VECTOR COMPETENCE IN LYME DISEASE : STUDIES ON IXODES SCAPULARIS , DERMACENTOR VARIABILIS , AND AMBLYOMMA AMERICANUM

Ticks in the Ixodes ricinus-species complex have been implicated as vectors of Lyme disease spirochetes (Borrelia burgdorferi). A variety of other ticks, notably American dog ticks (Dermacentor variabilis) and Lone Star ticks (Amblyomma americanum) in the U.S., appear incapable of transmitting Lyme disease spirochetes despite ingesting these bacteria. In this study, I confirmed that deer ticks, but not dog ticks or Lone Star ticks became infected with Lyme disease spirochetes after feeding on infected hosts. In addition, I assessed several potential physiologic factors that could affect the ability of ticks to acquire, maintain, and transmit B. burgdorferi. One factor in particular, found in tick saliva, appears responsible for preventing spirochete killing in I. scapularis but not D. variabilis or A. americanum. In one study, I assessed anti-microbial activity of the tick's immune system. In particular, I determined whether ticks exhibited measurable phenoloxidase and antimicrobial activity. Phenoloxidase activity in the hemolymph of larvae of the greater wax moth (Galleria me/lone/la L) was compared to phenoloxidase activity in I. scapularis, D. variabilis and A. americanum. Although activity was detected in wax moth hemolymph, no phenoloxidase activity was detected in three species of ixodid ticks. In addition, Enterobacter cloacae was used in an attempt to induce anti-microbial activity in punched cuticle of three species of ixodid ticks. Using this assay, no anti-microbial activity was detected in ticks. Moreover, no anti-microbial substances were found in the midgut of I. scapularis, D. variabilis and A. americanum. Factors associated with the tick's humoral immunity do not appear to play a role in vector competence for Lyme disease spirochetes. In another study, host-associated borreliacidal factors were examined in vitro. Spirochetes survived in the presence of pre-immune rabbit serum but were killed in the presence ofimmune rabbit serum. Heat-inactivation (56 °c, 1 hr) destroyed the killing effect of immune serum but the borreliacidal activity of heat-inactivated immune serum could be restored by the addition of pre-immune serum. Using a similar assay, spirochetes survived in tick midgut extract cultures when ticks were removed from preimmune hosts. Spirochetes also survived in I. scapularis midgut extract cultures, but not in D. variabilis or A. americanum midgut extract cultures when ticks were removed from immune hosts. As it did in immune serum, heat-inactivation destroyed the killing effect of D. variabilis and A. americanum midgut extract cultures. An indirect immunofluoresence assay indicated that anti-B. burgdorferi antibody concentrations were similar in both the host and the tick midgut extract. Taken together, these experiments suggest that borreliacidal activity of host immune serum is mediated by the activity of complement. Furthermore, this same activity is found in the midgut of D. variabilis and A. americanum feeding on immune hosts but is not found in I. scapularis. Since vector competence appeared to be related to the activity of host antibody and complement in the tick midgut, I determined whether substances existed in I. scapularis that might inhibit or inactivate antibody or complement. We tested for the presence of an antibody-cleaving enzyme in the gut extract of I. scapularis and were unable to demonstrate any activity in several dilutions of gut extract after 30 minutes at 37oc. A spirochete survival assay was performed in cobra venom, which possesses powerful anticomplement activity, and in the saliva collected from I. scapularis, D. variabilis and A. americanum. Spirochete survival of over 95% was observed in immune serum with cobra venom and I. scapularis saliva but no spirochete survival was found in cultures containing saliva from D. variabilis and A . americanum. We suggest that an anticomplement factor in the saliva of I. scapularis inactivates complement in the midgut of engorged ticks. Finally, the duration of tick attachment may serve as a useful predictor of risk for acquiring various tick-transmitted infections such as Lyme disease and babesiosis. We measured three tick engorgement indices (El) at known time intervals after tick attachment and used these indices to determine the length of time that ticks were attached to tick-bite victims in selected Rhode Island and Pennsylvania communities where the agents of Lyme disease and human babesiosis occur. Regression equations developed correlate tick engorgement indices with duration of feeding. More than 60% of tick-bite victims removed adult ticks by 36 hours of attachment, but only 10% found and removed the smaller nymphal ticks within the first 24 hours of tick feeding. A table containing specific EI prediction intervals were calculated for both nymphs and adults allowing practitioners or clinical laboratories to use easily-measured tick engorgement indices to predict transmission risk by determining the duration of feeding by individual ticks.

ticks became infected with Lyme disease spirochetes after feeding on infected hosts. In addition, I assessed several potential physiologic factors that could affect the ability of ticks to acquire, maintain, and transmit B. burgdorferi. One factor in particular, found in tick saliva, appears responsible for preventing spirochete killing in I. scapularis but not D. variabilis or A. americanum.
In one study, I assessed anti-microbial activity of the tick's immune system. In particular, I determined whether ticks exhibited measurable phenoloxidase and antimicrobial activity. Phenoloxidase activity in the hemolymph of larvae of the greater wax moth (Galleria me/lone/la L) was compared to phenoloxidase activity in I. scapularis, D. variabilis and A. americanum. Although activity was detected in wax moth hemolymph, no phenoloxidase activity was detected in three species of ixodid ticks. In addition, Enterobacter cloacae was used in an attempt to induce anti-microbial activity in punched cuticle of three species of ixodid ticks. Using this assay, no anti-microbial activity was detected in ticks. Moreover, no anti-microbial substances were found in the midgut of I. scapularis, D. variabilis and A. americanum. Factors associated with the tick's humoral immunity do not appear to play a role in vector competence for Lyme disease spirochetes.
In another study, host-associated borreliacidal factors were examined in vitro.
Spirochetes survived in the presence of pre-immune rabbit serum but were killed in the presence ofimmune rabbit serum. Heat-inactivation (56 °c, 1 hr) destroyed the killing effect of immune serum but the borreliacidal activity of heat-inactivated immune serum could be restored by the addition of pre-immune serum. Using a similar assay, spirochetes survived in tick midgut extract cultures when ticks were removed from preimmune hosts. Spirochetes also survived in I. scapularis midgut extract cultures, but not in D. variabilis or A. americanum midgut extract cultures when ticks were removed from immune hosts. As it did in immune serum, heat-inactivation destroyed the killing effect of D. variabilis and A. americanum midgut extract cultures. An indirect immunofluoresence assay indicated that anti-B. burgdorferi antibody concentrations were similar in both the host and the tick midgut extract. Taken together, these experiments suggest that borreliacidal activity of host immune serum is mediated by the activity of complement. Furthermore, this same activity is found in the midgut of D. variabilis and A. americanum feeding on immune hosts but is not found in I. scapularis.
Since vector competence appeared to be related to the activity of host antibody and complement in the tick midgut, I determined whether substances existed in I. scapularis that might inhibit or inactivate antibody or complement. We tested for the presence of an antibody-cleaving enzyme in the gut extract of I. scapularis and were unable to demonstrate any activity in several dilutions of gut extract after 30 minutes at 37oc. A spirochete survival assay was performed in cobra venom, which possesses powerful anticomplement activity, and in the saliva collected from I. scapularis, D. variabilis and A.
americanum. Spirochete survival of over 95% was observed in immune serum with cobra venom and I. scapularis saliva but no spirochete survival was found in cultures containing saliva from D. variabilis and A . americanum. We suggest that an anticomplement factor in the saliva of I. scapularis inactivates complement in the midgut of engorged ticks.
Finally, the duration of tick attachment may serve as a useful predictor of risk for acquiring various tick-transmitted infections such as Lyme disease and babesiosis. We measured three tick engorgement indices (El) at known time intervals after tick attachment and used these indices to determine the length of time that ticks were attached to tick-bite victims in selected Rhode Island and Pennsylvania communities where the agents of Lyme disease and human babesiosis occur. Regression equations developed correlate tick engorgement indices with duration of feeding. More than 60% of tick-bite victims removed adult ticks by 36 hours of attachment, but only 10% found and removed the smaller nymphal ticks within the first 24 hours of tick feeding. A table containing specific EI prediction intervals were calculated for both nymphs and adults allowing practitioners or clinical laboratories to use easily-measured tick engorgement indices to predict transmission risk by determining the duration of feeding by individual ticks.

INTRODUCTION
Insects and ticks are well known vectors of human and animal pathogens, transmitting a wide variety of microorganisms. These microorganisms replicate in vertebrate and invertebrate hosts. The invertebrate host is a blood-sucking arthropod that is competent to transmit the pathogen between susceptible animals. Microorganisms transmitted by ticks must adapt to the peculiar physiological and behavioral characteristics of ticks, particularly with regard to blood feeding, bloodmeal digestion, and molting (Nuttall et al. 1994). What is still poorly understood, however, is why certain of these arthropods are incompetent as vectors of the pathogen which they encounter during bloodfeeding.
Innate barriers to infection resulting in a potential vector's inability to transmit ingested pathogens have largely been characterized in ambiguous terms such as midgut or salivary gland barriers. Despite considerable research on such important infections as mosquitotransmitted malaria and tick-transmitted rickettsiae and viruses, we cannot yet identify a single molecule responsible for vector competence.
Ticks in the lxodes ricinus species complex have been implicated as vectors of Lyme disease spirochetes (Borrelia burgdorferi) worldwide. Other varieties of ixodid ticks, notably dog ticks (Dermacentor variabilis) and Lone Star ticks (Amblyomma americanum) in the U.S. have been shown in the laboratory to be incompetent vectors for Lyme disease spirochetes , Mather & Mather l 990a, Lindsay et al. 1991. AlthoughD. variabilis and A.
americanum may be capable of acquiring spirochetes from the host, infection disappears and is not detectable after molting. Jxodes scapularis, however, is capable of maintaining the infection and efficiently passing B. burgdorferi transstadially. Since it has been speculated that D. variabilis and A. americanum may possess some factor inside their gut 1 that destroys the spirochetes, while B. burgdorferi survives in the midgut during I. scapularis transstadial transformation, it would appear that factors promoting vector competence for this bacteria reside within the midgut . lxodes scapularis possess potent pharmacological armaments in their saliva, including apyrase, a kininase, an anaphylatoxin-destroying capacity, vasodilatory prostaglandins and an anticomplement protein. It has been well described how these substances serve the tick in successful blood feeding, and perhaps in modifying the tick feeding site so as to facilitate spirochete infection of the host .
Further research was needed on the adaptation and replication of B. burgdorferi in gut and saliva of ticks to determine the effect of pharmacological factors in gut and saliva on spirochetes. These factors may be involved in the mechanism of vector competence for Lyme disease.

LITERATURE REVIEW
Lyme disease was first recognized as a clinical entity in 1909 by a Swedish dermatologist and was termed erythema migrans. The first recognized American case of Lyme disease was reported in Wisconsin in 1969 by a physician who was familiar with the European literature . He sought spirochetes in biopsy material and used penicillin to effect a cure. The first recognized epidemic of Lyme disease appeared in coastal Connecticut in 1975. The aflluent suburban community of Old Lyme gave its name to the disease then afilicting so many of its inhabitants. Some 51 residents of the site were found to be suffering from an atypical arthritic condition that was generally preceded by an annular rash. lxodes ticks immediately became suspect as vectors for this disease (Steere et al. 1977). Ixodes dammini was recognized as a distinct species in 1979 . Outbreaks of infection seem to occur solely where I. dammini is abundant, especially in the northeastern and north central portion of the United States, the regions in which cases cluster most intensely. However, Oliver et al. (1993) suggested that I. dammini , is not a valid species separate from I. scapularis Say, 1821. No major divergence could be demonstrated between I. scapularis and I. dammini in experiments involving hybridization, assortative mating, morphometrics, chromosomes, isozymes, life cycles, host preferences, vector competencies, and DNA sequences. The pathogen of this disease has been isolated and named Borre/ia burgdorferi . From 1986 to 1990, Lyme disease cases accounted for 81 % of all reported cases of arthropod-transmitted diseases in the United States (Centers for Disease Control 1992). Lyme disease is now the most important arthropod-transmitted disease in this country.

Pathogen (Spirochete)
The genus Borrelia is composed of21 species, the majority of which are asociated with relapsing fever illnesses that are transmitted by soft-bodied ticks {Argasidae) . In contrast to all the other species oftickassociated borreliae, B. burgdorferi is transmitted by hard-bodied ticks (Ixodidae ), namely the lxodes ricinus complex , such as I. scapularis, I. ricinus, I. pacificus, and I. persulcatus Ai et al. 1988). Borrelia burgdorferi is morphologically similar to Treponema. Cells are irregularly coiled with tapered ends and possess 6 to 8 axial fibrils located beneath the outer membrane; the cell diameter is about 0.2 um and ranges from 4 to 30 um in length ). The spirochetes tend to remain in the midgut of unfed ticks, where they are sequestered in the microvillar brushborder and intercellular spaces of the epithelium  a marked cross-antigenicity with other spirochetes and even more distantly related bacteria, antibodies against the 94 KDa, 3 1 KDa, and 21 KDa protein are largely speciesspecific. The early immune response in Lyme borreliosis is triggered mainly by the flagellin. In the later stage a wide range of immunogenic proteins is involved, with the 94 KDa antigen being the best marker for late immune response , Zoller et al. 1993). Five to seven HSPs (heat shock proteins) of B. burgdorferi were detected and HSP may result in an autoimmune reaction causing arthritis (Carreiro et al. 1990).

4
The major proteins of Bo"e/ia isolated from I. scapularis collected on Shelter Island, New York were similar to proteins in Bo"e/ia from I. pacificus, wild animals, and humans in the United States (Anderson et al. 1983, Steere et al. 1983a, from!. ricinus, wild animals, and humans in Europe , and from!. persulcatus and humans in Asia (Ai et al. 1988). Variants with different major outer surface proteins from the seminal B 31 strain have frequently been isolated from humans and I. ricinus in Europe Wilske et al. 1985Wilske et al. , 1986Wilske et al. , 1988Anderson et al. 1986a), but in the United States the isolates from humans, rodents, and most ticks have been remarkably similar to one another and the B 31 strain . Exceptions have included isolates from the songbird (Catharusfuscescens), Cottontail rabbits (Sylvilagusjloridanus), I. dentatus, I. scapularis, I. neotomae, and I. pacificus Bisset & Hill 1987;Anderson et al. , 1989Lane & Pascocello 1989). While major proteins vary considerably among European Bo"e/iae isolated from humans, specific illnesses linked to particular variants have not been documented.

Animal Reservoir:
Bo"elia burgdorferi has been isolated from or detected in tissues of three domestic and wild mammals and eight birds in United States, three rodents in Europe, and one rodent in Asia (Anderson & Magnarelli 1993). The white-footed mouse is a particularly important host for the spirochete in the northeastern and midwestem United States (Anderson et al. 1983(Anderson et al. , 1986c(Anderson et al. , 1987b). These mice apparently harbor the spirochete throughout their lives , and where Lyme disease is prevalent 70 % to 80 % or more of the mice may be infected. Prevalence of infection is highest in summer, following peak feeding by nymphs, and lowest in winter, when immature ticks are inactive (Anderson et al. 1987a).
Other species of rodents, such as eastern chipmunks, may also be important reservoirs, but these animals have not been extensively studied . Apodemus may be an important reservoir in Europe and Asia , Miyamoto et al. 1991).
Deer were reported as reservoir incompetent (Telford et al. 1988). Antibody to B. burgdorferi has been detected in deer (Magnarelli et al. , 1986a and these animals are extensively parasitized by I. scapularis , Anderson & Magnarelli 1980 (Magnarelli et al. ,b,c, 1985Rawlings 1986;Teitler et al. 1988;Lane l 990b;. Birds also are parasitized by I. scapularis and are naturally infected with B. burgdorferi ). Only one isolate of this bacterium has been obtained from a field-caught bird, a veery (Catharusfuscescens) 6 (Anderson et al. 1986), and this was antigenically different from strain B 31 . Larval ticks feeding on birds, as on mammals, often harbor Borre/iae ).

Tick vectors:
The principal vectors of Lyme disease are ticks belonging to the I. ricinus complex.
pacificus. The Eurasian species is identified as I. persulcatus and the European species is I. ricinus.
The two species in North America tend to occur in different geographical areas.
lxodes pacificus extends along the west coast of the United States from the California-Mexico border northward into southwestern Canada , Anderson 1989. Ixodes scapularis extends northward from Virginia into New Hampshire, Vermont, and Maine and westward through southern Ontario to its westernmost border in Minnesota and Iowa, including southern Illinois, and southward from Florida to central Texas and the southeastern United States Anderson et al. , 1990aWilson et al. 1988;Anderson 1989;; .
Jxodes ricinus extends from England to about 500 to 550 east longitudinally , Anderson 1989. It has been collected in North Africa and as far north as 650 latitude. lxodes persulcatus overlaps with I. ricinus in eastern Europe and is found in a relatively broad band across Asia , Anderson 1989.
Each tick feeds three times in its life, initially as a larva, then as a nymph, and finally as an adult. The duration of the life cycle is a year or more. lxodes scapularis and /.
pacificus may take one to two years to complete their lives . Ixodes ricinus and I.
persu/catus often complete their life cycles in two years. The seasonal distribution of the four life stages of I. scapularis is as follows : larvae hatch from eggs predominately in midsummer, and after feeding fully from a host they detach and drop to the ground. Larvae either molt into nymphs or remain in an engorged state in the duff layer of the soil throughout the winter (Yuval & Spielman 1990). Blood-fed larvae molt into nymphs the following spring. Host-seeking nymphs become abundant in late spring and early summer. After ingesting blood from animals, they fall to the ground and molt into adults. Males and females seek large mammals. The fully engorged females lay the eggs and repeat the cycle. Seasonal abundance of the three feeding stages of the other species differs from I. scapularis. lxodes pacificus may be collected throughout the year in California , Westrom et al. 1985, Lane & Loye 1989. In unfed lxodes ticks the Lyme disease spirochetes are usually found sequestered within the tick's midgut. In fed ticks, infection is disseminated; spirochetes might be found in saliva, hemolymph, or other tick tissue. Disseminated infection is rare in flat ticks , Ribeiro et al. l 987b, Ewing et al. 1994. The direct inoculation of spirochetes into vertebrates via 8 regurgitated midgut contents has been suggested as one possible route of transmission  but has not been critically examined. However, B. burgdorferi have been directly observed in the salivary gland , Zung et al. 1989), e.g. in pilocarpine-stimulated salivary secretions collected from infected ticks, specifically feeding nymphal and adult I. scapularis, thus documenting that spirochetes are inoculated via saliva . Following a tick's attachment to its host, spirochetes multiply ) and may congregate in intercellular pits near the microvillar brushborder of midgut epithelial cells  where some leave the midgut by both inter-and intracellular penetration of the gut epithelial cells (Zung et al. 1989). Once free of the gut's basement membrane, spirochetes disseminate, traveling throughout the tick's hemocoel via hemolymph  ). Disseminated spirochetes may infect various tick tissues, including salivary glands and ducts, ovaries, and the central ganglion (Zung et al. 1989). However spirochetes were observed in saliva only of ticks having hemolymph infection  ). In adult ticks, spirochetes were observed in the hemolymph of gut-infected adult I. scapularis as early as 24 hours after tick attachment and spirochetes were observed in saliva as soon as 72 hours following tick attachment. Spirochetes were frequently observed within the salivary glands of nymphal ticks after 48 hours of attachment and less frequently in nymphs attached for only 24 hours (Zung et al. 1989).
So, the duration of tick attachment on animals is an important factor which allows spirochetes to be transmitted from ticks to animals. The ability of I. scapularis to transmit B. burgdorferi, for example, is obtained only after a certain amount of feeding.
Transmission of this infectious agent is most efficient when ticks feed to repletion . Spirochetes were rarely transmitted sooner than 24 hours in nymphal ticks feeding on white-footed mice and hamsters ) and were rarely transmitted sooner than 36 hours in adult ticks feeding on rabbits . It is important to investigate the time of attachment in order to predict the risk of Lyme disease and removing attached ticks within 24 hours is an important strategy in the management of Lyme disease.
The prevalence of spirochete infection (tick infection rate) varies widely among distinct populations of vector ticks. In I. scapularis, adult ticks generally display about twice the rate of B. burgdorferi infection as do nymphs. In highly endemic areas, spirochete prevalence in populations of adult I. scapularis typically ranges between 3 5 % and 75 %, while that for nymphs is between 15 % and 35 % (Anderson et al. 1983;Piesman et al. 1986bPiesman et al. , 1987a).
In contrast to I. scapularis in the eastern United States, spirochete prevalence in the western I. pacificus is quite low. Borrelia burgdorferi typically infect between 0 % to 5. 9 % of adult ticks collected from vegetation in northern California and Oregon , Bisset & Hill 1987. However, transovarial passage of spirochetes appears to occur more efficiently in I. pacificus than in I. scapularis. One of three field-collected female I. pacificus, whose ovaries contained spirochetes, successfully infected 100 % of its progeny . Furthermore, these same larval ticks maintained their spirochetal infections, passing them transstadially into the nymphal and adult stages. These adult ticks eventually passed their infection transovarially, also with great efficiency.
Ticks not belonging to the I. ricinus c omplex have only rarely demonstrated competence as vectors of B. burgdorferi , Magnarelli et al. l 986a, Anderson et al. 1985, Teltow et al. 1991)-Despite their ability to become infected by ingesting spirochetes during feeding, ticks such as Dermacentor variabilis, Amblyomma americanum, I. holocyclus, and Haemaphysalis leporispalustris appear to lose their infection either before or during transstadnal passage .
Determinant factors influencing vector competence are yet unknown. Many factors, including physiological factors, probably i::nfluence the ability of ticks to become infected with, transstadially pass, and ultimately tra nsmit pathogens. As a physiological factor, vector-competent lxodes may possess a sp irochetal growth-promoting factor that other groups of ticks lack. Spirochetes can be observed in preparations made from several species of ticks as well as other arthropod s, such as mosquito and fleas, soon after the ingestion ofinfected blood (Magnarelli et al. l 986b, Rawlings 1986). Spirochetes have also been shown to multiply rapidly in I. s capularis during the first two weeks following ingestion ). Spirochete survival in vectorincompetent arthropods is ephemeral, ho\Wever, usually lasting less than two to four weeks. As an alternative hypothesis, vect or-incompetent ticks may possess spirochete growth-inhibitory factors that may not occ ur, or occur at lower concentrations, in the vector-competent lxodes. lxodes scapularis secretes copious amounts of saliva containing a variety of antihemostatic, anti-inflammatory, immun osuppressive properties, and an anticomplement factor (Ribeiro et al. 1985 & l 987b ). Active immunosuppressive chemicals, including prostaglandin E2 (PGE2), blockr macrophage activation and neutrophil activity 11 and may inhibit T -cell activation, all early precursors in the cascade of cellular events leading to antibody production and antigen processing. Thus, salivary components injected by I. scapularis may promote infection of the host by Lyme disease spirochetes.
By deactivating macrophages, PGE2 may serve to protect B. burgdorferi during its initial phase of adaptation in the skin of infected hosts, allowing spirochetes to escape into tissues. Anti-complement may serve to interrupt the complement pathway and spirochetes can survive even when exposed to the host's humoral immune system. All these pharmacological factors may play a role in vector competence for I. scapularis.
Among the principal and potential vectors of B. burgdorferi, it is perhaps more important to consider intrinsic vector competence. Internal physiological ability to become infected with and transmit spirochetes and other innate behavioral characteristics promoting infection or transmission are major factors in vector competence  . For example, in laboratory studies, both I. scapularis and I. pacificus are highly competent vectors of B. burgdorferi , while the natural prevalence of infection among these ticks in the southeastern and western United States is relatively low (less than 4 % ) . Because immature stages of I. scapularis and I. paci.ficus frequently infest lizards , which are inhospitable hosts for B. burgdorferi , it is most likely that perceived differences in the vector potential of these ticks when compared with that of I. scapularis result more from their predilection for feeding on noninfective hosts than on any physiological vector incompetence. Thus, the intrinsic vector competence and, more specifically in this case, the feeding behavior exhibited by biologically competent vectors, may significantly influence the prevalence of infection in   Comparing the relative potential of rodents as reservoirs of the Lyme disease spirochete (Bo"e/ia burgdorjeri

Deer ticks (Ixodes scapularis) but not American dog ticks (Dermacentor variabilis)
or lone star ticks (Amblyomma americanum) become infected with and transmit Lyme disease spirochetes (Bo"e/ia burgdoiferi). We evaluated factors associated with the midgut of these three ixodid ticks that might affect their ability to serve as vectors for this spirochete.
When larvae of all three species of ticks ingested blood from a spirochete-infected rabbit, only nymphal deer ticks(!. scapularis) derived from these larvae retained the infection (infection rate = 82% ). In vitro experiments demonstrated that spirochetes survived in the presence of pre-immune rabbit serum but were were killed in the presence of spirochete-immune rabbit serum. Moreover, spirochetes also survived in the presence of immune rabbit serum heated to 560C for 1 hour to remove the activity of serum complement. Additionally, spirochetes survived in the presence of midgut extracts from all three species of ticks fed on pre-immune rabbits, and in the presence of deer tick gut extract from ticks fed on immune rabbits. However, spirochetes were killed in the presence of dog tick (D. variabilis) and lone star tick (A. americanum) gut extract when these ticks fed on immune hosts. Heat inactivation of the gut extract restored spirochete survival regardless of the tick species. The a.-B. burgdoiferi antibody titer in immune serum and tick gut extract was similar (1 : 1024), even after incubation at 37oc for up to 90 minutes, suggesting that these ticks lack an antibody-cleaving enzyme in their midgut.
Thus, inactivating serum complement appeared to be critical for spirochete survival.
Using a borreliacidal micro assay composed of immune serum and spirochetes, the addition of as little as 2 µg protein of cobra venom (  . Because B. burgdoiferi survive in the midgut during I. scapularis' transstadial transformation, it would appear that factors promoting vector competence for this bacteria reside within the tick midgut. However, these ticks also possess a potent pharmacological armament in their saliva, including apyrase, a kininase, an anaphylatoxin-destroying capacity, vasodilatory prostaglandins and an anticomplement protein (Ribeiro , 1988Ribeiro & Spielman 1986). It has been well described how these substances serve the tick in successful blood feeding, and perhaps in modifying the tick feeding site to facilitate spirochete infection of the host. It may be that re-ingested saliva also exhibits some activity in the tick's midgut. In particular, we observed antagonism of the in vitro borreliacidal activity of serum from immune hosts after heat treatment, treatment with a known complement inhibitor (cobra venom), midgut extracts from engorged/.
scapularis, as well as with I scapularis saliva. Anti-complement activity in the saliva of I. scapularis may well serve to determine this tick's ability to become infected with B. burgdorferi.

Jn vivo experiments
To assess vector competence for Lyme disease spirochetes (B. burgdorferi) in lxodes scapularis, D. variabilis and A. americanum feeding on rabbits, larval ticks of all three species derived from a spirochete-free laboratory colony were allowed to feed simultaneously on the ears of a spirochete-infected New Zealand white rabbit (Charles River Laboratories, Wilmington MA). The rabbit had been infected three weeks previously by allowing more than 50 field-collected female I. scapularis to engorge to repletion. Field-collected ticks were from Prudence Island, Rhode Island, where the prevalence of infection in adult I. scapularis was 40%. All immature ticks in the laboratory colony were derived from adult ticks also collected on Prudence Island.
Engorged larvae were collected in cloth bags affixed to the rabbit's ears, and then stored in plastic vials at 23oc and >95% relative humidity until molting into nymphs. Nymphs of all three species, derived in this manner, were examined for the presence of B.
burgdorferi by dissecting midgut tissues onto a glass slide and treating them directly with fluorescein isothiocyanate-conjugated (FITC) antibodies to B. burgdorferi in a direct fluorescent antibody assay (Piesman et al.1986.
To prepare tick midgut homogenates, field-collected adult I. scapularis, D. variabilis and A. americanum were fed to repletion in groups of 30 on the ears of NZ white rabbits.
Engorged ticks were collected following detachment, weighed and then dissected to remove all midgut tissue. Pools composed of midguts from five ticks were stored at -100 C until used in an assay.
To obtain serum from pre-immune and B. burgdoiferi-immunized rabbits, blood was collected by venapuncture either just before allowing ticks to feed, or 4 weeks following tick feeding. Blood was allowed to clot for 1 hour at 4oc, and then was centrifuged at 2000 g for 15 min. Serum was immediately removed and stored at -7ooc. Antibody titers were assessed by an indirect immunofluorescent assay .
Tick saliva was collected by allowing field-collected adult I. scapularis, D. variabilis and A. americanum to partially engorge on NZ white rabbits. Ticks were harvested on the fifth through seventh day of engorgement Gust prior to their rapid engorgement phase), rinsed in distilled water and then prepared for collecting saliva as described previously (Tatchell 1967. Briefly, ticks were affixed to glass slides using double-faced tape and their mouthparts inserted into a sterile, glass micropipette.
Salivation was induced by topically applying 2 µl of pilocarpine to the ticks' scutum. All saliva collected in about two hours from a group of 15-30 ticks was pooled and stored at -7ooc until used.

In vitro experiments
We developed a simple in vitro microassay to assess borreliacidal activity of host serum components as well as spirochete survival in the presence of potential borreliacidal antagonists. In this assay, 25 µl of host serum, either pre-immune, immune, or heatinactivated (56°C, 1 hr) sera from New Zealand white rabbits or laboratory-raised whitefooted mice (Peromyscus /eucopus) was added to an equal volume ofBSKII containing 2 X 106 actively-growing B. burgdoiferi. In some experiments, varying amounts of tick midgut extract, tick saliva, or cobra venom (a complement antagonist) were added to these cultures. The mixture was incubated at 33oc for 4-18 hours depending on the experiment, at which time cultures were centrifuged (14,000 x g for 12 s), and examined for spirochete survival on the basis of motility, refractivity, and extensive surface blebbing using darkfield and phase contrast microscopy.
To assess the borreliacidal activity of host antibody as compared to serum complement I designed an experiment with three treatments:(!) Two-fold serial dilutions (1:2 to 1:4046) of immune rabbit serum, (2) dilutions of pre-immune rabbit serum in the presence of undiluted, heat-inactivated immune serum (titer: 20481/X), and (3) dilutions of heat-inactivated immune serum in the presence of undiluted, pre-immune serum.
Spirochete survival was determined after overnight incubation at 33oc. The immune rabbit serum has antibodies and complement, both of which are diluted in treatment (1).
In treatment (2), the complement in pre-immune serum is diluted, but antibody concentrations remain high in the undiluted, heat-inactivated immune serum. Antibodies are diluted in treatment (3), but complement concentrations remain high in the undiluted, pre-immune serum.
To evaluate the effect of tick midgut factors on spirochete survival, extracts were prepared from a homogenate of midguts dissected from pools of 5 rabbit-engorged female ticks. Briefly, frozen midguts were macerated in enough PBS to make a 20% w/vol homogenate, centrifuged (14,000 x g for 12 s) to pelletize the gut tissue, and the supernatant passed through a 0.45 µm filter. Gut extracts from/. scapularis, D. burgdorferi antigen (strain B31) was prepared by centrifuging (35, 000 x g for 45 minutes) 8 ml of a two-week old culture. The resulting pellet was washed three times by resuspending in equal parts of phosphate-buffered saline (PBS). After the final wash, the spirochete concentration was adjusted to 10 7 cells per ml . One drop of the antigen solution was placed into wells on an eight-well immunofluorescence microscope slide. Slides were air-dried, then fixed in acetone for 10 min. Slides prepared in this way were overlaid with either heated (560C, 1 hr) or non-heated host serum or tick midgut extract in two-fold serial dilutions, incubated for 30 min at 37oc, then washed in PBS for 20 min. After washing, slides were treated with a goat a-rabbit lgG antibody conjugated to FITC (Sigma Chemical Co., St. Louis, MO) for 30 min at 37oc, then washed again in PBS for 20 min. Prepared slides were air-dried, mounted in buffered glycerol and examined for fluorescence. The antibody titer was determined to be the dilution prior to the fluorescence end-point.

RESULTS
When larvae of deer ticks(/. scapularis), dog ticks (D. variabilis) and lone star ticks (A . americanum) were simultaneously fed on an infected New Zealand white rabbit, B.
burgdorferi was detected in 82% of the derived nymphal I. scapularis but could not be detected in either nymphal D. variabilis or A. americanum (Table 1). Additionally, deer ticks took more than twice as long to molt from the larval to the nymphal stage than either the dog ticks or lone star ticks.
Spirochetes survived in vitro in pre-immune rabbit sera but were killed in sera collected 4-6 weeks after infection of the host by tick bite up to a dilution of 1 : 3 2 ( Figure   1 ). In more dilute sera, an increasing proportion of spirochetes survived; 100% survival was noted at a serum dilution of 1 :4096. At lower dilutions, little change in the spirochetes could be seen after just 30 min of incubation with immune serum but by 1.5 hr, both spirochetal form and motility were lost in 68.6% of those bacteria counted.
Spirochete killing increased by 3 hr and by 4 hrs of incubation all spirochetes were killed.
Heat-inactivation (560C, 1 hr) of immune sera prevented spirochete killing, but borreliacidal activity was restored by the addition of pre-immune rabbit sera (Table 2).
Heat-inactivation of the serum appeared to have no effect on antibody titer (Table 3).
The assay for spirochete survival in immune serum showed the following survival 0 %, 33 .9 %, 91.5 %, 97.9 %, and 100 % at 1 : 16, 1 : 32, 1 : 64, 1 : 128, and 1 : 256, respectively. The last dilution for 0 % survival of the serial dilution of pre-immune serum mixed with undiluted heat-inactivated immune serum was 1 : 16, but the last dilution for 0 % survival of the serial dilution of the heat-inactivated immune serum mixed with undiluted pre-immune serum was 1 : 256. The results showed that the killing of spirochetes was more efficient in undiluted pre-immune serum mixed with the serial dilution of heat-inactivated immune serum, however 1 : 2048 dilution, the three treatments all had 100 % survival ( Figure 2). It seemed that the killing of spirochetes required a large quantity of complement but only a small quantity of antibody.
Spirochetes survived when incubated with midgut extracts from all three species of ticks bloodfed on pre-immune rabbits (Table 4). However, spirochetes incubated with extracts from A. americanum and D. variabilis taken from immune rabbits were killed while those incubated with extracts from I. scapularis survived. Spirochetes survived exposure to heat-inactivated midgut extracts from all three tick species fed on immune rabbits. Serum antibody titers in pre-immune and immune rabbits, and in tick gut extracts from ticks derived from those rabbits were comparable (Table 5) In treatments 5, 6, 7, and 8, the spirochetes were killed in immune rabbit serum and pre-immune rabbit serum mixed with heat-inactivated immune rabbit serum. Complement may be involved in the killing of spirochetes, because the antibody of heat-inactivated immune rabbit serum was 204g llxand was the same titer as non-heat treated serum with complement intact. The only difference was that heat treated serum may have the heatlabile complement destroyed. It was reported that immune hamster serum failed to kill spirochetes in the absence of complement . In treatment 7, the heat-inactivated immune rabbit serum (Antibody) was restored by the addition of pre-immune rabbit serum (Complement) to kill the spirochetes.
The antibody titer of tick gut extract depends on the antibody titer of the host. In our experiments, there were no changes in the antibody titer of rabbit serum and gut extract of three kinds of ixodid ticks during the first feeding and the second feeding. It is interesting that the spirochete can be killed by immune rabbit serum in vitro, but the larval ticks of I. scapularis still picked spirochete up from the immune rabbit in xenodiagnosis. This may suggest that the spirochete has a mechanism to protect itself from an effective borreliacidal response and to evade host defenses , Barbourr et al. 1986.
The other interesting factor which may be involved in vector competence was some component in the gut extract of I. scapularis. This factor might allow the spirochetes to survive in vitro and in vivo in this species. In experiments 1 to 8, we concluded that the killing of spirochetes required antibody and complement in rabbit serum. Similar results were reported in human and hamsters; immune serum was able to kill B. burgdorferi when the presence of complement . It also seems that the killing of spirochetes needs a great amount of complement, compared to antibody ( Figure 2). It has been reported that B. burgdorferi activated both the alternative and classical complement pathways in normal human serum (NHS), however, the killing of spirochetes was demonstrated through the classical complement pathway. It is not clear why the alternative pathway does not participate in the killing mechanism. It was suggested that the failure of the alternative pathway to mediate killing in the presence of the antibody must ultimately be related to inappropriate interaction of the membrane attack complex (MAC) with the bacterial surface as a consequence of a different localization of the alternative pathway C5 convertase. Thus, the alternative pathway C5 convertase may result in Csb-9 complexes that are lytically inefficient because of deposition of Csb at inappropriate sites not susceptible to the action of the antibody , Kochi et al. 1991 (Appendix I). If antibody and complement were involved in killing of spirochetes, a factor in I. scapularis may inhibit or cleave the antibody or complement allowing the survival of spirochetes. One possible factor is an enzyme in I. scapularis which may cleave the antibody very quickly; this enzyme is also heat-labile. In this manner, the spirochetes could survive even in ticks fed on immune rabbits. The test for digesting antibody from gut extract of I. scapularis at different concentrations showed that antibody titers were the same in different treatments at 37oc for 30 minutes, compared to the control group. These data indicated that the enzyme for digesting antibody may not be the factor allowing the survival of spirochetes.
The other possible factor is anti-complement in the gut of I. scapularis. The complement from rabbit serum in tick gut extract may lose the ability to kill the spirochetes through the addition of an anti-complement factor from I. scapularis. In this case, the spirochetes could survive, therefore anti-complement may be involved in vector competence.
The saliva-activated transmission (SAT) factor was demonstrated in the salivary glands of competent vectors of Tho goto virus, but not in a limited number of noncompetent vectors (Jones et al. 1992). An SAT factor may exist in the saliva of I.
scapularis which determines the vector competence of I . scapularis.  reported that anti-complement activity was found in the saliva of I. scapularis. The anticomplement activity may exist in the gut of I. scapularis, too, because the channel for saliva secreted by the salivary gland passing through feeding ticks is the same channel in which the blood meal is ingested (Odhiambo 1982). The saliva, with anti-complement, may possibly mix with the blood meal in the feeding channel of the tick and pass through to the gut or get added into the blood of the cavity which is formed in the host's tissue during feeding by I. scapularis. When spirochetes appear in the blood cavity, the blood mixed with anti-complement may be sucked into the gut. The immune serum will fail to kill spirochetes when anti-complement exists in the blood. In the spirochete survival assay, spirochete survival reached over 95 % at an antibody titer of 2048x in immune rabbit serum with the presence of 20 ul of cobra venom and the saliva of I. scapularis ( Figure 3). If anti-complement exists in the gut homogenate of the tick, the function of complement will be lost and spirochetes can survive in the gut homogenate. It has been reported that cobra venom and the saliva of I. scapularis contain anti-complement factors. The cobra venom factor has been used extensively in the depletion of complement , Cochrane et al. 1970, Drake et al. 1974, Rudofsky et al. 1975, Wikel & Allen 1977. The addition of cobra venom to immune serum resulted in the depletion of complement; therefore spirochetes could survive. Anti-complement in the saliva of I. scapularis that prevents C3 hydrolysis even after C3 is fixed to a surface has been described. The complement pathway may be disrupted with consequent and inhibition of the associated C3 convertase . Jxodes scapularis also secretes copious amounts of saliva containing salivary apyrase, a kininase, an anaphylatoxin inactivating activity, and vasodilatory prostaglandins (Ribeiro et al. 1985a. Active immunosuppressive chemicals, including prostaglandin E2 (PGEz) and anaphylatoxin inactivating activity, block macrophage activation and neutrophil activity and may inhibit T-cell activation, all early precursors in the cascade of cellular events leading to antibody production and antigen processing. Spirochetes could be detected in nearly all of the nymphal I. scapularis derived from larvae bloodfed on an infected rabbit, and prevalence of infection was similar regardless of the tick's incubation temperature. With one exception, spirochetes were not detected in any nymphal D. variabilis or A. americanum; three spirochetes were observed in 1 of 57 A. americanum held at 1soc. Although extrinsic incubation temperature did affect the time required for ticks to molt, it appears to have no effect on vector competence in Lyme disease.

INTRODUCTION
Environmental temperature can have a significant effect on pathogen transmission by mosquito vectors ). However, little is known about the effect of environmental temperature on pathogen replication and transmission in ticks. The rate of Theileria parva parva transmission to cattle by Rhipicephalus appendiculatus was more rapid at higher rather than lower ambient temperatures  Dermacentor variabilis and the lone star tick, Amblyomma americanum also have been implicated as potential secondary vectors , Magnarelli et al. 1986). However, in studies directly comparing their ability to become infected, I. scapularis readily are infected but D. variabilis and A. americanum fail to maintain B. burgdorferi beyond the transtadial molt ). It has been suggested that I. scapularis possesses a potent a-complement protein in their saliva that disables complement-mediated spirochete killing in the tick's gut (Yeh et al., unpublished--Manuscript 1 ). Both D. variabilis and A. americanum lack such a mechanism.

45
In this study, we compared replication of Borre/ia burgdorferi growing in culture with tick midgut homogenates, and in infected ticks at different ambient temperatures.
Specifically, we investigated whether temperature affected the borreliacidal activity of serum antibody and complement in the presence of tick midgut extracts and in vivo.
Other innate midgut factors besides immune serum may affect the growth of spirochetes in a temperature-dependent manner.

In vivo experiments:
To assess the effect of the temperature on replication success of Lyme disease spirochetes (B. burgdorferi) in I. scapularis, D. variabi/is and A. americanum, larval ticks of all three species derived from a spirochete-free laboratory colony were allowed to feed simultaneously on the ears of a spirochete-infected New Zealand white rabbit (Charles River Laboratories, Wilmington MA). The rabbit had been infected three weeks previously by allowing more than 50 field-collected female I. scapularis to engorge to repletion. Field-collected ticks were from Prudence Island, Rhode Island, where the prevalence of infection in adult I. scapularis was 40%. All immature ticks in the laboratory colony were derived from adult ticks also collected on Prudence Island.
Engorged larvae were collected in cloth bags affixed to the rabbit's ears, and then stored in plastic vials at 1soc, 23oc, 280C, and 33oc and >95% relative humidity until molting into nymphs. Nymphs of all three species, derived in this manner, were examined for the presence of B. burgdorferi by dissecting midgut tissues onto a glass slide and treating them directly with fluorescein isothiocyanate-conjugated (FITC) antibodies to B.
burgdorferi in a direct fluorescent antibody assay .

In vitro experiments
To evaluate the effect of temperature on replication success of B. burgdorferi in gut homogenates of I. scapularis, D. variabilis and A. americanum, extracts were prepared from midguts dissected from pools of 5 engorged female ticks bloodfed on New Zealand white rabbits. Briefly, frozen midguts were macerated in enough PBS to make a 20% w/vol homogenate, centrifuged (14,000 x g for 12 s) to pelletize the gut tissue, and the supernatant passed through a 0.45 µm filter. Gut extracts were added to cultures of B.
burgdorferi B31 and diluted with BS.Kii media. The final tick gut homogenate dilution was 1: 200 which was determined to be the 50% killing endpoint of spirochetes in immune serum. The final concentration of spirochetes was 3. 0 ± 1. O x 1 o5 in each tube.
Experimental controls for each trial were B. burgdorferi growing in BSKII. Spirochete cell concentrations were determined by dark field and phase contrast microscopy after 8 days of incubation using a Petroff-Hausser counting chamber .
To compare spirochete replication in cultures containing mid gut extracts incubated at different temperatures we used a Sx4 factorial design with two-way analysis of variance  variabilis and A. americanum at 2soc and 33oc (Figure 4).
Ticks required less time to molt from engorged larvae to nymphs at higher temperatures. Enzymatic activity involved in the molting process is likely enhanced at higher temperatures, and other studies have shown temperature to be a major determinant affecting the molting process  burgdorferi. Further study would be required to assess whether such conditions might occur in nature.
Although three spirochetes were observed in one of 57 derived nymphal A.
americanum at 18 oc by DF A, there was no statistical difference, compared to the infection rate at 23oc, 2soc, and 33oc. Experiments have been performed using whitefooted mice ), Syrian hamsters , and New Zealand white rabbits , with similar results. In our experience, larval A.
americanum had difficulty feeding on white-footed mice and Syrian hamsters.  reported immature stages of A. americanum have never been observed on white-footed mice. White-footed mice and Syrian hamsters may not be the natural host for A. americanum, but we found that larval I. scapularis, D. variabilis and A.
americanum could feed on New Zealand white rabbits simultaneously in this experiment.
Although these three species of larval ticks were fed on spirochete infected rabbits, the infection rate was high only in derived nymphal I. scapularis. The spirochetes survived in derived nymphal /. scapu/aris and were killed in derived nymphal D. variabi/is and A.
americanum. We suggest that as soon as the rabbit was infected with spirochetes, the immune response began with antibody production, while spirochetes were simultaneously replicating in the tissue of the rabbit, but that some mechanism protects the spirochete from borreliacidal antibody, and thus the bacteria can evade host defenses (Duray & , Barbourr et al. 1986).
When three species of larval ticks were fed on the infected host, they should have had the same possibility of acquiring spirochetes from the host. But the spirochetes may have mixed with antibody in the blood, and in this manner, spirochetes may be killed in the blood cavity which is formed in the host's tissue during feeding, or spirochetes may have passed through the feeding channel and were killed in the gut of D. variabilis and A.
americanum by borreliacidal antibody. A factor may exist in the gut of/. scapularis preventing killing by antibody. An anti-complement factor in saliva of/. scapularis which interrupts the complement pathway was reported by . The anticomplement factor in saliva may mix with body fluid in the blood cavity during feeding on the host and this factor travels to the gut of/. scapularis allowing the survival of spirochetes.
From in vitro and in vivo results, we conclude that a factor in /. scapularis may inhibit killing of B. burgdorferi by spirochete immune serum of rabbits. Information from this experiment combined with the experiments in manuscript I, suggest that anti-complement factor may play a role in determining vector competence for Lyme disease among these three species of ticks.

INTRODUCTION
There is little information available on the immune mechanisms in chelicerate groups such as the Acari and the Araneidae. Most results concerning invertebrate immunity were obtained with insects and crustaceans. Because some tick species have hemocytes that are morphologically similar to insects , we examined the hemocytes of three tick genera to confirm any morphological similarities with insects. We were especially interested in the presence of granulocytes because they involved in prephenoloxidase activity of insects and crustaceans, ). The phenoloxidase system is roughly equivalent to the vertebrate complement system, directing and working in concert with the hum oral and cytological portions of the immune system to initiate phagocytosis, encapsulation and nodule formation (Appendix 11). As with any enzymatic system it is possible to measure the amount of activity initiated by particular non-self stimuli depending on the chemical composition and quantity of antigenic material . The immue response of the ticks may be similar to insects and crustaceans when stimulated by pathogen invasion.
A standard phenoloxidase activity assay was used to determine the presence of phenoloxidase in whole homogenates of non-engorged adult female I. scapularis, D.
variabilis, and A. americanum and compared to activity in the larvae of the greater wax moth (G. me/lone/la L.) as a positive control.
In addition, antimicrobial factors in the hemolymph have been demonstrated for a number of arthropods Dunphy 1986, Brey et al. 1993) and some ixodid ticks  infected by pathogens. An antimicrobial assay would reveal 57 antimicrobial factors in the form ofhomogenous substances and together with phenoloxidase activity, help gain a better understanding of immune factors in tick vector competence.

MATERIALS AND METHODS
Hemocyte preparation: Ticks were bled by amputating the distal portion of one or more legs with dissecting scissors . The wounded ticks were placed on a slide containing one drop of PBS, were covered with a cover slip and examined under bright field, phase contrast, and differential interference contrast (Nomarski method) microscopy.
Assay of Phenoloxidase Activity: Phenoloxidase activity was assayed according to the method described by .
Galleria mellonella larvae (Grubco Co., Hamilton, Ohio) were used as positive controls to standardize minimum significant incubation periods. Hemolymph from fifteen G. Phenoloxidase activity is expressed as the change in absorbance at 520 nm caused by 20 ul hemolymph/5 min (A520-20ul-1_5 min-1).

F 2 adult ticks were derived from parental adults collected by flagging on Prudence
Island , Rhode Island during the spring and fall of 1993 . Ticks were raised in a 23 °c incubator under relative 95% humidity. Thirty adult female ticks were weighed and macerated to form a 50 % w/v homogenate in 100 mM sodium phosphate, pH 7.2. The homogenate was then centrifuged at 14,000 rpm in an Eppendorf centrifuge 5415 (Brinkmann Instruments, Inc., Westbury, N .Y.) for 12 seconds to pellet the tissue. The supernatant was collected and incubated on ice for 30 minutes for the phenoloxidase activity assay as described above. The positive control was G. me/lone/la hemolymph from larvae of G. me/lone/la and the negative control contained no hemolymph or tick homogenate. Both control treatments were incubated on ice for 30 minutes. The experiments were repeated three times.

Preparation of tick homogenate:
lxodes scapularis, D. variabilis, and A. americanum adult ticks were blood fed in groups of 12 mating pairs on the ears of New Zealand white rabbits (Charles River Laboratories, Wilmington, MA). Cloth ear bags were taped to the rabbits' ears to prevent tick escape. An Elizabethan collar was placed on each rabbit to prevent grooming.
After the partial engorgement of ticks on the right ear of the rabbit, at the fourth day of attachment on I. scapularis, at the fifth day of attachment on D. variabilis, and at the sixth day of attachment on A. americanum, a 28 G 1/2 micro-fine IV needle (Becton Dickinson & Co., Franklin Lakes N.J.) with E. cloacae (cell concentration: 1 o9 /ml) was used to punch a hole inside the tick body behind the scutum. Ticks on the left ear of the rabbit were punched only with a needle as a control group. After treatment for 24 hours, the partially engorged female ticks were collected, weighed and macerated in PBS to make a 25% w/v homogenate, then centrifuged at 14,000 rpm 12 sec (Eppendorf, centrifuge 5415) to pelletize the gut tissue. The supernatant from each type of tick gut homogenate was passed through a 0.45 um filter for the antimicrobial activity assay.
Antimicrobial activity assay: Escherichia coli pop 3 (Institut Pasteur, Paris, France) were cultured in 5 ml nutrient broth. After two days, the E. coli pop 3 were mixed with streptomycin sulfate, 1 mg/ml in NaCl (0.9 % solution), and 15 ml of nutrient agar in a water bath at 45oc. This mixture was swirled to mix and poured into a petri dish. When the agar had solidified, three mm wells were punched using a sterile pasteur pipette and a laboratory suction line.
Fifteen microliters of 50 % tick homogenate was placed in each of six wells per species.
All detection plates were placed at 4oc for 30 minutes and then incubated at 3ooc for 24 hours, at which time inhibition zones were scored. A clear zone (> 1 mm from the edge of the well) inhibiting bacterial growth is considered indicative of antimicrobial activity ).

RESULTS AND DISCUSSION
Based on morphological studies, there are at least three types of hemocytes in ticks that are structurally similar to insects . There are prohemocytes (Figure 6), plasmatocytes (Figure 7), and granulocytes ( Figure 8) and granulocytes degranulate by exocytosis in vitro (Figure 9).
Similar findings were described in Ixodes ricinus Amosova 1983;. Although three species of ixodid ticks have hemocytes morphologically similar to insects, phenoloxidase and antimicrobial activity was not detected.
The F values of optical densities of hemolymph of the larvae of greater wax moths incubated at different times had overall significance, E (8, 18) = 12.17, at~< 0.05 by the Tukey multiple range. Optical density was significant after incubation ofhemolymph on ice for 20 minutes (m = 0.05), compared to 0.5 minutes (m = 0.04) ( Figure 5). This information supplied us with the proper incubation time for the phenoloxidase activity assay for ticks. Compared to the positive controls, there was no phenoloxidase activity detected in three species of ixodid ticks (Table 7). Similar results were reported by Schwagerl (1991). The absence of activity suggest either the absence ofphenoloxidase, the existence of factors inhibiting phenoloxidase or the inability to detect phenoloxidase activity by this method.
Antimicrobial activity was also not detected using our methods. Either no antimicrobial factors are present or the method is not effective. For example, blood fed ticks diluted whatever factors were present prior to challenge with E. cloacae. Podoronov ( 1991) used an unspecified number of unfed ticks which may have provided enough test material to detect antimicrobial activity.
Agglutination activities of lectins in the hemolymph of four tick species, Ixodes ricinus, Omithodoros tartakovskyi, Omithodoros papillipes, and Argas polonicus have been reported  and it is possible that these tick lectins play an important role in the processes of self and non-self recognition and defense reactions.
The absence of phenoloxidase coupled with the degranulation of cells by exocytosis and the presence of lectins suggests that the tick immune system may more closely resemble that of Limulus polyphemus than that of insects. Limulus has a system functionally analogous to the prophenoloxidase cascade of insects. Exposure of granulocytes of Limulidae to nonself molecules causes rapid release of their contents . Like the phenoloxidase system, Limulus immune activity depends on a divalent cation to stimulate proteolytic cleavage of a serine protease which in tum converts zymogen precursors sequestered in granulocytes to active forms. The proteolytic cascade in Limulus is best stimulated by bacterial lipopolysaccharide with granulocyte aggregation mediated by a lectin. Furthermore, the resulting protein monomers polymerize like those formed in the prophenoloxidase cascade to form the mesh of the clot . Limulus is more closely related to Chelicerates than to other arthropods .

Hemocytes of ticks are structurally similar to insects but functionally are analogous to
Limulus, therefore future research should determine the degree of homology between tick and Limulus immune systems.

ABSTRACT
Objectives: The duration of tick attachment is one factor associated with risk for human infection caused by several tick-borne pathogens. We measured tick engorgement indices (EI) at known time intervals after tick attachment and used these indices to determine the length of time that ticks were attached to tick-bite victims in selected Rhode Island and Pennsylvania communities where the agents of Lyme disease and human babesiosis occur.
Methods: The total body length and width as well as the length and width of the scutum was measured on nymphal and adult northern deer ticks (Ixodes scapularis = /. dammini) removed from laboratory animals at 0, 12, 24, 36, 48, 60, and 72 hours after their attachment. Three engorgement indices were calculated at each time interval. In addition, EI measurements were recorded for 504 ticks submitted to a commercial laboratory for pathogen detection testing between 1990-1992.
Results: No detectable change was observed in the average El's for either nymphal or adult ticks between 0 and 24 hours of attachment using any of the engorgement indices (P > 0.05). After 24 hours of tick attachment, all El's continuously increased; average indices for nymphs attached 36, 48 and 60 hours were significantly different from those attached~ 24 hours and from each other (P < 0.05). Similarly, average El's for adult ticks attached~ 36 hours were significantly different from those attached for 48 hours or more (P < 0.05). More than 60% of tick-bite victims removed adult ticks by 36 hours of attachment, but only 10% found and removed the smaller nymphal ticks within the first 24 hours of tick feeding.
Conclusions: The duration of tick attachment may serve as a useful predictor of risk for acquiring various tick-transmitted infections such as Lyme disease and babesiosis.
Regression equations developed herein correlate tick engorgement indices with duration 68 of feeding. A table containing specific EI prediction intervals calculated for both nymphs and adults will allow the practitioner or clinical laboratory to use easily-measured tick engorgement indices to predict transmission risk by determining the duration of feeding by individual ticks.

INTRODUCTION
The risk of infection with several tick-borne pathogens, including the agents of Lyme disease, human babesiosis and Rocky Mountain spotted fever, depends directly on the duration of vector tick attachment. For example, before transmission of Borrelia burgdorjeri can occur, nymphal and adult lxodes scapu/aris (=I. dammini) usually must be attached to a host for more than 24 and 36 hours, respectively (Piesman et al.1987, Piesman et al.1991. Sporozoites of Babesia microti generally require a minimum of36-48 hours to complete maturation in salivary glands of nymphal I. scapularis, and babesial infections in hamsters were transmitted most efficiently after 54 hours of tick attachment . Rickettsia rickettsii become "reactivated" during the tick feeding process and infectious forms are transmitted by Dermacentor andersoni only after a short feeding period (usually 8 -10 hrs) . These examples of transmission or reactivation delays between tick attachment and transmission may partially explain why the risk of acquiring Lyme disease, and perhaps other tick-borne infections, by people with recognized tick bites is less than might be expected based on the relatively high infection prevalence in ticks . Ticks can be removed prior to pathogen transmission, and many probably are.
During tick feeding, the alloscutal length and width of ixodid ticks increases markedly while the dimensions of their hard scutal plate remain unchanged. Tick engorgement indices have been created from ratios that compare scutal and full-body dimensions, and have been used in previous studies to describe the time course of tick feeding in relation to parasite development and transmission to animals (Obenchain et al.1980). Thus, it may be possible to assess risk for human infection with various 70 tick-transmitted agents, especially those with lengthy delays between tick attachment and transmission, by determining a tick's engorgement index. Accordingly, we evaluated various engorgement indices of/. scapularis and compared them with the length of time a tick was attached to a host. In particular, we developed regression equations explaining the relationship between the duration of attachment and various engorgement indices for both nymphal and adult ticks fed on hamsters and rabbits, respectively. We then used these equations to predict the duration of attachment of ticks removed by tick-bite victims in two communities in Rhode Island and Pennsylvania.

MATERIALS AND METHODS
Host-seeking, adult ticks of I. scapularis used for experimental infestations were collected from clothing after walking through vegetation ) at a heavily tick-infested site located in South Kingstown (Washington Co.), RI, during the spring of 1993. All ticks were separated by sex into different vials and stored at 4oc under relative humidity > 95% until used. Nymphal I. scapularis for experimental infestations were derived from larval ticks blood-fed on laboratory-raised white-footed mice (Peromyscus /eucopus) . The larval ticks were from field-derived adult females collected on Prudence Island, RI during the spring and fall of 1993 . Engorged larvae and subsequently-derived nymphs were held in vials at 23oc under relative humidity> 95%.
To obtain partially-engorged adult I. scapularis, ticks were placed onto the ears of a New Zealand white rabbit (Charles River Laboratories, Wilmington, MA) in two groups of 50 mating pairs. Ticks were contained on the rabbit's ears using cloth bags affixed at the ear base with tape. An Elizabethan collar was placed on the rabbit to prevent excessive grooming. After allowing 2-3 hours for tick attachment, the ear bags were opened and all non-attached ticks were removed. At time intervals of 12, 24, 36, 48, 60, and 72 hours after ticks were attached, ten ticks were removed from the rabbit's ear by traction using fine-pointed forceps, taking care not to damage the ticks' mouth parts.
Partially-engorged nymphal I. scapularis were obtained in a similar way, except that they were attached to three Syrian golden hamsters (Charles River Laboratories, Wilmington, MA) held in small, wire-mesh restraining cages and wrapped in paper. Totals of 100 nymphs were placed on hamsters; after 2-3 hours, all non-attached ticks were removed, and animals were placed singly into larger cages with wire-mesh bottoms held over pans of water. At time intervals of 12, 24, 36, 48, and 60 hours following their attachment, 30 nymphs were removed (10 nymphs per time interval from each hamster).
All ticks removed from hosts were anesthetized by transferring them to a 2 ml vial containing cotton and a drop ofhalothane (Halocarbon Laboratories, North Augusta, S. Step-wise regression analysis with true stepping (a. to enter= .15, a. to remove= .20) was used to evaluate relationships between the known duration of tick attachment and each engorgement index. Seventy-five percent prediction intervals (for predicting outcomes from individual tick El's) were calculated .

73
To predict the duration of attachment of ticks removed by tick-bite victims, the total body length and scutum length (Index 1) was obtained from 128 adult female ( (Paul et al.1986) while the infection prevalence in ticks, at least in the northeastern U.S., generally ranges from 20-35% for nymphs and 50-75% for adults . Duration oftick attachment is commonly cited as an important factor determining the probability of infection, although studies have yet to be conducted that relate the length of time that a tick feeds to the incidence of Lyme disease, or any other tick-borne infection.
Lack of a convenient, reliable, and objective measure for determining the duration of tick attachment is often cited by clinicians as a principal limitation to assessing risk of infection following a tick bite (Magid et al.1992). The notion of using engorgement indices (El's) to measure duration of tick feeding is not new ). However, the missing parameter, for infections transmitted by/.
scapularis (and for most other vector ticks), has been in comparing length of tick attachment with a particular index of engorgement. In this experiment, we made four simple measurements on ticks ( Figure 12) assisted only by a dissecting microscope and ocular micrometer, developed regression equations for three separate El's at relevant time intervals for both nymphal and adult ticks, and then generated a table presenting the 75% prediction intervals (Table 8). For the practitioner, only one EI need be calculated.
There was no real advantage of one EI over any other in their ability to discern differences in length of tick attachment. We associated all three measures with their respective attachment times merely to provide flexibility in making at least one accurate measurement in the event that a particular patient's tick was damaged. Commonly, after someone removes a tick, the proximal part of the scutal plate will be missing making accurate measurement of Indices 1 and 2 difficult. In practice, we found Index 1 (total body length/scutal length) to be the simplest and most accurate EI to measure as it required no additional manipulation of the tick once it was on the microscope stage.  . Similarly, adult I. scapularis feeding for~ 36 hours failed to infect rabbits, whereas two of three rabbits became infected when ticks were attached for 48 hours, and five of five rabbits were infected when attachment was > 120 hours (Piesman et al.1991 ). Variability in tick feeding rates in relation to spirochete transmission among different tick hosts has not been determined making it impossible to know if findings from these animal studies are applicable to humans. Similarly designed studies utilizing a wider array of animal hosts may provide some evidence toward the degree of interspecies variation that exists in tick feeding rates.
We propose that the true likelihood of becoming infected can be predicted by (1)   . It may be that human exposure to nymphs is greater making nymph bites more common, as this life stage can be more abundant than adult ticks by an order of magnitude . Moreover, peak abundance of nymphs during the summer coincides with higher levels of human activity outdoors than during the late fall and early spring when adult I. scapularis are most abundant (Piesman et al.1987). Since our findings suggest that the risk ofinfection following any deer tick bite is similar, it is likely that people are bitten more frequently by nymphs than by adults.
Prophylactic antibiotic treatment of tick bites continues to be controversial, particularly in areas where Lyme disease occurs. Proponents of prophylactic therapy argue that even though B. burgdorferi and B. microti are mainly transmitted in the latter stages of tick feeding, earlier transmission of a small number of organisms has not been ruled out ). Moreover, decision analysis has suggested that empirical treatment ("treat all") be recommended when the probability of infection is above 0. 01 . In contrast, opponents of prophylactic treatment cite the apparent low probability of infection even after tick bite and risk of adverse drug reactions in defending their position. In addition, inappropriate tetracycline treatment of a tick-bite leading to Rocky Mountain spotted fever could prove problematic as such treatment may postpone the onset of symptoms perhaps leading to a delay in diagnosis and proper treatment ofthis serious tick-transmitted infection . In this study, we developed a simple and objective means for determining the duration of tick attachment. Since attachment of nymphs for less than 36 hrs and adult ticks for less than 81 48 hrs pose relatively little risk, at least for transmission of Lyme disease spirochetes, prophylactic therapy reasonably could be reserved for those patients presenting with a deer tick whose EI suggests a longer period of attachment. Additional testing, to detect infection in the tick, would provide an even better estimate of the infection probability, and thus, is also recommended.