Gut Microbiota and Liver Regeneration: A Synthesis of Evidence on Structural Changes and Physiological Mechanisms

Abstract

Liver regeneration (LR) is a unique biological process with the ability to restore up to 70% of the organ. This allows for the preservation of liver resections for various liver tumors and for living donor liver transplantation (LDLT). However, in some cases, LR is insufficient and interventions that can improve LR are urgently needed. Gut microbiota (GM) is one of the factors influencing LR, as the liver and intestine are intimately connected through the gut–liver axis. Thus, healthy GM facilitates normal LR, whereas dysbiosis leads to impaired LR due to imbalance of bile acids, inflammatory cytokines, microbial metabolites, signaling pathways, etc. Therefore, GM can be considered as a new possible therapeutic target to improve LR. In this review, we critically observe the current knowledge about the influence of gut microbiota (GM) on liver regeneration (LR) and the possibility to improve this process, which may reduce complication and mortality rates after liver surgery. Although much research has been done on this topic, more clinical trials and systemic reviews are urgently needed to move this type of intervention from the experimental phase to the clinical field.

The liver has a unique capacity to regenerate, with the ability to restore 2/3 of its mass with functional tissue.^1,2 Researchers are using an experimental model of liver regeneration (LR), partial hepatectomy (PHx), to study the details of this process. Up to two-thirds of the liver is removed during this surgery, as LR is inhibited when a larger amount of liver is harvested due to insufficient residual liver mass.^{3, 4, 5} Human liver regeneration (LR) studies have been based on partial hepatectomy for liver tumors or liver transplantation from living donors. It is now known that LR consists of priming, progression, and termination phases.^{6, 7, 8} Immediately after surgery, portal venous pressure increases, resulting in shear stress on hepatic sinusoidal cells⁹; besides, portal flow contains various signaling molecules that initiate LR.

After degradation of extracellular matrix proteins, modulation of gene expression results in release of various growth factors, signaling molecules, interleukins, etc., leading to hepatocyte proliferation or liver progenitor cells, depending on the type of liver injury.^10,11 Failed LR causes postoperative hepatic dysfunction,^12,13 therefore, understanding certain mechanisms that initiate and maintain LR and hepatic homeostasis is very important.¹⁴ Numerous cells, signaling and metabolic pathways, molecules, growth factors, immune responses orchestrate the process of LR and may be impaired by various factors, but in the current review we will focus on the role of gut microbiota.^{15, 16, 17, 18, 19}

Gut Microbiota And Gut–Liver Axis

Since 1998, the term “gut–liver axis” has been used to refer to the bidirectional communication between the liver and gut microbiota.^20,21 The gut–liver axis maintains host homeostasis via the integrity of the intestinal barrier, the transfer of certain microbes and their metabolites and intestinal contents via portal vein flow, and immune responses resulting in differential host responses.^{22, 23, 24, 25} The liver and intestine are anatomically connected by portal venous flow (which accounts for 70% of total hepatic blood flow) and the common bile duct.^26,27 Bile acids, bacterial metabolites, various toxins, and nutrients flow through these anatomical pathways, resulting in multiple cellular responses in the liver and intestine.^{28, 29, 30} A healthy liver and immune system can easily cope with this stress,^21,31 while various diseases may disrupt the gut microbiota (GM) and lead to dysbiosis, impaired intestinal barrier function, and increased translocation of pro-inflammatory mediators.^23,32

There are more than 100 trillion microorganisms representing gut microbiota, a “new virtual metabolic organ” as it is now well worthy called by some authors.^26,33,34 It has been widely demonstrated that gut microbiota regulates multiple functions in the host body, affecting mental health, gastrointestinal tract, endocrine, respiratory, cardiovascular, and reproductive systems, and the research pool is still widening.^{35, 36, 37} Gut microbiota contains bacteria, fungi, viruses, eukarya, archaea, and protozoa that cooperate with each other and environmental factors,^{38, 39, 40} gut microbiome represents their genome (which contains 150 times more genes than the human genome).^{41, 42, 43}

The majority of GM representatives reside in distal segments of the intestinal tract, in healthy individuals Bacteroidetes and Firmicutes represent 90% of GM with smaller amounts of Proteobacteria, Actinobacteria, Verrucomicrobia, Lactobacilli, Fusobacteria, Cyanobacteria, Streptococci and Escherichia coli.^{43, 44, 45} Most studies of the intestinal microbiota explore the bacterial compartment, although many recent studies have focused on fungi and viruses, so our knowledge of genetically engineered microbe is gradually growing.⁴⁵ Feces contain 10⁵–10⁶ fungal cells per gram, dominant phyla are Ascomycota and Basidiomycota, while Candida, Saccharomyces, Penicillium, Aspergillus, and others are present in smaller amounts but also form the core microbiota.^{45, 46, 47, 48} Within the archaea, Methanobrevibacter smithii Haloferax alexandrinus and Haloferax massiliensis spp. predominate in the gut.^43,49 The gut virome includes eukaryotic and procaryotic viruses, with bacteriophages representing 90% of the total gut virome. The most common taxa include the families Ackermannviridae, Chaseviridae, Herelleviridae and others.^{50, 51, 52}

Liver dysfunction leads to disturbances in immunity and intestinal metabolism, and dysbiosis can lead to progression of liver disease.^28,53 For example, in non-alcoholic fatty liver disease (NAFLD), alcoholic liver disease (ALD), cholestatic liver diseases are associated with impaired GM that worsens during disease progression.^45,54 There are decreased Prevotella and increased Bacteroides and Ruminococcus in patients with non-alcoholic steatohepatitis (NASH) compared to controls, and higher amounts of Escherichia genus and Bacteroides vulgatus in NAFLD and NASH patients. Increased Enterobacteriaceae and decreased Lactobacillus and Bacteroidetes were observed in ALD patients. Cirrhotic patients have higher abundances of Veillonella, Megasphaera, Dialister, Atopobium, and Prevotella.^24,29,53

It has been shown that “healthy” GM should have a certain microbial composition, as microbes produce different patterns of molecules that affect the host's immune system, intestinal epithelium, gene expression, and many more, thus disturbances in GM lead to various changes, including LR impairment.^{41,55, 56, 57, 58} GM composition is formed with the host's cellular and humoral immune system, diet, medication intake, intestinal diseases, liver diseases, etc., therefore disturbances in GM modulators can lead to impaired LR resulting in liver failure.^{33,59, 60, 61}

Gut Microbiota Influence On Liver Regeneration

The role of gut microbiota (GM) in liver regeneration (LR) was first suggested in 1990 when Cornell et al. demonstrated depressed LR after partial hepatectomy (PHx) in germ-free mice.⁶² The authors hypothesized that the reduction in LR was explained by decreased cytokine release after surgery due to the absence of gut-derived endotoxin. In 1995, Miyazaki et al. demonstrated depressed LR with delayed hepatocyte proliferation after PHx in mice that underwent concomitant bowel resection.⁶³

Yin et al. studied the role of gut microbiota in LR after PHx in mice.⁶⁴ The first group of mice received broad-spectrum antibiotics for 3 days prior to PHx, resulting in dysbiosis (there was a significant increase in fecal aerobic colony-forming units and a reduction in total bacterial load compared to the control group). Proteobacteria were predominant on Firmicutes and Bacteroides in dysbiotic mice with Gammaproteobacteria prevalence, whereas a high abundance of Firmicutes and a low abundance of Proteobacteria were seen in controls prior to PHx. Antibiotic-treated mice showed significantly decreased LR after PHx compared to controls, and dysbiotic mice showed significantly higher mortality.

After the PHx the marker of hepatocyte proliferation Ki67 was decreased and critical cell cycle regulators such as cyclin D1, cyclin E1, cyclin A2 and cyclin B1 showed delayed activation in the dysbiotic group, whereas cell cycle inhibitors p21, p27 and p15 were increased. The authors suggested that the reason for impaired regeneration in antibiotic-treated mice was a significant reduction in short-chain fatty acids (SCFAs), especially acetate, resulting in a significant reduction in phospholipid biosynthesis in the liver (especially phosphatidylinositol and phosphatidylglycerol), which are crucial for hepatic membrane synthesis. Microscopically, hepatocytes tend to be smaller in the dysbiotic group. The authors also showed significant impairment of LR and increased mortality in germ-free (GF) mice with its improvement after intestinal recolonization.⁶⁴

Antibiotic cocktail was also administered to mice prior to PHx by Wu et al.⁶⁵ These mice showed reduced LR at 24, 48, and 72 h and 7 days after PHx compared to the control group. The researchers found that ampicillin alone had the most dramatic effect on LR, suggesting that ampicillin-sensitive bacteria, including a variety of gram-positive and gram-negative species, play a role in LR. Certain mechanism of LR promotion by GM seemed to be the regulation of natural killer T cells (NKT cells) by bacteria, in particular, prevention of its over-activation by maintaining Kupffer cells in a tolerant state.

Similar results with germ-free (GF) mice were shown by Amin et al.⁶⁶ The investigators examined LR and hepatic matrisome (non-fibrillar proteins of the extracellular matrix that serve as a reservoir for growth factors and cytokines necessary for liver homeostasis and appropriate response to stress and injury) after partial hepatectomy (PHx) in GF mice, GF mice recolonized with normal GM (ex-GF mice), and controls.⁶⁷ LR was impaired in GF mice with reduced hepatocyte regeneration as assessed by Ki67 and hematoxylin and eosin (H&E) staining.

Hepatocyte growth factor (HGF) levels were reduced in GF mice, hepatic matrisome gene expression pattern of GF mice was different from GF mice after recolonization and controls with lower expression of differentially expressed genes (DEGs): 69 (47 upregulated and 22 downregulated genes) at 3 h and 71 (46 upregulated and 25 downregulated genes) at 72 h post PHx, whereas expression of DEGs in ex-GF mice was 140 (106 upregulated and 22 downregulated genes) at 3 h and 139 (111 upregulated and 28 downregulated genes) at 72 h post PHx. In particular, GF mice showed lower levels of collagen gene expression, which is required for the restoration of normal liver architecture. Ex-GF mice showed improved LR, comparable to the LR of controls. The authors concluded that gut microbiota (GM) plays a role in liver regeneration (LR) and interacts with the hepatic matrisome, noting that LR is a complex process and the true importance of GM remains a question.

In 2020, Leng et al. investigated the proteins transferred to the liver by portal flow and their role in LR.⁶⁸ The investigators performed PHx in mice that had received ampicillin, neomycin, metronidazole, and vancomycin for 1 week prior to surgery. The mice were divided into three groups according to the postoperative interventions: mice from the control group received bottled water, while the second group received antibiotics, and the third group received antibiotics and probiotics (Bifidobacterium and Lactobacillus) for 2 weeks, and their portal vein (PV) serum were collected 48 h after PHx. 1271 proteins were found in PV serum, of which 406 were upregulated at 48 h after surgery, but 139 of them were not significantly upregulated in mice with dysbiosis.

In addition, probiotic-fed mice showed that 95 of these 139 proteins were upregulated in mice after probiotic treatment. These 95 proteins were considered to be potential promoters of LR and mostly included extracellular matrix proteins (receptors, regulators, contact-related proteins, promoters of angiogenesis, etc.) and some signaling pathway proteins (phosphatidylinositol 3-kinase– protein kinase B [PI3K-Akt], nuclear factor kappa B [NF-Κb], Janus kinase (JAK)/signal transducer and activator of transcription [JAK-STAT]). 524 proteins were found to be downregulated after 48 h and 35 of them were upregulated in mice with dysbiosis and the authors considered them as potential inhibitors of LR. Among these 35 proteins, proteins related to lipid metabolism were enriched. These proteins may play a negative role in the first 48 h of LR, as lipid accumulation in the liver leads to steatosis and oxidative stress. Proteins related to acute and immune responses were also represented.

The above-said demonstrates the changes in level of different proteins and cytokines synthesis during liver regeneration (LR) depending on gut microbiota (GM) state. For example, interleukin 6 (IL-6) and tumor-necrosis factor alpha (TNF-α), major mediators of the LR priming phase,^{69, 70, 71, 72} have been shown to be reduced in germ-free mice, as their release is influenced by GM metabolites such as lipopolysaccharide (LPS) from Gram-negative bacteria (Escherichia coli, Proteus, Pseudomonas aeruginosa, etc.).⁷³ LPS is known to play a role in LR, as germ-free mice, LPS neutralization and induction of LPS tolerance significantly impair LR in rats. LPS enters the systemic circulation through the portal vein and induces hepatotrophic factors such as insulin, glucagon, epidermal growth factor (EGF), and also activates Kupffer cells, stimulating the production of IL-6 and TNF-α. IL-6 then activates the production of hepatocyte growth factor (HGF), inducing the onset of the priming phase of LR.⁷⁴ It was also shown that HGF alone resulted in 4.5 times less hepatocyte replication compared to HGF combined with LPS in mice after partial hepatectomy (PHx).⁷⁵ Although the beneficial role of LPS is clear, it also has a negative effect on the liver, as it can over-activate TLR4 and induce liver injury, fibrosis and tumor formation, so it cannot be safely used as a therapeutic molecule for LR stimulation, while GM modulation for achieving appropriate LPS levels seems to be a possible strategy.⁷⁴

The termination phase of LR depends mainly on IFN-γ and transforming growth factor beta (TGF-β). IFN-γ inhibits the proliferation of hepatocytes and activates their apoptosis. IFN-γ synthesis can be activated by Fusobacterium nucleatum subsp. Polymorphum via its outer membrane vesicles, whereas Christensenellaceae, Lactobacillus, Bifidobacterium bifidum and Parabacteroides distasonis downregulate this cytokine. TGF-β plays anti-proliferation role, its level can be decreased by Lactobacillus acidophilus and Lactobacillus salivarius, whereas Bifidobacterium longum, Bifidobacterium lactis, Lactobacillus johnsonii and some Clostridium species can promote TGF-β release.⁷³

Therefore, different microbes affect different LG phases and should gradually change during the process, which was confirmed in 2015 by Liu et al.⁷⁶ Authors investigated GM changes during LR in mice after PHx. It was shown that Bacteroides and Firmicutes were the most abundant phyla before surgery, followed by a gradual increase in Bacteroides and decrease in Firmicutes. The abundance of Ruminococcus, Bifidobacterium, Lactobacillus, Clostridium peaked 1 h and 7 days after PHx. Although further studies in animals and humans are needed, similar changes in Bacteroides and Firmicutes were observed in two different studies, which may have a therapeutic opportunity.

Interestingly, the authors found that Clostridiaseae, Lachnospiraceae and Ruminococcaceae were associated with the same gene expression in mice after PHx, which means that these bacteria have similar biological functions during LR. Moreover, Lachnospiraceae and Ruminococcaceae had a very similar pattern of correlations with the genes, while the pattern of Bacteroidetes S24-7 family was opposite. It was also shown that specific microbiota representative had correlation with the genes responsible for host metabolism and immune response.⁷⁶

In 2019 by Bao et al. also investigated changes in the gut microbial community during liver regeneration (LR) in rats undergoing partial hepatectomy (PHx) and sham surgery (control group).⁷⁷ Some rats from the PHx group received ampicillin, vancomycin, metronidazole, and neomycin for 4 weeks prior to PHx for selective gut decontamination, and another group of rats then underwent fecal microbiota transplantation (FMT) with 1 g fecal suspension per 1 kg for 7 days. The authors showed a shift from Firmicutes to Bacteroides during LR and an increase in Proteobacteria in the first 48 h. Analysis of 16 S rRNA sequences revealed the prevalence of bacteria with metabolic functions (carbohydrate, glycan and lipid metabolism) in the first 12 h after PHx, whereas in 48 h after PHx there was a prevalence of bacteria whose metabolites can induce cellular proliferation (cofactors, vitamins, pyrimidine, purine, nicotinate and nicotinamide). Rats that were exposed to antibiotics showed significantly depressed LR, but fecal microbiota transplantation (FMT) led to its normalization.

Bacterial Bioactive Metabolites In Liver Regeneration

Gut microbes metabolize primary BAs (cholic acid and chenodeoxycholic acid) via deconjugation, oxidation/epimerization, (7-α-) dehydroxylation, and esterification, producing secondary BAs and modulating the affinity of BAs for their receptors.^{78, 79, 80} Deconjugation occurs via bile salt hydrolases (BSH) found in Firmicutes, Bacteroidetes and Actinobacteria, leading to better reabsorption of BAs and bacterial nutrition (source of sulfur, nitrogen and carbon).⁸¹ The most important secondary BAs are ursodeoxycholic acid (UDCA), deoxycholic acid (DCA) and lithocholic acid (LCA).⁵⁵ The production of LCA and DCA depends on 7α-dehydroxylation by Clostridium and Eubacaterium.^74,82

Both primary and secondary BAs can modulate gene expression, control BAs synthesis through the farnesoid X receptor (FXR), inhibit the bile salt export pump, and LCA can also induce the expression of genes involved in xenobiotic detoxification.^55,78 LCA also controls BA absorption by regulating the apical sodium-dependent BA transporter and basolateral transporters. Thus, Lactobacillus and Lactobacillaceae convert primary BAs into secondary and deconjugate to free BAs, whereas Enterococcus, Erysipelatrichales and Enterobacteriales have the opposite effect on the bile acids (BAs) profile.

Another study showed the correlation between Chenodeoxycholic acid (CDCA) and Enterobacteriaceae, DCA and Ruminococcaceae, DCA/CA ratio and Ruminococcaceae, LCA/CDCA ratio and Blautia.⁷³ However, elevated levels of BAs are toxic to the liver, so DCA can induce hepatocyte apoptosis, oxidative stress, so a normal level of gut microbes is also needed for proper BAs metabolism.⁵⁵ In cirrhotic patients correlations between the level of primary BAs with hydrophobic sites and Enterobacteriaceae and Ruminococcaceae growth.⁷⁵

The effect of BAs on liver homeostasis and LR has been demonstrated in numerous studies.⁸³ Hepatocyte proliferation can be triggered by farnesoid X receptor (FXR) activation, FGF15/FGF19 secretion, which bind to fibroblast growth factor receptor 4 (FGFR4)/β-klotho complex on hepatocyte membranes.⁸⁴ For example, cholestyramine-induced BAs depletion resulted in suppressed LR in mice; patients with external biliary drainage who underwent major hepatectomy had worse LR rates compared to controls.⁷⁴ A retrospective cohort study by Pravisani et al. showed a significantly higher prevalence of early allograft failure in patients with T-tube who underwent liver transplantation (LT), regardless of graft quality and pre-existing liver disease, because T-tube led to external bile drainage and prevented passage of BAs into the intestine.⁸⁵ Secondary BAs play a role of TAKEDA G protein-coupled receptor 5 (TGR5) agonists, leading to prevention of BAs overload, reduced hydrophobic BAs pool, suppression of LPS-induced cytokine release in the liver.⁷⁴

BAs are necessary for nutrient absorption, although the effect depends on their concentration, hydrophilic/hydrophobic ratio, as BAs may also cause damage to the liver under certain circumstances.^80,86 The hydrophilic/hydrophobic ratio reflects the concentration of hydrophilic BAs, such as UDHA, and hydrophobic bile acids (BAs), including lithocholic acid (LCA), chenodeoxycholic acid (CDCA), deoxycholic acid (DCA), taurine conjugated BAs, and so on. The more hydrophobic BA is, the more toxic effects it has, so hydrophilic and hydrophobic BAs should be balanced to maintain GI health.⁸⁷ Hepatic resection leads to BAs overload of the liver and may cause oxidative stress, which prevents with the activation of expression of multidrug resistant protein 2 and 3 (Mdr2 and Mdr3), bile salt export pump (Bsep) after PHx. Moreover, activated expression of small heterodimer partner (SHP) nuclear receptor, fibroblast growth factor 15 (FGF15) lead to BA synthesis arrest for 48 h after surgery.⁷⁶

Liu et al. demonstrated the interaction between gut microbiota (GM) and bile acids (BAs) in liver regeneration (LR).⁸⁸ There were 2 groups of mice exposed to partial hepatectomy (PHx): the first group received retinoic acid (RA) 48 h before surgery, the second group was control. RA-primed mice were shown to have a higher rate of hepatocyte regeneration with increased levels of Ki67-positive cells and the absence of lipid droplets normally seen in regenerating liver. The difference in GM composition in both groups involved Verrucomicrobia phyla (Akkermansia), which increased on Day 1 and became undetectable on Day 2 in controls, whereas it was detected on Day 1.5 in RA-primed mice.

In addition, RA-primed mice showed a 4-fold increase in Lactobacillus immediately after surgery, a 2-fold increase in Bacteroides, and a 9-fold increase in Bifidobacterium on day 1 after PHx, with a prevalence of Ruminococcaceae and Lachnospiraceae in the fermicutes on Day 2 compared to controls. At this time, RA-primed mice had a shift in BAs during LR with increased hydrophilic/hydrophobic ratio, solubility and lipid emulsification, and improved BAs recirculation. The authors suggested that RA treatment led to a GM-modulated BA composition towards hydrophilic and conjugated BAs, improved BAs recirculation and lipid metabolism, resulting in enhanced energy metabolism (via more robust fibroblast growth factor 21- liver kinase B1-AMP-activated protein kinase [FGF21-LKB1-AMPK] pathway activation) in the regenerating liver. Thus, gut microbiota (GM) can modulate bile acids (BAs) synthesis and BAs can modulate GM. Conjugated BAs have antimicrobial effects, T-β-MCA induces the population of Lactobacillus, Cholic acid (CA) administration leads to increased Firmicutes, pathogenic Clostridia and Erysipelotrichia populations, which should be mentioned as a possible therapeutic target.^79,86,89

In 2022, Zhang et al. showed that GM also affects LR via the production of H₂, which leads to increased levels of antioxidant enzymes (superoxide dismutase) and decreased levels of malondialdehyde (MDA), IL-6, and TNF-α, resulting in accelerated LR in mice.⁹⁰ H₂ levels depend on the balance between H₂-producing bacteria (such as Bacteroides and Firmicutes) and H₂-consuming bacteria, which means that gut dysbiosis-associated alteration of LR also lies in the depletion of H₂ levels.

In 2022, certain metabolites of GM that affect LR were identified by Oh et al. using network pharmacology.⁹¹ Twelve bacterial metabolites were shown to be related to AKT1 (RAC-alpha serine/threonine protein kinase), one of the major accelerators of LR, including flavonoids (representing 50% of the metabolites), coumarins, indoles, steroids, and carboxylic acids. The authors revealed a significant role of Escherichia sp. in LR via inhibition of signaling pathways.

Critical to the host functions of GM are the biosynthesis and metabolism of some essential vitamins that our bodies are capable of producing. For example, germ-free mice are deficient in vitamins K2 and B. Vitamins B are known to modulate GM because they are essential for B cell immune metabolism, host-microbe drug co-metabolism, and microbial fitness, and in the liver they prevent inflammation, improve hepatocyte survival, and control lipid and glucose levels. Vitamin K2 has been shown to play a role in the differentiation and maturation of human pluripotent stem cell (hPSC)-derived hepatocytes in vitro, so it may play a role in improving LR.⁵⁵

Important bacterial metabolites include short-chain fatty acids (SCFAs), the product of carbohydrate and protein fermentation, among which acetate, propionate and butyrate account for about 95% of SCFAs, regulate intestinal barrier integrity, provide energy to colonocytes, etc. and are transported to the liver through the portal flow. Acetate and propionate are mainly produced by Bacteroidetes, while butyrate is mainly produced by Firmicutes.⁷³ In the liver, SCAFs regulate hepatic metabolism and may also play a role in stem cell specification according to in vitro studies.⁵⁵ SCFAs are known to be involved in LR. Acetate, propionate, and butyrate promote the differentiation of T-reg cells that maintain the intestinal mucosal epithelial barrier, limit LPS translocation, and thereby inhibit inflammation in the regenerating liver.

SCFAs may also activate JNK1, which has been shown to facilitate the activation of cyclin D1, a promoter of hepatocyte proliferation.⁷³ During LR, SCFAs regulate hepatic glucose and lipid homeostasis via peroxisome proliferator-activated receptor gamma (PPAR-γ) pathway, improve metabolic phenotype through GLP-1 and peptide YY production, acetate and propionate reduce hepatic free fatty acid flux, propionate and butyrate activate gluconeogenesis. Butyrate also inhibits macrophage activation in the lamina propria, prevents nuclear factor kappa B (NF-κB) activation and ischemia-reperfusion injury.⁷⁴ Other important products of bacterial fermentation include indoles, skatole, and tryptamine, which are known to improve mucosal defense and decrease intestinal permeability,⁹² as well as reduce inflammation in the liver (via modulation of the NF-κB pathway).⁵⁵

Gut Microbiota Modulation And Liver Regeneration

There is a possibility to modulate the composition of the gut microbiota either with a special diet or with the use of pre-, pro-, synbiotics or even fecal microbiota transplantation. Several liver diseases have been shown to be improved with GM modulation,^{93, 94, 95, 96, 97} however the effect of GM modulation on LR is not clear due to very limited research.⁹⁸ We could find 5 clinical trials, 1 systematic review and 3 studies in mice.

The first randomized double-blind pilot study was conducted in 2012 by Rayes et al. and demonstrated a positive effect of pre and probiotics on LR in 19 patients undergoing right hepatectomy. Patients were divided into two groups, the first group received synbiotic containing four lactic acid bacteria and four bioactive fibers twice daily. The treatment was started the day before surgery and continued for 10 days after surgery. The second group received only the fibers. It was shown that the functional capacity of the liver was significantly increased in the first group on the 3^rd and 5^th days after surgery, although patients with postoperative complications had impaired LR in both groups.⁹⁹

The only systematic review and meta-analysis on this topic was performed by Gan et al. and included 4 studies and the study by Rayes et al. with a total of 205 participants (103 treated with pro- or prebiotics and controls).¹⁰⁰ Probiotics were administered on the day of surgery in 3 studies and on the day before surgery in one study. It was shown that the administration of pro- and/or prebiotics one day before surgery reduces the postoperative infection rate, shortens the duration of hospitalization and the use of antibiotics.

Huang et al. also demonstrated a beneficial effect of probiotics in 50 Child-Pugh grade B patients undergoing partial hepatectomy for hepatocellular carcinoma (HCC) or hemangioma.¹⁰¹ Patients were divided into control and probiotic groups, with the second group receiving Clostridium butyricum for 3 days before surgery and 4 days after surgery. The probiotic group showed decreased early bacterial infection rates and promoted recovery of liver functions. However, Child-Pugh A patients did not benefit from probiotic treatment compared to the control group. This study is a single-center study, further research is needed.

Xie et al. investigated the role of probiotics on LR after liver surgery in mice.¹⁰² Lactiplantibacillus plantarum AR113 at a dose of 1 × 1010 CFU/mL was administered to mice 14 days before PHx and continued until 3 or 7 days after surgery. Pretreatment with L. plantarum AR113 resulted in increased hepatocyte proliferation (as determined by Ki67 staining) at 3 days post-PHx, reduced mortality, and accelerated regeneration rate at 7 days post-PHx compared to controls. Researchers found increased serum levels of tumor-necrosis factor alpha (TNF-α) and hepatocyte growth factor (HGF) at 3 days post PHx, and increased levels of tumor-necrosis factor alpha (TNF-α), interleukin 6 (IL-6), transforming growth factor beta (TGF-β), promoting LR. Pretreatment with L. plantarum AR113 also increased levels of l-leucine, which has been reported to have a proliferative effect on hepatocytes.

The beneficial effect of probiotics on LR after PHx was also demonstrated by Cohen et al.¹⁰³ The researchers performed PHx and administered microbiota and probiotic membrane vesicles (pbMVs) to mice at a dose of 0.5 mg/kg, which resulted in a significant decrease in the hepatic neutrophil-to-lymphocyte ratio Ly6G+/CD3+, increased levels of IL-6, decreased levels of IL-10, vascular cell adhesion molecule-1 (VCAM1), and intercellular adhesion molecule-2 (ICAM2), resulting in improved LR at 72 h after PHx compared to controls.

Fecal microbiota transplantation (FMT) refers to the transfer of healthy donor stool to the recipient to restore eubiosis. FMT has recently become one of the hottest topics in medical research, as it was shown to be the answer not only to recurrent Clostridioides difficile infection (rCDI), but also to other gastrointestinal and non-gastrointestinal diseases.^{104, 105, 106, 107} Many studies have shown that various liver diseases may also improve after FMT. For example, clinical trials on patients with NAFLD, ALD, primary sclerosing cholangitis showed the beneficial effect of FMT.^{108, 109, 110, 111} In addition, FMT can improve the metabolism of BAs, which is known to have a positive effect on LR.¹¹² It is still unknown whether FMT can improve LR as relevant clinical studies are scares, however there are researches on animal models and some clinical cases that prove that FMT may be used in immunosuppressed patients after LT.

Meyerovich et al. demonstrated improved LR after FMT from healthy mice in PHx mice pretreated with antibiotics. In addition, FMT from obese mice was associated with decreased LR.¹¹³ Schneider et al. demonstrated a successful case of FMT in a patient with severe therapy-refractory Clostridioides difficile infection (CDI) who underwent LT.¹¹⁴ This case demonstrated that FMT is safe for patients on immunosuppression. There is also a systematic review confirming the statement that FMT is safe for immunosuppressed patients undergoing solid organ transplantation.¹¹⁵ We summarize the evidence and mechanisms in Table 1, Table 2 and Figure 1.

Table 1.

Summary of Research on the Impact of the Gut Microbiome on Liver Regeneration.

Author	Materials and methods	Results	Conclusion
Cornell et al.62	Male germ-free and control euthymic and athymic mice underwent Partial (67%) hepatectomy.	Depressed LR after PHx in germ-free mice.	Insufficient gut-derived endotoxin in germ-free mice lead to impaired cytokine release by macrophages and/or monocytes after acute liver injury by partial hepatectomy and impaired LR.
Miyazaki et al.63	264 male Wistar rats underwent simultaneous 70% PHx and Bowel resection.	Suppressed hepatic DNA synthesis, increased LPS levels and depressed LR, compared with control group.	LR might be suppressed by excessive endotoxin levels after bowel resection.
Yin et al.64	PHx was carried out in mice treated with broad-spectrum antibiotics, in germ-free mice and mice colonized with minimal microbiota.	Antibiotic treatment leads to decreased SCFAs, delayed LR, increased mortality	GM regulates hepatic membrane phospholipid biosynthesis and LR.
Amin et al.66	Partial hepatectomy (PHx) in germ-free (GF) mice, and in GF mice reconstituted with normal gut microbiota (XGF).	Reduced liver mass restoration in germ-free mice at 72 h post PHx.	Reduced liver mass restoration is suggested due to impaired immune response. Interaction of GM and matrisome may play a significant role in hepatic remodeling during LR.
Leng et al.68	Three groups, each consisting of 10 male mice. After PHx, mice received bottled water (control group), antibiotic-containing water, or antibiotic- and probiotic-containing water.	Probiotic supplementation restored liver proliferation and function in mice with intestinal flora disorder after PHx. 95 and 35 proteins potentially promoting or inhibiting liver regeneration, respectively, were identified.	LR can be improved by altering the intestinal flora or intravenously administering potential regeneration-promoting proteins.
Bao et al.77	Male Sprague–Dawley rats were subjected to surgical removal of 2/3 of the liver and samples were collected over a 14-day period.	Partial hepatectomy caused fluctuating changes in gut microbiome, which paralleled the biological processes of LR.	Gut flora may play a central role in manipulating metabolic responses in LR.
Wu et al.65	Two-thirds (approximately 67%) PHx was performed in male mice.	LR was impaired in antibiotic water–treated mice. Increased number of Kupffer cells and higher IFN-γ and Il-12 were observed in ATB-treated mice.	Commensal bacteria play a critical role in maintaining Kupffer cells in a tolerant state, preventing subsequent NKT cell over-activation during liver regeneration.
Liu et al.76	n/a	Partial hepatectomy led to rapid changes in gut microbiota. Specific bacteria were closely associated with the concentration of certain bile acids and expression of hepatic genes.	The presented data established, for the first time, an intimate relationship between intestinal microbiota and the expression of hepatic genes in regenerating livers.
Oh et al.91	Exploring the complex microbiome and its network with phytochemicals via public databases.	12 metabolites related to a key target (AKT1) of LR were found.	This study demonstrated promising effectors (AKT Serine/Threonine Kinase 1 [AKT1] and mitogen-activated protein kinase 1 [MAPK1]) in the treatment of LR.

Table 2.

The Effect of Gum Microbiome Modulation on Liver Regeneration.

Author	Materials and methods	Results	Conclusion
Rayes et al.99	Randomized double-blind pilot study. 19 patients underwent right hepatectomy and were divided into 2 groups: a group receiving a combination of four probiotics and four fibers and a placebo group receiving the fibers only starting the day before surgery and continuing for 10 days.	Liver function capacity was better in the patients with synbiotics. No severe side effects occurred.	Synbiotics might be able to increase liver function capacity in patients after liver resection, but patient numbers were too small and the clinical courses too heterogeneous to draw any definite conclusions.
Gan et al.100	Systematic review and meta-analysis. 205 patients underwent hepatic resection.	Wound infection rate in the probiotic group was significantly lower than that in the placebo group. Furthermore, probiotics/prebiotics decreased the duration of hospital stay.	Our findings show that administration of probiotics and/or prebiotics prior to operation day decreases the infection rate post-liver resection and could shorten the duration of hospitalization and antibiotics use.
Huang et al.101	100 patients (50 cases in Child-Pugh A grade and 50 cases in Child-Pugh B grade) underwent hepatectomy. Subsequently, Child-Pugh A grade and Child-Pugh B grade patients were set as probiotics group and control group.	In Child-Pugh B grade subgroup patients, the procalcitonin, alanine aminotransferase, and prothrombin time of the probiotics group were statistically significantly lower than that of the control group. In Child-Pugh A grade subgroup patients, there were no significant differences between probiotics group and control group after operation.	Child-Pugh B grade subgroup patients taking probiotics after hepatectomy could reduce postoperative infection and accelerate recovery of liver function.
Xie et al.102	Sprague–Dawley (SD) rats were gavaged with L. plantarum AR113 suspensions (1 × 1010 CFU/mL) both before and after partial hepatectomy.	L. plantarum AR113 administration 2 weeks before partial hepatectomy can accelerate liver regeneration by increased hepatocyte proliferation and tumor necrosis factor-α (TNF-α), hepatocyte growth factor (HGF), and transforming growth factor-β (TGF-β) expression.	The colonization of L. plantarum AR113 enriched the normal gut microbiota, especially Lactobacillus. The decrease in potential pathogens and restoration of the normal gut microbiota by L. plantarum AR113 might improve host metabolism and accelerated liver regeneration.
Cohen et al.103	Probiotic-derived vesicles (pbMVs) were isolated and characterized from probiotics, mainly from the bacteria genus Lactobacillus, Bifidobacterium, and Lactococcus. Mice underwent bile duct ligation, sham laparotomy (SHAM), or 70% partial hepatectomy (70%PH).	The administration of pbMV improved liver regeneration 72 h following surgical liver resection with a significant decrease in IL17 expression.	Probiotic vesicles can reduce adhesion molecule expression and affect immune cell invasion and liver immunity, resulting in improved liver recovery following hepatic surgery.
Meyerovich et al.113	70% PH model in mice to induce LR. Intestinal microbiome was manipulated by aggressive antibiotic treatment followed by fecal microbiota transplantation (FMT). Normal, as well as obese mice received FMT by gavage from normal and obese mice.	Fecal microbiota transplantation (FMT) from healthy normal mice, compared to vehicle resulted in improved LR, and that FMT from obese mice was associated with decreased LR and increased postoperative mortality. Obese mice receiving FMT from obese mice had decreased LR.	Microbiota transfer from healthy donor mice enhances LR, microbiota transfer of dysbiotic microbiota from high-fat diet (HFD) donor mice impairs LR, following PH.

Figure 1.

Gut-liver axis in liver regeneration. LR: liver regeneration; CBD: common bile duct; PV: portal vein; pBAs: primary bile acids; sBAs: secondary bile acids; LPS: lipopolysaccharides. A – LR in eubiosis. The eubiotic condition maintains a healthy state in the intestine and liver, with a thick layer of mucus, functioning tight junctions of enterocytes, preventing excessive entry of metabolites into the blood. In eubiosis, bacteria, fungi, viruses, eukarya, archaea and protozoa species predominate in the intestinal lumen and interact with each other. Via the “gut–liver axis”, microbial metabolites such as short-chain fatty acids (SCFAs), secondary bile acids (BAs), fibroblast growth factor (FGF)15/FGF19, lipopolysaccharides (LPS), coumarins, indoles, steroids, carboxylic acids, vitamins K2 and B, H2, interleukin 6 (IL-6) and tumor-necrosis factor alpha (TNF-α) and etc. reach the portal vein and affect hepatocytes, inducing LR. Each stage of LR depends on specific microbial metabolites and is impaired in dysbiosis. B – LR in dysbiosis. An imbalance in the microbial community leads to the proliferation of pathogenic microorganisms, impaired mucus layer, enterocytes and tight junctions, resulting in a “leaky gut”. Excessive amounts of various pro-inflammatory agents (such as LPS, TNF-α, IL-6, interferon gamma [INF-γ], etc.) and primary BAs induce inflammation in the liver. The amounts of SCFAs, vitamins, indoles, secondary BAs and other microbial metabolites are smaller than in the eubiotic stage, which results in insufficient activation of various pathways that activate hepatocyte proliferation and differentiation. Therefore, dysbiosis leads to impaired LR.

Gut Microbiome And Liver Regeneration I Liver Transplantation

Therapeutic strategies to improve LR are urgently needed not only for better recovery after liver resection, but also to improve the outcome of liver transplantation. Studies observing the changes of GM composition in the peri-transplantation period in LT are limited, although we know that various hepatic diseases lead to serious GM disturbances in the terminal stage and also such patients usually require antibiotics, thus liver recipients would definitely have dysbiosis that may disturb the post-LT period.^{116, 117, 118} Therefore, further studies are needed to investigate the role of GM in LT.

At present, the shortage of donor livers is an emerging problem,^119,120 so the acceptance criteria for liver transplantation are broadening and older liver donors, steatotic grafts, viral hepatitis C are no longer the absolute contraindication for liver transplantation (LT).¹²¹ The problem is that even good quality grafts are susceptible to damage via the ischemia/reperfusion (I/R) mechanism resulting from the cessation of circulating blood during organ harvest and transport,^122,123 while non-alcoholic fatty liver disease (NAFLD) livers are even more sensitive to these injurious effects.^124,125 Therefore, one of the main goals to improve liver resection and LT outcomes is to reduce I/R.⁹²

One of the possible strategies to reduce I/R could be an intervention on the gut microbiota. It is well known that NAFLD is associated with changes in the gut microbiota, such as reduced Bacteroidetes, Ruminococcaceae and Faecalibacterium prausnitzii and increased Prevotella, Porphyromas, Lactobacillus, Escherichia and Streptococcus compared to healthy individuals. Hepatic inflammation and progression to non-alcoholic steatohepatitis (NASH) are dependent on an impaired primary/secondary bile acids (BAs) profile, elevated plasma BAs and lipopolysaccharides (LPS) levels, and an impaired intestinal barrier found in dysbiosis.^{126, 127, 128, 129}

Thus, glucagon-like peptide-1 (GLP-1) may be a potential therapeutic target. It has been shown that increased levels of GLP-1 in steatotic rat livers resulted in reduced inflammatory response and liver damage and improved regeneration in donors of cardiovascular death.¹³⁰ It is known that GLP-1 levels increase in the presence of SCFAs, products of bacterial fermentation, so decreased GLP-1 levels in steatotic rats seems obvious. Zhao et al. demonstrated that GLP-1 can modulate the GM composition in obese and diabetic obese rats, resulting in reduced body weight.¹³¹ There were higher levels of Firmicutes and lower levels of Bacteroidetes and an overall reduction in microbial diversity. Treatment with liragludite (GLP-1 receptor agonist) increased the Bacteroidetes/Firmicutes ratio and decreased obesity-related phenotypes such as Romboutsia, Ruminiclostridium, and Erysipelotrichaceae. Therefore, modulation of GM via GLP-1 receptor agonists prior to surgery may lead to better outcomes, although this hypothesis requires further research.⁹²

Branched-chain amino acids (BCAAs) are involved in protein and glucose metabolism and fatty acid oxidation, reduce oxidative stress and hepatic inflammation, and protect against ischemia/reperfusion mechanism (I/R).^92,132,133 Beppu et al. demonstrated a beneficial effect of BCAA supplementation on LR in patients undergoing portal vein embolization (PVE) and liver resection.¹³⁴ Therapy with 4 g of BCAAs, including l-isoleucine, l-leucine, and l-valine, was started before PVE and continued for 6 months after liver resection, and was immediately discontinued if severe adverse events developed. The authors concluded that branched-chain amino acid (BCAA) supplementation improved functional LR, but larger studies are needed. Nanno et al. demonstrated that preoperative BCAA therapy significantly improved blood lactate levels during and after major hepatectomy.¹³⁵ BCAA supplementation included 1 g valine, 2 g leucine, and 1 g isoleucine and was administered 2 h before anesthesia. Lowering blood lactate levels may be helpful in patients at increased risk for oxidative stress.

Recently, it has been shown that BCAAs derived from the gut microbiota can enter the bloodstream and exert systemic effects. In this study, plant-based diet was associated with enrichment of BCAAs and positive correlation with Prevotella, whereas omnivorous diet was associated with Bacteroides, Ruminococcus, and Faecalibacterium and increased SCFAs synthesis.¹³⁶ This means that improving GM via diet or other modifications may lead to increased serum BCAAs and better LR.

Although some recent clinical studies did not find differences in GM in the first seven days post-LT, recipients with post-LT acute cell resection and bloodstream infection, many other studies supported the idea of GM changes during liver surgery that may affect liver inflammation and LR.⁹² Several longitudinal studies showed dysbiosis in patients post-LT and its persistence in recipients. HBV and cirrhotic LT patients had similar GM changes with decreased Feacalibacterium (which have anti-inflammatory effects), increased Enterococcus and Enterobacteriaceae family. Subsequent partial recovery of GM was observed 12–24 months after surgery.¹³⁷ Perioperative GM changes were also demonstrated, such as increased Clostridiales cluster XIVa, Blautia, Akkermansia and Eubacteriaceae family and decreased Enterobacteriaceae family at 3 months post-LT.¹³⁷ Another study demonstrated decreased SDI in the first 7 days post-LT with subsequent improvement in the following 8 weeks. Low SDI was associated with ACR and bloodstream infections.¹³⁸

In 2012, it was also shown that increased Bacteroides was associated with acute rejection as well as increased Clostridium bolteae and decreased F. prausnitzii and Lactobacillius.^139,140 Many studies have shown that LT patients have dysbiosis in the post-LT period and recover in the first month, suggesting that these GM disturbances are associated with end-stage liver disease. In addition, immunosuppressive drugs also modulate GM composition, such that cyclosporine decreases some Clostridium and Enterobacteriaceae pathogenic bacteria, increases Feacalibacterium prausnitzii, and improves graft function.¹⁴¹ Medium doses of tacrolimus maintain a stable GM balance, whereas high doses are associated with increased endotoxemia.¹⁴²

In 2019, Lu et al. investigated the GM composition in liver recipients with normal and abnormal liver function and found a significant decrease in GM diversity, increased Fusobacteria and Proteobacteria, and decreased Firmicutes and Lentisphaerae in the second group, whereas Bacteroidetes were increased in the first group.¹⁴³ Compared to the healthy subjects, an abundance of Negativicutes, Gammaprotobacteria and Bacilli and a decrease of Firmicutes were found in both post-LT groups. In addition, butyrate-producing bacteria were negatively correlated with serum markers of liver dysfunction. The authors concluded that microbiota-directed therapeutic strategies should be considered after LT. A systematic review of 4 controlled trials showed an improved infection rate, decreased length of hospital stay and ICU stay in the groups receiving probiotics (4 studies included Lactobacillus and 2 studies added Bifidobacterium).¹⁴⁴

Dysbiosis leads to fibroblast growth factor 15 (FGF15) reduction via decreased Lactobacillus and Bifidobacterium, producers of secondary BAs, which could also be seen in brain dead donors.¹⁴⁵ Gulfo et al. demonstrated that therapy with FGF 15 resulted in improved LR after LT of brain-dead donor grafts.¹⁴⁶ FGF 15 inhibits BA synthesis and prevents BA overload of the graft, leading to better LT outcomes.^146,147

In 2018, Bajaj et al. studied GM in 40 patients who underwent deceased donor LT.¹⁴⁸ Stool samples were collected before and 7 ± 3 months after surgery. Postoperative stool samples showed a significant reduction in potentially pathogenic genera, increased microbial diversity, and elevated potentially beneficial bacteria levels (Ruminococcaceae and Lachnospiraceae) compared to pre-LT samples. BAs profiles were also evaluated. Total BAs levels were increased after LT with higher levels of secondary BAs. The authors concluded that successful LT resulted in positive changes in GM within 6 months after LT. Therefore, GM disturbances, among others, may affect LT outcomes and should be taken into consideration before surgery, dysbiosis should be corrected before LT and should be under control during post-LT period.

Future Prospects For Studying The Role Of Gut Microbiota In Liver Regeneration

Regulation of LR may lead to significant progress in various liver diseases, so many researchers are investigating this process. Many factors can affect LR and one of them is dysbiosis as shown in the current review. Nowadays, GM remains a potential therapeutic target in various diseases and insufficient LR is one of them.

We describe the results of studies that have demonstrated an important role of GM state in LR, found possible mechanisms of this interaction, and potential therapeutic targets. Now we need to explore whether we can integrate this knowledge into clinical practice, so more detailed studies are needed. In our review, we showed that research on LR related to GM disorders is mostly based on animal models and PHx, which are performed on healthy livers, while diseases such as ALD, NAFLD, viral hepatitis, etc. remain the main problem of insufficient LR in our patients. Therefore, future studies should clarify how GM dysfunction affects LR with underlying liver diseases and whether GM modulation leads to better outcomes.

It is known that liver diseases lead to LR disturbances, especially in cirrhotic patients, and that adding probiotics to the therapy may be helpful, but the issue is still questionable. The problem is that we do not fully understand how GM affects LR in a healthy state. Another obstacle is the variety of probiotics and prebiotics and their combination. Different strains, microbial content, additives can seriously alter the therapeutic effects, so the results of studies can vary drastically. The selection of preferred strains could help researchers from different countries to conduct comparable studies and integrate them into systemic reviews.

The gut microbiota influences hepatic homeostasis and supports liver regeneration in the “healthy” state, whereas dysbiosis impairs this process. Several liver diseases are associated with dysbiosis, and preoperative antibiotic treatment may also induce GM dysfunction, so perioperative stool examination and eubiosis induction with pre/probiotics or fecal microbiota transplantation may be a possible strategy to improve LR and decrease mortality. Further multicenter clinical trials and systemic reviews are needed to integrate GM testing and interventions into clinical practice.

Credit authorship contribution statement

All the authors – Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Validation, Visualization, Writing (original draft, review, and editing).

Conflicts of interest

There are no conflicts of interest.

Funding

N/A.