Vol. 29, No. 12

JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 1991, p. 2798-2804

0095-1137/91/122798-07$02.00/0 Copyright © 1991, American Society for Microbiology

Rickettsiae and Borrelia burgdorferi in Ixodid Ticks LOUIS A. MAGNARELLI,1* THEODORE G. ANDREADIS,1 KIRBY C. STAFFORD III,' AND CYNTHIA J. HOLLAND2t Department of Entomology, Connecticut Agricultural Experiment Station, P.O. Box 1106, New Haven, Connecticut 06504,1 and Hemotropic Disease Section, Department of Veterinary Pathobiology,

University of Illinois, Urbana, Illinois 618012 Received 13 June 1991/Accepted 26 September 1991

Nymphs and adults of hard-bodied ticks were collected in Connecticut and tested by direct and indirect immunofluorescence staining methods for rickettsiae and Borrelia burgdorferi. Of the 609 Ixodes dammini ticks examined, 59 (9.7%) harbored rickettsialike microorganisms in hemocytes (blood cells). These bacteria reacted with fluorescein-conjugated antiserum to Ehrlichia canis, the etiologic agent of canine ehrlichiosis. Prevalence of infection ranged from 6.8 to 12.7% for males and females, respectively. Although the specific identities of the hemocytic rickettsialike organisms are unknown, they share antigens with ehrlichiae. Electron microscopy revealed rickettsiae in ovarian tissues of I. dammini that also had infected hemocytes. Rickettsialike organisms were also observed in the hemocytes of 5 (6.9%) of 73 Dermacentor variabiis ticks. In analyses for B. burgdorferi, 146 (23.7%) of 617 I. dammini ticks harbored these spirochetes in midguts. Hemocytic rickettsialike microorganisms coexisted with B. burgdorferi in 36 (6.7%) of the 537 nymphs and adults of I. dammini examined. dammini, with its broad host range, has the potential to acquire multiple microorganisms. J.

Public awareness of tick-borne pathogens and the diseases they cause is rising. The association between American dog ticks, Dermacentor variabilis, and Rocky Mountain spotted fever is well recognized (9, 27). During the 1970s, increased populations of Ixodes dammini and the occurrence of babesiosis were reported (5, 17, 39). The subsequent discovery of Borrelia burgdorferi, the causative agent of Lyme borreliosis (10, 43), in I. dammini and the apparent widespread geographic distribution of this disease on different continents (40, 41) greatly elevated scientific interest in tick-related zoonotic diseases, particularly in North America and Europe. More recently, there is evidence of yet another tickassociated human disease-ehrlichiosis (26). Human ehrlichiosis has been reported in at least 16 states (12, 26, 36). Although the etiologic agent has not been isolated from ticks, it is believed to be Ehrlichia canis or a closely related organism (14, 44). The brown dog tick, Rhipicephalus sanguineus, has been shown to be a vector of E. canis for dogs (20), but this tick rarely bites humans. Thus, the vector(s) of human ehrlichiosis is unknown. During an epidemiological investigation of human ehrlichiosis among army reservists (31), some patients recalled being bitten by "tiny ticks." The tick species was not identified, but the training exercises were conducted in areas of New Jersey where I. dammini, Amblyomma americanum (lone star ticks), and Lyme borreliosis occur. Moreover, reports of human and canine ehrlichiosis in southern states coincide with the distribution of Ixodes scapularis, a tick that is closely related to I. dammini and Ixodes ricinus. As a group, these ticks have a wide host range that includes mammals, birds, and even reptiles (1). The purpose of this study was to determine whether I. dammini nymphs and adults and other ixodid (hard-bodied) ticks harbor rickettsiae and whether these microorganisms coexist with B. burgdorferi.

* Corresponding author. t Present address: Protatek Reference Laboratory, Chandler, AZ

85225.

2798

MATERIALS AND METHODS

Tick collection. Adults of I. dammini and D. variabilis collected mainly by dragging flannel cloth over vegetation in or near woodlands during the period from 1989 to 1991. Additional specimens were obtained from white-tailed deer (Odocoileus virginianus) killed during fall hunting seasons, woodchucks (Marmota monax) captured in Tomahawk traps, and people who had detached ticks from themselves. Nymphs of I. dammini and D. variabilis were also removed from white-footed mice (Peromyscus leucopus) and chipmunks (Tamias striatus) caught in Sherman box traps (2). Other nymphs of I. dammini were removed from passerine birds caught in Japanese mist nets. The woodchucks and other rodents were released unharmed into their natural habitats following tick removal and recovery from anesthesia. Birds were examined without the use of anesthetics and were likewise released unharmed after examination. I. dammini and D. variabilis were collected in 25 towns in Connecticut, including East Haddam, Chester, Lyme, Old Lyme, and Waterford, communities in southeastern and south-central parts of the state where Lyme borreliosis is highly endemic (32). Preparation of tick cells and tissues. Hemolymph (blood) and midgut tissues were obtained from living ticks and processed for serologic testing. Hemolymph was obtained from ticks by amputating a leg (8) and then allowed to air dry on glass microscope slides. Midgut tissues were removed from ticks, and impression smears were prepared as reported earlier (10). All preparations were then air dried at 37°C for at least 6 h and fixed in cold acetone for 10 min. If cell and tissue preparations could not be stained by immunofluorescence methods within 24 h, the slides were stored at -60°C until analyzed. FA staining. Direct fluorescent antibody (FA) staining methods were primarily used to detect ehrlichiae in tick hemocytes. At the University of Illinois, dogs with no cross-reactive antibodies prior to immunization were inoculated with E. canis. When homologous antibody titers to E. canis reached 1:5,120 or greater, as determined by indirect were

VOL. 29,

1991

FA staining, serum was obtained. Details on inoculations and antiserum production, including purification of the immunoglobulin G (IgG) antibody fraction by sodium sulfate precipitation procedures, have been described previously (29). These antibodies were conjugated with fluorescein isothiocyanate (FITC). This stock FA reagent was diluted 1:30 in phosphate-buffered saline (PBS) (pH = 7.2), applied to tick hemocyte preparations on glass microscope slides, incubated at 37°C for 30 min, washed twice in PBS solution (pH = 7.2), and then washed once in distilled water. Slides were then mounted with buffered glycerol and coverslips and were examined by fluorescence microscopy. In our tests of controls, there was no reactivity of the direct FA reagent with uninfected canine monocytes. Therefore, the anti-E. canis serum was not an antimonocytic serum and adsorption with uninfected canine monocytes was unnecessary. Indirect FA staining of tick hemocytes was used to confirm direct FA staining results. However, there was insufficient hemolymph from any given tick to routinely perform both tests in parallel experiments. Therefore, indirect FA staining was usually applied to hemolymph preparations of ticks collected from the same populations included in studies for which direct FA staining methods were used. For indirect FA staining, the same unlabeled canine antiserum, having a homologous antibody titer of 1:10,240 to E. canis, was diluted in PBS to 1:64 before application to tick cell preparations. Following a 30-min incubation period and washes with PBS and distilled water, polyvalent FITClabeled goat anti-dog IgG antibodies (specific for heavy and light chains; Cooper Biomedical, Malvern, Pa.) were diluted in PBS solution to 1:40, applied to the slide, incubated, and washed as described above. Seventy tick hemocyte preparations were tested with convalescent-human serum by indirect FA staining. The serum sample was from a person who was clinically diagnosed as having ehrlichiosis. The titer of the antibody to E. canis was 1:10,240. This sample, also obtained from the University of Illinois, was used at a working dilution (in PBS) of 1:80; polyvalent FITC-conjugated goat anti-human immunoglobulin (GIBCO Laboratories, Grand Island, N.Y.), diluted to 1:40, were added as a second antibody. Duplicate preparations of hemolymph were also tested by indirect FA staining with a monoclonal antibody to Ehrlichia species (134-5B1). The production and use of this genusspecific antiserum, which reacted with a 57- to 58-kDa antigen of Ehrlichia species, has been reported (13). Detection of B. burgdorferi in tick midgut tissues was likewise accomplished by indirect FA staining. Murine monoclonal antibodies (H5332), directed to outer surface protein A of B. burgdorferi, and a 1:60 dilution of FITCconjugated goat anti-mouse IgG (specific for heavy and light chains) antibodies (Organon Teknika Corp., Durham, N.C.) were used as previously reported (22). Specificity tests. Indirect FA staining methods were applied to assess serologic cross-reactivity between the Ehrlichia spp. and related microorganisms. The canine and human sera, with antibodies to E. canis, were tested against E. canis (infected canine monocytes were used as the antigen) in homologous tests and against the following purified spotted fever group (SFG) rickettsiae: Rickettsia rickettsii (R strain), Rickettsia montana (M15-6), and Rickettsia rhipicephali (3-7 9 -6). Formalin-treated antigens of these SFG rickettsiae were prepared at the Rocky Mountain Laboratories (Hamilton, Mont.). Details of the preparation of standardized concentrations of stock rickettsial antigens, stabilization procedures, and use of antigens in serologic

RICKETTSIAE AND B. BURGDORFERI IN TICKS

2799

TABLE 1. Examination of nymphs and adults of I. dammmini tested for rickettsialike microorganisms and B. burgdorferi Rickettsialike organisms" Year and tick stage or sex

1989 Nymph Male Female 1990 Nymph Male Female

Total

B. burgdorferib

Total no. of ticks tested

% Positive

Total no. of ticks tested

% Positive

194 158 98

12.4 5.7 7.1

195 85

173

13.9 32.9 20.0

21 78 60

0 9.0 21.7

21 80 63

4.8 37.5 36.5

609

9.7

617

23.0

aExamination of tick hemocytes by direct or indirect FA staining with antiserum to E. canis. Note that positive reactions indicate shared antigens with ehrlichiae. b Examination for B. burgdorferi found in tick midgut tissues.

testing have been published previously (23, 24, 30, 33). The direct FA reagent for E. canis was tested against Wolbachia persica (a rickettsial symbiote), Ehrlichia risticii (the causative agent of Potomac horse fever), the SFG antigens, and E. canis. Monoclonal antibodies to W. persica (134-5C9) and to Ehrlichia species (134-5B1) were included to verify antigen reactivity in homologous tests. Sera from six patients who had clinical manifestations of Rocky Mountain spotted fever were tested by indirect FA staining methods against E. canis, the SFG rickettsiae, W. persica, and hemocytes of I. dammini. Titers of antibody to R. rickettsii ranged from 1:4,096 to 1:16,284 by indirect microimmunofluorescence staining (33). Sera were obtained from persons in Connecticut from 1980 to 1983 and had been stored at -60°C at the Connecticut Agricultural Experiment Station. Electron microscopy. Tissues from unfed female I. dammini that had rickettsialike microorganisms in hemocytes were dissected and were fixed for 2 h at room temperature (22 to 24°C) in a solution consisting of 2.5% glutaraldehyde, 2% paraformaldehyde, 0.1% CaCl2, and 1% sucrose, buffered with 0.1 M sodium cacodylate (pH = 7.2). Specimens were postfixed for 2 h at room temperature in 1% OSO4 in the same buffer, stained en bloc with 2% uranyl acetate, rapidly dehydrated through an ethanol-and-acetone series, and embedded in an LX 112-araldite mixture. Ultrathin sections were stained with 5% methanolic uranyl acetate and Reynold's lead citrate and examined with a Zeiss EM-10 electron microscope at 80 kV. RESULTS

Ticks from Connecticut contained rickettsialike microorganisms in hemocytes that reacted with conjugated and unlabeled antisera to E. canis. From 1989 to 1991, 609 I. dammini ticks were examined for hemocytic microorganisms, and of these, 537 specimens were also tested for B. burgdorferi. Hemocytes from 59 (9.7%) nymphs, males, or females contained morphologically similar microorganisms that reacted by direct or indirect FA staining to E. canis antisera (Table 1). The mean prevalence of infection ranged from 6.8 to 12.7% for males (n = 236) and females (n = 158), respectively. Rickettsialike organisms, having shared antigens with ehrlichiae, were observed in the hemocytes of ticks collected in the following towns of Connecticut: Burlington, Chester, Durham, Eastford, East Haddam, Lyme,

2800

J. CLIN. MICROBIOL.

MAGNARELLI ET AL.

TABLE 2. Examination of nymphs and adults of I. dammini for rickettsialike microorganisms in hemocytes by direct and indirect FA staining methods using antisera to E. canis Indirect FA (canine sera)

Direct FA Sampling method or source

Total no. of ticks tested

No. (%) positive'

Total no. of ticks tested

50 (20.8)

158 7

Rodents' Deer or humans

240 0 0 129

Total

369

57 (15.5)

Flagging

Birds'

7 (5.4)

5 0 170

Indirect FA (human sera)

No. (%) positive

Total no. of ticks tested

No. (%) positive

6 (3.8)

70

4 (5.7)

0 °

0 O

0 6 (3.5)

70

4 (5.7)

" Positive reactions indicate shared antigens with ehrlichiae. b Wood thrush (Hylocichla mustelina), ovenbird (Seiurus aurocapillus), and veery (Catharus fuscescens). Chipmunks and white-footed mice. d Ticks found attached to hosts.

Old Lyme, and Wilton. Burlington and Eastford are in northern Connecticut, while Wilton is in the southwestern part of the state. The remaining towns are in south-central or southeastern Connecticut. In analyses for B. burgdorferi, 146 (23.7%) of 617 I. dammini ticks harbored these spirochetes in their midguts. The prevalence of spirochetal infection was more than twofold greater for adults (28.9%) than that of nymphs (13%). Despite relatively high infection rates for B. burgdorferi, the number of ticks simultaneously carrying this spirochete and hemocytic rickettsialike microorganisms was low. Of the 537 nymphs and adults tested for both bacteria, 36 (6.7%) specimens had coexisting organisms. Examination of 43 females and 30 males of D. variabilis, collected in East Haddam and Lyme, Connecticut, revealed hemocytic rickettsialike organisms in 1 female tick and 4 male ticks. No B. burgdorferi were detected. Rickettsialike microorganisms were observed in the cytoplasm of tick hemocytes by direct or indirect FA staining procedures. Although the percentage of positivity was nearly threefold greater by direct FA staining methods (Table 2), intense fluorescence of microorganisms was noted, regardless of the assay method used (Fig. 1). Parallel tests of hemolymph preparations were conducted on 9 female and 12 male I. dammini ticks. The hemocytes of three females and four males were positive by indirect FA staining methods using antiserum to E. canis or to a monoclonal antibody to Ehrlichia species (134-5B1). Hemocytes from another male tick reacted by indirect FA staining with this monoclonal antibody and by direct FA staining. Hemocytes

FIG. 1. Rickettsialike organisms in the cytoplasm of a hemocyte from a female I. dammini following direct FA staining with antibodies to E. canis. Bar 5 ,um. =

from the remaining 13 ticks were negative by these methods. Not all hemocytes of positive ticks were infected. Morulae, a growth stage described for ehrlichial infection (26, 36), were not observed. The number of positive I. dammini included questing ticks (i.e., ticks searching for vertebrate hosts) as well as those that had been feeding on mammals. However, positivity for rickettsialike microorganisms was nearly fourfold greater for the unfed, questing ticks. In specificity studies, canine antiserum to E. canis did not react with W. persica or with the SFG rickettsiae by indirect FA staining methods. Similar results were recorded when human sera from spotted fever patients were tested against E. canis antigen and when the direct FA reagent for E. canis was screened against W. persica and the SFG rickettsiae. However, homologous reactions for E. canis, W. persica, and R. rickettsia were positive at titers of 1:5,120 or greater. Cross-reactivity was noted when conjugated or unlabeled antiserum to E. canis was tested against E. risticii antigen (titer = 1:1,280). Parallel tests were conducted to determine whether human serum containing antibodies to R. rickettsii (titer = 1:4,096) would react with the hemocytic rickettsialike microorganisms in female I. dammini. Of the 20 hemolymph preparations positive by direct FA staining with antiserum to E. canis, none of the duplicate preparations from these ticks reacted when sera from spotted fever patients were tested by indirect FA staining. Numerous rickettsiae were observed by electron microscopy in ovarian tissue of adult female I. dammini (Fig. 2). The reactivity of these organisms to conjugated E. canis antiserum is unconfirmed. However, hemocytes from these ticks contained rickettsialike organisms that reacted positively with FITC-labeled antibodies to E. canis. Ovarian tissues were identified by the presence of intercellular cytoplasmic bridges and extensive microtubules within the host cell cytoplasm (Fig. 2B), as described previously by Brinton (6) and Raikhel (35). No rickettsiae were found by examining serial sections of other internal organs, including the Malpighian tubules, midgut epithelium, nerves, salivary glands, and tracheae. By electron microscopy, the individual rickettsiae found in tick ovaries were uniformly rod-shaped and measured 0.9 to 1.5 ,um long and 0.4 to 0.5 ,um wide (Fig. 3A). The rickettsiae were surrounded by a prominent electron-lucent halo zone and were delineated by two distinct unit membranes: an outer cell wall that was characteristically rippled and an inner cytoplasmic membrane, separated by an electron-lucent periplasmic space (Fig. 3B). The various shapes

RICKETTSIAE AND B. BURGDORFERI IN TICKS

VOL. 29, 1991

2801

4 -I r> .f.

`i

4:

,-II

2R.

* .:.

s:w

N.

.

.. * *

\

/o_...se 8

.....R * n~~ f. id:

'N

oJ,

f

t'

-s .Rl

t"'

'It

w

N

à* Il

MT ;

A.,

A

f

.....;

',,. "sw

r~~~~~~~~~~~~~~~~.

Ir#

,r.

. '1t

'`

'* ; ......la. ,

%

B

FIG. 2. (A) Oocyte from an unfed female I. dammini showing numerous rickettsiae (R) within the cytoplasm. N, host cell nucleus. Bar 2 ,um. (B) Longitudinal section of an intercellular bridge (IB) and associated microtubules (MT) within rickettsia-infected ovarian tissue. R, rickettsia. Bar = 0.5 ,um. =

of the rickettsiae observed in ovarian cells were due to the orientation of microorganisms at the time they were sectioned for electron microscopy. True pleomorphic forms were not observed. Also, electron-lucent halo zones sometimes overlapped and appeared deceptively like vacuoles but were clearly without membranes. The cytoplasm was amorphous and composed of a dense granular matrix. The cytoplasm was noticeably void of any fibrillar reticulation, nuclear inclusion bodies, or well-defined vacuoles but often contained a single small (65- to 120-nm diameter) crystalline inclusion body that possessed a distinct lattice structure (Fig. 3C). Rickettsiae were confined to the host cell cytoplasm, and reproduction appeared to occur by binary fission (Fig. 3D). A variable number of lysosomelike organelles that contained small vessicles surrounded by concentric membranes were also observed (Fig. 3E). These were randomly dispersed throughout the cytoplasm and appeared to engulf individual rickettsiae. DISCUSSION Rickettsialike microorganisms are present in the hemocytes of I. dammini and D. variabilis. However, it is unclear whether or not these organisms, which apparently share antigens with ehrlichiae, are pathogenic to humans or domesticated animals. They could be avirulent. Initial attempts to isolate these organisms from I. dammini in cell culture have been unsuccessful thus far. Ticks are suspected vectors of ehrlichiae, but conclusive experimental evidence linking the pathogen, vector, and vertebrate host for Potomac horse fever and human ehrlichiosis in the United States is lacking.

Isolation of etiologic agents and subsequent studies with ticks and laboratory animals are required to demonstrate transmission and pathogenicity. Such experiments have been performed for E. canis, R. sanguineus, and dogs (20) and can be performed for other ehrlichial agents as well. However, brown dog ticks, unlike I. dammini, are relatively host specific (28) and prefer dogs over other mammals. If the etiologic agent of human ehrlichiosis is E. canis, then it is likely that additional tick vectors are involved. Canine ehrlichiosis occurs in Connecticut (25) and several other states where there are multiple tick species. Aside from dogs and other wild canids, such as coyotes and red and gray foxes, that are reservoirs for E. canis, it is unknown if other wildlife (e.g., rodents, lagomorphs, or deer) are likely reservoirs for this organism or other ehrlichiae. In Europe, I. ricinus is a vector of Ehrlichia phagocytophila, the etiologic agent of tick-borne fever in sheep and cattle (37). Deer are believed to be natural hosts for this granulocytic ehrlichia. Thus, alternative reservoir hosts for E. canis and E. risticii should be considered. The number of I. dammini harboring hemocytic rickettsialike microorganisms was larger than the number of D. variabilis adults infected with SFG rickettsiae (24) but was smaller than the number infected with B. burgdorferi (21). Although ticks do not efficiently pass ehrlichiae or B. burgdorferi transovarially (22, 26), transstadial transmission of these bacteria does occur (3, 26). Therefore, as in Lyme borreliosis, the competence of vertebrate reservoirs for ehrlichiosis is probably an important epidemiological factor and warrants further study. In tests of the specificity of conjugated or unlabeled antisera to E. canis, there was no cross-reactivity with W.

2802

J. CLIN. MICROBIOL.

MAGNARELLI ET AL.

4.

?,r

bmil

.

IL,

9

4,

e.w

r

\p.

I..,



A

42W4+

r

f

E

Rickettsiae and Borrelia burgdorferi in ixodid ticks.

Nymphs and adults of hard-bodied ticks were collected in Connecticut and tested by direct and indirect immunofluorescence staining methods for rickett...
3MB Sizes 0 Downloads 0 Views