TARGETING CYCLIC GMP SIGNALING for the TREATMENT of GASTROINTESTINAL DISEASES by October COPYRIGHT© 2017 by Sarah Sharman

Home , Linaclotide, Plecanatide

TARGETING CYCLIC GMP SIGNALING FOR THE TREATMENT OF GASTROINTESTINAL DISEASES

Sarah K. Sharman

Submitted to the Faculty of the Graduate School

of Augusta University in partial fulfillment

of the Requirements of the Degree of

Doctor of Philosophy

October

2017

TARGETING CYCLIC GMP SIGNALING FOR THE TREATMENT OF GASTROINTESTINAL DISEASES

This thesis/dissertation is submitted by Sarah K. Sharman and has been examined and approved by an appointed committee of the faculty of the Graduate School of

Augusta University.

The signatures which appear below verify the fact that all required changes have been incorporated and that the thesis/dissertation has received final approval with reference to content, form and accuracy of presentation.

This thesis/dissertation is therefore in partial fulfillment of the requirements for the degree of Doctor of Philosophy.

______Date Major Advisor

______Departmental Chairperson

(Nursing Only)______Associate Dean for Graduate Programs

______Dean, Graduate School

ACKNOWLEDGEMENTS

This document represents the culmination of over four years of graduate school

work and a lifetime of training and molding me into the person I am today. Words cannot begin to describe the gratitude and love that I have for everyone who has helped me along the way. I would first like to thank my mentor Dr. Darren Browning for not only accepting me into his lab family but more importantly for breaking me out of my shell and making me a more confident and outspoken professional. Above all else that is one of the most valuable things that I have gained during my time in graduate school. I also owe many thanks to my advisory committee—Drs. Kebin Liu, Nita Maihle, Lynnette

McCluskey, and Nagendra Singh— for helping steer me in the right direction and hone in on my presentation skills throughout the years. I also offer my sincerest thanks to Dr.

Honglin Li for taking the time to serve as my PhD dissertation reader.

Thank you to my BCB family for your tireless support and advice throughout the years. In addition, I owe sincere gratitude to the Graduate School for everything they have helped me with during my tenure at Augusta University. Special thanks to Dr.

Patricia Cameron for helping me to reach my potential as a leader in the AU community, this is something that I will continue to pursue throughout my professional career. I would also like to thank my lab family—Bianca Islam, Yali Hou, and Allison Bridges—

iii for their friendship, guidance, and help throughout our time together in the lab. They really were a second family to me while I was living so far away from my own family.

Above all else, I could not have gotten where I am today without the unfaltering love and support from my family, my boyfriend, and my closest friends. They have seen me at my lowest of lows during graduate school but somehow picked me up and got me back on the track of greatness. My mom and dad distilled a fearless independence and individualism in me early in my life that has allowed me to follow my dreams and pursue my education to the highest level and for this I am eternally grateful. To my sister and best friend Emily, thank you for keeping me laughing and smiling even on the days when all my experiments failed. I would not have made it this far in life without your friendship and support.

ABSTRACT

SARAH K. SHARMAN Targeting cyclic GMP signaling for the treatment of gastrointestinal diseases (Under the direction of DARREN D. BROWNING, PhD)

Continual renewal of the luminal epithelium in the gut is essential for the maintenance of

a healthy intestine as it sustains the barrier that protects underlying tissue from infiltration

of material passing through the lumen. Dysregulation of homeostatic processes involved

in maintenance of the barrier have been implicated in numerous gastrointestinal diseases.

The cGMP signaling axis has emerged as an important regulator of homeostasis in the

intestinal mucosa, and has been implicated in the suppression of visceral pain, colitis, and

colon cancer. While there is considerable interest in exploiting this pathway, until recently the approaches used to increase cGMP have been limited. The present study sought to test the hypothesis that elevation of cGMP in the intestinal epithelium using

PDE5 inhibitors will alter epithelial homeostasis and be therapeutic for constipation and preventative for colon cancer. Healthy mice treated with the PDE5 inhibitor sildenafil or the GC-C agonist linaclotide exhibited reduced proliferation and apoptosis, and increased numbers of differentiated secretory cells in the intestinal epithelium. In addition to these homeostatic effects, both drugs normalized intestinal transit and fecal water content in two mouse models of constipation. Furthermore, administration of sildenafil to mice treated with dextran sulfate sodium tightened the disrupted epithelial barrier. Treatment of ApcMin/+ mice with sildenafil or linaclotide significantly reduced the number of polyps

per mouse (67% and 50%, respectively). The effect of these cGMP-elevating agents was

not on the polyps themselves but was rather on the pre-neoplastic tissue, which was less

v proliferative and more apoptotic in the presence of the drugs. Taken together, the results of this study demonstrate that increasing cGMP with a pediatric dose of PDE5 inhibitors could be a potential alternative to GC-C agonists for the treatment of gastrointestinal diseases.

KEY WORDS: (cGMP, PDE5, GC-C agonist, IBS, constipation, motility, colon cancer)

Table of Contents

Page

Acknowledgements ...... iii

List of figures ...... x

List of tables ...... xi

Chapter

1 Introduction ...... 1

A Statement of problem ...... 1

2 Review of literature ...... 6

A Intestinal homeostasis ...... 6

i. Intestine structure and function ...... 6

ii. Proliferation and apoptosis ...... 9

iii. Differentiation ...... 10

B Functional gastrointestinal disorders ...... 13

i. Chronic idiopathic constipation ...... 13

ii. Irritable bowel syndrome ...... 14

iii. Current treatments ...... 15

C Colorectal cancer ...... 17

i. Statistics and etiology ...... 17

ii. Sporadic versus hereditary colorectal cancer ...... 18

iii. Chemoprevention ...... 19

D cGMP signaling in the intestine ...... 20

i. Overview ...... 20

ii. Secretion, homeostasis, and barrier ...... 20

iii. cGMP signaling in colorectal cancer ...... 21

E Pharmacological regulation of cGMP signaling in the intestine ...... 24

i. GC-C agonists ...... 24

ii. PDE5 inhibitors ...... 25

3 Materials and Methods ...... 27

A. Standard solutions ...... 27

B. Concentrated solutions from vendors ...... 28

C. Animals ...... 29

i. Wild-type mice ...... 29

ii. Apcmin mice...... 29

iii. Drug administration ...... 30

iv. Euthanasia ...... 31

D. cGMP assay...... 31

E. Immunohistochemistry ...... 32

i. Tissue preparation ...... 32

ii. Antigen blocking ...... 32

iii. Protein detection ...... 33

iv. ABC detection ...... 33

vii

v. Antibodies used ...... 34

F. Western blotting...... 35

i. Acrylamide gel preparation ...... 35

ii. Sample preparation ...... 35

iii. Preparation of blots ...... 36

iv. Antibodies used ...... 36

G. DSS-induced inflammation recovery model ...... 37

H. Intestinal transit assay ...... 37

I. Fecal water content ...... 38

J. Barrier permeability assay ...... 38

K. Apcmin/+ model of colon cancer ...... 39

i. Apcmin/+ colon cancer study ...... 39

ii. Histopathology ...... 40

iii. Analysis of gene expression ...... 41

L. Statistical analysis ...... 41

4 Results ...... 43

The effect of cGMP-elevating agents on intestinal homeostasis ...... 43

cGMP elevation for regulation of intestinal transit ...... 59

cGMP elevation for colon cancer prevention ...... 72

5 Discussion ...... 85

Regulation of intestinal homeostasis by cGMP-elevating agents ...... 85

viii

cGMP-elevating agents for the treatment of IBS-C and constipation ...... 87

cGMP-elevating agents for chemoprevention of colorectal cancer ...... 91

References ...... 96

Appendices ...... 105

A List of commonly used abbreviations ...... 105

List of Figures

Figure 1: Structure of intestines ...... 8

Figure 2: Intestinal epithelial homeostasis ...... 12

Figure 3: cGMP signaling in the intestinal epithelium ...... 23

Figure 4: Sildenafil increases cGMP in the colon mucosa ...... 44

Figure 5: cGMP increases secretory cell number in colon epithelium ...... 47

Figure 6: cGMP regulates proliferation and apoptosis in colon epithelium ...... 48

Figure 7: cGMP regulates secretory cell density in intestinal epithelium ...... 49

Figure 8: cGMP does not regulate proliferation or apoptosis in intestine...... 50

Figure 9: PDE5 inhibitors increase EC cells in colon epithelium ...... 51

Figure 10: Dose effect of sildenafil on homeostasis in the intestine ...... 54

Figure 11: Dose effect of sildenafil on homeostasis in the colon ...... 55

Figure 12: Dose effect of linaclotide on homeostasis in the intestine and colon ...... 56

Figure 13: Time course of sildenafil ...... 57

Figure 14: Homeostatic changes require 5 days of continuous treatment ...... 58

Figure 15: cGMP has no effect on transit time or fecal water in healthy mice ...... 61

Figure 16: DSS-recovery model schematic ...... 62

Figure 17: DSS-recovery mice have increased transit time and fecal water content .63

Figure 18: Sildenafil recovers 5HT cell deficit in DSS-recovery mice ...... 66

Figure 19: cGMP decreases intestinal transit time in DSS-recovery mice ...... 67

Figure 20: cGMP regulates fecal water content in DSS-recovery mice ...... 68

Figure 21: Sildenafil repairs leaky barrier in DSS-recovery mice ...... 69

Figure 22: Sildenafil normalizes transit in loperamide-induced constipation ...... 71

Figure 23: cGMP suppresses intestinal tumorigenesis in ApcMin/+ mice ...... 73

Figure 24: cGMP elevation has no effect on polyps ...... 74

Figure 25: Sildenafil treatment has no effect on inflammation in ApcMin/+ mice ...... 78

Figure 26: Expression of cGMP signaling components in ApcMin/+ mice ...... 79

Figure 27: Non-neoplastic tissue from ApcMin/+ mice have reduced GC-C peptides ...80

Figure 28: cGMP regulates homeostasis in non-neoplastic tissue ...... 81

Figure 29: cGMP elevation has no effect on tumorigenesis in ApcMin/+ colon ...... 83

Figure 30: Expression of cGMP signaling components in colon polyps ...... 84

List of Tables

Table 1: Mouse primers used for PCR ...... 42

I. INTRODUCTION

A. Statement of problem

Gastrointestinal diseases affect millions of Americans and negatively impact their quality

of life due to disruptive symptoms like abdominal pain, alterations in bowel habits, and

weight change. Diseases like irritable bowel syndrome, chronic constipation, and ulcerative colitis are not life threatening but do diminish a person’s quality of life with crippling symptoms of abdominal pain and changes in bowel habits. There are currently very few effective treatments for these diseases, and many patients are forced to take multiple medications to control their symptoms. Colorectal cancer is the second leading cause of cancer-related death in the United States, with an estimated 50,000 deaths predicted this year alone. Early screening and treatment is the cornerstone of colorectal cancer prevention; but the high mortality rate associated with diagnosis at advanced stages makes it one of the leading causes of cancer related deaths. In addition, chemopreventative efforts for colorectal cancer have uncovered very few effective drugs and are currently limited to low-dose aspirin. There is a growing need for new, more effective drugs to combat these gastrointestinal diseases.

The cyclic guanosine monophosphate (cGMP) signaling pathway has emerged as a potential therapeutic target for the treatment of gastrointestinal diseases, with

implications in the suppression of visceral pain, colitis, and colon cancer. Uroguanylin

and guanylin are endogenous peptide hormones in the intestine and colon that activate

epithelial guanylyl cyclase C (GC-C), causing intracellular cGMP production. It is well-

established that the cGMP signaling axis regulates water and electrolyte balance in the

colon; however, several labs have found that the cGMP signaling axis is also involved in

the control of homeostasis in the intestinal mucosa. Homeostasis is a highly regulated

process that is important in maintaining a healthy colon as it sustains the barrier which

protects underlying tissue from material passing through the lumen.

A critical barrier to exploiting cGMP signaling for the treatment of gastrointestinal

diseases is the limited options for increasing cGMP in the colon epithelium. Recently, a

new class of drugs has gained attention in the field of cGMP signaling for their ability to

alleviate constipation in patients with chronic idiopathic constipation (CIC) and

constipation predominant irritable bowel syndrome (IBS-C). Linzess® (linaclotide) is a

GC-C agonist prescribed for the treatment of IBS-C and CIC that increases cGMP in the

intestinal mucosa by binding and activating intestinal epithelial GC-C. While it has been highly successful in the clinic for the treatment of constipation, the main side effect is diarrhea making it a poor treatment for subsets of IBS and colitis patients. Linzess is also contraindicated in juvenile patients due to adverse effects during preclinical trials. Cyclic

GMP can also be pharmacologically elevated in the intestine with the use of phosphodiesterase 5 (PDE5) inhibitors which block the breakdown of endogenous cGMP. These drugs confer fewer gastrointestinal side effects and can be administered to children, making them available to a wider cohort of patients. Growing evidence from

numerous groups coupled with the success of Linzess indicate that pharmacological targeting of the cGMP signaling axis could be beneficial in the prevention and treatment of many gastrointestinal diseases.

Our objectives are to 1) determine the most effective pharmacological method for increasing cGMP in the intestine and colon epithelium, 2) to determine if cGMP signaling can normalize motility in preclinical mouse models of constipation, and 3) to determine if pharmacological elevation of cGMP in the colon epithelium is chemopreventative in a preclinical mouse model of colon cancer. This project’s impact will be the identification of the most effective method to increase cGMP in the colon epithelium, and demonstration that the cGMP signaling pathway can be targeted for treatment of constipation and colon cancer chemoprevention by using FDA-approved

PDE5 inhibitors and GC-C agonists. Because PDE5 inhibitors and GC-C agonists work on different components of the cGMP signaling pathway to increase cGMP in the colon, there could be potential for combination therapy using the two drugs. If additive effects existed, patients could be prescribed low doses of each drug that are under the threshold for side effects but still cause a maximum alleviation of symptoms. Combination therapy would be beneficial for patients with chronic conditions such as IBS and familial adenomatous polyposis who take drugs for prolonged periods of time.

The central hypothesis is that elevation of cGMP in the intestinal epithelium using

PDE5 inhibitors will alter epithelial homeostasis and be therapeutic for constipation and preventative for colon cancer.

Aim 1: Test the hypothesis that stimulation of cGMP production and inhibition of degradation will elevate cGMP signaling, improving barrier integrity and function in the intestinal epithelium. We used the GC-C agonist linaclotide to stimulate cGMP production and the PDE5 inhibitor sildenafil to block the degradation of cGMP. We performed time course and dose response studies to determine the most effective treatment strategy for each drug. Our approach to this aim was based on preliminary data that sildenafil increases cGMP levels in the mouse colon epithelium.

Aim 2: Test the hypothesis that sildenafil can regulate intestinal motility in two experimental models of constipation. We determined the extent to which cGMP regulates intestinal transit in post-inflammatory and opioid-induced constipation models.

We characterized intestinal motility in the DSS-recovery model and examined the effect of sildenafil on motility. We also used Loperamide to induce constipation and examined the ability of sildenafil to normalize opioid-induced constipation. Our approach to this aim was based on preliminary data suggesting both linaclotide and sildenafil work similarly through epithelial GC-C; therefore, we hypothesize that sildenafil will normalize motility in preclinical models of constipation.

Aim 3: Test the hypothesis that pharmacological elevation of cGMP in the colon

epithelium is chemopreventative in an ApcMin/+ murine colon cancer model. We

determined the extent to which cGMP can reduce polyp burden in the intestines of sildenafil and linaclotide-treated ApcMin/+ mice. We also analyzed proliferative and

apoptotic indices of polyps. We determined the expression levels of cGMP signaling

components within polyps and surrounding non-polyp tissue. Our approach to this aim

was based on our preliminary data demonstrating that cGMP activation inhibits

proliferation in the colon mucosa in vivo and published data indicating that exogenous

uroguanylin can reduce tumor burden in ApcMin/+ mice.

II. REVIEW OF LITERATURE

A. Intestinal homeostasis

i. Intestine structure and function

The major function of the gastrointestinal tract is the digestion of food into nutrients which are then absorbed into the bloodstream to maintain overall health. The intestinal tract is composed of four tissue layers: the mucosa, the submucosa, the muscularis externa, and the adventitia (Fig. 1). The mucosa is the inner-most layer and is comprised of epithelial cells, connective tissue called the lamina propria, and a thin layer of muscle that forms the boundary between the mucosa and submucosa [1]. The intestinal

epithelium is a specialized tissue responsible not only for the absorption of food and

water but also to serve as a physical barrier that separates luminal contents from the

underlying tissue [2]. Below the mucosa is the submucosa which provides support to the

mucosa and attaches it to the underlying smooth muscle. The muscularis layer of the

intestine are responsible for peristalsis which occurs through concerted contractions of

the longitudinal and circular muscles [3]. All of the layers of the intestinal tract work in concert to ensure proper digestion and absorption of nutrients into the body to maintain health.

The intestinal tract is anatomically and functionally divided into the small intestine and the large intestine, or colon. The major function of the small intestine is the digestion and absorption of nutrients [1]. It is divided into three sections, the duodenum, the jejunum, and the ileum. The small intestinal mucosa is organized into crypts with finger-like projections called villi that extend into the lumen and increase surface area for absorption. The colon carries out the final stages of water absorption and waste removal.

Because it does not carry out as much absorption as the intestine, the colon mucosa lacks villi and is comprised only of crypts.

Figure 1. Structure of the intestines. An illustration of a cross-section of the human intestines. The intestinal tract is composed of four tissue layers: the mucosa, the submucosa, the muscularis externa, and the adventitia. The mucosa is divided into three sections: the inner most layer called the epithelium, the connective tissue called the lamina propria, and the muscularis mucosa. The next layer is called the submucosa which provides support to the mucosa. Underlying the submucosa is the muscular layer, the inner-most called the circular muscle, and the outermost being the longitudinal muscle. The muscles are regulated by nerves and nerve plexus that reside between the two muscle layers. These nerves also branch into the mucosal layer where they are influenced by cells of the epithelium. The adventitia is the external covering of the intestines that holds the nerves and vasculature in place.

ii. Proliferation and apoptosis

The intestinal epithelium is a dynamic tissue, being constantly renewed every 3-5 days.

The process of continual renewal is orchestrated by pluripotent stem cells at the base of

the crypts [4]. Stem cells and their daughter transit amplifying cells make up the

proliferating compartment of the intestinal epithelium. Stem cells divide into daughter

cells which are pushed up the crypt into the transit-amplifying zone where they rapidly

replicate and receive signals that define their cell lineage [5]. The major cell types of the

intestinal tract include three types of secretory cells (goblet cells, enteroendocrine cells,

and Paneth cells) and absorptive enterocytes. Terminally differentiated cells no longer

have the capacity to proliferate and will move up and out of the crypt where they will

eventually be shed into the intestinal lumen.

The major driving force behind proliferation in the intestine is the Wnt pathway [6]. Wnts

exist in a gradient, with the most expression at the base of the crypt where it is essential

for the maintenance of crypt progenitor cells [7]. β-catenin is the central player in the

Wnt signaling pathway. In the absence of Wnt stimulation, β-catenin is phosphorylated

and marked for proteasomal degradation by a degradation complex. Wnt binding to its

receptor inactivates the degradation complex, allowing β-catenin to accumulate in the cell

and translocate into the nucleus. Once inside the nucleus β-catenin binds to transcription

factors of T cell factor/lymphocyte enhancer factor (TCF/LEF) family. Wnt/ β-catenin target genes are mainly involved in cell proliferation, differentiation, and migration.

Mutations in Wnt signaling components that lead to hyperactivation of the signaling

pathway are found in the majority of colorectal cancers [8].

As terminally differentiated cells reach the tip of the villus or the surface epithelium of

the colon they undergo cell death and are shed into the lumen [9]. Apoptosis, or anoikis,

is thought to be induced by a loss of anchorage to neighboring cells [10]. A balance exists

between cell proliferation at the base of the crypt and cell extrusion at the top of the villus

or crypt which is paramount in the maintenance of the crypt structure. Changes in the rate

of apoptosis has been linked to many pathological conditions in the gastrointestinal tract,

including ulcerative colitis and colorectal cancer [11].

iii. Differentiation

Once cells have been pushed out of the proliferating zone into the transit amplifying zone, they terminally differentiate into one of four epithelial cell types. Absorptive enterocytes make up the majority of cells in the intestine and are responsible for absorbing and transporting nutrients across the epithelium [12]. The remainders of the cells belong to the secretory lineage. Goblet cells produce mucins which form a protective barrier along the apical side of the intestinal epithelium [13]. They also produce and secrete other protective peptides like trefoil factors which are thought to have a role in mucosal healing [14]. Enteroendocrine (EE) cells make up 1% of the population of the intestinal epithelium. EE cells secrete various peptide hormones that modulate gastrointestinal functions [15]. There are roughly 15 different types of EE cells, defined by their morphology and the type of hormones they produce. Paneth cells exist solely in the base of crypts in the small intestine where they produce and release antimicrobial peptides [16]. Once committed to Paneth cell lineage, the cells migrate

down to the base of the crypt in opposition of other cells in the crypt. They are also more long lived than other epithelial cells, surviving for up to three weeks.

Cells are signaled to transition into one of these cells fates by various signaling pathways.

The Notch pathway is central to the cell fate decisions in the intestines. Like Wnt, Notch is essential in the maintenance of the undifferentiated, proliferative compartment [17]. It is believed to control the absorptive versus secretory cell fate decision. Inhibition of

Notch signaling results in the expansion of secretory cells (mainly goblet cells) at the expense of absorptive enterocytes [18, 19]. Downstream targets of Notch further determine the fate of secretory cells; Math1 is required for commitment to the secretory lineage [20], Neurogenin 3 is required for endocrine cell specification [21], and kruppel- like-factor 4 fates cells to goblet lineage [22].

Within the intestinal crypt, there is a balance between Wnt-mediated proliferation at the base of the crypt, lineage commitment to become either secretory or enterocyte cells, and terminal differentiation and programmed cell death at the luminal surface (Figure 2).

Intestinal epithelial homeostasis is a highly regulated process that is essential to maintaining a healthy colon as it sustains the barrier which protects underlying tissue from material passing through the lumen. The distal parts of the intestinal tract (ileum and colon) are colonized by large numbers of commensal bacteria that produce essential metabolites. Breaches in the epithelium can lead to infiltration of these and other bacteria into the underlying tissue which can cause inflammation which can lead to several gastrointestinal diseases [23].

Figure 2. Intestinal homeostasis. Schematic overview of a colonic crypt and the cells which reside within it. Stem cells (light blue) reside at the base of the crypt and divide into transit-amplifying cells. As cells are pushed up out of the proliferative zone and away from Wnt signal, they terminally differentiate into one of three cell type: absorptive enterocytes (light pink), enteroendocrine cells (green), or goblet cells (purple). The terminally differentiated cells continue to be pushed up the crypt axis until they reach the luminal surface and are discarded into the gut lumen via apoptosis. The entire process from stem cell division to apoptosis at the luminal surface takes 3-5 days in mammals.

B. Functional gastrointestinal disorders

Functional gastrointestinal disorders are a group of diseases which are characterized by

chronic abdominal complaints without a structural or biochemical cause to explain the

symptoms. The most common symptoms of functional gastrointestinal disorders are

visceral hypersensitivity and impaired gastrointestinal motility. There are a number of

functional gastrointestinal disorders which have been reviewed in great detail elsewhere

[24, 25]. For the purposes of this study, only chronic idiopathic constipation and irritable

bowel syndrome will be discussed further.

i. Chronic idiopathic constipation

Constipation is one of the most common gastrointestinal complaints in the United States, affecting about 42 million Americans per year [26]. Chronic idiopathic constipation

(CIC) is a functional gastrointestinal disorder characterized by infrequent bowel

movements and hard, lumpy stools. CIC is diagnosed if patients have experienced two of

the following symptoms in the past three months: fewer than 3 bowel movements in a

week, straining, lumpy or hard stool, abdominal symptoms such as bloating and

abdominal discomfort, and sensation of incomplete defecation [24]. It is not a life- threatening disease but it does decrease a person’s quality of life. In addition there are other complications that can arise from chronic constipation including hemorrhoids, anal fissures, rectal prolapse, or fecal impaction.

While the organic cause of CIC is currently unknown, there are a number of reasons that constipation may arise in patients. Slow intestinal transit will cause stool to remain in the

intestines longer which can dehydrate them leading to painful defecation. Diets that are low in fiber and water and certain medications like opioids for pain management can also

cause constipation.

ii. Irritable bowel syndrome

Irritable bowel syndrome (IBS) is a functional gastrointestinal disorder characterized by

alterations in bowel habits and recurrent abdominal pain [24]. It is estimated that 10-15%

of Americans suffer from IBS [27]. Although the disease is not life-threatening, it does

adversely affect a patient’s quality of life. The disease is sub-classified according to the type of bowel alterations into constipation-predominant (IBS-C), diarrhea-predominant

(IBS-D), or mixed symptom (IBS-M). IBS affects women twice as frequently as men, and young people under the age of 45 tend to be affected more than older adults [27].

The underlying cause of IBS is unknown; however, there are many potential causes of the symptoms associated with IBS. The major clinical hallmarks of IBS include alterations in intestinal motility, secretion, and visceral sensation. Some of the potential causes for IBS include brain-gut signaling problems, GI motility problems, pain sensitivity, bacterial infections, small intestinal bacterial overgrowth, or genetics [28-33].

Numerous studies indicate a role for serotonin signaling in IBS. Alterations in serotonin levels and enterochromaffin cell densities are common in patients with IBS, and serotonin has an established role in the regulation of intestinal motility and enteric nociception [34-36]. In support of this idea, targeting the serotonin system

pharmacologically has had moderate success in the clinic. 5HT-4 receptor agonists such as prucalopride stimulate peristaltic reflex, accelerating gastrointestinal transit and inhibiting visceral hypersensitivity [37-39]. However, due to adverse effects of the 5HT-4 agonist tegaserod, prucalopride does not have FDA approval in the United States. 5HT-3 receptor antagonists delay transit but also alleviate visceral pain in IBS-D patients [40].

While there have been moderate successes with targeting serotonin, they do not resolve all symptoms of IBS and are not specific enough to the gut.

Currently, there is no cure for IBS, and treatment strategies focus on controlling the symptoms associated with the disease. This often means that patients may take multiple medications to control their symptoms which can lead to additional health problems.

Bulking agents, laxatives, and anti-diarrheal drugs aim to help normalize alterations in bowel habits, while tricyclic antidepressants and antispasmodics are prescribed to minimize the visceral pain associated with IBS [41]. While this symptom-targeted approach to treating IBS works for some patients, the variable effectiveness of this strategy underscores the need to alternative treatments.

iii. Current treatments for IBS-C and chronic constipation

First line treatment for constipation focuses on non-prescription treatment options.

Lifestyle modifications such as dietary change and increased exercise are recommended before supplement or drug intervention [42]. The most successful first line treatment for constipation is fiber supplementation. Many studies provide strong evidence that soluble fiber supplements can alleviate symptoms in CIC and IBS-C [43]. Fiber does this by

creating bulky stool which naturally stimulates intestinal muscles to contract, thus

decreasing transit time and increasing motility. Although fiber is successful for many

patients, it can also exacerbate symptoms of constipation in other patients.

Over-the-counter (OTC) options are recommended when diet and lifestyle changes do not

help to alleviate constipation. The most common OTC treatment is laxatives, either

osmotic or stimulant [44]. Osmotic laxatives are hypertonic solutions that draw fluid into

the intestinal lumen by osmosis, creating softer stool. PEG3350, the most studied osmotic

laxative, has been shown to improve stool frequency and consistency while also

alleviating visceral pain [45]. Stimulant laxatives like Ex-Lax and Dulcolax induce

propagated colonic contractions to accelerate colonic transit [46]. They are not widely prescribed because of early studies that suggested senna and bisacodyl cause damage to the enteric nervous system. Recent studies report this is not true and stimulant laxatives are safe to use as a part of a long-term treatment strategy. OTC laxatives are effective for many patients; however, they can become resistant to them if they are used in the long term so patients with chronic constipation often need alternative treatment strategies.

There are a growing number of prescription treatment options for patients with CIC and

IBS-C who do not respond to non-prescription options. Intestinal secretagogues are some of the most promising classes of drugs on the market to treat constipation today [47].

Lubiprostone (Amitiza®) is a bicyclic fatty acid derived from prostaglandin E1 that

activates chloride channels in the intestinal epithelium [48]. The chloride-rich secretions soften stool and increase motility. In clinical trials, patients reported increased numbers

of bowel movements, improvements in straining, and improvements in stool consistency

[49-51]. The most common complaint from patients taking lubiprostone is nausea (30%).

Another class of intestinal secretagogues is the guanylate cyclase C (GC-C) agonists.

Currently linaclotide (Linzess®) and plecanatide (Trulance®) are available for patients with CIC. GC-C agonists work by activating GC-C which produces cyclic guanosine monophosphate (cGMP) that regulates downstream ion channels to increase secretion and soften stool.

C. Colorectal cancer i. Statistics and etiology

Colorectal cancer (CRC) is the third most commonly diagnosed cancer and second leading cause of cancer related death in the United States with roughly 135,000 new diagnoses and 50,000 deaths estimated this year [52]. There are many risk factors that increase a person’s chance of getting CRC including hereditary factors, environmental factors, and lifestyle.

It is now widely accepted that CRC arises through step-wise accumulation of genetic mutations that drive the transformation and progression of normal epithelial cells to an adenoma and, over time, a carcinoma. Bert Vogelstein described this in detail in a 1988 publication, and outlined key genetic events in the progression of colon cancer [53].

About 85% of all colon cancers have a loss of the tumor suppressor adenomatous polyposis coli (Apc), which Vogelstein thinks is the first mutational event in many colon cancers followed by mutations in KRAS, DCC, and Tp53. These pathways regulate

important cellular processes including proliferation and apoptosis, which are hijacked in

neoplastic cells, giving them a growth advantage. In support of the genetic basis of

colorectal cancer proposed by Fearon and Vogelstein, Hanahan and Weinberg suggested

eight essential alterations in cell physiology that must arise for malignant growth to

occur. These include self-sufficiency in growth factors, insensitivity to growth-inhibitory

signals, evasion of apoptosis, limitless replicative potential, sustained angiogenesis, the

ability to invade and metastasize, reprogramming of energy metabolism, and evasion of

the immune system [54, 55].

ii. Sporadic versus hereditary colorectal cancer

The majority of cases of CRC are attributed to sporadic mutations that occur over the

person’s lifetime. The incidence of CRC increases with age, with risk remaining

relatively low until the age of 50 and then rapidly accelerating [52]. Other factors that can contribute to sporadic CRC include dietary, environmental, and lifestyle factors.

Smoking, drinking 3 or more alcoholic beverages per day, being overweight, and eating an excess of red meat are all examples. In addition, people with ulcerative colitis and

Crohn’s disease are at an increased risk of CRC. Inherited susceptibility accounts for roughly 20% of all cases of CRC, meaning a person has an increased chance of CRC if they have a first-degree relative who has been diagnosed with the disease.

While the majority of colon cancer cases are sporadic, 5-10% of the cases are hereditary forms of colon cancer resulting from high-penetrance germline mutations. The two most common forms of hereditary colon cancer are familial adenomatous polyposis (FAP) and

hereditary non-polyposis colorectal cancer (HNPCC), also called Lynch Syndrome [56].

FAP patients are born with only one functional copy of the tumor suppressor gene

adenomatous polyposis coli (Apc). Apc is involved in the β-catenin degradation complex

in Wnt signaling [57]. Loss of heterozygosity of Apc fully inactivates the gene,

constitutively activating β-catenin. This leads to the formation of hundreds of polyps in

the colon [58]. People with FAP typically begin to have detectable polyps in their colon

as early as their mid-teens and will have multiple polyps by age 35. For these patients,

routine screening is paramount but many still require a colectomy when the growth of the

polyps is too aggressive. Only about 1% of colorectal cancers are attributed to FAP.

Lynch syndrome is a slightly more common type of hereditary colorectal cancer affecting

between 3-5% of colon cancer patients. It is an autosomal dominant form of colon cancer

that occurs when mutations arise in several DNA repair genes including

MLH1, MSH2, MSH6, PMS2, and EPCAM [59]. In additional to colorectal cancer, Lynch

patients are also at an increased risk of forming other types of cancer.

iii. Chemoprevention

Current treatment options for CRC are limited to surgery to remove the cancer,

chemotherapy, radiation therapy, and targeted therapy. The earlier the cancer is found,

the more likely it will be able to be eradicated. The 5 year survival rate for people with

Stage I CRC is about 93%, dropping to only 8% for Stage IV [60]. Therefore it is paramount that people predisposed to CRC because of family history, prior cancer, or age have routine screening to monitor for early-stage, treatable disease.

Because treatment for CRC is so limited and ineffective at treating late-stage disease, prevention has emerged as a new strategy for the amelioration of CRC in people at an elevated risk of CRC. Currently the only approved methods of prevention are limited to exercise, aspirin, and colonoscopy to remove pre-cancerous polyps [61]. This underscores the need for more effective and widely available prevention modals for CRC.

D. cGMP signaling in the intestine i. Overview

The cGMP signaling pathway has recently emerged as a potential therapeutic target for the treatment of gastrointestinal diseases. Uroguanylin and guanylin are endogenous peptide hormones in the intestine and colon, respectively. They are synthesized as prepropeptides in goblet cells, enteroendocrine cells, and enterocytes [62, 63]. The guanylin peptides are produced in response to a salty meal and then released into the intestinal lumen wherein they are proteolytically cleaved into active forms. The active peptides then bind epithelial GC-C, which produces cGMP as a second messenger [64,

65]. cGMP then activates downstream targets such as protein kinases (PKG1 and PKG2).

The action of cGMP can be discontinued by phosphodiesterase 5 (PDE5), an enzyme that

cleaves cGMP into its inactive form GMP (Figure 2).

ii. Secretion, homeostasis, and barrier

Secretion of water into the intestinal lumen is important for the lubrication and

breakdown of food. Dysregulation of fluid balance in the intestines can lead to diarrhea

or constipation. It is well established that cGMP signaling regulates water and electrolyte

balance in the intestine through regulation of ion channels in the intestinal epithelium

[64]. Cyclic GMP, through PKG2, activates the cystic fibrosis transmembrane conductance regulator (CFTR) causing increased secretion of chloride. Increased cyclic

- - GMP levels also activate bicarbonate secretion through stimulation of the Cl /HCO3

exchanger. There is also a cGMP mediated inhibition of the Na+/H+ exchanger which

results in reduced reabsorption of sodium. Together, this results in a net movement of

water into the lumen of the gut, driving water secretion [66].

It is now widely accepted that cGMP signaling regulates more than just secretion. Many

studies over the past decade have implicated cGMP signaling in the regulation of

intestinal epithelial homeostasis. Knockout studies with mice deficient in cGMP

signaling components exhibit crypt hyperplasia, increased luminal apoptosis, and reduced

differentiation of secretory cells in the intestinal epithelium [67-70]. In addition it was shown that knockout mice have intestinal barrier defects [71, 72]. While the mechanism behind the regulation of secretion by cGMP is understood, the mechanistic details of the cGMP effect on homeostasis are far from clear.

iii. cGMP signaling in colorectal cancer

Growing evidence suggests that activating cGMP signaling in the colon epithelium could have therapeutic potential for colon cancer. A derivative of the NSAID sulindac called

Exisulind, lacks COX-inhibitory activity but was found to block growth in colon cancer cell lines by inhibiting phosphodiesterases [73, 74]. The underlying mechanism was shown to involve inhibition of β-catenin signaling, leading to apoptosis [75, 76]. In

clinical trials, Exisulind suppressed tumorigenesis in familial adenomatous polyposis

(FAP) patients [77]. Although Exisulind induced regression of rectal polyps in FAP

patients it did not get FDA-approval due to harmful side effects reported in Phase III

trials [78]. Although Exisulind was ultimately not successful, it did highlight the

therapeutic potential of cGMP signaling.

In addition to information gained through the Exisulind trials, there is a growing amount

of preclinical data that suggests a therapeutic role for cGMP signaling in colorectal

cancer. Mice deficient in GC-C were found to have an exaggerated proliferative

compartment [67, 79] and exhibit increased tumorigenesis in both AOM and ApcMin/+

mouse models of colon cancer [68]. The endogenous GC-C peptides uroguanylin and

guanylin are commonly lost gene products in most human colon cancers suggesting

hormone replacement could be a potential treatment strategy [80, 81]. In support of this

idea, it was reported that exogenous uroguanylin treatment or treatment with the

uroguanylin analog plecanatide can reduce tumor incidence in ApcMin/+ mice that are genetically predisposed to intestinal cancer [82, 83]. It was recently reported that the

PDE5 inhibitor Vardenafil can reduce proliferation in the colon and protect against inflammation and inflammatory-driven colorectal cancer in mice, although the protective mechanism has not been elucidated at this point [69, 84]. Taken together, these results suggest that pharmacological elevation of cGMP signaling in the colon epithelium with

PDE5 inhibitors or GC-C agonists could be effective for colon cancer treatment without the negative side effects of Exisulind.

Figure 3. cGMP signaling in the intestinal epithelium. A schematic representation of cGMP signaling in an intestinal epithelial cell. The endogenous ligands uroguanylin (Ugn) and guanylin (Gn) or the GC-C agonist linaclotide (Lin) bind and activate epithelial GC-C. GC-C makes cGMP which activates downstream targets like protein kinases and ion channels like CFTR. cGMP is degraded into its inactive form GMP by phosphodiesterase 5 (PDE5) which can be inhibited by PDE5 inhibitors such as sildenafil (Sild).

E. Pharmacological regulation of cGMP signaling in the intestine

i. GC-C agonists

A novel class of drugs called GC-C agonists has emerged for the treatment of IBS-C and

CIC. Linzess® (linaclotide) was the first FDA-approved member of this family and increases cGMP levels in the intestinal epithelium by stimulating GC-C receptors [85,

86]. Linaclotide is a peptide analog of the heat-stable bacterial enterotoxin, ST. ST peptides contain three disulfide bonds and are highly potent super activators of GC-C.

They irreversibly bind the receptor causing hypersecretion, making them the main cause of traveler’s diarrhea [87]. Increased fluid secretion in response to linaclotide is likely to be central to the therapeutic effect of the drug on constipation [85]. However, linaclotide has also been shown to affect neuromuscular function and reduce visceral pain in human patients as well as in rodents, suggesting an additional role for cGMP signaling in the gut

[88, 89]. In clinical studies, IBS-C and CIC patients treated with linaclotide both reported improvements in the number of spontaneous bowel movements per week and alleviation of their abdominal pain [90-92]. However, the most common side effect reported during the studies was diarrhea, being reported in 20% of patients with IBS-C and 16% of patients with CIC. In addition, linaclotide is contraindicated in pediatric patients under the age of 18 because of adverse findings in preclinical trials with juvenile mice [93].

In contrast to ST peptide, uroguanylin and guanylin only have two disulfide bonds and bind GC-C in a pH-sensitive manner [94]. Uroguanylin is most potent in acidic conditions like that of the proximal small intestine. As material is moved through the

intestine and the pH becomes more basic, uroguanylin can no longer exert its effect on

GC-C and guanylin becomes the active peptide in those regions. Trulance® (plecanatide),

another GC-C agonist, is structurally similar to uroguanylin with the exception of the

replacement of one pH sensing residue with another near the N-terminus [95]. This means that like uroguanylin, plecanatide also binds GC-C in a pH sensitive manner, acting predominantly in the proximal small intestine. Plecanatide is currently approved for use in patients with CIC only. In clinical studies, it was shown that plecanatide improved the number of spontaneous bowel movements and alleviated the abdominal pain associated with constipation [96]. The most commonly reported side effect was diarrhea, which was reported in only 5% of patients. Like linaclotide, plecanatide is contraindicated in pediatric patients due to adverse findings in juvenile mice [97].

ii. PDE-5 inhibitors

Phosphodiesterase 5 (PDE5) is an enzyme that breaks down cGMP into the inactive form

GMP [98], serving as a regulator of cGMP signaling. However, PDE5 inhibitors can block the action of PDE5, thus causing an increased accumulation of endogenously produced cGMP [99]. Currently PDE5 inhibitors are approved to treat erectile dysfunction, benign prostate hyperplasia, and pulmonary arterial hypertension [99-101].

However, there is growing evidence that PDE5 inhibitors could have therapeutic potential for many gastrointestinal diseases.

The PDE5 inhibitor vardenafil was recently shown to regulate homeostasis in the intestinal epithelium of mice [69]. Treatment of mice with vardenafil led to increased

differentiation of secretory cells (goblet and enteroendocrine) and decreased proliferation and apoptosis. It was also shown that the PDE5 inhibitors vardenafil and sildenafil can block DSS-induced colitis and inflammatory-drive colorectal cancer in mice [69, 84].

This study sought to test the hypothesis that pharmacological elevation of cGMP signaling in the intestinal epithelium will regulate intestinal homeostasis and be therapeutic for IBS-C, constipation, and colon cancer.

III. MATERIALS AND METHODS

A. Standard Solutions

10x Sodium-Citrate Buffer for antigen retrieval (pH 6): 100mM sodium-citrate (2H2O),

0.5% (v/v) Tween 20. For 100ml stock, 2.94 g sodium-citrate, 500 µl Tween 20, diH2O to

100 ml, adjust to pH 6 using citric acid. Dilute to 1X with diH2O before use.

20X Phosphate Buffered Tween Buffer (PTS): 2.2g monobasic sodium phosphate, 6.6 g

dibasic sodium phosphate, 360 g NaCl, 10 g sodium azide, 20ml Tween 20, diH2O to 2L.

Store at room temperature. Dilute to 1x with diH2O before use.

4X Lower Running Acrylamide Gel Buffer (pH 8.8): 90.85 g Tris base, 20 ml 10%

(w/v) sodium dodecyl sulfate (SDS), diH2O to 500 ml, adjust to pH 8.8. Store at room

temperature.

4X Upper Loading Acrylamide Gel Buffer (pH 6.8): 30.3 g Tris base, 20 ml 10% (w/v)

sodium dodecyl sulfate (SDS), diH2O to 500 ml, adjust to pH 6.8. Store at room

temperature.

1X Bovine Serum Albumin Blocking Buffer: 5% (w/v) solution, 5 g bovine serum

albumin fraction (BSA, ethanol precipitated, biotech grade) in 100 ml 1x PTS. Filter and

store at 4°C.

5X SDS Polyacrylamide Gel Electrophoresis (SDS-PAGE) Buffer: 625mM Tris buffer pH 6.8, 10% (w/v) SDS, 25% (v/v) glycerol, 0.015% (w/v) bromophenol blue, 5% (v/v)

β-mercaptoethanol. For 50 ml solution, 31.3 ml of 1M Tris buffer stock, 5 g SDS, 12.5 ml glycerol, 7.5 mg bromophenol blue, 2.5 ml β-mercaptoethanol, diH2O to 50 ml.

Aliquot for long term storage at -80°C and short storage at -20°C.

Tissue Lysis Buffer (pH 7.4): For 100 ml, 3 ml of 5M NaCl stock, 5 ml of 1M Tris-HCl

stock pH 8, 10 ml of 10% (v/v) NP-40 stock, 0.25g of sodium deoxycholate, 0.2ml of

0.5M EDTA stock, 1 ml of 10% SDS stock. Add diH2O to 100ml, adjust pH 7.4, aliquot

for storage at -20°C.

B. Concentrated solutions purchased from vendors

TAE buffer for DNA gels: For 2L, dilute 40 mL of 50X TAE buffer (Biorad, #161-0773)

in 1.6 L of diH2O.

Phosphate buffered saline (PBS): For 10L, dilute 1L of 10X PBS (Fisher BioReagents,

BP3994) into 9L of diH2O; store at room temperature.

1X Tris/Glycine/SDS Western blot running buffer: For 10L, dilute 1L of 10X

Tris/Glycine/SDS (BioRad, #161-0772) into 9L of diH2O; store at room temperature.

1X Tris/Glycine Western blot transfer buffer: For 10L, dilute 1L of 10X Tris/Glycine

(BioRad, #161-0771) into 9L of diH2O; store at room temperature.

C. Animals

All animals used throughout this study were maintained in compliance with recommendations in the Guide for the Care and Use of Laboratory Animals of the

National Institutes of Health. Additionally all animal procedures were approved by the

Institutional Animal Care and Use Committee at Augusta University. Mice were housed in the Augusta University animal facility in a temperature and humidity controlled room with free access to food and water, unless otherwise stated in individual protocols. i. Wild-type Mice

Initial studies in this lab used strain CD-1 mice to test the effects of the PDE5 inhibitor vardenafil on homeostasis in the intestinal epithelium. Therefore, the preliminary sildenafil studies also used strain CD-1 mice to ensure that comparable results could be obtained. Six-to eight-week old CD-1 mice were purchased from Harlan Laboratories

(Indianapolis, IN, USA). Once the sildenafil effect on homeostasis was confirmed, the mouse strain was changed to C57/BL6 mice because most of the disease models in this lab use that strain. Breeding pairs of C57/BL6 mice were purchased from Jackson

Laboratories (Bar Harbor, ME) and a colony was maintained in our mouse facility.

ii. ApcMin/+ Mice

Male C57/BL6J-ApcMin/+ mice and female C57/BL6J mice were obtained from Jackson

Laboratory at age 6 weeks for breeding. Because of potential pregnancy complications in

Apcmin/+ females, we chose to use male C57/BL6J-ApcMin/+ throughout our breeding for

this colony. Genotyping was carried out by PCR using 2µl genomic DNA from mouse

tails (purified by Puregene Mouse Tail kit, QIAGEN, Valencia, CA) with a wild-type

APC forward primer (GC-CATCCCTTCACGTTAG), an APCmin forward primer

(TGAGAAAGACAGAAGTTA), and a common reverse primer

(TTCCACTTTGGCATAAGC). The APC min primer detects a T to A transversion of

nucleotide 2549.

iii. Drug administration

Pharmaceutical grade sildenafil citrate (Revatio®) was ground with a mortar and pestle

and stored as a suspension in di-H2O at -20°C. To calculate dosing for sildenafil in

drinking water, daily water intake of mice was monitored for one-week. Daily water

intake (ml) per gram of animal was calculated. For example, CD1 mice drink an average

of 0.18 ml/g and weigh on average 25g. A cage of five CD1 mice would drink 4.5ml x 5

mice, or 22.5ml per day. Based upon this daily mean water intake, stock sildenafil was added to the drinking water at concentrations that approximated daily doses of 0.36-17

mg/kg per mouse. Unless otherwise stated, the dose of sildenafil administered in the

drinking water was equivalent to 5.7 mg/kg per day. For oral gavage and intraperitoneal

injection, mice were administered 100 µl of 1.4 mg/ml sildenafil. For intraperitoneal

injections of vardenafil, pharmaceutical grade vardenafil was stored frozen in DMSO at

1.5 µg/ml and diluted 1/10 in PBS prior to injection (100 µl twice daily). Linaclotide was

prepared by grinding the contents of Linzess® capsules in di-H2O using a tissue

homogenizer and stored in aliquots at -80°C. Mice were gavaged daily with 100 µl of the

linaclotide suspension at a concentration of 2.07 µg/ml. Loperamide HCl (Sigma-

Aldrich) was dissolved in di-H2O at a concentration of 2.5 mg/ml. Mice were gavaged

with 100 µl of loperamide.

iv. Euthanasia

Two methods of euthanasia were used, in accordance to the American Veterinary

Medical Association guidelines. Mice were first placed in a chamber devoid of CO2 that

has a filter lid to allow the flow of CO2 gas into the chamber. The flow rate of 2L/min allows for an anesthetic effect prior to CO2 asphyxiation. After the allotted time (3

minutes after the last sign of breath) decapitation was performed as the secondary method

of euthanasia. Both procedures were approved by the Augusta University Institutional

Animal Care and Use Committee.

D. cGMP assay

To measure cGMP in the mucosa of untreated and treated mice, a cGMP EIA Kit

(Cayman Chemical) was used. Briefly, mucosal tissue from control and treated mice was

scraped into 100µl 0.1M HCl. The samples were incubated on ice for 30 minutes,

vortexed every 10 minutes. The samples were then spun down at 14,000 RPM for 10

minutes. The supernatant was diluted 1:3 in EIA buffer. 50µl of each sample and

standard was loaded into a well of an ELISA strip with 50µl of cGMP AChE tracer and

50µl EIA antisera. The plate was incubated overnight at 4°C. The next morning, the wells

were washed 5 times with 200µl of washing buffer. After the washing step, 200µl of

Elman’s reagent was added to each well and the plate was incubated in the dark for 90

minutes, until the samples and standards began to turn color. The plate was read by a

multi-plate reader at a wavelength of OD405. A standard curve was created using high

and low cGMP concentrations for each experiment. A serial dilution of cGMP stock

(ranging from 30 pmol/ml to 0.23 pmol/ml) was measured alongside samples at OD405.

The measurements were plotted on a scatter plot where the Y axis was OD405 and the X- axis was log([cGMP] in pmol/ml). To determine the cGMP concentration of a sample,

[cGMP]= 10[(Y-b)/a]. Each sample was standardized to protein per sample, measured by

OD280. Relative level of cGMP per sample was presented as [cGMP]/OD280.

E. Immunohistochemistry i. Tissue Preparation

Intestine and colon from control and treated mice were removed, flushed with ice-cold

PBS, and Swiss rolled then placed in 10% formaldehyde. After 24 hours in formaldehyde the tissues were switched to 70% ethanol, embedded in paraffin blocks, and sectioned to 5µm by the Augusta University histology core. Sections were then processed as follows: deparaffinization with xylene (3x 10min), rehydration with decreasing concentrations of ethanol (100%, 90%, 80%, 70%, 50%, 5 minutes each), antigen-retrieval by boiling in citrate buffer (microwaved until boiling, then incubated with slides in a 135°C oven for 30 minutes). ii. Antigen blocking

After boiling in citrate buffer, the slides were allowed to cool in the citrate buffer. Slides were then washed 3x 5 minutes in 1x PBS. Slides were then incubated in 3% hydrogen peroxide for 10 minutes, then washed 3x 5 minutes in 1x PBS. Then slides were incubated in 5% goat serum at 37°C for 1 hour.

iii. Protein Detection

A hydrophobic pen was used to draw a barrier around the tissue sections to hold the

primary and secondary antibodies in place. Slides were incubated at 4°C overnight in

primary antibody diluted in blocking serum (various concentrations used depending on

antibody). The next morning slides were washed 3x 10 minutes in 1x PTS. Secondary

antibody was added to each slide for 1 hour at room temperature. Slides were washed 3x

10 minutes in 1x PTS. To detect the protein on slides, 3,3′-Diaminobenzidine (DAB) was

applied to each slide until brown color appeared on the tissue. Hematoxylin was used as a

counterstain to detect nuclei in the tissue, with sodium bicarbonate used to “blue” the

hematoxylin. Briefly, slides were placed in hematoxylin for 10-15 seconds each, then

immediately rinsed with water. The slides were placed in a slide apparatus in a tank in the

sink where running water was allowed to gently wash over them for 3 minutes. Slides

were placed in 0.1% sodium bicarbonate for 1 minute to cause bluing of the hematoxylin.

After counterstaining, the hydrophobic pen was carefully removed from around the tissue sections on each slide. The slides were then dehydrated with increasing concentrations of

ethanol (50%-100%, 1 minute each). Coverslips were mounted onto the slides with

permanent mounting media and allowed to dry overnight before imaging.

iv. ABC Detection

Visualization of Ki-67 and cleaved caspase antibodies was done using the ImmunoCruz

ABC kit (Santa Cruz Biotechnology). After deparaffinization, rehydration, and antigen

retrieval, slides were washed in 1X PBS (3x 5 minutes) then incubated in 1% hydrogen

peroxide for 10 minutes. Slides were then incubated in 1.5% blocking serum (goat serum

included in the kit) at 37°C for 1 hour. The slides were incubated in primary antibody

overnight at 4°C. The next morning, slides were washed 3x 5 minutes in PBS then

incubated for 30 minutes with a biotinylated secondary antibody. After the incubation,

slides were washed 3x 5 minutes in PBS, then incubated for 30 minutes with AB enzyme

reagent (avidin and biotin, included in the kit). Slides were then washed 3x 5 minutes in

PBS and processed according to the DAB protocol in section iii.

v. Antibodies Used

The tissues were probed using antibodies to CgA (rabbit, 1:200; Immunostar), Ki-67 (rat,

1:100; Dako Cytomation), cleaved caspase-3 (rabbit, 1:500; Cell Signaling, Danvers,

MA), and 5-HT (rabbit, 1:5000; Immunostar). Secondary antibodies were diluted 1:3000

in blocking buffer and were anti-rabbit (BioRad) and biotinylated anti-rabbit (Vector

Labs). To ensure the specificity of the 5-HT and CgA antibodies, we used a commercially available mixed IgG pool as an isotype control that stained nothing in our sections.

Hematoxylin and eosin (H&E) staining to visualize tissue structure and Alcian Blue

Periodic acid schiff (AB/PAS) staining to visualize goblet cells were performed by the

Augusta University histology core.

F. Western blotting

i. Acrylamide gel preparation

Before setting up the gel apparatus, twelve 1.5mm plates and glass were cleaned with

70% ethanol. Twelve sets of plates and glass were loaded into the gel apparatus,

separated with plastic plates. To make separating gel for one set of twelve 1.5mm midi

gels the following were mixed together: 59.19 ml diH2O, 30 ml 4x lower gel buffer,

30.15 ml (for 10%) ml (for 12%) 40% acrylamide 37.5:1 (Bio-Rad), 600 µl 10%(w/v)

ammonium persulfate, 60 µl TEMED. The mixture was carefully poured into the gel

apparatus and left to polymerize for 1 hour at room temperature. Stacking gel was

prepared by mixing the following: 25.19 ml diH2O, 10 ml 4x upper gel buffer, 3.9 µl

40% acrylamide 37.5:1 (Bio-Rad), 200 µl 10%(w/v) ammonium persulfate, 20 µl

TEMED. The mixture was poured into the top of the gel apparatus and combs were

carefully inserted to avoid causing bubbles under the combs. The gels were allowed to

polymerize for 1 hour at room temperature or 4°C overnight. Gels were stored in plastic

wrap at 4°C until use.

ii. Sample preparation

Mucosal tissue was scraped into tubes with ice-cold lysis buffer containing a protease and phosphatase inhibitor cocktail (Calbiochem). The tissues were incubated at 4°C for 30

minutes, vortexing every 10 minutes. The tubes were then centrifuged at 14,000Xg for 10

minutes at 4̊C. The supernatant was transferred to a new tube, reserving 25 µl for a

protein assay. 5X SDS-PAGE buffer was added to the remaining supernatant and the

samples were boiled for 8 minutes at 100°C. Samples were stored at -80°C for long-term storage and at -20°C for short-term storage. iii. Preparation of blots

The prepared samples were run out on 1.5mm gels at 30mA/gel for protein separation.

Gels were blotted to 0.45µm nitrocellulose membrane (Bio-Rad) at 100V for 1 hour.

Membranes were blocked with 5% BSA blocking buffer for 1 hour at room temperature.

Primary antibodies for target proteins were diluted in blocking buffer (concentration varies depending on antibody) and placed on the membrane at room temperature for 1 hour or 4°C overnight. Membranes were washed with 1X PTS by rocking 3x 10minutes before adding secondary antibodies. Secondary antibody was diluted in blocking buffer and added to the membrane for 1 hour at room temperature. Membranes were washed again 3x 10 minutes with 1x PTS. Clarity Western ECL Substrate (BioRad) was mixed at a ratio of 1:1 and added to the membrane for 1 minute. The chemiluminescent substrate was blotted off of the membrane with filter paper and the membrane was placed between plastic sheets and loaded into a cassette for x-ray film imaging. iv. Antibodies used

Blots were probed with antibodies against pVASP, Ser 239 (rabbit, Cell Signaling,

1:1000), total VASP (rabbit, Cell Signaling,1:1000), claudin-4 (mouse, Santa Cruz,

1:1000), occludin (rabbit, AbCam, 1:50000), and β-actin (mouse, Sigma, 1:3000). All purchased antibodies were stored at the temperature suggested by the manufacturer. They were diluted before use in Western blot blocking buffer, stored at 4°C, and recycled. Goat

anti-mouse/rabbit secondary antibodies (Biorad) were diluted 1:3000 in western blot

blocking buffer.

G. DSS-induced inflammation recovery model

To induce inflammation in the colon of mice, eight week old C57/BL6 mice were treated with 3% DSS (m.w. 36000-50000; MP Biomedicals, Santa Ana, CA, USA) ad libitum in drinking water for five days. The DSS was changed out every three days. Body weight and disease symptoms were monitored throughout the treatment period and during the entire recovery period. After the DSS treatment, mice were allowed to recover with normal drinking water for three weeks. After two weeks of recovery, mice were randomized into control and treatment groups. Long term sildenafil (ad libitum in drinking water) or linaclotide (daily oral gavage) treatment was started during the last week of recovery. Acute sildenafil and linaclotide treatments (oral gavage) were administered one hour prior to beginning the transit assay. Intestine and colon from 3 week recovered mice were collected for histology to measure the amount of residual inflammation, and RNA/protein to measure levels of 5HT and SERT.

H. Intestinal transit assay

Intestinal transit assay was modified from previously reported methods [102, 103].

Briefly, animals were fasted overnight before the experiments, with free access to water.

Mice were gavaged with 100 µl of charcoal meal (10% charcoal, 5% gum acacia in di-

H2O) and placed individually into clean cages for monitoring. Food was given back to the

mice immediately following the charcoal gavage. For total GI transit, mice were

monitored every five minutes and the time from gavage of the charcoal meal to expulsion

of a charcoal-marked fecal pellet was recorded [104]. For upper GI transit, mice were

sacrificed ten minutes after administration of the charcoal meal and the small intestines

were excised. The distance to the charcoal front and total length of small intestine from

the stomach to the cecum were both measured to calculate upper intestinal transit. For

acute drug studies, mice were gavaged with sildenafil or linaclotide one hour before the

charcoal meal. Long term administration of the drugs started five days before the transit

assay was to be performed.

I. Fecal Water Content

To determine the water content of the feces, individual fecal pellets were collected from

each mouse, weighed, and placed into a PCR tube strip. The pellets were dried in a 105°C

oven for 24 hours. Each individual pellet was then re-weighed. Water content was

calculated using the following formula: ((Wet weight - dry weight)/wet weight)*100.

J. Barrier permeability

Six to eight week old C57/BL6 mice were either untreated, or treated with 2% DSS for five days. The DSS-treated mice were randomized into two groups: control and sildenafil.

Mice were treated with sildenafil by oral gavage the night before and one hour prior to fluorescein isothiocyantate-dextran (FITC-dextran) administration. To measure barrier permeability, each group of mice was gavaged with 100µl of FITC-Dextran (100 mg/mL;

4kD; Sigma-Aldrich) after an overnight fast. 90 minutes after gavage, blood was collected from the submandibular vein from each mouse and placed in BD Microtainer

SST serum collection tubes (Becton, Dickinson and Company, Franklin Lakes, NJ). The tubes were inverted 5 times and allowed to clot for thirty mins. The tubes were then

centrifuged for 90 seconds at 13000Xg. The serum was pipetted into a 96-well clear

bottom plate and was analyzed by fluorimetry at 490 nm to determine the concentration

of FITC-dextran.

K. Apcmin/+ model of colon cancer

The Apcmin/+ mouse model of colon cancer is a well-respected model of the human condition familial adenomatous polyposis (FAP). The Apcmin/+ colony from Jackson Labs

was developed following N-ethyl-N-nitrosourea (ENU) treatments. The mice were found to develop anemia and multiple intestinal neoplasia, predominantly in the small intestine

[105]. Min mice develop more than 30 polyps in their small intestine and many die by

120 days of age. Sequencing of the mice showed a T to A transversion at nucleotide

2549, which results in a Stop codon nonsense mutation which creates a truncated form of

Apc [106]. Although Min mice develop polyps in the small intestine in contrast with

human FAP patients who develop hundreds of colonic polyps, it is still a widely accepted

murine model of colon cancer and FAP for studying chemoprevention.

i. Apcmin/+ colon cancer study

Male C57/BL6J-ApcMin/+ mice and female C57/BL6J mice were obtained from Jackson

Laboratory at age 6 weeks and separated into breeding cages. Pups were genotyped at 3

weeks of age using primers to detect wild-type and mutated Apc. Mice heterozygous for mutant Apc are used in this study, as homozygotes die in utero. ApcMin/+ mice were

randomly separated into three groups- control, sildenafil, and linaclotide. Mice started

receiving drug treatment at age 4 weeks. Body weight was monitored weekly to track

disease progression, as anemia and weight loss are symptoms of late-stage disease. At the

end of 14 weeks, the animals were sacrificed by CO2 asphyxiation. The entire GI tract

from stomach to rectum was removed and flushed out with ice-cold PBS. The GI tract was then divided into four sections- three portions of small intestine, from stomach to duodenum, and the entire colon. The tissue was opened longitudinally and spread out on a light box. High-resolution images were captured and enlarged for identification and selection of polyps. ImageJ software was used to count and determine the size of polyps.

After capturing images, representative polyps were collected for RNA and protein analysis, as well as histological analysis. ii. Histopathology

Tissue from ApcMin/+ mice were formalin-fixed flat between two sheets of filter paper

overnight. The next day the formalin was removed and replaced with 70% ethanol. The

strips of intestine were embedded in low melt agar and loaded into a tissue cassette so

that when cut the sections will show cross sections of the intestine. The tissue cassettes

were then embedded in paraffin and sectioned at 5µm thickness by the Augusta

University Histology Core. Sections were then processed as follows: deparaffinization

with xylene, rehydration with decreasing concentrations of ethanol (100%-50%), antigen-

retrieval by boiling in citrate buffer. Tissues were stained directly by the Augusta

University histology core with H&E and Alcian blue/periodic acid Schiff (AB/PAS) to

visualize polyps and goblet cells. The tissues were probed using antibodies against Ki-67

(1:100; Dako Cytomation) and cleaved caspase-3 (1:500; Cell Signaling, Danvers, MA,

USA). Antibody visualization was done using an anti-rat/anti-rabbit ImmunoCruz ABC

kit (Santa Cruz Biotechnology) to enhance DAB staining. Mucus density and

proliferative and apoptotic indices were quantitated using ImageJ software. Briefly, for

surrounding tissue, 10 different sections containing approximately 8 crypts per section

were analyzed for each mouse. For polyps, three sections from four different polyps per

mouse were analyzed.

iii. Analysis of gene expression

Mucosal tissue was scraped into TRIzol reagent (Life Technologies) and polyps were removed with scissors and forceps, placed in TRIzol and flash-frozen. RNA was DNAse treated (TURBO DNA-free kit, Life Technologies) then converted to cDNA using M-

MLV reverse transcriptase (Invitrogen™). Quantitative PCR (qRT-PCR) analysis of the cDNA was performed using SYBR Green PCR Master Mix (Applied Biosystems).

Relative expression levels were calculated using the 2–ΔΔCT method with β-actin

(ACTB) as a reference. Amplifications were performed in triplicate wells and melt curve

analysis was done to confirm the specificity of the primers used. Primers were designed

using Primer Blast Software (NCBI; Table 1)

L. Statistical analysis

All statistical analysis was performed using GraphPad Prism 7.0 (GraphPad Software,

Inc.). The statistical significance was set at p<0.05. All data were expressed as mean ±

SEM, unless otherwise stated in the figure legend. Either a Student’s t-test or One-Way

ANOVA with post analysis were used for comparison of means.

Table 1. Mouse primers used for real-time quantitative PCR.

Gene Name Sequence (forward/reverse, 5’-3’) Slc6a4 CCAAGAGCTAGTCAGGGTCC / CCAACGGCACTAACCTCCAC Tph1 CAGCAAGGACGGGATCAACT / AACGGGGTCCCCATGTTTG Actb GTACTCTGTGTGGATCGGTGG / AACGCAGCTCAGTAACAGTCC Tnf CCCACTCTGACCCCTTTACT / TTTGAGTCCTTGATGGTGGT Il1b CCCAACTGGTACATCAGCAC / TCTGCTCATTCACGAAAAGG Ifng CCATCAGCAACAACATAAGCGTC / TCTCTTCCCCACCCCGAATCAGCAG Guca2a TGTCCTGGTAGAAGGGGTCA / CGGGGAGCAAACTTCTTGTG Guca2b TGGAAGC-CATGGTACTTGATG / AAGCGAGGC-CATGTCAAGAA Gucy2c ACCGACAGTGGAGAACCAAC / TCTTATAGCTGGC-CACCCGA Prkg1a TGAGCTTCCAGC-CCATTCAG / AACTTTTCCTGC-CGAGACGG Prkg1b GGAAAGTTGATCCGAGAGGG / TTCTCCTGGAGCGCATACTG Prkg2 GGGCATCTGTGATGAGCTTT / CAACCACCCGAACCTATGAC Pde5a GGCAAGCACCATGGAACGAG / TGCGTTGACCATGTCTCTGG Pde9 ATGTGGATGAACCGTGGGAT / CGGGCACCGTTCTGACATTTA Pde10a AAGATGGACCCTCTAACAATGCG / TCCACAGTCTCTGCACTAACA

IV. RESULTS

A. The effect of cGMP-elevating agents on intestinal homeostasis

i. PDE-5 inhibitors increase cGMP levels in the mouse colon mucosa

Guanylin and uroguanylin are intestinal hormones that increase cGMP by binding and

activating epithelial GC-C. With this endogenous cGMP-generating system, blocking

cGMP degradation using PDE5 inhibitors can increase cGMP levels. It was recently

reported that increasing cGMP levels by intraperitoneal (IP) injection of the PDE5 inhibitor vardenafil had profound effects on homeostasis in the colon epithelium [69]. In

the present study, a single-dose of the PDE5 inhibitor sildenafil administered by IP

injection or oral gavage increased cGMP levels in the colon mucosa (15.2-fold and 11.7-

fold respectively; p<0.001; Fig. 4A). Furthermore, sildenafil administered to mice ad

libitum in drinking water also increased cGMP levels in the colon (8.6-fold; p< 0.001;

Fig. 4A). At the doses used in our studies, inclusion of sildenafil in the water did not significantly affect daily water intake compared to controls (Fig. 4B). Vasodilator- stimulated phosphoprotein (VASP) is widely used to measure PKG activation by cGMP.

Treatment of mice with sildenafil led to increased phosphorylation of VASP at the

Ser239 site which is a target site for PKG (Fig. 4C). The most abundant increase occurred

after 3 days of ad libitum treatment in the drinking water.

Figure 4. Sildenafil increases cGMP levels in the colon epithelium via three delivery methods. (A) cGMP levels in the colon mucosa of untreated mice (Ctrl), 5 hours after oral gavage (Gav) or intraperitoneal injection (IP) of sildenafil, or mice provided sildenafil in water ad libitum for 5 days. (B) Daily intake by mice provided water alone (Control) or water containing sildenafil (5.7 mg/kg). (C) Western blot of pVASP expression in colon mucosa from mice treated with sildenafil for various amounts of time. Data are shown as means, error bars represent SEM, n=6. (A,B) n=6 mice per group, * p<0.05, ** p< 0.0005 by two-tailed student’s t-test.

ii. cGMP-elevating agents regulate homeostasis in the mouse colon mucosa.

In agreement with the previous study using IP administered vardenafil, administration of

sildenafil to mice for five days ad libitum in drinking water significantly increased the number of differentiated goblet cells and enteroendocrine cells in the colon epithelium

(Fig. 5). Sildenafil treatment also reduced the proliferating compartment and the number of apoptotic cells at the luminal surface (Fig. 6). Previous work in our lab using PDE5 inhibitors have primarily focused on its protective effects in the colon epithelium. This study sought to expand the current knowledge of PDE5 inhibitors in the colon by looking at their effects in the intestine as well. In agreement with the data collected from the colon, sildenafil treatment significantly increased the number of secretory cells

(enteroendocrine and goblet cells) within the villi of the intestine (Fig. 7). Interestingly, sildenafil treatment did not alter proliferation or apoptosis in the intestinal epithelium

(Fig. 8) as we see in the colon.

Guanylin and GC-C knockout mice exhibit altered homeostasis in the colon, but the effects of exogenous GC-C ligands on intestinal epithelial homeostasis have not previously been examined. It was observed here that administration of linaclotide to mice had a similar effect on homeostasis as the PDE5 inhibitors. Within the colon epithelium, linaclotide treatment increased secretory cell density (Fig. 5) and decreased proliferation and apoptotic cells (Fig. 6). Linaclotide treated mice showed increased numbers of secretory cells in the intestinal villi, although the goblet cell increase did not reach statistical significance (Fig. 7). The proliferation rate within the intestine was not affected

by linaclotide; interestingly, we observed an increase in apoptotic cells at the tip of the

villi in mice treated with linaclotide (Fig. 8). While there is currently not an explanation

for this increase in apoptosis, other groups using GC-C agonists have reported similar

findings.

Chromogranin-A recognizes generalized enteroendocrine cells; however, there are many

different types of enteroendocrine cells which are classified depending on what hormone

they make and secrete. One of the most widely studied subtypes of enteroendocrine cells

are enterochromaffin (EC) cells which synthesize and secrete serotonin (5HT). Serotonin has many functions in the gut including intestinal motility and enteric nociception. The increase in enteroendocrine cells with PDE5 inhibitors is made up in large part by an increase in EC cells (Fig. 8). In addition to seeing an increase in EC cell density, we also observed an increase in the rate-limiting enzyme in serotonin synthesis, tryptophan hydroxylase (Tph1) in the mucosa of mice treated with vardenafil. This suggests that

PDE5 inhibitors are increasing enterochromaffin cells and serotonin signaling in the gut.

Figure 5. cGMP elevating agents increase secretory cell number in colon epithelium. . (A,C) Quantification and representative images of enteroendocrine cells (CgA) and (B,D) goblet cells (AB/PAS) in the colon epithelium of CD1 mice untreated (Ctrl) or treated with sildenafil (Sild) or linaclotide (Lin) for five days. Scale bars represent 50µm. Data are shown as means, error bars represent SEM. n=3 mice per group. * p<0.05, ** p<0.005, by two-tailed students t-test.

Figure 6. cGMP-elevating agents decrease apoptosis and proliferation in the colon epithelium. (A,C) Quantification and representative images of apoptotic cells (CC3) and (B,D) proliferating cells (Ki-67) in the colon epithelium of CD1 mice untreated (Ctrl) or treated with sildenafil (Sild) or linaclotide (Lin) for five days. Scale bars represent 50µm. Data are shown as means, error bars represent SEM. n= 3 mice per group. * p<0.05 by two-tailed students t-test.

Figure 7. cGMP-elevating agents regulate secretory cell density in intestinal epithelium. (A,C) Quantification and representative images of enteroendocrine cells (CgA) and (B,D) goblet cells (AB/PAS) in the intestinal epithelium of CD1 mice untreated (Ctrl) or treated with sildenafil (Sild) or linaclotide (Lin) for five days. Scale bars represent 50µm. Data are shown as means, error bars represent SEM. n= 3 mice per group. * p<0.05 by two-tailed students t-test.

Figure 8. cGMP-elevating agents do not effect proliferation and apoptosis in the intestinal epithelium. (A,C) Quantification and representative images of apoptotic cells (CC3) and (B,D) proliferating cells (Ki-67) in the intestinal epithelium of CD1 mice untreated (Ctrl) or treated with sildenafil (Sild) or linaclotide (Lin) for five days. Scale bars represent 50µm. Data are shown as means, error bars represent SEM. n=3 mice per group. * p<0.05 by two-tailed students t-test.

Figure 9. PDE5 inhibitors increase enterochromaffin cells in the colon epithelium of mice. (A) Dual- staining of enteroendocrine cells (CgA, red) and enterochromaffin cells (5-HT, green) in the colon epithelium of CD-1 mice untreated (Ctrl) or treated with vardenafil for five days. Scale bars represent 50µm. (B) Quantification of enteroendocrine cells (CgA) and enterochromaffin cells (5HT) in the colon epithelium of CD-1 mice untreated (-) or treated with vardenafil (+) for five days. The inset shows expression of tryptophan hydroxylase (Tph1) in colon mucosa scraped from untreated (-) or vardenafil treated (+) mice. Data are shown as means, error bars represent SEM. n=3 mice per group. * p<0.05 by two-tailed student’s t-test.

iii. Pharmacodynamics of sildenafil and linaclotide in the intestine

The initial studies investigating the effect of sildenafil on gut homeostasis only looked at one effective dose of the drug. This study sought to expand our knowledge of the effects of sildenafil on gut homeostasis by investigating different doses of the drug. The original dose from preliminary studies was originally calculated to represent four-times the human dose prescribed for erectile dysfunction, 100mg per day. Based on allometric scaling of the mouse doses into an equivalent human dose (factor of 12.3), the dose was actually calculated to be a pediatric dose of 32.4 mg. We found that the doses of sildenafil equivalent to a pediatric human dose were able to decrease proliferation and apoptosis and increase secretory cells in both the intestine and the colon (Fig. 10, 11).

The effects were saturable at the human dose equivalent prescribed for erectile dysfunction.

Linaclotide is prescribed for constipation as a once-daily 145 µg pill. Daily oral gavage of various doses of linaclotide to mice for five days sought to determine the maximal effective dose of the drug on homeostasis. Linaclotide exerted the strongest effect on homeostasis at a very low human dose equivalent of 2.8 µg per day, with a weaker effect being observed at higher doses of the drug (Fig. 12 A, B). This effect is also reported in human clinical trials with linaclotide, where low doses induced more diarrhea and side effects that the higher doses that are prescribed. While the explanation behind this phenomenon has not been investigated, it is possible that higher doses of the drug promote excess secretion into the lumen which dilutes out the remaining drug in the lumen, thereby causing weaker homeostatic effects.

Intraperitoneal administration of vardenafil was reported to increase cGMP levels in the

colon mucosa after only 4 hours; however, the majority of the homeostatic parameters

were measured after three to five days of once-daily vardenafil treatment [69]. This

study sought to determine the dynamics of the homeostatic changes that are observed

with continual ad libitum treatment with sildenafil in drinking water. Sildenafil takes

three days to induce its maximal effect on secretory cell number (Fig. 13A). However,

sildenafil only takes twenty-four hours to effect proliferation and apoptotic rate (Fig. 13

B, C). These effects take continual treatment, as giving one dose of the drug and waiting

for five days is not enough to alter homeostasis (Fig. 14A). In addition, the homeostatic changes are not permanent, as the gut will return to normal after a week without exposure to the drug (Fig. 14B).

Figure 10. Dose effect of sildenafil on homeostasis in the intestinal epithelium. Quantification of (A) goblet cells (AB/PAS) and (B) apoptotic cells in the intestinal epithelium of mice treated with various doses of sildenafil ad libitum in drinking water for five days. Data are shown as means, error bars represent SEM. n= 3 mice per group. * p<0.05 by two-tailed student’s t-test.

Figure 11. Dose effect of sildenafil on homeostasis in the colon epithelium. Quantification of (A) goblet cells (AB/PAS) and (B) apoptotic cells in the colon epithelium of mice treated with various doses of sildenafil ad libitum in drinking water for five days. Data are shown as means, error bars represent SEM. n= 3 mice per group. ** p<0.005 by two-tailed student’s t-test.

Figure 12. Dose effect of linaclotide on homeostasis in the intestinal epithelium. Quantification of goblet cells (AB/PAS) in the (A) intestinal epithelium and (B) the colon epithelium of mice treated with various doses of linaclotide via oral gavage for five days. Data are shown as means, error bars represent SEM. n= 3 mice per group. * p<0.05, ** p<0.005 by two-tailed student’s t-test.

Figure 13. Sildenafil effect on homeostasis in the colon epithelium over time. Quantification of (A) enteroendocrine cells (CgA), (A) goblet cells (AB/PAS), (B) proliferating cells (Ki-67), and (B) apoptotic cells (CC3) in the colon epithelium of mice treated with sildenafil ad libitum in drinking water for 24, 48, 72, or 120 hours. Data are shown as percentage of maximum change. n=3 mice per group. * p<0.05 by two- tailed student’s t-test.

Figure 14. Homeostatic changes require five days of continuous sildenafil treatment. (A) Mice were given one dose of sildenafil via oral gavage and allowed to recover for five days. Colons were harvested for histological analysis of goblet cells (AB/PAS). (B) Mice were treated with sildenafil ad libitum in drinking water for seven days, then allowed to recover with normal drinking water for seven days. Colons were harvested after seven days of treatment and after seven days of recovery for histological analysis of goblet cells (AB/PAS). Data are shown as means, error bars represent SEM. n= 3 mice per group. * p<0.05 by two-sided students t-test.

B. cGMP elevation for regulation of intestinal transit

i. Mice recovered from DSS-induced intestinal damage mimic post-infectious IBS

Since linaclotide is known to be beneficial in the treatment of IBS-C and CIC, it was hypothesized that sildenafil could also be used to increase bowel transit in constipation models. Since the therapeutic effects of linaclotide are thought to involve regulation of secretion and motility, it was of interest to determine whether sildenafil could affect bowel transit. However, it was found that neither sildenafil nor linaclotide treatment affected transit in healthy mice (Fig. 15A, B). One of the most common side effects of linaclotide is diarrhea due to over-activation of GC-C and subsequent secretion.

However, at the doses of the drugs used here, no change in fecal water content was observed (Fig. 15C).

As reported previously [107, 108], administration of DSS (3%) to mice for five days in drinking water induced a well-characterized inflammatory response that was accompanied by loss of body weight (Fig. 16A). Mice were allowed to recover for three weeks. At the time of the transit assay, body weight had normalized and disease symptoms were largely abrogated, suggesting that the mice had recovered from the DSS- induced intestinal inflammation. Histological analysis revealed that the recovered animals continued to exhibit evidence of a residual but stable disease marked by patches of edema and crypt loss (Fig. 16B). Intestinal transit was significantly slower in mice recovered from DSS compared to healthy mice (Fig. 17A). This change in transit was stable over several weeks (Fig. 17A). Although damaging effects of DSS are more pronounced in the

distal colon, the recovered mice exhibited slower upper intestinal transit as well as total intestinal transit (Fig. 17B). Despite the reduced transit in the recovered mice, fecal water content was higher than matched control animals (Fig. 17C).

Figure 15. Sildenafil and linaclotide have no effect on transit time or fecal water content in healthy mice. Mice were treated with sildenafil (ad libitum in water) or linaclotide (oral gavage) for five days prior to intestinal transit assay of (A) total transit or (B) upper intestinal transit. (C) Fecal pellets were collected from control and treated mice during the total transit assay and measured for fecal water content. Data are shown as means, error bars represent SEM, n=12.

Figure 16. DSS recovery model schematic. (A) 6-8 week old C57/BL6 mice were administered 3% DSS in drinking water for five days and allowed to recover for three weeks. Body weight was recorded throughout the study to monitor disease progression and recovery. (B) H&E and AB/PAS staining of a section of distal colon in a healthy and DSS-recovery mouse shows residual inflammation in the recovery mice. Arrows indicate edema (upper panel) and crypt loss (lower panel) in the recovery mice. Scale bars represent 100µm.

Figure 17. Mice recovering from DSS-induced inflammation have increased transit time and fecal water content. (A) Total transit time for healthy mice and DSS-recovery mice at different stages of recovery. Briefly mice were gavaged with a bolus of charcoal and the time from administration of charcoal to the expulsion of the charcoal in a fecal pellet was recorded. (B) Distance that a gavaged bolus of charcoal traveled in 10 minutes in mice recovered from DSS for three-weeks, compared to healthy controls. (C) Fecal water content of healthy and DSS-recovered mice. Data are shown as means, error bars represent SEM, n=6 mice per group. * p< 0.05, ** p< 0.001 by two-tailed Student’s t-test.

ii. Sildenafil reduces IBS symptoms in a DSS-recovery mouse model

The DSS-recovery mice exhibited reduced numbers of enterochromaffin (EC) cells in the

colon epithelium compared to healthy mice (Fig. 18A). Sildenafil treatment for five days

reduced the deficit of these 5-HT-staining cells, and partially returned the density to that

of healthy mice although it did not reach significance (Fig. 18B). In support of this

finding, mucosa from recovery mice showed lower expression of SERT and Tph1, two

major components in serotonin signaling (Fig. 18C). Both acute and long-term treatment

with sildenafil returned the levels to that of healthy mice. In addition, treatment of mice

with either sildenafil or linaclotide reduced transit times in the DSS-recovery mice almost

to the normal levels observed in healthy controls (Fig. 19A). Notably, even an acute dose

of either drug was effective at reducing intestinal transit time in this model. This result

indicates that EC cell density is unlikely to contribute to the therapeutic effect of either

drug in this model because the normalization of EC cell numbers required several days of

sildenafil treatment. In addition, small intestinal transit time was unaffected by sildenafil

treatment, suggesting that the regulation of transit in this model is largely due to

increased cGMP in the colon (Fig. 19B).

Despite the slower transit of the DSS-recovered mice, the total fecal output was paradoxically slightly higher than untreated animals (Fig. 20A). However, the total dry mass was significantly less in the diseased animals, indicating that the extra mass was

attributed to water in the stool. Indeed, the average percentage of water per fecal pellet

was higher in the diseased animals (Fig. 20B). Administration of either sildenafil or

linaclotide normalized the total fecal output and water content per fecal pellet. This effect

was observed whether the drugs were given acutely (one hour prior to measurement), or

longer term (5 days). Consistent with partial rescue of intestinal motility, sildenafil and

linaclotide partially restored the total dry mass. These results suggest that residual

inflammation in the colon of DSS-recovery mice more closely models PI-IBS, or low

grade ulcerative colitis. Indeed, barrier-disruption by acute DSS treatment was restored by sildenafil treatment (Fig. 21A). Knockout animals that are deficient in cGMP signaling components have a compromised intestinal barrier that has been suggested to be caused by reduced expression of junctional proteins [69, 71, 109]. Consistent with this idea, occludin and claudin-4 levels in the colon epithelium were increased by sildenafil treatment relative to untreated controls (Fig. 21B).

Figure 18. DSS recovery mice have a deficit of 5HT cells which is recovered with sildenafil treatment. (A, B) Quantification of 5-HT-positive cells in stained sections of colon from healthy mice, or those recovered from DSS treatment (DSS Recovery). Mice were either untreated controls (Ctrl) or treated with sildenafil for 5 days (Sild). Scale bars represent 100µm. (C) Expression of SERT and Tph1 in colon mucosa from healthy mice (WT), or those recovered from DSS treatment (DSS recovery). Mice were either untreated controls (Ctrl) or treated with sildenafil for 5 days (LT Sild) or one-hour (Acute Sild). Data are shown as means, error bars represent SEM, n=6 mice per group. ** p<0.005 by one-way ANOVA and Tukey’s post hoc analysis.

Figure 19. Sildenafil decreased intestinal transit time in a DSS-recovery model of IBS. (A) Total transit time and (B) upper intestinal transit were measured in healthy and DSS-recovery mice. Mice were either untreated (Ctrl) or treated with sildenafil or linaclotide acutely (1 hr; Acute Sild, Acute Lin respectively) and 5 days (Sild, Lin, respectively). Data are shown as means, error bars represent SEM, n=12 mice per group. * p<0.05, ** p<0.005, *** p<0.001 by one-way ANOVA and Tukey’s post hoc analysis.

Figure 20. Sildenafil regulates fecal water content in DSS-recovery mice. Fecal pellets from healthy mice, or those recovered from DSS treatment were collected to measure (A) total wet and dry fecal mass and (B) fecal water content. Mice were either untreated (Ctrl) or treated with sildenafil or linaclotide acutely (1 hr; Acute Sild, Acute Lin respectively) and 5 days (Sild, Lin, respectively). Data are shown as means, error bars represent SEM, n=12 mice per group. * p<0.05 by one-way ANOVA and Tukey’s post hoc analysis.

Figure 21. Sildenafil repairs leaky barrier and increases expression of tight junctional proteins in DSS-treated mice. (A) Barrier function was by gavaging mice with FITC-dextran and measuring the amount of FITC that has leaked into the blood after 90 minutes. Mice exposed to DSS were either untreated (Ctrl) or treated with an acute dose of sildenafil (Sild). (B) Levels of tight junctional proteins claudin 4 and occludin were measured by western blot in mucosa from untreated mice (Ctrl) and mice treated with sildenafil for 5 days (Sild). Data are shown as means, error bars represent SEM, n=6 mice per group. * p<0.05 by one-way ANOVA and Tukey’s post hoc analysis.

iii. Sildenafil suppresses loperamide-induced constipation

Linaclotide has been reported to relieve constipation by regulating both secretion and motility. It is unlikely that the normalization of transit conferred by sildenafil treatment of

DSS-recovered mice was due to increased secretion, since the fecal wet weight was reduced in treated animals. This suggests that the regulation of intestinal transit by sildenafil treatment was more likely due to a direct effect on intestinal motility. To further explore this idea, the ability of sildenafil to normalize transit in loperamide treated animals was examined. Loperamide is a µ-opioid receptor agonist that acts on the myenteric plexus of the large intestine to decrease smooth muscle contractions and therefore models opioid-induced constipation [110]. In addition to regulating transit in the DSS-recovery model, sildenafil was also able to normalize transit in loperamide treated mice. Administration of loperamide to mice nearly doubled the intestinal transit time (Fig. 22A). Both acute (one hour) and five day treatment of mice with sildenafil were able to block the loperamide-induced increase in intestinal transit time (Fig. 22A).

Loperamide also dramatically reduced transit in the small intestine, but this was not affected by sildenafil treatment (Fig. 22B). As expected for mice administered loperamide, the reduced transit was accompanied by a decrease in fecal water content

(Fig. 22C). In contrast to the DSS-recovery model described above, but consistent with the increased transit, sildenafil administration increased fecal water content in the loperamide treated mice.

Figure 22. Sildenafil normalizes transit in Loperamide-induced constipation model. Mice were either provided vehicle (healthy) or 10mg/kg loperamide (constipated). (A) Total intestinal transit and (B) distance traveled by a gavaged bolus of charcoal in 10 minutes were measured in untreated (Ctrl) or sildenafil treated (Sild) mice. Fecal pellets were collected from mice in the total transit experiment to measure (C) fecal water content. Data are shown as means, error bars represent SEM, n=12 mice per group. * p<0.05, ** p<0.001 by one-way ANOVA and Tukey’s post hoc analysis.

C. cGMP elevating agents for prevention of colorectal cancer

i. cGMP-elevating agents reduce tumor burden in intestine of Apcmin mice

Treatment with the GC-C agonist plecanatide or the PDE5 inhibitor sildenafil has recently been shown to suppress polyp multiplicity in inflammation-driven models of colon cancer in mice, although the mechanism of suppression is poorly understood [83,

84]. To better understand the inhibitory mechanism, the present study compared the two approaches of cGMP elevation in a non-inflammatory model of intestinal carcinogenesis.

Loss of heterozygosity at the Apc locus causes ApcMin/+ mice to develop numerous adenomatous polyps primarily in the small intestine (Fig 23A, B). Daily administration of sildenafil in drinking water or the GC-C agonist linaclotide by gastric gavage reduced the number of intestinal polyps per mouse (56% and 67%, respectively) compared to controls given water alone (Fig 23C). The polyps from treated mice did not differ significantly in size relative to control polyps (Fig 23D). Further analysis of polyp sections by IHC showed that sildenafil or linaclotide treatment of the host did not significantly affect proliferation or apoptosis of the polyps (Fig 24 A, B). The polyps from both control and treated animals were undifferentiated with respect to the presence of goblet cells but showed an increase in goblet cell density in the non-neoplastic periphery of the polyps (Fig. 24 C).

Figure 23. cGMP-elevating agents suppress tumorigenesis in the intestine of ApcMin/+ mice. Upon weaning, ApcMin/+ mice were randomly assigned to three groups, untreated (Ctrl) or treated for 10 weeks with sildenafil (sild) or linaclotide (lin). (A) Representative intestines from 14-week old ApcMin/+ mice and (B) representative image of an H&E stained intestinal section with a polyp. (C) Plot shows the effect of sildenafil and linaclotide on median polyp number per animal (n=10 per group). (D) Average polyp size from ApcMin/+ mice untreated (Ctrl) or treated for ten-weeks with sildenafil (sild) or linaclotide (lin). Data are shown as means, error bars represent SEM. n= 10 mice per group. ** p<0.005, *** p<0.001 by one- way ANOVA with Tukey’s post hoc analysis.

Figure 24. cGMP elevation has no effect on the phenotype of intestinal polyps from ApcMin/+ mice. Polyps from untreated and sildenafil (sild) or linaclotide (lin) treated animals were collected and processed for histological analysis of (A) proliferative index (Ki-67), (B) apoptotic index (CC3), and (C) mucus density (AB/PAS). Data are shown as means, error bars represent SEM. n= 3 mice, 6 polyps per mouse. ** p<0.005 by two-tailed student’s t-test.

ii. Sildenafil does not affect inflammation in the ApcMin/+ mouse intestine

Sildenafil has been reported to suppress myeloid-derived suppressor cell (MDSC) function and reduce inflammation in colonic polyps [84, 111, 112]. In order to determine

whether this phenomenon played a role in the present study, RT-qPCR was used to measure inflammatory cytokine expression in the ApcMin/+ mouse intestinal tissues. These

studies showed that the expression of IL-1β, TNFα, and IFNγ in the non-polyp tissue

from the ApcMin/+ mice intestine was comparable to the levels in wild-type mice (Fig 25

A-C). In contrast, the intestinal polyps showed dramatic increases in the mRNA levels of these inflammatory cytokines. Treatment of mice with sildenafil did not significantly reduce the cytokine expression in either the polyp or non-polyp tissue. In support of this, the slightly higher density of CD11b cells associated with polyps did not change in response to sildenafil treatment (Fig 25 D-E). These results support the idea that mucosal inflammation is not a central driver of intestinal carcinogenesis in the ApcMin/+ mouse,

and that the suppressive effect of cGMP is not due to effects on polyps.

iii. The expression of cGMP-signaling components in the ApcMin/+ intestinal polyps.

While both cGMP-elevating drugs examined in the present study reduced polyp

multiplicity in the ApcMin/+ intestine, neither drug affected polyp size, proliferation or

apoptosis. Similar observations were reported recently using the AOM/DSS

inflammatory carcinogenesis model, where a dramatic loss of cGMP-generating machinery in the polyps partially explained the phenomenon [84]. To determine whether

this also occurs in the intestinal polyps in ApcMin/+ mice, the expression of cGMP- signaling components was examined using RT-qPCR. These studies revealed that the mRNA levels of uroguanylin and GC-C were unchanged between non-polyp epithelium and the polyp tissue (Fig. 26A). Surprisingly, guanylin mRNA levels were significantly higher in the polyps compared to the surrounding tissue, but the physiological significance is not clear as the relative transcript levels were much lower than uroguanylin. The expression of the effector protein kinases PKG1α, β did not differ significantly between polyp and non-polyp tissue in the ApcMin/+ mouse intestine, but the relative mRNA level of the PKG2 isoform was elevated in the polyps (Fig 26B).

Expression-analysis of phosphodiesterases with activity toward cGMP (PDE5, 9, and

10a), revealed little change in PDE5 or PDE9, but significantly higher levels of the dual- specificity PDE10a were observed in the polyps compared to non-polyp tissue (Fig 26C).

However, the increased expression of PDE10a in the ApcMin/+ polyps was independent of sildenafil treatment.

iv. cGMP-elevating agents regulate homeostasis in pre-neoplastic tissue of ApcMin/+ mice

The suppression of tumorigenesis by both sildenafil and linaclotide in the absence of any effect on initiated polyps indicates that the principal effect of cGMP is on the preneoplastic epithelium. To explore this further, RT-qPCR analysis of cGMP signaling component mRNA expression in the non-polyp epithelium of the ApcMin/+ intestine was compared to wild-type animals. These studies revealed significantly less uroguanylin and

guanylin, but elevated levels of GC-C in the ApcMin/+ intestine (Fig 27A). To determine whether the altered expression of these components affected epithelial homeostasis, histological analysis of proliferation, apoptosis, and differentiation was examined. These results showed that the goblet cell numbers were reduced, but the intestine of ApcMin/+ mice exhibited more crypt proliferation and more apoptosis at the villus tip compared to wild-type animals (Fig 27B-D).

It has been reported that increasing cGMP in the intestine with PDE5 inhibitors can suppress proliferation and apoptosis in the wild-type mouse intestines but this has not been examined in ApcMin/+ mice. To better understand the effect of cGMP-elevating agents in the ApcMin/+ mice, the effects of linaclotide and sildenafil on homeostasis in the non-neoplastic intestinal epithelium was measured. Similar to the effect of these agents on the colon [69, 113], both drugs increased goblet cell density and suppressed proliferation (Fig. 28A, B). In contrast to previously reported effects on the colon however, both linaclotide and sildenafil treatment caused increased apoptosis at the tip of the villi (Fig. 28C).

Figure 25. Sildenafil treatment does not reduce inflammation within polyps of ApcMin/+ mice. (A, B, C) Real-time qPCR analysis of inflammatory cytokine gene expression in mucosa from C57BL/6 mice and mucosa and polyps from untreated or sildenafil-treated ApcMin/+ mice. n= 6 each. (D, E) Quantification and representative images of Cd11b+ cells in intestines of C57BL/6 mice and untreated or sildenafil-treated ApcMin/+ mice. Scale bars represent 50µm. Data are shown as means, error bars represent SEM. n= 3 mice, 3 polyps per mouse.

Figure 26. Expression of cGMP signaling components in intestinal polyps from ApcMin/+ mice. (A) The expression of cGMP generators was measured by real-time qPCR in non-neoplastic tissue and untreated and treated polyps from ApcMin/+ mice. (B) The expression of cGMP effectors were measured by real-time qPCR in non-neoplastic tissue and untreated and treated polyps from ApcMin/+ mice. (C) The expression of various phosphodiesterases were measured by real-time qPCR in non-neoplastic tissue and untreated and treated polyps from ApcMin/+ mice. n= 6 mice; * p<0.05, ** p<0.005 by One-way ANOVA and Tukey’s post hoc analysis.

Figure 27. Non-neoplastic epithelium from ApcMin/+ mice has reduced levels of endogenous GC-C peptides and increased proliferation and apoptosis. (A) Real-time qPCR analysis of uroguanylin, guanylin, and GC-C gene expression in mucosal tissue from C57BL/6 mice and non-neoplastic mucosal tissue from ApcMin/+ mice. (B, C, D) Quantification and representative images of goblet cells (AB/PAS), proliferation (Ki-67), and apoptosis (CC3) in stained sections of intestine from C57BL/6 mice and ApcMin/+ mice. A, n=6; B-D, n=3. * p<0.05, ** p<0.005, p< 0.0005 by two-tailed Student’s t-test.

Figure 28. Linaclotide and sildenafil treatment regulate homeostasis in non-neoplastic epithelium of ApcMin/+ mice. Non-neoplastic tissue from ApcMin/+ mice treated with sildenafil or linaclotide for ten weeks were collected and prepared for histological analysis. (A, B, C) Tissue was stained for goblet cell density (AB/PAS), proliferation (Ki-67), and apoptosis (CC3). Representative pictures of the staining are shown at the right of the graph. Arrows on Panel C indicate CC3 positive cells. * p<0.05, ** p<0.005, *** p<0.0005 by One-way ANOVA and Tukey’s post hoc analysis.

v. cGMP elevation has no effect on colon tumorigenesis in Apcmin/+ mice

While we saw a robust inhibition of intestinal tumorigenesis in Apcmin/+ mice treated with sildenafil or linaclotide, the same was not true for the colon. Mice treated with sildenafil or linaclotide did not show any difference in polyp number in the colon (Fig. 29 C). In the classical Apcmin/+ mouse model, the majority of the tumorigenesis occurs in the small intestine with very few polyps occurring in the colon [114]. However, the colony used in this study exhibited an average of five polyps in the colon compared to an average of less than one recorded in the model historically. The polyps in the colon were larger and more advanced than those in the intestines of the same mice (Fig. 29 A, B). While we do not have an explanation for the unusually high number of polyps we have recorded in the colon, it is of interest to note that at the time the drug administration began mice already exhibited a high number of lesions (Fig. 29 D). Since sildenafil is not acting on the intestinal polyps themselves, it is not surprising that pre-existing lesions in the colon are not inhibited by the drugs either.

Recent observations from the AOM/DSS inflammatory carcinogenesis model showed that sildenafil inhibited tumorigenesis in the colon but did not have a dramatic effect on the polyps themselves. In that model there was a dramatic loss of cGMP-generating machinery in the polyps which may explain why the drug had not effect on the polyps [84]. To determine if this was also a phenomenon in the colonic polyps of this model, the expression of cGMP-signaling components was examined using RT-qPCR. The colonic polyps from the Apcmin/+ mice were reflective of both the AOM/DSS polyps and human polyps as they had an almost complete loss of guanylin expression (Fig. 30). In addition, the polyps had a significant reduction of GC-C expression. The loss of these two key components of the cGMP signaling pathway would explain why sildenafil and linaclotide were not able to affect the pre-existing polyps in the colon.

Figure 29. Linaclotide and sildenafil treatment do not affect polyp number in the colon of ApcMin/+ mice. Upon weaning, ApcMin/+ mice were randomly assigned to three groups, untreated (Ctrl) or treated for 10 weeks with sildenafil (sild) or linaclotide (lin). (A) Representative colons from 14-week old ApcMin/+ mice and (B) representative image of an H&E stained colon section with a polyp. (C) Plot shows the effect of sildenafil and linaclotide on median polyp number per animal (n=10 per group). (D) Median polyp number from the colons of ApcMin/+ mice over time. Data are shown as medians, error bars represent standard deviation.

Figure 30. Expression of cGMP signaling components in colon polyps from ApcMin/+ mice. (A) The expression of cGMP generators was measured by real-time qPCR in non-neoplastic tissue and untreated and treated polyps from ApcMin/+ mice. (B) The expression of cGMP effectors were measured by real-time qPCR in non-neoplastic tissue and untreated and treated polyps from ApcMin/+ mice. n= 6 mice; * p<0.05by One-way ANOVA and Tukey’s post hoc analysis.

V. DISCUSSION

A. cGMP elevating agents regulate homeostasis in the intestinal epithelium

There is growing interest in exploiting the cGMP signaling pathway for the treatment of

gastrointestinal diseases. Cyclic GMP can be pharmacologically elevated in the intestine

by increasing its production by stimulating GC-C with agonists or by blocking the degradation of endogenously produced cGMP with PDE5 inhibitors. This study sought to determine the most effective way to increase cGMP signaling in the intestines and whether elevating cGMP is therapeutic for IBS-C or colorectal cancer. Data shown here demonstrate that the PDE5 inhibitor sildenafil can confer homeostatic changes in the intestinal epithelium when administered orally at pediatric doses prescribed to treat pulmonary arterial hypertension. The ability of the low dose sildenafil to regulate homeostasis in the intestinal epithelium is paramount to its potential use as a long-term treatment or prevention agent in various gastrointestinal diseases. Through studies of men taking sildenafil on a daily basis for benign prostatic hyperplasia, it has been shown that sildenafil can be prescribed in the long-term without detrimental side effects [115-117].

Combined with the new knowledge that a pediatric dose of the drug is as effective as a higher dose at regulating homeostasis within the intestine, this makes a promising safety profile for a preventative drug.

In addition to the new knowledge about sildenafil, this study also shows for the first time that the GC-C agonist linaclotide can confer similar homeostatic changes in the intestinal epithelium. The biological effects of linaclotide occur luminally by activating GC-C expressed specifically in the intestinal epithelium. This contrasts with sildenafil, which is absorbed by the gut and has systemic effects on many tissues in addition to the intestinal epithelium. Because linaclotide and sildenafil have similar effects on gut homeostasis, it is likely that the effect of sildenafil in this tissue is due to its ability to increase cGMP in the intestinal epithelium by amplifying the effects of endogenous GC-C ligand, rather than indirectly by affecting other tissues.

The mechanism underlying the effect of cGMP on intestinal homeostasis is poorly understood, but since both drugs required multiple days for maximal effect it is likely that gut turnover is an important component of the mechanism. We do not believe that cGMP is altering the stem cell compartment as removing the drug returns the proliferation rate and differentiated cell density to normal levels after a week. Because the proliferation and apoptotic rates were altered with the drugs after twenty-four hours but the increases in differentiated cells took three to five days, it is possible that cGMP is affecting the proliferating compartment, and the changes in differentiated cells are a result of the slower proliferation. There are a number of previous reports using mainly cell lines that suggest cGMP can regulate β-catenin activity and TCF target gene expression through activation of PKG1 [118-121]. However, we have shown in our lab that PKG1 is not expressed in the colon epithelium, rather in the smooth muscle and vasculature underlying the epithelium. While this does not fully rule out the possible involvement of

PKG1 in the reduction of proliferation, it is not highly likely that it is the mechanism

behind the changes we see. In addition to Wnt/β-catenin signaling, Notch also plays an important role in maintenance of the proliferative compartment. It is possible that cGMP elevation is inhibiting Notch signaling, thus causing a shift toward the secretory lineage of differentiated cells and a reduction in the proliferating compartment.

B. cGMP-elevating agents for the treatment of IBS-C and constipation

Irritable bowel syndrome is a significant health problem in the United States and there is a need for more effective treatments. The clinical success of the GC-C agonist linaclotide

demonstrates the utility of cGMP elevation in the intestinal epithelium for the treatment

of IBS-C and CIC patients. However, the mechanism of action is not fully understood

and its contraindication in pediatric patients is an important limitation.

The similar effects of sildenafil and linaclotide on intestinal homeostasis prompted us to

determine whether these drugs were also equally effective at treating constipation. There

are no rodent models that fully mimic human disease, but a widely used approach

involves allowing mice recover from bacterial infection, which mimics some aspects of

post-infectious IBS (PI-IBS), [122, 123]. Recovery from intestinal inflammation induced by chemical agents can also induce symptoms of IBS [124-126], and we demonstrate

here that mice recovered from DSS-induced colitis exhibit slower intestinal transit,

barrier dysfunction, and reduced EC cell density.

Treatment of DSS-recovered animals with either sildenafil or linaclotide increased intestinal transit in this model, but the mechanism of action downstream of cGMP is not clear. Sildenafil is absorbed systemically and has well-established relaxation effects on smooth muscle that might affect transit. Several studies have reported a reduction in esophageal contractile force by sildenafil treatment, particularly in patients with dysmotility disorders, but this effect is less pronounced in the normal esophagus and in the lower intestine [127-131]. Reduced muscular contractility cannot explain the results reported here because this would be expected to further increase transit time in our disease models, but as shown here, sildenafil reduced transit time. Moreover, sildenafil did not affect transit in healthy mice, nor did it affect transit in the upper gastrointestinal tract in the constipation models.

Altered serotonin function may play an important role in IBS; IBS-D patients often exhibit increased 5-HT signaling, and IBS-C patients typically have reductions in 5-HT signaling [132, 133]. The DSS-recovered animals were shown here to be deficient in EC cells, and sildenafil treatment partially restored their number. Our initial hypothesis was that restoration of EC cell density contributed to the normalization of transit in these animals. However, this is unlikely because the effect of the cGMP-elevating drugs on transit was rapid, whereas the effect on EC cells required several days. Linaclotide was found to attenuate visceral pain in mice and in human IBS and CIC patients [89, 134].

While the mechanism of analgesia remains unclear, it has been suggested that cGMP efflux by multi-drug resistance proteins (MRP4, 5) might directly suppress submucosal neuron excitability [134, 135]. It is possible that this phenomenon might also mediate the

effects of sildenafil on bowel transit in the work shown here. While linaclotide therapy

has been reported to relieve constipation symptoms rapidly, improvement in visceral pain

can take several days [88, 91]. This suggests that cellular changes in the gut such as

increased EC cell density might underlie the analgesic effects of linaclotide. This

intriguing idea further suggests that sildenafil treatment might also confer analgesia in

IBS patients, but whether the homeostatic changes observed in the mouse colon also

occur in human patients treated with cGMP-elevating agents has not yet been examined.

Chronic, low-grade inflammation is common in IBS patients [136], and a defective intestinal barrier is likely to contribute to increased diarrhea and visceral hypersensitivity

in IBS-D [133]. Several groups have shown that mice deficient in cGMP signaling

components (GC-C and guanylin) have a dysfunctional intestinal barrier [71, 109]. We

demonstrate here for the first time that pharmacologically increasing cGMP in the colon

epithelium can augment a dysfunctional epithelial barrier. Linaclotide is a secretagogue

and has been reported to increase fecal water content in mice and humans [88, 137, 138].

This effect of cGMP in the gut has prevented GC-C agonist use in patients with IBS-D or

IBD where diarrhea is a hallmark symptom. However, changes in fecal water content

were not observed in healthy mice with the doses of sildenafil and linaclotide used here,

indicating that low doses of either drug could affect homeostasis and barrier without

over-activating secretion. This suggests that PDE-5 inhibitors and GC-C agonists might

also be effective in treating PI-IBS or IBD patients, where the diarrhea is associated with

a dysfunctional epithelial barrier rather than hypersecretion. However, the potential risks

of systemic circulation of GC-C agonists in barrier-compromised patients will need

further study.

In contrast to the damaging effects of DSS on the distal colon, loperamide does not affect

intestinal barrier integrity, but works principally on the myenteric plexus to decrease tone

of intestinal smooth muscles. We show here that loperamide increased transit time in

mice, and as expected, this was associated with reduced fecal water content. Sildenafil

was shown to normalize both transit and fecal water content in the loperamide model.

While further study is necessary to identify the underlying mechanism, it is worthy to

note that sildenafil did not affect small intestinal transit that was slowed by loperamide

treatment. This suggests that the neuro-regulatory machinery that is responsive to cGMP regulation might be preferentially localized to the colon.

Taken together, the preclinical results shown here underscore the potential therapeutic value of PDE-5 inhibitors for the treatment of gastrointestinal disease. We demonstrated that sildenafil can normalize bowel transit in both a post-inflammatory and opioid- induced constipation models, suggesting possible benefit for human patients afflicted with IBS-C, CIC, and OIC. According to the prescribing information, preclinical studies revealed that linaclotide is lethal in juvenile mice due to increased fluid secretion leading to severe dehydration. The mechanism may be due to increased expression of GC-C in

the intestinal epithelium, which would amplify the effect of exogenous ligand. It is likely

that the increased GC-C expression is balanced by reduced endogenous ligand. This suggests that PDE-5 inhibitors are less likely to cause hypersecretion because they

increase the basal cGMP level regardless of GC-C expression. Indeed, sildenafil is well-

tolerated even long term in pediatric patients where it is used to treat pulmonary

hypertension. Results described here suggest that sildenafil might also benefit pediatric

patients suffering from CIC and IBS-C.

C. cGMP-elevating agents for chemoprevention of colorectal cancer

Recent reports show that the GC-C agonist plecanatide can reduce DSS-induced

tumorigenesis in the colon of ApcMin/+ mice [83], and that sildenafil can inhibit tumorigenesis in colon of AOM/DSS treated mice [84, 111]. However, these two classes of drugs have not been compared in a single model, and the tumor suppressive mechanism remains poorly understood. Results shown here demonstrate that the GC-C agonist linaclotide and the PDE5 inhibitor sildenafil were equally able to suppress polyp multiplicity in the intestine of ApcMin/+ mice. Moreover, the magnitude of the tumor suppressive effect in the intestine was the same as reported for sildenafil in the colon of

AOM/DSS-treated mice. Despite differences in tumor initiation and location, these observations underscore a common cGMP-dependent mechanism. PDE5 inhibitors have been shown to suppress DSS-induced inflammation, suggesting that this could play an important role in the tumor prevention effect of sildenafil [69]. However, the AOM/DSS study concluded that the tumor suppressive effect of sildenafil was mostly during the mutagenesis/initiation phase, and based on the timing of drug administration it was independent of inflammation [84, 111]. A separate study that used a very high dose of sildenafil in the AOM/DSS model suggested that the tumor suppressive effect is due to

blockade of myeloid cell function [84, 111]. However, the results shown here do not

support this idea because endogenous inflammation is not a driver of tumorigenesis in the

small intestine of ApcMin/+ mice, and in contrast to the colonic polyps in AOM/DSS mice,

inflammation in the intestinal polyps was not affected by sildenafil.

Sildenafil only slightly reduced proliferation in the colonic polyps of AOM/DSS-treated

mice, but notably increased mucus differentiation [84, 111]. This contrasts with the

intestinal polyps from ApcMin/+ mice, which were largely devoid of goblet cells and this

was not altered by treatment with either drug. The observation that neither linaclotide nor

sildenafil affected proliferation or apoptosis of established polyps is in contrast with

numerous reports describing antitumor effects of cGMP [119, 139, 140]. The previous

study of sildenafil in AOM/DSS treated mice, partly attributed the relatively small effect

of the drug on established polyps to reduced guanylin expression [84]. Guanylin levels are also reduced in human colorectal tumors relative to non-tumor tissue [80, 81, 141].

As GC-C ligands have been shown to have cytostatic effects in colon cancer cell lines

[68, 73, 140], “hormone replacement therapy” has been suggested as an approach to use

exogenous ligands to treat colon tumors [142-144]. The results shown here do not support this idea, as linaclotide did not affect mean size or proliferation of polyps. Moreover, the mRNA levels of uroguanylin and guanylin were not reduced in the intestinal polyps from

ApcMin/+ mice. These observations indicate that the loss of GC-C ligands is not the central

reason for a lack of effect of cGMP-elevating drugs on initiated polyps. In support of several previous studies, it was shown here that the expression of PDE10a increased dramatically in the intestinal polyps [84, 120, 121]. It is possible that this dual-specificity

PDE impedes the activation of cGMP signaling in tumors. Since inhibition of PDE10a

inhibits proliferation of colon cancer cell lines [120], whether such inhibitors might be

used as a treatment strategy for colon cancer will be an important area for future

investigation.

The lack of effect of either sildenafil or linaclotide on established polyps strongly

suggested that the reduced polyp formation resulting from treatment with these agents

was due to an effect on the preneoplastic tissue. A growing body of evidence

demonstrates that cGMP signaling controls intestinal homeostasis, and that increasing

cGMP can suppress proliferation while increasing goblet cell density [67, 70, 71, 113]. It

was shown here that treatment with either linaclotide or sildenafil caused an increase in

epithelial apoptosis in the intestine, which contrasts with previous reports of reduced

apoptosis in the colon epithelium [69, 113]. However, because the apoptosis is restricted

to the luminal border, and because these agents prevent polyp formation in both colon

and intestine, it is unlikely that the apoptosis contributes to the mechanism of tumor

suppression. A novel observation in the present study was the reduced expression of GC-

C ligands in the mucosa of ApcMin/+ mice relative to wild-type animals, and increased proliferation that is consistent with reduced cGMP signaling. While further study is needed to determine the significance of this to human FAP patients, both linaclotide and sildenafil were found to reduce proliferation in the intestine of ApcMin/+ mice. The

carcinogenic mechanisms underlying the AOM/DSS and ApcMin/+ colon cancer models

are different. In the AOM/DSS model carcinogenesis is due to mutations caused by

azoxymethane, whereas in the ApcMin/+ model it is due to bi-allelic loss of the tumor

suppressor gene Apc [114, 145]. The suppression of proliferation in response to cGMP

elevating agents would affect both the loss of heterozygosity and mutagenic efficacy, and

is therefore likely to be part of the tumor preventative mechanism. The mechanism by

which cGMP can reduce the proliferative compartment of both the intestine and colon of

mice is presently unknown. However, this novel interpretation predicts that increasing

epithelial cGMP is also likely to reduce tumorigenesis in sporadic and Lynch syndrome

associated lesions that are also a function of proliferation.

Taken together, the results shown here demonstrate that PDE5 inhibitors and GC-C

agonists can equally suppress intestinal tumorigenesis in mice. The equal efficacy of the

two drugs to suppress polyp multiplicity underscores the common mechanism involving

elevation cGMP in the preneoplastic epithelium resulting in a reduced proliferative

compartment. A recent report demonstrated that linaclotide administration can reduce

proliferation in the colon epithelium of human patients [146]. The present preclinical study therefore underscores the potential for targeting cGMP signaling for chemoprevention of colorectal cancer in human patients at increased risk.

VI. SUMMARY

1. cGMP-elevating agents regulate epithelial homeostasis in the

intestine and the colon.

We have shown that a “baby dose” of the PDE5 inhibitor sildenafil and the GC-C agonist linaclotide can regulate homeostasis in both the intestinal and colonic epithelium. In

addition, it was shown that homeostatic changes require five days of continual treatment

with cGMP-elevating agents.

2. cGMP-elevating agents regulate intestinal transit

While GC-C agonists are already available as a treatment option for patients with IBS-C

or CIC, they have many drawbacks and an alternative treatment is needed for some

patients. We show here that the PDE5 inhibitor sildenafil has similar efficacy as linaclotide for alleviating constipation in two mouse models by reducing intestinal transit time. In addition, we have found that sildenafil can repair leaky intestinal barrier.

3. cGMP-elevating agents suppress intestinal tumorigenesis

Both sildenafil and linaclotide were able to suppress tumorigenesis in the intestine of

ApcMin/+ mice. We believe that the effect is not on the polyps themselves but rather on the

preneoplastic tissue, specifically a reduction in aberrant proliferation.

REFERENCES

1. Shen, L., Functional Morphology of the Gastrointestinal Tract, in Molecular Mechanisms of Bacterial Infection via the Gut, C. Sasakawa, Editor. 2009, Springer Berlin Heidelberg: Berlin, Heidelberg. p. 1-35. 2. Turner, J.R., Intestinal mucosal barrier function in health and disease. Nat Rev Immunol, 2009. 9(11): p. 799-809. 3. Olsson, C. and S. Holmgren, The control of gut motility. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology, 2001. 128(3): p. 479-501. 4. Barker, N., et al., Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature, 2007. 449(7165): p. 1003-7. 5. Crosnier, C., D. Stamataki, and J. Lewis, Organizing cell renewal in the intestine: stem cells, signals and combinatorial control. Nat Rev Genet, 2006. 7(5): p. 349- 59. 6. Clarke, A.R., Wnt signalling in the mouse intestine. Oncogene, 2006. 25(57): p. 7512-21. 7. Gregorieff, A., et al., Expression pattern of Wnt signaling components in the adult intestine. Gastroenterology, 2005. 129(2): p. 626-38. 8. Giles, R.H., J.H. van Es, and H. Clevers, Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta, 2003. 1653(1): p. 1-24. 9. Gunther, C., et al., Apoptosis, necrosis and necroptosis: cell death regulation in the intestinal epithelium. Gut, 2013. 62(7): p. 1062-71. 10. Marchiando, A.M., et al., The epithelial barrier is maintained by in vivo tight junction expansion during pathologic intestinal epithelial shedding. Gastroenterology, 2011. 140(4): p. 1208-1218.e2. 11. Ramachandran, A., M. Madesh, and K.A. Balasubramanian, Apoptosis in the intestinal epithelium: its relevance in normal and pathophysiological conditions. J Gastroenterol Hepatol, 2000. 15(2): p. 109-20. 12. van der Flier, L.G. and H. Clevers, Stem cells, self-renewal, and differentiation in the intestinal epithelium. Annu Rev Physiol, 2009. 71: p. 241-60. 13. Birchenough, G.M., et al., New developments in goblet cell mucus secretion and function. Mucosal Immunol, 2015. 8(4): p. 712-9. 14. Mashimo, H., et al., Impaired defense of intestinal mucosa in mice lacking intestinal trefoil factor. Science, 1996. 274(5285): p. 262-5. 15. Schonhoff, S.E., M. Giel-Moloney, and A.B. Leiter, Minireview: Development and differentiation of gut endocrine cells. Endocrinology, 2004. 145(6): p. 2639- 44. 16. van Es, J.H., et al., Wnt signalling induces maturation of Paneth cells in intestinal crypts. Nat Cell Biol, 2005. 7(4): p. 381-6. 17. Vooijs, M., Z. Liu, and R. Kopan, Notch: Architect, Landscaper, and Guardian of the Intestine. Gastroenterology, 2011. 141(2): p. 448-459. 18. Milano, J., et al., Modulation of notch processing by gamma-secretase inhibitors causes intestinal goblet cell metaplasia and induction of genes known to specify gut secretory lineage differentiation. Toxicol Sci, 2004. 82(1): p. 341-58.

19. van Es, J.H., et al., Notch/gamma-secretase inhibition turns proliferative cells in intestinal crypts and adenomas into goblet cells. Nature, 2005. 435(7044): p. 959- 63. 20. Yang, Q., et al., Requirement of Math1 for secretory cell lineage commitment in the mouse intestine. Science, 2001. 294(5549): p. 2155-8. 21. Jenny, M., et al., Neurogenin3 is differentially required for endocrine cell fate specification in the intestinal and gastric epithelium. EMBO J, 2002. 21(23): p. 6338-47. 22. Ng, A.Y., et al., Inactivation of the transcription factor Elf3 in mice results in dysmorphogenesis and altered differentiation of intestinal epithelium. Gastroenterology, 2002. 122(5): p. 1455-66. 23. Camilleri, M., et al., Intestinal barrier function in health and gastrointestinal disease. Neurogastroenterol Motil, 2012. 24(6): p. 503-12. 24. Drossman, D.A., Functional Gastrointestinal Disorders: History, Pathophysiology, Clinical Features, and Rome IV. Gastroenterology, 2016. 150(6): p. 1262-1279.e2. 25. Camilleri, M., et al., Pharmacologic, Pharmacokinetic, and Pharmacogenomic Aspects of Functional Gastrointestinal Disorders. Gastroenterology. 150(6): p. 1319-1331.e20. 26. Pinto Sanchez, M.I. and P. Bercik, Epidemiology and burden of chronic constipation. Canadian Journal of Gastroenterology, 2011. 25(Suppl B): p. 11B- 15B. 27. Canavan, C., J. West, and T. Card, The epidemiology of irritable bowel syndrome. Clinical Epidemiology, 2014. 6: p. 71-80. 28. Ibeakanma, C., et al., Brain–Gut Interactions Increase Peripheral Nociceptive Signaling in Mice With Postinfectious Irritable Bowel Syndrome. Gastroenterology, 2011. 141(6): p. 2098-2108.e5. 29. Kellow, J.E., et al., Dysmotility of the small intestine in irritable bowel syndrome. Gut, 1988. 29(9): p. 1236-43. 30. Farzaei, M.H., et al., The Role of Visceral Hypersensitivity in Irritable Bowel Syndrome: Pharmacological Targets and Novel Treatments. J Neurogastroenterol Motil, 2016. 22(4): p. 558-574. 31. Spiller, R. and K. Garsed, Postinfectious irritable bowel syndrome. Gastroenterology, 2009. 136(6): p. 1979-88. 32. Kassinen, A., et al., The fecal microbiota of irritable bowel syndrome patients differs significantly from that of healthy subjects. Gastroenterology, 2007. 133(1): p. 24-33. 33. Henström, M. and M. D’Amato, Genetics of irritable bowel syndrome. Molecular and Cellular Pediatrics, 2016. 3: p. 7. 34. Coates, M.D., et al., Molecular defects in mucosal serotonin content and decreased serotonin reuptake transporter in ulcerative colitis and irritable bowel syndrome. Gastroenterology, 2004. 126(7): p. 1657-64. 35. Costedio, M.M., et al., Mucosal serotonin signaling is altered in chronic constipation but not in opiate-induced constipation. Am J Gastroenterol, 2010. 105(5): p. 1173-80.

36. Bearcroft, C.P., D. Perrett, and M.J. Farthing, Postprandial plasma 5- hydroxytryptamine in diarrhoea predominant irritable bowel syndrome: a pilot study. Gut, 1998. 42(1): p. 42-6. 37. Hoffman, J.M., et al., Activation of colonic mucosal 5-HT(4) receptors accelerates propulsive motility and inhibits visceral hypersensitivity. Gastroenterology, 2012. 142(4): p. 844-854 e4. 38. Camilleri , M., et al., A Placebo-Controlled Trial of Prucalopride for Severe Chronic Constipation. New England Journal of Medicine, 2008. 358(22): p. 2344- 2354. 39. Coremans, G., et al., Prucalopride Is Effective in Patients with Severe Chronic Constipation in Whom Laxatives Fail to Provide Adequate Relief. Digestion, 2003. 67(1-2): p. 82-89. 40. Cremonini, F., S. Delgado-Aros, and M. Camilleri, Efficacy of alosetron in irritable bowel syndrome: a meta-analysis of randomized controlled trials. Neurogastroenterol Motil, 2003. 15(1): p. 79-86. 41. Halland, M. and Y.A. Saito, Irritable bowel syndrome: new and emerging treatments. BMJ, 2015. 350: p. h1622. 42. Chang, L., A. Lembo, and S. Sultan, American Gastroenterological Association Institute Technical Review on the pharmacological management of irritable bowel syndrome. Gastroenterology, 2014. 147(5): p. 1149-72 e2. 43. Eswaran, S., J. Muir, and W.D. Chey, Fiber and functional gastrointestinal disorders. Am J Gastroenterol, 2013. 108(5): p. 718-27. 44. Ford, A.C. and N.C. Suares, Effect of laxatives and pharmacological therapies in chronic idiopathic constipation: systematic review and meta-analysis. Gut, 2011. 60(2): p. 209-18. 45. DiPalma, J.A., et al., A randomized, placebo-controlled, multicenter study of the safety and efficacy of a new polyethylene glycol laxative. Am J Gastroenterol, 2000. 95(2): p. 446-50. 46. Pare, P. and R.N. Fedorak, Systematic review of stimulant and nonstimulant laxatives for the treatment of functional constipation. Can J Gastroenterol Hepatol, 2014. 28(10): p. 549-57. 47. Thomas, R.H. and D.R. Luthin, Current and emerging treatments for irritable bowel syndrome with constipation and chronic idiopathic constipation: focus on prosecretory agents. Pharmacotherapy, 2015. 35(6): p. 613-30. 48. Rivkin, A. and L. Chagan, Lubiprostone: chloride channel activator for chronic constipation. Clin Ther, 2006. 28(12): p. 2008-21. 49. Barish, C.F., et al., Efficacy and safety of lubiprostone in patients with chronic constipation. Dig Dis Sci, 2010. 55(4): p. 1090-7. 50. Fukudo, S., et al., Efficacy and safety of oral lubiprostone in constipated patients with or without irritable bowel syndrome: a randomized, placebo-controlled and dose-finding study. Neurogastroenterol Motil, 2011. 23(6): p. 544-e205. 51. Johanson, J.F. and R. Ueno, Lubiprostone, a locally acting chloride channel activator, in adult patients with chronic constipation: a double-blind, placebo- controlled, dose-ranging study to evaluate efficacy and safety. Aliment Pharmacol Ther, 2007. 25(11): p. 1351-61.

52. Siegel, R.L., et al., Colorectal cancer statistics, 2017. CA: A Cancer Journal for Clinicians, 2017. 67(3): p. 177-193. 53. Fearon, E.R. and B. Vogelstein, A genetic model for colorectal tumorigenesis. Cell, 1990. 61(5): p. 759-67. 54. Hanahan, D. and Robert A. Weinberg, Hallmarks of Cancer: The Next Generation. Cell, 2011. 144(5): p. 646-674. 55. Hanahan, D. and R.A. Weinberg, The hallmarks of cancer. Cell, 2000. 100(1): p. 57-70. 56. de la Chapelle, A., Genetic predisposition to colorectal cancer. Nat Rev Cancer, 2004. 4(10): p. 769-780. 57. van Es, J.H., R.H. Giles, and H.C. Clevers, The many faces of the tumor suppressor gene APC. Exp Cell Res, 2001. 264(1): p. 126-34. 58. Galiatsatos, P. and W.D. Foulkes, Familial Adenomatous Polyposis. American Journal of Gastroenterology, 2006. 101(2): p. 385-398. 59. Lynch, H.T., et al., Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clinical genetics, 2009. 76(1): p. 1-18. 60. O’Connell, J.B., M.A. Maggard, and C.Y. Ko, Colon Cancer Survival Rates With the New American Joint Committee on Cancer Sixth Edition Staging. JNCI: Journal of the National Cancer Institute, 2004. 96(19): p. 1420-1425. 61. Bibbins-Domingo, K. and U. S. Preventive Services Task Force, Aspirin use for the primary prevention of cardiovascular disease and colorectal cancer: U.s. preventive services task force recommendation statement. Annals of Internal Medicine, 2016. 164(12): p. 836-845. 62. Brenna, O., et al., Cellular localization of guanylin and uroguanylin mRNAs in human and rat duodenal and colonic mucosa. Cell Tissue Res, 2016. 365(2): p. 331-41. 63. Ikpa, P.T., et al., Guanylin and uroguanylin are produced by mouse intestinal epithelial cells of columnar and secretory lineage. Histochem Cell Biol, 2016. 146(4): p. 445-55. 64. Forte, L.R., Jr., Uroguanylin and guanylin peptides: pharmacology and experimental therapeutics. Pharmacol Ther, 2004. 104(2): p. 137-62. 65. Steinbrecher, K.A. and M.B. Cohen, Transmembrane guanylate cyclase in intestinal pathophysiology. Curr Opin Gastroenterol, 2011. 27(2): p. 139-45. 66. Arshad, N. and S.S. Visweswariah, The multiple and enigmatic roles of guanylyl cyclase C in intestinal homeostasis. FEBS Lett, 2012. 586(18): p. 2835-40. 67. Li, P., et al., Homeostatic control of the crypt-villus axis by the bacterial enterotoxin receptor guanylyl cyclase C restricts the proliferating compartment in intestine. Am J Pathol, 2007. 171(6): p. 1847-58. 68. Li, P., et al., Guanylyl cyclase C suppresses intestinal tumorigenesis by restricting proliferation and maintaining genomic integrity. Gastroenterology, 2007. 133(2): p. 599-607. 69. Wang, R., et al., Type 2 cGMP-dependent protein kinase regulates homeostasis by blocking c-Jun N-terminal kinase in the colon epithelium. Cell Death Differ, 2014. 21(3): p. 427-37.

70. Wang, R., et al., Type 2 cGMP-dependent protein kinase regulates proliferation and differentiation in the colonic mucosa. Am J Physiol Gastrointest Liver Physiol, 2012. 303(2): p. G209-19. 71. Lin, J.E., et al., GUCY2C opposes systemic genotoxic tumorigenesis by regulating AKT-dependent intestinal barrier integrity. PLoS One, 2012. 7(2): p. e31686. 72. Steinbrecher, K.A., et al., Murine guanylate cyclase C regulates colonic injury and inflammation. J Immunol, 2011. 186(12): p. 7205-14. 73. Thompson, W.J., et al., Exisulind induction of apoptosis involves guanosine 3',5'- cyclic monophosphate phosphodiesterase inhibition, protein kinase G activation, and attenuated beta-catenin. Cancer Res, 2000. 60(13): p. 3338-42. 74. Whitt, J.D., et al., A novel sulindac derivative that potently suppresses colon tumor cell growth by inhibiting cGMP phosphodiesterase and beta-catenin transcriptional activity. Cancer Prev Res (Phila), 2012. 5(6): p. 822-33. 75. Chang, W.C., et al., Sulindac sulfone is most effective in modulating beta-catenin- mediated transcription in cells with mutant APC. Ann N Y Acad Sci, 2005. 1059: p. 41-55. 76. Richter, M., et al., Growth inhibition and induction of apoptosis in colorectal tumor cells by cyclooxygenase inhibitors. Carcinogenesis, 2001. 22(1): p. 17-25. 77. Giardiello, F.M., et al., Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med, 1993. 328(18): p. 1313-6. 78. Arber, N., et al., Sporadic adenomatous polyp regression with exisulind is effective but toxic: a randomised, double blind, placebo controlled, dose-response study. Gut, 2006. 55(3): p. 367-73. 79. Steinbrecher, K.A., et al., Targeted inactivation of the mouse guanylin gene results in altered dynamics of colonic epithelial proliferation. Am J Pathol, 2002. 161(6): p. 2169-78. 80. Wilson, C., et al., The paracrine hormone for the GUCY2C tumor suppressor, guanylin, is universally lost in colorectal cancer. Cancer Epidemiol Biomarkers Prev, 2014. 23(11): p. 2328-37. 81. Steinbrecher, K.A., et al., Expression of guanylin is downregulated in mouse and human intestinal adenomas. Biochem Biophys Res Commun, 2000. 273(1): p. 225-30. 82. Shailubhai, K., et al., Uroguanylin treatment suppresses polyp formation in the Apc(Min/+) mouse and induces apoptosis in human colon adenocarcinoma cells via cyclic GMP. Cancer Res, 2000. 60(18): p. 5151-7. 83. Chang, W.L., et al., Plecanatide-mediated activation of guanylate cyclase-C suppresses inflammation-induced colorectal carcinogenesis in Apc+/Min-FCCC mice. World J Gastrointest Pharmacol Ther, 2017. 8(1): p. 47-59. 84. Islam, B.N., et al., Sildenafil Suppresses Inflammation-Driven Colorectal Cancer in Mice. Cancer Prev Res (Phila), 2017. 10(7): p. 377-388. 85. Bryant, A.P., et al., Linaclotide is a potent and selective guanylate cyclase C agonist that elicits pharmacological effects locally in the gastrointestinal tract. Life Sci, 2010. 86(19-20): p. 760-5. 86. Busby, R.W., et al., Linaclotide, through activation of guanylate cyclase C, acts locally in the gastrointestinal tract to elicit enhanced intestinal secretion and transit. Eur J Pharmacol, 2010. 649(1-3): p. 328-35.

100

87. Field, M., et al., Heat-stable enterotoxin of Escherichia coli: in vitro effects on guanylate cyclase activity, cyclic GMP concentration, and ion transport in small intestine. Proc Natl Acad Sci U S A, 1978. 75(6): p. 2800-4. 88. Johnston, J.M., et al., Linaclotide improves abdominal pain and bowel habits in a phase IIb study of patients with irritable bowel syndrome with constipation. Gastroenterology, 2010. 139(6): p. 1877-1886 e2. 89. Eutamene, H., et al., Guanylate cyclase C-mediated antinociceptive effects of linaclotide in rodent models of visceral pain. Neurogastroenterol Motil, 2010. 22(3): p. 312-e84. 90. Lacy, B.E., et al., Linaclotide in Chronic Idiopathic Constipation Patients with Moderate to Severe Abdominal Bloating: A Randomized, Controlled Trial. PLoS One, 2015. 10(7): p. e0134349. 91. Rao, S., et al., A 12-week, randomized, controlled trial with a 4-week randomized withdrawal period to evaluate the efficacy and safety of linaclotide in irritable bowel syndrome with constipation. Am J Gastroenterol, 2012. 107(11): p. 1714- 24; quiz p 1725. 92. Rao, S.S., et al., Effect of linaclotide on severe abdominal symptoms in patients with irritable bowel syndrome with constipation. Clin Gastroenterol Hepatol, 2014. 12(4): p. 616-23. 93. Ironwood Pharmaceuticals. Linzess (R) [prescribing information]. 2016. 94. Hamra, F.K., et al., Regulation of intestinal uroguanylin/guanylin receptor- mediated responses by mucosal acidity. Proc Natl Acad Sci U S A, 1997. 94(6): p. 2705-10. 95. Shailubhai, K., et al., Plecanatide, an oral guanylate cyclase C agonist acting locally in the gastrointestinal tract, is safe and well-tolerated in single doses. Dig Dis Sci, 2013. 58(9): p. 2580-6. 96. Miner, P.B., Jr., et al., A Randomized Phase III Clinical Trial of Plecanatide, a Uroguanylin Analog, in Patients With Chronic Idiopathic Constipation. Am J Gastroenterol, 2017. 112(4): p. 613-621. 97. Synergy Pharmaceuticals. Trulance (R) [prescribing information]. 2017. 98. Francis, S.H., J.L. Busch, and J.D. Corbin, cGMP-Dependent Protein Kinases and cGMP Phosphodiesterases in Nitric Oxide and cGMP Action. Pharmacological Reviews, 2010. 62(3): p. 525-563. 99. Rotella, D.P., Phosphodiesterase 5 inhibitors: current status and potential applications. Nat Rev Drug Discov, 2002. 1(9): p. 674-82. 100. Kukreja, R.C., et al., Emerging new uses of phosphodiesterase-5 inhibitors in cardiovascular diseases. Experimental & Clinical Cardiology, 2011. 16(4): p. e30-e35. 101. Archer, S.L. and E.D. Michelakis, Phosphodiesterase Type 5 Inhibitors for Pulmonary Arterial Hypertension. New England Journal of Medicine, 2009. 361(19): p. 1864-1871. 102. Pol, O., E. Planas, and M.M. Puig, Peripheral effects of naloxone in mice with acute diarrhea associated with intestinal inflammation. J Pharmacol Exp Ther, 1995. 272(3): p. 1271-6. 103. Puig, M.M., O. Pol, and W. Warner, Interaction of morphine and clonidine on gastrointestinal transit in mice. Anesthesiology, 1996. 85(6): p. 1403-12.

101

104. Marona, H.R. and M.B. Lucchesi, Protocol to refine intestinal motility test in mice. Lab Anim, 2004. 38(3): p. 257-60. 105. Moser, A.R., H.C. Pitot, and W.F. Dove, A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science, 1990. 247(4940): p. 322-4. 106. Su, L.K., et al., Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene. Science, 1992. 256(5057): p. 668-70. 107. Cooper, H.S., et al., Clinicopathologic study of dextran sulfate sodium experimental murine colitis. Lab Invest, 1993. 69(2): p. 238-49. 108. Wirtz, S., et al., Chemically induced mouse models of intestinal inflammation. Nat Protoc, 2007. 2(3): p. 541-6. 109. Han, X., et al., Loss of guanylyl cyclase C (GCC) signaling leads to dysfunctional intestinal barrier. PLoS One, 2011. 6(1): p. e16139. 110. Heel, R.C., et al., Loperamide: a review of its pharmacological properties and therapeutic efficacy in diarrhoea. Drugs, 1978. 15(1): p. 33-52. 111. Lin, S., et al., Phosphodiesterase-5 inhibition suppresses colonic inflammation- induced tumorigenesis via blocking the recruitment of MDSC. Am J Cancer Res, 2017. 7(1): p. 41-52. 112. Serafini, P., et al., Phosphodiesterase-5 inhibition augments endogenous antitumor immunity by reducing myeloid-derived suppressor cell function. The Journal of Experimental Medicine, 2006. 203(12): p. 2691-2702. 113. Sharman, S.K., et al., Sildenafil normalizes bowel transit in preclinical models of constipation. PLoS One, 2017. 12(4): p. e0176673. 114. Luongo, C., et al., Loss of Apc+ in intestinal adenomas from Min mice. Cancer Res, 1994. 54(22): p. 5947-52. 115. McMurray, J.G., et al., Long-term safety and effectiveness of sildenafil citrate in men with erectile dysfunction. Therapeutics and Clinical Risk Management, 2007. 3(6): p. 975-981. 116. Oelke, M., et al., Efficacy and safety of tadalafil 5 mg once daily in the treatment of lower urinary tract symptoms associated with benign prostatic hyperplasia in men aged >/=75 years: integrated analyses of pooled data from multinational, randomized, placebo-controlled clinical studies. BJU Int, 2017. 119(5): p. 793- 803. 117. Rubin, L.J., et al., Long-term Treatment With Sildenafil Citrate in Pulmonary Arterial Hypertension: The SUPER-2 Study. Chest, 2011. 140(5): p. 1274-1283. 118. Kwon, I.K., et al., PKG inhibits TCF signaling in colon cancer cells by blocking beta-catenin expression and activating FOXO4. Oncogene, 2010. 29(23): p. 3423-34. 119. Deguchi, A., W.J. Thompson, and I.B. Weinstein, Activation of protein kinase G is sufficient to induce apoptosis and inhibit cell migration in colon cancer cells. Cancer Res, 2004. 64(11): p. 3966-73. 120. Lee, K., et al., beta-catenin nuclear translocation in colorectal cancer cells is suppressed by PDE10A inhibition, cGMP elevation, and activation of PKG. Oncotarget, 2016. 7(5): p. 5353-65. 121. Li, N., et al., Phosphodiesterase 10A: a novel target for selective inhibition of colon tumor cell growth and beta-catenin-dependent TCF transcriptional activity. Oncogene, 2015. 34(12): p. 1499-509.

102

122. Grover, M., et al., On the fiftieth anniversary. Postinfectious irritable bowel syndrome: mechanisms related to pathogens. Neurogastroenterol Motil, 2014. 26(2): p. 156-67. 123. Qin, H.Y., et al., Systematic review of animal models of post-infectious/post- inflammatory irritable bowel syndrome. J Gastroenterol, 2011. 46(2): p. 164-74. 124. Eijkelkamp, N., et al., Increased visceral sensitivity to capsaicin after DSS- induced colitis in mice: spinal cord c-Fos expression and behavior. Am J Physiol Gastrointest Liver Physiol, 2007. 293(4): p. G749-57. 125. Hughes, P.A., et al., Post-inflammatory colonic afferent sensitisation: different subtypes, different pathways and different time courses. Gut, 2009. 58(10): p. 1333-41. 126. Zhou, Q., et al., Visceral and somatic hypersensitivity in a subset of rats following TNBS-induced colitis. Pain, 2008. 134(1-2): p. 9-15. 127. Bortolotti, M., et al., Effects of sildenafil on esophageal motility of patients with idiopathic achalasia. Gastroenterology, 2000. 118(2): p. 253-7. 128. Eherer, A.J., et al., Effect of sildenafil on oesophageal motor function in healthy subjects and patients with oesophageal motor disorders. Gut, 2002. 50(6): p. 758- 64. 129. Fox, M., et al., Sildenafil relieves symptoms and normalizes motility in patients with oesophageal spasm: a report of two cases. Neurogastroenterol Motil, 2007. 19(10): p. 798-803. 130. Milone, M. and J.K. DiBaise, A pilot study of the effects of sildenafil on stool characteristics, colon transit, anal sphincter function, and rectal sensation in healthy men. Dig Dis Sci, 2005. 50(6): p. 1005-11. 131. Xu, X. and J.D. Chen, Inhibitory effects of sildenafil on small intestinal motility and myoelectrical activity in dogs. Dig Dis Sci, 2006. 51(4): p. 671-6. 132. Atkinson, W., et al., Altered 5-hydroxytryptamine signaling in patients with constipation- and diarrhea-predominant irritable bowel syndrome. Gastroenterology, 2006. 130(1): p. 34-43. 133. Dunlop, S.P., et al., Abnormal intestinal permeability in subgroups of diarrhea- predominant irritable bowel syndromes. Am J Gastroenterol, 2006. 101(6): p. 1288-94. 134. Silos-Santiago, I., et al., Gastrointestinal pain: unraveling a novel endogenous pathway through uroguanylin/guanylate cyclase-C/cGMP activation. Pain, 2013. 154(9): p. 1820-30. 135. Sager, G., Cyclic GMP transporters. Neurochemistry International, 2004. 45(6): p. 865-873. 136. O'Sullivan, M., et al., Increased mast cells in the irritable bowel syndrome. Neurogastroenterol Motil, 2000. 12(5): p. 449-57. 137. Chey, W.D., et al., Linaclotide for irritable bowel syndrome with constipation: a 26-week, randomized, double-blind, placebo-controlled trial to evaluate efficacy and safety. Am J Gastroenterol, 2012. 107(11): p. 1702-12. 138. Lembo, A.J., et al., Two randomized trials of linaclotide for chronic constipation. N Engl J Med, 2011. 365(6): p. 527-36. 139. Basu, N., et al., Intestinal cell proliferation and senescence are regulated by receptor guanylyl cyclase C and p21. J Biol Chem, 2014. 289(1): p. 581-93.

103

140. Pitari, G.M., et al., Guanylyl cyclase C agonists regulate progression through the cell cycle of human colon carcinoma cells. Proc Natl Acad Sci U S A, 2001. 98(14): p. 7846-51. 141. Birbe, R., et al., Guanylyl cyclase C is a marker of intestinal metaplasia, dysplasia, and adenocarcinoma of the gastrointestinal tract. Hum Pathol, 2005. 36(2): p. 170-9. 142. Lin, J.E., et al., Guanylyl cyclase C in colorectal cancer: susceptibility gene and potential therapeutic target. Future Oncol, 2009. 5(4): p. 509-22. 143. Li, P., et al., Can colorectal cancer be prevented or treated by oral hormone replacement therapy? Curr Mol Pharmacol, 2009. 2(3): p. 285-92. 144. Li, P., et al., Colorectal cancer is a paracrine deficiency syndrome amenable to oral hormone replacement therapy. Clin Transl Sci, 2008. 1(2): p. 163-167. 145. Robertis, M.D., et al., The AOM/DSS murine model for the study of colon carcinogenesis: From pathways to diagnosis and therapy studies. Journal of Carcinogenesis, 2011. 10: p. 9. 146. Blomain, E.S., A.M. Pattison, and S.A. Waldman, GUCY2C ligand replacement to prevent colorectal cancer. Cancer Biol Ther, 2016. 17(7): p. 713-8.

104

Appendix A: List of commonly used abbreviations

5-HT: 5-hydroxytryptophan (serotonin) Apc: adenomatous polyposis coli CC3: cleaved caspase 3 CFTR: cystic fibrosis transmembrane conductance regulator cGMP: cyclic guanosine monophosphate DSS: dextran sulfate sodium GC-C: guanylate cyclase C CIC: chronic idiopathic constipation CRC: colorectal cancer EC cell: enterochromaffin cell EE cell: enteroendocrine cell FAP: familial adenomatous polyposis Gn: guanylin IBS: irritable bowel syndrome IHC: immunohistochemistry NSAID: non-steroidal anti-inflammatory drug PCR: polymerase chain reaction PDE: phosphodiesterase PKG: cGMP-dependent protein kinase SEM: standard error of the means Sild: sildenafil Ugn: uroguanylin

105