SHI Da Wei,WANG Dong Mei,NING Li Hua,LI Jing,DONG Yan,ZHANG Zhi Kun,DOU Hai Wei,WAN Rui Jie,JIA Chun Mei,#,and XIN De LI,#

1.Department of Pediatrics, Beijing Friendship Hospital, Capital Medical University, Beijing 100050, China;2.Department of Pediatrics, Baotou Fourth Hospital (Baotou Children’s Hospital), Baotou 014030, Inner Mongolia Autonomous Region, China; 3. Department of Pediatrics, Beijing Chang Ping Hospital of Integrated Chinese and Western Medicine, Beijing 100096, China; 4. Tropical Medicine Research Institute, Beijing Friendship Hospital,Capital Medical University, Beijing 100050, China

Abstract Objective We investigated changes in the intestinal flora of children with Mycoplasma pneumoniae pneumonia (MPP).Methods Between September 2019 and November 2019, stool samples from 14 children with MPP from The Fourth Hospital of Baotou city, Inner Mongolia Autonomous Region, were collected and divided into general treatment (AF) and probiotic (AFY) groups, according to the treatment of“combined Bifidobacterium, Lactobacillus, Enterococcus, and Bacillus cereus tablets live”. Highthroughput 16S rDNA sequencing was used to identify intestinal flora.Results Intestinal flora abundance and diversity in children with MPP were decreased. Both Shannon and Simpson indices were lower in the AF group when compared with healthy controls (P ＜ 0.05). When compared with healthy controls, the proportion of Enterorhabdus was lower in the AF group, while the proportion of Lachnoclostridium was higher (P ＜ 0.05). The proportion of Bifidobacteria and Akkermansia was lower in the AFY group but Enterococcus, Lachnoclostridium, Roseburia, and Erysipelatoclostridium proportions were higher. The proportion of Escherichia coli-Shigella in the AFY group after treatment was decreased (P ＜ 0.05).Conclusions The intestinal flora of children with MPP is disturbed, manifested as decreased abundance and diversity, and decreased Bifidobacteria. Our probiotic mixture partly improved intestinal flora disorders.

Key words: Intestinal flora; Mycoplasma pneumoniae pneumonia; Probiotics; Respiratory tract infection

INTRODUCTION

Mycoplasmapneumoniae(M.pneumoniae,MP) is a pathogen that causes respiratory tract infections in children and accounts for 10%-40% of communityacquired pneumonia (CAP) in hospitalized children[1,2]. Clinical manifestations can be mild or severe; severe cases cause necrotizing pneumonia and even dangerous extrapulmonary complications.In recent years, many studies have shown that changes in intestinal microecology have important roles in pulmonary infection[3]. Maladjustment of the intestinal flora aggravates respiratory infections caused by pathogens such as influenza viruses,Staphylococcus aureus,Klebsiella pneumoniae, andStreptococcus pneumoniae[3]. Mouse studies have shown that antibiotic-induced intestinal dysbacteriosis aggravates MP respiratory tract infections and suggested that intestinal microflora has a regulatory effect on respiratory tract infections[4]. In this study, stool samples from children with MP pneumonia (MPP) were examined by 16S rDNA sequencing. Probiotics were administered to investigate changes in intestinal microflora in these children, and identify possible preliminary roles during MP infection.

MATERIALS AND METHODS

Patients

Between September 2019 and November 2019,14 children with MPP were admitted to Baotou Fourth Hospital (Baotou Children’s Hospital) in Inner Mongolia Autonomous Region, China and divided into probiotic group (AFY,n= 8) and general treatment group (AF,n= 6). Simultaneously, nine healthy children from Beijing Chang Ping Hospital of Integrated Chinese and Western Medicine in China were included as a healthy control group (KHJB).

MPP was diagnosed by consensus using the Diagnosis and Treatment of Children’s Mycoplasma Pneumoniae Pneumonia 2015 edition: (i) acute respiratory infection symptoms (fever, cough or wheezing) upon physical examination, and chest imaging with infiltrates; (ii) the MP infection was confirmed using serological tests (MP-IgM-positive(Diagnostic kit for Antibody toMycoplasma pneumoniae, HAITIANLANBO.BIO-TECH.CO., Ltd,Fujian, China) and an antibody titer ≥ 1:160 or a fourfold or greater increase in titer (SERODIA? -MYCO II,FUJIREBIO INC. Tokyo, Japan) and MP nucleic acid detection in nasopharyngeal aspirates[1,5].

Inclusion criteria: (i) patients were 3-14 years old; (ii) patients were diagnosed with MPP; and (iii)the disease course was ≤ 7 days. Exclusion criteria:(i) patients with measles, whooping cough, chicken pox, or other infectious diseases; (ii) severely malnourished children; (iii) patients with underlying diseases such as asthma, chronic heart and lung disease, rheumatic disease, kidney disease, or immunodeficiency; (iv) children with other pathogenic infections (bacterial, fungal, and/or viral); (v) children treated with antibiotics,hormones, intestinal microbial preparations, or other immunological preparations in the previous month; and (vi) children who did not cooperate with sampling regimens or whose parents refused to participate.

Children in both groups received (i) azithromycin(Pfizer Pharmaceutical Co., Ltd，New York, USA),(ii) Pediatric Feirekechuan Oral Liquid (Heilongjiang Sunflower Pharmaceutical Co., Ltd, Heilongjiang,China), and (iii) other symptomatic treatments if necessary. Children in the probiotics group also received combinedBifidobacterium, Lactobacillus,Enterococcus,andBacillus cereuslive tablets(Hangzhou Grand Biologic Pharmaceutical INC,Zhejiang, China). Each tablet contained ＞ 1.0 × 106colony forming unit (CFU) ofBifidobacterium,Lactobacillus, and Enterococcusand ＞ 1 × 105CFU/gB. cereus.Children treated with probiotics received two tablets three times a day.

Study protocols were approved by the Ethics Committee of the Affiliated Beijing Friendship Hospital at Capital Medical University (Beijing,China), and methods were conducted in accordance with approved guidelines (Number: 2019-P2-206-02). Written informed consent was obtained from the parents or guardians of participants prior to enrollment.

Specimens

Two stool specimens and two throat swabs were collected from AF and AFY groups, the specimens were collected on the first day of treatment (AF_A,AFY_A) and 7 ± 1 days after treatment (AF_B,AFY_B). One stool specimen and one throat swab were collected from KHJB group.

After defecation, samples were comprehensively collected from stools using a sampling spoon, quickly placed in a specimen box, and frozen at -80 °C.Specimens were tested at the Institute of Microbiology of the Chinese Academy of Sciences.Fluorescence quantitative polymerase chain reaction(PCR) was used to amplify23S ribosomal RNA (23S rRNA)in MP throat swab specimens[6].

DNA Extraction and 16S rDNA Amplicon Sequencing

DNA was extracted from fecal samples (0.5 g)using a QIAamp PowerFecal DNA Kit (QIAGEN,Germany) according to manufacturer’s protocols.The V3-V4 region of bacterial16S rDNAwas then amplified using the primers: 1stF3:CCTACGGGNBGCASCAG and 1stR4:GACTACNVGGGTATCTAATCC (Beijing Liuhe Bgi Co.,Ltd. Beijing, China). PCR was performed in a 25 μL mixture containing 5 μL 5 × GC buffer, 0.5 μL KAPA dNTP mix, 0.5 μL KAPA HiFi HotStart DNA polymerase, (Roche, USA) 0.5 μL each primer(10 pmol/L), and 50-100 ng template DNA. PCR cycling parameters were: 95 °C for 3 min, followed by 25 cycles at 95 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s, and a final extension at 72 °C for 5 min. We then used AMPure XP (Beckman Coulter, USA) beads to purify amplicons from free primers and primerdimer species. A second specific linker amplification step was next used to construct a library that fulfilled Illumina requirements. Each linker contained a unique eight base barcode sequence. PCR was performed in a 25 μL mixture: 5 μL 5× GC buffer,0.75 μL KAPA dNTP mix, 0.5 μL KAPA HiFi HotStart DNA polymerase, 1.5 μL each primer (10 pmol/L),and 5 μL purified product. PCR cycling parameters:95 °C for 3 min, followed by eight cycles at 95 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s, and a final extension at 72 °C for 5 min. Amplicons were purified as described to clean up the final library before quantification. Finally, purified amplicons were pooled in equimolar quantities and paired-end sequenced (2 × 250) on an Illumina MiSeq platform(Illumina, USA) according to standard protocols.

Bioinformatics and Statistical Data Analyses

Fast Length Adjustment of Short Reads was used to merge paired-end reads from next-generation sequencing[7]. Low quality reads were filtered out using fastq_quality_filter (-p 90 -q 25 -Q33) in the FASTX Toolkit 0.0.14, and chimera reads were removed using USEARCH 64 bit (Ver. 8.0.1517). The number of reads per sample was normalized using random subtraction based on the smallest sample size. Operational taxonomic units (OTUs) were aligned using the UCLUST algorithm with a 97%identity, and taxonomically classified using the SILVA 16S rRNA database (Ver. 128). Both α and β diversities were generated in the Quantitative Insights Into Microbial Ecology database and calculated based on weighted and unweighted Unifrac distance matrices[8]. The α diversity included an index of observed species, Chao1 estimator,Shannon and Simpson; the β diversity included principal coordinate analysis (PCoA)、Adonis analysis and UPGMA clustering tree analysis. We used the linear discriminant analysis effect size (LEfSe)method to identify species with statistically significant abundance between groups[9].

Proportions of bacteria in AF and AFY groups were compared in each patient before and after treatment, the bacteria was considered as increased/decreased with a ratio higher/lower than 1.2/0.83.When the bacteria was changed in more than half of the patient within groups, it was considered as being influenced by the treatment.

Statistical Methods

SPSS 20.0 software (IBM Corp, Armonk, NY, USA)was used for statistical data analysis.P＜ 0.05 was considered statistically significant.

RESULTS

Gener al Information

In this study, 23 children were enrolled, aged 4 years and 8 months to 9 years and 5 months, of which one child was 4 years old and 8 months old,and the rest were ＞ 5 years old. The mean age was 5 years and 8 months. No statistical differences in age were observed between AFY group, AF group, and KHJB group (P= 0.361).

MP Diagnosis

Antibody titers were ≥ 1:160 in the general treatment and probiotics groups. MP nucleic acid levels in the first throat swabs were all positive.

16S rDNA Sequencing Results

After high-throughput sequencing of fecal samples from probiotic and healthy control groups,we generated 10,268,989 valid sequences in total.The average number of sequences/sample was 277,540.24.

Intestinal Flora Dilution Curve Analysis

Some sequences were randomly selected from our data, the number of species represented by sequences counted, and a dilution curve constructed using sequence and species numbers. As shown(Figure 1), as the sample volume increased, sample flora OTUs tended to be stable and the curve tended to be flat. This indicated that sequencing data were sufficient and the sequencing depth of the sample had been reached.

Intestinal Microflora α-diversity Analysis

OTUs were higher in KHJB group than in AF and AFY groups (P＜ 0.05). The Chao1 index was higher in the KHJB group than in the AFY group (P＜ 0.05); no significant differences were observed between KHJB and AF groups. The Shannon index was higher in the KHJB group than in the AF group (P＜ 0.05); no significant differences were observed between KHJB and AFY groups. The Simpson index was higher in the KHJB group than in the AF group (P＜ 0.05); no significant differences were observed between KHJB and AFY groups.

The mean number of OTUs in stools was lower in the AF group than in the KHJB group, but differences were not statistically significant. The average Chao1 index of the KHJB group was higher than in the stool sample AF_A, which was higher than in the second stool sample AF_B; however, differences were not statistically significant. The average Shannon index was higher in the KHJB group than in the AF group,but differences were not statistically significant. The average Simpson index was higher in the KHJB group than AF_B which was higher than AF_A; however,differences were not statistically significant.

OTUs in the AFY group were lower than in the KHJB group (P＜ 0.05). The Chao1 index was higher in the KHJB group than in the AFY group (P＜ 0.05). The average Chao1 index of the stool sample AFY_B was lower than in AFY_A, but differences were not statistically significant. The average Shannon index was higher in the KHNB group than in the stool sample AFY_B, which was higher than in AFY_A;however, differences were not statistically significant.The average Simpson index of the KHJB group was higher than in the stool sample AFY_B, which was higher than in AFY_A; however, differences were not statistically significant (Figure 2).

Figure 1. Intestinal flora dilution curve. AF,general treatment group; AFY, probiotic group;and KHJB, healthy control group.

Intestinal Microbiota β-diversity Analysis

PCoA showed that the intragroup differences of bacterial community structures increased in AF and AFY groups comparing to KHJB group, while structural differences between AF and AFY groups were significant (Figure 3). Adonis analysis showed no significant differences in bacterial community structures between AF, AFY, and KHJB groups (R2=0.097,P= 0.087).

UPGMA clustering tree analysis based on the Weighted UniFrac distance showed that the bacterial community structures in AFY group are tending to be clustered together after treatment; While similarity in bacterial community structures were observed pre- and post-treatment in AF group. These results suggested that probiotics effectively stabilized intestinal flora (Figure 4).

Bacterial Analysis at the Phylum Level

In total, ten phyla were detected across all samples:Firmicutes, Proteobacteria, Bacteroides,Actinobacteria, Verrucomicrobia, Patescibacteria,Tenericutes,Euryarchaeota, Fusobacteria,andEpsilonbacteraeota(Figures 5 and 6).

Figure 2. Comparison of α-diversity among groups. AF, general treatment group; AF_A, the first stool sample; AF_B, the second stool sample; AFY, probiotic group; AFY_A, the first stool sample; AFY_B, the second stool sample; KHJB, healthy control group. The intragroup differences of Shannon and Simpson index increased after MP infection and reduced after treatment in AFY group.

Figure 3. Principal coordinate analysis (PCoA) plots of individual fecal microbiota based on weighted UniFrac (A) and unweighted UniFrac distances (B). AF, general treatment group; AFY, probiotic group;KHJB, healthy control group.

Figure 4. UPGMA clustering tree analysis. The bacterial community structures in AFY group are tending to be clustered together after treatment; While similarity in bacterial community structures were observed pre- and post-treatment in AF group.

Figure 5. The top five bacteria at phylum level.

Figure 6. The top 6-10 bacteria at phylum level.

Among groups, the proportion ofVerrucomicrobiaandEuryarchaeotawas lower in the AFY group than in the AF group, and the proportion ofFusobacteriawas higher in the AFY group than in the AF group (P＜ 0.05). When compared with the KHJB group, the proportion ofActinobacteriaandVerrucomicrobiadecreased, while the proportion ofBacteroidetesandFusobacteriaincreased (P＜ 0.05).

The proportion ofActinomycetesin stool sample AF_B was lower in the AF group than in the KHJB group (P＜ 0.05). In the AFY group, the proportion ofEpsilonbacteraeotawas lower in the stool sample AFY_B than in AFY_A (P＜ 0.05). The proportion ofActinomycetesandFirmicutesin the stool specimen AFY_B was lower than in samples from the KHJB group (P＜ 0.05).

Bacterial Analysis at the Genus Level

The proportion ofEnterorhabduswas lower in the AF group than in the KHJB group, while the proportion ofLachnoclostridiumwas higher in the AF group than in the KHJB group (P＜ 0.05). The proportion ofAkkermansiawas lower in the AFY group than in the AF group, and the proportion ofBifidobacteriaandAkkermansiawas lower in the AFY than in the KHJB group. The proportion ofEnterococcus,Lachnoclostridium,Clostridium erysipelas,andErysipelatoclostridiumincreased when compared with levels in KHJB group (P＜ 0.05).

The proportion ofFaecalibacteriumandEubacteriumhalliiin stool samples AF_A was lower group than in the KHJB group (P＜ 0.05). The ratio ofBifidobacteriaandRomboutsiain stool samples AF_B was lower than in the KHJB group (P＜ 0.05). In the AFY group, the proportion ofEscherichia-ShigellaandButyrivibrioin stool samples AFY_B was lower than in AFY_A. The proportion ofBifidobacteriain stool samples AFY_A was lower than in KHJB group.The proportion ofE.coli-ShigellaandSubdoligranulumin stool samples AFY_B was lower than in KHJB group. When compared with the KHJB group, the proportion ofEnterococcus,Lachnoclostridium,Roseburia, andErysipelatoclostridiumwas lower in AFY group (P＜0.05, Figures 7 and 8).

In the probiotic group (AFY), after receiving“combined liveBifidobacterium, Lactobacillus,Enterococcus,andB. cereustablets”, the proportion ofBifidobacterium,Lactobacillus,andEnterococcusin stool samples AFY_B was higher than that in AFY_A, although differences were not statistically significant (P= 0.454,P= 0.113, andP= 0.463,respectively), but the proportion ofEnterococcusin stool specimens (AFY_B) after probiotic treatment increased significantly when compared with healthy controls (P＜ 0.05). The proportion ofBifidobacteriumin pre-treatment stool specimens(AFY_A) was lower than in healthy controls (P＜0.05), but no differences inBifidobacteriumratios in stool specimens after probiotics treatment were noted when compared with healthy controls(Figure 9).

To analyze the influence of different treatment to intestinal flora, the changes in proportions of bacteria in each individual were analyzed at genus level. Fourteen bacterial species in the general treatment group increased and 21 species decreased after treatment; 13 species in the probiotic treatment group were increased and 21 species decreased (Table 1).

DISCUSSION

The intestinal microecosystem is composed of billions of microorganisms which maintain a dynamic physiological balance and promote host immunity,metabolism, energy balance, and neural development[10,11]. Many studies have reported that the intestinal flora significantly alters after respiratory tract infections[12-14]. A study of 11 children (4-5 years) with CAP showed that gut the microbiome had increased forEscherichia/Shigella,Bifidobacterium, Streptococcus,andPsychrobacterabundance and decreased forFaecalibacterium,Bacteroides, Lachnospiraceae,andRuminococcusabundance when compared with matched healthy controls[12]. Children with pulmonary tuberculosis had reduced intestinal microbial diversity, with an enrichment of pro-inflammatoryPrevotellaand the opportunistic pathogenEnterococcus, and decreasedRuminococcaceae,Bifidobacteriaceae, andFaecalibacteriumprausnitziiprobiotics when compared with healthy peers[13]. Mice withstreptococcus pneumoniae-induced pneumonia had lower gut bacterial community diversity (lower phylogenetic diversity and Shannon indices)[14]. The number of intestinal probiotics in Avian influenza A(H7N9)-infected patients decreased while pathogens increased, thereby inducing intestinal injury and mucosal immune dysfunction[15]. Previous studies identified intestinal microbiota as a protective mediator during pneumococcal pneumonia, which enhanced primary alveolar macrophage function[16].In our study, OTU’s, and Shannon and Simpson indices were lower in the general treatment group when compared with the healthy control group (P＜0.05). OTU’s and the Chao1 index in the probiotic group were lower than in the healthy control group(P＜ 0.05) and suggested that intestinal flora abundance and diversity in children with MPP were lower, and that intestinal flora community structures had changed when compared with healthy children,suggesting intestinal flora disorders occurred in children with MPP. Due to their drug actions and low side effects, macrolide antibiotics are the first treatment choice for children with MPP[1]. Previous studies reported that macrolides reduced intestinal flora richness in these children, significantly reducing the proportion ofBifidobacteriaandLactobacillusand significantly increasing the proportion ofProteobacteria,such asE. coli[17]and suggesting these antibiotics affected the intestinal microecology of children with MPP. Therefore, intestinal flora disorders in these children are not only caused by the condition but also by the therapy itself.

Figure 7. The top 30 different operational taxonomic units (OTUs) in bacterial communities. The graph compares the average proportion of different bacteria in each group and indicates overall changes in bacterial communities. The y-axis represents percentages. On the whole, bacteria were sorted according to the proportion and size of groups. In this figure, they were sorted according to bacteria in the control group.

Healthy intestinal flora exhibits a high diversity,whereFirmicutesandBacteroidetesare dominant flora. Previous studies reported that the proportion ofFirmicutesin the stool of children with sepsis, and patients with severe pneumonia, was significantly lower in normal control groups[18,19]. In our study, the highest relative abundant bacteria among groups wereFirmicutes, but differences were not statistically significant. We hypothesized this was related to the small number of specimens and the condition of the children; therefore, further research on the fecal and intestinal flora of children with severe MPP is warranted.Actinobacteriais one of the main bacteria living in the intestines of healthy humans, and together withBacteroidetesandFusobacteria, they cover most obligate anaerobe bacteria and have a dominant position in the hypoxic environment of the colon, with key physiological roles[20,21]. We showed that at the phylum level, the proportion ofActinomycetesin second stool samples in the general treatment group was lower than in healthy controls. The proportion ofBifidobacteriain second stool samples of the generic treatment group was lower than in healthy controls and suggested thatActinomyceteswere more strongly affected over a prolonged disease course. After the probiotic group received probiotics, differences within the group decreased and the Simpson diversity also decreased, but the proportion ofActinobacteria,Bifidobacteria, andAkkermansiadecreased when compared with the healthy controls and suggested that probiotics improved but did not completely recover the intestinal flora disorder in children with MPP in the short term.Bifidobacteriumis the dominant symbiotic bacterium in the colon microbiome, accounting for 25% of culturable fecal bacteria in adults and 80% in infants, and is widely clinically studied[22]. Studies have shown that infantBifidobacteria,alone or in combination with other bacteria, specifically relieve different irritable bowel syndrome symptoms and reduce the incidence and severity of necrotizing enterocolitis in very-lowbirth-weight infants[23,24]. Studies have also reported that short-term use ofClostridium butyricumplusBifidobacteriuminfantile preparations effectively prevented antibiotic-associated diarrhea in hospitalized children receiving azithromycin treatment for MPP, and that the probiotic mixture partially reconstructed intestinal microbiota and restored bacterial diversity[25]. Probiotics could increase the proportion of beneficial bacteria such asBifidobacteriaandFaecalibacterium prausnitziiin children after adjuvant therapy and reduce opportunistic pathogens such asEnterococcus[26].Short-chain fatty acids such as butyric acid in feces are significantly increased and appear to stabilize blood glucose levels, effectively improving immune indicators, and reducing the chance of secondary infections in children[26,27]. In our study, the proportion ofEscherichia-Shigellain stool specimens in the probiotic group decreased after treatment,and the proportion ofEscherichia-ShigellaandSubdoligranulumdecreased when compared with healthy controls.Enterococcus,Lachnoclostridium,Roseburia, andErysipelatoclostridiumwere increased when compared with healthy controls.Roseburiaproduces butyric acid and is a key bacteria involved in dietary fiber xylan degradation in the human intestinal tract. Our results suggest that probiotics improve beneficial bacteria production in the intestinal flora of children with MPP and reduce pathogens such asEscherichia-Shigella.

Figure 9. Bifidobacterium, Lactobacillus, and Enterococcus in probiotic (AFY) and healthy control groups (KHJB).

Figure 8. The top 31-60 different operational taxonomic units (OTUs) in bacterial communities. The graph compares the average proportion of different bacteria in each group and indicates overall changes in bacterial communities. The y-axis represents percentages.

Study Strengths and Limitations

This was a preliminary study of intestinal flora in children with MPP; however, some limitations were identified. Firstly, we only observed intestinal flora in patients with ordinary MPP, numbers were small,and the sampling area was limited. Secondly, we did not investigate the influence of cytokines and intestinal function on intestinal flora differences in children with MPP. Therefore, our report is preliminary in nature; further studies are required to compare changes in stool flora in children with MPP in different conditions, and explore relationships between MP infection and intestinal flora, andpossible underlying mechanisms should be explored.

Table 1. The bacteria of inter-individual flora changes after treatment in general treatment (AF) and probiotic treatment groups (AFY)

In conclusion, intestinal flora disorders occur in children with MPP. Such changes were manifested by decreased flora abundance and diversity and changes in community structures. The proportion of beneficialBifidobacteriawas decreased, while the proportion of pathogenicEnterococcusandC. erysipeliswas increased. However, probiotic supplementation improved intestinal flora. In future studies, fecal flora in children with MPP must be investigated to elucidate possible immune mechanisms between intestinal flora and the condition. Such studies could identify significant therapies for treating sick children with MP infections.

CONFLICT OF INTEREST

The authors declare none.

Received: March 31, 2022;

Accepted: June 9, 2022