1
Measurement of Cryoprotectant Permeability in Adherent Endothelial Cells and
Applications to Cryopreservation
Running Title: Membrane Permeability of Adherent Endothelial Cells
Allyson K. Fry and Adam Z. Higgins*
School of Chemical, Biological and Environmental Engineering, Oregon State University,
Corvallis, Oregon 97331-2702, USA
*Address correspondence to:
Adam Higgins
School of Chemical, Biological and Environmental Engineering
Oregon State University
102 Gleeson Hall
Corvallis, OR 97331-2702, USA
Email: adam.higgins@oregonstate.edu
Phone: +1-541-737-6245
Fax:
+1-541-737-4600
2
Abstract – Vitrification is a promising approach for cryopreservation of adherent cells because it
allows complete avoidance of ice formation. However, high cryoprotectant (CPA)
concentrations are required to prevent freezing, and exposure to high CPA concentrations
increases the risk of osmotic and toxic damage. Although cell membrane transport modeling
can be used for rational design of CPA equilibration procedures, the necessary permeability
data is extremely scarce for adherent cells. This study validates a method for in situ
measurement of water and CPA permeability in adherent cells based on the fluorescence
quenching of intracellular calcein. Permeability parameters for endothelial monolayers were
measured during exposure to four common cryoprotectants (dimethyl sulfoxide, ethylene glycol,
propylene glycol and glycerol) at temperatures of 4°C, 21°C and 37°C. Propylene glycol
exhibited the highest permeability and glycerol the lowest. The data was fit using an Arrhenius
model, yielding activation energies ranging from 45 kJ/mol to 61 kJ/mol for water transport and
84 kJ/mol to 99 kJ/mol for CPA transport. These permeability parameters will facilitate the
development of mathematically-optimized CPA equilibration procedures for vitrification of
adherent endothelial cells. Our results establish calcein fluorescence quenching as an effective
method for measurement of CPA permeability in adherent cells.
Keywords: adherent cells, vitrification, calcein, fluorescence quenching, membrane
permeability
3
INTRODUCTION
The ability to store biological material without a loss of viability is required for mass
production of cell- and tissue-based products and for banking of tissues and organs.24
Cryopreservation increases the potential shelf-life of such materials from days to decades
because metabolic processes virtually cease at cryogenic temperatures.34 Several types of
suspended cells are routinely and successfully cryopreserved. However, some cell types can
be difficult to cryopreserve in suspension (e.g., stem cells4) and in some cases it is desirable to
preserve characteristics of the adherent cultured cells (e.g., neuronal networks32). In these
instances, it would be advantageous to cryopreserve cells in the adherent state. The ability to
cryopreserve adherent cells would also be advantageous for the many cell types that are
cultured in the adherent state because it would enable improved efficiency and flexibility of the
experimental workflow. Adherent cells have been cryopreserved previously with varying
degrees of success,32,37,39,40,42 but this is not yet an established practice.
The conventional cryopreservation approach involves the use of low cryoprotectant
(CPA) concentrations and slow cooling procedures to inhibit intracellular ice formation.
However, this approach does not prevent formation of extracellular ice. Although the
conventional cryopreservation approach is often successful for suspended cells, it has proven
less effective for adherent cells. Vitrification is an alternative cryopreservation approach that
entails the use of high CPA concentrations to completely suppress ice formation. Vitrification
methods are particularly promising for cryopreservation of adherent cells.4,13,33 However,
exposure to high CPA concentrations can lead to cell damage due to osmotically-driven cell
volume changes and toxicity.12,28,30,52 For vitrification methods to be successful, CPA
equilibration procedures that avoid these issues are required.
Mathematical models of cell membrane transport have been used previously to design
CPA addition and removal procedures. The most widely used approach involves the use of cell
membrane transport predictions to design multi-step procedures that avoid excessive cell
4
volume changes.17,38 Recently, this rational design approach has been extended to address
avoidance of toxicity as well.5,6,25 In particular, we recently developed a new mathematical
optimization approach based on minimization of a toxicity cost function. The cost function is an
equation based on kinetic toxicity data that represents the cumulative toxicity incurred during the
CPA addition or removal process. Procedures designed using this approach have the potential
to dramatically reduce toxicity compared with conventional methods.6 Successful
implementation of these rational design approaches requires accurate knowledge of cell
membrane permeability parameters.
Water and CPA permeability data is available for several types of suspended cells.10,35,41
However, permeability data for adherent cells is relatively scarce. Permeability parameters are
typically estimated by fitting a membrane transport model to measurements of transient cell
volume changes. A convenient technique has been reported for in situ measurement of volume
changes in adherent cells based on the fluorescence quenching of intracellular calcein.19,45
With this technique, cells are loaded with calcein acetoxymethyl ester (calcein AM), a molecule
that is nonfluorescent but membrane permeable. Once inside the cell, the calcein AM molecule
is hydrolyzed by intracellular esterases creating calcein, a molecule that is fluorescent and
membrane impermeable. The fluorescence emitted by intracellular calcein is quenched by
interactions with endogenous quencher molecules, resulting in a total fluorescence intensity that
is dependent on the concentration of intracellular quenchers, and hence the cell volume.45 This
fluorescence quenching technique has been used to measure the water permeability in several
types of adherent cells,18,29,36,45 but measurements of CPA permeability in adherent cells are
essentially nonexistent. To our knowledge, there are only two reports where volume changes in
adherent cells were measured after exposure to CPA, and both utilized methods that are more
cumbersome than the fluorescence quenching technique.14,15 In these studies, volume changes
were only measured after exposure to glycerol at room temperature, and only one study used
the data to estimate the glycerol permeability.
5
In the present study, we have used the fluorescence quenching technique to measure
water and CPA permeability parameters for adherent monolayers of endothelial cells.
Permeability parameters were measured for the common cryoprotectants dimethyl sulfoxide
(DMSO), ethylene glycol (EG), glycerol (GLY), and propylene glycol (PG) at 4°C, 21°C, and
37°C. The results of this study make it possible to predict changes in cell volume and
intracellular composition during CPA addition and removal, which will allow development of CPA
addition and removal procedures using the rational design strategies described above.
Furthermore, our results validate the fluorescence quenching technique for in situ measurement
of CPA permeability, which represents an important step towards development of successful
vitrification procedures for adherent cells.
MATERIALS AND METHODS
Cell Culture
Cell culture supplies were purchased from Invitrogen (Carlsbad, CA) unless otherwise
noted. Bovine pulmonary artery endothelial cells (Cambrex, San Diego, CA) were cultured on
tissue culture treated plastic T25 flasks in a 5% CO2 environment at 37°C. The culture medium
was composed of Dulbecco’s Modified Eagle Medium low glucose nutrient mixture
supplemented with 5% v/v fetal bovine serum, 100 U/ml penicillin, and 100 µg/mL streptomycin
in cell culture grade UltraPure water (ThermoFisher, Waltham, MA). Medium was changed
every 48 hours. Flasks were subcultured when they reached approximately 80% confluency
and split at a ratio of 1:5 using 0.05% trypsin-EDTA solution. To prepare samples for
fluorescence quenching experiments, 25 mm diameter glass coverslips were placed into
separate 30 mm petri dishes and sterilized with 70% ethanol. Cells were seeded onto
coverslips at a density of 1x105 cells per coverslip and cultured for 3 days before
experimentation, at which point they had reached a confluency of 80%.
6
Perfusion Chamber
Fluorescence quenching measurements were obtained by exposing cells to test
solutions using the custom acrylic perfusion chamber illustrated in Figure 1. The perfusion
chamber was designed for temperature control and rapid solution exchange. To control
temperature, a heat exchanger built into the base of the perfusion chamber was circulated with
fluid from a temperature-controlled water bath. Test solutions were delivered through a 1 m
length of tubing contained within the heat exchanger shell, resulting in nearly complete thermal
equilibration with the shell fluid. A rectangular perfusion channel was created by inverting a 25
mm diameter coverslip with adherent endothelial cells, onto a gasket of 100 µm thick medical
grade silicone (BioPlexus, Ventura, CA) with a 3 mm x 15 mm (width x length) cutout. The base
of the rectangular channel was formed by the lid of the heat exchanger. Test solutions were
delivered to the perfusion channel through a Y-connector affixed to the lid of the heat
exchanger. To assure a good seal around the perfusion channel, an acrylic sheet with a
viewing window was placed on top of the coverslip and the entire apparatus was tightly bolted
together. A single syringe pump (Harvard Apparatus, Holliston, MA) was used to deliver test
solutions, which were contained in two separate 60 mL syringes. Pinch valves were used to
prevent backflow to the inactive syringe.
Test Solutions
Hyper- and hypotonic solutions containing only nonpermeating solutes (i.e., without
CPA) were prepared by adding sucrose (Mallinkrodt, Hazelwood, MO) or distilled water to
isotonic (300 mOsm/kg) phosphate buffered saline (PBS; Mediatech, Manassas, VA) to create
solution osmolalities of 200, 230, 430, 600, and 1000 mOsm/kg. CPA solutions were made by
adding ethylene glycol (EG), propylene glycol (PG), glycerol (GLY) or dimethyl sulfoxide
(DMSO) to isotonic PBS to obtain final concentrations of 1000, 2000, and 3000 mOsm/kg.
Propylene glycol was obtained from AlfaAesar (Ward Hill, MA) while all other CPAs were
7
purchased from Mallinkrodt. The actual osmolalities were confirmed to be within 1% of the
nominal values using a freezing point depression osmometer (Advanced Micro Osmometer
Model 3300, Advanced Instruments, Norwood, MA). Solutions with osmolalities above 1000
mOsm/kg were diluted 1:2 in distilled water before measuring the osmolality to ensure that
measurements were taken within the linear range of the osmometer. The osmolality of the
original, undiluted CPA solution was estimated by accounting for the change in water content
upon dilution.
Fluorescence Microscopy
The perfusion chamber was assembled onto the stage of a DM2500 upright microscope
(Leica, Bannockburn, IL) for time-lapse microscopy. Epifluorescence images were acquired at a
rate of 1 Hz with an exposure time of 100 msec using a Fast 1394 camera (Qicam, Surrey, BC
Canada) and Lambda SC-10 Smart Shutter (Sutter Instruments, Novato, CA). Images were
collected using a GFP filter in the light path (465/535 nm ex/em). Neutral density filters (50%,
25%, 12.5%, and 6.25%) were used to decrease the intensity of light from the mercury lamp.
Images with 8x8 binning were collected using ImagePro software (Media Cybernetics,
Bethesda, MD), and the average fluorescence intensity for each image was determined using
built-in software functions.
To prepare endothelial cell monolayers for fluorescence imaging, coverslips were treated
with a solution of 1.25 μg/mL calcein-AM (Sigma-Aldrich, St. Louis, MO) in isotonic PBS for 15
minutes at 37°C. Coverslips were then inverted onto the channel of the flow chamber and
exposed to isotonic PBS at a flow rate of 20 mL/hr for ten minutes in order to equilibrate the
cells to the experimental temperature. After the cells were equilibrated to test conditions, the
flow rate was increased to 200 mL/hr, and time-lapse imaging was initiated. Perfusion with
isotonic solution continued for three minutes, at which point the syringe was switched and cells
were perfused with anisotonic solution for at least three additional minutes. The experimental
8
solution was then switched back to isotonic PBS until chemical equilibrium was achieved.
Experiments were performed at 4°C, 21°C, and 37°C.
The raw fluorescence data from time-lapse images was converted to a nondimensional
cell fluorescence in a similar manner to that described previously.21 The background
fluorescence intensity resulting from the perfusion chamber apparatus and inherent cell
fluorescence was measured each day using an endothelial monolayer that had not been
exposed to calcein dye. This background intensity was subtracted from raw fluorescence data
gathered in the same day. Data was also corrected for non-volume-dependent fading of
fluorescence (e.g., photobleaching, dye leakage) by fitting an exponential decay model to
measurements made while cells were in equilibrium with isotonic solution. The nondimensional
cell fluorescence F was determined by normalizing the background-corrected fluorescence
intensity measurements to the best-fit exponential model.
Calibration of Cell Fluorescence Measurements
To determine cell membrane permeability parameters from cell fluorescence
measurements, it is necessary to convert the fluorescence into cell volume. Exposure to CPA
results in movement of both water and CPA across the cell membrane. Therefore we must
define the relationship between cell fluorescence and the intracellular volumes of water and
CPA. When only nonpermeating solutes are present, the equilibrium volume of intracellular
water (Vw) can be determined from the Boyle-van’t Hoff relationship:
(1)
Vw 
Vw M 0

Vw0 M e
where Vw0 is the intracellular water volume at isotonic conditions, M0 is the isotonic osmolality,
and Me is the osmolality of the extracellular solution. To determine the relationship between the
water volume and the fluorescence, cells were allowed to equilibrate with anisotonic solutions
9
that contained nonpermeating solutes only, and the resulting nondimensional fluorescence was
measured.
To determine the relationship between the fluorescence and the intracellular volume of
CPA, the nondimensional cell fluorescence was measured after cells had equilibrated with CPA
solutions. For solutions prepared by adding CPA to isotonic PBS, the equilibrium volume of
intracellular water and CPA are described by the following equations:
Vw  1
(2)
(3)
VCPA 
VCPA
  CPA  w M CPA
Vw0
where VCPA is the intracellular CPA volume,  CPA is the molar volume of CPA, ρw is the density
of water (assumed to be 1 kg/L) and MCPA is the extracellular CPA osmolality.
A calibration curve was created by plotting the nondimensional cell fluorescence ̅
against the nondimensional volume V , which was defined as
V  Vw  VCPA .
(4)
An expression based on the Stern-Volmer equation44 was developed to describe the
relationship between these variables:
(5)
 V    1   
F 


 V     1   
where α and β are parameters that describe quenching behavior within the system (see
appendix for details). The best-fit values of α and β were determined using a least-squares
Levenberg-Marquardt nonlinear fitting algorithm.
Estimation of Membrane Permeability Parameters
10
The transport of water and CPA across the cellular membrane can be described using a
two-parameter model.26 Typical use of this model to determine membrane permeability
parameters requires knowledge of the cell surface area, isotonic cell volume, and osmotically
inactive cell volume. However, these parameters are difficult to determine for cells in the
adherent state. To adapt this model for use with adherent cells, we expressed the model in
terms of the nondimensional volumes V w and VCPA as follows:
dVw L p ART

dt
Vw 0
(6)
  CPA  w M 0  VCPA


  w M e 
 CPAVw


dVCPA PCPA A 
V 
 CPA  w M CPA  CPA 

dt
Vw 0 
Vw 
(7)
where A is the cellular surface area, R is the ideal gas constant, T is the absolute temperature,
Lp is the membrane water permeability and PCPA is the CPA permeability. Also, we defined
effective water and CPA permeability parameters as follows:
(8)
Lp 
(9)
PCPA 
Lp A
Vw0
PCPA A
Vw0
The nondimensional volume V during exposure to CPA was determined from the sum of Eq. 6
and 7, which results in
(10)
  M  V


V 
dV
 L p RT  CPA w 0 CPA   w M e   PCPA  CPA  w M CPA  CPA 
dt
 CPAVw
Vw 



.
For determination of the best-fit values of L p and P CPA , a subset of the experimental data was
used, comprising fluorescence measurements preceding exposure to the CPA solution (while
the cells were in equilibrium with isotonic perfusate), as well as fluorescence measurements
corresponding with the shrink-swell response after exposure to the CPA solution. The
11
exchange of isotonic PBS for CPA solution was assumed to result in a step change in solution
composition. Because the time, t0 at which this step change occurred was not known, it was
used as a variable parameter in the fitting algorithm, along with the effective permeability
parameters L p and P CPA . The nondimensional fluorescence was predicted by first numerically
integrating Eq. 10 and then substituting the result into Eq. 5. The sum of the squared residuals
between the measured and predicted values of the nondimensional fluorescence was then
computed. This error was minimized by systematically iterating the values of the three
adjustable parameters, yielding best-fit values of t0, L p , and P CPA .
The effective water permeability was also estimated from fluorescence measurements
after exposure to 1000 mOsm/kg solutions containing sucrose but lacking CPA. In these
experiments, water is the only substance that crosses the cell membrane, and Eq. 10 can be
simplified by recognizing that MCPA and V CPA are both equal to zero. The best-fit value of L p
was determined by fitting this model to the data as described previously.21
The temperature dependence of the permeability parameters can be described using an
Arrhenius relationship. For water transport, the Arrhenius equation is
(11)
  EL p
L p  L p ,ref exp 
 R
 1 1 
 
 
 T Tref  
where E L p is the activation energy for water transport and L p ,ref is the effective water
permeability at a reference temperature Tref. For CPA transport, the Arrhenius equation is
(12)
  EP
CPA
PCPA  PCPA,ref exp 
 R
1 1
 
 T Tref

 
 
where E PCPA is the activation energy for CPA transport and P CPA,ref is the effective CPA
permeability at a reference temperature Tref. The values of E L p and E PCPA were determined
12
from linear regressions to Arrhenius plots of the log transformed permeability parameters.
Statistics
Before performing parametric statistical analysis, the underlying assumptions of
normality and homogenous variances were verified. All permeability data was grouped by
temperature and CPA. Normality for each group was confirmed using the Shapiro-Wilk W test
and variances between groups were tested for equality using the Bartlett test. Differences in the
membrane permeabilities at a given temperature were determined using analysis of variance
(ANOVA) and Tukey-Kramer HSD tests. Water and CPA permeabilities were tested separately.
Activation energies for each solute were compared using multivariate techniques with analysis
of covariance (ANCOVA) and Tukey-Kramer HSD tests. Water and CPA activation energies
were tested separately. All analysis was performed using JMP Pro 9 statistical software.
RESULTS
Figure 2 shows representative fluorescence intensity measurements during exposure of
endothelial monolayers to 2000 mOsm/kg PG solution at 21°C. The raw fluorescence intensity
continuously decreased, even when the cells were in equilibrium with isotonic solution. We
accounted for this non-volume-dependent fading by fitting an exponential decay model to the
data, as illustrated in Figure 2. The nondimensional cell fluorescence F was determined by
normalizing the raw fluorescence to the exponential decay fit. The nondimensional cell
fluorescence exhibited transient behavior that was qualitatively similar to the expected cell
volume changes. When cells are at equilibrium with an isotonic solution, exposure to a
hypertonic CPA solution causes the cells to transiently shrink. This is because water will initially
leave the cell, and at a much faster rate than the movement of CPA into the cell. As the
intracellular concentration reaches osmotic equilibrium with the extracellular solution, water and
13
CPA will both enter the cell, causing the volume to return to near that of isotonic equilibrium.
This shrink-swell behavior is evident in the nondimensional fluorescence data directly after the
cells were exposed to the PG solution (see Figure 2). When cells are at equilibrium with a
hypertonic CPA solution, exposure to isotonic solution causes the cells to transiently swell and
then shrink. This swell-shrink behavior is also evident in our nondimensional fluorescence data.
To determine the cell membrane permeability parameters, only a portion of the nondimensional
fluorescence data was used (black squares, Figure 2). This portion corresponds with the period
of isotonic equilibrium preceding exposure to the CPA solution, as well as the shrink-swell
response after exposure to the CPA solution.
To calibrate between fluorescence and cell volume, monolayers were allowed to come to
chemical equilibrium after exposure to anisotonic solutions. The relationship between
fluorescence and cell volume is depicted in Figure 3. We first examined fluorescence intensity
changes after exposure to hypo- and hypertonic solutions containing only nonpermeating
solutes. Exposure to hypotonic solution is known to cause water influx and a concomitant
increase in cell volume; this corresponded with an increase in the cell fluorescence.
Conversely, exposure to hypertonic sucrose solution, which causes a decrease in cell volume,
yielded a decrease in fluorescence. The data was fit using a modified Stern-Volmer relationship
(Eq. 5), yielding best-fit parameters α = 0.81 and β = 0.46 and a correlation coefficient R2 = 0.97.
We also examined the relationship between cell volume and fluorescence after the cells had
equilibrated with CPA solutions. Since solutions were prepared by addition of CPA to isotonic
PBS, the equilibrium water volume is equivalent to the isotonic water volume, and the presence
of intracellular CPA causes the nondimensional volume V to be slightly larger than the
nondimensional volume under isotonic conditions. Changes in cell volume after equilibration in
CPA solutions follow approximately the same trend as volume changes after exposure to
solutions containing nonpermeating solutes (see Figure 3). In fact, if the entire data set is used
14
when determining the Stern-Volmer fit, the values of the parameters α and β remain the same
(within the two significant digits reported here).
The Stern-Volmer calibration trend was used with Eq. 10 to make predictions of
nondimensional cell fluorescence after exposure to CPA solutions, allowing determination of the
best-fit permeability parameters. Typical examples of fluorescence intensity trends after
exposure to CPA solutions at 21°C can be seen in Figure 4. PG permeation was slightly faster
than permeation of DMSO and EG, but intensity profiles of the three data sets were relatively
similar compared to the very slow permeation of GLY. Model fits based on Eq. 10 agreed well
with the data. Figure 5 illustrates the effect of temperature on the rate of CPA transport. As
expected, the rate of EG permeation decreased as the temperature was lowered. Again, the
best-fit model predictions are in good agreement with the experimental data.
Best-fit values of
the effective water and CPA permeability are listed in Table 1 for the four CPAs (DMSO, EG,
GLY, PG) at temperatures of 4°C, 21°C and 37°C. The data was analyzed by ANOVA, revealing
significant differences in the effective permeability parameters at specified temperatures. Most
notably, the GLY permeability was significantly lower than the permeability of EG, PG and
DMSO at all of the temperatures investigated.
The activation energies for water and CPA transport were determined from Arrhenius
plots of the log-transformed permeability parameters, as shown in Figs. 6 and 7. To determine
the permeability values, cells were exposed to CPA solutions with a total osmolality of
1000 mOsm/kg, with the exception of PG at 37°C. At this high temperature the cells reached
equilibrium with the PG solution very quickly and a higher concentration (i.e., 2000 mOsm/kg)
was needed in order to produce detectable changes in the cell volume for permeability
measurement. The trends in the Arrhenius plots are linear, as expected, though the linear
correlations for GLY were relatively weak (R2 =0.87 and R2 = 0.73 for plots of L p and P CPA ,
respectively) compared to the other CPA types (R2 ≥ 0.93). The activation energies E L p and
15
E PCPA were determined from the slope of the linear regressions to the Arrhenius plots. Their
values can be found in Table 2.
DISCUSSION
In the present study, we have demonstrated the in situ measurement of water and CPA
permeability parameters for cells in the adherent state. This capability represents an important
advance, with implications for the rational design of cryopreservation procedures for adherent
cells. Cryopreservation of adherent cells using conventional methods which utilize low CPA
concentrations has been attempted with many cell types including hepatocytes,37,48 embryonic
stem cells,37,39 keratinocytes,40 endothelial cells,10,42 kidney cells,3 and fibroblasts.31 However,
problems including cell detachment, loss of viability, and decreased functionality are often seen
post-thaw. Vitrification methods are particularly promising for cryopreservation of adherent
cells.4,13, 33
However, in order to vitrify, the sample needs to be exposed to high
concentrations of CPA, where 40% or more of the solution is composed of CPA. Such highly
concentrated CPA solutions can cause cell damage because of osmotically-driven cell volume
changes, as well as toxicity. The design of CPA equilibration procedures that avoid both of
these mechanisms of damage is a major obstacle to the development of successful vitrification
procedures. In order to safely equilibrate cells with vitrification solutions, the concentration of
the extracellular solution is typically changed in a gradual or step-wise manner.5,6,17,25 To predict
the optimal CPA concentration and equilibration time for each segment of the procedure, the
values of the cell membrane permeability to water and CPA must be known. Thus, the
permeability parameters determined in this study will enable the design of CPA addition and
removal procedures for adherent endothelial cells using computer simulations.5,6,17,25
Cryopreservation of adherent monolayers is a valuable intermediate step between
cryopreservation of suspended cells and cryopreservation of three-dimensional tissues. While
16
there are obvious differences between cells in a three-dimensional matrix and two-dimensional
monolayer, such as sample thickness and geometric complexity, there are also important
similarities. Most notably, cells in both monolayers and tissues form connections with
neighboring cells and the extracellular matrix. Consequently, adherent cells are often used as
model systems for investigating the response of tissue to cryopreservation. 1,22,23,49 Previous
studies have demonstrated that cell-cell and cell-substrate connections can increase the risk of
intracellular ice formation1,22 and may increase susceptibility to damage caused by extracellular
ice.49 These results are qualitatively consistent with observations of the response of threedimensional tissues to cryopreservation.43,46,53 Because of the similarities between monolayers
and tissues, it is reasonable to expect the permeability parameters for endothelial monolayers
reported in this study to better approximate the permeability of endothelial cells within threedimensional tissue than would the corresponding permeability parameters for suspended
endothelial cells. Nonetheless, we acknowledge that the permeability properties of cells within
intact tissue could differ from those of cells cultured in monolayer. A future direction for this
work will be to adapt the fluorescence quenching technique for use with three-dimensional
tissues by using confocal microscopy to detect changes in fluorescence at specific depths within
a tissue sample. Such a technique would be useful for determining membrane permeability and
diffusion parameters for thin tissues, such as corneal endothelium and precision-cut tissue
slices. Thus, this study represents an important preliminary step towards the rational design of
cryopreservation procedures for tissue as well as adherent cells.
While the CPA permeability properties of cells in the suspended state have been
extensively studied using Coulter counter methods,9,10,42,52 this method cannot be used for cells
in the adherent state. Moreover, there is evidence that surface dissociation and disaggregation
of adherent cells alters their membrane permeability properties.20,54 These studies emphasize
the importance of in situ measurement of membrane permeability in adherent cells and tissues.
Several different in situ methods have been reported for measuring the water permeability,
17
including confocal microscopy,55,56 light scattering,47 interferometry,16 total internal reflection
microfluorimetry,15 light microscopy with spatial filtering,14 and calcein fluorescence
quenching.18,19,29,45 However, to the authors’ knowledge, the CPA permeability of adherent cells
has only been measured using total internal reflection microfluorimetry,15 and only for exposure
to glycerol at room temperature. Drawbacks exist with many of these techniques, such as
costly equipment, technically difficult procedures, and a dependence on cell shape that results
in low quality signals.45,50 Development of a quantitative method that is relatively inexpensive
and straightforward would allow accurate measurement of membrane water and CPA
permeability for the prediction of optimal CPA addition and removal protocols. Solenov et al
(2004) found that calcein fluorescence quenching was preferred over other cell volume
measurement methods for several reasons, including the ability to distinguish signal changes
even with low profile cells of complex shape. In the current study, we have presented a
systematic validation of measurement of the CPA permeability in adherent cells using the
calcein fluorescence quenching method.
Our experimental results strongly support the validity of the fluorescence quenching
method. First, we correlated the cellular fluorescence at steady state with the expected cell
volume after equilibration in solutions containing nonpermeating solutes. In these experiments,
we used the Boyle-van’t Hoff relationship (Eq. 1) to predict the steady-state cell volume in
various hypo- and hypertonic solutions. The accuracy of the Boyle van’t Hoff relationship has
previously been validated experimentally for a large number of cell types, including suspended
BPAEC.21 We found that hypotonic exposure (which causes cell swelling) results in an increase
in fluorescence and hypertonic exposure (which causes shrinkage) results in a decrease in
fluorescence. This result is consistent with the expected fluorescence quenching response,
and, as described below, the Stern-Volmer fit to the data yielded coefficients that were
physically reasonably. Second, we used this Stern-Volmer fit to compare the expected steadystate cell volume to the steady-state cell fluorescence after exposure to CPA solutions. The
18
steady-state data after exposure to CPA solutions was consistent with the best-fit Stern-Volmer
model (Figure 3). Together these steady-state results demonstrate that the Stern-Volmer model
can be used to accurately relate the measured cell fluorescence to the cell volume. Third, we
compared the changes in cell fluorescence after exposure to CPA solutions to the cell volume
changes predicted using the two-parameter membrane transport model. Addition of CPA is
known to cause a transient decrease in the cell volume, whereas removal of CPA is known to
result in a transient increase in the cell volume. In our study, we detected changes in
fluorescence that were consistent with cell volume changes induced by addition and removal of
CPA solutions (Figure 2). Lastly, the membrane permeability trends in our study are consistent
with the published literature. For instance, we observed lower membrane permeability values
for glycerol than for the other CPAs (DMSO, EG, and PG). In addition, we observed
permeability values that decreased as the temperature decreased, linear trends in Arrhenius
plots, and higher activation energy values for CPA transport than water transport. All of these
observations are consistent with published studies of cell membrane permeability. 10,17,21,26,51,52
The modified Stern-Volmer expression that we used to relate fluorescence intensity to
cell volume is relatively complicated compared to the linear models used in most previous
studies that used fluorescent dyes to detect volume changes.2,8,15,18,19,21,36,45, 56 However,
several of these studies reported data that appeared to deviate from linearity, and the studies
that did report data that was well-described using a linear model investigated cell volume
changes over a relatively narrow range. In the current study, the calibration curve exhibited
downward concavity and a linear fit to the data was not very accurate (R2 = 0.90). A similar
downward concave trend was also observed previously.7 The authors of this study used a
polynomial expression to account for the nonlinearity in their data. However, we sought to
describe the nonlinear trend with an expression that could be supported with scientific
reasoning.
19
The Stern-Volmer expression describes the reduction in fluorescence intensity caused
by interactions between a fluorophore and a quencher molecule. By normalizing to isotonic
conditions we were able to create an expression with two parameters, α and β, which describes
the quenching behavior within the system (see appendix for details). Physically, the parameter
α describes the extent of quenching that occurs under isotonic conditions. It is comprised of the
product of the isotonic quencher concentration Q0 and the quenching constant K. Previous
studies have shown that the fluorescence of intracellular calcein is quenched by molecules
present in the cytoplasm, but the exact identity of the intracellular quenchers is unknown.45
Albumin quenches calcein fluorescence with a quenching constant of about K = 0.02 mL/mg.45
Using this quenching constant with our best-fit α-value yields an intracellular quencher
concentration of 41 mg/mL. This quencher concentration is entirely reasonable, given that the
total macromolecule concentration in the cytoplasm is about 250 mg/mL.11 The parameter β
describes the fraction of fluorophores that are not influenced by the intracellular quencher
concentration or the cell volume. We obtained a best-fit value β = 0.46, indicating that 46% of
the intracellular calcein molecules do not interact with cytoplasmic quencher molecules. Under
isotonic conditions, this corresponds with 61% of the total cellular fluorescence originating from
non-interacting calcein molecules. This result is consistent with a previous study of calcein
fluorescence in neuroblastoma cells, which identified two separate pools of intracellular calcein
molecules.8 After membrane permeabilization with alpha toxin, the neuroblastoma cells
retained 67% of their original fluorescence intensity, indicating that a large portion of intracellular
calcein molecules were bound, compartmentalized, or otherwise trapped within the cell.
Only one previous study investigated the permeability of endothelial cells in the adherent
state,20 allowing direct comparison with the current study. In this previous study, the water
permeability of BPAEC monolayers was measured using the fluorescence quenching technique;
the CPA permeability was not measured. The water permeability values from the previous
20
study were nearly two-fold higher than those in the present study.20 Nonetheless, the activation
energies for water transport were almost identical. The most likely explanation for the
discrepancy in the water permeability values is the differences in culture conditions between the
two studies. In the previous study,20 cells were cultured in a medium containing several growth
factor supplements, whereas the medium in the present study did not contain any supplements.
A possible consequence is that cell suspensions derived from BPAEC cultured using
supplement-free medium exhibited a larger average cell volume than the BPAEC from the
previous study (unpublished observations). This larger cell volume would be expected to
correspond with a lower value of the effective water permeability parameter L p .
There have been several previous studies of membrane permeability properties for
endothelial cells in the suspended state. 10,20,51,52 In particular, one previous study investigated
the membrane water permeability of suspended BPAEC, yielding a room temperature water
permeability that was over 30% lower than that in the present study, and an activation energy
that was over 50% higher than that in the present study.20 These differences in permeability
parameters are consistent with previous observations that adherent cells exhibit permeability
properties that differ from the corresponding cell suspensions.20,51 To the authors’ knowledge,
there are no published reports of CPA permeability parameters for suspended BPAEC.
However, CPA permeability values for human corneal endothelial cells have been measured
during exposure to DMSO and PG at three different temperatures,10 for porcine aortic
endothelial cells during exposure to DMSO at two different temperatures,51 and for an
immortalized vascular endothelial cell line (ECV304) during exposure to PG, DMSO, and EG at
two different temperatures.52 The room temperature permeability parameters in these studies
differed from those in the present study by as much as three-fold, and the activation energy
values differed by as much as 40%. Such differences in permeability parameters would be
expected to yield substantial differences in the predicted response to the cryopreservation
21
process, underscoring the importance of in situ measurement of permeability properties for
adherent cells and tissues.
CONCLUSION
We have validated a calcein fluorescence quenching method for in situ measurement of
water and CPA permeability parameters in adherent cells using experiments with endothelial
monolayers. Permeability parameters were determined after exposure to four common CPAs at
three different temperatures, allowing determination of the activation energies for water and
CPA transport. The permeability parameters determined in this study will allow mathematical
optimization of CPA addition and removal procedures for vitrification of endothelial monolayers.
In addition, the fluorescence quenching method is expected to be applicable to other cell types
and adaptable to three-dimensional tissues. The ability to measure the permeability parameters
for adherent cells and tissues is an important step toward the rational design of vitrification
strategies for these complex cellular systems.
APPENDIX
We described the relationship between fluorescence intensity and cell volume using a
modified Stern-Volmer model.
The Stern-Volmer relationship describes the change in
fluorescence intensity due to physical interaction between a fluorophore and quenching
molecule as follows:
F
F*
1  KQ
(A1)
where F is the fluorescence intensity of the fluorophore in the presence of quenching
*
interactions, F is the fluorescence intensity in the absence of quenching, K is the quenching
constant, and Q is the quencher concentration. Previous studies suggest that a portion of
intracellular calcein molecules are not free to interact with quencher molecules because of
22
binding or compartmentalization within the cell.8,19
Modifications to the Stern-Volmer
relationship have been described previously that account for accessible and inaccessible
fluorophores, where it is assumed that any change in fluorescence intensity is due to the
physical interaction of quenchers with accessible fluorophores only and the fluorescence
intensity of inaccessible fluorophores remains constant.27,44
Similarly, we can divide the
intracellular calcein molecules into two groups, one of which is free to interact with intracellular
quencher molecules, and one that does not interact with the quenchers. The calcein molecules
that are free to interact with quenchers yield a fluorescence intensity described by Eq. A1,
whereas non-interacting fluorophores give rise to a constant intensity, resulting in the following
equation for the total fluorescence:
F
F*
 FB
1  KQ
(A2)
where FB is the fluorescence intensity due to non-interacting calcein molecules (i.e., the portion
of the fluorescence that is not sensitive to quencher concentration).
The quencher
concentration can be expressed in terms of the cell volume by assuming that the endogenous
quencher molecules (e.g., intracellular proteins 45) are membrane impermeable and trapped
within the cell.
This assumption results in the following expression for the quencher
concentration:
Q
Q0Vw0
Vw  VCPA
(A3)
where Q0 is the quencher concentration under isotonic conditions. Substituting Eq. A3 into Eq.
A2 and normalizing by the fluorescence intensity under isotonic conditions (F0) results in:
F
F  V    1   



F0  V     1   
(5)
23
where   KQ0 and  
FB
. The parameter  describes the extent of quenching that
F  FB
*
occurs under isotonic conditions, and β is the fraction of total fluorophores that is insensitive to
quencher concentration and therefore insensitive to cell volume.
ACKNOWLEDGMENTS
This work was supported by funding from the Medical Research Foundation of Oregon
(MRF grant #1015).
Allyson Fry received support from the Shirley Kuse Fellowship and
Diversity Advancement Pipeline Fellowship. The authors would also like to acknowledge Austin
Rondema and Nadeem Houran for their assistance with fluorescence quenching experiments.
24
REFERENCES
1.Acker, J. P., A. Larese, H. Yang, A. Petrenko, and L. E. McGann. Intracellular ice formation is
affected by cell interactions. Cryobiology 38:363-371, 1999.
2.Altamirano, J., M. S. Brodwick, and F. J. Alvarez-Leefmans. Regulatory volume decrease and
intracellular Ca2+ in murine neuroblastoma cells studied with fluorescent probes. J. Gen.
Physiol. 112:145-160, 1998.
3.Armitage, W. J., and B. K. Juss. Freezing monolayers of cells without gap junctions.
Cryobiology 46:194-196, 2003.
4.Beier, A. F. J., J. C. Schulz, D. Dörr, A. Katsen-Globa, A. Sachinidis, J. Hescheler, and H.
Zimmermann. Effective surface-based cryopreservation of human embryonic stem cells by
vitrification. Cryobiology 63:175-185, 2011.
5.Benson, J. D., C. C. Chicone, and J. K. Critser. A general model for the dynamics of cell
volume, global stability, and optimal control. J. Math. Biol. 63:339-359, 2010.
6.Benson, J. D., A. J. Kearsley, and A. Z. Higgins. Mathematical optimization of procedures for
cryoprotectant equilibration using a toxicity cost function. Cryobiology, In press ,
2012.doi:10.1016/j.cryobiol.2012.01.001
7.Chen, P. Y., D. Pearce, and A. S. Verkman. Membrane water and solute permeability
determined quantitatively by self-quenching of an entrapped fluorophore. Biochemistry
27:5713-5718, 1988.
8.Crowe, W. E., J. Altamirano, L. Huerto, and F. J. Alvarez-Leefmans. Volume changes in single
N1E-115 neuroblastoma cells measured with a fluorescent probe. Neuroscience 69:283-296,
1995.
9.Ebertz, S. L., and L. E. McGann. Osmotic parameters of cells from a bioengineered human
corneal equivalent and consequences for cryopreservation. Cryobiology 45:109-117, 2002.
10.Ebertz, S. L., and L. E. McGann. Cryoprotectant permeability parameters for cells used in a
bioengineered human corneal equivalent and applications for cryopreservation. Cryobiology
49:169-180, 2004.
11.Ellis, R. J. Macromolecular crowding: obvious but underappreciated. Trends Biochem. Sci.
26:597-604, 2001.
12.Fahy, G., B. Wowk, J. Wu, and S. Paynter. Improved vitrification solutions based on the
predictability of vitrification solution toxicity. Cryobiology 48:22-35, 2004.
13.Fan, W.-X., X.-H. Ma, T.-Q. Liu, and Z.-F. Cui. Vitrification of corneal endothelial cells in a
monolayer. J. Biosci. Bioeng. 106:610-613, 2008.
25
14.Farinas, J., M. Kneen, M. Moore, and A. S. Verkman. Plasma membrane water permeability
of cultured cells and epithelia measured by light microscopy with spatial filtering. J. Gen.
Physiol. 110:283-296, 1997.
15.Farinas, J., V. Simanek, and A. S. Verkman. Cell volume measured by total internal
reflection microfluorimetry: application to water and solute transport in cells transfected with
water channel homologs. Biophys. J. 68:1613-1620, 1995.
16.Farinas, J., and A. S. Verkman. Cell volume and plasma membrane osmotic water
permeability in epithelial cell layers measured by interferometry. Biophys. J. 71:3511-3522,
1996.
17.Gao, D. Y., J. Liu, C. Liu, L. E. McGann, P. F. Watson, F. W. Kleinhans, P. Mazur, E. S.
Critser, and J. K. Critser. Prevention of osmotic injury to human spermatozoa during addition
and removal of glycerol. Hum. Reprod. 10:1109-1122, 1995.
18.Hamann, S., J. F. Kiilgaard, M. la Cour, J. U. Prause, and T. Zeuthen. Cotransport of H+,
lactate, and H2O in porcine retinal pigment epithelial cells. Exp. Eye Res. 76:493-504, 2003.
19.Hamann, S., J. F. Kiilgaard, T. Litman, F. J. Alvarez-leefmans, B. R. Winther, and T.
Zeuthen. Measurement of cell volume changes by fluorescence self-quenching. J. Fluoresc.
12:139-145, 2002.
20.Higgins, A. Z., and J. O. M. Karlsson. Comparison of cell membrane water permeability in
monolayers and suspensions. Cryo letters, 33:95-106, 2012.
21.Higgins, A. Z., and J. O. M. Karlsson. Analysis of Solution Exchange in Flow Chambers with
Applications to Cell Membrane Permeability Measurement. Cell. Mol. Bioeng. 3:269-285,
2010.
22.Irimia, D., and J. O. M. Karlsson. Kinetics and Mechanism of Intercellular Ice Propagation in
a Micropatterned Tissue Construct. Biophys. J. 82:1858-1868, 2002.
23.Irimia, D., and J. O. M. Karlsson. Kinetics of intracellular ice formation in one-dimensional
arrays of interacting biological cells. Biophys. J. 88:647-60, 2005.
24.Karlsson, J. O. M., and M. Toner. Long-term storage of tissues by cryopreservation: critical
issues. Biomaterials 17:243-256, 1996.
25.Karlsson, J. O. M., A. I. Younis, A. W. S. Chan, K. G. Gould, and A. Eroglu. Permeability of
the rhesus monkey oocyte membrane to water and common cryoprotectants. Mol. Reprod.
Dev. 76:321-333, 2009.
26.Kleinhans, F. W. Membrane permeability modeling: Kedem–Katchalsky vs a two-parameter
formalism. Cryobiology 37:271-289, 1998.
27.Laws, W. R., and P. B. Contino. Fluorescence quenching studies: analysis of nonlinear
Stern-Volmer data. Meth. Enzymol. 210:448-463, 1992.
26
28.Lawson, A., H. Ahmad, and A. Sambanis. Cytotoxicity effects of cryoprotectants as singlecomponent and cocktail vitrification solutions. Cryobiology 62:115-122, 2011.
29.Levin, M. H., and A. S. Verkman. Aquaporin-dependent water permeation at the mouse
ocular surface: in vivo microfluorimetric measurements in cornea and conjunctiva. Invest.
Ophthalmol. Vis. Sci. 45:4423-32, 2004.
30.Levin, R. L. A generalized method for the minimization of cellular osmotic stresses and
strains during the introduction and removal of permeable cryoprotectants. J. Biomech. Eng.
104:81-87, 1982.
31.Liu, K., Y. Yang, and J. Mansbridge. Comparison of the stress response to cryopreservation
in monolayer and three-dimensional human fibroblast cultures: Stress proteins, MAP
kinases, and growth factor gene expression. Tissue Engineering 6:539-554, 2000.
32.Ma, W., T. O’Shaughnessy, and E. Chang. Cryopreservation of adherent neuronal networks.
Neurosci. Lett. 403:84-9, 2006.
33.Magalhaes, R., A. K. Pr, F. Wen, X. Zhao, H. Yu, and L. L. Kuleshova. The use of vitrification
to preserve primary rat hepatocyte monolayer on collagen-coated poly(ethyleneterephthalate) surfaces for a hybrid liver support system. Biomaterials 30:4136-4142, 2009.
34.Mazur, P. Freezing of living cells: mechanisms and implications. Am. J. Physiol. 247:125142, 1984.
35.McGrath, J. J. Quantitative measurement of cell membrane transport: technology and
applications. Cryobiology 34:315-334, 1997.
36.Mitchell, C. H., J. C. Fleischhauer, W. D. Stamer, K. Peterson-Yantorno, and M. M. Civan.
Human trabecular meshwork cell volume regulation. Am. J. Physiol., Cell Physiol. 283:315326, 2002.
37.Miyamoto, Y., S. Enosawa, T. Takeuchi, and T. Takezawa. Cryopreservation in situ of cell
monolayers on collagen vitrigel membrane culture substrata: ready-to-use preparation of
primary hepatocytes and ES cells. Cell. Transplant. 18:619-626, 2009.
38.Mukherjee, I. N., Y. C. Song, and A. Sambanis. Cryoprotectant delivery and removal from
murine insulinomas at vitrification-relevant concentrations. Cryobiology 55:10-18, 2007.
39.Nie, Y., V. Bergendahl, D. J. Hei, J. M. Jones, and S. P. Palecek. Scalable Culture and
Cryopreservation of Human Embryonic Stem Cells on Microcarriers. Biotechnol. Prog. 25:2031, 2009.
40.Pasch, J., A. Schiefer, I. Heschel, and G. Rau. Cryopreservation of keratinocytes in a
monolayer. Cryobiology 39:158-168, 1999.
41.Paynter, S. J., K. J. Andrews, N. N. Vinh, C. M. Kelly, A. E. Rosser, N. N. Amso, and S. B.
Dunnett. Membrane permeability coefficients of murine primary neural brain cells in the
presence of cryoprotectant. Cryobiology 58:308-314, 2009.
27
42.Pegg, D. E. Cryopreservation of vascular endothelial cells as isolated cells and as
monolayers. Cryobiology 44:46-53, 2002.
43.Pichugin, Y., G. M. Fahy, and R. Morin. Cryopreservation of rat hippocampal slices by
vitrification. Cryobiology 52:228-240, 2006.
44.Shetlar, M. D. A generalized form of the Stern-Volmer equation and its application. Mol.
Photochem. 6:191-205, 1974.
45.Solenov, E., H. Watanabe, G. T. Manley, and A. S. Verkman. Sevenfold-reduced osmotic
water permeability in primary astrocyte cultures from AQP-4-deficient mice, measured by a
fluorescence quenching method. Am. J. Physiol., Cell Physiol. 286:426-432, 2004.
46.Song, Y. C., B. S. Khirabadi, F. Lightfoot, K. G. M. Brockbank, and M. J. Taylor. Vitreous
cryopreservation maintains the function of vascular grafts. Nat. Biotechnol. 18:296-299,
2000.
47.Srinivas, S. P., J. A. Bonanno, E. Larivière, D. Jans, and W. Van Driessche. Measurement of
rapid changes in cell volume by forward light scattering. Pflugers Arch. 447:97-108, 2003.
48.Stevenson, D. J., C. Morgan, E. Goldie, G. Connel, and M. H. Grant. Cryopreservation of
viable hepatocyte monolayers in cryoprotectant media with high serum content: metabolism
of testosterone and kaempherol post-cryopreservation. Cryobiology 49:97-113, 2004.
49.Stott, S. L., and J. O. M. Karlsson. Visualization of intracellular ice formation using highspeed video cryomicroscopy. Cryobiology 58:84-95, 2009.
50.Verkman, A. S. Water Permeability Measurement in Living Cells and Complex Tissues. J.
Membr. Biol. 173:73-87, 2000.
51.Wusteman, M. C., and D. E. Pegg. Differences in the requirements for cryopreservation of
porcine aortic smooth muscle and endothelial cells. Tissue Eng. 7:507-518, 2001.
52.Wusteman, M. C., D. E. Pegga, M. P. Robinsonb, L.-H. Wanga, and P. Fitcha. Vitrification
media: toxicity, permeability, and dielectric properties. Cryobiology 44:24-37, 2002.
53.Xu, X., S. Cowley, C. J. Flaim, W. James, L. Seymour, and Z. Cui. The roles of apoptotic
pathways in low recovery rate after cryopreservation of dissociated human embryonic stem
cells. Biotechnol. Prog. 26:827-837, 2009.
54.Yarmush, M. L., M. Toner, J. C. Y. Dunn, A. Rotem, A. Hubel, and R. G. Tompkins. Hepatic
Tissue Engineering - Development of Critical Technologies. Ann. N. Y. Acad. Sci. 665:238252, 1992.
55.Yoshimori, T., and H. Takamatsu. 3-D measurement of osmotic dehydration of isolated and
adhered PC-3 cells. Cryobiology 58:52-61, 2009.
56.Zelenina, M., and H. Brismar. Osmotic water permeability measurements using confocal
laser scanning microscopy. Eur. Biophys. J. 29:165-171, 2000.
28
Table 1. Comparisons of effective permeability parameters.
Temp
4°C
21°C
37°C
ABC
Lp x 108
PCPA x 103
(Pa-1s-1)
(s-1)
DMSO
0.65 ± 0.08A
9.4 ± 0.90AB
EG
0.92 ± 0.05A
4.0 ± 0.20B
GLY
0.63 ± 0.05A
0.36 ± 0.2C
PG
0.96 ± 0.12A
13 ± 2.5A
Sucrose
0.79 ± 0.04A
DMSO
3.6 ± 0.37BC
61.0 ± 8.8B
EG
3.8 ± 0.15BC
46 ± 6.6B
GLY
4.4 ± 0.65AB
4.9 ± 2.8C
PG
5.4 ± 0.26A
160 ± 15A
Sucrose
2.60 ± 0.29C
DMSO
9.4 ± 0.74AB
460 ± 67B
EG
13 ± 1.4AB
350 ± 31B
GLY
7.3 ± 0.98B
14 ± 1.9C
PG
17 ± 2.8A
1300 ± 190A
Sucrose
6.5 ± 1.0B
Solute
Statistical differences in permeability values at the given temperatures are denoted by
distinct superscript letters. GLY permeability ( PCPA ) is significantly lower than DMSO, EG, and
PG permeability at all temperatures. There is no clear trend in statistical significance for water
permeability ( L p ).
29
Table 2. Comparisons of activation energies.
E Lp
E PCPA
(kJ/mol)
(kJ/mol)
DMSO
58 ± 3.3AB
84 ± 4.7A
EG
57 ± 2.1AB
98 ± 5.3A
GLY
51 ± 5.0AB
89 ± 13.5A
PG
61 ± 3.6A
99 ± 3.8A
Sucrose
45 ± 3.1B
Solute
AB
Statistical differences in activation energy values are denoted by distinct superscript letters.
E L p values for PG and sucrose are significantly different. There are no significant differences
between values for E PCPA .
30
Figure 1. (A) Schematic of temperature-controlled perfusion chamber. (B) Microscopy image
of calcein stained cell monolayer.
31
Figure 2. Representative trend from time-lapse video microscopy. Cells were first exposed to
isotonic solution, then 2000 mOsm/kg PG, and finally returned to isotonic solution at 21°C. The
average fluorescence was collected from each image to create the raw fluorescence trend (dark
grey squares). An exponential decay model was fit to the isotonic regimes (line) and used to
correct for non-volume-dependent fading, resulting in a nondimensional fluorescence F (light
grey squares). The initial isotonic regime and the transient regime directly after PG exposure
(black squares) were used to determine membrane permeability parameters.
32
Figure 3. Calibration of fluorescence intensity to cell volume. The equilibrium volume was
determined using Eqs. 1-3 after exposure to solutions containing nonpermeating solutes only
(inverted triangles), or solutions containing DMSO (circle), EG (squares), or PG (triangles). The
data was fit using a modified Stern-Volmer relationship (line). Each point represents the
average ± SEM of at least four samples with exposure performed at 21°C.
33
Figure 4. Representative transient fluorescence data (symbols) and best-fit model predictions
(lines) for exposure to 1000 mOsm/kg CPA solutions at 21°C.
34
Figure 5. Representative transient fluorescence data (symbols) and best-fit model predictions
(lines) for exposure to 1000 mOsm/kg EG solution at temperatures of 4°C, 21°C, and 37°C.
35
Figure 6. Arrhenius plot of the effective water permeability for exposure to solutions containing
sucrose or CPA. The activation energies for water transport were determined from the slopes of
the best-fit lines. All solutions had a total osmolality of 1000 mOsm/kg except for the PG
solution at 37°C which was 2000 mOsm/kg. Data represents the average ± SEM of at least four
samples.
36
Figure 7. Arrhenius plot of the effective CPA permeability. The activation energies for CPA
transport were determined from the slopes of the best-fit lines. All solutions had a total
osmolality of 1000 mOsm/kg except for the PG solution at 37°C which was 2000 mOsm/kg.
Data represents the average ± SEM of at least four samples.