THE PREVALENCE AND BIOLOGY OF SOME ENDOSYMBIONTS OF SIPHONAPTERA
FROM DOGS AND CATS IN NORTH CENTRAL FLORIDA
CHARLES B. BEARD
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
TO MY PARENTS AND MY WIFE
I am deeply grateful for the contributions of many people, without
which the completion of this dissertation would have been impossible.
My advisory committee consisting of Drs. Jerry Butler, Don Hall, Jim
Maruniak, and Ellis Greiner provided guidance, insight, and encourage-
ment throughout my program. I am especially grateful to Dr. Butler, who
as committee chairman, was someone I could work with daily, who was
always available to answer questions, and who has been both a likable
supervisor and a friend. Dr. David Young, likewise, was very helpful,
providing frequent advice and serving as a sounding board for ideas. I
am very grateful to Mr. Jim Becnel, Mrs. Debra Akin, and Drs. Stephen
Zam, Sam Telford, Guy Palmer, Tony Barbet, Charles Courtney, Grover
Smart, and Drion Boucias whose aid was essential for the development of
various laboratory protocols and the identification of some of the
symbionts of fleas observed in this study.
I am appreciative of John Snyder and the staff of the Alachua
County Animal Shelter for their cooperation in providing dogs and cats
from which were obtained a constant supply of fleas. I also thank
Smokey Boyd for flea contributions.
Sincere thanks are due the personnel in the IFAS Electron micro-
scopy facility, namely, Drs. Henry Aldrich, Greg Erdos, Howard Berg, and
Ms. Donna Williams, whose expertise and willingness to help were essential
in the completion of this project.
I am appreciative of my talented, congenial, and illustrious co-
workers at Buildings 40 and 62, including Bruce Alexander, Jim Need,
Diana Simon, Debbie and Tommy Boyd, Margo Duncan, Ndeweso Kiwia, Ed
Wozniak, Farida Mahmood, Chad Lee, Terry Heaton-Jones (former members)
Edna Mitchell, Eric Wilson, Brooks Ferguson, Clay Smith, Eric Milstrey,
Terry Klein, Phil Lawyer, Richard Johnson, and (associate member) Al
I am grateful to the management of the now-defunct Tailgator for
providing a location and atmosphere conducive to solving obscure
questions of insect physiology, vector biology, and universal complexi-
ties in general.
I am particularly grateful for the editorial aid of Bruce Alexander
and the computer expertise of Debbie Boyd and Tom Hintz, which were
necessary to complete the manuscript.
I am forever indebted to my parents for their consistent encourage-
ment, love, advice, and support over the years that were required to
complete this degree. And finally, I am especially grateful to my wife,
Linda, for her companionship, support, and unconditional love that is
unequaled by none but the God we both serve and who brought us together.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS.............................................. ... iii
LIST OF TABLES...................................................... vii
LIST OF FIGURES.................................................... viii
1 AN INTRODUCTION TO THE ORDER SIPHONAPTERA AND TO THE BASIC
PRINCIPLES OF SYMBIOSIS....................................... 1
A Basic Description of the Siphonaptera...................... 1
Geographic Distribution....................................... 1
General Biology............................................... 3
Basic life cycle............................................ 3
The Egg................................. ......... ............ 3
Larval Development.......................................... 4
Pupation............................ .................... .. 5
Emergence ........................................... 5
Evolutionary History.......................................... 8
Host Associations ............................................ 9
Adaptations and Coevolution.................................. 10
Medical and Veterinary Importance of Fleas..................... 13
As Disease Vectors......................................... 13
Typhus ........ ........................ ................... 15
Myxomatosis. .............................................. 16
Other pathogens.......................................... 16
Ectoparasitic Pests................................. ........ 16
Impact on Man.............................................. 17
Symbiosis in Blood-Sucking Arthropods.......................... 18
Importance of Symbiosis.................................... 18
Definition of Symbiosis...................................... 19
Symbiotic Associations ......... ........................... 20
Symbiosis and Vector-Borne Diseases ......................... 21
Symbiosis Within the Siphonaptera ............................ 22
Statement of Objectives........................................ 22
2 THE PREVALENCE OF ENDOSYMBIONTS IN FLEAS FROM LOCAL DOGS AND
CATS, WITH BRIEF NOTES ON THEIR BIOLOGY ....................... 30
Materials and Methods........................................ 31
Results............... ...................................... 33
Discussion .................................................... 64
3 BIOLOGICAL CHARACTERIZATION OF A LEPTOMONAS SPECIES IN LOCAL
POPULATIONS OF THE FLEA, PULEX SIMULANS....................... 75
Introduction............. ...... ................ 75
Materials and Methods................................. 76
Strain Isolation............................................ 76
Morphologic Studies................................ .... 77
Electron Microscopy. ............ ........................... 77
Temperature Effects on Growth.............................. 77
Transmission and Host-Specificity Studies................. 78
Morphologic Studies ...................................... 79
Electron Microscopy........................................ 85
Temperature Effects on Growth............................... 92
Transmission and Host-Specificity Studies................... 92
Discussion....................... .......... .. ...... 96
4 NOLLERIA PULICIS (N. GEN., N. SP.) A MICROSPORIDIAN PARASITE
OF THE CAT FLEA CTENOCEPHALIDES FELIS........................ 102
Introduction................ ..... ............................ 102
Materials and Methods ........ .............. .................. 103
Results ................... ............................... ... 103
Discussion .................................................... 119
5 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS.................... 125
BIOGRAPHICAL SKETCH....................................... ....... 145
LIST OF TABLES
1.1. Microorganisms and metazoa reported from the Siphonaptera... 24
2.1. Numbers and sex of dissected fleas from each of the three
species recovered from dogs and cats in the Gainesville
area ................................. .................... 34
2.2. Numbers and species of dissected fleas from dog and cats in
the Gainesville area....................................... 34
2.3. Prevalence of specified endosymbionts in wild-caught fleas
from dogs and cats in the Gainesville area.................. 37
3.1. Morphological parameters of culture forms of a Leptomonas
sp. isolate from Pulex simulans.......................... .... 80
3.2. The results of experimental transmission studies with
leptomonads from Pulex simulans............................. 95
LIST OF FIGURES
2.1. Some endosymbionts observed in fleas off dogs and cats in
the Gainesville area....................................... 36
2.2. Midgut epithelial cells of Pulex simulans infected with a
nonoccluded baculovirus.................... ............... 39
2.3. Nonoccluded baculoviruses in midgut epithelial cells of
Pulex simulans............................................. 41
2.4. Rickettsia-like symbionts in Ctenocephalides felis.......... 44
2.5. Gram-negative rod-shaped bacteria in the lumen of
Ctenocephalides felis............................ ......... 46
2.6. Amoebae from ruptured malpighian tubule of Ctenocephalides
felis..................... .......... ............. ...... 49
2.7. Amoebae in malpighian tubule of Ctenocephalides felis....... 51
2.8. Flagellates of the family Trypanosomatidae.................. 54
2.9. Flagellates in hindgut of Pulex simulans.................... 56
2.10. Cephaline gregarines from Ctenocephalides felis ............ 59
2.11. Cephaline gregarines in the midgut of Ctenocephalides
felis................ .............. ........ ... ..... 61
2.12. Microsporidia in the midgut epithelium of Ctenocephalides
felis..... ........................................ ... 63
3.1. Culture forms of the Leptomonas strain, isolated from Pulex
simulans........ ............... ................... ...... .. 82
3.2. Culture forms of a Leptomonas sp. from Pulex simulans........ 84
3.3. Leptomonads in situ in the hindgut of Pulex simulans......... 87
3.4. Leptomonad flagellate..................... .................. 89
3.5. Leptomonad flagellates.................................... 91
3.6. The effects of six different temperatures on growth of the
Leptomonas strain isolated from Pulex simulans............. 94
4.1. Nolleria pulicis in midgut epithelial cells of
Ctenocephalides felis....................................... 105
4.2. Sporogonic sequence in Nolleria pulicis.................... 107
4.3. Midgut epithelial cells of Ctenocephalides felis infected
with Nolleria pulicis................................ .... 109
4.4. Sporophorous vesicle of Nolleria pulicis.................... I1
4.5. Spores of Nolleria pulicis................................ 114
4.6. Spore of Nolleria pulicis.................................. 116
4.7. Spore of Nolleria pulicis................................. 118
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THE PREVALENCE AND BIOLOGY OF SOME ENDOSYMBIONTS OF SIPHONAPTERA
FROM DOGS AND CATS IN NORTH CENTRAL FLORIDA
CHARLES B. BEARD
Chairman: Dr. Jerry F. Butler
Major Department: Entomology and Nematology
A study was conducted to determine the prevalence and biology of
endosymbionts in local populations of fleas collected from dogs and
cats. All of the fleas observed belonged to one of three species (viz.
Ctenocephalides felis, Pulex simulans, or Echidnophaga gallinacea).
Four hundred and three C. felis were examined, 194 P. simulans and 44 E.
gallinacea. The fleas were collected from 52 dogs and 51 cats. From 1
to 20 fleas were dissected from each host. One hundred and ninety-four
C. felis, 185 P. simulans, and 33 E. gallinacea were recovered from
dogs, and 209 C. felis, 9 P. simulans, and 11 E. gallinacea were
recovered from cats.
A variety of microorganisms and metazoa were observed, including a
virus, bacteria, several protozoans, nematodes, and a cestode. A non-
occluded baculovirus was seen in midgut epithelial cells of P. simulans,
rickettsia-like organisms were observed in tissues of all three flea
species, and gram-negative rod-shaped bacteria were observed in the
hindgut of C. felis and P. simulans. The protozoa observed include
amoebae which were seen in the malpighian tubules of all three flea
species, flagellate infections in the hindgut, rectum and occasionally
in the malpighian tubules and midgut of C. felis and P. simulans,
cephaline gregarines in the midgut of C. felis, and microsporidians in
midgut epithelial cells of C. felis and P. simulans. Microfilariae of
the dog heartworm Dirofilaria immitis were seen in the midgut of P.
simulans, entomophilic nematodes in the hemocoel of C. felis, and
metacestodes of the dog tapeworm Dipylidium caninum in the hemocoel of
Light and transmission electron microscopy were performed in an
effort to examine the parasite-host interaction of fleas found naturally
infected with viral, bacterial, and protozoan symbionts. A morphologi-
cal and biological characterization was provided for what appears to be
a new Leptomonas sp. isolated in this study from P. simulans. A new
genus and species were described for the microsporidian parasite
observed in C. felis. The organism was named Nolleria pulicis after
Wilhelm Noller who, in 1912, first observed microsporidia in fleas.
AN INTRODUCTION TO THE ORDER SIPHONAPTERA
AND TO THE BASIC PRINCIPLES OF SYMBIOSIS
A Basic Description of the Siphonaptera
The order Siphonaptera is a group of small, wingless ectoparasitic
insects commonly known as fleas. Under the current classification,
there are approximately 15 families, 212 genera, and 2018 species, with
an estimated total fauna of around 3000 (Lewis 1972, 1973, 1974a,b,c,
1975, Kim 1985). Fleas are highly characteristic and homogeneous as an
order, not easily confused with any other insect group. The adults are
obligate hematophagous parasites of mammals and, to a lesser degree,
birds (Traub 1985). They are laterally flattened and heavily sclero-
tized, often bearing specialized combs, spines, and setae. The larvae
are vermiform and free-living, resembling the larvae of nematoceran
Diptera (Askew 1971).
Fleas, as an order, are worldwide in distribution. A number of
monographs have been published on the fauna from different zoogeographic
regions, including the European Palaearctic (Smit 1957, 1966), the
U.S.S.R. (loff & Scanlon 1954), China (Li 1956), Japan (Sakaguti &
Jameson 1962, Sakaguti 1962), Australia (Dunnet & Mardon 1974), New
Zealand (Smit 1965), New Guinea (Holland 1969), South Africa
(Haeselbarth 1966), Madagascar (Lumaret 1962), South America (Johnson
1957), Panama (Tipton & Mendez 1966), Mexico (Barrera 1953), the western
United States (Hubbard 1947), the eastern United States (Fox 1940,
Benton 1980), Canada (Holland 1949), and Alaska (Hopla 1965).
The worldwide geographical distribution and host preferences for the
entire order have been summarized by Lewis (1972, 1973, 1974a,b,c,
1975). A few species are generally cosmopolitan in distribution such as
Ceratophyllus gallinae (Schrank), Ctenocephalides felis (Bouche), Pulex
irritans L., and Xenopsylla cheopis (Rothschild)(Lewis 1972, 1975).
Other species are ubiquitously distributed within limited geographical
regions according to the distribution of their hosts. The distribution
of many flea species falls short of their host range largely due to cli-
matic limitations of the immature stages developing independently of the
host (Askew 1971).
The three most commonly occurring flea species on dogs and cats in
Florida are C. felis, Pulex simulans Baker, and Echidnophaga gallinacea
(Westwood). As mentioned previously, C. felis is worldwide in its
occurrence. There are eight species and six subspecies in the genus
Ctenocephalides which, with only a few exceptions, are known mainly from
the Ethiopian Region (Lewis 1972). The genus Pulex is most strongly
associated with the southern Nearctic and northern Neotropical Regions.
Pulex irritans is the only species in this genus which is also found in
the Old World, with the exception of P. simulans in Hawaii (Hopla 1980).
Echidnophaga gallinacea is distributed practically worldwide throughout
the temperate and subtropical climates. Ten species in this genus are
known from the Australian Region. Five species are limited to the
Palaearctic Region, and the remaining five species occur in the
Ethiopian Region (Lewis 1972).
Basic life cycle
Fleas are holometabolic insects. The eggs are usually laid on the
host and fall to the ground in areas frequented by the host, such as
nests, burrows, and resting places. In these locations the larval fleas
develop, feeding on a mixed diet of organic material and dried semidi-
gested blood expelled from feeding adults in the form of fecal pellets.
In most species, there are two larval molts. The third-instar larva
pupates within a silken cocoon which it spins from salivary gland
secretions (Askew 1971). In C. felis, the eggs take about 48 hrs to
hatch. Larval development requires approximately 7 to 10 days. Adult
emergence occurs in another 7 to 10 days; so that the entire life cycle,
from egg to adult, can be completed in approximately 18 days, at optimum
temperature, which is around 300C (Harwood & James 1979, Akin 1984).
Flea eggs are oval to round in shape and are approximately 0.5 to 2
mm in length which is relatively large with respect to adult size, a
factor thought to be of nutritional significance (Pausch & Fraenkel
1966, Benton 1982). The rate of embryogenesis is greatly affected by
temperature. Silverman et al. (1981) found that C. felis eggs incubated
at 32 to 35C hatched sooner (approximately 1.5 days for 50% hatch) than
eggs incubated at 210C (approximately 3.0 days for 50% hatch). Eggs
held at 130C required 6.0 days for 50% hatch, and eggs held at 100C
required 12 days.
Flea larvae are generally considered to be scavengers, feeding as
previously mentioned on adult fecal pellets and organic debris. They do
not have eyes; however, they respond to light by means of localized
dermal photosensitive receptors. Flea larvae have been reported to be
negatively phototropic, positively geotropic, and positively thigmo-
tropic (Marshall 1981, Sgonina 1935).
As with the eggs, flea larval development is also affected by tem-
perature. Only 35% of the larvae held at 350C in the study by Silverman
et al. (1981) formed pupal cocoons, none of which hatched; approximately
80% of the larvae held at 320C pupated with approximately 70% of these
hatching; and 88% of the larvae held at 210C formed cocoons with approx-
imately 85% of these hatching. The time required for development from
egg to adult, however, increased as the temperature decreased as indi-
cated by a mean development time of around 16 days at 320C and around 40
days at 210C. At 130C, pupation did not occur until at least day 26,
and the mean development time was around 130 days. At 100C, none of the
larvae survived beyond 10 days.
While temperature has a great effect on the rate of larval develop-
ment, humidity is the most critical abiotic factor on larval survival.
Silverman et al. (1981) found that larvae of C. felis could not survive
below 50% RH at any temperature. At 27C, larval development required
10 days at 50% RH and only 5 days at 90% RH. Humidity was also shown to
have an effect on resultant adult size. The body length of adult fleas
was significantly increased as the humidity was increased from 50% to
90% RH. A humidity of greater than 70% RH was found necessary for rear-
ing most species of North American rodent fleas. Low humidities may
decrease longevity, reduce larval activity, hinder cocoon formation, and
affect adult emergence (Marshall 1981).
Prior to pupation, the third-instar larva empties its alimentary
canal and spins a silken cocoon. Small particles of sand, dust, and
other debris adhere to the cocoon. Within the cocoon, the larva
develops into a U-shaped larval prepupa which molts to a true exarate
pupa, which then molts to a pharate adult. Inside the cocoon, the pupa
is protected from adverse environmental factors such as temperature,
humidity, predators, and pesticides (Marshall 1981). The lowered sus-
ceptibility to extremes of temperature and humidity, however, has been
attributed more to the lowered metabolic rate of the quiescent pharate
adult than to any physical protection afforded by the cocoon (Silverman
& Rust 1985).
Emergence of the pharate adult from the cocoon can be stimulated by
different environmental cues which indicate the presence of a favorable
host. Many fleas emerge in response to temperature, air currents, CO2,
or vibrations (Askew 1971, Marshall 1981). In the absence of such
stimuli, some flea species have been reported to remain quiescent for up
to 450 days (Bacot 1914). Emergence can occur in synchrony, following a
particular stimulation, or may be staggered over a period of time.
Females are generally larger and emerge before males (Marshall 1981).
Host-finding is a critical step in completion of the flea life
cycle, in that for the majority of species, mating can occur only after
feeding has begun (Marshall 1981). Host-finding is highly developed in
the Siphonaptera. In one study, 270 marked Spilopsyllus cuniculi (Dale)
were released in a 2000 square yard field. Three rabbits were released
into the field, and within a few days 45% of the marked fleas were
recovered from the rabbits (Rothschild 1965).
Fleas utilize a broad range of environmental cues for host-finding
including gravity, light, vibrations, noise, temperature gradients, and
odors and other chemical stimuli (Rothschild 1965). Ceratophyllus styx
jordani Smit overwinters as a pharate adult in old underground nests of
sand martins. Adult emergence is triggered by the warm spring weather
which also signals the return of the sand martin (Bates 1962). The
emergent fleas congregate at the entrance to the burrow. They are stim-
ulated to jump by the change in light intensity which indicates that a
host is hovering over the burrow (Humphries 1969). This behavior can be
artificially stimulated by dangling a mechanically flapping bird over
the hole (Bates 1962, Rothschild 1965). In the cat flea C. felis the
warm emanations of CO2 exhaled by the cat stimulate jumping. The
swallow fleas, Ceratophyllus riparius (Jordan & Rothschild) and C. idius
(Jordan & Rothschild) respond to air currents which indicate an
approaching host. Host odors have also been shown to be important cues
for fleas. Vaughan and Meade-Briggs (1970) showed that host urine is an
attractant for the rabbit flea S. cuniculi, and Rothschild (1965)
reported that the pungent odor of its host is an attractant for the rat
Both male and female fleas are obligate blood-feeders, and once a
suitable host is located, feeding is triggered still by other specific
cues. Galun (1966) discussed the role of chemoreceptors and osmorecep-
tors in the feeding process in Xenopsylla cheopis (Rothschild). The
nucleotide phosphate adenosine triphosophate (ATP) is the single most
important phagostimulant. ATP receptors and osmoreceptors sensitive to
the tonicity of the feeding solution are apparently located in the
cibarial region of fleas and provide a crude but effective sense of
Mating on the host occurs following the initiation of feeding. The
attraction of males to females has been shown to be a temperature-
related phenomenon in Nosopsyllus fasciatus (Bosc) (Iqbal 1975). Both
sexes normally require a blood meal before mating will occur; however,
mating can be stimulated if they are warmed to 30 to 350C. In N.
fasciatus, S. cuniculi, Xenopsylla astia Rothschild, and X. cheopis a
plug of cells occludes the lumen of the epididymis, blocking the testes
and inhibiting fertilization (Rothschild et al. 1970, Marshall 1981).
Temperature alone or some other physiological stimulus related to blood-
feeding apparently triggers the release of a pheromone by the female
which is perceived by the male. Resorption of the testicular plug and
copulation is apparently triggered in the male by temperature, the
female pheromone, or some unknown stimulus provided by the blood meal
(Iqbal 1975, Marshall 1981).
The Siphonaptera are a highly evolved order of insects. They are a
monophyletic group, thought to have evolved from a mecopteran-like
ancestor (Tillyard 1935, Hinton 1958, Smit 1972, Traub 1980a, Traub
1985). The similarities lie in shared larval features, ultrastructure
of the spermatozoa, the process of resilin secretion in the pleural arch
of fleas as it compares to the same process in the wing base of certain
mecopterans, and the morphology of proventricular spines in adults.
Both orders display multiple sex chromosomes, sexual dimorphism of the
nerve cord, examples of both panoistic and polytrophic ovaries, and a
variety of exoskeletal similarities (Rothschild 1975, Schlein 1980).
The Boreus-like ancestor is thought to have had detritus-feeding larvae
and adults that fed on plant material or other insects. They presumably
became associated with the nests of mammals and then with the mammals
themselves (Marshall 1981).
Except for a single species, Palaeopsylla klebsiana Dampf, preserved
in Baltic Amber and dated approximately 50-60 million years old, fleas
are completely absent from the fossil record (Marshall 1981). It is
speculated, however, that they had their beginning approximately 100-140
million years ago with the mammals. Flea associations with marsupials
probably date back to the Cretaceous period, and it is thought that they
must have accompanied marsupials from South America across Antarctica to
Australia over 100 million years ago. The theory that fleas coevolved
with the mammals is supported on the basis of the, host relations,
physiology, historical biogeography, and morphology of fleas (Traub
1980a, 1980b, 1985, Marshall 1981).
With regards to host records, approximately 74% of the known fleas
are parasites of rodents, 8% from the Insectivora, 5% from the
Marsupialia, 5% Chiroptera, 3% Lagomorpha, 3% Carnivora, and less than
1% each in Monotremata, Edentata, Philodota, Hyracoidea and
Artiodactyla. Avian hosts, in particular seabirds and passerines,
accommodate around 6% of the total number of flea species (Marshall
1981). Though there are many exceptions, mammals that live in nests,
holes, dens, and caves during at least a portion of their life cycle are
more likely hosts (Holland 1964).
Specific fleas do not appear to be associated with elephants,
hippopotomi, aardvarks, flying lemurs, primates (excluding tupaiids),
and marine, aquatic, or semiaquatic mammals including beavers, otters,
muskrats, sirenians, and whales. The aquatic environment is not thought
to be a favorable habitat for fleas; however, there are certain species
associated with penguins, puffins, and other sea birds, some of which
live in trapped air pockets beneath the plumage (Holland 1964, Traub
1983 and 1985, Haddow et al. 1983). Grazing or browsing mammals which
have extensive ranges and do not utilize dens for their young in general
lack their own specific fleas, though there are a few exceptions (Traub
Adaptations and Coevolution
A number of different life cycle strategies exist among fleas. Some
species are nest-burrow inhabitors, seldom traveling on the body of the
host. Many of these species have morphological similarities such as
reduced eyes, reduced thoraces, and weak legs adapted to crawling rather
than jumping (Holland 1964, Balashov 1984). An example of this type of
a strategist can be found in Ceratophyllus styx Rothschild.
Some species of fleas, such as the oriental rat flea, X. cheopis,
spend the majority of their adult life on the host, leaving only to
deposit eggs. Still other fleas live their entire adult life on the
host, some as intermittent and others as prolonged feeders (Balashov
1984). The cat flea, C. felis, is an example of the former and the
sticktight flea, E. gallinacea, and the chigoe, Tunga penetrans (L.),
are examples of the latter. The adult female of T. penetrans burrows
beneath the skin of its host and remains embedded where it feeds and
reproduces (Marshall 1981).
In some cases, the flea larvae also live in the fur or skin of the
host. In Euhoplopsyllus glacialis glacialis (Taschenberg) which is a
parasite of the Arctic hare, the eggs are laid and remain on the host
where the larvae actually live in mats of fur (Freeman & Madsen 1949).
In the genus Uropsylla the entire life cycle is completed on the host.
In Uropsylla tasmanica Rothschild, a flea which parasitizes the marsu-
pial cat, Dasyurus, the eggs are laid on the host, and the larvae hatch
and burrow in superficial tissues of the host. The adult fleas emerge
from the cocoon at a time coinciding with nesting in the host. There is
evidence that temperature, humidity, and possibly host hormonal status
all play a role in synchronizing the flea life cycle with that of its
host (Dunnet & Mardon 1974, Pearse 1981, Traub 1985). Benton et al.
(1979) give examples of several other species of fleas which have larvae
that will apparently take a bloodmeal directly from a mammalian host or
from another flea larva.
Perhaps the most finely tuned parasite-host relationship existing in
the Siphonaptera is found in S. cuniculi which parasitizes the European
hare (Meade-Briggs 1964). Coevolution has proceeded to the point that
the breeding cycle of the flea is regulated by the reproductive hormones
of the rabbit. These fleas are prolonged feeders, remaining embedded on
the host's ears for long periods of their adult life. The attached
fleas do not breed unless the rabbit becomes pregnant. Ten days prior
to birth of the young rabbits, vitellogenesis begins in the fleas, and
just a few hour before birth, the eggs are ready to be laid. At this
time, the gravid female fleas detach themselves from the host's ears and
migrate to her face. When the new rabbits are born, the gravid female
fleas migrate to the new hosts where they commence feeding voraciously.
Mating and egg-laying occurs on the nestlings. After about 12 days, the
adult fleas leave the young rabbits and return to the doe where they
remain until the rabbit becomes pregnant again. The developing progeny
fleas maturate, emerge, and occupy the young rabbits.
The physiological mechanisms at work begin when the male and female
rabbits mate. The temperature in the ears of the rabbits increases
dramatically, causing the fleas to move actively between the mating
pair. Following conception, hormonal changes occur in the doe which
stimulate the fleas to attach themselves more firmly to their host.
Flea exchange occurs naturally at a low level between rabbits, and when
fleas migrate to a pregnant rabbit from a non-pregnant one, they tend to
attach and remain on the latter host. As a result, pregnant does tend
to accumulate higher populations of fleas than do the non-pregnant does.
Ten days prior to birth, increasing levels of adrenocorticotrophic
hormone in the blood of the rabbit stimulate oogenesis in the female
fleas. Thyroxin and estradiol also have a significant role in this
process and in egg deposition. When the fleas return to the doe,
increasing levels of progestins and leutinizing hormone result in
ovarian regression and resorption of the yolk (Rothschild 1965,
Rothschild and Ford 1966).
In addition to reproductive system changes that are affected by host
hormonal status, changes occur in the gut and salivary glands which pre-
pare the flea physiologically for increased feeding activity. The
blood-feeding rate climbs and the defecation rate is increased approxi-
mately 10 times (Rothschild 1965).
This system has obvious benefits for the flea in that it assures
that flea progeny will be produced at a place and time where new hosts
are abundant for continued mating and for dispersal. The increased
level of blood feeding and defecation provide an abundance of dried
blood which is an essential dietary component for the immature fleas
The preceding examples support the theory of coevolution of fleas
and mammals (Traub 1985). Tremendous differences in the degree of host
specificity exist among the Siphonaptera, some species being ultraspeci-
fic and others more catholic (Holland 1964, Traub 1980b, 1985). Gener-
alizations about survival strategies are difficult to develop simply for
the relationship between fleas and mammals. There are fleas such as T.
penetrans which have a "highly evolved" host relationship but utilize
multiple hosts, and there are rodent fleas such as the flying squirrel
flea, Opisodasys vesperalis, which have a very "basic" life cycle but
are so ultraspecific that they only occur on a single host species. As
more fleas are described and more information gained about existing
species, new patterns of host relations are sure to emerge and new ideas
on the phylogenetic history of fleas.
Medical and Veterinary Importance of Fleas
As Disease Vectors
Plague. Fleas are important pests and vectors of disease. Histori-
cally, they are most well known as vectors of the plague bacillus,
Yersinia pestis. There have been three major pandemics of plague, which
occurred in the sixth, fourteenth, and nineteenth centuries respectively
(Service 1978). The first pandemic began in Arabia and culminated in
Egypt around 542 A.D. It is referred to as the Plague of Justinian and
is thought to have killed over 100 million people in 50 years, which was
approximately half of the entire population of the Byzantine empire
The second pandemic was the "Black Death" of the Middle Ages. It
began in Asia approximately 700 years after the end of the Justinian
Plague and was introduced into eastern Europe in 1347. Merchant sailors
en route to Italy from ports in the Black Sea apparently imported the
disease into other parts of Europe. Within two years it had spread from
Italy into France, England, central Europe, and Scandinavia. The Black
Death killed at least one fourth of the entire population of Europe. It
is still considered today to be the worst disaster ever to affect man-
kind (Rail 1985).
The third plague pandemic is thought to have originated in China
around 1870. It reached Canton and Hong Kong in serious proportions in
1894 and from there spread rapidly to other coastal cities of the world.
The first recorded outbreak of plague in the continental U.S. occurred
among Chinese residents in San Francisco in 1900 (Jellison 1959). It is
a point of controversy as to whether plague was pre-existent in the U.S.
at that time or if the epidemic was introduced through Chinese
immigrants (Rail 1985).
Plague is active today in many parts of the world. There were 10
cases reported in the U.S. in 1986 and 11 in 1987 (U.S. Department of
Health and Human Services 1987, 1988). The epidemiology of the disease
is very complex and varies between continents and ecosystems (Poland &
Barnes 1979). Three types of habitats have been described that involve
different flea to vertebrate host cycles: (1) domestic rodent habitats,
usually urban; (2) peridomestic habitats, semiurban to rural; and (3)
wild rodent or rural habitats. The urban cycle generally involves rats,
Rattus rattus and R. norvegicus, and humans, with the primary vector
being the oriental rat flea, X. cheopis. Transmission can also occur
directly between humans via respiratory droplets. Peridomestic cycles
usually result from associations between wild and domestic animals in
In wild rodent habitats, the sylvatic or campestral cycle of plague
is more independent of man's activities. Within rodent populations it
is generally spread by an interchange of fleas in nests or burrows or
occasionally while rodents are foraging. In this habitat, plague exists
in enzootic and epizootic cycles. The enzootic or maintenance plague
cycle involves a few genera of relatively resistant rodent hosts, such
as Microtus or Peromyscus spp. and fleas such as Diamanus montanus
(Baker). Plague-related mortality here is rarely observed. The epizo-
otic or amplification plague cycle involves rodent species that are
highly susceptible to disease, such as prairie dogs, Cynomys spp., and
flea vectors of the genus Opisocrostis. In this cycle, the disease
spreads quickly, is highly fatal to its hosts, and is more frequently
responsible for human outbreaks (Poland & Barnes 1979).
Typhus. Murine typhus, caused by Rickettsia mooseri (=R. typhi), is
another flea-borne disease that has been of major importance throughout
history (Zinsser 1963). The typhus cycle involves commensal rats as the
principal reservoir host and X. cheopis as the flea vector. The flea
becomes infected when it takes an infectious bloodmeal. The rickettsiae
propagate in the midgut epithelium of the flea and are excreted in the
feces, which become the vehicle of transmission to the vertebrate host
(Ito et al. 1975). Murine typhus antibodies or isolations have been
reported from a number of small mammals including members of the
Muridae, Cricetidae, Sciuridae, Leporidae, Didelphidae, and Mustelidae.
Natural infections have also been demonstrated in the following fleas:
N. fasciatus, Leptopsylla segnis (Schonherr), and E. gallinacea (Philip
1980). Additionally, experimental transmission has been achieved in C.
felis (Farhang-Azad et al. 1984) and transovarial transmission in X.
cheopis (Farhang-Azad et al. 1985). The disease is seldom fatal in rats
or in man. In the U.S. there were 48 reported cases in 1986 and 33 in
1987 (U.S. Department of Health and Human Services 1987, 1988).
Another of the typhus-group rickettsiae which has been recovered
from fleas is the epidemic typhus organism, R. prowazekii. The normal
cycle of this pathogen involves the body louse, Pediculus humanus
humanus (L.), and man as the only vertebrate reservoir (Philip 1980).
The recovery of this organism from the eastern flying squirrel,
Glaucomys volans volans, and from the flea, Orchopeas howardi (Baker),
served as the first real evidence of a true zoonotic cycle (Bozeman et
Myxomatosis. Myxomatosis is a viral disease of wild and domestic
rabbits which is transmitted by fleas (Jellison 1959). In Britain, the
rabbit flea, S. cuniculi, is the primary vector. The myxoma virus was
used successfully for the control of the European rabbit, Oryctolagus
cuniculus, which is major pest of British agriculture (Mead-Briggs &
Other pathogens. A number of other potential pathogens of man and
animals have been recovered from fleas including Francisella tularensis
(McCoy 1911), Listeria monocytogenes (Vaschenok 1980), Pasteurella
multocida (Quan et al. 1986), Salmonella enteritidis (Varela & Olarte
1946), S. typhimurium (Eskey et al. 1949), Yersinia entercolitica
(Vaschenok 1980), and Y. pseudotuberculosis (Vaschenok 1980). Fleas are
also vectors of the murine trypanosome, Trypanosoma lewisi (Minchin &
Thompson 1915). They serve as intermediate hosts for the dog tapeworm,
Dipylidium caninum (Joyeux 1920), and intermediate hosts and vectors of
the filarial nematode, Dipetalonema reconditum (Newton & Wright 1956).
Fleas are also suspected to be involved in the natural transmission
cycle of certain viral haemorrhagic diseases.
In addition to their importance as vectors and intermediate hosts of
disease-causing agents of man and animals, fleas can be a very serious
problem as ectoparasites. Dermatitis resulting from flea infestation is
the most common canine skin disorder in the southern U.S., accounting
for 15% to 35% of the total canine caseload in veterinary clinics in
Florida (Halliwell 1981). Military dermatitis due to flea infestation is
also a fairly common disorder in cats (Scott 1980, Gross et al. 1986).
In a study in Florida, C. felis accounted for over 92% of the fleas
collected from dogs and 99% of those collected from cats (Harman et al.
1987). Pulex simulans is also frequently encountered in dogs (Layne
The sticktight flea, E. gallinacea, is an important pest on poultry,
in some cases, causing significant morbidity (Harwood & James 1979).
Echidnophaga gallinacea has been reported from numerous hosts including
dogs, cats, rats, mice, rabbits, opossums, squirrels, wild carnivores,
pigs, and man, frequently occurring in large numbers on the host. Over
120 specimens were collected from a single bobcat, Lynx rufus (Layne
Tunga penetrans, the chigoe, is unique among fleas in its mode of
parasitism and is an important pest of man in the tropics. The adult
female penetrates and becomes imbedded in the skin of its host, usually
on the feet. It stays attached for a prolonged period and causes
intense itching and frequent ulceration, which is subject to secondary
infection (Jellison 1959).
Impact on Man
Diseases such as plague and typhus, though not common today, have
shaped the course of history. The growth of western civilization is
said to have been set back 200 years due to plague (Service 1978).
Because of their close relationship with mammals, fleas impact man
significantly, as vectors of disease, as intermediate hosts of para-
sites, as direct mediators of medical and veterinary disorders, and as
Symbiosis in Blood-Sucking Arthropods
Importance of Symbiosis
The knowledge of the presence of symbiotic microorganisms associated
with blood-sucking arthropods is not recent to science (Steinhaus 1947,
Buchner 1965, Richards & Brooks 1958, Puchta 1955, Baines 1956). Finely
tuned symbiont-host relationships have been well documented from
numerous groups of hematophagous arthropods including Diptera, Hemip-
tera, Mallophaga, Anoplura, Siphonaptera, and Acarina (Buchner 1965,
Dasch et al. 1984, Rehacek 1984). Buchner (1965) emphasized that the
symbiotic organism together with its insect host constitutes a single
biological unit. Natural selection of any such symbiotic relationship
is contingent on the principle that in a particular environment at a
particular time the complex organism manifests some trait or traits that
give it an advantage over each individual (Margulis 1970, 1981). Often,
survival systems are present in the complex that are lacking in individ-
uals, such as in nitrogen-fixing bacteria-plant symbioses where neither
organism alone can fix atmospheric nitrogen (Margulis 1981).
The ecology and evolution of symbiosis, in addition to being of
interest from a standpoint of general biology, is of great importance in
understanding the nature of vector competency and vector-pathogen rela-
tionships in hematophagous arthropod vectors of human disease (Brooks
1975). Much of what is known about the extent of symbiosis in
hematophagous arthropods has resulted from studies of arthropods as
vectors of vertebrate pathogens (Dasch et al. 1984, Ito & Vinson 1980,
Rehacek 1984). Mutualistic symbionts are often an important potential
source of error in studies attempting to incriminate an arthropod as the
vector of a mammalian pathogen (Ito & Vinson 1980, Smith et al. 1976).
Definition of Symbiosis
Symbiosis, though sometimes used in a broader sense, can be defined
as an intimate association between two organisms in which one organism,
the host, provides the environment wherein the other organism, the sym-
biont (or symbiote), lives and reproduces. A symbiotic relationship in
which both symbiont and host are benefited by their association is
termed mutualism, and the symbiont is referred to as a mutual. An
example of mutualism would be the case where the host provided for the
growth, reproduction, and dispersal of the microbe, and the microbe
provided some nutritional factor needed by the host. Mutualistic
symbionts are found in at least 10% of the insect species. They are
also common in ticks and mites (Dasch et al. 1984, Buchner 1965, Rehacek
1984). Great variation has been been observed in location, morphology,
and staining characteristics. These microbes have been reported both
from intracellular and extracellular locations. Intracellular organisms
are sometimes found in specialized cell types (mycetocytes) or organs
(mycetomes) (Richards & Brooks 1958). Others are found within more
general cell types such as epithelial cells and oocytes, free in the
cytoplasm or harbored in vacuoles. Extracellular organisms have been
seen primarily in the lumen of the gut (Dasch et al. 1984).
Where only the symbiont benefits from the association, at the cost
of the host, the term parasitism is used to describe the relationship.
The symbiont, in this case, is referred to as a parasite. Parasitism
frequently results in host morbidity and sometimes mortality. Parasites
account for perhaps as much as 50% of all described animal species
(Rothschild & Clay 1952). Commensalism has been used to describe a
symbiotic relationship in which neither organism profits in an obvious
way (Richards & Brooks 1958). In this dissertation, however, commen-
salism will be defined as an association wherein the symbiont (commen-
sal) benefits at no apparent cost or benefit to the host. This defi-
nition is preferred because it seems doubtful that selection would
maintain an association that is beneficial to neither organism.
Symbiosis is often used synonymously in the literature with mutualism
and less frequently with commensalism. In this dissertation, symbiosis
will be used in a more general way, and mutualism, parasitism, and
commensalism will be used to describe specific symbiotic relationships.
Two main generalizations have been made about symbionts associated
with arthropods: (1) the associations are primarily mutualistic, and
(2) the associations are most prevalent among arthropods living on
highly restricted diets (Buchner 1965, Trager 1970). The diets of
insects considered to be nutritionally restricted include (1) plant
juices, (2) stored grain products, (3) wood, and (4) blood (Buchner
1965, Trager 1970). A diet consisting exclusively of vertebrate blood
is deficient in the level of B-vitamins required for normal insect
development and reproduction. Therefore, as a general rule, arthropods
that feed exclusively on blood diets throughout their entire life cycle
are inhabited by symbiotic microorganism which provide the necessary
dietary supplements (Buchner 1965, Brooks 1964, Trager 1970). Included
in this group are cimicids, reduviids, mallophagans, anoplurans, siphon-
apterans, and some acarines. Nycterbiids, strebliids, hippoboscids, and
glossines, which feed on blood as adults and are adenothropically vivip-
arous, are included here. Arthropods that feed during immature stages
on mixed diets, such as culicids, psychodids, and tabanids, apparently
sequester the required dietary supplements, thereby allowing them to
utilize a blood diet asymbiotically as adults (Brooks 1964). It must be
noted, however, that symbiotes, involved in non-dietary processes may
nevertheless be present (Hertig 1936, Yen & Barr 1973).
Symbiosis and Vector-Borne Diseases
A discussion of insect symbiosis must also include the specialized
case in which the symbiotic microbe utilizes and even depends on two
different hosts, such as the case with many vector-borne diseases. An
insect symbiont living in a mammalian host has the additional problems
of survival in a different physiological environment, at a different
temperature, and in the presence of a highly evolved and effective
immune response. Bacteria of the family Rickettsiaceae are an inter-
esting example of such symbionts. Rickettsiae are often thought of as
mammalian pathogens which are transmitted by arthropods, primarily
ticks, mites, lice, and fleas (Weiss & Moulder 1984). They can also be
thought of, however, as arthropod symbionts which produce disease when
introduced into a mammalian host.
Symbiosis Within the Siphonaptera
The potential pathogens of fleas have been reviewed by Jenkins
(1964), Strand (1977), Castillo (1980), and Daoust (1983). When
considered collectively, these lists form a fairly complete review of
endosymbionts in general, up to the most recent citation. The endo-
symbionts of fleas have been reported from around 60 host species and
include 1 virus, 12-15 bacteria (including rickettsiae and spirochaetes,
3 fungi, approximately 23 protozoa, 10-20 nematodes, and 4 cestodes
(Table 1.1). Many of the descriptions, particularly of the protozoa,
are old and inaccurate taxonomically. Very little is known about their
prevalence, biology, relative degree of pathogenicity, and impact col-
lectively on wild populations of fleas.
Statement of Objectives
This study was designed in an effort to learn more about endosym-
biosis in local populations of fleas with a goal of providing infor-
mation for further study that would lead to the development of biologi-
cal control systems. The three primary objectives were as follows:
1. To study the variety and prevalence of microorganisms common in
the three most prevalent flea species on domestic animals in and around
the Gainesville area (viz. Ctenocephalides felis, Pulex simulans, and
2. To elucidate the life cycles of observed symbionts with atten-
tion given to reproduction, transmission, and general symbiont-host
interaction. Attempts were made to isolate and culture microorganisms
observed in infected fleas. Infections were followed throughout the
life cycle of the host. Experimental transmission was attempted when
possible and comparisons made of infected and noninfected individuals.
3. To provide accurate identifications of the endosymbionts
observed in the study and new taxonomic descriptions when possible.
Table 1.1. Microorganisms and metazoa reported from the Siphonaptera.
BACTERIA, RICKETTSIA, AND SPIROCHETES:
X. gerbilli minax
(Bull & Mules 1944)
(Yakunin et al. cited in
(Yakunin et al. cited in
(Yakunin et al. cited in
(Yakunin et al. cited in
(Yakunin et al. cited in
(Ito & Vinson 1980)
(Prince & McMahon 1946)
(Quan et al. 1986)
(Quan et al. 1986)
(Quan et al. 1986)
(Traub et al. 1980)
(Traub et al. 1980)
(Traub et al. 1980)
(Traub et al. 1980)
(Traub et al. 1980)
(Traub et al. 1980)
(Traub et al. 1980)
(Dyer et al. 1931)
(Eskey et al. 1949)
(Eskey et al. 1949)
(Golov & loff 1926)
(Wheeler & Douglas 1945)
(Wheeler & Douglas 1945)
(Kondrashkina et al. cited
in Strand 1977)
(Kondrashkina et al. cited
in Strand 1977)
(Mironov et al. cited
in Castillo 1980)
Beauveria bassiana (cont.)
(Mironov et al. cited
in Castillo 1980)
(Mironov et al. cited
in Castillo 1980)
(Mironov et al. cited
in Castillo 1980)
(Ershova et al. cited
in Castillo 1980)
(Mironov et al. cited
in Castillo 1980)
(Nel'zina et al. 1978)
(Minchin & Thomson 1915)
(Patton & Rao 1921a)
(Patton & Rao 1921a)
(Laveran & Franchini 1913)
(Chatton & Delanoe
(Chatton & Delanoe
(Tyzzer & Walker 1919)
(Laveran & Franchini 1915)
(Patton & Strickland 1909)
(Molyneux et al. 1981)
(Molyneux et al. 1981)
(Minchin & Thomson 1915)
(Linardi et al. 1981)
(Newton & Wright 1956)
(Deunff et al. 1985)
(Silverman et al. 1982)
(Laumond & Beaucournu 1978)
(Deunff & Launay 1984)
(Poinar & Nelson
(Poinar & Nelson
(Poinar & Nelson
(Poinar & Nelson 1973)
(Poinar & Nelson 1973)
(Launay & Deunff 1984)
(Joyeux 1916, 1920)
(Joyeux 1916, 1920)
(Joyeux 1916, 1920)
(Joyeux 1916, 1920)
THE PREVALENCE OF ENDOSYMBIONTS IN FLEAS FROM
LOCAL DOGS AND CATS, WITH BRIEF NOTES
ON THEIR BIOLOGY
Siphonaptera (fleas) are important pests and potential vectors of
disease in Florida (Koehler & Short 1979). Historically, they are best
known as vectors of bubonic plague (Jellison 1959). In Florida, they
are most important as the causal agents of flea-bite hypersensitivity
and as intermediate hosts of the dog tapeworm, Dipylidium caninum
(Koehler & Short 1979). Dermatitis resulting from flea infestation is
the most common canine skin disorder in the southern United States
At the present, the control of fleas on dogs and cats is almost
exclusively insecticidal (Koehler & Short 1979, Osbrink & Rust 1986). A
wide range of chemicals have been marketed for this use, including chlo-
rinated hydrocarbons, carbamates, organophosphates, pyrethroids, and
insect growth regulators (Schwinghammer et al. 1985, El-Gazzar et al.
1986a). Resistance to many of these insecticides has been documented
(Fox et al. 1968, Brown & Pal 1971, and El-Gazzar et al. 1986b).
A variety of microorganisms and metazoa has been observed in fleas
(Jenkins 1964, Strand 1977, Castillo 1980, Daoust 1983). As mentioned
previously, endosymbionts have been reported from around 60 host species
and include 1 virus, 12-15 bacteria (including rickettsiae and
spirochaetes), 3 fungi, approximately 23 protozoa, 10-20 nematodes, and
4 cestodes. Very little is known about their prevalence, biology, rela-
tive degree of pathogenicity, and collective impact on wild populations
The purpose of this study was to determine the prevalence of some of
these organisms in local populations of fleas, how they are transmitted
between fleas, and whether any of them have potential as biological con-
Materials and Methods
Fleas were collected off dogs and cats belonging to residents of
the Gainesville area or at the Gainesville Animal Shelter. Specimens
were combed directly into plastic bags or removed with forceps and
brought back to the lab where they were anesthesized with CO2 and
attached with an adhesive to the well of a glass depression slide. The
fleas were then dissected in a modified Insect Ringer's solution deve-
loped by Dr. Jerry F. Butler containing the following components:
Potassium phosphate monobasic sodium hydroxide
buffer, 0.06 M, pH 7.00 (obtained from Fisher
Scientific)...... .......................... 1000 ml
NaC1 ............................................. 7.5 g
KC1 .............................................. 0.35 g
CaC12 ......... .................... .............. 0.21 g
Dissections were carried out by the aid of a stereoscopic dissecting
scope and the various tissues observed by phase contrast microscopy.
Selected tissues were fixed for electron microscopy in a solution of
2.5% glutaraldehyde and 2.0% formaldehyde in 0.2 M sodium cacodylate
solutions, and embedded in Spurr's resin (Spurr 1969). Thin sections
buffer, postfixed in 1.0% osmium tetroxide, dehydrated in graded ethanol
were cut on a LKB Model 8800 Ultrotome III microtome and poststained in
an aqueous solution of 1.0% uranyl acetate and in Reynold's lead
citrate. The sections were viewed and micrographs taken on a Hitachi
HU-11E electron microscope.
Prevalence data were collected for endosymbionts identified in
dissected fleas. Attempts were made to culture organisms on a blood
agar medium made with the following ingredients:
BBL purified agar.............................. 10 g
Bacto beef extract (Difco Laboratories)........ 1.5 g
Neopeptone (Difco Laboratories)................ 2.5 g
NaC............................................. 2 g
D-glucose, anhydrous........................... 0.75 g
Deionized H20 ............................... 500 ml
A 1:1 mixture of defibrinated rabbit blood
(obtained from Gibco Diagnostic Laboratories)
and sterile deionized H20................... 60 ml
The first six ingredients were brought to a gentle boil, utilizing a
magnetic stirring bar. When most of the ingredients were dissolved, the
suspension was autoclaved at 15 psi for 20 min. The rabbit blood was
heat-inactivated at 56C for 30 min and diluted in the deionized water.
When the autoclaved agar had cooled to approximately 500C, the blood
mixture was added and mixed by swirling. Approximately 7 ml of agar
were dispensed into sterile plastic culture tubes which were cooled on a
slanted tube rack.
Experimental transmission of protozoan parasites to colony fleas was
attempted by feeding the larvae on a diet of flea feces collected from
animals which maintained adult flea populations positive for the organ-
ism being studied. Pathogenicity was assessed in terms of percent mor-
tality in experimentally infected fleas in comparison to that of a con-
trol group. The assays were performed in plastic disposable petri
dishes, and the control groups were given a diet of dried dog blood and
powdered brewer's yeast mixed in a 1:1 ratio by weight.
Attempts were made to establish an aposymbiotic (symbiont-free)
strain of the cat flea, Ctenocephalides felis, by rearing fleas on diets
containing antibiotics. Flea larvae were fed a 1:1 mixture of brewer's
yeast and dried blood which contained 1 mg/ml chlortetracycline. Emer-
gent adults were placed on cats which were being maintained on 10 mg/lb
tetracycline hydrochloride administered twice daily. Eggs were col-
lected and the F-I progeny reared on the larval diet described previ-
ously. Treatment was continued in this manner up to the emergent F-2
adults, which were examined by transmission electron microscopy for the
presence of rickettsial symbionts.
All of the fleas observed belonged to one of three species:
Ctenocephalides felis, Pulex simulans, or Echidnophaga gallinacea (Table
2.1). Four hundred and three C. felis were examined, 194 P. simulans,
and 44 E. gallinacea, for a total of 641 fleas. The fleas were col-
lected from 52 dogs and 51 cats (Table 2.2). From 1 to 20 fleas were
dissected from each host. One hundred and ninety-four C. felis, 185 P.
simulans, and 33 E. gallinacea were recovered from dogs; and 209 C.
felis, 9 P. simulans, and 11 E. gallinacea recovered from cats.
A range of microorganisms and metazoa was observed including a
virus, bacteria, several protozoa, nematodes, and cestodes (Table 2.3 &
Fig. 2.1). The virus particles displayed typical baculovirus morpho-
logy, rod-shaped and enclosed in an elliptical envelope (Figs. 2.2 &
2.3). The nucleocapsid measured approximately 30 x 90 nm and the
Table 2.1. Numbers and sex of dissected fleas from each of the
three species recovered from dogs and cats in the Gainesville
403 Ctenocephalides felis (323 F / 80 M)
194 Pulex simulans (160 F / 34 M)
44 Echidnophaga gallinacea (35 F / 9 M)
641 Total (518 F / 123 M)
Table 2.2. Numbers and species of dissected fleas from dog and
cats in the Gainesville area.
Flea Species: Ctenocephalides Pulex Echidnophaga
felis simulans gallinacea
Dogs (n=52) 194 185 33
Cats (n-51) 209 9 11
Total (n=103) 403 194--- -- -- -- -- -- -- -- -- --
Some endosymbionts observed in fleas off dogs and cats in
the Gainesville area. (A) amoebae in the malpighian tubules
of Ctenocephalides felis, (B) encysted gamonts of cephaline
gregarines in the midgut of C. felis, (C) sporophorous vesi-
cle of a microsporidian in midgut epithelial cells of
Ctenocephalides felis, (D) microfilaria of Dirofilaria
immitis, (E) anterior end of an entomophilic nematode from
Ctenocephalides felis, (F) metacestode of Dipylidium
a' ,a. u
I )3: ,
.. .. ...
*'*a:: : f -:~; ;
--S 0--S c
* C' ; *
ONC% 14 -
- -14-i 3o I
1 0 .
i a) ) m- a u
Sc tf-4 ( 4-i
0, a)0 1
a M O0 L )
-,4 c 1
o .-i I
CC tu CO
0)U 6 I
* (0 .-l I I
e ") I
e nt al
0 *- 0 *000 000
X C' Li-4 c 0 c c 0
C* xo r00 C~ 0 n C
(N 5-.' v O
S c- *
0 1 O 0 -
Midgut epithelial cells of Pulex simulans infected with a
nonoccluded baculovirus. (c) chromatin aggregrates, (cy)
cytoplasm, (m) mitochondria, (nN) nucleus of a noninfected
cell, (viN) nucleus of a virus-infected cell, (pointer)
individual virus particle.
: ,:,, "
uS.W'- d ;., "'-: -. .i ^"*^
~'r 1- a
,I. ~L.h: LY5~ic
Figure 2.3. Nonocclu .e baculoviruses in lig-At epithelial cells of
Pulex simulans. (A) within the nucleus, (B) and (C) budding
off the nuclear membrane into the cytoplasm, (c) chromatin,
(cy) cytoplasm, (m) nuclear membrane, (N) nucleus, (n)
nucleocapsid, (pointer) viral envelope, (arrow) virus
particle in membrane vesicle.
~I. i w*
I .- .. ,-. .
envelope, 50 x 120 nm. The particles replicate in the nuclei of midgut
epithelial cells (Fig. 2.2), bud off the nuclear membrane (Fig. 2.38,C),
and are shed into the lumen of the midgut. Virus particles observed
within the cytoplasm of the host cell were enclosed in vesicles formed
from the nuclear membrane as the particles bud out of the nucleus. The
vesicle containing the enveloped virus measured approximately 125 x 160
nm. No virus particles were observed in the act of passing from midgut
epithelial cells into the lumen, but particles observed free in the
lumen appeared to have only one of the membrane layers acquired when
budding off the nuclear membrane. The infection was seen in four fleas,
all P. simulans. No attempts have been made yet to isolate the virus,
and nothing is known about its pathogenicity to the flea host. Infected
fleas could be recognized by characteristic cytopathology discernable by
phase contrast microscopy. The midgut tissue typically appeared opaque
white with the nuclei of infected cells significantly swollen (Fig.
2.2). The chromatin of infected nuclei was typically condensed and
arranged peripherally (Fig. 2.2).
Rickettsia-like symbionts were observed in various tissues of every
flea examined, including foregut, midgut, hindgut, malpighian tubules,
and both male and female reproductive tissues (Fig. 2.4). The microbes
were pleomorphic, approximately 0.5 to 1.0 pm in length. They were par-
ticularly numerous in the ovaries and testes. The organisms could also
be seen throughout various tissues in larval and pupal fleas.
In the ovaries, rickettsial symbionts were common and generally
dispersed in the region of the germarium (Fig. 2.4A). In developing
previtellogenic oocytes, the organisms tended to cluster in the distal
pole region. In chorionated oocytes the symbionts were much less
Figure 2.4. Rickettsia-like symbionts in Ctenocephalides fells. (A)
germarium region of the ovaries, and (B) testis, (N) oocyte
nucleus, (R) rickettsia-like symbionts.
l ... "
'F o~31 'q'a .~
,. + -.
+.. + i
Figure 2.5. Gram-negative rod-shaped bacteria in the lumen of
Ctenocephalides felis. (b) bacteria, (hg) hindgut.
apparent. In the testes, they were most prevalent in the interstitial
areas surrounding the spermatozoa (Fig. 2.4B). No symbionts were
observed in the spermatozoa themselves. In developing embryos, the
microbes were generally restricted to the pole cell region.
Attempts to develop an aposymbiotic strain of C. felis through anti-
biotic selection were unsuccessful. Although the numbers of microbes
appeared to be significantly diminished, the infection was not eradi-
cated in any of the tetracycline-treated fleas examined.
Massive numbers of rod-shaped bacteria were observed dividing extra-
cellularly in the hindgut lumen of six (0.9%) of 641 fleas dissected
(Fig. 2.5). They were approximately 2 to 2.5 pm in length and appear to
be gram-negative on the basis of cell wall ultrastructure. The infec-
tion was seen in five P. simulans and one C. felis.
Amoebae were seen in the malpighian tubules of about 14% of all
fleas dissected (Figs. 2.1B & 2.6). There were 78 infections found in
C. felis, 11 in P. simulans, and one in E. gallinacea. Both trophozo-
ites and cysts were present in tubule sections (Fig. 2.7). The tropho-
zoites were irregularly shaped, usually around 6-8 pm in length. The
cysts were round to oval in shape and measured approximately 5-6 pm in
length. The trophozoites demonstrated a well-defined nucleus and
nucleolus, and abundant mitochondria. The cysts had thick laminated
cell walls, were much more electron-dense, and had abundant of cyto-
plasmic vacuoles and granules. Only one nucleus was observed in each
cyst, and it was usually weakly contrasted in thin sections. Both
trophozoites and cysts displayed numerous cytoplasmic refractile bodies.
In some infections, massive numbers of amoebae occupied host malpighian
tubules; in other infections, amoebae were very sparse.
Figure 2.6. Amoebae from ruptured malpighian tubule of Ctenocephalides
- .oI 'r II T '"
Figure 2.7. Amoebae in malpighian tubule of Ctenocephalides fells. (A)
cross-section of malpighian tubule, (B) trophozoite, (C)
cyst, (m) mitochondria, (N) nucleus, (nu) nucleolus, (r)
refractile body, (w) cyst wall, (pointer) amoebae.
AP. I .S
Flagellate infections were observed in approximately 4.5% of the P.
simulans and C. felis dissected. The parasites attached to the cuticle
of the hindgut and rectum and were occasionally seen in the midgut and
malpighian tubules (Fig. 2.8 & 2.9). The flagellates were characterized
by a prominent nucleus and nucleolus, abundant mitochondria, a rod-
shaped kinetoplast, and a single flagellum that arose from a flagellar
pocket at the base of the kinetoplast. A well-defined axial filament
could be seen within the intraflagellar space, which was often broadened
toward the distal end of the flagellum, resulting in a large intra-
flagellar space. Desmosomes were seen at attachment sites between the
flagellum and the flagellar pocket (Fig. 2.9). Massive numbers of
parasites could be observed in rosettes and swimming freely, when the
guts of infected fleas were dissected and observed by phase contrast
microscopy (Fig. 2.8A,B). The infection could be experimentally trans-
mitted by feeding feces from infected adult P. simulans fleas to larvae
of the same species. Experimental transmission could not be achieved by
feeding the same material to larvae of C. felis. Aflagellate stages of
the parasite were abundant in Giemsa-stained preparations of the infec-
tious feces. Emergence rates in experimentally infected fleas was very
similar to that seen in control fleas (Table 3.2).
About 12% of the C. felis examined were infected with cephaline
gregarines (Figs. 2.1B, 2.10, & 2.11). None of these large protozoan
parasites were observed in the other two flea species. Parasite burdens
ranged from 1 to 140 organisms in the midgut of a single flea host. In
heavy infections, these parasites occupied a great portion of the volume
of the lumen. The different stages in the life cycle observed in
Flagellates of the family Trypanosomatidae. (A) infection
of hindgut of Pulex simulans, (B) rosettes of parasites, (C)
flagellates bound to the hindgut cuticle of Pulex simulans,
(ax) axoneme, (hg) hindgut, (i) intraflagellar space, (k)
kinetoplast, (m) mitochondria, (N) nucleus, (nu) nucleolus,
Figure 2.9. Flagellates in hindgut of Pulex simulans. (ax) axoneme, (hg)
hindgut cuticle, (i) intraflagellar space, (k) kinetoplast,
(m) mitochondria, (N) nucleus, (nu) nucleolus, (p) flagellar
pocket, (pointer) hemidesmosome, (arrow) desmosome.
y I*- *?.'*'t
naturally infected fleas included trophozoites of different sizes,
encysted gamonts, gametocysts, and oocysts.
The mature trophozoites were acorn-shaped and composed of three body
regions, an anterior epimerite, a medial protomerite, and a posterior
deutomerite (Fig. 2.10A). Mature trophozoites were approximately 110 to
150 um in length and 60 to 80 um in width. They remained attached to
the midgut epithelium by means of the small collar-like epimerite, with
the majority of the body of the parasite suspended free in the lumen.
Electron microscopy of the trophozoite plasma membrane revealed ridge-
like protrusions running longitudinally down the surface of the parasite
The gametocysts were spherical, measuring approximately 150 to 170
um (Fig. 2.10B,C). Within the early gametocyst, the gamonts were
observed as paired bodies, fused together along a central axis (Fig.
2.10B). Mature gametocysts were full of lemon-shaped oocysts measuring
approximately 14 to 15 pm in length (Fig. 2.10D). In one flea, oocysts
were observed escaping from a gametocyst which was in the midgut, and
passing through the pylorous into the hindgut.
The infection was transmitted experimentally to colony fleas by
feeding the feces from infected C. felis to developing larvae. Of 25
flea larvae reared on a diet of feces from an infected adult flea popu-
lation, 21 emerged as adults and ten were infected. Small trophozoites,
approximately 45 um in length, were observed attached to the midgut epi-
thelium of freshly emerged adult fleas. In the control group, which
received a 1:1 mixture of dried dog blood and brewer's yeast, 23 of 25
fleas emerged and none were infected.
Figure 2.10. Cephaline gregarines from Ctenocephalides felis. (A)
trophozoites, (B) encysted gamonts, (C) gametocyst, (D)
*;~ -I. ~IY~
Figure 2.11. Cephaline gregarines in the midgut of Ctenocephalides
felis. (A) and (B) trophozoites, (C) high magnification of
trophozoite plasmalemma, (E) midgut epithelial cells, (N)
nucleus, (nu) nucleolus, (pointer) cuticular ridges.
- m *
Figure 2.12. Microsporidia in the midgut epithelium of Ctenocephalides
felis. (A) infected epithelial cell, (B) spores, (arrow)
sporophorous vesicle containing sporoblasts.
.; ". r .1;, A
i IP Q"
_, '.;l .iu
;..I ... ,.- .
' ;.. 4
S m ~~ ~lrrr~at~'%
About 1% of the C. felis and P. simulans dissected were found to be
infected with a microsporidium that parasitized midgut epithelial cells
(Fig. 2.12). Sporophorous vesicles containing spores which were
released from disrupted cells measured approximately 15-20 im (Figs.
2.1C & 2.12A). The spores were highly refractile and spherical in
shape, measuring approximately 2 pm in diameter (Fig. 2.12B).
A microfilarial infection was seen in one P. simulans, collected
from a dog not concurrently receiving filarial prophylaxis. The micro-
filariae were approximately 310 um in length and 6.5 pm in width with
the anterior end tapered slightly toward the head and fit the general
description of the dog heartworm, Dirofilaria immitis (Fig. 2.ID).
One C. felis was found infected with four immature entomophilic
nematodes. The largest nematode, an immature female which appeared to
be in its last larval stage, was approximately 1.47 mm in length (Fig.
2.1E). This specimen resembled members of the family Sphaerulariidae of
the order Tylenchida. The nematodes were in the hemocoel of the flea;
no obvious signs of pathogenicity were observed.
Metacestodes of the tapeworm, Dipylidium caninum were observed in
three female and one male C. felis. The immature parasites were approx-
imately 0.377 mm in length (Fig. 2.1F). No pathology was observed in
infected male fleas; however, the ovaries of the infected females were
All three of the flea species observed in this study are common
ectoparasites of dogs and cats in the southeastern United States (Morlan
1952, Layne 1971, Sanford & Hayes 1974, Benton 1980). Additional flea
species reported from these hosts in Alabama, Georgia, or Florida
include C. canis, from dogs and cats, and Cediopsylla simplex,
Leptopsylla segnis, Polygenis gwyni, and Xenopsylla cheopis from cats.
Harman et al. (1987) reported P. irritans from 20 dogs and one cat in
Gainesville, Florida. These specimens, however, were in all probability
P. simulans rather than P. irritans, the two of which have been confused
in the literature prior to the resurrection of P. simulans as a valid
species by Smit (1958). The monographs of fleas of Florida (Layne
1971), Alabama (Sanford & Hayes 1974), and the southeastern United
States (Benton 1980) are in accord on the absence of P. irritans from
Florida, Alabama, and presumably Georgia. The two species can be dis-
tinguished on the basis of the morphology of the dorsal aedeagal scle-
rites and aedeagal crochets (Smit 1958). In this study, all of the male
Pulex spp. in a sample of about 25 specimens from different hosts and
locations were P. simulans.
In each of the three species collected, females out-numbered males
by a ratio of 4:1 or greater. It is possible that this ratio is due to
a difference in longevity that exists between the sexes. It could also
be accounted for by the fact that as the females feed on a host and pro-
duce eggs, they become much larger and more noticeable than males resul-
ting in a sampling bias. It seems unlikely, however, that the dif-
ference in size alone could result in such a significant bias. Further-
more, it would seem that the more noticeable females would also be more
efficiently groomed by the host, a factor which would tend to balance
The list of pathogens known to infect arthropods of medical impor-
tance is extensive (Roberts et al. 1983). The major groups include
viruses, bacteria, fungi, protozoa, and nematodes. With the exception
of the viruses, examples of each of these major groups have been repor-
ted from fleas (Jenkins 1964, Strand 1977, Castillo 1980, Daoust 1983).
There are at least ten groups of insect pathogenic viruses (Tinsley
& Kelly 1985). Among these groups, baculoviruses have received consi-
derable attention as potential biological control agents (Kirschbaum
1985). Baculoviruses have been reported from six insect orders, i.e.
Neuroptera, Trichoptera, Lepidoptera, Diptera, Hymenoptera, and Cole-
optera (Tinsley & Kelly 1985). No insect-pathogenic viruses have been
reported from fleas in the literature. The nonoccluded baculovirus seen
in this study in specimens of P. simulans is the first record of a
baculovirus the order Siphonaptera.
Baculoviruses have been subdivided on a structural basis into
nuclear polyhedrosis viruses, granulosis viruses, and nonoccluded bacu-
loviruses (Kelly 1985). The nonoccluded baculoviruses do not form pro-
teinaceous crystaline granules or polyhedra, as are formed in the other
two groups. This characteristic is the basis for distinguishing the P.
simulans virus as a member of the nonoccluded baculovirus group. Bacu-
loviruses have been reported within the size range of 40-70 nm x 250-400
nm (Tinsley & Kelly 1985). The complete virion of the P. simulans virus
was approximately 50 x 120 nm which is in the general magnitude but
somewhat shorter in length.
Vaughn & Doughtery (1985) discuss nucleocapsid and envelope forma-
tion in nonoccluded baculoviruses. In some instances, the nuclei of
infected cells become enlarged and characterized by dense aggregates of
chromatin-like material referred to as virogenic stroma (Smith 1977).
Progeny virions apparently are produced in these regions. Several
variations on nucleocapsid formation have been reported, however,
including production of nucleocapsids from long tubules of capsid-like
material (Summers 1971) and from larger progenitor nucleocapsids which
are cleaved (Federici 1980). Observations of the P. simulans virus
suggest that a virogenic stroma is produced in the form of condensed
chromatin which is dispersed in discrete bundles around the periphery of
the nucleus (Figs 2.2 & 2.3). Newly formed virus particles can pass
through the nuclear membrane in several ways, including by a budding
process in which the virions become enclosed in membrane-bound vesicles
(Adams et al. 1977), similar to those observed in this study.
Although nothing is yet known about the gross pathogenicity of this
virus to its host, infected adult fleas show characteristic baculovirus
cytopathology of affected tissues, including the pale flaccid appearance
and swollen nuclei. In this study, viral infections were first observed
consequentially in P. simulans specimens that were being examined ultra-
structurally for other potential pathogens. Once the infection was rec-
ognized, it became practice to examine dissected fleas more closely for
signs of viral infections observable at the light microscopy level, such
as swollen nuclei and flacid appearance of the midgut epithelium, which
might have otherwise been overlooked. Consequently, viral infec-tions
in local flea populations probably occur at a higher frequency and
broader host range than reported here.
Intracellular rickettsia-like symbionts have been reported from
numerous insect species (Buchner 1965, Dasch et al. 1984). Hertig &
Wolbach (1924) observed these organisms in the reproductive tissues of
the mosquito Culex pipiens L. Described as Wolbachia pipientis (Hertig
1936), this rickettsial agent of mosquitoes has been shown to mediate a
phenomenon called cytoplasmic incompatibility (Yen & Barr 1973). In
this phenomenon, females from strains of the mosquito which are negative
for the symbiont cannot successfully mate with males from strains which
are positive. Wolbachia spp. and Wolbachia-like organisms have been
reported from other species of mosquitoes, from moths, sheep keds, sev-
eral tick species (Weiss et al. 1984), and from fleas (Akin 1984).
The rickettsia-like symbionts observed in this study are similar to
those reported by other investigators. With the exceptions of W.
persica which has been grown in chicken yolk sacs and select cell lines
(Dasch et al. 1984), and a rickettsia-like symbiont from Glossina spp.
which has been grown in a mosquito cell line (Welburn et al. 1987),
attempts to culture intracellular symbionts of insects have generally
been unsuccessful. No attempts were made to culture the rickettsia-like
symbionts observed in local populations of fleas. Originally, it was
planned to develop an aposymbiotic flea strain, perform cross-matings
between this strain and a control colony strain, and compare successful
and unsuccessful crosses by means of transmission electron microscopy.
Attempts to develop an aposymbiotic strain of C. felis by antibiotic
selection through two complete generations, involving both larval and
adult stages, failed; consequently, the function of these organisms can
only be speculated.
It is uncertain whether the bacterial infections seen in the hindgut
of six of the 641 fleas represent monoxenous insect symbionts or mam-
malian pathogens that are vectored by fleas. Bacterial skin infections
are common in dogs and cats and greatly influenced by disease states
(Muller et al. 1983). These infections can occur as complications
related to insect and parasite infestations. Dogs with dermatoses
demonstrate higher levels of Staphylococcus aureus infections and more
aerobic and gram-negative microorganisms in general. Fleas found in
this study harboring gram-negative rod-shaped bacteria were collected
from three dogs and one cat, all from different locations. Only one of
the four hosts was from a local resident; the others were seen at the
animal shelter and are consequently inaccessible for follow-up. The
owner of the one available host, a dog from northwest Gainesville, could
not recall any history of dermatoses in the dog.
Minchin (1910) described the amoeba Malpighiella refringens from the
adult malpighian tubules of the northern rat flea, Nosopsyllus
fasciatus. The cysts were reported as being ovoid to spherical in
shape, intensely refringent, having four nuclei (Brooks 1974), and capa-
ble of infecting Ctenocephalides canis (Lipa 1963). Nothing else is
apparently known about the biology, morphology, or significance of this
organism, or the presence of amoebae in other species of fleas. Lipa
(1963) reported two similar monospecific entomophilic genera,
Malpighamoeba (Prell) and Malamoeba Taylor & King, as pathogens of honey
bees and grasshoppers, respectively.
The life cycle (Evans & Elias 1970, Harry & Finlayson 1976) and
ultrastructure (Hanrahan 1975, Harry & Finlayson 1976) of Malamoeba
locustae have been documented. Morphologically, these parasites appear
to be very similar to the parasites observed in local flea populations.
The trophozoites are irregular in shape and measure approximately 5-9 pm
in diameter. The cysts are ovoid, measure approximately 7 x 12 pm, and
have a thick laminated cyst wall. They contain a single nucleus and
abundant cytoplasmic refractile bodies (Lipa 1963, Brooks 1974, Hanrahan
1975, Harry & Finlayson 1976).
The parasitic amoebae of grasshoppers and honey bees can be signi-
ficantly pathogenic to their hosts in heavy infections (Lipa 1963,
Brooks 1974). Damage to the microvilli of the malpighian tubule epi-
thelium has been cited as the most obvious cytopathology (Cantwell 1974,
Hanrahan 1975, Harry & Finlayson 1976). No damage could be detected in
the malpighian tubules of infected fleas in this study.
Trypanosomatid flagellates have been reported from approximately 17
species of fleas (Wallace 1966, Molyneux et al. 1981). The observation
of these parasites in naturally-infected Pulex simulans represents a new
host record. Although massive infections are commonly reported in which
the surface of the host hindgut, rectal ampulla, and malpighian tubules
are blanketed with parasites, there is no apparent pathology. These
observations are consistent with those of the present study of natu-
rally-infected C. felis and P. simulans. A more detailed discussion of
the biology, ultrastructure, and host-parasite interaction will be pre-
sented in the following chapter.
Gregarines are protozoan parasites commonly associated with arthro-
pods. Though they are relatively large in size, they are generally con-
sidered to be nonpathogenic commensals, presumably due to the lack of a
merogonic or asexual cycle in their host (Brooks 1974). Gregarines have
been reported from six species of fleas (Table 1.1), all belonging to
the family Ceratophyllidae.
On the basis of the general morphology of the C. felis gregarine and
what has been observed of the life cycle, it seems that this symbiont
should be placed in the family Actinocephalidae. Three species of
gregarines have been described from fleas, i.e. Actinocephalus parvus
(Wellmer 1910), Agrippina bona (Strickland 1912), and Steinina rotundata
(Ashworth & Rettie 1912), all belonging to the Actinocephalidae. The
gregarine observed in this study is very similar in size and morphology
to Steinina rotundata reported from Ceratophyllus farreni, C. gallinae,
and C. styx. The measurements of trophozoites, gamonts, and gametocysts
for the C. felis gregarines fell within the range of measurements repor-
ted for the ceratophyllid gregarines. The oocysts, however, were 14-15
jim in length for the C. felis gregarines as opposed to 11-12 pm for the
others. In both groups of gregarines, the oocysts were observed being
released from the gametocyst within the midgut, in contrast to the
gametocyst passing through the hindgut and out of the flea as is seen in
septate gregarines of other arthropod hosts.
Other investigators of gregarines have observed similar ultrastruc-
tural features to those reported here. Korn and Ruhl (1972) observed
ridge-like cuticular protrusions in the cephaline gregarines, Gregarina
polymorpha and G. cuneata, from the mealworm beetle, Tenebrio molitor,
suggesting a role in motility.
Two species of microsporidia have been described from fleas, i.e.
Nosema ctenocephali and N. pulicis (Sprague 1977b). The former species
was reported from Ctenocephalides canis, and the latter from
Archaeopsylla erinacei, C. canis, and C. felis. The microsporidia seen
in this study are very different from members of the genus Nosema. The
spores are spherical and more primi-tive, resembling those of the
chytridiopsid genera. Details concerning the ultrastructure, biology,
and pathology of this protozoan parasite will be discussed in a later
An increasing number of entomophilic nematodes are being reported
from fleas. At present, there are at least 13 described species of
nematodes recorded from approximately 27 flea hosts. The only entomo-
philic nematode observed in C. felis is Neoaplectana carpocapsae
(Silverman et al. 1982). Although the nematodes observed in this study
were all immature stages, making specific identification very difficult,
they were determined not to be of the genus Neoaplectana or even of the
order Rhabditida. One of the specimens, a sub-adult female, was tenta-
tively identified as a member of either the family Sphaerulariidae or
Allantonematidae. Members of these groups are reproductive system para-
sites of insects. Poiner & Nelson (1973) described Psyllotylenchus
viviparus as the causal agent of parasitic castration in the flea
Catallagia sculleni rutherfordi. Parasitic castration of fleas by ento-
mophilic nematodes has also been reported by Holland (1952), Smit
(1953), and Akopyan (1968).
The other symbionts observed in fleas from local dogs and cats,
microfilaria of Dirofilaria immitis and metacestodes of Dipylidium
caninum, are pathogens of dogs and cats. The filarial nematode,
Dipetalonema reconditum, a nonpathogenic parasite of dogs which is
transmitted by fleas is very similar to D. immitis and has been confused
with it, particularly in the early literature (Newton & Wright 1956).
Microfilaria of the two species can be distinguished on the basis of
size and morphology (Lindsey 1962). I have seen D. reconditum on
previous occasions in P. simulans collected from a dog in the
Gainesville area. Fleas serve as a dead end host for Dirofilaria
immitis and become infected as a consequence of the flea feeding on a
microfilaremic host (Soulsby 1982).
Fleas are also intermediate hosts of the tapeworm, Dipylidium
caninum (Chen 1934). Flea larvae become infected when they ingest the
tapeworm eggs which are eliminated within gravid proglottids in the
dog's stool. The eggs hatch within the intestine of the larval flea,
and the oncospheres migrate into the host hemocoel where they develop
into metacestodes. Metacestode development continues throughout meta-
morphosis of the flea from larva to pupa and then to adult. The defini-
tive host becomes infected when it ingests fleas which harbor the infec-
Only 1.0% of the 403 C. felis examined in this study were found to
be infected with D. caninum. This figure seems intuitively low in com-
parison with the high frequency of caninum dipylidiasis. Tapeworm
infection is pathogenic to the flea host (Chen 1934), and it is possible
that infected fleas have a shorter life span on the dog than uninfected
fleas. A moribund infected flea, however, might also be more easily
groomed thus facilitating transmission. Considering the total number of
fleas generated from a single host, the mixing that occurs between
populations of fleas due to host mobility, and the number of tapeworm
eggs that are generated from a single infected dog, the 1.0% infection
rate in fleas becomes very reasonable.
This study revealed a broad range of microorganisms and metazoa
endosymbiotic in local populations of fleas. The baculovirus seen in P.
simulans is the first entomophilic virus to be reported from the order
Siphonaptera. The primitive microsporidia, which will be discussed more
fully in a later chapter, is also recorded here for the first time. The
other symbionts have been observed in fleas but in some cases not well
documented. The organisms reported here were all seen in living fleas;
consequently, little can be said about the effects of parasitism on the
host population, other than that some hosts survive. Cytopathology may
not be obvious in some cases; however, parasitism may have significant
effects in terms of reduced longevity or reproductive potential.
Hopefully, the results generated in this study will provide a foundation
for further study that will lead to the development of effective
biological control systems for fleas.
BIOLOGICAL CHARACTERIZATION OF A LEPTOMONAS SPECIES
IN LOCAL POPULATIONS OF THE FLEA, PULEX SIMULANS
The family Trypanosomatidae is comprised of nine recognized genera
(Vickerman 1976, McGhee & Cosgrove 1980, Lee & Hunter 1985). Five of
these genera are monoxenous, completing their life cycle in a single
host. The genera, Blastocrithidia, Crithidia, Herpetomonas, Leptomonas,
and Rhynchoidomonas are referred to collectively as the lower trypano-
somatids (Lee & Hunter 1985). The genera Endotrypanum, Leishmania,
Phytomonas, and Trypanosoma are heteroxenous, requiring an animal or
plant host and an insect vector. Some of the species in these genera
are important pathogens of man and animals (Wallace et al. 1983).
There is considerable uncertainty concerning the taxonomic status of
the lower trypanosomatids, particularly those species reported from
fleas (Guttman 1963, Wallace 1966, Molyneux et al. 1981). The available
descriptions are old and generally considered to be inadequate. McGhee
and Cosgrove (1980) list eight species hosted by fleas, five in the
genus Leptomonas Kent and three in the genus Blastocrithidia Laird. In
the past, the insect host has been used for specific identification of
the parasite; however, the validity of this taxonomic criterion has been
subject more recently to controversy. Experimental transmission studies
have shown that cross-infectivity can occur between different hosts at
the family and sometimes even at the order level. Closer examination of
trypanosomatid development within the host and natural transmission from
infected host to progeny, however, often indicates that host specificity
is more highly confined (Wallace 1966, McGhee and Cosgrove 1980).
Leptomonas spp. have been reported from the pulicid fleas,
Ctenocephalides canis (Curtis), C. felis (Bouche), Pulex irritans L.,
Spilopsyllus cuniculi (Dale), and Xenopsylla cheopis (Rothschild), but
none has been reported from P. simulans (Baker)(Molyneux et al. 1981).
This study reports some basic biological parameters of a Leptomonas
isolate from the flea, Pulex simulans. No attempt is made here to for-
mally name what appears to be a new species, but rather to provide
observations and measurements of the various forms seen in culture and
to discuss host-parasite interactions, including the apparent mode of
Materials and Methods
The Leptomonas strain was isolated on blood agar, from the hindgut
of a female P. simulans collected off a dog in Newberry, Florida. The
methods used in dissecting the flea host and preparing the blood agar
medium have been described in the previous chapter. The culture isolate
was streaked on a blood agar plate for isolated colonies which were
reinoculated into blood agar slants overlaid with phosphate buffered
saline (PBS) pH 7.2, maintained at 280C, and subcultured at 14 day
intervals. The stock was grown in several different culture media
including Modified Schneider's Drosophila Medium (Gibco, stock no.
350-1720AJ), blood agar slants overlaid with PBS, Lactalbumin hydro-
lysate with Earl's Salts (Sigma, stock no. L 3762), HOSMEM (Berens et
al. 1976), and Eagle's Minimum Essential Medium (Sigma, stock no. M
4767). The defined media above were supplemented routinely with 10%
fetal bovine serum (Sigma, stock no. C 5280).
The following measurements were taken, according to the recommenda-
tions of Wallace et al. (1983), of 50 individuals of each form observed
in culture: cell length (exclusive of the free flagellum), cell width
(at widest point), length of nucleus, width of nucleus, distance of
kinetoplast from anterior end, and length of the free flagellum. The
number of twists in the cell body were also recorded. Measurements were
taken at 1000x magnification, under oil emersion, of parasites which
were grown on blood agar slants and stained with Giemsa stain.
The guts of infected fleas were dissected and placed into a fixative
consisting of 2.5% glutaraldehyde and 2.0% formaldehyde in 0.2 M sodium
cacodylate buffer, postfixed in 1.0% osmium tetroxide, dehydrated in
graded ethanol solutions, and embedded in Spurr's resin (Spurr 1969).
Culture forms were centrifuged at 1000 x G for 3 minutes in a Fisher
Model 59 centrifuge and prepared for electron microscopy as previously
described, except that the parasites were pelleted in low-melting-point
agar prior to dehydration. Thin sections were cut on a LKB Model 8800
Ultrotome III microtome and poststained in an aqueous solution of 1.0%
uranyl acetate and in Reynold's lead citrate. The sections were viewed
and micrographs taken on a Hitachi HU-11E electron microscope.
Temperature Effects on Growth
To determine the temperature range for growth, blood agar slant
culture tubes, overlaid with 1 ml PBS pH 7.2 and supplemented with
gentamicin sulfate and penicillin-streptomycin were inoculated with 1
x 105 log phase parasites which had been maintained previously at 280C.
Three tubes were prepared at each of six temperatures: 12, 18, 25, 30,
33, and 370C. Growth was assessed at days 2, 4, 6, 8, 10, 14, 20, 27,
and 34 by means of a hemacytometer.
Transmission and Host-Specificity Studies
Adult P. simulans along with flea fecal pellets were collected off
of the same dog from which the original isolate was obtained. Specimens
were verified to be infected with the leptomonad flagellates by dis-
section in the modified Insect Ringer's solution described in the previ-
ous chapter. Samples of flea feces were dissolved in PBS on glass
slides, air-dried, fixed with methanol, and Giemsa-stained to confirm
the presence of leptomonads.
In the first transmission experiment, two groups of 20 eggs from
colony-reared P. simulans were placed into 100 x 15 mm plastic petri
dishes containing 25 ml of fine white sand and maintained in a growth
chamber at 23C and 70% RH. The first group was supplied with a pul-
verized mixture of Purina Cat and Rabbit Chows and flea frass from the
previously-mentioned dog. The second group was given the pulverized
mixture and dried dog blood. Adult fleas were dissected upon emergence
and examined for leptomonad infections.
In the second transmission experiment, two groups of 25 eggs from
colony-reared C. felis were placed into petri dishes as described above.
The first group was provided with the pulverized mixture and flea feces
combed one day earlier from the previously mentioned dog. The second
group was given the chow mixture and frass from uninfected colony C.
felis. Emerging adults were dissected and observed for infections.
In the last experiment, two groups of 30 eggs from wild-caught
female P. simulans were placed in petri dishes as described. The first
group was supplied with the pulverized mixture and dried dog blood. The
second group was given the food mixture and 20.0 mg dried dog blood
inoculated with 20 ul of day 17 Leptomonas culture medium, containing
2.1 x 107 organisms per ml.
The Leptomonas stock grew well in all of the culture media. The
forms observed in culture could be placed into three primary groups
based on morphology (Table 3.1). The first form was a long moderately
slender promastigote with a mean cell length of 13.0 pm 2.62 sd (range
of 8.78 to 18.53), excluding the flagellum (Figs. 3.1A,B & 3.2). These
forms were present throughout the life of the culture, but they were
most prevalent up to day 5. The nucleus was located centrally, and the
rod-shaped kinetoplast was frequently located midway between the nucleus
and the anterior end of the cell. Dividing forms were common (Fig.
3.1C). Some individuals had flagella at each end of the cell body. The
common characteristic feature of this form was the presence of 1 to 4
The second form observed in culture was shorter and stubbier than
the first, with the nucleus more anteriorly positioned and the kineto-
plast usually located adjacent to the nucleus on the anterior side (Fig.
3.1D,F). This form had a mean length of 8.41 pm + 1.42 sd (range of
5.85-13.86), a much shorter, often rudimentary flagellum, and no twists.
It was most prevalent from days 7 to 20. Rosettes of 3 to 200 organisms
\C 0ND V
00 -I 0 00
- 01 C00 r
00 O 00 O
C 0.. H
o o o c 6
S a -0000
O 4 () 4-4 0 0 0
O C ~0 u s
b030 0) p J ^0
J MC C-l U u L-I 4J M
r- r-d w a4 I) -
3c '3t 0
Figure 3.1. Culture forms of the Leptomonas strain, isolated from Pulex
simulans. (df) dividing form, (fl) form 1, (f2) form 2, (f3)
form 3, (pointer) aflagellate forms in flea feces.
.* -. ..i "
I ... C f2-1,
l. ..-.. .
Figure 3.2. Culture forms of a Leptomonas sp. from Pulex simulans.
A ,, .:
'. R '
S-Rvj*l t.-i lytnA
,,. *. .,.
B ^r *"*, ,
C.* ,, ,
were commonly seen in cultures from day 1 onward and often included both
stubby and twisted forms (Fig. 3.1E).
The third form observed in culture was an aflagellate form that
appeared primarily from day 14 onward, becoming more abundant as the age
of the culture increased (Fig. 3.1G). Individuals were spherical to
ovoid with a mean length and width of 3.26 pm + 0.96 and 2.63 pm + 0.81,
respectively (ranges of 1.95 to 7.20 for both measurements). Some of
the smaller and more rounded parasites stained very densely with Giemsa.
In infected fleas, transmission electron microscopy revealed dense
numbers of parasites bound to the surfaces of the hindgut and rectal
pads (Fig. 3.3). The greatest concentration of parasites was observed
typically in the region of the pyloric valve at the junction between the
midgut and hindgut. In heavy infections, parasites were present in the
malpighian tubules and more rarely in the midgut. Adherence of the
leptomonads to the surface wall of the hindgut and rectal pads was
mediated by hemidesmosomes formed between the broadened distal flagellar
membrane of the parasites and the cuticular lining of the alimentary
canal. Desmosomes were present between the flagellar and cell body
plasma membranes of adjacent individuals (Figs. 3.3 & 3.4). Dividing
forms were seen attached to the gut wall (Fig. 3.3C). In the malpighian
tubules, no actual cellular junctions were observed, but the flagella of
the parasites interdigitated with microvilli of the tubule wall, sug-
gesting a possible adherence mechanism.
The ultrastructure of these organisms was typical of trypanosomatid
parasites (Fig. 3.3 & 3.4). They had a well-formed nucleus, prominent
mitochondria, and a characteristic kinetoplast situated at the base of
Leptomonads in situ in the hindgut of Pulex simulans. (ax)
axoneme, (d) desmosome, (hg) hindgut cuticle, (i) intra-
flagellar space (k) kinetoplast, (N) nucleus, (nu) nucle-
olus, (p) flagellar pocket.
~: g" '-
Figure 3.4. Leptomonad flagellate. (ax) axoneme, (i) intraflagellar
space, (k) kinetoplast, (M) mitochondria, (N) nucleus, (nu)
nucleolus, (p) flagellar pocket, (arrow) desmosome.
J. I NP
*I & '-