Impact of Direct-acting Antiviral Drugs on The Patterns of Gut Microbiota in Patients With HCV Related Chronic Liver Diseases

Overview

The recent introduction of direct-acting antiviral (DAA) therapy has dramatically changed the rates of cure of HCV infection even in difficult to treat subjects and in those with severe comorbidities . A growing number of studies have also demonstrated an improvement in liver stiffness, which is a surrogate marker of liver fibrosis, after HCV eradication. This may have a beneficial impact on the gut liver axis in these patients and may further contribute to slowing down the progression of the disease, potentially influencing the future occurrence of complications. The gut microbiota alpha diversity in cirrhotic patients, which was lower than that in healthy subjects, was significantly improved by the cure of HCV infection and a shift in the overall gut microbiota composition was observed compared to baseline. The abundance of potentially pathogenic bacteria (Enterobacteriaceae, Enterococcus, and Staphylococcus) was decreased after treatment. Aim of study: 1-To determine the changes in gut microbiota in patients with HCV infections with / without liver cirrhosis.

It is a case-control study that was conducted on 8 non-cirrhotic chronically infected patients with hepatitis C virus (group I), 4 patients successfully treated from HCV after receiving 12 weeks of DAAs (Sofosbuvir, Daclatasvir) who achieved sustained viral response (group II), and 4 patients who relapsed after 12 weeks of DAAs (Sofosbuvir/Velpatasvir/Voxilaprevir) (group III). Fourteen healthy control subjects who were matched for age and sex (group VI) were enrolled after obtaining their informed consent. Detailed history taking, clinical examination, abdominal ultrasound (U/S), and laboratory investigations including complete blood count (CBC), liver function tests, and serum transaminases were determined for all patients. Inclusion criteria: patients of any age or sex who was proved to be infected with hepatitis C virus without treatment, those who received treatment and sustained viral response and those who received treatment and relapsed and healthy controls were enrolled in the study Exclusion Criteria Patients receiving antibiotic treatment, probiotics, or any other medical treatment influencing intestinal microbiota 1 month before the start of the study as well as patients with any other viral infection as HBV, and HIV were excluded. Specimen collection Fresh stool samples were collected in the morning from all the participants and were processed within 1 h after defecation. An aseptic technique was used for the collection of all specimens, with care taken to avoid contamination. Also, 5ml of venous blood under aseptic conditions was collected from each patient for estimation of the different laboratory parameters. 1.4: Quantitative assessment of the RNA viral load HCV-RNA level was detected using real-time polymerase chain reaction (RT-PCR) (Bioline International, UK) with a lower limit of detection of 15 IU/ml. (Alter.,2007). 1.5: Microbiota Profiling 1.5.1.DNA Extraction Immediately after collection, Genomic DNA was extracted from the stool samples using the Invitrogen PureLink Microbiome DNA Purification Kit (Thermo Fisher Scientific, Cat #A29790). The stool samples were mixed thoroughly to create a homogenous sample before weighing and transferring the measured amount of the feces (0.2± 0.05 g) to the bead tube, then we followed instructions as directed by the manufacturer as follows: We performed the procedure at room temperature (20-25 ºC).1. The lysate was prepared as following: 1. Sample (0.2±0.05 g) and S1-Lysis Buffer (600µL) were added to the bead tube, the total mixture was 800µL, 2. The bead tubes were vortexed to ensure that the samples were thoroughly dispersed in the liquid. 3. One hundred µL of S2 -Lysis Enhancer was added and vortexed briefly. e. Bead tubes were vortexed for 10 minutes at maximum speed on the vortex mixer using hands-free adapter and horizontal agitation. f. The bead tubes were centrifuged at 14,000× g for 5 minutes at room temperature. g. Then, 400µL of the supernatant was transferred to a clean microcentrifuge tube. h. After that, 250µL of S3- cleanup Buffer was added to the tubes and vortexed immediately to ensure even dispersion of S3 and uniform precipitation of inhibitors. i. The bead tubes were centrifuged at 14,000× g for 2 minutes at room temperature. j. Then 500µL of the supernatant was transferred to a clean microcentrifuge tube, with care to avoid the pellet and any debris. 2\. Then, DNA was bound to the column as following : 1. A total of 900µL of S4-Binding Buffer was added to the tubes and vortexed briefly. 2. Then, 700µL of the sample mixture was loaded onto a spin column-tube assembly, and centrifuged at 14,000× g for 1 minute. 3. The flow-through was discarded and the previous step was repeated with the remaining sample mixture. We ensured that the entire sample mixture passed into the collection tube by inspecting the column and centrifuging again at 14,000 × g for 1 minute. 3\. The DNA was washed and eluted as following: 1. The spin column was placed in a clean collection tube, 500µL of S5- Wash Buffer was added, and then the spin column- tube was centrifuged assembly at 14.000 × g for 1 minute. 2. the flow-through was discarded-, and then the spin column-tube was centrifuged assembly at 14.000 × g for 30 seconds, we did the second centrifugation to optimize the removal of S5- wash Buffer which could interfere with downstream applications. 3. The spin column was placed in a clean tube, 100µL of S6- Elution Buffer was added, and then the tubes were incubated at room temperature for 1 minute. 4. The spin column-tube was centrifuged assembly at 14.000× g for 1 minute then the column was discarded. The purified DNA in the tube was used for the following downstream applications. 1.5.2.PCR amplification of 16S rRNA gene a. PCR optimization Immediately after DNA extraction, PCR was conducted to amplify hyper variable regions V3-V4 of 16S rRNA gene using the following primers with Illumina adapters (underlined): Forward Primer 5'TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG3' Reverse Primer 5'GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATC 3' PCR reactions were performed for preparing 16SrRNA for sequencing in 25μL reactions with 0.8μL for each forward and reverse primers (10μM, Metabion, Germany), 3 μL of template DNA, and 12.5 μL of 1×of Hot Master Mix (Genedirex PCR supermix). 1.5.3. Agarose gel electrophoresis All PCR amplicons were assessed by loading 5 μL of the PCR product into wells of 1% (w/v) Agarose. Five μL of Hyper Ladder l marker was loaded into the first well. Five μL of the PCR reaction mixture from each sample was inserted into the appropriate wells of the gel. The gel was then immersed in Tris-Borate-EDTA (TBE) buffer and electrophoresed at 90 V for separation of the fragments. The gel was visualized after excitation under UV transillumination. The amplified products were sent to IGA Technology Services (Udine, Italy) for further processing. 1.5.4. PCR product purification (PCR Clean-Up) Five μL of the PCR reaction mixture from each sample was mixed with 2 μL of loading buffer and after that PCR products were purified using the Agincourt XP Ampure Beads (Beckam Coulter). 1.5.5. PCR Product sequencing After amplicon purification, dual indices and Illumina sequencing adapters were attached using the Nextera XT Index Kit (Illumina). The quality of the final products was assessed using a Bioanalyzer 2100 (Agilent Technologies, USA) and, after their quantification with a Qubit (Invitrogen), the samples were pooled in equal proportions and sequenced paired-end in an Illumina MiSeq with 600 cycles (300 cycles for each paired read and 12 cycles for the barcode sequence) at IGA Technology Services (Udine, Italy). 1.6. Sequence analysis 1.6.1. Sequence analysis with Illumina MiSeq metagenomics workflow IGA Technology Services performed sequences analysis using the metagenomics workflow of MiSeq Reporter v2.3 (Illumina). Briefly, sequences were demultiplexed based on index sequences. FASTQ files were generated with a Quality Score Trim sample-sheet set, to make trimming. The classification step was performed using Classify Reads, a proprietary Illumina algorithm that provides a species-level classification for paired-end reads. The process involves matching short subsequences of the reads (words) to a set of 16S reference sequences. The accumulated word matches for each read were used to assign reads to a particular taxonomic classification. The taxonomy database for the metagenomics workflow was an Illumina-curated version of the Greengenes database (greengenes. secondgenome. com/downloads/ database/13\_5). 1.6.2. Sequence preprocessing and analysis Illumina platform generated two reads for the forward and reverse direction of the V3-V4 hyper-variable region of the 16S rRNA gene. Sequence filtering and analysis of sequence data were essentially performed with the MOTHUR (v. 1.35.1) software package (Schloss et al.,2009). 1.6.3. Data combination and preprocessing Firstly, Forward and reverse reads for each sample were combined in a single contig separately and samples from each site were combined in a single dataset for each site. This was performed using the (commands make. Contigs, make. file, and make. group) respectively. 1.6.4. Quality value checking and filtering Quality filtering included rejecting reads \<420nt and \>470 nt, excluding homopolymer runs \>8nt, and 0 ambiguous bases, and requiring minimum average Phred quality ≥25 using (command trim. seqs). Illumina may generate many copies of the same sequence which were considered as duplicated sequences. Because it's computationally wasteful to align the same thing a bazillion times, we unique our sequences using the (command unique.seqs). 1.6.5. Sequence aligning to a reference database SILVA database was customized to the V3-V4 region of 16s rRNA gene using the (command pcr.seqs), corresponding to the same regions in Escherichia coli starting from 6426 and ending at 27645. After that, our sequences were aligned to reference database reference sequences (http://www.arb-silva.de/) using (command align.seqs). (Quast et al.,2012) 1.6.6. Sequence deduplication, denoising, and chimera removal Duplicate sequences were merged with previous unique sequences by running (command unique.seqs). To denoise our sequences, precluster the sequences was done using the (command pre. cluster) allowing for up to 2 differences between sequences. This command split the sequences by group and then sorted them by abundance and went from most abundant to least and identified sequences that were within 2 nt of each other. Chimeras sequences were removed using the MOTHUR implementation of the UCHIME algorithm (Edgar et al.,2011). The (command chimera. chime) detected chimeric sequence which was removed by (command get.seqs). 1.6.7. Sequence classification Sequences were aligned to the SILVA taxonomy using a Bayesian classifier with the (command classify.seqs). (Wang et al.,2007). 1.7 Bacterial diversity analysis using QIIME2 1.7.1 Alpha diversity analysis The diversity analytic plugin within QIIME2 uses a random resampling of 32,084 sequences per sample for even sampling depth. Estimates of ecological indices included: rarefaction curves, the number of observed OTUs (species), and the Shannon index, representing common measures of alpha diversity. Rarefaction of the alpha diversity, defining the number of OTUs detected as a function of sampling effort, was performed to confirm that sampling depth gave sufficient coverage to accurately analyze the dominant members of bacterial communities. Alpha diversity was analyzed using two approaches; the number of observed species representing richness and the Shannon diversity index representing richness and evenness of communities. 1.7.2 Beta diversity analysis Beta diversity analyses were estimated to determine phylogenetic distances between communities including weighted UniFrac distance matrices (Lozupone et al.,2011). In the weighted UniFrac analysis, the relative abundance of taxa found in the sample is counted, so less abundant OTU has less weight in the analysis. UniFrac principal coordinates analysis (PCoA) was performed for the interpretation of the UniFrac distance matrix. Phyloseq and ggpubr, R packages were used for the calculation and plotting of α and β diversity (McMurdie and Holmes.,2013). Finally, significance testing of sample clustering was assessed by Permutational Multivariate Analysis of Variance (Adonis R,package Vegan) (Anderson.,2001) using the weighted UniFrac distance matrix in QIIME by compare\_categories.py script. 1.8. Core microbiome analysis: Discovering a core microbiome is very important for understanding the stable, consistent elements across complex microbic assemblages. A core is often defined as the suite of members shared among microbial consortia from similar habitats and is represented by the overlapping areas of circles in Venn diagrams, in which each circle contains the membership of the sample or habitats being compared (Shade and Handelsman.,2012). 1.9. Identification of Biomarkers and Discriminative Taxa LefSe exemplifies the Linear discriminant analysis (LDA) Effect Size method. LEfSe is an algorithm for high-dimensional biomarker identification and characterization that identifies features such as genes, pathways, or taxa that explain the differences between two or more biological states. The idea of LefSe is to start to define significantly differentially abundant OTUs, then use the Wilcoxon rank sum test to check for consistency in the differential abundance, and finally estimate an effect size per feature using linear discriminant analysis. This method assumes that microbial communities can be separated by a linear combination of OTUs (Segata et al.,2011). 1.10. Correlation between community member The Spearman's Rank Correlation Coefficient is used to discover the strength of a link between two sets of data. The correlation coefficient has a maximum value of 1, indicating a perfect positive association between the data sets, and a minimum value of -1 indicating a perfect negative association between data sets. A value of 0 indicates no association between the ranks for the observed values (Farlie., 1960).