LANGRETH ET AL-LElSHMANfA

TROPICA WITHIN HUMAN MACROPHAGES

of pentamidine and amphotericin B in cases of apparent antimony-resistance, and with the demonstration that Pentostarnoresistant organisms were susceptible to the other five drugs in
vitro (2). If the three experimental agents prove to be efficacious
in humans, they may also be useful in antimony-resistant cases.
LITERATURE CITED
I . Berman, J. D. 1981. Activity of imidazoles against Leishmania
tropica in human macrophage cultures. Am. J. Trop. Med. Hyg., 30:
566-569.
2. - 1982. I n vitro susceptibility of antimony-resistant Leishmania to alternative drugs. J. Infect. Dis., 145: 279.
3. Berman, J. D., Dwyer, D. M. & Wyler, D. J. 1979. Multiplication
of Leishmania within human macrophages in vitro. Infect. Immun., 26:
375-379.
4. Berrnan, J. D. & Wyler, D. J. 1980. An in vitro model for investigation of chemotherapeutic agents in leishrnaniasis. J. Infect. Ds.,
142: 83-86.
5. Carson, D. A. & Chang, K.-P. 1981. Phosphorylation and antileishmania1 activity of Formycin B. Biochem. Biophys. Res. Commun.,
100: 1377-1 38 1.
6. Crofts, S. L. & Brazil, R. P. 1982. Effect of pentamidine isethionate on the ultrastructure and morphology of Leishmania mexicana
amazonensis in vitro. Ann. Trop. Med. Parasit., 76: 3 7 4 3 .

56 1

7. Docampo, R., Moreno, S. N. J., Turrens, J. S., Katzin, A. M.,

Gonzalez-Cappa, S . M. & Stoppani, A. 0. M. 198 I . Biochemical and
ultrastructural alteration produced by miconazole and econazole in T r y
panosoma cruzi. Mol. Biochem. Parasitol., 3: 169-180.
8. Hentzer, B. & Kobayasi, T. 1977. The ultrastructural changes of
Leishmania tropica after treatment with pentamidine. Ann. Trop. Med.
Parasit., 71: 157-166.
9. Keithly, J. & Langreth, S. 1983. Inefficacy of metronidazole in
experimental infections of Leishmania donovani. L. mexicana, and TryI
Trop.
. Med. Hyg., 32: 485496.
panosoma brucei. Am. .
10. Kinnamon, K. E., Steck, E. A,, Loizeaux, P. S., Hanson, W. L.,
Chapman, W. L. & Waits, V. B. 1978. The antileishmanial activity
of lepidines. Am. J. Trop. Med. H,yg., 27: 751-757.
1 I . Macadam, R. F. & Williamson, J. 1972. Drug effects on the
fine structure of Trypanosoma rhodesiense: diamidines. Trans. Roy.
Sac. Trop. Med. Hyg,, 66: 897-904.
12. Mahmoud, A. A. F. & Warren, K. S. 1977. Algorithms in the
diagnosis and treatment of exotic diseases. XXIV. Leishmaniasis. J.
Infect. Dis., 136: 160-168.
13. Pearson, R. D., Romito, R., Symes, P. H. & Harcus, J . L. 198 1.
Interaction of Leishmania donovani promastigotes with human monocyte-derived macrophages: parasite entry, intracellular survival, and
multiplication. Infect. Immun., 32: 1244-1253.

Received 24 X I 82; accepted 1 I V 83

J. Protozoo/., 30(3), 1983, pp. 561-564

0 1983 by the Society of Protozoologists

Myxosoma cerebralis (Myxozoa: Myxosporea) Etiologic Agent of

Salmonid Whirling Disease Requires Tubificid Worm
(Annelida: Oligochaeta) in its Life Cycle
MARIA E. MARKIW and KEN WOLF1
U S . Fish and WiIdliJe Service, National Fish Health Research Laboratory, Kearneysville, West Virginia 25430
ABSTRACT. Studies of the life cycle of Myxosoma cerebralis showed that development of infectivity did not occur endogenously
but that the spore aging process required participation of an aquatic tubificid oligochaete. Data suggestive of such involvement were
derived from trials in which spores were aged in an array of inert, sterilized, pasteurized, or natural aquatic substrates and from
examination of aquatic soils from trout hatcheries in which whirling disease was epizootic. The role of the aquatic oligochaete was
confirmed two ways. First, signs of whirling disease developed, and M . cerebralis spores were produced in young rainbow trout (Salmo
gairdneri) that had been fed oligochaetes harvested from pond soil taken from two hatcheries where whirling disease was epizootic.
Second, when containers of pasteurized soil were populated with four genera of oligochaetes-Aeolosoma, Dero, Stylaria, or Tubtfexfrom a biological supply house, or with tubificid worms from trout hatcheries free of whirling disease, and then seeded with M. cerebralis
spores and aged for 4 months, whirling disease occurred only in trout held with TubifpX and with hatchery tubificids.

XOSOMA cerebralis (Hofer), the etiologic agent of

whirling disease (WD) of trout and salmon, has been
known since the early 1900s; but notwithstanding considerable
research during the intervening years, the life cycle has remained
largely unknown. As is the case with other myxosporeans, information on the life cycle of M. cerebralis is essentially confined
to development of the parasite within the fish host. Not even
the size and shape of the infective stage are known, and all that
happens outside the fish-fully half the life cycle duration- has
been an enigma.
Unlike many other protozoal diseases of fish, WD is not transmitted directly from fish to fish; in fact, no one has been able
to infect fish with freshly isolated spores (2). Instead, spores
must be aged in mud o r water for several months in order to
yield infectivity or become infectious. A little known fact is that
the discovery of the critically essential but uncharacterized
aging process should be credited to A. V. Uspenskaya. Her
discovery was made during 1954 and 1955, but use of the aging
Authorship is alphabetical.

procedure elsewhere occurred only after she described her method in personal communications to parasitologists during the
1960s. For the record, some details of the initial discovery were
reported by Uspenskaya in 1978 (6). The experimental aging
process was routinely used by others after Hoffman & Putz (3)
published in English and included in their report the fact that
cut-up heads of infected fish, added to tanks containing mud as
well as being supplied with running water, produced infectivity.
The literature on WD and M. cerebralis is voluminous, and
because the infection is an important problem in fish husbandry,
the condition is usually discussed in standard texts on fish diseases. The most detailed coverage of the disease, however, and
the parasite is in the review by Halliday (l), who included nearly
150 references.
Relevant to the new work reported here, our long range plans
were based on research needs and voids in knowledge that were
discussed in a 1974 overview of WD (7). Our beginning efforts
went into the development of a method of releasing, concentrating, and purifying spores from infected tissues. That procedure (4) provided a highly sensitive method of spore detection

562

J. PROTOZOOL., VOL. 30, NO. 3, AUGUST 1983

and, additionally, the antigens needed for preparation of antiserum. Rabbits were immunized and the resulting antiserum
was conjugated with fluorescein isothiocyanate for serological
identifications with direct fluorescent antibody technique (5).
Applications were made in diagnostics, and the methods were
used in searching for stages in the life cycle.
The purpose of the present paper is to document our finding
that a tubificid oligochaete plays an essential role during the
aging of spores and development of infectivity in M. cerebralis.

depopulated of invertebrates and vegetative microorganisms by

pasteurization. Also, well water was used in only one container,
and all containers but one (kept under nitrogen) were kept under
normal atmosphere. The positive control consisted of a 180liter fiberglass tank that held untreated spring soil, was provided
with a small flow of spring water, and was seeded with crushed
cartilage from WD-infected trout. Density of spores seeded was
7 X lo6 per glass container, and those added to a glass wool
substrate were disrupted in a tissue homogenizer to the extent
of 50-60% as a means of freeing the sporoplasm. At the start,
each container was stocked with several planarians, snails, and
MATERIALS AND METHODS
freshwater shrimps. The containers were held at 12.5-14.5C
Fish. Throughout our studies we used pathogen-free fry or for 4 months and then assayed for infectivity. Postulating possmall fingerling rainbow trout (Sulmogairdneri) as test animals sible waterborne infectivity at the time of assay, flowing spring
for assays of infectivity. Except where noted, all experiments water was supplied to each container, and 15 test fry were held
with test fish were conducted at 12.SoC, and fish were fed com- in the effluent for 2 weeks. Thereafter, other fry were placed
mercial trout rations.
directly in the containers in contact with the substrates, held
Spores. Spores used to initiate all experiments were obtained there for 77 days, and examined for spores 4 months after first
in part from naturally infected rainbow trout or brook trout exposure.
(Safvefinusfontinalis), but predominantly from small rainbow
Determination of the size of infective forms. We investigated
trout that had been infected experimentally by holding them in water or aquatic soils, or both, from two Pennsylvania trout
tanks containing pond or stream soils from trout hatcheries hatcheries (here identified as P1 and P2) having earthen ponds
where WD was epizootic. Alternatively, trout were infected by and raceways in which W D was epizootic.
holding them in tanks containing soil from a pond or a hatchery
Eighty liters of pond water were transported from the P1
effluent stream that initially bore no infectivity but samples of hatchery to the National Fish Health Research Laboratory and
which had been seeded in the laboratory with M . cerebralis sequentially passed through screens with mesh sizes of 200, 100,
spores and held for 3 or 4 months before susceptible trout were 50, and 25 pm, and then by pressure filtration through membranes (293-mm dia.) with mean pore diameters of 14 and 3
exposed.
Determinations of endogenous or exogenous infectivity lfirst pm. Screens and membranes were then individually washed in
trial). A first question we asked was whether development of separate containers holding 9 liters of pathogen-free spring water,
infectivity by or from spores was an endogenous process or and each container was aerated and stocked with 60 susceptible
whether other organisms were involved. Accordingly, various trout fry. The fry were exposed for 3 days then transferred to
test substrates were set up either under well water or spring 1 .S-liter containers supplied with flowing spring water, held for
water, and additionally under one of several gas atmospheres. 4 months, and then assayed for spores.
Pond soil was also transported from the P1 hatchery, and
Well water was not chlorinated, nor did it harbor invertebrates.
The chemically similar spring water occasionally cames fresh- about 7 liters of a slurry were poured through a 4000-pm mesh
water shrimps, miscellaneous aquatic insect larvae, and infre- screen and then by gentle abrasion passed through screens of
quent small amphibians. Sand purchased from a chemical sup- 520, 200, 100, 50, and 25 pm. As in the processing of pond
ply house was used as one substrate and soil from the laboratorys water, the screens were washed in containers holding 9 liters of
pathogen-free spring as another. That soil harbors plananam, continuously aerated spring water; then the containers were
isopods, freshwater shrimps, snails, and a few aquatic macro- stocked with susceptible fry for 4 days. Containers were covered
to prevent aerosol contamination. Fry were removed to 1.5annelids.
Substrates to a depth of 5 cm were placed in each of nine liter containers supplied with flowing spring water, held for 4
7-liter capacity glass containers that were provided with flow- months, then assayed for spores.
Soil samples taken from a stream at the P2 hatchery were
through well or spring water or kept under static spring water.
Six containers had normal atmosphere; one was made anaerobic transported to the laboratory, where a slurry was made and
with the BBL Gas Pak System; another was kept under a flow serially passed through screens with mesh sizes of 4000 to 25
of nitrogen, and yet another was provided with 100% oxygen. pm; however, in lieu of the abrasion used earlier, the screens
One container of spring soil and spring water was sterilized by were gently shaken. Materials retained on the screens were samautoclaving. One container of untreated spring soil was kept pled for examination, and the great bulk was transferred to
under normal atmosphere and supplied with spring water; it containers holding 9 liters of spring water. Nine liters of the 25was seeded with spores and served as the positive control. As pm filtrate were placed in a separate container. Containers were
quantified by hemocytometer count, all containers but one were each aerated and stocked with 50 susceptible fry for 2 weeks;
seeded with 6.1 X lo6 spores of M. cerebralis; the anaerobic the fry were then transferred to 1.5-liter containers supplied
container, because it was smaller, was seeded with 1.5 X lo6 with flowing spring water, held for 4.5 months, and assayed for
spores. All containers were kept at 12.5C and, after allowing spores.
Testing the role of oligochaetes in the life cycle. Stocks of
spores to age for 4 months, water samples were collected for
concentration and examination by fluorescent antibody tech- aquatic oligochaetes, marketed as the genera Aeolosoma, Dero,
nique with rabbit anti-M. cerebralis serum (5). Sixty susceptible Stylaria, and Tubifex, were obtained from a commercial supplier (Carolina Biological) and planted in four 7-liter containers
trout fry were then placed in glass containers for 5 days-sufficient exposure time if infectivity were present-and held for containing soil from ponds and an effluent stream of the Ridge
4 months. The fish were then examined for spores by the pepsin- Hatchery (West Virginia). The hatchery had no history of WD,
and previous research had shown that the soil and its invertetrypsin-dextrose centrifugation method (4).
Determination of endogenous or exogenous infectivity (second brate fauna bore no evidence of W D infectivity (8). As an adtrial). The second trial was patterned after the first but with ditional precaution, the soil was pasteurized before the oligomodification. In lieu of sterilization, the spring soil was simply chaetes were stocked. A fifth container of pasteurized soil was

563

MARKIW & WOLF- INTERMEDIATE HOST FOR MYXOSOMA CEREBRALIS

TABLE
I Lhrnparatrve number of Myxosoma cerebralis spores that
de\dopi.d in rarnbow trout frv exposed to graded fractions of pond soil
from u junlitv harboring whirling diseuse

Screen mesh size

(urn)

TABLE
11. Occurrence of whirling diseuse in rainbow trout fry exposed
to oligochaetes with and without Myxosoma cerebralis spores.
Whirling

Genus or type of oligochaete,

SDores/fish

and sources

Accidental loss
5300
70,600
250,300
558,000
2 1,000
None

Aeolosoma. commercial
Dero, commercial
Stylaria, commercial
Tubifex, commercialNC
Tubijiex, commercial
Tubificid, L., West Virginia"
Tubificid, PennsylvaniaK
Tubificid, R., West VirginiaNC
Tubificid, R., West Virginia

Spores of
M . cerebralis

disease and
confirmation

Yes
Yes
Yes
No
Yes

4000
5 20
200
I00
50
25
25 pm filtrate

repopulated with oligochaetes from the Ridge Hatchery. Spores

of M . cerebralis were added to the five containers, which were
then supplied with a small flow of spring water. Small amounts
of granular fish food were added occasionally to sustain the
oligochaetes. Three negative controls were used-stocks of tubificids held without the addition of M . c e r e b r a k 2 The tubificids came from: (1) the commercial source (Tubifex);(2) from
Leetown, West Virginia; and (3) from Ridge, West Virginia. The
positive control consisted of tubificids from the P2 hatchery
where W D was epizootic. All containers were held at 12.5"C for
4 months before 25 susceptible rainbow trout fry were placed
in each. After an additional 3 months, five fry from each container were assayed for M. cerebralis spores, and a second assay
was carried out on fish surviving for 4 months.
RESULTS AND DISCUSSION
The first attempt to determine if endogenous or exogenous
factors were operative in producing infectivity from spores of
M . cerebralis showed that neither the positive control nor any
of the eight test environments produced infectivity. N o evidence
appeared of an endogenous process, and a modified repetition
of the work was begun. Failure of the positive control to yield
infectivity was at the time unexplained.
The second trial yielded results similar to those obtained from
the first; intact or ruptured spores in all of the inert or pasteurized
or clean spring soil substrates failed to produce infectivity. The
lack of infectivity in the positive control was again puzzling.
Assay ofthe 80-liter water sample from the P1 hatchery showed
no evidence of infectivity. In contrast, tests of the 7 liters of PI
soil slurry showed infectivity, but only in materials retained on
the 4000- and 520-pm screens, not in fractions of smallersizesthose that had been abraded.
Examination of materials sampled from the 4000- and 520pm screens showed a few isopods, freshwater shrimps, larval
insect forms, and annelids of several sizes. Annelids were the
most numerous fauna, and when crushed and subjected to flu>

Prudencc dictates that we use the general term tubificid (from the

family Tubificidae),because stocks of aquatic oligochaetes that we collected at trout hatcheries and used in our studies were mixed populations
that consisted of several genera. Samples of two populations were submitted to two different consultants for identification, and a uniform
finding of multiple genera was reported. D. Kathman of E.V.S. Con-

sultants Ltd.. Marine Technology Centre, Sidney, B.C., Canada, and J.

K. Hiltunen, U S . Fish and Wildlife Service, Great Lakes Fishery Laboratory, Ann Arbor, Michigan, both identified TubiJex tub@& in the
two samples. Many of the oligochaetes were immature, however, and
could have been either 7'. fuhifex,or Ilyodrilus tempelfmi;the imrnalure
specimens o f the two cannot be distinguished. Both consultants identified Limnodrilus in the two samples. D. Kathman classified them as
immature L. hytfineisteriand Further identifed Quistadrilus rnultisetosus
as present but least abundant in both samples.

No
No
No

Yes

+++

Abbreviations: NC = negative control; PC = positive control; L =

Leetown; R = Ridge.
a

orescent antibody, some small tubificids showed focal areas of

reactivity to anti-M. cerebralis serum.
Next, living small tubificids were selected and force-fed to
small trout. The trout in time developed signs of WD and eventually produced M . cerebralis spores.
Fixed tubificids were also sectioned histologically and the
sections processed with direct fluorescent antibody technique.
Aggregates of small reactive bodies were found internally, but
we were unable to correlate the findings with conventional light
microscopy.
A similar trial with pooled samples taken from the P2 hatchery showed infectivity in the fractions retained on mesh sizes
of 520 through 25 pm; however, a gradient was evident in the
numbers of spores that were produced in the infected fish. The
greatest number of spores was produced in fish exposed to materials on the 50-pm mesh size (Table I).
Microscopic examination of fresh and preserved materials
from the 200- through 50-gm screens showed no tubificids or
recognizable fragments thereoc however, as was found in tubificids from the P1 hatchery, when histologic sections of tubificids were examined with fluorescent antibody, reactivity was
found within the body. Tubificids from the Ridge (West Virginia) trout hatchery (without WD) showed no such reactivity.
When living, intact tubificids from the P2 hatchery soil were
made available as food to susceptible fry, the fish developed
WD.
Additional features of the infectivity were obtained from investigation of the susceptibility of young trout to materials in
a tank holding pooled soil samples from the P2 hatchery. The
tank was supplied with a small but continuous flow of pathogenfree spring water and the effluent passed by gravity through 200pm mesh screen. All exposed trout contracted WD and developed spores. At another time, two cages holding young trout
were placed in the tank-one cage in contact with the soil and
the other suspended in water. Clinical signs of WD appeared
first among fish in contact with the soil, but also appeared several
days later among those held above the bottom; fish of both
groups developed M . cerebralis spores.
The results support an interpretation that infectivity is harbored in soil, but that it can also be waterborne, though to a
lesser intensity. Provisional incrimination of a tubificid being
involved and actually necessary in the life cycle of M. cerebralis
was obtained from definitive trials with four genera of oligochaetes held in pasteurized soil that had been seeded with M .
cerebralis spores. Signs of W D and development of serologically
identified M. cerebralis spores occurred only among young trout
held in the presence of annelids marketed as the genus Tubifex;
the genera Aeolosoma, Dero, and Stylaria proved incapable of

5 64

5 . PROTOZOOL., VOL. 30, NO.3, AUGUST 1983

yielding infectivity (Table 11). Whirling disease occurred in the

presence of Tubifex worms from the commercial source and in
the presence of tubificids from pond soil from the Ridge (West
Virginia) hatchery having no history of WD, but which had
been seeded with M . cerebralis spores. Three negative controls- Tuhlrex worms and tubificids from two sources free of W D
and held in the absence of spores-did not produce infectivity.
Also, the positive control -Tubificids from the P2 hatchery but
stocked in pasteurized soil-showed the expected transmission
of WD.
Although determinations of genus- Tubifex, Limnodrilus, or
Ilyodrilus- were not made, our results establish tubificid oligochaetes as an essential participant in the life cycle of M. cerehralis outside the fish host. Thus, it becomes evident that postulated life cycles involving only fish are no longer tenable.
Likewise, what has been termed aging cannot be effected if
the substrate is simply mud or water; the substrate must
harbor one or more of the aforementioned tubificid oligochaetes.
The requirement for tubificid oligochaetes helps explain the
lack of infectivity in the intended positive controls used in our
two early experiments. In those experiments we sought evidence
for endogenous or exogenous infectivity and used soil from the
laboratorys clean spring site as the principal substrate. Retrospectively, we examined that soil but were unable to find tubificid worms. Those experiments also did not include periodic
additions of fish food to the test containers-a measure we have
since learned to d o in order to maintain populations oftubificids.
Accordingly, even if some tubificid oligochaetes had been present in the original clean spring soil, there was little to sustain
them during the 4-month holding time. Tubificid worms are
abundant in organically polluted or enriched environments, such
as effluent streams or basins of fish hatcheries with their typical
load of waste products and uneaten food residues.
The need for an invertebrate intermediate is consistent with
previously known features of M. cerebralis; namely that fresh
spores are not infectious for fish and that WD occurs in problem
proportions in earthen ponds and raceways. Providing that the
water supply is free of infectivity, W D does not occur in clean
concrete raceways.
Requirement of tubificid oligochaetes in the life cycle of M .
cerebralis suggests that other myxosporeans- particularly those
that cause problems in fish husbandry (e.g., Ceratomyxa shasta
and Henneguya exilis)-could similarly require an aquatic annelid.
It is now possible to investigate in detail the events that take
place within the oligochaete. Inasmuch as WD was transmitted
by feeding oligochaetes to trout fry, the inference is that the

worms per se harbor infectivity. They might, however, also

release an infective stage, for susceptible fish kept in the water
column above an infective soil also contracted the disease;
moreover, our inability to recognize peak infectivity in the materials retained on the 50-pm screen is puzzling.
We speculate that the recognized high susceptibility of young
salmonids might not be a physiological phenomenon; instead,
it might be due to feeding habits of small fish. It seems reasonable to believe that trout fry would graze on small annelids to
a greater extent than would large fish.
The findings that tubificid worms are involved in the life cycle
of W D presents a dilemma to agencies concerned with dissemination and control of the infection. On the one hand, testing
of chemicals that are selectively lethal for annelids but nontoxic
to trout is a possibility that could lead to effective eradication
measures. On the other hand, a considerable industry is based
on harvesting tubificids for distribution and marketing as tropical fish food. Where tubificids are produced commercially at
salmonid facilities that have WD, the potential exists for accidental spread to areas not now known to have the disease. It
becomes essential that the methods used to preserve the annelids
be examined and, if indicated, that the final product be assayed
for possible infectivity.
LITERATURE CITED
I . Halliday, M. M. 1976. The biology of Myxosoma cerebralis: the
causative organism of whirling disease of salmonids. J. Fish Biol., 9:
339-342.
2. Hoffman, G. L. 1963. Whirling disease in trouts and its prevention in hatcheries. U S . Trout News, 8: 8-18.
3. Hoffman, G. L. & Putz, R. E. 1969. Host susceptibility and the
effect of aging, freezing, heat, and chemicals on spores of Myxosoma
cerebralis. Prog. Fish-Cult., 31: 35-37.
4. Markiw, M. E. &Wolf, K. 1974. Myxosorna cerebralis: isolation
and concentration from fish skeletal elements-sequential enzymatic
digestions and purification by differential centrifugation. J. Fish. Res.
Board Can., 31: 15-20.
5. - 1978. Myxosoma cerebralis: fluorescent antibody techniques for antigen recognition. J. Fish. Res. Board Can., 35: 828-832.
6. Uspenskaya, A. V. 1978. Biological peculiarities of the spore
stage of Myxosoma cerebralis (Myxosporida: Myxosomatidae). Parasitologiya, 12: 15-20. (in Russian with English summary)
7. Wolf, K. 1974. Whirling disease research-a delineation of needs.
Fish Health News, 3: 20-27.
8. Wolf, K. & Markiw, M. E. 1982. Myxosoma cerebralis: inactivation of spores by hot smoking of infected trout. Can. J. Fish. Aquat.
Sci., 39: 926-928.

Received 13 X I I 82; accepted 21 I I I 83