U.S. patent application number 15/045004 was filed with the patent office on 2016-08-25 for methods and systems for real-time sensing of viable and nonviable cells.
The applicant listed for this patent is General Electric Company. Invention is credited to Victoria E. Cotero, Jon A. Dieringer, Thomas J. Erdenberger, Radislav A. Potyrailo, Timothy A. Wortley.
Application Number | 20160244712 15/045004 |
Document ID | / |
Family ID | 52232190 |
Filed Date | 2016-08-25 |
United States Patent
Application |
20160244712 |
Kind Code |
A1 |
Potyrailo; Radislav A. ; et
al. |
August 25, 2016 |
METHODS AND SYSTEMS FOR REAL-TIME SENSING OF VIABLE AND NONVIABLE
CELLS
Abstract
Provided herein are techniques for identification of viable and
nonviable cells in a cell culture that include measuring a
resonance impedance spectral response of at least one resonator in
proximity to the cell culture and correlating the measured response
to the concentration of viable cells in cell culture and/or the
concentration of nonviable cells in cell culture.
Inventors: |
Potyrailo; Radislav A.;
(Niskayuna, NY) ; Cotero; Victoria E.; (Troy,
NY) ; Dieringer; Jon A.; (Des Plaines, NY) ;
Erdenberger; Thomas J.; (Arlington, MA) ; Wortley;
Timothy A.; (Lancaster, NY) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
General Electric Company |
Schenectady |
NY |
US |
|
|
Family ID: |
52232190 |
Appl. No.: |
15/045004 |
Filed: |
February 16, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14144118 |
Dec 30, 2013 |
9274071 |
|
|
15045004 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01N 33/48735 20130101;
G01N 27/026 20130101; C12M 41/46 20130101; C12M 41/36 20130101;
G01N 27/02 20130101 |
International
Class: |
C12M 1/34 20060101
C12M001/34 |
Claims
1. A system, comprising: a sensor comprising: a dielectric
substrate; at least one sensing region; and a plurality of tuning
circuits over a frequency range of analysis of a cell culture
reaction, wherein the sensor is configured to be compatible with at
least one a plurality of cell culture media types; and a reader
configured to receive a signal from the sensor across the frequency
range of analysis; and processing circuitry configured to perform
the steps of: analyzing impedance spectra of the sensor; and
determining one or more properties of the cell culture reaction
based on the analyzed impedance spectra.
2. The system of claim 1, where the system quantifies a
concentration of nonviable cells and a concentration of viable
cells.
3. The system of claim 1, where the system quantifies a
concentration of nonviable cells, a concentration of viable cells,
and a size of viable cells.
4. The system of claim 1, wherein the reader is configured to
communicate wirelessly with a bioprocess control system controlling
the cell culture reaction.
5. The system of claim 1, wherein the sensing pregion is protected
from direct contact with the cell culture reaction via a
biocompatible dielectric conformal layer.
6. The system of claim 1, wherein the sensor is inserted or mounted
to a non-rigid container via mounting device such that the sensor
is held aseptically within the internal space of the container and
in operational contact with fluid within the container.
7. The system of claim 1, wherein the mounting device comprises a
hose barb port.
8. A system, comprising: a bioreactor holding a cell culture
reaction; a sensor comprising at least one sensing region in
operational contact with the cell culture reaction; a reader
configured to receive a signal from the sensor, wherein the signal
is representative of impedance spectra of the sensing region over a
measured spectral frequency range; and processing circuitry
configured to perform the steps of: analyzing the impedance spectra
of the sensor; and determining one or more properties of the cell
culture reaction based on the analyzed impedance spectra.
9. The system of claim 8, where the system quantifies a
concentration of nonviable cells and a concentration of viable
cells.
10. The system of claim 8, where the system quantifies a
concentration of nonviable cells, a concentration of viable cells,
and a size of viable cells.
11. The system of claim 8, wherein the reader is configured to
communicate wirelessly with a bioprocess control system controlling
the cell culture reaction.
12. The system of claim 8, wherein the sensing pregion is protected
from direct contact with the cell culture reaction via a
biocompatible dielectric conformal layer.
13. The system of claim 8, wherein the sensor is inserted or
mounted to a non-rigid container via a mounting device such that
the sensor is held aseptically within the internal space of the
container and in operational contact with fluid within the
container.
14. The system of claim 13, wherein the mounting device comprises a
hose barb port.
15. The system of claim 8, wherein the sensor comprises a plurality
of tuning circuits outside a fluid of the cell culture
reaction.
16. A system, comprising: a sensor comprising at least one sensing
region in contact with a cell culture reaction; a reader configured
to receive a signal from the sensor, wherein the signal is
representative of impedance spectra of the sensing region over a
measured spectral frequency range; and processing circuitry
configured to perform the steps of: analyzing the impedance spectra
of the sensor; and determining one or more properties of the cell
culture reaction based on the analyzed impedance spectra.
17. The system of claim 16, wherein the sensor comprises at least
two electrodes.
18. The system of claim 17, wherein the at least two electrodes
form a sensing electrode structure.
19. The system of claim 18, wherein the sensing electrode structure
is part of the sensing region.
20. The system of claim 17, wherein the sensor comprises at least
one resonant circuit.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present application is a divisional of U.S. patent
application Ser. No. 14/144,118, entitled "METHODS AND SYSTEMS FOR
REAL-TIME SENSING OF VIABLE AND NONVIABLE CELLS", filed Dec. 30,
2013, the disclosure of which is incorporated herein by reference
in its entirety for all purposes.
BACKGROUND
[0002] The subject matter disclosed herein relates to cells grown
in cell culture and distinguishing between viable and nonviable
cells in cell culture in real time.
[0003] The self-replicating nature of biological systems may be
harnessed to produce materials of interest by growing individual
cells or organisms that produce such materials and purifying the
desired products from the cell culture. For example, certain drugs
or compounds may be produced by cells, either as a natural
metabolic product of the cells or via engineered cells, such as
cells engineered to produce recombinant proteins. These cells may
be grown in large scale reactors to produce high quality
biopharmaceuticals. In such reactors, the manufacturing parameters
are monitored using in-line sensors and off-line (i.e., not in-line
but in the laboratory, not connected to the reactors) analytical
systems. Monitoring of critical manufacturing parameters may be
performed upstream (before the bioreactor), in the bioreactor, and
downstream (after the bioreactor). Parameters that are typically
continuously monitored include parameters that influence growth
conditions, such as temperature, solution conductivity, pressure,
pH, glucose concentration, dissolved oxygen, and viable cell mass,
and are monitored in a similar manner in both single-use and in
conventional stainless steel manufacturing systems. Other
parameters, such as cell viability, lactate, glutamine, osmolality,
pyruvate, amino acids, product purity, and trace elements are
typically not measured in-line, because of difficulties of
measurements using existing sensors. Instead, these parameters are
typically measured "off line" by extracting samples from the
reactors and further analyzing these samples outside of the reactor
using laboratory analytical instruments.
BRIEF DESCRIPTION
[0004] In one embodiment, a method for analyzing a cell culture
reaction is provided. The method includes generating a plurality of
frequencies with a sensor comprising a sensing region in
operational contact with the cell culture reaction fluid and a
plurality of tuning circuits outside the fluid; receiving a signal
from the sensor, wherein the signal is representative of impedance
spectra of the sensing region in operational contact with the cell
culture reaction fluid over a measured spectral frequency range;
analyzing the impedance spectra; and determining one or more
properties of the cell culture reaction fluid based on the analyzed
impedance spectra. The method includes generating multiple
frequencies with a sensor comprising at least one resonant circuit,
wherein the sensor is in operational contact with a cell culture
reaction fluid; receiving a signal from the sensor, wherein the
signal is representative of an impedance spectrum of the cell
culture reaction over a measured spectral frequency range;
analyzing at least three spectral parameters of the impedance
spectrum; and correlating at least three spectral parameters to a
concentration of viable and nonviable cells in the cell culture
reaction based on the spectral parameters.
[0005] In another embodiment, a method for analyzing a cell culture
reaction with enhanced selectivity is provided. The method includes
generating multiple frequencies with a sensor comprising a sensing
region that is in operational contact with a cell culture reaction
fluid and a plurality of tuning circuits outside the cell culture
reaction fluid that provide resonances over the frequency range of
spectral beta-dispersion of the cell culture, wherein the sensor is
in contact with a cell culture reaction; receiving a signal from
the sensor, wherein the signal is representative of resonance
impedance spectra of the cell culture reaction over a measured
spectral frequency range; analyzing at least six spectral
parameters of the resonance impedance spectrum; and correlating the
linear combination of the spectral parameters to a concentration of
viable and nonviable cells in the cell culture reaction based on
the spectral parameters.
[0006] In another embodiment, a method for analyzing a cell culture
reaction is provided. The method includes generating multiple
frequencies with a sensor comprising a sensing region that is
protected from a direct contact with the analyzed fluid by the use
of a dielectric conformal layer or a dielectric protective layer
(e.g., 10 nanometers to 10 millimeters in thickness, in a
non-limiting example) and a plurality of tuning circuits outside
the cell culture reaction fluid that provide resonances over the
frequency range of spectral dispersions of the cell culture,
wherein the sensor with the dielectric conformal protective layer
is in contact with a cell culture reaction; receiving a signal from
the sensor, wherein the signal is representative of resonance
impedance spectra of the cell culture reaction over a measured
spectral frequency range; analyzing at least six spectral
parameters of the resonance impedance spectrum; and correlating the
linear combination of the spectral parameters to a concentration of
viable and nonviable cells in the cell culture reaction based on
the spectral parameters. A sensor comprising a sensing region and
at least one tuning circuit forms a resonant circuit structure or a
resonator. The dielectric conformal layer is biologically
compatible with the fluid and cell culture of the cell culture
reaction. Other parameters, such as cell viability, lactate,
glutamine, osmolality, pyruvate, amino acids, product purity, and
trace elements may also be assessed. For example, in an antibody
production reaction, increased osmolality in the context of
decreased cell growth occurs during increased antibody
production.
[0007] In another embodiment, a method for analyzing a cell culture
reaction in a single-use container with a single-use sensor is
provided. The method includes calibration of at least one sensor
from the fabricated batch of sensors using known calibration
standards, establishing calibration transfer relations between
sensors in the fabricated batch based on the manufacturing
tolerances of the sensors manufacturing, recording calibration
coefficients into the memory of each sensor in the fabricated
batch, integrating the sensors into single-use containers, applying
a sterilization step for the containers with the integrated
sensors, operating container for cell culture, and monitoring the
concentration of viable and nonviable cells in the cell culture
reaction using the integrated sensors. In other embodiments,
sensors may be multi-use and, further, may be tuned to one or more
growth media depending on their desired use.
[0008] In another embodiment, a method for analyzing a cell culture
reaction is provided. The method includes providing a sensor
comprising at least one resonant circuit; exposing the sensor to
the cell culture reaction; probing the cell culture reaction with
at least one resonance generated by the sensor; determining at
least one resonance impedance spectrum of the sensor response over
a measured spectral frequency range of the sensor; applying
multivariate statistical analysis to the at least one resonance
impedance spectrum of the sensor response to obtain multivariate
response factors; and relating the multivariate response factors to
concentrations of viable cells and nonviable cells in the cell
culture reaction.
[0009] In another embodiment, a method for analyzing a cell culture
reaction is provided. The method includes providing a sensor
comprising at least one resonant circuit; exposing the sensor to
the cell culture reaction; probing the cell culture reaction with
at least one resonance generated by the sensor; determining at
least one resonance impedance spectrum of the sensor response over
a measured spectral frequency range of the sensor; applying
multivariate statistical analysis to the at least one resonance
impedance spectrum of the sensor response to obtain multivariate
response factors; and relating the multivariate response factors to
concentration of viable cells, size of the viable cells, and
concentration of nonviable cells in the cell culture reaction.
[0010] In another embodiment, a system is provided that includes a
sensor comprising at least one resonant circuit and at least two
electrodes, wherein the sensor is configured to generate a signal
representative of an impedance spectrum over a measured spectral
frequency range. At least two electrodes form a sensing electrode
structure. The sensing electrode structure can be bare and in
direct contact with the cell culture reaction fluid or the sensing
electrode structure can be conformally coated with a dielectric
protective coating. A sensor comprising a sensing region and at
least one tuning circuit forms a resonant circuit structure. The
system also includes a memory that stores calibration coefficients
of the sensor for quantitation of viable and nonviable cells and a
memory that stores instructions for: receiving the signal from the
sensor; analyzing two or more spectral parameters of the impedance
spectrum based on the signal; and correlating the two or more
spectral parameters to a concentration of viable and nonviable
cells in the cell culture reaction based on the spectral
parameters; and a processor configured to execute the
instructions.
[0011] In another embodiment, a system is provided that includes a
sensor comprising at least one sensing region and a plurality of
tuning circuits that provide at least three resonances over the
frequency range of spectral dispersions of the cell culture,
wherein the sensor is in contact with a cell culture reaction. The
system also includes a memory storing instructions for: generating
multiple frequencies with the sensor; receiving a signal from the
sensor, wherein the signal is representative of resonance impedance
spectra of the cell culture reaction over a measured spectral
frequency range; analyzing at least six spectral parameters of the
resonance impedance spectrum; and correlating the linear
combination of the spectral parameters to a concentration of viable
and nonviable cells in the cell culture reaction based on the
spectral parameters; and a processor configured to execute the
instructions.
[0012] In another embodiment, a system is provided that includes a
sensor comprising at least one sensing region that is protected
from a direct contact with the analyzed fluid by the use of a
dielectric conformal layer that is biologically compatible with the
cells in the reaction and a plurality of tuning circuits that
provide at least three resonances over the frequency range of
spectral dispersions of the cell culture, a sensor reader to read
the response of the sensor across predetermined frequency ranges of
the resonances, and a processor for analyzing at least six spectral
parameters of the resonance impedance spectra and correlating the
linear combination of the spectral parameters to a concentration of
viable and nonviable cells in the cell culture reaction based on
the spectral parameters.
[0013] In another embodiment, a system for analyzing a cell culture
reaction is provided that includes a single-use sensor comprising
at least one resonant circuit and at least two electrodes, wherein
the sensor is configured to generate a signal representative of an
impedance spectrum over a measured spectral frequency range. The
system also includes a memory that stores calibration coefficients
of the sensor for quantitation of viable and nonviable cells and a
memory that stores instructions for: receiving the signal from the
sensor; analyzing two or more spectral parameters of the impedance
spectrum based on the signal; and correlating the two or more
spectral parameters to a concentration of viable and nonviable
cells in the cell culture reaction based on the spectral
parameters; and a processor configured to execute the
instructions.
[0014] In another embodiment, a system for analyzing a cell culture
reaction in a single-use container (e.g., flask, microtiter well
plate, etc.) with a single-use sensor is provided that includes a
resonant sensor assembly on a dielectric substrate having a sensing
region. The sensor assembly further comprises a plurality of tuning
elements operatively coupled to the sensing region, wherein the
sensing region is coupled to the plurality of tuning elements to
define a plurality of resonant circuits.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] These and other features, aspects, and advantages of the
present invention will become better understood when the following
detailed description is read with reference to the accompanying
drawings in which like characters represent like parts throughout
the drawings, wherein:
[0016] FIG. 1 is a general schematic for performing off-line and
in-line measurements according to an embodiment of the
disclosure;
[0017] FIG. 2 is a block diagram of a system for determining viable
and nonviable cells in cell culture according to an embodiment of
the disclosure;
[0018] FIG. 3 is a schematic view of a resonant sensor according to
an embodiment of the disclosure;
[0019] FIG. 4 is a schematic view of a portion of an example sensor
system employing a sensor assembly configured for in-line sensing
of a cell culture reaction fluid using a plurality of frequencies,
in accordance with embodiments of the present disclosure;
[0020] FIG. 5A is a sensing region of a sensor for detection of
viable and nonviable cells where the sensing region is an
interdigital electrode structure with the electrode width and the
spacing between electrodes of 2 mm and sensing area of 2.times.1
cm;
[0021] FIG. 5B is a sensing region of a sensor for detection of
viable and nonviable cells where the sensing region is an
interdigital electrode structure with the electrode width and the
spacing between electrodes of 0.45 mm and sensing area of 2.times.1
cm;
[0022] FIG. 5C is a sensing region of a sensor for detection of
viable and nonviable cells where the sensing region is an
interdigital electrode structure with the electrode width and the
spacing between electrodes of 0.30 mm and sensing area of 2.times.1
cm;
[0023] FIG. 6A is an example of a bioreactor-integrated sensor for
detection of viable and nonviable cells according to an embodiment
of the disclosure;
[0024] FIG. 6B is an example of a bioreactor-integrated sensor for
detection of viable and nonviable cells according to an embodiment
of the disclosure;
[0025] FIG. 6C is an example of a bioreactor-integrated sensor for
detection of viable and nonviable cells according to an embodiment
of the disclosure;
[0026] FIG. 6D is an example of a bioreactor-integrated sensor for
detection of viable and nonviable cells according to an embodiment
of the disclosure;
[0027] FIG. 7 is a flow diagram of a method of in-line
determination of viable and nonviable cells in cell culture
according to an embodiment of the disclosure;
[0028] FIG. 8 shows an example of a plot of operation of the sensor
at five different resonant frequencies with an inset of a plot of
operation of the sensor at two highest resonant frequencies;
[0029] FIG. 9 is a graph of measured impedance parameters of an
embodiment of the resonant sensor, in accordance with embodiments
of the present technique;
[0030] FIG. 10A is a plot of a comparison of conventional and
resonance impedance detection with a linear vertical axis;
[0031] FIG. 10B is a plot of a comparison of conventional and
resonance impedance detection with a logarithmic vertical axis;
[0032] FIG. 11 is a plot of off-line analysis for a cell
culture;
[0033] FIG. 12 is a plot of in-line analysis of the cell culture of
FIG. 11 according to an embodiment of the disclosure;
[0034] FIG. 13A depicts the PCA scores plot of PC1 vs. PC2 upon
exposure of sensor to four solutions and performing conventional
impedance measurements;
[0035] FIG. 13B depicts the PCA scores plot of PC1 vs. PC2 upon
exposure of sensor to four solutions and performing resonance
impedance measurements;
[0036] FIG. 14A displays a plot of an estimate of signal to noise
for each factor (principal component) for measurements of four
solutions using conventional impedance;
[0037] FIG. 14B displays a plot of an estimate of signal to noise
for each factor (principal component) for measurements of four
solutions using resonant impedance;
[0038] FIG. 15 depicts results of dynamic measurements of a
solution over time when the changes in the solution properties were
measured using conventional impedance (two dotted lines) and using
resonant impedance with four resonators (two solid lines per
resonator);
[0039] FIG. 16A depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and no dielectric protective coating
as measured with conventional impedance;
[0040] FIG. 16B depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and with a 0.25 micrometers thick
dielectric protective coating as measured with conventional
impedance;
[0041] FIG. 16C depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and with a 0.5 micrometers thick
dielectric protective coating as measured with conventional
impedance;
[0042] FIG. 16D depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and no dielectric protective coating
as measured with resonant impedance;
[0043] FIG. 16E depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and with a 0.25 micrometers thick
dielectric protective coating as measured with resonant
impedance;
[0044] FIG. 16F depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.45 mm and with a 0.5 micrometers thick
dielectric protective coating as measured with resonant
impedance;
[0045] FIG. 17A depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and no dielectric protective coating
as measured with conventional impedance;
[0046] FIG. 17B depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and with a 0.25 micrometers thick
dielectric protective coating as measured with conventional
impedance;
[0047] FIG. 17C depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and with a 0.5 micrometers thick
dielectric protective coating as measured with conventional
impedance;
[0048] FIG. 17D depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and no dielectric protective coating
as measured with resonant impedance;
[0049] FIG. 17E depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and with a 0.25 micrometers thick
dielectric protective coating as measured with resonant impedance;
and
[0050] FIG. 17F depicts the PCA scores plot of PC1 vs. PC2 upon
measurements of four solutions using a sensor with an interdigital
electrode structure with the electrode width and the spacing
between electrodes of 0.3 mm and with a 0.5 micrometers thick
dielectric protective coating as measured with resonant
impedance.
DETAILED DESCRIPTION
[0051] The present sensing techniques facilitate simultaneous
independent and real time determination of viable and nonviable
cells using a sensor directly in a cell culture. Real time
identification and quantitation of both viable and nonviable cells
in bioreactors using in-line sensors without the need for periodic
sampling of the bioreactor is beneficial for biopharmaceutical
process operations. At present, assessing a parameter such as cell
mass is performed with cell mass sensors that provide information
only about the viable cells and do not provide additional
information about the populations and concentrations of viable as
well as nonviable cells in cell culture. Further, integrating
increased functionality with in-line sensors provides the benefit
of increased sterility and decreased opening of reaction containers
for sampling. As depicted in FIG. 1, when sampling from a reaction
container (bioreactor) is performed, the sample is further
transferred to a stand-alone sensor for off-line analysis. The
off-line analysis has the risk of contamination of the reaction
container because of the periodic sampling. The off-line analysis
also requires the removal of a certain fluid volume from the
reaction container for the off-line analysis. Unlike these problems
with off-line analysis, the in-line analysis does not have these
limitations, eliminates the risk of culture contamination due to
the periodic sampling, and facilitates an increased sterility.
[0052] Conventional measurements of concentration of viable cells
(also known as cells biomass) rely on capacitance measurements at
given frequencies or rely on the scanning dielectric spectroscopy
or scanning radio-frequency spectroscopy. The viable cells have
intact plasma membranes and contribute to the change in the
relative permittivity of the cell culture of the beta-dispersion
frequency range of cells. In contrast, the nonviable cells have
leaky membranes that do not contribute to the change in the
relative permittivity of the cell culture of the beta-dispersion
frequency range of cells and, thus, do not produce change in the
capacitance signal.
[0053] One technique for assessing viable cell concentration is the
changes in capacitance or the relative permittivity. The relative
permittivity is calculated from the capacitance measurements taking
into the account the physical constant of the electrodes of the
sensor. Another technique is the inductive dielectric spectroscopy
where the electrodes are assembled as toroidal coils and are
separated from each other by the liquid. The geometrical parameters
of toroidal coils are inner and outer coil diameters, coil
thickness, and coil-to-coil separation distance. Yet another
technique of measurements of dielectric properties of samples
involves resonator methods. However, such techniques cannot
quantify concentrations of viable and nonviable cells. The bulk
conductivity of the cell culture depends on many fermentation
factors even at the constant concentration of viable cells when the
concentration of viable cells reaches the saturation.
[0054] Provided herein are techniques for identification and
quantitation of viable and nonviable cells in a culture that
include measuring a resonance impedance spectral response of at
least one resonator in proximity to the cell culture and
correlating the first measured resonance impedance sensor response
to the concentration of viable cells in cell culture; and
correlating a second measured sensor resonance impedance response
to the concentration of nonviable cells in cell culture.
[0055] The disclosed embodiments may be used for simultaneous
in-situ independent determination of viable and nonviable cells in
cell culture of any scale. A cell culture reaction may mean a fluid
undergoing a process to form a cell culture, such that sensors as
disclosed may be in contact with a fluid undergoing a reaction
process. The cell culture may be in an industrial or lab scale
reactor or, alternatively, may take place in small scale reactors
or microtiter well plates. The cell culture fluid may include the
medium (e.g., a liquid or gel containing suitable nutrient sources
such as amino acids, glucose, salts, etc.) and cells. The cell
culture fluid can be related to the fermentation of different
cells. Nonlimiting examples of fermentation of cells include animal
cells, mammalian cells, plant cells, bacteria, yeasts, fungi. The
fermentation is performed in a batch or in a continuous mode.
According to the present techniques, information about
concentrations of viable and nonviable cells is obtained without
disturbing the fermentation process and without taking samples for
off-line analysis of cells. For example, the embodiments may be
used in conjunction with systems that include a bioreactor, at
least one resonant sensor, a sensor reader, and a data processor.
To that end, FIG. 2 illustrates a system 10 for assessing cells
that includes a bioreactor 12 and an in-line resonant sensor 14.
The resonant sensor 14 may be disposed in or on the bioreactor 12
or may be coupled to in-line medical tubing or connectors in fluid
communication with the bioreactor 12. In particular embodiment, the
resonant sensor 14 is configured to provide continuous or
intermittent monitoring of the cell growth within the bioreactor.
The bioreactor 12 may be any suitable cell culture reaction vessel
that facilitates cell growth. For example, the bioreactor may be a
large or small-scale reactor, a bag reactor, a tank reactor,
etc.
[0056] The resonant sensor 14 is configured to detect chemical,
physical or biological parameters of a sample via the resonance
impedance spectral response of at least one resonator or resonant
circuit. As opposed to simple impedance measurements, the disclosed
embodiments use the sample to be probed with at least one resonant
electrical circuit. The resonance impedance spectrum of the sensor
in proximity to the sample (the sensor in the operational contact
with a cell culture reaction fluid) varies based on sample
composition. The measured resonance impedance values Z' (which is
the real part of impedance, Zre) and Z'' (which is the imaginary
part of impedance, Zim) reflect the response of the cell culture
sample (for example, the portion of the cell culture in proximity
to the sensor 14) to a stimulus of the electric field of the
resonant electrical circuit. The electrical field may be applied by
the sensor 14 via electrodes, which may be in direct or indirect
electrical contact with the sample. For example, a resonant sensor
may be a combination of a sensing region and tuning circuits. The
sensing region is either bare or coated with a protective
dielectric layer (e.g., as shown in the embodiment of FIG. 3). In
both cases, the sensing region may be considered to be in
"operational contact" with cell culture fluid. In such embodiments,
the tuning circuits are not in operational contact with cell
culture fluid. Indirect electrical contact with the sample is when
sensing electrode structure is conformally coated with a dielectric
protective coating and when the electric field that is generated
between the electrodes interacts with the cell culture reaction
fluid after penetrating through the dielectric protective
coating.
[0057] The resonant sensor 14 may be configured as disclosed in
U.S. patent application Ser. Nos. 13/729,800 and 13/729,851 to
Potyrailo et al., and filed on Dec. 28, 2012, which are
incorporated by reference herein in their entirety for all
purposes. In one embodiment, the resonant sensor 14 may be a single
use sensor that is used during all or part of a reaction process.
For example, during a single reaction, the resonant sensor 14 may
come into operational contact with the cell culture and, therefore,
be unsuitable for use in other reactions. Further, the resonant
sensor 14 may be associated with a particular reactor and may be
cleaned and sterilized or replaced according to a maintenance
schedule. For example, the resonant sensor 14 may include one or
more pairs of electrodes and one or more tuning elements, e.g., a
resistor, a capacitor, an inductor, a resonator, impedance
transformer, or combinations thereof to form an
inductor-capacitor-resistor (LCR) resonant circuit operated at at
least one resonant frequency. In certain embodiments, different
resonant circuits of a plurality of resonant circuits of a resonant
sensor 14 may be configured to resonate at different
frequencies.
[0058] The different resonant circuits may be configured to probe
the growing cells in the bioreactor 12 with a plurality of
frequencies. Further, the different frequencies may be used to
probe a fluid sample at different depths allowing for use in either
adhered or suspension cell culture systems. In certain embodiments,
an integrated circuit memory chip may be galvanically coupled to
the resonant sensor 14. The integrated circuit memory chip may
contain different types of information. Nonlimiting examples of
such information in the memory of the integrated circuit chip
include calibration coefficients for the sensor, sensor lot number,
production date, end-user information. In another embodiment, the
resonant sensor 14 is an interdigital structure that is a part of
the resonator and has a sensing region for detection of cells in
the culture. Turning back to FIG. 2, data from the resonant sensor
14 may be acquired via data acquisition circuitry 16, which may be
associated with the sensor 14 or which may be associated with a
control system, such as a monitor or workstation 22 including data
processing circuitry, where additional processing and analysis may
be performed. The data acquisition circuitry 16 can be within the
bioreactor 12 as shown in FIG. 2 or can be within the workstation
22. Further, the workstation 22 can be replaced with the control
system of the whole bioprocess factory (not shown in FIG. 2) where
the resonant sensor 14 and its data acquisition circuitry 16 are
connected to the control system of the whole bioprocess factory.
The data acquisition circuitry 16 may be in the form of a sensor
reader, which may be configured to communicate wirelessly with the
bioreactor 12 and/or the workstation 22. For example, the sensor
reader may be a battery-operated device.
[0059] In addition, the data acquisition circuitry 16 may receive
data for one or more resonant sensors 14, (e.g. multiple sensors
formed in an array or multiple sensors positioned at different
locations in or around the bioreactor 12). The data may be stored
in short or long term storage devices, such as archiving
communication systems, which may be located within or remote from
the system 10 and/or reconstructed and displayed for an operator,
such as at the operator workstation 22. Nonlimiting examples of
positioning and installations of sensors and sensor systems of the
present techniques include bioreactors, individual wells of
microtiter well plates, connectors, flow-through components, and
any other relevant bioprocess components.
[0060] In addition to displaying the data, the operator workstation
22 may control the above-described operations and functions of the
system 10. The operator workstation 22 may include one or more
processor-based components, such as general purpose or application
specific computers 24. In addition to the processor-based
components, the computer 24 may include various memory and/or
storage components including magnetic and optical mass storage
devices, internal memory, such as RAM chips. The memory and/or
storage components may be used for storing programs and routines
for performing the techniques described herein that are executed by
the operator workstation 22 or by associated components of the
system 10. Alternatively, the programs and routines may be stored
on a computer accessible storage and/or memory remote from the
operator workstation 22 but accessible by network and/or
communication interfaces present on the computer 24. The computer
24 may also comprise various input/output (I/O) interfaces, as well
as various network or communication interfaces. The various I/O
interfaces may allow communication with user interface devices,
such as a display 26, keyboard 28, mouse 30, and printer 32, that
may be used for viewing and inputting configuration information
and/or for operating the imaging system 10. The various network and
communication interfaces may allow connection to both local and
wide area intranets and storage networks as well as the Internet.
The various I/O and communication interfaces may utilize wires,
lines, or suitable wireless interfaces, as appropriate or
desired.
[0061] FIG. 3 illustrates a non-limiting example of a design of the
resonant sensor 14. A sensing electrode structure 34 of the sensor
is connected to the tuning circuits (see FIG. 4) and the data
acquisition circuitry 16. The sensing electrode structure 34 can be
bare and in direct contact with the cell culture reaction fluid.
The sensing electrode structure can be conformally coated with a
dielectric protective coating 36. The sensing electrode structure
34 without or with the conformal dielectric protective coating 36
forms a sensing region 38. The sensing electrode structure 34
without or with the conformal dielectric protective coating 36 that
forms a sensing region 38 is in operational contact with a cell
culture reaction fluid. An "operational contact" of a sensing
electrode with a cell culture reaction fluid may refer to
embodiments in which the sensing electrode structure 34 is either
without (bare) or with the conformal dielectric protective coating
36.
[0062] When the sensing electrode structure 34 is bare, the
electric field that is generated between the electrodes interacts
directly with the cell culture reaction fluid. When the sensing
electrode structure is conformally coated with a dielectric
protective coating 36, the electric field that is generated between
the electrodes interacts with the cell culture reaction fluid after
penetrating through the dielectric protective coating 36.
[0063] FIG. 4 illustrates a portion of a resonant sensor system
employing a sensor assembly 40 configured to probe a fluid sample
such as cell culture reaction fluid using a plurality of
frequencies. The resonant sensor assembly 40 includes the resonant
sensor 14. The sensor comprises a single sensing region 38. The
sensing region 38 may be disposed on a substrate. In some
embodiments, the substrate of the sensor 14 may be a dielectric
substrate. The substrate may be a microtiter well plate. In this
embodiment, the electrodes may be deposited on the microtiter well
plate. A well of a microtiter well plate is a non-limiting example
of an open sample container or an open flow-channel.
[0064] In certain embodiments, the sensor assembly 40 further
includes a plurality of tuning elements 42. The plurality of tuning
elements may be operatively coupled to the single sensing region 38
to define a plurality of resonant circuits. The tuning elements 42
along with the single sensing region 38 may define a plurality of
resonant circuits. Each resonant circuit of the plurality of
resonant circuits may include one or more tuning elements of the
plurality of tuning elements.
[0065] In the illustrated embodiment, the plurality of tuning
elements 42 is external to the sensor 14. However, in one
embodiment, the tuning elements 42 may be disposed on the substrate
of the sensor 14. In another embodiment, some of the plurality of
tuning elements 42 may be external to the sensor substrate, while
other tuning elements 42 may be disposed on the substrate. The
tuning elements 42 may comprise a resistor, a capacitor, an
inductor, a resonator, impedance transformer, or combinations
thereof.
[0066] Each resonant circuit may be configured to resonate at a
particular frequency. At least one resonant circuit may be
configured to resonate at a frequency that is different from the
resonating frequency of the other resonant circuits. By way of
example, if the sensing region 38 includes a pair of electrodes,
the tuning elements 42 may be a resistor, a capacitor, and an
inductor to form an inductor-capacitor-resistor (LCR) resonant
circuit. The tuning elements 42 may be electrically coupled to the
sensing region 38. In one embodiment, the tuning elements 42 may be
in parallel connection to the sensing region 38. In certain
embodiments, the different resonant circuits of the plurality of
resonant circuits may be configured to resonate at different
frequencies. The different resonant circuits may be configured to
probe the fluid sample with a plurality of resonant frequencies.
The different resonant frequencies may be used to probe a fluid
sample over the frequency range of spectral dispersions of the cell
culture. The spectral dispersions of the cell culture that are
monitored with the sensors of the present disclosure are over the
frequency range from 0.1 Hz to 100 GHz and include alpha, beta,
gamma, and delta spectral dispersions.
[0067] In the illustrated embodiment, the sensor assembly 10 may
also include a multiplexer 44. The multiplexer 44 may be configured
to facilitate electronic switching between the plurality of tuning
elements 42. The multiplexer 44 may be configured to select one or
more signals associated with the probing frequencies and forward
the selected signal to an output device or a reader. In one
embodiment, the multiplexer 44 may be configured to selectively
send signals to an output device or a reader. The multiplexer 44
may be configured to send a plurality of signals simultaneously to
a sensor reader.
[0068] FIGS. 5A, B, and C illustrate non-limiting examples of
electrode designs. These electrodes are interdigital
(interdigitated) electrode structures with the equal electrode
width and the spacing (gap) between electrodes where the spacing
between electrodes can be same or different in different directions
as well known in the art. FIG. 5A depicts a sensing region of a
sensor for detection of viable and nonviable cells where the
sensing region is an interdigital two-electrode structure with the
electrode width and the spacing between electrodes of 2 mm and
sensing area of 2.times.1 cm. FIG. 5B depicts a sensing region of a
sensor for detection of viable and nonviable cells where the
sensing region is an interdigital two-electrode structure with the
electrode width and the spacing between electrodes of 0.45 mm and
sensing area of 2.times.1 cm. FIG. 5C depicts a sensing region of a
sensor for detection of viable and nonviable cells where the
sensing region is an interdigital two-electrode structure with the
electrode width and the spacing between electrodes of 0.3 mm and
sensing area of 2.times.1 cm.
[0069] Other known examples of interdigital electrodes that can be
used include electrode structures with variable electrode width and
the spacing between electrodes, tapered electrodes, circular
electrodes, and others known in the art. The sensor as provided may
have at least two electrodes, but also can have four electrodes or
more electrodes.
[0070] The resonant sensor 14 is positioned in the operational
contact with the cell culture reaction fluid. At least one sensor
generating a plurality of frequencies is inserted or mounted onto a
non-rigid, flexible container via at least one hose barb port or
any other mounting device such that the sensor is held aseptically
within the internal space of the container and is in operational
contact with fluid within the container. The sensor 14 is further
connected to a reader device external to the non-rigid, flexible
container by direct electrical connection or remote means. In one
embodiment, the non-rigid, flexible container is a bioreactor. In
another embodiment, a bioreactor is a rigid, non-flexible
container.
[0071] FIG. 6A-D depict examples of bioreactor-integrated sensors
for detection of viable and nonviable cells. In one embodiment
depicted in FIG. 6A, the resonant sensor 14 is assembled into an
insert. The assembly of the resonant sensor with the insert is
further integrated into the bioreactor using a hose barb port. The
sensing region of this sensor desirably extends into the bioreactor
for at least 10 millimeters.
[0072] In another embodiment depicted in FIG. 6B, the resonant
sensor is assembled into an insert. The assembly of the resonant
sensor with the insert is further integrated into the bioreactor
using a hose barb port. The sensing region of this sensor desirably
extends into the bioreactor for at most 2 millimeters.
[0073] In another embodiment depicted in FIG. 6C, the resonant
sensor is assembled into an insert. The assembly of the resonant
sensor with the insert is a patch assembly. The sensing region of
this sensor desirably extends into the bioreactor for at least 10
millimeters.
[0074] In another embodiment depicted in FIG. 6D, the resonant
sensor is assembled into an insert. The assembly of the resonant
sensor with the insert is a patch assembly. The sensing region of
this sensor desirably extends into the bioreactor for at most 2
millimeters. In another embodiment, the resonant sensor is
assembled into an insert to form a patch where the patch has the
ratio of its diameter to its thickness of at least 10:1 and in one
embodiment at least 20:1, and in another embodiment at least 50:1.
Some additional non-limiting examples of integration of the sensor
into the bioreactor are disclosed in U.S. Pat. No. 8,508,368,
"Disposable sensing device having radio frequency based sensor," to
Potyrailo et al., which is incorporated by reference in its
entirety herein for all purposes.
[0075] In any of the disclosed embodiments, the sensor signal is
communicated out from the sensor in a suitable manner. For example,
in one approach, the sensor signal is communicated via a wired link
to the sensor reader and then via a wired link or a wireless link
to workstation. In another approach, the sensor signal is
communicated via a wired link or a wireless link to the sensor
reader and then via a wired link or a wireless link to workstation.
"Wireless link" may refer to the communication of an analog or a
digital signal via inductive coupling, capacitive coupling, or
electromagnetic coupling using radio waves in the radio frequency
range from about 30 kHz to 300 GHz.
[0076] FIG. 7 is a flow diagram 50 of a method 50 for simultaneous
independent determination and quantitation of viable and nonviable
cells in cell culture. At step 52, the sensor 14 measures the
resonance impedance spectral response of at least one resonator
coupled to the cell culture (e.g., associated with the bioreactor
12). A resonator is a sensor comprising a sensing region and at
least one tuning circuit that forms a resonant circuit structure or
a resonator. Thus, the resonator coupled to the cell culture is the
resonant circuit structure where the sensing region is in
operational contact with a cell culture reaction fluid.
[0077] At step 54, the method 50 performs feature selection of the
measured impedance spectral response and analysis of selected
features at step 56. The impedance spectral response can include
alpha, beta, gamma, and delta spectral dispersions of cells in a
cell culture fluid. Based on the analysis, the sensor response is
correlated to the concentration of nonviable cells (step 58) and
the concentration of viable cells (step 60). The method 50 may be
configured also to provide information about the size of the viable
cells. Information about size of viable cells is important for
different applications, for example in protein production for
vaccines.
[0078] The method 50 may be configured to provide outputs to
downstream devices or to operators of the system 10. For example,
based on the results of the method 50, the system 10 may provide
visual indication or readouts to an end user of total numbers of
cells, a number or percentage of viable cells, a number or
percentage of nonviable cells, a ratio of nonviable cells to viable
cells, plots or graphs of viable and nonviable cells over time, or
any combination thereof.
[0079] In another embodiment, the results of the method 50 may be
used to control reaction conditions in the bioreactor 12. For
example, if the number of nonviable cells is above a predetermined
threshold, the method 50 may also include steps for triggering
alarms. In other embodiments, if the total number of cells and the
number of viable cells are both above a threshold, the system 10
may be configured to provide instructions for harvesting the cell
culture reaction. Further, the results may be used to control
reaction conditions (e.g., changing gas and/or feed mixtures,
temperature, other conditions) to facilitate a desired ratio of
nonviable cells to viable cells.
[0080] In particular, the signal acquired from the resonant sensor
may include a resonance impedance response spectrum of the sensor
from each resonant circuit (e.g., if the sensor 14 includes or
operates only a single circuit, the resulting signal includes one
spectrum). A plurality of such impedance response spectra may be
generated by a plurality of resonant circuits. In one embodiment,
viable cells may exhibit a characteristic resonant response that
may be distinguished from the resonant response of nonviable cells
based on the impedance spectral response that can include alpha,
beta, gamma, and delta spectral dispersions of cells in a cell
culture fluid. Accordingly, the acquired impedance spectrum or
spectra for characteristic patterns may be associated with a
particular percentage viability vs. nonviability of cells.
[0081] FIG. 8 illustrates the operation of a sensor comprising a
sensing region and a plurality of tuning circuits that provide five
resonances over the frequency range of spectral dispersions of the
cell culture. In a non-limiting example, five resonances of the
single sensing region are provided by the electronic circuit of the
sensor that has five tuned inductor values of 100, 10, 1, 0.1 and
0.01 mH. These inductor values provide the resonances of the single
sensing region over the frequency range depicted in FIG. 8.
[0082] As depicted in FIG. 9, in one embodiment, the system 10 may
be configured to measure a resonant impedance {hacek over (Z)}(f)
(represented by Eq. (1)) of the sensor.
{hacek over (Z)}(f)=Z.sub.re(f)+jZ.sub.im(f) Eq. (1)
where Z.sub.re(f) is the real part of the resonant impedance and
Z.sub.im(f) is the imaginary part of the resonant impedance. In
certain embodiments, the impedance response of the sensor may be a
multivariable response as more than one frequency may be utilized
to measure sensor response across the resonance of the sensor. In
certain embodiments, the impedance response of the sensor may be a
multivariable response because more than one frequency may be
utilized to measure sensor response outside the resonance peak of
the sensor. In some embodiments, the sensor response is measured at
multiple frequencies across the resonance of the sensor. For
example, if the sensor resonates at about 1 MHz, the measured
frequencies and associated sensor responses are measured from about
0.25 MHz to about 2 MHz. This multivariable response is analyzed by
multivariate analysis. The multivariable response of the sensor
includes the sensor's full impedance spectra and/or several
individually measured properties, such as but not limited to
F.sub.p, Z.sub.p, F.sub.z, F.sub.1, F.sub.2, Z.sub.1, and Z.sub.2.
FIG. 9 depicts a graph of measured impedance parameters of an
embodiment of the resonant sensor, in accordance with embodiments
of the present technique. These and other measured properties are
"spectral parameters". These properties include the frequency of
the maximum of the real part of the impedance (F.sub.p, resonance
peak position), magnitude of the real part of the impedance
(Z.sub.p, peak height), zero-reactance frequency (F.sub.z,
frequency at which the imaginary portion of impedance is zero),
resonant frequency of the imaginary part of the impedance
(F.sub.1), and anti-resonant frequency of the imaginary part of the
impedance (F.sub.2), signal magnitude (Z.sub.1) at the resonant
frequency of the imaginary part of the impedance (F.sub.1), and
signal magnitude (Z.sub.2) at the anti-resonant frequency of the
imaginary part of the impedance (F.sub.2). Other parameters may be
measured using the entire impedance spectra, for example, quality
factor of resonance, phase angle, and magnitude of impedance.
Multivariable response spectral parameters are described in U.S.
Pat. No. 7,911,345 entitled "Methods and systems for calibration of
RFID sensors", which is incorporated herein by reference in its
entirety for all purposes.
[0083] By using multivariate analysis of full spectra and/or
calculated parameters of {hacek over (Z)}(f) spectra, quantitation
of concentrations of viable and nonviable cells is performed. Also,
by using multivariate analysis of full spectra and/or calculated
parameters of {hacek over (Z)}(f) spectra, determination of size of
viable cells and quantitation of concentrations of viable and
nonviable cells is performed.
[0084] FIGS. 10A and 10B is a plot of a comparison of conventional
and resonance impedance detection during in-line measurements of
the varying concentration of cells in the culture. Over the course
of these measurements, the conventional and resonance impedance
spectra were collected in a rapid sequence. The resonance impedance
spectra were generated using four different inductors 1-4 that
provided sensor resonances ranging from the smallest frequency to
the largest frequency resonances, respectively. Over the time of
this experiment, numerous spectra were collected. In FIGS. 10A and
B, five time points were chosen to depict the spectral features of
the conventional and resonance impedance spectra during the
monitoring of the varying concentration of cells in the culture.
For clarity, results are illustrated only for the real part Z' (or
Z.sub.re) of impedance. FIG. 10A shows the data with the linear
vertical axis, while FIG. 10B shows the data with the logarithmic
vertical axis for better depiction of the responses of resonators 3
and 4. Spectra from conventional impedance detection are labeled as
L0. Spectra from the resonance impedance detection using inductors
1-4 are labeled as L1, L2, L3, and L4. Individual groups of five
spectra of responses L0, L1, L2, L3, and L4 show an increase in
their spectral intensity (or impedance) upon an increase of the
concentration of cells in the culture.
[0085] The resonance impedance detection provides enhanced signal
diversity and sensitivity of measurements in cell cultures. First,
the signal changes of the resonant responses L1, L2, L3, and L4
(represented as data profiles) are more significant at any
particular frequency relative to the conventional impedance data L0
profile. Second, the resonant responses are more diverse in their
profiles as illustrates in the changes of the shapes of the
resonances L1, L2, L3, and L4.
[0086] FIG. 11 depicts results of the off-line analysis of growth
of cells in a typical cell culture run. The cells in this cell
culture run were mammalian cells. The off-line analysis was
performed using a NucleoCounter NC-100 and it served as a reference
data for the evaluation of the developed in-line sensor. Off-line
analysis was performed by periodic sampling of the cell culture and
performing measurements of the concentration of viable cells, the
concentration of nonviable cells, total number of cells (as the sum
of viable and nonviable cells), and percent viability of cells (as
the ratio of viable to total number of cells multiplied by 100%). A
resonant sensor was also positioned directly in this cell culture
and in-line measurements were performed with this sensor. FIG. 12
depicts results of the in-line analysis of growth of cells as
collected with the resonant sensor. Quantitation of the
concentration of viable cells, the concentration of nonviable
cells, total number of cells, and percent viability of cells was
performed using a well-known partial least squares (PLS) technique
of analysis of spectra. The PLS determines correlations between the
independent variables and the instrument response by finding the
direction in the multidimensional space of the instrument response
that explains the maximum variance for the independent variables. A
comparison of off-line measurement (FIG. 11) and in-line
measurement (FIG. 12) results illustrates a desired correlation
between measurements done by off-line and in-line instruments such
as measurements of the concentration of viable cells, the
concentration of nonviable cells, total number of cells, and
percent viability of cells.
[0087] In the past, PLS has been used successfully for multivariate
analysis of data from an in-line sensor for biomass monitoring of
cell culture with scanning dielectric spectroscopy. However, using
the earlier reported sensors and univariate and multivariate
analysis of data, concentrations of only viable cells were
determined quantitatively. Unfortunately, using the earlier
reported sensors and univariate and multivariate analysis of data,
concentrations of nonviable cells were not determined
quantitatively. Only qualitative information was possible to obtain
using earlier reported sensors and multivariate analysis of data
because the collected spectral data had only at most two
multivariable factors. When scanning dielectric spectroscopy data
was fitted to a capacitance change of the measured culture, only
viable cells were quantitated without ability to quantify nonviable
cells. In contrast to those earlier results, the sensors and
multivariate analysis of data according to the present techniques
provide the quantitative information of not only the viable cells
but also the nonviable cells.
[0088] In one embodiment, the resonant sensor 14 may be calibrated
based on the analysis of off-line parameters or exposure to
bioreactors 12 with known percentages of viable cells, upon
exposure to a cell culture medium without cells, and upon exposure
to a cell culture medium with a known concentration of viable
cells. The response of the one or more sensors 14 is measured and
an analytical relationship between the environmental parameter and
the sensor response is established. The analytical fit coefficients
may be calculated using multivariate calibration. In multivariate
calibration, more than one property of the sensor response is
related to the value of the environmental parameter of interest.
Multivariate calibration utilizes the full impedance spectra for
calibration, or at least two of individually measured parameters
(Zp, Fp, Fz, F1, F2, Z1, Z2), or at least two of any other
parameters that can be extracted from the response of the resonance
circuit of the sensor 14. Nonlimiting examples of these additional
parameters are quality factor of resonance, phase angle, and
magnitude of impedance of the resonance circuit response sensor 14.
Nonlimiting examples of multivariate analysis tools are canonical
correlation analysis, regression analysis, nonlinear regression
analysis, principal components analysis, discriminate function
analysis, multidimensional scaling, linear discriminate analysis,
logistic regression, pattern matching, and/or neural network
analysis. Multivariate calibration can be performed using spectra
from conventional impedance measurements, using spectra from
resonance impedance measurements, and/or their combinations.
Spectra from resonance impedance measurements can include all
utilized resonator responses or only a subset of these
responses.
[0089] The conventional and resonance impedance measurements were
compared by measuring sensor responses in model solutions with
different solution conductivity and dielectric constant. Such
measurements provided the opportunity to quantitatively compare the
response of the sensor in its conventional impedance measurement
mode and resonant impedance mode. For these measurements, four
model solutions were prepared, each having a volume of 200 mL for
testing of the sensor response. Solution 1 had 200 mL of water with
conductivity of 2 mS/cm. Solution 2 had 200 mL of water with
conductivity of 15 mS/cm. Solution 3 had 180 mL of water with
conductivity of 2 mS/cm and 20 mL of ethanol. Solution 4 had 180 mL
of water with conductivity of 15 mS/cm and 20 mL of ethanol. The
dielectric constants of pure water and pure ethanol at room
temperature (20.degree. C.) are 80.1 and 24.5, respectively. The
dielectric constants of water/ethanol mixtures have respectively
changing dielectric constants. Thus, these four model solutions
provided the ability to test the response of the sensor to the
changes in solution conductivity and dielectric constant. All
impedance measurements were done over the range from 1,000 Hz to
10,000,000 Hz. Real and imaginary conventional impedance and
resonance impedance spectra of sensor response in four solutions
were further processed using a well-known multivariate analysis
tool such as principal components analysis (PCA) to determine the
similarities and differences between these spectra. Multivariate
analysis was performed using a common PLS_Toolbox software
(Eigenvector Research, Inc., Wenatchee, Wash.) operated with Matlab
software (The Mathworks Inc., Natick, Mass.). The PCA pattern
recognition method explains the variance of the data as the
weighted sums of the original variables, known as principal
components (PCs) or factors. In the past, PCA has been used
successfully for multivariate analysis of data from an in-line
sensor for biomass monitoring of cell culture with scanning
dielectric spectroscopy, where only two PCs were observed from the
measured spectra.
[0090] FIGS. 13A and B depict the PCA scores plot of PC1 vs. PC2
upon exposure of sensor to four solutions and performing
conventional and resonant impedance measurements, respectively.
FIG. 13A and FIG. 13B show four clusters of sensor response to four
solutions. The clusters are labeled 1, 2, 3, and 4, corresponding
to the measured solution. Each cluster has two data points, which
are replicates (n=2) of spectral measurements.
[0091] FIG. 13A shows that measurements of four solutions using
conventional impedance gave good discrimination ability between
solutions of low and high conductivity (discrimination between
solutions 1 and 2 and solutions 3 and 4). Also, good discrimination
ability between different dielectric constants is observed when
conductivity of solutions was low (solutions 1 and 3).
Unfortunately, discrimination between different dielectric
constants when conductivity of solutions was high (solutions 2 and
4) significantly degraded. Data points in clusters 2 and 4 were
much closer in the PCA plot as compared to other clusters. In
contrast, measurements of four solutions using resonance impedance
gave good discrimination ability between all four solutions as
depicted in FIG. 13B. The spectral analysis of resonance impedance
was performed on all measured resonance peaks. Importantly, the
quality of discrimination between solutions with different
dielectric constants was similar when solutions were of low
dielectric constant (solutions 1 and 3) and high dielectric
constant (solutions 2 and 4) as evidenced by the large distances
between the respective clusters 1, 2, 3, and 4 (FIG. 13B).
[0092] The number of significant factors (principal components) in
multivariate data from conventional and resonance impedance
measurements was further determined using a common algorithm in
PLS_Toolbox software (Eigenvector Research, Inc., Wenatchee, Wash.)
operated with Matlab software (The Mathworks Inc., Natick, Mass.).
This common algorithm provides an estimate of the number of
significant PCA factors for the measured data based on their signal
to noise and displays a plot of an estimate of signal to noise for
each principal component (factor). The signal to noise of 2 or
below of a PC is known to be dominated by noise, while the signal
to noise above 3 of a PC is acceptable. The number of factors
needed to describe the data is the number of eigenvectors with
signal to noise of 3 or greater.
[0093] FIG. 14A displays a plot of an estimate of signal to noise
for each factor (principal component) for measurements of solutions
1-4 using conventional impedance. This plot shows that only the
first three factors (principal components) had the signal to noise
of three or more, acceptable for a reliable multivariate analysis
and quantitations.
[0094] FIG. 14B displays a plot of an estimate of signal to noise
for each factor (principal component) for measurements of solutions
1-4 using resonant impedance. This plot shows that the first five
factors (principal components) had the signal to noise of three or
more, acceptable for a reliable multivariate analysis and
quantitations. Further, additional factors (principal components)
such as PC 6 and PC7 can be also used for multivariate analysis and
quantitations because their signal to noise was 3.
[0095] Thus, the comparison of signal to noise of different PCs
from measurements performed using conventional impedance (FIG. 14A)
and resonant impedance (FIG. 14B) showed the higher dimensionality
of the multivariable response of the resonant impedance
measurements as compared to that of the conventional impedance
measurements. The higher dimensionality of the multivariable
response of the sensor is related to the higher selectivity of the
sensor and sensor ability to more selectively determine larger
number of independent environmental conditions affecting the sensor
as compared to the sensor with the lower dimensionality of its
multivariable response.
[0096] In the past, PCA has been used successfully for multivariate
analysis of data from an in-line sensor for biomass monitoring of
cell culture with scanning dielectric spectroscopy, where only two
PCs were observed from the measured spectra. Unlike those earlier
results, the sensors may have three (FIG. 14A) and even up to seven
(FIG. 14B) PCs useful for the multivariate analysis of frequency
spectral data collected using sensors according to the disclosed
embodiments.
[0097] Detection of viable and nonviable cells can be also
described by using an equivalent circuit model of the sensor that
includes the cell resistance and capacitance, double layer
resistance and capacitance, and solution bulk resistance, as well
as the initial sensor resistance, capacitance, and inductance. The
exact values of the components of the equivalent circuit are varied
in relation to the design of the electrode structure and the
existence and type of the protective layer on the sensing region to
separate the sensor electrodes from the direct contact with the
solution. The equivalent circuit is described by an analytical
expression that includes the terms for the real and imaginary
portions of at least one resonance impedance spectra of the sensor.
The terms in the derived analytical expression are correlated with
the total cell concentration and with the viable cell
concentration. The terms in the derived analytical expression can
be related to the spectral dispersions of the cell culture over the
frequency range from 0.1 Hz to 100 GHz and can include alpha, beta,
gamma, and delta spectral dispersions.
[0098] High detection sensitivity of the resonant impedance
measurements has been demonstrated in the present disclosure. When
compared to the detection sensitivity of the conventional impedance
measurements, sensitivity enhancement of 45-74 fold was achieved
depending on the resonator used for sensing. FIG. 15 depicts
results of dynamic measurements of a solution over time when the
changes in the solution properties were measured using conventional
impedance (two dotted lines) and using resonant impedance with four
resonators (two dotted lines per resonator). The enhancement of the
measurement sensitivity using resonant impedance over conventional
impedance was calculated as the impedance response Zre difference
between responses measured using a particular resonator divided by
the difference between responses measured using conventional
impedance at the same frequency range as the resonator. For
example, the resonator that operated at the smallest resonant
frequency had Zre responses ranging from 58,000 Ohm to 12,500 Ohm
over the time of the experiment, while Zre responses measured using
conventional impedance were in the range from 1,800 to 800 Ohm over
the same experiment time. Thus, the enhancement of sensitivity of
the resonant impedance over the conventional impedance measurements
was calculated as (58,000-12,500)/(1,800-800)=45. Similarly
calculated enhancements of the sensitivity of resonant impedance
over conventional impedance measurements for other resonators 2, 3,
and 4 were 62, 74, and 46 fold, respectively. Thus, the use of
resonant impedance measurements provided a very significant
enhancement in sensitivity of measurements over the range of
beta-dispersion frequency range of cells. The use of resonant
impedance measurements can further provide a significant
enhancement in sensitivity of measurements over the range of alpha,
beta, gamma, and delta spectral dispersions. The sensor used in
this experiment was an interdigital electrode structure coated with
a dielectric protective layer. Nonlimiting examples of electrode
structures include two- and four-electrode structures. Nonlimiting
examples of interdigital electrode structures include two- and
four-electrode structures. Nonlimiting examples of materials for
electrodes include stainless steel, platinum, gold, noble metals,
and others. Nonlimiting examples of materials of a dielectric
protective layer include silicon dioxide, silicon nitride,
parylene, silicone, fluorinated polymers, ceramics, and others.
Nonlimiting examples of fabrication methods of electrodes include
metal etching and mask-based metal deposition. Thickness of
fabricated electrodes on the substrates is in the range from 10
nanometers to 1000 micrometers.
[0099] The importance of the appropriate geometry of the sensing
electrode and the associated protective layer has been further
evaluated. Two geometries of interdigital electrode structures were
fabricated as 1.times.2 cm structures. These geometries of
electrodes are depicted in FIG. 5B and FIG. 5C. In these two types
of structures, the electrode width was same as the spacing between
electrodes and was 0.45 mm and 0.30 mm. Parylene coating of
different thickness was applied onto the electrodes. Coating
thickness was 0 micrometers (no coating), 0.25 micrometers, and 0.5
micrometers. For these measurements, four model solutions were
prepared, each having a volume of 200 mL for testing of the sensor
response. Solution 1 had 200 mL of water with conductivity of 2
mS/cm. Solution 2 had 200 mL of water with conductivity of 15
mS/cm. Solution 3 had 180 mL of water with conductivity of 2 mS/cm
and 20 mL of ethanol. Solution 4 had 180 mL of water with
conductivity of 15 mS/cm and 20 mL of ethanol. Measurements of
conventional impedance and resonance impedance were performed using
these sensors. Spectra were analyzed using PCA.
[0100] FIG. 16A-F summarize results for the interdigital electrode
structures with the electrode width and the spacing between
electrodes of 0.45 mm as measured with conventional impedance (FIG.
16 A, B, C) and with resonant impedance (FIG. 16 D, E, F). These
interdigital electrode structures had three parylene coating
thicknesses of 0, 0.25, and 0.5 micrometers. FIG. 16 A, D depicts
results for parylene coating thickness of 0 micrometers (no
coating). FIG. 16 B, E depicts results for parylene coating
thickness of 0.25 micrometers. FIG. 16 C, F depicts results for
parylene coating thickness of 0.5 micrometers.
[0101] Similarly, FIG. 17A-F summarize results for the interdigital
electrode structures with the electrode width and the spacing
between electrodes of 0.3 mm as measured with conventional
impedance (FIG. 17 A, B, C) and with resonant impedance (FIG. 17 D,
E, F). These interdigital electrode structures had three parylene
coating thicknesses of 0, 0.25, and 0.5 micrometers. FIG. 17 A, D
depicts results for parylene coating thickness of 0 micrometers (no
coating). FIG. 17 B, E depicts results for parylene coating
thickness of 0.25 micrometers. FIG. 17 C, F depicts results for
parylene coating thickness of 0.5 micrometers.
[0102] FIGS. 16A-F and 17A-F illustrate that applying a protective
coating improves the separation between the spectral response
clusters related to the different dielectric constants and
conductivities of measured solutions.
[0103] Technical effects of the disclosed embodiments include an
in-line and real-time technique for assessing cell viability in a
cell culture. Such techniques facilitate distinguishing between
viable and nonviable cells, which provides a benefit over
biomass-type sensors, which do not distinguish between viable and
nonviable cells.
[0104] This written description uses examples, including the best
mode, and also to enable any person skilled in the art to practice
the embodiments, including making and using any devices or systems
and performing any incorporated methods. The patentable scope is
defined by the claims, and may include other examples that occur to
those skilled in the art. Such other examples are intended to be
within the scope of the claims if they have structural elements
that do not differ from the literal language of the claims, or if
they include equivalent structural elements with insubstantial
differences from the literal languages of the claims.
* * * * *