UNDERSTANDING THE RELATIONSHIP BETWEEN BACTERIAL COMMUNITY

COMPOSITION AND THE MORPHOLOGY OF BLOOM-FORMING

MICROCYSTIS

A dissertation submitted

to Kent State University in partial

fulfillment of the requirements for the

degree of Doctor of Philosophy

by

Leighannah Akins

December 2018

©Copyright

All rights reserved

Except for previously published materials

Dissertation written by

Leighannah Akins

B.S., Spring Hill College, 2003

M.S., University of Nebraska-Lincoln, 2009

Ph.D., Kent State University, 2018

Approved by

Laura G. Leff, Ph.D. , Chair, Doctoral Dissertation Committee

Xiaozhen Mou, Ph.D., Members, Doctoral Dissertation Committee

Darren L. Bade, Ph.D.

Joseph Ortiz, Ph.D.

Alison Smith, Ph.D.

Accepted by

Laura G. Leff, Ph.D. , Chair, Department of Biological Sciences

James L. Blank, Ph.D., Dean, College of Arts and Sciences TABLE OF CONTENTS TABLE OF CONTENTS...... iii LIST OF FIGURES ...... v LIST OF TABLES ...... vi PREFACE…...... vii ACKNOWLEDGEMENTS ...... viii

I. OVERVIEW CHAPTER...... 1

II. COMPOSITION AND DIVERSITY OF CYANOBACTERIA- ASSOCIATED AND FREE-LIVING BACTERIAL COMMUNITIES DURING CYANOBACTERIAL BLOOMS...... 10

INTRODUCTION...... 10

METHODS...... 12

RESULTS...... 18

DISCUSSION...... 27

CONCLUSIONS...... 32

III. EXUDATES OF HETEROTROPHIC BACTERIA ENHANCE FREQUENCY AND SIZE OF MICROCYSTIS AERUGINOSA (CYANOPHYCEAE) COLONIES...... 34

INTRODUCTION...... 34

METHODS...... 37

RESULTS...... 43

DISCUSSION...... 55

CONCLUSIONS...... 60

IV. AI-2 QUORUM SENSING SIGNAL PROMOTES COLONY FORMATION IN MICROCYSTIS AERUGINOSA...... 61

INTRODUCTION...... 61

METHODS...... 64

iii

RESULTS...... 66

DISCUSSION...... 71

CONCLUSIONS...... 73

V. INTERACTIVE EFFECTS OF BACTERIA AND NUTRIENTS ON MICROCYSTIS MORPHOLOGY...... 75

INTRODUCTION...... 75

METHODS...... 77

RESULTS...... 80

DISCUSSION...... 89

CONCLUSIONS...... 92

VI. SUMMARY AND GENERAL DISCUSSION...... 93

REFERENCES...... 102

APPENDIX A. Supplementary Table 1. Abundance of Bacterial Families...... 124

iv

LIST OF FIGURES Figure 1. Study Sites ...... 13 Figure 2. Bacterial Community Diversity...... 19 Figure 3. Bacterial Community Compositional Similarity...... 21 Figure 4. Variation in Abundance of Proteobacteria...... 22 Figure 5. Variation in Abundance of Other Bacteria...... 23 Figure 6. Microcystis aeruginosa Colonies at 200x Magnification...... 44 Figure 7. Phylogeny of Colony-Promoting Isolates and Selected Other Bacteria...... 46 Figure 8. Treatment Effects on Morphology...... 48 Figure 9. Correlations of Colony Frequency with Colony Size and EPS...... 50 Figure 10. Components of Reflectance Across the Visible Spectrum ...... 53 Figure 11. Treatment Differences in Reflectance Components...... 54 Figure 12. Differences in Morphology at 24 Hours...... 69 Figure 13. Differences in Morphology and EPS at 48 Hours...... 70 Figure 14. Morphological Effects of EI-23...... 85 Figure 15. Morphological Effects of EI-7...... 86 Figure 16. Effects of EI-23 on EPS...... 87 Figure 17. Relationship of Colony Size and EPS Ratio...... 88

v

LIST OF TABLES Table 1. Physical and Chemical Conditions at Sampling Sites...... 27 Table 2. Colony-Promoting Effects of Bacterial Isolates…...... 45 Table 3. Cyanobacterial Growth Rates Varying with Isolate Treatment…...... 51 Table 4. Cyanobacterial Growth Rates Varying with AI-2 Treatment…...... 67 Table 5. Main and Interactive Effects of Four Independent Variables …...... 83

vi

Preface

The second chapter of this dissertation was previously published under the title

“Composition and Diversity of Cyanobacteria-Associated and Free-Living Bacterial

Communities During Cyanobacterial Blooms” in Annals of Microbial Ecology Volume

68, Issue 8, 493-503. My co-author Paul Ayayee made substantive contributions to the selection of analysis methods for the metagenomic dataset and to the design of Figures 3,

4, and 5. My dissertation advisor Laura G. Leff, also a co-author of the manuscript, advised me extensively on experimental design, analysis, and writing for publication.

The third chapter of this dissertation is a manuscript in preparation for submission for publication. My co-author Joseph Ortiz contributed greatly by advising me on the collection and analysis of reflectance data. My dissertation advisor Laura G. Leff, also a co-author, advised me on experimental set-up, presentation of figures and tables, and writing for publication.

vii

Acknowledgements

I would like to thank my committee members, Laura G. Leff, Xiaozhen Mou, Darren

Bade, and Joseph Ortiz for the all of time, effort, and expertise they have contributed to this project. I would also like to thank Paul Ayayee for his contributions to my first manuscript and to improving my understanding of genetic sequencing and statistical analysis of large datasets. Thanks to Mike Model for help with imaging many cyanobacterial cells and colonies, to Mahinda Gangoda for advice on analyzing microcystin, and to Chris Blackwood for statistical advice and the use of his lab’s instruments. Thanks to fellow graduate students Alescia Roberto, Jon Van Gray, Joe

Taura, and Anjali Krishnan for help with lab work, coding, and moral support and to my undergraduate assistant Emily Senderak for her careful work with my cell cultures and her enthusiastic attitude during trying times.

I thank my beloved mother Charlotte Ferrell for setting me on this journey and believing in me and encouraging me throughout it all.

viii

Chapter I: Overview

The Growing Challenge of Cyanobacterial Blooms

The geographic range and temporal duration of cyanobacterial harmful algal blooms (CyanoHABs) are expanding due to climate change and economic development, presenting challenges for ecological conservation and public health throughout the world

(Paerl & Huisman 2009, Paerl & Paul 2012). CyanoHABs occur in warm, thermally stratified, eutrophic waters (Paerl & Huisman 2009, Paerl & Paul 2012, Reynolds 1981).

During a bloom, buoyant cyanobacteria float at the surface, shading out other primary producers which have higher nutritional value for consumers in aquatic food webs (Paerl

& Paul 2012).

CyanoHABs have a negative impact on the environment in several ways. Some cyanobacterial strains release toxins which cause sickness and death of humans, livestock, and other animals (Paerl & Huisman 2009, Paerl & Paul 2012). Most zooplankton grazers find even the non-toxic strains of cyanobacteria less palatable or more difficult to consume than other phytoplankton, such as green algae and diatoms

(Paerl & Paul 2012). When blooms are severe, night time respiration or the decomposition of dead cyanobacteria can lead to anoxia in the water column and subsequent fish kills (Paerl & Paul 2012).

1

Physical and chemical factors that contribute to cyanoHABs are well-studied.

Previous research has established the importance of high water temperatures, stability of the water column, and high concentrations of bioavailable nitrogen and phosphorus as factors controlling cyanoHABs (Paerl & Huisman 2009, Reynolds 1981, Wang et al.

2010, Wilhelm et al. 2011). However, physical and chemical factors do not tell the whole story. Increasing attention is being paid to the role of biotic factors in determining the occurrence, duration, and toxicity of CyanoHABs.

Characteristics of Microcystis Blooms: Colony Morphology and the Polysaccharide

Matrix

Some of the most rapidly spreading nuisance cyanobacteria in the world belong to the genus Microcystis. In the environment, all known strains of Microcystis occur in aggregates or colonies embedded in a thick extracellular polysaccharide (EPS) matrix, with few free-floating single cells (Reynolds 1981). After a strain has been isolated and maintained in culture for several generations, it typically ceases to exhibit aggregate/colonial morphology (Bolch and Blackburn 1996). Most studies demonstrate a positive relationship between two key features of Microcystis morphology—the association of cells in colonies and the amount of polysaccharide mucilage. In laboratory experiments, Microcystis cells begin to aggregate and develop a mucilage matrix concurrently (Gan et al. 2012), and the size of aggregates or colonies is correlated with the amount of EPS surrounding the cells (Gan et al. 2012, Li et al. 2013, Ma et al. 2014,

Yang et al. 2008). Furthermore, extracting polysaccharides from Microcystis samples significantly diminishes the ability of the cyanobacteria to maintain colonies (Xu et al.

2014). However, more recent work has revealed that it is possible to stimulate colony

2 formation in Microcystis and other colonial cyanobacteria without a concomitant increase in polysaccharides (Tan et al. 2018).

There has been some speculation that typical culture conditions cause Microcystis populations to increase too rapidly to form as many colonies as stable or slow-growing populations. In one study, Microcystis cultures exhibited an inverse relationship between growth rate and EPS content (Li et al. 2013). Li et al. concluded that cultured Microcystis populations are found primarily as unicellular morphs because the warm temperatures and high light intensities under which they are maintained drive them to reproduce too fast to accumulate sufficient EPS to hold colonies together. However, other studies indicate that EPS accumulation and aggregate/colony size are strongly influenced by chemical factors that do not have significant effects on growth rate (Gan et al. 2012, Ma et al. 2013, Wang et al. 2011)

Chemical factors that potentially influence EPS production and aggregate/colony morphology include concentrations of ions and organic compounds. High concentrations of calcium ions are associated with increases in EPS production, aggregate/colony size, and buoyancy (Wang et al. 2011). The effect appears to be specific to calcium, as

Microcystis treated with magnesium behaves no differently from control cultures (Wang et al. 2011).

Complex organic molecules can also affect EPS production and colonial morphology. Microcystins, a family of hepatotoxic oligopeptides released by some strains of Microcystis, can also affect the morphology of other Microcystis strains, including non-toxic ones (Gan et al. 2012). For a non-toxic strain in monoculture, amendment with microcystin can induce a doubling of EPS production, accompanied by

3 significant increases in the size and buoyancy of aggregates/colonies (Gan et al. 2012).

Gan et al. found that adding exogenous microcystin to cultures of non-toxic Microcystis wesenbergii caused the cyanobacteria to upregulate transcription of some, but not all, of the genes involved in polysaccharide biosynthesis. As the cells produced more EPS, they also formed more frequent and larger colonies.

Other studies implicate nitrogen and phosphorus in morphology determination

(Ma et al. 2013, Yang & Kong 2013). In non-axenic microcosms, nutrient amendments rich in P were associated with decreased aggregate/colony size and increased prevalence of single Microcystis cells, whereas microcosms that received amendments of bioavailable N alone showed no significant morphological differences from controls (Ma et al. 2013). Axenic Microcystis cultures, in contrast, exhibit increased colony formation and EPS production when both N and P are low, although the colonies observed under such conditions appear not to be buoyant as colonies from the environment (Yang &

Kong 2013).

Relationship between Microcystis and Heterotrophic Bacteria

The relationship between Microcystis and attached heterotrophic bacteria is multi- faceted. EPS mucilage is home to bacteria such as Pseudomonas, which benefit

Microcystis by absorbing phosphorus that leaks from the cyanobacterial cells and recycling it into a form that can be taken up again (Jiang et al. 2007). Yet, microcystin- degrading bacteria may limit EPS accumulation and colony size by keeping environmental microcystin concentrations low (Gan et al. 2012). Bacteria are also capable of inducing aggregation of Microcystis cells that previously existed in unicellular form, as Shen et al. (2011) demonstrated when they detached bacteria from Microcystis

4 aggregates/colonies collected in Lake Taihu and introduced those bacteria to cultures of non-aggregated Microcystis; subsequent studies confirmed this effect could be achieved with specific bacterial isolates (Wang et al. 2015, Xu et al. 2012). Although studies that have examined the response of Microcystis EPS production and aggregation to bacterial isolates are less numerous than those that examine the effects of Microcystis and its exudates on heterotrophic bacteria, the literature available at present indicates that only a small number of the cultivable bacteria from a lake have this ability (Wang et al. 2015).

The composition of bacterial communities associated with cyanobacterial blooms is strongly influenced by the morphological and biochemical characteristics of the cyanobacteria. The mucilage surrounding Microcystis or other colonial or filamentous cyanobacteria provides a protected habitat for many types of bacteria that do not thrive in the open water column (Jiang et al. 2007, Li et al. 2011, Louati et al. 2015, Maruyama et al. 2003, Mou et al. 2013, Niu et al. 2011, Parveen et al. 2013). Furthermore, hepatotoxins released by toxic strains are a carbon source for microcystin-degrading bacteria such as Sphingomonas (Jones et al. 1994, Maruyama et al. 2003). Therefore, it might be expected that the microenvironment formed by cyanobacterial mucilage and cell surfaces would select for communities of heterotrophic bacteria that are compositionally distinct from free-living bacteria that thrive in open eutrophic waters. Several studies have examined the taxonomic composition of heterotrophic bacteria associated with a cyanoHAB within a single lake, often paying special attention to taxa that are known or suspected to degrade microcystins (Li et al. 2011, Louati et al. 2011, Maruyama et al.

2003, Mou et al. 2013, Niu et al. 2011, Parveen et al. 2013, Shen et al. 2011). However,

5 previous research has not compared cyanobacteria-associated and free-living bacterial communities collected from multiple lakes.

The ability of cyanobacterial colonies and EPS to shape communities of heterotrophic bacteria has received more attention than the ability of heterotrophic bacteria to influence the morphology and EPS production of cyanobacteria. Nevertheless, the morphological responses of Microcystis to colony-promoting heterotrophic bacteria

(Shen et al. 2011, Wang et al. 2015, Xu et al. 2012) may be equally ecologically important. Microcystis blooms begin with the mass accumulation of colonies in the upper water column (Wu & Kong 2009, Yamamoto et al. 2011, Zhu et al. 2014). Colonial morphology is crucial to the ability of Microcystis to maintain access to sunlight (Wu &

Kong 2009, Zhu et al. 2014) and escape predation (Jarvis et al. 1987, White et al. 2011,

White & Sarnelle 2014). Thus, bacterial interactions that increase the frequency and size of Microcystis colonies may be key to the development and maintenance of blooms.

Identifing colony-promoting bacteria and studying how different strains of Microcystis respond to them could be an important step in improving human understanding of

Microcystis blooms.

As reported in Chapter 2, this project investigated the composition and diversity of heterotrophic bacteria that occurred in association with cyanobacterial blooms in three eutrophic lakes in Ohio and compared them with free-living bacterial assemblages from the same lakes, with a view to assessing whether a distinctive cyanobacteria-associated microbial community structure could be identified across multiple lakes. It was hypothesized that the biomass of cyanobacterial blooms would select for distinctive cyanobacteria-associated communities that would be significantly different from free-

6 living bacterial communities and similar to other cyanobacteria-associated communities from different lakes.

In the next stage of the project, living bacteria were isolated from the biomass of a

Microcystis blooms. As reported in Chapter 3, bacterial isolates were identified by genetic sequencing and tested for the ability to enhance formation and growth of

Microcystis colonies in a controlled environment. Once this ability was confirmed in some of the bacterial isolates, those bacteria were selected for further testing to investigate the mechanism by which they influence Microcystis morphology.

Though previous experiments reported that co-cultures of Microcystis and colony- promoting bacteria contained more extracellular polysaccharides than pure cultures of

Microcystis (Shen et al. 2011, Wang et al. 2015), those studies did not reveal what how much of the polysaccharide content was produced by Microcystis cells or how much was produced by the heterotrophic cells. Moreover, it was not shown how colony-promoting bacteria were able to affect the morphology of Microcystis. One mechanism that has recently been proposed for the formation or maintenance of Microcystis colonies is that the cells might be held together by epibiontic bacteria that produce large amounts of strongly adherent polysaccharides (Zhang et al. 2018). Alternatively, colony-promoting bacteria may exude diffusible chemicals such as quorum-sensing signals that would induce other organisms to upregulate EPS production or to change the content of their mucilage or surface proteins in ways that increase adhesive abilities (Auger et al. 2006,

Federle & Bassler 2003, Pereira et al. 2013, Rickard et al. 2006, Ryan & Dow 2008).

This project tested the latter hypothesis with culture experiments described in Chapter 3, in which toxic and non-toxic strains of Microcystis aeruginosa were exposed to exudates

7 from colony-promoting bacteria. It was hypothesized that M. aeruginosa would exhibit morphological changes in response to the presence of colony-promoting bacteria even when the two microbial species were separated by a dialysis barrier permeable to small molecules but not to whole cells or EPS polymers. It was also hypothesized that any differences that might be observed between toxic and non-toxic M. aeruginosa strains would be attributable to the toxic strain producing microcystin which would then enhance

EPS production and cellular adhesion in this strain.

Another goal of the project was to learn whether the well-known cross-species quorum sensing molecule AI-2 (Auger et al. 2006, Federle & Bassler 2003, Pereira et al.

2013, Rickard et al. 2006, Ryan & Dow 2008) might be the means by which heterotrophic bacteria promote colonial morphology in Microcystis. In an additional experiment reported in Chapter 4, toxic and non-toxic cultures of M. aeruginosa were treated with AI-2 and assessed for changes in morphology and EPS content. It was hypothesized that AI-2 would promote colony formation in M. aeruginosa by increasing the total amount of EPS per cell and the ratio of cell-bound EPS over soluble EPS.

Furthermore, PCR primers targeting a gene crucial to AI-2 synthesis were employed to detect similar genes in DNA samples from colony-promoting bacteria. It was hypothesized that all colony-promoting bacteria would carry this gene.

Finally, as described in Chapter 5, a full factorial culture experiment was used to investigate how different nitrogen and phosphorus combinations interact with colony- promoting bacteria to affect Microcystis morphology. Laboratory experiments on

Microcystis strains are typically conducted in media where nitrogen is greatly in excess of phosphorus and where absolute concentrations of both nutrients are very high

8 compared to environments where Microcystis is found (Graham et al. 2004, Liu et al.

2011). Therefore, co-culture experiments with varying nutrient treatments needed to show how Microcystis strains respond to colony-promoting bacterial neighbors under different conditions. It was hypothesized that, across all nutrient treatments, colonies would be larger and more frequent in M. aeruginosa co-cultures with a colony-promoting bacterium than in control cultures of the same M. aeruginosa strain, and that growth media with high N:P ratios would enhance the colony-promoting effects of bacteria more than N-limited media. It was further hypothesized there would be significant interaction between N:P ratio and bacterial co-culture and that enhancement of colony size would be greater for toxic M. aeruginosa than for non-toxic M. aeruginosa.

9

Chapter 2: Composition and Diversity of Cyanobacteria-Associated and Free-Living Bacterial Communities During Cyanobacterial Blooms Introduction

Cyanobacterial harmful algal blooms (CyanoHABs) are common in growing numbers of freshwater ecosystems around the world due to climate change and nutrient loading (Paerl 1996; Paerl and Paul 2012). Typically, bloom-forming cyanobacteria occur as colonies embedded in mucilaginous matrices or as filaments within mucilaginous sheaths. The surfaces of cyanobacteria and their surrounding mucilage form microenvironments that make up the phycosphere, a distinctive habitat which supports heterotrophic bacterial communities that generally differ from the surrounding bacterioplankton (Li et al. 2011; Louati et al. 2015; Niu et al. 2011; Parveen et al. 2013a;

Parveen et al. 2013b; Shi et al. 2012). These microenvironments are protected from physicochemical fluctuations in the water column (Paerl 1996) and are rich in organic compounds, including polysaccharides (Parikh 2006; Pereira et al. 2009; Plude et al.

1991; Xu et al. 2013) and oligopeptides (such as microcystins and nodularins) that can be used as carbon sources by some bacteria (Imanishi et al. 2005; Jones et al. 1994;

Maruyama et al. 2003).

Previous studies of the composition and diversity of heterotrophic bacteria living on mucilaginous cyanobacteria have focused on the differences between cyanobacteria- associated (CA) communities and free-living (FL) bacterioplankton communities within a single lake rather than comparing CA bacterial assemblages from different lakes

10

(Louati et al. 2015; Niu et al. 2011; Parveen et al. 2013a; Shi et al. 2012). Comparisons across multiple lakes are necessary to understand whether differences between FL and

CA bacterial communities are attributable to innate selective properties imposed by the cyanobacteria-microhabitat. Alternatively, such differences may result from stochastic community assembly in combination with differences among lakes in physicochemical conditions and pools of potential bacterial colonizers. Ultimately, such information will reveal whether or not CA communities share common characteristics that could potentially serve as predictive or management tools.

Composition of bacterial communities associated with mucilaginous cyanobacteria varies, but often includes Bacteroidetes and Actinobacteria (Cai et al.

2014; Li et al. 2011; Louati et al. 2015; Niu et al. 2011; Parveen et al. 2013a; Shen et al.

2011; Shi et al. 2012). Actinobacteria tend to be predominantly free-living cells incidentally co-occurring with cyanobacteria (Louati et al. 2015; Parveen et al. 2013b), although some members of this phylum live within the mucilage (Zhang et al. 2016).

Bacteroidetes are found embedded in the mucilage (Parveen et al. 2013b) or associated with surfaces of non-mucilaginous cells (Velichko et al. 2015). Gammaproteobacteria, uncommonly found as free-living organisms in freshwater (Niu et al. 2011; Parveen et al.

2013a; Shi et al. 2012), are often abundant in communities attached to cyanobacteria

(Parveen et al 2013a; Velichko et al. 2015). Betaproteobacteria are typically well- represented in free-living bacterial communities in freshwater lakes and have even higher abundances in communities associated with mucilaginous cyanobacteria (Louati et al.

2015; Parveen et al. 2013a). While taxonomically coarse comparisons can be made

11 among CA bacterial communities across studies, whether or not there is a characteristic

CA bacterial community is unknown.

In this study, the composition and diversity of CA and FL bacterial communities from three temperate lakes (OH, USA) during toxic cyanobacterial blooms were investigated. We hypothesized that the protected and resource-rich microenvironments associated with mucilaginous cyanobacteria would select for a subset of the bacteria available in the water column. Thus, we anticipated that CA bacterial communities would be compositionally distinct from FL bacterial communities within the same lakes and would have lower alpha diversity than FL communities as suggested by prior studies (Li et al. 2011; Niu et al. 2011; Parveen et al. 2013a; Parveen et al. 2013b; Shi et al. 2012).

Furthermore, we anticipated that the conditions of CA microhabitats would select for a consistent subset of taxonomic groups from among the pool of potential colonizers.

Therefore, we hypothesized that CA communities from different lakes would be more similar in composition to each other than to FL communities from the same lakes and that

CA communities would exhibit a greater degree of cross-lake similarity than would FL communities.

Methods

Study Sites

In the summer of 2014, three lakes were examined based on the occurrence of cyanoHABs as reported by the Ohio Environmental Protection Agency (2014) and the presence of dense, visible green surface scum. High cyanobacterial cell counts, cell biovolumes, and gene sequence abundances in these lakes were also reported by Francy

12 et al. (2015). In recent years, the selected lakes have all developed annually recurring toxic cyanobacterial blooms which typically persist throughout the summer and into early autumn. The lakes were located in northeast, central, and southern Ohio (Figure 1).

Figure 1. Study Sites. Locations of three eutrophic lakes sampled during cyanobacterial blooms in 2014.

Buckeye Lake (39.93°N, 82.48°W; mean depth 2.5 m, maximum depth 7 m, surface area 11.6 km2) is a reservoir in central OH (Francy et al. 2016). For most of the year, the reservoir is fed by a small watershed (~ 70 km2) with 60% agricultural, 14% forest, and 15% urban land use (Francy et al. 2016; Taylor and Governor 2012). At times of high precipitation, it receives overflow from the headwaters of South Fork Licking

13

(Taylor and Governor 2012) and runoff from additional areas, draining a total of 127 km2 of predominately agricultural land (Francy et al. 2016; Taylor and Governor 2012).

William Harsha Lake (39.02°N, 84.11°W, mean depth 12.9 m, maximum depth

30 m, surface area 8 km2), formerly known as East Fork Lake, is a monomictic reservoir in southern OH (Beaulieu et al. 2014; Francy et al. 2016). Constructed on the East Fork of the Little Miami River, it drains a watershed of about 886km2. Land use is 64% agricultural and 27% forest, with the rest lightly urbanized (Beaulieu et al. 2014; Francy et al. 2016).

Maumee Bay (41.68°N, 83.38°W, mean depth <3 m, maximum depth ~3 m except for a dredged shipping channel of 8.5 m, surface area 70 km2) is a shallow embayment on the southwestern shore of Lake Erie (mean depth 7.4 m, max depth 19 m, surface area 19,830 km2). The 2014 HAB in Maumee Bay was part of a larger bloom in which cyanobacterial surface scum, consisting predominately of Microcystis spp., covered much of the lake’s western basin (3,284 km2). The Maumee River drains a watershed of 16,388 km2, of which 73.3% is agricultural land and 10.6% is urban, including the city of Toledo, OH (Baker et al. 2014; Moorhead and Morris 2008). The river discharges directly into Maumee Bay from the southwest (Francy et al. 2015;

Michalak et al. 2013, Moorhead and Morris 2008). To the northeast, the bay opens onto the western basin, but water flow patterns permit little mixing within Maumee Bay, leaving the Maumee River as the primary conduit of water, dissolved nutrients, and suspended sediment into the bay.

Sample collection and processing

14

Three replicate one liter water samples were collected from the top five centimeters of each lake in 2014. Each lake was sampled once in the period from July to

August, when the cyanobacterial bloom season is at its height in temperate North

America due to high temperatures and strong thermal stratification of lakes, and again in

September, when cyanobacterial blooms in the region are generally on the decline.

Buckeye Lake was sampled on July 24 and September 19, William Harsha Lake was sampled on July 17 and September 2, and Maumee Bay was sampled on August 18 and

September 28. At the time of sampling, temperature, conductivity, and dissolved oxygen were measured with a HQ40d multiprobe (Hach, Loveland, CO, USA). Samples were transported on ice to the lab, where they were first filtered through 3 µm nitrocellulose membranes (Millipore, Darmstadt, Germany) under vacuum to collect cyanobacteria- associated bacteria (CA) associated with larger sized cyanobacteria cell surfaces or embedded in cyanobacterial mucilage, and then through 0.2 µm polycarbonate membranes (Millipore, Darmstadt, Germany) (modified from Li et al. 2011 by using filters with a smaller pore size) to collect small sized free-living bacterial (FL) fractions.

Membranes were stored at -80° C until DNA extraction.

Inorganic nitrogen was measured with a Synergy 2 plate reader (BioTek,

Winooski, VT, USA) following the indophenol blue method for ammonium and the sulfanilamide method for nitrate/nitrite as adapted for microplates (Ringuet et al. 2011).

Soluble reactive phosphorus (SRP) was assayed by the ascorbic acid method (Murphy

1962), and absorbance measured with a DU 730 UV/Visible spectrophotometer

(Beckman Coulter, Brea, CA, USA). Nutrient data were tested for normality with the

Shapiro-Wilkes test and the Kruskal-Wallis test was used to check for significant

15 differences in the event of non-normal distribution. Statistical analyses were carried out in JMP (SAS, Inc., Cary, NC, USA).

Bacterial Community Analysis

DNA was extracted from filters using the Power Soil DNA extraction kit (MoBio,

Carlsbad, CA, USA) according to manufacturer’s protocol. Presence of 16S rRNA genes was confirmed and samples were subsequently submitted for high-throughput 2 x 300 bp paired-end sequencing of the V4-V5 hypervariable region (Sun et al. 2013) using an

Illumina MiSeq Series System (Illumina Inc, San Diego, CA, USA) at the Ohio State

University Molecular and Cellular Imaging Center (Wooster, OH, USA).

Following sequencing, paired reads were assembled into iTags (Degnan 2012). iTags were sorted by length, filtered for chimeras, and quality filtered in the pick_open_reference workflow with usearch61_ref as the operational taxonomic unit

(OTU) picking and classification method in QIIME Version 1.9.1 (Caporaso et al. 2010).

OTU clustering was performed at the 97% similarity level and taxonomy assigned based on partial 16S rRNA sequences in the 16S rRNA SSU_Ref_NR_99_128.1 reference database (SILVA_SSU_128.1, Release date, September 29, 2016) (Caporaso et al. 2010).

A total of 1,564,141 ‘iTags’ and 44,296 OTUs were obtained. Two samples with low

OTU counts, Maumee Bay Aug. 18 FL (33 iTags) and William Harsha Lake Sept. 2 FL

(52 iTags), were excluded from subsequent analysis. Singletons, OTUs unassigned at the basal level (D1), and OTUs assigned to Archaea, mitochondria, chloroplast, and cyanobacteria lineages were removed from all samples in the resulting OTU table, yielding 968,520 iTags with 19,306 OTUs. The final filtered OTU table was then summarized to 445 bacterial phylotypes at the family level.

16

Species richness (alpha diversity) across samples was assessed using Shannon diversity, Simpson’s index, and the unique OTU count (observed_species metric in

QIIME), following rarefaction of the family-level OTU table to 6,940 iTags per sample.

Samples were sorted into twelve a priori FL and CA groups representing bacterial communities collected from the three lakes on two sampling dates per lake. Differences in Shannon diversity, Simpson’s index, and OTU richness were evaluated using the

Wilcoxon non-parametric test. To estimate beta diversity, a Bray-Curtis distance matrix was generated using the rarefied family-level table (Bray and Curtis 1957; Anderson et al. 2006). A non-metric multidimensional scaling (NMDS) plot was generated to visualize dissimilarity in community composition among samples following NMDS analysis on the distance matrix (Kruskal 1964). Differences in community composition among samples were evaluated using the Multivariate Response Permutation Procedure

(MRPP) (Mielke 1984) on the distance matrix via the “compare_categories.py” command in QIIME with 1000 permutations.

An additional test of dissimilarity in microbial composition among samples

(permutational multivariate analysis of variance, PERMANOVA) was performed

(Anderson et al. 2006). An underlying assumption of PERMANOVA is that all groups are made up of replicates which exhibit the same level of dispersion around their group centroids (Anderson et al. 2013). Homogeneity of within-group dispersions was assessed using PERMDISP (Anderson et al. 2013), and a PERMANOVA-based F-value calculated from average distances among groups relative to average distances within groups for actual and permutated data. A pseudo-F statistic then tested the likelihood that permutated F-values were larger than the observed F-value. Finally, differences in

17 abundance values for each OTU among sampling groups and among a posteriori clusters were examined using the group_significance.py command followed by the default non- parametric Kruskal-Wallis test of significance in QIIME (Caporaso et al. 2010). The p- values were adjusted using the False Discovery Rate (FDR) approach.

Results

Sequences of cyanobacterial taxa dominated the three m pore size fraction of each sample. The bloom in Buckeye Lake consisted almost entirely of Planktothrix on both sampling dates. Maumee Bay was strongly Microcystis-dominated on August 18, but bloom composition shifted to a mix of Microcystis and Dolichospermum by

September 24. William Harsha Lake had the most diverse cyanobacterial assemblage, including Cylindrospermopsis, Dolichospermum, Synechococcus, and Microcystis, although Microcystis became much less abundant in September than in July.

In contrast to our prediction, there were no significant differences in bacterial diversity (cyanobacteria excluded) between CA and FL communities across the three lakes (OTU richness, Wilcoxon chi-square = 11.6, df = 11, P = 0.40; Shannon’s index,

Wilcoxon chi-square = 9.2, df = 11, P = 0.60; Simpson’s index, Wilcoxon chi-square =

10.2, df = 11, P = 0.52), although FL community diversity indices were generally larger than CA indices (Fig. 2). The CA community of Maumee Bay became slightly less diverse over time, with fewer unique OTUs detected in September than in August (Fig.

2A), but the change was statistically non-significant.

18

Figure 2. Bacterial Community Diversity. Bar plots of mean and standard error for (A)

OTU richness, (B) Shannon Diversity, and (C) Simpson’s Index for free-living (black bars) and cyanobacteria-associated (white bars) bacterial communities.

19

Community composition differed significantly among the twelve sampling groups representing three lakes, two sampling dates, and two community types (FL and CA;

MRPP, within-group agreement, effect size, A = 0.67, observed delta = 0.18, and expected delta = 0.54, P = 0.001, 1000 permutations; PERMANOVA, Pseudo-F = 21.5,

P = 0.001, 1000 permutations). The sampling groups fell into four clusters, with Cluster 1 made up of CA communities from Buckeye and Harsha Lakes and Cluster 2 composed of

FL communities from the same two lakes (Fig. 3). Both CA and FL communities sampled from Maumee Bay in August were represented by Cluster 3, along with the FL community from Maumee Bay sampled in September. Cluster 4 consisted of only the CA community from the September sampling date in Maumee Bay. The four clusters differed significantly in community composition (MRPP, within-group agreement, effect size, A

= 0.35, observed delta = 0.35, and expected delta = 0.54, P = 0.001, 1000 permutations), confirming differences between FL and CA habitat types and similarities among samples within habitat types for two of three lakes. Out of 445 family-level bacterial taxa, 178 differed significantly in abundance among sampling groups, including 80 Proteobacteria taxa (Fig. 4) and 98 other taxa (Fig. 5), and 128 family-level groups differed among the four clusters in Figure 3 at <0.05 (Appendix).

20

Figure 3. Bacterial Community Compositional Similarity. NMDS plot showing similarity of bacterial communities in free-living (FL) and cyanobacteria-associated (CA) samples from Buckeye Lake (BL), William Harsha Lake (WH), and Maumee Bay (MB) based on mean NMDS scores (stress = 0.07, MRPP, within-group agreement, effect size,

A = 0.67, observed delta = 0.18 and expected delta = 0.54, P = 0.001, 1000 permutations). Samples close to each other are more similar in composition than those farther apart.

21

Figure 4. Variation in Abundance of Proteobacteria. Relative abundances (%) of the

80 Proteobacteria families and family-level groups that differed significantly (<0.05) among the twelve sampling groups of free-living (FL) or cyanobacteria-associated (CA) communities collected from Buckeye Lake (BL), William Harsha Lake (WH), or

Maumee Bay (MB).

22

Figure 5. Variation in Abundance of Other Bacteria. Relative abundances (%) of the

98 families and family-level groups outside the Proteobacteria phylum that differed significantly (<0.05) among the twelve sampling groups of free-living (FL) or cyanobacteria-associated (CA) communities collected from Buckeye Lake (BL), William

Harsha Lake (WH), or Maumee Bay (MB).

With the exception of the CA community collected from Maumee Bay in August,

CA communities were separated from FL communities along the first axis of the NMDS plot (Fig. 3). Multiple taxa were abundant in Clusters 2 and 3, which consisted entirely or mostly of FL samples, that were relatively depleted in Clusters 1 and 4, made up exclusively of CA communities. The greatest difference was in the abundance of

Sporichthyaceae (Actinobacteria), a family more than ten times as abundant in the FL- dominated clusters than in the CA-only clusters (Appendix). Most other taxa that were

23 more abundant in the FL clusters were Alphaproteobacteria or Betaproteobacteria

(Appendix). However, unassigned Acidimicrobiales (Actinobacteria),

Sphingobacteriaceae (Bacteroidetes), and Leptospiraceae (Spirochaetes) were also found in the FL-dominated clusters at abundances that were, while low, still significantly greater than their abundances in Clusters 1 and 4 (Appendix). No taxa that were found to be abundant in Clusters 1 and 4 were significantly less abundant in Clusters 2 and 3.

Communities in Maumee Bay separated from communities in the two other lakes along the second NMDS axis (Fig. 3). Clusters 1 and 2 were enriched in

Chthoniobacterales (Verrucomicrobia), unclassified Verrucomicrobia, Planctomycetes families Phycisphaerae and Planctomycetaceae, Rickettsiaceae (Alphaproteobacteria),

Oceanospirillaceae (Gammaproteobacteria), and several families belonging to the

Deltaproteobacteria orders Bdellovibrionales and Oligoflexales, whereas these taxa were much less common in Clusters 3 and 4 (Appendix). Other Gammaproteobacteria families were rare overall but significantly more abundant in the Buckeye-Harsha clusters than in the Maumee Bay clusters (Appendix). In contrast, Maumee Bay clusters were relatively enriched in Caulobacteraceae and Hyphomonadaceae (Alphaproteobacteria),

Nitrosomonadaceae (Betaproteobacteria), and Chromatiaceae (Gammaproteobacteria), while unassigned Rickettsiales (Alphaproteobacteria) sequences were rare but most abundant in Maumee Bay (Appendix).

The abundance of unclassified Verrucomicrobia was much greater in Cluster 1 than in Cluster 2. Cluster 1 was further separated from all other clusters by high abundances of Blastocatellaceae (Acidobacteria), group OPB35 (Verrucomicrobia),

Flavobacteriales NS9 marine group (Bacteroidetes), and several members of the order

24

Sphingobacterales (Bacteroidetes), especially Saprospiraceae and group env. OPS 17

(Appendix). Cluster 1 also had significantly more Chloroflexi than other clusters, especially Caldilinaceae and unassigned Chloroflexi, but also the relatively rare

Anaerolineaceae (Appendix). Two families within the Bacillales (Firmicutes),

Bacillaceae and Paenibacillaceae, also contributed to the separation of Cluster 1, as did an unassigned Planctomycetes group OM190. This cluster was further distinguished by high abundances of a wide variety of rare taxa, including Fusobacteriaceae

(Fusobacteria), Ignavibacteriaceae (Ignavibacteriae), unassigned Hydrogenedentes,

Rhizobiales group A0839 (Alphaproteobacteria), unclassified Proteobacteria, and many

Deltaproteobacteria such as unassigned Bradymonadales, various members of the order

Myxococcales, and clade Sva0485 (Appendix).

Phycisphaerae and Planctomycetaceae were highly enriched in Cluster 2 but less enriched in Cluster 1 and relatively depleted in Clusters 3 and 4 (Appendix). Cluster 2 communities were also rich in unassigned Acidobacteria and several members of the

Actinobacteria, including the highly abundant Acidimicrobiaceae (Acidimicrobiales), other Acidimicrobiales, Mycobacteriaceae (Corynebacteriales), and multiple family-level groups within clade PeM15 and class Thermoleophilia (Appendix). Two

Gammaproteobacteria taxa, the Legionellaceae family and an unassigned

Xanthomonadales group, were significantly more abundant in this cluster than elsewhere, as were the Deltaproteobacteria groups Bacteriovoraceae and Oligoflexales 0319-6G20, plus a group of bacterial sequences that could not be assigned to any known phylum

(Appendix). Rare phylotypes found at their highest abundance in this cluster fell into

25 unassigned groups within Acidobacteria, Actinobacteria, Omnitrophica,

Alphaproteobacteria, and Gammaproteobacteria.

Microbacteriaceae (Actinobacteria), Cyclobacteriaceae (Bacteroidetes),

Chitinophagaceae (Bacteroidetes), Rhodobacteraceae (Alphaproteobacteria),

Xanthomonadaceae (Gammaproteobacteria), and unassigned Opitutae

(Verrucomicrobia) were all abundant in Cluster 3 and significantly less abundant in other clusters (Appendix). The SL56 marine group (Chloroflexi) and two family-level groups of Rhizobiales (Alphaproteobacteria) were moderately abundant in Cluster 3 but rare elsewhere (Appendix). Demequinaceae and unassigned Micrococcales, both belonging to the same order as Microbacteriaceae, were rare in all clusters but significantly less rare in Cluster 3 (Appendix). Another rare taxon, Candidatus Azambacteria (Parcubacteria), was also found mostly in this cluster (Appendix).

Few family-level groups were significantly more abundant in Cluster 4 than any other cluster. They included the moderately abundant Cytophagaceae (Bacteroidetes),

Parachlamydiaceae (Chlamydiae), Nitrosomonadaceae (Betaproteobacteria), and

Chromatiaceae (Gammaproteobacteria) (Appendix). Among the rare taxa, an unassigned

Bacteroidetes group, Rhizobiales group DUNssu044, and Desulfomonadales C8S-102

(Deltaproteobacteria) were detected mainly in Cluster 4 (Appendix).

The lakes exhibited variation in temperature, DO, conductivity, and Secchi depth

(Table 1). Inorganic P concentrations were similar across all lakes, below five µg/L except for a single sample collected from Buckeye Lake. Ammonium and nitrate/nitrite were below detection limits in all samples.

26

Table 1. Physical and Chemical Conditions at Sampling Sites. Water temperature, dissolved oxygen, conductivity, Secchi depth, and soluble reactive phosphorus (SRP) data collected for all sampling locations and dates.

Lake Sampling Temperature Dissolved Conduc- Secchi SRP Date (oC) Oxygen tivity Depth (g/L) (mg/L) (S/cm) (m) Mean (SE) Harsha Lake July 17 27.3 11.53 235 0.75 1.3 (0.31)

Sept. 2 26.7 6.93 242 1.0 1.06 (0.20)

Buckeye Lake July 24 25.8 7.52 270 0.25 1.65 (0.47)

Sept. 19 19.7 7.48 290 0.5 3.78 (3.78)

Maumee Bay Aug. 18 24.1 13.11 408 0.25 1.42 (0.54)

Sept. 24 23.0 13.18 374 0.5 1.77 (0.35)

Discussion

In William Harsha Lake and Buckeye Lake, cyanobacteria-associated communities were compositionally distinct from free-living communities but not significantly different in diversity. The CA communities from these two different lakes clustered together in NMDS analysis, while the FL communities from the same lakes fell outside the cluster, indicating that CA communities were more similar to each other than to their respective FL communities. Although the communities underwent some turnover between July and September, the separation between CA and FL communities within lakes and the clustering of CA communities across lakes remained consistent over time.

The FL communities from the same pair of lakes also fell into a single cluster. These

27 results support the hypothesis that cyanobacteria would provide a microhabitat that selected for a different set of bacteria than those dominant in the water column. They did not, however, support the hypothesis that bacterial communities selected by the CA habitat type would exhibit lower alpha diversity within lakes or less compositional variation between lakes than FL communities. Furthermore, the dissimilarity between the

CA community in Maumee Bay and other CA communities was equal to or greater than the dissimilarity between the FL community in Maumee Bay and other FL communities.

Previous studies revealed compositional differences between CA and FL communities within lakes (Li et al. 2011; Louati et al. 2015; Niu et al. 2011; Parveen et al. 2013a;

Parveen et al. 2013b; Shi et al. 2012) or differences among homogenized bacterial communities sampled from multiple lakes (Eiler and Bertilsson 2004), but this study is the first, to our knowledge, to distinguish between CA and FL communities and also compare them across multiple lakes.

Maumee Bay had highly similar FL and CA communities in August, unlike other lakes in this study and in prior studies (Li et al. 2011; Louati et al. 2015; Niu et al. 2011;

Parveen et al. 2013a; Parveen et al. 2013b; Shi et al. 2012). The similarity of the FL and

CA communities in this instance might be due to the relatively early stage of the cyanobacterial bloom. The initial sampling of each lake in this study occurred after the cyanobacterial bloom had developed thick green surface scums. This appearance coincided with the cyanobacterial biovolume of the bloom reaching a density of 1 x 107

µm3/mL, as reported in Francy et al. (2015). The Maumee Bay cyanoHAB of 2014 reached that density at the beginning of August, whereas the blooms in Harsha Lake and

Buckeye Lake had already met or exceeded that density more than a month earlier

28

(Francy et al. 2015). It is possible that significant dissimilarities between communities of free-living and cyanobacteria-associated bacteria in other lakes were the result of processes that occurred over the course of several weeks. Thus, the CA community in

Maumee Bay may have had less time to become distinct from its surrounding FL community, compared with the other CA communities in this study. By the next sampling date in September, the CA and FL communities in Maumee Bay were quite distinct, the CA community forming its own cluster while the FL community remained similar in composition to the samples collected in August. The compositional changes that made this CA community different from the other Maumee Bay communities in

Cluster 3 also increased its distance in the NMDS plot from the FL communities of

Cluster 2 and brought it closer to the CA communities of Cluster 1.

Differences among lakes in physicochemical conditions presumably contributed to differences among bacterial communities. In a number of ways, such as the predominantly agricultural land use of the surrounding watersheds and the depletion of inorganic nutrients during the cyanobacterial blooms, the lakes are similar. However, other factors such as temperature, size, and conductivity differed. During the sampling period of this study, the range of water temperatures recorded in Maumee Bay was not different from that in Buckeye Lake (Table 1). However, throughout the entire summer of

2014, temperatures tended to be slightly cooler in Maumee Bay than in the other lakes

(Francy et al. 2015). Lower temperatures early in the season may have slowed the development of the cyanoHAB in Maumee Bay and the differentiation of the CA community from the FL community. Another potential contributing factor is the large surface area of Maumee Bay. The cyanoHABs in the two reservoirs may have reached

29 high biovolume densities earlier because their surface waters covered roughly one- seventh of the area of Maumee Bay. Additionally, Maumee Bay had by far the highest conductivity readings and the greatest range of conductivity readings (Table 1). Over the course of the season, the conductivity in Maumee Bay exhibited an even greater range, with a maximum of 727 S/cm, more than double the maximum values for the other lakes (Francy et al. 2015). Although the range of variation in conductivity for these freshwater lakes is small compared to brackish waters, differences in conductivity within the range of 22-1,399 S/cm have significant effects on the structure of bacterial communities in streams (Lear et al. 2009). Dissolved oxygen was elevated in Maumee

Bay, especially relative to Buckeye Lake (Table 1), but the literature shows that this difference between lakes did not persist throughout the summer (Francy et al. 2015). It is unlikely that the separation of clusters in Fig. 2 was dependent upon differences in DO.

When Harsha Lake was sampled on July 17, the DO reading was closer to the values for

Maumee Bay than to those for Buckeye Lake, yet the bacterial communities, both CA and FL, clustered with the Buckeye Lake communities rather than the Maumee Bay communities (Fig. 3).

Because mucilaginous cyanobacterial offer ready access to complex organic molecules (Cottrell and Kirchman 2000; Imanishi et al. 2005; Jones et al. 1994;

Kirchmann 2002), it was anticipated that CA communities would be dominated by taxa that assimilate carbon and nutrients from these sources for rapid growth as is common among Gammaproteobacteria and Bacteroidetes (Cottrell and Kirchman 2000;

Kirchmann 2002; Newton et al. 2011), while FL communities would exhibit more evenness of OTU abundances and greater variation in the identities of dominant taxa.

30

Contrary to expectations, the two clusters that contained FL communities, Clusters 2 and

3, exhibited the greatest dominance of a single family, Sporicthyaceae (Table S2, Fig. 5).

The greater abundance of this and other Actinobacteria groups in these clusters is consistent with previous studies showing that members of this phylum were predominantly free-living (Allgaier and Grossart 2006; Louati et al. 2015; Parveen et al.

2013b). However, most of the Gammaproteobacteria that differed significantly among clusters and two of the highest-abundance Bacteroidetes families, Cyclobacteriaceae and

Chitinophagaceae, were affiliated with FL-dominated clusters rather than CA-only clusters. This was unexpected, given the tendency of Gammaproteobacteria and

Bacteroidetes to associate with cyanobacterial particles and other large organic particles in aquatic environments (Cai et al. 2014; Cottrell and Kirchmann 2000; Crump et al.

1999; Li et al. 2011; Louati et al. 2015; Niu et al. 2011; Parveen et al. 2013a; Shi et al.

2012).

Another unusual feature of the communities examined in this study was the importance of Deltaproteobacteria. Bacteriovoracaceae (Bdellovibrionales) were abundant in all three lakes, and this family and Bdellovibrionaceae, also a member of order Bdellovibrionales, were especially abundant in Clusters 1 and 2. Several

Deltaproteobacteria phylotypes in the order Myxococcales were detected in Cluster 1 at abundances that were significantly higher than in the other clusters. Myxococcales are more characteristic of soils and sediments than of surface waters (Basak et al 2015; Kim et al. 2016; Kou et al. 2016; Zlatković 2017). These bacteria may have entered the lakes along with sediment from surrounding agricultural lands. Both Bdellovibrionales and

Myxococcales are highly motile predators that attack and lyse other Gram negative

31 bacteria (Cai et al. 2014; Davidov and Jurkevitch 2004; Rotem et al. 2014; Velicer et al.

2014), including cyanobacteria (Caiola and Pellegrini 1984; Maruyama et al. 2003). Prior studies of lakes undergoing cyanoHABs reported Deltaproteobacteria as present but not abundant (Cai et al. 2014; Eiler & Bertilsson 2004; Li et al. 2011; Louati et al 2015; Niu et al. 2011). The high abundance of cell-lysing predatory bacteria in these lakes may also explain how large numbers of Gammaproteobacteria and Bacteroidetes were able to access high molecular weight organic compounds without associating with intact cyanobacteria. Widespread cell lysis would release into the water cellular products that would otherwise be exuded gradually, thus eliminating the need for bacteria that metabolize these compounds to associate directly with cyanobacterial cells.

Flavobacteriaceae have been known to become highly abundant among free-living bacteria when high molecular weight organic compounds occurred at high concentration in open water (Kirchmann 2002), a condition which could follow from high abundances of bacterial predators.

Conclusions

Results from this study support the hypothesis that the kind of microhabitat created by dense cyanoHABs selected for similarly structured assemblages of bacteria across multiple lakes, and that the selective pressures of this habitat type were different from those in the water column. Insights gained from this study further understanding of the ways in which cyanoHABs shape aquatic microbial communities. Cyanobacteria- associated communities from three lakes were significantly different from free-living communities in taxonomic composition. The twelve bacterial communities sampled fell into four clusters based on compositional similarity, with one cluster consisting of CA

32 communities from Harsha Lake and Buckeye Lake and a second cluster consisting of the

FL communities from these two lakes. Cross-lake similarities in the composition of CA communities or FL communities can be attributed to the similar sizes and physicochemical properties of these lakes as well as to similarities of the microhabitat types. Samples from Maumee Bay formed their own distinct clusters, one consisting of

CA and FL samples collected shortly after the bloom reached its maximum density by cyanobacterial biovolume as well as FL samples from a later date, while the CA samples from a later stage of the bloom clustered separately. As the bloom progressed in Maumee

Bay, the FL community remained similar to the earlier samples, while the composition of the CA community became significantly different. The hypothesis that the CA habitat type would select for a less diverse set of bacteria was not supported, as the CA communities overall had neither significantly lower measures of alpha diversity nor greater compositional similarity among lakes than did the FL communities. This is the first study to separate bacterial communities physically associated with mucilaginous cyanobacterial colonies or filaments from free-living bacteria and to compare the diversity and composition of the two types of communities across multiple lakes. Further studies are needed to explore the processes that differentiate CA from FL communities and the time scale on which these processes occur.

33

Chapter 3: Exudates of Heterotrophic Bacteria Enhance Frequency and Size of

Microcystis aeruginosa (Cyanophyceae) Colonies

Introduction

Cyanobacterial harmful algal blooms (cyanoHABs) pose a growing threat to water quality and ecosystem health throughout the world (Lürling et al. 2017, O’Neil et al. 2012, Paerl & Huisman 2009, Paerl & Paul 2012). One of the most widespread bloom- forming cyanobacteria is Microcystis aeruginosa, characterized by large, buoyant colonies of cells embedded in an extracellular polysaccharide (EPS) matrix (Reynolds et al. 1981, Wu & Kong 2009, Xu et al. 2013, Yamamoto et al. 2011, Zhu et al. 2014). M. aeruginosa typically loses colonial morphology when cultured in the laboratory, becoming predominantly unicellular (Bolch & Blackburn 1996, Zhang et al. 2007).

Likewise, prior to the start of the bloom season, M. aeruginosa exists in lake water columns as small colonies and unicells (Wu et al. 2009, Zhu et al. 2014). Blooms begin to form as colonies increase in size and become more frequent relative to unicells

(Yamamoto et al. 2011, Zhu et al. 2014). Larger colonies are more resistant to zooplankton predation (van Gremberghe et al. 2009, Yang & Kong 2012) and more buoyant and thus better suited to shading out competitors (Li et al. 2016, Wu & Kong

2009, Yamamoto et al. 2011, Zhu et al. 2014).

A variety of factors influence the size and frequency of colonies, including presence of the hepatotoxin microcystin (MC) (Gan et al. 2012, Sedmak & Eleršek 2006),

34 chemical exudates from zooplankton grazers (Bukert et al. 2001, Ha et al. 2004, van

Gremberghe et al. 2009, Yang & Kong 2012, Yang et al. 2005, Yang et al. 2006, Yang et al. 2008), low concentrations of antibiotics (Tan et al. 2018), and contact with heterotrophic bacteria (Shen et al. 2011, Wang et al. 2015). Although progress has been made in uncovering mechanisms by which some factors promote aggregation of

Microcystis cells, the process by which heterotrophic bacteria impact colony formation is not yet understood.

Effects of heterotrophic bacteria on the morphology of Microcystis may be mediated by diffusible chemical signals, as is the case for certain zooplankton (Ha et al.

2004). Bacteria may release exudates such as quorum-sensing molecules (Passos da Silva et al. 2017, Xue et al. 2012, Zhai et al. 2012) or antimicrobial compounds (Tan et al.

2018) which cause cyanobacterial cells to modify their surface properties and/or their mucilage in ways that make them more adhesive. Microcystin and zooplankton grazing promote colony formation by upregulating EPS production in Microcystis (Gan et al.

2012, Yang et al. 2008), and a similar mechanism might be involved in colony induction by heterotrophic bacteria. Bacteria that produce antibiotics may promote colony formation by inducing cyanobacterial cells to allocate more proteins to the cell surface or the extracellular mucilage (Tan et al. 2018). Alternatively, Microcystis colonies might form as a result of highly adhesive heterotrophic cells attaching directly to the surfaces of any cells around them. Some bacteria isolated from Microcystis colonies have strong adhesive properties and tendency to from EPS-rich aggregates among themselves in the absence of Microcystis (Zhang et al. 2018). It is possible that these bacteria and the EPS

35 they produce may bind aggregations of Microcystis and other bacteria together, rather than any properties of Microcystis itself or its own extracellular products.

Although previous studies showed that co-cultures of M. aeruginosa with colony- promoting bacteria were enriched in EPS compared to pure cultures of the same M. aeruginosa strain, it is not known if the excess EPS was produced by the cyanobacteria, the heterotrophs, or both (Shen et al. 2011, Wang et al. 2015). Co-cultures had a higher ratio of cell-bound EPS to soluble EPS than pure M. aeruginosa cultures (Wang et al.

2015), which could be attributed to the two species synthesizing different polysaccharides or to M. aeruginosa changing the composition of its endogenously produced polysaccharides in response to signals from bacterial neighbors. Furthermore, Wang et al.

(2015) observed morphological changes prior to changes in EPS, indicating that EPS differences were not the immediate cause of colony induction.

Understanding the factors that determine Microcystis morphology is a necessary precursor to understanding how Microcystis blooms begin and how to predict their onset.

This is of vital importance as changing climate patterns allow blooms to develop in environments from which they were previously absent (Paerl & Huisman 2009, Paerl &

Paul 2012). Although much has been learned about the effects of Microcystis blooms on other microbes in aquatic ecosystems by providing habitat (Li et al. 2011, Louati et al.

2015, Niu et al. 2011, Parveen et al. 2013a, Parveen et al. 2013b, Shen et al. 2011, Shi et al. 2012) and carbon sources (Casamatta & Wickstrom 2000, Jones et al. 1994,

Maruyama et al 2003, Mou et al. 2013), relatively little is known about how other members of the microbial community affect a key adaptive trait of Microcystis that makes the initiation and maintenance of blooms possible.

36

To examine the relationship between Microcystis and associated bacteria, bacteria were isolated from the 2014 M. aeruginosa bloom in Lake Erie (Francy et al. 2015) to determine if they enhanced the frequency and size of colonies in cultured strains of M. aeruginosa. We hypothesized that morphological effects of these bacteria would be mediated by chemical exudates that initiate changes in Microcystis mucilage and/or surface properties, as were the effects of zooplankton grazers (Ha et al. 2004) and thus would not be dependent on physical contact with heterotrophic bacteria such as those studied by Zhang et al. (2018). Finally, we hypothesized that any morphological differences between toxic and non-toxic M. aeruginosa strains may be caused by microcystin enhancing the effects of colony-promoting exudates, either by promoting additional EPS production as in Gan et al. (2012) or by directly enhancing intercellular adhesion as in Kehr et al. (2006).

Methods

Isolation of Bacteria

In August 2014, a toxic cyanobacterial bloom occurred in Maumee Bay (Lake

Erie, Ohio, USA) during which Microcystis constituted more than 90% of total near shore cyanobacterial biovolume (Francy et al. 2015). To obtain bacteria associated during this

Microcystis bloom, samples were collected from the upper five centimeters of water in

Maumee Bay on August 18, 2014. Cyanobacteria and their associated microorganisms were concentrated by filtering 100 mL samples through 3 m nitrocellulose membranes

(Millipore, Darmstadt, Germany). Organisms collected on membranes were then resuspended in sterile deionized water to dissolve EPS mucilage (Plude et al. 1991).

37

Aliquots of suspension were poured onto agar plates and spread evenly using sterilized metal spreaders. Three growth media were used: standard methods (plate count) agar and

R2A agar (Reasoner & Gelderich 1985) from Difco Laboratories (Detroit, MI, USA), and

0.9 g/L glucose and 1 g/L casamino acids with 15g/L agar. Plates were incubated at room temperature for four weeks in the dark, and colonies were isolated.

16S rRNA Gene Sequencing of Bacteria

To obtain DNA for 16S rRNA gene sequencing, isolates were incubated overnight at 37° C in nutrient broth. Subsequently, DNA was extracted using an UltraClean

Microbial DNA purification kit (MoBio Laboratories, Carlsbad, CA, USA) (La Duc et al.

2009). DNA was amplified with bacterial primer 357f (Muyzer et al. 1993) and universal primer 1391r (Lane et al. 1985). PCR consisted of an initial denaturing step at 96° C for five minutes, followed by thirty cycles of 96o C for fifty seconds, 57° C for one minute, and 72o C for one minute, then a final extension step of 72° C for eight minutes. PCR products were cleaned with an ENZA DNA purification kit (Omega Bio-Tek, Norcross,

GA, USA) and sent to the Ohio State University Molecular and Cellular Imaging Center

(Wooster, OH, USA) for Sanger sequencing. Sequences were opened in Mega 7 (Kumar et al. 2016, Kumar et al. 2004), and used as the template for a BLAST search of the

NCBI database (Altschup et al. 1990). Isolate sequences and the closest matching sequences from the database were aligned with Clustal W (Thompson et al. 1994), and a maximum likelihood tree was constructed using 1000 bootstrap iterations (Tamura et al.

2011).

Preliminary Isolate Screening

38

Isolates were screened for the ability to enhance colony size and frequency in direct co-culture with M. aeruginosa. The two M. aeruginosa strains used were from the

University of Texas Algal Culture Collection: toxic UTEX LB 2385 and non-toxic UTEX

LB 2386. Bacterial isolates were cultured overnight in nutrient broth at 37° C prior to the initiation of the experiment. M. aeruginosa was centrifuged at 2500 x g for 15 minutes and rinsed with BG-11 medium three times, then resuspended in fresh BG-11. Each bacterial isolate was centrifuged for 20 minutes at 4500 x g, rinsed with BG-11 medium three times, and added to M. aeruginosa suspension at a final density of 2 x 106 cells/mL for M. aeruginosa and 1.5 x 108 cells/mL for the bacterial isolate. Control cultures of M. aeruginosa contained 2 x 106 cells/mL with no added bacteria. All experiments were performed in triplicate. Cultures were incubated at 27° C under a 12H:12H light:dark cycle.

M. aeruginosa cells were collected after 48 hours and preserved with Lugol’s iodine for subsequent observation of morphology. Cells were viewed on wet mount slides at 200x with an Olympus BX53 microscope (Olympus, Center Valley, PA, USA), and ten fields per slide were recorded with an Olympus SC100 digital camera. Colonies and unaggregated Microcystis cells were counted manually. Colony frequency was calculated as the ratio of colonies to unaggregated Microcystis cells. Surface area of each

Microcystis colony was measured with MetaMorph version 7.7 (MetaMorph, Inc.,

Nashville, TN, USA). To determine colony size, equivalent spherical diameter (ESD) was calculated from visible surface area. Any replicate in which no colonies were found was recorded as having a frequency and an ESD of zero. Any bacterial isolate that caused

39 a significant increase in frequency or size of colonies in either M. aeruginosa strain compared to control cultures of the same were considered a colony-promoting isolate.

Assessment of Colony-Promoting Effects of Bacterial Exudates

To investigate whether exudates of these bacteria could affect M. aeruginosa morphology without direct contact between the cells, additional experiments were performed. First, cells were centrifuged and rinsed as described above. Then, cyanobacterial cells were resuspended at 2 x 106 cells/mL, and colony-promoting bacteria at 3 x 108 cells/mL. Thirty milliliters of each bacterial isolate suspension were loaded into a Slide-A-Lyzer dialysis cassette (Thermo Scientific, Waltham, MA, USA) with 10,000

Dalton molecular weight cut-off, and each cassette was submerged in a M. aeruginosa suspension. Control cultures of the M. aeruginosa suspension with no amendment.

Cultures were incubated as above for 48 hours. At the end of the experiment, aliquots of

M. aeruginosa suspension were preserved and colonies measured and enumerated as described above.

EPS Quantification

EPS was extracted at the end of the experiment; to separate soluble EPS from cell-bound EPS, aliquots were centrifuged at 2500 x g for 15 minutes (Xu et al. 2013).

Supernatants, containing soluble extracellular polysaccharides (SL-EPS), were transferred to dialysis tubing with a molecular weight cutoff 3,500 Da (Thermo Fisher

Scientific, Waltham, MA). Pellets were resuspended in five mL of 0.05% NaCl solution, incubated in a water bath at 60o C for 40 minutes, and centrifuged at 4500 x g for 60 minutes (Xu et al. 2013). Supernatants, containing the bound fraction of extracellular

40 polysaccharides (BD-EPS), were transferred to dialysis tubing as above. EPS fractions were dialyzed overnight and quantified by the phenol-sulfuric acid method using galactose as a standard (Dubois 1956).

Microcystin Analysis

To quantify microcystin (MC), cultures with the toxic stain were filtered through

GF/C membranes (Whatman, Maidstone, UK). Filtrates, containing extracellular MC, were purified according to the methods of Lawton et al. (1994). To extract intracellular

MC, membranes were submerged for one hour in methanol (Lawton et al. 1994).

Supernatants were decanted, and the process was repeated twice. The methanol was evaporated, and extracts from each subsample were resuspended in methanol and recombined with extracellular MC from the same subsample. After final evaporation of methanol, extracts were resuspended in 500 L methanol per subsample for quantification of MC by HPLC-UV following the methods of Shamsollahi et al. (2015) and using a Shimadzu Prominence HPLC system (Shimadzu, Kyoto, Japan) and standards from Cayman Chemical.

Reflectance Readings

To determine whether enhanced colony frequency or size might be associated with changes in the pigmentation of M. aeruginosa, samples were collected for spectral reflectance analysis. M. aeruginosa suspension from each culture was filtered through a

45 mm diameter GF/F membrane (Whatman, Maidstone, UK). Filters were dried at 60° C for 24 hours prior to freezing at -80° C and subsequently dried for an additional two hours at 60° C. For each filter, percent reflectance was measured for wavelengths 360-730 nm

41 with a CM-2600d spectrophotometer (Konica Minolta, Tokyo, Japan), with an uncontaminated GF/F membrane as a blank. The data were denoised by taking first order derivatives of blank-corrected reflectance readings at ten nm wavelength increment between 400 and 700 nm.

Statistical Analysis

Statistical analyses were performed in R 3.3.2 (R Core Team 2016). Initially, morphological and EPS data were tested for conformity to the assumptions of parametric

ANOVA. Because Shapiro-Wilk testing (Shapiro & Wilk 1965) confirmed that most dependent variables were non-normally distributed, the Kruskal-Wallis test was used as a non-parametric alternative to ANOVA. In this method, data are ranked without regard to group assignment, and a test statistic is calculated based on ranks and compared to a chi- squared probability distribution (Kruskal, Wallis 1952). When this test detected differences among treatments at <0.05, Dunn’s post hoc test was used to analyze pairwise differences between individual treatment and control groups (Dunn 1965). To explore correlations between colony frequency and EPS measurements, the strength of each pairwise relationship was summarized by an adjusted R-squared statistic and an F- test was used to determine its statistical significance.

Principal Component Analysis with varimax rotation (V-PCA) was used to explore the reflectance data, with wavelengths as rows and treatments as columns (Ortiz et al. 2013).

Principal components were extracted from the derivatives of percent reflectance readings, and a scree plot was used to determine which components should be retained for further analysis. Rotated component loadings converted to z-scores and compared to a library of spectral signatures of pigments (Bartley & Scolnik 1995, Gantt 1975, Schagerl &

42

Donabaum, Schagerl et al. 2003). Forward stepwise regression (Hamaker 1962) was used to infer which pigments were likely to have contributed to the observed loadings.

Results

Thirty-two bacteria were isolated and screened for colony-promoting effects. Though occasional small colonies were observed in negative control cultures (Fig. 6A), six isolates induced significantly greater colony frequency, colony size, or both in at least one M. aeruginosa strain (Table 2, Fig. 6B). Two of the colony-promoting isolates were affiliated with Gammaproteobacteria genus Pseudomonas, three with Firmicutes genus

Exiguobacterium, and the last with Bacillus, a Firmicutes genus closely related to

Exiguobacterium (Figure 7). However, many other isolates obtained also belonged to these genera yet did not have significant effects on the morphology of M. aeruginosa.

Subsequently, colony-promoting isolates were used in experiments to examine if exudates would promote colony formation without physical contact between the isolated bacterium and M. aeruginosa. Kruskal-Wallis revealed significant differences in colony frequency among bacterial exudate treatments (Chi-squared=30.409, p=0.0041). Post hoc tests showed that exudates of three isolates (PI-2, PI-6, and EI-23), enhanced the frequency of toxic M. aeruginosa colonies relative to controls (Table 2, Fig. 8A). PI-6 demonstrated a similar effect on the frequency of non-toxic M. aeruginosa colonies, while EI-23 exudates had a marginally significant effect. In contrast, colonies occurred in non-toxic cultures treated with PI-2 exudates at slightly lower frequency than in control cultures. Overall, the colony frequency-promoting effect of PI-2 was strain-specific whereas PI-6 and EI-23 were not.

43

Figure 6. Microcystis aeruginosa Colonies at 200x Magnification. Colonies of M. aeruginosa in negative control cultures (A) and in co-culture with EI-23 (B), a colony- promoting bacterium.

44

Table 2. Colony-Promoting Effects of Bacterial Isolates. The six bacterial isolates that promoted increased size or frequency of M. aeruginosa colonies across a dialysis barrier, presented with their significant (<0.05) effects on M. aeruginosa.

Isolate Difference from Difference from Toxic Control Non-Toxic Control

PI-2 Frequency Mean ESD z=-1.8636 z=-3.1947 p=0.0312 p=0.0007 Max ESD Max ESD z=-2.0632 z=-2.4958 p=0.0195 p=0.0063

PI-6 Frequency Frequency z=-2.0965 z=-1.7637 p=0.0180 p=0.0389

EI-12 Mean ESD z=-2.0632 p=0.0195 Max ESD z=-1.9634 p=0.0248

EI-15 Mean ESD z=-3.1281 p=0.0009 Greater max ESD z=-2.5624 p=0.0052

EI-23 Frequency Frequency z=-2.7621 z=-1.5308 p=0.0029 p=0.0629

BI-3 Mean ESD z=-2.7288 p=0.0032 Greater max ESD z=-1.8303 p=0.0336

45

Figure 7. Phylogeny of Colony-Promoting Isolates and Selected Other Bacteria.

Phylogenetic tree of heterotrophic bacterial isolates from the 2014 Microcystis bloom in

Maumee Bay and their closest matches in the GenBank, based on 16S rDNA sequences, with bootstrap confidence levels greater than 50% are indicated at internodes. Asterisks indicate isolates that promoted increased size or frequency of colonies under standard laboratory conditions.

46

Different bacterial exudate treatments also resulted in different M. aeruginosa colony sizes. Although colonies that received different exudate treatments were similar in median size, there were statistically significant differences in average size (Chi- squared=28.615, p=0.007421) and maximum size (Chi-squared=27.844, p=0.009515).

Investigation of pairwise differences revealed that exudates from four bacterial isolates

(PI-2, EI-12, EI-15, and BI-3) enhanced average colony size in non-toxic M. aeruginosa

(Table 2, Fig. 8B). Unlike the colony frequency-enhancing effect, this effect was always specific to the non-toxic strain of M. aeruginosa (Fig. 8B). The same four isolates all promoted greater maximum colony size in non-toxic M. aeruginosa cultures compared to non-toxic controls, whereas only one isolate, PI-2, was associated with a significant increase in maximum colony size for toxic M. aeruginosa (Table 2, Fig. 8C).

Results suggest a tradeoff between colony frequency and colony size. In control cultures, toxic M. aeruginosa formed larger colonies than non-toxic M. aeruginosa, while colonies occurred more frequently in cultures of the non-toxic strain than in the toxic strain (Fig. 8). Each treatment induced non-toxic M. aeruginosa to form colonies of either greater size or increased frequency compared to non-toxic controls, never both (Table 2,

Fig. 8). Of the three treatments that had observable effects on the morphology of toxic M. aeruginosa, two promoted increased colony frequency compared to toxic controls but smaller average colony size (Table 2, Fig. 8). There was a weak but statistically significant negative correlation between colony frequency and colony size across the entire dataset (Adjusted R-squared=0.07829, F=4.483, p=0.04051, Fig. 9A).

47

Figure 8. Treatment Effects on Morphology. Bar plots of morphological measurements of M. aeruginosa cultures exposed to exudates from bacterial isolates across a dialysis barrier: colony frequency (A), mean colony size (B), and maximum colony size (C).

48

The amount of EPS normalized to estimated number of Microcystis cells differed among treatments (Chi-squared=30.351, p=0.004191), largely due to non-toxic control cultures having higher EPS per cell than cultures treated with PI-2, BI-3, EI-12, or EI-15.

Over the entire dataset, normalized EPS was positively correlated with colony frequency

(Adjusted R-squared=0.1895, F=10.59, p=0.0023). When strains of M. aeruginosa were considered separately, the correlation was stronger among toxic cultures (Adjusted R- squared=0.3906, F=13.82, p=0.0015), whereas there was no significant correlation between frequency and normalized EPS values for non-toxic cultures (Fig. 9). This difference between strains implies that the polysaccharide component of the mucilage is more important to colonial morphology in the toxic strain than in the non-toxic strain.

HPLC-UV analysis of extracts from the cultures with the toxic strain did not show detectable amounts of MC, indicating that different responses of toxic and non-toxic M. aeruginosa strains cannot be attributed the presence of MC in the cultures.

Growth rates of M. aeruginosa cultures also varied among treatments (Chi- squared=31.215, p=0.003136). Non-toxic M. aeruginosa cultures exposed to PI-2, EI-15, and BI-3 grew at significantly higher rates than did control cultures, whereas growth rates for toxic cultures exposed to these and other treatments were all similar to growth rates of control cultures (Table 3). Furthermore, both M. aeruginosa strains exhibited higher growth rates in response to PI-2, EI-15, or BI-3—all isolates associated with increased colony size—than in response to PI-6 or EI-23, the two isolates which enhanced the frequency of colonies with no effect on size.

49

Figure 9. Correlations of Colony Frequency with Colony Size and EPS. Scatter plots of mean colony size (A) and normalized EPS concentration (B) against colony frequency.

50

Table 3. Cyanobacterial Growth Rates Varying with Isolate Treatment. Average daily percent increase in M. aeruginosa cell numbers for cultures receiving different bacterial exudate treatments. Numbers in parentheses indicate standard error. Asterisks indicate significant difference between the treatment and control cultures of the same M. aeruginosa strain (<0.05).

Treatment UTEX 2385 UTEX 2386

Control 1.29 (0.09) 0.65 (0.24)

PI-2 1.53 (0.09) 1.73 (0.07)*

PI-6 0.71 (0.12) 0.65 (0.05)

EI-12 1.30 (0.25) 1.33 (0.16)

EI-15 1.55 (0.17) 1.61 (0.10)*

EI-23 0.65 (0.18) 0.75 (0.35)

BI-3 0.78 (0.21) 1.69 (0.09)*

PCA extracted seven components from the transformed reflectance readings.

These components explained 46.7%, 31.3%, 8.4%, 4.6%, 3.6%, 2.5%, and 1% of variance, respectively, accounting for a total of 98% of variance in the dataset. Average communality for reflectance at wavelengths 400-700 nm was 0.98 out of 1.0, indicating that nearly all the shared variance was explained by the seven extracted components.

Treatment groups had significantly different scores on three of the seven components.

The first component had strong negative loadings at 400-420 nm, 650 nm, and 700 nm and strong positive loadings at 450-500 nm and 680 nm (Fig. 10). Stepwise regression

51 found that this component had a positive relationship with chlorophyll-a, carotenoids, and neoxanthin. Non-toxic M. aeruginosa exposed to PI-2, PI-6, EI-15, EI-23, or BI-3 scored significantly higher on this component than the non-toxic control group (Fig. 11). The second component had strong negative loadings at 620-650 nm and 690 nm as well as strong positive loadings 420-430 nm and at 560-600 nm (Fig. 10). It was negatively correlated with chlorophyll-a, carotenoids, phycocyanin, and peridinin. With the exception of those cultures treated with PI-2, toxic control M. aeruginosa consistently scored lower than non-toxic M. aeruginosa on this component (Fig. 11).

Treatment differences in the scores for this component (Chi-squared=28.907, p=0.007) could be attributed to the differences between strains and to the significant effect of PI-2 on toxic M. aeruginosa (Fig. 11). The fourth component exhibited high positive loadings at 540-550 nm (Fig. 10). It was positively correlated with chlorophyll-a, carotenoids, and fucoxanthin but negatively correlated with chlorophyll-b, myxoxanthophyll, and phycoerythrocyanin. Toxic and non-toxic controls had negative scores on the fourth component, but significantly higher scores (Chi-squared=23.915, p=0.032) were obtained for toxic M. aeruginosa in the PI-6 and EI-15 treatment groups and for non-toxic M. aeruginosa in the PI-2, EI-12, and EI-15 groups (Fig. 11).

52

Figure 10. Components of Reflectance Across the Visible Spectrum. Comparison of varimax-rotated component loadings with respect to wavelength for five principal components extracted from reflectance data.

53

Figure 11. Treatment Differences in Reflectance Components. Bar plot showing significant differences (<0.05) in component scores among treatment groups for

Components 1 (A), 2(B), and 4(C).

54

Discussion

Six bacterial isolates from a cyanoHAB community in Lake Erie enhanced colony size or frequency in at least one strain of M. aeruginosa. Colony-promoting isolates were affiliated with the Gammaproteobacteria genus Pseudomonas or with the Firmicutes genera Bacillus and Exiguobacterium. Likewise, Wang et al. (2015) isolated colony- promoting bacteria from some of these genera; specifically, Bacillus cereus and

Exiguobacterium acetylium from lakes in China. In this context, it is interesting that the colony-promoting bacteria isolated from Lake Erie included B. cereus and

Exiguobacterium, a genus closely related to Bacillus (Wang & Sun 2009).

Other colony-promoting bacteria identified by Wang et al. (2015) were classified as Gammaproteobacteria, as were the remaining colony-promoting isolates in this study.

Microcystis colonies observed by Wang et al. (2015) and Shen et al. (2011) may have been driven not by changes in the Microcystis cells but by their contact with highly adhesive co-cultured bacteria. Some heterotrophic bacteria that attach to Microcystis in the environment exhibit strong adhesive properties and a tendency to autoaggregate, leading some authors to speculate that they may play a role in holding together the large, multispecies colonial assemblages found in blooms (Zhang et al. 2018). However, those bacteria were not tested for the ability to restore colonial morphology to any predominantly unicellular strains of Microcystis. The bacteria that demonstrated autoaggregation belonged to a wider variety of phyla and classes than the known colony- promoting isolates, representing Actinobacteria, Bacteroidetes, and Alphproteobacteria in addition to Firmicutes and Gammaproteobacteria (Zhang et al. 2018). Although none of them were affiliated with the same genera as the known colony-promoters, one of the

55 bacteria with the strongest propensity for autoaggregation and general adhesiveness was

Staphylococcus caprae, belonging to Bacillales, the same order within the Firmicutes as

Bacillus and Exiguobacterium.

Our study is the first to show that exudates from heterotrophic bacteria enhance colony formation in Microcystis even in the absence of physical contact with heterotrophic cells. Genera that exhibited colony-promoting capabilities in this study are known to exude a variety of secondary metabolites, including antimicrobial compounds

(Haas & Défago 2010, Pathak et al. 2013, Molohon et al. 2011, Raaijmakers et al. 2002) and quorum sensing molecules (Biswa & Doble 2013, Fernandes et al. 2018, Jones &

Blaser 2003, Lee & Zhang 2015, Lombardia et al. 2006, Winson et al. 1995). Some antimicrobials enhance aggregation of Microcystis cells when added to cultures in low doses (Tan et al. 2018). Quorum sensing molecules such as the acyl homoserine lactones

(AHLs) produced by some Pseudomonas (Case et al. 2008, Juhas et al. 2005) and

Exiguobacterium (Biswa & Doble 2013) and the autoinducer 2 (AI-2) produced by B. cereus (Auger et al. 2006, Fernandes et al. 2018) take part in regulating EPS production and/or cell aggregation (Keller & Surette 2006, Rickard et al. 2006, Ryan & Dow 2008,

Xue et al. 2015, Zhang et al. 2016). All of the above types of molecules can diffuse through a liquid medium to be taken up by receptive cells that need not have direct contact with the cells that synthesized the molecules. In this study, the dialysis barrier that separated M. aeruginosa from heterotrophic bacteria ensured that the diffusion of cellular exudates was the only means by which the heterotrophs could have influenced morphology of the cyanobacteria.

56

Colony size is an important aspect of Microcystis success; colonies that are large

(80-100 m) can maintain position at the surface (Wu & Kong 2009, Zhu et al. 2014) and escape consumption by zooplankton grazers (Jarvis et al. 1987) and zebra mussels (White et al. 2011, White & Sarnelle 2014). In this study, colony size of non-toxic M. aeruginosa increased from approximately 30 m to 100 m within 48 hours of exposure to PI-2 and

EI-15, suggesting that inter-relationships with bacteria may impact bloom formation.

Because Microcystis blooms are initiated by rapid accumulation of colonies near the water surface (Wu & Kong 2009, Yamamoto et al. 2011, Zhu et al. 2014) and facilitated by consumers that reject Microcystis while feeding on its phytoplankton competitors

(Leitão et al 2018, Vanderploeg et al. 2001, Wang et al. 2010), these bacteria may contribute to the ability of a Microcystis strain to initiate a bloom.

Some isolates had notably different effects on the two M. aeruginosa strains.

Though it was known that different bacteria induced colonial morphology in different

Microcystis morphospecies (Wang et al. 2015), our experiment is the first to reveal different responses by strains of the same morphospecies. Strain-specific enhancement of colony size may have wider ecological consequences. For example, exudates of EI-15 induced non-toxic M. aeruginosa to form larger colonies than toxic M. aeruginosa (Fig.

6), a difference that could make the non-toxic strain less vulnerable than its toxic conspecific to light limitation (Wu & Kong 2009, Yamamoto et al. 2011, Zhu et al. 2014) and consumption (Jarvis et al. 1987, White et al. 2011, White & Sarnelle 2014). Toxic and non-toxic strains of Microcystis typically co-occur in the environment and microcystin concentration is often strongly correlated with the ratio of toxic Microcystis to total Microcystis (Davis et al. 2009, Joung et al. 2011, Kardinaal 2007, Kurmayer et al.

57

2003). The morphological responses observed in this study imply that interactions with heterotrophic bacteria may play a role in determining which Microsystis strains predominate within a bloom.

The prediction that colony-promoting bacteria would promote increased production of EPS was only tentatively supported for toxic M. aeruginosa and not at all supported for non-toxic M. aeruginosa. For toxic M. aeruginosa cultures, colony frequency was correlated with EPS concentration, but treatment differences in EPS concentration were not great enough to be statistically significant. For non-toxic M. aeruginosa, colony frequency was unrelated to EPS, while the relationship between colony size and EPS was the opposite of the relationship predicted. Cultures in which the size of non-toxic M. aeruginosa colonies was significantly enhanced relative to non-toxic control cultures contained significantly less EPS per cell than was found in non-toxic controls. Therefore, the mechanism by which the bacterial isolates in this study promoted the formation and growth of non-toxic M. aeruginosa colonies must not depend upon the adhesive properties of EPS. Perhaps colony-promoting bacteria induce non-toxic M. aeruginosa cells to increase their adhesive potential with cell surface proteins (Kehr et al.

2006, Zilliges et al. 2008). Alternatively, chemical signals from these bacteria may induce specific strains of M. aeruginosa to exude more extracellular proteins, thus increasing the adhesiveness of the mucilage without increasing extracellular polysaccharides (Tan et al. 2018).

The two cyanobacterial strains differed in optical properties as well as in their morphological responses to the treatments. Scores for the first PCA component indicated that all colony-promoting isolates except for EI-12 induced elevated levels of neoxanthin,

58 chlorophyll-a, and miscellaneous carotenoids in non-toxic M. aeruginosa compared to control cultures. Strain-specific differences in scores on the second component showed that toxic cultures generally contained higher levels of phycocyanin, peridinin, chlorophyll-a, and miscellaneous carotenoids than did non-toxic cultures. PI-2 was associated with depletion of these pigments in the toxic strain and higher levels of them in non-toxic M. aeruginosa. Scores for the fourth component implied that PI-2 and EI-12 induced non-toxic M. aeruginosa to become enriched in fucoxanthin, chlorophyll-a, and miscellaneous carotenoids relative to controls and depleted in phycoerythrocyanin, myxoxanthophyll, and chlorophyll-b. Most of the above pigments are typical of cyanobacteria, while chlorophyll-b (Schagerl & Donabaum 2003, Takaichi 2011) and peridinin (de Oliveira et al. 2015) are less common but known to occur at low levels.

Fucoxanthin, however, has previously been reported only in eukaryotic algae (Schagerl et al. 2003, Takaichi 2011).

Notably, all treatments were associated with heightened levels of chlorophyll-a and carotenoids in non-toxic M. aeruginosa, as well as morphological change in this strain. It can be inferred that colony-promoting bacteria induced metabolic changes in non-toxic M. aeruginosa that affected both pigment concentrations and cell-to-cell adhesion. In contrast, only one of the three isolates that promoted colony formation in toxic M. aeruginosa exhibited a simultaneous effect on its pigmentation. This implies that, even when a colony-promoter has similar effects on morphology in toxic and non- toxic strains, as in the case of PI-6, such effects do not necessarily result from identical processes within the cells.

59

Conclusions

Six bacterial isolates from a Microcystis bloom in Lake Erie enhanced the frequency or size of colonies in toxic and/or non-toxic strains of Microcystis aeruginosa.

Four of these isolates induced non-toxic M. aeruginosa to produce colonies sufficiently large to escape from predation and wind-mixing. This demonstrates that bacteria which occur in association with Microcystis blooms can influence the physical characteristics of

Microcystis strains in ways that potentially contribute to the initiation and maintenance of blooms or to the succession of strains within the bloom. Furthermore, bacterial isolates were able to affect the morphology of M. aeruginosa in the absence of direct contact with

M. aeruginosa cells, supporting the hypothesis that heterotrophic bacteria influence

Microcystis morphology via a mechanism mediated by diffusible chemical signals.

However, the additional hypothesis that colony-promoting bacteria would enhance both frequency and size of colonies by inducing any strain of M. aeruginosa to increase production of extracellular polysaccharides was shown to be false. Different responses of toxic and non-toxic M. aeruginosa strains, low EPS concentrations in cultures that exhibited increased colony size, and an overall negative correlation between colony frequency and colony size all imply that the mechanisms by which heterotrophic bacteria promote Microcystis colony formation are more complex and diverse than was initially hypothesized. Morphological responses of the non-toxic strain to colony- promoting bacteria were independent of the adhesive properties of extracellular polysaccharides. Additional study is needed to clarify the mechanisms by which heterotrophic bacteria contribute to Microcystis colony formation and to explore how these interactions play out in complex communities.

60

Chapter 4: AI-2 Quorum Sensing Signal Promotes Colony Formation in Microcystis

aeruginosa

Introduction

Microcystis is a growing environmental and public health problem throughout the world (Lürling et al. 2017, O’Neil et al. 2012, Paerl & Paul 2012) because this widely distributed freshwater cyanobacterium forms nuisance blooms when colonies are sufficiently large and buoyant (Wu & Kong, 2009, Yamamoto et al. 2011, Zhu et al.

2014). Because larger colonies are less vulnerable to sinking (Wu & Kong, 2009,

Yamamoto et al. 2011, Zhu et al. 2014) and grazing by zooplankton (Yang et al. 2008,

Yang & Kong 2012), factors that promote occurrence of colonies and/or increased size of colonies may enhance the ability of Microcystis to initiate and maintain nuisance blooms.

Microcystis colonies consist of many individual cells embedded in extracellular polysaccharide (EPS) mucilage (Reynolds 1981, Xu et al. 2013). In the environment, these colonies are inhabited by smaller microorganisms which live in the mucilage or attached to cell surfaces (Li et al. 2011; Louati et al. 2015; Maruyama et al. 2003, Niu et al. 2011; Parveen et al. 2013a; Parveen et al. 2013b; Shi et al. 2012). However, when

Microcystis is cultured in the laboratory there is typically a morphological shift from the colonial form to dominance by single and double cells with little EPS (Bolch and

Blackburn 1996, Zhang et al. 2007). Colonial morphology can be induced by culturing unicellular Microcystis with heterotrophic bacteria that co-occur with Microcystis blooms

61 in the environment (Shen et al. 2011, Wang et al. 2015). Colony-promoting bacteria are taxonomically diverse, including representatives of the Gram Positive genera Bacillus and Exiguobacterium and Gram Negative bacteria, such as Aeromonas, Enterobacteria, and Shewanella (Wang et al. 2015).

The mechanism by which these bacteria influence the formation of Microcystis colonies has not yet been identified (Shen et al. 2011, Wang et al. 2015). In general, number and size of colonies increase with total amount of EPS (Gan et al. 2012, Li et al.

2013, Ma et al. 2014, Sato et al. 2017, Shen et al. 2011, Xu et al. 2013, Yang et al. 2008).

However, other studies suggest that morphology may depend upon the relative amounts of EPS components (Wang et al. 2015) or interactions with cell surface proteins (Kehr et al. 2006, Zilliges et al. 2008) rather than on total amount of EPS. Furthermore, it remains unknown how heterotrophic bacteria, including representatives of distantly related phyla, such as Gram Negative Proteobacteria and Gram Positive Firmicutes (Wang et al. 2015), induce morphological change in organisms belonging to the phylum Cyanobacteria.

Many interactions among phylogenetically distant bacteria are mediated by diffusible chemical signals via quorum sensing (Bandara et al. 2012, Federle & Bassler

2003, Henke & Bassler 2004, Papenfort & Bassler 2016, Passos da Silva 2017, Pereira et al. 2008, Pereira et al. 2013, Ryan & Dow 2008). Although several quorum sensing signals coordinate interactions among bacteria belonging to different genera,

Autoinducer-2 (AI-2) is the only known signal to be shared by Gram Positive and Gram

Negative bacteria (Federle & Bassler 2003, Henke & Bassler 2004, Papenfort & Bassler

2016, Pereira et al. 2013, Ryan & Dow 2008, Surette et al. 1999). AI-2 plays a role in a variety of bacterial community processes, including regulation of EPS production and

62 initiation of cell aggregation during biofilm formation (DeLisa et al. 2001, Papenfort &

Bassler 2016, Pereira et al. 2013, Ryan & Dow 2008, Xue et al. 2015).

Synthesis of AI-2 is controlled by an enzyme encoded for by the luxS gene, which is widely conserved in bacteria but has not been detected in cyanobacteria (Pérez-

Rodríguez et a. 2015, Rao et al. 2016, Surette et al. 1999, Xavier & Bassler 2003). Some bacteria that cannot produce AI-2 nevertheless exhibit phenotypic responses to AI-2 from other sources (Duan et al. 2003, Pereira et al. 2008, Pereira et al. 2013), demonstrating that these phenotypic changes can be induced by bacterial neighbors. In fact, several

Microcystis strains respond to co-culture with heterotrophic bacteria by forming colonies more frequently than in pure culture (Shen et al. 2011, Wang et al. 2015). However, the chemical signal that induces this change is not known; a range of organic compounds are exuded by the colony-promoting bacteria including potentially AI-2. Some strains (but not all) of Bacillus cereus, a known promotor of colony formation by Microcystis (Wang et al. 2015; Akins, Chapter 3), exhibit AI-2 activity and harbor luxS genes (Auger et al.

2006, Fernandes et al. 2018).

The purpose of this study was to investigate the effects of AI-2 on Microcystis colonial morphology. We hypothesized that AI-2 would promote colony formation in M. aeruginosa by increasing the total amount of EPS per cell and the ratio of cell-bound EPS over soluble EPS. Additionally, we hypothesized that the genes which control synthesis of AI-2 would be present in heterotrophic bacteria that promote colony formation by

Microcystis. Therefore, it was predicted that bacterial isolates previously identified as having colony-promoting abilities would carry homologs to luxS, a gene crucial to AI-2 synthesis.

63

Methods

Two strains of M. aeruginosa, toxic UTEX 2385 and non-toxic UTEX 2386

(University of Texas Algal Culture Collection) were maintained in BG-11 medium at

27oC under a 12H:12H light:dark cycle. Prior to experimentation, cultures were centrifuged at 2500 x g and rinsed with BG-11 three times. Then, cells were resuspended at a density of 2 x 106 cells/mL and treated with 0, 0.042, 0.42, or 4.2, M AI-2 (Omm

Scientific, Dallas, TX, USA). Cultures were incubated at 27oC and 12H:12H light:dark and samples were collected after 24 and 48H.

Morphological Observation

Subsamples for microscopy were treated with Lugol’s iodine and were observed at

200x on wet mount slides with an Olympus BX53 microscope (Olympus, Center Valley,

PA, USA). Colonies and unaggregated cells were counted in ten fields of view per sample, using an Olympus SC100 digital camera (Olympus, Center Valley, PA, USA) and measured in MetaMorph 7.7 (MetaMorph, Inc., Nashville, TN USA). The visible surface area of each colony was used to derive its equivalent spherical diameter (ESD).

Total cell number was estimated by dividing colony surface area by average visible area of a M. aeruginosa cell and adding the number of unaggregated cells.

EPS Extraction and Quantification

Soluble extracellular polysaccharides (SL-EPS) was extracted by centrifuging subsamples collected after 48H of incubation for 15 minutes at 2500 x g (Xu et al. 2013).

Supernatants were transferred to dialysis tubing (molecular weight cutoff 35 kDa,

Thermo Fisher Scientific, Waltham, MA), and pellets were resuspended in 0.05% NaCl.

64

To extract bound extracellular polysaccharides (BD-EPS), resuspended cells were incubated in a 60o C water bath for 40 minutes, then centrifuged for 60 minutes at 4500 x g (Xu et al. 2013). Supernatants containing BD-EPS were transferred to dialysis tubing as above. Both EPS fractions were dialyzed overnight against DI water and quantified by the phenol-sulfuric acid method with a galactose standard (Dubois 1956). Concentrations of soluble, bound, and total EPS were normalized to estimated cell number.

MC Extraction and Quantification

To extract intracellular microcystin (MC), cultures were filtered through GF/C membranes (Whatman, Maidstone, UK), which were submerged for one hour in methanol (Lawton et al. 1994); extraction of the filters was then repeated twice and extracts were pooled. Extracellular MC was extracted from filtrates according to the method of Lawton et al. (1994). Combined MC extracts from each culture were resuspended in methanol and quantified by HPLC-UV following the protocol of

Shamsollahi et al. (2015) with a Shimadzu Prominence HPLC system (Shimadzu, Kyoto,

Japan) and standards from Cayman Chemical (Ann Arbor, MI, USA).

PCR

PCR was used to detect luxS gene homologs in bacterial isolates that were previously demonstrated to promote Microcystis colony formation (Akins, Chapter 3 and Chapter 5).

Two sets of luxS primers, one to amplify luxS gene homologs in Gram negative bacteria and the other targeting luxS in Gram positive bacilli, were used (Santiago-Rodriguez et al. 2014). To verify quality and quantity of DNA, 16S rDNA sequences were amplified with universal bacterial 16S primers to (Santiago-Rodriguez et al. 2014). PCR primers

65 were from Integrated DNA Technologies (Coralville, IA, USA) and amplifications followed the methods of Santiago-Rodrigues et al. (2014). Reactions included negative controls with no template DNA and positive controls with DNA from bacteria known to carry luxS; Bacillus subtilis and E. coli. PCR products were visualized on 1% agarose gels stained with ethidium bromide.

Statistical Analysis

Statistical analyses were performed in R 3.3.2 (R Core Team 2016). The Shapiro-Wilk test (Shapiro & Wilk 1965) was applied to each dependent variable to determine whether data conformed to a normal distribution. Because the data distributions were non-normal, thus violating a fundamental assumption of ANOVA, non-parametric Kruskal-Wallis test

(Kruskal & Wallis 1952) was used. When this test returned a chi-squared value with probability less than 0.05, Dunn’s post hoc test was used to analyze pairwise differences between treatments (Dunn 1965). Correlations between colony frequency and colony size were summarized by adjusted R-squared, and the F test was used to determine the statistical significance of such relationships.

Results

Treatment with AI-2 had a statistically significant effect on growth of M. aeruginosa (Chi-squared=15.227 p=0.03). Analysis of pairwise differences revealed that both strains of M. aeruginosa displayed significantly higher growth rates in cultures that received 0.42 or 4.2 M doses of AI-2 than in controls, while 0.042 M had no significant effect (Table 4).

66

Table 4. Cyanobacterial Growth Rates Varying with AI-2 Treatment. Average percent increase in M. aeruginosa cell number per day, with standard error indicated in parentheses. Asterisks indicate treatment cultures that were significantly different from control cultures of the same M. aeruginosa strain.

Treatment Toxic UTEX 2385 Non-Toxic UTEX 2386

Control 0.37 (0.05) 0.18 (0.17) 0.042 M AI-2 0.64 (0.07) 0.69 (0.29) 0.420 M AI-2 0.85 (0.14)* 0.86 (0.12)* 4.200 M AI-2 0.76 (0.05)* 1.16 (0.19)*

Significant differences in colony frequency were evident after incubation for 24H

(Chi-squared=19.693, p=0.0063). For non-toxic M. aeruginosa, colonies appeared at a frequency more than twice that of control cultures in all three AI-2 treatments (Fig. 12A).

For toxic M. aeruginosa, the highest AI-2 concentration enhanced colony frequency relative to controls, while cultures that received lower initial concentrations of AI-2 did not exhibit any significant difference from the controls or from the high AI-2 toxic cultures. Colonies occurred significantly more frequently in non-toxic M. aeruginosa than in the toxic strain at AI-2 concentrations of 0.42 or 4.2 M. By 48H, there were no longer any significant differences among treatments in colony frequency, and maximum frequency had decreased from 0.10 to 0.06 colonies per unaggregated Microcystis cell.

Treatments differed significantly in average colony size at 24H (Chi-squared=15, p=0.036) and at 48H (Chi-squared=16.667, p=0.0197). Among toxic cultures, at 0.42 M

AI-2 average colony size was significantly greater than that of the control group at 24H

(z=-2.1362, p=0.016) (Fig. 12B). This treatment difference persisted at 48H (z=-2.483,

67 p=0.0065) (Fig. 13A). Among non-toxic cultures, no differences in colony size were observed at 24H, but after 48 H incubation, the average size of colonies in the non-toxic

4.2 M AI-2 treatment group significantly exceeded that of colonies in the non-toxic control group (z=-1.963, p=0.0248) (Fig. 13A). Non-toxic M. aeruginosa colonies were larger on average than those of toxic M. aeruginosa. At 24H, the difference between strains was statistically significant in all treatments except the 0.42 M AI-2, where toxic colonies were at their largest, but by 48H the difference was significant only in the highest AI-2 treatment (Fig. 13A).

At 24 H, a significant positive correlation was observed between colony frequency and average colony size (adjusted R-squared=0.15, F=5.052, p=0.035).

Although differences in frequency were not statistically significant at 48H, the correlation between frequency and average colony size was stronger by this time (adjusted R- squared=0.422, F=30.93, p=1.951 x 10-6).

Maximum colony size also differed among treatments at 24H (Chi- squared=14.773, p=0.039), with toxic M. aeruginosa colonies reaching a greater maximum ESD in cultures treated with 0.42 M AI-2 than in toxic controls (z=-1.963, p=0.0248) (Fig. 12C). By 48H, significant treatment effects were observed for both M. aeruginosa strains (Chi-squared=15.52, p=0.0299). Within each strain, only one experimental treatment differed from the control (Fig. 13B). For non-toxic M. aeruginosa, maximum size of colonies was enhanced in the 4.2 M AI-2 treatment (z=-

2.252, p=0.0122), whereas maximum size of toxic M. aeruginosa colonies was enhanced in the 0.42 M AI-2 treatment (z=-2.771, p=0.0028). Within the 4.2 M AI-2 treatment, non-toxic colonies reached a greater maximum size than toxic colonies.

68

Figure 12. Differences in Morphology at 24 Hours. Bar graphs showing colony frequency (A), mean colony size (B), and maximum colony size (C) at 24 hours’ incubation.

69

Figure 13. Differences in Morphology and EPS at 48 Hours. Bar graphs showing mean colony size (A), maximum colony size (B), and normalized EPS (C) at 48 hours’ incubation.

70

The amount of EPS normalized to cell number differed among treatments (Chi- squared=17.253, p=0.0158). Surprisingly, cultures that had been treated with AI-2 yielded less EPS per cell than controls (Fig. 12C). Toxic M. aeruginosa treated with 0.42

M AI-2 had significantly lower amounts of total EPS per cell than toxic control cultures.

For non-toxic M. aeruginosa, there was less EPS per cell in both the 0.42 M AI-2 treatment and the 4.2 M AI-2 treatment than in non-toxic controls. No differences among treatments were detected in the ratio of soluble to bound EPS, indicating that this was not a determining factor in Microcystis morphology. Morphological changes also did not appear to be dependent on differences in MC production, as HPLC-UV revealed no detectable MC in extracts from any of the culture.

16S rRNA genes of all seven colony-promoting bacterial isolates examined were amplifiable (Akins, Chapter 3). Yet, six of seven isolates did not yield any amplification products when reactions were carried out with luxS primers. A gene sequence homologous to luxS was present in BI-3, an isolate previously identified as Bacillus cereus.

Discussion

In this study, AI-2 affected toxic and non-toxic M. aeruginosa strains in ways that could facilitate formation of a Microcystis bloom, including an increase in specific growth rate and an increase in colony size. Colonies greater than 80 m in diameter are too big for several common grazers (Jarvis et al. 1987) such as zebra mussels (White et al. 2011, White & Sarnelle 2014) to consume. In the environment, size-specific avoidance from grazing and filter-feeding pressures allows large cyanobacterial colonies

71 to form nuisance blooms while competition from other phytoplankton is minimized by consumers (Leitão et al 2018, Vanderploeg et al. 2001, Wang et al. 2010). Furthermore, colonies of diameter 100 m or larger have sufficient buoyancy to maintain position at the water surface despite wind-driven mixing, a key factor in the initial formation of

Microcystis blooms with the accumulation of large colonies at the surface and in the upper water column (Wu & Kong 2009, Yamamoto et al. 2011, Zhu et al. 2014). The impact of AI-2 on colony size has important implications; for example, at some concentrations, the maximum size of toxic M. aeruginosa colonies increased from less than 50 m to more than 100 m.

Although AI-2 is one factor that can contribute to the formation of large

Microcystis colonies, it is clearly not the only factor that contributes to this phenomenon in environments where Microcystis may interact with a variety of bacterial neighbors. Of the colony-promoting bacteria examined, (Akins, Chapter 3 and Chapter 5), only one carried the luxS gene necessary for production of AI-2. Beyond AI-2, these bacteria must influence Microcystis morphology by means of other small molecule signals, such as

AHLs (Bandara et al. 2012, Bassler 1999, Papenfort & Bassler 2016, Passos da Silva et al. 2017, Zhai et al. 2012), oligopeptides (Bassler 1999, Monnet & Gardan 2015, Waters

& Bassler 2005, Xavier & Bassler 2003), or antibiotics (Tan et al. 2018).

The effects of AI-2 on EPS production by Microcystis cultures were very different from of microcystin treatments and bacterial co-cultures which induced colonies in previous studies (Gan et al. 2012, Shen et al. 2011, Wang et al. 2015). AI-2 treatments that had larger colony size also had lower EPS, a finding contrary to prior studies involving co-culture between M. aeruginosa and heterotrophic bacteria (Shen et al. 2011,

72

Wang et al. 2015) but similar to a previous on exudates of colony-promoting bacteria

(Akins, Chapter 3). In earlier experiments, physical contact between M. aeruginosa cells and heterotrophic bacteria made it impossible to determine which species produced most of the EPS (Shen et al. 2011, Wang et al. 2015). However, the co-cultures produced more

EPS than either Microcystis or heterotrophic bacteria alone (Shen et al. 2011, Wang et al.

2015), implying that one or both of the organisms upregulated EPS production in co- cultures. Furthermore, experiments conducted by Gan et al. (2012) showed that pure cultures of Microcystis upregulated their EPS concentration and colony formation simultaneously in response to the presence of microcystin. In this study, however, AI-2 enhanced the size of Microcystis colonies without enhancing EPS concentrations. The mechanism by which AI-2 affects Microcystis morphology does not depend upon adhesive properties of polysaccharides. Instead, this signal may induce Microcystis to allocate resources to other adhesive molecules such as proteins (Kehr et al. 2006, Tan et al. 2018, Zilliges et al. 2008), perhaps at the cost of lower EPS production.

Conclusions

The factors that determine Microcystis morphology are not yet fully understood, and the role of large Microcystis colonies in the formation of nuisance blooms makes further exploration of this topic a matter of environmental, economic, and public health concern as well as an area of scientific interest. This study is the first to show that

Microcystis responds to the common quorum sensing signal AI-2 and the first to report significant effects of a specific chemical product of heterotrophic bacteria on the size of

Microcystis colonies. Toxic and non-toxic strains of Microcystis exhibited different responses to AI-2 at 24 and 48 hours after exposure, and the non-toxic strain appeared to

73 be sensitive to a wider range of AI-2 concentrations than the toxic strain. This study also showed that some bacteria known to promote colony development in one or both of these

Microcystis strains lack a key gene in the AI-2 synthesis process, indicating that AI-2 is not the only product of heterotrophic bacteria that can influence Microcystis morphology.

Further work is needed to clarify the mechanism by which AI-2 influences Microcystis morphology and to identify other factors which contribute to the development of large, bloom-forming colonies.

74

Chapter 5: Interactive Effects of Bacteria and Nutrients on Microcystis Morphology

Introduction

Blooms of potentially toxic cyanobacteria, such as Microcystis, are increasing in lakes and reservoirs around the world due to climate change and rising nutrient loads associated with economic development (Lürling et al. 2017, O’Neil et al. 2012, Paerl &

Paul 2012). The initiation and maintenance of Microcystis blooms are highly influenced by the ability of this cyanobacterium to form large colonies which can escape from sinking and consumption (van Gremberghe et al. 2009, Wu & Kong, 2009, Yamamoto et al. 2011, Yang et al. 2008, Yang & Kong 2012, Zhu et al. 2014). Therefore, advancing understanding of factors that contribute to the formation and growth of Microcystis colonies is an important component of learning how Microcystis blooms develop and what steps can be taken to prevent or mitigate them.

Microcystis colonies consist of many individual cells embedded in extracellular polysaccharide (EPS) mucilage (Reynolds 1981, Xu et al. 2013). Microcystis colonies in the environment are made up of many cyanobacterial cells surrounded by extracellular polysaccharide (EPS) mucilage (Reynolds 1981, Xu et al. 2013) and inhabited by smaller, heterotrophic bacterial cells (Li et al. 2011; Louati et al. 2015; Maruyama et al.

2003, Niu et al. 2011; Parveen et al. 2013a; Parveen et al. 2013b; Shi et al. 2012).

Microcystis strains isolated in the lab gradually lose the ability to maintain colonial form and shift to a different morphology characterized by predominance of single and double

75

cells and by low levels of EPS (Bolch & Blackburn 1996, Zhang et al. 2007). Colonial morphology can be restored by exposure to heterotrophic bacteria isolated from

Microcystis blooms (Shen et al. 2011, Wang et al. 2015).

Although high phosphorus loading is well-established as prerequisite of

Microcystis blooms, and the importance of nitrogen-phosphorus co-limitation has become a focus of study in recent years, little is known about the effects of N:P ratios on

Microcystis morphology. Ma et al. (2014) demonstrated that non-axenic Microcystis samples maintained colonial morphology longer when given nitrogen additions than when the mesocosms received additions of mixed nitrogen and phosphorus or of phosphorus alone. This is consistent with the role of N-rich MC in promoting colonies as described by Gan et al. (2012) and by Sedmak and Elersek (2006). However, culture experiments that explore biotic influences on the morphology of Microcystis are often conducted in BG-11 medium, with its high nitrate concentration of 17.94 mM and high

N:P ratio of 78:1 (Gan et al. 2012, Shen et al. 2011, Wang et al. 2015) or in WC medium with a 1 mM concentration of nitrate and N:P ratio of 20:1 (van Gremberghe et al.). The responses observed under the moderately to highly N-replete conditions of past experiments may not necessarily reflect how Microcystis would respond to these organisms in many environments where blooms occur under N-limitation (Graham et al.

2004, Liu et al. 2011).

The purpose of this study was to explore how nutrient conditions interact in co- culture with colony-promoting bacteria to determine the size and frequency of colonies in toxic and non-toxic Microcystis aeruginosa. I hypothesized that both size and frequency would be greater in M. aeruginosa co-cultures with a colony-promoting bacterium than in

76 control cultures of the same M. aeruginosa strain, across all nutrient treatments. It was further hypothesized that colonies would be larger and more frequent in co-cultures with higher N:P ratios in the growth medium than in more N-limited co-cultures, and that significant interactive effects between N:P ratio and co-culture would be detected.

Finally, it was hypothesized that the effects of N:P ratio and co-culture would be greater for toxic M. aeruginosa than for non-toxic M. aeruginosa.

Methods Selection of Colony-Promoting Bacteria

Heterotrophic bacteria isolated from a toxic cyanobacterial bloom in Maumee Bay

(Lake Erie, Ohio, USA) were evaluated for their ability to promote colony formation in

Microcystis aeruginosa as described in Chapter 2. Among the isolates, two were selected for use in this study: EI-23 (Exiguobacterium indicum) that promoted colony formation in

Microcystis aeruginosa and EI-7 (Exiguobacterium undae) that did not induce any morphological response in the cyanobacterium. EI-23 was used to investigate effects on toxic and non-toxic strains of M. aeruginosa under varying nutrient conditions while EI-7 was used as a positive control.

Preparation of Growth Media

BG-11 (a widely used Microcystis culture medium, Gan et al. 2012, Shen et al.

2011, Srivastava et al. 2016, Wang et al. 2015) was augmented with varying nitrogen and phosphorus concentrations. Two P concentrations were used: 0.23 mM, which is standard for BG-11 medium (Rippika et al. 1979) but extremely high compared to environmental concentrations in lakes (Nurnberg 1996), and 0.023 mM, which is more realistic for lakes where cyanoHABs occur (Nurnberg 1996). The two concentrations of P were used in

77 combination with three different ratios of N to P: the standard BG-11 medium ratio of

78:1 (Rippika et al. 1979), the Redfield ratio of 16:1 (Redfield 1958), and a nitrogen- limited ratio of 4:1. All media were adjusted to a pH of 8.1, as Microcystis colonies in the environment are generally found in waters of pH somewhat higher than neutral (Geng et al. 2013, Ma et al. 2014, Wu et al. 2009, Zhu et al. 2014).

Preparation of Cultures

Microcystis aeruginosa strains UTEX 2385 and UTEX 2386 (University of Texas

Algal Culture Collection) were maintained in standard BG-11 medium at 27o C under a

12:12 light:dark regime. At the beginning of each experiment, M. aeruginosa cultures were centrifuged at 2500x g for 15 minutes and washed three times with the appropriate medium. Pellets were resuspended at 3 x 106 cells/mL in the selected medium.

Exiguobacterium cultures were grown in nutrient broth at 27o C with shaking.

Cultures of EI-7 and EI-23 were centrifuged at 4500x g for 15 minutes and rinsed three times three times with the appropriate medium. Supernatants were discarded, and pellets were resuspended at 3 x 108 cells/mL.

M. aeruginosa suspensions was combined with an equal volume of EI-7 suspension, EI-23 suspension, or sterile growth medium, yielding a final concentration of

1.5 x 106 Microcystis cells/mL in all cultures and 1.5 x 108 Exiguobacterium cells/mL in all bacterial co-cultures. The design was full factorial, with two M. aeruginosa strains, two co-culture treatments plus one negative control (including negative controls), two P concentrations, and three N:P ratios. All flasks were incubated at 27o C on a 12:12 light:dark cycle for 48 hours. One subsample from each culture was stained and

78 preserved with Lugol’s iodine solution for cell imaging, while another subsample was saved for extraction and quantification of extracellular polysaccharides. All experiments were performed in triplicate.

Observation of Morphology

Samples stained with Lugol’s iodine were observed at 200x with an Olympus BX53 microscope (Olympus, Waltham, MA, USA), and ten fields of view were recorded with an Olympus C1000 digital camera (Waltham, MA, USA). Microcystis colonies and unaggregated Microcystis cells were counted manually, and the visible surface area of each particle was measured using Metamorph 7.7 (MetaMorph, Inc., Nashville, TN,

USA). The size of each colony was estimated as equivalent spherical diameter (ESD).

The number of cells in a colony was estimated by dividing its measured area by the average area for a single Microcystis cell.

Determination of Extracellular Polysaccharide Content

Polysaccharides were extracted following the method described by Xu et al.

(2013). Samples were first centrifuged at 2500 x g for 15 minutes, and supernatants, containing soluble extracellular polysaccharides (SL-EPS), were transferred to dialysis tubing with a molecular weight cutoff 35 kDa (Thermo Fisher Scientific, Waltham, MA).

Cells were resuspended in 0.05% NaCl solution and incubated in a 60o C water bath for

40 minutes, then centrifuged for 60 minutes at 4500 x g (Xu et al. 2013). Supernatants, containing bound extracellular polysaccharides (BD-EPS), were transferred to dialysis tubing as above. All EPS samples were submerged in deionized water and dialyzed

79 overnight. Polysaccharide concentrations were quantified by the phenol-sulfuric acid method using galactose as a standard (Dubois 1956).

Statistical Analysis

Because Bartlett’s test (Bartlett 1937) revealed the data distributions were heteroscedastic and the Shapiro-Wilk test showed they were non-normal (Shapiro & Wilk

1965), generalized least squares regression was applied to reduce the spread of the data

(Aitken 1936). Multi-way ANOVAs were applied to the residuals to test for significant main effects of P concentration, initial N:P ratio, Microcystis strain, and co-culture, as well as interactions between factors. All analyses were performed in R version 3.3.3 (R

Core Team 2016).

Results

The frequency with which colonies occurred relative to single Microcystis cells varied significantly among treatments (F=836.6172, p<0.0001). Bacterial co-culture and

N:P ratio each had a significant main effect on frequency, but P concentration and

Microcystis strain did not. However, all four factors were involved in significant interactive effects, including a four-way interactive effect (Table 5).

Across all high P nutrient treatments, regardless of N:P ratio, colony frequencies for toxic M. aeruginosa colonies in co-culture with EI-23 were similar to each other and significantly higher than frequencies in negative controls (Fig. 14A). In standard BG-11 medium, non-toxic M. aeruginosa in co-culture with EI-23 exhibited a colony frequency that was intermediate between that of toxic M. aeruginosa treated with EI-23 and the frequency observed in control cultures. However, the combination of high P x low N:P x

80

EI-23 had a positive effect on the colony frequency of non-toxic M. aeruginosa. The combination of a more environmentally realistic P concentration, a high N:P ratio, and bacterium EI-23 resulted in a significantly lower frequency of toxic colonies compared to the combination of EI-23 and standard BG-11. Furthermore, the frequency of non-toxic colonies in the moderate P x high N:P ratio x EI-23 treatment was lower than that in the high P x low N:P ratio x EI-23 treatment. At moderate P and a 16:1 N:P ratio, EI-23 induced high colony frequency in toxic M. aeruginosa but not on non-toxic M. aeruginosa.

Although bacterium EI-7 was selected for use in positive controls due to its lack of effect on colony frequency in standard BG-11, this isolate induced colonies to form with greater frequency than observed in controls under all other nutrient conditions. The effect was especially strong in cultures of toxic M. aeruginosa under conditions of moderate P and a nutrient ratio of 16:1, where colonies occurred more frequently than in other treatment (Fig. 15).

Differences among treatments in median colony size were also significant

(F=1288.4346, p<0.0001). All four factors had significant main effects, but no statistically significant interactions were detected (Table 5). Although median size tended to be greater in co-culture treatments than in negative control cultures of the same strain under the same nutrient condition, there were no significant differences in median colony size among co-culture treatments.

Average colony size varied across treatments (F=915.1742, p<0.0001), and all four factors had significant main effects, as well as contributing to interactions (Table 5).

In standard BG-11 medium, both toxic and non-toxic M. aeruginosa strains formed larger

81 colonies in co-culture with EI-23 than with EI-7 or in negative controls. Average size of non-toxic M. aeruginosa was significantly larger in cultures with EI-23 under high P and high N:P ratio than in all other treatments (Fig. 14B). Under conditions of high P and low

N:P ratios, colonies formed by toxic M. aeruginosa co-cultured with EI-23 were larger than those formed by non-toxic M. aeruginosa under the same conditions and larger than those in any of the treatments that received moderate P. All cultures in moderate P media were similar to each other in average colony size. Colony size was significantly correlated with colony frequency (adjusted R-squared=0.284, F=43.43, p=1.75 x 10-9).

Maximum colony size differed among treatments (F=505.34, p<0.0001).

Phosphorus concentration, co-culture, and M. aeruginosa strain had significant main effects, but N:P ratio did not (Table 5). As with average size, all four factors contributed to interactive effects. In standard BG-11 medium, toxic M. aeruginosa co-cultured with

EI-23 exhibited greater maximum colony size than toxic M. aeruginosa alone (Fig. 14C).

Non-toxic M. aeruginosa colonies co-cultured with EI-23 reached a maximum size not only greater than that in controls but also greater than toxic M. aeruginosa with the same co-culture. Under all other conditions, maximum size of non-toxic M. aeruginosa colonies was significantly lower. Among high P treatments, maximum size of toxic M. aeruginosa colonies with bacterial co-cultures increased as N:P ratio decreased. At 16

N:1 P, toxic M. aeruginosa colonies with EI-23 reached a size significantly greater than that observed in the moderate P x 4 N:1 P x EI-23 treatment. Under conditions of high P and low N:P, and in the presence of bacterium EI-23, toxic M. aeruginosa colonies reached their greatest size, significantly larger than those in standard BG-11 as well as those in any moderate P treatment.

82

Table 5. Main and Interactive Effects of Four Independent Variables. Measurements of colony morphology presented

with the factors that had significant main or interactive effects on each measurement (=0.05).

Dependent Main Effects Two-Way Interactions Three-Way Four-Way Variable Interactions Interactions Colony Co-Culture P Level x Co-Culture None P Level x Co-Culture Frequency F=143.043, F=12.6921, p<0.0001 x N:P Ratio x Ma. p<0.0001 P Level x N:P Ratio Strain F=3.309, N:P Ratio F=11.1737, p=0.0001 p=0.0152 F=3.423, p=0.0380 P Level x Ma. Strain F=6.1456, p=0.0155 N:P Ratio x Co-Culture F=3.1487, p=0.0192 N:P Ratio x Ma. Strain F=8.2967, p=0.0006 Median Colony P Level None None None Size F=6.467, p=0.0131 Co-Culture F=56.511, p<0.0001 N:P Ratio F=7.504, p=0.0011 Ma. Strain F=13.129, p=0.0005

83

Mean Colony P Level P Level x Co-Culture P Level x N:P Ratio x P Level x Co-Culture Size F=31.955, F=10.7610, p=0.0001 Ma. Strain x N:P Ratio x Ma. p<0.0001 N:P Ratio x Co-Culture F=5.513, p=0.0059 Strain F=3.013, Co-Culture F=4.4019, p=0.0031 Co-Culture x N:P Ratio p=0.0234 F=76.5133, N:P Ratio x Ma. Strain x Ma. Strain p<0.0001 F=4.2421, p=0.0181 F=3.4612, p=0.0121 N:P Ratio F=3.6521, p=0.0393 Ma. Strain F=10.7053, p=0.0016 Maximum P Level P Level x N:P Ratio P Level x N:P Ratio x P Level x Co-Culture Colony Size F=45.9228, F=3.5953, p=0.0325 Ma. Strain x N:P Ratio x Ma. p<0.0001 P Level x Co-Culture F=9.4303, p=0.0002 Strain F=5.6247, Co-Culture F=20.4896, p<0.0001 Co-Culture x N:P Ratio p=0.0005 F=84.1767, N:P Ratio x Co-Culture x Ma. Strain p<0.0001 F=3.9497, p=0.0059 F=5.9712, p=0.0003 Ma. Strain N:P Ratio x Ma. Strain F=11.828, F=11.2372, p=0.0001 p=0.0010

84

Figure 14. Morphological Effects of EI-23. Frequency (A), mean size (B), and maximum size (C) of M. aeruginosa colonies in co-culture with EI-23. Letters mark significant differences (=0.05).

85

Figure 15. Morphological Effects of EI-7. Frequency of M. aeruginosa colonies in co- culture with EI-7. Letters mark significant differences (=0.05).

Extracellular Polysaccharides

ANOVA revealed significant differences in the amount of EPS per M. aeruginosa cell

(F=596.5, p<0.0001) over the entire set of cultures. Post hoc testing for pairwise differences showed that co-cultures containing both M. aeruginosa and either EI-23 or

EI-7 produced EPS in amounts that did not differ from pure cultures that received the same nutrient treatment. Instead, the significant differenced detected by ANOVA reflected differences between non-toxic M. aeruginosa cultures that were incubated with the same bacterial isolate under different nutrient conditions (Fig.16A). This indicates that, within each nutrient treatment, M. aeruginosa cells in pure culture had the same amount of EPS available to them as M. aeruginosa cells co-cultured with colony- promoting bacteria.

86

Figure 16. Effects of EI-23 on EPS. Normalized EPS content (A) and ratio of soluble to bound EPS (B) for co-cultures of non-toxic M. aeruginosa with EI-23. Letters mark significant differences (=0.05).

Similarly, the ratio of soluble EPS to bound EPS varied (F=156.785, p<0.0001), but significant pairwise differences were found only between the different nutrient treatments of non-toxic M. aeruginosa co-cultured with EI-23 (Fig. 16B). The relationship between EPS ratio and colony size depended on the P concentration of the growth medium. There was a positive correlation in high P treatments (Adjusted R- squared=0.1165, F=7.987, p=0.0067) (Fig. 17A) but no significant correlation in moderate P treatments (Fig. 17B).

87

Figure 17. Relationship of Colony Size and EPS Ratio. Correlations of mean M. aeruginosa colony size in growth medium with 0.23 mM P (A) and growth medium with

0.023 mM P (B).

88

Discussion

As anticipated, bacterium EI-23 induced M. aeruginosa to form colonies more frequently than in negative control cultures across all nutrient conditions. However, colony frequency did not vary significantly with N:P ratio. The colony-promoting bacterium had positive effects on colony size only in co-cultures where initial P concentrations extremely high, whereas colonies in co-cultures that received a more moderate dose of P were of similar size to those in negative controls. Colony frequency was similar across all cultures treated with EI-23 and a high initial P concentration, regardless of N:P ratio or potential toxicity of the M. aeruginosa strain. The highest colony frequency among the moderate P treatment groups was associated with an intermediate N:P ratio of 16:1 rather than the N-replete ratio of 78:1. Likewise, bacterium

EI-7 demonstrated a significant frequency-enhancing effect in combination with moderate P and an intermediate N:P ratio. Under these conditions, both isolates induced more frequent colony formation in toxic M. aeruginosa but not in non-toxic M. aeruginosa. Thus, colony-promoting bacteria may render the toxic strain better suited to forming surface blooms in environments with moderately high P levels and a nutrient ratio around 16:1.

Treatment differences in colony size are of interest both for what they reveal about mechanisms of colony formation and for what they imply about the role of colony- promoting bacteria in ecosystems with variable nutrient ratios. In this study as in the pilot study from which bacterial isolates were selected, EI-23 significantly enhanced both frequency and size of colonies in both strains of M. aeruginosa. In contrast, cell-free exudates of the same isolate had no significant effect on size in either strain of M.

89 aeruginosa and only a marginally significant effect on frequency of non-toxic colonies

(Akins, Chapter 3). It can be inferred that, while EI-23 can affect frequency of toxic M. aeruginosa colonies through diffusible chemicals, its influence on colony size is mediated by a mechanism that requires physical contact, such as the direct attachment of

EI-23 cells to M. aeruginosa cells and to each other (Zhang et al. 2018).

There may be ecologically significant consequences to the fact that the combined effects of EI-23 and high P included formation of colonies larger than 150 m in diameter. In the environment, colonies greater than 100 m can maintain positions high in the water column (Wu & Kong 2009, Zhu et al. 2014) and avoid being consumed by zebra mussels (White et al. 2011, White & Sarnelle 2014) and several common zooplankton grazers (Jarvis et al. 1987). Microcystis blooms begin with the size- dependent accumulation of colonies near the surface of a water body is (Wu & Kong

2009, Yamamoto et al. 2011, Zhu et al. 2014), and consumers that reject Microcystis colonies and consume potential competitors contribute to the dominance of Microcystis among phytoplankton (Leitão et al 2018, Vanderploeg et al. 2001, Wang et al. 2010).

Therefore, conditions which drive a strain of M. aeruginosa to produce colonies greater than 100 m across may enable that strain to initiate a bloom or to become competitively dominant over other strains within a pre-existing cyanoHAB.

It is notable that, when the P concentration was high, varying the N:P ratio had strain-specific effects on colony size in the presence of bacterium EI-23. It is common for toxic and non-toxic strains of Microcystis to co-occur in natural environments (Davis et al. 2009, Joung et al. 2011, Kardinaal 2007, Kurmayer et al. 2003), where large colony size provides competitive advantages for Microcystis (Jarvis et al. 1987, Leitão et al

90

2018, Vanderploeg et al. 2001, Wang et al. 2010, White et al. 2011, White & Sarnelle

2014, Wu & Kong 2009, Yamamoto et al. 2011, Zhu et al. 2014). These results imply that, where P is sufficiently available and certain bacteria are present, a high N load would induce the non-toxic strain to form larger colonies and thus render this strain less susceptible to sinking and predation. Under these conditions, the toxic strain would be more vulnerable and presumably less competitive as longs N-replete conditions persist.

Should the system become N-limited while the P load remained high, the toxic M. aeruginosa strain would gain the advantages of large colony size while the non-toxic strain was relatively vulnerable. Real aquatic ecosystems undergo changes in nutrient ratios throughout the season as water flows into and out of the system and nutrients are taken up, utilized, and remineralized within the system. Thus, interactive effects of nutrient ratios and a bacterium such as EI-23 may play a role in determining succession of strains within a Microcystis bloom.

Previous research has indicated that co-cultures of M. aeruginosa with colony- promoting bacteria had a lower ratio of soluble EPS to bound EPS than pure M. aeruginosa cultures (Wang et al. 2015). In this study, such variation was limited to the non-toxic strain of M. aeruginosa and was not significantly different between co-cultures and controls under most nutrient conditions. Furthermore, this ratio tended to be higher in treatments where colonies were larger, rather than lower as in Wang et al. (2015), and uncorrelated with colony frequency. Therefore, it can be inferred that changes in the ratio of SL-EPS to BD-EPS initiated by colony-promoting bacteria are strain-specific and not necessarily the cause of colony formation in co-cultures. Similarly, differences in colony size and frequency could not be attributed to variation in the amount of EPS per

91

Microcystis cell, as this variation limited to only one of M. aeruginosa strains that exhibited morphological differences. Rather than producing excess EPS or inducing

Microcystis cells to upregulate their EPS production, it is possible that some colony- promoting bacteria induce changes in the cell surface properties of Microcystis or upregulate the production of extracellular proteins, as suggested by Tan et al. (2018).

Alternatively, polysaccharides exuded by these bacteria may be chemically distinct from polysaccharides produced by M. aeruginosa in ways that support greater cell-to-cell adhesion like that observed by Zhang et al. (2018).

Conclusions

Co-culture with bacterium EI-23 enhanced colony frequency in toxic and non-toxic strains of M. aeruginosa at two different phosphorus levels and a wide range of N:P ratios. However, colony size was enhanced only in treatments with an extremely high initial concentration of phosphorus. Identity of the M. aeruginosa strain, N:P ratio, and overall P level, as well as co-culture with a colony-promoting bacterium, interacted to determine the size and frequency of colonies in each treatment. Although bacterium EI-7 did not promote colony formation under the nutrient-rich conditions of BG-11 medium, it did have positive effects on colony frequency of toxic M. aeruginosa relative to negative controls under other nutrient conditions which are more likely to be found in natural environments. This study illustrates that the strain-specific effects of colony-promoting bacteria on Microcystis may be exacerbated or mitigated by nutrient availability. Further research is needed to understand how these factors interact to affect Microcystis morphology and, ultimately, the development and succession of Microcystis blooms.

92

Chapter 6: Summary and General Discussion

Interactions between cyanobacteria and co-occurring heterotrophic bacteria are important processes that influence the ecology of cyanobacterial blooms. Once established, blooms can create microenvironments that support bacterial communities which are just as diverse as free-living bacterial communities of the surrounding water column but distinct in composition from free-living communities. Heterotrophic bacteria found in these microenvironments can, in turn, alter the morphology of cyanobacterial strains through a variety of mechanisms and could, potentially, influence cyanobacterial succession through their strain-specific effects. Some heterotrophic bacteria can shape strains of Microcystis in ways that contribute to the formation and persistence of

Microcystis blooms. Further exploration of these interactions may lead to the identification of signature bacterial taxa whose abundance in a nutrient-rich lake presages a Microcystis bloom.

In Chapter 2, analysis of data obtained through high throughput sequencing of bacterial communities from three lakes uncovered evidence that bacterial communities associated with the biomass of cyanobacterial blooms differed from the surrounding bacterioplankton in ways that were consistent across lakes. However, the bloom in

Maumee Bay supported a bacterial community that was different from communities in other lakes throughout the sampling period (Fig. 3). Moreover, the bacterial community associated with this bloom did not differ from the surrounding free-living bacterial

93 community early in the sampling period (Fig. 3). Such results suggest that any community structure uniquely characteristic of microbes living upon cyanoHAB biomass may take weeks to develop after the beginning of the bloom (Francy et al. 2015). Because the Maumee Bay cyanoHAB consisted primarily of Microcystis on the earlier of two sampling dates, these results also indicate that any bacteria necessary to initiate formation of large Microcystis colonies and the start of a Microcystis bloom were not observably more abundant among epibiontic bacteria associated with the bloom biomass than among bacterioplankton in the surrounding water column. Although different cyanobacterial taxonomic groups made up the cyanoHABs in the three lakes, the greatest similarity of cyanobacterial assemblage was observed between Maumee Bay in September and Harsha

Lake in July. Nevertheless, the heterotrophic bacterial communities associated with the cyanobacterial assemblages of these two sampling groups were significantly different, while the very different cyanobacteria of Buckeye Lake supported a CA community like the one in Harsha Lake. This implies that the specific identity of the cyanobacteria is less important in determining CA community composition than lake-specific conditions or the time passed since the cyanoHAB reached a high bloom density.

As reported in Chapter 3, six of thirty-two bacterial strains isolated from the early

Microcystis bloom in Lake Erie promoted colony formation in standard BG-11 medium, compared to only one or two isolates per lake from Taihu, Dongu, and Dianchi (Wang et al. 2015). These six isolates from Lake Erie, as well as a seventh Lake Erie isolate that promoted colony formation under less enriched nutrient conditions, belonged to the families Pseudomonadaceae and Bacillaceae—neither of which was detected in samples from Maumee Bay in the Chapter 2 study whereby 16S rRNA genes were used to assess bacterial community composition. It is possible that the absence of these taxa from the community sequencing data may be simply reflective of low relative abundance in

Maumee Bay. However, there is reason to suspect that the colony-promoting taxa may not have been rare in Maumee Bay at all.

Colony-promoting Pseudomonadaceae were affiliated with the genus

Pseudomonas, and the Bacillaceae included Bacillus cereus and multiple strains affiliated with Exiguobacterium, a genus closely related to Bacillus (Wang & Sun 2009).

Failure to detect these taxonomic groups with high throughput sequencing may have been due to biases of the Illumina sequencing process or the primers used. MiSeq primers, including those for the V4-V5 region, have a reported bias against Bacillus cereus and some bacteria of the genus Pseudomonas leading to underreporting of these groups in laboratory samples where they were known to be present at the same abundance as other organisms that were reported as highly abundant (Fouhy et al. 2016). Bacillus cereus and

Exiguobacterium acetylium were also among the bacteria identified as colony-promoters by Wang et al. (2015). Clearly, the primers typically used to target the V4-V5 region for

MiSeq are not suitable for detecting most of the bacteria that have demonstrated morphological effects on Microcystis. In the future, high throughput sequencing with primers that are sensitive to these taxa may be used to evaluate the abundance of colony- promoting bacteria in at-risk lakes before Microcystis blooms begin. From relative abundances of colony-promoting bacteria, it may be possible to extrapolate the likelihood of a Microcystis bloom developing. Given the strain-specific nature of many of the responses reported, it may also be possible to predict when conditions will favor toxic

Microcystis strains over non-toxic strains.

95

The two strains of Microcystis aeruginosa used in the culture experiments, UTEX

2385 and UTEX 2386, are both established laboratory strains that have long been predominantly unicellular (Baxa et al. 2010, Bozarth et al. 2010, Geh et al. 2010, Ger et al. 2010, Haney & Lampert 2013, Lee et al. 2015, Otten et al. 2015). Though they are visually indistinguishable, UTEX 2385 contains the mcy gene cluster (Baxa et al. 2010,

Lee et al. 2015, Otten et al. 2015) and has been observed to produce the MC-LR congener of microcystin in culture (Haney & Lampert 2015, Yang & Park 2017), whereas the mcy genes are absent from the UTEX 2386 genome (Baxa et al. 2010, Lee et al. 2015, Otten et al. 2015). These strains have often been paired in studies that require comparison between toxic and non-toxic strains that are otherwise similar in phenotype

(Geh et al. 2015, Ger et al. 2010, Yang & Park 2017) or genome (Baxa et al. 2010, Lee et al. 2015, Otten et al. 2015).

Both strains exhibited morphological responses to diffusible chemical signals, as described in Chapters 3 and 4. Signals by which heterotrophic bacteria can influence the morphology of Microcystis include the “universal” quorum sensing signal AI-2 (Federle

& Bassler 2003, Pereira et al. 2013, Rickard et al. 2006, Ryan & Dow 2008) as well as one or more as-yet unidentified chemicals produced independently of AI-2 by Lake Erie bacteria. Although colony frequency and colony size exhibited a positive relationship under AI-2 treatment, the opposite was true for M. aeruginosa cultures exposed to unidentified exudates from Pseudomonas, Bacillus, and Exiguobacterium isolates. Thus, it can be inferred that AI-2 induces changes in M. aeruginosa cells that are different from those induced by the unidentified bacterial exudates.

96

Chemical signals that induced Microcystis aeruginosa to form colonies larger than those in control cultures were also associated with lower levels of extracellular polysaccharide compared to controls (Akins, Chapter 3 and Chapter 4). These results contrast with mechanisms of colony size enhancement discussed in previous literature which attributed colonial morphology to increased availability of polysaccharides (Gan et al. 2012, Shen et al. 2011, Wang et al. 2015). Instead, the findings of this project are more consistent with studies in which cyanobacterial cell were induced to adhere to each other by changes to cell surface properties (Kehr et al. 2006, Tan et al. 2018, Zilliges et al. 2008). Further research is necessary to learn which bacterially generated signals other than AI-2 enhance colonial morphology and what changes they trigger in Microcystis cells that lead to greater cell-to-cell adhesion.

Diffusible chemicals may not be the only means by which heterotrophic bacteria affect

Microcystis morphology. In Chapter 3, colony-promoting bacterium EI-23 affected only colony frequency but not colony size when it was it was separated from M. aeruginosa by a barrier that permitted the passage of diffusible chemicals but not of entire cells.

However, this bacterium enhanced both size and frequency when direct contact with M. aeruginosa cells was possible, as in Chapter 5. This may indicate that a single isolate is capable of increasing adhesion among Microcystis cells by both direct (Zhang et al. 2018) and indirect (Federle & Bassler 2003, Ryan & Dow 2008) mechanisms. Alternatively, the stronger effects of EI-23 in Chapter 5 might have been due to higher cell density in a smaller culture volume or to cells of both species being evenly mixed throughout the culture volume rather than confined to opposite sides of a barrier. Further investigation is needed to determine how colony enhancement works and under what circumstances

97

Microcystis morphology changes in response to bacterial neighbors. Greater understanding of these interactions may enable humans to assess the risk of Microcystis colonies accumulating based on the abundance and physical dispersal of colony- promoting bacteria.

Chapter 5 also showed that the nutrient content of a growth medium can modify the effects of a colony-promoting bacterium on Microcystis morphology. Although media with extremely high levels of phosphorus and nitrogen and a high N:P ratio were used for prior studies (Shen et al. 2011, Wang et al. 2015, Xu et al. 2012), this experiment revealed that lowering the N:P ratio, the absolute concentration of P, or both can magnify or diminish colony size in the presence of EI-23 or even allow a bacterium such as EI-7 which exhibited no effect on Microcystis morphology under typical laboratory conditions to enhance colony frequency. To explore the extent of colony-promoting abilities among

Microcystis-associated bacteria, future studies should assess the morphological responses of Microcystis strains under a variety of nutrient conditions. Attempts to predict the long- term effects of such interactions must consider how changes in nutrient levels over time may affect the organisms.

In all the culture experiments in this project, some treatments induced different morphological responses in the two M. aeruginosa strains, even when EPS concentrations were similar. Unidentified bacterial exudates that promoted increased colony size in the non-toxic strain had no effect on colony size in toxic cultures. AI-2 enhanced colony size in both strains, but the toxic strain responded to a narrower range of

AI-2 concentrations than did the non-toxic strain. Additionally, the nature of the non- toxic cultures’ responses to AI-2 varied over the course of 48 hours in ways the responses

98 of toxic cultures did not. Bacterial co-cultures that induce similar morphological responses in both M. aeruginosa strains under some nutrient conditions can have significantly different effects on the size of toxic and non-toxic colonies at a different nutrient ratio. Strain-specific effects on morphology, especially on colony size, and on the time at which colonies begin to grow larger than 100 m in diameter, raise the possibility that bacterial communities associated with a cyanobacterial bloom may influence succession of strains as the bloom progresses.

During a Microcystis bloom, succession of toxic and non-toxic strains can drive changes in the overall toxicity of the bloom (Bozarth 2010). It has been observed that larger Microcystis colonies, especially those over 100 m in diameter, are better suited to position themselves near the water surface (Wu & Kong 2009, Yamamoto et al. 2011,

Zhu et al. 2014) and avoid consumption (Jarvis et al. 1987, Leitão et al 2018,

Vanderploeg et al. 2001, Wang et al. 2010, White et al. 2011, White & Sarnelle 2014).

Therefore, bacteria that drive one strain of Microcystis, but not another co-occurring strain, to form colonies greater than 100 m in size could determine which strains will have an opportunity to shade out competitors and which strains will be lost to sinking or grazing. Succession of OTUs within the cyanobacteria-associated bacterial community could be a determining factor in the dominance of one Microcystis strain over another, as different Microcystis strains form larger colonies in response to exudates from different bacterial neighbors and different concentrations of AI-2. Changes in availability of N and

P as nutrients are taken up and remineralized may also affect this process, as interactive effects of nutrients and colony-promoting bacteria lead to differences in colony size between toxic and non-toxic Microcystis strains. Future directions must include both

99 environmental studies which take possible effects of bacteria on Microcystis blooms into consideration and laboratory experiments which explore complex assemblages of multiple Microcystis strains and colony-promoting bacteria under controlled conditions.

Humankind is just beginning to understand how heterotrophic bacteria influence the growth habits of cyanobacteria and the roles that such microbial interactions play in the development and persistence of cyanoHABS. As the abiotic conditions that are prerequisites for Microcystis blooms become increasingly widespread, it is important to explore how biotic factors may contribute to or inhibit the dominance of Microcystis, especially toxin-producing Microcystis strains, among the phytoplankton of nutrient-rich lakes. Several strains of heterotrophic bacteria belonging to multiple phyla are now known to induce morphological changes in one or more strains of Microcystis which would confer adaptive advantages upon those Microcystis strains. Moreover, Microcystis aeruginosa undergoes strain-specific morphological changes in response to the widespread quorum sensing molecule AI-2. Thus, any bacteria that produce this molecule can impact the morphology of Microcystis in ways that may potentially determine the relative competitive success of toxic and non-toxic strains. The strain-specific morphological effects of some bacteria can be modified by varying nutrient conditions, raising the possibility that changes in the relative abundance and competitive success of different Microcystis strains over the course of a season may be driven by interactions with colony-promoting bacteria and changes in nutrient availability. Further research must be undertaken to reveal the colony-promoting capabilities of cyanobacteria- associated epibiontic bacteria under a wider range of nutrient conditions and to identify the molecular and genetic basis of colony-promoting exudates beyond AI-2. A future

100 research goal of special practical importance is determining whether the observed differences between UTEX 2385 and UTEX 2386 are specific to these two strains or broadly representative of differences between toxic and non-toxic Microcystis strains.

Studies of communities that include multiple strains of Microcystis, other cyanobacteria, and colony-promoting heterotrophic bacteria will be necessary to reveal how these bacteria might be used to predict the development, succession, and toxicity of cyanoHABs.

101

References

1. Aitken, A. C. (1936). IV.—on least squares and linear combination of

observations. P Roy Soc Edinb: 55, 42-48

2. Allgaier, M., and Grossart, H. P. (2006). Seasonal dynamics and phylogenetic

diversity of free-living and particle-associated bacterial communities in four lakes

in northeastern Germany. Aquat Microb Ecol, 45(2), 115-128

3. Altschup, S. F., Gish, W., Miller, W., Myers, E. W., Lipman, D. J. (1990). Basic

local alignment search tool. J Mol Biol: 215(3), 403-10

4. Anderson, M. J., Ellingsen, K. E., McArdle, B. H. (2006). Multivariate dispersion

as a measure of beta diversity. Ecol Lett 9(6): 683-693

5. Anderson, M. J., and Walsh, D. C. (2013). PERMANOVA, ANOSIM, and the

Mantel test in the face of heterogeneous dispersions: what null hypothesis are you

testing? Ecol Monograph 83(4): 557-574

6. Auger, S., Krin, E., Aymerich, S., Gohar, M. (2006). Autoinducer 2 affects

biofilm formation by Bacillus cereus. Appl Environ Microbiol, 72(1), 937-941

7. Baker, D. B., Confesor, R., Ewing, D. E., Johnson, L. T., Kramer, J. W.,

Merryfield, B. J. (2014). Phosphorus loading to Lake Erie from the Maumee,

Sandusky and Cuyahoga rivers: The importance of bioavailability. J Great Lakes

Res: 40(3), 502-517

8. Bandara, H. M. H. N., Lam, O. L. T., Jin, L. J., Samaranayake, L. (2012).

Microbial chemical signaling: a current perspective. Crit Rev Microbiol: 38(3),

217-249

102

9. Bartlett, M. S. (1937). Properties of sufficiency and statistical tests. Proc R Soc

Lond A: 160(901), 268-282

10. Basak, P., Majumder, N. S., Nag, S., Bhattacharyya, A., Roy, D., Chakraborty,

A., SenGupta, S., Roy, A., Mukherjee, A., Pattanayak, R., Ghosh, A.,

Chattopadhyay, D., Bhattacharyya, M. (2015). Spatiotemporal analysis of

bacterial diversity in sediments of Sundarbans using parallel 16S rRNA gene tag

sequencing. Microb Ecol: 69(3), 500-511

11. Bassler, B. L. (1999). How bacteria talk to each other: regulation of gene

expression by quorum sensing. Curr Opin Microbiol: 2(6), 582-587

12. Bassler, B. L., Wright, M., Silverman, M. R. (1994). Multiple signalling systems

controlling expression of luminescence in Vibrio harveyi: sequence and function

of genes encoding a second sensory pathway. Mol Microbiol: 13(2), 273-286

13. Baxa, D. V., Kurobe, T., Ger, K. A., Lehman, P. W., & Teh, S. J. (2010).

Estimating the abundance of toxic Microcystis in the San Francisco Estuary using

quantitative real-time PCR. Harmful Algae: 9(3), 342-349

14. Beaulieu, J. J., Smolenski, R. L., Nietch, C. T., Townsend-Small, A., Elovitz, M.

S. (2014). High methane emissions from a midlatitude reservoir draining an

agricultural watershed. Environ Sci Technol: 48(19), 11100-11108

15. Biswa, P., and Doble, M. (2013). Production of acylated homoserine lactone by

Gram-positive bacteria isolated from marine water. FEMS Microbiol Lett: 343(1),

34-41

103

16. Bolch, C.J.S. and Blackburn, S.I. (1996) Isolation and purification of Australian

isolates of the toxic cyanobacterium Microcystis aeruginosa Kutz. J Appl Phycol:

8, 5-13

17. Bozarth, C. S., Schwartz, A. D., Shepardson, J. W., Colwell, F. S., Dreher, T. W.

(2010). Population turnover in a Microcystis bloom results in predominantly

nontoxigenic variants late in the season. Appl Environ Microbiol: 76(15), 5207-

5213

18. Bray, J. R., and Curtis, J. T. (1957). An ordination of the upland forest

communities of southern Wisconsin. Ecol Monograph, 27(4), 325-349

19. Burkert, U., Hyenstrand, P., Drakare, S., Blomqvist, P. (2001). Effects of the

mixotrophic flagellate Ochromonas sp. on colony formation in Microcystis

aeruginosa. Aquat Ecol: 35(1), 11-17

20. Cai, H., Jiang, H., Krumholz, L. R., Yang, Z. (2014). Bacterial community

composition of size-fractioned aggregates within the phycosphere of

cyanobacterial blooms in a eutrophic freshwater lake. PloS One: 9(8), e102879

21. Caiola, M. G., and Pellegrini, S. (1984). Lysis of Microcystis aeruginosa (Kütz.)

by Bdellovibrio‐like bacteria. J Phycol: 20(4), 471-475

22. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D.,

Costello, E.K., Fierer, N., Gonzalez Pena, A., Goodrich, J.K., Gordon, J.I.,

Huttley, G.A., Kelley, S.T., Knights, D., Koenig, J.E., Ley, R.E., Lozupone, C.A.,

McDonald, D., Muegge, B.D., Pirrung, M., Reeder, J., Sevinsky, J.R., Turnbaugh,

P.J., Walters, W.A., Widman, J., Yatsunenko, T., Zaneveld, J., Knight, R. (2010).

104

QIIME allows analysis of high-throughput community sequencing data. Nat

Methods: 7, 335-336

23. Casamatta, D. A., and Wickstrom, C. E. (2000). Sensitivity of two disjunct

bacterioplankton communities to exudates from the cyanobacterium Microcystis

aeruginosa Kützing. Microb Ecol: 40(1), 64-73

24. Case, R. J., Labbate M., Kjelleberg, S. (2008). AHL-driven quorum-sensing

circuits: their frequency and function among the Proteobacteria. ISME J: 2(4),

345

25. Cottrell, M. T., and Kirchman, D. L. (2000). Natural assemblages of marine

Proteobacteria and members of the Cytophaga-Flavobacter cluster consuming

low-and high-molecular-weight dissolved organic matter. Appl Environ

Microbiol: 66(4), 1692-1697

26. Crump, B. C., Armbrust, E. V., Baross, J. A. (1999). Phylogenetic analysis of

particle-attached and free-living bacterial communities in the Columbia River, its

estuary, and the adjacent coastal ocean. Appl Environ Microbiol: 65(7), 3192-

3204

27. Davidov, Y., and Jurkevitch, E. (2004). Diversity and evolution of Bdellovibrio-

and-like organisms (BALOs), reclassification of Bacteriovorax starrii as

Peredibacter starrii gen. nov., comb. nov., and description of the Bacteriovorax–

Peredibacter clade as Bacteriovoracaceae fam. nov. Int J Syst Evol Microbiol:

54(5), 1439-1452

28. Davis, T. W., Berry, D. L., Boyer, G. L., Gobler, C. J. (2009). The effects of

temperature and nutrients on the growth and dynamics of toxic and non-toxic

105

strains of Microcystis during cyanobacteria blooms. Harmful Algae: 8(5), 715-

725

29. Degnan, P.H., and Ochman, H. (2012). Illumina-based analysis of microbial

community diversity. ISME J: 6, 183-194

30. DeLisa, M. P., Wu, C. F., Wang, L., Valdes, J. J., Bentley, W. E. (2001). DNA

Microarray-Based Identification of Genes Controlled by Autoinducer 2-

Stimulated Quorum Sensing in Escherichia coli. J Bacteriol: 183(18), 5239-5247

31. de Oliveira, V. E., Miranda, M. A. N., Soares, M. C. S., Edwards, H. G., de

Oliveira, L. F. C. (2015). Study of carotenoids in cyanobacteria by Raman

spectroscopy. Spectrochim Acta A: Mol Biomol Spectrosc: 150, 373-380

32. Duan, K., Dammel, C., Stein, J., Rabin, H., Surette, M. G. (2003). Modulation of

Pseudomonas aeruginosa gene expression by host microflora through interspecies

communication. Mol Microbiol: 50(5), 1477-1491

33. Dubois, M., Gilles, K. A., Hamilton, J. K., Roberts, P. A., Smith, F. (1956).

Phenol sulphuric acid method for carbohydrate determination. Ann Chem: 28,

350-359

34. Dunn, R. M. and Manoylov, K. M. (2016). The Effects of Initial Cell Density on

the Growth and Proliferation of the Potentially Toxic Cyanobacterium

Microcystis aeruginosa. J Environ Protection: 7(09), 1210

35. Eiler, A., and Bertilsson, S. (2004). Composition of freshwater bacterial

communities associated with cyanobacterial blooms in four Swedish lakes.

Environ Microbiol: 6(12), 1228-1243

106

36. Federle, M. J., and Bassler, B. L. (2003). Interspecies communication in bacteria.

J Clin Invest: 112(9), 1291-1299

37. Fernandes, M. D. S., Esper, L. M. R., Kabuki, D. Y., Kuaye, A. Y. (2018).

Quorum sensing in Enterococcus faecium, Enterococcus faecalis and Bacillus

cereus strains isolated from ricotta processing. Ciênc Rural, 48(2)

38. Fouhy, F., Clooney, A. G., Stanton, C., Claesson, M. J., Cotter, P. D. (2016). 16S

rRNA gene sequencing of mock microbial populations-impact of DNA extraction

method, primer choice and sequencing platform. BMC microbiology, 16(1), 123.

39. Francy, D. S., Brady, A. M., Ecker, C. D., Graham, J. L., Stelzer, E. A.,

Struffolino, P., Dwyer, D.F., Loftin, K. A. (2016). Estimating microcystin levels

at recreational sites in western Lake Erie and Ohio. Harmful Algae: 58, 23-34

40. Francy, D.S., Graham, J.L., Stelzer, E.A., Ecker, C.D., Brady, A.M.G.,

Struffolino, P., Loftin, K.A. (2015) Water quality, cyanobacteria, and

environmental factors and their relations to microcystin concentrations for use in

predictive models at Ohio Lake Erie and inland lake recreational sites, 2013–14:

U.S. Geological Survey Scientific Investigations Report 2015–5120, 58 p.,

http://dx.doi.org/10.3133/sir20155120

41. Gan, N., Xiao, Y., Zhu, L, Wu, Z., Liu, J., Hu, C., Song, L. (2012) The role of

microcystins in maintaining colonies of bloom-forming Microcystis spp. Environ

Microbiol: 14(3), 730-742

42. Geh, E. N., Ghosh, D., McKell, M., de la Cruz, A. A., Stelma, G., Bernstein, J. A.

(2015). Identification of Microcystis aeruginosa peptides responsible for allergic

sensitization and characterization of functional interactions between

107

cyanobacterial toxins and immunogenic peptides. Environ Health Perspect:

123(11), 1159

43. Ger, K. A., Arneson, P., Goldman, C. R., Jooteh, S. (2010). Short Communication

Species specific differences in the ingestion of Microcystis cells by the calanoid

copepods Eurytemora affinis and Pseudodiaptomus forbesi. J Plankt Res: 0(0), 1-

6

44. Graham, J. L., Jones, J. R., Jones, S. B., Downing, J. A., Clevenger, T. E. (2004).

Environmental factors influencing microcystin distribution and concentration in

the Midwestern United States. Water Res: 38(20), 4395-4404

45. Ha, K., Jang, M. H., Takamura, N. (2004). Colony formation in planktonic algae

induced by zooplankton culture media filtrate. J Freshwater Ecol: 19(1), 9-16

46. Haas, D., and Défago, G. (2005). Biological control of soil-borne pathogens by

fluorescent pseudomonads. Nat Rev Microbiol: 3(4), 307

47. Haney, J. F., and Lampert, W. (2013). Spatial avoidance of Microcystis

aeruginosa by Daphnia: fitness consequences and evolutionary implications.

Limnol Oceanogr: 58(6), 2122-2132

48. Imanishi, S., Kato, H., Mizuno, M., Tsuji, K., Harada, K. I. (2005). Bacterial

degradation of microcystins and nodularin. Chem ResToxicol: 18(3), 591-598

49. Jarvis, A. C., Hart, R. C., Combrink, S. (1987). Zooplankton feeding on size

fractionated Microcystis colonies and Chlorella in a hypertrophic lake

(Hartbeespoort Dam, South Africa): implications to resource utilization and

zooplankton succession. J Plankton Res: 9(6), 1231-1249

108

50. Jiang, L., Yang, L., Xiao, L., Shi, X., Gao, G., Qin, B. (2007) Quantitative studies

on phosphorus transference occurring between Microcystis aeruginosa and its

attached bacterium (Pseudomonas sp.). Hydrobiologia: 581, 161-165

51. Jones, G. J., Bourne, D. G., Blakeley, R. L., Doelle, H. (1994). Degradation of the

cyanobacterial hepatotoxin microcystin by aquatic bacteria. Natural Toxins: 2(4),

228-235

52. Joung, S. H., Oh, H. M., Ko, S. R., Ahn, C. Y. (2011). Correlations between

environmental factors and toxic and non-toxic Microcystis dynamics during

bloom in Daechung Reservoir, Korea. Harmful Algae: 10(2), 188-193

53. Juhas, M., Eberl, L., Tümmler, B. (2005). Quorum sensing: the power of

cooperation in the world of Pseudomonas. Environ Microbiol: 7(4), 459-471

54. Kardinaal, W. E. A., Janse, I., Kamst-van Agterveld, M. P., Meima, M., Snoek, J.,

Mur, L. R., Huisman, J., Zwart, G. Visser, P. M. (2007). Microcystis genotype

succession in relation to microcystin concentrations in freshwater lakes. Aquat

Microb Ecol: 48

55. Kehr, J. C., Zilliges, Y., Springer, A., Disney, M. D., Ratner, D. D., Bouchier, C.,

Seeberger, P.H., Tandeau de Marsac, N. Dittmann, E. (2006). A mannan binding

lectin is involved in cell–cell attachment in a toxic strain of Microcystis

aeruginosa. Mol Microbiol: 9(3), 893-906

56. Keller, L. and Surette, M. G. (2006). Communication in bacteria: an ecological

and evolutionary perspective. Nat Rev Microbiol: 4(4), 249

57. Kim, J. M., Roh, A. S., Choi, S. C., Kim, E. J., Choi, M. T., Ahn, B. K., Kim, S.

K., Lee, Y. H., Joa, J. H., Kang, S. S., Lee, S. A., Ahn, J. H., Song, J., Weon, H.

109

Y. (2016). Soil pH and electrical conductivity are key edaphic factors shaping

bacterial communities of greenhouse soils in Korea. J Microbiol: 54(12), 838-845

58. Kirchman, D. L. (2002). The ecology of Cytophaga–Flavobacteria in aquatic

environments. FEMS Microbiol Ecol: 39(2), 91-100

59. Kou, W., Zhang, J., Lu, X., Ma, Y., Mou, X., Wu, L. (2016). Identification of

bacterial communities in sediments of Poyang Lake, the largest freshwater lake in

China. SpringerPlus: 5(1), 401

60. Kruskal, J.B. (1964). Nonmetric multidimensional scaling: a numerical method.

Psychometrika: 29, 115-129

61. Kruskal, W. H., and Wallis, W. A. (1952). Use of ranks in one-criterion variance

analysis. J Amer Stat Assoc: 47(260), 583-621

62. Kumar, S., Stecher, G., Tamura, K. (2016). MEGA7: molecular evolutionary

genetics analysis version 7.0 for bigger datasets. Mol Biol Evol: 33(7), 1870-1874

63. Kumar, S., Tamura, K., Nei, M. (2004). MEGA3: integrated software for

molecular evolutionary genetics analysis and sequence alignment. Brief

Bioinform: 5(2), 150-163

64. Kurmayer, R., Christiansen, G., Chorus, I. (2003). The abundance of microcystin-

producing genotypes correlates positively with colony size in Microcystis sp. and

determines its microcystin net production in Lake Wannsee. Appl Environ

Microbiol: 69(2), 787-795

65. Lawton, L. A., Edwards, C., Codd, G. A. (1994). Extraction and high-

performance liquid chromatographic method for the determination of

microcystins in raw and treated waters. Analyst: 119(7), 1525-1530

110

66. Lear, G., Niyogi, D., Harding, J., Dong, Y., Lewis, G. (2009). Biofilm bacterial

community structure in streams affected by acid mine drainage. Appl Environ

Microbiol: 75(11), 3455-3460

67. Lee, T. A., Rollwagen-Bollens, G., Bollens, S. M., Faber-Hammond, J. J. (2015).

Environmental influence on cyanobacteria abundance and microcystin toxin

production in a shallow temperate lake. Ecotoxicol Environ Safety: 114, 318-325

68. Leitão, E., Ger, K. A., Panosso, R. (2018). Selective Grazing by a Tropical

Copepod (Notodiaptomus iheringi) Facilitates Microcystis Dominance. Front

Microbiol: 9, 301

69. Li, H., Xing, P., Chen, M., Bian, Y., Wu, Q. L. (2011). Short-term bacterial

community composition dynamics in response to accumulation and breakdown of

Microcystis blooms. Water Res: 45(4), 1702-1710 Li, M., Zhu, W., Guo, L., Hu,

J., Chen, H., & Xiao, M. (2016). To increase size or decrease density? Different

Microcystis species has different choice to form blooms. Sci Rep: 6, 37056

70. Li, M., Zhu, W., Guo, L., Hu, J., Chen, H., Xiao, M. (2016). To increase size or

decrease density? Different Microcystis species has different choice to form

blooms. Sci Rep: 6, 37056

71. Louati, I., Pascault, N., Debroas, D., Bernard, C., Humbert, J.-F., Leloup, J.

(2015). Structural Diversity of Bacterial Communities Associated with Bloom-

Forming Freshwater Cyanobacteria Differs According to the Cyanobacterial

Genus. PLoS ONE: 10(11), e0140614

72. Lürling, M., Van Oosterhout, F., Faassen, E. (2017). Eutrophication and Warming

Boost Cyanobacterial Biomass and Microcystins. Toxins, 9(2), 64

111

73. Ma, J., Brookes, J. D., Qin, B., Paerl, H. W., Gao, G., Wu, P., Zhang, W., Deng,

J., Zhu, G., Zhang, Y., Xu, H., Niu, H. (2014). Environmental factors controlling

colony formation in blooms of the cyanobacteria Microcystis spp. in Lake Taihu,

China. Harmful Algae: 31, 136-142

74. Maruyama, T., Kato, K., Yokoyama, A., Tanaka, T., Hiraishi, A., Park, H. D.

(2003). Dynamics of microcystin-degrading bacteria in mucilage of Microcystis.

Microb Ecol: 46(2), 279-288.

75. Michalak, A. M., Anderson, E. J., Beletsky, D., Boland, S., Bosch, N. S.,

Bridgeman, T. B., Chaffin, J.D., Cho, K., Confesor, R., Daloglu, I., DePinto, J.

V., Evans, M.A., Fahnenstiel, G.L., He, L., Ho, J.C., Jenkens, L., Johengen, T.H.,

Kuo, K.C., LaPorte, E., Liu, X., McWilliams, M.R., Moore, M.R., Posselt, D.J.,

Richards, R.P., Scavia, D., Steiner, A.L., Verhamme, E., Wright, D.M., Zagorski,

M.A. (2013). Record-setting algal bloom in Lake Erie caused by agricultural and

meteorological trends consistent with expected future conditions. Proc Natl Acad

Sci: 110(16), 6448-6452

76. Mielke, P.W. (1984) Meteorological applications of permutation techniques based

on distance functions. In: Krishnaiah, P.R., and Sen, P.K., (eds) Handbook of

Statistics, 4th end., Elsevier Science, New York pp. 813-830

77. Molohon, K. J., Melby, J. O., Lee, J., Evans, B. S., Dunbar, K. L., Bumpus, S. B.,

Kelleher, N.L., Mitchell, D. A. (2011). Structure determination and interception

of biosynthetic intermediates for the plantazolicin class of highly discriminating

antibiotics. ACS Chem Biol: 6(12), 1307-1313

112

78. Monnet, V. and Gardan, R. (2015). Quorum‐sensing regulators in G ram‐positive

bacteria:‘cherchez le peptide’. Mol Microbiol: 97(2), 181-184

79. Moorhead, D., Bridgeman, T., Morris, J. (2008). Changes in water quality of

Maumee Bay 1928-2003. In: Munawar, M., and Heath, R. T. (eds) Checking the

Pulse of Lake Erie, Ecovision World Monograph Series, AEHMS, pp. 123-158

80. Mou, X., Lu, X., Jacob, J., Sun, S., Heath, R. (2013) Metagenomic identification

of bacterioplankton taxa and pathways involved in microcystin degradation in

Lake Erie. PLoS ONE: 8(4), e61890

81. Murphy, J., and Riley, J. (1962). A modified single solution method for

determination of phosphate in natural waters. Analytica Chimica Acta: 37, 31-36

82. Muyzer, G., De Waal, E. C., Uitterlinden, A. G. (1993). Profiling of complex

microbial populations by denaturing gradient gel electrophoresis analysis of

polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ

Microbiol: 59(3), 695-700

83. Niu, Y., Shen, H., Chen, J., Xie, P., Yang, X., Tao, M., Ma, Z., Qi, M. (2011).

Phytoplankton community succession shaping bacterioplankton community

composition in Lake Taihu, China. Water Res: 45(14), 4169-4182

84. Ohio Environmental Protection Agency, 2014. State of Ohio—Harmful Algal

Bloom Response Strategy for Recreational Waters.

http://epa.ohio.gov/habalgae.aspx (Accessed June 2014)

85. O’Neil, J. M., Davis, T. W., Burford, M. A., Gobler, C. J. (2012). The rise of

harmful cyanobacteria blooms: the potential roles of eutrophication and climate

change. Harmful Algae: 14, 313-334

113

86. Ortiz, J. D., Witter, D. L., Ali, K. A., Fela, N., Duff, M., Mills, L. (2013).

Evaluating multiple colour-producing agents in Case II waters from Lake Erie. Int

J Remote Sens: 34(24), 8854-8880

87. Otten, T. G., Crosswell, J. R., Mackey, S., Dreher, T. W. (2015). Application of

molecular tools for microbial source tracking and public health risk assessment of

a Microcystis bloom traversing 300 km of the Klamath River. Harmful Algae: 46,

71-81

88. Paerl, H. W. (1996). Microscale physiological and ecological studies of aquatic

cyanobacteria: macroscale implications. Microsc Res Tech: 33(1), 47-72

89. Paerl, H.W., and Huisman, J. (2009). Climate change: a catalyst for global

expansion of harmful cyanobacterial blooms. Environ Microbiol Rep: 1(1), 27-37

90. Paerl, H.W., and Paul, V.J. (2012). Climate change: Links to global expansion of

harmful cyanobacteria. Water Res: 46, 1349-1363

91. Papenfort, K. and Bassler, B. L. (2016). Quorum sensing signal–response systems

in Gram-negative bacteria. Nat Rev Microbiol: 14(9), 576

92. Parikh, A., and Madamwar, D. (2006). Partial characterization of extracellular

polysaccharides from cyanobacteria. Bioresour Technol: 97(15), 1822-1827

93. Parveen, B., Ravet, V., Djediat, C., Mary, I., Quiblier, C., Debroas, D., Humbert,

J. F. (2013). Bacterial communities associated with Microcystis colonies differ

from free‐living communities living in the same ecosystem. Environ Microbiol

Rep: 5(5), 716-724

94. Parveen, B., Mary, I., Vellet, A., Ravet, V., Debroas, D. (2013). Temporal

dynamics and phylogenetic diversity of free-living and particle-associated

114

Verrucomicrobia communities in relation to environmental variables in a

mesotrophic lake. FEMS Microbiol Ecol: 83, 189–201

95. Passos da Silva, D., Schofield, M. C., Parsek, M. R., Tseng, B. S. (2017). An

update on the sociomicrobiology of quorum sensing in gram-negative biofilm

development. Pathogens: 6(4), 51

96. Pathak, R., Kumar, R., Gautam, H. K. (2013). Cross-species induction and

enhancement of antimicrobial properties in response to gamma irradiation in

Exiguobacterium sp. HKG 126. Ind J Microbiol: 53(2), 130-136

97. Pereira, C. S., McAuley, J. R., Taga, M. E., Xavier, K. B., Miller, S. T. (2008).

Sinorhizobium meliloti, a bacterium lacking the autoinducer‐2 (AI‐2) synthase,

responds to AI‐2 supplied by other bacteria. Mol Microbiol: 70(5), 1223-1235

98. Pereira, C. S., Thompson, J. A., Xavier, K. B. (2013). AI-2-mediated signalling in

bacteria. FEMS Microbiol Rev: 37(2), 156-181

99. Pereira, S., Zille, A., Micheletti, E., Moradas-Ferreira, P., De Philippis, R.,

Tamagnini, P. (2009). Complexity of cyanobacterial exopolysaccharides:

composition, structures, inducing factors and putative genes involved in their

biosynthesis and assembly. FEMS Microbiol Revi: 33(5), 917-941

100. Pérez-Rodríguez, I., Bolognini, M., Ricci, J., Bini, E., Vetriani, C. (2015). From

deep-sea volcanoes to human pathogens: a conserved quorum-sensing signal in

Epsilonproteobacteria. ISME J: 9(5), 1222

101. Plude, J. L., Parker, D. L., Schommer, O. J., Timmerman, R. J., Hagstrom, S. A.,

Joers, J. M., Hnasko, R. (1991). Chemical characterization of polysaccharide from

115

the slime layer of the cyanobacterium Microcystis flos-aquae C3-40. Appl

Environ Microbiol: 57(6), 1696-1700

102. R Core Team (2016). R: A language and environment for statistical computing.

R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-

project.org/

103. Raaijmakers, J. M., Vlami, M., De Souza, J. T. (2002). Antibiotic production by

bacterial biocontrol agents. Antonie Leeuwenhoek: 81(1-4), 537

104. Rao, R. M., Pasha, S. N., Sowdhamini, R. (2016). Genome-wide survey and

phylogeny of S-Ribosylhomocysteinase (LuxS) enzyme in bacterial genomes.

BMC Genomics: 17(1), 742

105. Reasoner, D. J., and Geldreich, E. E. (1985). A new medium for the

enumeration and subculture of bacteria from potable water. Appl Environ

Microbiol: 49(1), 1-7

106. Reynolds, C. S., Jaworski, G. H. M., Cmiech, H. A., Leedale, G. F. (1981). On

the annual cycle of the blue-green alga Microcystis aeruginosa Kütz. emend.

Elenkin. PhilosTrans R Soc Lond Ser B: 419-477

107. Rickard, A.H., Paler, R.J., Blehert, D.S., Campagna, S.R., Semmelhack, M.F.,

Egland, P.G., Bassler, B.L., Kolenbrander, P.E. (2006) Autoinducer 2: a

concentration-dependent signal for mutualistic bacterial biofilm growth. Mol

Microbiol: 60(6), 1446-1456

108. Ringuet, S., Sassano, L., Johnson, Z. I. (2011). A suite of microplate reader-

based colorimetric methods to quantify ammonium, nitrate, orthophosphate and

116

silicate concentrations for aquatic nutrient monitoring. J Environ Monit: 13(2),

370-376

109. Rotem, O., Pasternak, Z., Jurkevitch, E. (2014). Bdellovibrio and like organisms.

In: Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H., Stackebrandt, E.

(eds) The Prokaryotes, 3rd edn., Springer, Berlin Heidelberg pp. 3-17

110. Ryan, R.P. and Dow, J.M. (2008) Diffusible signals and interspecies

communication in bacteria. Microbiology: 154, 1845-1858

111. Santiago-Rodriguez, T. M., Patrício, A. R., Rivera, J. I., Coradin, M., Gonzalez,

A., Tirado, G., Cano, R. J., Toranzos, G. A. (2014). luxS in bacteria isolated from

25-to 40-million-year-old amber. FEMS Microbiol Lett: 350(1), 117-124

112. Sato, M., Omori, K., Datta, T., Amano, Y., Machida, M. (2017). Influence of

extracellular polysaccharides and calcium ion on colony formation of unicellular

Microcystis aeruginosa. Environ Eng Sci: 34(3), 149-157

113. Schagerl, M., and Donabaum, K. (2003). Patterns of major photosynthetic

pigments in freshwater algae. 1. Cyanoprokaryota, Rhodophyta and Cryptophyta.

Ann Limnol-Int J Limnol: 39(1), 35-47

114. Schagerl, M., Pichler, C., Donabaum, K. (2003). Patterns of major

photosynthetic pigments in freshwater algae. 2. Dinophyta, Euglenophyta,

Chlorophyceae and Charales. Ann Limnol-Int J Limnol: 39(1), 49-62

115. Sedmak, B., and Eleršek, T. (2006). Microcystins induce morphological and

physiological changes in selected representative phytoplanktons. Microb Ecol:

51(4), 508-515

117

116. Shapiro, S. S., and Wilk, M. B. (1965). An analysis of variance test for

normality (complete samples). Biometrika: 52(3/4), 591-611

117. Shen, H., Niu, Y., Xie, P., Tao, M. I. N., Yang, X. I. (2011). Morphological and

physiological changes in Microcystis aeruginosa as a result of interactions with

heterotrophic bacteria. Freshwater Biol: 56(6), 1065-1080

118. Shi, L., Cai, Y., Kong, F., Yu, Y. (2012). Specific association between bacteria

and buoyant Microcystis colonies compared with other bulk bacterial

communities in the eutrophic Lake Taihu, China. Environ Microbiol Rep: 4(6),

669-678

119. Sun, D. L., Jiang, X., Wu, Q. L., Zhou, N. Y. (2013). Intragenomic

heterogeneity of 16S rRNA genes causes overestimation of prokaryotic diversity.

Appl Environ Microbiol: 79(19), 5962-5969

120. Surette, M. G., Miller, M. B., Bassler, B. L. (1999). Quorum sensing in

Escherischia coli, Salmonella typhimurium, and Vibrio harveyi: a new family of

genes responsible for autoinducer production. Proc Natl Acad Sci USA: 96(4),

1639-1644

121. Takaichi, S. (2011). Carotenoids in algae: distributions, biosyntheses and

functions. Marine Drugs: 9(6), 1101-1118

122. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S. (2011).

MEGA5: molecular evolutionary genetics analysis using maximum likelihood,

evolutionary distance, and maximum parsimony methods. Mol Biol Evol: 28(10),

2731-2739

118

123. Tan, L. R., Xia, P. F., Zeng, R. J., Li, Q., Sun, X. F., Wang, S. G. (2018). Low-

level concentrations of aminoglycoside antibiotics induce the aggregation of

cyanobacteria. Environ Sci Poll Res: 1-9

124. Taylor, M., and Governor, L. (2012). Biological and Water Quality Study of the

Licking River. Ohio EPA Division of Surface Water.

http://www.epa.state.oh.us/portals/35/documents/2008lickingtsd.pdf. Accessed 11

December 2017

125. Thompson, J. D., Higgins, D. G., Gibson, T. J. (1994). CLUSTAL W:

improving the sensitivity of progressive multiple sequence alignment through

sequence weighting, position-specific gap penalties and weight matrix choice.

Nucleic Acid Res: 22(22), 4673-4680

126. Vanderploeg, H. A., Liebig, J. R., Carmichael, W. W., Agy, M. A., Johengen, T.

H., Fahnenstiel, G. L., Nalepa, T. F. (2001). Zebra mussel (Dreissena

polymorpha) selective filtration promoted toxic Microcystis blooms in Saginaw

Bay (Lake Huron) and Lake Erie. Can J Fish Aquat Sci: 58(6), 1208-1221

127. Van Gremberghe, I., Vanormelingen, P., Vanelslander, B., Van der Gucht, K.,

D'hondt, S., De Meester, L., Vyverman, W. (2009). Genotype‐dependent

interactions among sympatric Microcystis strains mediated by Daphnia grazing.

Oikos: 118(11), 1647-1658

128. Velicer, G. J., Mendes-Soares, H., Wielgoss, S. (2014). Whence comes social

diversity? Ecological and evolutionary analysis of the myxobacteria.

Myxobacteria: Genomics, Cell Mol Biol: 1-29

119

129. Velichko, N., Chernyaeva, E., Averina, S., Gavrilova, O., Lapidus, A., Pinevich,

A. (2015). Consortium of the ‘bichlorophyllous’ cyanobacterium Prochlorothrix

hollandica and chemoheterotrophic partner bacteria: culture and metagenome‐

based description. Environ Microbiol Rep: 7(4), 623-633

130. Wang, W., Shen, H., Shi, P., Chen, J., Ni, L., Xie, P. (2015). Experimental

evidence for the role of heterotrophic bacteria in the formation of Microcystis

colonies. J Appl Phycol: 1-13

131. Wang, W., and Sun, M. (2009). Phylogenetic relationships between Bacillus

species and related genera inferred from 16s rDNA sequences. Braz J Microbiol:

40(3), 505-521

132. Wang, X., Qin, B., Gao, G., Paerl, H. W. (2010). Nutrient enrichment and

selective predation by zooplankton promote Microcystis (Cyanobacteria) bloom

formation. J Plankton Res: 32(4), 457-470

133. Wang, Y.W., Zhao, J., Li, J.H., Li, S.S., Zhang, L.H., Wu, M. (2011) Effects of

Calcium Levels on Colonial Aggregation and Buoyancy of Microcystis

aeruginosa. Curr Microbiol: 62, 679-683

134. Waters, C. M., and Bassler, B. L. (2005). Quorum sensing: cell-to-cell

communication in bacteria. Annu Rev Cell Dev Biol: 21, 319-346

135. White, J. D., Kaul, R. B., Knoll, L. B., Wilson, A. E., Sarnelle, O. (2011). Large

variation in vulnerability to grazing within a population of the colonial

phytoplankter, Microcystis aeruginosa. Limnol Oceanogr: 56(5), 1714-1724

120

136. White, J. D., and Sarnelle, O. (2014). Size‐structured vulnerability of the

colonial cyanobacterium, Microcystis aeruginosa, to grazing by zebra mussels

(Dreissena polymorpha). Freshwater Biol: 59(3), 514-525

137. Winson, M. K., Camara, M., Latifi, A., Foglino, M., Chhabra, S. R., Daykin, M.,

Bally, M., Chapon, V., Salmonds, G.P.C., Bycroft, B. W., Lazdunski, A., Stewart,

G.S.A.B., Williams, P. (1995). Multiple N-acyl-L-homoserine lactone signal

molecules regulate production of virulence determinants and secondary

metabolites in Pseudomonas aeruginosa. Proc Natl Acad Sci USA: 92(20), 9427-

9431

138. Wu, X., and Kong, F. (2009). Effects of light and wind speed on the vertical

distribution of Microcystis aeruginosa colonies of different sizes during a summer

bloom. Int Rev Hydrobiol: 94(3), 258-266

139. Xavier, K. B. and Bassler, B. L. (2003). LuxS quorum sensing: more than just a

numbers game. Curr Opin Microbiol: 6(2), 191-197

140. Xu, H., Yu, G., Jiang, H. (2013). Investigation on extracellular polymeric

substances from mucilaginous cyanobacterial blooms in eutrophic freshwater

lakes. Chemosphere: 93(1), 75-81

141. Xue, T., Ni, J., Shang, F., Chen, X., Zhang, M. (2015). Autoinducer-2 increases

biofilm formation via an ica-and bhp-dependent manner in Staphylococcus

epidermidis RP62A. Microb Infect: 17(5), 345-352

142. Yamamoto, Y., Shiah, F. K., Chen, Y. L. (2011). Importance of large colony

formation in bloom-forming cyanobacteria to dominate in eutrophic ponds. Ann

Limnol-Int J Limnol: 7(2), 167-173

121

143. Yang, Z., and Kong, F. (2012). Formation of large colonies: a defense

mechanism of Microcystis aeruginosa under continuous grazing pressure by

flagellate Ochromonas sp. J Limnol: 71(1), 5

144. Yang, Z., and Kong, F. (2013). Abiotic factors in colony formation: effects of

nutrition and light on extracellular polysaccharide production and cell aggregates

of Microcystis aeruginosa. Chinese J Oceanol Limnol: 31(4), 796-802

145. Yang, Z., Kong, F. X., Cao, H. S., Shi, X. L. (2005). Observation on colony

formation of Microcystis aeruginosa induced by filtered lake water under

laboratory conditions. Ann Limnol-Int J Limnol: (3), 169-173

146. Yang, Z., Kong, F., Shi, X., Zhang, M., Xing, P., Cao, H. (2008). Changes in the

morphology and polysaccharide content of Microcystis aeruginosa

(cyanobacteria) during flagellate grazing. J Phycol: 44(3), 716-720

147. Yang, D. and Park, S. (2017). Freshwater anostracan, Branchinella

kugenumaensis, as a potential controlling consumer species on toxic

cyanobacteria Microcystis aeruginosa. Aquat Ecol: 51(3), 449-461

148. Zhai, C., Zhang, P., Shen, F., Zhou, C., Liu, C. (2012). Does Microcystis

aeruginosa have quorum sensing? FEMS Microbiol Lett: 336(1), 38-44

149. Zhang, M., Kong, F., Tan, X., Yang, Z., Cao, H., Xing, P. (2007). Biochemical,

morphological, and genetic variations in Microcystis aeruginosa due to colony

disaggregation. World J Microbiol Biotechnol: 23(5), 663-670

150. Zhang, P., Chen, M., Zhang, Y., Li, Y., Lu, S., Li, P. (2018). Autoaggregation

and adhesion abilities in bacteria associated with colonies of Microcystis.

Hydrobiologia: 1-1

122

151. Zhu, W., Li, M., Luo, Y., Dai, X., Guo, L., Xiao, M., Huang, J., Tan, X. (2014).

Vertical distribution of Microcystis colony size in Lake Taihu: its role in algal

blooms. J Great Lakes Res: 40(4), 949-955

152. Zilliges, Y., Kehr, J. C., Mikkat, S., Bouchier, C., de Marsac, N. T., Börner, T.,

Dittmann, E. (2008). An extracellular glycoprotein is implicated in cell-cell

contacts in the toxic cyanobacterium Microcystis aeruginosa PCC 7806. J

Bacteriol: 190(8), 2871-2879

123

Appendix

Supplementary Table 1. Abundance of Bacterial Families. The 128 bacterial families and family-level groups that differed significantly in abundance (FDR-corrected p-value <0.05) among the four NMDS clusters shown in Fig. 3 Phylum Class Order Family 2 P Cluster Cluster Cluster Cluster 1 2 3 4 Acidobacteria Blasto- Blasto- Blasto- 22.5 0.0007 385.5 14 9 18.5 catellia catellales catellaceae (Subgroup 4) Acidobacteria Subgroup 6 Unassigned Unassigned 26.9 0.0004 7.1 75.5 1.5 0 Acidobacteria Subgroup 6 Unassigned Unassigned 14.1 0.0121 0.9 4.6 0.3 3 Actinobacteria Acidi- Acidi- Acidi- 18.2 0.0028 391.8 4380 880.6 180.5 microbiia microbiales microbiaceae Actinobacteria Acidi- Acidi- Acidi- 18.2 0.0028 8.2 31.5 0.9 2 microbiia microbiales microbiales Incertae Sedis Actinobacteria Acidi- Acidi- Unassigned 21.6 0.0008 7.4 57.9 90.1 3 microbiia microbiales Actinobacteria Actino- Coryne- Myco- 24.2 0.0007 4.8 776.5 23.1 4.5 bacteria bacteriales bacteriaceae Actinobacteria Actino- Frankiales Sporich- 23.6 0.0007 1105 11530 12215 498.5 bacteria thyaceae Actinobacteria Actino- Frankiales Unassigned 12.2 0.0258 0.8 6.2 0 0 bacteria Actinobacteria Actino- Micro- Demequin- 14.1 0.0121 0.00 0.9 3.8 0 bacteria coccales aceae Actinobacteria Actino- Micro- Micro- 26.7 0.0004 17.4 251.6 2029 30.5 bacteria coccales bacteriaceae Actinobacteria Actino- Micro- Unassigned 16.3 0.0053 0 0.2 1.6 0 bacteria coccales

124

Actinobacteria Actino- PeM15 Unassigned 25.7 0.0006 10.8 138.1 19.9 2 bacteria Actinobacteria Actino- PeM15 Unassigned 23.9 0.0007 5.8 386.8 53.6 9 bacteria Actinobacteria Actino- PeM15 Unassigned 20.9 0.0011 0.8 59 3.3 0 bacteria Actinobacteria Thermoleo- Gaiellales Unassigned 27.2 0.0004 2.3 159.6 35.5 3.5 philia Actinobacteria Thermoleo- Solirubro- TM146 24.8 0.0007 4.7 425.9 24.8 9 philia bacterales Actinobacteria Unassigned Unassigned Unassigned 13 0.0183 0.8 8.8 1.9 0 Armati- Armati- Armati- Armati- 11.3 0.0363 0.7 0.2 10 4 monadetes monadia monadales monadaceae Armati- Armati- Armati- Unassigned 22.3 0.0007 0 0 0.1 2 monadetes monadia monadales Bacteroidetes Bacteroidia Bacter- Porphyro- 14.7 0.0094 10.3 1.8 1.6 1 oidales monadaceae Bacteroidetes Bacteroidia Bacteroidia Draconi- 17.1 0.0040 11.6 0.2 6.4 2.5 Incertae bacteriaceae Sedis Bacteroidetes Bacter- Unassigned Unassigned 11.4 0.0348 5.3 0.3 2.5 0.0 oidetes vadinHA17 Bacteroidetes Cytophagia Cyto- Cyclo- 22.9 0.0007 3.3 21.2 869 227 phagales bacteriaceae Bacteroidetes Cytophagia Cyto- Cytophagaceae 23.1 0.0007 19.9 8.6 539.6 1513 phagales Bacteroidetes Cytophagia Cyto- Flammeo- 11 0.0417 1.1 0.00 5.8 0 phagales virgaceae

125

Bacteroidetes Flavo- Flavo- NS9 marine 17.2 0.0038 165.6 30.9 73.5 74.5 bacteriia bacteriales group Bacteroidetes SM1A07 Unassigned Unassigned 32 0.0002 0 0 0 29 Bacteroidetes SM1A07 Unassigned Unassigned 17.5 0.0035 0.2 0 7.4 5 Bacteroidetes Sphingo- Sphingo- Chitino- 21.3 0.0009 628.8 1358 2232 365 bacteriia bacteriales phagaceae Bacteroidetes Sphingo- Sphingo- Lentimicrobi- 17 0.0042 9.4 0.5 1.9 0 bacteriia bacteriales aceae Bacteroidetes Sphingo- Sphingo- Saprospiraceae 22.9 0.0007 1390 9.5 182.1 108.5 bacteriia bacteriales Bacteroidetes Sphingo- Sphingo- Sphingo- 18.2 0.0028 4.7 22.7 20.8 1.5 bacteriia bacteriales bacteriaceae Bacteroidetes Sphingo- Sphingo- env.OPS 17 16.5 0.0051 254.4 36.6 55.8 112.5 bacteriia bacteriales Bacteroidetes Sphingo- Sphingo- LiUU-11-161 24 0.0007 22.5 0.5 1.4 2.5 bacteriia bacteriales Bacteroidetes Sphingo- Sphingo- PHOS-HE51 22.3 0.0007 33.1 2.6 0.8 3.5 bacteriia bacteriales Bacteroidetes Sphingo- Sphingo- Unassigned 20.3 0.0014 47.5 17.6 2.6 1 bacteriia bacteriales Chlamydiae Chlamydiae Chlamydi- Parachlamydi- 17.9 0.0030 0 0.2 5 57 ales aceae Chlamydiae Chlamydiae Chlamydi- Simkaniaceae 14.3 0.0112 4.8 21.2 0.1 0 ales Chloroflexi Anaero- Anaero- Anaero- 13.6 0.0141 25.8 2.8 7 2 lineae lineales lineaceae Chloroflexi Caldilineae Caldilineales Caldilineaceae 22.8 0.0007 271.4 1.6 1.9 18 Chloroflexi SBR2076 Unassigned Unassigned 11.8 0.0291 371.3 0.2 0 23

126

Chloroflexi SL56 Unassigned Unassigned 23 0.0007 0.8 8.7 68.5 1.5 marine group Firmicutes Bacilli Bacillales Bacillaceae 12.2 0.0249 21.1 1.1 0.3 0 Firmicutes Bacilli Bacillales Paeni- 23.6 0.0007 72.7 6.6 3.6 0 bacillaceae Firmicutes Clostridia Halanaerob- ODP1230B8.2 17.7 0.0032 0 0 0.9 2 iales 3 Firmicutes Erysipelo- Erysipelo- Erysipelo- 15.1 0.0084 2.9 0.2 8.8 0 trichia trichales trichaceae Fusobacteria Fuso- Fuso- Fuso- 16.3 0.0053 4.4 0.2 0.4 0 bacteriia bacteriales bacteriaceae Hydrogen- Unassigned Unassigned Unassigned 17.5 0.0035 1.8 0 0 0 edentes Ignavibacteriae Ignavi- Ignavi- Ignavi- 16.9 0.0044 2.3 0 0.8 0 bacteria bacteriales bacteriaceae Omnitrophica Unassigned Unassigned Unassigned 11.9 0.0287 1.2 13.9 5.1 0 Parcubacteria Candidatus Unassigned Unassigned 18.7 0.0024 0 0.1 1.6 0 Azam- bacteria Planctomycetes OM190 Unassigned Unassigned 22.9 0.0007 64.7 6.9 1.3 6 Planctomycetes OM190 Unassigned Unassigned 15.5 0.0071 0 0 0 1 Planctomycetes Phycis- Phycis- Phycis- 21.3 0.0009 1922 4446 99.9 363 phaerae phaerales phaeraceae Planctomycetes Plancto- Plancto- Plancto- 14.1 0.0121 354.8 454.5 85.3 30.5 mycetacia mycetales mycetaceae Proteobacteria Alpha- Caulo- Caulo- 21.8 0.0008 11.7 21.9 179.6 216 proteo- bacterales bacteraceae bacteria

127

Proteobacteria Alpha- Caulo- Hypho- 15.6 0.0069 8.9 7 75.8 86 proteo- bacterales monadaceae bacteria Proteobacteria Alpha- Rhizobiales Bradyrhizobi- 14 0.0124 0.2 2.7 0.4 11.5 proteo- aceae bacteria Proteobacteria Alpha- Rhizobiales Rhizobiales 16.4 0.0051 8.1 49.6 31 0.5 proteo- Incertae Sedis bacteria Proteobacteria Alpha- Rhizobiales Methylo- 15 0.0087 1.9 0.7 39 19.5 proteo- bacteriaceae bacteria Proteobacteria Alpha- Rhizobiales Methylo- 25.2 0.0007 21.8 218 7.5 0 proteo- cystaceae bacteria Proteobacteria Alpha- Rhizobiales alphaI cluster 22.4 0.0007 1.8 6.3 57.5 6 proteo- bacteria Proteobacteria Alpha- Rhizobiales A0839 22.8 0.0007 27.7 12.9 0.9 9.5 proteo- bacteria Proteobacteria Alpha- Rhizobiales DUNssu044 15.5 0.0071 0 0 0 1 proteo- bacteria Proteobacteria Alpha- Rhizobiales MNG7 23 0.0007 32.3 71.1 7.1 6.5 proteo- bacteria Proteobacteria Alpha- Rhodo- Rhodo- 22.4 0.0007 59.6 174.6 991.5 98 proteo- bacterales bacteraceae bacteria

128

Proteobacteria Alpha- Rhodo- AT-s3-44 14.1 0.0121 0 2.8 0 0 proteo- spirillales bacteria Proteobacteria Alpha- Rhodo- I-10 16.4 0.0053 139.7 262.3 38.5 44.5 proteo- spirillales bacteria Proteobacteria Alpha- Rickettsiales Rickettsiaceae 16.1 0.0057 122.2 144.6 20.9 24.5 proteo- bacteria Proteobacteria Alpha- Rickettsiales [Caedibacter] 16.3 0.0053 1 5.6 0.3 0 proteo- caryophilus bacteria group Proteobacteria Alpha- Rickettsiales Rickettsiales 19.8 0.0016 81.7 361.6 27.9 40 proteo- Incertae Sedis bacteria Proteobacteria Alpha- Rickettsiales LWSR-14 24.1 0.0007 0 0.2 53.1 2.5 proteo- bacteria Proteobacteria Alpha- Rickettsiales Unassigned 21.5 0.0009 0 0 1.5 3.5 proteo- bacteria Proteobacteria Alpha- Rickettsiales Unassigned 19.4 0.0018 0 0 1.1 2.5 proteo- bacteria Proteobacteria Alpha- Sphingo- Erythro- 12.4 0.0236 0.8 12.8 12.8 0 proteo- monadales bacteraceae bacteria Proteobacteria Alpha- Sphingo- Sphingo- 13.8 0.0136 52.3 193.1 193 23 proteo- monadales monadaceae bacteria

129

Proteobacteria Beta-proteo- Burkhol- Alcaligenaceae 22.2 0.0007 237.6 1444 1283 177 bacteria deriales Proteobacteria Beta-proteo- Burkhol- Burkhol- 23.4 0.0007 360.7 1204 3675 160 bacteria deriales deriaceae Proteobacteria Beta-proteo- Burkhol- Comamonad- 18.2 0.0028 597.7 2033 2533 471.5 bacteria deriales aceae Proteobacteria Beta-proteo- Burkhol- Oxalo- 19.8 0.0016 204.3 625.9 663.4 90 bacteria deriales bacteraceae Proteobacteria Beta-proteo- Burkhol- Unassigned 15.2 0.0081 37.2 599.1 626.3 0.5 bacteria deriales Proteobacteria Beta-proteo- Methylo- Methylo- 24.2 0.0007 97.6 1166 806.8 50 bacteria philales philaceae Proteobacteria Beta-proteo- Nitroso- Nitroso- 20.1 0.0015 34.3 39.8 196. 279.5 bacteria monadales monadaceae Proteobacteria Beta-proteo- Unassigned Unassigned 17.9 0.0030 1.5 13.2 16 4 bacteria Proteobacteria Gamma- Cell- Spongii- 18.3 0.0028 19.8 17.6 0.9 4.5 proteo- vibrionales bacteraceae bacteria Proteobacteria Gamma- Chromatiales Chromatiaceae 20.3 0.0014 1.7 1.3 29.3 120 proteo- bacteria Proteobacteria Gamma- HTA4 Unassigned 11.3 0.0360 0 1.3 0 0 proteo- bacteria Proteobacteria Gamma- KI89A clade Unassigned 14.7 0.0094 7.1 2.1 0 0 proteo- bacteria

130

Proteobacteria Gamma- Legionel- Legionellaceae 22.5 0.0007 13.6 390.6 14.1 9 proteo- lales bacteria Proteobacteria Gamma- Methylo- CABC2E06 11.2 0.0370 3 0 4.3 1 proteo- coccales bacteria Proteobacteria Gamma- Oceano- Oceano- 13.7 0.0141 34.2 72.7 5 2 proteo- spirillales spirillaceae bacteria Proteobacteria Gamma- Xantho- Xantho- 13.3 0.0160 11.8 18.9 273.4 97 proteo- monadales monadaceae bacteria Proteobacteria Gamma- Xantho- Xantho- 22.1 0.0007 8.9 31.2 3.5 2 proteo- monadales monadales bacteria Incertae Sedis Proteobacteria Gamma- Xantho- Unassigned 27 0.0004 135.3 1021 26.3 7.5 proteo- monadales bacteria Proteobacteria Gamma- Xantho- Unassigned 22.7 0.0007 3.7 18.5 0.9 0 proteo- monadales bacteria Proteobacteria Gamma- Unassigned Unassigned 17.7 0.0032 11.3 10.1 0.3 0 proteo- bacteria Proteobacteria Delta- Bdello- Bacterio- 16.7 0.0047 124.1 373.1 117.4 40 proteo- vibrionales voracaceae bacteria Proteobacteria Delta- Bdello- Bdello- 19.4 0.0018 93.4 64 9 20 proteo- vibrionales vibrionaceae bacteria

131

Proteobacteria Delta- Brady- Brady- 13.8 0.0134 13.9 2 3 5 proteo- monadales monadales bacteria Proteobacteria Delta- Desulfo- Desulfo- 11.3 0.0360 3.4 0.4 2.4 1.5 proteo- bacterales bulbaceae bacteria Proteobacteria Delta- Desulfuro- C8S-102 15.5 0.0071 0 0 0 1 proteo- monadales bacteria Proteobacteria Delta- Myxo- Phaselicys- 29.3 0.0004 31.3 0 0 2 proteo- coccales tidaceae bacteria Proteobacteria Delta- Myxo- Blfdi19 14.5 0.0104 13.7 5.8 1.9 0 proteo- coccales bacteria Proteobacteria Delta- Myxo- KD3-10 12.3 0.0244 1 0 0 0 proteo- coccales bacteria Proteobacteria Delta- Myxo- mle1-27 28.2 0.0004 36.1 0 0.1 5.5 proteo- coccales bacteria Proteobacteria Delta- Myxo- P3OB-42 23.4 0.0007 19.9 0.9 0.4 1.5 proteo- coccales bacteria Proteobacteria Delta- Myxo- Polyangiaceae 19.8 0.0016 9.4 0.9 0.1 0 proteo- coccales bacteria Proteobacteria Delta- Myxo- Sandara- 14.8 0.0091 6.5 0.2 0.6 0 proteo- coccales cinaceae bacteria

132

Proteobacteria Delta- Myxo- Unassigned 22.3 0.0007 25.9 7.3 0.4 0 proteo- coccales bacteria Proteobacteria Delta- Oligo- Oligoflexaceae 15.2 0.0081 143.4 148.2 11.1 52.5 proteo- flexales bacteria Proteobacteria Delta- Oligo- 0319-6G20 19.8 0.0016 101.5 352.8 22 13.5 proteo- flexales bacteria Proteobacteria Delta- SAR324 SAR324 18.2 0.0028 39.1 43.2 0.6 3 proteo- Marine Marine group bacteria group B Proteobacteria Delta- Sva0485 Unassigned 12.8 0.0204 1.8 0 0.1 0 proteo- bacteria Proteobacteria Delta- Sva0486 Unassigned 12.5 0.0224 1.2 0 0.4 0 proteo- bacteria Proteobacteria SPOYSOCT Unassigned Unassigned 24.3 0.0007 21.1 8.1 0.9 1.5 -00m83 Sacchari- Unassigned Unassigned Unassigned 12.1 0.0261 9.2 16.4 0.8 0 bacteria Spirochaetae Spirochaetes Spiro- Leptospiraceae 19.4 0.0018 9.9 64.6 35.3 2 chaetales Verruco- OPB35 soil OPB35 soil Unassigned 21.2 0.0009 243.1 11.6 3.3 11.5 microbia group group Verruco- OPB35 soil OPB35 soil Unassigned 20.7 0.0012 150.7 31 18.5 18.5 microbia group group Verruco- OPB35 soil OPB35 soil Unassigned 14.9 0.0091 1.3 0 0 0 microbia group group

133

Verruco- OPB35 soil OPB35 soil Unassigned 12.5 0.0228 1.9 0.3 0.9 0 microbia group group Verruco- Opitutae Opitutales Opitutaceae 19.3 0.0018 54 71.8 4.9 4 microbia Verruco- Opitutae Opitutae Unassigned 13.3 0.0160 6 3.9 247.9 88 microbia vadinHA64 Verruco- Sparto- Chthonio- Chthonio- 19.2 0.0019 117.5 73.8 2.3 8.5 microbia bacteria bacterales bacteraceae Verruco- Sparto- Chthonio- DA101 soil 16.3 0.0053 244.1 108.8 17.3 10 microbia bacteria bacterales group Verruco- Verruco- Unassigned Unassigned 16.7 0.0047 1682 436.2 43.6 73.5 microbia microbia Verruco- Incertae microbia Sedis Unassigned Unassigned Unassigned Unassigned 24.8 0.0007 66.7 738.9 12 51 bacteria

134