Human microbiome is a diagnostic biomarker in hepatocellular carcinoma

Ben-Chen Rao , Jia-Min Lou , Wei-Jie Wang , Ang Li , Guang-Ying Cui ,Zu-Jiang Yu , Zhi-Gang Ren , *

a Department of Infectious Diseases, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China

b Gene Hospital of Henan Province; Precision Medicine Center, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China

c Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China

ABSTRACT

Keywords:

Human microbiome

Gut microbiome

Oral microbiome

Hepatocellular carcinoma

Diagnostic biomarker

Introduction

Liver cancer is one of the most common malignancies worldwide with 841 080 new cases and 781 631 deaths in 2018 [1] . The incidence is listed as 7th among all tumors, and its mortality ranks as 3rd of all cancers, leading to a large global cancer burden [1-3] .Primary liver cancer includes hepatocellular carcinoma (HCC)(comprising 75%-85% of cases) and intrahepatic cholangiocarcinoma (comprising 10%-15% of cases) as well as other infrequent types. HCC is primarily attributed to the progression of chronic liver disease, but the causes and risk factors vary by region. In China, HCC is mainly caused by HBV infection [ 4 , 5 ]. As we all know that there are multiple diagnostic methods for liver cancer, imaging methods including ultrasound, computed tomography, magnetic resonance imaging and selective hepatic angiography [ 6 , 7 ],and tumor biomarkers such as alpha-fetoprotein and its heterogeneity are widely used in patients. Liver biopsy is the gold standard for the diagnosis of liver cancer, however, it’s an invasive procedure. Notably, novel biomarkers, including des-gammacarboxyprothrombin, glutamine synthetase and serum fucosidase in proteomics [ 8 , 9 ], bile acid in metabolomics, microRNA in transcriptomics [10] , BCORL1-ELF4 and CTNND1-STX5 fusion genes in genomics [11] , and the gut microbiome [3] and oral microbiome [12] in human microbiology, are reported. All these show a promising diagnostic potential for liver cancer.

The human microbiome is a leading and influential research topic because it is the sum of genetic information of microorganisms, which are located in and on the surface of the human body, a large number of which are commensal microorganisms located in the intestine. The human genome and microbiome are coordinated and harmonious to ensure the health of the human body [13-16] .Studies about the human microbiota and metagenomics have extremely expanded our knowledge of the component of microbiome communities within the human body [17-20] . Thus, the human microbiome has certain potential as a new diagnostic tool. Many compelling studies about the gut microbiota and the oral microbiota have been published. Studies have shown that the intestinal microbiota has been extensively researched and has proven to be a diagnostic biomarker for many diseases, such as colorectal cancer [21] , type 2 diabetes [22] , liver cirrhosis [23] , pancreatic carcinoma [24] , heart failure [25] and central nervous system diseases [26] . As for the oral microbiome, the Human Microbiome Project was the first to study it in healthy people [19] , and succedent studies have proposed that the oral microbiome also has diagnostic potential [27] , such as in inflammatory bowel diseases [28] ,oropharyngeal squamous cell carcinoma [29] , gastritis [30] and pancreatic cancer [31] . These microbiomes have many advantages when applied to disease diagnosis, such as high accuracy and effi ciency, among which the most noteworthy is their non-invasive nature. We will focus on the application of the human microbiota in the early diagnosis of HCC in this review.

Gut microbiome for precancerous diseases of HCC

Hepatocarcinogenesis is a multistep process, and almost all of the HCC progresses from chronic liver diseases. Hepatitis - cirrhosis - HCC are trilogy of progression of chronic liver diseases, and 80%-90% of HCC patients experience advanced liver fibrosis and cirrhosis. Importantly, about 30% of patients with compensated cirrhosis will develop HCC [32-35] . Moreover, chronic viral hepatitis,alcoholic liver disease (ALD), non-alcoholic steatohepatitis (NASH)and non-alcoholic fatty liver disease (NAFLD) have the potential to develop into cirrhosis [34] . All these diseases are considered precancerous diseases of HCC. Thus, identifying microbiome biomarkers through changes in the gut microbiome in cases of chronic liver disease could be beneficial for the early diagnosis of HCC.

Chronic viral hepatitis

Chronic viral hepatitis are global public health problems that cause progressive liver damage, which might result in liver cirrhosis and HCC [ 36 , 37 ]. Wangetal.reported that the gut microbiota of patients with chronic hepatitis B (CHB) was significantly different from the healthy control group.Alistipes,Bacteroides,Asaccharobacter,Butyricimonas,Ruminococcus,Clostridiumcluster IV,ParabacteroidesandEscherichia/Shigellawere decreased significantly. In contrast,Megamonas, unclassifiedLachnospiraceae,ClostridiumsensustrictoandActinomyceswere increased significantly [38] . Another study [39] has also shown a difference in intestinal microbiota between patients with decompensated hepatitis B cirrhosis and HBV infection. An obvious increase inEnterobacteriaceae,EnterococcusfaecalisandFaecalibacteriumprausnitziiwere also observed in patients with decompensated hepatitis B cirrhosis. Conversely, they found a significant decrease inBifi dobacteriaandLacticacidbacteria. As for patients with chronic hepatitis C, after a comparison of stool samples from healthy individuals and patients with chronic hepatitis C, the results indicated that patients with HCV infection have lower bacterial diversity with an increase ofStreptococcus,LactobacillusandBacteroidetesand a decrease ofClostridialesandBifidobacterium[ 40 , 41 ]. Interestingly, the gut microbiota of HCV-infected patients was significantly improved after cure by direct-acting antivirals, and the abundance of potentially pathogenic bacteria (Enterobacteriaceae,EnterococcusandStaphylococcus) was reduced [42] . If we continue to study the changes in the abovementioned gut microbiota, we can then determine the effects of microecological changes in patients with HBV and HCV on the outcomes of patients with liver cancer.

Alcoholic liver disease

Alcoholic liver disease (ALD) has been one of the leading causes of cirrhosis and liver-related death worldwide. Shalikianiet al.[43] proposed that a shift occurred in the gut microbiota of patients with ALD. Subsequently, Dubinkinaetal.[44] found that the gut microbiome of patients with ALD was rich inBifidobacteriaandLactobacilli. Relevant studies [44-46] suggested that patients with ALD developed intestinal flora dysbiosis including higher levels ofProteobacteriaand lower levels ofBacteroides. The discrimination in the gut microbiota between ALD patients and healthy individuals were explained by Purietal.[46] , and they emphasized that, compared with healthy individuals, the relative amount of the phylumBacteroideteswas obviously decreased in patients with ALD. On the contrary, all alcohol-consuming patient groups exhibited enrichment ofFusobacteria. In addition, clinical trials have indicated that the use of antibiotics and probiotics is beneficial for patients with ALD because of the microbiome dysbiosis that occurs [47] . Experimental studies in mice models have demonstrated that ALD causes bacterial overgrowth and intestinal flora dysbiosis that primarily manifest as a significant decrease in the ratio of probiotics such asLactococcus,Leuconostoc,LactobacillusandPediococcus[ 48 , 49 ]. Alteration in the gut microbiome is closely related to the progression of ALD [50] . Alcohol consumption destroys the intestinal epithelial barrier and increases intestinal permeability, which increases bacterial endotoxin and lipopolysaccharide in the portal vein, and promotes ALD inflammation by activating toll-like receptor 4 [50] .Based on this mechanism, Wangetal.over-regulated regenerating islet-derived protein 3 gamma (Reg3g) in intestinal epithelial cells to limit bacterial colonization on the mucosal surface and reduce bacterial translocation. The result was that the balanced gut microbiome protected mice from alcohol-induced steatohepatitis [51] .

Non-alcoholic steatohepatitis/non-alcoholic fatty liver disease

Non-alcoholic fatty liver disease (NAFLD) is an important global health burden not only for adults, but also for adolescents and even children. A significant increase in prevalence over the past few decades has been observed [52] . Shenetal.[53] identified the relationship between gut microbiota and NAFLD. NAFLD patients had lower gut microbiota diversity than healthy subjects.They had moreProteobacteriaandFusobacteriaphyla. In addition,theErysipelotrichaceae,Enterobacteriaceae,Lachnospiraceaeand theEscherichiaShigellaas well asStreptococcaceaefamilies were enriched in the patients with NAFLD. But a lower amount ofPrevotellawas observed in the NAFLD patients compared with healthy controls. In daily life, the BMI values of many NAFLD patients do not meet the criteria for obesity. Wangetal.[54] studied the patients with nonobese fatty liver disease and found that NAFLD patients had lower diversity and a phylum-level change in the gut microbiota. Additionally, an increase inBacteroidetesand a decrease inFirmicuteswere found in NAFLD patients. Later, Duarteetal.[55] conducted a similar study. Compared with the healthy control, lean NASH patients had one third abundance ofFaecalibacteriumandRuminococcus, obese NASH patients had more abundance inLactobacilli, and overweight NASH patients had less abundance inBifidobacterium. Moreover, lean NASH patients showed a deficiency inLactobacilluscompared with overweight and obese NASH patients. For children with NAFLD, clinical trials have indicated that they have more abundantGammaproteobacteria,PrevotellaandLactobacillus[ 56 , 57 ]. Later research demonstrated that NASH is a precursor for HCC development [58-61] . Gut microbiota dysbiosis and increased intestinal mucosa permeability promote the passage of bacterial derivatives to the portal vein. The most important bacterial derivative is lipopolysaccharide, which activates toll-like receptors (TLRs) and produce pro-inflammatory cytokines and chemokines that promote NASH development [62-64] .

Liver fibrosis/cirrhosis

Cirrhosis plays a critical role in hepatocarcinogenesis. In 2014,Bajajetal.[65] revealed the gut microbiome composition through 219 patients with cirrhosis and 25 matched healthy controls. They suggested the occurrence of gut microbiota dysbiosis in patients with cirrhosis. Progressive changes in the intestinal microbiota were associated with cirrhosis, and in particular, these changes become more serious in cirrhotic patients with decompensation. In addition, the intestinal flora of cirrhosis patients before and after infection were significantly different. Later, a strong relationship was reported between alterations in the gut microbiota and liver cirrhosis. Compared with the healthy group, the cirrhosis group displayed a significant increase inEnterobacteriaceaeandEnterococcus[66] . Chenetal.[67] found that there were a significant decrease inBacteroidetesand a significant increase inProteobacteriaas well asFusobacteriain patients with cirrhosis.Streptococcaceae,VeillonellaceaeandEnterobacteriaceaesignificantly abounded at the family level. Furthermore, the amount ofStreptococcusandLachnospiraceaefamilies were related to the Child-Pugh score.Qin and colleagues recently reported a Chinese cirrhosis cohort and observed specific gut microbiome signs in patients with cirrhosis [23] . They constructed a gut microbiota gene set of 98 cirrhotic patients and a total of 2.69 million genes were obtained.All 75 245 genes can be divided into 66 gene clusters which were significantly different between cirrhotic patients and healthy individuals. Among these gene clusters, 28 significantly increased in patients with cirrhosis, and the remaining 38 were significantly reduced. They created an accurate patient discrimination index on the basis of 15 microbial biomarkers, which indicates that microbial biomarkers are a powerful tool for the diagnosis of cirrhosis. Loombaetal.[68] recently figured out a set of 40 characteristics including 37 bacterial species. A random forest classifier model was constructed to distinguish mild/moderate NAFLD from advanced fibrosis by the identified microbiota. These findings represent a major advancement in the field of gut microbiome research, and specifically, its use as a diagnostic marker for disease.

Gut microbiome and HCC

HCC is the end-stage of various chronic liver diseases. In 2017,we first proposed the theory that the gut microbiome serves as non-invasive biomarker of early HCC [3] . We constructed an HCC classifier model based on a random forest model and fivefold cross-validation. We identified 30 optimal operational taxonomic units for diagnosis. Notably, we have validated our diagnostic model in another two centers, which also have higher the area under the ROC curve values (the AUC values are 0.792 and 0.817, respectively). This study made a significant contribution to the early diagnosis as well as therapy for HCC. In a recent study,Ponzianietal.[69] compared the features of the intestinal microbe of HCC and NAFLD-related cirrhosis. They studied patients with consecutive series of NAFLD-associated cirrhosis, NAFLD-associated cirrhosis with HCC and healthy individuals. They found thatEnterobacteriaceaeandStreptococcuswere enriched, whileAkkermansiawas reduced in patients with cirrhosis. As for HCC group, there was an increase inBacteroidesandRuminococcaceaeas well as a decrease inBifidobacterium.

Therefore, early identification of changes in the gut microbiome and early diagnosis are of great significance to prevent the progression of cirrhosis and HCC.

Oral microbiome for precancerous diseases of HCC

Several studies [ 70 , 71 ] have shown that patients with chronic liver disease have oral microbial dysbiosis. Moreover, some important bacteria of the gut microbe in patients with chronic liver disease are from the oral cavity [ 23 , 72 , 73 ]. Thus, it would be worthwhile to explore the diagnostic function of the oral microbiome in liver diseases.

Chronic viral hepatitis

Lingetal.[70] characterized the oral microbiome of CHB and revealed a decrease in oral bacterial diversity and a significant increase in the ratio of theFirmicutes/Bacteroidetes. They also found that the changing patterns of the oral microbe in patients with HBV-induced cirrhosis were similar to those of CHB patients. Based on previous studies of changes in the gut microbe of CHB, we speculated that the dysbiosis of oral microbe may promote the progression of CHB or the progression of CHB lead to the dysbiosis of oral microbe. But the specific mechanism needs further research.

Autoimmune liver disease

In 2018, Japanese scholars described the characteristics of oral microbiota in patients with autoimmune liver disease [74] . They proposed that compared with healthy controls, patients with autoimmune hepatitis showed an obvious decrease inStreptococcusand a significant increase inVeillonellain the oral microbiome. In addition, patients with primary biliary cholangitis showed a significant decrease inFusobacteriumwith a concurrent increase inVeillonellaandEubacteriumin their oral microbiota. Besides, the relative amount ofVeillonellain the feces was positively correlated withLactobacillus.

Liver fibrosis/cirrhosis

Bajajetal.[71] conducted a clinical study to evaluate the oral microbiome in cases of cirrhosis with hepatic encephalopathy and to compare the gut microbiome. In the salivary microbiome, microbiota dysbiosis was obvious. Compared with healthy controls, the relative amount of potentially pathogenic ones (EnterococcaceaeandEnterobacteriaceae) were increased whereas autochthonous families were decreased in saliva. A study [75] of oral microbial characteristics in cirrhotic patients with pneumonia indicated that there were significant increase in evenness index, the Shannon’s diversity and species richness in cirrhotic patients with pneumonia versus healthy controls. Further, cirrhotic patients with pneumonia had moreActinomycetes,NeisseriaandBacteroidesand lessStreptococcusversus healthy controls. These studies on the changes in the oral microbiome in cirrhosis patients will facilitate the development of the oral microbiome as a diagnostic biomarker for cirrhosis.

Oral microbiome for HCC

Luetal.[12] were the first to analyze the 16S ribosomal RNA gene sequence of tongue coating microorganisms of cirrhotic patients with HCC. They found that there was a significant increase in the microbe diversity. Furthermore, the linear discriminant analysis revealed significant microbial dysbiosis of tongue coat in HCC patients. The oral microbiome of HCC patients was described by a preponderance ofEpsilonproteobacteria,Actinobacteria,ClostridiaandFusobacteria, whereasGammaproteobacteriaandBacteroideteswere enriched in healthy controls. Strikingly, they suggested that differences of the amount inFusobacteriumandOribacteriumcould differentiate HCC patients from healthy controls. These findings identified oral microbiome dysbiosis in HCC patients and provided a non-invasive potential diagnostic biomarker of HCC ( Fig. 1 ).

Fig. 1. Human microbiome includes gut microbiome, oral microbiome and other human microbiome. Gut microbiome, oral microbiome and other human microbiome have certain potential as diagnostic biomarkers for the precancerous diseases of HCC. At the same time, human microbiome can serve as diagnostic biomarkers for HCC.

Diagnostic value of other human microbiomes in liver diseases

In addition to the gut and oral microbiome ( Table 1 ), there are other human microbiomes related to liver diseases. Lelouvieretal.[76] evaluated the difference in blood microbiota between NAFLD-related cirrhosis and healthy individuals. An analysis of blood microbiome provided potential biomarkers for the distinguishing the cirrhosis from a NAFLD group.

As for the ascites microbiota, currently, some scholars believe that the ascites microbiota could aid in the evaluation of the prognosis of decompensated cirrhosis [ 77 , 78 ].

However, it is unfortunate that the examination of the blood and ascites microbiomes for diagnosis requires an invasive procedure. At present, although data on the relationship between the human microbiome (except for intestinal flora and oral flora) and liver diseases are very limited, it is a research area that should be further investigated.

Prospect of clinical application of the human microbiome for the diagnosis of HCC

In recent years, the human microbiome and metagenomics sequencing have made an indelible contribution to the early diagnosis of HCC and for therapy of HCC. However, the human microbiome has certain limitations. For example, regional variation limits the wide application of human microbiome models [79] .

Moreover, heterogeneity in different populations, alteration caused by therapies, technical differences and specimen quality are also its limitations. In later studies, expanding clinical sample validation, cross-regional validation, and even cross-time validation can reduce the limitations. Presently, the combination of microbe biomarkers and currently used diagnostic methods may lead to more benefits in HCC populations. The human microbiome has great clinical application potential in the diagnosis of liver cancer,and based on the subsequent well-designed clinical studies of multiple large cohorts, the human microbiome may become an effective independent diagnostic tool for HCC.

Table 1Microbial genus of diagnostic biomarkers for liver diseases.

CRediT authorship contribution statement

Ben-Chen Rao:Investigation, Writing - original draft.Jia-Min Lou:Data curation, Methodology.Wei-Jie Wang:Formal analysis,Visualization.Ang Li:Resources, Software.Guang-Ying Cui:Conceptualization, Validation.Zu-Jiang Yu:Project administration, Supervision.Zhi-Gang Ren:Conceptualization, Funding acquisition,Writing - review & editing.

Funding

The study was supported by grants from the National Key Research and Development Program of China (2018YFC20 0 0501),National Natural Science Foundation of China ( 81600506 ), National S&T Major Project of China (2018ZX10301201-008), China Postdoctoral Science Foundation ( 2017M610463 and 2018M632814 ).

Ethical approval

Not needed.

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.