Résumé

Les protozoaires sont les prédateurs des bactéries, mais certaines évitent la digestion une fois ingérées. Certaines bactéries résistantes peuvent être enrobées dans les corps fécaux des protozoaires, les protégeant de certains stress. Selon la bactérie et le protozoaire impliqués, les corps fécaux présentent différentes morphologies. Nous avons évalué comment certaines caractéristiques bactériennes (forme, taille, Gram, hydrophobicité) influencent cette morphologie. La résistance à la digestion par les ciliés Tetrahymena pyriformis et T. thermophila de quatre bactéries non pathogènes, Microbacterium oxydans, Micrococcus luteus, Cupriavidus sp. et Cellulosimicrobium funkei, a été étudiée. Pour chaque souche, les corps fécaux avaient une morphologie différente. Parmi les caractéristiques examinées, la taille et l’hydrophobicité semblaient influencer davantage la morphologie, car l’ingestion de petites bactéries hydrophobes conduisait à des corps fécaux plus ronds, denses et réguliers. Ces résultats indiquent que l’enrobage de bactéries n’est pas restreint aux bactéries pathogènes et que les interactions bactéries- protozoaires sont uniques et complexes.

2.2 Abstract

Objective: Protozoa are natural predators of bacteria, but some bacteria can evade

digestion once phagocytosed. Some of these resistant bacteria can be packaged in fecal pellets produced by protozoa, protecting them from physical stresses and biocides.

Depending on the bacteria and protozoa involved in the packaging process, pellets can have different morphologies. In the present descriptive study, we evaluated how some bacterial characteristics (shape, size, Gram staining, hydrophobicity) could influence pellet morphology.

Results: To assess this, four non-pathogenic bacteria, Microbacterium oxydans,

Micrococcus luteus, Cupriavidus sp. and Cellulosimicrobium funkei, were studied for their capacity to resist digestion and be packaged by the ciliates Tetrahymena pyriformis and T. thermophila. A different pellet morphology was obtained for each bacterial strain studied. Of the bacterial characteristics chosen, size and surface hydrophobicity seemed to influence pellet morphology the most, as the ingestion of small, hydrophobic bacteria resulted in rounder, denser and more regular pellets. These results support the idea that bacteria packaging is not restricted to pathogenic species and that bacteria-protozoa interactions, including those leading to bacteria packaging, are unique and complex.

2.3 Introduction

A vast number of protozoa, which are ubiquitous unicellular eukaryotes, prey on bacteria. Some bacteria can resist predation by protozoa and survive digestion in the phagocytic pathway (Greub and Raoult, 2004). This can result in bacteria being included in fecal pellets expelled by protozoa, a phenomenon known as bacteria packaging. This forms a typically spherical cluster of bacteria, usually covered by one or more layers of membrane, depending on both bacterial and protozoan species (Berk et al., 2008; Denoncourt et al., 2014; Paquet and Charette, 2016). Packaging of different bacterial species by the same protozoan can result in different pellet morphologies; pellets can include very little or abundant membrane layers surrounding bacteria (Denoncourt et al., 2017b; Trigui et al., 2016). Pellet morphology also differs between protozoan species as those produced by amoebae tend to contain fewer bacteria and more membrane than ciliate-produced ones. (Berk et al., 1998; Berk et al., 2008).

Bacteria packaging, which grants resistance to chemical and physical stresses, has mostly been studied with pathogens so far (Brandl et al., 2005; Espinoza-Vergara et al., 2019; Gourabathini et al., 2008; Koubar et al., 2011; Raghu Nadhanan and Thomas, 2014; Trigui et al., 2016). However, non-pathogenic bacteria have been shown to be packaged by the social amoeba Dictyostelium discoideum (Paquet and Charette, 2016).

Tetrahymena ciliates are useful organisms to study bacteria packaging, as pellets can be expelled within less than an hour (Denoncourt et al., 2017a). With the exception of M. smegmatis packaging by Tetrahymena (Denoncourt et al., 2017b), no extensive study of non-pathogenic bacteria packaging by ciliates is available. Consequently, there is little evidence that ciliates can package non-pathogenic bacteria like they do pathogenic ones. Bacterial packaging could promote the survival of bacteria in harsh environments or act as a reservoir, as pellets can stay intact for weeks (Espinoza-Vergara et al., 2019; Koubar et al., 2011).

In the present study, four non-pathogenic bacterial species resistant to digestion in amoeba (Paquet and Charette, 2016) were co-cultured with two ciliates, Tetrahymena pyriformis and T. thermophila. The pellets produced were characterized with regards to bacterial features.

2.4 Materials and methods

Strains. Tetrahymena strains used were T. pyriformis ATCC 30202, grown axenically in SPP medium at 25°C, and T. thermophila CU428.2, grown axenically in PP medium at 30 °C (Gorovsky et al., 1975; Orias et al., 2000). Cells were subcultured when reaching confluence. Bacterial stock cultures were grown as needed on Tryptic Soy Agar (TSA) (EMD, Canada) and incubated for 48h, at the appropriate temperature. Strains used were Microbacterium oxydans US1 (30 °C), Micrococcus luteus (37 °C), Cellulosimicrobium funkei (30°C), and Cupriavidus sp. (25 °C) (see Additional file 1 for details).

Hydrophobicity assay. The surface hydrophobicity assay was conducted as described in Rosenberg et al. and results were analyzed as presented in Bonifait et al. (Bonifait et al., 2010; Rosenberg et al., 1980). The assay was conducted 5 times for each species.

Co-cultures. Co-culture assays were conducted in 6-well plates. Bacteria were suspended in plate count broth (PCB) (Berk et al., 2008; Brandl et al., 2005; Gourabathini et al., 2008; Trigui et al., 2016) at an OD595 of 1. Tetrahymena cells were transferred from their growth

medium to PCB medium. 3 x 105 ciliates were added to each co-culture (one strain per co-

culture) along 300 µl of one of the bacterial suspensions, for a ratio of 1 protozoa:1000 bacteria in a total volume of 3 ml of PCB. Co-cultures were incubated for 4h, at 25 °C with T. pyriformis and at 30 °C with T. thermophila.

Cell fixation and 4-,6-diamidino-2-phenylindole (DAPI) staining. This procedure was performed as previously described (Trigui et al., 2016). Once stained, samples were

observed with an Axio Observer Z1 microscope equipped with an Axiocam MRm camera (Zeiss).

LIVE/DEAD staining. 1 ml of co-culture was centrifuged for 2 min at 600 x g. The supernatant containing the fecal pellets was collected. Staining with LIVE/DEAD bacterial viability stain (Life Technologies, Invitrogen) was done as previously described (Trigui et al., 2016). LIVE/DEAD staining was reproduced twice for each co-culture.

Transmission electron microscopy (TEM) processing. Samples from co-cultures, some incubated longer than 4h to ensure optimal pellet retrieval, were fixed for 3 h in 0.1 M sodium cacodylate buffer (pH 7.3) containing 2% glutaraldehyde and 0.3% osmium tetroxide. After centrifugation, sample pellets were resuspended with a micropipette in cooled molecular biology agarose (BioRad) diluted to 3% in phosphate buffered saline. Once solidified, samples were cut into cubes of about 3-5 mm and processed for TEM as previously described (Paquet et al., 2013).