPE is 3.6M NaCI,

0.2M sodium phosphate, 0.02M EDTA, pH 7.7), 50% formamide, 5x

Denhardt's solution [50x Denhardt's solution is 1% each of Ficoll 400

(Sigma), polyvinylpyrrolidone (Kodak), and bovine serum albumin

(Fraction V, USB-Amersham)], 0.5% SDS, 200 ug/ml tRNA at 600C for

1 hour. Radiolabeled riboprobes were synthesized from linearized

gene-specific clones using T3 and T7 RNA polymerases (Stratagene),

added to a fresh aliquot of the same buffer @ 50-100 ng/ml, and

hybridized overnight at 600C. Following hybridization, the blots were

washed in 40 mM Na', 1% SDS, 1 mM EDTA at 600C for 1.5 hr with 3-

4 changes. Following the washes, excess wash solution was removed

and the blots exposed to both film and phosphorimaging screen.


Dot Blot Quantitation of Mitochondrial DNA


The same aliquot of mitochondrial nucleic acids used for the

Northern analysis was used to quantitate the level of mtDNA in that

sample, so that the steady state levels could be corrected to unit






56

mitochondrial genome. For this, the samples were incubated with 3

ul of RNase Cocktail (Ambion, Inc.) at 370C for 15' to degrade RNA.

The samples were then adjusted to 0.4 M NaOH and denatured at

960C for 10', cooled on ice and SSC added to lOx. The samples were

then filtered through a pre-wetted nylon N+ membrane in a dot blot

apparatus (Bio-Rad). The dots were washed three times with 500 ul

of lOx SSC, air-dried, uv-crosslinked, and hybridized with mtDNA

specific riboprobes using the hybridization buffer system of Church

and Gilbert (1984) described in the footprinting section. The

Phosphorimager quantitation of mtDNA on the dot blots were used to

normalize the levels of the mitochondrial RNAs (measured by

Northern analyses) to unit mitochondrial genome.


Mitochondrial Run On Transcription


Oligonucleotides


WH540. 22 base oligonucleotide CCG ACG TCT GCA GTA AGT

CAT G. Corresponds to the L-strand sequence 1603-1624 in Dunon-

Bluteau et al. (1985).

WH541. 21 base oligonucleotide CAG ATT CAC TAC CGA CAG

ATG. Corresponds to the H-strand sequence 3078-3058 in Dunon-

Bluteau et al. (1985).








Plasmid Clones


Most of the plasmid clones used for the production of the non-

radioactive gene- and strand-specific riboprobes have been

described in the steady state RNA analysis section above. The one

plasmid clone which has not been described earlier is the pBSD-Loop

clone, which is described below.

pBSD-Loop. The Xenopus D-loop region (from 1603-3078 in

Dunon-Bluteau et al., 1985) including the two putative TAS elements

and the 3' ends, were PCR amplified using primers WH540 and

WH541, and the PCR product cloned into the Hincil site of pBS (+/-)

phagemid (Stratagene). The insert was cloned in the forward

orientation. Therefore, T3 RNA polymerase makes the L-strand

sense and T7 RNA polymerase makes the H-strand sense of cRNAs.


Preparation of Riboprobes for Strand-specific Probing


In order to obtain gene- and strand-specific hybridization targets

for analyzing the labeled run-on transcripts, only riboprobes are

suitable. Large-scale synthesis of cold riboprobes for strand-specific

probing was done with the MegaScript kit (Ambion, Inc.) using the

mitochondrial gene-specific clones linearized with the appropriate

restriction enzyme. The large-scale reactions were allowed to






58

incubate for 4-6 hours and typically yielded 100-120 ug of RNA per

reaction. The reactions were phenol-extracted and ethanol

precipitated twice in the presence of 0.5M ammonium acetate. The

RNA yield was calculated spectroscopically, 1 ug of each transcript

electrophoresed in parallel lanes of 1.2% formaldehyde-agarose gels,

and blotted onto nylon membranes as described in the section on

steady state analyses of mitochondrial RNA. In some of the earlier

experiments, combinations of RNAs were run in the same lane as L-

strand or H-strand sets. This was not followed in later experiments

due to difficulties in quantitating the bands after hybridization.

Since in vitro transcription reactions always result in shorter

transcripts of the same specific sequence, gel purification of the RNA

was attempted. Although effective, the method is labor-intensive

and inefficient. To overcome this problem, the membranes carrying

the RNA blots were trimmed around the specific RNA bands prior to

hybridization, to aid in ease of quantitation. This method was

preferred over dot-blotting the probe panel due to the following

reasons: (1) electrophoresing the RNA samples each time serves as a

checkpoint for RNA degradation, (2) gene assignments are

unambiguous due to the pattern of specific RNA sizes on the blot, and

(3) removal of template DNA and shorter transcripts is automatically

facilitated.






59

Mitochondrial Run-On Transcription Assay


The incubation conditions for the run-on transcription assays are

modifications of the protocols of Gaines et al. (1987) and Gaines

(1996). Mitochondria from eggs and staged-embryos were isolated

in MSE buffer as outlined in the mitochondrial isolation section. After

one wash with MSE buffer in the microcentrifuge, the pellets were

resuspended in 1 ml of MSE buffer and spun at 750 x g in a

microcentrifuge to pellet pigment and debris which were carried

over during handling of the bigger volumes. The supernatant was

then spun at 10,000 x g for 1 min to pellet mitochondria. The pellet

was then resuspended in run-on wash buffer (40 mM Tris, 7.2, 50

mM NaCI, 10% glycerol, 5 mM MgCl2, 1 mg/ml BSA) and the

mitochondria re-pelleted. The final pellets were resuspended in the

basic run-on buffer (40 mM Tris, 7.2, 50 mM NaCI, 10% glycerol, 5

mM MgCl2,, 2 mg/ml BSA).

Supplementation of this basic buffer with salts, nucleotides, TCA

cycle intermediates or drugs interfering with mitochondrial

metabolism was done by adding the components to a 1.2x or 1.25 x

Basic buffer, mixing the components, and then resuspending the

mitochondria in the final mixture. Run-on reactions were initiated

by the addition of [a32P]UTP (50-100 mci) to the mitochondrial






60

suspension on ice and then incubating at 300C for 30 min. Following

the incubation, the samples were cooled on ice, and then pelleted in a

microcentrifuge at 10,000 x g for 1 min. The mitochondrial pellets

were either used immediately for probing mtDNA-specific cold

probes or stored at -800C till necessary.

To detect and quantify the run-on RNA, the mitochondrial pellets

were resuspended in 50 mM Tris, 7.5, 200 mM NaCI, 25 mM EDTA.

After adding SDS to 1% to lyse the mitochondria, the lysate was spun

for 2 min at 14, 000 x g in a microcentrifuge to remove pigment. The

supernatant was added to the pre-hybridized riboprobe panel in

fresh hybridization solution and allowed to hybridize in a rolling

hybridization oven (Biometra) at 600C overnight. The pre-

hybridization, hybridization, and wash conditions were the same, as

that used for the steady state RNA analyses. All quantitations were

performed using Imagequant software (Molecular Dynamics) on the

Phosphorimager.

To correct for differences in the yield of mitochondria obtained

from the different developmental stages, the mtDNA levels in each

run-on reaction was measured as follows. Prior to the addition of the

radiolabel, 10% of the mitochondrial suspension in basic buffer was

removed, pelleted and lysed in mitochondrial lysis buffer (described






61

in the footprinting section). The lysate was then extracted twice with

phenol:chloroform:isoamyl alcohol, once with chloroform, and then

processed for dot blotting (without ethanol precipitation) as

described in the steady state analysis section.


Measurement of Mitochondrial Adenine Nucleotides


Extraction Of Adenine Nucleotides


The total egg/embryonic and mitochondrial adenine nucleotides

were measured by reversed-phase, ion-paired HPLC (RP-HPLC).

Adenine nucleotides were extracted from eggs, embryos, and isolated

mitochondria by an adapted protocol of Weinberg et al., 1989. For

eggs and embryos, 5-25 de-jellied eggs or embryos were

homogenized in 225-400 ul of 12% TCA with an Eppendorf pestle

(Kontes) and incubated at room temperature for 5 minutes. The

tubes were then vortexed and spun in an Eppendorf centrifuge for 5

minutes. The clarified supernatant was removed and extracted twice

with 2-6 volumes of 0.5 M trioctylamine in cyclohexane, centrifuged,

the aqueous layer filtered through a 0.22 micron nylon centrifuge

filter (Micron Separations Inc.), and stored at -700C.

For extraction from isolated mitochondria, mitochondria from 300

eggs or embryos were prepared by differential centrifugation and






62

sucrose gradient sedimentation as described. The final

mitochondrial pellets were lysed in 12% TCA by repeated pipeting

and then processed as above.


HPLC Measurements


All HPLC runs were done on a CI8 silica HPLC column using an

injection volume of 100 ul, the mobile phase being 0.2M sodium

phosphate titrated to pH 5.75 with tetrabutyl ammonium hydroxide

(TBAH), by the isocratic elution mode. Signal amplification, peak

detection, and area quantitations were done using Nelson PE

hardware and software. Standard calibration curves for ATP, ADP,

and AMP were generated using ultrapure ATP, ADP, and AMP

(Sigma), and used to calculate the amounts of the three adenine

nucleotides in the total egg/embryo and mitochondrial TCA extracts.


Estimation of protein concentrations


The adenine nucleotide levels found in each developmental stage

were normalized to the amount of mitochondrial protein in the

respective extract. For this, protein pellets obtained after TCA

precipitation were resuspended in resuspension buffer (Tris, pH 8.5,

0.5% SDS, ImM EDTA, ImM PMSF) using an Eppendorf pestle (Kontes)

and repeated pipeting. After thorough resuspension, the insoluble






63

debris was removed by a brief centrifugation. The clarified

supernatant was used for protein estimation by the Lowry method

(1951), using BSA as a standard.


Electrophoretic Mobility Shift Assays


Oligonucleotides


WH 615. 40 base oligonucleotide AGA TTA GGG CTA GCT AGC

GTG GCA GAG CCT GGC TAA TGC G. Corresponds to the native L-

strand sequence 4712-4751 (Roe et al., 1985).

WH 616. 41 base oligonucleotide GCT TAG GTC TTT CGC ATT

AGC CAG GCT CTG CCA CGC TAG CT. Corresponds to the native H-

strand sequence 4763-4723.

WH 639. 46 base oligonucleotide AGA TTA GGG CTA GCT AGC

GTG GCA GGG CCT GGC TAA TGC GAA AGA. Corresponds to the L-

strand sequence 4712-4756, with a point mutation at position 4737

(A to G) which is analogous to the MELAS transition at position 3243

(A to G) in the human mitochondrial genome.

WH 640. 44 base oligonucleotide GCT TAG GTC TTT CGC ATT

AGC CAG GCC CTG CCA CGC TAG CTA GC. Corresponds to the H-strand

sequence 4763-4720, with a point mutation at position 4737 (T to C

transition), to match the mutation in WH 639.






64

WH 654. 46 base oligonucleotide AGA TTA GGG CTA GCT AGC

GAG CCA GGG CCT GGC TAA TGC GAA AGA. Corresponds to the L-

strand sequence 4712-4756, with three point mutations in the

mTERF consensus sequence (Valverde et al., 1994) at positions 4731

(T to A), 4733 (G to C), and 4737 (A to G).

WH 655. 20 base oligonucleotide GCT TAG GTC TTT CGC ATT AG.

Corresponds to the H-strand sequence 4763-4744.


Preparation of Band Shift Probes and Competitors


To make the 52 bp wild type mTERF bandshift probe, 1 nmol of

WH 615 and WH 616 were denatured in 10mM Tris, pH 7.5, 10 mM

NaCI at 950C for 4 min, followed by annealing at 510C for 10 min.

The annealed templates were kept on ice and used to perform a fill-

in reaction with either Klenow fragment (New England Biolabs) or

Sequenase (USB-Amersham) DNA polymerase. The use of Klenow

was discontinued in later experiments due to incomplete filling-in of

the termini which led to production of two major products in native

gels. Radiolabeled probes were made using either [a- 'P]TTP or

dATP. Unincorporated radiolabel was removed using spin-columns

(Sephadex 25, Pharmacia) and the probe specific activity calculated.






65

The specific competitor with a single point mutation at 4737 (the

MELAS competitor) was prepared by annealing WH 639 and WH 640

as described above. The specific competitor with three point

mutations in the mTERF consensus sequence was prepared using WH

654 & WH 655. Annealing was done at room temperature in this

case and processed as above. The filled-in probes were diluted to a

concentration of 10 picomoles/ul and frozen at -200C.


Preparation of S-13 Mitochondrial Lysate


Mitochondria from 400 eggs or embryos were purified as

described in the mitochondrial isolation section. The S-13 lysate was

prepared by an adaptation of the protocol of Fernandez-Silva et al.

(1996) for preparing S-13 mitochondrial lysates from HeLa cells.

The high speed mitochondrial pellet was resuspended in 500 ul

lysis buffer (10% v/v glycerol, 25 mM Hepes-KOH, pH 7.6, 5 mM

MgCl,. 0.5 mM EGTA, 1 mM DTT, 1 mM PMSF, 1 ug/ml of Aprotinin,

Leupeptin and Pepstatin A). Tween 20 and KCI were added to final

concentrations of 0.5% and 0.5 M, respectively. The mitochondria

were homogenized with an Eppendorf pestle (Kontes) and repeated

pipeting. The lysate was then kept on ice for 5 min, followed by a 30

sec vortexing. After a further incubation on ice for 5 min, the lysate






66

was centrifuged at 13,000 x g for 45 min in a microcentrifuge at 40C.

The clear supernatant was collected carefully, avoiding the fluffy

layer floating on top. Aliquots were rapidly frozen in a dry ice-

ethanol bath and stored at -800C. Protein concentrations were

determined by the Lowry method.

The extracts were assayed for presence of DNA endonuclease

activity by incubating 20 ug of the extract from each stage with 1 ug

of 100 bp ladder (GIBCO BRL) and incubated at 370C for 30'. The

reactions were run on a 1% agarose gel and stained with ethidium

bromide. By this assay, no detectable DNA endonuclease activity was

detected (data not shown).


Band Shift Assay


The protocol for the mobility shift assay for mTERF is essentially

the one described by Fernandez-Silva et al. (1996) for human

mitochondrial lysates, with some modifications. The DNA-binding

reactions were assembled on ice and consisted of the following

components: 0.1 pMol of the mTERF bandshift probe (0.2-lx

10"cpm/pMol), 5 ug BSA, 5 ug poly (dl-dC).(dI-dC) as non-specific

competitor, and the desired amounts of the S-13 mitochondrial

extract (usually from 10 to 60 ug). The buffer and salt conditions





67

were modified accordingly to obtain final concentrations of 25 mM

HEPES, pH 7.6, 100 mM KCI, 12.5 mM MgC92, 1 mM DTT, 12% glycerol.

The samples were gently mixed, spun down briefly, and incubated in

a room temperature water bath for 15 min. After the incubation, the

samples were placed on ice and then loaded onto a 5%

polyacrylamide:bisacrylamide (80:1) gel in Tris-glycine buffer

containing 2.5% glycerol. The gel had been pre-run for at least 2 hr

at 150 volts. Samples on the gel were run at 100-200 volts at room

temperature under conditions which did not heat the gel. After the

run, the gel was transferred to 3 mm paper, dried, and exposed to

film or phosphor screens.

For competition experiments, the specified molar excess of cold

competitors were added to the reactions after a 15 min pre-

incubation with the wild type probe.














CHAPTER 3
STEADY STATE ANALYSIS OF MITOCHONDRIAL RNA LEVELS


Introduction


Xenopius eggs, as those of other animals, contain large numbers of

mitochondria and consequently large amounts of mtDNA and mtRNA

(Dawid. 1965; 1966). This stockpile of mitochondria, along with

other organelles and macromolecules, are partitioned into the rapidly

dividing embryos, bypassing the need for fresh biogenesis of these

components until later in embryogenesis (Chase and Dawid, 1972).

Active synthesis of mtDNA does not occur till about 40-48 hours of

development. Chase and Dawid (1972) used '"C incorporation to

measure mitochondrial rRNA synthesis during early development in

Xenopus laevis by polyacrylamide gel electrophoresis and

scintillation counting of gel slices. Before the gastrula stage (10

hours of development), the rate of synthesis was very low. It

increased 4-5 fold after the gastrula stage and -8-9 fold by

neurulation (by 22 hours post-fertilization). On the basis of these






69

data, they concluded that after the gastrula stage, there is a "switch-

on" of mitochondrial rRNA synthesis at gastrulation (10 hours post-

fertilization). Accumulation then proceeded at a constant rate, and

by 96 hours post-fertilization, the content of both mitochondrial

rRNAs per embryo had doubled.

Young and Zimmerman (1973) measured the synthesis of mtRNA

by "P labeling of disaggregated Xenopus embryos at different

developmental stages, followed by electrophoresis and scintillation

counting of gel slices. They could detect mitochondrial RNA synthesis

in the blastula, albeit at a very low level. Older embryos (26-32

hours post-fertilization) had greater incorporation (2.5 fold more)

over the gastrula stage (10 hours post-fertilization). Differences in

the extent of disaggregation and resultant differences in '"P uptake

by the embryos were inherent problems in their system. However,

their results support the general conclusion that the older the

embryos. the greater is the incorporation into mitochondrial RNA.

Dawid et al. (1985) microinjected ["3P]GTP into fertilized eggs and

analyzed labeled RNA by methyl mercury hydroxide agarose gel

electrophoresis. They could not detect RNA synthesis of any kind

before the mid-blastula (6-7 hours) stage, with the reservation that





70

they did not show whether the microinjected GTP could enter the

mitochondria from the fertilized egg cytoplasm. Shortly after the

mid-blastula transition (MBT), RNA synthesis was detected both in

total RNA and in RNA prepared from isolated mitochondria.

Webb et al. (1975) investigated mitochondrial RNA synthesis

during oocyte maturation by injecting [3H]GTP into full-grown (Stage

VI) and in vitro matured (progesterone treated stage VI oocytes)

oocytes of Xenopus laevis and studied RNA synthesis in a

polyacrylamide gel system. They found that the rate of mtRNA

synthesis remained essentially constant throughout maturation.

Furthermore, their measured rates of mitochondrial rRNA synthesis

were similar to those reported by Chase and Dawid (1972) for pre-

gastrula embryos. They concluded that mtRNA synthesis progresses

throughout the terminal stages of Xenopus oogenesis and its rate of

synthesis remains essentially unchanged by the events associated

with fertilization.

Meziane et al. (1989) directly measured the steady state levels of

mitochondrial RNAs during X. Iaevis development by Northern

hybridization. In this study, they isolated total RNA from

unfertilized egg and embryos at distinct stages of development,





71

resolved the RNA on formaldehyde agarose gels, blotted them onto

nylon filters, and probed them with gene-specific DNA probes. They

found that the levels of all mitochondrial mRNAs decreased abruptly

within a few hours after fertilization (by a factor of 5-10), remained

at a very low level up to the late neurula stage (24 hours of

development), and increased again during organogenesis. However,

the levels of mitochondrial rRNAs remained essentially constant

during early development. On the basis of these data, they suggested

that the mitochondrial genome was completely inactivated at

fertilization and during early embryonic development and that the

reactivation of mitochondrial transcription in the developing embryo

was decoupled from active DNA replication, since the mtDNA levels

do not vary during this time (Chase and Dawid, 1972).

The early part of the results obtained by Meziane et al. (1989) are

in direct contradiction of the studies reported by the earlier groups.

Chase and Dawid (1972) did not find any evidence for differences in

mtRNA synthesis rates in unfertilized eggs vs. early embryos. Dawid

et al. (1985) found increased incorporation of microinjected ["P]GTP

into mtRNA only after the mid-blastula transition (after 7 hours of

development). In all these studies, increasing rates of mtRNA





72

synthesis were observed only after the mid-blastula transition.

Thus, earlier studies do not support the observation by Meziane et al.

(1989) that mRNA levels decrease after fertilization and remain at a

very low level till the late neurula stage. Hence, the findings of

Meziane et al. (1989) remain a solitary piece of evidence suggesting

that fertilization in Xenopus laevis induces a rapid shutdown of

mitochondrial transcription.

To resolve these issues and extend our understanding of

mitochondrial transcriptional regulation in the Xenopus laevis

embryogenesis system, I re-investigated the steady state levels of

mitochondrial RNAs during early development of X. laevis. The

pattern of early developmental steady state levels of mitochondrial

mRNAs I observed were not in agreement with their results. My

interpretation of these discrepancies and additional experiments

which clarify the situation are presented in the following sections.


Results


Xenopus eggs were fertilized in vitro as described in Materials

and Methods. Developmental stages were identified according to the

Nieuwkoop-Faber tables (Nieuwkoop and Faber, 1967). Total RNA





73

from batches of 100 viable eggs or embryos were extracted as

described in Materials and Methods. Identical amounts of total RNA

(2 ug) extracted from unfertilized eggs and staged embryos were run

on a 1.2% formaldehyde-agarose gel, blotted to nylon membranes,

and probed with gene-specific riboprobes. The results of such an

analysis for six different mRNA species spanning early development

are shown in Figure 3-1, panels A-F. The mRNA levels were

quantitated using a Phosphorimager. The level of each transcript at

each developmental stage is relative to the egg level (Table 3-1),

setting the egg level to unity. All mRNAs showed a 1.5 to 2 fold

decrease in transcript levels by 14 hours of development over that of

the unfertilized egg, returned to egg levels by 20 hours in the case of

ND1, Cyt b, CO II and ATPase 6 (Figure 3-1, panels A, D, E, F and

Table 3-1). and increased .1.rIir..i1. thereafter. In contrast, the

levels of ND4 and ND6 (Figure 3-1, panels B and C) continued to be

low even in the tadpole stage (50 hours of development). The 12S

and 16S rRNA levels continued to decrease marginally during this

window of development (Figure 3-1, panels G & H and Table 3-1).

The complex pattern of bands in the two rRNA blots have been

observed by other groups working on different systems (Williams,













Figure 3-1. Steady state levels of mitochondrial RNA during early development of
Xenopus laevis. Total RNA was isolated from unfertilized eggs and embryonic stages at 5, 13, 20, 30,
38, and 50 hours post-fertilization as described in Materials and Methods. 2 ug of total RNA from each
developmental stage was run in a 1.2% formaldchyde-agarose gel, blotted on nylon N' membranes,
hybridized with gene-specific riboprobes synthesized by in vitro transcription from linearized clones,
washed, and exposed to x-ray film or phosphorimager (described in Materials and Methods). Each lane
(labeled accordingly) corresponds to the respective mRNA or rRNA level in the corresponding
developmental stage. Panel A: NDI mRNA; Panel B: ND4 mRNA; Panel C: ND6 mRNA; Panel D: Cyt b
mRNA; Panel E: CO II: Panel F: ATPase 6 mRNA; Panel G: 12S rRNA; Panel H: 16S rRNA. The complex
pattern of bands seen in panels G and H represent multiple products corresponding to the 12S and 16S
sequences, and is found in ribosomal RNA blots from other animal systems (see text).














0 ,,00 0 O 0 00 0

A ** D


B w -9 -e -rry


S* *w 9 E


1C ** 4-1 F










G

p, L LL~I-
DDffC
v, U) 0 0
U)nv


H


U) U


ii::;;;

Cu*1 3
I~aem


Figure 3-1...continued














Table 3-1. Steady state levels** of mitochondrial RNA in total egg or
embryonic RNA during early development of X. laevis.


Egg 5 hr* 14 hr 20 hr 30 hr 38 hr 50 hr
ND 1 1.0 0.9 0.7 1.3 1.3 1.4 1.2
ND 4 1.0 0.6 0.5 0.8 0.8 0.8 0.6
Cyt b 1.0 0.8 0.4 1.0 1.0 1.4 1.2
ATPase 6 1.0 0.6 0.4 1.0 1.1 1.5 1.4
CO II 1.0 0.8 0.6 1.0 1.2 1.2 1.2
ND 6 1.0 1.1 0.7 0.7 1.0 0.5 0.5
16 S rRNA 1.0 0.9 0.8 0.9 0.8 0.6 0.7
12 S rRNA 1.0 0.9 0.9 0.9 0.7 0.7 0.8


* Hours post-fertilization
** Steady state levels are expressed as fold over egg, with the egg
levels set to unity.





78

1986; Williams et al., 1986) but they were not discussed except in

one case (Mazo et al., 1983) where they have been accurately

analyzed. In rat liver mitochondria, it has been unambiguously

demonstrated that six abundant poly(A) transcripts ranging from 1.3

to 0.7 kb were non-random fragments of the 16S rRNA, all carrying

its 5' terminus (Mazo et al., 1983). The mechanism for this non-

random degradation or premature termination of synthesis has not

yet been established.

The above results do not agree with the findings reported by

Meziane et al. (1989), where they found a 5-10 fold decrease in

steady state levels of most mtRNAs analyzed within 5-7 hours of

development. All transcripts remained at a low level until 24-28

hours of development, and increased thereafter. One of the

possibilities for this discrepancy between labs is that there may be

strain-specific differences in the frogs used for analyses.

Commercially sold Xenopus laevis is highly heterozygous. After the

mid-blastula transition at 7 hours of development, there is a

dramatic increase in nuclear transcription, with the concomitant

appearance of new species of nuclear RNAs. It is conceivable that

the quantity and quality of specific nuclear RNAs may be under the






79

influence of both genetic and environmental factors. Meziane et al.

(1989) analyzed mitochondrial mRNAs in the context of rapidly

fluctuating total RNA (nuclear plus mitochondrial) levels. Hence, a

normalization based on spectroscopic measurement of total RNA is

not the optimal choice in this case, because the actual steady state

levels of mitochondrial RNAs could be camouflaged by strain and

animal-specific variations in total RNA metabolism.

To remove the interference from differential amounts of nuclear

RNA in the samples, mitochondrial RNA was prepared from sucrose

gradient purified mitochondria as described in Materials and

Methods. Equal amounts (0.5 ug) of mitochondrial RNA were

electrophoresed and blotted as above with mtDNA specific

riboprobes (Figure 3-2, panels A-F). Since the bulk of the

mitochondrial RNA is rRNA. the mitochondrial mRNA levels were

normalized to the steady state levels of mitochondrial 16S rRNA, and

the level of each transcript at each developmental stage expressed

relative to the level in the unfertilized egg, setting the egg level to

unity (Table 3-2). With the bulk of the interference from nuclear

RNA removed, Cox II, Cyt b. ND1 and ND4 levels now showed a 2-2.5














Figure 3-2. Steady state levels of mitochondrial RNA isolated from purified mitochondria
during early development of Xenopus laevis. Mitochondrial RNA was isolated from sucrose-
gradient purified miiocliondria prepared from unfertilized eggs and embryonic stages at 13, 22, and 43
hours post-fertilization. 0.5 ug of RNA from these distinct stages were run on 1.2% formaldehyde-
agarose gels, blotted on nylon N' membranes, hybridized with gene-specific riboprobes synthesized by
in vilro transcription from linearized clones, washed, and exposed to x-ray film or phosphorimager
(described in Materials and Methods). Each lane (labeled accordingly) corresponds to the respective
mRNA or rRNA level in the corresponding developmental stage. Panel A: NDI mRNA; Panel B: ND4
moRNA; Panel C: ND6 mRNA; Panel D: Cyt b mRNA; Panel E: CO II mRNA; Panel F: 16S rRNA. The complex
pattern of bands seen in panels F represent multiple products corresponding to the 16S sequence, and is
found in ribosomal RNA blots from other animal systems (see text).






















C *


C


1P T at

tlJ -~ -~f -
D




E






F















Table 3-2. Mitochondrial mRNA steady state levels** in isolated
mitochondria from distinct developmental stages of X. laevis.


Egg 13 hr* 22 hr 43 hr
ND1 1.0 0.9 1.3 3.4
ND4 1.0 0.6 0.8 2.8
Cyt b 1.0 0.5 0.8 3.1
CO II 1.0 0.9 1.2 3.0
ND6 1.0 0.8 0.8 0.8


* Hours post-fertilization.
** The steady state level of each
16S rRNA level in that stage and
the egg levels set to unity.


stage has been normalized to the
is expressed as fold over egg, with





83

fold increase over egg levels in the 43 hour-old embryo. ND6 levels,

however, showed a marginal decrease during this time.

Based on the above results, I conclude that there are no

substantial reductions in steady state levels of mitochondrial mRNAs

soon after fertilization as reported by Meziane et al. (1989).

Although these results seem to agree well with the observations of

Chase and Dawid (1972), Young and Zimmerman (1975) and Dawid et

al. (1985), they suffer from the following drawback. The mRNA

levels were normalized to rRNA levels based on the assumption that

the rRNA levels do not change drastically during development.

However, if these levels do change due to active transcription in the

later stages of development, the mRNA levels would be

underestimated in these stages. Therefore, a more static reference is

needed.

One macromolecule whose level does not change during the first

two days of development in Xenopus laevis is the mtDNA molecule.

The mtDNA remains at a concentration of 4 ng per embryo during

the first 48 hours of development after fertilization (Chase and

Dawid, 1972). Net increases in its level could be detected only at 60

hours, after which it continued to increase, reaching 8 ng per embryo






84

by 96 hours of development (Chase and Dawid, 1972). Hence the

perfect macromolecular species to normalize mitochondrial RNA

levels is the mitochondrial DNA itself. This also makes teleological

sense, since it is the template for mitochondrial transcription.

In order to normalize mtRNA levels to mtDNA levels, it is crucial

to isolate DNA and RNA simultaneously using a single method from

the same embryos. If they are isolated in parallel rather than

simultaneously, the chances of introducing additional experimental

errors are increased. Hence, mitochondrial DNA and RNA were

simultaneously isolated from purified mitochondria of distinct

developmental stages using pH 7.0 buffers as described in Materials

and Methods. Equal volumes of the DNA/RNA mixture from each

developmental stage were RNAase treated to remove RNA, alkali

denatured and dot blotted as described in Materials and Methods to

quantitate the level of mtDNA in each sample using a

Phosphorimager. This quantitation was used to approximately

normalize the level of mtDNA in each stock sample to that of the

sample with the lowest amount of mtDNA by adding sterile water.

In control experiments, DNAse I digestion of the same samples prior

to alkali denaturation and dot blotting completely eliminated the






85

signal (not shown), confirming that the blots truly represented the

levels of just mtDNA, without any RNA carryover.

Equal volumes of these approximately normalized samples were

then electrophoresed on formaldehyde-agarose gels, blotted and

hybridized as described earlier (Figure 3-3, panels A-F). In a

parallel experiment, the same volume of sample from each stage was

again dot blotted for mtDNA quantitation as described above (Figure

3-3, panel G). The Phosphorimager quantitation of the mRNA levels

in the Northern blots were then corrected for the mtDNA levels in

each stage, and the steady state level of each transcript at distinct

developmental stages expressed relative to the level in the

unfertilized egg, setting the egg level to unity (Figure 3-4 and Table

3-3). As seen in Figure 3-4 and Table 3-3, the steady state levels of

ND1, ND4, Cyt b, Coll and ATPase 6 (panels A, B, D, E, F) remained at

more or less egg levels until 9 hours of development. The levels of

these transcripts were about 4-fold more than egg levels by 20

hours, 6-8 fold more by 30 hours, and 17-23 fold more by 48 hours

of development. The high levels seen at 48 hours drops down to the

levels seen at 30 hours in the 5-day old tadpole (7 days from the

time of fertilization). ND6 (Figure 3-3, panel C) showed lower steady






Figure 3-3. Steady state levels of mitochondrial RNA per unit mitochondrial genome
during early development of Xenopus laevis. Mitochondrial DNA and RNA were simultaneously
isolated in the same solution from mitochondria prepared from unfertilized eggs and embryonic stages
at 6 hours, 9 hours 20 hours 30 hours, 48 hours, and 7 days post-fertilization. An equal volume of
sample from each developmental stage was RNAase-treated, alkali-denatured at 960C for 5 minutes,
and dot blotted in lOx SSC as described in Materials and Methods. The dot blots were hybridized with
an antisense riboprobe to Cyt b, washed, and the mtDNA levels quantitated using the Phosphorimager.
The Phosphorimager quantilation was used to approximately normalize (lie samples fromii each
developmental stage to the mtDNA level in that respective stage. The DNA dot blot experiment was
again repeated as above with normalized samples to accurately estimate the actual mtDNA levels for
correcting the RNA levels (Panel G). Digestion with DNAse I completely eliminated this signal (not
shown). Each lane in Panel G (labeled accordingly) corresponds to the mtDNA level of the respective
developmental stage. In parallel experiments, equal volumes of these approximately normalized
RNA/DNA samples were then run on 1.2% formaldehyde-agarose gels, blotted to nylon N' membranes,
hybridized with gene-specific riboprobes synthesized by in vitro transcription from linearized clones,
washed, and exposed to x-ray film or Phosphorimager (described in Materials and Methods). Each lane
(labeled accordingly) corresponds to the respective mRNA or rRNA level of the corresponding
developmental stage. Panel A: NDI mRNA; Panel B: ND4 mRNA; Panel C: ND6 mRNA; Panel D: Cyt b
mRNA; Panel E: CO II mRNA; Panel F: ATPase 6 mRNA; Panel H: 12S rRNA; Panel 1: 16S rRNA. The
complex pattern of bands seen in panels H and I represent multiple products corresponding to the 16S
and 12S sequences, and is found in ribosomal RNA blots from other animal systems (see text). The shift
in mobility seen in some of the lanes is an electrophoretic artifact, and does not represent a higher
molecular weight species.















A .e9


B i *1



C 0 *< f*a*


*00
,. .s*9 9


M ^ ^ c L l i

*i 4 *1.
OC mC
i~h oe 00














H



(Kb) CD 0 0
CD 0) r CO ') t-
9.5- I I I I I I
7.5-


4.4-



2.4-



1.35-


12 4-





0.24-


(Kb) 0

9.5- I I
7.5-

4.4-



2.4-


165 4









0.24-


Figure 3-3...continued















Table 3-3. Steady state levels** of mitochondrial RNA (normalized to
mtDNA) during early development of X. laevis.


Egg 6 hr* 9 hr 20 hr 30 hr 48 hr 7 days
ND 1 1.0 1.3 1.5 4.6 7.2 19.5 6.1
ND 4 1.0 1.4 1.6 3.7 6.2 17.7 4.9
Cyt b 1.0 1.2 1.3 4.0 8.1 23.1 6.6
ATPase 6 1.0 1.3 1.2 3.7 8.1 28.7 8.5
CO II 1.0 1.0 1.3 3.7 6.8 19.7 8.0
ND 6 1.0 0.7 1.1 2.8 3.7 8.4 2.3
16 S rRNA 1.0 0.9 1.1 1.9 2.2 3.8 1.3
12 S rRNA 1.0 0.8 1.0 1.7 2.2 4.2 1.9


* Time post-fertilization.
** Steady state level of each developmental stage has been
normalized to the level of mtDNA in that stage and is expressed as
fold over the level in the egg, with the egg level set to unity.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EK52XEFF7_XBAA4E INGEST_TIME 2013-02-07T18:15:57Z PACKAGE AA00012971_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES