FOLIA PARASITOLOGICA 55: 100–104, 2008

The marine herring myxozoan auerbachi (: Ceratomyxidae) uses Chone infundibuliformis (Annelida: Polychaeta: Sabellidae) as invertebrate host

Marianne Køie1, Egil Karlsbakk2 and Are Nylund3

1Marine Biological Laboratory, University of Copenhagen, DK-3000 Helsingør, Denmark; 2Institute of Marine Research, P.O. Box 1870, Nordnes, N-5817 Bergen, Norway; 3Department of Biology, University of Bergen, P.O. Box 7800, N-5020 Bergen, Norway

Key words: Myxozoa, Ceratomyxa auerbachi, Clupea harengus, Chone infundibuliformis, actinospores, life cycle, Denmark

Abstract. Sequencing of SSU rDNA showed that actinospores of the tetractinomyxon type, which develop in Chone infundibuli- formis Krøyer (Annelida, Polychaeta, Sabellidae) from the northern Øresund, Denmark, are identical with Ceratomyxa auerbachi Kabata, 1962 (Myxozoa, Ceratomyxidae). This myxosporean was found in the gallbladder of the Atlantic herring Clupea harengus L. from the northern Øresund, Denmark, and from the Bergen area, western Norway. The pansporocysts and actinospores of C. auerba- chi are described. This is the third elucidated two-host life cycle of a marine myxozoan, and the first involving a marine ceratomyxid.

Previously undescribed actinospores from the collective template DNA and a reaction mixture consisting of 5 µl 10 × group Tetractinomyxon were found in Chone infundibuli- PCR buffer (NE Biolabs), 5 µl 10 mM dNTP, 2 µl (10 mM) of formis Krøyer (Annelida, Polychaeta, Sabellidae). Marine the reverse and forward primer, 2 U of thermostable DNA poly- actinospores of this type have never been recorded in ma- merase (Qiagen) and 34 µl dH2O. The PCR conditions were as rine sabellids before, whereas the freshwater sabellid Ma- follows: after an initial 5 min denaturation step at 95°C, samples were taken through 40 amplification cycles, each consisting of a nayunkia speciosa Leidy harbours actinospores of Cera- 30 sec denaturation step at 94ºC, a 30 sec primer annealing step tomyxa shasta (Noble, 1950) and Parvicapsula minibicor- at a temperature dependent of the primer combination used, and nis Kent, Whitaker et Dawe, 1997 (Bartholomew et al. a 1 min and 30 sec extension step at 72ºC. A prolonged exten- 1997, 2006). The novel actinosporean stages from C. in- sion step of 10 min at 72ºC completed each reaction. Usually, fundibuliformis are described in the present paper. Their the annealing temperature of the combinations MyxF2/Myx- partial SSU rDNA sequences were compared with those of gen4r, MarF1/ RosR2 and Erib1/MybiR was set to 58°C. The fish myxosporeans found in the northern Øresund, and PCR products were purified on Qia-quick PCR Purification col- found identical to those of Ceratomyxa auerbachi Kabata, umns (Qiagen) and then sequenced using the BigDye Terminator 1962 from herring. Sequencing kit. Sequencing (direct) was done using the amplifi- cation primers described above. The sequence data were assem- bled with the Vector NTI AdvanceTM 10.3 software (Invitrogen, MATERIALS AND METHODS Carlsbad, CA) and GenBank searches were done with BLAST The polychaete Chone infundibuliformis Krøyer, 1856 (An- (2.0). Alignments were done with Clustal_W (Vector NTI). nelida, Sabellidae), 4–7 cm long without the tentacle crown, was The phylogenetic position of C. auerbachi was examined dredged at 20 m in the northern Øresund, in August 2007. with an alignment of other Ceratomyxa spp. and other selected Smears of different parts of the worm were examined. Speci- members of the ‘marine clade’ (Fiala 2006) based on BLAST mens of the Atlantic herring Clupea harengus L. were dredged searches. Alignments from AlignX (Vector NTI) were manually in the Øresund, Denmark, at 25 m depths in August 2004, and edited in Genedoc; hypervariable or ambiguous regions were examined immediately upon capture. Pieces of infected C. in- deleted from the alignment to ensure comparison of homologous fundibuliformis and the gallbladder of infected herring were positions. Phylogenetic analysis on the data sets were performed fixed in 100% ethanol. DNA was extracted from two pieces of with Paup* (4.0b10) (Maximum likelihood [ML], bootstrapped infected tissue per host type (duplicate DNA samples) using the 100 replicates), TreePuzzle (5.0) (ML, quartet puzzling algo- DNeasy® Tissue Kit protocol for tissues (Qiagen, Hilden, rithm, 10,000 replicates). The Paup maximum likelihood analy- Germany). The PCR primers used were the forward primers sis used a heuristic search algorithm with 10 random sequence Erib1, MarF1 and MyxF2, and the reverse primer Myxgen4r additions and TBR branch swapping. A General time reversible (Barta et al. 1997, Kent et al. 2000, Nylund et al. 2005, Køie et gamma+ pinvar (GTR+Γ+I) model of nucleotide substitution al. 2007a) with the addition of two new reverse primers MybiR with a 6 category gamma distribution was used following the (5’AAT TTC ACC TCT CGC GGC AA) and RosR2 (5’ATC examination of the data set with the program Modeltest (Posada CTT CCG CAG GTT CAC CTA CGG). The PCR amplifica- and Crandall 1998). Phylogenetic trees were drawn using Tree- tions were performed in a total volume of 50 µl using 2 µl of View.

Address for correspondence: M. Køie, Marine Biological Laboratory, University of Copenhagen, DK-3000 Helsingør, Denmark. Phone: ++45 353 21 981; Fax: + 45 353 21 951; E-mail: [email protected]

100 Køie et al.: Life cycle of Ceratomyxa auerbachi

RESULTS Two of 25 specimens of Chone infundibuliformis were (5.4–6.3) µm in diameter (n = 20). Three identical spheri- infected with pansporocysts and actinospores (Figs. 1, 2, cal polar capsules 2.0 (1.8–2.3) µm in diameter. Polar fila- 6). Both developmental and mature stages were found free ments 13 (11–15) µm long (Fig. 5). Three valve cell nuclei in the coelomic fluid in the abdominal segments of the seen as small knobs on surface of young spores (Fig. 2). polychaete. Both infected polychaetes were mature males. Nuclei not seen in older thick-walled, slightly smaller and Most pansporocysts with eight actinospores. Most pan- often somewhat angular spores. Sinuous crests on spore sporocysts 20 (19–21) µm long and 12 (11–13) µm wide surface probably indicate suture line of valve cells (Figs. (n = 10), with a weak constriction (Figs. 1, 6). All actino- 4–6). Sporoplasm binucleate (Fig. 3). Mature spores did spores (Figs. 1–6) in same pansporocysts at same level of not change size or shape when kept for two days in sea- development. Fully developed actinospores spherical to water at 10ºC. slightly ellipsoidal, 6.7 (6.3–7.2) µm in length and 5.9

Figs. 1–5. Live actinosporean stages of Ceratomyxa auerbachi from Chone infundibuliformis. Interference contrast. Fig. 1. Pansporo- cyst with six actinospores. Arrows indicate the restriction. Fig. 2. Actinospore showing shell valve cell nuclei (nv). Fig. 3. Frontal view of actinospore with binucleate sporoplasm (ns). Fig. 4. Apical view of actinospore. Fig. 5. Actinospore with one extruded polar filament (pf). Arrowheads – crest-like structures on spore surface. All to same scale. Scale bar = 5 µm.

Ceratomyxa auerbachi Kabata, 1962 was found in two of 10 examined specimens of Clupea harengus from the northern Øresund, Denmark, and in one of six from the Bergen area and two of six from Storfjorden, Norway. Three partial 18S rDNA sequences of Ceratomyxa au- erbachi were obtained from each of two infected herring from Øresund, Denmark (isolates ‘CAug04’ EU616731 1,596 nt and ‘Cau07’ EU616730 1,733 nt), and one from Storfjorden, Norway (isolate Si2GB EU616732 1,778 nt). These did not differ over 1,596 nt aligned. Partial myxo- sporean sequences were also obtained from two pieces of a single tetractinomyxon-infected Chone infundibuliformis (isolate ‘Chon’, EU616733 1,743 nt) and one piece of a second tetractinomyxon-infected C. infundibuliformis (iso- late ‘NyChon’, EU616734 1,762 nt). These were identical over 1,743 nt aligned, and also identical to CAug04 and Cau07 (1,596 and 1,733 nt compared). Based on the com- plete identity of the Ceratomyxa auerbachi sequences from C. harengus with the tetractinomyxon sequences from C. infundibuliformis, these are considered different developmental stages of the same species. Fig. 6 shows a Fig. 6. Schematic diagram of the life cycle of Ceratomyxa auer- schematic diagram of the life cycle of Ceratomyxa auer- bachi. The polychaete Chone infundibuliformis acts as inverte- bachi. brate host and the Atlantic herring Clupea harengus acts as fish The most similar SSU rDNA sequence in GenBank to host. a – pansporocyst with mature actinospores; b – actino- C. auerbachi is that of Ceratomyxa sp. (DQ333430), a spore; c – myxospore. Not to scale. parasite of the marine fish Siganus rivulatus in Israel. Phy-

101 Leptotheca fugu AB195805

Sphaerospora truttae AM410773

Myxidium truttae AF201374

Kudoa thyrsites AY941819 other marine clade members Auerbachia pulchra DQ377703 94: 93:91 91 Coccomyxa jirilomi DQ323044

99:100 Sinuolinea phyllopteryxa DQ645952

Myxidium gadi DQ377711

Ellipsomyxa gobii AY505127

Ceratomyxa shasta AF001579 100:98

Parvicapsula pseudobranchicola AY308481

Ceratomyxa sp. ex Notacanthus DQ377699 100:100

Palliatus indecorus DQ377712

99:99 Ceratomyxa sp. ex Variola DQ333432

100:100 Ceratomyxa Tetractinomyxon ex Chone infundibuliformis EU616733 Ceratomyxa auerbachi ex Clupea harengus EU616730 53:90 Ceratomyxa labracis AF411472 clade

Ceratomyxa sp. ex Siganus DQ333430 62:96 Ceratomyxa sp. 0.1 89:92 ex Lophius DQ301510 83:- Ceratomyxa sparusaurati AF411471

Ceratomyxa sp. ex Siganus DQ333431

74:77 Ceratomyxa sp. ex Siganus DQ333429 Fig. 7. Phylogenetic position of Ceratomyxa auerbachi among other Ceratomyxa spp. and selected members of the ‘marine clade’ (Fiala 2006). Maximum likelihood (ML) distance tree, based on alignment of SSU rDNA sequences from 21 taxa with 1,267 charac- ters (692 constant), constructed using PAUP 4.0b10. Rooted with Leptotheca fugu and Sphaerospora truttae. Support values at nodes; ML bootstrap (Paup): quartet puzzling (QP) support values (SV). Topology of the ML and QP trees differed in the terminal branching of the ‘other marine group members’ and support values are not shown there. Elsewhere, only support values above 50% are shown. logenetic analyses place C. auerbachi robustly in a clade The spores found by Kabata (1962) were similar to those dominated by Ceratomyxa spp. from Mediterranean and found in Clupea harengus from the coastal area near Ber- Red Sea perciform fishes (Fig. 7). An exception is Cera- gen by Auerbach (1909). Auerbach (1909) temporarily tomyxa sp. from Lophius litulon in Japan (DQ301510), identified his specimens as Ceratomyxa sphaerulosa Thé- which in preliminary analyses showed a variable position lohan, 1892, which originally was described from elasmo- in the Ceratomyxa clade, and also shows insertions/de- branchs. Ceratomyxa orientalis (Dogiel, 1948) and Cera- letions in its SSU sequence not shared with other members tomyxa acadiensis Mavor, 1915 from C. harengus and of the clade. Ceratomyxa shasta is not a member of the White Sea herring Clupea pallasi marisalbi Berg may Ceratomyxa clade. represent misidentifications of C. auerbachi, which ap- pears to be the only species of Ceratomyxa in the gall- DISCUSSION bladder of Clupea harengus (MacKenzie 1987, Tolonen Ceratomyxa auerbachi has a broad distribution in the and Karlsbakk 2003, Rahimian 2007). Tolonen and Karls- northern hemisphere. It was described from the northern bakk (2003) found C. auerbachi in herring from the win- North Sea and the west coast of Scotland (Kabata 1962). tering areas in fjords in Northern Norway, but did not find

102 Køie et al.: Life cycle of Ceratomyxa auerbachi it in offshore oceanic nursery and feeding areas. Shulman the Øresund. It is most likely that the actinospores infect and Shulman-Albova (1953) found C. auerbachi (as C. herring in shallow water. The actinospores are unlikely to orientalis) in 7–33% of White Sea herring. Ceratomyxa infect herring in offshore oceanic areas as also verified by auerbachi was found in Pacific herring C. pallasi Valen- the negative findings of C. auerbachi in herring caught ciennes from the Pacific coast of North America (Arthur over great depths (Tolonen and Karlsbakk 2003). and Arai 1980a) and in three bays in central California Herring parasites, including myxozoans, have been sug- (Moser and Hsieh 1992). It was found in all size groups of gested as biological indicators or biological tags. How- C. harengus from the Gulf of St. Lawrence area, eastern ever, Ceratomyxa auerbachi was found unsuitable as bio- Canada (McGladdery and Burt 1985). indicator, even though it was earlier believed to have a Rahimian (2007) found that 15% of fingerling herring, direct life cycle (Arthur and Arai 1980b, McGladdery and 4–11 cm in length, from western Sweden were infected Burt 1985, MacKenzie 1987, Moser and Hsieh 1992). with C. auerbachi. The prevalence increased to 98% after With the present knowledge, Ceratomyxa auerbachi may 2 weeks in tanks supplied with filtered seawater. The be used as biological tag if C. infundibuliformis is the only prevalence remained at this level for the 24 months stay in invertebrate host, and if the geographical distribution of C. the tanks (Rahimian 2007). According to Rahimian infundibuliformis is known. (2007), these observations suggest a direct life cycle for C. As is the case with many myxosporean genera, Cera- auerbachi. However, our findings indicate that the inver- tomyxa is polyphyletic and needs revision (Fiala 2006, tebrate host Chone infundibuliformis lives close to the Holzer et al. 2007). The interpretation of the implications tanks and that the filters did not retain the small actino- from the SSU rDNA phylogenies of Ceratomyxa spp. on spores. the systematics of these myxosporeans is hampered by the The subspherical actinospores of C. auerbachi most lack of sequence information for the type species, Cera- closely resemble the actinospores found in Spirorbis spp. tomyxa arcuata Thélohan, 1892 and the related genera and Manayunkia speciosa (Køie 2002, Bartholomew et al. Leptotheca Thélohan, 1892 and Alatospora Shulman, 2006). They differ from all known tetractinomyxon acti- Kovaleva et Dubinina, 1979. Furthermore, sequence in- nospores by the external ridges. These actinospores proba- formation is only available for 4 of the more than 160 bly enter the environment through the polychaete nominal species. gonopores. They probably enter the fish host through the Even though five actinosporean stages have been found mouth or gills, as it is unlikely that the small actinospores in polychaetes in the Øresund, Denmark, only three have may attach to the external surface of a swimming herring. been paired with a fish host (Køie et al. 2004, 2007b, pre- Chone infundibuliformis has been recorded from the sent study). With the present contribution, three marine Arctic, the north Pacific, the north Atlantic, the northern myxosporean life cycles are known, all involving poly- North Sea, the Skagerrak, Kattegat, Øresund and western chaetes: Ellipsomyxa gobii Køie, 2003, a member of the Baltic (Hartmann-Schröder 1996). The distribution of ‘marine Myxidium clade’ of Fiala (2007), Gadimyxa atlan- Ceratomyxa auerbachi in herring coincides with the dis- tica Køie, Karlsbakk et Nylund, 2007, a member of the tribution of C. infundibuliformis. This may indicate that C. ‘Parvicapsula clade’ and herein C. auerbachi, a member infundibuliformis probably is the only invertebrate host, of the ‘Ceratomyxa clade’. The sabellid polychaetes are even though it may not be excluded that related poly- known to act as invertebrate hosts to three members of the chaetes may act as invertebrate hosts for Ceratomyxa au- ‘marine clade’, C. shasta, C. auerbachi, and Parvicapsula erbachi. No other sabellid polychaete in the Øresund has minibicornis (Bartholomew et al. 1997, 2006, present been found infected with actinospores. Chone infundibuli- study), and it is likely that members of this cosmopolitan formis has been recorded from shallow water down to a filter-feeding family play an important role as invertebrate depth of 3,500 m (Hartmann-Schröder 1996). The infected hosts for myxosporeans. specimens of C. infundibuliformis were found at 20 m in

REFERENCES

ARTHUR J.R., ARAI H.P. 1980a: Studies on the parasites of Pacific FOUS-JUCHELKA H. 1997: Phylogenetic relationships among herring (Clupea harengus pallasi Valenciennes): survey results. eight Eimeria species infecting domestic fowl inferred using Can. J. Zool. 58: 64–70. complete small subunit ribosomal DNA sequences. J. Parasitol. ARTHUR J.R., ARAI H.P. 1980b: Studies on the parasites of Pacific 83: 262–271. herring (Clupea harengus pallasi Valenciennes): a preliminary BARTHOLOMEW J.L., ATKINSON S.D., HALLETT S.L. 2006: In- evaluation of parasites as indicators of geographical origin for volvement of Manayunkia speciosa (Annelida: Polychaeta: Sa- spawning herring. Can. J. Zool. 58: 521–527. bellidae) in the life cycle of Parvicapsula minibicornis, a myxo- AUERBACH M. 1909: Bemerkungen über Myxosporidien. Zool. Anz. zoan parasite of Pacific salmon. J. Parasitol. 92: 742–748. 34: 65–82. BARTHOLOMEW J.L., WHIPPLE M.J., STEVENS D.G., FRYER J.L. BARTA J.R., MARTIN D.S., LIBERATOR P.A., DASHKEVICZ M., 1997: The life cycle of Ceratomyxa shasta, a myxosporean para- ANDERSON J.W., FEIGHNER S.D., ELBRECHT A., PERKINS- site of salmonids, requires a freshwater polychaete as an alterna- BARROW A., JENKINS M.C., DANFORTH H.D., RUFF M.D., PRO- tive host. J. Parasitol. 83: 859–868.

103 FIALA I. 2006: The phylogeny of (Myxozoa) based on crops (Teleostei: Gobiidae) uses Nereis spp. (Annelida: Poly- small subunit ribosomal RNA gene analysis. Int. J. Parasitol. 36: chaeta) as invertebrate host. Folia Parasitol. 51: 14–18. 1521–1534. MACKENZIE K. 1987: Relationship between the herring, Clupea HARTMANN-SCHRÖDER G. 1996: Annelida, Borstenwürmer, Poly- harengus L., and its parasites. Adv. Mar. Biol. 24: 263–319. chaeta. Die Tierwelt Deutschlands 58. Gustav Fischer Verlag, MCGLADDERY S.E., BURT M.D.B. 1985: Potential of parasites for Jena, 648 pp. use as biological indicators of migration, feeding, and spawning HOLZER A.S., WOOTTEN R., SOMMERVILLE C. 2007: The secondary behavior of Northwestern Atlantic herring (Clupea harengus). structure of the unusually long 18S ribosomal RNA of the myxo- Can. J. Fish. Aquat. Sci. 42: 1957–1968. zoan Sphaerospora truttae and structural evolutionary trends in MOSER M., HSIEH J. 1992: Biological tags for stock separation in the Myxozoa. Int. J. Parasitol. 37: 1281–1295. Pacific herring Clupea harengus pallasi in California. J. Parasi- KABATA Z. 1962: Five new species of Myxosporidia from marine tol. 78: 54–60. fishes. Parasitology 52: 177–186. NYLUND A., KARLSBAKK E., SÆTHER P.A., KOREN C., LARSEN T., KENT M.L., KHATTRA J., HEDRICK R.P., DEVLIN R.H. 2000: Tetra- NIELSEN B.D., BRØDERUD A.E., HØSTLUND C., FJELLSØY K.R., capsula renicola n. sp. (Myxozoa: Saccosporidae); the PKX LERVIK K., ROSNES L. 2005: Parvicapsula pseudobranchicola myxozoan – the cause of proliferative kidney disease of sal- (Myxosporea) in farmed salmon Salmo salar: tissue distribution, monid fishes. J. Parasitol. 86: 103–111. diagnosis and phylogeny. Dis. Aquat. Org. 63: 197–204. KØIE M. 2002: Spirorbid and serpulid polychaetes are candidates as POSADA D., CRANDALL K.A. 1998: MODELTEST: testing the invertebrate hosts for Myxozoa. Folia Parasitol. 49: 160–162. model of DNA substitution. Bioinformatics 14: 817–818. KØIE M., KARLSBAKK E., NYLUND A. 2007a: Parvicapsula bicornis RAHIMIAN H. 2007: Parasites of fingerling herring Clupea harengus n. sp. and Parvicapsula limandae n. sp. (Myxozoa, Parvicapsu- L.: ecology and fine morphology. J. Helminthol. 81: 199–217. lidae) in Pleuronectidae (Teleostei, Heterosomata) from Den- SHULMAN S.S., SHULMAN-ALBOVA R.E. 1953: [Parasites of Fishes mark. Dis. Aquat. Org. 76: 123–129. of the White Sea.] Izdat. Akad. Nauk SSSR, Moskva, 198 pp. (In KØIE M., KARLSBAKK E., NYLUND A. 2007b: A new genus Gadi- Russian.) myxa with three new species (Myxozoa, Parvicapsulidae) para- TOLONEN A., KARLSBAKK E. 2003: The parasite fauna of the Nor- sitic in marine fish (Gadidae) and the two-host life cycle of wegian spring spawning herring (Clupea harengus L.). ICES J. Gadimyxa atlantica n. sp. J. Parasitol. 93: 1459–1467. Mar. Sci. 60: 77–84. KØIE M., WHIPPS C.M., KENT M.L. 2004: Ellipsomyxa gobii (Myxo- zoa: Ceratomyxidae) in the common goby Pomatoschistus mi-

Received 15 January 2008 Accepted 28 February 2008

104