Note: Descriptions are shown in the official language in which they were submitted.
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
SELECTIYB CELL ANALYSIS
FIELD OF THE INVENTION
This invention relates to devices and methods used in
selectively separating cells of varying types and
characteristics from each other for analysis and particularly to
such methods and devices used with laser scanning cytometers.
BACKGROUND OF THE INVENTION
Many important cell based assays for various medical
conditions have been developed through the use of flow
cytometry. Flow cytometry instruments measure multiple optical
properties of cells as they are made to flow in a cuvette in
single file through a light beam, often a laser. As the cell
interacts with the light, both fluorescent and scatter emissions
are given off that can be measured by photo sensors. If the
cells are appropriately reacted prior to measurement, with
fluorescent dyes that bind to specific cellular constituents,
the amount of fluorescence measured can be directly related to
each of a specific cellular constituent, such as DNA, RNA or
specific proteins. These dyes can be conjugated by well
established techniques to antibodies which can in turn bind to
specific proteins in the cell. The resulting fluorescence,
measured by a flow cytometer's sensors, can be used to quantify
the amount of a specific protein per cell or to differentially
count one or more types of cell in a heterogeneous cell
population such as blood. Additionally, fluorescence dyes can
be attached to specific nucleic acid sequences, which in turn,
will bind to specific DNA or RNA sequences within the cell's
nucleic acids to mark cells with specific genotypes.
This latter technique, because preparations are difficult
to perform on cells in suspension (as required in flow
cytometry), can be more advantageously be done by the more
recently developed laser scanning cytometry (Described in U.S.
Patent No. 5,107,422), currently a laser scanning cytometer is
commercially available from CompuCyte Corporation, Cambridge,
MA. under the trademark LSC~, which is designed for measuring
CA 02321203 2000-08-18
WO 99!45094 PGT/US99/04415
cells prepared and stationarily positioned on a microscope slide
surf ace .
Although flow cytometry technology has been widely used in
biomedical research, it has not been widely applied to clinical
use, with the exceptions of diagnosis and monitoring of leukemia
and AIDS. Laser scanning cytometry, with its stationary
samples, is more appropriate for use in all other clinical
settings. In laser scanning cytometry, fluorescence and scatter
is measured in cells placed on a solid substrate such as a
microscope slide. Prepared cells (as described above with
respect to flow cytometry), are measured by moving a laser beam
over the substrate and moving the substrate (without separate
sample movement) under computer control, to find the cells on
the substrate and to measure their fluorescence and scatter
emissions. Data obtained by scanning cytometry has been shown to
be equivalent to that of flow cytometry but with greater
efficiencies of use and with less margin for operator error (see
e.g. Kamentsky et al Slide-Based Laser Scanning Cytometry, Acta
Cytologica 41:123-143 (1997)). In addition, laser scanning
cytometry has been shown to have a number of benefits not
available with use of flow cytometry.
First, heterogeneous populations of cells on a slide can be
scanned, multiple properties of the cells measured, and the
instrument under user direction can relocate cells with specific
defined sets of properties for visual observation or. photography
or to combine data from multiple assays using the location of
each cell as the data merge key.
Second, since each cell is scanned with many data elements,
constituents of cells can be localized to cell compartments such
as the nucleus or cytoplasm and tests such as in-situ
hybridizations, in which fluorescent spots are counted, can be
performed.
Third, since the cells are measured on the surface of a
substrate such as a slide, certain assays such as in-situ
hybridizations are possible because they are difficult to do vn
2
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
suspended cells because cells do not remain intact they are lost
due to the multiple centrifugations required.
It is also possible to perform complex procedural steps on
the cells without separating the cells from the medium in which
they are presently contained. This is presently done primarily
by centrifugation to bring down the cells so that the
supernatant can be poured off. For users of flow cytometry,
heterogeneous populations of cells such as blood must be
pretreated prior to analysis by separating specific cell types
by centrifugations, columns, settling or lysis. The cells must
then be reacted with reagents such as antibody conjugated
fluorochromes, and the cells must be washed between and after
these reactions using laboratory centrifuges. As a result, flow
cytometry has not been adopted for many useful diagnostic and
prognostic applications because the preparative procedures are
complex, requiring highly trained personnel to perform and
control the various protocols and run the instrument.
Furthermore, for a number of applications, these tests must be
performed in emergency rooms, in operating theaters, near
neonatal units or in physician's offices where highly trained
operators are not available and yet results are needed in
minutes to determine patient treatment. Laser scanning cytometry
obviates many of such problems.
There are however, limitations unique to the use of laser
scanning cytometers. For example, an important characteristic
of such use is that the time required to measure a specimen is
directly proportional to the area scanned. Thus, if 10,000
cells are placed over a 10 millimeter square area it will take
ten times as long to process the specimen than if it were placed
in a one millimeter square area. Because of present computer
speeds, data processing is so rapid that the time to process a
specimen is primarily controlled and limited by the time it
takes to scan the laser beam over the specimen area. However,
in many assays, including all of those that use blood as the
specimen, the specimen is heterogeneous, with the cells of
3
CA 02321203 2000-08-18
WO 99/45094 PCTNS99/04415
normal interest for the assay being only a small fraction of the
total specimen's cells. For example there are normally 1000
times as many red blood cells as each specific type of white
blood cell. The presence of irrelevant cells therefore results
in increased processing time to analyze a given number of cells
of interest, if irrelevant cells must also be measured and
analyzed, since a larger area is required to place all the cells
on a surface without overlap of the cells. An expedient of
simply placing the entire sample in a desirably small scan area
will however result in multiple layers of irrelevant cells being
present, with overlaying of the cells of interest, thereby
actually interfering with the measurement process itself. It is
thus highly advantageous if only cells of interest remain in the
small area to be scanned and irrelevant cells are removed prior
to measurement.
In prior art methods, specifically selected analytes are
made to adhere to a surface by capture of the analyte on a
surface by means of a reagent which binds thereto, and
thereafter the quantity of the analyte itself is measured (see
U.S. Patent No. 5,637,469). However, often the applicable
reagent required to cause cell adherence will not always result
in the proper subset (i.e., cells of interest) adhering to a
surface without irrelevant cells.
sY of THE =NVErrr=orr
It is therefore an object of the present invention to
provide a means for increasing the efficiency of laser scanning
cytometry by limiting the scanning area of a heterogeneous cell
specimen, with removal of irrelevant cells and the like from a
designated limited area, and rendering cells of interest in the
limited area statistically sufficient to provide meaningful
measurements.
It is another object of the present invention to separate
from a heterogeneous specimen, a set of cells containing the
cells of interest, by causing the set of cells to be located in
the limited area, thereafter discriminating between a subset of
4
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
cells of interest and other cells of the set, not of interest,
and then performing measurements only on the subset of cells of
interest. Hereinafter, reference to "cells of interest" includes
a "cell set (e. g., red blood cells, white blood cells,
microorganisms, etc.) containing a subset of cells of interest"
(e. g., monocytes) unless otherwise indicated.
It is another object of the present invention to identify a
subset of cells of interest from among cells adhering to a
limited area of a surface by a marker reagent which can be
measured by light sensors of a laser scanning cytometer.
It is yet another object of the present invention to use
one or more additional marker reagents that will identify
characteristics of cells of interest which will be measured by
additional light sensors in the laser scanning cytometer.
It is a further object of the present invention to thereby
enable laser scanning cytometers to be used as a point-of-care
instrument to be used by unskilled operators, wherein a specimen
is processed in the laser scanning cytometer within a disposable
cartridge, without user intervention, and wherein the cartridge
contains all necessary reagents needed to react the specimen
with fluorescent dyes, as well as providing means for removing
irrelevant cells and placing relevant cells in a small area on a
surface that will be scanned by the cytometer.
Generally the present invention comprises a method and
device for adhering cells of interest (or a set of cells
containing cells of interest as a subset) to a limited area of a
stationary specimen substrate such as a microscope slide, for
examination by a laser scanning cytometer. The method of the
present invention comprises the steps of:
a) treating a heterogeneous cell specimen sample with
adhering means which bind to cells of interest or to a selected
set of cells including the cells of interest as a subset
thereof;
b) adhering cells bound with the adhering means to a
limited area substrate by secondary adhering means;
5
CA 02321203 2000-08-18
WO 99/45094 PCTNS99/04415
c) if relevant, discriminating with discrimination means
between adhered cells of the subset of interest from other cells
of a set if a selected set of cells is adhered; and
d) analyzing cells of interest with a laser scanning
cytometer in the limited area of the substrate.
Alternatively, steps "a" and "b" in the above method, can
be replaced with the step of the surface itself being initially
provided with the adhering means, whereby the cells of interest,
or set of cells containing the cells of interest as a subset,
are adhered to the surface in situ, without pre-treatment of the
specimen sample. Additionally, irrelevant cells can be separated
out with the use of a porous membrane.
In a particularly preferred embodiment attuned to actual
usage, the analysis and separation is conducted in three
separate steps:
a) there is an initial separation of a general class of
cells from the overall heterogenous sample for adhesion to a
surface by means of a relevant reagent;
b) cells of interest which are adhered among other related
but irrelevant cells (i.e., the actual cells for which
measurements are exclusively contemplated) are marked with a
marker reagent (different from the reagent used to adhere the
cells) for the identification thereof, separate from the other
adhered cells; and
c) using one or more reagents for the determination of
specific properties (not merely a count) of the marked cells of
interest.
As a result of this procedure, in accordance with the
present invention, clinically important complex properties of
cells, such as their activation status or binding to other cells
can be determined.
In a specific embodiment of the device of the present
invention, used in effecting the above method, a slide is
provided with a fluid specimen loading conduit, wherein non-
impeding gating means, having the secondary adhering means, is
6
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
provided in a minimal area of the conduit, Said gating means
effects adhering of the cells having the adhering means attached
thereto with application of the secondary adhering means in a
pre-selected limited gating area.
In accordance with the method of the present invention,
there are three different techniques for adhering specific cells
of a type to limited area surfaces of a specimen substrate.
In the first technique, cells of a heterogeneous specimen
are reacted with paramagnetic particles that have been coated
with material that will bind the particles to an antibody and
the antibody is selected so that it will bind to an antigen on
the surface of the cell type of interest in the specimen. The
heterogeneous specimen is then reacted with the coated
paramagnetic particles whereby the particles bind to the cells
of interest (or cells of a set including a subset of cells of
interest). Techniques for manufacturing and using such particles
in biologic assays are described in U.S. Patents Nos. 3,933,997;
3,970,518; 4,018,886; 4,047,814; 4,219,411; 4,230,685; 4, 695,
393; 5,411,863; and 5,543,289.
The cells of interest are adhered to the substrate and
thereby separated from irrelevant cells by any of the following
procedures:
a) the suspension can be made to flow over a magnetic field
in a flow chamber to retain cells of interest on the surface. In
all of the above, the magnetic material comprises the secondary
adhering means (this is similar to procedures described in U.S.
Patents Nos. 4,219,411; 4,731,337 and 5,053,344).
b) the cells of interest can be separated from the
irrelevant cells by moving a magnetic field through the
suspended cells within the stationary fluid, thereby causing the
cells to follow the magnetic field to an area within or outside
the original stationary suspension (see U.S. Patents Nos.
3,985,649, 3,712,472, 4,272,510, 4,292,920, and 5,567,326).
Prior art also describes moving particles and cell from one
7
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
liquid phase to a second liquid phase by a magnetic field (e. g.,
U.S. Patents Nos. 4,777,145 and 5,279,936).
The second method of adhering selected cells to a surface
area is to directly deposit antibodies on the substrate surface
in the desired limited area. The specimen is made to come into
contact with the surface, whereupon antigens on the surface of
cells of interest bind to corresponding antibodies on the
surface, and irrelevant cells are washed away. Reagents such as
cellulose binding domain are commercially available for
enhancing -antibody to surface binding to achieve antigen
capture. This technique has been reviewed by Norden et al, An
Experimental Model of Affinity Cell Separation, Cytometry 16:15-
33 (1994) .
A third method of adhering selected cells to a surface
area in accordance with the present invention is to use a
membrane filter with pores that will selectively allow specific
types of cells to pass through the membrane while retaining
others. This methodology has been commonly employed for
separating blood cells and advantageously for separating
leukocytes from red blood cells and platelets, which readily
pass through a 5 micron sized pore while retaining leukocytes on
the membrane surface. Such ~ilters are commercially available
from numerous suppliers such as Osmotics, Livermore, CA. These
membranes may be layered over an absorbent material such as MFS
Borosilicate microfiber manufactured by Micro Filtration
Systems, Dublin, CA, with the specimen being made to flow over
the membrane surface, thereby capturing selected cells, while
passing irrelevant cells and fluid through the pores. Capillary
action may be utilized to cause flow of specimen, after it has
been mixed with marker reagents, to the filter surface. The
absorbent material may be placed below the membrane in a chamber
that serves as the waste reservoir.
Though the prior art is replete with magnetic, antibody
adhesion, and filtering techniques, the prior art has not taught
or suggested the use of magnetic, affinity adherence or
8
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
filtering for cell constituent analysis such as with laser
scanning cytometry.
Commonly, with flow cytometry, cells are reacted with
fluorescent dye conjugated antibodies and then washed. Multiple
centrifugations are employed to separate cells from reacting
reagents to perform the reaction and washing steps. Also,
physical separation techniques such as Ficoll-Hypaque or cell
analysis techniques are employed to reduce the numbers of
irrelevant cells. In accordance with the present invention,
laser scanning cytometry is utilized in order to exploit the
utility of solid phase assay as described.
Other objects, features and advantages of the present
invention will become more evident from the following
description and drawings in which:
svr~n~Y oi~ ~xE nl~mrras
FIGS. la and lb are schematic representations of disposable
specimen loading devices constructed in accordance with a
preferred embodiment of this disclosure.
FIG. 2 is a schematic representation of the optical
measurement system used in effecting measurements of the
specimen on the devices shown in Figures la and lb.
FIG. 3 is a rendering of photo sensor data from a laser
scanning cytometer, as the optical measurement system of FIG. 2.
FIG. 4 is a schematic depiction of a disposable slide
device as used in Figures la and lb, with loading conduit and
gated area shown.
FIGS. 5a and 5b are data from a laser scanning cytometer of
an example of an assay employing an embodiment of the disclosure
as set forth in a scattergram and histogram respectively.
FIGS. 6a-c are cell scattergrams of various parameters
prepared from data from a laser scanning cytometer of an example
of an assay of a specimen sample.
FIG. 6d is a histogram obtained from the data of FIG. 6a.
9
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
FIGS. 7 are two histograms obtained from laser scanned
cytometer data from platelet bound CD42b-PE-separated and
superimposed to obtain a third histogram.
FIGS. 8, 9a and 9b are additional cell scattergrams
obtained from data obtained from specimens assayed by a laser
scanning cytometer.
FIGS. l0a-b are histograms of superimposed histograms after
assay of two specimens.
FIG. 11 is a rendering of an analysis instrument used in
the specimen preparation and analysis in accordance with the
present invention.
FIG. 12a is a dot plot of the position of each granulocyte
of a sample on the filter of another embodiment of the present
invention.
FIG. 12b is a histogram of the laser scanning cytometer
CDllb-FITC fluorescence per cell of the sample shown in FIG 12a
and a second sample.
FIGS. 13a and 13b are two sets of superimposed histograms
of the orange fluorescence values per cell of other specimens.
DETAILED DESCRIPTION OF THE INVENTION
The method and device of the present invention, when
combined with use of laser scanning cytometry technology allows
the development of an instrumental system that can perform a
number of specific cell constituent assays, whereby an unskilled
user can obtain rapid results as a point-of-care instrument. A
number of clinically important assays are described below, by
way of Examples showing use of the present invention, which have
been developed using flow cytometry but have not been employed
in routine clinical medicine, because of problems inherent with
the use of flow cytometry.
EXAMPLE I
Among assays performable by laser scanning cytometers are
tests for activation of specific blood cells caused by sepsis in
newborns or adults, which can be detected by measuring the
activation of blood granulocytes resulting from interaction of
CA 02321203 2000-08-18
WO 99/45094 PC'TNS99/04415
these cells with bacteria. ( See e.g., Davis et al Neutraphil
CD64 expression: Potential diagnostic indicator of accuse
inflamation and therapeutic monitor of interferon - y therapy,
Laboratory Hematology 1:3-12 (1995). Leukocytes from whole blood
are captured on a small area surface of a specimen substrate by
a CD45 antibody, a fluorescent dye is conjugated to a CD15 to
bind to granulocytes and thereby used to identify granulocytes.
The fluorescence of a second dye conjugated to the antibody
CDllb which is expressed on the granulocyte surface as a
function of the cell's activation state is measured by a laser
scanning cytometer. A test for adult sepsis is made by
substituting the antibody CD64 for CDllb.
Additionally, in adults, CDllb granulocyte activation is
used to determine the condition of patients following
angioplasty or stent surgery. The activation of another
constituent of blood, platelets as measured by antibodies such
as AAC-2 to p-selectin, after their capture to a surface by an
anti-platelet antibody or thrombin, and their identification by
antibodies such as CD42b is used in an assay for cardiovascular
disease.
88AMPLE II
A sensitive assay for early detection of myocardial
infarctions (AMI) as well as other vascular pathologies has been
devised and tested using flow cytometry (described in U.S.
Patent No. 5,503,982). This assay relies on the discovery that
after arterial damage, circulating platelets become activated
resulting in p-selectin's transport to the platelet surface
where it binds to blood monocytes. A high ratio of monocytes
that are complexed with platelets to uncomplexed monocytes in a
patient blood specimen is an indicator of an AMI.
In accordance with the present invention, leukocytes from
whole blood are captured on a small area surface by a CD45
antibody and a~fluorescent dye is conjugated to a CD14 antibody
which binds to monocytes, and can be used to identify them. The
fluorescence of a second fluorescence dye conjugated to a CD42b
11
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
or CD61 antibody, which binds to platelets, is measured. The
ratio of the number of monocytes with second fluorescence to the
total number of monocytes is used in the assay determination.
EBAMPLE III
The method of the present invention is used to assay for
drug occupancy on a cell surface. One important clinically
useful assay of this type is to measure whether platelet protein
GPIIb/IIIa is bound to a drug or is capable of binding to
fibrinogen (described in U.S. Patent No. 5,440,020). Such drugs
are being developed for treatment of various cardiovascular
diseases. It is necessary to monitor the level of these drug's
occupancy, since over-treating will lead to other conditions
such as stroke. Techniques for performing diagnostic assays of
these drugs are disclosed in U.S. Patent No. 5,114,842.
In accordance with the present invention, blood is
activated by an ADP agonist to cause fibrinogen binding to non
drug occupied platelets. Thereafter, platelets from the whole
blood specimen are captured on a small area substrate surface by
a CD42b antibody. A fluorescent dye is conjugated to a CD41
antibody which binds to the GPIIb/IIIa protein in a region
separated from the fibrinogen binding region (U.S. Patent No.
5,372,933). The fluorescence of a second fluorescent dye
conjugated to a RIBS antibody which binds to fibrinogen when it,
in turn, is bound to platelet GPIIb/IIIa, is measured. The ratio
of the second fluorescence to first fluorescence intensity for
each captured platelet is measured and the distribution of this
ratio and computed derivatives of it are reported as the assay
determination.
The method of the present invention may also be used for
immunophenotyping, for example, counting T4 cells for AIDS
patient monitoring or for diagnosing various leukemias. This is
presently the major application for centralized flow cytometry
instruments and is clinically employed because results are not
required immediately so that an expensive, labor intensive flow
cytometer can be employed.
12
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
The present invention, as used with a laser scanning
cytometer, is however separately useful because of the very high
volume clinical assay of white blood differential counts since
laser scanning allows for the relocation of events for visual
observation, which is not possible with flow cytometry. With
use of the present invention, leukocytes from whole blood are
captured on a small area surface by an antibody such as CD45, a
fluorescent dye is conjugated to an antibody such as CD3 that
will bind to T lymphocytes and can be used to identify them, and
the fluorescence of a second fluorescence dye conjugated to an
antibody such as CD4 which binds to T4 helper lymphocytes, is
measured. The ratio of the number of CD4 positive cells with
second fluorescence to the total number of T lymphocytes is used
as the assay determination. This can be extended to many other
CD antibodies to detect other cell types. It is contemplated
that the methods of the disclosure can be extended to more than
detection of one type of cell by using more than two fluorescent
dyes conjugated to antibodies.
DETAILED DESCRIPTION OF THE DRAWINGS
AND THE PREFERRED EMBODIMENTS
With reference to Figs. la and lb, a specimen such as blood
from a patient is placed in a closed end tube 1. Tube 1 may be
a standard vacutainer used to draw blood, with a rubber sealing
insert 12 at its top. Tube 1 is inserted into sleeve 13, of the
disposable 2, thereby causing tubes 3 and 4 to pierce insert 12.
The disposable 2 (preferably made of molded plastic material)
contains three chambers 5, 7, and 8 and two ports 6 and 9.
Reagents are contained within chamber 5 of the disposable. In
one embodiment, in which magnetic particles coated with an
antibody are used to cause cells to adhere to a surface, chamber
5 contains coated magnetic particles. In a preferred embodiment,
chamber 5 contains two or more antibodies to cellular antigens,
each of which is conjugated to a different fluorescent dye.
Chamber 5 is connected to tube 4 as well as port 6. The assembly
consisting of the specimen tube 1 and disposable 2 is inserted
13
CA 02321203 2000-08-18
WO 99/45094 PGT/US99/04415
into the specimen entry port 81 of the analysis instrument 80
shown in Fig 11.
The assay is activated by the user by pressing the enter
key of the instrument beginning the test cycle. Port 6 which is
connected to a pressure source in the instrument is pressurized
then cycled in pressure, to cause the reagent contained in
chamber 5 to enter through tube 4 and mix with the specimen in
tube 1. The reagent contains two or more antibodies, each
conjugated to a different fluorescent dye. In preferred
embodiments these are fluorescein isothiocyanate (FITC) and
phycoerythrin (PE) which are excited by the wavelength energy of
an Argon ion laser. Other dyes such as CY3, CY5, APC, CY5.5. or
CY7 could be used for excitation by other gas or solid state
lasers. The reagent of the preferred embodiment in which cells
are adhered by a magnetic field, also contains magnetic
particles coated with an antibody such as CD45 which will bind
to cell surface antigens such as those of white blood cells for
CD45. Such particles are commercially available from Perceptive
BioSystems, Framingham, MA. In another embodiment in which
cells are directly captured by an antibody coated on a surface,
the magnetic particles are not part of the reagent.
After an incubation period, during which time the specimen
is mixed by varying the pressure through port 6, the pressure
through port 6 is maintained while port 9 is opened. The
specimen mixture flows through tube 3 to chamber 7. Chamber 7
has dimensions of 50 to 400 microns in depth so that the
specimen cells will flow close to the lower surface where there
is a magnetic field gradient. The chamber 7 width is set at a
value depending on the number of cells necessary to be captured
for good statistical results, as cells will be captured
perpendicular to the flow. A typical chamber width is about 5
mm. A magnet is placed in the instrument so as to produce a
strong magnetic gradient along a line perpendicular to the flow
at the area 10. This is accomplished in the Examples by using a
rectangular bar magnet with the pole face directed against the
14
CA 02321203 2000-08-18
- WO 99/45094 PCT/US99/04415
chamber 7, and the magnet edge placed at the centerline of area
10. To enhance the magnetic gradient, the chamber 7, can contain
a paramagnetic wire with a triangular cross section embedded in
the disposable perpendicular to the flow, so that the wire apex
is near the cell flow and the base is in contact with a magnet.
As the specimen flows through chamber 7, cells that have bound
to the magnetic particles will adhere in the area 10 (i.e. which
is, in effect, a pre-determined "gating area" for the cells).
The port 6 is then connected to a source of phosphate buffered
saline which is made to flow through chamber 5, and into chamber
7, thereby washing irrelevant cells from surface area 10. Waste
is collected in chamber 8. Cells are collected over the area 10
(a non-flow impeding gate) defined by the width of chamber 7 and
the distribution width characteristics of the magnetic field
gradient. The field and chamber dimensions are appropriately
adjusted so that the cells of interest are captured as a
monolayer for the specimen used. In other embodiments, cells
can be captured using an antibody coated surface area 10, in
place of a magnetic field and magnetic particles.
After completion of the cell capture phase, surface area
10, is scanned by a laser scanning cytometer as schematically
shown in FIG. 2. The scanning may be preferably done in the same
instrument used to capture the cells, or can be done within a
second independent instrument. The disposable 26, is held to
movable stage 33. The instrument contains a laser 20., such as an
Argon ion, HeNe, or solid state laser and the choice will depend
on the fluorescent dyes used. The laser beam 21 is reflected by
dichroic mirror 22, and computer controlled scanning mirror 23.
Such scanning mirrors are commercially available, for example,
from General Scanning, Watertown, MA. The beam is passed
through lenses 24 and 25, to produce a line scan focused at area
10. The dimensions of this line scan are preferably the width of
the cell capture area in scan extent and.5 to 10 microns in beam
diameter. The disposable 26 is moved perpendicular to the scan
by the stage 33, to raster scan all of area 10. As each cell
CA 02321203 2000-08-18
WO 99/45094 PCTNS99/04415
encounters the laser beam it emits fluorescent light
proportional to the amount of each fluorochrome in the scan
spot. This fluorescence is collected by lens 25 and imaged back
through the laser light path 21, where it is collimated as it
passes through dichroic mirror 22. In a preferred embodiment,
two fluorescence emissions are measured. Dichroic mirror 27
splits the emission into two wavelength ranges, optical filters
28 and 31 further define the wavelength ranges detected by photo
sensors 29 and 32. In an embodiment using an antibody coated
area 10, it is contemplated that a third sensor measuring
forward angle scatter will be used to find cell data and isolate
data belonging to individual cells. This sensor will consist of
a light blocking bar and photo sensor placed along the laser
light path below disposable 26.
The data signals from all photo sensors are simultaneously
digitized at a fixed rate, such as at 625,000 Hz. As the scan
beam passes over area 10 due to the motion of the scan mirror
and disposable platform, a raster of digital data or pixel
values is generated and stored in memory of a computer. FIG. 3
is a representation of data from one sensor as fluorescent
emitting cells are scanned where the density at each pixel
position is representative of the data value. The data from a
typical cell is represented by the pixels 30. Cells are found if
a set number of contiguous values above a set threshold are
found for any sensor's data. Alternatively a scatter sensor may
be used for detecting the presence of and isolating cell data.
The computer program first isolates the data belonging to each
cell found. It generates a contour 31' surrounding the cell at
the threshold value. The largest contour of the sensor contours
is selected and this is enlarged so that all cell data is used.
The pixel values within each contour are then summed. Two
additional contours 32' and 33 are constructed by the computer a
set distance from contour 31'. For each sensor, the pixel values
between these contours are averaged to determine a background
level which is subtracted from the corresponding cell sensor
16
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
sums. The integrated value result is proportional to the
fluorescence emission from that cell as detected by each sensor
and in turn is proportional to the number of a specific antigen
or other constituent bound to fluorescent dye molecules on the
cell.
Additional information, such as the contour area equal to
the number of pixels in the contour of each cell or the maximum
pixel value in the contour can be computed and may be used to
distinguish single cells from cell clusters and distinguish cell
states.
The integrated values determined as above are used to
determine the contents of information displayed to the user.
This data may be displayed as two parameter scattergrams,
histograms, or numeric values as described in the examples. The
result can be displayed on a screen, alphanumerical display, or
printed form.
The following examples are provided as evidencing test
results from laser scanning cytometry of various cell
determinations and characterizations in accordance with the
present invention.
BXAMPLE 1
This example describes an experiment to measure the cell
capture efficiency for a specific disposable design using
magnetic particle capture to adhere cells to a surface in a
defined area. Whole blood was diluted 1:1 with phosphate
buffered saline (PBS) containing an antibody conjugated to
fluorescein isothiocyanate (FITC). For this example the mouse
anti-human antibody CD45 which binds to all leukocytes was used.
After an incubation period the cells were washed with PBS and 50
microliters of Perceptive Biosystems magnetic particles
conjugated to an anti-mouse antibody were added to 100
microliters of the specimen.
A disposable device 40 shown diagrammatically in Fig. 4 was
constructed by adhering a 200~t thick adhesive coated mylar
template 41 with a 5 mm wide channel to a standard microscope
17
CA 02321203 2000-08-18
WO 99/45094 PGT/US99/04415
slide. A standard glass cover slip 43 was adhered to the center
of the slide so as to form a 200, x 5 mm flow channel 42. A bar
magnet was glued below the slide so that one pole was in direct
contact with the back surface of the slide and perpendicular to
the flow path. In this way two strong magnetic gradients are
placed along the specimen flow path and one would expect that
maximum capture of magnetically coated cells would occur in
proximity to the two edges of the magnet.
The specimen was pipetted into the right side of the flow
channel and flowed by capillary action through the channel. A
cotton wick was placed at the left end of the flow channel and
PBS was then pipetted into the flow channel to wash away all
irrelevant cells not adhering to the slide. This resulted in the
appearance of a line 44 from the excess magnetic particles
adhering to the slide surface at a position coinciding with the
first edge of the magnet where the magnetic field gradient is
strongest.
The slide disposable was placed on the microscope stage of
a CompuCyte LSC~, CompuCyte Corporation, Cambridge, MA., laser
scanning cytometer and a portion of the cover slip area was
scanned to locate the position of each cell found based on
detection of FITC fluorescence. The location of each cell can be
displayed as a scattergram 50 of Fig. 5a, in which each dot
represents a cell at the X and Y coordinates of the axes. The
number of cells as a function of position along the flow path
can be plotted as a histogram 51 of Fig. 5b, in which numbers of
cells at each X position is plotted.
If it is assumed that the magnetic gradient at the leading
and trailing edges of the magnet are equal, it is possible to
calculate the efficiency of setting and holding cells that can
be captured in place during the specimen and wash fluid flows
through the chamber. In this experiment the number of cells
captured near the first gradient is approximately twice the
number near the second gradient. With the small number of cells
captured between the gradients, the efficiency is calculated as
18
CA 02321203 2000-08-18
WO 99145094 PCTNS99/04415
approximately 50%. It should be noted that provided there is no
biased capture of a subset of the cells of interest, practice of
this invention does not require high efficiencies. It requires
capturing statistically significant numbers of cells within a
small area that can be scanned.
EBAMPhE 2
A disposable configured as described in Example 1 was used
in all of the remaining examples. In this example, cells are
captured with the same antibody used to detect their presence
with a first fluorochrome, and a property of the cells is
measured using a second antibody and fluorochrome.
A specimen of whole blood was mixed 1:1 with PBS containing
a FITC conjugated mouse anti-human antibody CD42b which binds to
platelets and a phycoerythrin (PE) conjugated mouse anti-human
antibody ACC-2 which binds to p-selectin antigen found on the
surface of activated platelets. PE fluoresces at a wavelength
that can be separated from FITC fluorescence by appropriate
optical filters. This example is intended to show the number and
extent of activated platelets in a patient blood specimen. After
incubation, 50 microliters of anti-mouse antibody conjugated
magnetic particles were added to 100 microliters of the mixture,
incubated and added to the flow chamber followed by a PBS wash.
This was repeated with a second control blood specimen without
the ACC-2 antibody. The resulting LSC displays after assaying
these two specimens is shown in Figs. 6a-d. The position
distributions of the detected platelets in the control specimen
are shown as a scattergram 60 of Fig. 6a and a histogram 61 of
Fig. 6d and are similar to Example 1. The spectral overlap
compensated scattergram 63 of Fig. 6c shows the level of
activation based on ACC-2 antigen measurement plotted versus
CD42b antigen per platelet. The data was compensated for filter
spectral overlap as practiced in flow cytometry phenotyping. The
number of events in each of the four quadrants of display 62.of
Fig. 6b were counted and 83% of the platelet events had a
relatively high level of ACC-2 p- selectin expression as
19
CA 02321203 2000-08-18
WO 99145094 PCT/US99/04415
compared to the control specimen's scattergram display 63 of
Fig. 6c in which the count in this quadrant was 12%.
It is understood that other antibodies can be used to
measure cell activation such as antibodies that will bind to
GPIIb/IIIa or fibrinogen, as well as p-selectin of platelets. It
is also contemplated that this method will be used to measure
granulocyte activation using either CDllb or CD64, or other
antibodies expressed on activated granulocytes. It is also
understood that the first antibody can be different than the
antibody used to capture the cells of interest.
EBAMPhE 3
This is an example of application of the preferred
embodiment in which cells are captured to a small surface area
with a first antibody, a subset of cells is identified with a
second antibody, and a characteristic of the cells of the subset
is measured with a third antibody. This example demonstrates an
assay to determine the percentage of monocytes complexed with
platelets. It has been shown in the prior art that
cardiovascular injury results in the activation in circulating
blood, of platelets which then in turn bind to monocytes. An
assay of the ratio of platelet bound to non-platelet bound
monoctes is an indicator of cardiovascular injury such as an
acute myocardial infarction.
Two samples of whole blood from the same individual were
obtained. Two micromoler ADP and 5 millimolar GPRP were added to
one sample to activate the platelets. Platelets, when activated,
bind to some leukocytes including monocytes. Each sample was
mixed in a 1:1 ratio with PBS containing three antibodies, the
first mouse anti-human CD45 conjugated to Perceptive Biosystems
magnetic particles. CD45 binds to a common leukocyte antigen on
white blood cells. The second antibody, mouse anti-human CD14
was conjugated to FITC and binds to monocytes. The third
antibody, mouse anti-human CD42b, which binds to platelets, was
conjugated to the dye phycoerythrin (PE). The samples were then
fixed by adding paraformaldehyde to a 1% concentration to
CA 02321203 2000-08-18
WO 99/45094 PGT/US99/04415
further inhibit platelet activation. Each mixture was placed in
a disposable as described in Example 1, followed by PBS to wash
irrelevant cells from the surface. Each slide was assayed on the
LSC laser scanning cytometer, measuring fluorescence from each
of the dyes. Cells were detected and data from each cell was
contoured based on green fluorescence from the FITC bound to
CD14 on monocytes. Fig. 7 shows two superimposed histograms of
the maximum value within each contour of orange fluorescence
from the platelet bound CD42b-PE. The non-activated sample curve
71 is clearly differentiated from the assay result 72 of the
activated sample indicating that the preferred embodiment can be
used to assay for platelet monocyte complexes in a blood
specimen.
This experiment was repeated using the same dye conjugated
antibodies and capture reagent, however fresh undiluted and
unfixed blood was used for both the unstimulated and stimulated
specimens. Ten micromolar ADP was used to stimulate adhesion of
platelets to monocytes in this experiment. The results are shown
in Fig. 7 in which curve 73 represents the data from the
unstimulated specimen and 74 represents the data from the ADP
stimulated specimen, measuring the total fluorescence of CD42b-
PE of each found cell. In the separate graph of Fig. 7, 75
represents the data from the unstimulated specimen and 76
represents the data from the ADP stimulated specimen, measuring
the maximum fluorescence of CD42b-PE of each found cell. The two
populations from the stimulated and unstimulated specimens of
the same fresh blood are clearly distinguishable. It is noted
that other platelet antibodies such as CD61 and CD 41 may be
preferable to CD42b for the assay of this example because the
antigen to which CD42b binds may be decreased when platelets are
activated.
EXAMPLE 4
Using the method of the preferred embodiment as in Example
3, the percentage of lymphocyes expressing the T4 antigen was
21
CA 02321203 2000-08-18
WO 99145094 PCTNS99/04415
assayed. This is a common assay performed on flow cytometers for
monitoring AIDS patients. In this example, the specimen was
mixed with CD45 magnetic particles to capture leukocytes as
above, The reagent mixture, in this example, contained mouse
anti-human CD3 antibody conjugated to PE and mouse anti-human
antibody CD4 conjugated to FITC. CD3 binds to antigens found on
T lymphocytes while CD4 binds to T4 or helper T lymphocytes.
The slide was assayed as described in Example 4 resulting
in the scattergram data of Fig. B. CD4 expressing lymphocytes
have higher green fluorescence than the remaining T lymphocytes
and appear in the upper cluster of the scattergram of CD4 versus
CD3. Using gating regions this assay showed that approximately
58% of T cells were T4.
EXAMPLE 5
This example uses the preferred embodiment to demonstrate
the common white blood cell differential count. A whole blood
specimen was prepared as in examples 3 and 4, using CD45
conjugated to magnetic particles and the monocyte specific mouse
anti-human antibody CD14 and the granulocyte specific mouse
anti-human antibody CD15. CD14 was conjugated to FITC and CD15
was conjugated to PE. The slide was assayed twice, the first
time using green fluorescence data to determine the contour and
the second time using orange fluorescence to determine the
contour. This was done this way because the CompuCyte LSC does
not have the means to simultaneously use two sensors. to segment
cell data. The resulting data is shown in Figs. 9a and 9b in
which the two scattergrams represent data from each of the two
assays respectively. The orange fluorescence gated scattergram
shows 95% of the cells expressing orange fluorescence are CD15
positive and are granulocytes. Approximately 2920 cells were
counted in quadrant region 4. The green fluorescence gated
scattergram shows 92% of the cells in quadrant region 1
expressing high levels of CD14 and are monocytes. Approximately
800 cells were counted in region 1.
EXAMPLE 6
22
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
This example uses the preferred embodiment to demonstrate
the determination of activation of whole blood granulocytes by
the agonist f-Met-Leu-Phe (fMLP). Two samples of the same blood
specimen were obtained. To one sample at 37° C, 10 ~,M fMLP was
added and that sample was incubated for 10 minutes to cause
activation of the sample's granulocytes. Both samples were mixed
1:1 with PBS containing two antibodies. Mouse anti-human CD15
was conjugated to FITC and binds to granulocytes. The second
antibody, mouse anti-human CDllb which binds to an activation
antigen of granulocytes was conjugated to the dye phycoerythrin
(PE). After an incubation period, paraformaldehyde was added to
a final concentration of 1%. Fifty microlitera of anti-mouse
antibody conjugated magnetic particles was then added to 100
microliters of each mixture, incubated and added to the flow
chamber followed by a PBS wash. Each specimen was assayed on the
LSC cytometer, detecting and contouring events based on green
fluorescence from the FITC bound to the granulocytes. The orange
PE fluorescence resulting from the binding of CDllb to the
granulocyte activation antigen was measured. The resulting
superimposed LSC histogram displays after assaying these two
specimens is shown in Fig. 10a. The distribution of fluorescence
per cell of the unactivated sample had the result shown as 91
and the activated sample had the distribution result 92. These
distributions are clearly distinguishable from each other.
This experiment was repeated with undiluted blood using the
same dye conjugated antibodies and 10 ~Ni fMLP stimulant.
However, in this experiment the specimens were not washed and
the specimens were incubated for 30 minutes with magnetic
particles conjugated to CD45 antibody to capture them on to a
surface. The results of assaying these specimens on a CompuCyte
Corporation LSC cytometer are shown in Fig. lOb, in which blood
assayed within 2 hours of draw is represented as curve 93, blood
assayed 6 hours later is represented as curve 94, and a
stimulated specimen of the same patient's blood is represented
23
CA 02321203 2000-08-18
- WO 99/45094 PCT/US99/04415
as curve 95 in which the distribution of per cell CDllb-PE
fluorescence is shown. The stimulated specimen assay data is
clearly distinguishable from both non stimulated specimens.
It should be understood that other means of capturing cells
not using antibodies such as coating magnetic particles or
surfaces with lectins are contemplated. Also means of
distinguishing cells or subsets of cells using dyes not
conjugated to antibodies are contemplated. For example, dyes
such as propidium iodide taken up by cellular nucleic acids are
contemplated as are chemical substrates producing fluorescent
end products as a result of interaction with cellular
constituents. The use of nucleic acid probes as a marker is also
contemplated.
Fvr example, bacteria contained in blood, pharmaceuticals,
or food may be captured onto a surface using lectins coated to
the surface or conjugated to magnetic particles ( See e.g.
Payne et al The use of ilmmobolized lectins in the separation of
Staphylococcus aureus, Escherichia coli, Listeria and Salmonella
spp. From pure cultures and foods, Journal of Applied
Bacteriology 73:41-52 (1992). Bacteria can be counted by using
nucleic acid specific dyes taken up by all bacteria. ( See the
review, Kepner et al, Use of Fluorochromes for Direct
Enumeration of Total Bacteria in Environmental Samples: Past and
Present, Microbiological Reviews, 58:603-615 (1994). The surface
can be scanned and in this case fluorescent events counted.
Additionally, a second dye (available from Molecular Probes,
Eugene, OR.) can be used to determine if the bacteria are, alive
or dead and can be added to differentially count live bacteria.
Third, bacterial specific antibodies can be used to determine
the type of bacteria on the surface (See e.g. Vesey et al,
Evaluation of Fluorochromes and Excitation Sources for
Immunofluorescence in Water Samples, Cytometry 29:147-154
(1997). Fourth, ribosomal RNA specific nucleic acid probes can
be used to identify bacteria. This is reviewed by Amann et al,
Phylogenetic Identification and In Situ Detection of Individual
24
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
Microbial Cells without Cultivation, Microbiological Reviews
59:143-169 (1995).
It is also contemplated that the surface used to capture
the bacteria may be embedded in a growth medium, allowing the
live bacteria to remain viable. In this case, after a period of
time to allow the bacteria to reproduce, the capture area may be
scanned a second time and the specimen reassayed. Since laser
scanning cytometry records the position of each cell, as
described in the examples, it is possible to determine changes
in fluorescence of any event found. Event contouring can be
adjusted to encompass a large enough area so proximate bacteria
are within the same contour. If nucleic acid per bacteria, for
example, is measured, viable bacteria will have increased
amounts of DNA per event measured during the second assay, and
their viability indicated. The patterns of located events ( as
shown and described in EXAMPLE 1) for the first and second
assays can be compared by software to determine the
correspondence of each events data in the first and second
assays. The corresponding fluorescence values will then indicate
bacterial growth by virtue of increases in total fluorescence or
size properties between the two assays.
It is also contemplated that, with the antibody coating
cell capture method, more than one capture area will be used,
each coated with a different antibody, so that different types
of cells will be captured in different areas of the surface.
This will allow for more complex assays of multiple cell types
without the need to employ multiple fluorescent dyes and
multiple instrument sensors to detect these fluorescences.
Additionally one specimen may be tested for multiple assays.
The following examples are indicative of the manner in
which the third embodiment of the cell separation is effected in
accordance with the present invention. In such embodiment the
selective cell separation is effected by means of a filtration
system as described.
EXAMPhE 7
CA 02321203 2000-08-18
WO 99/45094 PCT/US99/04415
This example exemplifies the embodiment in which all
leukocytes from a whole blood specimen are separated from red
blood cells by a membrane with 5 micron pores. The example
demonstrates an assay for determination of the activation of
whole blood granulocytes by the agonist f-Met-Leu-Phe (fMLP).
Two samples of the same blood specimen were obtained. lOM
fMLP was added to one sample at 37°C, and the sample was
incubated for 10 minutes to cause activation of the sample's
granulocytes. Each sample was then separately mixed in a 1:1
ratio with PBS containing two antibodies. A first antibody,
mouse anti-human CD15 that binds to granulocytes was conjugated
to the dye phycoerythrin (PE). The second antibody, mouse anti-
human CDllb that binds to an activation antigen of granulocytes
was conjugated to the dye fluorescein isothiocyanate (FITC).
After an incubation period, 100 microliters of each
mixture, was pipetted into the center of an Osmotics (Catalog
no. 10572), 5 micron pore size, 13 mm membrane filter placed
over MFS Borosilicate microfiber absorbent material (Micro
Filtration Systems, Dublin, CA). This was immediately followed
by the pipetting of 200 microliters of PBS onto the membrane
center surface. Each filter was removed, placed on a microscope
slide and covered with a slip and assayed on the laser scanning
cytometer. Granulocytes were detected and contoured based on
orange fluorescence from the PE bound to the granulocytes. The
green FITC fluorescence resulting from the binding of CDllb to
the granulocyte activation antigen was measured for every
contoured event. Since the LSC cytometer is capable of recording
and displaying the position of each found event, it was used for
such purpose. The positions of each granulocyte on the filter,
for the first specimen, is displayed as a dot plot 101 of Fig.
12a. The resulting superimposed LSC histogram displays of CDllb-
FITC fluorescence per cell, after assaying these two specimens
are shown as Fig. 12b. The distribution of fluorescence per cell
of the unactivated sample is shown as 103 and the activated
sample is shown as distribution 104. These distributions are
26
CA 02321203 2000-08-18
WO 99/45094 PCTNS99/04415
clearly distinguishable from each other. The histogram display
was divided into two regions by the boundary line 105 and the
counts of cells on the right (positive CD llb) side of the
boundary was determined for each of the unstimulated and
stimulated specimen. For these assays the unstimulated count was
87 of 15,882 or 5.1% of the total cells counted and the
stimulated specimen's count was 7342 of 9043 or 81.2%.
EXAMPLE 8
This is an example of an application of the third
embodiment in which cells are captured on a membrane of small
surface area, wherein a subset of cells is identified with a
first antibody, and wherein a characteristic of the cells of the
subset is measured with a second antibody. The example
illustrates an assay of determination of the percentage of
monocytes complexed with platelets.
It has been shown in the prior art that cardiovascular
injury results in the activation, in circulating blood, of
platelets which in turn bind to monocytes. An assay of the ratio
of platelet bound to non-platelet bound monocytes is an
indicator of cardiovascular injury, such as an acute myocardial
infarction.
Four samples of whole blood from the same individual
were obtained. Two micromolar ADP and 5 millimolar GPRP were
added to two samples to activate the platelets. Platelets, when
activated, bind to some leukocytes including monocytes. Each of
the four samples was separately mixed in a 1:1 ratio with PBS
containing two antibodies. The first antibody, mouse anti-human
CD14 was conjugated to FITC, which binds to monocytes, was used
in all specimens The second antibody, mouse anti-human CD42b,
was added to one activated and one non-activated sample, and the
antibody, mouse anti-human CD62, was added to one activated and
one non-activated sample. Both antibodies, which bind to
platelets, were conjugated to the dye phycoerythrin (PE). 100
microliters of each of the mixtures were pipetted to the center
of a filter as described in Example 7, followed by PBS to wash
27
CA 02321203 2000-08-18
WO 99/45094 PCTNS99/04415
irrelevant cells from the surface. The membranes were
transferred to a slide and a slip was placed over each membrane.
Each slide was assayed on the LSC laser scanning cytometer,
measuring fluorescence from each of the dyes. Cells were
detected and data from each cell was contoured based on green
fluorescence from the FITC bound to CD14 on monocytes. Figs.
13a and 13b show two sets of superimposed histograms of the
orange fluorescence values per cell, the histograms 106 from the
platelet bound CD42b-PE specimen, and the histograms 109 from
the platelet bound CD62-PE specimen. The non-activated sample
curve 1077 is clearly differentiated from the assay result 108
of the activated sample as is the non-activated curve 110,
differentiated from activated curve 111. The histogram displays
were divided into two regions by the boundary lines 109a and 112
respectively and the counts of cells on the right side of the
boundary consisting of complexed monocytes and platelets
(positive CD42b or CD62) was determined for each of the
unstimulated and stimulated specimens. For these assays the
unstimulated count using CD42b was 111 of 2408 or 4.6% of the
total cells counted and the stimulated specimen's count was 1681
of 2267 or 74.2%. Using CD62 it was 448 of 6053 or 7.4% of the
total cells counted and the stimulated specimen's count was 3614
of 4616 or 78.3%. This embodiment for effecting the requisite
separation is a preferred embodiment, as one which can also be
used to assay for platelet monocyte complexes in a blood
specimen.
It is understood that the above discussion with specific
examples as well as the drawings are merely illustrative of the
present invention and that changes in procedure, materials and
processing steps are possible without departing from the scope
of the present invention as defined in the following claims.
28
