Abnormal Fecal Microbiota in Patients With Hepatitis B
Abnormal Fecal Microbiota in Patients With Hepatitis B
The Child-Turcotte-Pugh (CTP) scoring system was used to assess the severity of cirrhosis. A total of 240 individuals, including 120 hepatitis B liver cirrhosis (HBLC) patients (40 with CTP score A, 40 with CTP score B, 40 with CTP score C) and 120 healthy individuals, 40–60 years old, with a body mass index (BMI) = 18.5–24.9 kg m and without food preferences, were enrolled in this study. Cirrhosis was diagnosed histologically in all patients. None of the patients had comorbid diseases. The control group consisted of 120 normal individuals who visited the People's Liberation Army 302 Hospital in Beijing for routine health examinations. All healthy individuals had normal liver biochemistry test results with no evidence of hepatic or other diseases. None of the subjects had received antibiotics, probiotics, steroids or other hormones (including oral, intramuscular or intravenous injection) for at least 3 months before sampling. Characteristics of the two groups are given in Table 1 and Additional file 1: Table S1. Patients and normal individuals were each asked to provide a fresh stool sample, which was frozen immediately for DNA extraction. Samples from 20 patients with HBLC (6 with CTP score A, 7 with CTP score B, 7 with CTP score C) and 20 normal individuals were subjected to metagenomic analysis and the other 200 samples were subjected to real-time qPCR analysis. All participants signed an informed consent form prior to entering the study. The study conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
A frozen aliquot (200 mg) of each fecal sample was suspended in 250 μL guanidine thiocyanate, 0.1 M Tris (pH 7.5) and 40 μL of 10% thiocyanate. Metagenome extraction was conducted as described previously. DNA library preparation was performed according to the manufacturer's instruction (Illumina). Workflows were designed to perform cluster generation, template hybridization, isothermal amplification, linearization, blocking and denaturation, and hybridization of the sequencing primers. High-throughput Illumina/Solexa sequencing of a 350-bp library for each sample was conducted using an Illumina Genome Analyzer IIx with read lengths of 90 bp. The base-calling pipeline (version Illumina Pipeline-0.3) was used to process the raw fluorescence images and call sequences. High-quality reads were extracted by filtering out low quality reads with 'N' bases, adapter contamination, or human DNA contamination from the Illumina raw data.
The short read assembler SOAPdenovo 2.20, with the same parameters used to construct the MetaHIT gene catalogue in 2010, was used to assemble the high-quality short reads from the samples. Gene prediction was performed using GeneMark v2.7. All the predicted genes were aligned pairwise using BLASTN. To construct a non-redundant gene catalogue, sequences that aligned over at least 90% of their length with more than 95% identity (no gaps allowed) were removed as redundancies.
Genes were aligned to the NCBI nr database (e-value < 1e-10, identity > 90%) and taxonomic assignment of the predicted genes was performed using an in-house pipeline as described previously. In this study, samples were clustered using Hellinger distance and then principle component analysis (PCA) was performed to identify the significant features of the intestinal microbes that distinguished patients from healthy individuals. The top five principal components (P value in Tracy-Widom test < 0.05 and contribution > 3%) were tested. Using the Wilcoxon rank-sum test method based on the pair-wise comparison matrix, genes with significantly differential abundance between the patients and normal subjects were assessed and subjected to taxonomic assignment at the species level.
Genes were aligned to the eggNOG (v 3.0) and KEGG (release 59.0) databases using BLASTP (e-value ≤1e-5) for functional annotations. Each sequence was assigned to a KEGG orthologous group or to an eggNOG orthologous group based on the highest scoring annotated hit(s) that contained at least one high-scoring segment pair (HSP) scoring over 60 bits. For genes with no hits in the eggNOG database, novel gene families were identified based on all-against-all BLASTP results and clustered using the Markov Cluster (MCL) algorithm with an inflation factor of 1.1 and a bit-score cutoff of 60.
Real-time qPCR was performed using a DNA Engine Opticon 2 apparatus (Bio-Rad, Hercules, CA) with the associated Opticon Monitor software (version 3.0, Bio-Rad). All primer sets used are listed in Table 2. Each primer mixture (25μL) contained 12.5μL of SYBR Premix Ex Taq (Takara, Dalian, China) containing MgCl2, Tris–HCl, KCl, deoxynucleoside triphosphate, SYBR Green I, and Taq DNA polymerase, 0.5μM primer, and 2μL of the template DNA. Amplifications were performed under the following temperature profiles: one cycle at 95°C for 3 min, 30 cycles of denaturation at 95°C for 30 s, annealing for 40 s, and extension for 45 s, followed by a final elongation step at 72°C for 7 min. Fluorescence was measured after the extension phase of each cycle at an appropriate temperature for 10 s to avoid interference of primer-dimers, secondary structure, or spurious priming. The DNA template concentration was determined by comparison with serially diluting standards (10 to 10 copies of plasmid DNA containing the respective amplicon for each set of primers) running on the same plate. Each reaction was repeated in triplicate on separate run plates.
A two-tailed Wilcoxon rank-sum test was used to identify the association between microbiota abundance and cirrhosis. Correlation between variables was computed using Spearman rank correlation. The Wilcoxon rank-sum test, Spearman rank correlation and Student t test were conducted using SPSS version 11.0 for Windows (SPSS Inc., Chicago, IL, USA). We applied the "q-value" method proposed previously to estimate the FDR (false discovery rate) instead of a sequential P-value rejection method. The statistical hypothesis tests were performed on a large number of features of the gene, KEGG orthologue, and eggNOG orthologue profiles.
Methods
Human Fecal Sample Collection
The Child-Turcotte-Pugh (CTP) scoring system was used to assess the severity of cirrhosis. A total of 240 individuals, including 120 hepatitis B liver cirrhosis (HBLC) patients (40 with CTP score A, 40 with CTP score B, 40 with CTP score C) and 120 healthy individuals, 40–60 years old, with a body mass index (BMI) = 18.5–24.9 kg m and without food preferences, were enrolled in this study. Cirrhosis was diagnosed histologically in all patients. None of the patients had comorbid diseases. The control group consisted of 120 normal individuals who visited the People's Liberation Army 302 Hospital in Beijing for routine health examinations. All healthy individuals had normal liver biochemistry test results with no evidence of hepatic or other diseases. None of the subjects had received antibiotics, probiotics, steroids or other hormones (including oral, intramuscular or intravenous injection) for at least 3 months before sampling. Characteristics of the two groups are given in Table 1 and Additional file 1: Table S1. Patients and normal individuals were each asked to provide a fresh stool sample, which was frozen immediately for DNA extraction. Samples from 20 patients with HBLC (6 with CTP score A, 7 with CTP score B, 7 with CTP score C) and 20 normal individuals were subjected to metagenomic analysis and the other 200 samples were subjected to real-time qPCR analysis. All participants signed an informed consent form prior to entering the study. The study conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
Metagenomic DNA Extraction, DNA Library Construction and Sequencing
A frozen aliquot (200 mg) of each fecal sample was suspended in 250 μL guanidine thiocyanate, 0.1 M Tris (pH 7.5) and 40 μL of 10% thiocyanate. Metagenome extraction was conducted as described previously. DNA library preparation was performed according to the manufacturer's instruction (Illumina). Workflows were designed to perform cluster generation, template hybridization, isothermal amplification, linearization, blocking and denaturation, and hybridization of the sequencing primers. High-throughput Illumina/Solexa sequencing of a 350-bp library for each sample was conducted using an Illumina Genome Analyzer IIx with read lengths of 90 bp. The base-calling pipeline (version Illumina Pipeline-0.3) was used to process the raw fluorescence images and call sequences. High-quality reads were extracted by filtering out low quality reads with 'N' bases, adapter contamination, or human DNA contamination from the Illumina raw data.
Gene Catalogue Construction
The short read assembler SOAPdenovo 2.20, with the same parameters used to construct the MetaHIT gene catalogue in 2010, was used to assemble the high-quality short reads from the samples. Gene prediction was performed using GeneMark v2.7. All the predicted genes were aligned pairwise using BLASTN. To construct a non-redundant gene catalogue, sequences that aligned over at least 90% of their length with more than 95% identity (no gaps allowed) were removed as redundancies.
Bioinformatic Analysis
Genes were aligned to the NCBI nr database (e-value < 1e-10, identity > 90%) and taxonomic assignment of the predicted genes was performed using an in-house pipeline as described previously. In this study, samples were clustered using Hellinger distance and then principle component analysis (PCA) was performed to identify the significant features of the intestinal microbes that distinguished patients from healthy individuals. The top five principal components (P value in Tracy-Widom test < 0.05 and contribution > 3%) were tested. Using the Wilcoxon rank-sum test method based on the pair-wise comparison matrix, genes with significantly differential abundance between the patients and normal subjects were assessed and subjected to taxonomic assignment at the species level.
Genes were aligned to the eggNOG (v 3.0) and KEGG (release 59.0) databases using BLASTP (e-value ≤1e-5) for functional annotations. Each sequence was assigned to a KEGG orthologous group or to an eggNOG orthologous group based on the highest scoring annotated hit(s) that contained at least one high-scoring segment pair (HSP) scoring over 60 bits. For genes with no hits in the eggNOG database, novel gene families were identified based on all-against-all BLASTP results and clustered using the Markov Cluster (MCL) algorithm with an inflation factor of 1.1 and a bit-score cutoff of 60.
Real-time qPCR
Real-time qPCR was performed using a DNA Engine Opticon 2 apparatus (Bio-Rad, Hercules, CA) with the associated Opticon Monitor software (version 3.0, Bio-Rad). All primer sets used are listed in Table 2. Each primer mixture (25μL) contained 12.5μL of SYBR Premix Ex Taq (Takara, Dalian, China) containing MgCl2, Tris–HCl, KCl, deoxynucleoside triphosphate, SYBR Green I, and Taq DNA polymerase, 0.5μM primer, and 2μL of the template DNA. Amplifications were performed under the following temperature profiles: one cycle at 95°C for 3 min, 30 cycles of denaturation at 95°C for 30 s, annealing for 40 s, and extension for 45 s, followed by a final elongation step at 72°C for 7 min. Fluorescence was measured after the extension phase of each cycle at an appropriate temperature for 10 s to avoid interference of primer-dimers, secondary structure, or spurious priming. The DNA template concentration was determined by comparison with serially diluting standards (10 to 10 copies of plasmid DNA containing the respective amplicon for each set of primers) running on the same plate. Each reaction was repeated in triplicate on separate run plates.
Statistical Analysis
A two-tailed Wilcoxon rank-sum test was used to identify the association between microbiota abundance and cirrhosis. Correlation between variables was computed using Spearman rank correlation. The Wilcoxon rank-sum test, Spearman rank correlation and Student t test were conducted using SPSS version 11.0 for Windows (SPSS Inc., Chicago, IL, USA). We applied the "q-value" method proposed previously to estimate the FDR (false discovery rate) instead of a sequential P-value rejection method. The statistical hypothesis tests were performed on a large number of features of the gene, KEGG orthologue, and eggNOG orthologue profiles.
Source...