Session 6 – Moderated by Süleyman Yıldırım

Ryan Kerney

09.00- 09.45

Egg capsule microbial symbionts attenuate salamander tail tip regeneration

The amphibian egg capsule is unique in harbouring symbiotic prokaryotic and eukaryotic microbiota. We have been exploring the diversity of this microbiota in Ambystoma maculatum through amplicon sequencing and culturing, and testing the potential benefits of these microbes through selective co-culturing with sterilized embryonic hosts. By modifying a tail tip regeneration assay developed by the Voss lab at the University of Kentucky, we have found that none of the cultured bacteria affect rates of tail tip regeneration. However, the mutualistic algae Oophila amblystomatis does increase the rate of tail tip regeneration in controlled trials. We are exploring the mechanistic basis of this increase by modifying pO2 levels and a chemical fraction screen of Oophila-derived metabolites.


Maria del Rosario Sanchez-Gonzalez

09.45-10.05

Axolotl, a new animal model for a “leaky” blood brain barrier

The blood brain barrier (BBB) represents a physical interface that tightly regulates the trafficking of molecules between the blood and the neural tissue, thereby maintaining the physiological homeostasis of the brain. Little is known about how regenerative-competent vertebrates such as amphibians establish the BBB, in particular in view of the lack of protoplasmic astrocytes in these animals, a key cell type regulating BBB permeability in mammals. Here, we analyzed the BBB permeability in the brain of Xenopus laevis and Ambystoma mexicanum (Axolotl). Surprisingly, we observed remarkable differences between the two species. While the BBB was similarly tight for 1kDa molecules in Xenopus as in zebrafish and mammals, Axolotl showed a complete leakiness for the 1kDa tracer and an increased endothelial transcytosis, despite the absence of any obvious neurological deficits. This suggests that Axolotl might have implemented specific mechanisms to protect the brain from detrimental consequences of “leaky blood vessels”.


Kıvanç Kök

10.05-10.20

Mining axolotl microbiome: Novel insights into axolotl limb regeneration by using machine learning

Aim: The axolotl (Ambystoma mexicanum) is a common vertebrate model organism in regeneration research. Previously, Demircan et al. (2019) published research on axolotl limb regeneration by monitoring the temporal microbiome dynamics of this process. Interestingly, the three main phases of axolotl limb regeneration were reflected in the uncovered longitudinal microbiome profile. The present study followed up on this observation by mining the resulting ASV abundance dataset. The main purpose was to examine in more detail variation in microbiome structure and detect elusive microbial aspects.

Material and methods: The aforementioned microbiome dataset was retrieved from a dedicated repository and bioinformatically analyzed. Hereby, data mining was performed using the following consecutive steps: explanatory data analysis (EDA), unsupervised ML (including hierarchical clustering and dimension reduction) and supervised ML (using a Random Forest classifier).

Results: The EDA and Unsupervised ML steps revealed grouping of samples according to the respective regeneration phases. Interestingly, the combined analysis of alpha and beta diversity delineated a hitherto hidden overarching microbiome pattern, characterized by concurrent changes in the microbial community structure in the course of axolotl limb regeneration, namely simultaneous microbiome reshaping and alpha diversity reduction. The separation of samples based on regeneration phase was readily evident and statistically significant (p<0.05). Concordantly, the Random Forest-based classification model successfully predicted the phase labels in the subsequent supervised ML step. Importantly, the classification highlighted a distinct list of bacterial taxa as important features (predictors), which can be considered as new biomarker candidates for distinguishing between the phases.

Conclusions: In this work, we evaluated patterns in the microbime variation in the context of axolotl limb regeneration using a machine learning approach. Furthemore, we identified novel potential discerning microbial biomarkers of regeneration phase. Overall, the insights gained from the microbime data minig complement available experimental results and contributes to the progress in this research field.