DEVELOPMENT AND APPLICATIONS OF A MICROFLUIDIC VASCULAR MIMIC
By
Paul A. Vogel

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Chemistry
2012

ABSTRACT
DEVELOPMENT AND APPLICATIONS OF A MICROFLUIDIC VASCULAR MIMIC
By
Paul A. Vogel
Vascular wall biology and interactions with bloodstream components are areas of continued
interest due to the high prevalence of cardiovascular disease and stroke. Cell types comprising
the vascular wall, as well as those found in the bloodstream, are frequently isolated and
cultured in vitro to determine their biological roles, including their involvement in disease
progression. Recently, microfluidic systems have emerged as desirable platforms for mimicking
biological systems, offering precise control of cellular environments. Here, work is presented
that advances microfluidic technology and improves the utility of microfluidic systems for cellbased assays and cell culture.
The endothelium is an important component of the vascular wall due to its role as a
protective barrier, as well as its involvement in several important vascular functions. In vivo, the
endothelium plays a vital role in the regulation of vascular tone by releasing several factors that
cause vessel dilation, of which nitric oxide (NO) is one of the most potent. The endothelial cells
comprising the endothelium are influenced by many environmental factors, such as shear stress
generated by flowing blood and the degree of confluence of the endothelial monolayer.
Presented here is the development of novel microfluidic system capable of mimicking the
vasculature by incorporating microfluidic channels that enable blood flow past cultured
endothelial monolayers. Importantly, this microfluidic vascular mimic is successfully integrated
with a transendothelial electrical resistance (TEER) measurement system that allows for the

monitoring of endothelial monolayer confluence and barrier integrity. This microfluidic system
is used to highlight the impact of monolayer confluence on cellular behavior by showing that, in
response to flowing red blood cells (RBCs), confluent endothelial monolayers produce
significantly more NO than less confluent monolayers.
C-peptide, a potential drug for type 1 diabetes, is known to improve blood flow by a
previously unknown mechanism. This dissertation hypothesizes that C-peptide is capable of
indirectly inducing vessel dilation by stimulating the release of adenosine triphosphate (ATP)
from flowing RBCs, which can diffuse to the endothelium to bind purinergic receptors on the
endothelial cell surface, ultimately resulting in endothelial NO production. To investigate this
hypothesis, the microfluidic vascular mimic is utilized to monitor cell-cell communication,
revealing C-peptide is capable of stimulating endothelial NO production by a mechanism
mediated by the RBC which requires P2Y purinoreceptor activation by ATP. Importantly, it is
2+

shown that in order to observe this NO production, C-peptide must be prepared with Zn .
2+

This proposed mechanism describes how Zn /C-peptide may improve blood flow indirectly
as a result of activity within the bloodstream. However, since C-peptide has never been
observed exogenous of the bloodstream in vivo, it is unclear whether all C-peptide bioactivity
results similarly, or if C-peptide can also escape the bloodstream to directly stimulate cells in
surrounding tissue. To assess the ability of C-peptide to penetrate the vascular wall, the
microfluidic TEER system is employed to determine the ability of C-peptide to permeate the
endothelium, revealing that C-peptide can permeate confluent endothelial monolayers in vitro,
suggesting that C-peptide may escape the bloodstream in vivo. Collectively, this work shows
that C-peptide may have a more comprehensive biological role than is currently assumed.

ACKNOWLEDGEMENTS
Firstly, I would like to thank the organizations responsible for funding this research,
including Michigan State University for research and travel funding, some of which I have repaid
in the form of late registration and enrollment fees. Additionally, I am grateful for research
funding provided by the National Institutes of Health and Merck. I appreciate the investments
made by each of these organizations in this research, as well as my development as a scientist.
Family members, friends and faculty have helped me complete this dissertation and the
work described within. In fact, I was coaxed into attending graduate school by my dear wife
Erica, who always has great ideas. In the absence of her persuasion I would not have attended
graduate school and, therefore, Erica is primarily responsible for this accomplishment. Luckily,
Erica did not abandon me during the course of this work and instead provided consistent
encouragement and support throughout. Erica’s influence has significantly improved my life,
and she was also fundamental to the completion of this dissertation to the extent that she
should probably be listed as an author, but this acknowledgement will have to suffice.
Aside from Erica, my family members are the only people who I am certain believed that I
was capable of my achievements. Julie, my mother, worked incredibly hard during my
childhood, and without this example I am not sure I would have had the dedication required to
complete this dissertation. My mom has always urged me to do better, academically and in
general, often to no avail. I cannot begin to fathom the stress I have caused my mother over the
past quarter century, so I am thankful that she was well enough to endure more during my
years as a graduate student. Richard and Sue, my grandparents, have always stressed the
importance of an education, both verbally, and by example as career educators. During the

iv

numerous periods of time that I was bored and distracted from school, it was always advice
from my grandparents that would motivate me to be a better scholar. Without their advice and
support, financial and otherwise, I would most likely not be college educated. My grandpa is the
wisest man I know, and I believe that every word he says before 6:00 pm should be written
down and studied. Also, my brother Danny has always been excited about my accomplishments
and he truly cares about my success, which I find particularly motivating. However, his
contributions to this dissertation are questionable when considering the hours we spent playing
video games in lieu of studying.
The faculty of this department has provided me with a tremendous graduate education
which has enabled me to think about complex problems and approach them as a scientist. As a
result of my graduate training, I have gained the skills and confidence necessary for
independent research. I am grateful for the advice and critique from my guidance committee.
Importantly, I would like to thank Dana Spence, my advisor, for accepting me into his group
with my sterling reputation and providing me with exciting scientific opportunities. I truly
appreciate his guidance through the work presented in this dissertation. I believe the training I
have received from him is invaluable, which will undoubtedly benefit me throughout life, as will
the one-liners and limericks that will be forever in my memory.
I would also like to thank the Spence research group for their contributions to this work and
for being fun, which made my graduate experience enjoyable. In particular, I would like to
thank Kari Anderson and Suzanne Letourneau for always reminding me of my obligations and
allowing me to study their superior class notes. I would also like to thank Stephen Halpin for
showing me the professional etiquette of national conferences.

v

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................................... iv
LIST OF FIGURES...................................................................................................................... v
Chapter 1 - Introduction.......................................................................................................... 1
1.1 - Monitoring Cell Monolayer Barrier Integrity ...................................................................... 3
1.1.1 - Optical Microscopy ....................................................................................................... 5
1.1.2 - Permeability Assays ...................................................................................................... 8
1.1.3 - Measuring Transendothelial Electrical Resistance (TEER) ......................................... 11
1.1.4 - Importance of Monitoring Barrier Integrity in Vascular Mimics ................................ 16
1.2 - Mimicking the Vasculature................................................................................................ 17
1.2.1 - The Endothelial Cell .................................................................................................... 23
1.2.2 - The Red Blood Cell...................................................................................................... 28
1.2.3 - C-peptide .................................................................................................................... 31
1.2.4 - Current Hypothesis Regarding Vasodilation and the Influence of C-peptide ............ 39
1.3 - Microfluidics ...................................................................................................................... 42
1.3.1 - Fabrication and Background ...................................................................................... 42
1.3.2 - Microfluidics to Mimic Biological Systems ................................................................. 44
1.3.3 - Monitoring Barrier Integrity on a Microfluidic Device ............................................... 45
1.3.4 - Cell-Cell Communication in Microfluidic Systems...................................................... 46
1.4 - Goals and Motivation ........................................................................................................ 48
REFERENCES .................................................................................................................................. 52
Chapter 2 - Development of a Microfluidic Device Capable of Measuring TEER ..................... 71
2.1 - TEER Measurements on a Microfluidic Vascular Mimic ................................................... 71
2.2 - Methods ............................................................................................................................ 73
2.2.1 - Device Fabrication ...................................................................................................... 73
2.2.2 - Pumping and Attachments ......................................................................................... 78
2.2.3 - bPAEC Culture ............................................................................................................ 78
2.3 - Results ............................................................................................................................... 80
2.3.1 - TEER Measurement Method Development ............................................................... 80
2.3.2 - Validation of TEER Measurements ............................................................................. 84
2.3.3 - Monitoring bPAEC Monolayer Barrier Integrity on a Microfluidic Device ................. 86
2.4 - Discussion .......................................................................................................................... 90
2.5 - Conclusions........................................................................................................................ 99
REFERENCES ................................................................................................................................ 102
Chapter 3 - Monitoring Cell-Cell Communication in a Microfluidic-Based Vascular Mimic ... 106
3.1 - Microfluidics to Mimic the Vasculature .......................................................................... 106
3.2 - The Potential Role of C-Peptide in Vessel Dilation ......................................................... 109
vi

3.3 - Effects of Metal Binding on C-Peptide Bioactivity .......................................................... 110
3.4 - Measuring Intracellular Nitric Oxide ............................................................................... 111
3.5 - Methods .......................................................................................................................... 116
3.5.1 - Device Fabrication .................................................................................................... 116
3.5.2 - bPAEC Culture .......................................................................................................... 116
3.5.3 - TEER to Verify bPAEC Monolayer Confluence in Device .......................................... 117
3.5.4 - Sample Preparation .................................................................................................. 117
3.5.5 - PPADS to Inhibit Purinergic Signaling ....................................................................... 118
3.5.6 - Sample Introduction in Microfluidic Flow Experiments ........................................... 118
3.5.7 - Measuring NO Production of bPAECs Immobilized on Device................................. 119
3.6 - Results ............................................................................................................................. 121
3.6.1 - Meeting bPAEC Culture Requirements .................................................................... 121
3.6.2 - bPAEC Monolayer Confluence Influences NO Production ....................................... 123
3.6.3 - Mechanism Determination: C-peptide Stimulated Endothelial NO Production ...... 125
3.7 – Discussion ....................................................................................................................... 127
3.8 - Conclusions...................................................................................................................... 135
REFERENCES ................................................................................................................................ 138
Chapter 4 - Determining Endothelial Permeability to C-peptide .......................................... 146
4.1 - Biological Activity of C-peptide: Direct and/or Indirect Mechanisms............................. 146
4.2 - Enzyme-Linked Immunosorbent Assay (ELISA) for C-peptide Detection ........................ 148
4.3 - Methods .......................................................................................................................... 150
4.3.1 - Bovine Pulmonary Artery Endothelial Cell Culture .................................................. 150
4.3.2 - TEER Measurements ................................................................................................ 151
4.3.3 - Microfluidic C-peptide Permeability Assay: Sample Preparation ............................ 151
4.3.4 - Microfluidic C-peptide Permeability Assay: Flow Studies ........................................ 152
4.3.5 - Microtiter C-peptide Permeability Assay: Sample Preparation ............................... 153
4.3.6 - Microtiter C-peptide Permeability Assay: Static Studies ......................................... 154
4.4 - Results ............................................................................................................................. 155
4.4.1 - Microfluidic Assay of C-peptide Endothelial Permeability ....................................... 155
4.4.2 - Static Microtiter Assay of C-Peptide Endothelial Permeability ................................ 158
4.5 - Discussion ........................................................................................................................ 161
4.6 - Conclusions...................................................................................................................... 167
REFERENCES ................................................................................................................................ 169
Chapter 5 - Conclusions and Future Directions .................................................................... 172
5.1 - Conclusions...................................................................................................................... 172
5.1.1 - Microfluidic TEER Measurement Device .................................................................. 172
5.1.2 - C-peptide Induced Vasodilation ............................................................................... 174
5.1.3 - Endothelial Permeability to C-peptide ..................................................................... 179
5.2 - Future Directions ............................................................................................................. 181
5.2.1 - Microfluidic TEER System ......................................................................................... 181
5.2.1.1 - Polystyrene (PS) as a Material for Microfluidic Fabrication ............................. 182
vii

5.2.2 - C-peptide Induced Vasodilation ............................................................................... 186
2+
5.2.2.1 - Elucidate Zn /C-peptide Interactions .............................................................. 186
2+
5.2.2.2 - Investigation of Zn /C-peptide Activity In Vivo ............................................... 189
5.2.3 - Endothelial Permeability to C-peptide ..................................................................... 189
5.2.3.1 - Choosing Endothelial Cell Type......................................................................... 189
5.2.3.2 - Kinetics of Permeation and Comparisons......................................................... 190
5.2.3.3 - Permeability of C-peptide Mutants .................................................................. 191
REFERENCES ................................................................................................................................ 194

viii

LIST OF TABLE
Table 4.1 - The permeability of an endothelial monolayer to C-peptide was investigated by
performing microtiter permeability assays, the results of which are reported here. C-peptide
was quantified in the basolateral and apical compartments via ELISA 3.5 hours after adding
varying quantities of C-peptide to apical compartments. Prior to each permeability assay,
proper bPAEC barrier function was confirmed by measuring TEER. (n=4, Errors are SD) ......... 161

ix

LIST OF FIGURES
Figure 1.1 - A microtiter cell culture insert (membrane insert) is shown resting in a microtiter
well, forming a diffusion cell most commonly used for permeability assays. Typically, a
monolayer of cells is cultured upon a collagen or fibronectin coated membrane at the bottom
of the membrane insert. The permeation of analytes or drugs of interest from the apical
solution to the basolateral solution can be monitored to estimate the in vivo permeability of
particular cellular barriers, represented by the cultured cells, to the molecules under
investigation. Also, the confluence and barrier integrity of the cultured cellular monolayer can
be determined by monitoring the permeation of a tracer molecule from the apical
compartment to the basolateral compartment and calculating a permeability coefficient, which
is proportional to layer barrier integrity. ........................................................................................ 9
Figure 1.2 - A traditional TEER measurement utilizes a microtiter membrane insert that fits into
a traditional well plate. The cell layer is cultured on the polycarbonate membrane at the
bottom of the insert and placed into the well, which contains cell media. The TEER
measurement is performed by placing electrodes in the basolateral compartment on the
outside of the insert and in the apical compartment on the inside of the insert containing the
cell layer, and measuring the resistance between the two, which is primarily determined by the
barrier integrity of the cultured cells. ........................................................................................... 15
Figure 1.3 - A cross-sectional view of a resistance vessel. Flowing blood components, such as
the RBC, are contained in the core of the vessel. Surrounding the flowing blood and lining the
interior of all blood vessels is a confluent and continuous monolayer of endothelial cells, which
collectively form the endothelium. The endothelium forms a barrier, separating the flowing
blood from exterior smooth muscle, which is responsible for contracting and relaxing the blood
vessel to regulate vascular tone. In the smooth muscle, the potent vasodilator, NO, is capable of
causing vessel relaxation; however the source of this NO is under investigation. In two
prominent mechanisms, the primary source is either NO directly released from the RBC, or NO
produced in the endothelium in response to ATP released from the RBC. ................................. 19
Figure 1.4 - Mechanism describing the production of nitric oxide (NO) in the endothelium. Free
intracellular calcium binds calmodulin to form the calcium calmodulin complex (Ca-CaM), which
stimulates endothelial nitric oxide synthase (eNOS) by facilitating electron transfer through the
complex and thereby increasing the oxidation rate of NADPH to NADP+. The oxygen dependent
oxidation of NADPH by eNOS provides electrons for the enzymatic conversion of L-arginine to
citruline and NO. Therefore, the increase in electron transfer kinetics resulting from Ca-CaM
binding ultimately increases NO production in the endothelium. For simplicity, other essential
proteins and signaling molecules which are involved in the signaling complex formed during
endothelial NO synthesis are discussed previously, but not pictured. ........................................ 26

x

Figure 1.5 - The amino acid sequence of C-peptide. Subsequent to cleavage from insulin, the 31
residue peptide is capable of increasing ATP release from RBCs, an activity which is proposed to
mediate NO production in the endothelium. ............................................................................... 37
Figure 1.6 - The proposed mechanism of RBC mediated vasorelaxation. In response to a
stimulus, such as C-peptide interaction or shear stress, the RBC releases ATP, which is capable
of diffusing to the endothelium. As an agonist of the G-coupled P2 purinoreceptors, ATP binds
the P2Y purinergic receptors located on the endothelial cell membrane, which initiates a
purinergic signaling cascade. In response to ATP agonist, P2Y increases calcium flux into the cell,
leading to an increase in free intracellular Ca2+, which can bind CaM to form the Ca-CaM
complex. In this active conformation, Ca-CaM is capable of binding the calmodulin-binding
domain of eNOS to stimulate the enzymatic conversion of L-arginine to L-Citruline and NO
(Figure 1.4). NO produced in the endothelium is able to diffuse to the smooth muscle cells to
activate soluble guanylyl cyclase (sGC), resulting in the enzymatic production of cyclic
guanosine monophosphate (cGMP) from guanosine triphosphate (GTP) and leading to
relaxation of smooth muscle, dilating the blood vessel. .............................................................. 41
Figure 2.1 - Soft lithography techniques are used for the rapid prototyping of PDMS slabs from
silicon masters. Briefly, as shown in (a), a silicon wafer is spin-coated with SU-8 photoresist, and
a negative mask is placed over the photoresist prior to UV light exposure. The photoresist
exposed to light cross links to form a rigid polymer structure, and the remaining SU-8 polymer is
dissolved by a developing agent, resulting in a silicon master with raised features. As shown in
(b), PDMS polymer and curing agent is mixed in a desired ratio and degased prior to pouring
over the silicon master. The PDMS is then cured at 75° C for 15 minutes and subsequently
removed from the master, yielding a PDMS slab with embedded features. ............................... 75
Figure 2.2 - The microfluidic TEER device is fabricated from 3 PDMS slabs. The device contains a
series of flow channels integrated with a common electrode below the cell layer, which is
cultured on the polycarbonate membrane. A second copper top electrode (not pictured) is
placed in the wells of the top layer of PDMS to perform a TEER measurement directly across the
cell monolayer............................................................................................................................... 77
Figure 2.3 - Shown in the top graph is the bipolar pulsed perturbation signal used in the system,
a 20 Hz square wave. The resulting current is then measured as a function of time as shown in
the bottom graph, yielding current decays that resemble those typically seen in an RC circuit.
These current decays are then averaged and integrated, and the resulting charge value is
presented as a TEER signal in units of charge. .............................................................................. 83
Figure 2.4 - A cross-sectional view of the microfluidic device is shown in (a). Solutions
containing various concentrations of potassium chloride (KCl) are pumped into the channels of
the microfluidic device, and a solution of the same concentration is pipetted into the overlying
wells. TEER measurements are shown to correlate linearly with KCl concentration in (b),
displaying the ability of the device to measure conductance, which is inversely proportional to
xi

resistance, and therefore indicative of resistance between the two electrodes. (n=4, errors are
SD) ................................................................................................................................................. 85
Figure 2.5 - A cross-sectional view of one channel of the microfluidic device, where TEER
measurements are performed directly across an endothelial monolayer cultured on a
polycarbonate membrane. Shown in (b) are the results of TEER measurements performed on
endothelial cells over the course of 17 h in units of μCoulombs, demonstrating the cells reach
confluence at approximately 8 h. (n=4, errors are SD)................................................................. 88
Figure 2.6 - TEER measurements performed on endothelial cells over the course of 17 h are
reported in conventional TEER units of ohms·cm2. These values are estimated from the TEER
measurements displayed in Figure 2.5b. These results again demonstrate that the cells reach
confluence at approximately 8 h, as suggested by the peak in TEER signal at that time. (n=4,
errors are SD) ................................................................................................................................ 89
Figure 3.1 - Shown in the top of the image is the structure of DAF-FM, a fluorogenic probe for
NO. Below is a potential mechanism for the formation of fluorescent DAF-FM-NO. This
mechanism illustrates the reaction between NO and molecular oxygen to form a peroxynitrite
intermediate before forming DAF-FM-NO. DAF-FM-NO fluorescence is detectable at
wavelengths similar to fluorescein, with an excitation wavelength of 485 nm and emission of
515 nm. ....................................................................................................................................... 114
Figure 3.2 - The structures of DAF-FM and DAF-FM DA are shown on the left, the difference
being the acetate groups of DAF-FM DA. As illustrated here, in contrast to DAF-FM, DAF-FM DA
is capable of penetrating the cell membrane to the intracellular space where esterases cleave
the acetate groups, which results in intracellular DAF-FM. Nitric oxide can then react with
intracellular DAF-FM, producing the fluorescent DAF-FM T in amounts proportional to
intracellular NO concentration. .................................................................................................. 115
Figure 3.3 - An integrated microfluidic TEER device is shown in (a) prior to use. Panel (b) shows
the microfluidic TEER device during a flow experiment prior to using fluorescence microscopy to
study confluent cells cultured in the wells of the device. An examination of intracellular NO
production of cells cultured in the wells of the microfluidic TEER device by fluorescence
microscopy is shown in (c). ......................................................................................................... 120
Figure 3.4 - A bright-field image of a monolayer of endothelial cells is shown. Cells were
cultured to a high degree of confluence and verified to have the appearance shown here prior
to seeding in microfluidic wells. During examination, the cells are checked for a cobblestone
appearance, a lack of vacuoles (generally indicates healthy cells), as well as a lack of any visible
bacteria. ...................................................................................................................................... 122

xii

Figure 3.5 - A side view of the measurement system utilized to detect NO in the endothelial
layer is shown in (a). A solution of RBCs in PSS is pumped under a membrane containing an
endothelial layer cultured for 1.5 or 8 h. A significant increase (P < 0.05, n = 3, errors are SD) in
NO production is observed when the cells reached confluence at 8 h, as shown in the inset bar
graph. .......................................................................................................................................... 124
Figure 3.6 - A cross sectional view of a single channel of the microfluidic device utilized to
detect NO produced by the endothelial layer is shown (a). Images of wells of a microfluidic
device containing endothelial layers that had been exposed for 30 min. to untreated flowing
RBCs, or flowing RBCs treated with Zn2+/C-peptide, are compared in (b). An 88.6% ± 7.5%
increase (n = 3, p < 0.001) in intracellular endothelial NO production was observed when RBCs
were pumped in the presence of Zn2+/C-peptide and compared to RBCs alone in (b).
Interestingly, a statistical change in endothelial NO production was not observed when RBCs
exposed to C-peptide alone or Zn2+ alone are compared to RBCs alone. A statistical decrease
(n=3, p<0.02) in endothelial NO production was observed when comparing PSS lacking RBCs to
RBCs alone. Additionally, when endothelial cells were incubated with PPADS, a purinergic
receptor inhibitor, prior to flow of RBCs treated with Zn2+-activated C-peptide, an increase in
NO production was no longer observed. (* denotes p<0.001, ** denotes p<0.02) .................. 126
Figure 4.1 - Shown is a diagram illustrating C-peptide detection by sandwich ELISA. Complex
biological samples containing C-peptide are added to wells of a 96-well plate, the bottoms of
which have been coated with a primary antibody for human C-peptide. C-peptide is allowed to
bind to the antibody for a period of time, after which a secondary antibody is added to the
solution, which also binds to sandwich C-peptide between the two antibodies. Following
incubation, an enzyme is added that binds to the secondary antibody and, subsequently, a
substrate is added that allows the catalytic production of a detectable substance by the
enzyme. In the case of C-peptide ELISA, this reaction yields a product which changes the color
of the solution during the enzymatic process, which is stopped after an interval of time, causing
the amount of color produced to be proportional to C-peptide concentration. The absorbance
of the solution is read at 450 nm using a multiwell plate reader, allowing for quantification of Cpeptide in complex biological samples. ...................................................................................... 149
Figure 4.2 - A cross-sectional view of one channel of the microfluidic vascular mimic used to
assess the permeability of a pulmonary endothelium to C-peptide under flow conditions is
shown in (a). The permeation of C-peptide was monitored by detecting, via ELISA, C-peptide
that permeates to basolateral solutions (in 30 minutes) from flowing apical solutions containing
20 nM C-peptide. As shown in (b), it was found that basolateral solutions overlying a cultured
endothelium (black bars) reached 17.09 ± 1.16 nM and 18.91 ± 1.11 nM C-peptide
concentrations when flowing C-peptide solutions either contained, or did not contain RBCs,
respectively. Additionally, when endothelial layers were not cultured on polycarbonate
membranes (red bars), and flowing apical solutions either contained, or did not contain RBCs,
the basolateral solutions reached 18.31 ± 1.12 nM and 19.91 ± 0.94 nM C-peptide

xiii

concentrations, respectively. Also, as a negative control, if C-peptide is not added to apical
solutions, negligible C-peptide concentrations were found. (n=3, errors are SEM) .................. 157
Figure 4.3 - The permeability of the pulmonary endothelium to C-peptide was investigated by
monitoring C-peptide permeation from the apical compartment of a microtiter diffusion cell,
pictured in the upper right, to the basolateral compartment. 9 pmol of C-peptide is added to
the apical compartment and, after 3.5 hours of rest, C-peptide in each compartment is
quantified by ELISA and reported. As the graph shows, when 9 pmol of C-peptide is initially
added and allowed to permeate a confluent bPAEC layer (black bars), 7.67 ± 0.37 pmol and 1.12
± 0.08 pmol of C-peptide is detected in the apical and basolateral regions, respectively.
However, if 9 pmol of C-peptide is added to the apical side of a bare polycarbonate membrane
(red bars), the amounts of C-peptide detected in the apical and basolateral regions totaled 5.50
± 0.27 pmol and 3.98 ± 0.27 pmol, respectively. (n=4, Errors are SD) ....................................... 160
Figure 5.1 - Side-view of microchip with integrated TEER measurement. For simplicity, a 4reservior array is shown. Also shown is a micrograph of a 4-microelectrode (gold) array that is
embedded in polystyrene, which was fabricated in the lab of Scott Martin. The potential of the
preliminary 4-microelectrode gold array as an effective platform for TEER analysis is illustrated
by the previous utilization of the array in a microfluidic platform to detect analytes by
amperometry, an electrochemical technique that is experimentally similar to the method used
to determine TEER. ..................................................................................................................... 185
Figure 5.2 - The amino acid sequence of C-peptide is shown, which arrows highlighting residues
that are negatively charged in vivo, and thus likely capable of binding free Zn2+ ions. ............. 188

xiv

Chapter 1 - Introduction
In vitro cell-based assays have become widely used in the bioscience, biotechnology, and
drug discovery fields as they allow precise control of the environment surrounding the cells,
enabling the observation of cellular responses to specific environmental queues of interest.
Replicating a biological environment in vitro greatly simplifies the system under investigation,
and also provides an opportunity to monitor specific phenomenon that may not be observable
in the more sophisticated environments found in vivo. However, in order to successfully
extrapolate results derived from in vitro studies to the biology of the intact system found in
vivo, all necessary components that influence the biological activity of interest should be
included in an in vitro biological mimic. Microfluidic technology provides a platform which most
adequately meets these requirements as microfluidic devices can mimic microenvironments
found in vivo dimensionally, while also allowing precise control of important aspects of the
microenvironment, such as chemical signaling[1-3], physiological stresses[4-6], temperature[79], and degree of cell-cell interactions,[10, 11].
Biomedically, microfluidic systems have provided valuable platforms for the modeling of
cellular monolayers in vitro. In order to investigate cell monolayer biology with an in vitro
model, the layer of cells comprising the wall must be confluent, display a restrictive paracellular
pathway, high electrical resistance, physiologically relevant cell architecture, and functional
expression of transporter mechanisms while being easy to culture.[12, 13] Microfluidic
technology can meet these requirements, providing the successful development of a method to
determine monolayer confluence of cells cultured on a microfluidic device. Cellular monolayers
are of particular interest because they form barriers in vivo that often regulate solute transport.
1

These barriers are responsible for selectively trafficking some solutes and macromolecules
between the distinct regions defined by the barrier, while completely excluding others.
Importantly, this activity is often responsible for the exclusion of harmful toxins, maintaining
concentration gradients, influencing drug pharmacodynamics, and influencing the physiology of
various cell types by manipulating the cellular environment and regulating the introduction of
molecules that impact cell behavior. Not surprisingly, the malfunction of cells comprising
monolayers, including breakdown of monolayer barrier integrity, has been implicated in various
disease states, such as: congenital deafness[14-16], inflammatory bowel disease[17],
diabetes[18, 19], multiple sclerosis[20, 21], and even allergies.[22, 23]
Depending on the tissue they developed from in vivo, these cellular barriers are comprised
of endothelial or epithelial cells. Each of these cell types is similar in function, and, in fact,
endothelial cells are often thought of as specialized epithelial cells. Epithelial cells line most
structures and cavities in the body, such as the digestive, respiratory, and urinary tracts, as well
as many glands, whereas endothelial cells line the lymph and cardiovascular systems.
Importantly, endothelial cells line all blood contacting surfaces in the body, from the internal
lining of all types of blood vessels, to the outer lining of heart valves. These cells are involved in
several distinct and important functions including, but not limited to: regulation of vascular
tone[24-26], fluid filtration[27, 28], neutrophil recruitment[29, 30], hemostasis[31, 32],
pathogen exclusion and host response[33-35], as well as hormone trafficking.[36, 37]
Importantly, since many endothelial functions like these are influenced by the degree of cell
confluence, a determination of cell confluence prior to experimentation is imperative and
should be routine. Nevertheless, the overwhelming importance of endothelial functions is

2

continuing to be realized, and as a result, the unsurpassed ability of microfluidic technology to
mimic biological systems has been exploited for in vitro research of important vascular
processes.[38, 39]
1.1 - Monitoring Cell Monolayer Barrier Integrity
Endothelial permeability is frequently studied because it is fundamental for understanding
pathological and physiological cardiovascular processes. Molecules can permeate the
endothelium by transcellular or paracellular mechanisms.[40] For example, molecules can
permeate the endothelium paracellularly by passive diffusion between interendothelial
junctions (the space between adjacent endothelial cells). Additionally, the transport of
macromolecules, such as plasma proteins, is often accomplished by fluid-phase or absorptive
(receptor mediated) transcellular transport mechanisms where the molecule trafficked through
the cell. Generally, macromolecules are encapsulated and transported from the apical side
(facing the bloodstream) of the cell, through the cellular cytosol, to be endocytosed on the
abluminal, or basolateral side (facing the interstitium, or the region opposite of the
bloodstream), a process that is receptor mediated. In a similar transcellular mechanism of
transport, the fusion of multiple caveolae can transiently form a transcellular channel spanning
the entire length of the endothelial cell allowing for passive diffusion of plasma proteins
through the cell in a mechanism not requiring receptor mediation. Additionally, aquaporins can
form channels across the lipid bilayer that is highly specific for water soluble molecules, thus
creating a transendothelial pathway for water and water solutes.
Endothelial cells molecularly regulate permeability by the formation of tight junctions
between adjacent endothelial cells, causing the paracellular diffusion of most molecules to be
3

prohibited.[41, 42] In fact, tight junctions have been discovered to be primarily responsible for
the regulation of endothelial permeability.[43] Surprisingly, tight junctions can limit paracellular
transport so much that only small water soluble molecules with an ionic radius of less than 3
nm are typically capable of permeating these tight junctions, predominantly limiting
interendothelial diffusion to only ions.[40] Tight junctions form a continuous, circumferential
barrier surrounding endothelial cells that serves as a boundary between the apical and
basolateral domains, as well as a fence within the plasma membrane, offering protection to the
cells from the external environment while mediating the paracellular exchange of
substances.[43] Structurally, tight junctions are composed of clusters of integral proteins, some
of which span the intercellular space and interact with adjacent integral intracellular proteins of
nearby cells to complete the protein cluster and form an intercellular boundary.
Several integral transmembrane proteins are responsible for forming cell-cell contacts,
including: the claudin family of proteins (1-24), occludin, tricellulin, JAM-A, CAR, CLMP, ESAM,
JAM4, and CRB3.[44] These integral proteins are organized based on structure in three classes.
The first class includes claudins, occludin, and tricellulin, all of which include four
transmembrane regions, two extracellular loops, and two cytoplasmic tails. Another class
includes CRB3, which is characterized by a short extracellular domain, a single transmembrane
domain, and a short cytoplasmic tail. CRB3 differs from the other integral membrane proteins
as the function of the extracellular domain is not yet clear. The remaining integral proteins
make up another class, and are characterized by two immunoglobulin-like domains. These
integral proteins are linked to the actin cytoskeleton of the cell by classic scaffolding proteins,
such as the zonula occludens (ZO) proteins (ZO-1, ZO-2, and ZO-3) among others.[44] Together,

4

the protein network forms a barrier capable of mediating paracellular transport. Importantly,
the complexity of the tight junctional network depends on cell type, yielding differences in
barrier integrity and function between different tissues and organs.[45] It is also important to
note that tight junctional control of permeability is not statically controlled, but is an organized,
dynamic process controlled by molecular signaling.[46] While tight junctions are primarily
responsible for regulating monolayer permeability, another protein cluster can also play a role
in paracellular transport by forming adherens, but these junctions are primarily responsible for
providing structural support to tissue. However, an ideal measure of endothelial barrier
integrity will account for any restriction in permeability caused by tight junctions and adherens
junctions, since adherens junctions are essential to the maintenance of endothelial barrier
function.[47]
Interest in cell monolayer integrity has led to the development of several techniques to
evaluate this integrity. Cell confluence and integrity can be determined in vitro by a variety of
methods, including: the acquisition of an image of the cells by optical microscopy, measuring
the permeability of radiolabeled or fluorescently tagged molecules, immunostaining for
proteins characteristic of tight junctions, and measuring transendothelial/epithelial electrical
resistance. Each of these methods has utility in various situations, however they are drastically
different and, therefore, the significance and meaning of results acquired by each method
should be well understood.
1.1.1 - Optical Microscopy
A simple method for verifying confluence of cultured cells is by the visual assessment of a
bright-field image of the cells acquired via optical microscopy. This method is quite fast and
5

simple, and requires minimal training to learn. Additionally, visualization of the cells allows for
the inspection of cell morphology and detection of undesirable bacteria. While the general
health and extent of confluence can be estimated by this method, it is a qualitative and fairly
subjective technique. Conclusions resulting from the visual assessment of a cell culture by
different individuals are never quantitative and can vary, inevitably introducing bias to the
method. Furthermore, it is difficult to detect voids between non-manipulated cells and
impossible to verify the presence of tight junctions by optical microscopy. These issues are
often compounded when only a small area of the cell culture is examined, as is typical for this
method. For these reasons, only a subjective and qualitative estimation of cell health and
confluence can be obtained by this method without modification of the cell, whereas a
quantitative measure of monolayer barrier integrity is preferred.
Ideally, a quantitative measure of cell confluence is attainable so that bias does not
influence the measurement, yielding more accurate and objective conclusions than those
derived from qualitative measurements. Furthermore, an analysis that encompasses the area of
the entire cell culture is preferable so that areas of compromised barrier integrity are not
overlooked. Not surprisingly, developments have been made in order to improve these issues
and broaden the utility of the technique, often by modifying the cells prior to analysis, but not
always. Recently, a method for quantitatively determining monolayer confluence was reported
that involves using optical phase microscopy in conjunction with computer software to detect
intercellular voids.[48, 49] Detection of these voids by commonly used bright-field optical
microscopy is not plausible due to the minimal variations in light intensity between paracellular
voids and areas covered by the transparent cells. In contrast, optical phase microscopy results

6

in the destructive interference of light that passes through a specimen, causing the areas
covered by cells to be dark with a light background allowing for the differentiation and
quantification of these different phases by computational image analysis. However, the
successful detection of small paracellular voids requires a detailed image, an issue that
potentially increases the likelihood that areas of compromised barrier integrity are unnoticed
since the analysis of an entire cell culture with a single image is often not feasible. Furthermore,
these techniques are limited to verifying monolayer confluence since the modes of transport
across the monolayer, or factors associated with proper barrier function, are not investigable
by these methods and therefore are not ideal techniques for determining monolayer
permeability and barrier function.
An additional method for verifying cell confluence is immunostaining proteins characteristic
of tight junctions (occludins, ZO-1, and ZO-2) for detection by fluorescence microscopy.[50] This
method has proven useful for verifying barrier function of endothelial monolayers,[51-53] in
addition to detecting reductions in barrier function in response to various biological signals.[5456] Furthermore, this method allows for the visualization of paracellular voids so that areas of
barrier compromise can be detected by eye. Undoubtedly, this is a powerful method for
determining tight-junctional control of paracellular transport and the extent of monolayer
barrier integrity; however this labeling technique is not quantitative. While bias in this
measurement is minimized, a visual detection of paracellular voids is still required, which may
be accidentally overlooked. Furthermore, this method is limited to a single determination
observation for a specific cell culture, eliminating the possibility of a time-lapse analysis of
barrier integrity or confluence of the cultured cell monolayer, unless compared to a separate

7

cell culture that serves as a control. Additionally, a label-free determination of barrier integrity
is preferred as labeling techniques can sometimes alter cell behavior, an important
consideration if further experimentation is necessary post-labeling. Additionally worth noting is
a microscopy technique, ion conductance microscopy, and while it is not an optical technique, it
has been adapted for evaluating monolayer barrier integrity since an image in very fine detail of
practically any surface can be obtained.[57, 58] However, this method is not yet capable of
providing a quantitative measurement of barrier integrity for the complete cell layer.
1.1.2 - Permeability Assays
A straightforward method for the determination of barrier integrity is to monitor the
pervasion of an analyte through the monolayer of cells.[59] Typically, endothelial cells are
cultured upon a polycarbonate membrane that has been coated with an adhesion molecule
such as fibronectin or collagen.[60] After the cells are thought to reach confluency, the
membrane is placed in a diffusion cell as shown in Figure 1.1 (typically a cell culture insert
placed in the well of a microtiter multiwell plate), while permeation of an analyte from the
apical (abluminal) to the basolateral (luminal) compartments is monitored.[60-63] These
analytes, or tracers, are typically fluorescently or isotopically labeled.[64] Resulting
permeability coefficients are useful for verifying the barrier integrity of a biological mimic, or
for estimating the ability of a drug of interest to permeate a barrier in vivo.[65] Some typical
tracers include sodium fluorescein, lucifer yellow, fluorescein isothiocyanate (FITC) labeled
dextrans, [14C] labeled sucrose[66], and [3H] labeled albumin[67]. By determining permeation
coefficients of an analyte across a monolayer of cells, a quantitative determination of barrier
integrity is achieved that is dependent on the entire cell culture area.
8

Figure 1.1 - A microtiter cell culture insert (membrane insert) is shown resting in a microtiter
well, forming a diffusion cell most commonly used for permeability assays. Typically, a
monolayer of cells is cultured upon a collagen or fibronectin coated membrane at the bottom
of the membrane insert. The permeation of analytes or drugs of interest from the apical
solution to the basolateral solution can be monitored to estimate the in vivo permeability of
particular cellular barriers, represented by the cultured cells, to the molecules under
investigation. Also, the confluence and barrier integrity of the cultured cellular monolayer can
be determined by monitoring the permeation of a tracer molecule from the apical
compartment to the basolateral compartment and calculating a permeability coefficient, which
is proportional to layer barrier integrity. For interpretation of the references to color in this and
all other figures, the reader is referred to the electronic version of this dissertation.

9

The determination of permeation coefficients has become a dominant method for
predicting drug absorption in vivo.[68] Monitoring drug permeation across a cultured
monolayer of cells is often superior to other methods for predicting drug absorption in vivo
because it accounts for all modes of transport of a drug through a cellular barrier. Drugs are
able to permeate cellular barriers in order to reach a desired organ of interest by paracellular
and transcellular mechanisms as discussed previously.[40] All possible transcellular mechanisms
are accounted for when determining permeation coefficients of a particular analyte, as well as
paracellular transport around individual endothelial cells through interendothelial junctions.
However, if a confluent endothelial monolayer is intact, transport through interendothelial
junctions will be under tight junctional control where typically only small molecules with a
molecular radius of less than 3 nm, such as glucose, urea, and small ions, will successfully
permeate the monolayer paracellularly.[40]
In addition to drug absorption studies, permeation coefficients of a molecule are
particularly also useful for determining factors that can alter or influence monolayer
permeability in vivo.[69, 70] However, cell monolayer confluence is dependent on complete
endothelial coverage and the development of tight junctional intercellular contacts, and since a
healthy endothelial barrier will still allow transcellular transport, a superior measurement for
monitoring monolayer growth and confluence should involve the permeation of molecules
whose permeation is primarily restricted by interendothelial junctions as opposed to
transcellular transport. For example, interendothelial junctions under tight junctional control
restrict all molecules aside from very small molecules (molecular radius < 3 nm), the
interendothelial flux of small water soluble molecules are generally determined primarily by cell

10

confluence as opposed to transcellular transport and, therefore, a measurement of that flux
will adequately gauge cell growth, confluence, and barrier integrity.
1.1.3 - Measuring Transendothelial Electrical Resistance (TEER)
A continuous endothelial monolayer under tight junctional control forms a barrier capable
of restricting the motion of molecules as small as ions,[71] and because of this, a
straightforward way to determine cell monolayer confluence and barrier integrity is by
measuring the ability of the monolayer to restrict the motion of ions. TEER measurements
determine the resistance to current induced by an endothelial monolayer when a potential
difference is applied across the cellular barrier. The ability of the monolayer to restrict the
interendothelial flux of ions, measured by current, is dependent upon the degree of endothelial
confluence and barrier integrity induced by the formation of tight junctions between the cells.
In this construct, a fully confluent cell monolayer with a high degree of barrier integrity will
have a decreased current response, and therefore increased electrical resistance, when
compared to a cell monolayer with compromised barrier integrity, assuming that potential
applied across each monolayer is of equal magnitude.
By measuring the ability of a monolayer of cells to restrict the motion of ions, a quantitative
measure of monolayer confluence and barrier function can be obtained. For example, the most
common TEER evaluation is the measurement of the electrical resistance induced by the
cellular monolayer. Not only is measuring TEER a quantitative technique, these measurements
encompass the entire monolayer and thereby offer a characterization of the overall barrier
function of the collective cell culture as opposed to only a portion of the cell layer, as is typical
of most optical techniques for determining confluence. Importantly, TEER measurements are
11

quite useful for verifying monolayer confluence, as TEER values change exponentially during
periods of cell growth and death, but remain constant while the cells are at a confluent state.
Therefore, comprehensive cell confluence can be verified by multiple sustained TEER
measurements over a period of time.
It should be noted that the relationship between transport of solutes (as measured by
permeability coefficients) and electrical resistance is nonlinear.[65, 72] This is because solute
transport depends on the sum of transport across all transcellular and paracellular
pathways[64], while electrical resistance is believed to be dependent on the paracellular
permeability of small ions through the interendothelial junctions between endothelial cells.[65]
These junctions generally restrict the paracellular transport of hydrophilic and large
molecules.[72, 73] While both measurements are indicative of barrier integrity, one must
acknowledge the meaning of each measurement when performing experiments.
Obtaining TEER of a cell monolayer to evaluate confluence and barrier integrity is
particularly beneficial as it does not damage the cell or alter cell phenotype, and requires only a
matter of seconds to perform each measurement. Experimentally, these advantages make TEER
measurements uniquely suitable for time-lapse analysis of monolayer integrity due to the
ability to repeatedly acquire a quantitative measure of monolayer integrity in a timely fashion
without altering cell behavior or phenotype. Consequentially, TEER measurements are
frequently involved in investigations of factors affecting cell monolayer barrier solidification[7476], as well as compromise[77-79], in studies where a basal monolayer integrity can be
achieved and verified prior to measuring change in barrier integrity in response to stimuli of
interest. These studies have the experimental advantage of confirming monolayer barrier

12

integrity prior to investigation, yielding the associated benefits mentioned previously, while
also eliminating the need for separate cell cultures for negative controls that are often
necessary for methods that only allow a one-time measurement of barrier integrity, such as
immunostaining tight junctional proteins. This ensures that a change in barrier integrity is truly
in response to the intended stimuli, eliminating the assumption that the control cell monolayer
exhibits identical barrier integrity as the experimental cell monolayer prior to experimentation.
These experimental benefits are in addition to the fundamental advantages granted by the
ability to perform multiple measurements without altering cell behavior, resulting in a method
that allows for the verification of cell monolayer confluence prior to experimentation which, in
turn, ensures proper cellular behavior and increased reproducibility between biological models.
As a result of these important experimental advantages, measuring TEER has become the most
commonly used method for investigating cell growth and confluence.
The first reports of monitoring TEER across a cultured monolayer of brain endothelial cells
were in 1987[80, 81]. Since then, performing TEER measurements has become a common and
simple technique with the production of commercial instruments. Most of the commercially
based methods for measuring TEER across endothelial cell cultures adhered to microporous
filters or culture inserts employ Volt-Ohm resistance meters equipped with “chopstick”
electrodes, or chamber electrodes as shown in Figure 1.2. The same microtiter technology used
to create a diffusion cell for permeability studies may be used for measuring TEER across a
monolayer of cells. However, endothelial cells cover all blood-contacting surfaces of the
cardiovascular system and, due to their location, they experience constant exposure to
hemodynamic shear stress generated by flowing blood. In vitro, applying shear stress on

13

endothelial cells is critical for the expression of certain cellular responses because shear stress
has been shown to regulate many functions of the endothelium in vivo.[82] Additionally, many
functions of the endothelium are mediated by cell-cell communication,[30, 83-90] so the
presence and the degree of cell-cell interactions should be considered in an in vitro model of
the vasculature.

14

Figure 1.2 - A traditional TEER measurement utilizes a microtiter membrane insert that fits into
a traditional well plate. The cell layer is cultured on the polycarbonate membrane at the
bottom of the insert and placed into the well, which contains cell media. The TEER
measurement is performed by placing electrodes in the basolateral compartment on the
outside of the insert and in the apical compartment on the inside of the insert containing the
cell layer, and measuring the resistance between the two, which is primarily determined by the
barrier integrity of the cultured cells.

15

1.1.4 - Importance of Monitoring Barrier Integrity in Vascular Mimics
It is fundamentally important that models of biological systems be comprised of fully
confluent monolayers of cells to form barriers that each exhibit physiologically relevant
paracellular permeability because healthy tissue barriers are comprised of confluent
monolayers of cells in vivo. Determination of cellular monolayer barrier integrity, specifically
tight junctional control of transendothelial permeability, is quite important since cell behavior,
including responses to agonists/antagonists and transport of molecules, has been shown to
vary with cell confluence and integrity.[91-100] The dependence of cell behavior on barrier
integrity highlights the importance of accurately reproducing conditions found in vivo when
modeling biological systems in vitro. As a result, cell monolayer integrity should be determined
to more accurately reproduce realistic cell physiology, granting a more appropriate cellular
response to stimuli and more realistic monolayer permeability.
Physiologically, conclusions stemming from experimental results involving cellular
monolayers that are confirmed to be confluent are generally more accurate due to added
physiological relevance in terms of cell behavior and response, adding credibility to the work.
Also, measuring barrier integrity to determine monolayer confluence reduces variability
between different cell culture models, which is often rather large, allowing for the direct
comparison of a particular model to different models, as well as to in vivo conditions.
Furthermore, verifying monolayer confluence increases reproducibility between models of the
same type, which is sometimes low, by ensuring the cultured cells are at the same stage of
growth at the time of experimentation. Collectively, these experimental advantages have
caused the widespread and common practice of obtaining a determination of confluence in
16

studies involving monolayers of cells, and furthermore, measurements of barrier integrity are
now often requisite for acceptance of results from such studies. Additionally, the ability to
monitor cell confluence and barrier integrity in a vascular model introduces a new scope of
potential research, specifically the investigation of factors affecting cell proliferation, death,
and permeability and corresponding alterations in cellular behavior.
1.2 - Mimicking the Vasculature
The human cardiovascular system is composed of the heart, blood, and the circulatory
network of blood vessels, or the vasculature. The blood consists of several components, the red
blood cell (RBC), being the most common blood cell, the platelet, the leukocyte, and the
plasma, in which the other blood components are suspended. The plasma is primarily water
(over 90% by volume), and makes up about 55% of whole blood by volume with the average
human having about 5 to 6 liters of whole blood. The blood vessels can be described simply as a
network of tubes in which blood can flow throughout the body. Blood vessels are surrounded
by smooth muscle that is capable of expansion and contraction. This smooth muscle is lined
with a continuous monolayer of endothelial cells to form the endothelium, which completely
lines the interior of all blood vessels and surrounds flowing blood as depicted in Figure 1.3.
Together, the smooth muscle and the endothelium form the vascular wall, providing separation
between flowing blood and the surrounding tissue. The heart pumps blood through the
vasculature, first through arteries, which are oxygen rich, to arterioles and finally capillaries. As
the blood is pumped farther through the vasculature and away from the heart, the blood
vessels begin to decrease in diameter from arterioles to capillaries, which have the smallest
diameter. Arterioles and capillaries are considered resistance vessels because they resist the
17

flow of blood at varying degrees depending on vessel diameter, which ranges from 20 µm to
250 µm, with arterioles generally having larger diameters than capillaries. Importantly, the
hematocrit, or percentage of blood occupied by RBCs by volume is typically about 45%, but can
be reduced to less than 10% in resistance vessels.[101]

18

Figure 1.3 - A cross-sectional view of a resistance vessel. Flowing blood components, such as
the RBC, are contained in the core of the vessel. Surrounding the flowing blood and lining the
interior of all blood vessels is a confluent and continuous monolayer of endothelial cells, which
collectively form the endothelium. The endothelium forms a barrier, separating the flowing
blood from exterior smooth muscle, which is responsible for contracting and relaxing the blood
vessel to regulate vascular tone. In the smooth muscle, the potent vasodilator, NO, is capable of
causing vessel relaxation; however the source of this NO is under investigation. In two
prominent mechanisms, the primary source is either NO directly released from the RBC, or NO
produced in the endothelium in response to ATP released from the RBC.

19

Interestingly, it is within these resistance vessels, particularly capillaries that nutrients such
as oxygen are provided to surrounding tissues and organs, with RBCs being the principal carrier
of oxygen in the blood stream. In order to adequately supply nutrients to surrounding tissue,
blood vessels are capable of relaxing to accept cardiac output, especially in regions where RBCs
(and therefore oxygen supply) may be scarce ( often ≤10% hematocrit). After the exchange of
nutrients, blood is circulated back up to the heart through veins that increase in diameter as
the heart is approached. Importantly, the deoxygenated blood is pumped by the heart to the
lungs where the entire cardiac output is accepted for reoxygenation before returning to the
heart, completing the cycle, to be once again pumped throughout the body with a renewed
oxygen supply.
Perhaps the most important aspect of the cardiac cycle is the ability of resistance vessels to
dilate, allowing blood to flow to nutrient poor tissue, and thereby selectively regulating blood
flow to increase distribution efficiency. In 1988, groundbreaking work reported that nitric oxide
(NO) plays a primary role in regulating vessel dilation through activity in the smooth muscle
that results in relaxation of the muscle, increasing vessel diameter and increasing local blood
flow. However, in the time since this work, there has been considerable debate as to the source
of this NO. Regardless of the specific source of this NO, the most important determinant of
continuous NO production, and thus the most important factor in local blood flow and vessel
dilation, is shear stress generated by flowing blood. Therefore, the source of this NO is a topic
of intense research focus due to its fundamental importance in vascular function and disease.
As a result of this research focus, two general arguments have primarily been put forth, the first
hypothesis citing the RBC as a direct source of NO, and the second citing the endothelium as

20

the source of NO as depicted in Figure 1.3. These hypotheses and associated evidence will be
discussed in detail within the following sections.
Vascular wall biology, and interactions with bloodstream components (e.g., RBCs, platelets,
and leukocytes), is an area of continued interest due to the high prevalence of cardiovascular
disease and stroke, as well as the growing realization of its increasingly important roles in many
fundamental vascular functions, such as vessel dilation. As stated previously, in order to
investigate vascular wall biology with an in vitro model, the layer of cells comprising the wall
must be confluent and display a restrictive paracellular pathway, high electrical resistance,
physiologically relevant cell architecture, and functional expression of transporter mechanisms
while being easy to culture.[12, 13] Unfortunately, meeting these criteria is challenging due to
the intricate environment found in vivo in which various aspects of the environment and
sometimes multiple cell types collectively contribute in complex mechanisms to facilitate
biological phenomena of interest. As a result, important biological functions are often
multifaceted in nature, often being influenced by several environmental queues and cell types
while, in turn, influencing other subsequent biological activity. This collective regulation of
biological activity where the interplay between several factors ultimately dictates the
conception and magnitude of a single biological event is common in the vasculature. As a result,
many factors must be considered when attempting to replicate the environment found in vivo
because the exclusion of elements affecting the desired activity in question will yield
biologically irrelevant behavior, potentially resulting in incorrect conclusions based on
extraneous results. Unfortunately, the factors affecting a specific biological event are not
always known and, in fact, are often the focus of in vitro research. Thus, the adequate

21

reproduction of the biological system of interest is a persistent challenge. Regardless, the fact
remains that obtaining meaningful finings in vitro that accurately correlate to activity observed
in vivo hinges on successful reproduction of aspects of the environment that are required for,
or play a role in, the activity under investigation.
In order to address these challenges, technologies and methods, including microfluidic
technology, have been developed that enable more accurate mimicry of biological systems. For
example, microtiter technology is the most commonly used platform to mimic the vasculature,
and is described in section 1.1.2. Briefly, endothelial cells are cultured upon a well insert, which
is placed in a microtiter well containing a buffer solution, ultimately resulting in a diffusion cell.
Alternatively, cells can be simply cultured in the wells of a microtiter plate. These methods
allow for the culture of an intact endothelium where many components of an in vivo
environment can be reproduced, such as pH, molecular content, temperature, and gas content.
Additionally, cellular response to added stimuli is easily observable by several methods,
including fluorescence or chemiluminescence monitoring, absorption changes, alterations in
cell morphology and scintillation counting. Conveniently, a large microtiter based infrastructure
exists so that high throughput monitoring of cellular response by these methods is achievable
via instrumentation commonly found in typical clinical or biological laboratories. Unfortunately,
microtiter technology cannot investigate many vascular functions and activities.
Microtiter technology inadequately reproduces several important biological characteristics
of the vasculature that are crucial for producing or regulating an abundance of vascular events
involved in normal vascular function, in addition to activities resulting from various diseases, or
those involved in the progression of cardiovascular disease. Specifically, microtiter technology

22

fails to account for the flow of blood, or in some cases the presence of blood whatsoever.
Importantly, the fluid shear stress resulting from blood flow is critical for essential vascular
functions, such as vasodilation as previously discussed, and is determinant in several
fundamental functions of the endothelial cell, as well as the red blood cell. Unfortunately, the
majority of studies involving the vascular wall are performed in static systems that do not
adequately represent dynamic in vivo conditions where shear stress is essential for important
biological phenomenon. Furthermore, the inherent geometries of microtiter systems are not
capable of reproducing the microenvironments found in vivo. For example, these geometries do
not adequately reproduce physiologically relevant distances between cells, limiting the ability
to derive meaningful conclusions from results where cell-cell communication plays a role. The
importance of each of these aspects will be discussed in further detail in the subsequent
sections describing important components of the vascular system, especially in the regulation
of vascular tone.
1.2.1 - The Endothelial Cell
In order to properly describe various measurements of barrier integrity and the importance
of such measurements, the endothelial cell was introduced in Section 1.1 where several cellular
functions observed in vivo were discussed in addition to important cellular architecture
involved in regulating endothelial permeability. To briefly reiterate, endothelial cells cover all
blood-contacting tissue in a continuous monolayer which forms a barrier that separates flowing
luminal blood from the surrounding tissue. Interestingly, endothelial cells form tight junctional
cell-cell contacts that restrict interendothelial permeability, and collectively, these endothelial
cells form a continuous monolayer that serves as a protective barrier called the endothelium.
23

As such, the molecular regulation of vascular permeability prevents pathogen transport to and
from the bloodstream, while facilitating the selective transport of macromolecules and solutes
between the tissue interstitium and the luminal blood, and vice versa.
In addition to these important biological functions, the endothelium plays a vital role in
regulating vascular tone.[102] The first evidence that the endothelium is crucial for vascular
tone regulation was the observation by Furchgott and Zawadzki that the endothelial lining was
required for the observation of acetylcholine induced relaxation in isolated sections of rabbit
aorta.[103] Relaxation induced in the presence of the endothelium was attributed to an
endothelial-derived relaxing factor, or EDRF, and that soluble guanylyl cyclase in the
surrounding smooth muscle cells was the target of the EDRF upon its calcium-dependent
release from the endothelium.[104, 105] Eventually, several observations led Furchgott and
Ignarro to independently determine that nitric oxide (NO) was the EDRF, marking the discovery
of the first gaseous signaling molecule. Subsequently, it was determined that endothelial NO
release indeed did account for the biological activity of the EDRF.[106] Finally, the L-arginine –
NO-synthase pathway was identified as the enzymatic source of endothelial NO
production.[107, 108] Importantly, endothelial nitric oxide synthase (eNOS) was identified as
the enzyme responsible for the conversion of L-arginine to L-citruline and NO as shown in
Figure 1.4. [109] Furthermore, eNOS activity was found to be dependent on calmodulin activity
2+

and an increase of free intracellular Ca

levels.[110] In accordance with the impact of the

discovered role of the endothelium in vasorelaxation and subsequent identification of NO as
the EDRF, Furchgott, Ignarro, and Murad won the Nobel Prize in Medicine in 1998.

24

Following these important findings, external factors inducing or influencing endothelial NO
production are still being investigated, however several stimuli have been identified, including:
bradykinin, acetylcholine, prostacyclin, estrogen and adenosine triphosphate (ATP).[111-114]
These factors stimulate a signaling cascade resulting in the enzymatic oxidation of L-arginine to
L-citruline and NO by eNOS. The catalysis of this reaction requires a number of essential
cofactors such as calmodulin (CaM), FAD, NADPH, tetrahydrobiopterin (H4B), and flavin
mononucleotide.[115] Upon stimulation by an external factor, these proteins and transduction
molecules required for NO production are co-localized in membrane compartments to jointly
form the active “eNOS signaling complex” that is responsible for enzymatic endothelial NO
production.[115] It is important to note that eNOS exhibits a CaM-binding motif where CaM
binds eNOS to facilitate NADPH-mediated electron flux required for NO production by the
2+

complex, however, Ca

binding to CaM enhances the binding of Ca-CaM to eNOS and electron

flux of the complex is dependent on CaM phosphorylation and dephosphorylation.[116, 117]

25

Figure 1.4 - Mechanism describing the production of nitric oxide (NO) in the endothelium. Free
intracellular calcium binds calmodulin to form the calcium calmodulin complex (Ca-CaM), which
stimulates endothelial nitric oxide synthase (eNOS) by facilitating electron transfer through the
complex and thereby increasing the oxidation rate of NADPH to NADP+. The oxygen dependent
oxidation of NADPH by eNOS provides electrons for the enzymatic conversion of L-arginine to
citruline and NO. Therefore, the increase in electron transfer kinetics resulting from Ca-CaM
binding ultimately increases NO production in the endothelium. For simplicity, other essential
proteins and signaling molecules which are involved in the signaling complex formed during
endothelial NO synthesis are discussed previously, but not pictured.

26

Prior to the identification of NO as the EDRF, it was shown that its production required an
2+

increase in intracellular Ca

concentration as previously discussed. Additionally, after CaM was
2+

shown as a necessary cofactor for eNOS activity, it was shown that Ca

enhances CaM binding

to eNOS. Subsequently, it has been shown that shear induced NO production can be
2+

accomplished at basal [Ca ] levels when eNOS is phosphorylated to increase activity, however
2+

free intracellular Ca

was still required for NO production. Summarily, these results indicate

2+

that intracellular [Ca ] plays an important role in endothelial NO production and regulation of
vascular tone. Interestingly, P2 purinergic receptors, specifically the P2Y purinoreceptor located
2+

2+

on the endothelial cell membrane,[118] regulates intracellular [Ca ] by mediating Ca influx in
response to agonist binding.[119]
The P2 purinoreceptors are G-coupled protein receptors that have a wide distribution of
physiological roles throughout the body and in various cell types, most of which have been well
established.[120, 121] However, purinergic signaling has recently been implicated in the
regulation of vessel tone by a mechanism that is not yet well established.[122] Upon binding of
an external agonist, the endothelial P2Y purinoreceptor influences vasodilation by initiating a
purinergic signaling cascade that results in the production of smooth muscle relaxing factors by
the endothelium, including NO, which seems to be the most potent.[106, 123-130] This activity
can be achieved by extracellular ATP levels, as ATP is the primary agonist of the P2Y
receptor.[131, 132] Additionally, the resulting endothelial NO upon P2Y activation can be
attributed to eNOS activity resulting from purinergic signaling.[133-135] It has been shown that
ATP in the blood plasma can increase endothelial NO production and induce vasodilation in

27

humans via purinergic signaling,[136] and malfunction of this ATP mediated P2 purinergic
pathway is observed in several disease states, such as hypertension and diabetes, where
vasodilation is dysfunctional leading to blood flow issues.[137]
Upon successful production of NO by endothelial cells, the gaseous radical can only diffuse a
short distance as the half-life of NO in physiological solution is short, being on the order of
seconds.[138] Within this time frame, NO can diffuse to smooth muscle cells since the
endothelium is surrounded by, and in direct contact with, smooth muscle in vivo. In smooth
muscle cells, NO can activate the soluble guanylyl cyclase resulting in the enzymatic production
of cGMP from GTP. Increases in intracellular cGMP levels result in vasorelaxation. This
regulation of vascular tone by cGMP may be mediated by a number of mechanisms, but is
2+

thought to be primarily due to cGMP induced changes in the concentration of free Ca

in

smooth muscle cells.[139] Physiologically, the most important determinant of the continuous
production of NO, and therefore the most important factor in regulating local blood flow and
vascular tone is shear stress generated by flowing blood.[115] However, the mechanism of
vasodilation in response to shear stress had not yet been well established, although it is clear
that shear induced NO is regulatory of vascular tone via its activity in the smooth muscle
cell.[140] Regardless of this knowledge, the source of this NO is a topic of considerable debate
in the literature. To date, the leading proposed mechanisms of shear induced vasorelaxation
involve the red blood cell.
1.2.2 - The Red Blood Cell
RBCs, or erythrocytes, are the most common type of blood component and are the primary
carrier of nutrients, primarily oxygen (O2), through the bloodstream to replenish more hypoxic
28

(oxygen deprived) tissue and organs as discussed formerly. RBCs consist mainly of water and
hemoglobin, a metalloprotein consisting of four subunits. Hemoglobin contains heme groups
whose iron reversibly binds oxygen that is subsequently released throughout the body in
regions of lower oxygen content and usually replaced by carbon dioxide, a common waste
product of tissue. Upon oxygen binding or release, hemoglobin fluctuates between two
confirmations called the relaxed and tense states, respectively.[141, 142] Furthermore, NO can
also bind oxygenated hemoglobin at a specific cysteine residue as s-nitrosothiol. Interestingly,
this NO is released readily when hemoglobin is deoxygenated and in the tense state.[143] As a
result of these findings, it has been suggested that the red blood cell regulates vasodilation
directly by releasing this NO in regions where the RBC is exposed to low oxygen content.
Interestingly, because the RBC releases oxygen, and therefore NO, primarily in resistance
vessels where shear stress is experienced by the RBC and the endothelium alike, it is proposed
that this RBC derived NO could be the source of shear related NO that is important for blood
flow and oxygen replenishment, even though this NO is released in response to hypoxia, or low
oxygen content, and not shear.[144]
An additional hypothesis implicating the RBC as the direct source of shear induced NO
production is based on the ability of deoxygenated hemoglobin to act as a nitrite reductase,
producing NO by the enzymatic reduction of nitrite.[145] This hypothesis is supported by
research showing that nitrite injections reduce blood pressure in the human arm.[146]
However, it has been shown that nitrite is capable of stimulating ATP release from the RBC, a
finding that adds credibility to the hypothesis discussed below.[147] Interestingly, each of these
hypotheses that suggest that NO produced in association with shear stress that regulates vessel

29

dilation is actually a result of the RBCs response to hypoxia and not shear stress. Furthermore,
in each hypothesis, it is suggested that this NO originates from the RBC directly. In this
construct, the NO released from the RBC would be required to diffuse from a location in the
vessel and through the endothelium to reach the smooth muscle cell where NO is known to
induce vasodilation. Unpublished data resulting from studies in the Spence lab suggest that NO
produced by hypoxic RBCs is essentially blocked by an endothelium. These results are
supported by the short half-life of NO in physiological solutions.[148] However, it is important
to note that endothelial permeability varies between organs as stated formerly, so the
importance of RBC derived NO could potentially vary based on the anatomical location of
release.
In a final alternate hypothesis, the ability of the RBC to release ATP can indirectly result in
NO production by the endothelium that induces vasodilation. It has been shown that the RBC
contains millimolar levels of ATP[149] that can be released in response to several stimuli,
2+

including shear stress[150], hypoxia[151, 152], pharmaceutical agents[90, 153], and Zn activated C-peptide[154]. Notably, the mechanism of this ATP release is not a passive process,
or a result of cell lysis, but rather an active process mediated by a G-protein coupled receptor,
of which the mechanism has been elucidated by Sprague et al.[155-158] Also, it has been
shown that this ATP release can be inhibited pharmacologically, verifying that this process is
indeed regulated.[155, 159]
As shown in Figure 1.6, ATP is proposed to be an important regulator of vascular tone
resulting from its activity in purinergic signaling, which is discussed in Section 1.2.1. In this
construct, ATP released in the bloodstream by RBCs can activate the endothelial P2Y
30

purinoreceptor, resulting in NO production by a mechanism shown to result in vasodilation as
discussed previously. Furthermore, by this mechanism, NO is produced in response to shear
stress, adding probability that this mechanism is the source of NO produced in association with
shear stress, the origin of which has been, and is currently, debated in the literature.
Furthermore, the production of endothelial NO has been shown to be capable of, and in some
cases, necessary for the relaxation of blood vessels as mentioned earlier. In addition to
determining the mechanism of shear induced vasorelaxation, identifying other factors
influencing vessel dilation is an active area of research due to the biological importance of the
regulation of vascular tone.
1.2.3 - C-peptide
C-peptide, whose sequence is shown in Figure 1.5, is a 31 amino acid peptide produced in
the beta cells of the islets of Langerhans within the pancreas and released in equimolar
amounts with insulin into the bloodstream.[160, 161] For the first few decades after its
discovery in the late 1960's, C-peptide was thought to have minimal physiological effects after
release into the bloodstream.[162] However, numerous reports since the mid 1990's have
confirmed that C-peptide is a bioactive peptide that can have significant beneficial effects in
organs, tissue, and even cultured cells such as endothelial cells.[163-165] Studies involving Type
1 diabetic subjects who are unable to naturally produce C-peptide due to beta cell destruction
have shown that C-peptide improves nerve conduction and regeneration[166], renal
function[167], inflammation,[168] and blood flow[169, 170] via unknown mechanisms. Diabetic
patients are known to develop hypertension as well as other cardiovascular problems, which
are often ameliorated by C-peptide treatment, although the mechanism by which C-peptide
31

impacts diabetic blood flow is currently unknown, as are any influences on blood flow in
healthy individuals. However, in light of the importance of NO in the regulation of vascular
tone, it is clear that if C-peptide is capable of stimulating endothelial nitric oxide production, it
could account for the C-peptide induced improvement in blood flow observed in Type 1
diabetic subjects.
Interestingly, in separate studies, diabetic complications related to blood flow have been
shown to be related to endothelial dysfunction,[171] and nitric oxide production, suggesting
that C-peptide may be capable of stimulating NO production in the bloodstream that is most
likely derived from the endothelium.[172] The fundamental importance of endothelial derived
NO for the regulation of vessel tone has been well-established and, as a result, the
identification of factors influencing NO production in the endothelium has become a highly
active area of research which garners much attention due to the associated biological impact.
For these reasons, a potential role for C-peptide in endothelial NO regulation is particularly
exciting. This potential activity would not only provide an explanation for observed
improvements in diabetic blood flow by a process which has so far eluded identification, but
also reveal a fundamentally important and overlooked role of C-peptide in regulating vascular
tone.
A relationship between C-peptide and eNOS, the enzyme responsible for regulating
endothelial NO concentration, has been reported in many different types of complications
associated with diabetes, as well as in various tissues and organs, both in vivo and in vitro.[173,
174] Reports of C-peptide-stimulated NO production in the endothelium could have great
importance, especially in the pulmonary endothelium, where the flow of RBCs through the

32

pulmonary bed allows re-oxygenation of the RBCs prior to delivery of that oxygen to organs and
tissues throughout the body as described previously. In accordance with the importance of
pulmonary NO production, Sprague et al. has shown that ATP derived from RBCs stimulates NO
production in the pulmonary circulation.[175] Interestingly, C-peptide stimulates the release of
ATP from RBCs.[154] However, in order to observe increased ATP release from RBCs, C-peptide
was introduced to a metal source, such as zinc, prior to incubation with RBCs. In this construct,
C-peptide may have the ability to stimulate NO production in the endothelium indirectly by
purinergic signaling, as mediated by C-peptide stimulated release of ATP from RBCs.
Importantly, the potential involvement of RBC derived ATP in the proposed mechanism again
highlights the importance of flow in in vitro vascular experiments, as flowing RBCs under the
influence of shear stress release more ATP than static RBCs.[155, 159, 176, 177] However, in
order to thoroughly investigate these prospective roles of C-peptide, multiple cell types,
separated by physiologically relevant distances, would need to be investigated in a biologically
accurate mimic of the vasculature.
Similar to the proposed mechanism for C-peptide induced vasorelaxation, much of the
observed biological effects of C-peptide can potentially be attributed to processes occurring in
the cardiovascular system resulting indirectly from C-peptide activity within the bloodstream.
However, it is unclear if the broad influence of C-peptide is elicited entirely from activity in the
bloodstream, or if C-peptide is also capable of escaping the bloodstream to influence cell
behavior directly. Some activity stimulated by C-peptide extending beyond the vascular system
is difficult to explain in paradigms where C-peptide is confined to the bloodstream, such as the
observed improvements in diabetic neuropathy and nerve conduction. While the capability of

33

C-peptide to escape the bloodstream has yet to be investigated, nor has C-peptide ever been
detected exterior to the vasculature in vivo, some reports insinuate that C-peptide is potentially
trafficked across the endothelium to leave the bloodstream.
For example, It has been previously shown that C-peptide specifically binds to endothelial
cells, which is of importance since the transport of macromolecules is generally by transcellular
and specific mechanisms.[178] The existence of a binding structure for C-peptide is still in
question, however, since this specific binding has not yet been shown by conventional
radioligand binding and, furthermore, a specific receptor for C-peptide has not been identified.
Additionally, the homology pattern of C-peptide is markedly different from most signaling
peptides.[179] However, C-peptide has been shown to be internalized by endosomes to enter
human endothelial cells.[180] Additionally, this internalization can be inhibited by nocodazole
and monodansylcadaverine (MDC), signifying that C-peptide enters the cell by a mechanism
which is receptor mediated and requires microtubule assembly. Interestingly, microtubules
have been shown to facilitate transcellular transport and also influence endothelial
permeability and integrity.[181-184] Also, some endothelial cell types express PEPT membrane
proteins, which promiscuously facilitate the transcellular transport of peptides.[185, 186] In
addition to human endothelial cells, C-peptide can also enter fibroblasts and human embryonic
kidney cells where it is capable of influencing gene transcription,[187, 188] suggesting that Cpeptide is capable of directly influencing cells exogenous of the bloodstream and again
demonstrating the ability of C-peptide to permeate cellular membranes. Collectively, these
results indicate that C-peptide is potentially selectively trafficked across the endothelium by a
transcellular transport mechanism.

34

When considered jointly, these results indicate that C-peptide can plausibly exit the
bloodstream but do not, however, provide sufficient evidence to conclusively confirm or reject
this hypothesis. The ability or inability of C-peptide to permeate a cultured endothelium can
provide valuable insight as to whether some biological effects of the peptide are initiated
directly, or if all of these effects are result indirectly from activity in the bloodstream. However,
in order to determine the potential involvement of each of these potential mechanisms, further
investigation of the ability of C-peptide to escape the bloodstream is required. In order to more
decisively determine the aptitude for C-peptide to leave the bloodstream, an assay of
endothelial permeability to C-peptide should be accomplished, since the endothelium serves as
the primary barrier separating flowing blood from the interstitium.
Importantly, these permeability studies must account for transcellular transport, which is
the probable mode of transport of C-peptide across the endothelium in vivo. This requirement
is most adequately satisfied by permeability assays, which can be performed to quantitatively
monitor the permeation of C-peptide across a cultured endothelium. However, shear has been
shown to influence changes in endothelial permeability,[189] and therefore, shear will ideally
be incorporated in the permeability assay. More importantly, the endothelial cellular barrier
should be characterized in terms of barrier function and confluence to ensure that C-peptide is
required to permeate an endothelial monolayer that closely resembles the true endothelium. A
more conclusive and accurate deduction of the ability of C-peptide to permeate the
endothelium in vivo may be accomplished via the utilization of a biological mimic that includes
the flow capabilities required to replicate shear stress, as well as RBCs. Unfortunately, these
important biological aspects are not accounted for by the microtiter diffusion cell, which is the

35

conventional platform utilized to perform permeability assays as previously discussed.
Therefore, a contrasting platform that accounts for these biological aspects is ideal since
conclusions stemming from results obtained using this alternate biological model will likely be
significantly more accurate and credible as a result of the enhanced physiological relevance
achieved when compared to conventional microtiter platforms.

36

Figure 1.5 - The amino acid sequence of C-peptide. Subsequent to cleavage from insulin, the 31
residue peptide is capable of increasing ATP release from RBCs, an activity which is proposed to
mediate NO production in the endothelium.

37

Recently, our group has shown that if C-peptide is incubated in a solution containing certain
divalent metal ions prior to addition to a RBC solution, the resulting metal activated C-peptide
was able to facilitate glucose uptake into, and increase ATP release from RBCs.[154, 190]
Remarkably, these results could not be reproduced in the absence of a metal, indicating that Cpeptide must first be introduced to a metal source for some, if not all bioactivity. These findings
were further supported when mass spectrometric analysis of a solution containing active Cpeptide not deliberately introduced to a metal, revealed that the peptide had indeed formed an
at least one adduct with a metal.[154, 191] Although C-peptide has been shown to be active in
the presence of multiple metal ions[154, 190, 192, 193], it has been shown that zinc is present
at millimolar levels within beta cells where C-peptide is produced and cleaved from proinsulin
prior to release into the bloodstream.[194] Therefore, zinc is the most bioavailable metal
source for C-peptide prior to secretion, increasing the probability that zinc is responsible for the
bioactivity of C-peptide in vivo.
To date, most studies involving the exogenous addition of C-peptide to humans, animal
models, and cell models have not directly co-administered a metal as the influence of metal on
C-peptide bioactivity is largely under-recognized. Interestingly, some studies report that Cpeptide only elicits bioactivity when co-administered with insulin, which often contains Zn

2+

at

levels as high as 0.5% by mass.[195] Furthermore, it has been shown that C-peptide is capable
of increasing RBC membrane deformability, however these effects were abolished in the
presence of EDTA, which is known to strongly chelate metals.[196] Hypothetically, it is feasible
that C-peptide samples often unintentionally contain metals, sometimes rendering C-peptide
bioactive, which could explain the inconsistent activity observed in clinical trials. If this is the
38

case, it could provide an explanation for the surprising lack of findings regarding mechanistic
aspects of C-peptide bioactivity, despite the discovery that C-peptide is responsible for many
beneficial functions in vivo. Due to the development of a now well-established method for C2+

peptide preparation with Zn , our lab is uniquely well suited for investigating the mechanisms
of C-peptide bioactivity as indicated by our ability to reproducibly demonstrate various Cpeptide bioactivies in the presence of a metal.
1.2.4 - Current Hypothesis Regarding Vasodilation and the Influence of C-peptide
Recent findings have shown that circulating ATP impacts vascular tone by purinergic
signaling, specifically as an agonist of the endothelial P2Y receptor, which results in eNOS
mediated production of endothelium NO, a potent vasodilator.[197-200] These findings
corroborate previous work by Sprague showing that RBC-derived ATP stimulated NO production
and vasodilation in the rabbit pulmonary circulation. While it is debated whether this
mechanism is primarily responsible for the regulation of vascular tone, it is the only proposed
mechanism where NO is derived from the endothelium as opposed to the RBC. Consequential
to the short half-life of NO in vivo, endothelial derived NO can likely more efficiently stimulate
vasorelaxation than NO derived from the RBC since the diffusional distance to the smooth
muscle is reduced. Furthermore, RBC derived NO is required to permeate the barrier formed by
the endothelium before reaching the smooth muscle, a problem that is largely avoided by NO
produced by the endothelium. This assumption is supported by unpublished data from the
Spence lab suggesting that RBC derived NO is blocked from reaching the smooth muscle by the
pulmonary endothelium. In addition, the continuous production of NO most important for
regulating vascular tone has been repeatedly associated with shear stress, which is known to
39

influence circulating ATP levels by stimulating its release from the RBC. In contrast, the known
primary mechanisms proposed in the literature for NO release from the RBC are in response to
hypoxia as opposed to shear stress.
Considering these recent developments implicating ATP in the purinergic regulation of
vascular tone, factors affecting circulating levels of ATP may be fundamentally important for
maintaining healthy blood flow by association. It is well established that the RBC contains
millimolar levels of ATP, which can be released into the blood stream by several stimuli.
Interestingly, C-peptide, when carefully prepared with a metal source, has been shown to
significantly increase ATP release from the RBC. These results are exciting because in patients
with type 1 diabetes who have impaired C-peptide production, vascular ailments arise,
including hypertension, which can be ameliorated by improved blood flow resulting from Cpeptide therapy. However, the mechanism by which C-peptide improves diabetic blood flow is
unknown. Figure 1.6 displays the proposed mechanism by which C-peptide regulates vascular
tone. In this construct, C-peptide may have the ability to stimulate NO production in the
endothelium by purinergic signaling, as mediated by C-peptide stimulated release of ATP from
RBCs. As discussed earlier, endothelial derived NO has been shown to be crucial for the
regulation of vascular tone.

40

Figure 1.6 - The proposed mechanism of RBC mediated vasorelaxation. In response to a
stimulus, such as C-peptide interaction or shear stress, the RBC releases ATP, which is capable
of diffusing to the endothelium. As an agonist of the G-coupled P2 purinoreceptors, ATP binds
the P2Y purinergic receptors located on the endothelial cell membrane, which initiates a
purinergic signaling cascade. In response to ATP agonist, P2Y increases calcium flux into the cell,
2+

leading to an increase in free intracellular Ca , which can bind CaM to form the Ca-CaM
complex. In this active conformation, Ca-CaM is capable of binding the calmodulin-binding
domain of eNOS to stimulate the enzymatic conversion of L-arginine to L-Citruline and NO
(Figure 1.4). NO produced in the endothelium is able to diffuse to the smooth muscle cells to
activate soluble guanylyl cyclase (sGC), resulting in the enzymatic production of cyclic
guanosine monophosphate (cGMP) from guanosine triphosphate (GTP) and leading to
relaxation of smooth muscle, dilating the blood vessel.
41

If C-peptide can indeed stimulate endothelial NO production, C-peptide may be
revolutionary for the treatment of type 1 diabetes due to the ability to improve blood flow, but
also due to its potential ability to mediate glucose clearance from the bloodstream via
increased glucose uptake by the RBC. Potentially even more importantly, if shown to influence
endothelial NO production, C-peptide may be crucially important for maintaining healthy
vascular tone in general. Specifically, C-peptide may play an important role in maintaining
appropriate levels of circulating ATP by influencing the degree of ATP release from RBCs, since
healthy individuals typically have concentrations of C-peptide varying in the single-digit
nanomolar range circulating in the bloodstream at any given time,[201, 202]. Importantly, if Cpeptide is shown to potentially influence purinergic regulation of vascular tone, by inference, Cpeptide may be fundamentally important for maintaining healthy blood flow, which has so far
been overlooked. Of course, in order to verify this hypothesis, an in vitro model of the
vasculature will need to be developed that is capable of reproducing important aspects of the
vasculature. This model will require geometries that mimic resistance vessels and the inclusion
of multiple cell types separated by physiological diffusional distances so that biologically
relevant cell-cell communication can be monitored. Ideally, a measure of endothelial
monolayer confluence will be possible in conjunction with flow capability to reproduce shear
stress that is so influential in the vasculature.
1.3 - Microfluidics
1.3.1 - Fabrication and Background
Microfluidic technology has been largely and broadly influential in various fields of science,
leaving behind a rich history. For example, microfluidic technology famously made significant
42

contributions to the successful sequencing of the human genome.[203-207] Microfluidic
technology offers several general experimental advantages that are attractive for a vast range
of research, causing the preferential use of microfluidic devices in lieu of other well established
technology and instrumentation which, in turn, has led to significant and diverse scientific
contributions. For example, some of the most appreciated advantages offered by microfluidics
are: small sample volume, low reagent consumption, minimal waste, and high throughput.[208210] In fact, microfluidic technology is commonly called “lab on a chip” technology since often
microfluidic devices can accomplish tasks, sometimes with improved results, that usually
require whole instruments, of which microfluidic devices are a small fraction of the size. Due to
these conveniences, since the 1980s microfluidic technology has been applied in research
throughout various scientific fields, however more recently, microfluidic systems have begun to
emerge as useful platforms for cell based assays.[30, 88, 113, 211-217]
Arguably, microfluidic technology first became amenable to cell culture research when a
soft lithographic technique for the rapid prototyping of polydimethylsiloxane (PDMS)
microfluidic devices was developed in the late 1990s by the Whitesides laboratory.[218, 219]
PDMS, being a gas permeable polymer, was the first suitable material for use in microfluidic
fabrication that also allowed for gas exchange, a requirement for cell survival.[220, 221]
Additionally, the fabrication of PDMS microfluidic devices by soft lithography is generally less
complicated for the wide range of other materials often utilized for microfluidic
construction.[222] While PDMS, as a material for microfluidic fabrication, has several
disadvantages, [223] such as hydrophobicity, absorption and adsorption of substances, as well
as high flexibility, it has several advantages for cell-based research. Firstly, PDMS is gas

43

permeable, allowing cellular oxygen and carbon dioxide exchange, which is required for cell
survival. Furthermore, PDMS is a transparent material, allowing for visualization of cells, optical
analysis, and imaging.[210]
1.3.2 - Microfluidics to Mimic Biological Systems
The inherent geometries and dimensions of microfluidic systems are well suited for
mimicking in vivo physiological conditions and monitoring cell behavior within these
microenvironments. As mentioned previously, microfluidic technology offers precise control of
important aspects of the microenvironment, leading to more physiologically relevant models of
in vivo systems. For these reasons, when considering all technology available for modeling
biological systems, a microfluidic platform would serve as an accurate and physiologically
relevant model of the vasculature. For example, an in vitro cell layer mimic capable of
incorporating flow, and other biologically relevant cells, such as the RBC, would be paramount
for mimicking a true vascular barrier. However, the various cells would have to be in close
proximity to allow for cell-cell communication. These issues are well addressed with
microfluidic systems, as it is possible to mimic the small distance between the blood vessel and
the endothelium.[211] Blood flow is also easily accomplished by incorporation of channels that
have diameters similar to resistance vessels (capillaries and arterioles) found in vivo,[211] and
thereby incorporating shear stress at a biologically relevant magnitude, a crucially important
aspect since shear stress is influential to cell behavior. In accordance with the importance of
shear stress, and the ability to simulate this force in microfluidic systems, a myriad of shearrelated microfluidic experiments have been accomplished, such as studies on migration,[224]
adhesion,[30, 225, 226], enrichment,[227] and shear-induced NO production.[90, 113]
44

Not surprisingly, microfluidic technology has been used by the Spence group and others to
mimic the vasculature.[30, 88, 113, 228-232] Also, as a result of these same beneficial
properties of microfluidic systems, several microfluidic organ mimics have been accomplished,
including the lung[233, 234], kidney[235, 236], heart[237], and digestive system[238] in work
coined “organ on a chip” research. Additionally, microfluidic technology has been utilized to
model neurons.[239, 240] Interestingly, a majority of these microfluidic mimics of biological
systems feature cellular monolayers comprised of epithelial or endothelial cells, as do
numerous other microfluidic biological models, indicating their importance in vivo. As
previously discussed, it has been shown that the degree of confluence can alter the behavior of
these cells, and thus a method for the determination of monolayer barrier integrity is beneficial
in all experimental situations and necessary in some. Also, since epithelial and endothelial cells
often form barriers in vivo that often influence drug delivery, and are implicated in various
diseases, measurements of monolayer barrier integrity and permeability are the basis of a
majority of research involving these processes.
1.3.3 - Monitoring Barrier Integrity on a Microfluidic Device
Using microfluidic systems for the analysis of cellular monolayers has been contributive and
progressive towards understanding complex biological systems, although reliably measuring
cell monolayer integrity in these microfluidic devices is challenging. Cell monolayer integrity can
be measured in a number of ways that are described previously in Section 1.1. Interestingly,
these methods are used in the vast majority of experimentation regarding monolayers, except
for those performed using microfluidic systems. Subsequent to the start of the work presented
in the following chapters, microfluidic platforms have been reported that are capable of
45

determining cell monolayer integrity of various cell types by monitoring permeation,[241-243]
as well as TEER,[244-246] demonstrating the need for such devices. However, the microfluidic
device featured in chapter 2 is the first microfluidic vascular mimic capable of measuring TEER,
which is relatively simple and timely to fabricate compared to similar devices.[97] Similarly,
microfabricated cell culture chambers have been developed capable of measuring cell
monolayer confluence, however these models lack flow capabilities.[247-249] Also, additional
devices with this important capability have been reported subsequent to publication of the
device presented in Chapter 2.[250-253]
Importantly, the ability to determine monolayer barrier integrity without affecting cells
cultured within a microfluidic device, and also simultaneously eliciting shear stress generated
by flowing blood, offers the potential for dynamic experiments while accurately mimicking a
true biological system. In order to achieve this goal, TEER is a preferred method for determining
monolayer barrier robustness since it allows for repeated quantitative measuring without
altering cell phenotype or damaging the cells. In addition to the potential for time-lapse
microfluidic studies, such a device would further ensure proper cellular behavior by confirming
the confluence of cultured cells, adding credibility and accuracy to results. For these reasons,
the unprecedented development of a microfluidic vascular mimic with integrated TEER
capabilities could be significantly useful for elucidating mechanisms of important physiological
events and functions of the vascular system, providing cell-cell interactions are permitted.
1.3.4 - Cell-Cell Communication in Microfluidic Systems
A substantial advantage of microfluidic systems yet unparalleled by other technologies is
the ability to incorporate varying cell types separated by physiologically relevant distances,
46

which are often exceedingly diminutive and therefore challenging to reproduce with alternative
technology.[254] Furthermore, microfluidics allows for cell culture, in addition to the flow of
cells, reproducing the shear forces experienced by cells in vivo in a manner not easily
accomplished by other techniques. In the cardiovascular system, it has been shown that shear
stress is necessary for certain cellular responses and impacts cellular behavior, including that of
the endothelial cell and RBC, and it is one mode by which these cells interact. By employing
microfluidics, the Spence group has shown repeatedly that for cell types found in the
vasculature, various stimuli, such as shear force, are capable of eliciting a response from one
cell type that, in turn, influences the behavior of a second.
For example, it was shown that in addition to C-peptide, shear stress can cause an increase
in ATP release from RBCs,[150, 155] which can subsequently stimulate platelet NO
production,[90] as well as endothelial NO production.[89, 97, 113, 211, 255] Interestingly, NO is
capable of decreasing ATP release from RBCs by a potential feedback mechanism of the
endothelial cell and platelet in response to ATP release from the RBC.[256] Additionally, the
effects of an anti-adhesion drug were determined by monitoring platelet adhesion to a cultured
endothelium under flow conditions.[30] These examples of monitoring cell-cell interactions to
elucidate mechanisms of biological events by the Spence group adds to an increasing amount of
literature highlighting the utility of microfluidic technology for mimicking biological systems and
monitoring cell-cell communication.[257-259] Due to the architectural advantages and flow
capabilities of microfluidic systems, a microfluidic device capable of characterizing monolayer
barrier integrity can feasibly be developed to elucidate the mechanism by which C-peptide and
shear play a role in influencing vascular tone. The displayed ability of microfluidic devices to

47

mimic biological systems in a manner accurate enough to monitor cell-cell interactions
evidences the plausibility of this endeavor.
1.4 - Goals and Motivation
Motivated by the associated impact of elucidating the mechanism by which C-peptide
influences vascular tone, specifically by the potential ability to stimulate NO production in the
endothelium, the development of a vascular mimic will be attempted in order to investigate the
potential involvement of C-peptide in this fundamentally important biological process. Also, the
ability of C-peptide to escape the bloodstream is potentially of high biological importance since
the direct involvement of C-peptide in biological processes is possible, not only in the vascular
system, but throughout the entire body. For example, the biological influence of C-peptide may
be restricted to effects that are accomplished indirectly as a result of activity within the
bloodstream, or C-peptide may be able to both indirectly and directly stimulate cell types
exogenous of the vasculature, provided C-peptide can permeate the vascular wall to physically
interact these cells. In the latter case, the scope of the physiological influence of C-peptide may
be extensive and diverse, similar to insulin, but also severely under-recognized by the scientific
community.
In order to investigate these important possibilities, many important biological aspects must
be considered when developing a model of the vasculature in order to achieve meaningful and
biologically relevant results. For example, due to the involvement of multiple cell types in the
proposed mechanism for C-peptide induced vasorelaxation, specifically the RBC and endothelial
cell, a vascular model including these multiple cell types is required. Furthermore, since it is
proposed that ATP released from the RBC acts as an extracellular signaling molecule that
48

directly interacts with endothelial cells to stimulate NO production in the endothelium, the RBC
and the endothelial cells must be separated by physiologically relevant distances in the vascular
mimic in order to accurately replicate the diffusional distance to the endothelium experienced
by ATP released from circulating RBCs in vivo. In addition to the release of signaling molecules,
endothelial cells and RBCs both respond to shear stress generated from the drag force
produced by the direct contact of flowing blood with the stationary endothelium. Importantly,
shear stress greatly impacts each cell type in vivo, and also directly influences several
fundamental aspects of the processes under investigation. For example, shear stress has been
shown to impact ATP release from the RBC, and also greatly increase endothelial barrier
integrity. Also, since endothelial cells comprising the endothelium form a confluent barrier in
vivo, and endothelial behavior and response to environmental queues has been shown to vary
with monolayer barrier integrity and confluence and, therefore, the properties of endothelial
cells cultured within the vascular mimic should ideally replicate the true endothelium by
forming a completely confluent monolayer that exhibits proper barrier integrity. This criterion is
important for the successful observation and elucidation of biologically relevant cell-cell
communication mechanisms, since physiologically correct cellular responses to stimuli are often
dependent on monolayer barrier integrity and confluence. Additionally, the most important
factor determining the permeability of cellular monolayers to molecules is the integrity of the
barrier formed by the monolayer. Therefore, in order to accurately determine ability of a
molecule to permeate a cellular monolayer, conditions found in vivo should be replicated and
confirmed prior to experimentation by first measuring the barrier integrity and confluence of
cultured cells.

49

To successfully address each of these important aspects of biological systems observed in
vivo, a microfluidic vascular mimic will be developed to determine a potential mechanism of Cpeptide induced endothelial NO production, as well as the endothelial permeability to Cpeptide. While microfluidics have recently proven useful for cell based assays, a vascular mimic
with integrated flow capabilities, as well as cell layer confluence measurement capabilities, has
not yet been developed. The design of the microfluidic vascular mimic will ideally allow for the
incorporation of each of the important biological aspects that are often not addressed by
conventional microtiter biological models, such as: multiple cell types, physiologically relevant
architecture, and shear stress. Additionally, the ability to confirm cell monolayer confluence,
which is rarely accomplished in microfluidic platforms, and often easily accomplished in
microtiter platforms, will also ideally be integrated in the microfluidic vascular mimic. The
development of such a vascular mimic can potentially be utilized to reproduce and decipher
highly complicated biological phenomena, including those involving cell-cell communications,
with unparalleled accuracy. Therefore, such a device may be capable of replicating conditions
found in vivo to such an extent that the potential mechanism by which C-peptide stimulates NO
production in the endothelium may be elucidated. Furthermore, such a device could be utilized
to investigate the ability of C-peptide to permeate the endothelium in vivo with a high degree
of biological relevance in comparison to conventional techniques.

50

REFERENCES

51

REFERENCES
1.

Atencia, J. and D.J. Beebe, Controlled microfluidic interfaces. Nature, 2005. 437(7059): p.
648-55.

2.

Kuczenski, B., et al., Probing cellular dynamics with a chemical signal generator. PLoS
ONE, 2009. 4(3): p. e4847.

3.

Faley, S., et al., Microfluidic platform for real-time signaling analysis of multiple single T
cells in parallel. Lab Chip, 2008. 8(10): p. 1700-12.

4.

Vickerman, V., et al., Design, fabrication and implementation of a novel multi-parameter
control microfluidic platform for three-dimensional cell culture and real-time imaging.
Lab Chip, 2008. 8(9): p. 1468-77.

5.

Walker, G.M., H.C. Zeringue, and D.J. Beebe, Microenvironment design considerations
for cellular scale studies. Lab Chip, 2004. 4(2): p. 91-7.

6.

Huh, D., et al., Acoustically detectable cellular-level lung injury induced by fluid
mechanical stresses in microfluidic airway systems. Proc Natl Acad Sci U S A, 2007.
104(48): p. 18886-91.

7.

Lao, A.I.K., et al., Precise temperature control of microfluidic chamber for gas and liquid
phase reactions. Sensors and Actuators a-Physical, 2000. 84(1-2): p. 11-17.

8.

Guijt, R.M., et al., Chemical and physical processes for integrated temperature control in
microfluidic devices. Lab on a Chip, 2003. 3(1): p. 1-4.

9.

Roper, M.G., et al., Infrared temperature control system for a completely noncontact
polymerase chain reaction in microfluidic chips. Anal Chem, 2007. 79(4): p. 1294-300.

10.

Bhatia, S.N., et al., Effect of cell-cell interactions in preservation of cellular phenotype:
cocultivation of hepatocytes and nonparenchymal cells. FASEB J, 1999. 13(14): p. 1883900.

11.

Ayliffe, H.E., A.B. Frazier, and R.D. Rabbitt, Electric impedance spectroscopy using
microchannels with integrated metal electrodes. Journal of Microelectromechanical
Systems, 1999. 8(1): p. 50-57.

12.

Gumbleton, M. and K.L. Audus, Progress and Limitations in the Use of In Vitro Cell
Cultures to Serve as a Permeability Screen for the Blood-Brain Barrier. J. of Pharm. Sci.,
2001. 90(11): p. 1681-1698.

13.

Crone, C. and S.P. Olesen, Electrical resistance of brain microvascular endothelium. Brain
Res FIELD Full Journal Title:Brain research, 1982. 241(1): p. 49-55.

52

14.

Wilcox, E.R., et al., Mutations in the gene encoding tight junction claudin-14 cause
autosomal recessive deafness DFNB29. Cell, 2001. 104(1): p. 165-72.

15.

Gow, A., et al., Deafness in Claudin 11-null mice reveals the critical contribution of basal
cell tight junctions to stria vascularis function. J Neurosci, 2004. 24(32): p. 7051-62.

16.

Riazuddin, S., et al., Mutations in TRIOBP, which encodes a putative cytoskeletalorganizing protein, are associated with nonsyndromic recessive deafness. Am J Hum
Genet, 2006. 78(1): p. 137-43.

17.

Xavier, R.J. and D.K. Podolsky, Unravelling the pathogenesis of inflammatory bowel
disease. Nature, 2007. 448(7152): p. 427-34.

18.

Chronopoulos, A., et al., High glucose-induced altered basement membrane composition
and structure increases trans-endothelial permeability: implications for diabetic
retinopathy. Curr Eye Res, 2011. 36(8): p. 747-53.

19.

Sward, P. and B. Rippe, Acute and sustained actions of hyperglycaemia on endothelial
and glomerular barrier permeability. Acta Physiol (Oxf), 2012. 204(3): p. 294-307.

20.

D'Aversa, T.G., et al., Myelin Basic Protein Induces Inflammatory Mediators from Primary
Human Endothelial Cells and Blood-Brain-Barrier Disruption: Implications for the
Pathogenesis of Multiple Sclerosis. Neuropathol Appl Neurobiol, 2012.

21.

Alvarez, J.I., R. Cayrol, and A. Prat, Disruption of central nervous system barriers in
multiple sclerosis. Biochim Biophys Acta, 2011. 1812(2): p. 252-64.

22.

Perrier, C. and B. Corthesy, Gut permeability and food allergies. Clin Exp Allergy, 2011.
41(1): p. 20-8.

23.

Pushparaj, P.N., et al., The cytokine interleukin-33 mediates anaphylactic shock.
Proceedings of the National Academy of Sciences of the United States of America, 2009.
106(24): p. 9773-9778.

24.

Ignarro, L.J., G. Buga, and G. Dhaudhuri, EDRF generation and release from perfused
bovine pulmonary artery and vein. Eur. J. Pharmacol, 1988. 149: p. 79-88.

25.

Ignarro, L.J., et al., Endothelium-derived relaxing factor from pulmonary artery and vein
possesses pharmacologic and chemical properties identical to those of nitric oxide
radical. Circulation Research, 1987. 61(6): p. 866-79.

26.

Rees, D.D., R.M. Palmer, and S. Moncada, Role of endothelium-derived nitric oxide in the
regulation of blood pressure. Proc Natl Acad Sci U S A, 1989. 86(9): p. 3375-8.

27.

Eremina, V., et al., Glomerular-specific alterations of VEGF-A expression lead to distinct
congenital and acquired renal diseases. J Clin Invest, 2003. 111(5): p. 707-16.
53

28.

Tolins, J.P., et al., Role of endothelium-derived relaxing factor in regulation of renal
hemodynamic responses. Am J Physiol, 1990. 258(3 Pt 2): p. H655-62.

29.

Witko-Sarsat, V., et al., Neutrophils: molecules, functions and pathophysiological
aspects. Laboratory Investigation, 2000. 80(5): p. 617-53.

30.

Ku, C.J., T. D'Amico Oblak, and D.M. Spence, Interactions between multiple cell types in
parallel microfluidic channels: monitoring platelet adhesion to an endothelium in the
presence of an anti-adhesion drug. Anal Chem, 2008. 80(19): p. 7543-8.

31.

Becker, B.F., et al., Endothelial function and hemostasis. Z Kardiol, 2000. 89(3): p. 160-7.

32.

Kim, J.A., et al., Brain endothelial hemostasis regulation by pericytes. Journal of Cerebral
Blood Flow and Metabolism, 2006. 26(2): p. 209-217.

33.

Aird, W.C., The role of the endothelium in severe sepsis and multiple organ dysfunction
syndrome. Blood, 2003. 101(10): p. 3765-3777.

34.

Amar, S., et al., Periodontal disease is associated with brachial artery endothelial
dysfunction and systemic inflammation. Arterioscler Thromb Vasc Biol, 2003. 23(7): p.
1245-9.

35.

Tuomanen, E., Entry of pathogens into the central nervous system. FEMS Microbiol Rev,
1996. 18(4): p. 289-99.

36.

Luttun, A. and P. Carmeliet, De novo vasculogenesis in the heart. Cardiovasc Res, 2003.
58(2): p. 378-89.

37.

Hoekstra, M., et al., Specific gene expression of ATP-binding cassette transporters and
nuclear hormone receptors in rat liver parenchymal, endothelial, and Kupffer cells. The
Journal of biological chemistry, 2003. 278(28): p. 25448-53.

38.

van der Meer, A.D., et al., Microfluidic technology in vascular research. J Biomed
Biotechnol, 2009. 2009: p. 823148.

39.

Wong, K.H., et al., Microfluidic Models of Vascular Functions. Annu Rev Biomed Eng,
2012.

40.

Mehta, D. and A.B. Malik, Signaling mechanisms regulating endothelial permeability.
Physiological Reviews, 2006. 86(1): p. 279-367.

41.

Gumbiner, B.M., Breaking through the Tight Junction Barrier. Journal of Cell Biology,
1993. 123(6): p. 1631-1633.

42.

Gonzalez-Mariscal, L., et al., Tight junction proteins. Progress in Biophysics & Molecular
Biology, 2003. 81(1): p. 1-44.
54

43.

Forster, C., Tight junctions and the modulation of barrier function in disease. Histochem
Cell Biol, 2008. 130(1): p. 55-70.

44.

Ebnet, K., Organization of multiprotein complexes at cell-cell junctions. Histochem Cell
Biol, 2008. 130(1): p. 1-20.

45.

Staehelin, L.A., Further observations on the fine structure of freeze-cleaved tight
junctions. J Cell Sci, 1973. 13(3): p. 763-86.

46.

Dejana, E., et al., Organization and signaling of endothelial cell-to-cell junctions in
various regions of the blood and lymphatic vascular trees. Cell Tissue Res, 2009. 335(1):
p. 17-25.

47.

Alcaide, P., et al., p120-Catenin regulates leukocyte transmigration through an effect on
VE-cadherin phosphorylation. Blood, 2008. 112(7): p. 2770-9.

48.

Curl, C.L., et al., Quantitative phase microscopy: a new tool for investigating the
structure and function of unstained live cells. Clin Exp Pharmacol Physiol, 2004. 31(12):
p. 896-901.

49.

Kobayashi, H., et al., Angiocrine factors from Akt-activated endothelial cells balance selfrenewal and differentiation of haematopoietic stem cells. Nature Cell Biology, 2010.
12(11): p. 1046-U32.

50.

Furuse, M., et al., OCCLUDIN - A NOVEL INTEGRAL MEMBRANE-PROTEIN LOCALIZING AT
TIGHT JUNCTIONS. Journal of Cell Biology, 1993. 123(6): p. 1777-1788.

51.

Furuse, M., et al., A single gene product, claudin-1 or -2, reconstitutes tight junction
strands and recruits occludin in fibroblasts. Journal of Cell Biology, 1998. 143(2): p. 391401.

52.

Chen, Y.H., et al., Restoration of tight junction structure and barrier function by downregulation of the mitogen-activated protein kinase pathway in Ras-transformed MadinDarby canine kidney cells. Molecular Biology of the Cell, 2000. 11(3): p. 849-862.

53.

Wang, W., W.L. Dentler, and R.T. Borchardt, VEGF increases BMEC monolayer
permeability by affecting occludin expression and tight junction assembly. American
Journal of Physiology-Heart and Circulatory Physiology, 2001. 280(1): p. H434-H440.

54.

Harhaj, N.S. and D.A. Antonetti, Regulation of tight junctions and loss of barrier function
in pathophysiology. international journal of biochemistry & cell biology, 2004. 36(7): p.
1206-1237.

55.

Bolton, S.J., D.C. Anthony, and V.H. Perry, Loss of the tight junction proteins occludin and
zonula occludens-1 from cerebral vascular endothelium during neutrophil-induced bloodbrain barrier breakdown in vivo. Neuroscience, 1998. 86(4): p. 1245-1257.
55

56.

Hirase, T., et al., Occludin as a possible determinant of tight junction permeability in
endothelial cells. Journal of Cell Science, 1997. 110: p. 1603-1613.

57.

Chen, C.C., Y. Zhou, and L.A. Baker, Scanning ion conductance microscopy. Annu Rev
Anal Chem (Palo Alto Calif), 2012. 5(1): p. 207-28.

58.

Morris, C.A., A.K. Friedman, and L.A. Baker, Applications of nanopipettes in the analytical
sciences. Analyst, 2010. 135(9): p. 2190-202.

59.

Deli, M.A., et al., Permeability studies on in vitro blood-brain barrier models: physiology,
pathology, and pharmacology. Cell Mol Neurobiol, 2005. 25(1): p. 59-127.

60.

Borchardt, R.T., et al., Models for Assessing Drug Absorption and Metabolism. [In:
Pharm. Biotechnol., 1996; 8]1996. 444 pp.

61.

Dehouck, M.P., et al., Drug transfer across the blood-brain barrier: correlation between
in vitro and in vivo models. Journal of Neurochemistry, 1992. 58(5): p. 1790-7.

62.

Dehouck, M.P., et al., In vitro reconstituted blood-brain barrier. Journal of Controlled
Release, 1992. 21(1-3): p. 81-91.

63.

Audus, K.L. and R.T. Borchardt, Characterization of an in vitro blood-brain barrier model
system for studying drug transport and metabolism. Pharmaceutical Research, 1986.
3(2): p. 81-7.

64.

Deli, M.A., et al., Permeability studies on in vitro blood-brain barrier models: Physiology,
pathology, and pharmacology. Cellular and Molecular Neurobiology, 2005. 25(1): p. 59127.

65.

Gaillard, P.J. and A.G. de Boer, Relationship between permeability status of the bloodbrain barrier and in vitro permeability coefficient of a drug. European Journal of
Pharmaceutical Sciences, 2000. 12(2): p. 95-102.

66.

Bowman, P.D., et al., Brain microvessel endothelial cells in tissue culture: a model for
study of blood-brain barrier permeability. Annals of Neurology, 1983. 14(4): p. 396-402.

67.

Plateel, M., E. Teissier, and R. Cecchelli, Hypoxia dramatically increases the nonspecific
transport of blood-borne proteins to the brain. Journal of Neurochemistry, 1997. 68(2):
p. 874-877.

68.

Audus, K.L., et al., Brain microvessel endothelial cell culture systems. Pharm Biotechnol,
1996. 8: p. 239-58.

69.

Brown, R.C., et al., Protection against hypoxia-induced increase in blood-brain barrier
permeability: role of tight junction proteins and NFkappaB. J Cell Sci, 2003. 116(Pt 4): p.
693-700.
56

70.

Qaum, T., et al., VEGF-initiated blood-retinal barrier breakdown in early diabetes. Invest
Ophthalmol Vis Sci, 2001. 42(10): p. 2408-13.

71.

Butt, A.M., H.C. Jones, and N.J. Abbott, Electrical resistance across the blood-brain
barrier in anaesthetized rats: a developmental study. The Journal of physiology, 1990.
429: p. 47-62.

72.

Madara, J.L., Regulation of the movement of solutes across tight junctions. Annual
Review of Physiology, 1998. 60: p. 143-159.

73.

Wong, V. and B.M. Gumbiner, A synthetic peptide corresponding to the extracellular
domain of occludin perturbs the tight junction permeability barrier. Journal of Cell
Biology, 1997. 136(2): p. 399-409.

74.

Madsen, K.L., et al., Interleukin 10 prevents cytokine-induced disruption of T84
monolayer barrier integrity and limits chloride secretion. Gastroenterology, 1997.
113(1): p. 151-9.

75.

Garcia, J.G., et al., Sphingosine 1-phosphate promotes endothelial cell barrier integrity by
Edg-dependent cytoskeletal rearrangement. J Clin Invest, 2001. 108(5): p. 689-701.

76.

Birukova, A.A., et al., Afadin controls p120-catenin-ZO-1 interactions leading to
endothelial barrier enhancement by oxidized phospholipids. J Cell Physiol, 2012. 227(5):
p. 1883-90.

77.

Hurst, R.D., et al., Nitric-oxide-induced inhibition of glyceraldehyde-3-phosphate
dehydrogenase may mediate reduced endothelial cell monolayer integrity in an in vitro
model blood-brain barrier. Brain Research, 2001. 894(2): p. 181-8.

78.

Imaizumi, S., et al., The influence of oxygen free radicals on the permeability of the
monolayer of cultured brain endothelial cells. Neurochem Int, 1996. 29(2): p. 205-11.

79.

Xu, M., et al., Ethanol disrupts vascular endothelial barrier: implication in cancer
metastasis. Toxicol Sci, 2012. 127(1): p. 42-53.

80.

Hart, M.N., et al., Differential opening of the brain endothelial barrier following
neutralization of the endothelial luminal anionic charge in vitro. Journal of
Neuropathology and Experimental Neurology, 1987. 46(2): p. 141-53.

81.

Rutten, M.J., R.L. Hoover, and M.J. Karnovsky, Electrical resistance and macromolecular
permeability of brain endothelial monolayer cultures. Brain Research, 1987. 425(2): p.
301-10.

82.

Davies, P.F., Flow-mediated endothelial mechanotransduction. Physiol Rev, 1995. 75(3):
p. 519-60.

57

83.

Nagaraja, S., A. Kapela, and N.M. Tsoukias, Intercellular communication in the vascular
wall: A modeling perspective. Microcirculation, 2012.

84.

Sprague, R.S., et al., ATP: the red blood cell link to NO and local control of the pulmonary
circulation. Am. J. Physiol, 1996. 271: p. H2717-H2722.

85.

Ellsworth, M.L., et al., Erythrocytes: Oxygen Sensors and Modulators of Vascular Tone.
Physiology, 2009. 24(2): p. 107-116.

86.

Sprague, R.S., Olearczyk, Jeffrey J., Spence, Dana M., Stephenson, Alan H., Sprung,
Robert W., Lonigro, Andrew J., Extracellular ATP signaling in the rabbit lung:
erythrocytes as determinants of vascular resistance. American Journal of PhysiologyHeart and Circulatory Physiology, 2003. 285: p. H693-H700.

87.

Dietrich, H.H., et al., Red blood cell regulation of microvascular tone through adenosine
triphosphate. Am. J. Physiol, 2000. 278: p. H1294-H1298.

88.

Tolan, N.V., et al., Personalized Metabolic Assessment of Erythrocytes Using Microfluidic
Delivery to an Array of Luminescent Wells. Analytical Chemistry, 2009. 81(8): p. 31023108.

89.

Halpin, S.T., et al., The Red Blood Cell and Nitric Oxide: Derived, Stimulated, or Both? The
Open Nitric Oxide Journal, 2011. 3(Suppl 1-M2): p. 8-15.

90.

Carroll, J.S., et al., Red blood cell stimulation of platelet nitric oxide production indicated
by quantitative monitoring of the communication between cells in the bloodstream. Anal
Chem, 2007. 79(14): p. 5133-8.

91.

Masszi, A., et al., Integrity of cell-cell contacts is a critical regulator of TGF-beta 1induced epithelial-to-myofibroblast transition: role for beta-catenin. Am J Pathol, 2004.
165(6): p. 1955-67.

92.

Pentland, A.P., et al., Cellular confluence determines injury-induced prostaglandin E2
synthesis by human keratinocyte cultures. Biochim Biophys Acta, 1987. 919(1): p. 71-8.

93.

Fleisher, L.N., et al., Stimulation of arterial endothelial cell prostacyclin synthesis by high
density lipoproteins. The Journal of biological chemistry, 1982. 257(12): p. 6653-5.

94.

Drexler, H.C., W. Risau, and M.A. Konerding, Inhibition of proteasome function induces
programmed cell death in proliferating endothelial cells. FASEB J, 2000. 14(1): p. 65-77.

95.

Yalcin, H.C., S.F. Perry, and S.N. Ghadiali, Influence of airway diameter and cell
confluence on epithelial cell injury in an in vitro model of airway reopening. Journal of
applied physiology (Bethesda, Md. 1985), 2007. 103(5): p. 1796-807.

58

96.

Govers, R., et al., Endothelial nitric oxide synthase activity is linked to its presence at cellcell contacts. The Biochemical journal, 2002. 361(Pt 2): p. 193-201.

97.

Vogel, P.A., et al., Microfluidic transendothelial electrical resistance measurement device
that enables blood flow and postgrowth experiments. Anal Chem, 2011. 83(11): p. 4296301.

98.

Corvera, S., C. DiBonaventura, and H.S. Shpetner, Cell confluence-dependent remodeling
of endothelial membranes mediated by cholesterol. The Journal of biological chemistry,
2000. 275(40): p. 31414-21.

99.

Vinals, F. and J. Pouyssegur, Confluence of vascular endothelial cells induces cell cycle
exit by inhibiting p42/p44 mitogen-activated protein kinase activity. Mol Cell Biol, 1999.
19(4): p. 2763-72.

100.

Lampugnani, M.G., et al., Cell confluence regulates tyrosine phosphorylation of adherens
junction components in endothelial cells. J Cell Sci, 1997. 110 ( Pt 17): p. 2065-77.

101.

Pries, A.R. and T.W. Secomb, Microvascular blood viscosity in vivo and the endothelial
surface layer. American journal of physiology. Heart and circulatory physiology, 2005.
289(6): p. H2657-64.

102.

Feletou, M., Y. Huang, and P.M. Vanhoutte, Endothelium-mediated control of vascular
tone: COX-1 and COX-2 products. Br J Pharmacol, 2011. 164(3): p. 894-912.

103.

Furchgott, R.F. and J.V. Zawadzki, The Obligatory Role of Endothelial-Cells in the
Relaxation of Arterial Smooth-Muscle by Acetylcholine. Nature, 1980. 288(5789): p. 373376.

104.

Ignarro, L.J., et al., Association between cyclic GMP accumulation and acetylcholineelicited relaxation of bovine intrapulmonary artery. J Pharmacol Exp Ther, 1984. 228(3):
p. 682-90.

105.

Rapoport, R.M. and F. Murad, Agonist-induced endothelium-dependent relaxation in rat
thoracic aorta may be mediated through cGMP. Circ Res, 1983. 52(3): p. 352-7.

106.

Palmer, R.M., A.G. Ferrige, and S. Moncada, Nitric oxide release accounts for the
biological activity of endothelium-derived relaxing factor. Nature, 1987. 327(6122): p.
524-6.

107.

Palmer, R.M., et al., L-arginine is the physiological precursor for the formation of nitric
oxide in endothelium-dependent relaxation. Biochem Biophys Res Commun, 1988.
153(3): p. 1251-6.

108.

Palmer, R.M., D.S. Ashton, and S. Moncada, Vascular endothelial cells synthesize nitric
oxide from L-arginine. Nature, 1988. 333(6174): p. 664-6.
59

109.

Pollock, J.S., et al., Purification and characterization of particulate endothelium-derived
relaxing factor synthase from cultured and native bovine aortic endothelial cells. Proc
Natl Acad Sci U S A, 1991. 88(23): p. 10480-4.

110.

Forstermann, U., et al., Calmodulin-dependent endothelium-derived relaxing
factor/nitric oxide synthase activity is present in the particulate and cytosolic fractions of
bovine aortic endothelial cells. Proc Natl Acad Sci U S A, 1991. 88(5): p. 1788-92.

111.

Chataigneau, T., et al., Acetylcholine-induced relaxation in blood vessels from endothelial
nitric oxide synthase knockout mice. Br J Pharmacol, 1999. 126(1): p. 219-26.

112.

Osanai, T., et al., Cross talk of shear-induced production of prostacyclin and nitric oxide
in endothelial cells. Am J Physiol Heart Circ Physiol, 2000. 278(1): p. H233-8.

113.

D'Amico Oblak, T., P. Root, and D.M. Spence, Fluorescence monitoring of ATPstimulated, endothelium-derived nitric oxide production in channels of a
poly(dimethylsiloxane)-based microfluidic device. Anal Chem, 2006. 78(9): p. 3193-7.

114.

Chen, Z., et al., Estrogen receptor alpha mediates the nongenomic activation of
endothelial nitric oxide synthase by estrogen. J Clin Invest, 1999. 103(3): p. 401-6.

115.

Fleming, I. and R. Busse, Molecular mechanisms involved in the regulation of the
endothelial nitric oxide synthase. Am J Physiol Regul Integr Comp Physiol, 2003. 284(1):
p. R1-12.

116.

Busse, R. and A. Mulsch, Calcium-dependent nitric oxide synthesis in endothelial cytosol
is mediated by calmodulin. FEBS Lett, 1990. 265(1-2): p. 133-6.

117.

Fisslthaler, B., et al., Phosphorylation and activation of the endothelial nitric oxide
synthase by fluid shear stress. Acta Physiol Scand, 2000. 168(1): p. 81-8.

118.

Olsson, R.A. and J.D. Pearson, Cardiovascular purinoceptors. Physiol Rev, 1990. 70(3): p.
761-845.

119.

Raqeeb, A., et al., Purinergic P2Y2 receptors mediate rapid Ca(2+) mobilization,
membrane hyperpolarization and nitric oxide production in human vascular endothelial
cells. Cell Calcium, 2011. 49(4): p. 240-8.

120.

Abbracchio, M.P. and G. Burnstock, Purinergic signalling: pathophysiological roles. Jpn J
Pharmacol, 1998. 78(2): p. 113-45.

121.

Burnstock, G., Purinergic signalling--an overview. Novartis Found Symp, 2006. 276: p.
26-48; discussion 48-57, 275-81.

122.

Burnstock, G., Purinergic regulation of vascular tone and remodelling. Auton Autacoid
Pharmacol, 2009. 29(3): p. 63-72.
60

123.

Saenz de Tejada, I., et al., Cholinergic neurotransmission in human corpus cavernosum. I.
Responses of isolated tissue. Am J Physiol, 1988. 254(3 Pt 2): p. H459-67.

124.

Kimoto, Y., R. Kessler, and C.E. Constantinou, Endothelium dependent relaxation of
human corpus cavernosum by bradykinin. J Urol, 1990. 144(4): p. 1015-7.

125.

Kim, N., et al., A nitric oxide-like factor mediates nonadrenergic-noncholinergic
neurogenic relaxation of penile corpus cavernosum smooth muscle. J Clin Invest, 1991.
88(1): p. 112-8.

126.

Azadzoi, K.M. and I. Saenz de Tejada, Diabetes mellitus impairs neurogenic and
endothelium-dependent relaxation of rabbit corpus cavernosum smooth muscle. J Urol,
1992. 148(5): p. 1587-91.

127.

Burnett, A.L., et al., Nitric oxide: a physiologic mediator of penile erection. Science, 1992.
257(5068): p. 401-3.

128.

Trigo-Rocha, F., et al., Nitric oxide and cGMP: mediators of pelvic nerve-stimulated
erection in dogs. Am J Physiol, 1993. 264(2 Pt 2): p. H419-22.

129.

Saiag, B., et al., Study of the mechanisms involved in adenosine-5'-O-(2-thiodiphosphate)
induced relaxation of rat thoracic aorta and pancreatic vascular bed. Br J Pharmacol,
1996. 118(3): p. 804-10.

130.

Shalev, M., et al., Stimulation of P2y purinoceptors induces, via nitric oxide production,
endothelium-dependent relaxation of human isolated corpus cavernosum. J Urol, 1999.
161(3): p. 955-9.

131.

Needham, L., et al., Characteristics of the P2 purinoceptor that mediates prostacyclin
production by pig aortic endothelial cells. Eur J Pharmacol, 1987. 134(2): p. 199-209.

132.

Mitchell, J.A., et al., Different patterns of release of endothelium-derived relaxing factor
and prostacyclin. Br J Pharmacol, 1992. 105(2): p. 485-9.

133.

Kalinowski, L., et al., Third-generation beta-blockers stimulate nitric oxide release from
endothelial cells through ATP efflux: a novel mechanism for antihypertensive action.
Circulation, 2003. 107(21): p. 2747-52.

134.

da Silva, C.G., et al., Mechanism of purinergic activation of endothelial nitric oxide
synthase in endothelial cells. Circulation, 2009. 119(6): p. 871-9.

135.

Xu, H.L., et al., Agonist-specific differences in mechanisms mediating eNOS-dependent
pial arteriolar dilation in rats. American journal of physiology. Heart and circulatory
physiology, 2002. 282(1): p. H237-43.

61

136.

Mortensen, S.P., et al., ATP-induced vasodilation and purinergic receptors in the human
leg: roles of nitric oxide, prostaglandins, and adenosine. Am J Physiol Regul Integr Comp
Physiol, 2009. 296(4): p. R1140-8.

137.

Erlinge, D. and G. Burnstock, P2 receptors in cardiovascular regulation and disease.
Purinergic Signal, 2008. 4(1): p. 1-20.

138.

Griffith, T.M., et al., The nature of endothelium-derived vascular relaxant factor. Nature,
1984. 308(5960): p. 645-7.

139.

Lincoln, T.M., N. Dey, and H. Sellak, Invited review: cGMP-dependent protein kinase
signaling mechanisms in smooth muscle: from the regulation of tone to gene expression.
Journal of applied physiology (Bethesda, Md. 1985), 2001. 91(3): p. 1421-30.

140.

Nossaman, B., E. Pankey, and P. Kadowitz, Stimulators and activators of soluble
guanylate cyclase: review and potential therapeutic indications. Crit Care Res Pract,
2012. 2012: p. 290805.

141.

Vucetic, M., P.K. Jensen, and E.C. Jansen, Diameter variations of retinal blood vessels
during and after treatment with hyperbaric oxygen. Br J Ophthalmol, 2004. 88(6): p.
771-5.

142.

Lampe, J. and K. Pommerening, [Oxygen binding of hemoglobin following covalent
fixation in the deoxy- and oxy- conformation]. Acta Biol Med Ger, 1975. 34?710: p. 16038.

143.

Stamler, J.S., et al., Blood flow regulation by S-nitrosohemoglobin in the physiological
oxygen gradient. Science, 1997. 276(5321): p. 2034-2037.

144.

Jia, L., et al., S-Nitrosohemoglobin: a dynamic activity of blood involved in vascular
control. Nature (London), 1996. 380(6571): p. 221-6.

145.

Doyle, M.P., D.M. LePoire, and R.A. Pickering, Oxidation of hemoglobin and myoglobin
by alkyl nitrites inhibition by oxygen. The Journal of biological chemistry, 1981. 256(23):
p. 12399-404.

146.

Cosby, K., et al., Nitrite reduction to nitric oxide by deoxyhemoglobin vasodilates the
human circulation. Nat Med, 2003. 9(12): p. 1498-505.

147.

Garcia, J.I., et al., Nitrite and nitroglycerin induce rapid release of the vasodilator ATP
from erythrocytes: Relevance to the chemical physiology of local vasodilation. J Inorg
Biochem, 2010. 104(3): p. 289-96.

148.

Aisaka, K., et al., NG-methylarginine, an inhibitor of endothelium-derived nitric oxide
synthesis, is a potent pressor agent in the guinea pig: does nitric oxide regulate blood
pressure in vivo? Biochem Biophys Res Commun, 1989. 160(2): p. 881-6.
62

149.

Beutler, E., Biochemistry of the erythrocyte. Experientia, 1969. 25(5): p. 470-3.

150.

Edwards, J.L., et al., Chemiluminescence detection of ATP release from red blood cells
upon deformation in microbore tubing. Analyst, 2001. 126: p. 1257-1260.

151.

Faris, A. and D.M. Spence, Measuring the simultaneous effects of hypoxia and
deformation on ATP release from erythrocytes. Analyst (Cambridge, United Kingdom),
2008. 133(5): p. 678-682.

152.

Bergfeld, G.R. and T. Forrester, Release of ATP from human erythrocytes in response to a
brief period of hypoxia and hypercapnea. Cardiovasc. Res., 1992. 26: p. 40-47.

153.

Raththagala, M., et al., Hydroxyurea stimulates the release of ATP from rabbit
erythrocytes through an increase in calcium and nitric oxide production. Eur J Pharmacol,
2010. 645(1-3): p. 32-8.

154.

Meyer, J.A., et al., Metal-activated C-peptide facilitates glucose clearance and the
release of a nitric oxide stimulus via the GLUT1 transporter. Diabetologia, 2008. 51(1): p.
175-82.

155.

Sprague, R.S., et al., Deformation-induced ATP release from red blood cells requires
cystic fibrosis transmembrane conductance regulator activity. Am. J. Physiol, 1998. 275:
p. H1726-H1732.

156.

Olearczyk, J.J., et al., Heterotrimeric G protein Gi is involved in a signal transduction
pathway for ATP release from erythrocytes. American Journal of Physiology, 2004.
286(3, Pt. 2): p. H940-H945.

157.

Sprague, R.S., et al., Participation of cAMP in a signal-transduction pathway relating
erythrocyte deformation to ATP release. American Journal of Physiology, 2001. 281(4, Pt.
1): p. C1158-C1164.

158.

Olearczyk, J.J., Stephenson, Alan H., Lonigro, Andrew J., Sprague, Randy S., Receptormediated activation of the heterotrimeric G-protein Gs results in ATP release from
erythrocytes. Medical Science Monitor, 2001. 7(4): p. 669-674.

159.

Price, A., K., et al., Deformation-induced release of ATP from erythrocytes in a
poly(dimethylsiloxane)-based microchip with channels that mimic resistance vessels.
Analytical Chemistry, 2004. 76(16): p. 4849-55.

160.

Steiner, D.F., et al., Proinsulin and the biosynthesis of insulin. Recent Prog Horm Res,
1969. 25: p. 207-82.

161.

Rubenstein, A.H., et al., Proinsulin and C-peptide in blood. Diabetes, 1972. 21(2 Suppl):
p. 661-72.

63

162.

Hoogwerf, B.J., et al., Infusion of synthetic human C-peptide does not affect plasma
glucose, serum insulin, or plasma glucagon in healthy subjects. Metabolism, Clinical and
Experimental, 1986. 35(2): p. 122-5.

163.

Jensen, M.E. and E.J. Messina, C-peptide induces a concentration-dependent dilation of
skeletal muscle arterioles only in presence of insulin. Am J Physiol, 1999. 276(4 Pt 2): p.
H1223-8.

164.

Scalia, R., et al., C-peptide inhibits leukocyte-endothelium interaction in the
microcirculation during acute endothelial dysfunction. FASEB J, 2000. 14(14): p. 2357-64.

165.

Young, L.H., et al., C-peptide exerts cardioprotective effects in myocardial ischemiareperfusion. Am J Physiol Heart Circ Physiol, 2000. 279(4): p. H1453-9.

166.

Pierson, C.R., W. Zhang, and A.A. Sima, Proinsulin C-peptide replacement in type 1
diabetic BB/Wor-rats prevents deficits in nerve fiber regeneration. Journal of
Neuropathology and Experimental Neurology, 2003. 62(7): p. 765-79.

167.

Maezawa, Y., et al., Influence of C-peptide on early glomerular changes in diabetic mice.
Diabetes/metabolism research and reviews, 2006. 22(4): p. 313-22.

168.

Sima, A.A., et al., Inflammation in Diabetic Encephalopathy is Prevented by C-Peptide.
Rev Diabet Stud, 2009. 6(1): p. 37-42.

169.

Lindstrom, K., et al., Acute effects of C-peptide on the microvasculature of isolated
perfused skeletal muscles and kidneys in rat. Acta Physiol Scand, 1996. 156(1): p. 19-25.

170.

Hansen, A., et al., C-peptide exerts beneficial effects on myocardial blood flow and
function in patients with type 1 diabetes. Diabetes, 2002. 51(10): p. 3077-3082.

171.

Schalkwijk, C.G. and C.D. Stehouwer, Vascular complications in diabetes mellitus: the
role of endothelial dysfunction. Clin Sci (Lond), 2005. 109(2): p. 143-59.

172.

Johansson, B.L., J. Wahren, and J. Pernow, C-peptide increases forearm blood flow in
patients with type 1 diabetes via a nitric oxide-dependent mechanism. Am J Physiol
Endocrinol Metab, 2003. 285(4): p. E864-70.

173.

Wallerath, T., et al., Stimulation of endothelial nitric oxide synthase by proinsulin Cpeptide. Nitric Oxide, 2003. 9(2): p. 95-102.

174.

Kitamura, T., et al., Proinsulin C-peptide activates cAMP response element-binding
proteins through the p38 mitogen-activated protein kinase pathway in mouse lung
capillary endothelial cells. Biochemical Journal, 2002. 366(3): p. 737-744.

64

175.

Sprague, R.S., et al., Extracellular ATP signaling in the rabbit lung: erythrocytes as
determinants of vascular resistance. Am J Physiol Heart Circ Physiol, 2003. 285(2): p.
H693-700.

176.

Sprung, R.J., R.S. Sprague, and D.M. Spence, Determination of ATP release from
erythrocytes using microbore tubing as a model of resistance vessels in vivo. Anal.
Chem., 2002. 74: p. 2274-2278.

177.

Wan, J., W.D. Ristenpart, and H.A. Stone, Dynamics of shear-induced ATP release from
red blood cells. Proc Natl Acad Sci U S A, 2008. 105(43): p. 16432-7.

178.

Rigler, R., et al., Specific binding of proinsulin C-peptide to human cell membranes. Proc.
Natl. Acad. Sci., 1999. 96: p. 13318-13323.

179.

Henriksson, M., et al., Separate functional features of proinsulin C-peptide. Cell Mol Life
Sci, 2005. 62(15): p. 1772-8.

180.

Luppi, P., et al., C-peptide is internalised in human endothelial and vascular smooth
muscle cells via early endosomes. Diabetologia, 2009. 52(10): p. 2218-28.

181.

Minshall, R.D. and A.B. Malik, Transport across the endothelium: regulation of
endothelial permeability. Handb Exp Pharmacol, 2006(176 Pt 1): p. 107-44.

182.

Minshall, R.D., et al., Vesicle formation and trafficking in endothelial cells and regulation
of endothelial barrier function. Histochem Cell Biol, 2002. 117(2): p. 105-12.

183.

Lossinsky, A.S. and R.R. Shivers, Structural pathways for macromolecular and cellular
transport across the blood-brain barrier during inflammatory conditions. Review. Histol
Histopathol, 2004. 19(2): p. 535-64.

184.

Le Sueur, L.P., C.B. Collares-Buzato, and M.A. da Cruz-Hofling, Mechanisms involved in
the blood-brain barrier increased permeability induced by Phoneutria nigriventer spider
venom in rats. Brain Research, 2004. 1027(1-2): p. 38-47.

185.

Groneberg, D.A., et al., Distribution and function of the peptide transporter PEPT2 in
normal and cystic fibrosis human lung. Thorax, 2002. 57(1): p. 55-60.

186.

Meredith, D. and C.A. Boyd, Dipeptide transport characteristics of the apical membrane
of rat lung type II pneumocytes. Am J Physiol, 1995. 269(2 Pt 1): p. L137-43.

187.

Lindahl, E., et al., Cellular internalization of proinsulin C-peptide. Cell. Mol. Life Sci.,
2007. 64: p. 479-486.

188.

Lindahl, E., et al., Proinsulin C-peptide regulates ribosomal RNA expression. J Biol Chem,
2010. 285(5): p. 3462-9.

65

189.

Jo, H., et al., Endothelial albumin permeability is shear dependent, time dependent, and
reversible. Am J Physiol, 1991. 260(6 Pt 2): p. H1992-6.

190.

Meyer, J.A., et al., Zinc-activated C-peptide resistance to the type 2 diabetic erythrocyte
is associated with hyperglycemia-induced phosphatidylserine externalization and
reversed by metformin. Mol Biosyst, 2009. 5(10): p. 1157-62.

191.

Keltner, Z., et al., Mass spectrometric characterization and activity of zinc-activated
proinsulin C-peptide and C-peptide mutants. Analyst, 2010. 135(2): p. 278-288.

192.

Brooks, P., J. Meyer, and D.M. Spence, Metal (Zn ) Activated C-Peptide Induced ATP
Release Via Red Blood Cells with the Addition of Albumin. Abstracts, 60th Southeast
Regional Meeting of the American Chemical Society, Nashville, TN, United States,
November 12-15, 2008: p. SERM-680.

193.

Medawala, W., et al., A Molecular Level Understanding of Zinc Activation of C-peptide
and its Effects on Cellular Communication in the Bloodstream. Rev Diabet Stud, 2009.
6(3): p. 148-58.

194.

Hutton, J.C., The insulin secretory granule. Diabetologia, 1989. 32(5): p. 271-81.

195.

Shafqat, J., et al., Proinsulin C-peptide elicits disaggregation of insulin resulting in
enhanced physiological insulin effects. Cell Mol Life Sci, 2006. 63(15): p. 1805-11.

196.

Kunt, T., et al., The effect of human proinsulin C-peptide on erythrocyte deformability in
patients with type I diabetes mellitus. Diabetologia, 1999. 42(4): p. 465-471.

197.

Rieg, T., et al., P2Y(2) receptor activation decreases blood pressure and increases renal
Na(+) excretion. Am J Physiol Regul Integr Comp Physiol, 2011. 301(2): p. R510-8.

198.

Gonzalez-Alonso, J., ATP as a mediator of erythrocyte-dependent regulation of skeletal
muscle blood flow and oxygen delivery in humans. The Journal of physiology, 2012.

199.

Burnstock, G., Control of vascular tone by purines and pyrimidines. Br J Pharmacol, 2010.
161(3): p. 527-9.

200.

Sprague, R.S. and M.L. Ellsworth, Erythrocyte-derived ATP and perfusion distribution:
Role of intracellular and intercellular communication. Microcirculation, 2012.

201.

Kuzuya, H., et al., Heterogeneity of circulating C-peptide. The Journal of clinical
endocrinology and metabolism, 1977. 44(5): p. 952-62.

202.

Horwitz, D.L., et al., Proinsulin, insulin, and C-peptide concentrations in human portal
and peripheral blood. J Clin Invest, 1975. 55(6): p. 1278-83.

2+

66

203.

Dovichi, N.J. and J. Zhang, How Capillary Electrophoresis Sequenced the Human Genome
This Essay is based on a lecture given at the Analytica 2000 conference in Munich
(Germany) on the occasion of the Heinrich-Emanuel-Merck Prize presentation. Angew
Chem Int Ed Engl, 2000. 39(24): p. 4463-4468.

204.

Kidd, J.M., et al., Haplotype sorting using human fosmid clone end-sequence pairs.
Genome Res, 2008. 18(12): p. 2016-23.

205.

Kidd, J.M., et al., Mapping and sequencing of structural variation from eight human
genomes. Nature, 2008. 453(7191): p. 56-64.

206.

Easley, C.J., et al., A fully integrated microfluidic genetic analysis system with sample-inanswer-out capability. Proc Natl Acad Sci U S A, 2006. 103(51): p. 19272-7.

207.

Easley, C.J., et al., Rapid DNA amplification in glass microdevices. Methods Mol Biol,
2006. 339: p. 217-32.

208.

Dittrich, P., S. and A. Manz, Lab-on-a-chip: Microfluidics in Drug Discovery. Nature
Reviews Drug Discovery, 2006. 5: p. 210-218.

209.

Dittrich, P.S., K. Tachikawa, and A. Manz, Micro Total Analysis Systems. Latest
Advancements and Trends. Anal. Chem., 2006. ASAP.

210.

Whitesides, G.M., The origins and the future of microfluidics. Nature, 2006. 442(7101):
p. 368-373.

211.

Genes, L., I., et al., Adressing a Vascular Endothelium Array with Blood Components
using Underlying Microfluidic Channels. Lab on a Chip, 2007. 7(10): p. 1256-1259.

212.

Song, J.W., et al., Microfluidic endothelium for studying the intravascular adhesion of
metastatic breast cancer cells. PLoS ONE, 2009. 4(6): p. e5756.

213.

Barbulovic-Nad, I., et al., Digital microfluidics for cell-based assays. Lab on a Chip, 2008.
8(4): p. 519-526.

214.

Pihl, J., et al., Microfluidics for cell-based assays. Materials Today, 2005. 8(12): p. 46-51.

215.

Oblak, T.D., J.A. Meyer, and D.M. Spence, A microfluidic technique for monitoring
bloodstream analytes indicative of C-peptide resistance in type 2 diabetes. Analyst, 2009.
134(1): p. 188-193.

216.

Hung, P.J., et al., Continuous perfusion microfluidic cell culture array for high-throughput
cell-based assays. Biotechnol Bioeng, 2005. 89(1): p. 1-8.

217.

Kim, M.S., J.H. Yeon, and J.K. Park, A microfluidic platform for 3-dimensional cell culture
and cell-based assays. Biomed Microdevices, 2007. 9(1): p. 25-34.
67

218.

McDonald, J.C., et al., Fabrication of microfluidic systems in poly(dimethylsiloxane).
Electrophoresis, 2000. 21(1): p. 27-40.

219.

Duffy, D.C., et al., Rapid prototyping of microfluidic systems in poly(dimethylsiloxane).
Analytical Chemistry, 1998. 70(23): p. 4974-4984.

220.

Mehta, G., et al., Quantitative measurement and control of oxygen levels in microfluidic
poly(dimethylsiloxane) bioreactors during cell culture. Biomed Microdevices, 2007. 9(2):
p. 123-34.

221.

Leclerc, E., Y. Sakai, and T. Fujii, Microfluidic PDMS (polydimethylsiloxane) bioreactor for
large-scale culture of hepatocytes. Biotechnol Prog, 2004. 20(3): p. 750-5.

222.

Soper, S.A., et al., Polymeric microelectromechanical systems. Anal Chem, 2000. 72(19):
p. 642A-651A.

223.

Regehr, K.J., et al., Biological implications of polydimethylsiloxane-based microfluidic cell
culture. Lab Chip, 2009. 9(15): p. 2132-9.

224.

Shamloo, A., et al., Endothelial cell polarization and chemotaxis in a microfluidic device.
Lab on a Chip, 2008. 8(8): p. 1292-1299.

225.

Young, E.W., A.R. Wheeler, and C.A. Simmons, Matrix-dependent adhesion of vascular
and valvular endothelial cells in microfluidic channels. Lab Chip, 2007. 7(12): p. 1759-66.

226.

Lu, H., et al., Microfluidic shear devices for quantitative analysis of cell adhesion. Anal
Chem, 2004. 76(18): p. 5257-64.

227.

Sin, A., et al., Enrichment using antibody-coated microfluidic chambers in shear flow:
model mixtures of human lymphocytes. Biotechnol Bioeng, 2005. 91(7): p. 816-26.

228.

Wong, K.H.K., et al., Microfluidic Models of Vascular Functions. Annual Review of
Biomedical Engineering, 2012. 14(1): p. null.

229.

Srigunapalan, S., et al., A microfluidic membrane device to mimic critical components of
the vascular microenvironment. Biomicrofluidics, 2011. 5(1): p. 13409.

230.

Kaihara, S., et al., Silicon micromachining to tissue engineer branched vascular channels
for liver fabrication. Tissue Eng, 2000. 6(2): p. 105-17.

231.

Lim, D., et al., Fabrication of microfluidic mixers and artificial vasculatures using a highbrightness diode-pumped Nd:YAG laser direct write method. Lab Chip, 2003. 3(4): p. 31823.

232.

Borenstein, J.T., et al., Microfabrication of three-dimensional engineered scaffolds.
Tissue Eng, 2007. 13(8): p. 1837-44.
68

233.

Huh, D., et al., Reconstituting organ-level lung functions on a chip. Science, 2010.
328(5986): p. 1662-8.

234.

Tavana, H., et al., Epithelium damage and protection during reopening of occluded
airways in a physiologic microfluidic pulmonary airway model. Biomed Microdevices,
2011. 13(4): p. 731-42.

235.

Suwanpayak, N., et al., Blood cleaner on-chip design for artificial human kidney
manipulation. Int J Nanomedicine, 2011. 6: p. 957-64.

236.

Jang, K.J. and K.Y. Suh, A multi-layer microfluidic device for efficient culture and analysis
of renal tubular cells. Lab Chip, 2010. 10(1): p. 36-42.

237.

Grosberg, A., et al., Ensembles of engineered cardiac tissues for physiological and
pharmacological study: heart on a chip. Lab Chip, 2011. 11(24): p. 4165-73.

238.

Kim, H.J., et al., Human gut-on-a-chip inhabited by microbial flora that experiences
intestinal peristalsis-like motions and flow. Lab Chip, 2012. 12(12): p. 2165-74.

239.

Chronis, N., M. Zimmer, and C.I. Bargmann, Microfluidics for in vivo imaging of neuronal
and behavioral activity in Caenorhabditis elegans. Nat Methods, 2007. 4(9): p. 727-31.

240.

Chung, K., M.M. Crane, and H. Lu, Automated on-chip rapid microscopy, phenotyping
and sorting of C. elegans. Nat Methods, 2008. 5(7): p. 637-43.

241.

Shao, J., et al., A microfluidic chip for permeability assays of endothelial monolayer.
Biomed Microdevices, 2010. 12(1): p. 81-8.

242.

Young, E.W., et al., Technique for real-time measurements of endothelial permeability in
a microfluidic membrane chip using laser-induced fluorescence detection. Anal Chem,
2010. 82(3): p. 808-16.

243.

Imura, Y., et al., A microfluidic system to evaluate intestinal absorption. Anal Sci, 2009.
25(12): p. 1403-7.

244.

Douville, N.J., et al., Fabrication of two-layered channel system with embedded
electrodes to measure resistance across epithelial and endothelial barriers. Anal Chem,
2010. 82(6): p. 2505-11.

245.

Ferrell, N., et al., A microfluidic bioreactor with integrated transepithelial electrical
resistance (TEER) measurement electrodes for evaluation of renal epithelial cells.
Biotechnol Bioeng, 2010. 107(4): p. 707-16.

246.

Nalayanda, D.D., et al., An open-access microfluidic model for lung-specific functional
studies at an air-liquid interface. Biomed Microdevices, 2009.

69

247.

Sun, T., et al., On-chip epithelial barrier function assays using electrical impedance
spectroscopy. Lab on a Chip, 2010. 10(12): p. 1611-1617.

248.

Hediger, S., et al., Biosystem for the culture and characterisation of epithelial cell tissues.
Sensors and Actuators B-Chemical, 2000. 63(1-2): p. 63-73.

249.

Hediger, S., et al., Modular microsystem for epithelial cell culture and electrical
characterisation. Biosens Bioelectron, 2001. 16(9-12): p. 689-94.

250.

Booth, R. and H. Kim, Characterization of a microfluidic in vitro model of the blood-brain
barrier (muBBB). Lab Chip, 2012. 12(10): p. 1784-92.

251.

Esch, M.B., et al., Characterization of in vitro endothelial linings grown within
microfluidic channels. Tissue Eng Part A, 2011. 17(23-24): p. 2965-71.

252.

Gaspar, S., et al., Simultaneous impedimetric and amperometric interrogation of renal
cells exposed to a calculus-forming salt. Analytica Chimica Acta, 2012. 713: p. 115-20.

253.

Nan, Y.S., et al., Different effects of dibutyryl cAMP on monolayer permeability in human
aortic and coronary arterial endothelial cells. African Journal of Microbiology Research,
2012. 6(5): p. 897-903.

254.

Khademhosseini, A., et al., Microscale technologies for tissue engineering and biology.
Proceedings of the National Academy of Sciences of the United States of America, 2006.
103(8): p. 2480-2487.

255.

Letourneau, S., et al., Evaluating the effects of estradiol on endothelial nitric oxide
stimulated by erythrocyte-derived ATP using a microfluidic approach. Anal Bioanal
Chem, 2010. 397(8): p. 3369-75.

256.

Olearczyk, J.J., et al., NO inhibits signal transduction pathway for ATP release from
erythrocytes via its action on heterotrimeric G protein Gi. American Journal of
Physiology, 2004. 287(2, Pt. 2): p. H748-H754.

257.

El-Ali, J., P.K. Sorger, and K.F. Jensen, Cells on chips. Nature, 2006. 442(7101): p. 403-11.

258.

Skelley, A.M., et al., Microfluidic control of cell pairing and fusion. Nat Methods, 2009.
6(2): p. 147-52.

259.

Paguirigan, A.L. and D.J. Beebe, Microfluidics meet cell biology: bridging the gap by
validation and application of microscale techniques for cell biological assays. BioEssays,
2008. 30(9): p. 811-21.

70

Chapter 2 - Development of a Microfluidic Device Capable of Measuring TEER
2.1 - TEER Measurements on a Microfluidic Vascular Mimic
Studying vascular wall biology and interactions with blood components is an important area
of research due to the prevalence of cardiovascular diseases in which these interactions are
implicated in disease progression. Efforts to understand the cardiovascular system has led to
the development of vascular models that replicate conditions found in vivo. The goal of these
models is to simplify the system so that a more detailed and convenient analysis can be
performed with focus placed completely on certain components of the system that are
convoluted in whole organisms. In an in vivo situation, flowing blood is always surrounded by
an endothelial cell monolayer that forms a vascular wall. The endothelial cells and the blood
components are continuously interacting in a dynamic fashion via cell-cell communication
where flowing blood is able to elicit endothelial cellular responses, and the endothelial cells are
also capable of influencing cellular behavior of various blood components. Examples of
important biological events caused by these interactions between different cell types in the
vascular system are continually emerging. These interactions between various cell types are
biologically important and, as a result, a drive for the development of technologies capable of
elucidating the mechanisms of cell-cell communications is underway. Also, cell types that are
found in close proximity in vivo are often cultured together in vitro so that interactions between
the cells are accounted for, yielding more physiologically relevant cellular responses and,
therefore, increasing the validity of experimental results.

71

TEER measurements have been universally established as a non-invasive and reliable
method for evaluating and monitoring the growth of cellular monolayers. As such, a simple way
of performing TEER measurements on in vitro biological systems was made commercially
available by World Precision Instruments in the mid-1980’s as discussed previously in section
1.1.3. A version of this instrumentation is still available and is commonly used for cell based
assays. This TEER system was previously integrated into microfabricated cell culture systems in
an important step towards integrating TEER measurements with microfluidic technology.[1, 2]
Unfortunately, the resulting microsystem did not allow for flow, an important requirement for
adequately mimicking certain biological systems as discussed in section 1.3.2.
Microfluidic systems have recently proven useful for mammalian cell culture and cell-based
assays,[3-6] including the culture and subsequent cellular assay of cell types found in the
vasculature.[5, 7-10] The inherent geometries and dimensions of microfluidic systems are well
suited for mimicking in vivo physiological conditions and monitoring cell behavior within these
microenvironments. An in vitro cell layer mimic capable of incorporating flow, and other
biologically relevant cells, such as the RBC, would be paramount for mimicking a true vascular
barrier. However, the various cells would have to be in close proximity to allow for cell-cell
communication. These issues are well addressed through microfluidic systems, as it is possible
to mimic the small distance between the blood vessel and the endothelium.[7] Blood flow is
also easily accomplished by incorporation of channels that have diameters similar to
capillaries.[7] Furthermore, microfluidic systems lend themselves to high throughput
techniques,[11-14] and could provide a more realistic model of a vascular wall. Microfluidics
have the unique ability to adequately model in vivo conditions to the extent that cell-cell

72

communications can be observed, an ideal property that is absent in many conventional
biological mimics.
Microfluidics have recently been utilized for vascular research by our group and others,
highlighting the utility of microfluidics as in vitro models of the vasculature.[5, 10, 15-20]
However, since many vascular mimics include an endothelial monolayer of cells, it is important
to remember that cellular responses have been shown to vary with cell monolayer confluence
as discussed in Section 1.1. Because of this, a measure of cell barrier integrity, such as TEER
determination, is common practice during studies involving cell monolayers, including those
often found in models of biological systems, prior to experimentation involving these
monolayers. A confirmation of cell monolayer confluence prior to experimentation adds
credibility to experimental results by ensuring more physiologically appropriate cellular
behavior, which is quite advantageous for elucidating the mechanisms of cell-cell
communication where cell responses may be difficult to detect. These benefits offer ample
incentive for the integration of a TEER measurement system with a microfluidic vascular model.
However, this work illustrates the first microfluidic vascular mimic capable of performing TEER
measurements.[21] Perhaps more importantly, this work describes the first microfluidic device
capable of measuring TEER that also includes flow capabilities.[22] Since this work began, other
examples of microfluidic devices capable of monitoring barrier integrity of various cell types
have appeared in the literature, suggesting the potential importance of such devices.[23-25]
2.2 - Methods
2.2.1 - Device Fabrication

73

Soft lithography techniques are used to produce PDMS slabs from silicon masters. Briefly,
silicon masters are spin coated with 4 mL of SU8-50 photoresist (MicroChem Corp.) at 500 rpm
for 15 seconds, followed by a 1000 rpm spin cycle for 30 seconds. The wafer is then allowed to
rest on a hotplate at 95°C for 15 minutes. A transparency mask is placed over the wafer, and
subsequently exposed to UV light for 1 minute for a 100 µm thick feature[26]. The wafer is then
transferred again to a hotplate at 95°C for 7 minutes before it is developed in propylene glycol
methyl ether acetate (Sigma) to remove excess photoresist. The resulting completed silicon
master is suitable for rapid prototyping of PDMS slabs with embedded features. The process for
fabricating a silicon master is depicted in Figure 2.1a.
PDMS slabs are prepared by mixing Dow Corning’s Sylgard 184 bulk polymer (Ellsworth
Adhesives, Germantown WI) with curing agent in a desired ratio by mass. A 20:1 ratio of bulk
polymer to curing agent mixture is used where sealing is desired, whereas a more rigid 5:1
mixture is utilized on edges, or wherever a more rigid area was desired. The PDMS mixture is
well stirred and then degassed by vacuum prior to pouring onto a silicon master. Each layer of
PDMS is subsequently cured by heating for 12 minutes at 75°C.[27] The cured PDMS slab can be
removed from the silicon master. Any raised features on the silicon master will be embedded in
the PDMS slab. The process of rapid prototyping PDMS slabs from a silicon master is depicted in
Figure 2.1b. Outlets and wells are punched in the PDMS slabs using 1/8 in. and 1/4 in. paper
hole punches. A 23 gauge luer stub adapter is employed to punch inlets, and then plumbed
using 20 gauge stainless steel tubing to allow for connectivity to syringe pumps. After the
microfluidic device is constructed, it is irreversibly sealed by heating for 30 minutes at 75°C.

74

Figure 2.1 - Soft lithography techniques are used for the rapid prototyping of PDMS slabs from
silicon masters. Briefly, as shown in (a), a silicon wafer is spin-coated with SU-8 photoresist, and
a negative mask is placed over the photoresist prior to UV light exposure. The photoresist
exposed to light cross links to form a rigid polymer structure, and the remaining SU-8 polymer is
dissolved by a developing agent, resulting in a silicon master with raised features. As shown in
(b), PDMS polymer and curing agent is mixed in a desired ratio and degased prior to pouring
over the silicon master. The PDMS is then cured at 75° C for 15 minutes and subsequently
removed from the master, yielding a PDMS slab with embedded features.

75

The construction of the microfluidic device requires three PDMS slabs as shown in figure
2.2. The middle slab of PDMS has channels whose dimensions are 200 μm wide by 100 μm tall,
which are prepared using standard soft lithographic techniques as described above. At one end
of each channel, inlets are punched, and in the center of each channel, a 1/8 in. hole is punched
to act as a well. The top PDMS slab has no channels or other features, but does contain two
columns of wells, one column that aligns with the wells punched in the center of each channel,
and one column that aligns with the end of each channel on the middle slab of PDMS. The
column of wells in the center of each channel on the middle PDMS slab is separated by the
column of wells on the upper PDMS slab by a polycarbonate membrane. The pore size of the
membrane is typically 0.22 μm, but 0.1-1.0 μm pore sizes have been used in our group. The
column of wells on the upper PDMS slab lying above the end of the channels on the middle
PDMS slab act as waste wells for excess solution that has passed beneath the polycarbonate
membrane. An aluminum strip is placed underneath the column of wells that are punched in
the center of the channels on the middle PDMS slab. The aluminum strip is then sealed in place
with a bottom slab of PDMS that has been halved using a scalpel, also without any channels or
other features. The bottom slab of PDMS is cut in half in order to allow access to the inlets that
are punched in the middle slab of PDMS, which is required for plumbing of the channels.
The aluminum strip running beneath each of the membranes serves as a stationary
common electrode that is sufficient for TEER measurements across every membrane. The well
punched in the center of each channel on the middle PDMS slab allows solution that is flowing
through the channel to contact the electrode. The electrode can be polished by sanding the
electrode surface with a Dremel tool to remove debris and subsequently reused.

76

Figure 2.2 - The microfluidic TEER device is fabricated from 3 PDMS slabs. The device contains a
series of flow channels integrated with a common electrode below the cell layer, which is
cultured on the polycarbonate membrane. A second copper top electrode (not pictured) is
placed in the wells of the top layer of PDMS to perform a TEER measurement directly across the
cell monolayer.

77

2.2.2 - Pumping and Attachments
Syringe pumps (Harvard Apparatus) are employed to drive 500 µL Hamilton Gastight
syringes for controlled flow of solution in the microfluidic channels. Tygon tubing is interfaced
to the syringes using an adapter connecting a male luer syringe to a female 1/16 in.
compression fitting (Upchurch/IDEX scientific). The Tygon tubes are interfaced with the
microfluidic device using 20 gauge stainless steel tubing (New England Small Tube Company),
which fits securely into both the Tygon tubing and the microfluidic inlets described in 2.2.1.
A top electrode consists of a 16 gauge copper wire that has been secured with epoxy to a
pipet tip, forming a disk electrode. The epoxy is used to secure the wire in the pipet tip to
prevent capillary effects of the solution and likely improving the reproducibility between
measurements by maintaining a constant effective electrode area (area of the electrode
allowed to contact solution). The top electrode is then clamped in place using a ring stand, and
the microfluidic device is raised until the top disk electrode rests inside of the well in which a
TEER measurement is desired. The outer diameter of the pipette tip and the inner diameter of
the 1/8 in. well are nearly identical, so the electrode rests directly in the center of the well for
every measurement.
2.2.3 - bPAEC Culture
Bovine pulmonary artery endothelial cells (bPAEC) are purchased from Lonza and arrive
frozen. They are cryogenically frozen for storage, and are thawed to room temperature and
then added to a T-75 tissue culture flask containing 12 mL of cell culture media that has been
equilibrated at 37°C. Cell culture media is prepared from Lonza’s EGM MV media containing 5%
fetal bovine serum with gentamicin and phenol red. Subsequently, bPAECs are incubated at 37
78

°C and 5% CO2 atmosphere until confluence is reached. Media is changed the day after plating
the cells, and then every two to three days thereafter. When the cells reach approximately 80%
confluence as identified by optical microscopy, subculturing is performed.
Seeding cells in the wells of a microfluidic device is accomplished by first coating the
polycarbonate membrane with fibronectin. Fibronectin (Invitrogen) is dissolved in phosphate
buffered saline (PBS) to prepare a 1000 µg/mL stock solution. The stock solution is aliquoted
and frozen until use. When cell seeding is desired, an aliquot of stock solution is thawed and
diluted to 100 µg/mL and 10 µL of solution are added to each microfluidic well. The solution is
allowed to completely evaporate to yield a fibronectin coating on the polycarbonate
membrane.
bPAECs are cultured in T-75 tissue culture flasks, and are removed by pipetting 2 mL of a
0.25% trypsin solution onto the cells and incubated for 2 minutes. The trypsin solution is
removed by aspiration and immediately replaced with 6 mL of cell media. The cells are
repeatedly rinsed and the bottom of the flask is gently rapped to promote removal of the
adhered cells. The resulting cell suspension is centrifuged at 500 g for 5 min. and the
supernatant removed. The cells are then resuspended in 1 mL of cell media and 10 µL of the
suspension are added to each well of the microfluidic device. Cell media is replaced after 1.5
hours, and then again every 3 hours afterwards. If the cells are allowed to incubate overnight,
18 µL of cell media are added to each well, and the device is stored in a covered petri dish with
a Kimwipe saturated with water in order to prevent evaporation of the cell media. Media is also
replaced prior to each TEER measurement in order to maintain electrolyte concentration in the
electrochemical cell for each measurement.

79

2.3 - Results
2.3.1 - TEER Measurement Method Development
A custom LabView (National Instruments) program was written in-house and implemented
to perform TEER measurements. The software applies a 20 Hz square wave with a magnitude of
0.75 V across the membrane for 2 seconds as shown in Figure 2.3a. This potential is applied
using the analog output of a DAQ board (National Instruments), and the resulting current is
measured at an acquisition rate of 1000 Hz. The resultant current signal (Figure 2.3b) is then
analyzed via MATLAB (Mathworks), automatically performing curve fitting and integration in
real time. The resultant current signal is averaged and integrated to provide charge, which is
plotted as a function of time as in Figure 2.4.
A typical TEER device, discussed in section 1.2.3, is comprised of a small volume cup that
contains a membrane. This cup then sits in a larger well and, when both are filled with liquid
(e.g., a physiological buffer), a voltage can be applied and the current passing through the
membrane can be determined. Specifically, a standard TEER measurement is performed with a
four electrode system, and consists of an instrument that applies a potential through a batterypowered square wave generator at approximately 12 Hz. Additional sensing electrodes are
employed to measure the resulting current through a built-in traditional multimeter.
To simplify fabrication of our device and eliminate the need for four electrodes, the device
shown in Figure 2.2 was constructed. Current resulting from a potential applied across cellular
monolayers cultured in the wells of the completed device can be measured and characterized.
When the system is operational with added electrolyte solutions and endothelial cell layers

80

cultured upon the polycarbonate membranes, it can be modeled as an RC circuit. As such,
applying a steady potential should yield a current profile as predicted by Equation 2.1,

I (t ) = I oe
where I is current,

−

t
RC

(2.1)

t is time, Io is the initial current, R is resistance, and C is capacitance.

The resulting current decays were averaged and integrated with respect to time, the result of
which is the average charge passed through the circuit in a single pulse as described by
equation 2.2

∫ I ∂t =

∂Q
V
+c= .
∂t
R

(2.2)

where Q is charge, and V is voltage. Equation 2.2 describes the inverse relationship between
resistance and average charge. Therefore, TEER values are reported in units of coulombs, which
are inversely proportional to resistance, and represent the average charge passed through the
electrochemical cell during a single potential pulse as described by Equation 2.2. Due to the
inverse relationship of average charge and resistance, the average charge will reach a minimum
charge at maximum cell confluence, which is indicative of the maximum resistance measured
by commercial systems when maximum cell confluence is achieved.
In order to report an estimate of resistance in the electrochemical cell, total current was
determined by
t

Q = ∫ I ∂t + c
0

81

(2.4)

where

Q is the average charge determined by the TEER system, I is the total current, and ∂t is

the time interval of the applied square wave. After the total current is determined, resistance
can be determined by

R=

V
I

(2.5)

where R is resistance, V is the magnitude of the applied square wave potential, and I is the
total current determined by Equation 2.4. While measurements acquired using the microfluidic
TEER system introduced in this chapter are typically reported in units of charge, TEER values are
conventionally reported in units of ohms·cm2, which requires normalization to the area of the
cultured cell monolayer. In order to accurately determine the area of microfluidic wells, calipers
are employed to measure the diameter of each well, from which the cross-sectional area of the
well is determined in units of cm2. It was determined that the microfluidic wells had an average
well diameter of 1.76 ± 0.08 mm (n=10 wells), which is used to calculate TEER estimates in
Figure 2.6. The TEER data can then be presented in ohms·cm2, by finding resistance as discussed
previously, and multiplying by the area of the microfluidic well.

82

1.0
0.8

Applied Volts (V)

0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.20

0.25

0.30

Time (seconds)

6.0

Current (mA)

4.0
2.0
0.0
-2.0
-4.0
-6.0
0.00

0.05

0.10

0.15
Time (seconds)

Figure 2.3 - Shown in the top graph is the bipolar pulsed perturbation signal used in the system,
a 20 Hz square wave. The resulting current is then measured as a function of time as shown in
the bottom graph, yielding current decays that resemble those typically seen in an RC circuit.
These current decays are then averaged and integrated, and the resulting charge value is
presented as a TEER signal in units of charge.
83

2.3.2 - Validation of TEER Measurements
The ability to measure resistance using the integrated TEER system was first examined by
performing measurements across solutions containing various KCl concentrations. Since salts,
such as KCl, completely dissociate in aqueous solutions to form ions, the concentration of ions
in the resulting solution is proportional to the amount of salt dissolved, assuming the solvent
does not contain ions initially. Therefore, by applying a constant potential across an ionic
solution, a resulting current will be induced that can be measured, which is directly
proportional to the conductivity of the solution and inversely proportional to solution
resistance. The data in Figure 2.3a represents the applied quare wave potentials (ranging from 750 mV to 750 mV at 20 Hz for 2 seconds), while the data in Figure 2.3b is the resultant current
decay profiles that coincide with the applied potentials. Charge is obtained by averaging the
current decays resulting from the 20 negative applied potential pulses, and integrating with
respect to time. The curve in Figure 2.4 shows charge plotted as a function of KCl concentration
above and below the membrane. Summarily, as the salt concentration increases in the device,
the current passed per potential pulse increases, resulting in an increased charge as shown in
2.4. The waveforms shown in 2.3a, the current profiles in 2.3b and the resultant integration of
averaged area in 2.4 were all performed using the computer controlled interface to generate
the waveforms, collect the resultant current data, and analyze the data.

84

Figure 2.4 - A cross-sectional view of the microfluidic device is shown in (a). Solutions
containing various concentrations of potassium chloride (KCl) are pumped into the channels of
the microfluidic device, and a solution of the same concentration is pipetted into the overlying
wells. TEER measurements are shown to correlate linearly with KCl concentration in (b),
displaying the ability of the device to measure conductance, which is inversely proportional to
resistance, and therefore indicative of resistance between the two electrodes. (n=4, errors are
SD)

85

The methods for performing conventional TEER measurements using microtiter based
techniques vary somewhat from the methodology we used to accomplish TEER measurements
on a microfluidic platform. For instance, microtiter based measurements utilize a four electrode
system in order to eliminate junction potentials. The necessity to simplify the system in a
smaller microfluidic based platform required the incorporation of only two electrodes, so
junction potentials are present in the microfluidic system. Additionally, a two electrode system
restricts the addition of a proper reference electrode. These inherent differences account for
TEER values being reported in units of coulombs, which are easier to calculate in our case, and
are still a satisfactory measurement of barrier integrity, and relative changes in that integrity, as
shown in Figure 2.5. However, the data is also presented in Figure 2.6 using conventional TEER
units of ohms·cm2.
2.3.3 - Monitoring bPAEC Monolayer Barrier Integrity on a Microfluidic Device
After it was determined that the total charge passed correlated with conductivity of an
electrochemical cell, the device was employed to estimate confluence of a monolayer of
bPAECs. In order to culture the cells within the wells of the microfluidic device, a fibronectin
coating was first added to the polycarbonate membrane representing the bottom of the well to
promote cell adhesion, as described in section 2.2.3. Cells are then added to each well, and cell
barrier integrity is monitored by performing TEER measurements as the cells become confluent.
A side view of the device with the cells in place on the polycarbonate membrane is shown in
figure 2.5. After seeding on the membrane, the bPAECs were given time to adhere to the
fibronectin matrix. During adhesion, the cells ability to develop into a confluent layer was
monitored at various time points using the integrated TEER measurement system.
86

The

resulting data, shown in Figures 2.5, demonstrates that the cells are able to form a more
resistive barrier with time, reaching a plateau after approximately 8 hours. Initially, after cell
seeding on the fibronectin-coated membrane, the applied potential resulted in approximately
22.3 ± 1.6 μC (40.9 ± 2.9 ohms·cm2). However, after 8 h, this value dropped to 3.5 ± 0.4 μC
(259.1 ± 27.4 ohms·cm2), a statistically significant decrease (p<0.001, n = 4 devices).

A

statistically non-significant change between 8 h and 17 h suggested that the cells had reached
an optimum resistivity.

87

Figure 2.5 - A cross-sectional view of one channel of the microfluidic device, where TEER
measurements are performed directly across an endothelial monolayer cultured on a
polycarbonate membrane. Shown in (b) are the results of TEER measurements performed on
endothelial cells over the course of 17 h in units of μCoulombs, demonstrating the cells reach
confluence at approximately 8 h. (n=4, errors are SD)

88

300

250

Ohms x cm2

200

150

100

50

0
0

2

4

6

8

10

12

14

16

18

Time (hours)

Figure 2.6 - TEER measurements performed on endothelial cells over the course of 17 h are
reported in conventional TEER units of ohms·cm2. These values are estimated from the TEER
measurements displayed in Figure 2.5b. These results again demonstrate that the cells reach
confluence at approximately 8 h, as suggested by the peak in TEER signal at that time. (n=4,
errors are SD)

89

2.4 - Discussion
The method for constructing the microfluidic device is depicted in Figure 2.2. Utilizing a
single aluminum strip as a single electrode for each microfluidic well eliminated the need for
including an individual electrode for every microfluidic well, simplifying the design and software
requirements compared to other microfluidic devices capable of TEER measurements that have
emerged since the development of our device.[23, 24, 28] Additionally, minimizing the number
of individual electrodes also eliminates the need to make numerous electrical connections,
simplifying experimental setup while decreasing the probability of errors that arise from device
handling and interfacing the device to external hardware. Designing the microfluidic TEER
device in a non-complex manner was intentional so that the device can be easily reproduced
and implemented for work in our laboratory, and in other research laboratories that intend to
culture cell monolayers in microfluidic devices. Also, since the device can be fabricated and
utilized simply and quickly, with minimal experimental setup and automated data collection
and analysis, the device can be prepared and used by researchers with only modest experience
in microfluidics.
In general, microfluidics are evolving towards more intricate devices that can be used for
complicated analyses involving increasingly complex research interests. As a result, microfluidic
devices are better suited for solving sophisticated and interesting questions in research, but
often entail a complicated and timely fabrication procedure that can only be accomplished by
microfluidics experts. Additionally, microfluidic devices that include sophisticated design
features requiring difficult and time consuming fabrication processes often limits the
widespread use of the technology. Furthermore, important experimental results are often

90

convoluted by complex microfluidic designs, as is the likelihood of device errors arising from
design flaws. For these reasons, the microfluidic TEER device was designed simply, however it is
still capable of performing complex analyses to study multifaceted research problems as shown
by experimental results obtained by the device over the course of this work. Fortunately, the
microfluidic device presented in this work can be constructed relatively quickly, requiring only
3-4 hours for fabrication, and is comprised of materials, such as PDMS, that are well studied
and frequently used in microfluidics. The microfluidic design showcased in this work was,
however, modified to maximize simplicity, reliability and reproducibility of the devices.
In order to fabricate the device, PDMS was selected as a fabrication material for several
reasons. Firstly, the device is intended to contain biological samples containing live cells, which
require a supply of gases, such as oxygen and carbon dioxide, in order to survive and behave
properly.[29] Healthy cell culture is most amenable in PDMS microfluidic devices because it is a
non-toxic, gas permeable polymer that does not compromise cell health by restricting the
diffusion of gases to and from cells within the device. Also, PDMS is well-studied and frequently
used for microfluidic device construction because soft lithography permits highly detailed
microfluidic features and rapid prototyping.[30] PDMS is also optically clear, enabling optical
measurements.
Furthermore, track etched polycarbonate membranes were selected as a cell culture
support because the support should be porous, allowing for diffusion of molecules between the
flowing solution and the cultured cells, which is required for cell-cell communication and
permeability assays. Additionally, polycarbonate membranes have been integrated into PDMS

91

microfluidic devices for years in the Spence group.[20] These aspects of the device remained
unchanged during development of the device.
Microfluidic devices with integrated electrodes commonly entail a complicated design in
order to successfully incorporate the electrodes, as well as lengthy fabrication times (on the
order of days). However, the benefits provided by the ability to measure TEER in a vascular
mimic provide incentive to overcome these challenges. Integration of the TEER system
ultimately increased the quality of experimental results obtained using the device, and also
addressed the overall absence of TEER measurements in microfluidic research. Initially, the
device was constructed with wire electrodes sealed in PDMS that addressed each channel
individually, a design that was modified for several reasons. Additionally, each wire had to be
interfaced to external hardware sequentially, which forces repeated handling of the device and
often compromised structural integrity. Here, the design was improved by using a single wire
that ran perpendicular to the channels so that each well could be addressed using one
electrode, similar to the aluminum electrode in the optimized final design, which reduced
fabrication time, device complexity, and handling during experiments. Unfortunately the
resulting microfluidic device was unreliable because upon solution contacted with the wire, the
solution would leak along the length of the wire due to capillary effects resulting from the
inability to completely seal a round wire in a PDMS microfluidic device.
The issues that arise by attempting to seal a metal electrode in PDMS were also
encountered by others when constructing a microfluidic TEER device.[23] The problem was
ultimately resolved by adding a channel to the PDMS design that was completely dedicated to
containing the wire. The wire is pushed into the channel, and the voids in the channel

92

surrounding the wire were filled with uncured PDMS mortar by syringe injections. The wire was
then able to be secured into place after the completed device was cured. This method
successfully resolves the issue, but requires a complicated fabrication procedure coupled with a
long fabrication time where the device must be allowed to cure overnight for two consecutive
nights. This issue was alternatively avoided this work by choosing an aluminum foil as an
electrode material because it is able to lie flush between two PDMS slabs, and with a thickness
of less than 200 µm, adequate sealing is accomplished. The width of the aluminum foil strip is
also always prepared in dimensions so that the width of the foil is greater than that of the
overlying wells. This helps prevent leakage in the device, and ensures that the electrode area
that makes contact with the solutions contained in the channels is as similar as possible from
channel to channel, leading to more reliable and reproducible TEER measurements.
The electrode areas resulting from using foil instead of wire were greatly increased. The
average effective electrode areas of the bottom electrodes in the microfluidic device are a
result of the size of well that is punched in the center of each channel. These wells are punched
with a 1/16 in. hole punch, resulting in an average electrode area of about 2.4 mm2. The
resulting macroelectrodes are often avoided in electrochemistry due to the production of large
charging currents required to form a double layer at the electrode surface. These charging
currents are sometimes so large that any faradaic current is not observable, which is not ideal
when using faradaic current to determine characteristics of an electrochemical cell, such as
resistance. However, the perturbation potential used in the TEER system, displayed in Figure
2.4a, is a 20 Hz square wave applied for 2 seconds. Because a potential pulse is applied for only
0.05 sec before being reversed, the resulting current response provides primarily kinetic

93

information. It is hypothesized that cell monolayer confluence in the electrochemical cell
affects the kinetics of faradaic processes, as well as the kinetics of double layer formation,
resulting in an ideal situation where even charging current can be used to measure TEER. The
validity of this hypothesis is supported by the TEER measurements performed using the device
where successful determination of conductivity and cell monolayer barrier integrity were
accomplished as shown in Figures 2.4 and 2.5, respectively. In a situation that is unique to our
microfluidic TEER system due to our measurement methods, the integration of
macroelectrodes in the device design was acceptable since charging current can be used in
addition to faradaic current the determine TEER. For these reasons, we were able to forego the
complications of integrating microelectrodes into the device, such as sealing wires in PDMS.
Interestingly, the applied perturbation potential, shown in Figure 2.4a, was not chosen
because of the advantages discussed previously, but rather for three other reasons. Firstly, the
measurement technique was implemented as one of the first successful methods for
determining resistance in a small electrochemical cell, including the microscale electrochemical
cells that are basically inevitable in microfluidic devices.[31, 32] Secondly, the square wave
potential is an ideal choice because it can only polarize the cells for a short time interval, an
important feature because polarization of cells for extended periods of time has been shown to
influence cell behavior[33, 34], and even cause cell lysis.[35] Lastly, since TEER measurements
have become a commercialized and standard measurement technique outside of microfluidic
research, we were able to use well established methods to develop our method for determining
TEER. A simple way of performing TEER measurements on in vitro biological systems was made
commercially available by World Precision Instruments in the mid-1980’s. A version of this

94

instrumentation is still available and is commonly used in microtiter based experiments as
discussed in section 2.1. The square wave potential applied by the microfluidic TEER system
shown closely resembles the perturbation potential used in the commercially available TEER
system, which applies a 12 Hz square wave potential and measures current to determine TEER.
However, the methodology used to determine TEER in the microfluidic system varies somewhat
from the traditional TEER device, where a four electrode system is employed to determine TEER
rather than the two electrode system utilized in the microfluidic TEER system.
The development of a two electrode system arose from the necessity to simplify the
method on a smaller microfluidic platform. A four electrode system is used in a standard TEER
system in order to eliminate junction potentials, which are present in the microfluidic TEER
device. Also, the resistance in an electrochemical cell in response to an amperometric potential
is typically determined by analyzing steady state current. However, for reasons discussed
previously, the nature of the applied square wave potential restricts the system from reaching a
steady state. Consequently, estimations of resistance in the microfluidic electrochemical cell
are determined when the system is kinetically controlled and charging current contributes to
the current. As a result of these differences, we can only estimate TEER in units of ohms·cm2
because determination of resistance is required. Regardless, TEER data is presented in
ohms·cm2 in figure 2.6 in an effort to compare our system to traditional measurements,
resulting in measurements similar to literature values.[36, 37] Even as estimates of TEER values
with units of ohms·cm2 tend to match literature values, presenting TEER values in units of
µCoulombs avoid estimation, and still provide a measurement of cell monolayer confluence,
and changes in that confluence, as shown in Figure 2.5.

95

The current response to the applied square wave in Figure 2.4b resembles a current decay
commonly seen in a typical RC circuit. These types of responses to a constant DC potential are
common and well characterized, allowing for simple isolation of important properties of the
electrochemical cell as discussed in section 2.3.1. Due to the predictability of the current
response observed, the manipulation of data to determine TEER values was able to be
completely automated allowing for the development of a simple user interface in which a TEER
value can be determined in seconds without the need for additional manipulation of data.
The data presented in Figure 2.5 displays the ability of the microfluidic device to
successfully perform measurements to quantitatively determine conductance in an
electrochemical cell. A microfluidic system capable of measuring conductance is not novel, but
it is a powerful tool used previously in vascular mimics, primarily to isolate rare cells from
circulating blood samples.[38] The purpose of determining the conductance of our microfluidic
electrochemical cell was to verify that our specific method for determining TEER was valid, as
the technique for determining conductance in the microfluidic device is the same as used to
determine TEER. The ability to measure conductance with a wide linear range indicates that the
software employed to perform TEER measurements, as well as the microfluidic design, was
robust enough to interrogate the barrier integrity cell monolayer.
The data presented in Figures 2.5 and 2.6 demonstrates the capacity for quantitatively
determining endothelial cell barrier integrity using the microfluidic TEER system. The initial
TEER measurement at 0 hours was taken immediately after the cells had been added to the
microfluidic well, and before seeding on the fibronectin network was possible. This initial TEER
measurement yielded an average charge of 22.3 ± 1.6 µC (or 40.9 ± 2.9 ohms·cm2) , a value over

96

6 fold greater than that obtained at 8 hours of 3.5 ± 0.4 µC (or 259.1 ± 27.4 ohms·cm2) when
the cells had reached optimal confluence. These findings illustrate the sensitivity of the TEER
system, and the ability to detect slight variations in cell monolayer confluence. Furthermore, a
non-statistical change in TEER measurements obtained from 8 h to 17 h shows that the bPAEC
monolayer had reached an optimal confluence that was verifiable by the device. Prior to the
these findings, our group would allow 1.5 to 2 hours for bPAEC seeding on microfluidic devices,
which does not allow for complete confluence according to the data in Figure 2.5. After
obtaining this data, cells were allowed to seed overnight to ensure confluence, which greatly
increased the reproducibility of cell based experiments in our lab.
An additional important feature of the device, aside from the ability to measure TEER, is the
ability to culture endothelial cells that are separated in a biologically relevant distance from a
capillary that allows for the flow of solution, such as a blood sample, producing a potential
microfluidic vascular mimic. The ability to mimic the vasculature using microfluidics has been
accomplished previously[20, 39, 40], however this work is the first example of a microfluidic
vascular mimic capable of measuring TEER.[21] Utilization of the microfluidic TEER device as a
vascular mimic will be the topic of the upcoming chapters. Since publication of this work, one
other microfluidic vascular mimic capable of performing TEER measurements has appeared in
literature[28], re-emphasizing the importance of confirming and monitoring cell monolayer
integrity, even in microfluidic devices. Since the primary incentives for inventing the device was
to perform endothelial permeability studies and to study cell-cell communication between
flowing RBCs and a vascular wall comprised of a monolayer of endothelial cells, integrating
TEER measurement capabilities was a priority in the development and fabrication of the device.

97

Often, when performing research involving monolayers of cells, a quantitative measure of
monolayer barrier integrity is essentially prerequisite to any investigation involving the cell
layer, an aspect not previously addressed in microfluidic research. Additionally, confirmation of
cell monolayer confluence is important for proper cellular behavior as discussed in Section
1.2.4.
Data presented in Figure 2.5 exhibits the ability of the microfluidic device to confirm the
barrier integrity of bPAECs, which is a cell type particularly interesting to our group due to our
intention to mimic the cardiovascular system. However, other biological research has recently
been performed using microfluidics to mimic specific organs, resulting in a new genre of
research termed “organ on a chip” research. Organ on a chip research involves employing
microfluidic devices in order to create crude reproductions of important organs, ideally
reducing the need for animal models. Several microfluidic organ mimics have been
accomplished, including the lung[41], kidney[42, 43], heart[44], and digestive system[45].
Importantly, all of these models join numerous previously developed microfluidic systems that
require a monolayer of cells. While the importance of verifying cell monolayer confluence prior
to experimentation has been well understood for decades[46], an alarmingly small number of
microfluidic platforms capable of performing such measurements have been developed,
especially considering that cell monolayers have been cultured in microfluidic devices for over a
decade.[47] Hypothetically, the lack of microfluidic devices capable of performing these simple
measurements is due to the engineering challenge of including the required components in a
microfluidic platform. Therefore, a simple method for including TEER measurement capabilities
in microfluidic experiments where cell monolayers are required would be paramount for

98

confirming the validity of experiments performed. We anticipate that the microfluidic TEER
device is capable of performing TEER measurements across cell barriers comprised of any cell,
in addition to bPAEC monolayers. The integration of a TEER measurement system would surely
aid in validating experiments performed in organ on a chip research and any research involving
cell monolayers cultured in microfluidic devices.

2.5 - Conclusions
Summarily, the work presented here describes the fabrication of a novel microfluidic
vascular mimic that can determine TEER of endothelial monolayers, providing a reliable and
sensitive method for monitoring cell barrier integrity. When performing microfluidic
permeability assays, this ability to measure cell monolayer barrier integrity is essentially
required to validate results by confirming proper barrier function of the cell layer. Importantly,
this novel device is the first example a microfluidic platform with integrated TEER capabilities
that also enables flow. Therefore, this device, in contrast to other similar devices, takes
advantage of desirable flow characteristics that are typical of microfluidic platforms, and that
have led to the widespread use of microfluidics. Furthermore, the flow of solutions through
microfluidic channels which are dimensionally similar to resistance vessels is easily
accomplished, yielding the first microfluidic vascular mimic with integrated TEER capabilities.
These desirable properties are unique to the microfluidic TEER system presented here and,
therefore, this device can be used to mimic the vasculature with unparalleled accuracy.
Verifying cell confluence prior to experimentation is crucial to validating results, and since
microfluidics have become a popular platform for cellular assays, the device presented here is a
powerful tool for providing credibility to microfluidic research involving cell monolayers. Since

99

the microfluidic TEER system is simply designed and fabricated, integration into other
microfluidic platforms used to study cell monolayers should be trivial. Additionally, using the
microfluidic TEER system is intuitive since experimental setup is minimal, and obtaining TEER
values is completely automated. Also, the device can potentially be integrated with microtiter
technology since the microfluidic wells are arranged in an array format, similar to a 96-well
plate. In this construct, conventional microtiter instrumentation can be used to perform high
throughput cell-based assays, while still taking advantage of the flow and TEER measurement
capabilities of the microfluidic device that are not permitted by microtiter platforms. This work
provides a simple method for ensuring proper cellular response in a vascular mimic, thus
increasing the value and accuracy of experimental results, including the results presented in
upcoming chapters. These benefits are particularly favorable when studying cell-cell
communication, cell monolayer permeation, or any other cellular response involving the
monolayer.

100

REFERENCES

101

REFERENCES
1.

Hediger, S., et al., Modular microsystem for epithelial cell culture and electrical
characterisation. Biosens Bioelectron, 2001. 16(9-12): p. 689-94.

2.

Hediger, S., et al., Biosystem for the culture and characterisation of epithelial cell tissues.
Sensors and Actuators B-Chemical, 2000. 63(1-2): p. 63-73.

3.

Lee, P.J., et al., Microfluidic System for Automated Cell-based Assays. JALA Charlottesv
Va, 2007. 12(6): p. 363-367.

4.

Barbulovic-Nad, I., S.H. Au, and A.R. Wheeler, A microfluidic platform for complete
mammalian cell culture. Lab Chip, 2010. 10(12): p. 1536-42.

5.

Tolan, N.V., et al., Personalized Metabolic Assessment of Erythrocytes Using Microfluidic
Delivery to an Array of Luminescent Wells. Analytical Chemistry, 2009. 81(8): p. 31023108.

6.

Yu, Z.T., et al., Integrated microfluidic devices for combinatorial cell-based assays.
Biomed Microdevices, 2009. 11(3): p. 547-55.

7.

Genes, L., I., et al., Adressing a Vascular Endothelium Array with Blood Components
using Underlying Microfluidic Channels. Lab on a Chip, 2007. 7(10): p. 1256-1259.

8.

Song, J.W., et al., Microfluidic endothelium for studying the intravascular adhesion of
metastatic breast cancer cells. PLoS ONE, 2009. 4(6): p. e5756.

9.

Halpin, S.T. and D.M. Spence, Direct Plate-Reader Measurement of Nitric Oxide Released
from Hypoxic Erythrocytes Flowing through a Microfluidic Device. Analytical Chemistry,
2010. 82(17): p. 7492-7497.

10.

Ku, C.J., T. D'Amico Oblak, and D.M. Spence, Interactions between multiple cell types in
parallel microfluidic channels: monitoring platelet adhesion to an endothelium in the
presence of an anti-adhesion drug. Anal Chem, 2008. 80(19): p. 7543-8.

11.

McClain, M.A., et al., Microfluidic devices for the high-throughput chemical analysis of
cells. Analytical Chemistry, 2003. 75(21): p. 5646-5655.

12.

O'Neill, A.T., N.A. Monteiro-Riviere, and G.M. Walker, Microfabricated curtains for
controlled cell seeding in high throughput microfluidic systems. Lab Chip, 2009. 9(12): p.
1756-62.

13.

White, A.K., et al., High-throughput microfluidic single-cell RT-qPCR. Proc Natl Acad Sci U
S A, 2011. 108(34): p. 13999-4004.

102

14.

Lecault, V., et al., High-throughput analysis of single hematopoietic stem cell
proliferation in microfluidic cell culture arrays. Nat Methods, 2011. 8(7): p. 581-6.

15.

Wong, K.H.K., et al., Microfluidic Models of Vascular Functions. Annual Review of
Biomedical Engineering, 2012. 14(1): p. null.

16.

Srigunapalan, S., et al., A microfluidic membrane device to mimic critical components of
the vascular microenvironment. Biomicrofluidics, 2011. 5(1): p. 13409.

17.

Kaihara, S., et al., Silicon micromachining to tissue engineer branched vascular channels
for liver fabrication. Tissue Eng, 2000. 6(2): p. 105-17.

18.

Lim, D., et al., Fabrication of microfluidic mixers and artificial vasculatures using a highbrightness diode-pumped Nd:YAG laser direct write method. Lab Chip, 2003. 3(4): p. 31823.

19.

Borenstein, J.T., et al., Microfabrication of three-dimensional engineered scaffolds.
Tissue Eng, 2007. 13(8): p. 1837-44.

20.

D'Amico Oblak, T., P. Root, and D.M. Spence, Fluorescence monitoring of ATPstimulated, endothelium-derived nitric oxide production in channels of a
poly(dimethylsiloxane)-based microfluidic device. Anal Chem, 2006. 78(9): p. 3193-7.

21.

Vogel, P.A., et al., Microfluidic transendothelial electrical resistance measurement device
that enables blood flow and postgrowth experiments. Anal Chem, 2011. 83(11): p. 4296301.

22.

Vogel, P.A., et al., Microfluidic Transendothelial Electrical Resistance Measurement
Device that Enables Blood Flow and Postgrowth Experiments. Analytical Chemistry,
2011. 83(11): p. 4296-4301.

23.

Douville, N.J., et al., Fabrication of two-layered channel system with embedded
electrodes to measure resistance across epithelial and endothelial barriers. Anal Chem,
2010. 82(6): p. 2505-11.

24.

Ferrell, N., et al., A microfluidic bioreactor with integrated transepithelial electrical
resistance (TEER) measurement electrodes for evaluation of renal epithelial cells.
Biotechnol Bioeng, 2010. 107(4): p. 707-16.

25.

Young, E.W., et al., Technique for real-time measurements of endothelial permeability in
a microfluidic membrane chip using laser-induced fluorescence detection. Anal Chem,
2010. 82(3): p. 808-16.

26.

Duffy, D.C., et al., Rapid prototyping of microfluidic systems in poly(dimethylsiloxane).
Analytical Chemistry, 1998. 70(23): p. 4974-4984.

103

27.

McDonald, J.C., et al., Fabrication of microfluidic systems in poly(dimethylsiloxane).
Electrophoresis, 2000. 21: p. 27-40.

28.

Booth, R. and H. Kim, Characterization of a microfluidic in vitro model of the blood-brain
barrier (muBBB). Lab Chip, 2012.

29.

Obradovic, B., et al., Gas exchange is essential for bioreactor cultivation of tissue
engineered cartilage. Biotechnol Bioeng, 1999. 63(2): p. 197-205.

30.

Qin, D., Y.N. Xia, and G.M. Whitesides, Rapid prototyping of complex structures with
feature sizes larger than 20 mu m. Advanced Materials, 1996. 8(11): p. 917-&.

31.

Johnson, D.E. and C.G. Enke, Bipolar Pulse Technique for Fast Conductance
Measurements. Analytical Chemistry, 1970. 42(3): p. 329-&.

32.

Caserta, K.J., et al., Computer controlled bipolar pulse conductivity system for
applications in chemical rate determinations. Anal. Chem., 1978. 50(11): p. 1534-41.

33.

Kralj, J.M., et al., Optical recording of action potentials in mammalian neurons using a
microbial rhodopsin. Nat Methods, 2012. 9(1): p. 90-5.

34.

McCaig, C.D., et al., Controlling cell behavior electrically: current views and future
potential. Physiol Rev, 2005. 85(3): p. 943-78.

35.

Lee, D.W. and Y.H. Cho, A continuous electrical cell lysis device using a low dc voltage for
a cell transport and rupture. Sensors and Actuators B-Chemical, 2007. 124(1): p. 84-89.

36.

Ali, M.H., et al., Endothelial permeability and IL-6 production during hypoxia: role of ROS
in signal transduction. Am J Physiol, 1999. 277(5 Pt 1): p. L1057-65.

37.

Wilhelm, I., C. Fazakas, and I.A. Krizbai, In vitro models of the blood-brain barrier. Acta
Neurobiol Exp (Wars), 2011. 71(1): p. 113-28.

38.

Adams, A.A., et al., Highly efficient circulating tumor cell isolation from whole blood and
label-free enumeration using polymer-based microfluidics with an integrated
conductivity sensor. Journal of the American Chemical Society, 2008. 130(27): p. 86338641.

39.

Letourneau, S., et al., Evaluating the effects of estradiol on endothelial nitric oxide
stimulated by erythrocyte-derived ATP using a microfluidic approach. Anal Bioanal
Chem, 2010. 397(8): p. 3369-75.

40.

Huang, J.H., et al., Rapid Fabrication of Bio-inspired 3D Microfluidic Vascular Networks.
Advanced Materials, 2009. 21(35): p. 3567-+.

104

41.

Huh, D., et al., Reconstituting organ-level lung functions on a chip. Science, 2010.
328(5986): p. 1662-8.

42.

Suwanpayak, N., et al., Blood cleaner on-chip design for artificial human kidney
manipulation. Int J Nanomedicine, 2011. 6: p. 957-64.

43.

Jang, K.J. and K.Y. Suh, A multi-layer microfluidic device for efficient culture and analysis
of renal tubular cells. Lab Chip, 2010. 10(1): p. 36-42.

44.

Grosberg, A., et al., Ensembles of engineered cardiac tissues for physiological and
pharmacological study: heart on a chip. Lab Chip, 2011. 11(24): p. 4165-73.

45.

Kim, H.J., et al., Human gut-on-a-chip inhabited by microbial flora that experiences
intestinal peristalsis-like motions and flow. Lab Chip, 2012. 12(12): p. 2165-74.

46.

Hart, M.N., et al., Differential opening of the brain endothelial barrier following
neutralization of the endothelial luminal anionic charge in vitro. Journal of
Neuropathology and Experimental Neurology, 1987. 46(2): p. 141-53.

47.

Patel, N., et al., Spatially controlled cell engineering on biodegradable polymer surfaces.
FASEB J, 1998. 12(14): p. 1447-54.

105

Chapter 3 - Monitoring Cell-Cell Communication in a Microfluidic-Based Vascular
Mimic
3.1 - Microfluidics to Mimic the Vasculature
As discussed in Chapter 1, vascular wall biology, and interactions with bloodstream
components, such as the red blood cell, is an area of continued interest due to the high
prevalence of cardiovascular disease and stroke. In order to investigate vascular wall biology
with an in vitro model, the layer of cells comprising the wall must be confluent, display a
restrictive paracellular pathway, high electrical resistance, physiologically relevant cell
architecture, and functional expression of transporter mechanisms while being easy to
culture.[1, 2] Additionally, a determination of cell confluence prior to experimentation is
important because cellular activity (i.e., responses to agonists/antagonists, transport of
molecules, etc.) has been shown to vary with cell confluence and integrity.[3-12] While these
criteria for an in vitro mimic of the vascular wall seem straightforward, reproducing the actual
environment found in vivo is quite challenging.
Most in vitro models reported in the literature are based on stationary cultures or cocultures (multiple cell types cultured together) to attempt to reproduce the intricate
environment of the true vascular barrier. Generally, isolated endothelial cells are cultured
within a microtiter insert upon a polycarbonate membrane that has been coated with an
adhesion molecule such as fibronectin or collagen to promote cell adhesion to the
membrane.[13] The cells are allowed to reach confluence, and the membrane is placed in a
diffusion cell where a cell-based assay is performed.[13] These methods are discussed in detail
within section 1.1.2. While this methodology has proven useful, the use of buffer in the

106

diffusion cell and lack of flow on the apical side of the model fail to adequately mimic the
microvasculature. In an even simpler experimental design, cell-based assays are commonly
performed on cells cultured within wells of a microtiter plate. These methods for studying cell
biology are adequate in many situations, resulting in their widespread use, however these
platforms again lack important properties required to accurately mimic the vasculature.
Endothelial cells cover all blood contacting areas of the cardiovascular system and, due to
their location, are constantly exposed to the hemodynamic shear stress of flowing blood. Shear
stress is known to regulate endothelial cell behavior, and is a determinant in the expression of
certain cellular responses.[14] It is also known that other cell types, such as the RBC, can alter
endothelial cell behavior and vice versa via cell-cell communication.[15-18] Therefore, an
accurate vascular mimic should include flow to account for shear stress, multiple cell types in
close proximity to account for cell-cell communication, as well as a method for determining cell
monolayer confluence to ensure proper cell behavior. Unfortunately, the most commonly used
aperture for measurements involving the integrity of a cell layer is the diffusion cell, which does
not account for the shear stress due to blood flow, or the presence of blood. The ability to
measure cell monolayer integrity, however, is easily accomplished in diffusion cells by
measuring TEER or performing permeation studies, which has become common practice.
The shortcomings of conventional vascular mimics are well addressed with microfluidic
systems, as it is possible to mimic the small distance between the blood vessel and the
endothelium.[19] Blood flow is also easily accomplished by incorporation of channels that have
diameters similar to resistance vessels (capillaries and arterioles) found in vivo.[19] For these
reasons, microfluidic systems have been employed for shear-related endothelial cell

107

experiments involving cell-cell communication, such as studies on migration[20], adhesion[15],
and shear-induced NO production[18]. The advantages of using microfluidic technologies to
model the vasculature have been realized as indicated by numerous reports of microfluidic
devices capable of mimicking the vasculature.[21, 22] However, the ability to measure cell
monolayer confluence on a microfluidic device was yet to be accomplished at the start of this
work. Very recently, however, microfluidic devices have been developed to monitor cell
monolayer permeation [23] and measure TEER [24, 25], illustrating the importance of such
devices.
The integrated microfluidic TEER system reported in Chapter 2 is capable of mimicking the
vasculature by enabling the flow of RBCs past an immobilized layer of endothelial cells whose
barrier integrity has been characterized by the integrated TEER system. This device is the first
microfluidic vascular mimic capable of determining cell monolayer confluence, and also the first
microfluidic device capable of measuring TEER that also enables flow.[26, 27] This device
provides a vascular mimic with multiple cell types separated by physiological distances, allowing
for cell-cell communication. Additionally, this device incorporates flowing blood to account for
shear stress, as well as a method for determining TEER to ensure cell monolayer confluence
prior to experimentation. Importantly, the TEER measurement capabilities of the device allow
for a confirmation of monolayer barrier integrity and confluence of cell layers cultured within
the wells of the device prior to experimentation. Measuring TEER to verify layer confluence and
barrier integrity prior to cellular investigations helps ensure physiologically appropriate cellular
responses of cultured cells, thus improving reproducibility between experiments and also
increasing the value and accuracy of experimental results, as well as conclusions stemming

108

from these results. These benefits are particularly favorable when studying cell-cell
communication, cell monolayer permeation, or any other cellular response involving the
monolayer. As such, cell monolayer confluence was confirmed prior to experiments carried out
during the course of this work.

3.2 - The Potential Role of C-Peptide in Vessel Dilation
C-peptide is a 31 amino acid peptide produced in the pancreas and released in equimolar
amounts with insulin as discussed in Chapter 1.[28, 29] In brief summary, for the first few
decades after its discovery in the late 1960's, C-peptide was thought to have minimal
physiological effects after release into the bloodstream.[30] However, numerous reports since
the mid 1990's have confirmed that C-peptide is a bioactive peptide that can have significant
beneficial effects in organs, tissue, and even cultured cells such as endothelial cells.[31-33]
Studies involving Type 1 diabetic subjects who are unable to naturally produce C-peptide have
shown that C-peptide improves nerve conduction and regeneration[34], renal function[35],
inflammation,[36] and blood flow[37, 38] via unknown mechanisms.

In vivo, the endothelium plays a vital role in regulating blood flow and pressure throughout
the entire vascular system. It is able to produce and release NO, a potent vasodilator that can
participate in the relaxation of the smooth muscle cells surrounding blood vessels.[31, 39-41] In
smooth muscle cells, NO produced by the endothelium stimulates the activation of soluble
guanylyl cyclase (sGC) that converts guanidine triphosphate (GTP) to cyclic guanidine
monophosphate (cGMP), resulting in relaxation of the smooth muscle and subsequent dilation
of the vessel.[42, 43] Indeed, if C-peptide is capable of stimulating endothelial NO production, it

109

could account for the C-peptide induced improvement in blood flow observed in Type 1
diabetic subjects.
A relationship between C-peptide and the enzyme responsible for regulating endothelial NO
concentration, namely endothelial nitric oxide synthase (eNOS), has been reported in many
different types of complications associated with diabetes, as well as in various tissues and
organs, both in vivo and in vitro.[44, 45] Reports of C-peptide-stimulated NO production in the
endothelium could have great importance, especially in the pulmonary endothelium, where the
flow of RBCs through the pulmonary bed allows re-oxygenation of the RBCs prior to delivery of
that oxygen to organs and tissues throughout the body. In accordance with the importance of
NO production, Sprague et al. has shown that ATP derived from RBCs stimulates NO production
in the pulmonary circulation.[46] Interestingly, C-peptide stimulates the release of ATP from
RBCs.[47] In this construct, C-peptide may have the ability to stimulate NO production in the
endothelium directly and indirectly, as mediated by C-peptide stimulated release of ATP from
RBCs via a mechanism proposed and elaborately discussed within Section 1.2.4. Importantly,
the potential involvement of RBC derived ATP in the proposed mechanism again highlights the
importance of flow, as flowing RBCs under the influence of shear stress release more ATP than
static RBCs.[48-51] Elucidating these potential mechanisms of cell-cell communication will be a
primary focus of the work presented in this chapter.

3.3 - Effects of Metal Binding on C-Peptide Bioactivity
Recently, our group has shown that if C-peptide is incubated in a solution containing
divalent metal ions prior to addition to a RBC solution, the resulting metal activated C-peptide
was able to facilitate glucose uptake into, and increase ATP release from RBCs.[47, 52]

110

Interestingly, these results could not be reproduced in the absence of a metal, indicating that Cpeptide must first be introduced to a metal source for some, if not all bioactivity. This was
further supported when mass spectrometric analysis of a solution containing active C-peptide
not deliberately introduced to a metal, revealed that the peptide had indeed formed an at least
one adduct with a metal.[47, 53] C-peptide has been shown to be active in the presence of
multiple metal ions[47, 52, 54, 55], however it has been shown that zinc is present at millimolar
levels within insulin secretory granules where C-peptide is produced and cleaved from
proinsulin.[56] Therefore, Zn

2+

is used to activate C-peptide throughout this work because

when C-peptide is produced in vivo, zinc is the most bioavailable metal source, increasing the
probability that zinc binds C-peptide for bioactivity in the body.

3.4 - Measuring Intracellular Nitric Oxide
The production and release of NO from bPAECs has been observed by the Spence group
frequently, by electrochemistry [57, 58] and through the use of the fluorogenic probes DAF-FM
DA[17-19, 59] and DAF-FM[60]. Electrochemistry is a useful technique for measuring NO
because it provides direct detection of the analyte as opposed to a derivative. However,
electrode fouling is often an issue in biological systems, which is compounded by the high
oxidizing potentials required for NO detection. Also, the electrochemical detection of NO is not
entirely selective, often resulting in the detection of peroxynitrite and nitrite in addition to NO.
Furthermore, the relatively poor detection limits that are typical for electrochemical techniques
could also be problematic. As a result of these shortcomings, electrochemistry was not chosen
as a method for measuring NO during this work.

111

The use of fluorogenic probes to measure NO is an indirect detection method often
requiring long derivatization times. However, a fluorogenic determination of NO provides the
significant advantage of detection by fluorescence microscopy or flow cytometry, which is
easily achieved in a common laboratory. Additionally, since NO can be measured optically by
this method, it is amenable to high throughput screening, an advantage easily taken advantage
of when coupled with microfluidic technology. Also, DAF-FM and DAF-FM DA have been
reported to have extremely low detection limits at around 3 nM.[61] It should also be noted
that DAF probes have been shown to be susceptible to reductants such as ascorbate, 2mercaptoethanol, and glutathione, thereby diminishing NO induced fluorescence.[62] However,
DAF-FM DA, a cell permeable probe, has been shown to produce emission that is highly
dependent on intracellular NO concentration, as it is not sensitive to O2-, NO2-, NO3-, H2O2, or
ONOO-, all of which are common interferences in most methods for determining NO.[63]
DAF-FM reacts with NO in the presence of oxygen to form triazole derivatives, which emit
green fluorescence similar to fluorescein as shown in Figure 3.1. This DAF probe is not cell
permeable, and therefore is used in our group to measure changes in NO release from cells.
Alternatively, DAF-FM DA has an identical structure to DAF-FM, with the addition of two
acetate groups as shown in Figure 3.2. The acetate groups of DAF-FM DA increase cell
permeability compared to DAF-FM, allowing DAF-FM DA to enter the cell where the acetate
groups are cleaved by intracellular esterases, rendering the molecule confined to the cell with a
structure identical to DAF-FM. The DAF-FM which has been confined to the cell can then react
with NO in the same fashion described previously, yielding fluorescence that is proportional to
the intracellular NO concentration. The DAF probe is confined to the cell in large quantities and,

112

therefore, any NO subsequently produced by the cell will react with the probe, enabling the
detection of intracellular NO production in response to a given stimulus.

113

DAF-FM
O

O

O

F

F
O

O
NH

H2N

O

DAF-FM

O

F

O

F
O

:N

O

+ O2

O

O
N

O
O

N
N

N

Figure 3.1 - Shown in the top of the image is the structure of DAF-FM, a fluorogenic probe for
NO. Below is a potential mechanism for the formation of fluorescent DAF-FM-NO. This
mechanism illustrates the reaction between NO and molecular oxygen to form a peroxynitrite
intermediate before forming DAF-FM-NO. DAF-FM-NO fluorescence is detectable at
wavelengths similar to fluorescein, with an excitation wavelength of 485 nm and emission of
515 nm.

114

Extracellular
H2N

Cell membrane

Intracellular

HN

O
O

F

F

Esterases

H2N

N

HN

O
AcO
OAc
O
DAF-FM DA - Non fluorescent
H2N

HN

O

O

O
DAF-FM

O

F

N

O

NO

O
O
O
DAF-FM - Weakly fluorescent

O
F

F

N

F

O

F

O
O
O
DAF-FM T - Fluorescent

F
O

Figure 3.2 - The structures of DAF-FM and DAF-FM DA are shown on the left, the difference
being the acetate groups of DAF-FM DA. As illustrated here, in contrast to DAF-FM, DAF-FM DA
is capable of penetrating the cell membrane to the intracellular space where esterases cleave
the acetate groups, which results in intracellular DAF-FM. Nitric oxide can then react with
intracellular DAF-FM, producing the fluorescent DAF-FM T in amounts proportional to
intracellular NO concentration.

115

DAF-FM DA was selected to determine whether endothelial NO production is altered by cell
2+

monolayer confluence, or by the actions of Zn -activated C-peptide in the bloodstream. As
2+

discussed previously, it is hypothesized that Zn -activated C-peptide can ultimately increase
endothelial NO production by increasing ATP release from RBCs via a mechanism which involves
cell-cell communication between the RBC and the endothelial cell. Since we could not predict
the magnitude of C-peptide induced endothelial NO production, DAF-FM DA was a desirable
choice for NO determination as a result of optimal detection limits that enable the observation
of even slight changes in NO production. Also, due to the permeability of PDMS to gas, NO can
diffuse in any direction from the endothelial monolayer and avoid detection by an extracellular
probe, so an intracellular probe was desired. Additionally, optical detection schemes easily
integrate with microfluidic technology, enabling simple measurements in a high-throughput
fashion using common laboratory instrumentation.

3.5 - Methods
3.5.1 - Device Fabrication
The microfluidic device developed during the work described in Chapter 2 was used to
accomplish the research presented throughout this chapter. Section 2.2.1 details the
fabrication of the microfluidic vascular mimic. All polycarbonate membranes had a pore size of
0.6 µm.

3.5.2 - bPAEC Culture
The methods for maintaining bPAEC cultures, as well as the subsequent seeding and culture of
these cells within microfluidic devices, are described in detail in section 2.2.3.

116

3.5.3 - TEER to Verify bPAEC Monolayer Confluence in Device
TEER measurements were performed routinely prior to flow experimentation to ensure
bPAEC monolayer confluence. Section 2.3.1 includes the method used for acquiring TEER
measurements, including the associated software and data manipulation. Cell monolayer
confluence was confirmed by multiple sustained TEER measurements.

3.5.4 - Sample Preparation
Human blood samples were obtained from healthy volunteers by venipuncture.

All

procedures were followed according to approved Michigan State IRB protocol. Briefly, whole
blood was collected from the donor into plastic blood collection vials (Becton, Dickinson and
Company), each containing 158 USP units of spray-dried lithium heparin to prevent coagulation.
The whole blood was then consolidated and centrifuged. After removal of the plasma and
buffy layer by aspiration, RBCs were washed 3 times with a physiological salt solution (PSS, pH
7.4) containing (all from Sigma-Aldrich): 4.7 mM KCl, 2.0 mM CaCl2, 140.5 mM NaCl, 12.0 mM
MgSO4, 21.0 mM tris(hydroxymethyl)aminomethane, 5.6 mM glucose, and 5% bovine serum
albumin.[47] Crude C-peptide (Genscript) was purified by HPLC and prepared in distilled and
deionized water (DDW, 18.2 MΩ), resulting in an 8.3 µM stock solution with is stored at 4°C.
2+

For samples containing Zn -activated C-peptide, a 2.0 µM working solution of human C2+

peptide was prepared in DDW and combined with the same volume of a 2.0 µM Zn solution
2+

also prepared in DDW.[47] For samples containing C-peptide without Zn , only C-peptide was
2+

added, followed by the same volume of DDW in the absence of Zn . Similarly, samples lacking
C-peptide contain only the desired volume of Zn

2+

117

solution with an equal volume of DDW. For

2+

samples lacking C-peptide and Zn , a volume of DDW was added equaling the volume of the
other samples, ensuring that the volume of DDW was constant throughout all samples used for
a given experiment. After the Zn

2+

and C-peptide solutions were allowed to rest for around 5

min, PSS was added to the solution, immediately followed by the appropriate volume of
washed RBCs, resulting in a solution of RBCs at 7% hematocrit. The appropriate volumes of 2.0
µM Zn

2+

and 2.0 µM C-peptide working solutions were added so that final concentrations in the

7% RBC samples were always 20 nM. For samples lacking RBCs, PSS was added in lieu of RBCs to
yield a 20 nM C-peptide solution.

3.5.5 - PPADS to Inhibit Purinergic Signaling
PPADS (pyridoxalphosphate-6-azophenyl-2',4'-disulfonic acid) (Sigma-Aldrich) is dissolved in
Hanks balanced salt solution (HBSS) for a 100 µM stock solution. The stock solution was stored
at -20°C as single use aliquots to prevent repetitive thawing. Aliquots were thawed at room
temperature immediately before use and diluted in HBSS for a 10 µM working solution.
Selected bPAECs were incubated with 10 µL of 10 µM PPADS solution for 30 minutes. After the
incubation period, the solution was removed and rinsed with HBSS multiple times to remove
color prior to the addition of DAF-FM DA. Care was taken to prevent light exposure by PPADS,
as it is a light sensitive molecule.

3.5.6 - Sample Introduction in Microfluidic Flow Experiments
Syringe pumps (Harvard Apparatus) are employed to drive 500 µL Hamilton Gastight
syringes for controlled flow of solution in the microfluidic channels. Tygon tubing is interfaced
to the syringes using an adapter connecting a male luer syringe to a female 1/16 in.

118

compression fitting (Upchurch/IDEX scientific). The Tygon tubes are interfaced with the
microfluidic device using 20 gauge stainless steel tubing (New England Small Tube Company),
which fits securely into both the Tygon tubing and the microfluidic inlets described in previously
in Section 2.2.1. RBC samples were prepared as described above and propelled via syringe
pump at a rate of 1.0 µL·minute-1 for 30 minutes in all flow experiments. The syringes were
rotated periodically throughout experiments to prevent RBC settling.

3.5.7 - Measuring NO Production of bPAECs Immobilized on Device
For measurements of intracellular NO production in endothelial cells immobilized in the
wells of a microfluidic device, the cell permeable fluorescent probe DAF-FM DA (Invitrogen) was
used. DAF-FM DA is dissolved in anhydrous dimethyl sulfoxide (DMSO) for a 5 mM stock
solution that could be stored at -20°C, but was often used immediately in experiments. A 100
µM working DAF-FM DA solution was prepared immediately before experimentation by diluting
stock solution in HBSS. 10 µL of DAF-FM DA working solution were added to each well and
allowed to incubate with bPAECs for 30 minutes. After incubation, the probe was removed and
replaced with phosphate buffered saline (PBS), prior to flow of the RBCs.
Fluorescence images of the bPAECs were obtained using an Olympus MVX10 microscope
(Olympus America Inc.) prior to flow, and again after flow completion, and the difference in
mean gray value analyzed. The difference in fluorescent signal is presented as the average gray
value intensities of the integrated areas taken from each acquired image. Images acquired
before and after flow for each well were analyzed sequentially so that the areas chosen for
both integrations were the same. Figure 3.3 displays an image of the microfluidic vascular
mimic during bPAEC NO determination via fluorescence microscopy.

119

Figure 3.3 - An integrated microfluidic TEER device is shown in (a) prior to use. Panel (b) shows
the microfluidic TEER device during a flow experiment prior to using fluorescence microscopy to
study confluent cells cultured in the wells of the device. An examination of intracellular NO
production of cells cultured in the wells of the microfluidic TEER device by fluorescence
microscopy is shown in (c).

120

3.6 - Results
3.6.1 - Meeting bPAEC Culture Requirements
The success of the work described here requires complete confluence of bPAEC
monolayers, as well as the observation of cell-cell communication, requiring proper and
reproducible behavior of the cultured cells. Due to these requirements, bPAEC culture was held
to high standards as shown in Figure 3.4. Prior to immobilization in a microfluidic device,
bPAECs cultured in tissue flasks were examined to confirm optimal confluence, proper
morphology, absence of bacteria, and minimal vacuoles, which is generally indicative of healthy
cells. Also, since bPAEC monolayers were cultured on a polycarbonate membrane, the cells
could not be imaged because the polycarbonate membrane is not optically clear, again
highlighting the advantages of TEER capabilities on a microfluidic device. TEER was used to
ensure confluence, but full coverage of the membrane by bPAECs was also often confirmed by
cell staining with a nuclear dye, Hoechst 33342. This dye binds to the nuclear material of viable
cells, allowing for visualization of the cells by optical microscopy when excited with 350 nm
light and resulting in emission at 461 nm. Cells were often stained with this dye in addition to
DAF-FM DA because their optical properties do not interfere. This method can be used to
confirm full coverage of the polycarbonate membrane by bPAECs, but is incapable of
confirming confluence as it does not allow for the confirmation of tight junctions between the
cells.

121

Figure 3.4 - A bright-field image of a monolayer of endothelial cells is shown. Cells were
cultured to a high degree of confluence and verified to have the appearance shown here prior
to seeding in microfluidic wells. During examination, the cells are checked for a cobblestone
appearance, a lack of vacuoles (generally indicates healthy cells), as well as a lack of any visible
bacteria.

122

3.6.2 - bPAEC Monolayer Confluence Influences NO Production
In order to display the importance of confirming cell monolayer confluence, a real
physiological event was monitored, namely intracellular NO production within bPAEC
monolayers of varying confluence as determined by the TEER system. Cells were allowed to
adhere for either 1.5 h or 8 h, and their barrier integrity, or lack thereof, was verified via TEER
measurements. After barrier integrity was confirmed, human RBCs were hydrodynamically
pumped through the channel underlying the endothelial cell-modified membrane for 30 min at
a flow rate of 1.0 µL·min-1. The movement of the RBCs through these channels has been shown
by our group to stimulate the release of ATP from the RBC.[19, 48] This released ATP is able to
migrate through the pores of the membrane and interact with the endothelial cell, stimulating
the resultant production of NO by the endothelium as shown in Figure 3.5.
Intracellular NO production was measured using DAF-FM DA as the fluorogenic intracellular
probe in conjunction with fluorescence microscopy. The NO production was measured in
endothelial cells that had not reached confluence (1.5 h after seeding on the membrane) and
for cells that were deemed confluent by the TEER measurement (8 h after seeding on the
membrane). The difference in NO production (as indicated by fluorescence emission intensity)
prior to pumping the RBCs to that 30 min after was measured. As shown in figure 3.5, those
cells that had reached confluence (8 h post-seeding) produced 34% more NO than those cells
that had not reached confluence (1.5 h post-seeding), a difference that is statistically significant
(p<0.05). Importantly, data not shown here also demonstrated that measurements of NO
production obtained at various points after confluence had been reached showed statistically
insignificant changes.

123

Figure 3.5 - A side view of the measurement system utilized to detect NO in the endothelial
layer is shown in (a). A solution of RBCs in PSS is pumped under a membrane containing an
endothelial layer cultured for 1.5 or 8 h. A significant increase (P < 0.05, n = 3, errors are SD) in
NO production is observed when the cells reached confluence at 8 h, as shown in the inset bar
graph.
124

3.6.3 - Mechanism Determination: C-peptide Stimulated Endothelial NO Production
Using the microfluidic vascular mimic, RBCs treated with 20 nM C-peptide and Zn
Zn + C-peptide), 20 nM C-peptide (RBC + C-peptide), and 20 nM Zn

2+

2+

(RBC +

(RBC + Zn) were pumped

beneath bPAECs cultured on a polycarbonate membrane as shown in Figure 3.6. PSS containing
20 nM C-peptide with Zn

2+

was also pumped beneath bPAECs to determine if the mechanism

leading to NO production is RBC-mediated. Fluorescence images were taken prior to flow, and
again after flow completion and the difference in mean gray value was analyzed.

The

difference in fluorescent signal is presented in Figure 3.6 as the average gray value intensities of
the integrated areas taken from each acquired image. It was observed that the endothelial
cells that were allowed to interact with RBCs treated with C-peptide and Zn

2+

had an 88.6 ±

7.5% increase in fluorescence intensity, indicative of NO production, when compared to
untreated RBCs. The RBCs treated with only C-peptide or only Zn

2+

did not generate statistical

change in intracellular NO production within the endothelial cells when compared to untreated
RBCs. Also, a statistically significant decrease in endothelial NO production was observed in the
absence of RBCs (p<0.02) (Buffer + C-peptide + Zn). To confirm the possibility that the increase
in endothelium NO production was due to ATP binding to the endothelial P2Y receptor, PPADS,
a purinergic receptor inhibitor was incubated with the endothelial cells for 30 min prior to flow
2+

of RBCs that had been incubated with Zn /C-peptide. This PPADS incubation resulted in a
complete reversal of the increase in NO production measured when comparing the flow of
2+

RBCs incubated with Zn /C-peptide to RBCs alone.

125

Figure 3.6 - A cross sectional view of a single channel of the microfluidic device utilized to
detect NO produced by the endothelial layer is shown (a). Images of wells of a microfluidic
device containing endothelial layers that had been exposed for 30 min. to untreated flowing
2+

RBCs, or flowing RBCs treated with Zn /C-peptide, are compared in (b). An 88.6% ± 7.5%
increase (n = 3, p < 0.001) in intracellular endothelial NO production was observed when RBCs
2+

were pumped in the presence of Zn /C-peptide and compared to RBCs alone in (b).
Interestingly, a statistical change in endothelial NO production was not observed when RBCs
exposed to C-peptide alone or Zn

2+

alone are compared to RBCs alone. A statistical decrease

(n=3, p<0.02) in endothelial NO production was observed when comparing PSS lacking RBCs to
RBCs alone. Additionally, when endothelial cells were incubated with PPADS, a purinergic
2+

receptor inhibitor, prior to flow of RBCs treated with Zn -activated C-peptide, an increase in
NO production was no longer observed. (* denotes p<0.001, ** denotes p<0.02)

126

3.7 – Discussion
As shown in figure 3.5, bPAECs that had reached confluence after 8 hours produced 34%
more NO than those cells that had not reached confluence (1.5 h post-seeding), a difference
that is statistically significant (p<0.05).

Also, data not shown also demonstrated that

measurements of NO production obtained at various points after confluence had been reached
showed statistically insignificant changes.

Coupled with the data in figure 3.5, the TEER

measurements showing no changes in barrier integrity after confluence was reached represent
the importance of TEER measurements prior to investigating the cellular activity of the layered
cells. For example, the increase in fluorescence intensity could be due to improved cell-cell
communication between the endothelium and the RBCs, altered endothelial cell behavior, or it
could simply be due to more cells in the membrane after a small period of growth. In all
likelihood, the behavior of the endothelial cells was altered, resulting in increased NO
production upon reaching a confluent state, since eNOS activity has been shown to be
associated with the presence of endothelial cell-cell contacts.[64] In any case, one seemingly
indisputable experimental advantage to making the TEER measurements prior to the cellular
investigation is having a quantitative measure of cell growth prior to addition of agonists or
antagonists. Such knowledge has the potential to increase reproducibility of cellular assays
performed on a specific cell type, as well as enhance precision between assays from different
cultures. Performing TEER measurements prior to experiments with agonists or antagonists
allows for more realistic, and in this case, enhanced cellular responses, allowing for more
accurate conclusions stemming from experimental results.

127

The flow capabilities of this device enable hemodynamic shear stress on RBC’s, creating a
more biologically relevant mimic of the vasculature than conventional platforms capable of
measuring TEER. Perhaps more beneficial is the ability to study cell to cell interactions and
communication in vitro, while maintaining an environment that geometrically resembles the
vasculature in vivo. In order to display the importance of performing TEER measurements prior
to cellular investigations, we chose to study a cellular response that changes as a function of
cell confluence, while showcasing the ability of the device to closely mimic and monitor
biological events, namely endothelial intracellular NO production in response to flowing RBCs.
However, we expect the ability to measure cellular barrier integrity, and relative changes in that
integrity on a microfluidic platform, to enable studies in a broad variety of fields. Drug
permeability studies, as well as elucidating the mechanisms of cellular barrier breakdown in
disease states are some examples of studies made possible by this device. Additionally, since
TEER measurements do not damage cells, or influence cell phenotype, this device could be used
to monitor the influence of agonists or antagonists on cell barrier integrity in real time.
After it was determined that the TEER capability of the device allowed for an increase in
reproducibility between cell assays and helped ensure proper cell behavior, we were prepared
to elucidate the mechanism of C-peptide induced endothelial NO production, which would help
clarify the observed improvements of blood flow in type 1 diabetic subjects since NO is a potent
vasodilator. The first observation suggesting that C-peptide stimulates NO release was that in
rat models of type 1 diabetes, C-peptide induced glucose utilization was altered by the
inhibition of endothelial nitric oxide synthase (eNOS) via the eNOS inhibitor L-NNA.[65] It has
also been demonstrated that C-peptide-mediated arteriolar vasorelaxation is dependent on

128

concentrations of NO in the vessel.[31, 66] Specifically, in the presence of insulin, C-peptide
induced a vasodilatory effect in isolated rat cremaster muscle arterioles that could be reversed
by L-NNA, again implicating eNOS in the mechanism.[31]
Subsequent to these findings, results from other studies have emerged that suggest a
relationship between C-peptide and NO production. For example, in diabetic polyneuropathy,
it has been shown that impaired expression of eNOS is the cause of decreased endoneurial
blood flow resulting from decreased amounts of NO in diabetic models, ultimately resulting in
slowed acute nerve conduction.[67-70] However, complete C-peptide replacement in these
models normalized blood flow and vascular conductance.[67] Importantly, C-peptide had no
effect on lipid peroxidation or superoxide dismutase activity, showing that the effects are not
due to reduced oxidative stress and, instead, may be due to a NOS-related mechanism. Also,
during the diabetic control and complications trial (DCCT), a relationship was established
between circulating C-peptide and risk for vascular complications[71], which is thought to be
primarily mediated by NO release from endothelial cells.[45, 72-74] Summarily, there is growing
evidence that C-peptide improves blood flow of diabetic subjects by an eNOS dependent
increase in NO production.
In previous microfluidic studies by our group, it was revealed that flowing RBCs are able to
stimulate endothelial NO production.[19] In a separate microfluidic study conducted by our
group, it was demonstrated that ATP could also stimulate endothelial NO production by a
process that could be inhibited by L-NAME, a competitive inhibitor of NOS.[18] This work builds
upon previous work demonstrating that the RBC itself can increase NO production by a process
requiring RBC derived ATP, implicating the RBC as a regulator of vascular tone.[75-77]

129

Additionally, it has been shown that patients with pulmonary hypertension have impaired ATP
release from their RBCs, further implicating RBC derived ATP in vasodilation.[78] Moreover, ATP
release from diabetic RBCs is significantly less than that released from RBCs obtained from
healthy, non-diabetic subjects.[79] However, C-peptide has been shown to normalize this
release.[47] Interestingly, our group has revealed the ability of C-peptide to increase ATP
release from RBCs, but only if C-peptide had first been carefully prepared in a solution
containing Zn

2+

prior to interaction with RBCs.[47, 52] Collectively, results from these studies

suggest that C-peptide, upon zinc activation, could increase ATP release from RBCs that, in turn,
could indirectly stimulate an increase in endothelial NO production.
The mechanism of ATP release from the RBC has been studied as discussed in Section 1.2.3.
Also, the mechanism of ATP induced endothelial NO production has been elucidated, and is also
described in depth in Chapter 1. Importantly, the purinergic P2Y receptor, located on the
endothelial cell membrane[80], influences vasodilation in response to an agonist by producing
smooth muscle relaxing factors, including NO, which seems to be the most potent.[41, 81-88]
This activity of the P2Y purinoreceptor can be achieved by extracellular ATP levels, as ATP is an
agonist of P2Y.[89, 90] Additionally, upon P2Y activation, an intracellular signaling cascade is
initiated that results in the activation of eNOS [91-93], which has already been shown to play a
role in C-peptide activity as discussed before. Understanding P2Y signaling is quite important
for this work because the P2Y receptor is responsible for sensing and responding to
extracellular signaling molecules, such as ATP, as is the case in the proposed mechanism for
metal activated C-peptide induced endothelial NO production. Stated simply, the P2Y
purinoreceptor is the proposed link between the RBC and the endothelium, and in order to

130

show that C-peptide can increase endothelial NO production as a result of increased ATP
release from the RBC, the receptor will have to be implicated in the mechanism experimentally.
The proposed mechanism for endothelial NO stimulation by C-peptide is RBC mediated, and
as such, is proposed to be an indirect, downstream consequence of C-peptide’s action in the
bloodstream. However, previous studies involving C-peptide application to cultured bovine
aortic endothelial cells (bPAECs) resulted in an increased NO production and eNOS mRNA
expression by an unknown mechanism not mediated by the RBC.[44] However, these results
could not be replicated in our lab by directly applying C-peptide solutions at varying
concentrations (physiologically relevant up to 10 times that amount) to cultured bPAECs. In
contrast, results using other known stimuli such as ATP or a calcium ionophore did stimulate a
significant production of NO, [44] thus providing evidence that the cells could produce NO
which we could successfully measure.
Based on the importance of NO production in the pulmonary bed, it was somewhat
surprising that C-peptide would evoke changes in aortic endothelium, but not in bPAECs.
Unlike other organs, the lungs must accept the entire cardiac output, requiring the blood
vessels in the lung bed to dilate and accept this output while simultaneously maintaining a
reasonable blood pressure. Consequentially, the ability of the endothelial cells in the
pulmonary bed to produce NO becomes especially important as all RBCs must return to the
lungs for reoxygenation. Not surprisingly, there have been many reported stimuli of NO in the
pulmonary endothelium, including but not limited to bradykinin, acetylcholine, prostacyclin,
and adenosine triphosphate (ATP).[18, 94, 95] Regardless of the inability to directly stimulate
NO production in bPAECs, C-peptide may be capable of indirect stimulation via its ability to

131

increase the release of ATP from RBCs as discussed previously. Of course, a system capable of
investigating such cell to cell communication would be requisite to answer such a question.
The microfluidic TEER system introduced in chapter 2 enables the flow of blood components
in a channel with dimensions mimicking a resistance vessel that is separated from a cultured
endothelium by physiologically relevant distances. Therefore, ATP released from the flowing
RBCs can diffuse to the endothelial layer, as it would in vivo. Here, RBCs in the presence and
2+

absence of C-peptide (with and without the addition of Zn ) were pumped underneath a
monolayer of bPAECs that had been incubated with DAF-FM DA for NO detection. In the
2+

presence of Zn -C-peptide, a significant increase (p<0.001) in NO production by the bPAECs
was measured as shown in Figure 3.6. When RBCs with C-peptide alone (or Zn

2+

alone) were

pumped underneath the bPAECs, there was no increase in NO production observed.
2+

Furthermore, when a PSS solution containing Zn -activated C-peptide, but lacking RBCs, was
pumped beneath bPAECs, a decrease in NO production was observed when compared to basal
2+

levels produced by RBCs alone (p<0.02), proving that the Zn -activated C-peptide induced NO
production is mediated by the RBC. These results also suggest that flowing RBCs play a role in
maintaining basal NO production in the bloodstream. Additionally, when bPAECs were
2+

incubated with PPADS to inhibit the P2 purinoreceptors, flowing RBCs containing Zn -activated
C-peptide could no longer cause an increase in NO production. Collectively, these results prove
that C-peptide increases endothelial NO production by stimulating the release of a known
endothelial P2Y agonist, ATP, from RBCs and thereby enhancing P2Y purinoreceptor activity by
increasing levels of ATP antagonist, which stimulates purinergic signaling cascade known to

132

result in increased endothelial NO production. Therefore, even though C-peptide may not be
able to stimulate NO directly in all cell types, its ability to stimulate the release of ATP from the
RBC may enable C-peptide to stimulate NO production in certain types of endothelial cells
through purinergic receptor signaling, just as it does in bPAECs.
C-peptide therapy has been shown to significantly reduce renal complications by increasing
glomerular hyperfiltration,[96] microalbuminuria[35] and eNOS expression.[97] C-peptide also
2+

increases the Ca
receptors

are

influx to the renal tubular cells.[98] It is well-established that renal P2

determinants

in

the

function

of

glomerular

filtration

and

renal

hemodynamics.[99] In light of the findings presented in Figure 3.6 showing the ability of Cpeptide to influence purinergic signaling in bPAECs, it is quite possible that C-peptide influences
renal function by purinergic signaling as well, potentially explaining the observed C-peptide
induced improvements in diabetic renal function.
The potential impact of C-peptide’s effect on ATP release and purinergic receptor-based
signaling reaches beyond the endothelium alone. For example, previous studies have shown
that C-peptide appears to have no direct effect on platelets.[100] To date, unpublished studies
in our group involving platelets and C-peptide interacting directly, with our without any metal
added, have resulted in similar conclusions. However, our group and others have recently
2+

reported that the P2X purinoreceptor on the platelet surface (an ATP-gated Ca

channel) may

be play a more crucial role in platelet function than originally perceived.[101, 102] In this
paradigm, the ability of C-peptide to stimulate ATP release from RBCs may indirectly affect
purinergic receptor signaling on the platelet, similar to that shown here for the
endothelium.[27] Indeed, recent reports have shown that RBC-derived ATP has a profound

133

effect on platelet adhesion to the endothelium, [15] and it is already well known that platelet
activation is a result of purinergic signaling.[103-105]
Another important aspect of the results presented here is the role of the metal in C-peptide
activity. It has been reported that chromium and iron have similar effects to zinc in rendering
C-peptide “bioactive”.[47] However, it is important to remember that the studies reported for
all three of these metals are more pharmacological than physiological as they were performed

in vitro where a metal was selected by researchers. Specifically, C-peptide is found in the
bloodstream at single-digit nanomolar levels, and bioavailable levels of chromium and iron are
sub-nanomolar concentrations as well.

Assuming the metal binds C-peptide, at these

concentrations, chromium, iron, or any other metal capable of binding to C-peptide’s
carboxylate-containing amino acids (glutamate and aspartate) could not outcompete sodium,
potassium and other metals found in the bloodstream found at concentrations 6 orders of
magnitude higher in the millimolar range. However, in the pancreatic beta cell granules where
C-peptide is produced, the concentration of zinc is known to be in the millimolar range.[56]
Therefore, it is possible that zinc would have the ability to bind to C-peptide after cleavage from
the proinsulin C-peptide hormone in the beta cell granules prior to excretion into the
bloodstream.
This work clearly shows that C-peptide is incapable of eliciting endothelial NO production
2+

unless Zn is introduced, indicating that C-peptide can only influence purinergic signaling in the
presence of a metal. Of course, most studies to date involving the exogenous addition of Cpeptide to humans, animal models, and cell models have not directly co-administered a metal.
Interestingly, some studies report no effects from C-peptide unless co-administered with

134

insulin, which often contains Zn

2+

at levels as high as 0.5% by mass.[106] Additionally, our

group has found by mass spectrometric analysis that C-peptide produced commercially often
contains metal contaminants. It is feasible that C-peptide samples often unintentionally contain
metals, frequently rendering C-peptide bioactive, which could explain the inconsistent activity
observed in clinical trials.

3.8 - Conclusions
In summary, the results presented here show that the degree of pulmonary endothelial
monolayer confluence is a determinant in NO production, where confluent cells produce more
NO relative to non-confluent cells. This finding highlights the importance of performing TEER
measurements prior to experimentation involving cell monolayers to ensure proper behavior,
which adds credibility to results and increases reproducibility between experiments.
Additionally, through a mechanism that has been expounded during this work, C-peptide is
identified as another factor influencing NO production in the pulmonary endothelium. It was
2+

determined that C-peptide, only in the presence of Zn , stimulates RBC mediated NO
production resulting from purinergic signaling in a cultured pulmonary endothelium. C-peptide
is able to increase ATP release from RBCs, and, as an agonist of the P2Y purinoreceptor, ATP is
capable of triggering a purinergic signaling cascade resulting in endothelial NO production.
While the involvement of C-peptide in purinergic signaling provides an explanation for the
improvements in blood flow observed in diabetic patients during C-peptide therapy, this
involvement may provide an explanation for the collective bioactivity of C-peptide observed
throughout the body. Additionally, the requirement of a metal for C-peptide activity could
explain inconsistencies in activity observed in C-peptide clinical trials. After all, the proper

135

treatment of C-peptide with Zn

2+

reproducibly increases RBC mediated NO production in

pulmonary endothelial cells cultured in vitro within the microfluidic vascular mimic. Due to the
well-established ability of NO derived from the endothelium to regulate vessel dilation in vivo,
2+

Zn -activated C-peptide can potentially relieve, or prevent, the common type 1 diabetic
ailments resulting from a compromised ability to properly regulate vessel tone and, therefore,
would be a promising drug candidate. Furthermore, C-peptide may be a determinant involved
in the fundamentally important role of maintaining appropriate vascular tone in healthy
individuals, a potential activity that has been previously overlooked.

136

REFERENCES

137

REFERENCES
1.

Gumbleton, M. and K.L. Audus, Progress and Limitations in the Use of In Vitro Cell
Cultures to Serve as a Permeability Screen for the Blood-Brain Barrier. J. of Pharm. Sci.,
2001. 90(11): p. 1681-1698.

2.

Crone, C. and S.P. Olesen, Electrical resistance of brain microvascular endothelium. Brain
Res FIELD Full Journal Title:Brain research, 1982. 241(1): p. 49-55.

3.

Witko-Sarsat, V., et al., Neutrophils: molecules, functions and pathophysiological
aspects. Laboratory Investigation, 2000. 80(5): p. 617-53.

4.

Mehta, D. and A.B. Malik, Signaling mechanisms regulating endothelial permeability.
Physiological Reviews, 2006. 86(1): p. 279-367.

5.

Staehelin, L.A., Further observations on the fine structure of freeze-cleaved tight
junctions. J Cell Sci, 1973. 13(3): p. 763-86.

6.

Wang, W., W.L. Dentler, and R.T. Borchardt, VEGF increases BMEC monolayer
permeability by affecting occludin expression and tight junction assembly. American
Journal of Physiology-Heart and Circulatory Physiology, 2001. 280(1): p. H434-H440.

7.

Madsen, K.L., et al., Interleukin 10 prevents cytokine-induced disruption of T84
monolayer barrier integrity and limits chloride secretion. Gastroenterology, 1997.
113(1): p. 151-9.

8.

Xu, M., et al., Ethanol disrupts vascular endothelial barrier: implication in cancer
metastasis. Toxicol Sci, 2012. 127(1): p. 42-53.

9.

Nagaraja, S., A. Kapela, and N.M. Tsoukias, Intercellular communication in the vascular
wall: A modeling perspective. Microcirculation, 2012.

10.

Atencia, J. and D.J. Beebe, Controlled microfluidic interfaces. Nature, 2005. 437(7059): p.
648-55.

11.

Kuczenski, B., et al., Probing cellular dynamics with a chemical signal generator. PLoS
ONE, 2009. 4(3): p. e4847.

12.

Ayliffe, H.E., A.B. Frazier, and R.D. Rabbitt, Electric impedance spectroscopy using
microchannels with integrated metal electrodes. Journal of Microelectromechanical
Systems, 1999. 8(1): p. 50-57.

13.

Borchardt, R.T., et al., Models for Assessing Drug Absorption and Metabolism. [In:
Pharm. Biotechnol., 1996; 8]1996. 444 pp.

138

14.

Davies, P.F., Flow-mediated endothelial mechanotransduction. Physiological Reviews,
1995. 75(3): p. 519-60.

15.

Ku, C.J., T. D'Amico Oblak, and D.M. Spence, Interactions between multiple cell types in
parallel microfluidic channels: monitoring platelet adhesion to an endothelium in the
presence of an anti-adhesion drug. Anal Chem, 2008. 80(19): p. 7543-8.

16.

Tissot Van Patot, M.C., et al., Endotoxin-induced adhesion of red blood cells to
pulmonary artery endothelial cells. Am J Physiol, 1996. 270(1 Pt 1): p. L28-36.

17.

Letourneau, S., et al., Evaluating the effects of estradiol on endothelial nitric oxide
stimulated by erythrocyte-derived ATP using a microfluidic approach. Anal Bioanal
Chem, 2010. 397(8): p. 3369-75.

18.

D'Amico Oblak, T., P. Root, and D.M. Spence, Fluorescence monitoring of ATPstimulated, endothelium-derived nitric oxide production in channels of a
poly(dimethylsiloxane)-based microfluidic device. Anal Chem, 2006. 78(9): p. 3193-7.

19.

Genes, L., I., et al., Adressing a Vascular Endothelium Array with Blood Components
using Underlying Microfluidic Channels. Lab on a Chip, 2007. 7(10): p. 1256-1259.

20.

Shamloo, A., et al., Endothelial cell polarization and chemotaxis in a microfluidic device.
Lab on a Chip, 2008. 8(8): p. 1292-1299.

21.

van der Meer, A.D., et al., Microfluidic technology in vascular research. J Biomed
Biotechnol, 2009. 2009: p. 823148.

22.

Wong, K.H., et al., Microfluidic Models of Vascular Functions. Annu Rev Biomed Eng,
2012.

23.

Young, E.W., et al., Technique for real-time measurements of endothelial permeability in
a microfluidic membrane chip using laser-induced fluorescence detection. Anal Chem,
2010. 82(3): p. 808-16.

24.

Douville, N.J., et al., Fabrication of two-layered channel system with embedded
electrodes to measure resistance across epithelial and endothelial barriers. Anal Chem,
2010. 82(6): p. 2505-11.

25.

Ferrell, N., et al., A microfluidic bioreactor with integrated transepithelial electrical
resistance (TEER) measurement electrodes for evaluation of renal epithelial cells.
Biotechnol Bioeng, 2010. 107(4): p. 707-16.

26.

Vogel, P.A., et al., Microfluidic transendothelial electrical resistance measurement device
that enables blood flow and postgrowth experiments. Anal Chem, 2011. 83(11): p. 4296301.

139

27.

Halpin, S.T., et al., The Red Blood Cell and Nitric Oxide: Derived, Stimulated, or Both? The
Open Nitric Oxide Journal, 2011. 3(Suppl 1-M2): p. 8-15.

28.

Steiner, D.F., et al., Proinsulin and the biosynthesis of insulin. Recent Prog Horm Res,
1969. 25: p. 207-82.

29.

Rubenstein, A.H., et al., Proinsulin and C-peptide in blood. Diabetes, 1972. 21(2 Suppl):
p. 661-72.

30.

Hoogwerf, B.J., et al., Infusion of synthetic human C-peptide does not affect plasma
glucose, serum insulin, or plasma glucagon in healthy subjects. Metabolism, Clinical and
Experimental, 1986. 35(2): p. 122-5.

31.

Jensen, M.E. and E.J. Messina, C-peptide induces a concentration-dependent dilation of
skeletal muscle arterioles only in presence of insulin. Am J Physiol, 1999. 276(4 Pt 2): p.
H1223-8.

32.

Scalia, R., et al., C-peptide inhibits leukocyte-endothelium interaction in the
microcirculation during acute endothelial dysfunction. FASEB J, 2000. 14(14): p. 2357-64.

33.

Young, L.H., et al., C-peptide exerts cardioprotective effects in myocardial ischemiareperfusion. Am J Physiol Heart Circ Physiol, 2000. 279(4): p. H1453-9.

34.

Pierson, C.R., W. Zhang, and A.A. Sima, Proinsulin C-peptide replacement in type 1
diabetic BB/Wor-rats prevents deficits in nerve fiber regeneration. Journal of
Neuropathology and Experimental Neurology, 2003. 62(7): p. 765-79.

35.

Maezawa, Y., et al., Influence of C-peptide on early glomerular changes in diabetic mice.
Diabetes/metabolism research and reviews, 2006. 22(4): p. 313-22.

36.

Sima, A.A., et al., Inflammation in Diabetic Encephalopathy is Prevented by C-Peptide.
Rev Diabet Stud, 2009. 6(1): p. 37-42.

37.

Lindstrom, K., et al., Acute effects of C-peptide on the microvasculature of isolated
perfused skeletal muscles and kidneys in rat. Acta Physiol Scand, 1996. 156(1): p. 19-25.

38.

Hansen, A., et al., C-peptide exerts beneficial effects on myocardial blood flow and
function in patients with type 1 diabetes. Diabetes, 2002. 51(10): p. 3077-3082.

39.

Ignarro, L.J., et al., Endothelium-derived relaxing factor from pulmonary artery and vein
possesses pharmacologic and chemical properties identical to those of nitric oxide
radical. Circulation Research, 1987. 61(6): p. 866-79.

40.

Furchgott, R.F. and J.V. Zawadzki, The Obligatory Role of Endothelial-Cells in the
Relaxation of Arterial Smooth-Muscle by Acetylcholine. Nature, 1980. 288(5789): p. 373376.

140

41.

Palmer, R.M., A.G. Ferrige, and S. Moncada, Nitric oxide release accounts for the
biological activity of endothelium-derived relaxing factor. Nature, 1987. 327(6122): p.
524-6.

42.

Ignarro, L.J., et al., Nitric oxide and cyclic GMP formation upon electrical field stimulation
cause relaxation of corpus cavernosum smooth muscle. Biochem Biophys Res Commun,
1990. 170(2): p. 843-50.

43.

Rapoport, R.M. and F. Murad, Agonist-induced endothelium-dependent relaxation in rat
thoracic aorta may be mediated through cGMP. Circ Res, 1983. 52(3): p. 352-7.

44.

Wallerath, T., et al., Stimulation of endothelial nitric oxide synthase by proinsulin Cpeptide. Nitric Oxide, 2003. 9(2): p. 95-102.

45.

Kitamura, T., et al., Proinsulin C-peptide activates cAMP response element-binding
proteins through the p38 mitogen-activated protein kinase pathway in mouse lung
capillary endothelial cells. Biochemical Journal, 2002. 366(3): p. 737-744.

46.

Sprague, R.S., et al., Extracellular ATP signaling in the rabbit lung: erythrocytes as
determinants of vascular resistance. Am J Physiol Heart Circ Physiol, 2003. 285(2): p.
H693-700.

47.

Meyer, J.A., et al., Metal-activated C-peptide facilitates glucose clearance and the
release of a nitric oxide stimulus via the GLUT1 transporter. Diabetologia, 2008. 51(1): p.
175-82.

48.

Sprung, R.J., R.S. Sprague, and D.M. Spence, Determination of ATP release from
erythrocytes using microbore tubing as a model of resistance vessels in vivo. Anal.
Chem., 2002. 74: p. 2274-2278.

49.

Sprague, R.S., et al., Deformation-induced ATP release from red blood cells requires
cystic fibrosis transmembrane conductance regulator activity. Am. J. Physiol, 1998. 275:
p. H1726-H1732.

50.

Price, A., K., et al., Deformation-induced release of ATP from erythrocytes in a
poly(dimethylsiloxane)-based microchip with channels that mimic resistance vessels.
Analytical Chemistry, 2004. 76(16): p. 4849-55.

51.

Wan, J., W.D. Ristenpart, and H.A. Stone, Dynamics of shear-induced ATP release from
red blood cells. Proc Natl Acad Sci U S A, 2008. 105(43): p. 16432-7.

52.

Meyer, J.A., et al., Zinc-activated C-peptide resistance to the type 2 diabetic erythrocyte
is associated with hyperglycemia-induced phosphatidylserine externalization and
reversed by metformin. Mol Biosyst, 2009. 5(10): p. 1157-62.

141

53.

Keltner, Z., et al., Mass spectrometric characterization and activity of zinc-activated
proinsulin C-peptide and C-peptide mutants. Analyst, 2010. 135(2): p. 278-288.

54.

Brooks, P., J. Meyer, and D.M. Spence, Metal (Zn2+) Activated C-Peptide Induced ATP
Release Via Red Blood Cells with the Addition of Albumin. Abstracts, 60th Southeast
Regional Meeting of the American Chemical Society, Nashville, TN, United States,
November 12-15, 2008: p. SERM-680.

55.

Medawala, W., et al., A Molecular Level Understanding of Zinc Activation of C-peptide
and its Effects on Cellular Communication in the Bloodstream. Rev Diabet Stud, 2009.
6(3): p. 148-58.

56.

Hutton, J.C., The insulin secretory granule. Diabetologia, 1989. 32(5): p. 271-81.

57.

Hulvey, M.K. and R.S. Martin, A microchip-based endothelium mimic utilizing open
reservoirs for cell immobilization and integrated carbon ink microelectrodes for
detection. Anal. Bioanal. Chem., 2009. 393(2): p. 599-605.

58.

Hulvey, M.K., et al., Fabrication and evaluation of a 3-dimensional microchip device
where carbon microelectrodes individually address channels in the separate fluidic
layers. Analyst, 2007. 132(12): p. 1246-1253.

59.

Ku, C.-J., et al., Fluorescence Determination of Nitric Oxide Production in Stimulated and
Activated Platelets. Analytical Chemistry, 2007. 79(6): p. 2421-2426.

60.

Halpin, S.T. and D.M. Spence, Direct Plate-Reader Measurement of Nitric Oxide Released
from Hypoxic Erythrocytes Flowing through a Microfluidic Device. Analytical Chemistry,
2010. 82(17): p. 7492-7497.

61.

Itoh, Y., et al., Determination and bioimaging method for nitric oxide in biological
specimens by diaminofluorescein fluorometry. Anal Biochem, 2000. 287(2): p. 203-9.

62.

Nagata, N., K. Momose, and Y. Ishida, Inhibitory effects of catecholamines and antioxidants on the fluorescence reaction of 4,5-diaminofluorescein, DAF-2, a novel indicator
of nitric oxide. J Biochem, 1999. 125(4): p. 658-61.

63.

Kojima, H., et al., Direct evidence of NO production in rat hippocampus and cortex using
a new fluorescent indicator: DAF-2 DA. NeuroReport, 1998. 9(15): p. 3345-8.

64.

Govers, R., et al., Endothelial nitric oxide synthase activity is linked to its presence at cellcell contacts. The Biochemical journal, 2002. 361(Pt 2): p. 193-201.

65.

Li, L., et al., Rat C peptide I and II stimulate glucose utilization in STZ-induced diabetic
rats. Diabetologia, 1999. 42(8): p. 958-964.

142

66.

Joshua, I.G., et al., Mechanisms of endothelial dysfunction with development of type 1
diabetes mellitus: role of insulin and C-peptide. J Cell Biochem, 2005. 96(6): p. 1149-56.

67.

Stevens, M.J., et al., C-peptide corrects endoneurial blood flow but not oxidative stress in
type 1 BB/Wor rats. Am. J. Physiol. Endocrinol. Metab., 2004. 287(3): p. E497-505.

68.

Cotter, M.A., et al., Effects of proinsulin C-peptide in experimental diabetic neuropathy:
vascular actions and modulation by nitric oxide synthase inhibition. Diabetes, 2003.
52(7): p. 1812-7.

69.

Stevens, M.J., et al., The linked roles of nitric oxide, aldose reductase and, (Na+,K+)ATPase in the slowing of nerve conduction in the streptozotocin diabetic rat. The Journal
of clinical investigation, 1994. 94(2): p. 853-9.

70.

Kitamura, T., et al., Proinsulin C-peptide increases nitric oxide production by enhancing
mitogen-activated protein-kinase-dependent transcription of endothelial nitric oxide
synthase in aortic endothelial cells of Wistar rats. Diabetologia, 2003. 46(12): p. 1698705.

71.

Steffes, M.W., et al., Beta-cell function and the development of diabetes-related
complications in the diabetes control and complications trial. Diabetes Care, 2003. 26(3):
p. 832-6.

72.

Delaney, C., J. Shaw, and T. Day, Acute, local effects of iontophoresed insulin and Cpeptide on cutaneous microvascular function in Type 1 diabetes mellitus. Diabetic
medicine a journal of the British Diabetic Association, 2004. 21(5): p. 428-33.

73.

Forst, T., et al., Effects of proinsulin C-peptide on nitric oxide, microvascular blood flow
and erythrocyte Na+,K+-ATPase activity in diabetes mellitus type I. Clin Sci (Lond), 2000.
98(3): p. 283-90.

74.

Forst, T., et al., Biological activity of C-peptide on the skin microcirculation in patients
with insulin-dependent diabetes mellitus. J. Clin. Invest., 1998. 101(10): p. 2036-41.

75.

Sprague, R.S., et al., ATP: the red blood cell link to NO and local control of the pulmonary
circulation. Am. J. Physiol, 1996. 271: p. H2717-H2722.

76.

Ellsworth, M.L., et al., The erythrocyte as a regulator of vascular tone. Am. J. Physiol,
1995. 269: p. H2155-H2161.

77.

Dietrich, H.H., et al., Red blood cell regulation of microvascular tone through adenosine
triphosphate. Am. J. Physiol, 2000. 278: p. H1294-H1298.

78.

Sprague, R.S., et al., Impaired release of ATP from red blood cells of humans with
primary pulmonary hypertension. Experimental Biology and Medicine, 2001. 226(5): p.
434-439.

143

79.

Carroll, J., et al., An altered oxidant defense system in red blood cells affects their ability
to release nitric oxide-stimulating ATP. Mol Biosyst, 2006. 2(6-7): p. 305-11.

80.

Olsson, R.A. and J.D. Pearson, Cardiovascular purinoceptors. Physiol Rev, 1990. 70(3): p.
761-845.

81.

Saenz de Tejada, I., et al., Cholinergic neurotransmission in human corpus cavernosum. I.
Responses of isolated tissue. Am J Physiol, 1988. 254(3 Pt 2): p. H459-67.

82.

Kimoto, Y., R. Kessler, and C.E. Constantinou, Endothelium dependent relaxation of
human corpus cavernosum by bradykinin. J Urol, 1990. 144(4): p. 1015-7.

83.

Kim, N., et al., A nitric oxide-like factor mediates nonadrenergic-noncholinergic
neurogenic relaxation of penile corpus cavernosum smooth muscle. J Clin Invest, 1991.
88(1): p. 112-8.

84.

Azadzoi, K.M. and I. Saenz de Tejada, Diabetes mellitus impairs neurogenic and
endothelium-dependent relaxation of rabbit corpus cavernosum smooth muscle. J Urol,
1992. 148(5): p. 1587-91.

85.

Burnett, A.L., et al., Nitric oxide: a physiologic mediator of penile erection. Science, 1992.
257(5068): p. 401-3.

86.

Trigo-Rocha, F., et al., Nitric oxide and cGMP: mediators of pelvic nerve-stimulated
erection in dogs. Am J Physiol, 1993. 264(2 Pt 2): p. H419-22.

87.

Saiag, B., et al., Study of the mechanisms involved in adenosine-5'-O-(2-thiodiphosphate)
induced relaxation of rat thoracic aorta and pancreatic vascular bed. Br J Pharmacol,
1996. 118(3): p. 804-10.

88.

Shalev, M., et al., Stimulation of P2y purinoceptors induces, via nitric oxide production,
endothelium-dependent relaxation of human isolated corpus cavernosum. J Urol, 1999.
161(3): p. 955-9.

89.

Needham, L., et al., Characteristics of the P2 purinoceptor that mediates prostacyclin
production by pig aortic endothelial cells. Eur J Pharmacol, 1987. 134(2): p. 199-209.

90.

Mitchell, J.A., et al., Different patterns of release of endothelium-derived relaxing factor
and prostacyclin. Br J Pharmacol, 1992. 105(2): p. 485-9.

91.

Kalinowski, L., et al., Third-generation beta-blockers stimulate nitric oxide release from
endothelial cells through ATP efflux: a novel mechanism for antihypertensive action.
Circulation, 2003. 107(21): p. 2747-52.

92.

da Silva, C.G., et al., Mechanism of purinergic activation of endothelial nitric oxide
synthase in endothelial cells. Circulation, 2009. 119(6): p. 871-9.

144

93.

Xu, H.L., et al., Agonist-specific differences in mechanisms mediating eNOS-dependent
pial arteriolar dilation in rats. American journal of physiology. Heart and circulatory
physiology, 2002. 282(1): p. H237-43.

94.

Chataigneau, T., et al., Acetylcholine-induced relaxation in blood vessels from endothelial
nitric oxide synthase knockout mice. Br J Pharmacol, 1999. 126(1): p. 219-26.

95.

Osanai, T., et al., Cross talk of shear-induced production of prostacyclin and nitric oxide
in endothelial cells. Am J Physiol Heart Circ Physiol, 2000. 278(1): p. H233-8.

96.

Samnegard, B., et al., Effects of C-peptide on glomerular and renal size and renal
function in diabetic rats. Kidney Int, 2001. 60(4): p. 1258-65.

97.

Kamikawa, A., et al., Proinsulin C-peptide abrogates type-1 diabetes-induced increase of
renal endothelial nitric oxide synthase in rats. Diabetes Metab Res Rev, 2008. 24(4): p.
331-8.

98.

Ohtomo, Y., et al., Differential effects of proinsulin C-peptide fragments on Na+, K+ATPase activity of renal tubule segments. Diabetologia, 1998. 41(3): p. 287-91.

99.

Graciano, M.L., et al., Purinergic receptors contribute to early mesangial cell
transformation and renal vessel hypertrophy during angiotensin II-induced hypertension.
Am J Physiol Renal Physiol, 2008. 294(1): p. F161-9.

100.

Ekberg, K., et al., C-Peptide replacement therapy and sensory nerve function in type 1
diabetic neuropathy. Diabetes Care, 2007. 30(1): p. 71-6.

101.

Karunarathne, W., C.J. Ku, and D.M. Spence, The dual nature of extracellular ATP as a
concentration-dependent platelet P2X1 agonist and antagonist. Integr Biol (Camb),
2009. 1(11-12): p. 655-63.

102.

Anderson, K.B., W. Karunarathne, and D.M. Spence, Measuring P2X1 receptor activity in
washed platelets in the absence of exogenous apyrase. Analytical Methods, 2012. 4(1):
p. 101-105.

103.

Clemetson, K.J., Platelet activation: signal transduction via membrane receptors.
Thrombosis and haemostasis, 1995. 74(1): p. 111-6.

104.

Heemskerk, J.W., E.M. Bevers, and T. Lindhout, Platelet activation and blood
coagulation. Thrombosis and haemostasis, 2002. 88(2): p. 186-93.

105.

Dorsam, R.T. and S.P. Kunapuli, Central role of the P2Y12 receptor in platelet activation. J
Clin Invest, 2004. 113(3): p. 340-5.

106.

Shafqat, J., et al., Proinsulin C-peptide elicits disaggregation of insulin resulting in
enhanced physiological insulin effects. Cell Mol Life Sci, 2006. 63(15): p. 1805-11.

145

Chapter 4 - Determining Endothelial Permeability to C-peptide
4.1 - Biological Activity of C-peptide: Direct and/or Indirect Mechanisms
C-peptide has been shown to act in vivo to maintain general health, having influence in
various organs and throughout the body. This is evident by C-peptide induced improvements in
type 1 diabetic patients who are incapable of producing endogenous C-peptide, including
improved nerve conduction and regeneration[1], renal function[2], inflammation,[3] and blood
flow.[4] In spite of the importance and scope of these physiological actions, until the work that
was presented in Chapter 3 elucidating a mechanism by which C-peptide can improve blood
flow, the mechanisms by which C-peptide exerts its beneficial effects in vivo were largely
unknown. Work presented in Chapter 3 describes how C-peptide is capable of indirectly
stimulating NO production in the endothelium via its ability to enhance ATP release from the
red blood cell. NO is unique in that it was the first gaseous signaling molecule, inducing vessel
dilation by diffusing to the smooth muscle cells that surround blood vessels. C-peptide is
secreted into the bloodstream with insulin, however it has never been determined if, like
insulin, C-peptide can leave the bloodstream. The work presented in Chapter 3 shows that Cpeptide is capable of influencing cell-cell communication and purinergic signaling, allowing Cpeptide to expand its influence beyond the bloodstream. However, whether the entirety of Cpeptide stimulated bioactivity is a result from its activity in the bloodstream, or if particular
activity is a result of direct stimulation by C-peptide in tissue exterior of the bloodstream is still
indefinite. In order to potentially influence biological functions directly, the ability of C-peptide
to permeate the endothelium to exit the bloodstream will be of obvious importance.

146

It has been previously shown that C-peptide is able to specifically bind to endothelial cells,
which is saturated concentrations of ~ 1 nM, which is of importance since the transport of
macromolecules is generally transcellular and specific.[5] This specific binding, however, has
not yet been shown by conventional radioligand binding. Currently, a specific receptor for Cpeptide has not been identified as the alleged binding structure is elusive. Furthermore, the
homology pattern of C-peptide is markedly different from most signaling peptides.[6] However,
C-peptide has been shown to be internalized by endosomes to enter human endothelial
cells.[7]

Additionally,

this

internalization

can

be

inhibited

by

nocodazole

and

monodansylcadaverine (MDC), signifying that C-peptide enters the cell by a mechanism which is
receptor mediated (clathrin) and requires microtubule assembly. Interestingly, microtubules
have been shown to facilitate transcellular transport and influence endothelial permeability and
integrity.[8-11] Additionally, some endothelial cell types have been shown to express PEPT
membrane proteins, which are proteins that promiscuously facilitate the transcellular transport
of peptides.[12, 13] These results mutually suggest that C-peptide can potentially leave the
bloodstream and directly stimulate various cell types exterior of the cardiovascular system. In
addition to human endothelial cells, C-peptide can also enter fibroblasts and human embryonic
kidney cells where it is capable of influencing gene transcription,[14, 15] suggesting that Cpeptide is capable of directly influencing cells exogenous of the bloodstream and again
demonstrating the ability of C-peptide to permeate cellular membranes. The ability or inability
of C-peptide to permeate a cultured endothelium can provide insight as to whether some
biological effects of the peptide are initiated directly, or if all of these effects are a result of
activity in the bloodstream.

147

4.2 - Enzyme-Linked Immunosorbent Assay (ELISA) for C-peptide Detection
ELISA is a powerful method for the detection and quantification of proteins or peptides in
complex biological samples. As shown in Figure 4.1, antibodies are used to bind a particular
protein or peptide antigen that is of interest, allowing for their detection with great specificity.
Briefly, at least one antibody with specificity for a particular antigen of interest is immobilized
at the bottom of a well of a microtiter plate. A sample with an unknown amount of antigen is
then added to the well where binding occurs to the antibody. After binding for a specific
duration of time, the remaining sample is removed, leaving the antigen of interest bound to the
antibody, thereby extracting it from often complex biological samples. In order to detect the
antigen, generally a secondary antibody that has been bioconjugated to an enzyme is added,
and the secondary antibody specifically binds the immobilized antigen. Subsequently, an
enzymatic substrate is added to produce a detectable signal such as color change, which is
proportional to the quantity of antigen in the sample. Described here is a sandwich ELISA,
however other ELISA methods with similar general principles to exist. Importantly, ELISA not
only detects proteins and peptides with high specificity, impressive lower detection limits are
also commonly achieved. For example, detection limits as low as 200 pg/mL have been
reported.[16] Since C-peptide exists at low nanomolar levels in vivo, ELISA is a suitable
technique for the detection and quantification of C-peptide in complex physiological samples
resembling those found in biological systems.

148

Figure 4.1 - Shown is a diagram illustrating C-peptide detection by sandwich ELISA. Complex
biological samples containing C-peptide are added to wells of a 96-well plate, the bottoms of
which have been coated with a primary antibody for human C-peptide. C-peptide is allowed to
bind to the antibody for a period of time, after which a secondary antibody is added to the
solution, which also binds to sandwich C-peptide between the two antibodies. Following
incubation, an enzyme is added that binds to the secondary antibody and, subsequently, a
substrate is added that allows the catalytic production of a detectable substance by the
enzyme. In the case of C-peptide ELISA, this reaction yields a product which changes the color
of the solution during the enzymatic process, which is stopped after an interval of time, causing
the amount of color produced to be proportional to C-peptide concentration. The absorbance
of the solution is read at 450 nm using a multiwell plate reader, allowing for quantification of Cpeptide in complex biological samples.

149

4.3 - Methods
4.3.1 - Bovine Pulmonary Artery Endothelial Cell Culture
General methods for bPAEC culture are described in Chapter 2. Briefly, seeding cells in the
wells of a microfluidic device or microtiter insert is accomplished by first coating the
polycarbonate membrane with fibronectin. Fibronectin (Invitrogen) and dissolved in phosphate
buffered saline (PBS) to prepare a 1000 µg/mL stock solution. The stock solution is aliquoted
and frozen until use. When cell seeding is desired, an aliquot of stock solution is thawed and
diluted to 100 µg/mL and 10 µL of solution are added to each microfluidic well, or 50 µL are
added for microtiter inserts. The solution is allowed to completely evaporate to yield a
fibronectin coating on the polycarbonate membrane.
As described in Section 2.2.3, bPAECs are cultured in T-75 tissue culture flasks, and are
removed by pipetting 2 mL of a 0.25% trypsin solution onto the cells for 2 minutes. The trypsin
solution is removed by aspiration and immediately replaced with 6 mL of cell media. The cells
are repeatedly rinsed and the bottom of the flask is gently rapped to promote removal of the
adhered cells. The resulting cell suspension is centrifuged at 500 g for 5 min. and the
supernatant removed. The cells are then resuspended in 1 mL of cell media and 10 µL or 200 µL
of the suspension are added to each well of the microfluidic device or microtiter insert,
respectively. Cell media is replaced after 1.5 hours, and then again every 3 hours afterwards.
For microtiter cell culture inserts, 450 µL of cell media is added for each replacement. If the
cells are allowed to incubate overnight, 18 µL of cell media are added to each microfluidic well,
and the device is stored in a covered petri dish with a Kimwipe saturated with water in order to

150

prevent evaporation of the cell media. Microtiter cell culture inserts are also incubated
overnight with 450 µL of cell media.

4.3.2 - TEER Measurements
Immediately prior to TEER measurements or experimentation, 1 mL of media is added to
each microtiter well containing an insert. Media contained in either the microtiter insert,
microtiter well, or microfluidic well was replaced immediately prior to permeability assay or
TEER measurement. TEER measurements on a microfluidic device were obtained using the
integrated TEER system by methods described in Chapter 2. TEER evaluation of bPAECs cultured
in microtiter cell culture inserts was accomplished using a commercially available TEER meter
(Millicell-ERS, EMD Millipore). TEER values obtained by the device were normalized to the 1.13
cm2 cell culture area of the microtiter cell culture inserts by multiplying each TEER value by the
cell culture area. Prior to all permeability assays, TEER confirmation of bPAEC monolayer
confluence was performed by measuring two sustained TEER measurements. Typically these
measurements were obtained 1.5 hours apart, and upon a positive confirmation of cell
monolayer confluence a permeability assay was subsequently performed, generally within the
following hour.

4.3.3 - Microfluidic C-peptide Permeability Assay: Sample Preparation
Human blood samples were obtained from healthy volunteers by venipuncture.

All

procedures were followed according to approved Michigan State IRB protocol. Whole blood
was collected from the donor into plastic blood collection vials (Becton, Dickinson and
Company), each containing 158 USP units of spray-dried lithium heparin to prevent coagulation.
The whole blood was then consolidated and centrifuged. After removal of the plasma and

151

buffy layer by aspiration, RBCs were washed 3 times with a physiological salt solution (PSS, pH
7.4) containing (all from Sigma-Aldrich): 4.7 mM KCl, 2.0 mM CaCl2, 140.5 mM NaCl, 12.0 mM
MgSO4, 21.0 mM tris(hydroxymethyl)aminomethane, 5.6 mM glucose, and 5% bovine serum
albumin.[17] Crude human C-peptide (Genscript) was purified by HPLC and prepared in distilled
and deionized water (DDW, 18.2 MΩ), resulting in an 8.3 µM stock solution which is stored at
2+

4°C. For samples containing Zn -activated C-peptide, a 2.0 µM working solution of human C2+

peptide was prepared in DDW and combined with the same volume of a 2.0 µM Zn solution
2+

also prepared in DDW.[17] For samples lacking C-peptide and Zn , a volume of DDW was
added equaling the volume of the other samples, ensuring that the volume of DDW was
constant throughout all samples used for a given experiment. After the Zn

2+

and C-peptide

solutions were allowed to rest for around 5 min, PSS was added to the solution, immediately
followed by the appropriate volume of washed RBCs, resulting in a solution of RBCs at 7%
hematocrit. For samples lacking RBCs, PSS was added in lieu of RBCs to yield a 20 nM C-peptide
solution. The appropriate volumes of 2.0 µM Zn

2+

and 2.0 µM C-peptide working solutions were

added so that final concentrations in the 7% RBC samples were always 20 nM.

4.3.4 - Microfluidic C-peptide Permeability Assay: Flow Studies
After bPAEC monolayer confluence was determined by the integrated TEER microfluidic
device, existing media was removed from the well and replaced with 12 µL of fresh cell media.
The solutions described above were propelled by syringe pumps (Harvard Apparatus) equipped
with 500 µL Hamilton Gastight syringes for controlled flow of solution in the microfluidic
channels. These solutions were propelled at a flow rate of 1.0 µL/min for 1 hour beneath the

152

cultured endothelial monolayers where C-peptide may permeate the bPAEC layers and finally
reach the overlying cell media that is contained within the microfluidic wells as shown in Figure
4.2. During flow, the microfluidic device and a Kimwipe saturated with water were covered by a
petri dish to prevent evaporation. After flow completion, 7 µL of solution were removed from
each microfluidic well and diluted 11 or 12 fold in cell media prior to C-peptide determination
by ELISA (EMD Millipore). The dilutions were performed so that the final concentration of the
solutions rested within the linear range of the assay, typically 0 nM – 2 nM for these particular
human C-peptide ELISA kits. ELISA was performed as instructed by included directions, though
purified stock C-peptide solutions were used to prepare standard solutions as opposed to Cpeptide standard solutions included with the kit in order to accurately represent the matrix of
the experimental solutions. A calibration curve, with standards ranging from 0-2 nM, was
obtained on each day that microfluidic permeability assays were performed, from which Cpeptide concentrations were determined in the solutions within microfluidic wells. These
solutions contain only C-peptide that successfully permeates the cultured endothelium from
solution flowing in the underlying microchannel.

4.3.5 - Microtiter C-peptide Permeability Assay: Sample Preparation
2+

For samples containing Zn -activated C-peptide, a 2.0 µM working solution of human C2+

peptide was prepared in DDW and combined with the same volume of a 2.0 µM Zn solution
2+

also prepared in DDW.[17] For samples lacking C-peptide and Zn , a volume of DDW was
added equaling the volume of the other samples, ensuring that the volume of DDW was
constant throughout all samples used for a given experiment. After the Zn

153

2+

and C-peptide

solutions were allowed to rest for around 5 min, appropriate amounts of cell media were added
to achieve 10 nM, 20 nM, and 30 nM C-peptide solutions. Unfortunately, these experiments do
not permit the use of blood samples since RBCs settle in static systems, rendering the system
physiologically irrelevant and potentially adversely affecting analyte diffusion and permeation,
again highlighting the advantages of microfluidic platforms for permeability assays and
biological mimicry.

4.3.6 - Microtiter C-peptide Permeability Assay: Static Studies
Following the confirmation of bPAEC monolayer confluence by the Millicell-ERS TEER meter,
the cell media contained in the microtiter inserts and wells was removed. The cell media
removed from the cell culture insert was replaced by 450 µL of 10 nM, 20 nM, or 30 nM Cpeptide solutions, resulting in the addition of 4.5 pmol, 9 pmol, or 13.5 pmol of C-peptide,
respectively. After the addition of C-peptide to the cell culture inserts containing the bPAEC
monolayers, 1.0 mL of cell media was added to the wells of the microtiter multiwell plate which
contained the inserts, so that the added cell media surrounds the cell culture inserts,
completing the diffusion cell as depicted in Figure 4.3. In this experimental platform, C-peptide
is added to the apical region of the diffusion cell represented by cell culture inserts, and
allowed to pervade the cultured endothelium to the basolateral region, represented by the
microtiter wells. The microtiter multiwell plate is then sealed with a plastic multiwell plate
sealer to prevent solution evaporation and then allowed to rest in the incubator (37°C) for 3.5
hours, during which C-peptide permeation from the apical region of the monolayer to the
basolateral region.

154

Following this incubation, dilutions of the solutions contained within the cell culture inserts
and microtiter wells were performed so that the resulting C-peptide concentrations fell within
the linear range of ELISA. Typically, for example, if 4.5 pmol of C-peptide was added, the
solutions in the cell culture insert were diluted 6 fold and the solutions from the surrounding
wells were diluted 2 fold. Similarly, when a total of 9 pmol of C-peptide is added, the solutions
in the microtiter inserts and wells underwent 11 fold and 2 fold, respectively. Finally, in cases
where the amount of C-peptide added totaled 13.5 pmol, 18 fold and 3 fold dilutions of the
solutions contained in the microtiter inserts and wells were performed, respectively. The Cpeptide content of the solutions resulting from these dilutions was determined by ELISA and
correlated back to determine the amount of C-peptide contained in the original solutions
extracted from the diffusion cell prior to dilution.

4.4 - Results
4.4.1 - Microfluidic Assay of C-peptide Endothelial Permeability
The ability of C-peptide to permeate a cultured endothelium was investigated by the
microfluidic vascular mimic, in conjunction with ELISA. Specifically, a physiological solution
containing C-peptide was allowed to flow beneath a cultured monolayer of bPAECs for 1 hour,
and C-peptide was allowed to permeate the cell layer to reach an overlying cell media solution
contained by a microfluidic well that initially lacked C-peptide. Importantly, these bPAEC
monolayers are characterized by the integrated TEER microfluidic system and deemed
confluent by multiple sustained TEER measurements prior to permeability studies. Initially, a
collective average TEER value for the bPAEC monolayers included in permeability assays of 3.69
± 0.26 µC was obtained, followed by an average value of 3.67 ± 0.22 µC approximately 1.5

155

hours later, resulting in an average change in TEER of 0.15 ± 0.06 µC. The content of C-peptide
contained by the basolateral solutions within the microfluidic wells was investigated by
determining C-peptide concentration via ELISA. Any C-peptide contained within these
basolateral solutions was capable of permeating the monolayer and, therefore, their C-peptide
concentration was used to gauge ability of C-peptide to permeate the endothelium.
The C-peptide concentrations determined within these solutions are displayed in Figure 4.2.
In the case where a 7% RBC solutions containing 20 nM C-peptide were flowed, the final Cpeptide concentration of the basolateral cell media solutions overlying the confluent
endothelial layers reached 17.09 ± 2.00 nM, as determined by ELISA. However, when a
physiological solution lacking RBCs was streamed beneath the cultured endothelial layers, the
overlying cell media solutions reached 18.91 ± 1.91 nM C-peptide concentrations. These 20 nM
C-peptide solutions that either included RBCs, or lacked RBCs, were also flowed beneath bare
polycarbonate membranes, and C-peptide that was allowed to permeate these membranes
that were devoid of any bPAEC layers to finally reach the overlying cell media solutions. By
flowing 20 nM C-peptide solutions containing 7% RBCs, the overlying solutions were verified by
ELISA to contain 18.31 ± 1.93 nM C-peptide in the absence of bPAEC monolayers. Also, in the
absence of RBCs, the flowing 20 nM C-peptide solution beneath bare polycarbonate
membranes caused the basolateral solutions to reach 19.91 ± 1.62 nM C-peptide
concentrations. In order to confirm that the biological matrices of the solutions investigated by
ELISA were not influencing experimental results, a physiological solution lacking C-peptide, but
containing RBCs at 7% hematocrit was flowed beneath the cell layer, and the basolateral
solution overlying the bPAECs was found to contain 0.12 ± 0.69 nM C-peptide.

156

Figure 4.2 - A cross-sectional view of one channel of the microfluidic vascular mimic used to
assess the permeability of a pulmonary endothelium to C-peptide under flow conditions is
shown in (a). The permeation of C-peptide was monitored by detecting, via ELISA, C-peptide
that permeates to basolateral solutions (in 30 minutes) from flowing apical solutions containing
20 nM C-peptide. As shown in (b), it was found that basolateral solutions overlying a cultured
endothelium (black bars) reached 17.09 ± 1.16 nM and 18.91 ± 1.11 nM C-peptide
concentrations when flowing C-peptide solutions either contained, or did not contain RBCs,
respectively. Additionally, when endothelial layers were not cultured on polycarbonate
membranes (red bars), and flowing apical solutions either contained, or did not contain RBCs,
the basolateral solutions reached 18.31 ± 1.12 nM and 19.91 ± 0.94 nM C-peptide
concentrations, respectively. Also, as a negative control, if C-peptide is not added to apical
solutions, negligible C-peptide concentrations were found. (n=3, errors are SEM)
157

4.4.2 - Static Microtiter Assay of C-Peptide Endothelial Permeability
The ability of C-peptide to permeate a cultured endothelium was also investigated with
microtiter techniques. In this construct, bPAECs were cultured in microtiter cell culture inserts,
and confirmed confluent by multiple sustained TEER measurements obtained using a MillicellERS TEER meter. Collectively, initial TEER values to confirm bPAEC confluence averaged 190.26
± 7.65 Ω·cm2, which were followed by measurements obtained approximately 1.5 hours later,
averaging 190.92 ± 4.90 Ω·cm2, resulting in an average change in TEER of 4.90 ± 2.23 Ω·cm2 over
the specified time period. In contrast, for cell culture inserts not containing bPAEC monolayers,
average TEER values of 139.74 ± 3.96 Ω·cm2 were obtained. By using this experimental
platform, solution flow and the inclusion of RBCs was not permitted. Instead, after TEER
evaluation, 10 nM, 20 nM, and 30 nM C-peptide solutions prepared in cell media were added to
the microtiter cell culture inserts, resulting in the total addition of 4.5 pmol, 9 pmol, or 13.5
pmol of C-peptide, respectively. For 3.5 hours following C-peptide addition, C-peptide was
allowed to potentially permeate the confluent cell layer and reach the solution contained by
the microtiter well in which the cell culture insert rests. C-peptide content was determined in
the apical and basolateral compartments of the diffusion cell by ELISA and presented in Table
4.1.
When a total of 4.5 pmol of C-peptide is added to the diffusion cell, 4.11 ± 0.44 pmol are
recovered from the apical region and 0.54 ± 0.6 pmol are recovered from the basolateral region
after the allotted time period. Similarly, when 9 pmol of C-peptide is initially added, 7.67 ± 0.37
pmol and 1.12 ± 0.08 pmol of C-peptide is detected in the apical and basolateral regions,
respectively. Finally, when 13.5 pmol is initially added to the apical region of the diffusion cell,

158

11.92 ± 0.46 pmol and 1.57 ± 0.26 pmol of C-peptide is ultimately detected in the apical and
basolateral regions, respectively. Alternatively, in diffusion cells not containing cultured bPAEC
monolayers, if 9 pmol of C-peptide is added to the apical side of the bare polycarbonate
membrane, the amounts of C-peptide detected in the apical and basolateral regions totaled
5.50 ± 0.27 pmol and 3.98 ± 0.27 pmol, respectively.

159

Figure 4.3 - The permeability of the pulmonary endothelium to C-peptide was investigated by
monitoring C-peptide permeation from the apical compartment of a microtiter diffusion cell,
pictured in the upper right, to the basolateral compartment. 9 pmol of C-peptide is added to
the apical compartment and, after 3.5 hours of rest, C-peptide in each compartment is
quantified by ELISA and reported. As the graph shows, when 9 pmol of C-peptide is initially
added and allowed to permeate a confluent bPAEC layer (black bars), 7.67 ± 0.37 pmol and 1.12
± 0.08 pmol of C-peptide is detected in the apical and basolateral regions, respectively.
However, if 9 pmol of C-peptide is added to the apical side of a bare polycarbonate membrane
(red bars), the amounts of C-peptide detected in the apical and basolateral regions totaled 5.50
± 0.27 pmol and 3.98 ± 0.27 pmol, respectively. (n=4, Errors are SD)

160

Table 4.1 - The permeability of an endothelial monolayer to C-peptide was investigated by
performing microtiter permeability assays, the results of which are reported here. C-peptide
was quantified in the basolateral and apical compartments via ELISA 3.5 hours after adding
varying quantities of C-peptide to apical compartments. Prior to each permeability assay,
proper bPAEC barrier function was confirmed by measuring TEER. (n=4, Errors are SD)
C-peptide added
bPAECs
(pmol)

Apical C-peptide
(pmol)

Basolateral C-peptide
(pmol)

Average TEER
(Ω·cm2)

4.5

yes

4.11 ± 0.44

0.54 ± 0.06

190.4 ± 5.4

9

yes

7.67 ± 0.37

1.12 ± 0.08

188.1 + 6.4

13.5

yes

11.92 ± 0.046

1.57 ± 0.26

194.2 ± 6.2

9

no

5.50 ± 0.27

3.98 ± 0.27

-

4.5 - Discussion
Determining the ability of C-peptide to permeate an endothelial layer in vitro would be a
significant contribution towards determining if the peptide can escape the bloodstream in vivo.
Importantly, since C-peptide is confirmed to be a bioactive peptide,[18] and is important for
regulating many fundamental biological functions throughout the body as discussed earlier, a
mechanistic understanding of these activities could be quite lucrative, scientifically. However,
most of the discovered anatomically diverse biological impacts of C-peptide on the
cardiovascular system and other organs are not currently understood mechanistically, or
otherwise. In fact, aside from the work presented in Chapter 3, it has not yet been determined
if C-peptide acts to induce biological events in various anatomical locations by directly
interacting with the cell types that are primarily responsible for a specific event, or if these
events are indirect consequences of C-peptide activity in the blood stream. The capability of Cpeptide to act by direct mechanisms throughout the body is dependent on the ability of C-

161

peptide to permeate the endothelium in vivo. Therefore, all direct mechanisms can potentially
be ruled out by showing that C-peptide is incapable of penetrating an endothelial layer since
such findings would strongly suggest that C-peptide is confined to the bloodstream in vivo. For
these reasons, in addition to the results presented in Chapter 3 showing that C-peptide acts
indirectly to influence purinergic signaling, the ability of C-peptide to permeate a cultured
endothelium was investigated.
In order to determine the ability of C-peptide to permeate an endothelial layer, a
microfluidic vascular mimic was used in conjunction with conventional microtiter techniques.
These two platforms have stark differences however, with microfluidic technology allowing for
the flow of RBCs and, in contrast, microtiter technology is a static system and, therefore, does
not permit flow. Additionally, the inclusion of RBCs is physiologically irrelevant in a static
system, especially considering the tendency of RBCs to sink in static solutions, which would
cause RBCs to coat the apical membrane and influence the diffusion and permeation of
analytes between the apical and basolateral compartments. Furthermore, the work presented
in Chapter 3 describes how RBCs can drastically influence endothelial cells, namely by their
ability to release ATP, which ultimately stimulates endothelial nitric oxide synthase (eNOS).
Importantly, eNOS has been shown to play a predominant role in regulating vascular
permeability in vivo,[19-21] and furthermore, flowing RBCs have been shown to release more
ATP than static RBCs.[22, 23] Since endothelial barrier integrity is the primary determinant of
vascular permeability,[24] flowing RBCs will ideally be present in a vascular model, an important
shortcoming to consider when performing microtiter based permeability assays.

162

Perhaps the most important aspect of a physiologically relevant vascular model is the
incorporation of shear stress generated by flowing blood. For instance, shear stress has been
shown to increase eNOS activity that, as stated earlier, plays a role in vascular permeability by
influencing the barrier integrity of the endothelium. Also, in addition to increasing ATP release
from RBCs, shear influences the metabolic state of RBCs,[25] as well as interactions between
flowing RBCs that can lead to changes in viscosity and aggregation.[26] While the ability to
increase eNOS activity is of fundamental importance, especially for permeation research, shear
stress impacts endothelial cell behavior even more drastically. For example, shear stress alters
endothelial morphology,[27] tissue factor expression,[28] signal transduction between adjacent
endothelial cells,[29] intracellular signaling,[30, 31] and gene expression.[32, 33] Collectively,
this myriad of influences makes shear stress a principal determinant of endothelial cell behavior
and health.[34] Summarily, in contrast to microfluidic systems, microtiter technology does not
account for the dramatic role of the RBC or shear stress in vascular function.
Furthermore, a microfluidic platform enables the flow of a solution with a specific C-peptide
concentration which, in this case, is more physiologically correct than adding a finite amount
since the concentration of circulating C-peptide is relatively constant in vivo. For these reasons,
recovered C-peptide is reported as a concentration and an amount for microfluidic and
microtiter permeability studies, respectively. Not surprisingly, the inherent differences between
the two platforms influence experimental results, as well as perception of these results. For
example, when employing the microfluidic vascular mimic to determine the ability of C-peptide
to permeate a cultured endothelium, a 20 nM C-peptide solution flowing apical to the cell
monolayer and not containing RBCs resulted in enough C-peptide permeating the cultured

163

endothelium to ultimately raise the overlying basolateral solution, which initially lacks Cpeptide, to 18.91 ± 1.91 nM within an hour. In contrast, after adding a 20 nM C-peptide
solution also not containing RBCs, for a total of 9 pmol, to the apical compartment of a
microtiter diffusion cell, 1.12 ± 0.08 pmol of C-peptide successfully permeated the endothelial
monolayer to the basolateral compartment in 3.5 hours.
Interestingly, the microfluidic permeability assay provides the unique perception that the
basolateral solution nearly equilibrates with the apical flowing solution within an hour. In
contrast, in over triple the time allotted for the microfluidic assay, the microtiter permeability
assay reveals that only about 10% of the total C-peptide added to the apical compartment of
the diffusion cell successfully reaches the basolateral compartment, which actually contains a
higher volume of solution than the apical compartment. In fact, this was the case for each
microtiter based permeability assay, regardless of the quantity of C-peptide added. However,
these results are to be expected since shear has been shown to influence changes in
endothelial permeability.[35] As an example, one study shows that shear stress can increase
the permeability of a cultured endothelium 4-fold compared to an endothelium under static
conditions.[36] Interestingly however, since the microfluidic wells contained only 12 µL of
solution, only about 0.23 pmol of C-peptide would have to permeate the endothelial layer to
result in a 19 nM solution. Hypothetically, if it took a full hour for this to occur, only 0.8 pmol
would permeate to the basolateral solution in the same timeframe allotted in the microtiter
permeability assay, suggesting the two results may not be so contrasting and also further
highlight the varying perspectives that each platform provides. Obviously, in order to confirm
this, the kinetics of this equilibration would need to be determined for the basolateral solution

164

in the microfluidic vascular mimic. Regardless, due to the small sample volumes typical of
microfluidic systems, C-peptide permeation can be determined in a shorter duration than with
microtiter systems. However, even considering the short duration of the microfluidic
permeability assay, the ability of the endothelial monolayer to hinder C-peptide permeation to
the basolateral region was not observed as it was microtiter permeability assay when 1.12 ±
0.08 pmol of C-peptide was detected in the basolateral chamber when the diffusion cell
contained a cultured endothelium, compared to 3.98 ± 0.27 pmol detected when it did not.
Another aspect of the microfluidic approach towards studying permeation is the ability to
determine the influence of blood or other common biological matrices on analyte permeation.
For example, if blood was included in the flowing physiological C-peptide solution, the
basolateral solution reached 17.09 ± 2.00 nM, compared to 18.91 ± 1.91 nM for a flowing
solution lacking RBCs. While not statistically different, the basolateral solutions slightly differed
in concentration by around 2 nM, depending on the presence of RBCs. This effect was again
observed when bPAECs were not cultured on the polycarbonate membrane and the flowing 20
nM C-peptide solutions containing RBCs, or lacking RBCs, resulted in basolateral solution
concentrations of 18.31 ± 1.93 nM and 19.91 ± 1.62 nM, respectively. Interestingly,
unpublished data from our group has reproducibly shown that in a RBC solution at 7%
hematocrit, C-peptide is able to interact with the RBC and thus lower the concentration of free
C-peptide in the solution by 2 nM. Collectively, these data suggest that interactions between
the RBC and C-peptide may act to lower the concentration of bioavailable C-peptide in vivo,
which may be considerable since C-peptide is typically found at single-digit nanomolar levels in
the bloodstream. Similarly, a small change can be observed that is dependent on the presence

165

of endothelial cells. While not statistically significant, after 1 hour of flow in the absence of
endothelial cells, each measurement of C-peptide concentration in basolateral solutions is
approximately 1 nM lower than in the presence of endothelial cells. While this change is slight,
it is interesting that C-peptide has been shown to specifically bind to endothelial cells via an
event that is saturated at approximately 1 nM levels.[5, 37]
Despite such vast differences, a strength shared by each platform is the ability to confirm
endothelial monolayer confluence and barrier integrity. Clearly, a confirmation of barrier
integrity is incredibly advantageous, especially for permeability assays, adding significant
credibility to experimental results and conclusions stemming from these experimental results.
In the permeability studies presented, multiple sustained TEER measurements were obtained as
verification of bPAEC confluence. These measurements, typically obtained 1.5 hours apart,
yielded dismal changes over time, averaging to 0.15 ± 0.06 µC and 4.90 ± 2.23 Ω·cm2 for the
microfluidic and microtiter assays, respectively. These confirmations that every cultured
endothelium used to assay C-peptide permeation formed confluent barriers increase the
likelihood that results derived from these assays will correlate to true activity in vivo.
An additional similarity between the two experimental platforms is that each technique
conclusively confirms that C-peptide is capable of permeating a cultured endothelium. These
results suggest that C-peptide may be able to perform biological functions by direct and indirect
stimulation of cell types throughout the body. If C-peptide is truly capable of escaping the
bloodstream, it may act as an important hormone that functions universally in the body, similar
to insulin. This activity would support the recent concept that type 1 diabetes is actually a
double-hormone deficiency disorder, with late complications potentially being derived from

166

chronic lack of C-peptide.[18] These results, in conjunction with the ability of C-peptide to
specifically bind,[5, 37] and also become internalized by endothelial cells,[7] suggest that Cpeptide likely has an endothelial transcellular transport mechanism. This is further supported
by C-peptide’s ability to be internalized by,[7, 14] and induce biological events in cell types
exterior to the bloodstream.[15]

4.6 - Conclusions
In summary, these results are the first to suggest that C-peptide is capable of permeating
the endothelium in vitro, implying that C-peptide can likely also permeate the endothelium in
vivo. These findings suggest that, in addition to indirect influences shown in Chapter 3, Cpeptide may have direct influences on various cell types throughout the body. Since C-peptide
has never been detected exterior of the bloodstream, these novel findings suggest that Cpeptide may have a much more comprehensive role in physiology and health than is currently
assumed.

167

REFERENCES

168

REFERENCES
1.

Pierson, C.R., W. Zhang, and A.A. Sima, Proinsulin C-peptide replacement in type 1
diabetic BB/Wor-rats prevents deficits in nerve fiber regeneration. Journal of
Neuropathology and Experimental Neurology, 2003. 62(7): p. 765-79.

2.

Maezawa, Y., et al., Influence of C-peptide on early glomerular changes in diabetic mice.
Diabetes/metabolism research and reviews, 2006. 22(4): p. 313-22.

3.

Sima, A.A., et al., Inflammation in Diabetic Encephalopathy is Prevented by C-Peptide.
Rev Diabet Stud, 2009. 6(1): p. 37-42.

4.

Lindstrom, K., et al., Acute effects of C-peptide on the microvasculature of isolated
perfused skeletal muscles and kidneys in rat. Acta Physiol Scand, 1996. 156(1): p. 19-25.

5.

Rigler, R., et al., Specific binding of proinsulin C-peptide to human cell membranes. Proc.
Natl. Acad. Sci., 1999. 96: p. 13318-13323.

6.

Henriksson, M., et al., Separate functional features of proinsulin C-peptide. Cell Mol Life
Sci, 2005. 62(15): p. 1772-8.

7.

Luppi, P., et al., C-peptide is internalised in human endothelial and vascular smooth
muscle cells via early endosomes. Diabetologia, 2009. 52(10): p. 2218-28.

8.

Minshall, R.D. and A.B. Malik, Transport across the endothelium: regulation of
endothelial permeability. Handb Exp Pharmacol, 2006(176 Pt 1): p. 107-44.

9.

Minshall, R.D., et al., Vesicle formation and trafficking in endothelial cells and regulation
of endothelial barrier function. Histochem Cell Biol, 2002. 117(2): p. 105-12.

10.

Lossinsky, A.S. and R.R. Shivers, Structural pathways for macromolecular and cellular
transport across the blood-brain barrier during inflammatory conditions. Review. Histol
Histopathol, 2004. 19(2): p. 535-64.

11.

Le Sueur, L.P., C.B. Collares-Buzato, and M.A. da Cruz-Hofling, Mechanisms involved in
the blood-brain barrier increased permeability induced by Phoneutria nigriventer spider
venom in rats. Brain Research, 2004. 1027(1-2): p. 38-47.

12.

Groneberg, D.A., et al., Distribution and function of the peptide transporter PEPT2 in
normal and cystic fibrosis human lung. Thorax, 2002. 57(1): p. 55-60.

13.

Meredith, D. and C.A. Boyd, Dipeptide transport characteristics of the apical membrane
of rat lung type II pneumocytes. Am J Physiol, 1995. 269(2 Pt 1): p. L137-43.

14.

Lindahl, E., et al., Cellular internalization of proinsulin C-peptide. Cell. Mol. Life Sci.,
2007. 64: p. 479-486.

169

15.

Lindahl, E., et al., Proinsulin C-peptide regulates ribosomal RNA expression. J Biol Chem,
2010. 285(5): p. 3462-9.

16.

Leith, A.G., G.D. Griffiths, and M.A. Green, Quantification of ricin toxin using a highly
sensitive avidin/biotin enzyme-linked immunosorbent assay. J Forensic Sci Soc, 1988.
28(4): p. 227-36.

17.

Meyer, J.A., et al., Metal-activated C-peptide facilitates glucose clearance and the
release of a nitric oxide stimulus via the GLUT1 transporter. Diabetologia, 2008. 51(1): p.
175-82.

18.

Wahren, J., K. Ekberg, and H. Jornvall, C-peptide is a bioactive peptide. Diabetologia,
2007. 50: p. 503-509.

19.

Fukumura, D., et al., Predominant role of endothelial nitric oxide synthase in vascular
endothelial growth factor-induced angiogenesis and vascular permeability. Proc Natl
Acad Sci U S A, 2001. 98(5): p. 2604-9.

20.

Hutcheson, I.R., B.J. Whittle, and N.K. Boughton-Smith, Role of nitric oxide in
maintaining vascular integrity in endotoxin-induced acute intestinal damage in the rat.
Br J Pharmacol, 1990. 101(4): p. 815-20.

21.

Fukumura, D., et al., Role of nitric oxide in tumor microcirculation. Blood flow, vascular
permeability, and leukocyte-endothelial interactions. Am J Pathol, 1997. 150(2): p. 71325.

22.

Sprung, R.J., R.S. Sprague, and D.M. Spence, Determination of ATP release from
erythrocytes using microbore tubing as a model of resistance vessels in vivo. Anal.
Chem., 2002. 74: p. 2274-2278.

23.

Price, A., K., et al., Deformation-induced release of ATP from erythrocytes in a
poly(dimethylsiloxane)-based microchip with channels that mimic resistance vessels.
Analytical Chemistry, 2004. 76(16): p. 4849-55.

24.

Cines, D.B., et al., Endothelial cells in physiology and in the pathophysiology of vascular
disorders. Blood, 1998. 91(10): p. 3527-61.

25.

Murphy, J.R., Erythrocyte Metabolism. Vi. Cell Shape and the Location of Cholesterol in
the Erythrocyte Membrane. J Lab Clin Med, 1965. 65: p. 756-74.

26.

Brooks, D.E., J.W. Goodwin, and G.V. Seaman, Interactions among erythrocytes under
shear. Journal of applied physiology (Bethesda, Md. 1985), 1970. 28(2): p. 172-7.

27.

Davies, P.F., Flow-mediated endothelial mechanotransduction. Physiol Rev, 1995. 75(3):
p. 519-60.

170

28.

Grabowski, E.F. and F.P. Lam, Endothelial cell function, including tissue factor expression,
under flow conditions. Thrombosis and haemostasis, 1995. 74(1): p. 123-8.

29.

Lehoux, S. and A. Tedgui, Signal transduction of mechanical stresses in the vascular wall.
Hypertension, 1998. 32(2): p. 338-45.

30.

Li, Y.S., J.H. Haga, and S. Chien, Molecular basis of the effects of shear stress on vascular
endothelial cells. J Biomech, 2005. 38(10): p. 1949-71.

31.

Lehoux, S., Y. Castier, and A. Tedgui, Molecular mechanisms of the vascular responses to
haemodynamic forces. Journal of internal medicine, 2006. 259(4): p. 381-92.

32.

Malek, A.M. and S. Izumo, Control of endothelial cell gene expression by flow. J Biomech,
1995. 28(12): p. 1515-28.

33.

Resnick, N., et al., Endothelial gene regulation by laminar shear stress. Adv Exp Med Biol,
1997. 430: p. 155-64.

34.

Chatzizisis, Y.S., et al., Role of endothelial shear stress in the natural history of coronary
atherosclerosis and vascular remodeling: molecular, cellular, and vascular behavior. J
Am Coll Cardiol, 2007. 49(25): p. 2379-93.

35.

Jo, H., et al., Endothelial albumin permeability is shear dependent, time dependent, and
reversible. Am J Physiol, 1991. 260(6 Pt 2): p. H1992-6.

36.

Sill, H.W., et al., Shear stress increases hydraulic conductivity of cultured endothelial
monolayers. Am J Physiol, 1995. 268(2 Pt 2): p. H535-43.

37.

Pramanik, A., et al., C-Peptide Binding to Human Cell Membranes: Importance of Glu27.
Biochem. Biophys. Res. Comm, 2001. 284: p. 94-98.

171

Chapter 5 - Conclusions and Future Directions
5.1 - Conclusions
5.1.1 - Microfluidic TEER Measurement Device
Since the 1980s, microfluidic technologies have filled important niches in many scientific
fields.[1, 2] More recently, microfluidics have been shown to be very useful for in vitro cell
based assays, in particular the study of cellular monolayers or single cells, due to their
unsurpassed ability to control cellular environments and mimic biological systems found in

vivo.[3-7] However, in vitro cell culture and analysis is a longstanding science that has been
consistently advancing since the 1910s,[8, 9] with cellular monolayers being cultured and
studied as early as the 1920s.[10, 11] Over time, as the understanding of the physiological
importance cellular monolayers increased, monolayer barrier integrity became a well-studied
research topic by the 1950s, especially in regards to the impact of confluence and barrier
integrity on cell behavior.[12-14] To meet this demand, TEER measurement devices have been
made commercially available since the mid-1980s and have been further developed for the high
throughput analysis of monolayer barrier integrity. However, regardless of the knowledge
acquired from ongoing work building on research initiated in the 1950s showing the importance
of quantitatively measuring barrier integrity to ensure appropriate cellular behavior and
response, the copious amounts of microfluidic assays of cellular monolayers have not included
such measurements. This is especially surprising since cellular monolayers have been routinely
cultured in microfluidic devices for over a decade.[15, 16] Within the past two years, the first
microfluidic devices capable of determining barrier integrity by measuring analyte

172

permeability,[17] or TEER,[18-21] have appeared in literature, which includes the microfluidic
TEER system presented in Chapter 2, indicating the necessity for such devices.
Collectively, these devices and the microfluidic TEER measurement device presented in
Chapter 2 finally addresses the long withstanding need for a microfluidic platform with
monolayer barrier characterization capabilities. However, the microfluidic device presented in
Chapter 2 is novel as it is the first microfluidic capable of measuring barrier integrity that serves
as a vascular mimic, and also the first microfluidic device capable of monitoring barrier integrity
that has been further utilized to observe real biological events.[21] For example, work
presented in chapter 3 shows that confluent endothelial cells, as determined by the microfluidic
TEER system, are capable of producing significantly more nitric oxide (NO) in response to
flowing red blood cells (RBCs) than non-confluent cells, a difference that is statistically
significant. These results highlight the importance of confirming monolayer confluence prior to
investigations involving the monolayer, since important physiological responses are sometimes
drastically altered by the extent of cell layer confluence.
Additionally, the simplicity of the design and construction of the device adds to its novelty,
as similar devices are considerably more complicated in comparison, facilitating the unique
potential for the successful integration of the TEER system with preexisting microfluidic
platforms via the same design strategy employed during this work to develop the TEER device.
Since measuring TEER of a cellular monolayer cultured in a microfluidic device can provide the
same fundamental information and physiological accuracy as in conventional platforms, it is
expected that the development of this technology has provided the necessary foundation so
that TEER measurements can, and optimistically will be, used as extensively and routinely in

173

microfluidics as they are in other platforms. In any event, the work described throughout
involving cellular monolayers is benefitted by the TEER measurement capabilities of the
microfluidic vascular mimic. Given the ability to confirm and monitor cell monolayer
confluence, cellular responses were consistently of optimal physiological relevance and
occasionally enhanced, such as it was for RBC stimulated NO production in confluent
endothelial cells mentioned previously. Additionally, a dramatic increase experimental
reproducibility was accomplished as a result of the ability to investigate cellular responses with
consistently optimal physiological relevance, which surely enhances the accuracy and
applicability of conclusions stemming from these investigations, adding credibility to the work.
While these advantages illustrate the importance of confirming monolayer confluence prior to
experimentation, they also exhibit the ability of this particular device to mimic the vasculature
with unparalleled biological accuracy as a result of its integrated TEER measurement
capabilities.

5.1.2 - C-peptide Induced Vasodilation
Previously, C-peptide has been shown to improve blood flow in diabetic patients by an
unknown mechanism.[22, 23] Interestingly, separate studies have shown that, diabetic
complications related to blood flow are related to endothelial dysfunction,[24] and a potent
vasodilator, nitric oxide.[25] Also previously, our group has established that C-peptide, if
2+

activated by Zn , is capable of increasing ATP release from RBCs.[26] Interestingly, ATP has
been shown to bind endothelial P2Y, a purinoreceptor capable of initiating a purinergic
signaling cascade which results in endothelial nitric oxide synthase (eNOS) activation, the
enzyme responsible for producing and regulating NO in the endothelium. Mindful of these

174

previous findings, it was hypothesized that C-peptide is capable of indirectly participating in
purinergic signaling, as mediated by C-peptide stimulated release of ATP from RBCs, which
ultimately results in increased endothelial NO production.

Since NO derived from the

endothelium has been identified as one of the most potent vasodilators yet discovered,[27] if Cpeptide was determined to increase endothelial NO production, it would provide an
explanation for C-peptide induced improvements in diabetic blood flow. The microfluidic
vascular mimic with integrated TEER capabilities developed during the work presented in
Chapter 2 was employed to examine the proposed mechanism of C-peptide induced endothelial
NO production in work described in Chapter 3.
2+

In order to examine the proposed mechanism, RBC solutions containing Zn -activated Cpeptide were flowed in the channels of the vascular mimic beneath a cultured endothelium,
and intracellular endothelial NO production was measured and compared to NO production
resulting from flowing RBC solutions not containing C-peptide. It was determined that flowing
2+

RBC solutions containing Zn -activated C-peptide are capable of increasing endothelial NO
production by almost 90% relative to RBC solutions lacking C-peptide. Interestingly, this effect
2+

was not observed if physiological RBC solutions containing only Zn , or only C-peptide were
flowed beneath the endothelial layer. This result shows that C-peptide requires Zn

2+

to

stimulate endothelial NO production, adding to other findings in the Spence group
demonstrating the necessity of a metal for C-peptide bioactivity.[26, 28-30] Additionally, it was
discovered that C-peptide is indeed capable of increasing endothelial NO production, however
these results alone do not show conclusively that this NO production is a result of purinergic

175

2+

signaling mediated by ATP released from RBCs as a result of Zn -activated C-peptide
stimulation. Instead, the RBC was revealed to mediate the NO production when physiological
2+

solutions lacking RBCs, but still containing Zn -activated C-peptide did not stimulate an
increase in NO production when flowed beneath the cultured endothelium. In order to show
that RBC derived ATP is involved in the proposed mechanism, endothelial cellular monolayers
were incubated with PPADS, a P2Y purinoreceptor inhibitor, flowing solutions containing RBCs
2+

and Zn -activated C-peptide were no longer able to increase NO production, resulting in a
complete reversal of the increase in NO production measured when comparing the flow of
2+

2+

RBCs incubated with Zn /C-peptide to RBCs alone. Collectively, these results prove that Zn activated C-peptide increases endothelial NO production by stimulating the release of a known
endothelial P2Y agonist, ATP, from RBCs and thereby enhancing P2Y purinoreceptor activity by
the stimulated increase in ATP antagonist levels, which prompts a purinergic signaling cascade
known to result in increased endothelial NO production.
The novel findings, presented in Chapter 3, revealing the involvement of C-peptide in
purinergic signaling and endothelial NO production provides an explanation for the observed
improvements in blood flow of diabetic patients. Additionally, these findings show the potential
for C-peptide as a drug candidate for the treatment of type 1 diabetes. Of course, in order to
elicit these desired effects as a drug, C-peptide would have to be properly introduced to a
2+

metal, particularly Zn , prior to administration. Unfortunately, this vital step has been
overlooked previously, potentially explaining irreproducible activity of C-peptide observed
during clinical trials that contrast with the reproducible stimulation of endothelial NO

176

2+

production by Zn -activated C-peptide. Plausibly, since C-peptide is often co-administered with
insulin, and insulin often contains substantial (0.5% by mass) zinc impurities, C-peptide may be
unintentionally introduced to a metal source eliciting varying degrees of biological activity upon
administration, ultimately leading to inconsistent and irreproducible outcomes. Nevertheless,
2+

this work conclusively shows that, only in the presence of Zn , C-peptide stimulates nitric oxide
production in a cultured pulmonary artery endothelium via RBC-mediated purinergic signaling.
Interestingly, the ability of C-peptide to treat diabetic complications suggests that a chronic
lack of C-peptide leads to the development of these complications and, more importantly, Cpeptide plays a vital physiological role in maintaining human health. Therefore, C-peptide may
not only improve diabetic blood flow, but it may be vital for regulating vascular tone in healthy
individuals via the mechanism confirmed by this work. As a result, C-peptide may be of far more
physiological importance than originally thought, which is supported by the discovered
involvement of C-peptide in purinergic signaling, since purinergic signaling is responsible for
regulating numerous biological processes throughout the body.[31] For these reasons, RBCs
may play an equally important role in regulating vascular tone. As discussed in Chapter 1, it is
well-known that NO produced in the blood vessel is primarily responsible for the fundamentally
important task of regulating vessel tone, however the source of this NO is debated in the
literature. Generally, the consensus is that this NO is produced in vascular regions of high shear
stress generated by flowing blood, such as in resistance vessels. However, generally these
regions are also where oxygen is exchanged from the blood to the surrounding tissues in
response to hypoxia, which has been shown to cause NO release from RBCs. In this construct,
NO is either derived from RBCs in response to hypoxia, or from the endothelium in response to

177

shear stress, however the mechanism by which the endothelium is still under discussion.
However, one promising hypothesis largely developed by Sprague et al. links NO production in
the endothelium to the RBC by ATP,[32] specifically by the ability of the RBC to release ATP.
This hypothesis is supported by the known vasodilatory effects of circulating ATP achieved by
stimulating NO production in the endothelium via the same purinergic signaling event involved
in mechanism elucidated during this work for C-peptide stimulated NO production. Also
supporting this hypothesis is work performed in the Spence lab showing that flowing RBCs
release more ATP than static RBCs.[33] Collectively, these results, and others not discussed,
support the hypothesis that ATP released from RBCs in response to shear stress is able to
stimulate endothelial NO production by purinergic signaling.[34]
Further supporting this hypothesis are the results discussed earlier, and obtained while
investigating endothelial NO production in the microfluidic vascular mimic during work
presented in Chapter 3. Specifically, when a physiological solution not containing RBCs was
allowed to flow beneath the cultured endothelium, a significant decrease in endothelial NO
production was observed (p<0.1) in comparison to NO production resulting from a flowing
solutions that contain RBCs. This result shows that flowing RBCs under the influence of shear
stress can cause an increase in NO produced in the cultured endothelium. However, this
activity, which was reported by the Spence group previously,[35] does not conclusively show
that this NO is produced as a result of the previously identified ability of shear stress to
stimulate ATP release from RBCs. In the same study, however, if selected endothelial layers
were incubated with PPADS, a purinergic receptor inhibitor, the NO production in response to
flowing solutions containing RBCs was statistically the same as NO produced in uninhibited

178

endothelial layers in response to flowing physiological solutions not containing RBCs.
Collectively, these results suggest that shear induced ATP released from RBCs is capable of
stimulating NO production in the pulmonary endothelium via purinergic signaling.
2+

In summary, these studies show that C-peptide, in the presence of Zn , is capable of
indirectly stimulating NO production in the pulmonary endothelium by mechanism mediated by
the red blood cell. Furthermore, these studies display the involvement of C-peptide in
purinergic signaling, and since purinergic signaling is involved in numerous biological processes,
it may be by this mechanism that C-peptide accomplishes vast biological activity throughout the
body. Additionally, since C-peptide is capable of stimulating NO production in the endothelium,
the peptide may play a crucial and overlooked role in maintaining basal NO levels in the
bloodstream and regulating vascular tone. Also, by the same purinergic mechanism, RBCs under
the influence of shear stress are capable of stimulating endothelial NO production. The source
of shear induced NO production is under considerable debate in the literature, however this
work shows reveals that sheared RBC are at least partially responsible for generating shear
induced NO, and it is by this mechanism that RBCs may regulate vascular tone.

5.1.3 - Endothelial Permeability to C-peptide
Determining the ability of C-peptide to permeate an endothelial layer in vitro would be a
significant contribution towards determining if the peptide can escape the bloodstream in vivo.
This is important to determination since, aside from the work presented in Chapter 3, it has not
yet been determined if C-peptide acts to induce biological events in various anatomical
locations by directly interacting with the cell types that are primarily responsible for a specific
event, or if these events are indirect consequences of C-peptide activity in the blood stream.

179

Since the capability of C-peptide to act by direct mechanisms throughout the body is dependent
on the ability of C-peptide to permeate the endothelium in vivo, an in vitro study was
performed to determine the likelihood that C-peptide can escape the bloodstream to act by
direct mechanisms, or if the peptide is likely confined to the bloodstream, implying that all
bioactivity is carried out indirectly by influences within the vasculature. To accomplish this
investigation, the ability of C-peptide to permeate an endothelial layer was observed using the
microfluidic vascular mimic in conjunction with conventional microtiter techniques.
Despite such vast differences between these models, a strength shared by each platform is
the ability to confirm endothelial monolayer confluence and barrier integrity. As such, complete
endothelial monolayer confluence was determined prior to each permeability assay, thereby
ensuring the cultured cells most accurately mimic the true pulmonary endothelium. In work
presented in Chapter 4, each experimental platform conclusively shows that C-peptide is
capable of permeating a cultured endothelium. These results suggest that C-peptide may be
able to perform biological functions by direct and indirect stimulation of cell types throughout
the body. For example, work discussed previously suggests that C-peptide is capable of
indirectly stimulating smooth muscle cell to induce vasorelaxation as a result of activity in the
bloodstream. However, if C-peptide is truly capable of escaping the bloodstream, it may act as
an important hormone that functions universally in the body by direct and indirect
mechanisms, similar to insulin. This activity would support the recent concept that type 1
diabetes is actually a double-hormone deficiency disorder, with late complications potentially
being derived from chronic lack of C-peptide.[36] The ability of C-peptide to induce biological
events in cell types exogenous of the bloodstream has been previously shown,[37] however it is

180

currently unclear if C-peptide is capable of reaching these cells. These results suggest that Cpeptide may indeed be capable of direct interactions with cells external of the cardiovascular
system as indicated by the ability to permeate a cultured endothelium. Since C-peptide has
never been detected exterior of the bloodstream, these novel findings suggest that C-peptide
may have a much more comprehensive role in physiology and health than is currently assumed.

5.2 - Future Directions
5.2.1 - Microfluidic TEER System
Currently, a strongpoint of the TEER system is the simplicity of device design and
fabrication, which is easily accomplished by using soft lithography for the rapid prototyping of
polydimethylsiloxane (PDMS) devices as discussed in Chapter 2. While PDMS offers many
advantages for microfluidic cell based assays and culture, as a material for microfluidic
fabrication PDMS has several disadvantages,[38, 39] such as hydrophobicity, absorption and
adsorption of substances, as well as high flexibility. These disadvantages render PDMS
microfluidic devices undesirable in various situations. For example, a future direction of the
TEER device is to study the pharmacokinetics and pharmacodynamics (PK/PD) of various drugs,
however unpublished data from the Spence lab shows that several drugs absorb in PDMS,
rendering the current device incapable of analyses of those drugs. Additionally, the flexibility of
PDMS facilitates unintentional wrinkling of the polycarbonate membranes, which leads to
leaking in the microfluidic device. Another issue regarding the current device design not related
to PDMS as a fabrication material is the throughput attained using a single moving top
electrode for TEER determination. Currently, the electrode must be manually inserted in the
solution overlying the cultured cells for each measurement. This method does not achieve the

181

desirable high throughput generally accomplished by microfluidic analyses. However, these
disadvantages may be addressed by compromising simplicity, a tradeoff likely favorable in this
case.

5.2.1.1 - Polystyrene (PS) as a Material for Microfluidic Fabrication
Recently, methods for micromolding and rapid prototyping of PS by soft lithography have
been accomplished by work performed in the lab of Nancy Albritton,[40] as well as in the labs
of Dana Spence and Scott Martin that was recently submitted for publication. PS is amenable to
cell culture and is very well characterized for such, as it is the material most commonly used to
construct tissue culture flasks and Petri dishes. Remarkably, PS is capable of alleviating the
current disadvantages of PDMS as the material used to construct the microfluidic TEER system,
such as absorption of small molecules and high degrees of flexibility. PS would greatly increase
the robustness of the device and surely decrease the frequency of leaking, however the
construction of PS microfluidic devices is substantially more complicated and time consuming
than constructing PDMS devices. However, this issue is alleviated by the robustness of PS
microfluidic platforms, potentially allowing the device to be reused for multiple experiments. Of
course, a method for high throughput TEER analysis would need to be developed for the PS
microfluidic device.
Recent advances in the Spence and Martin labs have showed promise for the successful
development of a PS microfluidic device capable of high throughput TEER analysis. Figure 5.1
shows a potential design for such a device, which includes an array of electrodes integrated into
a bottom layer of PS for use in in the microfluidic TEER measurement system 2.0. While
microfluidic-based TEER systems have been reported, these systems have yet to be used in an

182

array format allowing for high throughput analysis. Here, it is proposed that such an array can
facilitate proper timing of cell subculturing in the microfluidic system that will allow for the
culture of cells over a period of time that may extend from days to weeks, during which various
cell based assays are possible, including PK/PD determinations for drugs of interest.
The approach shown in Figure 5.1 can be used to integrate TEER measurement capabilities
in a PS device. As is shown in Figure 5.1, the group of Scott Martin has recently been able to
embed microelectrode arrays in polystyrene-based microchips. This was accomplished by
melting polystyrene powder in the presence of a gold SEM grid that was affixed to a connecting
wire. Once the chip was cooled, it could be polished to a flat surface. Subsequently, it was
found that these embedded electrode arrays can be used for amperometric detection in
polystyrene-based microchips, suggesting that they can be used to successfully determine
TEER. In this study, the arrays will be made to reside below each reservoir (Figure 5.1), and a
separate set of reservoir electrodes (conventional wires) will be directed through a relay
(controlled by LabView) into the reservoir so that the TEER value for each reservoir can be
monitored in an automated fashion. This separate set of reservoir electrodes will be used to
replace the single moving top electrode described previously, and importantly, these reservoir
electrodes will be arranged in an array format. By using relays, a specific relay can be triggered,
completing the circuit and allowing for the TEER reading for that particular reservoir to be
attained.
The successful development of this proposed microfluidic TEER system would increase the
robustness of the current design to a great extent. The proposed device would likely be
reusable and prevent the absorption of small molecules into the device. Most importantly, the

183

throughput of the proposed device would be greatly increased relative to the current TEER
system, since an array of reservoir electrodes can be used in lieu of the single moving reservoir
electrode requiring the tedious manual placement in each reservoir for every TEER
measurement. In contrast, reservoir electrodes in an array format would allow for automated
high throughput TEER analysis without human involvement. Collectively, this increase in
throughput, in addition to minimal absorption and increased device sturdiness will likely greatly
increase the utility and applicability of the microfluidic TEER device.

184

TEER

Figure 5.1 - Side-view of microchip with integrated TEER measurement. For simplicity, a 4reservior array is shown. Also shown is a micrograph of a 4-microelectrode (gold) array that is
embedded in polystyrene, which was fabricated in the lab of Scott Martin. The potential of the
preliminary 4-microelectrode gold array as an effective platform for TEER analysis is illustrated
by the previous utilization of the array in a microfluidic platform to detect analytes by
amperometry, an electrochemical technique that is experimentally similar to the method used
to determine TEER.

185

5.2.2 - C-peptide Induced Vasodilation
2+

5.2.2.1 - Elucidate Zn /C-peptide Interactions
The specifics of how Zn

2+

is capable of rendering C-peptide capable of biological activity are

currently unknown. Speculatively, Zn

2+

is capable of binding C-peptide, possibly inducing a

biologically active conformation of the peptide. In support of this proposed interaction, circular
2+

dicrhroism (CD) spectroscopic analysis performed in the Spence lab of Zn -activated C-peptide
2+

and C-peptide alone suggest that in the presence of Zn , a conformational change may occur.
Also, a Scatchard binding curve has been obtained in the Spence lab for Zn
peptide, which suggests that a single Zn

2+

2+

binding to C-

atom is capable of binding C-peptide. Additionally,

mass spectrometric analysis of the active complex has revealed that C-peptide was likely bound
to a metal, as discussed in Chapter 1. Aside from these results, as well as the observed change
2+

in the biological activity of C-peptide in the presence of Zn , there is no further evidence for
2+

Zn /C-peptide binding. These results do have not identified specific binding regions, or
conclusively shown that binding results in an active conformation of the peptide.
As shown in Figure 5.2, C-peptide exhibits five negatively charged residues in vivo that are
2+

likely capable of Zn

binding and, therefore, provide insight as to where Zn

peptide. However, since preliminary data shows that only 1 Zn

2+

2+

may bind the

atom binds C-peptide, it is still

unclear which residue or residues it may bind. In order to conclusively show that C-peptide
does bind Zn

2+

to achieve an active conformation, nuclear magnetic resonance (NMR)

186

2+

spectroscopic analysis of C-peptide, as well as the Zn /C-peptide complex should be obtained.
NMR is commonly used for the structural analysis of proteins and peptides,[41, 42] and could
be used to determine Zn

2+

induced conformational changes, and even elucidate specific binding

regions. If it is ultimately shown that Zn

2+

does not bind C-peptide, other interactions are

possible that may result in C-peptide bioactivity. For example, C-peptide may not itself be
responsible for bioactivity. Instead, zinc may be responsible for this observed bioactivity and Cpeptide could simply be a carrier of the metal, responsible for delivering zinc to various cell
types and thereby facilitating zinc transport through the bloodstream and into cells. This
hypothesis is supported by the vast and fundamental physiological roles of zinc, including links
to diabetes and nitric oxide production.[43-45]

187

Figure 5.2 - The amino acid sequence of C-peptide is shown, which arrows highlighting residues
2+

that are negatively charged in vivo, and thus likely capable of binding free Zn ions.

188

2+

5.2.2.2 - Investigation of Zn /C-peptide Activity In Vivo
Currently, several animal models of human type 1 diabetes exist.[46] To conclusively show
2+

the important potential bioactivity of C-peptide in the presence of Zn , C-peptide should be
administered alone, or co-administered with Zn

2+

and the resultant physiological effects

determined. Resulting improvements in blood pressure, nerve conduction, inflammation, etc.
2+

could be measured and attributed to Zn -activated C-peptide. This study would be the first to
2+

determine the potential for Zn -activated C-peptide as a drug for type 1 diabetes. Also, the
necessity of zinc activation can truly be evaluated in a comparative clinical investigation for the
first time. This investigation would be the most fundamental and conclusive method for
determining the physiological effects of C-peptide, as well as the necessity of zinc activation.

5.2.3 - Endothelial Permeability to C-peptide
5.2.3.1 - Choosing Endothelial Cell Type
The work presented earlier revealing the ability of C-peptide to permeate a cultured
pulmonary endothelium is novel in the sense that it is the first investigation into the potential
ability of C-peptide to escape the bloodstream in vivo, and is, therefore, the first evidence
suggesting that C-peptide does indeed leave the bloodstream in vivo. Since these results
suggest that C-peptide may have a much more comprehensive role in physiology and health
than is currently assumed, it is an important first step in determining the true scope of Cpeptide’s physiological influence throughout the body. However, to determine the true scope
of biological activity, more research is required. These findings suggest that C-peptide is capable
of permeating the pulmonary vasculature. However, it is well known that endothelial cells form

189

barriers of varying integrity, depending on their function in a specific organ or anatomical
location.[47] In fact, endothelial permeability has been shown to be heterogeneous within the
same organ.[48] Therefore, endothelial cell types that are specific to the particular organ or
anatomical location influenced by C-peptide should be investigated to determine their
particular permeability to C-peptide.
For example, C-peptide has been shown to improve diabetic neuropathy,[49] suggesting
that C-peptide may be able to leave the bloodstream to enter the brain and directly improve
nerved function. However, the endothelial cells lining blood vessels in the brain are highly
specialized, forming an endothelial barrier called the blood brain barrier, which is far less
permeable than endothelial layers found elsewhere in the body.[50] In consideration of this, in
order to examine whether C-peptide directly facilitates improvements in nerve conduction and
myelination, brain microvascular endothelial cells should be used to identify their permeability
to C-peptide.

5.2.3.2 - Kinetics of Permeation and Comparisons
The kinetics of permeation of a particular analyte can be measured by the permeation
coefficient of a particular analyte. The permeation assays reported in chapter 4 were
performed merely to determine whether C-peptide was capable of penetrating a cultured
endothelium. Since it was determined that C-peptide is capable of permeating the cultured
pulmonary endothelium, permeation coefficients of C-peptide, as well as another well studied
tracer molecules should be determined. By determining two permeation coefficients, the
relative kinetics of permeation can be established, allowing for comparative studies. Several
tracer molecules have been identified, such as phenol red and FITC-albumin, which are semi-

190

impermeable to endothelial monolayers.[51, 52] An experiment to support the results showing
that C-peptide is capable of permeating an endothelium would be a comparative permeability
assay, where the permeation coefficient of C-peptide obtained and compared to a semiimpermeable tracer molecule. If the permeability coefficient of C-peptide is higher than that of
the semi-impermeable tracer, then it would be expected that C-peptide is able to permeate
that particular endothelial cell type in vivo. However, since there exists no true negative control
for endothelial permeation, or in other words, there is not a specific tracer molecule that is
truly incapable of permeating endothelial cells in vitro, these semi-impermeable tracer
molecules are generally used to monitor changes, or differences, in endothelial barrier
integrity.

5.2.3.3 - Permeability of C-peptide Mutants
While a specific receptor for C-peptide has yet to be identified, it has been previously
shown that C-peptide is able to specifically bind to endothelial cells by a binding event that is
saturated at around 1 nM concentrations, which is of importance since the transport of
macromolecules across the endothelium is generally accomplished by a transcellular and
specific mechanism.[53] Also, C-peptide has been shown to be internalized by endosomes to
enter human endothelial cells,[54] by a process which can be inhibited by nocodazole and
monodansylcadaverine (MDC), signifying that C-peptide enters the cell by a mechanism which is
receptor mediated and requires microtubule assembly. Interestingly, microtubules have been
shown to facilitate transcellular transport and influence endothelial permeability and
integrity.[55-58] These results suggest that there is a receptor responsible for trafficking Cpeptide across the endothelium. In data not presented here, endothelial permeability to C-

191

2+

peptide was shown not to be dependent on the presence of Zn , suggesting that the proposed
receptor may be promiscuous. However, in order to gain insight as to whether the mechanism
of C-peptide transport is receptor mediated, the endothelial permeability to C-peptide mutants
can be determined. Since receptors are often specific, the endothelial permeation of a Cpeptide mutant may be halted or hindered due to inability to specifically bind the supposed
receptor.

192

REFERENCES

193

REFERENCES
1.

Kim, Y., W.C. Messner, and P. LeDuc, Disruptive Microfluidics: From Life Sciences to
World Health to Energy. Disruptive Sciences and Technology, 2012. 1(1): p. 41-53.

2.

Whitesides, G.M., The origins and the future of microfluidics. Nature, 2006. 442(7101):
p. 368-73.

3.

Szita, N., et al., Microfluidic approaches for systems and synthetic biology. Curr Opin
Biotechnol, 2010. 21(4): p. 517-23.

4.

Shuler, M.L. and M.B. Esch, Body-on-a chip: Using microfluidic systems to predict human
responses to drugs. Pure and Applied Chemistry, 2010. 82(8): p. 1635-1645.

5.

Genes, L., I., et al., Adressing a Vascular Endothelium Array with Blood Components
using Underlying Microfluidic Channels. Lab on a Chip, 2007. 7(10): p. 1256-1259.

6.

Price, A.K., R.S. Martin, and D.M. Spence, Circulation of erythrocytes through
microfluidic channels: Investigations toward a circulatory vessel network on-chip.
Abstracts of Papers, 229th ACS National Meeting, San Diego, CA, United States, March
13-17, 2005, 2005: p. ANYL-208.

7.

Spence, D.M., Automation and Microfluidic Assays: In Vitro Models of the Mammalian
Microcirculation. Journal of the Association for Laboratory Automation, 2005. 10(4): p.
270-275.

8.

Carrel, A., Pure Cultures of Cells. J Exp Med, 1912. 16(2): p. 165-8.

9.

Carrel, A. and M.T. Burrows, Cultivation of adult tissues and organs outside of the body.
Journal of the American Medical Association, 1910. 55: p. 1379-1381.

10.

Fischer, A., Cultures of Organized Tissues. J Exp Med, 1922. 36(4): p. 393-7.

11.

Fischer, A., A Three Months Old Strain of Epithelium. J Exp Med, 1922. 35(3): p. 367-72.

12.

Curtis, A.S. and M. Varde, Control of Cell Behavior: Topological Factors. J Natl Cancer
Inst, 1964. 33: p. 15-26.

13.

Abercrombie, M. and J.E. Heaysman, Observations on the social behaviour of cells in
tissue culture. II. Monolayering of fibroblasts. Exp Cell Res, 1954. 6(2): p. 293-306.

14.

Weiss, P., Cell Contact, in International Review of Cytology, G.H. Bourne and J.F. Danielli,
Editors. 1958, Academic Press. p. 391-423.

15.

Kaihara, S., et al., Silicon micromachining to tissue engineer branched vascular channels
for liver fabrication. Tissue Eng, 2000. 6(2): p. 105-17.

194

16.

Andersson, H. and A. van den Berg, Microfluidic devices for cellomics: a review. Sensors
and Actuators B-Chemical, 2003. 92(3): p. 315-325.

17.

Young, E.W., et al., Technique for real-time measurements of endothelial permeability in
a microfluidic membrane chip using laser-induced fluorescence detection. Anal Chem,
2010. 82(3): p. 808-16.

18.

Ferrell, N., et al., A microfluidic bioreactor with integrated transepithelial electrical
resistance (TEER) measurement electrodes for evaluation of renal epithelial cells.
Biotechnol Bioeng, 2010. 107(4): p. 707-16.

19.

Douville, N.J., et al., Fabrication of two-layered channel system with embedded
electrodes to measure resistance across epithelial and endothelial barriers. Anal Chem,
2010. 82(6): p. 2505-11.

20.

Booth, R. and H. Kim, Characterization of a microfluidic in vitro model of the blood-brain
barrier (muBBB). Lab Chip, 2012. 12(10): p. 1784-92.

21.

Vogel, P.A., et al., Microfluidic transendothelial electrical resistance measurement device
that enables blood flow and postgrowth experiments. Anal Chem, 2011. 83(11): p. 4296301.

22.

Lindstrom, K., et al., Acute effects of C-peptide on the microvasculature of isolated
perfused skeletal muscles and kidneys in rat. Acta Physiol Scand, 1996. 156(1): p. 19-25.

23.

Hansen, A., et al., C-peptide exerts beneficial effects on myocardial blood flow and
function in patients with type 1 diabetes. Diabetes, 2002. 51(10): p. 3077-3082.

24.

Schalkwijk, C.G. and C.D. Stehouwer, Vascular complications in diabetes mellitus: the
role of endothelial dysfunction. Clin Sci (Lond), 2005. 109(2): p. 143-59.

25.

Johansson, B.L., J. Wahren, and J. Pernow, C-peptide increases forearm blood flow in
patients with type 1 diabetes via a nitric oxide-dependent mechanism. Am J Physiol
Endocrinol Metab, 2003. 285(4): p. E864-70.

26.

Meyer, J.A., et al., Metal-activated C-peptide facilitates glucose clearance and the
release of a nitric oxide stimulus via the GLUT1 transporter. Diabetologia, 2008. 51(1): p.
175-82.

27.

Ignarro, L.J., et al., Endothelium-derived relaxing factor from pulmonary artery and vein
possesses pharmacologic and chemical properties identical to those of nitric oxide
radical. Circulation Research, 1987. 61(6): p. 866-79.

28.

Brooks, P., J. Meyer, and D.M. Spence, Metal (Zn2+) Activated C-Peptide Induced ATP
Release Via Red Blood Cells with the Addition of Albumin. Abstracts, 60th Southeast

195

Regional Meeting of the American Chemical Society, Nashville, TN, United States,
November 12-15, 2008: p. SERM-680.
29.

Medawala, W., et al., A Molecular Level Understanding of Zinc Activation of C-peptide
and its Effects on Cellular Communication in the Bloodstream. Rev Diabet Stud, 2009.
6(3): p. 148-58.

30.

Meyer, J.A. and D.M. Spence, A perspective on the role of metals in diabetes: past
findings and possible future directions. Metallomics, 2009. 1(1): p. 32-41.

31.

Burnstock, G., Purinergic signalling. Br J Pharmacol, 2006. 147 Suppl 1: p. S172-81.

32.

Sprague, R.S., et al., ATP: the red blood cell link to NO and local control of the pulmonary
circulation. Am. J. Physiol, 1996. 271: p. H2717-H2722.

33.

Price, A., K., et al., Deformation-induced release of ATP from erythrocytes in a
poly(dimethylsiloxane)-based microchip with channels that mimic resistance vessels.
Analytical Chemistry, 2004. 76(16): p. 4849-55.

34.

Ellsworth, M.L., et al., Erythrocytes: Oxygen Sensors and Modulators of Vascular Tone.
Physiology, 2009. 24(2): p. 107-116.

35.

D'Amico Oblak, T., P. Root, and D.M. Spence, Fluorescence monitoring of ATPstimulated, endothelium-derived nitric oxide production in channels of a
poly(dimethylsiloxane)-based microfluidic device. Anal Chem, 2006. 78(9): p. 3193-7.

36.

Wahren, J., K. Ekberg, and H. Jornvall, C-peptide is a bioactive peptide. Diabetologia,
2007. 50: p. 503-509.

37.

Lindahl, E., et al., Proinsulin C-peptide regulates ribosomal RNA expression. J Biol Chem,
2010. 285(5): p. 3462-9.

38.

Regehr, K.J., et al., Biological implications of polydimethylsiloxane-based microfluidic cell
culture. Lab Chip, 2009. 9(15): p. 2132-9.

39.

Toepke, M.W. and D.J. Beebe, PDMS absorption of small molecules and consequences in
microfluidic applications. Lab Chip, 2006. 6(12): p. 1484-6.

40.

Wang, Y., et al., Benchtop micromolding of polystyrene by soft lithography. Lab Chip,
2011. 11(18): p. 3089-97.

41.

Baldwin, A.J. and L.E. Kay, NMR spectroscopy brings invisible protein states into focus.
Nat Chem Biol, 2009. 5(11): p. 808-14.

42.

Grzesiek, S. and H.J. Sass, From biomolecular structure to functional understanding: new
NMR developments narrow the gap. Curr Opin Struct Biol, 2009. 19(5): p. 585-95.

196

43.

Prasad, A.S., Clinical, immunological, anti-inflammatory and antioxidant roles of zinc.
Exp Gerontol, 2008. 43(5): p. 370-7.

44.

Pluth, M.D., E. Tomat, and S.J. Lippard, Biochemistry of mobile zinc and nitric oxide
revealed by fluorescent sensors. Annu Rev Biochem, 2011. 80: p. 333-55.

45.

Jansen, J., W. Karges, and L. Rink, Zinc and diabetes--clinical links and molecular
mechanisms. J Nutr Biochem, 2009. 20(6): p. 399-417.

46.

von Herrath, M. and G.T. Nepom, Animal models of human type 1 diabetes. Nat
Immunol, 2009. 10(2): p. 129-32.

47.

Pries, A.R. and W.M. Kuebler, Normal endothelium. Handb Exp Pharmacol, 2006(176 Pt
1): p. 1-40.

48.

Moldobaeva, A. and E.M. Wagner, Heterogeneity of bronchial endothelial cell
permeability. Am J Physiol Lung Cell Mol Physiol, 2002. 283(3): p. L520-7.

49.

Johansson, B.L., et al., Beneficial effects of C-peptide on incipient nephropathy and
neuropathy in patients with type 1 diabetes mellitus. Diabetic Medicine, 2000. 17(3): p.
181-189.

50.

Rubin, L.L. and J.M. Staddon, The cell biology of the blood-brain barrier. Annu Rev
Neurosci, 1999. 22: p. 11-28.

51.

Wang, Y. and J.S. Alexander, Analysis of endothelial barrier function in vitro. Methods
Mol Biol, 2011. 763: p. 253-64.

52.

Sun, M., et al., A dynamic real-time method for monitoring epithelial barrier function in
vitro. Anal Biochem, 2012. 425(2): p. 96-103.

53.

Rigler, R., et al., Specific binding of proinsulin C-peptide to human cell membranes. Proc.
Natl. Acad. Sci., 1999. 96: p. 13318-13323.

54.

Luppi, P., et al., C-peptide is internalised in human endothelial and vascular smooth
muscle cells via early endosomes. Diabetologia, 2009. 52(10): p. 2218-28.

55.

Minshall, R.D. and A.B. Malik, Transport across the endothelium: regulation of
endothelial permeability. Handb Exp Pharmacol, 2006(176 Pt 1): p. 107-44.

56.

Minshall, R.D., et al., Vesicle formation and trafficking in endothelial cells and regulation
of endothelial barrier function. Histochem Cell Biol, 2002. 117(2): p. 105-12.

57.

Lossinsky, A.S. and R.R. Shivers, Structural pathways for macromolecular and cellular
transport across the blood-brain barrier during inflammatory conditions. Review. Histol
Histopathol, 2004. 19(2): p. 535-64.

197

58.

Le Sueur, L.P., C.B. Collares-Buzato, and M.A. da Cruz-Hofling, Mechanisms involved in
the blood-brain barrier increased permeability induced by Phoneutria nigriventer spider
venom in rats. Brain Research, 2004. 1027(1-2): p. 38-47.

198