Can colorectal cancer be prevented or treated by oral hormone replacement therapy?

Abstract

Guanylyl cyclase C (GCC) is the receptor specifically expressed by intestinal cells for the paracrine hormones guanylin and uroguanylin and diarrheagenic bacterial heat-stable enterotoxins. This tissue-specific receptor coordinates lineage-dependent regulation of epithelial homeostasis, and its disruption contributes to intestinal tumorigenesis. It coordinates regenerative and metabolic circuits by restricting the cell cycle and proliferation and programming metabolic transitions central to organizing the dynamic crypt-surface axis. Further, mice deficient in GCC signaling are more susceptible to colon cancer induced by Apc mutations or the carcinogen azoxymethane. Moreover, guanylin and uroguanylin are gene products most commonly lost, early, in colon cancer in animals and humans. The role of GCC as a tumor suppressing receptor regulating proliferation and metabolism, together with the universal loss of guanylin and uroguanylin in tumorigenesis, suggests a model in which colorectal cancer is a paracrine hormone deficiency syndrome. In that context, activation of GCC reverses the tumorigenic phenotype by limiting growth of colorectal cancer cells by restricting progression through the G1/S transition and reprogramming metabolic circuits from glycolysis to oxidative phosphorylation, limiting bioenergetic support for rapid proliferation. These observations suggest a pathophysiological hypothesis in which GCC is a lineage-dependent tumor suppressing receptor coordinating proliferative homeostasis whose dysregulation through hormone loss contributes to neoplasia. The correlative therapeutic hypothesis suggests that colorectal cancer is a disease of hormone insufficiency that can be prevented or treated by oral supplementation with GCC ligands.

Colorectal cancer is the 4th most common neoplasm and the 2nd leading cause of cancer-related mortality, producing 10% of cancer-related deaths, in the U.S.[1–3] While surgery is the mainstay of treatment, occult metastases produce relapse in ~50% of patients[2, 4, 5] and adjuvant chemotherapy only marginally improves survival.[4, 6] The poor prognosis and paucity of effective therapy underscore the clinical need for new approaches for earlier prevention and therapy. Conventionally, colon cancer is considered a genetic disease, reflecting sequential accumulation of mutations in oncogenes, tumor susceptibility genes or tumor suppressors,[7] most frequently (>80%) including sporadic or inherited alterations in the gene encoding adenomatous polyposis coli (APC).[8, 9] In that context, a novel paradigm is emerging which suggests that at initiation, intestinal neoplasia evolves from a state of paracrine hormone insufficiency,[10–12] providing a novel therapeutic opportunity for colorectal cancer prevention by ligand supplementation to reconstitute disrupted tumor-suppressing signaling pathways.[13, 14]

Intestinal homeostasis and colorectal cancer

The intestinal surface is covered by a single layer of epithelial cells organized in vertical anatomical units underlying specialized organ functions. This epithelium undergoes continuous homeostatic cycles of proliferation, migration, differentiation, apoptosis and shedding absolutely required for the maintenance of digestion, absorption, secretion, and barrier function (Fig. 1A).[13–18] Transit cells (Fig. 1E) originate from slowly proliferating stem cells near the crypt base and undergo cycles of division. Both rapid proliferating transit cells and slowly replicating stem cells form the proliferating zone, crypts in small intestine and the bottom of crypts in large intestine, providing cells for continuous renewal. Proliferating transit cells migrate along the vertical axis to the differentiated compartment where homo- and heterotypic interactions reprogram nuclear and cytoplasmic circuits, coordinating proliferative restriction, lineage commitment, and metabolic programming (Fig.1B–D, F). Multiple signaling pathways such as NOTCH 1, MATH 1 or HES 1 regulate lineage commitment to program the transition from proliferation to terminal differentiation.[19–22] Enterocytes, comprising the principle lineage of intestinal epithelial cells, develop well-organized microvillus brush borders (Fig. 1B), containing key functional proteins mediating cognate digestive and absorptive functions. Goblet cells secrete mucin, establishing the mucus layer (Fig. 1C) protecting intestinal surfaces and facilitating enterocytes nutrient digestion and absorption.[23] Enteroendocrine cells (Fig. 1D) belong to the neuro-endocrine system and secrete paracrine and autocrine hormones supporting intestinal neuromuscular function, secretion, and central regulation of food intake.[23] Paneth cells, (Fig. 1F) located only in small intestine, secrete antimicrobial peptides and growth factors into the lumen,[24] forming a physical and functional barrier defending against bacterial invasion and intestinal tumorigenesis through innate immune responses.[24] Consistent with cell cycle exit, mature cells also reprogram their metabolic requirements from glycolysis to mitochondria-dependent oxidation-phosphorylation. The established gradient of metabolism, with glycolysis prevailing in crypts and mitochondrial metabolism predominating in villi, parallels the proliferative gradient along that vertical axis. Metabolic programming reflects the energy demands for different compartments, subserving their balance of proliferation and differentiation. Thus, in the crypt, cell division requires rapidly available energy supplies from glycolysis, while in the differentiated compartment, mitochondria-dependent metabolism exploits the efficiency of ATP production by oxidative phosphorylation, supporting catabolic demands in mature cells. [25], ^a

Figure 1. Intestinal crypt to villus axis.

Intestinal epithelium, as a rapidly regenerating tissue, undergoes highly organized cycles of proliferation, migration, differentiation and apoptosis that maintain the structural and functional integrity of the crypt-surface axis. (A) Stem cells and rapidly proliferating transit cells are located in the proliferating crypt compartment. Cells migrate up to the villus, the differentiated compartment, or down to the bottom of the crypt to become terminally differentiated cells. Transit cells (Ki67 IHC staining, E) give rise to four principal mature cell lines characterizing the crypt-villus axis and supporting their digestive, absorptive and defense functions: absorptive enterocytes (villin IHC staining, B), goblet cells (alcian blue staining, C), enteroendocrine cells (chromatin A IHC staining, D), and Paneth cells (lysosome IHC staining, F). GCC promotes intestinal crypt-surface homeostasis by restricting the proliferative compartment and promoting cell differentiation.

Disruption of homeostatic renewal with the corruption of proliferative and metabolic circuits, especially the transition from proliferation to differentiation, produces crypt hyperplasia and a maturational shift, as well as a “Warburg”-like metabolic phenotype, contributing to the development of colon cancer.[15–18] Intestinal epithelial cell renewal is highly regulated and restricted by multiple antiproliferative mechanisms. Disruption of these mechanisms results in continuous cycling of DNA replication and proliferation, resulting in hyperplasia in crypts and accumulation of mutations which potentiate dysregulated division, failure to terminally differentiate, and failure to undergo programmed cell death, which ultimately establishes the invasive carcinoma phenotype. Cell division is predominantly dependent on controlling entry and progression through the G1 checkpoint. Accelerated progression through G₁ and premature entry into S is associated with amplification of genetic instability.[26–29] Furthermore, transformed cells fail to commit to lineage specification. Rather, they undergo metabolic reprogramming, enhancing glycolysis to meet energy demands associated with rapid proliferation. This hyperplasia generates a microenvironment for amplifying mutations potentiating transformation, resulting in colorectal cancer initiation, progression and ultimately malignant invasion and metastasis. [30–32]

Guanylyl cyclase C signaling in human colorectal cancer

Guanylyl cyclase C (GCC) is the only identified receptor, specifically expressed in the apical membrane of intestinal epithelial cells and tumors originated from colons (Fig. 2A, B),^[33–38] for diarrheagenic bacterial heat-stable enterotoxins (STs) and the endogenous hormones guanylin and uroguanylin (Fig. 2C). STs are structurally and functionally homologous to the endogenous ligands of GCC, with 10-fold greater potency than guanylin or uroguanylin.[39, 40] Interaction of GCC and its ligands activates the intracellular domain and converts GTP to cyclic GMP (cGMP; Fig. 2D). This cyclic nucleotide second messenger conveys signaling information from GCC to its downstream effectors[40], including cGMP-dependent protein kinase (PKG) which phosphorylates the cystic fibrosis transmembrane conductance regulator, resulting in efflux of salt and water and, in the case of ST, diarrhea.[41] Heretofore, GCC and its ligands were considered regulators of intestinal fluid and electrolyte homeostasis.[42] In humans, guanylin and uroguanylin are organized in a tail-to-tail configuration on chromosome 1p.[43, 44] Sparse patient sampling suggested that these hormones are the most commonly lost gene products in colorectal cancer in animals and humans,[45–49] and their loss early in transformation, with hormone deficiency detected in dysplastic crypts,[46, 48–51] may reflect chromosomal instability[43, 44] or epigenetic silencing.[52] More than 500 normal adjacent tissue (NAT) and tumors from more than 300 colon cancer patients, including 125 matched NAT and tumor specimens have been analyzed recently (unpublished data). Guanylin (~20-fold) and uroguanylin (~20-fold) expression was reduced in tumors, compared to NAT (p<0.0001), in 81% and 76% of patients, respectively, a frequency comparable to APC mutation rates in colon cancer.[53, 54] Most patients exhibited reductions in both hormones, eliminating compensatory reciprocal over-expression. Thus, GCC hormone loss generally is associated with sporadic colorectal cancer. Recent analyses of matched tumor and normal adjacent tissues revealed maladaptive epigenetic imprinting attenuating GCC hormone expression in tumorigenesis and the detailed mechanisms remain unclear. Studies are directed at defining the etiology of this acquired epigenesis as a key contributor to intestinal neoplasia.

Figure 2. GCC is restricted to the apical membrane of intestinal epithelial cells and its expression is retained following neoplastic transformation.

GCC is expressed specifically in apical membranes of intestinal epithelial cells (A), and human colorectal cancer cells (B) revealed by GCC immunohistochemical staining. The exogenous ligand (STs) or the endogenous ligands (guanylin and uroguanylin) (C) bind to the extracellular domain, activating the intracellular catalytic domain which converts GTP to cyclic GMP (D) and the cyclic nucleotide, as the second messenger, activates downstream effectors.

Of significance, GCC is selectively expressed in apical membranes of intestinal epithelial cells from the duodenum to the rectumbut not in tissues or tumors originating outside the gastrointestinal tract.[34, 55] Further, unlike its ligands, this receptor continues to be expressed after normal epithelial cells undergo neoplastic transformation and become colorectal cancer cells (Fig. 2B).[34, 55–57] Moreover, GCC expression has been identified in every primary and metastatic colorectal tumor obtained from patients regardless of tumor grade or anatomic location.[34, 55–57] Indeed, GCC is over-expressed in colon cancers [56, 57] and its metastasis compared to normal adjacent tissues (Fig. 2B), but not in tumors originating from other tissues.[34] Although some studies suggest that, in addition to intestine, GCC is expressed in epithelia from gut accessory glands including gallbladder,[58] pancreas,[59, 60] and salivary glands;[61, 62] respiratory epithelia;[63] and reproductive epithelium[64] in human and rats, these results may reflect the relatively low stringency of primers and non-specific cross-reactivity of antibodies, since functional GCC has not been detected in tissues other than intestine[34, 36]. Tissue-restricted expression of GCC in intestine and derivative tumors fulfills the criteria for a specific therapeutic target for colon cancer prevention and therapy.[56, 57] The novel paradigm in which colon cancer initiates as a hormone deficiency syndrome suggests a new cancer prevention strategy in the form of oral supplementation of ligands to the specific target, GCC.

GCC signaling maintains intestinal homeostasis opposing tumorigenesis

Recent studies suggest that GCC signaling restricts proliferative dynamics through reciprocal coordination of cell cycle and metabolic programs, and disruption of GCC signaling contributes to intestinal tumorigenesis.[12, 65], ^a Eliminating GCC (GCC^−/−)[65], ^a or guanylin[66] in mice expands the number of crypts along the rostral-caudal axis, with a higher proliferative index, quantified by immunohistochemistry (IHC) for PCNA and Ki67, associated with an accelerated cell cycle, quantified by 5′-bromo-2′-deoxyuridine (BrdU) kinetics.[65] Hyperproliferation in the absence of GCC signaling is associated with increased expression of cell cycle mediators, including cyclin D and pRb, which regulate transit through the G1-S transition; and decreased expression of suppressors, including p27.[12, 65], ^a Moreover, expansion of the proliferating compartment is coupled with reciprocal bioenergetic remodeling and the induction of a glycolytic metabolic phenotype across the entire crypt-surface axis, required for energy independence and anabolic biosynthesis supporting hyperproliferation.[67, 68] Indeed, rate-limiting enzymes are increased for glucose transport, including the glucose transporter 1; glycolysis, including hexokinase II, phosphofructokinase I, the regulatory enzyme phosphofructokinase 2P, pyruvate kinase and lactate dehydrogenase; and fatty acid synthesis, including phosphorylated ATP citrate lyase and acetylCoA carboxylase.^a Glycolytic reprogramming increases glucose uptake and aerobic glycolysis reflected by lactate accumulation.^a Conversely, mitochondrial biogenesis and function are inhibited in GCC-deficient mice, which exhibit a reduction in mitochondrial proteins including cytochrome oxidase, core II, and ATP synthase; loss of mitochondrial DNA; and diminished mitochondria-specific oxygen consumption, dehydrogenase activity and ATP concentrations. ^a Of significance, the gradient of mitochondria along the crypt to villus axis reflecting the transition from proliferation to differentiation is disrupted in GCC^−/−mice. ^a These data suggest a role for GCC in coordinating cell cycle and metabolic circuits integral to homeostatic programs organizing the crypt-surface axis. Normally, GCC and its ligands induce cGMP production in an ascending gradient along the crypt-surface axis (Fig. 3), lowest in the proliferating compartment and greatest in the differentiated compartment,[40, 42] restricting proliferation and reprogramming metabolism in crypts. Disruption of those programs, reflecting loss of paracrine hormone expression, induces hyperproliferation characterized by cell cycle acceleration and glycolytic remodeling that universally characterize tumorigenisis.[67, 68]

Figure 3. GCC regulates crypt-villus homeostasis whose disruption contributes to intestinal tumorigenesis.

In contrast to GCC, whose expression is uniform along the crypt to villus axis, the endogenous ligands guanylin and uroguanylin are expressed and secreted along an ascending gradient, greatest in the differentiated villus compartment and lowest in the proliferating crypt compartment. GCC signaling opposes the proliferative gradient along that vertical axis by restricting cell growth in the crypts and promoting maturation of differentiated cells by reprogramming cell metabolism from glycolysis to mitochondria-dependent oxidative phosphorylation (OxPhos). Disruption of GCC signaling reflecting hormone loss corrupts homeostatic proliferative and metabolic circuits, promoting neoplastic transformation.

GCC^−/− mice are more sensitive to Apc^Min/+ and azoxymethane (AOM)-induced carcinogenesis, with increased tumor incidence, multiplicity, and burden, compared to wild-type mice.[12] GCC-dependent tumorigenesis is driven by dysregulated proliferation, quantified by Ki67 IHC staining, and increases in cell cycle mediators.[12] Of significance, human colon cancers exhibit alterations in cell cycle mediators coordinated with glycolytic reprogramming identical to changes induced by disruption of GCC in mice. ^a Hyperproliferation together with glycolytic reprogramming in GCC^−/− mice, mutually reinforce accumulation of DNA damage, including DNA oxidation, ^a producing chromosomal instability leading to malignant transformation.[12], ^a These observations suggest that GCC is a tumor susceptibility gene product that regulates proliferative and metabolic homeostasis in intestine. In the context of the loss of GCC hormones during tumorigenesis, these data provide a mechanistic basis for hormone replacement as a strategy for preventing and treating colorectal cancer.[14], ^b

Taken together, these data suggest that GCC signaling suppresses intestinal tumorigenesis by coordinating cell proliferation and metabolic programming. In that context, disruption of GCC signaling reflecting early loss of guanylin and uroguanylin expression might be a key initiating event in human colorectal neoplastic transformation by releasing cell cycle and metabolic restriction, thus compromising genetic integrity, producing downstream accumulation of mutations in tumor suppressors or oncogenes promoting tumor progression.[12], ^a

Targeted prevention and therapy for colorectal cancer

The contributions of GCC to intestinal proliferative and metabolic homeostasis and intestinal tumorigenesis, together with loss of ligands but retention of receptor expression in human colorectal carcinogenesis, underscore a novel therapeutic paradigm for targeted colon cancer prevention and treatment through oral supplementation of lost hormones. This presumes that in the pathophysiological state, GCC is a dormant tumor-suppressing receptor whose reactivation by ligand coordinately rescues cell cycle restriction and reprograms glycolytic metabolism to limit proliferation.

Indeed, GCC signaling inhibits proliferation of normal human colonocytes and human colon carcinoma cells in vitro and ex vivo quantified by cell growth, colony formation, and ³H-thymidine incorporation for DNA synthesis.[11, 12, 65, 69], ^a Activation of GCC by ST induces a G1-S delay, quantified by flow cytometry and BrdU incorporation, that delays progression through the cell cycle, without apoptosis, measured by TUNEL analysis and DNA laddering, or necrosis, by trypan blue exclusion and lactate dehydrogenase release. Cytostasis induced by ligands is specifically mediated by GCC, associated with accumulation of cGMP, mimicked by the cell-permeant analog 8-Br-cGMP, and reproduced and potentiated by the cGMP-specific phosphodiesterase inhibitor zaprinast, but not the inactive ST analog TJU1-103. .[11, 12, 65, 69], ^a Cytostasis induced by cGMP was associated with altered expression of cell cycle mediators including cyclin D, pRb, and p27 regulating the transition through G1-S phase.[11, 12, 65, 69]

Also, activation of GCC signaling by ligands or cGMP reverts the tumorigenic metabolic phenotype in colon cancer cells. GCC signaling inhibits glycolysis and fatty acid synthesis by reducing rate-limiting enzymes including glucose transporter 1, hexokinase, pyruvate kinase and acetyl-CoA carboxylase and acid citrate lyase, respectively, ^a associated with a decrease in glucose uptake and lactate production. ^a Further, activation of GCC signaling in human colon cancer cells induces expression of critical transcription factors required for mitochondrial biogenesis, including PGC1α, mtTFA, and NRF1. ^a Moreover, GCC signaling promotes mitochondrial biogenesis by increasing the content of mitochondria, ^a associated with enhanced mitochondrial oxygen consumption, dehydrogenase activity and ATP production. ^a Reversion of the tumorigenic phenotype by GCC is generalizable, reproduced in numerous human and mouse colon cancer cell lines. ^a Of significance, activation of GCC signaling suppresses ROS production, ^a reflecting increased function of the electron transport chain[70–72] or increased production of ROS scavengers, [73, 74]and promotes genetic stability. ^a

Interestingly, a study[10] using Apc^Min/+ mice suggested that supplementation of uroguanylin significantly suppressed intestinal tumorigenesis, including tumor initiation and progression. Of significance, this study suggested that induction of GCC prevented tumorigenesis by promoting cell apoptosis, rather than suppressing proliferation. In that context, it is noteworthy that GCC signaling prevents, rather than induces apoptosis in human colon cancer cells [75]. Thus, mechanisms underlying inhibition of tumorigenesis by uroguanylin administration remain undefined.[10, 69, 76]

In summary, these observations suggest that re-activation of GCC signaling in human colorectal cancer cells by supplementing GCC ligands inhibits cell proliferation through cGMP-dependent mechanisms by coordinated regulation of cell cycle and metabolic circuits. In the context of universal hormone loss early in neoplasia, reconstitution of dormant receptor signaling by oral ligand supplementation may prevent the initiation and progression of colon cancer by suppressing proliferation and the associated metabolic reprogramming restricting premature entry into S phase and, therefore, maintaining genomic integrity. [10, 12], ^a

Future perspectives

The emerging paradigm in which colon cancer is initiated by hormone insufficiency suggests that, at initiation intestinal neoplasia could be prevented or treated by supplementing hormones to reconstitute disrupted tumor suppressing signaling pathways.[13, 14] Specifically, GCC signaling, which is compromised early in intestinal tumorigenesis reflecting near-universal hormone loss, maintains intestinal homeostasis by coordinating proliferation, metabolism, and genomic integrity along the crypt-surface axis (Fig. 4). Disruption of this paracrine hormone axis compromises the regenerative dynamics of the intestinal epithelium, leading to crypt hyperplasia, failure of lineage specification and commitment, and neoplastic metabolic programming. Disruption of those fundamental processes result in accumulation of mutations co-opting homeostatic circuits to produce a selection advantage permitting the emergence of progressive neoplastic transformation.

Figure 4. Molecular events underlying the colorectal adenoma-carcinoma sequence.[78].

Beyond sequential genetic mutations in oncogenes, tumor susceptibility genes or tumor suppressors in colorectal carcinogenesis, a novel paradigm is emerging which suggests that at initiation, intestinal neoplasia evolves from a state of paracrine (guanylin and uroguanylin) hormone insufficiency, providing a novel therapeutic opportunity for colorectal cancer prevention by ligand supplementation to reconstitute disrupted tumor-suppressing signaling pathways.

Beyond the view of pathogenesis, this emerging model suggests a correlative therapeutic hypothesis in which colorectal cancer can be prevented and treated employing the well-established paradigm originating in endocrinology in which diseases of hormone insufficiency are resolved by hormone replacement therapy.^{10, 12, a} In that context, the near-universal loss of GCC ligands early along the transformation continuum,[46, 48–51] and the specificity and compensatory over-expression of GCC in colorectal cancer cells[36, 56, 57, 77] support the utility of oral GCC ligand replacement to prevent and treat colorectal cancer. Oral supplementation of GCC ligands should restore and preserve intestinal paracrine hormone functions and defend homeostatic processes including proliferative restriction, lineage specification, metabolic programming and genomic stability that oppose intestinal neoplastic transformation. In that context, preliminary studies suggest that supplementation of GCC ligands is effective[10, 11, 65, 69, 76] and safe¹⁰ for colon cancer prevention and therapy. However, more studies in mouse models of colon cancer and clinical trials in human need to be performed to demonstrate the safety, efficacy, specificity and possible side-effects of oral supplementation of GCC ligands for targeted colon cancer prevention and therapy.

Abbreviations

AOM

azoxymethane

APC

adenomatous polyposis coli

BrdU

5′-bromo-2′-deoxyuridine

CFTR

cystic fibrosis transmembrane conductance regulator

cGMP

cyclic GMP

CNG

cyclic nucleotide-gated channel

GCC

guanylyl cyclase C

IHC

immunohistochemistry

PKA

cyclic AMP-dependent protein kinase

PKG

cyclic GMP-dependent protein kinase

pRb

phosphorylated retinoblastoma

RT-PCR

reverse transcriptase-polymerase chain reaction

ST

bacterial heat-stable enterotoxin