NOVEL APPROACHES TO TESTING GASTRO INTESTINAL
FUNCTION IN VITRO: CONTROLLING SIGNAL ACQUISITION,
TISSUE COMPOSITION, OR THE PLATFORM
BY
Dylan Thomas Knutson
A Thesis Submitted to the Graduate Faculty of
WAKE FOREST UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES
In Partial Fulfillment of the Requirements
for the Degree of
Master of Science
Biomedical Engineering
May, 2017
Winston-Salem, North Carolina
Approved by:
Khalil N. Bitar, Ph.D., AGAF, Advisor, Chair
Adam R. Hall, Ph.D
Aleksander Skardal, Ph.D
iii
ACKNOWLEDGEMENTS
I would like to thank my advisor, Dr. Khalil N Bitar for the opportunity to
work in his lab and full support. This project was his vision for me, and I thank
him for the guidance, expertise, mentorship. I would also like to thank my
committee, Dr. Adam R. Hall and Dr. Aleks Skardal for their willing collaboration,
mentorship, and commitment to join my committee. I am so grateful for your kind
help.
I am also quite thankful to both Dr. Prabhash Dadhich and Elie Zakhem for
their mentorship during my studies. Without their guidance, the success of my
project would not have been possible. Last, the support of family and friends
gave me the strength to continue pushing myself to work harder.
iv
TABLE OF CONTENTS
DEDICATION ............................................................................................................................................. II
ACKNOWLEDGEMENTS ........................................................................................................................... III
TABLE OF CONTENTS ............................................................................................................................... IV
LIST OF FIGURES ....................................................................................................................................... V
LIST OF TABLES ........................................................................................................................................ IX
LIST OF ABBREVIATIONS........................................................................................................................... X
INTRODUCTION ........................................................................................................................................ 1
CHAPTER I: DO INTERSTITIAL CELLS OF CAJAL PLAY A ROLE IN PYLORIC FUNCTION? ................................ 7
CHAPTER II: AN IN VITRO MODEL OF THE GASTRIC NEUROMUSCULAR APPARATUS USING ENGINEERED
PYLORUS TO UNDERSTAND GASTRIC PATHOPHYSIOLOGY ..................................................................... 25
CHAPTER III: ELECTROPHYSIOLOGY OF AN IN VITRO MODEL OF THE GASTRIC NEUROMUSCULAR
APPARATUS ........................................................................................................................................... 51
CHAPTER IV: MICRO-SENSITIVE MOLDED SILICONE TISSUE PILLAR PLATFORM, FABRICATED BY 3-D
PRINTING FOR SIMPLER PHYSIOLOGICAL STUDIES ................................................................................. 76
CHAPTER V: SUMMARY AND CONCLUSIONS ........................................................................................ 100
APPENDIX ............................................................................................................................................ 101
SCHOLASTIC VITA ................................................................................................................................. 102
v
LIST OF FIGURES
Figure 1- An explanation on the methods for analyzing tissues for slow-wave associated
contractions. Ultimately, the signal is filtered and can be analyzed for both amplitude and
frequency. 14
Figure 2- Cryosections of pyloric tissue, stained with DAPI (Blue) and ICC marker Ano1.
Several ICC were visualized in the tissue, including ICC-SM (submucosal), ICC-MY
(Myenteric), and ICC-IM (Intramuscular). Scale Bars are 100 micrometers. 16
Figure 3- An explanation on the methods for physiology. The thick pyloric sphincter is
isolated from the distal stomach (A). It is stretched and equilibrated in the tissue bath
(B), where an isometric force transducer senses strain and reports changes in stress
(newtons) from stimuli (C). The data is acquired real–time and sent to a computer (D). 17
Figure 4- A(A) Maximum Response to Ach and EFS with and without inhibitor T16Ainh-
A01 incubation. Contraction following exogenous ACh was significantly reduced (p<.05,
n=5). EFS-induced relaxation remained similar. (B) Tissue Response to incubation of
Inhibitor. There was a large increase in basal tone upon the incubation of T16. This
response makes the assessment of gross tissue responses unfair. 18
Figure 5- One minute segments from baseline recordings. It was observed that small
oscillations were being reduced when adding the inhibitor. This would be investigated
quantitatively. 19
Figure 6- Frequency and amplitude with and without inhibition. There was a significant
decrease in both frequency and amplitude of phasic contraction when incubated with
T16Ainh-A01 (25% and 20% respectively, n=5). 20
Figure 7- Characterization of isolated GFP-ICC in Culture. (A) ICCs attached and
acquired a star-shaped morphology. (B) ICCs stained positive for calcium-activated
chloride channel Ano1 confirming a population of isolated ICC (red). DAPI stained nuclei
with blue. All scale bars are 100 µm. 36
Figure 8- Microscopic Evaluation of Engineered Constructs. (A) Engineered pylorus
formed dense muscular tissues around the central post in culture vessel. (B) GFP-ICC
was visualized in the engineered tissues. (C) ICC formed networks within the tissues. All
scale bars are 100 µm. 37
Figure 9- Organ Bath Studies: Basal Tone and KCl: (A trace, before dashed line) All
engineered pylorus constructs established spontaneous basal tone. (B) Tone was similar
in maximum, area, and the rate of establishment (p>.05, n=5) among all constructs. (A,
after dashed line) KCl induced tissue contraction. Tissues contracted to a similar extent
(maximum, area, and the rate (p>.05, n=4)). 38
Figure 10- Organ Bath Studies: Electrical Field Stimulation EFS: EFS induced smooth
muscle relaxation in (A) SMC + NS + ICC and (C) SMC + NS constructs only. There was
vi
no relaxation in (B) SMC + ICC or (D) SMC only constructs. (E) EFS-induced relaxation
was significantly larger in SMC + NS + ICC than all others (1.9x, p<.05, n=5). 40
Figure 11- Organ Bath Studies: Inhibitors for EFS response: TTX abolished EFS-
induced relaxation in the engineered constructs (p<.05, n=3). nNOS neuronal blocker L-
NAME yielded similar extent of inhibition in the SMC + NS + ICC and SMC + NS
constructs, confirming there was a large functional population of differentiated NO donor
neurons (p>.05, n=3). 41
Figure 12- Kinetics EFS-Induced Relaxation: All relaxation responses over time were
averaged, computed and graphed. (A) A representative EFS-induced relaxation graph in
SMC + NS + ICC (purple trace) and SMC + NS (blue trace) constructs. EFS induced
relaxation was faster in (B) area and (C) rate in the SMC + NS + ICC constructs
compared to SMC + NS constructs (1.7x n=5, p=.30 and 2.5x, n=5 p=.12 respectively).
42
Figure 13 L-NAME inhibition of EFS-Induced relaxation: L-NAME completely abolished
EFS-induced relaxation in (A) SMC + NS constructs (dashed blue line) and (B) SMC +
NS + ICC constructs (dashed purple line). 44
Figure 14- Leads for detecting a differential current on the engineered tissue within the
organ bath. 60
Figure 15- Raw recordings from force transducer (blue trace) and/or electrodes (voltage,
mV, green) of within tissue bath. It was necessary for the electrodes to remain in the
bath; otherwise a differential current could not be achieved (arrowed, A) through tissue.
Voltage pulses from EFS (5V, 5Hz, 0.5ms) were detected upon stimulation (B). Without
tissue or electrical activites, the liquid reading from electrodes remained approximately 0
volts (C). 61
Figure 16- Recording from electrodes (voltage, mV, green) of SMC + ICC + NS within
tissue bath. Perturbances of either washing (A, after dashed line) or adding reagents (B,
highlighted) could obscure recordings for a short period. Washing increased perceived
voltage during and for several seconds following. Adding reagents sometimes resulted in
a voltage drop. Analyses and tests were performed following. 62
Figure 17- Recording from electrodes (voltage, mV, green) of SMC only tissue within
tissue bath. Perturbances of either washing (A, after dashed line) or adding reagents (B,
highlighted) could obscure recordings for a short period. Only with tissue did voltage
exceed 1 mV in these tests, and nifedipine blocked electrical activity recorded from the
tissues. Therefore the recordings of activity from the bath are accurate. 63
Figure 18- Raw simultaneous recordings of force (blue trace) and voltage of engineered
tissues within tissue bath. Both graphs show establishment of basal tone, with expected
electrical spikes (arrowed, green trace) preceeding contractions (arrowed, blue trace).
Engineered tissue with ICC had a higher frequency and amplitude, were more rhythmic.
64
Figure 19 - Raw simultaneous recordings of engineered neuromuscular tissues from
EFS response within tissue bath. Scales (force and voltage) are NOT the same. EFS
Response Is much larger, and more immediate in ICC tissues Electrical Changes: SMC
+ NS- increased peak activities following EFS, then nothing. SMC + NS + ICC, slowing
vii
of rhythm, maintain rhythm. Upon the release of NO, ICC slow waves may reduce in
order to facilitate relaxation from tonic state. 67
Figure 20- Spectral analyses graphs of electrical Activity of both SMC + NS and SMC +
NS + ICC tissues 69
Figure 21- Final CAD drawings to create Stomach on a chip design. Features: (1) Top
ABS plate provides windowed view, inlet, and outlet, and thermal resistance. (2) PDMS
Center channel protects 3-D printed parts from leaks, and width for low stress, lowered
flow around tissues. Cutouts allow for gating when culturing tissues. (3) 3 sets of PDMS
tissue pillars allow for the combination of 3 different tissues, and report strains visually to
quantify force. (4) Bottom plate allows sufficient room for added elements and thermal
resistance. 85
Figure 22- CAD drawings are printed using a desktop 3-D printer. Molds for pillars (A)
and bath walls (B) fit into stage (C, shown with part of mold component A). After printing,
M5 screws are tapped through the bottom of the stage for raising molded substrate out
of the assembly. Some bonding surfaces are sanded or epoxy-coated for smooth casts.
The complete assembly allows for the molding of either the pillars or walls in the stage
(D). 85
Figure 23- CAD drawings are printed using a desktop 3-D printer. Final products are
rigid, accurate, and fit together tightly. Right: top plate and bottom plate before
assembly. Lower Left: Assembled molds for PDMS products. 86
Figure 24- The stage allows for the fitment of either the pillar (A) or bath wall mold (B). 87
Figure 25- Molding process to make tissue pillars- (1) desired negative mold is placed in
stage and PDMS is poured over mold. (2) After curing, molded piece and negative are
driven out of stage by set screws. (3) The base is loosely peeled from face of product.
(4) For tissue pillars, the negative mold is disassembled into pieces. (5) Molded pillars
are revealed. (6) A PDMS base with a single set of pillars. 87
Figure 26- All components of the assembly. The bath with pillars (above) or the full
assembly for fluidic experiments (below). 88
Figure 27- (A) Gates were produced from 3D printed mold which fit into slotted walls to
isolate/separate hydrogels cultured in vessel. (B) Gates isolated 1 mL of 1.9 mg/mL
Collagen hydrogel. Gates could be removed after 40 minutes of gelation at 37˚C to add
medium. 89
Figure 28- (A) The bath with pillars representing the assembly to be characterized by
formula, B. Three mixtures of hardener to elastomer were characterized (C). Ultimately
the 1:30 mixture was chosen, and a linear curve was plotted to represent the relationship
of force to pillar distension (D). 90
Figure 29- Microscopy of tissues in culture reveal muscle cells are captured in the tissue,
and are aligned between the silicone pillars. Staining with Smooth muscle Actin
confirmed alignment of muscle filaments. 91
Figure 30- Calibration using a set of masses was used to deflect the pillars and
determine Hooke’s constant for each. 92
Figure 31- A Stereomicroscope can measure the deflection of the pillars in response in
order to calculate forces (A). A force transducer measures strain and reports force
generated by tissues over time (B). The basal tone measured by both devices was
viii
similar (C, n=3, paired, p>.05), but the measured response to KCl was less by the pillar
baths (D). 93
Figure 32- Comparison of the KCl response to sphincteric and non-sphincteric (small
intestine) muscle tissues (IAS). Tissues were washed with fresh buffer, and 60mM KCl
was added following equilibration period. Sphincteric tissues exerted significantly more
force within the device compared to non-sphincteric (n=3-4, p<.05). Therefore the
contractility of muscle when depolarized is different between sphincteric and non-
sphincteric muscle. 94
Figure 33- KCl responses from 3 human pylorus sphincteric tissues. Force from the
tissues was measured every 15 seconds while electrical activity was measured
continuously. Large spikes in EMG activity are associated with the onset of contraction.
Larger or more frequent activites resulted in higher measured forces. 95
ix
LIST OF TABLES
Table 1- Summary of the maximum, duration, and rate of basal tone and KCl
contractions. 39
Table 2- Summary of the minimum, duration, and rate of EFS-induced relaxation in the
absence and presence of inhibitors. 44
x
LIST OF ABBREVIATIONS
Ach Acetylcholine
CAD Computer aided design
ChAT Choline acetyltransferase
DMEM Dulbecco’s modified Eagle medium
EFS Electrical Field Stimulation
ENS Enteric nervous system
GI Gastrointestinal
HBSS Hank’s balanced salt solution
KCl Potassium chloride
LES Lower esophageal sphincter
L-NAME Nω-Nitro-L-arginine methyl ester hydrochloride
nNOS neuronal nitric oxide synthase
PBS Phosphate buffer saline
SEM Scanning electron microscopic
TTX Tetrodotoxin
VIP Vasoactive Intestinal Peptide
1
INTRODUCTION
Motility in the gut involves a complex interplay of tissues and cells to break
down and propel matter throughout each segment. Propulsion is orchestrated by
a complex called the neuromuscular apparatus which includes smooth muscle,
enteric neurons, and interstitial cells of Cajal (ICC) (Kenton M. Sanders, Hwang,
& Ward, 2010). Disruption of this apparatus characterizes some gastrointestinal
(GI) diseases, however both the etiology and pathophysiology of dysmotility
patterns are not well understood (Goldstein, Thapar, Karunaratne, & De Giorgio,
2016). Gastroparesis, or delayed gastric emptying, is a disease in which patients
are left with debilitating symptoms such as nausea and vomiting, early satiety
and malnutrition, and general discomfort. Treatment options are limited and
include surgeries, prokinetics, and pacemaking devices (Bielefeldt, 2012). These
interventions have limited efficacy and are only effective on specific subsets of
patients.
Depletions of ICC and/or enteric neurons have been associated with
clinical cases of gastroparesis. Histological analysis shows patients may be
missing one or both cell types (Grover et al., 2011). It is estimated that at least
50% of Type 1 diabetics and 30% of Type 2 develop this disease. It is well known
that diabetics suffer from neuropathies and other complicated symptoms, but a
significant amount of gastroparesis patients have no other diseases present and
are classified as ‘idiopathic’. Newer studies have mapped altered electrical
waveform patterns, recorded associated symptoms, and the extent of ICC
2
depletions in gastric tissues (Angeli et al., 2015). Furthermore, the specific
location of cell depletions is associated with different symptoms and disease
phenotypes (Moraveji et al., 2016). However, at present there are no cause and
effect studies that can attribute the effect of these cell depletions to disease
pathophysiology.
Furthermore, it is established that ICC generate electrical activity, which
regulates and transmits muscle potentials for contraction. Yet, the role that both
ICC and enteric neurons perform in transmitting cholinergic and nitrergic signals
to direct muscle motility is very controversial (Kenton M Sanders, Kito, Hwang, &
Ward, 2016; Wood, 2016). Some investigators argue that ICC have no role in
transducing responses (Chaudhury, 2016), while others have demonstrated that
they enhance it (Lies et al., 2014). Although tremendous strides have been made
in investigating the role of these cells using clinical populations, isolated tissues,
and gene knockouts (Mashimo, Kjellin, & Goyal, 2000; Sivarao, Mashimo, &
Goyal, 2008; Wagner, Sullins, & Dunn, 2014; Zarate et al., 2003), there is still a
lack of in vitro models to quantitatively deplete and reinstate cell populations in
order to relate cell depletions and measure functional changes in tissues.
In the thesis, provided first is a novel tissue engineering approach to
develop an in vitro model of gastric neuromuscular tissues using different
combinations of sphincteric smooth muscle, enteric neurons, and for the first
time- ICC. Previously, autologous intrinsically innervated pyloric sphincters were
bioengineered using human pyloric smooth muscle and neural progenitor cells.
3
The cells in the constructs demonstrated normal phenotypes, muscular
alignment, and function (Rego, Zakhem, Orlando, & Bitar, 2015). The first goal
of this study was to incorporate ICC into the model. The second goal was to
analyze the contribution of each cell type to basic functions. Physiology of the
engineered constructs showed a specific contribution from each cell type. (i)
Tone was myogenic trait and independent of neuronal or ICC contribution, (ii)
relaxation of tone was dependent on a functional population of nitrergic neurons,
(iii) ICC significantly increased the neural-mediated relaxation in both peak and
rate, and increased electrical rhythms. This study provides a promising new
model to understand, generate, and study new hypotheses regarding
gastrointestinal pathophysiology.
Similarly, mechanisms of gastric emptying disorders are characterized by
an inability for tissue relaxation such as pylorospasm or diabetic neuropathies
and are associated with abnormal manometric data and gastroparesis (Camilleri,
2016). Studying persistent relaxation in ICC knockouts tissues led to the
dismissal of the role of ICC in transducing muscle relaxations (Huizinga et al.,
2008). We confirmed that while these functions indeed persist, the addition of
ICC changes the magnitude of the relaxation response significantly. The
magnitude of relaxation may be an important mechanism for accomodation and
expedition of gastric emptying. Therefore, the ‘persistence’ of responses may not
be physiologically relevant, and so the specific character of these responses
should be investigated before ruling out these key cells.
4
The final aim of the thesis was to develop an open physiology platform to
make studying tissues more accessible and to test innovative hypotheses
regarding functional tissue interactions. In biomedical research, the study of
molecular markers is well characterized, but physiological analysis of tissues
remains less investigated. This is because testing physiological responses takes
tremendous timing, resources for equipment, and multidisciplinary skills. Clever
designs using silicone micro pillars which sense tissue forces have emerged, but
they are too small for simple culture, require complex lithography to produce, and
enhanced imaging using fluorescence to detect forces (Boudou et al., 2011). The
present investigation provides a simpler open-source platform for 3-D printing
molds to make micro-sensitive pillars in a novel fluidic bath. The pillar device
allows for controlled engineering of multiple muscle tissues, and records
muscular tone and response phenomena similar to expensive physiology
equipment. Further novelty includes the capacity for the interaction of several
tissues, measurement of electrical or other phenomena simultaneously, and flow
perfusion for studies like high throughput drug testing on functional tissues.
In conclusion, this study provides novel insights on the effects of gastric
cell depletions on physiological function and offers more accessible and
hypothesis generating platforms which greatly impact future research
propositions and the rectification of disease pathophysiology.
REFERENCES
Angeli, T. R., Cheng, L. K., Du, P., Wang, T. H.-H., Bernard, C. E., Vannucchi, M.-G., . . . O’Grady, G. (2015). Loss of Interstitial Cells of Cajal and Patterns of Gastric Dysrhythmia in Patients
5
With Chronic Unexplained Nausea and Vomiting. Gastroenterology, 149(1), 56-66.e55. doi:http://dx.doi.org/10.1053/j.gastro.2015.04.003
Bielefeldt, K. (2012). Gastroparesis: Concepts, Controversies, and Challenges. Scientifica, 2012, 19. doi:10.6064/2012/424802
Boudou, T., Legant, W. R., Mu, A., Borochin, M. A., Thavandiran, N., Radisic, M., . . . Chen, C. S. (2011). A microfabricated platform to measure and manipulate the mechanics of engineered cardiac microtissues. Tissue Engineering Part A, 18(9-10), 910-919.
Camilleri, M. (2016). Novel Diet, Drugs, and Gastric Interventions for Gastroparesis. Clinical Gastroenterology and Hepatology, 14(8), 1072-1080. doi:http://dx.doi.org/10.1016/j.cgh.2015.12.033
Chaudhury, A. (2016). Furthering the debate on the role of interstitial cells of Cajal in enteric inhibitory neuromuscular neurotransmission. American Journal of Physiology - Cell Physiology, 311(3), C479-C481. doi:10.1152/ajpcell.00067.2016
Goldstein, A. M., Thapar, N., Karunaratne, T. B., & De Giorgio, R. (2016). Clinical aspects of neurointestinal disease: Pathophysiology, diagnosis, and treatment. Developmental Biology, 417(2), 217-228. doi:http://dx.doi.org/10.1016/j.ydbio.2016.03.032
Grover, M., Farrugia, G., Lurken, M. S., Bernard, C. E., Faussone–Pellegrini, M. S., Smyrk, T. C., . . . Pasricha, P. J. (2011). Cellular Changes in Diabetic and Idiopathic Gastroparesis. Gastroenterology, 140(5), 1575-1585.e1578. doi:http://dx.doi.org/10.1053/j.gastro.2011.01.046
Huizinga, J. D., Liu, L. W., Fitzpatrick, A., White, E., Gill, S., Wang, X.-Y., . . . Starret, T. (2008). Deficiency of intramuscular ICC increases fundic muscle excitability but does not impede nitrergic innervation. American Journal of Physiology-Gastrointestinal and Liver Physiology, 294(2), G589-G594.
Lies, B., Gil, V., Groneberg, D., Seidler, B., Saur, D., Wischmeyer, E., . . . Friebe, A. (2014). Interstitial cells of Cajal mediate nitrergic inhibitory neurotransmission in the murine gastrointestinal tract. American Journal of Physiology - Gastrointestinal and Liver Physiology, 307(1), G98-G106. doi:10.1152/ajpgi.00082.2014
Mashimo, H., Kjellin, A., & Goyal, R. K. (2000). Gastric stasis in neuronal nitric oxide synthase–deficient knockout mice. Gastroenterology, 119(3), 766-773.
Moraveji, S., Bashashati, M., Elhanafi, S., Sunny, J., Sarosiek, I., Davis, B., . . . McCallum, R. (2016). Depleted interstitial cells of Cajal and fibrosis in the pylorus: novel features of gastroparesis. Neurogastroenterology & Motility.
Rego, S. L., Zakhem, E., Orlando, G., & Bitar, K. N. (2015). Bioengineered human pyloric sphincters using autologous smooth muscle and neural progenitor cells. Tissue Engineering Part A, 22(1-2), 151-160.
Sanders, K. M., Hwang, S. J., & Ward, S. M. (2010). Neuroeffector apparatus in gastrointestinal smooth muscle organs. The Journal of Physiology, 588(23), 4621-4639. doi:10.1113/jphysiol.2010.196030
Sanders, K. M., Kito, Y., Hwang, S. J., & Ward, S. M. (2016). Regulation of Gastrointestinal Smooth Muscle Function by Interstitial Cells. Physiology, 31(5), 316-326.
Sivarao, D. V., Mashimo, H., & Goyal, R. K. (2008). Pyloric sphincter dysfunction in nNOS−/− and W/W V mutant mice: animal models of gastroparesis and duodenogastric reflux. Gastroenterology, 135(4), 1258-1266.
Wagner, J. P., Sullins, V. F., & Dunn, J. C. Y. (2014). A novel in vivo model of permanent intestinal aganglionosis. Journal of Surgical Research, 192(1), 27-33. doi:http://dx.doi.org/10.1016/j.jss.2014.06.010
6
Wood, J. D. (2016). Enteric Nervous System: Neuropathic Gastrointestinal Motility. Digestive Diseases and Sciences, 61(7), 1803-1816. doi:10.1007/s10620-016-4183-5
Zarate, N., Mearin, F., Wang, X., Hewlett, B., Huizinga, J., & Malagelada, J. (2003). Severe idiopathic gastroparesis due to neuronal and interstitial cells of Cajal degeneration: pathological findings and management. Gut, 52(7), 966-970.
7
CHAPTER I: DO INTERSTITIAL CELLS OF CAJAL PLAY A
ROLE IN PYLORIC FUNCTION?
Dylan T. Knutson1,2, B.S., Elie Zakhem2, Ph.D, Khalil N. Bitar1,2,3, Ph.D., AGAF.
1Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences, Winston Salem, NC
2Wake Forest Institute for Regenerative Medicine, Wake Forest School of Medicine, Winston Salem, NC
3Section on Gastroenterology, Wake Forest School of Medicine, Winston-Salem, NC, USA
This chapter explores the physiological contribution of ICC in ex vivo Gastric tissues
8
ABSTRACT
BACKGROUND
Ano1 (TMEM16A) is a Ca2+-activated Cl− channel on Interstitial Cells of
Cajal (ICC) which regulates gastrointestinal motility. Absences of ICC and
consequently Ano1 are associated with several motility disorders, since they
pace smooth muscle with electrical conduction called the slow-wave. Several
Ano1 inhibitors have been identified such as T16Ainh-A01. Using the inhibitor has
therapeutic potential to help control cancerous proliferation of cells
overexpressing Ano1; however, it also has been shown to reduce slow waves in
isolated ICC cultures.
AIM
To investigate the effect of Ano1 inhibitor, T16Ainh-A01, on pyloric
function, a gastrointestinal sphincter responsible for regulating the passage of
stomach contents into the duodenum.
METHODS
Whole rat pyloric sphincter was excised and placed in an organ bath to
measure force generation. The effect of T16Ainh-A01 (10μM) was added to the
buffer. The changes in frequency and amplitude of phasic contractions were
measured. Muscarinic response to Ach (10μM) and electrical field stimulation
(EFS) to induce relaxation was measured in organ bath.
9
RESULTS
Phasic contractions in normal pylorus indicated a frequency of 4.2 +/- 0.25
contractions per minute. T16Ainh-A01 significantly reduced the frequency of
phasic contractions of the pyloric tissue to 3.2 +/- 0.11 (72% of control, n=5, P <
.05, t-test). The amplitude of phasic contractions was reduced. Muscarinic
response from Acetylcholine also decreased with the presence of T16Ainh-A01
(n=5). Pyloric relaxation remained similar with T16Ainh-A01 (n=4). We think the
effect of the inhibitor was not specific to ICC only. Some physiological responses
were retained, but reduced rhythmic contractions suggest the underlying slow
wave was being diminished with T16Ainh-A01.
CONCLUSION
The Ano1 inhibitor T16Ainh-A01 inhibited Ca2+-activated Cl− currents,
reduced the phasic contractions, increased tone, and altered some
neurotransmitter responses in healthy pyloric tissue. These data support the
notion that T16Ainh-A01 could alter pyloric function, inhibit ICC function in tissue,
and consequently gastric emptying in therapeutic uses. However, better models
must be made to further understand the ICC contribution.
Keywords: Interstitial cells of Cajal, ICC, TMEM16A, T16Ainh-A01, cancer,
pylorus
10
INTRODUCTION:
Ano1 (TMEM16A) is a Ca2+-activated Cl− channel on Interstitial Cells of
Cajal (ICC) which regulates gastrointestinal motility by activating pacemaker
currents (Zhu, Sung, O'Driscoll, Koh, & Sanders, 2015). Absences of ICC are
associated with several motility disorders, since they pace smooth muscle with
electrical conduction, also called the ‘slow wave’. Recently, it has also been
shown that Ano1 is responsible for regulating cellular proliferation and is up-
regulated in several forms of gastrointestinal cancer. Therefore, several Ano1
inhibitors have been identified such as T16Ainh-A01 (Hwang, Basma, Sanders, &
Ward, 2016; Mazzone et al., 2012). Using the inhibitor has therapeutic potential
to help control cancerous proliferation of cells overexpressing Ano1. However, it
also has been shown to reduce slow waves in isolated ICC cultures. Thus, the
inhibitor’s effects on whole gastrointestinal tissue function should be studied as
this may provide insight on the contribution of ICC within gastric tissues. The role
that ICC perform in transmitting cholinergic and nitrergic signals and regulating
muscular tissue function is controversial (Sanders, Kito, Hwang, & Ward, 2016;
Wood, 2016).
In the stomach, motility is dependent on the effective accomodation,
trituration, and emptying of reduced foods into the duodenum- carried out by the
fundus, antrum, and pylorus respectively. It is thought that phasic muscular
patterns are driven by electrical activity that is propagated by ICC, which
originate in the greater curvature of the stomach and diminish after the pylorus
11
(O'Grady et al., 2010). The pylorus contracts and maintains tone to facilitate
mixing, and subsequently relaxes to enhance emptying of broken down particles.
In the present investigation, the purpose was to inhibit ICC and study the
function of pyloric tissues in an organ bath. The first goal of this study was to
ensure the presence of ICC in the isolated pylorus. The second goal was to
analyze the contribution of ICC to basic pyloric physiology. The results show that
the Ano1 inhibitor T16Ainh-A01 inhibited Ca2+-activated Cl− currents, reduced the
phasic contractions associated with ICC, and altered neurotransmitter responses
in healthy pyloric tissue. These data support the notion that T16Ainh-A01 could
alter pyloric function and consequently gastric emptying in therapeutic uses.
12
MATERIALS AND METHODS:
Pylorus Tissues
Pylori from rat were removed by sharp dissection. Pylorus tissues were
manually cleaned by removing fat and mucosa with a surgical blade. The pyloric
sphincter was studied in its entirety (the band of muscle). Tissues were
approximately 6-7mm in diameter and 3mm in width.
Preparation of Cryosections and Ano1 Staining
Cryosections were prepared as previously described (Chen et al., 2011).
Briefly, tissue was placed on filter paper in 4% paraformaldehyde (PFA) for 2
hours at room temperature, and then placed in 30% sucrose in PBS overnight at
4 °C. The following day, cryosections were prepared in cryomolds and O.C.T.
Tissues were stained with 1:200 Anti-TMEM16A (Abcam, Cambridge, UK)
overnight and then incubated with 1:200 secondary antibody (tritC) diluted in
blocking buffer for 2 hours at room temperature.
Physiological Analysis of Ex Vivo Pylorus
All different combinations of engineered pyloric constructs were analyzed
for physiological functionality. Constructs were hooked in an organ bath between
a fixed arm, and the measuring arm of an isometric, magnetoresistive force
transducer in a horizontal tissue bath (F10; Harvard Apparatus, Holliston, MA).
Force data was acquired using LabChart 7 software (ADInstruments, Colorado
Springs, CO). Constructs were maintained in 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES) buffered solution at 37°C throughout all
13
experiments. All force generation studies were performed after the establishment
of stable basal tone. Then, acetylcholine (ACh; 1 μM) and electrical field
stimulation (EFS; 5 Hz, 0.5 ms, for 30 seconds) were administered. Constructs
were washed between each treatment, incubated in fresh buffer and allowed to
return to baseline. Pretreatment with T16Ainh-A01 was administered 8 minutes
prior to EFS stimulation to inhibit ICC. GraphPad Prism 7.00 software for
Windows (GraphPad Prism 7.00, San Diego, CA) was used to analyze collected
data. Second-order Savitsky–Golay smoothing was applied. Quantification of
physiologic data was performed relative to basal tone for contraction and
relaxation as maximum/minimum peak response (Delta Force of basal tone),
area (magnitude * time), and rate (linear regression from onset to peak
response).
14
Filtering Slow Wave Associated Contractions
Figure 1
As shown in Figure 1, in order to analyze the amount of slow-wave
associated contractions in the tissue, a FFT style analysis and reconstruction
were perfomed in MATLAB (Mathworks, Natick, MA). 3x1-minute segments, at
3x10 minute intervals, were taken to establish a baseline frequency and
amplitude. Each individual segment was de-trended, smoothed, filtered, and
reconstructed. Peak analyses could then count the amount of peaks and their
respective amplitudes. The baseline was compared to 3 1-minute segments after
T16Ainh-A01 incubation.
Statistical analysis:
Data were expressed as mean ± standard error of mean (SEM) unless
noted otherwise. Alpha was set at p<.05. One-way ANOVA followed by Tukey’s
Oscillations/min
Average Amplitude
1. De-trended (removal of floating baseline).
2. Smooth, Fast-Fourier transform and low-pass (.01 cut-off) filtering.
3. Reconstruction of filtered myograph and analysis.
3
15
test was used to compare all groups. Only up to three t-tests were used to
evaluate a priori hypotheses on the groups. Normality was assessed by fisher
skewness, and for ANOVA a Brown-Forsythe test was used to ensure
homogeneity of variance (Prism 7.00, GraphPad Software, Inc., La Jolla, CA,
USA).
16
RESULTS:
Immunofluorescence evaluation of ICC:
Figure 2
First, cryosections of pyloric tissue were prepared to ensure that the
excised tissues contained ICC (Figure 2). Then the sections were stained with
DAPI (Blue) and ICC marker Ano1. Several ICC could be visualized in the tissue,
including ICC-SM (Submucosal), ICC-MY (Myenteric), and ICC-IM
(Intramuscular). Scale Bars are 100 micrometers. After the presence of ICC was
confirmed in excised tissues, the main procedures of the study could be
performed.
17
Figure 3
Thick pyloric sphincter was isolated from the distal stomach for analyses
(Figure 3, A). It was stretched and equilibrated in the tissue bath (B), where an
isometric force transducer sensed strain and reported changes in stress
(newtons) from stimuli (C). The data was acquired real–time and sent to a
computer (D).
A. B.
C. D. Tensile Strain / Compressive Force
18
Figure 4
After the establishment of a stable baseline, the maximum response to
Ach was assessed. The same experiment was conducted with incubation of
T16Ainh-A01. Contraction following exogenous ACh was significantly reduced
(Figure 4, A, p<.05, n=5). EFS-induced relaxation remained similar with and
without the inhibitor.
In regards to the experiments with the incubation of T16Ainh-A01 inhibitor,
there was a large increase in basal tone upon the addition of inhibitor (B). This
would stabilize after five minutes at a new baseline. The increase was 390.16 ±
36.11 µN on average.
The addition of the inhibitor with exogenous Ach contraction leads to a
contraction change in tone higher than the response to only exogenous Ach.
*
Ex Vivo Rat Pylorus
T16Ainh-A01 response A. B. Ex Vivo Rat Pylorus
NT responses
19
Therefore, comparing gross responses with or without the inhibitor may not be a
suitable comparison.
Analysis of Slow-Wave Associated Contractions
Instead, the presence of slow-wave associated contractions was assessed
using fast Fourier transforms and peak analysis for both the amplitude of
contraction and frequency. Samples were taken with and without the presence of
inhibitor at baselines. Examining samples independently revealed that baselines
had more oscillations before the addition of the T16Ainh-A01 inhibitor,
suggesting that pacing, associated with the regulation of muscle contractility, was
being blocked (Figure 5, 3 different samples with excerpts of force over time).
Figure 5
Pylo
rus S
am
ple
1 minute
100 µN
1
t = 10 min 20 min T16Ainh-A01 30 min
2
3
20
Analyzing phasic contractions in normal pylorus indicated a frequency of
4.2 +/- 0.25 contractions per minute. This validates the filter methods since
normal gastric slow waves in the rat are recorded from three to five cycles per
minute. T16Ainh-A01 significantly reduced the frequency of phasic contractions
of the pyloric tissue to 3.2 +/- 0.11 (Figure 6, 72% of control, n=5, P < .05, t-test).
The amplitude of phasic contractions was reduced.
Figure 6
DISCUSSION:
The present investigation provides an advanced method for studying the
contribution of ICC to pyloric tissue function. The pylorus is a sphincteric tissue
which controls trituration periods and facilitates the passage of foods to the
duodenum from the stomach. The advanced analysis of force signals revealed
that the phasic contraction of the pylorus might be reliant on Ano1 and ICC
function. Physiology of the responses to exogenous Ach and EFS was
inconclusive, since the baseline was altered before the addition of other stimuli.
*
Frequency Amplitude
21
The first purpose of this study was to ensure that the target cells could be
found within the tissue being excised. After making cryosections and staining
them with ICC marker Ano1, ICC could be visualized within several layers of
pylorus tissues including the submucosa, myenteric plexus, and muscle layers.
The next purpose was to assess the contribution of ICC to pyloric
functions of contraction and relaxation, by incubating the tissue with an Ano1
inhibitor, T16Ainh-A01. The results were inconclusive because the incubation of
the inhibitor caused a tremendous increase in basal tone. The conclusion that the
blockade of ICC caused an increased in basal tone contradicts the results of
many prior experiments (Cobine et al., 2017). Rather, we doubt the inhibitors
specificity to target calcium activated chloride channels in only ICC. At this point,
this inhibitor has only been used on isolated ICC, and not whole gastrointestinal
tissues.
Furthermore, the use of this inhibitor as a therapeutic agent could have
tremendous GI side effects, as it caused a robust contraction in the tissue.
Perhaps, it could cause gastroparesis in patients by inducing pylorospasm, or
prolonged pyloric contractions. Therefore, we moved to assessing the function of
the tissue at baseline, instead of to responses only.
A method of examining phasic contractions at baseline was developed.
The novelty in this method was that it allowed us to not only derive the specific
frequency of contractions, but also measure the amplitude of each oscillation. For
us, this was a better method than examining only frequency and power, as we
22
would lose the measurement of force using traditional analyses. Analyzing phasic
contractions in normal pylorus indicated a frequency of 4.2 +/- 0.25 contractions
per minute, validating the filter methods since normal gastric slow waves in the
rat are recorded from 3-5 cycles per minute (Albertí et al., 2007). Here, we were
able to significantly decrease the amount of phasic contractions when incubated
with the inhibitor. There was a slight reduction in amplitude, but it was not
significant.
Other studies have shown that the slow wave or pacemaker activity
frequencies decrease when incubating ICC cultures with T16 Ainh-A01. We were
studying phasic contractions, which are understood to be the result of coupling
pacemaker activity of ICC with smooth muscle, in order to depolarize populations
of smooth muscle in coordination (Sun et al., 1995). At the concentration used in
this study (1 µM), the results of phasic contractions match rigorous assays of
pacemaker activity within ICC cultures using patch clamp techniques (Hwang et
al., 2016).
This study provides promising insights on GI pathophysiology, but there
are several limitations that need to be considered in the future. The data from this
study confirms that the phasic contractions were reduced, but a better method of
measuring ICC activity should instead measure electrical activity or frequency in
tissues, and match it with contractions or changes in tone. Also, the results are
inconclusive on the results of gross physiological responses such as contraction
and relaxation. Thus, better methods are needed to control specific cell
populations in tissues. In this study, as the specificity of the inhibitor was
23
uncertain, possible conclusions are limited. However, this study provides some
promising insights using advanced analyses of phasic contractions.
To conclude, the Ano1 inhibitor T16Ainh-A01 inhibited Ca2+-activated Cl−
currents, reduced the phasic contractions associated with ICC, and altered some
neurotransmitter responses in healthy pyloric tissue. Clinically, T16Ainh-A01
could alter pyloric function, and consequently gastric emptying, in therapeutic
uses. Future research should consider new models to study the contribution of
these key cells to gastrointestinal physiology.
ACKNOWLEDGMENTS:
This work was supported by Wake Forest School of Medicine Institutional Funds.
24
REFERENCES:
Albertí, E., Mikkelsen, H. B., Wang, X. Y., Díaz, M., Larsen, J. O., Huizinga, J. D., & Jiménez, M. (2007). Pacemaker activity and inhibitory neurotransmission in the colon of Ws/Ws mutant rats. American Journal of Physiology - Gastrointestinal and Liver Physiology, 292(6), G1499-G1510. doi:10.1152/ajpgi.00136.2006
Chen, Y., Shamu, T., Chen, H., Besmer, P., Sawyers, C. L., & Chi, P. (2011). Visualization of the Interstitial Cells of Cajal (ICC) Network in Mice. Journal of Visualized Experiments : JoVE(53), 2802. doi:10.3791/2802
Cobine, C. A., Hannah, E. E., Zhu, M. H., Lyle, H. E., Rock, J. R., Sanders, K. M., . . . Keef, K. D. (2017). ANO1 in intramuscular interstitial cells of Cajal plays a key role in the generation of slow waves and tone in the internal anal sphincter. The Journal of Physiology, n/a-n/a. doi:10.1113/JP273618
Hwang, S. J., Basma, N., Sanders, K. M., & Ward, S. M. (2016). Effects of new‐generation inhibitors of the calcium‐activated chloride channel anoctamin 1 on slow waves in the gastrointestinal tract. British journal of pharmacology.
Mazzone, A., Eisenman, S. T., Strege, P. R., Yao, Z., Ordog, T., Gibbons, S. J., & Farrugia, G. (2012). Inhibition of Cell Proliferation by a Selective Inhibitor of the Ca(2+)-activated Cl(−) Channel, Ano1. Biochemical and biophysical research communications, 427(2), 248-253. doi:10.1016/j.bbrc.2012.09.022
O'Grady, G., Du, P., Cheng, L. K., Egbuji, J. U., Lammers, W. J. E. P., Windsor, J. A., & Pullan, A. J. (2010). Origin and propagation of human gastric slow-wave activity defined by high-resolution mapping. American Journal of Physiology - Gastrointestinal and Liver Physiology, 299(3), G585-G592. doi:10.1152/ajpgi.00125.2010
Sanders, K. M., Kito, Y., Hwang, S. J., & Ward, S. M. (2016). Regulation of Gastrointestinal Smooth Muscle Function by Interstitial Cells. Physiology, 31(5), 316-326.
Sun, W. M., Smout, A., Malbert, C., Edelbroek, M. A., Jones, K., Dent, J., & Horowitz, M. (1995). Relationship between surface electrogastrography and antropyloric pressures. American Journal of Physiology - Gastrointestinal and Liver Physiology, 268(3), G424-G430.
Wood, J. D. (2016). Enteric Nervous System: Neuropathic Gastrointestinal Motility. Digestive Diseases and Sciences, 61(7), 1803-1816. doi:10.1007/s10620-016-4183-5
Zhu, M. H., Sung, T. S., O'Driscoll, K., Koh, S. D., & Sanders, K. M. (2015). Intracellular Ca2+ release from endoplasmic reticulum regulates slow wave currents and pacemaker activity of interstitial cells of Cajal. American Journal of Physiology - Cell Physiology, 308(8), C608-C620. doi:10.1152/ajpcell.00360.2014
25
CHAPTER II: AN IN VITRO MODEL OF THE GASTRIC
NEUROMUSCULAR APPARATUS USING ENGINEERED
PYLORUS TO UNDERSTAND GASTRIC
PATHOPHYSIOLOGY
Dylan T. Knutson1,2, B.S., Elie Zakhem2, Ph.D, Khalil N. Bitar1,2,3, Ph.D., AGAF.
1Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences, Winston Salem, NC
2Wake Forest Institute for Regenerative Medicine, Wake Forest School of Medicine, Winston Salem, NC
3Section on Gastroenterology, Wake Forest School of Medicine, Winston-Salem, NC, USA
This chapter describes the validation and physiological analysis of engineered neuromuscular tissues
This manuscript was submitted for publication
26
ABSTRACT
INTRODUCTION
Neuro-muscular disorders of the gut can result from the depletion of
neurons and/or interstitial cells of Cajal (ICC) populations. There is a lack of
controlled in vitro models to study the controversial contribution of these cellular
populations to gastrointestinal physiology.
OBJECTIVE
The aims of this study were (1) to bioengineer pyloric sphincter tissues
containing different combinations of cells and (2) to compare their physiological
functions.
METHODOLOGY
Neuro-muscular tissues were engineered using different combinations: (1)
SMC, (2) SMC + ICC, (3) SMC + NPC, and (4) SMC + ICC + NPC. Engineered
tissues were characterized by immunohistochemistry and physiology.
RESULTS
Isolated ICC exhibited normal morphology in culture and stained positive
for Ano1. Live microscopy of engineered tissues revealed GFP-positive ICC
networks in respective constructs. Organ bath studies demonstrated: (i)
Establishment of basal tone in all tissues to a similar extent. (ii) Increase in basal
tone in response to potassium chloride to a similar extent. (iii) Electric field
stimulation (EFS) induced relaxation in tissues that contained NPC (-244.3 ±
33.41 µN. p<.05, n=10). (iv) Relaxation in response to EFS was blocked by
27
inhibitors Tetrodotoxin and Lω-Nitro Arginine, confirming relaxation was mainly
dependant on a functional nNOS neuronal population. (v) Neural-mediated
relaxation was amplified in the presence of ICC in both peak and rate (1.9x,
p<.05, and 2.5x respectively, p=.12, n=5), indicating that there was an interaction
between the two populations to promote relaxation.
CONCLUSION
This study demonstrates the first successful engineering of a neuro-
muscular apparatus that incorporates the three major cells (smooth muscle,
enteric neurons and ICC). Real time force generation showed functional
contribution from all cell types, and an interaction from neurons and ICC which
enhanced relaxation. This model enables a new way to quantitatively deplete and
reinstate cell populations, in order to model, study, and repair gastrointestinal
pathophysiology.
Keywords: Gastroparesis, Pylorus, Diabetes, Bioengineering, Interstitial Cells of
Cajal (ICC), Disease Models
28
INTRODUCTION
Motility in the gut involves a complex interplay of tissues and cells to break
down, aid in effective absorption, and propel matter throughout each segment.
Propulsion is orchestrated by a complex interplay of the components of the
neuromuscular apparatus which include smooth muscle, enteric neurons, and
interstitial cells of Cajal (ICC) (Kenton M. Sanders, Hwang, & Ward, 2010).
Disruption of this apparatus characterizes some gastrointestinal (GI) diseases;
however, both the etiology and pathophysiology of dysmotility patterns are not
well understood (Goldstein, Thapar, Karunaratne, & De Giorgio, 2016).
Gastroparesis, or delayed gastric emptying, is a disease in which patients suffer
from debilitating symptoms consisting of nausea and vomiting, early satiety and
malnutrition, and discomfort. These incapacitating symptoms contribute to a
decreased health-related quality of life among these patients. Treatment options
are limited and include surgeries, prokinetics, and pacemaking devices
(Bielefeldt, 2012). Not only is the efficacy of these interventions limited, but they
are also only effective on certain subsets of patients.
Normal function of the stomach is dependent on the effective
accomodation, trituration, and emptying of reduced foods into the duodenum-
carried out by the fundus, antrum, and pylorus respectively. Phasic muscular
patterns are driven by electrical activity that is propagated by ICC which originate
in the greater curvature of the stomach and diminish after the pylorus (O'Grady et
al., 2010). The pylorus contracts and maintains tone to facilitate mixing, and
subsequently relaxes to enhance emptying of broken down particles.
29
Gastroparesis is thought to occur from a dysfunction occurring in one or more of
these tissues, including the pylorus, but this hypothesis is not entirely clear
(Camilleri, 2016).
Depletions of ICC and/or enteric neurons have been associated with
clinical cases of gastroparesis. Histological analysis shows patients may be
missing one or both cell types (Grover et al., 2011). It is well known that diabetics
suffer from neuropathies and complicated symptoms, including gastroparesis.
However, a significant amount of patients diagnosed with gastroparesis have no
other diseases present, and are thus classified as ‘idiopathic’. Recent studies
have mapped altered waveform patterns and associated them with both clinical
symptoms and with the extent of ICC depletions in stomach tissues (Angeli et al.,
2015). Furthermore, it has also been shown the specific location of cell
depletions is associated with different symptoms and disease phenotypes
(Moraveji et al., 2016). Yet, at present there are no cause and effect studies that
can attribute the effect of these cell depletions to disease pathophysiology.
The role that both ICC and enteric neurons perform in transmitting
cholinergic and nitrergic signals, and regulating muscular tissue function is
controversial (Kenton M Sanders, Kito, Hwang, & Ward, 2016; Wood, 2016).
Some investigators argue that ICC do not have a role in transducing nitrergic
responses (Chaudhury, 2016), while others have demonstrated that they
enhance it (Lies et al., 2014). Even though tremendous strides have been made
in investigating the role of these cells using clinical studies, isolated tissues, and
knockout models (Mashimo, Kjellin, & Goyal, 2000; Sivarao, Mashimo, & Goyal,
30
2008; Wagner, Sullins, & Dunn, 2014; Zarate et al., 2003), there is a lack of in
vitro models that quantitatively deplete and reinstate cell populations in order to
relate cell depletions with real functional changes in tissues.
The present investigation provides a novel tissue engineering approach to
develop an in vitro model of pyloric neuromuscular tissues using different
combinations of sphincteric smooth muscle, enteric neurons, and for the first
time- ICC. Prior to this study, autologous intrinsically innervated pyloric
sphincters were bioengineered, using human pyloric smooth muscle and neural
progenitor cells. The cells in these constructs demonstrated normal phenotypes,
muscular alignment, and function (Rego, Zakhem, Orlando, & Bitar, 2015). The
first goal of the present study was to incorporate ICC into the aforementioned
model. The second goal was to analyze the contribution of each cell type to basic
pyloric physiology. The results showed that physiology of the engineered
constructs consisted of a significant contribution from all cell types. (1) Tone was
myogenic trait and independent of neuronal or ICC contribution, (2) relaxation of
tone was dependent on a functional population of nitrergic neurons, (3) ICC
significantly increased the neural-mediated relaxation in both peak and rate. This
study provides a promising new model to aid in further research regarding the
investigation and understanding of gastrointestinal pathophysiology.
MATERIALS AND METHODS:
Isolation of cells:
Primary Smooth Muscle Cells (SMC)
31
Human tissues were ethically obtained from organ donors through
Carolina Donor Services and Wake Forest Baptist Medical Center (IRB#:
IRB00007586).
Smooth muscle cells were isolated as previously described (Rego et al.,
2015). Briefly, human pylori were removed by sharp dissection. Pylorus tissues
were manually cleaned by removing fat and mucosa with a surgical blade.
Tissues were extensively washed with HBSS solution containing 2X
antibiotics/antimycotic and then minced in sterile conditions. Tissues were
subjected to 2 digestions with HBSS containing 1 mg/mL collagenase type II
(Worthington Biochemicals, Lakewood, NJ) at 37°C for 1 hour each. Tissue
pellets were then resuspended in SMC growth media and plated in tissue culture
dishes at 37°C with 5% CO2.
Neural Progenitor Cells (NS)
Neural progenitors were isolated as previously described (Raghavan,
Gilmont, & Bitar, 2013). Briefly, duodenum from rats was harvested and
extensively washed with HBSS containing 2X antibiotics/antimycotic, followed by
mincing and additional washing. Minced tissue was subjected to 2 digestions in a
mixture containing 0.85 mg/ml type II Collagenase, 0.85 mg/ml Dispase II with 40
µg/ml DNAase I at 37oC for 45-60 minutes. The cells were then passed through a
70 µm nylon cell strainer and then through a 40 µm nylon cell strainer. Cells were
plated in Neurobasal Medium containing 1X N2 supplement, 20 ng/ml
recombinant human Epidermal Growth Factor, 20 ng/ml recombinant basic
Fibroblast Growth Factor, 1.0mM L-glutamine and 1X antibiotics/antimycotic.
32
Primary ICC
GFP-expressing mice were kindly provided by Dr. Frank C. Marini at
Wake Forest School of Medicine. GFP-ICC were isolated as previously described
(Zhu et al., 2009). Briefly, small intestine was harvested and washed extensively
with sterile HBSS containing 2X antibiotics/antimycotic in sterile conditions. After
mincing, tissues were again washed with HBSS three times. Tissues were
digested in HBSS containing 1.3 mg/mL collagenase type II (Worthington
Biochemicals), 2 mg/mL bovine serum albumin, 2 mg/mL trypsin inhibitor, and
0.27 mg/mL Adenosine triphosphate (Sigma-Aldrich, St. Louis, MO) for 23
minutes at 37°C . After digestion, tissue pellets were centrifuged at 600 g for 5
minutes and washed three more times. Pellets were then resuspended in SMC
growth media containing 0.01 µg/mL human stem cell factor (Peprotech, Rocky
Hill, NJ) and plated on 2.5 µg/mL collagen coated dishes at 37°C with 5% CO2.
Microscopic and immunofluorescence evaluation of isolated ICC:
Isolated ICC were grown on collagen-coated plates and evaluated
microscopically. At 80% confluency, cells were fixed with 10% normal buffered
formalin followed by permeabilization and blocking. ICC were stained against
ANO1 (Rabbit anti-Mouse T16mem/ANO1), followed by incubation with
secondary antibody (goat anti-Rabbit TRITC conjugate) (Abcam, Cambridge,
UK). Negative control received secondary antibody only. Cells were mounted
with mounting medium containing DAPI (Vector Labs, Burlingame, CA). Images
were processed using a fluorescence microscope (Nikon Instruments, Melville,
NY).
33
Bioengineered in vitro models of Pylorus
The process of engineering the sphincters was described previously
(Gilmont, Raghavan, Somara, & Bitar, 2014; Raghavan et al., 2011; Raghavan et
al., 2014). For all sphincters, 35 mm dishes were coated with sylgard. Eight
millimeter cylindrical sylgard posts were placed in the center of each plate,
followed by sterilization using ethanol and UV. Four types of sphincters were
engineered using different combinations of cells. Combinations included the
following and are described below: 1) SMC only, 2) SMC + ICC, 3) SMC + NS,
and 4) SMC + NS + ICC.
SMC only
Smooth muscle cells were trypsinized and 500,000 cells were collected
per construct. Cells were resuspended in a collagen gel of 0.4 mg/ml final
concentration. The mixture was then poured in the culture dish around the post.
The muscle gel was left to gel at 37ºC and then supplemented with differentiation
media.
SMC + ICC
Smooth muscle cells (500,000 per construct) and ICC (75,000 per
construct) were obtained and resuspended in collagen gel mixture with a final
concentration of 0.4 mg/ml. The mixture was left to gel 37ºC and then
supplemented with differentiation media.
SMC + NS
Neural progenitor cells (NS) were dissociated with Accutase (Invitrogen,
Carlsbad, CA) and 200,000 cells per construct were obtained. The cells were re-
34
suspended in a solution with a final concentration of 0.4mg/ml type I rat tail
Collagen (BD Biosciences) and 10ug/ml mouse laminin (Invitrogen, Carlsbad,
CA). The mixture was poured on the prepared Sylgard culture dishes. The
mixture was allowed to gel at 37ºC. Smooth muscle cells were then prepared by
collecting 500,000 cells and mixing them in the collagen gel mixture. The smooth
muscle mixture was poured on top of the first NS gel layer and allowed to gel at
37C. Following gelation of both layers, differentiation media was supplemented.
SMC + NS + ICC
Neural progenitor cells (NS, 200,000 cells per construct) were obtained
and re-suspended in a collagen/laminin solution as described above. The mixture
was poured on the prepared culture dishes coated with Sylgard and allowed to
gel. A total of 500,000 smooth muscle cells and 75,000 ICC per construct were
obtained and suspended in a collagen mixture as described above. The mixture
was overlaid on top of the first NS layer and allowed to gel, followed by the
addition of differentiation media.
Microscopic evaluation of the constructs:
Constructs were evaluated microscopically prior to physiological testing on
day 10 of culture. Constructs were imaged under FITC filter to visualize GFP-
expressing ICC. Images were processed using NIS Elements software (Nikon
Instruments, Melville, NY).
Physiological analysis of bioengineered pylorus
All different combinations of engineered pyloric constructs were analyzed
for physiological functionality. Constructs were hooked in an organ bath between
35
a fixed arm, and the measuring arm of an isometric, magnetoresistive force
transducer in a horizontal tissue bath (F10; Harvard Apparatus, Holliston, MA).
Force data was acquired using LabChart 7 software (ADInstruments, Colorado
Springs, CO). Constructs were maintained in 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES) buffered solution at 37°C throughout all
experiments. All force generation studies were performed after the establishment
of stable basal tone. Force generation was evaluated following the addition of
potassium chloride (KCl; 60 mM), acetylcholine (ACh; 1 μM), and electrical field
stimulation (EFS; 5 Hz, 0.5 ms, for 30 seconds). Constructs were washed
between each treatment, incubated in fresh buffer, and allowed to return to
baseline. Pretreatment with the neuronal blocker tetrodotoxin (TTX; 1 μM) was
done to evaluate the relative contribution of neurons to EFS response. Nitric
oxide synthase (NOS) inhibitor Nω-Nitro-l-arginine methyl ester hydrochloride
(LNAME; 300 μM) was administered for 12 minutes prior to EFS stimulation to
inhibit nitrergic neurons. GraphPad Prism 7.00 software for Windows (GraphPad
Prism 7.00, San Diego, CA) was used to analyze collected data. Second-order
Savitsky–Golay smoothing was applied. Quantification of physiologic data was
performed relative to basal tone for contraction and relaxation as
maximum/minimum peak response (delta force of basal tone), area (magnitude *
time), and rate (linear regression from onset to peak response).
Statistical analysis
Data were expressed as mean ± SEM unless noted otherwise. Alpha was
set at p<.05. One-way ANOVA followed by Tukey’s test was used to compare all
36
groups. Only up to three t-tests were used to evaluate a priori hypotheses on the
groups. Normality was assessed by fisher skewness, and for ANOVA a Brown-
Forsythe test was used to ensure homogeneity of variance (Prism 7.00,
GraphPad Software, Inc., La Jolla, CA, USA).
RESULTS:
Microscopic and immunofluorescence evaluation of isolated ICC
Isolated GFP-ICC were grown on collagen-coated plates. ICC adhered
and proliferated. Microscopic images at day 7 showed normal star-like
morphology with multiple projections, and exhibited bright green fluorescence
under FITC filter (Figure 7, A). Expression of ICC functional channel Ano1 was
confirmed by a positive ANO1 stain on the cells, (Figure 7, B, red) indicating a
pure ICC population in the culture.
Figure 7
DAPI ANO1 Merge
A. B.
37
Incorporation of ICC into bioengineered pylorus
Pyloric constructs were bioengineered. Smooth muscle cells (SMC),
neural progenitor cells (NS), and interstitial cells of Cajal (ICC) were
incorporated, representing structures as described in the methods; i.e. constructs
with 1) SMC, 2) SMC and ICC, 3) SMC and NS, and 4) SMC, NS, and ICC. All
four types of gels containing the combinations contracted and formed dense
circular muscle tissues around the central post (Figure 8, A). At day 10,
microscopic evaluation of SMC + ICC constructs showed GFP expressing ICC
incorporated in the construct with evidence of ICC forming networks (Figure 8, B
and C).
Figure 8
A. B. C.
38
Physiology of bioengineered pylorus
Basal tone:
All four types of constructs established a spontaneous basal tone (Figure
9, A, before dashed line, representative). SMC constructs generated a basal tone
of 384.6 ± 31.58 µN, SMC + ICC generated 328.9 ± 5.26 µN, SMC + NS
generated 384.9 ± 39.9 µN, and SMC + NS + ICC generated 294.4 ± 14.21 µN.
There was no significant difference in the peak of tone established in all
constructs. Both the rates and area of tone establishment were also similar
(Figure 9, B, p>.05). Results for the maximum, duration, and rate of basal tone
are summarized in Table 1. Establishment of maximum basal tone was similar
among all tissues regardless of the composition of the constructs. This indicates
that the establishment of basal tone is purely myogenic, and is dependent only
on the presence of a functional smooth muscle population.
Figure 9
Contractile response to potassium chloride:
All constructs established baseline before any treatment. Addition of 60
mM potassium chloride resulted in a robust increase of the tone which reached a
Response to KCl
A. B.
C.
Spontaneous Basal Tone
39
plateau (Figure 9, A, after dashed line, representative). Maximal average
contraction in each type of construct was 382.3 ± 49.17 µN for SMC only, 320.5 ±
51.68 for SMC + ICC µN, 269.4 ± 61.28 for SMC + NS µN, and 342.1 ± 107.9 for
SMC + NS + ICC µN. This additional increase in tone in response to KCl was
similar between the different groups (Figure 9, C, p>.05). The area and the rate
of contraction were also similar (p>.05). Results for the maximum, duration, and
rate of KCl contractions are summarized in Table 1.
SMC SMC + ICC SMC + NS SMC + NS + ICC
Tone
n= 5 5 5 5
Maximum (µN)
384.6 ± 31.58 328.9 ± 5.26 384.9 ± 39.9 294.4 ± 14.21
Area (µN x s)
106865 ± 34581 86116 ± 16136 106683 ± 37681 82694 ± 16511
Rate (dF/dt)
1.327 ± 0.3487 1.25 ± 0.2737 1.394 ± 0.226 1.141 ± 0.3952
KCl
n= 4 4 4 4
Maximum (µN)
382.3 ± 49.17 320.5 ± 51.68 269.4 ± 61.28 342.1 ± 107.9
Area (µN x s)
110018 ± 19642 102614 ± 14492 69430 ± 21438 98870 ± 34947
Rate (dF/dt)
0.8524 ± 0.233 0.5191 ± 0.1221 0.5885 ± 0.2624 0.7227 ± 0.1523
Table 1
40
EFS-induced Relaxation of Engineered Tissues
After the constructs established baseline, EFS was applied (Figure 10, A-
D, dashed lines). Average smooth muscle relaxation response in the SMC group
was -94.83 ± 10.01 µN, SMC + ICC was -115.3 ± 23.37 µN, SMC + NS was -
170.1 ± 25.97 µN, and SMC + NS + ICC was -303.7 ± 40.37 µN (Figure 10, E) .
Full EFS analyses are summarized in Table 2. While all minimums were
recorded over 10 minutes, and there was no perceivable relaxation (a decrease
in force) in constructs without neural progenitor cells (SMC and SMC + ICC),
EFS did induce relaxation in the constructs that contained neural progenitor cells
(SMC + NS and SMC + NS + ICC groups). This suggests a necessity of the
neural component for relaxation.
Figure 10
To confirm the dependence of EFS-induced relaxation response on
innervation, neural-blocker TTX was incubated with tissues before EFS
stimulation. TTX abolished EFS-induced relaxation in constructs, bringing
minimum innervated responses similar to those without (SMC, SMC + ICC).
*
41
nNOS neuronal blocker L-NAME yielded similar inhibition, confirming there was a
large functional population of differentiated neurons that were NO donors (Figure
11, p>.05). These results suggest that a differentiated population of nNOS
donated neurons, which were the driving force in the EFS response.
Figure 11
The innervated constructs (SMC +NS and SMC + NS + ICC) responded
with relaxation and returned to stable baseline. Further analysis of relaxation
responses revealed SMC + NS + ICC constructs had the most prominent
relaxation versus SMC + NS alone (Figure 4, E, 1.9x SMC + NS group, p<.05),
suggesting a role of ICC by magnifying EFS-induced relaxation.
42
Figure 12
To investigate the role of ICC on relaxation, rate and area of relaxation
were examined on the innervated groups only. An average of all relaxation
responses over time was computed and graphed to investigate the specific
cellular contributions to relaxation (Figure 12, A, traces, n=5). Here, the shape of
the response is similar- containing distinct relaxation with an effect ending at ten
minutes. However, the rate at which innervated tissues relaxed was markedly
increased with ICC (-0.27 ± 0.08 µN /s for SMC + NS and -0.64 ± 0.20 µN/s for
SMC + NS + ICC peak, figure 6, C, p=.12). Area of the responses remained
increased in similar fashion to the peak response (1.7x, SMC + NS: 49918 ±
14570 µNxs versus SMC + NS + ICC: 82639 ± 23594 µNxs, p=.30). The
difference between the onset rate and duration are represented graphically: The
SMC + NS + ICC tissues have faster rates, are time advanced, and have greater
magnitudes at virtually all time points. In summary, the addition of ICC created
A.
B.
C.
43
neuromuscular tissues which relaxed to a greater extent and reached peak
values at a faster rate.
The average response to nNOS blocker L-NAME was computed, graphed,
and compared to the corresponding EFS response (Figure 7, A and B, dashed
lines). L-NAME inhibition completely abolished relaxation in both construct types.
Summarizing, nNOS neurons were responsible to initiate most prominent
relaxation. ICC transduced inputs, including NO to increase relaxation throughout
the entirity of the response.
44
Figure 13
SMC SMC + ICC SMC + NS SMC + NS + ICC
EFS
n= 4 4 5 5
Minimum (µN)
-94.83 ± 10.01 -115.3 ± 23.37 -170.1 ± 25.97 -303.7 ± 40.37
Area (µN x s)
12905 ± 7733 24274 ± 8888 49918 ± 14570 82639 ± 23594
Rate (dF/dt)
0.26 ± 0.23 -0.09 ± 0.06 -0.27 ± 0.08 -0.64 ± 0.20
TTX EFS
n= 3 3 3 3
Minimum (µN)
-77.24 ± 14.22 -65.00 ± 26.00 -111.7 ± 33.17 -93.4 ± 16.5
L-NAME EFS
n= 3 3 3 3
Minimum (µN)
-58.59 ± 29.32 -60.00 ± 4.00 -50.11 ± 18.17 -72.92 ± 36.11
Table 2
A.
B.
45
DISCUSSION:
In this study, we provide a tissue engineering approach to model clinically
relevant cell depletions in engineered neuromuscular pylorus tissues using
smooth muscle, enteric nerves, and interstitial cells of Cajal. As previously
mentioned, the pylorus is a sphincteric tissue which controls trituration periods
and facilitates the passage of foods to the duodenum from the stomach. The
engineering technique used in this study allowed the formation of pyloric
constructs with different cell combinations for physiological studies. All cell types
contributed significantly to functional changes. Physiology of the engineered
sphincters demonstrated that capacity to establish tone was myogenic, since all
groups results were similar. Prominent relaxation was dependent on the
presence of NO-donating neurons, and the addition of ICC enhanced this neural
mediated relaxation in both the onset rate and peak response.
The first objective of this study was to create and validate a bioengineered
pylorus model similar to those in previous studies, in order to build additional
cells onto it. Following electric field stimulation, sphincters that were innervated
relaxed significantly more than those engineered only with smooth muscle. This
response could also be blocked by TTX and L-NAME, indicating that the base
model by Rego et al. in 2016 was both repeatable and functioned similarly.
Responses were also within magnitudes and shape of similarly sized ex vivo
tissues.
At the same time, we isolated and characterized ICC positive for Ano1.
Engineered ICC could be visualized in the constructs forming dense networks on
46
the day of physiology. These results indicate that the engineered constructs are
feasible, and could be used as representative models for pyloric tissues.
Conveniently, relaxations were initiated from similar basal tone, which reduces
error that can be introduced when studying changes in physiological responses.
ICC were involved in increasing rate and peak components of neural relaxation.
This indicates that relaxation is a fully incorporated mechanism and any change
in these populations creates a measurable change in function. Therefore, the
results of this model can be used to understand what may happen to a tissue as
a result of a complete ICC or neuron depletion.
Many pyloric diseases are characterized by an inability for pyloric
relaxation, such as pylorospasm or diabetic neuropathies, and are associated
with abnormal manometric data and gastroparesis (Camilleri, 2016). Studying
persistent relaxation in ICC knockout tissues has led to the dismissal of the role
of ICC in muscle function (Huizinga et al., 2008). The results of the present
investigation confirm that while these functions indeed persist, the addition of ICC
change the magnitude of the relaxation response. The magnitude of relaxation
may be an important mechanism for the pylorus to expedite gastric emptying.
Thus, the ‘persistence’ of responses may not be physiologically relevant, and the
character of these responses should be investigated further.
Previous research and models rely on the use of inhibitors, knockouts, or
toxins to produce quantitative observations on tissue function (Cobine et al.,
2017; Fujimura et al., 2016; Hwang, Basma, Sanders, & Ward, 2016). While
these may be useful tools, they lack specificity and enhanced control over tissue
47
cellular compositions. This contributes to the controversy regarding the role of
cells in tissue function. The in vitro model of the present study allows for
complete or partial depletion of cell populations, and is also capable of matching
and modeling clinically relevant conditions unique to patient populations.
This study provides a promising model to alter and test GI
pathophysiology, but there are several limitations that need to be considered in
the future. The data confirms that the ICC were positive for Ano1 when isolated
and cultured. Yet, the different connectivities between cells in the constructs
were not assessed. Also, there was no change in the establishment of basal
tone, but there may be cellular contributions to the maintenance of this tone. A
key attribute of ICC is the generation and transduction of electrical signals, and
this was not addressed in this study. Further investigations should examine
electrical phenomena in these tissues. Nevertheless, this study provides proof
that the engineered neuromuscular constructs may provide a basis for modeling
and researching pathophysiology resulting from cell depletions.
In conclusion, this tissue engineering approach has the potential to
generate further advanced research surrounding gastrointestinal
pathophysiology, by allowing cell types associated with disease to be
quantitatively depleted and replenished. The present investigation is the first of
its kind to demonstrate physiology of gut-derived neuromuscular tissues
engineered with ICC. Future research characterizing other qualitative and
quantitative changes, by adding, altering, or partially depleting cells should be
performed using this promising model.
48
ACKNOWLEDGMENTS:
This work was supported by Wake Forest School of Medicine Institutional
Funds. The authors would like to thank Dr. Frank Marini at Wake Forest School
of Medicine for providing the GFP mice.
49
REFERENCES:
Angeli, T. R., Cheng, L. K., Du, P., Wang, T. H.-H., Bernard, C. E., Vannucchi, M.-G., . . . O’Grady, G. (2015). Loss of Interstitial Cells of Cajal and Patterns of Gastric Dysrhythmia in Patients With Chronic Unexplained Nausea and Vomiting. Gastroenterology, 149(1), 56-66.e55. doi:http://dx.doi.org/10.1053/j.gastro.2015.04.003
Bielefeldt, K. (2012). Gastroparesis: Concepts, Controversies, and Challenges. Scientifica, 2012, 19. doi:10.6064/2012/424802
Camilleri, M. (2016). Novel Diet, Drugs, and Gastric Interventions for Gastroparesis. Clinical Gastroenterology and Hepatology, 14(8), 1072-1080. doi:http://dx.doi.org/10.1016/j.cgh.2015.12.033
Chaudhury, A. (2016). Furthering the debate on the role of interstitial cells of Cajal in enteric inhibitory neuromuscular neurotransmission. American Journal of Physiology - Cell Physiology, 311(3), C479-C481. doi:10.1152/ajpcell.00067.2016
Cobine, C. A., Hannah, E. E., Zhu, M. H., Lyle, H. E., Rock, J. R., Sanders, K. M., . . . Keef, K. D. (2017). ANO1 in intramuscular interstitial cells of Cajal plays a key role in the generation of slow waves and tone in the internal anal sphincter. The Journal of Physiology, n/a-n/a. doi:10.1113/JP273618
Fujimura, T., Shibata, S., Shimojima, N., Morikawa, Y., Okano, H., & Kuroda, T. (2016). Fluorescence visualization of the enteric nervous network in a chemically induced aganglionosis model. PloS one, 11(3), e0150579.
Gilmont, R. R., Raghavan, S., Somara, S., & Bitar, K. N. (2014). Bioengineering of physiologically functional intrinsically innervated human internal anal sphincter constructs. Tissue Engineering Part A, 20(11-12), 1603-1611.
Goldstein, A. M., Thapar, N., Karunaratne, T. B., & De Giorgio, R. (2016). Clinical aspects of neurointestinal disease: Pathophysiology, diagnosis, and treatment. Developmental Biology, 417(2), 217-228. doi:http://dx.doi.org/10.1016/j.ydbio.2016.03.032
Grover, M., Farrugia, G., Lurken, M. S., Bernard, C. E., Faussone–Pellegrini, M. S., Smyrk, T. C., . . . Pasricha, P. J. (2011). Cellular Changes in Diabetic and Idiopathic Gastroparesis. Gastroenterology, 140(5), 1575-1585.e1578. doi:http://dx.doi.org/10.1053/j.gastro.2011.01.046
Huizinga, J. D., Liu, L. W., Fitzpatrick, A., White, E., Gill, S., Wang, X.-Y., . . . Starret, T. (2008). Deficiency of intramuscular ICC increases fundic muscle excitability but does not impede nitrergic innervation. American Journal of Physiology-Gastrointestinal and Liver Physiology, 294(2), G589-G594.
Hwang, S. J., Basma, N., Sanders, K. M., & Ward, S. M. (2016). Effects of new‐generation inhibitors of the calcium‐activated chloride channel anoctamin 1 on slow waves in the gastrointestinal tract. British journal of pharmacology.
Lies, B., Gil, V., Groneberg, D., Seidler, B., Saur, D., Wischmeyer, E., . . . Friebe, A. (2014). Interstitial cells of Cajal mediate nitrergic inhibitory neurotransmission in the murine gastrointestinal tract. American Journal of Physiology - Gastrointestinal and Liver Physiology, 307(1), G98-G106. doi:10.1152/ajpgi.00082.2014
Mashimo, H., Kjellin, A., & Goyal, R. K. (2000). Gastric stasis in neuronal nitric oxide synthase–deficient knockout mice. Gastroenterology, 119(3), 766-773.
Moraveji, S., Bashashati, M., Elhanafi, S., Sunny, J., Sarosiek, I., Davis, B., . . . McCallum, R. (2016). Depleted interstitial cells of Cajal and fibrosis in the pylorus: novel features of gastroparesis. Neurogastroenterology & Motility.
50
O'Grady, G., Du, P., Cheng, L. K., Egbuji, J. U., Lammers, W. J. E. P., Windsor, J. A., & Pullan, A. J. (2010). Origin and propagation of human gastric slow-wave activity defined by high-resolution mapping. American Journal of Physiology - Gastrointestinal and Liver Physiology, 299(3), G585-G592. doi:10.1152/ajpgi.00125.2010
Raghavan, S., Gilmont, R. R., & Bitar, K. N. (2013). Neuroglial differentiation of adult enteric neuronal progenitor cells as a function of extracellular matrix composition. Biomaterials, 34(28), 6649-6658. doi:http://dx.doi.org/10.1016/j.biomaterials.2013.05.023
Raghavan, S., Gilmont, R. R., Miyasaka, E. A., Somara, S., Srinivasan, S., Teitelbaum, D. H., & Bitar, K. N. (2011). Successful implantation of bioengineered, intrinsically innervated, human internal anal sphincter. Gastroenterology, 141(1), 310-319.
Raghavan, S., Miyasaka, E. A., Gilmont, R. R., Somara, S., Teitelbaum, D. H., & Bitar, K. N. (2014). Perianal implantation of bioengineered human internal anal sphincter constructs intrinsically innervated with human neural progenitor cells. Surgery, 155(4), 668-674.
Rego, S. L., Zakhem, E., Orlando, G., & Bitar, K. N. (2015). Bioengineered human pyloric sphincters using autologous smooth muscle and neural progenitor cells. Tissue Engineering Part A, 22(1-2), 151-160.
Sanders, K. M., Hwang, S. J., & Ward, S. M. (2010). Neuroeffector apparatus in gastrointestinal smooth muscle organs. The Journal of Physiology, 588(23), 4621-4639. doi:10.1113/jphysiol.2010.196030
Sanders, K. M., Kito, Y., Hwang, S. J., & Ward, S. M. (2016). Regulation of Gastrointestinal Smooth Muscle Function by Interstitial Cells. Physiology, 31(5), 316-326.
Sivarao, D. V., Mashimo, H., & Goyal, R. K. (2008). Pyloric sphincter dysfunction in nNOS−/− and W/W V mutant mice: animal models of gastroparesis and duodenogastric reflux. Gastroenterology, 135(4), 1258-1266.
Wagner, J. P., Sullins, V. F., & Dunn, J. C. Y. (2014). A novel in vivo model of permanent intestinal aganglionosis. Journal of Surgical Research, 192(1), 27-33. doi:http://dx.doi.org/10.1016/j.jss.2014.06.010
Wood, J. D. (2016). Enteric Nervous System: Neuropathic Gastrointestinal Motility. Digestive Diseases and Sciences, 61(7), 1803-1816. doi:10.1007/s10620-016-4183-5
Zarate, N., Mearin, F., Wang, X., Hewlett, B., Huizinga, J., & Malagelada, J. (2003). Severe idiopathic gastroparesis due to neuronal and interstitial cells of Cajal degeneration: pathological findings and management. Gut, 52(7), 966-970.
Zhu, M. H., Kim, T. W., Ro, S., Yan, W., Ward, S. M., Koh, S. D., & Sanders, K. M. (2009). A Ca2+-activated Cl− conductance in interstitial cells of Cajal linked to slow wave currents and pacemaker activity. The Journal of Physiology, 587(20), 4905-4918. doi:10.1113/jphysiol.2009.176206
51
CHAPTER III: ELECTROPHYSIOLOGY OF AN IN
VITRO MODEL OF THE GASTRIC NEUROMUSCULAR
APPARATUS
Dylan T. Knutson1,2, B.S., Elie Zakhem2, Kenneth L. Koch, M.D.3, Ph.D, Khalil N. Bitar1,2,3, Ph.D., AGAF.
1Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences, Winston Salem, NC
2Wake Forest Institute for Regenerative Medicine, Wake Forest School of Medicine, Winston
Salem, NC
3Section on Gastroenterology, Wake Forest School of Medicine, Winston-Salem, NC, USA
This chapter describes the validation and electrophysiological analysis of engineered neuromuscular tissues
52
ABSTRACT
INTRODUCTION
Neuro-muscular disorders of the gut can result from the depletion of
neurons and/or interstitial cells of Cajal (ICC) populations. There is a lack of
controlled in vitro models to study the controversial contribution of these cellular
populations to gastrointestinal physiology.
OBJECTIVE
The aims of this study were to develop a measurement method to record
force and physiology simultaneously, and to study responses from an in vitro
model of the pyloric neuromuscular apparatus.
METHODOLOGY
Recordings were made using silver coated wire in an organ bath. Neuro-
muscular tissues were engineered using different combinations: (1) SMC, (2)
SMC + ICC, (3) SMC + NPC, and (4) SMC + ICC + NPC. Engineered tissues
were characterized by force and electrical activity.
RESULTS
The recording methodology was validated using a barrage of tests.
Tissues containing ICC had higher frequency (A, 6.0 ± .34 Cpm) and amplitude
(B, 1029 ± 65 µV) at baseline compared to tissues without ICC (frequency; 2.0 ±
.34 Cpm and amplitude; 425 ± 8.3 µV). The increase was significant (p<.05,
53
n=2). ICC tissues had more precise rhythm, and EFS prompted further decrease
in rhythmic frequency and increase in rhythmic precision.
CONCLUSION
This study was the first to demonstrate simultaneous electrical and force
recordings of a neuro-muscular apparatus that incorporated the three major cells
(smooth muscle, enteric neurons, and ICC). These results show that ICC play a
direct role in generating electrical phenomena and transducing neural relaxation
to muscle.
Keywords: Gastroparesis, Pylorus, Diabetes, Bioengineering, Interstitial Cells of
Cajal (ICC), Disease Models
54
INTRODUCTION
A hallmark symptom of gastroparesis is abnormal pacemaker activity in
the stomach, detected by electrogastrogram (EGG) (Koch & Stern, 2004). Phasic
muscular patterns are driven by electrical activity that is propagated by interstitial
cells of Cajal (ICC), which originate in the greater curvature of the stomach and
diminish after the pylorus (O'Grady et al., 2010). It has been shown that
depletions of ICC and/or enteric neurons are associated with clinical cases of
gastroparesis (Bashashati & McCallum, 2015; Penagini, 1998). Patients may
have bradygastria (slowing of rhythms), tachygastria (increased rhythm), or an
absence of activity altogether (Penagini, 1998). More recent studies have
mapped altered waveform patterns, and shown them to be associated with both
clinical symptoms and the extent of ICC depletions in stomach tissues (Angeli et
al., 2015). Furthermore, the specific location of cell depletions is associated with
different symptoms and disease phenotypes (Moraveji et al., 2016). As
previously explained, ICC may play a role in transducing nitrergic signaling in
order to boost sphincteric relaxation in the pylorus model. However, the
aforementioned study did not consider pacemaker activity.
In this continuation, for the first time, electrical phenomena of engineered
tissue with ICC were examined. The first goal of this study was to validate that
true signals were being received. The second goal was to evaluate the
contribution of each cell type to electrical activity, and then investigate changes
that may occur during responses of interest. The data showed that recorded
electrical activity was indeed true, and ICC were the main contributor to electrical
55
phenomena at baseline. When treated with EFS, ICC rhythm decreased in
frequency in order to facilitate a relaxation. This study provides a promising
foundation for understanding comprehensive physiology of GI tissues.
MATERIALS AND METHODS
Isolation of cells:
Primary Smooth Muscle Cells (SMC)
Human tissues were ethically obtained from organ donors through
Carolina Donor Services and Wake Forest Baptist Medical Center (IRB#:
IRB00007586). Smooth Muscle cells were isolated as previously described
(Rego, Zakhem, Orlando, & Bitar, 2015). Briefly, human pylori were removed by
sharp dissection. Pylorus tissues were manually cleaned by removing fat and
mucosa with a surgical blade. Tissues were extensively washed with HBSS
solution containing 2X antibiotics/antimycotic and then minced in sterile
conditions. Tissues were subjected to two digestions with HBSS containing 1
mg/mL collagenase type II (Worthington Biochemicals, Lakewood, NJ) at 37°C
for 1 hour each. Tissue pellets were then resuspended in SMC growth media and
plated in tissue culture dishes at 37°C with 5% CO2.
Neural Progenitor Cells (NS)
Neural progenitors were isolated as previously described (Raghavan,
Gilmont, & Bitar, 2013). Briefly, duodenum from rats was harvested and
extensively washed with HBSS containing 2X antibiotics/antimycotic, followed by
mincing and additional washing. Minced tissue was subjected to two digestions in
a mixture containing 0.85 mg/ml type II Collagenase, and 0.85 mg/ml Dispase II
56
with 40 µg/ml DNAase I at 37oC for 45-60 minutes. The cells were then passed
through a 70 µm nylon cell strainer and then through a 40 µm nylon cell strainer.
Cells were plated in Neurobasal Medium containing 1X N2 supplement, 20 ng/ml
recombinant human Epidermal Growth Factor, 20 ng/ml recombinant basic
Fibroblast Growth Factor, and 1.0mM L-glutamine and 1X antibiotics/antimycotic.
Primary ICC
GFP-expressing mice were kindly provided by Dr. Frank C. Marini at
Wake Forest School of Medicine. GFP-ICC were isolated as previously described
(Zhu et al., 2009). Briefly, small intestine was harvested and washed extensively
with sterile HBSS containing 2X antibiotics/antimycotic in sterile conditions. After
mincing, tissues were again washed with HBSS three times. Tissues were
digested in HBSS containing 1.3 mg/mL collagenase type II (Worthington
Biochemicals), 2 mg/mL bovine serum albumin, 2 mg/mL trypsin inhibitor, and
0.27 mg/mL Adenosine triphosphate (Sigma-Aldrich, St. Louis, MO) for 23
minutes at 37°C . After digestion, tissue pellets were centrifuged at 600 g for 5
minutes and washed three more times. Pellets were then resuspended in SMC
growth media containing 0.01 µg/mL human stem cell factor (Peprotech, Rocky
Hill, NJ) and plated on 2.5 µg/mL collagen coated dishes at 37°C with 5% CO2.
Bioengineered in vitro models of Pylorus
The process of engineering the sphincters was described previously
(Gilmont, Raghavan, Somara, & Bitar, 2014; Raghavan et al., 2011; Raghavan et
al., 2014). For all sphincters, 35 mm dishes were coated with sylgard. Eight
millimeter cylindrical sylgard posts were placed in the center of each plate
57
followed by sterilization using ethanol and UV. Four types of sphincters were
engineered using different combinations of cells. Combinations included the
following and are described below: 1) SMC only, 2) SMC + ICC, 3) SMC + NS,
and 4) SMC + NS + ICC.
SMC only
Smooth muscle cells were trypsinized and 500,000 cells were collected
per construct. Cells were resuspended in a collagen gel of 0.4 mg/ml final
concentration. The mixture was then poured in the culture dish around the post.
The muscle gel was left to gel at 37ºC and then supplemented with differentiation
media.
SMC + ICC
Smooth muscle cells (500,000 per construct) and ICC (75,000 per
construct) were obtained and resuspended in collagen gel mixture with a final
concentration of 0.4 mg/ml. The mixture was left to gel 37ºC and then
supplemented with differentiation media.
SMC + NS
Neural progenitor cells (NS) were dissociated with Accutase (Invitrogen,
Carlsbad, CA) and 200,000 cells per construct were obtained. The cells were re-
suspended in a solution with a final concentration of 0.4mg/ml type I rat tail
Collagen (BD Biosciences) and 10ug/ml mouse laminin (Invitrogen, Carlsbad,
CA). The mixture was poured on the prepared Sylgard culture dishes. The
mixture was allowed to gel at 37ºC. Smooth muscle cells were then prepared by
58
collecting 500,000 cells and mixing in the collagen gel mixture. The smooth
muscle mixture was poured on top of the first NS gel layer and allowed to gel at
37C. Following gelation of both layers, differentiation media was supplemented.
SMC + NS + ICC
Neural progenitor cells (NS, 200,000 cells per construct) were obtained
and re-suspended in a collagen/laminin solution as described above. The mixture
was poured on the prepared culture dishes coated with Sylgard and allowed to
gel. A total of 500,000 smooth muscle cells and 75,000 ICC per construct were
obtained and suspended in a collagen mixture as described above. The mixture
was overlaid on top of the first NS layer and allowed to gel followed by the
addition of differentiation media.
Microscopic evaluation of the constructs:
Constructs were evaluated microscopically prior to physiological testing on
day 10 of culture. Constructs were imaged under FITC filter to visualize GFP-
expressing ICCs. Images were processed using NIS Elements software (Nikon
Instruments, Melville, NY).
Physiological analysis of bioengineered pylorus
All different combinations of engineered pyloric constructs were analyzed
for physiological functionality. Constructs were hooked in an organ bath between
a fixed arm, and the measuring arm of an isometric, magnetoresistive force
transducer in a horizontal tissue bath (F10; Harvard Apparatus, Holliston, MA).
Force and electrical data was acquired using LabChart 7 software
59
(ADInstruments, Colorado Springs, CO). Electrical activity was recorded
simultaneously in bath using electrode fixture using 3 silver coated wires, with a
differential amplifier. The electrode sat in a fixed position, to prevent changes in
amplitudes between tissues. Band Pass filters were applied: High pass: 0.03 Hz
and Low pass: 3 Hz, sampled at 2kHz.Constructs were maintained in 4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffered solution at 37°C
throughout all experiments. All studies were performed after the establishment of
stable basal tone. Force and electrical phenomena were evaluated following the
addition of potassium chloride (KCl; 60 mM), and electrical field stimulation (EFS;
5 Hz, 0.5 ms, for 30 seconds). Constructs were washed between each treatment,
incubated in fresh buffer and allowed to return to baseline. Pretreatment with the
nifedipine (100 μM) was done to evaluate the risk of false signals. GraphPad
Prism 7.00 software for Windows (GraphPad Prism 7.00, San Diego, CA) was
used to analyze collected data. Second-order Savitsky–Golay smoothing was
applied. Quantification of physiologic data was performed relative to basal tone
for contraction and relaxation as maximum/minimum peak response (Delta Force
of basal tone), area (magnitude * time), and rate (linear regression from onset to
peak response).
Statistical analysis:
Data were expressed as mean ± SEM unless noted otherwise. Alpha was
set at p=.05. One-way ANOVA followed by Tukey’s test was used to compare all
groups. Only up to 3 t-tests were used to evaluate a priori hypotheses on the
groups. Normality was assessed by fisher skewness, and for ANOVA a Brown-
60
Forsythe test was used to ensure homogeneity of variance (Prism 7.00,
GraphPad Software, Inc., La Jolla, CA, USA).
RESULTS:
Validation of Electrical Signals
Figure 14
Silver coated wire leads were placed for detecting a differential current on
the engineered tissue within the organ bath (Figure 14). They remained fixed to
the bath for the following experiments.
61
Figure 15
First, a series of phenomena were recorded and documented, which was
unrelated to tissues within the bath. Raw recordings from force transducer
(Figure 15, B, blue trace) and/or electrodes (voltage, mV, green) of within tissue
bath were recorded simultaneously. It was necessary for the electrodes to
remain in the bath, otherwise a differential current could not be achieved
(arrowed, A) through tissue. Voltage pulses from EFS (5V, 5Hz, 0.5ms) were
detected upon stimulation (B). Without tissue or electrical activites, the liquid
reading from electrodes remained approximately 0 volts (C).
A. B. C.
62
Figure 16
Some disturbances within the bath could be perceived via the electrodes.
Recordings from electrodes of SMC + ICC + NS within the tissue bath (Figure 16,
mV, green). Perturbances of either washing (A, after dashed line) or adding
reagents (B, Highlighted) could obscure recordings for a short period. Washing
increased perceived voltage during and for several seconds following. Adding
reagents occasionally resulted in a voltage drop. Tests or analyses and were
performed following or by removing these false signals.
A.
B.
63
Figure 17
To fully validate and isolate any false recordings, a sequence of tests was
performed in a bath with electrodes with and without tissues (Figure 17). Since
the contractility of the tissues was dependent only on muscle, SMC Only tissue
was prepared for the series. First, perturbances of either washing or adding
reagents (1, after dashed line, carrier only) were recorded. This was done using
only buffer. Only small perturbances were recorded. Next, KCl was added, and
there was a tremendous increase in electrical signals (2 and 3, dashed line). This
test was repeated. Finally, the activity was blocked by nifedipine, confirming the
electrical depolarization was blocked (4, dashed lines). Finally, the same tests
were performed without any tissue in the same bath (5 and 6). No activity was
recorded, except for a small, short 2 mV peak upon the addition of KCl. This
64
barrage confirmed that the electrical signals were true, and that the electrical
signals were coming from functioning muscle cells in engineered tissue.
Physiology of bioengineered pylorus
Basal Tone:
Figure 18
We bioengineered pyloric constructs smooth muscle (SMC), neural
progenitor cells (NS), and interstitial cells of Cajal (ICC). The constructs are
previously described in the methods; i.e. constructs with 1) SMC, 2) SMC and
ICC, 3) SMC and NS, and 4) SMC, NS, and ICC. First, raw simultaneous
recordings of force (Figure 18, blue trace) and voltage (green trace) of
engineered tissues within tissue bath were recorded. Both graphs show
establishment of basal tone, with expected electrical spikes (arrowed, green
trace) preceeding contractions (arrowed, blue trace).
All four types of constructs established a spontaneous basal tone (Figure
3, A, before dashed line, representative). In review, SMC constructs generated a
basal tone of 384.6 ± 31.58 µN, SMC + ICC generated 328.9 ± 5.26 µN, SMC +
*
* *
* A.
B.
SMC + NS SMCC
SMC + ICC SMC + NS + ICC 1 min
2 m
V
65
NS generated 384.9 ± 39.9 µN, and SMC + NS + ICC generated 294.4 ± 14.21
µN. There was no significant difference in the peak of tone, area, and rate
established in all constructs. Establishment of maximum basal tone was similar
among all tissues regardless of the composition of the constructs. This indicates
that the establishment of basal tone was purely myogenic, and was dependent
only on the presence of a functional smooth muscle population.
However, there was a difference in the electrical activity of the tissues.
Tissues containing ICC had higher frequency (A, 6.0 ± .34 Cpm) and amplitude
(B, 1029 ± 65 µV) at baseline compared to tissues without ICC (frequency; 2.0 ±
.34 Cpm and amplitude; 425 ± 8.3 µV). The increase was significant (p<.05,
n=2). This change resulted in voltages with continuous rhythmic peaks each
minute. Therefore the ICC were generating electrical (pacemaker) activity within
the gastric tissue.
EFS-induced Relaxation of Engineered Tissues
Next, further investigation was made of the changes resulting from EFS.
Raw simultaneous recordings of engineered neuromuscular tissues from the EFS
response within the tissue bath were made. As explained in the previous chapter,
EFS did induce relaxation in the constructs that contained neural progenitor cells
(SMC + NS and SMC + NS + ICC groups). This suggests a necessity of the
neural component for relaxation. The relaxation of SMC + NS was -170.1 ± 25.97
µN, and of SMC + NS + ICC was -303.7 ± 40.37 µN.
66
After the constructs established baseline, EFS was applied (Figure 4, A-D,
dashed lines). EFS did induce relaxation in the constructs that contained neural
progenitor cells (SMC + NS and SMC + NS + ICC groups). The results in the
previous chapter confirmed a necessity of the NO donating neural component for
relaxation using inhibitors. Further analysis of relaxation responses revealed
SMC + NS + ICC constructs had the most prominent relaxation versus SMC +
NS alone (Figure 10, E, 1.9x SMC + NS group, p<.05). This suggests that ICC
play a role in relaxation by magnifying EFS-induced relaxation. The SMC + NS +
ICC tissues had faster rates, were time advanced, and had greater magnitudes
at virtually all time points. In summary, the addition of ICC created
neuromuscular tissues which relaxed further and reached peak values at a faster
rate.
67
Figure 19
Electrical Activity was recorded following EFS simultaneously with force
(Figure 19, n=2 per group). In the raw graphs, are two tissues tested on the same
day. scales (force and voltage) are the same. The EFS relaxation response is
greater in rate and magnitude in ICC tissues. In regards to Electrical Changes:
SMC + NS increased peak activities following EFS, then there was a drop-off in
activity. In SMC + NS + ICC, a slowing of rhythm occurs, and then rhythm is
SMC + NS + ICC
SMC + NS
68
maintained. Upon the release of NO, reduction of ICC slow waves may occur in
order to facilitate relaxation from a tonic state.
Following EFS stimulation, a plateau of activities occurs in SMC + ICC
+NS tissue. When comparing it to the control SMC +NS tissue, this .5 mV
plateau can be attributed to the resultant of both the 4-5 CPM, 1 mV signal
recorded following EFS, and the 4-6 CPM, 1 mV ICC signal recorded. Therefore,
each signal does involve the initiation of a neural component, but then the ICC
continue with a slowed, cyclic signal that occurs throughout the duration of
relaxation. This data confirms ICC transduced inputs, including NO to increase
relaxation throughout the entirity of the response, and perhaps became coupled
with muscle to facilitate a coordinated reduction in tone.
69
To confirm these qualitative observations, running spectral analysis was
performed. This analysis shows the changes and powers of different frequencies
over time.
Figure 20
Spectral analyses graphs of electrical Activity of both SMC + NS and SMC
+ NS + ICC tissues (figure 7, A). Strong frequencies are highlighted in bright
colors. Dense blue is indicative of a signal which has more rhythm. A 4 minute
segment was analyzed for spectral density(shaded), and is shown in graph A on
top. The dominant frequencies of these 4 minute segments before and after were
graphed (B).
There was a decrease in frequency in only ICC tissues following EFS from
3.7 cpm to 2.22 cpm. This lage decrease was also accompanied by a prominent
SMC + NS
SMC + NS + ICC
A. B.
1. 2.
70
increase in rhythm (2). This change was not present in SMC + NS only tissues,
who had a continuous, and abberrant signal. At any point, the dominant
frequency of these tissues was .72 CPM. There was no change before or after
EFS. This is summarized in figure 7, graph B.
There was also higher frequncy, low rhythm band in both tissues following
EFS, which could be attributed to Neural Activity (1). This activity is present in
both signals. Since we know that EFS caused no response in muscle only
tissues, and was present in both of the tissues incorporating neurons, then it is
suitable to attribute this activity to a neural component
71
DISCUSSION
The present investigation provides insight on a tissue engineering model
of clinically relevant cell depletions in neuromuscular pylorus tissues using
smooth muscle, enteric nerves, and interstitial cells of Cajal. The pylorus is a
sphincteric tissue which controls trituration periods and facilitates the passage of
foods to the duodenum from the stomach. The acquisition of electrical signals
within the organ bath yielded results that were valid, containing minimal false
recordings. In the prior study, physiology of the engineered sphincters
demonstrated that the capacity to establish tone was myogenic, and prominent
relaxation was dependant on the presence of NO-donating neurons. Also, the
addition of ICC enhanced this neural mediated relaxation in both the onset rate
and peak response.
The first aim of this study was to create and validate a method for
measuring electrical phenomena within the organ bath. After identifying
perturbances in signals, isolation of signals from tissues versus the environment
could reliably be recorded, observed, and analyzed.
Next, both electrical and force phenomena of the engineered tissues were
simultaneously measured. These results indicate that the engineered constructs
exhibited increased cyclic peaks and increased amplitudes, characteristic of the
pacemaker activity of both in vivo and cultured ICC. It is known that ICC play a
role in increasing rate and peak components of neural relaxation. This indicates
that relaxation is a fully comprehensive mechanism, and any change in these
populations creates a measurable change in function. However, when examining
72
the force and electrical phenomena simultaneously, a decrease in the cyclic
pacemaker pattern following the depolarization of the neural component by EFS
occurs.
It is well known that ICC and smooth muscle cells are electrically coupled
(Daniel & Wang, 1999). Though we found previously establishment of tone was
myogenic, maintenance of tone or changes in tone may depend on this
interaction. ICC contain guanylate cyclase, which is involved in a muscle
relaxation cascade in response to reception of NO (Groneberg, Aue, Lies, &
Friebe, 2015). The slowing of cycles from ICC may coordinate the relaxation of
muscle, through electrical coupling. A greater amount of muscle is hypopolarized
simultaneously, and the tissue relaxes at a greater magnitude and faster rate
(Furness & Costa, 1987).
Many pyloric diseases are characterized by an inability of pyloric
relaxation, such as pylorospasm or diabetic neuropathies, and are associated
with abnormal manometric data and gastroparesis (Camilleri, 2016). Studying
persistent relaxation in ICC knockouts tissues has led to the dismissal of the role
of ICC in muscle function (Huizinga et al., 2008). The present investigation
confirms that while these functions indeed persist, the addition of ICC change the
magnitude of the relaxation response. The magnitude of relaxation may be an
important mechanism for the pylorus to expedite gastric emptying. Therefore, the
‘persistence’ of responses may not be physiologically relevant, and the character
of these responses should be further investigated.
73
The innovation of the in vitro model used in the present study allows for
complete or partial depletion of cell populations. In addition, it allows for
controlled phenomena, which makes it possible to make further advancements in
the study of physiological responses. This model also makes it possible to
simulate clinically relevant conditions unique to patient populations. This
investigation is the first of its kind to comprehensively examine the contribution of
each key cell type, in both force and electrical phenomena.
Furthermore, even though this study provides a promising new model to
alter and test GI pathophysiology, there are several limitations that need to be
considered in future research. The data from this study confirm that the ICC were
likely responsible for increasing the electrical activity. Yet, there are currently no
suitable inhibitors available for entirely providing additional controls for this
contribution of the ICC without affecting other components. In this study, only
EFS of a NO donating (nNOS) neuron population was examined. There are
several other neural components in GI tissue including cholinergic, VIPergic, and
nicotinic neurons and receptors (Furness & Costa, 1987). Controlling the
differentiation of other neural populations could provide other models to study.
The paucity of data regarding cellular contributions to tissue function could
also arise from the complicated process of recording physiological phenomena.
As analysis becomes more in-depth, the modalities, expense, and skill required
become greater. Newer methods to create tissue engineered controls and record
phenomena must be produced. Nonetheless, this study provides proof that
74
engineered neuromuscular constructs could provide the basis for modeling and
studying pathophysiology resulting from cell depletions.
In conclusion, this tissue engineering approach has the potential to
generate and test unexplored hypotheses surrounding gastrointestinal
pathophysiology, by quantitatively depleting and replenishing cell types
associated with disease. This study is the first to demonstrate physiology of gut-
derived neuromuscular tissues engineered with ICC. Future studies should use
this promising model to characterize other qualitative and quantitative changes,
by adding, altering, or partially depleting cells.
ACKNOWLEDGMENTS:
This work was supported by Wake Forest School of Medicine Institutional
Funds.
REFERENCES:
Angeli, T. R., Cheng, L. K., Du, P., Wang, T. H.-H., Bernard, C. E., Vannucchi, M.-G., . . . O’Grady, G. (2015). Loss of Interstitial Cells of Cajal and Patterns of Gastric Dysrhythmia in Patients With Chronic Unexplained Nausea and Vomiting. Gastroenterology, 149(1), 56-66.e55. doi:http://dx.doi.org/10.1053/j.gastro.2015.04.003
Bashashati, M., & McCallum, R. W. (2015). Is Interstitial Cells of Cajal–opathy Present in Gastroparesis? Journal of Neurogastroenterology and Motility, 21(4), 486-493. doi:10.5056/jnm15075
Camilleri, M. (2016). Novel Diet, Drugs, and Gastric Interventions for Gastroparesis. Clinical Gastroenterology and Hepatology, 14(8), 1072-1080. doi:http://dx.doi.org/10.1016/j.cgh.2015.12.033
Daniel, E. E., & Wang, Y. F. (1999). Gap junctions in intestinal smooth muscle and interstitial cells of Cajal. Microscopy research and technique, 47(5), 309-320.
Furness, J. B., & Costa, M. (1987). The enteric nervous system: Churchill Livingstone Edinburgh etc.
75
Gilmont, R. R., Raghavan, S., Somara, S., & Bitar, K. N. (2014). Bioengineering of physiologically functional intrinsically innervated human internal anal sphincter constructs. Tissue Engineering Part A, 20(11-12), 1603-1611.
Groneberg, D., Aue, A., Lies, B., & Friebe, A. (2015). Cell-specific modulation of gastrointestinal NO-induced relaxation by phosphodiesterases. BMC Pharmacology and Toxicology, 16(1), A56. doi:10.1186/2050-6511-16-s1-a56
Huizinga, J. D., Liu, L. W., Fitzpatrick, A., White, E., Gill, S., Wang, X.-Y., . . . Starret, T. (2008). Deficiency of intramuscular ICC increases fundic muscle excitability but does not impede nitrergic innervation. American Journal of Physiology-Gastrointestinal and Liver Physiology, 294(2), G589-G594.
Koch, K. L., & Stern, R. M. (2004). Handbook of electrogastrography: Oxford University Press. Moraveji, S., Bashashati, M., Elhanafi, S., Sunny, J., Sarosiek, I., Davis, B., . . . McCallum, R.
(2016). Depleted interstitial cells of Cajal and fibrosis in the pylorus: novel features of gastroparesis. Neurogastroenterology & Motility.
O'Grady, G., Du, P., Cheng, L. K., Egbuji, J. U., Lammers, W. J. E. P., Windsor, J. A., & Pullan, A. J. (2010). Origin and propagation of human gastric slow-wave activity defined by high-resolution mapping. American Journal of Physiology - Gastrointestinal and Liver Physiology, 299(3), G585-G592. doi:10.1152/ajpgi.00125.2010
Penagini, R. (1998). Practical Guide to Gastrointestinal Function Testing. European Journal of Gastroenterology & Hepatology, 10(7), 623.
Raghavan, S., Gilmont, R. R., & Bitar, K. N. (2013). Neuroglial differentiation of adult enteric neuronal progenitor cells as a function of extracellular matrix composition. Biomaterials, 34(28), 6649-6658. doi:http://dx.doi.org/10.1016/j.biomaterials.2013.05.023
Raghavan, S., Gilmont, R. R., Miyasaka, E. A., Somara, S., Srinivasan, S., Teitelbaum, D. H., & Bitar, K. N. (2011). Successful implantation of bioengineered, intrinsically innervated, human internal anal sphincter. Gastroenterology, 141(1), 310-319.
Raghavan, S., Miyasaka, E. A., Gilmont, R. R., Somara, S., Teitelbaum, D. H., & Bitar, K. N. (2014). Perianal implantation of bioengineered human internal anal sphincter constructs intrinsically innervated with human neural progenitor cells. Surgery, 155(4), 668-674.
Rego, S. L., Zakhem, E., Orlando, G., & Bitar, K. N. (2015). Bioengineered human pyloric sphincters using autologous smooth muscle and neural progenitor cells. Tissue Engineering Part A, 22(1-2), 151-160.
Zhu, M. H., Kim, T. W., Ro, S., Yan, W., Ward, S. M., Koh, S. D., & Sanders, K. M. (2009). A Ca2+-activated Cl− conductance in interstitial cells of Cajal linked to slow wave currents and pacemaker activity. The Journal of Physiology, 587(20), 4905-4918. doi:10.1113/jphysiol.2009.176206
76
CHAPTER IV: MICRO-SENSITIVE MOLDED SILICONE
TISSUE PILLAR PLATFORM, FABRICATED BY 3-D
PRINTING FOR SIMPLER PHYSIOLOGICAL STUDIES
Dylan Knutson1, 2, and Khalil N. Bitar1, 2, 3
1Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences, Winston Salem, NC
2Wake Forest Institute for Regenerative Medicine, Wake Forest School of Medicine, Winston Salem, NC
3Section on Gastroenterology, Wake Forest School of Medicine, Winston-Salem, NC, USA
This chapter describes a new innovation for physiological studies
77
ABSTRACT
INTRODUCTION
Physiological studies are a tremendous resource for understanding cell
or tissue function, repair, or engineering. However, these studies require
expensive equipment to record signals, vast knowledge to acquire the right
signals, and have limited capabilities.
OBJECTIVE
The aim of this study were to fabricate a multifunctional tissue pillar
platform using 3-D printing and silicone molding.
METHODOLOGY
Pillars were molding using a printed device. Muscle tissues were
engineered inside using a collagen gel mixture.
RESULTS
The assembly and pillars could be made with high success. (>90%). The
sensitivity of the pillars fabricated reached 1.91 µN/µm of deflection. The
engineering process demonstrated control over the tissues, and cells expressed
contractile phenotype Actin and aligned uniformly between the pillars. Organ bath
studies validated that the pillars were almost as sensitive traditional techniques.
Electrical activity could be recorded simultaneously, and the pillars could
differentiate between sphincteric and non-sphincteric tissues.
78
CONCLUSION
This method provides a simple, powerful, multifunctional tool to investigate
physiology using engineered tissue.
Keywords: micro pillars, sensitive, physiology, tissue engineering, fluidic,
disease Models
79
INTRODUCTION:
Physiological studies are a tremendous resource for understanding cell or
tissue function, repair, or engineering. However, these studies require expensive
equipment to record signals, vast knowledge to acquire the right signals, and
have limited capabilities. Due to the challenge and inaccessibility of these
technologies, there is a lack of research involving the characterization of the
phenomena that whole tissues or cells generate (Polacheck & Chen, 2016). Even
though new techniques have emerged to record forces, these same hurdles of
cost, complexity, and inadequate equipment still exist.
Moreover, one favorable solution involves micro pillars, which provide a
simpler approach to quantifying cellular forces (Boudou et al., 2011; Stephen L
Rego, Raghavan, Zakhem, & Bitar, 2015; Stephen L Rego, Zakhem, Orlando, &
Bitar, 2016). Small silicone pillars were fabricated using lithography. Their
deflection was characterized, and then small collagen tissues containing
myocytes were engineered around them. The data from this abovementioned
study showed that the alignment of the cells was dependent on the experienced
tensile strain, and that each cell exerted nanonewton scale forces in order to
deflect the pillars. These results demonstrate simpler physiology approaches can
be used to investigate the formation of tissues and cellular function.
Nevertheless, not only is lithography an advanced method of fabrication, but the
tissues were engineered in a complex manner in that centrifugation of the entire
pillar platform in very small isolated wells was performed. Therefore, challenges
80
exist among fabricating the pillars, engineering the tissue, and isolating each
bath containing the pillars.
Recently, the accessibility of 3-D printers has increased. Simple and
accurate printers that can be used out-of-the-box can be acquired for fewer than
300 dollars. This modality provides new opportunities and methods for capturing
physiological data. Thus, an aim of the present investigation was to mold silicone
pillar tissues, using a basic 3-D printer, and subsequently characterize and
validate them. The final aim was to perform small physiological studies using
these silicone pillar tissues. Micro-sensitive pillars in a fluidic-ready bath were
created, which allowed for differentiation between muscle tissue types, and
simultaneous recordings of both electrical and forces phenomena. This type of
technology provides promising new techniques for further research regarding
cellular and tissue function.
METHODS:
Primary Smooth Muscle Cells (SMC)
Human tissues were ethically obtained from organ donors through
Carolina Donor Services and Wake Forest Baptist Medical Center (IRB#:
IRB00007586). Smooth muscle cells were isolated as previously described
(Stephen Lee Rego, Zakhem, Orlando, & Bitar, 2015). Briefly muscle tissues
were removed by sharp dissection. Tissues were manually cleaned by removing
fat and mucosa with a surgical blade. Tissues were extensively washed with
HBSS solution containing 2X antibiotics/antimycotic and then minced in sterile
conditions. Tissues were subjected to two digestions with HBSS containing 1
81
mg/mL collagenase type II (Worthington Biochemicals, Lakewood, NJ) at 37°C
for 1 hour each. Tissue pellets were then resuspended in SMC growth media and
plated in tissue culture dishes at 37°C with 5% CO2. Other smooth muscle
tissues were isolated using the same process.
Bioengineered Muscle Tissues
The process of engineering the sphincters was similar to methods
described previously (Gilmont, Raghavan, Somara, & Bitar, 2014; Raghavan et
al., 2011; Raghavan et al., 2014), with changes outlined below. Smooth muscle
cells were trypsinized and 500,000 cells were collected per construct. Cells were
resuspended in 1 mL of collagen gel of 1.9 mg/ml final concentration. The
mixture was then poured into the pillar vessel with ‘gates’. The muscle gel was
left to gel at 37ºC and then supplemented with differentiation media after 1 hour.
Microscopic evaluation of the constructs:
Constructs were evaluated microscopically prior to physiological testing on
day 4 of culture. A Stereomicroscope was used to collect data on the force of
tissues. Tissues were monitored by a time-lapse of 15 second intervals over 10
minutes. Changes in pillar deflection were measured at each time point, force
was then calculated, and then the output was plotted. Images were processed
using NIS Elements software (Nikon Instruments, Melville, NY).
Characterization of Pillar Stiffness
The stiffness of pillars was characterized using the following equation:
Some pillar deflection, d (µm) that corresponds to a force, F, µN; directly
82
determined by constant, k (µN / µm). Defined, 𝐹 = 𝑘𝑑. Pillars were submerged in
water, to account for swelling of silicone elastomer in liquids. Then, a hanging
mass was hooked to a string and pulley which was fastened to the pillar. The
deflection was measured and a k-value was computed for the translation of
deflection to a force. The k-value was characterized for each pillar set.
Physiological Analysis of Bioengineered Tissues
All muscle tissues were analyzed for physiological functionality. Tissues
were monitored by a time-lapse of 15 second intervals over 10 minutes. Changes
were measured at each point and then plotted following the calculation of force.
For the validation using the organ bath, constructs were hooked between a fixed
arm, and the measuring arm of an isometric, magnetoresistive force transducer
in a horizontal tissue bath (F10; Harvard Apparatus, Holliston, MA). Force data in
the organ bath and electrical activity was acquired using LabChart 7 software
(ADInstruments, Colorado Springs, CO). Electrical phenomena were acquired
using band pass filter (High pass: 0.03 Hz and Low Pass: 3 Hz, at 2000 kHz
sampling rate). Constructs were maintained in 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES) buffered solution at 37°C throughout all
experiments. All force generation studies were performed after the establishment
of stable basal tone. Force generation was evaluated following the addition of
potassium chloride (KCl; 60 mM). Constructs were washed between each
treatment, incubated in fresh buffer and allowed to return to baseline. GraphPad
Prism 7.00 software for Windows (GraphPad Prism 7.00, San Diego, CA) was
used to plot and analyze collected data. Second-order Savitsky–Golay smoothing
83
was applied. Quantification of physiologic data was performed relative to basal
tone for contraction and relaxation as maximum/minimum peak response (Delta
force of basal tone).
Fabrication:
CAD Drawings were made using AutoDesk Fusion 360 (AutoDesk, Mill
Valley, CA). Assembly and molds were made on a Printrbot Simple Metal using
PLA filament (Printrbot, Lincoln, CA). Pillars were molded using Sylgard 184
(Dow Corning, Greensboro, NC).
Statistical analysis:
Data were expressed as mean ± SEM unless noted otherwise. Alpha was
set at p.05. One-way ANOVA followed by Tukey’s test was used to compare all
groups. Only up to three t-tests were used to evaluate a priori hypotheses on the
groups. Normality was assessed by fisher skewness, and for ANOVA a Brown-
Forsythe test was used to ensure homogeneity of variance (Prism 7.00,
GraphPad Software, Inc., La Jolla, CA, USA).
84
RESULTS:
Fabrication:
CAD drawings were made to create the final design, which needed to be
simple, and allow for fluid flow and tissue interaction (Figure 21). The design
featured: 1. a top ABS plate providing thermal resistance and a windowed view,
inlet, and outlet. 2. A PDMS center channel allowed 3-D printed parts to be
protected from leaks, provided width for lower stresses, and lowered flow velocity
around tissues. Notches in the side allowed for ‘gating’ with silicone walls, so
that when tissues were cultured, barriers could be put in between each tissue. 3.
Several pairs of PDMS tissue pillars allow for the construction of different tissues,
and report strains visually to quantify force. (4) The bottom plate provides
sufficient room for added elements and thermal resistance.
37.5 mm
Tissue Pillar Device
1 2 3 4
85
Figure 21
CAD drawings are printed using a desktop 3-D printer. Final products are rigid,
accurate, and fit together tightly. Right: top plate and bottom plate before
assembly. Figure 22 shows assembled molds for PDMS products and parts for
the assembly.
Figure 22
86
Figure 23
CAD drawings are printed using a desktop 3-D printer. Molds for pillars
and bath walls (Figure 23, A and B, respectively) fit into the stage (C, shown with
part of mold component A). Some molds were coated with epoxy to provide a
smoother finish on silicone parts. After printing, M5 screws were tapped through
the bottom of the stage in order to raise molded pillars out of the stage. The
stage allows for the fitment of either the pillar (figure 24, A) or bath wall mold (B).
CAD drawings of mold components to be 3-D printed:
Mold assemblies after printing
A.
B.
C.
D.
87
Figure 24
The following figure outlines the molding process used to make tissue
pillars (Figure 25)- (1) the desired negative mold is placed in stage and PDMS is
poured over mold. (2) After curing, molded piece and negative are driven out of
stage by set screws. (3) The base is loosely peeled from face of product. (4) For
tissue pillars, the negative mold is disassembled into pieces. (5) Molded pillars
are revealed. (6) A PDMS base with a single set of pillars. Molding was complete
after only a day and the rate of success for this process was 90%.
Figure 25
A.
B.
1 2 3
4 5 6
88
All components of the assembly (Figure 26). The two silicone parts were
bonded with a final layer of silicone. The bath with pillars (below, top) or the full
assembly for fluidic experiments (below, bottom) is shown.
Figure 26
89
Tissue Engineering and Characterization:
Gates were produced from 3D printed mold which fit into slotted walls to
isolate/separate hydrogels cultured in vessel (Figure 27, A). These silicone‘gates
isolated 1 mL of 1.9 mg/mL Collagen hydrogel with a 92% percent success rate
(B). Gates could be easily removed with forceps after 40 minutes of gelation at
37˚C to add medium.
Figure 27
A.
B.
90
Figure 28
The bath with pillars, (Figure 28, A), representing the assembly to be
characterized for sensing force. These small pillars would operate like cantilevers
in order to report forces as they are deflected. Hooke’s Law describes the
deflection of an elastic material which corresponds to a linear change in force
(B). This approximation is generally accurate to a threshold, called the elastic
limit, where the deflection in relation to force loses its linearity. PDMS is a tunable
substrate where solids of different elasticity can be molded.
Two common mixtures of hardener to elastomer were characterized as an assay-
1:20 and 1:30 (Figure 29). In this experiment each pillar deflection was tested
against a series of known calibration masses to determine linear constant, k, and
the elastic limit of each.
Hooke’s Law Some pillar deflection, d (µm)
that corresponds to a force, F, µN; directly determined by constant, k (µN / µm).
Defined: 𝐹 = 𝑘𝑑
A. B.
91
0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0
0
2 0 0
4 0 0
6 0 0
8 0 0
1 0 0 0
P D M S M ix tu re S e n s it iv ity
(H a rd e n e r :E la s to m e r )
C a lib ra t io n F o rc e ( N )
Dis
ten
sio
n o
f P
illa
r (
m
)
1 :2 0
1 :3 0
1:2
0
1:3
0
0 .0
0 .5
1 .0
1 .5
2 .0
2 .5
H o o k e s L a w :
1 s t-O rd e r A p p ro x im tio n o f
S t if fn e s s , K
H a rd e n e r : E la s to m e r o f P il la r
N
/
m
0 2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0
0
2 0 0
4 0 0
6 0 0
1 :2 0
C a lib ra t io n F o rc e ( N )
Dis
ten
sio
n o
f P
illa
r (
m
)
0 2 0 0 4 0 0 6 0 0
0
2 0 0
4 0 0
6 0 0
8 0 0
1 0 0 0
1 :3 0
C a lib ra t io n F o rc e ( N )
Dis
ten
sio
n o
f P
illa
r (
m
)
1:2
0
1:3
0
0
2 0 0
4 0 0
6 0 0
8 0 0
E la s tic L im it o f
P D M S M ic ro P illa r
E la s to m e r : H a rd e n e r
Ela
sti
c L
imit
(
N )
Figure 29
Constants from each calibration test were calculated. For example a mass which
exerts 102 µN was hung from a tether to the pillar, and the change deflection was
recorded on the stereomicroscope. At some point, masses no longer deflected
the pillars within a linear confidence, representing their elastic limits. The 1:20
mixture resulted in constant, k= 0.58 ± 0.03 µN/µm, that is- each micrometer of
deflection measured represented 0.58 micronewtons. The 1:30 mixture resulted
in a more sensitive pillar at 1.909 ± 0.22 µN/µm. Linearity fell off at 600 and 400
micronewtons respectively, representing each vessels elastic limit. For the
experiments that followed a 1:30 pillar mixture was used, for the anticipation of
forces that were less than 400 micronewtons.
92
Microscopy of tissues in culture revealed muscle cells captured in the
tissue and aligned between the silicone pillars (Figure 30). Alignment was
uniform, and could be seen at high objectives. Staining with smooth muscle actin
(green) confirmed alignment of muscle filaments, and expression of a contractile
phenotype. Filaments in the tissues, with only 500k cells culture for 4 days, were
only beginning to connect.
Figure 30
93
Validation and investigation:
Forces were compared between a traditional organ bath and the pillars
(Figure 31). Tissues were tested first on the pillars they were cultured in, and
then they were shifted to an organ bath to perform the same test. A
stereomicroscope could measure the deflection of the pillars in response in order
to calculate forces (A). Deflections were small, but easy to perceive and measure
on the microscope. In the bath a force transducer measures strain and reports
force generated by tissues over time (B). The forces measured on the pillars
were: Basal Tone: 346.7 ± 80.91 and KCl: 67.39 ± 10.71. Forces measured in
the bath from the same tissues were 291 ± 12.31 and 95 ± 20.82 respectively.
The basal tone measured by both devices was similar (C, n=3, paired, p>.05),
but the measured response to KCl was slightly less by the pillar baths and was
approaching significance (D, n=3, paired, p=0.16).
Figure 31
A
.B
C
DD
94
With the forces measured by the pillars compared to a standard,
comparison of the KCl response of sphincteric and non-sphincteric (rabbit small
intestine) muscle tissues (rabbit internal anal sphincter) was made. Tissues were
washed with fresh buffer, and 60mM KCl was added following 30 minute
equilibration period. Sphincteric tissues exerted significantly more force within the
device compared to non-sphincteric (Figure 32, n=3-4, p<.05). Therefore, the
contractility of muscle when depolarized is different between sphincteric and non-
sphincteric muscles.
Figure 32
95
Next both force and electrical phenomena were measured simultaneously
in the bath. Below are KCl responses from three muscular tissues (Figure 33).
Force from the tissues was measured every 15 seconds while electrical activity
was measured continuously. Peak forces and large spikes in EMG activity are
associated with the onset of contraction. The peak EMG activity was 1.676 ±
0.8522 mV, and peak forces were 45 ± 7.211 µN. Larger or more frequent
activites resulted in higher measured forces.
Figure 33
96
DISCUSSION:
This study provides a simple and easy to fabricate platform for
investigating physiology of engineered tissues and cells. It allows for control over
engineering and reported phenomena of the tissues. In the study, the tissues had
muscle cells aligned between the pillars, and the pillars themselves could read
forces, though slightly lower than other physiology device standards. With this in
mind, the pillars still had the capacity to read micro-scale forces, which allows for
open ended investigations such as those of electrical activity.
The first aim of this investigation was to effectively fabricate the device.
This yielded high success, and provided a strong foundation for tissue
engineering. Parts were large enough to print successfully with a desktop printer,
and then molded silicone pillars were easily retrieved from the stage with high
success. These molded pillar vessels were sensitive enough to detect micro
scale forces.
Next, the tissues were engineered, which contracted into neatly aligned
muscular tissues which had a contractile phenotype. In the present studies, only
500K smooth muscle cells were used in the constructs. While these tissues were
functional, future studies might look to improve the density and consequently
force exerted by increasing cell numbers.
Last, validation of the apparatus was made by testing previously known
hypotheses on the tissues. For example, it is well know that the contractility of
sphincteric and non-sphincteric muscles are different (Stephen L Rego et al.,
97
2015; Stephen L Rego et al., 2016). The pillars in the apparatus were able to
differentiate between tissues cultured with either muscle type. We confirmed that
sphincteric smooth muscle exerted significantly more force than non-sphincteric
muscle when depolarized. Increased force is important to controlling the passage
of food between segments by increasing luminal pressure and limiting the size by
decreasing the luminal diameter. Ultimately, this study demonstrates the
feasibility and practicality of such technologies for generating and testing new
research regarding contractile tissues.
However, one drawback to these pillars was the discrepancy between
muscular responses when comparing it to a force transducer. Testing between
the standard horizontal bath and the pillar tissues showed that the pillar tissues
read a response on the same tissues that was less. This discrepancy could be
due to the stretch of the tissues. Upon setting tissues in an organ bath it is
recommended that the tissue be stretched, so that the transducer is immediately
feeling strain from the tissue. While the pillar tissue are cultured on the device,
and have some small ‘slack’ which is not accounted for immediately. So any
strain generated at first with the tissue is not experienced by the pillars, but
instead is used to pick up ‘slack’.
Another challenge is the seal of 3-D printed parts to water. Printing with a
large nozzle at lower infills produced parts that had to be patched with epoxy. So
waterproofing is another potential drawback of this fabrication. More involved
investigators performing long-term fluidic studies might produce platens made
from sheets of plastic instead.
98
In conclusion, this technology offers a simpler, novel approach to studying
the physiology of both engineered tissues and cells. It also provides an open
source platform for future research on tissues.
99
REFERENCES
Boudou, T., Legant, W. R., Mu, A., Borochin, M. A., Thavandiran, N., Radisic, M., . . . Chen, C. S. (2011). A microfabricated platform to measure and manipulate the mechanics of engineered cardiac microtissues. Tissue Engineering Part A, 18(9-10), 910-919.
Gilmont, R. R., Raghavan, S., Somara, S., & Bitar, K. N. (2014). Bioengineering of physiologically functional intrinsically innervated human internal anal sphincter constructs. Tissue Engineering Part A, 20(11-12), 1603-1611.
Polacheck, W. J., & Chen, C. S. (2016). Measuring cell-generated forces: a guide to the available tools. Nat Meth, 13(5), 415-423. doi:10.1038/nmeth.3834
Raghavan, S., Gilmont, R. R., Miyasaka, E. A., Somara, S., Srinivasan, S., Teitelbaum, D. H., & Bitar, K. N. (2011). Successful implantation of bioengineered, intrinsically innervated, human internal anal sphincter. Gastroenterology, 141(1), 310-319.
Raghavan, S., Miyasaka, E. A., Gilmont, R. R., Somara, S., Teitelbaum, D. H., & Bitar, K. N. (2014). Perianal implantation of bioengineered human internal anal sphincter constructs intrinsically innervated with human neural progenitor cells. Surgery, 155(4), 668-674.
Rego, S. L., Raghavan, S., Zakhem, E., & Bitar, K. N. (2015). Enteric neural differentiation in innervated, physiologically functional, smooth muscle constructs is modulated by bone morphogenic protein 2 secreted by sphincteric smooth muscle cells. Journal of tissue engineering and regenerative medicine.
Rego, S. L., Zakhem, E., Orlando, G., & Bitar, K. N. (2015). Bioengineered human pyloric sphincters using autologous smooth muscle and neural progenitor cells. Tissue Engineering Part A, 22(1-2), 151-160.
Rego, S. L., Zakhem, E., Orlando, G., & Bitar, K. N. (2016). Bioengineering functional human sphincteric and non-sphincteric gastrointestinal smooth muscle constructs. Methods, 99, 128-134.
100
CHAPTER V: SUMMARY AND CONCLUSIONS
In conclusion, these tissue engineering approaches have the potential to
generate and test new hypotheses surrounding pathophysiology, prompted by
the first study. By quantitatively depleting and replenishing cell types associated
with disease, the second study was the first of its kind to demonstrate physiology
of gut-derived neuromuscular tissues engineered with ICC. The development of
the final platform using micro pillars provides a promising and innovative model
to characterize other qualitative and quantitative changes of tissues.
101
APPENDIX
MATLAB CODING FOR SLOW WAVE ASSOCIATED CONTRACTION ANALYSIS:
close all %Detrends Sample Data force = force5; time = time5; [p,s,mu] = polyfit((1:numel(force))',force,6); f_y = polyval(p,(1:numel(force))',[],mu); %plot detrended data x = force - f_y; figure plot(x); hold on %Fast-Fourier Transform and Low-Pass Filter %figure [a, b] = butter(4, .01, 'low'); %filter cutoff at frequency fx = filter(a,b,x); %application of butterworth filter %plot(fx) %hold on % Reconstruction of Plot [Maxfreq, Loc] = max(periodogram(x)); figure periodogram(fx) hold on %Counting Peaks figure findpeaks(fx, time,'MinPeakProminence',20,'MinPeakDistance',5); xlabel('time(s)'); ylabel('microNewtons'); ylim([-200 200]); title('1 Min'); [pks, locs] = findpeaks(fx),
time,'MinPeakProminence',20,'MinPeakDistance',5);
102
SCHOLASTIC VITA
Dylan Knutson
Minneapolis, MN
Graduate Student + Biomedical Engineer
Education
Master of Science Candidate (Biomedical Engineering) 2017
Virginia Tech-Wake Forest School of Biomedical Engineering Sciences (SBES, Winston Salem,
NC)
Thesis:
Novel Approaches to Testing Gastro Intestinal Function In Vitro: Controlling Signal
Acquisition, Tissue Composition, or The Platform.
Advisor: Dr. Khalil Bitar.
Bachelor of Science (Health Sciences, Mathematics) 2014
Lees-McRae College (Banner Elk, NC)
Thesis:
Static and Kinematic Slip Properties of Wax Emulsions
Publications
DT Knutson, E Zakhem, KN Bitar, An In Vitro Model of the Gastric
Neuromuscular Apparatus Using Engineered Pylorus to Understand Gastric
Pathophysiology (Submitted).
DT Knutson, KN Bitar, Micro-sensitive Molded Silicone Tissue Pillar Platform,
fabricated by 3D printing, for simple physiological analysis (in preparation).
Awards
Awarded Alvarez Award, for ‘Best Overall Abstract’ and invited to do an extended
oral presentation by International Gastrointestinal Electrophysiology Society
Awarded 1st place poster presentation, people’s choice for “An Engineered In
Vitro Model of Pylorus Neuro-Muscular Apparatus.”, WFIRM Retreat, 01/2017
103
Presentations
DT Knutson, E Zakhem, KN Bitar, Interstitial Cells of Cajal Increase Neural-
Mediated Relaxation and Electrical Phenomena within In Vitro Model of Pylorus
Neuromuscular Apparatus.
o (Extended Oral presentation, International Gastrointestinal
Electrophysiology Society, 05/2017)
DT Knutson, E Zakhem, KN Bitar, An Engineered In Vitro Model of Pylorus
Neuro-Muscular Apparatus.
o (Oral, Wake Forest-Virginia Tech Graduate Research Seminars, 02/2017)
o (Poster, WFIRM Retreat, 01/2017)
DT Knutson, E Zakhem, KN Bitar, Incorporation of Interstitial Cells of Cajal in
Engineered Innervated Smooth Muscle Pyloric Sphincters.
o (Poster, International Conference of Biofabrication, 10/2016)
DT Knutson, E Zakhem, KN Bitar, Role of Interstitial Cells of Cajal (ICC) on
Sphincteric Tone in the Pylorus.
o (Poster, Graduate Symposium, 02/2016)
KN Bitar, E Zakhem, J Bohl, R Tamburrini, P Dadhich, C Scott, DT Knutson, J Gilliam Implantation of Autologous Biosphincters in a Non-Human Primate (NHP) Model of Fecal Incontinence.
o (Oral, Digestive Disease Week, 04/2017)
E. Petran, P. Dadhich, DT Knutson, E. Zakhem, K. N. Bitar, Cell Therapy for
Neo-innervation and Restoration of Neural Function in the Pylorus.
o (Poster and Oral presentation, WFIRM, 08/2016)
Departmental/Institutional Citizenship
Mentored summer student to complete a research project in 3 months.
o Taught and oversaw safe lab practices necessary for project.
o Assisted with collection, analysis, and interpretation of data.
o Edited and reviewed abstract and presentations.
Collaborates with other investigators to quickly engineer new components and
assemblies for projects
o Modeled and fabricated components to house innovative device which
will be used for high throughput drug screening.
o Created molding kits which can expedite the processing of bioengineered
tissues.
Restarted, refurbished, and manages entire physiology core.
o Instructs other scientists how to setup, record and analyze tissue
functionality.
o Improve scientific outreach by explaining and displaying physiology
equipment in departmental tours
o Saved institute tens of thousands of dollars on equipment and expensive
installation of new devices.
Volunteered at international and institutional conferences