Michael Roggenbuck:
Characterization and manipulation of the reticulated microbiome in vertebrates

Date: 15-09-2014    Supervisor: Søren J. Sørensen

The term microbiome - “The ecological community of commensal, symbiotic, and pathogenic microorganisms that literally share our body space” - was first described by Professor Joshua Lederberg of the Rockefeller University. With the beginning of the golden age of High-throughput- Sequencing, is has become more evident that animals and their microbial communities are metabolically and immunologically tightly connected and highly dependent on each other. Today the complex microbial flora is often considered as an organ – with a healthy and a diseased stage. Currently the human microbiome is most intense evaluated. However, mechanistically questions often cannot be studied in humans, therefor animal research is applied.

In the first part of this thesis, the diet intervention on the “total” microbial community of two animal model organisms – mice and lambs - was characterized using 16S rRNA gene amplicon sequencing.

We aimed to co-correlate the lung microbial composition of vitamin D supplemented and depleted diet in mice. Vitamin D deficiency has been recognized for its role in chronic allergic diseases, but a link between vitamin D status and the development of asthma has not been established. We focused our study on the lung sampling from bronchoalveolar lavage (BAL) fluids and lung tissue. The BAL fluids are commonly used to determine inflammatory responses and/or the diseased stage of the lung. The microbial flora of the lower respiratory tract in mice has never been described due to the low microbial DNA yield and the high risk of contamination from the upper respiratory tract during sampling – therefor we evaluated different experimental approaches (Manuscript 1 - published). We consistently amplified microbial DNA from the lung that clusters in bacterial composition apart from other parts of the body, such as the mice intestine. In the second experiment we exposed the mice to vitamin D and ovalbumin (OVA) – a common model to induce asthmaallergic responses in the lung. We investigated the variation between the OVA sensitized and naïve mice under vitamin D supplemented and depleted conditions (Manuscript 2 – in preparation). We found that the lung microbial floral composition is significantly affected by the OVA treatment, but responses inconsistent to vitamin D exposure.

The second animal-microbiome model focused on the foregut of young ruminants (lambs). The livestock methane production is a substantial part of the annual anthropogenic greenhouse gas production. There is a large interest in reducing the methane emission since the worldwide meat consumption will double until 2050. Interestingly calves and lambs emit little methane in the early phase of live since the major foregut methanogenesis in herbivores is associated to fiber rich diet of adult ruminants. Therefor we designed a “colostrum”-like diet (maternally produced fat-rich newborn feed) and changed the development of the rumen in lambs from birth until the age of 6 months. The diet reduced the methane production of 87% compared to the hay fed control group and resulted in the development of a rudimentary rumen (Manuscript 3 - published). The fat-rich diet significantly reduced the relative counts of methane producing bacteria (methanogens) in the ruminal fluid and decreased the methanogenic diversity. However the diet manipulation increased the frequency of a single methanogenic species (Methanobrevibacter sp.) in the ruminal solid fraction compared to the control group (Hay fed). Additionally we evaluated the complex microbial community of the undeveloped rumen system. Alternative hydrogen oxidations pathways (alternatives to methanogenesis) were predicted based on significantly elevated species previously described by cultivation studies. Finally we estimated the interactions of Methanobrevibacter sp. with static co-correlation and co-exclusion models (Manuscript 4 – in preparation). The second part of this thesis was of a more descriptive nature. Wild animals can highlight new microbial-host relations with health or environmental impact.

Vultures are carrion feeders. As forensic studies revealed, vultures can wait up to 48 hours of decay prior preying, possibly because the advanced decomposition softens the tissue and easies the birds to feed large carrions. However, feeding on carcass is associated to the potential exposure of bacterial toxins, with severe outcome such as avian botulism. As Ley et al (2008) described – diet is the key driver of microbial composition of the digestive system in vertebrates. Therefore we raised the question of what microbes would be found in the vulture gut community. We observed that 50 wild vulture hindgut samples (large intestine), of the species Coragyps atratus and Cathartes aura. Their hindgut microbiome were highly similar and dominated by Clostridium sp (putative C. perfringens) and Fusobacterium sp (putative F. nucleatum) with a low species diversity (Manuscript 5 - submitted). We speculated that the vulture digestive system has co-evolved with these two microbes. To address this question we compared our results from the wild with birds in captivity from the Copenhagen zoo. We evaluated zoo vultures as well as carnivore, herbivore and omnivore birds. The carnivorous birds, including the vultures, were fed fresh meet. However, despite of the similar diet between bird types, Zoo-vulture feces contained a more statistically similar microbial composition to wild vultures, than compared to other fresh meat fed carnivorous birds. In this study we propose for first time the term carrionovore – as “animals that prey on carcasses”.

As part of a larger rumen screening approach we have also analyzed the microbial flora of giraffes (Manuscript 6 - accepted) and wallabies (marsupial) foreguts (in progress of analysis). Giraffes are herbivores that prefer leaves and fruits with higher crude protein and lignin content compared to the common livestock animals of cattle and sheep fed with hay. The purpose of this study was to uncover new microbes previously undescribed. Our results reveal that the major part of the giraffe rumen environment contains mostly sequences not assigned to any known genera.