Two seasonal surveys were conducted for the environmental presence of Hematodinium sp. Eighteen sites within Maryland and Virginia coastal bays were investigated, encompassing the Sinnepuxent, Newport, and Chincoteague Bays, south of the Ocean City, MD inlet (Figure 1). The time frame for sampling was initiated with emergence of blue crabs from winter hibernation in April 2010, through late summer to early autumn, and into winter hibernation beginning in November 2010. This sampling regimen was repeated in 2011 and permitted for investigation of known Hematodinium-positive sites during periods prior to, and throughout, a seasonal pattern of high prevalence in summer and low to non-existent infections during the winter. Sediment and plankton samples were collected from each site, and the DNA from the biota present was extracted for further analysis (see Materials and Methods).

Endpoint PCR-based analysis, using Hematodinium-specific primers, detected 48 of 546 sample sites (8.8%) as positive, based upon observation of a definitive band corresponding to the predicted amplicon length of 285 bp (data not shown). The distribution was relatively even throughout all months, and also between plankton and sediment samples, although positive detections for 2010 were higher than in 2011 (Table 1). In addition, Hematodinium sp. was never detected in water samples in May or November. Varying environmental parameters, such as dissolved oxygen and water temperature differences, did not correlate to any patterns in detection (data not shown). Three widely dispersed locations (sites 12, 16, and 17) were never positive in our assays.

Shown are the sites and dates in which Hematodinium sp. detection occurred, based upon gene-specific PCR analysis. The sites correspond to those labeled in Figure 1B. Dates are indicated by the month, followed by the year. Typically sampling was conducted in the second week of each month, unless weather conditions precluded collection to the third week.

Collectively, our data suggested that putative free-living Hematodinium sp. maintain a continuous environmental presence in water and/or sediment in the Delmarva coastal ecosystem. During 2010 Sinnickson, VA (site 10) was positive in our PCR screen for all months in either water or sediment, and in addition it was one of four sites in which water samples were positive in April (Table 1). The April water data was of particular interest as during this month crab infections have never been reported, and thus release of Hematodinium sp. cells from diseased crabs was unlikely. As Sinnickson represented a potential hotspot for environmental transmission, we sought to further explore the dinoflagellate population structure from water samples at this site.

To detect Hematodinium sp., along with other dinoflagellates, we generated an 18 S rRNA clone library that targeted a conserved region of the gene. We reasoned that enumeration of Hematodinium sp. within the library would yield a rough estimate of its abundance relative to other free-living dinoflagellate species, in addition to confirming our temporal observations. The seasons in which various dinoflagellate species are known to bloom has been established in the related ecosystem of Chesapeake Bay 22, and thus their presence in specific months provided an internal control for dinoflagellate population structure analysis.

Dinoflagellate 18 S rRNA clone library results for Sinnickson in 2010 are presented in Table 2 and Figure 2, subdivided into categories of early spring (April), late spring (June), summer (July/August), and autumn (October). In total, Hematodinium sp. was detected in 40/70 (57%) of water samples. This percentage was unexpectedly high, and may represent some unknown sampling bias. The plankton filter used has a pore size of 20 μm which in theory should allow released amoeboid trophonts, in addition to individual micro- and macrodinospores, to pass though the net 17, but would still retain cells either in the process of division or in clumps. Alternatively, cells may associate with small plankton or particulates. Surprisingly, analysis of individual months revealed that April has the highest relative prevalence (15/16 or 94%) for Hematodinium sp. (see Figure 2). The temporal distribution of other dinoflagellates was consistent with the succession pattern from Chesapeake Bay 22. Heterocapsa rotundata, commonly present in Chesapeake Bay through winter and spring, represented 77% (10/13) of the clones from June sequences. In summer and early autumn a mix of species that included H. rotundata and Gymnodinium sp. was observed, although the majority of dinoflagellates remained Hematodinium sp. (23/41 or 56%).

Shown are sequencing results from analysis of libraries generated using dinoflagellate-specific 18 S rRNA.

A phylogenetic analysis of the 2010 Sinnickson samples was also performed (Figure 3). Of the 40 Hematodinium sp. clones in the library, 35 were identical and are labeled in the phylogenetic tree as “The consensus sequence.” Five other Hematodinium sp. clones, from the months indicated, contained single nucleotide polymorphisms, with the exception of April clone b which had two (data not shown). The Hematodinium sequences detected in the Sinnickson samples were more closely related to an isolate from Callinectes sapidus than to Hematodinium spp. isolated from other non-portunid host crustacean species such as Cancer pagurus and Carcinus maenas. Other representative dinoflagellate clones from the 2010 Sinnickson library are also included, and they are related to a diverse group of dinoflagellates (Figure 3).

It has been suggested that the species of Hematodinium infecting crabs in the waters of the Delmarva Peninsula is a host generalist, based on the high sequence identity of the ITS-1 region from various infected crustaceans 16. This observation is consistent with similar data derived from analysis of both clade A and clade B Hematodinium disease systems 23. We were thus interested to analyze the population structure in our study, particularly as the temporal distribution patterns suggests a continuous environmental presence at hotspot sites such as at Sinnickson, VA in 2010. In addition to Sinnickson, we concurrently analyzed Tom’s Cove, VA and Newport Bay, MD (sites 10, 13, and 3 respectively). These locations produced 37.5% (18/48) of the total positive identifications, and also represent distributions that are in the northwest, southwest, and southeast ends of Chincoteague Bay (see Figure 1B).

Clone libraries were generated from PCR products amplified by the Hematodinium-specific ITS1/5.8 S rRNA primers used in our initial detection assays. In Table 3 is a comparison of the relative identity of our clones derived from three hotspots of environmental presence. The populations were homogeneous, with approximately 83.2% of all clones identical (109/131). Of the remaining 22, 18 contained single nucleotide polymorphisms (SNPs). Surprisingly, these SNPS did not predominantly cluster in the ITS-1 region of the amplicon, as 10 of the 18 clones contained substitutions in the 5.8 S rRNA gene.

Clone libraries were generated from Hematodinium-specific ITS1/5.8 S rRNA primers used in our initial detection assays, with locations and dates of analysis. Identical sequences were present in 83% of the clones in the libraries. p/w: plankton-water samples; Sed: sediment samples.

An alignment of the dominant clones in our libraries is shown in Figure 4 contrasted to sequence from a Hematodinium sp. isolated from an infected blue crab. A SNP between these is located at position 40 of the 5.8 S rRNA, with G substituting for A. However, it should be noted that a single clone in our library matched identically with these alternative sequences. Also shown in the alignment are the four clones containing two substitutions. Despite these minor differences, all samples were >98.5% identical or higher in sequence, suggesting a genetically homogenous population that is likely derived from a single species.

From April to November of 2010 and 2011, monthly sediment and water samples were collected from 18 sites between the northern end of the Maryland coastal bays and the southern end of the Virginia coastal bays (Figure 1). The only exception was September of 2010, where only sites 2, 3, and 5 were sampled. Typically sampling was accomplished in the second week of each month, over a two-day time span, unless weather conditions delayed collection to the third week.

For sediment collection, a Ponar grab was used to collect ~ 40 g of sediment, with two replicates per site. To collect phytoplankton samples, a plankton net with a 30 cm diameter was trawled through sub-surface water for 3 minutes at boat speeds ranging from 3–4 mph. The total water volume sampled was thus estimated to be approximately 25 m³ for each collection. The retention cup was fitted with a 20 μm filter to capture phytoplankton and microzooplankton. A final volume of between 30 to 50 ml of concentrated plankton was obtained in each trawl. Samples were frozen overnight at 20°C, and thawed at room temperature without shaking (to permit for settling of larger particulate matter prior to DNA isolation).

Isolation of DNA from sediment and water samples was accomplished by using the UltraClean Soil DNA Isolation Kit (MoBio Laboratories) and the Illustra Tissue and Cells Genomic Prep Kit (GE Healthcare), respectively, according to the manufacturers’ protocols. One gram of sediment was used per isolation, and DNA was re-suspended in a final volume of 100 μl of water. For DNA isolation from water samples, 200 μl was used per preparation. After DNA isolation, the samples were re-suspended in water to a final volume of 100 μl.

Environmental PCR detection assays used the following primers: forward Hematodinium primer: (5’-CGCCTACCACTGAACTCCTC-3’); reverse Hematodinium primer: (5’-TGAACAGACGCTGAGACCAG-3’). Primer design was based upon Hematodinium sp. sequence derived from a blue crab infection (Eric Schott, personal communication, Accession # JQ815886). The forward primer anneals to a region within the 3’ end of ITS-1, and was predicted to hybridize only to clade A Hematodinium species 26, as the ITS-1 of clade B sequences available at the time we began our study showed significant divergence at this site. The reverse primer hybridizes to the junction between 5.8 S rRNA and ITS-2. One μl was used in each reaction, and PCR parameters were set at 58°C and 45 s for annealing, extension at 72  C for 30s, with amplification for 35 cycles. Electrophoresis of PCR products in 1% agarose gels was followed by visualization with UV light after staining with ethidium bromide. For sediment samples, a positive score for detection was given for amplification in either of the two replicates.

For cloning purposes amplicons were purified from gels using the Gene Clean Kit (MP Biomedicals). Isolated products were cloned using the TOPO TA system (Invitrogen), and were sequenced at the University of Maryland Center for Marine and Environmental Sciences at the Institute of Marine and Environmental Technology (UMCES@IMET).

For PCR reactions designed to detect dinoflagellate species, the following primers from Oldach et al., 2000 27 were used: universal dinoflagellate SSU forward primer (5’-CGATTGAGTGATCCGGTGAATAA-3’);universal eukaryotic SSU reverse primer (5’-TGATCCTTCTGCAGGTTCACCTAC-3’). Reaction conditions were set at 58°C and 45 s for annealing, extension at 72°C for 30 s, with amplification for 35 cycles. Products were identified, excised from gels, and cloned as described above.

The partial 18 S rRNA gene sequences obtained from the clone libraries were carefully checked for chimeric artifacts using the BLASTN program (http://www.ncbi.nlm.nih.gov/BLAST) and chimeric sequences were excluded from the phylogenetic analysis. Sequence alignment and phylogenetic reconstruction were performed using MEGA 5.05 software 28.