Occurrence of a Cyanobacterial Neurotoxin, Anatoxin-A, in New

OCCURRENCE OF THE CYANOBACTERIAL NEUROTOXIN, ANATOXIN-A, IN

NEW YORK STATE WATERS

Xingye Yang

A dissertation submitted in partial fulfillment of the requirements for the Doctor of Philosophy Degree

State University of New York College of Environmental Science and Forestry Syracuse, New York

January 2007

Approved:

Faculty of Chemistry

------Gregory L. Boyer, Major Professor William Shields, Chairperson, Examination Committee

------John P. Hassett, Faculty Chair Dudley J. Raynal, Dean, Instruction and Graduate Studies UMI Number: 3290535

UMI Microform 3290535 Copyright 2008 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, MI 48106-1346 Acknowledgements

I would like to express my sincerest gratitude to Dr. Gregory L. Boyer, my major

professor and academic advisor for his guidance, support, and assistance over the past

years. He has provided me with invaluable knowledge and skills.

I thank Dr. David J. Kieber for his advice and instrument support. I acknowledge

Dr. John P. Hassett for his support on both my research and my career. I appreciate Dr.

Francis X. Webster for his help on chemical characterization. I thank Dr. James P. Nakas

for advice on my career development. Thanks are also due to Dr. William Shields for

serving as chairman of this examination committee. I appreciate critical reviews and

comments on the thesis from all the examiners on this committee.

I would like to thank Dr. Israel Cabasso and Dr. Paul Caluwe for their help and

advice on my study. I would also like to thank past and present members of the Boyer

lab. To Michael F. Satchwell for his help whenever I need any; to Steve Ragonese,

Amber Hotto, Elizabeth Konopko, Dr. Guozhang Zou, John Usher, Karen Haward, and

Nick Smith. Special thanks to Dave Kiemle for his help on LCMS and NMR analysis

and to Emily White and other members at Dr. Kieber’s lab for help on photochemistry, and to Dr. Hui Zhao, Suoding Li, Xinwei Wang, Ju Feng, and all my colleagues and friends. They make my life and study at SUNY-ESF memorable.

This study was supported by the National Oceanic and Atmospheric Agency

Coastal Oceans Program through their MERHAB-LGL project # NA160 P2788 and New

York Sea Grant Award NA16R61645.

This thesis is dedicated to my parents, whose endless love and support always

motivate me. Without them, none of my accomplishments could have come true.

i Abstract

YANG, XINGYE. Occurrence and distribution of a cyanobacterial neurotoxin, anatoxin- a, in New York State waters. Typed and bound thesis, 232 pages, 19 Tables, 63 Figures,

2006.

Cyanobacterial blooms are a serious environmental and health problem throughout the world. In New York State, cyanobacterial toxins such as the hepatotoxic microcystins and the neurotoxic anatoxin-a have been detected in various lakes including the lower Great Lakes. Compared to microcystins, the occurrence and distribution of anatoxin-a in New York State is poorly studied. This study provided the first systematic monitoring for anatoxin-a in New York State waters.

Anatoxin-a is a bicyclic alkaloid (MW 165) which has potent nicotinic agonistic activity and has caused several animal fatalities around the world and in New York State.

It may be produced by several genera of cyanobacteria including Anabaena,

Aphanizomenon, Planktothrix and Microcystis. Anatoxin-a is mainly retained within cells but can be released into the surrounding water column during cell senescence and lysis. To better understand its behavior in natural waters, the stability of anatoxin-a was examined under natural and artificial light and different pH conditions. Under the influence of light and/or pH, anatoxin-a readily degraded with a first order or a pseudo- first order kinetics. Anatoxin-a has a half-life of 4-10 hours under natural and artificial solar radiation, and several days up to several months in the absence of light. Several degradation products were formed during photolysis of anatoxin-a. The primary degradation product of anatoxin-a under artificial light was identified by NMR as

ii tricycloanatoxin-a and the degradation was proposed to proceed via an intramolecular

rearrangement process.

In New York State, anatoxin-a was periodically detected at concentrations above

an cautionary level of 0.1 μg L-1 in the western basin of Lake Erie, the embayments along

the southern shoreline of Lake Ontario, nearshore sites on Lake Champlain, and in other

smaller inland lakes such as Onondaga Lake, Lake Neatahwanta and Lake Agawam.

Anatoxin-a exceeded 1 μg L-1 in Lake Ontario, Lake Champlain, Onondaga Lake and

Lake Agawam. In these lakes, the locations where anatoxin-a was observed are primarily

shallower waters and are used extensively for recreational purposes. Exposure of humans

or pets to cyanobacterial blooms in these areas during summer and fall is possible. The

detection of anatoxin-a in waters in proximity to human habitats at concentrations higher than the cautionary level indicates the health risk associated with anatoxin-a is real.

Author Name Xingye Yang

Candidate for the degree of Doctor of Philosophy Date ______

Major Professor Gregory L. Boyer

Faculty of Chemistry

State University of New York College of Environmental Science and Forestry,

Syracuse, New York

Signature of Major Professor ______

iii Table of Contents

ACKNOWLEDGEMENTS i

ABSTRACT ii

TABLE OF CONTENTS iv

LIST OF FIGURES vii

LIST OF TABLES xii

CHAPTER 1. INTRODUCTION AND STATEMENT OF RESEARCH

HYPOTHESIS 1

CHAPTER 2. LITERATURE REVIEW 7 2.1. Cyanobacteria and cyanobacterial blooms 7 2.2. Cyanobacterial toxins 16 2.2.1. Introduction to cyanobacterial toxins 16 2.2.2. The neurotoxic cyanobacterial toxin, anatoxin-a 27 2.2.3. Detection of anatoxin-a in natural waters 38 2.3. The New York State lakes 39 2.4. References 49 CHAPTER 3. MATERIALS AND METHODS 71 3.1. Preparation of anatoxin-a standards 71 3.1.1. Preparation of anatoxin-a standards for toxin analysis 71 3.1.2. Synthesis of two non-toxic degradation products of anatoxin-a 72 3.2. Analysis of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a 78 3.3. Sampling in New York State lakes and processing of water samples for toxin analysis 81 3.3.1. Sampling in the lower Great Lakes 81 3.3.2. Sampling in Lake Champlain 84 3.3.3. Sampling in other New York State lakes 84 3.3.4. Extraction of anatoxin-a from water samples 87 3.4. Stability studies of anatoxin-a 88

iv 3.4.1 Effects of temperature and pH on anatoxin-a in the dark 88 3.4.2. Effects of pH, natural and simulated solar radiation on the stability of anatoxin-a 88 3.4.3. Photolysis of anatoxin-a in natural lake water 94 3.4.4. Characterization of degradation product(s) of anatoxin-a 95 3.5. References 95 CHAPTER 4. EXPERIMENTAL RESULTS 97

4.1. Stability of anatoxin-a under different environmental factors 97 4.1.1. Introduction 97 4.1.2. Materials and methods 98 4.1.3. Results 101 4.1.4. Discussion 121 4.1.5. References 127 4.2. Occurrence and distribution of anatoxin-a in the lower Great Lakes 129 4.2.1. Introduction 129 4.2.2. Materials and methods 130 4.2.3. Results 132 4.2.4. Discussion 150 4.2.5. Acknowledgement 155 4.2.6. References 155 4.3. Identification, occurrence and spatial distribution of anatoxin-a in Lake Champlain, New York 160 4.3.1. Introduction 160 4.3.2. Materials and methods 163 4.3.3. Results 165 4.3.4. Discussion 172 4.3.5. Acknowledgments 178 4.3.6. References 178 4.4. Occurrence and distribution of anatoxin-a in other New York State lakes 183

v 4.4.1. Introduction 183 4.4.2. Materials and methods 185 4.4.3. Results 187 4.4.4. Discussion 205 4.4.5. Acknowledgements 209 4.4.6. References 210 CHAPTER 5. DISCUSSION 214 APPENDIX 219 VITAE 232

vi List of Figures

Figure 1.1. Chemical structure of anatoxin-a (protonated form). 2 Figure 2.1. Three common bloom-forming cyanobacteria. Anabaena flos- aquae, Microcystis aeruginosa and Oscillatoria sp. 14 Figure 2.2. Warning flyer released by Vermont Department of Health regarding blue-green algal blooms in Lake Champlain and the potential health risks 17 Figure 2.3. Structures of microcystins 19 Figure 2.4. Structures of the alkaloid cyanobacterial toxins 20 Figure 2.5. Conformation of anatoxin-a and structures of the two known non- toxic degradation products, epoxyanatoxin-a and dihydroanatoxin-a 30 Figure 2.6. Mechanism of anatoxin-a poisoning 31 Figure 2.7. Sections of the 125 MHz 13C NMR spectrum of anatoxin-a 34 Figure 2.8. A plausible biosynthetic scheme for the neurotoxic anatoxin-a and its precursor 11-carboxy anatoxin-a 35 Figure 2.9. New York State water bodies 42 Figure 2.10. The Great Lakes Basin 43 Figure 2.11. A view of the lower Great Lakes from SeaWiFS satellite showing an algal bloom in the western basin of Lake Erie and a lakewide algal bloom in Lake Ontario 46 Figure 2.12. Map of Lake Champlain Basin 47 Figure 3.1. Anatoxin-a standard curves by HPLC-FD 73 Figure 3.2. Typical standard curves for anatoxin-a analysis by LCMS 74 Figure 3.3. Sample HPLC-FD chromatograms of anatoxin-a (ATX) and its degradation products 75 Figure 3.4. Typical sample LCMS chromatograms and mass spectra for anatoxin-a and l-phenylalanine 76 Figure 3.5. Typical sample LCMS chromatographs of the two non-toxic degradation products of anatoxin-a: epoxyanatoxin-a (epoxyATX) (top) and dihydroanatoxin-a (dihydroATX) 77

vii Figure 3.6. Sampling sites in the lower Great Lakes between 2001 and 2005 83 Figure 3.7. Sampling locations in Lake Champlain between 2000 and 2005 86 Figure 3.8. Schematic illustration of experimental design for evaluating stability of anatoxin-a under natural and simulated solar radiation 90 Figure 3.9. Comparison of the transmittance of both quartz and optical glass tubes used in the experiments 91 Figure 3.10. Spectrum of the anatoxin-a fumarate standard 92 Figure 4.1. The effect of pH on the degradation of anatoxin-a over time. 104 Figure 4.2. Loss of anatoxin-a in deionized water over time under natural sunlight and simulated solar radiation in optical glass tube and quartz tube 105 Figure 4.3. Loss of anatoxin-a at different pH under laboratory simulated solar radiation in short term (12 hr) experiments as a function of exposure time and light dose 107 Figure 4.4. The decay of anatoxin-a in natural lake water (pH 9.6) under natural sunlight and artificial solar radiation versus exposure time and cumulative light doses 108 Figure 4.5. Total ion chromatograph of anatoxin-a in aqueous solution before and after 12-hour exposure to simulated UV radiation 109 Figure 4.6. Mass spectra of the detected four degradation products 110 Figure 4.7. Formation of the degradation products of anatoxin-a under artificial solar radiation. as a function of time in deionized water and pH 10 buffer solution 112 Figure 4.8. Formation of the degradation products of anatoxin-a as a function of cumulative light dose in deionized water and pH 10 buffer solution under artificial solar radiation. 113 Figure 4.9. 13C NMR of the degradation product of anatoxin-a, unknown 3, in

H2O with 5% D2O as solvent 117 Figure 4.10a. 1H-COSY NMR of the primary degradation product of anatoxin-a 118 Figure 4.10b. 1H-13C correlation NMR (HMBC) of the primary degradation product of anatoxin-a 119

viii Figure 4.11. Possible degradation product of anatoxin-a via [2+2] intermolecular photocycloaddition 123 Figure 4.12. Proposed degradation pathway of anatoxin-a under experimental conditions 124 Figure 4.13. Occurrence of anatoxin-a in Lake Erie between 2002 and 2005 133 Figure 4.14. Percent occurrence and the maximum concentration of anatoxin-a in Lake Erie over the 4-year period, and the monthly distribution and maximum concentration of anatoxin-a in Lake Erie in summer 2005 135 Figure 4.15. Comparison of anatoxin-a concentration at Station 1163 Sandusky Bay and at a sampling site outside Sandusky Bay 137 Figure 4.16. Occurrence of anatoxin-a in a transect down the Maumee River into the Maumee Bay in June of 2004 138 Figure 4.17. The HPLC-FD and LCMS chromatograph of an anatoxin-a- containing water sample collected from the Maumee River on July 14th, 2004 139 Figure 4.18. Relationship bwteeen the concentration of anatoxin-a and chlorophyll a in samples collected from all basins of Lake Erie between 2002 and 2005 140 Figure 4.19. Relationship between the concentration of anatoxin-a and chlorophyll a in samples collected from the western basin of Lake Erie 141 Figure 4.20. Occurrence and distribution of anatoxin-a in Lake Ontario between 2001 and 2005 143 Figure 4.21. The percentage occurrence and the maximum concentration of anatoxin-a in Lake Ontario between 2001 and 2005 146 Figure 4.22. Percent occurrence of anatoxin-a in the Long Pond, an embayment of Lake Ontario during summer 2003 and 2004 147 Figure 4.23. Relationship between the concentration of anatoxin-a and chlorophyll a in Lake Ontario 148 Figure 4.24. Variation in concentrations of anatoxin-a and chlorophyll a in

ix Long Pond, Lake Ontario in 2003 and 2004 149 Figure 4.25. Sampling sites and the occurrence of anatoxin-a in Lake Champlain 164 Figure 4.26. The HPLC-FD chromatograph of the water sample containing anatoxin-a collected during a cyanobacterial bloom that caused a dog toxicosis event at Lake Champlain on August 3rd 2000 167 Figure 4.27. LCMS screening for anatoxin-a in a Lake Champlain sample 169 Figure 4.28. Percent occurrence of anatoxin-a and its maximum concentrations in Lake Champlain between 2000 and 2005 173 Figure 4.29. Chlorophyll a concentration and production of anatoxin-a in Lake Champlain 174 Figure 4.30. Occurrence and maximum concentrations of anatoxin-a at the two sampling sites in Onondaga Lake in 2002 and 2003 190 Figure 4.31a. The HPLC-FD chromatograph of an anatoxin-a-containing water sample collected from the southern basin of Onondaga Lake on September 8th, 2003 191 Figure 4.31b. LCMS chromatograph of an anatoxin-a-containing water sample collected from the southern basin of Onondaga Lake on September 8th, 2003 192 Figure 4.32. Relationship between concentration of anatoxin-a and the total cyanobacterial biomass as estimated by the pigment phycocyanin in Onondaga Lake 194 Figure 4.33. Changes in concentrations of anatoxin-a and phycocyanin at the two sampling sites in Onondaga Lake in summer and fall 2002 and 2003 195 Figure 4.34. Concentration of anatoxin-a confirmed by LCMS versus concentration of chlorophyll a in Oneida Lake 199 Figure 4.35. Variation in anatoxin-a concentration over time at the three sampling sites on Lake Neatahwanta, New York in summer 2004 and 2005 201 Figure 4.36. Variation in chlorophyll a concentration at the three sampling sites

x on Lake Neatahwanta in summer 2004 202 Figure 4.37. Distribution of anatoxin-a in Lake Agawam 204 Appendix 229 Appendix 1. Discrete sampling stations on Lake Erie and their occurrence of anatoxin-a between 2002 and 2005 219 Appendix 2. Discrete sampling stations on Lake Ontario and their occurrence of anatoxin-a between 2002 and 2005 223 Appendix 3. Discrete sampling stations on Lake Champlain and their occurrence of anatoxin-a between 2000 and 2005 228 Appendix 4. New York State lakes monitored for anatoxin-a between 2000 and 2005 230

xi List of Tables

Table 2.1. Global occurrence of toxic cyanobacterial blooms 10 Table 2.2. Cyanobacterial toxins and their acute toxicity 22 Table 2.3. Anatoxin concentrations detected in various water bodies 37 Table 3.1. Experimental conditions for the electrospray ionization mass spectrometry (ESI-MS) analysis of anatoxin-a and its degradation products 82 Table 3.2. Locations of the sampling stations on Onondaga Lake, Oneida Lake and Lake Neatahwanta 85 Table 3.3. Physical, chemical and biological parameters of raw lake water collected from Lake Neatahwanta for assessing stability of anatoxin-a 93 Table 4.1. Degradation of anatoxin-a under natural and artificial light 103 Table 4.2. Chemical shifts of carbons and protons in anatoxin-a 114 Table 4.3. Chemical shifts of carbons and protons in the degradation product 3, tricycloanatoxin-a 116 Table 4.4. The occurrence of anatoxin-a in Lake Erie between 2002 and 2005 134 Table 4.5. Summery of occurrence of anatoxin-a in Lake Ontario between 2000 and 2005 145 Table 4.6. Concentrations of cyanobacterial toxins in four water samples collected during the year 2000 dog toxicosis event near Whallon Bay, Lake Champlain 168 Table 4.7. Occurrence of anatoxin-a in Lake Champlain during a 6-year monitoring period by sub-basin 171 Table 4.8. The Occurrence of anatoxin-a in Onondaga Lake between 2000 and 2003 188 Table 4.9. Occurrence of anatoxin-a in Oneida Lake from 2000 to 2005 198 Table 4.10. Occurrence of anatoxin-a in Lake Neatahwanta during summer 2004 and 2005 200 Table 4.11. Occurrence of anatoxin-a in Lake Agawam, Long Island, New York 203

xii Chapter 1. Introduction and Statement of Research Hypothesis

INTRODUCTION

Anatoxin-a, 2-acetyl-9-azabicyclo[4,2,1]non-2-ene (Figure 1.1), is one of the

neurotoxins produced by cyanobacteria. It is a low molecular weight alkaloid (MW 165)

with strong nicotinic agonistic activity (Molloy et al. 1995; Carmichael 1997; Codd et al.

1997). This molecule was first isolated from a toxic cyanobacterial bloom dominated by a strain of Anabaena flos-aquae, and caused deaths of birds and animals (Carmichael et al. 1975; Edwards et al. 1992; Gugger et al. 2005). Anatoxin-a can also be produced by other cyanobacterial species including Anabaena planktonica, Planktothrix/Oscillatoria sp., Aphanizomenon sp. Cylindrospermum sp, and Microcystis aeruginosa. It has been linked to deaths of birds, dogs, and other domestic stock throughout the world (Chorus and Bartram 1999; Sivonen 2000; Chorus et al. 2001; USEPA 2001; Yang et al. 2001;

WHO 2003, 2004). Although there are no documented human fatalities associated with anatoxin-a poisoning, the widespread occurrences of this neurotoxin in water bodies and its high toxicity have caused great concern over its potential health risks to human beings.

Anatoxin-a, like the hepatotoxic cyanobacterial toxin microcystins, is mostly held within

the cyanobacterial cells. The aging and lysis of algal cells and the subsequent

decomposition of algal blooms releases anatoxin-a into the surrounding water bodies

(Watanabe et al. 1992; Sivonen 1996). These dissolved toxins may pose a great threat to

birds, animals and human beings that are feeding or recreating in the water. To better

estimate the possible environmental and human health risks associated with anatoxin-a,

and to prevent future anatoxin-a poisonings, the occurrence, distribution and fate of this

neurotoxin in natural freshwater bodies needs to be examined.

Figure 1.1. Chemical structure of anatoxin-a (protonated form).

2 In New York State, cyanobacterial blooms have plagued many freshwater

systems, including the Great Lakes. This is mainly due to the continuous input of

nutrients and the consequent eutrophication in these water bodies (Shanley and Denner

1999; McGucken 2000). These lakes are important in both the economics of the region and the quality of the lifestyle. Although not all cyanobacterial blooms are toxic, illnesses in animals have been linked to toxic cyanobacteria and their toxins in New York State.

Monitoring for the occurrence of toxic cyanobacterial blooms and their toxins is necessary, especially for freshwater systems extensively used for drinking or recreational purposes.

HYPOTHESIS AND OBJECTIVES

It is well established that the cyanobacterial neurotoxin anatoxin-a, can lead to severe illness and even death when consumed by birds and animals, and that anatoxin-a

producing cyanobacterial blooms occur throughout the world (Chorus and Bartram

1999). However in contrast to the hepatotoxic cyanobacterial toxins such as microcystins

and cylindrospermopsin, human illness and fatalities associated with anatoxin-a have

only been suspected and later excluded. The actual risk anatoxin-a may pose to human

health is unknown. Anatoxin-a occurs in New York State waters less frequently and at

lower concentrations than microcystins, and consequently has received much less

attention. Considering the high toxicity of this neurotoxin, the diversity and complexity of anatoxin-a-producing organisms, and their cosmopolitan distribution in freshwater systems, I propose that the risk from anatoxin-a intoxication is real. Furthermore, I hypothesize the occurrence of anatoxin-a has been underestimated in New York State

3 waters. This toxin poses greater risks to environment, animal and human health than

currently recognized. To understand and evaluate this risk, I had the following objectives.

1. To determine the occurrence and distribution of anatoxin-a in New York State

waters, especially in the lower Great Lakes (Lake Erie and Lake Ontario) and

Lake Champlain during 2000 to 2005;

2. To examine the behavior and stability of anatoxin-a under changing

environmental factors such as light and pH in the laboratory;

3. using this information, to estimate the current risk status associated with

anatoxin-a in New York State lakes.

A general literature review is included in the second chapter of this thesis. This

introduces the global appearance of cyanobacterial blooms, their related environmental

and health problems, the different cyanobacterial toxins, and their modes of action with

emphasis on anatoxin-a. In chapter 3, the methods for sampling, sample treatment,

analysis by high performance liquid chromatography (HPLC) with fluorescence detection

and electrospray ionization mass spectrometry (LCMS), and a stability study of anatoxin-

a are presented. Chapter 4 contains the results of a study on the stability of anatoxin-a

under different conditions. The studies on occurrence of anatoxin-a in the lower Great

Lakes, Lake Champlain and some other New York State lakes are also presented in

chapter 4. The interpretation of these studies is given in Chapter 5.

4 REFERENCES

Carmichael, W. W. 1997. The cyanotoxins. Advances in Botanical Research 27:211-256. Carmichael, W. W., Briggs, D. F., and Gorham, P. R. 1975. Toxicology and pharmacological action of Anabaena flos-aquae toxin. Science 187:542-544. Chorus, I., and Bartram, J., eds. 1999. Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring, and Management. London: E & FN Spon (for the World Health Organization). 416 p. Chorus, I., Bumke-Vogt, C., Lindenschmidt, K.-E., and Jaime, E. 2001. Cyanobacterial neurotoxins. In Cyanotoxins - Occurrence, Causes, Consequences, edited by I. Chorus: Springer-Verlag. p37-45. Codd, G. A., Ward, C. J., and Bell, S. G. 1997. Cyanobacterial toxins: occurrence, modes of action, health effects and exposure routes. In Applied Toxicology: Approaches Through Basic Science, Archives of Toxicology Supplement. 19, edited by J. P. Seiler and E. Vilanove. Berlin: Springer. p399-410. Edwards, C., Beattie, K. A., Scrimgeour, C. M., and Codd, G. A. 1992. Identification of anatoxin-a in benthic cyanobacteria (blue-green algae) and in associated dog poisonings at Loch Insh, Scotland. Toxicon 30 (10):1165-1175. Gugger, M., Lenoir, S., Berger, C., Ledreux, A., Druart, J.-C., Humbert, J.-F., Guette, C., and Bernard, C. 2005. First report in a river in France of the benthic cyanobacterium Phormidium favosum producing anatoxin-a associated with dog neurotoxicosis. Toxicon 45 (7):919-928. McGucken, W. 2000. Lake Erie Rehabilitated. Controlling Cultural Eutrophication, 1960s-1990s. Edited by J. Stine and J. Tarr. 1st ed, Technology and the Environment. Akron, Ohio: The University of Akron Press, 318 p. Molloy, L., Wonnacott, S., Gallagher, T., Brough, P. A., and Livett, B. G. 1995. Anatoxin-a is a potent agonist of the nicotinic acetylcholine receptor of bovine adrenal chromaffin cells. European Journal of Pharmacology, Molecular Pharmacology Section 289 (3):447-453. Shanley, J. B., and Denner, J. C. 1999. The hydrology of the Lake Champlain basin. In Lake Champlain in Transition: from Research toward Restoration, edited by T. O.

5 Manley and P. L. Manley. Washington, DC: Americal Geophysical Union. p41- 67. Sivonen, K. 1996. Cyanobacterial toxins and toxin production. Phycologia 35 (6):12-24. ———. 2000. Freshwater cyanobacterial neurotoxins: Ecobiology, chemistry, and detection. Food Science and Technology (New York, NY, United States) 103:567- 581. USEPA. 2001. Creating a cyanotoxin target list for the unregulated contaminant monitoring rule. 17 p. Watanabe, M., Tsuji, K., Watanabe, Y., Harada, K., and Suzuki, M. 1992. Release of heptapeptide toxin (microcystin) during decomposition process of Microcystis aeruginosa. Natural Toxins 1:48-53. WHO. 2003. Guidelines for Safe Recreational Water Environments. Volume 1: Coastal and Fresh Waters. Geneva: World Health Organization, 219 p. ———. 2004. Guidelines for Drinking Water Quality. 3rd ed. Geneva: WHO, 515 p. Yang, X., Satchwell, M. F., and Boyer, G. L. 2001. The identification of anatoxin-a from a toxic blue-green algae bloom in Lake Champlain, USA. Abstract. In Fifth International Conference on Toxic Cyanobacteria, Moosa Lakes, Queensland, Australia. July 15-20. 1 p.

6 Chapter 2. Literature Review

2.1. CYANOBACTERIA AND CYANOBACTERIAL BLOOMS

Cyanobacteria, also known as blue-green algae, are the predominant group of

gram-negative oxygen-evolving photosynthetic prokaryotes on earth and play a crucial role in successional processes, global photosynthetic biomass production and nutrient cycling (Garcia-Pichel 1998). Although referred to as blue-green algae for their superficial resemblance to eukaryotic green algae, the cyanobacteria are eubacteria. They have a cell wall containing peptidoglycan and lack a discrete membrane-bound nucleus, internal membrane-bound organelles and the histone proteins associated with their nucleic acid (Fay and Van Baalen 1987). The term “algae” now merely refers to the fact that they are aquatic organisms capable of photosynthesis. Based on the traditional

International Code of Botanical Nomenclature for the class Cyanophyceae, about 150 genera and about 2000 species have been identified (Skulberg et al. 1993; van den Hoek et al. 1995). Current efforts are being made to categorize cyanobacteria based on the

International Code of Bacteria Nomenclature.

In addition to the photosynthetic pigment chlorophyll a, cyanobacteria also contain other characteristic water-soluble photosynthetic pigments called phycobilins.

Phycobilins give the group their blue-green coloration and is the origin of the name

“blue-green algae”. Not all of the blue-green algae or cyanobacteria are necessarily blue as some are red or pink from the pigment phycoerythrin. Cyanobacteria are primitive organisms which were involved in oxygenation of the primordial earth and their existence

can be dated back to 3.5 billion years ago. With a cosmopolitan distribution in nearly all

kinds of environments, cyanobacteria have been found in fresh-, brackish, and marine

7 waters, and terrestrial environments ranging from hot springs to Arctic and Antarctic

regions. They play a key role in the biogeochemical cycling of matter and in the

structure, maintenance, and biodiversity of microbial and higher organism communities

(Berman-Frank et al. 2003).

Because of the presence of chlorophyll a and accessory photosynthesis pigments

i.e., the phycobilins, phycocyanin and phycoerythrin, their unique ability to fix

atmospheric nitrogen, and their ability to compete effectively for low levels of nutrients

and trace metals (phosphorus, iron, etc.) (Paerl and Millie 1996; Ferber et al. 2004),

cyanobacteria can undergo rapid growth under favorable environmental conditions. In

contrast to other planktonic species, many planktonic cyanobacteria possess specialized

intracellular gas vesicles (Reynolds et al. 1987) to regulate their buoyancy and thus

actively exploit the optimal position in the water column for gathering light needed for

growth. As a result, cyanobacterial species often dominate other phytoplankton during

the summer and fall season. Especially in calm eutrophic waters, in warmer climates

(water temperature of 15-30oC) with ample sunlight, the proliferation and accumulation

of cyanobacteria may become very dense. This is referred to as a cyanobacterial bloom

(Skulberg et al. 1984; Bumke-Vogt et al. 1999; Downing et al. 2001). These blooms can

either form as a dense distribution throughout the water column or as a surface scum, or

as a mat/biofilm for benthic cyanobacterial species (Codd et al. 2005).

Although water bodies can naturally become more productive and eventually

eutrophic, and thus prone to the formation of cyanobacterial blooms, this natural process

normally takes a long time to progress. In contrast, cultural or anthropogenic

eutrophication, i.e., excessive loading of nutrients into natural water bodies caused by

8 human activities such as agricultural runoff, detergents and municipal wastewater discharges, has led to the accelerated eutrophication of freshwater and marine environments and the subsequent frequently occurring cyanobacterial blooms (Codd and

Bell 1985; Moss 1988; Paerl and Millie 1996). The formation of cyanobacterial blooms is dependent on many other environmental factors, especially climatic conditions.

Surface scums or blooms of cyanobacteria can be quickly dispersed by wind mixing or dissipated by wave action in open water, yet in near shore areas, especially in shallow bays, algal blooms may persist for longer times. If conditions remain favorable, successive blooms may overlap and appear as one continuous bloom persisting for several months (Rapala and Sivonen 1998; Falconer 1999).

Cyanobacterial blooms occur globally (Table 2.1). In recent history, mass accumulations of cyanobacteria have been observed more frequently in freshwaters such as inland lakes or reservoirs and in seas and oceans such as the Pacific Ocean, Atlantic

Ocean and Indian Ocean. More importantly, these blooms have formed largely along the shorelines of inland freshwater bodies and coasts of North America, South America, Asia,

India and Europe, i.e., they have formed near human habitats (Sivonen et al. 1990;

Willen and Mattsson 1997; Bumke-Vogt et al. 1999; Falconer 2001; Mazur et al. 2003).

Consequently, the formation of blooms in near-shore waters can have a serious impact on both the environment and public health since the time of mass occurrence of cyanobacteria often coincides with the time when the demand for recreational water is highest. The environmental and health risks caused by the proximity of these cyanobacterial blooms and human inhabitants have elicited great public concern (Francis

1878; Ekman-Ekebom et al. 1992; Falconer 1999; Codd 2000; Carmichael 2001b).

Table 2.1 Global Occurrences of Toxic Cyanobacterial Blooms (Codd et al. 2005). Americas USA, Canada, Mexico, Brazil, Argentina, Venezuela

Europe Belgium, Czech Republic, Denmark, Estonia, Finland, France,

Germany, Greece, Hungary, Ireland, Italy, Latvia, Netherlands,

Norway, Poland, Portugal, Russia, Slovakia, Slovenia, Spain, Sweden,

Switzerland, Ukraine, United Kindom

Middle East Peoples’ Republic of China, Philippines, Saudi Arabia, Sri Lanka, and Asia South Korea, Thailand, Turkey, Vietnam, Japan

Australasia Australia, New Zealand

Africa Botswana, Egypt, Ethiopia, Kenya, Morocco, South Africa, Zimbabwe

10 During a cyanobacterial bloom, the cyanobacterial cells can produce a variety of natural

products, which may or may not be excreted into the surrounding water column. Some cyanobacterial blooms are harmless. Other blooms can lead to taste and odor problems

by production of the secondary metabolites 2-methylisoborneol and geosmin (Izaguirre et

al. 1982; Wnorowski 1992). Blooms can also produce a variety of extremely bioactive

secondary metabolites that may have detrimental effects on consuming organisms, cause

allergic responses such as swimmers itch, and lead to acute poisonings of both aquatic

and terrestrial animals and even human beings (Edwards et al. 1992; Kenefick et al. 1992;

Persson 1996; Sivonen 1996; Falconer 1998, 1999; Oberemm et al. 1999; Fromme et al.

2000; WHO 2003). These bioactive compounds are referred to as "cyanobacterial toxins"

or "cyanotoxins" (Carmichael 1994, 1997). One of the earliest records on a toxic effect

from a cyanobacterial bloom was in China 1000 years ago (Chorus and Bartram, 1999).

The first scientifically-documented toxic cyanobacterial bloom in modern history was in

Australia in 1878 (Francis 1878). Currently, at least 40 of the 150 known genera of

cyanobacteria have been reported to be toxicogenic (Skulberg et al. 1993; van den Hoek

et al. 1995). Systematic surveys concerning the occurrence of cyanobacterial blooms in

Europe, Australia and North America showed that about 30-80% of the natural blooms

were toxic and contained one or more types of toxic natural products (Berg et al. 1986;

Sivonen et al. 1989; Sivonen et al. 1990; Willen and Mattsson 1997; Falconer 1998). It

is recommended that all cyanobacterial blooms be considered potentially toxic until

tested otherwise (Crayton 1993). As climatic conditions can significantly affect the

development of these blooms, toxic events often cannot be predicted with accuracy in

advance. Most blooms in temperate zones happen in the late summer and can last for 2 to

11 4 months. In some tropical or subtropical areas, cyanobacterial blooms can be present all

year round. Blooms of cyanobacteria once present tend to reoccur in the same water

body, posing a risk of repeated exposure.

Production of cyanobacterial toxins occurs in genera such as Anabaena,

Anabaenopsis, Aphanizomenon, Cylindrospermopsis, Microcystis, Nodularia,

Planktothrix, Raphidiopsis and more (Figure 2.1) (Sivonen 1996; Carmichael 1997;

Wiegand and Pflugmacher 2005). In inland lakes and other freshwater systems, surface

scum production is particularly common with Anabaena, Aphanizomenon and

Microcystis. Blooms of Anabaenopsis and Planktothrix are planktonic in location where

Phormidium, Oscillatoria, and Lyngbya, are often involved in mat and benthic biofilm

formation. All have also been reported to be associated with animal intoxications (Mez

et al. 1997; Hamill 2001). Dense accumulations of cyanobacteria in eutrophic waters are

usually caused by planktonic species. In contrast, the benthic cyanobacterial species

grow in oligotrophic water where the low nutrients and water clarity allow penetration of

sunlight to the bottom to support photosynthesis of the benthic cyanobacteria. During

sunny days, the high photosynthetic rate and the produced oxygen may drive the benthic

algal mats to the surface of the water. Mats of toxic benthic cyanobacteria have washed

on shore and caused deaths of dogs and cattle (Edwards et al. 1992; Mez et al. 1997; Mez

et al. 1998). These mats or biofilms may not be recognized as potentially toxic

cyanobacterial blooms and the environmental and health risks associated with these

benthic cyanobacteria may be underestimated. Efforts to establish the occurrence,

distribution and frequency of these toxic cyanobacterial blooms, their toxins, and their impacts on wild animals, livestock and humans are underway (Carmichael et al. 1975;

12 Hawkins et al. 1985; Wang and Zhu 1995; Codd et al. 1997; Kuiper-Goodman et al.

1999; Oberemm et al. 1999; Codd 2001; Rao et al. 2002; Reinikainen et al. 2002; Wolf and Frank 2002; Oberholster et al. 2004; MacPhail and Jarema 2005).

Toxicoses associated with cyanobacterial blooms have included intoxication of

amphibians, fish, birds, wild and domestic mammals by drinking or bathing in contaminated water, or by consuming other aquatic organisms that feed on toxic cyanobacteria and may contain and/or accumulate cyanobacterial toxins (Codd et al.

1989; Carmichael 1992; Codd 1995; Negri et al. 1995; Carmichael 1997; Codd et al.

1997; Chorus and Bartram 1999; Mohamed 2001; Engstrom-Ost et al. 2002; Sipia et al.

2002). The toxic effects on human beings range from mild to fatal (Jackson et al. 1984;

Negri et al. 1995; Matsunaga et al. 1999). Disease due to cyanobacterial toxins varies

according to the type of toxin and the type of water or water-related exposure (drinking,

skin contact, etc.). A range of adverse health effects resulting from contact with and/or

ingestion of cyanobacterial cells and toxins include skin irritations, allergic responses,

mucosa blistering, muscular and joint pains, gastroenteritis, pulmonary consolidation,

liver and kidney damage and a range of neurological effects (Carmichael 2001b, 2001a;

Haider et al. 2003). The most frequent and serious health effects for human beings are

caused by drinking water containing the toxins or by ingestion of toxins during

recreational water contact. Exposure to toxins through consumption of food or food

supplements contaminated with cyanobacterial toxins may also be important (Draisci et

al. 2001a; Draisci et al. 2001b). Long-term studies of the chronic toxicity of

cyanobacterial toxins are in progress in China and suggest that drinking of untreated

Figure 2.1. Three common bloom-forming cyanobacteria. Anabaena flos-aquae,

Microcystis aeruginosa and Oscillatoria sp. Pictures courtesy of UTEX Culture

Collection of Algae.

14 surface water containing cyanobacterial toxins is associated with an increased incidence of primary liver cancer (Yu 1989, 1995; Falconer and Humpage 1996). In 1996, a toxic cyanobacterial bloom occurred in a water supply in Caruaru, Brazil. Although extensive pretreatments with sand, carbon, resin and microfiltration were applied prior to the usage of haemodialysis water, over one hundred patients were affected and 52 patients (44%) died. The deaths of the patients were later attributed to exposure to cyanobacterial hepatotoxins, microcystins and the intoxication by microcystins is now referred to as the

“Caruaru Syndrome” (Pouria et al. 1998; Carmichael et al. 2001; Azevedo et al. 2002;

WHO 2003). This event has increased concern over the health effects of toxic cyanobacteria. These toxic events emphasize the importance in monitoring toxic cyanobacterial blooms for the purpose of minimizing potential damages to the environment and to the health of animals and human beings. It has been estimated that harmful cyanobacterial blooms cost the United States about $50 million per year

(Reynolds 2004).

Besides the direct and acute toxic effects caused by consumption of toxic blue- green algae polluted water, intoxication can occur through other routes of exposure

(Codd et al. 1997). Overall, exposure to toxic cyanobacteria include (1) oral exposure, i.e., ingestion of toxic cyanobacterial cells from drinking and recreational waters, and food and food supplements that contain or are contaminated with toxic cyanobacterial cells; (2) pulmonary route, exposure to toxic cyanobacteria occurs via inhalation of polluted water aerosols; (3) dermal route, exposure occurs via skin and mucosal contact; and (4) exposure via water used for haemodialysis, which occurred and caused human fatalities in Brazil in 1996 (Codd et al. 2005). Illness of human beings have been caused

15 by toxic cyanobacterial cells via all of these exposure routes while most of the reported animal toxic incidents occurred via ingestion of toxic cyanobacterial cells during a bloom.

Health advisories have been posted by various local health departments during cyanobacterial blooms to avoid human exposure to toxic cyanobacteria (Figure 2.2).

2.2 CYANOBACTERIAL TOXINS.

2.2.1. Introduction to cyanobacterial toxins.

The cyanobacterial toxins are a diverse group of extreme bioactive secondary metabolites produced by a variety of cyanobacterial species. Cyanobacterial toxins can be categorized into different groups based on their chemical structures, their mode of action, or th physiological systems (organs, tissues, or cells) affected. Structurally they can be classified into three major chemical structure groups: cyclic peptides, alkaloids and lipopolysaccharides (LPS) (Figure 2.3 and Figure 2.4). More commonly, cyanobacterial toxins are classified according to their toxic effects: (i) Hepatotoxins, the most often implicated in cyanobacterial toxicoses. These include the cyclic heptapeptide microcystins, of which over 90 structural variants are recorded, and the cyclic pentapeptide nodularins, of which about 6 variants are known. These peptides inhibit protein phosphatases, cause changes in membrane integrity and conductance, damage the liver and cause animal and human deaths due to hypovolemic shock and excessive blood pooling in the liver. In addition, these cyanobacterial toxins are tumor promoters (Yu

1989; Carmichael 1992; Marsalek 1995; Falconer and Humpage 1996; Sivonen 1996;

Codd 2001; Haider et al. 2003). The pentapeptide nodularin is also a carcinogen (Ohta et al. 1994); (ii) Neurotoxic alkaloids, which interfere with the functioning of the

Figure 2.2. Warning flyer released by Vermont Department of Health regarding blue- green algal blooms in Lake Champlain and the potential health risks.

17 neuromuscular system of the animals and often result in paralysis and very rapid death.

The most common neurotoxins are anatoxin-a, homoanatoxin-a, anatoxin-a(s), the PSP

toxins (paralytic shellfish poisoning toxins) and the non-protein amino acid toxin BMAA

(β-methylamino L-alanine). Anatoxin-a and homoanatoxin-a are postsynaptic, cholinergic neuromuscular blocking agents (Spivak et al. 1980; Wonnacott et al. 1992).

Anatoxin-a(s) is a guanidine methyl phosphate ester, which inhibits acetylcholinesterase

(Mahmood and Carmichael 1986, 1987; Matsunaga et al. 1989; Holzgrabe 1994). The

PSP toxins, including saxitoxin and about 20 structural variants, block sodium channels

(Carmichael 1997; Kaas and Henriksen 2000; Chorus et al. 2001). BMAA is an agonist

of the glutamate receptor by mimicking the effect of the neurotransmitter glutamate. It is

suspected to be a possible cause of the amyotrophic lateral sclerosis/Parkinsonism-

dementia complex (ALS/PDC) and has been found in patients’ brain tissue in Canada and

Guam (Copani et al. 1990; Cox et al. 2003; Cox et al. 2005; Banack et al. 2006); (iii)

Cytotoxic toxins. The most often encountered cytotoxin is cylindrospermopsin. It is a

guanidine alkaloid inhibitor of protein synthesis and causes widespread necrotic injury in

mammals (liver, kidneys, lungs, spleen, and intestine). Cylindrospermopsin is sometimes

also categorized into the group of hepatotoxins. It is genotoxic and can cause

chromosome loss and DNA strand breakage (Carmichael 1992, 1997; Codd et al. 1997;

Humpage et al. 2000; Shen et al. 2002); (iv) Dermatotoxic toxins. This category includes

aplysiatoxin, debromoaplysiatoxin, and lyngbyatoxin-a, which are produced by marine

cyanobacteria. They can cause skin irritation and are tumor promoters; and (v)

Lipopolysaccharide endotoxins (LPS), which are widely produced by cyanobacteria, may

R 2 7 6 CH N 2 CO2H

R O NH NH O 1 O CH3 5 R H3C O S OCH3 S NH R1 X 2 S R NH CH3 CH3 Y S 4 O 3 CO2H

1 2 MCYST-LA: X=Leu; R =CH3; Y=Ala; R =CH3

1 2 MCYST-M(O)R: X=Met(O); R =CH3; Y=Arg; R =CH3

1 2 MCYST-LR: X=Leu; R =CH3; Y=Arg; R =CH3

1 2 MCYST-YA: X=Tyr; R =CH3; Y=Ala; R =CH3

MCYST-RR: X=Arg; R1=H; Y=Arg; R2=H

1 2 MCYST-YR: X=Tyr; R =CH3; Y=Arg; R =CH3

Figure 2.3. Structure of microcystins. The general structure is a 7 amino group peptide: cyclo(D-Ala1-X2-D-MeAsp3/D-Asp3-Y4-Adda5-D-Glu6-Mdha7/Dha7). X and Y are variable L-amino acids; D-MeAsp is d-erythro-β-methylaspartic acid; Adda is (2S, 3S,

8S, 9S)-3-amino-9-methoxy-2,6,8-trimethyl-10-phenyldeca-4,6-dienoic acid and Mdha is

N-methyldehydroalanine (Dha is dehydroalanine). Several variations are listed above.

From (Carmichael 1992; Sivonen 1996).

Nodularin Saxitoxins (PSP toxins)

Cylindrospermopsin

Figure 2.4. Structures of the alkaloid cyanobacterial toxins: anatoxin-a, homoanatoxin-a, anatoxin-a(s), saxitoxin, cylindrospermopsin and nodularin. Anatoxin-a, homoanatoxin-a, anatoxin-a(s) and saxitoxin are neurotoxins. Cylindrospermopsin is a cytotoxin. The

Saxitoxins are a large toxin family referred as the PSP toxins. Nodularin is a hepatotoxin.

20 contribute to inflammatory and gastrointestinal incidents (Dow and Swoboda 2000; Codd

2001; USEPA 2001; Codd et al. 2005; Wiegand and Pflugmacher 2005).

Among the five different groups of cyanobacterial toxins, the first two categories,

hepatotoxins and neurotoxins are the most frequently encountered and also the most studied toxins due to their high toxicity, frequent presence and wide distribution in the aquatic environment. The World Health Organization has set an advisory limit of 1 μg

L-1 for the hepatotoxic microcystins in drinking water due to the previously-mentioned

"Caruaru Syndrome" and acute and chronic toxicity tests on laboratory animals. Other

countries, namely Canada, Australia and New Zealand have adopted similar guideline

values for microcystins. The neurotoxins target the neuromuscular system, paralyzing

skeletal and respiratory muscles and causing death within minutes. Although some of

these neurotoxins are more toxic (Table 2.2) than other toxins, and are responsible for a

series of lethal poisonings of livestock and birds (Negri et al. 1995; Bumke-Vogt et al.

1999; Yang et al. 2001; Haider et al. 2003), there are fewer “official” standards for these

toxins due to the inadequacy of the experimental toxicity data, and especially human

toxicity data. Several countries have issued advisory limits for neurotoxins in drinking and recreational waters (MHNZ 2000; USEPA 2001).

The physiological role of cyanobacterial toxins is not clear. Toxin production

may be an induced defense response to stress factors (Lampert 1981; Carmichael 1986).

Toxin production is influenced by environmental stresses such as pH, light, trace metals

etc. (Van der Westhuizen and Eloff 1983; Sivonen 1990; Lukac and Aegerter 1993).

Cyanobacterial toxins also have a negative impact on feeding and development of aquatic

Table 2.2. Cyanobacterial Toxins and Their Acute Toxicity (WHO 2003)

Cyanobacterial toxins LD50 i. p. mouse Mechanism of toxicity

Hepatotoxins

Microcystins in general (ca. 90 45-1000 μg kg-1 Blocks protein phosphatases

known congeners) by covalent binding and cause

Microcystin-LR 60 (25-125) μg kg-1 hemorrhaging of the liver;

Microcystin-YR 70 μg kg-1 cumulative damage may

Microcystin-RR 300-600 μg kg-1 occur.

Neurotoxins

Anatoxin-a 200-250 μg kg-1 Blocks post-synaptic

depolarization

Anatoxin-a(s) 40 μg kg-1 Blocks acetylcholinesterase

Saxitoxins 10-30 μg kg-1 Block sodium channels

Cytotoxin

Cylindrospermopsin 200-2000 μg kg-1 Blocks protein synthesis;

substantial cumulative toxicity

22 organisms including fish, invertebrates, and zooplankton (DeMott 1999; FerrAo-Filho et

al. 2000; Dionisio-Pires and Van Donk 2002). Increased growth of toxic cyanobacteria

and toxin production have been associated directly or indirectly with herbivores

(Vanderploeg et al. 2001; Jang et al. 2003), and cyanobacterial toxins may provide an

advantage to cyanobacteria by inhibiting zooplankton grazing (Lampert 1981; Nizan et al.

1986), bacterial settlement (Chrost 1975a, 1975b) and the growth of competing microbes

(Srivastava et al. 1998; Schlegel et al. 1999). However, most studies have shown that

toxins were produced under favorable growth conditions. Further examination on the

role and factors controlling toxin production is needed.

Aside from the acute toxicity of cyanobacterial toxins, some toxins such as

microcystins show chronic toxicity. Cyanobacterial toxins have chronic negative impacts on growth, development and reproduction of aquatic organisms (Oberemm et al. 1999;

Pietsch et al. 2001a; Rogers et al. 2005) and may have genotoxicity (Bouaicha et al. 2005;

Wu et al. 2006). Some cyanobacterial toxins are bioaccumulated and transferred to

higher trophic levels via the food web (Karjalainen et al. 2003; Karjalainen et al. 2005;

Xie et al. 2005). These emerging environmental and health problems caused by

cyanobacterial blooms and their toxins in recreational and drinking waters have caused

several countries to formulate risk management strategies. For example, Canada and

Australia have initiated guidelines for several of the cyanobacterial toxins. While there

are no official guidelines for cyanobacterial toxins in the U.S, the WHO’s guideline for

microcystins has been widely accepted and applied during risk assessment and

management of the potentially hepatotoxic bloom events. In spring of 2001, the United

States Environmental Protection Agency (USEPA) convened a panel of scientists to

23 assist in identification of a target list of cyanobacterial toxins that are likely to pose a

health risk in source and finished waters of the drinking water utilities in the United

States for regulatory consideration (USEPA 2001; Maizels and Budde 2004). The meeting was in accordance with the requirement of the 1996 Safe Drinking Water Act

(SDWA), and summarized background information on the USEPA drinking water contaminant candidates list, known cyanobacterial toxins, toxic cyanobacterial species, the occurrence and health effects of different cyanobacterial toxins, analytical methods for detecting different toxins, applicable treatment options for removal of cyanobacterial blooms and their toxins from water, and the stabilities as well as environmental fates of these toxins. Based on the potential health effects, occurrence in the United States, susceptibility to drinking water treatment and toxin stability, the neurotoxins anatoxin-a, the cytotoxin cylindrospermopsin and the cyclic heptapeptide hepatotoxins microcystins were recommended for further research (USEPA 2001; Maizels and Budde 2004). In

September 2005, another interagency symposium on cyanobacterial harmful algal blooms

(ISOC-HAB) was organized. This was in response to the reporting requirements called for by HABHRCA (Harmful Algal Bloom and Hypoxia Research and Control Act) and focused on occurrence of blooms and toxins; toxin kinetic, dynamics and OMICS

(research areas including genomics, proteomics, transcriptomics, metabolomics and physiomics); health and ecological effects; causes, prevention and mitigation; exposure assessment methods and risk assessment. The symposium was co-sponsored by USEPA,

National Oceanic and Atmospheric Administration (NOAA), Food and Drug

Administration (FDA), United States Geological Survey (USGS), the United States Army

Corps of Engineers (US ACE), National Institute of Health (NIH), National Institute of

24 Environmental Health Sciences (NIEHS), Center for Disease Control and Prevention

(CDC), US Department of Agriculture (USDA), and the Institute of Marine Sciences of

University of North Carolina at Chapel Hill (UNC-CH IMS) (USEPA 2005).

To date, efforts have been focused on monitoring the occurrence and effects of

toxic cyanobacterial blooms and their toxins. Although the neurotoxins anatoxin-a,

saxitoxin and the hepatotoxins microcystins are typically produced by Anabaena flos-

aquae, Aphanizomenon flos-aquae, and Microcystis aeruginosa respectively, it is not

enough to determine toxicity of an ongoing cyanobacterial bloom by merely determining its cyanobacterial composition. Both toxic and non-toxic strains of the same

cyanobacterial species can be present in the same bloom, and the same cyanobacterial

species isolated from different blooms can produce different toxins. Even if the bloom is

dominated by a known toxic cyanobacterial species, different strains of the same

cyanobacterial species may behave as neurotoxic, hepatotoxic or non-toxic (Gorham et al.

1964; Sivonen et al. 1990). Furthermore, not all cyanobacterial strains capable of toxin

production will produce cyanobacterial toxins in natural aquatic environments. The

production of multiple toxins, even toxins of different classes, have been observed in

toxic cyanobacterial strains (Mahmood and Carmichael 1987; Al-Layl et al. 1988;

Matsunaga et al. 1989; Harada et al. 1991; Carmichael 1997; Beltran and Neilan 2000;

Fromme et al. 2000; Kaas and Henriksen 2000; Rao et al. 2002; Namikoshi et al. 2003;

Viaggiu et al. 2003; Oberholster et al. 2004; Araoz et al. 2005). This illustrates the uncertainty in evaluating and determining the toxicity of cyanobacterial blooms. Little is known about why cyanobacteria produce toxins, why toxin production varies and why some toxic cyanobacterial species do not present toxicity all the time.

25 To better monitor the occurrences of cyanobacterial toxins and to prevent their potential toxicosis, the effects of environmental factors and other parameters on growth of toxic cyanobacteria and their production of toxins has been examined (Stevens and

Krieger 1991b; Rapala et al. 1993; Rapala et al. 1994; Rapala and Sivonen 1998; Chorus and Bartram 1999; Gupta et al. 2002; Xie and Xie 2002; Yamamoto and Nakahara 2005).

Light is essential for photosynthesis by cyanobacteria and toxic strains produce the most toxin when growing under optimum light conditions. Temperature also has an impact on toxin production. Different cyanobacterial species and strains differ slightly in their optimum growth temperatures, but generally toxin production is highest at temperatures ranging from 18 to 25oC whereas lower or higher temperatures may decrease toxin production (Sivonen 1990; Plinski and Jozwiak 1999). High nitrogen and /or phosphorus concentrations in the lake water are considered important factors in promoting eutrophication and stimulate growth of the cyanobacteria (Sivonen 1990). The limitation of algal growth by a lack of nutrients (nitrogen and phosphorus) is a well characterized phenomenon in nature (Rapala et al. 1993). However when the effects of these nutrients on toxin production were examined, the results were often conflicting. Nitrogen-fixing species do not depend on nitrogen in the media for toxin production (Rapala et al. 1993).

High concentrations of phosphorous normally favored the production of hepatotoxins in these cases, while no direct relationship between phosphorous level and anatoxin-a production was observed (Sivonen et al. 1990; Rapala et al. 1993; Rapala and Sivonen

1998; Chorus and Bartram 1999). Production of cyanobacterial toxins by Anabaena flos- aqua in lakes with low phosphorus concentrations (Willen and Mattsson 1997) indicated that toxin production was not always related to the nutrient status of the water body.

26 Other micronutrients such as trace metals affect the growth of cyanobacteria and may affect their production of toxins (Lukac and Aegerter 1993; Lyck et al. 1996; Chorus and

Bartram 1999). The presence of competing phytoplankton species and other aquatic

organisms may also influence production of cyanobacterial toxins and the fate of toxins

in aquatic environments (Watanabe et al. 1996; Kearns and Hunter 2000, 2001).

There are several ways to determine whether a bloom is toxic. Bioassays are easy

to use and can detect the qualitative and quantitative characteristics of toxins present

(Vezie et al. 1996). Bioassays can be easily interfered with by other factors, are unable

to measure low levels of toxins, and cannot distinguish different toxins within a given

group, e.g., different neurotoxins. Chemical techniques such as HPLC with either UV

(Harada et al. 1999), fluorimetric (James et al. 1997) or electrochemical detection (Boyer

and Goddard 2000), thin layer chromatography (Ojanpera et al. 1992), GC-ECD (Bumke-

Vogt et al. 1999), GCMS or LCMS (Harada et al. 1993; Zotou et al. 1993) have been

used to detect and distinguish toxins. Only a few of these techniques have been

developed into standard methods that can be easily used by monitoring agencies or test

laboratories. Immunoassays have also been developed or are in development for

cyanobacterial toxins. Immunoassay techniques for detecting hepatotoxic microcystins,

neurotoxic PSP toxins and cylindrospermopsin are already in use (Brooks and Codd 1988;

Chu et al. 1989; Metcalf et al. 2001; Usleber et al. 2001).

2.2.2. The neurotoxic cyanobacterial toxin, anatoxin-a

The neurotoxin anatoxin-a has been found in natural waters in North America,

Australia, New Zealand, many European countries and Asia. Although there are no

27 human poisonings directly related to anatoxin-a, the potential environmental and health

risks should not be underestimated due to its widespread occurrence and high toxicity.

Death usually occurs due to respiratory failure within a period of several minutes up to several hours depending upon the size of the animal and the amount of toxin ingested

(Carmichael et al. 1979; Carmichael and Gorham 1980; Spivak et al. 1980; Swanson et al.

1986). Historically called “very fast death factor” (VFDF), anatoxin-a was first reported by Gorham and co-workers as a potent neurotoxin produced by the freshwater cyanobacteria Anabaena flos-aquae (Lyngb.) de Breb. This culture was isolated in 1961

(Gorham et al. 1964) and deposited as clone NRC-44-1. Anatoxin-a was named after its origin from A. flos-aquae and to distinguish it from other neurotoxins produced by other strains of A. flos-aquae (Carmichael et al. 1975; Carmichael and Gorham 1978). The X- ray crystal structure and absolute configuration of anatoxin-a has been determined (Huber

1972). Anatoxin-a is a low molecular weight alkaloid (MW=165), semirigid, bicyclic secondary amine (2-acetyl-9-azabicyclo[4,2,1]non-2-ene, Figure 2.5). The bicyclic structure and the conjugated ketone group of anatoxin-a severely restrict its conformation and make anatoxin-a a potent agonist for the nicotinic acetylcholine receptor (Spivak et al. 1980; Swanson et al. 1986; Brough et al. 1992; Thomas et al. 1993). Pharmacolo- gically, anatoxin-a exerts its neurotoxic effect by mimicking the neurotransmitter acetylcholine and irreversibly binding the nicotinic acetylcholine receptor (nAChR)

(Figure 2.6) (Aronstam and Witkop 1981; Thomas et al. 1993). When acetylcholine is released by neurons that impinge on muscle cells, it binds to the receptor’s nACh neurotransmitter binding site. Attachment of acetylcholine to the receptors opens an adjacent ion channel, allowing ionic movement across the membrane that induces the

28 muscle cell to contract. Soon after, acetylcholinesterase degrades acetylcholine, the

channel closes, and the receptor prepares for a new signal. Degradation of acetylcholine

prevents the overstimulation and allows relaxation of the muscle cell. Anatoxin-a binds

to the nicotinic acetylcholine receptor and stimulates muscle contraction, but is not

degraded or cleaved by acetylcholinesterase. Overstimulation of the muscles induces

muscle twitching and cramping followed by fatigue and paralysis. If the respiratory

muscles are affected, the animals may suffer convulsions due to lack of oxygen to the

brain and eventually die of suffocation (Carmichael 1994). Indications of neurotoxicosis

in terrestrial animals include hypersalivation, muscle twitching, rigors, cyanosis,

convulsions and coma. Acute toxicity of anatoxin-a has been determined using mice in

laboratory intraperitoneally (i.p.) to be 200-250 μg kg-1 body weight (Stevens and

Krieger 1991a; Fawell et al. 1999). Efforts have also been made to determine the oral

acute toxicity of anatoxin-a since it is the route of toxin administration in natural waters.

Although preliminary oral toxicity was proposed as 5-16 mg kg-1 body weight using mice

in laboratory, no dosage-mortality relationship was developed so that the toxicity data

were merely based on the observation that fatalities occurred at toxin dosage higher than

5 mg kg-1 body weight (Fitzgeorge et al. 1994). Anatoxin-a was responsible for the death

of several dogs in 1990 and 1991 in Scotland (Edwards et al. 1992; Gunn et al. 1992) and

France (Gugger et al. 2005). In the United States, the death of several dogs in Lake

Champlain, NY in 1999 and 2000 were attributed to anatoxin-a poisoning (Chapter 4)

(Yang et al. 2001; Boyer et al. 2004). Insufficient data are available for derivation of a

chronic toxicity value for anatoxin-a.

29 H O H O N N

Dihydroanatoxin-a Epoxyanatoxin-a

Figure 2.5. Conformation of anatoxin-a and structures of the two known non-toxic degradation products, epoxyanatoxin-a and dihydroanatoxin-a.

Figure 2.6. Mechanism of anatoxin-a poisoning (Carmichael 1994).

31 Besides the detrimental effects, anatoxin-a and related synthetic analogues have

become widely used in medicinal and pharmacological applications due to their

irreversible binding to the nicotinic acetylcholine receptor (nAChR). Anatoxin-a is used

to study how acetylcholine binds and influences the activity of the nAChR, and the

mechanism of neuromuscular action in the peripheral and central nervous systems

(Swanson et al. 1989; Swanson et al. 1990; Swanson et al. 1991; Thompson et al. 1992;

Thomas et al. 1994; Clementi et al. 1998). Acetylcholine itself cannot be administrated

as a therapy since it is cleaved by acetylcholinesterase and quickly disappears. Because

of the high toxicity of anatoxin-a, the N-substituted derivatives of the toxin with greatly

reduced toxicity have been synthesized (Swanson et al. 1989; Swanson et al. 1991;

Wonnacott et al. 1991; Thomas et al. 1994). These modified analogues were used to

elucidate the nicotinic acetylcholine receptor subtypes (Aracava et al. 1988; Alkondon

and Albuquerque 1991; Wonnacott et al. 1991; Amar et al. 1993), and may lead to new

therapeutics acting as acetylcholine replacement candidates (Haider et al. 2003). These

studies of nAChRs may help our understanding neurodegenerative CNS disorders, such

as Alzheimer's disease, associated with an inability of neurons to produce acetylcholine.

A number of studies have been conducted to elucidate the biosynthesis of

anatoxin-a in cyanobacteria. Anatoxin-a appears to be related to the tropane class of

alkaloids found in higher plants. It has been suggested the biosynthesis of anatoxin-a is

analogous to that of tropanes and formed from ornithine/arginine via the amine putrescine.

This amine would be oxidized to a pyrroline, the precursor of anatoxin-a (Gallon et al.

1990; Gallon et al. 1994). Later labeling studies using 13C NMR (nuclear magnetic

resonance) spectrometry indicated the carbon skeleton of anatoxin-a was derived from

32 acetate and glutamate. The carbon atoms C-1, -5, -6, -7, -8 in anatoxin-a were derived

from glutamic acid while the other carbon atoms came from acetate. The C-1 of glutamic

acid was retained during transformation of anatoxin-a (Figure 2.7). This finding was not

compatible with the tropane alkaloid theory (Hemscheidt et al. 1995), suggesting

anatoxin-a was not produced as proposed. A new biosynthetic reaction sequence for

anatoxin-a and its analogues has been proposed (Figure 2.8) (Gallon et al. 1994;

Hemscheidt et al. 1995; Namikoshi et al. 2004). Most recently, an immediate precursor

of anatoxin-a, 11-carboxy anatoxin-a has been identified and characterized (Selwood et al.

2006), and further confirms the proposed biosynthesis of anatoxin-a. The genes and

enzymes involved in biosynthesis of anatoxin-a are still under investigation.

The fate of anatoxin-a varies in different aquatic environments. Once released

into the water column from cyanobacterial cells during cell senescence and lysis, the

concentration of anatoxin-a is influenced by many environmental factors. When exposed

to strong sunlight and high pH, anatoxin-a quickly undergoes photochemical degradation

to non-toxic degradation products (Himberg 1989; Stevens and Krieger 1991b; Sivonen

2000) and disappears from the water column. The half-life of anatoxin-a under strong

solar radiation is in the order of hours both in sunlight and in the laboratory. Two

degradation products, dihydroanatoxin-a and epoxyanatoxin-a, have been identified from

ageing blooms of cyanobacteria (Smith and Lewis 1987; Harada et al. 1993; Furey et al.

2003). Alkalinity also affects the stability of anatoxin-a in natural waters. In absence of

sunlight, anatoxin-a can undergo non-photochemical degradation with a half life (t1/2) of several weeks under basic conditions (pH >8) (Stevens and Krieger 1991b; Smith and

Sutton 1993). Lake sediments may adsorb anatoxin-a and reduce its bioavailability

33 9 NH O 1 10 2 11 8 6 5 3 7 4

Figure 2.7. Sections of the 125 MHz 13C NMR spectrum of anatoxin-a. (a) Sample

isolated from an experiment which incorporated [1-13C]glutamate/[2-13C]acetate showed

that C-5 derived from the carboxyl group of glutamic acid and C-4 derived from the

13 methyl group of acetate. (b) Sample isolated from an experiment incorporating (S)-[ C5] glutamate. The homonuclear coupling pattern suggested that the complete carbon chain of the precursor amino acid is incorporated intact into the toxin. Reprinted from

(Hemscheidt et al. 1995).

Figure 2.8. A plausible biosynthetic scheme for the neurotoxic anatoxin-a (1) and its precursor 11-carboxy anatoxin-a (2). From Hemscheidt et al. (1995) and Selwood et al.

(2006).

35 (Rapala et al. 1994). Microorganisms such as Pseudomonas sp., may also decrease the concentration of anatoxin-a through biodegradation (Rapala et al. 1994).

Besides the original Anabaena flos-aqua, anatoxin-a is also produced by other species of Anabaena and other genera of cyanobacteria including Aphanizomemon and

Planktothrix/Oscillatoria. Microcystis species, which are well-known microcystin producers, may also produce anatoxin-a (Edwards et al. 1992; Park et al. 1993; Bruno et al. 1994; Bumke-Vogt et al. 1999; Krienitz et al. 2003; Viaggiu et al. 2004). More recently, anatoxin-a was isolated from mat- and biofilm-forming benthic cyanobacteria

Phormidium favosum and Raphidiopsis mediterranea (Krienitz et al. 2003; Namikoshi et al. 2003). As new anatoxin-a-producing species are isolated each year, the number of cyanobacterial species able to produce this neurotoxin may be greater than currently estimated.

When an anatoxin-a-producing cyanobacterial bloom occurs, the maximum anatoxin-a concentration in cyanobacterial cells is found during the logarithmic growth phase. Anatoxin-a is released into the water column during the ageing and decomposition of the bloom. The concentration of anatoxin-a in natural waters is of key importance in its toxic effects. While high concentrations of anatoxin-a have been occasionally detected from some cyanobacterial blooms, e.g., 4,400 μg g-1 dry weight from a cyanobacterial bloom in Finland, most reported occurrences of anatoxin-a contained relatively lower toxin concentrations (Table 2.3) (Sivonen et al. 1989; Harada et al. 1993; James et al. 1997; Chorus and Bartram 1999). There is no universally accepted guideline for the maximum allowed concentration of anatoxin-a in drinking and recreational waters. Using mouse toxicity data, Fawell et al. suggested that a guideline

Table 2.3. Anatoxin-a concentrations detected in various water bodies.

Location Year Concentration of anatoxin-a References

China 2003 ca. 0.007 μg L-1 (Zhang et al. 2003)

France 2003 Up to 8,000 μg g-1 dry weight (Gugger et al. 2005)

Finland 1985-87 10 – 4,400 μg g-1 dry weight (Sivonen et al. 1989)

Germany 1999 Up to 13.1 μg L-1 (Bumke-Vogt et al.

1999)

Germany 1995-96 0.02 – 0.36 μg L-1 (Bumke-Vogt et al.

1999)

Ireland 1995 10 – 100 μg g-1 dry weight (James et al. 1997)

Italy 2003 12 μg g-1 (Viaggiu et al. 2004)

Japan 1988-92 0.3 – 16 μg g-1 dry weight (Harada et al. 1993)

Japan 1993 0.4 – 2,600 μg g-1 dry weight (Park et al. 1993)

37 value for anatoxin-a in drinking water of 1 μg L-1 would provide an adequate margin of safety (Fawell et al. 1999). Australia has suggested a guideline of 3 μg L-1 and New

Zealand has a provisional maximum acceptable value of 6 μg L-1 for anatoxin-a in drinking water (USEPA 2001; WHO 2004).

2.2.3. Detection of anatoxin-a in natural waters

A number of analytical methods have been devised and applied for detection of anatoxin-a in natural waters. Beside bioassays, chromatographic methods including thin- layer chromatography, gas chromatography (GC), high-performance liquid chromatography (HPLC) with ultraviolet detection or fluorimetric detection, gas chromatography-mass spectrometry (GCMS), and liquid chromatography-mass spectrometry (LCMS) using electrospray ionization (Stevens and Krieger 1988; Skulberg et al. 1992; Rapala et al. 1993; Zotou et al. 1993; Jefferies et al. 1994; James and

Sherlock 1996; Powell 1997; Takino et al. 1999; Draisci et al. 2001b; Pietsch et al.

2001b; Namera et al. 2002; Ruseva et al. 2003; Dell'Aversano et al. 2004; Ghassempour et al. 2005; James et al. 2005). Most of these methods are time consuming and can only be performed in laboratories with the appropriate instrumentation and training. Real-time in situ monitoring of this neurotoxin has yet to be achieved.

Aside from the availability and complexity of these methods, other difficulties also exist for monitoring anatoxin-a. Forensic investigations of suspected anatoxin-a poisonings are often impeded by the labile property of the toxin and interferences from other organic compounds in the biological and environmental matrices. The rapid

38 degradation of anatoxin-a in natural waters hinders accurate estimation of toxin concentrations. Several recent incidents highlighted the possible confusion in suspected anatoxin-a poisonings due to the presence of the amino acid phenylalanine (phe). In summer 2002, the death of a young adult in the USA, following exposure to cyanobacterial bloom contaminated lake water, was ascribed to anatoxin-a poisoning in the coroner’s report. This was based mainly on identification of anatoxin-a using liquid chromatography-single quadrupole MS (LCMS). This identification was shown to be incorrect and was the result of confusion between anatoxin-a and phenylalanine. These compounds are isobaric (same molecular weight) and have similar LC retention times under the analytical instrumental conditions used. In September 2003, another investigation of the fatal intoxication of two dogs in a lake in eastern France again found the presence of anatoxin-a in benthic cyanobacteria Planktothrix along the shoreline. The presence of high concentrations of phenylalanine prevented confirmation and quantification of the neurotoxin in the tissue samples of these dogs (Dagnino and

Schripsema 2005; Gugger et al. 2005). Thus, caution should be taken in monitoring anatoxin-a in natural water samples as interference from this common amino acid will inevitably occur. Combinational analysis such as sensitive determination of anatoxin-a using fluorimetric LC methods together with mass spectrometry can help distinguish anatoxin-a from this phenylalanine interference (Furey et al. 2005).

2.3. THE NEW YORK STATE LAKES.

New York State is rich in natural water resources (Figure 2.9). These water bodies serve as important drinking water resources for New York State residents and/or

39 important recreational resources for fishing, swimming and other water related activities.

A number of other New York lakes, ponds and reservoirs are experiencing increasingly serious problems of cyanobacterial blooms and await monitoring of toxic cyanobacteria and their toxins. Among these water bodies, the lower Great Lakes and Lake Champlain have received the most attention.

The lower Great Lakes (Lake Erie and Lake Ontario) are uniquely different in both their physical and biological properties and in the degree of human use and shoreline development than the other Great Lakes, namely Lake Superior, Michigan and Huron

(Stephenson 2003). Together, the Laurentian Great Lakes constitute the largest continuous mass of fresh water on earth; these freshwater systems contain about 21 percent of the world’s unfrozen fresh water supply and nearly 90 percent of the

U.S. supply, and represent an enormous cultural and economic resource for both the United States and Canada (Figure 2.10) (Fuller and Shear 1995; Stephenson 2003).

In spite of their large sizes, both Lake Erie and Ontario are sensitive to a wide range of exogenous influences. Lake Erie is larger in surface area than Lake Ontario, but it is the shallowest of all five lakes with an average depth of 19 meters. It is the smallest of the Great Lakes in volume and the most productive. Lake Erie supports the largest sport fishery in the Great Lakes and the largest freshwater commercial fishery in the world. The lake also provides drinking water to 11 million people each day. Because of the fertile soils surrounding the lake, i.e., southwestern Ontario, Ohio, Indiana and

Michigan, are intensively used for agriculture, the lake receives large amounts of runoff from these agricultural areas. There are also major metropolitan areas within the Lake

Erie basin. As a result, Lake Erie is exposed to the greatest effects from urbanization and agriculture, receives more pollutants, and more nutrients than the other lakes. This has led to increased anthropogenic eutrophication. Massive blooms and anoxia (depletion of

40 oxygen) occur throughout most of the lake and the lake was considered “dead” during the

1970s. The Great Lakes Water Quality Agreement (GLWQA) and its amendment signed by the United States and Canada provided the basis for managing these boundary waters.

The GLWQA called for phosphorus load reduction in all of the Great Lakes and tightened pollution laws have reduced nutrients and in particular, phosphorous-containing pollutants, loading into the lake (Fuller and Shear 1995; McGucken 2000). Despite the successful phosphorous abatement and reversal of eutrophication in Lake Erie (Hasler 1969; Edmondson 1970; Vallentyne et al. 1970; McGucken 2000; Nicholla et al. 2001), the excessive growth of phytoplankton, especially toxic cyanobacteria still plague the lake annually and many environmental and health problems have been generated accordingly. In 1990s cyanobacterial blooms with the ability to produce potent toxins re-emerged in the lake.

The previously dominant filamentous algae, primarily Cladophora, were replaced by blooms of toxic Microcystis, especially in the western basin (Brittain et al. 2000; Eadie et al. 2004). These Microcystis blooms were associated with the hepatotoxic microcystins in the western basin at concentrations even exceeding 20 µg L-1 (Rinta-Kanto et al. 2005).

Lake Ontario is slightly smaller in surface area but much deeper than its upstream neighbor Lake Erie. The lake serves as a major drinking, recreational and industrial water source for New York State and Ontario. Although several major urban industrial centers such as Rochester, Hamilton and Toronto, are located in the basin, the shore on

41 C

A D

North B

Figure 2.9. Selected water bodies in New York State. A and B indicate Lake Ontario and

Lake Erie. Smaller lakes include Lake Champlain (C), Oneida Lake (D) . Oonondaga

Lake and Lake Neatahwanta are near Oneida Lake, and Agawam Lake (E).

Figure 2.10. The Great Lakes Basin. From Great Lakes Information Network (GLIN), http://www.great-lakes.net/lakes/.

43 the New York State side is less urbanized and is not intensively farmed compared to Lake

Erie. However, as the last lake of the Great Lakes chain before the water heads out the St.

Lawrence River, Lake Ontario receives pollutants and nutrients from other lakes. This contributes to the eutrophication process in the lake and is one of the reasons why Lake

Ontario has a higher pollution load than the other Great Lakes. Enormous algal blooms were frequently observed lakewide in the 1930s. Their occurrence has decreased somewhat, but in 1970s algal proliferation increased again. In the last decade, toxic algal blooms have occurred annually in the lake, especially in the near-shore and embayments near human habitants on both New York and Ontario side. Figure 2.11 showed a lakewide algal bloom in the lake in summer 2000. Large masses of decaying algae generated problems along nearly all of the shoreline and cyanobacterial toxins

(microcystins and anatoxin-a) were identified (Makarewicz et al. 2006). Environment

Canada has identified the appearance of blue-green algae in the Great Lakes as ''a growing threat to water quality in Canada and around the world''.

Lake Champlain is the largest body of fresh water in the United States aside from the Great Lakes. At 200 km long, with 1124 square kilometers of surface area, over 900 km of shoreline, and a drainage basin of nearly 20,000 square kilometers, the lake is regarded as the “sixth Great Lake”. It is shared by the states of Vermont and New York, as well as a small part in the Canadian province of Quebec (Figure 2.12). Over 20% of the basins’ land use is classified as developed or agricultural, the majority of which is concentrated along the shoreline. Approximately 200,000 people or 35% of the basin population depend on Lake Champlain for a source of drinking water. Over 4000 people draw their water directly from the lake for daily individual use. Lake Champlain is also a

44 significant recreational resource with numerous public beaches, parks, and marinas.

Recreational use such as fishing, boating, swimming and tourism are a major part of the local economies and contributed around $4 billion USD in 1998-1999 (Lake Champlain

Basin Program, 2005).

Geologically, Lake Champlain is divided into five distinct sub-basins, each of which has different limnological conditions (combination of physical and chemical characteristics) and water quality as shown in Figure 2.12 (Shanley and Denner 1999).

Missisquoi Bay (A) lies on the northern end of Lake Champlain, is very shallow and turbid and has suffered from eutrophication due to nutrient loading from the Missisquoi

River and Pike River. Missisquoi Bay flows south into the Inland Sea (B), which represents the northeast arm of the lake and lies to the east the Champlain Islands and is also high in nutrients. South of the Inland Sea is Malletts Bay (C), a rather small basin with restricted exchange with the rest of the lake due to a highway causeway on its north side and a former railroad causeway on its west side. The South Lake (F) is narrow, shallow, and turbid, accounting for 40% of the length of the lake but only less than 1% of its volume. It is also eutrophic. Over 80% of the total surface area of Lake Champlain is contained within the relatively deep Main Lake (rest of the lake), which is also known as the Broad Lake. The Main Lake is oligotrophic. For the purpose of monitoring anatoxin- a, Lake Champlain was re-divided into a slightly different set of regions other than the geological sub-basins. Mallets Bay was included as part of the Inland Sea while the Main

Lake was subdivided into two parts, the Northwest Arm (D) and the Main Lake (E)

(Figure 2.12).

Figure 2.11. A view of the lower Great Lakes from SeaWiFS satellite showing an algal bloom in the western basin of Lake Erie and a lake-wide algal bloom in Lake Ontario.

Satellite: OrbView-2. Sensor: SeaWiFS. Image Date: 9-06-2000. Image from NASA

Visible Earth web site: http://visibleearth.nasa.gov. Note that lake-wide algal blooms have also been frequently observed in Lake Erie.

46 A D B

Figure 2.12. Map of Lake Champlain Basin. The lake is geologically divided into 5 distinct sub-basins according to their different limnological conditions. They are

Missisquoi Bay (A), Inland Sea (B), Mallet Bay (C), South Lake (F) and Main Lake (the rest of the lake area). For the purpose of monitoring anatoxin-a in Lake Champlain,

Mallet Bay was grouped together with Inland Sea as the new Inland Sea, and the Main

Lake was subdivided into two parts: Northwest Arm (D) and Main Lake (E).

47 Cyanobacteria, or blue-green algae, are a common and natural component of the phytoplankton community in Lake Champlain (Shambaugh et al. 1999; Mihuc et al.

2005). Anabaena species were the predominant cyanobacteria taxon in Lake Champlain before 1980s (Muenscher 1930). In recent years, potentially toxic cyanobacterial blooms dominated by Microcystis have become a common problem in Lake Champlain. Both

Vermont Department of Health and Quebec Ministry of Health posted Health advisories in recent years regarding toxic cyanobacterial blooms at public beaches and in water supplies. There were no human poisonings reported in Lake Champlain Basin, but animal fatalities linked with toxic cyanobacteria have been observed. In 1999, two dogs died soon after consuming toxic cyanobacterial cells. In 2000, two additional dog fatalities occurred on Lake Champlain. In these latter cases, the lethal agent was identified as anatoxin-a (Yang et al. 2001). In 2002, at least one dog death in the United

States and two in Quebec, Canada were attributed to the ingestion of toxic Microcystis scums from Missisquoi Bay (Boyer et al. 2004). These toxic cyanobacterial blooms led to closure of Missisquoi Bay beaches and public water supplies in Quebec, and issuance of public alerts in the U.S. Recognizing these hazards, the Vermont Department of Health and Quebec Ministry of Health have periodically issued health advisories regarding using of Lake Champlain during summer and fall to avoid exposure to the potential toxic cyanobacterial blooms. Several institutes in both the United States (New York and

Vermont) and Canada (Quebec) have initiated programs to monitor for the presence of toxic cyanobacteria in Lake Champlain. However, the current research on cyanobacterial toxins in Lake Champlain are mostly focused on the occurrence of microcystins (Watzin et al. 2005).

48 2.4. REFERENCES

Al-Layl, K., Poon, G. K., and Codd, G. A. 1988. Isolation and purification of peptide and alkaloid toxins from Anabaena flos-aquae using high performance thin-layer chromatography. Journal of Microbiological Methods 7:251-258. Alkondon, M., and Albuquerque, E. X. 1991. Initial characterization of the nicotinic acetylcholine receptors in rat hippocampal neurons. Journal of Receptor Research 11 (6):1001-1021. Amar, M., Thomas, P., Johnson, C., Lunt, G. G., and Wonnacott, S. 1993. Agonist pharmacology of the neuronal alpha 7 nicotinic receptor expressed in Xenopus oocytes. FEBS Letters 327 (3):284-288. Aracava, Y., Swanson, K. L., Rozental, R., and Albuquerque, E. X. 1988. Structure- activity relationships of (+)anatoxin-a derivatives and enantiomers of nicotine on the peripheral and central nicotine acetylcholine receptor subtypes. International Congress Series 832 (NEUROTOX '88):157-184. Araoz, R., Nghiem, H.-O., Rippka, R., Palibroda, N., Tandeau de Marsac, N., and Herdman, M. 2005. Neurotoxins in axenic oscillatorian cyanobacteria: Coexistence of anatoxin-a and homoanatoxin-a determined by ligand-binding assay and GC/MS. Microbiology (Reading, United Kingdom) 151 (4):1263-1273. Aronstam, R. S., and Witkop, B. 1981. Anatoxin-a interactions with cholinergic synaptic molecules. Proceedings of the National Academy of Sciences of the United States of America 78 (7):4639-4643. Azevedo, S. M. F. O., Carmichael, W. W., Jochimsen, E. M., Rinehart, K. L., Lau, S., Shaw, G. R., and Eaglesham, G. K. 2002. Human intoxication by microcystins during renal dialysis treatment in Caruaru--Brazil. Toxicology 181-182:441-446. Banack, S. A., Murch, S. J., and Cox, P. A. 2006. Neurotoxic flying foxes as dietary items for the Chamorro people, Marianas Islands. Journal of Ethnopharmacology 106 (1):97-104. Beltran, E. C., and Neilan, B. A. 2000. Geographical segregation of the neurotoxin- producing cyanobacterium Anabaena circinalis. Applied and Environmental Microbiology 66 (10):4468-4474.

49 Berg, K., Skulberg, O. M., Skulberg, R., Underdal, B., and Willen, T. 1986. Observations of toxic blue-green algae (cyanobacteria) in some Scandinavian lakes. Acta Veterinaria Scandinavica 27:440-452. Berman-Frank, I., Lundgren, P., and Falkowski, P. 2003. Nitrogen fixation and photosynthetic oxygen evolution in cyanobacteria. Research in Microbiology 154:157-164. Bouaicha, N., Maatouk, I., Plessis, M.-J., and Perin, F. 2005. Genotoxic potential of microcystin-LR and nodularin in vitro in primary cultured rat hepatocytes and in vivo in rat liver. Environmental Toxicology 20 (3):341-347. Boyer, G. L., and Goddard, G. D. 2000. High performance liquid chromatography (HPLC) coupled with post-column electrochemical oxidation (ECOS) for the detection of PSP toxins. Natural Toxins 8:1-7. Boyer, G. L., Watzin, M. C., Shambaugh, A. D., Satchwell, M. F., Rosen, B. H., and Mihuc, T. 2004. The occurrence of cyanobacterial toxins in Lake Champlain. In Lake Champlain: Partnership and Research in the New Millennium, edited by T. Manley, P. L. Manley and T. Mihuc: Kluwer Academic/Plenum Publishers. p241- 257. Brittain, S. M., Wang, J., Babcock-Jackson, L. K., Carmichael, W. W., Rinehart, K. L., and Culver, D. A. 2000. Isolation and characterization of microcystins, cyclic heptapeptide hepatotoxins from a Lake Erie strain of Microcystis aeruginosa. Journal of Great Lakes Research 26:241-249. Brooks, W. P., and Codd, G. A. 1988. Immunoassay of hepatotoxic cultures and water blooms of cyanobacteria using Microcystis aeruginosa peptide toxin polyclonal antibodies. Environmental Technology Letters 9:1343-1348. Brough, P. A., Gallagher, T., Thomas, P., Wonnacott, S., Baker, R., Abdul Malik, K. M., and Hursthouse, M. B. 1992. Synthesis and x-ray crystal structure of 2-acetyl-9- azabicyclo[4.2.1]nonan-3-one. A conformationally locked s-cis analog of anatoxin-a. Journal of the Chemical Society, Chemical Communications (15):1087-1089.

50 Bruno, M., Attard, B. D., Pierdominici, E., Serse, A. P., and Ioppolo, A. 1994. Anatoxin- a and a previously known toxin in Anabaena planctonica from blooms found in Lake Mulargia (Italy). Toxicon 32 (3):369-373. Bumke-Vogt, C., Mailahn, W., and Chorus, I. 1999. Anatoxin-a and neurotoxic cyanobacteria in German lakes and reservoirs. Environmental Toxicology 14 (1):117-125. Carmichael, W. W. 1986. Algal toxins. Advances in Botanical Research 12:47-101. ———. 1992. Cyanobacteria secondary metabolites---the cyanotoxins. Journal of Applied Bacteriology 72:445-459. ———. 1994. The toxins of cyanobacteria. Scientific American 270:78-86. ———. 1997. The cyanotoxins. Advances in Botanical Research 27:211-256. ———. 2001a. The cyanotoxins-bioactive metabolites of cyanobacteria: occurrence, ecological role, toxanomic concerns and effects on humans. Journal of Phycology 37 (3):9-9. ———. 2001b. Health effects of toxin-producing cyanobacteria: "The CyanoHABs". Human and Ecological Risk Assessment 7 (5):1393-1407. Carmichael, W. W., Azevedo, S. M. F. O., An, J. S., Molica, R. J. R., Jochimsen, E. M., Lau, S., Rinehart, K. L., Shaw, G. R., and Eaglesham Geoffrey, K. 2001. Human fatalities from cyanobacteria: chemical and biological evidence for cyanotoxins. Environmental Health Perspectives 109 (7):663-668. Carmichael, W. W., Biggs, D. F., and Peterson, M. A. 1979. Pharmacology of anatoxin-a, produced by the freshwater cyanophyte Anabaena flos-aquae NRC-44-1. Toxicon 17 (3):229-236. Carmichael, W. W., Briggs, D. F., and Gorham, P. R. 1975. Toxicology and pharmacological action of Anabaena flos-aquae toxin. Science 187:542-544. Carmichael, W. W., and Gorham, P. R. 1978. Anatoxins from clones of Anabaena flos- aquae isolated from lakes of western Canada. Mitteilungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 21:285-295.

51 ———. 1980. Freshwater cyanophyte toxins: types and theri effects on the use of micro algae biomass. In Algae Biomass, edited by G. Shelef and C. J. Soeder: Elsevier/North-Holland Biomedical Press. p437-448. Chorus, I., and Bartram, J., eds. 1999. Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring, and Management. London: E & FN Spon (for the World Health Organization). 416 p. Chorus, I., Bumke-Vogt, C., Lindenschmidt, K.-E., and Jaime, E. 2001. Cyanobacterial neurotoxins. In Cyanotoxins - Occurrence, Causes, Consequences, edited by I. Chorus: Springer-Verlag. p37-45. Chrost, R. J. 1975a. Inhibitors produced by algae as an ecological factor affecting bacteria in water ecosystems. I. Dependence between phytoplankton and bacteria development. Acta Microbiologica Polonica Seria B 7:125-133. ———. 1975b. Inhibitors produced by algae as an ecological factor affecting bacteria in water. II. Antibacterial activity of algae during blooms. Acta Microbiologica Polonica Seria B 7 (3):167-176. Chu, F. S., Huang, X., Wei, R. D., and Carmichael, W. W. 1989. Production and characterization of antibodies against microcystins. Applied and Environmental Microbiology 55:1928-1933. Clementi, F., Dolly, O. J., Onali, P., and Bagetta, G. 1998. Neurotoxins in neurobiology: from basic research to clinical application. Toxicon 36 (12):2043-2046. Codd, G. A. 1995. Cyanobacterial toxins: occurrence, properties and biological significance. Water Science and Technology 32:149-156. ———. 2000. Cyanobacterial toxins, the perception of water quality, and the prioritisation of eutrophication control. Ecological Engineering 16:51-60. ———. 2001. Cyanobacterial toxins: their actions and multiple fates in microbes, animals and plants. Journal of Phycology 37 (3):13-13. Codd, G. A., and Bell, S. G. 1985. Eutrophication and toxic cyanobacteria. Journal of Water Pollution Control 34:225-232. Codd, G. A., Bell, S. G., and Brooks, W. P. 1989. Cyanobacterial toxins in water. Water Science and Technology 21:1-13.

52 Codd, G. A., Morrison, L. F., and Metcalf, J. S. 2005. Cyanobacterial toxins: risk management for health protection. Toxicology and Applied Pharmacology 203:264-272. Codd, G. A., Ward, C. J., and Bell, S. G. 1997. Cyanobacterial toxins: occurrence, modes of action, health effects and exposure routes. In Applied Toxicology: Approaches Through Basic Science, Archives of Toxicology Supplement. 19, edited by J. P. Seiler and E. Vilanove. Berlin: Springer. p399-410. Copani, A., Canonico, P. L., and Nicoletti, F. 1990. β-N-Methylamino-L-alanine (L- BMAA) is a potent agonist of ‘metabolotropic’ glutamate receptors European Journal of Pharmacology 181 (3):327. Cox, P. A., Banack, S. A., and Murch, S. J. 2003. Biomagnification of cyanobacterial neurotoxins and neurodegenerative disease among the Chamorro people of Guam. Proceedings of the National Academy of Sciences 100 (23):13380-13383. Cox, P. A., Banack, S. A., Murch, S. J., Rasmussen, U., Tien, G., Bidigare, R. R., Metcalf, J. S., Morrison, L. F., Codd, G. A., and Bergman, B. 2005. Diverse taxa of cyanobacteria produce β-N-methylamino-L-alanine, a neurotoxic amino acid. Proceedings of the National Academy of Sciences 102 (14):5074-5078. Crayton, M. A. 1993. Toxic cyanobacteria blooms: A field/laboratory guide. Office of Toxic Substances, Washington Department of Health, 21 p. Dagnino, D., and Schripsema, J. 2005. 1H NMR quantification in very dilute toxin solutions: application to anatoxin-a analysis. Toxicon 46 (2):236-240. Dell'Aversano, C., Eaglesham, G. K., and Quilliam, M. A. 2004. Analysis of cyanobacterial toxins by hydrophilic interaction liquid chromatography-mass spectrometry. Journal of Chromatography, A 1028 (1):155-164. DeMott, W. R. 1999. Foraging strategies and growth inhibition in five daphnids feeding on mixtures of a toxic cyanobacterium and a green alga. Freshwater Biology 42 (2):263-274. Dionisio-Pires, L. M., and Van Donk, E. 2002. Comparing grazing by Dreissena polymorpha on phytoplankton in the presence of toxic and non-toxic cyanobacteria. Freshwater Biology 47 (10):1855-1865.

53 Dow, C. S., and Swoboda, U. K. 2000. Cyanotoxins. In The Ecology of Cyanobacteria: Their Diversity in Time and Space, edited by B. A. Whitton and M. Potts. Boston: Kluwer Academic. p613-632. Downing, J. A., Watson, S. B., and McCauley, E. 2001. Predicting cyanobacteria dominance in lakes. Canadian Journal of Fisheries and Aquatic Sciences 58 (10):1905-1908. Draisci, R., Ferretti, E., Marchiafava, C., and Quadri, F. D. 2001a. Occurrence of anatoxins in blue-green algae food supplements. Advances in Mass Spectrometry 15:607-608. Draisci, R., Ferretti, E., Palleschi, L., and Marchiafava, C. 2001b. Identification of anatoxins in blue-green algae food supplements using liquid chromatography- tandem mass spectrometry. Food Additives and Contaminants 18 (6):525-531. Eadie, B. J., Ludsin, S., Schwab, D. J., and DePinto, J. 2004. Lake Erie research planning workshop. GLERL/NOAA, 28 p. Edmondson, W. T. 1970. Phosphorus, nitrogen and algae in Lake Washington after diversion of sewage. Science 169:690-691. Edwards, C., Beattie, K. A., Scrimgeour, C. M., and Codd, G. A. 1992. Identification of anatoxin-a in benthic cyanobacteria (blue-green algae) and in associated dog poisonings at Loch Insh, Scotland. Toxicon 30 (10):1165-1175. Ekman-Ekebom, M., Kauppi, M., Sivonen, K., Niemi, M., and Lepisto, L. 1992. Toxic cyanobacteria in some Finnish lakes. Environmental Toxicology and Water Quality 7 (2):201-213. Engstrom-Ost, J., Lehtiniemi, M., Green, S., Kozlowsky-Suzuki, B., and Viitasalo, M. 2002. Does cyanobacterial toxin accumulate in mysid shrimps and fish via copepods? Journal of Experimental Marine Biology and Ecology 276:95-107. Falconer, I. R. 1998. Algal toxins and human health. Handbook of Environmental Chemistry 5 (Pt. C):53-82. ———. 1999. An overview of problems caused by toxic blue-green algae (cyanobacteria) in drinking and recreactional water. Environmental Toxicology 14:5-12. ———. 2001. Toxic cyanobacterial bloom problems in Australian waters: risks and impacts on human health. Phycologia 40 (3):228-233.

54 Falconer, I. R., and Humpage, A. R. 1996. Tumour promotion by cyanobacterial toxins. Phycologia 35 (6):74-79. Fawell, J. K., Mitchell, R. E., Hill, R. E., and Everett, D. J. 1999. The toxicity of cyanobacterial toxins in the mouse: II Anatoxin-a. Human & Experimental Toxicology 18 (3):168-173. Fay, P., and Van Baalen, C., eds. 1987. The Cyanobacteria. New York, USA: Elsevier. 560 p. Ferber, L. R., Levine, S. N., Lini, A., and Livingston, G. P. 2004. Do cyanobacteria dominate in eutrophic lakes because they fix atmospheric nitrogen? Freshwater Biology 49:690-708. FerrAo-Filho, A. S., Azevedo, S. M. F. O., and DeMott, W. R. 2000. Effects of toxic and non-toxic cyanobacteria on the life history of tropical and temerate cladocerans. Freshwater Biology 45 (1):1-19. Fitzgeorge, R. B., Clark, S. A., and Keevil, C. W. 1994. Routes of intoxification. In Detection Methods for Cyanobacterial Toxins, edited by G. A. Codd, T. M. Jefferies, C. W. Keevil and E. Potter. Cambridge, UK: The Royal Sociesty of Chemistry. p69-74. Francis, G. 1878. Poisonous Australian lake. Nature 18:11-12. Fromme, H., Kohler, A., Krause, R., and Fuhrling, D. 2000. Occurrence of cyanobacterial toxins- microcystins and anatoxin-a-in Berlin water bodies with implications to human health and regulations. Environmental Toxicology 15 (2):120-130. Fuller, K., and Shear, H., eds. 1995. The Great Lakes - An Environmental Atlas and Resource Book. Third ed. Washington, DC and Downsview, Ontario: United State Environmental Protection Agency and the Government of Canada. 46 p. Furey, A., Crowley, J., Hamilton, B., Lehane, M., and James Kevin, J. 2005. Strategies to avoid the mis-identification of anatoxin-a using mass spectrometry in the forensic investigation of acute neurotoxic poisoning. Journal of Chromatography A 1082:91-97. Furey, A., Crowley, J., Lehane, M., and James, K. J. 2003. Liquid chromatography with electrospray ion-trap mass spectrometry for the determination of anatoxins in

55 cyanobacteria and drinking water. Rapid Communications in Mass Spectrometry 17 (6):583-588. Gallon, J. R., Chit, K. N., and Brown, E. G. 1990. Biosynthesis of the tropane-related cyanobacterial toxin anatoxin a: role of ornithine decarboxylase. Phytochemistry 29 (4):1107-1111. Gallon, J. R., Kittakoop, P., and Brown, E. G. 1994. Biosynthesis of anatoxin-a by Anabaena flos-aquae: examination of primary enzymic steps. Phytochemistry 35 (5):1195-1203. Garcia-Pichel, F. 1998. Solar ultraviolet and the evolutionary history of cyanobacteria. Origins of life and evolution of the biosphere 28:321-347. Ghassempour, A., Najafi, N. M., Mehdinia, A., Davarani, S. S. H., Fallahi, M., and Nakhshab, M. 2005. Analysis of anatoxin-a using polyaniline as a sorbent in solid-phase microextraction coupled to gas chromatography-mass spectrometry. Journal of Chromatography A 1078:120-127. Gorham, P. R., MacLachlan, J., Hammer, U. T., and Kim, W. K. 1964. Isolation and culture of toxic strains of Anabaena flos-aquae (Lingb.). Verhandlungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 15:796-804. Gugger, M., Lenoir, S., Berger, C., Ledreux, A., Druart, J.-C., Humbert, J.-F., Guette, C., and Bernard, C. 2005. First report in a river in France of the benthic cyanobacterium Phormidium favosum producing anatoxin-a associated with dog neurotoxicosis. Toxicon 45 (7):919-928. Gunn, G. J., Rafferty, A. G., Rafferty, G. C., Cockburn, N., Edwards, C., Beattie, K. A., and Codd, G. A. 1992. Fatal canine neurotoxicosis attributed to blue-green algae (cyanobacteria). The Veterinary Record 130 (301-302). Gupta, N., Bhaskar, A. S. B., and Rao, P. V. L. 2002. Growth characteristics and toxin production in batch cultures of Anabaena flos-aquae: effects of culture media and duration. World Journal of Microbiology & Biotechnology 18 (1):29-35. Haider, S., Naithani, V., Viswanathan, P. N., and Kakkar, P. 2003. Cyanobacterial toxins: a growing environmental concern. Chemosphere 52:1-21.

56 Hamill, K. D. 2001. Toxicity in benthic freshwater cyanobacteria (blue-green algae): first observation in New Zealand. New Zealand Journal of Marine and Freshwater Research 35:1057-1059. Harada, K.-I., Kondo, F., and Lawton, L. 1999. Laboratory analysis of cyanotoxins. In Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring and Management, edited by I. Chorus and J. Bartram. London: E & FN Spon (for the World Health Organization). p369-405. Harada, K., Nagai, H., Kimura, Y., Suzuki, M., Park, H. D., Watanabe, M. F., Luukkainen, R., Sivonen, K., and Carmichael, W. W. 1993. Liquid chromatography/mass spectrometric detection of anatoxin-a, a neurotoxin from cyanobacteria. Tetrahedron 49 (41):9251-9260. Harada, K., Ogawa, K., Kimura, Y., Murata, H., Suzuki, M., Thorn, P. M., Evans, W. R., and Carmichael, W. W. 1991. Microcystins from Anabaena flos-aquae NRC 525- 17. Chemical Research in Toxicology 4 (5):535-540. Hasler, A. D. 1969. Cultural eutrophication is reversible. Bioscience 19:425-431. Hawkins, P. R., Runnegar, M. T. C., Jackson, A. R. B., and Falconer, I. R. 1985. Severe hepatotoxicity caused by the tropical cyanobacterium Cylindrospermopsis raciborskii (Woloszynska) Seenaya and Subba Raju isolated from a domestic water supply reservoir. Applied and Environmental Microbiology 50:1292-1295. Hemscheidt, T., Rapala, J., Sivonen, K., and Skulberg, O. M. 1995. Biosynthesis of anatoxin-a in Anabaena flos-aquae and homoanatoxin-a in Oscillatoria formosa. Journal of the Chemical Society, Chemical Communications (13):1361-1362. Himberg, K. 1989. Determination of anatoxin-a, the neurotoxin of Anabaena flos-aquae cyanobacterium, in algae and water by gas chromatography-mass spectrometry. Journal of Chromatography 481:358-362. Holzgrabe, U. 1994. Anatoxin-a(s). A bacterial organophosphate. Pharmazie in Unserer Zeit 23 (5):301-302. Huber, C. S. 1972. The crystal structure and absolute configuration of 2,9-diacetyl-9- azabicyclo[4,2,1]non-2,3-ene. Acta Crystallographica B28:2577-2582. Humpage, A. R., Hardy, S. J., Moore, E. J., Froscio, S. M., and Falconer, I. R. 2000. Microcystins (cyanobacterial toxins) in drinking water enhance the growth of

57 aberrant crypt foci in the mouse colon. Journal of Toxicology and Environmental Health Part A 61 (3):155-165. Izaguirre, G., Hwang, C. J., Krasner, S. W., and McGuire, M. J. 1982. Geosmin and 2- methylisoborneol from cyanobacteria in three water supply systems. Applied and Environmental Microbiology 43 (3):708-714. Jackson, A. R. B., Mclnnes, A., Falconer, I. R., and Runnegar, M. T. C. 1984. Clinical and pathological changes in sheep experimentally poisoned by the blue-green alga Microcystis aeruginosa. Veterinary Pathology 21:102-113. James, K. J., Crowley, J., Hamilton, B., Lehane, M., Skulberg, O., and Furey, A. 2005. Anatoxins and degradation products, determined using hybrid quadrupole time- of-flight and quadrupole ion-trap mass spectrometry: Forensic investigations of cyanobacterial neurotoxin poisoning. Rapid Communications in Mass Spectrometry 19 (9):1167-1175. James, K. J., and Sherlock, I. R. 1996. Determination of the cyanobacterial neurotoxin, anatoxin-a, by derivatization using 7-fluoro-4-nitro-2,1,3-benzoxadiazole (NBD-F) and HPLC analysis with fluorimetric detection. Biomedical Chromatography 10 (1):46-47. James, K. J., Sherlock, I. R., and Stack, M. A. 1997. Anatoxin-a in Irish freshwater and cyanobacteria, determined using a new fluorometric liquid chromatographic method. Toxicon 35 (6):963-971. Jang, M.-H., Ha, K., Joo, G.-J., and Takamura, N. 2003. Toxin production of cyanobacteria is increased by exposure to zooplankton. Freshwater Biology 48 (9):1540-1550. Jefferies, T. M., Brammer, G., Zotou, A., Brough, P. A., and Gallagher, T. 1994. Determination of anatoxin-a, homoanatoxin and propylanatoxin in cyanobacterial extracts by HPLC, GC-mass spectrometry and capillary electrophoresis. In Detection Methods for Cyanobacterial Toxins, edited by G. A. Codd, T. M. Jefferies, C. W. Keevil and E. Potter. Cambridge: The Royal Society of Chemistry. p34-39. Kaas, H., and Henriksen, P. 2000. Saxitoxins (PSP toxins) in Danish lakes. Water Research 34 (7):2089-2097.

58 Karjalainen, M., Reinikainen, M., Lindvall, F., Spoof, L., and Meriluoto, J. A. O. 2003. Uptake and accumulation of dissolved, radiolabeled nodularin in Baltic Sea zooplankton. Environmental Toxicology 18 (1):52-60. Karjalainen, M., Reinikainen, M., Spoof, L., Meriluoto, J. A. O., Sivonen, K., and Viitasalo, M. 2005. Trophic transfer of cyanobacterial toxins from zooplankton to planktivores: Consequences for pike larvae and mysid shrimps. Environmental Toxicology 20 (3):354-362. Kearns, K. D., and Hunter, M. D. 2000. Green algal extracellular products regulate antialgal toxin production in a cyanobacterium. Environmental Microbiology 2 (3):291-297. ———. 2001. Toxin-producing Anabaena flos-aquae induces settling of Chlamydomonas reinhardtii, a competing motile alga. Microbial Ecology 42 (1):80-86. Kenefick, S. L., Hrudey, S. E., Prepas, E. E., Motkosky, N., and Peterson, H. G. 1992. Odorous substances and cyanobacterial toxins in prairie drinking water sources. Water Science and Technology 25 (2):147-154. Krienitz, L., Ballot, A., Kotut, K., Wiegand, C., Putz, S., Metcalf, J. S., Codd, G. A., and Pflugmacher, S. 2003. Contribution of hot spring cyanobacteria to the mysterious deaths of Lesser Flamingos at Lake Bogoria, Kenya. FEMS Microbiology Ecology 43 (2):141-148. Kuiper-Goodman, T., Falconer, I., and Fitzgerald, J. 1999. Human health aspects. In Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring, and Management, edited by I. Chorus and J. Bartram. London: E & FN Spon (for the World Health Organization). p113-153. Lampert, W. 1981. Toxicity of the blue-green Microcystis aeruginosa: effective defense mechanism against grazing pressure by Daphnia. Verhandlungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 21:1436-1440. Lukac, M., and Aegerter, R. 1993. Influence of trace metals on growth and toxin production of Microcystis aeruginosa. Toxicon 31 (3):293-305.

59 Lyck, S., Gjolme, N., and Utkilen, H. 1996. Iron starvation increases toxicity of Microcystis aeruginosa CYA 228/1 (Chroococcales, Cyanophyceae). Phycologia 35 (6):120-124. MacPhail, R. C., and Jarema, K. A. 2005. Prospects on behavioral studies of marine and freshwater toxins. Neurotoxicology and Teratology 27:695-699. Mahmood, N. A., and Carmichael, W. W. 1986. The pharmacology of anatoxin-a(s), a neurotoxin produced by the freshwater cyanobacterium Anabaena flos-aquae NRC 525-17. Toxicon 24 (5):425-434. ———. 1987. Anatoxin-a(s), an anticholinesterase from the cyanobacterium Anabaena flos-aquae NRC-525-17. Toxicon 25 (11):1221-1227. Maizels, M., and Budde, W. L. 2004. A LC/MS method for the determination of cyanobacteria toxins in water. Analytical Chemistry 76 (5):1342-1351. Makarewicz, J. C., Boyer, G. L., Guenther, W., Arnold, M., and Lewis, T. W. 2006. The occurrence of cyanotoxins in the nearshore and coastal embayments of Lake Ontario. Journal of Great Lakes Research (submitted). Marsalek. 1995. Toxins in cyanobacterial water blooms - general overview. Vodni Hospodarstvi 45 (6-7):183-184. Matsunaga, H., Harada, K.-I., Senma, M., Ito, Y., Yasuda, N., Ushida, S., and Kimura, Y. 1999. Possible cause of unnatural mass death of wild birds in a pond in Nishinomiya, Japan: sudden appearance of toxic cyanobacteria. Natural Toxins 7:81-84. Matsunaga, S., Moore, R. E., Niemczura, W. P., and Carmichael, W. W. 1989. Anatoxin- a(s), a potent anticholinesterase from Anabaena flos-aquae. Tennen Yuki Kagobutsu Toronkai Koen Yoshishu 31st:252-259. Mazur, H., Lewandowska, J., Blaszczyka, A., Kot, A., and Plinski, M. 2003. Cyanobacterial toxins in fresh and brackish waters of Pomorskie Province (northern Poland). Oceanological and Hydrobiological Studies 32 (1):15-26. McGucken, W. 2000. Lake Erie Rehabilitated. Controlling Cultural Eutrophication, 1960s-1990s. Edited by J. Stine and J. Tarr. 1st ed, Technology and the Environment. Akron, Ohio: The University of Akron Press, 318 p.

60 Metcalf, J. S., Bell, S. G., and Codd, G. A. 2001. Colorimetric immuno-protein phosphatase inhibition assay for specific detection of microcystins and nodularins of cyanobacteria. Applied and Environmental Microbiology 67 (2):904-909. Mez, K., Beattie, K. A., Codd, G. A., Hanselmann, K., Hauser, B., Naegeli, H., and Preisig, H. R. 1997. Identification of a microcystin in benthic cyanobacteria linked to cattle deaths on alpine pastures in Switzerland. European Journal of Phycology 32:111-117. Mez, K., Hanselmann, K., and Preisig, H. R. 1998. Environmental conditions in high mountain lakes containing toxic benthic cyanobacteria. Hydrobiologia 368:1-15. MHNZ. 2000. Drinking-water standards for New Zealand 2000. Wellington, New Zealand: Ministry of Health, 145 p. Mihuc, T. B., Boyer, G. L., Satchwell, M. F., Pellam, M., Jones, J., Vasile, J., Bouchard, A., and Rob, B. 2005. 2002 phytoplankton community composition and cyanobacterial toxins in Lake Champlain, U.S.A. Verhandlungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 29:328-333. Mohamed, Z. A. 2001. Accumulation of cyanobacterial hepatotoxins by Daphnia in some Egyptian irrigation canals. Ecotoxicology and Environmental Safety 50:4-8. Moss, G. 1988. Ecology of Fresh Waters. Oxford: Blackwell, 417 p. Muenscher, W. G. 1930. Plankton studies in Lake Champlain Watershed. In A Biological Survey of the Champlain Watershed: Supplemental to Nineteenth Annual Report, 1929. Albany: New York Conservation Department. p164-185. Namera, A., So, A., and Pawliszyn, J. 2002. Analysis of anatoxin-a in aqueous samples by solid-phase microextraction coupled to high-performance liquid chromatography with fluorescence detection and on-fiber derivatization. Journal of Chromatography, A 963 (1-2):295-302. Namikoshi, M., Murakami, T., Fujiwara, T., Nagai, H., Niki, T., Harigaya, E., Watanabe Mariyo, F., Oda, T., Yamada, J., and Tsujimura, S. 2004. Biosynthesis and transformation of homoanatoxin-a in the cyanobacterium Raphidiopsis mediterranea Skuja and structures of three new homologues. Chemical Research in Toxicology 17 (12):1692-1696.

61 Namikoshi, M., Murakami, T., Watanabe, M. F., Oda, T., Yamada, J., Tsujimura, S., Nagai, H., and Oishi, S. 2003. Simultaneous production of homoanatoxin-a, anatoxin-a, and a new non-toxic 4-hydroxyhomoanatoxin-a by the cyanobacterium Raphidiopsis mediterranea Skuja. Toxicon 42 (5):533-538. Negri, A. P., Jones, G. J., and Hindmarsh, M. 1995. Sheep mortality associated with paralytic shellfish poisons from the cyanobacterium Anabaena circinalis. Toxicon 33 (10):1321-1329. Nicholla, K. H., Hopkins, G. J., Standke, S. J., and Nakamoto, L. 2001. Trends in total phosphorus in Canadian near-shore waters of the Laurentian Great Lakes: 1976- 1999. Journal of Great Lakes Research 27 (4):402-422. Nizan, S., Dimentman, C., and Shilo, M. 1986. Acute toxic effects of the cyanobacterium Microcystis aeruginosa on Daphnia magna. Limnology and Oceanography 31:497-502. Oberemm, A., Becker, J., Codd, G. A., and Steinberg, C. 1999. Effects of cyanobacterial toxins and aqueous crude extracts of cyanobacteria on the development of fish and amphibians. Environmental Toxicology 14 (1):77-88. Oberholster, P. J., Botha, A.-M., and Brobbelaar, J. U. 2004. Microcystis aeruginosa: source of toxic microcystins in drinking water. African Journal of Biotechnology 3 (3):159-168. Ohta, T., Sueoka, E., Iida, N., Komori, A., Suganuma, M., Nishiwaki, R., Tatematsu, M., Kim, S.-J., Carmichael, W. W., and Fujiki, H. 1994. Nodularin, a potent inhibitor of protein phosphatases 1 and 2A, is a new environmental carcinogen in male F344 rat liver. Cancer Research 54:6402-6406. Ojanpera, I., Vuori, E., Himberg, K., Waris, M., and Niinivaara, K. 1992. Screening of algal blooms for anatoxin-a by thin-layer chromatography and Fast Black K salt. Recent Adv. Toxinol. Res., [World Conf. Anim., Plant Microb. Toxin], 10th 3:337- 343. Paerl, H. W., and Millie, D. F. 1996. Physiological ecology of toxic aquatic cyanobacteria. Phycologia 35 (6):160-167. Park, H. D., Watanabe, M. F., Harada, K., Nagai, H., Suzuki, M., Watanabe, M., and Hayashi, H. 1993. Hepatotoxin (microcystin) and neurotoxin (anatoxin-a)

62 contained in natural blooms and strains of cyanobacteria from Japanese freshwaters. Natural Toxins 1 (6):353-360. Persson, P.-E. 1996. Cyanobacteria and off-flavor. Phycologia 35 (6):168-171. Pietsch, C., Wiegand, C., Ame, M. V., Nicklisch, A., Wunderlin, D., and Pflugmacher, S. 2001a. The effects of a cyanobacterial crude extract on different aquatic organisms: Evidence for cyanobacterial toxin modulating factors. Environmental Toxicology 16 (6):535-542. Pietsch, J., Fichtner, S., Imhof, L., and Schmidt, W. 2001b. Analytical determination of algae toxins (hepato- and neurotoxins) by HPLC/MS/MS. Biologische Abwasserreinigung 16:35-50. Plinski, M., and Jozwiak, T. 1999. Temperature and N:P ratio as factors causing blooms of blue-green algae in the Gulf of Gdansk. Oceanologia 41 (1):73-80. Pouria, S., de Andrade, A., Barbosa, J., Cavalcanti, R. L., Barreto, V. T. S., Ward, C. J., Preiser, W., Poon, G. K., Neild, G. H., and Codd, G. A. 1998. Fatal microcystin intoxication in haemodialysis unit in Caruaru, Brazil. The Lancet 352:21-26. Powell, M. W. 1997. Analysis of anatoxin-a in aqueous samples. Chromatographia 45:25-28. Rao, P. V. L., Gupta, N., Bhaskar, A. S. B., and Jayaraj, R. 2002. Toxins and bioactive compounds from cyanobacteria and their implications on human health. Journal of Environmental Biology 23 (3):215-224. Rapala, J., Lahti, K., Sivonen, K., and Niemela, S. I. 1994. Biodegradability and adsorption on lake sediments of cyanobacterial hepatotoxins and anatoxin-a. Letters in Applied Microbiology 19 (6):423-428. Rapala, J., and Sivonen, K. 1998. Assessment of environmental conditions that favor hepatotoxic and neurotoxic Anabaena spp. strains cultured under light limitation at different temperatures. Microbial Ecology 36 (2):181-192. Rapala, J., Sivonen, K., Luukkainen, R., and Niemela, S. I. 1993. Anatoxin-a concentration in Anabaena and Aphanizomenon under different environmental conditions and comparison of growth by toxic and non-toxic Anabaena-strains - a laboratory study. Journal of Applied Phycology 5 (6):581-591.

63 Reinikainen, M., Lindvall, F., Meriluoto, J. A. O., Repka, S., Sivonen, K., Spoof, L., and Wahlsten, M. 2002. Effects of dissolved cyanobacterial toxins on the survival and egg hatching of estuarine calanoid copepods. Marine Biology (Berlin, Germany) 140 (3):577-583. Reynolds, C. S., Oliver, R. L., and Walsby, A. E. 1987. Cyanobacterial dominance - the role of buoyancy regulation in dynamic lake environments. New Zealand Journal of Marine and Freshwater Research 21:379-390. Reynolds, K. A. 2004. Cyanobacteria: natural organisms with toxic effects. Water conditioning & Purification 46 (3):78-80. Rinta-Kanto, J. M., Ouellette, A. J. A., Boyer, G. L., Twiss, M. R., Bridgeman, T. B., and Wilhelm, S. W. 2005. Quantification of toxic Microcystis spp. during the 2003 and 2004 blooms in western Lake Erie using quantitative real-time PCR. Environmental Science and Technology 39:4198-4205. Rogers, E. H., Hunter III, E. S., Moser, V. C., Phillips, P. M., Herkovits, J., Munoz, L., Hall, L. L., and Chernoff, N. 2005. Potential development toxicity of anatoxin-a, a cyanobacterial toxin. Journal of Applied Toxicology 25:527-534. Ruseva, E., Pietsch, J., Fichtner, S., Imhof, L., and Schmidt, W. 2003. Determination of cyanobacterial hepato- and neurotoxins in water samples by HPLC-ESI-MS-MS. Chimia 57 (1/2):32. Schlegel, I., Doan, N. T., de Chazal, N., and Smith, G. D. 1999. Antibiotic activity of new cyanobacterial isolates from Australia and Asia against green algae and cyanobacteria. Journal of Applied Phycology 10:471-479. Selwood, A. I., Holland, P. T., Wood, S. A., Smith, K. F., and McNabb, P. S. 2006. Production of anatoxin-a and a novel biosynthetic precursor by the cyanobacterium Aphanizomenon issatschenkoi. Environmental Science & Technology: In print. Shambaugh, A., Duchovnay, A., and McIntosh, A. 1999. A survey of Lake Champlain's plankton. In Lake Champlain in Transition: from Research toward Restoration, edited by T. O. Manley and P. L. Manley. Washinton, DC: Americal Geophysical Union. p41-67.

64 Shanley, J. B., and Denner, J. C. 1999. The hydrology of the Lake Champlain basin. In Lake Champlain in Transition: from Research toward Restoration, edited by T. O. Manley and P. L. Manley. Washington, DC: Americal Geophysical Union. p41- 67. Shen, X., Lam, P. K. S., Shaw, G. R., and Wickramasinghe, W. 2002. Genotoxicity investigation of a cyanobacterial toxin, cylindrospermopsin. Toxicon 40:1499- 1501. Sipia, V. O., Kankaanpaa, H. T., Pflugmacher, S., Flinkman, J., Furey, A., and James, K. J. 2002. Bioaccumulation and detoxication of nodularin in tissues of flounder (Platichthys flesus), mussels (Mytilus edulis, Dreissena polymorpha), and clams (Macoma balthica) from the northern Baltic Sea. Ecotoxicology and Environmental Safety 53:305-311. Sivonen, K. 1990. Effects of light, temperature, nitrate, orthophosphate, and bacteria on growth of and hepatotoxin production by Oscillatoria agardhii strains. Applied and Environmental Microbiology 56 (9):2658-2666. ———. 1996. Cyanobacterial toxins and toxin production. Phycologia 35 (6):12-24. ———. 2000. Freshwater cyanobacterial neurotoxins: Ecobiology, chemistry, and detection. Food Science and Technology (New York, NY, United States) 103:567- 581. Sivonen, K., Himberg, K., Luukkainen, R., Niemela, S. I., Poon, G. K., and Codd, G. A. 1989. Preliminary characterization of neurotoxic cyanobacteria blooms and strains from Finland. Toxicity Assessment 4 (3):339-352. Sivonen, K., Niemela, S. I., Niemi, R. M., Lepisto, L., Luoma, T. H., and Rasanen, L. A. 1990. Toxic cyanobacteria (blue-green algae) in Finnish fresh and coastal waters. Hydrobiologia 190:267-275. Skulberg, O. M., Carmichael, W. W., Andersen, R. A., Matsunaga, S., Moore, R. E., and Skulberg, R. 1992. Investigations of a neurotoxic oscillatorialean strain (Cyanophyceae) and its toxin. Isolation and characterization of homoanatoxin-a. Environmental Toxicology and Chemistry 11 (3):321-329.

65 Skulberg, O. M., Carmichael, W. W., Codd Geoffrey, A., and Skulberg, R. 1993. Taxonomy of toxic Cyanophyceae (Cyanobacteria). In Algal Toxins in Seafood and Drinking Water, edited by I. R. Falconer. London: Academic Press. p145-164. Skulberg, O. M., Codd Geoffrey, A., and Carmichael, W. W. 1984. Toxic blue-green algal blooms in Europe: a growing problem. Ambio 13:244-247. Smith, C., and Sutton, A. 1993. The persistence of anatoxin-a in reservoir water. UK Report No. FR0427. Foundation for Water Research. Smith, R. A., and Lewis, D. 1987. A rapid analysis of water for anatoxin-a, the unstable toxic alkaloid from Anabaena flos-aquae, the stable non-toxic alkaloids left after bioreduction and a related amine which may be nature's precursor to anatoxin-a. Veterinary and Human Toxicology 29 (2):153-154. Spivak, C. E., Witkop, B., and Albuquerque, E. X. 1980. Anatoxin-a: a novel, potent agonist at the nicotinic receptor. Molecular Pharmacology 18 (3):384-394. Srivastava, V. C., Manderson, G. J., and Bhamidimarri, R. 1998. Inhibitory metabolite production by the cyanobacterium Fischerella muscicola. Microbiological Research 153:309-317. Stephenson, J. B. 2003. Great Lakes - An Overall Strategy and Indicators for Measuring Progress Are Needed to Better Achieve Restoration Goals. Report for Congressional Request. GAO-03-505. United States General Accounting Office, 90 p. Stevens, D. K., and Krieger, R. I. 1988. Analysis of anatoxin-a by GC/ECD. Journal of Analytical Toxicology 12 (3):126-131. ———. 1991a. Effect of route of exposure and repeated doses on the acute toxicity in mice of the cyanobacterial nicotinic alkaloid anatoxin-a. Toxicon 29 (1):134-138. ———. 1991b. Stability studies on the cyanobacterial nicotinic alkaloid anatoxin-a. Toxicon 29 (2):167-179. Swanson, K. L., Allen, C. N., Aronstam, R. S., Rapoport, H., and Albuquerque, E. X. 1986. Molecular mechanisms of the potent and stereospecific nicotinic receptor agonist (+)-anatoxin-a. Molecular Pharmacology 29 (3):250-257. Swanson, K. L., Aracava, Y., Sardina, F. J., Rapoport, H., Aronstam, R. S., and Albuquerque, E. X. 1989. N-Methylanatoxinol isomers: derivatives of the agonist

66 (+)-anatoxin-a block the nicotinic acetylcholine receptor ion channel. Molecular Pharmacology 35 (2):223-231. Swanson, K. L., Aronstam, R. S., Wonnacott, S., Rapoport, H., and Albuquerque, E. X. 1991. Nicotinic pharmacology of anatoxin analogs. I. Side chain structure-activity relationships at peripheral agonist and noncompetitive antagonist sites. Journal of Pharmacology and Experimental Therapeutics 259 (1):377-386. Swanson, K. L., Rapoport, H., Aronstam, R. S., and Albuquerque, E. X. 1990. Nicotinic acetylcholine receptor function studied with synthetic (+)-anatoxin-a and derivatives. ACS Symposium Series 418:107-118. Takino, M., Daishima, S., and Yamaguchi, K. 1999. Analysis of anatoxin-a in freshwaters by automated on-line derivatization-liquid chromatography- electrospray mass spectrometry. Journal of Chromatography, A 862 (2):191-197. Thomas, P., Brough, P. A., Gallagher, T., and Wonnacott, S. 1994. Alkyl-modified side chain variants of anatoxin-a: a series of potent nicotinic agonists. Drug Development Research 31 (2):147-156. Thomas, P., Stephens, M., Wilkie, G., Amar, M., Lunt, G. G., Whiting, P., Gallagher, T., Pereira, E., Alkondon, M., and et al. 1993. (+)-Anatoxin-a is a potent agonist at neuronal nicotinic acetylcholine receptors. Journal of Neurochemistry 60 (6):2308-2311. Thompson, P. E., Manallack, D. T., Blaney, F. E., and Gallagher, T. 1992. Conformational studies on (+)-anatoxin-a and derivatives. Journal of Computer- Aided Molecular Design 6 (3):287-298. USEPA. 2001. Creating a cyanotoxin target list for the unregulated contaminant monitoring rule. 17 p. ———. 2005. Paper read at International Symposium on Cyanobacterial Harmful Algal Blooms (ISOC-HAB), September 6-10, at Durham, NC. Usleber, E., Dietrich, R., Burk, C., Schneider, E., and Martlbauer, E. 2001. Immunoassay methods for paralytic shellfish poisoning toxins. Journal of AOAC International 84 (5):1649-1656. Vallentyne, J. R., Johnson, W. E., and Harris, A. J. 1970. A visual demonstration of the beneficial effects of sewage treatment for phosphate removal on particulate matter

67 in waters of Lakes Erie and Ontario. Journal of the Fisheries Research Board of Canada 27:1493-1496. van den Hoek, C., Mann, D. G., and Jahns, H. M., eds. 1995. Algae. An Introduction to Phycology. Cambridge: Cambridge University Press. 637 p. Van der Westhuizen, A. J., and Eloff, J. N. 1983. Effect of culture age and pH of culture medium on the growth and toxicity of the blue-green alga Microcystis aeruginosa. Zeitschrift fur Pflanzenphysiologie 110:157-163. Vanderploeg, H. A., Liebig, J. R., Carmichael, W. W., Agy, M. A., Johengen, T. H., Fahnenstiel, G. L., and Nalepa, T. F. 2001. Zebra mussel (Dreissena polymorpha) selective filtration promoted toxic Microcystis blooms in Saginaw Bay (Lake Huron) and Lake Erie. Canadian Journal of Fisheries and Aquatic Sciences 58:1208-1221. Vezie, C., Benoufella, F., Sivonen, K., Bertru, G., and Laplanche, A. 1996. Detection of toxicity of cyanobacterial strains using Artemia salina and Microtox assays compared with mouse bioassay results. Phycologia 35 (6):198-202. Viaggiu, E., Calvanella, S., Mattioli, P., Albertano, P., Melchiorre, S., and Bruno, M. 2003. Toxic blooms of Planktothrix rubescens (Cyanobacteria/Phormidiaceae) in three waterbodies in Italy. Archiv fuer Hydrobiologie, Supplement 148:569-577. Viaggiu, E., Melchiorre, S., Volpi, F., Di Corcia, A., Mancini, R., Garibaldi, L., Crichigno, G., and Bruno, M. 2004. Anatoxin-a toxin in the cyanobacterium Planktothrix rubescens from a fishing pond in northern Italy. Environmental Toxicology 19 (3):191-197. Wang, H., and Zhu, H. 1995. Toxicity and pollution of freshwater phytoplankton. Shanghai Huanjing Kexue 14 (8):38-41. Watanabe, M. M., Zhang, X., and Kaya, K. 1996. Fate of toxic cyclic heptapeptides, microcystins, in toxic cyanobacteria upon grazing by the mixotrophic flagellate Poterioochromonas malhamensis (Ochromonadles, Chrysophyceae). Phycologia 35 (6):203-206. Watzin, M. C., Miller, E. B., Shambaugh, A. D., and Kreider, M. A. 2005. Applicaiton of the WHO alert level framework to cyanobacterial monitoring of Lake Champlain, Vermont. Environmental Toxicology 21:278-288.

68 WHO. 2003. Guidelines for Safe Recreational Water Environments. Volume 1: Coastal and Fresh Waters. Geneva: World Health Organization, 219 p. ———. 2004. Guidelines for Drinking Water Quality. 3rd ed. Geneva: WHO, 515 p. Wiegand, C., and Pflugmacher, S. 2005. Ecotoxicological effects of selected cyanobacterial secondary metabolites a short review. Toxicology and Applied Pharmacology 203:201-218. Willen, T., and Mattsson, R. 1997. Water-blooming and toxin-producing cyanobacteria in Swedish fresh and brackish waters, 1981-1995. Hydrobiologia 353:181-192. Wnorowski, A. U. 1992. Tastes and odours in the aquatic environment: A review. Water South Africa 18 (3):203-214. Wolf, H. U., and Frank, C. 2002. Toxicity assessment of cyanobacterial toxin mixtures. Environmental Toxicology 17 (4):395-399. Wonnacott, S., Jackman, S., Swanson, K. L., Rapoport, H., and Albuquerque, E. X. 1991. Nicotinic pharmacology of anatoxin analogs. II. Side chain structure-activity relationships at neuronal nicotinic ligand binding sites. Journal of Pharmacology and Experimental Therapeutics 259 (1):387-391. Wonnacott, S., Swanson, K. L., Albuquerque, E. X., Huby, N. J. S., Thompson, P., and Gallagher, T. 1992. Homoanatoxin: a potent analog of anatoxin-A. Biochemical Pharmacology 43 (3):419-423. Wu, J.-Y., Xu, Q.-J., Gao, G., and Shen, J.-H. 2006. Evaluating genotoxicity associated with microcystin-LR and its risk to source water safety in Meiliang Bay, Taihu Lake. Environmental Toxicology 21 (3):250-255. Xie, L., and Xie, P. 2002. Long-term (1956-1999) dynamics of phosphorus in a shallow, subtrophical Chinese lake with the possible effects of cyanobacterial blooms. Water Research 36:343-349. Xie, L., Xie, P., Guo, L., Li, L., Miyabara, Y., and Park, H. D. 2005. Organ distribution and bioaccumulation of microcystins in freshwater fish at different trophic levels from the eutrophic Lake Chaohu, China. Environmental Toxicology 20 (3):293- 300.

69 Yamamoto, Y., and Nakahara, H. 2005. The formation and degradation of cyanobacterium Aphanizomenon flos-aquae blooms: the importance of pH, water temperature, and day length. Limnology 6:1-6. Yang, X., Satchwell, M. F., and Boyer, G. L. 2001. The identification of anatoxin-a from a toxic blue-green algae bloom in Lake Champlain, USA. Abstract. In Fifth International Conference on Toxic Cyanobacteria, Moosa Lakes, Queensland, Australia. July 15-20. 1 p. Yu, S.-Z. 1989. Drinking water and primary liver cancer. In Primary Liver Cancer, edited by Z. Y. Tang, M. C. Wu and S. S. Xia. New York: China Academic Publishers. p30-37. ———. 1995. Primary prevention of hepatocellular carcinoma. Journal of Gastroenterology and Hepatology 10:671-682. Zhang, Z.-h., Zheng, L.-x., Qu, W.-d., and Zhu, H.-g. 2003. Distribution state of microcystin-LR and anatoxin-A in Dianshan Lake. Zhongguo Huanjing Kexue 23 (4):403-406. Zotou, A., Jeffries, T. M., Brough, P. A., and Gallagher, T. 1993. Determination of anatoxin-a and homoanatoxin in blue-green algal extracts by high-performance liquid chromatography and gas chromatography-mass spectrometry. Analyst (Cambridge, United Kingdom) 118 (7):753-758.

70 Chapter 3. Materials and methods

3.1. Preparation of anatoxin-a standards

3.1.1. Preparation of anatoxin-a standards for toxin analysis.

Commercially available anatoxin-a fumarate (BIOMOL International, L. P., USA) was used as the standard for both high performance liquid chromatography with fluorescence detection (HPLD-FD) and liquid chromatography mass spectrometry

(LCMS) analysis. One mg of anatoxin-a fumarate was dissolved in 50% methanol to make a 1 mg mL-1 stock solution. This 1 mg mL-1 solution was diluted 1:50 with 50% methanol to make a 20 μg mL-1 stock solution. The 20 μg mL-1 solution was used to generate a series of standard solutions for LCMS analysis. For HPLC-FD analysis, further dilution with 50% methanol was performed to make a 1 μg mL-1 anatoxin-a working standard solution, and this 1 μg mL-1 anatoxin-a working standard solution was used to create a series of standard solutions for analysis of anatoxin-a by HPLC-FD. The standard curves created by HPLC-FD and LCMS are shown in Figure 3.1 and 3.2.

Besides anatoxin-a, a natural amino acid, phenylalanine (Sigma-Aldrich, USA), was used as a reference compound by dissolving 1 mg of compound in 1 mL deionized water.

Phenylalanine was analyzed together with anatoxin-a standard by both HPLC-FD and

LCMS techniques and was separated from anatoxin-a in biological matrices using both techniques (Figure 3.3 and 3.4). Methyl pipecolinate hydrochloride (Sigma-Aldrich,

USA) was used as an internal standard. This internal standard (10 μL of a 5 μg mL-1 solution in 50% aqueous methanol) was spiked into both the anatoxin-a standards and field samples.

71 3.1.2. Synthesis of dihydroanatoxin-a and epoxyanatoxin-a.

Water samples were also analyzed for two known non-toxic degradation products of anatoxin-a: epoxyanatoxin-a and dihydroanatoxin-a. These compounds were synthesized from commercial anatoxin-a standard as a reference material for analysis using both HPLC-FD and LCMS. Both degradation products were prepared following the method described by James et al. (1998). For epoxyanatoxin-a, 140 μl of 1 mg mL-1 anatoxin-a solution was added to a 2 mL flat-bottomed amber reaction vial, dried under nitrogen and was re-dissolved in 200 μl of acetone. Thirty μl each of 2 M NaOH and

o 30% H2O2 were added into the vial. The vial was capped, placed in a 50 C water bath, and stirred for 45 min using a flea stir bar. After completion of the reaction, the solution was evaporated to dryness under nitrogen and the residue was reconstituted in 0.5 mL of

50% methanol in water. The product was centrifuged (10,000 g, 5 min) and 20 μl of the solution was examined using HPLC-FD and electrospray ionization mass spectrometry

(Figure 3.3 and Figure 3.5). The synthetic yield of epoxyanatoxin-a was approximately

95% (135 μg) and epoxyanatoxin-a was the sole product ([M+H]+ m/z 182.1).

Epoxyanatoxin-a has a different retention time than anatoxin-a during both HPLC-FD and LCMS analysis.

To prepare the dihydroanatoxin-a standard, 250 μl of 1 mg mL-1 of anatoxin-a fumarate standard was added to a 2-mL amber vial and evaporated to dryness under nitrogen. The residue was re-dissolved in 250 μl of glacial acidic acid and approximately

0.2 mg of palladium oxide/carbon powder (Sabather catalyst) added to the vial. The vial was capped with a septa cap and 2 needles pierced through the septa, one needle was attached to a hydrogen gas cylinder and the other acted as a vent tube. The solution

2.5e+5 r2 = 0.998 2.0e+5

1.5e+5

1.0e+5 Response

5.0e+4

0.0

-5.0e+4 0.00 0.05 0.10 0.15 0.20 0.25 Concentration (μg mL-1)

Figure 3.1. A standard curve of NBD-derivatized anatoxin-a by HPLC-FD (Ex 470 nm,

Em 530 nm). Data are presented as mean ± SD (n = 5). The response was obtained using a 20 μL injection and the following HPLC conditions at room temperature: Ace 5 μ C18 column, isocratic 55% aqueous acetonitrile at flow rate of 0.8 mL min-1.

1e+7 r2 = 0.974 8e+6

6e+6

4e+6 Response

2e+6

024681012 Anatoxin-a concentration (μg mL-1)

Figure 3.2. A standard curve for anatoxin-a analysis by LCMS. The response was obtained by extracting m/z 166 from the total ion chromatogram using a 20 μL injection and the following LCMS conditions: Ace 5 μ C18 column, 5-50% acetonitrile containing

0.1% TFA (trifluoroacetic acid) linear gradient in 20 min at flow rate 0.8 mL min-1.

55000

EpoxyATX 50000

45000

40000 ATX Response Phe 35000

IS DihydroATX 30000

25000 0 5 10 15 20 Time (min)

Figure 3.3. Sample HPLC-FD chromatograms of anatoxin-a (ATX), dihydroanatoxin-a

(dihydroATX) and epoxyanatoxin-a (epoxyATX) from an aqueous standard. The internal standard (IS) is methyl pipecolinate and Phe is phenylalanine. HPLC conditions are the same as described in Figure 3.1.

1.3e+6 120 a c. ATX

100 166 1.2e+6 ATX

1.1e+6 80

1.0e+6 60 Response % Abundance% 9.0e+5 40 135

8.0e+5 20 208

7.0e+5 0 0 2 4 6 8 101214 50 100 150 200 250 300

Retention time (min) m/z (mAU) 8e+6 120 Phe d. Phe 7e+6 b 166 100

6e+6

80 5e+6

4e+6 60

Response Impurity

3e+6 Abundance% 40 208

2e+6

20 1e+6 120

0 0 02468101214 50 100 150 200 250 300

Retention time (min) m/z (mAU)

Figure 3.4. Sample LCMS chromatograms and mass spectra for anatoxin-a (a) and phenylalanine (b). Anatoxin-a had a retention time of 10.5 min and phenylalanine had a retention time of 11.7 min under the experimental conditions. The mass spectra for anatoxin-a (c) and phenylalanine (d) are shown on the right. LCMS conditions are same as described in Figure 3.2.

1e+6 120 DihydroATX ATX DihydroATX 9e+5 168 100

8e+5 80

7e+5 Impurity 60 6e+5 Response

% Abundance 40 5e+5

20 4e+5 210

3e+5 0 0 2 4 6 8 101214 50 100 150 200 250 300

Retention time (min) m/z (mAU)

1.8e+6 120 EpoxyATX EpoxyATX 182 1.6e+6 100

1.4e+6 80

1.2e+6 60 Response

1.0e+6 % Abundance 40

8.0e+5 20

6.0e+5 0 02468101214 50 100 150 200 250 300

Retention time (min) m/z (mAU) Figure 3.5. Sample LCMS chromatographs of dihydroanatoxin-a (dihydroATX) and epoxyanatoxin-a (epoxyATX). Dihydroanatoxin-a had an [M+H]+ m/z 168 with retention time of 11.0 min and epoxyanatoxin-a had an m/z 182 with retention time of 9.7 min.

Anatoxin-a had a retention time of 10.5 min under the current conditions. The mass spectra of epoxyATX and dihydroATX are given in the right hand panels.

77 mixture was purged with hydrogen gas at room temperature for 2 hours with stirring using a flea stir bar. Once the reaction was complete, the solution was blown to dryness under nitrogen and the residue was reconstituted by addition of 0.5 mL 50% methanol in water. The reconstituted solution was centrifuged and the supernatant was transferred to a 1 mL vial and stored at –20oC. Ten μl of the synthetic product was diluted to 200 μL with 50% methanol and analyzed using both HPLC-FD and LCMS to evaluate the synthetic yield and the purity of the product. The yield and the recovery of dihydroanatoxin-a was more than 90% with trace amount of anatoxin-a remaining.

Dihydroanatoxin-a has a different retention time from anatoxin-a using both the HPLC-

FD and LCMS methods (Figure 3.3). It also can be distinguished from anatoxin-a by its

[M+H]+ ion peak at m/z 168.1 (Figure 3.5).

3.2. Analysis of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a using high performance liquid chromatography with fluorescence detection (HPLC-FD) and electrospray ionization mass spectrometry (LCMS)

Pre-analysis treatments of water samples are described in section 3.4. For analysis of anatoxin-a using HPLC-FD, blanks, anatoxin-a standards or processed water samples were brought to 100 μL using sample buffer (0.05 M pH 9.3 sodium borate buffer containing 50 μg internal standard, methyl pipecolinate). The solution was derivatized by adding 50 μL of a 1 mg mL-1 NBD-F (7-fluoro-4-nitrobenzofuranzan) dissolved in acetonitrile. The mixture was vortexed and incubated in the dark at room temperature for 60 min. The reaction was terminated by addition of 50 μL of a 1 M hydrochloric acid solution. The mixture (total volume 200 μL) was centrifuged (15,000 g

78 for 5 min) to settle particulate material and the supernatant removed for anatoxin-a analysis. All samples were processed in duplicates. Anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a were analyzed by HPLC-FD according to James et al. with modifications (James and Sherlock 1996; James et al. 1998). HPLC-FD analysis was performed using an isocratic LC system equipped with a Water’s 510 HPLC pump, a

Hitachi AS-2000 autosampler, and a Shimadzu RF-535 dual monochromator fluorescence detector. ELAB software was used for instrument control, data acquisition and processing. The system was equipped with a Dupont Ace C18 column (4.6 x 250 mm,

5 μm; MAC-MOD Analytical, Inc., PA) with a C18 Phenomenex guard column before the analytical column. A solvent flow rate of 0.8 mL min-1 was maintained using a mobile phase of isocratic 55% acetonitrile in water for 20 min. The fluorescence detector was set to an excitation wavelength of 470 nm and an emission wavelength of 530 nm. The limit of detection (LOD) and the limit of quantification (LOQ) of this technique are defined as signal to noise ratio (S/N) of 3:1 and 10:1 and were determined to be 1 ng mL-

1 and 3 ng mL-1. Both values were based on a standardized injection volume of 20 μL.

Calibration was performed on the same day of field sample analysis by analyzing a series of anatoxin-a standard solutions with different anatoxin-a concentrations which were spiked with known amount of internal standard. Anatoxin-a in field samples was identified by comparing the retention time with those in standards. For quality control purpose, anatoxin-a standards were re-analyzed after every 5 field samples analyses.

The samples positive by HPLC-FD were further analyzed using LCMS-ESI

(liquid chromatography coupled with electrospray ionization mass spectrometry). An

Agilent 1100 series LC-MSD equipped with binary pump, autosampler, on-line vacuum

79 degasser and photodiode array detector, coupled to a quadrupole mass spectrometer

(Agilent, USA) through an electrospray ionization (ESI) interface was used for identification of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a. Two different

LCMS methods were used for the analysis of anatoxin-a. The first method utilized the higher sensitivity obtained via derivatization of anatoxin-a by NBD-F. Water samples were processed as described in section 3.4 and were derivatized with NBD-F as for

HPLC-FD analysis. The analytical column was a Phenomenex Luna C18 column (4.6 x

150 mm, 5 μm; Phenomenex Inc., USA). An isocratic flow of 55% acetonitrile in water was used at flow rate of 0.8 mL min-1 for 20 min. The mass scan range was from m/z 100

– 1000, with selective scanning of ion abundance at m/z 329. The LOD and LOQ of this technique were determined as 3 and 10 ng mL-1 based on a 20 μL injection. This method was used for samples collected in 2000 and was replaced by a second LCMS method to avoid precipitation of the derivatizing agent, NBD-F, in the mass spectrometer. The second LCMS method used an Ace C18 column (4.6 x 250 mm, 5 μm) with a C18

Phenomenex guard column and analyzed extracts directly without derivatization. An initial mobile phase of 95% water containing 5% acetonitrile and 0.1% trifluoroacetic acid (TFA) with a flow rate of 0.8 mL min-1 was used. Between 2 and 12 min, the mobile phase was changed linearly to 50% aqueous acetonitrile. After 12 min, the solvent was ramped back to the original condition (1 min) and equilibrated for 7 min. The mass scan range was from m/z 70-300, with selective scan of the ion abundance at m/z 166. The

LOD and LOQ of this method were 10 and 30 ng mL-1 based on a 20 μL injection. All experiments were performed in the positive-ion mode. The ESI-MS conditions are given

80 in Table 3.1. Agilent Chemstation® software was used for instrumental control, data acquisition and processing.

Dihydroanatoxin-a and epoxyanatoxin-a were analyzed using LCMS with extraction of the ions at m/z 168 and 182 respectively from a full scan. Samples collected from Lake Champlain were later re-examined for 11-carboxyl anatoxin-a, a biosynthetic precursor of anatoxin-a (Selwood et al. 2006) at m/z 210.

3.3. Sampling in New York State lakes and processing of water samples for toxin analysis

3.3.1. Sampling in the lower Great Lakes

The lower Great Lakes (Lake Ontario and Lake Erie) were only sparsely sampled for cyanobacterial toxin in 2000. Cyanobacterial toxins monitoring in Lake Ontario started in 2001, whereas monitoring Lake Erie started in 2002. Water samples were collected during several week-long monitoring cruises on the CCGS Limnos operated by

Canadian Coast Guard or during the cruises on the RV Lake Guardian operated by the

U.S. Environmental Protection Agency. Near-shore water samples have also been collected from embayments along the southern shoreline of Lake Ontario from Port

Niagara to the St. Lawrence River, and along the far eastern shoreline of Lake Erie. The sampling stations are shown in Figure 3.6. From each sampling site, a known volume of water (usually 20 L) was collected and filtered on site onto a 90 mm diameter glass fiber filter (Whatman 934AH, pore size 1.5 μm, USA). The filter was folded and stored in a vial on dry ice for later toxin analyses in laboratory. The identity and location of the sampling stations are given in Appendix 1 and 2.

Table 3.1. Experimental Conditions for the Electrospray Ionization Mass Spectrometry

(ESI-MS) Analysis of Anatoxin-a and its Degradation Products.

HPLC conditions 5-50% aqueous acetonitrile containing

0.1% TFA in 20 min

Ion mass range scan: 70-300 m/z (LCMS);

100-1000 m/z (LCMS-NBD)

Drying gas flow rate: 12.0 L min-1

Nebulizer pressure: 50 PSI

Drying gas temperature: 300oC

Capillary voltage: 3000 V

Fragmentation voltage: 70 V

82 Maumee Bay

Maumee River Lake Erie Sandusky Bay

Lake Ontario

Figure 3.6. Sampling sites in the lower Great Lakes between 2001 and 2005, taken from multiple cruises. GPS coordinates for the stations are given in Appendix 1 and 2. Not all stations were sampled during each cruise.

83 3.3.2. Sampling in Lake Champlain

Water samples (10-20 L) were collected from Lake Champlain at both nearshore and offshore locations at 0.5 m depth (Figure 3.7) and were filtered on site using a 90 mm diameter glass fiber filter. The filtered samples were immediately stored on dry ice.

More water samples were collected by University of Vermont, USDEA and other governmental agencies from Lake Champlain throughout the summer-fall season and were shipped on dry ice to SUNY ESF for toxin analysis. In addition, in 2000 four water samples were collected by a private camp owner during a toxic cyanobacterial bloom immediately after death of several dogs on Lake Champlain.

For those water samples collected in 2000 during a toxic cyanobacterial bloom, analysis for other cyanobacterial toxins were also conducted to identifying the responsible agent for the dog poisonings. Microcystins were analyzed using a protein phosphatase inhibition assay (PPIA) and a LCMS technique (Hotto et al. 2005).

Paralytic shellfish poisoning (PSP) toxins were analyzed using HPLC after electrochemical oxidation (ECOS) or chemical oxidation to form a fluorescent derivative

(PCRS) (Boyer and Goddard 2000).

3.3.3. Sampling in other New York State lakes

Between 2000 and 2005, more than 1000 water samples were collected from over

75 lakes, rivers, reservoirs and other water bodies (Appendix 4). Primary monitoring efforts have been focused on several inland lakes (Table 3.2). Water samples (3-20 L) collected from a depth of 1.0 m were filtered on site as previously described and stored on dry ice for lab analysis. In Onondaga Lake, the bulk of samples was collected from

84 Table 3.2. Location of the sampling stations on Onondaga Lake, Oneida Lake, Lake

Neatahwanta and Lake Agawam.

Lake Station Latitude (oN) Longitude (oW)

Onondaga Lake* North Deep 43.0997 76.2282

South Deep 43. 0806 76.1966

Harbor 43.0713 76.1792

Rose Garden Station 43.0911 76.1933

Allied Point 43.0719 76.2078

Oneida Lake Muskrat Bay 43.2285 76.1044

3 Mile Bay 43.2210 76.0460

Shackelton Station 43.1826 75.9275

Station 125 43.2130 75.9236

Station 117 43.1937 75.8526

Station 109 43.1855 75.7701

Lake Neatahwanta Campground 43.3150 76.4424

Bullhead Point 43.3141 76.4296

Beach Shore 43.3108 76.4249

Lake Agawam 40.8741 72.3922

* Indicates that not all stations on Onondaga Lake were sampled during all sampling years. North Deep and South Deep are two offshore stations that were sampled in 2002 and 2003, while the other three stations are nearshore sites that were sampled in 2000 and

2001.

85 Figure 3.7. Sampling locations in Lake Champlain between 2000 and 2005. Identities and GPS coordinates of the stations are given in Appendix 3.

86 north and south basins in 2002 and 2003 while selected shoreline locations were sampled in 2000 and 2001. Surface samples of cyanobacteria were collected in 2002 and 2003 with a 10-μm plankton net and stored in 0.2% v/v gluteraldehyde for laboratory phycological analysis. Oneida Lake was sampled at six offshore stations weekly in summer from 2000 to 2005. Besides from these two lakes, in response to a toxic event linked to cyanobacterial bloom in 2004, Lake Neatahwanta was sampled weekly from late spring (May) to early fall (October) in 2004 and 2005 at three separate nearshore sites (referred as “Campground”, “Bullhead Point” and “Beach Shore”). Lake Agawam is a eutrophic lake on Long Island, NY. It is a small, shallow and well-mixed lake near

Southampton, NY (Gobler et al. 2006). Samples were collected biweekly in summer and fall 2003 and 2004 by Professor Chris Gobler at SUNY-Stony Brook and sent to SUNY-

ESF for toxin analysis. Additional water samples including water samples from the

Finger Lakes were obtained from Professor Joseph Makarewicz at SUNY-Brockport and

New York Department of Environmental Conservation.

3.3.4. Extraction of anatoxin-a from water samples

The frozen water sample filters were extracted for toxin analysis using an extraction procedure developed by our laboratory as described below. The thawed filter was transferred to a 40-mL polypropylene centrifuge tube containing 10 mL of a 50% methanol/water mixture containing 1% acetic acid. The centrifuge tube was kept in an ice bath during the extraction to prevent heating. Sample filter was extracted using ultrasound (Branson Sonic Power Company, Model 950) in three 20-sec bursts with 20 sec intervals between bursts to allow for cooling down of the solution. This extraction

87 technique resulted in > 90% recovery for anatoxin-a as well as microcystins and PSP toxins as determined by spiking non-toxic samples with known amounts of different toxins and subsequent analyses by HPLC-FD and LCMS. The extracts were centrifuged at 15,000 g for 10-15 min in a Sorvall RC-5B Refrigerated Superspeed Centrifuge (Du

Pont Instruments, USA) to remove the insoluble material. The supernatant was further passed through a 13 mm diameter, 0.45 μm pore size syringe filter equipped with nylon membrane (Pall Life, USA) to remove small particulates. The final sample solution was stored at –20oC for cyanobacterial toxins analysis. Prior to derivatization with NBD-F, 1 mL of the extracted sample was placed in a 1.5 mL microcentrifuge tube and dried in a

SpeedVac (Model # SC110, Savant, USA) at low heat.

3.4. Stability studies of anatoxin-a

3.4.1. Effect of temperature and cool white fluorescence light on anatoxin-a in the dark.

Anatoxin-a fumarate standard was dissolved in deionized water at 10 μg mL-1 and stored in the dark. Solutions were stored at -20°C and analyzed weekly for 3 months.

Solutions were also stored in the dark at 4oC and 22oC and analyzed after 0, 1, 2, 3, 4, 5,

6, 7, 10, 21 and 31 days. Another anatoxin-a solution was exposed to laboratory cool white fluorescence light at 22oC.

3.4.2. Effect of pH, natural and simulated solar radiation on the stability of anatoxin-a

Anatoxin-a (10 μg mL-1) was dissolved in 0.05 M pH 7 and pH 8 Tris-HCl buffers and, 0.05 M pH 9 and pH 10 borate buffers, and stored in the dark at room

88 temperature. Aliquots (200 μL) were removed at the same time intervals as described above.

Anatoxin-a (10 μg mL-1) in Milli-Q deionzed water was placed in 5-mL quartz and optical glass tubes (Starna Cells, Inc., CA), capped and submerged in a 22oC water bath equipped with an isotemp refrigerated circulator (Figure 3.8) (Model 910, Fisher

Scientific, USA). The transmittance of both quartz and optical glass tubes are shown in

Figure 3.9. One tube was wrapped in aluminum foil to serve as a dark control. The others were exposed to natural sunlight on August 30th, 2005, a sunny day with an ambient air temperature of 32- 36oC. Light dose was measured using a chemical actinometer (described in later section). Aliquots (200 μL) of anatoxin-a solutions were removed for LCMS analysis at regular intervals up to 6-hr after exposure. In a separate experiment, artificial solar radiation was generated using a solar simulator with alternating UVA-340 lamps (Q-Panel Lab Products, USA) and cool white fluorescence lamps. Anatoxin-a exposed to only cool white fluorescence light was included as a control. The spectrum of anatoxin-a fumarate was presented in Figure 3.10. Aliquots were removed for LCMS analysis as described above. Anatoxin-a (10 μg mL-1) were prepared in the 50 mM buffered solutions at pH 5 and 6 (acetate), pH 7, 8 (Tris-HCl) and pH 9, 10 (borate) in both quartz and optical glass tubes. The tubes were exposed to artificial solar radiation at room temperature and aliquots removed for LCMS analysis as described above. Another set of anatoxin-a solutions in their individual buffer solutions was wrapped in aluminum foil and served as dark control.

Figure 3.8. Schematic illustration of experimental design for evaluating stability of anatoxin-a under natural and simulated solar radiation.

100

20 Transmittance (%) Transmittance

0 Quartz Optical glass

-20 200 300 400 500 600

Wavelength (nm)

Figure 3.9. Comparison of the transmittance of both quartz and optical glass tubes used in the experiments.

3.5 234 nm 0.30

3.0 0.25

0.20 2.5 0.15

2.0 0.10

0.05 1.5 0.00

Absorbance 1.0 260 280 300 320 340 360 380 400 420

0.5

0.0

-0.5 200 300 400 500 600 700 800 Wavelength (nm) Figure 3.10. Spectrum of anatoxin-a fumarate standard. Anatoxin-a fumarate has a maximum absorbance at 234 nm. The inserted chart is the blown-out of the spectrum between 280 to 400 nm (boxed).

Table 3.3. Physical, chemical and biological parameters of raw lake water collected from

Lake Neatahwanta for assessing stability of anatoxin-a.

Collection date August 29, 2005

Water temperature (oC) 24.3

Visual Dense algal cells suspension dominated by Anabaena sp.

and Microcystis sp.

Secchi depth (m) < 1 pH 9.6

DO (mg/L) 15.8

Nitrate and nitrite (mg/L) < 0.25 μg L-1

Total P (mg P/L) 1.42

Cyanobacterial toxins Microcystins ca. 12 μg L-1, other toxins including

anatoxin-a were not detected (detection limit = 0.001 μg

L-1).

93 3.4.3. Photolysis of anatoxin-a in natural lake water.

Natural lake water was collected from Lake Neatahwanta (Fulton, NY) and used to prepare a 10 μg mL-1 anatoxin-a solution. The physical, chemical and biological parameters of the lake water are described in Table 3.3. Anatoxin-a was exposed to both natural and artificial solar radiation both quartz and optical glass tubes at 22oC as described above. Aliquots of the anatoxin-a solutions were removed and centrifuged at

15,000 g, filtered through an 0.45 μm nylon syringe filter to remove insoluble material and stored at -20°C until analyses. Anatoxin-a and its degradation products were analyzed by electrospray ionization mass spectrometry (LCMS) using an Ace 5 C18 column (4.6 x 250 mm, 30 min gradient of 20-50% aqueous acetonitrile containing 0.1%

TFA). Ion intensities between 100 and 500 amu were collected over 100 msec and the protonated molecular ion of anatoxin-a (m/z 166), dihydroanatoxin-a (m/z 168), and epoxyanatoxin-a (m/z 182) extracted from the total ion chromatograph. Anatoxin-a was quantified using a standard curve generated from commercially available anatoxin-a fumarate.

The light dose in all radiation experiments was measured using a chemical actinometer developed by Jankowaski et al. (1999; 2000) with modifications. One mM benzoic acid with either 10 mM sodium nitrate of 1 mM sodium nitrite was prepared in

2.5 mM pH 7.2 bicarbonate solution. The actinometers were exposed to natural and simulated solar radiation at the same time as the anatoxin-a solutions. Alliquotes were removed for analysis at different time intervals. An ISS PC1 photon-counting spectrofluorometer (Champaign, IL) equipped with a 300 W xenon lamp was used to measure the salicylic acid produced in the actinometers. A standard curve was generated

94 by preparing 0.05-2 μM salicylic acid in the actinometer solutions. Photon exposure was calculated using the equation:

SA][ Ep = − 303.2 avg ε avg ××Φ× NOx ][

Where Ep is the photon exposure (μEinstein cm-2); [SA] is the concentration of salicylic acid produced in the actinometer solution during the irradiation; Φavg the mean quantum yield for SA production, which is wavelength and temperature dependent (Jankowski et

- al. 1999); εavg the mean molar absorption coefficient for nitrate or nitrite; and [NOx ] is the concentration of nitrate of nitrite in the actinometer solution.

3.4.4. Characterization of the degradation product(s) of anatoxin-a

For identification of the degradation product, 1-mg of anatoxin-a fumarate standard was dissolved in 100 μl of 50 mM pH 9.3 borate buffer in D2O, and exposed to artificial light for 24 hr. This anatoxin-a solution was analyzed using 600 mHz NMR prior to light exposure, after 12 hr and after 24 hr of exposure. Also, aliquots were removed at the same 3 time intervals and analyzed by LCMS to determine the mass of the degradation products.

3.5. References

Boyer, G. L., and Goddard, G. D. 2000. High performance liquid chromatography (HPLC) coupled with post-column electrochemical oxidation (ECOS) for the detection of PSP toxins. Natural Toxins 8:1-7.

95 Gobler, C. J., Davis, T. W., Coyne, K. J., and Boyer, G. L. 2006. Interactive influences of nutrient loading, zooplankton grazing and microcystin synthetase gene expression on cyanobacterial bloom dynamics in a eutrophic New York lake. Harmful Algae 6:119. Hotto, A., Satchwell, M. F., and Boyer, G. L. 2005. Seasonal production and molecular characterization of microcystins in Oneida Lake, New York, USA. Environmental Toxicology 20 (3):243-248. James, K. J., Furey, A., Sherlock, I. R., Stack, M. A., Twohig, M., Caudwell, F. B., and Skulberg, O. M. 1998. Sensitive determination of anatoxin-a, homoanatoxin-a and their degradation products by liquid chromatography with fluorimetric detection. Journal of Chromatography, A 798 (1 + 2):147-157. James, K. J., and Sherlock, I. R. 1996. Determination of the cyanobacterial neurotoxin, anatoxin-a, by derivatization using 7-fluoro-4-nitro-2,1,3-benzoxadiazole (NBD-F) and HPLC analysis with fluorimetric detection. Biomedical Chromatography 10 (1):46-47. Jankowski, J. J., Kieber, D. J., and Mopper, K. 1999. Nitrate and nitrite ultroviolet actinometers. Photochemistry and Photobiology 70 (3):319-328. Jankowski, J. J., Kieber, D. J., Mopper, K., and Neale, P. 2000. Development and intercalibration of ultroviolet solar actinometers. Photochemistry and Photobiology 71 (4):431-440. Selwood, A. I., Holland, P. T., Wood, S. A., Smith, K. F., and McNabb, P. S. 2006. Production of anatoxin-a and a novel biosynthetic precursor by the cyanobacterium Aphanizomenon issatschenkoi. Environmental Science & Technology: In print.

96 Chapter 4. Experimental Results

4. 1. Stability of Anatoxin-a Under Different Environmental Factors

4. 1. 1. INTRODUCTION

Anatoxin-a is primarily retained within actively growing cyanobacterial cells, and can be released into the surrounding water during cell senescence, lysis and death

(Sivonen and Jones 1999). Once released, the free anatoxin-a would be prone to degradation in natural aquatic environments. Anatoxin-a degraded rapidly in vitro under sunlight and its breakdown was accelerated under alkaline conditions (Stevens and

Krieger 1991). The reported half-life of anatoxin-a in aqueous solutions ranged from 1-2 hours under sunlight to several days in the absence of sunlight. Besides sunlight and pH, other environmental parameters may affect the stability of anatoxin-a in water. This liability of anatoxin-a may explain why it is rarely detected in natural blooms, and why when detected, the concentration of this toxin are mostly lower than 1 μg L-1 (this thesis)

However, even if anatoxin-a dissipates rapidly, the reports of animal fatalities suggest that the potential health risk should not be underestimated. Smith and Sutton (1993) suggest that anatoxin-a could persist for up to several weeks in natural reservoir waters under normal day-night light conditions after its release into water column. This observation further emphasis the necessity for understanding those factors affecting the stability of anatoxin-a in natural blooms.

The degradation process of anatoxin-a is poorly understood. Although various degradation products have been reported, only a few were chemically identified, and the mechanisms through which degradation of the toxin occurs are not understood. Two compounds, dihydroanatoxin-a and epoxyanatoxin-a, have been isolated and identified

97 from aging cyanobacterial blooms. It has been suggested that these compounds are the major natural degradation products of anatoxin-a in aquatic environments (Smith and

Lewis 1987; Harada et al. 1993). Thus the presence of these degradation products in waters might provide useful information on the occurrence and persistence of anatoxin-a- producing blooms. Here we re-examined the effect of solar radiation and pH on the stability of anatoxin-a in aqueous solution and identified the major degradation product of anatoxin-a under the laboratory conditions.

4. 1. 2. MATERIAL ANS METHODS

Effect of temperature and cool white fluorescence light on anatoxin-a in the dark.

One mg of commercial anatoxin-a fumarate standard (Biomol Inc. , U. S. A. ) was dissolved in deionized water (ca. pH 6) and diluted to make a 10 μg mL-1 solution. Five- mL of this solution were kept at -20°C and analyzed by LCMS weekly for 3 months.

Additional aliquots of this solution (5 mL each ) was kept at 4oC and 22oC in the dark and analyzed after 0, 1, 2, 3, 4, 5, 6, 7, 10, 21 and 31 days. A 10-mL 10 μg mL-1 solution was exposed to laboratory cool white fluorescence light at 22oC and analyzed at the same time intervals. No effort was made to purge the solution of oxygen. However, the solutions were capped tightly to prevent evaporation and gas exchange.

Effect of pH, natural and simulated solar radiation, on the stability of anatoxin-a

Anatoxin-a solutions (10 μg mL-1) were prepared in pH 7 and pH 8 Tris-HCl buffers (0.05 M), and pH 9 and pH 10 borate buffers (0.05 M). The solutions were kept

98 in the dark at room temperature for one month. Aliquots (200 μL each) were removed at different time intervals as described before for analysis by LCMS.

Individual 10 μg mL-1 anatoxin-a solutions prepared in deionzed water were transferred to quartz and an optical glass vials (5 mL, Starna Cells, Inc. , CA). These vials were capped and submerged in a circulating water bath. Water temperature was controlled at 22oC by an isotemp refrigerated circulating bath (Model 910, Fisher

Scientific, USA). Anatoxin-a solutions were exposed to natural sunlight for 7 hours and aliquots of 200 μL were removed every 30-min for later analysis. Another set of anatoxin-a solutions were exposed to artificial solar radiation in the laboratory and sampled at the same time.

Anatoxin-a solutions, 10 μg mL-1 each, were prepared in different pH buffers (50 mM pH 5 and 6 acetate buffer, 50 mM pH 7 and 8 Tris-HCl buffer, and 50 mM pH 9 and

10 borate buffer). These anatoxin-a solutions were exposed to continuous artificial solar radiation in the laboratory for 12 hr. Aliquot of 200 μL was removed from each anatoxin-a solution at 30-min intervals for LCMS analysis. A second set of matching solutions was wrapped in aluminum foil to serve as dark controls.

Photolysis of anatoxin-a in natural lake water

Natural lake water was collected from Lake Neatahwanta (Fulton, NY) in summer

2005, and the raw water was used directly to prepare a 10 μg mL-1 anatoxin-a solution.

The physical, chemical and biological parameters of the lake water are described in Table

3.4 in Chapter 3. Anatoxin-a prepared in the lake water in both quartz and optical glass

99 tubes was exposed to both natural and artificial solar radiation at 22oC as described above.

Aliquots of 200 μL solution were removed at 30-min intervals, centrifuged at 15,000 g, filtered through a 0.45 μm nylon syringe filter to remove insoluble material and stored at

-20°C until further analyses by LCMS.

Analysis of anatoxin-a and its derivative by LCMS and measurement of light dose

LCMS analysis of anatoxin-a and its derivatives used an Ace 5μ C18 column (4. 6 x 250 mm, 30 min gradient of 20-50% aqueous acetonitrile containing 0.1% TFA). Ion intensities between 100 and 500 amu were collected over 100 msec and the protonated molecular ion of anatoxin-a (m/z 166), dihydroanatoxin-a (m/z 168), and epoxyanatoxin-a

(m/z 182) extracted from the total ion chromatograph. Anatoxin-a was quantified using a standard curve generated from commercially available anatoxin-a fumarate (Biomol Inc,

Plymouth Meeting, PA).

The light intensity in all radiation experiments was measured using a chemical actinometer developed by Jankowaski et al. (1999; 2000) with modifications. One mM benzoic acid with either 10 mM sodium nitrate or 1 mM sodium nitrite was prepared in

2.5 mM pH 7.2 bicarbonate solution. The actinometers were exposed to natural and simulated solar radiation and removed for analysis at the same time as anatoxin-a solutions. An ISS PC1 photon-counting spectrofluorometer (Champaign, IL) equipped with a 300 W xenon lamp was used to measure the salicylic acid produced in the actinometers. A standard curve was generated by preparing 0.05-2 μM salicylic acid in the actinometer solutions. Photon exposure was calculated using the equation:

100 SA][ Ep = − 303.2 avg ε avg ××Φ× NOx ][

where Ep is the photon exposure (μEinstein cm-2); [SA] is the concentration of salicylic acid produced in the actinometer solution during the irradiation; Φavg the mean quantum yield for SA production (2.08e-3 or 2.52e-3 mol μEinstein-1 for SA from nitrate and nitrite respectively); εavg the mean molar absorption coefficient for nitrate or nitrite

2 -1 - (2.93e3 or 1.84e4 cm mol respectively); and [NOx ] is the concentration of nitrate of nitrite in the actinometer solution (10 or 1 nM respectively).

Characterization of the degradation product(s) of anatoxin-a

One milligram of anatoxin-a fumarate (Biomol Inc, USA) was dissolved in 500

μL of NMR grade H2O whose pH was adjusted to 9.3 by borate and was exposed to artificial light in the laboratory for 24hr. After exposure, a small aliquot was removed for

LCMS analysis and the rest of the solution was vacuum-dried. The dried sample was rehydrated in NMR grade H2O, vacuum-dried again, and reconsititued in H2O containing

5% D2O. The sample was then analyzed using a 600 mHz NMR.

4.1.3. RESULTS

Effect of pH and light of the stability of anatoxin-a

Anatoxin-a was stable at –20oC and 4oC in water in the dark throughout the examination period of 1-3 months. At 22oC, a slight, but non-significant declining trend in the concentration of anatoxin-a was observed both in the dark and under cool white fluorescence light. However, more than 90% of initial anatoxin-a was still retained in the

101 solution after completion of the experiment (data not shown). Little change was observed in the dark controls in all light-exposing experiments with more than 90% of the starting material remaining after the experimental period.

Changing pH markedly affected the stability of anatoxin-a. Greater degradation of anatoxin-a was observed at basic pHs than at neutral pH conditions in the dark.

Increasing the pH increased the loss rate of anatoxin-a (Table 4.1). More than 92% of the anatoxin-a remained after 30 days at pH 7, while only 37% of the initial anatoxin-a remained at pH 10 (Figure 4.1). Although the loss of anatoxin-a was not a strict first order process over the entire experimental period, a log-linear relationship between loss of anatoxin-a and the exposure time was obtained during the first 15 days. Using first or pseudo-first order kinetics, the half-life of anatoxin-a in the dark was 19 days at pH 10,

37 days at pH 9, four months (128 days) at pH 8, and more than 600 days at pH 7.

Anatoxin-a degraded rapidly in deionized water under both natural and artificial solar radiation (Figure 4.2). Within several hours of exposure to natural sunlight, approximately 40% of the initial anatoxin-a had disappeared (Figure 4.2a). No significant difference in the degradation rate of anatoxin-a was observed between solutions kept in quartz and optical glass vials. Increased light dose increased the degradation of anatoxin-a with the measured half-life (t1/2) of anatoxin-a being 7 hr.

Under laboratory simulated natural radiation, anatoxin-a degraded more quickly than under natural sunlight (Table 4.1, Figure 4.2b) with only 20% of initial anatoxin-a remaining after the 12-hr exposure. Anatoxin-a degraded exponentially with increased light dose and, although the overall degradation was not a first order process, a log-linear decay was obtained during the first 6-hr of exposure. The calculated half-life of

102 Table 4.1. Degradation of anatoxin-a under natural and artificial light. The loss rate was calculated as a proportion of original anatoxin-a, and is given in hours (light reaction) or days (dark reaction). The half-life (t1/2) is also given in hours (light reaction) or days

(dark reaction).

Experiment % ATX left Const. t1/2 Degradation products after 24hr (hr-1) (hr) 1 2 3 4 Temp. -20oC > 95 - - 4oC > 95 - - 22oC > 95 - - Dark pH 5 > 95 - - pH 6 > 95 - - pH 7 92 0.001d >600d pH 8 83 0.005d 128 d pH 9 60 0.019d 37 d pH 10 40 0.035d 19 d Natural DI water 42 0.076 7-8 9 light Lake water 40 0.085 8 9 Artificial pH 5 44 0.068 10 9 9 light pH 6 44 0.076 10 9 9 pH 7 23 0.143 5 9 9 9 9 pH 8 20 0.144 5 9 9 9 pH 9 19 0.138 5 9 9 9 pH 10 10 0.193 4 9 9 DI water 21 0.176 4 9 9 9 9 Lake Water 20 0.125 5 9 9 9 9 indicates the degradation product was detected by LCMS.

Degradation products were numbered as shown in Figure 4.5.

-1 (d) t1/2 is given in days and rate has unit of day .

103

100 90 80 70

60 xin-a 50 nato l a 40 Initia %

DIH2O 30 pH 7 pH 8 pH 9 pH 10 20 0102030 Time (day s) Figure 4.1. The effect of pH on the degradation of anatoxin-a over time. Experim ents were run at room t emperature in the dark. The individual conditions with a r2 in parenthesis were deionized water (0.24), 50 mM pH 7 Tris bu ffer (0 .09), 5 0 mM pH 8

Tris buffer (0.44), 50 mM pH 9 borate buffer (0.93) and 50 mM pH 10 borate buffer

(0.98). Data are shown as mean±SD.

104 100 90 80 70 60 50

30 % Initial ATX

20 A. Natural solar radiation

ATX in glass, r2 = 0.977 ATX in quartz, r2 = 0.977

10 0 50 100 150 200 250 300 350 Light dose (μEinstein cm-2) 100 90 80 70 60 50

40 l 30 % Initia ATX

20 B. Artificial solar radation

ATX in glass, r2 = 0.995 ATX in quartz, r2 = 0.995

10 0 50 100 150 200 250 Light dose (μEinstein cm-2)

Figure 4. 2. Loss of anatoxin-a over the course of 6.5 hr under (A) natural sunlight and

12 hr under (B) simulated solar radiation in deionized water (pH 6) in optical glass tube and quartz tube. Rates of loss are given in Table 4.1.

105 anatoxin-a when exposed to 1.70 μEinstein cm-2 of artificial light was 4 hr. Similar to the exposure under natural sunlight, no difference was observed between quartz and optical glass container.

The effect of pH on the degradation of anatoxin-a under artificial light is shown in

Figure 4.3. Between 20-40% of the initial anatoxin-a remained after exposing for 12 hr with an increased decay of anatoxin-a with increasing pH (Table 4.1). Anatoxin-a degraded significantly more quickly (p<0.1, student’s t-test) under basic and neutral conditions (pH 7, 8, 9 and 10) than under acidic conditions (pH 5 and 6). The degradation showed a log-linear relationship with accumulated light dosage, and was a first order or pseudo-first order process. The half-life of anatoxin-a ranged from 4 to10 hr depending on the pH (Table 4.1).

Similar degradation was observed for anatoxin-a in natural lake water under exposure to both artificial and natural solar radiation (Figure 4.4). Anatoxin-a degraded rapidly and only 20% and 40% of the initial anatoxin-a were retained respectively. The loss of anatoxin-a was again log-linear with a half-life of 8 hr under natural sunlight and

5 hr under artificial solar radiation (Table 4.1).

Exposure of anatoxin-a to light led to the formation of several different degradation products. Four major degradation products were detectable by LCMS at retention times 5.1, 6.7, 7.1 and 8.0 min respectively (Figure 4.5). All four products eluted before anatoxin-a under reverse phase conditions and were assigned as unknowns

1, 2, 3, and 4. Unknown 1, 2 and 4 had a molecular ion [M+H]+ of m/z 184, 18 amu higher than that of anatoxin-a, while unknown 3 had a mass ion [M+H]+ of m/z 166, the same as anatoxin-a (Figure 4.6). Analysis by LCMS of a more concentrated anatoxin-a

106 120

100

60 t X

% Ini ial AT pH 5, r2 = 0.96 40 pH 6, r2 = 0.94 pH 7, r2 = 0.99 pH 8, r2 = 0.98 20 pH 9, r2 = 0.97 pH 10, r2 = 0.98

0 02468101214 Time of exposure (hr) 100 90 80 70 60 50

30 % Initial ATX% pH 5, r2 = 0.96 20 pH 6, r2 = 0.94 pH 7, r2 = 0.98 pH 8, r2 = 0.98 pH 9, r2 = 0.97 pH 10, r2 = 0.98 10 0 100 200 300

Light dose (μEinstein cm-2)

Figure 4.3. Loss of anatoxin-a at different pH under laboratory simulated solar radiation in short term (12 hr) experiments as a function of exposure time (upper figure) and light dose (bottom figure). Buffers were 50 mM acetate (pH 5 and 6), 50 mM Tris-HCl (pH 7 and 8), and 50 mM borate (pH 9 and 10).

107 120

100

60 % Initial ATX 40

A. ATX vs. time 20 Under natural light, r2 = 0.98 Under artificial light, r2 = 0.98

0 02468101214 Time of exposure (hr) 100 90 80 70 60

30 % InitialATX

20 B. ATX vs. light dose Under natural light, r2 = 0.96 Under artificial light, r2 = 0.98

10 0 100 200 300 Light dose (μEinstein cm-2) Figure 4.4. The decay of anatoxin-a in natural lake water (pH 9.6) under natural sunlight and artificial solar radiation versus exposure time (A) and cumulative light doses (B).

Detailed information about the lake water was described in Chapter 3. The regression coefficients (r2) for the individual curves are given in the figure legends.

108 After exposure ATX Before exposure lative response lative 3 Re 4 ATX 1 2

02468101214

Retention time (min)

Figure 4.5. Total ion chromatograph of anatoxin-a in aqueous solution before (upper figure) and after (lower figure) 12-hour exposure to simulated UV radiation (light intensity 950 μW cm-2). Anatoxin-a had a retention time of 9.1 min. The numbers labeled four new peaks assigned as unknown 1, 2, 3 and 4 with retention time of 5.5, 6.7,

7.1 and 8.0 min respectively. HPLC conditions are described in Chapter 3 (Materials and

Methods). Anatoxin-a and its degradation products were run by LCMS without prior derivatization

109 % Abandance

m/z (mua) m/z (amu)

Figure 4.6. Mass spectra of the d e tecte d four degradation products. (a) Unknown 3 at retention time 7.1 min; (b) unknown 1, 2 and 4 at retention time 5.5, 6.7 and 8.0 min, these latter three compounds showed identical mass spectra.

110 sample after 24hr exposure to artificial light indicated several additional products with molecular mass m/z 330 at low abundance. These minor products were not characterized further.

The solution pH had a strong influence on what products were formed. Under acidic to neutral pH conditions, unknown 1 and 2 were the ma jor products. As the pH increased, unknowns 3 and 4, especially unknown 3, increased in abundance and became the more predominant. The formation of these degradation products with time (Figure

4.7) and with cumulative light dose (Figure 4.8) is shown in both deionized water, where all four products have been detected, and at pH 10 where only unknown 3 and 4 were detected. Formation of the degradation products in natural lake water (pH 9.6) was similar to in pH 10 deionized water with the minor difference that unknown 1 was also present as a minor product.

Chemical characterization of the major degradation p roduc t of an atoxin-a

The predominant degradation product of anatoxin-a at pH 10, Unknown 3, was characterized using 600 MHz NMR. The 600 MHz 1H-NMR s pect ra for the anatoxin-a, the starting material is summarized in Table 4.2. Most of the p roton signals are distinct with the C-3 proton at 7.48 ppm, the two bridgehead protons at 4.25 and 5.02 ppm, the methyl protons at 2.3 ppm, and the other methylene protons between 1.9 to 2.7 ppm. The protons at 6.5 ppm were from the fumarate counter-ion and not due to the anatoxin-a free base. Integration of these protons indicated the starting material was in the ratio of 2 molecules of ATX per molecule of fumarate.

111 3e+6 A. In deionized water (pH 6.5)

3e+6 Unknown 1 Unknown 2 Unknown 3 2e+6 Unknown 4

2e+6

e onse 1e+6 R sp

5e+5

-5e+5 02468101214 4e+6 B. At pH 10 3e+6 Unknown 3 Unknown 4 3e+6

2e+6 nse 2e+6 Respo 1e+6

5e+5

02468101214 Time of exposure (hour) Figure 4. 7. Formation of the degradation products of anatoxin-a under artificial solar radiation as a function of time in (A) deionized water and (B) pH 10 buffer solution.

112 3e+6 A. In deionized water (pH 6.5) Unknown 1, r2 = 0.83 3e+6 Unknown 2, r2 = 0.84 Unknown 3, r2 = 0.83 Unknown 4, r2 = 0.95 2e+6

2e+6

1e+6 Response

5e+5

0 100 200 300 3.5e+6 B. At pH 10 Unknown 3, r2 = 0.96 3.0e+6 Unknown 4, r2 = 0.90

2.5e+6

2.0e+6

1.5e+6 Response 1.0e+6

5.0e+5

0.0

0 100 200 300 Light dose (μEinstein cm-2)

Figure 4.8. Formation of the degradation products of anatoxin-a as a function of cumulative light dose in deionized water (A) and pH 10 buffer solution (B) under artificial solar radiation.

113

9 NH O 1 10 2 11 8 6 5 3 7 4

Table 4.2. Chemical shifts of carbons and protons in anatoxin-a.

Carbon δ (ppm) Proton δ (ppm) Connectivity

C-1 58 H-1 5. 02 H-8 (2 protons)

C-2 145

C-3 134 H-3 7. 48 H-4(2 protons)

C-4 24 H-4 2. 6, 2. 7 H-3, H-5 (3 protons)

C-5 40 H-5 1. 9 H-4, H-6 (3 protons)

C-6 61 H-6 4. 25 H-5, H-7 (4protons)

C-7 30 H-7 2. 2, 2. 2 H-6, H-8 (3 protons)

C-8 27 H-8 2. 0, 2. 4 H-7, H-1 (3 protons)

C-10 190

C-11 25 -CH3 2. 3 -

This result has been compared to the study by Koskinen and Rpoport (1985).

114 The 1H NMR spectrum of the degradation product formed in distilled water is summarized in Table 4.3. The protons at the two bridgehead positions have shifted upfield to 3.7 and 4.4 ppm, indicating bicyclic structure remains intact. The olefinic proton at 7.48 ppm was gone, and new proton signals were present at 5.6 ppm and 2.9 ppm. The methylene and methyl protons were clustered together between 1.77 and 1.92 ppm. Tthe protons due to fumarate rema ined intact at 6.2 ppm and integration showed the same 2:1 ratio between protons from fumarate and protons from the product, suggesting that fumarate was not involved in the degradation of anatoxin-a. Running the degradation reaction in deuterium oxide simplified the spectrum by exchange and loss of the protons on the methyl group and at H-3 position of anatoxin-a.

The 13C NMR spectrum of unknown 3 was presented in Figure 4.9. Only nine carbon signals were observed, two of which (135 and 174 ppm) could be assigned to fumarate. The remaining seven 13C peaks accounted for all the 13C atoms in the product.

Given the molecular ion at m/z 166 was only two amu less than the starting anatoxin-a, this indicate d an increase in symmetry in the molecule. The two-dimensional correlation

NMR (1H-COSY, and HMBC) used for connectivity of the protons and carbons are shown in Figures 4.10a and 4.10b). The 1H NMR of the product showed that the methylene and methyl protons were all clustered between 2.15 to 2.3 ppm, suggesting the all methylene groups in the product had similar connectivity, and again confirming the highly symmetrical nature of the final compound. This symmetry was supported by 13C

NMR where the two 13C peaks at 30.2 and 30.4ppm almost have the same chemical shift and are larger than other peaks. They likely comprise of two equivalent 13C atoms each.

The disappearance of methyl protons in D2O and its

115

Table 4. 3. Chemical shifts of carbons and protons in the degradation product 3, tricycloanatoxin-a.

Carbon δ (ppm) Proton δ (ppm)

C-1 62 H-1 4. 40

C-2 130 H-2 5. 61

C-3 52 H-3 2. 90

C-4 30 H-4 1. 77-1. 92

C-5 30 H-5 1. 77-1. 92

C-6 66 H-6 3. 72

C-7 30 H-7 1. 77-1. 92

C-8 30 H-8 1. 77-1. 92

C-10 -

C-11 27 -CH3 1. 86

116

13 Figure 4.9. C NMR of the degradation product of anatoxin-a, unknown 3, in H2O with

5% D2O as solvent. The insert shows that the peak at 30 ppm is actually comprised of two closely adjacent peaks.

117

Figure 4.10a. 1H-COSY NMR of the primary degradation product of anatoxin-a.

118

Figure 4.10b. 1H-13C correlation NMR (HMBC) of the primary degradation product of anatoxin-a.

119 reappearance in H2O indicated these protons are exchangeable and likely adjacent to a ketone, even though the corresponding carbonyl carbon was not observed in the 13C spectrum. The new bridgehead protons at 3.7 and 4.4 ppm were more shielded than in anatoxin-a, but still indicated the original cyclic ring structure was retained. The olefinic

H-3 proton (7.48 ppm) was missing, and a new H-3 proton at 2.9 ppm suggested an additional bridgehead. This proton was not observed when the reaction was run in D2O, and was not exchangeable in the final product, suggesting it was derived via neucleophilic attack on the solvent during formation of a new ring. Based on these observations and the COSY and HMBC NMR data, the structure of unknown 3 was thus determined to be the 2 acetyl-9-azatriclyclo[2.2.1]nonane (Table 4.3) , called tricycloanatoxin-a for short.

4.1.4. DISCUSSION

Previous studies have shown that anatoxin-a can undergoes photo- and non-photo breakdown in the water column (Stevens and Krieger 1991). In this study, anatoxin-a degraded readily both with and without light exposure, but at significantly different rates.

Degradation of anatoxin-a in the dark was much slower than under lighted conditions, with significant amounts of anatoxin-a still present at the end of the 30-day experiments.

Anatoxin-a degraded exponentially in the light, however the degradation process did not follow first or pseudo-first order kinetics over the entire 30-day period. A pseudo log- linear relationship was observed in the first 10 days, which might explain the first order decay kinetics suggested by Stevens and Krieger (1991).

120 Anatoxin-a was exposed to natural or artificial sunlight in both quartz and optical glass containers showed no significant difference in terms of the rate of photodegradation.

As the optical glass used here only transmit light with wavelength greater than 300 nm

(Chapter 3, Figure 3.9), the most effective wavelength range associated with photolysis of anatoxin-a was likely in the UV-A range of 300-400 nm. Anatoxin-a lacks a chromophore which absorbs above 400 nm, but the α, β-unsaturated ketone in anatoxin-a can undergo a π→ π* transition at λ = 300 nm in water (Schuster et al. 1993; Wilsey et al.

2000).

Near log-linearality obtained between the concentration of anatoxin-a and the cumulative light dosage early in the experiments, with a deviation from first order to second order kinetics under basic solutions (Figure 4.2, 4.3) suggest that anatoxin-a may undergo different degradation pathways. First or pseudo-first order kinetics would suggest that anatoxin-a primarily underwent self-fragmentation or rearrangement, or, it reacted with another reactant which was not significantly consumed in the reaction (e. g., water). Proton and carbon NMR analysis indicated that the major degradation product was formed by an intramolecular rearrangement followed by addition of a solvent proton to form the tricyclo-anatoxin-a. (Figure 4.12)

Second order kinetics would require a second reactant which was not water.

Possible reactants included oxygen, hydroxide ion and anatoxin-a itself. A previous study has shown that oxygen does not affect the stability of anatoxin-a (Stevens and

Krieger, 1991). Addition of a hydroxide ion to the double bond of anatoxin-a would generate degradation product(s) 18 amu higher than anatoxin-a in molecular mass.

Indeed, one such degradation product (unknown 4) was detected at high pH conditions,

121 suggesting hydroxide addition ion was likely. Anatoxin-a may also undergo photodimerisation. In presence of light, the [2+2]-photodimerisaiton of alkenes is a common reaction. Two molecules of anatoxin-a could react with each other to form anatoxin-a dimers with an expected molecular weight of 330 amu (Figure 4.11).

Products with molecular mass m/z 330 were observed in the LCMS traces, suggesting that the formation of such dimmers may occur. These dimmer were not characterized further as they appeared to represent a minor degradation pathway under our conditions.

Under acidic conditions, two other minor products with molecular mass 18 amu higher than anatoxin-a (unknown 1 and 2) were detected. Addition of hydroxide ion would unlikely reaction under these conditions. Instead, addition of a water molecule to anatoxin-a in a pseudo-first order process would be a more plausible explanation.

In conclusion, the photodegradation of anatoxin-a in the laboratory was not comprised of any single mechanism but was a combination of both first order and second order reactions. The degradation occurred via uncharacterized hydration of anatoxin-a under acidic conditions, but under basic conditions, more complex degradation process was involved including an internal rearrangement, hydration, addition of a hydroxide ion and photodimerisation of anatoxin-a. Interestingly, the two reported degradation products of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a, were not detected. Since these latter degradation products have been isolated from ageing cyanobacterial blooms, these two compounds may represent biological degradation products of anatoxin-a rather than photolytic or abiotic products. This suggests the abiotic degradation pathways of anatoxin-a are significantly different than the biotic pathways. In this study, raw natural lake water was used to simulate the aquatic environment yet the

122

NH O NH O

D2O

O NH O NH

CHNO20 30 2 2 330.47

Figure 4.11. Possible degradation product of anatoxin-a via [2+2] intermolecular photocycloaddition.

123

O O N N H - H O H Anatoxin-a

O N

Product (unknown 3) MW =165.24 C10H15NO

Figure 4.12. Proposed degradation pathway of anatoxin-a under experimental conditions.

124 formation of the putative biological degradation products (anatoxin-a epoxide and dihydroanatoxin-a) negligible when compared with the rapid abiotic degradation pathways.

The presence of sunlight appears to be the major factor affecting the stability of anatoxin-a in natural environments. Although the effect of light was not explicitly examined, cool white fluorescence light was utilized in the laboratory to roughly simulate

PAR (photosynthetic active radiation, 400-700 nm), and under prolonged exposure, anatoxin-a did not show any significant degradation. This observation suggested the visible range of solar radiation is less important for the stability of anatoxin-a than the

UVA irradiation, an expected result given the lack of a visible chromophore in anatoxin-a that absorbs at wavelengths > 400 nm (Chapter 3, Figure 3.10). This may not necessarily be the case in natural environments, where humic materials may serve as photosynthesizing agents to facilitate light absorption and energy transfer to the molecule

(Kiviranta et al. 1991; Rapala et al. 1994).

Although biodegradation of anatoxin-a was negligible in this laboratory study in comparison with the abiotic processes, the effect of biodegradation in natural aquatic environment should not be underestimated or ignored. During a bloom, visibility of lake water can drop dramatically and light would penetrate a very short distance into the water column. Anatoxin-a produced by those cells dwelling in deeper water, o shaded within the cell photopigments, may be resistant to photodegradation and can pose continuous risk to aquatic organisms. Future, more focused studies need to be conducted to examine the effect of the natural biota on stability of anatoxin-a in lake waters.

125 The abiotic degradation of anatoxin-a, especially the photolysis, may have an important effect on lake management. Both particulate and dissolved cyanobacterial toxins potentially pose a serious threat to human health. Multiple treatment procedures have been implemented to remove these toxins from water column (Hitzfeld et al. 2000;

Newcombe et al. 2000; Drikas et al. 2001). Cyanobacterial toxins such as microcystins and cylindrospermopsin can undergo photocatalytic degradation using titanium dioxide under UV radiation and this process has been used in water treatment facilities to remove cyanobacterial toxins from drinking water (Robertson et al. 1999; Senogles et al. 2001).

The facile photodegradation degradation of anatoxin-a described here suggests that anatoxin-a may also be effectively removed during this water treatment processes, thereby minimizing the health risk associated with anatoxin-a via drinking water .

4. 1. 5. REFERENCE

Drikas, M., Newcombe, G., and Nicholson, B. 2001. Water treatment options for cyanobacteria and their toxins. Proceedings - Water Quality Technology Conference:2006-2033. Harada, K., Nagai, H., Kimura, Y., Suzuki, M., Park, H. D., Watanabe, M. F., Luukkainen, R., Sivonen, K., and Carmichael, W. W. 1993. Liquid chromatography/mass spectrometric detection of anatoxin-a, a neurotoxin from cyanobacteria. Tetrahedron 49 (41):9251-9260. Hitzfeld, B. C., Hoger, S. J., and Dietrich, D. R. 2000. Cyanobacterial toxins: removal during drinking water treatment, and human risk assessment. Environmental Health Perspectives Supplements 108 (1):113-122. Jankowski, J. J., Kieber, D. J., and Mopper, K. 1999. Nitrate and nitrite ultroviolet actinometers. Photochemistry and Photobiology 70 (3):319-328.

126 Jankowski, J. J., Kieber, D. J., Mopper, K., and Neale, P. 2000. Development and intercalibration of ultroviolet solar actinometers. Photochemistry and Photobiology 71 (4):431-440. Kiviranta, J., Sivonen, K., Lahti, K., Luukkainen, R., and Niemela, S. I. 1991. Production and biodegradation of cyanobacterial toxins - a laboratory study. Archiv fuer Hydrobiologie 121 (3):281-294. Koskinen, A. M. P., and Rapoport, H. 1985. Synthetic and conformational studies on anatoxin-a: a potent acetylcholine agonist. Journal of Medicinal Chemistry 28 (9):1301-1309. Newcombe, G., Rositano, J., Brooke, S., and Nicholson, B. 2000. Treatment of cyanotoxins in drinking water using ozone. Proceedings - Water Quality Technology Conference:1494-1502. Rapala, J., Lahti, K., Sivonen, K., and Niemela, S. I. 1994. Biodegradability and adsorption on lake sediments of cyanobacterial hepatotoxins and anatoxin-a. Letters in Applied Microbiology 19 (6):423-428. Robertson, P. K. J., Lawton, L. A., Munch, B., and Cornish, B. 1999. The destruction of cyanobacterial toxins by titanium dioxide photocatalysis. Journal of Advanced Oxidation Technologies 4 (1):20-26. Schuster, D. I., Lem, G., and Kaprinidis, N. A. 1993. New insights into an old mechanism: [2+2] photocycloaddition of enones. Chemical Review 93:3-22. Senogles, P. J., Scott, J. A., Shaw, G., and Stratton, H. 2001. Photocatalytic degradation of the cyanotoxin cylindrospermopsin, using titanium dioxide and UV irradiation. Water Research 35 (5):1245-1255. Sivonen, K., and Jones, G. J. 1999. Cyanobacterial toxins. In Toxic cyanobacteria in water, edited by I. Chorus and J. Bartram. London and New York: E & FN Spon. p41-111. Smith, C., and Sutton, A. 1993. The persistence of anatoxin-a in reservoir water. UK Report No. FR0427. Foundation for Water Research. Smith, R. A., and Lewis, D. 1987. A rapid analysis of water for anatoxin-a, the unstable toxic alkaloid from Anabaena flos-aquae, the stable non-toxic alkaloids left after

127 bioreduction and a related amine which may be nature's precursor to anatoxin-a. Veterinary and Human Toxicology 29 (2):153-154. Stevens, D. K., and Krieger, R. I. 1991. Stability studies on the cyanobacterial nicotinic alkaloid anatoxin-a. Toxicon 29 (2):167-179. Wilsey, S., Gonzalez, L., Robb, M. A., and Houk, K. N. 2000. Ground- and excited-state surfaces for the [2+2]-photocycloaddition of α,β-enones to alkenes. Journal of American Chemical Society 122:5866-5876.

128 4.2. Occurrence and distribution of anatoxin-a in the lower Great Lakes

4.2.1. INTRODUCTION

The lower Great Lakes (Lake Erie and Lake Ontario) are important resources for

New York State in terms of drinking water, recreation and economic development.

These two lakes have a long history of cyanobacteria accumulations (Haffner et al. 1984;

McGucken 2000). During the past several decades, Lake Erie has undergone a switch in the dominant species during the summer cyanobacterial blooms. Previously, nitrogen- fixing taxa such as Anabaena spp. and Aphanizomenon spp. dominated the summer cyanobacterial blooms (Munawar and Munawar 1976; Makarewicz 1993; Conroy et al.

2005b). Since the 1990s, toxic species, especially the non-nitrogen-fixing Microcystis spp., have become more abundant and now often appeared as the predominant taxa

(Brittain et al. 2000; Ouellette et al. 2006). This switch may be related to invasion of dreissenids (Makarewicz et al. 1999). Invasive zebra mussels have been proposed to either modify the nutrient regime to facilitate the growth of Microcystis in well-mixed shallow waters (Raikow et al. 2004; Conroy et al. 2005a), or selectively reject

Microcystis cells in their filtering process (Vanderploeg et al. 2001; Sarnelle et al. 2005).

Microcystis sp. are well-known producers of microcystins (Brittain et al. 2000;

Oberholster et al. 2004), and these toxins have been detected in the western basin of Lake

Erie at concentrations exceeding the provisional guideline concentration of 1.0 μg L-1 for drinking water set by the World Health Organization (WHO) (Brittain et al. 2000; Rinta-

Kanto et al. 2005). Toxic blooms of Microcystis have also been observed in Lake

Ontario but microcystins occur at much lower concentrations (Boyer 2006a; Makarewicz et al. 2006). In addition to Microcystis species, other potentially toxigenic cyanobacteria

129 such as Anabaena sp., Aphanizomenon sp. and Planktothrix/Oscillatoria sp. are also commonly present in these lakes. Some of these species produce microcystins (Rinta-

Kanto and Wilhelm 2006), however, they are also capable of producing other cyanobacterial toxins such as anatoxin-a. Anatoxin-a has been detected in nearby Lake

Champlain and in a number of smaller NY ponds and embayments (this thesis). Much less is known about the distribution of this toxin in Great Lakes ecosystems though a preliminary report has indicated that anatoxin-a is found in both lakes at concentrations greater than 0.1 μg L-1 (Yang and Boyer 2005).

Anatoxin-a, once called “very fast death factor” (VFDF), is a potent neurotoxin

-1 with a LD50 of 200 μg kg (i.p. mouse) (Spivak et al. 1980; Fawell et al. 1999). It has been associated with fatalities of animals and birds in Europe, Africa and the USA

(Edwards et al. 1992; Yang et al. 2001; Krienitz et al. 2003; Gugger et al. 2005; Metcalf et al. 2006). Currently, the health risk to humans from anatoxin-a is unknown. However, the widespread presence of potential anatoxin-a-producing cyanobacteria, its high toxicity and the observation that anatoxin-a is present in the lower Great Lakes all indicate the necessity for increased monitoring for this toxin. Here we report on the occurrence and distribution of anatoxin-a in Lake Erie and Lake Ontario during a 5-year period.

4.2.2. MATERIALS AND METHODS

During 2001 to 2005, Lake Erie and Lake Ontario were sampled using the CCGS

Limnos operated by Canadian Coast Guard and the RV Lake Guardian operated by the

U.S. Environmental Protection Agency. Over 600 samples were collected from more

130 than 100 sampling locations throughout Lake Erie (Appendix 1) and nearly more than

900 samples were collected throughout Lake Ontario (Appendix 2) between the months of May to October during this 5-yr period. At each station, 20 L of water was collected from a depth of 1 m via a submersible pump, filtered through a 90-mm glass fiber filter

(Whatman 934AH) and immediately stored on dry ice for later toxin analysis in the laboratory.

Toxin filters were extracted by probe sonification (Branson model 450) in 10 mL of 50% aqueous methanol containing 1% acetic acid. Extracts were centrifuged at 15,000 g, filtered through an 0.45 μm nylon membrane syringe filter to remove insoluble material and stored at -20oC until analysis. Control experiments using non-toxic culture of cyanobacteria with spiked toxin indicated >90% recovery using this protocol (Chapter

3).

Anatoxin-a and two of its degradation products, dihydroanatoxin-a and epoxyanatoixn-a were analyzed using high performance liquid chromatography with fluorescent detection (HPLC-FD) using the method of James et al. (1998) with modifications. Duplicate 1-mL samples were evaporated to dryness and reconstituted in

100 μL of 0.1 M borate buffer containing 0.6 mM methyl pipecolinate as an internal standard. Fifty μL of the 5 mM fluorescent reagent NBD-F (7-fluoro-4-nitro-2,1,3- benzoxadiazole, TCI America, CA) was added and allowed to react at room temperature for 1 hr. The reaction was stopped by addition of 1 M HCl, centrifuged at 15,000 g and the supernatant analyzed by HPLC-FD using an Ace C18 column (4.6 x 250 mm; 55% aqueous acetonitrile; 470 nm excitation, 530 nm emission). Samples containing toxin by

HPLC-FD were analyzed by electrospray ionization mass spectrometry (LCMS) using an

131 Ace C18 column (5 μ, 4.6 x 250 mm; 30 min gradient of 20-50% aqueous acetonitrile containing 0.1% TFA). Ion abundances from 100 to 500 amu were collected and the protonated molecular ion of anatoxin-a (m/z 166) was extracted from the total ion chromatograph. Anatoxin-a was quantified using a standard curve generated from commercially available anatoxin-a chloride (Biomol Inc., Plymouth Meeting, PA).

4.2.3. RESULTS

Occurrence and distribution of anatoxin-a in Lake Erie

Between 2002 and 2005, 597 samples were collected and analyzed by HPLC-FD for anatoxin-a (Figure 4.13). Anatoxin-a was detected in 85 (13%) of these samples at concentrations greater than the limit of detection of 0.001 μg L-1. Nine samples contained anatoxin-a at concentrations greater than 0.1 μg L-1 (Table 4.4). Presence of anatoxin-a was confirmed in 65 (77%) of these positive samples by LCMS including all 9 samples with higher anatoxin-a concentrations. In those samples where anatoxin-a was not confirmed by LCMS, the concentrations measured by HPLC-FD were often lower than the detection limit (0.003 μg L-1) of the LCMS technique. The two known degradation products of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a, were not detected in any of the samples analyzed.

Anatoxin-a was detected annually in Lake Erie during the 4-year monitoring period. The occurrence and maximum concentration of anatoxin-a varied yearly (Figure

4.14a). Early in this work (2002 to 2004), samples were only collected during several lake-wide cruises in July and August and the monthly variation of occurrence of

132

Lake Erie

Central Basin

2005 Western Basin

2004

Maumee Bay Eastern Basin

< 0.001 μg ATX L-1 Maumee River > 0.001 μg ATX L-1 Sandusky Bay > 0.1 μg ATX L-1

2002 2003

Figure 4.13. Occurrence of anatoxin-a in Lake Erie between 2002 and 2005. The dots indicated the sampling stations. Not all stations were sampled each year, and positive samples may represents multiple detections of anatoxin-a at the same station. Lake Erie was divided into three sub-basins, western basin, central basin and eastern basin. The circles indicate the locations by year where anatoxin-a was detected.

133 Table 4.4. The occurrence of anatoxin-a in Lake Erie between 2002 and 2005. Only those anatoxin-a samples positive (>0.001 μg L-1) by HPLC-FD were analyzed by LCMS for confirmation.

Year of analysis

2002 2003 2004 2005 Overall

Samples analyzed by HPLC-FD

Number collected 117 86 80 314 597

Number (%)1 with 17(15%) 4(5%) 17(21%) 46(15%) 84(14%) > 0.001 μg ATX L-1 1(1%) 2(2%) 6(8%) 0(0%) 9(2%) Number (%)1 with > 0.1 μg ATX L-1 0. 10 0. 20 0. 66 0. 094 0. 066 Maximum concentration (μg ATX L-1)

Samples analyzed by LCMS

Number analyzed 17 4 17 46 84

Number (%)2 with >0.003 15(88%) 4(100%) 16(94%) 30(65%) 65(77%) ATX μg L-1

1 % is based on the total number of samples analyzed by HPLC-FD;

2 % is based on the number of samples analyzed by LCMS.

134 25 0.7

a ) -1

n = 80 0.6 g L μ

n-a 20

0.5 nato

a15 xi n = 117 0.4 ning

n = 314 0.3 10 s contai

0.2 n = 86 5

% sample % 0.1 ximu Ma anatoxin-a concentration ( m 0 0.0 2002 2003 2004 2005

25 0.10 )

b -1 n = 85 g L μ 20 n = 55 0.08

15 0.06 ng anatox

10 0.04 xin-a

n = 77

sam 5 0.02 % ples containi in-a n = 54

n = 43 anatoMaximum concentration ( 0 0.00 May June July August September

Figure 4.14. (a) Percent occurrence (bars) and the maximum concentration (dots) of anatoxin-a in Lake Erie over the 4-year period; (b) Monthly distribution and maximum concentration of anatoxin-a in Lake Erie in summer 2005. The bars indicate % occurrence of anatoxin-a and the solid dots indicate maximum anatoxin-a concentration in each year/month. The limit of detection was 0.001 μg L-1.

135 anatoxin-a in summer could not be determined. In 2005 samples were collected monthly from May to September. Anatoxin-a occurred primarily in late summer (August and

September) and its concentration peaked in early September (Figure 4.14b). The maximum concentration in all samples from 2005 was lower than 0.1 μg L-1.

Anatoxin-a was not detected in the eastern basin of Lake Erie and was rarely observed in the central basin. Most anatoxin-a-containing samples were collected in the western basin of Lake Erie where anatoxin-a was distributed widely across the basin

(Figure 4.13). Concentrations of anatoxin-a in the open water were lower than 0.1 μg L-1 while higher concentrations (>0.1 μg L-1) were observed in some embayments.

Anatoxin-a was detected in Sandusky Bay during all four years of the study at concentrations significantly higher (p = 0.01, Paired Student’s t-Test) than those in the open water (Figure 4.15). Concentrations of anatoxin-a reached 0.1 μg L-1 in 2002, exceeded this concentration in 2003, and were only slightly lower in later years.

Anatoxin-a was detected at significantly lower concentrations in a sampling location outside the bay. Anatoxin-a was also measured at concentrations greater than 0.1 μg L-1 in the Maumee Bay region. Six samples collected in July 2004 in the Maumee River outflow contained anatoxin-a at concentrations greater than 0.1 μg L-1 with a maximum concentration of 0.66 μg L-1. These concentrations decreased gradually as the sampling location moved farther into the lake (Figure 4.16). The HPLC-FD chromatograph and

LCMS spectrum of an anatoxin-a containing water sample collected from the Maumee

River is shown in Figure 4.17.

The relationship between anatoxin-a concentration and phytoplankton productivity, as estimated by chlorophyll a, are shown in Figures 4.18 and 4.19. No

136 0.20 Sandusky Bay 0.18 Station 885

0.16

) 0.14 -1

g L 0.12 μ

0.10

0.08

Anatoxin-a ( Anatoxin-a 0.06

0.04

0.02

0.00 2002 2003 2004 2005 Sampling year Figure 4.15. Comparison of anatoxin-a concentration at Station 1163 Sandusky Bay

(41.4778oN, 82.7064oW) and at a sampling site outside Sandusky Bay (Station 885,

41.5186ºN, 82.6422ºW). Data shown are mean ± standard deviation. The shaded area indicates the range of anatoxin-a concentrations found at open water sites of the western basin. No toxin was detected at station 885 in 2003.

137

0.8

0.6 ) -1 g L

0.4 Anatoxin-a ( μ 0.2

0.0 MR 1 MR 2 MR 3 MR 4 MR 5 MR 6 MR 7 Sampling sites Figure 4.16. Occurrence of anatoxin-a in a transect down the Maumee River into the

Maumee Bay in June of 2004. MR1 was located in the Maumee River near Toledo

(41.6547oN, 83.5328oW), MR7 was located in the outflow of the river in the Maumee

Bay (41.7450oN, 83.2620oW). GPS coordinates for all sites are listed in Appendix 2.

Error bars represent standard deviation for duplicate analysis.

138 1.2e+5

1.0e+5

ATX 8.0e+4 Response

6.0e+4

4.0e+4 0246810121416 Retention time (min) 3e+6 120 166 b 100

80 3e+6 60

40 % Abundance 20 2e+6 0 100 150 200 250 300 m/z (amu) sp

Re2e+6 onse

1e+6

5e+5 02468101214 Retention time (min)

Figure 4.17. (a) The HPLC-FD chromatograph of an anatoxin-a-containing water sample collected from the Maumee River on July 14th, 2004. The peak at 10 min was anatoxin-a.

(b) The LCMS total ion chromatograph of the same sample and the mass spectrum

(inserted figure) of the anatoxin-a peak at 10.7 min. Both HPLC-FD and LCMS chromatographic conditions were described in the Material and Methods.

139

0.7

r2 = 0.10 0.6

0.5 ) -1 0.4 g L μ

0.3

0.2 Anatoxin-a ( Anatoxin-a 0.1

0.0

-0.1 0 20406080 Chlorophyll a (μg L-1)

Figure 4.18. Relationship between the concentration of anatoxin-a and chlorophyll a in samples collected from all basins of Lake Erie between 2002 and 2005. The line represents the linear regression through the data (n = 569).

140 0.7 0.10 (a) 2002-2005 (b) 2002 0.6 2 r2 = 0.68 r = 0.18 0.08

0.5 )

-1 0.06 0.4 g L μ

0.3 0.04

0.2 0.02 Anatoxin-a ( Anatoxin-a 0.1 0.00 0.0

-0.1 -0.02 0 204060800 20406080

-1 0.06

g L 0.4 μ

0.04

0.2

Anatoxin-a ( Anatoxin-a 0.02

0.0 0.00

-0.2 0 102030405060 0 20406080

Chlorophyll a (μg L-1) Chlorophyll a (μg L-1)

Figure 4.19. Relationship between the concentration of anatoxin-a and chlorophyll a in samples collected from the western basin of

Lake Erie (a) overall from 2002 to 2005; (b) in 2002; (c) in 2004; and (d) in 2005. Solid lines are the linear regression through the data.

141 correlation was observed between concentrations of anatoxin-a and chlorophyll a on a lake-wide basis (Figure 4.18, r2=0.10) or when only samples from the western basin were included in the analysis (Figure 4.19a, r2=0.18). When examined yearly, a moderate correlation was observed between the concentrations in 2002 and 2004 (Figure 4.19b-c, r2=0. 68 and 0.55). Anatoxin-a was only weakly correlated with chlorophyll a in 2005

(Figure 4.19d, r2=0.15).

Occurrence and distribution of anatoxin-a in Lake Ontario

Nine hundred and thirty-six water samples were collected from Lake Ontario between 2001 and 2005 and analyzed for anatoxin-a (Table 4.5). Among these samples,

113 (12%) samples contained detectable concentrations of anatoxin-a (> 0.001 μg L-1) including 18 samples with anatoxin-a concentrations exceeding 0.1 μg L-1. Due to a lack of material, only 86 of these 113 positive samples were re-analyzed by LCMS for confirmation. This included 7 of the 18 samples with anatoxin-a concentrations greater than 0.1 μg L-1. Presence of anatoxin-a was confirmed in 36 samples including all 7 samples exceeding 0.1 μg L-1. The two known anatoxin-a degradation products, dihydroanatoxin-a and epoxyanatoxin-a were not detected in any of the samples tested.

Anatoxin-a was detected yearly in Lake Ontario from 2001 to 2005 (Table 4.5).

The occurrence and concentrations of anatoxin-a varied both by year and month (Figure

4.21). In 2001, anatoxin-a was observed at similar abundance in all three sampling months and the concentration of anatoxin-a peaked in July. In 2003, sampling was extended throughout the sampling season and anatoxin-a was detected in all months except May. In 2003 and 2004, anatoxin-a occurred in mid-summer with the maximum

142

Lake Ontario

2005

2004

2001

2003 <0.001 μg ATX L-1 >0.001 μg ATX L-1 >0.1 μg ATX L-1 >1 μg ATX L-1

Figure 4.20. Occurrence and distribution of anatoxin-a in Lake Ontario between 2001 and 2005. The dots indicated the sampling stations on Lake Ontario. Occurrence of anatoxin-a in each year was circled and labeled accordingly. Sampling station s are indicated as circles, not all stations were sampled each year.

143 occurrence and concentration in August (Figure 4.21). In 2005, the distribution of anatoxin-a differed from the previous years. Despite extensive sampling, anatoxin-a was only detected in 3 samples collected in September, all at concentrations lower than 0.1 μg

L-1.

Spatially, anatoxin-a was detected in nearshore locations along the southern and eastern shoreline, in the western lake near Toronto, at the entrance of the Bay of Quinte, as well as in lower St. Lawrence River (Figure 4.20). Anatoxin-a was rarely detected in offshore locations. The change in anatoxin-a was mainly due to the change in Long Pond, a small embayment near Rochester. Anatoxin-a was detected in 13 (39%) of 33 samples collected from June to September in 2003 (Figure 4.22), and in 17 (43%) of 40 samples collected from May to October in 2004. Five samples in 2003 had anatoxin-a concentrations exceeding 0.1 μg L-1. Anatoxin-a occurred mainly in mid-summer (July or August) and accounted for more than 60% of the total occurrence in the Lake Ontario samples. Compared to other sampling locations, concentrations of anatoxin-a measured in Long Pond were significantly higher (p < 0.05) than positive samples from the lake.

Anatoxin-a was not detected in any of the 25 samples collected in 2005 from Long Pond.

The relationship between concentration of anatoxin-a and chlorophyll a is shown in Figure 4.23. Although anatoxin-a was detected in some samples with high concentrations of chlorophyll a, other samples having very high chlorophyll a concentrations (up to 1600 μg L-1) contained no anatoxin-a. Neither did Long Pond, when examined separately from the whole lake, show a strong correlation (r2<0.2) between anatoxin-a and chlorophyll a. However, the occurrence of anatoxin-a in early August 2003 and July

2004 was associated with a peak in chlorophyll a (Figure 4.24) and, following the

144 Table 4.5. Summary of occurrence of anatoxin-a in Lake Ontario between 2000 and

2005. Only anatoxin-a positive samples determined by HPLC-FD were analyzed by

LCMS for confirmation. * indicates that data were based on a limited amount of sample and those samples with anatoxin-a concentration greater than 0.1 μg L-1 were not re- analyzed by LCMS due to a lack of material.

Year of analysis

2001 2002 2003 2004 2005 Overall

Samples analyzed by HPLC-FD

Number collected 137 7 290 260 242 936

Number (%)1 with > 47 5(71%) 20 38 (15%) 3(1%) 113 0.001 μg ATX L-1 (34%) (7%) (12%)

Number (%)1 with > 0.1 2(1%) 0(0%) 5(2%) 11(5%) 0(0%) 18(2%) μg ATX L-1

Maximum concentration 0. 16 0. 006 0. 37 1. 36 0. 07 1. 36 (μg ATX L-1)

Samples confirmed by LCMS

Number analyzed 45 5 20 13* 3 86

Number (%)2 with 13 0(0%) 18 2(15%) 3(100%) 36(42%) >0.003 ATX μg L-1 (29%) (90%)

1 % is based on the total number of samples analyzed by HPLC-FD;

2 % is based on the number of samples analyzed by LCMS.

145

2001 2002 2003 2004 2005 60% 1.6

n=4 1.4 50% ) -1

1.2 g L μ a

x 40%

n=38 1

30% 0.8 30 3 n= 0.6 34 mum a oxin-a ncentr on ( n=4 Samp20% containing anato in- 115 n= % les 66 n= Maxi nat co ati n=

n=8 0.4 33 64 31 = n= 9 n n= 10% n=25

n=10 0.2 40 55 10 3 n=114 n=108 n= n= n= n= n=7 0% 0

y t n p n l g y Jul ug Jul e un Jul ct Ju u ep ct Jul ep Ma Jun A Sep Oc May Ju Aug S Oct May J Aug Sep O May Ju A S O Ma Jun Aug S Oct

Figure 4.21. The percentage occurrence (bars) and the maximum concentration

(diamonds) of anatoxin-a in Lake Ontario between 2001 and 2005. The numbers on top of the bars indicate % occurrence of anatoxin-a in samples analyzed in each specific month. Blank months were not sampled.

146 100 1.6 ) -1 2003 1.4 g L 80 n = 12 μ 1.2 tratio 1.0 60 rrence n = 2 0.8 -a 40 n = 3 0.6 % ATX% occu

n = 9 0.4 anat 20

0.2 imum oxin concen n ( n = 7 Max 0 0.0 May June July August September October

120 1.6 )

2004 n = 10 n = 2 -1 1.4 g L

100 μ

1.2 ion (

ce 80 1.0 n = 8 curren 60 0.8 -a X oc 0.6

40 at % AT n = 8 0.4

20 0.2 n = 4 n = 8 Maximum an oxin concentrat 0 0.0 May June July August September October

Figure 4.22. Percent occurrence of anatoxin-a in the Long Pond, an embayment of Lake

Ontario during summer 2003 and 2004. Solid dots indicate the highest concentration of anatoxin-a detected in each month.

147

0.4

0.3 ) -1

g L 0.2 μ

0.1 Anatoxin-a ( Anatoxin-a

0.0

0 200 400 600 800 1000 1200 1400 1600 1800 Chlorophyll a (μg L-1)

Figure 4.23. Relationship between the concentration of anatoxin-a and chlorophyll a in

Lake Ontario. The line represents linear regression through the data (n = 480).

148 0.4 200 2003

Anatoxin-a Chlorophyll a 0.3 150 ) ) -1 -1 g L g L μ μ 0.2 100 Anatoxin-a ( Chlorophyll a ( a Chlorophyll 0.1 50

0.0 0 Jun Jul Aug Sep Oct Nov 1.6 200 2004 1.4 Anatoxin-a Chlorophyll a 1.2 150 ) ) -1 -1

1.0 g L g L μ μ 0.8 100

0.6 Anatoxin-a ( Chlorophyll a ( a Chlorophyll 0.4 50

0.2

0.0 0 May Jun Jul Aug Sep Oct Nov

Figure 4.24. Variation in concentrations of anatoxin-a and chlorophyll a in Long Pond,

Lake Ontario in 2003 and 2004.

149 decrease in chlorophyll a concentration, the concentration of anatoxin-a drastically decreased. In mid-August 2003, anatoxin-a concentration increased again following the increase in chlorophyll a, yet decreased and disappeared from the water soon even though chlorophyll a was still rising. When the relationships between concentration of anatoxin- a and nutrients (total phosphate, nitrate, and N/P ratio) were examined, no univariant correlation was observed (r2<0.01, data not shown).

4.2.4. DISCUSSION

This is the first systematic investigation on occurrence and distribution of anatoxin-a in the lower Great Lakes. For the purpose of monitoring for anatoxin-a, Lake

Erie was divided into the western, central and eastern basins (Figure 4.13). When compared to the central and eastern basins, the western basin of Lake Erie is much shallower (less than 10 m) and has higher nutrient levels (McGucken 2000). The western basin is more eutrophic than the oligo/mesotrophic central and eastern basins. As a result, cyanobacterial blooms, especially toxic blooms, are more likely to occur in the western basin. Cyanobacterial toxins, including microcystins and anatoxin-a, have been detected annually, especially in nearshore embayments such as Sandusky Bay and the

Maumee River area, where high concentrations of both cyanobacterial toxins were commonly measured. A similar distribution was observed in Lake Ontario between the nearshore/ embayments and offshore waters. Nearshore waters were shallower, with higher levels of nutrients than offshore waters and algal blooms occurred frequently

(Grey 1985; Edsall and Charlton 1997). Anatoxin-a in Lake Ontario was mainly localized in these nearshore or embayment regions along the southern shoreline. The

150 highest concentrations of anatoxin-a were measured in Long Pond near Rochester, New

York.

The western basin of Lake Erie and the nearshore/embayments of Lake Ontario generally have higher nutrient levels than other parts of these two lakes (Sherwook 1999;

Conroy et al. 2005a), and are favorable for growth of cyanobacteria. Cyanobacterial toxins have been proposed to be induced as a defense mechanism to inhibit herbivores and the growth of competing microbes (Lampert 1981; Nizan et al. 1986; Srivastava et al. 1998; Schlegel et al. 1999). If true, cyanobacteria may produce toxins to maintain their blooms. The observed distribution of anatoxin-a was largely restricted to these shallower high biomass waters and was rarely observed in deeper, more oligotrophic waters. A similar distribution has been observed for microcystins (Rinta-Kanto et al.

2005). These observations suggest that toxic cyanobacterial blooms may preferentially occur in shallower, more eutrophic waters in lower Great Lakes, however future studies are needed to determine if these observations are merely a reflection of the higher biomass in these regions on the detection limit for anatoxin-a. Unfortunately, lack of a molecular probe for the anatoxin-a biosynthetic operon make it difficult to determine if toxigenic species producing low levels of toxin are present in the offshore waters. Given that the western basin of Lake Erie and the nearshore or embayments of Lake Ontario are important drinking and/or recreational resources, the potential exposure of animals and humans to toxic cyanobacterial blooms and their toxins is a significant problem. High concentrations of cyanobacterial toxins in these waters could pose a serious human health risk. These shallow water areas should be more carefully monitored for occurrence of toxic cyanobacteria and their toxins to minimize this risk.

151 During our monitoring for anatoxin-a in both the lower Great Lakes, a threshold of anatoxin-a (0.1 μg L-1) was used as a critical threshold for this toxin in natural water.

These studies filtered large volumes of water (20 L) for particulate toxins. Coupled with the use of the highly sensitive HPLC-FD techniques, this gave us a extremely low detection limit for anatoxin-a (0.001 μg per liter lake water). Thus use of “detectable levels” of anatoxin-a probably overestimates the risk from this toxin. Although there are no official guidelines in the United States, New Zealand has adopted a provisional maximal allowed value (PMAV) of 6 μg L-1 and Australia has adopted a guideline of 3

μg L-1 (MHNZ 2000). Based on concentrations of anatoxin-a and chlorophyll a in toxic samples, and an estimation that chlorophyll a constitutes about 1-2% of the dry weight of phytoplankton algae, calculations in New York State waters resulted in 1-20000 μg anatoxin-a g-1 dry weight. By comparing the above data with the reported anatoxin-a concentration in animal fatalities in other countries, we feel these PMAVs may underestimate the risk associated with anatoxin-a. As EPA’s human health concerns on cyanobacterial toxins require concentrations being presented as per liter of water, a cautionary level of 0.1 μg anatoxin-a L-1 lake water was proposed by our laboratory.

This value can provide a least 10-100 fold safety margin for lethal acute toxicity using above calculation and the toxicity data of anatoxin-a. It has been found that the majority of samples collected from the Lower Great Lakes either did not contain anatoxin-a or had only trace amounts of the toxin. The occurrence of anatoxin-a in these environments is rare, and the few samples with higher concentrations of anatoxin-a were usually from locations where dense cyanobacterial blooms were present.

152 Not all cyanobacterial blooms in the lower Great Lakes are toxic, and anatoxin-a concentrations were not correlated with bloom events as measured by chlorophyll a.

Presence of a cyanobacterial bloom, even a bloom dominated by potential anatoxin-a- producing species, does not necessarily suggest production of anatoxin-a. Thus cell- based or biomass-based monitoring of anatoxin-a would likely misrepresent its potential risk.

In contrast to work with Microcystis and microcystins where ~50% of the potential blooms are toxic (Boyer 2006a), most potential anatoxin-a producers, e. g. ,

Anabaena and Aphanizomenon, are not producing anatoxin-a in Lake Erie and Lake

Ontario. Despite the lack of a relationship between anatoxin-a and chlorophyll a, the observation that anatoxin-a was mostly detected in eutrophic waters suggests a certain threshold amount of algal biomass is needed to have measurable levels of anatoxin-a.

This may explain why anatoxin-a was not found in the more oligotrophic offshore waters.

In Long Pond, the variation in chlorophyll a and subsequent changes in anatoxin-a concentration suggest that multiple cyanobacterial blooms occurred during the summer, but only isolated events produced anatoxin-a. These blooms may have produced other cyanobacterial toxins, though other studies have indicated microcystin production in these embayments is extremely low (Makarewicz et al. 2006).

To date, we have not identified the species responsible for toxin production in these lakes. A number of potentially toxicogenic species of cyanobacteria including

Microcystis, Anabaena, Aphanizomenon, and Planktothrix (Oscillatoria) species were commonly present. The western basin of Lake Erie is characterized by high biomass

153 blooms of Microcystis aeruginosa which have produced microcystins concentrations exceeding 20 μg L-1, and blooms of toxic Planktothrix that produce microcystins in

Sandusky Bay. The identification of these species as the toxin producers was based on molecular evidence using the mcy operon (Rinta-Kanto and Wilhelm 2006).

Unfortunately the characterization of the anatoxin-a operon has not been completed at this time, and similar tools are not available for the identification of anatoxin-a-producing species.

Bridgeman (2005) described the occurrence of Microcystis blooms in western

Lake Erie in 2003 and 2004. Heavy Microcystis blooms dominated in the Maumee River

Plume. All the samples collected from the Maumee River region in the summer of 2004 contained both microcystins and anatoxin-a, and both toxins showed similar decreasing trend as the sampling sites moved further off shore. A strong correlation (r2 = 0.77, data not shown, Boyer et al. in preparation) was observed between the concentrations of these two cyanobacterial toxins. This suggests both the microcystins and anatoxin-a may be produced by the same species, or at least the abundance of the different species co-varied.

Molecular studies on these samples showed that Microcystis spp. was the sole species carrying microcystin genes in this region (Hotto, personal communication). Microcystis species have been reported to simultaneously produce microcystins and anatoxin-a (Park et al. 1993). However, the possibility that an anatoxin-a-producing species co-dominated with Microcystis is equally possible. Both anatoxin-a and microcystins were also consistently detected in Sandusky Bay where Aphanizomenon and Anabaena are often the dominating species (Boyer 2006b). Production of anatoxin-a and microcystins in

Sandusky Bay were not correlated, suggesting that these toxins were produced by

154 different species of cyanobacteria. Molecular studies have shown that microcystins were produced by Planktothrix sp. in this bay (Rinta-Kanto et al. 2005; Rinta-Kanto and

Wilhelm 2006). As molecular tools are not yet available for identification of anatoxin-a- producing species, isolation of toxic cyanobacteria from the water body and laboratory cultivation may help identify the producer(s) of anatoxin-a. This would require timely sampling from a toxic bloom and near real-time analysis of anatoxin-a. Given that the current laboratory strains of anatoxin-a-producing cyanobacteria often lose their ability to make toxins in laboratory culture, identification of new anatoxin-a-producing species would be beneficial to future studies on this neurotoxin.

4.2.5. ACKNOWLEDGEMENT

This study was supported by the National Oceanic and Atmospheric Agency

Coastal Oceans Program through their MERHAB-LGL project # NA160 P2788, and New

York Sea Grant awards NA16R61645. Special thanks to Mike Satchwell and Amber

Hotto for microcystins analysis, and to Steve Ragonese, Elizabeth Konopko, Professor

Steven Wilhelm at University of Tennessee, and Professor Joseph Markarewicz at

SUNY-Brockport for sample collection and water chemistry analysis.

4.2.6. REFERENCES

Boyer, G. L. 2006a. Cyanobacterial toxins in New York and the Lower Great Lakes ecosystems. Advances in Experimental Medicine & Biology Proceedings of the Interagency, International Symposium on Cyanobacterial Harmful Algal Blooms: In press. ———. 2006b. Toxic cyanobacteria in the Great Lakes: More than just the western basin of Lake Erie. Great Lakes Research Review 7:2-7.

155 Bridgeman, T. B. 2005. The Microcystis blooms of western Lake Erie 2003-2004. Abstracts. In 48th Annual Conference of the International Association of Great Lakes Research, Ann Arbor, Michigan. May 23-27. Brittain, S. M., Wang, J., Babcock-Jackson, L. K., Carmichael, W. W., Rinehart, K. L., and Culver, D. A. 2000. Isolation and characterization of microcystins, cyclic heptapeptide hepatotoxins from a Lake Erie strain of Microcystis aeruginosa. Journal of Great Lakes Research 26:241-249. Conroy, J. D., Edwards, W. J., Pontius, R. A., Kane, D. D., Zhang, H., Shea, J. F., Richey, J. N., and Culver, D. A. 2005a. Soluble nitrogen and phosphorous excretion of exotic freshwater mussels (Dreissena spp.): potential impacts for nutrient remineralization in western Lake Erie. Freshwater Biology 50 (7):1146-1162. Conroy, J. D., Kane, D. D., Dolan, D. M., Edwards, W. J., Charlton, M. N., and Culver, D. A. 2005b. Temporal trends in Lake Erie plankton biomass: roles of external phosphorus loading and dreissenid mussels. Journal of Great Lakes Research 31 (Suppl. 2):89-110. Edsall, T. A., and Charlton, M. N. 1997. Nearshore waters of the Great Lakes. EPA 905- R-97-015a. USEPA, GLNPO, 179 p. Edwards, C., Beattie, K. A., Scrimgeour, C. M., and Codd, G. A. 1992. Identification of anatoxin-a in benthic cyanobacteria (blue-green algae) and in associated dog poisonings at Loch Insh, Scotland. Toxicon 30 (10):1165-1175. Fawell, J. K., Mitchell, R. E., Hill, R. E., and Everett, D. J. 1999. The toxicity of cyanobacterial toxins in the mouse: II Anatoxin-a. Human & Experimental Toxicology 18 (3):168-173. Grey, I. M. 1985. Differences between nearshore and offshore phytoplankton communities in Lake Ontario. Canadian Journal of Fisheries and Aquatic Sciences 44 (12):2155-2163. Gugger, M., Lenoir, S., Berger, C., Ledreux, A., Druart, J.-C., Humbert, J.-F., Guette, C., and Bernard, C. 2005. First report in a river in France of the benthic cyanobacterium Phormidium favosum producing anatoxin-a associated with dog neurotoxicosis. Toxicon 45 (7):919-928.

156 Haffner, G. D., Griffith, M., and Hebert, P. D. N. 1984. Phytoplankton community structure and distribution in the nearshore zone of Lake Ontario. Hydrobiologia 114 (1):51-66. James, K. J., Furey, A., Sherlock, I. R., Stack, M. A., Twohig, M., Caudwell, F. B., and Skulberg, O. M. 1998. Sensitive determination of anatoxin-a, homoanatoxin-a and their degradation products by liquid chromatography with fluorimetric detection. Journal of Chromatography, A 798 (1 + 2):147-157. Krienitz, L., Ballot, A., Kotut, K., Wiegand, C., Putz, S., Metcalf, J. S., Codd, G. A., and Pflugmacher, S. 2003. Contribution of hot spring cyanobacteria to the mysterious deaths of Lesser Flamingos at Lake Bogoria, Kenya. FEMS Microbiology Ecology 43 (2):141-148. Lampert, W. 1981. Toxicity of the blue-green Microcystis aeruginosa: effective defense mechanism against grazing pressure by Daphnia. Verhandlungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 21:1436-1440. Makarewicz, J. 1993. Phytoplankton biomass and species composition in Lake Erie, 1970 to 1987. Journal of Great Lakes Research 19 (2):258-274. Makarewicz, J. C., Boyer, G. L., Guenther, W., Arnold, M., and Lewis, T. W. 2006. The occurrence of cyanotoxins in the nearshore and coastal embayments of Lake Ontario. Journal of Great Lakes Research (submitted). Makarewicz, J. C., Lewis, T. W., and Bertram, P. E. 1999. Phytoplankton composition and biomass in the offshore waters of Lake Erie: Pre- and post-Dreissena introduction (1983-1993). Journal of Great Lakes Research 25 (1):135-148. McGucken, W. 2000. Lake Erie Rehabilitated. Controlling Cultural Eutrophication, 1960s-1990s. Edited by J. Stine and J. Tarr. 1st ed, Technology and the Environment. Akron, Ohio: The University of Akron Press, 318 p. Metcalf, J. S., Morrison, L. F., Krienitz, L., Ballot, A., Krause, E., Kotut, K., Putz, S., Wiegand, C., Pflugmacher, S., and Codd, G. A. 2006. Analysis of the cyanotoxinx anatoixn-a and microcystins in Lesser Flamingo feathers. Toxicological and Environmental Chemistry 88:159-167. MHNZ. 2000. Drinking-water standards for New Zealand 2000. Wellington, New Zealand: Ministry of Health, 145 p.

157 Munawar, M., and Munawar, I. F. 1976. A lakewide study of phytoplankton biomass and its species composition in Lake Erie, April-December, 1970. Journal of the Fisheries Research Board of Canada 33:581-600. Nizan, S., Dimentman, C., and Shilo, M. 1986. Acute toxic effects of the cyanobacterium Microcystis aeruginosa on Daphnia magna. Limnology and Oceanography 31:497-502. Oberholster, P. J., Botha, A.-M., and Brobbelaar, J. U. 2004. Microcystis aeruginosa: source of toxic microcystins in drinking water. African Journal of Biotechnology 3 (3):159-168. Ouellette, A. J. A., Handy, S. M., and Wilhelm, S. W. 2006. Toxic Microcystis is widespread in Lake Erie: PCR detection of toxin genes and molecular characterization of associated cyanobacterial communities. Microbial Ecology 51 (2):154-165. Park, H. D., Watanabe, M. F., Harada, K., Nagai, H., Suzuki, M., Watanabe, M., and Hayashi, H. 1993. Hepatotoxin (microcystin) and neurotoxin (anatoxin-a) contained in natural blooms and strains of cyanobacteria from Japanese freshwaters. Natural Toxins 1 (6):353-360. Raikow, D. F., Sarnelle, O., Wilson, A. E., and Hamilton, S. K. 2004. Dominance of the noxious cyanobacterium Microcystis aeruginosa in low-nutrient lakes is associated with exotic zebra mussels. Limnology and Oceanography 49 (2):482- 487. Rinta-Kanto, J. M., Ouellette, A. J. A., Boyer, G. L., Twiss, M. R., Bridgeman, T. B., and Wilhelm, S. W. 2005. Quantification of toxic Microcystis spp. during the 2003 and 2004 blooms in western Lake Erie using quantitative real-time PCR. Environmental Science and Technology 39:4198-4205. Rinta-Kanto, J. M., and Wilhelm, S. W. 2006. Diversity of microcystin-producing cyanobacteria in spatially isolated regions of Lake Erie. Applied and Environmental Microbiology 72 (7):5083-5085. Sarnelle, O., Wilson, A. E., Hamilton, S. K., Knoll, L. B., and Raikow, D. F. 2005. Complex interaction between exotic zebra mussels and the noxious phytoplankter, Microcystis aeruginosa. Limnology and Oceanography 50:896-904.

158 Schlegel, I., Doan, N. T., de Chazal, N., and Smith, G. D. 1999. Antibiotic activity of new cyanobacterial isolates from Australia and Asia against green algae and cyanobacteria. Journal of Applied Phycology 10:471-479. Sherwook, D. A. 1999. Phosphorus loads entering Long Pond, a small embayment of Lake Ontario near Rochester, New York. U.S. Geology Survey Fact Sheet FS 128-99. NYUSGS, 4 p. Spivak, C. E., Witkop, B., and Albuquerque, E. X. 1980. Anatoxin-a: a novel, potent agonist at the nicotinic receptor. Molecular Pharmacology 18 (3):384-394. Srivastava, V. C., Manderson, G. J., and Bhamidimarri, R. 1998. Inhibitory metabolite production by the cyanobacterium Fischerella muscicola. Microbiological Research 153:309-317. Vanderploeg, H. A., Liebig, J. R., Carmichael, W. W., Agy, M. A., Johengen, T. H., Fahnenstiel, G. L., and Nalepa, T. F. 2001. Zebra mussel (Dreissena polymorpha) selective filtration promoted toxic Microcystis blooms in Saginaw Bay (Lake Huron) and Lake Erie. Canadian Journal of Fisheries and Aquatic Sciences 58:1208-1221. Yang, X., and Boyer, G. L. 2005. Occurrence of the cyanobacterial toxin, anatoxin-a, in the lower Great Lakes. Abstract. In 48th Annual Conference of the International Association of Great Lake Research, Ann Arbor, Michigan. May 23-27. Yang, X., Satchwell, M. F., and Boyer, G. L. 2001. The identification of anatoxin-a from a toxic blue-green algae bloom in Lake Champlain, USA. Abstract. In Fifth International Conference on Toxic Cyanobacteria, Moosa Lakes, Queensland, Australia. July 15-20. 1 p.

159 4.3. Identification, Occurrence and Spatial Distribution of Anatoxin-a in Lake

Champlain, New York

4.3.1. INTRODUCTION

Lake Champlain, with 1124 square kilometers of surface area, over 900 km of shoreline, and a drainage basin of nearly 20,000 square kilometers, is regarded as the

“sixth Great Lake”. It is shared by the states of Vermont and New York, as well as a small part in the Canadian province of Quebec (Shanley and Denner 1999). The lake is heavily used for recreational purposes such as fishing, boating, swimming and tourism, and is also an important drinking water source for New York and Vermont.

Cyanobacteria have long been a common part of the phytoplankton communities of Lake Champlain (Muenscher 1930; Myer and Gruendling 1979; Brown et al. 1991;

Shambaugh et al. 1999; Mihuc et al. 2005). Cultural eutrophication has resulted in elevated nutrient concentrations in the lake over the last several decades (Jokela et al.

2004). These elevated nutrients have led to an increased frequency and size of cyanobacterial blooms. They have also led to a shift in species composition. Early blooms were often dominated by Anabaena and Aphanizomenon in mid- to late-summer of each year (Muenscher 1930; Shambaugh et al. 1999; Shanley and Denner 1999; Mihuc et al. 2005). In recent years, the northern portion of the lake has switched to Microcystis spp. dominated blooms, especially in Missisquoi Bay and Inland Sea (Muenscher 1930;

Shambaugh et al. 1999; Watzin et al. 2003). Anabaena, Aphanizomenon and Microcystis are all potentially toxigenic genera capable of producing potent neurotoxins such as anatoxin-a or saxitoxin, or hepatotoxins such as the microcystins (Codd et al. 1997;

Carmichael et al. 2001). As a result, the increased occurrences of these cyanobacterial

160 blooms in Lake Champlain has been accompanied by increased risk of cyanobacterial toxin production (Boyer et al. 2004).

Toxic cyanobacterial blooms in Lake Champlain were first documented in the summer of 1999. Several dogs died shortly after ingesting shoreline scum on both the

New York and Vermont sides of Lake Champlain (Watzin et al. 2005). These animals experienced convulsions, limb twitching, muscle spasm, paralysis and respiratory arrest, typical symptoms of anatoxin-a poisoning. The following summer, toxic cyanobacterial blooms was once again observed and two dogs died after swimming in a cyanobacterial bloom located in an area south of Whallon Bay on the New York side of Lake Champlain.

Similar to the toxic incidents in the previous year, both dogs showed signs of neurotoxicosis (Rosen et al. 2002). In 2002, two additional dog fatalities occurred in

Missisquoi Bay near Highgate Springs, Vermont after reportedly ingesting water from an intense bloom of Microcystis.

Although human beings are unlikely to consume water with a high toxin content, accidental or recreational exposure to toxic cyanobacteria may lead to accompanying health problems (Codd et al. 1997; Haider et al. 2003; Codd et al. 2005). In recognition of this threat, efforts are currently underway to monitor the occurrence of potentially toxic cyanobacterial blooms in Lake Champlain. The Vermont State Department of

Health, through their “Healthy Vermonter” website, provides weekly information about blue-green algae conditions around Lake Champlain. High risk areas, especially near

Burlington Bay and Missisquoi Bay, were monitored biweekly for the occurrence of cyanobacteria and their toxins. Mihuc et al. (2005), found that Microcystis species dominate the northern lake and can account for more than 50% of the total phytoplankton

161 density in summer. These blooms are often associated with production of microcystins where in Missisquoi Bay, microcystins levels can approach 1000 μg L-1, dramatically exceeding the World Health Organization (WHO) advisory level for drinking water of 1

μg L-1. In general, microcystins levels coincided well with Microcystis densities though some locations had high Microcystis densities but not elevated toxin levels. Watzin et al.

(2005) have developed a monitoring and alert framework for microcystins-containing cyanobacterial blooms in Lake Champlain. This alert is dependent on cell densities of potentially toxic cyanobacteria including Microcystis spp., Anabaena spp,

Aphanizomenon spp. and Gloeotrichia spp., and uses elevated cell densities to trigger the need for toxin analyses to lower the potential risks to humans and animals.

Although both microcystins and anatoxin-a producing species are included in this framework, the efforts have primarily focused on monitoring for microcystins-producing cyanobacteria. Less effort has been expended on the cyanobacterial neurotoxins, despite the fact that the neurotoxin anatoxin-a is probably produced by a number of species including Anabaena spp., Aphanizomenon spp. and Microcystis spp. (Codd et al. 1997;

Carmichael 2001), all of which are common in Lake Champlain. Anatoxin-a toxicosis was responsible for multiple fatalities of dogs and birds in Europe and Africa (Edwards et al. 1992; Krienitz et al. 2003; Gugger et al. 2005; Metcalf et al. 2006). These dogs typically showed symptoms of neurotoxicosis and died within a short period. Anatoxin-a has been detected from both lake water and dogs’ tissue samples at concentrations as high as 0.8 mg g-1. However, identification of anatoxin-a in biological matrices is complicated by the presence of the naturally occurring amino acid, phenylalanine. A suspected human intoxication in 2002 initially confused phenylalanine and anatoxin-a due to their

162 similar masses (Carmichael et al. 2004). Despite this error, and the generally rare occurrence of anatoxin-a in the United States, the high toxicity of anatoxin-a (200 μg kg-1 body weight, i.p. mouse) and the prior animal fatalities in Lake Champlain suggest that the potential health risk of anatoxin-a to human and animals should not be underestimated. Currently, little is known about the occurrence of anatoxin-a in Lake

Champlain. Here we report on the identification of anatoxin-a in samples from the initial animal poisonings and report on the occurrence and distribution of this toxin over a 6- year period in Lake Champlain.

4.3.2. MATERIALS AND METHODS

Lake Champlain was sampled during 2000 to 2005 from both nearshore (along the shoreline of both New York and Vermont) and offshore locations (Figure 4.25). At each site, a 20 L water sample was collected from a depth of 1.0 m using a peristaltic pump, filtered through a 90-mm glass fiber filter (Whatman 934AH) and the filter was immediately stored on dry ice for later toxin analysis in the laboratory. Following the dog fatalities in 2000, four samples (ca. 1 L each) were collected the same day from the site of the toxic incident and surrounding water by the camp owner and shipped to

SUNY-ESF for identification of responsible toxins.

Toxin filters were extracted by probe sonification (Branson 450) in 10 mL of 50% aqueous methanol containing 1% acetic acid. Extracts were centrifuged at 15,000 g, filtered through a 0.45 μm nylon syringe filter to remove insoluble material and stored at

-20°C until analyses. Anatoxin-a and two of its degradation products, dihydroanatoxin- a and epoxyanatoxin-a, were analyzed using high performance liquid chromatography

163 MISSISQUOI BAY

Site of dog fatality in 2002 NORTHWEST ARM

INLAND SEA Site of a dog fatality in 1999

MAIN LAKE

Site of dogs Site of a dog fatalities in 2000 fatality in 1999

<0.001 μg L-1 0.001 – 0.1 μg L-1 0.1 – 1.0 μg L-1 >1.0 μg L-1

SOUTH LAKELAKE

Figure 4.25. Sampling sites (open and solid dots) and the occurrence of anatoxin-a in

Lake Champlain. The solid dots indicated locations where anatoxin-a was identified in the lake between 2000 and 2005. Sites of the dog fatalities in 1999, 2000 and 2002 are indicated by arrows.

164 with fluorescence detection (HPLC-FD) using the method of James et al. (1998) with modifications. Extracts (1 mL) were evaporated to dryness and reconstituted in 100 μL of 0.1 M borate buffer containing 0.6 mM methyl pipecolinate as an internal standard.

Fifty μL of the fluorescent reagent, 5 mM NBD-F (7-fluoro-4-nitro-2,1,3-benzoxadiazole,

TCI America, CA) was added and allowed to react at room temperature for 1 hr. The reaction was stopped by addition of 1 M HCl, centrifuged at 15,000 g and the supernatant analyzed by HPLC-FD using an Ace 5μ C18 column (4.6 x 250 mm; 55% aqueous acetonitrile, 470 nm excitation, 530 nm emission). Positive samples by HPLC-FD were analyzed by electrospray ionization mass spectrometry (LCMS). Early extracts (before

2001) were analyzed by LCMS after derivatization with NBD-F using a Luna C18 column (4.6 x 150 mm, 55% aqueous acetonitrile with 1% acetic acid), later samples

(2001-2005) were analyzed directly without derivatization using the Ace C18 column (30 min gradient of 20-50% aqueous acetonitrile containing 0.1% TFA). Ion intensities from

100 to 500 amu were collected and the protonated molecular ion of anatoxin-a (m/z 166) was extracted from the total ion chromatograph. Anatoxin-a was quantified using a standard curve generated from commercially available anatoxin-a chloride (Biomol Inc,

Plymouth Meeting, PA).

4.3.3. RESULTS

Identification of anatoxin-a in samples from the 2000 toxicosis event

HPLC-FD analysis of the water samples collected at the site of the dog fatalities in 2000 indicated elevated levels of anatoxin-a were present in 2 of the 4 samples (Figure

4.26). The sample collected from the swimming area adjacent to the dock contained 1.66

165 μg anatoxin-a per liter lake water, whereas in a second sample from the same area farther off the dock contained 0.4 μg anatoxin-a L-1 (Table 4.6). Other cyanobacterial toxins, microcystins and paralytic shellfish toxins (PSP toxins) were not detected in these samples. The presence of anatoxin-a in these two samples was confirmed by LCMS analysis of the NBD-anatoxin-a derivatives (Figure 4.27). The mass spectra of the samples contained a molecular ion [M+H]+ at m/z 329 and were identical to the spectra obtained from a NBD-anatoxin-a standard. The concentration derived from LCMS analysis for the first sample was 1.47 μg L-1, in good agreement with 1.67 μg L-1 detected by HPLC-FD. Presence of anatoxin-a in the second water sample was also confirmed by

LCMS of the NBD derivative; however the concentration of anatoxin-a in the extract was below the limit of quantification of the LCMS method.

Occurrence of anatoxin-a in Lake Champlain from 2000 to 2005

A total of 788 water samples were collected between 2000 and 2005 from more than 70 different locations (Appendix 3) and analyzed by HPLC-FD for anatoxin-a.

Sixty-one (8%) of these samples contained detectable concentrations of anatoxin-a

(>0.001 μg L-1) while 8 samples contained levels of anatoxin-a above 0.1 μg L-1 (Table

4.7). Due to limited availability of these samples, only 51 of these toxin-containing samples were re-analyzed using LCMS for confirmation of anatoxin-a (Table 4.7).

Anatoxin-a was confirmed by LCMS in 26 of the 51 samples tested including all 8 samples with anatoxin-a concentrations above 0.1 μg L-1 as measured by HPLC-FD. The two degradation products of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a,

166

Figure 4.26. The HPLC-FD chromatograph of the water sample containing anatoxin-a collected during a cyanobacterial bloom that caused a dog toxicosis event at Lake

Champlain on August 3rd 2000. The arrow marks the peak corresponding to NBD- anatoxin-a derivative, IS stands for internal standard. The sample was first derivatized with the fluorescent reagent NBD-F before analysis. HPLC condition: 55% aqueous acetonitrile, 0.8 mL min-1 isocratic on an Ace C18 column (4.6 x 250 mm, 5 μ).

167

Table 4.6. Concentrations of cyanobacterial toxins in four water samples collected during the year 2000 dog toxicosis event near Whallon Bay, Lake Champlain.

Water sample Microcystins analysis Anatoxin-a analysis PSP toxins analysis

PPIA LCMS HPLC-FD LCMS HPLC- HPLC- ECOS PCRS (μg L-1) (μg L-1)

#1 (adjacent to ND ND 1.66 1.47 ND ND the dock)

#2 (swimming ND ND 0.4

#3 (nearshore) ND ND ND ND ND ND

#4 (offshore) ND ND ND ND ND ND ND means toxicity was not detected by the specific method;

168 7000 6000 A 11.4 5000 4000 3000 2000 1000 0 8000 B

Response 6000 11.4

4000

2000

0 2 4 6 8 10121416 Time (min) 120 C 100 186

40 329 20 351 0 120 D sponse 100 186

Re 80

40 329

20 351 0 150 200 250 300 350 400 m/z, amu

Figure 4.27. LCMS screening for anatoxin-a in a Lake Champlain sample. (A) A water sample collected at the site of dog toxicosis event on August 3rd 2000. (B) Anatoxin-a-

NBD standard. (C) ESI-MS spectrum of the Lake Champlain sample. (D) ESI-MS spectrum of anatoxin-a-NBD standard. Anatoxin-a had a retention time of 11.4 min.

[M+H]+ m/z 329, [M+Na]+ m/z 351. LCMS conditions are described in the Materials and

Methods.

169 were not detected in any of the samples collected.

The distribution of anatoxin-a showed a strong variation by month and year

(Figure 4.28). The maximum concentration of anatoxin-a peaked in August or September and was highly variable between years. One sample from a surface cyanobacterial bloom in early September 2001 heavily dominated by Gloeotrichia and Microcystis had a concentra tion of 6.3 μg L-1. The correl ation between a natoxin-a c oncentration s and the chlorophyll a concentrations is shown in Figure 4.29. The chlorophyll a conce ntration varied significantly in dif ferent sub-ba sins, within the same sub- basin, and be tween different years. Although relatively high chlorophyll a concentrations (up to 10 μg L-1) were detected in those sam ples contai ning anatoxin -a, other samp les containe d extremely high chlorophyll a concentrations (up to 1,000 μg L-1) without measurable levels of anatoxin-a.

Spatial distribution of anatoxin-a in different sub-basins of Lake Champlain

Anatoxin-a was detected in all parts of Lake Champlain except the South Lake, which was only sampled twice (Table 4.7). In Missisquoi Bay, occurrence of anatoxin-a was confirmed only once in 128 water samples collected, despite the high levels of cyanobacteria biomass and microcystins that routinely occur in that sub-basin. Anatoxin- a occurred more frequently in the other sub-basins. Although anatoxin-a were detected by HPLC in 18 samples, six out of the 335 water samples collected from Main Lake were confirmed to contain detectable concentrations of anatoxin-a using LCMS, though at concentrations significantly lower than 0.1 μg L-1 (p<0.01, Student’s t-Test). The

Northwest Arm and Inland Sea had the highest frequency of anatoxin-a occurrence

170

Table 4.7. Occurrence of anatoxin-a in Lake Champlain during a 6-year monitoring period by sub-basin. Only anatoxin-a positive samples determined by HPLC-FD were analyzed by LCMS technique for confirmation of anatoxin-a.

Missisquoi Inland Northwest Main South Overall Bay Sea Arm Lake Lake

Samples analyzed by HPLC-FD

Number collected 128 249 74 335 2 788

Number (%) with 5 (4%) 21 (8%) 6 (8%) 30 (9%) 0 (0%) 62 (8%) >0.001 ATX μg L-1

Samples analyzed by LCMS

Number analyzed 5 21 6 18 - 51

Number (%) with 1 (20%) 14 (67%) 5 (83%) 6 (33%) - 26 (51%) >0.003 ATX μg L-1

Number (%) with 1 (1%) 5 (2%) 1 (1%) 3 (1%) - 10 (1%) >0.1 ATX μg L-1

171 (8% and 8% respectively). Anatoxin-a was distributed widely within the Northwest Arm and was not detected repeatedly at any one location. In most cases, the concentrations of anatoxin-a were below 0.1 μg L-1 except in September 2003 when a concentration of 0.26

μg L-1 was detected in King Bay (Figure 4.25). In the Inland Sea, the toxin was first observed in 2001 in scums in Maquam Bay at a concentration of 6.3 μg L-1. In later years, St. Albans Bay appeared to become a primary area of concern as anatoxin-a was detected in this bay between the summer of 2002 and 2003. Anatoxin-a was also occasionally observed at other locations in the Inland Sea yet at concentrations significantly lower than those observed in St. Albans Bay (p<0.05, Student’s t-Test).

4.3.4. DISCUSSION

Here we show that the dog neurotoxicosis in 2000 was associated with anatoxin-a and that this neurotoxin is distributed widely throughout the lake. The dog fatalities in summer 2000 occurred with the clinical symptoms such as vomiting, paralysis of the legs and respiratory failure (Craig Russel, personal communication; Gregory L. Boyer, personal communication) typical of anatoxin-a intoxication. Anatoxin-a was the only cyanobacterial toxin detected in these samples and neither the hepatotoxic microcystins nor the neurotoxic paralytic shellfish toxins were detected in any of these samples. The amino acid phenylalanine commonly interferes with the identification of anatoxin-a in natural waters (Furey et al. 2005; Gugger et al. 2005). Here, anatoxin-a and phenylalanine were effectively separated by both HPLC-FD and LCMS. While the responsible organism for this intoxication was not identified, these dog fatalities were

172 20 7.0 2000 2001 2002 2003 2004 2005

18 n = 34

6.0 16 2.0 ) -1 g L

14 μ

1.6

12 81 n =

10 1.2

8 n = 16 n = n = 49 n = 65 n =

% Occurrence of anatoxin-a Occurrence % 0.8 6 n = 58 Maximum anatoxin-a concentration ( concentration anatoxin-a Maximum 4 0.4 27 17 n = 65 n = 2 n = 17 n = n = n = 38 n = 30 n = n = 27 n = n = 68 n = n = n = 4

2 n = 18 n = n = n = 23 n = n = 82 n = 3 n = 33 n = n = 29 n =

0 0.0 Jun Sep De c Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun Sep

Figure 4.28. Percent occurrence of anatoxin-a and its maximum concentrations in Lake Champlain between 2000 and 2005. The percentage of anatoxin-a occurrence are given by the vertical bars. The solid dots show the maxim um anatoxin-a concentration during that time period. “n” indicates the total number of water samples collecte d and analyzed in an indi vidual month.

173

1.2

1.0

0.8 ) -1 g L

μ 0.6 ( a in- 0.4 Anatox 0.2

0.0

1e-2 1e-1 1e+0 1e+1 1e+2 1e+3 1e+4 1e+5 Chlorophyll a (μg L-1)

Figure 4.29. Chlorophyll a concentration and production of anatoxin-a in Lake

Champlain.

174 likely associated with anatoxin-a based on both their clinical symptoms and the high level of toxin found in the water. The observation that a higher concentration of anatoxin-a was measured adjacent to the dock than in the swimming area away from the dock (Table

4.6) suggested a wind and/or wave-borne accumulation of toxic cyanobacterial bloom.

All three cyanobacterial species known to produce anatoxin-a (Anabaena,

Aphanizomenon and Microcystis) are commonly present in Lake Champlain.

Furthermore, a toxic bloom of Anabaena was reported in the Burlington marina about this same time. This observation was also likely due to a similar wind-borne accumulation of cyanobacteria against the retaining walls of the marina.

The observation that surface accumulations of cyanobacteria may concentrate toxigenic species is important for future anatoxin-a monitoring on the lake. The production of anatoxin-a was not clearly associated with algal density as measured by chlorophyll a and it would be difficult for monitoring to predict toxicity of a spec i fic bloom based on chlorophyll a or cell abundance alone. Current monitoring prog ra ms for toxic cyanobacterial blooms have focused on production of microcystins. In Lak e

Champlain, anatoxin-a was distributed widely and was detected not only in s outh e rn

Lake Champlain where Anabaena and Aphanizomenon species dominated, but also in northern Lake Champlain, especially in St. Albans Bay, where Microcystis dominated.

The 2002 dog fatalities occurred in Missisquoi Bay at a time when there were high levels of microcystins in the water (Satchwell et al. 2005b) suggesting microcystin intoxication; however the detection of anatoxin-a at a concentration greater than 1 μg L-1 in th is bay raises the possibility of future anatoxin-a poisoning in Missisquoi Bay. This issue needs to be addressed in future monitoring efforts.

175 During this 6-year period, HPLC-FD was used for routine monitoring of anatoxin- a. Positive samples were later analyzed by LCMS for confirmation. The reported interference from phenylalanine could be effectively eliminated by both analytical techniques. The HPLC-FD technique was less expensive, more sensitive and more selective than LCMS for the detection of anatoxin-a. However, the HPLC-FD technique is time consuming and is not suitable for high volume or in situ monitoring. Currently an immunoassay-based analytical technique is under development, this real-time or near real-time monitoring technique would provide an additional monitoring tool for lake managers.

In most cases, the anatoxin-a levels present in Lake Champlain were at concentrations lower than 0.1 μg L-1. These concentrations are difficult to detect using standard sampling techniques (1—2 L grab samples) and LCMS techniques. These anatoxin-a levels would be negligible for human and animal health per se. However, vigilance is still necessary in these cases. Anatoxin-a is not stable in aquatic environment

(Stevens and Krieger 1991; Rapala and Sivonen 1998) (Chapter 3) and the formation of several degradation products of anatoxin-a have been reported elsewhere (James et al.

1998; James et al. 2005). Anatoxin-a rapidly disappear from surface waters and current levels may not accurately reflect past risks. Interestingly, neither of the two known degradation products of anatoxin-a, dihydroanatoxin-a and epoxyanatoxin-a, were found in any samples. This suggests there may be additional degradation products of anatoxin-a in natural aquatic environments. We were also concerned about the presence of possible precursors to anatoxin-a such as the recently reported anatoxin-a carboxylate (Selwood et al. 2006), as these may reflect a potential pool of toxin waiting to be released under the

176 proper environmental conditions. An examination of those samples run by LCMS for the presence of this precursor showed that this anatoxin-a carboxylate was either not present in the methanolic extracts from these samples or present at concentrations lower than the detection limit of LCMS.

Microcystins are also common in Lake Champlain and molecular studies indicate that Microcystis and other cyanobacterial species capable of producing microcystins are distributed throughout Lake Champlain (Mihuc et al. 2005; Satchwell et al. 2005a).

However, microcystins are primarily observed in northern regions of Lake Champlain.

Microcystis blooms occurred consistently and dominated the summer cyanobacterial blooms in Missisquoi Bay in recent years, and have often lasted throughout the entire summer-fall season. Microcystins were commonly measured at concentrations above the

WHO drinking water guideline of 1 μg L-1 in these regions (Boyer et al. 2004; Satchwell et al. 2005a; Watzin et al. 2005). This production of microcystins often coincided with high Microcystis densities (Mihuc et al. 2005) and molecular analysis of the algal populations indicates a widespread abundance of the Microcystis mcy biosynthetic genes.

(Satchwell et al. 2005a; 2005b). In the Main Lake and southern part of Lake Champlain, microcystins were only intermittently detected and the concentrations were significantly lower than those in Missisquoi Bay despite the presence of microcystin-producing cyanobacteria. In contrast, anatoxin-a was more commonly found in sites outside

Missisquoi Bay; it was found in all basins except in South Lake where we only collected two samples, and was found at sites with relatively low chlorophyll a content. Based on the rare occurrence of anatoxin-a in the lake, the common presence of potential producers such as Anabaena spp. and Aphanizomenon spp., and the lack of correlation with

177 indicators of total algal biomass such as chlorophyll a, it is likely that anatoxin-a is produced by a minor member of the phytoplankton flora and that most blooms in this lake are not toxic for anatoxin-a. Identification of the toxicogenic blooms is hampered by their spotty occurrence and the lack of molecular tools such as those used for microcystins. Unfortunately, anatoxin-a may re-occur in the lake when the favorable conditions for toxin production or growth of this minor species occur. Therefore, these

“rare” anatoxin-a toxic events necessitate continuous monitoring of the lake.

4.3.5. ACKNOWLEDGEMENTS

This study was supported by the National Oceanic and Atmospheric Agency

Coastal Oceans Program through their MERHAB-LGL project # NA160 P2788 and New

York Sea Grant Award NA16R61645. Special thanks to Mike Satchwell, Amber Hotto,

Elisabeth Konopko for their help in microcystins analysis, samples collection and water chemistry analysis. Professor Tim Mihuc (SUNY-Plattsburg) and Professor Mary

Watzin (University of Vermont) provided assistance with sample collection.

4.3.6. REFERENCES

Boyer, G. L., Watzin, M. C., Shambaugh, A. D., Satchwell, M. F., Rosen, B. H., and Mihuc, T. 2004. The occurrence of cyanobacterial toxins in Lake Champlain. In Lake Champlain: Partnership and Research in the New Millennium, edited by T. Manley, P. L. Manley and T. Mihuc: Kluwer Academic/Plenum Publishers. p241- 257.

Brown, E. A., Duchovnay, A., McIntosh, A., Shambaugh, A., and Williams, A. 1991. 1991 Lake Champlain biomonitoring report. Vermont Water Resources and Lake Studies Center. School of Natural Resources, University of Vermont, 54 pp.

178 Carmichael, W. W. 2001. The cyanotoxins-bioactive metabolites of cyanobacteria: occurrence, ecological role, toxanomic concerns and effects on humans. Journal of Phycology 37 (3):9-9.

Carmichael, W. W., Azevedo, S. M. F. O., An, J. S., Molica, R. J. R., Jochimsen, E. M., Lau, S., Rinehart, K. L., Shaw, G. R., and Eaglesham Geoffrey, K. 2001. Human fatalities from cyanobacteria: chemical and biological evidence for cyanotoxins. Environmental Health Perspectives 109 (7):663-668.

Carmichael, W. W., Yuan, M., and Friday, C. F. 2004. Human mortality from accidental ingestion of toxic cyanobacteria - a case re-examined. In Sixth Internatioanl Conference on Toxic Cyanobacteria, Bergen, Norway. June 21-27.

Codd, G. A., Morrison, L. F., and Metcalf, J. S. 2005. Cyanobacterial toxins: risk management for health protection. Toxicology and Applied Pharmacology 203:264-272.

Codd, G. A., Ward, C. J., and Bell, S. G. 1997. Cyanobacterial toxins: occurrence, modes of action, health effects and exposure routes. In Applied Toxicology: Approaches Through Basic Science, Archives of Toxicology Supplement. 19, edited by J. P. Seiler and E. Vilanove. Berlin: Springer. p399-410.

Edwards, C., Beattie, K. A., Scrimgeour, C. M., and Codd, G. A. 1992. Identification of anatoxin-a in benthic cyanobacteria (blue-green algae) and in associated dog poisonings at Loch Insh, Scotland. Toxicon 30 (10):1165-1175.

Furey, A., Crowley, J., Hamilton, B., Lehane, M., and James Kevin, J. 2005. Strategies to avoid the mis-identification of anatoxin-a using mass spectrometry in the forensic investigation of acute neurotoxic poisoning. Journal of Chromatography A 1082:91-97.

Gugger, M., Lenoir, S., Berger, C., Ledreux, A., Druart, J.-C., Humbert, J.-F., Guette, C., and Bernard, C. 2005. First report in a river in France of the benthic

179 cyanobacterium Phormidium favosum producing anatoxin-a associated with dog neurotoxicosis. Toxicon 45 (7):919-928.

Haider, S., Naithani, V., Viswanathan, P. N., and Kakkar, P. 2003. Cyanobacterial toxins: a growing environmental concern. Chemosphere 52:1-21.

James, K. J., Crowley, J., Hamilton, B., Lehane, M., Skulberg, O., and Furey, A. 2005. Anatoxins and degradation products, determined using hybrid quadrupole time- of-flight and quadrupole ion-trap mass spectrometry: Forensic investigations of cyanobacterial neurotoxin poisoning. Rapid Communications in Mass Spectrometry 19 (9):1167-1175.

James, K. J., Furey, A., Sherlock, I. R., Stack, M. A., Twohig, M., Caudwell, F. B., and Skulberg, O. M. 1998. Sensitive determination of anatoxin-a, homoanatoxin-a and their degradation products by liquid chromatography with fluorimetric detection. Journal of Chromatography, A 798 (1 + 2):147-157.

Jokela, W. E., Clausen, J. C., Meals, D. W., and Sharpley, A. N. 2004. Effectiveness of agricultural best management practices in reducing phosphorus loading to Lake Champlain. In The Science and Management of Soil and Water: Lake Champlain. p39-53.

Krienitz, L., Ballot, A., Kotut, K., Wiegand, C., Putz, S., Metcalf, J. S., Codd, G. A., and Pflugmacher, S. 2003. Contribution of hot spring cyanobacteria to the mysterious deaths of Lesser Flamingos at Lake Bogoria, Kenya. FEMS Microbiology Ecology 43 (2):141-148.

Metcalf, J. S., Morrison, L. F., Krienitz, L., Ballot, A., Krause, E., Kotut, K., Putz, S., Wiegand, C., Pflugmacher, S., and Codd, G. A. 2006. Analysis of the cyanotoxinx anatoixn-a and microcystins in Lesser Flamingo feathers. Toxicological and Environmental Chemistry 88:159-167.

Mihuc, T. B., Boyer, G. L., Satchwell, M. F., Pellam, M., Jones, J., Vasile, J., Bouchard, A., and Rob, B. 2005. 2002 phytoplankton community composition and

180 cyanobacterial toxins in Lake Champlain, U.S.A. Verhandlungen - Internationale Vereinigung fuer Theoretische und Angewandte Limnologie 29:328-333.

Muenscher, W. G. 1930. Plankton studies in Lake Champlain Watershed. In A Biological Survey of the Champlain Watershed: Supplemental to Nineteenth Annual Report, 1929. Albany: New York Conservation Department. p164-185.

Myer, G. E., and Gruendling, G. K. 1979. Limnology of Lake Champlain, Lake Champlain Basin study. New England River Basins Commission, 407 p.

Rapala, J., and Sivonen, K. 1998. Assessment of environmental conditions that favor hepatotoxic and neurotoxic Anabaena spp. strains cultured under light limitation at different temperatures. Microbial Ecology 36 (2):181-192.

Rosen, B. H., Shambaugh, A., Ferber, L., Smith, F., Watzin, M. C., Eliopoulos, C., and Stangel, P. 2002. Evaluation of potential blue-green algae toxins in Lake Champlain, summer 2000 for the Lake Champlain Basin Program and the Centers for Disease Control and Prevention. 27 p.

Satchwell, M. F., Hotto, A., Damon, R. M., and Boyer, G. L. 2005a. Distribution and molecular analysis of microcystins in Lake Champlain. Abstract. In 48th Annual Conference of the International Association of Great Lake Research, Ann Arbor, Michigan. May 23-27, 1 p.

Satchwell, M. F., Hotto, A., Yang, X., Damon, R. M., Mihuc, T., and Boyer, G. L. 2005b. Cyanobacterial toxins in Lake Champlain - A five year review. In The Third Symposium on Harmful Algae in the US, Monterey, CA. October 2-7, 1 p.

Selwood, A. I., Holland, P. T., Wood, S. A., Smith, K. F., and McNabb, P. S. 2006. Production of anatoxin-a and a novel biosynthetic precursor by the cyanobacterium Aphanizomenon issatschenkoi. Environmental Science & Technology: In print.

Shambaugh, A., Duchovnay, A., and McIntosh, A. 1999. A survey of Lake Champlain's plankton. In Lake Champlain in Transition: from Research toward Restoration,

181 edited by T. O. Manley and P. L. Manley. Washinton, DC: Americal Geophysical Union. p41-67.

Shanley, J. B., and Denner, J. C. 1999. The hydrology of the Lake Champlain basin. In Lake Champlain in Transition: from Research toward Restoration, edited by T. O. Manley and P. L. Manley. Washington, DC: Americal Geophysical Union. p41- 67.

Stevens, D. K., and Krieger, R. I. 1991. Stability studies on the cyanobacterial nicotinic alkaloid anatoxin-a. Toxicon 29 (2):167-179.

Watzin, M. C., Miller, E. B., Shambaugh, A. D., and Kreider, M. A. 2005. Application of the WHO alert level framework to cyanobacterial monitoring of Lake Champlain, Vermont. Environmental Toxicology 21:278-288.

Watzin, M. C., Shambaugh, A., Brines, E., and Boyer, G. L. 2003. Monitoring and evaluation of cyanobacteria in Lake Champlain (Summer 2002). Report to Lake Champlain Basin Program. Rubenstein Ecosystem Science Laboratory, University of Vermont and SUNY-ESF, 38 pp.

182 4.4. Occurrence and distribution of anatoxin-a in other New York State lakes

4.4.1. INTRODUCTION

New York State is rich in water resources, and cyanobacteria are common components of the natural phytoplankton communities in its freshwaters (Greeson 1971;

Munawar and Munawar 1996). In recent decades, eutrophication has led to increased nutrients levels in various natural freshwater systems and consequently has led to excessive proliferation of cyanobacteria and consequent bloom formation.

Cyanobacterial blooms are not only associated with taste and odor problems, but some blooms can also produce various potent toxins. These cyanobacterial toxins were associated with numerous cases of animal mortalities, and as well as human illness and fatalities across the world (Codd 1995; Ueno et al. 1996; Jochimsen et al. 1998; Chorus and Bartram 1999; Falconer 1999; Carmichael 2001; Azevedo et al. 2002).

In New York State, there were also several cases of animal fatalities associated with toxic cyanobacterial blooms. In 1999, 2000 and 2002, several dogs died after swimming in Lake Champlain and these incidents were later attributed to the neurotoxic anatoxin-a (1999 and 2000) and the hepatotoxic microcystins (2002) (Yang et al. 2001;

Boyer et al. 2004; Watzin et al. 2005). More recently, in June 2004, a toxic cyanobacterial bloom in Lake Neatahwanta near Fulton claimed the lives of a dog and several geese (Boyer 2006). These toxic incidents, and the widespread abundance of cyanobacteria in New York State, have led to beach closings, and raised public concern over the potential health effects of cyanobacterial toxins on environmental and human health. In New York State, Microcystis spp. have become increasingly abundant in freshwater systems following the invasion of zebra mussels and resulted in bloom

183 formation. As a result, cyanobacterial toxins, such as microcystins are more frequently encountered. Current monitoring efforts for cyanobacterial toxins now focus on these hepatotoxic microcystins. In contrast, the occurrence of neurotoxins such as anatoxin-a has not been significantly studied despite the observed toxic incidents. Anatoxin-a is a potent neurotoxin produced by a number of cyanobacterial species, including Anabaena,

Aphanizomenon, Planktothrix and Microcystis (Bumke-Vogt et al. 1999; Sivonen 2000;

Viaggiu et al. 2004), all of which are common in New York State. The potential toxin- producing genera are all common members of the New York State phytoplankton flora.

Here we report on the distribution of anatoxin-a in New York State water systems and the occurrence and distribution of anatoxin-a in several selected lakes over a 6-year period. These lakes include three upstate lakes, Onondaga Lake, Oneida Lake and Lake

Neatahwanta, and one downstate lake, Lake Agawam. All these lakes are eutrophic and have well-established annual cyanobacterial blooms in summer and fall. Onondaga Lake is located along the northern edge of the City of Syracuse in Onondaga County, New

York. It was heavily contaminated with industrial pollutants and, despite extensive cleanup efforts, is still experiencing serious contamination due to the presence of mercury, organic hydrocarbons, excessive nutrients and a high level of salinity (Effler 1996).

Cyanobacterial blooms in Onondaga Lake are mainly composed of Aphanizomenon,

Planktothrix/Oscillatoria and Microcystis (OCDWEP 2004). Oneida Lake is located north of Syracuse and is a shallow (mean depth of 7 m), well-mixed eutrophic lake that is isothermal during summertime (Greeson 1971; Hotto et al. 2005). It has a history of summer blooms dominated by the cyanobacterial genera Anabaena, Aphanizomenon,

Microcystis, Gleotrichia, and Lyngbya (Greeson 1971) and microcystins have been

184 detected in this lake (Hotto et al. 2005). Lake Neatahwanta is a small lake located in

Fulton, New York. It is used for recreational purposes such as fishing and boating.

Cyanobacterial blooms occur annually starting from late-spring to early fall and species of Anabaena, Aphanizomenon, and Microcystis are commonly identified. Microcystins have been detected at concentrations as high as 4,000 μg L-1 from a surface scum

(Satchwell, personal communication). Following the animal fatalities during the summer of 2004, the local health department has posted advisories against swimming and public bathing in Lake Neatahwanta. Lake Agawam is a small shallow eutrophic lake located on Long Island, NY. It experiences dense blooms of Microcystis, Anabaena and

Aphanizomenon sp. in summer and microcystins occur in the lake at concentrations up to

20 μg L-1 (Gobler et al. 2005, 2006).

4.4.2. MATERIAL AND METHODS

Between 2000 and 2005, more than 1000 water samples were collected from over

75 lakes, rivers, reservoirs and other water bodies (Appendix 4). Water sample (3-20 L), collected from a depth of 1.0 m, were filtered on site through a 90-mm glass fiber filter

(Whatman 934AH) and immediately stored on dry ice for lab analysis. In Onondaga

Lake, the bulk of samples were collected from the north and south basins in 2002 and

2003 while selected shoreline locations were sampled in 2000 and 2001. Oneida Lake was sampled at six offshore stations weekly in the summer from 2000 to 2005 and Lake

Neatahwanta was sampled weekly from late spring (May) to early fall (October) in 2004 and 2005 at three separate nearshore sites (referred as “Campground”, “Bullhead Point” and “Beach Shore”) (Table 3.2 in Chapter 3). Lake Agawam samples were collected by

185 Professor Chris Gobler at SUNY-Stony Brook and sent to SUNY-ESF for toxin analysis.

All four lakes were sampled once every one to two weeks during the summer. Surface samples of cyanobacteria were collected with a 10-μm plankton net and stored in 0.2% v/v glutaraldehyde for laboratory phycological analysis.

-20°C until analyses. Anatoxin-a and two of its degradation products, dihydroanatoxin-a and epoxyanatoxin-a, were analyzed using high performance liquid chromatography with fluorescence detection (HPLC-FD) using the method of James et al. (1998) with modifications. Extracts (1 mL) were evaporated to dryness and reconstituted in 100 μL of 0.1 M borate buffer containing 0.6 mM methyl pipecolinate as an internal standard.

Fifty μL of the fluorescent reagent, 5 mM NBD-F (7-fluoro-4-nitro-2,1,3-benzoxadiazole,

TCI America, CA) was added and allowed to react at room temperature for 1 hr. The reaction was stopped by addition of 1 M HCl, centrifuged at 15,000 g and the supernatant analyzed by HPLC-FD using an Ace 5μ C18 column (4.6 x 250 mm; 55% aqueous acetonitrile, Shimadzu RF-535 fluorescence detector: 470 nm excitation, 530 nm emission). Positive samples by HPLC-FD were again analyzed by electrospray ionization mass spectrometry (LCMS). Early extracts (Before 2001) were analyzed by

LCMS after derivatization with NBD-F using a Luna 5μ C18 column (4.6 x 150 mm,

55% aqueous acetonitrile with 1% acetic acid); later samples (2001-2003) were analyzed directly without derivatization using the Ace C18 column (30 min gradient of 20-50% aqueous acetonitrile containing 0.1% TFA). Ion intensities from 100 to 500 amu were

186 collected and the protonated molecular ion of anatoxin-a (m/z 166) was extracted from the total ion chromatograph. Anatoxin-a was quantified using a standard curve generated from commercially available anatoxin-a chloride (Biomol Inc, Plymouth Meeting, PA).

Surface samples for phycological analysis were examined using a phase-contrast light microscope under 250 x magnification to identify species according to Whitford and

Schumacher (1969). Cyanobacterial colonies and filaments were estimated using a

Palmer-Malony counting chamber. Only samples collected from Onondaga Lake are reported here.

4.4.3. RESULTS

More than 75 New York water bodies have been tested for the occurrence of anatoxin-a during the monitoring period of 2000 to 2005 (Appendix 4). Most of these water systems were tested in selected years only. Among these lakes, rivers and reservoirs, anatoxin-a was detected in 8 lakes and at concentrations ranging from 0.001 to

17.3 μg L-1. Here, only Onondaga Lake, Oneida Lake, Lake Neatahwanta and Lake

Agawam were selected and monitored more systematically to determine the periodicity of anatoxin-a occurrence.

Occurrence of anatoxin-a in Onondaga Lake

A total of 96 water samples were collected from Onondaga Lake between 2000 and 2003 and analyzed for anatoxin-a by HPLC-FD. Anatoxin-a was detected in 23 of these samples at concentrations greater than 0.001 μg L-1. Six samples exhibited

187 Table 4.8. The Occurrence of anatoxin-a in Onondaga Lake between 2000 and 2003.

Year of analysis

2000 2001 2002 2003 Total

Samples analyzed by HPLC-FD

Number analyzed 13 3 45 35 96

Number (%) with 0 (0%) 0 (0%) 11 (24%) 12 (34%) 23(24%)

>0.001 ATX μg L-1

Samples analyzed by LCMS

Number analyzed 0 0 11 12 23

Number with >0.003 - - 11 6 17

ATX μg L-1

Number with >0.1 μg 0 0 0 6 6 ATX L-1

Maximum concentration < 0.001 < 0.001 0.056 17.3 17.3

(μg ATX L-1)

188 anatoxin-a at concentrations exceeding 0.1 μg L-1 and the concentration in 4 of these 6 samples exceeded 1 μg L-1 (Table 4.8). The presence of anatoxin-a was confirmed in 17 of these 23 positive samples including all 6 samples with higher anatoxin-a concentrations. Neither of the two known anatoxin-a degradation products

(dihydroanatoxin-a and epoxyanatoxin-a) was detected in any of the samples analyzed.

In 2000 and 2001, samples were collected from near-shore locations and anatoxin-a was not detected. In later years, anatoxin-a was detected in samples collected from two offshore sites located in the north and south basin of Onondaga Lake (North

Deep and South Deep). Anatoxin-a was observed in 24% of the 2002 samples and 17% of the 2003 samples. In both years, the majority of anatoxin-a-containing samples were collected in late summer (August 2002 and September 2003) (Figure 4.30). Anatoxin-a distributed evenly in abundance between North and South Deep, and was detected on same days and at comparable concentrations at both sampling sites. The concentrations of anatoxin-a detected in 2003 were nearly 100-fold higher than those in 2002, and all the positive samples in 2003 had anatoxin-a concentrations greater than 0.1 μg L-1. Figure

4.31 shows the HPLC-FD and LCMS chromatographs of one sample collected in summer

2003.

The relationship between concentration of anatoxin-a and cyanobacterial density, as estimated by phycocyanin, is shown in Figure 4.32. No positive correlation was observed when samples from all years were examined. However, a weak correlation was observed for the offshore samples collected in 2003 (Figure 4.32) when the concentration of anatoxin-a peaked concurrently with a peak of phycocyanin (Figure 4.33). Following

189 100 0.015 )

north deep 2002 -1 south deep 0.013 g L

2002 μ 80

0.011 ( tion

nato 60 0.009

0.007 enc 40 rr e of a xin-a

ccu 0.005 anato

% O 20 0.003 Maximum concentra ixn-a 0 0.001 06/2002 07/2002 08/2002 09/2002 10/2002

100 20 ) north deep 2003 -1 south deep g L

2003 μ

i 80 16 anatox n-a 60 12

urrence of urrence 40 8 Occ %

20 4 Maximum anatoixn-a concentration ( concentration anatoixn-a Maximum 0 0 06/2003 07/2003 08/2003 09/2003 10/2003 Sampling month

Figure 4.30. Occurrence and maximum concentrations of anatoxin-a at the two sampling sites (North Deep and South Deep) in Onondaga Lake in 2002 and 2003. Bars indicate number of samples containing anatoxin-a, dots indicate maximum anatoxin-a concentration. N = 4 for each month.

190

1e+6

ATX-NBD

8e+5

6e+5

Response 4e+5

2e+5

024681012141618 Retention time (min) Figure 4.31a. The HPLC-FD chromatograph of an anatoxin-a-containing water sample collected from the southern basin of Onondaga Lake on September 8th, 2003. The sample was first derivatized with the fluorescent agent NBD-F (see Chapter 3 for details).

191 1.6e+6

1.4e+6

1.2e+6

ATX 1.0e+6

8.0e+5

Response 6.0e+5

4.0e+5

2.0e+5

0.0 02468101214 Retention time (min) 120

166.1 100

Abundance (%) Abundance 40

20 207.1

0 50 100 150 200 250 300 m/z (amu)

Figure 4.31b. LCMS chromatograph (upper) of an anatoxin-a-containing water sample collected from the southern basin of Onondaga Lake on September 8th, 2003. The lower figure is mass spectrum of the peak at 10.78 min showing the presence of anatoxin-a with a m/z 166. The peak at 12.36 min was from an impurity in this sample.

192 a rapid decrease in phycocyanin, concentrations of anatoxin-a a reached a maximum and soon plummeted to levels below the detection limit.

The cyanobacterium responsible for toxin production was unknown. Visual examination showed that the dominant species in Onondaga Lake during summertime in both 2002 and 2003 was Aphanizomenon spp. and this species accounted for more than

50% of the phytoplankton biomass. Planktothrix was commonly present while

Microcystis and Anabaena existed at much lower abundance. During the toxic cyanobacterial bloom in September 2003, Aphanizomenon accounted for more than 99.

9% of total phytoplankton biomass (> 200,000 cells per milliliter) at both north and south basins of Onondaga Lake. Planktothrix, Microcystis, and Anabaena contributed less than

0.1%. In late September, cell density of Aphanizomenon drastically decreased (ca.

20,000 cells per milliliter). This decrease was in accordance with the decrease and disappearance of anatoxin-a. Planktothrix abundance increased significantly at that time and co-dominated with Aphanizomemon. Microcystis and Anabaena were still present in low abundance (< 0.1%).

Occurrence and distribution of anatoxin-a in Oneida Lake, Lake Neatahwanta and

Lake Agawam

Oneida Lake was more extensively monitored than Onondaga Lake with more than 350 water samples collected between 2000 and 2005 and analyzed for anatoxin-a.

Anatoxin-a was detected at trace concentrations (greater than 0.001 μg L-1 yet significantly lower than 0.1 μg L-1) in 25 samples by HPLC-FD. These samples were

193 20 18 a 16

) 14 -1

g L 12 μ 10 8 6

Anatoxin-a ( 4 2 0

0102030 0.06 b 0.05 )

-1 0.04 g L μ 0.03

0.02

Anatoxin-a ( 0.01

0.00

0102030 20 c r2 = 0.34 16 ) -1 12 g L μ

Anatoxin-a ( 4

0102030 Phycocyanin (μg L-1)

Figure 4.32. Relationship between concentration of anatoxin-a and the total cyanobacterial biomass as estimated by the pigment phycocyanin in Onondaga Lake (a) during the entire monitoring period, (b) in 2002 and (c) in 2003.

194 14 2.5 Anatoxin-a North Deep 12 Phycocyanin 2.0

10 ) ) -1

-1 1.5

8 g L g L μ μ (

a 6 1.0

Anatoxin- ( 0.5 Phycocyanin 2

0.0 0

Jun-02 Oct-02 Feb-03 Jun-03 Oct-03 20 3.5

18 Anatoxin-a South Deep Phycocyanin 3.0 16 2.5 14 ) ) -1 -1 12

2.0 g L g L μ μ 10

-a ( -a 1.5 8 cocyanin ( cocyanin atox 6 1.0 An in Phy 4 0.5 2 0.0 0

Jun-02 Oct-02 Feb-03 Jun-03 Oct-03

Figure 4.33. Changes in concentrations of anatoxin-a and phycocyanin at the two sampling sites (North Deep and South Deep) in Onondaga Lake in summer and fall 2002 and 2003.

195 widely distributed among the 6 sampling stations and were collected in different months and years. Later analysis by LCMS confirmed the presence of anatoxin-a in only 4 of these 25 samples due to their low concentrations (Table 4.9). These 4 samples were also from different sampling stations in different years. Throughout the 6-year monitoring, despite the observation that chlorophyll a commonly occurred at concentrations higher than 5 μg L-1 and could even be as high as 40 μg L-1, anatoxin-a rarely occurred and the concentrations of anatoxin-a and chlorophyll a were not positively correlated (Figure 4.34, r2 = 0.07).

The monitoring for anatoxin-a in Lake Neatahwanta started in 2004. During the sampling seasons in 2004 and 2005, 132 samples were collected at three nearshore sites as described previously. Detectable concentrations (> 0.001 μg L-1) of anatoxin-a were measured in 49 of these samples by HPLC-FD (Table 4.10), these included four samples with anatoxin-a at concentrations greater than 0.1 μg L-1. The presence of anatoxin-a was confirmed by LCMS in 44 of these 49 samples including all 4 samples with higher concentrations. Occurrence of anatoxin-a differed significantly between 2004 and 2005

(Figure 4.35). Anatoxin-a occurred more frequently and lasted much longer in 2005.

However, anatoxin-a occurred at significantly lower levels in comparison with those measured in 2004. Anatoxin-a was mostly detected in mid- to late-summer in 2004 while it was consistently detected from late July till late October in 2005. In 2004, besides the observed anatoxin-a in water samples, low concentrations of anatoxin-a (< 0.1 μg g-1 wet weight) were also detected in animal tissue samples in early June (Table 4.10). These animal tissue samples were from a dog and geese that died on Lake Neatahwanta during a toxic cyanobacterial bloom. However, anatoxin-a was not detected in water samples

196 collected at the same time from the lake. Spatially, anatoxin-a was detected concurrently at all three nearshore sampling locations, and at comparable concentrations and abundances (Figure 4.35). At all three sampling sites, chlorophyll a concentrations greater than 50 μg L-1 were measured throughout the sampling season (Figure 4.36), and no positive correlation was observed between the concentration of anatoxin-a and chlorophyll a (data not shown, r2 < 0.01).

We received and analyzed 136 water samples from Lake Agawam which were collected in summer 2003 and 2004. Anatoxin-a was detected in 21 of these samples at concentrations greater than 0.001 μg L-1 by HPLC-FD (Table 4.11). Among these positive samples, 19 contained anatoxin-a greater than 0.1 μg L-1. When LCMS was used to confirm the presence of anatoxin-a, anatoxin-a was identified in 20 of the 21 samples.

The occurrence of anatoxin-a varied between the two monitoring years (Figure 4.37).

Anatoxin-a occurred more frequently in 2003 (28%) than in 2004 (4%) and lasted much longer. Yet the peak concentrations of anatoxin-a measured in 2003 were significantly lower (p < 0.001, Student’s t-Test) than those observed in 2004. In 2004, all the anatoxin- a-containing samples exhibited anatoxin-a at concentrations greater than 1.0 μg L-1. In both years, the majority of anatoxin-a was observed in August and the concentration of anatoxin-a reached maximum in this month. In this eutrophic lake, chlorophyll a concentrations exceeding 50 μg L-1 were commonly measured in summer and fall seasons (data not shown). No positive correlation was observed between the concentrations of anatoxin-a and chlorophyll a.

197

Table 4.9. Occurrence of anatoxin-a in Oneida Lake from 2000 to 2005. Only those anatoxin-a samples positive (>0.001 μg L-1) by HPLC-FD were analyzed by LCMS for confirmation.

Year of analysis

2000 2001 2002 2003 2004 2005 Total

Samples analyzed by HPLC-FD

Number collected 22 49 89 95 74 37 366

Number >0.001 μg ATX L-1 0 11 4 7 3 0 25

Samples analyzed by LCMS

Number analyzed 0 11 4 7 3 0 25

Number >0.003 μg ATX L-1 - 0 1 2 1 0 4

Number (%) >0.1 μg ATX L-1 - 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

Maximum concentration (μg <0.01 0.007 0.008 0.019 0.023 <0.01 0.023

ATX L-1)

198

0.025 r2=0.07

0.020 )

-1 0.015 g L μ

0.010 n-

Anatoxi a ( 0.005

0.000

0 1020304050

Chlorophyll a (μg L-1) Figure 4.34. Concentration of anatoxin-a confirmed by LCMS versus concentration of chlorophyll a in Oneida Lake (n = 295).

199

Table 4.10. Occurrence of anatoxin-a in Lake Neatahwanta during summer 2004 and

2005. Only those anatoxin-a samples positive (>0.001 μg L-1) by HPLC-FD were analyzed by LCMS for confirmation.

Samples analyzed by HPLC-FD Samples analyzed by LCMS Number Number (%)1 Number (%)1 Number Number (%)2 collected >0.001 μg L-1 >0.1 μg L-1 analyzed >0.003 μg L-1

2004 Campground 17 4 (24%) 1 (6%) 3 3 (100%) Bullhead Point 17 5 (29%) 1 (6%) 4 4 (100%) Beach Shore 16 5 (31%) 1 (6%) 4 3 (75%) Animal tissue 6 2 (33%) 0 (0%) 2 2 (100%) Total 56 16 (29%) 3 (5%) 13 12 (92%)

2005 Campground 23 11 (48%) 0 (0%) 11 11 (100%) Bullhead Point 23 11 (48%) 0 (0%) 11 11 (100%) Beach Shore 23 11 (48%) 0 (0%) 11 11 (100%) Others 7 0 (0%) - 0 - Total 76 33 (43%) 0 (0%) 33 33 (100%) 1 % is based on the number of total samples analyzed by HPLC-FD; 2 % is based on the number of samples analyzed by LCMS.

200 0.35 Campground 2004 0.30 Bullhead Point Beach Shore

0.25 ) -1 0.20 g L μ

0.15

0.10

Anatoxin-a ( Time of animal fatalities 0.05

0.00

Jun Jul Aug Sep Oct 0.07 Campground 2005 0.06 Bullhead Point Beach Shore

0.05 ) -1 0.04 g L μ

0.03

0.02 Anatoxin-a ( Anatoxin-a

0.01

0.00

May Jun Jul Aug Sep Oct Nov

Figure 4.35. Variation in anatoxin-a concentration over time at the three sampling sites on Lake Neatahwanta, New York in summer 2004 and 2005. The arrow in the upper

Figure indicates the time when animal fatalities occurred on Lake Neatahwanta.

201

3000 2000

1000 )

-1 500 400 g L

μ 300 200 ll a ( ll a

phy 100 loro

Ch 50 40 30 Campground 20 Bullheak Point Beach Shore 10 Jun Jul Aug Sep Oct

Figure 4.36. Variation in chlorophyll a concentration at the three sampling sites on Lake

Neatahwanta in summer 2 004. Chlorophy ll a concentrations are plotted on a logarith mic scale.

202

Table 4.11. Occurrence of anatoxin-a in Lake Agawam, Long Island, New York. Only those anatoxin-a samples positive (>0.001 μg L-1) by HPLC-FD were analyzed by LCMS for confirmation.

Samples analyzed by HPLC-FD Samples analyzed by LCMS Number Number (%)1 Number (%)1 Number Number (%)2 collected >0.001 μg L-1 >0.1 μg L-1 analyzed >0.003 μg L-1

2003 60 18 (30%) 15 (25%) 18 17 (94%) 2004 76 3 (4%) 3 (4%) 3 3 (100%) Total 136 21 (15%) 18 (13%) 21 20 (95%) 1 % is based on number of samples analyzed by LCMS;

2 % is based on the total number of samples analyzed by HPLC-FD.

203

3.0

2.5 )

-1 2.0 g L μ

1.5

1.0 Anatoxin-a ( Anatoxin-a

0.5

0.0 Jun-2003 Oct-2003 Feb-2004 Jun-2004 Oct-2004

Figure 4.37. Distribution of anatoxin-a in Lake Agawam, Long Island, New York between 2003 and 2004. The multiple points on same date indicated samples from different sampling sites. Chlorophyll a was measured at concentrations greater than 50

μg L-1 throughout the summers.

204 4.4.4. DISCUSSION

In this study, anatoxin-a was detected in only 10% of the New York lakes examined and was not detected at most of the sampling sites. This low occurrence and distribution of anatoxin-a might suggest these New York State lakes are not currently at risk for this neurotoxin, and anatoxin-a was not present in most of the lakes. However these lakes were not monitored extensively, and it is equally possible that the spatial and temporal variation in anatoxin-a production may result in toxic cyanobacterial blooms being overlooked. Only four lakes in this study were monitored with enough spatial and temporal density so as to detect a rare occurrence of an anatoxin-a producing species. In all four lakes that were extensively examined, anatoxin-a was detected in each lake though with different frequ ency. In Onondaga Lake, the occurrence of several cyanobacterial “blooms” was indicated by an increase in the phycocyanin concentrations throughout the summer of 2002 and 2003. In both years, anatoxin-a was detected shortly after formation of a bloom and disappeared soon after the dissappearan ce of the bloom, suggesting production of anatoxin-a might be closely associated with the specific late- summer cyanobacterial bloom events. Phycological analysis showed Aphanizomenon spp. was the dominating species in Onondaga Lake at both sampling locations (North and

South Deep) with comparable abundance in the late-summer of 2003. Similar observation have been reported by the Onondaga County Department of Water

Environment Protection (OCDWEP 2004). Aphanizomenon species have been previously identified as anatoxin-a producers (Rapala et al. 1993; Bumke-Vogt et al.

1999), and the co-occurrence of Aphanizomenon spp. and anatoxin-a at both sites in a bloom in September 2003 suggest that Aphanizomenon spp. might be associated with

205 anatoxin-a-production in this lake. Future isolation of Aphanizomenon from Onondaga

Lake and its cultivation in the laboratory would be necessary to determine whether this species produces anatoxin-a.

Although anatoxin-a was observed at very high levels (17 μg L-1) in this lake, the health risk associated with this toxin is limited. Onondaga Lake is considered highly polluted and only sparsely used for contact recreation. It is not used as a drinking water source and dogs and human exposure to toxic cyanobacteria would likely only occur during summer recreational activities. Other cyanobacterial toxins such as microcystins have also been detected in this lake. Microcystins concentrations also exceeded the

WHO advisory level of 1 μg L-1 in 2002, but were not detected in summer 2003 (Boyer et al. 2002) (Boyer, personal communication). While Aphanizomenon can produce other cyanobacterial toxins such as the neurotoxic anatoxin-a and saxitoxin, or the hepatotoxin cylindrospermopsin (Mahmood and Carmichael 1986; Rapala et al. 1993; Banker et al.

1997), it is generally not associated with microcystin production. Phycological analysis indicated that Planktothrix sp. and Microcystis sp. are major bloom-forming species in

Onondaga Lake in addition to Aphanizomenon sp. Both Microcystis and Planktothrix sp. have been identified as microcystin-producing species in New York lakes (Hotto et al.

2005; Rinta-Kanto et al. 2006), and are likely the source of microcystins in this lake.

Both of these two potential microcystin-producing species contributed only a small portion to the total cyanobacterial community during the summer seasons and the current risk from microcystins was small. However under situations where these species dominate the phytoplankton community, their production of microcystins may represent a health risk.

206 Oneida Lake also has a long history of nuisance cyanobacterial blooms (Mills and

Holeck 2001). Much larger than Onondaga Lake, it has a similar cyanobacterial community composed of Anabaena, Aphanizomenon, and Microcystis species with a trimodal trend in dominance (Hotto et al. 2005). All three dominant cyanobacterial species in Oneida Lake are potentially toxigenic. However, in contrast to Onondaga

Lake, anatoxin-a was rarely detected in water samples from Oneida Lake. During the 6- year monitoring period, anatoxin-a was detected in only ca 1% of the samples analyzed and at concentrations significantly lower than the cautionary level of 0.1 μg L-1. This suggests that, although species of Anabaena and Aphanizomenon are potential anatoxin-a producers, the summer blooms of these cyanobacteria in Oneida Lake did not produce anatoxin-a. Oneida Lake is heavily used for recreational purposes, but the rare occurrence and low concentrations of anatoxin-a suggest the potential health risk associated with this toxin would be minimal. Microcystins were much more commonly detected at concentrations exceeding the WHO advisory guidelines (Hotto et al. 2005) and probably represent a more significant risk. Molecular studies suggest that microcystins in Oneida Lake were produced by toxic Microcystis. Thus, the production of cyanobacterial toxins in Oneida Lake differed significantly from Onondaga Lake.

The monitoring of Lake Neatahwanta and Lake Agawam started much later (2004 and 2003 respectively) and were in response to animal fatalities. Samples collected from both lakes showed high concentrations (> 1 μg L-1) of anatoxin-a and microcystins.

Considering the consistent detection of cyanobacterial toxins in these lakes, health advisories are being posted regarding safe usage of the lake during summertime. The common cyanobacteria in Lake Neatahwanta, Microcystis, Anabaena and

207 Aphanizomenon, were again all potentially toxigenic and the actual species responsible for producing anatoxin-a and microcystins remains to be identified. A toxic incident in early summer 2004 resulted in deaths of a dog and several geese. Analysis of water samples collected soon after had microcystins at concentrations higher than 1 μg L-1, but anatoxin-a was not detected. Both anatoxin-a and microcystins were detected in animal tissue samples (Satchwell, personal communication), however the clinical symptoms of the intoxicated animal/birds suggested neurotoxicosis. Anatoxin-a would be a more likely responsible agent for this toxic incident than microcystins. Its measured toxicity was too low to account for the animal fatalities via consumption of contaminated lake water. Instead, the animal/birds may have ingested toxic cyanobacterial cells from a wind-induced scum. The monitoring in this study targeted particulate toxin and it is possible that the bloom decomposed prior to the time of sample collection. This could release any cell-bound particulate anatoxin-a into the surrounding water column. Despite the presence of cyanobacterial toxins in animal tissue samples, the possibility those fatalities were caused by some other agents cannot be excluded.

In Lake Agawam, anatoxin-a was measured in both years at concentrations exceeding the cautionary level of 0.1 μg L-1. Microcystins were also detected at the same time (Gobler et al. 2006). In comparison with the occurrence of anatoxin-a, microcystins occurred more frequently and lasted throughout the entire summer to fall seasons (Gobler et al. 2006). Algal biomass in this lake was again dominated by the three potentially toxic genera, Microcystis, Anabaena and Aphanizomenon and Gobler et al. (Gobler et al.

2006) identified Microcystis as the responsible agent for microcystin-production. The species responsible for anatoxin-a is unknown. The high levels of anatoxin-a found in

208 this lake suggest this toxin was actively synthesized during summer. The pattern of occurrence of anatoxin-a in 2003 matched that of microcystins, with the concentrations of both toxins peaking at similar times and then decreasing during July and August. This led to a strong correlation (r2 = 0. 89, data not shown) between the concentrations of these two toxins. These two toxins probably were produced by different species that co-varied in response to similar environmental conditions. Later in October, microcystins peaked once more while anatoxin-a was only detected at trace concentration. In summer 2004, the occurrence of these two toxins differed. Microcystins occurred throughout the entire sampling season while anatoxin-a was only detected in August. Gobler et al. (2006) proposed anatoxin-a could be produced by Anabaena and/or Aphanizomenon since both genera increased their densities in August. Yet further isolation and cultivation of these two genera from this lake is necessary to determine the actual source of anatoxin-a This identification of the anatoxin-a-producing cyanobacteria could help minimize the potential health risk associated with this toxin by avoiding usage of the lake during occurrence of a toxic cyanobacterial bloom.

4.4.5. ACKNOWLEDGEMENTS

This study was supported by the National Oceanic and Atmospheric Agency

Coastal Oceans Program through their MERHAB-LGL project # NA160 P2788, and New

York Sea Grant awards NA16R61645. Special thanks to Mike Satchwell and Amber

Hotto for microcystins analysis, and to Steve Ragonese, Elizabeth Konopko, Professor

Christopher Gobler at Stony Brook University and New York State Department of

Environmental Conservation for sample collection and water chemistry analysis.

209 4.4.6. REFERENCES

Azevedo, S. M. F. O., Carmichael, W. W., Jochimsen, E. M., Rinehart, K. L., Lau, S., Shaw, G. R., and Eaglesham, G. K. 2002. Human intoxication by microcystins during renal dialysis treatment in Caruaru--Brazil. Toxicology 181-182:441-446. Banker, P. D., Carmeli, S., Hadas, O., Teltsch, B., Porat, R., and Sukenik, A. 1997. Identification of cylindrospermopsin in Aphanizomenon ovalisporum (Cyanophyceae) isolated from Lake Kinneret, Israel. Journal of Phycology 33:613-616. Boyer, G. L. 2006. Cyanobacterial toxins in New York and the Lower Great Lakes ecosystems. Advances in Experimental Medicine & Biology Proceedings of the Interagency, International Symposium on Cyanobacterial Harmful Algal Blooms: In press. Boyer, G. L., Watzin, M. C., Shambaugh, A. D., Satchwell, M. F., Rosen, B. H., and Mihuc, T. 2004. The occurrence of cyanobacterial toxins in Lake Champlain. In Lake Champlain: Partnership and Research in the New Millennium, edited by T. Manley, P. L. Manley and T. Mihuc: Kluwer Academic/Plenum Publishers. p241- 257. Boyer, G. L., Yang, X., Szprygada, K., and Satchwell, M. F. 2002. A re-examination of the occurrence of cyanobacteria toxins in Onondaga Lake. In Fourth Annual Onondaga Lake Forum, Liverpool, NY. November 22, 1 p. Bumke-Vogt, C., Mailahn, W., and Chorus, I. 1999. Anatoxin-a and neurotoxic cyanobacteria in German lakes and reservoirs. Environmental Toxicology 14 (1):117-125. Carmichael, W. W. 2001. The cyanotoxins-bioactive metabolites of cyanobacteria: occurrence, ecological role, toxanomic concerns and effects on humans. Journal of Phycology 37 (3):9-9. Chorus, I., and Bartram, J., eds. 1999. Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring, and Management. London: E & FN Spon (for the World Health Organization). 416 p. Codd, G. A. 1995. Cyanobacterial toxins: occurrence, properties and biological significance. Water Science and Technology 32:149-156.

210 Effler, S. W., ed. 1996. Limnological and Engineering Analysis of a Polluted Urban Lake: Prelude to Environmental Management of Onondaga Lake, New York. New York: Springer-Verlag. 854 p. Falconer, I. R. 1999. An overview of problems caused by toxic blue-green algae (cyanobacteria) in drinking and recreational water. Environmental Toxicology 14:5-12. Gobler, C. J., Davis, T. W., Coyne, K. J., and Boyer, G. L. 2005. The contrasting impacts of nutrient loading and zooplankton grazing on the growth and toxicity of cyanobacteria blooms in a eutrophic New York lake. In The Third Symposium on Harmful Algae in the U.S., Monterey, California. October 2-7, 1 p. ———. 2006. Interactive influences of nutrient loading, zooplankton grazing and microcystin synthetase gene expression on cyanobacterial bloom dynamics in a eutrophic New York lake. Harmful Algae 6:119. Greeson, P. E. 1971. Limnology of Oneida Lake with Emphasis on Factors Contributing to Algal Blooms, U.S. Geological Survey. Albany, NY: New York State Department of Environmental Conservation, 187 p. Hotto, A., Satchwell, M. F., and Boyer, G. L. 2005. Seasonal production and molecular characterization of microcystins in Oneida Lake, New York, USA. Environmental Toxicology 20 (3):243-248. James, K. J., Furey, A., Sherlock, I. R., Stack, M. A., Twohig, M., Caudwell, F. B., and Skulberg, O. M. 1998. Sensitive determination of anatoxin-a, homoanatoxin-a and their degradation products by liquid chromatography with fluorimetric detection. Journal of Chromatography, A 798 (1 + 2):147-157. Jochimsen, E. M., Carmichael, W. W., An, J., Cardo, D. M., Cookson, S. T., Holmes, C. E. M., Antunes, M. B., de Melo Filho, D. A., Lyra, T. M., Barreto, V. S. T., Azevedo, S. M. F. O., and Jarvis, W. R. 1998. Liver failure and death after exposure to microcystins at a hemodialysis center in Brazil. New England Journal of Medicine 338:873-878. Mahmood, N. A., and Carmichael, W. W. 1986. Paralytic shellfish poisons produced by the freshwater cyanobacterium Aphanizomenon flos-aquae NH-5. Toxicon 24:175-186.

211 Mills, E. L., and Holeck, K. T. 2001. Oneida Lake: undergoing ecological change. Clearwaters 31 (4):22-25. Munawar, M., and Munawar, I. F., eds. 1996. Phytoplankton dynamics in the North American Great Lakes. Vol. 1: Lakes Ontario, Erie and St. Clair. Amsterdam: SPB Academic Publishing. 282 p. OCDWEP. 2004. Onondaga Lake ambient monitoring program - 2003 annual report. Onondaga County Department of Water Environment Protection, 164 p. Rapala, J., Sivonen, K., Luukkainen, R., and Niemela, S. I. 1993. Anatoxin-a concentration in Anabaena and Aphanizomenon under different environmental conditions and comparison of growth by toxic and non-toxic Anabaena-strains - a laboratory study. Journal of Applied Phycology 5 (6):581-591. Rinta-Kanto, J. M ., Boyer, G. L., Satchwell, M. F., Smith, M. T., Li, R., and Wilhelm, S. W. 2006. Analysis of toxic Microcystis blooms on Lake Erie using quatitative real-time PCR. In 2006 ASLO Annual Meeting Victoria, BC, Canada. July 4-9, 1 p. Sivonen, K. 2000. Freshwater cyanobacterial neurotoxins: Ecobiology, chemistry, and detection. Food Science and Technology (New York, NY, United States) 103:567- 581. Ueno, Y., Nagata, S., Tsutsumi, T., Hasegawa, A., Watanabe, M. F., Park, H. D., Chen, G.-C., Chen, G., and Yu, S.-Z. 1996. Detection of microcystins, a blue-green algal hepatotoxin, in drinking water sampled in Haimen and Fusui, endemic areas of primary liver cancer in China, by highly sensitive immunoassay. Carcinogenesis 17:1317-1321. Viaggiu, E., Melchiorre, S., Volpi, F., Di Corcia, A., Mancini, R., Garibaldi, L., Crichigno, G., and Bruno, M. 2004. Anatoxin-a toxin in the cyanobacterium Planktothrix rubescens from a fishing pond in northern Italy. Environmental Toxicology 19 (3):191-197. Watzin, M. C., Miller, E. B., Shambaugh, A. D., and Kreider, M. A. 2005. Application of the WHO alert level framework to cyanobacterial monitoring of Lake Champlain, Vermont. Environmental Toxicology 21:278-288.

212 Whitford, L., and Schumacher, G. 1969. A manual of the fresh-water algae in North Carolina. Raleigh, NC: The North Carolina Agricultural Experiment Station, 313 p. Yang, X., Satchwell, M. F., and Boyer, G. L. 2001. The identification of anatoxin-a from a toxic blue-green algae bloom in Lake Champlain, USA. Abstract. In Fifth International Conference on Toxic Cyanobacteria, Moosa Lakes, Queensland, Australia. July 15-20. 1 p.

213 Chapter 5. Discussion

The hypothesis of this work was “The occurrence of anatoxin-a in New York

State has been underestimated and that this neurotoxin may consequently pose a serious threat in New York lakes to animals and humans”. For anatoxin-a to be a health risk, anatoxin-a-producing blooms must occur in proximity to human habitats with a frequency and duration so that human exposure to toxic blooms would likely occur.

Anatoxin-a must also be produced in high enough quantities to pose a health risk.

The occurrence of cyanobacterial blooms in New York State was increasingly observed during the summer and fall seasons. Cyanobacterial toxins, including anatoxin- a, were widely detected. Anatoxin-a was repeatedly detected in the lower Great Lakes

(Lake Erie and Lake Ontario), Lake Champlain and smaller inland lakes such as

Onondaga Lake and Lake Neatahwanta. Anatoxin-a was primarily found in shallower nearshore sites in these natural water bodies, or in shallower areas such as the western basin of Lake Erie, embayments along the southern shoreline of Lake Ontario, or in nearshore regions or embayments in Lake Champlain. These nearshore areas are often heavily used for recreational activities such as swimming, boating and fishing, thus exposure of animal and human to toxic cyanobacteria and anatoxin-a would be possible.

However, would this neurotoxin be produced in quantities large enough to cause health concerns?

Anatoxin-a concentration measured in this study were related directly to a defined volume of water in accordance with EPA’s health guideline values which express cyanobacterial toxins per liter of water. Over the past several years, anatoxin-a has been measured at concentrations greater than 1 μg L-1 and even 10 μg L-1 in lake water

214 samples. However, anatoxin-a data from other countries are mainly given in terms of amount toxin per dry weight of cells. Anatoxin-a is largely retained within the cell when the conditions for the cell growth are favorable, and the maximum concentration is usually found during the logarithmic phase (Chorus and Bartram 1999). The wind-borne accumulation of toxic cyanobacterial cells during a bloom event into nearshore or shallow waters could further increase the total toxin quantity in these areas over those values in the general water column. Expression of anatoxin-a concentration in per dry weight terms would be better for the purpose of assessing its potential health risk. In this study, algal biomass was evaluated by measuring chlorophyll a concentration.

Chlorophyll a is a common indicator of phytoplankton biomass and constitutes about 1 to

2% of the dry weight for planktonic algae (Huang and Cong 2007). This suggests that

Anatoxin-a concentrations in new York State waters range from 1-20,000 μg anatoxin-a g-1 dry weight. These higher concentrations were primarily found in the Lower Great

Lakes (up to 900 μg ATX g-1 dry weight), Lake Champlain (up to 6,000 μg ATX g-1 dry weight) and Onondaga Lake (as high as 20,000 μg ATX g-1 dry weight). These estimated concentrations were even higher than those reported to be associated with animal fatalities in European countries (Sivonen and Jones 1999), suggesting that the risk of intoxication during a severe bloom event may be real.

In addition to its occurrence in the bloom, anatoxin-a must also be consumed at a level exceeding the health threshold for acute poisonings to occur. Currently there is no well established oral toxicity data for anatoxin-a, despite several efforts. Fitzgeorge et al.

(1994) and Stevens and Krieger (1991) estimated an acute oral LD50 of 5 to 16 mg anatoxin-a kg-1 body weight in mice based on the observation that mortalities occurred at

215 higher dosage. Although no relationship was observed between the dosage administered and the rate of mortality, these preliminary toxicity information can be used to approximately assess the potential toxicity of an anatoxin-a producing bloom in natural waters. Based on the reported concentrations of anatoxin-a in New York State water (this thesis), and a rough estimation that dry weight biomass is 10% of the wet weight biomass, a typical 20-kg dog would need to consume between 50-1000 grams of toxic cyanobacterial cells to reach a fatal dose. Considering the number of occurrences of animal fatalities in the United States and other countries, and the measured concentration of anatoxin-a, it seems likely that anatoxin-a may be more toxic orally to animals in natural waters than predicted, or that the concentration of anatoxin-a in cyanobacterial cells may be higher than measured.

Selwood et al. (2006) reported that anatoxin-a might be present in algal cells as its immediate biosynthetic precursor, 11-carboxyl anatoxin-a. Commonly used extraction protocols may not be effective in releasing 11-carboxy anatoxin-a from biological matrices and it behaves differently in current analytical procedures. This may lead to its concentration being overlooked during routine monitoring, which may lead to an underestimation of the actual anatoxin-a pool present in natural material. Consumption of toxic cyanobacterial cells containing a large concentration of this precursor, followed by its subsequent decarboxylation while in the gut may produce additional quantities of anatoxin-a and resulted in increased opportunity of mortality of the affected animals.

Further study on presence of 11-carboxyl anatoxin-a in natural blooms, its transformation to anatoxin-a, and its behavior under typical extraction protocols would be beneficial for assessing health risk of anatoxin-a.

216 Laboratory studies presented in Chapter 4 indicated that anatoxin-a can undergo rapid degradation under simulated and natural aquatic environments. This would further complicate monitoring of anatoxin-a, as the levels measured in older blooms may significantly underestimate the concentrations at peak exposure. The loss of anatoxin-a

(and 11-carboxyl anatoxin-a) during transport, storage and processing may also result in underestimate of the anatoxin-a concentration and subsequently health risk. In this study, a cautionary level of 0.1 μg anatoxin-a L-1 lake water was proposed. Samples were largely collected from open waters in lakes and few or no severe anatoxin-a-producing scums were sampled. The occurrence of such scums along the downwind shorelines would concentrate the samples from the open water and, in turn, may concentrate a bloom containing minor levels of anatoxin-a to a hazardous concentration. Using the average algal biomass during the monitoring, we feel the measured value of anatoxin-a of

0.1 μg L-1 lake water would provide a sufficient safety margin to protect against lethal acute toxicity. Current efforts using remote sensing and hydrodynamic modeling are underway to provide timely information on the occurrence of cyanobacterial blooms in

New York State lakes. These efforts may help predict when a bloom might develop and be concentrated by wind and wave action. In addition, studies on biosynthesis and regulation of anatoxin-a in cyanobacterial cells are being conducted to identify the genes responsible for production of anatoxin-a and identify potentially toxic blooms. By combining these techniques, occurrence of toxin-producing blooms could be predicted and the health risks associated with anatoxin-a effectively minimized.

217 REFERENCES

Chorus, I., and Bartram, J., eds. 1999. Toxic Cyanobacteria in Water: A Guide to Their

Public Health Consequences, Monitoring, and Management. London: E & FN

Spon (for the World Health Organization). 416 p.

Fitzgeorge, R. B., Clark, S. A., and Keevil, C. W. 1994. Routes of intoxification. In

Detection Methods for Cyanobacterial Toxins, edited by G. A. Codd, T. M.

Jefferies, C. W. Keevil and E. Potter. Cambridge, UK: The Royal Sociesty of

Chemistry. p69-74.

Huang, T.-L., and Cong, H.-B. 2007. A new method for determination of chlorophylls in

freshwater algae. Environmental Monitoring and Assessment 129:1-7.

Selwood, A. I., Holland, P. T., Wood, S. A., Smith, K. F., and McNabb, P. S. 2006.

Production of anatoxin-a and a novel biosynthetic precursor by the

cyanobacterium Aphanizomenon issatschenkoi. Environmental Science &

Technology: In print.

Sivonen, K., and Jones, G. J. 1999. Cyanobacterial toxins. In Toxic cyanobacteria in

water, edited by I. Chorus and J. Bartram. London and New York: E & FN Spon.

p41-111.

Stevens, D. K., and Krieger, R. I. 1991. Effect of route of exposure and repeated doses on

the acute toxicity in mice of the cyanobacterial nicotinic alkaloid anatoxin-a.

Toxicon 29 (1):134-138.

218 Appendix 1. Discrete sampling stations on Lake Erie and their occurrence of anatoxin-a between 2002 and 2005. “-“ indicates station was sampled at least once during the year, “trace” and “+”indicate anatoxin-a was detected at concentration less and greater than 0.1 μg L-1 respectively. Blank cells indicate the station was not sampled.

Station Latitude Longitude 2002 2003 2004 2005 23 42.5067 -79.8911 - - - 84 41.9339 -81.6581 - - - 205 42.3325 -80.3673 - 311 41.6665 -82.5000 trace - 338 41.7017 -82.6264 - - 339 41.7303 -82.5142 - 340 41.7592 -82.3919 - - 341 41.7836 -82.2822 - 343 41.8481 -83.0828 - 344 41.7819 -82.8426 trace 357 41.8122 -82.9828 trace - - - 358 41.8927 -82.8567 - - 442 42.8424 -79.3920 - 478 41.6570 -82.8240 - - 496 41.5672 -82.7212 trace trace 558 41.6993 -83.4598 - trace - 580 41.8483 -83.1065 - 589 42.1495 -80.0837 - 590 42.3856 -79.5428 - 835 41.7555 -83.3400 - 881 41.9692 -83.2085 - - 882 41.7339 -83.3839 trace - - - 885 41.5186 -82.6422 - trace trace

219 886 42.5358 -79.6186 - 889 42.5892 -81.4417 - 896 42.4278 -81.2017 - 916 42.2811 -81.6719 - 931 42.8514 -78.9428 - 932 42.7881 -79.2072 - 933 42.8250 -79.5667 - - 934 42.7083 -79.5083 - 935 42.5906 -79.4681 - 936 42.4719 -79.4106 - 937 42.7169 -80.2486 - 938 42.6333 -80.0572 - 939 42.5667 -79.9167 - - 940 42.4 386 -79.8356 - 941 42.3244 -79.8356 - 942 42.2586 -79.8317 - 943 42.3586 -80.9167 - 944 42.5750 -80.6417 - 945 42.5333 -80.6417 - 946 42.1672 -80.6433 - 949 42.1572 -81.2594 - - 950 42.4736 -81.4403 - 951 42.3583 -81.4442 - 952 42.2081 -81.4419 - 953 42.02 44 -81.4419 - 954 41.7 997 -81.4442 - - 955 41.6922 -81.4419 - - 956 41.5250 -81.7086 - - 957 41.9342 -81.6519 - 958 41.5245 -81.7078 - - 961 41.9071 -82.1840 - 965 41.5025 -82.4997 - 966 41.9822 -82.6402 - 967 41.8914 -82.6664 - - - 968 41.7158 -82.7253 - - - trace 969 41.6078 -82.9235 - - trace 971 41.9494 -83.0503 - - - - 972 41.8592 -83.2011 - - - - 973 41.7911 -83.3311 trace - - 974 41.7236 -83.1511 trace - - - 1138 42.1657 -80.1614 - 1156 42.0468 -83.1348 - 1163 41.4778 -82.7064 + + trace trace 1165 41.9996 -82.7300 - 1191 41.6385 -83.5328 - - 580D 41.8480 -83.1067 -

220 Ashtabula 41.9278 -80.9149 - Barcelona 42.3360 -79.7026 - Buffalo 42.8519 -78.9425 - EPA 43 41.7744 -82.0042 - ER15M 42.5168 -79.8933 - ER37M 42.1102 -81.5753 - ER43 41.7883 -81.9603 - ER58 41.6842 -82.9325 - ER59 41.7265 -83.1490 - ER60 41.8910 -83.1977 - ER61 41.9467 -83.0450 - - ER63 42.4165 -79.7999 - ER78M 42.1167 -81.2500 - ER91 41.8408 -82.9167 - ER91M 41.8403 -82.9165 - ER92 41.9500 -82.6863 trace ER93B 42.6165 -79.9999 - GR Dock 42.8644 -79.5742 - GR mouth 42.8519 -79.5800 - LE1 42.0972 -83.1133 - LE2 41.6072 -82.7569 trace LE3 41.5242 -82.6719 trace LE4 41.7081 -83.0333 trace LE5 41.7339 -83.2186 trace LE6 41.8061 -82.9822 trace LE7 41.7461 -82.8375 - LE8 41.8578 -82.0000 - LE9 41.8942 -81.8333 - LE13 42.2503 -81.1056 - LE14 42.4019 -80.6414 - LE15 42.4386 -80.3833 - LE16 42.1956 -80.0167 - LE17 42.3306 -79.9886 - LE18 42.6889 -80.1667 - LE19 42.6900 -80.1308 - LE20 42.7292 -79.9500 - LE21 42.7742 -79.7500 - LE22 42.5375 -79.3333 - LE23 42.5539 -79.2500 - LE24 42.5872 -79.1681 - LE25 42.6833 -79.0931 - LE26 42.7306 -79.0000 - LE27 42.7786 -78.9000 - LE28 42.8250 -78.9167 - LE29 42.8361 -79.0167 - LE30 42.8228 -79.1000 - LE31 42.7908 -79.3272 - Long Point Bay 42.5869 -80.3871 - - LPB mouth 42.6185 -80.3574 - LPB2 42.5915 -80.4367 -

221 MB 1 41.7191 -83.4167 - MB 2 41.7387 -83.3750 - MB 3 41.7588 -83.3333 - MB 4 41.7783 -83.2917 - MB 5 41.7978 -83.2500 - MB 6 41.8179 -83.2083 - MB 7 41.8356 -83.1667 - MR 1 41.6389 -83.5328 + MR 2 - - + MR 3 41.6547 -83.5239 + MR 4 41.6667 -83.5011 + MR 5 41.6789 -83.4856 + MR 6 41.7047 -83.4464 + MR 7 41.7450 -83.3620 trace Port Colborne 42.8800 -79.2492 - Pt. Alma 42.0823 -82.1080 - Pt. Dover 42.7134 -80.1073 - Pt. Stanley 42.6321 -81.3191 - Rhondo 42.2806 -81.6731 - Sandusky Basin 41.5330 -82.4505 - Transect from 23 to 1163 41.4693 -82.7154 - US65-E 42.0015 -80.2663 - WB1 41.5830 -82.8782 - WB2 41.6009 -82.9184 - WB2.5 41.6517 -82.9500 - WB3 41.7005 -83.0036 - WB4 41.7427 -83.0849 - WB5 41.7068 -82.9085 -

222 Appendix 2. Discrete sampling stations on Lake Ontario and their occu rrence of anatoxin-a between 2002 and 20 05. “-“ ind icates station was sampled at least once during the year, “trace”, “+” and “++” indic ate ana toxin-a was detected at concentration less and greater than 0.1 μg L-1, and greater than 1 μg L-1 respectively. Blank cells indicate the station was not samp led.

Stations Latitude L ongitude 2001 20 02 2003 2004 2005 1 43.3132 -79.7517 - - - - 2 43.3410 -79.6659 - - 3 43.2680 -79.6209 - - - - 5 43.4256 -79.6572 - - - - 6 43.4660 -79.5306 - - 7 43.5466 -79.4912 - - 8 43.6243 -79.4524 - - - - 9 43.5875 -79.3953 - - - - 11 43.5852 -79.3128 - 12 43.5055 -79.3543 - - - - 13 43.4166 -79.3998 - - 14 43.3949 -79.4891 - - 15 43.3176 -79.4439 - - - - 16 43.2726 -79.3611 - - 18 43.3064 -79.2773 - 19 43.3835 -79.2846 - - 20 43.3385 -79.1920 - - 21 43.3030 -79.1194 - - 22 43.2967 -79.0061 - - - - 23 43.3717 -79.0664 - - 24 43.4414 -79.1274 - 25 43.5170 -79.0792 - 26 43.6066 -79.0147 - -

223 27 43.7036 -78.9574 - - 28 43.7754 -78.8548 - - - 29 43.8300 -78.8697 - - - - 30 43.8304 -78.6636 - - 31 43.8860 -78.4629 - 32 43.7829 -78.4395 - 33 43.5982 -78.8003 - - - 34 43.4628 -78.7594 - - - - 35 43.3611 -78.7289 - - - 36 43.4917 -78.3827 - - 37 43.3920 -78.3681 - 38 43.3819 -77.9908 - 39 43.4870 -78.0007 - - 40 43.5895 -78.0121 - - 41 43.7166 -78.0264 - - 42 43.8390 -78.0381 - 43 43.9489 -78.0478 - 45 43.8209 -77.7806 - 46 43.8853 -77.6898 - 48 43.8605 -77.5240 - 49 43.7728 -77.4380 - 52 43.4322 -77.7082 - - 53 43.3487 -77.7143 - 55 43.4439 -77.4383 - - - 57 43.2741 -77.5923 - - 58 43.3271 -77.4383 - 61 43.7858 -77.1578 - - - - 62 43.8799 -76.9990 - 63 43.7308 -77.0169 - - 64 43.5206 -76.9086 - - - 65 43.4225 -76.8811 - - 66 43.3331 -76.8403 - - 67 43.4074 -76.7952 - 68 43.5320 -76.7345 - - 70 43.5416 -76.6178 - - 71 43.4769 -76.5272 - - 72 43.5500 -76.5229 - - 73 43.6327 -76.2871 - 74 43.7510 -76.5161 - - 75 43.8419 -76.3547 - - - 77 43.9578 -76.4114 - - - 78 44.0830 -76.4078 - - - 79 44.0753 -76.5214 - - 80 44.1417 -76.6103 - - - 81 44.0156 -76.6714 - - - 82 44.0664 -76.8117 - - - 83 44.0000 -76.8431 - - - 84 43.8531 -76.7319 - - - 86 43.2550 -79.1942 - - - - 87 43.2970 -77.5187 -

224 88 43.5881 -76.4172 - - 89 43.6983 -76.4161 - - - 90 44.1364 -76.8267 - - - 93 43.3256 -78.8664 - 94 43.3234 -77.2161 - - 95 43.3081 -76.9986 - - 96 43.2238 -79.4476 - 98 43.9342 -76.2317 - - - 100 44.1386 -76.3292 - - 101 44.1911 -76.3083 - - 102 44.2036 -76.2369 - - 103 44.2039 -76.5431 - - - 403 43.5930 -78.2353 - 498 44.1519 -77.3367 trace 737 43.6086 -79.4303 - - - 738 43.5639 -79.3869 trace - 739 43.4236 -79.2578 - - 740 43.3403 -79.1564 - - 741 43.2567 -79.0597 - - - 742 43.3847 -78.1922 - - - 743 43.5219 -78.1903 - - 744 43.6658 -78.1811 - - 745 43.8067 -78.1728 - - 746 43.9475 -78.1672 - - 747 43.6333 -77.2919 - 750 43.5539 -79.5353 - - 751 43.5417 -79.5034 - 752 43.5008 -79.4819 - trac e - 753 44.2414 -76.2992 - - 754 44.2371 -76.4066 - - 756 43.2336 -79.4078 - - - 757 43.2158 -79.3342 - - - 826 44.1519 -77.2564 trace 1001 43.2878 -79.8417 - - - 1183 43.4650 -76.5133 - - 1184 43.8575 -76.2064 - - 1185 44.2681 -76.1817 - 1193 44.0606 -77.0814 - - 1194 44.1072 -77.0303 - - 1195 44.0969 -77.0731 - - 1196 44.1761 -77.0467 - - 1235 44.2031 -76.5130 - 1237 44.2093 -76.4838 - 1238 44.2157 -76.4572 - 1240 44.2263 -76.4295 - 1243 44.2415 -76.3497 - 1245 44.2778 -76.2744 - 1247 44.2484 -76.2499 - 1249 44.2513 -76.2250 - 1251 44.2509 -76.1997 -

225 12 mile 43.3172 -78.8354 trace - - 18 mile 43.3370 -78.7172 - - - 33c 43.5753 -78.5310 - 4 mile 43.2770 -78.9978 - - 64b 43.5231 -76.9248 - 81a 43.9818 -76.6552 - 98a 43.9470 -76.3224 - B.Forman Park 43.2852 -77.1951 trace BC 5c 44.2675 -76.3340 - BQ1 44.0594 -76.9658 - BQ2 44.0594 -77.0864 - BQ3 44.1069 -77.0306 - BQ4 44.1006 -77.0739 - BQ5 44.1756 -77.0461 - BR1 43.3540 -77.8913 - trac e - BR100 43.4250 -77.8939 trace - - BR30 43.3770 -77.8939 - - - Braddock Bay 43.3068 -77.7060 + + trace - Cape Vincent 44.1368 -76.3303 - Cedar Point 44.2031 -76.1956 - Chamount Bay 44.0590 -76.2242 - - - Eastman Park 43.2357 -77.5526 - EPA64 43.5186 -79.0848 - Fair Haven 43.3464 -76.6883 trace - Fort Niagra 43.2603 -79.0511 - Genesee River 43.2563 -77.6062 - trac e - Golden Hill 43.3683 -78.4697 - Grass Point 44.2847 -76.0036 trace Hamlin 43.3614 -77.9586 - Henderson Harbor 43.8577 -76.2059 - - - Irondequoit Bay 43.1781 -77.5349 - trace - - Kring Point 44.3763 -75.8554 - Lake View 42.7111 -78.9364 - - Lakeshore Beach 43.3669 -78.2375 - Little Sodus Bay 43.3342 -76.7089 trace - Long Point 44.0589 -76.1517 - Long Pond 43.2870 -77.7055 + ++ - Niagara River 43.2611 -79.0591 - - - Nine Mile 43.5423 -76.4131 - Nine Mile Point 43.5241 -76.3697 - Oak Orchard 43.3712 -78.1916 - - - - Olcott 43.3377 -78.7150 - ON68B 43.5839 -79.4165 - Oswego 43.4702 -76.5008 trace - - Pine Grove 43.5491 -76.2176 trace Port Bay 43.2790 -76.8282 - - trac e - Port Dalh. 43.2252 -79.2718 - - Salmon River 43.5756 -76.2558 + - - Sandy Creek 43.3496 -77.8947 trace - - Sandy Pond 43.6499 -76.1802 trace - -

226 Selkirk SP 43.5572 -76.2047 - - - Sodus Bay 43.2232 -76.9300 - - - - Southern Chamount Bay 44.0337 -76.2661 - Sunset Bay 43.5242 -76.3843 - - Tibbett's Point 44.6972 -75.4994 - Tuscarora 43.3109 -78.8519 - Wescott Beach 43.8969 -76.1139 - - WH7 44.2250 -76.4736 - Wrights Landing 43.4641 -76.5159 -

227 Appendix 3. Discrete sampl ing stations on Lake Champlain and their occurrence of anatoxin-a between 2000 and 2005. “-“ in dicates station was sampled at least once during the year, “trace”, “+”, and “++” ind icate anatoxin- a was detected at concentration less than 0.1 μg L-1, and grea ter than 0.1 a nd 1 μg L-1 respectively. Blank cells indicate the station was not sampled.

Lake Champlain Stations Latitude Long itude 2000 2001 2002 2003 2004 2005 Alburg westside off Rt. 2 44 .9660 -73.3 050 - - Alburg-Missisquoi Bay 44 .9779 -73.2 190 - Apple Tree Bay 44 .4972 -73.2 533 - AuSable River 44 .5537 -73.4 082 ------AuSable Shoal 44 .5884 -73.4 266 - trace - - - - Boquet River 44 .3557 -73.3 451 - - - - - Brady's Camp 44 .2267 -73.3 615 ++ Brochets River 45.0645 -73.1 039 - - - Burlington Bay 44.4862 -73.2 480 - - trace - - - Burlington Harbor 44 .4704 -73.2 286 ++ - - - - - Button Bay 44.1776 -73.3 577 - - - Carry Bay 44.8389 -73.2 806 - Cedar Island 44.6592 -73.2 544 trace - - Cedar Island Causeway 44 .6473 -73.2 577 - - - Champlain Water Intake 44.4417 -73.2 301 ------Chapman Bay 45.0012 -73.2 062 ++ - Cole Bay 44.1380 -73.4 209 - - - Corlear Bay/ Port Douglas 44 .4867 -73.4 051 - - - - - Cumberland Bay 44 .7084 -73.4 237 - - - - - DEC Site 19 44.3377 -73.2 992 - DEC Site 33 44.7017 -73.4 183 - - - - - DEC Site 36 44.7549 -73.3 550 - - - - - DEC Site 46 44 .9485 -73.3 490 - - - - - DEC Site 50 45.0122 -73.1 744 - - - - Essex Ferry/McNeil Cove 44.4741 -73.2 216 - Goose Bay 44.9865 -73.1 198 - - - - Great Chazy river launch 44.9322 -73.3 848 - Highgate Springs 44.9797 -73.1 081 - - - Horseshoe Shoal 44.8573 -73.3 007 - - - Isle la Motte 44.8628 -73.3 417 - Jones Point 44.3986 -73.3 669 - - - - - Juniper Island 44 .4589 -73.2 760 - Keeler Bay 44.6525 -73.2 976 - - - King Bay 44.9411 -73.3 797 - - + - - Knight Island 44 .7971 -73.2 675 - - - - Knight Point State Park 44 .7703 -73.2 936 - - LaMotte Passage 44 .8877 -73.3 107 trace - - - - Lapans Bay 44.8158 -73.1 850 trace - - Lime Kiln Pt. 44.9858 -73.0 983 - Long Point 44.2556 -73.2 822 ------Mallets Bay Inner 44.5667 -73.2 085 - - -

228 Mallets Bay Outer 44.5916 -73.2 747 - - - Maquam Bay 44 .9191 -73.1 876 ++ - - - - Missisquoi Bay Center 45.0392 -73.1417 - - - - Monty Bay 44.8378 -73.3 949 trace - - - - North Beach 44.4931 -73.2 408 - - - - North Hero marina 44 .8175 -73.2 9897 - North Hero State Park 44 .9153 -73.2 357 - - Northwest Bay 44.1843 -73.4 176 - - - Pelots Bay 44.8286 -73.3 003 - Point Au Roche launch 44 .8193 -73.3 733 - Port Henry 44.0409 -73.4452 - - - - Province Point 45.0093 -73.1884 - Quaker Smith Point 44.3755 -73.2781 - - - - - Ransoms Bay 44.9333 -73.2330 - - Red Rocks Beach 44.4420 -73.2243 - - Route 78 Bridge 44.9715 -73.2129 - - - - Sandbar State Forest Park 44.6267 -73.2406 - Savage Island 44.6989 -73.2475 trace - - - Shelburne Bay Inner 44.4314 -73.2405 ------Shelburne Bay Outer 44.4295 -73.2355 - - - - Snake Den Harbor 44.2261 -73.3608 - - - St Albans Bay Inner 44.7979 -73.1470 + trace - - St Albans Bay Outer 44.7633 -73.1806 + trace - - Swanton station 44.9200 -73.1620 - The Gut 44.7600 -73.3025 - Thompson's Pt. 44.2722 -73.2917 - Ticonderoga Ferry 43.8125 -73.4374 - - Treadwell Bay 44.7542 -73.3912 trace - - - - Valcour/Crab Island 44.6473 -73.4219 trace - - - - Venise Bay 45.0758 -73.1443 - - - Westport Bay 44.1838 -73.4173 - - Whallon bay 44.2710 -73.3339 ------Willsboro Bay Inner 44.3987 -73.3976 ------Willsboro Bay Outer 44.4544 -73.3914 ------

229 Appendix 4. New York State lakes monitored for anatoxin-a between 2000 and 2005.

“-“ indicates station was sampled at least once during the year, “trace”, “+”, and “++” indicate anatoxin-a was detected at concentration less than 0.1 μg L-1, and greater than

0.1 and 1 μg L-1 respectively. Blank cells indicate the station was not sampled.

Lakes Longitude Latitude 2000 20012 002 2003 2004 2005 Alcove River 42.437 73.937 - Allegheny River 42.097 78.848 - Almond Lake 42.349 77.711 - Auburn pond 42.932 76.542 - Basic Creek 42.354 74.955 - Bear Lake 42.349 79.380 - Beaver Lake 43.181 76.402 - Blue Mont. Lake 43.854 74.438 - Brant Lake 43.721 73.676 - Canadaigua Lake 42.796 77.272 - - - Canadice Lake 42.735 77.551 - - Cassadaga Lake 42.350 79.161 - Cayuga Lake 42.457 76.505 - - - Cayuta Lake 42.361 76.744 + Cazenovia Lake 42.904 76.313 - Chadwick Lake 41.503 74.011 - Chautauqua Lake 42.173 79.416 - Chenango Lake 42.583 75.441 - Conesus Lake 42.776 77.712 - - Cranberry Lake 41.711 74.837 - Crellin Pond 42.364 73.595 - Cross Lake 43.133 76.475 - Cuba Lake 42.237 78.307 - Deforest Lake 41.145 73.959 - DeReuyter Reservoir 42.801 76.313 - Dryden Lake 42.464 76.280 - Fourth Lake 43.755 74.794 - Genesee River 43.258 77.602 - - Gifford Reservoior 43.151 75.874 - Trace Great Sacandaga Lake 43.104 74.174 - Heiberg Forest Pond 42.904 76.313 Hemlock Lak e 42.763 77.611 - - Hincley Reservoir 42.319 75.107 - Honeoye Lake 42.726 77.507 - - Hornell Reservoir 42.349 77.711 - Hudson River 40.703 74.026 - Indian Lake 43.652 74.388 - Jamesville Reservoir 42.978 76.066 - Keuka Lake 42.582 77.136 - - Kiamesha Lake 41.677 74.667 -

230 Laborador Pond 42.903 76.313 - Lake Agawam 40.882 72.392 + ++ Lake Durant 43.828 74.384 - Lake George 43.313 74.055 - Lake Louise Marie 41.616 74.587 - Lake Neatahwanta 43.313 76.431 - - + Trace Lamoka Lake 42.413 77.072 - Monhagen Lake 41.446 74.423 - Niagara River 43.262 79.071 - - Northeast River 40.841 73.214 - Oneida Lake 43.034 76.137 - - Trace Trace Trace - Onondaga Lake 43.094 76.193 - - Trace ++ Oswego River 43.465 76.514 - - Otisco Lake 42.903 76.313 - - - Otter Lake 43.140 76.528 - Owasco Lake 42.903 76.540 - - - Owasco River 42.812 76.437 - Piseco Lake 43.446 74.529 - Raquette Lake 43.816 74.598 - Rushford Lake 42.385 78.207 - Sacandaga Lake 43.652 74.388 - Salmon River 43.575 76.204 - - - Sandy Creek 43.352 77.892 - - Saratoga Lake 43.050 73.719 - Schroon Lake 43.727 73.809 - Seneca Lake 42.387 76.977 - - Seneca River 42.903 76.540 - Seventh Lake 43.755 74.794 - Silver Lake 42.706 78.022 - - - Skaneateles Lake 42.978 76.066 - - - Sleepy Hollow Lake 42.281 73.806 - Southampton Lake 40.937 72.437 ++ St. Lawrence River 44.890 75.113 - - Upper Little York Lake 42.714 76.149 - Waneta Lake 42.438 77.098 + Whitney Point 42.351 75.970 - Reservoir Woodland Reservoir 43.027 76.177 -

231 Vita

Name: Xingye Yang

Date and Place of Birth: November 27, 1975 in Qingyang, Gansu, China

Education:

Name and L ocation Dates Degree State University of New York, College of 2000-2007 Ph.D. Environmental Science and Forestry, Biochemistry Syracuse, New York

Nanjing University, 1996-1999 M.S. Nanjing, China Environ. Biol.

Nanjing University, 1992-1996 B.S. Nanjing, China Environ. Sci.

Employment:

Employer Dates Degree SUNY-ESF 2000-2007 Research/Teaching Syracuse, New York Assistant

Nanjing University, 1996-1999 Research/Teaching Nanjing, China Assistant

232