 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention relates to a method for attaching nucleic acids to surfaces
to permit structural analysis of the nucleic acids. More particularly,
this invention relates to a method for attaching DNA or RNA to a gold
surface for structural analysis of the DNA or RNA by Auger electron
spectroscopy, angle-dependent x-ray photoelectron spectroscopy, scanning
tunneling microscopy, atomic force microscopy, and the like.
Structural analysis of nucleic acids suddenly became important in 1953 when
Watson and Crick discovered that deoxyribonucleic acid (DNA) is the
biological molecule that stores genetic information and transfers that
information from generation to generation. In succeeding years,
ribonucleic acids (RNAs) were also shown to play a role in genetics as
messenger RNAs and transfer RNAs. Further, studies of viruses such as
tobacco mosaic virus (TMV) demonstrated that RNA was capable of serving as
a repository of genetic information.
Prior to 1977, structural analysis of DNA and RNA was difficult, expensive,
and time-consuming. However, in that year two methods for determining the
sequence of bases in DNA were discovered independently that made
nucleotide sequencing of DNA easier, cheaper, and faster. These two
methods are the chemical degradation method of Maxam and Gilbert, 74 Proc.
Nat'l Acad. Sci. USA 560-64 (1977), and the chain termination method of
Sanger, 74 Proc. Nat'l Acad. Sci. USA 5463-67 (1977). At about the same
time, techniques of molecular cloning were developed so that DNA copies of
RNA molecules could be cloned and sequenced by these two methods. Direct
RNA sequencing techniques were also developed.
The chemical degradation method of DNA sequencing is based on the concept
of partially degrading DNA fragments through four base-specific
degradation reactions, one for each of the four bases. Degradation of a
base makes the phosphodiester backbone more susceptible to chemical
cleavage. Thus, after the bases are specifically degraded, the DNA
fragments are subjected to another reaction to break the phosphodiester
backbone. The specifically degraded fragments are then size fractionated.
Labeling of the fragments with radioactive, fluorescent, chemiluminescent
or other labels permits the fragments to be detected. Thus, a nucleotide
can be identified and assigned to each position in the nucleotide chain by
the reaction that specifically degraded the base at that position.
The chain termination method also relies on four different reactions to
deduce the identity of each nucleotide in a chain. However, these
reactions involve synthesis of nucleotide chains rather than their
degradation. DNA is normally double-stranded and each base in a nucleotide
chain is bonded to a complementary base in the other nucleotide chain.
Guanine (G) pairs with cytosine (C) and adenine (A) pairs with thymine
(T). Thus, if one knows the sequence of one strand of the DNA, the
sequence of the other strand is also known. DNA polymerases are available
which will synthesize a complementary strand of DNA if provided with a
single-stranded DNA template and an oligo- or polynucleotide primer for
providing a hydroxyl group to which the next nucleotide in the chain is
attached. Addition of chain terminating nucleotide analogs, such as
2',3'-dideoxynucleoside triphosphates that lack the hydroxyl group to
which the next nucleotide in the chain would ordinarily be attached, makes
it possible to terminate a chain at every possible nucleotide position. By
using four chain-terminating reactions, each one, respectively, containing
a chain terminating analog of one of the four nucleotides in DNA, the
sequence of nucleotides in the chain can be determined. The chains are
labeled as in the chemical degradation technique with radioactive,
fluorescent, chemiluminescent, or other labels. The chains are then
fractionated by length. In this way the sequence of nucleotides in the
chain can be deduced by identifying the nucleotide analog that terminates
a chain at each position in the DNA.
Despite the huge improvement that these two techniques have been to
determining the structure of DNA molecules at the nucleotide sequence
level, significant additional improvements are still needed to increase
the speed and reduce the cost of sequencing large nucleic acids. Without
such improvements it will not be feasible to sequence the entire three
billion basepairs of DNA that comprise the entire genetic complement of a
human being in a timely and economical manner. Other large sequencing
projects will, likewise, be impractical. Alternative methods of nucleotide
sequencing to those just described are being developed.
A method that has been suggested for rapid sequencing of nucleic acids is
through imaging of the nucleic acid with techniques such as scanning
tunneling microscopy (STM) and atomic force microscopy (AFM). These
recently developed methods utilize microscopes that make it possible to
resolve or visualize individual atoms in some samples. In principle, it
should be possible to attach nucleic acids to suitable substrates and
visualize the nucleic acids by STM or AFM. The nucleotide sequence of a
nucleic acid could be read by visually identifying each base from an image
of the nucleic acid, or by detecting some other non-visual signal
(spectroscopic) which is unique to the various bases. In practice, the
potential for sequencing nucleic acids by STM or AFM imaging has suffered
from several technical stumbling blocks. For example, suitable substrates
are needed that are both atomically flat and free of contaminants or other
artifacts that would interfere with producing readable images. A further
problem has been that the nucleic acid must adhere to the substrate so
that the nucleic acid does not move during the minutes that are needed to
produce the image. Clearly, a moving target is not an ideal subject for a
readable image. Thus, a method of attaching nucleic acids to an atomically
flat, contaminant free substrate so that the nucleic acids are firmly
anchored and do not move would be very important to structural analysis of
nucleic acids using methods that distinguish the atomic or molecular
structure of the nucleic acid. Further, a method is needed for binding the
nucleic acid to the substrate so that the nucleic acid is oriented for
imaging of the bases. Binding the phosphodiester backbone of the nucleic
acid to the substrate would, thus, seem to be the best way of attachment
for these purposes.
Binnig and Rohrer, the inventors of STM, were the first to image DNA
molecules, G. Binnig et al., 49 Phys. Rev. Lett. 57-60; G. Binnig & H.
Rohrer in Trends in Physics at 38-46, J. Janta & J. Pantoflicek, eds.
(European Physical Society, The Hague, 1984), however progress has gone
only as far as resolving the major and minor grooves of uncoated DNA, T.
Beebe et al., 243 Science 370-72 (1989); G. Lee et al., 244 Science 475-77
(1989), and distinguishing purines from pyrimidines in uncoated
polynucleotides, D. Dunlap and C. Bustamante, 342 Nature 204-06 (1989).
These experiments were conducted with highly oriented pyrolytic graphite
(HOPG) substrates. HOPG was originally the obvious substrate of choice
because of its advantages of having limited reactivity, low surface
roughness (less than 5 .ANG. vertical deviation over hundreds of .ANG.
lateral distance), reproducibility of flatness, low cost, and ease of
preparation. However, many ambiguous HOPG surface structures could be
confused with deposited biomolecules, C. Clemmer & T. Beebe, 251 Science
640 (1991), thus making HOPG an undesirable surface for this work. Thus,
other substrates were sought.
Prior to 1983, gold had not been extensively investigated as a substrate
for chemisorption studies due to its relative inertness towards molecular
oxygen, and even carbon monoxide interacted only weakly. Nuzzo and Allara,
105 J. Am. Chem. Soc. 4481-83 (1983), discovered that thiols and
disulfides could be adsorbed from solution to form ordered monolayers on
gold films. The monolayers are formed because of relatively strong (30-40
kcal/mole) covalent bonds between the sulfur and gold molecules.
Bain et al., 111 J. Am. Chem. Soc. 321 (1989), and Bain et al., 111 J. Am.
Chem. Soc. 7155 (1989), described further the nature of the bonds formed
between organosulfur compounds and gold. The following information is
extracted from these two articles. There is a specific interaction of gold
with sulfur and other "soft" nucleophiles and a low reactivity toward most
"hard" acids and bases. The strong specific interaction between gold and
sulfur atoms in thiols, disulfides, and certain other sulfur-containing
compounds induces spontaneous assembly of an adsorbed monolayer at the
gold-solution interface. It is the formation of strong, coordinative
gold-sulfur bonds that drives the spontaneous assembly of these
monolayers. It is the position of Bain et al. that the species ultimately
formed on the gold surface by adsorption of thiols from solution is a
thiolate (Au-SR). It is stated, however, that the mechanism by which an
initially physisorbed thiol is converted to a chemisorbed thiolate remains
unclear. In other words, while the mechanics of the chemistry are not
clear, the fact that bonds are formed is known. Monolayers of alkanethiols
on gold appear to be stable indefinitely in air or in contact with liquid
water or ethanol at room temperature.
Concerning the kinetics of formation of monolayers, Bain et al., 111 J. Am.
Chem. Soc. 321, 328 (1989), stated that the rate of formation of a
self-assembled monolayer is influenced by many factors, some of which can
be controlled relatively easily, such as temperature, solvent,
concentration and chain length of the adsorbate, and cleanliness of the
substrate. Other factors, such as the rate of reaction with the surface
and the reversibility of adsorption of the components of the monolayer,
are inherent to the system. They concluded that experimental conditions
must be established for each new system studied. At moderate
concentrations (ca. 1 mM), the adsorption process is characterized by two
distinct phases, an initial period of rapid adsorption lasting a few
minutes in which the monolayer reaches a thickness of 80-90% of its
maximum, and a slower period lasting several hours, during which the
thickness slowly approaches its final value. This behavior can be
rationalized by rapid adsorption of an imperfect monolayer followed by a
slower process of additional adsorption and consolidation, possibly
involving displacement of contaminants, expulsion of included solvent from
the monolayer, and lateral diffusion on the surface to reduce defects and
enhancing packing.
This discovery led to work involving a wide range of applications in fields
including electrochemistry, biology, and microlithography. Related
chemical systems have been utilized for STM of DNA molecules. In L.
Bottomley et al., 10 J. Vac. Sci. & Tech. 591 (1992), a gold surface was
activated by reaction with N,N-dimethyl-2-mercaptoethylamine to create a
monolayer of exposed cationic groups. DNA was then bound to the monolayer
by coulostatic interactions. This result offered one possible resolution
to the major problem of holding nucleic acids in place during STM and AFM
imaging.
U.S. Pat. No. 5,106,729 to Lindsay et al. describes a method of attaching
base-substituted, phosphate-substituted, or sugar-substituted
polynucleotides to gold substrates for analysis by STM or AFM. In this
method, oxygen atoms in the base, phosphate, or sugar of the
polynucleotide are replaced by sulfur atoms. The sulfur-containing
polynucleotide is then treated with a mercury compound or other
metal-containing compound to form complexes between the mercury or other
metal and the sulfur atoms for enhancing contrast during the imaging
process. Then, the metal-complexed polynucleotide is attached to the gold
substrate by Faradaic deposition (electrodeposition) by holding the gold
substrate about 1-2 V positive with respect to a reference electrode. It
is believed that the mercury forms an amalgam with the gold substrate,
thus binding the polynucleotide to the substrate. Subsequent imaging is
performed in water. This method does not ensure that covalent bonds are
formed directly between the nucleic acid and the substrate to firmly
attach the nucleic acid to the substrate nor that the bases are exposed
and unreacted so that they can be imaged or otherwise analyzed.
Herein is described a method of using gold-thiol monolayer chemistry to
anchor nucleic acids for imaging by STM and AFM. Instead of activating the
gold substrate for binding of the nucleic acids by coulostatic
interactions, or binding the nucleic acids to the gold substrate by
forming an amalgam between a nucleic acid-metal complex and the gold
substrate, the nucleic acids are activated to permit covalent bonding of
the phosphate backbone of the nucleic acid to the gold substrate, thus
leaving the bases exposed and unreacted for imaging and analysis. In view
of the foregoing discussion, it will be appreciated that these advantages
are a significant advancement in the art.
OBJECTS AND SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method of anchoring
nucleic acids to a surface to permit structural analysis of the nucleic
acid by scanning tunneling microscopy, atomic force microscopy, Auger
electron spectroscopy, and the like.
It is another object of the invention to provide a method of attaching
nucleic acids to a surface so that the DNA does not move during the
structural analysis.
It is also an object of the invention to provide a method of attaching
nucleic acids to a surface so that attachment is through the
phosphodiester backbone.
It is another object of the invention to provide a method of attaching
nucleic acids to a surface so that the nitrogenous bases are exposed for
structural analysis.
It is a further object of the invention to provide a method of attaching
nucleic acids to a surface so that the nitrogenous bases may be analyzed
by scanning tunneling microscopy and atomic force microscopy.
It is still another object of the invention to provide a method of
attaching nucleic acids to a surface for analysis by Auger electron
spectroscopy.
It is yet another object of the invention to provide a method of attaching
nucleic acids to a surface that is atomically flat and free of artifacts
that interfere with scanning tunneling microscopy and atomic force
microscopy.
It is yet a further object of the invention to provide a method of
attaching nucleic acids to a gold surface.
A still further object of the present invention is to provide a method and
means of anchoring thiolated DNA or RNA to a gold surface to permit the
sequencing of nucleotides in the DNA or RNA chain by scanning tunneling
microscopy, atomic force microscopy, Auger electron spectroscopy, and the
like.
These and other objects may be accomplished by a method for adsorbing
nucleic acids to the surface of a gold substrate for analyzing the
structure of the nucleic acid that includes the steps of thiolating the
nucleic acid by substituting at least one non-bridging internucleotide
oxygen of each phosphodiester moiety with sulfur, depositing the thiolated
nucleic acid on the gold surface, and subjecting the nucleic acid to
analysis by means for determining the atomic or molecular structure
thereof. The gold surface is prepared by subjecting a gold single crystal
to mechanical polishing, electropolishing, cleaning by cycles of Ar.sup.+
sputtering and annealing under vacuum until no contamination is detected
by Auger electron spectroscopy, and flame annealing and quenching in
methanol. A gold surface can also be prepared by vapor deposition on an
atomically flat surface such as mica. The thiolated DNA is deposited on
the gold surface for a sufficient time for covalent bonds to form between
the sulfur and the gold. Scanning tunneling microscopy, atomic force
microscopy, and Auger electron spectroscopy are used to analyze the
nucleic acid structure. As used herein the terms DNA and RNA can be used
interchangeably because the invention is applicable to the thiolation and
attachment to gold surfaces of both molecules. By "thiolation" is meant
the substituting of at least one non-bridging internucleotide oxygen of
each phosphodiester moiety in the nucleotide chain with sulfur for the
subsequent formation of an Au--S--P covalent bond and not the formation of
thiol or mercapto ("--SH") groups on the nucleotide bases.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a graph of results from Auger electron spectroscopy of a clean
gold single crystal and a gold single crystal which had been treated with
water for 17 hours.
FIG. 2 is a graph of results from Auger electron spectroscopy of a gold
surface on which a pentadecanucleotide (15-mer) containing a thiolated
backbone was deposited.
FIG. 3 is a graph showing peak ratios characteristic of DNA deposition and
ratioed with respect to the principal gold peak for modified and
unmodified DNA samples compared to clean and blank gold samples.
FIG. 4 is a graph showing peak ratios characteristic of DNA deposition and
ratioed with respect to a secondary gold peak for modified and unmodified
DNA samples compared to clean and blank gold samples.
FIG. 5 shows volume density ratios of the elements with respect to Au(4f)
that comprises the S-DNA plotted versus take-off angle. The upper panel
shows N(1s)/Au(4f), ; O(1s)/Au(4f), ; C(1s)/Au(4f), ; and the lower
panel shows P(2p)/Au(4f), ; and S(2p)/Au(4f), .
FIG. 6 shows elemental depth profiles for S-DNA on Au(111). The upper panel
shows C, ; O, ; and Au, ; and the lower panel shows P, ; S, ; and N,
+.
FIG. 7 shows a comparison of elemental composition for the adlayer (bars)
and S-DNA theoretical mean values for the bulk DNA (line) as a function of
depth into the surface.
FIG. 8 shows a diagram drawn to scale of the model structure of S-DNA bound
to Au(111) projected onto the depth profiles of FIG. 7.
FIG. 9. shows a 1.2 .mu.m.times.1.2 .mu.m AFM image showing an isolated
strand of sulfur-modified DNA positive contrast on Au/mica.
DETAILED DESCRIPTION
Scanning Tunneling Microscopy
The scanning tunneling microscope used in this work was a coaxial
double-tube piezo design built at the University of Utah. A similar design
has been described in J. Lyding et al., 152 J. Microscopy 871-78 (1988),
and D. Zeglinski et al., 61 Rev. Sci. Inst. 8769-74 (1990), which are
hereby incorporated by reference. The microscope consists of an outer
piezoelectric tube responsible for sample x, y, z offset and macroscopic
initial sample approach, and an inner piezoelectric tube responsible for
the x, y, z scanning of the STM tip. The tips were made of a
platinum/rhodium (10% rhodium) alloy wire (0.51 mm diameter), and were
prepared by either ac etching or mechanical cutting. All tips were
subjected to quality control experiments. Only tips capable of obtaining
atomic resolution in preliminary experiments on a freshly cleaved HOPG
surface were used.
STM images were obtained at room temperature in air at varying tip scan
speeds. A tunneling current set point of less than 1 nA was used. A
differential input preamplifier, in which a "noise" input is
electronically subtracted from a "signal plus noise" input, was
incorporated into the setup. The "noise" input is connected to a wire
which makes a path that is nearly identical to the path made by the
"signal plus noise" wire leading to the tip. This "noise" wire terminates
near the STM tip without making electrical contact. This preamplifier
operates with a gain of 0.10 nA/V on a ten volt full scale with a unity
gain bandwidth of approximately 2 kHz and a noise level of approximately 2
pA peak-to-peak. The STM was controlled using feedback, scanning, and
offset electronics built at the University of Utah, and images were
acquired using an 80386/387 based 20 MHz AT compatible computer system
equipped with a 12-bit, 150 kHz analog-to-digital converter. Images
consist of 256.times.256 arrays of 12-bit data obtained in the constant
current imaging mode. Image figures were photographed from the computer
screen.
Auger Electron Spectroscopy
Auger electron spectroscopy (AES) analyses were performed in an ultrahigh
vacuum (UHV) surface science chamber designed and built at the University
of Utah. The standard surface analysis equipment was obtained from
Leybold-Heraeus as bolt-on components, which were attached to the custom
designed UHV chamber. The UHV chamber has a base pressure of less than
5.0.times.10.sup.-11 Torr which is achieved with a combination of ion,
titanium sublimation, and turbomolecular pumps. A sample transfer
interlock facilitates the transfer of samples in and out of ultrahigh
vacuum in approximately ten minutes with only a momentary rise of the base
pressure. The sample temperature can be varied from 77 K to 1500 K while
being monitored with a chromel/alumel thermocouple attached to the sample.
The instrumentation and data collections are controlled using data
acquisition and graphics computer programs.
The typical chamber pressure for this work was less than 1.times.10.sup.-9
Torr, and the plane of the sample was analyzed normal to the entrance axis
of the hemispherical kinetic energy analyzer with the electron gun
positioned at a 60 degree angle with respect to the surface normal and
working distances of approximately 25 mm. The AES spectra were obtained in
the derivative mode at a primary beam energy of 3000 eV and 5 V
peak-to-peak modulation of the pass energy.
Preparation of Biological Samples
A pentadecanucleotide (15-mer) with a sulfur group bonded to every
phosphorus atom in the phosphodiester backbone was obtained from Amersham
(Arlington Heights, Ill.). The generalized structure of any nucleotide of
this 15-mer, including the location of the sulfur group, is:
##STR1##
For comparative purposes, an octanucleotide (8-mer) containing one
thiolated guanine moiety was prepared according to the method of M.
Christopherson & A. Broom, 19 Nucl. Acids Res. 5719-24 (1991). The
structure of this thiolated guanine is:
##STR2##
Also, for comparative purposes, brominated poly(dA) was obtained
commercially. The structure of brominated adenylate is:
##STR3##
Oligonucleotides or polynucleotides having the desired modifications, in
which non-bridging internucleotide oxygen atoms are replaced with sulfur,
may be prepared in several ways. These methods would be obvious to one of
ordinary skill in the art. Illustrative of these methods is solution-phase
synthesis and solid-phase synthesis by automated DNA synthesizers.
Hydrogen-phosphonate, phosphorothioamidite, and phosphoramidite
chemistries may be selected. Detailed procedures for the phosphoramidite,
phosphorthioamidite, and hydrogen-phosphonate methods of oligonucleotide
synthesis are described in the following references, which are
incorporated by reference: Caruthers et al., U.S. Pat. Nos. 4,458,066 and
4,500,707; Koester et al., U.S. Pat. No. 4,725,677; Matteucci et al., 103
J. Amer. Chem. Soc. 3185-91 (1981); Caruthers et al., 4 Genetic
Engineering 1-17 (1981); Jones, chapter 2, and Atkinson et al., chapter 3,
in Gait, ed., Oligonucleotide Synthesis: A Practical Approach (IRL Press,
Washington, D.C., 1984); Froehler et al., 27 Tetrahedron Letters 469-72
(1986); Garegg et al., 27 Tetrahedron Letters 4051-54 and 4055-58 (1986);
Andrus et al., U.S. Pat. No. 4,816,571; Brill et al., 111 J. Amer. Chem.
Soc. 2321 (1989); and Froehler et al., 14 Nucleic Acids Res. 5399-5407
(1986).
Methods of producing sulfurized oligonucleotide analogs from products of
the synthetic schemes are described in the following references, which are
incorporated by reference: Stec et al., U.S. Pat. No. 5,151,510; Vu and
Hirshbein, 31 Tetrahedron Letters 3005-08 (1991); Marugg et al., 12
Nucleic Acids Res. 9095-9110 (1984); Andrus et al., 8 Nucleosides &
Nucleotides 967-68 (1989); Froehler, 27 Tetrahedron Letters 5575-78
(1986); and Stein et al., 188 Analytical Biochemistry 11-16 (1990).
Synthesis of sulfurized oligonucleotide analogs by enzymatic methods is
also contemplated. For example, the Klenow fragment of Escherichia coli
DNA polymerase I and the reverse transcriptase of avian myeloblastosis
virus will incorporate 5'-.alpha.-thio-triphosphates into DNA by
enzyme-catalyzed in vitro DNA synthesis. The following reference describes
such synthesis and is incorporated by reference: Atrazhev et al., 13
Bioorg. Khim. 1045-52 (1987).
Preparation of Gold Single Crystals with Low Miller Indices
Crystals were cut from 0.25 inch diameter gold rods obtained from Metal
Research, England. Rods of Au(111), Au(110), and Au(100), respectively,
were used. The crystals were oriented using a Laue camera. Then the
crystals were mechanically polished with successively smaller alumina or
diamond grit, ending with 0.05 micron polishing grit, according to
standard metal single crystal polishing procedures. Following mechanical
polishing, the crystals were electropolished according to the method
described in W. Peck & S. Nakahara, 11 Metallurgy 347-54 (1978), which is
hereby incorporated by reference. The crystals were then cleaned in the
UHV chamber by cycles of Ar.sup.+ sputtering (500 eV, 75 minutes) and
annealing (500 degrees C.) until no contamination could be detected by
AES. The Au(111) surface was then flame annealed with a Bunsen burner and
quenched in methanol. Caution must be exercised at this step because the
methanol usually ignites and burns with a nearly invisible flame. Properly
prepared gold films with the appropriate cleanliness and flatness can also
be employed.
Adsorption to Gold Crystals
The samples of oligo- and poly-nucleotides in aqueous solution were
deposited on gold crystals by three methods. In the preferred method, at
least 15 .mu.l of sample solution containing 3 to 10 .mu.g/mL of oligo- or
poly-nucleotide was pipetted onto a clean 5 mm diameter Au(111) single
crystal and left for at least 10 hours. To prevent evaporation of the
solution on the crystal during this extended time, the crystal was
enclosed in a covered Petri dish containing a small amount of water to
saturate the atmosphere. Following incubation, the gold crystal was washed
three times with distilled ("Nanopure") water and then the crystal was
allowed to air dry. The second method involved pipetting 15 .mu.l of
sample solution onto the surface of a 5 mm Au(111) single crystal and then
incubating for about 5 minutes. Then, the remaining liquid was removed by
blotting with filter paper and the crystal was washed three times in
distilled water and permitted to air dry. The third method consisted of
pipetting 15 .mu.l of sample solution onto a 5 mm diameter Au(111) single
crystal and permitting the liquid to evaporate to dryness in air.
Results of AES on Clean, Blank, and DNA Deposited Gold Crystals
Nuzzo and Allara, 105 J. Am. Chem. Soc. 4481-83 (1983), discovered that the
soft acid/soft base interactions of sulfur and gold result in monolayers
of thiols and disulfides on surfaces of gold films. The present invention
uses this sulfur/gold chemistry to covalently bond thiolated DNA to a gold
surface, such as the surface of a gold single crystal.
Referring to FIG. 1, there is shown the results of AES analysis on a clean
gold single crystal, in the upper trace, and a gold single crystal that
had been exposed to distilled ("Nanopure") water for 17 hours, in the
lower trace. The spectra are from the range of 50-550 eV and were signal
averaged 50 times. The lower spectrum was displaced by -5 on the y-axis
for clearer display. The crystal that was exposed to distilled
("Nanopure") water for 17 hours exhibits additional intensity of the C
(272 eV, KLL) peak and the O (504 eV, KLL) peak compared to the clean
crystal. No nitrogen, sulfur, phosphorus, or bromine peaks were observed.
The following AES signals could be used to measure DNA adsorption:
nitrogen (379 eV, KLL), sulfur (152 eV, LMM), phosphorus (120 eV, LMM),
and bromine (1396 eV, LLM). Carbon and oxygen signals which exceed the
levels found in the blank (gold single crystal exposed to distilled
("Nanopure") water for 17 hours) would also indicate DNA deposition.
FIG. 2 shows an AES spectrum obtained with the 15-mer containing a
thiolated phosphodiester backbone that was deposited on the gold single
crystal by interaction of the gold crystal and the sample solution for
more than 12 hours. The spectrum was signal averaged 50 times. The
spectrum contains a significant nitrogen (379 eV, KLL) peak and, thus,
illustrates a positive verification of nucleic acid deposition on the gold
crystal. Because of the overlap between the sulfur (152 eV, LMM) signal
and the gold (150 eV, NO0) signal and the complete absence of a bromine
(1396 eV, LMM) signal, the nitrogen (379 eV, KLL) signal was the most
straightforward for detecting adsorption of DNA to the gold crystal.
DNA adsorption to the gold crystal might be detected by means of the sulfur
(152 eV, LMM) signal if the spectral overlap of the sulfur (152 eV, LMM)
peak with a secondary gold (150 eV) peak could be accounted for. A method
to accomplish this goal was developed. A clean gold crystal in UHV
exhibits a gold (150 eV) peak intensity that is 3% of the gold (69 eV)
peak intensity. This contribution of the gold (150 eV) peak to the
combined gold (150 eV) +sulfur (152 eV) peak was removed according the
equation:
I.sub.152eV.sup.S .varies.I.sub.152eV.sup.S+Au
-(0.03.multidot.I.sub.69eV.sup.Au)
The resulting quantity was expressed as a ratio against the gold (69 eV)
peak intensity to yield the results shown in FIG. 3. These results
illustrate that the DNA containing a thiolated backbone adsorbed to a
greater extent than the other forms of DNA tested: the 8-mer containing a
thiolated guanine, brominated poly(dA), and an unmodified 9 kilobasepair
plasmid double-stranded DNA.
It is known that the intensity of the gold (69 eV, NOO) peak is sensitive
to sample position, thus the secondary gold (240 eV) peak has been used as
a measure of the gold signal, D. Jaffey & R. Madix, 258 Surf. Sci. 359
(1991). However, carbon also exhibits an AES peak in the 240 eV range. The
contribution of carbon to the AES spectrum obtained when DNA is bound to a
gold single crystal can be removed in a manner similar to that just
described for removing the contribution of the gold (150 eV) peak. The
carbon signal at 240 eV in the absence of gold is 8.1% of the primary
carbon signal at 272 eV. This contribution can be removed from the
observed intensity at 240 eV, leaving the intensity due to gold, according
to the following equation:
I.sub.240eV.sup.Au .varies.I.sub.240eV.sup.Au+C
-(0.081.multidot.I.sub.272eV.sup.C)
The results of this correction, shown in FIG. 4, are analogous to those of
FIG. 3 except that peak intensity ratios were calculated relative to the
gold signal at 240 eV as corrected.
The AES spectrum (not shown) obtained from analysis of deposition of the
brominated poly(dA) exhibited a carbon signal significantly above the
background level, however the intensities of phosphorus, nitrogen, and
oxygen peaks remained low. Thus, binding of brominated DNA appears less
extensive than that of DNA thiolated in the phosphodiester backbone. The
AES spectrum (not shown) obtained from analysis of the 8-mer containing a
thiolated guanine exhibited a sulfur signal slightly above background
levels, but the carbon, nitrogen, and oxygen signals were
indistinguishable from background. This result suggests that little, if
any, DNA containing a single thiol group adsorbed to the gold substrate.
The AES spectrum obtained with an unmodified 9 kilobasepair
double-stranded DNA suggested no binding at all to the gold single
crystal. These results indicate that, among the modifications studied,
backbone sulfur modifications are the key to successful binding of DNA to
Au(111).
The adsorption of DNA to gold single crystals by placing the DNA solution
on the crystal and withdrawing the liquid with a filter paper was also
examined with AES. Occasionally a small nitrogen peak could be detected
when the 15-mer with thiolated backbone and brominated poly(dA) were
tested. However, peaks indicative of DNA adsorption were never detected
with the 8-mer containing thiolated guanine. Thus, an extended period of
time is required for the reaction between the thiol group and the Au(111)
surface.
The adsorption of DNA to the gold surface by completely evaporating the
liquid from the sample was also examined by AES. The 8-mer containing
thiolated guanine was the only DNA used in this example. Characteristic
DNA peaks were readily discernible (not shown). However, calculations
based on the concentration of the sample, size of the DNA molecule, size
of the crystal, and ideal packing conditions indicate that the DNA was
deposited in a film about 100 monolayers thick on the average. The nature
of this sample is not suitable for STM and AFM imaging because most of the
molecules are not directly attached to the surface, thus motion and
conductivity problems are expected.
Angle-Dependent X-ray Photoelectron Spectroscopy of Sulfur-Modified DNA on
Au(111)
Angle-dependent x-ray photoelectron spectroscopy (ADXPS) was used to
investigate the role of sulfur in the binding of a 7250-base
backbone-sulfur-modified DNA on a Au(111) film on mica. Even though the
presence of sulfur facilitated adsorption of DNA, as shown above, the
issue of whether the sulfur was directly bound to the Au(111) surface was
not answered.
The technique of ADXPS is well-suited for non-destructive quantitative
depth profiling of adsorbates on surfaces. Nefedov, 17 Surf. Interface
Anal. 825 (1991); Baschenko et al., 53 Electron Spectrosc. 1 (1990);
Andrade, 1 Surface and Interfacial Aspects of Biomedical Polymers 140 (J.
Andrade ed., 1985); Paynter et al., 2 Surface and Interfacial Aspects of
Biomedical Polymers 189 (J. Andrade ed., 1985); Tillman et al., 6 Langmuir
1512 (1990); Bain et al., 111 J. Am. Chem. Soc. 321 (1989); Holloway et
al., 18 Surf. Interface Anal. 251 (1992). The technique relies on the
known attenuation of photoelectron intensities as a function of depth when
emitted from within a sample. By changing the take-off angle of the sample
with respect to the detector, the relative depths of the elements of the
adsorbate can be obtained. ADXPS has been used for relative depth
profiling of organic molecules and biomolecules adsorbed onto surfaces.
Andrade, 1 Surface and Interfacial Aspects of Biomedical Polymers 140 (J.
Andrade ed., 1985); Paynter et al., 2 Surface and Interfacial Aspects of
Biomedical Polymers 189 (J. Andrade ed., 1985). ADXPS has been applied to
self-assembled monolayers of alkanethiols adsorbed onto gold. Tillman et
al., 6 Langmuir 1512 (1990). The atomic composition of a monolayer of
HS--(CH.sub.2).sub.10 --CO.sub.2 CH.sub.3 on gold was derived from ADXPS
and the results suggested the monolayer structure was comprised of ester
groups at the surface of the layer and sulfur on the substrate. Bain et
al., 111 J. Am. Chem. Soc. 321 (1989). The results should be regarded as
semi-quantitative, since many assumptions must be made about the adlayer
structure and the nature of the photoelectron ejected from a solid,
indicating primarily the relative positions of the elements of the
adsorbate deposited on a surface.
The ADXPS technique has been extended to the reconstruction of quantitative
depth profiles of multielement surfaces. Nefedov, 17 Surf. Interface Anal.
825 (1991); Baschenko et al, 53 Electron; Spectrosc. 1 (1990); Holloway et
al., 18 Surf. Interface Anal. 251 (1992); Tyler et al., 14 Surf. Interface
Anal. 443 (1989). These methods are variations of tomographic
reconstruction | | |
