Interactions with Heterotrophic Bacteria in the Microcystis Phycosphere

James Madison University JMU Scholarly Commons

Masters Theses, 2010-2019 The Graduate School

5-2-2019

The Microcystis microbiome: Interactions with heterotrophic bacteria in the Microcystis phycosphere

Alexa Hoke

Follow this and additional works at: https://commons.lib.jmu.edu/master201019

Recommended Citation Hoke, Alexa, "The Microcystis microbiome: Interactions with heterotrophic bacteria in the Microcystis phycosphere" (2019). Masters Theses, 2010-2019. 637. https://commons.lib.jmu.edu/master201019/637

This Thesis is brought to you for free and open access by the The Graduate School at JMU Scholarly Commons. It has been accepted for inclusion in Masters Theses, 2010-2019 by an authorized administrator of JMU Scholarly Commons. For more information, please contact [email protected].

The Microcystis microbiome: Interactions with heterotrophic bacteria in the Microcystis phycosphere

Alexa Hoke

A thesis submitted to the Graduate Faculty of

JAMES MADISON UNIVERSITY

Partial Fulfillment of the Requirements

for the degree of

Master of Science

Department of Biology

August 2019

FACULTY COMMITTEE:

Committee Chair: Dr. Morgan Steffen

Committee Members/ Readers:

Dr. Louie Wurch

Dr. Raymond Enke

Table of Contents List of Tables ...... iv List of Figures ...... v Abstract ...... vi Chapter I. Literature Review ...... 1 1. Introduction ...... 2 2. Cyanobacterial Harmful Algal Blooms ...... 2 2.1 Ecological Impact ...... 3 2.2 Economic and Public Health Impact...... 4 2.3 Geographic Distribution of Cyanobacterial Harmful Algal Blooms ...... 5 3. Microcystis ...... 6 3.1 Genomics of Microcystis ...... 7 4. Drivers of Cyanobacterial Harmful Algal Blooms ...... 8 4.1 Nutrients ...... 8 4.2 Climate Change ...... 9 5. Microbial Interactions in Bloom Communities ...... 11 5.1 Nutrient Exchange in the Phycosphere ...... 12 6. Works Cited ...... 14 Chapter II. The heterotrophic bacterial community of a western Lake Erie Microcystis bloom and impacts of co-culture ...... 19 Abstract ...... 20 1. Introduction ...... 21 2. Methods ...... 22 2.1 Sample Collection ...... 22 2.2 Culturable Microbiome ...... 23 2.3 Bacterial Growth Curves ...... 25 2.4 Co-Culture Conditions ...... 26 2.5 Metagenomics ...... 27 3. Results and Discussion ...... 27 3.1 Culturable Microbiome ...... 27 3.2 Response to Co-culture ...... 37 3.3 Response to Filtrate ...... 44 4. Metagenomics ...... 52 5. Conclusions ...... 55

6. Works Cited ...... 56 Chapter III. Comparative genomics of heterotrophic bacteria isolated from a Lake Erie Microcystis bloom ...... 58 1. Abstract ...... 59 2. Introduction ...... 60 3. Methods ...... 61 3.1 DNA Extraction and Sequencing ...... 61 3.2 Genome Assembly and Annotation ...... 62 4. Results and Discussion ...... 62 4.1 Sequencing Output, Assembly, and Annotation ...... 62 4.2 Nitrogen ...... 63 4.3 Carbon Utilization/Carbohydrates ...... 69 4.4 Iron ...... 70 4.5 Heavy Metal and Antibiotic Resistance ...... 72 4.6 Secondary Metabolites ...... 72 4.7 Quorum Sensing and Signaling ...... 73 4.8 Other Genes ...... 73 5. Conclusions ...... 74 6. Works Cited ...... 75 7. Chapter IV. Conclusions ...... 79

iii

List of Tables

Chapter II

Table 1. Characterization of heterotrophic bacteria chosen to be grown with axenic

Microcystis

Table 2. Description of co-culture treatments

Table 3. Identification of the 55 isolates that were cultured based on 16S sequencing

Table 4. Comparison of the Fv/Fm of Microcystis in co-culture at three different growth phases

Table 5. Comparison of the Fv/Fm of Microcystis in filtrate at three different growth phases

Chapter III

Table 1. DNA concentration and quality from sequenced heterotrophic bacteria and the number of raw and corrected reads obtained from sequencing

Table 2. Characteristics of genome assemblies

Table 3. Number of coding sequences obtained from RAST

Table 4. Nitrogen genes called by RAST

List of Figures

Chapter II

Figure 1. Phylogenetic tree of organisms in the Exiguobacterium genus

Figure 2. Phylogenetic tree of organisms in the Enterobacter genus

Figure 3. Phylogenetic tree of organisms in the Deinococcus genus

Figure 4. Phylogenetic tree of organisms in the Paenibacillus genus

Figure 5. Phylogenetic tree of organisms in the Acidovorax genus

Figure 6. Fluorescence of Microcystis grown with different concentrations of bacterial cells

Figure 7. Direct cell counts of Microcystis grown with different concentrations of bacterial cells

Figure 8. Fv/Fm of Microcystis grown with different concentrations of bacterial cells

Figure 9. Fluorescence of Microcystis grown in different bacterial filtrate treatments

Figure 10. Direct cell counts of Microcystis grown in different bacterial filtrate treatments

Figure 11. Fv/Fm of Microcystis grown in different bacterial filtrate treatments

Figure 12. Percentage of OTUs belonging to different prokaryotic phyla at each sampling site

Figure 13. Percentage of OTUs belonging to different classes within the Proteobacterium phylum at each sampling site

Abstract

Harmful algal blooms are a growing problem globally in both freshwater and marine systems. Cyanobacterial harmful algal blooms (cHABs) can have numerous environmental and economic impacts. cHABs can reduce oxygen levels in the water column, ultimately leading to the death of larger organisms, and reduce light penetration, impacting aquatic vegetation. cHABs can also release toxins into the water and efforts to prevent toxins from entering drinking water can cost cities millions of dollars in added filtration. Microcystis is a freshwater cyanobacterium that is globally distributed. While many of the abiotic factors that can impact bloom formation are known, biotic factors, such as microbial interactions within Microcystis blooms, are not as widely studied. To assess the impact of certain heterotrophic bacteria on the growth of Microcystis, heterotrophic bacteria were isolated from a Microcystis bloom on Lake Erie. The culturable microbiome of the bloom was compared to the unculturable microbiome. Five bacterial isolates were chosen to be grown with Microcystis to determine the impact of co-culture. The whole genomes of these five isolates were also sequenced with PacBio to determine if they contained any genes that might suggest interactions with Microcystis.

Co-culture with Acidovorax cells had a significant impact on the growth rate of

Microcystis cells. The filtrate of Enterobacter, Exiguobacterium, and Paenibacillus appeared to have a protective impact on Microcystis cells in high salt concentrations. The whole genomes of the sequenced heterotrophic bacteria contained genes that could be used in potential interaction with Microcystis. These were genes related functions such as nitrogen metabolism, iron scavenging, carbohydrate metabolism, and vitamin production.

The data obtained from this research can be used to further explore interactions between

Microcystis and heterotrophic bacteria and the potential mechanisms for protection provided by bacteria in environments that are not ideal for Microcystis growth.

vii

Literature Review

1. Introduction

Cyanobacterial harmful algal blooms (cHABs) occur annually in both marine and freshwater systems worldwide. These bloom events have become more common in recent years and are expected to increase in frequency and biomass as global temperatures rise

(Lurling et al., 2017). One of the most important contributors to the formation of cHABs is anthropogenic nutrient loading, specifically of nitrogen and phosphorus (Paerl, 2014).

Microcystis is one of the most common genera of bloom-forming cyanobacteria worldwide. Microcystis blooms can negatively impact ecosystems by releasing toxins and reducing light penetration in the water column (Anderson et al., 2002). This can also negatively impact local economies in areas near freshwater systems where blooms occur as noxious bloom biomass can reduce waterfront property values in addition to causing a loss of revenue for local fisheries and other businesses (Dodds et al., 2009; US EPA,

2015).

2. Cyanobacterial Harmful Algal Blooms

Cyanobacteria are a phylum of photosynthetic bacteria that are also known as blue-green algae. Members of the Cyanobacteria phylum are extraordinarily diverse in both morphology and function. In freshwater systems, there are numerous freshwater genera of cyanobacteria that are capable of forming blooms, including Dolichospermum

(Anabaena), Aphanizomenon, Lyngbya, Microcystis, Nodularia, Planktothrix, and

Oscillatoria (Paerl et al., 2001). Due to both biomass and toxin production, blooms formed by cyanobacteria can have numerous ecological, economic, and health impacts

(Anderson et al., 2002; De Figueiredo et al., 2004; Backer et al., 2015).

2.1 Ecological Impact

cHABs impact ecosystem health and function in numerous ways. As blooms decay, oxygen concentrations can be reduced (Anderson et al., 2002) and this can cause fish and plant deaths (Landsberg, 2002; Kaebernick and Neilan, 2001). Oxygen depletion is caused by bacterial respiration as the cyanobacterial biomass is broken down. cHAB biomass can also reduce light in the water column (Anderson et al., 2002), which can impact aquatic vegetation (Landsberg, 2002). Cyanobacteria can produce a diverse array of toxins, known as cyanotoxins (Paerl et al., 2016). Certain species of Microcystis can produce the cyanotoxin microcystin, which is a hepatotoxin (Paerl et al., 2016).

Microcystin inhibits protein phosphatase 1 and protein phosphatase 2A in liver cells

(Paerl et al., 2016), which can cause severe liver damage and ultimately lead to death of fish (Kaebernick and Neilan, 2001). Cyanotoxins have also been linked to the deaths of domestic and wild animals that generally come into contact with the toxins through drinking water. These wild animals include deer, birds, squirrels, skunks, and bats while domestic animals include cattle, poultry, houses, cats, dogs, and pigs (Huisman et al.,

2005). Toxin-producing cyanobacterial cells can also be taken up by zooplankton and filter-feeders such as shellfish (Landsberg, 2002). Cyanotoxins can also accumulate in fish species, although some studies have shown that toxin accumulation by fish may vary by species (Peng et al., 2010). Toxins can then be acquired through the food web as fish and other animals consume organisms that have ingested the toxins (Huisman et al.,

2005).

2.2 Economic and Public Health Impact

HABs can have a devastating impact fishing and tourism industries (Wolf et al.,

2017). Toxins released by cyanobacteria can be ingested by fish and oxygen depletion caused by blooms can lead to fish mortality. Mass deaths of commercially desirable fish impact fishing revenues (Wolf et al., 2017). Bloom events can lead to increased processing costs for fisheries and some fisheries have even had to stop production during bloom events (US EPA, 2015). The economic cost to fisheries during bloom events can be millions of dollars (Sanseverino et al., 2016). A 2005 diatom bloom in the Gulf of

Maine caused a loss of approximately $6 million in shellfish sales (Sanseverino et al.,

2016).

Economic losses are higher when loss of tourism is considered. The 2005 Gulf of

Maine bloom was estimated to cost local businesses $14.8 million (Sanseverino et al.,

2016). During bloom events, warnings to avoid contact with water can deter tourists

(Wolf et al., 2017). Tourists can also be repelled by the fish mortality, the water discoloration, and the algal decomposition smells that are caused by algal blooms

(Sanseverino et al., 2016). At Lake Erie, fishing permit sales have been shown to decrease during bloom events that exceed the World Health Organization’s moderate level (Wolf et al., 2017). A 2009 study estimated that the recreational loss caused by freshwater eutrophication could cost up to $1.16 billion annually in the United States

(Dodds et al., 2009). Lakefront property values have also been shown to decrease as water clarity decreases (Dodds et al., 2009; US EPA, 2015). The monitoring and management of algal blooms can also have major economic impacts. The City of Toledo,

Ohio, on Lake Erie has spent $3,000-$4,000 per day on water filtration during bloom events (US EPA, 2015).

Cyanotoxins released from blooms can impact human health in different ways.

Skin contact with cyanotoxins can cause rashes, while swallowing contaminated water can have gastrointestinal or neurological effects (Backer et al., 2015). Consumption of fish and shellfish that have accumulated cyanotoxins, including microcystins, can also be harmful (Peng et al., 2010). There have been multiple bloom events that have left people without access to potable water. In 2007, excessive nutrient loading and an unusually warm spring caused a massive Microcystis bloom in Lake Tai, the third largest freshwater lake in China (Qin et al., 2010). This toxin-forming bloom left approximately 2 million people without access to drinking water for a week (Qin et al., 2010). Measurable levels of microcystin in water originating from Lake Erie prevented about 500,000 residents of

Toledo, Ohio from being able to use tap water in August 2014 (Backer et al., 2015;

Steffen et al., 2017).

2.3 Geographic Distribution of Cyanobacterial Harmful Algal Blooms

Cyanobacteria capable of forming blooms are globally distributed (Harke et al.,

2016). To date, Microcystis species have formed blooms on every continent except

Antarctica (Zurawell et al., 2015). In the United States, Microcystis blooms have notoriously been a problem in Lake Erie for several decades (Brittain et al, 2000; Steffen et al., 2017). As Lake Erie provides millions of dollars in revenue through the tourism and fishing industries, these blooms have become increasingly problematic as their frequency and distribution continues to increase (Michalak et al., 2013).

The Chinese lake, Lake Tai, is also severely impacted by regularly occurring

Microcystis blooms. This lake provides water to approximately 40 million people and is also a fishing and tourism resource for the area (Paerl et al., 2011).

3. Microcystis

Microcystis spp. have spherical cells (1-9µm) that form colonies that can range in size from microscopic to macroscopic (<20 to >200 cells for Microcystis aeruginosa)

(Komárek and Komárková, 2002; Cao and Yang, 2010). Microcystis spp. can form a variety of morphologies and colony formation is influenced by both biotic and abiotic factors (Xiao et al., 2017). Antagonists, such as microbial predators or parasites, can trigger an increase in colony size by producing more extracellular polysaccharides

(Wichelen et al., 2016). Increases in colony sizes can defend the Microcystis cells against flagellates that may struggle to break apart the aggregation of cells (Wichelen et al.,

2016). Colonies of Microcystis can produce a mucilage, which is a carbohydrate-rich extracellular polymeric substance in which bacteria can become embedded (Pereira et al.,

2009). It is notoriously difficult to achieve an axenic culture of Microcystis, or a culture of Microcystis without any other bacterial species present, because other heterotrophic bacteria are usually tightly coupled with the mucilage (Shirai et al., 1989).

As described above, some species of Microcystis produce the toxin microcystin.

Microcystins are cyclic heptapeptides (Sivonen and Jones, 1999). The chemical structures of microcystisns can vary and there are at least 85 different varieties of the toxin

(Sivonen and Jones, 1999). Temperature, light intensity, and nutrient availability are important factors in the production of microcystins. Production of these toxins at certain temperatures is species dependent (Sivonen, 1990). In lab studies with Microcystis

7 aeruginosa, researchers have found a positive correlation between microcystins and increased irradiance (Wiedner et al., 2003). Microcystin is a nitrogen-rich compound and the concentration of microcystin in cells has been shown to decrease when nitrogen is less available (Harke and Gobler, 2013).

3.1 Genomics of Microcystis

Microcystis aeruginosa NIES-843 was the first Microcystis genome to be sequenced and to date the only representative of the genus to have its genome fully closed (Kaneko et al., 2007). This was followed by Microcystis aeruginosa PCC 7806

(Frangeul et al., 2008). In 2015, there were 15 draft or closed sequences available from

Microcystis isolated from countries around the world (Harke et al., 2016). There was a high content of DNA mobile elements found in the Microcystis aeruginosa NIES-843 genome, as well as the genomes of other Microcystis strains, suggesting high genome plasticity (Kaneko et al., 2007). This genome plasticity may help Microcystis adapt in eutrophic ecosystems (Frangeul et al., 2008).

The genes involved in the production of the toxin, microcystin, are encoded by the mcy gene cassette and can be found in both NIES 843 and PCC 7806 strains.

However, not all individual cells in a given population have the genetic capability to produce toxins. While Microcystis spp. lack nif genes, and are therefore unable to fix nitrogen, they are able to degrade urea, and contain the genes necessary for construction of urease (ureA-ureG) (Mobley et al., 1995; Kaneko et al., 2007; Solomon et al., 2010).

The transcription of nitrogen assimilation genes has been shown to increase under low nitrogen conditions (Harke and Gobler, 2013). The nitrogen regulatory gene, ntcA, is one of the genes with increased transcript abundance, as well as genes involved in

8 nitrate/nitrite transport (Harke and Gobler, 2013). Urease transport genes (urtA-urtE) also have increases in transcript abundance under low nitrogen conditions (Harke and Gobler,

2013) and in field studies (Steffen et al., 2017).

4. Drivers of Cyanobacterial Harmful Algal Blooms

4.1 Nutrients

Nitrogen and phosphorus are essential and often limiting nutrients for Microcystis field populations. The increase in Microcystis blooms around the world appears to be connected to increases in nitrogen and phosphorus due to anthropogenic activity (Paerl,

2014). Because Microcystis spp. cannot fix nitrogen, they must rely on external sources, or they must regenerate combined nitrogen sources internally (Harke et al., 2016).

Microcystis has also been found to prefer reduced nitrogen over oxidized nitrogen sources (Flores and Herrero, 2005).

Urea is a nitrogen source that can influence bloom communities in freshwater systems. In the 1960s, urea represented only 5% of nitrogen fertilizer and by the 1990s it represented over 40% of global nitrogen fertilizer (Glibert et al., 2006). Urea is water soluble (Belisle et al., 2016) and more cost effective than other nitrogen fertilizers

(Glibert et al., 2006). Urea also contains twice the amount of nitrogen as ammonium sulfate (Glibert et al., 2006), which was the common form of nitrogen fertilizer before the use of urea. Agricultural runoff carries urea and other nitrogen sources into freshwater systems. Increased levels of urea in freshwater environments can favor the growth of non-nitrogen-fixing bacteria, such as Microcystis, and inhibit the growth of diazotrophic

(nitrogen-fixing) species (Finlay et al., 2010).

Phosphorus is another important nutrient for cyanobacteria, and Microcystis is known to be efficient in phosphorus utilization (Baldia et al., 2007). Microcystis is extremely efficient at accumulating phosphorus in low phosphorus concentrations (Baldia et al., 2007). Microcystis can store phosphate so that is can survive when phosphorus is less available (Carr and Whitton, 1982). Like nitrogen, phosphorus can enter freshwater systems through agricultural runoff as well as wastewater (Hart et al., 2004; Conley et al.,

2009).

Research in Canadian Experimental Lakes led to the determination in the 1970s that phosphorus was the primary limiting nutrient for eutrophication in these freshwater lakes (Schindler, 1974). Original management strategies then focused on the limitation of phosphorus inputs. It was not until the 1990s that researchers began to form a consensus about the need to control nitrogen inputs (Howarth and Marino, 2006). The recognition of nitrogen as a limiting nutrient of eutrophication lead to the need for a dual nutrient management strategy (Conley et al., 2009; Paerl et al., 2011). In lakes such as Lake Tai and Lake Okeechobee, non-nitrogen-fixing cyanobacteria dominate the phytoplankton population and phosphorus is rapidly recycled (Conley et al., 2009). These lakes require strategies to control both nitrogen and phosphorus simultaneously to ultimately control the cyanobacterial populations (Conley at el.. 2009).

4.2 Climate Change

Fresh water resources are a concern with global rising temperatures as harmful algal bloom events are becoming more common and biomass is increasing. Higher temperatures and eutrophication –higher inputs of nutrients, especially nitrogen (N) and phosphorus (P) – are known to enhance the development of cyanobacterial blooms in

10 freshwater (De Figueiredo et al., 2004). Rising global temperatures are expected to cause earlier onset stratification and an increase in the duration of stratification in some lakes

(Carey et al., 2012). The variation of temperatures on a diel scale are also expected to become smaller. This will cause any microstratification that occurs during the day to be maintained overnight (Carey et al., 2012). This microstratification can reduce mixing in the water column to a shallow top layer which could ultimately favor the formation of cyanobacterial algal blooms (Carey at al., 2012). In many climate change scenarios, different cyanobacteria are expected to dominate phytoplankton communities (Carey et al., 2012).

The growth rates of different strains of Microcystis have been tested at temperatures ranging from 27°C to 36°C (Mowe et al., 2015). Out of the five strains tested in one study, Microcystis flos-aquae had a higher growth rate at 30°C and

Microcystis ichthyloblabe had a higher growth rate at 33°C (Mowe et al., 2015). In another study, the growth rate of Microcystis aeruginosa increased significantly with increased temperature and light intensity (Li et al., 2014). The microcystin concentration of in the cells of some Microcystis strains has also been shown to increase with increasing temperatures (27°C to 33°C) (Mowe et al., 2015). Temperatures over 36°C can decrease the microcystin content in cells. This suggests that rising temperatures in tropical regions could lead to a decrease in bloom toxins (Mowe et al., 2015). As bloom events intensify, it has been predicted that toxin-producing species could dominate bloom communities during different times of the year (Trolle et al., 2015).

5. Microbial Interactions in Bloom Communities

cHABs are comprised of a multitude of bacterial species that communicate and interact with each other in different ways. Interactions between species within a bloom are species-specific and can be beneficial or potentially antagonistic (Ramanan et al.,

2015). Cyanobacteria and heterotrophic bacteria within bloom communities rely on each other for biogeochemical cycling (Steffen et al., 2012; Ramanan et al., 2015).

Heterotrophic bacterial partners are tightly coupled with the extracellular polymeric substance that is secreted by Microcystis cells (Pereira et al., 2009). The interactions between Microcystis and its heterotrophic bacterial partners have been shown to affect the growth of both (Shen et al., 2011). Heterotrophic bacterial partners are ubiquitous in cyanobacterial harmful algal blooms and are important for the growth of cyanobacteria

(Ramanan et al., 2016). This makes it difficult to obtain axenic cultures of cyanobacteria.

The growth of the heterotrophic partners with Microcystis is limited during the exponential phase and increased during the stationary phase (Shen et al., 2011).

The term phycosphere was coined by Bell and Mitchell in 1972 to describe the zone around phytoplankton that is analogous to the rhizosphere in terrestrial environments. The size of the phycosphere is dependent on the size of phytoplankton cells (Seymour et al., 2017). The phycosphere has higher concentrations of organic matter due to the release of organic molecules, including dissolved organic carbon, by phytoplankton into the surrounding water (Seymour et al., 2017). The growth phase of phytoplankton cells can impact the molecular weight of molecules released from the cells. In the early phases of cell growth, phytoplankton release lower molecular weight molecules, such as amino acids and carbohydrates, while higher molecular weight

12 molecules, such as polysaccharides, nucleic acids, and proteins, can be released into the phycosphere during lysis (Seymour et al., 2017). The release of molecules by phytoplankton can attract heterotrophic bacteria to the phycosphere, which can lead to a potential exchange of nutrients between the bacteria and the phytoplankton (Amin et al.,

2015; Seymour et al., 2017).

5.1 Nutrient Exchange in the Phycosphere

Eukaryotic phytoplankton and cyanobacteria can excrete extracellular organics that can support the bacteria in the phycosphere (Paerl and Pinckney, 1996). Some of the molecules excreted by phytoplankton act as chemoattractants for heterotrophic bacteria

(Seymour et al., 2017). Heterotrophic bacteria within the phycosphere also rely on the dissolved organic carbon (DOC)that is released by the phytoplankton for growth

(Seymour et al., 2017). The bacteria within bloom communities can contribute to the growth of phytoplankton as well. In marine ecosystems, B12 has been shown to be an important vitamin for the growth of phytoplankton (Helliwell et al., 2016; Seymour et al.,

2017). Many eukaryotic phytoplankton are unable to synthesize vitamin B12 so these phytoplankton rely on the heterotrophic bacteria to provide the vitamin (Seymour et al.,

2017). Another important nutrient for marine phytoplankton is iron. Some marine heterotrophic bacteria are able to bind iron to molecules they produce known as siderophores (Amin et al., 2009). The binding of iron to siderophores ultimately increases the solubility of the iron and eukaryotic phytoplankton are able to access the iron through iron transporters in their outer cell membranes (Amin et al., 2009).

Evidence of quorum sensing has been found in phycosphere communities

(Rolland et al., 2016). Quorum sensing is a mechanism of cell-to-cell communication.

This can aid in the formation of biofilms and the acquisition of nutrients (Rolland et al.,

2016). Acylated homoserine lactones (AHLs) are quorum sensing related signals that have been demonstrated in axenic M. aeruginosa (Zhai et al., 2012).

Many studies involving Microcystis relate to the abiotic factors that impact bloom formation or the toxins produced by some species of Microcystis. This thesis focuses on the biotic factors that influence the growth and fluorescence of Microcystis, primarily microbial interactions. The impact of individual bacterial isolates on Microcystis was tested to try to further the understanding of the role of Lake Eire bacteria in cyanobacterial bloom communities and the genomes of these bacteria were sequenced to understand the mechanism of the potential interactions with Microcystis.

6. Works Cited Amin SA, Green DH, Hart MC, Kupper FC, Sunda WG, Carrano CJ. 2009. Photolysis of iron-siderophore chelates promotes bacterial-algal mutualism. PNAS 106(40):17071- 17076. Amin SA, Hmelo LR, van Tol HM, Durham BP, Carlson LT, Heal KR, Morales RL, Berthiaume CT, Parker MS, Djunaedi B, Ingalls AE, Parsek MR, Moran MA, Armbrust EV. 2015. Interaction and signaling between a cosmopolitan phytoplankton and associated bacteria. Nature 522:98-101. Anderson DM, Glibert PM, Burkholder JM. 2002. Harmful algal blooms and eutrophication: nutrient sources, composition, and consequences. Estuaries 25(4):704- 726. Backer LC, Baptiste DM, LePrell R, Bolton B. 2015. Cyanobacteria and algae blooms: review of health and environmental data from the Harmful Algal Bloom-Related Illness Surveillance System (HABISS) 2007-2011. Toxins 7:1048-1064. Baldia SF, Evangelista AD, Aralar EV, Santiago AE. 2007. Nitrogen and phosphorus utilization in the cyanobacterium Microcystis aeruginosa isolated from Laguna de Bay, Philippines. J. Appl. Phycol. 19(6):607-613. Belisle BS, Steffen MM, Pound HL, Watson SB, DeBruyn JM, Bourbonniere RA, Boyer GL, Wilhelm SW. 2016. Urea in Lake Erie: organic nutrient sources as potentially impacted drivers of phytoplankton biomass. J. Great Lakes Res. 42(3):599-607. Bell W and Mitchell R. 1972. Chemotactic and growth responses of marine bacteria to algal extracellular products. Biol. Bull. 143(2):265-277. Brittain SM, Wang J, Babcock-Jackson L, Carmichael WW, Rinehart KL, Culver DA. 2000. Isolation and characterization of microcystins, cyclic heptapeptide hepatotoxins from a Lake Erie strain of Microcystis aeruginosa.J. Great Lakes Res. 26(3):241-249. Cao H and Yang Z. 2010. Variation in colony size of Microcystis aeruginosa in a eutrophic lake during recruitment and bloom formation. J. Freshwater Ecol. 25(3):331- 335. Carey CC, Ibelings BW, Hoffmann EP, Hamilton DP, Brookes JD. 2012. Eco- physiological adaptations that favour freshwater cyanobacteria in a changing climate. Water Res. 46(5):1394-1407. Carr NG and Whitton BA. 1982. The biology of cyanobacteria. Blackwell Scientific, Oxford. Conley DJ, Paerl HW, Howarth RW, Boesch DF, Seitzinger SP, Havens KE, Lancelot C, Likens GE. 2009. Controlling eutrophication: Nitrogen and phosphorus. Science 323(5917):1014-1015.

De Figueiredo DR, Azeiteiro UM, Esteves SM, Goncalves JM, Pereira MJ. 2004. Microcystin-producing blooms – a serious global public health issue. Ecotoxicol. Environ.Saf. 59: 151-163. Dodds WK, Bouska WW, Eitzmann JL, Pilger TJ, Pitts KL, Riley AJ, Schloesser JT, Thornbrugh DJ. 2009. Eutrophication of U.S. freshwaters: analysis of potential economic damages. Environ. Sci. Technol. 43:12-19. Finlay K, Patoine A, Donald DB, Bogard MJ, Learitt PR. 2010. Experimental evidence that pollution with urea can degrade water quality in phosphorus-rich lakes of the Northern Great Plains. Limnol. Oceanogr. 55(3):1213-1230. Flores E and Herrero A. 2005. Nitrogen assimilation and nitrogen control in cyanobacteria. Biochem. Soc. Trans. 33(10):164-167. Frangeul L, Quillardet P, Castets AM, Humbert JF, Matthijs H, Cortez D, Tolonen A, Zhang CC, Gribaldo S, Kehr JC, Zilliges Y, Ziemert N, Becker S, Talla E, Latifi A, Billault A, Lepelletier A, Dittmann E, Bouchier C, Tandeau de Marsac N. 2008. Highly plastic genome of Microcystis aeruginosa PCC 7806, a ubiquitous toxic freshwater cyanobacterium. BMC Genomics 9:274. Glibert PM, Harrison J, Heil C, Seitzinger S. 2006. Escalating worldwide use of urea – a global change contributing to coastal eutrophication. Biogeochemistry 77(3):441-463. Harke MJ and Gobler CJ. 2013. Global transcriptional responses of the toxic cyanobacterium, Microcystis aeruginosa, to nitrogen stress, phosphorus stress, and growth on organic matter. PLOS ONE 8(7):e69834. Harke MJ, Steffen MM, Gobler CJ, Otten TG, Wilhelm SW, Wood SA, Paerl HW. 2016. A review of the global ecology, genomics, and biogeography of the toxic cyanobacterium, Microcystis spp. Harmful algae 54:4-20. Hart MR, Quin BF, Nguyen ML. 2004. Phosphorus runoff from agricultural land and direct fertilizer effects. J. Environ. Qual. 33(6):1954-1972. Helliwell KE, Lawrence AD, Holzer A, Kudahl UJ, Sasso S, Krautler B, Scanlan DJ, Warren MJ, Smith AG. 2016. Cyanobacteria and eukaryotic algae use different chemical variations of vitamin B12. Curr. Biol. 26(8):999-1008. Howarth RW and Marino R. 2006. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: Evolving views over three decades. Limnol. Oceabogr. 51(1.2):364-376. Huisman J, Matthijs HCP, Visser PM. 2005. Harmful Cyanobacteria, p. 1-23. Springer, Dordrecht, Netherlands. Kaebernick M, Neilan BA. 2001. Ecological and molecular investigations of cyanotoxins production. FEMS Microbiol. Ecol. 35:1-9.

Kaneko T, Nakajima N, Okamoto S, Suzuki I, Tanabe Y, Tamaoki M, Nukamura Y, Kasai F, Wantanabe A, Kawashima K, Kishida Y, Ono A, Shimizu Y, Takahashi C, Minami C, Fujishiro T, Kohara M, Katoh M, Nakazaki N, Nakayama S, Yamada M, Tabata S, Wantanabe MM. 2007. Complete genobic structure of the bloom-forming toxic cyanobacterium Microcystis aeruginosa NIES-843. DNA Research 14:247-256. Komárek J and Komárková J. 2002. Review of the European Microcystis-morphospecies (Cyanoprokaryotes) from nature. Czech Phycol. Olomouc 2:1-24. Landsberg JH. 2002. The effects of harmful algal blooms on aquatic organisms. Rev. Fish. Sci. 10(2):113-390. Li M, Nkrumah PN, Xiao M. 2014. Biochemical composition of Microcystis aeruginosa related to specific growth rate: insight into the effects of abiotic factors. Inland Waters 4:357-362. Lurling M, van Oosterhout F, Faassen E. 2017. Eutrophication and warming boost cyanobacterial biomass and microcystins. Toxins 9(2), 64:1-16. Michalak AM, Anderson EJ, Beletsky D, Boland S, Bosch NS, Bridgeman TB, Chaffin JD, Cho K, Confesor R, Daloglu I, DePinto JV, Evans MA, Fahnenstiel GL, He L, Ho JC, Jenkins L, Johengen TH, Kuo KC, LaPorte E, Liu X, McWilliams MR, Moore MR, Posselt DJ, Richards RP, Scavia D, Steiner AL, Verhamme E, Wright DM, Zagorski MA. 2013. Record-setting algal bloom in Lake Erie caused by agricultural and meteorological trends consistent with expected future conditions. PNAS 110(16):6448-6452. Mobley HL, Island MD, Hausinger RP. 1995. Molecular biology of microbial ureases. Microbiol. Rev. 59:451-480. Mowe MAD, Porojan C, Abbas F, Mitrovic SM, Lim RP, Furey A, Yeo DCJ. 2015. Rising temperatures may increase growth rates and microcystin production in tropical Microcystis species. Harmful Algae 50:88-98. Paerl HW. 2014. Mitigating harmful cyanobacterial blooms in a human – and climatically – impacted world. Life 4:988-1012. Paerl HW, Fulton RS, Moisander PH, Dyble J. 2001. Harmful freshwater algal blooms, with emphasis on cyanobacteria. Sci. World. J. 1:76-113. Paerl HW and Otten TG. 2016. Duelling ‘CyanoHABs’: unraveling the environmental drivers controlling dominance and succession among diazotrophic and non-N2-fixing harmful cyanobacteria. Environ. Microbiol. 18(2):316-324. Paerl HW, Otten TG, Joyner AR. 2016. Moving towards adaptive management of cyanotoxin-impaired water bodies. Microb. Biotechnol. 9(5):641-651. Paerl HW, Pinckney JL. 1996. A mini-review of microbial consortia: Their roles in aquatic production and biogeochemical cycling. Microb. Ecol. 31(3):225-247.

Paerl HW, Xu H, McCarthy MJ, Zhu G, Qin B, Li Y, Gardner WS. 2011. Controlling harmful cyanobacterial blooms in a hyper-eutrophic lake (Lake Taihu, China): The need for a dual nutrient (N & P) management strategy. Water Res. 45(5):1973-1983. Peng L, Liu Y, Chen W, Liu L, Kent M, Song L. 2010. Health risks associated with consumption of microcystin-contaminated fish and shellfish in three Chinese lakes: Significance for freshwater aquacultures. Ecotoxicol. Environ. Saf. 73(7):1804-1811. Pereira S, Zille A, Micheletti E, Moradas-Ferreira P, De Philippis R, Tamagnini P. 2009. Complexity of cyanobacterial exoploysaccharides: composition, structures, inducing factors and putative genes involved in their biosynthesis and assembly. FEMS Microbiol. Rev. 33:917-941. Qin B, Zhu G, Gao G, Zhang Y, Li W, Paerl HW, Charmichael WW. 2010. A drinking water crisis in Lake Taihu, China: Linkage to climatic variability and lake management. Environ. Manage. 45(1):105-112. Ramanan R, Kang Z, Kim B-H, Cho D-H, Jin L, Oh H-M, Kim H-S. 2015. Phycosphere bacterial diversity in green algae reveals an apparent similarity across habitats. Algal Res. 8:140-144. Ramanan R, Kim B-H, Cho, Oh H-M, Kim H-S. 2016. Algae-bacteria interactions: evolution, ecology and emerging applications. Biotechnol. Adv. 34(1):14-29. Rolland JL, Stien D, Sanchez-Ferandin S, Lami R. 2016. Quorum sensing and quorum quenching in the phycosphere of phytoplankton: a case of chemical interactions in ecology. J. Chem. Ecol. 42:1201-1211. Sanseverino I, CondutoD, Pozzoli L, Dobricic S, Lettieri T. 2016. Algal bloom and its economic impact. EUR 27905 EN; doi:10.2788/660478. Schindler DW. 1974. Eutrophication and recovery in experimental lakes: Implications for lake management. Science 184(4139):897-899. Seymour JR. Amin SA, Raina JB, Stocker R. 2017. Zooming in on the phycosphere: the ecological interface for phytoplankton-bacteria relationships. Nat. Microbiol. 2(17065): https://doi.org/10.1038/nmicrobiol.2017.65. Shen H, Niu Y, Xie P, Tao M, Yang X. 2011. Morphological and physiological changes in Microcystis aeruginosa as a result of interactions with heterotrophic bacteria. Freshwater Biol. 56(6):1065-1080. Shirai M, Matumaru K, Ohtake A, Takamura Y, Aida T, Nakano M. 1989. Development of a solid medium for growth and isolation of axenic Microcystis strains (cyanobacteria). Appl. Environ. Microbiol. 55(10):2569-2771. Sivonen K and Jones G. 1999. Cyanobacterial toxins. Toxic Cyanobacteria in Water: A Guide to Their Public Health consequences, Monitoring and Management. E & FN Spon, London, England. 1:40-111.

Sivonen K, Niemela SI, Niemi RM, Lepisto L, Luoma TH, Rasanen LA. 1990. Toxic cyanobacteria (blue-green algae) in Finnish fresh and coastal waters. Hydrobiologia 190(3):267-275. Solomon CM, Collier JL, Berg GM, Glibert PM. 2010. Role of urea in microbial metabolism in aquatic systems: a biochemical and molecular review. Aquat. Microb. Ecol. 59:67-88. Steffen MM, Zhou L, Effler TC, Hauser LJ, Boyer GL, Wilhelm SW. 2012. Comparative metagenomics of toxic freshwater cyanobacteria bloom communities on two continents. PLoS ONE. 7(8): e44002. Steffen MM, Davis TW, McKay RML, Bullerjahn GS, Krausfeldt LE, Stough JMA, Neitzey ML, Gilbert NE, Boyer GL, Johengen TH, Gossiaux DC, Burtner AM, Palladino D, Rowe MD, Dick GJ, Meyer KA, Levy S, Boone BE, Stumpf RP, Wynne TT, Zimba PV, Gutierrez D, Wilhelm SW. 2017. Ecophysiological examination of the Lake Erie Microcystis bloom in 2014: Linkages between biology and the water supply shutdown of Toledo, OH. Environ. Sci. Technol. 51(12):6745-6755. Trolle D, Nielsen A, Rolighed J, Thodsen H, Anderson HE, Karlsson IB, Refsgaard JC, Olesen JE, Bolding K, Kronvang B, Sondergaard M, Jeppesen E. 2015. Projecting the future ecological state of lakes in Denmark in a 6 degree warming scenario. Clim. Res. 64: 55-72. U.S. Environmental Protection Agency (U.S. EPA). 2015. A compilation of cost data associated with the impacts and control of nutrient pollution. Wichelen JV, Vanormelingen P, Codd GA, Vyverman W. 2016. The common bloom- forming cyanobacterium Microcystis is prone to a wide array of microbial antagonists. Harmful Algae 55:97-111. Wiedner C, Visser PM, Fastner J, Metcalf JS, Codd GA, Mur LR. 2003. Effects of light on the microcystin content of Microcystis strain PCC 7806. Appl. Environ. Microbiol. 69(3):1475-1481. Wolf D, Georgic W, Klaiber HA. 2017. Reeling in the damages: Harmful algal blooms’ impact on Lake Erie’s recreational fishing industry. J. Environ. Manage. 199:148-157. Xiao M, Willis A, Burford MA, Li M. 2017. Review: a meta-analysis comparing cell- adhesion in Microcystis colony formation. Harmful Algae 67:85-91. Zhai C, Zhang P, Shen F, Zhou C, Liu C. 2012. Does Microcystis aeruginosa have quorum sensing? FEMS Microbiol. Lett. 336(1):38-44. Zurawell RW, Chen H, Burke JM, Prepas EE. 2005. Hepatotoxic cyanobacteria: a review of the biological importance of microcystins in freshwater environments. J. Toxicol. Environ. Health B8(1):1-37.

Chapter II

The heterotrophic bacterial community of a western Lake Erie Microcystis bloom

and impacts of co-culture

Publication Note

A version of this chapter is in preparation for submission to an academic journal.

Abstract

Heterotrophic bacteria were isolated from a Microcystis bloom in the western

Lake Erie basin. The culturable microbiome of the bloom was characterized.

Metagenomic sequencing was used to characterize the unculturable microbiome of the bloom, which was then compared to the culturable microbiome. Based on characteristics such as urease production, glycolate dehydrogenase production, and pigment production, five isolates were then chosen to be grown with axenic Microcystis to determine the impact of co-culture on Microcystis cells. Co-culture with Acidovorax had a significant impact on the growth rate of Microcystis cells while the filtrate of Enterobacter,

Exiguobacterium, and Paenibacillus appeared to have a protective impact on the

Microcystis cells in higher salt concentrations. Further research is needed to determine the mechanisms of protection of the Microcystis cells by the heterotrophic bacteria in high salt concentrations.

1. Introduction

Cyanobacteria are primary producers that are critical for biogeochemical cycling in freshwater and marine ecosystems (Ramanan et al., 2015). Some cyanobacterial genera are able to form cyanobacterial harmful algal blooms (cHABs). These blooms are a growing global problem in aquatic systems, as the biomass of these blooms, along with the toxins that are sometimes produced can be detrimental to local ecosystems.

Microcystis is a globally distributed, toxic, freshwater and brackish cyanobacterium that is expected to bloom more frequently and with higher biomass as global temperatures continue to rise (Lurling et al., 2017).

Microcystis spp. form blooms in systems such as Lake Erie (North America) annually. Some species of Microcystis are able to produce hepatotoxins known as microcystins. A Microcystis bloom in Lake Erie in 2014 led to a water shutdown in

Toledo, OH that left approximately half a million citizens without water for a weekend due to microcystin levels above the WHO limit for drinking water (Steffen et al., 2017).

These blooms have ultimately cost the city millions of dollars in water treatment

(Associated Press, 2013; US EPA, 2015). While the exact triggers of Microcystis bloom formation are not known, there are many factors that can contribute. A major contributor to the formation of Microcystis blooms is anthropogenic nutrient loading (Paerl, 2014).

This provides freshwater systems with large amounts of nitrogen and phosphorus, both important nutrients for Microcystis. Dual nutrient management strategies have been suggested to combat the formation of cHABs (Paerl et al., 2016). The control of the input of both nitrogen and phosphorus is needed for cHAB prevention as both nutrients are essential for bloom formation.

Microcystis spp. are unique in several ways compared to other cyanobacteria.

Unlike other microcystin-producing cyanobacteria, they are colonial rather than filamentous, forming large colonies surrounded by a thick, sticky mucilage.

Heterotrophic bacteria are known to be tightly coupled with this exopolysaccharide layer that is excreted by Microcystis cells (Shirai et al., 1989). Heterotrophic bacteria have also been shown to impact the colony formation of Microcystis (Shen et al., 2011; Wang et al., 2015). The area that surrounds phytoplankton cells is known as the phycosphere (Bell and Mitchell, 1972). In recent years, interactions between phytoplankton and heterotrophic bacteria in the phycosphere have been shown to be important in shaping the physiology of both eukaryotic and prokaryotic phytoplankton (Amin et al., 2015;

Seymour et al., 2017). However, no work has been done to characterize such interactions in the Microcystis phycosphere, despite its ecological and economic importance.

To resolve the impact of phycosphere bacteria on the physiology of Microcystis, we measured the effect of co-culture on the growth and photosynthetic efficiency of

Microcystis aeruginosa NIES 843 using bacteria isolated from a Microcystis bloom on

Lake Erie in August 2017. Over 100 bacterial strains were isolated from the Lake Erie samples. Metagenomics was also utilized to gain an understanding of the microbial community of the Microcystis bloom and compare the culturable and unculturable bacterial communities.

2. Methods

2.1 Sample Collection

Water samples were collected from Lake Erie on August 9, 2017. Samples were taken from four different stations: WE02 (N 41° 45.777’, W 83° 12.931’), WE04 (N 41°

49.634’, W 83° 11.659’), WE13 (N 41° 44.619’, W 83° 08.081’), and MB18 (N 41°

44.886’, W 83° 24.061’). A 20µm and 80µm mesh plankton net were used to collect samples from each site to ensure only those bacteria tightly associated with the

Microcystis colonies would be isolated. 150 mL were collected and stored on ice until transport back to the laboratory. Additionally, whole water samples were collected in duplicate at each site for background community analysis via metagenomics. 750 mL were filtered onto 0.22µm Sterivex™ (Millipore Sigma) from stations WE02, WE04, and

WE13 while 600 mL were filtered from station MB18. Filters were stored on ice until transported back to the laboratory, where they were stored at -20°C.

2.2 Culturable Microbiome

Isolation and Identification. Two different types of media were used for the isolation of bacterial samples. Both Luria broth (LB) agar and CT (Watanabe and Ichimura, 1977) agar with different nitrogen sources were used. Selective CT plates were made using 100

µM of urea (CT-Urea) and without nitrogen (CT no N). About 500 µL of each water sample were pipetted onto LB agar and CT agar and spread with sterile cell spreaders and the plates were then incubated at 26 °C. LB plates were incubated for 48 hours while CT plates were incubated for seven days. Single colonies were re-streaked onto new plates of the same media type and this was repeated until Gram stains showed that the isolates were pure.

Purified single colonies were used to inoculate 5 mL of LB broth for DNA extraction. The tubes of LB were then incubated in a shaking incubator overnight at 26

°C and 200 rpm. From the broth tubes, 1500 µL were transferred to 1.5 mL microcentrifuge tubes, which were centrifuged at 13000 rpm and the supernatant was

24 removed without disturbing the pellet. The pellet was then resuspended in 500 µL of sterile water and heated at 95 °C in a dry bath for 15 minutes. After heating, the tubes were centrifuged again for one minute. The supernatant, the source of the template DNA, was removed from the tube without disturbing the pellet and was pipetted into a clean microcentrifuge tube. The 16S rRNA gene was then amplified using PCR primers 27f (5’

AGAGTTTGATCMTGGCTCAG 3’) and 1492r (5’ TACGGYTACCTTGTTACGACTT

3’) (Delong, 1992). The reaction mixture included 1.25 µL of 10 µM 27f primer, 10 µL of 10 µM 1492r prime, 12.5 µL of 2x Econo Taq Master Mix (Lucigen), 8.5 µL of PCR water, and 1.5 µL of template DNA (DeLong, 1992). The PCR program was run on a

BioRad C1000 Touch thermocycler and included an initial 95 °C for 5 minutes, and 35 cycles of 95 °C for 1 minute, 50 °C for 30 seconds, and 72 °C for 90 seconds. Products were screened for amplification using a 1% agarose/TAE gel and stained with 3x Gel Red

(Phenix Research Products).

After successful amplification of the 16S sequences, the Qiagen QIAquick PCR purification kit was used to purify the PCR products for Sanger sequencing at Eurofins

Genomics. Sequences were then run in BLAST (blastn) with environmental isolates excluded to identify at the genus level.

Nitrogen Utilization. Isolates that were grown on CT-Urea were tested for urea utilization capabilities. These isolates were used to inoculate urea broth and urea agar slants (Hardy

Diagnostics) and observed for color change to indicate urease activity. Isolates were grown overnight in 5 mL of LB broth at 26 °C for about 24 hours for DNA extraction and

DNA was extracted using the method previously described. A fragment of the ureC gene

25 was amplified using primers IGKAGNP (5’ATHGGIAARGCIGGIAAYCC3’) and

HEDWGA (5’IGYICCCCARTCYTCRTG 3’) (Collier et al., 2009). The PCR program was as follows: 94 °C for one minute, 25 cycles of 94 °C for 30 seconds, 53 °C for 30 seconds, and 72 °C for 45 seconds, with a final extension of 72 °C for 10 minutes. The

PCR products were run on a 2% agarose gel to observe amplification of the ~300-400 bp ureC gene product.

2.3 Bacterial Growth Curves

Five isolates were chosen to be grown in co-culture with axenic Microcystis to measure the physiological impact of these isolates on the photobiont host (Table 1).

Bacterial growth curves were generated to determine the bacterial cell densities at different optical densities (ODs) during the growth phases. A single bacterial colony was used to inoculate CT-TY and the culture was incubated for 24-48 hours at 26°C. The OD of the culture was determined, and 100 mL of fresh CT-TY was inoculated to an OD of

0.05 with the bacterial culture. The OD was again recorded, and a sample of the freshly inoculated culture was diluted and plated onto CT-TY agar in duplicate to determine cell counts. The culture was incubated on a shaking incubator at 26°C (32°C for

Paenibacillus) and samples were taken every hour (Exiguobacterium and Enterobacter) or every two hours (Acidovorax, Deinococcus, and Paenibacillus), diluted, and then plated. OD readings were also taken at the same time. The samples were taken until the culture reached stationary phase. The CT-TY plates were incubated at 26°C for 48 hours.

The colonies were then counted, and the cell densities were calculated to generate the growth curves.

2.4 Co-Culture Conditions

Pure bacterial cultures were grown for 24-48 hours at 26°C in CT-TY and these cultures were used to inoculate the co-culture treatment Microcystis flasks. All treatments had an initial inoculum of 105 cells/mL of Microcystis in CT. A total of six experimental treatments were applied (Table 2). Flasks were inoculated with 105 cells/mL of bacterial culture (ratio of 1:1 Microcystis:bacteria), 107 calls/mL of bacterial culture (ratio of 1:100

Microcystis:bacteria), the filtrate of 105 cells/mL, the filtrate of 107 cells/mL, 0.5% filtrate (0.75 mL), or 1% filtrate (1.5 mL). Filtrate was generated by pushing the bacterial culture of the desired concentrations through a 0.2 µm syringe filter. Control treatments included Microcystis in CT with 0.5% CT-TY (0.75 mL), Microcystis in CT with 1% CT-

TY (1.5 mL), and Microcystis in CT alone. Each flask had a final volume of 150 mL and each treatment was inoculated in triplicate. The flasks were incubated at 26°C with a

12:12 light:dark cycle.

As a proxy for growth, fluorescence readings (625nm) were taken every 48 hours with a Phyto-PAM-II Compact fluorometer (Walz) with a gain setting of 20.

Photosynthetic efficiency was measured via Fv/Fm. Flasks were dark adapted for 20 minutes and an initial fluorescence reading was collected (Fo). After the initial reading

(Fo), 6 µL of 5 µM DCMU was added to the sample to close the Photosystem II reaction centers and a second reading was taken (Fm) (Campbell et al., 1998). The Fo and Fm readings were used to calculate Fv/Fm, or variable fluorescence (Fv/Fm = (Fm-Fo)/Fm). One mL of culture was also harvested every 48 hours and was preserved with Lugol’s iodine for Microcystis cell counts. A hemocytometer was used for Microcystis cell counts. Ten

µL of the culture preserved in Lugol’s Iodine were loaded into the hemocytometer and the slide was viewed under a brightfield microscope at 100x magnification for counts.

2.5 Metagenomics

Sterivex filters were opened using pipe cutters and the filters were cut using sterile syringes and removed with sterile forceps (Cruaud et al., 2017). DNA was extracted using the Qiagen PowerWater DNA kit according to the manufacturer’s instructions. The DNA concentration and quality were assessed using a NanoVue Plus

(GE Healthcare). The DNA was sequenced on the Illumina platform by GENEWIZ.

Demultiplexed FASTQ files were provided by GENEWIZ and the sequence analysis was performed using Qiime 2 (Caporaso et al., 2010; Kuczynski et al., 2012). Sequence correction and trimming was done using the DADA2 pipeline. The need for trimming was determined by drops in quality scored below a predetermined threshold. The Qiime diversity plugin was used for the alpha and beta diversity analysis of the sequence data.

3. Results and Discussion 3.1 Culturable Microbiome Over 100 bacterial isolates were isolated from the Lake Erie samples. Of these isolates, the 16S rRNA gene of 55 isolates was sequenced for identification to the genus level (Table 3). Thirty-two of the isolates were identified as Aeromonas spp. At the phylum level, the majority (80%) of the bacterial isolates were members of the

Proteobacteria phylum. Forty-one of those 44 members of the Proteobacteria phylum belonged to the Gammaproteobacteria class. After the Proteobacteria, bacteria belonging to the phylum Firmicutes were the second most common, with eight isolates all belonging to the class Bacilli. Four genera that were isolated have previously been described as colony-inducing bacteria for Microcystis. These were isolates belonging to the genera

Exiguobacterium, Shewanella, Enterobacter, and Aeromonas (Wang et al., 2016). This may explain why the majority of the identified isolates belonged to the genus Aeromonas.

The five isolates chosen to be grown with Microcystis were an Exiguobacterium sp., an Enterobacter sp., a Deinococcus sp., a Paenibacillus sp., and an Acidovorax sp

(Table 1). Exiguobacterium is a member of the class Bacilli (Figure 1). This bacterium is

Gram-positive with rod-shaped cells. This isolate produced orange pigmented colonies and tested negative for urease. The Enterobacter sp. is a Gram-negative rod-shaped bacterium from the class Gammaproteobacteria (Figure 2) and was positive for urease production. The Deinococcus sp. is a Gram-positive rod-shaped bacterium belonging to the class Deinococci (Figure 3). This isolate produced pink-orange pigmented colonies and tested positive for urease production. Deinococcus was also able to grow on CT-Urea plates. The Paenibacillus sp. is a Gram-negative rod-shaped bacterium that produces spores (Figure 4). This isolate grew best at 32°C while the other organisms were able to grow at 26°C. Paenibacillus belongs to the Firmicutes class Bacilli. The Acidovorax sp. is a Gram-negative rod-shaped bacterium belonging to the class Betaproteobacteria

(Figure 5). This isolate was able to produce urease and grow on CT-Urea.

Table 1. Characterization of heterotrophic bacteria chosen to be grown with axenic

Microcystis.

Isolate Gram Status Cell Morphology Urease Exiguobacterium Gram Positive Rods Negative Enterobacter Gram Negative Rods Positive Paeinibacillus Gram Negative Rods Negative Deinococcus Gram Positive Rods Positive Acidovorax Gram Negative Rods Positive

Table 2. Descriptions of co-culture treatments.

Treatment Description 105 Bacterial Cells CT inoculated with 105 Microcystis cells/mL and 105 bacterial cells/mL for a 1:1 cell ratio 107 Bacterial Cells CT inoculated with 105 Microcystis cells/mL and 107 bacterial cells/mL for a 1:100 cell ratio Filtrate of 105 Bacterial CT with the filtrate of 105 bacterial cells/mL, inoculated Cells with 105 Microcystis cells/mL Filtrate of 107 Bacterial CT with the filtrate of 107 bacterial cells/mL, inoculated Cells with 105 Microcystis cells/mL 0.5% Bacterial Filtrate CT with 0.5% (0.75 mL) bacterial filtrate, inoculated with 105 Microcystis cells/mL 1% Bacterial Filtrate CT with 1% (1.5 mL) bacterial filtrate, inoculated with 105 Microcystis cells/mL 0.5% CT-TY CT with 0.5% (0.75 mL) CT-TY, inoculated with 105 Microcystis cells/mL 1% CT-TY CT with 1% (1.5 mL) CT-TY, inoculated with 105 Microcystis cells/mL CT Control CT inoculated with 105 Microcystis cells/mL

Table 3. Identification of the 55 isolates that were cultured and identified based on 16S sequencing. Phylum Class Genus Number of Isolates Proteobacteria Alphaproteobacteria Rhizobium 1 Proteobacteria Betaproteobacteria Comamonas 1 Proteobacteria Betaproteobacteria Acidovorax 1 Proteobacteria Gammaproteobacteria Aeromonas 32 Proteobacteria Gammaproteobacteria Enterobacter 6 Proteobacteria Gammaproteobacteria Shewanella 1 Proteobacteria Gammaproteobacteria Klebsiella 1 Proteobacteria Gammaproteobacteria Pseudomonas 1 Firmicutes Bacilli Exiguobacterium 5 Firmicutes Bacilli Lactococcus 1 Firmicutes Bacilli Staphylococcus 1 Firmicutes Bacilli Paenibacillus 1 Bacteroidetes Flavobacteria Chryseobacterium 2 Deinococcus- Deinococci Deinococcus 1 Thermus

Figure 1. Phylogenetic tree of organisms in the Exiguobacterium genus that were compared through BLAST to the isolate isolated from Lake Erie that was identified as Exiguobacterium (labelled Exiguobacterium sp. JMULE1).

Figure 2. Phylogenetic tree of organisms in the Enterobacter genus that were compared through BLAST to the isolate isolated from Lake Erie that was identified as Enterobacter (labelled Enterobacter sp. JMULE2).

Figure 3. Phylogenetic tree of organisms in the Deinococcus genus that were compared through BLAST to the isolate isolated from Lake Erie that was identified as Deinococcus (Deinococcus sp. JMULE3).

Figure 4. Phylogenetic tree of organisms in the Paenibacillus genus that were compared through BLAST to the isolate isolated from Lake Erie that was identified as Paenibacillus (labelled Paenibacillus sp. JMULE4).

Figure 5. Phylogenetic tree of organisms in the Acidovorax and Pseudomonas genera that were compared through BLAST to the isolate isolated from Lake Erie that was identified as Acidovorax (labelled Acidovorax sp. JMULE5).

3.2 Response to Co-Culture Axenic Microcystis was grown with five different bacterial isolates to determine if the bacteria had any impacts on the growth and physiology of the Microcystis cells.

The Microcystis cells were able to grow well with all of the bacterial isolates, and none of the isolates appeared to be antagonistic toward the Microcystis cells. The growth rate of the Microcystis grown in co-culture was highest with 107 cells/mL of Acidovorax cells

(Figure 6). When testing for significant differences between the growth rate of all the treatments, all of the treatments passed the Shapiro-Wilk normality test and an ANOVA was performed. There was a significant difference between the Microcystis growth rate of all the treatments (including the CT control) (p=<0.001). A multiple comparison test versus the CT control treatment was done (Holm-Sidak method). The growth rate of the

Microcystis grown with 107 Acidovorax cells/mL was the only significant treatment compared to the CT control (p=<0.001). Growth of Microcystis with this strain of

Acidovorax appears to confer a growth advantage when compared to Microcystis grown alone, indicating a potential exchange of beneficial metabolites or some other stimulating function of the Acidovorax cells.

The Microcystis cells/mL on the final day of the experiment (day 30; day 28 for for Microcystis grown with Paenibacillus only) were compared between the treatments

(including CT control) (Figure 7). The Microcystis cells/mL was based on direct cell counts from the samples. The data for these cell counts on day 30 were all normally distributed. Based on an ANOVA to determine differences between all of the groups

(including the CT control), there was a significant difference between the final

Microcystis cell counts in all of the treatment groups (p = 0.005). A multiple comparisons test of co-culture treatment groups versus the CT control was done (Holm-Sidak method).

There was a significant difference between the Microcystis cells/mL on the final day of the CT control and the Microcystis cells/mL in the 105 Exiguobacterium cells/mL treatment (p=0.048).

The Fv/Fm shows the photosynthetic stress of the Microcystis cells (Figure 8). The

Fv/Fm of the bacterial co-culture treatments were compared during three stages of growth

(based on Microcystis cell counts): the beginning of exponential phase, mid exponential phase, and the beginning of stationary phase (Table 4). The Fv/Fm of the Microcystis in the co-culture treatments at the beginning of exponential phase failed the Shapiro-Wilks normality test. A Kruskal-Wallis one-way analysis was performed to determine if there was a difference between the Fv/Fm of the Microcystis in the co-culture treatment groups

(including the CT control) on day 6. There was a significant difference between the Fv/Fm of the Microcystis in all of the treatment groups (p = 0.016). When compared directly to the CT control group in a multiple comparisons test (Dunn’s method), the Fv/Fm of the

Microcystis in the 105 Acidovorax cell treatment was significantly different (p = 0.024).

5 On this day (day 6), the Fv/Fm of the Microcystis in the 10 Acidovorax was the lowest of all the treatments, indicating that the Microcystis in this co-culture was more stressed than the Microcystis in the other treatments.

The Fv/Fm data during mid exponential phase (day 14) and the start of stationary phase (day 28) were normally distributed based on Shapiro-Wilks test of normality. An

ANOVA was run to determine if there were differences in the Fv/Fm of the Microcystis in the co-culture treatments (including the CT control) during those growth stages. At mid exponential phase (day 14) there was a significant difference between the Fv/Fm of the

Microcystis in the treatment groups (p = <0.001). When compared to the CT control

treatment in a multiple comparisons test (Holm-Sidak method), the Fv/Fm of the

Microcystis in the 105 Exiguobacterium cells treatment was significantly different (p =

5 <0.001). The FvFm of the Microcystis in the 10 Paenibacillus cells/mL treatment was also significantly different (p=<0.001). At the start of stationary phase (day 28), there was no significant difference between the between the Fv/Fm of the Microcystis in all of the co-culture treatments (p = 0.342).

Figure 6. Fluorescence of Microcystis when grown with different concentrations of bacterial cells.

Figure 7. Direct cell counts of Microcystis grown with different concentrations of bacterial isolates.

Figure 8. Fv/Fm of Microcystis when grown with different concentrations of bacterial cells. The green horizontal lines indicate a healthy zone while the red horizontal line shows a stressed zone (below 0.3).

Table 4. Statistical difference between all co-culture groups (co-culture treatments and CT control) during different phases of growth. Start of exponential phase was day 6, mid exponential phase was day 14, and start of stationary phase was day 28. Normal distribution was based on Shapiro-Wilks normality test. If data failed the Shapiro-Wilks normality test, p-value was based on Kruskal-Wallis one-way analysis of variance. If data were normally distributed, p-value was based on one-way analysis of variance (ANOVA). Growth Phase Normal Distribution P-value Start of Exponential No 0.016 Mid Exponential Yes <0.001 Start of Stationary Yes 0.342

3.3 Response to Filtrate

Microcystis cells were grown in different concentrations of bacterial filtrate to differentiate between any impacts of the bacterial isolates on the growth and physiology of Microcystis caused by to cell-to-cell communication or potential compounds released by the bacteria (Figure 9). Microcystis cells were grown with the filtrate of 105 bacterial cells and the filtrate 107 bacterial cells because those are the same cell concentrations with which the co-culture treatments were done. Microcystis was also grown in CT media with 0.5% bacterial filtrate and 1.5% bacterial filtrate to be able to compare a standard amount between the filtrate of the different bacterial isolates. The growth rates of the

Microcystis in the different filtrate treatments were compared. The growth rates failed the

Shapiro-Wilk normality test and a Kruskal-Wallis one-way analysis of variance was performed. When the growth rates of the Microcystis that was able to grow in all the different treatments were compared, they were found to be statistically different

(p=0.002). Treatments excluded from the test due to lack of Microcystis growth included

0.5% and 1% Exiguobacterium filtrate, 1% filtrate and filtrate of 107 cells/mL of

Deinococcus, 0.5% and 1% filtrate and filtrate of 107 cells/mL of Acidovorax, and the

CT-TY controls. A multiple comparisons test versus the CT control was performed and the growth rate of the Microcystis grown in the filtrate of 106 cells/mL of Paenibacillus was found to be significantly lower than the growth rate of the Microcystis in the CT control treatment (p=0.038).

The final Microcystis cell counts (cells/mL) on day 30 of the experiment (day 28 for the Microcystis in Paenibacillus treatments) were compared (Figure 10). When only the filtrate treatments and control treatments that grew (no final cell counts were 0) were

45 compared, the data were normally distributed. An ANOVA compared the Microcystis cells/mL in all of these treatments that were able to grow (105 Exiguobacterium cell filtrate, 105 Enterobacter cell filtrate, 107 Enterobacter cell filtrate, 0.5% Enterobacter filtrate, 1% Enterobacter filtrate, 105 Deinococcus cell filtrate, 0.5% Deinococcus filtrate,

105 Acidovorax cell filtrate, 105 Paenibacillus cell filtrate, 107 Paenibacillus cell filtrate,

0.5% Paenibacillus filtrate, 1% Paenibacillus filtrate, and CT control) and there was a significant difference between all of the treatments (p = 0.009). A multiple comparisons test versus the CT control was performed and there was no significant difference between the final cell counts of the Microcystis in any of the filtrate treatments and the

Microcystis in the CT control.

The Fv/Fm of the Microcystis in the filtrate treatments (Figure 11) were also compared during three stages of growth (based on Microcystis cell counts): the beginning of exponential phase, mid-exponential phase, and the beginning of stationary phase

(Table 5). The data at all three phases of growth failed the Shapiro-Wilks normality tests and a Kruskal-Wallis one-way analysis of variance was performed for each of the stages.

The Fv/Fm of the Microcystis in all of the treatments (including CT-TY and CT controls) at the beginning of the exponential phase (day 6) were significantly different (p =

<0.001). A multiple comparisons test versus the CT control was done and the Fv/Fm of the

Microcystis in six treatments was significantly different that the Fv/Fm of the Microcystis in the CT control. These treatments included 0.5% Exiguobacterium filtrate (p=<0.001),

1% Exiguobacterium filtrate (p=0.002), 0.5% Acidovorax filtrate (p=0.016), 1%

Acidovorax filtrate (p=0.005), and the filtrate of 107 Acidovorax cells/mL (p=0.026). The

0.5% and 1% Exiguobacterium filtrate treatments were inoculated with a different culture

46 of Microcystis than the other treatments and this, not the filtrate, was likely the cause of the photosynthetic stress and ultimate cell death in those cultures. During the mid- exponential phase (day 14), there was a significant difference between the Fv/Fm of the

Microcystis in all of the treatment groups (including the CT-TY and CT controls) (p =

<0.001). The Fv/Fm of the Microcystis in 0.5% Exiguobacterium filtrate (p=0.002), 1%

Exiguobacterium filtrate (p=0.003), 0.5% Acidovorax filtrate (p=0.028), 1% Acidovorax filtrate (p=0.005), and filtrate of 107 Deinococcus cells/mL (p=0.002) treatments were all significantly different than the Fv/Fm of the Microcystis in the CT control treatment on day 14. At the beginning of stationary phase (day 28), the Fv/Fm values of the Microcystis in all the treatments were significantly different (p=<0.001). The Fv/Fm of the

Microcystis in the 1% Acidovorax (p=0.038), filtrate of 107 Acidovorax cells/mL

(p=0.046), and 107 Deinococcus cells/mL (p=0.002) treatments were all significantly different than the Fv/Fm of the Microcystis in the CT control treatment on day 28 of the experiment.

The filtrate of some of the bacterial isolates appeared to have protective effects on the Microcystis. The Microcystis in all of the Enterobacter filtrate treatments survived.

Of the four Enterobacter filtrate treatments, the Microcystis in the 107 cell filtrate treatment had the highest final fluorescence, followed by the Microcystis in the 105 cell filtrate treatment. Because the Microcystis was able to grow in these filtrate treatments but was unable to grow in the CT-TY treatments, it is likely the Enterobacter cells either utilize the compounds in the CT-TY that are harmful to Microcystis or are able to release compounds that protect Microcystis in the higher salt concentrations of CT-TY media.

The Microcystis in all of the Paenibacillus filtrate treatments was also able to grow.

While the growth rates and Fv/Fm of the Microcystis in some of these treatments were lower than the CT control, the ability of the Microcystis cells to grow in these filtrate treatments and not the CT-TY treatments indicates protective impacts of the

Paenibacillus. The Exiguobacterium filtrate also appeared to have some protective impact on the Microcystis. The Microcystis in the 105 Exiguobacterium cell filtrate treatment was able to grow and the Microcystis in the 107 Exiguobacterium cell filtrate treatment was able to grow as well (Figure 9; Figure 10). The Microcystis in the 107

Exiguobacterium cell treatment had a lower Fv/Fm, indicating photosynthetic stress

(Figure 11). While the Microcystis cells were more stressed, their ability to grow in the treatment and not in the 1% CT-TY treatment indicated a protective impact. The

Microcystis in the 0.5% and 1% Exiguobacterium filtrate was unable to grow.

Figure 9. Fluorescence of Microcystis cells grown with bacterial filtrate treatments.

Figure 10. Direct cell counts of Microcystis cells in different bacterial filtrate treatments.

Figure 11. Fv/Fm of Microcystis cells grown in different filtrate conditions. The green horizontal lines indicate a healthy zone while the red horizontal line shows a stressed zone (below 0.3).

Table 5. Statistical difference between all filtrate groups (filtrate treatments, CT-TY controls, and CT control) during different phases of growth. Start of exponential phase was day 6, mid exponential phase was day 14, and start of stationary phase was day 28. Normal distribution was based on Shapiro-Wilks normality test. All data failed the Shapiro-Wilks normality test, so p-value were based on Kruskal-Wallis one-way analysis of variance. Growth Phase Normal Distribution P-value Start of Exponential No <0.001 Mid Exponential No <0.001 Start of Stationary No <0.001

4. Metagenomics A total of 4042 OTUs were identified from the four Lake Erie stations.

Actinobacteria was the most common phylum found in the first sample from station

WE02 (Figure 12). The second sample from that station (2B), had the most OTUs belonging to Cyanobacteria, followed by Actinobacteria and Proteobacteria. The majority of OTUs from the first sample from station WE04 were also Cyanobacteria, followed by

Proteobacteria and Actinobacteria. The second sample from station WE04 (4B) had the majority of OTUs belonging to Actinobacteria, followed by Proteobacteria and

Cyanobacteria. The largest number of OTUs from station WE13 belonged to the phylum

Actinobacteria, followed by Proteobacteria, Bacteroidetes, and Cyanobacteria. This sample was also the only one with OTUs belonging to the phylum Gemmatimonadetes.

The second sample from that station (13B) had mainly Cyanobacteria, followed by

Proteobacteria and Actinobacteria. The majority of bacteria belonging to the

Proteobacteria phylum were Alphaproteobacteria and Betaproteobacteria (Figure 13).

This was contrary to the bacterial community that was isolated, as the majority of these isolates belonged to the Gammaproteobacteria class. Steffen et al., 2012 previously found

Proteobacteria to be the largest phylum of heterotrophic bacteria in a Lake Erie

Microcystis bloom. Shotgun metagenomics was used in that study, while 16S metagenomics was applied here. The use of different sequencing techniques may account for some of the differences in these results.

Figure 12. The percentage of OTUs belonging to different prokaryotic phyla at each of the sampling sites.

Figure 13. The percentage of OTUs belonging to different classes within the

Proteobacteria phylum at each sampling station.

5. Conclusions Bacterial strains were isolated from a Microcystis bloom on Lake Erie in 2017.

The results of the culturable microbial community differed from those of the metagenomic data. The differences found in the community composition of culturable bacteria compared to the unculturable community demonstrates the importance of sequencing data. The metagenomic data provided a potentially more accurate view of overall community structure compared to culturing isolates in the lab.

Axenic Microcystis aeruginosa NIES 843 cells were grown with different bacterial cell and filtrate treatments to determine if bacteria isolated from a Lake Erie

Microcystis bloom had any effect on the growth and physiology of the Microcystis cells.

The growth of Microcystis cells was impacted by the presence of 107 Acidovorax cells/mL. The filtrate of Enterobacter and Exiguobacterium isolates also appeared to have some protective impact on Microcystis. These results could suggest that bacteria in a natural bloom environment may protect Microcystis cells from undesirable environmental conditions, such as higher salt concentrations. As it was demonstrated here, the protective impact of the bacteria does not necessarily require direct cell contact. Bacterial utilization of compounds harmful to Microcystis of the release of compounds that promote

Microcystis growth would give Microcystis an advantage in a natural bloom setting.

Works Cited Amin SA, Hmelo LR, van Tol HM, Durham BP, Carlson LT, Heal KR, Morales RL, Berthiaume CT, Parker MS, Djunaedi B, Ingalls AE, Parsek MR, Moran MA, Armbrust EV. 2015. Interaction and signaling between a cosmopolitan phytoplankton and associated bacteria. Nature 522:98-101. Associated Press (2013). Algae Blooms in Lake Erie becoming a Threat to Drinking Water. Cleveland, OH: Associated Press. Bell W and Mitchell R. 1972. Chemotactic and growth responses of marine bacteria to algal extracellular products. Biol. Bull. 143(2):265-277. Campbell D, Hurry V, Clarke AK, Gustafsson P, Oquist G. 1998. Chlorophyll fluorescence analysis of cyanobacterial photosynthesis and acclimation. MMBR 62(3):667-683. Caporaso JG, Kuczynski, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Pena AG, Goodrich JK, Gordon JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J, Sevinsky JR, Turnbaugh PJ, Walters WA, Widmann J, YatsunenkoT, Zaneveld J, Knight R. 2010. QIIME allows analysis of high-throughput community sequencing data. Nature Methods 7:335-336. Collier JL, Baker KM, Bell SL. 2009. Diversity of urea-degrading microorganisms in open-ocean and estuarine planktonic communities. Environ. Micobiol. 11(12):3118-3131. Cruaud P, Vigneron A, Fradette MS, Charette SJ, Rodriguez MJ, Dorea CC, Culley AI. 2017. Open the SterivexTM casing: An easy and effective way to improve DNA extraction yields. Limnol. Ocreanogr.-Meth. 15(12):1015-1020. DeLong EF. 1992. Archaea in coastal marine environments. PNAS. 89(12):5685-5689. Harvey EL, Deering RW, Rowley DC, Gamal AE, Schorn M, Moore BS, Johnson MD, Mincer TJ, Whalen KE. 2016. A bacterial quorum-sensing precursor induces mortality in the marine coccolithophore, Emiliania huxleyi. Front. Microbiol. 7(59):1664-302x. Kuczynski J, Stombaugh J, Walters WA, Gonzalez A, Caporaso JG, Knight R. 2012. Using QIIME to analyze 16S rRNA gene sequences from microbial communities. Curr. Protoc. Microbiol. 27(1):1E.5.1-1E.5.20. Lurling M, van Oosterhout F, Faassen E. 2017. Eutrophication and warming boost cyanobacterial biomass and microcystins. Toxins 9(2), 64:1-16. Paerl HW. 2014. Mitigating harmful cyanobacterial blooms in a human – and climatically – impacted world. Life 4:988-1012. Paerl HW, Otten TG, Joyner AR. 2016. Moving towards adaptive management of cyanotoxin-impaired water bodies. Microb. Biotechnol. 9(5):641-651.

Ramanan R, Kang Z, Kim B-H, Cho D-H, Jin L, Oh H-M, Kim H-S. 2015. Phycosphere bacterial diversity in green algae reveals an apparent similarity across habitats. Algal Res. 8:140-144. Seymour JR. Amin SA, Raina JB, Stocker R. 2017. Zooming in on the phycosphere: the ecological interface for phytoplankton-bacteria relationships. Nat. Microbiol. 2(17065): https://doi.org/10.1038/nmicrobiol.2017.65. Shen H, Niu Y, Tao M, Yang X. 2011. Morphological and physiological changes in Microcystis aeruginosa a result of interactions with heterotrophic bacteria. Freshw. Biol. 56(6):1065-1080. Shirai M, Matumaru K, Ohtake A, Takamura Y, Aida T, Nakano M. 1989. Development of a solid medium for growth and isolation of axenic Microcystis strains (cyanobacteria). Appl. Environ. Microbiol. 55(10):2569-2771. Steffen MM, Zhou L, Effler TC, Hauser LJ, Boyer GL, Wilhelm SW. 2012. Comparative metagenomics of toxic freshwater cyanobacteria bloom communities on two continents. PLoS ONE. 7(8): e44002. U.S. Environmental Protection Agency (U.S. EPA). 2015. A compilation of cost data associated with the impacts and control of nutrient pollution. Wang W, Zhang Y, Shen H, Xie P, Ju J. 2015. Changes in the bacterial community and extracellular compounds associated with the disaggregation of Microcystis colonies. Biochem. Syst. Ecol. 61:62-66. Wang W, Shen H, Shi P, Chen J, Xie P., 2016. Experimental evidence for the role of heterotrophic bacteria in the formation of Microcystis colonies. J. Appl. Phycol. 28(2):1111-1123. Watanabe MM and Ichimura T. 1977. Fresh- and salt-water forms of Spirulina platensis in axenic cultures. Bull. J. Soc. Phycol. 25:371-377.

Chapter III

Comparative genomics of heterotrophic bacteria isolated from a Lake Erie

Microcystis bloom

Publication Note

A version of this chapter is in preparation for submission to Environmental Microbiology.

1. Abstract

Heterotrophic bacteria were isolated from a Microcystis bloom in the western

Lake Erie basin. Five isolates were chosen for whole genome sequencing with PacBio.

The genomes were assembled with the CLC Genomics Workbench plugin, CLC Genome

Finishing Module. The genomes were then annotated with RAST and PGAP. A single contig was obtained for the Acidovorax spp. The genomes were then analyzed for genes that could be used in potential interactions and communication with Microcystis. Genes that have been previously described in bacteria-phytoplankton interactions were found in the bacterial genomes. Genes involved in functions such as nitrogen metabolism, carbohydrate metabolism, quorum sensing, and vitamin production were found in the bacterial genomes. This data could be used to further investigate the interactions between

Microcystis and heterotrophic bacteria and the potential impacts heterotrophic bacteria can have on the growth and physiology of Microcystis cells.

2. Introduction

Cyanobacterial harmful algal blooms (cHABs) occur annually in both freshwater and marine systems. These blooms are formed by multiple species of cyanobacteria and have the potential to be destructive to aquatic ecosystems due to their biomass or the release of toxins into the environment (Anderson et al., 2002; De Figueiredo et al., 2004;

Backer et al., 2015). Microcystis is a common genus of bloom-forming cyanobacteria.

Some species of Microcystis are able to produce hepatotoxins known as microcystins.

Bacteria are capable of degrading microcystins produced by cyanobacteria (Jones et al.,

1994; Kato et al., 2007). Microcystis spp are colonial cyanobacteria and these colonies are surrounded by an exopolysaccharide layer with which bacteria are tightly coupled.

Different strains of bacteria have been found to impact the formation of

Microcystis colonies and exopolysaccharide production (Shen et al., 2011; Wang et al.,

2016). The zone that surrounds the phytoplankton where bacteria reside is known as the phycosphere (Bell and Mitchell, 1972). It is expected that bacteria and phytoplankton within the phycosphere interact and exchange nutrients. Heterotrophic bacteria are able to utilize the carbon released by Microcystis while Microcystis may be able to benefit from products released by the heterotrophic bacteria, such as vitamins or other nutrients.

In the last decade, metagenomic studies have given us new insight into the community makeup of the bacteria associated with Microcystis blooms in sites such as

Lake Erie and Lake Tai, however, there has been limited information on the whole genomes of individual members of these bloom communities (Xie et al., 2016). Here, we report on the genomic content of five individual bacteria isolated from a Microcystis bloom in western Lake Erie in August 2017. Access to the genomic information of

61 individual heterotrophic bacteria has provided new insight into the potential microbial interactions and metabolic pathways that occur within Microcystis blooms.

3. Methods

3.1 DNA Extraction and Sequencing

The five isolates chosen for whole genome sequencing were isolated according to the protocol outlined in Chapter 2 of this thesis. These isolates were Exiguobacterium sp.,

Enterobacter sp., Deinococcus sp., Paenibacillus sp., and Acidovorax sp. The isolates were grown in Luria Broth (LB) and the Paenibacillus sp. was grown in Tryptic Soy

Broth (TSB) for DNA extraction. The DNeasy UltraClean Microbial Kit (Qiagen) was used for DNA extraction according to manufacturer’s instructions. A NanoVue Plus (GE

Healthcare) was used to check the quantity and purity of the DNA. The genomic DNA was sent to Genewiz for sequencing on the PacBio Sequel System.

At the Genewiz facility, the quality and quantity of the genomic DNA was assessed with a Nanodrop, Qubit DNA assay, and a gel. After it was determined that the

DNA passed the purity and quality requirements, the PacBio SMRTbell library was prepared according to the manufacturer’s instructions. Briefly, the library preparation began with fragmenting the DNA. The DNA then went through damage repair, followed by end repair. The next step was ligation of the barcoded adapters and the pooling of the libraries. The library pool was purified followed by the primers annealing and the polymerase binding. The SMRTbell libraries were then sequenced with the PacBio

Sequel System.

BAM files were generated from the PacBio Sequel platform. The SMRTLink suite was used to demultiplex the subreads. The default parameters were used to run the

62 demultiplexing algorithm. These demultiplexed sequence files (BAM and FASTQ) were then provided by the GENEWIZ sequencing facility for assembly and annotation.

3.2 Genome Assembly and Annotation

The genomes were assembled using the PacBio de novo assembly pipeline on the

CLC Genomics Workbench plugin CLC Genome Finishing Module (Qiagen). The assembly workflow was run with the default settings. First, the FASTQ files with raw

PacBio reads were imported into CLC Genomics Workbench. The first tool in the de novo assembly pipeline was the tool for correcting reads. This tool corrected sequencing errors in the raw reads and resolved untrimmed adapters. The error-corrected reads were then assembled into contigs with the “de novo Assemble PacBio Reads” tool. The corrected reads were then mapped to the contigs to close gaps and join contigs. The corrected reads were then mapped to the larger contigs and the final contigs were analyzed to find any problematic regions.

The genomes were annotated with the NCBI prokaryotic genome annotation pipeline (PGAP) (Tatusova et al., 2016) and RAST (Aziz et al., 2008; Brettin et al.,

2015). The annotated genomes were viewed on the RAST SEED Viewer (Overbeek et al., 2014). Maps of the genomes were generated with DNAPlotter (Carver et al., 2009).

4. Results and Discussion

4.1 Sequencing Output, Assembly, and Annotation

The number of raw reads in the files provided by GENEWIZ ranged from

831,129,345 reads (Deinococcus) to 1,818,930,107 reads (Enterobacter) (Table 1). The sequences were then corrected in the first step of the PacBio de novo assembly pipeline in

63 the CLC Genome Finishing module. The output of corrected reads was between

202,560,865 reads (Deinococcus) to 452,019,305 reads (Enterobacter) (Table 1).

The number of contigs obtained from the genome assemblies ranged from 1 to 20 contigs. Acidovorax was the only isolate for which a single contiguous contig was obtained (Table 2). Enterobacter had the longest genome, at 5.7Mb, which is slightly longer than other sequenced isolates in the Enterobacter genus. The genome lengths of the other isolates were comparable to other sequenced isolates within those genera. The

GC contents were also similar to other sequenced isolates.

4.2 Nitrogen

Nitrogen is a limiting nutrient for Microcystis. In cyanobacteria, including

Microcystis, nitrogen assimilation is regulated by the ntcA gene. This gene has been shown to be repressed when levels of ammonium in the water exceed 1µmol L-1 (Chaffin and Bridgeman, 2014). Microcystis favors ammonium as a nitrogen source, but in times when ammonium is scarce, nitrate- and nitrate-reductase, as well as urease production is promoted (Chaffin and Bridgeman, 2014). The presence of several different genes suggests the potential for these organisms to act as an ammonium source for Microcystis in the environment. All of the genomes except Exiguobacterium contain genes for the reduction of nitrate and nitrite to ammonium. The denitrifying reductase gene clusters in the Acidovorax genome includes genes for nitric oxide reductase and nitrous oxide reductase. Acidovorax also has genes for cyanate hydrolysis. Cyanate is a by-product of the urea cycle and the products of cyanate hydrolysis include bicarbonate and ammonium ions (Wen and Brooker, 1994). The ammonium produced by the heterotrophic bacteria

64 from the breakdown of other compounds could potentially be utilized by Microcystis to help satisfy its nitrogen requirements.

Table 1. The concentration of nucleic acid (Nanodrop 2000) and 260/280 of samples sent to GENEWIZ followed by the total number of raw reads for each isolate and the number of corrected reads after the sequences were corrected in the PacBio de novo assembly pipeline.

Isolate DNA (ng/µL) 260/280 Total Raw Reads Corrected Reads Exiguobacterium 462.8 1.91 1,297,742,008 329,791,775 Enterobacter 479.9 1.90 1,818,930,107 452,019,305 Deinococcus 693.0 1.93 831,129,345 202,560,865 Paenibacillus 176.4 1.89 1,422,433,251 340,274,897 Acidovorax 459.9 1.92 1,490,700,321 370,362,796

Table 2. Characteristics of genome assemblies obtained from the PacBio de novo assembly pipeline in the CLC Genome Finishing Module. Isolate Number of N50 GC % Total Length Contigs Exiguobacterium 2 3,111,939 47.13 3,148,435 Enterobacter 20 613,519 54.79 5,742,593 Deinococcus 5 3,281,798 69.75 4,224,197 Paenibacillus 19 392,282 49.95 5,404,135 Acidovorax 1 5,455,809 64.48 5,455,809

Table 3. The number of coding sequences called by the genome annotation tool RAST. Isolate Coding Sequences Exiguobacterium 3289 Enterobacter 5736 Deinococcus 4208 Paenibacillus 5888 Acidovorax 5101

Table 4. Nitrogen genes called by RAST and their roles in the bacterial isolates.

Isolate(s) Gene Role Enterobacter, nirD Nitrate reductase small subunit Deinococcus, Paenibacillus, Acidovorax Enterobacter, nirB Nitrate reductase large subunit Deinococcus, Paenibacillus, Acidovorax Enterobacter, narG Respiratory nitrate reductase alpha chain Acidovorax Enterobacter, narH Respiratory nitrate reductase beta chain Acidovorax Enterobacter, narI Respiratory nitrate reductase gamma chain Acidovorax Enterobacter narJ Respiratory nitrate reductase delta chain Enterobacter, norR Anaerobic nitric oxide reductase transcription Acidovorax regulator Enterobacter, nsrR Nitrite-sensitive transcriptional repressor Exiguobacterium Deinococcus, gltB Glutamate synthase large chain Exiguobacterium, Acidovorax Deinococcus, gltD Glutamate synthase small chain Exiguobacterium, Acidovorax Deinococcus glnN Glutamine synthetase type III Acidovorax cynS Cyanate hydratase Acidovorax cynR Cyn operon transcriptional activator Acidovorax nosFn nosR, nosY Nitrous oxide reductase maturation protein Acidovorax norB Nitric-oxide reductase subunit B

Urease is an enzyme that catalyzes the hydrolysis of urea to ammonia and carbamate. The genomes of Enterobacter, Paenibacillus, and Acidovorax contains genes for the alpha (ureC), beta (ureB), and gamma (ureA) subunits of the urease enzyme complex. The Deinococcus genome contains genes for the alpha (ureC) and gamma

(ureA) urease subunits. The Deinococcus and Acidovorax genomes contain all of the urease accessory genes (ureEFGD), while Enterobacter and Paenibacillus have some of the accessory genes (ureEFD and ureFGD respectively). Urease is a metalloenzyme that binds to nickel (Dixon et al., 1975; Boer et al., 2013). The Enterobacter and Acidovorax contain the nickel-binding accessory gene ureJ-hupE. Genes for nickel incorporation proteins (hypAB) were found in the Enterobacter and Deinococcus genomes. The utilization and decomposition of urea to ammonium and CO2 by heterotrophic bacteria during a bloom could supply Microcystis with some of its nitrogen (Cho et al., 1996) and carbon demand.

4.3 Carbon utilization/Carbohydrates

The monosaccharide composition of Microcystis extracellular polysaccharides has been characterized in the past (Nakagawa et al., 1987; Forni et al., 1997). Xylose was one of the monosaccharides found to be in Microcystis EPS. The genomes of Paenibacillus, and Enterobacter contain genes for xylose utilization. Paenibacillus contained genes for xylose isomerase, transporters, and binding components while Enterobacter had genes for the XylFGHR proteins for transcriptional regulation, transport, and ATP binding. All of the isolates had genes for mannose utilization. All of the isolates could produce mannose-6-phosphate isomerase. D-mannose isomerase is another enzyme for mannose metabolism that Acidovorax has genes for. Both xylose and mannose have been described

70 in bacteria-algae interactions (Giroldo et al., 2007) and like xylose, mannose is found in

Microcystis EPS (Nakagawa et al., 1987; Forni et al., 1997). Deinococcus, Paenibacillus, and Acidovorax have the glcD gene for glycolate dehydrogenase. Glycolate is produced from the oxygenase activity of rubisco during photorespiration and is therefore released by phytoplankton. The potential for utilization of glycolate released by phytoplankton has been examined in marine systems and lakes (Lau and Armbrust, 2006; Paver and Kent,

2010).

4.4 Iron

Iron is a trace element that is essential for phytoplankton growth. Microcystis spp. have been shown to have impacted phytopigments and photochemical efficiency during iron deprivation (Xing et al., 2007). Siderophores are vital for iron scavenging and acquisition during times of iron deprivation. Microcystis spp, are also produce siderophores in times of iron deprivation (Imai et al., 1999; Xing et al., 2007). The

Enterobacter, Paenibacillus, and Deinococcus, genomes all contain genes encoding various siderophores.

The Enterobacter genome has genes encoding FepBCDEG proteins for the transport of ferric enterobactin. Genes encoding the enterobactin biosynthesis pathway proteins EntBSH were also present. Enterobactin siderophores are characteristic of the

Enterobacteriaceae family. These are the strongest siderophores with a high affinity for iron sequestration (Carrano and Raymond, 1979; Fischbach et al., 2006). The

Enterobacter genome also contains iucA-D genes for aerobactin synthesis. Aerobactin siderophores do not have as high an affinity for iron as enterobacins, but they have been found to have an advantage for bacterial growth in iron limited conditions (Williams,

1979; Williams and Warner, 1980). Enterobacter spp have also been described with both types of siderophores (Van Tiel-Menkveld et al., 1982).

The genome of Deinococcus contains genes for isochorismate synthase.

Isochorismate synthase is a precursor of siderophores, including enterobactin (Yokoo et al., 2018). Isochorismate synthase has also been found to be important for the synthesis of salicylic acid for plant defense (Wildermuth et al., 2001). The Paenibacillus genome contained genes related to bacillibactin and anthrachelin siderophores. The genome genes for FeuA-C proteins for Fe-bacillibactin transport (Ollinger et al., 2006). Bacillibactin are catechol-based siderophores that are structurally similar to enterobactin siderophores.

Bacillibactin siderophores are produced by different members of the Bacillus genus, including Bacillus anthracis (Wilson et al., 2006). These siderophores have also been described in a Paenibacillus honey bee pathogen (Hertlein et al., 2014). The

Paenibacillus genome also contained genes for anthrachelin uptake transporters.

Anthrachelin siderophores have been previously described in Bacillus anthracis

(Cendrowski et al., 2004). The presence of siderophores in the bacterial genomes could indicate the potential of those bacteria to release bioavailable iron into the system, which could then be used by Microcystis.

The Ton and Tol transport systems are used to transport ferric-siderophore complexes and vitamin B12 across the cell membrane (Braun, 1995). The genomes of the

Acidovorax, Deinococcus, and Enterobacter isolates contained genes involved in the Ton and Tol transport systems. All three organisms had genes for TonB-dependent receptors.

Enterobacter and Acidovorax had genes for the TolA protein. The Enterobacter genome contains the gene encoding aerobactin siderophore receptor, iutA.

4.5 Heavy Metal and Antibiotic Resistance

All of the bacterial genomes contained genes for resistance to fluoroquinolones, antimicrobial agents that inhibit DNA synthesis (Wolfson and Hooper, 1989).

Exiguobacterium and Enterobacter have genes for streptothricin resistance. Deinococcus and Paenibacillus have genes for vancomycin resistance. Enterobacter have genes for fosfomycin resistance.

The isolate genomes also contain genes for resistance against the metals cobalt, zinc, and cadmium, as well as genes for either the tolerance of or resistance to copper.

Exiguobacterium and Paenibacillus have genes for mercury resistance.

4.6 Secondary Metabolites

Auxins are plant growth promoting hormones that are critical for plant development. The role of auxins in microbial interactions have been observed in marine diatom interactions with associated bacteria (Amin et al., 2015). Amin et al. (2015) found that a bacterial consortium promoted diatom cell division due to the secretion of an auxin synthesized from tryptophan released by the diatom. All five of the sequenced genomes contained genes for the biosynthesis of auxins. All of the genomes also contain genes for tryptophan synthesis. Tryptophan has been found to be an important precursor for indole-

3-acetic acid (IAA), the main auxin that occurs in plants. There are tryptophan-dependent pathways in some organisms to produce IAA from tryptophan. The production of auxins by these isolates could indicate potential growth promoting effects of the bacteria on

Microcystis.

Species of Paenibacillus have been found to be growth promoting in the rhizosphere of some plants due to the production of growth promoting secondary

73 metabolites, such as auxins, and antifungal compounds (Beatty and Jensen, 2002; Selim et al., 2005; Lebuhn et al., 1997). The genome of the Paenibacillus sp. isolated from

Lake Erie contain genes for the synthesis of thiazole/oxazole-modified microcins

(TOMMs).

4.7 Quorum Sensing and Signaling

Quorum sensing, or cell-to-cell communication in bacteria, relies on the production of signaling molecules known as autoinducers (Federle, 2009). Autoinducer 2

(AI-2) is an autoinducer produced by many different bacterial species. The Enterobacter genome contains genes for AI-2 transport and processing. This lsrACDBFGE operon has also been described in Salmonella enterica and Escherichia coli (Vendeville et al., 2005;

Xavier and Bassler, 2005), both members of the Enterobacteriaceae family with

Enterobacter. In these organisms, the lsrACDB genes encode AI-2 transporter components while the rest of the genes in the operon are needed from processing AI-2 once it is internalized (Xavier and Bassler, 2005). Tryptophan and the auxin IAA can also serve as signaling molecules (Amin et al., 2015). As mentioned previously, all five of the sequenced genomes contain genes for auxin and tryptophan synthesis.

4.8 Other genes/Vitamin Production

Many algae require vitamin B12 for growth yet are unable to produce it and must rely on exogenous B12 (Croft et al., 2005). It has been shown that microalgae can obtain vitamin B12 directly from bacterial interactions (Croft et al., 2005). The Paenibacillus,

Acidovorax, and Deinococcus genomes all contain genes for cobalamin (vitamin B12) synthesis (MTR, cobY, cobU, cobQ, bluB). All five genomes also contain genes for the synthesis of pyridoxin (vitamin B6) (PHGDH, serC, pdxA, pdxK).

5. Conclusions

The whole genomes of five bacteria isolated from a Lake Erie Microcystis bloom were sequenced. These genomes were analyzed for genes that could be important in potential interactions with Microcystis. Genes involved in vitamin production, iron sequestration, utilization of certain nitrogen sources, and the use or carbohydrates that are part of Microcystis EPS all suggest interactions with Microcystis are likely. The sequencing of these genomes opens up further possibility of understanding the mechanism of these interactions through future studies employing techniques such as transcriptomics.

Works Cited

Amin SA, Hmelo LR, van Tol HM, Durham BP, Carlson LT, Heal KR, Morales RL, Bethiaume CT, Parker MS, Djunaedi B, Ingalls AE, Parsek MR, Moran MA, Armbrust EV. 2015. Interaction and signaling between a cosmopolitan phytoplankton and associated bacteria. Nature. 522:98-101. Anderson DM, Glibert PM, Burkholder JM. 2002. Harmful algal blooms and eutrophication: nutrient sources, composition, and consequences. Estuaries 25(4):704- 726. Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, Formsma K, Gerdes S, Glass EM, Kubal M, Meyer F, Olsen GJ, Olson R, Osterman AL, Overbeek RA, McNeil LK, Paarmann D, Paczian T, Parrello B, Pusch GD, Reich C, Stevens R, Vasseiva O, Volstein V, Wilke A, Zagnitko O. 2008. The RAST server: Rapid annotations using subsystems technology. BMC Genomics. 9:75. Backer LC, Baptiste DM, LePrell R, Bolton B. 2015. Cyanobacteria and algae blooms: review of health and environmental data from the Harmful Algal Bloom-Related Illness Surveillance System (HABISS) 2007-2011. Toxins 7:1048-1064. Beatty PH and Jensen SE. 2002. Paenibacillus polymyxa produces fusaricidin-type antifungal antibiotics active against Leptosphaeria maculans, the causative agent of blackleg disease of canola. Can. J. Microbiol. 48(2):159-169. Bell W, Mitchell R. 1972. Chemotactic and growth responses of marine bacteria to algal extracellular products. Biol. Bull. 143(2):265-277. Boer JL, Mulrooney SB, Hausinger RP. 2013. Nickel-dependent metalloenzymes. Arch. Biochem. Biophys. 0:142-152. Braun V. 1995. Energy-coupled transport and signal transduction through the gram- negative outer membrane via TonB-ExbB-ExbD-dependent receptor proteins. FEMS Microbiol. Rev. 16:295-307. Brettin T, Davis JJ, Disz T, Edwards RA, Gerdes S, Olsen GJ, Olson R, Overbeek R, Parrello B, Pusch GD, Shukla M, Thomason JA 3rd, Stevens R, Volstein V, Wattam AR, Xia F. 2015. RASTtk: A modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes. 9:75. doi: 10.1186/1471-2164-9-75. Carrano CJ and Raymond KN. 1979. Ferric ion sequestering agents. 2. Kinetics and mechanism of iron removal from transferrin by enterobactin and synthetic tricatechols. J. Am. Chem. Soc. 101(18):5401-5404. Carver T, Thomson N, Bleasby A, Berriman M, Parkhill J. 2009. DNAPlotter: circular and linear interactive genome visualization. Bioinformatics. 25(1):119-120.

Cendrowski S, MacArthur E, Hanna P. 2004. Bacillus anthracis requires siderophore biosynthesis for growth in macrophages and mouse virulence. Mol. Microbiol. 51(2):407- 17. Chaffin JD, Bridgeman TB. 2014. Organic and inorganic nitrogen utilization by nitrogen- stressed cyanobacteria during bloom conditions. J. Appl. Phycol. 26(1):299-309. Cho BC, Park MG, Shim JH, Azam F. 1996. Significance of bacteria in urea dynamics in coastal surface waters. Mar. Ecol. Prog. Ser. 142:19-26. Croft MT, Lawrence AD, Rauz-Deery E, Warren MJ, Smith AG. 2005. Algae acquire vitamin B12 through a symbiotic relationship with bacteria. Nature. 438:90-93. De Figueiredo DR, Azeiteiro UM, Esteves SM, Goncalves JM, Pereira MJ. 2004. Microcystin-producing blooms – a serious global public health issue. Ecotoxicol. Environ.Saf. 59: 151-163. Dixon NE, Gazzola C, Blakeley RL, Zerner B. 1975. Jack Bean Urease (EC 3.5.1.5). A metalloenzyme. A simple biological role for nickel?. JACS. 97(14):4131-4133. Federle MJ. 2009. Autoinducer-2-based chemical communication in bacteria: Complexities of interspecies signaling. Contrib. Microbiol. 16:18-32. Fischbach MA, Lin H, Liu DR, Walsh CT. 2006. How pathogenic bacteria evade mammalian sabotage in the battle for iron. Nat. Chem. Biol. 2(3):132-138. Forni C, Telo FR, Caiola MG. 1997. Comparative analysis of the polysaccharides produced by different species of Microcystis (Chroococcales, Cyanophyta). Phycologia. 36(3):181-185. Giroldo D, Ortolando PIC, Vieira AAH. 2007. Bacteria-algae association in batch cultures of phytoplankton from a tropical reservoir: the significance of algal carbohydrates. Freshwater Biol. 52(7):1281-1289. Hertlein G, Muller S, Garcia-Gonzalez, Poppinga L, Sussmuth RD, Genersch E. 2014. Production of the catechol type siderophore bacillibactin by the honey bee pathogen Paenibacillus larvae. PLoS One 61(5):390-395. Imai A, Fukushima T, Matsushige K. 1999. Effects of iron limitation and aquatic humic substances on the growth of Microcystis aeruginosa. Can. J. Fish. Aquat. Sci. 56(10):1929-1937. Jones GJ, Bourne DG, Blakeley RL, Doelle H. 1994. Degradation of the cyanobacterial hepatotoxin microcystin by aquatic bacteria. Nat. Toxins, 2: 228-235. doi:10.1002/nt.2620020412. Kato H, Imanishi SY, Tsuji K, Harada K. 2007. Microbial degradation of cyanobacterial cyclic peptides. Water Res. 41(8):1754-1762.

Lau WWY, Armbrust EV. 2006. Detection of glycolate oxidase gene glcD diversity among cultured and environmental marine bacteria. Environ. Microbiol. 8(10):1688- 1702. Lebuhn M, Heulin T, Hartmann A. 1997. Production of auxin and other indolic and phenolic compounds by Paenibacillus polymyxa strains isolated from different proximity to plant roots. FEMS Microbiol. Ecol. 22(4):325-334. Nakagawa M, Takamura Y, Yagi O. 1987. Isolation and characterization of the slime from cyanobacterium, Microcystis aeruginosa K-3A. Agric. Biol. Chem. 51(2):329-337. Ollinger J, Song KB, Antelmann H, Hecker M, Helmann JD. 2006. Role of the Fur regulon in iron transport in Bacillus subtilis. J. Bacteriol. 188(10):3664-73. Overbeek R, Olson R, Pusch GD, Olsen GJ, Davis JJ, Disz T, Edwards RA, Gerdes S, Parrello B, Shukla M, Vonstein V, Wattam AR, Xia F, Stevens R. 2014. The SEED and the rapid annotation of microbial genomes using subsystems technology (RAST). Nucleic Acids Res. 42: doi: 10.1093/nar/gkt1226. Paver SF, Kent AD. 2010. Temporal patterns in glycolate utilizing bacterial community composition correlate with phytoplankton population dynamics in humic lakes. Microb. Ecol. 60(2):406-418). Selim S, Negrel J, Govaerts C, Gianinazzi S, van Tuinen D. 2005. Isolation and partial characterization ofantagonistic peptides produced by Paenibacillus sp. Strain B2 isolated from the sorghum mycorrhizosphere. Environ. Microbiol. 71(11):6501-6507. Shen H, Niu Y, Xie P, Tao M, Yang X. 2011. Morphological and physiological changes in Microcystis aeruginosa as a result of interactions with heterotrophic bacteria. Freshwater Biol. 56(6):1065-1080 Tatusova T, DiCuccio M, Badretdin A, Chetvernin V, Nawrocki EP, Zaslavsky L, Lomsadze A, Pruitt KD, Borodovsky M, Ostell J. 2016. NCBI prokaryotic genome annotation pipeline. Nucleic Acids Res. 44(14):6614-6624. Van Tiel-Menkveld GJ, Mentjox-Vervuut JM, Oudega B, de Graaf FK. 1982. Siderophore production by Enterobacter cloacae and a common receptor protein for the uptake of aerobactin and cloacin DF13. 150(2):490-7. Vendeville A, Winzer K, Heurlier K, Tang CM, Hardie KR. 2005. Making ‘sense’ of metabolism: autoinducer-2 LUXS and pathogenic bacteria. Nat. Rev. Microbiol. 3:383- 396. Wang W, Shen H, Shi P, Chen J, Xie P., 2016. Experimental evidence for the role of heterotrophic bacteria in the formation of Microcystis colonies. J. Appl. Phycol. 28(2):1111-1123. Wen N and Brooker MH. 1994. Rate constants for cyanate hydrolysis to urea: A Raman study. Can. J. Chem. 72(4):1099-1106.

Wildermuth MC, Dewdney J, Wu G, Ausubel FM. 2001. Isochorismate synthase is required to synthesize salicylic acid for plant defence. Nature. 414(6863):562-5. Williams PH. 1979. Novel iron uptake system specified by ColV plasmids: an important component in the virulence of invasive strains of Escherichia coli. Infect. Immun. 26:925-932. Williams PH and Warner PJ. 1980. ColV plasmid-mediated, colicin V-independent iron uptake system of invasive strains of Escherichia coli. Infect. Immun. 29:411-416. Wilson MK, Abergel RJ, Raymond KN, Arceneaux JE, Beyers BR. 2006. Siderophores of bacillus anthracis, Bacillus cereus, and Bacillus thuringiensis. Biochem, Biophys. Res. Commun. 348(1):320-5). Wolfson JS, Hooper DC. 1989. Fluoroquinolone antimicrobial agents. Clin. Microbiol. Rev. 2(4):378-424. Xavier KB, Bassler BL. 2005. Regulation of uptake and processing of the quorum- sensing autoinducer AI-2 in Escherichia coli. J. Bacteriol. 187:238-248. Xie M, Ren M, Yang C, Yi H, Li Z, Li T, Zhao J. 2016. Metagenomic analysis reveals symbiotic relationship among bacteria in Microcystis-dominated community. Front. Microbiol. 7: doi: 10.3389/fmicb.2016.00056. Xing W, Huang W, Li D, Liu Y. 2007. Effects of iron on growth, pigment content, photosystem II efficiency, and siderophores production of Microcystis aeruginosa and Microcystis wesenbergii. Curr. Microbiol. 55(2):94-98. Yokoo S, Inoue S, Suzuki N, Amakawa N, Matsui H, Nakagami H, Takahashi A, Arai R, Katou A. 2018. Comparative analysis of plant isochorismate synthases reveals structural mechanisms underlying their distinct biochemical properties. Biosci. Rep. 38(2): BSR20171457.

Chapter IV Conclusions

In this thesis, heterotrophic bacteria were isolated from a 2017 Microcystis bloom in the western basin of Lake Erie. The culturable microbiome of this bloom was characterized and then compared to the unculturable population of the bloom via metagenomics. It was surprising that no species from the phylum Actinobacteria were isolated, as members of this phylum were the most common phylum of heterotrophic bacteria in the metagenomic sequences from most of the stations. The isolates were also grown directly with axenic Microcystis to determine any impacts on the growth and physiology of the Microcystis cells. The co-culturing of bacteria with axenic Microcystis suggested that Microcystis does not necessarily need direct cell contact with heterotrophic bacteria to benefit from protective impacts of the bacteria. The protective impacts of the filtrate of some of the isolates suggests that bacterial interactions in a natural bloom community could protect Microcystis cells in harsh conditions such as higher salt concentrations. The mechanisms of the protection of this filtrate on

Microcystis in these conditions was not explored in this thesis.

The same five isolates that were grown in co-culture with Microcystis were chosen for whole genome sequencing with PacBio. The genomes of these five isolates were analyzed and compared to identify any genes that could be used in potential interactions with Microcystis. The genomic data from the bacterial isolates provided insight on the potential ways the isolates could benefit Microcystis and interact with the phytoplankton, as well as ways the phytoplankton may help the bacteria.

While the co-culturing of the bacteria with Microcystis and whole genome sequencing of the isolates provided insights into potential interactions, further studies are needed to determine the mechanisms of communication and interactions between the

81 bacteria and Microcystis. And while co-culture suggested protective impacts of the filtrate in high salt concentrations, the mechanisms of this protection were not explored and need to be further examined. The protective impacts could be as simple as the utilization of salt by the bacteria, therefore preventing osmotic stress of the Microcystis cells when exposed to the filtrate.