Culture-dependent characterization of the microbial community associated with epizootic shell disease lesions in American lobster, Homarus americanus.

ABSTRACT Epizootic shell disease in the American lobster is an important factor affecting lobster fisheries in and around the Long Island Sound. It is a strictly dermal disease, because no correlation was observed between occurrence of epizootic shell disease and hemolymph infection. The culturability of bacteria from lesions was variable and averaged around 1.1%. The lesions contained two to four orders of magnitude more bacteria than healthy carapace surfaces of the same animal. Chitinoclastic bacteria comprised a very small traction of bacteria present in the lesions, suggesting that their role in epizootic shell disease may be limited. Phylogenetic analysis of bacteria isolated from the lesions showed no typical bacterial pathogens of lobsters such as Aerococcus viridans or Vibrio fluvialis. Moreover, bacteria commonly associated with shell disease of other Crustacea or other forms of shell disease of the American lobster were not found. Two common groups of bacteria were isolated from lesions of all lobsters used in this research: one belonging to a species complex affiliated with the Flavobacteriaceae family and the second belonging to a series of closely related if not identical strains of Pseudoalteromonas gracilis. Bacteria isolated from only a few lobsters were related to Shewanella frigidimarina, Alteromonas arctica, Vibrio lentus, Shewanella fidelia, Pseudoalteromonas tunicata and Vibrio spp. Based on the analyses of culturable isolates, overall microbial communities found in lesions of lobsters from eastern Long Island Sound and Buzzards Bay appear to be similar to each other.

KEY WORDS: Homarus americanus, lobster, epizootic shell disease, bacterial infection, chitinolytic bacteria, 16S rDNA

INTRODUCTION

Pathologic erosions of carapace have been described for various species of Crustacea and are known as shell disease (Fisher 1974, Getchell 1989). The etiology of shell disease is generally considered to be bacterial, although fungi have been implicated in shell disease pathogenesis as well (Bruns et al. 1979). Vogan at al. (2002) have isolated a number of chitinolytic bacteria from lesions of the edible crab, Cancer pagurus, which have been assigned as Vibrio and Xanthomonas species. Vibrio and Pseudomonas spp. have been isolated from lesions of the blue crab, Callinectes sapidus (Noga et al. 1994). Bacteriologic investigation of shell disease in the deep sea red crab, Geryon quinquedens, has revealed the presence of Vibrio and Flavobacterium spp. and Escherichia coli in lesions (Bullis et al. 1988). Bacteria tentatively assigned to Vibrio spp. have also been isolated from shell disease lesions of the Pacific blue shrimp, Litopenaeus stylirostris (Goarant et al. 2000).

Shell disease syndrome has been described for at least three species of lobsters (i.e., Homarus americanus, Nephrops norvegicus and Panulirus argus). Porter et al. (2001) have isolated representatives of Vibrio, Pseudoalteromonas and Shewanella spp. from lesions of the spiny lobster, Panulirus argus.

Shell disease has been a commonly recognized problem of the American lobster, Homarus americanus, H. Milne Edwards (1837). It was originally reported in lobsters held in impoundments during the winter months (Hess 1937, Smolowitz et al. 1992). Carapace lesions have also been associated with offshore canyon lobsters (Ziskowski et al. 1996). In both cases the disease starts as a black spot or "cigarette burn" mark, which then may further progress to more generalized erosions of the carapace. The microbiology of impoundment shell disease has been examined by several authors (Fisher 1977, Malloy 1978, Stewart 1980 and Getchell 1989). The list of bacterial genera isolated from the lesions of American lobster with impoundment shell disease includes Vibrio, Aeromonas, Beneckea and Pseudomonas spp.

During the last 5 y, however, a new form of shell disease has been found at unusually high levels in free-living lobster populations (Smolowitz et al. 2002). This form of shell disease is characterized by moderate to deep erosions in the carapace along the dorsum of the cephalothorax and abdomen. As the lesions deepen, the exocuticle is eroded leaving behind cuticular matrix that forms skeletonal pillars in the lesions. This contrasts with impoundment shell disease, where erosions do not routinely occur as thin vertical erosions into the carapace, but rather appear as progressive areas of cleanly removed cuticle that leave little to no cuticular matrix behind (Smolowitz et al. 1992). We proposed to term this novel form of shell disease "epizootic shell disease" (Smolowitz et al. 2002).

Epizootic shell disease has been found at 22.7% and 11.6% in wild lobster populations from eastern Long Island Sound to Buzzards Bay, respectively (Castro & Angell 2000). Incidence of epizootic shell disease has also been reported for Cape Cod Bay and central Long Island Sound, indicating further geographic expansion of the disease. Except for the 2003 outbreak of shell disease in Kittery, Maine, other northern lobster populations have not yet been afflicted by epizootic shell disease. However, the observed tendency of shell disease expansion among wild lobster populations may suggest that epizootic shell disease will likely spread even there.

The microbial community involved in development of epizootic shell disease is also dominated by bacteria (Chistoserdov et al. 2002). The goal of the present research is to characterize bacteria comprising this community. Culture-based techniques were used to isolate bacteria from lesions followed by 16S rDNA sequencing and identification of phylogenetic affiliation of the lesion bacteria. Some data presented here were reported earlier at the National Shellfisheries Association Meeting in Mystic, CT (Chistoserdov et al. 2002).

MATERIALS AND METHODS

Sampling

Fourteen lobsters with lesions were collected from eastern Long Island Sound (ELIS) in 2001 by the Connecticut Department of Environmental Protection. ELIS lobsters were designated with the numbers ELIS1 through ELIS 14. One and four lobsters were provided by New York Department of Environmental Conservation in 2001 and 2002, respectively. These were collected from the coastal waters of central Long Island Sound (CLIS). The only 2001 lobster was assigned the number CLIS1. The 2002 lobsters received the numbers CLIS2 through 5 and were used only in a limited number of experiments. Ten lobsters with shell disease collected from Buzzards Bay (BB) were assigned the numbers BB1 through BB10. All diseased lobsters had various degrees of severity of cuticular lesions and primarily substages C3 and C4 of cuticular development (determined as per Smolowitz et al. 1992). The lobsters were sacrificed and used for collection of hemolymph and lesion material for microbiologic analyses.

Generally, a half of the carapace lesion(s) was used to collect bacterial biomass and a half was preserved in 4% formaldehyde prepared in sterile seawater for further histopathologic examination. Tail lesions were also used to collect bacterial biomass in selected animals. Before dissection, lobsters were gently rinsed with sterile seawater (autoclaved and then filtered through a 0.2 [micro]m filter). Microbiologic material from lesions and healthy carapace surfaces was collected by scraping with a sterile razor blade. A repeated scraping of the same areas with a second sterile razor blade indicates that this procedure removes >99% of all microorganisms along with some semidegraded shell material. The scraped material was resuspended in 7 mL of sterile seawater, vigorously vortexed for 1 min and then processed. A half-milliliter of the suspension was fixed with formaldehyde (3.7% final concentration), centrifuged for 1 min at x 14,000g and then the pellet was resuspended in 50% ethanol and 1 x PBS (130 mM sodium chloride and 10 mM sodium phosphate buffer, pH 7.2) solution. This suspension was stored at -20[degrees]C and later used for the total count of bacteria. The remaining material was used for isolation of bacteria (1.5 mL) and isolation of total nucleic acids (5 mL). The scraped material from healthy carapaces was also suspended in sterile seawater to optical densities similar to those of lesion material suspensions and processed in the fashion similar to that of lesion material.

Approximately 5 mL of hemolymph were drawn directly from the heart of each lobster either into sterile Vacutainer tubes (Becton-Dickinson, Rutherford, N J) with anticoagulant and refrigerated or immediately plated on Petri dishes with appropriate media. To prevent contamination of hemolymph samples, the puncture area was washed and wiped with sterile seawater. Five [micro]l and 100 [micro]L of hemolymph were plated on two sets of Petri dishes containing Brain Heart Infusion Agar with 10% sheep blood or Rabbit Blood Agar (both from Hardy Diagnostics, Santa Maria, CA). These two sets were incubated at 2 different temperatures, 22[degrees]C and 37[degrees]C. Growth observations were made every 24 h.

Bacterial Abundance

This was determined using microscopic census. Preserved samples were stained with fluorochrome (DAPI or acridine orange) and captured on 0.2-[micro]m Nuclepore filters. Filter-retained cells were enumerated and measured by epifluorescent microscopy according to Taylor et al. (1986) using an Olympus BX51 epifluorescent microscope.

Culture-dependent Microbiological Analyses of Microbial Communities in Shell Lesion Material

All media components, except agar, were from Sigma (Saint Louis, MO); agar was from Gibco BRL (Gaithersburg, MD). Four media were tested: Difco Marine Agar 2216 (MA; Becton-Dickinson, Rutherford, MD) and three types of seawater-based media designed for detection of chitinolytic bacteria. Seawater Agar (SA; 1.7% agar prepared with 35 ppt sterilized seawater), Artificial Seawater Agar (ASA; 3.5% artificial seawater salt mix from Sigma [Saint Louis, MO]--equivalent to 35 ppt seawater and 1.7% agar) and Seawater Agar II (SAII; 35 ppt seawater with 1.7% agar, 0.1% peptone and 0.01% Tween 80). All seawater-based media were supplemented with 1 mL of a filter-sterilized vitamin mix per 1 L of medium. The vitamin stock solution contained biotin (2 mg/L), folic acid (2 mg/L), thiamine HCl (5 mg/mL), Ca pantothenate (5 mg/mL), B12 (0.1 mg/L), riboflavin (5 mg/L) and nicotinamide (5 mg/L). Petri dishes with the seawater-based media were overlaid with 5 mL of "soft" or "hard" chitin suspension. Soft and hard chitin suspensions contained 3% chitin, 3.5% sodium chloride and 0.7 or 1.7% of agar Difco, respectively, in distilled water.

Serial dilutions ([10.sup.-2], [10.sup.-4], [10.sup.-6], [10.sup.-8]) of lesion material were prepared in sterile seawater and plated on Petri dishes with MA and seawater-based media in triplicate. Plates were incubated at room temperature (22[degrees]C) and growth observations were made every 24 h. Individual colonies were picked with sterile toothpicks and re-streaked several times on new plates with appropriate media to isolate/verify that pure cultures were obtained.

16S rRNA Analysis of Bacterial Isolates

DNA from isolated bacteria was purified using the Marmur procedure (Marmur 1961). DNA from lesions was isolated using a method of Xu and Tabita (1996), except that an RNase treatment was omitted. Isolated DNA was separated in agarose gels; images of the gels were acquired using a CCD camera and band intensities were quantified using SigmaGel 1.0 (Jandel Scientific, San Rafael, CA, USA). An approximately 1,500 kb fragment of the 16S rRNA gene was amplified using primers fD1 and rP3 (Weisburg et al. 1991) with the following modifications: fD1 was AGATTTGATCHTGGCTYAG and rP3 was ACGGNTACCTTGTTACGACTT. Restriction fragment length polymorphism (RFLP) analysis of the 1,500 kb fragment was carried out using HaeIII, HhaI and HinfI. Unique 16S rRNA gene sequences selected by RFLP were sequenced using the PCR protocol and primers as described by Borneman et al. (1996). On average a 700 bp portion of the 16S rRNA gene was sequenced (approximately bases 600 through 1300, E. coli numbering). The SEQUENCE MATCH program from the Ribosome Database Project II (RDP) and BLAST were used to search for closest phylogenetic neighbors in the RDP and GenBank databases, respectively. Chimeric 16S rDNA sequences were identified by the CHIMERA CHECK program from the RDP. Sequences were aligned using the CLUSTAL X program, and alignments were checked and corrected manually. Regions of ambiguous alignment were masked out for further analysis. The phylogenetic trees were constructed using 2 independent algorithms: distance analyses with Jukes-Cantor (programs DNA-DIST [Jukes-Cantor] and FITCH from the PHYLIP package, version 3.5c, [http://evolution.genetics.washington.edu/phylip.htmll) or Saitou and Nei corrections (CLUSTAL X program) and by maximum-likelihood (DNAml program also from PHYLIP) with randomizing sequence input order and with and without global rearrangements. To determine confidence values for individual branches of each final tree, bootstrap analysis was applied to each generated tree using the SEQBOOT and CONSENSE programs from the PHYLIP package or the bootstrap option in the CLUSTAL X program.

RESULTS

Optimization of Growth Medium

Bacteria on the lobster surfaces (normal carapace and lesions) are exposed to 2 types of nutrients: those present in seawater and those that bacteria can extract from the carapace. We tested several media for isolation of bacteria from shell lesions and normal carapaces (ASA, SA, SAII and MA). The ASA medium is very poor, because its only nutrients are organic contaminants from the agar. The SA medium contains organic compounds found in natural seawater but lacks compounds leached from lobster shell. The SAII medium best reflects the natural composition of lobster shell, because, in addition to chitin, it contains proteins and lipids also found in lobster shell, and it is prepared with seawater. The MA medium is rather unnatural, because it is very rich in nutrients in comparison with the environment in which lesion bacteria exist. The ASA medium showed poor bacterial recovery compared with other types of media (data not shown) and was not used in further analyses. Numbers of bacterial colonies emerged under different growth conditions are summarized in Table 1. Bacterial growth developed on MA plates within 72 h of incubation and no new colonies appeared following further incubation. It took more than one week for bacteria to develop fully-grown colonies (>2 mm in diameter) on SA plates. The numbers and morphotypes of colonies on SA plates were identical to those on SAII plates; however, bacteria grew faster on SAII plates. One week incubation was always sufficient. The numbers of colonies and number of dominant colony morphotypes were similar but not always identical for MA and SA plates incubated aerobically (see Table 1). After initial experiments, we typically used MA and SAII media for further microbiologic analysis. In addition to colony forming unit counts, we also counted bacteria using epifluorescent microscopy (see Table 1).

Individual colony morphotypes of bacteria, which grew on MA or SAII media, were counted. Bacteria from MA plates were streaked on SAII plates and visa versa. For a simplicity of the analysis, colony morphotypes were divided into "rare" (<1% of all colonies) and "common" (>5% of all colonies). There were no bacteria with an intermediate number (i.e., > 1% and <5%) of colonies. The "rare" colonies were one-two colonies on plates with low dilutions and were likely seawater- or sediment-associated contaminants. As we argue below, the "common" colonies are actual members of the lesion community. Most of the "common" colony morphotypes isolated from different lobsters were similar and were found on both MA and SAII plates. They included: (1) yellow to slightly reddish, round opaque colonies with smooth edges on MA tentatively assigned to Pseudoalteromonas or Vibrio spp., same colonies appeared white, opaque and slimy with smooth edges on SAII; (2) yellow to slightly bluish, round translucent colonies on MA also tentatively assigned to Pseudoalteromonas or Vibrio spp., same colonies appeared whitish, translucent with smooth edges; (3) small slow growing bright-yellow colonies that look identical on MA and SAIl and likely belonging to the Flavobacterium-Cytophaga-Bacteroides division.

Bacteria in Lesions Versus Healthy Carapace

In a separate experiment, we compared the number of bacteria present on healthy carapaces and in lesions. Four lobsters (CLIS2 through CLIS5) had both lesions and healthy surfaces on their carapaces, which allowed us to sample approximately 5 [cm.sup.2] of lesion as well as healthy areas of carapace of each animal. The scrapings were used to determine viable counts of bacteria and to isolate total DNA. Both viable counts and quantities of isolated DNA were normalized to the surface area of sampled carapace (Table 2). These data indicate that "normal" carapace contains two to four orders of magnitude less of bacteria than lesions. There were no correlations between the diversity of colony morphotypes and the source of colonies (i.e., lesions or healthy carapace). The three "common" colony types (see earlier section) were isolated from both healthy carapace and lesions.

Isolation of Chitinoclastic Bacteria

Most marine bacteria are able to use colloid chitin. We focused, however, on chitinoclastic bacteria able to use crude chitin. A powder from crab shells is a more adequate imitation of lobster shell material in its biochemical composition than colloid chitin.

Although bacterial colonies were formed within 1 wk of incubation on SAII plates with chitin overlays, it took 2 to 4 wk for noticeable chitinolytic activity to develop around the colonies. During this long incubation period, many gliding and swarming bacteria would cover the agar, sometimes completely obscuring halos formed around colonies of chitinoclastic bacteria. In several cases, we failed to detect the presence of chitinoclastic bacteria in lesions of several lobsters (see Table 1), despite some indirect indications that they might be present. Isolation of pure cultures of chitinolytic bacteria was also hampered by gliding and swarming bacteria. Through multiple restreaking, however, we managed to isolated pure cultures of chitinolytic bacteria from 5 lobsters, one of which was from ELIS and four from BB (see Table 3). Only two types of colonies of chitinoclastic bacteria were isolated. One of the two is a motile (gliding) bacterium forming bright yellow colonies, indicative of a member of the Flavobacterium Cytophaga-Bacteroides division. It was isolated from all five lobsters. These bacteria could completely dissolve particulate chitin, rendering the medium around their colonies completely translucent. The second belonged to bacteria forming whitish translucent colonies and it was isolated from only two lobsters. The clearing of chitin created by these bacteria was never complete: it appeared that smaller particles of chitin were removed, leaving large particles intact.

Culture-dependent Microbiological Analyses of Hemolymph

Five and 200 microliters of hemolymph from 29 lobsters were streaked or spread onto rabbit or sheep blood agar plates. No growth was observed on plates incubated at 37[degrees]C. Bacterial growth was detected on duplicate plates incubated at room temperature with plated hemolymph from the lobsters ELIS2, ELIS3, ELIS5, ELIS11, ELIS12, CLIS4 and CLIS5 (Table 4). Hemolymph from five healthy (not numbered) CLIS lobsters was sterile. Thus, hemolymph of seven lobsters contained some bacteria and only one lobster, ELIS3, was heavily infected.

16S rRNA Analysis of Representative Bacterial Isolates

Based on colony morphologic and tinctorial properties and microscopic observation of the cells, isolates belonging to one "common" type (see earlier) but isolated from different lobsters appeared to be identical. RFLP analyses of 16S rRNA genes confirmed that all isolates belonging to the same "common" type but isolated from different lobsters are indeed identical. We also analyzed by RFLP 16S rRNA genes from the isolated chitinoclastic bacteria as well as bacteria isolated from hemolymph. Comparison of RFLP patterns of "common" isolates and chitinoclastic bacteria suggested that chitinoclastic bacteria with yellow colonies were closely related to "common" isolates with bright-yellow colonies and that the chitinoclastic bacteria with white translucent colonies were closely related to the two other "common" morphotypes. The hemolymph isolates, on the other hand, were completely unrelated to bacteria isolated from lesions but some were related to each other (see Table 4).

For 16S rDNA sequence analysis, we selected to sequence portions of the 16S rRNA genes from all bacteria isolated from the lobsters ELIS6 and ELIS10 as representatives, including both "common" and "rare" isolates 16S rRNA genes from all chitinoclastic isolates and most bacteria isolated from hemolymph were also sequenced. The sequence information was used to identify in the GenBank and Ribosomal DNA project II databases closest relatives of our isolates. The closest relatives of bacteria from the lesions are shown in Table 3 and closest relatives of hemolymph isolates are shown in Table 4. We also carried out two types of phylogenetic analysis of 16S rRNA sequences of bacteria isolated from lesions. Both maximum likelihood and neighbor-joining algorithms generated identical trees in which lesion isolates clustered with their closest relatives (not shown).

Chitinoclastic bacteria forming yellow colonies belong to one of the three closely related species, which we propose to assign tentatively to "the lobster Flavobacteriaceae species complex." The first species, identified as Aquimarina muelleri (Nedashkovskaya et al. 2005) was isolated from 3 lobsters (BB3, BB8 and BB9) from Buzzards Bay. The second, identified as another Aquimarina sp., was isolated from lobster BB4. The third, most closely related to Cellulophaga baltica (99% identity of 16S rRNAs), was isolated from one Buzzards Bay and one ELIS lobster (i.e., BB9 and ELIS10). Two other isolates from the same species complex were identical strains MA7.3 and MA 11.3 and the isolate MA7.1 from lobsters ELIS6 and ELIS10, respectively. The former two belong to a Psychroserpens species and the latter is closely related to Lacinutrix copepodicola. Bacteria belonging to the lobster Flavobacteriaceae species complex formed type 3 "common" colonies and were isolated from each lobster used in analysis.

All Pseudoalteromonas gracilis isolates had identical 16S rRNA sequences, although substantial variations in colony morphology were observed. For example, four isolates from lobster ELIS10 were initially considered as different based on colony morphology. MA11.1 and SA11.3 belonged to colony type 1, with yellow to slightly reddish, round opaque colonies with smooth edges, and MA11.2 and SA11.4 belong to colony type 2 with yellow to slightly bluish, round translucent colonies. Two chitinoclastic strains of P. gracilis were isolated from lobster ELIS10 and lobster BB4. Their 16S rRNA sequences were identical to each other and to all other P. gracilis isolates. It appears that P. gracilis strains are also ubiquitous in lobster lesions.

We also sequenced and analyzed the 16S rRNA genes from several "rare" isolates. Two "rare" isolates from lobster ELIS10 (i.e., SA11.1 and SA 11.2) were shown to be related to Shewanella frigidimarina and Alteromonas arctica. Other "rare" isolates include strains related to Vibrio lentus, Shewanella fidelia and Vibrio spp. On occasion, we also observed conspicuous melanin-pigmented colonies belonging to Pseudoalteromonas tunicata.

DISCUSSION

Epizootic shell disease in the American lobster may become an important factor affecting lobster fisheries in and around the Long Island Sound. Although generally nonlethal, epizootic shell disease substantially decreases the value of landed lobsters due to their nonaesthetic appearance. The etiology of epizootic shell disease is considered to be bacterial; however, no detailed description of the bacterial community associated with lobster lesions has been carried out.

There is evidence suggesting that epizootic shell disease is not a systemic infection or an immunocompromised condition caused by a latent infection but it is rather a strictly dermal disease. First, tissues of lobsters with epizootic shell disease had no pathologic changes other than on the carapaces. Second, lobsters with epizootic shell disease effectively fight off the disease by developing an adequate immune response to the carapace infection (see Smolowitz et al. 2005). Third, there was no correlation between epizootic shell disease and the presence of bacteria in hemolymph. Hemolymph of only six out of 29 lobsters contained some bacteria and only one lobster (ELIS3) was heavily infected. Most lobsters had no detectable bacteria in the hemolymph (Table 4).

Since epizootic shell disease appears to be a dermal condition, this research was primarily focused on the analysis of cultivable bacteria from the lesions. The culturability of bacteria from lesions was variable and averaged around 1.1%. Based on the number of colony forming bacteria, the lesions contained about four orders of magnitude more bacteria than healthy carapace surfaces of the same animal. Quantitation of DNA isolated from healthy carapace surfaces and lesions also showed that at least two orders of magnitude more bacterial cells were present in lesions than on the healthy carapace. Although it was assumed that all isolated DNA derives from bacteria, it is possible that some lobster DNA and/or DNA from protozoa was coisolated from healthy carapace material. Such contamination may explain why the two methods (i.e., colony count and DNA isolation) gave a different proportion for bacterial abundances on healthy carapaces and in lesions. The diversity of isolates from healthy carapace surfaces and lesions was variable but comparable with each other. Similar "common" and "rare" colony morphology types were observed in both cases. Porter et al. (2001) analyzed the microbial community associated with healthy carapace and lesions of the spiny lobster from which they isolated similar members of Vibrionaceae. They concluded "... that the natural flora is responsible for the lesions seen in P. argus." Although it seems that the normal flora from American lobsters may be involved in the development of epizootic shell disease, another explanation is also likely. Because healthy carapaces contain much fewer bacteria, bacteria isolated from them could be either transient or indicative of developing shell disease lesions.

The chitinoclastic bacteria able to degrade chitin have long been suggested to be culprits of shell diseases in Crustacea (Fisher 1977, Malloy 1978, Stewart 1980, Getchell 1989, Vogan et al. 2002). Our data, however, indicate that chitinolytic activity is attributed to a very small fraction of the bacteria in the lesions (Table 1). Most of the bacteria present in the lobster lesions are not able to degrade lobster chitin. This observation is also in agreement with pathologic observations (Smolowitz et al. 2005). The pillars of chitin remaining in epizootic shell disease lesions indicate that bacteria mostly attack other polymers in the carapace, leaving chitin more or less intact. It is also possible that these few bacteria, which are able to degrade crude chitin, form the leading edge of the infection. The remaining bacteria, comprising most of the bacterial biomass in the lesion, use other nutrients in the lobster shell, principally protein.

Phylogenetic analysis of bacteria isolated from the lesions showed no typical bacterial pathogens such as Aerococcus viridans or Vibrio fluvialis associated with epizootic lobster shell disease. Bacteria found in lesions of lobsters with epizootic shell disease were also different from those isolated from lesions of other forms of shell disease of the American lobster, as well as other Crustacea. In lesions of lobsters with epizootic shell disease, we observed 2 common groups of bacteria: one belonging to a species complex affiliated with Flavobacteriaceae and the second belonging to a series of closely related if not identical strains of Pseudoalteromonas gracilis. Vibrio species commonly found in shell lesions of other crustaceans or other forms of shell disease in lobsters were isolated only occasionally and always as "rare" isolates.

Based on the analyses of culturable isolates, overall microbial communities found in lesions of lobsters from ELIS and BB are similar to each other and contain at least two groups of bacteria: one belonging to the lobster Flavobacteriaceae species complex and to one of the strains of P. gracilis. Challenge experiments are required to elucidate the role of the two in the development and progression of epizootic shell disease in the American lobster.

ACKNOWLEDGMENTS

The authors thank Penelope Howell (Connecticut DEP), Carl LoBue (New York DEC) and Bruce Estrella (MA Division of Marine Fisheries) for the gift of lobsters for this research. This research was supported by the NY Sea Grant #R/FBM-25 to AYC and RS. This publication was also supported by the National Sea Grant College Program of the US Department of Commerce's National Oceanic and Atmospheric Administration under award #NA16RG1354 to the Research Foundation of State University of New York for New York Sea Grant. The views herein do not necessarily reflect the views of any of those organizations.

LITERATURE CITED

Borneman, J., P. W. Skroch, K. M. O'Sullivan, J. A. Palus, N. G. Rumjanek, J. L. Jansen, J. Nienhuis & E. W. Triplett. 1996. Molecular microbial diversity of an agricultural soil in Wisconsin. Appl. Environ. Microbiol. 62:1935-1943.

Bruns, C. D., M. E. Berrigan & G. E. Henderson. 1979. Fusarium sp. infection in freshwater prawn Macrobrachium rosenbergii (De Man). Aquaculture 16:193-198.

Bullis, R., L. Leibovitz, L. Swanson & R. Young. 1988. Bacteriologic investigation of shell disease in the deep-sea red crab, Geryon quinquedens. Biol. Bull. 175:304.

Castro, K. M. & T. E. Angell. 2000. Prevalence and progression of shell disease in American lobster, Homarus americanus, from Rhode Island waters and the offshore canyons. J. Shellfish Res. 19:691-700.

Chistoserdov, A., F. Mirasol & R. Smolowitz. 2002. Characterization of microbial assemblages involved in the development of shell disease in the American Lobster, Homarus americanus. J. Shellfish Res. 21:410.

Fisher, W. S. 1977. Shell disease of lobsters. In: C.J. Sindermann, editor. Disease diagnosis and control in North American aquaculture. Developments in aquaculture and fisheries science. Vol 6. Amsterdam: Elsevier. pp. 158-167.

Getchell, R. G. 1989. Bacterial shell disease in crustaceans: a review. J. Shellfish Res. 8:1-6.

Goarant, C., R. Brizard & A.-L. Marteau. 2000. A white spot disease-like syndrome in the Pacific blue shrimp (Litopenaeus stylirostris) as a form of bacterial shell disease. Aquaculture 185:25-30.

Hess, E. 1937. A shell disease in lobsters (Homarus americanus) caused by chitinovorus bacteria. J. Biol. Board Can. 3:358-362.

Malloy, S. C. 1978. Bacteria induced shell disease of lobsters. J. Wildlife Dis. 14:2-10.

Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3:208-218.

Nedashkovskaya, O. I., S. B. Kim, A. M. Lysenko, G. M. Frolova, V. V. Mikhailov, K. H. Lee & K. S. Bae. 2005. Description of Aquimarina muelleri gen. nov., sp. nov., and proposal of reclassification of (Cytophaga) latercula Lewin 1969 as Stanierella latercula gen. nov., comb. nov. Int. J. Syst. Evol. Microbiol. 55:225-229.

Noga, E. J., D. P. Engel, T. W. Arroll, S. McKenna & M. Davidian. 1994. Low serum antibacterial activity coincides with increased prevalence of shell disease in blue crabs Callinectes sipadus. Dis. Aquat. Org. 19: 121-128.

Porter, L., M. Butler & R. H. Reeves. 2001. Normal bacterial flora of the spiny lobster Panulirus argus and its possible role in shell disease. Mar. Freshwater Res. 52:1401-1405.

Smolowitz, R. M., R. A. Bullis & D. A. Abt. 1992. Pathologic cuticular changes of winter impoundment shell disease preceding and during intermolt in the American lobster, Homarus americanus. Biol. Bull. 183:99-112.

Smolowitz, R., A. Hsu, E. Summers & A. Chistoserdov. 2002. Lesions associated with recent epizootic shell disease in Homarus americanus on the northeast coast. J. Shellfish Res. 21:412.

Smolowitz, R., A. Y. Chistoserdov & A. Hsu. 2005. A description of the pathology of epizootic shell disease in the American lobster, Homarus americanus. J. Shellfish Res. 24(3):749-756.

Stewart, J. E. 1980. Diseases. In: J. S. Cobb & B. F. Phillips, editors. The biology and management of lobsters, Vol. 1: Physiology and behavior. New York: Academic Press. pp. 321-329.

Taylor, G. T., D. M. Karl & M. L. Pace. 1986. Impact of bacteria and zooflagellates on the composition of sinking particles: an in situ experiment. Mar. Ecol. Prog. Ser. 29:141-155.

Weisburg, W. G., S. M. Barns, D. A. Pelletier & D. J. Lane. 1991. 16S ribosomal DNA amplification for phylogenetic studies. J. Bacteriol. 173:697-703.

Vogan, C. L., C. Costa-Ramos & A. F. Rowley. 2002. Shell disease syndrome in the edible crab, Cancer pagurus--isolation, characterization and pathogenicity of chitinolytic bacteria. Microbiol-UK 148:743-754.

Xu, H. H. & F. R. Tabita. 1996. Ribulose-1,5-bisphosphate carboxylase/ oxygenase gene expression and diversity of Lake Erie planktonic microorganisms. Appl. Environ. Microbiol. 62:1913-1921.

Ziskowski, J., R. Spallone, D. Kapareiko, R. Robohm, A. Calabrese & J. Pereira. 1996. Shell disease in American lobster in the offshore Northwest-Atlantic region around the 106-mile sewage-sludge disposal site. J. Mar. Environ. Eng. 3:247-271.

ANDREI Y. CHISTOSERDOV, (1) * ROXANNA SMOLOWITZ, (2) FELIZA MIRASOL (3) AND ANDREA HSU (2,4)

(1) Department of Biology, University of Louisiana at Lafayette, P.O. Box 42451, Lafayette, Louisiana 70504; (2) Marine Biological Laboratory, 7 MBL Drive, Woods Hole, Massachusetts 02543; (3) Marine Sciences Research Center, Stony Brook University, Stony Brook, New York 11794; (4) Boston University Marine Program, 7 MBL Drive, Woods" Hole, Massachusetts 02543

* Corresponding author. E-mail: [e-mail]

TABLE 1.
Total bacterial counts, number of different colony morphotypes
and total counts of chitinolytic bacteria in epizootic shell
disease lesions.

                        Colony Forming Units ([ml.sup.-1]) *
                           Number of Colony Morphotypes

Animal
Number                    MA Medium            SA Medium

ELIS1     carapace    9.6 x [10.sup.9]/5   8.0 x [10.sup.9]/3
ELIS2     carapace    6.0 x [10.sup.8]/3   5.4 x [10.sup.8]/4
ELIS3     carapace    3.1 x [10.sup.8]/3   3.0 x [10.sup.8]/3
ELIS4     carapace    2.5 x [10.sup.8]/4   2.4 x [10.sup.8]/6
CLIS1     carapace    2.4 x [10.sup.8]/2   2.5 x [10.sup.8]/4
          tail        3.0 x [10.sup.8]/2   2.5 x [10.sup.8]/4
ELIS5     carapace    5.0 x [10.sup.7]/4   7.0 x [10.sup.7]/4
          tail        1.2 x [10.sup.8]/4   1.3 x [10.sup.8]/2
ELIS6     carapace    9.0 x [10.sup.7]/4   8.0 x [10.sup.7]/3
ELIS7     carapace    7.0 x [10.sup.7]/2   8.0 x [10.sup.7]/4
ELIS8     carapace    4.0 x [10.sup.8]/4   4.0 x [10.sup.8]/3
ELIS9     carapace    1.7 x [10.sup.8]/4   1.9 x [10.sup.8]/4
ELIS10    carapace    2.8 x [10.sup.8]/3   3.2 x [10.sup.8]/4
ELIS11    carapace    1.8 x [10.sup.8]/4   1.7 x [10.sup.8]/3
ELIS12    carapace    1.4 x [10.sup.7]/3   1.1 x [10.sup.7]/2
ELIS13    carapace    9.2 x [10.sup.8]/4   6.1 x [10.sup.8]/4
ELIS14    carapace    4.0 x [10.sup.7]/2   3.9 x [10.sup.7]/3
          tail        4.4 x [10.sup.6]/6   4.3 x [10.sup.6]/3

Animal     Chitinolytic Colony       Microscopic Counts
Number    Numbers ([ml.sup.-1])   of Bacteria ([ml.sup.-1])

ELIS1       7.0 x [10.sup.5]          1.6 x [10.sup.10]
ELIS2             ND **               1.2 x [10.sup.10]
ELIS3             ND                  1.5 x [10.sup.10]
ELIS4       2.0 x [10.sup.3]          2.5 x [10.sup.10]
CLIS1       1.0 x [10.sup.4]          1.7 x [10.sup.10]
            1.0 x [10.sup.4]          1.7 x [10.sup.10]
ELIS5              ND                 2.5 x [10.sup.10]
            1.0 x [10.sup.4]          2.0 x [10.sup.10]
ELIS6       4.5 x [10.sup.4]          1.9 x [10.sup.10]
ELIS7       5.0 x [10.sup.3]          2.3 x [10.sup.10]
ELIS8       1.8 x [10.sup.4]          2.2 x [10.sup.10]
ELIS9       6.0 x [10.sup.5]          2.6 x [10.sup.10]
ELIS10      1.4 x [10.sup.4]          1.7 x [10.sup.10]
ELIS11            ND                  2.3 x [10.sup.10]
ELIS12            ND                  2.6 x [10.sup.10]
ELIS13            ND                  2.3 x [10.sup.10]
ELIS14            ND                  3.4 x [10.sup.10]
            1.0 x [10.sup.3]          2.0 x [10.sup.10]

* = per 1 mL bacterial suspension.

** = ND, not detected.

TABLE 2.
Total number of colony forming units from lesions and healthy
portions of lobster carapaces.

                                                     Yield of
                               Colony Forming        Bacteria
                            Units ([cm.sup.-2]) *       DNA
Animal                         on MA Number of      ([micro]g/
Number                       Colony Morphotypes     [cm.sup.2])

CLIS2    Lesions             9.4 x [10.sup.9]/6        1.11
CLIS2    Healthy carapace    4.0 x [10.sup.5]/6        0.01
CLIS3    Lesions             3.0 x [10.sup.9]/9        1.00
CLIS3    Healthy carapace            **                0.013
CLIS4    Lesions             4.0 x [10.sup.9]/9        0.67
CLIS4    Healthy carapace    2.2 x [10.sup.5]/6        0.00
CLIS5    Lesions             8.0 x [10.sup.9]/5        1.03
CLIS5    Healthy carapace    7.0 x [10.sup.5]/10       0.013

* = per one square centimeter of carapace surface.

** = ND. not determined.

TABLE 3.
Isolates from lesions of representative lobsters with shell
disease, Long Island Sound and Buzzards Bay.

Animal                                             Closest Relative
Number         Isolate *               (Sequence Accession Number)

ELIS6    MA7.1                      Lacinutrix copepodicola (AY694001)
         SA7.1 = MA7.3              Psychroserpens sp. (AY167341)
         SA7.2 = SA7.3 =            Pseudoalteromonas gracilis
         = MA7.2 = MA7.4              (AF038846)
ELIS10   11a2 = 11b1                Cellulophaga baltica (AJ005972)
         MA11.3                     Psychroserpens sp. (AY167341)
         MA11.1 = MA11.2 =          Pseudoalteromonas gracilis
         SA11.3 = SA11.4 =            (AF038846)
         11b1white
         SA11.1                     Shewanella figidimarina (AJ300833)
         SA11.2                     Psychromonas arctica (AF374385)
BB3      18a = 18b1 = 18b2 = 18c2   Aquimarina muelleri (AY608407)
BB4      19b2 = 19a2 = 19b1         Aquimarina muelleri (AY608407)
         19b1 white                 Pseudoalteromonas gracilis
                                      (AF038846)
BB8      23a1 = 23b1 = 23c1         Aquimarina muelleri (AY608407)
BB9      24c1 = 24e1                Cellulophaga baltica (AJ005972)
         24d1 = 24d2 = 24f1         Aquimarina muelleri (AY608407)

Animal                   Degradation of
Number   % of Identity    Crude Chitin

ELIS6         98
              98
              99            positive
ELIS10        99            positive
              98
              99            positive
              99            negative
              98            negative
BB3           99            positive
BB4           97            positive
              99            positive
BB8           99            positive
BB9           99            positive
              99            positive

* -strain designation beginning with SA means that strain was isolated
on the SA or SAII medium; strain designation beginning with MA means
that strain was isolated on MA; strain designations beginning with a
number indicate that strains were isolated as chitin degraders.

TABLE 4.
Presence of bacteria in hemolymph of lobsters.

                                                       Density in
                                                       Hemolymph
  Lobster #                 Isolates               (Cells m[L.sup.-1]

ELIS1           sterile                                   <10
ELIS2           Pseudomonas fragi H2                      1200
ELIS3           Brochothrix thermosphacta H3        >2 x [10.sup.6]
ELIS4           sterile                                   <10
ELIS5           Pseudomonas sp. H6                        7200
ELIS6-ELIS10    sterile                                   <10
ELIS11          H12.1 *, H12.2, H12.3, H12.4 **      2 x [10.sup.4]
ELIS12          H13.1 *, H13.2, H13.3 **                  3200
ELIS13-ELIS14   sterile                                   <10
BB1-BB10        sterile                                   <10
CLIS1-CLIS3     sterile                                   <10
CLIS4           unidentified                               25
CLIS5           unidentified                               5

* = isolates H12.1 and H13.1 are identical;

** = isolates H13.3 and H12.4 are identical.