Mechanisms and modeling of wound repair in the intestinal epithelium

ABSTRACT

The intestinal epithelial barrier is susceptible to injury from insults, such as ischemia or infectious disease. The epithelium’s ability to repair wounded regions is critical to maintaining barrier integrity. Mechanisms of intestinal epithelial repair can be studied with models that recapitulate the in vivo environment. This review focuses on in vitro injury models and intestinal cell lines utilized in such systems. The formation of artificial wounds in a controlled environment allows for the exploration of reparative physiology in cell lines modeling diverse aspects of intestinal physiology. Specifically, the use of intestinal cell lines, IPEC-J2, Caco-2, T-84, HT-29, and IEC-6, to model intestinal epithelium is discussed. Understanding the unique systems available for creating intestinal injury and the differences in monolayers used for in vitro work is essential for designing studies that properly capture relevant physiology for the study of intestinal wound repair.

KEYWORDS:

• IPEC-J2 cells
• transepithelial electrical resistance
• injury barrier function
• cell model

Introduction

The intestinal mucosa is lined by a single layer of epithelium that serves as a crucial barrier between the gut lumen and the body. Numerous insults are known to disrupt the epithelial barrier. For example, SARS-CoV2 (also known as COVID-19) has profound effects on the intestinal epithelium. This is believed to be in part due to the affinity SARS-CoV2 S for the angiotensin converting enzyme 2 receptors (ACE2) which is extensively expressed on the apical surface of small intestinal epithelium. While the exact mechanism is not entirely clear, the ability of SARS-CoV2 to damage the intestinal mucosa and result in increased permeability may result in the onset of diarrhea and malabsorption.¹ Numerous other infectious diseases, such as Rotavirus, also disrupt epithelial barrier function, and disease processes such as ischemia/reperfusion are known to result in epithelial lifting from the basement membrane, and loss of epithelial monolayer integrity.^2–6 Therefore, a crucial component of epithelial barrier function is its ability to rapidly repair using a mechanism called restitution. This involves cell spreading and extension of lamellipodia in order to bridge the wound that results from sloughing of epithelium⁷ Reductionist models of restitution have been employed to develop a mechanistic understanding of restitution, including in vitro cell models. However, restitution remains incompletely understood, and this is partly related to the anatomical and physiological complexity of the intestinal mucosa and the limitations of cell models and methods of this reductionist barrier modeling. The wounding process itself can also present a technical dilemma. For instance, wounding typically involves ‘scratching’ in a mechanical fashion, but this differs from in vivo wounding more typically attributable to inflammation from infectious disease or hypoxia and inflammation induced by ischemia/reperfusion. Therefore, this review outlines different models of epithelial wounding and the characteristics of the cell lines used to aid investigators in designing optimal epithelial wounding models for the study of barrier restitution.

The intestinal barrier

The epithelial barrier functions as a physical boundary between luminal contents and the remainder of the body while also serving as a filter for crucial nutrients, ions, and water. Effective barrier function is therefore essential to maintaining intestinal homeostasis and this should be recapitulated in in vitro intestinal modeling systems. Tight junctional proteins bridge across the paracellular space of subjacent epithelial cells and are largely responsible for controlling the flow of ions and macromolecules between cells and create measurable differences in transepithelial electrical resistance (TEER) and macromolecular flux. However, tight junctional proteins retain their own unique size-selective properties which are classified as either pore or leak pathways. Pore pathways are defined as those permeable to molecules with a radius of ~4 Å or less, whereas leak pathways are defined as allowing the passage of large non-charged solutes.⁸ To accurately assess the integrity of these pathways, methods to assess TEER have been developed for use in cells. One pitfall of using TEER to assess wound healing is that both transcellular and paracellular transport contributes to this measurement, whereas macromolecular fluxes of labeled macromolecules, such as FITC-dextran, are more indicative of changes to the paracellular space.⁹ Therefore, to avoid confusion as to whether cellular crawling or closure of tight junctions is responsible for differences in barrier function, restitution studies continue to largely rely on the area of epithelial wound coverage.

Mechanisms of barrier repair

In the small intestine, following barrier injury, there are three phases of wound healing: villus contraction, cell migration (restitution), and paracellular space closure.⁷ Given that cell migration is a major component of the reparative response to injury, different wound models have been developed to closely examine this process during the restitution of epithelial barrier defects. Epithelial restitution utilizes the ability of epithelial cells adjacent to injured portions of mucosa to flatten and migrate over the intact basement membrane following epithelial loss attributable to insults, such as infectious disease, or ischemia.¹⁰ This process is accomplished when cell–cell and cell-matrix adhesions are temporarily remodeled, allowing for cells to extend forward and cover denuded surfaces and can occur rapidly before the onset of increased enterocyte proliferation to ultimately replace lost epithelial cells.

Originally, investigators focused on the mechanisms that initiate restitution. One pivotal study in epithelial restitution performed by Nusrat et al. evaluated cellular migration, while cell monolayers recovered from scratch wounding in vitro. The study ultimately concluded that cell flattening, spreading, and formation of lamellipodia-like protrusions were among the first steps involved in restitution.¹¹ However, a greater understanding of the mechanisms of restitution is needed to screen new potential therapeutics and extend the findings to the clinic. One study utilized IEC-6 cells in order to evaluate the effects of various cytokines on wound closure, and found enhanced epithelial restitution when cells were exposed to TGF-α, EGF, and IL-1β, whereas there was no change in migration of cells surrounding wounds in the presence of IL-6, TNF-α, PDGF, and lipopolysaccharide.¹² Other studies have focussed on the role of hypoxic conditions at the site of mucosal injury. Activation of hypoxia-inducible factor-1α (HIF-1α) was found to initiate protection of the mucosal barrier to allow for epithelial wound repair.^13–15 Specifically, one study found an increase in expression of fibroblast integrin β1 (ITGB1) through transcriptional mechanisms dependent on HIF, which is critical to epithelial wound restitution.¹⁶ In contrast to previous studies, cytokines such as IFN-γ have been studied in relation to integrin-mediated restitution and were found to have an inhibitory effect on restitution as a result of dysfunction in the F-actin lamellipodia on T84 cells at the edge of epithelial wounds, raising further questions about the implications of inflammatory cytokines and precursors on restitution.¹⁷ While these are only some areas of study surrounding restitution mechanics, much remains still unknown. Therefore, the restitution process continues to be studied extensively using scratch wounding models, cell exclusion assays, neutrophil-induced epithelial wounding models, and electrical wound-healing assays to study healing events, and the mechanisms that regulate cell crawling.

Following restitution, wound healing is completed by resealing of tight junctions. Select tight junction proteins have been studied to refine our understanding of the reparative mechanisms of tight junctions. For example, claudins have been investigated and found to have seemingly complementary functions in barrier function, with some claudins acting as “leaky” or pore-forming claudins, and others as “tight” sealing claudins.¹⁸ Works evaluating the ClC-2 pathway in ischemic-injured murine jejunum have shown a critical role for claudin-1 and occludin in resealing tight junctions following epithelial restitution.¹⁹ Additionally, an in vitro study using Caco-2 cells overexpressing ClC-2 resulted in a decrease of the pore-forming claudin-2 protein while maintaining claudin-1 and claudin-4 expression.²⁰ Other tight junction proteins have been studied as it appears their reassembly plays a role in epithelial repair; however further investigation is needed to fully understand the extent of their role in healing.

InVitro injury models

A number of in vitro cellular wounding and repair models have been created to assess epithelial restitution, all of which have strengths and weaknesses when attempting to recapitulate intestinal wound repair in clinical patients (Table 1). In addition, models to explore the more complex effects of multiple cell types on wound repair have also been studied, particularly the role of epithelial neutrophil transmigration, in order to provide more physiologically relevant models.

Table 1. Comparison of in vitro injury models.

Injury model	Advantages	Disadvantages	Notes
Scratch Assay	Can be used on any cell line or primary culture able to form monolayers Technically simple Ability to adjust size and location of wound	Inability to standardize each scratch, creating variability Can disrupt ECM	Concerns remain regarding the ability to differentiate between cell migration, restitution and proliferation. Edu labeling can be used to clarify results
Cell Exclusion Zone Assay	Does not actively injure monolayer, therefore does not damage ECM Uniform pattern, standardized	Because it is a passively created hole, some do not consider this a ‘wound’	Only used to evaluate migratory or cell ‘crawling’ ability
Neutrophil wounding models	Ability to evaluate secondary effects at the site of injury other than the physical wound Able to manipulate chemoattractant levels as well as oxygen level	Only evaluates wound formation and repair specific to neutrophil transmigration	Method can be used to examine mechanisms of interepithelial junction repair
Electric cell-substrate impedance sensing	Allows for direct observation of wound healing events Allows of electrophysiological quantification of each phase of repair	Heat from electrical current can cause additional damage Density of monolayers can complicate measurement of impedance	Opportunity to alter wound severity
Organoid models	Enables study of the reparative process with crypt and villus domains 3-D structure comparable to in vivo environment	Difficult to access apical surface of epithelium Can be difficult to assess repair due to the complex structure	Is able to repair post- irradiation

Scratch assay

Scratch wounding is one of the most frequently utilized methods to study barrier restitution in vitro. It offers a simple procedure to induce cell crawling similar to that which happens in vivo post injury. Due to its simplicity, it can be used on any cell line or primary culture that can successfully be grown into a monolayer. This method involves growing confluent monolayers, scratch wounding, and wound measurement. Scratch processes differ by lab, but the two most popular methods are scratches formed by razor blades and those created by pipette tips.^21–31 Typically, razor blades are carefully pressed into the confluent monolayer of cells, while pipette tips are dragged across it (Figure 1). More advanced methods of scratching have been developed to standardize the scratch. One technique resembles a multichannel pipette but has guiding rails on the outside to line up with the plate. This method scratches eight wells at once and has adjustable hex screws to set varying angles.³² Another method does not use any pipette tips, but instead uses a 96 floating-pin array. The operator sets the device on top of the plate, engages the pins, and drags the device toward one corner of the plate scratching 96 wells at once.³³ These advancements have helped to standardize scratches within experiments.

Figure 1 Diagrammatic representation of the use of epithelial monolayers for studies of injury and repair. a. Scratch wounding is accomplished with a pipette tip. b. The cell wound exclusion assay involves creation of a ‘wound’ by growing cells in the presence of a silicone stopper.

Initially, cells are grown to confluence, after which monolayers are either serum deprived or serum reduced from anywhere between 18 and 48 hours before the wound is inflicted. The level of serum present depends on the type of cell. Transformed cell lines can accommodate complete serum deprivation, while primary cells continue to need small amounts of serum to function properly.³⁴ The intention of serum deprivation is to coerce the cells into a quiescent state, so that cell proliferation is not mistaken for cell migration. Although serum deprivation is a standard step in many scratch wound assays, serum deprivation is somewhat controversial. In some cell lines, it can cause unpredictable effects such as phenotypic changes or varying responses to certain stimuli.³⁵ Given this controversy, researchers need to decide whether serum deprivation or reduction is a necessary component of their scratch wound assay in order to optimally study restitution. In addition, proliferation assays like nuclear labeling with Edu incorporation may be used to ensure rapid proliferation into the wound bed is not being misinterpreted as purely cellular migration.³⁶ Edu labeling has been utilized in place of BrdU as it uses a simple chemical reaction instead of DNA denaturation in order to analyze cellular proliferation, population homeostasis, and cell marking procedures.³⁷

Once the monolayers are injured, the wells can be washed to ensure there are no remaining cells in the scratched area. Immediately following this, the cells and the scratch are imaged to assess the size of the wound. The wells are then imaged at different time intervals to record the migration process. Quantification of wound closure can also vary; some researchers choose to do a simple count of nucleated cells across the wound margins,^24,25,27,28 while others use image analysis software to compare the area of the scratch before, during, and after cell migration.^{7,22,26,38,39} Overall, the process of scratch assays is relatively straightforward and does not require advanced technology.

Due to the simplicity of this process, it has been used for many animal cell line studies to determine whether different growth factors,^23,25,38,40 cytokines,²⁷ mRNA proteins,²⁶ nucleic acids,^28,39 or supplementations^22,24 improve the epithelial restitution by enhancing cell migration. For example, when TGF-β was added to medium, although it is known for the inhibition of proliferation, it actually sped up the migration process post scratching.²⁵

Although scratch wounding is both quick and simple, it does have its drawbacks. In order for comparison among studies and among samples to be made, the scratches need to be of similar length, width, and depth. There is extensive variability between scratching tools and techniques. Even within the same procedure, variation occurs due to varying operator dexterity patterns (pressure, angle, length) between each scratch. However, additional tools previously mentioned have been developed to minimize these discrepancies. Another problem with scratch assays is that the scratching instrument disrupts more than just the cells. Specifically, most experiments require culture surfaces to be coated with components of the extracellular matrix (ECM). This matrix is often also damaged during wound infliction. This damage can affect the migration of the cells and adds yet another variable impacting restitution.⁴¹ In order for scratch assays to be compared, standard procedures including scratch methods and coating materials should be established.

Cell exclusion zone assay

Given its injurious nature to epithelial cells, the scratch wounding method has a multitude of limitations that may influence restitution such as an influx of intracellular fluid or ECM damage.⁴² Therefore, the cell exclusion zone assay was created to analyze cell crawling in response solely to neighboring barren space, thereby eliminating some of the confounding factors generated by physically wounding a monolayer. Cell exclusion models use patterned cell seeding, created by molds that are placed on the surface before the cells are seeded (Figure 1). Before this assay was even named, the concept was first used in 1984 to study cytoskeletal regulation in aortic endothelial cells.⁴³ In that experiment, a Teflon fence was used to create the cell barren areas. The experiment that coined the name ‘Cell Exclusion Zone Assay’ used a microfabricated Polydimethylsiloxane (PDMS) mold with rectangular holes in it.⁴² The mold was placed on ECM coated plastic, and cells were seeded into the holes. The cells were grown to confluence, and the molds were peeled off. This resulted in geometrically perfect ‘wounds’ into which the cells began crawling. The migration rate or area coverage was captured microscopically and then analyzed. This method has been utilized amongst a wide variety of epithelial cell crawling studies involving kidney cells,⁴² esophageal cells,⁴⁴ and bronchial cells.⁴⁵ Due to the fast-healing effects of the intestine, this assay was also optimized in the intestine. In this study, researchers employed the cell exclusion zone model to study cell crawling of fetal small intestine (FHs-74 int) cells.⁴⁶ Silicone stoppers were used to create the cell barren spaces, after which cell crawling was microscopically examined. This model was compared to the standard scratch wounding assays, with the key finding that the cell exclusion zone assay created a more uniform wound with no damage to the underlying ECM components or leakage of intracellular contents as would occur during scratching, making the quantification of cell crawling easier and more accurate.⁴⁶

However, this model does not perfectly recreate in vivo wound healing. Researchers have been hesitant to call this cell excluded ‘wound’ an actual wound because of the non-injurious way that it is created. Nevertheless, this model is still thought to be relevant because cell migration is one of the biggest components of epithelial restitution, and this assay has helped to narrow down mechanisms of cell crawling. For example, researchers using a cell exclusion zone assay with canine kidney cells verified that the cells did not need to be damaged in order to migrate. They found that exposing the cells to barren space was enough to trigger their movement. This finding supports the hypothesis that there is a mechanical type of communication between cells and exposed matrix, which induces migration.⁴²

Neutrophil wounding models

Intestinal disease processes, such as ischemia/reperfusion injury, inflammatory bowel disease, and infectious disease that damages intestinal epithelium, are all known to attract significant numbers of neutrophils to the sites of epithelial injury.^6,47 This is in part due to neutrophilic chemoattractants accumulating in the area when oxidants are released by the damaged tissue. For example, LTB₄ is generated from oxidation of lipid membranes, and is a potent neutrophil chemoattractant. As neutrophils are chemoattracted to the basal side of epithelium via chemoattractant gradients, they continue to migrate through paracellular pathways, disrupting interepithelial junctional structures, ultimately entering the intestinal lumen. This process is regulated in part by CD44v6 and CD55 as well as ICAM-1.⁴⁸ As the number of neutrophils migrating across the epithelium increases, paracellular spaces are initially widened, followed by lifting and loss of epithelium, with wound formation.⁴⁹

To model this process in vitro, advances in existing cell motility models have occurred to observe neutrophilic migration across epithelial monolayers. Notably, Brazil et al. developed an advanced model for investigating mediators associated with increased neutrophil transmigration and luminal migration. To achieve this, T84 monolayers were grown on collagen-coated transwells, and a chemotactic gradient was created by adding 100 nM formyl-methionyl-leucyl-phenylalanine to the media on the apical side of a cell monolayer in order to establish a basolateral to apical direction of migration. Next, 1 × 10⁶ polymorphonuclear leukocytes (PMN; (neutrophils) were added to the upper chambers of transwells (in order to achieve technical simplicity and accuracy, application of neutrophils to the apical surface was adopted), and transmigrated PMNs were measured by a colorimetric enzyme activity assay specific for the PMN azurophilic marker myeloperoxidase (MPO). The resulting round wound was observable by using the epithelial desmosomal marker, desmoglein.⁵⁰ This technique has been further employed to investigate the role of interepithelial junctional complexes in signaling and guiding neutrophil transmigration. For example, a study evaluating junctional-adhesion molecule like protein (JAML) found that binding of JAML to epithelial coxsackie and adenovirus receptor (CAR) plays a role in the regulation of neutrophil transmigration. The bound JAML/CAR fusion proteins specifically inhibited neutrophil migration when introduced into PMN-rich T84 monolayers. While this model can be used to represent wounding following transmigration of PMNs, the restitution that follows is markedly different from that described in other monolayer wound assays. In particular, the wound develops a rim of actinomyosin that appears to help close the wound by contraction, rather than relying solely on epithelial migration.⁵¹

Electric cell-substrate impedance sensing

Modern advances have sought to alleviate the problems of inconsistency noted in scratch wounding systems of modeling. This has led to the development of in vitro systems that are able to monitor monolayer behavior after wounding the epithelial cells via high electrical current pulses. The electric cell-substrate impedance sensing (ECIS) system utilizes gold film electrodes that are inserted into the bottom of cell culture plates.⁵² Cell monolayers are then established and grown on the surface of this gold film electrode, with a larger counter electrode inserted into the culture media in order to create an electrical circuit. In addition to wounding the cell monolayer, the electrode is then able to quantify and chart changes in impedance resulting from cellular behavior as they migrate to heal the wound.⁵³ This technique allows for the investigator to manipulate the wound severity by altering the duration and current strength to fit study parameters.

One study seeking to understand the connection between integrins and epithelial injury in the intestine employed ECIS to help better understand the role of epithelial CD98 in colonic inflammation. This group injured Caco2-BBE cell monolayers apically with application of 3% Dextran Sodium Sulfate (DSS) treatment prior to ECIS with voltage pulses of 40 kHz and 4.5 V for 30s and then allowed the wound to heal. The resulting examination of CD98 during wound infliction and restitution revealed that the upregulation of epithelial CD98 mediated by INF-γ during intestinal inflammation in DSS-treated monolayers hampered cellular restitution in this model.⁵⁴ Another study performed with Caco2-BBE cell monolayers explored the role of ADAM-15 in intestinal wound healing. Monolayers with overexpressed ADAM-15 and control groups were injured with a voltage pulse of 40 Hz and 4.5-V amplitude for 30s to cause cell death. Ultimately, cells with upregulated ADAM-15 did not recover as well as the control group, which was able to completely heal the wound after 19 hrs.⁵⁵

Where many of the existing models fail to provide consistent injury that can be modified to exacerbate or change the level of injury, ECIS produces a consistent injury via gold film electrodes while also allowing investigators to track restitution in real-time via impedance measurements. However, the heat created from the high current can alter the viability of the surrounding cells and cause additional damage. In addition, the density of cell monolayers and the extent to which interepithelial junctions are intact can influence impedance measurements and make the interpretation of study results challenging.

Organoid models

Recent advances in 3-D organoid models of the intestine have created a unique opportunity to evaluate epithelial injury and repair. Organoids display architecture and physiologic similarity to in vivo tissues including crypt and villus regions and homeostasis of stem cell populations.^56–58 One study subjected small intestinal organoids to γ-radiation to support previous reports indicating hepatocyte nuclear factor 4α as an upstream regulator of repair.⁵⁹ Another investigated injury-responsive stem cell populations to identify VPA and EPZ6438 as features critical to regeneration after epithelial damage.⁶⁰ However, organoids pose a challenge for investigators evaluating physical damage to the epithelial monolayer. Due to the closed nature of the enteroid, accessing the apical surface is extremely difficult without inadvertently disrupting the barrier of the organoid. This model does still provide a means of studying important questions regarding the reparative process of intestinal epithelium in a physiologically relevant 3-D system.

Cell lines

In addition to knowledge of the characteristics of epithelial wound models, an important component of any study is the epithelial cell type. There is significant utility in highly reductionist model systems that allow precise study of biological mechanisms. Although these models do not fully recapitulate the complexity of the intestinal mucosa, they do enable the selective study of epithelial wound repair, which is arguably the most vital component of the intestinal mucosal barrier.^7,21,61 Although transformed epithelial cell lines have been developed for convenient usage in the lab, primary cell culture techniques have also undergone major advances to not only be more convenient for lab use, but also represent the multiple cell lineages that make up the epithelial barrier. While primary cultured epithelium more closely represents the in vivo environment, they have drawbacks related to passage time, confluence abilities, and susceptibility to experimental variation. These disadvantages can prevent investigators from understanding mechanisms of restitution, and therefore must be fully understood before embarking on further studies. In this section, some of the more commonly used cell lines that can be used for the study of restitution are described.

IEC-6

Quaroni et al. were one of the first labs to characterize an epithelial cell line that could be passaged multiple times when they developed IEC-6 cells.⁶² IEC-6 cells are a non-transformed cell line derived from rat small intestinal crypt cells that mimic small intestinal epithelium in vitro. They are polygonal in shape with large nuclei and have microvilli present on one side of the cell demonstrating polarization.⁶² They also express tight junction proteins, such as occludin, claudin-6, and ZO-1, as well as adherens junction proteins such as E-cadherin that are all crucial to adhere cells together into a uniform barrier.⁶³ In addition, their average peak TEER is ~30 Ω·cm² after 6 days of seeding,⁶⁴ which mimics the ‘leakier’ TEER of small intestine. Originally, serial passages were made for ~6 months. From these passages, other researchers have been able to use these cells for various barrier function studies involving iron absorption,⁶⁵ mitogen activated protein (MAP) kinases’ role in apoptosis,⁶⁶ and enteric protozoan’s effect on TEER.⁶⁴ It must be noted that these cells mimic undifferentiated small intestinal crypt cells and therefore need additional coercion to differentiate. Fortunately, certain cell-medium additions such as transforming growth factor beta (TGF-β),⁶⁷ interferon gamma (IFN-γ), tumor necrosis factor alpha (TNF-α),⁶⁸ and gastrin⁶⁹ have been found to successfully differentiate IEC-6 cells. Although differentiated IEC-6 cells seem to have only been used in one scratch wounding study investigating the role of polyamines and myosin II in cell motility,⁷⁰their undifferentiated counterparts have been widely used for different types of in vitro restitution studies. Most commonly, scratch wound models have utilized IEC-6 monolayers to investigate how various treatments like polyamine depletion,⁷¹ cytokine modulation,¹² and Astragalus polysaccharides⁷² can affect cell migration. These cells have also been used in in vitro hypoxia studies investigating how hypoxia affects barrier function⁷³ and different ways to protect against these hypoxic injuries.⁷⁴

Caco-2

Researchers have found multiple tumor-based cell lines that have the ability to differentiate. One example is the Caco-2 cell line which was derived from a human colon adenocarcinoma and was developed in part due to the inability of researchers to passage and culture differentiated primary epithelial cells.⁹ This transformed cell line spontaneously differentiates, and forms monolayers connected by tight junctions. These cells are absorptive, polarized, and have distinct apical, lateral, and basal cell membrane characteristics.⁹ On their apical side, they have microvilli similar to both enterocytes and colonocytes.⁷⁵ Furthermore, Caco-2 cells do exhibit mature enterocytic properties, such as disaccharidase and peptidase expression, more typical of small intestinal epithelium.⁷⁶ However, Caco- 2 cells are a transformed cell line, and have been shown to have gene mutations in p53, APC, ββ-catenin, and Smad4, which are uncharacteristic of intestinal epithelial cells.⁷⁷ Even with these mutations, Caco-2 cells are still widely used for intestinal studies focused on barrier functions, such as barrier maintenance in the absence of glutamine,⁷⁸ tight junction modulation,⁷⁹ and permeability response to short-chain fatty acids.⁸⁰ They have also been a part of multiple restitution studies involving varying model types. Caco-2 scratch assays have been performed to investigate how certain actions or factors enhance wound healing like the activation of protease activated receptor-2.⁸¹ Caco- 2 cells were also used to optimize the cell exclusion zone assay⁴⁶ as well as investigate how the loss of lipolysis-stimulated lipoprotein receptors affect barrier integrity.⁸² In addition, these monolayers have undergone hypoxic injury to better understand the effect on barrier integrity.^83,84

There also exists a Caco-2 derived cell line, C2BBe1, that is more similar to the colon due to its brush border enzyme (BBe) expression. The BB of this cell line contains proteins such as villin, fimbrin, BB myosin I, fodrin, myosin II, and sucrase-isomaltase, all of which are found in the BB of colonocytes.⁸⁵ Given these unique properties, C2BBe1 has been used in a variety of intestinal studies. For example, C2BBe1 was used to study the effects of essential oils on inflammatory diseases⁸⁶ and recently used to investigate how SARS-Cov-2 affects the gastrointestinal tract.⁵ However, C2BBe1 cells have not been used for restitution studies involving the models previously stated.

HT29

Another cell line originating from a human colon adenocarcinoma is HT29 cells. HT29 cells are more versatile than Caco-2 cells because they can accurately represent multiple cell lineages in vitro. In the presence of glucose concentrated media, these cells remain undifferentiated. When glucose is replaced by galactose, they differentiate into enterocyte-like cells.⁷⁶ In media highly concentrated with glucose and methotrexate (MTX), they begin producing large amounts of mucin, mimicking goblet cells.⁷⁵ This led to the subcell line called HT29-MTX. This cell line has been used to study drug transport⁸⁷ and protein digestion.⁸⁸ This cell line has also been used in conjunction with Caco-2 cells to study drug absorption⁸⁹ and toxicity.⁹⁰ In these co-culture studies, HT29-MTX cells are cultured with Caco-2 cells on a transwell at a biologically informed ratio of 1:9, respectively. In regard to wound studies, the HT29 cell line has been used to analyze how different biological components like MUC2,⁹¹ beta-Adrenoreceptor antagonists,⁹² and protease activated receptor 2 agonists⁹³ influence restitution given a cell barren space either caused by scratch wound or cell exclusion zone assay. In addition, they have also been used to study epithelial integrity following hypoxic injury and the mechanisms whereby calcium channel blockade following this injury affects barrier integrity.⁹⁴

T84

With similar origins to Caco-2 and HT29 cells, T84 cells are also derived from human colon adenocarcinoma; however, these cells originated from a metastatic lung tumor. T84 cells possess similarities to Caco-2 cells, such as their absorptive properties, spontaneous differentiation post confluence, and columnar cell physiology with apical microvilli.⁹⁵ These similarities have led researchers to use both cell lines interchangeably as a monolayer model of intestinal epithelial cells. However, some studies have shown that T84 cells act more like undifferentiated crypt cells and colonocytes. T84 cells, like many intestinal cell lines, also exhibit atypically high measures of barrier resistance (Table 2). Madara et al. indicated that the T84 cell Cl⁻ secretory response was more similar to crypt cells’ secretory response than mature absorptive epithelial cells.¹⁰³ In 2017, Devriese et al. built on this argument by showing the difference between T84 models and Caco-2 models.⁹⁵ More specifically, this study showed that T84 cells had 3-fold shorter microvilli than Caco-2 cells, which is a characteristic of colon epithelial cells. In addition, the colonic marker, monocarboxylate transporter 1 (MCT1), was upregulated in differentiated T84 cells, while enterocyte brush border enzymes were not.⁹⁵ Based on these findings, and the fact that Cl⁻ secretion plays a large role in the regulation of diarrhea and constipation, T84 cells have been widely used as models for electrogenic Cl⁻ secretion in the colon.^96,104–107 This cell line has also been used to model restitution following injury. Scratch assays have been conducted to see how factors such as galectins¹⁰⁸ or microbiota-derived butyrate¹⁰⁹ affect re-epithelization. In addition, T84 cells have been used in conjunction with neutrophils or PMNs, to better understand their role following intestinal injury. One study coupled a scratch wound assay with a PMN migration assay and found that PMN adhesion to epithelial cells post-migration enhances cell proliferation thereby aiding in wound healing. They believed that although transmigrating PMNs do cause additional damage to the tissue, once they adhere to epithelial cells, ICAM-1, a luminal epithelial surface receptor, is engaged, and this initiates Akt and β-catenin signaling, both of which are important signaling pathways for cell proliferation.⁹⁷ In addition, as previously mentioned, Nusrat et al. utilized T84 monolayers to model a wound caused by PMN infiltration. From this model, they found that wound closure methods vary with regard to wound size. Lamellipodia extensions are the main method of closure for larger wounds, while smaller wounds tend to form and contract an actinomyosin cytoskeletal ring.⁹⁸

Table 2. Comparison of intestinal cell lines.

Cell line	Origin	Transformed (Y/N)	Average TEER	Notes	Reference
IPEC-J2	Neonatal porcine jejunum	N	500–4000 (Ω X $c{m}^2)$	-Not known to possess mutations -Differentiates into epithelial cell lineage without coercion	^Citation96
Caco-2	Human colorectal adenocarcinoma	Y	150–400 (Ω X $c{m}^2)$	-Form distinctly polarized monolayer -Possess apical microvilli and are phenotypically similar to small intestinal epithelial cells -Known mutations	^Citation97,Citation98
T84	Human colon adenocarcinoma	Y	2,000–4,000 (Ω X $c{m}^2)$	-Derived initially from lung -Used in study of colonic $C{l}^{-}$ secretion	^Citation99
HT-29	Human colon adenocarcinoma	Y	500–600 (Ω X $c{m}^2)$	-Can produce mucin, act as goblet cells -Differentiates into enterocytes in presence of glucose	^Citation100,Citation101
IEC-6	Rat small intestinal epithelium	N	42 > (Ω X $c{m}^2)$	-Exhibit crypt-like phenotype, need media additives to differentiate	^Citation102

IPEC-J2

The IPEC-J2 cell line is a non-transformed line derived from jejunal crypts of a neonatal pig, developed in 1989.⁹⁹ Historically, these cells have been used for a variety of functional studies due to their similarity to small intestinal epithelia. Specifically, IPEC-J2 cells contain several tight junctional proteins, are capable of secreting mucins (specifically Muc1 and Muc3), and cytokines such as TNF-α, and express a number of toll like receptors (TLR-1, −2, −3, −4, −5, −6, −8, −9, and −10).^100,101 However, one of the largest limitations of the IPEC-J2 line grown conventionally in 5–10% Fetal Bovine Serum has been its unusually high barrier resistance, with TEER values averaging upwards of 1000 ${\mathrm{\Omega }}^2\times c{m}^2$.¹⁰² Previous studies performed by Zakrewski et al. showed a dramatic decrease in TEER values of IPEC-J2 cells when grown in 10% Pig Serum, with average TEER being 200–400 Ω·cm²after 14 days.¹⁰² Recently, studies in our laboratory evaluating the effects of media containing Wnt3a, R-spondin, Noggin (WRN) on IPEC-J2 cells were performed demonstrating decreased TEER values in IPEC-J2 cells grown in WRN-conditioned media. TEER values in these studies were 50.3 ± 2.3 Ω·cm² over 12 days as opposed to IPEC-J2s grown in conventional media which exceeded 2000 Ω·cm2.¹¹⁰ This suggests potential for IPEC-J2 cells to be manipulated using media typically used for organoid culture, and may represent a more physiologically relevant model of barrier function in vitro in the future.¹¹⁰ IPEC-J2 cells have also been used when studying restitution mechanisms. For example, one study utilized scratch wounded IPEC-J2 cell monolayers to observe restitution during tryptophan supplementation and its effects on the CaSR/Rac1/PLC-γ1 signaling pathway.¹¹¹ Additionally, IPEC-J2 cells have been employed to explore the positive effect of sodium butyrate and identify it as an agent capable of recovering intestinal tissues after assault or injury.¹¹²

Conclusion

The in vitro models used to study gastrointestinal pathophysiology are an integral component of mechanistic research. For example, the HT-29 and IPEC-J2 cell lines have contributed heavily to research in restitution mechanisms by performing scratch wounding assays and cell exclusion models. Similarly, IEC-6 cells have been utilized in neutrophil exclusion studies. The T84 cell line has been utilized to better understand neutrophilic injury during the reparative process. Caco-2 cells have also been invaluable in studies seeking to understand restitution through ECIS and its novel system for modeling injury through electrical pulses. However, the success of these models lies in the ability of researchers to properly choose the cell line and model best suited for their study.

Future work in optimizing these modeling systems to create increasingly relevant models of the in vivo environment is necessary to speed the pace toward clinical application. In addition, the field of in vitro modeling systems of intestinal epithelium is expanding rapidly and new systems which rely on primary cells derived from individual patients and grown in monolayers or 3-D organoids are becoming commonplace. These systems will inevitably require alternative or adapted methods of in vitro injury and repair to assess their viability properly and accurately as a model of restitution after insult.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Institute of Child Health and Human Development [R01 HD095876].