Mucosal defense: gastroduodenal injury and repair mechanisms

Abstract

Purpose of Review:

The mucosal barrier serves as a primary interface between the environment and host. In daily life, superficial injury to the gastric or duodenal mucosa occurs regularly but heals rapidly by a process called “restitution”. Persistent injury to the gastroduodenal mucosa also occurs but initiates a regenerative lesion with specific wound healing mechanisms that attempt to repair barrier function. If not healed, these lesions can be the site of neoplasia development in a chronic inflammatory setting. This review summarizes the past year of advances in understanding mucosal repair in the gastroduodenal mucosa, which occurs as a defense mechanism against injury.

Recent Findings:

Organoids are an emerging new tool that allows for the correlation of in vivo and in vitro models; organoids represent an important reductionist model to probe specific aspects of injury and repair mechanisms that are limited to epithelial cells. Additionally, proof-of-concept studies show that machine learning algorithms may ultimately assist with identifying novel, targetable pathways to pursue in therapeutic interventions. Gut-on-chip technology and single cell RNA-sequencing contributed to new understanding of gastroduodenal regenerative lesions after injury by identifying networks and interactions that are involved in the repair process.

Summary:

Recent updates provide new possibilities for identifying novel molecular targets for the treatment of acute and superficial mucosal injury, mucosal regeneration, and regenerative lesions in the GI tract.

INTRODUCTION

The ability of stomach mucosa to heal rapidly after superficial injury was first described in frog mucosa [1] and was subsequently reproduced in guinea pig, rat, and mouse mucosa [2,3]. Later, this process was also noted to occur in duodenal and colonic mucosa. Termed “restitution,” this healing ability is an important aspect of mucosal defense against superficial injury in gastrointestinal (GI) tract tissues. Restitution occurs after damaging substances exfoliate cells at the mucosal surface and the wound is rapidly repaired (within 15 min to 1 h) by cells at the wound edge that migrate to cover the denuded area without cell proliferation [2]. The initial wound is accompanied by high mucosal to serosal flux, a drop in mucosal resistance, and disruption of ion transport mechanisms, which are normalized as the epithelial barrier is reestablished [4–6]. Many papers have been published in the past year that utilize ex vivo tissues and/or cell culture systems to highlight new substances that cause surface wounds or facilitate restitution. Notably, several investigators have used the mucoadhesive biopolymer Chitosan as a delivery platform to load compounds that are released in an acid environment [7]. Chitosan containing allantoin [8], curcumin [9], selenium [10], or anthrocyanin [11] showed very good protective profiles against mucosal injury in the stomach.

Although more difficult to study experimentally, superficial injury also occurs in human mucosa as highlighted in a recent case report “Lesson of the Month” [12], whereby a patient presented in the emergency room with recurrent vomiting, abdominal pain, and hematemesis as the result of extensive superficial injury to both stomach and duodenal mucosa. It was subsequently determined that the lesions occurred from drinking a sub-lethal amount of ethanol-based hand disinfectant for fear of contracting severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Shortly after withdrawal of the damaging agent, both stomach and duodenal mucosa returned to a normal appearance and the symptoms subsided [12].

Persistent injury, which is often associated with chronic inflammation, requires a different set of mucosal defense mechanisms to restore barrier integrity. Rather than repair by restitution, these lesions heal after forming a regenerative lesion by mechanisms that are not well understood. This process requires signals from mesenchyme and immune cells, making the entire network important for full regeneration of the mucosa after injury.

SUPERFICIAL INJURY AND RESTITUTION

Numerous models have been used to study restitution including ex vivo models, Ussing chambers, in vivo models using exteriorized mouse stomach or intestine, cultured gastric or intestinal cells, and most recently, organoids made from isolated gastric (gastroids) or intestinal (enteroids) epithelial cells.

Organoid models to study restitution

Epithelium-derived organoid models allow investigators to conduct more experiments compared to tissues from individual animals. Additionally, information derived from organoid experiments may be more representative compared to cultured cells from cancer sources. But organoids contain many different cell types with few mature cells. As a result, there may be significant variability in the data, so large datasets are required. Despite this limitation, organoids are emerging as a valuable reductionist model with which to study restitution. Gastroids were recently validated to respond to injury similarly to in vivo models [13,14].

Calcium (Ca²⁺) and restitution

Initial studies to assess the requirements for restitution utilized Ca²⁺-free solutions (or solutions with EDTA to chelate extracellular Ca²⁺) in frog mucosa. These experiments found that that both physiologic and morphologic restitution was impeded in the absence of Ca²⁺ [15]. Cells migrated in the absence of Ca²⁺ to cover most, but not all, of the wounded area after injury, but did not re-form tight junctions so that permeability remained high [15]. To revisit the role of calcium in restitution using molecular technology and an in vivo mouse stomach model, Aihara et al used transgenic mice expressing Yellow Cameleon 3.0 protein as a reporter of intracellular calcium, fura-red dye to localize extracellular Ca²⁺, and membrane-associated Ca²⁺-ATPase 1 and 2 transgenic mice. They demonstrated that endogenous Ca²⁺, mobilized by cell signaling, and calcium transport via membrane-associated Ca²⁺-ATPase 1, are essential for restitution to occur [16]. Additionally, studies using Fura-red showed that extracellular Ca²⁺ next to the injured cells assisted with wound repair after injury [16]. Like the frog studies [15], chelation of intracellular or extracellular Ca²⁺ impeded but did not inhibit restitution [16].

The role of Ca²⁺ in restitution was also studied in gastroids made from mice that express an intracellular membrane-associated Ca²⁺ reporter, Yellow Cameleon-Nano15 (YC Nano) [13]. In comparing the mRNA signature for calcium-related genes from normal and ulcerated mouse stomach with gastroids from control (non-transgenic) mice, some overlap was found. Pathways from this unique set of genes were used to evaluate the role of Ca²⁺ in restitution using gastroids from the YC Nano reporter mice [14]. The pathways chosen were 1) voltage-gated Ca²⁺ channels, 2) phospholipase C, 3) inositol 3-phospate receptors, and 4) store-operated channels represented by Stim1 and Ori1, with the strategy to evaluate repair of the injured gastroid after blockade of these pathways with pharmacological inhibitors [14]. All pathways were needed for restitution to occur in the gastroid model [14].

Pathways inhibited with pharmacological mediators may be non-specific; utilizing organoids from pathway knockout and/or knock-in mice could improve specificity for drug-inhibitor experiments on epithelial-based functions. This was the strategy used to examine the role of trefoil factor 2 (Tff2; also known as spasmolytic polypeptide) in restitution, which uncovered a link between Tff2 and Ca²⁺ signaling in regulating restitution after injury in the stomach [17]. Tff2 peptide activated AKT signaling (a Ca²⁺-mediated cellular signaling pathway) in a rat model of stress-induced injury [18]. Trefoil factor peptides increased the rate of migration of gastric and intestinal epithelial cells after wounding [19]. Tff2 peptide was also shown to be a Muc6 binding lectin that requires Ca²⁺ to stabilize the mucus/bicarbonate barrier [19], which also facilitates restitution and is required for gastroduodenal mucosal protection in vivo [20]. Thus, studies using organoid models lent important insights into mechanisms specific for epithelial cells in comparison to those of other essential aspects of restitution that are found in ex vivo or in vivo models.

RNA-binding proteins and restitution

To evaluate the role of human antigen R (HuR; Elavl1 gene) in restitution, Elavl1 was genetically silenced in intestinal IEC-6 cells and in vivo in Elavl1 conditional knockout (KO) mice, both of which showed a significant down-regulation of the vitamin D receptor (VDR; Vdr gene) [21]. In the gastric epithelium, little is known about the role of VDR in cellular functions, particularly in mucosal defense mechanisms [22]. In contrast, VDR is a highly expressed nuclear receptor in the intestinal epithelium that mediates vitamin D3 functions by regulating downstream target proteins important for mucosal homeostasis.

Experiments using IEC-6 cells demonstrated that silencing Vdr inhibited the repair of wounded epithelial cells and that the response was dependent on HuR expression [21]. Mechanistically, HuR interacted with the 3’-untranslated region (UTR) of Vdr mRNA and enhanced VDR-mediated translation in IEC-6 cells. Vdr silencing inhibited the migration of IEC-6 cells in response to wounding but proliferation was not affected [21].

Recent work also demonstrated that silencing HuR inhibits the expression of caveolin-1, which inhibited restitution in IEC-6 cells after injury [22]. In contrast to nuclear VDR (nVDR) signaling, membrane-associated VDR (mVDR) signaling is linked to caveoli [23]. mVDR’s are associated with a rapid, non-genomic response to vitamin D3 including the liberation of Ca²⁺ and the activation of numerous cell signaling pathways [23], which also regulate restitution. Further research is needed determine the relationship between these HuR-mediated events on restitution and whether the effects of HuR are specific for intestinal epithelial cells or also regulate restitution of the stomach mucosa after superficial injury.

Dynamic modeling and machine learning

Restitution is a dynamic process that depends on well-timed interactions between the damaged cells, neighboring cells, and underlying matrix/substratum to facilitate rapid repair of the wounded area. One emerging area in the field is to provide data that are interrogated by machine learning aimed to provide new insights into how the repair process is accomplished. Proof-of-concept work was presented recently to determine mechanisms that underlie actomyosin dynamics in restitution [24], using experimental data from organoids subjected to a laser-induced wound [25]. Using a defined computational pipeline, simulations were modeled to recapitulate actual data. The modeled simulations patterned well after the original data, and provided new insight that suggested actin polymerization timing plays a critical role in determining the behavior of cells involved in the repair process after injury [24].

PERSISTENT INJURY AND MUCOSAL REGENERATION

Persistent injury to the gastroduodenal mucosa challenges mucosal defense mechanisms and results in deeper lesions including ulcers in the stomach and duodenum, intestinal metaplasia in the stomach, and spasmolytic polypeptide-expressing metaplasia in the stomach. Although many models have been used to examine the pathogenesis of lesions occurring after persistent injury, new on the forefront is gut-on-chip technology. Gut-on-chip models contain mechanically active, continuously perfused microchannels inhabited by different human intestinal cell populations that form tissues with in vivo-like morphology, which can be functionally interfaced with each other. This technology was highlighted in a recent study of SARS-CoV-2 infection using a human intestinal model in vitro, which demonstrated direct evidence for viral disruption of villus structure, epithelial barrier integrity, mucous secretion, and the activation of immune responses [26]. The gut-on-chip model was amenable to RNA-sequencing and immunostaining analysis of the epithelium after injury so that granular information on pathogenesis could be obtained, but also allowed for probing injury responses and interactions of the injured epithelium with other cellular compartments [26] that cannot be done using organoids alone. Gut-on-chip technology is now available to use with mouse cells, where it is possible to grow mouse intestinal villi and crypts using cells from normal or transgenic mice [27]. Although it will be necessary to figure-out how to get the mouse cells to pattern villus and crypt compartments properly, with cell types in the correct spatial position, it will be exciting to see this new model progress toward mechanistic studies of acute and persistent injury and repair in the near future [28].

Gastroduodenal ulcers

An update to the clinical guidelines for management of peptic ulcer disease was recently published [29]. In addition to an extensive algorithm to assist with ulcer treatment in the clinic, consideration was highlighted from studies in Japan concerning the use of vonoprazan fumarate (P-CAB; Takada) to treat ulcer disease [29]. This acid-blocking drug was reported to effectively heal peptic and duodenal ulcers, and prevent the recurrence of NSAID-induced and low-dose aspirin-induced ulcers [29]. However, more global utilization and reporting on the efficacy of vanoprazan is needed to understand its full benefits and contraindications [30].

Even when gastroduodenal ulcers appear to have healed histologically, they have a high rate of recurrence [31,32], because the repaired epithelium is not normally restored [33]. One ion transport mechanism that is slow to recover after gastric ulceration is the sodium/hydrogen exchanger 2 (NHE2; Slc9A2 gene), which is required for intracellular pH (pHi) regulation in the mouse stomach mucosa during restitution [34]. Slc9A2 KO mice fail to recover from acute injury to the gastric or intestinal epithelium ex vivo [34, 35], and the lack of transporter expression for months after the initial injury was proposed to contribute to re-injury and ulceration in the stomach mucosa [34]. In the intestinal mucosa from Slc9A2 KO mice, full recovery of the intestinal morphology after injury appeared to occur but paracellular permeability defects were sustained [35]. Unfortunately, this aspect of the repair process was not investigated in stomach mucosa [34]. With organoids and with gut-on-chip technology in place for both human and mouse, it will be possible to determine whether NHE2 regulates pHi in human stomach after wounding, whether repair mechanisms are similar between human and mouse, and whether a link exists between barrier function, tight junctions, and NHE2 in the regenerating stomach mucosa.

It is not surprising that repair mechanisms for duodenal ulcers are more challenging, since ulceration usually occurs in areas of gastric metaplasia that have high acid output and are sites of H. pylori infection with chronic duodenitis [36]. In re-evaluating the relationship between gastric metaplasia and duodenal ulcers, a recent study characterized markers for gastric metaplasia compared to heterotopic gastric mucosa (HTG) to show that 55% of duodenal ulcers had HTG, and 59% of HTG tissues were also positive for H. pylori [37,38]. Within the ulcerated area, gastric metaplasia per se did not seem to heal after H. pylori eradication, despite resolution of inflammation. Additionally, the use of acid-blockers for acid hypersecretion did not seem to influence the extent of gastric metaplasia [36]. Whether HTG subsequently develops into gastric metaplasia, particularly in an inflammatory environment, was not investigated but may lend new insight into adenocarcinoma development in duodenal mucosa [39,40]. Furthermore, comparison of low-grade versus high grade neoplasms in the duodenum revealed that the vast majority of high-grade/submucosal tumors were gastric-type tumors [41]. Current thinking about the pathway to gastric-type duodenal tumors also suggest that they begin with gastric metaplasia and progress through a mutation pathway to duodenal gastric-type adenocarcinoma [42], although this proposed pathway needs to be supported experimentally. Overall, these data suggest that areas of duodenal ulcers with gastric metaplasia fail to fully repair and are prone to further ulceration, but may also be sites of tumorigenesis by a proposed mutational program in the duodenum.

Spasmolytic polypeptide-expressing metaplasia (SPEM)

Unlike superficial injury that involves surface epithelial cells, acute or chronic injury involving the death of gastric parietal and/or chief cells causes a unique metaplastic lesion expressing wound healing markers including Tff2 [43–48]. SPEM is a critical acute phase program required for tissue reconstitution following injury. However, new transcriptional information also suggests that during the repair process, effectors are expressed that have a similar gene expression signature found in cancer [49], suggesting that SPEM lesions may represent potential cancer drivers. Studies this year continued to divide investigators concerning the role of mature chief cells in the formation of SPEM. For this, studies evaluating paligenosis (i.e. – reversion of differentiated chief cells to proliferative metaplastic cells) clearly suggest that mature chief cells transdifferentiate into SPEM [50–54]. Conversely, several rigorous publications argued that chief cells die rather than transdifferentiate into SPEM cells and that expansion of stem cells to mucous neck cells constitute SPEM [55–56]. Although it is unclear how the ongoing debate will be settled, it is clear that new technology and data are pushing a re-definition of SPEM, its origin, cells involved in repair after injury, and the transcriptional program of SPEM cells.

CONCLUSIONS

Injury to the gastroduodenal mucosa can be superficial, affecting surface/pit epithelial cells, or persistent, causing injury/death of epithelial cells deeper in the mucosa. Each injury heals differently. Recent studies that utilized genetically-modified mice, organoid models, and RNA sequencing have provided new information on mechanisms that promote epithelial restitution after superficial injury. Proof-of-concept studies were successful in demonstrating the feasibility of modeling restitution by novel machine learning mechanisms.

Repair processes for deeper wounds that occur after acute or persistent injury also provided a wealth of new information in the past year. Gut-on-chip technology has the potential to significantly enhance mechanistic investigations of gastroduodenal injury and repair. Experimental results suggested that wounded areas do not reconstitute and thus remain sites of further injury and ulceration. Furthermore, unhealed mucosa show metaplastic changes in epithelial constitution and may be sites of cancer formation. Investigations in the past year on stem cells in stomach mucosa also challenged current paradigms and provided new ideas for mechanisms of reconstitution after injury and cellular loss. The controversy about chief cells and their direct role in the formation of regenerative lesions after injury remained strong, with compelling work to both support and to refute chief cell trans-differentiation in the process.

Key points.

• Gastric and duodenal mucosa are subject to superficial injury that heals by restitution, which requires both endogenous and transported calcium, is regulated by RNA-binding proteins, and can be studied in vivo, ex vivo, or in organoids made from transgenic mice or humans.

• Gastric and duodenal mucosa subjected to persistent injury heals by forming a regenerative lesion that only partially recovers lending to repeat injury and ulceration.

• New insights from studies in the past year highlight re-thinking stem cell lineages and which cells contribute to mucosal regeneration after injury.

• Understanding the mechanisms and pathways involved in mucosal defense after superficial or persistent injury may open the door for discovery of novel therapeutic targets and agents to prevent mucosal injury and promote mucosal repair in the gastroduodenal mucosa.