Homologous pairing of DNA molecules promoted by rec1 protein of Ustilago Maydis

MISSING IMAGE

Material Information

Title:
Homologous pairing of DNA molecules promoted by rec1 protein of Ustilago Maydis
Physical Description:
xiv, 174 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kmiec, Eric B
Publication Date:

Subjects

Subjects / Keywords:
Basidiomycota   ( mesh )
DNA   ( mesh )
Microbiology   ( mesh )
Immunology and Medical Microbiology thesis Ph.D   ( mesh )
Dissertations, Academic -- Immunology and Medical Microbiology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida.
Statement of Responsibility:
by Eric B. Kmiec.
General Note:
Photocopy of 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 - 000413248
oclc - 11123151
notis - ACG0279
System ID:
AA00009104:00001


This item is only available as the following downloads:


Full Text











HOMOLOGOUS PAIRING OF DNA MOLECULES PROMOTED
BY REC1 PROTEIN OF USTILAGO MAYDIS





By

ERIC B. KMIEC


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


1984

















To my parents, Emil and Catherine, and my wife, Jennifer













ACKNOWLEDGEMENTS


I would like to express my sincerest appreciation to my mentor,

Dr. William K. Holloman, for the superb guidance he has given me

throughout this effort. His sterling example of how research should be

approached and carried out will remain with me always. He has been a

friend, an advisor, and a collaborator.

In addition, I should like to thank the members of my dissertation

committee, Dr. Bruce Alberts, Dr. Kimon Angelides, Dr. Kenneth I.

Berns, Dr. James B. Flanegan, Dr. Nicholas Muzyczka, and Dr. Chris

West. Each of them contributed ideas and expertise throughout the

course of this work.

I am grateful to my co-workers, Dr. Michael Brougham, Dr. Thomas

Rowe, Dr. James Rusche, Michele Yarnall, and Roy Bohenzky for their

support and advice. I would like to thank Paul Kroeger for his

outstanding efforts and collaboration and for making the laboratory a

fun place to be. I would like to thank Sandy Ostrofsky for typing and

editing this dissertation.














TABLE OF CONTENTS

Page

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

LIST OF TABLES ....................... ...................... vii

LIST OF FIGURES .............................................. viii

KEY TO ABBREVIATIONS ............. ................... ... ...... x

ABSTRACT .................................................... xiii

CHAPTER

ONE INTRODUCTION AND DEVELOPMENT OF ASSAY SYSTEMS ........ 1

Purification of 0 Protein ............................ 2
a Protein Promotes DNA Reannealing ................... 7
Reannealing Activity in Recombination-Deficient
Mutants ............................................ 10
Discussion ............................ ............... 14

TWO USTILAGO MAYDIS CONTAINS AN ENZYME WHICH CATALYTICALLY
PAIRS HOMOLOGOUS DNA MOLECULES ...................... 15

Isolation and Purification of a Pairing Protein ...... 17
DNA-Dependent ATPase Activity ........................ 18
ATP-Dependent Reannealing Activity ................... 24
Homologous Pairing of Single Strand Fragments with
Superhelical DNA ........... ........... ............ 24
Recl Mutant of Ustilago .............................. 30
Discussion ...................... ....... ......... .... 30

THREE HETERODUPLEX FORMATION PROMOTED BY REC1 PROTEIN ...... 34

Initiation of Strand Transfer ........................ 36
Energy Requirements for Heteroduplex Growth .......... 42
Mechanism of Strand Transfer ......................... 42
Discussion ................................... ....... 50







Page


FOUR TOPOLOGICAL LINKAGE OF CIRCULAR MOLECULES CATALYZED
BY REC1 PROTEIN AND TOPOISOMERASE .................... 53

Homologous Pairing and Topological Linkage of Single
Stranded Circles and Closed Circular Duplex DNA .... 54
Reaction Requirements ................................ 59
Catenation of Intact Homologous Duplex DNA Molecules 59
Characterization of Complexes Formed Between Intact
Homologous Cirucular Duplexes ...................... 65
Discussion .............. ......... ................. 77

FIVE SYNAPSIS OF DNA MOLECULES PROMOTED BY REC1 PROTEIN .... 80

Formation of Stable Complexes ........................ 81
Paranemic Joints Contain Left-Handed, Z-DNA .......... 85
Discussion ........................................... 89

SIX HYSTERETIC REGULATION OF REC1 PROTEIN PRIOR TO DNA
SYNAPSIS ............................ ................. 94

Kinetics of Synapsis are Anomalous ................... 95
Inactivation of Recl Protein During Presynapsis ...... 98
Presynaptic Inhibition Can be Blocked ................ 101
Low Temperature Reverses Presynaptic Inhibition ...... 107
Discussion ........................................... 110

SEVEN INDUCTION OF REC1 PROTEIN BY U.V. IRRADIATION OR
HEAT SHOCK ............................................ 117

Irradiation of Ustilago Cells ........................ 118
Heat Shock of Ustilago Cells ......................... 121
Discussion ........................................... 130

EIGHT INTERACTION OF REC1 PROTEIN WITH LEFT-HANDED, Z-DNA .. 132

Single Stranded DNA Outcompetes Z-DNA for Recl
Protein ............................................ 136
Binding of Z-DNA is Stimulated by B-DNA ........... 142
Poisoning of Paranemic Joint Molecule Formation is
Blocked by Z-DNA ................................ ... 142
Discussion ............................. .... ... 145

NINE SIGNIFICANCE ......................................... 149

APPENDIX

A BACTERIAL AND FUNGAL STRAINS ......................... 156

B PREPARATION OF DNA STOCKS ........................... 157

C PREPARATION OF DNA SUBSTRATES ........................ 158








Page

D PROTEIN PURIFICATION ................................. 160

E ASSAYS ............................................ ... 161

F METHODS AND MATERIALS ................................ 166

REFERENCES ................................................... 168

BIOGRAPHICAL SKETCH ............. ............................. 174











































vi













LIST OF TABLES


Page
TABLE

1-1 Levels of Protein and Exonuclease in Red Mutants
Lysogens .................. ...... ..... ........... 13

2-1 Purification of Recl Protein ........................ 19

2-2 ATPase Activity Associated with the Ustilago Protein 25

2-3 Level of Reannealing Activity in Mutants ............. 31

4-1 Extent of Heteroduplex Formation ...................... 58













LIST OF FIGURES


Page
FIGURE

1-1. Chromatography of red gene products ............... 4

1-2. SDS gel electrophoresis of purified a protein ........ 6

1-3. Requirements of Reannealing Reaction ................. 9

1-4. Activity of X exonuclease on denatured DNA treated
with a protein ................. ................... 12

2-1. Co-chromatography of reannealing and ATPase
activity ............................................. 21

2-2. Purification of the Ustilago pairing protein ......... 23

2-3. Kinetics of reannealing ............................ 27

2-4. Uptake of homologous single-strand fragments by
superhelical DNA ........................... ......... 29

3-1. Formation of joint molecules assayed by two different
methods .... ... ........................... ... 39

3-2. Formation of joint molecules measured by agarose gel
electrophoresis ................................... 41

3-3. ATP requirement in formation of joint molecules ...... 44

3-4. Restriction maps of DNA from phages M13 Goril and fd 47

3-5. Formation of joint molecules at a preferred end ...... 49

4-1. Linkage of single-stranded circles and form I DNA .... 56

4-2. Linkage of single stranded circles with duplex circular
DNA in the absence of homology or superhelicity ...... 61

4-3. Linkage of form I DNA molecules promoted by homologous
single stranded fragments ............................ 64

4-4. Linkage of form I DNA molecules promoted by RNA
polymerase ................................ ... ...... 67


viii








Page
FIGURE

4-5. Analysis of product formed with homologous or
heterologous combinations of form I DNA .............. 69

4-6. Linked pairs of circular DNA molecules ...... ...... 71

4-7. Maps of DNA from phages M13 Goril, M13, fd, and G4 ... 73

4-8. Analysis by agarose gel electrophoresis of the catenated
product formed between homologous DNA duplexes ....... 75

5-1. Nitrocellulose filter assay for synaptic complexes ... 84

5-2. Paranemic joint molecules are dissociated by ADP ..... 87

5-3. Paranemic joints contain left-handed DNA ............. 91

6-1. Anomalous kinetics of synapsis ....................... 97

6-2. Inactivation of recl protein during a presynaptic step 100

6-3. Parameters of presynaptic inhibition ................. 103

6-4. Stoichiometry of protector DNA ....................... 106

6-5. Fate of presynaptic single-stranded DNA during
synapsis ............................................. 109

6-6. Temperature dependent binding of recl protein to
single-stranded DNA .................................. 112

7-1. Induction of recombinants by ultraviolet light ....... 120

7-2. Induction of ATPase activity in cells irradiation with
UV light ............................................. 123

7-3. Purification of 110 K Protein ....................... 124

7-4. Phosphocellulose chromatography: Reannealing Activity
and ATPase ........................................... 127

7-5. The effect of end structure in strand exchange
reactions ............................................ 129

8-1. The paranemic joint .................................. 135

8-2. Binding of Z-DNA requires ATP ........................ 138

8-3. Recl protein has a higher binding affinity for a single
stranded DNA versus Z-DNA ............................ 141








Page
FIGURE

8-4. DNA in the B-form stimulates the binding of DNA in the
Z-form by recl protein ............................... 144

8-5. Poisoning of paranemic joint molecules is blocked by
Z-DNA ................................................ 147

9-1. The pairing of homologous DNA molecules promoted by
recl protein ............................. .......... 155













KEY TO ABBREVIATIONS


AMPPNP

ADP

ATP

bp

BSA

CTP

d

dCTP

DEAE

DNA

E. coli

EDTA

Kd

PMSF

poly(dC)

poly(dG)

RF I

RF II

RF III

RF IV

RNA

rpm


adenylyl-imidodiphosphate

adenosine-5'-diphosphate

adenosine-5'-triphosphate

base pair

bovine serum albumin

cytosine-5'-triphosphate

dalton

deoxycytosine-5'-triphosphate

diethylaminoethyl

deoxyribonucleic acid

Escherichia coli

ethylenediamine tetra-acetic acid

1,000 daltons

phenylmethylsulfonylfluoride

polydeoxycytidylic acid

polydeoxyguanylic acid

form I, superhelical DNA

form II, nicked circular DNA

form III, linear DNA

form IV, relaxed closed circular DNA

ribonucleic acid

revolutions per minute







SDS sodium dodecyl sulfate

SSC standard saline citrate (0.15 M NaCi, 15 mM sodium
citrate)

u units

Z-DNA left-handed helical double stranded DNA













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


HOMOLOGOUS PAIRING OF DNA MOLECULES PROMOTED
BY REC1 PROTEIN OF USTILAGO MAYDIS

By

Eric B. Kmiec

April 1984

Chairman: William K. Holloman,.Ph.D.
Major Department: Immunology and Medical Microbiology

An enzyme was isolated from mitotic cells of the fungus Ustilago

maydis based on its ability to reanneal complementary strands of DNA.

The purified protein of molecular weight 70,000 daltons catalyzes the

ATP-dependent pairing of a wide range of DNA substrates including the

uptake of single stranded DNA by a homologous superhelical duplex. In

addition, circular single strands of DNA are paired with homologous

linear duplex molecules in a reaction occurring in two distinct and

experimentally separable stages. In the first phase, synapsis, the

enzyme aligns the DNA molecules and brings them into homologous

register. The second phase, strand exchange, extends the length of the

heteroduplex joint in a protein-directed polar fashion. Synapsis

requires only the presence of ATP, while strand exchange utilizes the

energy of ATP hydrolysis. Furthermore, synapsis entails the formation

of a ternary complex composed of the two DNA molecules and recl

protein. When the synaptic complexes are formed in the absence of a


xiii








free end, the topological constraint is relieved by the development of

stretches of left-handed DNA. A complex series of steps leading to

synapsis comprises a third phase of the strand transfer reaction.

During the presynapsis phase, ADP and single stranded DNA regulate the

synaptic pairing activity of recl protein. Mutations in the recl gene

of Ustilago lack this ATP-dependent pairing activity. Because it is

the first enzyme isolated in eukaryotes having such recombinational

properties and because the recl mutation from Ustilago is deficient in

some forms of recombination, the enzyme has been called recl protein.


xiv













CHAPTER ONE
INTRODUCTION AND DEVELOPMENT OF ASSAY SYSTEMS



Genetic recombination is a process by which linkage relationships

among genes are altered by a breakage and reunion of homologous chromo-

somes (Meselson, 1967). The rejoining mechanism involves the pairing

of complementary sequences of single strands of DNA contributed by

different parent duplexes. The structure created by this molecular

splicing is often referred to as a heteroduplex joint. On a molecular

basis, a variety of models have been proposed to account for the

formation of the heteroduplex joint (Holliday, 1964; Whitehouse, 1973;

Meselson and Radding, 1975). It is currently believed that recombina-

tion begins with the transfer of a single strand of DNA from one duplex

molecule to another. Subsequently, this strand pairs with its

complement and displaces its homolog. The structure, thus created, is

commonly referred to as a D-loop.

The biochemical nature of genetic recombination has only recently

come into focus. The remarkable E. coli protein recA, indispensable

for reciprocal recombination, has been shown to align DNA molecules

into homologous register and promote the formation of heteroduplex

molecules (McEntee and Epstein, 1977; Shibata et al., 1979; Cunningham

et al., 1980; Kahn et al., 1981; Cox and Lehman, 1981; Gonda and

Radding, 1983). The pairing of homologous DNA molecules is a crucial

event in the overall process of reciprocal recombination.





2


The bacteriophage X, can carry out homologous recombination in the

absence of functional recA protein relying instead on its own red

system (Signer and Weil, 1968). This two gene system was shown to code

for an exonuclease with a 5' to 3' specificity (Little, 1967), and a

protein, known as a protein, whose function was unknown. An attractive

possible activity for 8 protein would be that the pairing of DNA

strands, a function required for the initiation of recombinational

processes.

Ta- study this possibility an assay was designed in which
[3H]-labeled P22 single strands of DNA become resistant to the single

strand specific nuclease, S-1 (Weinstock et al., 1979). The develop-

ment of such resistance could be a reflection of the strands assuming a

duplex or paired DNA form. P22 was chosen as a substrate because of

its complex nature and slow natural renaturation. To avoid misrepre-

senting results, DNA from the single stranded circular bacteriophage jX

was digested by S-1 nuclease after 0 protein treatment. In this way,

the resistance to S-1 nuclease due to simple protein-DNA binding could

be accounted for.

Purification of 8 Protein

The procedure for purification of a protein outlined by Radding et

al. (1971) was followed with slight modifications (see Appendix A). In

general, the X exonuclease and a protein co-chromatograph throughout

the purification procedure. They were fractionated and separated by

chromatography on phosphocellulose which produced the activity profile

illustrated in Figure 1-1. SDS gel electrophoresis (Fig. 1-2) of two

peaks (I and II) depicted in Figure 1-1 produces a pattern consistent

with that reported by Radding et al. (1971). 8 protein was identified



























Fig. 1-1. Chromatography of red gene products.
Fraction IV containing partially purified exonuclease and a
protein was chromatographed on phosphocellulose as described in
Appendix D. The column was run at 30 ml/hour and fractions of 5 ml
each were collected and monitored for exonuclease (o), reannealing
activity (*), and A280 (A).






































25 50 75 100
Fraction Number


0


2E
0
00
*Cj


C:
0
(11

IC
.c
0
(f)
-Q
O_
0.



























Fig. 1-2. SDS gel electrophoresis of purified a protein.
Samples containing 10 to 20 ug of protein from Peaks I and III
shown in Figure 1-1 were subjected to SDS gel electrophoresis as
described in Appendix F. Molecular weight standards include ovalbumin
(43,000) a-chymotrypsinogen (25,700) and lysozyme (14,300).
a) Standards, b) Peak III, c) Peak I.










0a


m's


,
s/
.'e


b


C








by its characteristic molecular weight (28,000 daltons) and found to be

present in Peak I (lane c). Peak II contained the X exonuclease. Peak

III in the phosphocellulose profile contained the a exonuclease a

protein complex as described by Radding et al. (1971). Both this

complex and the a exonuclease alone contained high levels of nuclease

activity (65,000 units/mg).

0 Protein Promotes DNA Reannealing

Fractions from Peaks I, II, and III were assayed for their

capacity to reanneal complementary strands of DNA as determined by the

S-1 nuclease assay. Peak I and III, but not II, contained high levels

of this activity. Neither peak catalyzed the formation of S-1

resistant DNA molecules when incubated with single strand circular DNA.

One unit of reannealing was defined as the amount of s protein capable

of reannealing 10 picomoles of P22 single stranded DNA in 30 minutes at

370C.

The reaction requirements included a phosphate buffer near pH 6.0

and a divalent cation. Magnesium-chloride (MgCl2) was the most

effective metal used although calcium chloride (CaCl2) produced

approximately 30% reannealing activity. The fraction containing the

most homogeneous preparation of 0 protein was used in the characteriza-

tion studies. The specific activity of this fraction was approximately

1.26 x 10-5 units/mg (Reannealing). The titration of a protein as

a function of its reannealing activity is depicted in Figure 1-3B. The

curve rises sharply and finally plateaus. The reaction is also time

dependent reaching a maximum level at 20 minutes. The amount of a

protein used to determine the time course of reannealing was 5 ug. At

saturation, there was approximately 1 molecule of 8 protein/65






























Fig. 1-3. Requirements of Reannealing Reaction.
Kinetics of reannealing. A. Reaction mixtures prepared as
described in Appendix F containing the indicated amounts of $ protein
were incubated at 370C for 30 minutes. B. Individual reaction mixtures
were prepared and incubated with 5 Ug 0 protein. At the indicated time
each sample was processed as described in Appendix E.




















,40
-o
_0)
-5

020


2 4 6 8 10 10 20 30 40 50
,6 Protein (pg) Time (min)







nucleotides of DNA (Kmiec and Holloman, 1981). The maximum amount of

single stranded P22 DNA becoming S-1 nuclease resistant was only 44% '

input. This somewhat low value may be attributed to the fact tnat P22

DNA may form large aggregates in solution incapable of becoming

reannealed. When single stranded P22 DNA was thermally reannealed at

650C for 3 hours, approximately 55% of input became resistant to S-1

nuclease digestion.

To explore the possibility that both gene products of the red

system may work in a coordinated fashion and to confirm that tne DNA

was actually becoming reannealed, single stranded [3H]-labeled, P22

DNA was treated with 0 protein for 30 minutes at:370C. After incuba-

tion, the DNA was digested with a exonuclease, whose specificity is "-

duplex DNA. Figure 1-4 illustrates that the P22 DNA became increas-

ingly sensitive to a exonuclease digestion as a function of exposure

time to a protein.

Reannealing Activity in Recombination-Deficient Mutants

A variety of phage X recombination mutants are available and

provide the opportunity to directly implicate a protein as an enzyme

which catalyzes the reannealing of DNA molecules. Cells in mini-

culture (2 liter volumes) were processed through the phosphocellulose

step in the purification scheme of a protein. As shown in Table 1-1,

the induced red+ lysogen contained high levels of exonuclease and

reannealing activity. However, mutations in either the exonuclease

gene (X red X314) or B protein genes (B113) (Radding et al., 1971)

showed significant loss of nuclease or reannealing activity

respect vely.



























Fig. 1-4. Activity of X exonuclease on denatured DNA treated with 0
protein.
A reaction mixture (300 pl) containing 10 mM potassium phosphate
pH 6.0, 10 mM MgC12 21 nmol denatured P22 [3H]DNA and 45 Ug 5
protein was incubated at 370C. At the indicated times, aliquots
(0.05 ml) were removed, mixed with 0.25 ml of 50 mM glycine-KOH,
pH 9.5, 10 mM MgCl2, 6.5 units of X exonuclease (a protein) and
incubated an additional 30 minutes. DNA rendered acid soluble was
determined as described in Appendix E.

















03-


.0

0


410-
Z
0


5 10 15 20 25
Time of reannealing (min)








TABLE 1-1

Levels of a Protein and Exonuclease in Red Mutants Lysogens


Exonuclease Reannealing
Lysogen units/mg units/mg x 10-3


AB2463 (Xred+) 900 126
AB2463 (xred+) uninduced 2.0 < 2.0
AB2463 (Xred X314) 0.8 186
AB2463 (Ared B113) 830 < 2.0


CelI extracts were prepared from lysogens as described in
Appendix D and chromatographed on DEAE-cellulose and phosphocellulose.
Specific activities for exonuclease were determined in crude lysates.
Specific activities for a protein promoted reannealing were determined
after phosphocellulose chromatography.








Discussion

s protein from phage X was found to catalyze the reannealing of

complementary single strands of DNA in an assay which measured

increased resistance to S-1 nuclease. The enzyme was also found to be

capable of cooperative enzymatic activity with another gene product

from the same genetic loci. 8 protein's enzymatic activity was

verified by mutant studies in which recombination deficient phage X

strains lacked homologous pairing activity (due to the loss of

functional 0 protein). These results also validate the use of the S-1

nuclease assay in probing extracts of cells for DNA pairing activity.

The results presented here provide a-function for a protein in

generalized recombination events in phage X. Although the recA gene

product could substitute for 8 protein within cells undergoing

recombination, the reaction mechanisms of the two proteins are quite

different. For example, recA protein requires adenosine triphosphate

(ATP) for renaturation reactions, while a protein does not. This

difference may ensure that specific DNA metabolic processes are able to

occur independent of environmental limitations. Furthermore, the red

(phage x) and rec (E. coli) systems of recombination are functionally

unrelated to a large extent (Shulman et al., 1970). Other proteins,

such as the gene 32 protein from T4 phage, have also been found to

catalyze the reannealing of single strands of DNA (Alberts and Frey,

1970). Mutations in this gene product are defective in T4

recombination.













CHAPTER TWO
USTILAGO MAYDIS CONTAINS AN ENZYME WHICH CATALYTICALLY
PAIRS HOMOLOGOUS DNA MOLECULES



Yeast and fungal cells have served as model systems for the study

of genetic recombination in eukaryotes. Since both these organisms can

be studied in the haploid and diploid states, results gained from such

analysis develop our appreciation for the complexities of recombination

in higher organisms. These studies have revealed a number of different

types of recombination events. For example, haploid mitotic cells from

yeast can undergo both intrachromosomal recombination and sister

chromatid exchange (Jackson and Fink, 1981). Gene conversion and

reciprocal recombination occurring in fungi are often closely

associated in meiosis (Case and Giles, 1969). Lesions in DNA molecules

have been found to induce recombination events in mitotic cells,

probably by inducing certain recombination-repair pathways (Fabre and

Roman, 1977). Such an association between DNA breaks and meiosis has

not been established. Resnick and Martin (1976) showed that double

strand breaks may be initiation sites for recombination and recently

Szostak et al. (1983) showed that double strand breaks are recombi-

nogenic.

Mitotic recombination occurs prior to DNA replication with the

creation of symmetric Holliday structures initiated by the asymmetric

transfer of a single strand of DNA into a sister duplex (Meselson and

Radding, 1975). Recently, a mutation in yeast cells has been








identified that diminishes meiotic recombination, but enhances mitotic

recombination. Furthermore, it provides strong genetic evidence for

the existence of an asymmetric exchange event (Bruschi and Esposito,

1983).

This asymmetric exchange event of DNA strands produces a structure

known as a heteroduplex joint, a molecular splice in which two homolo-

gous DNA strands contributed by different parent duplexes are paired

via hydrogen bonds. How the formation of the heteroduplex joint occurs

has been a puzzling question. The discovery and purification of the

recA gene product from Escherichia coli has provided insight into a

possible mechanism for the formation of this crossover junction

(Cunningham et al., 1981). This enzyme of molecular weight 40,000

daltons pairs a wide range of homologous DNA molecules using the energy

generated by the hydrolysis of ATP (Shibata et al., 1979; Weinstock et

al., 1979; DasGupta et al., 1980; Cox and Lehman, 1981; West et al.,

1981). The underlying feature of all recA protein-catalyzed reactions

is its ATP-dependent reannealing of complementary DNA strands. The

wide range of structures created from this activity resemble structures

highlighted in the recombination models of Holliday (1964), Meselson

and Radding (1975), Wilson (1979) and Szostak et al. (1983).

Biochemical studies of heteroduplex formation in eukaryotes have

lagged far behind those in prokaryotes. However, the genetic and

physical analyses of eukaryotic recombination amassed by numerous

laboratories provide a solid basis for such enzymological investiga-

tions. Using the experimental approach devised in studying the recA

protein, but realizing the increased complexity of recombination








mechanisms in eukaryotes, extracts from the fungus Ustilago maydis were

examined for the presence of a DNA pairing activity.

Isolation and Purification of a Pairing Protein

As previously established, a convenient way to measure the

reannealing of complementary single stranded DNA is by using S-1

nuclease to digest any unpaired DNA (Weinstock et al., 1979; Kmiec and

Holloman, 1981). There are a number of limitations in using this

particular assay. First, DNA binding proteins could protect the DNA

against-S-1 nuclease digestion. Second, a variety of other cellular

components, histones and polyamines (Cox and Lehman, 1981) promote the

noncatalytic renaturation of DNA contributing significantly to higher

assay backgrounds. As a control X single strand circular DNA,

incapable of being reannealed itself, was used to measure the blocking

of S-1 nuclease digestion. In addition, only reannealing activity

which was dependent on the presence, and subsequently the hydrolysis of

ATP, was characterized further. Reaction mixtures generally contained

recently denatured P22 DNA and ATP.

The cell extract resuspended in a buffer containing one molar KC1

was initially fractionated by polyethylene glycol phase partition to

remove DNA, polyamines, and DNA associated proteins such as histones.

A small, but detectable amount of ATP-dependent reannealing activity was

observed in this initial purification step. Isolation of the pairing

protein continued by ion exchange chromatography on DEAE-cellulose and

phosphocellulose. Activity was detected by elution with salt gradients

on both columns. The anion exchanger, DEAE, was effective in removing

much of the remaining nonspecific DNA binding activities. Phospho-

cellulose chromatography proved to be an important purification step in








that it concentrated the ATP-dependent reannealing activity which

eluted at a NaCl concentration of 0.35 molar. The main peak of

activity co-eluted with a DNA-dependent ATPase activity. The purifica-

tion scheme and phosphocellulose column profile of activity are

depicted in Table 2-1 and Figure 2-1.

By using SDS polyacrylamide gel electrophoresis (PAGE) the protein

responsible for the DNA pairing and ATPase activities was tentatively

identified. A protein band with molecular weight 70,000 daltons was

enhanced after chromatography on phosphocellulose, corresponding to the

increased level of ATP-dependent reannealing (Figure 2-2, lane e). To

remove a contaminating endonuclease activity, molecular sieve chroma-

tography was employed. From this final purification step, the DNA-

dependent ATPase and reannealng activity co-eluted with a molecular

weight of 70,000 daltons. This observation was verified by electro-

phoresing the active fraction in an SDS PAGE system, again producing a

70,000 dalton protein (lane e) (see Appendix C).

DNA-Dependent ATPase Activity

The observation that a DNA dependent ATPase activity coeluted with

the DNA reannealing activity proved to be important. Because this

ATPase activity can be assayed conveniently, it provided supporting

data throughout the purification procedure for identifying the

appropriate pairing activity. Throughout the course of subsequent

protein preparations, both assays were used simultaneously.

A distinctive characteristic of this ATPase was its dependence on

the presence of single stranded DNA for high activity. No form of

duplex DNA supported appreciable levels of activity when the most





19


TABLE 2-1

Purification of Recl Protein


Fraction Volume Protein ATPase Reann. Specific Activity
(ml) (mg/ml) (u/ml) ATPase Reann.


Crude Extract 180 3.3 420 127.27

P.E.G. 200 1.2 548 456.66

DEAE-
Cellulose 75 .6 822 1370.0

Phosphoso-
cellulose 25 .018 672 720 37,333 40,000

Agarose A
1.5 M 10 .006 914 390 152,333 60,000






























Fig. 2-1. Co-chromatography of reannealing and ATPase activity.
A phosphocellulose column (1 x 12 cm) was loaded, washed and
eluted as described in Appendix F. The flow rate was maintained at
30 ml/hour and fractions of 2 ml were collected. (o) Reannealing.
(*) ATPase.







































10 20 30 40
Fraction Number


C
"7
-I

-5




C
-
men
cz

0
ci)
zr


0.4-0-
o
0
z

0


75>
+-

5.0-
a)

250-
H-




























Fig. 2-2. Purification of the Ustilago pairing protein.
Electrophoresis in a polyacrylamide gel containing SDS was
performed as described in Appendix F. Channel a, standards for estima-
tion of molecular weight include, from top to bottom, phosphorylase b
(Mr 94,000), bovine serum albumin (Mr 67,000), ovalbumin (Mr 43,000),
carbonic anhydrase (Mr 30,000), soybean trypsin inhibitor (Mr 21,000),
channel b, polyethylene glycol Fraction II, 24 pg; channel c, DEAE-
cellulose Fraction III, 18 pg; channel d, phosphocellulose Fraction IV,
0.8 pg; channel e, BioGel A 1.5 M Fraction V, 0.7 pg.









a b


c d


phos
bsa

ov

chy


lys








purified fraction was used (Table 2-2). The specific activity of the

ATPase was 150,000 units/mg.

ATP-Dependent Reannealing Activity

After chromatography on phosphocellulose, DNA reannealing was

reproducibly achieved without contamination of nonspecific binding

activity. The kinetics of reannealing in the presence and absence of

ATP are illustrated in Figure 2-3. The level of reannealing achieved

both as a function of time and protein concentration was approximately

40%. jlIthough this level of activity is somewhat low, it is consistent

with the amount of reannealing observed when protein from phage X was

used under the same reaction conditions (Kmiec and Holloman, 1981) (see

Chapter One). Therefore, the level of reannealing obtained in both

these studies may be a reflection of the aggregation properties of the

P22 DNA (Weinstock et al., 1979).

Homologous Pairing of Single Strand Fragments with Superhelical DNA

The formation of a triple stranded DNA structure by the uptake of

a homologous DNA fragment by a superhelical molecule is thought to be

an important part of the recombination pathway. Various studies have

shown that the recA protein can use ATP as an energy cofactor to

unstack the bases of the duplex while assimilating the homologous

single strand fragment (Shibata et al., 1979; Cunningham et al., 1979).

The D-loop structure thus formed contains both double and single

stranded DNA regions. To assay for the formation of such structures,

nitrocellulose filters were used because of their inherent quality of

retaining single but not double stranded DNA (Beattie et al., 1977).

The Ustilago protein catalyzed the formation of D-loops in a manner

dependent on homology and, to a lesser extent, on ATP (Fig. 2-4).








TABLE 2-2

ATPase Activity Associated with the Ustilago Protein


DNA Cofactor ADP Formed
nmoles

single stranded X174 5.50
OX174 form I 0.35
OX174 form II 0.36
none < 0.20


Reactions were carried out as described in Appendix E. Individual
mixtures contained 1 nmol of the indicated DNA and 30 ng of protein
from Fraction V.



























Fig. 2-3. Kinetics of reannealing.
A. Reaction mixtures (0.30 ml) containing 12 nmol denatured P22
[3H]DNA, 6 Ug of protein from fraction IV and ATP where indicated
were incubated at 370 as described in Appendix E. Aliquots (50 Ul)
were removed at the times shown and the DNA resistant to hydrolysis by
S1 was determined.
B. Reaction mixtures (50 pl) containing either 2 nmol denatured
P22 [3H]DNA or viral strand 0X174 [3H]DNA (3.2 x 104 cpm/nmol)
were incubated with the indicated amount of protein from Fraction IV.
DNA remaining susceptible to S1 was determined as described in
Appendix E.































Time (minutes)


Protein (jg)

























Fig. 2-4. Uptake of homologous single-strand fragments by superhelical
DNA.
Reaction mixtures containing either 4X174 RFI[3H]DNA (2.0 x
104 cpm/nmol) or fd RFI[3H]DNA (1.0 x 104 cpm/nmol) and the homologous
or nonhomologous combination of single strand fragments were prepared
and incubated with the indicated amount of protein from Fraction V.
OX174 RF DNA was approximately 85% Form I and fd RF DNA was approxi-
mately 95% Form I as estimated by a nitrocellulose filter assay
(Kuhnlein et al., 1976). OX RFI and 4X fragments (A); fd RFI and fd
fragments T);1 -X RFI and fd fragments (A); fd RFI and OX fragments
(*).






















-o

C



c-


cc


Q2
Protein








Nonhomologous combinations were not effective substrates for the

formation of joint molecules. A characteristic of joint molecules

containing D-loops is that they are stable to 0.1% SDS or exposure to

75C (Beattie et al., 1977). Over 90% of the D-loop molecules formed

in reactions catalyzed by the Ustilago pairing protein were stable to

either treatment.

Recl Mutant of Ustilago

A variety of phenotypically identifiable Ustilago mutants are

available. Some of them, including the recl mutant, are deficient in

certain types of recombination. The recl mutation and phenotype could

be explained by loss of a regulatory function (Holliday et al., 1976).

Because recA protein functions not only in homologous recombination but

also in the regulation of cellular response to DNA damage (Clark,

1973), the recl mutant and two radiation-sensitive mutations were

examined for altered levels of DNA pairing activity. As illustrated in

Table 2-3, lower levels of DNA reannealing activity were observed in

all the mutations. However, only in the recl mutant stain was the

enzyme catalyzed DNA pairing activity diminished significantly.

Therefore, the Ustilago pairing protein was designated recl protein.

Discussion

The experiments, outlined in this chapter, show that the lower

eukaryote, Ustilago maydis, contains a protein which catalytically

pairs DNA molecules in a reaction dependent on ATP. By using two

assays which measure the reannealing of complementary single strands of

DNA and the DNA-dependent hydrolysis of ATP respectively, the isolation

and purification of this enzyme has been achieved. The uptake of

single strand fragments by superhelical duplexes to form a








TABLE 2-3

Level of Reannealing Activity in Mutants


Strain
haploids


Reannealing Activity
units/mg


58 (rec+) 23,500
293 (recl) < 400
221 (rec2) 10,400
387 (uvs3) 20,600

diploids

M133 (recl-i or recl-2/+) 38,500
M133 S Trecl-1/recl-2) < 400

Cells were grown in two liters of medium and processed as
described in Appendix D. Specific activities indicated are the values
obtained after phosphocellulose chromatography.








three-stranded structure known as a D-loop can be catalyzed by recl

protein. Most molecular models describing the mechanism of recombina-

tion include the formation of D-loops as a vital step (Meselson and

Radding, 1975; Wilson, 1979). Taken together these observations

suggest that the recl protein may function directly in recombination

pathways. An interesting observation concerning the formation of

D-loops is that ATP is not an absolute requirement for the reaction.

For these lower levels of D-loop formation, the recl protein may

utilize the energy of the superhelix, thereby bypassing the need for

ATP.

Evidence for the role ATP plays in contributing an energy source

for such pairing reactions catalyzed by recA protein is somewhat

dichotomous. Early studies on strand uptake reactions indicated that

ATP was an absolute requirement (Shibata et al., 1979). Recently, it

has been reported that the energy generated by the hydrolysis of ATP is

used as the enzyme dissociates from a D-loop molecules (Shibata et al.,

1982a). The strand uptake reaction may rely at least in part on the

energy of the superhelix (Shibata et al., 1982b).

The formation of joint molecules catalyzed by the recl protein

occurs at a much lower stoichiometry than those reactions catalyzed by

recA protein. Calculation based on D-loop formation data indicates

that one recl monomer is required per 200 nucleotides of single

stranded DNA (Kmiec and Holloman, 1982). The recA protein-DNA nucleo-

tide ratio is approximately one to three (Shibata et al., 1979). Part

of this difference may be attributed to the larger size of the recl

protein (70,000 daltons). However, a more interesting possibility is

that the recl protein may possess an enzymatic feature absent in the








recA protein. For example, the amount of recA protein required for

pairing reactions can be substantially reduced if single strand binding

protein is present (Soltis and Lehman, 1983). Perhaps the character-

istic reaction requisite of SSB in pairing reactions promoted by recA

protein is inherent in the recl protein.

Mutations in the recl gene produce phenotypes are deficient in

recombination, repair and cell division (Holliday et al, 1976).

Because of the similarities to recA mutant phenotypes (Clark, 1974;

Gottesraan, 1981; Roberts et al., 1978), it is attractive to postulate a

role for the recl protein in certain cellular regulation pathways. At

this time no regulatory function, such as an endopeptidase activity,

has yet been associated with the recl protein. However, in the absence

of a specific substrate, such studies are particularly difficult.

Recombinational deficiencies in the recl mutation include a lack

of gene conversion events (Holliday, 1967). Although the molecular

basis for gene conversion is not completely understood, it is generally

believed that a heteroduplex molecule is first created followed by a

correction process known as mismatch repair. The formation of this

heteroduplex molecule is nonreciprocal and therefore may begin via an

asymmetrical exchange of DNA strands. The formation of D-loops

catalyzed by the recl protein in vitro is an asymmetric exchange

reaction.













CHAPTER THREE
HETERODUPLEX FORMATION PROMOTED BY USTILAGO REC1 PROTEIN



The formation of a heteroduplex joint is a key feature in the

overall process of reciprocal recombination. This structure, a hybrid

molecule consisting of complementary DNA strands from different

parents, has also been applied to the processes of bacterial conjuga-

tion (Lacks, 1970). The biochemical and molecular events preceding

heteroduplex formation are becoming more evident due mainly to the

characterization of the properties of the recA protein from E. coli

(Weinstock et al., 1979; Shibata et al., 1979). It has previously been

reported that the lower eukaryote Ustilago maydis contains a protein

which promotes the pairing of a variety of homologous DNA substrates in

vitro (Kmiec and Holloman, 1982). One of these pairing reactions

involves the uptake of a single strand fragment by a superhelical

duplex. The joint molecule formed, known as a D-loop, resembles a

molecular intermediate postulated to exist in an array of recombination

models (Meselson and Radding, 1975; Wilson, 1979). However, this

particular strand transfer reaction is somewhat difficult to

quantitate.

Circular, single stranded DNA and linear duplex DNA constitute a

pair of substrates potentially useful in studying the pairing reaction.

This reaction has the advantage that the substrates and the products

can be easily identified and quantified by a variety of appropriate








assays. Recent experiments in several laboratories have led to the

conclusion that the recA protein efficiently transfers a single strand

circle onto a linear duplex in a reaction coupled to the hydrolysis of

ATP (Cox and Lehman, 1981; Cox et al., 1982). The recA protein-

promoted DNA strand transfer reaction occurs in two, experimentally

distinguishable, phases (Cox and Lehman, 1981). The first, known as

synapsis brings the two DNA molecules into homologous alignment (Wu et

al., 1982; Gonda and Radding, 1983). This structure can be measured by

a filter-binding assay using nitrocellulose filters in which duplex DNA

is retained as a function of its single strandedness (Beattie et al.,

1977). The reaction requires ATP, but not its hybrolysis and stoichio-

metric amounts of recA protein (Cox and Lehman, 1981). Synapsis can

take place in the absence of any homologous free ends, although stable

hybrid molecules cannot form under these conditions (Wu et al., 1983).

Topological linkage of such paired molecules by the use of topoiso-

merase can produce stable joint molecules called hemicatenanes in which

the circular single strand is interwound with its complementary strand

(Cunningham et al., 1981).

The second phase of the strand transfer reaction is known as

strand exchange or branch migration. Whereas synapsis occurs quickly,

the assimilation of the circle and the concurrent displacement of its

homolog-strand in the duplex is slow. The branch migration reaction

can be measured using the nuclease, S-1, which digest the displaced

strand of the duplex and therefore is a true measure of heteroduplex

formation. By using this assay, a variety of investigators have

determined that the strand exchange process follows sigmoidal kinetics

(Cox and Lehman, 1981). Furthermore, recA protein promoted branch








migration occurs in a 3' to 5' polarity relative to the minus strand

(Kahn et al., 1981). Presumably, synapsed molecules formed at the

"unfavored end" of the linear duplex are quickly eliminated by interac-

tion with ADP (Wu et al., 1982).

RecA-protein promoted strand exchange is stimulated 5-fold to

20-fold by single-strand binding protein (SSB) (Cox and Lehman, 1982).

In the presence of SSB and ATP, the binding of single stranded DNA by

recA-protein is stabilized and the complex, therefore, is more able to

interact with duplex DNA (Cox and Lehman, 1982; Soltis and Lehman,

1983). This stimulation alters the rate-determining phase of DNA

strand exchange.

The branch migration phase of strand exchange requires the

concurrent hydrolysis of ATP (Cox and Lehman, 1981). Approximately, 10

to 15 ATP molecules are hydrolyzed per heteroduplex base pair (Cox et

al., 1982). ATP hydrolysis actually reflects the dissociation of the

recA protein from the heteroduplex molecule (Ohtani et al., 1982). How

chemical energy is assimilated and utilized in the strand transfer

reaction is currently unknown.

Since the recl protein from Ustilago maydis carries out strand

transfer reactions similar to those described for the recA protein, it

is particularly important to test the eukaryotic protein under such in

vitro recombination criteria. A more detailed understanding of the

mechanisms of strand transfer will lead to a fuller appreciation of

heteroduplex formation promoted by recl protein.

Initiation of Strand Transfer

The formation of joint molecules can be measured in two ways. The

first is the assay used to detect D-loops; that is binding to








nitrocellulose filters in washes of high salt (see Chapter Two). The

second involves the sensitivity of S-1 nuclease digestion of a

displaced strand of DNA during the assimilation of a single strand

circle. When circular single strands of DNA were incubated with

homologous linear duplex in the presence of recl protein and ATP, joint

molecule formation was evident using either assay. The filter binding

assay showed that the formation of joint molecules is immediate and

increase in a linear fashion (Fig. 3-1A). However, the kinetics of

joint Golecule formation as measured by the S-1 nuclease assay are

different. In this assay the labeled DNA molecule is the circular

single strand and the determination is the development of S-1 nuclease

resistance as a function of time. Here sigmoidal kinetics are observed

(Fig. 3-1A). After a lag of several minutes, a rapid rise in the

labeled DNA resistant to S-1 is seen. Yet the maximum level of joint

molecules formed was the same in both assays. The initial high rate of

joint molecules formation, detected by the nitrocellulose filter assay

may indicate that formation of joint molecules can occur in the absence

of extensive heteroduplex.

Heating the reaction mixtures prior to filtration eliminates a

large fraction of the joint molecules formed early in the reaction

(Fig. 3-1B). The formation of heat-stable joint molecules follows

sigmoidal kinetics.

Joint molecule formation can also be measured by agarose gel

electrophoresis (Fig. 3-2). During early reaction times, no product

molecules are seen (lanes d-i). However, after 20 minutes of

incubation a band appeared which migrated near the form II DNA marker

(lane a, upper band). Subsequent joint molecule formation occurred


























Fig. 3-1. Formation of joint molecules assayed by two different
methods.
A. Reaction mixtures of 0.5 ml contained 15 ug/ml recl protein
and either 15 pM 3H-labeled fd form III DNA (1.2 x 104 cpm/nmol)
and 7.5 yM fd circular single stranded DNA or 15 pM fd form III DNA and
7.5 pM 3H-labeled fd circular single stranded DNA (1.6 x 104
cpm/nmol). Aliquots of 40 pl were removed at the indicated times and
formation of joint molecules was measured by retention on nitrocell-
ulose filters (o) or by resistance to S-1 nuclease (*).
B. Aliquots from a reaction mixture as in A were held at 500C or
370C for 4 minutes before filtering through nitrocellulose. The ratio
indicates the fraction of joint molecules surviving treatment at 500C
compared with that at 370C.




















I I I I I I
^1.0 -B-----A-

S0.54


60 -600
A )

L40- -40
o ClT
d-5-
0
C

S20- -20 m
r-



15 30 45 60 75 90 r2-
Time (min)




























Fig. 3-2. Formation of joint molecules measured by agarose gel
el ectrophoresi s.
A reaction mixture of 0.32 ml contained 15 AM fd form III DNA,
7.5 UM fd circular single stranded DNA, and 15 ug/ml recl protein.
Aliquots of 40 pl were removed at the indicated times, mixed with 10 pl
of a solution containing 0.25% bromophenol blue, 0.5% SDS, 25 mM EDTA,
and 20% glycerol, loaded into slots of an agarose gel and electro-
phoresed as described, a) fd form I DNA contaminated with a small
amount of slower moving form II DNA, b) fd form III DNA, c) fd circular
single stranded DNA, d) reaction mixture minus recl protein, e-o)
course of reaction at indicated times (minutes).














a b c de f a h i i k I m n o


0 I 2 5 7510 20304560120


Controls


Time Course


M -- ----q








after 20 minutes of incubation (lanes I-o). The formation of a form II-

like DNA molecule and the formation of extensive heteroduplexed DNA

correlate quite well as a function of time.

Strand transfer reactions required the presence of ATP, although

the nonhydrolyzable ATP analog adenylyl-imidodiphosphate (AMPPNP) could

serve as a adequate substitute when product molecules were measured by

the filtering-binding assay (Kmiec and Holloman, 1983). The level

attained using AMPPNP in place of ATP was three- to fourfold lower, and

the complexes formed in such reactions were unstable and did not

require homologous partners.

Energy Requirements for Heteroduplex Growth

Extention of the initiation complex joint required ATP hydrolysis.

When AMPPNP was added to an ongoing reaction, strand exchange stopped

immediately (Fig. 3-3). The addition of glucose and the enzyme

hexokinase, which catalytically converts ATP to adenosine diphosphate

(ADP) in a rapid manner, also stopped the strand transfer reaction

(Fig. 3-3). This continual requirement for ATP in its hydrolyzable

form suggests that its energy is used by recl protein to drive strand

exchange reactions.

Mechanism of Strand Transfer

The formation of a heteroduplex joint molecule involves strand

transfer of the single strand circle into the duplex linear molecule.

As the circle is assimilated, the homolog duplex strand would be

displaced. Does the displacement of this strand occur in a preferred

direction? That is, is there a favored end to start the strand

exchange reaction and is there a subsequent polarity to heteroduplex

formation?



























Fig. 3-3. ATP requirement in formation of joint molecules.
Reaction mixtures of 0.4 ml contained 15 pg/ml recl protein, 15 UM
3H-labeled fd form III DNA (1.2 x 104 cpm/nmol) 7.5 pM circular
single stranded DNA and either ATP (e), adenylyl-imidodiphosphate (o),
adenylyl-imidodiphosphate and G4 circular single stranded DNA in place
of fd circular DNA (A), or no ATP (&). Aliquots of 40 1l were removed
and formation of joint molecules was measured by the nitrocellulose
filter assay.





















8

zO

0
+--
+--





O
)1
mu


15 30 45 60 75 90
Time (min)








To first explore the role of ends in the pairing of single strand

circle and linear duplexes, DNA from the chimeric phage M13 Goril was

used (Kaguni and Ray, 1979) (Fig. 3-4). This phage consists of 2216

bases of G4 inserted into the M13 phage genome. A single Xhol restric-

tion site exists in the G4 region, while the M13 region contains a

unique BamHI cleavage site. When M13 Goril duplex linear, cleaved by

the enzyme Xhol, was used in reactions containing recl protein and G4

single stranded circular DNA, high levels of joint molecules were

formed,(Fig. 3-5). However, under the same reaction conditions, but

substituting fd single strand circles very low levels of joint

molecules were formed (Fig. 3-5A). Phage.fd :is 97% homologous to the

M13 DNA sequence (van Wesenbeck et al., 1980). The reverse reaction

using BamHI restriction of M13 Goril DNA which contains M13 sequences

at its termini produced joint molecules only when fd single strand

circles were used (Fig. 3-5B).

These results emphasize the importance of ends in the production

of joint molecules stable under reaction conditions which involved

washes of high salt (see Appendix E). However, this does not suggest

that the strand transfer reaction begins at a unique, preferred end.

To investigate this possibility, duplex linear molecules were made as

described by Kahn et al. (1981). In this procedure duplex linear

molecules complementary to the (+) viral circular single strands

at either the 5' or the 3' end can be created. The restriction enzyme

Hpal cuts M13 Goril DNA in two sites. One site is located within the

M13 region, the other in the G4 sequence. The cut within the G4

sequence produces a fragment 6583 base pairs long with G4 DNA at the 5'

end of the minus strand. The cut within the M13 region produces a






























Fig. 3-4. Restriction maps of DNA from phages M13 Goril and fd.
Sites of cleavage by restriction endonucleases HpaI, Xhol, BamHI,
and Sau96I have been redrawn from the maps shown by Kahn et al. (1981)
and Cunningham et al. (1981).














Ppo I Xho I

G4 2216

MI3Gori I *-Hpa I

M13 6407

BamH I


Sau96





















Fig. 3-5. Formation of joint molecules at a preferred end.
Since the sizes of linear duplex molecules and circular single
stranded molecules varied from reaction to reaction in this experiment,
concentrations indicated refer to moles of the full length molecule.
Calculations were based on the following sizes: 8623 base pairs, M13
Goril (Kaguni and Ray, 1979); 6583 base pairs, M13 Goril Hpal large
fragment (Kahn et al., 1981); 2040 base pairs, M13 Goril Hpal small
fragment (Kahn et al., 1981); 6408 bases, fd DNA (Beck et al., 1978);
5577 bases, G4 DNA-TGodson et al., 1978). A. Reaction mixtures of
320 Ul contained 0.6 nM 3H-labeled M13 Goril form III DNA (prepared
with Xhol) and either 0.6 nM G4 or fd circular single strands, plus
12 Ug/ml recl. B. Reaction mixtures of 320 pl contained 0.6 nM
3H-labeled M13 Goril form III DNA (prepared with BamHI) and either
0.6 nM G4 or fd circular single strands plus 12 ug/ml recl.
C. Reaction mixtures of 320 pl contained 0.6 nM 3H-labeled M13 Goril
DNA Hpal large fragment, and either 0.6 nM G4 or fd circular single
strands, plus 6 ug/ml recl. D. Reaction mixtures of 320 pl contains
0.6 nM H-labeled M13 Goril DNA Hpal small fragment, and either
0.6 nM G4 or fd circular single strands, plus 6 ug/ml recl. Aliquots
of 40 1l were removed at the indicated times and joint molecules were
determined using nitrocellulose filter assay. (o) G4 DNA, (*) fd DNA.















































20 30 40
Time (min)


4-

eo
0








fragment 2040 base pairs long with M13 DNA at the 5' end of the minus

strand. When the larger fragment was reacted with either G4 or fd

single stranded circles, joint molecules were formed only with the G4

single stranded circles (Fig. 3-5C). The fd single stranded circles

paired only with the smaller duplex fragment (Fig. 3-5D). Taken

together these results indicate that single stranded circular DNA is

transferred onto a homologous linear duplex to create a stable

heteroduplex by pairing with the 5' end of duplex complementary strand.

Accompanying the base pairing of the circle with the 5' end of the

duplex minus strand is the concurrent displacement of the 3' end, and

subsequently the entire length of the positive duplex strand (Kmiec and

Holloman, 1983).

Discussion

The recl protein catalyzes the homologous pairing of circular

single stranded and linear duplex DNA in a reaction occurring in at

least two distinguishable phases. The first involves synapsis of the

two DNA molecules with the alignment of homologous sequence, but little

heteroduplex formation. The reaction is dependent on the presence of

ATP as well as DNA homology. In the second phase of the reaction,

strand exchange, the nascent heteroduplex lengthens to form a stable

joint molecule. This progressive branch migration is slower than

synapsis, requires ATP hydrolysis and occurs in a polar direction. The

mechanism by which the two DNA molecules become homologously aligned is

currently unknown. In recA protein promoted reactions, the enzyme

polymerizes first on single stranded DNA then processively searches the

recipient duplex molecule for complementary sequences (Gonda and

Radding, 1983). Once aligned the formation of heteroduplex DNA








proceeds in a 3' to 5' polar direction, with respect to the minus

strand of the duplex (Kahn et al., 1981; West et al., 1982).

Remarkably, its direction is opposite to that of recl protein-promoted

strand transfer reactions.

The polar movement of strand transfer reactions may be a function

of the recl protein's unwinding activity. Although no direct evidence

has been produced showing that this enzyme unwinds duplex DNA, it can

be assumed that this activity takes place prior to homologous pairing

in D-loop formation. There are a number of DNA enzymes whose activity

involves polar motion in relation to the duplex DNA. These include DNA

polymerases and in particular, DNA helicases (Yarrington and Gefter,

1979).

The significance of the polar movement inherent in heteroduplex

formation reactions promoted by recl protein in vitro is unclear. In

meiotic recombination, heteroduplex DNA, once formed seems to grow in a

polar fashion (Rossignol et al, 1978). In fungi, polarity in recombi-

nation events has been reported (Catcheside and Angel, 1974). In yeast

there is evidence indicating that recombination begins at a defined

site and extends into a locus from a particular direction (Fogel et

al., 1978).

In recA-protein promoted strand exchange reactions, formation of

D-loops, nascent heteroduplexes, does not exhibit a polarity (Cox and

Lehman, 1981). Wu et al. (1983) found that nascent heteroduplexes may

occur at either end of the linear duplex molecule, but under the

regulatory mechanism of ADP only those at the favored end persist,

eventually elongate into a stable recombinant molecule. The control of

heteroduplex formation promoted by recl protein may also involve ADP





52


since this molecule is an inhibitory reaction component to the

extension of heteroduplex joints.













CHAPTER FOUR
TOPOLOGICAL LINKAGE OF CIRCULAR MOLECULES CATALYZED
BY REC1 PROTEIN AND TOPOISOMERASE


It is currently believed that the mechanism of genetic recombina-

tion involves the breakage and reunion of homologous chromosomes

(Tayloi, 1965). This concept poses a fundamental and still largely

unanswered question. Does recognition of homology precede breakage or

follow it? To produce the cross-stranded-structure, known as a

Holliday structure (1964), molecular events have been proposed which,

by and large, embody single strand breaks (Hotchkiss, 1974; Meselson

and Radding, 1975), because generation of single strand ends prior to

synapsis avoids topological problems. Furthermore, free single strand

(Benbow et al., 1974) and double strand (Szostak et al., 1983) ends

have been observed to be recombinogenic. The discovery of a group of

enzymes known as topoisomerases which can relieve the topological

constraints of interwound DNA molecules (Kirkegaard and Wang, 1978)

have strengthened the theory that pairing precedes breakage (Wilson,

1979; Kikuchi and Nash, 1979).

The recl protein from Ustilago maydis pairs single stranded

circular DNA with homologous linear duplex in a two step reaction. The

first, known as synapsis, conjoins homologous DNA sequences while the

second, strand exchange, slowly increases the heteroduplex length

(Chapter 3; Kmiec and Holloman, 1983). In those studies, the impor-

tance of homologous free ends in the creation of stable heteroduplex

molecules was established. However, the possibility still existed that








synapsis began in regions of homology lacking free ends and then

incorporated free ends to increase the stability of the joint

molecules. The synapsis of the DNA molecules and the search for

homology comprise a set of reactions which in vitro correlate with the

initiation of homologous reciprocal recombination in vivo. To explore

this possibility, experiments were designed to test the ability of recl

protein to homologously pair molecules lacking free ends. Such a

synapsed pair of molecules would be subject to catenation by the

enzymet topoisomerase. Successful enzymatic pairing of two circular

homologous molecules followed by topological linkage would demonstrate

that during the initiation of recombination pairing precedes breakage.

Homologous Pairing and Topological Linkage of Single Stranded Circles
and Closed Circular Duplex DNA

E. coli recA protein has been observed to pair single stranded DNA

with homologous duplex DNA in a side-by-side fashion (Wu et al., 1983).

Previously, Cunningham et al. (1981) showed that molecules paired in a

side-by-side manner could be topologically linked after addition of E.

coli topoisomerase I. Circular single stranded DNA, homologous

superhelical DNA, ATP and recl protein were incubated together for 45

minutes. This was followed by addition of Ustilago topoisomerase I

(Rowe et al., 1981) and the product molecule identified by sedimenta-

tion in gradients of alkaline sucrose. As illustrated in Figure 4-1C,

linked molecules comprised of [3H]-labeled superhelical duplex DNA

and [32P]-labeled single stranded circles were observed to sediment

between each of the parent molecule populations. No product was

observed in reactions lacking recl protein, ATP or topoisomerase

(Fig. 4-1A, B, D). Nonhomologous combinations of DNA were ineffective


























Fig. 4-1. Linkage of single-stranded circles and form I DNA.
Complete reaction mixtures of 100 pl contained 35 mM Tris-HC1,
RH 7.5, 10 mM MgCl2, 1 mM ATP, 1 mM dithiothrietol, 10 yM
H-labeled fd form I DNA, 10 pM 32P-labeled fd circular single-
stranded DNA, and 40 vg/ml recl protein. After 45 minutes at 370C 20
units of topoisomerase were added and incubation was continued for an
additional 30 minutes. The reactions were stopped by addition of EDTA
to 0.1 M and NaOH to 0.3 M. Mixtures were layered on top of alkaline
sucrose gradients and the gradients centrifuged for 2 hours at 45,000
rpm in a Beckman SW50.1 ti rotor. (A) minus recl protein (B) minus
topoisomerase (C) complete reaction (D) minus ATP (E) 32P-labeled
X174 DNA (2.8 x 104 cpm/nmol) in place of fd single stranded DNA
(F) control, DNA alone. (*) 3H-labeled DNA. (A) 32P-labeled
DNA.



































0



I



o
0'


Fraction Number







substrates for joint molecule formation (Fig. 4-1E). The sedimentation

pattern observed for the product joint molecule is expected due to the

increase in the frictional coefficient of the catenane.

To analyze the structure of the product made when circles and

superhelical duplexes are linked, we once again employed the chimeric

phage M13 Goril (Kaguni and Ray, 1979). Using [32p]-labeled G4

phage DNA and [3H]-labeled M13 Goril superhelical duplex DNA (form

I), recl protein and topoisomerase acting in concert created joint

molecules topologically linked within the G4 stretches of the M13 Goril

DNA as determined by alkaline sucrose gradients. Approximately 27% of

the input M13 Goril form I DNA became completed with G4 phage DNA.

Restriction enzyme digests of the joint molecule, after isolation and

renaturation, within the G4 region of M13 Goril dissociated the

complexes. Restriction enzyme digests inside the unpaired M13 region

did not dissociate the complexes. Therefore, the structure of the

joint molecules containing single stranded circles and superhelical

duplex is a hemicatenane (Cunningham et al., 1981) in which the circle

is linked to the duplex through interwinding with its complement in the

duplex (Kmiec et al., 1983). When the extent of the heteroduplex

formation was measured by S-1 nuclease digestion approximately 32% of

the input G4 single strand circle became resistant (Table 4-1). This

amount is equivalent to the formation of a length of 1780 base pairs,

accounting for nearly the entire length of the G4 insert in M13 Goril

DNA.

Reaction Requirements
Since the energy from superhelix formation or from ATP hydrolysis

might be expected to drive the catenation, reactions were carried out








TABLE 4-1

Extent of Heteroduplex Formation


Enzymes Added DNA Resistant to S-1 Digestion

Topoisomerase Recl Protein % Length in Nucleotides
+ <0.5 <30

+ 2.0 110

+ + 32.0 1780


Reaction mixtures of 100 pU containing 35 mM Tris-HCI, pH 7.5,
10 mM MgC12 1 mM dithiothrietol, 1 mM ATP, 50 pM M13 Goril form I
DNA (8 x 103 cpm/nmol), 25 kM G4 single stranded circular DNA (2 x
104 cpm/nmol), and 15 ug/ml recl protein were incubated at 370C.
After 20 minutes 200 units/ml topoisomerase was added and incubation
continued for 15 minutes. The reaction was stopped by addition of
0.4 ml of 0.2 M sodium acetate buffer pH 4.5, 2 mM ZnCl2 and 20 units
of S-1 nuclease, incubation was continued for 30 minutes, then 0.25 ml
of 1 mg/ml carrier DNA and 0.25 ml of 10% trichloroacetic acid were
added. The precipitates were washed onto Whatman GF/C glass fiber
filters and the filters washed with 10 ml of 5% trichloroacetic acid,
2 ml ethanol, then dried and the 32p radioactivity bound was
determined.








using relaxed closed circular duplex DNA and the nonhydrolyzable analog

adenylyl-imidodiphosphate. Form I DNA, relaxed by topoisomerase and

freed of protein by phenol extraction, was an effective substrate for

catenation (Fig. 4-2B). The ATP analog could substitute as a cofactor

in the reaction, although the hemicatenanes were formed without regard

to homology (Fig. 4-2C). This result concurs with earlier observations

(Chapter Three) which indicated that joint molecules could form without

regard to homology when the analog was used (determined by the filter

binding assay).

Catenation of Intact Homologous Duplex DNA Molecules

Circular duplex molecules can also be linked, albeit, independent

of homology by type I topoisomerases (Tse and Wang, 1980; Brown and

Cozzarelli, 1981). These joined molecules are catenated like links in

a chain (full catenane; contrasted with hemicatane). Homology

dependent topological linkage of duplex DNA molecules has not been

reported. When [32P]- and [3H]-labeled M13 superhelical DNA was

reacted with recl protein, ATP and topoisomerase either concurrently or

added in succession, no catenated molecules were observed to form.

Previously, it was shown that recl protein catalyzes the uptake of

a homologous single strand fragment by superhelical DNA to form a

D-loop structure (Kmiec and Holloman, 1982). Since this joint molecule

could be an intermediate in recombination and a possible starting point

for crossover junctions between two duplexes, an experiment was

designed in which D-looped DNA duplex was catenated with superhelical

duplex DNA. D-loop DNA can be prepared by either thermal annealing

(Beattie et al., 1977) or by recl protein catalysis. [3H]-labeled

and [32P]-labeled fd form I DNA was reacted with recl protein and

























Fig. 4-2. Linkage of single stranded circles with duplex circular DNA
in the absence of homology or superhelicity.
Reaction mixtures of 100 ul containing 35 mM Tris-HC1, pH 7.5,
10 mM MgCl2, 1 mM dithiothrietol, 1 mM ATP or adenylyl-imidodi-
phosphate, 10 aM 3H-labeled fd form I DNA or form IV DNA (form I
relaxed with topoisomerase), 10 uM 32P-labeled fd or G4 single-
stranded circular DNA, and 40 ug/ml recl protein were incubated at 370C
for 20 minutes. After topoisomerase was added to 200 units/ml incuba-
tion was continued for 20 minutes. Reactions were stopped by addition
of 50 ul 0.25 M EDTA and 5 pl 10 M NaOH. Mixtures were loaded on
alkaline sucrose gradients and centrifuged at 45,000 rpm for 2 hours in
a Beckman SW50.1 ti rotor. Reaction mixtures contained (A) fd circles,
fd form I DNA and ATP (B) fd circles, fd form IV DNA and ATP (C) G4
circles, fd form I DNA and adenylyl-imidodiphosphate. (*)
3H-labeled DNA. (A) 3 P-labeled DNA.









































10 15 20
Fraction Number


b 4

E 3
a-
2
I
T


iro
20 o
X
15 E
CL
0
IO~

5 Q
rE1








ATP for 30 minutes af 370C followed by incubation for 30 minutes with

topoisomerase. A catenated pair of duplex molecules should collapse in

alkali and sediment as a dimer, faster than denatured monomeric form I

(West et al., 1982). When D-loop DNA was used, made by either thermal

annealing or recl protein catalysis, a rapidly sedimenting peak of DNA

containing both [3H]- and [32p]-labels was observed to form

(Fig. 4-4). This population of molecules is probably two DNA duplex

molecules linked together. The product of this reaction was charac-

terized further and will be described below. Interestingly, in

reactions where the topoisomerase was added with the recl protein and

ATP, no dimer size DNA molecules were produced. This can be attributed

to the fact that relaxation of form I DNA is fast and produces duplex

molecules incapable of D-loop formation. These results indicate that

the formation of D-loop molecules is a prerequisite for the catenation

of two duplex molecules.

The cellular process of transcription naturally produces DNA

molecules of three-stranded configurations. With increasing super-

helicity, both the level of transcription and the stability of RNA-DNA

hybrids (R-loops) are increased (Wang, 1974). RNA polymerase was added

to reactions containing [3H]- and [32P]-labeled form I DNA of

M13, recl protein and all four ribonucleotides, followed by addition of

topoisomerase. The concerted action of three DNA enzymes produced

diner size DNA hybrid molecules containing both the [3H]- and the
[32p]-labels (Fig. 4-4). Addition of 20 pM rifampicin completely

eliminated the formation of joint molecules. The level of dimer sized

molecules formed in part by the action of RNA polymerase was at a lower

level than those achieved using D-loop DNA. This was probably due to a

























Fig. 4-3. Linkage of form I DNA molecules promoted by homologous
single stranded fragments.
Reaction mixtures of 200 pl containing 35 mM Tris-HC1, pH 7.5,
10 mM MgCl2, 1 mM dithiothrietol, 1 mM ATP, 15 yM 3H-labeled M13
form I DNA, 15 pM 32P-labeled M13 form I DNA, 10 pM M13 single-
stranded fragments, and 20 pg/ml recl protein were incubated at 37*C.
After 20 minutes 200 units/ml topoisomerase was added and incubation
continued for 20 minutes. Reactions were stopped and the DNA analyzed
by centrifugation in alkaline sucrose gradients as in Figure 5.
(A) minus fragments (B) topoisomerase added simultaneously with recl
protein (C) complete mixture (D) G4 fragments added in place of M13
fragments. (e) 3H-labeled DNA. (A) 32P-labeled DNA.





















4

3
02
xl

<4

3
2


25 5
Fraction Number


4

3
2-

0

4

3

2E
I








contamination of endonuclease activity in the RNA polymerase

(Fig. 4-4C, D). Taken together these results indicate that a three-

stranded structure in either the D-loop or R-loop form can promote the

homologous pairing of two duplex DNA molecules.

Characterization of Complexes Formed Between Intact Homologous Circular
Duplexes

Based on the results obtained from the characterization of

hemicatenanes formed between single stranded circles and superhelical

duplex DNA, the possibility that fully catenated molecules were formed

in the duplex-duplex reaction was unlikely. However, experiments were

designed to directly test this possibility and to elucidate the

structure formed in such a pairing reaction. [3HJ-labeled M13 Goril,

[32p]-labeled G4 form I and G4 single stranded fragments were

prepared. These substrates were reacted in a mixture containing recl

protein, ATP and then topoisomerase. Dimer-sized product molecules

were purified by centrifugation in a neutral sucrose gradient

(Fig. 4-6A). To rule out the possibility that the duplexes were joined

as full catenanes (Fig. 4-6D), the complex was cut at the unique BamHI

site within the M13 Goril DNA or at the PstI site within the G4 DNA

(Fig. 4-7). Because the DNA did not dissociate into full length linear

molecules (Fig. 4-8, lanes g, h), the complex is not in the form of a

full catenane. The change in mobility of the complex after restriction

enzyme digestion is consistent with the pattern we would expect of a

structure composed of two joined circular duplexes that had been cut in

one duplex and then in the other.

Since it was now believed that the joint molecule made between two

duplex DNA supercoils was in the form of a hemicatenane, experiments


























Fig. 4-4. Linkage of form I DNA molecules promoted by RNA polymerase.
Reaction mixtures of 200 pl containing 35 mM Tris-HC1, pH 7.5,
10 mM MgC12 1 mM dithiothrietol, 1 mM ATP, 0.15 mM CTP, GTP, and
UTP, 40 UM 3H-labeled M13 form I DNA, 40 yM 32P-labeled M13 form
I DNA, 40 pg/ml recl protein and RNA polymerase were incubated at 37*C.
After 20 minutes 200 units/ml topoisomerase was added and incubation
continued for an additional 20 minutes. DNA was centrifuged in
alkaline sucrose gradients as described in Figure 5. (A) No RNA
polymerase (B) 0.5 units RNA polymerase, no recl protein (C) complete
reaction, 0.5 units RNA polymerase (D) complete reaction, 1 unit RNA
polymerase (E) complete reaction containing 1 unit RNA polymerase and
20 pM rifampicin (F) complete reaction containing 1 unit RNA polymerase
minus UTP, CTP, GTP. (o) 3H-labeled DNA. (A) 32P-labeled DNA.





67



















20 A D 40
" 15- 30

10- A 20
5- 10


-20- E 40 '0
X
S15 -30 E

< I 20
Z
5 I10


20- c F 40
15- % 30

10 20
5 -10

5 10 15 20 25 5 10 15 20 25
Fraction Number














Fig. 4-5. Analysis of product formed with homologous or heterologous
combinations of form I DNA.
(A) Reaction mixture of 0.5 ml containing 35 nM Tris-HC1 pH 7.5,
10 mM MgC12, 1 mM dithiothrietol, 1 mM ATP, 200 MM 3H-labeled M13
Goril form I DNA, 200 pM 32P-labeled G4 form I DNA, 100 pM
unlabeled G4 single stranded fragments, and 50 pg/ml recl protein was
incubated at 370C. After 20 minutes 300 units/ml topoisomerase was
added and incubation continued for 20 minutes. The reaction was
stopped by addition of 50 mM EDTA, the mixture layered on a neutral
sucrose gradient in 1 M NaCl, and centrifuged at 25,000 rpm for 18
hours in a Beckman SW41 ti rotor. Fractions of 0.4 ml were collected
and aliquots of 10 pl were removed for determining radioactivity.
Fractions 11-14 containing the catenated product were pooled and the
DNA was collected after precipitation with ethanol. (B) The reisolated
catenated product was cleaved with BamHI in a reaction mixture of
100 pl containing 10 mM Tris pH 7.9, 6 mM MgC12, 150 mM NaCl, 30 pM
DNA and 10 units/ml BamHI. After 2 hours at 37 C the DNA was analyzed
by centrifugation in an alkaline sucrose gradient as in Figure 5.
(C) 32P-labeled G4 form I DNA and 3H-labeled M13 Goril DNA
converted to the linear form III by cleavage with BamHI were run as
markers in an alkaline sucrose gradient in parallel with B. In D-F
catenated product was prepared with a heterologous combination of form
I DNAs. The reaction mixture of 200 Ul containing 35 mM Tris-HCl,
pH 7.5, 10 mM MgC12, 1 mM dithiothrietol, 1 mM adenylyl-imidodi-
phosphate, 125 pM 'H-labeled fd form I DNA, 125 pM 2P-labeled G4
form I DNA, and 30 ug/ml recl protein was incubated at 370C. After 20
minutes 250 units/ml topoisomerase was added and incubation continued
for 20 minutes. The reaction was stopped by addition of EDTA and the
product was isolated as in A. Reisolated catenated product was cleaved
with restriction endonuclease PstI in a reaction mixture of 100 pl
containing 10 mM Tris-HC1, pH 7.-4, 6 mM MgC12, 50 mM NaCl, 2 pM DNA,
and 2 units/ml PstI. After 2 hours at 300C the mixture was divided
into two equal portions. To one was added 50 mM EDTA and 1 M NaCl.
After heating at 750C for 10 minutes it was centrifuged in a neutral
sucrose gradient containing 1 M NaCl for 18 hours at 25,000 rpm in the
SW41 ti rotor. The other portion was denatured with alkali and
centrifuged for 2 hours at 34,000 rpm. (D) Reisolated catenated
product centrifuged in a neutral sucrose gradient 18 hours at 25,000
rpm in the SW41 ti rotor. (E) Product cleaved with PstI centrifuged in
a neutral sucrose gradient. (F) Product cleaved with Pstl centrifuged
in an alkaline sucrose gradient. (*) 3H-labeled DNA. TA)
32P-labeled DNA.












































5 10 15 20 25 5
Fraction Number



























Fig. 4-6. Linked pairs of circular DNA molecules.
(A) Fully catenated molecules (B) Hemicatenated molecules with
multiple links (C) Hemicatenated molecules with a single link (D) Fully
catenated molecules (E) Hemicatenated molecules with one pair of
strands interlinked (F) Hemicatenated molecules with both pairs of
strands interlinked (G) Hemicatenated molecules with a single link
(H) Covalently linked molecules.










A


co


DO


Go



H





























Fig. 4-7. Maps of DNA from phages M13 Goril, M13, fd, and G4.
The restriction sites shown have been taken from the data of Beck
et al. (1978), Godson et al. (1978), Kaguni and Ray (1979), and van
Wesenbeek et al. (1980T.












Xhol EcoRI
Hpa1 Xhol EcoRI

M3Gor/ (fd, M13 G4
/ Ps^Xho I
PstI


BamHI


























Fig. 4-8. Analysis by agarose gel electrophoresis of the catenated
product formed between homologous DNA duplexes.
The linked complex of M13 Goril form I DNA and G4 form I DNA was
prepared as described in Figure 4-6A. Each lane was loaded with
approximately 2.5 nmol of DNA. Electrophoresis was carried out as
described in Appendix F. (A) M13 Goril form I DNA (B) M13 Goril DNA
cut with EcoRI (C) G4 form I DNA (DT-G4 DNA cut with EcoRI (ET
reisolated catenated product (F) product cut with EcoRI (G) product cut
with BamHI (H) product cut with PstI (I) product cut with BamHI and
Pstl.








ab


c d







were designed to test this hypothesis. Several structures were

possible, all of which would be defined theoretically as hemicatenanes.

The first (Fig. 4-6E) contains a region in which only one set of

complementary strands from each parent duplex are interwound. An

alternative structure (Fig. 4-6F) would be one in which both sets of

complementary strands are heteroduplexed. To look for the formation

of singly or doubly interwound hemicatenanes, the joint molecules were

made as described using M13 Goril and G4 form I DNA. The catenated

molecules formed in this instance would be heteroduplexed only within

the 2200 base pairs of the G4 section in M13 Goril. A single EcoRI

restriction site lies within this region in both substrates (Fig. 4-7).

When the complex was cut with EcoRI, the molecules resolved into two

full length linear molecules. Therefore, the hemicatenated dimer is

likely to be a structure in which there are two sets of heteroduplexed

strands (Fig. 4-7F).

There is an alternative explanation for the results obtained after

digestion with EcoRI, depicted in Figure 4-6H, and referred to as a

true Holliday structure. To eliminate this structure as a candidate

for the complex, the joined molecules were cut with the restriction

endonuclease BamHI and the products analyzed by centrifugation in

gradients of alkaline sucrose. If the complex was in the Holliday

structure, only linear and circular single strands would be liberated.

The results, depicted in Figure 4-5 (B, C) reveal the liberation of G4

covalently closed duplex circles and full length linear strands. These

results are consistent with the view that the complex is in the form of

a double hemicatenane. Analysis of joined molecules of nonhomologous

partners G4 and fd form I molecules, paired by recl protein in the








presence of adenylyl-imidodiphosphate, and toposiomerase revealed that

this complex is a hemicatenane in which strands are linked but not base

paired (Kmiec et al., 1983).

Discussion

Previous studies on the homologous pairing activity of the recl

protein illustrated that superhelical duplexes and single stranded

fragments, and linear duplexes and single stranded circles could be

united in an efficient manner. The present studies add to the list of

succes-sful pairing partners, circular single stranded and circular

duplex DNA molecules. The significance of these latest pairing

reactions lies in the fact that neither substrate contains free ends.

In addition; both combinations, single strand circle-duplex and

duplex-duplex, require a single stranded stretch of DNA in a region of

homology.

The recA protein of E. coli can pair duplex molecules if one of

them contains a single stranded gap (Cunningham et al., 1980; West et

al., 1981,1982). Using EM DasGupta et al. (1980) showed that single

stranded DNA could be paired to its complement in a homologous duplex

in the absence of free ends. Linkage of such molecules, using E. coli

topoisomerase I, was demonstrated by Wu et al. (1983). These results

coupled with the present observations strengthen the view that pairing

can precede breakage during initiation of recombination.

In recl protein promoted pairing of homologous duplexes several

reaction requirements must be met. Formation of the dimeric DNA

complex, topologically linked, was dependent on the presence of ATP,

single stranded DNA and the order of enzyme addition. The role ATP

plays in the pairing reaction is as yet unknown. One can speculate








that the energy required to unwind the superhelix may be derived from

ATP hydrolysis in the presence of single stranded DNA. Since pairing

reactions were unsuccessful in the absence of single stranded DNA, a

cofactor in recl protein's ATPase activity, this hypothesis is

attractive. When topoisomerase was added to reaction mixtures with the

recl protein, no linked dimeric molecules were formed. This may be a

reflection of the fact that fully relaxed covalently closed duplex

molecules are not good substrates for the uptake of a single stranded

fragment. These results implicate the energy of the superhelix itself

in overall pairing reactions.

The governing role played by single -stranded DNA in the initiation

of recombination has been clearly illustrated (see Radding, 1982 for

review). The present results reinforce this idea. Successful topolog-

ical linkage of duplex DNA molecules relied heavily, not only on the

presence of single stranded fragments, but also on the location of the

three stranded region. It must be pointed out that no direct data has

been presented which unequivocally shows that juxtaposition of the

duplexes begins at a D-loop site. However, it is clear that this

D-loop region is an essential part of the pairing pathway. Futhermore,

heteroduplexed regions within dimeric molecules were present only at

sites previously made three stranded. These results indicate that

pairing of homologous sequences in the absence of a free end can occur;

again suggesting that pairing before breakage is a viable mechanism in

the initiation of recombination.

Although D-loop DNA is a well-known feature of mitochondrial DNA

(Clayton, 1982), it is not difficult to envision three stranded

structures occurring throughout the DNA chromosome. For example,







D-loops might form along linear chromosomes as a result of an uptake of

redundant single strands during DNA synthesis (Lundquist and Olevera,

1982) or from short stretches of undirected replication. The

transcriptional process can generate stable RNA-DNA hybrids (R-loops)

and from the data presented here these R-loops can catalyze the forma-

tion of duplex-duplex complexes. Other laboratories have presented

evidence, using X phage, that chain elongation during transcription is

important in homologous recA-independent recombination (Ikeda and

Matsumoto, 1979). Furthermore, evidence has accumulated from studies

of X phage integration (Davies et al., 1972), mating-type switching

(Klar et al., 1981) and rearrangements during immunoglobulin develop-

ment (Van Ness et al., 1981) suggesting a relationship between

transcription activities and site specific recombination.

Finally, Shibata et al. (1982) suggested that pairing along

homologous regions of chromosomes aids in accurate segregation in

meiosis. This naturally occurring activity may facilitate the creation

of crossover junctions since the chromosomes are in homologous

alignment. Once chromosomes are juxtaposed, the conjoining of the DNA

duplexes to initiate recombination begins. The events which lead to

the synapsis of DNA molecules are presently undefined.














CHAPTER FIVE
SYNPASIS OF DNA MOLECULES PROMOTED REC1 PROTEIN


The synapsis of homologous chromosomes prior to karyokinesis is a

cellular event which assures a proper segration pattern. As'early as

1911, Morgan theorized that such synapsed chromosomes could undergo

genetic recombination. Until recently the forces bringing the DNA

duplexes into homologous register have remained unknown and outside the

realm of biochemistry. The pioneering work on the recA gene and its

product have transformed cytogenetic observations into defined

enzymological processes (for review see Radding, 1982). The eukaryotic

recl protein, has been shown to carry out a variety of DNA pairing

reactions crucial for initiation of recombination (Kmiec and Holloman,

1982, 1983, 1984; Kmiec et al., 1983). Pairing or synapsis of DNA

molecules requires only an energy cofactor and sequence homology. It

was previously established that homologous free ends are not a require-

ment (Kmiec et al., 1983) (see Chapter 4). It was clear, however, that

stable heteroduplex molecules could be formed only in the presence of a

homologous free end. Still a possibility existed that recl protein may

be able to place two DNA molecules into homologous register regardless

of the topological constraints.

In the absence of free ends, the recA protein can pair a single
strand with its complement in a duplex molecule promoting the formation

of a different kind of joint in which the strands are paired, but not

interwound (Bianchi et al., 1983). This so-called paranemic joint is








unstable and has a linking number of zero. In contrast, the plecto-

nemic joint in which the DNA strands are interwound has a linking

number of non-zero and is stable.

Experiments were designed to explore the side-by-side pairing of

DNA molecules promoted by the recl protein in an assay which measured

the initiation complex consisting of the two DNA molecules and the

enzyme. This ternary complex may closely resemble the molecular frame-

work present when recl protein brings the molecules into homologous

alignment. Information gained from this study will contribute to the

understanding of recl-protein promoted DNA synapsis, the initial step

in strand transfer reactions.

Formation of Stable Complexes

It had previously been observed that recl protein could pair DNA

molecules lacking free ends in preparation for topological linkage by

Ustilago topoisomerase (Chapter 4). The initial event of the synapsis

reaction was reasoned to be an interaction between recl protein and the

two DNA molecules. To study the formation of synaptic ternary

complexes, a nitrocellulose assay was used similar to the one developed

in studies on lac repressor (Riggs et al., 1968). This assay monitors

the formation of protein-DNA complexes because only labeled DNA bound

by protein is retained on the nitrocellulose filters. In washings of

low ionic strength buffer, DNA itself is not retained. Calf thymus H-1

histone was used to measure the efficiency and accuracy of the assay.

When 0.1 ug of H-1 histone was incubated under standard reaction

conditions with labeled duplex linear DNA, approximately 80% of the DNA

was retained by the filter (data not shown).








When recl protein was added separately to binding reactions,

linear duplex DNA was bound to a low level, but single stranded DNA was

retained to a much higher level. However, duplex linear DNA could be

retained at high levels when homologous single stranded DNA was

included in the reaction mixture (Fig. 5-1). The stimulated binding of

duplex was dependent on ATP. Treatment with proteinase K dissociated

complexes already formed.

When recl protein was reacted with single stranded DNA and homolo-

gous linear duplex in the presence of ADP, the formation of complexes

was inhibited, confirming an earlier observation in which ADP was found

to inhibit heteroduplex formation early in the reaction.

The type of linear duplex DNA used up to this point has been

completely homologous to the single stranded circle and therefore could

facilitate the formation of stable heteroduplexes. These joint

molecules have been found to be stable under a variety of conditions

including the removal of protein molecules (Kmiec and Holloman, 1983).

Therefore, to study synapsis separately from heteroduplex formation and

to avoid complications arising after recl protein completed synapsis

and began to drive strand exchange, DNA molecules which would be

topologically barred from forming authentic Watson-Crick base pairs

were employed. By cutting the chimeric phage M13 Goril with the

restriction endonuclease Xhol, a duplex linear molecule is created

which contains M13 sequences internally flanked by nonhomologous G4

stretches (Kaguni and Ray, 1979; Godson et al., 1978). In reaction

mixtures where M13 single stranded DNA was used, pairing could only

occur at internal sites without the benefit of a homologous free end.

When recl protein was added to a reaction containing these two DNA


























Fig. 5-1. Nitrocellulose filter assay for synaptic complexes.
A. Reactions (500 pl) contained 7.5 pM M13 single-stranded
circular DNA, 15 PM 3H-labeled M13 linear duplex DNA, 1 mM ATP and
15 yg/ml recl protein (*). Aliquots (50 il) were removed at the
indicated times, brought to 25 mM EDTA, and washed onto nitrocellulose
filters and radioactivity determined. In one reaction the aliquots
were treated with treated with proteinase K (100 g/ml) for 10 minutes
before filtering through nitrocellulose (A). In another reaction
single stranded M13 DNA was omitted (A) or replaced by single stranded
4X174 DNA (0). In other reactions 2 mM ADP was added either initially
(n) or at the time indicated by the arrow (w).















80



60



40



20


20 40 60


Time (min)








molecules and ATP, complexes formed rapidly (Fig. 5-2). However, after

20 minutes, the level of complexes decreased steadily. In sharp

contrast totally homologous M13 duplex linear DNA and M13 single

stranded circles formed complexes with recl protein, but continued

rising to a plateau. When ADP was added to completely homologous

complexes it stopped further complex formation but those already formed

remained stable (Fig. 5-1). ADP rapidly dissociated complexes that

contained molecules paired in the absence of free ends (Fig. 5-2C).

Removal of ADP by regeneration back to ATP (see Appendix E) increased

the level of complex formation (Fig. 5-2D). Dissociation of complexes

could be prevented by adding topoisomerase or adenylyl-imidodiphosphate

(Fig. 5-2B). Stability of synaptic complexes appeared to be greatly

enhanced when true Watson-Crick base pairs were allowed to form.

Paranemic joints, conjoined molecules lacking Watson-Crick base

pairing, are less stable than their plectonemic counterparts (Bianchi

et al., 1983). These data reinforce that observation and extend it by

implicating ADP in the dissociation process.

The size of the paranemic joint formed between circular single

stranded DNA and linear duplex was measured by a simple nuclease

digestion assay. These experiments showed that the formation of joint

molecules occurring in the middle of the duplex is accompanied by

complete unwinding of the homologous region (Kmiec and Holloman, 1984).

Furthermore, these same sutdies indicated that ADP prevented the

unwinding of duplex DNA.

Paranemic Joints Contain Left-Handed Z-DNA

An array of molecular frameworkd could be envisioned for the

structure of the paranemic joint. First, the two complementary























Fig. T-2. Paranemic joint molecules are dissociated by ADP.
A. Reaction mixtures (0.45 ml) containing either 15 yM 3H-
labeled linear duplex M13 DNA (*) or M13 Goril DNA (o), 7.5 pM single-
stranded circular M13 DNA, 1 mM- ATP, and 15 --g/ml recl protein were-
incubated at 370C. Radioactive DNA retained by nitrocellulose filters
was determined from aliquots removed at the indicated times.
B. Reaction mixture (0.85 ml) containing 15 UM 3H-labeled linear
duplex M13Goril DNA, 7.5 pM single stranded circular M13 DNA, 1 mM ATP,
and 15 ug/ml recl protein was incubated at 370C. After 15 minutes the
reaction was split into portions. To one was added 2 mM adenylyl-
imidodiphosphate, and to another was added 200 units/ml topoisomerase.
Aliquots were removed at the indicated times and radioactivity retained
by nitrocellulose was determined. No addition (o); topoisomerase added
(A); adenylyl-imidodiphosphate added (*). C. To a reaction identical
to that in B was added ADP. Control (o); 0.5 mM ADP (A); 2 mM ADP (*).
D. To a reaction identical to that in B was added 10 mM creatine
phosphate and 20 units/ml creatine phosphokinase. Control (o); plus
ATP regenerating system (e).




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 E8CQWGCLF_JZDFPG INGEST_TIME 2012-02-20T22:37:06Z PACKAGE AA00009104_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES