Note: Descriptions are shown in the official language in which they were submitted.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
1
METHOD OF PROCESSING AND TESTING POWDERED
SAMPLES USING IMMUNOCHROMATOGRAPHIC STRIP
TESTS
Field of the Invention
This relates to the field of immunology and more
specifically relates to a method for the analysis of a powdered
sample.
Background of the Invention
Many situations exist in which it is important to have rapid,
reliable, inexpensive tests that can be run by untrained individuals
in the field, home, or other non-laboratory setting. Immunoassays
have many of these characteristics. One type of immunoassay,
referred to as an immunochromatographic or lateral flow strip test,
has been successfully developed into 'one-step' tests and
employed for on-site analysis. The over-the-counter home
pregnancy test is an example of a simple, one-step
immunochromatographic strip test.
Immunoassays have been used to detect substances in many
different kinds of samples in many different markets, including the
agriculture market. Modern biotechnology methods are being used
to genetically modify plants. These genetically modified plants,
and the seeds, grain and food derived from them all may contain
novel or recombinant proteins. It is important to determine the
presence of such proteins for regulatory, environmental, safety,
and world trade issues. Very large amounts of grain and seed are
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
2
harvested, transported, mixed, stored, distributed and traded
throughout the world, and it is important to have tests that can
detect these novel proteins in order to address these issues. It is
desirable to have a rapid, simple and inexpensive method that can
be used to test grain samples at many points along the distribution
channel including trucks, elevators, barges, ships, etc. A strip test
is ideally suited for this purpose. For example, recently, an
immunochromatographic strip test (strip test or lateral flow device,
LFD) that detects a specific protein in genetically modified
1 o soybeans was developed.
Strip tests are comprised of multiple porous components,
membranes and filters, through which liquid sample is drawn by
capillary action. Analyte in the sample reacts with the test reagents
contained within the test strip as it traverses the length of the strip.
To detect an analyte (such as a protein or mycotoxin) in grain or
seed (e.g., corn, soybean, rice, wheat, etc.), it is necessary to grind
the grain into a powder and then extract the protein from the
powder with a liquid that is then separated from the solid material
and assayed using the test. To achieve the highest sensitivity test
possible, it is important to extract as much of the analyte from the
grain as feasible. This requires that the grain be ground to a fine
powder to facilitate efficient extraction. However, very fine
particulate matter suspended in the liquid sample clogs the pores
of the strip test components causing the strip to run slowly,
erratically or not at all. Additional steps or test components
employed to remove the particulates only serve to complicate the
testing process and increase the cost of the test and time-to-result.
What is needed is a means for preparing and testing a finely
ground powdered sample that maximizes extraction efficiency,
3o sensitivity, speed and reliability of the strip test while minimizing
the complexity and cost.
Summary of the InvQntion
A method and kit for processing and testing powdered
samples are provided herein. The powdered sample is combined
with a fluid to produce a slurry or liquid suspension, and the slurry
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
3
is contacted with an immunochromatographic test strip containing
detectable reagents that are immunoreactive with an analyte to be
detected in the sample. The method is particularly useful for the
analysis of agricultural products, crops, or food fractions, such as
the identification of recombinant grains.
It is therefore an object of the present invention to provide a
rapid, reliable, and inexpensive method for analyzing an analyte in
powdered samples using an immunoassay.
It is a further object of the present invention to provide an
to immunoassay method for the detection of analyte in a sample that
is simple, portable and user-friendly and can be utilized
successfully by non-scientific personnel in the laboratory and the
field.
It is a further object of the present invention to provide an
immunoassay device and method for the detection of analyte in a
sample without requiring precise sample or reagent measurements.
It is a further object of the present invention to provide an
immunoassay method for the detection of analyte in a powdered
sample that provides results that are accurate and reproducible.
2o These and other objects of the present invention will
become apparent after reading the following detailed description
of the disclosed embodiments and the appended claims.
Brief Description of the Drawings
Figure 1 is a table showing chromatographic strip test
results for the analysis of corn kernels using the method described
herein.
Detailed Description of the Disclosed Embodiments
A method for the detection of analyte in a sample using an
immunochromatographic strip test is described herein. In
accordance with the method, the sample is provided in powdered
form or is processed from a solid or semi-solid sample such as a
slice, granule, seed, grain, particle, or the like, to produce a
powder. The powdered sample is then mixed with a fluid, such as
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
4
an aqueous solution, to form a slurry or liquid suspension. The
slurry is then contacted with an immunochromatographic strip.
The method and kit are useful for the detection of a wide
variety of analytes including, but not limited to, agricultural
products, biological and environmental contaminants or additives,
industrial chemicals, toxins (particularly mycotoxins), and
biological analytes, such as antigenic determinants of proteins,
polysaccharides, glycoproteins, lipoproteins, nucleic acids and
hormones, of organisms such as viruses, bacteria, fungi, parasites,
plants and animals, including large and small molecules, polymers
and haptens. The method is particularly useful for the
identification or quantification of proteins or peptides in
genetically modified plants or recombinant agricultural products,
such as grains. For example, the method is useful for the detection
of the Bacillus thuringiensis CrylAb or CrylAc proteins in
genetically modified plants, such as corn, as described below in
the examples.
Definitions
The terms "a", "an" and "the" as used herein are defined to
mean "one or more" and include the plural unless the context is
inappropriate.
The term "analyte" refers to a drug, hormone, chemical,
toxin, compound, receptor, nucleic acid molecule or other
molecule and fragments thereof to be measured by the method
described herein.
Sample Processing Method Development
During the development of an immunochromatographic
strip test to detect the Bacillus thuringiensis CrylAb protein in
genetically enhanced corn kernels, great difficulty was
3o encountered in achieving the sensitivity and time-to-result
specifications that are required of the commercial product. Data
demonstrated that increasing grinding time of the corn kernels
increased extraction efficiency and test sensitivity. However, the
more finely ground particles in the liquid sample caused the test to
3s flow too slowly.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
To increase sensitivity, less liquid was added to the ground
powder to minimize the dilution of the analyte in the sample, but
this resulted in even more particulate matter in the liquid sample
and poorer performing tests. In fact, in some experiments, the
s amount of liquid added to the powdered corn was reduced to the
point that there was only enough liquid to suspend the powder but
no remaining liquid to pull off and test. At this point, lacking any
liquid sample to test, the end of the strip test was inserted directly
into the wet, powdered corn slurry. The expectation was that the
thick slurry would clog the strip, but, surprisingly, the strip flowed
significantly faster than when tested with liquid pulled from the
sample. Further experimentation demonstrated that the sensitivity
and reliability of the test were also significantly improved and
addition of the strip directly to the powdered slurry eliminated the
step of transferring the liquid sample to a separate vessel, thereby
reducing the number of components and complexity of the test.
Sample Processing
A sample is collected or obtained using methods well
known to those skilled in the art. The sample containing the
2o analyte to be detected may be obtained from any biological or
environmental source. For example, the sample may be any plant
tissue or extract including root, stem, leaf, or seed. The sample
may be diluted, purified, concentrated, filtered, or otherwise
manipulated prior to being mixed with the fluid to create a slurry
or liquid suspension to optimize the assay results.
Analytes to be detected using the immunoassay method
described herein include, but are not limited to, the following
analytes: molecules, such as organic and inorganic molecules,
peptides, proteins, glycoproteins, carbohydrates, nucleic acids,
lipids, toxins, and the like. Analytes also include but are not
limited to neurotransmitters, hormones, growth factors,
antineoplastic agents, cytokines, monokines, lymphokines,
nutrients, enzymes, receptors, antibacterial agents, antiviral agents,
and antifungal agents. The term analyte also means detectable
components of structured elements such as cells, including all
animal and plant cells, and microorganisms, such as fungi, viruses,
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
6
bacteria including but not limited to all gram positive and gram
negative bacteria, and protozoa. The term analyte also means
detectable components of organelles and cells.
The sample is ground, homogenized, chopped, blended, or
otherwise manipulated to create a powder. The size and texture of
the powder or fine granules can be adjusted by modifying the
technique or increasing the amount of manipulation time, such as
grinding or blending, as needed.
The fluid with which the powdered sample is combined or
mixed is one capable of forming a slurry or liquid suspension. The
preferred fluid is an aqueous solution, such as a buffer that is
compatible with the immunochromatographic strip.
Immunochromato~raphic Strip
The immunochromatographic strip, or detection membrane,
is a membrane or strip having reagents deposited in zones along
the longitudinal length of the membrane. Immunochromatographic
strips are readily available from commercial suppliers or can be
customized by laboratory personnel skilled in the art, or a
commercial immunodiagnostic supplier, to include
immunoreagents specific for the analyte to be detected. The
immunoassay reactions conducted on the strip may be in
competitive or sandwich immunoassay formats
The reagents suspended or immobilized on the membrane
provide means for detecting analyte, preferably by visually
detecting a labeled substance or substances, such as colloidal gold,
that have been bound to analyte. Alternatively, the reagents may
detect the absence of labeled substance, and the label may be
detected using instrumentation know to those skilled in the art
such as a spectrophotometer or fluorescence detector. The reagents
on the membrane may be immobilized or may be diffusible but
contained on the membrane in a solid or semi-solid state that,
when contacted with the sample, becomes mobile and moves with
the sample toward the distal end of the membrane. Reagents may
also be included on the immunochromatographic strip to enhance
or clarify the signal produced by the label being detected.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
7
Additional reage~its are optionally incorporated in zones on the
detection membrane for calibration.
The membrane is preferably a non-woven substrate upon
which the reagents can be immobilized or deposited, and which is
capable of conveying sample in a fluid flow direction generally
parallel to the longitudinal length of the chromatographic strip.
Desirable chromatographic strips are composed of a fluid-
conducting material including, but not limited to, nylon,
polyethylene, glass fiber, nitrocellulose, cellulose, and other
l0 common membrane matrices or bibulous materials. The preferred
chromatographic strip is composed of nitrocellulose. The
membrane of the chromatographic strip is optionally backed with,
or laminated to, another material. Desirable backing or laminating
material is polyethylene or vinyl, although other suitable materials
known in the art may be used.
A reagent sink is optionally included at the distal end of the
immunochromatographic membrane or strip for enhancing the
flow of the fluids, including reagents and sample, along the
longitudinal length of the strip. The sink may be composed of an
absorbent material, such as blotting paper.
Optionally included in flow communication with the distal
end of the strip is an end of test indicator for indicating completion
of the assay. The end of test indicator may be included in the
reagent sink described above or may be a separate component
located at the distal end of the immunochromatographic strip or
reagent sink, if present.
The end of test indicator is preferably composed of an
absorbent material containing an indicator, or dye, that travels
with the liquid to indicate that the liquid has traveled to the distal
3o end of the strip and that the strip is ready for analysis. Analysis
can be conducted either by visual inspection or with the aid of
instrumentation. Preferably, movement of the dye through the
absorbent material to a predetermined location is visually
detectable.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
8
Kit
An immunoassay kit for the detection of analyte in a sample
provided herein contains a fluid, such as a buffer to be combined
with the powdered sample to form a slurry or liquid suspension,
and an immunochromatographic strip or membrane. The kit may
optionally contain one or more antibodies or other reagents to
combine with the sample prior to contacting the sample slurry with
the membrane.
The kit may additionally contain additional reagents or
l0 buffers, equipment for obtaining or collecting the sample, a vessel
for containing the sample and reagents, a timing means, and a
colorimeter, reflectometer, or standard against which a color
change may be measured. A simple, inexpensive reflectometer is
preferred.
The reagents, including any antibody reagents may be
lyophilized, in a single vessel or in individual vessels. Addition of
the slurry sample to the vessel results in solubilization of the
lyophilized reagents, causing them to react.
2o The method and kit as described above will be further
understood with reference to the following non-limiting examples.
Example 1
Genetically Modified Corn Sample Preparation
and Analysis
This example illustrates the use of the method described
herein to detect the CrylAb protein in a genetically modified,
insect-resistant corn sample.
Each sample was weighed to obtain the proper number of
corn kernels. Samples containing 60 kernels weighed 18.4 g and
samples containing 125 kernels weighed 39.3 g.
Each sample was placed in a clean and dry glass, 4 oz.
MASONTM jar.
A 70 mm blender base and blade assembly (Eberbach Co.,
Ann Arbor, MI) was attached. The MASONT'~' jar was placed onto
the laboratory grade blender (blaring Model 31 BL91 ) and covered
with a safety shield. The sample was ground on high speed for 30
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
9
seconds to produce powdered corn. The safety shield, adapter and
blade were removed.
The following specified amounts of buffer (Trait~M, also
referred to as "Traitcheck", sample buffer, Strategic Diagnostics
Inc., Newark, DE) was added to the powdered corn directly in the
MASONTM jar. For the test tube procedure, 60 kernel samples
were combined with 18.0 mL of buffer. The 125 kernel samples
required 40.0 mL of buffer to produce a slurry or liquid
suspension. For samples run directly in the MASONTM jar, 60
kernel samples used 16.0 mL of buffer, and 125 kernel samples
required only 31.0 mL of buffer.
The MASONTM jar lid was attached to the MASONTM jar
and the contents shaken for approximately 30 seconds, until all the
ground corn was evenly and thoroughly wet.
For samples run in the test tube, the MASONTM jar lid was
removed, and the sample allowed to settle approximately 30
seconds. A 500 ~,L aliquot of extract was pipetted off and placed
in a test tube. A chromatographic test strip, upon which was
immobilized antibody having immunospecificity for the C'ryl Ab
protein of Bacillus thuringiensis, was inserted into the extract until
it reached the bottom of the test tube.
For samples run directly in the MASONTM jar, the
MASONTn'' jar lid was removed and the chromatographic strip
inserted into the corn mixture. The strip was immersed until the
check mark (~ on the filter cover was reached (approximately 16
mm).
Each strip was allowed to run for 10 minutes at room
temperature. Various readings were taken, including time until
control line was visible, time until test line was visible, times for
3o control and test, readings at 3 minutes and 10 minutes, and time
until the reagents reached the sink.
The results are shown in Figure 1. A detailed user guide
describing the products and procedures for analysis of the Cry 1 Ab
protein in a genetically modified corn sample is provided in
Example 2, below.
CA 02394196 2002-06-12
WO 01/44779 PCTNS00/33774
Example 2
Lateral Flow Test Kit and Procedure for the Detection of
Cryl Ab Protein
This example a detailed procedure and kit to detect a
5 CrylAb protein in a genetically modified, insect-resistant corn
sample.
Kit Description
A kit is used to detect the CrylAb protein produced by a
gene derived from Bacillus thuringiensis. This gene has been
10 incorporated into insect-resistant corn including YIELDGARDT""
brands from both Monsanto and Novartis. The intended uses of the
kits include the qualitative (yes/no) determination of the Cry 1 Ab
protein in plant leaf and seed tissue and corn grain samples. The
lateral flow strips and other components provided in the kit are
sufficient to make qualitative determinations for the presence of
absence of the Cry 1 Ab protein in both field and laboratory
environments. Different application protocols are required for le~.t~,
seed and bulk grain determinations.
Principle of the Assay
2o The assay uses a double antibody sandwich format.
Antibodies specific to the CrylAb protein are coupled to a color
reagent and incorporated into the lateral flow strip. When the
lateral flow strip is placed in a small amount of an extract from
plant tissue that contains Cry 1 Ab protein, binding occurs between
the coupled antibody and the protein. A sandwich is formed with
some, but not all the antibody that is coupled to the color reagent.
The membrane contains two capture zone, one captures the bound
CrylAb protein and the other captures unreacted antibodies
coupled to the color reagent. These capture zones display a reddish
color when the sandwich and/or unreacted colored reagents are
captured in the specific zones on the membrane. The presence of
only one line (control line) on the membrane indicates a negative
sample and the presence of two lines indicates a positive sample.
Contents of Kit
100 Bacillus thuringiensis (Btl) chromatographic Test Strips
100 or more Sample cups
100 or more Wooden spatulas
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
11
1 User guide
Material Require. but not Supplied
Trait~M, also referred to as "Traitcheck", sample buffer (Strategic
Diagnostics Inc., Newark, DE)
Laboratory grade blender (Waring Model 31 BL91 recommended)
Waring adapter for MASONTM-type jars
MASONTM 4 oz. Blender jars*
Graduated cylinder, 50 ml (60 ml)
*Caution: A shield should be used over the blender jars while
l0 grinding. Safety glasses should be used.
Preparation and Storage of Reagents
Trait~M sample extraction buffer for grain is shipped as a
concentrate. Follow the procedure below to prepare the product to
be used for sample extraction:
1. Pour the contents of a one liter bottle of Trait~M Sample Buffer
Concentrate into a 5-8 liter carboy or other suitable container.
2. Add four liters of water to the sample buffer. Tap water may be
used.
3. Mix well and label as Trait~M Corn Sample Buffer. Label
buffer expiration as six months from date of preparation.
The Trait~M Corn Sample Buffer may be stored at room
temperature.
Btl Corn Grain Test Kit: User Guide
The Lateral Flow Test Kit should be stored at room
temperature. The Bt 1 Test Strips must be kept in the foil pouch
with desiccant. Storage conditions higher than room temperature
may adversely affect performance.
Sampling
The samples used for the Bt1 Corn Grain Test Kit can be
sub-samples of those "representative samples" collected from
trucks, railcars, barges, etc. for other tests. The size of the sub-
samples to be used for the test will depend on the level of
genetically modified corn kernels at which the screening is being
conducted and an acceptable level of risk that the genetically
modified level is close to the screening level. The number and size
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
12
of the sub-samples is discussed in more detail in the application
protocol section below.
It is assumed that the samples collected are representative of
the contents of the truck or container and are sufficiently mixed to
contain a random distribution of the sample contents.
Sample Preparation: Wei hind the Sample
The statistical sampling plan (see Principle of the Screening
Application) is dependent on the number of corn kernels used. It is
more practical to weigh corn kernels to obtain the correct sub-
sample size, however, the number of corn kernels in a unit weight
depends on the variety of corn and environmental conditions.
The table below is a guideline for the weight to corn kernel
number ration for seed corn.
Average Weight of Seeds (Iowa State University)
Seed Size Average Grams/Seed
Verv small 0.188
Small 0.219
Medium 0.257
2o Large 0.3 I 6
However, it is recommended that the ratio for each variety
be determined as follows:
1. Count 100 corn kernels of the variety to be tested.
2. Weigh the 100 kernels to the nearest 0.01 gram
3. Divide the weight of the corn kernels by 100 to get the average
grams per corn kernel (the result will be less than one gram per
corn kernel).
4. Multiply this average weight by the desired number of corn
kernels in the sub-sample to determine the weight for the sub
samples.
5. Construct a table for each variety for the different sub-sample
sizes to be used.
Example: One hundred corn kernels of Variety X weigh 35.00
grams. Each corn kernel then weighs 0.35 grams. Multiply the
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
13
0.35-gram per corn kernel times the number of corn kernels in
each sample size to get the following table:
Example of Weight-to-Corn kernel Ratio Table
Variety X:Grams per Sample
Corn kernels 40 60 70
125
Weight (g) 14.0 21.00 24.5 43.75
Note: The National Corn Growers Association (web site) states
that typical corn is 56 lbs. For a bushel of about 72,800 kernels or
0.350 g/kernel.
This average weight is then used to determine the weight of
corn kernels that will be used to represent the number of corn
kernels (for this variety) selected in the Probability Tables.
Sample Preparation: Processing the Sample
The preparation of the sample is very important for the
proper function of the test, especially the ratio of extraction buffer
to the weight of the corn sample. The following table contains
2o recommendations for the amount of Trait~M Corn Sample Buffer
for extraction based on the sample weight. A 4 oz MASONTM jar
and a grinding time of 15 seconds on high speed is used for all the
sample sizes (40, 60, 70, and 125 kernels) used in this example.
The processing parameters were determined using the laboratory
grade Waring Model 31 BL91 food processor. Other food
processors or extreme sized corn kernels will require further
validation.
Parameters for Prenarin~ Samples
3o Weight Range Buffer Volume
~ r~ ams) (ml)*
5-10 10
13-21 20
22-29 25
30-39 35
40-48 45
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
14
Note: the buffer volume is very important and should be measured
carefully.
1. Weigh two sub-samples from each truck or container.
2. Place each sub-sample in a dry MASONTM jar.
3. Attach the jar adapter and dry cutting blades.
4. Place the jar onto the food processor, place a shield over the jar
and grind the sub-samples for 15 seconds on high speed. (It is very
important to shield the jars since breakage can occur.)
l0 5. Remove the adapter and cutting blades.
6. Add Trait~M Corn Sample Buffer, prepared as described above,
t the ground corn kernels in the jar, place a lid on the jar and shake
the jar until all the ground corn kernels are well wetted (about 10-
20 seconds)
7. Use this sample extract for the Test Procedure.
Note: it is important to clean and dry the jars and cutting blades
between samples.
Eguipment Cleaning and D .rying
1. The MASONTM jar should be emptied, rinsed well with water
and completely dried with a paper towel between uses.
2. The cutting blades for the blender should be rinsed, wiped
clean, sprayed or rinsed with methanol and dried with a paper
towel between uses.
Test Procedure
1. With a wooden spatula from the kit, transfer sufficient sample to
the sample cup to fill about level full and smooth the top surface
of the sample.
2. Place one Btl Test Strip into the sample cup with the arrows
pointing into the cup. The test strip should just touch or be very
3o close to the bottom of the sample cup. Let sit until the control line
(top line) is clearly visible then determine if the test line is visible.
This may require 5-8 minutes.
3. The appearance of a single line (control) near the top of the strip
indicates a negative result.
4. The appearance of two lines on the strip indicates a positive
result.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
Reading the Lateral Flow Test Strip
1. Insert the labeled filter cover of the test strip into the sample
cup containing the wet corn sample. The arrows on the filter cover
should point into the cup. The test trip contains a test strip top
5 (reservoir pad where sample is contacted), a result window below
the test strip top, and a test strip bottom (labeled filter cover).
2. Check the result window frequently after adding the strip. At
least one line, the control line, should always develop
approximately 1 cm down from the reservoir pad. The control line
to contains an antibody specifically immunoreactive with a common
corn protein, but not a Bt protein. A red line in this position
indicates that the device is functioning properly. A red line
appearing below the control line is the test line, which contains
monoclonal antibodies specifically immunoreactive with the Bt
15 protein, and, if visible, indicates a positive results. If the test strip
displays two red lines, the test is complete and the sample is
positive for CrylAb Bt corn kernels. 1f, after approximately eight
minutes, the test strip only shows a clearly visible control line,
then the sample is negative for CrylAb Bt corn.
Principle of the Screening_Method
The test strip contains immobilized immunologic (antibody)
reagents that react with the recombinant Bt proteins present in
genetically modified corn that has been engineered to include one
or more genes for Bt proteins. As the sample migrates down the
chromatographic strip, it reacts with the immobilized antibodies
and generates a visible, colorimetric signal, thereby providing a
yes/no answer for the presence or absence of CrylAb Bt corn
kernels in a given sample. One red line at the top of the test strip
indicates a negative result and two red lines, the second line lower
on the test strip, indicates a positive result for the presence of
CrylAb corn kernels. Testing two statistically selected sub-
samples allows an estimate of the percent of Cry 1 Ab Bt corn
kernels. Te test results can provide information about the
probability of the percent in the representative sample being
within certain ranges. (Note: this test protocol will not specifically
determine the percent of genetically modified corn kernels.)
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
16
The statistical model used for this method is the Poisson
Probability Distribution which provides good approximations to
binomial (yes/no) probabilities when the number of items tested
(i.e. corn kernels) is large but the probability of a positive result is
expected to be small (i.e. low level of genetically modified corn
kernels). Using the Poisson Probability Distribution, it is possible
to determine the probability of having zero genetically modified
corn kernels in a random sample of a given size (number of corn
kernels) at a given percent genetically modified corn. For example,
to a random sub-sample of 100 corn kernels selected from a larger
sample at one percent genetically modified corn has a 36.8%
probability of containing no genetically modified corn kernels.
The probability of a 75 corn kernel sub-sample (at one percent
genetically modified) containing zero genetically modified corn
kernels is 47.2%.
Alternatively, for a 100 corn kernel sub-sample of unknown
percent genetically modified, a single negative: test result provides
the information that the representative sample has a 60.7%
probability of having 0.5% or less genetically modified corn
kernels, 36.8% probability of 1 % or less genetically modified
corn kernels and 13.5% probability of having 2% or less
genetically modified corn kernels.
Using two sub-samples instead of one increases the
confidence of estimating the genetically modified range. The
application protocols developed for screening corn by using this
test kit use two sub-samples of equal number of corn kernels. The
size of the two sub-samples is determined by the desired screening
level and the level of risk tolerance. If screening at 1 % genetically
modified level or higher, a passing results is achieved if one or
both sub-samples are negative for CrylAb Bt corn kernels. If
screening at 0.5% (or lower) the decision criterion is changed so
that both sub-samples must be negative to pass the sample.
The following probability tables provide guidelines for
selection of the two sub-sample sizes for screening at different
genetically modified (GM) corn kernel levels.
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
17
Probability Tables
Screening at 3.0% genetically modified (GM) corn:
Pass is one or two negative results
Probability of sample passing (%)
GM% 1.0 2.0 3.0 4.0 5.0 7.5
30 kernels 83.3 77.6 64.8 51.2 39.7 20.0
40 kernels 89.1 69.7 51.2 36.6 25.2 9.7
50 kernels 84.5 60.0 39.7 25.2 15.7 4.7
l0 Screening at 2.0% genetically modified (GM) corn:
Pass is one or two negative results
Probability of sample passing (%)
GM% 0.5 1.0 2.0 3.0 4.0 5.0
50 kernels 95.1 84.5 60.0 39.6 25.2 15.7
60 kernels 93.3 79.6 51.2 30.3 17.3 9.7
75 kernels 90.2 72.2 39.6 20.0 9.7 4.6
Screening at 1.0% genetically modified (GM) corn:
Pass is one or
two negative results
Probability of sample %)
passing (
GM% 0.25 0.5 1.0 2.0 3.0 4.0
100 kernels 95.1 84.5 60.0 25.2 9.7 3.6
125 kernels 92.8 78.4 49.1 15.7 4.6 1.3
Screening at
0.5% genetically
modified (GM)
corn:
Pass is two negative
results
Probability of sample %)
passing (
GM% 0.1 0.25 0.5 1.0 2.0 3.0
50 kernels 90.5 77.9 60.7 36.8 13.5 5.0
70 kernels 86.9 70.5 49.7 24.7 6.1 1.5
90 kernels 83.5 63.8 40.7 16.5 2.7 0.5
Note: the 0.5% screening level requires both of the two sub-
samples to be negative to pass the sample.
Interpretation of Test Results
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
18
Using the Probability Table for the 2% genetically modified
corn kernel screening level as an example, the recommended size
of the two sub-samples is 60 corn kernels. The two sub-samples of
60 corn kernels each are tested. The test can determine if either
zero corn kernels (negative) or one or more genetically modified
corn kernels (positive) are in each sub-sample. The test cannot
determine if a positive sub-sample has only one genetically
modified corn kernel or several genetically modified corn kernels.
The "fail" criterion is when both sub-samples are positive.
Using two sub-samples of 60 corn kernels with the above
decision criteria, samples at the following genetically modified
corn kernel levels will have the indicated statistical probabilities
of passing the test:
GM Corn Kernels%) Probability of Pass (%)
0.5 93.3
1.0 79.6
2.0 51.2
3.0 30.3
5.0 9.7
A 60-corn kernel sub-sample for a 2% screening level has
the risks associated with the above statistical probabilities. If a
lower risk of accepting a "more contaminated" sample is desired,
then a 75-corn kernel sub-sample can be chosen for the 2% screen.
With this size for the two sub-samples, the probability of passing a
5% contaminated sample, for example, is reduced to 4.6%.
However, probabilities of passing "good" samples (i.e. below 2%
genetically modified) are also reduced. Likewise, if more risk can
be tolerated, 50-corn kernel sub-samples used with this decision
criterion will pass 60% of samples at 2% and 15.7% at 5%.
Screening at below 1 % genetically modified screenin lgevel
Because of the sensitivity of the strip test described in this
example, screening at levels lower than 10.% genetically modified
corn requires different criteria. The decision criteria are that a
sample passes if both sub-samples results are negative for
CA 02394196 2002-06-12
WO 01/44779 PCT/US00/33774
l9
genetically modifif~d corn. The sample will fail if one or both sub
samples are positive. Using two 70-kernel sub-samples and the
criteria that two negative results is a "pass", the probability of
passing a trick at 0.5% is 49.7%, at 0.25% is 70.5% and at 1 % is
24.7%.
The use of two sub-samples is recommended for screening
at all genetically modified levels. In all cases, the recommended
sub-sample size has been chosen to provide about a 50/50
probability of passing a representative sample at the screening
to level. However, by making changes to the sub-sample size, those
probabilities can be changed.
Modifications and variations of the present method and kit
will be obvious to those skilled in the art from the foregoing
detailed description. Such modifications and variations are
intended to come within the scope of the appended claims.
