«The Ecology of Bacterial Individuality ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van ...»
The Ecology of Bacterial Individuality
ter verkrijging van de graad Doctor aan
de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus
prof.dr. L.M. Bouter,
in het openbaar te verdedigen
ten overstaan van de promotiecommissie
van de faculteit der Aard- en Levenswetenschappen
op dinsdag 27 maart 2012 om 13.45 uur
in de aula van de universiteit,
De Boelelaan 1105 door Mitja Nandi Paul Remus-Emsermann geboren te Siegburg, Duitsland promotoren: prof.dr. G.A. Kowalchuk prof.dr. J.H.J. Leveau The Ecology of Bacterial Individuality Mitja Nandi Paul Remus-Emsermann Ph.D. Thesis The research presented in this thesis was carried out at the Department of Microbial Ecology at the Netherlands Institute of Ecology (NIOO-KNAW), The Netherlands.
The research was financially supported by a vidi-grant from the Netherlands Research Foundation awarded to JHJ Leveau.
Production of the thesis was partly financed by the Netherlands Institute of Ecology (NIOO- KNAW).
Printed by: GVO printers & designers B.V. | Ponsen & Looijen Ede, The Netherlands ISBN: 978-90-6464-538-9 Table of Contents Page Introduction 1 Chapter 1 - Linking environmental heterogeneity and repro- 19 ductive success at single-cell resolution Chapter 2 - Variation in local carrying capacity and the in- 33 dividual fate of bacterial colonizers in the phyllosphere Chapter 3 - Single-cell experience of bacterial immigrants to 49 pre-colonized leaf surfaces Chapter 4 - Quantification of lateral heterogeneity in carbohy- 59 drate permeability of isolated plant leaf cuticles Chapter 5 - ASiMoPh – Agent-based Simulation of Microbial 71 Phyllosphere Colonization Chapter 6 - Draft sequence and partial genome annotation of 83 the phyllosphere model bacterium Erwinia herbicola strain 299R General discussion and synthesis 87 References 95 Appendix 109 About the author 119 Acknowledgements 121 Summary 125 Samenvatting 129 1 General Introduction General Introduction It is long known that bacteria exhibit phenotypical heterogeneity within clonal popu- lations and that heterogeneous environments have an impact on bacterial individuals that inhabit it. Nevertheless microbiologists tend to investigate bacteria as averages of populations: Be it by observing the average phenotypic properties of a population such as colony morphologies on an agar plates or the growth rate in a shaken liquid culture. A classical approach in microbiology is to determine the total number of colony forming units that can be recovered from an environment to evaluate the capability of the population of a bacterium to grow and survive in the environment.
More recently, scientists turned to molecular tools to determine bacterial responses to an environment on a molecular level such as gene-expression of a population.
The concept of bacterial individuality is a recent development in microbiology. In essence it is built around the ideas that bacteria should be treated as individuals and investigated at scales that actually matter for them. In this thesis I followed these ideas and developed methods that show the value of the information that can be extracted from single-cell observations and how they increase our knowledge about the ecology of bacteria.
As the study environment, I used the leaf surface, or phyllosphere, as a natural system for bacterial colonization. By using bacterial bioreporter technology and fluorescence microscopy coupled with image cytometry I was aiming to determine colonization success and the impact of chance on colonization processes. In another bioreporter-based approach I aimed to determine local differences in fructose permeance through isolated cuticles. I then used the data derived from the previous studies to formulate a spatial-explicit, agent-based model for phyllosphere colonization.
A brief introduction to bacterial ecology The term ecology describes the study of the distribution, abundance, and interaction of organisms with their biotic and abiotic environment (Begon, Harper et al. 1996).
Bacterial ecology deals with questions on what and how Earth’s ecosystems are inhabited and influenced by bacteria. Some examples for bacterial ecosystems are: biofilms in aquatic sediments (Torsvik, Sorheim et al. 1996), oceans (Gugliandolo and Maugeri 1993), the gastro intestinal tract of animals (Brune 1998; Zoetendal, Collier et al. 2004), soil (Torsvik, Sorheim et al. 1996)and plants (Andrews and Harris 2000;
Lindow and Brandl 2003; Leveau 2006). Additionally, a variety of anthropogenic types of bacterial ecosystems exist, e.g. wastewater treatment plants (Kapley, De Baere et al. 2007), vinegar or acetic acid bioreactors (Sueki, Kobayashi et al. 1991), glutamic acid bioreactors (Hermann 2003), and silage (Denoncourt, Caillet et al.
2 The Ecology of Bacterial Individuality 2007). It is generally accepted that bacteria colonize and influence nearly all habitats on Earth (Whitman, Coleman et al. 1998).
Introduction of plants as a habitat for bacteria As the experimental model system in this thesis involves plant-associated bacteria, a more detailed introduction into plants as a habitat for bacteria will be given in the following section. Plants are common habitats for microorganisms. Their roots and leaves are naturally colonized by a broad spectrum of bacteria, fungi and protozoa (Andrews and Harris 2000), which have a variety of interactions with the host plant ranging from mutualism to commensalism and pathogenicity (Singh, Millard et al. 2004). Especially the root system, the so-called rhizosphere, is colonized very densely. One gram of forest soil may typically contain up to 109 bacterial cells (Whitman, Coleman et al. 1998), and bacterial densities in the rhizosphere may be one to two orders of magnitude greater (Liljeroth and Baath 1988; Duineveld and Van Veen 1999; Jaeger, Lindow et al. 1999). Compared to underground organs, aerial parts of plants are less densely inhabited (Holm and Jensen 1972). Leaf surfaces, for example, feature limited availability of water and nutrition, increased exposure to UV-radiation, and the waxy cuticle covering the leaf is a substrate that is very hard to penetrate (Kolattukudy 1985) and offers few niches to populate (Monier and Lindow 2004) ¤7. Also, several plants posses a self-cleaning effect, which prevents the successful colonization of leaves (Barthlott and Neinhuis 1997). Therefore, only specialized microorganisms can manage to survive the conditions on leaf surfaces.
Leaves are typically inhabited by 106 -107 bacterial cells per cm2 or up to 108 bacteria cells per gram of leaves (Lindow and Brandl 2003), and the cumulative number of bacteria on plant leaves world-wide has been estimated to add up to the astonishing number of 1026 (Morris and Kinkel 2002). Most of these so-called epiphytic bacteria have developed strategies to adhere to their substrate: some synthesize and secrete extracellular polymers, whereas others possess pili to facilitate adhesion and prevent removal (Beattie and Lindow 1999). Some bacteria can also influence the substrate structure itself, e.g. by producing indole-3-acetic acid (IAA), a phytohormone. In high doses, bacterially produced IAA is thought to cause russet on pear fruits (Lindow, Desurmont et al. 1998), which result from an increased division activity of plant cells. These cells act as sinks for nutrients from which the bacteria might benefit. Also, it is known that bacteria are able to increase the permeability of leaf cuticles (Schreiber, Krimm et al. 2005), thereby increasing their access to plant-derived nutrients.
Leaves are not colonized in a uniform manner. The most frequently colonized strucGeneral Introduction tures are the bases of trichomes, glandular cells, leaf veins, epidermal cell wall junctions, hydathodes, wounds, and stomata (Beattie and Lindow 1999; Monier and Lindow 2004). Leaf permeance for nutrients, mostly photoassimilates such as fructose, glucose and sucrose, can differ between these features (Schlegel, Schönherr et al.
2005; Schreiber 2005). This is in part due to the heterogeneous compositions and thickness of the cuticle, which overlays the epidermis of plant leaves (Fernández, Khayet et al. 2011). The cuticle has also been shown to contain locally clustered hydrophilic pores, which allow higher rates of diffusion for hydrophilic substances through the cuticle (Schönherr and Schreiber 2004; Schlegel, Schönherr et al. 2005;
Schreiber 2005). Water droplets can adhere longer at trichomes, glandular cells, leaf veins and other structures on the leaf and act as a sink for nutrients that leach out of the leaf (Linskens 1950). In addition, the nutrient flow in leaf veins is much higher than in the rest of the leaf, resulting in a steeper gradient, increasing the driving force with which nutrients permeate across the cuticle (Schreiber and Schönherr 2009).
Differences in leaf geography are also thought to contribute to differential adherence and protection of bacteria (Monier and Lindow 2004). As a result, the distribution of cells on a single leaf surface can be very heterogeneous and can differ by 100 fold even between segments of 9 mm2 (Kinkel, Wilson et al. 1995; Monier and Lindow 2004).
Leaves are colonized by airborne, waterborne, or biotic vector mediated (e.g. insects) events or during the emergence of the shoot through the soil (Leveau 2006).
Usually, bacteria initially arrive on leaves as single cells or cell aggregates. These colonizers may encounter a variety of conditions, depending on their particular location. In laboratory experiments with bean plants, it was found that uncolonized leaf surfaces carry sufficient amounts of sugar (0.2 to 10 µg; the majority consisting of sucrose, fructose and glucose) to sustain a population of 1.7 x 107 bacteria per gram (Mercier and Lindow 2000). In another study, it was shown that most bacterial immigrants to bean leaves encounter nutrient conditions that allow them to divide several times, but after this initial phase, sugar is only available in a limited number of sites on the leaf (Leveau and Lindow 2001). Bacteria that have not already colonized these sinks likely starve and die off unless they have specialized adaptations to this environment, i.e. high tolerance to desiccation, starvation, UV-radiation, or adaptations to avoidance strategies these stresses altogether, e.g. by entering the leaf (Beattie and Lindow 1999).
Studying the ecology of bacterial communities The study of bacteria in their natural environment involves experimental techniques 4 The Ecology of Bacterial Individuality that typically have been developed for the study of bacteria in the laboratory, e.g.
cultures grown in flasks or colonies and biofilms grown on agar plates. Many of these techniques are population-based measurements. Some of these techniques will be discussed here in more detail, using the phyllosphere as a model bacterial ecosystem.
A basic but valuable method to determine the number of bacteria on a leaf, as well as to investigate the general composition of the bacterial phyllosphere community, is plate counting. Leaves of known size or weight are washed in a known volume of a wash buffer, and samples of this buffer are spread onto a defined growth medium.
After a sufficient time of incubation, colony forming units (CFUs) can be counted and used to estimate the bacterial numbers per unit of leaf surface (i.e. population size). A disadvantage of plate counting is the selectivity of the medium upon which the samples are plated, as some bacteria do not grow under artificial conditions, or they grow very slowly e.g. the pink- pigmented facultative methylotrophic bacteria (Holland 1997). Those organisms will be missed in standard plate count analyses.
Another disadvantage of plate counting is that it is a semi-quantitative method, because bacterial aggregates will form only one colony. These two problems may lead to an underestimation of the bacterial count in a sample.
Microscopic techniques can be utilized as an alternative to estimate bacterial counts.
For this, leaves are washed in a known amount of buffer, and the leaf is weighed and/or the surface area is determined. Afterwards, samples are placed on a Neubauer counting chamber with a defined volume between slide and coverslip, which allows estimation of the cell count of a community. Microscopy can also be used to analyze colonized plant surfaces directly, either to examine the success of leaf washing procedures or to investigate the spatial distribution of bacteria on the leaf. To improve bacterial cell analysis, bacterial cells are commonly stained, for example with 4’,6-diamidino-2-phenylindole (DAPI) or Acridine orange, both fluorescent dyes that interact with nucleic acids and stain all bacteria including those bacteria that would be missed by plate counting due to their non-culturability.
Molecular community fingerprinting techniques can also be used to characterize the composition of an overall leaf community. For this, the total community DNA, or metagenome, is extracted and a subsequent polymerase chain reaction (PCR) with primers for conserved DNA-sequences that are connected by variable DNAsequences is performed (Handelsman, Rondon et al. 1998). Afterwards, a step to separate PCR products of the dominant populations, such as denaturating gradient gel electrophoresis (DGGE) (Muyzer and Smalla 1998) and terminal-restriction fragment length polymorphism (T-RFLP) (Marsh 1999), is conducted. Each band 5