BRIEF REVIEWRobert F. Schwabe and John W. Wiley, Section Editors
Tissue Engineering in the Gut: Developments in NeuromusculatureKhalil N. Bitar,1,2 Shreya Raghavan,1,2 and Elie Zakhem1,2
1Wake Forest Institute for Regenerative Medicine; 2Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences,Wake Forest School of Medicine, Winston-Salem, North Carolina
The complexity of the gastrointestinal (GI) tract lies in itsanatomy as well as in its physiology. Several different celltypes populate the GI tract, adding to the complexity of cellsourcing for regenerative medicine. Each cell layer has aspecialized function in mediating digestion, absorption,secretion, motility, and excretion. Tissue engineering andregenerative medicine aim to regenerate the specific layersmimicking architecture and recapitulating function.Gastrointestinal motility is the underlying program thatmediates the diverse functions of the intestines, as anorgan. Hence, the first logical step in GI regenerativemedicine is the reconstruction of the tubular smoothmusculature along with the drivers of their input, theenteric nervous system. Recent advances in the field of GItissue engineering have focused on the use of scaffoldingbiomaterials in combination with cells and bioactive fac-tors. The ability to innervate the bioengineered muscle is acritical step to ensure proper functionality. Finally, in vivostudies are essential to evaluate implant integration withhost tissue, survival, and functionality. In this review, wefocus on the tubular structure of the GI tract, tools forinnervation, and, finally, evaluation of in vivo strategies forGI replacements.
Torgan responsible for the transport and digestion offood, absorption of nutrients, and excretion of waste. Theactivity of the GI tract is a summation of several complexcell types that include smooth muscle cells, neurons, glia,interstitial cells, and different kinds of intestinal epithelialcells. The outer layer of the GI tract is composed of 2 typesof smooth muscle tissues: circular and longitudinal smoothmuscle. The sphincters of the GI tract allow unidirectionaland directed flow of luminal contents. Apart from thesmooth musculature, the GI tract contains several kinds ofintestinal epithelial cells that mediate absorption andsecretion within the gut. Smooth muscle tissues are theprimary effectors of motility in the gut, mediating themovement of luminal content. The function of the muscletissue is dictated by the enteric nervous system (ENS),which is the intrinsic innervation of the gut. Several classesof functional neurons (sensory, motor, secretory, and soforth) and glia are present in the ENS, with a diversity
paralleled only by the central nervous system.1 The ENS isresponsible for the variety of gastrointestinal motor pat-terns produced in different parts of the gut, as well as thecoordination of function between various segments of thegut. The interstitial cells of Cajal also additionally areimplicated in pacemaking function within the gut,2 roundingout the primary players responsible for gastrointestinalmotility.
Gastrointestinal motility can be altered postnatally as aresult of disease, damage, surgical or obstetric trauma, andage. Congenital defects of GI motility include but are notlimited to Hirschsprung disease, intestinal pseudoob-struction, and achalasia.3 Although the therapeutic main-stay for motility disorders has remained pharmacologic,surgical correction also does not provide a long-lastingsolution. Regenerative medicine seeks to replace GI seg-ments, preferably using the patient’s own cells while usingthe optimal route of delivery. Advances in biomaterialsand tissue engineering have catapulted regenerativemedicine strategies forward, bringing them closer to thebedside.
This review focuses on regenerative medicine strategiesaimed at the restoration of the neuromuscular anatomyand/or function of the neuromusculature of the GI tract.This review highlights both biomaterial-based and celltransplantation–based methods. Finally, a future perspec-tive is provided indicating the complexities of sourcing andmaintaining phenotypes of many constituent cells, neo-innervation, and neovascularization.
Tissue Engineering of GI TubularOrgans: Where Do We Start?Anatomy and Function
Tissue engineering the GI tract has fundamental chal-lenges that one would encounter when faced with mostbiological systems—anatomic and physiological complexity.The complexity of the GI tract lies in the different cell layersthat exist within the tract. These cells work in coordination
June 2014 Regenerative Medicine of the GI Tract 1615
REVIEW
SAN
DPE
RSPE
CTIVES
to respond appropriately to different stimuli. In GI tissueengineering, each of the different cell types must beconsidered. The first question that arises in any tissue en-gineering application is the appropriate source of cells. Canthe several cell types required to duplicate physiologicalcomplexity be sourced? If yes, can they be sourced inadequate numbers from a biopsy specimen, which is pref-erably minimally invasive?
Musculature. The GI tract is a complex, highly regu-lated, multilayered system. Although the muscularis propriais divided into several different layers, its complexity isdetermined by its anatomy. For example, different muscletypes exist along the length of the esophagus. The first onethird is composed of skeletal muscle and the lower one thirdis composed of both circular and longitudinal smoothmuscle. The middle part of the esophagus is a mixture ofboth skeletal and smooth muscle. The variety of muscletypes is essential to ensure swallowing and propulsion offood into the stomach. In addition to the circular and lon-gitudinal smooth muscle layers that make up the gastricmusculature, another oblique smooth muscle layer exists.4
Taking into account the different muscle types and theirorientation is critical when engineering GI segments. In thesmall and large intestine, the muscular layer is divided againinto longitudinal and circular muscle layers. The basic unitof intestinal musculature is the smooth muscle. It receivesregulatory inputs from different levels to perform its con-tractile functions. Smooth muscle contraction is initiated bymembrane depolarization, which activates voltage-gatedcalcium channels and leads to calcium (Ca2þ) influx intothe cell. Entry of Ca2þ stimulates complex signaling cascadeswithin the smooth muscle cell, leading to contraction.Different intracellular proteins are involved in mediating thecontractile response via a series of phosphorylation anddephosphorylation (G proteins, phospholipases, calmodulin,myosin light chain kinase, Rho kinases, and phosphatases).Although similar pathways exist for both circular and lon-gitudinal smooth muscle layers, mechanisms for Ca2þ
mobilization differ between the two. Depolarization ofsmooth muscle is controlled at different cellular levels in theGI tract. Contractile response is divided into 2 phases: initialand sustained contractions. A balance between phosphory-lation and dephosphorylation events is responsible for theoccurrence of the 2 phases. A complete regeneration of themusculature of the GI tract requires the maintenance of allthe intracellular pathways. Smooth muscle cells also can beclassified as phasic or tonic smooth muscle based on thecontents of contractile protein isoforms and their levels ofexpression. A higher level of contractile protein expressionis observed in smooth muscle cells present in the high-pressure zones of the GI tract, namely the sphincters ofthe GI tract. For example, circular smooth muscle of thelower esophageal sphincter (LES) or the internal analsphincter (IAS) differ significantly from the smooth muscleof the esophageal body or the colorectum, respectively.5,6
Epithelium and defense functions. The epitheliumof the gut performs various critical functions such asenzyme secretion, nutrient absorption, and acts as a phys-ical barrier to perform a highly sophisticated defense
function. Secretion and absorption functions require a largesurface area, which is provided by finger-like villi structuresthat face the luminal side of the gut. The epithelial mono-layer is characterized by apicobasolateral polarity and isdivided into several specialized cell types: enteroabsorptivecells, goblet cells, Paneth cells, and neuroendocrine cells.7
Polarity of the epithelium is essential for function, it al-lows cells to sense and respond to stimuli.8 The epitheliumdiffers in structure and function along different parts of theGI tract.9 Apart from nutrient absorption, the epitheliumalso provides a defense function. The gut has to have anappropriate defense system capable of protecting itself fromcommensal bacteria as well as foreign antigens in its luminalcontent. A single layer of epithelial cells makes up theepithelial barrier as a primary defense wall. A coordinatedinteraction between the different epithelial cells contributesto the defense function of the epithelium. The integrity ofthe epithelial barrier is maintained by tight junctions. Inaddition to tight junctions, adherens junctions and gapjunctions also are involved in cell–cell interaction. Whatmakes the epithelial barrier more complex is the fact that itis a dynamic structure that is renewed continuously in thecontext of epithelial cell shedding and proliferation.10,11
Thus, the gut is vulnerable and is a potential site for infec-tion, inflammatory diseases, and loss of the epithelial barrierintegrity. Part of the regeneration process must involvegenerating a polarized epithelium with tight junctionintegrity to reinstate mucosal function.
Regulatory apparatus. The regulatory apparatus ofthe GI tract is multilayered—with input arising from theintramural innervation, integrated inputs from the centraland autonomic nervous system, as well as input frominterstitial cells of Cajal (ICCs). Ultimately, GI physiology is acomplex addition and interpretation of several signals thatlead to smooth muscle motor activity, mucosal secretion/transport, local blood flow/vasodilation, and intestinal im-mune and endocrine function. The ENS is contained entirelywithin the musculature of the GI tract and is arrangedwithin ganglionated plexuses. The myenteric plexus islocated anatomically between the circular and longitudinalmuscle layers, and extends the full length of the digestivetract from the esophagus to the rectum. The submucosalplexus lies between the mucosa and the inner circularmuscle layer. Together, the neurons of the ENS regulatediverse functions such as control muscle activity, secretoryactivity of intestinal glands, motility of the blood vessels,and sensory functions that include reflex pathways.Depending on the region of the gut, the size and composi-tional diversity of the ganglia within the plexuses vary. Forexample, the submucosal plexus is far more obvious in thesmall and large intestines than in the stomach, and thenumber and neuron density of myenteric ganglia are higherin the colon near the mesenteric attachment.12–14 Differentclasses of enteric neurons exist with diverse neurochemicalcoding.
Neurotransmitters in the intestine are similar to those ofthe central nervous system (CNS), and include acetylcholine,tachykinins, serotonin, nitric oxide, purines, and severalneuropeptides.15 Motor neurons are responsible for
1616 Bitar et al Gastroenterology Vol. 146, No. 7
REVIEWSAND
PERSPECTIVES
excitatory and inhibitory transmission to smooth muscle ofboth the muscularis externa and the muscularis mucosa.Neuroneuronal transmission is mediated by 2 types of in-terneurons: ascending and descending interneurons.Although ascending interneurons are thought to mediatepropulsive reflexes, descending interneurons have beenimplicated in local motility reflexes, secretomotor reflexes,as well as conducting migrating myoelectric complexes. Thebasis for these functions arises in complex neurochemicalcoding for these neurons. Intrinsic sensory neurons(intrinsic primary afferent neurons) are responsible forseveral reflexes that mediate mucosal chemo-sensing,mechano-sensing, and stretch responsiveness. Intrinsicsecretomotor neurons are controlled both by local reflexcircuits and from sympathetic input outside of the ENS.Extrinsic vasodilation is known to occur through sensoryneurons that originate in the dorsal root ganglia with pro-cesses directed toward the GI microvessels.16,17 Submucosalsecretomotor neurons also have been implicated in neuro-genic vasodilation and control of mucosal blood flow.16,18
Although a majority of the neuronal input arises fromnerves of local origin, afferent sensory inputs from the in-testine are transmitted to the CNS via the nodose ganglia ofthe vagus or the dorsal root ganglia. For example, sensationsof fullness from the stomach are transmitted through thispathway.19 Efferent impulses from the CNS primarily areparasympathetic in nature from the vagus and pelvic nerves,and sympathetic from prevertebral ganglia. An example ofintegrated neurogenic response is the vago-vagal reflex inthe esophagus that mediates swallowing-induced peristalsis.In addition to the neural input, smooth muscle intrinsicelectrical activity is mediated by the ICCs. Three populationsof ICCs have been identified in the human colon using c-Kitimmunohistochemistry. These include the ICCs along thesubmucosal surface of the circular muscle, ICCs within thedeep muscular layers, and ICCs in the myenteric region.20
The role of ICCs in neurotransmission, slow wave elec-trical activity, and functions in the gut have been reviewedextensively.21,22 ICCs have a syncytial pacemaker activitythrough which they regulate smooth muscle and slow-waveelectrical activity. Significant improvement has been madein the past decade for the isolation of ICC populations.23–25
Incorporation of these cells during regeneration of the in-testine might allow more efficient neurotransmission andtimely peristaltic activity. Small steps are being made to-ward recapturing the diversity of the intrinsic innervation tothe gut as well, which is discussed in detail in the section,Regenerative Medicine and Neoinnervation.
Tissue Engineering the GI Tract: HowCan the Physical Structure BeAchieved?
Once the biological challenges have been defined, amaterials and structural question arises. Despite the factthat tubular organs have simple geometry, their structuraland mechanical properties are complex. The esophagus is along muscular tube that mediates the transport of food to
the stomach via peristalsis. The stomach functions as areservoir that grinds and mixes food. The small and largeintestines are tubular structures that function to enhancethe absorption of nutrients, transit of food, and, finally,excretion of waste via coordinated rhythmic peristalticwaves. The GI segments are interspersed with sphinctersthat ensure unidirectional flow and prevent any backflow. Ifanatomic, structural, and physiology complexity were to beachieved, even one step at a time, or focused on single celllayers, the next rate-limiting step in tissue engineering isneovascularization. Successful implantation necessitates cellsurvival and maintenance of biological functions post-implantation. Vascularization guarantees transport ofoxygen and nutrients and excretion of waste products.Several tissue engineering approaches can be considered tofacilitate vascularization including delivery of angiogenicfactors or prevascularization. Vascularization strategieshave been discussed extensively in several reviews,26,27 andis not a subject of the current review.
BiomaterialsA wide spectrum of biomaterials can be used to fabricate
tubular scaffolds that mimic the GI tubular organs. Naturalas well as synthetic biomaterials have been investigated inGI regenerative medicine for their ability to support cellattachment, proliferation, and differentiation. In any tissueengineering application, scaffolds are characterized by theextent and degrees of biocompatibility, porosity and poresizes, and mechanical properties, among several other spe-cific material properties.28 Those characteristics dictate thekind of interaction cells will have with the material in termsof cell behavior, alignment, phenotype maintenance, andfunction in a 3-dimensional setting. The key point for suc-cessful tissue regeneration and function is to mimic thenative alignment of the cells. Early studies failed to restorethe proper alignment of different layers of the GI tract usingcell-seeded scaffolds.29 Described within this section areinstances in which specific biomaterial interactions wereused for achieving cellular alignment during the tissue en-gineering process.
Natural materials. Collagen, a major component ofthe extracellular matrix, is one of the most common naturalbiomaterials used in the field of GI tissue engineering.Collagen scaffolds were fabricated and evaluated for smoothmuscle attachment and growth in gastric and intestinalregeneration.30–33 In addition, collagen was used as acoating material in gastric reconstruction.34 Collagen sup-ported cell attachment and differentiation; however, it failedto support the recapitulation of the native architecture.Chitosan is another natural polymer commonly used intissue engineering applications.35,36 We have shown thebiocompatibility of chitosan in GI applications.37 Intestinalsmooth muscle cells attached, proliferated, and maintainedtheir contractile phenotype when cultured on chitosan–collagen membranes.
Our group has bioengineered circular smooth muscletissue constructs using fibrin gel.38 The smooth muscle cellswere aligned circumferentially in a pattern that mimics theorientation of circular smooth muscle cells in native GI tract.
June 2014 Regenerative Medicine of the GI Tract 1617
REVIEW
SAN
DPE
RSPE
CTIVES
We also bioengineered composite tubular scaffolds made ofchitosan and collagen. We fabricated highly porous scaffoldswith different lengths and diameters. A lumen also wascreated to obtain a hollow tube that could be used to replacetubular organs of the GI tract. The advantage of using chi-tosan is its ability to be immobilized to glycosaminogly-cans35 such as heparan sulfate, which is an abundantglycosaminoglycans in intestinal extracellular matrix.39 Weplaced several bioengineered smooth muscle tissue con-structs around the chitosan scaffolds and showed theirphysiological functionality in response to potassium chlo-ride, acetylcholine, and vasoactive intestinal peptide (VIP).When designing these experiments toward clinical trans-lation, a normalized comparison with native tissue for ki-netics and magnitude of physiological responses will be arequirement. Physiological analyses must include mandatesto resolve phasic or tonic activity of the smooth muscle typeunder investigation. In addition, the underlying mechanicalproperties of the biomaterial scaffold need to be evaluated,to elucidate the evolution of the structure-function rela-tionship of physiologically functional smooth muscle. Mov-ing forward, a critical review of the biomaterial itself alsoneeds to be undertaken, for instance, how much remodelingdoes the material allow, how much infiltration does thescaffold support, and what kind of immunogenic responseswould be acceptable for the material to be considered safefor transplantation.
Investigations have shifted toward modifying the struc-ture and chemistry of the scaffolds. OptiMaix collagenscaffolds (Optimaix 3D; Cellon, Luxembourg) were fabri-cated by unidirectional porous structures that alloweddirectional smooth muscle growth.40 Totonelli et al41 suc-cessfully optimized a decellularization protocol using ratsmall intestine to completely remove cellular componentsand preserve the native architecture of the intestine. Themaintenance of architecture helps guide the cells to orientthemselves on the matrix. Our group also has bioengineeredlongitudinal smooth muscle tissues.42 Uniaxial smoothmuscle alignment was facilitated by substrate micro-topographies. The bioengineered tissues mimicked thealignment of the longitudinal layer of the small and largeintestines. The tissues also showed physiological function-ality (contraction and relaxation).
Synthetic materials. Synthetic materials also repre-sent strong candidates owing to their biocompatibility,biodegradability, mechanical properties, and ease of fabri-cation. However, because of the lack of binding domains insynthetic polymers, addition of natural polymers has beeninvestigated to enhance cell attachment and survival.43 In arecent study, poly(L-lactide-co-caprolactone) scaffolds wereelectrospun and immobilized with fibronectin to enhancecell attachment and growth for esophageal reconstruction.44
Synthetic polymers also were investigated in their ability toregenerate different layers in the stomach. Polyglycolic acidscaffolds were coated with collagen and seeded with orga-noid units.34 Epithelial and muscularis layers were regen-erated; however, the native architecture was notrecapitulated. Recently, a tissue-engineered small intestinewas formed using polyglycolic acid scaffolds coated with
collagen and seeded with postnatally derived progenitorcells.45 The scaffolds were implanted in the omentum ofmice for a period of 28 days. Immunostaining studiesshowed the presence of differentiated epithelial cells, mus-cularis layer, and nerve tissue. The use of organoid units toregenerate segments of the GI tract requires further opti-mization in regards to the source of the units, the age of thedonor, and the initial level of differentiation of the units.
Mechanical characterization of tissues.Studying the mechanical properties of GI tissues is valuablein providing insight into the developmental stages ofregenerative medicine–based therapies; it allows us to (1)understand the biophysical mechanisms behind the trans-port of luminal contents; (2) understand the pathophysi-ology of diseases associated with an underlying change inthe mechanical properties of the tissue; and (3) to developscaffolds suitable for replacements. Extensive mechanicalstudies have been conducted on tissue-engineered tubularorgans such as blood vessels,46–48 but little emphasis isdirected to the study of the mechanics of GI tissues. Previ-ous tensile strength studies on rats showed a trend ofincreased tensile strength from proximal to distal colon.49
This change was correlated to the exposure of the colon tohigher stress because the fecal pellets become more solid.Egorov et al50 reported tensile properties of humanesophagus, stomach, and small and large intestines. In thisreport, the small and large intestines were studied asmultilayered structures. It was concluded from this studythat the mechanical strength of the intestinal wall is deter-mined predominantly by the submucosa and muscularislayers with little contribution from the serosa and the mu-cosa. Another study compared the mechanical behavior ofthe whole esophagus with the mucosa alone, implicating amucosal contribution to the strength of the esophagealwall.51 Another study has shown a decrease in tensilestrength of the colon as a function of age.52 This correlatesto a change in the connective tissue composition that occurswith age. Recently, a new technique has been used tomeasure mechanical properties of the esophageal wall as away to provide therapy for patients.53 The study showed adecrease in esophageal distensibility and esophagealcompliance in patients with eosinophilic esophagitis. Theoutcome of this study helps correlate the mechanicalproperties of the esophageal wall with tissue remodelingand fibrosis in pathologic conditions. Distinguishing themechanics of the gut wall between the different layers helpsin the design and synthesis of the scaffolds to match thenative tissue.
Regenerative Medicine andNeoinnervation
Gastrointestinal function is controlled predominantly bythe ENS, which is considered as the intrinsic innervation ofthe GI tract. ENS consists of an enormous diversity of entericneurons as well as enteric glia, both of which derive fromthe neural crest. Neurodegenerative conditions are charac-terized by the loss of neuronal circuitry, which results indysmotility in the GI tract.
1618 Bitar et al Gastroenterology Vol. 146, No. 7
REVIEWSAND
PERSPECTIVES
Stem Cell TransplantationNeural stem cell transplantation provides a cell-based
therapy to reinstate innervation in segments of the GI tract.The main goal of cell transplantation is to restore neuronalfunction in cases of neurogenic disorders. However, the long-term fate of the transplanted cells requires further in-vestigations. It is ideal to obtain sufficient quantities of stemcells from an easily accessible source, using minimally inva-sive procedures.54 Different sources of neural stem cells arebeing investigated actively, including the central nervoussystem and the enteric nervous system. Delivery of stem cellsis another factor that needs to be taken into account. Prog-ress in tissue engineering and regenerative medicine pro-vides different materials as delivery vehicles for the cells.Delivery of the cells must ensure target specificity andcomplete repopulation with the specific neuronal subtypesappropriate for the location along the GI tract. Availability ofcell source, viability, and differentiation of the stem cells aftertransplantation are critical factors that must be accounted forwhen considering clinical use.
Enteric nervous system as a source for progenitorcells. Techniques for isolation of enteric neural progenitorcells from embryonic, fetal, postnatal, and adult rodent, aswell as from human, GI tract already have been establish-ed.55–61 They can be isolated from full-thickness, muscu-laris, and mucosal biopsy specimens with the ability todifferentiate into mature neurons. Recently, neural crestprogenitor cells were isolated form neonatal rats and weredelivered to a denervated colon. The cells differentiatedinto neurons and glia and were able to restore motility.62
In another study, fetal and postnatal neural progenitorcells derived from mice intestines were transplanted intothe external muscle layer of mice distal colon.63 Thetransplanted cells migrated, proliferated, and differenti-ated into excitatory and inhibitory motor neurons.Although motility studies were not performed in theseanimals, there was no sign of dysmotility. Longitudinalsmooth muscle function in the piebald mouse improvedafter implantation of an ENS-derived cell line.64 Intraperi-toneal injection of enteric neuronal stem cells into lethalspotting rats (ls/ls) showed engraftment from the stomachto the cecum, with an associated expression of markersindicating neuronal differentiation.65 Enteric neuronalstem cells carrying the green fluorescent protein transgeneand transplanted directly into recipient mouse colons wereable to migrate, differentiate into functional neurons andglial cells, and form ganglion-like structures.63
Embryonic mouse and neonatal human enteric neuronalprogenitor cells differentiated into mature neuronal phe-notypes after transplantation into embryonic mouse hindgutorganotypic cultures.59 ENS-derived neural stem cells iso-lated from human or rodent bowels have been transplantedsuccessfully into recipient embryonic aganglionic bowelexplants.54,59 These cells survived, migrated, and differen-tiated and were able to impart neuronally dependentcontraction and calcium mobilization.59
Central nervous system as a source for stemcells. Neural stem cells derived from the central nervoussystem also have been tested for their ability to treat
gastrointestinal disorders.66 Micci et al67 transplanted CNS-derived neuronal stem cells isolated from embryonic mousebrain into mouse pylorus. Neuronal stem cells differentiatedupon transplantation, and rescued gastric function in theneuronal nitric oxide synthase knockout mice. Dong et al68
transplanted CNS-derived neuronal progenitor cells intoaganglionic rodent rectum. These cells showed differentia-tion into glial and neuronal (nitrergic and cholinergic)subtypes, and restored the rectoanal inhibitory reflex in therat. In addition, ex vivo studies have been conducted. CNS-derived neuronal stem cells can acquire an enteric-likeneuronal phenotype when co-cultured with mouse longitu-dinal muscle–myenteric plexus preparations.60,66,69
Neoinnervation of Smooth MuscleOur group has shown the ability to neoinnervate bio-
engineered smooth muscle tissue constructs. An intrinsi-cally innervated smooth muscle tissue construct wasbioengineered using smooth muscle cells and neural pro-genitor cells isolated from full-thickness adult rabbit colons.The intrinsically innervated smooth muscle construct wasplaced around chitosan scaffold next to a bioengineeredsmooth muscle construct that initially lacked innervation.The latter became neoinnervated after a period of 14 days inculture. Both constructs responded to electrical field stim-ulation and stained positive for b-III tubulin. This studyprovides a methodology whereby a combination of bio-materials and cells are used to reinstate innervation inneuromuscular diseases of the gut.70
Currently, regenerative medicine is directed at restoringinnervation of gut smooth muscle with the goal of differ-entiating stem cells into functional subtypes of neurons.Providing chemical cues to the enteric neural progenitorcells can modulate their fate in vitro. The benefit of directingthe differentiation of neural progenitor cells into specificneuronal subtypes is to obtain an enriched population ofcertain types of neurons. This is promising when it comes totreating GI disorders that are caused by the lack of a specificpopulation of neurons.71–73 Gastric emptying and gastricneuromuscular dysfunctions are characterized by theabsence of intrinsic inhibitory neurons. Achalasia is a con-dition in which the sphincter (LES or IAS) is in a toniccontractile state and is unable to relax. In all cases, anincreased population of nitrergic neurons is beneficial inrestoring motility, facilitating gastric emptying, andreducing sphincteric tone.
We isolated enteric neural progenitor cells from full-thickness adult rabbit guts following a previouslydescribed protocol.54 The enteric neural progenitor cellswere positive for p75 (neural-crest lineage) and for sox2and nestin (progenitor status). The extracellular matrix is animportant factor in determining the lineage fate of stemcells. We evaluated the in vitro differentiation capacity ofenteric neural progenitor cells on different extracellularmatrix components known to be abundant in the myentericplexus.74 The extent of neuronal and glial differentiationvaried among different substrates. Composite collagenmixtures (collagen IV, laminin, and heparan sulfate)
June 2014 Regenerative Medicine of the GI Tract 1619
REVIEW
SAN
DPE
RSPE
CTIVES
improved neuronal differentiation, however, glial differen-tiation was reduced.
In a recent study, we bioengineered intrinsically inner-vated human IAS constructs. We isolated human IAS smoothmuscle cells and human enteric neural progenitor cells fromthe adult human small intestine. Cells were co-cultured incollagen hydrogel and formed 3-dimensional tissue con-structs. Immunostaining studies showed the presence ofcontractile smooth muscle, glial cells, and mature neuronalsubtypes. The mature neurons were able to produce excit-atory (choline acetyltransferase) or inhibitory (VIP andneuronal nitric oxide synthase) neurotransmitters. In addi-tion, the constructs responded to appropriate physiologicalstimuli using neurotransmitters that normally are present inthe gut. Neuronal integrity also was shown by the responseto electric field stimulation. We show the ability to innervatebioengineered human smooth muscle tissues using gut-derived enteric neural progenitor cells.75
Enteric glia are distributed in all layers of the GI tractwall and play a regulatory role at different levels. During theregeneration process, a glial population could be beneficialin providing trophic and structural support for neurons.They would ensure neuronal survival, protection, and dif-ferentiation. In addition, glial cells also are involved inenteric neurotransmission by regulating neurotransmitterssynthesis. At the mucosal level, glial cells play a role inregulating the integrity of the epithelial barrier. Althoughthere are no established GI disorders that specifically affectenteric glia, it is beneficial to identify, in future studies,factors that selectively induce glial differentiation.76,77
In Vivo Studies: Current StatusStudies involving implantation offer a basis for clinical
translation. Several studies have reported different attemptsto reconstruct segments of the GI tract in animal models.Neovascularization and neoinnervation are among the mostchallenging factors to be considered after implantation. It isof paramount importance that the implant integrates withthe host tissue, receives adequate blood supply for cellsurvival, and becomes innervated for proper functionality.Many studies were conducted in rodents and resulted inpromising outcomes; however, larger animal models areprerequisites for approaching clinical settings.
Tubular Organ ReconstructionIn an attempt to improve scalability of bioengineered
products to make appropriate for humans, large animalstudies have been conducted. A synthetic bioabsorbablepolymer patch was used to reconstruct the esophageal wallin pigs.78 Up to 12 weeks of follow-up evaluation after im-plantation showed regeneration of the esophageal muscularlayer with alignment similar to native tissue. Other studiesalso have used decellularized matrices (urinary bladder andsmall intestine submucosa) in an attempt to reconstruct theesophagus in dog models.79,80 Gastric patches, in addition topolymeric scaffolds, represent potential methods forreconstructing the stomach. In vivo studies have resulted inregeneration of the epithelial layers but failed to regenerate
the muscularis layer in terms of architecture and func-tion.34,81 In regenerating the small intestine, syntheticscaffolds seeded with organoid units were successful inregenerating the epithelial layer after implantation.82–84 Thechallenge remains in recapitulating the organization of bothcircular and longitudinal muscle layers. In another study,small intestine submucosa seeded with smooth muscle cellsresulted in partial epithelialization after implantation butwas not successful in regenerating the muscularis layer.85
Sphincter ReconstructionSphincters are composed of tonic circular smooth mus-
cle and show high pressures as a result of the contracturestate of the muscle. Sphincteric deficiencies caused bydamage of the smooth muscle lead to different disordersthat can have overwhelming impacts on patients’ lives.These disorders include gastroesophageal reflux disease,which affects the LES, fecal incontinence, which affects theIAS, achalasia (LES and IAS), and stenosis (pylorus). Firstattempts to repair defects in sphincters used cell-basedtherapies that involved injection of skeletalmuscle–derived stem cells.86,87 Functional contractilesmooth muscle was not reported. Tissue engineering offersa different approach to repair sphincteric degeneration.
Our group has well-established protocols for engineeringsphincteric smooth muscle. We have bioengineered intrin-sically innervated internal anal sphincter smooth muscletissue constructs. The constructs were bioengineered usinghuman IAS smooth muscle cells co-cultured with immora-tilized cultures of fetal enteric neurons.88 The presence ofexcitatory and inhibitory neurons in the constructs wasshown by reverse-transcription polymerase chain reactionand immunostaining studies. The constructs were implan-ted subcutaneously in mice. After 4 weeks, the harvestedconstructs were neovascularized and preserved theneuronal network as depicted by immunostaining. Theharvested constructs maintained physiological functionality(myogenic and neuronal components) in response toacetylcholine, VIP, and electrical field stimulation. Bio-engineered human IAS constructs, by us and others, haveshown basal tone values ranging from 0.4 to 0.7 mN,measured in organ bath studies.89–91 Forces measured inorgan bath studies are not directly comparable with clinicaldata obtained for basal tone from anorectal manometry. Thediscrepancy when this type of direct comparison is madewith human IAS is thought to arise from not having anadequate number of cells within the engineered constructs.Furthermore, several neurohumoral mechanisms and addi-tional muscle groups are responsible for the generation ofbasal tone in the human IAS, which are not captured whilemeasuring isometric force from bioengineered smoothmuscle constructs.92 Although the magnitude of the forceswere reduced substantially when compared with forces thatwould be produced in an intact human IAS tissue, bio-engineered constructs retained key aspects of IAS physi-ology by generating spontaneous increased basal tone.Although considering these constructs for clinical replace-ment of IAS, rigorous quality standards have to be
1620 Bitar et al Gastroenterology Vol. 146, No. 7
REVIEWSAND
PERSPECTIVES
established and met, which are discussed in later sections.However, orthoptic transplantation to augment sphinctericfunction in place of repeated injections of inert bulkingagents such as Deflux (Salix Pharmaceuticals Inc, Raleigh,NC) and botulinum toxin into the IAS appears to be aninviting option.
In a separate study, we studied the effect of differentangiogenic factors on the viability and functionality of bio-engineered smooth muscle tissue constructs.93 The bio-engineered IAS constructs were implanted subcutaneouslyalong with osmotic pumps in the back of mice. The pumpswere delivering 1 of the 3 different factors: fibroblastgrowth factor, vascular endothelial growth factor, orplatelet-derived growth factor. All growth factors promotedneovascularization, survival, and maintenance of contractilesmooth muscle phenotype in the constructs when comparedwith control (implanted constructs without delivery ofgrowth factors). However, platelet-derived growth factorand vascular endothelial growth factor resulted in superiorbasal muscle tone generated by the implanted constructs.
Future PerspectivesAdvances in biomaterials and tissue engineering have
brought regenerative medicine of the neuromusculature ofthe GI tract to fruition. Over the past decades, biomaterialcharacteristics have been refined to imitate near-native ar-chitecture and promote optimal interaction with several celltypes. Several material- and bioreactor-based methods havebeen reported to facilitate smooth muscle cell alignment,recapitulating native muscle organization similar to the GItract. From a biological point of view, however, the celltypes within the native GI tract are diverse to a point whereisolation and characterization of individual cell types musthave defined standards. A standard set of markers toidentify cell types has to be established, to standardize cellsourcing for transplantation or tissue replacements. Ofspecific interest, given the notoriety of smooth muscle cellsto dedifferentiate into synthetic phenotypes during prolif-eration, care must be taken to ensure that replacementtissue constructs have constituent smooth muscle cells thatexpress specific markers of the contractile phenotype.Reinstatement of GI function must compulsorily includecells other than smooth muscle and neurons. As the role ofthe interstitial cells of Cajal in GI function emerges, thesecells may be critical in reinstatement of the pacemakingactivity.
Quality Control of Smooth Muscle PhenotypeSeveral distinct smooth muscle phenotypes have been
identified during the development of the intestinal smoothmusculature, including myoblasts, immature myocytes, andmature smooth muscle myocytes.94,95 This continuum ofsmooth muscle phenotypes results from hierarchicalinductive events of specific gene products including iso-actins, calponin, and a myosin heavy chain isoform, amongseveral others.96,97 Mature smooth muscle cells retain thenotorious ability to de-differentiate into synthetic, prolifer-ative, and possibly migratory phenotypes when they are
isolated from their native in vivo milieu and placed in cul-ture. However, these cells also are documented to recapturetheir myogenic program in vitro. The challenge in derivingsufficient numbers of smooth muscle cells for GI tissue en-gineering is that the induction of smooth muscle prolifera-tion in low-density cultures with serum-supplementedmedium is an automatic result of smooth muscle de-differentiation. Although this developmental flexibility per-mits a high proliferative index, de-differentiated smoothmuscle cells show reduced contractility and acetylcholineresponsiveness.98,99 The recovery of the contractile smoothmuscle phenotype is paramount to its function in generatingforces required for the various motility patterns of the GItract. Although research has shown that high-density cul-tures of smooth muscle cells with cellular syncytium andhigh cell–cell contact result in a more contractile pheno-type,100 establishing reproducible standards for analyzingthe quality of the constituent cells is important. Researchersoften report the expression of a-smooth muscle actin asconfirming smooth muscle phenotype. However, in somecases, populations of cultured fibroblasts and myofibro-blasts express this marker, making it less reliable in iden-tifying the smooth muscle phenotype.101,102 Alternatively,the smooth muscle–specific heavy isoform of caldesmon,calponin, and smoothelin are all induced and observed in“contractile” smooth muscle cells or fully mature smoothmuscle cells.97,103,104 These markers may be more reliablewhile ensuring the contractile phenotype of the smoothmuscle cells. Several studies now document differentialtranscriptional activities and gene expression profiles ofgenes encoding both contractile proteins as well as thoseassociated with the extracellular matrix in the variousphenotypes of smooth muscle.105-110 Although aiming fortranslational ability, GI tissue engineering must take a cuefrom vascular tissue engineering in establishing rigorousquality control based on complementary DNA screening, toensure contractile smooth muscle phenotypes.
Integration of NeoinnervationThe principal roadblock in clinical translation remains
adequate and appropriate reinstatement of ENS innervation.Autologous sources of stem cells derived from the GI tracthave been identified and isolated in sufficient numbers toshow the feasibility of this cell source for therapy. Currenttherapies focus on injecting various types of stem and pro-genitor cells, in the hope that environmental cues will drivethem to differentiate into phenotypes appropriate for thelocation. Trophic factor regimens that drive subtype differ-entiation and maintain phenotype need to be identifiedin vitro, for incorporation into the next generation of bioac-tive scaffolds. Of particular interest are extracellularmatrix–based microenvironmental cues that have the capa-bility of modulating both trophic and morphogeneticsignaling to the encapsulated neuronal stem and progenitorcells. The roadmap to neoinnervation must include neuronalnetwork formation and connectivity, otherwise discordancebetween neoinnervation and existing complex reflex path-ways of the gut may persist. More fundamental research
June 2014 Regenerative Medicine of the GI Tract 1621
REVIEW
SAN
DPE
RSPE
CTIVES
studying the interactions and communication of entericneurons with peripheral neurons and among themselves isrequired to be able to establish standards for neural con-nectivity. Overall, in view of clinical translation, basic ques-tions pertaining to phenotypes of isolated cell types (smoothmuscle, neurons, stem/progenitor cells, interstitial cells)have to be investigated. Several studies outlined within thisreview show promise for the identification of some of thesefactors, driving regenerative medicine of the GI neuro-musculature to be a reality in the coming decades.
References
1.Furness JB. The enteric nervous system and neuro-gastroenterology. Nat Rev Gastroenterol Hepatol 2012;9:286–294.
2.Sanders KM, Koh SD, Ro S, et al. Regulationof gastrointestinal motility—insights from smoothmuscle biology. Nat Rev Gastroenterol Hepatol 2012;9:633–645.
3.Chumpitazi B, Nurko S. Pediatric gastrointestinal motilitydisorders: challenges and a clinical update. Gastro-enterol Hepatol 2008;4:140.
4.Smith GT, Moran TH, Coyle JT, et al. Anatomic locali-zation of cholecystokinin receptors to the pyloricsphincter. Am J Physiol 1984;246:R127–R130.
5.Szymanski PT, Chacko TK, Rovner AS, et al. Differencesin contractile protein content and isoforms in phasic andtonic smooth muscles. Am J Physiol 1998;275:C684–C692.
6.Patel CA, Rattan S. Spontaneously tonic smooth musclehas characteristically higher levels of RhoA/ROKcompared with the phasic smooth muscle. Am J PhysiolGastrointest Liver Physiol 2006;291:G830–G837.
7.Pott J, Hornef M. Innate immune signalling at the intes-tinal epithelium in homeostasis and disease. EMBO Rep2012;13:684–698.
8.Martin-Belmonte F, Perez-Moreno M. Epithelial cell po-larity, stem cells and cancer. Nat Rev Cancer 2012;12:23–38.
9.Karlsson J, Putsep K, Chu H, et al. Regional vari-ations in Paneth cell antimicrobial peptide expressionalong the mouse intestinal tract. BMC Immunol2008;9:37.
10.Marchiando AM, Shen L, GrahamWV, et al. The epithelialbarrier is maintained by in vivo tight junction expansionduring pathologic intestinal epithelial shedding. Gastro-enterology 2011;140:1208–1218.
11.Ulluwishewa D, Anderson RC, McNabb WC, et al.Regulation of tight junction permeability by intestinalbacteria and dietary components. J Nutr 2011;141:769–776.
12.Sternini C. Structural and chemical organization of themyenteric plexus. Ann Rev Physiol 1988;50:81–93.
13.Furness JB. Types of neurons in the enteric nervoussystem. J Auton Nerv Syst 2000;81:87–96.
14.Wedel T, Roblick U, Gleiss J, et al. Organization of theenteric nervous system in the human colon demon-strated by wholemount immunohistochemistry with
special reference to the submucous plexus. Ann Anat1999;181:327–337.
15.Costa M, Brookes SJ, Hennig GW. Anatomy and physi-ology of the enteric nervous system. Gut 2000;47(Suppl4):iv15–v19. discussion iv26.
16.Vanner S, Surprenant A. Neural reflexes controlling in-testinal microcirculation. Am J Physiol 1996;271:G223–G230.
17.Timmermans JP, Hens J, Adriaensen D. Outer sub-mucous plexus: an intrinsic nerve network involved inboth secretory and motility processes in the intestine oflarge mammals and humans. Anat Rec 2001;262:71–78.
19.Konturek SJ, Konturek JW, Pawlik T, et al. Brain-gut axisand its role in the control of food intake. J PhysiolPharmacol 2004;55:137–154.
20.Ward SM, Sanders KM, Hirst GD. Role of interstitial cellsof Cajal in neural control of gastrointestinal smoothmuscles. Neurogastroenterol Motil 2004;16(Suppl 1):112–117.
21.Sanders KM, Koh SD, Ward SM. Interstitial cells of Cajalas pacemakers in the gastrointestinal tract. Annu RevPhysiol 2006;68:307–343.
22.Sanders KM. A case for interstitial cells of Cajal aspacemakers and mediators of neurotransmission in thegastrointestinal tract. Gastroenterology 1996;111:492–515.
23.Thomsen L, Robinson TL, Lee JC, et al. Interstitial cells ofCajal generate a rhythmic pacemaker current. Nat Med1998;4:848–851.
24.Epperson A, Hatton WJ, Callaghan B, et al. Molecularmarkers expressed in cultured and freshly isolatedinterstitial cells of Cajal. Am J Physiol Cell Physiol 2000;279:C529–C539.
25.Langton P, Ward SM, Carl A, et al. Spontaneous elec-trical activity of interstitial cells of Cajal isolated fromcanine proximal colon. Proc Natl Acad Sci U S A 1989;86:7280–7284.
26.Lovett M, Lee K, Edwards A, et al. Vascularization stra-tegies for tissue engineering. Tissue Eng Part B Rev2009;15:353–370.
27.Rouwkema J, Rivron NC, van Blitterswijk CA. Vasculari-zation in tissue engineering. Trends Biotechnol 2008;26:434–441.
28.Yang S, Leong K-F, Du Z, et al. The design of scaffoldsfor use in tissue engineering. Part I. Traditional factors.Tissue Eng 2001;7:679–689.
29.Hori Y, Nakamura T, Kimura D, et al. Experimental studyon tissue engineering of the small intestine by mesen-chymal stem cell seeding. J Surg Res 2002;102:156–160.
30.Lee M, Wu BM, Stelzner M, et al. Intestinal smoothmuscle cell maintenance by basic fibroblast growthfactor. Tissue Eng Part A 2008;14:1395–1402.
31.Hori Y, Nakamura T, Kimura D, et al. Functional analysisof the tissue-engineered stomach wall. Artif Organs2002;26:868–872.
32.Araki M, Tao H, Sato T, et al. Development of a newtissue-engineered sheet for reconstruction of the stom-ach. Artif Organs 2009;33:818–826.
33.Nakase Y, Hagiwara A, Nakamura T, et al. Tissue engi-neering of small intestinal tissue using collagen spongescaffolds seeded with smooth muscle cells. Tissue Eng2006;12:403–412.
34.Speer AL, Sala FG, Matthews JA, et al. Murine tissue-engineered stomach demonstrates epithelial differentia-tion. J Surg Res 2011;171:6–14.
35.Madihally SV, Matthew HW. Porous chitosanscaffolds for tissue engineering. Biomaterials 1999;20:1133–1142.
36.Kim IY, Seo SJ, Moon HS, et al. Chitosan and its de-rivatives for tissue engineering applications. BiotechnolAdv 2008;26:1–21.
37.Zakhem E, Raghavan S, Gilmont RR, et al. Chitosan-based scaffolds for the support of smooth muscle con-structs in intestinal tissue engineering. Biomaterials2012;33:4810–4817.
38.Somara S, Gilmont RR, Dennis RG, et al. Bioengineeredinternal anal sphincter derived from isolated human in-ternal anal sphincter smooth muscle cells. Gastroenter-ology 2009;137:53–61.
39.de Toledo OM, Marquezini MV, Jia KB, et al. Biochemicaland cytochemical characterization of extracellular pro-teoglycans in the inner circular smooth muscle layer ofdog small intestine. IUBMB Life 2002;54:19–25.
40.Saxena AK, Kofler K, Ainodhofer H, et al. Esophagustissue engineering: hybrid approach with esophagealepithelium and unidirectional smooth muscle tissuecomponent generation in vitro. J Gastrointest Surg 2009;13:1037–1043.
41.Totonelli G, Maghsoudlou P, Garriboli M, et al. A ratdecellularized small bowel scaffold that preserves villus-crypt architecture for intestinal regeneration. Bio-materials 2012;33:3401–3410.
42.Raghavan S, Lam MT, Foster LL, et al. Bioengineeredthree-dimensional physiological model of colonic longi-tudinal smooth muscle in vitro. Tissue Eng Part CMethods 2010;16:999–1009.
43.Zhu Y, Ong WF. Epithelium regeneration on collagen (IV)grafted polycaprolactone for esophageal tissue engi-neering. Mater Sci Eng C 2009;29:1046–1050.
44.Zhu Y, Leong MF, Ong WF, et al. Esophageal epitheliumregeneration on fibronectin grafted poly(L-lactide-co-caprolactone) (PLLC) nanofiber scaffold. Biomaterials2007;28:861–868.
45.Levin DE, Barthel ER, Speer AL, et al. Human tissue-engineered small intestine forms from postnatal pro-genitor cells. J Pediatr Surg 2013;48:129–137.
46.Seliktar D, Black RA, Vito RP, et al. Dynamic mechanicalconditioning of collagen-gel blood vessel constructs in-duces remodeling in vitro. Ann Biomed Eng 2000;28:351–362.
47.Konig G, McAllister TN, Dusserre N, et al. Mechanicalproperties of completely autologous human tissue engi-neered blood vessels compared to human saphenousvein and mammary artery. Biomaterials 2009;30:1542–1550.
48.L’Heureux N, Dusserre N, Konig G, et al. Human tissue-engineered blood vessels for adult arterial revasculari-zation. Nat Med 2006;12:361–365.
49.Watters DA, Smith AN, Eastwood MA, et al. Mechanicalproperties of the rat colon: the effect of age, sex anddifferent conditions of storage. Q J Exp Physiol 1985;70:151–162.
50.Egorov VI, Schastlivtsev IV, Prut EV, et al. Mechanicalproperties of the human gastrointestinal tract. J Biomech2002;35:1417–1425.
51.Goyal RK, Biancani P, Phillips A, et al. Mechanicalproperties of the esophageal wall. J Clin Invest 1971;50:1456–1465.
52.Watters DA, Smith AN, Eastwood MA, et al. Mechanicalproperties of the colon: comparison of the features of theAfrican and European colon in vitro. Gut 1985;26:384–392.
53.Kwiatek MA, Hirano I, Kahrilas PJ, et al. Mechanicalproperties of the esophagus in eosinophilic esophagitis.Gastroenterology 2011;140:82–90.
54.Almond S, Lindley RM, Kenny SE, et al. Characterisationand transplantation of enteric nervous system progenitorcells. Gut 2007;56:489–496.
55.Suárez-Rodríguez R, Belkind-Gerson J. Culturednestin–positive cells from postnatal mouse small boweldifferentiate ex vivo into neurons, glia, and smoothmuscle. Stem Cells 2004;22:1373–1385.
56.Schafer KH, Hagl CI, Rauch U. Differentiation of neuro-spheres from the enteric nervous system. Pediatr SurgInt 2003;19:340–344.
57.Belkind-Gerson J, Carreon-Rodriguez A, Benedict LA,et al. Nestin-expressing cells in the gut give rise toenteric neurons and glial cells. Neurogastroenterol Motil2013;25:61–69.e7.
58.Silva AT, Wardhaugh T, Dolatshad NF, et al. Neuralprogenitors from isolated postnatal rat myenteric ganglia:expansion as neurospheres and differentiation in vitro.Brain Res 2008;1218:47–53.
59.Lindley RM, Hawcutt DB, Connell MG, et al. Human andmouse enteric nervous system neurosphere transplantsregulate the function of aganglionic embryonic distalcolon. Gastroenterology 2008;135:205–216.e6.
60.Metzger M, Caldwell C, Barlow AJ, et al. Enteric nervoussystem stem cells derived from human gut mucosa for thetreatment of aganglionic gut disorders. Gastroenterology2009;136:2214–2225. e1–3.
61.Rauch U, Hänsgen A, Hagl C, et al. Isolation and culti-vation of neuronal precursor cells from the developinghuman enteric nervous system as a tool for cell therapyin dysganglionosis. Int J Colorectal Dis 2006;21:554–559.
62.Pan WK, Zheng BJ, Gao Y, et al. Transplantation ofneonatal gut neural crest progenitors reconstructsganglionic function in benzalkonium chloride-treatedhomogenic rat colon. J Surg Res 2011;167:e221–e230.
June 2014 Regenerative Medicine of the GI Tract 1623
REVIEW
SAN
DPE
RSPE
CTIVES
improvement in intestinal neural function. Gastroenter-ology 2008;134:1424–1435.
65.Tsai YH, Murakami N, Gariepy CE. Postnatal intestinalengraftment of prospectively selected enteric neuralcrest stem cells in a rat model of Hirschsprung disease.Neurogastroenterol Motil 2011;23:362–369.
66.Kulkarni S, Zou B, Hanson J, et al. Gut-derived factorspromote neurogenesis of CNS-neural stem cells andnudge their differentiation to an enteric-like neuronalphenotype. Am J Physiol Gastrointest Liver Physiol 2011;301:G644–G655.
67.Micci MA, Kahrig KM, Simmons RS, et al. Neural stemcell transplantation in the stomach rescues gastricfunction in neuronal nitric oxide synthase-deficient mice.Gastroenterology 2005;129:1817–1824.
68.Dong YL, Liu W, Gao YM, et al. Neural stem cell trans-plantation rescues rectum function in the aganglionic rat.Transplant Proc 2008;40:3646–3652.
69.Metzger M, Bareiss PM, Danker T, et al. Expansion anddifferentiation of neural progenitors derived from thehuman adult enteric nervous system. Gastroenterology2009;137:2063–2073.e4.
70.Zakhem E, Raghavan S, Bitar KN. Neo-innervation of abioengineered intestinal smooth muscle constructaround chitosan scaffold. Biomaterials 2014;35:1882–1889.
71.Holloway RH, Dodds WJ, Helm JF, et al. Integrity ofcholinergic innervation to the lower esophageal sphincterin achalasia. Gastroenterology 1986;90:924–929.
72.De Giorgio R, Di Simone MP, Stanghellini V, et al.Esophageal and gastric nitric oxide synthesizing inner-vation in primary achalasia. Am J Gastroenterol 1999;94:2357–2362.
74.Raghavan S, Gilmont RR, Bitar KN. Neuroglial differen-tiation of adult enteric neuronal progenitor cells as afunction of extracellular matrix composition. Biomaterials2013;34:6649–6658.
75.Gilmont RR, Raghavan S, Miyasaka EA, et al. In situimplanted bioengineered human internal analsphincter innervated with human enteric neuronalprogenitor cells maintain myogenic and neurogenicphysiological functionality. Gastroenterology 2012;142. S-17.
76.Ruhl A. Glial cells in the gut. Neurogastroenterol Motil2005;17:777–790.
77.Cabarrocas J, Savidge TC, Liblau RS. Role of enteric glialcells in inflammatory bowel disease. Glia 2003;41:81–93.
78.Aikawa M, Miyazawa M, Okamoto K, et al.A bioabsorbable polymer patch for the treatment ofesophageal defect in a porcine model. J Gastroenterol2012:1–8.
79.Badylak SF, Vorp DA, Spievack AR, et al. Esophagealreconstruction with ECM and muscle tissue in a dogmodel. J Surg Res 2005;128:87–97.
80.Tan B, Wei RQ, Tan MY, et al. Tissue engineeredesophagus by mesenchymal stem cell seeding for
esophageal repair in a canine model. J Surg Res 2013;182:40–48.
81.Sala FG, Kunisaki SM, Ochoa ER, et al. Tissue-engi-neered small intestine and stomach form from autolo-gous tissue in a preclinical large animal model. J SurgRes 2009;156:205–212.
82.Grikscheit TC, Siddique A, Ochoa ER, et al. Tissue-engineered small intestine improves recovery aftermassive small bowel resection. Ann Surg 2004;240:748–754.
83.Grikscheit TC, Ochoa ER, Ramsanahie A, et al. Tissue-engineered large intestine resembles native colon withappropriate in vitro physiology and architecture. AnnSurg 2003;238:35–41.
84.Sala FG, Matthews JA, Speer AL, et al. A multicellularapproach forms a significant amount of tissue-engineered small intestine in the mouse. Tissue EngPart A 2011;17:1841–1850.
85.Qin HH, Dunn JC. Small intestinal submucosa seededwith intestinal smooth muscle cells in a rodent jejunalinterposition model. J Surg Res 2011;171:e21–e26.
86.Pasricha PJ, Ahmed I, Jankowski RJ, et al. Endoscopicinjection of skeletal muscle-derived cells augments gutsmooth muscle sphincter function: implications for anovel therapeutic approach. Gastrointest Endosc 2009;70:1231–1237.
87.Sung-Bum Kang M, Lee HN, Lee JY, et al. Sphinctercontractility after muscle-derived stem cells autograftinto the cryoinjured anal sphincters of rats. Dis ColonRectum 2008;51:1367–1373.
88.Raghavan S, Gilmont RR, Miyasaka EA, et al. Successfulimplantation of bioengineered, intrinsically innervated,human internal anal sphincter. Gastroenterology 2011;141:310–319.
89.Gilmont RR, Raghavan S, Somara S, et al. Bioengi-neering of physiologically functional intrinsically inner-vated human internal anal sphincter constructs. TissueEng Part A 2014. Epub ahead of print.
90.Singh J, Rattan S. Bioengineered human IAS re-constructs with functional and molecular propertiessimilar to intact IAS. Am J Physiol Gastrointest LiverPhysiol 2012;303:G713–G722.
91.Raghavan S, Miyasaka EA, Gilmont RR, et al. Perianalimplantation of bioengineered human internal analsphincter constructs intrinsically innervated withhuman neural progenitor cells. Surgery 2014;155:668–674.
92.Rattan S, Puri RN, Fan YP. Involvement of rho and rho-associated kinase in sphincteric smooth musclecontraction by angiotensin II. Exp Biol Med 2003;228:972–981.
93.Miyasaka EA, Raghavan S, Gilmont RR, et al. In vivogrowth of a bioengineered internal anal sphincter: com-parison of growth factors for optimization of growth andsurvival. Pediatr Surg Int 2011;27:137–143.
94.Chamley-Campbell J, Campbell GR, Ross R. The smoothmuscle cell in culture. Physiol Rev 1979;59:1–61.
95.Halayko AJ, Solway J. Molecular mechanisms ofphenotypic plasticity in smooth muscle cells. J ApplPhysiol 2001;90:358–368.
96.Brittingham J, Phiel C, Trzyna WC, et al. Identification ofdistinct molecular phenotypes in cultured gastrointestinalsmoothmusclecells.Gastroenterology1998;115:605–617.
97.Frid MG, Shekhonin BV, Koteliansky VE, et al. Pheno-typic changes of human smooth muscle cells duringdevelopment: late expression of heavy caldesmon andcalponin. Dev Biol 1992;153:185–193.
98.Halayko AJ, Camoretti-Mercado B, Forsythe SM, et al.Divergent differentiation paths in airway smooth muscleculture: induction of functionally contractile myocytes.Am J Physiol 1999;276:L197–L206.
99.McHugh KM. Molecular analysis of gastrointestinalsmooth muscle development. J Pediatr GastroenterolNutr 1996;23:379–394.
100.Chamley-Campbell JH, Campbell GR. What controlssmooth muscle phenotype? Atherosclerosis 1981;40:347–357.
101.Darby I, Skalli O, Gabbiani G. Alpha-smooth muscleactin is transiently expressed by myofibroblasts duringexperimental wound healing. Lab Invest 1990;63:21–29.
102.Desmouliére A, Rubbia-Brandt L, Abdiu A, et al.a-Smooth muscle actin is expressed in a subpopulationof cultured and cloned fibroblasts and is modulated byg-interferon. Exp Cell Res 1992;201:64–73.
103.van der Loop FT, Gabbiani G, Kohnen G, et al. Differ-entiation of smooth muscle cells in human blood vesselsas defined by smoothelin, a novel marker for the con-tractile phenotype. Arterioscler Thromb Vasc Biol 1997;17:665–671.
104.Niessen P, Rensen S, van Deursen J, et al. Smoothelin-ais essential for functional intestinal smooth musclecontractility in mice. Gastroenterology 2005;129:1592–1601.
105.Shanahan CM, Weissberg PL. Smooth muscle cell het-erogeneity: patterns of gene expression in vascular
smooth muscle cells in vitro and in vivo. ArteriosclerThromb Vasc Biol 1998;18:333–338.
106.Liddell RA, Syms M, McHugh KM. Heterogeneous iso-actin gene expression in the adult rat gastrointestinaltract. Gastroenterology 1993;105:347–356.
107.Sobue K, Hayashi K, Nishida W. Expressional regulationof smooth muscle cell-specific genes in association withphenotypic modulation. Mol Cell Biochem 1999;190:105–118.
108.Shanahan CM, Weissberg PL, Metcalfe JC. Isolationof gene markers of differentiated and proliferatingvascular smooth muscle cells. Circ Res 1993;73:193–204.
109.Davis BN, Hilyard AC, Nguyen PH, et al. Inductionof microRNA-221 by platelet-derived growth factorsignaling is critical for modulation of vascularsmooth muscle phenotype. J Biol Chem 2009;284:3728–3738.
110.Bourgier C, Haydont V, Milliat F, et al. Inhibition of Rhokinase modulates radiation induced fibrogenic pheno-type in intestinal smooth muscle cells through alterationof the cytoskeleton and connective tissue growth factorexpression. Gut 2005;54:336–343.
Received October 31, 2013. Accepted March 20, 2014.
Reprint requestsAddress requests for reprints to: Khalil N. Bitar, PhD, AGAF, Wake ForestInstitute for Regenerative Medicine, 391 Technology Way, Winston-Salem,North Carolina 27101, Winston-Salem, North Carolina 27101. e-mail:[email protected]; fax: (336) 716-2011.
AcknowledgementsThis work was supported by National Institutes of Health (RO1 DK 071614).
Conflicts of interestThe authors disclose no conflicts.
