Identification of Nucleotide Patterns in Meconium Microbiota to Improve the Management of Meconium Aspiration Syndrome in Clinical Practice

For many years, airways, lungs and meconium, which is a thick green substance and accumulates in the baby’s intestines during pregnancy, have been assumed to be sterile. However, this assumption has been revised by the development of microbiota analysis, which is used for the identification of microorganisms as a new promising technique; it has been found that various microorganisms exist in the lungs and in the meconium fluid. Meconium Aspiration Syndrome (MAS) as a common neonatal problem can give rise to the development of lung infections in newborns and even cases leading to death are encountered. Various strategies regarding antibiotic therapy have been developed against the risk of infection, but positive outcome could not be obtained for each case. In this study, 16S rRNA gene sequences of bacteria existing in meconium microbiota and infected lung microbiota were subjected to comparative sequence analysis. Our aim was to identify positions of nucleotide patterns between the hypervariable regions of the bacterial 16S rRNA gene sequences that could provide similar functions among bacterial groups. Furthermore, similarity analysis was conducted to identify molecular signatures via phylogenetic framework to understand the etiology of the infections after MAS. Interestingly, Bifidobacteria which are used as probiotics were found to be similar to Actynomyces which are known as opportunistic pathogens. Furthermore, Clostridium leptum was associated with pulmonary inflammation for the first time. This study proposes the usage of microbiota analysis to improve the MAS management in clinical practice. *Correspondence E-mail: seyma.is@tau.edu.tr

For many years, airways, lungs and meconium, which is a thick green substance and accumulates in the baby's intestines during pregnancy, have been assumed to be sterile. However, this assumption has been revised by the development of microbiota analysis, which is used for the identification of microorganisms as a new promising technique; it has been found that various microorganisms exist in the lungs and in the meconium fluid. Meconium Aspiration Syndrome (MAS) as a common neonatal problem can give rise to the development of lung infections in newborns and even cases leading to death are encountered. Various strategies regarding antibiotic therapy have been developed against the risk of infection, but positive outcome could not be obtained for each case. In this study, 16S rRNA gene sequences of bacteria existing in meconium microbiota and infected lung microbiota were subjected to comparative sequence analysis. Our aim was to identify positions of nucleotide patterns between the hypervariable regions of the bacterial 16S rRNA gene sequences that could provide similar functions among bacterial groups. Furthermore, similarity analysis was conducted to identify molecular signatures via phylogenetic framework to understand the etiology of the infections after MAS. Interestingly, Bifidobacteria which are used as probiotics were found to be similar to Actynomyces which are known as opportunistic pathogens. Furthermore, Clostridium leptum was associated with pulmonary inflammation for the first time. This study proposes the usage of microbiota analysis to improve the MAS management in clinical practice.

Introduction
Newborn microbiota is interesting to many scientists since the colonization of many bacteria starts in this period of human life. Despite the studies of Hymanson andHertz in 1917 andBurrage in 1927, which demonstrated the presence of bacteria in meconium -which has a thick consistency and accumulates during pregnancy in the baby's intestine-of some newborns, it was commonly accepted that the airways, lungs and meconium are sterile (Burrage 1927;Chotirmall and Burke 2015;Hall 1934;Hymanson and Hertz 1917;Jiménez et al. 2008;Snyder 1936). However, recent studies have shown that this acceptance is not true and findings are promising to use microbiota as a biomarker (Dietert and Silbergeld 2015;Huang and Lynch 2011;Shukla et al. 2017;Zemanick et al. 2013). Besides, the first non-culture based microbiota analysis of meconium was -to the best of our knowledge-conducted by Mshvildadze et al. (Mshvildadze et al. 2010). In a study, Macfarlane and Heaf have observed that newborns develop asthmatic symptoms due to the aspiration of meconium, suggesting that aspiration leads to abnormal respiratory functions through long-term effects on the respiratory system (Macfarlane and Heaf 1988). Following meconium aspiration, multiple life-threatening complications such as the development of hypoxia, respiratory and/or metabolic acidosis, pulmonary hypertension, hypoxic-ischemic encephalopathy and pneumothorax can be encountered (Espinheira et al. 2011;Shaikh et al. 2016). Other diseases potentially related to Meconium Aspiration Syndrome (MAS) include necrotizing enterocolitis, late-onset sepsis, autism and autism-like behavior disorders, fetal macrosomia, obesity, food sensitization and allergy, neonatal jaundice. Also autoimmune diseases such as type 1 diabetes, multiple sclerosis or Morbus Crohn were found to be related to MAS (Dobbler et al. 2017;Dong et al. 2018;Koleva et al. 2015;Stewart et al. 2012;Wilczyńska et al. 2019). In addition, the pro-inflammatory components of meconium lead to pneumonia and pulmonary inflammation, which in turn contribute to the development of bronchopulmonary dysplasia (Kopincova and Calkovska 2016;Speer 2003).
Thus, the aspiration of meconium by the baby due to maternal hypertension, maternal diabetes, hypoxia etc. often leads to the development of lung infections which make the prescription of partially non-specific prophylactic antibiosis mandatory (Yurdakök 2011). Antibiotics are used not only for prophylaxis, but also for treatment of MAS. It is difficult to distinguish between MAS and pneumonia, which leads to an increased prescription of antibiotics due to the high susceptibility to infection. However, antibiotherapy in newborns has many concerns since there are many factors that influence the effects of antibiotics including gestational age, birth weight, intrauterine growth restriction, chronological age and, especially, kidney and liver function immaturity (Chirico et al. 2009). Another aspect is that some studies revealed that antibiosis impairs the immune system of the patient and thus leads to bacterial growth (Benoun et al. 2016;Ubeda and Pamer 2012). It should be added that antibiotherapy needs to be optimized concerning dosing, timing and route of administration if the bacterial composition of the microbiota is known due to the fact that optimization is strongly related to the appropriate selection of antibiotics. However, this is often not the case since traditional culturing techniques fail to identify the existence of many important bacterial clades. But principally, Gosalbes et al. detected extraordinarily high expression rates of antibiotic resistance genes in meconiumassociated bacteria which helps us to understand the non-responsiveness to antibiotherapy in some cases and the associated mortality (Gosalbes et al. 2016). The present study is an attempt to reveal the relation of meconium microbiota and lung microbiota in the presence of pneumonia. It is aimed to shed light on the advancement of more specific antibiotherapies or to reduce the usage of antibiotics in some MAS cases, respectively.

Targeting Bacteria and Retrieving 16S rRNA Sequences from Databases
After extensive database and literature search, meconium-and pneumonia-associated bacteria were obtained from the studies of Nagpal et al. and Chamberlain, respectively (Chamberlain 2016;Nagpal et al. 2016). The 16S rRNA gene sequences of these bacteria were retrieved from the National Center for Biotechnology Information (NCBI) Nucleotide Database and the European Bioinformatics Institute (EMBL-EBI) Database. Subsequent analyses were performed with 16 bacteria found in meconium microbiota of vaginallyborn infants and 33 bacteria found in lung microbiota of patients diagnosed with pneumonia. The sequences of some of these bacteria (Streptococcus agalactiae, Nocardia sp. and Serratia sp.) could not be retrieved from a single database entry, thus these sequences were merged from several database entries. BLASTn was performed for these sequences and the similarity rates were ascertained as >90%. All these bacteria and the corresponding accession numbers are given in Supplementary S1. All supplementary tables and data (S1-S10) are available online at http://dx.doi.org/10.17632/w5653pr7yh.1.

Extraction of the Hypervariable Regions
V-Xtractor version 2.1 was used on Linux (Ubuntu 18.04.1 LTS, GNOME 3.28.2, OS Type 64-bit) to extract the hypervariable regions V1-V9 of the bacterial 16S rRNA genes (Hartmann et al. 2010). Necessary to that end was a sensitive method; we preferred HMMER version 3.2.1 which allows biological sequence analysis using profile Hidden Markov Models (Eddy 2011).

Comparative Sequence Analysis
The progressive alignment tool T-Coffee version 11.00.8cbe486 was performed on Linux in order to align multiple sequences. The results of the multiple sequence alignments were then displayed and analysed by using Jalview version 2.10.4b1 (Waterhouse et al. 2009). In this way, global and local similarities of each of the extracted hypervariable regions V1-V9 of the bacterial 16S rRNA sequences were curated (Notredame et al. 2000). The results of each hypervariable region was further analysed and 9 bacteria associated with pneumonia (namely Streptococcus pneumoniae, Streptococcus pyogenes, Streptococcus agalactiae, Staphylococcus aureus, Peptostreptococcus sp., Bacillus anthracis, Nocardia sp., Actynomyces sp. and Neisseria meningitidis) were found to be similar to meconium-associated bacteria.

Results and Discussion
In recent years, many platforms were established, and pipelines were designed to analyse human microbiota. In 2007, the Human Microbiome Project (HMP) was launched which further contributes to our understanding of the human microbiota composition. However, there is no data concerning the lung, pneumonia or meconium available on the website of HMP (https://hmpdacc.org/hmp/); thus, much needs to be investigated in future. Because of the limited microbiome data regarding lung or pneumonia or meconium, respectively, we were forced to use the existing limited data and tried to indicate an approach for the management of MAS with our study with the aim to reveal that there is the pressing need for much more studies on this field. One in every seven pregnancies ends with meconium-stained amniotic fluid and 5% of these infants develop MAS (van Ierland and Beaufort 2009). MAS has a wide range of severity and even in some cases, it is the leading cause of morbidity and mortalitydepending on the conditions and complications which develop after parturition. Although many studies show that antibiotic use in infants cause many problems, antibiotic administration is one of the general applications to counteract this syndrome in newborn follow-up clinics (Miller et al. 2018;Rogawski et al. 2017).
In this study, the role of meconium microbiota in the development of lung infections was investigated by using the new data emerged from the microbiota analyses. Identifying molecular signatures via phylogenetic framework helps us to understand the etiology of the infection after MAS. Clinical usage of antibiotics in newborns could be redesigned by using these molecular signatures. Staphylococcus and Streptococcus exist both in meconium and in infected lungs according to the existing data. Abundance of these bacterial clades may indicate a target for diagnosis. Quantitative analyses on these bacteria of newborn meconium immediately after birth may help to enhance rational use of antibiotics in newborns. The results of the comparative sequence analysis propose several significant similar nucleotide patterns which may indicate useful targets for diagnostic assays. To analyze those similarities, phylogenetic trees were constructed with the sequences of each hypervariable region V1-V9 altogether of the corresponding bacteria (Figure 1). Positions of nucleotide patterns varying between the respective hypervariable regions of the 16S rRNA genes of bacteria found in meconium and bacteria leading to lung infections have been analysed and are shown in Figure 2. Peptostreptococcus that shows pathogenic activity for pneumonia under traumatic conditions has been found 95% similar to Clostridium leptum in meconium according to the phylogenetic tree generated by using the hypervariable regions V1-V9  all merged together consecutively (Figure 3). Although C. leptum's role in inflammatory bowel disease and in the development of asthma was shown in some studies, it was never associated with pulmonary inflammation (Huang et al. 2015;Kabeerdoss et al. 2013). Another detected similarity was between Actynomyces and Bifidobacterium which is a bit surprising at first glance (Figure 1g  and 1i). Besides being part of the human microbiota, Actynomyces are known for causing several diseases, whereas Bifidobacterium species are known to be beneficial and used as probiotics. However, some cases were reported in which Bifidobacterium species caused urinary, pleuropulmonary, obstetric and gynecologic infections, dental caries and even bacteremia (Bertelli et al. 2015;Weber et al. 2015). Moreover, Bifidobacterium species may possess multiple genes associated with virulence factors, harmful metabolites and antibiotic resistance, whereas one of these antibiotic resistance genes is possibly transferable (Wei et al. 2012). In order to confirm the similarity of Actynomyces and Bifidobacterium, we performed additional comparative sequence analysis with the appropriate hypervariable regions of both bacteria and ascertained that there was a significant similarity: These bacteria's hypervariable regions V9 and V7 were the most similar sequences with similarity values 91% and 90%, respectively (Table 1). Thus, we propose that the pathogenicity of Bifidobacterium may be ascribed particularly to V9 and V7, respectively, since hypervariable regions may play an important role in the pathogenicity of bacteria. This pathogenicity could be in the form of antibiotic resistance, since antibiotic-binding sites are usually located within important structures of the bacterial 16S rRNA in order to achieve blockage of protein synthesis on the ribosome. As it can be obtained from Table 1, V1-V3 are less similar when compared with the other hypervariable regions. We concluded that V1-V3 could be more suitable for the differentiation of bacterial species to the genus level. In line with our conclusion, Chakravorty et al. reported that V2 and V3 were the most suitable regions for bacterial differentiation (Chakravorty et al. 2007). Besides, the hypervariable regions V2 and V3 are also more appropriate for bacterial differentiation due to the high mutation rates in these regions (Bukin et al. 2019). This study may be the first regarding the molecular evidence of the similarity between Actynomyces and Bifidobacterium. In fact, it would be appropriate if these analyses will be repeated with a larger data set and the results confirmed in future. Besides, further studies regarding the function of 16S rRNA hypervariable regions in the pathogenicity, but especially in the antibiotic resistance of bacteria should be performed in order to optimize or even to eliminate antibiotic treatment after MAS.

Conclusion
The successful treatment of MAS depends on the choice of the appropriate antibiotic, which may only be possible after a microbiota analysis. Even if it is not possible to carry out individual microbiota analyses in the immediate future, at least one largescale study can improve the prospects for successful treatment of MAS. In addition, identifying which hypervariable regions of the 16S rRNA gene are responsible for the pathogenicity, especially the resistance to antibiotics, of bacteria is also of great importance for the optimization of antibiotherapy.

Acknowledgement
This research did not receive any specific grant from funding agencies in the public, commercial, or notfor-profit sectors.

Conflict of interest
Authors declare no conflict of interest.