Histology

Fish were preserved in either 10% buffered formalin (Lab 1) or Dietrich’s fixative (Labs 2 and 3) and processed for routine histology. Transverse sections were made of the fish from Lab 1; sagittal sections were prepared of the fish from Labs 2 and 3. All sections were stained with hematoxylin and eosin. Additional tissue sections of the one infected fish from Lab 1 were stained with Lillie-Twort Gram stain (Culling 1974) and sections from several infected fish from Lab 3 were also stained with Accustain^™, (Sigma-Aldrich, St. Louis, MO, USA), a commercial Brown-Hopps stain.

rRNA Gene Sequencing

After diagnosis of the infection by histology in several fish from Lab 3, additional live fish were euthanized and skeletal muscle was examined by wet mount. Infected muscle tissues from three fish were processed for sequencing. Approximately 25 mg of muscle tissue was extracted using the QIAgen Blood and Tissue kit (QIAgen, Valencia, CA) following the manufacturers protocol for extraction from tissue with an initial proteinase K digestion at 56°C for 3 hr. Approximately 50 juvenile neon tetras with a suspected history of the infection were obtained from a private retail fish store in the Corvallis, OR area. Muscle tissue from one infected fish was frozen and processed for sequencing as above.

PCR was performed using the general microsporidian small subunit rRNA gene primers V1F (5′-CACCAGGTTGATTCTGCCTGAC-3′) and 1492R (5′-GTTACCTTGTTACGACTT-3′). Amplifications were performed on a Peltier 200 thermocycler (MJ Research, Watertown, MA, USA) with an initial denaturation at 94°C for 2 min, 35 cycles at 94°C for 1 min, 55°C for 1 min, and 68°C for 1 min with a final extension at 68°C for 7 min. PCR products were cloned into TOPO TA Cloning vectors (Invitrogen, Carlsbad, CA, USA) and sequenced in both directions using primers flanking the inserted sequence. In order to exclude concomitant Pseudoloma neurophilia infection, a reverse primer was designed using the Primer-BLAST program available online from the NCBI (http://www3.ncbi.nlm.nih.gov/). The novel primer, PleistR (5′-TCTCGCTTGTTCGCGCCTGA-3′), was used with the forward primer, V1F to perform the PCR on samples from Lab 3 using the same thermocycling conditions as described above. PCR products from these samples were sequenced directly. All DNA analyzed in the study was sequenced on an ABI Prism®3730 Genetic Analyzer with the BigDye® Terminator v. 3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, California).

Phylogenetic Analysis

The 16S rRNA gene sequence of P. hyphessobryconis was aligned with several Pleistophora sequences and other closely related genera returned by BLAST query of the GenBank database (Altschul 1990) using the Clustalw2 software (Larkin et al 2007). Pseudoloma neurophilia and Glugea anomala (Microsporidia) were selected as outgroup species. The jModelTest software (Posada & Buckley 2004) was used to determine the most likely model of sequence evolution for the dataset and phylogenetic analyses were performed using Bayesian inference as implemented in MrBayes3.1.2 (Huelsenbeck et al 2001) and maximum likelihood algorithms (Guindon et al 2009) as implemented in the PHYML webserver (http://www.atgc-montpellier.fr/phyml/). MrBayes was run using the General Time Reversible (GTR) model of nucleotide substitution with γ-distributed rate variation across sites and a proportion of invariable sites (GTR+G+I) for 1,000,000 generations. PHYML was run using the GTR model of nucleotide substitution and bootstrap support was based on 100 replicates.

The following sequences were obtained from GenBank and used for alignment and subsequent phylogenetic analysis: Pleistophora ovariae (AJ252955.1), P. mirandellae (AJ252954.1), Ovipleistophora mirandellae (AF356223.1), Heterosporis anguillarum (AF387331.1), P. mulleri (AJ438985.1), P. hippoglossoideos (AJ252953.1), Pleistophora sp. (AJ252957.1), P. typicalis (AJ252956.1), P. anguillarum (U47052.1), Pleistophora sp. 2 (AF044389.1), Trachipleistophora (AJ002605.1), Vavraia culicis (AJ252961.1), Pleistophora sp. 3 (AF044390.1), G. anomala (AF056016.1), G. stephani (AF056015.1), G. atherinae (U15987.1), Loma acerinae (AJ252951.1), Pleistophora sp. 1 (AF044394.1), and Pseudoloma neurophilia (AF322654.1).

Case History: Lab 2

A microsporidian infection consistent with P. hyphessobryconis was detected in 1 of 30 moribund fish submitted to the diagnostic laboratory of the Zebrafish International Resource Center (ZIRC), University of Oregon, Eugene, OR. The fish was a wild-type strain purchased from a commercial tropical fish vendor and was being held in quarantine at the time it became ill. The fish were hatched Sept 2008 and submitted to the diagnostic laboratory March 2009.

Case History Lab 3

The infection was initially detected in 2 of 13 moribund fish submitted to the ZIRC diagnostic service in June of 2009. The infected animals were 18-month-old fish from the CG1 strain: A clonal, homozygous strain of fish used exclusively for tissue transplantation studies and developed using AB fish obtained from a laboratory colony and a brass mutant line obtained from a pet store (Mizgireuv & Revskoy 2006).

The “founding” CG1 fish in Lab 3 (i.e., grandparents of the infected fish) were originally imported from another institution into Lab 3’s quarantine facility as surface-disinfected eggs. This quarantine facility is physically separated from the main fish holding room, with isolated recirculating water systems, ultraviolet treatment of effluent post-filtration, restricted access, and dedicated equipment. Fish are moved from quarantine into the main facility as eggs after being disinfected in 30 ppm sodium hypochlorite for 2 min and transfered to sterile water.

After this discovery, more fish in this facility were screened for the parasite. First, 10 live fish from the same stock in which the initial infection was detected were shipped live to Oregon State University for histology. All of these animals, many of which presented obvious clinical signs of infection, tested positive for the parasite by histology. An additional 13 individuals from the subsequent (F3) generation of CG1 in this facility were examined and confirmed as positive for the infection by histology. Finally, infections were detected in three more fish in August 2009, two more from the F3 generation of the CG1 strain and another fish from a separate population of WIK strain zebrafish. All three of these animals had been exposed to sublethal doses of gamma radiation to suppress the immune system before tissue transplantation (Traver et al 2004). The WIK fish in this facility were imported as described for the CGI line and had been maintained internally for many generations, with no known history of exposure to fish outside of the colony.

In October 2009, 46 AB strain fish from Lab 3 were submitted to ZIRC for histological analysis as part of the parent institution’s health sentinel monitoring program. These fish had been directly imported into the facility from ZIRC as surface disinfected eggs and reared in small groups on each of the facility’s recirculating systems. At 6 months, all of these fish were euthanized and examined by histology. A single fish from this group was positive for P. hyphessobryconis and it had been housed in the same rack as the infected CG1 fish.

Macroscopic changes and clinical signs

The one infected fish from Lab 1 showed no obvious clinical or macroscopic changes. The one infected fish from Laboratory 2 was sluggish, appeared bloated, and exhibited a white area in the integument below the dorsal fin that was obvious in the swimming fish when examined from above.

In Lab 3, large, depigmented regions in the central dorsal fin area were observed in infected fish while still swimming in tanks. Some fish also displayed spinal curvatures. Examination of infected fish with a dissecting microscope revealed multifocal to coalescing white-gray, slightly raised regions where the skin appeared depigmented (Fig 1). Removal of the skin showed that the underlying skeletal muscle was white, soft, and edematous (Fig 1). The one positive AB fish from the Lab 3 sentinel group appeared normal when euthanized.

Open in a separate window

Figure 1

Zebrafish infected with P. hyphessobryconis. Upper image shows mottled appearance with light areas (arrows) on flanks. Lower image is same fish with skin removed, exhibiting opaque regions in muscle (arrows) representing massive infection. Bar = 500 μm.

Microscopic Examinations

Wet-mount preparations of skeletal muscle from infected fish revealed numerous sporophorous vesicles, many which contained fully developed spores. Spores in wet mounts were 7 μm by 4 μm (n=20), and contained a prominent posterior vacuole measuring approximately 3 μm by 3 μm (Fig 2). Histopathologic changes were consistent from all three laboratories and thus are described together. As previously described by other investigators (Canning et al 1986, Dyková & Lom 1980, Schäperclaus 1941), intramuscular infection by P. hyphessobryconis microsporidian organisms caused disruption of skeletal muscle comprising myomeric units. However, in several of the fish examined as a part of this study, the severity of infection was much more pronounced and was associated with severe chronic inflammation. In these fish, on average, 50 to 80% of affected skeletal muscle myofibers displayed extensive liquefactive necrosis and marked expansion by the intramuscular microsporidial parasite. Skeletal muscle degeneration, denoted by loss of cross-striations as well as fragmented, hypereosinophilic myofibers and centralized nuclei, was also observed among the affected myofibers. In sections, muscle myofibers contained from 2 to 30-plus P. hyphessobryconis organisms at varying developmental stages, ranging from rounded-up multinucleated meronts to sporophorous vesicles containing sporoblast mother cells and vesicles with mature spores (Fig. 3). The affected skeletal muscle myofibers were frequently surrounded by large numbers of intermixed histiocytes and lymphocytes with fewer eosinophilic granular cells, which tracked along the endomysial connective tissue and were accompanied by marked endomysial edema that widely separated myofibers. Regenerating myofibers characterized by enhanced sarcoplasmic basophilia and nuclear rowing were frequently adjacent to many of the necrotic myofibers (data not shown). Numerous liberated mature spores within the endomysial spaces were surrounded by dense aggregates of histiocytes and lymphocytes (Fig. 3b). Phagocytized mature spores were occasionally seen within activated histiocytes.

Open in a separate window

Figure 2

Wet mount of P. hyphessobryconis spores. Note prominent posterior vacuole (arrow). Bar = 10 μm.

Open in a separate window

Figure 3

Histological sections of zebrafish infected with P. hyphessobryconis. A. Numerous sporophorous vesicles with spores and developmental stages within myocytes. Hematoxylin and eosin (H&E). Bar = 20 μm B. Spores in phagocytes associated with chronic infections and myolysis (demarked by arrows). H&E. Bar = 20 μm. C. Mixed infection with P. hyphessobryconis in muscle and xenomas of Pseudoloma neurophilia in spinal cord (arrows). Accustain Gram. Bar = 20 μm. D. Meronts (m), and developing and mature spores (sp) in sporophorous vesicles. Some fully developed spores stain deep blue (arrows). Accustain Gram. Bar = 10 μm. E. Mature spores stain deep blue to purple (arrows), developmental stages are light orange (arrow heads). Lillie Twort Gram. Bar = 10 μm. F. Numerous spores (arrows) throughout all layers of the intestine and mesenteries. Accustain Gram. Bar = 20 μm.

G. Masses of spores in phagocytes (arrows) in ovaries. Accustain Gram. Bar = 10 μm.

The parasitic infection in the single fish from Lab 1 was limited to the skeletal muscles; however in the fish from the other two laboratories, spores within phagocytes were also detected in various other organs, including the kidney interstitium, spleen, ovaries, intestine, and mesenteries. Massive numbers of spores within phagocytes were frequently observed in the connective tissues of the ovary, but never within ova (Fig. 3g). Aggregates of spores were observed in all layers of the intestine, extending through the serosal surface and into mesenteric connective tissue. Smooth and cardiac muscle were not affected in any of the fish examined. Both Gram stain methods were effective for distinguishing spores. With the Accustain method, mature spores stained deep blue while immature spores appeared to take up less of the stain (Fig. 3d). With the Lillie-Twort method, fully formed spores stained deep blue to purple (Fig 3e).

All seven of the fish from Lab 3 examined by histology exhibited a mixed infection with Pseudoloma neurophilia (Fig 3c). This microsporidium was easily distinguished from P. hyphessobryconis by its location in the central nervous system and lack of a sporophorous vesicle with a prominent wall.

Ribosomal DNA sequence

1361 bp of the small subunit rRNA gene was sequenced from P. hyphessobryconis obtained from an infected neon tetra and is available in the GenBank database under accession number GU126672. The maximum likelihood and Bayesian analyses produced phylogenetic trees with identical topology (Figure 4). The topology is also consistent with other phylogenetic analyses of fish microsporidian parasites (Moran et al 1999). P. hyphessobryconis was placed in the clade containing Ovipleistophora ovariae and Ovipleistophora mirandellae as well as Heterosporis anguillarum.

Open in a separate window

Figure 4

Phylogenetic tree of Pleistophora hyphessobryconis obtained from neon tetra and related microsporidia based on small subunit rRNA gene sequences. 1361 bp small subunit rRNA gene sequences from 20 microsporidia infecting fish were used to reconstruct phylogeny. Maximum likelihood tree is shown. Branch numbers are maximum likelihood bootstrap support based on 100 replicates/Bayesian posterior probabilities. Genus names shown are as recorded in GenBank with new genus designations in parentheses. The microsporidian parasite, Pseudoloma neurophilia was selected as an outgroup taxa.

A total of 1224 bp (position 1 to 1224) of the small subunit rRNA gene was sequenced from P. hyphessobryconis obtained from three zebrafish submitted by Lab 3. All three of these sequences were identical, with no insertions/deletions nor transitions/transversions. Comparisons of the 1224 bp subset of the sequence obtained from the neon tetra and the three sequences obtained from zebrafish showed no insertions/deletions and two transitions: One at site 149 (T to C) and at site 180 (G to A) for an overall paired distance of 0.002.

Transmission

The three facilities had no history of sharing fish and are located in different areas of the United States. Thus we presume that the infections arose in three independent instances by exposure to other species of infected aquarium fishes. Indeed, the CG1 line was actually derived from zebrafish purchased from a pet store (Mizgireuv & Revskoy 2006). The affected fish from Lab 2 had also been purchased from a pet store and the maternal parent of the infected fish at Lab 1 had been purchased from a vendor that supplied fish to the commercial pet trade. As stated above, transmission of P. hyphessobryconis, ostensibly per os, was achieved by placing large numbers of spores in water with fish, a method that has been used by others (Canning et al 1986). As with other microsporidia, it is assumed that infection is initiated by ingestion of spores. Spores released from dead fish are a likely source of infection, but could also be released from the intestines. Schäperclaus (1941) suggested that spores may be released from the skin or urinary tract of infected fish.

Zebrafish spawn frequently within aquaria, and tank mates quickly eat available eggs (Lawrence 2007, Spence et al 2008). Although spores were not found within eggs, the massive numbers seen within ovaries of some infected zebrafish suggest that infectious spores could be released during spawning and thus would be available to fish feeding on eggs or to infect the next generation of fish. Maternal transmission, including true vertical transmission within eggs, has been verified or implicated for other microsporidia of fishes (Docker et al 1997, Kent & Bishop-Stewart 2003, Phelps & Goodwin 2008). Schäperclaus (1941) suggested the possibility of maternal transmission as he found infections in 8-day-old neon tetras derived from infected parents. Once established in zebrafish, it is conceivable that the infection could be maternally transferred to the next generation. This could even occur with spores outside the egg as microsporidian spores are very resistant to chlorine (Ferguson et al 2007). This provides one explanation for the occurrence of the infection in fish derived from surface-disinfected eggs. Alternatively, these fish may have become infected by an unrecognized breach in biosecurity protocols.

The infected fish from Lab 1 was the 13-month-old offspring of a female fish that had been purchased from an outside vendor and brought into the laboratory. If this fish was infected as an embryo, this would mean that the fish was subclinically infected for more than 1 year. All of the fish purchased from the vendor had been removed from Lab 1 more than 7 months before the infected fish was detected. Even if the fish had not been infected as an embryo but instead at a later time through accidental cross-contamination with an infected fish from the group purchased from the vendor, it indicates that this fish was subclinically infected for more than 7 months.

As with other microsporidian infections, it is likely that immune status influences susceptibility of zebrafish to P. hyphessobryconis. Recently, Ramsay et al.(Ramsay et al 2009) showed that stress enhances infections of P. neurophilia in zebrafish. Infections by P. hyphessobryconis were widespread in only one strain (CG1) in one of the labs with the infection suggesting that this line may be particularly susceptible to the parasite. The microsporidium was also found in fish (CG1 and WIK) that were irradiated and thus were immune compromised. Although immune status likely affects the susceptibility and progression of the disease in zebrafish, apparently immunocompetent fish can become infected as demonstrated by reports of P. hyphessobryconis in zebrafish from aquaria (Opitz 1942, Steffens 1962) and experimental infections in presumably healthy zebrafish are reported here. The latter were exposed at 15-days old, a time at which fish have innate immunity but adaptive immunity has not fully developed (Lam et al 2004, Trede et al 2004).

This study was supported by grants from the National Institutes of Health (NIH NCRR 5R24RR017386-02 and NIH NCRR P40 RR12546-03S1). Mention of a brand name does not imply endorsement of the product by the U.S. Federal Government. Opinions, interpretations, conclusions, and recommendations stated are those of the authors and are not necessarily endorsed by the U.S. Army. We thank Dr. G. Sanders, University of Washington for manuscript review, Dr. M. Schuster for assistance in translation of German references, and K. Berkenkamp for histology slide preparation and technical support.