Explorative metabarcoding of Abies balsamea L. Mill. endomycobiota
|Advisor:||Bernier, Louis; Bérubé, Jean|
|Abstract:||As for the studies of other members of the plant microbiota, fungal endophytology has vastly benefited from the development of High Throughput Sequencing techniques in the late 2000s. This technological progress has notably allowed for the popularization of metabarcoding, i.e. a DNA-based approach to identify biodiversity components from environmental samples and study the community composition and distribution. The massive production of data, and the standardization in the sample preparations associated with such methods, have deeply modified the perception of the extent of the fungal biodiversity. Yet fungal endophytology precepts remain largely inherited from culture-dependent methods which have been shown to yield a more fractioned portion of the biodiversity than the molecular-based approach, as many fungi are not amenable to standard culturing. HTS techniques are not without drawbacks either as they tend to inflate the biodiversity estimates even with state of the art analysis. The main goals of this thesis were first to develop a more rigorous approach to analyse data obtained from 454 pyrosequencing, one of the original HTS techniques, in order to estimate conservatively the biodiversity; and then to develop a better understanding of the structure of forest trees endomycobiota and challenge earlier conclusions based on culture-dependent methods. Inflation of the biodiversity is mostly due to remaining undetected erroneous sequences partially forming the large number of singletons and doubletons generally observed with HTS based studies. Three sources of error are significant: PCR chimeras, PCR single base substitutions, and sequencing error. Here we hypothesized that the selection of a sub-region of the fungal barcode displaying particular characteristics might, if not formally assess erroneous sequences as such, at least limit their impact on the estimation of the diversity. We thus considered a fragment composed of the partial ribosomal small sub-unit immediately following the ITS1F primer in addition of the ITS1 sub-locus (pSSU-ITS1). We showed that basing the analysis on the pSSU-ITS1 fragment enhances the sensitivity of chimera detection. As PCR single base substitutions and sequencing errors remain rare events, spurious sequences are rare too and somewhat similar to true abundant sequences. We hypothesized that the presence of the pSSU, whose variability is lower than that of the ITS1 sub-locus, might buffer these errors. Putative rare spurious sequences were grouped with the true abundant sequences they deviated from, thus reducing the proportion of singletons and doubletons. We then developed an approach to readily extract this pSSU-ITS1 fragment from fungal ITS amplicons. We observed from the endomycobiota of a single balsam fir that we produced to test our data treatment in the first chapter that considering the pSSU-ITS1 fragment did not alter the conclusions on the structure of the fungal endophytic community from ITS1 analysis. While it has to be considered with appropriate reservations due to the limited sampling, we also estimated, for the first time to the best of our knowledge, the extent of the fungal endophyte biodiversity harboured by a single tree at a precise time with an extrapolation of 2 536 ± 73 mOTUs. In the second chapter on the endomycobiota present in the different tissue types of balsam fir branches, we confirm that some tissue specificity is exhibited by fungal endophytes as our results suggest that the aerial endomycobiota of balsam fir trees might be fractioned in distinct communities depending on the tissue types. Finally, in the third chapter, we reveal that the mechanisms of colonization of the host plant by fungal endophytes might be more complex and dynamic that the suggested passive accumulation hinted by culture-dependent methods. The last four cohorts of needles from balsam fir sampled displayed relatively similar diversities, but harboured distinct communities.|
|Document Type:||Thèse de doctorat|
|Open Access Date:||2 October 2019|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.