- Email: [email protected]
Intestinal microbiota dysbiosis in children with recurrent respiratory tract infections
Microbial Pathogenesis 136 (2019) 103709
Contents lists available at ScienceDirect
Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath
Intestinal microbiota dysbiosis in children with recurrent respiratory tract infections
T
Lei Lia, Fang Wanga, Yanni Liub, Feng Gua,∗ a b
Department of Pediatrics, The Affiliated Hospital of Qingdao University of Shandong Province, Qingdao, China Department of Obstetrics and Gynecology, Binzhou Medical University Hospital of Shandong Province, Binzhou, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Recurrent respiratory tract infection Gut microbiome Bacterial diversity 16S rRNA
Background: The impact of the gut microbiota on recurrent respiratory tract infection (RRTI) remains to be fully elucidated. Methods: To characterize the gut microbiota in patients with RRTI, fecal samples from 26 patients with RRTI and 23 healthy volunteers were profiled using the Illumina MiSeq platform. Beta diversity (Principal Component Analysis (PCA), Principal Co-ordinates Analysis (PCoA), Non-metric multidimensional scaling (NMDS)) analysis showed that the bacterial community structure segregated differently between the RRTI and control groups. Results: Results from alpha diversity analysis revealed lower microbiota diversity in samples from RRTI patients than in normal controls. Taxonomic analysis illustrated that the abundance of six phyla (Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, Verrucomicrobia, Tenericutes) and four genera (Enterococcus, Faecalibacterium, Bifidobacterium, Eubacterium were significantly different between these two groups. In addition, Enterococcus (P < 0.001) was more enriched in the RRTI group, whereas the abundances of Eubacterium (P < 0.001), Faecalibacterium (0.01 < P < 0.05) and Bifidobacterium (0.01 < P < 0.05) were reduced in the RRTI group compared to those in the normal control group. The performance of the model was assessed using ROC analysis, and Enterococcus, Eubacterium and Bifidobacterium achieved AUC values of 0.860, 0.820, and 0.689, respectively. Conclusions: These results provide fundamental evidence in support of intestinal microbiota dysbiosis in children with RRTI.
1. Introduction Recurrent respiratory tract infections (RRTIs) are one of the most common and most frequent diseases in children and adolescents [1]. According to the Italian Pediatric Society, the diagnostic criteria for RRTI in children are as follows: ≥ 6 respiratory tract infections in a year, ≥ 1 upper respiratory tract infection in a month between September and April, or ≥3 lower respiratory tract infections in a year [2]. RRTI is a common disease in pediatrics, occurring in children under 5 years of age, and accounting for 10%–30% of all pediatric respiratory infections, an incidence rate that increases each year [3]. In countries with high mortality rates, such as southeast Asia and Africa, RRTIs have a 23% mortality rate and impact two million people [4]. RRTIs are dangerous diseases in patients with antibody defects, which can be treated with antibody replacement therapy but the cost, estimated to be approximately 20,000–30,000 euros per patient per year, is relatively
high [5]. If not properly treated, RRTIs can lead to asthma, myocarditis, nephritis, and other diseases, which may seriously affect a child's growth and development [6]. Therefore, the pathogenesis of RRTIs has become the focus of clinicians to promote new approaches for the management and prevention of RRTI [7]. The human gut has a huge number of microbes, which are collectively referred to as the ‘‘microbiota’’ [8]. The gut microbiota constitutes approximately 70% of a person's entire microbiota, and the gut microbiota is dominated by the Bacteroidetes and Firmicutes phyla [9]. The dynamic composition of the human gut microbiota is determined by multiple factors, including the mode of delivery, diet, the environment, and antibiotics. A healthy gut microbiota is beneficial for the host in many ways, including by providing nutrients and protection from pathogens, and by inducing the maturation of immune responses [10]. Dysbiosis in the diversity of the microbiota can lead to physiological changes including in the gastrointestinal, respiratory, cardiovascular, or
∗ Corresponding author. Department of pediatrics, the Affiliated Hospital of Qingdao University, No.1677 Wutaishan Road, Huangdao Distict, Qingdao, 266555, China. E-mail address: [email protected] (F. Gu).
https://doi.org/10.1016/j.micpath.2019.103709 Received 5 May 2019; Received in revised form 29 August 2019; Accepted 2 September 2019 Available online 05 September 2019 0882-4010/ © 2019 Elsevier Ltd. All rights reserved.
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
2.3. Illumina MiSeq sequencing
neurological systems, in autoimmunity or hyperimmunity, and in chronic, cancerous, and psychiatric diseases [11]. In recent years, studies have found that respiratory diseases in children are also associated with the composition of the intestinal microbiota. Suarez-arrabal MC et al. [12] found that compared with those of control children, the nasopharyngeal and nasal areas of the children infected with RSV (respiratory syncytial virus bronchiolitis) were more likely to be colonized with Haemophilus influenzae, Pneumococcus sp. and mucositis (81% vs. 65%). It has been reported that the number of Bifidobacteria in the intestinal tract of Spaniards with long-term allergic asthma is significantly lower than in healthy Spaniards [13]. All of these studies have shown that the intestinal microbiota play a crucial role in the development of respiratory diseases, however, the impact of the gut microbiota in RRTI has not been adequately described. In this study, we performed a microbiome analysis of fecal samples from RRTI patients and healthy volunteers to elucidate the bacterial changes that might induce or accompany RRTI using a sequencing platform.
The PCR products were quantified using the QuantiFluor™-ST Blue Fluorescence Quantitative System (Promega) and the initial quantitative results from electrophoresis. The corresponding proportions were then mixed according to the sequencing requirements of each sample. Amplicons were extracted from 2% agarose gels and purified using the AxyPrep DNA Gel Extraction Kit according to the manufacturer's instructions and quantified using QuantiFluor™-ST (Promega, USA). Purified amplicons were pooled in equimolar amounts and sequenced (paired-end; 2 × 250) on the Illumina MiSeq platform according to standard protocols.
2.4. Processing of sequencing data Raw FASTQ files were demultiplexed and quality-filtered using FLASH Trimmomatic with the following criteria: i) Using a 50 bp sliding window, reads with adaptors under 20 bp, reads with an average quality value of less than 20, reads less than 50 bp, and reads containing N bases were filtered out; ii) PE reads with a minimum overlap of 10 bp were merged into a single sequence; iii) Sequences with a mismatch ratio higher than 0.2 in the overlapping area were filtered out; iv) Sequences were identified and oriented according to their barcodes and primers; sequences with barcode mismatches greater than 0 or primer mismatches greater than 2 were filtered out [17].
2. Methods 2.1. Sample collection Fecal samples were collected from 26 children with RRTIs and 23 healthy volunteers of the Affiliated Hospital of the Medical College Qingdao University. The diagnosis of RRTI was performed according to the criteria issued by the Pediatric Society of the Chinese Medical Association in 2007. Additionally, the children with RRTIs were eligible for enrollment only if they had no history of other diseases and no history of antibiotic or probiotic use. Exclusion criteria were as follows: (1) Those with chronic liver disease, kidney disease or basic diseases of the digestive tract were excluded; (2) Patients who had recently received antibiotics were also excluded; (3) Patients who were diagnosed with RRTI caused by organic or congenital lesions, or primary immunodeficiency were excluded [14]. Two weeks before fecal sample collection, patients and healthy volunteers had to have not taken antibiotics, probiotics, Chinese herbal medicines, hormones or other medications that would affect bacterial structure. Furthermore, the participants (> 5 years old) were not permitted to eat food in the 12 h before fecal sample collection, and the sample collection environment was confirmed to be clean. All samples were collected in accordance with the relevant guidelines and regulations, and the research was approved by the Research Ethics Boards of Qingdao General Hospital. Informed consent was provided by the parents of all participants. We collected 3–5 g of fresh fecal samples and immediately placed them into clean, sterilized centrifuge tubes. All samples were frozen immediately after sampling and stored at −80 °C [15].
2.5. Bacterial community characterization The high-throughput sequencing reads from 49 fecal samples were reassigned to their respective samples according to their barcodes. Operation taxonomic units (OTUs) were clustered with a 97% similarity cutoff using Usearch (Version 7.0, http://drive5.com/uparse/), and chimeric sequences were identified and removed using UCHIME. The OTUs with a 97% similarity level were used for alpha diversity analysis, which analyzed the species diversity of individual samples through the evaluation of Chao, abundance-based coverage estimators (ACE), Shannon, and Simpson parameters. The rank-abundance curves were also analyzed using mothur and R software. A Venn diagram was generated to show the shared and unique OTUs among the groups, based on the occurrence of OTUs in a sample group and regardless of their relative abundance; these were also analyzed using R software.The Wilcoxon rank-sum test was used to identify significant differences in the alpha diversity indices between the different groups (P < 0.05). Beta diversity was analyzed to investigate the similarity of the bacterial community structures among the groups using Bray Curtis distances. The beta diversity was visualized via principal component analysis (PCA), principal coordinate analysis (PCoA) and Nonmetric Multidimensional Analysis (NMDS). The Wilcoxon rank-sum test was used to detect significant differences in abundance between the groups. LEfSe (http://huttenhower.sph.harvard.edu/galaxy/root?tool_id= lefse_upload) [18] uses linear discriminant analysis (LDA) to estimate the effect that the abundance of each component has on the differences. The Circos sample and species relationship map is a visual circle diagram that describes the correspondence between sample and species. To evaluate the discriminatory ability of the bacteria, operating characteristic curves (Receiver Operating Characteristic, ROC) were constructed and the area under curve (AUC) was calculated.
2.2. DNA extraction and PCR amplification Microbial DNA was extracted from fecal samples using the OMEGAsoil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to manufacturer's protocols. The V3–V4 region of the bacterial 16S ribosomal RNA gene were amplified by PCR (ABI GeneAmp® 9700; 95 °C for 2 min, followed by 25 cycles at 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s and a final extension at 72 °C for 5 min) using primers 338F and 806R, which included unique eight-base barcode sequences for each sample. PCRs were performed in triplicate in 20 μL mixtures containing 4 μL of 5 × Fast Pfu buffer, 2 μL of 2.5 mM dNTPs, 0.8 μL of each primer (5 μM), 0.4 μL of Fast Pfu polymerase, and 10 ng of template DNA. All samples were subjected to the formal test conditions, and each sample was duplicated. The duplicate PCR products from the same sample were mixed and tested by 2% agarose gel electrophoresis. The PCR products were recovered using 2% agarose gel electrophoresis and the AxyPrepDNA Gel Recovery Kit (Axygen Biosciences, Union City, CA, US A), with elution in Tris-HCl [16].
2.5.1. Statistical analysis All statistical analyses were performed using R packages (V.2.15.3). The Student's t-test was performed using SPSS version 20 for Windows. For correlation analysis, Spearman's rank test was performed. Multiple hypotheses tests were adjusted using the Benjamini and Hochberg false discovery rate (FDR). 2
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Table 1 Alpha diversity between patients had family history of RRTI (RRTI with FH) and patients did not have family history of RRTI (RRTI without FH). Group
samples
RRTI with FH RRTI without FH P-value
Sobs
Feces (N = 7) Feces (N = 19) –
ACE
Chao
Shannon
Simpson
Coverage
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
113 100.89 0.5031
38.962 40.698
137.49 132.05 0.7477
37.766 37.811
137.22 128.22 0.6421
39.151 44.559
2.3787 2.1594 0.5748
0.96333 0.83928
0.24056 0.26035 0.861
0.24552 0.25547
0.99962 0.99962 0.997
0.00013523 0.00015349
Note: 0.01 < p < = 0.05, marked *, 0.001 < p < = 0.01, marked as * *, p < = 0.001, marked as * *.
separation between the healthy and RRTI groups in the first two principal component scores, which accounted for 28.5% and 15.3% of the total variations, respectively. This suggested that disease may be an important factor that accounts for the changes in the structure of the intestinal microbiota (Fig. 1). PCoA also showed a clear separation between the RRTI and healthy groups based on weighted analysis (Fig. S4). The figure shows that the percent of variation explained by PC1 and PC2 was 19.59% and 10.37%, respectively. The NMDS of the bacterial community also showed a separation between the RRTI and healthy groups (Fig. S5).
3. Results 3.1. Characterization of study population A total of 49 subjects were analyzed. Basic participant information was recorded, including age, sex, gestational age, feeding mode, mode of production, family history of RRTI, and secondhand smoke exposure (Supplementary Table S1 and Table S2). Based on the characteristics of family history of RRTI, we analyzed the alpha and beta diversity between patients had family history of RRTI (RRTI with FH) and patients did not have family history of RRTI (RRTI without FH). The Sobs, Chao, ACE, Shannon, and Simpson alpha diversity indices, as well as the coverage evenness and SD values are shown in Table 1. There was no statistically significant difference in the Sobs, ACE, Chao, Shannon and Simpson indices between RRTI with FH and RRTI without FH, as determined by the t-test (P > 0.05). Beta diversity analysis (PCA, PCoA and NMDS) showed no separation between RRTI with FH and RRTI without FH (Fig. S1). These results may indicate that family history of RRTI will not influence the following analysis.
3.4. Comparison of bacterial communities at the phylum level from RRTI and healthy groups When comparing the alpha diversity between samples, except for the Simpson and Shannon indices, significant differences were observed. Furthermore, PCA showed distinct bacterial community structures between the groups. To investigate the specific changes in the microbiota in RRTI samples, we assessed the relative abundance of specific taxa in the RRTI and healthy control groups. Fig. 2A shows the taxonomic distributions of the predominant bacteria (relative abundance > 1% of the total sequences) at phylum level. The bacterial microbiota analysis showed that the Firmicutes were the most predominant phylum, contributing 56.7% and 54.57% of the microbiota in healthy volunteers and RRTI patients, respectively. The second most dominant phyla in healthy volunteers and RRTI patients was the Bacteroidetes (24.85% and 23%) (Fig. S6). However, there was a clear decrease in both the Verrucomicrobia (0.001 < P < 0.01) and Tenericutes (0.01 < P < 0.05) phyla in the RRTI patients compared to the control group (Fig. 2B). Otherwise, there was a moderate decrease in the Firmicutes and Actinobacteria phyla in the RRTI group (P > 0.05). Of the remaining taxa, two phyla, including the Bacteroidetes and the Proteobacteria, were more enriched in RRTI patients (P > 0.05).
3.2. Alpha diversity analysis between RRTI and healthy groups In total, 3068255 usable sequences were obtained from 49 fecal samples by sequencing. Additionally, 499 OTUs were delineated at a 97% similarity level. The Good's coverage values for all libraries were above 99%. The Sobs, Chao, ACE, Shannon, and Simpson alpha diversity indices, as well as the coverage evenness and SD values are shown in Table 2. There was a statistically significant decrease in the Sobs, ACE, and Chao indices in the RRTI group, as determined by the ttest (P < 0.001), while there was a moderate difference in the Simpson and Shannon indices between the groups. In terms of the rank-abundance curve, the curve for the healthy group was higher on the horizontal axis and in the horizontal direction, and the bacterial abundance in the healthy group was higher than that of the RRTI group (Fig. S2). As seen from the Venn diagram, there were 280 common OTUs between the two different groups, but 31 OTUs were unique to the control group (Fig. S3). All of these results showed a decrease in the bacterial diversity in patients with RRTI compared to that in healthy volunteers.
3.5. Comparison of bacterial communities at the genus level in the RRTI and healthy groups As shown in Fig. 3, we obtained 68 genera from 26 RRTI fecal specimens and 23 healthy fecal specimens. The seven most abundant genera were Bacteroidetes, Bifidobacterium, Faecalibacterium, Enterococcus, Escherichia-Shigella, Blautia and Veillonella (Fig. 3A). As shown in Fig. 3B, it is worth noting that Enterococcus was overrepresented in the RRTI group relative to the controls (P < 0.001). Conversely, the abundance of Eubacterium (P < 0.001) was significantly decreased in the RRTI group. In addition, there was a
3.3. Characteristics of the beta diversity analyses between RRTI and healthy samples Beta diversity analysis represents the extent of similarity between different microbial communities and was calculated based on PCA. PCA based on 499 OTUs (grouped at 97% sequence identity) revealed a
Table 2 Richness and alpha diversity of the bacterial community between RRTI and healthy individuals. Group
RRTI Healthy
samples
Feces (N = 26) Feces (N = 23)
Sobs
ACE
Chao
Shannon
Simpson
Coverage
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
89.462*** 140
33.349 47.606
109.97*** 159.25
33.031 49.592
104.79*** 162.67
36.024 51.995
2.193 2.6348
0.866 0.783
0.25885 0.19329
0.252 0.157
0.99946* 0.99961
0.00016 0.00023
Note: 0.01 < p < = 0.05, marked *, 0.001 < p < = 0.01, marked as * *, p < = 0.001, marked as * * *. 3
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 1. PCA plots based on Bray-Curtis metrics between RRTI and healthy groups. Each symbol represents one sample.
H et al. found that patients with asthma and COPD had a higher abundance of Pseudomonas and Streptococcus than healthy people [22]. Korean scholars have reported that patients with acute respiratory distress syndrome caused by Clostridium infection were cured after intestinal microbiota transplantation [23]. Dutch researchers found that patients with pneumonia could be effectively distinguished from healthy people using five groups of bacteria, with a specificity of 0.95 and a sensitivity of 0.84 [24]. Although there have been many reports on the relationship between respiratory diseases and the intestinal microbiota, no studies have determined the relationship between RRTI and the intestinal microbiota. In our study, the gut microbiome of RRTI patients had a reduced species richness, as well as a significant shift in the overall microbial diversity. There were statistically significant decreases in the Sobs, ACE, and Chao indices in the RRTI group as determined by the t-test (P < 0.001). This was also supported by the Venn diagram and rankabundance curves. Our PCA was based on the first two principal component scores, which indicated that disease (RRTI) may be the most important factor contributing to changes in the gut bacterial structure. In other words, the development of RRTI may contribute to the microbiota imbalance. It is worth mentioning that there has been little to no research on the relationship between the intestinal microbiota and RRTI. Local microbiota disturbances may further promote the development of RRTI. Disturbances in the balance between RRTI and the microbiota, whether causal or consequential, may still be a problem and further studies are required to clarify their relationship [25]. When comparing the alpha diversity between samples, except for the Simpson and Shannon indices, there were significant differences observed. Furthermore, the PCA showed distinct bacterial community structures between the groups. To further clarify the specific differences in the intestinal microbiota between the RRTI and control groups, taxonomic analyses at phylum and genus levels and beta diversity analyses were carried out. At the phylum level, the Firmicutes were the most predominant phylum. The second most dominant phylum was Bacteroidetes. As we all know, the gut microbiota is dominated by the Bacteroidetes and Firmicutes phyla [26]. In addition, there was a significantly decrease in the Verrucomicrobia and Tenericutes phyla (P < 0.01) in RRTI patients compared to healthy volunteers. Furthermore, at the genus level, we observed statistically significant
moderate decrease in the Bacteroidetes (0.01 < P < 0.05) and the Faecalibacterium (0.01 < P < 0.05) in the RRTI group compared to the healthy group. These differences were further supported by the LDA analysis. Fig. S6 shows bacterial taxa that were differentially represented between the RRTI and healthy participants. In healthy feces, 80 bacterial taxa were significantly more abundant (e.g., Eubacterium, Faecalibacterium, and Bifidobacterium), while only 24 taxa were overrepresented in the RRTI group (e.g., Enterococcus). Enterococcus, Eubacterium, Bifidobacterium and Faecalibacterium were among the top 10 effective genera that contributed to the bacterial community differences (Fig. S7). 3.6. The accuracy of RRTI discrimination based on three genera A random forest model was used to identify the ability of the gut microbiota to discriminate RRTI status, which was based on the gut microbiota signature consisting of 17 RRTI-associated genera. Using ROC analysis, we assessed the performance of Enterococcus, Eubacterium and Bifidobacterium, achieving AUC values of 0.860, 0.820, and 0.689, respectively (Fig. 4). These results confirmed that using Enterococcus as a biomarker has the highest accuracy in diagnosing RRTI, which is followed by Eubacterium. However, as a probiotic, the accuracy of the diagnosis for RRTI using Bifidobacterium was much lower. This result indicated that gut microbiota-based classifiers are able to distinguish RRTI from healthy controls. 4. Discussion In recent years, the incidence rate of RRTI has been on rise, affecting the growth of children [19]. Many studies have shown that imbalance in the intestinal microbiota is associated with a variety of respiratory diseases, such as asthma, chronic obstructive pulmonary disease, acute distress and respiratory syndrome [20]. In our study, we assessed the microbiome of fecal samples, and there was a relative correlation between RRTI and the intestinal tract microbiota in children. The following studies are representative of the research on the relationship between the bacterial microbiota and respiratory diseases in recent years. In a study of the intestinal microbiota in asthmatic patients, it was found that there was a low abundance of Bifidobacteria in the asthmatic patients compared to healthy control patients [21]. Park 4
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 2. Distribution of the predominant bacteria at phylum level (A). The predominant taxa (> 1% relative abundance) in each level are shown. Bar chart of multispecies difference by Wilcoxon rank-sum test (B). Note: 0.01 < P < = 0.05, marked *, 0.001 < P < = 0.01, marked as * *, P < = 0.001, marked as * * *.
contrast, certain therapies that are often used, such as antibiotics and chemotherapy, may negatively affect the composition and function of the gut microbiota, and as a result, the well-being of patients. As the microbiota research field is currently moving from association studies to intervention studies and even clinical trials, the implementation of this new knowledge into clinical practice is approaching. Several therapeutic interventions that target the gut microbiota are being evaluated, ranging from supplementation of food components to transplantation of the fecal microbiota [29]. Emerging experimental and epidemiological evidence underscores an important crosstalk between the intestinal microbiota and the lungs, referred to as the ‘gut–lung axis’ [30]. Aspiration of the content of the bronchus may deliver approximately 1011 live bacteria per day into the gut. Gut-derived injury factors can reach the lung and systemic circulation via the intestinal lymphatic system. Changes in the constituents of the gut microbiome, through either diet, disease or medical interventions (such as antibiotics) are linked to altered immune responses and homeostasis in the airways. The protective role of short-chain fatty acids (SCFAs) has been extensively studied in the murine gut. Accumulating evidence supports an anti-inflammatory and
differences in Enterococcus, Faecalibacterium, Bifidobacterium and Eubacterium between RRTI patients and healthy volunteers. In particular, the abundance of Enterococcus in the RRTI group was much higher than in the healthy group (P < 0.001). Previous studies have found that Enterococcus is an opportunistic pathogen, which can cause communityacquired and nosocomial infections [25,26]. Upon further analysis, we found that using Enterococcus as a biomarker for RRTI had the highest accuracy among the three selected genera in diagnosing RRTI. Bifidobacterium is the most well-known probiotic and is an important part of the gut microbiota. Probiotics are colonizers of the intestinal tract and form biofilms, which prevent adhesion and invasion of pathogens. Furthermore, probiotics can maintain the tight protein structures of the host and reduce cytokine production, thus regulating inflammation and the immune system [27]. In our studies, we observed a lower abundance of Bifidobacterium in the RRTI group. Previous studies in populations of children found lower levels of Bacteroidetes in the feces of allergic subjects than in non-allergic individuals [28]. Interestingly, the amount of Eubacterium was significantly higher in healthy volunteers than in RRTI patients. Therefore, it is essential to seek a biological approach to maintain the intestinal microecological balance. By 5
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 3. Distribution of the predominant bacteria in genus level in Circos (A). Bar chart of multi-species difference by Wilcoxon rank-sum test (B).
6
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 4. Gut microbiota signature which can be used to discriminate RRTI patients from healthy controls. Receiving operational curve analysis was performed in Enterococcus, ((area under curve) AUC = 0.860) and in Bifidobacterium (AUC = 0.689), in Eubacterium_rectale (AUC = 0.820) respectively. Diagonal lines represent random classification (AUC = 0.5).
immunomodulatory action of SCFAs in the periphery, in particular, along the gut–lung axis. Similar to the intestinal tract, SCFAs are able to generate an extrathymic peripheral Treg cellular pool, which is linked to the dampening of allergic airway diseases through HDAC inhibition. SCFAs are mainly composed of acetate, propionate and butyrate. In the present study, the abundance of major butyrate-producers, Eubacterium and Faecalibacterium species, was decreased in RRTI patients, which may result in decreased SCFAs and disorders in intestinal functions. In conclusion, we have identified a structural imbalance in the gut microbiota by comparing the intestinal microbial composition of RRTI patients to healthy individuals. These aberrant microbial profiles were represented by an increased incidence of opportunistic pathogens, such as Enterococcus, and a reduction in Eubacterium_rectale in patients. PCA and PCoA analysis suggested that respiratory disease may be the primary factor leading to changes in the structure of the bacterial community. These results provide fundamental evidence supporting that intestinal microecological dysbiosis may lead to the decline in immune function in children.
[10] M.K. Gibson, T.S. Crofts, G. Dantas, Antibiotics and the developing infant gut microbiota and resistome, Curr. Opin. Microbiol. 27 (2015) 51–56. [11] P. Garagnani, C. Pirazzini, C. Franceschi, Colorectal cancer microenvironment: among nutrition, gut microbiota, inflammation and epigenetics, Curr. Pharmaceut. Des. 19 (2013) 765–778. [12] M.C. Suarez-Arrabal, C. Mella, S.M. Lopez, N.V. Brown, M.W. Hall, S. Hammond, et al., Nasopharyngeal bacterial burden and antibiotics: influence on inflammatory markers and disease severity in infants with respiratory syncytial virus bronchiolitis, J. Infect. 71 (2015) 458–469. [13] A.J. Klecha, J. Salgueiro, M. Wald, J. Boccio, M. Zubillaga, N.M. Leonardi, et al., In vivo iron and zinc deficiency diminished T- and B-selective mitogen stimulation of murine lymphoid cells through protein kinase C-mediated mechanism, Biol. Trace Elem. Res. 104 (2005) 173–183. [14] R. Tang, Y. Wei, Y. Li, W. Chen, H. Chen, Q. Wang, et al., Gut microbial profile is altered in primary biliary cholangitis and partially restored after UDCA therapy, Gut 67 (2018) 534–541. [15] Y. Lu, J. Chen, J. Zheng, G. Hu, J. Wang, C. Huang, et al., Mucosal adherent bacterial dysbiosis in patients with colorectal adenomas, Sci. Rep. 6 (2016) 26337. [16] E. Aguirre, A. Galiana, A. Mira, R. Guardiola, L. Sanchez-Guillen, E. Garcia-Pachon, et al., Analysis of microbiota in stable patients with chronic obstructive pulmonary disease, APMIS : APMIS 123 (2015) 427–432. [17] K.R. Amato, C.J. Yeoman, A. Kent, N. Righini, F. Carbonero, A. Estrada, et al., Habitat degradation impacts black howler monkey (Alouatta pigra) gastrointestinal microbiomes, ISME J. 7 (2013) 1344–1353. [18] M. Hamady, C. Lozupone, R. Knight, Fast UniFrac: facilitating high-throughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data, ISME J. 4 (2010) 17–27. [19] W. Hua, C.B. Shen, Observations on the effect of ferrous gluconate on children with RRTI, J. Bengbu Med. Coll. 29 (2004) 428–429. [20] E.S. Gollwitzer, B.J. Marsland, Microbiota abnormalities in inflammatory airway diseases - potential for therapy, Pharmacol. Ther. 141 (2014) 32–39. [21] A. Hevia, C. Milani, P. Lopez, C.D. Donado, A. Cuervo, S. Gonzalez, et al., Allergic patients with long-term asthma display low levels of Bifidobacterium adolescentis, PLoS One 11 (2016) e0147809. [22] H. Park, J.W. Shin, S.G. Park, W. Kim, Microbial communities in the upper respiratory tract of patients with asthma and chronic obstructive pulmonary disease, PLoS One 9 (2014) e109710. [23] G. Li Volti, G. Malaponte, V. Bevelacqua, A. Messina, S. Bianca, M.C. Mazzarino, et al., Persistent high plasma levels of interleukins 18 and 4 in children with recurrent infections of the upper respiratory tract, Transplant. Proc. 35 (2003) 2911–2915. [24] W.A. de Steenhuijsen Piters, E.G. Huijskens, A.L. Wyllie, G. Biesbroek, M.R. van den Bergh, R.H. Veenhoven, et al., Dysbiosis of upper respiratory tract microbiota in elderly pneumonia patients, ISME J. 10 (2016) 97–108. [25] Z.J. Yu, Z. Chen, H. Cheng, J.X. Zheng, W.G. Pan, W.Z. Yang, et al., Recurrent linezolid-resistant Enterococcus faecalis infection in a patient with pneumonia, Int. J. Infect. Dis. : IJID Off. Publ. Int. Soc. Infect. Dis. 30 (2015) 49–51. [26] S. Ager, K. Gould, Clinical update on linezolid in the treatment of Gram-positive bacterial infections, Infect. Drug Resist. 5 (2012) 87–102. [27] S. Macfarlane, J.F. Dillon, Microbial biofilms in the human gastrointestinal tract, J. Appl. Microbiol. 102 (2007) 1187–1196. [28] M.M. Gronlund, M. Gueimonde, K. Laitinen, G. Kociubinski, T. Gronroos, S. Salminen, et al., Maternal breast-milk and intestinal bifidobacteria guide the compositional development of the Bifidobacterium microbiota in infants at risk of allergic disease, Clin. Exp. Allergy : J. Br. Soc. Allergy Clin. Immunol. 37 (2007) 1764–1772. [29] S.N. Chen, Treatment of central serous chorioretinopathy: might medical treatment play arole? Taiwan J. Ophthalmol. 3 (2013) 1–2. [30] A.T. Dang, B.J. Marsland, Microbes, metabolites, and the gut-lung axis, Mucosal Immunol. 12 (2019) 843–850.
Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.micpath.2019.103709. References [1] A. Lemanska-Perek, B. Polanska, D. Krzyzanowska-Golab, I. Katnik-Prastowska, Occurrence of soluble supra-molecular FN-fibrin complexes in the plasma of children with recurrent respiratory infection, Ann. Clin. Biochem. 52 (2015) 441–447. [2] S. Esposito, G. Marseglia, A. Novelli, M. de Martino, G. Di Mauro, C. Gabiano, et al., [Acute, subacute and recurrent bacterial rhinosinusitis in pediatrics: guidelines of the study group of the Italian society for pediatric infectious diseases (SITIP)], Minerva Pediatr. 59 (2007) 474–475. [3] M.T. van den Aardweg, C.W. Boonacker, M.M. Rovers, A.W. Hoes, A.G. Schilder, Effectiveness of adenoidectomy in children with recurrent upper respiratory tract infections: open randomised controlled trial, BMJ 343 (2011) d5154. [4] J. Pasnik, Vaccines nonspecific - immunostimulation in patients with recurrent respiratory infections, Otolaryngol. Pol. 70 (2016) 31–39. [5] D.A. van Kessel, T.W. Hoffman, H. van Velzen-Blad, P. Zanen, J.C. Grutters, G.T. Rijkers, Long-term clinical outcome of antibody replacement therapy in humoral immunodeficient adults with respiratory tract infections, EBioMedicine 18 (2017) 254–260. [6] S. Mao, A. Zhang, S. Huang, Meta-analysis of Zn, Cu and Fe in the hair of Chinese children with recurrent respiratory tract infection, Scand. J. Clin. Lab. Investig. 74 (2014) 561–567. [7] X. Jiang, L. Sun, B. Wang, X. Yang, L. Shang, Y. Zhang, Health-related quality of life among children with recurrent respiratory tract infections in Xi'an, China, PLoS One 8 (2013) e56945. [8] C.Y. Lu, Y.H. Ni, Gut microbiota and the development of pediatric diseases, J. Gastroenterol. 50 (2015) 720–726. [9] J.T. Ho, G.C. Chan, J.C. Li, Systemic effects of gut microbiota and its relationship with disease and modulation, BMC Immunol. 16 (2015) 21.
7
Contents lists available at ScienceDirect
Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath
Intestinal microbiota dysbiosis in children with recurrent respiratory tract infections
T
Lei Lia, Fang Wanga, Yanni Liub, Feng Gua,∗ a b
Department of Pediatrics, The Affiliated Hospital of Qingdao University of Shandong Province, Qingdao, China Department of Obstetrics and Gynecology, Binzhou Medical University Hospital of Shandong Province, Binzhou, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Recurrent respiratory tract infection Gut microbiome Bacterial diversity 16S rRNA
Background: The impact of the gut microbiota on recurrent respiratory tract infection (RRTI) remains to be fully elucidated. Methods: To characterize the gut microbiota in patients with RRTI, fecal samples from 26 patients with RRTI and 23 healthy volunteers were profiled using the Illumina MiSeq platform. Beta diversity (Principal Component Analysis (PCA), Principal Co-ordinates Analysis (PCoA), Non-metric multidimensional scaling (NMDS)) analysis showed that the bacterial community structure segregated differently between the RRTI and control groups. Results: Results from alpha diversity analysis revealed lower microbiota diversity in samples from RRTI patients than in normal controls. Taxonomic analysis illustrated that the abundance of six phyla (Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, Verrucomicrobia, Tenericutes) and four genera (Enterococcus, Faecalibacterium, Bifidobacterium, Eubacterium were significantly different between these two groups. In addition, Enterococcus (P < 0.001) was more enriched in the RRTI group, whereas the abundances of Eubacterium (P < 0.001), Faecalibacterium (0.01 < P < 0.05) and Bifidobacterium (0.01 < P < 0.05) were reduced in the RRTI group compared to those in the normal control group. The performance of the model was assessed using ROC analysis, and Enterococcus, Eubacterium and Bifidobacterium achieved AUC values of 0.860, 0.820, and 0.689, respectively. Conclusions: These results provide fundamental evidence in support of intestinal microbiota dysbiosis in children with RRTI.
1. Introduction Recurrent respiratory tract infections (RRTIs) are one of the most common and most frequent diseases in children and adolescents [1]. According to the Italian Pediatric Society, the diagnostic criteria for RRTI in children are as follows: ≥ 6 respiratory tract infections in a year, ≥ 1 upper respiratory tract infection in a month between September and April, or ≥3 lower respiratory tract infections in a year [2]. RRTI is a common disease in pediatrics, occurring in children under 5 years of age, and accounting for 10%–30% of all pediatric respiratory infections, an incidence rate that increases each year [3]. In countries with high mortality rates, such as southeast Asia and Africa, RRTIs have a 23% mortality rate and impact two million people [4]. RRTIs are dangerous diseases in patients with antibody defects, which can be treated with antibody replacement therapy but the cost, estimated to be approximately 20,000–30,000 euros per patient per year, is relatively
high [5]. If not properly treated, RRTIs can lead to asthma, myocarditis, nephritis, and other diseases, which may seriously affect a child's growth and development [6]. Therefore, the pathogenesis of RRTIs has become the focus of clinicians to promote new approaches for the management and prevention of RRTI [7]. The human gut has a huge number of microbes, which are collectively referred to as the ‘‘microbiota’’ [8]. The gut microbiota constitutes approximately 70% of a person's entire microbiota, and the gut microbiota is dominated by the Bacteroidetes and Firmicutes phyla [9]. The dynamic composition of the human gut microbiota is determined by multiple factors, including the mode of delivery, diet, the environment, and antibiotics. A healthy gut microbiota is beneficial for the host in many ways, including by providing nutrients and protection from pathogens, and by inducing the maturation of immune responses [10]. Dysbiosis in the diversity of the microbiota can lead to physiological changes including in the gastrointestinal, respiratory, cardiovascular, or
∗ Corresponding author. Department of pediatrics, the Affiliated Hospital of Qingdao University, No.1677 Wutaishan Road, Huangdao Distict, Qingdao, 266555, China. E-mail address: [email protected] (F. Gu).
https://doi.org/10.1016/j.micpath.2019.103709 Received 5 May 2019; Received in revised form 29 August 2019; Accepted 2 September 2019 Available online 05 September 2019 0882-4010/ © 2019 Elsevier Ltd. All rights reserved.
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
2.3. Illumina MiSeq sequencing
neurological systems, in autoimmunity or hyperimmunity, and in chronic, cancerous, and psychiatric diseases [11]. In recent years, studies have found that respiratory diseases in children are also associated with the composition of the intestinal microbiota. Suarez-arrabal MC et al. [12] found that compared with those of control children, the nasopharyngeal and nasal areas of the children infected with RSV (respiratory syncytial virus bronchiolitis) were more likely to be colonized with Haemophilus influenzae, Pneumococcus sp. and mucositis (81% vs. 65%). It has been reported that the number of Bifidobacteria in the intestinal tract of Spaniards with long-term allergic asthma is significantly lower than in healthy Spaniards [13]. All of these studies have shown that the intestinal microbiota play a crucial role in the development of respiratory diseases, however, the impact of the gut microbiota in RRTI has not been adequately described. In this study, we performed a microbiome analysis of fecal samples from RRTI patients and healthy volunteers to elucidate the bacterial changes that might induce or accompany RRTI using a sequencing platform.
The PCR products were quantified using the QuantiFluor™-ST Blue Fluorescence Quantitative System (Promega) and the initial quantitative results from electrophoresis. The corresponding proportions were then mixed according to the sequencing requirements of each sample. Amplicons were extracted from 2% agarose gels and purified using the AxyPrep DNA Gel Extraction Kit according to the manufacturer's instructions and quantified using QuantiFluor™-ST (Promega, USA). Purified amplicons were pooled in equimolar amounts and sequenced (paired-end; 2 × 250) on the Illumina MiSeq platform according to standard protocols.
2.4. Processing of sequencing data Raw FASTQ files were demultiplexed and quality-filtered using FLASH Trimmomatic with the following criteria: i) Using a 50 bp sliding window, reads with adaptors under 20 bp, reads with an average quality value of less than 20, reads less than 50 bp, and reads containing N bases were filtered out; ii) PE reads with a minimum overlap of 10 bp were merged into a single sequence; iii) Sequences with a mismatch ratio higher than 0.2 in the overlapping area were filtered out; iv) Sequences were identified and oriented according to their barcodes and primers; sequences with barcode mismatches greater than 0 or primer mismatches greater than 2 were filtered out [17].
2. Methods 2.1. Sample collection Fecal samples were collected from 26 children with RRTIs and 23 healthy volunteers of the Affiliated Hospital of the Medical College Qingdao University. The diagnosis of RRTI was performed according to the criteria issued by the Pediatric Society of the Chinese Medical Association in 2007. Additionally, the children with RRTIs were eligible for enrollment only if they had no history of other diseases and no history of antibiotic or probiotic use. Exclusion criteria were as follows: (1) Those with chronic liver disease, kidney disease or basic diseases of the digestive tract were excluded; (2) Patients who had recently received antibiotics were also excluded; (3) Patients who were diagnosed with RRTI caused by organic or congenital lesions, or primary immunodeficiency were excluded [14]. Two weeks before fecal sample collection, patients and healthy volunteers had to have not taken antibiotics, probiotics, Chinese herbal medicines, hormones or other medications that would affect bacterial structure. Furthermore, the participants (> 5 years old) were not permitted to eat food in the 12 h before fecal sample collection, and the sample collection environment was confirmed to be clean. All samples were collected in accordance with the relevant guidelines and regulations, and the research was approved by the Research Ethics Boards of Qingdao General Hospital. Informed consent was provided by the parents of all participants. We collected 3–5 g of fresh fecal samples and immediately placed them into clean, sterilized centrifuge tubes. All samples were frozen immediately after sampling and stored at −80 °C [15].
2.5. Bacterial community characterization The high-throughput sequencing reads from 49 fecal samples were reassigned to their respective samples according to their barcodes. Operation taxonomic units (OTUs) were clustered with a 97% similarity cutoff using Usearch (Version 7.0, http://drive5.com/uparse/), and chimeric sequences were identified and removed using UCHIME. The OTUs with a 97% similarity level were used for alpha diversity analysis, which analyzed the species diversity of individual samples through the evaluation of Chao, abundance-based coverage estimators (ACE), Shannon, and Simpson parameters. The rank-abundance curves were also analyzed using mothur and R software. A Venn diagram was generated to show the shared and unique OTUs among the groups, based on the occurrence of OTUs in a sample group and regardless of their relative abundance; these were also analyzed using R software.The Wilcoxon rank-sum test was used to identify significant differences in the alpha diversity indices between the different groups (P < 0.05). Beta diversity was analyzed to investigate the similarity of the bacterial community structures among the groups using Bray Curtis distances. The beta diversity was visualized via principal component analysis (PCA), principal coordinate analysis (PCoA) and Nonmetric Multidimensional Analysis (NMDS). The Wilcoxon rank-sum test was used to detect significant differences in abundance between the groups. LEfSe (http://huttenhower.sph.harvard.edu/galaxy/root?tool_id= lefse_upload) [18] uses linear discriminant analysis (LDA) to estimate the effect that the abundance of each component has on the differences. The Circos sample and species relationship map is a visual circle diagram that describes the correspondence between sample and species. To evaluate the discriminatory ability of the bacteria, operating characteristic curves (Receiver Operating Characteristic, ROC) were constructed and the area under curve (AUC) was calculated.
2.2. DNA extraction and PCR amplification Microbial DNA was extracted from fecal samples using the OMEGAsoil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to manufacturer's protocols. The V3–V4 region of the bacterial 16S ribosomal RNA gene were amplified by PCR (ABI GeneAmp® 9700; 95 °C for 2 min, followed by 25 cycles at 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s and a final extension at 72 °C for 5 min) using primers 338F and 806R, which included unique eight-base barcode sequences for each sample. PCRs were performed in triplicate in 20 μL mixtures containing 4 μL of 5 × Fast Pfu buffer, 2 μL of 2.5 mM dNTPs, 0.8 μL of each primer (5 μM), 0.4 μL of Fast Pfu polymerase, and 10 ng of template DNA. All samples were subjected to the formal test conditions, and each sample was duplicated. The duplicate PCR products from the same sample were mixed and tested by 2% agarose gel electrophoresis. The PCR products were recovered using 2% agarose gel electrophoresis and the AxyPrepDNA Gel Recovery Kit (Axygen Biosciences, Union City, CA, US A), with elution in Tris-HCl [16].
2.5.1. Statistical analysis All statistical analyses were performed using R packages (V.2.15.3). The Student's t-test was performed using SPSS version 20 for Windows. For correlation analysis, Spearman's rank test was performed. Multiple hypotheses tests were adjusted using the Benjamini and Hochberg false discovery rate (FDR). 2
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Table 1 Alpha diversity between patients had family history of RRTI (RRTI with FH) and patients did not have family history of RRTI (RRTI without FH). Group
samples
RRTI with FH RRTI without FH P-value
Sobs
Feces (N = 7) Feces (N = 19) –
ACE
Chao
Shannon
Simpson
Coverage
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
113 100.89 0.5031
38.962 40.698
137.49 132.05 0.7477
37.766 37.811
137.22 128.22 0.6421
39.151 44.559
2.3787 2.1594 0.5748
0.96333 0.83928
0.24056 0.26035 0.861
0.24552 0.25547
0.99962 0.99962 0.997
0.00013523 0.00015349
Note: 0.01 < p < = 0.05, marked *, 0.001 < p < = 0.01, marked as * *, p < = 0.001, marked as * *.
separation between the healthy and RRTI groups in the first two principal component scores, which accounted for 28.5% and 15.3% of the total variations, respectively. This suggested that disease may be an important factor that accounts for the changes in the structure of the intestinal microbiota (Fig. 1). PCoA also showed a clear separation between the RRTI and healthy groups based on weighted analysis (Fig. S4). The figure shows that the percent of variation explained by PC1 and PC2 was 19.59% and 10.37%, respectively. The NMDS of the bacterial community also showed a separation between the RRTI and healthy groups (Fig. S5).
3. Results 3.1. Characterization of study population A total of 49 subjects were analyzed. Basic participant information was recorded, including age, sex, gestational age, feeding mode, mode of production, family history of RRTI, and secondhand smoke exposure (Supplementary Table S1 and Table S2). Based on the characteristics of family history of RRTI, we analyzed the alpha and beta diversity between patients had family history of RRTI (RRTI with FH) and patients did not have family history of RRTI (RRTI without FH). The Sobs, Chao, ACE, Shannon, and Simpson alpha diversity indices, as well as the coverage evenness and SD values are shown in Table 1. There was no statistically significant difference in the Sobs, ACE, Chao, Shannon and Simpson indices between RRTI with FH and RRTI without FH, as determined by the t-test (P > 0.05). Beta diversity analysis (PCA, PCoA and NMDS) showed no separation between RRTI with FH and RRTI without FH (Fig. S1). These results may indicate that family history of RRTI will not influence the following analysis.
3.4. Comparison of bacterial communities at the phylum level from RRTI and healthy groups When comparing the alpha diversity between samples, except for the Simpson and Shannon indices, significant differences were observed. Furthermore, PCA showed distinct bacterial community structures between the groups. To investigate the specific changes in the microbiota in RRTI samples, we assessed the relative abundance of specific taxa in the RRTI and healthy control groups. Fig. 2A shows the taxonomic distributions of the predominant bacteria (relative abundance > 1% of the total sequences) at phylum level. The bacterial microbiota analysis showed that the Firmicutes were the most predominant phylum, contributing 56.7% and 54.57% of the microbiota in healthy volunteers and RRTI patients, respectively. The second most dominant phyla in healthy volunteers and RRTI patients was the Bacteroidetes (24.85% and 23%) (Fig. S6). However, there was a clear decrease in both the Verrucomicrobia (0.001 < P < 0.01) and Tenericutes (0.01 < P < 0.05) phyla in the RRTI patients compared to the control group (Fig. 2B). Otherwise, there was a moderate decrease in the Firmicutes and Actinobacteria phyla in the RRTI group (P > 0.05). Of the remaining taxa, two phyla, including the Bacteroidetes and the Proteobacteria, were more enriched in RRTI patients (P > 0.05).
3.2. Alpha diversity analysis between RRTI and healthy groups In total, 3068255 usable sequences were obtained from 49 fecal samples by sequencing. Additionally, 499 OTUs were delineated at a 97% similarity level. The Good's coverage values for all libraries were above 99%. The Sobs, Chao, ACE, Shannon, and Simpson alpha diversity indices, as well as the coverage evenness and SD values are shown in Table 2. There was a statistically significant decrease in the Sobs, ACE, and Chao indices in the RRTI group, as determined by the ttest (P < 0.001), while there was a moderate difference in the Simpson and Shannon indices between the groups. In terms of the rank-abundance curve, the curve for the healthy group was higher on the horizontal axis and in the horizontal direction, and the bacterial abundance in the healthy group was higher than that of the RRTI group (Fig. S2). As seen from the Venn diagram, there were 280 common OTUs between the two different groups, but 31 OTUs were unique to the control group (Fig. S3). All of these results showed a decrease in the bacterial diversity in patients with RRTI compared to that in healthy volunteers.
3.5. Comparison of bacterial communities at the genus level in the RRTI and healthy groups As shown in Fig. 3, we obtained 68 genera from 26 RRTI fecal specimens and 23 healthy fecal specimens. The seven most abundant genera were Bacteroidetes, Bifidobacterium, Faecalibacterium, Enterococcus, Escherichia-Shigella, Blautia and Veillonella (Fig. 3A). As shown in Fig. 3B, it is worth noting that Enterococcus was overrepresented in the RRTI group relative to the controls (P < 0.001). Conversely, the abundance of Eubacterium (P < 0.001) was significantly decreased in the RRTI group. In addition, there was a
3.3. Characteristics of the beta diversity analyses between RRTI and healthy samples Beta diversity analysis represents the extent of similarity between different microbial communities and was calculated based on PCA. PCA based on 499 OTUs (grouped at 97% sequence identity) revealed a
Table 2 Richness and alpha diversity of the bacterial community between RRTI and healthy individuals. Group
RRTI Healthy
samples
Feces (N = 26) Feces (N = 23)
Sobs
ACE
Chao
Shannon
Simpson
Coverage
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
89.462*** 140
33.349 47.606
109.97*** 159.25
33.031 49.592
104.79*** 162.67
36.024 51.995
2.193 2.6348
0.866 0.783
0.25885 0.19329
0.252 0.157
0.99946* 0.99961
0.00016 0.00023
Note: 0.01 < p < = 0.05, marked *, 0.001 < p < = 0.01, marked as * *, p < = 0.001, marked as * * *. 3
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 1. PCA plots based on Bray-Curtis metrics between RRTI and healthy groups. Each symbol represents one sample.
H et al. found that patients with asthma and COPD had a higher abundance of Pseudomonas and Streptococcus than healthy people [22]. Korean scholars have reported that patients with acute respiratory distress syndrome caused by Clostridium infection were cured after intestinal microbiota transplantation [23]. Dutch researchers found that patients with pneumonia could be effectively distinguished from healthy people using five groups of bacteria, with a specificity of 0.95 and a sensitivity of 0.84 [24]. Although there have been many reports on the relationship between respiratory diseases and the intestinal microbiota, no studies have determined the relationship between RRTI and the intestinal microbiota. In our study, the gut microbiome of RRTI patients had a reduced species richness, as well as a significant shift in the overall microbial diversity. There were statistically significant decreases in the Sobs, ACE, and Chao indices in the RRTI group as determined by the t-test (P < 0.001). This was also supported by the Venn diagram and rankabundance curves. Our PCA was based on the first two principal component scores, which indicated that disease (RRTI) may be the most important factor contributing to changes in the gut bacterial structure. In other words, the development of RRTI may contribute to the microbiota imbalance. It is worth mentioning that there has been little to no research on the relationship between the intestinal microbiota and RRTI. Local microbiota disturbances may further promote the development of RRTI. Disturbances in the balance between RRTI and the microbiota, whether causal or consequential, may still be a problem and further studies are required to clarify their relationship [25]. When comparing the alpha diversity between samples, except for the Simpson and Shannon indices, there were significant differences observed. Furthermore, the PCA showed distinct bacterial community structures between the groups. To further clarify the specific differences in the intestinal microbiota between the RRTI and control groups, taxonomic analyses at phylum and genus levels and beta diversity analyses were carried out. At the phylum level, the Firmicutes were the most predominant phylum. The second most dominant phylum was Bacteroidetes. As we all know, the gut microbiota is dominated by the Bacteroidetes and Firmicutes phyla [26]. In addition, there was a significantly decrease in the Verrucomicrobia and Tenericutes phyla (P < 0.01) in RRTI patients compared to healthy volunteers. Furthermore, at the genus level, we observed statistically significant
moderate decrease in the Bacteroidetes (0.01 < P < 0.05) and the Faecalibacterium (0.01 < P < 0.05) in the RRTI group compared to the healthy group. These differences were further supported by the LDA analysis. Fig. S6 shows bacterial taxa that were differentially represented between the RRTI and healthy participants. In healthy feces, 80 bacterial taxa were significantly more abundant (e.g., Eubacterium, Faecalibacterium, and Bifidobacterium), while only 24 taxa were overrepresented in the RRTI group (e.g., Enterococcus). Enterococcus, Eubacterium, Bifidobacterium and Faecalibacterium were among the top 10 effective genera that contributed to the bacterial community differences (Fig. S7). 3.6. The accuracy of RRTI discrimination based on three genera A random forest model was used to identify the ability of the gut microbiota to discriminate RRTI status, which was based on the gut microbiota signature consisting of 17 RRTI-associated genera. Using ROC analysis, we assessed the performance of Enterococcus, Eubacterium and Bifidobacterium, achieving AUC values of 0.860, 0.820, and 0.689, respectively (Fig. 4). These results confirmed that using Enterococcus as a biomarker has the highest accuracy in diagnosing RRTI, which is followed by Eubacterium. However, as a probiotic, the accuracy of the diagnosis for RRTI using Bifidobacterium was much lower. This result indicated that gut microbiota-based classifiers are able to distinguish RRTI from healthy controls. 4. Discussion In recent years, the incidence rate of RRTI has been on rise, affecting the growth of children [19]. Many studies have shown that imbalance in the intestinal microbiota is associated with a variety of respiratory diseases, such as asthma, chronic obstructive pulmonary disease, acute distress and respiratory syndrome [20]. In our study, we assessed the microbiome of fecal samples, and there was a relative correlation between RRTI and the intestinal tract microbiota in children. The following studies are representative of the research on the relationship between the bacterial microbiota and respiratory diseases in recent years. In a study of the intestinal microbiota in asthmatic patients, it was found that there was a low abundance of Bifidobacteria in the asthmatic patients compared to healthy control patients [21]. Park 4
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 2. Distribution of the predominant bacteria at phylum level (A). The predominant taxa (> 1% relative abundance) in each level are shown. Bar chart of multispecies difference by Wilcoxon rank-sum test (B). Note: 0.01 < P < = 0.05, marked *, 0.001 < P < = 0.01, marked as * *, P < = 0.001, marked as * * *.
contrast, certain therapies that are often used, such as antibiotics and chemotherapy, may negatively affect the composition and function of the gut microbiota, and as a result, the well-being of patients. As the microbiota research field is currently moving from association studies to intervention studies and even clinical trials, the implementation of this new knowledge into clinical practice is approaching. Several therapeutic interventions that target the gut microbiota are being evaluated, ranging from supplementation of food components to transplantation of the fecal microbiota [29]. Emerging experimental and epidemiological evidence underscores an important crosstalk between the intestinal microbiota and the lungs, referred to as the ‘gut–lung axis’ [30]. Aspiration of the content of the bronchus may deliver approximately 1011 live bacteria per day into the gut. Gut-derived injury factors can reach the lung and systemic circulation via the intestinal lymphatic system. Changes in the constituents of the gut microbiome, through either diet, disease or medical interventions (such as antibiotics) are linked to altered immune responses and homeostasis in the airways. The protective role of short-chain fatty acids (SCFAs) has been extensively studied in the murine gut. Accumulating evidence supports an anti-inflammatory and
differences in Enterococcus, Faecalibacterium, Bifidobacterium and Eubacterium between RRTI patients and healthy volunteers. In particular, the abundance of Enterococcus in the RRTI group was much higher than in the healthy group (P < 0.001). Previous studies have found that Enterococcus is an opportunistic pathogen, which can cause communityacquired and nosocomial infections [25,26]. Upon further analysis, we found that using Enterococcus as a biomarker for RRTI had the highest accuracy among the three selected genera in diagnosing RRTI. Bifidobacterium is the most well-known probiotic and is an important part of the gut microbiota. Probiotics are colonizers of the intestinal tract and form biofilms, which prevent adhesion and invasion of pathogens. Furthermore, probiotics can maintain the tight protein structures of the host and reduce cytokine production, thus regulating inflammation and the immune system [27]. In our studies, we observed a lower abundance of Bifidobacterium in the RRTI group. Previous studies in populations of children found lower levels of Bacteroidetes in the feces of allergic subjects than in non-allergic individuals [28]. Interestingly, the amount of Eubacterium was significantly higher in healthy volunteers than in RRTI patients. Therefore, it is essential to seek a biological approach to maintain the intestinal microecological balance. By 5
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 3. Distribution of the predominant bacteria in genus level in Circos (A). Bar chart of multi-species difference by Wilcoxon rank-sum test (B).
6
Microbial Pathogenesis 136 (2019) 103709
L. Li, et al.
Fig. 4. Gut microbiota signature which can be used to discriminate RRTI patients from healthy controls. Receiving operational curve analysis was performed in Enterococcus, ((area under curve) AUC = 0.860) and in Bifidobacterium (AUC = 0.689), in Eubacterium_rectale (AUC = 0.820) respectively. Diagonal lines represent random classification (AUC = 0.5).
immunomodulatory action of SCFAs in the periphery, in particular, along the gut–lung axis. Similar to the intestinal tract, SCFAs are able to generate an extrathymic peripheral Treg cellular pool, which is linked to the dampening of allergic airway diseases through HDAC inhibition. SCFAs are mainly composed of acetate, propionate and butyrate. In the present study, the abundance of major butyrate-producers, Eubacterium and Faecalibacterium species, was decreased in RRTI patients, which may result in decreased SCFAs and disorders in intestinal functions. In conclusion, we have identified a structural imbalance in the gut microbiota by comparing the intestinal microbial composition of RRTI patients to healthy individuals. These aberrant microbial profiles were represented by an increased incidence of opportunistic pathogens, such as Enterococcus, and a reduction in Eubacterium_rectale in patients. PCA and PCoA analysis suggested that respiratory disease may be the primary factor leading to changes in the structure of the bacterial community. These results provide fundamental evidence supporting that intestinal microecological dysbiosis may lead to the decline in immune function in children.
[10] M.K. Gibson, T.S. Crofts, G. Dantas, Antibiotics and the developing infant gut microbiota and resistome, Curr. Opin. Microbiol. 27 (2015) 51–56. [11] P. Garagnani, C. Pirazzini, C. Franceschi, Colorectal cancer microenvironment: among nutrition, gut microbiota, inflammation and epigenetics, Curr. Pharmaceut. Des. 19 (2013) 765–778. [12] M.C. Suarez-Arrabal, C. Mella, S.M. Lopez, N.V. Brown, M.W. Hall, S. Hammond, et al., Nasopharyngeal bacterial burden and antibiotics: influence on inflammatory markers and disease severity in infants with respiratory syncytial virus bronchiolitis, J. Infect. 71 (2015) 458–469. [13] A.J. Klecha, J. Salgueiro, M. Wald, J. Boccio, M. Zubillaga, N.M. Leonardi, et al., In vivo iron and zinc deficiency diminished T- and B-selective mitogen stimulation of murine lymphoid cells through protein kinase C-mediated mechanism, Biol. Trace Elem. Res. 104 (2005) 173–183. [14] R. Tang, Y. Wei, Y. Li, W. Chen, H. Chen, Q. Wang, et al., Gut microbial profile is altered in primary biliary cholangitis and partially restored after UDCA therapy, Gut 67 (2018) 534–541. [15] Y. Lu, J. Chen, J. Zheng, G. Hu, J. Wang, C. Huang, et al., Mucosal adherent bacterial dysbiosis in patients with colorectal adenomas, Sci. Rep. 6 (2016) 26337. [16] E. Aguirre, A. Galiana, A. Mira, R. Guardiola, L. Sanchez-Guillen, E. Garcia-Pachon, et al., Analysis of microbiota in stable patients with chronic obstructive pulmonary disease, APMIS : APMIS 123 (2015) 427–432. [17] K.R. Amato, C.J. Yeoman, A. Kent, N. Righini, F. Carbonero, A. Estrada, et al., Habitat degradation impacts black howler monkey (Alouatta pigra) gastrointestinal microbiomes, ISME J. 7 (2013) 1344–1353. [18] M. Hamady, C. Lozupone, R. Knight, Fast UniFrac: facilitating high-throughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data, ISME J. 4 (2010) 17–27. [19] W. Hua, C.B. Shen, Observations on the effect of ferrous gluconate on children with RRTI, J. Bengbu Med. Coll. 29 (2004) 428–429. [20] E.S. Gollwitzer, B.J. Marsland, Microbiota abnormalities in inflammatory airway diseases - potential for therapy, Pharmacol. Ther. 141 (2014) 32–39. [21] A. Hevia, C. Milani, P. Lopez, C.D. Donado, A. Cuervo, S. Gonzalez, et al., Allergic patients with long-term asthma display low levels of Bifidobacterium adolescentis, PLoS One 11 (2016) e0147809. [22] H. Park, J.W. Shin, S.G. Park, W. Kim, Microbial communities in the upper respiratory tract of patients with asthma and chronic obstructive pulmonary disease, PLoS One 9 (2014) e109710. [23] G. Li Volti, G. Malaponte, V. Bevelacqua, A. Messina, S. Bianca, M.C. Mazzarino, et al., Persistent high plasma levels of interleukins 18 and 4 in children with recurrent infections of the upper respiratory tract, Transplant. Proc. 35 (2003) 2911–2915. [24] W.A. de Steenhuijsen Piters, E.G. Huijskens, A.L. Wyllie, G. Biesbroek, M.R. van den Bergh, R.H. Veenhoven, et al., Dysbiosis of upper respiratory tract microbiota in elderly pneumonia patients, ISME J. 10 (2016) 97–108. [25] Z.J. Yu, Z. Chen, H. Cheng, J.X. Zheng, W.G. Pan, W.Z. Yang, et al., Recurrent linezolid-resistant Enterococcus faecalis infection in a patient with pneumonia, Int. J. Infect. Dis. : IJID Off. Publ. Int. Soc. Infect. Dis. 30 (2015) 49–51. [26] S. Ager, K. Gould, Clinical update on linezolid in the treatment of Gram-positive bacterial infections, Infect. Drug Resist. 5 (2012) 87–102. [27] S. Macfarlane, J.F. Dillon, Microbial biofilms in the human gastrointestinal tract, J. Appl. Microbiol. 102 (2007) 1187–1196. [28] M.M. Gronlund, M. Gueimonde, K. Laitinen, G. Kociubinski, T. Gronroos, S. Salminen, et al., Maternal breast-milk and intestinal bifidobacteria guide the compositional development of the Bifidobacterium microbiota in infants at risk of allergic disease, Clin. Exp. Allergy : J. Br. Soc. Allergy Clin. Immunol. 37 (2007) 1764–1772. [29] S.N. Chen, Treatment of central serous chorioretinopathy: might medical treatment play arole? Taiwan J. Ophthalmol. 3 (2013) 1–2. [30] A.T. Dang, B.J. Marsland, Microbes, metabolites, and the gut-lung axis, Mucosal Immunol. 12 (2019) 843–850.
Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.micpath.2019.103709. References [1] A. Lemanska-Perek, B. Polanska, D. Krzyzanowska-Golab, I. Katnik-Prastowska, Occurrence of soluble supra-molecular FN-fibrin complexes in the plasma of children with recurrent respiratory infection, Ann. Clin. Biochem. 52 (2015) 441–447. [2] S. Esposito, G. Marseglia, A. Novelli, M. de Martino, G. Di Mauro, C. Gabiano, et al., [Acute, subacute and recurrent bacterial rhinosinusitis in pediatrics: guidelines of the study group of the Italian society for pediatric infectious diseases (SITIP)], Minerva Pediatr. 59 (2007) 474–475. [3] M.T. van den Aardweg, C.W. Boonacker, M.M. Rovers, A.W. Hoes, A.G. Schilder, Effectiveness of adenoidectomy in children with recurrent upper respiratory tract infections: open randomised controlled trial, BMJ 343 (2011) d5154. [4] J. Pasnik, Vaccines nonspecific - immunostimulation in patients with recurrent respiratory infections, Otolaryngol. Pol. 70 (2016) 31–39. [5] D.A. van Kessel, T.W. Hoffman, H. van Velzen-Blad, P. Zanen, J.C. Grutters, G.T. Rijkers, Long-term clinical outcome of antibody replacement therapy in humoral immunodeficient adults with respiratory tract infections, EBioMedicine 18 (2017) 254–260. [6] S. Mao, A. Zhang, S. Huang, Meta-analysis of Zn, Cu and Fe in the hair of Chinese children with recurrent respiratory tract infection, Scand. J. Clin. Lab. Investig. 74 (2014) 561–567. [7] X. Jiang, L. Sun, B. Wang, X. Yang, L. Shang, Y. Zhang, Health-related quality of life among children with recurrent respiratory tract infections in Xi'an, China, PLoS One 8 (2013) e56945. [8] C.Y. Lu, Y.H. Ni, Gut microbiota and the development of pediatric diseases, J. Gastroenterol. 50 (2015) 720–726. [9] J.T. Ho, G.C. Chan, J.C. Li, Systemic effects of gut microbiota and its relationship with disease and modulation, BMC Immunol. 16 (2015) 21.
7