Study On Microbiota In Patients With Henoch Schonlein Purpura Based On High Throughput Sequencing

Background:Henoch-Schnlein purpura?HSP?,also known as IgA vasculitis?IgAV?,is a vascular allergic hemorrhagic disease mediated by immunoglobulin A1?IgAl?,often involving the skin,gastrointestinal tract,joints and kidneys.HSP is the most common autoimmune vasculitis in childhood and one of the most common forms of vascular disease.The annual incidence of HSP is reported tobe underestimated due to underreporting.HSP is prevalent in children and has be 10-20/100,000.Because of the complexity of diagnosis,the incidence of HSP may a peak incidence from 4 to 6 years old.Although HSP has been known for more than 200 years and is one of the most common vasculitis,its pathogenesis remains unsolved.Studies have shown that HSP may be associated with allergies caused by infection,vaccination,pollen,Mosquito bites,food,drugs and/or other factors.Many reports suggest that infection,especially beta-hemolytic streptococcal infection,may be a potential cause of the disease.It may be a potential cause of this disease,leading to the production of large amounts of IgA,IgE and so on in the human body,thus causing a series of immune responses in the body.Human health is determined by both internal and external factors.Internal factors are composed of the genome and the second genome.The "second" genome is also known as the microbiome.The unity of interaction between intestinal flora and its host is called intestinal microecology.They have an inseparable symbiotic relationship with the human body that directly affects everyone's health.Intestinal flora plays an important role in promoting digestion and absorption of nutritious food,producing beneficial nutrients,resisting invasion of foreign pathogenic bacteria and regulating immune mechanism.Gut microbial dysbiosis are associated with many immune diseases,including eczema,food allergy,allergic asthma and so on.In recent years,gut microbial dysbiosis in HSP patients have also been reported.However,the number of people involved in the study is small,the classification of various diseases is not in-depth study,the use of backward technology and other issues.Aim:The aim of this study was to investigate the gut microbial dysbiosis in children with HSP.The diversity,composition and characteristics of intestinal microflora were analyzed,and the correlation between intestinal microflora and clinical indexes such as hospitalization days,IgA,IgE and immune indexes was analyzed.Objective to explore the role of intestinal biomarkers in the diagnosis and guidance of children with HSP.Methods:290 hospitalized children with HSP in Qilu Children's Hospital of Shandong University in 2017 were collected and 225 children who met the inclusion criteria were included in the study.Clinical information was collected and 16S rRNA sequencing of stool flora was performed using high-throughput sequencing technique.The results were compared with those of 70 healthy controls.LEfSe,SPSS and other software were used to analyze and compare the difference of intestinal flora.Random forest machine learning model was used to screen intestinal microbial biomarkers,and receiver operating characteristics?ROC?curve was used to evaluate the diagnostic ability of HSP.Results:This study included 225 children with Henoch-Schonlein Purpura,ofthem,29 with simple type of HSP?S?,49 with GI type of HSP?GI?,57 with arthrosis type HSP?A?,5 HSP with glomerulonephritis?GL?and 85 with mixed type of HSP?Mix?.Clinical information was collected and 16S rRNA sequencing of stool flora was performed using high-throughput sequencing technique.The results were compared with those of 70 healthy controls.Chi square test showed no significant difference in sex ratio between the two groups?P>0.05?.The mean age of HSP group?7.1 � 2.5y?was similarwith that of healthy control group?7.0 � 2.7y??P=0.58?.ANOVA analysis showed that there was no significant difference between different age subtypes and healthy control group?P = 0.07?.However,the age of different subtypes was age-related?P=0.04?.Based on the high-throughput sequencing,we analyzed the alpha diversity of intestinal flora in HSP patients and healthy children.The results showed that intestinal flora in HSP patients was lower than that in healthy controls in Shannon index?2.98d�0.67 vs 3.35�0.60,P<0.0001?,Ace index?481.03�92.41 vs 540.12�90.52,P<0.0001?.Differences.We further analyzed the alpha diversity of intestinal flora in patients with different HSP analysis.The Sobs,Shannon,Chao and Ace index of GL group were not significantly different from the healthy control group?P>0.05?.In terms of sobs index,the healthy control group was significantly higher than S group?375.66 �68.18 vs 299.86 � 65.20,P<0.0001?,GI group?375.66 � 68.18 vs 293.80 �72.53,P<0.0001?,A group(375.66 �68.18 vs 300.72 � 58.06,P<0.0001],mixed HSP?375.66 �68.18 vs 297.28 � 82.77,P<0.0001?.In the aspect of Shannon index,the healthy control group was higher than the S group?3.35�0.60 vs 3.04�0.53,P<0.05?,GI group?3.35�0.60 vs 2.90�0.70,P<0.001?,A group?3.35�0.60 vs 3.02�0.62,P<0.01?,Mix group?3.35�0.60 vs 2.95�0.74,P<0.0001?.In the aspect of Ace index and Chao index,the healthy control group was higher than the S group,GI group,A group and Mix group?P<0.05?,respectively.From the above results,except for GL group,the diversity and abundance of intestinal flora were significantly decreased in different types of HSP patients.After analyzing the diversity of bacteria,we used PCoA to analyze the structure of the microbiota.The results showed that the intestinal microbial structure of HSP patients was significantly different from that of healthy controls.Anosim analysis showed that the difference of intestinal microbial structure between the two groups was statistically significant?r = 0.228,P = 0.001?.We further analyzed the differences of the flora structure between the subtype of HSP and the healthy controls,and found that there was no significant difference in the intestinal microbial structure between different subtype groups?Anosim analysis,P>0.05?,but the Anosim analysis showed that the differences between the different subtypes and control group were statistically significant?r = 0.16,P=0.001?.Microbial composition is an important part of intestinal microbiota.Based on the analysis of microbial diversity,we compared the composition of HSP with that of healthy controls.Firstly,we analyzed the number of OTUs in the two groups.It was found that the HSP group contains 1740 OTUs and the control group contains 1675 OTUs.Further analysis revealed that there were 1672 OTUs in the two groups,68 in the HSP group and only 3 in the healthy control group.After clustering analysis of OTU,it was found that there were three main phylums in both groups,mainly Bacteroides,Firmicutes and Proteobacteria.The results of multi-group phylum level also showed that the HSP analysis was mainly composed of Bacteroides,Firmicutes and Proteobacteria.The results of genus-level analysis showed that the main HSP components?with an average content of more than 1%?incl uded Bacteroides,Faecalibacterium,Prevotella 9,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterium,Alistipes,Megamonas,Lachnoclostridium,Veillonella and Enterococcus.The main members of the control group?with an average content of more than 1%?included:Bacteroides,Faecalibacterium,Prevotella₉,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterizum,Christensenellaceae_R-7_group,Alistipes,Roseburia,Dialister,Lachnospiraceae_NK4A136_group and Suterella.We also analyzed the main genus of intestinal mcirobiota of different types of HSP.Among them,the main genus members of HSP patients included 15 genera:Bacteroides,Faecalibacterium,Prevotella 9,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterium,Megamonas,Alistipes,Lachnoclostridium,Christensenellaceae_R-7_group,Veillonella,Anaerostipes,Intest Inibacter and Prevotellaceae NK3B31_group.The main intestinal microflora members?average content more than 1%?of GI group patients included 13 genera:Bacteroids,Faecalibacterium,Prevotella 9,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterium,Megamonas,Alistipes,Lachnoclostridium,Enterococcus,Veillonella,and Streptococcus.The main intestinal microflora?average content of more than 1%?in A group included 12 genera:Bacteroides,Faecalibacterium,Prevotella 9,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterium,Megamonas,Alistipes,Lachnoclostridium,Sutterella,and Anaerostipes.The main intestinal microflora members?average content more than 1%?of Mix group included 11 genera:Bacteroides,Faecalibacterium,Prevotella 9,Parabacteroides,Escherichia-Shigella,Parasutterella,Phascolarctobacterium,Megamonas,Alistipes,Lachnoclostridium,and Enterococcus.The main intestinal microflora members?average content more than 1%?of GL group included 8 genera:Bacteroides,Faecalibacterium,Prevotella 9,Parabacteroides,Parasutterella,Phascolarctobacterium,Sutterella,and Ruminococcaceae_UCG-002.It can be seen that there are obvious differences in the composition of the different types of bacteria.We further constructed the Heatmap map of different HSP microbial communities and found that there were significant differences in intestinal microflora among differentHSPtypes.For example,Roseburia,Dialister,Lachnospiraceae_NK4A136_group,increased significantly in healthy controls,while Escherichia-Shigella and Megamonas decreased significantly in healthy controls.From the above analysis,we can see that the intestinal microbial composition of HSP patients and their different types are different from that of healthy controls.We used LefSe to analyze the difference of intestinal microbial composition.The results showed that the order of Enterobacteriales,Selenomonadales and Fusobacteriales increased in the HSP group.At the generic level,genus of Parabacteroides,Megamonas,Enterococcus,Veillonella,Anaerostipes,Megasphaera,Enterbacter and Streptococcus also increased significantly?LDA>3.0?.In the healthy group,the target microorganisms increased in Clostrideales,Burkholderiales and Pasteurellales,Roseburia,Dialister,Lachnospiraceae NK4A136_group,Parasutterella,Lachnospira,and Haemophilus?LDA>3.0?at the generic level.On this basis,we further analyzed the significantly enriched microorganisms among the subtypes.Megamonas,Prevotellacear_NK3B31_group and Paraprevotella increased significantly?LDA>3.0?in patients with simple HSP.Enterococcus,Veillonella,Streptococcus,unclassified_olactobacillales,Klebsiella,Weissella,Prevotella₇,Abiotrophia,and Barnesiella increased significantly?LDA>3.0?in patients of GI group.In the patients of A group,Bacteroides and Anaerostipes were significantly increased?LDA>3.0?.Ruminococcaceae_UCG₀02,g_unclassified_flachnospiraceae,g_Eubacterium_ruminantium group and g Ruminococcus_torques_group were significantly increased in patients of GL group.No LDA value greater than 3 was found in mixed HSP,but the microorganisms of g norank_fErysipelotrichaceae,Provetella 2,Acidaminococcus and Lachnoanaerobaculum increased significantly?LDA>2.0?.Because of the significant differences in intestinal microbiota between HSP and healthy controls,we want to identify biomarkers that distinguish the two group.Firstly,we use Random Forest in machine learning to construct a decision tree to screen OTU sequences which can be used to distinguish two groups,and 50 OTU which can be used to distinguish them were selected.Then we use these OTU to construct the work curve of ROC.The AUC of ROC curve is 0.95,which shows that the model can be used to distinguish two groups.The sensitivity and specificity of the model are 97.1%and 82.7%respectively.At the same time,we selected the top 20 OTUs with the greatest difference from 50 biomarkers,and constructed the ROC subject working curve with AUC of 0.96,which showed that the model had high accuracy.Meanwhile,the sensitivity and specificity of the model were 89.3%and 90.7%,respectively,with high diagnostic efficiency.Furthermore,OTU of the first 10 abundances was selected from 50 biomarkers and the ROC subject working curve was constructed.AUC was 0.94,indicating that the model still had high accuracy.Meanwhile,the sensitivity and specificity of the model were 90%and 87.6%respectively,and still had high diagnostic efficiency.There are significant immune disorders in patients with HSP,including elevated IgA and IgE levels.We calculated the Spearman's rank correlation coefficient between the gut microbiota of Mix group and the clinical indices including LOS,IgA,IgM,IgE and IgG.Strong correlations?correlation r>0.22 or<-0.22,P<0:05?were detected among 11 genus of Mix group and the 5 clinical indices.A significant negative correlation can be observed between the LOS and the gunes Paraprevotella?r=-0.245,P<0.05?and Roseburia?r=-0.233,P<0.05?.Meanwhile,the IgA exhibited significant negative?r=-0.341,P<0.01?correlation with the genus Bifidobacterium.Additionally,IgM was negatively associated with Genus Escherichia-Shigella?r =-0.234,P<0.05?,Enterococcus?r =-0.28,P<0.05?and norank_fErysipelotrichaceae?r =-0.257,P<0.05?.IgE was negatively associated with Genus Escherichia-Shigella?r =-0.222,P<0.05?,Megamonas?r =-0.222,P<0.05?,and Roseburia?r =-0.256,P<0.05?while positively associated with Genus[Eubacterium]_coprostanoligenes-group?r=0.287,P<0.01?,and Parabacteroides?r=0.287,P<0.05?.The genus Coprobacter was found to be positively associated with IgG?r=0.335,P<0.01?.We also found that the LOS was positively related to IgA?r=0.27,P<0.05?and negatively related to IgM?r=0.269,P<0.05?based on Spearman's rank correlation analysis.Conclusions:1.The results showed that there were significant differences in the composition and structure of intestinal microbiota between children with Henoch-Schonlein purpura and healthy children.2.Except for GL group of HSP,different subtypes of Henoch Schonlein purpura are similar in composition and structure.3.Intestinal microbial biomarkers might be used for the distinguish HSP and healthy children with a high diagnosis efficiency.4.The intestinal microbiota is correlated with the clinical indexes of HSP patients,which may be related to its role in the pathogenesis of HSP.