Retrospective Theses and Dissertations
1994
G-wires: The growth and characterization of a
G4-DNA nanostructure
Thomas Cosgrove Marsh
Iowa State University
Follow this and additional works at: http://lib.dr.iastate.edu/rtd
Part of the Biochemistry Commons, and the Molecular Biology Commons
Recommended Citation
Marsh, Thomas Cosgrove, "G-wires: The growth and characterization of a G4-DNA nanostructure " (1994). Retrospective Theses and
Dissertations. Paper 11291.
This Dissertation is brought to you for free and open access by Digital Repository @ Iowa State University. It has been accepted for inclusion in
Retrospective Theses and Dissertations by an authorized administrator of Digital Repository @ Iowa State University. For more information, please
contact [email protected].
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter £ace, while others may
be from ax^ type of con^uter printer.
Hie quality of this reproduction is dqiend^t upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard maTync
and improper alignment can adverse^ affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note wiD indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing frt>m left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photogr:q>hs inchided in the original manuscr^ have been reproduced
xerogr^hically in this copy. Higher quali^ 6" x 9" black and white
photographic prints are available for any photogr^hs or illustrations
qypearing in this cqjy for an additional charge. Contact UMI directly
to order.
A Bell & Howell Information Company
300 North Zeeb Road. Ann Art5or. Ml 48106-1346 USA
313/761-4700 800/521-0600
Order Number 9518414
G-wires: The growth and characterization of a G4-DNA
nanostructure
Marsh, Thomas Cosgrove, Ph.D.
Iowa State University, 1994
UMI
300 N. Zeeb Rd.
Ann Art)or, MI 48106
G-wires: The growth and characterization of a G4-DNA nanostructure
by
Thomas Cosgrove Marsh
A Dissertation Submitted to the
Graduate Faculty in Partial Fulfillment of the
Requirements for the Degree of
DOCTOR OF PHILOSOPHY
Department: Zoology and Genetics
Interdepartmental Major: Molecular, Cellular, and Developmental Biology
Approved:
Signature was redacted for privacy.
In Charge of Major Work
Signature was redacted for privacy.
For the Inf^epartmental
Signature was redacted for privacy.
For the Major Department
Signature was redacted for privacy.
For the Graduate College
Iowa State University
Ames Iowa
1994
Copyright © Thomas Cosgrove Marsh, 1994. All rights resented.
ii
This work is dedicated to
Sharon, Alex and Entropy
iii
TABLE OF CONTENTS
GENERAL INTRODUCTION
1
Dissertation organization
1
Literature cited
2
LITERATURE REVIEW
3
Introduction
3
Guanine Structures
3
G-rich Sequences
5
Nanotechnology
10
Scanning Probe Microscopy
10
Molecular Self-assembly
11
Literature Cited
11
G-WIRES; SELF-ASSEMBLY OF A TELOMERIG
OLIGONUCLEOTIDE. d(GGGGTTGGGG) INTO LARGE SUPERSTRUCTURES.
16
Abstract
16
Introduction
16
Materials & Methods
18
Results
19
Discussion
21
Literature Cited
24
A NEW DNA NANOSTRUCTURE, THE G-WIRE, IMAGED
BY SCANNING PROBE MICROSCOPY
33
Abstract
33
Introduction
33
Materials & Methods
34
Results
35
Discussion
36
Literature Cited
38
GENERAL CONCLUSIONS
Literature Cited
45
50
1
GENERAL INTRODUCTION
The field of nanotechnology involves creating and controlling structures at the atomic level. In
theory, nanometer scale machines will assemble other molecular machines atom by atom (Drexler
1986). The origins of nanotechnology stem from Richard Feynman who in 1959 gave a lecture titled
"There's plenty of room at the bottom" (Feynman 1960). In this prophetic talk he introduced the
concepts for construction of nanometer scale devices that would revolutionize the fields of
computation, medicine and engineering. One of his sources of inspiration was derived from
biological systems. The nanostaictures and molecular machines of living systems provide numerous
examples for the creation of molecular switches, synthetic machinery, information storage and scaffold
structures. Exploiting the experience of nature is one of the strategies for creating nanostructures
Nucleic acids form a wide variety of different structures found in vivo and in vitro which could
be useful for creating unique nanostuctures. Nucleic acids have been proposed to have various
nanotechnology applications functioning primarily as scaffold stmctures. A nucleic acid nanostructure
with scaffold potential is the double helical DNA cube (Chen & Seeman, 1991). The DNA cube
exists as a single unit with the potential of further modification to form two and three dimensional
arrays. Creation of these proposed structures has not yet been achieved. Other methods have
been proposed for creating nucleic acid nanostaictures. Forming aggregates by molecular selfassembly has been a major goal of designers of nanoscale stmctures (Whitesides et al., 1991).
Nucleic acids are good candidates for forming nanostructures utilizing this synthesis strategy (Amato,
1993).
A great deal of research in the study of G-quartet DNA structures from oligonucleotide
studies indicates that they have the basic requirements for fomning of nanostructures. Custom
synthesis of oligonucleotides is reliable and relatively inexpensive which makes them ideal for
creating self-complimentary monomers for molecular self-assembly. Formation of G-DNA
superstructures limited to 4 or 5 repeats has been demonstrated for 3' end self-complimentary G4ONA (Sen & Gilbert, 1992). We have focused on G4-DNA as a system for creating self-assembling
nanostructures using a simple 10 nucleotide sequence d(GGGGTTGGGG).
Dissertation Organization
This dissertation contains two papers. The self-assembly process and characterization of Gwires by gel electrophoresis are examined in the first paper which has been accepted for publication
in Biochemistry. The second paper concerns the length, height and stability of G-wires. Conditions
that influence those attributes have been analysed by atomic force microscopy. The second paper
and has been submitted to Nucleic Acid Research. The two papers are preceded by a Literature
review then followed by the General Conclusions.
2
Literature Cited
Amato, I. (1993). Haute Coutre in chemistry. Sc/ence 260 p. 753-755
Feynman, R. (1961) There's Plenty of Room at The Bottom" reprinted in Miniaturization. H.D.
Gilbert, ed., New York, Reinhold.
Chen, J. H. & Seeman, N. C. (1991). "Synthesis from DNA of a molecule with the connectivity of a
cube." Nature 350 p 631-3.
Drexler, K. E. (1986). Engines of Creation. New York, Anchor Press/Doubleday.
Sen, D. & Gilbert, W. (1992). "Novel DNA superstructures formed by telomere-like oligomers."
Biochemistry
): 65-70.
Whitesides, G. M., et al. (1991). "Molecular Self-Assembly and Nanochemistry: A Chemical Strategy
for the Synthesis of Nanostructures." Science 254(29 November): 1312-1318.
3
LITERATURE REVIEW
Introduction
The self-associative properties of guanine nucleotides were first discovered in 1910. Heated
in the presence of KOH, a solution of GMP formed a viscous gel that showed unusual stablity (Bang
1910). These GMP gels were later studied by X-ray fiber diffraction and NMR which revealed that
their self-association was derived from four cyclic hydrogen bonded guanines in a single plane
(Zimmerman 1975). The ability of guanine to recognize itself and fonri a novel G-tetrad structures
suggested a biological role. Attention then turned to DNA sequences that could form these unique
guanine structures.
There are many G-rich motifs in the genomes of living organisms. Telomeric repeats,
mutational and recombinational hot spots and transcriptional regulation sequences have stretches of
contiguous guanine. Oligonucleotides derived from these sequences have been studied in the form
of oligonucleotides and form quadruple helical DNA. The tenn G-DNA was coined to describe
sequences that were capable of forming the novel quadmple helical G-tetrad stmctures. Detailed
analysis of sequence variation and monovalent cation stabilization has revealed that G-DNA is a highly
polymorphic family of structures. There is ample evidence that indicates that G-DNA stmctures play
an important role in biological systems (Blackburn 1991b; Sundquist 1991).
The growth of the new field of nanotechnology has created a demand for methods of creating
molecular machines from a bottom up approach. Biological systems have many examples of
materials which can be used to create molecular electronic devices, molecular machines and scaffold
structures (Amato 1993). Nucleic acids, specifically G-DNA can potentially perform those roles.
Guanine Structures
The structures formed by guanosine monophosphate (GMP), polyguanylic acid (poly(G))
and polyinosinic acid (poly(l)) have provided insight into the potential structure and function of
guanine-rich (G-rich) sequences found in the cell. Guanine: guanine (G:G) Interactions were first
reported by Bang (1910) who observed that a high concentration of 3' GMP formed a gel in an acid
solution (reviewed in Sundquist 1991). X-ray fiber diffraction data from GMP gels established that the
guanines formed a planer tetramer with two hydrogen bonds per guanine (Figure la). The guanine
tetrads were stacked on top of each other creating a right handed quadruple helix (Gellert et al.
1962). X-ray fiber diffraction analysis of poly(rG) and poly(l) show the formation of right handed
quadruple helices (Figure 1c) with helical parameters that match those of GMP (Amott et al. 1974;
Zimmerman Cohen et al. 1975; Zimmennan 1976).
4
Figure 1. The structure of the G-quadruplex and coordination of monovalent cation in the right handed
G quadruplex. (a.) The cyclic Hoogsteen hydrogen bonding scheme of the four guanines fomiing a
planer tetrad (Gellert, 1962). (b) The coordination of potassium in an eight fold chelation cage created
by the eight 06 oxygens of co-axially stacked G-tetrads (adapted from Sunquist & Klug, 1989). (c) A
stereo rendering of quadruple helical poly(rG) that was derived from X-ray fiber diffraction data and
model building (adapted from Zimmermann, et al., 1975). The parallel strands form a right handed
quadruple helical stmcture. The helical parameters are a 3.4A rise between adjacent tetrads and a
helical twist of 31° that yield a helix with a repeat of 11.5 bases and a pitch of 39.2°.
5
The stability of G quadruplexes is modulated by class la and class lla metal cations. The
order of stabilization proceeds as follows: class la (K > Rb > Na > Li & Cs) and class lla (Sr > Ca >
Ba > Mg) (Miles & Frazier 1978; Pinnavaia 1978; Howard & Miles 1982; Venczel & Sen 1994). The
degree of stabilization has been correlated to the size of the cation. The size of the cation is
important as it is predicted to reside in the axial channel between adjacent tetrads. The ion is held in
place by a chelation cage formed by the eight 06 oxygen groups of the guanines (Figure lb).
Potassium is the ideal size to fit into the chelation cage, based on metal to oxygen distances
calculated from the X-ray data, and therefore confers the greatest stability to the quadruplex
(Sundquist 1991). Sodium, with a low stabilizing effect, has two proposed binding schemes. The
first places Na'*' within the plane of the G-tetrad (Howard & Miles 1982; Williamson et al. 1989). The
second proposed binding scheme places Na'*' in the chelation cage between tetrads. However, this
would not maintain as strong an interaction as K"*" (Sundquist 1991). Additional forces that contribute
to quadruplex stability are the eight hydrogen bonds between guanines, and the base stacking
between tetrads.
G-rich Sequences
The capacity of G-rich sequences to self-associate into hairpin orquadaiplex structures
suggest that these structures may arise in the cell. Telomeres (Blackburn 1991; Price 1992),
immunoglobin switch regions (Shimuzu & Honjo 1984), and gene regulatory regions (Nickol et al.,
1983) are examples of such G-rich motifs. The essential dimerization of the HIV ribonuclear genome
has also been found to involve a G-quartet structure responsible for mediating the association
(Marquet et al. 1991). Sequences of telomeres have been the most extensively examined
structures. Telomeres are protein /DNA complexes at the termini of chromosomes. They are
responsible for presenting chromosome structural integrity and ensuring the complete replication of
the chromosome. The DNA component of telomeres are short repetitive double stranded
sequences consisting of a guanine rich strand running 5' to 3' toward the chromosomal terminus and a
complimentary cytosine rich strand (Blackbum 1991). Telomeres are replicated by a ribonuclear
protein terminal transferase called telomerase (Greider & Blackbum 1985). At the very end of the
telomere the guanine strand forms a 3' overhang of approximately two sequence repeats (Figure 2)
(Henderson & Blackbum 1989). Telomerase carries with it an RNA sequence complimentary to the
G-strand and functions as its own primer (Greider & Blackbum 1989). Telomeric repeat sequences
vary between different species but all have a G-rich strand.
G
Figure 2. Telomeric repeats at the termini of linear chromosomes. The G-strand of the telomere has
an approximate 16 base 5' overhang. Telomeric ends are therefore available for structure fonnation
with another telomeric G-strand overtiang (see figure 3).
7
Figure 3. Possible roles for G4-DNA and G-quartets in the eukaryotic chromosome, (a) The proposed
interaction of telomeres on homologous chromosomes associated with the nuclear matrix causing the
alignment of G-rich motifs A with A", B with B', and C with C during meiosis. These G-rich motifs may
assist in aligning sister chromatids during meiosis (from Sen & Gilbert, 1988). A parallel arrangement
(G4-DNA) is suggested for the association of telomeric DNAand the G-rich motifs, (b) The proposed
association of chromosomal termini facilitated by the 3' overhanging telomeric G-strand (from Sundquist
& Klug, 1989). Two hairpins dimerize to a quadruple helical structure with anti-parallel strand orientations.
This structure is referred to as the G-quartet (Williamson, et al., 1989).
8
The conserved nature of the G-rich 3' overhang suggests that some aspect of the sequence plays an
integral roll in telomere function. This assumption has directed research toward determining the
structure of G-rich sequences and their interaction with the other components of telomeres.
Telomeric G-strand sequences from various species were capable of forming novel hairpin
structures stabilized by G:G base pairs (Henderson et al. 1987). Subsequently, another unusual
DNA stmcture was reported for an oligo of the IgG switch region G-rich strand. The oligonucleotide
generated a four stranded structure referred to as G4-DNA (Sen & Gilbert 1988). Sen and Gilbert
proposed that G4-DNA could allow for the alignment of sister chromatids during meiosis and that
telomeric G-strand could form a quadruplex structure (Figure 3a). Oxytricha macronuclear DNA will
cohere in the absence of any protein and involves the interaction of telomeric DNA sequences (Oka &
Thomas 1987). Telomeric G-strand 3' overhangs form a quadruplex stmcture in the presence of
monovalent cation (Sundquist & Klug 1989; Williamson et al. 1989; Hardin et al. 1991) The model
features the folding of a single molecule to form a G-quartet (Figure 3b) (Williamson et al. 1989).
X-ray crystallography and NMR studies (Kang et al. 1992; Smith & Feigon 1992; Laughlan et
al. 1994) of the G-quartet have confirmed the predictions of previous models that were based on
chemical modification and model building. The structure is an anti parallel quadruple helix with
altemating syn and anti glycosidic linkages along each strand (Figure 4). NMR data of G4-DNA show
variability in structural details. G4-DNA is a parallel stranded quadruple helical structure with all the
glycosidic linkages in the anti conformation (Wang & Patel 1992; Aboul-ela et al. 1992; Jin et al. 1990;
Lu et al. 1993; Laughlan et al. 1994). Hairpin and parallel quartet structural transitions are dependent
on sequence composition, length, and the species of monovalent cation (Sen & Gilbert 1990; Hardin
et al. 1991; Balagummoorthy et al., 1992; Jin et al. 1992).
G4-DNA also exhibits additional self-association that is concentration dependent and requires
monovalent alkaline metal cations. The complexes proposed are end to end associations of the
base G4-DNA unit yielding G8, G12, and G16 structures (Sen & Gilbert 1992). The sequences used
were S'-TgGs and S'-TgGsT. The key feature necessary for self-association appears to be a
requirement for a 3' guanine (Lu et al. 1992; Sen & Gilbert 1992) as the presence of a 3' T residue
disrupts self-association. Another requiremnet for the formation of these superstructures was a
relatively high salt concentration, 1M KCI. The interpretation of this activity is that chromosome end
association could be mediated by G4-DNA.
9
a
4 A
G'2-DNA
mwm
G4-DNA
Figure 4. A cartoon of the two predominant conformations of G-DNA. Oxytrlcha 1.5 telomeric G-strand
repeat is capable of forming an antiparallel hairpin dimer (a) or a parallel stranded tetrsmer (b). The
conformations are termed G'2-DNA and G4-DNA resectively. The preference for one conformationover
the other is determined by the cation stabilizing the stmcture. This is known as the Na+ZK"*" switch (Sen
& Gilbert, 1990).
10
Nanotechnology
Nanotechnology, also known as molecular manufacturing, seeks to develop nanometer scale
machines that can manipulate atoms and assemble them into replicating assemblers, switches and a
variety of other molecular machines (Drexler 1936). The potential of this type of technology stands to
benefit science, medicine and the environment. Atom by atom assembly of structures requires the
development of technology that spans the fields of physics, chemistry, biology, and engineering to
produce faster and smaller computers, cellular repair machines, nanoscale sensors and methods for
cleaning environmental waste.
Living cells provide many good examples of molecular machines. Restriction enzymes
reconize specific sequences of DNA and make a precise cut. The protein synthetic machinery of the
cell, ribosomes, reads instructions and converts that information into a nascent polypeptide chain
molecule by molecule. The capability to create a molecular assembler like the ribosome does not yet
exist, however, the current prediction of achieving this is estimated to be within the next 50 years.
Progress toward this goal has been demonstrated by manipulation of atoms by scanned probe
microscopy (SPM) (Foster et al. 1988; Eigler & Schwizer 1990), production of atomic scale switches,
the development of supramolecular chemistry (Lehn 1988) and molecular self-assembly of highly
ordered structures (Whitesides et al. 1991).
Scanning Probe Microscopy
A generic SPM has a stylus probe that scans a sample in a rastor pattern and provides
topographical and other proximal force information as the probe interacts with surface features on the
sample. On extreamiy flat surfaces, atomic resolution topographical information may be obtained
(Hansma et al. 1988). An SPM may also be used to etch a surface (Stalling & Lagally 1994) and
manipulate individual atoms on a surface (Eigler & Schwizer 1990). The two primary types of SPM are
the scanning tunneling microscope (STM) and the atomic force microscope (AFM) (Binnig et al. 1986).
STM obtains topographic information by means of variation in a tunneling current between the
sample and the stylus probe attached to a piezoelectric crystal. Consequently samples need to be
conductive to be detected by STM. Non-conductive samples may be imaged by AFM. A highly
localized contact force t)etween the probe and the sample enable the AFM to image non-conductive
samples. Frequently, optical deflection of a laser beam off a flexible cantilever is used to monitor
this interaction.
11
Molecular Self-assembly
Nanotechnology requires the development of materials which allow the controlled positioning
of atoms in three dimensional space. One way to accomplish this goal is to use the strategy of
molecular self-assembly. Molecular self-assembly can produce large scale structures from relatively
small monomers. The structures formed in this way can exhibit a high degree of order and stability.
Molecular self-assembly is an equilibrium process that yields structures held together by weak forces
such as H-bonding, van der Walls interaction, hydrophobic effects and electrostatic interactions
(Whitesides et al. 1991).
Many biological structures are created by this type of process. For example,
oligonucleotides modified with hydrophobic constituents self-assemble to form aggregate structures
(Letsinger et al 1991). The formation of micelles, lipid bilayers, lipid tubules and even viral particles
such as TMV (Lauffer 1975) form by means of self-assembly. Structures that have recently been
developed, such as nanotubes, have been suggested as potential benefits to nanotechnology.
Protein nanotubes, self-assembled from a cyclic polypeptide, have been proposed as electronic
devices, a molecular trap and a channel (Ghadiri et al. 1993). Cyclodextrins also assemble into tube
structures which have been suggested to funtion in similar roles as other nanotubes (Li & McCown
1994).
Other biological molecules may have an impact on nanotechnology. Examples of these
include knots and cubes (Chen & Seeman 1991; Wang et al. 1993; Du & Seeman 1994). The ability of
nucleic acids to self-assemble into a variety of structures allow them to be employed in the bottom
up construction of nanoscaffolding. In addition to a scaffold, self-assembling nucleic acid junction
systems have been proposed to form the structural basis of a biochip memory device (Robinson &
Seeman 1987).
Literature Cited
Aboul-ela, F., Murchie, A. & Lilley, D. (1992). "NMR study of parallel-stranded tetraplex fomriation
by the hexadeoxynucleotide d(TG4T)." Nature 360:280-282.
Amato, I. (1993). "Haute Coutre in Chemistry" Sc/ence 260 p 753-755.
Amott, S., Chandraaekaran, R. & Maitilla, C. (1974). "Structures for polyinosinic acid and polyguanylic
acid." Biochem. J. 141:537-543.
12
Bang, I. (1910). "Untersuchungen uberdie guanylsaure." Bioch. Ztschr. 26:293-311.
Binnig, G., Quate, C. & Gerber, C. (1986). "Atomic force microscope." Phys. Rev. Lett. 56: 930-933.
Blackbum, E. H. (1991). "Structure and function of telomeres." Nature 350:569-573.
Blackburn, E. H. (1991). "Telomeres." Trends Biochem. Sci. 16(10): 378-381.
Chen, J. H. & Seeman, N. C. (1991). "Synthesis from DNA of a molecule with the connectivity of a
cube." Nature 3S0{6319): 631-3.
Drexler, K. E. (1986). Engines of Creation. New York, Anchor Press/Doubleday.
Du, S. M. & Seeman, N. C. (1994). The construction of a trefoil knot from a DNA branched junction
motif." S/opo/y/7>ers34(1): 31-7.
Eigler, D. M. & Schwizer, E. K. (1990). "Positioning of single atoms with a scanning tunneling
microscope." Nature 344:524-526.
Foster, J. S., Frommer, J. & Amett, P. (1988). "Molecular manipulation using a tunneling
microscope." Nature 33V. 324-326.
Gellert, M., Lipsett, M., Davies, D. (1962). "Helix fonnation by guanylic acid." Proc. Nat. Acad. Sci.
48:2013-2018.
Ghadiri, M. R., Granja, J., Milligan.R., Mc Ree, D., Khazanovich, N.. (1993). "Self-assembling organic
nanotubes based on a cyclic peptide architecture." Nature 366(25th November): 324-327.
Greider, C. W. & Blackbum, E. H. (1985). "Identification of a specific telomere terminal transferase
activity on Tetrahymena eyXracAs." Cell {A3): 405-413.
Greider, C. W. & Blackbum, E. H. (1989). "A telomeric sequence in the RNA of Tetrahymena
telomerase required for telomere repeat synthesis." Nature 337(6205): 331-337.
13
Hansma, P. K., Elings,V., Marti, O. & Bracl<er, C. (1988). "Scanning tunneling microscopy and atomic
force microscopy: application to biology and technology." Science 242(4876): 209-216.
Hardin, C. C., Henderson, E.,Watson, T. & Prosser,J.K. (1991). "Monovalent cation-induced structural
transitions in telomeric DNAs: G-DNA folding intermediates." Biochem. 30:4460-4472.
Henderson, E. & Blackburn, E. H. (1989). "An overhanging 3' terminus is a conserved feature of
telomeres." Mot. Cell. Biol. 9(1): 345-348.
Henderson, E., Hardin, C.C., Walk, S.K., Tinoco, I. & Blackburn, E. H. (1987). "Telomeric DNA
oligonucleotides form novel intramolecular structures containing guanine-guanine base pairs." Cell
51(6): 899-908.
Howard, F. B. & Miles, H. T. (1982). "A stereospecific complex of poly(l) with ammonium ion."
Biopolymers 21:147-157.
Jin, R., Breslauer, K., Jones, R.A. & Gaffney, 8. L. (1990). Tetraplex Formation of a GuanineContaining Nonameric DNA Fragment." Science 250:543-546.
Jin, R., Gaffney, B.L, Wang, C., Jones, R. A. & Bresluaer, K. J. (1992). Thermodynamics and
structure of a DNA tetraplex: A spectroscopic and calorimetric study of the tetramolecular complexes
of d(TG3T) and d(TG3T2G3T)." Proc. Natl. Acad. Sci. USA 89(18): 8832-8836.
Kang, 0. H., Zhang, X., atliff, R., Moyzis,R. & Rich, A. (1992). "Crystal structure of four-stranded
Oxytricha XB\omenc DNA." Nature 356:126-131.
Lauffer, M. A. (1975). Entropv-Driven Processes in Biologv. New York, Springer-Verlag.
Laughlan, G., Murchie, A.H., Norman, D. G., Moore, M.H., Moody, P., Lilley, D. & Luisi, B. (1994).
"The high-resolution crystal structure of a parallel-stranded guanine tetraplex." Science 265(5171): 520524.
Lehn, J.-M. (1988). "Supramolecular chemistry-Scope and Perspectives Molecules,Supermolecules
and Molecular Devices." Angewande Chemie Int. Ed. Engl. 27:89-112.
14
Li, G. & l\4cCown, L. B. (1994). "Molecular nanotube aggregates of b- and g- cylodextrins linked by
diphenylhexatrienes." Science 264:249-251.
Lu, M., Gou, Q. & Kallenbach, N. R. (1992^. Biochemistry 3V. 2455-2459.
Lu, M., Gou, Q. & Kallenbach, N. R. (1993). Thermodynamics of G-tetraplex formation by telomeric
DNAs." Biochemistry 32(2): 598-601.
Marquet, R., Baudin, F., Gabus, C., Darlix, J., Mougel, M., Ehresmann, C. & Ehresmann, B. (1991).
"Dimerization of human immunodeficiencey virus (type 1) RNA: stimulation by cations and possible
mechanism." Nucleic Acids Res. 19:2349-2357.
Oka, Y. & Thomas, C. A. (1987). The cohering telomeres of Oxytricha." Nuc. Acids Res. 15(21):
8877-8898.
Price, C. M. (1992). "Centromeres and telomeres." Curr. Opin. Cell Biol. 4(3): 379-384.
Robinson, B. H. & Seeman, N. C. (1987). The design of a biochip: a self-assembling molecular
memory device." Protein Engineering 1:295-300.
Sen, D. & Gilbert, W. (1988). "Formation of parallel four-stranded complexes by guanine-rich motifs
in DNA and its implications for meiosis." Nature 334:364-366.
Sen, D. & Gilbert, W. (1990). "A sodium-potassium switch in the formation of four-stranded G4-DNA."
Nature 344(6265): 410414.
Sen, D. & Gilbert, W. (1992). "Novel DNA superstaictures formed by telomere-like oligomers."
Biochemistry65-70.
Smith, F. W. & Feigon, J. (1992). "Quadruplex structure of Oxytricha telomeric DNA
oligonucleotides." Nature 356:164-168.
Stalling, C. T. & Lagally, M. G. (1994). "Fabrication of atomic scale structures on Si (001) surfaces."
Science 265:502-506.
15
Sundquist, W. I. (1991). The structures of telomeric DNA. Nucleic Acids and Molecular Biology. Berlin,
Springer-Veriag. 1-24.
Sundquist, W. I. & Klug, A. (1989). Telomeric DNA dimerizes by formation of guanine tetrads
between hairpin loops." Nature 342; 825-829.
Wang, H., Du, S. M. & Seeman, N. C. (1993). Tight single-stranded DNA knots." J. Biomol. Struct.
Dyn. 10(5): 853-63.
Whitesides, G. M., Mathias, J. P. & Seto, C.T. (1991). "Molecular Self-Assembly and Nanochemistry:
A Chemical Strategy for the Synthesis of Nanostructures." Science 254:1312-1318.
Williamson, J. R., Raghuraman, M. K. & Cech, T. R. (1989). "Monovalent cation-induced structure of
telomeric DNA: the G-quartet model." Cell 59(5): 871-880.
Zimmennan, S. B. (1976). "X-ray study by fiber diffraction methods of a self-aggregate of guanosine5'-phosphate with the same helical parameters as poly r(G)." J. Mol. Biol. 106: 663-672.
Zimmemian, S. B., Cohen, G. H. & Davies, D. R. (1975). "X-ray fiber diffraction and model-building
study of polyguanylic acid and polyinosinic acid." J. Mol. Biol. 92:181-192.
16
G-WIRES: SELF-ASSEMBLY OF A TELOMERIC OLIGONUCLEOTIDE, d(GGGGTTGGGG)
INTO LARGE SUPERSTRUCTURES.
A paper published in Biochemistry
Thomas Marsh and Eric Henderson
ABSTRACT: The telomeric DNA oligonucleotide 5'-G4T2G4-3' (Tet 1.5) spontaneously assembles
into large superstructures we have termed G-wires. G-wires can be resolved by gel electrophoresis
as a ladder pattern. The self-association of Tet 1.5 is non-covalent and exhibits characteristics of G4DNA, a parallel four-stranded structure stabilized by guanine tetrads. Formation of G-wires is
dependent upon the presence of Na+ and/or K+ and, once formed, G -wires are resistant to
denaturation. The assembly process further defines the proposed Na+/K+ switch mechanism
observed for G4-DNA. The results described here extend our understanding of the structural
potential of G-rich nucleic acids and may provide insight into the possible roles of G-rich sequences,
and the novel structures they can form, in biological systems.
INTRODUCTION
Telomeres are specialized protein-DNA complexes at the ends of linear eukaryotic
chromosomes. They are responsible for presen/ing chromosome structural integrity and stability,
ensuring complete replication of chromosomal termini and association of chromosomes with the
nuclear matrix (Blackburn, 1991; Henderson and Larson, 1991; Price, 1992). In most cases, telomeric
DNA consists of short repetitive sequences containing a guanine-rich strand running 5' to 3' toward the
end of the chromosome and a complimentary cytosine rich strand (Greider, 1991; Zakian, et al., 1990).
At the very end of the chromosome the guanine strand forms a 3' overhang of approximately two
sequence repeats (Henderson and Blackburn, 1989; Klobutcher, et al., 1981; Pluta, et al., 1982). In
many instances, telomeric repeat sequences from diverse species are very similar in composition.
Examples of telomeric sequences are TTAGGG in vertebrates and Trypanosomes, TTTTGGGG in
the ciliates Oxytricha and Euplotes, TTGGGG in Tetrahymena and TG1-3S. cerevisiae.
In the presence of monovalent cations, guanosine monophosphates, guanosine derivatives
(Guschlbauer and Chantot, 1976) and guanine rich oligonucleotides (Sundquist, 1991) including
telomeric sequences, have the capacity to form higher order structures. These higher order structures
are stabilized by G-quartets; planar structures composed of four Hoogsteen base-paired guanines
in a cyclical array (Henderson, et a!., 1987; Sen and Gilbert, 1988; Williamson, et al., 1989; Sundquist
17
and Klug, 1989; Kang, et al., 1992; Smith and Feigon, 1992). Many G-rich telomeric oligomers can
form a family of related structures containing G-quartets (G-DNA). For example, a single repeat of
the Tetrahymena telomeric G-strand sequence TGGGGT will adopt a four stranded parallel
quadruplex (Aboul-ela, et al., 1992) while two repeats, (1264)2, can adopt either a four stranded
parallel quadruplex known as G4-DNA or an anti-parallel hairpin dimer quadruplex known as G'2-DNA
(Sen and Gilbert, 1990). For the two repeat molecules the preference of one conformation over the
other is most influenced by the monovalent cation present (Na+ or K+) and DNA concentration. The
ability of different monovalent cations to affect G-DNA conformation has been termed the Na+/K+
switch (Sen and Gilbert, 1990).
Stabilization of G-quartets by the class la cations is ranked K>Rb>Na>Li=Cs. The class lla
metal cations are also capable of stabiling G-quartets and are ranked Sr>Ba>Ca>Mg (Venczel and
Sen 1994). The balance between the kinetics of structure formation and the stability of a structure is
an important factor in determining the ultimate favored confomiation. G'2-DNA, the anti-parallel
hairpin dimer quadruplex (Sen and Gilbert, 1988), is less stable than G4-DNA but faster forming
while the tetrameric parallel stranded quadruplex is more stable but slower to form (Lu, et al., 1993).
K"*" facilitates the formation of G'2-DNA by G-rich sequences due to its superior stabilizing ability.
However, in Na"*" the same sequences will adopt the G'2-DNA conformation, but will also be found in
the more stable G4-DNA conformation (Sen and Gilbert, 1990). The length of the G-rich oligomer will
also affect which conformation it will adopt. Specifically, single repeats of telomeric G-strand
sequences are too short to fold back into hairpins and therefore can only adopt the G4-DNA
conformation.
Superstructures formed by G-rich DNA's have been observed in the presence of K""". An
oligomer with a G-rich telomere-like sequence at its 3' end and a 5' T tail can fomi G4-DNA
superstructures (Sen and Gilbert, 1992; Lu, et al., 1992). The superstructures are composed of 8,12,
and 16 strands (G8, G12, and G16 DNA). G4-DNA with a 3' overhang of a single G promoted
formation of head to tail intemioiecular structures (Sen and Gilbert, 1992). The limited number of
superstructures that were obtained may have resulted from steric hindrance caused by the 5' T-tails
protruding from the superstructures. Formation of these superstructures usually required high salt
concentrations and oligonucleotide at concentrations up to 7.5 ^g/nL It is important to note that
superstructure fonmation was achieved in buffer containing KOI but, in contrast to the data presented
here, not in buffer containing NaCI.
The oligonucleotide 5'-GGGGTTGGGG-3' (Tet 1.5) mimics the 3' overhang from
Tetrahymena thermophila (Henderson and Blackburn, 1989). Here we report the self-assembly of
Tet 1.5 into long linear structures termed G-wires. Formation of these large structures is facilitated by
18
the presence of the blocks of guanines at both the 5' and 3' ends of Tet 1.5. This is the first
demonstration of spontaneous self-assembly of a telomeric sequence into large polymers (i.e..
greater than G16-DNA) in NaCI buffer and may provide new insight into the form and function of Grich sequences in biological systems.
MATERIALS & METHODS
Synthesis and purification of oligonucleotides. Oligonucleotides were synthesized using
standard phosphoramidite chemistry on an ABI 320 oligonucleotide synthesizer, boiled for 10' in 53
mM Tris-HCI, 53 mM Boric acid, 0.9 mM EDTA (0.6X TEE, final pH 8.3), 40% formamide, and
separated at 50 V/cm on a 20% acrylamide, 8 M urea, 0.6X TBE gel. Oligonucleotide bands were
visualized by UV shadowing against an intensifying screen and cut from the gel. DNA was eluted by
passive diffusion from the gel fragments in 10 mM Tris HCI (pH 8.0), 1 mM EDTA (TE). The
oligonucleotide was desalted, concentrated by CI8 (Waters) column chromatography and eluted in
50:50 methanolrwater. Eluant was lyophilized and dissolved in ddH20.
Non-denaturing gel electrophoresis was performed at either 4°C or room temperature (1820°C) in 10% acrylamide, 0.6X TBE gels with a field strength of less than 10 V/cm. Gel bands were
quantitated using a Phosphorlmager (Molecular Dynamics).
5' end and internal labeling of Tet 1.5. Tet 1.5 was radiolabeled either at the 5' end or
intemally near the 3' end with
jhe 5' end of Tet 1.5 was labeled using the fonvard reaction of T4-
polynucleotide kinase and y-^^p-ATP at 37°C for 30'. The concentration of oligonucleotide in the
reactions was typically 2 |xg/^L in a total volume of 20 ^L. T4 polynucleotide kinase reaction buffer
(referred to as KB throughout the text) conditions were 50 mM Tris-HCI (pH 7.5), 10 mM MgCl2, 0.1
mM EDTA, 0.1 mM spermidine, 0.1 mM DTT.
The internal 3' end of Tet 1.5 was radiolabeled using the T4 DNA polymerase replacement
synthesis reaction (adapted from Maniatis, et al., 1982) to incorporate 32p into the oligonucleotide.
Tet 1.5 was annealed to an equimolar amount of complimentary (C4A2)2 (500 nM in a 20 nL reaction
volume) in 50 mM Tris-HCI (pH 8.0), 50 mM MgCl2,1 mM DTT and treated with T4 DNA polymerase
at 11°C for 1 minute prior to adding a-32p-dGTP to the reaction. The reaction was then terminated
after 10 minutes by adding 1 ^.L of 0.5 M EDTA (pH 8.0) and boiling the sample for 5 minutes.
The products of the T4 polynucleotide kinase reaction and the T4 DNA polymerase
replacement synthesis reaction were gel purified in the manner described above with the exception
that autoradiography rather than UV shadowing was used to locate the radiolabeled oligomer bands
in the gel.
19
Inhibition of G-wire assembly by the presence of a 5' phosphate on Tet 1.5. Internally
labelled Tet 1.5 and non-radiolabelled Tet 1.5 were phosphorylated at the 5' end with cold ATP at
room temperature for 2 hrs. The oligonucleotides were then purified using the above procedure.
The denaturing gels included Tet 1.5 5'end labeld with 32P as a control. The 5' phosphorylated
intemally labelled Tet 1.5 was mixed with 2.5 ng of Tet 1.5 with (5'P-Tet1.5) or with out a 5'
phosphate. The samples were diluted to 10 ^1 H20, heated in a boiling water bath for 10' rapidly
cooled on ice then dried. The samples were dissolved in either 50 mM Na Buffer or 50 mM KCI
Buffer and incubated at 37°C for 24 hrs. The samples were then examined by denaturing gel
electrophoresis with 15% acrylamide gels at room temperature.
Sample preparation. The following description for setting up the incubations using non­
radioactive Tet 1.5 and the intemally labeled Tet 1.5 was standard procedure unless stated otherwise.
1 jiL of !ow concentration (10 nM) intemally labeled Tet 1.5 was added to 10 ng of Tet 1.5, diluted in
10 m-L 0.1 X TE and heated in a boiling water bath for 10 min. The denatured sample was cooled
rapidly on ice, then dried by vacuum evaporation and dissolved in the appropriate buffer to a final
concentration of 0.6 mM.
Monovalent cation selective stabilization of G-wires. Tet 1.5 (2^g/)iL) was 5' end labeled and
then diluted with buffer containing either 50 mM KCI KB (Kinase Buffer, see preceding section) or 50
mM NaCI KB. Control samples did not have any additional salt added to them. The samples were
diluted 1:1 in loading buffer (80% formamide, 0.6X TBE, 0.2% bromophenol blue, 0.2% xylene
cyanole FF) containing either 50 mM NaCI or 50 mM KCI and divided into two tubes. One half of the
sample was loaded directly onto a 20% acrylamide, 8M urea, 0.6X TBE gel and the second half of the
sample was boiled prior to loading on the gel.
Temperature of incubation and formation of G-wires. Intemally labeled Tet 1.5 was
resuspended in either 50 mM NaCI KB or 50 mM KCI KB at 4°, 25°, 37°, and 65° C for 24 hrs.
Samples were overlaid with mineral oil to prevent evaporation. An equal volume of loading buffer
was added to the samples, which were then separated on denaturing and non-denaturing 10%
acrylamide gels.
RESULTS
Generation of G-wires. It was found that moderately denaturing conditions could be used to
isolate a relatively clean distribution of higher order structures from 5' end labeling with 32P.
Denaturing gel electrophoresis of 5" phosphorylated Tet 1.5, which had been boiled in 40%
formamide, revealed a regular pattem of bands (Figure 1A). The bands represent DNA
20
superstructures that are resistant to the moderate denaturation. The incremental, regular size increase
suggested that a monomer unit was added at one end or both ends of the structures. These higher
order structures are referred to as G-wires.
To confirm that the G-wires were composed of Tet 1.5 monomer oligonucleotide individual Gwire bands were cut from the gel, eluted and desalted as described in Materials and Methods. The
desalted G-wire bands were denatured in 60% formamide loading buffer and separated on a
denaturing gel. All of the G-wire bands were composed of the Tet 1.5 oligonucleotide (Figure 1B).
A small percentage of the 9mer premature termination product d(GGGGTTGGG) (Tet1.4), was
associated with the purified G-wire bands. The proportion of Tet 1.5 to Tet 1.4 increased with the
increase in G-wire size suggesting that Tet 1.4 could be involved in termination of G-wire growth.
To test the stability of the G-wires, purified G-wire bands were isolated and eluted from
acrylamide gels. Upon analysts of single isolated bands by electrophoresis, the banding pattern that
was originally observed was, to a large extent, regenerated. The maximum size of G-wires in the
regenerated banding pattern was related to the original size of the purified G-wire but was slightly
greater than that band size (Figure 10). In addition, the amount of monomeric Tet 1.5 decreased as
the size of the G-wire increased. We were unable to obtain G-wire formation from Tet 1.5 that was
homogenously 5' phosphorylated (Tet 1.5 monomer band in Figure 1A). 5' phosphorylated
oligonucleotides have a faster rate of mobility than non-5' phosphorylated oligonucleotides. This
suggested the presence of a 5' phosphate on Tet 1.5 may hinder G-wire formation.
5'phosphate inhibition of G-wire fonnation. Homogeneous 5'P-Tet 1.5 did not redily fomi Gwires under conditions that ordinarily yield many G-wires (Figure 2, lanes 1 and 2). When compared to
5'P-Tet 1.5 intemally labeled 5'P-Tet 1.5 mixed with Tet 1.5 demonstrated a significantly imporved
formation of G-wires (Figure2, lanes 3 and 4). It is not clear if the 5'P-Tet 1.5 higher order structures
are G-wires. The high molecular wieght bands have similar mobilities between the +P and -P but the
maximum size is diminished and at least 2 bands are abscent from the +P samples. The 5'P-Tet 1.5
easily participated in G-wire fomriation when combined with Tet 1.5 but it is not known where it is
inserted. The presence of a 5' phosphate somehow disfavors assembly into G-wires while a small
amount of 5'P-Tet 1.5 can incorporate into G-wires.
Monovalent cation dependent assembly of G-wires. To test the hypothesis that G-wires are
G-DNA structures we examined their growth and stability in Na''' and K"**. A feature of G-DNA's in
general is stabilization by the monovalent cations with K"*" conferring greater stability than Na***. Gwires grown in the presence of 50 mM NaCI and then incubated for 10 minutes in either 50 mM NaCI,
50 mM KCI or no additional salt showed a dramatic difference in resistance to denaturation when
boiled in 40% formamide and examined by denaturing gel electrophoresis (Figure 3A). The t>oiled
21
sample treated with K"*" was highly resistant to denaturation (Figure 3A, lane 5) however, the G-wire
samples treated with Na'*' or no additional salt were totally denatured (Figure 3A, lanes 3 and 1
respectively). Interestingly, longer treatment with 50 mM KCi gave a different result. G-wires were
grown in 50 mM NaCI then treated for 1 hour in KCI concentrations that ranged from 0 mM to 100 mM.
The treated samples were boiled in 40% fomiamide and examined by denaturing gel electrophoresis
(Figure 3). Treatments with as little as 5 mM KCI were capable of providing some resistance to
denaturation (Figure 3B, lane 3). Smearing of the G-wire banding pattern occurred in the 50 and 100
mM KCI treatments (Figure 3B, lanes 6 and 7).
Temperature dependent formation of G-wires. Tet 1.5 was incubated in the presence of
either 50 mM NaCI KB or 50 mM KCI KB at 4°, 23°, 37° and 65° C for 24 hrs. The growth of G-wires
was detected by native (Figure 4A) and denaturing (Figure 4B) electrophoresis. The samples
incubated in NaCI showed a mari<ed improvement in G-wire formation with increasing temperature,
peaking at 37°C. At 65°C there was very little G-wire formation in Na**". Comparing the native and
the denaturing gels for the Na+ grown G-wires, it was observed that they had a more complex
banding pattern under non-denaturing conditions. Apparently denaturation removed the less stable
conformations contributing to the complexity of the banding pattem observed under non-denaturing
conditions. Thus, comparison of native and denaturing gels reveals the most stable structures, i.e.,
perfectly aligned G-wires.
The samples incubated in K+ showed little growth of G-wires at 4°C or at 23°C. Under nondenaturing conditions, incubation at 37°C and 65°C in K"*" created a large smear of irregular higher
order structures. The elimination of less stable conformations under conditions of denaturing
electrophoresis revealed a G-wire banding pattem like that observed in the presence of Na'*'. In
addition, an apparent compression of the K"*" banding pattem is observed at 65°C under denaturing
conditions.
These results show that it is possible to manipulate the incubation conditions to generate Gwires in either Na"*" or K""" and the temperature at which G-wires form depends on the type of
movalent cation that is present.
DISCUSSION
General summary. The telomeric oligonucleotide d(GGGGTTGGGG) (Tet 1.5) is capable of
spontaneous self-assembling into superstructures that are most likely stabilized by G-quartets.
These structures, termed G-wires, can be resolved as a regular pattem of bands by gel
electrophoresis. Growth of G-wires with Tet 1.5 is much more extensive and occurs at lower cation
22
concentrations than that of previously reported superstructures (Sen and Gilbert, 1992; Lu, et al.,
1992).
Hypothetical G-wire structure. On the basis of observations presented here we propose a
model that represents the most probable G-wire conformation (Figure 5). In this model the G-wire
consists of G4-DNA domains punctuated by T nodes. The initial starting structure consists of a G4
domain containing four quartets fonned by the association of the 5' half of a duplex Tet 1.5 with the 3'
half of another Tet 1.5 duplex, forming a slipped tetrapiex structure with G-duplex "sticky-ends".
The strands run parallel to each other and can accept additional duplex Tet 1.5 at either end. The
slipped tetrapiex association of Tet 1.5 is consistent with the regular ladder pattem seen by gel
electrophoresis. Sen & Gilbert (1992) proposed a slipped architecture for the superstructures they
observed that is conceptually similar to that proposed here for Tet 1.5. The increased stacking
interactions would enhance the stability of the G-wire.
G-wires have G-DNA characteristics. In the presence of Na+ and K+ Tet 1.5 assumed
different conformations. At low temperatures K+ stabilized conformations that did not lead to G-wire
fomriation. Assuming the behavior of Tet 1.5 was consitant with cun-ent literature (Sen and Gilbert
Venczel and Sen Williamson et la), these structures could have been hairpin dimers (G'2-DNA).
Although Na"*" forms the same conformations as K'^, the weaker stabilizing ability of Na"*" could not
stabilize them as well as K"*". Therefore, random structures formed in Na**" are short lived compared
to K"*" while the more stable four-stranded structure leading to formation of G-wires predominates.
Assembly of G-wires occurred most efficiently in Na'*', yet a greater degree of stability was
acquired by the addition of K***. A similar phenomenon was reported for the association of
chromosomal DNA containing telomeric repeats (Oka and Thomas, 1987) and for a yeast telomeric Gstrand oligonucleotide (Venczel and Sen, 1993). Growing G-wires in Na'*' followed by a brief
incubation in K+ vividly demonstrated the strong stabilizing effect of K+ on G-wires. What is most
striking about this is the resistance to thermal denaturation and imperviousness to denaturing agents
(8 M urea and 40% formamide). The stabilizing effect was only observed for brief (i.e. less than 30
min.) treatments with K+. Exposing G-wires to K+ for longer periods of time resulted in their
disassembly. Thus, G-wires are not static structures. They undergo assembly and disassembly in a
relatively short time scale in comparison to some telomeric G4 structures, which can have half lives on
the order of many hours (Raghuraman and Cech, 1990). This is cleariy illustrated by the regeneration
of the G-wires from a single isolated band (Figure 10). A variety of transient structures arise in
solution as part of the dynamic equilibrium of G-wire assembly. Stabilization of the resulting
structures by addition of potassium would account for the smearing of the G-wires on gels.
23
Other factors influencing G-wire fomnation. The major factors that influence the fomiation of
G-wires are the species of monovalent cation, 5' phosphorylation, DNA concentration and the
temperature at which the formation occurs. The ionic species effect was discussed in the preceding
section. The 5' phosphate effect, in which phosphorylation of Tet 1.5 inhibited formation of G-wires,
is an interesting phenomenon that is difficult to explain at this time. A 5' phosphate may cause steric
hindrance and or electrostatic interference with adjacent G4-domains.
The effect of temperature is related to the requirement for reducing the stability of structures
that impede G-wire growth and allowing stabilization of the G-wire precursors. Incubation in
at
65°C and above resulted in the formation of stable high molecular weight species, the structural
details of which are unknown, but are likely to be irregular G-wires. These stmctures appear as a
smear on a non-denaturing gel. Regular G-wires (containing only blocks of 4 G-tetrads) maximize the
number of possible quartets that stabilize the entire superstructure. Denaturing gel electrophoresis
revealed that regular G-wires, the most stable structural species, were present within the smear of
irregular structures. A similar obsen/ation was reported in which thermal denaturation of a telomeric
sequence in a K"*" buffer induced structure formation as the temperature approached 95°G (Lu, et al.,
1993). This is also consistent with the phenomenon of high temperature guanosine monophosphate
gel formation in the presence of K+ and poly-G resistance to thermal denaturation (Pinnavaia, et al.,
1975:Zimmemnan, 1976).
Biological relevance of G-wires. Are structures related to G-wires biologically relevant? G4
binding proteins exist in eukaryotic cells (Chung, etal., 1992; Liu, etal., 1993; Pearson, etal., 1993;
Walsh and Gualberto, 1992; Weisman-Shomer and Fry, 1993; Schiererand Henderson, 1993; Fang
and Cech, 1993). It has been shown explicitly in two cases (Schierer and Henderson, 1993; Fang and
Cech, 1993) and implied in others, that G4 binding proteins bind best when there is a singlestranded domain adjacent to the G4 structure. This is the case with G-wires, which are bound by a
G4-DNA binding protein from Tetrahymena thermophila (Schierer and Henderson, 1993). This
suggests that G-wires at the very least mimic some DNA conformation that may exist in vivo. A
number of phenomenon, such as recombination, meiotic chromosome pairing, gene rearrangement
and chromosome polytenization require self-recognition similar to that observed with G-wires and all
other G4 DNA structures. Thus, the further elucidation of the structural variability of G4 DNA will shed
light on its role in these, and perhaps other, cellular phenomena. Finally, the conformational dynamics
and special features of G-wires provide new and provocative twists in the study of the biology and
chemistry of self-recognizing nucleic acid structures.
24
ACKNOWLEDGMENTS
We would like to thank S. Ahmed, T.P. Schierer, S.L.K. Marsh, A.M. Myers, A. Kintinar and the others who
assisted in the preparation of this manuscript through critical evaluation and discussion. This work was
supported by a grant from the NIH to E.H.
LITERATURE CITED
Aboul-ela, F., Murchie, A. I. H. and Lilley, D. M. J. (1992) Nature 360, 280-282.
Blackburn, E. H. (1991) Nature 350,569-573.
Chung, I. K., Mehta, V. B., Spitzner, J. R. and Muller, M. T. (1992) Nuc. Acids Res. 20,1973-1977.
Fang and Cech, T. R. (1993) Cell 74,875-885.
Greider, C. W. (1991) Curr. Opin. Cell Biol. 3,444-451.
Guschlbauer, W. and Chantot, J. F. (1976) Proceedings of the conference of synthesis, structure and
chemistry of the tRNAs and their components. (M Wiewion3wski, Ed.)96-^^\4.
Henderson, E. and Blackburn, E. H. (1989) Mol. Cell. Biol. 9, 345-348.
Henderson, E., Hardin, C. C., Walk, S. K., Tinoco, I. J. and Blackburn, E. H. (1987) Cell 51, 899-908.
Henderson, E. and Larson, D. D. (1991) Curr. Opin. Gen. Dev. 1,538-543.
Kang, C. H., Zhang, X., Ratliff, R., Moyzis, R. and Rich, A. (1992) Nature 356,126-131.
Klobutcher, L. A., Swanton, M. T., Donini, P. and Prescott, D. M. (1981) Proa. Natl. Acad. Sci. USA
78,3015-3019.
Liu, Z., Frantz, J. D., Gilbert, W. and Tye, B. K. (1993) Proc. Natl. Acad. Sci. USA 90, 3157-3161.
Lu, M.. Guo, Q. and Kallenbach, N. R. (1992) Biochemistry 31,2455-2459.
Lu, M., Guo, Q. and Kallenbach, N. R. (1993) Biochemistry 32, 598-601.
Maniatis, T., Fritsch, E. F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold
Spring Harbor University Press, Cold Spring Harbor.
Oka, Y. and Thomas, C. A. (1987) Nuc. Acids Res. 15,8877-8898.
Pearson, A. M., Rich, A. and Krieger, M. (1993) J. Biol. Chem. 268,3546-3554.
Pinnavaia, T. J., Miles, H. T. and Becker, E. D. (1975) J. Am. Chem. Soc. 97,7198-7200.
Pluta, A. F., Kaine, B. P. and Spear, B. B. (1982) Nuc. Acids Res. 10, 8145-8154.
Price, C. M. (1992) Curr. Opin. Cell Biol. 4, 379-384.
Raghuraman, M. K. and Cech, T. R. (1990) Nuc. Acids Res. 18,4543-4552.
Schierer, T. P. and Henderson, E. R. (1994) Biochemistry, in press.
Sen, D. and Gilbert, W. (1988) Nature 334,364-366.
25
Sen, D. and Gilbert, W. (1990) Nature 344,410-414.
Sen, D. and Gilbert, W. (1992) Biochemistry 31,65-70.
Smith, F. W. and Feigon, J. (1992) Nature 356,164-168.
Sundquist, W. I. and Klug, A. (1989) Nature 342, 825-829.
Venczel, E. A. and Sen, D. (1993) Biochemistry 32,6220-6228.
Walsh, K. and Gualberto, A. (1992) J. Biol. Chem. 267,13714-13718.
Weisman-Shomer, P. and Fry, M. (1993) J. Biol. Chem. 268,3306-3312.
Williamson, J. R., Raghuraman, M. K. and Cech, T. R. (1989) Cell 59, 871-880.
Zakian, V. A., Runge, K. and Wang, 8. S. (1990) Trends. Genet. 6,12-6.
Zimmerman, S. B. (1976) J. Mol. Biol. 106, 663-672.
Zimmerman, S. B., Cohen, G. H. and Davies, D. R. (1975) J. Mol. Biol. 92,181-192.
26
FIGURE LEGENDS
Figure 1. A) Non-denaturing gel showing G-wire banding pattern after labeling Tet 1.5 at the 5' end
with 32p_ The concentrations of oligonucleotide were 1ng/(il and 5^lg/nJ. The individual bands are
numbered 1-9 while 10 is the collection of all the lowest mobility G-wires. B) Denaturing gel
electrophoresis of the purified G-wires. Desalted G-wires were boiled for 10 minutes in 60%
formamide loading buffer and run on a 20% acrylamide, 8M urea, 0.6X TBE gel and mn at 50 V/cm.
The isolated G-wire bands are listed 1-10. Tet 1.5 (10 mer) is indicated by ** and Tet 1.4 (9 mer) is
indicated by * on the gel. C) Native electrophoresis of individual isolated G-wire bands on a 10%
acrylamide, 0.6X TBE gel run at lOV/cm and 4°C. The G-wires examined were loaded in 50 mM
NaCI, 40% sucrose, 0.6X TBE, 0.1% bromophenol blue, 0.1% xylene cyanole FF. Abbreviations: M,
oligonucleotide size marker; n-1, Tet1.4: Tetl.5,5'-32p end labeled Tetl.5.
Figure 2. The 5' phosphate inhibition of G-wire growth. Samples incubated at 37°C fro 24 hrs were
mixed with a equal volume of 40% formamide loading buffer and run 5V/cm at room temperture on a
15% acrylamide, 8 M urea, 0.6 X TBE gel. Lanes 1 and 2 are homogeneous 5'P-Tet 1.5,including
internally radiolabelled tracer, and lanes 3 and 4 are intemally radiolabelled 5'P-Tet 1.5 tracer mixed
with Tet 1.5. Abbrevitions: N, 50 mM NaCI KB; K, 50 mM KCI KB; +P, bulk 5' phosphoryated Tet
1.5; -P, bulk non-5' phosphorylated Tet 1.5; M, Kpn lligated linkers of 8 bp increments.
Figure 3. Potassium stabilization of G-wires. A) G-wires grown in the presence of 50 mM NaCI KB
were incubated with 50 mM KCI KB 10 minutes prior to boiling in 40% fonnamide loading buffer and
separation by gel electrophoresis. Denaturing gel electrophoresis shows the selective stabilization
and maintenance of the Na'^ G-wire banding pattem by K'''. B) Denaturing electrophoresis gel of
G-wires grown in 50 mM NaCI buffer and then treated for 1 hour with an increasing concentration of
KCI prior to boiling and separation on a denaturing gel. Samples were run on a 20% acrylamide, 8 M
urea, 0.6X TBE gel at 50 V/cm.
27
Figure 4. Temperature dependence of G-wire formation. A) Native 10% acrylamide gel run in 0.6X
TBE. Samples were either incubated in 50 mM NaCI buffer or 50 mM KCI buffer at the indicated
temperatures. Prior to loading, the samples were divided into equal portions and mixed 1:1 with the
appropriate loading buffer. B) Denaturing gel of the temperature incubation samples.
Abbreviations: C, Tetl.5 in the absence of additional cation; M, oligonucleotide size markers.
Figure 5. Proposed G-wire model. Tet 1.5 adopts an out of register (slipped) tetrameric association
forming G4-DNA domains stabilized by monovalent cations. The overlapping slipped G4-DNA
structures are responsible for formation of the G-wire. This model is speculative but consistent with
the data presented herein.
mgAnl Tetl.5
1
5
G-wires
M 1 2 3 4 5 6 7 8 910 M
T^tl.5
Figure 1
•«
P A
G-wires
5
1 3 5 6 8 1 0 P
M
29
Na+P
Na-P
K+P
Figure 2
K-P
M
30
mMKCl
NS Na* S*.
10'Boil + - + • I* • I
M
IbtlJS
12 3 4 5 6
1 2
Figure 3
31
Native
Lane 123456789
123456789
Figure 4
32
Figure 5.
33
A NEW DNA NANOSTRUCTURE, THE G-WIRE, IMAGED BY
SCANNING PROBE MICROSCOPY
A paper submitted to Nucleic Acids Research
Thomas C. Marsh, James Vesenka and Eric Henderson
ABSTRACT
G-DNA is a polymotphic family of quadruple helical nucleic acid structures containing guanine
tetrad motifs (G-quartets) (1, 2). Guanine rich oligonucleotides that are self-complimentary, as found
in many telomeric G-strand repeat sequences, form G-DNA in the presence of monovalent and/or
divalent metal cations. In this report we use the atomic force microscope (AFM) to explore the
structural characteristics of long, linear polymers formed by the telomeric oligonucleotide
d(GGGGTTGGGG) in the presence of specific metal cations. In the AFM these polymers, termed
G-wires, appear as filaments whose height and length are detemiined by the metal ions present
during the self-assembly process. The highly ordered, controllable self-assembly of G-wires could
provide a basis for developing advanced biomaterials.
INTRODUCTION
Formation of ordered structures by molecular self-assembly has been a pursued as an
effective strategy for development biomaterials for nanotechnology applications (3, 4). Nucleic acids
have useful properties that may be exploited. The variety of conformations that nucleic acid
sequences can adopt provide numerous opportunities to create useful nanostructures, primarily as
scaffolds (5). Custom synthesis of oligonucleotides is reliable and relatively inexpensive. Moreover,
the methods by which nucleic acids can be derivatized (6) provides a mechanism for construction of
nanostructures with diverse chemical and physical characteristics. Cube-shaped, branched, knotted
and other structures have been constructed using double helical DNA (7, 8, 9). Those structures are
based upon Watson-Crick complimentarity and have the potential to form two and three dimensional
arrays.
The ability of some nucleic acids to recognize not only their Watson-Crick compliment but also
their twin enhances their potential to form useful nanostructures. G-DNA is a very stable fourstranded structure containing a planar, tetrameric arrangement of guanines termed the G-quartet
(Figure 3, inset) (1). Short G-rich oligonucleotides with at least three consecutive guanines can form
34
G-DNA structures in the presence of specific class la (Na+, K"'", Rb+) and lla (Mg"''+, Ca+"'", Ba++,
Sr*"*") metal cations (10). Formation of G-DNA superstructures limited to 4 or 5 repeats has been
demonstrated for 3' end self-complimentary G4-DNA (11,12). We have utilized G4-DNA as a system
for creating large, self-assembling nanostructures using the simple 10 nucleotide sequence,
d(GGGGTTGGGG) (Teti .5), that is self-complimentary at its 5' and 3' ends (13). We call these
structures G-wires.
Numerous studies have proven the utility of the AFM (14) as a powerful tool for imaging and
analyzing nucleic acids (15,16,17,18). In the AFM, interactions between a sharp stylus and the
sample are detected as the tip scans the sample. The tip may be in constant or transient contact with
the sample and the force applied precisely controlled. The data collected can be displayed as a 3D
topograph of the sample, with sub-Angstrom accuracy in the Z (height) dimension. Here, the AFM is
used for the first time to make direct measurements of a new DNA structure, the G-wire .
MATERIALS AND METHODS
Preparation of oligonucleotide
Oligonucleotides were synthesized using standard phosphoramidite chemistry. Purification of
the oligonucleotide was performed by denaturing gel electrophoresis as previously described (19).
Lyophilized oligonucleotide was dissolved in 80% formamide, 0.6 X TBE loading buffer and
denatured in a boiling water bath for 5 min. The oligonucleotide was loaded on a 30 cm long 1.5 nnm
thick 20% polyacrylamide, 0.6 X TBE gel and electrophoresed at 800 V. The gel was stopped when
the bromophenol blue tracking dye in a separate lane reached 5 cm from the gel bottom. The
oligonucleotide was visualized by UV shadowing, cut from the gel and eluted overnight in TE, pH 8.0.
The oligonucleotide was concentrated and desalted by CI8 (Waters) column chromatography. 2.5 ^g
aliquots of Teti.5 were diluted in water and denatured by boiling followed by rapid cooling and
lyophilization.
Atomic force microscopy (AFM)
Teti .5 was dissolved in the appropriate buffer and incubated at 37°C for 12-24 h. After
incubation the samples were diluted 1:100 in 10 mM Tris-HCI, pH 7.5,1 mM MgCl2 and 10m.L was
deposited onto a freshly cleaved mica substrate. The G-wires were allowed to adsorb for 5 min.,
washed with 1 mL sterile water, and rapidly dried in a stream of N2 gas. Imaging was performed with
125^m Si tapping probes at relative humidity < 10% using a Nanoscope III equipped with a Dscanner (Digital Instmments, Inc.).
35
Gel electrophoresis
Teti .5 was radiolabeled internally with 32p as described previously (13). The extent of G-wire
fomiation in 50 mM NaCI, 5 mM KCI; 10 mM MgCl2: 50 mM NaCI, 10 mM MgCl2; 50 mM KCI, 10
mM MgCl2; and TE, pH 7.5 was determined by native and denaturing polyacrylamide gels. Salt
solutions were buffered with 50 mM Tris-HCI, pH 7.5. All samples were incubated 24 hours at 37°C
and then electrophoresed through a 15% acrylamide, 0.6 X TBE gel (native) and 15% acrylamide, 0.6 X
TBE, 8 M urea gel (denaturing). Loading buffer for native electrophoresis contained 20% sucrose, 1.2
X TBE, while loading buffer for denaturing electrophoresis contained 80% formamide, 1.2 X TBE,
0.05% bromophenol blue and 0.05% xylene cyanole FF. Each G-wire incubation was mixed 1:1 with
loading buffer and loaded on the gels without boiling.
RESULTS
G-wires self-assemble into long filamentous polymers
The self-assembly of G-wires was found to occur efficiently at 37°C in buffer containing 50
mM NaCI, 10 mM MgCl2, IjiM spermidine and 50 mM Tris-HCI pH 7.5 (Na+ZMg+^/spermidine) (13).
G-wires grown in Na+ZMg+^/spermidine buffer were examined by AFM which revealed they were
polydisperse linear polymers (Figure 1). The length of G-wires ranged from 10 nm to greater than
1000 nm. The height and width of the G-wires was found to be uniform with few bends, kinks or
branches. This indicated that G-wires were ordered, relatively rigid polymers. The height of G-wires
was found to be 2-3 times greater than the height plasmid DNA in the AFM (Figure 1, g-i and Table
1).
Magnesium induced synergy of G-wire self-assembly
In the AFM G-wires appeared as relatively straight, variable length polymers, as expected
for a tubular structure with a central canal filled with a tight fitting cation. The length of G-wires is a
function of the incubation conditions with particular dependence upon ionic environment, temperature
and time. This dependence is shown by AFM imaging (Figure 2A) and gel electrophoresis (Figure
2B) of G-wires grown under various conditions. In Na"*" or K+ alone, relatively short G-wires were
generated. In Mg''"*', longer G-wires were observed at low frequency. The longest and most
plentiful G-wires were grown in Na+/Mg+"'' buffer. K+/Mg++ grown G-wires were somewhat shorter
in length than Na'''/Mg''"'' G-wires and more structurally heterogeneous. The corresponding gel
electrophoresis of G-wire incubations reflected these results.
AFM height measurements and G-wire stability
The heights of G-wires depended upon the growth conditions. These data are summarized
in Table 1. K+/Mg++ grown G-wires had an average height of 23.1 ± 2.0 A. G-wires grown in
36
Na+/Mg++ buffer have an average height of 16.9 ± 1.4 A. Those formed in Mg++ alone had an
average height of 12.7 ± 1.6 A. This difference in heights suggests that G-wire compressibility is a
function of cationic growth conditions. Additional buttressing of the structure may be conferred by
cation (e.g., Mg''"'') binding to the external surface. Although spermidine was not required for G-wire
formation (Figure 2 A & B), a significant increase in G-wire height was observed when spermidine
was present in the growth buffer (Table 1). The gallery of G-wires grown in Na+ZMg+^/spennidine
buffer is shown in Figure 1. Spermidine and spermine are known to bind to double helical regions of
tRNA (20) and duplex DNA (21). Triple helical DNA formation is also enhanced by the presence of
such polycations (22). The apparent increase in G-wire height in the presence of spermidine could
be indicative of a true increase in molecular diameter due to spermidine binding or improved
resistance to compression forces from the scanning probe.
DISCUSSION
The height of G-wires measured in the AFM under some conditions was close to the
diameter of G-quartets determined by X-ray crystallography (-28A) (23.24). In contrast, B-fomri DNA
appears much flatter in the AFM than predicted from its molecular diameter of approximately 20 A,
with observed heights typically in the range of 5-10 A (Figure 2, panels g, h and i) (17, 25,26). The
reduced height of B-DNA is probably caused by a number of factors which include sample
compression (27) and/or friction related artifacts (28). Friction artifacts can be revealed and
compensated for by examination of apparent heights in both scan directions (29) and do not play a
role in the observed height of G-wires (data not shown). Thus, G-DNA appears to be more
resistant to compression in the AFM than does B-DNA. The reduced compressibility of G-wires
implies that structures based on the G-quartet motif are structurally robust, which is not unexpected
for an ion filled tube. An alternative explanation for reduced height of B-DNA imaged by AFM is
relaxation of the helix such that it lies flat on the substrate in a non-helical 'railroad track" configuration
(30), which might facilitate sequence determination by scanning probe methods. Nicked circle and
linear B-DNA could topologically accommodate unwinding of the helix if surface binding forces were
sufficient to induce this degree of distortion. In this case, a quadruple helical structure would have
greater resistance to this type of conformation change and, therefore, appear taller. This idea is
illustrated in Figure 3.
Formation of G-wire polymers depends upon G-quartets as a basis of self-complimentarity
and stability. The ability of Teti .5 to self-assemble into long polymers clearly demonstrates a
potential as scaffold structures for nanotechnology applications. G-wires are structurally more robust
than duplex DNA when imaged by AFM. This feature is important when considering manipulation of
37
structures by the scanning tip. Nucleic acids can be readily chemically modified with functional
moieties, which vastly increases their potential as scaffolds for nanostructure design and construction.
In addition to a scaffold role, some of the physical properties of G-wires, such as the central canal
within which cations reside and an electron-rich extemal surface, suggest that they may be useful as
functional components of nanoscale electronic devices.
Table 1
Height comparison of G-wires and plasmid DNA. Measurements of the heights
of G-wires grown in the listed buffers encompass data from at least two different
sample preparations. The number of individual data points per sample used to
calculate averages is indicated (n).
G-wire growth conditions
Measured Height (A)
Na+/Mg++
16.9 ±1.4
n==90
K+/Mg++
23.1 ±2.0
n==100
Mg++
12.7 ±1.6
n==50
Na+/Mg++/spermidine
23.9 ±1.1
n==150
Plasmid (pSK II)
6.3 ±1.1
n==161
38
ACKNOWLEDGMENTS
The authors thank L. Ambrosio, S. Ahmed, T. Schierer, and S. Marsh for helpful comments and
criticisms. This work was supported in part by grants from NIH and NSF. The authors are members
of the ISU Signal Transduction Training Group and Laboratory for Cellular Signaling.
REFERENCES
1.
Williamson, J.R., Raghuraman, M.K. and Cech, T.R. (1989) Cell 59,871-880.
2.
Williamson, J.R. (1993) Proc. Natl. Acad. Sci. USA 90,3124-3124.
3.
Whitesides, G.M., Mathias, J.P. and Seto, C.T. (1991) Science 254,1312-1318.
4.
Ghadiri, M.R., Granja, J.R., Milligan, R.A., McRee, D.E. and Khazanovich, N. (1993) Nature
366,324-327.
5.
Amato, I. (1993) Science 260,753-755.
6.
Letsinger, R.L., Chaturvedi, S.K. and Farooqui, F. (1993) J. Am. Chem.Soc. 115,7535-7536.
7.
Chen, J.H. and Seeman, N.C. (1991) Nature 3S0, 631-3.
8.
Du, S.M. and Seeman, N.C. (1994) BiopofymersZA, 31-7.
9.
Seeman, N.C. (1991) Dna Cell Biol 10,475-86.
10.
Venczel, E.A. and Sen, D. (1993) Biochemistry 32,6220-6228.
11.
Sen, D. and Gilbert, W. (1992) Biochemistry31,65-70.
12.
Lu, M., Guo, Q. and Kallenbach, N.R. (1992) Biochemistry 31,2455-2459.
13.
Marsh, T.C. and Henderson, E. (1994) Biochemistry33,10718-10724.
14.
Binnig, G., Quale, C.F. and Gerber, C. (1986) Phys. Rev. Lett. 56, 930-933.
15.
Allen, M.J., Fan Dong, X., O'Neill, T.E., Yau, P., Kowalczykowski, S.C., Gatewood, J.,
Balhom, R. and Bradbury, E.M. (1993) Biochem. 32,8390-8396.
16.
Bustamante, C., Keller, D. and Yang, G. (1993) Cum Opin. Struct. Biol. 3, 363-372.
17.
Hansma, H., Vesenka, J., Siegerist, C., Kelderman, G., Morret, H., Sinsheimer, R.L.,
Elings, V., Bustamante, C. and Hansma, P.K. (1992) Science 2S6,1180-1184.
18.
Hansma, H.G., Bezanilla, M., Zenhausem, F., Adrian, M. and Sinsheimer, R.L. (1993) Nucleic
Acids Res. 21, 505-512.
19.
Henderson, E., Hardin, C.C., Walk, S.K., Tinoco, I.J. and Blackbum, E.H. (1987) CellSI, 899908.
20.
Teeter, M.M., Quigly, G.J. and Rich, A. (1980) In Spiro, T.G. (ed.). Nucleic Acid-Metal Ion
Interactions. John Wiley & Sons, Inc., Neiv York, pp. 145-177.
21.
Schmid, N. and Behr, J.-P. (1991) Biochem. 30,4358-4361.
39
22.
Hampel, K.J., Crosson, P. and Lee, J.S. (1991) Biochem 30,4455-4459.
23.
Kang, C.H., Zhang, X., Ratliff, R., Moyzis, R. and Rich, A. (1992) Nature 356,126-131.
24.
Laughlan, G., Murchie, A.I.H., Nomnan, D.G., Moore, M.H., Moody, P.C.E., Lilley, D.M. and
Luisi, B. (1994) Science 2BS, 520-524.
25.
Bustamante, C. (1991) Annu Rev Biophys Biophys Chem 20,415-446.
26.
Thundat, T., Allison, D.P., Wannack, R.J. and Ferrell, T.L. (1992) UltramicroscopyA2-44,1101.
27.
Bustamante, C., Vesenka, J., Tang, C.L., Rees, W., Guthold, M. and Keller, R. (1992)
Biochemistry 3^, 22-26.
28.
Yang, J. and Shao, Z. (1993) Ultramicroscopy 50,157-170.
29.
Vesenka, J., Manne, S., Gibberson, R., Marsh, T. and Henderson, E. (1993) Biophys. J. 65,
30.
Yagil, G. and Sussman, J.L (1986) EMBOJ. 5,1719-1725.
31.
Zhong, Q., Inniss, D., Kjoller, K. and Elings, V.B. (1993) Surf. Sci. Lett. 290,688-692.
AO
FIGURE LEGEND
Figure 1. AFM gallery of G-wires grown in the presence of spermidine. The average height of the
G-wires is 23.9 ± 1.1 A and the lengths range from 10 nm to >1000 nm. Panels a-i are G-wires grown
with Teti.5 at a concentration of 0.25 |xg/M.L in buffer containing 50 mM NaCI, 50 mM Tris-HCI, pH 7.5,
10 mM MgCl2, and 1 nM spemiidine at 37°G for 12-24 hours. Panel f is an image of G-wires
deposited with 15 nm colloidal gold (indicated by arrow) as an intemal height standard. Panels g-i are
G-wires deposited with pSKII (Stratagene) plasmid DNA. The apparent height of the plasmid DNA
was -7A. Images were collected with a Nanoscope III D-scanner (Digital Instruments, Inc.) using
contact and tapping modes (31). Scale bars: a and b =500 nm, c = 200 nm, d = 50 nm, e-i = 100 nm.
Figure 2. Tapping mode (31) AFM images and electrophoretic analysis of Tetf.s G-wires grown in
different buffer conditions. Conditions used for G-wire growth were identical for AFM and gel
analyses. A. After incubation the samples were diluted 1:100 in 10 mM Tris-HCI, pH 7.5,1 mM
MgCl2 and deposited onto a freshly cleaved mica substrate. The samples were incubated for 5 min.,
washed with 1 mL sterile water, and rapidly dried in a stream of N2 gas. Imaging was performed with
125^m Si tapping probes at relative humidity <10% using a Nanoscope III with a D-scanner (Digital
Instruments, Inc.). The scale bar = 250 nm. 6. Native and denaturing gels show the extent of Gwire formation in 50 mM NaCI, 5 mM KCI; 10 mM MgCl2; 50 mM NaCI, 10 mM MgCl2; 50 mM
KCI, 10 mM MgCl2; and TE, lanes 1-6 respectively. Lane 7 contains 8 bp Kpnl linker oligonucleotide
ligation ladder. Samples were incubated 24 hours at 37°C and then analyzed on 15% acrylamide, 0.6
X TBE gels (denaturing gels included 8 M urea).
Figure 3. Model of differential stability of duplex DNA and tetraplex DNA during imaging by AFM.
The height of plasmid DNA is significantly diminished compared to the height of the G-wire. A
possible explanation is that an uncoiling of the double helix occurs as the DNA molecule is adsort)ed
to the substrate and dried (30). The quadruple helical G-wires have a coordinated cation that
provides additional stability to the structure, thereby providing resistance to conformation change
upon surface binding. The heights of the G-wires correlate with the stabilizing capacity of the
coordinated cation (K"'">Na"'">Mg"'"+) (10).
41
Figure 1
42
Figure 2A
Na/Mg
44
G-wire
G-quartet
Figure 3
45
GENERAL CONCLUSIONS
The oligonucleotide Teti.s
capable of ordered self-assembly into a filamentous polymer
that we term a G-wire. Electrophoretic studies imply that G-wires are based on a G4-DNA slipped
register conformation that self-assemble in the presence of monovalent and divalent metal cations.
The cation dependant difference in structural stabilty, the regular electrophoretic pattern, binding by
the G4-DNA specific protein TGP strongly suggest G-wires are a G4-DNA nanostructure. AFM
studies show G-wires to be relatively uniform polymers with a significantly greater stabilty compared
to plasmid DNA. The conditions that lead to the self-assembly of G-wires can be manipulated to
obtain different length distributions and stability. A diagram of the G-wire assembly pathway is
depicted in Figure 1a.
The self-assembly process of G-wires could be explained as an entropy driven process
(Lauffer 1975). Water is structured in the vicinity of hydrophobic surfaces and charged molecules. This
reduces the entropy of the water. When ions and hydrophobic surfaces are removed from the water
there is a increase in entropy (Israelachvili 1992). During G-wire assembly, a net increase in entropy
occurs when Tetl.5 and the coordinating cation release water. G-wires associate by means weak
forces such as H-bonding, base stacking and hydrophobic effect. The combined stabilizing effect of
these forces increases with an increase in G-wire length and is greatly increased by the presence of
cations like K+ and Na"*". G-wires are novel structures and the data collected thus far suggests that Gwires could benefit the development of nanotechnology.
The goal of nanotechnology is the ability to controlis the position of atoms. Replicating
assemblers are the theoretical nanomachines that would assemble molecules of any design and
purpose atom by atom. The ability to create such a device is one problem and controlling it is
another. Scaffolding (nanoscaffolding) structures that would support assemblers switches are
needed. G-wires have the potential to serve as a nanoscaffold. The self-assembly of G-wires is a
convenient method for creating a polymer upon which other structures or machines may be attached.
Synthesis of functionally modified Tet-j.s can be used to augment the self-assembly properties of Gwires and create more stable structures. Tet 1.5 decorated with a number of useful moieties such as
intercalating drugs, cationic substituents, hydrophobic moieties, biotin, etc could be synthesised. For
example the 5' end of an oligonucleotide may be amino modified and subsequently biotinylated.
The biotin may be bound by streptavidin conjugated to a paramagnetic bead or colloidal gold
particle (Figure 2). Cross linking groups will be included that can be activated after self-assembly to
increase the stability of the G-wires and produce two dimensional G-wire nets. Streptavidin contains
four binding sites for biotin and could generate a three dimensional network. A G-wire
46
Tell^ monomer
5*
y
•
•*•
Dead end 1
\
nil iiii»
Dead end 2
If
G*2-DNA
G4-DNA
G^uartet
G-wires
B
>v>
Tier <.6
ttrrtnim
<;as^
«« (T)
Figure 1. A) Proposed self-assembly pathway of G-wires. Teti .5 can adopt a number of different
conformations. The hairpin dimer (G'2-DNA) shown on the left and the in register parallel tetramer (G4DNA) shown on the right are dead ends in temris of polymer formation. The central path of out of register
conformations would lead to G-wire fonnation. The structures formed in the central path would produce
a regular ladder pattern on an electrophoretic gel and have G4-DNA "sticky ends' that are analogous
to those produced by some type II restriction endonucleases on duplex DNA. B) Water associated with
ions and Tetl.5 oligonucleotide is resleased from a confined state to a mobile state as a consequence
of G-wire formation. This results in a net increase in entropy favoring formation of long G-wire polymers.
47
<G-wire)
P*t'"-"Z^.^-*(Streptavidiri)
Figure 2. G-wires attached to a streptavidin anchor by a biotin/aminolinker. Streptavidin has four biotin
binding sites and therefore can link G-wires into a three dimensional network.
4Q
nanoscaffolding could also support nanomachines, molecular switches and molecular sensors. G-wires
attached to paramagntic beads would allow them to t>e manipulated with great precision by a
magnetic AFM probe. By this method nm-scale electronic circuit templates and structures such as
molecular corrals could be made. Structures constructed in the way could be used as masks for the
production of nanoscale electronic circuits. Synthesis of a bifurcated oligo based on Tetl.5 would also
be a method for creating self-assembling G4-DNA networks. A triskelion G-DNA molecule could
form a hexagonal network (Figure 3).
An intriuging feature of G-wires is the inhibition of self-assembly by the presence of a 5'
phosphate. The cause of the 5' phosphorylation inhibition remains speculative but indicates that
initiation of G-wire growth may be controlled. Such cotroll could be achieved by blocking the 5' end
of Teti.5 with a photolabile protecting group. Blocked Tet1.5 monomers could be introduced into an
environment that would ordinarily induce the formation of G-wires. Lipid tubules would be a suitable
environment for the testing of the conductive properties of G-wires.
Future experiments may also explore the biological relevance of G-wires. Biology is
replete with examples of macromolecular structures constructued from smaller molecules. Viral
particle assembly and formation of lipid bilayers occur spontaneously in aqueous environments by
means of self-assembly. Formation of ordered macromolecular structures by means of self-assembly
is likely to have played an important role in developing the first cellular organisms. It has been
theorized that the chemical processes of life were originally based on RNA (Gech 1985; Gilbert 1986;
Weiner 1986) That is to say prior to the time of proteins, RNA and nucleic acids in general functioned
as enzymes, structures etc. In vitro molecular evolution has demonstrated that, when selected for,
RNA sequences can evolve to bind a specific ligand (Wang, McCurdy et al. 1993), catalyse reactions
and inhibit enzymes (Cech 1986). Ribosomal RNA, tRNA and other cellular RNA's are the
remaining representatives the RNA world. Forming assemblies and maintaing stabilty could have
employed self-assembling structures such as G-wires. The oligonucleotide Tet-j.s is 10 bases long
could conclevably be created by a ribozyme and or a self-replicating strategy (Hoffmann 1992). A
system such as this may generate the supply of Tetf .5 to form G-wires in an in vitro molecular
evoltion system geared toward forming nanoscaffolding.
Another question that may pertain to G-wires is the origin of telomeres in eukaryotic
organisms. Most eukaryotes have a telomeric repeat that is short (-6bp) and G-rich. The selection
of this type of sequence for a protective end of a linear chromosome may possibly be due to the
ability to form G-quartets. It would be interesting to explore how telomeres evolved and how the
self complimentary nature of G-rich sequences like Teti.5 3*^ involved.
49
j34-DNA>
Linker
(j34-DNA Hexagon)
Figure 3. A bifurcated Tet1.5 or tris!<elion G-rich oligonucleotide could form hexagonal G4-DNA structures
that would function as self-assembling two and three dimensional networks.
50
Literature Cited
Cech, T. R. (1985). "Self-splicing RNA: implications for evolution." Int Rev Cytol93{3): 3-22.
Cech, T. R. (1986). "RNA as an enzyme." Sci Am 255(5): 64-75.
Gilbert, W. (1986). "Origin of Life The RNA worid." Nature 319: 618.
Hoffmann, S. (1992). "Artificial Replication Systems." Angew. Chem. Int. Ed. Engl. 31:1013-1016.
Israelachvili, J. (1992). Intermolecular &Surface Forces. San Diego, Academic Press.
Lauffer, M. A. (1975). Entropy-Driven Processes in Biology. New York, Springer-Verlag.
Wang, K. Y., S. McCurdy, et al. (1993). "A DNA aptamer which binds to and inhibits thrombin exibits
a new strucural motf for DNA." Biochemistry32:1899-1904.
Weiner, A. M. (1988). "Eukaryotic nuclear telomeres: molecular fossils of the RNP world?" Cell 52{2):
155-8.