Altered intestinal microbiome and epithelial damage aggravate intestinal graft-versus-host disease

ABSTRACT

Despite significant achievements in hematopoietic stem cell transplantation (HSCT), graft-versus-host disease (GVHD), especially intestinal GVHD, remains a major obstacle to this procedure. GVHD has long been regarded as a pathogenic immune response, and the intestine has been simply considered as a target of immune attack. In effect, multiple factors contribute to intestinal damage after transplantation. Impaired intestinal homeostasis including altered intestinal microbiome and epithelial damage results in delayed wound healing, amplified immune response and sustained tissue destruction, and it may not fully recover following immunosuppression. In this review, we summarize factors leading to intestinal damage and discuss the relationship between intestinal damage and GVHD. We also describe the great potential of remodeling intestinal homeostasis in GVHD management.

KEYWORDS:

• graft-versus-host disease
• hematopoietic stem cell transplantation
• intestinal microbiome
• intestinal epithelium

Introduction

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) has been established as an effective therapy for hematological malignancies. The reconstituted immune system after transplantation is efficient to eliminate malignant cells, while it may lead to targeted host tissue injuries termed as graft-versus-host disease (GVHD).¹ The pathology of GVHD can be categorized into three steps: first, conditioning regimens cause tissue damage, in which the release of damage-associated molecular patterns (DAMPs), pathogen-associated molecular patterns (PAMPs), together with abundant pro-inflammatory cytokines activate the innate immune system; subsequently, host antigen-presented antigen-presenting cells (APCs) activate donor T cells, leading to proliferation, differentiation, migration and secretion of cytokines; finally, allogeneic T cells infiltrate and damage the target organs.^2,3

The intestine not only is the major target of GVHD but also modulates the development of systemic GVHD.² The intestinal epithelium serves as the major site of nutrient digestion and absorption, and it is also crucial for preventing pathogen invasion and generating tolerance to intestinal microbes. In turn, microbes residing in the intestinal tract are tightly regulated, and their metabolic products benefit intestinal growth and shape mucosal immunity.^4,5 While in the context of GVHD, intestinal barrier loss and dysbiosis break the balance of the intestinal homeostasis and activate the immune system (Figure 1).^6–8

Figure 1. Intestinal epithelium in homeostasis and GVHD. a. The intestinal epithelium comprises crypts and villus. A mucous layer separates epithelial cells and intestinal bacteria, which is indispensable to intestinal homeostasis. ISCs give rise to all kinds of epithelial cells, and the activities of ISCs are tightly regulated by various cells and signals; b. Conditioning regimens induce epithelial damage and dysbiosis. Antibiotics and diet further affect the microbiome components and metabolites. Accumulated DAMPs and PAMPs activate the innate immune system and trigger donor lymphocyte infiltration, leading to impaired ISC niche and sustained tissue damage.(ISC, intestinal stem cell; EP, enterocyte progenitor; SP, secretory progenitor; EC, enterocyte; PC, Paneth cell; GC, goblet cell; EEC, enteroendocrine cell; TC, tuft cell; Mø, macrophage; Neu, neutrophil; Lym, lymphocyte; APC, antigen-presenting cell; SCFA, short-chain fatty acid; BA, bile acid; TMAO, trimethylamine N-oxide; DAMPs, damage-associated molecular patterns; PAMPs, pathogen-associated molecular patterns.).

At present, immunosuppressive agents remain the preferred choice for GVHD treatment. With an evolving knowledge of T cell biology, therapeutics develop from broad immune suppression toward precision medicine, such as Ruxolitinib, a selective JAK1/2 inhibitor.⁹ Notwithstanding these advances, treatment resistance still occurs in a large proportion of patients (especially in patients with intestinal GVHD) and emerges as a new challenge.¹⁰ Indeed, the disturbance in intestinal ecosystem still exists after immunosuppression. Remodeling intestinal homeostasis, a field that we have long neglected, may become an important complementary strategy in GVHD management.¹¹

Here, we review the factors that leading to intestinal damage during HSCT. We also scrutinize the novel approaches to promoting tissue repair and intestinal flora intervention.

Intestinal damage in GVHD

Damage to the intestinal microbiome

The intestinal epithelium harbors large communities of microbes, including bacteria, archaea, fungi and viruses. Bacteria, the major of which belong to the Firmicutes and Bacteroidetes phyla and the minority of which belong to Proteobacteria, Actinobacteria, Fusobacteria and Verrucomicrobia phyla, represent the largest group within the intestinal microbes.^7,8 However, several factors such as conditioning regimens, antibiotic exposure and dietary changes contribute to the flora disturbance and exacerbate intestinal GVHD.

Conditioning regimens

Based on 16S rRNA gene sequencing, several preclinical studies have investigated the impact of cancer therapy on fecal flora. Most of these studies observed a reduced diversity of intestinal bacteria after radiotherapy or chemotherapy.^12–18 The major components of intestinal bacteria, Firmicutes ^{12–14,19,20} or Bacteroidetes ^{14,18,19,21–23}, are more likely to decrease, while the proportions of Proteobacteria ^13,22,23 and Verrucomicrobia ^12,13,20 tend to increase. On the genus level, Clostridium ^12,13, Lactobacillus ^15,19,22,23 Ruminococcus ¹³ and Blautia ²² may account for the loss of Firmicute, and Bacteroides may account for the decreased abundance of Bacteroidetes. ^19,21,23 Meanwhile, some opportunistic pathogens increase after cancer therapy, such as Akkermansia belonging to Verrucomicrobia,^13,19,20 and Enterococcus belonging to Firmicutes.^14,22,24 Moreover, the perturbation of intestinal microbiome increases inflammatory response and amplifies tissue damage. Using bacterial–epithelial coculture, radiation-altered microbes enhanced the secretion of IL-1β and TNF-α from HT29 epithelial cells. Mice inoculated with irradiated microbes also exhibited more severe intestinal damage compared to those which received naïve microbes.¹³

In clinical studies, the stool specimens collected from patients with hematological malignancies had significantly lower microbial diversity than healthy controls.^25–28 Chemotherapy further disrupted the bacteria diversity and its components^26,28–32, which was characterized by steep decrease in proportions of Blautia ^31,32 and Clostridium ^30,32,33, and an expansion of Enterococcus ^28,30,33,34.

Altogether, conditioning regimens dramatically shift the structure of the intestinal microbes.

Antibiotics

The intestine is a strict anaerobic environment under normal circumstances. Therefore, most intestinal bacteria are anaerobic, and they are 100–1000 times numerous than aerobic bacteria or facultative anaerobes (e.g., Enterococcus).^7,35 During HSCT, antibiotics are commonly prescribed to prevent bacterial infection. Nevertheless, antibiotics are a double-edged sword. Broad-spectrum antibiotics and heavy antibiotic exposure may shift the bacterial profiles by depleting commensal bacteria and inducing the expansion of pathogenic populations.

In murine transplant models, several studies have investigated the antibiotics’ effect on intestinal flora and GVHD. Ampicillin, imipenem-cilastatin, meropenem, etc., were used as broad-spectrum antibiotics. Aztreonam, cefepime, etc., were antibiotics with minor anaerobic activities. In these studies, broad-spectrum antibiotics increased epithelial damage and the GVHD-related mortality rate. Changes in bacterial compositions revealed that broad-spectrum antibiotics suppress the anaerobic commensals such as Lactobacillus, Clostridia, Blautia, and increase the abundance of pathogens like Enterococcus and mucus-degrading microbes such as Akkermansia muciniphila and Bacteroides thetaiotaomicron.^36–38 Enterococcus activates the mucosal immune system and secretes metalloprotease that impairs the epithelial barrier.³⁶ Mucus-degrading bacteria also compromise the mucus layer and increase intestinal permeability.^37,38 Conversely, antibiotics with reduced activities against anaerobes better preserved the intestinal bacterial diversity and spared more commensal bacteria.

The results from patients were similar to those from mice. Metronidazole, a drug highly active against anaerobes, increased the risk of Enterococcus dominance by threefold.³⁹ Broad-spectrum antibiotics or antibiotics that inhibit anaerobes exacerbated the loss of Blautia ⁴⁰, Lactobacillus ³⁷, Bacteroidetes ^37,41 and Clostridiales ^42,43 (Clostridium cluster XIVa)⁴⁴. An increased proportion of Enterococcus was more prominent in patients developing GVHD,^45,46 and the abundance of Blautia was negatively related to GVHD-related mortality.^41,46 Apart from this, numerous clinical studies indicated that antibiotic exposure during HSCT aggravated the incidence or severity of GVHD.^{37,41–43,47–51} Early and prolonged use of antibiotics even induced more severe dysbiosis.^42,48,52

In general, irrational use of antibiotics, especially broad-spectrum antibiotics, may contribute to dysbiosis and increase the risk of GVHD.

Diet and nutrient

Diet provides energy and nutrient, it also regulates the intestinal microbial ecology and the host-microbes interactions.^53,54

Oral and gastrointestinal mucositis frequently prevent eating and drinking in tumor treatment. To meet nutrient requirements, patients often receive enteral nutrition (EN) or parenteral nutrition (PN) during HSCT. Compared with PN, EN was associated with a reduced risk of GVHD.^40,55–58 Besides, patients in the EN group showed advantages in neutrophil engraftment⁵⁶, infection control^55,59 and overall survival^55–58 over those who received PN. Compared with patients who received EN, PN resulted in loss of Blautia, Dorea and Bacteroides (accompanied by decreased short-chain fatty acid (SCFA) production), and an increase in Streptococcus, which possibly aggravates the development of GVHD.^40,60

More recently, experiments have demonstrated that some dietary components directly affect the intestinal microbial structure and influence the risk of GVHD. Serum level of high stearic acid is a reliable biomarker for predicting GVHD in patients receiving HSCT. In murine aGVHD models, a high stearic acid diet aggravated GVHD, resulting in loss of commensals and enrichment of Akkermansia. Mice fed with a diet including Akkermansia muciniphila exacerbated GVHD, while antibiotics against Akkermansia muciniphila attenuated GVHD.⁶¹ Enterococcus dominance, another potential driver of GVHD, frequently occurred in mice and humans suffering from GVHD. As Enterococcus depends on lactose and galactose metabolism, dietary lactose depletion in mice suppressed Enterococcus and ameliorated the severity of GVHD.⁶² Besides, a high-fat diet and obesity in mice led to loss of Clostridiaceae, and expansion of Akkermansia muciniphila and Enterococcus, which may explain recipients with high body mass index are more susceptible to GVHD.⁶³ Tyrosine supplement in diet restored the intestinal microbes (e.g., the recovery of Lachnospiraceae) and ameliorated GVHD in murine models. Tyrosine deprivation worsened the GVHD-related outcomes.⁶⁴ In another report, Lachnospiraceae was found to accumulate in leukemia patients post-radiation and support gastrointestinal repair.⁶⁵

Damage to the intestinal epithelial cells (IECs)

The epithelial cell components originate from Lgr5⁺ ISCs, which reside at the bottom of the crypt. ISCs proliferate and give rise to progenitor cells, the latter cells further differentiate into enterocytes (ECs) that take charge of nutrient absorption, Paneth cells (PCs) that secrete antimicrobial peptides and lysozymes, goblet cells (GCs) that secrete mucus, enteroendocrine cells (EECs) that release hormones, etc. The intestinal epithelial barrier, which consists of different IECs with intact tight junctions and a chemical barrier that includes mucus and antibacterial substances, is critical for maintaining intestinal homeostasis.⁴

Conditioning regimens

Conditioning regimens are often used to deplete tumor cells and vacate bone marrow stem cell niches. The intestinal epithelium is extremely vulnerable to conditioning toxicity due to its high cellular turnover. Conditioning regimens prefer to kill crypt cells, leading to considerable cell death and reduced cell production, which is characterized by crypt hypoplasia, shortened villus and impaired tight junctions.^66–68 The loss of IECs results in impaired absorptive activity,⁶⁹ and also disturbs the intestinal integrity with elevated endotoxin translocation and macromolecule leakage.^69–71 Intestinal toxicity positively correlates with the conditioning intensity and predicts GVHD severity.^71,72

To explore these further, transgenic mice with improved intestinal integrity were studied in intestinal damage and GVHD. The Tg222 mice have a strengthened mucus layer. After intensive chemotherapy, Tg222 mice had lower tissue damage and better-preserved intestinal microbes than wild-type mice, which was linked to reduced bacterial translocations.²⁴ Cytoskeletal contraction, a process regulated by long myosin light chain kinase (MLCK210), is one of the major mechanisms of tight junction disruption. MLCK210 expression and myosin light chain phosphorylation increase in intestinal epithelium suffering from GVHD. Genetic knockout of MLCK210 promoted intestinal barrier recovery from radiation, prevented T cell infiltration and limited GVHD propagation.⁷³

Dysregulated microbial metabolites

Previously, we describe the impact of diet on intestinal microbes. On the other hand, intestinal microbes metabolize dietary components and secrete metabolites, which have a profound impact on human health.⁵³ Diet, microbial signatures and metabolite profiles are complex and closely related. For example, most genera in the Firmicutes prefer dietary protein that provides branched-chain fatty acids and aromatic metabolites, while Clostridium avidly uses dietary fiber as a carbon source; Bacteroides prefer dietary fiber over dietary protein, in contrast to other genera in Bacteroidetes.⁵⁴ Among diverse microbial metabolites, SCFAs, bile acids (BAs) and indole derivatives play vital roles in maintaining intestinal homeostasis.

SCFAs are 1–6 carbon volatile fatty acids that are mostly derived from dietary fibers.⁵⁴ Butyrate, the best-studied SCFA by far, can be produced by families belonging to the Clostridiales order such as Lachnospiraceae, Ruminococcaceae, Erysipelotrichaceae, and genera like Clostridium, Dorea, Blautia and Roseburia. Butyrate has diverse impacts on intestinal cells and host functions. First, butyrate can be metabolized via fatty acid oxidation and becomes the major energy source for IECs. Besides, butyrate serves as a ligand for G protein-coupled receptors (GPCR) and aryl hydrocarbon receptors (AhR). Acetyl-CoA which is metabolized from butyrate also takes part in transcriptional regulations by enhancing histone acetylation.⁷⁴ Previously, accumulating evidence has identified a reduced level of butyrate or butyrogenic bacteria in patients suffering from GVHD.^{40,48,52,75,76} In murine studies, researchers also found butyrate decreased in intestinal tissue and stools after allo-HSCT, together with the reduced level of histone acetylation in IECs. Intragastric gavage of butyrate in recipients restored the loss of butyrate and histone acetylation in IECs, accompanied by a protective effect in GVHD. Recipients treated with Clostridiales also showed an increased level of butyrate and attenuated GVHD. Mechanistically, histone acetylation directly acts on IECs to suppress the expression of apoptosis-related genes, and upregulate the production of junction proteins. In vitro, butyrate enhanced the survival and growth of intestinal organoids, and butyrate-pretreated organoids were more tolerant when cocultured with allogenic T cells.⁷⁷

As a downstream metabolite of cholesterol, BAs can be classified into two groups. Primary BAs (PBAs) that are conjugated with taurine or glycine are synthesized in the liver. When released in the intestine, 95% of PBAs recirculate to the liver by enterohepatic circulation, the rest are modified by bacteria secreting bile acid hydrolase (such as Bacteroides, Clostridium, Lactobacillus, Bifidobacterium and Enterococcus) to secondary BAs (SBAs). Except for facilitating the digestion and absorption of dietary lipids, BAs trigger inhibitory or active signals by binding farnesoid X receptor (FXR) and G protein-coupled bile acid receptor-1 (Gpbar-1 or TGR5).⁷⁸ For example, deoxycholic acid (DCA), also known as a strong FXR antagonist, induces intestinal inflammation and promotes stem cell expansion.^79,80 Lithocholic acid (LCA), another SBA, activates the TGR5 and stimulates the proliferation of ISCs and intestinal regeneration.⁸¹ BA-dysmetabolism has been linked with many diseases such as inflammatory bowel disease (IBD), which is characterized by reduced levels of SBAs in stool samples.^82,83 In a murine model, GVHD disrupted TGR5 and reduced the levels of BAs. Using cytokine-organoid coculture, TNF and IFN-γ killed cells and prevented intestinal organoid growth. Adding BAs to this coculture system, especially an SBA named tauroursodeoxycholic acid (TUDCA), rescued the intestinal organoid from cytokine-mediated cytotoxicity. In vivo, pre-administration of TUCDA attenuated GVHD, probably due to decreased intestinal antigen presentation and reduced T cell activation. ISCs and PCs were also better preserved as TUCDA downregulated the apoptosis-related genes.⁸⁴ Ursodeoxycholic acid (UDCA), another SBA used in several clinical trials, reduced the incidence of GVHD.^85,86

Tryptophan is an essential amino acid for humans. Tryptophan can be metabolized by both epithelial cells and immune cells via the kynurenine (Kyn) pathway, and it is also degraded by intestinal bacteria into indoles and indole derivatives. Many indole derivatives are ligands for AhR or pregnane X receptor (PXR), including indoleacrylic acid (IA), indole-3-aldehyde (IAld), indole-3-acid-acetic (IAA), indole-3-propionic acid (IPA), etc.⁸⁷ Activation of AhR or PXR promotes barrier integrity, partly by inducing GC differentiation and maintaining tight junctions.^88–90 In GVHD, a new protective mechanism of indole derivatives has been found, in which indole-3-carboxaldehyde (ICA) activates the type-I IFN signal, limits radiation damage and promotes epithelial regeneration.⁹¹

In addition, mounting evidence suggests that metabolites can regulate immune cell activities.⁹² For example, butyrate acts on T cells and increases H3 acetylation in the locus of Foxp3, thereby promoting the differentiation of regulatory T cells (Tregs).⁹³ LCA metabolites (3-oxoLCA and isoalloLCA) regulate host immune responses by modulating Th17 and Tregs differentiation.⁹⁴ Activated AhR promotes IL-22 production from lamina propria lymphocytes and accelerates intestinal epithelial proliferation.⁹⁵ Moreover, some metabolites even accelerate GVHD, such as trimethylamine N-oxide (TMAO) and retinoic acid.^96,97

Immune cell infiltration

Immune cell activation and infiltration mediate sustained tissue damage. Here, we describe that immune responses are closely related to disturbed intestinal homeostasis, allowing immune cells to invade and further destroy ISCs and the ISC niche.

As mentioned before, conditioning regimens cause tissue damage and bacterial translocation. DAMPs, PAMPs, as well as a bunch of cytokines rapidly accumulate and recruit neutrophils to the damaged intestine.^3,98,99 In contrast, germ-free condition reduces the level of neutrophil migration.⁹⁸ Neutrophils amplify tissue damage by releasing ROS, and they also express high levels of MHC-II and contribute to T cell activation. Defect in ROS production or lacking TLR compromises neutrophil function and reduces GVHD severity.^98,99 Besides, microbiome regulates MHC-II expression on IECs. Upon radiation or dysbiosis, the level of MHC-II on IEC elevates, and these MHC-II expressing IECs activate T cells and initiate GVHD. Meanwhile, MHC-II is absent in IECs of germ-free mice, even after radiation. Specific deletion of MHC-II blocks T cell activation and the subsequent GVHD.¹⁰⁰

Once activated, T cells migrate to the target organs and mediate killing effects. Early after HSCT, T cell density increases throughout the intestinal epithelium. Notably, the crypt region where MAdCAM-1⁺ vessels located are is the first and primary target of immune attack, resulting in rapid loss of ISCs before tissue destruction.¹⁰¹ PCs and GCs are also targets of GVHD, leading to flora disturbance, increased bacterial translocation and amplified GVHD severity.^102–104 Besides, innate lymphoid cells (ILCs) promote ISC regeneration via secretion of IL-22. Depletion of host ILCs and IL-22 by donor T cells dampens tissue healing and boosts the development of GVHD.^105–107

Additionally, T cells regulate ISC function through variant cytokines. In a murine T cell/cytokine-organoid coculture system, Tregs and IL-10 maintain ISC self-renew, while T helper (Th) cells and their cytokines (IFN-γ, IL-13 and IL-17A) promote ISC differentiation.¹⁰⁸ Interferon regulatory factor 2 (IRF2) is a negative regulator of the IFN signal. Mice completely lacking IRF2 or with a selective IRF2 deletion in IECs have significantly fewer ISCs and accumulated immature PCs.¹⁰⁹ Excessive IFN also activates the JAK/STAT pathway, promoting proapoptotic gene expression and depleting ISCs and PCs.¹¹⁰ Ruxolitinib, a JAK1/2 inhibitor, prevents T cell-mediated damage and restores the loss of ISCs and PCs.^110,111 Enrichment of another cytokine TNF-α enhances ISC expansion but induces differentiated cell necroptosis.¹¹²

Interplay between intestinal damage and GVHD

Peri-transplantation treatments break tissue homeostasis and immune tolerance. Epithelial damage and dysbiosis lead to massive leakage of macromolecules, bacterial products and food antigens, which further accelerates antigen presentation, T cell accumulation and GVHD propagation. Moreover, disturbances in metabolite profiles and cytokines result in cellular dysfunction, both in epithelial cells and immune cells. Altogether, the above results suggest that intestinal damage and GVHD are tightly linked and mutually facilitated.

Novel approaches targeting the intestinal ecosystem

The self-renew and differentiation of ISCs are precisely regulated by signals from the basal side including Wnt (stem cell maintenance and proliferation), Notch (lineage decision) and BMP (negative regulation of crypt proliferation), and signals from the apical side such as microbes and dietary components.¹¹³ ISCs are also affected by immune cells and metabolism.^108,114 The development of GVHD not only destroys the cellular components but also disrupts the complex regulatory network. Given that immunosuppression has not fully addressed GVHD, strategies that preserve the intestinal ecosystem should catch enough attention (Figure 2).^9,11,115

Figure 2. Outstanding issues and treatment strategies for GVHD.

Promoting tissue repair

Pretransplant conditioning and GVHD lead to a sharp reduction in the ISC pool. Restoring ISCs may promote tissue repair and limit GVHD development.

R-spondin1 has been identified as a potent Wnt agonist, which stabilizes β-catenin and activates the downstream genes.¹¹⁶ In murine transplant models, injection of R-spondin1 significantly ameliorated GVHD. Mechanistically, R-spondin1 boosts crypt cell expansion and increases the number of ISCs after radiation.¹¹⁷ Moreover, R-spondin1 promotes PC differentiation. The increased production of α-defensins from PCs prevents dysbiosis and restores bacterial diversity.¹¹⁸ Consistently, Reg3γ is an antibacterial protein that is secreted by PCs and is negatively related to GVHD severity. IL-22 administration restores the loss of Reg3γ and prevents the apoptosis of ISCs and PCs.¹¹⁹ IL-22 also enhances ISC recovery via STAT3 phosphorylation.¹²⁰ Combination of IL-22 and systemic corticosteroids showed good clinical efficacy as an initial treatment for lower intestinal GVHD, as 70% (19/27) of patients achieved a day-28 treatment response with an increased fecal microbial diversity.¹²¹ Additionally, lithium stimulates Wnt signal via GSK3 inhibition and drives epithelial regeneration.¹²² Teduglutide, a GLP-2 agonist, promotes the expansion of PCs and ISCs and improves the microbiome diversity.¹²³ IL25, a growth factor of GCs, inhibits bacterial translocation and attenuates GVHD.¹⁰⁴ IFN-I and RIG-I signals have also been reported to fuel ISC regeneration and benefit GVHD control.^124,125 It is worth noting that most of these drugs above do not impair T cell reconstitution and the graft versus leukemia (GVL) effect, indicating a non-immunosuppressive direction for GVHD management.

Intestinal flora intervention

Intestinal Flora disturbance in HSCT leads to immune activation and epithelial cell dysfunction, which finally fuel the development of GVHD. Interventions on microbes are widely investigated, including oral prebiotics/probiotics and fecal microbiota transplantation (FMT).

Probiotics are beneficial bacteria or yeast for the host, and prebiotics are digestive fibers that support probiotic colonization. Compared with historical control, clinical application of the prebiotic mixture GFO (glutamine, fiber and oligosaccharides) preserved the butyrate-producing bacteria, shortened the duration of diarrhea and reduced the risk of GVHD.¹²⁶ In another report, the microbial structure altered by FOS (fructo-oligosaccharides) could not be maintained after HSCT. No advantages in survival or GVHD were found in patients receiving FOS.¹²⁷ Besides, administration of Bacteroides fragilis and Lactobacillus rhamnosus GG through oral gavage reduced GVHD and improved survival in experimental models.^128,129 However, Lactobacillus rhamnosus GG failed to reduce the incidence of GVHD in clinical trials.^130,131

Although prebiotics/probiotics have clear constituents and controlled quality, FMT rapidly restores the composition and function of intestinal microbes in GVHD. In murine models, mice given FMT from healthy donors were characterized by reduced immune cell attack and GVHD severity.¹²⁹ Moreover, growing evidences suggest that either donor or third-party FMT is safe and effective for patients developing GVHD.^132–138 Most patients achieved clinical symptom improvement, and nearly 50–70% of the patients acquired clinical remission. Fecal samples after FMT displayed increased diversity and higher levels of beneficial bacteria. Microbiome functionality related to lipids, carbohydrates, amino acids, nucleotides and energy metabolism was also restored by FMT.¹³⁹ Additionally, only a small population of patients developed infection events that might have resulted from FMT. DeFilipp et al. reported that 1 out of 13 patients developed C. difficile colitis and 1 developed bacteremia.¹³³ In another study, the infection rate within the first month after FMT was 5/17, whereas none of the pathogens was detectable in the donor fecal sample.¹³⁴ Patients given FMT did not experience a higher infection rate compared to those without FMT (5/23 vs 7/18).¹³⁵

Moreover, patients suffering from tumors have massively different microbial structures from healthy people,^25–28 some commensals (such as Bifidobacterium, Ruminococcaceae and Bacteroides) promote the anti-tumor effect of immune checkpoint inhibitors^140–142 or correlate with immune responses in chimeric antigen receptor T cell (CAR-T) therapies,^143,144 suggesting that some species may fuel tumor growth or become powerful weapons against malignant cells. However, little research has explored the roles of these microbes in disease recurrence following HSCT. Only Peled, et al. reported a higher ambulance of Eubacterium limosum between day 0 and day 21 negatively correlated with relapse.¹⁴⁵ What’s the difference in microbial signature between relapsed patients and non-relapsed patients? Whether modifications of intestinal microbiome after HSCT improve the GVL effect require further research.

Outstanding issues and future perspective in intestinal GVHD

Despite progress with remodeling intestinal ecosystem in the treatment of GVHD, there are still many unknowns about the damaged ISC niche and its regulatory network. Except for the sharp decreased ISC pool in GVHD, the functional impairment of ISC may vary widely from mild cases to fatal outcomes, leading to distinct quantity, composition even function of IECs. The altered extracellular signal molecules, the disturbance of intracellular signal pathways, and the imbalance between cell proliferation and differentiation of ISC need to be answered in future research.

Additionally, 16S rRNA sequencing and metabolomics reveal that dysbiosis and dysregulated metabolites are closely related to the development of GVHD. However, intestinal flora intervention, especially prebiotics and probiotics, sometimes fails in GVHD treatment. One possible reason is that dysbiosis is not only affected by external factors but also by the internal interactions between different microbes. Therefore, not all aberrations in intestinal microbes are necessarily pathogenic. Selective removal or supplement of certain bacteria might not have a strong and long-lasting effect, as it could be interfered by other species or environmental factors. Uncovering interactions within microbiota community and remodeling microbial structure may ultimately improve microbial function and attenuate intestinal GVHD.

Conclusions

The intestine should not be merely considered as a target of immune attack. Many factors, especially epithelial damage and dysbiosis, are upstream regulators that influence the incidence and severity of GVHD. Since immunosuppression only partly resolves this disease, remodeling the intestinal ecosystem may overcome drug resistance and improve the therapeutic effect.

Author contributions

He Huang and Yanmin Zhao designed; Fei Gao performed the literature search and wrote this paper. All authors revised and approved the final manuscript.

Acknowledgments

Figures were created using BioRender (http://www.biorender.com). We really appreciate Runnan Chen for polishing the English of the final version.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (82170210) and the National Key R&D Program of China (2022YFA1103500).