U.S. patent application number 16/168081 was filed with the patent office on 2019-06-27 for systems, apparatuses and methods for reading an amino acid sequence.
The applicant listed for this patent is ARIZONA BOARD OF REGENTS ON BEHALF OF ARIZONA STATE UNIVERSITY. Invention is credited to Stuart LINDSAY, Peiming ZHANG, Yanan ZHAO.
Application Number | 20190195884 16/168081 |
Document ID | / |
Family ID | 48905827 |
Filed Date | 2019-06-27 |
View All Diagrams
United States Patent
Application |
20190195884 |
Kind Code |
A1 |
LINDSAY; Stuart ; et
al. |
June 27, 2019 |
SYSTEMS, APPARATUSES AND METHODS FOR READING AN AMINO ACID
SEQUENCE
Abstract
Embodiments of the present disclosure relate to amino acid,
modified amino acid, peptide and protein identification and
sequencing, by means of, for example, electronic detection of
individual amino acids or small peptides.
Inventors: |
LINDSAY; Stuart; (Phoenix,
AZ) ; ZHANG; Peiming; (Gilbert, AZ) ; ZHAO;
Yanan; (Tempe, AZ) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
ARIZONA BOARD OF REGENTS ON BEHALF OF ARIZONA STATE
UNIVERSITY |
SCOTTSDALE |
AZ |
US |
|
|
Family ID: |
48905827 |
Appl. No.: |
16/168081 |
Filed: |
October 23, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14376154 |
Aug 1, 2014 |
10139417 |
|
|
PCT/US2013/024130 |
Jan 31, 2013 |
|
|
|
16168081 |
|
|
|
|
61593552 |
Feb 1, 2012 |
|
|
|
61647847 |
May 16, 2012 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01N 33/48721 20130101;
G01N 33/6821 20130101 |
International
Class: |
G01N 33/68 20060101
G01N033/68; G01N 33/487 20060101 G01N033/487 |
Goverment Interests
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH &
DEVELOPMENT
[0002] This invention was made with government support under grant
RO1HG006323 awarded by the National Institutes of Health. The
government has certain rights in the invention.
Claims
1. An apparatus for sequencing and/or identifying at least one of
protein, peptide, amino acid and/or modified amino acid, the
apparatus comprising: a plurality of electrodes; at least a first
electrode of the plurality of the electrodes being functionalized
with a molecule strongly bonded to at least the first electrode and
that forms transient bonds with an amino acid; voltage bias means
for applying a voltage between the first electrode and a second
electrode of the plurality of electrodes, wherein the first and
second electrodes are arranged such that a gap is produced between
them; and current data monitoring means for monitoring current
passing between the first and second electrodes upon a solution
containing at least one of protein, peptide, amino acid, and/or
modified amino acid being passed in the gap established between the
first and the second electrodes.
2. The apparatus according to claim 1, further comprising means for
recording and/or storing data corresponding to the monitored
current.
3. The apparatus according to claim 1, further comprising a
database and processing means, wherein the database includes data
associated with characteristic current signatures for a plurality
of proteins, peptides, amino acids, and/or modified amino acids,
and/or derivatives thereof, and wherein the processing means
compares such stored signatures with the collected current data to
determine the identity of the protein, peptide, amino acid and/or
modified amino acid.
4. The apparatus according to claim 1, wherein the chirality of an
amino acid is determined from the current data.
5. The apparatus according to claim 1, further comprising one or
more proteases for digesting a peptide, the proteases being
arranged in proximity to the first and second electrodes.
6. The apparatus according to claim 5, wherein the proteases are
provided on a bead.
7. The apparatus according to claim 1, further comprising one
carboxypeptidase for digesting a peptide or protein from its
C-terminus, the peptidase being arranged in proximity to the first
and second electrodes.
8. The apparatus according to claim 1, further comprising one
aminopeptidase for digesting a peptide or protein from its
N-terminus, the peptidase being arranged in proximity to the first
and second electrodes.
9. The apparatus according to claim 1, further comprising one or
more proteins, peptides, amino acids and/or modified amino acids on
a bead to be sequenced using Edman chemistry.
10. The apparatus according to claim 1, further comprising one or
more proteases for digesting a protein and/or peptide, the
proteases being arranged in proximity to the first and second
electrodes.
11. The apparatus according to claim 1, wherein the distance
between electrodes is adjustable.
12. The apparatus according to claim 10, wherein the proteases are
provided on a bead.
13. A system for sequencing at least one of protein, peptide,
and/or amino acids comprising the apparatus according to claim 1,
and further comprising: one or more proteases arranged in proximity
to at least one of the electrodes for digesting a protein and/or
peptide, and injection means for injecting at least one reagent
into the solution, wherein the activity of the proteases are
synchronized via an injection of a reagent by the injection means
that activates said proteases.
14. (canceled)
15. An apparatus for capturing and reading negatively charged amino
acids and/or peptide fragments, the apparatus comprising: a cis
chamber; a trans chamber; a first apparatus comprising the
apparatus of claim 1; a second apparatus comprising the apparatus
of claim 1, wherein the second apparatus is arranged substantially
orthogonal to the first apparatus, and wherein the second apparatus
separates the cis chamber from the trans chamber; a plurality of
reference electrodes arranged among the chambers, wherein one of
the reference electrodes is provided in the trans chamber and is
positively biased with respect to the cis chamber, and wherein the
gap provided by the second apparatus acts to capture negatively
charged amino acid and/or peptide residues.
16-25. (canceled)
Description
RELATED APPLICATIONS
[0001] This is a divisional application of U.S. patent application
Ser. No. 14/376,154, filed Aug. 1, 2014, which is the U.S. national
stage of International Application No. PCT/US2013/024130, which was
filed Jan. 31, 2013, which claims benefit under 35 USC 119(e) of
U.S. provisional patent application nos. 61/593,552, filed Feb. 1,
2012, and entitled, "Apparatus And Method For Reading Amino Acid
Sequence," and 61/647,847, filed May 16, 2012 of the same title.
Each of these disclosures is herein incorporated by reference in
its entirety.
FIELD OF THE DISCLOSURE
[0003] Embodiments of the present disclosure relate to amino acid,
peptide and protein, sequencing and identification, by means of,
for example, electronic detection of individual amino acids or
small peptides.
BACKGROUND OF THE DISCLOSURE
[0004] The human proteome, which is encoded by just some 25,000
genes, consists of millions of proteins variants, due to RNA
splicing, post translational modifications (PTM), somatic DNA
rearrangements, and single nucleotide polymorphisms (SNP). Despite
the widespread interest in whole genome sequencing, the genome is
largely fixed. While sequencing of mRNA is more informative, giving
an average picture of the levels of gene expression, it yields
little of post-translational modifications that generate the
millions of proteins from just thousands of genes. There is a need
of a massively parallel method to read human proteomes with high
throughput and low cost. In contrast to the human genome in which
DNA exists as diploid, the proteome has a wide dynamic range, for
example the abundance of proteins in human plasma spans more than
10 orders of magnitude. Some proteins are expressed in a low
quantity. Proteomics has no tool equivalent to the polymerase chain
reaction (PCR) for protein sample amplification. There is no cost
effective way to faithfully reproduce a protein population from a
source. Thus, protein analysis must be carried out by extracting
materials from samples removed from humans in some quantity.
[0005] There have been remarkable advances in sample preparation,
and sequencing techniques, most notably based on mass spectroscopy
as a proteomic tool. However, mass-spectrometers are large, costly
machines. Their size is dictated by the need for very high mass
resolution to obtain accurate identification of the amino acid
components (and even then, readout is complicated by isobaric amino
acids). Accordingly, there is a need for an alternative method for
identifying amino acids, particularly in small quantities, and
ideally at the single-molecule level.
[0006] In a series of earlier disclosures, WO2009/117522A2, WO
2010/042514A1, WO 2009/117517, WO2008/124706A2, US2010/0084276A1,
and US2012/0288948, each of which is incorporated herein by
reference, a system was disclosed where nucleic acid bases could be
read by using the electron tunneling current signals generated as
the nucleobases pass through a tunnel gap functionalized with
adaptor molecules. A demonstration of the ability of this system to
read individual bases embedded in a polymer was given by Huang et
al..sup.1 This method is referred to as "Recognition Tunneling"
(RT)..sup.2 It was earlier recognized in these previous disclosures
that, because the method is purely physical, in that it did not
rely on the reactions of a DNA polymerase or ligase, any chemical
residue should be able to be recognized provided that it generates
a distinctive tunneling current signal.
[0007] Yet another problem in determining the human proteome arises
because of the stereochemistry of amino acids. Amino acids can
exist in two forms ("enantiomers") that are mirror images of each
other. Nature has chosen the so called "L" (for left-handed) form
in general, but the presence of "D" (for dextro or right handed)
amino acids is an important biomarker for diseases such as ALS or
schizophrenia. Since these enantiomers are isobaric, they cannot be
sensed by mass spectroscopy, so large amounts of additional sample
are needed for optical identification. Furthermore, control of
optical isomers is one of the most difficult problems in chemical
synthesis and separation. Since the isomers are chemically
identical, it is very hard to extract pure samples of one isomer or
the other. Thus, a method to read the relative concentrations of
isomers from very small amounts of sample would represent a major
advance. Present, optical techniques require large amounts of
sample. In addition, it would be advantageous to provide a simple
method for reading the identity of optical isomers at the single
molecule level.
SUMMARY OF THE DISCLOSURE
[0008] Accordingly, it is an object of some of the embodiments of
the present disclosure to provide a method for sequencing and/or
identifying amino acid molecules. It is another objective of some
of the embodiments of the present disclosure to provide a method to
read sequences of proteins and peptides. It is yet another
objective of some embodiments to detect enantiomers at the single
molecule level. Accordingly, it is an object of some of the
embodiments of the present disclosure to provide a method for
sequencing and/or identifying at least one protein, peptide, amino
acid and/or modified amino acid.
[0009] These and other objectives are achieved by at least some of
the embodiments of the present disclosure. In some embodiments, an
apparatus is provided which includes a plurality of electrodes
which are closely spaced, where at least one electrode (and, in
some embodiments preferably more than one electrode) is
functionalized with molecules that bind transiently to the analyte
target. In some embodiments, the apparatus includes two such
electrodes where each is functionalized with such molecules. In
some embodiments, at least two of the electrodes enclose a space
into which an analyte can be drawn, by diffusion, electrophoresis,
dielectrophoresis and/or electro-osmosis (for example).
Accordingly, in some embodiments, on binding an analyte to the
recognition molecules, a series of current spikes signal the
identity of the molecule upon a bias being applied between the
electrodes. In some embodiments, the small analyte molecules are
generated from a parent protein by placing the electrode(s) in
close proximity to a bead containing proteases that digest the
protein to be analyzed.
[0010] According to some of the embodiments of the present
disclosure, a computer system is additionally included, as well as
associated software for operating such a computer and operating
embodiments for identifying and, in some embodiments,
sequencing/identifying proteins anal/or peptides, analyzing amino
acids, as detailed herein.
[0011] According to some embodiments, an apparatus for identifying
anal/or sequencing at least one protein, peptide, amino acid and/or
modified amino acid is provided, where the apparatus includes a
plurality of electrodes, at least a first electrode of the
plurality of the electrodes being functionalized with a molecular
adaptor that is strongly bonded to at least the first electrode and
that forms non-covalent bonds with an amino acid, voltage means
(e.g., a voltage biasing device, which may comprise a power supply)
for applying a voltage between the first electrode and a second
electrode of the plurality of electrodes, where the first and
second electrodes are arranged such that a gap is produced between
them, and current data monitoring means (e.g., a current measuring
device) for monitoring current passing between the first and second
electrodes upon a solution containing at least one protein,
peptide, amino acid and/or modified amino acid being passed in the
gap established between the first and the second electrodes. In
some embodiments, a database and processing means (e.g., a computer
processor, which may also include a memory) may also be included
where the database includes data associated with characteristic
current signatures for a plurality of amino acids and amino acid
derivatives, where the processing means has computer instructions
operating thereon configured to compare the amino acid and the
derivative signatures (which are contained in the database, for
example) with the current data to determine the identity of the
amino acid, peptide and/or protein. Moreover, the chirality of the
amino acid may be determined from the current data.
[0012] In some embodiments, an assembly used in an apparatus for
sequencing and/or identifying at least one protein, peptide, amino
acid, and/or modified amino acid, may include a silicon substrate
covered with a thin layer of dielectric between about 1 nm and
about 100 nm in thickness, a window etched through the silicon
substrate, a metal electrode layer deposited onto the dielectric
layer which may comprise at least one of silver, gold, palladium or
platinum between about 1 nm and about 10 nm in thickness, an
adhesive layer of at least one of Ti or Cr of about 1 nm, a second
layer of dielectric deposited onto the metal electrode layer
comprising at least one of silicon nitride, silicon dioxide or
hafnium oxide, the layer having a thickness of between about 1 nm
and about 4 nm in thickness, a second metal electrode deposited on
top of the second dielectric layer comprising at least one of
silver, gold, palladium or platinum between about 1 nm and about 10
nm in thickness, a third layer of dielectric deposited on top of
the apparatus and is formed to contact each pair of electrode
planes at each apparatus, and a nanopore of between about 1 and 5
nm in diameter provided through the entire assembly.
[0013] In some embodiments, a method for sequencing/identifying at
least one protein, peptide, amino acid, and/or modified amino acid
is provided and may include one or more of the following steps:
flowing a fluid containing at least one protein, peptide, amino
acid, and/or modified amino acid in a gap provided between two
closely spaced electrodes, where: at least one of the electrodes is
functionalized with molecules that form transient, non-covalent,
bonds with a target amino acid, and the target molecules diffuse
freely into the gap from solution, and/or can be driven into the
gap upon the gap spanning a nanopore through which fluid flows.
Fluid flow may be accomplished via at least one of a pressure
gradient, electroosmosis, dielectrophoresis and electrophoresis.
Signals may be generated as the fluid flows through the gap which
is representative of the at least one of protein, peptide and amino
acid contained in the fluid. Additional steps may include
determining the at least one of protein, peptide and amino acid
based on the signals.
[0014] Determining the at least one of protein, peptide and amino
acid may comprise determining at least one of: a width of signal
spikes, a ratio of "on" to "off" time in a signal burst, a duration
of a signal burst, a degree of clustering of the peaks, a flatness
of the top of the spike of a signal, a frequency of the signal
spikes, a rise and/or fall time of the signal spikes, and a shape
of the signal spikes.
[0015] Some embodiments may include a computer system for
sequencing/identifying at least one protein, peptide, amino acid,
and/or modified amino acid which comprises a processor, where the
processor includes computer instructions operating thereon for
performing steps of a method for identifying and/or sequencing at
least one protein, peptide, amino acid and/or modified according to
any of the disclosed embodiments. Other related embodiments include
a computer program for identifying and/or sequencing at least one
protein, peptide, amino acid, and/or modified amino acid comprising
computer instructions for performing such steps, as well as
computer readable media (e.g., CD, DVD, flashdrive, and the like)
containing such a program operable on a computer system.
[0016] In some embodiments, a molecular recognition apparatus is
provided which may be configured to identify at least one portion
of a target molecule. The apparatus may include a microfluidic
channel containing first and second sensing electrodes, where the
electrodes are separated from one another by a gap for receiving a
target molecule, a first affinity element connected to the first
electrode, and a second affinity element connected to the second
electrode. When a portion of the target molecule is adjacent to the
gap and completes a circuit between the electrodes, the device is
configured to enable an electrical current to pass through the
first electrode, the first affinity element, the portion of the
target molecule, the second affinity element, and the second
electrode.
[0017] The above-noted embodiments, and embodiments disclosed
throughout this application, are merely examples of the many
embodiments disclosed herein. Such embodiments further include
embodiments which are similar to those set out above/below, but
which may include less than the recited features, and others still,
more than the recited features, thereof. Moreover, other
embodiments include a combination of features of the embodiments
exemplified above, below and throughout the application.
[0018] Some embodiments of the disclosure are further described in
the following detailed description and included drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0019] FIG. 1 illustrates a tunnel gap apparatus for detecting
anayltes that bind transiently to recognition molecules, generating
current spikes (I) when a bias (V) is applied across the gap,
according to some embodiments of the present disclosure.
[0020] FIG. 2 illustrates exemplary signals from glycine detected
using gold electrodes produced by some embodiments of the present
disclosure.
[0021] FIG. 3 illustrates exemplary signals from L-asparagine
detected using gold electrodes produced by some embodiments of the
present disclosure.
[0022] FIG. 4 illustrates exemplary signals from phenylalanine
detected using gold electrodes produced by some embodiments of the
present disclosure.
[0023] FIG. 5 illustrates distributions of pulse heights from
gylcine (red) asparagine (blue) and phenylalanine (green) fitted
with log-normal functions produced by some embodiments of the
present disclosure.
[0024] FIG. 6 illustrates peak currents (pA) and count rates from
glycine, asparagine and phenylalanine at a gap conductance of 8 pS
(I=4 pA, V=0.5V) produced by some embodiments of the present
disclosure.
[0025] FIG. 7 illustrates optical isomers of asparagine.
[0026] FIG. 8 illustrates distribution of peak heights from aqueous
buffer alone using gold electrodes at a bias of 0.8V and a tunnel
current of 6 pA, corresponding to a gap conductance of 7.5 pS,
produced by some embodiments of the present disclosure..sup.FN1
*Few of the background peaks are present at peak heights above 20
pA; data is accumulated over about 1 h (or about 1 count per
minute).
[0027] FIG. 9 illustrates signals from D-asparagine using gold
electrodes at a bias of 0.8V and a tunnel current of 6 pA over a
timescale of about 100 ms produced by some embodiments of the
present disclosure. .sup.FN2 *Frequent spikes are observed, most of
which greatly exceed the water background of 20 pA and occur at
much higher frequency.
[0028] FIG. 10 illustrates a distribution of pulse heights from D
asparagine using gold electrodes at a bias of 0.8V and a tunnel
current of 6 pA produced by some embodiments of the present
disclosure..sup.FN3 *The peak of a log-normal distribution fit is
33.5 pA.
[0029] FIG. 11 illustrates typical signals from L-asparagine using
gold electrodes at a bias of 0.8V and a tunnel current of 6 pA over
a timescale of about 2 s produced by some embodiments of the
present disclosure.
[0030] FIG. 12 illustrates a distribution of pulse heights from
L-asparagine using gold electrodes at a bias of 0.8V and a tunnel
current of 6 pA, with the peak of a log-normal distribution fit is
58 pA, produced by some embodiments of the present disclosure.
[0031] FIG. 13 illustrates a distribution of pulse heights from an
equimolar mixture of D- and L-asparagine using gold electrodes at a
bias of 0.8V and a tunnel current of 6 pA, produced by some
embodiments of the present disclosure, where the distribution is
fitted by two log-normal functions with peaks at 34 and 60 pA.
[0032] FIG. 14 illustrates an apparatus for protein, peptide, amino
acid, and/or modified amino acid identification and/or sequencing
according to some embodiments of the present disclosure.
[0033] FIG. 15 illustrates the construction of nanopore apparatus
articulated with electrodes according to some embodiments of the
present disclosure.
[0034] FIG. 16 illustrates an apparatus with a second nanopore and
trans chamber biased for collecting negatively charged amino acids
and peptide fragments according to some embodiments of the present
disclosure.
[0035] FIG. 17 is an illustration of peptide sequencing by directly
tunneling reading of amino acid residues, according to some
embodiments of the present disclosure.
[0036] FIG. 18 is an illustration of peptide sequencing by
enzymatically cleaving amino acids from a terminal of protein,
according to some embodiments of the present disclosure.
[0037] FIGS. 19A-C illustrate characteristic signal traces for 18
of the 20 amino acids together with two examples of control data
(upper left panels in 19A and 19C) obtained with buffered
electrolyte alone, with an apparatus according to some embodiments,
the data being obtained using palladium electrodes and free of
background signals produced by gold electrodes.
[0038] FIG. 20A illustrates a protein sequence reader apparatus
without a nanopore according to some embodiments of the present
disclosure.
[0039] FIG. 20B illustrates a measurement cell apparatus having a
nanoliter volume according to some embodiments of the present
disclosure.
[0040] FIG. 20C illustrates a view of a nanoliter scale reaction
system according to some embodiments of the present disclosure.
[0041] FIG. 20D illustrates a view down into the reaction chamber
703 of FIG. 20C.
[0042] FIG. 20E illustrates a view of the overall assembly of the
nanoliter scale reaction system according to some embodiments of
the present disclosure.
[0043] FIG. 21A illustrates a schematic view of a tunnel gap
apparatus, according to some embodiments, created by cutting a
palladium wire on a silicon nitride substrate with an electron
beam, and the overall assembly of the nanoliter scale reaction
system according to some embodiments of the present disclosure.
[0044] FIG. 21B is a TEM image of a tunnel gap, according to some
embodiments, prior to functionalizing with imidazole carboxamide
recognition molecules.
[0045] FIG. 21C is a graph illustrating the current signals
produced with a phosphate buffer (pH7) using a sequencing device
according to some embodiments of the present disclosure.
[0046] FIG. 21D is a graph illustrating the current signals
produced with the sequencing device used in the signals generated
in FIG. 21C, but after adding 100 micromolar glycine.
[0047] FIG. 22A is a graph illustrating current signals recorded of
L-leucine in a sequencing system according to some embodiments of
the present disclosure.
[0048] FIG. 22B is a graph illustrating current signals recorded of
L-isolleucine in a sequencing system according to some embodiments
of the present disclosure.
[0049] FIG. 22C is an image depicting the normalized peak width vs.
normalized amplitudes of the signals from FIG. 22A.
[0050] FIG. 22D is an image depicting the normalized peak width vs.
normalized amplitudes of the signals from FIG. 22B.
[0051] FIG. 22E shows typical spectra from a series of multiple
repeated measurements on seven amino acids (as listed on the
figure) according to some embodiments of the present disclosure
(panels a-i). The insets show features at higher resolution
(current scale 150 pA. time scale 20 ms). One of the analytes is
sarcosine (methyl glycine, panel e) an example of a modified amino
acid. Leakage current is controlled so that the tunnel gap is
substantially is the same on repeated runs, and a covariance
analysis is carried out to determine which signal parameters are
consistent across the different experiments. These parameters were
then used as input for support vector machine analysis.
[0052] FIG. 22F is a Table summarizing true positive rate for
identifying an amino acid based on a single peak in a data train
after an analysis using the support vector machine, according to
some embodiments of the present disclosure. The data is for a mixed
pool of data for all seven amino acids listed.
[0053] FIG. 23A is a graph of predicted accuracy (i.e., fraction of
wrong calls) of a sequencing system according to some embodiments
of the present disclosure.
[0054] FIG. 23B shows the measured true positive rate obtained by
requiring that a particular number of spikes make the same call
sequentially, according to some embodiments of the present
disclosure.
[0055] FIG. 23C shows a signal from a peptide consisting of three
repeated amino acids (gky-gly-gly) showing that reads can be
obtained from amino acid polymers according to some embodiments of
the present disclosure.
[0056] FIG. 24 is a perspective, schematic view of a sequencing
system according to some embodiments, illustrating pressure
injection (via, e.g., a pump, syringe, and the like) into the
tunnel gap.
[0057] FIG. 25 is a graph of an HPLC chromatogram for 25 amino
acids without derivatization.
[0058] FIG. 26 is a perspective view of a micro-reactor coupled to
an RT system through a microfluidics channel for determining
sequences of peptides, according to some embodiments of the present
disclosure.
[0059] FIG. 27 illustrates the sequencing a peptide by RT in
combination with enzymatic hydrolysis according to some embodiments
of the present disclosure.
[0060] FIG. 28 is a block diagram of a sequencing system according
to some embodiments of the present disclosure.
[0061] FIG. 29 is a block diagram of a sequencing system according
to some embodiments of the present disclosure.
DETAILED DESCRIPTION
[0062] In some embodiments of the present disclosure, RT relies on
passing an analyte between two closely spaced electrodes, each of
which is functionalized with molecules that form transient,
non-covalent, bonds with the target analyte. The target molecules
may diffuse freely into the gap from solution, and/or they can be
driven into the gap if the gap spans a nanopore through which fluid
flows. Fluid can be made to flow through the gap by means of, for
example, at least one of a pressure gradient electrophoresis,
dielectrophoresis and electroosmosis (e.g., if the walls of the
nanopore carry a charge). If the analyte molecules are charged,
they can also be driven through the gap by electrophoresis.
[0063] FIG. 1 illustrates an arrangement used for collecting
tunneling signals according to some embodiments. As illustrated,
two opposed electrodes 1, 2 (i.e., a plurality) are separated by a
gap 3 of about 2.5 nm, but which could be as small as about 1 nm
and as large as about 4 nm (still other embodiments include a gap
between about less than 1 nm and larger than 4 nm, for example).
Each electrode may be functionalized with a reagent 4 that is
chemically-bonded to the electrodes, to form non-covalent bonds
with the target molecule. Shown as 4 in FIG. 1 are two
4(5)-(2-mercaptoethyl)-1H imideazole-2-carboxamide molecules where
the SH group has lost the H in order to from a bond with the metal
electrode. Such molecules may be synthesized as adaptor molecules
for reading, for example, DNA bases..sup.3 Suitable metals for
electrodes include platinum, palladium, gold and silver, all of
which have been demonstrated theoretically to give excellent
coupling with the electronic states of molecules bonded to them in
a tunnel junction..sup.4 In some embodiments, the entire system is
immersed in an aqueous electrolyte 5.
[0064] For the data presented here, for some embodiments, the
buffer is a 1 mM phosphate buffer, having a pH=7.0. Provided that
only (sub-micron).sup.2 areas of one of the electrodes are exposed
to electrolyte, electrochemical leakage currents in some
embodiments are considerably smaller than tunnel current between
the electrodes. The gap 3 may be defined by the current I and the
voltage V (electrical bias being the voltage, being supplied by a
power supply device, and structure I can be a monitoring device for
monitoring current). For a voltage of 0.5V applied between the
electrodes, a current of 6 pA is indicative of a gap of about 2.5
nm.sup.5 (for example). A bias of 0.5V and a set-point current of 4
pA were used to collect the data shown in FIGS. 2-5. In these
conditions, the tunneling signal in aqueous buffered electrolyte
solution (i.e., absence of any amino acids) is free of features,
having a noise amplitude of about 3 pA. If the gap is made a
smaller (e.g., by increasing the current to 6 pA for example) noise
spikes of up to about 20 pA appear in the presence of aqueous
buffer alone, but these are rare (e.g., a few counts per hour).
[0065] In some embodiments, when amino acids are added to the
solution ("aa" in FIG. 1), relatively large signals immediately
appear, in this case because the analyte molecules drift into the
tunnel gap by diffusion. Referring to FIG. 1, data for glycine, 6,
asparagine, 7 and phenylalanine, 8 will be described. These amino
acids were added at a concentration of 100 micromolar, but smaller
concentrations (down to 1 nM for example, in some embodiments) can
be used. Each amino acid gives a different characteristic tunneling
signal. FIG. 2 illustrates a signal from glycine, produced
according to some embodiments of the present disclosure. In this
example, the spikes have an average peak height of about 30 pA and
a frequency of about 600/hr at this concentration. FIG. 3
illustrates signals from asparagine. Here, the amplitude is higher
(about 40 pA on average) as is the frequency (.about.1000 counts
per hour), and the duration of the spikes is also longer. FIG. 4
illustrates data for phenylalanine, where the peaks are large
(.about.100 pA) but infrequent (.about.100 counts/hr).
[0066] FIG. 5 shows distributions of spike intensities for all
three amino acids plotted together and fitted with log-normal
distribution functions according to some embodiments. The peak
values of the fitted currents are listed in FIG. 6 together with
the count rate. In calculating these distributions, all peaks with
a width less than or equal to the current amplifier's response time
(0.15 ms) were discarded. These data establish that distinct
electronic signals are generated by different amino acids in a RT
junction.
[0067] While there is considerable overlap between the peak height
distributions from these three exemplary amino acids, other
parameters can be used to make assignments. Examples of such
parameters may include: [0068] Width of the spikes; [0069] Ratio of
"on" to "off" time in a signal burst; [0070] Duration of a burst of
signal; [0071] Degree of clustering of the peaks; [0072] Flatness
of the top of the spike; [0073] Frequency of the spikes; and [0074]
Rise and fall times of the spikes [0075] In some embodiments, a
combination of these parameters may be used.
[0076] In some embodiments, the shape of the spikes may be
characterized by the size of frequency components in a wavelet
analysis, and any number of components can be included in a
principle component analysis, for example, conveniently carried out
with Support Vector Machine (SVM) code. Accordingly, with adequate
training of a recognition code, an assignment of signals from all
20 of the amino acid residues may be obtained.
[0077] In some embodiments, the system may be sensitive to
chirality (i.e., optical isomerism) of individual molecules.
Specifically, in some embodiments, the amino acid is trapped by
bonding to the recognition molecules in the gap. Thus, the three
dimensional details of its non-covalent bonding to the recognition
molecules becomes important. To demonstrate this, exemplary data
for L-asparagine and D-asparagine is illustrated in FIGS. 9 to 13.
The stereo chemistry of the two isomers is displayed in FIG. 7. In
D-asparagine, the amine (NH.sub.2) group points into the plane of
the paper, while in L-asparagine, the amine (NH.sub.2) group points
out of the plane of the paper. In RT measurements, according to
some embodiments, the difference is that D-asparagine produces
virtually no signals in the tunneling conditions used to detect
L-asparagine (0.5V bias, 4 pA tunnel current, corresponding to a
gap conductance of 8 pS). Increasing the tunnel current to 6 pA
with the same bias (gap conductance=12 pS) also produced no reads
from D-aspargine. However, increasing the bias to 0.8V while
leaving the current at 6 pA produced copious reads from both D and
L apsaragine. In some embodiments, the gap conductance in these
conditions is the smallest (7.5 pA), so the ability of the gap to
trap a specific type of molecule is not only a function of its size
(which is proportional to the logarithm of conductance), but also a
function of the electric field in the gap (for example). In these
conditions, in some embodiments, the clean background in the
absence of analyte may be lost (small current peaks being observed
in the absence of an analyte). The distribution of the heights of
these background peaks is illustrated in FIG. 8, where the count
rate is about 1 peak per minute (e.g., low count rate). A typical
signal train produced from some embodiments, from D-asparagine (100
.mu.M D-asparagine in 1 mM phosphate buffer, pH 7.0) is illustrated
in FIG. 9, where the count rate is much higher (tens to hundreds of
counts per second). The corresponding distribution of peak heights
is illustrated in FIG. 10, where the peak value of this
distribution is 33.5 pA.
[0078] In some embodiments, L-asparagine produces larger signals in
these conditions. FIG. 11 illustrates a typical signal train when
100 .mu.M D-asparagine in 1 mM phosphate buffer (pH 7.0) is added
to the solution in which the tunnel junction is immersed. The
signals may now be less frequent than those from D-asparagine in
these tunneling conditions, but the peak heights are larger (for
example). The distribution of peak heights is illustrated in FIG.
12, where the peak of a log-normal fit is 58 pA.
[0079] The ability of the methods and systems according to some
embodiments of the present disclosure to distinguish between two
enantiomers at the single molecule level is illustrated in FIG. 13.
This figure shows the distribution of pulse heights from an
equimolar mix of D- and L-asparagine. The distribution shows two
well separated peak currents of 34 pA (D-asparagine) and 60 pA
(L-asparagine). On the basis of peak height alone, each peak may be
assigned (according to some embodiments). Thus, one can identify
one amino acid from another on the basis of its RT signal, and
identify enantiomers--and to do so at the single molecule
level.
[0080] A readout apparatus, according to some embodiments, which
enables a scheme for sequencing and/or identifying one or more
proteins and/or one or more peptides is illustrated in FIG. 14 (in
some embodiments, also including sequencing and/or identifying one
or more amino acid, and/or modified amino acid). As shown, a
silicon or silicon nitride membrane 130 spans a channel 117
containing an electrolyte solution 108, 109. The membrane,
typically about 10 to about 100 nm thickness, divides the channel
into two chambers, a cis chamber 108 and a trans chamber 109, in
fluid communication with each other only by means of a nanopore 102
that may be drilled through the membrane (and all the other layers
103, 104, 101, 120) by means of, for example, an electron beam.
Ideally, the nanopore may be between about 1 and about 5 nm in
diameter at the point where it passes through the electrodes. The
pore passes through planar electrodes 103, 104 separated by the
dielectric layer 101 (for example). The dielectric layer may be
between about 1 nm and about 4 nm thickness between the electrodes
and may be made by depositing silicon nitride, silicon dioxide or
hafnium oxide. The electrodes may be made by depositing between
about 1 nm to about 5 nm of silver, gold, palladium or platinum
onto the top of the first membrane 130 (for example). The
dielectric layer 101 may then be deposited, followed by deposition
of a second metallic layer, 104. This "top" electrode may then be
covered by a second layer of dielectric 120. Fabrication of the
assembly according to some embodiments is completed by drilling the
nanopore. One or more of the electrodes (in two-electrode
embodiments, e.g., both; in plurality of electrode embodiments, two
or more) may then be functionalized by immersing the assembly in a
solution of the recognition molecules 107. The molecules is linked
to a thiol group via a one or two carbon atom chain, and electrodes
may be functionalized using a 0.1 mM to 10 mM solution in ethanol
to which the nanopore is exposed overnight (for example).
[0081] Each chamber of the apparatus may be filled with an
electrolyte such as KCl or NaClO.sub.3 (for example), in
concentrations that range from about 1 mM to about 1M, and to which
may be added Mg ions and/or ATP as required to activate enzymes 113
attached to one or more beads 112 which are fixed in turn to the
walls of the channel 117 in close proximity to the nanopore 102.
Alternatively, the enzyme is directly attached to the channel in
close proximity to the nanopore. The enzymes may then be any one of
the well-known proteases which include carboxypeptidase,
aminopeptidase, trypsin, chymotrypsin, pepsin, papain or elastase
(for example). Since each of these proteases is somewhat selective
in their hydrolysis of peptide bonds, the beads may ideally contain
a mixture of these enzymes. In some embodiments, the bead may be
functionalized with proteosomes, assemblies that sequentially
degrade proteins into their component amino acids. The isolated
protein or peptide 114 to be identified/sequenced may also bound to
the bead. In some embodiments, digestion of the protein may be
initiated by binding the protein to the bead in the absence of Mg
or other chemicals needed to initiate digestion, the bead placed in
the channel, and then digestion is initiated by the addition of Mg
or other chemicals (needed to initiate digestion). The resulting
small fragments, small peptides, or optimally amino acids 115, may
be released into the solution. A bias 118 may be applied between
the cis and trans chambers by means of the reference electrodes
111, 110, for example. In the figure, this is shown with the
negative electrode in the trans chamber 109. Accordingly, this
draws positively charged amino acids and small peptides through the
nanopore where they bind transiently to the recognition molecules
107 generating current spikes recorded by a transconductance
amplifier 106. A bias 105 of between about 0.1V and about 1V is
preferably applied between the electrodes 103, 104. The walls of an
oxide layer (such as the surfaces of silicon, silicon nitride or
silicon dioxide or hafnium oxide) in aqueous electrolyte are
negatively charged 116 owing to the accumulation of OH groups on
the surface. This negative charge causes an electro-osmotic flow of
water towards the negative electrode 115. Thus, neutral amino acids
that diffuse into the vicinity of the nanopore may be swept through
it by the electro-osmotic flow of the water.
[0082] All the neutral and positively charged amino acid residues
can be read by the same nanopore apparatus, according to some
embodiments. Accordingly, the amino acid mixture will generate a
characteristic set of tunneling signals that will allow the protein
or peptide to be identified. To the extent that the digestion of
the target proteins is sequential and synchronized, the train of
signals could also be used to deduce protein sequence, while the
sequence of small peptides may be read directly from the time
series of signals generated as each amino acid in the chain passes
through the tunnel gap.
[0083] Fabrication of the apparatus according to some embodiments
is illustrated in more detail in FIG. 15. Accordingly, a silicon
substrate 201 is covered with a thin layer of dielectric 202, such
as low-stress silicon nitride. This layer may be between about 1 nm
and 100 nm in thickness. A window 208 may be etched through the
silicon support by an anisotropic etch. A metal electrode layer 203
may be deposited onto the dielectric layer 202. This layer may be
any one of silver, gold, palladium or platinum and be between about
1 nm and about 10 nm thick. A layer of Ti or Cr of about 1 nm may
be used as an adhesion layer for this metal electrode. A second
layer of dielectric 204 may then be deposited onto the metal
electrode 203. This dielectric may be silicon nitride, silicon
dioxide or hafnium oxide and be between about 1 nm and about 4 nm
in thickness. A second metal electrode 205 may then be deposited on
top of the dielectric layer 204. This layer may be of the same or
similar composition and thickness to that of layer 203. Finally, a
layer of dielectric 206 may be deposited on top of the apparatus
and is formed (not shown) to contact, for example, each pair of
electrode planes at each apparatus. 203, 205. A nanopore 207 of
between about 1 and about 5 nm in diameter may then be drilled
through the entire assembly. A perspective view of the apparatus is
illustrated on the right of FIG. 15, indicating where bias may be
applied and/or where current may be read. Each nanopore apparatus
in an array may have a separately addressable pair of electrodes,
and the area of opposed electrodes may be minimized in order to
reduce leakage through the dielectric. Typical opposed areas of
electrodes may be about 0.1 micron by about 0.1 micron square.
[0084] FIG. 16 illustrates an arrangement according to some
embodiments for capturing and reading the negatively charged amino
acids and peptide fragments. A second nanopore apparatus 301 may be
placed in close proximity to the first apparatus 305 and separates
the cis chamber 108 from a second trans chamber 302 isolated from
the first trans chamber 109. A third reference electrode 303 may be
placed in the second trans chamber 303 but may be biased positively
304 with respect to the cis chamber. Thus, this second nanopore
acts to capture the negatively charged residues 306.
[0085] FIG. 17 is an illustration of peptide sequencing by directly
tunneling reading of amino acid residues, according to some
embodiments of the present disclosure.
[0086] FIG. 18 is an illustration of peptide sequencing by
enzymatically cleaving amino acids from a terminal of protein,
according to some embodiments of the present disclosure.
[0087] FIGS. 19A-C illustrate characteristic signal traces for 18
of the 20 amino acids together with two examples of control data
(upper left panels in 19A and 19C) obtained with buffered
electrolyte alone. These data were obtained using palladium
electrodes. In this example, the baseline tunnel current was set to
4 pA with the probe biased at +0.5V with respect to the substrate.
100 micromolar aqueous solutions of the amino acids containing 1 mM
phosphate buffer (pH-7) were also used. FIG. 19A illustrates
recordings for the charged amino acids argenine (Arg), histidine
(His), Lysine (Lys), Aspartic acid (Asp) and glutamic acid (Glu).
FIG. 19B illustrates recordings for the hydrophobic amino acids,
alanine (Ala), Valine (Val), isoleucine (IIL), leucine (Leu),
methionine (Mfet) and phenylanlanine (Phe). FIG. 19C illustrates
recordings for serine (Ser), threonine (Thr), asparagine (Asn),
glutamine (Gln), cysteine (Cys), glycine (gly) and proline (pro).
The two examples of control data (top left panel in FIGS. 19A and
19C) are free of features with the exception of few sharp spikes
seen in a percent or so of the runs (an example is pointed to by an
arrow on the panel in FIG. 19C).
[0088] In this example, the two most aromatic amino acids, tyrosine
and tryptophan, appear not to give signals in these conditions.
Signals can be obtained by increasing the tunneling set point to 6
pA and 10 pA respectively, but the signals contain some spurious
background from water. This can be removed using a SVM
analysis..sup.7 Thus, in some embodiments, there may be an
advantage in having an adjustable tunnel gap (as described in an
additional embodiment detailed below).
[0089] An SVM.sup.7, 8 trained on 18 amino acid data sets
simultaneously, and 3000 peaks were sampled for each of 17 amino
acids. The exception was isoleucine, where only 800 peaks were
measured. 400 of the 3000 were used for training, and then the
remaining 2,600 peaks were called using the resulting support
vectors. The accuracy with which the peaks were called is listed in
Table 1 below. Tyrosine and tryptophan were excluded from the
analysis shown here, but also give distinctive signals.
[0090] Thus, individual amino acids may be read with high fidelity
based on, in some circumstances, a single signal spike. A final
accuracy of a call may depend on how many amino acids cluster near
a correct one in parameter space. Should the missed calls be spread
among many other amino acids, then the probability of a majority
call being incorrect would be even smaller than suggested by the
numbers in Table 1.
TABLE-US-00001 TABLE 1 Accuracy with which each signal spike is
called. Amino Acid Accuracy PHE 93% ASP 95% SER 87% ARG 84% THR 93%
GLU 87% HIS 84% LEU 83% ASN 88% GLY 99% CYS 78% IIL 52% MET 81% PRO
92% GLN 82% ALA 92% LYS 83% VAL 89%
[0091] These data were taken with just one run each, and the SVM
was trained on a single amino acid at a time. Run to run variations
in the tunnel gap were not accounted for, and the problem of
identifying multiple analytes from a mixed pool of signals was not
addressed. These problems are addressed for a limited pool of amino
acids below. Note that the ability to separate leucine and
isoleucine is another example of how RT may be used to separate
isobaric isomers which are not distinguished by mass spectroscopy
in addition to the earlier case of L- and D-Apargine.
[0092] In some embodiments, an amino acid/protein sequence reader
which eliminates the requirement for a nanopore is provided. The
principle of this embodiment is illustrated in FIG. 20A.
Accordingly, a protein to be sequenced is immobilized in a reaction
chamber 501, where it is degraded step by step from the N-terminus
using a heat/acid/base cycle with the Edman reagent..sup.6 As each
aliquot of an amino acid from the chains is released, it is passed
to the measurement chamber 503 by means of a microfluidic
connection 502. If the measurement and reaction chambers are of
similar volume (V1=V2) then all of the sample will be passed into
the measurement chamber. An example of a chip for carrying out
Edman degradation is given by Chen et al..sup.9 with aspects of the
chip described further in Chen, Shen et al..sup.10 The chip
described uses Edman chemistry to remove one amino acid at a time
from the N-terminal chain of a peptide, ejecting an aliquot of each
amino acid in a volume as small as 3.5 nanoliters starting from
femtomoles of peptide. The removal of N-terminal amino acids
remains synchronized across a population for about 30 to about 50
amino acids. Thus, the chip ejects each amino acid in turn starting
from the N-terminus into a microfluidic channel. In such
embodiments of the peptide sequencer described here, the tunneling
electrodes may be placed in a microfluidic channel of such a
dimension that the final concentration of amino acid at each read
step is in the range of about 10 to about 100 micromolar, though
lower concentrations may work, but may also give lower count
rates.
[0093] In some embodiments, smaller volumes can be utilized, with
reaction chambers down to the femtoliter range (for example), given
the relatively small tunneling volume between electrodes. In the
case of a 3.5 nanoliter reaction volume, for example, a final
concentration of 10.sup.-4 moles/liter (which yields hundreds of
counts per second in our detector) would require just 350 fM of
starting protein. If the volume was reduced to 10 picoliters (a
cell of 20 microns on each side), then less than a fM would be
needed. In some embodiments, the reading volume and the reaction
volume are substantially the same (and in some embodiments, the
same) or similar. With appropriate treatment of the microfluidic
channels, the only dilution may be caused by diffusion of the
aliquot of molecules along the channels. In the case of a 3.5
nanoliter volume, a characteristic channel length is 150 microns.
Given that the diffusion constant of a typical amino acid is
8.times.10.sup.-4 cm.sup.2/s, the aliquot requires nearly 30 s to
diffuse over this distance. In the case of a 10 picoliter volume,
this time is reduced to 0.5 s. Thus, in some embodiments, there is
an ample amount of time to move the sample and record recognition
signal peaks.
[0094] As shown in FIG. 20B, an example of a measurement cell
according to some embodiments, with nanoliter volume, is shown and
described herein. Accordingly, the aliquot from the reaction
chamber is injected via a microfluidic coupling channel 601 into a
channel 602 that has a width in the region of 100 microns and a
similar height, set by the etching depth 607 into the glass
substrate 608. The lower surface 606 is coated with a film of
suitable metal such as Pd (for example), which is contacted via an
electrical connection (not shown). A metal probe 603, made of a
similar metal (usually Pd) having a sharp exposed apex 604 and
otherwise covered with insulation (as described by Tuchband et
al..sup.11) is positioned through a hole 610 in an upper plate 609
that is bonded over the lower assembly to seal the channel 602. The
probe apex 604 is placed within tunneling distance of the substrate
606 by means of a scanning tunneling microscope controller as is
well known in the art. The apex of the probe 604 is placed a short
distance 605 away from the end of the channel 602. If this distance
is about 100 microns or less, then a nanoliter of sample passed
down the channel 602 floods the tunnel junction between the probe
apex 604 and the substrate 606. Providing that the hydrostatic
pushing pressure is removed once the aliquot is delivered, it will
remain in the vicinity of the junction for at least 30 seconds (as
described earlier). Please note, the dimensions described above are
merely exemplary, as the scale is readily reducible from about 100
micron distances down to about 20 micron distances or less.
[0095] In operation (for example), an aliquot to be measured is
passed to the tunneling junction, measured by means of its
characteristic tunneling signal, and then flushed out by passing
clean buffer into the measurement cell via the channel 601/602. The
cycle is then repeated as needed.
[0096] Another advantage of using a microfluidic channel with a
scanning-tunneling micropscopy (STM) apparatus, according to some
embodiments, is that an adjustable tunnel gap can be included for
added versatility and ease of manufacture. Such embodiments also
may allow identification of species that require a different
tunneling gap. For example, in the case of tryrosine and
tryptophan, a "null" read in the standard tunneling conditions
(0.5V, 4 pA) can be followed rapidly by a sample taken at a smaller
gap (6 or 10 pA current) to see if the aliquot of sample that
produced no signal was tyrosine or tryptophan. In that way, all 20
amino acids can be identified.
[0097] A view of an exemplary nanoliter scale reaction system,
according to some embodiments, is shown in FIG. 20C. Two
microfluidic channels 702 are etched in the form of a cross in a
substrate 700 sealed with a top plate 720. One of the channels 703
contains the nanoliter reaction chamber. Reagents are pushed into
this chamber by the pump/dispenser 712. Reaction products are
released by opening the valves 708 and 710, where they pass into
the output channel 705 to the coupler to the measurement cell 601.
For rinse cycles of the reaction cell, valves 708 and 709 are
opened, and buffer dispensed by the pump 712. For rinse cycles of
the measurement cell, valves 711 and 710 are opened and clean
buffer flowed from the dispenser 713 out via the coupler 601.
[0098] FIG. 20D illustrates a view down into the reaction chamber
(703 in FIG. 20C) according to some embodiments. A weir 810 is
formed in the reaction channel 800 by masking part of the channel
during etching leaving a space of a few microns at the top of the
channel to pass fluid and small molecules but trapping reaction
column beads. A nanoliter or so of column material (such as
C12.sup.9) is flowed into the channel where it piles up against the
weir to form the reaction cell 802. A heater (not shown) is placed
underneath this reaction cell so that the standard cycle of base,
acid and heat can be applied to complete the Edman reaction.
[0099] The overall assembly, according to some embodiments, is
shown in FIG. 20E. The measurement chamber 900 and reaction chamber
901 are coupled via the coupler 601 and may sit on a common base
902 that contains controls for the reaction chip and its heater.
The probe 603 is held above the substrate 606 by an actuator 903.
This consists of a coarse mechanical approach based on a stepper
motor in series with a fine adjustment that uses a piezeoelectric
material as is well known in the art. This may be coupled to the
base 902 via a rigid support 905 and a top housing 904 that holds
the actuator and control electronics for the tunnel gap.
[0100] Prior to sequencing, both the probe 603 and the substrate
606 may be functionalized with an adaptor molecule as previously
described. Buffered electrolyte may then be pushed from a reservoir
on the microfluidic chip 901 to fill the reading channel 602 and
the probe advanced to the desired tunneling current by means of the
actuator 903, and subsequently controlled by the servo circuit. A
set point at a bias of 0.5V between the probe and substrate is a
current of 4 pA (for example).
[0101] The first Edman degradation may then be carried out, and
control valves on the microfluidic chip may be set so as to release
the first amino acid aliquot released from the peptide into the
reading channel up to the point where it is preferably centered on
the junction region between the probe apex 604 and the substrate
606. RT signals may then be acquired for a period of between about
0.1 to about is and recorded for subsequent analysis. Valves on the
microfluidic chip may then be set to flush the sample out of the
junction area. Then a next cycle of Edman chemistry may be carried
out, releasing the next amino acid to the junction for the next
read. This cycle may be repeated out to the limit of reliable
cleaving by Edman chemistry, which may be (for example) up to about
50 amino acid residues. In some embodiments, mixtures of amino
acids can be analyzed by this technique with each read producing
hundreds or thousands of signal peaks, each one of which can be
assigned to a particular amino acid, enabling analysis of the data
well into a number of amino acids beyond 50 amino acid residues,
based on the identity of the last reliably called residue. This is
because contaminants from residues that may fail to have cleaved in
earlier reactions will have been previously recorded, so their
presence in a subsequent reaction can be recognized as an artifact
of the chemistry.
[0102] Accordingly, in some embodiments, the same ability to make
hundreds or even thousands of single molecule calls from even a few
femtomoles of sample for each amino acid, which, corresponds to
small amounts of post translationally modified amino acids, can be
identified.
[0103] Analysis of the signals by means of a multiparameter
characterization of each pulse is described in the publication by
Chang et al..sup.7 Wavelet and Fourier analysis may be used to
characterize the shape of each pulse in the signal train and a
SVM.sup.8 to assign the pulse using data obtained from known
calibration samples. Using a systematic search of parameter
combinations, the SVM may be trained to recognize new data with an
accuracy that can exceed 90% based on just one pulse of the data
train. Thus, many calls will be made for each aliquot of amino
acids that pass into the reader channel.
[0104] In some embodiments, a fixed tunnel gap can generate RT
signals from an amino acid. FIG. 21A illustrates an exemplary
apparatus for sequencing an amino acid according to some
embodiments. A tunnel gap crafted by cutting a palladium wire 2101
on a silicon nitride substrate with an electron beam. A drop of
analyte 2103 is placed onto the gap and signals collected. FIG. 21B
illustrates a transmission electron microscopy image (TEM) of a gap
prior to functionalizing with imidazole carboxamide recognition
molecules--the smallest part of the gap in this image is 2.07 nm.
FIG. 21C illustrates the signals generated with just phosphate
buffer (pH7), and FIG. 21D shows the signals generated after adding
100 micromolar glycine, which shows the current spikes when
molecules of glycine pass through the tunnel gap.
[0105] The data analysis shown in Table 1 was restricted to just
one run for each amino acid tested with the true positive rate
quoted for the SVM trained on a subset of the data in that run so
they do not reflect the important influence of variations in the
atomic details of the tunnel junction form experiment to
experiment. Here, it is shown that amino acids may be reproducibly
identified across different measurements with different tunnel
junctions set to the same set point current. In order to determine
transferability of data, a study was conducted on a limited number
of amino-acids. Since no two tunnel junctions are identical at the
atomic level, that is, signals will likely be different form device
to device, the following measurements correspond with four
different junctions to find characteristics of the signals that are
conserved from junction to junction. In addition, an investigation
was conducted to determine whether the technique would discriminate
between those molecules that are challenging for mass spectral
analysis, without the aid of other techniques.
[0106] Accordingly, leucine and isoleucine (isobaric isomers),
L-asparagine and D-asparagine (as an example of enantiomers),
L-arginine (with a charged side chain) and glycine (with no side
chain), were chosen for analysis. Data was generated according to
the following constraints: (1) control runs (phosphate buffer alone
with concentration in the range of 1.0 to 10.0 mM) were free of any
features (indicating that the buffer solution was free of
contaminants); (2) insulated STM probes had to show no
electrochemical leakage down to below the measurement limit
(<1pA)--electrochemical leakage is sensitive to the
tip-to-surface distance and cannot be simply backed out of the
signal. Thus, leakage often introduces an error into the tunneling
current set point, resulting in variability of the tunnel gap from
experiment to experiment.
[0107] Finally, data from four (4) runs (meeting the above
criteria) for each sample was collected. A key collected criterion
was to find one or more signal parameters that varied
systematically from analyte to analyte, but that remained constant
with repeated runs (using different junctions) on the same analyte.
Examples of signals from L-leucine and L-isoleucine are shown in
FIGS. 22A and 22B. In these figures, the parameters are shifted so
that zero (0) on each axis is the mean value for six (6) amino
acids measured (in four separate experiments each). The scale is
normalized to the standard deviation of the distributions for all
six (6) amino acids. In this two-parameter analysis, most of the
data is overlapped (white area of the probability density
function). A 14 component analysis separates isoleucine from the
other 4 amino acids to 90% true positive rate for each spike in the
signal train. When a large data set is analyzed, amplitudes are
exponentially distributed and cover a wide range, but the
distribution shows that there is a significant difference in the
tails of the distribution (FIGS. 22C and 22D).
[0108] The plots show scatter plots for peak widths and amplitudes
(symbols) overlaid on the fitted probability distribution functions
used by the SVM. The circled region shows that isoleucine can
generate significantly larger peaks than leucine. This is a notable
result, since the structural difference between these two amino
acids is small and subtle (see insets in FIGS. 22A and 22B). The
SVM can use up to 30 parameters that characterize each pulse in a
data train, including parameters that characterize the pulse shape
as well its relation to neighboring pulses in a signal
train..sup.7
[0109] The parameter sets were first subjected to a covariance
analysis and parameters that were highly correlated were rejected
(so do not convey new information). Other parameters were rejected
whose distributions vary from data set to data set with the same
analyte. This reduced the subset of useful parameters to 14. Using
these parameters, the SVM was trained on small subsets of the data,
and then analyzed all of the remaining data (pooled from all six
amino acids). The results of which are illustrated below in Table 2
(parameter combinations not optimized).
TABLE-US-00002 TABLE 2 SVM analysis of 318,000 signal spikes from
pooled data from 6 amino acids, using 4 tunnel junctions each
(Wrong Calls are the fraction of other amino-acid signals called as
the amino acids listed- numbers do not sum to 1 as wrong calls are
randomly distributed among all amino acids.). Signal parameters
analyzed were selected to be robust against variations from
junction to junction. Amino Acid True Positives Wrong Calls
Arginine 0.55 0.07 D-asparagine 0.78 0.09 L-asparagine 0.84 0.08
Isoleucine 0.51 0.05 Leucine 0.82 0.01 Glycine 0.84 0.08
[0110] Random calling of a peak would result in a 17% probability
of a correct call. Even at this early stage, a single peak can call
an amino acid with better than 0% true positive rate (with 90%
discrimination between pairs such as leucine and isoleucine). Each
trapped analyte generates many peaks.
[0111] Post translational modifications play an extremely important
role and detecting them can require very high resolution mass
spectroscopy. Here it is shown that a modified amino acid is
readily identified from within a larger pool of analytes. Signals
were obtained from sarcosine (methylglycine), again running at
least four repeated experiments with different tunnel junctions
(FIG. 22E, panel e). The entire pool was then subjected to the SVM
analysis as described above. The results for this pool (now of
seven analytes) are shown in FIG. 22F. Individual peaks are
assigned to sarcosine with a true positive rate that exceeds
80%.
Examples: Accuracy, Sample Concentrations and Amounts of Sample
[0112] Table 3 below illustrates a comparison of analytical methods
for detection of amino acids and peptides.
TABLE-US-00003 TABLE 3 Lowest Analytical detection Sample Quantity
Linear Method Sample Concentration Volume Detected Range Reference
LC-IMS- Fibrinopeptide A 0.7 nmole/L 5 .mu.L 3.6 fmole 0.001 to 10
.mu.g/mL J. Proteome MS (1.0 ng/mL)* Res. 2010, 9, 997-1006 ICE-MS
Amino acids 0.5 .mu.mole/L NR* -- 0.5 to 2500 .mu.mol/L J.
Chromatography B 2011, 879, 2695-2703 CE Amino acids 6 pmole/L 30
.mu.L 180 NR Anal. Chem. F-labeling attomole 2010, 82, 2373-2379
LC- Peptides 0.4 nmole/L 25 .mu.L 10 fmole 1-500 fmole/.mu.L Nature
QTRAP (0.4 fmole/.mu.L)* Biotechnology, 2009, 27, 633-641 RT Amino
Acids 1 nmole/L 5 .mu.L 5 fmole Digital (Targeted) (unlabeled)
Counting
[0113] While recognition tunneling has potential advantages in
terms of sensitivity, lack of labeling and single molecule
counting, the accuracy of the assignment of single molecule signals
can be improved. In some embodiments, accuracy is better than the
true positive rate shown in Table 3 since each trapped molecule
generates multiple spikes (e.g., at least 10 or more). In some
embodiments, one way to exploit the repeated data is to use, for
example, a majority vote algorithm. For example, if the case of a
pool of six (6) amino acids with a 0.5 true positive rate
considered, with the remaining calls being distributed amongst the
five (5) remaining miscalls with 0.1 probability each, then two (2)
correct calls will occur one time in four (4) reads (1/(0.5)2). Two
(2) incorrect calls of the same kind occur only one (1) in twenty
(20) times (e.g., 1/(0.5.times.0.1) The remaining outcomes consist
of ambiguous reads with two (2) different calls. The accuracy
improves rapidly with the number of spikes sampled as shown in FIG.
23a (probability of a wrong call vs. number of current spikes).
This may be at the cost of an increased number of reads that do not
pass the "majority vote" test, which, in a large train of spikes,
may be too stringent. Other strategies for optimizing the required
size of the winning block vote to maximize accuracy while
minimizing the amount of wasted data, can be used and determined by
testing them with Monte Carlo sampling of real data (for example).
In some embodiments, bursts with high information content may be
clustered, and thus, may be used to further refine data selection
for SVM training. This is illustrated with measurements of the
experimental true-positive rate for the entire pool of data for the
seven amino acids whose spectra were shown in FIG. 22F. FIG. 23b
shows the measured true positive rate as a function of the number
of sequential signal spikes that make the same call. Using only
those signals that are called the same way for 8 spikes in a row
increases the true positive rate from 70% to 99%.
[0114] Lower Concentrations. The collected data outlined in FIGS.
21 and 22 were taken with concentrations of between about 1 to
about 100 .mu.M of target molecule using a large volume liquid cell
on a conventional STM. The collected data may be a consequence of
the experimental design: in such experiments, the tunnel gap was
set up in clean buffer and thereafter, the sample of amino-acid
added after control signals were obtained. Accordingly, target
molecules have to diffuse large (many microns) distances to get
into the tunnel gap. Two strategies are explored: dielectophoresis
and pressure-flow.
[0115] With dielectophoresis, by applying alternating current (AC)
fields to the tunnel gap, charged and dipolar molecules can be
concentrated into the gap. Accordingly, in some embodiments, a
one-dimensional STM (FIG. 20B) is provided which may be stable
enough to permit periods of servo-control of the gap, which may be
interleaved with periods where the tip potential can be altered in
an arbitrary manner. In some embodiments, the solution potential
(controlled by a reference electrode) is altered to trap molecules
in the tunnel gap, which, in some embodiments, sub nM
concentrations may be possible.
[0116] With pressure-flow, microfluidic cells are provided
configured to inject the sample in .mu.l (for example) quantities
adjacent to the tunnel gap, and includes a flow profile configured
to maximize injection into the gap itself. In such embodiments,
this may reduce the absolute amount of sample needed. An exemplary
embodiment is show in FIG. 24, where pressure injection (arrow)
into the tunnel gap 2402 in a microfluidic cell 2400 is shown. A
probe 2404 penetrates a Polydimethylsiloxane (PDMS), microfluidic
cover (not shown).
[0117] In some embodiments, the recognition tunneling, amino-acid
analysis system may be integrated with a high-performance liquid
chromatography (HPLC) system. As illustrated in FIG. 10, alone HPLC
suffers from incomplete separation of the analytes, an issue
addressed by the addition of mass spectrometry (MS). The combined
HPLC-MS has achieved limit of detection (LOD) of 2 pico-mole for
amino acids without derivatization of amino acids. While it has
been demonstrated using isotope dilution and high-resolution mass
spectrometry without derivatization and chromatography of amino
acids, such a technique required hundreds of pico-mole of
peptide..sup.12 Accordingly, some embodiments of the present
disclosure combine recognition tunneling with HPLC to achieve an
improved system, which, in some embodiments, yields advantages in
the identification of steric isomers and isobaric molecules, a
concentration limit in the nM range, a sample volume of
microliters, and an LOD below pico-moles. Thus, recognition
tunneling (RT) coupled to an HPLC column is an alternative to
HPLC-MS. To that end, since some such embodiments of an RT-HPLC can
be effected as a simple sequential analysis system, the
one-dimensional STM of FIG. 20B may be used.
[0118] In some embodiments, the RP-HPLC includes a flow rate of
about 0.1 mL/min with the microfluidic flow cell shown in FIG. 24.
Accordingly, in some embodiments, amino acids injected into the
column generate a tunneling chromatogram of the sample. Liquid
chromatographic methods that may be used include, for example, ion
exchange (IC), reverse phase (RP), and hydrophilic interaction
liquid chromatography (HILIC). In some embodiments of the RP-HPLC,
the elution buffer is optimized to ensure that the tunnel junction
is cleared rapidly after each read.
Example--Reads of Peptides
[0119] Thus far, data for amino acids and modified amino acids has
been shown. Since the ends amino and carboxy ternmini might be
expected to interact strongly with the recognition molecules, it is
not obvious that peptide chains will generate signals. FIG. 23C
shows a typical signal train from a glycine trimer, demonstrating
that peptides will produce recognition tunneling signals.
Example--Sequencing
[0120] FIG. 26 illustrates a micro-reactor coupled to an RT system
which may be used to carry out peptide sequencing. It has been
reported that 19 amino acids of a peptide were sequenced by
monitoring its carboxypeptidase digest products on line using
electrospray ionization (ESI) mass spectrometry..sup.13 In some
embodiments, micro-beads functionalized with exopeptidase are
constrained in the micro-reactor to hydrolyze peptides to be
sequenced and release individual amino acids. The reaction solution
is flushed to the chamber of the RT system for analysis. A series
of RT spectra are recorded as a function of incubation time. As
illustrated in FIG. 27, which shows sequencing a peptide by RT in
combination with enzymatic hydrolysis, each spectrum contains
information on amino acid compositions released at a given time,
from which the sequence of a peptide is deduced. For example, in
one embodiment, peptides were devised to include either random or
repeated amino acid sequences and carboxypeptidase Y (CPY), the
most commonly used enzyme to determine amino acid sequences of
peptides will be used to sequence the peptide..sup.14 Other
embodiments may use carboxypeptidase P..sup.15
[0121] Various implementations of the embodiments disclosed above
(e.g., protein, amino acid and/or peptide sequencing), in
particular at least some of the processes discussed, may be
realized in digital electronic circuitry, integrated circuitry,
specially designed ASICs (application specific integrated
circuits), computer hardware, firmware, software, and/or
combinations thereof. These various implementations may include
implementation in one or more computer programs that are executable
and/or interpretable on a programmable system including at least
one programmable processor, which may be special or general
purpose, coupled to receive data and instructions from, and to
transmit data and instructions to, a storage system, at least one
input device, and at least one output device.
[0122] Such computer programs (also known as programs, software,
software applications or code) include machine instructions for a
programmable processor, for example, and may be implemented in a
high-level procedural and/or object-oriented programming language,
and/or in assembly/machine language. As used herein, the term
"machine-readable medium" refers to any computer program product,
apparatus and/or device (e.g., magnetic discs, optical disks,
memory, Programmable Logic Devices (PLDs)) used to provide machine
instructions and/or data to a programmable processor, including a
machine-readable medium that receives machine instructions as a
machine-readable signal. The term "machine-readable signal" refers
to any signal used to provide machine instructions and/or data to a
programmable processor.
[0123] To provide for interaction with a user, the subject matter
described herein may be implemented on a computer having a display
device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal
display) monitor and the like) for displaying information to the
user and a keyboard and/or a pointing device (e.g., a mouse or a
trackball) by which the user may provide input to the computer. For
example, this program can be stored, executed and operated by the
dispensing unit, remote control, PC, laptop, smart-phone, media
player or personal data assistant ("PDA"). Other kinds of devices
may be used to provide for interaction with a user as well; for
example, feedback provided to the user may be any form of sensory
feedback (e.g., visual feedback, auditory feedback, or tactile
feedback); and input from the user may be received in any form,
including acoustic, speech, or tactile input.
[0124] Certain embodiments of the subject matter described herein
may be implemented in a computing system and/or devices that
includes a back-end component (e.g., as a data server), or that
includes a middleware component (e.g., an application server), or
that includes a front-end component (e.g., a client computer having
a graphical user interface or a Web browser through which a user
may interact with an implementation of the subject matter described
herein), or any combination of such back-end, middleware, or
front-end components. The components of the system may be
interconnected by any form or medium of digital data communication
(e.g., a communication network). Examples of communication networks
include a local area network ("LAN"), a wide area network ("WAN"),
and the Internet.
[0125] The computing system according to some such embodiments
described above may include clients and servers. A client and
server are generally remote from each other and typically interact
through a communication network. The relationship of client and
server arises by virtue of computer programs running on the
respective computers and having a client-server relationship to
each other.
[0126] For example, as shown in FIG. 28, such a system 2800 may
include at least one sequencing device 2802 (or molecule detection
device, and other disclosed embodiments thereof), which is in
communication (wired or wireless) with controller/processor 2804.
Processor 2204 communicates with database 2806, which may store
signatures for various amino acids, peptides and proteins, as well
as collected data from sequencing runs. The processor may include
computer instructions operating thereon for accomplishing any and
all of the methods and processes disclosed in the present
disclosure, including comparing collected current spike data to
signatures stored in the database. Input/output means 2808 may also
be included, and can be any such input/output means known in the
art (e.g., display, printer, keyboard, microphone, speaker,
transceiver, and the like). Moreover, in some embodiments, the
processor and at least the database can be contained in a personal
computer or client computer which may operate and/or collect data
from the sequencer. The processor also may communicate with other
computers via a network (e.g., intranet, internet).
[0127] Similarly, FIG. 29 illustrates a sequencing system 2900
according to some embodiment which may be established as a
server-client based system, in which the client computers are in
communication with sequencers. The client computers may be
controlled by a server 2912, each of which may include the database
for storing current signatures of amino acid, peptide and proteins,
and also be used to collect data (e.g., either or both may include
the database). The client computers communicate with the server
2912 via a network 2910 (e.g., intranet, internet). Each sequencer
device 2902a, 2902b, may each be connected to a dedicated client
2904a, 2904b.
[0128] Any and all references to publications or other documents,
including but not limited to, patents, patent applications,
articles, webpages, books, etc., presented in the present
application, are herein incorporated by reference in their
entirety.
[0129] Although a few variations have been described in detail
above, other modifications are possible. For example, any logic
flow depicted in the accompanying figures and described herein does
not require the particular order shown, or sequential order, to
achieve desirable results. Other implementations may be within the
scope of at least some of the following exemplary claims.
[0130] Example embodiments of the devices, systems and methods have
been described herein. As noted elsewhere, these embodiments have
been described for illustrative purposes only and are not limiting.
Other embodiments are possible and are covered by the disclosure,
which will be apparent from the teachings contained herein. Thus,
the breadth and scope of the disclosure should not be limited by
any of the above-described embodiments but should be defined only
in accordance with claims supported by the present disclosure and
their equivalents. Moreover, embodiments of the subject disclosure
may include methods, systems and devices which may further include
any and all elements from any other disclosed methods, systems, and
devices, including any and all elements corresponding to
protein/peptide/amino-acid sequencing. In other words, elements
from one or another disclosed embodiments may be interchangeable
with elements from other disclosed embodiments. In addition, one or
more features/elements of disclosed embodiments may be removed and
still result in patentable subject matter (and thus, resulting in
yet more embodiments of the subject disclosure).
REFERENCES
[0131] 1. Huang, S.; He, J.; Chang, S.; Zhang, P.; Liang, F.; Li,
S.; Tuchband, M.; Fuhrman, A.; Ros, R.; Lindsay, S. M., Identifying
single bases in a DNA oligomer with electron tunneling. Nature
Nanotechnology 2010, 5, 868-73.
[0132] 2. Lindsay, S.; He, J.; Sankey, O; Hapala, P.; Jelinek, P.;
Zhang, P.; Chang, S.; Huang, S., Recognition Tunneling.
Nanotechnology 2010, 21, 262001-262013.
[0133] 3. Liang, F.; Li, S.; Lindsay, S.; Zhang, P., Chemical and
Hydrogen Bonding Properties of Imidazole-2-carboxamide, a reagent
for DNA Sequencing by Recognition Tunnelling Chemistry 2011,
submitted.
[0134] 4. Lawson, J. W.; Bauschlicher, C. W., Transport in
Molecular Junctions with different molecular contacts. Physical
Review B 2006, 74, 125401.
[0135] 5. Chang, S.; He, J.; Zhang, P.; Gyarfas, B.; Lindsay, S.,
Analysis of Interactions in a Molecular Tunnel Junction. J. Am Chem
Soc 2011, 133, 14267-14269.
[0136] 6. Hempel, J., An oritentation to Edman chemistry. In Modern
Protein Chemistry, Howard, G. C.; Brown, W. E., Eds. CRC Press:
2002; pp 103-122.
[0137] 7. Chang, S., S. Huang, H. Liu, P. Zhang, R. Akahori, S. Li,
B. Gyarfas, J. Shumway, B. Ashcroft, J. He, and S. Lindsay,
Chemical Recognition and Binding Kinetics in a Functionalized
Tunnel Junction. Nanotechnology, 2012.
[0138] 8. Chang, C.-C. and C.-J. Lin, LIBSVM: a library for support
vector machines. ACM Transactions on Intelligent Systems and
Technology, 2011: p. 2:27:1-27:27.
[0139] 9. Chen, W., X. Yin, J. Q. Mu, and Y. Yin, Subfemtomole
level protein sequencing by Edman degradation carried out in a
microfluidic chip. Chem. Commun., 2007: p. 2488-2490.
[0140] 10. Chen, W., J. Shen, X. Yin, and Y. Yin, Optimization of
microfabricated nanoliter-scale solid phase extraction device for
detection of gel-separated proteins in low abundance by matrix
assisted laser desorption/ionization mass spectrometrry. Rapid
Communications in Mass Spectrometry, 2007. 21: p. 35-43.
[0141] 11. Tuchband, M., J. He, S. Huang, and S. Lindsay, Insulated
gold scanning tunneling microscopy probes for recognition tunneling
in an aqueous environment. Rev, Sci. Instrum., 2012. 83: p.
015102.
[0142] 12. Louwagie, M.; Kieffer-Jaquinod, S.; Dupierris, V.;
Coute, Y.; Bruley, C.; Garin, J.; Dupuis, A.; Jaquinod, M.; Brun,
V., Introducing AAA-MS, a Rapid and Sensitive Method for Amino Acid
Analysis Using Isotope Dilution and High-Resolution Mass
Spectrometry. J. Proteome Res. 2012, 11, p. 3929-3936.
[0143] 13. Rosnack, K. J.; Stroh, J. G., C-Terminal Sequencing of
Peptides Using Electrospray Ionization Mas Spectrometry. Rapid
Commun. Mass Spectrom. 1992, 6, 637-640.
[0144] 14. Walker, J. M.; Winder, J. S., C-Terminal Sequence
Analysis with Carboxypeptidase Y. In The Protein Protocols
Handbook, Walker, J. M., Ed. Humana Press Inc.: 1996; pp
569-571;
[0145] 15. Yan-Fei, G.; Hong-Xia, W., Development of C-terminal
Sequencing Analysis of Protein and Peptide. Chin J Anal Chem 2007,
35, 1820-1826.
* * * * *