U.S. patent application number 12/660608 was filed with the patent office on 2011-02-24 for method for accurate measurement of enzyme activities.
Invention is credited to Stephen Joseph Kron, Juliesta Elaine Sylvester.
Application Number | 20110046919 12/660608 |
Document ID | / |
Family ID | 43606031 |
Filed Date | 2011-02-24 |
United States Patent
Application |
20110046919 |
Kind Code |
A1 |
Sylvester; Juliesta Elaine ;
et al. |
February 24, 2011 |
Method for accurate measurement of enzyme activities
Abstract
This invention relates to a novel method for the use of
specialized equipment to significantly improve the accuracy with
which enzyme activities are measured. The method establishes an
internal standard mathematical curve for the measurement of
relative values for each unit in a complex mixture. The novel
feature of the method is that the control is an integral part of
each and every sample, so that any variation in test conditions
affects the control in exactly the same manner as the sample.
Therefore, this novel method forces each measured sample to have a
separate control internal to each sample that is exactly matched as
to all possible variables. Although some prior methods claim an
internal control, the prior methods provide only a rough
calibration to an external control, which results in substantial
deficiencies in accuracy, reliability, and efficiency.
Inventors: |
Sylvester; Juliesta Elaine;
(Chicago, IL) ; Kron; Stephen Joseph; (Oak Park,
IL) |
Correspondence
Address: |
Juliesta Elaine Sylvester Unit 3
5618 South Kimbark Avenue
Chicago
IL
60637-1606
US
|
Family ID: |
43606031 |
Appl. No.: |
12/660608 |
Filed: |
March 1, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61208925 |
Mar 2, 2009 |
|
|
|
Current U.S.
Class: |
702/179 |
Current CPC
Class: |
G16B 40/00 20190201 |
Class at
Publication: |
702/179 |
International
Class: |
G06F 17/18 20060101
G06F017/18 |
Goverment Interests
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0002] A portion of the research activities involved in the
refinement of the methods described herein was supported by U.S.
government funding from the National Institutes of Health, listed
under NIH Funding Agreement Numbers HG003864, CA126764, and
CA103235, and NIH EIR number 1413601-09-0005.
Claims
1. A method for accurate measurement utilizing specialized
equipment, comprising the steps of: (a) Establish a foundation for
comparison in form of measured internal standards for each
individual experimental segment, through accurate measurement of
each standard segment, strata consisting of a group of segments,
and individual samples within each segment. (b) Separate
measurements of the number of samples and the attained score, for
each standard segment, each test segment, and each strata including
a specific group of segments. (c) Separate calculation of the mean
and the confidence interval, with modification based on the
specific sample size for each segment based on the data for each
segment and each strata, for both the standard and test
circumstances. (d) Calculation of the mathematical curve that
describes the test results, based on a least squares fit,
correlation coefficient, or other accepted criteria. (e) Accurate
measurement of the result of the change for each sample. (f)
Calculation of the statistical significance of the test results, to
establish standards that prevent distortion from the effect of
factors or circumstances that are uncontrolled.
2. A method according to claim 1 wherein: (a) the internal
standards are based on a strata of more than one point per sample,
and the standard circumstance is a known percentage of expected
product; (b) the attained score is the measured amount of readout
from a detector; (c) each segment is an individual test; (d) each
strata is a group of segments that are intended to be identical,
except for uncontrolled variables (e) the score that is measured is
the individual readout from each individual product in a test (f)
components of the sample are each individual products monitored
within a single test; (g) the equipment used has the basic
capabilities to support the accuracy and particularity of
measurement necessary for the method described herein.
3. A method according to claim 1 wherein: (a) the internal standard
curves established as controls are based on measured amounts for a
known standard condition for application with alternative
specialized equipment; (b) the alternative equipment may include,
but is not limited to, mass spectrometry, Western blot,
chromatography, nuclear magnetic resonance, or various other
chemical or biochemical equipment, instruments, fixtures, or
ingredients; (c) the amount of change is measured by observation of
change in the sample from the control, including but not limited to
changes in mass, function, fluorescence, radioactivity, intensity
in arbitrary units, or other observable information. (d) the test
samples are material with amounts and characteristics suitable for
measurement by the selected method, (e) the data results are
calculated by any suitable method, software, or procedure, that
would produce results that provide meaningful data useful for
inference, including but not limited to a measured amount for each
control, and for each segment, each strata, and each sample, with
separate identification of each group of samples and each sample,
and a physical count of the number of samples tested.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefits of U.S. Provisional
Application No. 61/208,925 filed Mar. 2, 2009, the teachings of
which are herein incorporated in their entirety, by reference.
COMPUTER PROGRAM LISTING APPENDIX
[0003] A computer software program is attached in two replacement
compact discs, which are identical, and which contain no new matter
from the original compact disc which was created Feb. 25, 2010. The
compact disc contains one ASCII file, named Patent-Computer
Program, size 37 kilobyte, that contains the format and sequence of
computer instructions that are the preferred embodiment for
processing of equipment data output into statistically significant
data for functional curves. The contents of the compact are
incorporated by reference as part of this application. The data on
the compact disc is an exhibition of results from flow cytometry
measurements using internal controls, with detailed calculations of
the confidence level for each measurement based on statistical
inference. A portion of this patent document contains material that
is subject to copyright protection. The copyright owner has no
objection to the facsimile reproduction by anyone of the patent
document or the patent disclosure, as it appears in the Patent and
Trademark Office patent file or records, but otherwise reserves all
copyright rights whatsoever.
FIELD OF THE INVENTION
[0004] The present invention relates to the field of improved
methods for accurate measurement in laboratory and clinical
conditions. More specifically, it relates to a method for the
accurate measurement of components in a test sample using internal
standards. The internal standards are used to define a mathematical
function that relates calibrated reference points to each other and
to an experimental sample. The unique feature of the method is that
the relationship between internal standards is defined by a
mathematical function with a statistically relevant fit; this
mathematical relationship is used to describe experimental samples
with accuracy and precision.
DESCRIPTION OF THE RELATED ART
[0005] The prior art reflects the continuing challenge to obtain
accurate and precise measurements under conditions wherein many
experimental conditions are difficult to control. Significantly,
the prior art is silent with regard to the influence of random
factors on accuracy, statistical significance, and the information
content of data used to derive relevant mathematical curves. The
methods and protocols presented in relevant technical papers
describe a typical embodiment of quantitative analysis that
displays wide data dispersion and low correlation coefficients. The
prior art highlights an important point that the obstacles to
precision become more difficult for test measurements of
biological, chemical, and physical functions at the molecular
level.
[0006] The prior art is shown in the listing of patents and
relevant technical publications on pages 2-4. The failures of prior
art are demonstrated by the omissions in prior patents. Prior art
recognizes the necessity of constructing a calibration curve
(Akhavan-Tafti, 2007, paragraph 0044) and the importance of an
internal standard (Chandler 2001, paragraph 0025). However, the
methods described result in only an external standard (Chandler
2001, claims 1 through 15) such as daily calibration for
machine-to-machine differences (Chandler 2005, paragraph 0129).
[0007] The use of even basic internal positive and negative
standards for calibration has been shown to improve assay
reliability in biochemical experiments using bead arrays (Martins
2002; Hanley 2007). Internal standards are widely used in
analytical chromatography, Western blots, quantitative PCR, and
quantitative mass spectrometry. The reason for the inclusion of
internal standards is that they are especially useful for analyses
in which the quantity and quality of sample varies from run to run
for reasons that are difficult to control.
[0008] For example, gas chromatography/mass spectrometry-based
analysis of pollutants demonstrated that internal standards
corrected for systematic errors while external standards introduced
bias that affected measurement accuracy (Kirchmer 1983). In
quantitative polymerase chain reactions (PCR) an internal control
gene is used to normalize samples for measures of relative
abundance (Livak 2001). Similarly, in quantitative mass
spectrometry a control sample is detected simultaneously to provide
relative abundance by direct comparison to experimental samples
(Ong 2002).
[0009] In these cases, quantitation is carried out by a direct
ratio of signal obtained from the experimental sample divided by
the signal obtained from the control sample. These methods do not
define mathematical functions established by statistical parameters
that correct for detector response and uncontrollable variables in
handling that influence abundance between samples. Indeed, in the
case of quantitative mass spectrometry, control samples are
prepared and handled independently from experimental samples,
thereby excluding them from providing accurate quantitation beyond
instrument calibration.
[0010] Although internal positive and negative controls have been
included in prior art, internal standard curves that define a
mathematical function through more than one calibrated internal
standard have not been implemented to quantitatively analyze
experimental samples. Although prior art includes internal
standards, the standards are used to compare detector performance
between samples. Thus, the internal standards serve as basic
comparisons for detector calibration instead of quantitative tools
for sample-specific analyses.
[0011] For assays that measure multiple components of a reaction,
an additional concern is the normalization of output contributions
by various components of the reaction. For example, while multiple
peptides are readily detected in a single scan by mass
spectrometry, the difference in ionization potential between
phosphorylated and un-phosphorylated forms results in
over-representation of one species over another (Busman, Schey et
al. 1996). Internal standards for one molecular species are
irrelevant for the other. Accordingly, only ratios of observed
abundance can be obtained to describe relative relationships
without statistical significance.
[0012] Generally, the prior art does not provide a sufficient
number of internal controls to ensure accurate measurements of test
samples against calibrated standards. The prior art does not use
internal standards to quantitatively analyze the test samples with
statistical confidence. Although the prior art uses calibrated
standards, the prior art does not force all measurement conditions
for the standard and the test sample to be identical. Technical
challenges associated with standardizing reagents used in chemical
and biological methods result in inherent problems caused by
uncontrolled variations between samples and between
laboratories.
[0013] For example, high-throughput biological screens often note
that the organization of samples within plates and small variations
between plates can lead to strong sample bias (Koren, Tirosh et al.
2007). Significantly, even basic quantitation in chemistry and
biology requires external standards for a combination of background
subtraction resulting from non-specific interactions in complex
samples, and the calculation of proportions relative to a baseline
(Xiaoming, Syuhei et al. 2008).
[0014] Although several papers discuss internal controls, the
typical methods described depend on external standards and are
therefore not sensitive to uncontrollable variables between
samples. For many circumstances, an internal standard is only used
as a baseline to provide a proportionate measure of abundance for
experimental samples. In most cases, the prior art fails to provide
a method to simultaneously measure both the control standard and
the test sample simultaneously, as an integral part of the test
sample. Because prior art necessarily includes inherent variation
in test conditions that involve separate controls and test samples,
the result is significant inaccuracy. The prior art that does
include internal controls uses them as calibration standards and
does not allow multiple internal control standards to provide
multiple simultaneous comparisons of the test sample versus the
control standards or the control standards versus each other.
The Need for a New Method
[0015] The following is a discussion of problems with the prior art
and the resulting need for a new method that would resolve the
important measurement issues.
The Requirement for Accuracy
[0016] Extreme precision is required for measurements at the
molecular level, such as enzyme activity. Despite the importance of
accuracy, the prior art shown in current publications result in
high data dispersion and resulting low confidence in the test
results, even after selective exclusion of outlier data. As a
result, many authors fail to show the statistical confidence
intervals in data plots, and typical functional curves exhibit low
correlation coefficients. Typically, the resulting wide data
dispersion supports only a rough figure of merit (Zhang, Chung et
al. 1999).
[0017] This invention solves these difficulties by establishment of
a standard curve based on multiple data points for each reaction to
provide a calibrated standard response. In this way, the previously
uncontrolled factors are measured with precision, so that sample
data from each test can be meaningfully compared. This novel method
uses an internal standard curve for each test and for triplicate
groups of tests during an experiment. This method provides a
standard for comparison that prevents imprecision caused by subtle
but uncontrolled variables. Therefore, this method provides a
practical solution to the critical need for high accuracy
measurements under conditions wherein many variables are extremely
difficult to control.
[0018] This novel method is applicable to the broad range of test
equipment and procedures for precision measurements required for
biochemistry, biophysics, and chemical engineering. By contrast
with prior art, the specific protocol described by this novel
method results in minimal data dispersion, calculation of the mean
with narrow confidence intervals at the 0.01 level, and derivation
of precise mathematical curves based on the least-squares fit with
an unusually high correlation coefficient. This method establishes
calibrated internal controls with statistical relevance to avoid
the data dispersion caused by uncontrollable variables.
The Requirement for Comprehensive Internal Standards
[0019] By definition, an internal standard must be based on each
sample, and measure all relevant variables that could reasonably
affect the test. To qualify as an internal standard, the calibrated
standard control must be established for the test conditions for
each sample. For the accuracy required for measurements at the
molecular level, an internal calibration standard must be based on
samples drawn from a specific segregated population under
controlled test conditions. As a critical flaw, the calibration to
external standards has the inherent risk of changed test
circumstances due to variables that are very difficult to
control.
[0020] The methods to develop and apply an internal standard curve
for quantitation of individual sample data are unique.
Significantly, the internal standard curves provide the basis for
reaction-specific quantitation while controlling for detector bias
and uncontrollable factors from handling procedures such as
non-homogenous distributions. Internal standards are a practical
requirement for analyses that exhibit differences in sample
quantity and quality for different runs. For many types of test
equipment, internal standard curves are rare because most platforms
do not allow for the analytical separation of multiple components
in the same test sample.
[0021] For many types of laboratory tests, internal controls are a
practical necessity due to many variables that are difficult to
control, such as sample variables or test conditions. For example,
establishment of internal standards is necessary for quantitative
chromatography and mass spectrometry, where injector and detector
performance with small volumes is not reproducible. By contrast,
establishment of internal standards remains a challenge for many
test methods, such as Western blot, antibody-based studies, and
array equipment.
[0022] Statistical inference is a practical necessity to establish
the accuracy of the test measurements. Definitions of terms used
for statistical inference depend on the circumstances under which
tests are performed. As applied to array equipment, the population
is defined as each reaction in the array from which samples are
drawn. The score is the measured signal intensity. The sample is
the number of individual units analyzed by a single use of the
sampling device. The stratified sample is defined as the combined
results of triplicate tests designed to exhibit identical
conditions that can be controlled.
Problems Caused by Uncontrolled Variables
[0023] It is noted that uncontrolled variables typically cause
minor variations in test results. For example, even with the same
test equipment, for reactions with the same mixture and tested at
the same time, there are variations in the number of samples (n)
and the score of each sample (x). Therefore, even minor variations
in the sample data may result in major changes in the mean,
confidence interval, and mathematical function curve.
[0024] A critical requirement for an internal standard curve is
that an assay must be able to measure more than one component in a
mixture. Although new equipment, including mass spectrometers and
high-density matrix arrays, has been developed to measure multiple
components of a biochemical reaction, allowing improved test
measurements at the molecular level, the test methodology typically
results in ambiguous results. This ambiguity is caused by the
presence of variables that are difficult to control.
[0025] For example, although there are several types of new
equipment and reagents that aim to measure enzyme activity, the
results of the methods and protocols described in recent technical
papers typically display wide data dispersion and a low correlation
coefficient, and resulting low confidence in the mathematical
functions derived from the ambiguous data. Typical recent
scientific papers exhibit data results that are not statistically
significant and reflect wide random variation. Generally, an
underlying reason for this wide dispersion of test data is test
conditions that are difficult to control.
[0026] As applied to equipment that samples a given reaction
multiple times, the internal standard curve effectively corrects
for factors that are not feasible to control between samplings,
including but not limited to the number of measurements acquired
and variability in the reaction mixture between samplings. Examples
of conditions that are difficult to control in a multiplexed bead
array include the number of microspheres sampled and analyzed and
identical reaction mixtures in each sampled unit.
[0027] Current technology allows the measurement of multiple
components in a reaction and multiple reactions in an experiment,
providing the opportunity for rapid analysis of many samples.
Current analytical techniques based on this technology have
resulted in a wide dispersion of data that results from conditions
that are very difficult to control due to measurements on a very
small scale. The invention solves these difficulties by
establishment of a standard curve based on a set of known data
points in each reaction unit to provide a set of calibrated
standards with statistically relevant response. In this way,
previously uncontrolled factors are measured with precision, so
that results from each component in each reaction can be
meaningfully compared.
DISCUSSION OF UNRESOLVED ISSUES WITH THE PRIOR ART
[0028] This method is applied for the accurate measurement of
enzyme activity, which is an important issue and essential for
future biochemical research. Various enzyme activities are
responsible for intracellular signaling cascades that lead to
changes in cellular physiology. The post-translational
phosphorylation of proteins, considered the most common means of
intracellular signal propagation and amplification, is
traditionally surveyed by a series of experiments, each querying a
single kinase. New technologies have made it possible to measure
more than one kinase activity in a single reaction.
[0029] Kinases are enzymes that transfer the .gamma. phosphate of
adenosine triphosphate (ATP) to tyrosine residues on substrate
proteins (Robinson, Wu et al. 2000). Kinases mediate critical
growth and survival signaling pathways in response to cell-to-cell
contact, peptide hormones, and cell stress. Inappropriately
activated kinases play a role in cancer initiation and progression.
As such, they are key pharmaceutical targets and research
pertaining to their activities involves the investment of hundreds
of millions of dollars per year.
[0030] As an example of the accuracy issues involved with prior
art, the following discussion focuses on a common example of
measuring kinase activity to monitor intracellular signaling. The
classic approach of measuring kinase activity is to measure the
incorporation of radioisotope-labeled phosphate. This method is not
easily applied to complex samples with multiple components, such as
cell lysates, and cannot be used to measure more than one kinase
activity in a single sample.
[0031] Several current methods, such as Western blots or
microscopic techniques using immunohistochemistry and
immunocytochemistry, only infer kinase activity by indexing changes
in kinase expressions and phosphorylation states over time. In
these cases, the observed signal is a compiled average of the total
molecular content in the sample. Triplicate experiments are
performed when possible to derive some measure of predictability in
the results; however, data are only qualitative and cannot be used
to provide statistical inference. With the ability to calculate the
average signal per population of cells, flow cytometry provides
quantitative measures of phosphorylation events in cells (Perez and
Nolan 2002).
[0032] However, internal standard curves have not been implemented
to increase confidence in measurements. In general, these methods
are heavily dependent on existing protein-specific antibodies and
detection is focused on a single reagent, for example the
phosphorylation state of a particular kinase. Therefore, these
methods have a limited capacity for measuring multiple components
in a single reaction. In principle, a phospho-specific antibody
could be matched to each substrate and detected independently in
solution, but this creates significant challenges. The development
of internal standards by extension of this strategy is challenging
but can be done.
[0033] A critical requirement for the embodiment of the invention
is that multiple components of a reaction must be monitored
simultaneously. To monitor multiple components in a reaction,
methods have been developed that provide unique tags to capture
multiple substrates from a solution-phase reaction (Shults, Kozlov
et al. 2007). Kinase assays are also performed with peptide and
protein substrates tethered to surfaces (Henderson and Bradley
2007). For example, peptide microarrays offer detection of multiple
substrates with spatial addressing. These formats allow the
interrogation of multiple kinase activities in cell extracts with
the detection of a single generalized label, such as a fluorescent
anti-phosphotyrosine antibody (Houseman, Huh et al. 2002). This
chip-based approach can also be adapted to a multi-well format (Wu,
Mand et al. 2008).
[0034] An alternative format for monitoring multiple components in
a reaction is the use of bead arrays. Bead arrays can be used to
monitor the phosphorylation of multiple endogenous substrates in
cell lysates by immobilizing a different capture antibody on
different types of beads and detecting phosphorylation with a
second phospho-specific antibody (Du, Bernasconi et al. 2009).
Kinase activity assays are readily implemented using kinase
substrates immobilized on beads (Wu, Nair-Gill et al. 2005). It is
therefore straightforward to design activity assays using beads to
simultaneously monitor more than one component of a reaction
(Bernsteel, Roman et al. 2008). Although the use of beads in an
assay allows sample counting and population-based statistics, these
principles have not been previously implemented to increase
measurement accuracy and confidence.
[0035] The solution to the problem of assigning statistical
significance to data that displays high variability is to increase
the number of replicate samples analyzed. Generally, prior art is
limited to the analysis of experiments performed in triplicate.
Data are sorted by qualitative comparisons and outliers are removed
without explanation. The results may or may not be predictive of
future attempts and statistical validation is not available to
provide a measure of confidence. The present invention solves this
issue by providing quantitative internal standard curves for each
reaction performed in triplicate. Therefore, the invention allows
for each of the triplicate test samples to be (a) simultaneously
measured, (b) identical as to all known controllable conditions,
and (c) include the calibrated control standard within each
sample.
[0036] For practical application, the issue is whether this method
provides additional information that cannot be acquired through
other means. Several companies have produced equipment designed to
allow high-throughput assays of kinase phosphorylation, including
Luminex, Cell Signaling Technologies, Kinexus Bioinformatics Corp.,
and Qiagen. Established sources can supply control beads for
testing bead I.D. and reporter I.D; however, these controls only
validate the working condition of the instrument and do not provide
an analytical tool for experimental tests. Although there is a wide
range of specialized equipment and activity assay formats, accurate
measurement of kinase activity has remained a persistent challenge.
The use of internal standard curves for experimental quantitation
and statistical validation significantly improves the reliability
of sample comparisons.
SUMMARY OF THE INVENTION
[0037] The invention is a novel procedure for use with test
equipment that substantially increases the accuracy and reliability
of the equipment measurements. The test equipment includes, but is
not limited to, machines that utilize multiplexed arrays, mass
spectrometry, liquid chromatography, or other chemical and
biochemical devices typically used in a laboratory or clinical
setting. The novel procedure is a method to utilize the specialized
equipment in a unique way, so that a control is established as a
foundation for comparison.
[0038] The control data is based on precision measurement and
statistical inference for the unique characteristics of each
individual reaction or sample group. This measurement method
prevents errors based on variation that is not feasible to control
in test samples. This measurement method allows accurate
measurements at the molecular level, under conditions wherein
uncontrolled random differences may produce data scatter that
prevent accurate measurement.
[0039] For both the control data and the comparison test data, the
subject matter to be tested is divided into separate reaction
vessels that are intended to be identical, excepting only one
controlled variable. Significantly, each reaction typically has
random differences that cannot be controlled with existing
technology. Then, samples are taken from each reaction, under
various conditions ranging from a baseline condition to an extreme
test condition. For each reaction and each condition, there is a
physical count of the number of samples (n), and a measurement of
the attained score for each sample in each group (x). From this
data, statistical inference methods, with modification for the
specific sample size (t-distribution), provides the mean and the
confidence interval for the mean.
Control of Random Variations
[0040] The random differences between nominally identical reactions
are controlled through an additional step. The nominally identical
reactions are grouped into a strata, with separately calculated
statistical results. Because this detailed data is provided for
each sample, each reaction, and each strata, the random differences
are controlled because the variations between nominally identical
reactions are measured.
[0041] Terminology varies with the specific equipment. For example,
for multiplexed bead arrays, a reaction vessel is defined as one of
the wells in a 96-well array, a score (x) is the measured
florescence at the surface of each microsphere, the mean is
calculated based on attained scores for each microsphere in a
specific well, and the confidence interval is calculated from the
mean and the deviation from the mean. Then, the strata are defined
as groups of three nominally identical wells per controlled
variable.
Calibrated Standard for each Experimental Sample
[0042] As applied to the measurement of kinase activities, the
principle and distinct feature of the invention is to add a set of
four internal standards to each experimental sample prior to
antibody labeling for the accurate measure of kinase activity. The
internal standard curves are generated by at least four points per
well of a known percentage of substrate phosphorylation; these
curves are used to translate the fluorescence readout from bound
anti-phosphotyrosine antibody to a meaningful scale. Standard
curve-based calculation of well-to-well variations in antibody
binding allows for the measurement and validation of small changes
in substrate phosphorylation by un-fractionated cell lysates as
well as purified recombinant enzymes.
[0043] The accuracy of each measurement ensures that the assay is
sensitive enough to be used with very small amounts of cell lysates
and/or concentrations of additional reagents, including ATP. As a
functional assay, examples of this method are limited by the
endogenous activity of the enzyme of interest and the specificity
of the tested substrate for that enzyme. The prior test methods
typically result in wide data dispersion, which has resulted in the
adoption of selective editing of outlier data points and
non-dimensional figures of merit. By contrast, this method
establishes a calibrated internal standard curve for each test
sample, which effectively controls the previously uncontrollable
variables, and results in highly accurate test results, with a
precise fit to standard mathematical functions and a narrow
confidence interval at the 0.01 level of significance.
ADVANTAGES OF THE NEW METHOD
[0044] By contrast with prior art, the specific protocol described
by this novel method results in minimal data dispersion,
calculation of the mean with narrow confidence intervals at the
0.01 level, and derivation of precise mathematical curves based on
the least-squares fit with an unusually high correlation
coefficient. It has been demonstrated that this method provides a
high level of accuracy and resulting confidence in the test
results. This method of establishment of calibrated internal
controls to avoid uncontrollable variables is expected to be
applicable to a wide variety of test conditions that require
extreme precision and high confidence in the results.
[0045] The method measures both the control and the test variable
simultaneously under identical conditions. Importantly, calibrated
reference points and experimental samples are processed in a single
test run so that random and intentional conditions affect all
measured values identically. Therefore, the broad range of
variables which are difficult to control are integrated into the
calibrated standards, so that the measured results for the test
condition are caused by the changed circumstances instead of the
uncontrolled variables.
[0046] Using internal standards, samples can be monitored with
sufficient sensitivity to allow accurate measurements using very
small amounts of test material. As a typical embodiment, the method
is applied to the measurement of enzyme activity for the functional
analysis of biological signaling events and the identification of
effective inhibitors for the treatment of disease; however, the
method can be applied to any chemical or physically modifying
reaction and may be useful in the establishment of industrial
processes.
[0047] Due to the precision of measurements using internal
controls, the test data exhibit very narrow confidence intervals
and unusually precise fit to the derived mathematical curves. By
contrast, as shown by many recent publications of tests of protein
functions, prior art typically results in a wide data scatter with
inherent high risk of error. This novel method offers unique and
valuable advantages over prior art, based on the demonstrated high
accuracy. With this method, the broad range of variables which are
difficult to control are integrated into the calibrated standard
controls, so that the measured results for the test condition are
effectively segregated from the uncontrollable variables.
[0048] This novel method provides the foundation for prompt,
accurate measurements that are necessary for high-throughput and
clinical assays, including enzyme inhibitor screens and diagnostic
testing. For example, this method is useful for accurate assays of
kinase activity as measured by substrate phosphorylation and
derived from small changes in bound antibody fluorescence over time
in a large array of samples. Typical applications include, but are
not limited to, the analysis of cellular signaling pathways and
assessment of the effectiveness of pharmaceutical inhibitors.
[0049] Accordingly, this novel method has potential for wide
applications to address biochemical issues that require accurate
measurement for reliable comparisons, combined with high
throughput. Included in the method is custom software that allows
for the accurate calculation of confidence intervals for the
specific sample size for each unit, using widely accepted
statistical inference criteria. A typical embodiment of this method
would include equipment that allows multiple components of a sample
to be measured simultaneously. The method is applicable to several
types of laboratory and clinical equipment, such as but not limited
to, bead arrays, chip arrays, mass spectrometry, and liquid
chromatography
BRIEF DESCRIPTION OF THE DRAWINGS
[0050] FIG. 1 is a summary of the primary elements of the test
procedure. This figure describes the essential features for the
establishment of test conditions that offset the effects of
uncontrolled factors. This is done by precise physical measurement
including a physical count of the number of units per sample. The
internal calibration standards are established based on test data
from the same test equipment and the same test mixture.
[0051] FIG. 2 is a summary of the primary elements of the method,
with emphasis on accurate measurement and the use of statistical
inference to provide confidence in the results. Because the test
method results in minimal scattering for the data points, the
mathematical curve that describes the functional effects exhibits a
high correlation coefficient. Mathematical curves are established
for each test condition to serve as the calibrated control
standards.
[0052] FIG. 3 is a summary of the typical embodiment of the method
for measurement of kinase activity using equipment that can
simultaneously monitor more than one component of a reaction. A
standard curve is derived for each reaction in each well in a
96-well array. Peptide substrates are covalently immobilized on
specialized beads that can be analyzed in distinct populations. The
test sample is compared to an internal standard curve constructed
from synthetically phosphorylated peptide substrates.
[0053] FIG. 4 summaries the essential characteristics of the
internal standard curve in a typical embodiment. For array-based
embodiments, the standard curve describes each reaction in an
array. In a typical embodiment, standards are added to each well
after termination of the kinase reaction and before labeling the
phosphorylated substrate with fluorescent antibodies. The internal
standard curve corrects for differences in antibody binding and
uncontrollable reaction conditions between wells.
[0054] FIG. 5 demonstrates that for typical biochemical equipment
replicate wells and replicate runs of the same well result in
variability and scattered data distributions with low confidence.
In FIG. 5A, synthetic standards with increasing phosphorylation
were arranged in separate wells, as in traditional calibration
curves. In FIG. 5B, standards constructed from three different
forms of background signal were compared to estimate the variance
in non-specific detection. For both cases, standards were arranged
in a 96-well plate, labeled with fluorescent anti-phosphotyrosine
antibody, and the detector was run three consecutive times on
standard laboratory equipment (BioPlex 200, BioRad). The only
change to the plate between runs was the added Luminex running
buffer from the BioPlex 200 upon replacement of the sampled beads.
Triplicate columns are grouped to show separate measurements for
runs 1, 2, and 3 per well, for triplicate wells, in arbitrary units
of raw fluorescence intensity. While some wells had consistent
values across three runs, others showed large variability from run
to run. Error bars represent the 99% confidence interval around the
mean and are a function of the number of beads sampled per run.
This variability between runs and between wells undermines the
effectiveness of an external standard but does not affect the
validity of an internal standard. This data strongly support
well-specific internal standards for accurate measurements that are
independent of plate-to-plate and run-to-run fluctuations.
[0055] FIG. 6 shows the preferred embodiment of test data. FIG. 6A
shows the calculation of the mean for 7 data points, showing the
confidence interval (CI) at the 0.01 level, which is a 99%
probability that the results are not due to random variables. The
mathematical curve shows the correlation coefficient (R.sup.2) of
0.98, which shows a very high confidence that the curve describes
the underlying test data. Replicates of the same well were not
required because measurement confidence can be derived from each
sample. FIG. 6B shows the effect of lapse of time on the percentage
of substrate phosphorylation by kinase in cell lysates. Data were
transformed through internal standard curves from raw fluorescence
intensity to the accurate percentage of phosphorylation, and fit to
a saturating hyperbolic curve with a very high correlation
coefficient of 0.94 to show the accuracy of the underlying data.
This method allows a mathematical analysis of the data to confirm
enzyme activity with high confidence based on a single objective
experiment.
[0056] FIG. 7 is a plot of the unique standard curve for each of 48
wells in an array. The plot shows the relationship of fluorescence
intensity to the percentage of substrate phosphorylation. For each
well, the Boltzmann-Sigmoidal curve provides a very good fit,
emphasizing a concentration-based dose response. Significantly, the
variability between internal standard curves in each well
highlights the weaknesses of prior methods, which use only one
external standard curve per plate or per day.
[0057] FIG. 8 shows the low data dispersion and resulting accurate
calculation of the mathematical function curves based on test data
derived using the novel test procedure, based on internal
calibration standards. FIG. 8A illustrates the raw data from a
typical embodiment using three distinct peptide substrates,
Abltide, Srctide, and a peptide derived from Btk, immobilized on
Luminex beads and treated with dilutions of the inhibitors imatinib
(.mu.M range, at the right) and dasatinib (nM range, at the left).
Results for multiple substrates detected from the same reaction
mixture are analyzed with 99% statistical confidence about the
mean. Each curve is self-normalized with a minimum at 0 and maximum
at 1, to facilitate visual comparisons between substrates with
absolute values that differ by up to 100-fold. FIG. 8B demonstrates
the effect of using internal standard curves to transform each data
point according to well-specific parameters established by the
calibrated control standards. The phosphorylation of each substrate
is related to the internal standards by a Boltzmann-sigmoidal curve
that defines the system response. The sigmoidal inhibitory curves
demonstrate altered slopes when defined in relation to the internal
calibrated standards. This highlights accurate relative differences
between components in a single reaction. This figure shows that the
test method allows sufficient accuracy for a detailed comparison of
multiple variables in a single test.
[0058] FIG. 9 presents a selection of data organized by the custom
software included in the invention and used to analyze results from
a typical embodiment for statistical significance. In this
selection an internal calibrated standard for 50% phosphorylation
is described in 24 wells of an array. In each well, the number of
beads sampled and the median, mean and standard deviation of the
signal are reported by the standard laboratory equipment. For each
sample, the standard error and 99% confidence interval is
calculated based on the standard error (SE), which is calculated
from the sample size (n), mean intensity (X), and the standard
deviation (SD), which are the basic outputs from typical equipment.
The custom software performs the critical calculations to establish
the statistical significance of the measured data. This calculation
is performed for each component of a reaction, including each
calibrated standard so that the statistical significance is
described for each and every measurement. The preferred embodiment
of this software is shown in the computer program listing appendix
submitted on a compact disc.
[0059] FIG. 10 is a description of the specific calculation
procedure provided by the custom software that results in the
accurate calculation of the standard error (SE). This figure shows
the specific keystrokes within Excel (Microsoft Office, 2008) that
comprise the critical calculations. Standard statistical software
based on a normal distribution is not used, because the small
sample size requires the t-distribution. Within Excel, the
t-distribution values are arranged in a detailed lookup table.
Thus, this custom software results in an accurate calculation of
the statistical significance of each measurement. The details of
this software are shown in the computer program listing appendix
submitted on a compact disc.
DESCRIPTION OF THE PREFERRED EMBODIMENT
[0060] The following provides a description of the present
invention, a typical application of the method including the
procedure for establishment of the calibrated internal standard
curves, and an example of results of the test procedure. The
application described utilizes groups of microspheres arranged in a
96-well plate to facilitate parallel measurements on the effect of
different kinase inhibitors used in the treatment of leukemia.
[0061] The aim is to improve the resolution of diagnostic tests for
chronic myelogenous leukemia (CML) by providing more than one
marker of patient progress. To achieve statistical significance for
accurate sample comparisons, we developed an embodiment of the
invention that used a platform of internally fluorescent
microspheres (acquired from the Luminex corporation) for
simultaneous measurements of experimental and control parameters to
monitor kinase reactions in human cell lysates.
[0062] Bead array analysis, implemented using the Luminex platform
of internally-fluorescent polymeric microspheres, offers an ideal
format for monitoring multiple analytes in a single sample (Fulton,
McDade et al. 1997). The platform uses a dedicated flow cytometer
to track up to 100 components per reaction in each well, while
processing multiple conditions in 96-well plates. This is
accomplished by a pair of lasers, one to excite internal red
fluorescence for bead identification and count and the second to
excite green fluorescence at the bead surface to measure analyte
levels.
[0063] The platform allowed for high throughput and quantitative
analysis of multiple kinase activities in a single experiment,
using internal standard curves to accurately compare samples. This
embodiment of the invention, which uses an internal standard curve
to accurately measure systematic and intentional changes, provided
a quantitative assay for profiling tyrosine kinase activity in a
biological context. Usefulness of the invention is demonstrated by
unusually high accuracy measurements of multiple tyrosine kinase
activities in cell lysates from a single well of a 96-well plate,
using standard laboratory equipment.
[0064] The embodiment focuses on tyrosine kinases and directly
calculates the percentage of substrate phosphorylation via
non-linear regression from internal standard curves. The synthetic
kinase substrates Abltide, Srctide and a peptide derived from Btk
are immobilized on Luminex beads to facilitate handling procedures
and enable the analysis of more than one component per reaction.
Fluorescent antibodies are used to label substrate phosphorylation
sites.
[0065] Antibody fluorescence intensity is translated directly to
enzyme phosphotransferase activity through internal standards
included in each well of a 96-well plate. Serially diluted
inhibitors are applied to different wells of a 96-well plate to
measure the effect of variable inhibitor concentrations on kinase
activity. Because of the precise quality control implemented
through internal standards, the test data exhibit an extremely low
probability of error, with statistical significance at the 0.01
level.
[0066] The subject invention demonstrates substantially improved
information content with a measurement scale that provides equal
intervals for improved precision. As a result, data are analyzed by
statistical correlations to calibrated standards and not merely
ordered by rank. The measurement accuracy allows for standard
methods of statistical inference, such as the t-test for level of
confidence. By contrast, external standards are generally used for
assays based on the Luminex technology and lack individual
calibration for any well.
[0067] Typically, arrayed data fail to meet the standards required
for standard statistical methods. This lack of an established
baseline results in information content that is limited to rank
order sorting, with insufficient information content in individual
reactions to allow calculation of the mean, the standard deviation,
and resulting levels of confidence. With an external standard and
resulting accuracy limited to ordinal measurement, the information
content is limited to the use of a dimensionless figure of merit,
such as the Z' value, or alternative standards. Although
measurement precision is not feasible for some situations, and
alternative protocols have been developed to allow reasoned
decisions based on limited information content, improved precision
is widely recognized as the preferred foundation for test
results.
[0068] The application of the invention is this embodiment defines
the first reported use of well-specific internal standard curves to
calculate the accurate percentage of substrate phosphorylation
following reactions with kinases in cell lysates. The invention has
been used to reliably measure the simultaneous phosphorylation of
Abltide, Srctide and the peptide derived from Btk for an in-depth
view of intracellular network dynamics during treatment with
clinically relevant inhibitors. With high-throughput formatting and
requiring only hours for completion, this assay is expected to be a
valuable tool in clinical settings.
[0069] This embodiment improves the resolution of diagnostic tests
by providing statistical relevance to observed differences and
sample comparisons. The benefits of the invention are demonstrated
by an embodiment that can be routinely used for quantitative,
high-throughput screening of kinase inhibitors and is easily be
applicable in the clinic to assess CML patients undergoing
treatment.
EXAMPLE OF THE PREFERRED EMBODIMENT
[0070] The following is an example of a typical practical
application of the invention for measurement of enzyme activity by
the quantitative phosphorylation of peptide substrates. Expanded
application of the novel method would allow parallel applications
for circumstances that require accurate quantitative analysis and
assay robustness, such as diagnostic testing of patient samples,
modeled here using a human cell line for chronic myelogenous
leukemia, and pharmaceutical screens using a small set of
clinically-relevant inhibitors. Persons familiar with the art would
be aware of useful application of this basic method with other
laboratory equipment, test conditions, or test samples.
[0071] The method results in a quantitatively robust assay for the
functional analysis of intracellular signaling events, based on a
covalently immobilized set of synthetic peptide substrates on
fluorescent Luminex beads. Phosphorylation of the peptide
substrates by active kinases present in cell lysates is detected by
a phycoerythrin-labeled anti-phosphotyrosine antibody.
[0072] The Luminex system uses two orthogonal lasers to display
both internal bead fluorescence, which identifies the bead region
and counts the number of beads analyzed, and phycoerythrin
fluorescence at the bead surface, bound by interaction with the
phosphorylated substrate. Only phycoerythrin that is bound to a
bead surface is recorded. Results from Luminex assays are typically
displayed as the median fluorescence intensity, in arbitrary units,
per a minimum of 100 beads. This method reports the mean
fluorescent intensity of the total number of beads counted to allow
statistical analysis of the population results, providing robust
99% confidence intervals for each sample.
Substrate Immobilization on Luminex Beads
[0073] To measure the characteristic Bcr-Abl activity profile of
CML, the standard high-affinity peptide substrate for c-Abl and its
oncogenic relative Bcr-Abl (CEAIYAAPFAKKK) is synthesized. The
established core recognition sequence was modified only by the
inclusion of an amino-terminal cysteine, for specific covalent
attachment to Luminex beads.
[0074] To enable separation of the substrate from the reaction
components, the synthetic peptide on Luminex beads is immobilized.
Luminex beads are supplied in bulk with free carboxyl groups and
can be modified with primary amines using standard methods. To
provide distance between the bead surface and the site of
phosphorylation on the peptide substrate, a biologically
passivating N-(3-Aminopropyl)methacrylamide linker was introduced
using EDC/NHS crosslinking.
[0075] Abltide peptide substrate was covalently attached to the
bead surface by Michael addition of the sulfhydryl at its
amino-terminal cysteine to on-bead acryl groups. All conjugation
steps were carried out in filtered microcentrifuge tubes to enable
easy removal of excess reagents and wash steps. Modified beads were
counted using a hemacytometer and stored refrigerated for up to a
year in phosphate-buffered saline, pH 7.4.
Substrate Phosphorylation by Cell Lysates
[0076] The K-562 cell line was established from a CML patient in
terminal blast crisis and is characterized by highly
undifferentiated cells of the granulocytic series. With a low
frequency of the Philadelphia chromosome but highly up-regulated
Abl kinase, the K-562 line serves as an ideal model for testing CML
diagnostics. Kinase assays were performed in 96-well filter plates
to accommodate high throughput processing. A liquid handling robot
was used to efficiently transfer approximately 1000 beads to each
well of a 96-well filter plate. A 50 .mu.L/well reaction mixture,
containing kinase buffer (50 mM Tris-HCl pH 7.5, 10 mM MgCl2, 1 mM
EGTA, 0.01% Brij-35, 2 mM DTT, Complete protease inhibitor
cocktail), 10 .mu.M ATP and variable concentrations of purified
enzyme or cell lysates, were incubated with the beads for up to one
hour.
[0077] The kinase reaction was terminated by the addition of 250 mM
EDTA, pH 8.0, which chelates the cofactor MgCl.sub.2 that is
required for Bcr-Abl activity. A brief wash with 2% SDS was used to
remove non-specific adsorption of cellular proteins and 1% BSA was
used to block nonspecific binding of labeling antibodies to
un-phosphorylated peptides and the polystyrene bead surface.
Phosphorylated substrate was labeled sequentially with biotinylated
4G-10, an anti-phosphotyrosine antibody, and phycoerythrin-coupled
streptavidin. After phospho-specific labeling the beads are given
an optional final wash and re-suspended in the Luminex system
running buffer.
[0078] The Luminex system removes 50 .mu.L of the suspension and
queries a random sample of at least 100 beads out of the estimated
total 1000 beads per well. Peptide-modified Luminex beads are
phosphorylated by purified c-Abl kinase or by kinases in K-562
lysates, resulting in a change in phycoerythrin fluorescence from x
to y in one hour at 30 degrees Celsius. Experiments are performed
in triplicate wells. Unmodified carboxyl-coated Luminex beads
typically display a background fluorescence of 4 units.
Generation of Internal Standard Curves and Measure of Bcr-Abl
Activity
[0079] The standards are produced by mixing synthetic Abltide
(CEAIYAAPFAKKK) and synthetic phospho-Abltide (CEAI-pY-AAPFAKKK) in
known molar ratios and immobilizing those peptide mixtures on
Luminex beads. In order to ensure accurate relative concentrations,
purified peptides were analyzed separately by absorbance of the
peptide backbone at 214 nm with analytical reverse-phase
high-performance liquid chromatography (RP-HPLC). The integrated
peak areas were plotted versus injection volume per peptide and the
ratio of the slopes was used as the calibration factor for relative
peptide concentration.
[0080] The points at 0% and 100% substrate phosphorylation are
generated from pure synthetic Abltide and synthetic
phospho-Abltide, while the points at 25% and 50% are produced by
corresponding molar ratios of phospho-Abltide to Abltide. Pure
peptides and peptide mixtures were covalently conjugated to
region-specific Luminex beads by carbodiimide chemistry. Four
distinct bead regions, modified with standard phosphorylated
peptides, were added to each well of a 96-well plate after the
kinase reactions were quenched and the experimentally
phosphorylated beads were washed with 2% SDS. All five bead regions
were blocked with 1% BSA prior to anti-phosphotyrosine antibody
labeling.
[0081] The fluorescence readout from the well-specific internal
standard curves over an entire 96-well plate is not linear with
increasing phosphorylation. The slope and shape of the functional
relationship between percent phosphorylation and observed
florescence intensity is based on the test data and the calculation
of the means and confidence intervals for each cell, and for the
entire group of 96 cells. The sample size is the bead count for
each well, which ranged from 171 to 668, with a mean sample size of
326. The extent of experimentally phosphorylated substrate is
calculated by non-linear regression from the internal standard
curve.
DETAILS OF THE PREFERRED EMBODIMENT
[0082] The following is a description of the basic materials,
equipment, and test procedure that are a practical necessity to
assure accurate measurements for kinase activity.
Materials and Reagents
[0083] Reagents for peptide synthesis were purchased from Peptides
International (Louisville, Ky.). K-562 cells were obtained from
American Type Culture Collection (Manassas, Va.). RPMI-1640 media,
L-glutamine, and the Kaiser test kit were purchased from
Sigma-Aldrich (St. Louis, Mo.). FBS was purchased from Gemini
Bio-products (West Sacramento, Calif.). Phosphosafe Extraction
Reagent was purchased from Novagen EMD Biosciences (Madison, Wis.)
and Complete protease inhibitor cocktail was purchased from Roche
Diagnostics (Mannheim, Germany).
[0084] The Coomassie (Bradford) protein assay kit, HALT protease
inhibitor, NHS, and EDC were purchased from Pierce (Rockford,
Ill.). Imatinib and dasatinib were purchased from LC Laboratories
(Woburn, Mass.). Purified recombinant human Abl kinase (EC
2.7.10.2), biotin-conjugated anti-phosphotyrosine clone 4G10,
phycoerythrin-conjugated streptavidin, and 0.22 .mu.m-filtered
microcentrifuge tubes were purchased from Millipore (Billerica,
Mass.). N-(3-Aminopropyl)methacrylamide was purchased from
Polysciences (Warrington, Pa.). Luminex (Austen, Tex.) generously
provided Luminex beads with free carboxyl groups, in bead regions
27, 34, 42, 45, 56, 61, 65, and 73.
Instrumentation
[0085] Peptides were synthesized on a Prelude.TM. parallel peptide
synthesizer from Protein Technologies (Tucson, Ariz.), purified on
a Waters 6000S HPLC system (Milford, Mass.), and analyzed by
MALDI-TOF (4700 Voyager, Applied Biosystems). Modified Luminex
beads were distributed to filter plates using a Precision
Microplate Pipetting System purchased from BioTek (Winooski, Vt.).
Data were acquired with a minimum target of 100 bead counts per
region per well using the BioPlex 200 system from BioRad (Hercules,
Calif.), calibrated separately at both high (15993) and low (3515)
targets to determine the maximum range of detector linearity per
plate.
Cell Culture and Lysis
[0086] K-562 cells were cultured at 37.degree. C. and 5% CO.sub.2
in RPMI-1640 media with 4 mM L-glutamine and 10% FBS (v/v). Lysates
were prepared from confluent cells using Phosphosafe Extraction
Reagent with Complete protease inhibitor cocktail and tested for
total protein content by Bradford analysis.
Peptide Synthesis and Substrate Sequences.
[0087] Peptides were synthesized at the 40 .mu.mmol scale using a
5-fold excess of Fmoc amino acids (200 .mu.mol per coupling)
relative to Rink-Amide-CLEAR resin (87 mg at 0.47 mmol/g). Fmoc
protecting groups were removed with 20% piperidine in DMF for 20 m.
After 6 washes amino acids were coupled using 1:1:2 amino
acid/HCTU/NMM in DMF for 30-45 m. Phosphotyrosine was coupled in
HBTU/HoBt/DIPEA for 2 h, and complete coupling was confirmed by the
Kaiser test. Both the amino acid N-terminal to phosphotyrosine and
the final amino-terminal cysteine were coupled twice for 45 m each.
Peptides were cleaved from the resin with 94.5:2:2:1.5
TFA/water/EDT/TIS for 3 h, precipitated with diethyl ether,
re-suspended in 5% CH.sub.3CN and lyophilized. Crude peptides were
purified by HPLC using a preparative 10.times.250 mm 10 .mu.m
.sup.18C column. Both crude and purified peptides were analyzed by
MALDI-TOF in linear positive and negative modes using a 1:1 (v/v)
mixture of 10 mg/mL CHCA matrix in 75% CH.sub.3CN with 0.1%
TFA.
[0088] Established kinase substrate recognition sequences were
modified by the inclusion of an amino-terminal cysteine for
specific covalent attachment to Luminex beads. While Abltide
(CEAIYAAPFAKKK) (Songyang, Carraway et al. 1995) and Srctide
(CAEEEIYGEFEAKKKK) (Songyang, Carraway et al. 1995) are optimized
synthetic substrates, the peptide substrate for Btk kinase was
derived from its tyrosine auto-phosphorylation site
(CKKVVALYDYMPMN) (Bence, Ma et al. 1997; Yamadori, Baba et al.
1999).
Generation of Internal Standards
[0089] Internal standards were generated from synthetic Abltide and
phospho-Abltide (CEAI-pY-AAPFAKKK). To ensure accurate relative
concentrations between Abltide and phospho-Abltide for 15%, 25%,
30%, 45%, and 50% molar mixtures, purified synthetic peptides were
analyzed separately by absorbance of the peptide backbone at 214 nm
with analytical C.sub.18 RP-HPLC. Integrated peak areas were
plotted versus injection volumes per peptide and the ratio of the
slopes was used as the calibration factor for relative peptide
concentration. Pure Abltide and phospho-Abltide were used for 0%
and 100% phospho-standards.
Covalent Substrate Immobilization
[0090] Luminex beads were modified with primary amines using
standard methods. Up to 300 .mu.L of carboxylated beads, supplied
at 1.25.times.10.sup.7 beads/mL, were added to a filtered
microcentrifuge tube, washed with water by centrifugation at 100 g,
and re-suspended in 100 mM NaH.sub.2PO.sub.4, pH 6.2. 50 .mu.L of
50 mg/mL NHS in water and 50 .mu.L of 50 mg/mL EDC in water were
added and the beads were incubated at room temperature for 20 m
with gentle shaking (Giavedoni 2005). The beads were washed three
times with 100 mM MES, pH 5.0, and re-suspended in 100 .mu.M
N-(3-Aminopropyl)methacrylamide in the same buffer. The primary
coupling reaction was mixed for 2 h at room temperature. Beads were
washed three times with 100 mM NH.sub.4HCO.sub.3, pH 8.0, and
re-suspended in 100 .mu.M peptide in the same buffer. The secondary
coupling reaction was mixed for 1 h at room temperature and allowed
to incubate 12-18 h at 4.degree. C. Modified beads were counted
using a hemacytometer and stored at 4.degree. C. for up to a year
in PBS, pH 7.4, supplemented with phosphatase inhibitor as
necessary.
Kinase Assays
[0091] Peptide-conjugated beads were diluted to 1.25.times.10.sup.6
beads per mL in 10 mM Tris-HCl, pH 7.4 and each bead region was
distributed into one row of a black, conical bottom 96-well plate.
Using a pipetting robot, 5 .mu.L from each well per row were
distributed to each of the 12 rows of a 96-well filter plate. The
96-well filter plate, containing approximately 1000
peptide-modified beads per region per well, was vacuum-washed three
times with 10 mM Tris-HCl, pH 7.4 with 50 mM MgCl.sub.2. A 50 .mu.L
reaction mixture, containing kinase buffer (50 mM Tris-HCl, pH 7.5,
10 mM MgCl.sub.2, 1 mM EGTA, 0.01% Brij-35, 2 mM DTT, and 1.times.
Complete protease inhibitor), 10 .mu.M ATP (unless otherwise
specified), and variable concentrations of purified kinase or cell
lysates, was incubated with beads for up to 60 m.
[0092] Lysates prepared from cells distributed in sterile 96-well
filter plates were diluted approximately five-fold for activity
assays, while lysates prepared in conical tubes were diluted ten-
to fifty-fold. Kinase reactions were terminated by the addition of
250 mM EDTA, pH 8.0. Three 5 m washes with 2% SDS and 5 successive
washes with water were used to remove non-specific adsorption and
detergent. 1 h incubation with 1% BSA in Tris-buffered saline with
Tween-20 (TBST; 20 mM Tris base, 137 mM NaCl, 0.1% Tween-20, pH
7.6) was used to block non-specific binding of labeling
antibodies.
[0093] Phosphorylated substrate was labeled sequentially with a
1:1000 dilution of biotinylated 4G10 and a 1:1000 dilution of
phycoerythrin-coupled streptavidin in TBST. Beads were given a
final wash with TBST and re-suspended in the Luminex system running
buffer prior to analysis. All steps, including bead handling and
labeling, were performed in reduced lighting.
Statistical Analysis
[0094] Several parameters were recorded for each bead region
analyzed: the number of beads per region in the queried sample, the
median, the mean, and the standard deviation of bound phycoerythrin
per bead region. Data were reviewed using widely accepted methods
of statistical inference (Snedecor and Cochran 1989). The following
is the standard calculation procedure for the confidence interval
and the standard error (Huntsberger and Billingsley 1987). Tables
of t-values used in the calculation of the confidence interval were
verified against published data from standard sources (Owen 1965;
Snedecor and Cochran 1989).
Confidence interval = ( X - t .alpha. 2 , n - 1 ) .times. .sigma. n
##EQU00001## Standard error = .sigma. n ##EQU00001.2##
[0095] Wherein:
[0096] X=the mean fluorescence intensity per bead region per
well
[0097] t.sub..alpha./2=2-tailed t distribution, for a specified
level of confidence (.alpha.)
[0098] n-1=degrees of freedom (df), sampled bead count per region
per well minus one
[0099] .sigma.=standard deviation of the fluorescence intensity per
bead region per well
[0100] n=sample size, the number of beads sampled per bead region
per well
[0101] Because of variations in sample sizes, it was necessary to
calculate confidence intervals based on specific t-values for each
bead region in each well of a 96-well plate. The sampled number of
beads per region per well was often less than 200, resulting in
t-values that were substantially different from the normal
distribution. Therefore, separate t-values were derived based on
the sample size (n) for each bead region in each well, using
published extended values for the t-distribution with six
significant digits for df from 40 to 200 within one well (Owen
1965) and 4 significant digits for df from 500 to 10,000 over an
entire plate (Federighi 1959).
Non-Linear Regression
[0102] Well-specific standard curves were constructed from the
observed mean fluorescence intensity of known ratios of
synthetically phosphorylated Abltide. Prism v4.0a (GraphPad
Software, Inc., La Jolla, Calif., USA) was used to calculate the
goodness of fit to non-linear models, where the criterion for
selection was the minimum absolute sum of squares. For comparison,
the correlation coefficient, R.sup.2, was also noted. The
Boltzmann-sigmoidal model best fit all of the data from acquired
standard curves, with a calculated least squares of zero and an
R.sup.2 no less than 0.95. To calculate the effect of inhibitors on
the observed fluorescence intensity and the calculated percentages
of phosphorylation, sigmoidal curves (variable slope) provided an
excellent fit after log-transformation of x-axis values, with an
R.sup.2 of 0.89-0.99. For kinetic rate relationships, such as the
amount of enzyme units versus percent phosphorylation, or the lapse
of reaction time versus phosphorylation, the best fit was either
linear for short time scales or hyperbolic for extended
concentrations or durations.
CONCLUSIONS
[0103] The test protocol described herein results in a carefully
calibrated internal standard for each well in the array.
Significantly, this protocol allows substantially improved accuracy
for use of microsphere arrays, so that standard statistical methods
can be used to establish the level of confidence in the test
results. The usefulness of this invention is established by
providing a method that could impact human health by integrating
basic biology with clinical science. The subject novel method can
allow expansion of simultaneous high-accuracy quantitative analysis
to one hundred or more kinase activities in a single experiment.
The limitations as to number of kinase activities tested in a
single experiment depend on the number of wells in the array, the
number of available substrates, and the capacity of the equipment,
but not on the subject method.
[0104] Those skilled in the art will recognize, or be able to
ascertain using no more than routine experimentation, many
equivalents to the specific embodiments of the invention described
herein. Such equivalents are intended to be encompassed by the
following claims.
* * * * *