Tandem mass spectrometry-based proteotyping allows characterizing microorganisms in terms of taxonomy and is becoming an important tool for investigating microbial diversity from several ecosystems. Fast and automatable sample preparation for obtaining peptide pools amenable to tandem mass spectrometry is necessary for enabling proteotyping as a high-throughput method. First, the protocol to increase the yield of lysis of several representative bacterial and eukaryotic microorganisms was optimized by using a long and drastic bead-beating setting with 0.1 mm silica beads, 0.1 and 0.5 mm glass beads, in presence of detergents. Then, three different methods to obtain greater digestion yield from these extracts were tested and optimized for improve efficiency and reduce application time: denaturing electrophoresis of proteins and in-gel proteolysis, suspension-trapping filter-based approach (S-Trap) and, solid-phase-enhanced sample preparation named SP3. The latter method outperforms the other two in terms of speed and delivers also more peptides and proteins than with the in-gel proteolysis (2.2 fold for both) and S-trap approaches (1.3 and 1.2 fold, respectively). Thus, SP3 directly improves tandem mass spectrometry proteotyping.

Several bacteria are able to degrade the major industrial solvent dichloromethane (DCM) by using the conserved dehalogenase DcmA, the only system for DCM degradation characterised at the sequence level so far. Using differential proteomics, we rapidly identified key determinants of DCM degradation for Hyphomicrobium sp. MC8b, an unsequenced facultative methylotrophic DCM-degrading strain. For this, we designed a pan-proteomics database comprising the annotated genome sequences of 13 distinct Hyphomicrobium strains. Compared to growth with methanol, growth with DCM induces drastic changes in the proteome of strain MC8b. Dichloromethane dehalogenase DcmA was detected by differential pan-proteomics, but only with poor sequence coverage, suggesting atypical characteristics of the DCM dehalogenation system in this strain. More peptides were assigned to DcmA by error-tolerant search, warranting subsequent sequencing of the genome of strain MC8b, which revealed a highly divergent set of dcm genes in this strain. This suggests that the dcm enzymatic system is less strongly conserved than previously believed, and that substantial molecular evolution of dcm genes has occurred beyond their horizontal transfer in the bacterial domain. Our study showed the power of pan-proteomics for quick characterization of new strains belonging to branches of the Tree of Life that are densely genome-sequenced.

The microbial diversity encompassed by the environmental biosphere is largely unexplored, although it represents an extensive source of new knowledge and potentially of novel enzymatic catalysts for biotechnological applications. To determine the taxonomy of microorganisms, proteotyping by tandem mass spectrometry has proved its efficiency. Its latest extension, phylopeptidomics, adds a biomass quantitation perspective for mixtures of microorganisms. Here, we present an application of phylopeptidomics to rapidly and sensitively screen microorganisms sampled from an industrial environment, i.e., a pool where radioactive material is stored. The power of this methodology is demonstrated through the identification of both prokaryotes and eukaryotes, whether as pure isolates or present as mixtures or consortia. In this study, we established accurate taxonomical identification of environmental prokaryotes belonging to the Actinobacteria, Bacteroidetes, Firmicutes, and Proteobacteria phyla, as well as eukaryotes from the Ascomycota phylum. The results presented illustrate the potential of tandem mass spectrometry proteotyping, in particular phylopeptidomics, to screen for and rapidly identify microorganisms.

Microbial life can thrive in the most inhospitable places, such as nuclear facilities with high levels of ionizing radiation. Using direct meta-analyses, we have previously highlighted the presence of bacteria belonging to twenty-five different genera in the highly radioactive water of the cooling pool of an operating nuclear reactor core. In the present study, we further characterize this specific environment by isolating and identifying some of these microorganisms and assessing their radiotolerance and their ability to decontaminate uranium. This metal is one of the major radioactive contaminants of anthropogenic origin in the environment due to the nuclear and mining industries and agricultural practices. The microorganisms isolated when sampling was performed during the reactor operation consisted mainly of Actinobacteria and Firmicutes, whereas Proteobacteria were dominant when sampling was performed during the reactor shutdown. We investigated their tolerance to gamma radiation under different conditions. Most of the bacterial strains studied were able to survive 200 Gy irradiation. Some were even able to withstand 1 kGy, with four of them showing more than 10% survival at this dose. We also assessed their uranium uptake capacity. Seven strains were able to remove almost all the uranium from a 5 µM solution. Four strains displayed high efficiency in decontaminating a 50 µM uranium solution, demonstrating promising potential for use in bioremediation processes in environments contaminated by radionuclides.

La biosphère microbienne constitue une source importante de catalyseurs biologiques, dont l’identification et la caractérisation restent à ce jour limitées. La phylopeptidomique est une approche de métaprotéomique innovante pour l’identification taxonomique sans a priori des micro-organismes procaryotes et eucaryotes. Cette thèse a pour objectif d’optimiser la méthodologie à des fins de criblage d’isolats microbiens, d’étude rapide de leur potentialité, et de sélection des variants les plus intéressants. Les améliorations technologiques portent sur l’optimisation de la préparation d’échantillon, la préparation simultanée de 96 échantillons, la diminution du temps d’analyse LC-MS/MS et le traitement informatique des données. La sensibilité et la rapidité de l’approche optimisée ont été exemplifiées avec l’identification au niveau taxonomique espèce d’un nombre important d’isolats issus de la plaque dentaire et d’eaux prélevées dans des installations industrielles. La phylopeptidomique permet d’identifier : i) les micro-organismes procaryotes et eucaryotes, tel que démontré sur l’analyse de diverses bactéries isolées ou en mélanges et l’identification des champignons eucaryotes Cordyceps confrogosa et Cladosporium herbarum ; ii) les bactéries cliniques avec une identification plus précise au rang taxonomique de l’espèce comparé à d’autres méthodes d’analyses telles que le MALDI-TOF, notamment pour certaines espèces opportunistes telles que Serratia marcescens et Stenotrophomonas maltophilia et ; iii) les espèces issues d’environnements peu étudiés, dont les souches de référence ne sont pas encore séquencées mais pour lesquelles des génomes d’organismes très proches sont disponibles dans la base de données et permettent d’établir la phylogénie de la souche d’intérêt jusqu’à l’espèce la plus proche. Une approche originale de pan-protéomique a aussi été conçue pour la caractérisation des voies métaboliques d’un micro-organisme dont le génome n’est pas encore connu. L’analyse protéomique de la bactérie Hyphomicrobium sp. MC8b à partir d’une base de données qui regroupe les informations génomiques de l’ensemble des souches proches référencées a permis d’identifier les protéines impliquées dans la dégradation du dichlorométhane et notamment reconstituer la séquence probable de la dichlorométhane déhalogénase. L’intérêt de la phylopeptidomique pour le criblage de micro-organismes environnementaux à grande échelle a été démontré par le traitement de 96 échantillons en 55 heures et leur identification taxonomique précise et sans équivoque que ce soit pour de nouveaux isolats de bactéries, de champignons, voire des mélanges de microorganismes.