探索斑腿樹蛙腸道菌以及其網絡關係

全文

(1)國立臺灣師範大學生命科學系中央研究院生物多樣性國際研究生學程 Department of Life Science, National Normal University Taiwan International Graduate Program on Biodiversity, Academia Sinica. 博士論文 Doctoral Dissertation. 探索斑腿樹蛙腸道菌以及其網絡關係 Exploring gut microbiota of Polypedates megacephalus and the inference of its microbial interactions. 研究生：翁正軒 Student: Francis Cheng-Hsuan Weng 指導教授：王達益博士 Advisor: Daryi Wang, PhD. 中華民國 108 年 2 月 February 2019.

(2) Abstract The concerted activity of intestinal microbes is crucial to the health and development of their host organisms. Studies have suggested that microbial assemblages in the intestine of animals are engines of globally important host physiological processes between hibernating and non-hibernating states. The advances in Next Generation Sequencing (NGS) technologies facilitate our understanding of gut microbiota with high resolutions in diversity and metabolic functioning between hibernating and nonhibernating seasons. Polypedates megacephalus is an invasive species in Taiwan since 2006. The approach of habitat usage and population dispersal of this invasive treefrogs across seasons have suggested a rapid expansion across counties in Taiwan. However, it still lacks an effective solution to control the expansion of invasive P. megacephalus. Due to the reciprocal interactions between gut microbiota and host physiology, I attempt to explore gut microbiota of P. megacephalus, decipher microbial interactions to understand the potential mechanisms of microbial ecosystem, and further manipulate host response by modulating gut microbiota according to the guidance by computational network analysis. This study not only delineated seasonal changes of gut microbiota in composition and metabolic functioning but demonstrated the potentials of computational network inference toward practical applications on animal systems. The compositional and predicted functional changes of gut microbiota across nonhibernating and artificial hibernating seasons were identified based on 16S rRNA amplicon analysis. The abundance profile and predicted functions of microbial community significantly change between artificial hibernating (AH) and nonhibernating (NH) treefrogs. Artificial hibernation significantly reduces microbial diversity and the level of Firmicutes and increases the level of Proteobacteria in the treefrog gut microbiota. In addition, AH treefrogs harbor core taxonomic units that are 2.

(3) rarely abundant in NH treefrogs. Moreover, artificial hibernation significantly increased relative abundance of red-leg syndrome-related genera such as Citrobacter and Aeromonas. Functional predictions via PICRUSt and Tax4Fun suggested that artificial hibernation has effects on most pathways including metabolism and signal transduction. These results suggest that artificial hibernation restructure gut microbiota in treefrogs and significantly reduce microbial complexity of gut microbiome. The use of computational methods to decipher microbial interactions have been applied on microbiome data in a time-series fashion. A time-series microbiome data could monitor the population changes of each bacterium in the community over time. Due to the adjustable gut microbiome complexity of hibernating animals, the growth microbiome time-series (GMT) dataset is proposed to apply on the computational network inference methods. Among varieties of network inference tools, regressionbased network model is selected and utilized due to its better performance tested by using in silico dataset. Lotka-Volterra models, also known as predator–prey equations, are the most currently used regression-based method, and predict both dynamics of microbial communities and how communities are structured and sustained. The interaction network of gut microbiota at the genus level in the treefrog was constructed using Metagenomic Microbial Interaction Simulator (MetaMIS) package. The interaction network contained 1,568 commensal, 1,737 amensal, 3,777 mutual, and 3,232 competitive relationships, e.g., Lactococcus garvieae has a commensal relationship with Corynebacterium variabile. To validate the interacting relationships, I took advantage of probiotic system to evaluate the responses of gut microbiota to the probiotic trials. The trials involved different groups including single strain (L. garvieae, C. variabile, and Bacillus coagulans, respectively) and a combination of L. garvieae, C. variabile, and B. coagulans, because of the 3.

(4) cooperative relationship among their respective genera identified in the interaction network. After a two-week trial, the combination of cooperative microbes yielded significantly higher probiotic concentrations than single strains, and the immune response (interleukin-10 expression) also significantly changed in a manner consistent with improved probiotic effects. Keywords: network, gut microbiota, artificial hibernation, probiotics, Polypedates megacephalus.. 4.

(5) Table of Contents Introduction............................................................................................. 13. 1 1.1. 1.1.1. Strong links between animal gut microbiome and its physiology and behaviour .......13 The amount and importance of microorganisms in the gastrointestinal tract (gut. microbiota) ..........................................................................................................................13 1.1.2. Phylogenetic diversity and structural assembly of gut microbiota correlates to host. physiology and behaviour....................................................................................................13 1.2. Seasonal behavioural observations of Polypedates megacephalus .............................15. 1.2.1. Population distribution of P. megacephalus in Taiwan ..........................................15. 1.2.2. Seasonal dispersal and habitat usage of P. megacephalus in Taiwan ...................15. 1.2.3. Invasive P. megacephalus, potential threat, and current solutions for invasion ....16. 1.3. Seasonal alterations on gut microbiota in hibernators ................................................17. 1.3.1. Compositional changes of gut microbiota during hibernation ...............................17. 1.3.2. Seasonal shifts of dominant bacterial populations in hibernators identified by. conventional methods .........................................................................................................18 1.3.3. Seasonal restructure of gut bacterial assembly in hibernators characterized by. advanced sequencing technology........................................................................................19 1.4. Network inference provides insight into the mechanisms of gut bacterial assembly ..21. 1.4.1. The potentials of top-down approach on microbial community ............................21. 1.4.2. Deciphering microbial interaction provides insights into causal relationships .......22. 1.4.3. Network inference theory .......................................................................................24. 1.4.4. From static network to dynamic network: time-series data ...................................26. 1.4.5. Current pitfalls of network inference ......................................................................28. 1.5. Well-developed intestinal probiotic system ................................................................29. 1.5.1. Lactic acid bacteria (LAB) – the wild used probiotics..............................................29. 1.5.2. The interactions between probiotic strains and other bacteria..............................30. 5.

(6) 1.5.3. Host immune responses by probiotics ....................................................................31 Probiotics used in real applications.............................................................................32. 1.6 1.6.1. Probiotic strains used in raniculture .......................................................................32. 1.6.2. The growing process of probiotics development ....................................................33 From network theory to practical probiotic system ....................................................33. 1.7 1.7.1. Network inference by the growth microbiome time-series (GMT) datasets...........33. 1.7.2. Network validation by probiotic system .................................................................34. 1.7.3. Research objectives ................................................................................................34. Materials and Methods ............................................................................ 37. 2 2.1. Animal individuals......................................................................................................37. 2.1.1. Treefrogs from season fall, winter, and spring .......................................................37. 2.1.2. AH treefrogs in the laboratory................................................................................38. 2.2. Microbiome dataset of seasonal changes ....................................................................39. 2.2.1. Feces collection and DNA extraction ......................................................................39. 2.2.2. 16S rRNA sequencing ..............................................................................................40. 2.2.3. 16S rRNA amplicon analysis ...................................................................................41. 2.2.4. Functional predictions ............................................................................................41. 2.3. The growth microbiome time-series dataset ...............................................................42. 2.3.1. Animal individuals ..................................................................................................42. 2.3.2. 16S rRNA sequencing ..............................................................................................43. 2.3.3. 16S rRNA amplicon analysis ...................................................................................44. 2.4. Network inference tools ..............................................................................................44. 2.4.1. Correlation-based network inference tool ..............................................................44. 2.4.2. Regression-based network inference tool ..............................................................46. 2.4.3. Training dataset and validation .............................................................................47. 2.5. Network analysis ........................................................................................................49. 6.

(7) 2.5.1. The probiotic sub-network for validation and application .....................................49. 2.5.2. Network visualization .............................................................................................49 Microbiome dataset of probiotic trials ........................................................................50. 2.6 2.6.1. Probiotic selection and culturing ............................................................................50. 2.6.2. Animal individuals ..................................................................................................51. 2.6.3. 16S rRNA sequencing and 16S rRNA amplicon analysis .........................................51. 2.6.4. Validation of network inference .............................................................................51. 2.6.5. Level of IL-10 ...........................................................................................................52 Statistical analysis.......................................................................................................53. 2.7 2.7.1. 3. Alpha and beta diversity .........................................................................................53. Results – compositional and metabolic functional changes of gut microbiota. over AH treefrogs ............................................................................................ 55 Sampling information in fall, winter, spring, and artificial hibernation ......................55. 3.1 3.1.1. Environmental parameters .....................................................................................55 Microbiome in fall, winter, spring, and artificial hibernation .....................................56. 3.2 3.2.1. Sequencing quality and throughput .......................................................................56. 3.2.2. Microbial diversity ..................................................................................................56. 3.2.3. Microbial compositions ..........................................................................................57 Functional prediction ..................................................................................................59. 3.3. 4. Results – inference, validation, and potential application of microbial. interactions ..................................................................................................... 63 4.1 4.1.1 4.2. Model selection...........................................................................................................64 Performance of correlation-based and regression-based model............................64 The growth microbiota time-series (GMT) dataset .....................................................64. 4.2.1. Preliminary test for the GMT dataset .....................................................................64. 4.2.2. Sequencing throughput ..........................................................................................65. 7.

(8) 4.2.3. Diversity changes in the GMT dataset ....................................................................65. 4.2.4. Compositional changes in the GMT dataset ...........................................................66. 4.3. The inferred interaction relations ................................................................................66. 4.4. Probiotic selection ......................................................................................................67. 4.5. The inferred positive relation facilitates the growth of probiotics, L. garvieae ..........68. 4.6. The up-regulated IL-10 caused by the increased L. garvieae .....................................69. 4.7. Validation on probiotic system ...................................................................................71. Discussion ................................................................................................ 73. 5 5.1. Seasonal impacts on gut microbiome of hibernators ..................................................74. 5.1.1. Hibernation decreases microbial diversity among all hibernating animals............74. 5.1.2. Hibernating animals shared similar phyla of gut microbiota .................................76. 5.1.3. Seasonal changes in composition of gut microbiota among hibernators...............77. 5.1.4. Seasonal changes in functions of gut microbiota among hibernators ...................77. 5.2. Habitat impacts on gut microbiome of amphibian ......................................................78. 5.2.1. Gut microbiota of adult amphibian harbor similar composition in aquatic animals. than terrestrial animals .......................................................................................................79 5.3. Artificial hibernation increases the risk of pathogenic infection in treefrogs .............81. 5.3.1. The increase of pathogenic bacteria in artificial hibernation .................................81. 5.3.2. High risk of pathogenic infection in artificial hibernation ......................................81. 5.4. The potential control agent for invasive treefrog: a perspective from gut microbiota 83. 5.5. Regression-based models provide better performance than correlation-based methods 84. 5.5.1. Performance comparison between regression-based and correlation-based. methods ...............................................................................................................................85 5.6 5.6.1. The breakthrough of gut microbiota in amphibians – microbial interactions .............86 The potentials of network inference in amphibian studies .....................................87. 8.

(9) 5.6.2. Inference of microbial interactions based on the GMT datasets ............................88 The interactions between probiotics and other gut microbes in amphibians...............90. 5.7 5.7.1. Consistency of microbial interactions between bottom-up and top-down research. and more insights from network inference .........................................................................90 The probiotic consortia ...............................................................................................91. 5.8 5.8.1. Potential probiotic consortia determined by network inference ............................91 Validation of network inference using probiotics .......................................................92. 5.9 5.9.1. Network validation according to real abundance of bacterial changes in the. community...........................................................................................................................92 The bridge between network theory and applications.................................................94. 5.10. Conclusions and future directions ............................................................ 95. 6 6.1. Future directions .........................................................................................................96. 6.1.1. Incorporate species network with metabolic network ...........................................96. 6.1.2. Pathogen resist gut microbiota ..............................................................................97. Figures and Tables .......................................................................................... 99 Figures .....................................................................................................................................99 Figure 1 – The diagram of correlation-based method. ........................................................99 Figure 2 – The flowchart of a null model. ......................................................................... 100 Figure 3 – Simulated time series dataset of 50 operational taxonomic units spending 10 time points........................................................................................................................ 101 Figure 4 – Weekly records of temperature and humidity in October, November, December, January, February, and March.......................................................................................... 101 Figure 5 – Alpha-diversity rarefaction plot of fecal microbioas between AH and NH treefrogs. .......................................................................................................................... 102 Figure 6 – Heatmap. ......................................................................................................... 103. 9.

(10) Figure 7 – Compositional variation in microbial communities between AH and NH treefrogs. .......................................................................................................................... 104 Figure 8 – Core genera of AH and NH treefrogs. .............................................................. 104 Figure 9 – Rarefaction analyses for the observed number of genera from 12 time points. .......................................................................................................................................... 105 Figure 10 – Time dependent taxonomic composition spanning 15 days over AH. ........... 106 Figure 11 – Inferred interaction partners of Bacillus, Corynebacterium, and Lactococcus. .......................................................................................................................................... 107 Figure 12 – Relative abundance of IIPs of Bacillus, Corynebacterium, and Lactococcus after two-week oral trials.......................................................................................................... 108 Tables ................................................................................................................................... 109 Table 1 – Summary of sample information for AH and NH treefrogs. .............................. 109 Table 2 – Sample information of preliminary time-series................................................. 109 Table 3 – Summary of sample information of the GMT dataset. ..................................... 110 Table 4 – The intrinsic growth rate and initial abundance of 50 OTUs from simulation dataset ............................................................................................................................. 111 Table 5 – Summary of sample information of oral administration for bacterial composition. .......................................................................................................................................... 112 Table 6 – Summary of sample information of oral administration for quantitative PCR. 112 Table 7 – Phylogenetic diversity indices of AH and NH treefrogs. .................................... 113 Table 8 – Relative abundance of dominant phyla in AH and NH treefrogs. ..................... 113 Table 9 – PICRUSt showing predicted relative abundance of KEGG ortholog groups (Level 2 KOs). ................................................................................................................................. 114 Table 10 – The performance of correlation-based and regression-based methods ......... 115 Table 11 – Time-dependent diversity spanning 15 days over AH ..................................... 115 Table 12 – Expression analysis of IL-10 and level of Lactococcus. .................................... 116. 10.

(11) Table 13 – Validation of IIPs that correlate with Lactococcus, Corynebacterium, or Bacillus. .......................................................................................................................................... 116 Supplementary Tables .......................................................................................................... 117 Supplementary Table 1 – Interacting matrix between 50 OTUs. ...................................... 117 Supplementary Table 2 – Tax4Fun showing predicted relative abundance of KEGG ortholog groups............................................................................................................................... 117 Supplementary Table 3 – Microbial interactions inferred by Pearson’s correlation coefficient based on simulated datasets. ......................................................................... 117 Supplementary Table 4 – Ten thousand and three hundred fourteen significant inferred interacting pairs (IIPs) identified by MetaMIS. ................................................................. 117 Supplementary Table 5 – The raw data for the interaction network at the species level. 117 Supplementary Table 6 – Fold change of the inferred relationships with Lactococcus, Corynebacterium, and Bacillus. ........................................................................................ 117 Supplementary Table 7 – Relative abundance of IIPs of Bacillus, Corynebacterium, and Lactococcus after two-week oral trials. ............................................................................ 118. References ..................................................................................................... 119. 11.

(12) 12.

(13) 1 Introduction 1.1 Strong links between animal gut microbiome and its physiology and behaviour 1.1.1 The amount and importance of microorganisms in the gastrointestinal tract (gut microbiota) Microorganisms are found in many parts of the animal body primarily on the internal (e.g., gastrointestinal tracts, saliva, and oral mucosa) and external surfaces (e.g., skin). In vertebrates such as human, the vast majority of bacteria reside in the intestine, also known as gut microbiota. Gut microbiota, used to be called the microflora of the gut, is the assembly of microbial cells including bacteria, viruses, and fungi, whereas gut microbiome refers to all the genes inside these microbial cells. The estimated ratio of gut microbiota to human cells is close to 1:1 (Sender, Fuchs et al. 2016). These enormous numbers of bacteria in the vertebrate gastrointestinal tract forms a community structure, contains varieties of genetic materials, and create high metabolic capacities that potentially contributed to the host metabolic pathways. These residential microorganisms in the intestine share nutrition with the host and form a variety of metabolic functions, suggesting a strong potential in connections between the hosts and their gut microbiota.. 1.1.2 Phylogenetic diversity and structural assembly of gut microbiota correlates to host physiology and behaviour Studies have suggested that gut microbiota are engines of globally important host physiological processes among varieties of animals. Lederberg has emphasized the importance of having the broad view on the relationship between human infections and the residential microbes (Lederberg 2000). Microbial diversity in a microbial assemblage is the central property in community ecology and is important to determine the functioning of ecosystem (Konopka 2009). By further taking the advantages of next-generation sequencing technology, researchers reveal the most comprehensive and high resolution of microbial diversity within the human gut based on 16S rRNA amplicon approach. Instead of bacterial cultivations approach heavily 13.

(14) relied on growth condition for each strain, these sequences are obtained directly from DNA extracted from intestinal feces using PCR primers targeted to characterize phylogenetic groups. These technologies help to reveal the low levels of deep bacterial and archaeal diversity and provide insights into the direct connections between host physiology and their gut microbiota. For example, intestinal dysbiosis drives metabolic dysfunctions in human studies. Among the huge body of approach describing disease-associated microbiota, the loss of microbial diversity (LOMD) has been constantly found as a common feature of most dysbioses, such as colorectal cancer (Ahn, Sinha et al. 2013), celiac disease (Schippa, Iebba et al. 2010), irritable bowel syndrome (IBS), with (Carroll, Ringel‐Kulka et al. 2012) or without diarrhea (Durbán, Abellán et al. 2012), and Crohn's disease (CD) (Sha, Xu et al. 2013, Matsuoka and Kanai 2015). LOMD is also identified as a risk factor of relapse in Clostridium difficile colitis (Chang, Antonopoulos et al. 2008). In addition to human diseases, the changes of diversity and composition of gut microbiota have also been addressed on ecological animal behavior. For example, invasive Asian carp shows significant differences in microbial diversity and distinct compositional microbial patterns compared with indigenous American one (Ye, Amberg et al. 2014). These approaches suggest that the intestines of animals reside diverse microbial ecosystems that have profound effects on host physiology. The diversity and composition of gut microbiota is important to the ways in symbiotic functions and further influences the host’s metabolic capabilities and numerous aspects of physiology and behavior (McFall-Ngai, Hadfield et al. 2013). These animals physiological or behavioural changes caused by the alterations of their gut microbiota suggest future potential solutions in varieties of scientific fields, such as the prevention of invasive species. In this dissertation, I will aim to focus on exploring gut microbiota of an invasive species – Polypedates megacephalus.. 14.

(15) 1.2 Seasonal behavioural observations of Polypedates megacephalus 1.2.1 Population distribution of P. megacephalus in Taiwan Polypedates megacephalus (Hallowell 1861), also known as Spot-legged Treefrog, Hong Kong Whipping Frog, White-lipped Frog, is a member of P. leucomystax complex of cryptic species. Although P. megacephalus has been reported since 1861, it is however, the taxonomic lineage has been identified later by Matsui et al., which demonstrated that P. megacephalus is distinct from P. leucomystax (Matsui, Seto et al. 1986). P. megacephalus has a natural distribution that extends through southern China, Hong Kong, and the Southeast Asian Peninsula. In Taiwan, the dispersal surveys showed that P. megacephalus has been distributed over Taipei City, New Taipei City, Taoyuan County, MiaoLi County (which is added to the list in 2013), Taichung City, Zhanghua County, Yunlin County, and Pingtung County. Their current studies even suggest that P. megacephalus has rapid dispersal into more than 80% of cities of Taiwan. It is therefore, P. megacephalus has been identified as an invasive species, and has potentials to influence the ecosystem in Taiwan. In 2006, individuals of P. megacephalus had been found in Changhua city, Taiwan. It was speculated that these treefrogs were introduced to Taiwan via aquatic plants through agricultural business (Tsai 2013, Ko-Huan Lee 2019). In the past decade, increased approach had focused on studying this invasive species on dietary analysis, effective dispersal accumulation in natural fields, physiology, and the potential impacts of P. megacephalus on ecological biodiversity in Taiwan. Results from these studies suggest that P. megacephalus has severe potentials to impact on the ecosystem in Taiwan.. 1.2.2 Seasonal dispersal and habitat usage of P. megacephalus in Taiwan Yang et al. sought to reveal the natural population dispersal of P. megacephalus over the seasons and their ecological habitat usage (Yang and Gong 2014). They found that P. megacephalus earliest breed starts from March and ends around October. The 15.

(16) breeding season of P. megacephalus is relatively longer than several native amphibians, which can have potential impacts on natural ecology of other species that are resident in the same habitats. P. megacephalus inhabit in a variety of habitats including grassland, ponds, forests, marshes, plantations, and cultivated fields. During their breeding season, they prefer ponds, streams, or water containers habitats and move toward shallow and still water to lay foam nests. The foam nests have reported to contain 300 – 400 eggs, and typically attached to vegetation (Christy, Savidge et al. 2007).. 1.2.3 Invasive P. megacephalus, potential threat, and current solutions for invasion To understand the potential threats of P. megacephalus on ecosystem in Taiwan, Yang et al. also collected stomach contents from 45 individuals in northern Taiwan in 2012. They suggested that P. megacephalus is a potential helminth carrier that carries multiple parasites in the intestine and may cause severe threats to the ecosystem in Taiwan (Yang, Norval et al. 2014). Three gastrointestinal helminths were identified, Pseudoacanthocephalus bufonis, Cosmocerca ornate, and Oswaldocruzia hoepplii. In addition to P. bufonis, they demonstrated that P. megacephalus is the new record to host C. ornate and O. hoepplii. Due to the increased tendency of rapid migrating dispersal of P. megacephalus across counties, these findings suggest a high risk of parasites spread toward the natural ecosystem and may cause severely potential threats in Taiwan by invasive P. megacephalus. Taiwan is not the first record to get invaded by P. megacephalus. In 2004, Christy et al. found P. megacephalus was invaded in Guam (Christy, Clark et al. 2007). Their study showed that the foam nests can be found in a wild range of habitats, including shallow, still water, typically attached to emergent vegetation. This invasive species also rapid extend their breeding across different cities such as Agat, Malojloj, Inarajan, Yona, and Ordnance Annex, suggesting strong dispersal potentials. 16.

(17) Although this wild dispersal creates impacts on the ecology, however, efficient strategies to inhibit or slow down the invasion are still limited. To control the dramatically increasing population of P. megacephalus in Taiwan, Yang et al. started research teams involving in projects for amphibian ecological conservation and evaluated the ecological dominance of P. megacephalus by counting the relative amounts of P. megacephalus, compared with other amphibian species, such as rain frogs etc., that are resident in the same habitat. Their findings between 2012 and 2013 showed that the ratio of P. megacephalus out of other amphibians dropped from 69% to 29% at Waziwei, from 54% to 41% at Taichung Metropolitan Park, but ascended from 15% to 16% at Bealong Temple. Although the amphibian ecological conservation projects by Yang et al. suggested an extraordinary achievement on the control of P. megacephalus. However, the performance of the ecological control strongly relies on the amounts of participants in the projects. Therefore, there is still an urgent need to develop new strategies to the conservation of biodiversity in Taiwan. Here, I propose a perspective on gut microbiome of P. megacephalus to deeply understand microbial compositions, the potential metabolic capacity generated by gut microbiota, and the possible mechanisms of microbial interactions. This dissertation first aims to reveal the alterations of gut microbiome of adult P. megacephalus in the natural habitat across different seasons and further investigates the possible microbial interaction network between these discovered bacteria in the intestine of P. megacephalus.. 1.3 Seasonal alterations on gut microbiota in hibernators 1.3.1 Compositional changes of gut microbiota during hibernation Hibernators, such as amphibians, few mammalians and birds, have evolved the physiological features to survive unfavourable winter environment. Hibernation is a biological adaptation for animals that helps animals to conserve energy during food shortage in winter and avoid extinction. During torpor, typical hibernators lower their 17.

(18) body temperature to only a few degrees above ambient temperatures to reduce metabolic activity and energy expenditure (Book 1974). Torpor bouts are interrupted by reactivating animal metabolic activities when body temperature is restored. Seasonal patterns in physiology of hibernators are accompanied by marked changes. For example, gastrointestinal tract of hibernators reduces in crypt depth, villus length, and total mucosal surface area. Epithelial permeability, a major component of the intestinal barrier that restricts passage of luminal molecules such as bacterial lipopolysaccharide, increases during hibernation (Carey and Physiology 1990). Hibernation also reduces digestive enzyme activities (Balslev-Clausen, McCarthy et al. 2003) and affects the function of the innate and the adaptive systems (Bouma, Carey et al. 2010).. 1.3.2 Seasonal shifts of dominant bacterial populations in hibernators identified by conventional methods Microbial diversity in the intestine of most hibernating animal varies temporally in response to seasonal changes due to the food availability across seasons. Food availability determines the temporal variation of dietary composition for these hibernators and their gut microbiota. As a result, phylogenetic compositions of gut microbiota in hibernators can be altered dramatically across seasons due to the variability of diets and patterns of nutrition intake (Goldizen, Terborgh et al. 1988, Overdorff, Strait et al. 1997, Sommer, Ståhlman et al. 2016). The nutritional shifts and the changes in morphology and physiology of intestine across seasons suggest strong impacts on gut microbiota in microbial composition, diversity, and metabolic functions in hibernators. There are varieties of conventional approach on exploring the impacts of hibernation on gut microbiota in amphibians. For example, studies on artificial hibernation of northern leopard frogs (Rana pipiens) and bullfrogs demonstrated a drop in bacterial counts and a change in the composition of gut microbiota (Carr, Amborski et al. 1976, 18.

(19) Gossling, Loesche et al. 1982). In addition, studies on hibernating amphibians also characterized a higher risk of pathogenic infections by identifying the increase of pathogenic populations in the intestine. Hibernating northern leopard frogs (Van der Waaij, Cohen et al. 1974) and chilled southern bullfrogs (Rana catesbiana) (Carr, Amborski et al. 1976) harbored potentially pathogenic facultative bacteria in the intestine. Facultative (preferentially aerobic but facultatively anaerobic) bacteria from the intestines of frogs have been investigated as a source of septicemia, often associated with chilling and hibernation (Gibbs, Gibbs et al. 1966, Carr, Amborski et al. 1976), which occasionally kills large numbers of frogs in the laboratory and in the wild (Gibbs, Gibbs et al. 1966, Hawksworth 1974). Carr et al. (Carr, Amborski et al. 1976) and Gibbs et al. (Gibbs, Gibbs et al. 1966) also found that hibernation can alter the relative concentrations and proportions of facultative versus anaerobic bacteria, leading to disease. These studies suggest that during hibernation, gut microbiota of amphibian reduces the amounts of bacterial populations, and it may increase the risk of pathogenic infections due to the increase of pathogenic populations.. 1.3.3 Seasonal restructure of gut bacterial assembly in hibernators characterized by advanced sequencing technology Hibernating ground squirrel has been studied extensively with regard to the gut microbiota. Research on ground squirrels extend the comparisons of gut microbiota between hibernating and non-hibernating stage into the seasonal changes of gut microbiota. Based on cultivated approach, Barnes and Burton reported that the total number of viable bacteria in the intestine decreases during hibernation (Barnes and Burton 1970). Later on, Carey et al. took advantage of the improvement of sequencing throughputs to extend the research on gut microbiota of ground squirrel into seasonal-wise scale, and explored the effect of the annual hibernation cycle across different seasons on microbial diversity and composition using deep sequencing of 16S rRNA genes from fecal contents (Carey, Walters et al. 2013). In 19.

(20) their work, they found that the phylogenetic diversity was the lowest in late winter, and the highest in spring after two weeks period of re-feeding, suggesting hibernation causes the reduction of microbial complexity, and the complexity is able to return back to the original status. This study also demonstrated the compositional changes of gut microbiota across the seasons. The most dominant phyla in ground squirrel are Bacteroidetes, Firmicutes, and Verrucomicrobia. The relative abundance of Bacteroidetes and Verrucomicrobia that contain species, e.g., Akkermansia muciniphila, capable of surviving on host-derived substrates such as mucins increases in hibernating ground squirrel. However, Firmicutes reduces relative abundance in hibernating ground squirrel. It may be due to the large amounts of species in Firmicutes prefer dietary polysaccharides and thus reflects the reductions in relative abundance of Firmicutes in hibernating ground squirrels. These results suggest that the gut microbiota of ground squirrel is restructured across seasons that reflects differences in microbial preferences for host-derived substrates or dietary intake across seasons. Another approach using high-throughput sequencing technology on exploring the changes of gut microbiota across seasons is brown bear (Sommer, Ståhlman et al. 2016). Compared with the active seasons, gut microbiota in hibernating season has reduced in microbial diversity, levels of Firmicutes and Actinobacteria, and increased levels of Bacteroidetes, which are also found in ground squirrels. In order to understand the connections between the changes of gut microbiota and host physiologies, they found that several metabolites involved in lipid metabolism were also affected by gut microbiota in hibernation shown in their study. Transplantation of the gut microbiota of brown bear from winter to germ-free mice also demonstrate some seasonal metabolic features observed in brown bear such as physiological energy absorption, suggesting that seasonal variation in the microbiota may contribute. 20.

(21) to host energy metabolism in the hibernating brown bear. These results concluded that gut microbiota of brown bear changes in composition and metabolic functions across the seasons, and manipulation of microbial composition in the intestine can modulate host physiology. The population of P. megacephalus have widely distributed in Asia. However, the knowledge of gut microbiota of this invasive species is still lacking, including composition and functions. Although evidences consistently indicated that gut microbiota of varies of hibernators restructure in microbial composition and metabolic functions, however, none of the studies had addressed gut microbiome in this invasive species. We have limited knowledge of how microbial composition could alter physiology of P. megacephalus. Furthermore, the impacts of seasonal changes on gut microbiome of P. megacephalus is also unclear. It is crucial to have a deep understanding of gut microbiota to reach a global view to understand this invasive species.. 1.4 Network inference provides insight into the mechanisms of gut bacterial assembly 1.4.1 The potentials of top-down approach on microbial community The use of genome sequences and related approach (Pace 1997, Giovannoni and Stingl 2005) has overcome the need for cultivation to characterize and identify microorganisms in nature. Advance in sequencing technologies couples with new bioinformatic developments have allowed the scientific researchers to investigate the microbes that inhabit everywhere (Gilbert and Dupont 2011). Over the last 10 – 15 years, studies for the compositions of microbial community have increased exponentially. High-throughput sequencing (i.e., next-generation sequencing, NGS) technology have generated massive amount of reads on small-subunit ribosomal RNA gene (16S rRNA), which elucidates the composition of microbial community. Several 16S rRNA processing pipelines were established using software packages, such as 21.

(22) MOTHUR (Schloss, Westcott et al. 2009) or Quantitative Insights Into Microbial Ecology (QIIME) (Caporaso, Kuczynski et al. 2010). Follow analytical pipelines of the 16S rRNA protocol, including removal of multiple sources of potential bias generated by 16S rRNA sequences using pyrosequencing (Kunin, Engelbrektson et al. 2010), I can describe the compositions of microbial communities, diversity, and how communities may change across time, geographical space, or varieties of experimental treatments. All pipelines result in high comparable view of the microbiomes between different experimental treatments (Hsiao, McBride et al. 2013, Ridaura, Faith et al. 2013, Kohl, Amaya et al. 2014, Sommer, Ståhlman et al. 2016). High throughput-based studies allow us to reveal microbial composition. Studies based on microbial compositional structure provides all genomic information in the community and deduce the potential metabolic capacity. However, one of the limitations of this approach is that it is hard to interpret the mechanisms of microbial structure only based on microbial compositions. Therefore, deciphering microbial interactions by using statistical algorithm or models currently provide potentials to understand the mechanisms of microbial community. It is currently attracting the interests to attempt to deduce the structure of microbial interaction networks based on 16S rRNA amplicons. The recent advances in system microbiology have been applied by the methods that characterize microbial interaction networks using longitudinal microbial studies or time-series data. Such microbiome data can be used for inferring an interaction structure to examine the static or dynamic of biological processes on a system level.. 1.4.2 Deciphering microbial interaction provides insights into causal relationships The traditional tools of microbiology to decipher microbial interactions, such as pure cultures and genetic studies, tend to study each microorganism in isolation and characterize direct interaction between different isolations by co-culturing on the 22.

(23) same plate. This reductionist approach provides valid information on pairwise or few interacting bacterial strains to understand microbial interactions in a small scale of community. However, it is not well suited for learning about microbial interactions in the real world based on such approach because few bacterial isolations do not reflect natural community. In addition, it is hard to mimic natural environmental conditions in the laboratory. Holistic approach, which conduct all bacteria in the community and study natural habitats directly, can yield comprehensive data to deduce more reliable microbial interactions. Gut microbiota are complex in both compositional structure and metabolic function due to their reproduction variability, their ability to self-reproduce, and even host temporal dynamics. This complexity can be well represented by using computational modeling approach, such as network analysis. Graph-theoretical and top-down system-oriented approach can facilitate deciphering microbial interactions, characterizing potential interacting modules as interpretations of microbial consortia, and enhance our understanding of complex evolutionary and ecological processes. Network approach that model the co-occurrence or causal relationships among microorganisms create insights into the applications on microbiome studies by finding microbial relationships essential for community assembly and further to deduce the potential influence of various interactions on the host health. Deciphering microbial interactions of gut microbiota has received much attention in recent years, because the inferred interaction relationships calls for better knowledge than we have today about how microbial community structure in the intestine relates to their symbiotic functions. To engineer gut microbiota for host physiological benefits, we must understand the rules of microbial community assembly and interacting structures, i.e., the direct interactions between community members. Microbes do not exist in isolation and are often found in a consortia of different. 23.

(24) microbial species populations (Handelsman 2004). An assemblage of microorganisms potentially interacts with each other or with the organic chemicals from the environments (Konopka 2009). Within a community of species sharing limited resources, interactions can be described in terms of “positive”, “negative” interaction or “no impact” determined by the type of relations between members. Therefore, deciphering microbe-microbe interactions in the gut is likely to extend the knowledge on the functions of microbial symbionts. However, the inference of microbial interactions is far from straightforward due to the high complexity of microbial community (Faust and Raes 2012). In the following section, we attempt to review the methods on characterization of the interactions among microorganisms and describe the current gap in the network research.. 1.4.3 Network inference theory A variety of methods have been used to infer microbial interaction network based on microbiome data. These methods have different performances in accuracy, speed, efficiency, and computational requirements, and can be categorized into simple pairwise correlation methods, such as Pearson or Spearman correlation measurements, and more complex multiple regression methods and probabilistic graphical models, e.g., Bayesian model. Some of the methods are very popular used in microbiome data due to the ease of use and better computational speed (Faust and Raes 2012), while probabilistic graphical models have not yet been applied extensively to address biological questions, although the performance of such models showed good success in other scientific fields with high accuracy and minimal bias. Here I discuss two of the different methods that popularly apply on microbiome data to infer interaction network. The first method, un-directional co-occurrence networks, is relatively simple and required low cost of computational time and resources. They use a pairwise dissimilarity index such as Bray–Curtis or Kullback–Leibler to infer co-occurrence 24.

(25) network from operational taxonomic unit (OTU) that defined from microbiome data. The statistical significance of pairwise dis-similarity index could be evaluated by permutation tests, all significant pairwise relationships are aggregated to infer microbial interaction network (Faust, Sathirapongsasuti et al. 2012). Faust et al. have developed a pipeline combining varieties of measurements of dependency, such as dissimilarity (e.g., Kullback–Leibler), similarity, and correlation to infer microbial interactions based on the oceanic plankton community (Lima-Mendez, Faust et al. 2015). Correlation-based methods, which still aim for un-directional co-occurrence networks, is a popular alternative to dissimilarity-based network inference. Different from calculating dissimilarity index, correlation-based methods use correlation coefficients such as Pearson's correlation coefficient or Spearman's rank correlation coefficient to detect significant pairwise interactions between OTUs. Correlation-based methods have been popularly used in human gut microbiome (Arumugam, Raes et al. 2011). For example, Arumugam et al. used correlation-based method to analyze combined datasets of human fecal microbiome from four different countries to conclude that there are three robust human gut enterotypes that were not nation-specific or continent-specific. However, correlation-based methods attempt to detect dependencies between OTUs in a microbiome and remains limitations, such as detecting spurious correlations among low-abundance OTUs closed to zero or being sensitive to compositionality (Chen and Li 2016). Weiss et al. (Weiss, Van Treuren et al. 2016) compared eight correlation-based methods and evaluated the performance of each method on both synthetic and real microbiome data in response to challenges by assessing their ability to identify a range of time-series ecological relationships and distinguish signals from noise. Their results provided the performance and. 25.

(26) shortcomings of each method and concluded specific recommendations for the use of correlation-based methods. Although correlation-based methods to infer interaction networks are fast, it is however not able to capture more complex forms of multiple microbial interactions (Faust, Sathirapongsasuti et al. 2012). Regression-based methods modify the calculation strategy from pairwise comparison to multiple relationships by inferring the abundance of one OTU from the combined abundance of other OTUs. Regressionbased methods are frequently used in biological fields; however, the regression results are usually difficult to interpret with biological meanings. For example, the interactions between OTUs inferred by regression-based methods might not always suggest an underlie biological basis for the relationships. In addition, regression-based methods suffer from overfitting that creates in the number of false positives. These false positives caused by overfitting can be reduced by using cross-validation and sparse regression. These computational network inference methods have been proposed to infer microbial interactions from OTUs abundance profiles. Some of the network inference methods are designed to infer static networks by using microbial profiles that do not involve temporal aspect into consideration. The other network methods are specifically designed for the dynamic aspects of OTUs dependencies. Microbial populations are not instantaneous but vary across time. These dynamic properties characterized from the temporal microbiome data should shed light on the causal dependencies between OTUs in the communities.. 1.4.4 From static network to dynamic network: time-series data The inference of microbial interactions based on network theories can be considered as static models of microbial communities. The static network modeling effectively provides a “snapshot” for the community status for a given time point. However, there are several phenomena, such as community dynamics, perturbation, and succession, 26.

(27) could not be studies only based on the static models. These biological phenomena could be further studied if applying specifically designed dynamic models, which required a time series data to describe how microbial population change over time. In macro-ecological system, dynamic models have been widely used to identify the stability and development of the communities (Holling 1973, May 1974, Ives and Carpenter 2007). The recent availability of time series microbiome data generated by NGS technologies make it possible to apply dynamic models on microbial communities and target the questions of community dynamics in micro-ecosystem. Studies have shown successful inference of microbial interactions in the communities by analyzing high-throughput 16S rRNA sequencing in a time series course (Faust and Raes 2012). Longitudinal microbial approaches are informative because they provide microbial dynamics of communities, and the response of bacterial structure to external perturbations (Dethlefsen, Huse et al. 2008). Monitoring all bacterial populations changes over time provides further potentials to discover the interacting mechanism between bacterial members, such as how commensal bacteria in the intestine resist the invasive pathogenic species (Stein, Bucci et al. 2013). These studies suggest the values of the longitudinal, transversal, or both measurements of microbial communities that can help to produce a set of significant dependent relationships and build dynamic models to predict interacting mechanism of microbial communities. A dynamic model composes of a set of Boolean functions or differential equations (Hecker, Lambeck et al. 2009), which can describes the abundance changes of bacterial members in the community over time. The Boolean function only describes bacterial present or absent, while differential equations describe absolute or relative abundance changes over time. Dynamic modeling already has a long history in single population ecology (May and Jaenike 1973), however, few approach attempted to. 27.

(28) include multiple species to build up dynamic models for microbial communities. Mounier et al. (Mounier, Monnet et al. 2008) included multiple microbial species in the analysis to model cheese fermentation community interactions by using generalized Lotka-Volterra (gLV) equations. In their study, the model predicted negative interactions among three different yeast species and conducted further cocultural experiments to confirm these negative relationships. The key feature that makes gLV model extensively used to construct microbial interactions is the model parameters could directly capture the growth rates and the pairwise interactions between all bacterial species in the community (Alshawaqfeh, Serpedin et al. 2017, Cao, Gibson et al. 2017). A variety of studies have been applied gLV models to infer microbial interaction network using 16S rRNA sequencing data over a timescale. For example, Stein et al. (Stein, Bucci et al. 2013) extended generalized Lotka-Volterra (gLV) equations to study the mechanism of C. difficile colonization in mice after antibiotic perturbation. They inferred that the genera Akkermansia, Blautia, and Coprobacillus had inhibitory interactions on C. difficile. In contrast, Enterococcus and Mollicutes could positively affect the growth of C. difficile, while the genus Barnesiella was predicted to inhibit growth of the genus Enterococcus. These results demonstrate that gLV model can be applied to high-resolution time series data to infer community structure and response to external perturbations. In addition, gLV models have already been extensively applied to characterize microbial interactions in varieties of studies (Mounier, Monnet et al. 2008, Fisher and Mehta 2014, Marino, Baxter et al. 2014, Buffie, Bucci et al. 2015, Bucci, Tzen et al. 2016).. 1.4.5 Current pitfalls of network inference Although gLV models are the most popular differential equations used in ecological modeling, it is still in its infancy to model microbial communities that consist of multiple species and faces varieties of challenges. For example, these models must handle a large amount of species or group of species in the real-world microbial 28.

(29) communities. Although there are increasing studies on deciphering microbial interactions by applying network theories on microbial ecosystems, little attention has been paid to deal with the high complexity of microbial communities. In addition, there is still a huge gap between computational theories and applications due to the lack of reliable validation systems. The major reason is the lack of benchmark metagenomics datasets (Berry and Widder 2014). For example, the gLV models are widely applied to interpret varieties of biological questions of interest via inferring microbial interactions (Mounier, Monnet et al. 2008, Stein, Bucci et al. 2013), however, the validations of advancing analytical algorithms and pipelines must reply on simulated data. It is therefore that network inference approaches remain a gap between analytical theories and biological applications.. 1.5 Well-developed intestinal probiotic system 1.5.1 Lactic acid bacteria (LAB) – the wild used probiotics Probiotics are originally defined as the organisms and substances which contribute to the intestinal microbial balance (Paker 1974). After oral administration, probiotics are able to colonize and multiply in the intestine of host and execute numerous beneficial effects to the host (Cross 2002). The most commonly seen probiotics are lactic acid bacteria (LAB). LAB have been reported to have beneficial effects to improve physiological functions in fish, amphibians, and mammals (Gill and Guarner 2004, Nayak 2010, Pasteris, Guidoli et al. 2011), and popular used in aquaculture to control pathogenic agents and the diseases (Verschuere, Rombaut et al. 2000), influence the immune enhancement (Panigrahi, Kiron et al. 2005), and a wide range of developments including growth performance and behavior (Yang, Cao et al. 2012, Borrelli, Aceto et al. 2016). Dietary probiotic treatments have already been investigated to successfully inoculate into the intestine in varieties of animals although they have to pass through high concentration of acid in the stomach before going to intestine. For example LAB have been shown the ability to resist low pH and 29.

(30) bile salt (Cotter and Hill 2003, Geraylou, Vanhove et al. 2014). In addition, LAB have been indicated to have good colonization capacity on gastrointestinal tract, including stomach and intestine, in simulated gastrointestinal conditions of fish (Amin, Adams et al. 2017). Therefore, LAB are good candidates as oral probiotic treatments.. 1.5.2 The interactions between probiotic strains and other bacteria There are tons of approach have shown that the relations between probiotic modulations of gut microbiota and improvement of health in human, such as treatment of diarrhoea, treatment of Helicobacter pylori infection, treatment of atopic disease, and enhancement of immune functions (Gill and Guarner 2004). Other approach highlighting the associations between resident microbiota and health benefits, including breakdown of complex molecules in food, protection from pathogens, and development of immune system were also documented (Bäckhed, Ley et al. 2005, Mazmanian, Round et al. 2008, Lee and Mazmanian 2010, Dethlefsen and Relman 2011). The use of probiotics is able to exert inhibitory effects against undesired microorganisms or pathogens, and to support microbial defense mechanisms of hosts (Hernandez, Cardell et al. 2005). Many members of probiotics have been reported to enhance animal resistance to bacterial pathogens by enhancement of host’s immune response (Panigrahi, Kiron et al. 2005), or by producing a variety of important molecules such as organic acids (Vázquez, González et al. 2005), hydrogen peroxide (H2O2) (Verschuere, Rombaut et al. 2000), diacetyl, antimicrobial peptides (AMPs), and bacteriocins (Lin, Chen et al. 2013). Furthermore, some authors suggest that probiotics could compete with pathogens on nutrient uptakes or adhesion site from gut mucosa (Nikoskelainen, Salminen et al. 2001, Harper, Monfort et al. 2011). Several probiotic strains are already described to inhibit pathogens in animals, including Lactobacillus plantarum, Lb. brevis, Pediococcus pentosaceus, Lactococcus lactis, L. garvieae and Enterococcus gallinarum (Pasteris, Pingitore et al. 2009, Pasteris, Roig Babot et al. 2009, Pasteris, Guidoli et al. 2011, Mendoza, Pasteris et al. 2012). These 30.

(31) results have demonstrated that the probiotic strains have varieties of ways to interact with other bacteria in the intestine.. 1.5.3 Host immune responses by probiotics Oral administration of probiotics not only directly increases probiotic concentration in the intestine, but also demonstrates the responses of intestinal immunity to the probiotics. For example, probiotics are claimed to have beneficial effects on gut immune system. The results of animal studies demonstrate that some probiotic strains can successfully modify the mucosal immune response and modulate the levels of specific activation molecules such as cytokine interleukin-10 (IL-10) (de Moreno de LeBlanc, Del Carmen et al. 2011). IL-10 is the most important anti-inflammatory cytokine and a pluripotent cytokine found within the immune response (Asadullah, Sterry et al. 2003). IL-10 is generally considered an immunosuppressive molecule with its main biological function being the limiting and termination of inflammatory responses by suppressing the secretion of inflammatory cytokines. IL-10 has the ability to differently affect the function of several immune cell subsets and is therefore considered to be a broad effective molecule in immunoregulation of host defense (Asadullah, Sterry et al. 2003). Large amounts of approach have shown that probiotics can increase IL-10 production to help in preventing certain diseases that are caused by abnormal inflammatory responses. For example, oral administration of a probiotic mixture that consist of Bifidobacterium and Lactobacillus strains prevented inflammation and mucosal ulcerations in colitis mouse model (Roselli, Finamore et al. 2009). Continuing in this beneficial strain of thought, it is also found that fermented foods containing probiotic strains could also be effective in inducing an antiinflammatory effect (de LeBlanc and Perdigón 2010). Other forms of probiotic administration, such as rectal or intranasal administration, have also been shown to be effective in stimulating IL-10 production (Mastrangeli, Corinti et al. 2009, D’Incà,. 31.

(32) Barollo et al. 2011). These studies demonstrated that probiotics administration exert anti-inflammatory effects through the up-regulation of IL-10 expression. To summarize, the probiotic treatment can provide a validation system on inferred interaction network by directly measure the abundance changes of each bacteria after the treatments. The increase of probiotic strain directly or indirectly influences other bacteria in the intestine and creates the modified abundance profile regulated by the introduced strain as a result of the impact of probiotic strain on other bacteria. It suggests that using probiotic system to validate the inferred interaction network facilitates the understanding of connection between analytical theories and applications.. 1.6 Probiotics used in real applications 1.6.1 Probiotic strains used in raniculture Red-leg syndrome (RLS) is one of the main infectious diseases that cause economic losses in raniculture, such as Lithobates catesbeianus hatcheries. Several pathogenic bacteria are identified to be etiological agents. Treatment or prevention with chemicals or therapeutics results in modifications of the indigenous microbiota and development of antibiotic resistance. Thus, probiotics could be used as an alternative therapy. Lactic acid bacteria are identified as the indigenous microbiota of healthy frogs and can prevent pathogen colonization. For example, Lactococcus lactis was identified as probiotic strain to inhibit etiological agent Citrobacter freundii (Pasteris, Guidoli et al. 2011). Furthermore, Leuconostoc mesenteroides, Pediococcus pentosaceus, and Lactococcus lactis were characterized as probiotics in bullfrog hatcheries (Pasteris, Pingitore et al. 2009, Pasteris, Roig Babot et al. 2009). Briefly, probiotics have been used to restore the balance of the gut microbial ecosystem and control pathogenic infections in raniculture.. 32.

(33) 1.6.2 The growing process of probiotics development Although probiotics have proven successful in the control of enteric pathogens, they still have limitations. The specifically designed probiotic treatment often fails to inhibit the certain pathogens at specific sites of infection and induce low levels of an immune response (McCarthy, O’mahony et al. 2003). To improve the efficiency of probiotic treatments, a variety of approach developed different strategies, such as probiotics mixture (Gill and Guarner 2004), complementary prebiotics for probiotics (Ooi and Liong 2010), and probiotic engineering (Mathipa and Thantsha 2017). Due to the lack of knowledge on a coherent picture of the potential mechanisms governing microbial community, yet limited approach attempts to develop a system that contains probiotic strains and their symbionts.. 1.7 From network theory to practical probiotic system 1.7.1 Network inference by the growth microbiome time-series (GMT) datasets To aim more focuses on dealing with the high complexity of microbial communities to investigate their interaction networks, first of all, we attempt to reduce microbial complexity. We choose a natural microbiome system that is able to specifically design a continuous time series window covering the growth of microbiome from simplicity. Thus, we propose the growth microbiome time-series (GMT) datasets, which represent the progress of how bacterial members growth over time and increase their population size from an initial or simple bacterial assembly. The design of GMT datasets aims to magnify the bacterial ecological interactions by reducing species abundance noises in the community, i.e., when environmental condition becomes harsh, most bacteria decrease their populations and only few bacteria dominate in the community (an initial or simple bacterial assembly). Providing nutrition afterward and designing a reasonable time window allow us to monitor the progress of which bacterial members increase their population step by step. We hypothesize that. 33.

(34) manipulating treefrogs behaviour between hibernating and non-hibernating states provides a natural microbial laboratory to investigate the growing progress of microbiome without any manual effects on microbiome by human experimental operations.. 1.7.2 Network validation by probiotic system To fill the gap between network theories and real-world applications, we further take advantages of probiotic system in treefrogs to validate the inferred networks generated by using the GMT datasets. The probiotic trials in raniculture have a long history to increase the colonization of certain probiotic bacteria in the intestine, as well as to improve the immune responses of amphibians (Pasteris, Guidoli et al. 2011, Mendoza, Pasteris et al. 2012). Due to the lack of metagenomic benchmark datasets for validation, the conventional way to validate the network models relies on in silico simulation datasets. However, simulation datasets have limitations to reflect real biological interactions in microbial ecosystem. An alternative method to validate the inferred interactions from network models is bottom-up experiments, which coculture the bacterial strains that inferred from the models. It still remains limitations due to the difficulties to mimic the real environment, such as nutrition and all related species that should involve in the interactions. It is therefore, we use well-developed probiotic trial system on amphibian to validate the inferred interaction network in our study. The benefits of using probiotics as the inferred network validation system are not only providing clear observation of probiotic strain abundance when follow welldeveloped dosage and duration, but we can validate the host immune response triggered by the bacterial abundance changes.. 1.7.3 Research objectives In this study, I attempt to construct microbial interaction networks of gut microbiota in treefrog (Polypedates megacephalus) and further validate the inferred networks via a probiotic system. The concept of dealing with current hurdle of high microbial 34.

(35) complexity is to design and use the GMT dataset covering the changes of microbial complexity from initial condition. To meet this criterium, we manipulate artificial hibernating treefrogs, which reduce gut microbial composition and diversity naturally without direct interfering by human or manipulation, rather than antibiotic treatment to destroy gut microbiota. To improve the resolution of gut microbial composition on the research of how hibernation impacts on amphibian by using conventional approach (Gossling, Loesche et al. 1982, Gossling, Loesche et al. 1982, Banas, Loesche et al. 1988, Banas, Loesche et al. 1988), we conduct new sequencing technology and 16S rRNA amplicon pipeline to study the differences of gut microbiota between hibernating and non-hibernating treefrogs. In the present thesis, our first objective is to confirm the dramatic changes in microbial diversity, composition, and functions of treefrogs across seasons and artificial hibernation phase. Due to the seasonal significant restructure of gut microbiota confirmed by previous studies, I tend to identify the impacts of seasons and artificial hibernation on gut microbiota, including the shifts of diversity indices, predominant members, pathogenic populations, and predicted symbiotic functions in treefrogs based on 16S rRNA amplicon. The second objective involves the design of the GMT dataset, network model selections, microbial interactions construction, and validation of inferred network by oral trials in an animal system. Choosing the time series datasets is critical in the network inference analysis. A time-series dataset with stable microbial communities remains difficult to infer convincing relationships between microbes due to the lack of statistical significances from the stable time-series communities. Dramatically dynamics in abundance of microbial distribution provides more statistical power to infer microbial interactions. It is therefore, we first test how fast could gut microbiome grow from artificial hibernation to non-hibernation phase of treefrogs to. 35.

(36) determine the time window of the GMT dataset. Then we compare the performance of correlation-based and regression-based network inference models by using training datasets. We aim to construct microbial interaction network by applying the GMT dataset on the regression-based network inference model. The inferred interaction networks delineate pairwise relation between members of treefrog gut microbiota. To validate the inferred relations, we took advantages of probiotic systems that have been well developed in raniculture to increase the colonization of probiotics and improve amphibian health. The responses of gut microbiota to the probiotic trials facilitate the validations of pairwise relations from the inferred network. Through the probiotic system, we evaluate the performance of inferred interaction network by measuring the level of both probiotic concentration and immune responses. The comparison between the groups of single probiotic strain, and a cocktail contains probiotic strain and two co-operated bacteria inferred by our network will provide insights into the potentials of network inference on biological applications. This study will not only provide concrete and validated interacting relations of gut microbiota, but also demonstrate the connection between theoretical network model and biological applications on animal system.. 36.

(37) 2 Materials and Methods In this chapter, I introduce the treefrog samples used in this study, including wild treefrogs collected in different seasons and lab-reared treefrogs that were used to manipulate artificial hibernation for the GMT datasets. All fecal samples of treefrogs were used to generate 16S rRNA amplicon for three different analyses. These three different analyses based on 16S rRNA amplicon datasets are used to investigate three different research purposes on treefrogs accordingly. The first dataset describes seasonal change of microbial community in the wild treefrogs. It reveals the compositional and functional alterations between artificial hibernation and nonhibernating seasons. The compositional identification and functional prediction will be introduced in this chapter. The second dataset describes the growth microbiome in a continuous time-series fashion and is used as input into network inference modelling. This dataset is carefully designed to refine and optimize the time period to track the growing progress of gut microbiome. The third dataset describes the microbial profile after oral administration of probiotics. It facilitates the validations of the inferred interacting network. All 16S rRNA amplicon datasets were followed well-developed pipeline that currently used to study community profiles of animal gut microbiome.. 2.1 Animal individuals 2.1.1 Treefrogs from season fall, winter, and spring The population of treefrog P. megacephalus is widely distributed in Taiwan, and the individuals we collected aim at two cities that have been reported to be dominant in the field. A total of 36 treefrog individuals were collected from Wazihwei Nature Reserve (WNR) (121.41432° E, 25.16775° N) in Bali (Dist., New Taipei City, Taiwan) and private botanic gardens (PBG) (120.31423° E, 23.53302° N) in Tianwei (Township, Changhua County, Taiwan). To understand hibernation effects on gut microbiota, the seasons for sample collections include winter, as well as non37.

(38) hibernating seasons fall and spring. The sample size, average snout-vent length, and body mass were shown in Table 1. 12, 18, and six individuals were collected in fall (October and November 2013), winter (December 2013, January, and February 2014), and spring (March 2014), respectively. There were three individual replicates collected in each month at each sampling city. The average of snout-vent length (SVL) in fall, winter, and spring is 5.7 ± 0.6 centimeter (cm), 5.6 ± 1.3 cm, and 6.5 ± 1.0 cm, respectively. The body mass in fall, winter, and spring in average is 12.2 ± 4.4 gram (g), 11.8 ± 6.3 g, and 18.7 ± 13.1 g. All wild treefrogs after capture were euthanized immediately and stored in -80°C before fecal collection and DNA extraction. All protocols of this study and animal use were reviewed and approved by the Academia Sinica Institutional Animal Care and Utilization Committee (Approved No. BSF0413-00002859).. 2.1.2 AH treefrogs in the laboratory Due to limited research about hibernation records for amphibians in Taiwan, I set up a lab-reared conditions for treefrogs to induce artificial hibernation. The candidate treefrogs that will artificially hibernate in the lab were firstly collected in the wild from WNR and PBG during June 2014 to September 2014. All captured treefrogs were at least acclimated for three months under the normal treefrog-reared system in the lab of Academia Sinica. The normal treefrog-reared system contains several 240liter glass tanks and maintained in a temperature controlled at 23°C within a variation of 2°C. All treefrogs were exposed a 8 h:16 h light:dark cycle. In order to mimic the major diet of treefrogs in the field, I used the same taxonomic category of insect, Turkestan cockroach nymphs (Finke 2013), as a good source of protein. Turkestan cockroach nymphs were fed to the treefrogs at a quantity of ~3% of treefrog biomass twice a week as suggested in the previous study (Zhu, Chen et al. 2015). To stimulate artificial hibernation, we modified an artificial hibernating program previously described in leopard frogs (Rana pipiens) (Gossling, Loesche et al. 1982). 38.