Association of Toxin-Producing Clostridium botulinum with theMacroalga Cladophora in the Great LakesChan Lan Chun,† Urs Ochsner,† Muruleedhara N. Byappanahalli,‡ Richard L. Whitman,‡

William H. Tepp,§ Guangyun Lin,§ Eric A. Johnson,§ Julie Peller,⊥ and Michael J. Sadowsky†,∥,*†BioTechnology Institute, University of Minnesota, St. Paul, Minnesota 55108, United States‡Lake Michigan Ecological Research Station, U.S. Geological Survey, Porter, Indiana 46304, United States§Department of Bacteriology, University of Wisconsin, Madison, Wisconsin 53706, United States∥Department of Soil, Water, and Climate, University of Minnesota, St. Paul, Minnesota 55108, United States⊥Department of Chemistry, Physics, Astronomy, Indiana University-Northwest, Gary, Indiana 46408, United States

*S Supporting Information

ABSTRACT: Avian botulism, a paralytic disease of birds, oftenoccurs on a yearly cycle and is increasingly becoming more commonin the Great Lakes. Outbreaks are caused by bird ingestion ofneurotoxins produced by Clostridium botulinum, a spore-forming,gram-positive, anaerobe. The nuisance, macrophytic, green algaCladophora (Chlorophyta; mostly Cladophora glomerata L.) is apotential habitat for the growth of C. botulinum. A high incidence ofbotulism in shoreline birds at Sleeping Bear Dunes NationalLakeshore (SLBE) in Lake Michigan coincides with increasinglymassive accumulations of Cladophora in nearshore waters. In thisstudy, free-floating algal mats were collected from SLBE and othershorelines of the Great Lakes between June and October 2011. Theabundance of C. botulinum in algal mats was quantified and the typeof botulism neurotoxin (bont) genes associated with this organism were determined by using most-probable-number PCR(MPN-PCR) and five distinct bont gene-specific primers (A, B, C, E, and F). The MPN-PCR results showed that 16 of 22 (73%)algal mats from the SLBE and 23 of 31(74%) algal mats from other shorelines of the Great Lakes contained the bont type E(bont/E) gene. C. botulinum was present up to 15 000 MPN per gram dried algae based on gene copies of bont/E. In addition,genes for bont/A and bont/B, which are commonly associated with human diseases, were detected in a few algal samples.Moreover, C. botulinum was present as vegetative cells rather than as dormant spores in Cladophora mats. Mouse toxin assaysdone using supernatants from enrichment of Cladophora containing high densities of C. botulinum (>1000 MPN/g dried algae)showed that Cladophora-borne C. botulinum were toxin-producing species (BoNT/E). Our results indicate that Cladophoraprovides a habitat for C. botulinum, warranting additional studies to better understand the relationship between this bacteriumand the alga, and how this interaction potentially contributes to botulism outbreaks in birds.

1.0. INTRODUCTION

Outbreaks of botulism, a paralytic and often fatal bacterialdisease, have caused large mortalities of birds and fish in theGreat Lakes.1−4 Outbreaks were first reported in this region in19635 and have gone through episodic cycles over the lastseveral decades. The USGS National Wildlife Heath Centerestimates that there were in excess of 100 000 bird mortalitiesfrom botulism-related outbreaks between 1963 and 2007.2

These outbreaks have become increasingly more common inLakes Michigan, Erie, Huron, and Ontario, with recentincreases and expansion of the affected areas and of birdspecies killed. Since 1999, it has been estimated that botulismtoxicity may be responsible for the deaths of more than 87 000Great Lakes’ birds.2,3 In 2006, a large die-off of nearly 3000native, fish-eating birds occurred in the areas of Lake Michiganwithin and near the Sleeping Bear Dunes National Lakeshore

(SLBE). This was not an isolated event, and bird die-offscontinue to be a problem in and around SLBE.6

Clostridium botulinum, the responsible pathogen, is anobligate anaerobe that is widespread in aquatic and soilenvironments, mostly as dormant spores.7 The production ofbotulism neurotoxin (BoNT) only occurs when environmentalconditions become suitable and allow for the germination ofspores and the subsequent growth of vegetative cells. Sevendistinct serotypes of botulism neurotoxin have been identifiedand classified (designated BoNT/A to/G) based on theirantigenic properties.8 BoNT types C and E are primarily

Received: November 28, 2012Revised: February 19, 2013Accepted: February 20, 2013Published: February 20, 2013

Article

pubs.acs.org/est

© 2013 American Chemical Society 2587 dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−2594

responsible for the deaths of waterfowl and fish-eating birds inthe Great Lakes and BoNT type E outbreaks have recentlyincreased in Lake Michigan.3,9 Laboratory studies10,11 haveprovided some insights into the abiotic factors favoringbacterial growth and some studies have attempted to correlateenvironmental factors with botulism outbreaks.12,13 Despite thisknowledge, however, it is still unclear how ecophysiologicalfactors in the Great Lakes lead to in situ growth, toxinproduction, and subsequent botulism outbreaks leading to birdmortalities.A few overlapping and hypothetical pathways have been

proposed to account for bird deaths, although there may besubtle differences in the etiology among the affected avianpopulation. One possible route is that shorebirds acquire toxinfrom insects or larvae that have been feeding on the carcasses ofdead fish and birds, referred to as the carcass-maggot cycle.14 Ithas also been postulated that birds become ill due toconsumption of C. botulinum spores, vegetative cells, or toxincontaminated mussels, fish, and/or algae. Aside from birds,BoNT/E can also be toxic to freshwater fish15 and has beenfound in wild fish living in and around Lake Erie and LakeOntario.16 Several studies suggest that the increased botulismmortality seen in Great Lakes fish and birds may be due toconsumption of contaminated invasive mussels (Dreissena spp.)and round gobies (Neogobius melanostomus).17,18

Recently, the filamentous, nuisance green alga Cladophorahas also been considered as a possible vector of C.botulinum.6,19 Byappanahalli and Whitman6 demonstrated thatCladophora collected from the shoreline in Lake Michigan andincubated under laboratory conditions contained a higherfrequency of C. botulinum type E (bont/E gene) than did fresh,nonincubated, samples. Likewise, C. perf ringens, which isphysiologically similar to C. botulinum,20 has been found inhigh densities in Cladophora mats suggesting that thedecomposing algal mats may provide a generally suitableniche (high nutrients, organic matter, and low Eh) for theserelated anaerobic bacteria.21

The potential for the association between Cladophora matsand C. botulinum has been increasingly speculated. A highincidence of botulism in shoreline birds at SLBE in LakeMichigan appears to parallel the increase in Cladophoraaccumulations in nearshore areas. The increased accumulationof Cladophora in water is thought to be largely due to improved

water clarity and light penetration depth as a result of thepresence of invasive filter feeders such as dreissenid mussels.22

Algal exudates and thalli are rich in nutrients and can sustain avariety of organisms including epiphytes (cyanobacteria anddiatoms), grazers (protozoa, mollusks, rotifers, micro andmacroflora, and young crayfish),23−25 and numerous bac-teria.26,27 This complex ecosystem, which includes benthicsediments, bacteria, spores, algae, and macroinvertebrates, likelyplay a role in the growth of C. botulinum and the transport oftoxin up the food chain.28,29 Furthermore, as the algae begin todecay or accumulate in dense mats along shorelines, this leadsto anaerobic conditions promoting the growth of C. botulinum.Because this bacterium lacks the ability to synthesize essentialgrowth factors (such as some amino acids),7 decomposing algalmatter (plus other aquatic vegetation) and the macro-invertebrate community most likely provide the necessarynutrients for the growth of toxin-producing C. botulinum. Theresultant toxins, we believe, are then transferred to fish directly,or via several food chain intermediates, eventually to fish andthen to birds. Byappanahalli and Whitman6 suggested that alikely pathway for piscivorous species is sediment − Cladophora− invertebrates − fish − birds.In this study, we examined the potential of the botulism

pathogen to grow in Cladophora by determining the relativeabundance and types of C. botulinum in algal mats collectedfrom SLBE and other shorelines of the Great Lakes. Analyseswere done by using the most probable number-polymerasechain reaction (MPN-PCR) technique using bont gene-specificprimers. Moreover, we examined the nature of the pathogenthat are present in Cladophora mats (viable/vegetative cells,spores only, or both) and determined if Cladophora-borne C.botulinum has the potential to produce botulism neurotoxin byusing mouse toxin-antitoxin assays.

2.0. EXPERIMENTAL SECTION

2.1. Site Description and Sampling. Cladophora sampleswere collected monthly between June and October 2011 fromfour beaches on Lake Michigan and from a shoreline on LakeOntario; Cladophora accumulations are common at theselocations. Our principal study area was SLBE in Empire,Michigan. Other study sites included shorelines along DoorCounty, WI, the City of Racine, WI, Porter and Cook Counties

Figure 1. Location of sampling sites on the Great Lakes. Cladophora samples were taken from the nearshore water.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942588

along southern Lake Michigan in Indiana and Illinois (IndianaDunes National Lakeshore, IDNL), and Hamlin Beach StatePark, NY on Lake Ontario (Figure 1). GPS coordinates of each

sampling site are listed in Table 1. Three replicate samples ofapproximately 150 g of free-floating Cladophora algal matter(thalli) were obtained from two locations at each sampling site.

Table 1. Clostridium botulinum Density (Most Probable Number (MPN)/g Dried Weight) in Cladophora Collected fromShorelines of the Great Lakes in 2011

location site (GPS point) date MPN (bont Type)awater/airtemp. °C

algaeconditionsb

fish/birddeath

Sleeping Bear National Lakeshore, MI County Road 651 N(44.947N, 85.812W)

6/30/11 ND 62 ± 23 (E) 15.4/20.0 FT, FR fish (∼1000gobies)

8/18/11 ND ND 17.2/25.5 FT, FRSunset Beach (44.931N,85.965W)

6/30/11 480 ± 153 (E) 310 ± 111 (E) 17.0/20.0 FT, FR fish (∼700gobies)

8/18/11 3500 ± 1200 (E) 230 ± 75 (E) 16.8/24.4 FT, FRGood Harbor Bay Trail(44.938N, 85854W)

8/08/11 5300 ± 1683 (E) 2250 ± 810(E)

18.2/22.7 FT, FR

8/18/11 300 ± 96 (E) 4600 ± 1479 (E) 16.8/24.4 FT, FR10/03/11 ND 110 ± 35 (E) 14.0/17.0 FT, FR

County Road 669(44.939N, 85.870W)

8/08/11 210 ± 64 (E) 18.2/22.7 FT, FR10/03/11 ND 14.0/16.0 FT, FR

Esch Road (44.760N,86.076W)

8/15/11 ND 1000 ± 316 (E) 16.4/21.2 FT, FR

Peterson Road (44.732N,86.103W)

8/15/11 9800 ± 3600 (E) 15000 ± 5100(E)/ 2000 ± 650 (A)

16.4/21.2 FT, FR

Dimmick’s Point(44.059N, 85.965W)

10/04/11 1000 ± 370 (E) 3800 ± 1300(E)

NM/15.5 FT, FR

Door County, WI Europe Bay #1 (45.112N,87.07W)

6/21/11 580 ± 183 (E) 15.0/14.6 FT, FR7/26/11 410 ± 128 (E) NM/25.5 AT, FR 2 birds

Europe Bay #2 (45.258N,86.97W)

6/21/11 80 ± 26 (E) 15.0/14.7 AT, FR7/26/11 1400 ± 464 (E) NM/25.5 FT, FR 2 birds8/25/11 1000 ± 331 (E) NM/18.5 FT, FR

Racine, WI Shoop Park (42.779N,87.764W)

6/20/11 630 ± 355 (E) 15.1/16.7 FT, FR fish (>100alewife)

7/26/11 1500 ± 550(E) 18.2/24.2 FT, FR fish (>100alewife)

7/26/11 640 ± 240 (E) 18.2/24.2 FT, SN fish (>100alewife)

8/25/11 270 ± 28 (E) 22.8/21.8 FT, FR10/10/11 78 ± 34 (E) 16.7/16.8 FT, SN fish

Light House (42.781N,87.758W)

6/20/11 190 ± 61 (E) 15.3/14.4 FT, FR fish (>100alewife)

7/26/11 830 ± 260 (E) 18.0/24.2 FT, FRSamuel Meyers Park(42.443N, 87.463W)

6/20/11 340 ± 112 (E) 15.0/20.1 FT, SN7/26/11 2900 ± 974 (E)/65 ± 22 (B) 21.0/22.0 FT, SN8/25/11 490 ± 150 (E) 22.8/21.8 FT, FR 1 fish and 1

birdEichelman (42.344N,87.383 W)

6/20/11 150 ± 50 (E) 15.0/17.7 FT, SN

Carre Hogle (42.711N,87.771W)

8/25/11 270 ± 88 (E) 21.0/21.8 FT, SN

Indiana Dunes National Lakeshore, IN,and Southern Lake Michigan, IL

Calumet Park(41.392N,87.399W)

6/21/11 12 ± 3 (E) 19.6/30.2 AT/SN7/21/11 NDc 19.4/31.1 AT/SN9/6/11 ND NMd/17.2 FT/SN

Ogden Dunes (41.622N,87.191W)

6/21/11 ND 17.5/32.4 AT/FR7/21/11 ND 19.0/31.5 AT/FR9/06/11 ND NM/17.2 FT/SN

63rd Street (42.344N,87.485W)

6/21/11 ND 17.4/30.2 FT/FR7/21/11 ND 19.0/31.5 FT/FR

Burns Ditch (41.58N,78.19W)

9/6/11 11 ± 2.7 (E) NM/17.2 FT/SN9/22/11 ND NM/17.6 FT/SN

Hamlin Beach State Park, NY HYC (43.364N, 77.951W) 7/21/11 3400 ± 1100 (E) 3600 ± 1200(E)

NM/30.0 FT, FR

8/24/11 4800 ± 1700 (E) 12000 ± 4400(E)

NM/21.1 FT, FR 3 birds

aMPN values are means ± SE on means. Numbers in parentheses refer to bont gene type detected. bAlgal condition abbreviation are as follows AT,attached; FT, detached and floating; SN, senescent (decomposing); FR, fresh (little to no decomposition). cND, not detected or below detectionlimit. dNM, not measured.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942589

The samples consisted of detached, floating algal matter innearshore waters but had not yet washed-up onto the drybeach. In one case, lake bottom-attached Cladophora, sand, anddeep water were collected in October at Good Harbor BayTrail in SLBE from approximately 10 m below the water surfaceby a diving team. Samples were collected by gloved hands,placed in Whirl-Pak bags, and transported to the laboratory onice at 4 °C. The samples were immediately shipped to theUniversity of Minnesota and were analyzed within 48 h ofcollection. Nearly axenic Pithophora spp. (a filamentous greenalga) grown in a fish tank in the lab was used as the negativecontrol because there was no contamination with clostridia.2.2. MPN-PCR Analyses. Five-tube-MPN analyses were

used to identify and quantify types of Cladophora-borne C.botulinum.30 Wet algae (50 g) from the pooled Cladophorasamples (16.6 g of each original replicate) were blended with100 mL of phosphate-buffered water (PBW; 10 mM sodiumphosphate, pH 6.8). Benthic organisms attached to Cladophorawere removed prior to homogenization. The blender wascontinually flushed with N2 gas to minimize exposure of C.botulinum to oxygen. The algal homogenate was serially diluted10-fold, three times, in PBW. From each dilution, a five-tubeMPN was set up by adding 1 mL aliquots of each dilution to 9mL of prereduced trypticase-peptone-glucose yeast extract(TPGY) broth as described by Hielm et al.31 In addition, someof the attached benthic organisms removed from Cladophorawere also incubated in TPGY broth (each organism in 9 mLbroth). Anaerobic tubes were sealed using gastight rubberstoppers and aluminum caps. Immediately after the assembly,the tubes were flushed for 3 min with N2 gas, passed through0.2 μm sterilized filter with an 18-gauge needle, to removeresidual oxygen in the headspace, and tubes were incubated at26 °C for 5 d. After incubation, 2 mL aliquots from each tubewere vigorously shaken, centrifuged at 8000 rpm for 10 min,and the pelleted cells were saved at −20 °C for DNAextraction. Around 15 g triplicate samples of the remaining algalhomogenates were placed overnight in an oven at 85 °C ovento measure the dried weight of algal matter.Genomic DNA from the pelleted cells prepared above was

directly extracted using the UltraClean Microbial DNAIsolation Kit (MO BIO Laboratories, Carlsbad, CA) as perthe manufacturer’s protocol, with slight modifications: thepellets were resuspended in buffer and transferred into amicrobead tube with MD1 solution (surfactant and otherdisruption agents for cell lysis) and tubes were incubated at 80°C for 10 min in a dry bath before a 10 min vortexing step toincrease yield and deactivate potential BoNTs that are heatlabile. The non-DNA organic and inorganic materials were alsoprecipitated with C3 solution before spin-filtering to furtherpurify DNA.Multiplex PCR was used for the detection of genes encoding

for C. botulinum toxins A, B, C, E, and F.32−34 Theoligonucleotide primer sets used are listed in Table S1 of theSupporting Information. Two multiplex PCRs were conductedfor toxins A, C, and F, and B and E based on the amplicons’size. PCR was performed in 25 μL reactions containing 1 μL oftemplate, 400 nM of each primer (Integrated DNATechnologies Inc., Coralville, Iowa), 220 μM of eachdeoxynucleotide triphosphate (Invitrogen, Carlsbad, CA), 1.5mM MgSO4, 10 mM KCl, 8 mM (NH4)2 SO4, 10 mM Tris-HCl, pH 9.0, 0.05% NP-40, 200 ng μL−1 BSA, and 0.08 U μL−1

of DNA Polymerase (Denville Scientific Inc.). PCR was doneusing 35 cycles of denaturation at 95 °C for 30 s, annealing at

60 °C for 25 s, and elongation at 72 °C for 1 min, followed by afinal extension at 72 °C for 3 min. Amplified PCR productswere visualized on 1.5% agarose gels, and stained with ethidiumbromide. Positive controls consisted of plasmids containing thetargeted bont genes35 and no-template negative controls wereused for all PCR reactions. Tubes showing positive PCRreactions for any of the tested genes, as determined by gelelectrophoresis of PCR products, were counted as positive foreach type of C. botulinum. Bacterial counts were expressed asMPN per gram dry weight of algae. The identity of all PCRproducts was determined by DNA sequence analysis done atthe University of Minnesota Biomedical Genomics Center.BLAST analyses were used to identify the homology of theamplified target sequence for bont genes to those in Genbank.

2.3. Heat Treatment. A comparative study of selectedCladophora samples was done to determine if vegetative cellsand/or spores of C. botulinum were present in Cladophora mats.Subsamples of Cladophora homogenates were held at 80 °C for10 min in a water bath to kill vegetative cells prior to three-tubeMPN-PCR analysis.36 The MPN counts from these sampleswere compared to controls with no treatment (total count).

2.4. Identification of C. botulinum NeurotoxinSerotypes Using Mouse Bioassay and Antibody Neu-tralization Analysis. MPN cultures from Cladophora sampleswith elevated MPN values were analyzed for actively growingand toxin-producing C. botulinum by using a mouse bioassay.37

Briefly, MPN cultures were diluted in 0.5 mL gelatin phosphatebuffer (30 mM sodium phosphate, pH 6.3, plus 0.2% gelatin),in 1:1 ratio and injected via i.p. (intraperitoneal) into mice.Female ICR (CD-1) mice were obtained from Harlanlaboratories. Each mouse within the group (2 mice) wereinjected with 0.5 mL of the toxin-antibody mixture andobserved for 4 d for symptoms of botulism poisoning. Samplescausing death indicated that neurotoxin was present. Todetermine the toxin serotypes produced in the samples,antibody neutralization analyses were employed. Anti-BoNT/A (10 international unit/ml; IU/ml), anti-BoNT/B (10 IU/ml), and/or anti-BoNT/E (100 IU/ml) antibodies used inneutralization analysis and were obtained from Centers ofDisease Control and Prevention, Atlanta, GA. One IU of anti-BoNT protects a mouse against a 10 000 mouse medium lethaldose units (LD50/mouse) of BoNT/A and/B or 1000 LD50/mouse of BoNT/E. The LD50 of BoNT is estimated to beabout 1 ng/kg in mice. For an initial neutralization analysis,toxicity was estimated to be about 10 LD50/mouse based on thetime of death (mice died within 24 h) and toxin was dilutedwith gelatin phosphate to achieve the appropriate LD50concentration. The mixtures of toxin and antibody wereincubated at 37 °C for 90 min prior to injection. All procedureswith mice were performed in accordance with the InstitutionalAnimal Care and Use Committee in the Department ofBacteriology at the University of Wisconsin.

3.0. RESULTS3.1. Type and Abundance of C. botulinum Associated

with Cladophora. A total of 53 Cladophora mats collectedfrom June to October in 2011 were tested for the presence andtype of C. botulinum using botulism toxin gene specific primerpairs (Table S1 of the Supporting Information). Among the fivebont genes examined, C. botulinum type E strains were prevalentin Cladophora mats collected from the Great Lakes. Sixteen of22 (73%) Cladophora mats from SLBE and 23 of 31 (74%)algal mats collected from other shorelines in the Great Lakes

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942590

contained bont/E genes (Table 1). In addition to type E toxingenes, we also detected bont types A and B genes in a fewCladophora samples from SLBE (Peterson Beach) and Racine’sSamuel Meyers Park. These BoNT types have been commonlyassociated with human diseases rather than birds.20 The PCRproducts of bont gene-positive algal samples were confirmed byDNA sequencing.The population density of C. botulinum type E strains in algal

mats ranged from 100−105 MPN per gram dried algae (Table 1and Figure 2) based on the assumption of one copy of eachgene per genome and that generally C. botulinum types A, B, E,and F neurotoxin genes are located on the chromosome orplasmids, whereas in types C and D the genes are found on abacteriophage.20,38 Whereas the population level of C.botulinum varied widely in Cladophora samples from SLBEbeaches, algal samples collected in August contained thegreatest number of C. botulinum, up to 15 000 MPN/g driedalgae. The Cladophora samples from Hamlin Beach State Parkin Lake Ontario also had a high population of C. botulinum typeE (104−105 MPN/g dried algae). In contrast, most samplescollected from locations along southern Lake Michigan werenegative for bont/E genes, and only 2 of 10 samples (20%)contained C. botulinum type E, with counts in other samplesgenerally below the detection limit (<20 MPN/g dried algae).Moreover, Cladophora mats collected in July and Augustgenerally had a greater number of C. botulinum type E thanthose collected in June and October (Figure 2).

A comparison of C. botulinum densities present in nearshorefloating Cladophora versus Cladophora attached to the bottomof lake is shown in Figure 3. Water, sediment, and mussels(Dreissena spp.) near Cladophora mats were also collected whensubmerged Cladophora was sampled from Good Harbor Bay

Figure 2. Population density of C. botulinum type E in Cladophora samples as determined by MPN-PCR. Legend: (A) SLBE, (B) the City of Racine,(C) Door County (C), and (D) Hamlin Beach State Park. The error bars indicate standard errors.

Figure 3. Comparison of density of C. botulinum type E in nearshorefloating Cladophora mats vs Cladophora in water, mussel internaltissue, and sand samples collected near the lake bottom (∼10 mdepth) at Good Harbor Bay Trail (SLBE). The error bars indicatestandard errors.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942591

Trail at SLBE in October. Generally, nearshore floatingCladophora harbored a higher population density of C.botulinum than did Cladophora collected at the lake bottomsuggesting that the population of C. botulinum likely increasesin algal mats when Cladophora becomes detached andaccumulates along the shoreline. Moreover, MPN values ofC. botulinum in floating algal mats were greater than thosefound in sediment and in mussel internal body parts (musseltissue), both of which have been suggested as sources of avian-acquired C. botulinum.16,29

3.2. Vegetative Cells vs Spores. To determine if C.botulinum was present in Cladophora mats as vegetative cells orspores, MPN-PCR analyses were done using Cladophorasamples treated with heat (80 °C for 10 min.) and comparedto MPN counts from Cladophora samples receiving notreatment. PCR results in Figure S1 of the SupportingInformation show that, whereas Cladophora samples (Racine’sSamuel Meyers Park) without treatment clearly have the bont/Egene, heat treatments resulted in elimination of vegetative cells.This suggests that C. botulinum is present as vegetative cellsrather than as dormant spores in the algal mats. We alsoconducted heat treatment analysis on Cladophora samples thatwere positive for bont/E genes and were collected from SLBEin August and City of Racine in July. All samples receiving heattreatment were negative for bont/E or had low MPN (<10)values of C. botulinum indicating that mostly vegetative cellswere present in the Cladophora mats.3.3. Mouse Bioassays. Four Cladophora samples were

examined for the ability of the attached C. botulinum to activelyproduce BoNT under favorable environmental conditions.Samples were selected based on the type of bont gene present,as determined by PCR, and those having high MPN values:SLBE-Peterson Beach (August), Racine-Samuel Meyers Park(July), Hamlin Beach State Park-HYC (August), and SLBE-Dimmick’s Point (October). Analyses were done onCladophora cultures enriched in TPGY broth. Results inTable S2 of the Supporting Information show that all samplesactively produced BoNT type E (<3000 LD50/ml) and werefound to be neutralized by BoNT/E antitoxin, except forCladophora obtained from SLBE-Dimmick’s Point, which had arelatively low population density of C. botulinum. The SLBE-Peterson Beach sample was positive for both bont/A gene byPCR, but the culture did not contain active BoNT/A. Similarly,the sample from Racine’s Samuel Meyers Park was negative forBoNT/B.

4.0. DISCUSSION

Whereas the potential for association between Cladophora matsand C. botulinum has been hypothesized in the past,6,19 directreports linking Cladophora as a possible carrier or anenvironmental sources of bont/E pathogen are limited. Inaddition, there are no studies showing the number of C.botulinum present in Cladophora, whether it has the capacity toproduce toxin, or if the pathogen exists as spores or asvegetative cells. Byappanahalli and Whitman6 detected thebont/E gene in Cladophora mats from three beaches at SLBEwhen the samples were incubated under anaerobic conditionwithout media. These results preliminarily supported thehypothesis that Cladophora may be a potential habitat for thegrowth of C. botulinum. However, the linkage between thepresences of C. botulinum in Cladophora remained speculativegiven the small number of Cladophora samples analyzed and a

lack of information concerning the population size of C.botulinum in the algal mats examined.In our study, 53 nearshore Cladophora mats were collected

from wide geographic areas in Lake Michigan and Lake Ontarioand were analyzed to determine whether C. botulinum canoccur and grow on the algal vegetation found in the GreatLakes. MPN-PCR analyses indicated that C. botulinum type Epredominated in nearshore floating algal mats compared toother types of bont genes (serotypes). Whereas only a few algalsamples contained bont/A and /B genes, which are commonlyassociated with human diseases, about 74% (39 of 53) of algalmats from shorelines of the Great Lakes harbored the bont/Egene and population densities were relatively high, up to 15 000MPN per gram of dried algae. Considering that the populationdensity of C. botulinum in soils or sediments across the USranges from 0.021 to 1.28 MPN per gram of soil,20 thepopulation densities of C. botulinum in Cladophora mats aresignificantly greater on a weight basis. The high incidence of thebont/E gene in nearshore Cladophora mats further supports theproposed association between Cladophora and C. botulinum.6

Moreover, Cladophora mats collected in July and Augustgenerally had a greater number of C. botulinum type E genesthan did those collected in June and October (Figure 2)suggesting that water temperature or algal growth character-istics may affect the population density of this bacterium. Sincebotulism outbreaks generally occur from June throughNovember and bird die-offs begin in the heat of the summerand reach their peak during the fall bird migrations,4 theincrease of C. botulinum populations in Cladophora inmidsummer may subsequently increase the likelihood ofbotulism outbreaks as the toxin moves up the food chain.Results of this study also established that the bont/E gene

was also present in the benthic environment (sediment,Cladophora attached to rocks on sediment, water nearsediment, and in mussel tissue), but the population densitiesof C. botulinum in benthic matrices were lower than thosefound in nearshore floating Cladophora, which may startdecaying. The low C. botulinum densities in Cladophora on thebottom of the lake is consistent with a previous report of nodetection of bont/E gene in benthic algal samples byquantitative PCR.29 Given the resilience of C. botulinum sporesto environmental stresses, this strongly suggests the C.botulinum likely proliferate on Cladophora mats. Presumably,spores, vegetative cells, and/or both in benthic sediment maybe transferred to fresh Cladophora attached to sediment andbacteria (and more vegetative cells) accumulate in the algae at atime when the algae becomes detached, stranded nearshore,and subsequently starts to decompose.This hypothesis is further strengthened by our finding that C.

botulinum is mostly present as vegetative cells in Cladophoramats. Because MPN-PCR analyses cannot provide informationon whether the causative agent is present as vegetative cells,spores, or both, the MPN counts from heat treated samplescompared to those of controls without treatment implied C.botulinum may actively grow on the algal mats. On the basis ofthe fact that BoNT is produced only when the pathogen isactively growing and proliferating, vegetative cells of C.botulinum likely produce and accumulate toxins within decayingCladophora mats, which provide necessary nutrients for theirgrowth. The findings that Cladophora harbors enteric bacteriaincluding pathogens (e.g., E. coli, enterococci, Salmonella,Campylobacter)26,27 and a range of microbial communities,

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942592

including sulfate-reducers such as clostridia,21 supports thisobservation.We tested four Cladophora samples that were positive for

bont/E gene and had different population densities (MPNvalues) of Cladophora-borne C. botulinum to determine if theywere able to produce active toxin given favorable environmentalconditions. Mouse toxin-antitoxin neutralization assay con-firmed that enriched cultures from three Cladophora samplesproduced detectable levels of BoNT/E. Cladophora samplesobtained in October at the SLBE-Dimmick’s Point site, whichwere negative for toxin may have been colonized by nontoxin-producing strains of C. botulinum. A significant portion (∼41%)of isolates from birds and fish obtained from a recent outbreakin Lake Erie and Ontario were found to be nontoxin-producingstrains.16 Another possibility is that the concentration of BoNTproduced by the enrichments perhaps was well below the lethalthreshold for mice because the population size of toxigenic C.botulinum on the Cladophora (∼1000 MPN/g dried algae) wasalso low. This same reasoning may apply to Cladophorasamples, which contained bont/A and /B genes but failed toproduce toxins. Therefore, our data indicate that finding thatthe presence of bont genes alone is not sufficient evidence toconclude that lethal amounts of botulism toxin are produced ina specific matrix.One surprising result of this study was that C. botulinum

likely exists on, or in, Cladophora mats as vegetative cells ratherthan spores. This indicates that Cladophora likely provides afavorable habitat for C. botulinum growth, and perhaps toxinproduction. We believe that large deposits of decaying algaeunder water, such as those found at Good Harbor Bay Trail atSBLE, (personal communication Chris Otto, Biologist, SLBE,National Park Service) can serve as sources of C. botulinum in asimilar manner to floating and beached algal mats. While notyet definitive based on one year’s data alone, our findingsstrongly lend support to the hypothesis that Cladophora playsan important role in the pathogen transmission pathway andmay potentially transfer spores, vegetative cells, or toxin tohigher trophic levels and ultimately to aquatic birds. On thebasis of these assumptions, a proposed pathway for theinvolvement of Cladophora in the botulism poisoning of birds isas follows: C. botulinum spores become incorporated intoCladophora via transfer from benthic sediment and sub-sequently germinate into actively growing and toxin-producingcells once the redox conditions become lower due toheterotrophic microbial growth occurring during the Clado-phora decaying process. Bacterial growth and/or toxinaccumulate in the algae and are transferred to fish directly, orvia invertebrates or mussels, to fish and then to birds. Thishypothesis is somewhat similar to that proposed byByappanahalli and Whitman (2009). We often observed thepresence of chironomids (larvae of Dicrotendipes spp. andPotthastia spp.) and dreissenids in Cladophora mats during ourinvestigations and our preliminary findings show that 6 of 9benthic organisms collected from Cladophora contained thebont/E gene by PCR. Since larvae of chironomids are detritus-feeders or scrape material from plants and other organicsubstrates, and are important food sources for fish and aquaticorganisms, floating, and beached mats of algae might exposebirds to botulism toxin-laden invertebrates from lower trophicfeeding guilds. Clearly, further studies are needed to elucidatethe dominant pathway of C. botulinum transfer from algae tobirds.

■ ASSOCIATED CONTENT

*S Supporting InformationPrimer sets used to detect BoNT genes, detection of bont/Egenes from samples with and without heat treatment, mousetoxin and neutralization assays. This material is available free ofcharge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Author*E-mail: sadowsky@umn.edu, tel: (612) 624-2706.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This work was supported by a grant from the U.S.Environmental Protection Agency (EPAR5-GL2010-1).Wethank Steve Yancho, Chris Otto, and Emily Kobernik(SLBE); Stephan Kurdas and Julie Kinzelman (City of Racine,WI); Karen Terbush, Jay G. Bailey, Jeff Meyer (Hamlin BeachState Park, NY); Dawn Shively (USGS, Great Lakes ScienceCenter); and Greg Kleinheinz, Kimberly Busse, and BrookeJansen (U. of Wisconsin-Oshkosh) for providing Cladophoramat samples and beach conditions. Any use of trade, product,or firm names is for descriptive purposes only and does notimply endorsement by the U.S. Government. This article isContribution 1735 of the USGS Great Lakes Science Center.

■ REFERENCES(1) Brand, C. J.; Duncan, R. M.; Garrow, S. P.; Olson, D.; Schumann,L. E. An outbreak of type E botulism among common loons (Gaviaimmer) in Michigan’s upper peninsula. J. Wildl. Dis. 1988, 24, 471−476.(2) USGS National Wildlife Disease Center, Disease Information,Avian Botulism , 2008 http://www.nwhc.usgs.gov/disease_information/avian_botulism/index.jsp (accessed October 2012).(3) Zuccarion-Crowe, C. Type E botulism. In Nearshore Area of theGreat Lakes 2009, USEPA and Environment Canada. EPA 905-A-09-013, p 99−103; 2009.(4) U.S. Environmental Protection Agency Great Lakes NationalProgram Office, Type E Botulism Outbreaks: A Manual for BeachManagers and the Public, 2012 http://www.michigan.gov/documents/dnr/Botulism_Manual_2012_390795_7.pdf (accessedOctober 2012).(5) Kaufmann, O. W.; Fay, L. D., Clostridium botulinum type E toxinin tissues of dead loons and gulls. In Michigan State UniversityAgricultural Experiment Station Quarterly Bulletin, p 236−242; 1964.(6) Byappanahalli, M. N.; Whitman, R. L. Clostridium botulinum typeE occurs and grows in the alga Cladophora glomerata. Can. J. Fish.Aquat. Sci. 2009, 66, 879−882.(7) Smith, L. D. S.; Sugiyama, H. Botulism. The Organism, Its Toxins,the Disease. Springfield, IL: Charles C. Thomas, 1988.(8) Hambleton, P. Clostridium botulinum toxins: A general review ofinvolvement in disease, structure, mode of action and preparation forclinical use. J. Neurol. 1992, 239, 16−20.(9) Domske, H. M.; Obert, E. C. Avian Botulism in Lake Erie. GreatLakes Research Review. 2001, 1−5.(10) Long, S. C.; Tauscher, T. Watershed issues associated withClostridium botulinum: A literature review. J. Water Health 2006, 4,277−288.(11) Wong, D. M.; Young-Perkins, K. E.; Merson, R. L. Factorsinfluencing Clostridium botulinum spore germination, outgrowth, andtoxin formation in acidified media. Appl. Environ. Microbiol. 1988, 54,1446−1450.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942593

(12) Rocke, T. E.; Samuel, M. D. Water and sediment characteristicsassociated with avian botulism outbreaks in wetlands. J. WildlifeManage. 1999, 63, 1249−1260.(13) Rocke, T. E.; Euliss, N. H., Jr.; Samuel, M. D. Environmentalcharacteristics associated with the occurrence of avian botulism inwetlands of a Northern California refuge. J. Wildlife Manage. 1999, 63,358−368.(14) U.S. Geological Survey, Avian Botulism, 1999 http://www.nwhc.usgs.gov/publications/field_manual/chapter_38.pdf (accessed Octo-ber 2012).(15) Yule, A. M.; Barker, I. K.; Austin, J. W.; Moccia, R. D. Toxicityof Clostridium botulinum type E neurotoxin to Great Lakes fish:Implications for avian botulism. J. Wildlife Dis. 2006, 42, 479−493.(16) Hannett, G. E.; Stone, W. B.; Davis, S. W.; Wroblewski, D.Biodiversity of Clostridium botulinum Type E associated with a largeoutbreak of botulism in wildlife from Lake Erie and Lake Ontario.Appl. Environ. Microbiol. 2011, 77, 1061−1068.(17) Getchell, R. G.; Culligan, W. J.; Kirchgessner, M.; Sutton, C. A.;Casey, R. N.; Bowser, P. R. Quantitative polymerase chain reactionassay used to measure the prevalence of Clostridium botulinum type Ein fish in the lower Great Lakes. J. Aquat. Anim. Health 2006, 18, 39−50.(18) Getchell, R. G.; Bowser, P. R. Ecology of type E botulism withindreissenid mussel beds. Aquatic Invaders 2006, 17, 1−8.(19) Hecky, R. E.; Smith, R. E.; Barton, D. R.; Guildford, S. J.; Taylor,W. D.; Charlton, M. N.; Howell, T. The nearshore phosphorus shunt:a consequence of ecosystem engineering by dreissenids in theLaurentian Great Lakes. Can. J. Fish. Aquat. Sci. 2004, 61, 1285−1293.(20) Simjee, S. Foodborne Diseases; Humana Press: Totowa, NJ, 41−70, 2007.(21) Olapade, O. A.; Depas, M. M.; Jensen, E. T.; McLellan, S. L.Microbial communities and fecal indicator bacteria associated withCladophora mats on beach sites along Lake Michigan shores. Appl.Environ. Microbiol. 2006, 72, 1932−1938.(22) Depew, D. C.; Houben, A. J.; Guildford, S. J.; Hecky, R. E.Distribution of nuisance Cladophora in the lower Great Lakes: Patternswith land use, near shore water quality and dreissenid abundance. J.Great Lakes Res. 2011, 37, 656−671.(23) Taft, C. E. History of Cladophora in the Great Lakes. InCladophora in the Great Lakes; Shear, H., Konasewich, D.E., Eds.;Great Lakes Research Advisory Board, International Joint CommissionRegional Office: Windsor, Ontario, p 5−16; 1975.(24) Stevenson, R. J.; Stoermer, E. F. Seasonal abundance patterns ofdiatoms on Cladophora in Lake Huron. J. Great Lakes Res. 1982, 8,169−183.(25) Chilton, E. W.; Lowe, R. L.; Schurr, K. M. Invertebratecommunities associated with Bangia Atropurpurea and CladophoraGlomerata in Western Lake Erie. J. Great Lakes Res. 1986, 12, 149−153.(26) Byappanahalli, M. N.; Shively, D. A.; Nevers, M. B.; Sadowsky,M. J.; Whitman, R. L. Growth and survival of Escherichia coli andenterococci populations in the macro-alga Cladophora (Chlorophyta).FEMS Microbiol. Ecol. 2003, 46, 203−211.(27) Ishii, S.; Yan, T.; Shively, D. A.; Byappanahalli, M. N.; Whitman,R. L.; Sadowsky, M. J. Cladophora (Chlorophyta) spp. harbor humanbacterial pathogens in nearshore water of Lake Michigan. Appl.Environ. Microbiol. 2006, 72, 4545−4553.(28) Perez-Fuentetaja, A.; Clapsadl, M.; Einhouse, D.; Bowser, P.;Getchell, R.; Lee, W. Influence of limnological conditions onClostridium Botulinum Type E presence in Eastern Lake Erie sediments(Great Lakes, USA). Hydrobiologia 2006, 563, 189−200.(29) Perez-Fuentetaja, A.; Clapsadl, M. D.; Getchell, R. G.; Bowser,P. R.; Lee, W. T. Clostridium botulinum type E in Lake Erie: Inter-annual differences and role of benthic invertebrates. J. Great Lakes Res.2011, 37, 238−244.(30) Standard Methods for the Examination of Water and Wastewater,22nd ed.; Eaton, A. D., Clesceri, L. S., Greenberg, A. E., eds.; AmericanPublic Health Association, American Water Works Association, andWater Environmental Federation: Washington D.C., 2012.

(31) Hielm, S.; Hyytia, E.; Ridell, J.; Korkeala, H. Detection ofClostridium botulinum in fish and environmental samples usingpolymerase chain reaction. Int. J. Food Microbiol. 1996, 31, 357−365.(32) Hill, B.; Skerry, J.; Smith, T.; Arnon, S.; Douek, D. Universaland specific quantitative detection of botulinum neurotoxin genes.BMC Microbiol. 2010, 10, 267.(33) Lindstrom, M.; Keto, R.; Markkula, A.; Nevas, M.; Hielm, S.;Korkeala, H. Multiplex PCR assay for detection and identification ofClostridium botulinum Types A, B, E, and F in food and fecal material.Appl. Environ. Microbiol. 2001, 67, 5694−5699.(34) Franciosa, G.; Fenicia, L.; Caldiani, C.; Aureli, P. PCR fordetection of Clostridium botulinum type C in avian and environmentalsamples. J. Clin. Microbiol. 1996, 34, 882−5.(35) Marshall, K. M.; Bradshaw, M.; Johnson, E. A. Conjugativebotulinum neurotoxin-encoding plasmids in Clostridium botulinum.PLoS ONE 2010, 5, E11087.(36) Dezfulian, M.; McCroskey, L. M.; Hatheway, C. L.; Dowell, J.,V.R. Selective medium for isolation of Clodstridium botulinum fromhuman feces. J. Clin. Microbiol. 1981, 13, 526−531.(37) U.S. Food and Drug Administration, Bacteriological analyticalmanual Chapter 17: Clostridium botulinun, 2001, http://www.fda.gov/F o o d / S c i e n c e R e s e a r c h / L a b o r a t o r y M e t h o d s /BacteriologicalAnalyticalManualBAM/ucm070879.htm. (accessed Jan-uary 2013).(38) Hielm, S.; Bjarkroth, J.; Hyytia, E.; Korkeala, H. Genomicanalysis of Clostridium botulinum group II by pulsed-field gelelectrophoresis. Appl. Environ. Microbiol. 1998, 64, 703−708.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304743m | Environ. Sci. Technol. 2013, 47, 2587−25942594