The interaction specificities of selected small molecules with nucleic acids

MISSING IMAGE

Material Information

Title:
The interaction specificities of selected small molecules with nucleic acids
Physical Description:
xiii, 129 leaves. : ill. ; 28 cm.
Language:
English
Creator:
DeStefano, Ronald Phillip, 1947-
Publication Date:

Subjects

Subjects / Keywords:
Molecules   ( lcsh )
Nucleic acids   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 120-124.
Statement of Responsibility:
By Ronald Phillip De Stefano.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000580687
notis - ADA8792
oclc - 14074410
System ID:
AA00003546:00001

Full Text












THE INTERACTION SPECIFICITIES OF SELECTED SMALL
MOLECULES WITH NUCLEIC ACIDS












By

Ronald Phillip DeStefano


A DISSERTATION PRESENTED TO THE GR'JADU.TIL COUiNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLU-NT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA
1973






















To my girls. ... Karen and Kimberly.














ACKNOWLEDGMENTS

The author extends his appreciation to Dr. E. J. Gabbay for

his guidance during the course of these studies. Thanks are also

due Dr. C. S. Baxter for his technical assistance in obtaining the

XL-100 pmr spectra. The encouragement and support of his wife and

daughter are also deeply appreciated.















TABLE OF CONTENTS


Page


ACKNOWLEDGMENTS. . .... .. .. iii

LIST OF TABLES . . . vi

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

ABSTRACT ... . ... .. .. xi

INTRODUCTION. .... . .... ...... 1

RESULTS AND DISCUSSION-PART I. . . ... 27


Proton Magnetic Resonance Studies
Circular Dichroism Studies. .
Viscosity Studies .
Visible Spectral Studies. .
Equilibrium Dialysis Studies. .
Consequences of Intercalation .
The Experimental Approach .
The Association Reaction. .
The Dissociation Reaction .


RESULTS AND DISCUSSION PART II .


. . 78


Ultraviolet-Visible Spectral Studies. . ... 79
Proton Magnetic Resonance Studies (Pmr) . 82
Melting femperature Studies . ... .89
Viscosity Studies . . 91
Equilibrium Dialysis Studies. . ... 94
Circular Dichroism Studies. . .. 98
Significance of the Results . .. 103

EXPERIMENTAL . . . 104


BIBLIOGRAPHY . . .


. 120


APPENDIX . . . .. 126


(Pmr)
. .








Page

BIOGRAPHICAL SKETCH............. ...... ...... 129














LIST OF TABLES


Table Page

1 Summary of Induced Circular Dichroism Spectra for Nucleic
Acid Complexes of 3 and 7 ....... .... . 29

2 Summary of Absorption Spectra of 3 and 7 in Buffer, 95%
Ethanol, and Nucleic Acid Solutions . ... .36

3 Equilibrium Dialysis Binding Studies for Molecules 3 7
with Nucleic Acids of Varying Base Content. ... 37

4 Observed First Order Rate Constants for Association of 3
and 7 with Nucleic Acids. .. . .45

5 Observed First Order Rate Constants for Association of 3
with Synthetic RNA of Repeating Sequence. ... 46

6 Observed First Order Rate Constants for the Formation of
Salmon Sperm DNA-3 Complex at 383rmand 410nm. .. .51

7 Observed First Order Rate Constants for Formation of Salmon
Sperm DNA-3 Complexes at Different Nucleic Acid
Concentrations. . ... 54

8 Observed Rate Constants for Association of 3 with DNA-
Polylysine Complexes. . ... ... .. 61

9 Dependence of Dissociation Rates on the Concentration of the
Nucleic Acid.3 Complex. ... . ...... 66

10 Observed First Order Rate Constants for the Dissociation of
Complexes of 3 with Native, Denatured, and Sonicated Salmon
Sperm DNA . . .. ... ... 69

11 Summary of Absorption Spectra of 3 in Buffer and in Native
and Denatured Salmon Sperm DNA Solutions. ... 70

12 Equilibrium Dialysis Results for Native and Denatured Salmon
Sperm DNA Complexes of 3 and 7. ............. 76

13 Observed First Order Rate Constants for Dissociation of
Complexes of 3 with Nucleic Acids of Differing Base Content 77

14 Summary of the Absorption Spectra of Reporter Molecules in
Buffer, 95% Ethanol, and Salmon Sperm DNA Solution. ... 80







Table Page

15 Chemical Shifts (6) in Hz from the Internal Standard Sodium
2,2-dimethyl-2-silapentanesulfonate (DSS) of Free and
DNA-Bound Reporters at Various Temperatures .. 83

16 Upfield Shifts in the Pmr Signals of the Aromatic Protons
of Reporter Molecules in the Presence of Salmon Sperm DNA 87

17 The Effect of Reporter Molecules on the Temperature of the
Helix-Coil Transition of Salmon Sperm DNA ... 90

18 Results of Viscometric Titrations of Salmon Sperm DNA with
Reporter Molecules. .... . .93

19 Summary of Equilibrium Dialysis Studies on the Binding of
Reporters to Calf Thymus and Salmon Sperm DNA ...... 96

20 Summary of Induced Circular Dichroism Spectra of Salmon Sperm
DNA-Reporter Complexes. . .99


vii













LIST OF FIGURES


Figure Page

1 The DNA Double Helix. . .. .. .. 2

2 Keto-Enol Tautomers for the Bases Guanine and Cytosine. 3

3 Base Pairing Schemes. . . 5

4 Nucleoside Stacking Schemes . .... ... 8

5 Exciton Coupling in DNA . .... 10

6 Intensity Interchange of Coupled Transition Moments 11

7 Possible Open Conformations of DNA. . ... 12

8 Propped Base Pairs at (a) Low pH and (b) High pH. ... 14

9 Structure of Actinomycin D. . 16

10 Intercalating Reporter Molecule .. . 17

11 Ring Current Effect of Aromatic Systems ... 20

12 Schematic Representation of Intercalation Geometries 20

13 Reporter Molecule . ..... .. 21

14 Molecules Incapable of Intercalation. . ... 21

15 Reporter Molecules Used in Kinetic Studies. .. .24

16 Substituted Reporter Molecules. . ... 26

17 Circular Dichroism Spectra of Salmon Sperm DNA Complexes
of 3 and 7 . . 30

18 Viscometric Titrations of Salmon Sperm DNA with 3 and 7 32

19 The Effect of DNA on the Absorption Spectrum of 3 34

20 The Effect of DNA on the Absorption Spectrum of 7 35

21 Relative Dimensions of 3 and the DNA Double Helix 39

viii







Figure Page

22 Oscilliscope Trace Obtained for Dissociation of C.
perfringens DNA'3 Complex . 42

23 First Order Rate Plot for Dissociation of C. perfringens
DNA-3 Complex. . ... ... 43

24 Rate Plots for Formation of Poly dAT-poly dAT-3 Complex
at Varying Phosphate/3 Ratios. .. .. 47

25 First Order Rate Plot for Association of 3 and Salmon
Sperm DNA. . . ... ..... .49

26 First Order Rate Plot for Association of 3 and Salmon
Sperm DNA. . . ... ..... .50

27 Log-log Plots for the Determination of Order with
Respect to Nucleic Acid. . .. 53

28 Possible Models for the Association Reaction of 3 with
Nucleic Acids. . .. 56

29 The Multiple Sites Model for the Association of 3 with
Nucleic Acids. . .. . 57

30 Effect of Varying Ionic Strength on the Association of
Salmon Sperm DNA and 3 . .. ..... 60

31 Rate Plot for Dissociation of Poly dAT-poly dAT-3 Complex 63

32 Rate Plot for Dissociation of C. perfringens DNA-3 Complex 64

33 Rate Plot for Dissociation of M. luteus DNA-3 Complex 65

34 Kinetic Model for the Dissociation of Nucleic Acid-3
Complexes. . . 68

35 Van't Hoff Plot for k and k2 Rate Processes in the
Dissociation of Denatured Salmon Sperm DNA'3 Complex .71

36 Van't Hoff Plot for k and k Rate Processes in the
Dissociation of Native Salmon Sperm DNA-3 Complex 72

37 Scatchard Plot for Equilibrium Dialysis of 3 with Native
and Denatured Salmon Sperm DNA . .. .74

38 Scatchard Plot for Equilibrium Dialysis of 7 with Native and
Denatured Salmon Sperm DNA . .... .75

39 The p-Nitroaniline Reporter Molecule. .. .. 78

40 Partial Pmr Spectra for 12 and Salmon Sperm DNA-12 Complex
at 90C. . .... ... 85

ix







Figure Page

41 Reporter-DNA Complex in the "In" Geometry ... 88

42 Viscometric Titrations of Salmon Sperm DNA with 8 and 8
Followed by Trimethylene Bis(trimethylammonium Bromide). 92

43 Scatchard Plot for Binding of 14 to Salmon Sperm DNA. 97

44 The Induced Circular Dichroism Spectra of the DNA-Bound
Reporters 8, 14, 15, and 17 at High and Low P/R Ratios,
i.e., at 70/1 and 5.7/1. . ... 100

45 The Ten Different Intercalating Sites of DNA. .. 101

46 Synthesis of Nitroaniline Reporter Molecules. ... ... .106

47 Synthesis of Naphthalene Reporter 1 . 111

48 Block Diagram of Stopped-Flow Spectrophotometer System. 118









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

THE INTERACTION SPECIFICITIES OF SELECTED SMALL
MOLECULES WITH NUCLEIC ACIDS

By

Ronald Phillip DeStefano

December, 1973

Chairman: Dr. E. J. Gabbay
Major Department: Chemistry

This research deals with the use of small molecules to probe the

structure of nucleic acids in solution. The first part of the work

deals with the assertion that nucleic acids exist in equilibrium with

a certain percentage of "open" structures even at temperatures far

below the melting temperature of the double helix. A new class of

reporter molecules was utilized to probe this open structure to deter-

mine possible specificity which might arise due to the somewhat lower

stability of sections of the DNA rich in adenine and thymine as opposed

to sections rich in guanine and cytidine. These reporter molecules

were so designed that intercalation in the usual sense of insertion

between base pairs of the intact double helix was impossible due to

steric bulk. The interactions were investigated in detail by subtle

variations in the structure of the probes and also by utilizing a

number of different nucleic acid systems.

The investigations consisted of a series of stopped-flow kinetic

measurements wherein rapid mixing of the reactants and observation of

the developing hypochromic effect on the spectral transitions of the

reporters were used to monitor the complex formation. The use of

xi







sodium dodecyl sulfate (SDS) to dissociate the premade complexes also

allowed the investigation of the reverse reaction. Variations in the

reaction temperature allowed an investigation of the thermodynamic

parameters associated with the reaction.

The data point to the following mechanism: (1) The nucleic acid

double helix, due to natural thermal motions such as twisting of the

phosphodiester chains relative to each other and consequent under-

winding, "cracks" to reveal an open but still stacked segment several

base pairs long which may then unstack in a manner dependent upon its

base content and sequence. (2) The bulky reporters then insert them-

selves between adjacent bases. (3) The strands of the double helix

reclose to form the original native structure, thus trapping the inter-

calated reporter molecule. (4) Kinetic evidence suggests that the

reclosure step is rate-controlling for the association reaction. (5)

Further, the experiments using SDS to dissociate the premade complexes

provide additional information on the interaction. It appears that

the nucleic acids richer in A-T base pairs dissociate more easily than

do those richer in G-C base pairs. This fits in with the experimental

observation that the association is three times faster with poly dAT-

poly dAT than with poly dG-poly dC.

Another line of investigation involved comparing the interactions

of a number of differently substituted E-nitroanilines with nucleic

acids. The techniques of proton magnetic resonance (pmr), circular

dichroism, ultraviolet-visible spectroscopy, viscometry, direct

binding studies, and melting temperature profiles were utilized. The

goals of this research were to investigate the possibility of specificity

arising from the intercalation process and to gain a better understanding







of the interactions of "reporter" molecules in general with nucleic acids.

A model utilizing opposing steric and hydrophobic forces serves to

correlate a great deal of the data. The presence of hydrophobic sub-

stituents tends to cause a more intimate complex to form, while the

increase in steric bulk obtained upon substitution with very large

groups makes the insertion between base pairs more difficult. In

addition, a number of results suggest that there is a certain degree

of specificity in the intercalation reaction. (1) A different circular

dichroism spectrum is seen depending upon the ratio of reporter molecules

to nucleic acid base pairs in solution. (2) The binding constants

decrease as the base pair-reporter concentration ratio drops. (3) The

binding constants depend upon the base content of the nucleic acid as

well as upon its sequence. In general, this class of molecules is found

to be specific for A-T base pairs in the intercalation reaction.


xiii













INTRODUCTION

The determination of the fiber structure of deoxyribonucleic acid

in 1953 by Watson and Crick '2 was undoubtedly one of the most signifi-

3
cant breakthroughs of the century. To quote Watson himself, "the

gene was no longer a mysterious entity whose behavior could be investi-

gated only by breeding experiments. Instead, it became a real molecular

object about which chemists could think objectively in the same manner

as smaller molecules." Their discovery generated a fantastic amount of

research into not only the role of the nucleic acids in the mechanism

of heredity but also the finer points of the structure itself. While

the x-ray structure is indeed very significant, it must also be realized

that DNA in aqueous solution is the real determinant of heredity in the

living organism. It is the determination of the solution structure of

nucleic acids that continues to busy countless researchers around the

world. It is becoming increasingly clear that DNA is a very complex

beast in solution, with eight different structures having been reported.4

While the use of x-ray diffraction has been applied to DNA structure

determination,4,56 serious questions have been raised about the re-

liability of such techniques, as the fibers do not provide enough data

points to unambiguously define a structure. Instead, a combination

of model-building studies and x-ray data is used to fit a structure to

the data. These techniques are undoubtedly open to question in many

instances due to their indirectness.7'8'9








In short, less direct methods of probing the structure of nucleic

acids in solution must, at least for the present, be depended upon.

The x-ray structure should, however, be carefully considered as a

sort of "jumping-off" point for the analysis of solution studies. Even

a cursory examination of the original structure, subsequently refined by

Wilkins,0 reveals the delicately balanced forces at work which might

be expected to cause complex changes in structure upon transfer to an

aqueous environment.

Briefly, the model of Watson, Crick, and Wilkins consists of

two helical arrays wound about a common axis with a right-handed sense.

Two types of forces maintain this geometry and consist of hydrogen

bonding between the complementary bases and hydrophobic stacking forces

generated by the release of solvent molecules due to the clustering

of bases in the nonpolar atmosphere of the helix interior. The gross

structure, shown in Figure 1, thus has both vertical and horizontal

stabilizing forces.

-0-
















Figure 1. The DNA Double Helix. The vertical line is the helical axis,
and the horizontal lines represent hydrogen bonds between
bases.







The bases are somewhat complex in themselves, allowing for a keto-enol

tautomerization equilibrium. The keto tautomers are used in the model,

with the G-C and the A-T base pairs having, respectively, three and

two hydrogen bonds. The model studies of Watson and Crick and also

the chemically determined relations due to Chargaff11 ruled out any

but A-T and G-C base pairs. The connecting chains between bases consist

of sugar phosphodiester links. The sugar moiety is D-deoxyribose and

exists in the furanoside form with two hydroxyl groups at the 3' and 5'

positions of the pentose ring. It was proven somewhat later that the
12
two sugar phosphate chains are oriented antiparallel to each other.2


OH




NHI NH



H HO


Figure 2. Keto-Enol Tautomers for the Bases Guanine and Cytosine.



Since the phosphodiesters are monoanions at physiological p1H's

and the interior of the helix is quite nonpolar, the charged oxygens

are directed outward into the solution. They are not identical, how-

ever--the one being at right angles and the other being parallel to

the helix axis. The charged nature of the phosphates is actually a

destabilizing force, since the charge repulsion between phosphates is

expected to be considerable. Due to the position of substitution of









the sugars on the bases, the distance from one phosphate chain to the

other is not the same on both sides of the helix. In other words, a

major (larger) groove and a minor (smaller) groove are created. The

bases are oriented perpendicular to the axis of the helix, and the dis-

tance between bases is 3.4 angstroms since the pitch is 34 angstroms

with ten bases per turn.13

It can be seen that a number of considerations enter into the

possible consequences of transfer of this complex assemblage to an

aqueous medium. Temperature, pH, counterions, relative humidity, and

even the A-T/G-C ratio have been found to cause structural alteration

to the DNA structure.5'6 For example, it has been found that a number

of A-T rich DNA's (e.g., Cl. perfringens, B. cereus) are quite different

in the solution state from their structure in the fiber state. The

pitch of the helix is at least 10 percent longer in the solution state.

Hydrogen Bonding Forces

While the geometrical constraints of colinear hydrogen bonds and

hydrogen bonding lengths of 2.8-3.0 angstroms would allow for the

formation of twenty-nine possible base pairs connected by two or three

hydrogen bonds,115 it has been found that even rather simple base

derivatives exhibit surprising specificity in the hydrogen bonding

schemes employed in practice. In fact, only three different schemes

have been found in the crystal state. Calculations based on dipole-

dipole interactions have shown that the Watson-Crick scheme is favored

markedly for guanine-cytosine base pairing,6 but adenosine-uracil

pairs can exist in three forms, two of them of approximately equal

energy. Indeed, two crystal structures have been found for substituted

A-T base pairs. Cocrystals of 9-methyl adenine and 1-methyl thymine







have been found to assume the Hoogstein base-pairing scheme (Figure 3b).17

It was postulated by these authors that this structure was not seen in




S -- --1 0- CH




MH----O C0


(a)




-- + 1C


Figure 3.


Base Pairing Schemes.

a) Adenine-thymine and guanine-cytosine Watson-Crick pairs.

b) Hoogstein adenine-thymine pair.

c) Anti-Hoogstein adenine-thymine pair.


nucleic acids due to the fact that the electrostatic repulsion between

strands was lower in the Watson-Crick base-pairing scheme. This asser-

tion was based on the fact that the two phosphodiester chains are

farther apart in the Watson-Crick scheme than they are in the Hoogstein

scheme. Further, adenosine and 5-bromouridine have been found to








assume the anti-Hoogstein base-pairing scheme (Figure 3c).1819 On the

other hand, single crystals of guanine and cytosine have yielded only

the expected Watson-Crick base-pairing scheme.20'21 This may well be

due to the fact that the Watson-Crick scheme has three hydrogen bonds

while the others have only two. Hydrogen bonding also appears to be

quite specific in the presence of aqueous solvent. It was found that

when a given nucleoside was covalently attached to an insoluble support

and a mixture of the other three nucleosides was passed through this

modified support, only the complement to the covalently bound nucleoside
22
was retarded.22 In addition, elution of the column with aqueous urea

instead of water abolished the association, thus strongly implicating

hydrogen bonding as the force responsible for the very specific binding

of the complementary nucleoside. Another study23 utilized proton

magnetic resonance in a variety of solvents to show that base pairing,

as monitored by the downfield shift due to the decrease in electron

density about the proton nucleus, decreases as the hydrogen bonding

ability of the solvent increases. Still another study using proton

magnetic resonance showed that the base pairing was quite specific.24

The N-1 proton of guanosine was shifted drastically only when the comple-

ment cytosine was added, with little effect being seen upon addition of

the other nucleosides. Likewise, the N-3 proton of uridine was shifted

only in the presence of adenine. While these studies show that hydrogen

bonding may be quite specific in organic solvents of various types, it

is hard to imagine that hydrogen bonding plays the major part in

stabilizing the double helical structure. In aqueous solution, the

competition for the hydrogen bonding sites on the bases by water would

minimize the importance of these forces. Further, though hydrogen

bonding is quite specific in nonpolar solvents, it is known that these








same solvents cause the disruption of the double helix. The answer to

the question of the importance of hydrogen bonding in the aqueous

environment lies in the realization that the double helix has a nonpolar

interior region where the bases are located. The aqueous solvent is

excluded from this region due to the hydrophobic nature of the bases, thus

making this region ideal for the formation of extensive hydrogen bonding.

Stacking Forces

A number of observations suggest that stacking interactions are

the main force for stabilizing the double helix. First, it is known

that the double helix can exist even in the presence of high concentra-

tions of protons which might be expected to cause disruption of hydrogen

bonding due to protoliation of the amino groups on the bases. A lowered

temperature to reduce the thermal motion is all that is necessary to

maintain this structure. Vapor pressure studies by Ts'o and coworkers526

showed that the bases do aggregate in water solution and that the associ-
27
ation is not due to hydrogen bonding.27 This type of interaction is

quite complex, as shown by further observations on these phenomena.

Methylation of the bases increases the aggregation, while methylation of

the sugar ring has no effect. However, if the purine nucleoside is

modified by adding an amino group to the 6-position, stacking increases.

Further, 5-bromouridine stacked better than the more hydrophobic

thymidine. A correlation of base stacking ability with polarizability
28
was noted by Hanlon.8 Pmr has also been quite useful for further

defining the stacking phenomenon. It was found that the ring protons of

the bases experienced increased shielding as the concentration of the
29
base was increased. By measuring the differential shielding of the

various positions of the aromatic systems of the bases, a geometry








was assigned to the stacked adenosine molecule.30 The two possible

arrangements are shown in Figure 4.




R







(a) (b)

Figure 4. Nucleoside Stacking Schemes.

a) Alternate Stack

b) Straight Stack.



Extension of these investigations to dinucleoside phosphates has pro-

vided further information on the details of the stacking interactions.
31
Chan and Nelson31 used pmr techniques to show that ApA exists in the

3'-anti-5'-anti right-handed stack. Further, striking differences

have been observed from one dinucleoside phosphate to another regarding
32
their stacking ability.32 It was found that UpU lost its ordered

structure readily upon elevation of the temperature, while ApA's struc-

ture persisted. This was explained as being due to differing solvation

capabilities for the two dinucleoside phosphates.

The stacking phenomenon is also of marked importance in more

complex systems. For example, single-stranded polynucleotides are

highly structured at ambient temperatures, as seen by the observation

that they possess a high degree of temperature-dependent hypochromism
33 34
and optical activity.334 Although originally thought to be due to

hydrogen bonding, the structuring is now known to be due to base











stacking,35,36 It was found that blockage of the hydrogen bonding

sites in polyriboadenylic acid and polyribocytidylic acid did not

abolish the temperature-dependent hypochromism and optical activity.33,34

The evidence for stacking in the double-stranded nucleic acidswill

be discussed in a separate section, as the methods for observing

these effects are not so straightforward.

Electrostatic Effects

The main destabilizing influence in the double helix is the

presence of the phosphate-phosphate ionic repulsions. In fact, it is

known that DNA spontaneously denatures if a minimum concentration of

positive ions is not present to at least partially shield the negative

charges from each other.37 Additional light is shed on the importance

of ionic interactions in the double-helical structures by the fact that

thermal denaturation of single-stranded polynucleotides is not too

dependent upon ionic strength,38 but that of double-stranded poly-

nucleotides is very dependent upon it.39 This is reasonable, as the

interactions between phosphates in the same chain do not change much

upon going from the stacked structure to the random coil form. This

is not to say that the intrachain repulsions are unimportant; it has

been shown that increased salt causes the helix to decrease in length

as evidenced by a decrease in the viscosity of the DNA solution.40

This is presumably due to the decreased repulsion resulting from shielding

of intrachain phosphates from each other, though neutralization of

interstrand repulsions could also play a part.








Electronic Effects

That the nucleic acid bases interact with each other electronical-
41-43
ly is quite well known. The exciton theory41-43 unifies a number of

experimental results and states that, for molecular crystals, energy is

not absorbed by a single molecule but is distributed over many molecules

instead. If the ordered polynucleotide is considered a one-dimensional

crystal, the exciton theory is applicable.







AEL



Figure 5. Exciton Coupling in DNA.



As seen in Figure 5, the excited state energy levels of the bases

are considered to be split to produce two new levels. The lower energy

transition has the electronic vectors of the bases antiparallel,

while the higher energy transition has these vectors parallel.

Quantum mechanical selection rules forbid the transition to the lower

excited level, so the only transition seen is the one of higher energy

than the original uncoupled transition. The absorption spectral maximum

is thus shifted toward the blue end of the.spectrum. Also seen is a

hypochromic effect, with the intensity of the new transition being

lower than that of the monomer. This fact is explained by the theory

of intensity interchange of coupled transition moments and is

illustrated in Figure 6.












I a) Card Stack, vi
v2 transition hyperchromnic
V1 V2


-- --- b) Head-to-Tail, v1 1V V2
V2 transition hypochromic



-- c) Herring bone, No interchange possible
V2


Figure 6. Intensity Interchange of Coupled Transition Moments.





If the transition moments are in a card stack orientation as

put forth in the model, the higher energy transition is hyperchromic,

and the lower energy transition is hypochromic. The technique of

circular dichroism also demonstrates the theory to be true. Contrary

to the situation in absorption spectroscopy, transitions to both the

upper and the lower states are allowed by the pertinent selection

rules. Thus, one sees both a red-shifted band and a blue-shifted

band as compared to the absorption maximum of the monomer. The signs

of the CD bands, however, cannot be predicted without the consideration
44
of more complicated quantum mechanical arguments. It may be stated,

however, that the two transitions are found to be opposite in sign, thus

yielding the very distinctive double Cotton effect for the polynucleotide.

Dynamic Structure of Nucleic Acids

Several lines of evidence point toward a rather remarkable

quality associated with the double-helical polynucleotides. It appears









that the hydrogen-bonded base pairs are in a continual state of flux

with respect to their non-hydrogen-bonded "open" forms. The work of
45-47
von Hippel and coworkers, utilizing the exchange-out of tritium

label has established a firm basis for the existence of the "open" forms

of the nucleic acids. Incubation of the nucleic acids in tritiated

water and rapid separation of the labeled polynucleotide from excess

radioactive solvent via gel filtration enabled these workers to show

that the kinetics of exchange-out of the label were much faster than

for protein systems. Further, it can be shown that the observable

exchange is taking place with the protons involved in the base-base

hydrogen bonding of the double helix. Three possible models were

examined by McConnell and von Hippel.46 These were (1) unstacking

without hydrogen bond breakage, (2) hydrogen bond breakage without

unstacking, and (3) hydrogen bond breakage with partial or complete

unstacking and strand separation. These alternatives are shown in

Figure 7.











(a) (b) (c) (d) (e)
Figure 7. Possible Open Conformations of DNA.

a) Stacked and hydrogen-bonded double helix,

b) Unstacked without hydrogen bond breakage,

c) Hydrogen bond breakage without unstacking,

d) Hydrogen bond breakage and unstacking with partial
strand separation,

e) Complete strand separation.








The evidence gathered from studies utilizing helix-destabilizing

salts such as sodium perchlorate indicated that there exist two possible

open states for the exchange reaction, a stacked one and an unstacked

one. At temperatures far below the melting temperature of the double

helix, the open state induced by thermally caused twisting and unwinding

of the strands relative to each other provides the exchange intermediate.

Since unstacking is minimized, little energy is required for this process.

If, however, the difference between the melting temperature and the

temperature used for the exchange is less than twenty Centigrade degrees,

the stacked form appears to be rapidly and cooperatively formed and

begins to dominate the exchange process. The use of large concentrations

of destabilizing salts was found to have the same effect as raising

the temperature at which the exchange was carried out. The implication

of the destabilizing salts in the disruption of the stacking interactions

between the bases is fairly well grounded. For example, Robinson and
48
Grant48 showed that the solubility of the bases in water was increased

in the presence of these salts in direct proportion to their effect

on the melting temperature of nucleic acids.

The form characterized by minimal unstacking also was shown to

be in equilibrium with the "propped" form which owes its existence to

mismatched hydrogen bonding sites caused by protonation or deprotonation

in the presence of added acid or base. These propped forms are shown in

Figure 8. Several studies 6'49'50 have shown that DNA can suffer

extensive protonation without an accompanying increase in hyper-

chromism. This indicates that the unstacking is minimal.







A H OH NH H C. CH-




(a) (b)



0 H


\ NH O /


Figure 8. Propped Base Pairs at a) Low pH and b) High pH.





Another method of establishing the existence of open forms of

the nucleic acids is based on the fact that formaldehyde reacts with

the single strands at a much faster rate than with the native double

helical form. Haselkorn and Doty1 established that the activation

energy of the reaction with the double helix was much greater than the

energy needed for the formylation of the single strands of denatured

DNA. They postulated that the helix had to be denatured before reaction

could take place. Von Hippel and Wong52 subsequently analyzed this

reaction in detail and found that the comparison of reaction rates for

native and denatured DNA enabled them to separate the reaction into a

conformational alteration and a subsequent chemical reaction. The

difference in activation energy between the two was attributed to the

conformational alteration as denatured DNA does not have this added

barrier to cross. Via this method, an estimate of nine base pairs was

advanced as the size of the "opening unit."








Further studies which are not unrelated have to do with the mode

of action of the antibiotic actinomycin. The structure of this drug is

shown in Figure 9. In a very thorough study, Mueller and Crothers53

used a host of different physical techniques to delineate the forces at

work in the complex of actinomycin and DNA. It was found that the very

bulky molecule was intercalated between base pairs of the nucleic

acid. Further, the nucleic acid was postulated to be altered in

conformation as a result of the complexation. It was shown that the

kinetics scheme was very complex, with several rate constants observable

in both the forward and reverse reactions. In addition, the possibility

of specific electronic interactions was considered by investigation of

the binding behavior with a number of different aromatic systems. It

was found that such specificity does indeed exist, and very subtle

alterations to the aromatic nucleus could drastically alter the affinity

for the antibiotic.

The Small Molecule Approach

The question naturally arises as to how one goes about investigating

the properties of a molecule as large and complex as DNA or a large

protein. The technique of attaching an innocuous and easily monitored
54,55
probe to these large molecules was pioneered by Koshland,5455 who

attached a p-nitrophenol group to the enzyme chymotrypsin. He was able

to draw conclusions regarding the enzyme conformation by monitoring the

concomitant spectral changes of the p-nitrophenol chromophore. This

technique has been applied to the problem of nucleic acid solution

structure by Gabbay and coworkers.56-59 Figure 10 illustrates the

general type of molecules utilized in these studies. It is actually a

diammonium salt substituted with a planar group capable of intercalation.






































CH3


Figure 9. Structure of Actinomycin D.


Lpro

D vql


Lthr


H3









R2 -2X-

02
R3
N + +
R (CH2) nN(CH 3)2 (CH2 Nm (CH3 )3



Figure 10. Intercalating Reporter Molecule.




This process will be discussed later in detail, but it essentially consists

of inserting the planar aromatic system between the base pairs of the

nucleic acid. The design of these molecules was based on the extensive

investigations into the interactions of simple diammonium salts with

nucleic acids. These simple salts were found60 to stabilize the double

helix with respect to the thermal melting into two separate strands.

Gabbay 61,62 studied these interactions in detail and found that

the salts were bound to adjacent phosphate anions on the same strand.

This conclusion was based on several lines of evidence. First, maximum

stabilization against thermal denaturation of the nucleic acid double
+ +
helix is found at n = 3 for the salts H3N(CH2) NH3 2 Cl If, however,

the ammonium protons are replaced by various alkyl groups, the maximum

stabilization effect occurs at n = 2. In addition, both the simple

diammonium salts and the alkylated derivatives were found to inhibit

the RNAse-catalyzed hydrolysis of single-stranded polyribonucleotides.

The effect was again maximized at n = 3 for the simple salts and at

n = 2 for the alkylated derivatives. These results are consistent

with binding to adjacent phosphates on the same strand. In the alkylated

derivatives, the steric requirements introduced by the bulky groups on









each end would force the molecule to be fully extended, while the simple

salts might be expected to be somewhat more compact. The fully extended

distance between the two positive charges for the case where n = 2 agrees

very closely with the intrastrand phosphate-phosphate distance, while the

interstrand phosphate-phosphate distance is much too large for effective binding.

This type of molecule was further modified by Gabbay56,57,59 to in-

clude the aromatic chromophore, and it was found to have very interesting

properties. The added interactions were seen to be those typical of an

intercalated system. Dyes such as acridine orange and proflavine have

been postulated to bind to nucleic acids by an intercalation mechanism

involving insertion of the planar dye molecule between the base pairs of

the nucleic acid.63 A number of observations suggested this mechanism.

Flow dichroism and polarized fluorescence studies indicated that the dyes

were lying at right angles to the helical axis. The viscosity of the

complex solution might be expected to increase due to the expansion in

the length of the nucleic acid required to accommodate the dye molecule,

and this is found to be the case.63,64 X-ray studies also show the mass

per unit length of the complex is less than that of the nucleic acid alone.65

Since the density of the intercalated molecule is greater than that of

the solvent which may be displaced, the only possible explanation is

that the polynucleotide has been extended.

In the case of the reporter molecules, the expected viscosity

increase is also seen.66 In addition, an abundance of other data

also supports the intercalation model. First, the visible spectrum of

the reporter is drastically altered in the presence of nucleic acids,

with the expected red shift and hypochromic effect both being in

evidence. The hypochromic effect is explained by the intensity interchange








theory discussed previously, with the transition moment of the reporter

being coupled to the transition moments of the nucleic acid bases

adjacent to it. The red shift is caused by the presence of nearby

negatively charged phosphate groups of the nucleic acid. It was shown by

Gabbay 56 that a positive charge placed near the aromatic system

of the reporter molecule caused a blue shift in the spectrum, so the

presence of a negative charge might be expected to counteract this

effect and cause a red shift in the absorbance maximum. Pmr studies67

show complete broadening of the ring protons of the reporter, but

sharpening of the signals results upon raising the temperature. The

resonances in the spectrum obtained at elevated temperatures remain

shifted upfield, indicating that the reporter is still bound to the nucleic

acid. The nucleic acids, being quite large, tumble very slowly, and the

local magnetic fields about the protons of the small molecule are not

effectively averaged. This behavior is well grounded in theory, and it

was used by Fischer and Jardetsky68 to determine the portions of the

small molecule (penicillin G in their case) which were involved in the

binding to the large molecule (bovine serum albumin). By noting the

portions of the small molecule's pmr spectrum which were broadened

the most by the presence of the larger molecule, the binding site

could be established. Likewise, all the protons of the DNA molecule

itself are broadened due to the very rigid structure.6 The upfield

shift is due to the ring current effect of the DNA bases above and below

the reporter as shown in Figure 11.

By far the most detailed evidence on the reporter-nucleic acid

interaction was the result of circular dichroism studies by Gab-

ba5659 The reporter molecules do not exhibit optical activity of
bay The reporter molecules do not exhibit optical activity of
























Figure 11. Ring Current Effect of Aromatic Systems.




their own, but the presence of the asymmetric binding site in nucleic

acids causes an induced optical activity. Due to the fact that the

sign of the induced CD effect differed between the DNA complex and the

RNA complex, it was postulated that two different geometries were

being seen. These are pictured below in Figure 12. The intercalation

was postulated to take place from the minor groove of the nucleic acids





SNO O0


2


IN OUT
Figure 12. Schematic Representation of Intercalation Geometries.
+ +
R=(CH2)2N(CH3) 2(CH2)3N(CH3) 2Br

for several reasons. The major groove of the DNA molecule contains

the methyl groups of thymine, and the minor groove of RNA contains the







added 2'-hydroxyl group of the ribose sugar system. Thus, the in

geometry could only logically be assumed to be in the minor groove of

DNA, which is less hindered than the minor groove of RNA. Further

evidence resulted from consideration of the molecule shown in Figure

13. This system possesses two distinct transitions, the 4-nitroaniline

at 440 nm and the naphthylamine at 322 nm. With DNA, both transitions

NO2



S + +
HN-(CH 2N(CH32 (CH2 3N(CH3)
Figure 13. Reporter Molecule.

were made optically active and yielded a CD signal. With RNA, however,

only the nitroaniline transition was made optically active, thus suggest-

ing that the fused phenyl ring points in in the DNA complex and out in

the RNA complex.

Some additional evidence regarding the intercalation model is of

an intuitive nature and relies on the idea that only so "thick" a

molecule can be expected to fit into the "niche" in the nucleic acid.

Indeed, molecules 1 and 2 below in Figure 14 were made by Gabbay and

coworkers771 and found not to intercalate, though they were found to

be bound to the nucleic acids.


+
(CH2)2NH3 NO2

Q Q 2Ci '2Br-
-CON(CH3 2
+ + +
1 (C 2)NH3 2 H-N(CH2) 2N (C3)2 (CH2 3N(CH3) 3


Figure 14. Molecules Incapable of Intercalation.








It has also been shown that the reporter molecules are capable

of exhibiting a certain degree of specificity. Gabbay and Gaffney59

were able to show that the magnitude of the induced CD signal was a

function of the base content of the nucleic acid. While these results

could be due to steric hindrance by the third hydrogen bond of the G-C

rich DNA's, it is also possible to invoke more sophisticated interactions

such as electronic energy differences among the different intercalating

sites. Another case in point deals with the methods whereby proteins

selectively interact with nucleic acids. It has been shown that aromatic

amino acids are capable of intercalation between base pairs of the

nucleic acids. Specifically, it has been shown that the aromatic

amino acids cause destacking of the polyadenylic acid molecule as

evidenced by downfield shifts for the protons of the adenine residues
72
upon complexation.72 Both pmr data and the fact that the 280 nm peak

in the circular dichroism spectrum of the polynucleotide decreased in

magnitude, indicating some destacking, were used to demonstrate

intercalation by tyramine, tyrosine, and tryptamine.7374 That the

intercalation of aromatic residues of a protein may contribute to the

selectivity exhibited by such molecules in their interactions with nucleic

acids is quite an intriguing suggestion. Indeed, detailed studies with

a large number of di- and tripeptide systems have indicated that

specificity may be influenced by the inclusion of aromatic amino acids

in the peptide system.75

Statement of Problem I

As stated previously, several lines of evidence make it clear

that "open" forms of the nucleic acids exist at temperatures far below

the melting temperature. In addition, the studies on actinomycin point








to the fact that even very large aromatic systems with quite bulky

substituents can be bound to nucleic acid systems via an intercalation

mechanism. Further, the possibility of specificity derived from the

stacking interactions in an intercalated complex has been demonstrated.

From consideration of these facts, it is quite apparent that these

open forms could be quite important for a numberof biologically important

processes. Since the bases of the nucleic acid are much more available

to solvent, other small molecules may also be able to interact in a much

more intimate and specific way with the macromolecule. Carrying this

theme a bit farther, the open form could be very sequence dependent

in its opening reaction or its geometry. Indeed, a very obvious way to

study such forms is to use the small probe approach so successfully

applied to other facets of the nucleic acid solution structure problem.

Unfortunately, the open forms are expected to be present in very small

amounts. Thus, methods which judge various structures only in proportion

to their relative amounts are not too useful. Any effects due to the

open structures will be swamped out by the very large excess of "normal"

forms. All of the methods discussed previously fall into this class.

One must use probes specific for the open forms alone. The molecules in

Figure 15 were made by Gabbay and Baxter76 and shown to behave in this

manner. Due to the fact that the distance between the side chains of I

is 11.2 angstroms, both side chains cannot fit into one groove of the

nucleic acid. Any intercalated complex would necessitate having one side

chain in each groove of the nucleic acid for this reason. Since it has

already been mentioned that bulky groups prevent intercalation between

the base pairs, it is concluded that local opening of the intact double

helix must occur prior to intercalation. The molecule II serves as a

































+^
CM





-N
u






+ z
C-')

Cr
U

N
N

H '
















H II




*d





03

r~- r-r

U U CM -' r-
CMi NM .. Ln o










II II II II






-
*L








standard, as it has a very similar planar system; it can, however, be

intercalated in the usual manner. A detailed kinetic investigation

was undertaken to gain a better understanding of the possible specificity

involved in the opening reaction of nucleic acids. To this end, both

the reporter molecule and the nucleic acid component were perturbed in

a number of ways to delineate the factors important in these interactions.

Statement of Problem II

Since little is known about what factors are responsible for

intercalation specificity either in terms of the small molecule or

the nucleic acid component, a detailed study of a large number of

variously substituted reporters was launched. The molecules were all

p-nitroanilines as shown in Figure 16, but R ,R2,R3, and n were varied.
The standard methods of investigation for such probes were utilized and

included ultraviolet-visible spectroscopy, pmr, viscosity measurements,

CD, melting curves, and direct binding studies. The nucleic acid

component was also varied to investigate the possibility of specificity

due to base sequence and base content.







NO2

O R1 .2Br


R2 + +
R3 N (CH2) N(CH3)2(CH2) N(CH3)3
III

Reporter R1 R2 R3 n m

8 H NO2 H 2 3

9 H NO2 CH3 2 3

10 H NO2 C2H 2 3

11 H NO2 C6H11 2 3

12 H H H 2 3

13 H H CH3 2 3

14 H CN H 2 3

15 H CH3 H 2 3

16 CH3 H H 2 3

17 H CF3 H 2 3

18 H CF3 H 3 3

19 H NO2 H 2 3

20 H H C6H11 2 3
--- 6 11


Figure 16. Substituted Reporter Molecules.














RESULTS AND DISCUSSION-PART I

A new class of reporter molecules has been utilized to probe the

dynamic aspects of nucleic acid conformation in aqueous solution. These

molecules are illustrated in Figure 15. Both the 1,8-napthylimide 7 and

the 1,8,4,5-naphthyldiimides 3-6 were found to exhibit physical properties

characteristic of an intercalated molecule in the presence of nucleic

acids. Intercalation consists of inserting a planar aromatic system

between the base pairs of a nucleic acid and has been shown to occur

with a number of systems such as acridine orange, proflavine, and

ethidium bromide.40,63'64 Several lines of evidence in support of an

intercalation mechanism for the binding of molecules 3-7 to nucleic

acids are discussed below.

Proton Magnetic Resonance Studies (Pmr)

The pmr spectra for complexes of both 3 and 7 with salmon sperm

DNA were found to be completely broadened and indistinguishable from
76
baseline noise. Raising the temperature as high as 90C did not

cause the signals to reappear. This broadening effect is explained as

being due to the very slow tumbling rates of the reporters bound to

the large nucleic acid molecule.77 If the rate of tumbling of molecules

in solution is lower than the typical Larmor frequencies W0 (of the
8 9 -1
order of 10 -10 radians sec- for protons in the conventional magnetic

field), then T2, the transverse relaxation time, is considerably

diminished, leading to substantial line broadening of the proton

signal. This situation is obtained if the proton is contained in a

27








rigid macromolecule, e.g., DNA, or if the proton is contained in a

slowly i.uTnh jing small molecule bound to a macromrolecule.

Simple external ionic binding of the reporters to the nucleic

acid phosphates would not explain the broadened signals, as the tumbling

rate would not be sufficiently lowered.71 Only binding of a type

causing very restricted tumbling, such as an intercalation process,

can account for the results.

Circular Dichroism Studies

Both molecule 3 and molecule 7 exhibit pronounced induced circular

dichroism effects in the presence of nucleic acids. Table 1 shows the

results of circular dichroism studies on the complexes of both 3 and

7 with several nucleic acids, and Figure 17 shows the entire induced

circular dichroism spectrum for salmon sperm DNA complexes of 3 and 7.

Since these molecules have no optical activity of their own, the induced

CD effect is attributed to the highly asymmetric binding site in the

nucleic acid double helix.

The data shown in Table 1 indicate that the induced CD in the

absorption band of 7 is not as sensitive to changes in the nucleic

acid used as the induced CD in the absorption band of 3. The shape of

the CD curve remains the same for complexes of 7 with salmon sperm DNA,

poly rI-poly rC, and poly rA-poly rU and consists of a trough at

312-325 nm and a peak at 355-360 nm. The molar ellipticity value,

however, seems to be somewhat higher for the poly rA-poly rU complex than

for the other complexes. With the complexes involving 3, however, the

CD spectrum depends to a much greater degree upon the identity of the

nucleic acid used. For example, the induced CD of salmon sperm DNA-3

complex exhibits two troughs, one at 355 nm ([0] = -5.47 x 103) and
14]= -54 1 n








Table 1


Summary of Induced Circular Dichroism Spectra
Acid Complexes of 3 and 7.a


for Nucleic


Complexb Xt [] X 10-3 X [e x 10-3



poly rl-poly rC'3 408 -1.63 319 6.25


poly rA-poly rU-3 365 -8.50


salmon sperm DNA-3 355 -5.47
407 -0.83


poly rl-poly rC-7 325 -1.63 360 3.25


poly rA-poly rU'7 321 -7.25 355 2.25


salmon sperm DNA-7 312 -1.50 357 1.83


aExperiments were carried
ambient temperature.


out in 0.01 M MES (pH 6.2, 0.005 M in Na ) at


Experiments with poly rl-poly rC and poly rA-poly rU were carried out
at base pair to reporter concentration ratios and nucleic acid con-
centrations of 15.6 and 1.0 x 10-3 M P/l, and those with salmon sperm
DNA were carried out at a base pair to reporter concentration ratio of
16.4 and a nucleic acid concentration of 1.9 x 10-3 M P/l.


















0
-- Ln








00 -
\ "* r



cd

II

44 r-a









Z f
\ -- --' -
S o0

-/ --i *-o






/4 P.L -
-a o-








0 -*















r-i C4 i-- 'L)
r-
w
,*-H *
I ci)















C-;
I o w o
*H 0


\ -- H N *rto







Ho k U
-^-cT ^!S -








another at 407 am ([0' = -0.83 x 103), while the poly rA-poly rU'3

complex exhibits a single trough at 365 nm ([0]M = -8.50 x 103) and

the poly rl-poly rC'3 complex exhibits a peak at 319 nm ([6]M = 6.25 x 103)

and a trough at 365 nr (?[]M = -1.63 x -10 ). The detailed interpreta-

tion of these results is rather difficult due to the fact that the CD

effect depends on both the magnitude of the interacting transition

moments and their relative directions. The only conclusions possible

are that intercalation is strongly suggested and molecule 3 may be more

able to discriminate among nucleic acids than is molecule 7.

Viscosity Studies

Viscosity studies also support an intercalation model for the

complexes of molecules 3 and 7 with nucleic acids. DNA is a long rigid

rod-like molecule, and the viscosity of a DNA solution is a function

of the length of the molecule. Since intercalation involves the in-

sertion of an aromatic residue between base pairs, the helix must unwind
66
and increase in length for this process to take place. Passero et al.6

showed that this effect was present in the formation of a number of nucleic

acid-reporter complexes and used this as partial proof for intercalation.

The effect on the viscosity of a salmon sperm DNA solution observed

upon adding increments of either 3 or 7 was studied, and the results are

shown in Figure 18. The relative specific viscosity (complexDNA) in-
sp sp
creases with increasing concentration of the reporters and levels off at

base pair to reporter concentration ratios of 2.30 and 1.74 for reporters

7 and 3, respectively.

Visible Spectral Studies

The visible spectra of molecules 3 and 7 were determined in both

the presence and the absence of a number of nucleic acids. Figures
















2.6




2.4




2.2





2.'0


1.8


1.6




1.4





1.2


0 2 4 6 8 10

Reporter Concentration (x 10-5 M)


Figure 18.


Viscometric Titrations of Salmon Sperm DNA with 3 ( [-- )
and 7 (O-- ). Experiments were done in 0.01 M MES
(pH 6.2, 0.005 M in Na+) at 370C. DNA concentration
was 2.65 x 10-4 M P/l.


W) U3
F-7 17-7








19 and 20 show the results for salmon sperm DNA complexes of 3 and 7,

respectively, and Table 2 shows the absorption maxima and molar

absorptivities for all systems studied. It is seen (Figures 19 and 20)

that the presence of the nucleic acid causes a dramatic decrease in the

intensity of the observed transition for both 3 and 7. The percent

hypochromicity (%H = bufferr -1]100%) varies from 110-171% for the
max max
complexes of 3, while the values for the complexes of 7 vary over a

wider r::'ge. The values for the complexes of 7 with poly rI-poly rC and

poly rA-poly rU are 36% and 42%, respectively, while the values obtained

with other nucleic acids are approximately the same as those for the

corresponding complexes of 3. The lower intensities for the transition

of the complexes as compared to that of the free reporter are predicted

by the intensity interchange theory discussed previously. Since the

reporter transitions are of lower energy than those of the nucleic acid

bases to which they are coupled, the card stack geometry postulated for

the intercalated complex implies that the reporter transition will be
41-43
decreased in intensity upon formation of the complex.

Another point worthy of note is the charge transfer band at the

low energy end of the spectra shown in Figures 19 and 20. This type of

band may be due to electron transfer between the pi system of the reporter
78
and that of the nucleic acid bases. The stacked geometry postulated

for the intercalated complex would allow for this kind of interaction

between the closely stacked pi systems of the two molecules involved.

Equilibrium Dialysis Studies

Equilibrium dialysis binding studies were performed on molecules

3-7 with a number of different nucleic acid st'steLi, and the results

are shown in Table 3. TI.i, method does not yield information on the








































N
N


340

Figure 19.


360 380 400

The Effect of DNA on the Absorption Spectrum of 3. Studies
were done using 3.2 x 10-5 M 3 in 0.01 M MES (pH 6.2, 0.005 M
in Na+) in the presence (---)and absence (--) of 1.8 x 10-3
M P/i salmon sperm DNA.


TIC















15



















10




















(nm)
Figure 20. The Effect of DNA on the Absorption Spectrum of 7. Studies
were done using 6.7 x 10-5 M & in 0.01 M MES (pH 6.2, 0.005 M
in Na+) in the presence (---)-and absence (__) of 1.8 x 10-3
M P/1 salmon sperm DNA.
N P/i salmon sperm DNA.

















0
OC
cy')e
&*< r-


a
0'
11)



0
Cl

C

H













rl
(0
W

















tH
0

cn








cfl
43
4J
M
to.















0"



(U


0C'
a Hr


0 0 0)
0o .0

,H '0 0
Hi o







0
o


Sn
r-l


Ln
I
04-4


0 0 0
rE r o *i-
SH '. cd C.4
p p, rI- 0 1
P. 4-4 4J U
0 o aa
C, 0
,0: I OU U
,-l Hri 0 0)

+ p- T
d co 4J *Hl

r o
$0 0 l

o o co m




Q) a 0 to

0 C Cj c l
-00 00 4

0 cda
S rl 4-4d 4J
en C o 0p
0 1 4co



*H C\a 0


*H0

o3 *H o



4-1 00 0
Q i 4-4 p -
No r- C 0 0 *rl

e p, 1 o nj
40 0 0 1)





o o 4J Cd
S4 c0 0
S* 0 rl




x CM 0n) 0 .f
0a)'i 0 -0

.co *i cO av



0 4J Ced r-
od c0





rO ) 0 40 (1)
0 t 4 4aJ 04


0) N 0 o
O) 43 *WC r
0 r o -1 fr
0 QO -H
a( o p, o
0 0 0t c








Cl 4J B r-lA 4
P ( -H w







4J M r-l x p
U) 1-4

0 a
toaaEr


NM
c4O LCO









-I co


mI r.I|


ml |I
















U)
0)
ca



44
60





1-4
0
t>










u-


0
-4
01



44


il





U
1-4


u








*H

E4 0
-44
1










TO
1-

















41
U) 0
4o








T-4
0

-1













l
-4 U





1-1


4J


0
u

ld

4-1




















































-4



0



-r
0


M

*H





















0

















































-
0
-4


0

(a



















r*








o
a


0




' u





*-


0 0 0 0 0 0 0 0D 0 0 0 0
-z C14 r- r-4 00 r-- C14 -4 C'4 r-0
c04 ir -I I'D ON o00 -4 c
0 D 1 L 0r c1 co ll <-I cl 1 a- r-
i-l r-l




























--T I3- Ln| %o| rl m | -I Tl | o| r | n|


(D
0








C-
aE


0
a



'o-

Ci
(C)


co
00
o



4
0




4-l





pq-

7-4




0


C
r-4







ma



0)a

0 C


ca
o
cH
00
P 0
0 *


C-4







o4
-4Q


*H Z
Scol
r-l



00






0H 0
4-r














0


S*0
U)




0C
PI-













CS|a
>-1

v17








type of binding in the complex but does give an indication of the

strength of binding. A semi-permeable membrane was used for these

studies and allowed the selective passage of small molecules. When

reporter molecule was placed on one side of this membrane and nucleic

acid was placed on the other side, the reporter molecule could pass

through the membrane to the DNA side until the concentration of unbound

reporter was the same on both sides. After equilibration was achieved,

the reporter concentration in the side free from DNA was determined

via absorption spectroscopy. From mass balance considerations, the

amount of reporter bound to the DNA was calculated and the equilibrium

constant for the complex determined. The pertinent calculations are

shown below in equations (1)-(3), where K is the equilibrium constant
a
for formation of the complex and 0. D.max is the optical density

(measured at the absorption maximum of the reporter) of the solution

in the DNA-free side of the dialysis vessel.

K a = [R]bound/[R]f [DNA] (1)


Xmax
free = 0. D. / max (2)



[R] bound= 2([R]total [R]free) (3)

The results indicate that the binding constants are a function of

the base content of the nucleic acid used. For instance, 3 has binding

constants of 11,810, 10,240, and 8,480 to poly dG-poly dC, salmon sperm

DNA, and poly dAT-poly dAT, respectively. The molecules 4-6 are also

found to bind more strongly to salmon sperm DNA than to poly dAT-poly

dAT. In all cases, 3 shows a higher binding constant than any of the

other molecules examined. It therefore seems that these molecules are







specific for nucleic acids, rich in G-C base pairs, and the binding is

sensitive to the substitution pattern of the reporter. The relationship

between the binding constant and the substitution pattern does not

seem to be straightforward, however.

Consequences of Intercalation

It is noted that the 1,8-naphthylimide ring of 7 may intercalate

readily between base pairs without the necessity of breaking hydrogen

bonds. However, in order to intercalate the 1,8,4,5-naphthyldiimide

ring system of 3 between base pairs of DNA, unstacking of adjacent

base pairs and local melting (hydrogen bond breakage) of the helix

appear to be necessary. If the 1,8,4,5-naphthyldiimide ring of 3

intercalates, the N-benzyl, N,N-dimethyl side chains must occupy

opposite grooves in DNA (i.e., one side chain in the minor, the other in

the major groove), since the distance between side chains is about

11.2 angstroms in 3. This argument is illustrated in Figure 21 below.









/ O-HN -- H
H



-----10.9AH


PfICJ2


Figure 21. Relative Dimnnsions of 3 and the DNA Double Helix.








Moreover, it has already been mentioned that molecules 1 and 2 (Figure 14)

do not intercalate due to the bulky substituents on their aromatic ring

systems. Since the N-benzyl, N,N-dimethyl group of 3 is larger than the

substituents employed in the molecules 1 and 2, it is concluded that

intercalation of the aromatic ring system of 3 must be proceeded by

local opening, i.e., melting of the helix.

As stated previously, existence of the open forms of the nucleic

acid double helix has been indicated by several lines of investigation.

The rates of exchange-out of tritium label from intact nucleic acid
45-47
double helices are unexpectedly fast compared to protein systems.

These results were explained as being due to the existence of transient

open forms which exposed the nucleic acid bases to the bulk solvent

normally excluded from the hydrophobic interior of the helix. A com-

parison of the rates of formylation of denatured and native DNA also indi-

cated the presence of the open forms.52

The Experimental Approach

The reason that the hydrogen exchange and formylation reactions

are able to delineate the open forms of the nucleic acid conformation

is that they do not measure an average conformation as the spectral

methods do. Instead, these reactions take place specifically with the

open forms. It was found that the reaction of molecule 7 with nucleic

acids was faster than the capability of the instrumentation employed

(1 msec), while the reactions of molecules 3-6 under the same conditions

were significantly slower.76 This fact indicated that these molecules

might be useful for investigating the dynamic aspects of the nucleic

acid conformation in aqueous solution. Detailed kinetic investigations

on the interaction of 3 with a number of different nucleic acid systems








were therefore carried out. The reaction was studied in both the

forward and the reverse directions, the latter involving dissociating

premade nucleic acid'3 complexes with low concentrations of the

detergent sodium dodecyl sulfate (SDS). The SDS acts by sequestering

the free reporter molecule and does not take an active part in the

reaction, as shown by the fact that varying the concentration of SDS

used by a factor of five did not influence the kinetics of the dis-

sociation reaction. Since the time scale of the reactions was too

fast for conventional methods, a stopped-flow method was used and is

described in the experimental section. It consisted of following the

reactions by observing either the hypochromic effect on the reporter's

absorption spectrum at 383 nm or the development of a charge transfer

band at 410 nm (see Figure 19) as a function of time.

It should be noted that there is an assumption made in the analysis

of the kinetic results. While the concentration of the reporter or

complex as a function of time is the desired quantity, the apparatus

used yields only a measure of the change in transmittance. This dif-

ficulty is circumvented by using concentrations of reporter or complex

which have very low absorbances (0.08 or less). For absorbances in this

range, the changes in transmittance are linearly related to the changes

in absorbance. In turn, the absorbance change is directly proportional

to the concentration change.

The photograph of the oscilloscope trace is used in a very direct

fashion to calculate the rate constants. The distances (D) from the

line representing the equilibrium value for the transmittance to the

points on the curves associated with the various time spans examined

are simply measured with a ruler and plotted as a logarithmic function




42



of time. Figure 22 shows a typical trace labeled with the appropriate

parameters.







DII












Time




0.5 sec/div. (--), 0.2 sec/div. (-..--), 0.1 sec/div. (---),
and 0.02 sec/div. ( ).





The corresponding rate plot is shown in Figure 23. The method of

obtaining the observed rate constant from the slope of the plot is

summarized by equation (4) below,

kbs 2.3/(t2 t ) log (D2/D1) (4)

where D2 and D1 are displacements from the equilibrium line on the

oscilliscope trace to points 1 and 2 on a particular reaction curve.

The times corresponding to points 1 and 2 are tl and t2, respectively.

The Association Reaction

Order with respect to reporter molecule

Synthetic DA. of epeno in seciuence.--A number of lines of evidence

indicate that the association reaction is first order with respect to the



























































Time


Figure 23.


First Order Rate Plot for Dissociation of C. perfringens
DNA-3 Complex. The time scale is 0.5 sec/div. ( 0-O ),
0.1 sc/div. ( -- ), or 0.02 sec/div. (A-A ), and
the reaction was followed at 410 nm.










concentration of 3 when DNA of repeating sequence is used. For

a first order reaction such as A+B, the rate of loss of A is given

by equation (5).79 Dividing both sides by the concentration of A

-dA/dt = kA (5)

and integrating yields the result shown in equation (6). A plot of

-ln A = kt + constant (6)

the natural logarithm of the concentration of A as a function of

time therefore yields the first order rate constant in a direct

manner. In the case at hand, the nucleic acid concentration does

not enter into the rate equation because it is present in excess

at all times during the course of the reaction and is included in

the observed first order rate constant. This result (equation (6))

implies that the first order plot of the concentration of the

reacting species will not depend on the initial value for the

concentration of this species. Further, the reaction should not

be influenced by the nucleotide concentration as long as the nucleic

acid is present in sufficient excess.

With the synthetic polymer duplexes poly dAT-poly dAT and

poly dG-poly dC, the reaction appears to be first order with

respect to the reporter molecule 3 for the following reasons.

(1) The observed rate constant does not depend upon the concen-

tration of the reporter used. For example, as seen in Table 4, the

observed rate constant for reaction of 2.50 x 10-6 M 3 with 1.29 x
-4 -i
10-4 M P/ of poly dAT-poly dAT is 11.95 sec -, while the

observed rate constant for association of 1.25 x 10-6 M 3
observed rate constant for association of 1.25 x 10 M 3





45







aa



cd a





o o o4
o c N.
rcl U! r-4 1






3li (1 (
z G 4-
0 C 0 *T(


Sa)

:$ 0 4-
aN






)10 4J 4 4
V 0 o C
'1 ciir UP
i to 9r- W C
4-4 co rZ- C-4 U4
0 0 Z ad
0 o O CO 4- C
tio co 0

o | 0 00 a
0 o mC *
O O c, i-:
w 0 4-J 0
SL e -1 44a

cl 0 0 a)
H 4- 0 r-l 00
cO o o u
0p r 0 '0-i
Cl) 4-4
co a E-

) t0 cd
rj Il o r0 o 4 U

o Ven a- cO 0 '
0 ** 0
0 4C10 04



S* C a) n
rt ol *



0 H a)H0 4-4

IfCO 00

4J .0 C A ri '
0) 0 a O m .









ai o : 3 u
> 4 43 -i 0 ei a
Pu p4 -H or-i B 4







0) 0 a) ,-1 4 03
0f 0 L o o o
o0 Lt C4 0 0 E0 1 5
ri H 0
) -0 ) 44 I 43 PU
u d C 0: (3 l
S1 CL. -- ; C0
PH h a cO- C
XvC PI *H C


e a u








-i
with the same concentration of poly dAT-poly dAT is 12.40 sec The

difference in rate constants is within the experimental error of the

method. (2) The first order plots of the reaction of 3 with poly dAT-

poly dAT are linear over a range of phosphate to reporter ratios from

3.5/1 to 103/1 (Figure 24).

Synthetic RNA of repeating sequence.--The kinetics of binding of

3 to the ribopolymers poly rl-poly rC and poly rA-poly rU were also

examined (Table 5).

Table 5

Observed First Order Rate Constants for Association of
3 with Synthetic RNA of Repeating Sequence.a

-4 -1
Nucleic Acid (conc. x 10 M P/ ) kb ,sec

poly rl-poly rC (0.25) 31.45
(0.75) 50.07
(1.50) 57.30
(2.50) 77.60

poly rA-poly rU (0.25) 78.80
(0.75) 94.65
(1.50) 169.50
(2.50) 200.00

aExperiments were carried out using 5 x 10-6 M of 3 at
a temperature of 20.5C in 0.01 M MES(pH 6.2, 0.025 M
in Na+). The reaction was followed at 383 nm.

Linear first order plots were obtained over a range of phosphate to

reporter concentration ratios from 5/1 to 50/1. It is important to

note that these studies were carried out at a much lower salt concen-

tration than were the studies with the synthetic DNA previously dis-

cussed. This was necessary due to the fact that the binding was

essentially eliminated in the presence of salt concentrations on the

order of those used in the DNA studies, as shown by the observation

that no hypochromic effect on the absorption spectrum of the reporter



















































.05 .10 .15 .20


Time (sec)


Figure 24.


Rate Plots for Formation of Poly dAT-poly dAT-3 Complex at
Varying Phosphate/3 Ratios. Experiments were carried out in
BPES (0.08 M Na H1P04, 0.02 M NaH2PO4. 0.18 M NaC1, 0.01 M
Na2EDTA), pH 6.9, at 200C and 410 nm. Phosphate/3 ratios
and poly dAT-poly dAT concentrations were respectively 103
and 12.9 x 10-5 M P/1 (A), 52 and 12.9 x 10-5 M P/I (C, 14
and 3.47 x 10-5 M P/l (a), or 3.5 and 8.7 x 10-6 M P/I ((.








molecule is noted in the presence of high salt concentrations. An

interesting fact seen in the data included in Table 5 is the much

faster reaction rate of poly rA-poly rU as compared to poly rl-poly rC.

The reporter 3 therefore appears to be able to discriminate between the

two polynucleotides.

Salmon sperm DNA.--The kinetics of binding of reporter 3 to salmon

sperm DNA were investigated and found to be complex, i.e., several

linear regions of different slopes were observed in the first order

rate plots. This finding is consistent with the presence of several

different classes of binding sites in the DNA molecule. The results of

kinetic experiments with poly dAT-poly dAT and poly dG-poly dC are

also compatible with the presence of different classes of binding

sites (Table 4). For example, the observed first order rate constant

with 6.93 x 10-5 M P/1 poly dAT-poly dAT is found to be 9.00 + 0.15 sec-,
-5
while the observed first order rate constant with 5.81 x 10-5 M P/1
-i
poly dG-poly dC under identical conditions is only 3.53 + 0.04 sec

In line with the postulate that reporter 3 is reacting with

several different classes of binding sites is the fact that the kinetic

plots may be simplified by increasing the ratio of the concentration of

nucleic acid base pairs to the concentration of reporter 3. Under

conditions where the intercalating sites are in huge excess, the re-

porter 3 might be expected to seek out the most favorable sites. Indeed,

as shown by Figures 25 and 26, this appears to be the case. Figure 25

shows the association reaction of salmon sperm DNA with 3 at a base

pair to reporter ratio of 15/1, under which conditions the reaction

gives linear first order kinetics for at'least three half-lives of

reaction. Figure 26 shows the same reaction done at a base pair to




































CN o



















-q0 0 0


HO
















un r(
00


(mm) CI


0
co

JJ -



41 U a)


EO
OO 0*




w 0 0
\o






O r-c


co q o
z O




H () Icl
}-i M4J '


00

-d -<






0 0 t
06 4- J





r- 0 *L






40 0)

0o I-i4 U
0a oI







oo :- )
CO o r









ar)a) (
0 0
0 0 *H






J 0 -



i 0 -
0 ( Z



0 E U-'-






4-1 4- Z





LU
C 4e H
0 0
uo a



(U C
o I E-a














*4l




























































































0 0 0O co
,~l -4


6m






0 Co
-H Trl C
C;s
C rd

o *o






0 0
co






S 0 04-
ed a)






m .o









0 u
*H
00 0






0 tor

ua 0


C O.
0 0o
oH PJ cl
0o (u
*O













4 O l4
co o 4
'4 0-.)













0 H A
H-





D CO: 0d
0)










$4 0 tf*
0 r-I<





















r$4
5P4
cH0 3
h 4- H-


*r


(unn) a








reporter ratio of only 7.5/1. Under these conditions, it is possible

to see at least two processes which contribute substantially to the

reaction. In fact, there are also traces of both faster and slower

processes in the experiment using a high base pair to reporter concen-

tration ratio, though the deviation from the straight line is more

certain at the very early times than at the very late times due to the

inaccuracy involved in measuring very short distances on the oscilliscope

trace.

A comparison of the kinetics at two different wavelengths (383 nm

and 410 nm) was also carried out. The first wavelength is the absorption

maximum for the reporter, while the second is in the area of the spectrum

due to the charge transfer band of the complex. It was found that

identical kinetics were obtained regardless of whether the development

of the hypochromic effect at the absorption maximum of the reporter or

the development of the charge transfer band for the complex was followed

(Table 6).

Table 6

Observed First Order Rate Constants for the Formation of Salmon Sperm
DNA-3 Complex at 383 nm and 410 nm.a

-1
k sec
obs _
Wavelength (nm) k k k3
-1 --2 --3
383 36.05 17.23 9.74

410 44.50 18.21 -

aStudies were done at 19.50C in 0.05 M MES (pH 6.2, 0.10 M in Na+4
using DNA and 3 concentrations of 1.5 x 10- M P/1 and 1.0 x 10- M,
respectively.








Order with respect to nucleic acid

Since the concentration of the nucleic acid is included in the

pseudofirst order rate constant, varying this concentration and following

the resultant change in the rate constant provides a measure of the

kinetic order with respect to the nucleic acid. Since the pseudofirst

order rate constant is actually the product of the true rate constant

and the nucleic acid to some power (equation(7)), a plot of the

k = k' nucleicc acid)n (7)
.obs
logarithm of the observed rate constant versus the logarithm of the

nucleic acid concentration should yield a line with a slope equal
81
to the true kinetic order with respect to the nucleic acid. Such

studies were done with poly dAT-poly dAT, poly rl-poly rC, poly rA-

poly rU, and salmon sperm DNA, and the resulting log-log plots are

shown in Figure 27. The most surprising point evident from a consideration

of thesedata is the fact that the order with respect to the nucleic

acid may vary depending upon the polynucleotide examined. It was es-

tablished that the order was 0.4 with poly dAT-poly dAT, 0.5 with poly

rl-poly rC and poly rA-poly rU, and 1.0 with salmon sperm DNA. In

the case of the latter, the rate constants used were those obtained from

portions of the rate plots which were linear over at least three

half-lives of reaction (Table 7). The varying order with respect to

the nucleic acid may be explained by assuming that the same concentrations

of nucleic acids may yield different concentrations of reactive sites for

interaction with the reporter 3.

Kinetic models

The main points considered in the formulation of a model to

explain the kinetic data were the first order dependence on the reporter















0



>H





0
4- 0 r-
c c


.-


o


o a
4J 04












4- 0 -
O.H











ca <]
Co














1
o 0-








04I
0 0

P' 04




0 0 c
0.
Qrl
0 (f


JTULSUo0 < >, J pn,;A-.sqI)










Table 7

Observed First Order Rate Constants for Formation of Salmon Sperm DNA'3
Complexes at Different Nucleic Acid Concentrationsa


Nucleic Acid Concentration x 10-4 M P/1)


-1
k sec
obs _


2.25

3.00

6.00

10.00


3.38

5.12

12.30

15.50


a -6
Experiments were carried out using 5 x 10 M of 3 at a temperature of
20.5C in 0.05 M MES (pH 6.2, 0.375 M1in Na+). The reaction was
followed at 410 nm.








concenitrtion, the varying order with respect to the nucleic acids,

and the nonlinear nature of the rate plots obtained with salmon sperm

DNA. A number of reaction schemes were considered. The detailed

kinetic equations are presented in Appendix 1, but the salient features

of the models are presented in simplified form in Figure 28. Model 1

(Rate-Determining Preequilibrium) is ruled out on the grounds that the

reaction is expected to be slower in its initial stages than in its

later stages; this is contrary to the observed results. Model 2

(Product Inhibition) was found to be in qualitative agreement with the

results, but it predicts a smooth decrease in the observed rate of

reaction as the product concentration increases. The observed kinetics,

however, consist of several rather extended linear regions of different

slopes. Model 3 (Catalysis by an Intermediate Complex) again predicts

a smoothly decreasing rate of reaction and is ruled out on this basis.

Model 4 (Rate-Determining Formation of Intermediate Complex) is not in

agreement with the data since it predicts a slower rate at the onset

of the reaction rather than a faster rate. Model 5 (Multiple Sites)

was the best explanation for the results and is shown in Figure 29 along

with a detailed derivation of the first order rate equation. It is

assumed that the reaction involves a conformational equilibrium between

closed and open forms of the nucleic acid double helix. Transient

twisting of one strand relative to the other is postulated to open the

double helix and expose a segment several base pairs long. This

process partially decouples the opened segment from the rest of the

structure and allows for dynamic properties depending upon the sequence

and base content of the opened segment. These open but still stacked

segments may then react with the reporter molecule by unstacking and








Model 1 Rate-Determining Nucleic Acid Preequilibrium
l(slow)
(CH) i n(OH)
-1

-2

Model 2 Product Inhibition


(CH) n (OH)
-1
2 3(slow)
(OH) + (R)7 ~(OH R) ---- (CHR)
-2
4
(CH.R) + n(OH)-- (CH.R) + (CH)


Model 3 Catalysis by an Intermediate Complex

1
(CH) n(OH)

2 3(slow)
(OH) + (R) (OH HR) --- (CH-R)

4
(OH-R) + (CH) -- -(OH.R) + n(OH)


Model 4 Rate-Determining Formation of Intermediate Complex
1
(CH) n n(OH)

2 3
(OH) + (R) --( (OH R) (CH.R)
(slow) 3-^



Figure'28. Possible Models for the Association Reaction of 3 with
Nucleic Acids. (CH) and (OH) are the closed and open
intercalating sites, (R) is the reporter.

















n(OH)


d(CH-R) -dR
dt dt


d (OH-R)
dt


(R) (OH R) (CH.R)


k3 (01HR)


= 0 -k3(OH.R) + k2(OH)(R) k.2(OH-R)


ko (OH)(R)
(OH*R) = k-2+k3


K (OH)n
K CH
CH


and (OH) = (K)i/n(CH)1/n


k2 (K) l/n (CIH)3 1 (R)
(OH*R) = k-2+k3


and -dR
dt


-dR k2k3Kl/n CH)/ndt
R k-2+k3

-InR + const. = k'(CH)I/nt

-inR + const. = kobst

kobs = k' (CH)I/n

log kobs = log k' + 1/n log(CH)


Figure 29.


The Multiple Sites Model for the Association of 3 with
Nucleic Acids. (CH) and (OH) refer to closed and open
variants of the double helix, respectively while (R)
refers to the reporter molecule.


(CH)


= k2k3 (K)/n (C)1/ /i)
k








then reclosing over the planar portion of the reporter system in the

slow rate-determining step.

It is worth noting that evidence exists for both the stacked and

the unstacked open variants of the nucleic acid conformation. The
52
main evidence comes from the formylation studies52 cited previously.

This investigation indicated that the unstacked conformation required

a much higher energy of activation for formation than did the stacked

open conformation utilized in the hydrogen exchange reaction.

The precise nature of the unstacking and intercalation events may

obviously differ from one nucleic acid to another or even from one

segment to another. This is indicated by the observation that complex

first order plots are observed for association of 3 with nucleic acids

of heterogeneous base sequence, while the association with synthetic

nucleic acids of repeating sequence yields only one first order rate

constant. Likewise, the results indicating that the order with respect

to the nucleic acid may differ from one nucleic acid to another are

also accomodated by this model. The opening segments may either be of

different length or of different capacities with regard to the number

of reporter molecules which may be intercalated in a given stretch of

the helix. In either case, the value for n in the rate equation

implied by this model indicates the number of reactive sites yielded

by the opening of a nucleic acid segment of a given size. For example,

the order with respect to poly dAT-poly dAT was found to be 0.4, and

this implies that the opening reaction with this polynucleotide supplies

open segments capable of intercalating 2.5 molecules of 3 on the

average. Poly rl-poly rC and poly rA-poly rU both need a value of

0.5 for the order and can apparently intercalate 2.0 molecules of 3








into an average opened segment. Salmon sperm DNA gives a value of 1.0

for the order and can therefore only intercalate one molecule of 3 in

the average opened segment. These findings indicate that a certain

degree of specificity dependent on the nature of the nucleic acid is

possible in the intercalation reactions.

Influence of ionic strength

A series of experiments was carried out to assess the effects of

varying ionic strength on the formation of the salmon sperm DNA'3

complex. The observed rate constants for the association reaction are found

to be lower at high ionic strength (Figure 30). There are at least

two explanations for these results. (1) The reaction may be slowed

due to a shift in the equilibrium between closed and open forms of

the nucleic acid double helix, i.e., K may be decreased at high ionic

strength. It has already been stated that the intact double helix is

favored by high ionic strength due to the decreased phosphate-phosphate

ionic repulsions. (2) The reaction may be slowed due to the lesser

importance of ionic interactions in solutions of high ionic strength.

Since the ionic attraction between the reporter and the opened nucleic

acid conformer is expected to be of some importance on intuitive grounds,

the lesser importance of these interactions at higher ionic strength,

i.e., the decreased value of k2/k_2, would also explain the results.

It is not possible to rule out either of these possibilities, but both

are in general agreement with the model discussed in the previous section.

Experiments with polylysine-DNA complexes

To get further information on the interactions between reporter

3 and nucleic acids, the kinetics of the binding of 3 to salmon sperm

DNA-polylysine complexes wereinvestigated. The observed first order



















































.05 .10 .15 .20 .25


Time (sec)


Figure 30.


Effect of Varying Ionic Strength on the Association of
Salmon Sperm DNA and 3. Experiments were done in 0.05 M
MES, pH 6.2 at 19.50C. Total Na+ was 0.2 M (0), 0.3 M (A),
or 0.4 M (0). The base pair to reporter ratio was 15/1,
and the nucleic acid concentration was 3.0 x 10-4 M P/i in
all cases.








rate constants for this reaction as a function of polylysine concentra-

tion are shown in Table 8. As with DNA alone, the rate plots exhibit


Table 8

Observed Rate Constants for Association of 3 with DNA-
Polylysine Complexesa


kobs, sec-1
Polylysine (pg/ml) Lysine/Phosphate kI k2 k3 k4


0 0 44.50 18.21 -

5 .15 19.40 8.68 -

10 .29 22.80 6.24 0.42 -

15 .44 24.60 3.99 0.90 0.48

20 .59 19.30 2.37 0.62 0.32

aExperiments were done at 410 nm in 0.05 M MES (pH 6.2, 0.10 M in
Na+) at a temperature of 19.50C. Salmon Sperm DNA and 3 were used
at concentrations of 1.5 x 10-4 M P/1 and 5 x 10-6 M, respectively.


several linear regions of different slopes. The effect of increasing

the polylysine concentration was twofold. (1) The various rates were

all slower than the rates shown with DNA alone. (2) There appear to

be more rate processes observed as the amount of protein increases.

While there were only two rate processes for the complex containing

5vg/ml of polylysine, the complex containing O1Pg/ml showed three rate

processes, and the complexes containing 15pg/ml and 20pg/ml showed four

rate processes. There are several ways to explain these results, but

all would be rather speculative due to the lack of detailed information

on the structure of the DNA-polylysine complexes.








The Dissociation Reaction

Order with respect to nucleic acid-3 complex

The kinetics of dissociation of nucleic acid-3 complexes were

followed at 410 nm using the SDS technique discussed previously. Low

concentrations of the detergent SDS were mixed with the complex, and

the resulting decrease in absorbance due to the freeing of the reporter

molecule from the nucleic acid double helix was followed as a function

of time. Two lines of evidence indicate that the dissociation is first

order with respect to the complexes. (1) The experiments using a nucleic

acid of repeating sequence yield only one rate constant, in agreement

with the model for the association reaction. Since the repeating

sequence of poly dAT-poly dAT provides only two very similar intercalating

sites and that of poly dG-poly dC provides only one intercalating site,

one might expect to see only one dissociation rate process with complexes

involving these nucleic acids. It is found that the first order rate

plots are linear over at least four half-lives of reaction in the case

of poly dAT-poly dAT-3 complexes (Figure 31). The typical first order

plots showing several long linear regions with different slopes are

obtained when natural DNA. is used for the reaction. The more complicated

cases seen when using the complexes of 3 with C. perfringens DNA and

M. luteus DNA are shown in Figures 32 and 33, respectively. (2) The

rate constant does not depend upon the initial concentration of complex

used. Table 9 shows the results obtained by varying the concentrations

of calf thymus DNA-3 and salmon sperm DNA-3 complexes. It is seen that

the observed first order rate constants do not change over as much as a

fourfold increase in the concentration of the complex. For example,

a complex containing salmon ,,Perma DNA at a concentration of 6.0 x 10-4 M P/I




























































Figure 31.


0.1 0.2 0.3 0.4
.02 .04 .06 .08

Rate Plot for Dissociation of poly dAT-poly dAT'3 Complex
at 410 nm. Time scale is either 0.02 sec/div. (0) or
0.1 sec/div. (0).
























































.02 .04 .06
.10 .20 .30
.50 1.0 1.5


Time (sec)


Figure 32.


Rate Plot for Dissociation of C. perfringens DNA 3
Complex at 410 nm. Time scale is 0.02 sec/div. (o-o),
0.10 sec/div. (A-A), or 0.50 sec/div. (0-0).


.08
.40
2.0













40


30





20








10
9
8

7

6

5


4



3


.04
.20
1.0


.08 .12
.40 .60
2.0 3.0


Figure 33.


Rate Plot for Dissociation of M. luteus DNA 3 Complex
at 410 nm. Time scale is 0.04 sec/div. (O--), 0.20 sec/div.
(A-A), or 1.0 sec/div. (C-).


.16
.80
4.0







Table 9

Dependence of Dissociation Rates on the Concentration of the
Nucleic Acid*3 Complex.a


Nucleic Acid (conc. x 10-4 M P/1) [3] kobs, sec-1
1 31
k1 k2 k3



Salmon Sperm (9.0) 4 x 10-5 M 1.61 1.48 1.23

(6.0) 4 x 10-5 M 1.86 1.65 1.05

(3.0) 2 x 10-5 M 1.63 1.54 1.13

(1.5) 1 x 10-5 M 1.65 1.54 -

Calf Thymus (6.0) 4 x 10-5 M 2.02 1.53 1.17

(3.0) 4 x 10-5 M 1.93 1.51 1.38

(3.0) 2 x 10-5 M 1.67 1.54 1.13


aStudies were done at 410 nm in 0.05 M MES (pH 6.2, 0.0 M in Na ) at 200C.








and 3 at a concentration of 4.0 x 10-5 M showed rate constants of 1.86,

1.55, and 1.05 sec-1, while a complex containing salmon sperm DNA at a

concentration of 3.0 x 10-4 M P/I and 3 at a concentration of 2.0 x 10-5 M

showed rate constants of 1.63, 1.54, and 1.13 sec--.

Kinetic model

The model for the dissociation reaction parallels the model pre-

sented for the association reaction in Figure 29. The mechanism involves

the same type of sequence-dependent opening reaction as that shown for

the association reaction, and this is followed by the release of the

reporter into the solution where it is rapidly sequestered by the SDS.

Since the reaction is followed at 410 nm where the free reporter has no

absorption (Figure 19), the effect seen is due to the time-dependent

decrease in absorbance due to the complex alone. Figure 34 shows the

model in schematic form along with the detailed derivation of the first

order rate equation. This model assumes that the opening of the

complex involving the double helix is the rate-determining step.

Experiments with sonicated and heat-denatured DNA

Kinetic studies were also done on complexes of 3 with both

sonicated low molecular weight DNA and heat-denatured DNA. While the

sonication had little effect on the observed rates of dissociation

measured by the SDS method, the denaturation of the DNA proved to have

a very marked effect (Table 10). It was found that the rate constants

were much faster for the denatured DNA than for the native DNA. On the

basis of the very small effect caused by sonication of the DNA, it may

be concluded that the opening event is not dependent upon proximity to

a chain end. To thoroughly understand the significance of the results

with denatured DNA, the details of the structure of this system must be












w~


(CH'R)


(OH*R)


(OH)


-d (CH.R)
dR/dt = k2 (OH-R) = dt
2. dt


d(OH"R)
d(R = = -k=(O -k_(OH)-k (OHR) + k (CH'R)



'. (OH'R) = k (CH'R)
k2+k_-


= klk2(CH-R)
k +k_
2 -1


o -d(CH R)
(CHorR)
(CHR)


-In(CH-R) = k 2
k+k t
k2+k-1


+ constant


= kbst + constant
obst


Figure 34.


Kinetic Model for the Dissociation of Nucleic Acid'3
Complexes. (CH) and (OH) refer to the closed and
open conformations of the nucleic acid double helix;
(R) refers to the reporter molecule.


(R)


-d(CH-R)
dt


klk2
Sk- dt
k +k
2 -1


~----
-e(.,








Table 10

Observed First Order Rate Constants for the Dissociation of Complexes
of 3 with Native, Denatured, and Sonicated Salmon Sperm DNA.a


kobs, sec-1
Nucleic Acid kl k2 k3

Salmon Sperm (Native) 1.95 1.70 1.46

Salmon Sperm (Sonicated) 2.12 1.86 1.63

Salmon Sperm (Denatured) 32.2 17.9 10.5

aThe reactions were followed at 410 nm in 0.05 M MES (pH 6.2,
0.10 M in Na+) at a temperature of 200C. Nucleic acids and 3
were used at concentrations of 3.0 x 10-4 M P/1 and 4 x 10-5 NM,
respectively.


considered. Printz and von Hippel have found that denatured DNA contains

a laric amount of short double helical segments of widely different

stabilities.82 They also found that the hydrogen exchange rates of na-

tive and denatured DNA were strikingly similar, indicating that the two

preparations are similar in conformational motility. An apparent contra-

diction exists between these results and our own, since our results

indicate that the conformational properties of the two DNA preparations

are quite different. As stated previously, the hydrogen exchange reaction

is thought to utilize a stacked open state of the nucleic acid conforma-

tion under mild experimental conditions. If it is assumed that the

reporter is not using the stacked open state but is using the unstacked

open state, the literature results may be reconciled with our own. The

denatured DJA. must evidently give rise to a higher steady state concen-

tration of the unstacked open form than does the native preparation.

Another possibility considered was that the structure of the denatured

DI;A was different enough to radically alter the types of binding sites








available. Three lines of evidence argue against this possibility.

(1) It was found that the absorption spectra of denatured and native

DNA-3 complexes are very similar (Table 11).

Table 11

Summary of Absorption Spectra of 3 in Buffer and in Native and Denatured
Salmon Sperm DNA Solutions.


Buffer Native DNA Denatured DNA
max max -max max max max H
max (nm) max max (nm) cax %H max (nm) max %H

381 31400 381 14330 119 381 15270 97
Experiments were carried out in 0.05 M MES (pH 6.2, 0.10 M in Na ) at
ambient temperature. The base pair to reporter ratio was 5.0 and the
nucleic acid concentration was 1.5 x 10-4 M P/1.

Since the percent hypochromicity is a crude measure of the closeness of

contact between the reporter and the nucleic acid bases in the complex,

it seems that the two types of complexes cannot be drastically different.

(2) The rates of dissociation of the two types of complexes were also

investigated as a function of temperature to get a measure of the

activation energies involved. The Arrhenius relationship shown in

equation (8) states that the derivative of the natural logarithm of the

dln k/d(l/T) = -E /R (8)
a
rate constant with respect to the reciprocal of the absolute temperature

is equal to the activation energy for the reaction multiplied by a

constant.83 This relationship indicates that a plot of the observed

rate constant for the dissociation reaction versus the reciprocal of

the absolute experimental temperature should yield a straight line with

a slope that gives a measure of the activation energy for the reaction.

The results of such studies on complexes of 3 with denatured and native

DNA are shown in Figures 35 and 36, respectively. The activation








































20








10




Figure 35.


3.2 3.3 3.4

1 (10-3
T
Van't Hoff Plot for kI (_ ) and k2 (---) Rate Processes in
the Dissociation of Denatured Salmon Sperm DNA-3 Complex.
Experiments were done at DNA and 3 concentrations of
3.0 x 10-4 M P/1 and 4.0 x 10-5 M, respectively, in 0.05 M
MES (pH 6.2, 0.10 M in Na+).



































h
C-)
('3
('3
0


Figure 36.


3.2 3.3 3.4

1 (10-3)
T

Van't Hoff Plot for k1 ( ) and k (---) Rate Processes in
the Dissociation of Native Salmon Sperm DNA-3 Complex.
Experiments were done at DNA and 3 concentrations of 3.0 x
10-8 M P/e and 4.0 x 10-5 M, respectively, in 0.05 M MES
(pH (.2, 0.10 M in Na+).







energies determined from these plots are all in the range 13-16 kilo-

calories per mole and indicate that the two types of DNA may be quite

similar. (3) The two types of complexes were also compared by equilibrium

dialysis techniques. A number of different concentrations of reporter

molecule are dialyzed against a single concentration of nucleic acid,

and the position of the equilibrium is assessed according to the methods

outlined previously. The concentration of the unbound reporter is

determined via measurement of the absorbance of the solution in the

side of the dialysis vessel containing no DNA and is inserted into the

Scatchard equation (9), where n is the number of moles of reporter bound

n = nmax (1/Ka) (n/Rf) (9)

per mole of phosphates, nmax represents maximum binding, Ka is the

association constant for the DNA-3 complex, and Rf is the concentration

of free reporter molecules.84 A plot of n versus n/RHf then gives the

values for nm (x-axis intercept) and nmaxKa (y-axis intercept).

Table 12 shows the results of these studies for complexes of 3 and 7

with both native and denatured salmon sperm DNA, and the data plots are

reproduced in Figures 37 and 38 for complexes of 3 and 7, respectively.

The calculated values for the minimum number of base pairs required to

provide one strong binding site for the reporter (1/2n ) are also shown
max
in Table 12. These data indicate still another similarity between the

two types of complexes. Both types yield the same value for the maximum

number of reporters able to be bound to the nucleic acid on a per phos-

phate basis (1/2nmax), although the binding constants (Ka) are quite

different. For example, 3 binds to both types of DNA to the extent of

0.166-0.167 molecules per nucleic acid phosphate, while the binding

constants are 26.7 x 104 and 12.6 x 104 for reaction with native and

denatured DNA, respectively.

























































.10 .12 .14 .16


Figure 37.


Scatchard Plot for Equilibrium Dialysis of 3 with Native
(0-0) and Denatured (3--) Salmon Sperm DNA.

































C"t
0


16 -





12 -




8





4










Figure 38.


.04 .06


Scatchard Plot for Equilibrium Dialysis of 7 with Native
(CO) and Denatured (L-H) Salmon Sperm DNA.








Table 12

Equilibrium Dialysis Results for Native and Denatured Salmon
Sperm DNA Complexes of 3 and 7.a



Reporter DNA K (10 ) n 1/2n
a max max



3 native 26.7 0.167 3.00

3 denatured 12.6 0.166 3.00

7 native 4.4 0.100 5.00

7 denatured 7.9 0.090 5.56



aExperiments were carried out in BPES (0.08 M Na2HP04,
0.02 M NaH2P04, 0.18 M NaC1, 0.01 M Na2EDTA, pH 6,9)
at ambient temperature. The DNA concentration was
3.0 x 10-4 M P/l, and the reporters were used at initial
concentrations of 2,3,4,5, and 6 x 10-5 M.



Experiments with nucleic acids of differing base content

Still further experiments showed that the dissociation rates

were a function of the base content of the nucleic acid used to form

the complex (Table 13). It can be seen that the rates of dissociation

are fastest for those nucleic acids which are richest in adenine and

thymine. This implies that the AT base pairs cause a higher steady

state concentration of the reactive open forms under the experimental

conditions. The fact that poly dAT-poly dAT associated with the reporter

3 about three times as fast as did poly dG-poly dC also fits in quite

well with these data.








Table 13

Observed First Order Rate Constants for Dissociation of
Complexes of 3 with Nucleic Acids of Differing Base Content.


-1
k ,osec
obs

Nucleic Acid %A+T k1 kk k3

Poly dAT-poly dAT 100 4.22 -

C. perfringens 76 2.35 1.99 1.74

Calf Thymus 58 1.93 1.45 1.15

Salmon Sperm 58 1.74 1.56 1.19

M. luteus 24 1.25 0.81 0.44



aKinetic measurements were carried out at 200C in 0.05 M MES
(pH 6.2, 0.10 M in Nat) and followed at 410 nm. Nucleic
acids and 3 were used at 3.0 x 10-4 M P/1 and 4.0 x 10~5 M,
respectively, except in the experiments with5poly dAT-poly
dAT, where the concentrations were 2.2 x 10- M P/1 and
2.0 x 10-5 M, respectively.



In summary, the results indicate that the dynamic aspects of

nucleic acid conformation depend upon the particular bases involved

in the opening segment. This finding is quite significant, as it

is easy to imagine the role of such open structures in the many

important biological processes where open single strand conformations

have been indicated to be necessary. Two examples are genetic

recombination and nucleic acid replication. The fact that the formation

of the open segment has been indicated to be base content-dependent presents

an obvious method for the introduction of specificity into these

processes.













RESULTS AND DISCUSSION-PART II

The second part of the dissertation deals with the investigation

of precisely which features of the p-nitroanilines (Figure 39) are





N0 2

R 3N(CH ) N(CH 3)(CH) N(CH-3)3

Figure 39. The p-Nitroaniline Reporter Molecule.

responsible for specificity in their interactions with nucleic acids.

That such specificity may exist is indicated by the results of Gabbay
59
and Gaffney.59 Circular dichroism studies showed that the reporter

molecules exhibited a more negative induced CD effect in the presence

of those nucleic acids richest in adenine-thymine base pairs. Further,

as stated previously, the drug actinomycin has been shown to exhibit

selectivity for those nucleic acids richest in guanine. This behavior

has been postulated to involve specificity in the stacking forces

present in the complex.53

A number of experimental methods including pmr, ultraviolet-

visible spectroscopy, melting temperature profiles, viscosity measure-

ments, circular dichroism studies, and equilibrium dialysis binding

studies were used to investigate the intercalation reactions of a

number of different p-nitroaniline reporter molecules in the presence

of nucleic acids. The results of these studies will be presented and








discussed in order.

Ultraviolet-Visible Spectral Studies

The results of comprehensive ultraviolet-visible spectral studies

show several points worthy of note (Table 14). (1) The hypochromicity

value (%H m= E mmax 1] 100) is most pronounced with those compounds

bearing only planar substituents. For example, 8 (2-NO2) and 14 (2-CN)

show percent hypochromicity values of 75% and 68%, respectively, in the

presence of salmon sperm DNA. These results suggest that the planarity

of the aromatic system of the reporter molecule is important for a close

contact of the reporter and the nucleic acid bases in the complex. The

hypochromic effect in the absorption spectrum of the reporters is due

to intensity interchange between the electronic transitions of the

latter and those of the nucleic acid bases and is probably a good

indicator of the contact distance in the complexes. (2) The percent

hypochromicity values for the 4-nitroaniline series and those for the

2,4-dinitroaniline series are affected differently by increases in the

size of the N-alkyl substituent. The percent hypochromicity values

for the 2,4-dinitroaniline series are 75% for 8 (2-NO2), 47% for

9 (2-NO2, N-CH3), 31% for 10 (2-NO2, N-C2H5), and 29% for 11 (2-NO2,

N-C6H11). On the other hand, the values for the 4-nitroaniline series

are 33% for 12 (unsubstituted), 30% for 13 (N-CH3), 45% for 20 (N-C6H11),

and 35% for 21 (N-C7H7). The results indicate that the presence of

progressively larger N-alkyl substituents in the 2,4-dinitroaniline

series causes a progressive increase in the contact distance between

the reporter and the nucleic acid bases in the complex, although

little effect is observed in the 4-nitroaniline series. The most

obvious explanation for the results is increasing steric interference
























'0
41
w
Cd














H



0

04.
a)
H









C4







44
H-


















0
41


0









4-H
0
0









Lo






'm4

4-i
U
a)
















C4Ca

o0
H43



















4d 0
o rl)
:jc












WU)
a)
a)





(Cd


w

ac


o









H X



V)
Cd












H4 w
o









Cd











4-4
p<



0
cd














$4







C










P.
a)h
4B
'-4v
0\








a)


80














tr) t -4 m c) 0 00 00 r % n f
rLO IT CY O\ ) Co ) %D Cc) N. Ln to) -to t






CA C) CA CA CA CA 0 CA 0 CA 0) CA C)
oA CA CA 00 CA CA m IT CA CA -It CA 0
r" N.- ko r- 0 -IT 00 C 'Tj r, N c -1-T C
00 cc 00 r- i m 00c ) 0 0 m Nq










~H H H H H H H
LA r, N. 00 c7% ) %-0 m\ m~ %D o C) m~
Mo t Mo MO M -t MO MO M Co) cn -~It co









o CA 0 CA CA C) CA 0) ) C) 0) 0 CA
00 to r- %.0 CA C) C -T C) C Ln C) C

-4 Mo C) C% t0 mO N. -It mO %D 10 cn
HA H 'A H Hq H- H Hq H H Hi H-






r- N1 cn MO CM 0c H- H- %o %D cn H N.
nf ul ul un I'D N. -T N. I'D It -It 0c to




Co 0o 0o CA Co co C CD C CA CA CA




C) C) C) C) CA C) CA WT CD CA to C) C)
C) 4 LA ON LA ) .1 C) CY) C) t0 Cr) C)

H- H H H H H H H







00N. c ON Nl m~ N1 N. H CA to Co C)
to) %t D %0 oo ON o ) 00 CIA cc o JO) CA m\
c) CO) Co) m Co) M. Co Co o C C )C
~ m o~~m ~ ~ ~ ci










r-4 r rI a\y, ?-I C%4 C14












rl HI u 1~I Q I u 1 0 1 q Q I


a)
(d

,0

tv
CA
















C:;
to



















41
ca
Q)
















N











r4J
Cd

41






co
U


0


0

ca
*0

Cd










a)


Cd

a)




4J




0)



Cd


















C
o
*-H

*rl







4-J
'O



0
















0
4-I
41



0

0










0-
0a





4 -

0

0)







0 )
I1





44
a0









4.J
0l c-
4 a
Otc o






M
Ma M
0 M
0.l (U
(U 4-
)- ^


I
4-4
0)
0


0
*r-f


C
4-I




C
c0
0)

,0










a >

0)
U a
Cd 3'








0







c o 4-.
g-
0)






r *=








ca



















W
0




I o

SO 0
<0 U










P, 04
z Cd
0)


C) U)




O o



0
s-I0 0






C
*
0. 0

0) *)








between the 2-nitro group and the N-alkyl group, resulting in increasing

nonplanarity of the aromatic system of the reporter. Consideration of

the molar absorptivity values ( max) for the reporters in buffer

solution lends further support to this explanation. Increasing the

size of the N-alkyl substituent in the 2,4-dinitroaniline series causes

a steady decrease in the molar absorptivity from 15600 for 8 (2-NO2) to

9250 for 11 (2-NO2, N-C6H11), while no such effect is seen in the

4-nitroaniline series. Twisting of the alkyl-substituted aniline

nitrogen atom with respect to the rest of the aromatic system of the

reporter would decrease the resonance interaction between the nonbonded

electrons of the former and the rest of the aromatic system due to poor

orbital overlap. Since a charge transfer transition contributes to the

intensity of the absorption spectra for systems of this kind, the

overall intensity will decrease upon partial removal of the delocali-

zation capability of the aniline nitrogen.85 (3) Compounds 13 (N-CH3),

15 (2-CH ), and 16 (3-CH ) have percent hypochromicity values of 30%,

38%, and 27%, respectively, in the presence of nucleic acid. The

results indicate that the position of the methyl substituent may

influence the contact distance between the reporter and the nucleic

acid bases in the complex.

Proton Magnetic Resonance Studies (Pmr)

Pmr spectra were obtained for a number of 2-nitroaniline

reporter molecules in both the absence and the presence of salmon

sperm DNA (Table 15). It is observed that the signals for the protons

of the reporters are both broadened and shifted upfield in the presence

of DNA (Figure 40). As discussed previously, the broadening is

attributed to the slow tumbling rate of the large nucleic acid molecule,























U1 0U

co ,-.


n mn

m
en


a)
4J
(3

0


In
H

ca


a)



1H

















0
r-'-
a)













I1


0)

N 4
3 o




Cl 4
























*H n
to o

a
4-1a)









o n4
0o
H Cu
Cum








a




4-'0





4-1
C u
4U

Nr 3l





U v


Eo 4
10

0
or-





en


I-,

or

o

r-


Uin Lf in
.0.


I II I '"a a t
'a '042
H r-I r-
cn cO o
cO o r~-


Ln


O'
*

a
01
or


NM I
-e-
CN
0


co
en


SC

con
U.0
H
N
ft


r-4
I I

N


m com


"- I
v"l
u'

I--


0
I'-

en
1,,


,- .... 0

r-l 0o '

00 N O 30 OcO DO
aocO en -Coo rH
oo r oo co r- co


0 *
II,


I 'U '-

CO CM
-:r -11
co oo


i-
CM

'

co co
O v


r- O r- OO O 0~O r aO 0 o
c y)o Jo CY oi c cL oa C cf, ON m to 0 CY c o0


ol

col CmI


o|
;?4i-l


-e,

H


en




+z
en




i 0
CM,





0 CM



u

Sz
-/
\ ^



















So o0 co


m ca c




m
C- 4
c Cai


0
Ch








0\



a





C-
0-







co
C-.





I I t I
co












'0
N






1 t


Co


f ,


m)V ccan in EmD


cn CYcoe) Cfl cY) C-)cY)


co


co in

en r
CM CM


COm C C
,0.o ,0 0o L


CO i- o
iH HH
c0
,t co inca i


Nl
CM4 CM
CIO C',


ON C ri.
. .0 '



-a cc
0co 0
co


r-0 0 r4 0
M Ln M> Oe CV ON


*AA
afl0 ,fl
en oo cr
K r-f r-l
S s 0
Lo Ho


'\

n a 6

33 '0 n o0
r rll P,
a ,t ,f '0


I'dlII '













-cO c o-."
co -e -.








0-%


I'0'0 00 -,3
ODco r- r-


4 C4
N (4
) cm
com

C) CM)
m









04



Sv







0o co
O OD







HC4





coco


r-l 0r
Vn co
co ini


0\
v

Lrf
I.














cc
\C



oo


o-N
C'











a00
I "0














uHH
r-i si


r-- 0 o r- 0C o0 r- o r0 o C) r- o
CM nI M CO' mMnM C m0'O'c m cnl McrO mm


*I Al
r-4 H V-4


z z


0I aI 0


0

C)

-1
Ha





*rl
'1










r-i
4-























ca
Cd









00
rz





0 a)

Co





O













(c
O
O

o
03
0
0














CLO
-v*I
0







00
0





os-






0


4-1



0


bo








0 0
Id m
04









0 o
00

oo


u4-i
0











m





u
0
0




o
cot







0 0








o
4-4






H *
OO
*H *i
3r-l 0




4C-

0
*H.

HM




0
CA]

00
Wcs
-H
Cl
4M


CO,






r )



4
4-1









































-AL


0



o













co

0o
4)j









<0 a
0


cd Q0
-a







r c
O.4
0c,









0 po
cd <

c














0a
PO4

cod

Un
cu to



BU



H 4-

P41 4U
^1 -il








and the upfield shifts are due to the ring current effect of the aromatic

nucleic acid bases. Both observations are compatible with an intercalation

mechanism for binding of the reporters to the nucleic acid.

In a manner analogous to that used by Jardetsky and Fischer to

determine the portions of the penicillin G molecule involved in binding
68
to the protein bovine serum albumin, a comparison of the upfield shifts

observed for the pmr signals of the different aromatic protons of the

reporter molecule in the presence of the nucleic acid can be used to

determine the geometry of the small molecule in the complex (Table 16)

The upfield shifts (A6) are calculated according to equation (10), where

free is the chemical shift for a particular proton at 90C in the

absence of DNA and 6b is the chemical shift for the proton at 90C
bound
in the presence of DNA. Comparisons of chemical shifts determined at

A = free bound (0)

lower temperatures are not meaningful due to the extensive line broadening

observed in the spectra of the DNA-reporter complexes under such

conditions. Due to the temperature conditions, the relevant DNA species

is the single strand, and the conclusions reached may or may not apply

to the native DNA-reporter complexes present at ambient temperature.

The "in" geometry postulated previously for the intercalated

reporter molecule is in agreement with the pmr evidence, as it would

explain the greater ring current effect experienced by the H-3 proton in

the 2,4-dinitroaniline series as compared to the H-6 proton (Figure 41).

The H-6 proton is located on the outside perimeter of the nucleic acid

structure and would be expected to experience less of the ring current

effect than does the H-3 proton located deep inside the nucleic acid.

The manner in which the upfield shifts for the various protons depend on








Table 16

Upfield Shifts in the Pmr Signals of the Aromatic Protons of
Reporter Molecules in the Presence of Salmon Sperm DNA.a

Reporter Upfield Shift (A6) in Hzb

H3 H5 H6


8 c c c

9 32 35 20

10 31 35 17

11 27 14 8

H3,5 H2,6

12 21 28

13 30 29

20 28 28


aSonicated salmon sperm DNA (MW 5
and reporters were 0.02 M.

bA = 6 6 where 6
b free bound' where free
the absence of DNA and 6bound is
bounding the presence of DNA.
in the presence of DNA.


x 105) was used at 0.16 M P/I in D20,


is the chemical shift for a proton in
the chemical shift for this proton in